] _ [(31.9 - 12.7)(5.3/2)J + (12.7 x,5.3)· + [<26 - 5.3><12.7/2>J - 50.9 + 67.3 + 131.4 - 250 psi-ms i
BL
< ira' use bi-1inear pressure-time as design blast load.
See Figure 2A-16 g. h.
, For Ps o - 12.6 psi, Lw/W1/3 = 2.10 ft/1b 1/3, Fig 2-15 Lw - (Lw/W1/ 3 ) (W1/3) - 2.10 x 18.1712 - 38.16, say 38.2 ft.
B.Interior Blast Load a.'.
From previous steps: - 20 ft.
Lw - 38.2 ft, L,- 30 ft, H - 15 ft, W
2A-56
II
FICTICIOUS INFINITE SURFACE IMPULSE
THEORETIC ACTUAL' SlJRFACE IMPULSE
,,
,. , . , I l'
!
.,, , , i ,
'-
,I
I.
, ,
, ;
'I
! t
!
!
I
!'
,
.
, ',
I
I
I I ;' I ,
I
III
a..
~
20
(J) (J)
UJ lr
a..
I
:
, I
i
'
I
:
I
I
I!
:.
"
I
;
I
I
UJ lr
::> PSO
I
I .'
I
I i :
i ' -i', I I': . I ..
:
'
I
t-
Co C\>
i
I
'
j
:
1
'
I I
I
I
I
,
PI I I
10
. I
,
I
: :
I 1 I',
I
,
,
I
o
~'C = 5'~
' j 20 . t r f = 19 '6 . 10
TIME (rns)
FIGURE
2A - 16
2A-57
tof
o
TM 5-1300/NAVFAC P-397/AFR 88-22 A ._ Door opening area + window opening area o _ (7 x 3) + (3 x 5:5) - 37.5 ft 2 Aw - H x W- 15 x 20 - 300 ft 2 Ao/Aw - 37.5/300 - 0.125
.Lw/L - 38.2/30 ~ 1.27 Lw/H - 38.2/15 - 2.54 W/H - 20/15 - 1.33 L/H - 30/15 b.
~
2.00.
For W/H - 1.33, P s o- 12.6 psi, Ao/Aw - 0.125, and Lw/H 2.54 summarize factored maximum average pressure, Pmax x Lw/H, for W/H and Lw/H equal to .75, 1.5, 3 and 6. Figure No ..
W/H
2-203
0.750.75 ·1.50 3.0 6.0
2-204
1. 50'· 0.75 1,50 3.0 6.0
15.0 ·38.5 88.0 190.0 4.70 17.0 55.0 ·131. 0
2-205·
.3.0
·0.75 1. 50 3.0 6.0
2.0 7.0 27.0 85.0
2-206
6,0
0.75
LO
1.50
3.20 10.0 35.0
3.0 6.0
2A-58
TH 5-1300/NAVFAC P-397/AFR 88-22
Plot Figure 2A-17 Cal, and interpolate to determine Pmax x lv/H at lv/H - 2.54 for W/H - .75, 1.5, 3 and 6. . W/H
.75
1.5'
3
74
44
19
Pmax x lv/H
6
7.5
Plot Figure 2A-17 Cb) from above values, and interpo1at~ to Pmax x lv/H - 48 for W/H - 1.33.
deter~ine
Determine Pmax - CP max x lv/H)/Clv/H) - 48/2.54 c.
18.9 psi.
For Ps o -:12.6 psi and Ao/~ - 1/8, .determine T l for W/H.75, 1.5"and 3. from Figures 2-207 and 2-208 .75 1:25
1.5 1.70
3 2 .26
Plot Figure 2A-18 with above values, and determine T1 - 1.60 for W/H - 1.33. d.
For Ps o- 12.6 psi, determine T2 - T1 for W/H and lv/H - .75, 1.5, and 3, from Figures 2-209 and 2-210 .75 .75
1.5
1.5
3
2.07
2.07
.75
3
1.5
3
.75
1.5
3
6.20
7.0
6.50
10.8
15.0
,
2.07
below.
4.70
Plot Figure 2A-19 with above values, and interpolate to determine T2 - T1 at Lw/H - 2.54 for W/H - .75, ·1.5 and 3, as summarized .75 2.07
1.5 6.75
3
14.0
Plot Figure 2A-19 with above values, and determine T2 - Tl ms for W/H - 1.33. e.
5.80
For Pso- 12.6 psi, determine T3 - T1 for W/H - .75, .1. 5, and 3. and Lw/H - .75, 1.5, and 3, from Figures 2-211 and 2-212, as summarized below.
2A-59
100 W/H=O.75
NOTE:' SEE EXAMPLE 2A-II. STEP 3,PART B. ITEM b
·W/H=I.5
80
74
FOR l.w/H =2.54 GRAPHIC INTERPOLATION YIELDS Pmox Lw/H = f,( W/H), AS SUMMARIZED BELOW.
WIH= qIQ (\J
44
:x:
.....
40
~
...J >C
,.20 19
/H= 6
a
s
7.
o
0
4
I
W/H
P,.oa !.wilt
0.75 . 1.5
74 44
3 6
19
7.5
5
Lw/H (0 )
100
"
FOR Lw tH· = 2.54, AND W/H = 1.33, GRAPHIC INTERPOL ATiONYI ELDS Pmo x Lw/H =48 ....
48
:x:
40
.....
J >C
><
20
a
L w/H=2'5
cf o
'"
: : o
2
3
5
4
W/H (b)
FIGURE 2A-17 2A·60
6
, ,,
, , , , ,, I ,, I , ,,, , ,, , , , I ,I , ,, , , , , ,, , ,, ,, ,, , , ,, I ,I , , ,, ,, I : , I ,, ,, ,,, , , , ,I , ,, ,,, , , , , , :, I
, , I , , I ,I , , ,, , , ,, , , ,,, ,I I , , , , , , I , I , I ,,, I I , , , , I , , ,, I ,,, ,, ,, 3 ,, , , , , , .r , , ,, : , ,, , , , ,! , ,, ,I ,I , , , ,', ,, ,, , ,, ,, , , , ,I , , , , ,, , ,, 2 , , I , ,, , ,, 1.6 ,I , , ,, ,, , , , , ,, ,I , ,, , ,. , , , , , , , , ,
"
,
,
i
, , ,,, , ,, I , , I , , , I , ,, , I ,,, ,, , , , , ,I , , ,,I , , ,, , ! ,,,I , , ,, , ,
" •• 1
i
i
i
,
i
i
i
i
,
:
, , ,i ,
,
, , , ,
i
,
,
,
,
, , I
o
,, : ",'
, ,r<>
,
"1
,
,i
i
I
,
i
...
, , ,, ,
I
, ,
,
I i
i
,
, , , , , ,
i
3
i
i
i
,
,
I I
,
,
,
, ,
, ,, , ,, , , ,,, ,, ,. , , , . , ,,I , ,
, ,
,
,
= 0.125 , ,, ,I , , , , , , , ,! : , , , , , , ,, , ,: ,, 1 , I I ,I , ,, , , ,, i
i
,
, ,:
,
,,
,, ,
I
4
"
r,
i
I
i
,
i
5
FOR Ao/A w = 0.125; FORW/H = 1.33, GRAPHIC INTERPOLATION YIELDS T, = 1.60ms
2A-61
,
,I ,,, ,, ,,,, , ,,, , , , ,, , , , ,, , , , I , ,I , ,
, , ,,
NOTE: SEE EXAMPLE 2A-II,:STEP 3,PART B ,ITEM c
2 A-IS
i
i
W/H
FIGURE
,
,
,
, , ,,, , , , ,. ,I , ,: ,
, ,
i
.,
I
i I !
I I
, I ,I , , , I I -r , , ,, , ,, , ,, ,I , , ,I , , , ,,, ,, , ! I , ,,, , ,, ,, ,, , , , ,I , : , , , , , , ,, , , , , , ,I , , , , ,: I I ,,: , ,
I
,
,
, ,
,
,! , , , , , , I
, , , , ,, ,,, I ,,,
,
i
,
,
, ,
2
, , , , , , , II , ,, I
,
, ,,
Ao/A w , , ,
,,, , ,: I ,, I , , , , i
!
I
,
, , ,, , , , , , ,, , , I , ,, ,,:
i
,
,,
,
, ,.
,i
,
.,
,
,
! ! I ,
i
i
,
,
i
i
I
i
i
,
i
i
i
"
i
i
i
I
,
, , , ,,, " I , ,I , ,, I , , , i ,,, I I , , , , , , I
,,,, ,,,, I t. , , ,, ,,, , ,. ,, ,,, , , ,, ,, ,, , ,
,
i
,
,
,
i
i
,
,,
,
,
,,,
6
NOTE: .
W/H=3 -,
14.0 ~
III
E
~
.=I
10
6.75
t\I
~
..
WIH=I.
l'H=O:TS ,
2.07' N
2
. I
3
SEE EXAMPLE 2A-11 STEP 3, PART B,ITEM d. FOR LW/H =2.54,GRAPHIC INCORPORATION YIELDS T2 - T1 SUMMARIZED BELOW
WIH
~li
0.715
2.07 '6.75
1.5 3
I·
14.0
,Lw IH (a)
20
Lw /H=25
,
'iii
.§
·F.OR W/H = I. 33
i-=" 10
T2 - TI = 5.8ms
I t\I
~
5.8m
,r<>
I
'"
2
W/H
~.
(b)
-;
".'.
NOTE:
FIGURE
2A - 19· 21\" liZ-
TK 5-l300/NAVFAC P-397/AFR 88-22 .75 .75 1. 5 5.28.2
.75 10.0
3
14.5
1.5 1.5 13.5
3
3' 19.5
.75 18.6
1. 5 24.8
3
34.8
Plot Figure 2A-20,with above values, and graphically interpolate to determine T3 - Tl at Lw/H - 2.54 for W/H'- .75, 1.5, and 3, as summarized below. , 3 .75 1.5 31.7' iz .a 17.5 .
Plot Figure 2A-20 with above values, and determine. T3 ms at W/H - 1.33.
Tl -'16.3
T1 - 1. 60 ms
T2
Tl - 5.80 ms
I3
I1 - 16.30 ms
T2 - 5.80 + 1.60
~'7.40 m~~
T3 - 16.30 + 1.60 - 17.9 ms Plot Figure 2A-23 using above values and Prnax - 18.9 psi. Step 4. a.
Sidewall idealized interior pressure-time blast load. Using equation 2·15, with Ps o- 12.6 psi; Ao/Aw ~ 1/8, Lw/L1.27, L/H - 2.0, solve for Pmax for. W/H - .75, 1.5. and 3. Equation 2·15:
Pmax - K/(Lw/L) For W/H - .75, K - A x Bx CE x FH x psoO.97l8
where:
A,:" [0.5422 (Lw/L) 1. 2944] - ,0.001829 "' 0.7385': B - [0.654 + 2.616 (Ao/Aw) - 4.928 (Ao/Aw) )2' x [2.209 (L/H)-0.34~1 ~. 0.739] - 0.9040
c - [0.829
x [L/H]D - 1.0388 D
-
+ 0.104 (Lw/L)l'~l + [0. Q0124 + -0_00414 (L",;/L) 3. 334.J
"
2.579 - .0534 (Lw/L)~·~9l =2.4428 -.1,
."
,
,
,
,
,
,
,,
40
, ,
,
, ,
,
, , ,
31.7
,, ,
I
,
30 ,
~
,
,,
'" E
,
~
f- 20
17.5
~
,, ,
, ,
-
12.8
10
,
"
,
, ,
,
,
.
, ,
It:,
, , ,, ,
, , ,,
,
,
W/H=015
,
,,
,
,, ,
, ,
, ,, , ,
, , ,,
W/H
T:3-T,
0.75
12.8 17.5
1.5.
, ,
, ,
o
:3
31.7
6
56.5
4
0.75
0
,
,
W/H=I.5
,,
,
, ,, , ,
,
I ,
W/H =3 ' .. ;
,
, ,
I
,
,
NOTE: SEE EXAMPLE 2A-II, STEP 3,PART B. ITEM e. FOR Lw/H = 2.54 I GRAPHIC INTERPOLATION YIELDS T3- TI SUMMARIZED BELOW
,"
40
,,
,,
,
,
,
,
,
,
30
NOTE:
,,
FOR W/H = 1.33 T 3 - T 1z I 6 .3 ms
,'
10
Ir<> r<>
o W/H ( b)
FIGURE
2A - 20
2A- 64 .
TM 5-1300/NAVFAC P-397/AFR 88-22
E - 999 (Ao/Aw)9.964 - 0.000001002 F - 1.468 - 1.6627 (Ao/Aw)0.78~1 + [1.8692 1.1735 (Ao/Aw)-0.2226] x [Lw/LjG - 1.1592 G - 0.2979 (Ao/Aw)-1.4872 - 0.8351 - 5.7286 • H - (5.425 x 10- 4) + (1.001 x 10- 3) (L/H)9.965 - 0.0005435
K - 0.7385 x 0.9040 x 1.03880.000001002 x 1.15920.0005435 x 12.60.9718 - 7.83 Pmax - 7.83/1.27 z 6.2 psi, for W/H - .75 For 1.5 $ W/H $ 6, K - [A + [8 x (Lw/L)C]] x D x E x pso1.025
A - 0.002 (W/H)1.4467 - 0.0213 8 - 2.2075 - ( 1.902 ( W/H )_0.085)
C - 1.231 + ( 0.0008 ( W/H )2.678 )
D - (2.573 (L/H )-0.444) -0.3911 E - 0.4221 + (1.241 (Ao/Aw )0.367)
, For W/H - 1.5, 3, and 6, determine values for A to E, and K and Pmax' as summarized below.
C
D
-E
W/H
A
1.5
,-0.0177
0.3700
1.2334
1.5003
1.0006
3
-0.0115
0.4751
1.2462
1.5003
1.0006 12.690
8
2A-65
9.6743
7.60 9.98
TM 5-1300/NAVFAC P-397/AFR
88~22
Plot Figure 2A-21 with above values of Pmax vsW/H- - .75, 1.5, and 3, and determine Pmax - 7.30 psi for W/H - 1.33. b.
For Ps o - 12.6 psi, determine Tl and T2 for'W/H - .75, 1.5, and 3 from Figure 2-213. as summarized below. W/H
.75
1.5
3
1.82
4.95
11.7
3.20
6.50
12.5
Plot Figure 2A-21 using above'values of'Y/H and Tl,and T 2, and determine Tl - 4.4 ms and T2 - 5.9 ms for W/H - 1.33. c.
For Ps o - 12.6 psi and'L/H - 2, determine T3 and,T 4 for W/H - .75, 1.5 and 3 and Lw/L - .75, I, 1.5, and 2, from Figures 2-218 to 2-221, as summarized below: Plot Figure 2A-21 with above values, and interpolate to determine T3 and T4, for Lw/L ~ 1.27 for W/H - .75, 1.5, and , .! 3, as summarized below . • i
W/H
.75'
1.5
3
35
18
24
71
71
74.5
Plot Figure 2A-21 with above values, and determine T - 19 3 ms and T4 - 71 ms, for W/H -' 1.33. d. Step' 5.
For Pmax - 7.3 psi, Tl - 4.4 ms, T2 - 5.9 ms, T3 - 19 ms and T4 - 71 ms' plot Figure 2A-23. , Back wall idealized interior pressure-time blast load.
a.
For L/H,- 2, Ps o - 12.6 psi and Ao/Aw -' 1/8, determine PRIB/Pso - 0.575 from Figure 2-234.
b.
PRI B - (PRIB/Pso) x Ps o - 0.575 x 12.6
2A-66
-.7.5 psi.
...
15
15
. T2
-
-;; 0. .......
7.3
)(
5
e,
I-
.5 4.4
en E .......
0
E
. T1
10 5.
Ii<) i<)
3
2
0
2
0
W/H
W/H
(a)
(b)
3
75 71.0 70 6
100
T4
90
60 5
80 74. 70 ~
en
710
E 60 ....... ILl
:::E I-
W/H=075
W/H=3
240
-so i<)
:::E
W/H=0.75
50
3
50
en 4 E ....... ILl 4
W/H=1.
50 40
-
W/H=3
I-
}T'
35 3
20
25
180 10 _
W/H=1.5 20
I"N
19.0
0
15 0
.5
.75
Lw
1.5
/
2
0
.5
1.5
W/H
L
(c)
(d)
FIGURE
2A-21
2A-67
.
2
2.5
3
TM 5-l300/NAVFAC P-397/AFR 88-22
c.
For Ps o - 12.6 psi, L/H - 2, and Ao/Aw - 1/8, determine Tl for W/H - .75, 1.5 and 3 from Figures 2-230 and 2-231, as is shown below . . 75 19.0
1. 5 21.5
3
24.0
Plot Figure 2A-22 using above values, and determine T _ l 21.1 ms for W/H - 1.33. d.
For Pi o - 12'.6 psi and,Ao/Aw - 1/8, determine T2.- Tl - 2•. 35 ms from Figure 2-232. Determine T2 - (T 2
e.
Step 6. a.
C
Tl) :+ Tl - 2.35 ms + 21.1 ms - 23. 5ms.,
Plot Figure 2A-23 using PR1B - 7.5 psi, Tl - 21.1 ms and T 2 - 23.5 ms. Roof idealized interior pressure-time blast load Sidewall W/H - 1-1/3 - 4/3 Roof W/H - 1/(4/3) - 3/4
b.
, Repeat Step 4 with W/H - 3/4 "
For W/H - 3/4, L/H - 2.0, Ao/Aw - 1/8, L,.,/L - 1.272 and Ps o-: 12.6, ~ Pmax - 6.2 psi For W/H
=
3/4 and Ps o - 12.6 psi
,
.
Tl - 1;82 and T2,- 3.20 ms For W/H - 3/4, Ps o - 12:6 psi, and L,.,/L -1.27 T3 - '35, and T4 ., 65 ms Plot Figure 2A-23 with above values.
2A-68
25 .
I
, , ~
, I
: ! I
,
I I I
,
,
21.1
,I
;
,
,
I
I
,,
,,
I
,
,,
,
I
,
I
~
11_.
I
20 !
1 I
I' 11
,
,
!
:
I I
' I
!
,, ,
,
I
,
,, , ,, I ,
I
,
,
,
, , , I
,,
I
,,
, I
,
,
I
,, , ,
,,
l"j
,
,
.r
,,
,
,
, ,,
,, ,,
10
o.
.5
,
1
:
!
,
I
,
!
I
I"
,,
I
i
2
FIGURE ',,'
I
I
,
.. '.,
:
,
,
,
I
,;
; \.:
!
:
:
I
I
I
I I
I
,
W/H
.'
,
,
+
:
'
t
:
,
~'
I
~
; i
,oro ' .
,
l~l
I
3
(0) EXTERIOR FRONT WALL
iii 20 Q.
12.7 10
o
5.3
26 30
I
40
50
60 '
7.
. 80
ms
(b) INTERIOR FRONT WALL 20
iii 10 Q.
01.6
.410
17.9 20
3
4
50
6
7
3
4
5
6
70
5
6
7
8
5
6
7
8
ms
8
(C) INTERIOR SIDE WALL10 7.3 en
Q.
4.4
.9 I
·19
ms
71
8
(d) INTERIOR BACK WALL 10 7.2
-
en
Q.
-t-
O
21.1
3.5
3
ms
(e) INTERIOR ROOF 10 6.2 '0; Q.
o
\3.2 1.8
2
30
35
4
ms
FIGURE 2A-23 BLAST LOAD SUMMARY 2A- 70 -
I
TM 5-l300/NAVFAC P-397/AFR 88-22
Problem 2A-12. Interior Pressure BUildup in a Structure -rProblem:
Determine the interior pressure· time curve for a structure
with an opening in one of its walls and subjected to an applied blast pressure. Procedure: Step 1.
I
I Determine
,
,
the pressure;time history of the applied blast
p~essure P acting on the:wall surrounding the opening ,in the
structure as presented in problem 2A~lO. Also the area of the opening Ao and the v?l~e of the structure Vo must be
. known.
Step
2.
'~
Divide the duration to ~f the applied pressure into n equal iriterVa~s ct, ~ach interval bei?g ~pproximately t~/lO to, ; t / 20 , and determine the pressures 'at the end of ~ach 'inter, val.
o
Step 3.
Compute 'the pressure differential P:P i where Pi'is the interior pressure. Obtain the leakage pressure coefficient CL for each,P-P i from Fig~re 2-235.
Step 4.
Calculate cP i from cP i - CL
Ao
,
.
at
Vo
using 'the proper values for CL and ct, Add cP i tO'P i for the interval being considered to obtain the new value of Pi for the next interval, Step 5.
Repeat steps 3 and 4 for each interval using the proper values of P and,P i. Plot curve of pressure buildup.
Note: When P-P i becomes negative, the value of CL must be taken as
negative also.
Example 2A-12. Interior Pressure Buildup in a Structure Required:
Interior pressure· time curve for a structure with an opening
in one of its walls and subjected to an applied blast pressure.
Solution:
t e
Step 1.
The curve of the applied blast pressure P for the wall in question is shown in Figure 2A-24. (Only the positive phase of the blast wave is considered in this example.)
2A-71
3 APPLIED BLAST PRESSURE ON WALL C/)
a,
LU
a:
::l
2
(/)
(/)
LU
a: c,
INTERIOR PRESSURE
o
O· .
10'
20
30 TI~E (ms)
FIGURE 2 A - 24
2A-72
40
50
60
TH 5-1300/NAVFAC P-397/AFR 88-22
Area of opening
Ao - 3' x 3' - 9 sq ft.
Volume of structure
Va - 10' x 10' x 10' Va - 1,000 cu ft
Step 2.
to -. 55 ms
Use n - lOot - 5.5 ms For the first interval, P - 3.5.psi at. t - 0 Pi - 0 for the first interval
Step 3.
.
.
. . P - Pi - 3.50 - 0 - 3.50 psi (fig. 2-235) (eq. 2-31)
Step 4. oP i - 8.75 (9.0 / 1000) (5.5) - 0.433 psi new Pi = 0 + 0.433 = 0.433 psi
The remainder of the analysis is presented in tabular form below and the pressure buildup within the structure is plotted in Figure 2A-24.
Step 5.
t(l)
P
Pi
P-P i
CL
(ms)
(psi) 3.50 3.15 2.80 2.45 2.10 1. 75 1.40 1. 05 0.70 0.35 O. O. O. O. O.
(psi) O. 0.433 0.780 1.05 1. 24 1. 36 1.41 1.41 1. 36 1.27 1.14 .97 . 82 .70 .59 .49
(psi) 3.50 2.72 2.02 1.40 0.86 0.39 -0.01 -0.36 -0.66 ~O. 92 -1. 14 -0.97 -0.82 . -0.70 -0.59 -0.49
(psi-ft/ms) 8.75 7.00 5.45 3.78 2.32 1'.05 -0.027 -0.972 -1. 78 -2.48 -3.50 -3.00 -2.50 . -2.25 -2.00 -1.80
o.
t
e
5.5 11.0 16.5 22.0 27.5 33.0 38.5 44.0 49.5 55.0 60.5 66.0 71.5 77 .0 82.5
o.
2A-73
oP i (psi) 0.433 0.347 0.270 0.187 0.115 0.052 -0.0013 -0.048 -0.088 -0.123 -0.173 -0.148 -0.124 -0.111 -0.099 -0.089
TM 5-1300/NAVFAC P-397/AFR 88-22
P
Pi
P-P i
CL
6P i
(ms)
(psi)
(psi)
(psi)
(psi-ft/ms)
(psi)
88.0 93.5 99.0 104.5 110.0 115.5 121.0 126.5 132.0
O. O. O. O. O. O. O. O. O.
.40 .32 .25 .19 .13 .08 .04 .01
- .40 -.32 -.25 -.19 - .13 -.08 - .04 -.01
-1.60 -1.45 -1. 30 -1.15 -1.0 - .85 - .05 - .2
- .079 - .071 - .064 -.056 -.050 - .042 - .025 -.01
O.
O.
O.
(1)
O.
Maximum Pi occurs between t - 27.5 and·t - 33.0 ms
Problem 2A-13 Problem:
.'
t(l)
Primary Fragments from Cased Cylindrical Charges
Determine the average fragment weight for a primary fragment ejected from a uniform cylindrical steel casing; the total number of fragments, the design fragment weight .arid the number of Er-agments weighing more than the design fragment.
Procedure:
Step 1. a.
Establish design parameters: .. type of explosive.
b.' average casing thickness,
t
e.
c.
average inside diameter of casing, dio
d.
total casing weight, Wc'
e.
confidence level, CL.
Step 2.
Determine the value of the explosive constant B for the given type of explosive from table 2-6. With this value and the values of t c and d i from step:1, calculate the fragment distribution MA from l
(eq. 2-37) Step 3.
With the value of MA from step 2, calculate the average weight of the fragments from (eq. 2-40)
2A-74
TH 5-1300/NAVFAC P-397/AFR 88-22
Step 4.
Calculate the total number of fragments using the value of Wc from step 1d and MA from step 2 and equation 2-37. (eq. 2·39)
or:
With the values of d i and t c from step 1, enter Figure 2-241 and determine B2NT/Wc ' From this value, the value of B from step 2 and Wc from step 1 find NT'
Step 5.
Find the design fragment weight for the confidence level CL , given in step 1, using the value of MA and (eq. 2-42)
or equation 2-43 if CL > 0.9999 Step 6.
Using the value of Wc from step 1, MA from step 2 and Wf from step 5, determine the number of fragments which weigh more than the design fragment from
(eq. 2-36)
Nf - - - - - - - : : : - - - - - -
M 2 A
or:
Calculate the number of fragments which weigh more than the design fragment using the confidence level of step 1, the total number of fragments from step 4 and equation 2-244.
Example 2A-13 Required:
Primary Fragments from Cased Cylindrical Charges
The average fragment weight, the total number of fragments, the design fragmient weight and the number of fragments weighing more than the design fragment.
Solution: Step 1.
Given:
a.
type of explosive:
Comp B
b.
average casing thickness:
c.
average inside diameter of casing:
2A-75
t c - 0.50 inch d i - 12.0 inches
TM 5-1300/NAVFAC P-397/AFR 88-22
d.
total casing weight:
e.
confidence level:
Step 2.
Wc - 65.0 1bs
CL - 0.95
For Comp B, B - 0.22 (table 2-7) MA - Btc5/6di1/3(1 + tc/d i)
(eq. 2-37)
- 0.22 (0.5)5/6 (12)1/3 (1 + 0.5/12) _ 0.294 Step 3.
Average weight of, fragments. (eq . 2-40)
Step 4.
Total number of fragments. NT - 8Wc / MA2 - 8 X 65/0.294 2 - 6016 fragments
or:
B2NT
- 0.28
(e q .
2·39)
(fig. 2-37)
Wc NT - 0.28 X (65 X 16)/;222 - 6016 fragments Step 5.
Design fragment weight. 2
Wf - 2MA
1n2(1 - CL) .., 0.294 2 1n 2 (l - 0.95) _ 0.78 oz (eq , 2-42)
Step 6 ..
Number of fragments weighing more than W - 0.78 oz. f
(8 X 65) e
. (0.78)1/2/ 0. 294
,Nr- - - - - - - ' - - - - - M 2. ·A
- 298 fragments
"'.
(eq. 2-36)
>,
2A-76
TN 5-1300/NAVFAC P-397/AFR 88-22 (eq. 2·44)
or: Nf
NT (1 • CL) - 6016 (1 • 0.95) - 301 fragments
Problem 2A-14 Primary Fragment Velocity Problem:
Determine the initial velocity of a primary fragment and its striking velocity.
Procedure:
Step 1.
Establish design parameters.
a.
shape of charge
b.
dimensions of charge
c.
type and density of explosive
d.
type and density of casing
e.
distance from center of charge to impact location
f.
weight of fragment
Step 2.
Calculate the total weight of the explosive Wand increase it 20%. Find the weight of the casing Wc' Also calculate the ratio of the explosive weight to the casing weight W/Wc'
Step 3.
Determine the Gurney Energy Constant (2E,)1/2 for the explo· sive charge from table 2·5. With this value and the value of W/Wc from step 2, calculate the initial Vo of the primary fragments from the equation chosen from table 2·6.
or:
Calculate the casing to charge weight ratio Wc/W. With Wc/W, find the initial velocity from Figure 2·237 for proper shape.
Step 4.
For the distance traveled by the fragment Rf , calculate the striking velocity V s using the initial velocity from step 3, the weight of the fragment from step Ig and (eq. 2·48)
vs -
2A-77
TM 5-1300/NAVFAC P-397/AFR 88-22
With the fragment weight Wf and striking distance Rf from step 1, enter Figure 2-243 and find the ratio of the striking velocity to initial velocity. Multiply the ratio by the initial velocity V o from step 3 to find the striking velocity v s'
or:
Example 2A-14 Required:
Primary Fragment Velocity
The initial velocity and striking velocity of a primary fragment.
Solution: Step 1.
Given:
a.
spherical charge
b.
inner diameter of charge:
di
average casing thickness:
t c - 0.25 inches
c.
type of explosive TNT 0.0558 Ib/in 3
density of explosive d.
6 inches
mild steel casing density of casing - 0.283 1b/in3
e.
striking distance Rf - 35 ft.
f.
weight of fragment, Wf - 2 oz
Step 2. a.
weight of the explosive W - 4/3 n (6/2)3
b.
X 0.0558 - 6.31 1bs
Increase weight of explosive 20 percent W = 1.20 X 6.31 - 7.57 1b
c.
Weight of casing Wc - 4/3 n (3.5 3 - 3 3)
d.
X 0.283 -
18.82 1b
Explosive weight to casing weight ratio. W/W c - 7.57/18.82 - 0.402
Step 3.
For TNT, (2E,)1/2 -
8000 (table 2-5)
Initial velocity from table 2-6 2A-78
TK 5-l300/NAVFAC P-397/AFR 88-22
v
o
_ (2E·)1/2
1/2
0.40 vo -
(8000) '[
1 + 0.40 (3/5)
1'
4500 ft/sec
18.82/7.57 . 2.49
or:
from Figure 2-237 v o/(2E·)1/2- 0.56 v o- 0.56 X 8000 - 4500 ft/sec, Step 4.
Striking velocity "s V
or:
voe
-0.004 Rf/Wfl/3
-0.004 x (35/2)1/3 - 4500 e
(eq. 2-48)
s - 4030 ft/sec
from Figure 2-243
V
s - 0.895 X 4500 - 4030 ft/sec
Problem 2A-15 , Unconstrained Secondary Fragments "Close" to a Charge Problem:
Determine the velocity of an unconstrained object close to an explosive charge and its maximum range.
Procedure:
Ie 2A-79.
TM 5-l300/NAVFAC P-397/AFR 88-22
Step 1.
Establish design parameters
a.
weight Wand shape of TNT equivalent explosive
b.
radius of explosive Re
c.
shape, dimensions and weight of target
d.
distance from the center of the explosive charge .to the surface of the target, R
e.
orientation of target with respect to the explosive charge
f.
mass density of air Po
Step 2.
Calculate the ratio of standoff distance to radius of the explosive R/R e using the values from step 1.
Step 3.
From Figure 2-249; determine target shape factor,
B.
Step 4. a.
If R/R e is less than or equal to 10 for cylindrical charges, or less than equal to 5.0 for spherical ,charges, determine the specific acquired impulse either from Figure 2-250 or ~quations 2-60, ~r 2-61.
b.
If 10 < R/R e 5 20 for a cylindrical charge, or 5 < R/R e 5 20 for a spherical charge, calculate the scaled standoff distance ZA - R/W 1 3 . With that value of ZA' obtain the normal reflected impulse from Figure 2-7. The normal reflected impulse is then used as the specific acquired impulse.
Step 5.
Calculate the mean presented area A.of 'the target and the mass M using the values of step 1.
Step 6.
With .the area and mass from step 4,' the target shape factor B from step 2 and the impulse from step 3, calculate the velocity from 1000 ABi vo -
Step 7.
(eq. 2-59) 12 M
Determine the drag coefficient CD from table 2-8. Using that value of CD' the area and mass of the target from step 5, the velocity from step 6 and the mass density of air from step 1f, evaluate the term:
2A·80
TK 5-l300/NAVFAC P-397/AFR 88-22
With the term calculated in step 7, enter Fig~re 2-252 and read the value of 12 poCOAOR/M from ~hich the range R is calculated.
Step 8.
Example 2A-15
Unconstrained Secondary Fragments "Close" to a Charge ".
Required:
':
.
The velocity and maximum range of a steel tool holder resting on a lathe when a charge being held by. the lathe explodes.
Solution: Step 1. a.
Given: spherical charge.of TNT 15 Lbs
W-
b.
c.
d.
e.
coo I holder is, resting so Sthat "its longitudinal axis is perpendicular to the radial line from the charge
f.
mass density of air:
Step 2.
Po - 0.115 Ib - ms 2/in 4
Ratio of standoff.distance to radius of explosive: R/R e - 1.0/0.33 - 3.03
Step 3. B -
Step 4.
Target shape factor.
n/4 from Figure 2-512 Oetermine specific acquired impulse' , psi-ms
a.
8000
(fig 2·250) ft
b.
Re f f - Re - 0.33 ft
2A-81'
(eq. 2·61)
TM 5-1300/NAVFAC P-397/AFR 88-22
c.
i -
"/4 (0.33 ft~ (0.083/0.33)0.158 X 8000 psi-ms/ft
i - 1667 ps L-ms Step 5. a.
.
',Calculate area and jnas s of target:
,
Mean presented area .A - 2.0,in·X 8.0 in - 16i'n2
b.
Mass
Wt M- - - - g
Step 6.
7.13 1b -:-_--;:-
32.2 x 12x10- 6in/ms 2
-
18,450 in
Find the velocity. 1000 ABi - 1000 X 16 X ("/4) X 1667 V
o -.----
12 M
Vo
Step 7. ,.' " a, b.
- 95 ft/sec
...
Evaluate· toe term
12 PoCoADvo 2/Mg.
CD - 1.2 12P oCoADVo2
( table 2-8) 0.115
Mg Step 8.
(eq. 2-59)
12 X 18450
(1. 2) (16) (95)2 12 18450 X 32.2
- 0.40
Calculate the range.
a.
- 0.33
b.
R - 0.33M/(12p oCoAo) - 0.33 (18450)/[12 (0.115) 1.2 (16)J - 230 ft.
2A-82
(fig. 2-252)
TM 5-J.300/NAVFAC P-397/AFR 88-22 Problem: 2A-16 Problem:
Unconstrained Secondary Fragments "Far" from a Charge
Determine the velocity of an unconstrained object nfar ll from an explosive charge.
Procedure:
Step 1.
Establish design parameters;
a.
weight W of TNT equivalent explosive
b.
shape, dimensions and weight of target
c.
distance from the center of explosive charge to surface of
the target d.
orientation and location of target with respect to the explosive charge
e.
velocity of sound in air, a o
f.
atmospheric pressure, Po
Step 2.
From
Calculate the scaled standoff distance from:
Figur~
2-7. and the scaled distance find the peak incident
overpressure and the incident specific impllise.
Step 3.
Determine the drag coefficient CD from table 2-8 based on the shape and orientation of target (step 1).
Step 4.
Calc;"late the mass of the target. Determine the distance from the front of the target to the location of its largest cross-sectional area, X. Also, determine the minimum transverse distance of the mean presented area,
H,
and the pre-
sented area.
Step 5.
Determine the constant K, which is equal
is on the ground or reflecting surface. the air, K is equal to 2.
Step 6.
to 4 if the object If the target is in
With the peak incident overpressure Pso from step 2 and the atmospheric pressure Po f r orn step If, find Pso/po'
2A-83
TM 5-l300/NAVFAC P-397/AFR 88-22
Step 7.
Evaluate the term l2CDisao/lO?[Pso(KH.+ X)] using is and Ps o from step 2 .. CD from step 3. a o from step Le , K from step 5 and H and X from step 4.
Step 8.
With two terms calculated in steps 6 and 7 enter Figure 2248 and read l44voMao/[106PoA(KH + X)] from which the veloc-. ity is calculated.
Example 2A-16 Required: Solution: Step 1.
Unconstrained'Secondary Fragments "Far" from Charge
The initial velocity of a steel tool holder resting on a nearby .table, when a charge explodes., Given: .a .
b.
"
weight of explosive:
W:.- l5·1bs of TNT
Rt
8.0 in 1.0 in
Wt -
7.13 lb
cylindrical target: length
c.
standoff distance:
d.
tool holder is resting on a table so that its longitudinal axis is perpendicular to the radial line from the charge
e.
velocity of sound in air:
f.
atmospheric pressure:
Step 2.
a o - 1100 ft/sec
Po - 14.7 psi
Find the peak incident overpressure and the' incident specific impulse.
a.
R - 10 ft
r,
,Scaled distance ZA - R/Wl/3 - 10/(15)1/3 - 4.05 ft/lb l/ 3
b.
Peak incident overpressure.
Ps o - 39 psi c.
(fig -, 2-7)
Incident specific impulse.' i s/W l/3 - ·12 psi-ms/lb l/ 3
'.
....
is - 12 (15)1/3 - 29.6 psi-ms
2A-84
(fig .. F7)
TM 5-1300/NAVFAC P-397/AFR 88-22
Drag coefficient.
Step 3.
(from table 2-8 for cylinder loaded perpendicular to axis.)
CD - 1 Step 4.
Mass of target.
a.
Wt
7.13 1b
M - _.- - -----.....,.....--_::_
'g ,
b.
18,450 in
Location of largest cross-section.
xc.
32.2 x 12x10- 6 in/ms 2
1 in (radius of object in this case - see fig. 2A-21)
Transverse distance of presented area.
H = 2.0 in
(diameter of object in this case - see fig. 2A25) .
;
d.
Mean 'presented area.. A - 2 X 8 - 16 in 2
Step 5.
Re f Le c t.Lon constant".
Target is resting on table which is a reflecting surface so: K -
Step 6.
4
Evaluate
Pso/po
Pso/po - 39/14.7 - 2.65 Step 7. 12 (1.2) (29.6) (1100) 10 3 (39) (2 X 4 + 1)
Step 8.
- 1.34
Calculate the velocity. - 6.0
(fig. 2-248)
10 6.(6.0)(14.7)(16)(2 X 4 + 1) vo -
(144)(18450)(1100) 4.34 ft/sec
I
2A-85
., .
'j
"
-,
"
FIGURE
2A-25
2A- 86"
.
TM 5-1300/NAVFAC P-397/AFR 88-22 Problerl. 2A-17 Problem:
Constrained Secondary Fragments
Determine .t.he velocity of a constrained object close to an explod-
ing
charg~. '~
Procedure:
Step 1.
:
Establish design parameters:
a.
fra~ment
b.
dim~nsions'~f object.
c.
boundary condition, cantilever or fixed-fixed.
d.
specific impulse imparted,to object.
material.
",.
,
"
Step 2.
"
Det~rmine the fragment toughness T from table 2-9 and the
fragment mass density Pi:. " Step 3.
Calculate the loaded area of the object.
Step 4.
Evaluate the term ib (2L/b)O.3/[A(Pf T)O.5] ,using the specific impulse and object dimensions from step 1, the fragment density and toughness from step 2 and the loaded area from step 3. With this term enter Figure 2-251 and read the value of l2V/1000 (pf/T)O.5 from which the velocity is calculated. I
or:
Using the specific impulse and object dimensions from step
1, the fragment density and toughness from step 2, and the loaded area from step 3. calculate the· velocity of the object from-equation 2-66. Example 2A-17 Constrained Secondary Fragments Required:
The velocity of a cylindrical tool holder after it breaks free of its moorings.
".,'
'.
Solution: Step 1.
Given:
a.
fragment material; A36 steel
b.
dimensions of object:
c.
boundary conditions; cantilever
d.
specific impulse: i - 1667 psi~ms (see step 4 of example 2A15).
I
b 2.0 inches L - 8.0 inches
,
2A-87 .
TM 5-1300/NAVFAC P-397/AFR 88-22 Step 2. Fragment· toughness ..
a.
T - 12,000 in-lb/in 3 b.
(table 2-9)
Mass density of steel.
Pf -
32.2 ft/s 2 (12 in/ft) (s2 /10 6ms 2) .
Step 3.
Loaded area.
Step 4.
Calculate the velocity.·
.
-
734 - - in4 .
"
ib
1667 x 2.0
A(p . f T)0.5
3.14(734 x 12,000)0,5
- 0.668
from Figure 2-251 "
12/1000
'.
[pf/T J1/2 V - 0.025
so 1000
1000 0.025 - - 12
V -
12
12 000
[ ~34]
1/2 0.025 - 8.4 fps
or from equation 2-66
V -
[~J 1/2
1000 12
Pf
1000 12
[ -.2369
12.000 [
] 1/2 (-.2369 + 0.3931 X 0.668)·
734 .
.
v - 8.6 fps
2A-88
TM 5-l300/NAVFAC P-397/AFR 88-22
Problem 2A-18 Problem:
Ground·Shock Load
Determine the air blast and direct induced ground shock parameters.
, Procedure:
Air blast-induced ground shock.
Step 1.
Determine the charge weight, ground distance R, height of burst He l if any, and structure dimensions.
Step 2.
Apply a 20% safety factor to the charge weight.
Step 3.
Calculate the scaled distance Z.
Read: From fig. 2-15 a.
Peak positive incident pressure Ps o'
b.
Scaled unit positive· incident impulse i s;Wl/3. scaled value by Wl!3 to obtain absolute value.
c.
Shock front velocity U.
Step 4. a.
Multiply
Determine the maximum vertical ground motions.
Calculate maximum vertical velocity.
(eq. 2-74)
where:
p - Mass density of soil
(table 2-10)
Cp - Compression wave seismic velocity in the soil . .' " .", (table 2-11)
b.
Calculate maximum vertical displacement (eq. 2-74)
1,000 p Cp
2A-89
TM 5-l300/NAVFAC P-397/AFR 88-22
c. Calculate maximum vertical acceleration of the ground surface. lOOP s o AV -
(eq. 2-76)
-----
where: g - Gravitational constant'equal to 32.2 ft/sec 2
Step 5.
Determine the .maximum horizontal ground motions parameters.
a.
Check Cp/12000 U > .707
b.
Calculate maximum horizontal velocity. (eq. 2-77)
.c.
Calculate maximum.horizontaldisplacement. (eq. 2- 78)
d.
Calculate maximum horizontal acceleration~
(eq. 2-79) Step 6. a.
Determine arrival time t
A
and duration to:
Read from fig. 2-15. t A/W l/ 3 scaled time of arrival of blast wave and, t ;Wl/3 scaled duration of positive phase. 0
b.
Multiply scaled value by Wl/ 3 to obtain absolute value.
Direct·Induced Ground Shock Step 7. a.
Determine the maximum vertical ground motions. Calculate maximum vertical displacement.
2A-90
TM 5-l300/NAVFAC P-397/AFR 88-22
0.025 Rel/3wl/3 (eq. 2-80, rock media) Ze1. 3 or 0.17 Rel/3wl/3 (eq. 2-83, dry or saturated soil)
Z 2.3
c
b.
Calculate the maximum vertical velocity.
vV _ l50/Z e1. 5 c.
(eq. 2-85)
Calculate the maximum vertical acceleration.
(eq. 2-87) Step 8. a.
Determine the maximum horizontal ground motion parameters.
Calculate the maximum horizontal displacement. (eq. 2-82, rock media)
or (eq. 2_84, dry or saturated soil) b.
Calculate the maximum horizontal velocity. (eq. 2-86, all ground media)
c.
Calculate the maximum· horizontal acceleration. (eq. 2-88, dry soil) (eq. 2-89, wet soil or rock media)
Step 9.
Determine arrival time tAG:
2A-91
TM 5-1300/NAVFAC P-397/AFR 88-22
tAG -. 12000 RG/C p
(eq. 2-92)
Example 2A-18 Ground Shock Loads Required:
Maximum acceleration, velocity, and displacement at a point 155 ft. away from a surface burst of, 5,000 lbs. Also required are: times of arrival and duration ofoair-blast induced ground shock.
Solution: Air blast-induced ground shock. Step 1.
Given:
Step 2.
w-
Step 3.
Calculate the scaled distance Z.
b.
Step 4. a.
1.2 (5,000) - 6,000 lbs.
is
9 psi-ms/lb l/3
Wl/3 is -
c.
Charge weight - 5,000 Ibs., R - 155 ft., He - 12 ft.
9 x Wl/ 3 _ 9 x (6,000)1/3
163.54 psi-ms
U - 1. 5 ft/ms .Determine the maximum vertical ground 'motion.
Vv -
Ps o (eq. 2-74)
p Cp
13
Vv -
b.
Dv-
1. 65 x 10- 4 x 70,000
- 1. 125 in/sec
is 1,000 p Cp
(eq. 2-75)
TM 5-1300/NAVFAC P-397/AFR 88-22
0.0142 in
(eq. 2-76)
3~44
Step 5. a.
Determine the maximum horizontal ground motion.
Check Cp/12000 U > .707 70,000/12,000 X 1.5 - 3.89 > .707 .
b.
VH - Vv - 1.125 in/sec
c.
DH - DV - 0.0142 in
d.
AH - AV - 3.44 g
Step 6. a.
Arrival time
t
A
Read from Figure 2-15. 3.35 ms/1b 1/3
to~1/3 b.
g
- 2.35 ms/1b 1/3
t A ~ 3.35 X W1/3 - 3.35 (6,000)1/3 t A - 60.90 ms to - 2.35
X
W1/3 - 2.35 x (6,000)1/3
2A-93
TM 5-1300/NAVFAC P,397/AFR 88-22
to - 42.70
IDS
Direct-induced ground shock. Step 7.
Maximum vertical ground motions.
0.17 RG1/3 W1/ 3
a.
(eq. 2-83)
Z 2.3 G
0.17 (155)1/3 '(6,000)1/3 -
(8.53)2.3
0.1198 in
,.
150/ZGt'·5
b.
(eq. 2-85)
150 / 8.531.5
6.020' in/sec
10,000 / W1/3 ZG 2
c.
(eq. 2·87)
10,000 / [ (6,000)1/3 (8.53)2 1 - 7.56 g Step 8. a.
Horizontal ground motions.
DH
~
D V
(eq. 2-84)
DH - 0.1198 in. b.
VH - Vv
vH -
(eq. 2-86)
6.020 in/sec
c.
(eq. 2-88)
J .•
2A-94
.'
TM 5"1300/NAVFAC P-397/AFR 88-22'
AH
0:' 5 (7.56)
3.78 g Step 9.
,
'
Arrival time tAG
, - )..2 ,000 RG/~p - (12,000 x 155) / -r 70;000 •. ., . .;.. 26.6
(eq. 2-92)
ms,
)
1
. , Problem 2Ac19 . J'
Problem:
-,
I.
Structure Mot10n
Du~
to 'Air'Shock
. ~l
Determine the maximum horizontal acceleration, displacement and
velocity of an above ground structure subjected to air shock. Procedure: Step 1.
Determine external loadings 'acting on 'the roof, front and rear walls according to the procedure outlined in problem 2A-10. j',
Step 2.
;,
Construct the horizontal force-time load curve by ~ombining the frbnt and'vre ar wall loadings from step 1 applied over the arJa of front and rear walls. Use times of arrival to phase these two loads.
"
Step 3.
Calculate the dead weight and mass of the structure.
Step 4.
Construct the downward force-time curye by adding the weight of the structure and total roof load. The roof load is the pressure time loading from step 1 applied 'over the total area of the roof.
.
Step
5.'
Step 6.
,
Determine·the coefficient of friction between soil and the structure from table 2-12. Determine maximum horizontal acceleration, displacement and velocity using the acceleration impulse extrapolation method outl t nad-Ln Chapter 3, Ar.t LcLe 3-19.2.1.2 of this manual. The res~sting force at each time interval is equal to the value of downward force curve of step 4 multiplied by the coeffic~ent of friction determined in step 5. The resisting force is assumed to be effective when the total horizontal movement is equal to or larger than 1/4 inch as mentioned in paragraph
2A'-95 '
M 5-l300/NAVFAC . :P-397/AFR 88-22 ~
Example 2A-19 tequired:
Structure Motion Due to Air Shock
Maximum horizontal acceleration, velocity and displacement of the square structure shown in Figure 2A-9 from problem 2A-10 for a surface burst of 5,000 lbs at a distance from the front wall of 155 ft. Assume a coarse and compact soil. Roof, floor slab and side walls are 1 foot thick reinforced concrete slabs and assume a 50 psf of internal dead load for the structure.
Step 1.
External loadings on the structure are determined according to the procedure in example .2A-10. See Figures 2A-10. 2A-12 and 2A-13. The arrival time (t A) for these loads are tabulated in step 3d of example 2A-10. tAf (front wall) - 60.9 ms tAb .(rear wall) - 83.6 ms
Step 2.
a.
Calculate area of front and rear walls.
Area (front and rea,) b.
Calculate the time difference between the rear and front walls from step 1 . . 'ot - tAb - tAf - 83.6 -·60.9
c.
22.7 ms .
. Construct the horizontal force-time load curve by multiplying the values of the front and rear wall curves from step 1 by the area from step 2a. Rear wall load starts at time equal to ot - 22.7 from step 2b. See' Figure 2A-26.
Step 3. a.
Calculate dead weight of structure Wd ... Assume concrete weight is lSO.psf.
Wd - 150 [4 (30 - 1) (12-1) + 2· (30)2]' + 50 (30-2)2 - 500,600 lbs
2A-96·
w u .. 0-
ll::.c u,
2,000,000
100
~I
~2
14 120·
~r-==
~I
-124,921 -3~7.786
1"104711
COMBINED
FIGURE
333
.00
'00
2A -26
2A-97
~I
TIME,ms
TM 5-l300/NAVFAC P-397/AFR 88-22
Calculate total mass.
b.
wd
500,600 (1000)2
1295.55 x 10 6
mg Step 4.
a.
32.2 x (12)
in
Cal.culate area of the roof.
Area (roof)
b.
lb ms2
-
30 2x (12)2
129,600 in 2
Construct the downward force-time curve by multiplying the values of the roof curve from step 1 by the roof area from
step 4a, and adding, the. dead weight of structure Wd 500,600 Lb s from step 3a. (If the resulting value is negative, assume zero), See Figure 2A-27 below.
'c::.c "
0"
oIL
sao
600
398,347
100
~I
200
500
....
'" "
69032
TIME,ms
Fig1.lre 2A-27 Coefficient of friction
Step 5.
~
from table 2-12 for coarse and
compact soil. ~
. Step 6.
0.60 Using the acceleration impulse extrapolation method from
'Chapter 3 of this manual determine max~murn'horizontal a~del eration; displacement and velocity of the structure due to
load curve (P)
from,~tep
2c.
Resisting force R is the
friction force produced due to the downward load curve F
from step 4b after an initial lateral translation of 1/4 inch. Use two ms time intervals for extrapolation. The following are the equations used in the extrapolation shown
in Table 2A-3.
2A-98'
TM 5-l300/NAVFAC P-397/AFR 88-22
II\, - IJ.Fn Xu+l - 2Xu - Xu-l + an (ot)2
where
Pn
Load at time of step n (step 2c)
I\,
Resisting Friction Force at time step n
m - Mass from (step 3b) an
Acceleration at time step n.
IJ. - Friction coefficient (step 5) force at time step n (step 4b) Fn - Downward , Xu
Deflection at time step n
Vn
Velocity at time step n
The maximum ,motions from Table 2A-3 are: .00122 in/~s2 «
.01231 in/ms «
101.67 ft/sec 2 «
3.16 g's
1.026 ft/sec
Dmax - .355 -i n Problem 2A-20 Shock Response Spectra Problem:
Construct the elastic shock response spectra for the interior components of an above ground structur; subject to an external explosion.
Procedure:
2A-99
TK 5-1300/NAVFAC P-397/AFR 88-22 Table ZA-3
p.-R"
8 n' '"
(Pn-R.,)/m
a,.,(6t)Z
x..•
Xn _ 1
2X.
v,
Vn _1
8 n Gt
t
p.
R"
ms
lbs
lb.
o
0
1.781,208
o
1,781,208
(.00138)/2
(.00550)/2
1
2
1.582,190
o
1.582,190
.00122
.00489
.00550
.01038
.00244
.00138
2
4
1.363,173
o
I,3B3,I?3
.00107
.00427
,02077.00275.02229
.00214
.00382.00596
3
6
1,184,155
o
1.184,155
.00091
.00366
.04458
.01038
.03766
.00182
.00596
.00778
4
8
985,137
o
985,137
.00076
,00304
.07571
.02229
.05646
.00152
.00178
.00930
5
10
786,120
o
786,120
.00061
.00243
.11292
.03786
.07749
.00122
.00930
,01052
6
12
587,102
o
587,102
.00045
. 00161
.15497.05646
.10032
.00090
.01052
.01142
7
14
388,084
o
388,084
.00030
.00120
.20064
,07749
.12435
.00060
.01142
.01202
8
16
189,067
o
189,067
.00015
.00058
.24870.10032
.14896
'.00029
9
18
0
o
a
::0
0
.29793
,12435·
.17358
0
10
20
0
o
O"'~
,0
a
.34716
.14896
.19820
0
o
o
0
0
'~39640'
.17358
.22282 .
a
.01231
.
-50,013'
-.00004
-.00015
.44564
.19820
.24728
-.00008
.01231
.01223
n
11
22
0
in/ms z
lbs
.
0
0.00275 0
in/lIls
in/ms
in
in
in
in
(.00275)/2
in/ms
0.00138 .00382
.01202. 01231
...
.01231
.01231
.01231
.' .01231 ~
.01231
12
24
-50,013
o
13
26
-126,956
o
-126,956
-. 00010
-,00039
.49456
.22282'
·.27135
-. 00020
.01223
.01203
14
28
-203,890
733,536
-937,426
-.00072
-.00289
.54270
.24728
.29253
-.00142
.012,03
.01061
15
30
-280,843
713,896
-994,739
-. 00077
-.00307
.58505
.27135
.310629
-.00153
. 010~1
.00908
16
32
-357,786
694,256
-1,052,042
-.00081
-.00325
,62126
.29253
.32546,'.,. 00162
.00908
.00746
17
34
-337,457
674,617
-1,012,074
-.00078
-,00312
.65096
.310629
.33721
-,00156
.00746
.00590
18
.36
-317,128
~54,977
-972,1~5
-.00075
-.00300
.6744'1
.32546, .34593
-.00150
.00590
. 00440
19
38
-296,780
635,337
-932,117-
-.00072
-.00288
.69166
.33721
.35177
-.00144
20
40
-276,471
615,697
-892,168
-.00069
-.00275
.70354
.34593
.35466
-.00136
.00296
.00158
21
42
-256,142
596,057
-852,199
-.00066
-,.00263
.70972
.35177
.35532
-.00132
.00158
.00026
22
44
-240,745
555,710
-796,455
-.00061
-.00244
.71064
.35486
.35332
-.00122
. 00026
,00096
23
46
-278,004
504,214
.70664
.35532
24
48
-215,263
452,718
.35332
.T
2A-lOO
.' . 00 Iro 4.
.. 00296
TM 5-l300/NAVFAC P-397/AFR 88-22
Step 1.
Determine maximwn acce l.e ra t Lon cr ve l.oc Lty and displacement. due (to ground shock according to procedure outlined in Pro~lem 2A-18.
Step 2.
Determine maximum acceleration, velocity and displacement
due .t o air shock according to the procedure outlined in Problems 2A-10 and 2A-19. Step 3.
Determine if the ground shock is outrunning or superseismic
(paragraph 2-23.2).
For outrunning' ground shock the maximum
values of displacement, velocity and acceleration in hori-
zon~~l and
Step 4.
For Fuperseismic ground shock the maximum values of displacement, velocity and acceleration are the numerically
largkr values of direct-induced groun~ shock or the algebraic'sum of the maximum motions from air'shock and airburst indu~ed ground shock.
I
Step 5.
.
Calculate the magnitude of, acceleration, velocity and dispiac¢ment for r~sponse ~pectra 'in horrzo~tal and vertical directions by multiplying the maximum values of motions from step!3 or step 4 by their appropriate factor from paragraph 2-24,3. ~
.'
Step 6.
Draw; the horizontal and vertical shock response spectras.
,
Example 2A-20 Required:
Shock. response and 2A~19.r
spec~ra
Shock Response Spectra for the structure defined in Examples 2A-18
,~
Solution: Step 1.
Maximum values of motion in vertical and horizontal direc-
tions due to ground shock according to the procedure outlined in Example 2A-18 are: . a.
Air blast-induced 3.44 g 1.125 in/sec .014 in
b.
Direct-induced
2A~lOl
",
TM 5-l300/NAVFAC P-397/AFR 88-22
- 3.78 g A'V - 7.56 g VH -, Vv - 6.02
AH
,,
~
,
in/sec
DV ;. :120 'in
DH
Step 2. '. " Maximum horizontal acceleration, velocity and displacement due to air shock following the'procedure outlined in Examples 2A-lO and 2A-19. ,. "l"~
'Amax _. 3.16 g ,
.
Vma" - 12.31' in/sec
"••
.... _'
.355 in a.
Step 3.
'}
Check for, 'outrunning ground shock
From Example 2A-18 TAG - 26.6 ms TA
=
60.9 ms
TAG
.
b ·;
,
Add the values of max Lmueimo tLons f rom-a t.ep l"and step' 2.' AH max - 3.16 + 3.44 + 3.78 - 10.38 g
,
VH max - 12.310 + 1.108 + 6.020 - 19.438 in/sec .355
,+ .014 + .120 -' .489' 'in
and
:
2A-102 '
..
~
.
~. ~
1M 5-1300/NAVFAC P-397/AFR 88-22
AV max -3.44+7.56
- 11.00 "
-
g
Vv max - 1.108 + 6.020 - 7.128 ' in/sec ;
Dv
max
- .014 +
.120 - .134 in
Step 4.
Does not apply, the ground shock is not superseismic.
Step 5.
Magnitude of the motions for response spectra.
AH - 10.38 x 2.0 - 20.76 g, VH -~9.438 xl,S - 29.157 in/sec' DH -
,
.'489 x 1.0 - .489 in
and ~
AV - 11.00 it 2.0 - 22.00 g Vv - 7.128 x 1.5- 10.692 in/sec ,
DV - .134 x 1.0 Step 6.
- .134
in
,
See the shock response spectra for the values from step 6 in Figure: 2A-28 ..
i
•
--
2A-103
MAXIMUM VELOCITY, iftch•• per .econd .
8I
o
!='-
$ o
~/ X'X""'/,0X'
~'k'
)~,
i".
/
r-,
"(//V,0X,
'h1'> '\j<'(
~
/
'/.
~~1"(l,"'"
>'\X//I.//><",,- ",IX' .
1,,: ~!lo.. () v:v "',. Ix X>0X(,V -, 1/ ~./i'J.A, D{ i(Y A'ii"-~-+-i/h"i,f"-yh,,\r, """",~~i1iO"l'--+-i /'lI","--Plli'+ll''boI X :"-: ",:>:' o ,x'X",' ~'«: V ",x' IX, eX".: X / ' f" ", , >.;I,
iX,
" ' ,
I:
•
"TI
!:!
G'l
e
I:
I;
C ::0
'"
-c0
rn N
l> r N
CD
c
'"r": • '"I: z ... ",
.• ~
",
§-
TM 5-1300/NAVFAC P-397/AFR 88-22
APPENDIX 2B LIST OF SYMBOLS
TK 5-1300/NAVFAC P-397/AFR 88-22·
(1)
a
(2) (3)
ao
velocity of sourid in air (ft./sec.)
ax
accelerationinix direction (in./ms 2 )
ay
acceleration in y direction (in./ms 2)
A
(1)
(2)
area (in. 2 ) explosive composition factor (oz.1/2:i n. -3/2)
Ab
area of diagonal' bars at the support wi thin a width b (in. 2) 1 c area of reinforc,ing bar (in. 2 )
Ad
(1)
Aa
(2)
e
acceleration ;(in./ms 2) depth of equivalent rectangular stress block (in.) long span of a panel (in.)
door area (1n. 2 ) area of diagonal bars at the support within a width b (in. 2 )
An
drag area (in. 2 )
Af
net area of wall excluding openings (ft. 2)
Ag
area of gross section (in. 2 )
AH
maximum horizontkl acceleration of the ground surface (g's)
Al
area of longitudinal torsion reinforcement (in. 2 )
AL
lift area (in. 2 ) .
~
(1)
Ao
area of openings (ft. 2 )
~s
area of prestressed reinforcement ':(in. 2)
As
area of tension reinforcement within a width b (in. 2 )
As '
area of compress~on reinforcement within a width b (in. 2)
As -
area of rebound
AsH
area of flexural.reinforcement within a width b in the horizontal direction on each face (in. 2 )*
Asp
area of spiral reinforceme~t (in. 2)
As t
total area of reinforcing steel (in. 2 )
AsV
area of flexural -reinforcement within a width b in the vertical direction on each face (in. 2 ) *
*
(2)
net area of section (in. 2 ) area of individual wall subdivision (ft. 2)
reinforce~ent
See note at end of symbols
(in. 2 )
2B-l
TM 5-1300/NAVFAC P-397/AFR 88-22'
area of one leg of a closed tie resisting torsion within a distance s (in,2) total area of sttrrups or lacing rein~orcement in tension within a
distance, Ss or sl and a width b s or b l (in. 2)
maximum vertical acceleration of ·the ground surface (g's) area of wall (ft. 2) area of sector I and II, respectively (in. 2) (1) width of compr e s's Lon face of flexural member (in.)
(2) width of concrete strip in which the direct shear stresses at the supports are resisted by diagonal bars (in.) (3) short span of a panel (in.) bf
width of fragment (in.)
b
s
width" concrete strip in which the diagonal tension stresses are resisted by stirrups of area Ay (in.)
l
width of concrete strip ih which the diagonal tension stresses are resisted by lacing of area Ay (in.)
b
of
bo
failure perimeter for punching shear (in.)
b
center-to-center dimension of a closed rectangular tie along b (in.)
t
..
'
B
explosive constant defined in table 2-7 (oz.1/2 in. -7/6)
c
(1) (2) (3)
distance from the resultant applied load to the axis of rotation (in. ) damping coefficient width of column capital (in.)
distance from the resultant applied load to ·the axis of rotation for sectors I and II, respectively (in.) dilatational velocity of concrete' (ft./sec.) C
(1) (2)
C
deflection coefficient for the center of interior panel of flat slab
Ccr
critical damping
Cd
shear coefficient for ultimate shear stress of one-way.elements
Cn
drag coefficient
Cnq
drag pressure (psi)
c
shear coefficient deflection coefficient for flat slabs
2B-2
TM 5-l300/NAVFAC P-397/AFR 88 c22
e
GDqo
peak drag pressure (psi)
GE
equivalent load factor
Gf
post-failure fragment coefficient (lb. 2_ms4/in. 8)
GH
shear coefficient for ultimate shear stress in horizontal direction , * for two-way elements
GL
(1) (2)
leakage pressure coefficient from Figure 2-235 deflection coefficient for midpoint of long side of interior
(3)
slab panel lift coeffi9ient
>,'
flat
e
GM
maximum shear coefficient
Gm
equivalent"moment correction factor
Gp
compression wave seismic velocity in the soil from Table 2-10 (in./sec.)
Gr
sound velocity in, reflected region from Figure 2-192 (ft./ms)
GR
force coefficient for shear at the corners of a window frame
Gra
peak reflected pressure coefficient at angle of incidence a'
Gs
shear coefficient for ultimate support shear for one-way elements
GsH
shear coefficient for ultimate support shear in horizontal direction
I I
for two-way
*
elem~nts
."
Gs V
shear coefficient for ultimate support shear in vertical direction , * for, two-way elements
Cs
deflection coefficient for midpoint of short 'side of interior flat slab p'!nel
Gu
impulse coefficient at deflection
~
(psi-ms2/in.. 2)
Gu '
impulse coefficient at deflection
~
(psi-ms 2/in}) "-',
Gv
shear coefficient'l.for ultimate shear stress in vertical direction for two-way elements *
Gx
e
.I
shear coefficient for the ultimate shear'along the long side of window frame
Gy
shear coefficient'for the ultimate shear along the short side of window frame
GL
confidence level I
*
J
I
See note at end of symbols
2B-3
TK 5-l300/NAVFAC'P-397/AFR88-22
(1) (2)
impulse coefficient at deflection Xl (psi-ms 2/in. 2 ) ratio of gas load to shock load
Cl '
impulse coefficient at deflection Xm.(PSi-ms 2/in. 2 )
C2
ratio of gas load duration to shock load duration
d
(1)
d'
",
i
(2)
distance from extreme compression fiber to. centroid of tension reinforcement (in.) diameter (in.)
(3)'
fragmentdiameter.~in.)
distance from extreme compression fiber to centroid of'compression reinforcement (in.) . , . diameter of reinforcing bar (in.)
de
distance between the centroidsiof:the:compression and tension
reinforcement (in.) .t
dcH d co
"';f
i'
distance between the centroids of the horizontal compression and tension reinforcement (in.) ., , . diameter of steel core (Ln .') distance between the centroids of.the vertical compression and tension reinforcement (in,) , ~ distance f r cm auppo r t i.and equal..;to distance d or ,de (in.) average inside diameter of
exp~osive
casing (in.)
d' i
adjusted inside diameter of casing (in.)
dp
distance between center lines of adjacent lacing.bends measured
normal to flexural ,reinforcement· (in.) distance from extreme compression fiber to centroid of prestressed reinforcement (in.) + depth of spalled concrete (in.) diameter of cylindrical portion. of primary fragment (in.) D
( 1) (2)
(3)
(4) (5)
(6) (7)
unit flexural rigidity (lb-in.) location of. shock front for maximum stress (ft.) minimum magazine separation distance (ft.) caliber density (lb/in. 3 ) overall diameter of circular·section (in.) damping force (lb.) displacement of mass from shock load (in.) ,.
DE
..
equivalent loaded width of structure for non-planar wave front (ft.) 28-4
TM 5-1300/NAVFAC P,-397/AFR 88-22
maximum horizontal displacement of the ground surface (in,) dynamic increase factor
diameter of the circle through centers of reinforcement arranged in a circular pattern (in.) diameter of the spiral measured through the centerline of the spiral " bar (in.) I
DLF
dynamic load factor maximum vertical displacement of the ground surface (in.) (1) (2)
e
(3)
base of natural logarithms and equal to 2.71828 ... distance from centroid of section to centroid of prestressed reinforcement (in,) actual eccentricity of load (in.) ecce~tricity (Ln . )
eb
balanced
2E,l/2
Gurney Energy\Constant (ft./sec.)
E
(1) (2)
Ec
modulus of elasticity of concrete (psi)
Em
modulus of elasticity of masonry units (psi)
Es
modulus of elasticity of reinforcement (psi)
f
(1) (2)
fc'
static ultimate compressive strength of concrete at 28 days (psi)
".
,
,
modulus of elasticity internallwork (in.-lbs.)
unit external force (psi) frequency of vibration (cps) j
dynamic ultimate compressive strength of concrete (psi) f
dm
•
dynamic ultimate compressive strength of masonry units (psi)
f ds
dynamic desig~'stress for reinforcement (a function of f y• f u and 8) (psi)
f du
dynamic ultimate stress of reinforcement (psi)
f dy
dynamic yield stress of reinforcement (psi)
f m'
static,ultimate compressive strength of masonry units (psi) . . .I· . ... . -
, natural
,,
freque~cy
of
vibr~tion
(cps)
average stress' in the prestressed ,reinforcement at ultimate load
2B-5
TM 5-l300/NAVFAC P-397/AFR 88-22
specified tensile strength of prestressing tendon (psi)
yield stress of prestressing tendon corresponding to a1 percent elongation (psi) reflection factor static design stress for reinforcement
~psi)
effective stress in prestressed reinforcement after allowances for prestress losses (psi) f
u
f
y
F
static ultimate stress of reinforcement (psi)
static yield stress of reinforcement (psi) (1)
(2) (3)
total external force (lbs.) coefficient for moment of inertia of cracked section function of C 2 and Cl for bilinear triangular load
Fo
force in the reinforcing bars (lbs.)
FE
equivalent external force (lbs.)
FD
drag force (lbs.)
FF
frictional force (lbs.)
FL
lift force (lbs.)
FN
vertical load supported by foundation (lbs.)
g
acceleration due to gravity (32.2 it. /sec. 2).
G
shear modulus (psi)
h
(1) charge location parameter (ft.')
(2) height of masonry wall
hn
average clearing distance for individual areas of openings from Section 2-15.4.2
ht
center-to-center dimension of a closed rectangular tie along h (in.)
h'
clear height between floor slab and roof slab
H
(1) (2) (3)
Hc
span height (in.)* distance between reflecting'sur~ace(s) and/or free edge(s) in vertical direction (ft.) minimum transverse dimension of mean presented area of object (ft. )
height of charge above ground (ft.)
* See note at end of symbols
28-6
TM 5-l300/NAVFAC P-397/AFR 88-22
Hs
height of structure (ft.) height of triple'point (ft.) height of wall (ft.) heat of combustion (ft.-lb./lb.) heat of detonation (ft.-lb./lb.)
i
unit positive impulse (psi-ms) sum of blast impulse capacity of the receiver panel and the least impulse absorbed· by the. sand (psi-ms) blast impulse capacity of receiver panel (psi-ms)
i
unit negative impulse (psi-ms) sum of scaled unit blast impulse capacity of receiver panel and scaled unit blast impulse attenuated through concrete and sand in a composite element (psi-msl,lb. l/ 3 )
unit blast impulse.(psi-ms) • scaled unit b l.as t impulse (ps Lvmsy Ib -, l / 3
i.
scaled. unit blast. impulse capacity of receiver' panel of composite element (psi'ms/lb. l/ 3 ) i bd element
scaled unit blast impulse capacity of donor panel of composite ... .;. ,,' '.
(psi-ms/lb. lI;3)
total scaled!unit blast impulse capacity" of' composite element (psi-ms/lb. l/,3) impulse id
capa~ity
of an element (psi-ms)
total.drag.and'diffraction· impulse (psiCms) ,,'
unit excess blast impulse (psi-ms) required
imp~lse'~ capaci ty
of fragment shield' (p's Loms) .
gas'. impulse (psi-ms) , unit positive normal reflected impulse (psi-ms) unit negative normal reflected impulse' (psi-ms) ira
peak reflected impulse at angle of incidence a (psi-ms)
is
unit positive incident impulse (psi-ms) 2B-7
I.
TM 5-l300/NAVFAC P-397/AFR 88-22
is·
unit negative incident impulse (psi-ms)
i
impulse consumed by fragment support connection (psi-ms)
st
(1) moment of inertia (in. 4/in. for slabs) (in. 4 for beams) (2) total impulse applied to fragment
I
average of gross and cracked moments of inertia (in. 4 for beams)
(in.4/in~
for slabs)
moment of inertia of cracked concrete section (in. 4/in. for slabs') (in. 4 for beams) moment of inertia of cracked concrete section in horizontal direction
(in. 4/in.)*
moment of inertia of direction (in. 4/in.)* moment of inertia of (in. 4 for beams)
cracked concrete section in vertical'
gross
concrete section
. 4/.1.0. for slabs) ( 1.0.
mass moment of inertia (lb.-ms 2-in.)
t'.
In
moment of inertia of net section of masonry unit (tn. 4)
Is
gross moment of inertia'of slab (in. 4/in.)
1s t
impulse consumed by the fragment support connection (psi-ms)
Iw
gross moment of inertia of wall (in. 4/in.)
j
ratio of distance between centroids of compression and tension forces to the depth d
k
(1) constant depending on the. casing metal (2) effective length factor
kv
velocity decay coefficient
K
(1)
(2) Ke
KE
KL
*
unit stiffness (psi/in. for slabs) (lb./in./in. for beams) (lb./in. for springs) constant defined ,in paragraph 2-18.2
elastic unit stiffness
,(psi/in.
for slabs)
(lb./in./in. for beams)
elasto-plastic unit stiffness (psi/in. for slabs) (lb./in./in. for beams) , ' (1) equivalent elastic unit stiffness (psi/in. for slabs) (lb./in./ in. for beams) (2) equivalent spring constant (lb./in.) load factor
See note at end of symbols
28-8
TN 5-l300/NAVFAC P-397/AFR 88-22 I" load-mass. factor
load-mass
fac~or
in the. ultimate range
(KLM)up
load-mass factor in the post-ultimate range
KM
mass factor
KR
resistance factor
KE
kinetic energy
1
charge location parameter (ft.)
1
(1) (2)
Id
basic
Idh
development lepgth of hooked bar (in.)
lc
length of cylindrical explosive (in.)
Ip
spacing of same type of lacing bar (in.)
Is
span of flat slab panel (in.)
L
(1) (2)
-,F·
,
length of, the yield line (in.) width of 1/2 of the column strip (in.)
..
'
devel~pnrent length of reinforcing bar (in'.) Ii
span length (in.)* distance between reflecting surface(s) and/or free edge(s) in horizontal direction (ft.)
length of cylivder (in.) length of fragment (in.) ,,
.
-
clear span in short direction (in.) length of Lac Lng bar requried in distance s1. (in.)" clear span in long direction (in.) embedment length of reinforcing bars (in.) unsupported
le~gth
of column (in.)
wave length of '":,positive pressure phase (ft.) wave length of .ne ga t Ive pressure phase (ft . .): clear span in I'ong direction (in.) clear span in short direction (in.) '
'J'
*
See note at end of symbols
I
2B-9
....
TM 5-l300/NAVFAG P-397/AFR'88-22 Lwb' Lwd wave length of positive pressure phase' at points b-and d,_ respectively (ft.) :
.'
..
L l
total length of sector of element normal to axis of rotation (in.)
m
(1) unit mass
(psi-ms2/i~. 'fo~
slabs) [beams, (2) ultimate unit moment (in. -lbs./in.) (3) mass of fragment (lbs.-ms 2/in.)
(lb.,iin-ms2)/i~.]
"
."
average of the effective elastic and plastic unit masses (psi-ms 2/in. for slabs) [beams, (lb./in-ms 2)/in] , effective unit mass' (psi-ms 2/in. ms p
for 'slabs) [beams, (lb/in-ms2.)/in]
mass of spalled fragments (psi-ms 2/in.) effective unit mass in the ultimate range [beams, (lb/in-ms 21/in.j . .'
~p M
(psi-~s2/in. ..
'.
for slabs) .,
effective unit mass in the post-ultimate'range (psi-ms 2/in.) (1)
unit bending moment (in. -lbs./in. for slabs)' (in. -lbs ..for
(2) (3)
total mass (lb.-ms 2/in.) design moment (in.-lbs.)
beams)
effective totai mass (lb.-ms 2/in.)
Me ~
u l t Lma t e unit :resisting 'moment (in ..-Lbs ./in. for' slabs) (in. -lbs. for beams)
M u-
ultimate unit rebound moment (in. -lbs./in. for slabs) (in. -Tbs ,' for beams) '.
Mc
moment of concentrated loads about line of rotation of sector (in.-lbs.)
MA
fragment distribution factor
ME
equivalent total mass (lb. -ms 2/in.') '..
M HN
ultimate ~n~t ~egative moment capacity in horizontal, directioTI,1 (in. -lbs'./in.)
MHP
u:timate u~it ~ositive moment capacity in horizontal direction (In.-lbs.jln.) .... : ' I '.-
MOH,MOL
total panel moment for direction Wand L respectively (in. -lbs.)
MN
ultimate unit negative moment capacity at supports (iri.lbs./in. for slabs) (in.-lbs. for beams)
Mp
ultimate unit positive moment capacity at midspan (in.-lbs./in. for slabs) (in.-lbs. for beams)
*
See note at end of symbols
2B-10·
.,
TK 5-l300/NAVFAC P-397/AFR 88-22
MVN
ultimate unit ~egative moment capacity in vertical direction (in.-lbs./in.)
MVp
ultimate unit ~ositive moment capacity in vertical direction (in.-lbs./in.) ,
M1
value of srnalle~ end moment on column
M2
value of larger, end moment on column
n
(1) (2) (3) (4)
modular ratio number of time intervals number of glass pane tests caliber radius of the tangent ogive of fragment nose
N
(1) (2)
number of adjacent reflecting surfaces nose shape factor
number of primary fragments larger than Wf axial load normal to the cross section total number of: fragments p
reinforcement ratio equal to Ag/bd or As/bdc
p'
reinforcement ratio equal to As'/bd ·or As·/bd c·
Ph
reinforcement ratio producing balanced conditions at ·ultimate
strength ambient atmospheric pressure (psi) prestressed
rei~forcement
ratio equal to
~s/bdp
Pm
mean pressure in a partially vented chamber (psi)
Pmo
peak mean pressure in a partially vented chamber (psi) average peak reflected pressure (psi) reinforcement ratio in horizontal direction on each face *
Pr
total reinforcement ratio equal to PH + Pv
Pv
reinforcement r~tio in vertical direction on each face*
p(x)
distributed load per unit length
p
(1) pressure (p~i) (2) concentrated load (lbs.) n~gative
pressure (psi)
critical axial load causing buckling (lbs.)
*
See note at end of symbols
28-11·
1M 5-1300/NAVFAC P-397/AFR 88-22
Pg
maximum gas pressure (psi)'
Pi
interior pressure within structure (psi)
oP
interior pressure increment (psi)
i
fictitious peak pressure (psi) .
"-;-
maximum average pressure acting on interior face. of wall (psi) Po
(1) (2) (3)
peak pressure (psi) maximum axial load (lbs.) atmospheric pressure (psi)
"
peak positive normal reflected pressure (psi) P r
peak negative normal reflected pressure (psi) peak reflected pressure at angle of incidence a (psi) maximum average pressure on backwall (psi) positive incident pressure (psi)
P
s b'
Ps e positive incident pressure at points band e, respectively (psi)
Ps o
peak positive incident pressure (psi)
Pso -
peak negative incident pressure .. (psi)
Psob,Psod' P peak positive incident pressure at points .b, do and e, respectively s oe (psi) P
u
ultimate axial load at actual eccentricity c (lbs.)
P x
ultimate load when eccentricity ex is present (lbs.)
P y
ultimate load when eccentricity e y is present (lbs.)
q
dynamic pressure (psi)
qb,qe
dynamic pressure at points band e, respectively (psi)
qo
peak dynamic pressure (psi)
qob' qoe
peak dynamic pressure at points band e " respectively (psi)
r
(1)
(2)
unit resistance (psi) radius of spherical TNT [density equals
(3)
radius of gyration of cross section of column (in.)
95 Ib./ft. 3 ] charge
(ft. )
r
unit rebound resistance (psi, for slabs) (lb./in. for beams)
28-12
•
TN 5-l300/NAVFAC P-397/AFR 88-22
dynamic resistance available (psi)
or
change in unit resistance (psi, for slabs) (lb./in. for beams) radius from center of impulse load to center of door rotation (in.) uniform dead load (psi) elastic unit resistance (psi, for slabs) (lb./in. for beams)
rep
elasto-plastic unit resistance (psi, for slabs)
(lb./in. for beams)
rfs
ultimate unit resistance of fragment shield (psi)
rr
tension membrane resistance (psi)
ru
ultimate unit'resistance (psi, for slabs) (lb./in. for beams) post-ultimate 'unit resistance (psi) I
radius of hemIspherical portion of primary fragment ·(in.) R
(1)
total internal resistance (lbs.)
(2)
slant distance (ft.) ratio of [SIC , standoff distance (ft.)
(3)
(4)
effective radius (ft.) (1) (2)
distance ~raveled by primary fragment (ft.) distance from center of detonation (ft.)
,
Rg
uplift force at corners of window frame (lbs.)
Rl
radius of lacing bend (in.)
Rt
target radius (ft.)
RA
normal distance (ft.)
RE
equivalent total internal resistance (lbs.)
RC
ground distance (ft.)
, I
total ultimate 'resistance (lb.)
,
total internal resistance of sectors I and'II, respectively (lbs.) s
(1)
(2) (3)
sample standard deviation spacing o~ torsion reinforcement in a direction parallel to the , longitudinal reinforcement (in.) pitch of spiral (in.)
.
28-13
TH 5-1300/NAVFAC P-397/AFR 88-22 .
spacing of stirrups in the direction parallel to the longitudinal reinforcement (in.) spacing of lacing in the direction parallel to the'longitudinal reinforcement (in.) height of front wall or one-half its width, whichever is smaller
S
(ft. ) S'
weighted average clearing distance with openings (ft.)
SE
strain energy
t
time (ms)
cSt
time increment (ms) any time (ms)
tb
I
t
e
I
tf time of arrival of blast wave at points b, e, and f, respectively
(ms)
,
tc
(1)
t c'
(1)
td
rise time (ms)
tE
time to reach maximum elastic. deflection (ms)
tg
fictitious gas duration (ms)
tm
time at which maximum deflection occurs (ms)
to
duration of positive phase
to
duration of negative phase of blast pressure (ms)
tof
fictitious positive phase pressure duration (ms)
tof
fictitious negative phase pressure duration (ms)
tr
fictitious reflected pressure duration (ms)
tu
time at which ultimate deflection occurs (ms)
ty
time to reach yield (ms)
tA
time of arrival of blast wave (ms).
tAG
time of arrival of ground shock (ms)
(2)
(2)
clearing time for reflected pressures (ms) average casing thickness of explosive charges (in.) adjusted casing thickness (in.) Clearing time for reflected pressures adjusted for wall openings (ms)
ofblastpressur~
2B-14 .
(ills)
TM
5:~300/NAVFAC
P"397/AFR,88-22
time at which partial failure occurs (ms). T
(1) duration of equivalent triangul~r loading (2) thickness of masonry wall (in.) (3) toughness of material (psi-in./in.)
thickness of glass (in.)
func~ion
(ms)
"
force in the continuous,reinforcement in the short span direction (lbs.) angular impulse load (lb.-ms-in.) force in the continuous reinforcement in the long span direction
(lbs.) effective natural period of vibration (ms) minimum thickness of concrete to prevent perforation by a given
fragment (in.) rise time (ms) (1) thickness of sand fill (in.) (2) thickness of slab (in.) Ts p
minimum concrete thickness to prevent spalling· (in.)
Ts
scaled thickness of sand fill (ft./lb. l/ 3)
Tu
total torsional moment at critical section (in.-lbs.)
Tw
thickness of wall (in.)
Ty
force of the continuous reinforcement in the short direction (lbs.)
u
particle velocity (ft./ms)
~
ultimate flexural or anchorage bond stress (psi)
U
shock front velocity (ft./ms)
Us
strain energy
v
velocity (in./ms)
va
instantaneous velocity at any time (in./ms)
vb
boundary velocity for primary fragments (ft./sec;)
"
Ii ...
,
. ,
, r
2B-15
.
r.
TM 5-l300jNAVFAC P-397jAFR 88_22
V
c
vf
ultimate shear stress permitted on an unreinforced web '(psi)"
maximum post'failure fragment velocity (in.jms)
vf(avg.) average post-failure fragment velocity (in.jms) velocity at incipient failure deflection (in.jms) initial velocity of primary fragment' (ft.jsec.) residual velocity of primary fragmenL after perfora~i.on (ft./sec.) striking velocity of primary fragment (ft:jsec.) maximum torsion capacity of an unreinforced web- (psi)
nominal torsion stress in the direction of
V
u (psi)
ultimate shear stress (psi) ultimate shear stress at distance de from the ho r Lz ou t.a l. support
(psi)* ultimate shear stress at distance de from the vertical support (psi)* velocity in x direction (in.jms.) velocity in y direction (in.jms.)
;-
v
(1) volume-of partially vented chamber (ft. 3) (2) velocity of compression wave through concrete (in.jsec.) (3) velocity of mass under shock load (in.jsec.)
Vd
ultimate direct shear capacity of the concrete of
VdH
shear at distance de from the vertical support on a unit width (lbs.jin.)*·
VdV
shear at distance de from the horizontal support on a unit width (lbs.jin. )*
Vf
free volume (ft. 3)
VH
maximum horizontal velocity of the ground surface (in.jsec.)
Vo
volume of structure (ft. 3)
Vs
shear at the support (lb.jin., for panels) (lbs. for beam)
Vs H
shear at the vertical support on a unit width (lbs./in.)*
VSV
shear at the horizontal support on a unit width (lbs.jin.)*
Vu
total shear on a width b (lbs.)
~idth
b (lbs.)
.,
,
* See note at end of symbols
28-16
TM 5-l300/NAVFAC P-397/AFR 88-22
maximum vert cal velocity of the ground surface (in./sec.) unit shear a ong the long side of window frame (lb./in.) unit shear along the short side of window frame, (lbs./in.) w
applied uniform load (lbs. -in. 2 ) (1) unit weight (psi, for panels) (lb./in. for beam) (2) weight density of concrete (lbs ./ft. 3) weight density of sand (lbs./ft. 3)
W
(1) design charge weight (lbs.) (2) external ~ork (in.-lbs.) (3) width of wall (ft.) weight of fluid (lbs.) actual quantity of explosives (lbs.) total weight of explosive containers (lbs.) effective charge weight (lbs.) effective charge weight for gas pressure (lb.) weight of explosive in question (lbs.)
Wf
,.
weight of primary fragment (oz.)
,
average fragment weight (oz.) ".
WF
'weight of frangible element (lb./ft. 2)
WCI
weight of inner casing (lbs.)
Wc o
total weight of steel core (lbs.)
Wco
weight of outer casing (lbs.)
WeI' Wc 2 total weight of plates 1 and 2, respectively (lbs.) Ws
width of structure (ft.) work
dOIlt
x
yield line lo~ation in horizontal direction (in,)*
x
(1) (2)
I
deflect.io'n (in.) distance ~rom front of object to location of largest cross section t? plane of shock front (ft.)
any deflection: (in.) I * See note at end of symbols
28-17
TM 5-1300/NAVFAC P-397/AFR.88-22 lateral deflection to which. a masonry wall develops no resistance (in. ) deflection due to dead load (in.) elastic deflection (in.) equivalent elastic deflection (in.) elasto-plastic deflection. (in.) maximum penetration into concrete of armo r.cp Le rc Lng fragments (in.) maximum penetration into concrete of. fragments other than armor-piercing (in.) maximum transient deflection (in.) plastic deflection (in.) Xs
(1) (2)
~
ultimate deflection (in.) (1) (2)
maximum penetration into sand of armor-piercing fragments (in.) static deflection (in.)'
partial failure deflection (in.) deflection at maximum ultimate resistance of masonry wall (in.)
.
' . -* ."
yield line location in vertical direction (in.)
y
distance from the top of section to centroid (in. ) scaled slant distance (ft./lb. l/ 3)
z
scaled normal distance (ft./lb. l/3) scaled ground distance (ft./lb. l/ 3) Q'
(1) (2) (3) (4)
Q
ec
angle formed by the plane of stirrups, lacing, or diagonal reinforcement and the plane of the longitudinal reinforcement (deg) angle of incidence of ·the pressure front (deg) acceptance coefficient trajectory angle (deg.)
ratio of flexural stiffness of exterior wall to flat slab
Q e cH ' Q e cL
..
ratio of flexural stiffness of exterior wall to slab in direc-tion H
and L respectively
;-
"
'.
'.
*
See note at end of symbols
25-18
,.
TM 5-l300/NAVFAC P-397/AFR 88-22
(1)
B
coefficient
for
determining
elastic
and -_elasto -plastic
resist~nces
(2) (3) (4)
particJlar support rotation angle (deg) , reject~on coefficient target .shape factor from Figure 2-212
factor equal to 0.85 for concrete strengths ~p to 4,000 psi and is reduced by 0.05' for each 1,000 psi in excess of 4,000 psi coefficient for determining elastic and elasto-plastic deflections • factor for type of prestressing tendon
y
moment magnifier ;
clearing factor deflection at sector's displacement (in.) €
c
•
average strain rate for concrete (in./in./ms) unit strain in'mortar (in./in.) I
€
s
•
average strain rate for reinforcement (in. lin,. /ms) rupture strain (in./in./ms)
8
(1) support rotation angle (deg) (2) angular acceleration (rad/ms 2 ) maximum support rotation angle (deg) horizontal rotation angle (deg)* vertical rotation angle (deg) *
(1) ductility factor (2) coefficient of friction v
Poisson's "ratio
P
(1) mass density (lbs.-ms. 2/in. 4 ) (2) density of air behind shock front (lbs/ft. 3 )
Pa
density of air (oz./in. 3 )
Pc
density of casing (oz./in. 3 )
Pf
mass density of fragment (oz./in. 3 )
Po
mass density of medium (lb.-ms. 2/in. 4)
au
fracture strength of concrete, (psi)
*
See note at end of symbols
28-19-
TK 5-1300/NAVFAC P-397/AFR 88-22 ' effective perimeter of reinforcing bars (in.)
~o
summation of moments (in.-lbs.) sum of the ultimate unit resisting yield lines (in.-lbs.)
-momerit s
acting along..the negative
sum of the ultimate unit resisting moments acting along the positive yield lines (in.-lbs.) (1) (2) (3)
capacity reduction factor bar diameter (in.) TNT conversion factor
¢r
assumed shape function for concentrated loads
¢(x)
assumed shape function for distributed loads free edge simple support fixed support
xxxxx
either fixed, restrained, or simple support
.: -.
*
Note.
walls.
This symbol was developed for two-way elements which are used as When. roof slabs or other horizontal elements are under consideration,
this symbol will also be applicable if the element is treated as being rotated into a vertical position.
2B-20
TK 5-1300/NAVFAC P-397/AFR 88-22
APPENDIX 2C - BIBLIOGRAPHY
TM 5-l300/NAVFAC P-391/AFR 88-22
1.
2. 3.
4.
Blast Loads Armendt, B. R., HipP'ensteel, R. G., Hoffman',' ,A. J., and Keefer, J. H., Pro; ect White Tribe:~ Air Blast From :Simulta;"e~usly Detonated Large'Scale Explosive Charges, BRL Report 1145, Aberdeen Proving Ground, " Maryland, September 1961. Armendt, B. R., Hippensteel, R. G... Hoffman, A. J., and Kingery, C. N., The' Air Blast From Simultaneously Detonated Explosive Spheres, BRL Report 1294, Aberdeen Proving Ground, Maryland, August 1960. Armendt, B. R., Hippensteel, R. G., Hoffman, A. J., and Schlueter, S. D. , The Air Blast From Simultaneously Detonated Explosive Spheres: Part II - Optimization, BRL Memorandum Report 1384, Department of the Army Project No. 503-04-002, Aberdeen Proving Ground, Maryland, January 1962. Ayvazyan, H.E.
t'
.Dobb s- N., Computer Program.Impres, special publication
ARLCD-SP-8400l, prepared by Ammann and Whitney, New York, New York, for U.S. Army Armament:Research and Development Center, Large Caliber 5. .6.
.7. 8.
e
9. 10.
11.
12. 13. 14 . 15. 16. 17.
e
18.
Weapon Systems Laboratory, Dover, New Jersey. Baker, W. E., Explosions in Air, University of Texas Press, Austin,
Texas 1973. Baker, W. E., Prediction and Scaling of Reflected Impulse From Strong Blast Waves, International·Journal of Mechanical Science, 9, 1967, pp. 45-51. Bleakney, W., The Diffraction of Shock Waves Around Obstacles and the Transient Loading of Structures, Princeton'University, March 1950, (Published and Distributed .by the Armed Forces Special Weapons Pr~ject). Coulter, G.A., Air Shock and Flow in Model Rooms, BRL Memorandum.Report 1987, Aberdeen Proving Ground, Maryland, June 1969. Coulter, G. A., Air Shock Filling of Model Rooms, BRL Memorandum Report 1916, Aberdeen Proving Ground, Maryland, March 1968. Coulter, G. A. and Peterson, R.L., Blast Fill Time of a One-Room ! Structure , Operation .Prairie Flat Project Officers Report;. Project LN 111 POR-2l02, Ballistic Research Laboratory, Aberdeen Proving Ground, Maryland, .November'1964. Coulter, G. A., Bulmash G., and Kingery C., Feasibility Study of Shock Wave Modification in the BRL 2.44 m Blast Simulator, Memorandum Report ARBRL-MR-0339,'prepared by U.S. Army Armament Research and Development Center, Ballistic Research Laboratory, Aberdeen Proving Ground, Maryland, March 1984. . . Cranz, C., Lehrbuchder Ballistik,Springer-Verlag, Berlin, 1926. Dewey, J.M., Johns 01') , O.T., and Patterson,.. J.D. II, Mechanical Impulse Measurements Close to Explosive Charges, BRL Report 1182, Aberdeen Proving Ground, Maryland, November 1962. . Dobratz, B.M., LLNL Explosives Handbook, Properties. of Chemical Explosives and ,Explosive Simulants, URCL-52997, Lawrence Livermore National Laboratory;. Li.ve rmor'e , California, March 1981. Engineering Design Handbook,. Principles of Explosive Behavior, AMCP 706180; Headquarters, U.S. Army Material Command, Washington, D.C., April 1972. , Goodman, H. ' j . , Compiled Free Air Blast Data on Bar Spherical Pentolite, BRL Report 1092, Aberdeen 'Proving Ground, Maryland, 1960. Granstorm, S. A., Loading Characteristics of Air ~lasts From Detonating Charges, Transaction~ of the Royal Institute of Techriology, Stockholm, Sweden, Nr .100, 1956;. Hokanson, J.q. ,Esparza, E. D., and Wenzel, A. B. ,I Blast Effects of Simultaneous Multiple-Charge Detonations, Contractor Report ARLCD-CR78032, ARRADCOM, Dover, New Jersey, AD E400232,October 1978. 2C-l
TM 5-t300/NAVFAC 19.
P-397/AF~
88-22
Hokanson, J.C., Esparza, E.D., Baker W.E., Sandoval, N.R. and Anderson, C.E., Determination of Blast Loads in the Damaged Weapons Facility, Vol 1, -Fd na L Report, Purchase Order ,F09-l3400, SwRl- 65 78" pr ep a r ediby Southwest Research Institute, San Antonio, Texas, for,Mason and Hanger
20. 21.
22. 23.
24. 25. 26. 27. 28.
29.
30.
Silas Mason Company, Inc" Pant ex Plant, Amarillo, Texas, July 1982. Hokanson, J.' C., Esparza, E. : D" and Wenzel, A. B., Renec ted BIas t Measurements Around Multiple Detonating Charges,Minutes of the Eighteenth Explosives Safety Seminar, Volume I. September 1978, pp.' 447-471. Hokanson, J.. C, ,Esparza, E. D", Wenzel, A, ,B. (Southwest Research Institute). Price, P. D, (ARRADCOM). Blast ·Effects of Simultaneous Multiple-Charge Detonations, Contractor Report ARLCD-CR-78032, prepared by U.S. Army Armament Research and Development Command, ,Large. Caliber Weapon Systems, Laboratory,. Dover; New Jersey, October 1978. Hopkinson,' B. British Ordnance Board MinutBs, 13565, 1915. Iwanski, E. C., et al., Blast Effects on Buildings and Structures Operation of 6-Foot and 2-Foot Shock Tubes: High Pressure Tests on Simple Shapes, Report No. 54, Final Test Report No. 10, ARF Project No. 0087, Armour Research Foundation of Illinois Institute of Technology. Jack,W. H., Jr., Measurements of Normally Reflected Shock Waves From Explosive Charges, BRL Memorandum Report 1499, Aberdeen'Proving Ground, Maryland, AD 422886, July 1963. Jack, W. H., Jr. and Armendt, B. F., Jr" Measurements of Normally Reflected Shock Parameters Under Simulated High Altitude Conditions, BRL Report 1280, Aberdeen Proving .Ground, Maryland, AD 469014, April 1965. Johansson, C. H. and Persson, P.A., .De t on i c s of High Explosives, Academic Press, London, England, an~ New York, New York, 1970. Johnson, O.T., Patterson, J.D. II, 'and 01s6n, W.C., A Simple Mechanical Method for Measuring the Reflected Impulse of Air Blast Waves, BRL Report 1099, Aberdeen Proving Ground, Maryland, July 1957. Kaplan, 'K., Lewis, K. S., and Morris, P. J:, Blast Loading' and Response of Military Eouipment, Draft Final Report URS 7339-6, URS Research Company, San Mateo, California, Prepared for Ballistic Research Laboraco ry , December 1973. Kaplan, K. and Price, P.D. s • Accidental Explosions and Effects of Blast Leakage Into Structures', Contractor Report ARLCD-CR- 79009, U. S. Army ARRADCOM, Dover, New Jersey, AD E400320, June 1979. Keenan, W.
A., and Tancreto, J.
E .. Blast Environment From Fully and
Partially Vented Explosions in Cubicles, Technical Report R 828, prepared by Civil Engineering Laboratory, Naval Construction Battalion
31.
Center, Port Hueneme, Galifornia,. sponsored by Department of the Army ,Picatinny Arsenal, November 1975. Keenan, W., Tancreto, J., Meyers, G., Johnson, F., Hopkins, J., Nickerson, H'
32.
33.
t
arid Armstrong, .W., NCEL Products SUDDorting DOD Revision
of NAVFAC P-397. Program No. Y0995-0l-003-20l, Technical Memorandum 2591TM, sponsored by Naval Facilities Engineering Command, Alexandria, Virginia, and Naval Civil Engineering Laboratory, Port Hueneme,California, March 1983. Kingery,. C.N., and Bulmash G., Airblast Parameters from TNT Spherical Air Burst and Hemispherical Surface Burst, Technical Report ARBRL-TR02555, prepared by U. S. 'Army Armament Research and Development Center, Ballistic Research Laboratory, .Aberdeen Proving Ground, Maryland, April 1984. . ' Kingery, C. N. and Coulter G.A., Reflected Overpressure Impulse on a' Finite Structure, Technical Report ARBRL-TR-02537. prepared by U. S. 2C-2
TM 5-1300/NAVFAC P-397/AFR 88-22 Army Armament Res~arch and Development Center, Ballistic Research Laboratory, Aberd~en 'Proving Cround Maryland, December 1983.,
34.
35.
36. 37, 38.
Kingery, C. N. and Coulter G. A.; Enhanced Blast as a Function of Multiple Detonatidns and Shape for Bare Pentolite Charges, Memorandum Report BRL-MR-353~, prepared by U.S. Army Ballistic Research· Laboratory, Aberdeen Proving Cround, Maryland, July 1986. Kingery, C. N., B~lmash G., and Muller P., Blast Loading on Above Ground Barricaded Munition Storage·Magazines, Technical Report ARBL-TR-Q2557, prepared by U.S. ~rmy Armament Research and' Deve~opment Center, Ballistic Research Laboi'atory, Aberdeen Proving Ground,Maryland, May 1984. Kriebel, A.R., Aip Blast in Tunnels and Chambers, Final Report DASA 1200-11, Suppleme~t I, Prepared for Defense Civil 'Preparedn~ss Agency, URS Research Company, 'San Mateo, California, September 1972, Makino, R. The Kirkwood-Brinkley Theory of the Propagation of Spherical Shock Waves and Its Comparison With Experiment, BRL Report 750, Aberdeen Proving Ground, Maryland, April 1951. • McIntyre, F. L., iNTEguivalency Test Results of Selected High Explosives, Propellants~, and Pyrotechnics ,in Surface Burst Configurations, Contractor Report,~ Prepared by Technical Services Laboratory , National
Space Technology Ijaboratories , NSTL MS 39529, for U.,S'.,
Army Research
and Development Engineering Center, Dover, New ,Jersey·, Sponsored by U.S.
Army - AMCCOM, July 1986. Melichar, J. F", :The Propagation of. Blast Waves Into Chambers, BRL Memorandum Report i920, Aberdeen Proving Ground, Maryland, March 1968. 40. Olson, W. C., Patt~rson, J. D. II, and Williams, J. S., The Effect of Atmospheric Pressure on the Reflected Impulse From Blast Waves, BRL Memorandum Report ~241, Aberdeen Proving Ground, Maryland, January 1960. 41. Reeves H., arid Rob'Lnson W. T., Hastings Igloo Hazards Tests for Small Explosive Charges,lMemora~dum Report ARBL-MRo03356, prepared by U. S. Army Armament·Research and Development Center, Ballistic Research' Laboratory, Aberdeen Proving Ground, Maryland, May 1984. 42. Reisler, R. E., K~nnedy.,. L. W., and Keefer; J. H., High Explosive Multi-Burst Air Blast Phenomena (Simultaneous and Non-Simuttaneous Detonations), BRL Technical Report ARBRL-TR-02142, Aberdeen Proving Ground, Maryland, February 1979. 43 . . Sachs, R. G., The' Dependence of Blast on Ambient Pressur~ and Temperature, BRL Report 4~6, Aberdeen Proving Ground, Maryland, 1944. 44. Shear, R. E., Incident and Reflected Blast Pressures for Pentolite, BRL Report 1262, Aberd~en Prov~ng Ground, Maryland, September 1964. 45. Shear, R. E. and McCane, P., Normally Reflected Shock Front Parameters, BRL Memorandum Rep?rt 1273, Aberdeen Proving Ground, Maryland, May 1960. 46. Swisdak, M.M., Jr., Explosion Effects and Properties: Part I - Explosion Effects in Air, NSWC/WOL/TR 75-116, Naval Surface Weapons Center, White Oak, Silver ~pring, Maryland, October 1975. 47. The Effects of Nuclear Weapons, S. Glastone, Editor, ·U. S. Atomic Energy Commission, Washington; D."C.; (Revised Edition), 1964. 48. Voltz, R. 'D., and ~iger; S.. A", An Evaluation of the Separated' Bay Concept for a Munition Assembly Complex: An Experimental Investigation of the Department of Energy Building 12-64 Complex, Technical Report SL83-6, prepared by Structures Laboratory; .U,S. Army Engineer Waterways Experiment Station! Vicksburg, Mississippi, for Department of Energy, Albuquerque Operattons, Amarillo, Texas, September 1983. 49. Ward, J. M. (DOD E~plosives Safety Board); Swisdak, M; M., Jr., Peckham P. J., and Soper '1. G. (NSWC), Lorenz, R. A. . (Boeing Military Air-' craft Company), Modeling of Debris and Airblast'Effects from Explosions 39.
2C-3.
TM 5-1300/NAVFAC P~397/AFR
50.
51.
52,
53.
54,
55,
88-22·
inside Scaled Hardened Aircraft Shelters, Final Repor.t NSWC TR 85-.470, prepared by Naval Surface Weapons, Center, Silver, Spring, Maryland, May 1985. _ Warnecke, ,C. Data Report: Support Test·Evaluation of a Pre "Engineered Building, TECOM Project No. 2-CO,PIC-004, U.S. Army Dugway Proving Ground, Dugway, Utah, June 1977. Wenz~l,
A. B'. and Esp ar z a , E,. D.. Measurements of Pressures and ,_Impulses
at Close Distances From Explosive Charges Buried and in Air, Final Report on Contract No., DAAK02-71-C-0393 with U,S. Army MERDC,.Fort Belvoir," Virginia, 1972. Wilton, C, and Gabrielsen, B, L., Shock Tunnel Tests of Pre loaded ,and Arched Wall Panels, Final.Report URS 7030~ 10" prepared for Defense Civil Preparedness Agency, URS Research Company; San Mateo,'California, June 1973. Wilton, C: and· Gabrielsen, B., ·L., Shock Tunnel Tests of Wall Panels ," 'Technical Report, Volume I, Test Information and Analysis, URS 7030-7,' prepared for Def'e ns e Civil Preparedne'ss Agency, URS Research Company," San Mateo; California, January 1972, . Wilton, C., Kaplan, K" and Gabrielsen, B.L., The Shock Tunnel: History and Results, Volume II· - Loading Studies, Final Report SSI 76l!8~1, .. prepared for Defense Civil Preparedness Agency, Scientific Service,' Inc., Redwood City, California, March 1978. Zaker" T:. A"Blast Pressures From Seguential "Explosions , Phase Report II, Project J6166,'IIT Research Institute, Chicago, Illinois, March,25, 1969. Primary.andS~condary Fragments
.'
56. 57,
58, 59, 60.
61.
,
Amos, C. W.,', Fragment Resistance of Martensitic St'eel Sheets
62.
York, :Annals of the New York Academy of Sciences, Conference on Preven- . tion of and' Protection Against Accidental Explosion of Munitions, Fuels and Other'Hazardous Mixtures, Vol. 152, Art. I, 1968. Doyle, J.M" Klein, M.J. and Shah, H., Design of Missile Resistant: Concrete Panels, Preprints-of the 2nd International Conference
on
'~,
Structural Mechanics in Reactor-Technology, Vol. 4, Commission of the' European Communf t i.e s , Brus sc Ls 1973, Paper No. J 3/3'. j
2C-4
TM 5-l300/NAVFAC P-397/AFR' 88-22 63. 64. 65. 66. 67. 68. 69. 70.
Effects of Impact and Explosion, Volume I, Office of Scientific Research and Development, National Defense Research Committee, Washington, D.C" 1946. Fundamentals of Protective Design for Conventional Weapons, prepared'by U.S. Army Engineer Waterways Experiment Station, Vicksburg, Mississippi, for Office, Chief ~f Engineers,. U.S. Army, Washington, D.C., July 198~. Gewaltney, R. C., Missile Generation and Protection in Light-Water- ... Cooled Power Reactor Plants, Nuclear Safety, 10(4):- July-August 1969. Giere, A. C., Calculating Fragment Penetration and Velocity Data for Use in Vulnerability Studies, NAVORD Report 6621, U.S. Naval Nuclear Evaluation Unit, Albuquerque, New Mexico, October 1959. Gurney, R. W., The Initial Velocities of Fragments from Bombs, Shells and Grenades, Report No. 405, Ballistic Research Laboratories, Aberdeen Proving Ground, M~ryland, September 1943. . Gurney, R. W., The Mass Distribution of Fragments from Bombs, Shells and Grenades, Report No. 448, Ballistic Research Laboratories, Aberdeen Proving Ground, Maryland, February 1944. Healey, J. J., and Weissman,S., Primary Fragment Characteristics and Impact Effects in Protective Design, Minutes of the Sixteenth Explosive Safety Seminar, September 1974. Hoffman, P. R., McMath, R.R., and Migotsky, E., Projectile Penetration Studies, AFWL Technical Report No. WL-TR-64-l02, Avco Corporation for the Air Force Weapons Laboratory, Kirtland Air Force Base, December
71.
72. 73. 74. 75. 76.
77. 78. 79.
80.
1964. Industrial Engineering Study to Establish Safety Design Criteria for Use in Engineering of 'Explosive Facilities and Operations; Ammann and Whitney, Consulting Engineers, New York, New York, Report for Picatinny Arsenal, Dover, New Jersey, April 1963. Johnson, C. and Moseley, J.W., Preliminary Warhead Terminal Ballistic Handbook. Part 1: Terminal Ballistic Effects, NAVWEPS Report No. 7673, U.S. Naval Weapons Laboratory, Dahlgren, Virginia, March 1964. Kolsky, H., Stress Waves in Solids, Dover Publications, New York .. New York, 1963. Kymer, J.R., Penetration Performance of Arrow Type Projectiles, Report R-18l4, U.S. Army (Frankford Arsenal, Philadelphia, Pennsylvania, May 1966. Mott, R.I., A Theoretical Formula for the Distribution of Weights of Fragments, AC-3642 (Britlsh). March 1943. Mott, R.I., A Theory of Fragmentation,-Army Operational Research Group Memorandum, l13-AC-6427, Great Britain, 1943. Non-Nuclear Weapons Effects on Protective Structures (U), Technical Report No. AFWL-TR-69-57, Air Force Weapons Laboratory, Kirtland Air Force Base, December 1964. Non-Nuclear Weapons Effects on Protective Structures (U>, Technical
Report No. AFWL-TR-69-57, (SECRET), Mechanics Research, Inc., for the Air Force Weapons Laboratory, Kirtland Air Force Base, September 1969. Recht, R. ,.et al. ,Application of Ballistic Perforation Mechanics to Target Vulnerability and Weapons Effectiveness Analysis (U), NWC TR 4333 (CONFIDENTIAL), Denver Research Institute for"the Naval Weapons Center, China Lake, California, October 1967. Reeves H., and Robinson W. T., Hastings Igloo Hazards Tests for Small Explosive Charges" Memorandum Report ARBL-MR-03356, prepared by U. S. Army Armament Research and Development Center, Ballistic Research
Laboratory, Aberdeen Proving Ground, Maryland, May 1984. 2C-5
1M 5-l300/NAVFAC P-397/AFR 88-22' 81. 82. 83.
Rinehart, J. S. and Pearson, J., Behavior of Metals Under Impulsive Loads, The American Society of Metals, Cleveland, Ohio, 1954 . . Robertson, H. P., Terminal Ballistics, Preliminary Report, Committee on 'Passive Protection Against Bombing, National Research Coun~il, January 1941. Sterne, T. E., A Note on the Initial Velocities of Fragments from Warheads, Report No ..648, Ballistic Research Laboratories, Aberdeen Proving Ground, Maryland, September 1947.
84.
Thomas, L. H., Computing the Effect of Distance on Damage by Fragments,
85,
Report NO .. 468, Ballistic Research,Laboratories, Aberd~e~ Proving Ground, Maryland, May 1944, Voltz, R,D,; and Kiger, S. A., An Evaluation of the Separated Bay Concept for a Munition Assembly' Complex: an Experimental Investigation
86,
of the Department of Energy Building 12-64 Complex, Technical Report SL83-6, prepared by Structures Laboratory, U, S, Army Engineer Waterways Experiment Station, Vicksburg, Mississippi, for Department of Energy, Albuquerque Operations, Amarillo, Texas: Ward, J, M, (DOD Explosives Safety Board), Swisdak, M, M" Jr" Peckham, P, J .', and Soper, W. G, (NSWC), Lorenz,' R, A, (Boeing Military Aircraft Company), Modeling of Debris and Airblast Effects from Explosions Inside Scaled Hardened Aircraft Shelters, Final Report NSWC TR 85-470, prepared by Naval Surface Weapons Center, Silver Spring, Maryland, May 1985. _ Shock Loads
87. 88. 89.
90. 91. 92. 93.
94.
95. 96.
American Society of Civil Engineers (ASCE), A Comparative Studv of Structural 'Response to Explosion-Induced Ground Motions, ASCE, New ,York, New York, 1975. Auld, H. E., A Study of Air-Blast-Induced Ground' Motions , Ph.D. Thesis, University of Illinois, May 1967. Baladi, G. Y., and P. F. Hadala, Ground Shock Calculation Parameter Study Report I Effect of Various Nonlinear Elastic Plastic Model Formulations, Waterways Experiment Station Technical Report S-7l-4, April 1971. Ballard, R. F., R. E. Leach, Middle North Series, Mixed Company Event, Strong Motion Seismic Measurements, Defense Nuclear Agency Report POR-6746, July 25,- 1973. Barkan, D. D., Dynamics of Bases and Foundations,'McGraw Hill Book Co., New York, New York· 1962. Baron, M. L., A Summary of Some Analytical Studies on Air Blast Induced Ground Motions, Defense Atomic Support Agency Report 2634, March 1971. Baron, M. L., I. Nelson, and I. Sandler, Investigation of Air Induced Ground Shock Effect Resulting from Various Explosive Sources, Report 2 Influence of Constitutive Models on Ground Motion Predictions,Waterways Experiments Station Contract Report S-7l-lO, November 1971. Batchelder, F. E" et al., Hardness Program'- Non-EMP, Hardness Program Plan for Safeguard Ground Facilities, Vol. 1, Management and Technical .Plan,U. S. Army Corps of 'Engineers, Huntsville Division, HNDDSP-73l53-ED-R, August 16, 1974. Carder, D. S., and W. K. Cloud. Surface Motion from Large Underground Explosions, J. Geophys, Res .. , 64, 1471-1487, 1959. Carnes, B. L" and J. A. Conway. Mine' Throw I: Cratering Effects of a Multiton Near Surface Detonation in Desert Alluvium, Waterways Experiment Station Technical Report N-73-3, May 1973. 2C- 6
TM 5-l300/NAVFAC P-397/AFR 88-22
97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111.
112.
113.
114. 115.
116.
117.
Christensen, W. J., Air Blast Induced Ground Shock, Navy Bureau of Yards and Decks Technical Study ,,27, September 1959. Cooper, H. Jr., and J. L. Bratton, Calculation of Vertical AirblastInduced Grou~d Motions from Nuclear Explosions in Frenchman Flat, Ai~ Force Weapons Laboratory Report AFWL-TR-73-lll, October 1973. coope r , H. F., Jr., Empirical Studies of Ground Shock and Strong Motions in Rock, Defens~ Nuclear Agency Report 3245F: October 1973. Crandell, F. J., Ground Vibrations Due to Blasting and Its Effect Upon Structures, Journal Boston Soc., Civil Engineers, Vol. 36, p. 245, 1949. Crandell, F. J., Transmissi~n Coefficient for Ground Vibrations Due to Explosions, J. Boston Soc. Civil Engineering, 47, 152-168, 1960. Crawford, R. E., C~ J. Higgins, and E. H. Bultmann., Air Force Manual for Design and Analysis of, Hardened Structures, Air Force Weapons Laboratory Report AFWL-TR-74~102, October 1974. .' Day, J. D., H. J,. Stout, and D. W. Murrell, Middle Gust Calibration Shots; Ground Motion Measurements, Wate!ways Experiment Station Technical Report N,-75-l, February.1975., . I; , Duvall, Wilbur I. andD. E. Fogelson, Review of Criteria for Estimating Damage to Residenc~s from Blasting Vibrations, Bureau of Mines Report of Investigation 5968\ 1962. _ . Dvorak, A.,
Seismic Effects. of Blas'ting on' Brick Houses
I
.
Proce
Geofyrikeniha Ustance Ceskoslavenski Akademie, Vol. No. 169, Geofyskialni Sbornik, pp. 189:202, 1962. , Edwards, A.T.' and T. D. Northwood, Experimental Studies of the Effects of Blasting on Structures, The Engineer, V~ 210, pp. 538-546, September 30; 1960. Eubanks, R. A. and B. R. Juskie, Shock Hardening of Equipment, Shock and Vibrations Bulletin, No. 32, Part III, December 1963. Galbraith, F. W., Operation Distant Plain Final Report Proj, 3,02b Shock Spectrum Measurements, TRW Systems Group Report 05318-600l-ROOO, January 1968. Galbraith, F. W. ,.Operation Prairie Flat, Proj. LN-306 Shock Spectrum Measurements, Defehse Atomic Support,Agency Report POR-2l07, February 1970. Grubaugh, R. E., and L. E. Elliot, Scaling of Ground Shock Spectra, Defense Atomic Support Agency Report DASA-192l, February 1967. Habberjam, G. M. and J. T. Whetton, On the Relationship Between Seismic and Amplitude and Charge 'of Explosive Fired in Routine Blasting Operations, Geophysics 17. l16~128, January 1952. Hendron, A. J. Jr., Correlation of Operation Snowball Ground Motions with"Dynamic Properties of· Test Site Soils, Waterways Experiment Station Miscellaneous Paper 1-745. October 1965. Hoffman, H. V.', F., M. sa'uer, and B. ,Barclay, Operation . Prairie Flat, " Proj. LN-308 Strong Ground Shock Measurements, Defense Atomic Support Agency Report POR-2108 '. April 1971. ., , Hudson, D. E., J. L. Al f or d ," and W. 'D. Iwan, Ground Accelerations Caused by Large Ouarry Bl;'sts, Bull., Seismic Soc.,'A, 51, 191-202, 1961. Ichiro, ,I., On the'Relationship Between Seismic Ground Amplitude and' the Ouantity of Explosives in Blasting, Reprint from Memoirs of the Faculty of Eng Inee r Ing , Kyo t o Tln Lve r s Lt.y , 15, 579-587, 1953.' Ingram, J. K., Project Officer's Final Report Operation Distant Plain Events 1. 2A, 3, 4, and 5, ·Pr'oj. 3.02A Earth Motion and Stress Measurements, Waterways Experiment Station Technical Report N- 7.1-3, May 1971. Ingram, L. F., Ground Motions from High Explosive Experiments, Waterways Experiment Station'Miscellaneous Paper N-72-10, December 1972. •
I
•
~
2C-7 ,
TM 5-l300/NAVFAC P-397/AFR 88-22 118. 119. 120. 121.
Jaramillo, E. E., and R. E. Pozega, Middle Gust Free-Field Data Analysis, Air Force Weapons Laboratory Report AFWL-TR-73-25l, April 1974. Joachim, C. E., Mine Shaft Series, Events Mine Under and Mine Ore:' Ground Motion and Stress Measurements, ' Waterways ,Experiment Station Technical Report N-72-l, January 1972. Kennedy, T. E., In-Structure Motion Measurements, Pro), LN-3l5 Operation Prairie Flat, Waterways Experiment Station Technical Report N-7011, July 1970. Kochly, J. A., and T. F. Stubbs, Mine Shaft Series, Mineral Rock Particle Velocity Measurements from' a· ·IOO-ton TNT Detonation on Granite,
122. 123. 124. 125.
126. 127.
128. 129. 130. 131. 132. 133. 134. 135. 136. 137. 138.
Defense Nuclear Agency Report POR-2l62, May 24, 1971. Lamb, H., On the Propagation of Tremors Over the Surface of an Elastic Solid, Philosophical Transactions of Royal Society London, Series A. Volume 203, September 1904. Langefors, Ulf, B. Kihlstron, and H.' Westerberg, Ground Vibrations in Blasting, Water Power, pp. 335-338, 390-395, 421-'424, February 1958 .. Lockard, D. M., Crater Parameters and Material Properties', Air Force Weapons Laboratory Report AFWL-TR-74-200, October 1974. Meireis, E. C. and A. R. Wright, Hardness Program - No-EMP, Hardness Program Plan for Safeguard Ground Facilities, Vol. 2, Safeguard Structures and TSE Description, U.S. Army Corps'of Engineers, Huntsville Division, HNDDSP-73-l53-ED-R, November 1973. . Morris, G., The Reduction of Ground Vibrations' from Blasting Operations, Engineering, pp. 460-465, April 1957. Morris, W. E., and others·, Operation Jangle Blast and Shock Measurements
I, Armed Forces Special Weapons Project Report WI 366, June 1952. Murphey, B. F~, Particle Motions Near Explosions in Halite, Journal of Geophysical Research, Vo. 66, No.3, March 1961, pp. 947-958. Murrell, D. W., Operation Mine Shaft, Mineral Rock Event, Far-out Ground Motions From a 100-ton Detonation on Granite, Waterways Experiment Station Technical Report N-72-6, April 1972. Murrell, D. Operation·SNOWBALL, Project 3.6 - Earth Motion Measurements, Waterways Experiment Station Technical Report TR 1-759. March 1967. Newmark, N. M., and others, Air Force Design Manual Principles and Practices for Design of Hardened' Structures" Air Force Special Weapons Center Report AFSWC-TDR-62-l38, December 1962. Nicholls, H. R., C. F. Johnson, and W. I. Duvall, Blasting Vibrations and Their Effects on Structures, Bureau of Mines Bulletin 656, 1971. Odello, R. and P. 'Price, Ground Shock Effects from Accidental Explosions, Picatinny Arsenal Technical Report 4995 .. November 1976 .. Palaniswamy, K., and J. L. Merritt', Evaluation of Middle Gust Data: Acoustic Path Length Scaling of Peak Velocities, Space and Missile Systems Organization Report 72-006-T4, November 1973. Parsons, R. M., and others, Guide for the Design of Shock Isolation Systems for Underground Protective Structures, Air Force Special Weapons Center Report AFSWC-TDR-62-64, October 1962. Phillips, B. R., and G. Y. Baladi, Results of Two Free-Field Code Calculations Versus Fiefd Measurements for the Distant Plain lA Event, Waterways Experiment Station Miscellaneous,Paper S-73-21, April 1973. Proceedings of the Mixed Company/Middle Gust Results Meeting 13-15 March 1973. Vol. II. Edited by General Electric-Tempo, Defense Nuclear Agency Report 3151-P2, May 1973. Rausch, Maschinenfundamente Und Andere Dynamische Bauaufgaben, Vertrieb VDE Verlag G.M.B.H., (Berlin), 1943. 2C-8
TIl 5-1300/NAVFAC P-397/AFR 88-22
139.
Reiher, H.
and F.
1
J.
Meister, Die Empfindlichkeit der Menschen ~e~en 2, No. II, pp.
Erschutterun~en, Forsch. Gebiete In~enieur wesen, Vol.
140. 141. 142.
143. 144.
145.
146. 147.
381-386, 1931. Richart, F. E. Jr.: J. R. Hall, and R. D. Woods, Vibration of Soils and Foundations, Prentice-Hall, Inc., Englewood Cliffs, New Jersey, 1970. Ricker, N., The Form and Nature of Seismic Waves and the Structure of Seismo~rams, Geophysics, 5, 348-366, 1940. Sachs, D. C., and C. M. Swift, Small Explosion Tests, Pro;ect Mole, Armed Forces Special Weapons Project Report 291, Vols. I and II, December 1955. Sager, R. A., Concrete Arch Studies Pro; 3,2. Operation Snowball, Waterways Experime~t Station Miscellaneous Paper 1-736, August 1965. Sandler, I.S., J. P. Wright, and M. L. Baron, Ground Motion Circulations for Events II and i l l of the Middle Gust Series, Defense Nuclear Agency Report DNA 3290T, April 1974. Steffens, R. J., The Assessment of Vibration Intensity and its Application to the Study 0f Building Vibrations, National Building Studies Special Report No.'19, Department of Scientific and Industrial Research, Building Research Station, London, England, 1952. Sauer, F. M., and C. T. Vincent, Ferris Wheel Series - Flat Top Event Earth Motion and P~essure Histories, Defense Atomic ,Support Agency Report Ferris Wheel POR-3002, April 1967. Sauer, F. M., editor, Nuclear Geoplosics, A Sourcebook of Under~round Phenomena and Effects of Nuclear Explosions, in 5 parts, Defense Atomic
Support Agency Report DASA-1285, May 1964. 148.
149. 150. 151. 152.
153.
154.
155.
156.
157.
Sauer, F.
M..
S~ary
Report on Distant Plain Events 6 and 18 Ground
Motion Experiments, Defense Atomic Support Agency Report DASA-2587, October 1970. Teichmann, G. A. and R. Westwater, Blastin~ and Associated Vibrations, Engineering, pp. 460-465. April 1957. Thoenen, J. R., and S. L. Windes, Seismic Effects of Ouarry Blasting, Bureau of Mines Bulletin, 442, 83, 1942. URS/John A. Blume, and Associates, Seismic Hazard and Building Structure Behavior at the Pant ex Facility, April 1976. Vincent, C. T" Operation Prairie Flat, Pro;. LN-304 Earth Pressure and Ground Shock Profile Measurements, Defense Atomic Support Agency Report POR-2ll3, May 1969. Weidlinger, P., and A. Matthews, A Method for the Prediction of Ground Shock Phenomena in"Soils, Air Force Special Weapons Center Report AFSWCTDR-6l-66, March 1962. Westline, P. S., Esparza, E. D., and Wenzel, A. B., Analysis and Testing of Pipe Response to Buried Explosive Detonations, SwRI Final Report for American Gas Association, July 1978. Willis, D. E. and,J. T. Wilson, Maximum Vertical Ground Displacement of Seismic Waves Generated by Explosive Blasts, Bulletin Seismic Safety of America, 50, 455-459, 1960. Zaccor, J. V., Procedures for Prediction of Ground Shock Phenomena Based on One-Dimensional~Shock Propagation Considerations: Procedures and Applications, Waterways Experiment Station Contract Report 3-171. April 1967. Zolasko, J. S., and G. Y. Baladi, Free-Field Code Predictions Versus Field Measurements: A Comparative Analysis for the Prairie Flat Event, Waterways Experiment Station Miscellaneous Paper S-7l-6, March 1971.
2C-9
"STRUCTURES TO RESIST THE EFFECTS OF ACCIDENTAL EXPLOSIONS"
CHAPTER 3. PRINCIPLES OF DYNAMIC ANALYSIS
TM 5-1300/NAVFAC
P-397/AF~·88-22
CHAPTER 3 PRINCIPLES OF DYNAMIC ANALYSIS INTRODUCTION 3-1. Purpose The purpose vf this manual is to present methods of design for protective construction used in facilities for development,
testing, production, storage,
maintenance, modification, inspection, demilitarization, and explosive materials.
dispo~al
'of
3-2. Objective The primary objectives are to establish design procedures and construction techniques whereby propagation of explosion (from one structure or part of a structure to another) or mass detonation can be prevented and to provide protection for personnel and valuable equipment.
The secondary objectives are to: (1)
Establish the blast load parameters required for design of protective structures.
(2)
Provide methods for calculating the dynamic response of structural elements including reinforcp.d concrete, and structural steel.
(3)
Establish construction details and procedures necessary to afford
(4)
Establish guidelines for siting explosive facilities to obtain
the required strength to resist the applied blast loads.
.
maximum cost effectiveness in both 'the planning and structural
arrangements, providing closures, and preventing damage to interior portions of structures because of structural motion, shock, and
fragment perforation.
3-3. 8ackground For the first 60 years of the 20th century, criteria and methods based upon results of catastrophic events were used for the design of explosive facilities. The criteria and~ethods_did not include a detailed or reiiable quantitative basis for assessing the degree of· protection afforded by the protective f"cility. In the late 1960's quantita~ive procedures were set forth'in the first edition of the present manual, "Structures· to Resist -the Eff'e ct s o f
Accidental Explosions.".
This manual was based on extensive' r e s e archvarid
development programs which permitted a more reliable approach to current.and future design requirements. Since the'original publication of this manual, more extensive testing and development programs have taken place. This additional research included'work with materials other than reinforced concrete which was the principal construction material referenced in the initial version of the manual.
Modern methods for'the manufacture and storage of explosive materials, which include many exotic chemicals; fuels, and pr~pellants, require less space for a given quantity of~exp~osive material than was previously needed. Such concentration of explosives increases the possibility of the propagation of
accidental explosions. (One accidental explosion causing the detonation of
3-1
T~
5-1300/NAVFAC P-397/AFR 88-22
other explosive materials.)
It is evident th~t a requirement for more. accu-
rate design techniques is essential. This manual describes rational design methods to provide the required structural protection. These design methods account for the close-in effects of a detonation including the high pressures and the nonuniformity of blast loading on protective str~ctures or barriers. These methods also account for intermediate and farrange effects for the design of structures located away from the explosion. The dynamic response of structures,
construct~d
of various materials, or
combination of materials, can be calculated, and. details are' given to provide the strength and ductility required by the design. The design approach is directed primarily toward protective structures subjected to the effects of a high explosive detonation. However, this approach is general, and it is applicable to the design of other explosive environments as well as other
explosive
ma~erials
as mentioned above.
The design techniques set forth in this manual are based upon the results of numerous ful1- and small-scale structural response and explosive effects tests
of various materials conducted in conjunction with the development of this manual and/or related projects. 3-4. Scope It is not the intent of this manual to establish safety criteria. documents should be consulted for this purpose ..
Applicable
Response predictions for
personnel and equipment are included for information. In this manual an effort is However, sufficient general been included in order that situations other than those
made to cover the more probable design situations. information on protective design techniques has application of the basic theory-can be made to which were fully-considered.
This manual"is appli~~ble to the design of protective structures subjected to the effects associated with high explosive detonations. For these design situations,the manual will apply for explosive quantities less than 25,000 pounds for c10se-in·effects. However, this manual is also applicable to other situations suc~ as far- or intermediate-range effects. For these latter cases the design procedures are applicable for explosive quantities in the order of 500,000 po~ndswhich is the maximum quantity of high expYosive approved for aboveground storage facilities in the Department of Defense manual, "Ammunition and Explosives. Safety Standards", DOD 6055.9-STD. Since tests were primarily directed toward the response of structural steel and reinforced concrete elements to blast overpressures, ~his manual concentrates on design proce~ures and techniques for these materials. However, this does not imply that concrete and steel are the only useful materials for protective construction. Tests to establish the response of wood, brick·blocks, and plastics, as well as the blast. attenuating and mass effects of soil are contemplated. The results of these tests may require, at a later date, the supplementation of these design methods for these and other materials. Other manuals are available to design protective structures against the effects of high explosive or nuclear detonations. The procedures in these manuals will quite often. complement this manual and should be consulted for specific applications. 3-2
TM 5-1300/NAVFAC P-397/AFR 88-22
Computer programs, which are consistent with procedures and techniques contained in the manual, have been approved by the appropriate representative of the US Army, the US Navy, the US Air Force and the Department of Defense Explosives Safety Board (DDESB). These programs are available through the following repositories: (1)
Department of the Army Commander and Director U.S. Army Engineer Waterways Experiment Station Post Office Box 631 Vicksburg, Mississippi 39180-0631 Attn: WESKA
(2)
"
Department of the Navy Commanding Officer Naval Civil Engineering Laboratory Port Hueneme, California 93043 Attn: Code L5l
.'
.,
"
'.'
(3) :. Department of the Air Force Aerospace Structures Information and Analysis Center
Wright Patterson Air Force Base Ohio 45433, Attn: AFFDL/FBR If any modifications to these programs are required, they will be submitted
for review by DDESB and ,the above services. Upon concurrence of the revisions, the necessary changes will be made and notification of the changes will be made by the individual repositories. 3-5. Format This manual is subdivided into six specific chapters dealing with various aspects of design. The titles of these chapters are as follows: Chapter Chapter Chapter Chapter Chapter Chapter
1
Introduction
2
Blast, Fragment, and Shock Loads Principles of Dynamic Analysis Reinforced Concrete Design Structural Steel Design Special Considerations in Explosive Facility Design
3 4 5 6
\
When applicable, illustrative examples are included in the Appendices. Commonly accepted symbols are used as much as possible. However, protective design involves many different scientific and engineeri.ng fields, and, therefore, no attempt is made to standardize completely all the symbols used.': Each symbol is defined where it is first used, and in the ll.st of symbols at the end of each chapter.
3-3
..
TM 5-1300/NAVFAC P-397/AFR 88-22 CHAPTER CONTENTS 3-6'. General This chapter contains the procedures for analyzing structural elements subject to blast overpressures. These procedures are contained in the next eleven sections; Section 3-7 deals with a simplified discussion of the basic principles of dynamics as well as the procedures for calculating the various components used to 'perform the dynamic analyses. .Pr e senced in Sections 3-8
through 3-15 are resistance-deflection functions for various elements including both'one- and two-way panels as well as beam elements, These functions . include the elastic, elasco-plastic, and plastic ranges of response.
In
addition, a discussion 9£ dynamic equivalent systems is presented in Sections 3-16, and 3-17. These include single- and multi-degree-of-freedom systems. Presented in this Section also are methods for calculating load and mass factors required to perform the dynamic analyses.
Sections 3-18 through 3-20 include both a step-by-step numerical integration of an element's motion ,under dynamic loads utilizing the Acceleration-Impulse. Extrapolation Method or the Average Acceleration Method and design charts for idealized loads. Presented also in these Sections are methods for analyzing elements subjected to impulse type loadings; that is, loadings whose durations 'are short in comparison to the time to reach maximum response of the elements.
3-4
1M 5-l300/NAVFAC P-397/AFR' 88-22. BASIC PRINCIPLES 3-7, General The principles ·used.~n the analysis of structures under' static load will be reviewed briefly, ·since the same principles are used in the analysis and design of struc'tures subjected to dynamic loads: Two· different methods are used either separately or concurrently in static analysis:
one is hased'on
the principle of equilibrium, and the other on work done and internal energy stored.
Under the application of external loads, a given structure is deformed and internal forces developed in its members. In order to satisfy static equilibrium, the vector sum of all the external and ·internal forces acting on any free body· portion of the structure must beequa~to zero, For· the equilibrium. of the structure as a wr.ole,: the vector sum of the· external forces and the rea~tions of the foundation must be equal to 'zero.
The method based on work done and energy considerations is sometimes used when it is necessary to determine the deformation of a structure. In this method,
use is made of' the fact that the deformation of the structure causes the point of application of the external load-to be.displaced.
The'.force then does work
on the structure. Meanwhile, because of the structural· deformations, potential energy is stored in the structure in the form of strain-energy. By the principle of energy conservation, the work done by the external 'force and the energy stored in the members-must be equal.
In static analysis, simplified
methods such as the methodof.virtual work and the method of the uni't load are derived from the general principle of energy conservation.
In the analysis of statically indeterminate 'structures, in addition to satisfying the equations of equilibrium, it is necessary to include a .
calculation of the deformation of the structure in order to arrive at a complete solution of the internal forces in the structure. The methods based on energy considerations such as the method of least work and the method based on Castigliano's theorems are generally used. For the analysis of structures under dynamic loading, the same two methods are basically used; but the load changes rapidly with time and the acceleration velocity and, hence, the inertia force and kinetic energy are of magnitudes requiring consideration.
Thus, in addition to the internal and external
forces, the equation of equilibrium includes the inertia force and the equation of dynamic equilibrium takes the form of Newton's equation of motion:
F - R - Ma
3-1
where
F
total external force as a function of time
R
total internal force as a function of time
M total mass a = acceleration of the mass As for the principle of conservation of energy, the work done must be equal to
the sum of the kinetic energy and the strain energy:
WD - KE + SE
3-2
3-5 .
,
TM 5-l300/NAVFAC
where WD
KE SE
P-397/AF~
88-22
work done
kinetic energy strain energy
and the strain energy includes both reversible elastic strain energy and the irreversible plastic strain energy . . Thus, the difference between structures
under static and dynamic loads is the presence, of inertial force (Ma) in the, equation of dynamic equilibrium, and of kinetic energy in the equatio~ of energy conservation. Both terms are related to the mass of the structure; hence, the mass of the structure becomes an important consideration in dynamic
analysis. In the dynamic analysis of structures, both the energy balance equation and the force balance equation are applied with explicit description of the external forcing function F, and the internal resisting forcing function R. The difference between these forcing functions is the inertia force as described above. The following is a discussion of the details of how these forces are utilized in the design of structures which respond in the ductile mode.
In the design of a structure to resist ,the blast from an HE explosion, the total external force acting on the structure can be obtained by the principles discussed Chapter 2. The design method also consists of the determination of the total internal force, i.e. the resistance of the structure required to limit calculated deflections of the individual members and the structure as a
whole under the external force (blast loading), to within prescribed maximum values. The determination of the resistance of the individual members of the
structure is presented in Sections 3-8 through 3-17. Subsequent sections of this manual present the principles and methods of dynamic analysis and equations, charts, and procedures for design.
3-6'
TM 5-l300/NAVFAC P-397/AFR 88-22
RESISTANCE - DEFLECTION FUNCTIONS . 3-8. Introduction' Under the action of external loads, a structural' element is 'deformed and internal forces set up. The sum of these internal forces tending to restore the element to its unloaded static, position is defined as' the 'resistance. The resistance of a s t ruc tur a I element is a reactive· force" associated with the deflection of the element produced by the -applied 'load. It is'convenient to consider 'the resistance as "an equ.LvaLe nt; 'load in the same manner as the applied load, but opposite in direction'.' "I'he 'variation 'of the resistance vs .
displacement is expressed bya resistance-deflection function and may be represented graphically. ,An idealized resistance-deflection function for an element spanning~in two directions and'covering .in the complete flexural range to incipient failure is shown in Figure 3-1. . " As load is applied to a structural element, the element deflects and"at any instant " exerts a resistance to further deformation, which is a function of its units stiffness K, until, the ultimate unit resistance ,r u (total resistance is r u where A is .t.he element area) of the element is reached "at deflection
~.
The initial portion of the'resistance-deflection diagram is composed of the elastic and elasto-plastic ranges, each with its corresponding stiffness, the transition from one range to another occurring-as plastic hinges are formed at po Ln t svo f max Lmum.vs t r e s s (yield .Lf.ne s ). 'The numbe-r of elasto-plastic ranges .
required before the ultimate resistance of 'a paz t Lcu l a r element 'is reached depends upon the type and number of supports. and the: placement of reinforcing
steel (in the case of reinforced concrete elements).
For example a .beam with
simple supports subjected to uniformly distributed loads needs only one
plastic hinge to develop the ultimate resistance (or full plastic strength) of the element; whereas for the same beam fixed at both ends, more than one
plastic hinge iscrequired. '
.
In subsequent paragraphs, various procedures, equations and illustrations are presented to enable the designer to. determine the resistances of both one- and two-way 'elements. The procedures outlined apply mainly to reinforced concrete elements and so do the equations appearing in the text, unless the equations are given as part of an illustrative example.
However, they can also be used
for structural steel elements as well as other structural elements such as alwnirturn, plastics,'etc. Equations have been derived for" specific 'cases mostoften encountered in practice.
These are applicable fio r structural s t ee L'
reinforced concrete elements of uniform thickness in both the horizontal and vertical directions. Before "the e qua ti.oris and :rfigures c-an be used for reinforced concrete element, however, the reinforcing steel across ·any yield iine must have a uniform distribution in both the vertical and horizontal direc~ tions; however, the reinforcement across the positive yield lines can be
different from that across the negative yield lines and the reinforcing' pattern in the vertical direction different from that in the horizontal direc ti on .
Regardless "o f whether it is reinforced concrete or structural steel 'e l.emenr , any opening in"the element must be compact in shape,and small in area, compared to the total area of the element.
3-7
TM 5-l300/NAVFAC P-397/AFR 88-22
3-9. Ultimate Resistance 3- 9.1. General The ultimate
(1) (2) (3)
(4)
resi~tance
of an element depends upon:
The distribution of. the applied .loads. The geometry of the element (length and width). The n~ber and type of supports. The distribution of the moment capacity or reinforcement in the case of reinforced concrete elements.
The distribution of the loads depends..upon the design range of the element; i.e., high, intermediate or low pressure. For intermediate and low pressure .anges, it can ,be assumed that the pressure is uniform across the surface of the element although it varies with time. At high pressure ranges, however, the blast loads are variable across the surface of the element . However, .for structural steel elements and concrete elements utilizing laced
reinforcem~nt,
or for concrete elements with·standard· shear reinforcement which sustain relatively small deflections, a good estimate of;the resulting deflections can be made using the resistance functions conforming to those of uniformly loaded elements.
The other factors that affect the ultimate resistance of an element are predetermined by the requirements of the protective structure (where the element is used) and the magnitude of the blast output. '. 3-9.2. One-Way Elements The ultimate resistance of a. one·way reinforced concrete elemen~ with an elastic distribution of its reinforcing steel is based on the moment capacity at first yield since all critical sections yield simultaneously. For one-way reinforced concrete.elements (such as beams .or slabs) with non-elastic distribution of reinforcing steel and for structural steel elements, the ultimate resistance is a function of the moment capacity at the first yield plus the added moment capacity due to subsequent· yielding at other critical sections. _.
Values of the ult~mate resistance for one-way elements are shown in Table where the fo~lowing symbols are used:
3-~
.ultimate negative unit moment capacity at the support. ultimate positive unit moment capacity at midspan. length ultimate unit resistance total ultimate resistance
Table 3-1 applies to both beams and slabs. However, special attention must be paid to the units used for the respective element. The moment capacity of a slab is expressed for a unit strip of the slab· (inch-pounds per inch) whereas the total moment capacity (inch_pounds) is considered fora beam. Consequently, the resistance of a slab is expressed in load per unit area (psi) where·',,, the resistance of a beam is expressed in load per length along the beam (pounds per inch). . 3-8
1M 5-1300/NAVFAC P-397/AFR 88-22
3-9.3.
Two-Wa~
Elements
.,
The amount of·data available on the limit analysis of rectangular steel plates is very'limited. However, an.elementary approach imagines a' mechanism formed of straight yield lines " 85-- is customary in reInforced concrete. This
approach for reinforced'concrete elements will be considered appropriate' for structural steel elements.
In the design of two-way reinforced concrete elements, it is not necessary to define accurately the stress distribution during the initial and intermediate
stages of loading since the ultimate load capacity can be readily determined by the use of yield line procedures. The yield line method assumes that after initial cracking of the concrete at points of maximum moment, yielding spreads until the full moment capacity is developed along the length of the cracks on which failure will take place. Several illustrative examples of the simplified yield or crack lines for two-way elements are illustrated in Figure 3-2. In using the yield line solution, the initial step is to assume a yield line pattern (as shown in Figure"3-2)" applying the following rul:es: (1)
To act as plastic hinges of a collapse'mechanism made 'up of plane segments, yield lines must be straight lines forming axes of
rotation for the movements of the segments. (2)
The ,supports of the' slabs will act as axes of rotation.
A yield line may form along a fixed support and an axis of rotation will
pass over a column.
(3)
For compatibility of deformations, a yield line must pass through the intersection of the axes of rotation of the adjacent slab segments.
'I
Tests indicate that the actual location and extent of these lines on rein-
forced concrete elements differ only slightly at failure from the theoretical ones. Use of the idealized yield lines results' in little error in the determination of the ultrmate resistance and the error is on the side of
safety. The corner sections of two-way elements are stiff in comparison to the remainder of the member; therefore, straining of the reinforcement which is
associated with the reduced rotations at these sections will be less.
To
account for the corner effects, the design of anyone particular section of a
two-way element should consider a variation of the moment capacity along the yield lines rather than a uniform distribution. . This variation is approximated by taking the full moment capacity along the yield lines, except in the corners where two-thirds of the moment capacity . ..' over the lengths described in Figure 3-3 are used. The variation applies to
both the negative moments olong the supports and the positive moments at the interior,The ultimate unit resistance can be determined from the yield line pattern using either the principle of virtual work or the equations of equilibrium. Each approach has" its advantages'; in general,' the virtual work method ,is easier in principle but difficult to manipulate algebraically since it involves differentiating a usually complex mathemati~al expression for a minimum value of resistance. The equilibrium method, which is used in this
3-9
TM 5-l300/NAVFAC P-397/AFR 88-22
manual, also has its disadvantages. Since equilibrium requires that ,the shear forces acting on each side of a yield line have to be equal and opposite, correction forces (also known as nodal forces) have to be introduced around openings in two-way members and at free edges, and~these correction forces may not be available from simple analysis. However, in three of the six cases shown in Figure 3-2,
•
(cases,c, e, f), nodal forces exist; but their effects
are negligible. In order to calculate the ultimate unit resistance r u of· a two·~ay element, for~ed by the yield lines is expressed in terms of the moments produced by the internal·and external forces. The sum of the resisting mo~ents acting along the yield lines ·(both positive and negative) of each sector is equated to the moment produced by the applied load about the axis of rotation (support of the sector), assuming that the shear forces are zero along the positive'yield lines. the equation of equilibrium of each sector
3-3
R
sum of the ultimate unit resisting moments acting along the support (negative yield lines) sum of the ultimate unit resisting moments acting along the interior failure lines (positive yie ld lines) total ultimate resistance of the sector
c
distance from the centroid of the load to
where
the line of rotation of the sector ultimate unit resistance of the sector
area of the sector
Once the equations of equilibrium are known for all sectors, the ultimate resistance is obtained either by solving the equations simultaneously or by a trial and error procedure noting that the unit resistance of all sectors must be equal. To illustrate the above procedure (Equation 3-3), consider the two-way concrete element shown in Figure 3-3 which is fix~d on three edges and free on the fourth, and where the nomenclature is as follows: L H
x
length of element height of element yield line location in horizontal direction yield line location in vertical direction ultimate unit negative moment capacity in the vertical direction ultimate unit positive moment capacity in the vertical direction ultimate unit neg~tive moment capacity in the horizontal direction ultimate unit positive moment capacity in the horizontal direction
The nomenclature as stated in the paragraph above is strictly applicable to two-way elements which are used as walls .. However, when roof slabs or other horizontal elements are under consideration, the preceding nomenclature will 3-10
•
TK 5-l300/NAVFAC P-397/AFR 88-22
also be applicable if the element is treated as being rotated into a vertical position.
The first step in the solution is to assume the location of the yield lines as defined by the coordinates x and y. It should be noted that in some cases, because of geometry, the value of x and y will be known and therefore need not be evaluated. In this example, the negative reinforcement in the horizontal direction at opposite supports is assumed to be equal; therefore, the vertical
yield line is located at the center of the span and the value of x is numerically equal to L/2 (a, Figure 3-3). However, in other cases, neither the location of x nor ywill be known, and the solution' will require the determination of both coordinates.
Once the yield lines have been assumed, the distribution of the resisting moments along the yield lines is determined. In the case at hand, the reduced moments,' as a result of the increased stiffness at the corners, act over
lengths equal to x/2 and y/2 in the horizontal and vertical directions, respectively (a, Figure 3-3). The equations'of equilibrium are then written for each sector with the use of the free body diagrams (b, Figure 3-3). For the triangular sector I: . MVN - (2/3)MVN (L/4 + L/4) + MVN (L/2) (5/6)MVN L MVp
3-4
(2/3)MVp (L/4 + L/4) + MVp (L/2)' - (5/6)MVp L
CI
y/3
RI
( MVN +
3-5 3-6
MVP)/C I
- [5L(MVN + MVp)1/2y
3-7
Ly/2
3-8
RI/A I [5(MVN + Mv p)1/y2
3-9
For the trapezoidal sector II, a similar procedure gives MHN
(2/3)MHN (y/2) + MHN (H - y/2) - MHN (H . y/6)
MHP
3-10
(2/3)MHP (y/2) + MHP (H - y/2) MHP (H - y/6)
.;
Cn
3-11
(l/3)(L/2)[2(H-y) + Hl/(H + H - y) "
[L(3H - 2y)]/6(2H
y) 3-11
3-l2
TH 5-1300/NAVFAC P-397/AFR 88-22
- [(6H - y)(2H
y)(MHN + MHP)]/L(3H
2y)
(1/2 ).(L/2) (H + H - y)
AlI
3-14
[L(2H - y) J/4 ru(Sector II)
3-13
~
RII/A I I
- [4(MHN + MHP)(6H -Y)J/L 2(3H
2y)
3-1?
Equations 3-9 and 3-15 are the equations of equilibrium for the triangular (I) and the trapezoidal (II) sectors, respectively. As mentioned previously, these equations can be solved simultaneously or by a trial and error proce-
dure. In the latter method, values of yare substituted into both equations until r u (sector 1) is equal to r u (sector II). If a numerical solution based on the above procedure (Equation 3-3)-yields, negative values for either x, y or r then the assumed yield line location is wrong. In this example, the only other possible yield line pattern (x ~ L/2) would be as shown in Figure 3-2c. , The solution of Equation 3-3 is universally applicable for any ,two-way element. If the negative reinforcement in the horizontal direction had been u~equal at the opposing supports, the value of x - L/2 would have changed, and all three sectors wold have had to be consider~d to determine x, y and hence, r u' Simultaneous solution of Equations 3-9 and 3-15 reveals that the locaticns of the yield lines are a function of the ratio of the spans L/H and the ra t i,o of the sum of the unit vertical to horizontal moment capacities as follows:
ru(Sector I) - ru(Sector II)
3-,16
5(MVN + MVp)/y 2 - [4(MHN + MHP)(6H - y)J/L 2 (3H - 2y) L2(MVN + MVp)/H2(MHN + MIIP) [ 4y2(6 y/H)J/[5H 2(3 - 2y/H»)
3-17a
3-17b
, 1/2 (L/H) [(MVN + MVN)/(MIIN +, MHP)J (y/H)[(4(6
y/H)/5(3 - 2y/H)]1/2
3-17c
Equation 3-17c, which relates the location of the yield lines to the' moment capacity of the element, is used to plot Figure 3-6. Kno~ing the location of the.yield lines, the resistance of the two-way element can be ,'obtained from
either Equation 3-9 or 3-15 which are also presented in Table ,3 - 2.
Using the procedure outlined above, the values of the .ultimate unit resistances for several two+way elements with various support c~nditions are given in Tables 3-2 and 3-3, the nomenclature confirming to that previously listed. Table 3-2 covers the special cases where opposite supports provide the same
3-12
TIl 5-1300/NAVFAC .p c 397/ AFR 88-22
degree of restraint thus resulting in symmetrical yield line patterns." Table 3-3 deals with the general cases when the yield line patterns are not· symmetrical,(that is; when opposite supports provide different restraints). Yield line location ratios x/L and y/H for the same elements are depicted in Figures 3-4.through, 3-20. '. Figures 3-4 and 3-5 show the location of the yield lines for two-way elements with two adjacent edges supported and the other two free.
In each of these
figures eight curves are shown which" represent different ratios of the '" positive to the negative moment capacities in both the vertical and horizontal
directions. Figures 3-6 through 3-16 illustrate the yield line location for two -way: elements with three edges supported and one edge free. .; Figures 3" 6 and 3-11 covers the case when the yield ,line' pattern is'symmetrical opposite supports' provide the same degree of restraint). "Figures 3-17 through 3"20 show the yield line location' fo r--two-way elements with four sides supported. f
" ' . '
Figure 3-17 covers the special case when· opposite ,supports provide ,the same degree of restraint thus resulting in ,a symmetrical yield pattern.' An example illustrating the use of some of these figures is provided in Appendix A, t:
3-9.4. Openings in TWO-Way Elements The use of openings in two -way elements, whether for access as a door' operri.ng or for visual communication,as in the case of observation ports, is permissible with certain reservations. It is difficult to state exact rules
concerning openings, but their effect on the design is generally a function of location, size and shape.
Small compact openings with approximate are a s.. of less than -5 percent of the
panel area and located away· from regions of high stress can usually be ignored in the design~ However, as' in the case of conventional design; reinforcement at least equal to the amount interrupted should be placed adjacent,to the opening. For example, in Figure 3-21, the openings shown in (a) and (b) can be disregaraed. If the opening in, (b) were made more rectangular' as in (c), then the dekign must be modified'to account for the change in the"yield 1ines and, hence,' the change in the resistance.
This change in resistance is a
function of both the shape and the location of the opening. Door openings invariably requireispecial analysis because,of their·size.~ As
depicted in (d), (e)·and (f), Figure 3-21, the presence of door openings causes gross relocations of the yield lines which generally propagate from the corners of the openings. Since the door also su~tains the blast loading, concentrated line loads are present around the periphery of such openings.
These concentrated loads must be'included in the analysis since they change the resistance, As previously outlined for solid elements and for this· case also,,'the'yield line'locations are assumed and each sector is individually analyzed. The presence of line'loads modifies Equation 3-3 to . DIN +
~
- ruAc, +
Me '..,
3,18
where Mc is the moment of the concentrated loads about the line of rotation of the sector being considered.
Solution of elements with openings is most
easily accomplished through a trial and error procedure by se tt Lng upvt he simultaneous equations for each sector and
until the
severa~
values of
.T
1assuming
various values of x and y
u agree 'to within a few percent.
3-13
TM 5-1300/NAVFAC P-397/AFR 88-22
3-10. 'Post-Ultimate Resistance In general, the two-way elements described in this manual exhibit a postultimate resistance after initial failure occurs as indicated in Fig~re 3-1. Prior to this partial failure, the element is spanning in two directions with a resistance equal to the ultimate resistance r u'
At a par.ti~ular deflection,
denoted as Xl' failure occurs along one side or two opposite sides, and'the element then spans in one direction with the reduced post-ultimate unit resistance r u until. complete failure occurs atldeflection One-way elements do not exhigit this behavior.. .. . ,
Xu.
TheiLo ca t Lon.ro f the yield lines determines the presence or absence "of this. range. If the yield lines emanating from the corners of the elements bisect the 90-degree. corner angle, then all· supports fail simultaneously and there is. no post-ultimate range.
,As previously shown, the location of the yield lines
for a particular element is a function of L/H and the ratio of the unit vertical to .horizontal momento·capacities. Post-ultimate resistances for twoway elements are shown in Table .3,4. ;, 3-11. Partial Failure and Ultimate Deflection Partial failure deflection Xl' for two-way elements and ultimate deflections ~
f o r sbo ch one-way and two-way elements are a function of the angle of
rotation of the element at its. supports. and the geometry of the sectors formed by the positioh yield lines.
-. Once the ultimate resistance r u is reached (full moment capacity developed
along the yield ~ines), the structural element becomes a mechanism which rotates with no further, increase in either .che -mome nt; or curvature between the hinges. -For one~way elements" the rotation continues and the deflection increases until ei ther ~ the maximum deflection ~,is reached o r- failure occurs
Xm
at e max, The equations for the maximum deflection in the range 0 $ X < for several one-way elements as a function of the'rotation angle e and the ultimate deflection ~ are given in Table 3-5. when the values for ~ are based on the,development of a maximum support rotation,
e
~
max., prior to,
failure. Actually, the maximum support rotation will vary with the material type and geometry of'the element. The criteria for partial and incipient faiiure for concrete and,structural steel elements can'be fouhd in Chapters 4 and 5
respectively. For two-way elements, the r o t at Lons.i o f all, the sectors rnus't; be considered. in
order to define the deflections of partial and. incipient failure, Prior to partial failure (0,$ $ Xl)" the maximum deflection is a function of the.
Xm
larger angle of rotation formed along either the,. vertical. or horizontal sup-
ports, At deflection Xl' this larger angle equals e max, and failure occurs along this support. Beyond this point, the element spans in one direction until the angle of rotation at the adjacent supports (in the direction opposite to that at which failure has already occurred) reaches
time total collapse occurs
(Xm
e
at 'which
~ ~),
To illustrate the above, consider a two-way'element (Figure 3-22) which is fully restrained on four edges ,and whose positive yield lines are' defined by H/2 < x < L/2 and y - H/2. Denoting·eH as the angle of rotation in the. I 3-14
TM 5-l300/NAVFAC P-397/AFR 88-22
ev
horizontal direction (along vertical supports) and as the vertical angle of rotation (along horizontal supports), the maximum deflection ~ at' the center of the element prior to reaching the deflection Xl is 3-19 and at the partial failure deflection Xl where 8 v max 3-20 Referring to Figure 3-22, the deflected shape at deflection Xl is indicated by the solid line and 8 H has value B which is defined as
B - tan-l(X1/X~ - tan-l(H tan 8max/2x)
3-21
As the'element continues to deflect the angle of 'rotation 8 H increases; its magnitude becoming equal to 8H -
A +
B
3-22
where A is the angular rotation in excess of B, For a two-way element which undergoes partial failure but does reach incipient failure~·rthe maximum deflection in the range X1S'~ S ~ becomes ~
- x tan 8 H + [(L/2)
When 9 H equal 8 ma x.
~ - x tan
x)] tan (8 H - B)
3-23
the ultimate deflection ~ at incipi~nt failure is
8 ma x +
[(L/2 - x)] tan [8ma x - tan- l (H tan 8max/2x)] 3-24
Equations 3-19 through 3-24 are specifically for two-way elements described in Figure 3-2 and will vary for other two-way elements with different material properties and geometry;
The maximum deflection ~ for several Xl' and Xl < ~ S ~ as a function of in Table 3-6 along with the values of deflections, The support which fails cated,
two-way elements in the ranges 0 S ~ S the rotation angles 8 H and are given partial failure (Xl) and ultimate (~) at partial deflection Xl is also indi-
ev
3-12, Elasto-Plastic Resistance As stated in Section 3-8, the· initial' portion of the resistance function (Figure 3-1) generally is, composed of an elastic and one or more elastoplastic ranges. The elastic unit resistance'~e' is defined as the resistance ~t which f~rst yield oc:urs; simila:ly, the e:asto-plastic unit resistanc: rep ~s
the reslstance at WhlCh second.ylelds subsequently occur:
Where all htnges
form in a member at one time,-r e will be equal to the ultimate unit resistance
r u; where two or more hinges are formed at separate times, the maximum value
of rep will be equal to r u (depending upon the hinges formed,'one or more
values of rep may exist).
These resistances for one-way elements are listed in Table 3-7. ~In' those cases where the elasto-plastic resistance is equal to the ultimate resistance,
the value can be determined from Table 3 cl, 3-15
TK 5-l300/NAVFAC P-397/AFR 88-22
The determinatio~ of the elasto-plastic resistances of two-way elements ~is more complicated than that for one-way elements, since the resistance varies with the span ratio and, in the case of. reinforced concrete elements, with the
placement of the reinforcement. Data for calculating the resistances of twoway elements during the elasto-plastic ranges (graphically summarized in Figure 3-23) are presented in Figures 3-24 through 3-38. Figures 3-24 through 3-26 are for a two-way element support~d'on two 'adjacent sides and free at the others. Figures 3-27 through 3-32 are for a tW0-way element supported on three sides and free on the fourth, while Figures 3-33 through 3-~8are for. elements f Ixed. on, four sides. The resistances in each range can readily be determined using the coef f LcLencs B, .the subscript referring to the points listed in the accompanying illustration. For example, in Figure 3-27, for an el~ment fixed on three sides and free on
the fourth,
ifth~
ultimate unit resisting moments
Ku
are known for points I,
2 and 3, a resistance r for each point can be calculated from
r -
Ku
/BH 2
3-25
where .the va Lue s of B are found in Figure .3-27. The smallest value of resistance r, (say. at poin~ 2) corresponds to. the first yield and is equal to reo Next, the moments at the remaining ~wo poi~ts are computed for this value of reo and the differences between "these and the ultimate values are determined.
These differences represent the remaining moment capacities. available for additional load. At r e, the elements supports become free, fixed, and simplysupported on opposite.sides. Using the moment differences and entering Figure 3-29 two values of the change in resistance can be calculated as above, the smaller being dr and therefore: 3-26 .'
In similar fashion, the resistance at the end of each range can be determined until the ultimate unit resistance r u is reached. 3-13. Elasto-Plastic Stiffnesses and Deflections The slopes of the elastic and elasto-plastic ranges of the resistance are defined by the stiffness K of the element:
K - r/X
fu~ction
3-27
where r is the unit resistance and X i s the deflection corresponding to the
value of r. Th~ elastic range stiffness is denoted as Ke , the elasto-plastic range.as Ke p ' .wh i Le in the plastic ..range the stiffness is zero. Typical resistance-deflection functions used for design are shown in Figure 339. One- and two-step systems are generally used for one-way elements while two- and three-step.systems are used for two-way elements. Two.wayelements fixed on all four s~des will exhibit a four step system. As can be seen from the figure, the elastic range stiffness 3-28 the elasto-plastic stiffness for a two,step system 3-16 :
I
I
TM,5-1300/NAVFAC' P-397/AFR 88-22
3-29 and for a three-step .system, the elasto-plastic s tLff'ne s s e s-
, and
3-30
,
n.
~'
3-31 ,,'
The elastic and e Laa t o vp Laa t.Lc stiffnesses of one-way. e~em_ents are, 'given" in Table 3-8 as a function of- the modulus of e Las t Lc Lt.y, E"nioment -o f .Lrie r t La I, and span length. Knowing" the resistances and t s, the i elastic and elasto-plastic deflections can be computed from the above equas
Lff'ne s
s
e
co r
r
e
spond
tions.
ng
..." f'
>
"
.
'.
The determination of the elasto-plastic stiffnesses and deflections of two-way elements is more complicated than for one-way elements. s i.nce .ano the r var Lab l.e , i
namely,
the a~pect ratio L/H, must he considered.
For two-way elements"
the
deflections at the end of each range of behavior is obtained from the y coefficients presented in Figures 3-24 through 3-38. The deflection for each range of behavior is obtained from '. , ',': 3-32
XD -
where D, the flexural rigidity of the element is. defined cas ,
,
".,..,.
,,"'
D ~ EI/(l-~~)
3 -33
E is the modulus of elasticity, ·1 is the moment of iner,tia, and v is Poisson's ratio . .
I~
must-be
~ealized
that except for the elastic'range',
the values ,of X
(the displacement) ,and r in .Equation 3- 32 represent. change.i Ln deflection and resistance from one r:,,?nge of behavior to ano t.he r.. Therefore, for two-way.'. members the change ~n deflection and resis~ance (as previou~ly explained) is obtained from .Figures 3-24 through 3-38 and the stiffnesses are computed from Equations 3-28 through 3-33. 3-14. Resistance-Deflection Functions for Design ,
,
3-14.1. General The r e s t s t ancev de f l.ect Lon function used for design depends' upon the maximum"
permitted deflection according to the design criteria ofrthe 'element .be Lng , considered.
This maximum
d e
f
L e c
t
i . o r i
~
can be
c
a
t
e
g o
r
L z
e d v a s
either limited
or large. In the limited deflection range"the .maximum deflection of the system is limited to the elastic, elasto-plastic and plastic r ange s , . When the maximum deflection falls in the large deflection range, the response of the system is mainly within the 'plastic range and the elastic ,and elasto-plastic ranges need not be considered. The error resulting from the omission of the elasti~ and elasto-plastic portions in this; analysis is negligible. The support rotation that corresponds .to limited deflection varies for· the different materials used in protection design. "The response criteria for each material is obtain~d from the chapter that describes the design procedures for that material, 3-17
TH 5-l300/NAVFAC P-'397/AFR 88-22. - ,
3-14.2. Limited Deflections When designing for limited deflections, the maximum deflection lSn' of the' element is kept within the elastic, elasto-plastic, and limited plastic ranges, and the resistance-deflection function for design takes the form shown in Figure 3-39 a, -b , and c for a one-step system, a two-step system, and a three-step system, respectively. The design charts presented in Section 319.3.were established for a one-step system; for two- and three-step systems, these.charts can be used if the resistance-deflection functions are replaced with equivalent' elastic resistance-deflection functions defined by KE and XE as indicatedby,the dotted'lines 'in Figure 3-39. The equivalent elastic stiffness KE and the equivalent maximum elastic deflection XE are calculated suchthat·.the 'area -unde r the·dotted curve is equal to the area" under the solid curve, thereby'producingthe same' potential energy in each system. The
•
equivalent maximum elastic deflection XE for the two-step and three-step
systems shown is expressed by Equations 3-34 and 3-35, respectively. XE .:. Xe' + , .... , .
The
X' E
- xe
equiva~ent
~(l',-,
3-34
re/r u)
(rep/r u) + Xep(l
re/r u) + l)j(l - rep/r u) , ,
,
3-35 '
elastic stiffness KE in each case is equal to 3-36
One-way elements exhibit one vvand' two vs t.ep resistance deflection curves
depending on the type of supports. Consequently, the equivalent elastic stiffness KE is given for one-way elements in Table 3-8. The equivalent elastic deflection can then be calculated from Equation 3-36. Two-way elements generally exhibit two-' and three-step resistance-deflection curves which are a function of not only the type of supports but also of the aspect ratio L/H of the element." The' equivalent elastic deflection XE of the element under consideration must be calculated"from Equations' 3-34 and 3-35 for twoand three-step systems, respectively. TIle value of KE for the system can then be obtained from Equation '3-36. 3-14.3. ,Large Deflections When designing for a large deflection, it can be assumed without significant error that the resistance ri'ses instantaneously from zero to its ultimate value, r u at the onset of the blast loading thus neglecting the elastic and elasto-plastic ranges. The design resistance function f'or 'one-way elements ,is' approximated by a constant. plastic range resistance as shown in Figure 3-'40a, while··for. two-way elements, the resistance function becomes that shown in Figure 3-40b i where the ,maximum deflection lSn can be either. smaller or larger than the partial fail~re deflection
xi.
,.
3-15. Support
~hears
or Reactions
Support shears or reactions are a function the applied load and the maximum resistance attained by an element, its geometry and yield line location. However, for short duration blast loads, the support shears can be reasonably·
estimated by neglecting,the applied load. Therefore, the, ultimate support shear can be assumed to be' developed when the resistance reaches' the ultimate value, r u'
3-18
•
TM 5-l300/NAVFAC P-397/AFR 88-22
Equations for the ultimate support shears Vs for one-way elements are given in
Table 3~9.
For those cases where an element does not reach its ultimate
resistance, the support shears are obtained based upon elastic theory for the actual resista~ce r attained by the element.
For two-way elements, the ultimate shears acting at each section are calculated by use of the "yield line procedure" previously outlined for the determination of the ultimate resistance r u' The shear along the support is assumed to vary in the same manner as the mo~ent varies (2/3 V at the corners and V elsewhere) to account for the higher stiffness of the corners (Figure 341). Since the shear is assumed to be zero along the interior (usually positive) yield lines, the total shear at any section 'of a sector is equal to the resistance r u times the area between the
secti~n ~eing
considered and the
positive yield lines. Referring to Figure 3-41, the support shear Vs V for the triangular sector I is (2/3)V s V(L/4 + L/4) + VsV(L/2) - r u(Ly/2)
3-37
Vs V - 3r uy/5
3-38
and for the trapezoidal sector II (2/3)V s H(y/2) + VsH(H r u(L/2)[(H - y
y/2)
+ H)/2j
3-39 3-40
Values of the ultimate support shears Vs H and Vs V for several two-way elements
derived as above are presented in Tables 3-10 and 3-11.
Table 3-10 gives the
ultimate support shears for the case where opposite supports proyide the same
degree of restraint resulting in symmetrical yield line pattern~: Table 3-11 is for completely general situations where opposing supports provide different degrees of restraint and result in the formation of unsymmetrical yield lines. In these cases, the ultimate support shears are not equal at opposing supports.For the situations where the ultimate resistance of an element is not
attained, the support shears are less than the ultimate value. The proportion of the total load along each support as well as the d~stribution of the shears along the support is assumed to be the same as previously cited for the ultimate support shear. Therefore, the support shears. corresponding to the resistance r of the element is obtained from the equations presented in Tables 3-10 and 3-11 by replacing r u with the actual resistance attained (r e, rep' etc.).
3-19
ELASTIC RANGE
;{ II
.
ECASTO-PLASTIC RANGE PARTIAL FAILURE
FULLY PLASTIC
~
'u
/-
,AK
w u
z
«t
t-
'e
(f) (f)
w
w ,
'" 0
a::
'up
.
ep
INCIPIENT FAILURE
I
.
/
-
I
/}e Xe
. -
Xp
XI
Xu
DEFLECTION (Xl
Figure 3-1
Typical resistance-deflection function for two-way element
;-""',-. NEGATIVE YIELD LINES - - " ' POSITIVE YIELD LINES
(al
• 'or
(b)
FOUR EDGES FIXED
. (c)
(d)
THREE EDGES FIXED AND ONE EDGE FREE
(fl
(e]
'TWO ADJACENT EDGES FIXED AND TWO EDGES FREE )
Fi gure 3- 2
Idealized yield line locations for several two-way elements
3-21
L
\
I
~ ~
........
...
, ...'
Ito~ ~
ITT
I
.... ' ...
. . J.. ., .......... I
L-lI. O j-
]I ...
... ......
...
.....
I&ICI)
>1&1
-z
~ ,
1--
:I:
Cl Q
~
<..J
-=. ~~ !;t>= ,...£.i
.to~
M....
IIII
AT NEGATIVE YIELD LINES
tMVN
~J.J.J.""""'.u..~L.LJ"""''''
~ Ull
!Myp
AT POSlTlVE YIELD LINES
0) YIELD LINES AND DISTRIBUTION OF MOMENTS
b) FREE - BODY DIAGRAMS FOR INDIVIDUAL SECTORS
Figure 3-3
Determination of ultimate uni.t 3-22
~esistance
I
e
e
• 1.0
H
0.9
03 _J ~
----
0.8 0.7
.1
--
0.6
, ,
0.5 x/L
::I t-
...,,
0.4
,
: :I
-0
0.3
..., ec
, .
~
:
r-
1 • !'
-"'1"
l-
~:;
;",1
-0'1
t
1
0.2
:j:jttumtIll=Itt HtW H-HttrlHHlt1
~-+ -1-
-r-
:J:ft
,,
;i"t·
0.1
0.2
0.4
0.6
0.8 1.0
'i.:
,Mvp' 1/2 H (MI-4N+MHP):-
..
Figure 3C4
2
4
-
0.1 '
6
,
Location of yield lines for two-way element with two adjacent edges supported and two edges free (values of x)
"
8
10
10
I
09' 0.'1" 0.1 06
,
ts:JTI < y
IL
H
I,
2 r I III Willi 11111 III I ' , , , ,1111111111111111
1'1
, III
y/H
0.5
w ,
"" ~
~ ~ri
.. , ~I
02 0.1
MHP :0.25 MHN
. 'H~j
0.125
H
mm 1'_11~! 0.1
0.2
0.3
0.4
2
3
•
Figure 3-5
4
5
I·~:
Locafi~n of yield lines for ~wo-way eiement with two adjacent edges supp'orted and two edges free (values of y)
7
10
e
•
e 1.0 0.9
-....
0.8
,
r""'''''''''', l .. , !Illf1lm r'
0.6 .
;'
..-A,.... .... ,
'I
L
·T
j
_. --
.
,
~
I
y/H 0.5T "
,I
'
II
I'
I.
0.4;, I ,,
I
P
I+H+tt+tttHt+tlttt!+~U:litWllH
,I
0.3\ttHiitlr 0.2
",.
. ..
jI
'" '"
I I
:r
0.7
w ,
~
I~""
0.1 ~;Jl t ..
o
Tj ••.
0,1
0.2
0.4
0.6
0,8 L
1.0
2
4
.:
( MVN+M Vp) 1/2
HMHN +MH~
Figure 3-6
Location of symmetrical yield lines for two-way element with three edges supported and .one edge free' (ve l ues of y) . . . .. '.
6
8
10
; . • • t ,t ., ;.
I I"Ii···.. ',"':' '1"" .;:. -:..... I·· "~ i
-r--j:'dh!i::,~; iU::I! r ! 1 ,. . , .;
i
1II1
,,:,:."
~=~HN} + MHP11/~ 0.1 XI
LM HNI
~P-i
J
t MHP
I
Il1
of
H
III'! i i: :
. ,,, .. ,,,,,,,,,,,,,,, ..
..
1.0
.. ..
,,,,,, ..... ,, ,,.,.
/
[
II rili I';: 'I
® \
I
\ /
I , ,
..... , ..
,
L
./
.,"";
111111111111111!111111m1l111111111111
0.9
I
1.000
0.8 I
-
0.7
I
; 1
,
'-' ;v
xl--l
1111Ii1111l111l!
,:::
0.6
I 1. 1 :. I : .
0.5
I III!: ~',
Co
0.4
,
I I11: , "
I
!
'"
""~'II' .:l:::.":::.:::':":":.:::::i:::!:::::::":. I d""'l"!"'l''':'''':'''~'' .'." : '.' "
I " i .
1
I I
0.3 ~
I .: . ,
:~::
0.2
11111111111111
II"
!
.
I
".....
.'
I""
: : ..
, . . '.,
,
:;:::;:~~
'.
'1"
0.125
~:'; ;;: .;. :-.u': '," '"
~!~
.I :!
oi' - ':
0.1
1
I"
02"0
..
'.II,I,·,···· ......······, .. ,,·!,·,.··:
II'
.
-- t ' : ' . . . . ,
.
~
~
1'.11'' ..
I
I
,II,lJ
0.500
'1"
•"
;
'I"
"....
...';-~'"rt r;: , :..,."";::Jr.....
"'::+-'l'"....""""..;
. 0 .I
"II ' .. I. ..., ..•;. ' I:-.
.
0.2 :. '
0.5
0.8 1.0
H
Figure 3-7
,"" : I' ,~
5
2 L [
(.
111111:111"" 1
. .;,. .~;
8
10
20
Mvp' JI/2
~HNltMH~
Location of unsymmetrical yield lines for two-way element with three edges supported and one edqe free {X?IXl~~.11
e
e
e
~ J!"HN3 • XI
~HNI
MHP] 1;2 t MHP
03 •
P1
P-i
0/
\G)
H
[
/ I
® \
\ \ /
I.
_[X
L
1.011111111111111111111111':11
0.9,__ ---_::...=..-
0.8 0.7
0.6
,
w
..... '"
xl..J
0.5
0.4 co
0.3
0.2
0.1 0
0_1
-0;2 -'
2 L [
20 MVp ]1/2'
H I!AHNI+MHPJ
Fi gure 3--8
Location of unsymmetrical yield lines for two-way element with three edges .supported and onee dqe f.ree (X2/Xl=O;3)
.1
~ "[M NN} • XI
l '!20.5 J
~
MNP lMNNI t MNP
A \0
(J), H
[
/
@
\
I
\
, /
I
.1
L
...
LO 0.9
.: ·F
--r-
.
0.8 . 0.7 -
-r-
.. +.
0.6
w
•
'" '"
xl..J
0.5 0.4 0.3
0.2 0.1 0 0.1
0.2
0.5
2
20
L [ H
Fi gure 3-9
MVI'
JI/2
..
~HNlt"~
Location of'unsymmetrical yield lines for two-way element with three edges supported and one edqe free (X2/Xl"O.5)
60
80 100
..
e , ,; I, ,
.
I
.
I
I
I
-Xz= U"'HNl +,
I I
XI
",li"I'"" -i-'-r -j-;-rr TITt1
,:'I
, . 1
1
.,
"'HNI
t
"'HPJ 1/2= 0.75 .
''''HP
. t
IIrt
I' j'
A
A
0/
\0
H
,
[
1-L-!J1JI
/
I'
'" "? 1:11-11 n
.-."
11'4.000
''I ' 'j'l
• • • '.
I
0,1
•
J:Ll - 8.000
o. ~ , ,L L.; ,I :""1' II'i ,,"i' 0.1 l:-,:';::.c-:, -4 -\ ~ : :
\ /
w ,
iWil o ':: I
® \
I
-------,
"")'
•
i.::
t
0,2
~':' ifm
• '''U:lill;
Iii ;I I, I
.II!
I " ' T . ' 1'1
!·"I'.·I'j'f!1I1"·I!'!!IIfJ!'1I1
0,5
I
I
0.8,1.0
5 , L
H Fi(lure 3-10
l1J!l11\1 1111, i "I" r I!' -IT· - ,-
I I I ! ! ! ' Ill!!!I!!!!1
~
:
8
10
20
MVp,':'~: 1/2
MHN1tMHP
"
Location of unsymmetrical yield 1 ines for twc -waye l emen t with three 'edges supported and one edge free (X2/Xl=b.7~) . -' ......
,L
,,:1·
<,: I!I~!~
,111~1I1·!illlillll.,liltllll!'·lli t~;li ili ~=r!"HN5+MHPJI/2= 1.0 l- :_:_. _' , .!l :" j 'j;jl ! , liil: -1--+ . _:.1.:,. XI ~HNI + MHP ' , 111 'I' ;'111m, 1 I I I lil , ,"III If,!", I 111 "I 'I ,11- I I" 'r j.::III':: 1,j' I uu- 'i.'i1. I "I,·, I I·: ,~-;-:: 'I';'I;!'I : II'I' I!I II I!I '" II' 'I I : ; I TTTl, I i II ITT, ' I I ill 'I!I ''IiIi, IIIi1'1' j . II!' I I, I i : " 1[Ii I 1"11,, I II I " ,·II III1,I1, - I-+ · -, I I -',""'·,·'·!I·ll"·,"';'·'·;'d!r'l I"" ! I' 'II '" I d;""liI 11 1'11 d'lll 'I ~ ~ 'jIll II I I I I i I"I ' I, I .II,:. '~)l ~! h 10 :.: ::! Ii ill ,jill L:: !::~.;~I,~i~11 ~i; il:: :i ill:! J-,. L+ t;JJ III 'I
." ","
t I
1,1'
lI"l l iI, ·
""
l
,
'II,
I
dlll l'
"1"'I1
I
1
I, , " ,I
!
'
•
;
I!
i!
!
ill'
'I'j rPII11
Ii
l I 11 I I'
'II
~:
1
j'
I'
I'
~I
l
1'1,11:;
I
unuuuu.rrrriru
II
II I
"
I
1
I
[
1111
11
/,
,'5'
'1'/1
I..!J I
H
I
nnnuuun
1
~I
I
I
t:\
•
II!I;!!'!'!
1
1
:'01'1'1 "'1"1''''''1'1 )',,11:,1111,,',,: "'I"'" ',",' II' '" '"~ ,:",',i,:,I,,"I"'!,,1
I",
,1
":'
I
"
't"
I!I, I
I
I-Ii; [: 1
Co> Co>
0
l
I
t
'
I
t
t
II t ,.
: ''1-'I' IIII II I I it rr '! :'1' ,r-,I,I I I 'i' ':1:"iTl'l,I'ii I'I[ Ii·''i ir!1I I" i i !I xl..-J 0,5 li'I', :T1',':ll'l ll li' ill1 ::ir,i~: :J:111,111 l'jnlli"ilillinl'l~f'l;N:- - M ' 111 l " 1' " "'I '"" "', , I ,. , ' [ l I~ " I, ", I:" . I'll ' , 0.4 I~"f-:-L' l , ' ., I 0.6 -~Lj--",! I : : i ,j
I ,
I
0.3
,I I' ,
I
I
+-_
$',ml'
'II' , " .tn 'liit 10' "~ I, I ll,I ! ill:ld' 'li!!Iiillll' 11,lllli! II)
i
1
II'
• 111'lil : i'" "I'l
['i'Il.:.:.: '
I
iii'
'1
li'l ! 1'1
I
i
',I: L .. , I~'· '!~"t~i:':;Ii':' iI'
II, ,! 0.2 '-i"-:---"T~1f Ill :~~rtt::1
,Iii :ill "I'
"...,
L.J::
:;':·:'1: I 0 ~ooI" "·
I'i·i
M
" I
-
111- 2.000 I 1;1
'I '"
_.
-
1-4.000 ::.
i l ' : 1 "II~ IIi!
,
,
-
"
i
1 : , ; 1,
J
0.1
0.2
0.5
0.8 1.0
,
Itt
II
·11
I'
L [
5· 8
Mvp "
.10
3~11
ill i
+ttHtt IItllll
m I I
urn
i
~I ~
",-".., l~
20
J1/2 ,
'H ~HNI+MH~ Figure
'II,I'!
I
III II II III
,':,' m 2
#
:1-1- . 1 :
IllHlliL-e.ooo ,
II III! Iii
I
1 11
, " ,,' ,...-co, Ili, i:1 :: :i:li':I::III: 1111:1:llli:II':; i,; "'I' IIIIH'I'~!" ".l!I, , ,,'~f:lji~ ~ !:, I! ' I; ',' "''''":: ' ' 1:' ,I!:I, Ii' 11,111, i.i I, : 1 'I, 1'1".11' 'I i,lS', 1tl1, ..d lttH-;Ii" I'j"!: "":;:"''''1,,''1'':'''''' I ~'I _. " -"- 1 "-/j. I tr"ft' :-1',l.J ,ttMc'1'''''''''r-+. ' 0.1 --r"t-' i 1-'-I·t' fT·l'i""J;, " 'r" -'1 -•••-. ··t t !--'rl,ll "11 ,.- _.•-. •• r-.I':T., .: I ' l l : I'" il"'!:';::"';::II",'I,'IIIII",II'" I ,1'1'1 : ' 1 "1' ; Ii' I ' I II ~'~r:" ... tJ.,;LT"t-i-,~ 1."1 ,I /'1 ,j,1 1111!11 1111"""" 1--,........... ~J, , 11' ;' I'jil I ill-'l;·"''''''''I'·" j!I·:!i-III!-"~"- \ill:'·I'I::,I.I, j · i · 1 I:!!,II: 11 11:11' II 'l:ijj';"111 r-rn OUI' ",.;,"'I""I""""':II",I:I,:,.. ,!: ,.: i :1i1"I"III:I.l1 ii ;'111':11' 11111'
.t. I I' ,
<;1._."
1 illl!1'
1
=::.':- :
II
. - - ,.
q, ',!WII 'III"
ililll !ill!!I!
.. -- --" I' --I·
l- 1.000
II
I
-'1:1::\
I
VN2
1
~I
I.I.!! J i-.; ., !;:: I!,H] I,14~ 1 L O.12~ -r,'l- " . I"'IIJII lli"i'l ll :;TI:,I',iI'l i,,: l iI IIIH ii Tlli O.2~O ~, I"" i', ',:;' 1'1 11 11: I " .. I: I "" :'" 'if' "'i1 11,11, '"
M' --..y!...
1
I
/
•
I'l'
I"
I
I
L:..._....
:I,rr:tlilidl II'iI:1:, 1 '~'~It 1 1~lilil !II'I~I 1 '1111'I"I-l-lllfl~'1rrl TI1lrrrrm 1 II'i! :,',II',i!,",'I: " I1,1',' I :["''''''''''1'1'1' i""I,I: 1'1'1"11'1"11111"1111'1 11"' I'",lill"!II,j'II'I!II! 111"11111"1111,111 '1 [1 'Iii' --' 1' 11/ 11 ;'1, '~'It· _. . .I I, 1,,1i 0,9 1 1 ,," 'Iill: :;i:1 1I'::lii,'''I:I:I!!i'J'jll'li',;!II''I'I'! ,.;. 1,1 1- l;j'll!liIJ'I:llil"IIIH'I\'1 'III' i 1 I ,i j 1 l 1 -~ 1 ,1,,1111' II"'" ""I'i 'I' 1I'I"I"I'lij'IJ I, ~"Ill 1 ' II,I ,~I '~II I i 'j '~'IIL'I" i! !',,': ' :"'II' l'I'll i i ' I \"1 Ili'"1I'I'ill!!111111 Ii'1 I : I: I"! i,: ;;';','1" I!, . 1'1 1 ,t III lIi I ! +i IHI11!' I i111 11'11 , 0.8 i-ii :-:1(1 :IF:lli' i'll' dF~n'; 1';1 ! I] :HTIi', ;I~ r[ -- ~"T~ I II 1'1 I,1i'1111111111111 IHIIII'I'I'1I I [lil!111 . 11I I I " ' : 11; ,tlll!!I' 'I dill "ill" II 11'11 i,I" I ~II 11/1- - iii: 1!!::!!;;iliJ;;~,l ji.ililii i!II).·~ 111111'1111 i I!IIII,III.:.:' 11, I' 0.7 ',! i::, 1 ll oI i: ;;1 1 11 :111 :ii'II:: l ;:i'li!I,II! '1'1:11111 !:! 1'1 I I I, 1I11 I1 - un ), I
i",'
,~
\!I,
Location of unsymmetrical ..yi.eld·l ines for two-way element with· tnree edges supported and one edge free (X2/X~=LO)
50
80100
,I.
~ =~HNS + MHPlI/~
,II'
XI
, , . ,1' , -r-r-r- -r-i- -i-qmf ,
,
"
~HNI
t
MH;J
~
1.25
o»
"I
[
H
/
/
~ \0 ® \ \
/
I
L
I
w
c., ;-
4,oo~0'11 lin:" -.
I I I \
~-I ·8.000 .j
-~.
i
~
til· 11. 1
',lmjll,liJJ[\',
"I iill, I
.>
en
i· : .
.
----_
..,
.
0.2
80 100
L [
Nvp,
JI/2
Ii ~HNltMH~ Figurk 3-12
, Location of unsymmetrical yield lines for two-way element with three.edges supported and one edge free (X2/X1=1.25)
-Xz= ~MHN3. XI
~
MHPJ 1/2 = 15 MHN1 t "'HP .
A
' [ CD/' H
\0
® \
/
I. I
I II;: I
0.81 'i
',I
0.7
O.GHI ' Illlllllllilll!dll:I'11l1
W I
W
ec
xl..J ! 1 ' : !
t
, ,
., •
!
~
~
I , ' II
--
. I.U va -_.-
P ~
1!1!!illl!!!;!!I!!!!I!!!!!!!!! 1I!l1!1I111 . i 1!,"Jl 2
L ,I
" ' Mvp I "'~1/2
~
~~ MHNltMHP '
Figure 3-13
.4J:111 Rl!llJ1i11ll
..
toce t ton of unsymmetrical yield lines for two-way element with. three edges supported and one edge free (X2/11=1.5)
" u
L
.•
r
e
·:·tir~f-+
1111 I I
I I jilll
--
e
- --,
~ JMHN~ + MHPl i/2 = I. 75
LLJ-+
XI
If,IHNI t MHpJ
~
A
CD/
\0
H
[
® \
/ I
l
,
\ /
I
L
1.0 0.9
Ii'
0.8 0.7
'", '" '"
0.6
, mmf .runn II
. , iii i ;
X-I-l 0.5 : : . ; -
__
0.4
0.3
-r
__I_
!
iii
i,1 II:; . ll I Ii
--
ILUlUllllllllllilllllWll1II
M vp' M VN2
0.125
-
-
0.2
4,000
I t t ! 1 I I'
1
0·11-:1 ' ' " , I
I' , ,
o ui: 1
2
0.1
Ii
L [ H
Figure 3-14
MVp
20
JI/2
~HNltMH~
Location of unsymmetrical yield line~ for two-way element with edges supported and one edge free (X2/Xl • 1.75)
~hree
50
80 100
ii' " jJ ! ~ I'"
I'
, I
-
I!-'T.' '1 ,I
I! I
.
I
1.0
I·
I
I'
I r
II
:I";!,[,, I
il'!
GI
11 ,,'1'I"~1• [Ijl!il~;; I ::1 'II Ii", '1
'
t
JI :[I
.
,'1"'11' I, iI III":1'.' I IIIII, I,I"
Iii '.' ""I '1'1,, 1'1'
.
' Ii
II
IiII'
-X2= ~MHN3 + MHP] 1/2= 20
I', I' i ; ''
I 1- t I+-~,I Iitil '
1j
.. ... .
~I
II
XI
MHNI tM HP '
[
H
.. . I.
ii' I III1 illl i il ,
/ I
\0 (i) \
, /
I.
I
I
A
o»
I'
!I~l ~ ~ ~,i I' I~' I
i jj il j nili'll i
~
•
II
I
L
1lllmt~lm
0.9 '0.8
!I111111111111111U1l1l1llt1
,
0.7 0.6
w , w
...
xl.J
0.5 0.4
'1. .,.
0.2
0.5
0.8 1.0
2
5
L [ H
Figure 3-15
Mvp
8
10
20
JI12
~HNltMH~
Location of unsymmetrical yield lines for two-way element with three edges supported and one edge free (X2/Xj = 2.0)
50
80100
e
•
lll~
',. X -= L
~
MHN1 + M HP l/2 MHN2 + M HP
[
I+GMHN1 + MHPJ 1/2 M HN2 + M HP
I
x
'I
01
/J.-
• I
® ~
-: (!) "<,
y /
I
1 _LLLLLLLLLLLU_I U.111 LLLLLLJ..LL1_1LLILlUl UlIllIllllI
I
L
1.0
O.8
~+l+lliHtHttlttf-fflilil:l:lffit1mtJjmmr
"",
"" '"
>-II
I
-! .-
O.4ffijt. 0.2FFIIIl1FRIIIIIfIIlIIII FIlTIITIIIflIIlIlIlIlIIIlIIi
"
.02
.05 ,~
,
H
Figure 3-16
~
L'
.08.1' . .
.5
.2 . '.,' ~
" .:' I/~ (M VN3 + MVp) "
". 1/2' . ' 1/2 (MHN2+MHP) qMHNI+MHP),
..
, .'
.
Location of unsymm~trical yield lines for two-way element with three edges supported and one edge free (values of y)
,.~.;:
05 ~
••
0.4
,... »> ......
.................
I~ ~ -r
0.3,T
w
,
,,
02
'" O. I
T -t T;'ld H1\
04
/
r---< ". '" ""
L
'j'mh:rrmml.mm;~~...J "! . ---
Y/ H w ,
~I','
i ,.. "....... .... ..... ....
I
0.3
L
-I
)(/L
il II j , "., J '
+11,1 iI I ~ j \ . , , ' J i; u t rI . C ill T~. i :j:. "i_H++++H I :l , tIl!
nTl ~W-li:
I
I lfltlfillli
~mmii :1,1:1" i1Wl
I.:: ll,
Wlm~
0.1
0.2
UIlHlIlIlIlHilllj tttttttttItJrtt
0.2
,:HH!m~_ 0.3
0.4
I+H+++ttlill
T
0,1
.r Illllillftjj
-. ,HHHttHHH
0.6
2
0.8 1.0
-'=- ( MYN+ Myp) H
Figure 3-17
1 _
MH~"
3
4
.6
8
10
1/2
MHP
, Location of symmetrical yield lines for two-way element with four edges suppqrted
e
o
e 0.7
r-:-;--:-;- ::- Til: "
.. , i I
t
il l iii"
::.; i i il
0.6
l ii
· I" III ,I II" i I:
·i_It!" ~'I
I
:1'1'1
X,
Ii' Ilil 1I Ifilii'
~J--~~3
-: II 0.5
X2
~ X2 : ,
I:
/ MtlNI + MtlP MtlN2 t MtlP
i
-
i 1Tfi
- .-
0.4 \ '.I "I i
I'
~
I
X'/L
MVNlt,MI,'p .. MVN2tMvp
= -':;'--';:=M=VN='=t:;:;M;::V:=P
". '.,+ 03
·
.
MVN2tMvp
,., 0.5
:I · ,
...,, ...,
J
1-,
..... 0.2
01
0.1
0.3
0'4
0.5
0.7
1.0'
2
'~(jMVNI t Mvp. +'/MVN2 t MVP) .': .' )5 (MtlNI + MtlP). . Figure 3-18
3
4
5678910 .''''
Location of -unsymme t r i ca l yield lines fo r two-way element wHh four edges supported (values of X))
"
0.7 ,
L I~CD.=--------'i
"+/
06
"", [: I:
Y,
/®" j r W
Y2
0.5
I
!.!.., 0.4
J MVN2 t Mvp,
X
IMHNI + MHP j MHN2+MHP
Hill;+)
I
'T
II I I 1I ~i1~1 ~I ~il !l!ili~I " " ",I!i , II ,II Ii • 1'1'1'
!I
T'
H
I
0,3
a>
'II
i
(MVNI + Mvp
Y2
I
Y'/
~
'I' I'1IiII'''I''I'I;';'~I;'l'''I'''~':~'''1 ':"'Ijl~' · :; : IliII':: :i ,'1 i i: :,: i!': :1:' r j;j,III' j i.1i Iii: III I ~I'
i jiill"'" ....,'""
HI I \.VI~ f.\1r.\
I
!
::
0.2
t
jMHNI+MHP, MHN2 t MHP
[i:. '1": ;:'.i!, ~ :::i:: ::
ill''
. I' ,' I
I
..: ,HI'[11 If!' , III i~ lf i ~i~~il~~~~~tHm~#H+lH~~ I, l ; : lill'i I III I :il lili l lilil l ,ih:J,;:t:JH lilillililililillilliii 'ft.,' II
'
II
'
jt I
!'
0.1 ,
, , I
H+++f+T++JH J++f+++H;
-_.'-j
01
0.2
0.3
'0.6
5
1:....= H ,
Figure 3-19
0.7
,
j
1.0
1< Mv. +
2
3
4
MVP)
MHNI+ MHP T /MHN2
+- MHP
,
Location of unsymmetrical yield lines for two-way element with four edges supported (values of VI)
5
6
7 8
10
e
e
•
I
+ MHP MHN2 + MHP
MHNI (
\
'12
)
~I ' ' 11 H+++HHH :i
.,'1 11·1
.•
I •.
0.6
L -x
0.5
'",
'" '"
0.4
111111!]!'IIII· !Ii 'j :fljil; I.i :Willl'nllllllllllll'I'I'II'I!iiili'~~i':IIII"I!i':'i!I'i'"Iii -''I' , , 1'" c'ri I fill 'm;' ,~ ,~ ., .. 'I :"11' I' I' 111 ' illll,,!"'II' 1:·lrr.lll·f• •frrrrrrl\\lrrlllll::1 'I.I ,IH:, " I---:-:-:,:-~LL~;P;-::!!il-:-:';-;-~-H~ ':'" r' r' IHH""I'jIlIT'~"II"'IIII'uSI ~I"II'! JJJ UWllillllj ~ II t ~l !! Il:!lII;~ Il! Ip<'::!'I:~I'
I::: . --
.1 '.
.,
"
t·
,
-ii',
,i II
'II
. ,'. .
,I
t
;,,
J
II " I If I
!
II
1
'oil 11'I1"'Ii' 1111:,II
I:
111,11
.-, -.
""I' , :
_
i I
0.3
I
TI
0.2
01
0.2
0.3
0.4
05 0,6
0.8
1.0
2
.(MVNI + MVp)I/2 MVN2
Figure 3-20
+
Mvp
Location of unsymmetrical yield lines for two-way element with four edges supported (values of x/L and Y/H)
10
'!.. H
3-40
B
,
I
,
....
........
'........
..... ..... ,
.....
lA » > -> ..:::t -:::. ,- - - - ¥ - - y o .....
,
........
L !,
~ B·
ELEVATION
/~
--,.
x
E
x
H
SECTION A-A
SECTION B-B
PARTIAL FAILURE - - INCIPIENT FAILURE
Fi gure 3-22
Deflection of two-way element
3-41
1JCJ' L FIG. 3-24
.1
L
1
FIG. 3-25
FIG. 3-26
FIG. 3-28
FIG. 3-29
2
.1 FIG. 3-27
.' L FIG. 3-30
FIG. 3-32
FIG. 3-31
{JO]l.
,
I 'L
,1
•
FfG. 3-33
FIG. 3-,34
FIG. 3-35
FIG. 3-37
FIG. 3-38
xDc:=J I,
.J
L FIG. 3-36
LEGEND: EDGE CONDITIONS
t
r<;"
~
1
FREE
Figure 3-23
<
4
FIXED
Graphical summary of two-way elements 3-42
.................
_-.- -_.
··f ..
'-'-'-'--,-1
-: . ..•...._....:::....: .. ;..;.:..:.::._ ........... ......
t.:...:.-==-==-----.-----
.e- .007 r en. .005 .e.
..
_.-.. _
..;
"
; .......•..
,
>
ext
__ . _ . . _.._. '-C-_'_ . ~':-""~-'-'-
.... ;.~:....:..:..:.....:-.:.....:....::....:.:.~,;::~:.:....:::..:;.. .:..~ .. ..:...
.. -.-.
-.-
... --..
_
__
..
. . -._----------------_._-.------... - . ... .....
........... -
,"-
..-,.-'--'--f 1----.----.0007 01 ~Hmax
.0005 I...-..:..-{
"
....•...•....•.
-
M = XD =
.0003 .0002 t
i-=
.0001
·10 7.0 5.0
3.0 2.0
1.0 0.7 0.5
L 0.3 0.2
0.1
H/L
Flgure 3-24
Moment and deflection coefficients for uniformly-loaded, two-way element with·two adjacent edges fixed andlwo edges free 3-43
L
·1
-
-:::,:-::::::-,,::::'::::;::::::,;:.~,;:"::;:;::
= =
M XD
... , :::::::::: :..:
:"':.':
.-_.- :.•-...... Ir--_~ ...•.. :.•. -- •...•.••.• - - - ._. __ .__ .,...... •..•.•..•.•• -+ •• -_ ••.• -~,-_
~-
, -:::::::::t::::==::::::::::~:~::: ::::: :::::, -' ...:-----~_._.-' - ' . - '_ :
•
..e..
> Q:1.
...
1..
. _...,...
~_
.__.__ . __
..07----·-· ...".·; ..; ... e . O~ C-: .L:: ..:J':±"S2EJI· :z:
. .: _.----.---_ .. _.
." - .. _:- ....:,;~:;~~~~-'.~: I
.003 .
.
~...
. : ...
:
__ :
._._~---'--
. .:... .
. :··_···-:-.:...-....:i---·-.---
.002 ;-
........ .. . i
I .001 .I'...: 10 7.0 5.0
---'-
....;
3.0 2.0
C,I
H/L
Fig!Jre 3-25
Moment and deflection coefficients for uniformly-loaded; two-way element with one edge fixed, an adjacent edge.' simply-supported'and two edges free' 3-44
,. M XD
L
"2
= .B r H 4 = yrt-i
1.0 t:-:-~:±Z:'
07 "
.. >·.~:ijjji!ii~~IU
~ '.. : .,.e".
i;~.;~.
• (: .... ':,.~.:,. -:::;:;;:~::;.;;-;: =':. 0.5;,~ ·····,:;"i::~i::;:f:~:,;
. at . a E
~~:::::.:::
..
~.3 ~:~~~:~~~~~!llrrr: ~~_=j::: .BH:
0.2 ~;=~ji~~~:ii~;~i: ~._,
--:c_.
a E
•
-".
. ,,,~;:.:'::-':-:;:..::
-- "----,.--··t---· ------r---. --,~--
0.1 ; - ~i::::::.=--.--'~7--·l;~i: '" .._..._...:......-
. :07"~'
:.....:.:•....1:.:
'..• '
..
' "'''p
.-
--_.
-'-.-. "____ r__ ".---., ... __ ,._. __... _----..,..-.,--~--,_.
.-'-'- - ..:'::----. C,.'.. - " - ' - ' - - _ ' _
:.._:~. ;;.;;?~~ -~~:'::.~~: -. ,-._·-L"':':~".-"-'---l
.Oll '-:-T~L0
cl
,'~=,;.;,..;..;;;;;;;;;.;;;,;;,;.;.;.;,,;,;.~
..:...L .. ~,_ ...:
.....•. .. . . ........... .............. .. _...
.01
i
.005 ' I 003:I '
__ j, , .. ,_
..,
.
.
L.
••• __ .0
.
•
... --, .. _.
-_ _ -- _ ..-~':.~.-- .-:-~..-:.--....
_-_._-_
_.' --'
.007 >' 1-":'#"-"- H
•
,
'-:--~~-
i
..
•
"
'.
--:
'.
~_
- ,.., ••
-.oo--"---.- - -
.:.::::;..~.:'::_i..:.:.:: .
~--t
.......-.:'-'-'-.-" ~ ....
3.0 2.9-
1.0 0.7 0.5
. -....._.. __ ... ~- .:. ----
0.3 0.2
H/L F~ure
3-26
Moment· and def1 ecti on coeffi ci ents for un i form1 y-1 oaded, two-way element witn'two adjacent edges simply-supported and two edges free' 3-45
1.0
, . , ., . ," ....... _-_....._ ......,._-,.,.t-:
2
0.7
ZI
0.5 0.3
. I·
0.2
M 0.1
~
3
=
XD =
L
' 2 ,BrH yr'H 4
_._-+:..:..,.,.:
L._ _ _.
YI
•
I',·
...
c:Q.
.0003:' ! ,
.._..._.__.__ .. ,.
.0002 ,-' ,C.cc.'-t-r-r-r-t-r-r--]:
+_._.~---
.
1-:'-: I: '
.0001 .'..."'--, 10 7.0 5.0'
-1..
-'
3.02.0
1.0 0.7 0.5
0.3
0.2
,0,1
H/L
'Fi gtl re 3- 27
Moment and deflection coefficients for uniformly-loaded, 'two-way element with three'edges fixed and one edge free
3·46 .
1.0,,""--'--....:.------,.....-,-,...-...,--,-..,....,-7""1
0.7 0.5
0.3 0.2+----:---~--+~j"'==/_:,-:.,."=~7'_'_'__;_j
0.1
.0 2=~, _ __ .
·_·_-t- --------
-
ext .007 (=-" .005
'., ·f-~J~+++,.:-I
H·Ll-f:Fi c~
1····.003
~-=-. ~
.:~ r ' ,
. •. .
.0007! .; .0005
.... :..
._..
/-....;,....L 2.__ ~j._l
.
---I
--~---.........,._,_----i
1
i
.0003 . .
I
.0002 :
••:
---~L...,
,.j,c.:_L~'." . . 0.0/.
.0001
1,-0.0_00.
.........
10 7.0 5.0
.
":.:-: . . :
3.0 2.0
. • _ _ 0•••.• _ ••
'_1 I
,: -,...._ _--.J!
1
1.0 0.70.5
0.3 0.2
0.1
H/L
Fcigure 3-28
Moment 'and deflection coefficients for uniformly-loaded, two~way element with two opposite edges fixed, one· edge simply-supported and one edge free 3·47
1.0,..-..-------..,.-......,."-,....,.--"..-,...,....,.,,,...,..,..,
0.7 ~
M = XD
.05
f' ".. --.: ":" t"
•. - . - -
= .. ..
"T-··--··~::-·
~-,'h~·"··
.03l~~~~~;~~~:~.. .iLi~" .02
":::::.::::::=~::::::=:::::
.
..
.
.. .
.
~ .~:~~~~f~j!"~E0~2~~~¥~lj .005' ,.. ,>/1-__-" I
..
. '.
,
.
, __ .;.:__ , L.
._'.:,'.: ::::::::::::::;;::" -. -'._,-.',.. .003 t,:.·.·.·.'. .T ,.. ',:.-
r :
.002
f
.
-:
.
",_
,
.... ,.,.
.
.
.~~I:Gjjt0t~~c",c-'c-~~,~~~;~ ..:__,:;.
_-_
-_._-._.---_..
"'_'-== -_.
-
.001 •!~.~-----
'
....
I.
;
::..:::_'-:..:.',:'. .
-
.
__. ..
..:-.- ---- .... --'- .. _-.----;---
~._--
_._-,_._--------_<._-.-'-
.
.0003· .0002--'-":"'-.-·1--=-=-=-'--··......--..- - - - - - -..-
.I· I
.
.000 I 10- .--::-:-:~-L..:-~:=:.:.:.:.:.:.:...:..::..::.:.:.:...:.::........:....:....:....:..:.::........:....:...:......J 705.0 3.0 2.0 1.0 0.70.5 0.3 0.2 0.1 H/L
'Fi gure 3-29
Moment and deflection coefficients for uniformly-Joaded. two-way element with two opposite edges simply-supported. one edge fixed. and one edge free 3-48
1.0 r---------.--,....,....,.-,....-.....,.,.-'"""""?---.., -. '_ ... .--
10 ~!,~~-,. _.;'""",-_._,--_._,----...,..~
0.7 0.5 0.3
... -".:~~
L
...... :-.
0.2+.,...--------+~:c/,~J'--:c:c"""'_~~~
M =
XD_ =
0.1
_--_ ...-.
"--"-'--"~-""--.
....
-
-.-~
-"-+--;--'----'--'-1
,.... :..~,._~:::. i· -: ..; ...
.01
----_.- ' .~._,.....:..;._--_:..-..:._.;.. _---_.
..
,....-,-~.,....---_:.-:-..: _~----
:~: [:~.-;;'-:='+L~~" Ij:: '-.2.--.:.;'-;'"-::.. . . .+;:-'-•.-'-.e.:.,- JH--"--'-'--'-,'-.-'---:-:-'.-'-. ... . ..
.:':
:"
,: , . :
_--_.-.-.....-~--
.003
i •.
.002
t
~,:
r'.
:..:': .. J.. p.8--+--;i-"""S-'- f,-;.c.-;-i-""--'---+--;--;..-+--;.,.,-'-+--;-'---'--'-1 .'
- <>:
. _.
::: ::::::.:':::":::' : .
• __ ..._ ::::.::::.-c..:-===cc.j
- ._ .. _ _-_._...__._
.001
.0007' .:.-.--.----:•.-.. :: .. . .--:-..,---- .-----.---.-----i·· .. ·.....:-.--~-.:-__+'· ----..----- .---~-.. --~.0005-····--- - ..-.-
;0003.
.
.._----_...
.0002-----'--.-- I--'-='-'-='--
• :' •.... I ._--_.. _- ..__.-:. ... _...:_----_._-~,
.0 .
.0001 10 7.0 5.0
3.0 2.0
1.0 0.7 0.5
0.3 0.2
H/L Figure 3c 30
Moment and deflection coefficients for uniformly-loaded. two-way element with three edges simply-supported and one . edge ,free
I
a
.005 ... ;:::;:
ext'" .001 .0007
e
.0005 .0003 .0002
!OOOI '--;-----.
.00007 .00005
.0002 .000 I
'-"-.'1
"r- .--.'.C
.=------1
::... ::.::::: ..
l:ut----u.C··-···:··~t--~
10 7.0 5.0
3.0 2.0
.
1.0 0.7 0.5
.00002
u_...._ '1.000 0.3 0.2
I 0
0.1
H/L
Fi,gure 3-31
Moment and deflection coefficients for uniformly-loaded, two-way element with two adjacent edges fixed, one edge simply-supported, ar.d one edge free 3-50
0.3
~~
0.2+----..,.----E'1
.03
F"'":"2~-:-:';'-:; .02
..
.
...
.003
~'--"'.~
(~'~'~~E-"EE'~-~-~ .001 .0007
.01..,- .---, ' . . - - - -..- - . ..-f-'c~-.00,7 '-----;-, --I ' --0-;:--
.005 l.::,~, :..,~/-7:..:..· ,i
;.
'. ' : . , . '
,
,
-. .0005
.000,3 .0002
'.'
.0001 .0007", - - , ....----.,(-'+......,... -.-..----~---._---
.oooe
/--- - - - - _ .
:", '.. ' -~_. L.~; ~-,
.0003
.~_."'-'-'-.:_...;._-,' - " - - ' ,
.00007 .00005
/, - : '
----:--L:.'.,' :,:...~ .
.00003 .00002.
7.0 5.0
3.0 2.0
0.3 .0.2 \
H/L Fi gure 3-32
Moment and deflection coefficients for unt f'orml y-Toaded , two-way elo.ment with two adjacent edges simply-supported, one edge f ixed, and one edge free ~ 3-51r
·e .003 .002
.007.··· .005
..... . _..
;.c;,,<' '.
_,
.
::::::::::::
...
...... y.
.0001 >
err·0 0 0 7
I,,-PP •.""": '-~-' ...:.:-: ..~-+-P~P-/·.00007 , ... ;".::' ',,,::;:: -;' :;:; ..: ','
i .0005'
CQ.
~
;
Jbr-'"T'"'-'-'-'-'-'--o--+-';--'-"-'--i-++-'--'~---J.00005
....
h-:.
.0003 .-'-7"-··/~~"-=Ib+~~~~~~~C-+-C-+-~~-'-I.00003' .0002-'- --'-."'"'=~-r:--',=~'-'-'-'-~-7'-~-,;--'-~-7'-'-=.'~~ . .00002 ... " . .. . . ..
...
.
_._----_._--_._----_.~~
.00007' _
2 M = .BrH :XD = )'rH 4
; ..
.00005
J
~00003
:I:
.00002 ....... -............ _---_
.00001 10 7,0 3.0
---
. 3.0 '2.0
.
_-
( ( l
j.
.,2
t.((( <
3
2"-
L
.,
..LA. 77,,',;,) '3
1.00.7 0.5
~
.•.1
F
0.3 0.2
H/L
. Figure 3-33
Moment and deflectiori'coeffic~ents for uniformly-loaded, two-way element w{th,a 11 edges fi xed·
. ....
"
..
·,
.00003 .00002 :.
.0003
.00000.3
.0002
,i .00000. 2
.0001
., -,
-
107.05.0
L
1.0 0.70.5 - 0.3 Q.2
-I
i
.000001
··,0.1,
H/L fi gure 3-34
Moment "arid 'deflection ccef f ic i en t s for uhiformly-loaded, two-way element with two opposite -edqe s fixed and two edges s impl y-supported . , 3-53
1.0,.-------,..-,--,----,..,,,......,;.,....--,-...,.,.,--, .0 I
0.7 0.5
T :I:
. I·
I
'
XD=
.005 .
3
~
M =
.007 . 2...
D H 2.. )" r H'4
}.Jr
.001 """".......~:.,., .0007
0.1., C._'_''--'-'--'-.' ...... I.
....... :::
." _ .
.0005
.03
N
H/L Fi gure 3- 35
Moment'and deflection. coefficients for uniform1y-loaded, two-way element with three edges fixed and ,one edge slmplysupported
..
3-54'
.03
0.1
~:++8++2'+±4~4J"F++--hS,2H++C-""8-+-:-H .003 E~====::=:Ed/,-':-':-::7:-~~~~~~~ .002
,.",....Jrjf.-r.-...,..,-;=-.,.....,.-r--.,........"...........,...~-.-.-,-,
.00 I
'.0007
-:.,;....:..:_._-:-.~-"""~
FHH+-~~·.;....-_·
.0005
I++-;-'-'.,--,----.-f.-.-.-,.-,.. ----,.-----•• :
:F2S4
6
j~ .:: : .;"...:.. .
:: •••• ;
••
•••
••
+'-'-'~-"-+~'P-:
~~:;;-~~:.~::.:~:~::~:.:.:
.. .... --. ~
0003
,
'T"7'=""",,,.,,..,...L;-,-,-,-,-,--=~~:,-;;.,,,,,,,-",--,-:7'-'.;..,-j .0002 ... .. ------------_ .. ..... ,.
.
.
.
_--.---~--
.001
."
.0001
2 M = ,Bri-l XD = yrH 4
.0007 . .0005 ..•....
.00005
__ ._-----
. ·:·i··" i·.;.
.0003, .
!:,;.~ ;;, "
.0002,
~TI
..L
..................... •
.00007
QJ
.00003 .00002
'_·- _ _ ••• 0 • • • • • • - - - _ . -
L
.0001
I
::10 7.0 5.0
3.0 2.0
.00001 1.0.0.7 0.5
0.3 '0.2
0.1
H/L Fi gure3- 36 '
Moment ahd defl ecti on. coeffi ci ents. for uni formly-l oaded, :two-way element with all edges simply-supported
3-55
il[J] I·
L
__
-_. _-.' __ -.-.- -- ._- - . ~~':''!f·~::~=~~:'"~~:~';~~ .002 .
M 0.1
XD
..- -
f32
...
c:o..
c:o..
.0001
>'
.00007 .00005
%
c:o..
.00003 .00002
:, .
"~'_':''::'_~-~';
3.0 2..0
:...-.:.- ~-:--
1.0 0.70.5
0.3 0.2
H/L
Figure 3-37
Moment and deflection coefficients for uniformly-loaded, two-waj. element with two, adjacent edges fixed, and two edges simplysupported 3-56
-
L·
I· M =
0.1
XD =
f3lv,/3 2 .01 .007
2:IJ2III~;j:j===1.005
~i~~~~~a;Pj~~;;;;;~~2~ .003 ...
.
Cl:l..
='="==""":''c'2-j~''E:~Ic~~==~~~~:-:-:-'-J .002
.001
~
.0007 •
--:. !".'
•..... ;
, c-_··-;
.0005
, - .• , .. c-~~.-~-,;
" ' . : . ' .,:
:
_
.. .
.
. ..
.0003
_
... . ... . ... .... . ... ..__
....•_.4 •.•. _. _.•.
_._" ...........
,
.0002
...
._-~_.~----~--
,~_._.,.;
..-
--.
...._ .._....,...•
"~:;==:=2~=::3::::E==8===~ J. . ---,.. _-.:.:. . - . .0001 --..- ~:::-.~~_-=_--=:. -'---'---'----j .00007 .1 _ _ . _.-~--.~ - =---.. -_.-'--;
.OOOO!
............
_---_ .. ------ ,-_ .... . . - .
.oooo: .ooooa -
-..
_'
-_ ..
'--_L-J
---J
3.0 2.0
1.0 .0.7 0.5
0.3 0.2
.00001
0./
H/L
Fi gllfe 3- 38
Moment and defl ecti on coeffi ci ents for uniforml y-Lcaded, two-way element with three edges simply-supported and one edge fixed
3-57
1&1 (,J
z
cl: I-
U), U)
1&1
II:
Xm
Xe
DEFLECTION
(a)
ONE STEP ELASTO-PLASTIC SYSTEM
r.
--:- "I
~k.p I-
1&1 (,J
Z cl:
re-
dkE
II
I-
U)
ke
U)
1&1
I
II:
Xp
Xm
DEFLECTION (b)
TWO STEP ELASTO-PLASTIC SYSTEM
r. 1&1 (,J
Z cl: I-
rep re
U)
,.
U)
1&1
II:
DEFLECTION (e)
-Figure 3-39
THREE STEP ELASTO-PLASTIC SYSTEM
Resistance-deflection functions for limited deflections
3-58
..,u
'u
I
I
z
I
~ en
I
Hia:
I I
• Xm DEFLECTION
a)
l&l
ONE-WAY ELEMENT
'u
• I
U Z
~ en
en l&l
I
'up
I I
a:
- I
I
I I I I
X'" DEFLECTION
b) TWO-WAY ELEMENT
Figure 3- 40
Resistance-deflection functions for large
3-59
~eflections
L
i ~I'" I
:r
1111111
Figure 3-41
II IIIII
Determination of ultimate support shear
I 3-60
Ultimate Unit Resistances for One-Way Elements
Table 3-1
Ultimate Resistance
Edge Conditions and Loading Diagrams
,
".
f
."
L "
f
"
j' "
L/2
.
f
'Ru
f
l.!/2
L
t
ru
{
Ru
=
ru
=
,
=
p'
~·w-
1
,
t •I .
~_._-
L/2 '
f
L/2
.
". L
.
LI2
,
-
. "
.
..
"'
-.
,
I
I
-
:
'
L/3
,
, .
_.
=
I
L
..
L2 4(MN+M pl L
iMN
=7
.
MN
Ru
=
R, u.
=~
"
L
• /2
1(2' "'. " I
L/3
I'
:~
Ll3"
...
"
i'
.'
.. 3-61
•
,
,
S(MN .. Mpl
ru
,f
:
2 (MN" 2M pl L
.,
I
L
.t
4M p
L2
•
.- ..
-~'~~,
• Ru
~
L/2
-.l L
=
4 (MN+ 2M pl
"
-
"
·1
if
r.I ~
SMp
=7
I
I
"~
';
t
ru
.
,
.
.
L
!
Ultimate Unit Resistances for Two-Way Elements (Symmetrical Yield Lines)
Table 3-2
Edge Conditions
" Yield Line Locations
Ultimate Unit Resistance
Limits
~
,
"
0
'\
Two adjacent edges supported and two edges free
II
•
Three edges supported and one edge free
II
.
I~ .
.... ...
I
Four edges supported
3.
3 <,
L < - 2
, y:S H
5 (M VN + Mvp) y2
OR
6H M HN + (5M HP - MHN)Y L 2(3H-2y)
5 (M HN + MHP) . ' II 2
OR
2 MYN(3L-ll) +IOllMyp H2 (3L -4x)
5(M
vN+M vp) y2
,
L--l
'~ , .
6LM vN + (5M vp-M yNh H2(3L-2x)
,
.1--4.
::I:
OR
t2-1
B
'"
N
y :S H
I
L
r!;
co
5(M HN+MHP) 112
"
xIl,;;;,:,:.J 3 I
:S L
OR
4 ( MtiN + MHP)( 6 H - y ) L2 (3H -2y) •
~,
,
I
/
/
2
,
3)----<4
~
II
L < - 2
-
5(MHN +MHP) x2
OR
8 (M VN + M VP )( 3 L -
5( MV'N + M vp ) y2
OR
8(M HN + MHP)(3H .y)
II )
H2(.3L - 4 x)
.8
IGID3 --t-_ .3
:I:
3'4
I L
• H y:S 2"
L2( 3H - 4y\
,
,
I
-----
e
•
-~
--
~
Table 3-3 Yield I=ine Locotions
Edge Conditions
e
Ultimate Unit Resistances for Two-Way Elements (Unsymmetrical Yield Lines) Ultimate Unit Resistance
Limits
~
,Two adjacent ,edges supported and two edges free . . ...
-- -
.-
0
xS L Same as In Table 3- 2
.
~El;.;;,z::~1 !
.'
yS H
-
..
· .. .
...
~.
-
. ..
..
'
· .
,
,
0
w ,
,
Pi
H .r
'"
w
Three edges supported and one edge free
2
\
L x< -'2
t-2--1
{~3<,
ySH
~
rl,
/3
I
; 3>----<4 / 2 v r-
L x<- 2
"
Four edges supported
H
IDJ]~ --t-_ .~ 3
:I:
I
4
L
I
X2{3H-2Y) OR 5(MYN3 + Myp) y2
I' ,
L
,
(5MVp - MVN2) (X, t X2) t 6 MVN2 L OR H2 (3L-.2X, - 2 X2)
I MHN I t MHP).(6H- Y)
. ",.-'" 3
I
51 MHN' t MHPI! OR 5 (MHN3 t MHP) X,2 X22
!!.
- 2
:r
(MHN2t MHP){6H-Y) (L-X)2 (3H-2Y) -., \
·
(MVN,tMvp){6L -X, -X2) (MVN2t Mvp){6L-X,-X2) y2 (3L- 2X, -2X2) OR (H-Y)2(3L-2X,-2X2) 5 (MHN' + MHP) OR X,2 5
y<
OR
I MVN' t,Mvp)
yr "
OR
5'(MHN2 t MHP)
xl 5( MVN2 + Mvp) Y2 2
(M HN1+ MHP)(6H -YI -Y2)' (MHN2 + MHP)(6H-Y,-Y2) X2(3H-2y,-2Y2) OR (L-X)2(3H-2y,-2Y2)
.
Post-Ultimate Unit Resistances for Two-Way Elements
Table 3-4 Edge Conditions
Yield Line· . Locations ,
Limits
H
w
-
·
o
H
t2--t
/
,\.
0 '.
'" ~
Three edges • supported and one edge free , · - .
:
'r-'--<
Four edges , supported
'
Irs:J3 ..,....
_.. . .J..._...:.
I
-'.j
x< H
2 MVN/H
L
i~
I
..'
2
-
-
--
..-.
~~._...
.
.. :
'
)/i:.2 .
• -
.
... -;_ . . .
~
-
-
8( MHNt MHP) /L2 2 MVN/H 2
~(
x < HI2 .
2
8 (MHN +Mtip
._.
MVN.t Mvp)
If;2. "'
.. y < LI2 .
,
--.
. 2 . 8 (MHNt MHP)/L
x> HI2 .
y> L/2
..
.., 2 MVN/H
~
-
..
.
y> L
_.
,
2 MHN/L2
-,
~.
-
~
.,
8
:I:
lin-.~r. •
.
.Y < L/2·
..
~
"
,"-
- x>. H
B3
,
2 - ..
,
.-
'I. : , y> i./a
L
8
,
--
y< H
r--=-1
.-: I
•
2MHN/L
.
, {Lj,;:~i;J~
.
-
x> H .
'·1 . L; I
•
-r . '>,.
2 MVN/H 2 "
{t-;;,:,:.J] ·.
,
x < H
E:J
Two adjacent edges supported and two edges free
Unit Resistance
Post - Ultimate
.8 ( MVN +
M~~) /L 2
8 (MH~t
MHP)/H 2
.
Genera 1. and Ultimate Def1 ecti ons . for One-Wai.El ement s
Table 3··5
Edge Conditions and LoodinO DiOCJrams .
f f
LI2
1 ~
l
L
f f
•I
L/2
r ...j
.
L/2
,
L
.
"2 ton
e max
. (Ll2) ton e
, L
"2
,
(Ll2)tan e max
-
,
,
t
L
•
'2
2" ton
ton e
e max
-
(Ll2) ton e max
L·
.
2" ton e max
ton e
,
,
,
L
(Ll2) ton e
.1.. L/2
. Ultimate Deflection. Xu
~ tan'e
'
L
"
Maximum Deflection, Xm
t ,t,
LI2
L
L/2
I
-. ,-
(L/2) ton e
(Ll2) ton e max
-I
•
t
,
I
L
L ton e
L ton e max
,
•
'I
t
I
L
f/2 L/3
i!'
f/2
L/3:: L/3
"
L ton e
L ton' e max -
,.
t:
(L/3) ton e
'.
-
-
•
i
.1
.. -
(Ll3)tane max
,
3-65
,
General, Partial Failure, and Ultimate Deflections for Two-Way Elements
Table 3-6 Edge Conditions
Yield Line Locolion
lim!(.
Suppor
Portial
Maximum Detlectl;)n, Xm .. (O'
Failure Deflection
-
Ultimate Deflection I Xu
Failing
O.f1ecti(
X,
X,
1
.,
EJ~
x ~'H HionSy
H
Two adjacent edges supporled and Iwo edges' free
f"",
H ton Sy
I
~
y ~ L LtonSH
.
Three edges supported and one edge free
~
~
I
,
E:J H {C5J -'>3 I· .: I'
x ~ H H ton Sy
. 'L y<-2 Yton Sy L ton SH 2
y~~
H x';2 x ion SH ~
8
B3 ,
Four edges supporled
,
}----< '
y Ion
Ely~ (H
. xionSH'
,
-!)
tanlsy- tori '(
..xton Sinax .. .ylonSmax
I~~~H~
H ton Ely
(~-x) tan[SH -ton'~( !oxn)~r~ L tonSH 2
. ( to;'SH~ ytonSy' H-y ) ton, ~Sv-ton-I (2y/L)
H ton Sv '2 L X ) ton,~SH-ton-I (2x/H) tonsv~ xionSH'('2-
·U::S:J';1 -- B
~--'---~
-
H
H tanSmax
. ~: ' . -.(tonsmOx)] HionSmax xton maxt(L)t -x on ~Smax~ton, x/H
xtonSmax
L
,,1
"
-
-
L tonSmax'
L
~ ',-I( !onSmal)] LtonSmax y!anSmox+ (. H:yltan emax- ton y/L H tonSmal
.,
L . ~ _1(IonSmal~ HlonSmax donSmal t('2,xlton Smox-ton x/H
L tonSmox 2
ytonSmax
ltonSmax I . ~ (!ons max)] Y.lonSmalt(H-yltan Sma.-ton-i 2y/L 2 H tonSmal 2
xionSmax
HtonSmax . xionSmal+(L) '2' x Ion ~Smadon-i(tonSmax~ 2x/H 2
H
H
L
L H
H
L
,
..
.
~
--
H x ,-2 - H lonSy ,2
~~~~Y~
L ton SH
x,;H x ton Sy
w ,
~.
y,;L y Ion Sy
I
L
~.
x10;' SH ( L - x! Ion [sw ton-I( .
-
{C2J] I
'" '"
;
x'; H xlon SH
'L y';2 y ton Sy
y,k. Lion SH "2 2
Lion SH 2 Hy) Ion ~SY'Ion•l ( !OnSH~ ylonSy' (22y/L )
ylonSmax
L tonSmal . 2
LlonSmax mal)] '2 -yI ton ~Smadon-I (tonS 2y/L ylonSmal+(H 2
L H
Table 3-7 Elastic and Elasto-Plastic Unit Resistances for .One-Way Elements. Edge Conditions ond Loading Diagrams
f f
.
L/2
f.
1:
:t
~
f
L/2
" •I ~ ~
f L/3
.:j
8MN L2
ru
16MN 3L.
Ru
I•
12MN 2 L
ru
I•
8MN L
Ru
Ll2 t ~
L/2
ru
I
L
.
L
f/2 f/2 LI3
--
t
( L/2
Ru
.
,
L
.
-
f
L/2
L/3
-
ru
l
L
Elasto - Plastic Resistance, rep
Elastic Resistance, re
J:
Ru
I
t
Ru
3-67 .
.
"
--
-
Table 3-8 Elastic, Elasto-Plastic and Equivalent Elastic Stiffnesses -for One-Way Elements Elasto- Plastic Equiv.Elastic Elastic Edge Conditions and Loading Diagrams Stiffness, K. Sti ffness,K ep Stiffness, KE
i
-
f .
:t
L
f
L/2
1
L
~
f
L/2
LI2
L/2
I
-,
r
L
L
I
.. -
L/3
fl2
f/2 L/3 L/3
.
48EI
-
48EI
t:
185EI L4
384El 5~
160El" L4
I07EI
48EI L3
I06EI" " 1.3
I
-
I
I
f
f
t
L/2
384EI 5L4
L}
I
·f L/2
t
,
L:
-
~
~
1
"•
t
384EI 5L4
I
384EI L4
384EI 5L4
307El"
192EI L3
48El.... L3
192EI" L3
8El ~
. -
3EI L3
-
56.4 EI L3
t
.
.
* Valid only if MN = Mp -* Valid only if MN < Mp
3-68
--
~
SEl
l4 3EI
L.3 56.4El L3
Table
3~9
Support
S~ears
for One-Way Elements
'I
Edge Conditions arid , Looding Diagrams
,
f f
,
t'
L
'I
L/2
Support Reactions. Vs '
f
L/2
,
~
-,
t
{ ,
J:
.'
t
.
5ruL -8-
R. Reaction
3r uL -8-
,
1'6
ru L
Ru
2
I
.
-
.
ruL
'f
Ru
.
Ru
I
L
16
5 Ru
2
I
.
,
II Ru
R. Reaction
, ·1 , -
,
'f/2
,
L/2
L
~
'.
.1 ..
~
.
L/2
L
L/2
'
L. Reaction
L. Reaction
II,
~
•I
t
,
2
,
L
L/2
2 Ru
'I
~
ru L
. fL/3
f/2
Ll3: L/3 t
,
, 2 ..
. 3·69'
.
Table 3-10 Ultimate Support Shears for Two-Way Elements (Symmetrical Yield lines) Edge Conditions
- Yield Line Locotions
limits
Horlzontol
sheet, V.H
Vertical shear, V.v
~
0
Two adjacent edges supported and two edges free
o
0
x S
\
~
y S H' I.
J .
L
B3 Four edges supported
'.x < _L -
'
2
8
-- --.3
{(::},~:J3 I I L
(6-!.) L
3r u y
5
.
-
.
~
3ruH(I(3 -
3ruL(2-~).
~
t)
.
H
yS -
.5
,
2
I
3ruH(I-~)
3r ux
3ruL(I-~) 2(3-~)
.
)
3r u y
,2(6.~)·
5
,
.~
'~
.
,.
~)
..
:r---< ,
/
)
5
2
,
{L;,;;~;:;J3
~
3r ux.
1.
H
I
.
(6-~ )
Pi
/
Three edges supported ond one edge free
yS H
I
L
t2-i
.....
....,,
3ruL (2-
It:;;;::,:.J 3 I
5
I
..
r
3ruH (2-
3rux
x S L
I
2 (3-
3ru y
5
~)
-
e
4; Table 3-11
: 'Edge Conditions
Ultimate Support Shears for Two-Way Elements (Unsymmetrical Yield Lines)
,
Yield Line Locations
Vertical shear, VaV
Horizontal shear, VaH
Limits
.
~
0
Two adjacent edges supported and two edges' free'
x
-'
"" .....
"
Some as inTo..~~ 3-10
Some as in Table 3-10
xIL,,;,;,:JJ I -, I ._ L
-
,
Y.
"" .... 3 < ,
I
.'
L
A -,
I
I'
/
2
"
3>----<4 Four _edges supported
.
H
'Irrm~
':I:
XI
. < L x2 - 2'
YI
3' 4 L
I
s 2'
Yz
:s -H2 H
s 2'
3ruH(.2L-XI- X2 ) 6L- XI'; x2
3ru x (2H-y) - .... 6H-y
H
.-
--t-_ .>EI
I-
s
J
3
..
~ 5
2'
, -L
t-=-t
>
5
- -
-I
'2
-
3xI ru
< L
Y
..
,-
-
s
-
x2-
I[£J3.
'
,-
"
:
H ::r.
s !i
L XI ~2'
0
Three edges , supported and one edge free
-;
.
r
'I M H r : \
-
L
--
'2
w
~
..
0
-
..
,
3rux ( L -'x )(2H -yl ,.:
l'
..
:
~H-y .;
3 ru XI
.
--
_ 5
.
,
. ~
.
3ru y (2L":XI -X2) 6L-x.- x2_
3rux2
3r u (H-y)(2L- XI-X2) 6L-xl -x2
5
3ru(L- x }(2H-YI-YZ) 6H-YI-Y2
.3ruY
,~
-'-5--
3ru X (2H -YI -Y2) 6H-YI-Y2
,
-
, -,
3ruYI
5 3ruY2
5
TM 5-l300/NAVFAC P-397/AFR 88-22 ,DYNAMICALLY EQUIVALENT SYSTEMS 3-16, Introduction In dynamic analysis, there are only three quantities to be considered: (1) the work done, (2) the strain energy, and·(3) the.kinetic energy. To evaluate the work done, the displacement at any point on'the structure under externally distributed concentrated lo~ds must be" known. The strain energy is equal to the summation of the! strain energies in all the structural elements which
or
may be
in hending,
involves, the structure.
tension, compression,-shear or torsion.
e~ergy
The kinetic ene~gy
of translation and rotation' of all the masses of the
The actual evaluation of these quantities for a given
under dynamic load would be complicated.
structur~
However, for practical problems,
this can be avoided by using appropriate assumptions. ;In most cases, a structure can be replaced by an idealized (or dynamically
equivalent) system which behaves. timewise in nearly .the same manner as. the actual structure. The distributed masses of the given structure are lumped' together into a number
of
concentrated masses.
The strain energy is assumed
to be stored in several'weightless springs which do not have to behave, elastically; similarly, the distribJted.load is replaced by a number of concentrated loads acting o~ the concentrated masses. Therefore, the e--~ quivalent system"consists merely' of a number of concentrated masses joined
together by weightless sprihgs and subjected to concentrated loads whicp vary with time. Tpis concentrated mass-spririg-l9ad system is defined as an equivalent dynamic system. ,Although all structures possess many degrees
of'free~om,
one mode usually
predominates in the response to short duration loads; -chus ,: for all practical purposes, this one mode may be considered to define the'behavior of the structure and the problem can be simplified by considering a single-degree-offreedom system whose pr6perties are those of the fundamental mode of the ·structures. A single-degree-of-freedom system is defined as one in which only one type of motion is possible 'or, in other words; only one coordinate is required to define its motion .. Such a system Ls shown in Figure 3-42 which, consists of a concentrated mass on a fr'ictionless surface attached to a: : -,
weightless spring arid subjected to a concentrated load. A single' displacement variable x is sufficient_to describe its motion. The' material presented herein will be limited to single-degree-of-freedom systems only. Many s t ruc ture s, however, exist which can ' not be adequately described by the first vibration,. particularly when cons i de r a t Lon of dynamic strain is ne~es sary. For these 'situations, a two- (or-more) -degree-of-freedom analysis may be required . . Procedures are presented in ,section 3-:19.2 for performing these 'ana Lys e s , .
Two fundamental methods are available for treating 'simple systems subjected'to dynamic forces. The first of these methods is concerned with solving the , differential equations of the system by either classical, numerical or :,: graphical means. The second meth~d of analysis 'and chart ;olutions dep~nd on solutions ~hich have been determined ,by the use of the first method and'ar~ approximate solutions to the problems in hand. There is a .third method which may also be very useful in vibrational problems; this involves ,the energy co
equat.Lon .
3-72
I
TM 5-1300/NAVFAC P-397/AFR 88-22
In the following paragraphs the design factors used to determine the equivalent dynamic system that is used to analyze structural elements are defined and the methods used to obtain these factors are discussed.
3-17. Dynamic Design Factors 3-17.1. Introduction The values of the mass and external force used in the equation of motion
(Equation 3-1) are the actual values only if all the elements of the mass of
as
the structure experience' the same force and, consequently, move a unit, in which case, the entire mass may be assumed to be concentrated at its center of gravity. In other cases:, the assumption of uniform motion of the .entire mass cannot.be~made without introducing serious error.' This is true' for members with uniformly distributed mass which ..bend or rotate under load. In such
cases,
member.
the motion of the par'ticles of mass varies along the length of the
Beams, slabs, etc., with such distributed mass actually have an
infinite number of degrees of freedom; however, such structural elements can
be represented by an equivalent single-degree-of-freedom system. In order to define an equivalent one-degree system, it is necessary to evaluate the parameters of that system; namely, the equivalent mass ME' the
equivalent spring constant KE and the equivalent load FE'
The equivalent
system is selected usually so that the deflection of the concentrated mass is
the same as that for a significant point of the· structure.
The single-degree-
of-freedom approximation:of the dynamic behavior of the structural element may
be achieved by assuming f defIes ted shape for the element which is usually taken as the shape resulting from the static application of the dynamic loads. The assumption of a deflected shape establishes an equation relating the relative deflection of all points of the element. The accuracy of th~ comp~ted deformation of the single-degree~of-freedom
system is dependent on the assumed deflected shape and the loading region. For example, in the impulsive loading realm, if the assumed deformed shape for a simply supported beam is taken to be a parabola, then the exact solution
will be obtained for the ,deflection of the beam.
In the'quasi-static loading
realm, however, the static deformed shape gives the exact answer. For blast resistance design, however, the differences in either a static deformed shape
or a first mode approximation is negligible and thus either way is permissible . . Usually, though, the static deformed shape is easier to use as it covers both symmetric and asymmetric deformation.
To permit rapid design of structural elements subjected to dynamic loads, it is convenient to introduce design or transformation factors. These factors are used to convert the real system to the equivalent system. -To obtain them, it is necessary to equate the relations expressing the kinetic energy, strain and work done on the actual structural element deflecting' according to the
assumed deflected shape, with the corresponding relations for the equivalent mass and spring system.
3-73
TM 5-1300/NAVFAC P-397/AFR 88-22
3-17.2. Load, Mass, and Resistance Factors 3-17.2.1. Load Factor The 1oad·factor is the design or transformation factor by which the total load applied on the structural element is multiplied to obtain the equivalent concentrated load for the equivalent single-degree-of-freedom system. If the actual total load on the structure is F and the equivalent load is FE' the load factorK L is defined by the equation 3-41 The load factor is derived by setting the external work done by the equivalent load FE on the equivalent system equal to the external work done by the actual
load F on the actual element deflecting to the assumed deflected shape. For a structure with distributed loads; L
r p(x)o(x)
o
where
dx
3-42a
J
0max - maximum deflection of actual structure
p(x)
distributed load per unit length
O(x)
deflection at any point of actual structure length of structure
L
rearranging terms L
rJ p(x)¢(x)
3-42b
dx
o where
¢(x) -
3-43
O(x)/omax
and is called the shape function. NOTE. The shape function, ¢(x) is different for the elastic range and the plastic range and·therefore the load factor, KL, will be different. For example; the shape factor for a simply supported beam with a uniformly distributed load, in the elastic. rangers is defined as 3-44a while for the plastic range ¢(x) - 2x/L
x < L/2
3-74
3-44b
I
TM 5-l300/NAVFAC P-397/AFR 88-22
For a structure with concentrated loads
i
3-45a
~
WD -
r-l r t h concentrated load
where
or -
deflection at load r
i-number of concentrated loads
rearranging terms i
3-45b
~tF~r
r-l where the shape factor is defined as ¢r ~
i ~
3-46
or lOmax
r-l
Again there is a KL for the elastic range and a KL for the plastic range.
However, for a single concentrated load located at the point of maximum
deflection (i.e. the center of a simply supported or fixed-fixed beam, or the end of a cantilever beam) ¢r is equal to one for both the elastic and plastic ranges and therefore KL is equal to one for both ranges. Values for KL for one-way elements are preserited in Table 3-12. Equation 3-41 through 3-46 were used to calculate these values. An example calculating KL is shown in Appendix A. 3-17.2.2. Mass Factor The mass factor is the design or transformation factor by which the total distributed mass of an element is multiplied to obtain the equivalent lumped mass of the equivalent single-degree-of-freedom system. If the total mass of the actual element is M and the mass of the equivalent system is ME' the mass factor KM is defined by'the equation 3-47 KM can be obtained by setting the kinetic energy of the equivalent ~ystem equal to the kinetic energy of the actual structure as determined from its deflected shape .. For a structure with continuous mass
L 1/2f
J
m(x)[~o(x)J2 dx
o
3-75
3-48a
TM 5-l300/NAVFAC P-397/AFR 88-22 . natural circular frequency
where
m(x)
distributed mass per unit length
rearranging terms L
f m(x)¢2(x)
o
3-48b
dx
J
,I ~.
where' the shape function ¢(x) is based on the deflected shape of the element due to the applied loading and not to the distribution of the mass. Since the deflected shape of the element is different for the ~lastic and plastic ranges, ¢(x). and therefore KM. will also be different. Using Equations 3-47 and 3-48. values for KM were calculated for one-way. elements with constant mass.
These values are shown in Table 3-12.
For a concentrated mass system,
1/2
where
i L r-l
3 -.49a
r t h mass
i
number of lumped masses
rearranging terms
3-491;>
An example of calculations for finding KM can be found in Appendix A. 3-17.2.3. Resistance Function Resistance.factor. is the design factor by which the resistance. of. the actual structural element must be multiplied to obtain the resistimce of the equivalent single-degree-of-freedom system. To obtain the resistance factor, it is necessary to equate the strain energy of ~he structural element, as
computed from the assumed deflection shape. and the strain energy of the equivalent single-degree-of-freedom system. If the computed total resistance of the structural element is R and the equivalent total resistance of the equivalent system is RE• then the resistance factor is defined by the equation 3-50
3-76
I
TK 5-l300/NAVFAC P-397/AFR 88-22
Since the resistance of an element is the internal.force.tending to restore the element to its unloaded static position, it can be shown that the resistance factor KR must always equal the load factor KL. 3-17.3. Load-Hass Factor
"
The load-mass factor is a factor formed by combining the two,basic transformation factors, KL and KM, It is merely the ratio of the mass factor to the load factor, and it is convenient since the equation of motion may be written in terms of that factor alone. The equation of motion of th~ actual system is given as 3-51
'F-R-Ma and for the equivalent system,
3-52a which can be r e-wr t t t en as . ,
.
or
F
R
3-52b
F
R
3-52c .'
where
load-mass factor 3-53
and
Me -.effective,total mass of the equivalent system
".
""
)
when expressed in terms of the unit area of the element, Equation 3-52c can be
written as
Values of load, mass andlload-mass factors are presented in Table 3-12. Equation 3-53 ,was used ,to calculate the load mass factor.·; Instead -o f., computing. the: several-factors above, the load-mass factors in the elastic and elasto-plastic ranges can be determined by relating the primary mode of vibration of the·1member to that of an equivalent single-degree-offreedom system. In the plastic range, it can be assumed that neither the moment .nor the curvat.ure vchanges between the plastic hinges under increasing deflection. This behavior results 'in a linkage action, consideration of which can be used to evaluate the effective plastic mass as follows: In Figure 3-43, a portion of a two:way element bounded.by the support and the . ' yield line is shown. The equation of angular motion for this section is
.", .
I
where
en
3-55
m
summation of moments 1m
mass
mo"~nt
of inertia about the axis of rotation .3-77
TM 5-1300/NAVFAC P-397/AFR 88-22
e" -
angular 'acceleration
Substituting in equation
3-56a" where a is the acceleration.
Dividing through by c· 3-.56b
Since the second term is the resistance R
3-56c and the load-mass factor KLM for the sector shown is
3-57 where c is the distance from the resultant applied load to the axis of rotation. L1 is the total length of the sector normal to the axis of rotation and M is the tota1.mass of the sector. When the. element is composed of several sectors, each,sector must be con-
sidered separately and the contributions then summed to determine the 10admass factor for the entire element.
KLM - L(I m /cL 1)/M
3-58
For elements of constant depth and, therefore, 'of constant'unit mass, the
above equation can be written in terms of the area moment of inertia I and the areas Avof .t.ha "individual sectors. r.'
3-59 Table 3-13 gives the load-mass factors for the elastic and elasto-plastic response ranges for various two-way elements. The-values given in the table apply for two-way elements of uniform thickness and supported as indicated. Also if there are operri.ngsvLn the element,- .they must be compact in shape and small in area compared to the total area of the element. . • Figure 3-44 gives the load-mass factors for the plastic response of various two-way elements.
The load-mass factors are a junction of support condition These two-way elements must conform to the following
and-yield line location. limitations:
'(1)
Element must be of constant thickness.
'(2)
Opposite support must provide the same restraint so that a symmet-
.,
rical yield line pattern occurs. (3)
Any openings must be compact in shape and small in area compared··
to the total area of the element.
3-78
I
TH 5-1300/NAVFAC P-397/AFR 88-22
3-17.4. Natural Period,of Vibration
by numerical methods or design charts (both methods are described in Section 3-19), the effective natural period of vibration is required. This effective natural period of vibration, when related to the duration of a blast loading of given intensity
To determine the maximUm response of a system either
and a given structural resistance, determines the maximum transient deflection ~
of the structural element.
The effective natural period of vibration is 3-60 the effective unit mass
where
KE
the equivalent unit stiffness of system
The values used for the effective mass and stiffness for a particular element
depends on the allowable maximum deflection. elastic behavior,
When designing for completely
the elastic stiffness is used.
In all other cases the
equivalent elasto-plas~ic stiffness, KE, is used. The elastic value of the effective mass is used for the elastic range, while in the elasto-plastic range the effective mass is the average of the elastic and elasto-plastic
values.
For small plastic deflections (8
~
2·), the value of the effective
mass is equal to the average of the plastic value and the equivalent elastic
value.
The plastfc effective mass is used for large plastic deflections.
3-79
---+~ K,O,V
R=Kx R .......- -
Figure 3-42
M
~F
Typical single-degree-of-freedom system
I 3-80
,
'i AXIS.OF ROTATION
e. F
C.g. AXIS OF ROTATION
'l'IELD LINE
b) SECTION
a) PLAN
Figure 3-43
Determination of load-mass factor in the plastic range
3-81
1.0
0.5
N d5 0.8
0.4
W
w a:
u,
Wo Ww a: Il.L.a: -~
d5ll.
S ~0.6 I-C/l
0.3
a:W ~C)
w wo a:w u.1-a: d5~ Oll.
W~
I-C/l a:C/l Ow ll.C) ll.0
ll.0
~W
~W
C/llC/lZ
C/l'lt C/l, 0.2 wJ: C) ....
W~ 0.4
C) 0<1:
W..,
0:>0
W ...
,.,~
,.,0
,..J ..J;('
I
IN
;:;.,j 0.1
;(' 0.2
...0 ....I
.
....
:>0
0
0.5
0.5
0.6
0.65
0
0.7
LOAD-MASS FACTOR KLiol 1 Figure 3-44
Load-mass factors in plastic range for two-way elements
3-82
Table 3-12 Transformation Factors for One-Way Elements ,
Edge Conditions and Loading Diagrams
..
f t
f
L/2
~
.
"
'I to
f
. L/2
.
1 ~
L/2
L/2
t ·1 '
L/2
L/2
.L
.
!/2 f/2 L/3 L/3' L/3
Load-Mass Factor KLM
Elastic
0.64
0.50
0.78
Plastic
0.50
0.33
0.66
Elastic
1.0
0.49
0.49
1.0
.
0.33
0.33
0.58
0045
0.78
0.64
0.50
0.78
0.50
0.33
0.66
1.0
0.43
0.43
1.0
0.49
0049
0.33
0.33
0.41
0.77
0.64
0.50
0.78
0.50
0.33
0.66
1.0
0.37
0.37
1.0
0.33
0.33
0040
0.26 0.33
0.65 0.66
0.24 0.33
0.24 033
0.52 0.56
0.60 0.56
Elastic Elasto- . Plastic Plastic
t
Elastic Elosto . Plastic Pla-stic .
,
-
.
1.0 0.53
.
"
Elastic Plastic
-I -
, Elastic Plastic
I
·1
Mass Factor KM
Elastic ElastoPlastic Plastic
f L,
Load Factor KL'
Plostic
•I
L'
.
t
t
L
,1 •I
t
L
..
~
.• Range af Behavior
Elastic Plastic.
.
"
Elastic Plastic
t
'.
3-83
, 0.50
1.0 .1.0
0.87 1.0
Table 3-13 Load-Mass Factors in the Elastic and Elasto-Plastic Ranges for Two-Way Elements Elastic and Elasto - Plastic Ranges (Support Conditions
'Value of L/H
Support Conditions
,
Two adjacent edges supported and two edges free
OJ I I
I
On. Support Simpl. Two Support. Simple Thre. Support. Simple All Support. FI••d Other Supports Fiud Other Support. Fiud Other Supports FI..d All S~pporll Simple
.-
.ALL
0.65
"
-
0.66
,
--
-
... 0.66 .
L
,
.'
w
e.
.
.-
-lIti <0.5
•
ex>
.".
Three edges supported and one edge free
OJ I. I . L
Four edges supported
03 I I L
,
'.
0.71
0.77
0.79
. L L .5s.L/HS2 0.65 -0.16 ( 2H-1) 0.66 ;O.l44(i"H-I) 0.65 -Q186(2H -I) LlH~2
0.65
0.66
. 0.65
L/H=I
0.61
0.61
0.62.
;
L ,ISL/HS2 0.61+0.16 (H-I) LlH
~
2
\
0.61+0.16(~
0.71
0.77
a
--
0;79
--
L 0.66-0.175 (H-I)
--
: 0:66
0.63
0.63
.
I-
-I) 0.62+0.16 ( 0.78
~"I )
, L 0.63 +0.16 ( H -II 0.63+0.i6 0.79
(~ -I)
0.79
1M
5-l~OQI.NAV.FAC
P-397/AFR 88;22
DYNAMIC ANALYSIS 3-18. Introduction
iJ~. ,
'
I)
", ., ,:;
1."1
,.,
I .
)
..
.,
"
•
.
.:
.
Structural-elements must develop an internal resistance. sufficient to maintain
all motion 'within th~' limits of deflection':prescrl.bed·' for 'the pa'rtic\ha~' , design. The load capacity of the member depends on the peak strength d~v~l· ' oped by the specific member (see Chapters 4 and 5) and on the ability of the
member to sustain its resistance for a specific .though r e La t i.ve Ly s ho r t i pe r Lod
of time. This section describes the methods used for the analysis of structural elements· subj e,cted, to dynamic. loads "a!,d,'1::h,e,s,e elements are d Lv Lded into three groups, namely: (l)f~lements,wh~ch"respondto,~he pressure o~ly (low pressure range), (2) eJem~I\ts,whi ch. r,espond, to" pressure - time relationship (intermediate-pressure design range) 'and (3) elements which respond to the impulse which corresponds to the high pressure design range. ,, The method of analysi~,is~~dit.fere!'tfor the groups of elements. Those elements that respond to pressure'only (this is still a pressure-time rela-, tion) and those that respond to pressure-time relationship may be analyz~d' using either numerical methods or design charts,
~hile
elements that respond
to the impulse are analyzed using either design charts for large deflection or an impulse method for designs with limited deflectip,ns.
,
The method of analysis employing numerical teph~iques provjdes a solution .that is obtained by a step-by-step integration, and is generally applicable 'for any type of load and resistan~,e function. This method, of analysis i.s presented in this section. Also presented is the result of a systematic' analysis of single-degree dynamis ~yste~s ~o, ,several idealized loadings plotted in nondimensional design charts.
.
3-19. Elements Which Respond to Press~re Only and Pressure-Time Relationship. 3-19.1. General To analyze these" elements ,I the true magnitudes for the .,pressure-time relationship of the applied blast' ~loads"must' be known, The de ce rmj.nat Lon of the-
dynamic response of these systems is accomplished using numerical techniques or design charts which relate the dynamic properties of the element (natural period of vibration, resistance, and deflection) to those of the blast overpressures. ~l
,
,
3-19.2. Analysis by Numerical Methods The numerical method of'a~alysi~involve~:~:tep~by-ste~'integrationprocedure, starting at zero time, when displacement and the velocity of the system are presumably known.
The time scale is divided into
sever~l.:discrete
intervals, and one progresses by successively extrapolating the displacement from one., time st a t i cn to, tp~ next. As the time interval in, the sequence of time is reduced.. _i.e., aSjt~he number of;discrete.'.intervals,.~s increased, the
accuracy of the numerical method is improved.
3-85
h
'.
.
TK 5-1300/NAVFAC P-397/AFR 88"22
3-19.2.1. Sing1e-Degree-Of-Freedom Systems As stated previously, a single degree of freedom system is one in which only one coordinate is essential to define its motion. The equation of motion for the single-degree-of-freedom system shown'in Figure 3-42 can be expressed as: 'I',
3-61
'F-kx-'"cv-Ma when c - damping constant ,
,
There are several methods that can be used to nUmerically integrate Equation 3-61, but the two procedures applied to this problem will be the average acceleration,' and the acceleration- impulse extrapolation methods.
'.',.:
3-19.2.1.1. Average Acceleration Method at any time In the average acceleration method, the velocity and displacement , t are expressed
~s
v t, - vt_l + 1/2 (at +
at_l)~t
3-62
~t
, 3-63
x - xt_l + 1/2 (v t + vt_l)
Substituting Equation 3-62 into Equation 3-63, , -
': .
x - xt_l + vt_l ~t + 1/4 (at + '~t-l) ~t2
3-64
Substituting Equations 3-62 and 3-63 into 'Equation '3-61, we have ·t •
F - k [xt_l + 1/2(vt + vt_l).Hl
c[vt_l + 1/2 (at +
at_l)~t]
- Ma 3-65
Equation 3-65 can be simplified to 1 a m + (1/2)c
~t
+ (1/4)k H 2-
- c (vt_l + 1/2 at_l
~[
F' - k ~Xt_l + vt-l
~t
+ 1/4a t_l ~t2) 3-66
~t)l
The integrating procedure can be summarized as follows: Step 1.
At t - 0: Compute: a o using Equation 3-61 and specified values ' for xo(t - 0); vo(t-- 0), and F(t -'0),
Step 2.
Increment time:
Step 3.
At-t
t + 63. Step 4.
t
~t)
+ At:
t - t +
~t
Compute aCt
"t
+ .dt) using Equation 3-66, v (t
using Equation 3-62 and x (t - t +
~t)
using Equation 3-
Repeat steps 2 and 3 until some specified time or until the maximum displacement is
rea~hed.
3-86
=
1M 5-l300/NAVFAC,P-397/AFR 88-22
It should be noted that it x (t - 0) and v (t - 0) are zero, then the system will not move as long as F (t - 0) is zero; however, if F (t - 0) is not zero, Equation 3-66 will give an acceleration and motion will start.' Thus the system of equation is self'starting and requires no special starting provisions.
:. I
,
The procedure outlined above for an elastic single-degree-of-freedom system" can be easily extended to I accommodate elastic·plastic behavior.
To do so,
Equation 3-61 is re-writt~n'to include the restoring force R which replaces the spring restoring force, -kx , Equation' 3-61 becomes F
R - cv - ma
3-67
Substituting Equations 3-62 and 3-63 into Equation 3-67, the following equation for acceleration \S obtained: 1
m + (cAt)/2
t.,. 0
[,"
""·1 / 2]J
3-68
"
Unlike Equation' 3 - 66,' Equation 3-68 has a non- l'inear term the displacement x at time -to
Therefore,
R: 'and
it depends on
instead of using the direct integra-
tion approach as in the previous case; a predictor-corrector method is utilized. Briefly, the displacement x at time t is estimated (predicted) and then corrected and convergence of this procedure can be obtained in a single iteration if the yaLue of At is' small enough. The step-by-step procedure is outlined as follows: I Step 1.
At t - 0: Compute at (t - to) from Equation 3-67 and the' initial conditions.
Step 2.
Increment time:
Step 3.
At t - t + At:
Step 4.
Compute vt and xt from Equations 3-62 and
Step 5.
Compute R,
Step 6.
Compute at from Equation 3-68
Step 7.
If you are in th~ predicting pass, return to Step 4, i f in the .. correcting pas s vse t xt-l - xt, vt-l - vt, at - at-l and go to Step ,
t - t + At Set at - at-l (t
-
1 - 0; initially)
"
.
'3~63
i
2.
3-19.2.1.2. Acceleration Impulse Extrapolation Method Suppose the acceleration of the system is defined by Figure 3-45a. The acceleration· impulse extr~polation method assumes that·the actual acceleration curve can be replaced by a series of equally spaced impulses occurring at to, tl' t2' . . ., tn' as shown in Figure 3-45b. The magnitude of the acceleration impulse at tl} is given by 3-87
TK 5-l300/NAVFAC P-397/AFR 88-22
L
3-69.
-
. where t- tl', - to -,' tz - tl - , , - t n - tn_l' This is shown in Figure ,3 - .t. 45b, Since an impulse is app Lt.ed-at; -. t n" there is a discontinuity in the value of velocity at tn' In the time interval from t to tn_i' the velocity is constant and the displacement varies linearly with time. The velocity and displacement thus obtained are .shovn in f~gures 3-45c and 3-45d. ,
.
"
"
Suppose ,tn-and t n + indicate the time immediately before.and after the application of the impulse at tn' and let v n- and ~n+ ·ind~cate respectively,. the velocity at t n and t n+; these two velocities are related by the following equation:
3 - 70, ~
The relationship between xn_l and x n• and between x n and xn+l are given by: ~ - ~_lvn-(At)
)- 7la
~+l - ~
3-71b
vn+(At)
Combining Equations 3 - 70 and 3 - 7.1 •.the three successive dd sp Lacemenus., are .. r e La t ed by.; '. r l~ I
..
'
'.
. ',~+l -.Zxn·,- ~-l + an(At)Z
3-72 ..... This i~ the basic Irec.urrence . . f ormula fo r •.the acceleration impulse ex t r apo La- .
tion method. Once the values of x at tn_l and t n are known, the value at tn+l can be directly computed without resorting to a trial and error procedure.
"'
ij
It i~ necessary, however, to use a special procedure in the first time
interval because. at t - O. no value of xn_l is available. Two different procedures may be used. The first assumes the acceleration varies linearly up to the first time station. in which case. the displacement of the system at that time can be expressed as: 3-73
The second procedure assumes that the acceleration is constant during. the first time interval and is equal to the initial value. This assumption results in the following expression for Xl: 3-74 It should be noted that Equation 3-73 has to be solved by trial and error since al depends on Xl' This trial-and-error procedure can be avoided by taking the .value of al to be f(t)/m instead of the exact value of [f(tl) R(tl) lIm. In Figure 3-46a if a o' - O. a o is equal to a o'/6 and in Figure 3-. 46b if a o and al .are appro~imately equal to each other. the value ao'/Z may be"" used for a O " .The recurrence formula is directly applicable for the evaluation of x2' x3' <:: ~n+l wheneve r the acceleration is a continuous .function. ,When there is a discontinuity in the acceleration,' this equation can still be used I
but not before some modifications are made to reflect this discontinuity.
In
Figure 3-46a there is a discontinuity at t 4 which occurs at the end of'the ."' time interval At. Under this condition. the value of an used in the numerical
3-88 .
I
TM 5-l300/NAVFAC P-397/AFR 88-22
procedure is the average value at discontinuity, namely l/Z (a4' + a4") for The a4' In Figure 3-46b the discontinuity occurs within time interval, t. correct value for a3 in this procedure is: 3-75 When there are more than"two discontinuities, this procedur~ becomes cumbersome and it is more convenient to replace the given acceleration by a smooth curve. Equation 3-73 must be used if there is zero force (and hence zero acceleration) at zero time, for in no other way can xl be determined. If acceleration at t
0
=
is~not
zero,' Equation 3-74"may be used without appreci-
able error, provided the force does not change greatly.in the first interval. The tedious method of numerical analysis is made somewhat easier if the
computation procedure is set up as shown in Table 3-14.
In the first time
interval, the value used for the acceleration, a o' -depends· on the forcing function. If the value the forcing func~ion is· zero at time zero, then a o
of
is taken as one-sixth of ai, 'where a, is the ratio of the value of the forcing function at t, to the mass of the system tao - f(tl)/6m]. When the value of the forcing function is not zero at time zero, and if the values of the forcing function at times t-O and t-t l are approximately equal, then a o may be taken as (Po-Ro)/Z. The recurrence formula (Equation 3-7Z) is used in the other intervals. It may be necessary during the computational procedure, to increase the value of t so as to reduce the number of time intervals. In Figure 3-47, for example, the forcing function varies linearly after time j. For the next time station (j + l)st,the value of ~t, can be increased (~tz - Z~tl for example)
without introducing any errors in the computations.
In such cases, caution
has to be exercised when using the recurrence formula. In Table 3-14, the value of ~t changes after the jth step, to ~tz - Z~tl" In the (j + l)st step, xn_l should be the value of x for the (j-l)nd step and x n will be the value of xn+l for the (j-l)st step. 3-19.2.2. A Two-Degree-of-Freedom"System ~i
A two-degree-of-freedom ~ystem 'is shown in Figure. 3-48 where the coordinates defining the configuration of. the system are Xl and X2. The equation of motion for each of the two masses can be expressed as follows:
Fl + RZ
Rl - Mlal
3;76a
FZ
MZaZ
3-76b
RZ
Equation 3-76 can be rewrittenras: al
Fl/M l + RZ/M l -'Rl/M l
3-77a
aZ
FZ/MZ
R2/M2
3-77b
The integration procedures described for a single-degree-of-freedom system can also be applied to two-degrees-of-freedom systems. In the average accel.eration procedure, if the equations are placed in the same form as Equation 3-61, i.e., if the acceleration at time t are written in terms of displacements and
velocities at time t -
~t,
then two equations are obtained which must be. 3-89
TH 5-l300/NAVFAC ·P-397/AFR 88-22 .
solved simultaneously for aland a2' In the other numerical integration procedure, the recurrence formula (Equation 3-72) applies to each of the two displacements independently.
,.
3-19.2.3. Damping
The effects of damping are hardly ever considered in blast design because of the following reasons: (1). (2) (3)
Damping has very little effect on, the first peak of response which, . is usually the.only cycle of response·that is of interest. The energy dissipated' through plastic'deformation is_much greater than that dissipated by normal structural damping. Ignoring damping is a conservative approach.
If damping has to be considered in pres~ed
analysis, however, it should be ex-
~n
as some percentage of critical damping.
For free vibration, this is
the amount of damping that would remove all vibration from the system and allow it to return to'. its neutral position. Critical damping is expressed as
..
2kM
where
-r :
k
stiffness of the system
M
mass of system
3-78
1·,-
The damping coefficient, c , in Equations 3-61 and 3-67 is expressed as a '.percentage of critical damping, Cc r' For steel structures, c should be taken as 0.05C c r and O.Ol'C e r for reinforced concrete structures. 3-19.3. Design Charts for Idealized Loadings
,"
3-19.3.1. General The response of single-degree-of-freedom systems subjected to idealized blast loadings is presented in the form of non-dimensional curves. In order to utilize these.response charts,. both the blast loads (pressure - time history) and the resistance-deflection curve of the structural system are idealized to linear or bilinear functions. Methods for computing these idealized blast loads are given in Chapter 2 while the methods for computing the resistancedeflection functions as well as converting the actual system to an elastic or elasto-plastic single-degree-of-freedom system have been presented in previous sections of this chapter. The response of a structural system subjected to a dynamic load is defined in terms of its maximum deflection Xm and the time tm to reach this maximum deflection. The dynamic load is defined by its peak value P and duration T while the single-degree-of-freedom system is defined in terms of its ultimate resistance r u' elastic deflection XE and natural period TN. Response charts relate the dynamic properties of the blast load (P and T) to those of the element (r , X, TN)' that is, Xm/X E and tm/T are plotted as aTunc t Lon of. ru/P and T/TN. Response· charts have been.prepared for simplified loads as well as for the more complex bilinear loadings. The simplified loadings include triangular, 3-90
I
TM 5-l300/NAVFAC P-397/AFR 88-22
rectangular"step load wit,h finite rise ,time, ~riangular with rise time and sinusoidal pulse. The id~alized triangular loading is utilized in the analyses of acceptor structures. These structures are subjected to the effects of unconfined explosions, such as free-air or surface bursts or
partially confined explosions in donor structures which are fully vented. On the other hand, idealized instantaneouslY applied or gradually applied flat top loads are usually encountered in the design of roof elements of acceptor structures. : These structures,are subjected.to
v~ry
long duration loads
produced by,Nery large.explosive quantitie~. The t!iangular loading with a rise time is similar to the previously mentioned triangular load except that it is an equivalent loading produced by the blast traversing an element. The sinusoidal loadings is usually associated with the vibrational type of load rather than ,blast induced,input. The· bilinear loading is utilized in the analysis of containment structures which allow partial venting of an explosion and in the analysis of, acceptor structures where.reflected pressures clear rapidly around the structure. ~'. The effectsiof damping have not been included in the preparation of the response charts.
In the design of the blast resistant structures, the first
peak of response is usua l-l.y the only cycle .of response that is of interest since the maximum resistance and deflection is;attained in ·that cycle. Damping has very little effect on this first peak and consequently, neglecting damping has negligible effects on maximum response calculations.' 3-19.3.2. Maximum Response Of Linear Elastic System To Simplified Loads To obtain t~e response of a linear elastic system, it is convenient to consider the concept of the dynamic load factor. ,This factor ,is defined as the ratio of the maximum dynamic deflection to the deflection which would have resulted fr~m the static application of the peak load P, which is used in specifying the load-time variation. Thus the dynamic load factor (DLF) is, given by: DLF
3-79a ,I
where
static ;deflection or, in other words, the displacement
produced in the system when the peak load is applied statically. 'J
maximum dynamic deflection Since deflections"spring ,forces, and stresses in an
elasti~
sy?tem are all
proportional, the dynamic load factor may be applied to any of these to determine the ratio·of dynamic to static effects. Therefore, the dynamic load factor may~e considered, as the. ratio of the maximum resistance attained to the peak load or: DLF where
3-79b
lr
maximum dynami~ resistance, p ,- "peak load used in specifying, the load-,time variation.
For a linear elastic system subjected to a simplified dynamic load, the maximum response is defi~ed by the dynamic load' factor, DLF and maximum response time, t m. The dynamic load factor and time ratio tmlT are plotted 3-91
TM 5-1300/NAVFAC P-397/AFR 88-22 versus the time ratio T/TN for a 'triangular load, rectangular load, step load with finite rise time (T r), triangular load with rise time, and sinusoidal pulse in Figures 3-49 through 3-53 . • In many structural problems only the maximum value of the DLF is of'interest. For the most prevalent load case, namely, the triangular load as well as the rectangular and step load with rise time, the maximum value of theDLF is 2. This immediately indicates that all maximum displacements, forces,'and stresses due to the dynamic load are twice the value that would be obtained from a static analysis for the maximum load P. The above response charts apply for elastic systems.
However, the charts can
be applied to the entire elasto-plastic range if the actual resistancedeflection curve ,(two step system, three step system, etc.) is replaced by the equivalent elastic system where tne equivalent elastic stiffness KE and the ' equivalent elastic deflection XE are obtained according to previously ,explained procedures. In a typical design example, the pressure-time loading is calculated and idealized to one of the simplified loads defined by P and T or Tr, "Astruc" tural member is assumed and its corresponding dynamic properties are calcu-
lated. For a completely elastic response, the'values of r e and TN are calculated while for a response in the elasto-plastic range, r u' XE and TN are obtained. It should be noted that for a response in the elasto-plastic range, the value of TN is calculated using the' effective 'mass which is an average of the elastic and elasto-plastic values. Knowing the ratio of T/TN the dynamic load factor (DLF) and the time ratio can be read from the appropriate figures. The maximum resistance r attained by the structural member ,is calculated from the DLF and the response time is obtained from the time ratio. If the resistance r is greater than r e for a completely elastic response in 'the elasto-plastic range, then the analysis'is not valid and the 'procedure is repeated.
The maximum deflection is obtained from the resistance r- and the
stiffness Ke or KE. 3-19.3.3. Maximum Plastic Response Of An Elasto-Plastic System To Simplified Loads An elasto-plastic system may have an elastic or plastic response depending upon the magnitude of the blast load. If the response is elastic, that is, the member attains a resistance 'r which is less than its ultimate resistance r u' then the charts of the preceding section are used. The response charts presented in this 'section are only for the plastic response of members where the ultimate resistance r u is attained. ' The maximum plastic response of an elasto-plastic system subjected to a blast load is defined by the maximum deflection, ~ it attains and the time, t m it takes to reach this deflection. The blast load is defined by its peak value P and duration T while the singe-degree-of-freedom system is defined by its ultimate resistance r u' elastic deflection XE and natural period TN' A nondimensional response chart is constructed 'by plotting the ductility ratio ~/XE and the time ratio tm/T as a function of ru/P and T/T N. Response charts are given for a triangular load, rectangular load, step load with finite rise time Tr and triangular load with 'rise 'time in Figures 3-54 through 3-61, It· should be noted that for the'step load with finite rise time, the load is 3-92
TK 5-1300/NAVFAC P-397!AFR 88-22
defined by the rise time ;Tr and, ,consequently, the response curves are plotted using Ti rather than T. " - If 1
In a typical example, the pressure-time loading is calculated and idealized to one of the simplified loads defined by either P and T ~r P and Tr, A structural member is assumed and is.. corresponding dynamic properties .(ru' XE, TN) are calculated. Knowing.:the ratios of, ru/P and T/TN (or Tr/TN), the ductility ratio Xm/XEand the time'ratio tm/T can be read from the appropriate figures. The maxim4,deflection X]j, and ,response time tm can readily be cal<;ula~ed., If the ductility ratio'Xm/X E or the maximum deflection Xm and corresponding response time t are uns~tisfactory, the procedure is repeated.
m
'It should be noted that the value of the natural period of vibration TN,used in conjunction.with the "response .charts varies according to_the magnitude of
the ductility ratio K,n/XE. For .sma Ll, plastic deformations ,(Xm/XE less than 5), the.calculation~,of TN is based on an average of the equivalent elastic and plastic masses., Whe"eas,' for larger plastic, deformations (Xm/X E greater than 5), the equivalent plastic mass is used to, obtain TN' ,"
.~
~
3-19.3.4. ~aximum Plastic Response Of,An Elasto-Plastic System to Idealized Bilinear Leads r Response charts have been prepared for bilinear loads in much the same manner as for simplified loads "1 However,' four parameters are required to define the bilinear loading rather than two parameters which are required for the simplified loads: These a~ditional parameters greatly increase the number of charts required for the·rap~d prediction of the dynamic response of an elasto-plastic system.
However"
the computational procedures remain comparatively simple.
A bilineartload" is illustrated,in Figure 3-62c. to define this load shape" namely:
Four parameters. are required
peak pressure of shock load (primary pulse)
P
peak pressure of gas load (secondary pulse) duration of shock load C2T
duration of gas load
I
It can be kee? from Figure 3-62c that the value of Cl will always be less than 1 while th~ value of C2 will always be greater than 1. Response charts for the bilinear load are prepared in, the same way as the simplified, loads except ,that, each chart is prepaied:for a given value of Cl and C2. The ductility ratio Xm/XEversus T/T N is plotted on each chart for various values of P/r u' tm/T N and tE/T (where t E is the time to reach the maximum elastic defle<;tion xE)' Therefore, each chart contains three families of curves. In addition, each chart contains one, two or three boundaries which are formed by symbols. These boundaries delineate regions of each chart where certain approximations'are preferable to chart interpolation .., These approximations involve
modificati~n
of the
bili~ear
load for the various
regions. The loads considered for each of the regions are given in Figure 362 while the boundary for the various r~gions are defined in Figure 3-63. 3-93' ,
TM 5"1300/NAVFAC P-397/AFR 88-22 Numerous response charts are required for bilinear loads. Table 3-15 lists the figure numbers of the response charts prepared for the selected'values of Cl and C2. These response charts are presented in Figures 3-64 through 3-266.
•
To use the response charts, the first step is to enter Table 3-15 with the required values of Cl and C2 . Locate the points in the table with coordinates Cl and C2. The number in the box containing the point 'is the appropriate. figure' number to use to solve the problem. ' If the point is not located in a box having a number, the two or, more frequently, four numbers bracketing the point are the appropriate figure numbers to use. "Interpolation for the; . required values of Cl' and C2 must be 'performed. A graphical and mathematical interpolation procedure .is presented in subsequent sections.
In a typical design, a structural member is assumed and the1idealized resistancefunction defined by r u' XE and K E can be 'determined along wi'th the
natural period. As previously explained for simplified loads; the equivalerit mass used to calculated TN varies according to the amount of plastic deformation. Knowing the ratio of P/r u and T/TN, note'if the intersection point corresponding'to these parameters, is located in.region A,B,C or D on the response chart(s) for the required values of Cl and C2. The 'boundaries for
these regions are represented by a line of asterisks, solid circles, and solid
squares, as illustrated in Figure 3-63. The final step in obtaining the solution depends on which, region of the response chart(s) is applicable. These regions were established to reduce the amount of calculations required for a solution by eliminating, where possible, interpolation between charts.
Except for
one
of- the four regions ,:- the actual
bilinear blast load is replaced by a simplified load. Solution of the response for these simplified loads will predict a solution which is within 10 percent of the' exact solution. Of course, the ,exact solution of the response for the structural member in these three regions can 'be obtained by using 'the required response charts and interpolating, where required, between charts.
3-19.3.4.1. Region A This region is defined as the area to the left of the solid circles on the response charts. If a chart does not have a line of solid circles, then region A does not exist. Figures 3-221, 3-222 and 3-223 'illustrate this case. In region A the maximum dynamic response depends primarily upon the total impulse in the blast loading (area under the pressure-time curve). The actual pressure-time distribution does not significantly affect the maximum dynamic response because in these regions the durations T and C2T are small in comparison to the response time, t m of the· structural member. In this region the response charts are not used and a solution is obtained by considering the modified loading shown in Figure 3-62a: This, load shape yields the following equations:
2 1
+ 2
2
[:,][:01 ' 3-94
3 - 80.
I
TM 5-1300/NAVFAC P-397/AFR 88-22
and
[-;-j ,[ :U ][ :n]
3-81
where, 3-82
F
3-19.3.4.2. Region B This region is defined as the area in the chart to the' left of the asterisks. If a chart has no line of .asterisks, then region B does not exist. Figure 365 illustrates the case :where region B does not exist. In region B the maximum dynamic respo~se depends primarily upon. the gas load which is, described by GlP and G2T. The shock load described by P and T is neglected. This load condition is shown in Figure 3-62b. Therefore the solution is obtained from consideration of a s LngLe i.t r Langu.Laz load. The dynamic response is obtained from Figure 3-64 which like Figures 3-54 and 3-55 is for a ,triangular load, and will yield the same results. When using Figure 3-64 it must be realized that the P and T used in the chart are actually G1P and G2T, respectively. I~ a typical design, enter Figure 3-64 with the normalized parameters, GlP/r u and G2T/T N, and read the solution, Xm/X E, tm/T N, and t E/G2T. Figure 3-63 depicts region.B as .the area between the line of solid circles and the line of asterisks. It should be understood' that the solution technique associated with region B ,(i.e., neglect the shock load and use Figure 3-64) applies everywhere to the left of the line of asterisks, including region A. In other words, two solution techniques are available in region A.
3-19.3.4.3. Region D
.
'
,
This region is defined.as the area in the charts to the right of the line of solid squares.. If a chart has no line of solid squares, then region D does not exist. Figure 3-69 illustrates the case where region·D does not exist. region D the maximum dynamic response depends primarily upon the shock load which is described by P an~ T. The gas load described by G1P and G2T is neglected. This load,condition is shown in Figure 3-62d. Therefore, the solution is obtained from qonsid~ration of a single triangular load. Similar to region B, the dynamic: response is obtained from Figure 3- 64 which is for a triangular Load . Unlike!'region B, the parameters, P and T which describe the load are .used·in the figure. Therefore, in a, typical design, enter Figure 364 with ,the normalized parameters P/r u and T/TN, ,and read.the solution, Xm/X E, tm/T N and t E/ T. ~n
3-19.3.4.4. Region C This is the region.in the charts which do not meet the definitions of regions A, Band D. In most charts, region G is the area to the left of the line of 3-95
TM5--13ooiNAVFACP-397/AFR1 88-22 .
solid squares and to the right of the line of asterisks, as illustrated in Figure 3-63. Region C does not exist in ,some charts, such as Figure 3-97 which have over lapping regions A and D. In region C, both the shock and gas load 'must be,considered. Replacement of the actual bilinear load with a simplified load, as done for the other regions, will yield an incorrect solution. Therefore, in this region the response charts must be used. If the required values of Cl and/or C2 do not
correspond to the response chart values,: interpolation "between response charts
wH·l be required to obtain a solution.
In a- few cases one 'or two response
charts are needed, however, in general four-response charts are required for a solution.
3-19.3.4.5, Response Chart Interpolation Interpolation between'four response, charts will usually be required for region For regions A, B, or D, if conditions warrant 'an exact solution rather than the approximate solution usually used in these regions, interpolation between charts,may,be required: Eitner a graphical of mathematical interpolation procedure may be employed. The'method selected ~epends upon personal choice. A brief description of each procedure Lsip r e senced below and an ' C.
example of each is
giv~n
in Appendix A.
. '
'
Graphical interpolation requires a·sheet of log-log, graph paper. A convenient size is 2 x 1 cycle with the single-cycle axis representing Cl and C2, and the' two-cycle axis 'representing the, desired parameter(~/XE' tm/T N or tE/T) , called Y for ease of presentation. The appropriate figures to be used for a solution are obtained form Table 3-15 for the required values of Cl and C2. The procedure will illustrate the interpolation between four figures since
this is by 'far the usual case. For ,the values of P/r u and T/TN corresponding to the structural system selected, determine the desired parameter Y for each of the four figures. Organize a table in the same format as Table 3-16 and enter each figure number and corresponding value of Cl, C2 and Y. In the table, Cll and C21 are the values of Cl and C2' respectively, from Figure 1. Likewise, C12 and C22 are the values of Cl and C2' respectivelj from Figure 2, etc. The symbol Yl, is the desired parameter, such as ~/AE' which is read from Figure 1. The symbol Y2 is the value read'from Figure 2, etc. The interpolation is first performed for the required value of rand then for the required'value·of C2\' That is, with C21 constant: grapl:.cally interpolate on the log-log paper between points (Yl' Cll) and (Y2, C12) to find Ya corresponding to Cl'and C21' as shown on Figure 3-267: Similarly, with C23·constaht; graphically 'interpolate between (Yi, C13) and (Y4 , C14) to find Yb·corresponding to Cl and C23' also shown in Figure 3-267. The values of Ya and:Yb are recorded in Table 3-16. Ei.nal'Ly , wHh Cl constant; graphically interpolate on the log-log paper between (Ya, C2l ) and \Yb,"C 23) to find Y corresponding to the required Cl and C2' as shown on 'Figure 3-267. The values of,Y is the solution.' Since there are three parameters to define the response of-a structural, namely, ~/XE' tm/T N and tE/T; the 'interpolation
procedure is repeated three times, once for each parameter.
'
Mathematical interpolation requires the same initial steps as graphical interpolation. That is, the appropriate figures to be used for a solution are obtained from Table.3-15 and the required parameters 'are determined and" 3-96
I
TK 5-l300/NAVFAC P-397/AFR 88-22
entered into Table 3·,16. Logar Lthmi.c equations are used to' obtain the values of Ya• Yb and Y. The value of~nYa is obtained from: In('Y 2 /Yl ) In(
ci
ICll) 3-83
In( C12 I Cll) and the value of lnYb is obtained·from: lnY 3 4-
In( Y4 /Y3 ) In( Cl IC 13)
Using the values of lnY a' and lnYb; the value _,r lnY
lnY~
4-
3-84
In( C14 I CU)' ~f
lnY is obtained from:
(lnYb - lnY a) In( C2 IC 2l)
3-85
In( C 23 I CZ1)-
The desired parameter Y is then obtained from: Y
e l nY
3-86a
The above equations us'e ' natural logarithms. Common logarithms can be used to solve the above equation~ and then· solve for Y using:' Y
1010gY
3-86b
It should be noted that 'if C2 is represented by a response chart, then only t",o charts will' be invor'ved', and only Equation 3 - 83 will app Ly. 3 -19.3.4'.6, Accuracy The prediction error is less than 10 percent for the approximate solutions obtained in regions A, B' and D. This'is true provided the recommended procedures are employed, that is, the actual load is replaced by the approximate loads as shown on Figure 3-62 and the solution is obtained using the equations provided for region A and using Figure 3-64 for regions B and Dc Exact solutions are obtained from' the response charts if the required values of Cl and C2 -cor r e spond to those given in the charts. This' is' t rue for all four regions. Errors result from interpolation between'the'response charts. The prediction error in ~/XE for region C, where interpolation between charts is required, is less than 10 percent provided response charts bounded by the dashed or solid lines on Table 3-15 are not used. The prediction error in ~/XE will range from 55 to 100 percent for solutions in region C, if the point (C l, C2 ) lies in the area bounded by the solid line in Table 3-15, and will range from 10 to 55 percent £or the area bounded by the dashed line. If the solution involves charts from both sides of either, the dashed or solid lines. the prediction error will range from 10 to 55 percent. The large interpolation error for these charts result from the big change in ~/XE for a small change in P/r u which is peculiar to this chart,
3-97
TM 5-1300/NAVFAC P-397/AFR 88-22
It should be noted that the large errors descr~b~d above will always be on the high side, that is, the predicted value of Xm/X E will always be greater than the exact value.
Hence, a conservative design will always result.
In
addition, it should be noted that the prediction error in tm/T N is about half the error in Xm/X E. . " While the errors produced from interpolation between the charts in Table 3-15 bounded by the dashed and solid lines is large, ,the .area in which they apply is small.
As can be seen, region C is rather small on these charts.
for points in this small area where
interpol~tion.must be
performed.
It is
For the
remaining areas (A, B and D). the approximate solutions may be used rather
than interpolation. These less than 10 percent.
approximat~
solutions will result in a error of
Solutions involving 0.920 < CI < 1.00 and C2 ~ 100 can result in very large errors if interpolating procedures are employed. In this region, the dynamic response is primarily due to the gas load. Therefore, Figure 3-64 ,is used to obtain Xm/X E for the gas load. However, this,value is too low and should be increased by 20 percent. The value of tm/T N corresponding to the increased, value of ~/XE is then obtained from Figure 3-64. 3-19.3.5. Determination of Rebound In the design of elements which, respond to,' the
p~essure
only and pressure-time
relationship, the element must he designed to resist the negative deflection or rebound which can occur after maximum positive deflection has been reached.
The ratio of the required unit rebound resistance to the ultimate unit resistance r/r u' such that the element will remain elastic during rebound is presented in Figure 3-268 for a single:degree-of-freedom ~ystem subjected to a triangular loading function. Entering with ratios ~/XE and t/T N previously determined for design, the required unit rebound resistance r- can be read in
terms of the originally designed ultimate unit resistance r u' .To, obtain: the rebound resistance for an element subjected to another form of load other than the triangular loading fun~tion, a time-his~ory analysis such as the one described in Section 3-19 has to be performed., It may be. noted that· if' the loading is applied in a relatively short time compared to the natural period of vibration of the system, the required rebound resistance can be equal to the resistance in the initial design direction. When the loading is applied for a re La t Ive Ly long time, the maximum deflection is reached when the positive forces are still large and the rebound resistance is reduced.
3-20. Elements Which Respoud to Impulse
-.".
3-20.1. General When an element responds to the impulse, the maximum response depends upon the area under the pressure time curve, (impulse of blast loading). The magnitude and time variation of the pressure; are. no t. -important. The response charts presented in Section 3-19, ,which are based on pressure-time relationship are therefore not r~quired for these problems.
Instead, the element resistance
required to limit the maximum deflection to a specific value is obtained through the use of a semigraphical method of analysis. 3-98
TM 5-l300/NAVFAC P-397/AFR 88-22 !
•
Consider the pressure-time and resistance·time functions shown in Figure 3-
269. The resistance curve depicted is for a two-way element with a resistancedeflection function having'a post-ultimate range.
From Newton's equation of
motion it can be shown that the summation of the areas (considering area A as positive and area B as negative) under the load-time curves up to any time t a divided by the cor r e sponddng effective masses is equal to the instantaneous
velocity of that time:
('
va 0
(f-r) dt 3-87 me
J
The displacement at time t a is found by multiplying each differential area divided by the appropriate effective mass by its distance to ta'and summing the values algebraically:
1'"
Xa
J
(f-r) t] dt
[t a
3-88
me
0
In each range the mass is the'effective plastic mass :. Time Interval
Effective Mass
Resistance
' ru
O:S t :S tl tl :S t :S t m
.~
r up
- [KLM
~p-
J ~
[KLM:] ~p
Time tl is. the time at which. the partial failure deflection Xl ·occurs. and· time t m is the time at which maximum deflection is reached ~).
Xm
For an element to be in equilibriwn at its maximum
(Xm<
deflectio~l
its impulse
capacity must be numerically equal to the impulse of the ~pplied blast load. With the use of the foregoing equations, the expressions yhich define the
motion and capacity of elements subjected to impulse type loads, ~an be
defined. These expressions are presented for both large and limited deflectioncriteria. This criteria varies forthe.different materials. used in protective .des Lgn . The.! criteria for each material is obtained from the chapter that descri!?es the design procedures for that material. Case 1 - large deflections. Utilizing Equation 3-88 and by taking moments of. the areas under the pressure-time"and resistance-time curves (Figure 3-269) about time t m• assuming! that the unit blast impulse i b is applied ins tan., taneously at time t-O. and' that, time to reach yield t y is also close to zero. the expression for the maximum deflection is
" 3-99 I !
il
,.
"
TM 5-1300/NAVFAC P-397/AFR 88-22 : ib
JSn
t
r u tl [tm-tll 2]
m
-----
~
3-89
In,1
If moments of the areas are taken;aboutt l, then the deflection at partial failure Xl' is r u t12
3-90
Using' Equation 3-87 and summing the areas of t m and recognizing that the . instantaneous velocity at t m equals zero ..
3-91
- 0 .~
Solution of the above three simultaneous equations is accomplished by solving Equation 3-91 for t m and substituting this expression into Equation 3-89, and solving Equation 3-90 for tl and substituting this expression into the modified Equation 3-89. After continuing and rearranging terms, the general response equation becomes
3-92 The left side of this equation is simply the initial kinetic energy resulting from the applied blast impulse and the right side is the modified potential energy of the element. The modification is required since the above.analysis requires the use of two equivalent dynamic systems (before and after time tl)' The modification factor ~/~ equates the two dynamic systems. If the'effective mass in each range was t~e same, ~/~p would equal one and the right side of the expression would be ruJSn which is the potential energy . .For one-way elements which do not exhibit the post-ultimate resistance range, 'or for two-way panels where the maximum deflection JSn is less than Xl' Equation 3-92 becomes
3-93,
.. The above solutions are'valid only for what is considered large deflection design since the variation of resistance with deflection in the elasto-plastic
range has been ignored. This limitation is based on the assumption that the" time t~ reach yield t y and the .duration of the impulse to are small in
.comparlson to t rn .
4.
'
.
Case 2 - limited deflections. For· elements which respond to the impulse with limited deflections where the time to'reach maximum deflection tm'is greater than three times the duration to of the load, but where the support rotations are equal to or less than the established criteria, the elasto-plastic·range, behavior of the element must be accounted for in determining the overall ·3-100
,
1M S-1300/NAVFAC P-397/AFR 88-22
response of the element to the applied blast load. 92 becomes
. 2 'b
For this case, Equation 3-
+
3-94
2
where rna is the average of the effective elastic and plastic masses and XE the equivalent deflection. When the response time.t m of an element is less than three times the load duration to' the element will respond to the pressure pulse rather than to the impulse alone. In this case the response of the element may be obtained through the use of a r~sponse ~hart considering a fictitious pressure pulse as outlined in Chapter 2. However, if the element's maximum. defleption is greater than given in the response charts (ductility ratio ~/XE greater than 100), the response of the element may be obtained through use.of the semigraphical method of analysis as outlined in this paragraph considering the fictitious pressure pulse. It should be noted that the resistance time relationship uSed in the analysis to express the element's response should include the elasto-plastic region.
3-20.2. Determination of Time to Reach Maximum Deflection The time t m for each of the·cases covered previously can be determined by applying Equations 3-87 and 3-88 to the particular problem. The resulting equations are as follows: Case 1 - large deflections.
A.
Xl <
~$
~
ib tm -
B.
~$
ru
~p
+ [
~rup
, ]
- ---- [i b ru
2
-
2~rUXl] 'I'
3-95
Xl
Case 2 - limited deflections. Since the variation of resistance with time is not known in the elasto-plastic range, t m can only be determined approximately by assuming a linear variation of resistance with time.
3-97
3-101
--
--
~
f /)
CI
&.l
f/)
z
j
I
oJ :J
o
I(ln)
Q.
~ Q:
Z
0 l-
&.l oJ:
I(ln-I)
ce
IAl
U
Q:' IAl
U
ce
oJ
..
IAl
tn-I
tA
U U
ce
tn+1
tn-I
+
v:
IoJ
>
~ V;.l
I-
Z
IAl ~
v~
U
IAl
->C
~~
>-
o
tn+1
(b) ACCELERATION PULSES TO REPLACE THE GIVEN ACCELERATION CURVE
(a) GIVEN ACCELERATION CURVE
:-
tn
TIMEO)
TIME 0)
'--
(In+l)
&.l
U
Xn-I
ce
~
oJ
Q.
VI
•
C
tn +1 TIME (t)
TIME (t)
(d) DISPLACEMENT
(e) VELOCITY
I Figure 3-45
Acceleration-impulse extrapolation method 3 -102.
I
00
Z 0
...«
I
Q:
I
III ..J III
I
u u
At I At
-e
I
I
tl
to
At
At
',-
I
I
tz
'.
I
t3
t4
, 0
TIME --
,
( 0,)
I
z
•,
00
...« 0
03
At'
Q:
III ..J III
u u
At
At
« to
tl t
tz
t3
t4
TIME (b)
Figure 3-46
Discontinuities in the acceleration curve
3-103
... .
. " .'
'
-... ,
, ,
tj
,tj OI
.
, TIME
Jjo2
tj+3
(tl:>
,<
Figure 3-47
Pressure-time function w1th two different time intervals
3-104
·
~'X"OI.VI
.
I
RZ=KzlXz-xI) ,......:..-... toll
Figure 3-48
A two-degrees-of-freedom system
;.
3-105'
'.
2.0,....,....,........,....---..,.-....,....-...,--..,.-.,---....,........,........,...-..,._ _-...,
···::·r:::·::':
. '1:
::
.... .. .. r .............!... .. .
I
t
-.
,
::,1.:
.:::::: ·:::r
...o
,
.. ...
.
.
.... ····1 .. t·
..
:::·······:r
If
-,:,
. . •.1i.:.•..· ·.:.:..:.i.:.:.:.I.. :. I. . :•.•. . •.•.•.•. :.•.. . ~....
o o
..J
.2
!
.•.. :.:: :,"
e
·.·.·.·t·.·.··.·.·
:::\::::1-/:._::: :::.1
::
::::::;.;r::: .:.:: ::: :::: ····T
rI:::: :•::::::::1:: • ::1::: .:::::. ···rrr ::r:"
:i:
..• [ :
I
"1. . . .
I
;
;
•.. • • •. .
; ..• ,.
~
j_ .•
:: ::::::.• : : ·:::1:::: .
o
...
c
Q
::tr~ :: L: ::::1:::: ::::~:t:.:::: .::: :::: .... =t --: ":l' --+-L:::: - :::;:::: :::::~t:::: :::: :::. ·~t ~:[:i ~.:'. :::1:::. ....+- ::::::::t:f _ j....;... -. t..; ·..·,·····.. ·,··1· +.+-
0.5 ..::
o_
0.05
~
.__..... - - .. ~
0.1
.-!
0.2
~
+
····1
+.-:...
0.5
2
5
10
T/TN (0)
Xm
tm Displacement Function
Triangular Load 1.5
..
,i . .:: .r
.. ':
. ~ ..
.~
0.: ~::: . t!:.:t-·ttl..::~·T·:.jti},
;.j.
"i'
0.05
0.1
0.2
0.5 TlTN
. . . . I ...
2
5
10
(b)
Figure 3-49
Maximum response of elastic one-degree-of-freedom system for t ri angu 1ar load 3-106
..
I
2.0
r----------...."...---,--,---,--.....,..-.....,.......,..---, , : ,.I.r ····~rliTI
.II/:.!, ••••.••• :Ji.;~ ••.••••. 1: .• l
.
'0 .~ '0 ~
o o
..J
E o
1.0 •••••••1••••'
•
\! .. ······1,
I
! I
· I
; .. ': '•• 1'" ···i' ! . Iti ?·•••.••.••• _ · , ,:L!:;·:,li
..........+ .:........... ·····I··L+
,IL
.2
II
·1 .. ·.,. .. t..
i,: It. ·.·.·.· ·.·.·.·t·~.·.-:.· .,./
+
..
. ·.•.•.•.••• 1··.·
• I,
L:'7!; .• .·•·.•. t+-~=J.;'.· ,- ······J .... 1r:.1~
•••••••.••....•.1--
: /.:'
I'
.•......·.; ·1...·. .... ····i:·..; I 1•.•.•...••.•.... ···i : .; ; j..:..-. ... . !,. ~
-
.-
0.5
• •.. :
!•..._....
.
I ••• , ••• :.LL
,
'1 "'-"-" ... J....···i·-;...............
~'VFE" '1
-,
........' ...;.1 1.. +; , · .....•.•,•.••••••!.j -lj ···1····1····:·· ·C.~::
.. :
c ,..
o
..
.; •.....:::).:.- ..•..••..t+
..
~ ':i+', ::::" ' .::; .t ,..+- .::: ·:::T~
o =~ ~rf:f. 0.05
i:<: I: . ··;:::L~.· ·tfl
O.L
', !
, 2
0.5
0.2
5
10
T/TN (0 I
:::::::t::.
·,..·1
..•"'1"".•-
.,!, ..
p l - - -....
"1"
2.5 T
I-
......
-
E 2.0
tm
Rectangular Load
-I""
1.5
:I. 1 . :! .. :
1.0
., I
!
0·T:r-··-t7n7:·t~t~~~~·: :1
0.5
0.05
0.1'
I
..
.. ' f '
0.2
0.5
Displacement Function
,i
., 2
5
10
TIT N (b)
Fi gure 3- 50
. Maximum' response of e l a s t i c, on·e-'degr;:;e... of-freedom 'system for rec t anqul a r load ' .• : 3-107
2.0~------""'-""""""",,""'---r----:--------.,
....
.,
o <>
If
...o
E .0 c
--71
p-
1.6 . :
:.:" ..
"::.:::::: .. ::: .r., ~ :;~: ~ ~ ~;
...,.._._. _.. _._.',-+~-':' .. ..
Xm .
..
1.4.·
... o
Gradually Applied Load
I I
·····:1:::.:··
.3
.!:!
••••
Displacement Function
.. i":
. ::"
1.2
1.0'--
~
....x_ ___'__
~
__'_..x.~
o
~
4
3
5.01"'".~-...,.....---....,.......,...--...,.........,......,..........--~-------.......,
._ _. __ __ _.__~~_-:...i.._ .._ .,__.. :. . .:.:. 4.0·
~
..
.....
E
-.- -
,
-,
.......... :
3.0····
...,.
.~_._.:.....~
.:
.._.._
.
--.c···---,-····-······-·-·c····--······ .. -.,
:
.
i_._....;.... ~.i_ .. ;~.~._._.:
~ __ :
. .... -... _-_.:.... _--_.:,_ .. -......
2.0
1.0L
o
_ _:::::::'-L_:::::::::::-JC:::::::===::::=o..c===-l 2·
3
4
Tr'/TN (b)
Figure 3-,51
Maximum response of elastic, one-degree-of-r'reedom system for graduillY'applied l~ad . 3-108
e
e
~i'?i;
-
1.5
!
I
'.
.J.:'; .... '
'.';.'
,
.1.
~
.
"
\
~.
..
, ,
.-
.~.~
-
.--
.. .. ~
-
" ..
.. ..
.. I
,
,
....
f -
:
.2 1.0, u .
,--I ,
If
i
- - - -.
-
.~
./ ~
-
~
,....
I
w ,
-
o
o E
o c
>. Cl
5
:
,
,
...J
..
.
-
\
:
\
,
-
..
..
..
'.
..
.
..
,
.
.
- -
-
•,
,
.
,
-,
..... ;
,
i
,
.....
'"
,
:
't:J
o o
'.
.-
;
J
/1
,
-."
.
.- .
.
,
..
.,
" t
°
'.
-
.
-
..
:
,
, 0,
~.
2
3
4
5
T/TN
Figure 3-52
Maximum response of elastic one-degree-of-freedom system for triangular pulse'load
.. 2.0 ..
r
r-,
1.6
..
0 0 0
u,
1.2
L-
1\
. I .~
""- i
____
'tJ
.
0
w ,
,.... ,.... 0
. .
0
'-'
.
- t-
.
.
~
:.-1 0
E
0
c
>.
0
.
.8
II .
,
.4; II - - - .. '
,
,
1
,
. I
,
,..
.
.
.
. .
o
o
2
3
4
5
6
7
T /TN
Figure3-53
Maximum response of elastic one-degree-of-freedom system for sinusoidal pulse load
0.5
0.2
5
10
20
0.7
·0.8
40
1.0
10
5 llJ X
.....
E
X
0.2 T TrIangular Load
0.1 0.1
. 0.2
2
0.5
XE
tm
ResIstance Function
5
o rsplacement FunctIon
10
20
40
TITN
Figure 3- 54
Maximum deflection of elasto-plastic. one-degree-of 7freedom system triangular load
3-111 .,,
il
fo~
0.2
0.5
5
10
20
40
10
5
2 I .
2.0 0.02
T Trlan9ular Load
f:
Resistance Function
1m
Orsplacement Function
0.01 0.1
Figure 3-55
0.2·
0.5
2
5
10
20
40
Maximum response time of elasto-plastic, one-degree-of-freedom system for triangular load
3-112
-,
02
01
tJ:
5
05 '._
fU
?
20
10
40
cf-o.
/P =0.20
0.40
.0.60'
.,
50 ~>-" ~u,.EHJ'_ ,. '. EE:i~~ ..~ ..... l~ . _.
-
1.00
6.80
,~~q:.....::.
._ ,r"
'
..
r ~ __ ';r.
:::::
-."
"
/P '1.02~
fU
20 "1
I
~.
.'
+ .
"
- 1.05::
10 10:
ILl
5
,. "
'/'., .... ',.~ ........ 1+, ·_·I'..f :... -::-, ~;
... - -"
life ,Ii,
:liW
~
1+t
~
to_ ~":"I.;::~
::'::..:::.:: I. 10.,' ..:~'" ~
....: "IF-" ".Ii ...
--
r
X .....
·:-::::-:i,;.1.20 .
E
X
l~~~~~:;t1.30~
§.• • •
~
i=
2
"
-
.:::;;;::::"
1.40~
';';
1.60:';
"
Ill!= ffiE
.--
m
0.5
,
t=,
'-1-'-.
.
~., ~
,-!=C., ...
'n T
Rectangular Laa~
0.1
0.1
Figure 3-56
0.2
1.80 Elastlc:::C 2.00'
o
q
I'
n:
rul-~
L/! xE
tm
Resistance Functl~n
0.5
xm----lJ
5
Displacement Function
10
20
40
Maximum deflection of elasto-plastic, one-degree-of-freedom for rectangular load
3-113
s~s~~m
20
IOilll O~.I_~!! 0.2
0.5
10
40
,r'
0,50 tilt"
0.60 1I11':HL.!1
0.80 1.00, , 1.0,1
I-
......
..
E 0.2
l~ Il ................
"i
"
0.05
.,':
"10001=-,,
----
X
1,40,
,., '"I,,·
~ I. 60 !
.2.00'
T
ReclanQular Laad
1.15 '11.20 '
'Q
0.02
1m
E
Resistance Funcllan
ur
Displacement Function
~~
0.01 0.1
0.2
0.5
2
T/TN
Figure 3-57
.
1,10
0.1
5
10
20
40
"
Maximum response time of elasto-plastic, one-degree-of-freedom for rectangular load
I 3-114
.
02 .
01
WI I
20
10 B
,
<,
r :
1.10
5
.
I
I
:
.
!
1.20
't---1,
w 2
X
.....
1.30 I-l'"'"
,...
,
1.50
E
-J
K
'-2.00
\
1\
..
-
0.5
~r-.
./
t--
V;
Tr
O. I 0.1
0.2
I
;
0.5
I
:
,
I
I
-
~
I--
l-
;
I I
i
--
1
I
f-- f--
/\
:, --, <,
~
I~ I:'"
~
Xn 1-
f--,-ru/P=
1.02 I - 1.10 ,...1.20 1.50 r-
c-
.
-
71
"
2.00J== 1
: I I--T-
,
tm
I
Displacement function
I
I
I
O.B I
2
5
"
I .
.
_I- -I .
:XE Resistance function
I
, ,
:
111
Gradually applie~ load I '
I
ru I--'
Pt---
I I
I
j
ll-
I
71\
./ "
,
;
~ /~ ......... 1'\1\ l\ /;}~. r-, r\'1\ 1\1\
I
---t--
"
;
,,'
1\ I~
......
0.8
0.2
i
I
,
I .f - - J. -'
r'\.
r-.... t- ~
I
1.80
I
"
.;- ~
20
I,
'I' I I~ I---L- ~-1- -- K-I---I- - - t - - _...1_ -- -- ~- ,-.I 1 I ' I , I I .... 1.05 I '
!
X
;
i
II ru/P = 1.02 I
10
5
f---l-I I I
8 10
20
Tr/TN
Figure 3- 58
Maximum deflection of elasto-plastic. for gradually applied load
3-115
one~degree-of-freedom system
02
01 100
80
0.5
0.8 I -
'\.
I
50 I "
""Z_I
.
I'\+~o'
1
I,
'N2
1
\
~~-
1\
8
~
,
-
I"\..: I I -, ~<'!
"Ir-,
I---l--I
2
'\.~/o "",'(
, I 1--1--
I '
"!:
I"'\
i r-, \
*-"
1.0
I
1 I
I-I-
--1-"
I-- I-- 1-
I
1
,
, I
I
1\
\
1\
:
:
,I
I
1\ f\ f--l\I\ I-,.r-
. I
1
I
1,u / P: I.Ol - I -I-
~f'-~. ~,~
1/1.05~\.
1\
~~l\ N
I
I
I
..
: --+-I I
f---t'u / P' l.OO-,--
, I
I
I
I
I I
0.5 f-~
'u
T/i
'-
'0.2 ' -
0.1 0.1
XE
Tr
f-
,
Gradually applied load
I 0.2
I
Xm
[7
I
I, . 0.5
"71
: f--.1I
tm
Resistcnce function
'
i· I----:-r-
,
1\
1'.. " ~:+,' I" , I
I
I
-,
! ,,
\
'I
0.8
i I,
,
1
"
I"
,-
1 _-:t-- -r 1--- -
i- i - ,
10
E
20
I
--1-- -1- -- .- . "
: i ' l\-
20
; ;
i 1
.0
~---.:..~<-:,,<>- , -I -11--
..: .....
8 10
I
'\.
5
5
2
Displacement
I
1---+-
fundion
I
I
I
0.8 I
2
5
I I
8 10
20
Tr/TN
..'
Figure 3-59
,.
Maximum response time ofelasto-plastic, one-degree-of-freedom system for gradually appiied load
3-116
02
05
5
2
08 I
20
8 10 0.7.
T 1
,
I
1
I
II
I
II
II
1A
I
I I·
IJ
I
I .•
1/ LtJ X
..... X
e , I
-II
I.00T.....
,/
/
V
A
/'/ 0.5 /
1/
1/ /
//VI
"//V~ 0.2
V~~t
I V I t--+-f-+ I
I
I
.I"
0.2
Figure 3-60
__
1.20
r-,
:· · 'V\ tL , '" ~ ~y
.
rC-
I I
I
-_
I
t n.-2.
Xm - - - -
0.5
Resistance
,
'lunClion'
0.8 I
Maximum ~flection of elasto-plastic, triangular pulse load 3-117
T-
, I
.
00
XE .
Triangularpulse, I Ipa~ , , I .
1.60
I I ,
ru -
.
~l/....+...
-+--
1m
Displacement
. 'I
5
. lunClion
l
8 10
I
I
r--
t-I I
.
20
one-degree~of-treedom system
for
.
5
05 .
Q2
01 10 8
I
T I
I
I
Q8 I
i
~ ~ ~ r-;;:: . - -
..= .....
E
-
1-- j-- l-
1.0
-
1-
0.8
0.2 l1,.'
'Vi< i I
I
.- .-
+- ~-
1 I
,--1-,-
I-
0.4 O.~
1
,
I
~
'u
.
-
~
I . I .' I' ; .
,
"71
fUlnelicn
"
~I
,
2
,
.
I I .
-
XE
0.8
I-
.
~esislcnee
""
0.5
2.0
Xm
.
:
0.8-+=
1.0
,
Id
0.2
0.3
I
2.0~
Tricngulcr pulseiccd
. 0.1 0.1
'- '- I-
r
" ;:::~
l-
..
; ;
_:1_
0.6
I I
0.5
-
;
. r ..
"r
I-
i
~ ~I--.
!
r
I I
~.~+-- -1"-
-
~ .....
!
20
8 10
i
'u / P'
I .
I
5
I
:
2
2
.'.
~
L
L_
! ... .~
1m
Dis~lcepment funeticn S'
1-':"'-
.
I
8 10
~-
i'
20 .
td /TN
Figure 3-61
Maximum response time of elasto~plastic, one-degree-of-freedom system for triangular pulse.. load
I 3-118
P ,-_SHOCK LOAD GAS LOAD
CZT
T CZT REGION A: NEGLECT CHARTS USE EQUATIONS 3-80 AND 3-81
REGION B: .NEGLECT. SHOCK LOAD USE FIGURE 3-64
SHOCK LOAD
SHOCK LOAD p
GAS LOAD
T
REGION C: BOTH SHOCK AND GAS
REGION 0: NEGLECT GAS LOAD
LOADS ARE SIGNIFICANT USE
USE FIGURE 3-64
APPROPRIATE FIGURES AND INTERPOLATE.
Figure 3-62
.. Various billnear triangular loads
3-119
IE T
t
Xm
XE w ,
ReglonB
.... o '"
Region 0 1m
Tn
T/Tn -
Figure \3-63
Regions of figures 3-G4 through 3-266. labeling of axis and curves.
•
• .
e Ir/r
flME or YIELD TO LOAD DURA nON: ':'"
aO
V"
0
~ S"
'7
0
-:" --: 0
--:
v"
0
o
'cP
0
a
0
.
0
0
0
0
0
0
'J'O'
-oJ'""po.P'
....,..
......
0
0 d' 0
V'
0
0
0
.,....,
0'
0
.~
.~
v'-
0
0
0
.....
:-\
0
....
1.1'
0
'....
.....
v'....
'0'0'0'0 J) fIJ _1 d" 0 0 0 0
-
'0 1.1'
0
0'0'0'0 ... V" V" t-J 0 1.1' 0 If\
0 t'..)
0
··P.PIl\ff\ r \ (
,.,
&
f\ u
><
'AI~
E
><
a
'" W
S 'J'
Ill,
0 f-
U.
-'
u,
'I.
~
\.111'...111\/
L.J
o
~NJ
0
-'
L.J
>0
~
rZ
0
i= u
I-I
'6'
w :>:
~K~~
":0
d
.f. s
cf-
~
0
~'6
a t: o
..1
<.J
W
1.7
-MX,:T,,",'iI:r"AI,,~;I!"
w
1.5
0
_. I. J
:>:
=>
:>: ><
« :>:
~u~~!um o.71~;F~[;'-~
r>
/I/I/I/V
o. ~
.00 I
.002.00.1 .005
.01
.02
.0.1
.05 .07.1
, /H//.Y//V .2.3.5.7
235
LOAD DURATION TO rUNDAMENTAL PERIOD, TIT N
Figure 3-64a
'"
Z
Y'
<.0
-' c-
-' f-
-
:::DX
W'
.... .... '"
0
.,.,
6'
Q.
z
...,,
/
~
<,
...E
->
<,
2 f-
Maximum response of elasto-plastic, one-degree-of-freedom system for bilinear-triangular pulse (Cl • 1.000, C2 • 1.0)
7
10
1.1 0.93 0.81 0.70 0.61 '0.53
-c
z =>
o 2
w
-' c-
w 0
:>:
=>
:>:
>< « :>:
u,
0
w :>:
c:
TIME or YIELD TO LOAD DURA TlON. 000 000 000
'£IT o
0 0 0
o o
__ _ """"
o o o
o o o
N
c-:
...z
<,
E
'::.
~
><
o
<,
o
E >< roo
<>:
w
n,
z o
-'
...z«
;:: Ll
w
w
-'
L.
C\
,..,
'" '"
o
.....
o
~
. fit
e_J
,
<.>
:::t 'If
O.
w
Z ::J
u,
...oz
7
r
O. 6
Co
Z
o
O. 5
o f-.
t: Ll
w
o
...J
u,
w
w
c:
o
w
... ~_.t,..._\. tin.
a
:::t
~~>_._
~
::J
::f
sc
~
::J
-R
:::t
sc
«
I\ - \
:::t
....
U./~
1\
o
w
0:J.o~~ ~ .~ ~
:::t
,J
0.7
;::
r==-. --'--- =f-=1=-f-=aE=~=1=Ww~t==I=$m -
1 - - ----_..-
.-j--jc-+-J-.-
0.5 \0
20
'0
~o
10
100
200
JOD
~oo
700
1000
2000
JOOO
~ooo
LOAD DURATION TO rUNDAMENTAL PERIOD, T/T N
Figure 3-64b
Maximum response of elasto-iJlastic. one-degree-of-freedom system for bilinear-triangular pulse (tl. ~ 1.000.'CZ ~ 1.0) (cant.)
toeoe
e
--
• or
TiME v
u'
~
'":' <1'
V'
,.,
/
/
<1'
0
0
.'":'
YIELD TO LOAD DURATiON, 0
0 .~
'"0
0
'&
0
..
<1'
0
0
0
u:,
0
0
..
0
.'":'
'":'
U'
'£IT
0
0
0 /
e-
/
S0
SOD 200 w
»c
.'-..
E
>(
z 0
100
'S!ll'\'!,' . y' ,,' X· A \ I .1'ir.L.L\,. .11/\ I v I ~I\'" ell J V W!K.l ' /, I Y. II 1/\ OJi'"1Y'~I\ " , / / \/ \ IV 1,/1\ _V /1/1)1 1'\ ....'ljil. ..f\Jti:' tIl', . 1\/ l.L\ /\ / ~j L\ f\ . \ 1\ 1/\ ,~L II \ fI II' 1\ V IV 1.!' \ K' J IV XLI\ I/IV I/\lt':i: IV 'IVI,'II . ~.
10
lJ
w
50
.0
w
,
.... w '"
:\T\ . ["1\j)( I--'b" '.x 11'\ 1\ .;
SO
a
I,L~
,~
w
20
,.0'" z 0
;= lJ
w
321,\',1<;.' ~' L~l ,Y X
\0
~;
a
::>
"
5
::>
Y
Yo if t
1\ 11\
r
1\
-VI\ ,.
Y
\
~..l,...L.::>..1'.
'
V- ill
\1
IN.V \/ \
.
.
'\
/ If ~
. ~V
'l:>.L
J
0
"N' :
bt
V'
~r--xf1l)r',vll ~ r"'\ \i " [ \1 t:::!....
IX.
'!~ .>
C\
'1
6;~
S:,j-
L
l"X .
t"I"""
IX
KK
f'...
\>(
•
.... K <,
0:.....X.~ I'\f'.~ I
0:6'
r- - . . t'-!V
I) '):
1" V
\
\.1·\
'!'Z'"IV
LX
~~ Y[L<; z 0.)
\
~
\/"--}'"
J
!\
4W y.."" .
\I,..lX Y. rx '> ix bll2(Avlx / J' 'Vt--.tliY'" ,Ytyf.l'/J 1'\
2
0.1
\
'1' ,
'~~
J/ /17" .r~.LJL~ ..¥_ j 'lL
,001
.002.003
.005
~lC~~ I)
/
J
L\ "\
~ II Ifu I'\'-
r-
/
0: 0
J. s
I'
.f
r-
~
t\!b«b(\ r\.Kv
~
Q -:;: >Q
K f>
.
Q»<
P-L/:7f.r; .0.
.02
//1/
03 .. 05.07.1
//Y/
.xl""'"
1·5
./
,)-<
1.3
I.'
I \
~1
~~
L
'
J.7
X T"-
:t:x:)(~><\~~kC
~=!"'
I'L
p/ ru -
,
'"
,2.3.5.7
_ LOAD DURA'TiON TOTUNOAM[NTAL /,-[RIOD,
Fi gure 3-65
o
~'6 ~'.J
J..
\
1'9
....2
<,
...E 0 0
oe
W 0.
AI X1/\ r-\r',J J JI,
1/-)
0
0
~·O
..:', \ .\ \
3
>(
-c
:y--'tf
~
.
0
'" '" '"0 '" ~ X" \1\ ~ \L~ V"\ ,
<1'
1
:>c ::J
X
/\ °1
'"
.
0
0 <1' 0
0'
0
A I
J\.""
IV J IL\.. .il\ rJ't>.
L
j
I
I LY,
.\
A ·'X
\.-'
--'
L. W
-,
.\r,
.--' L, W
:.J'~
I'
;=
0
0
-0
u'
I
2.l
ITN
T
.: Maxi mum response' .of.e.1 as to-p 1as ti c , .one-deqree-o t-f reedom system forbilinear-triangular'pulse (Cl = 0.681, ,C2'= 1.7)
5
7
.0
0.93 0.81 0.70 0.61 0.53
--'
....
w :::£
~
::> L.
....0 z
0
;::
o
w
.j
L.
w
0
::> ::J
:::£
><
4-
0
w :::£
.;::
'ciT
TIME or YIELD TO lOAD DURA nON,
•,J
v·
u·
,., ,., V'
0
-:V'
/
,.,
0
0
0 .V' V'
.J
"'0
0
0
0
0
V'
0
u·
~
V
0
,.,v
0
0 ·"'0
-0 /
0'
/
u"
0
0
0
~
0
0
0.;.
0 V' 0
0
0
0 .... ....V'
•0
0
u·
N
0
Z f-
<,
....E a
.,.,
6'
o
'"
w
n,
...J
Z W ~
« a z
~~Vf#'-~t-tt;zm
"", .....
... I'>
~,
R
A
r 2
" r ,
CJ
ofZ
o
;:: u·
f--·-
·w
u
...J
"W
w
--'
a
,~
W
~
C\
::>
:>c ::> y
:oF
><
«
><
~
"-
:i
o
w :oF
c:
..LJ....-L
__L-I.L-.L.JC_..-:"J
_ILl_J£J~-A
"~7LJLY+:J4-A4~ r Jf//V/Y/ ....
.002.00;
.005
.0 ,
/1"
.02
/l
0;
/1
A
~_I~J---l----.J
y
_
1/1
.05 .07
.,
.2
.;
.5
.7
2
.5
lOAD DURATION TO rUNDAMCNTAl PCRIOD, TIT N
Fi gure 3-66
Maximum response of elasto-plastic, one-degree-of-freedom system for bilinear-triangular pulse ICI = O.464,C2 = 1.7}
5
7
10
.-
• ,
,.,
~
v· v·
,., ,., <1'
T1~(,O~ YI(,.~D !O'LOXD
.', '0
0
0
"'<1' "',.,
'"0
0
.~.
0
0,.,
'oj' oj'
~
c:
oj'
0
'dr,
DURAnON, 0
o'
'S
',j'
0
~
0 ''''0
.0 /
•
e-
0
0
".r
0
0
-0
d'
0
e>'
oj'
0
'0
0
<1'
0
~
0
V'
'0 N
U'
N
0 N 0
Z f-
<,
IWI
•
~/\
1~1
I \I I 1lLLJ..1
V'~
n
iT
z a
..J l_ W
,
.....
'" '"
w
>-
'" -s
'J"
...".
,-
-J
'"
:>'
'" ...'" o Cl Z
--t't--,1'I+--\f-T\~lS
A----l\--
f-
,.a z a
W
Q.
Z
\
lD
'" w
X I XIX"r c:tl' 1\ 30
Cl '0
f-
-cr
W
w
6'
IJ'...
"f:' u
a a..J
...E
Z
o
IV
10
u
f:'
Ll
IX
W
X IX I V X /
..J l_ W
a
w u,
.-J
w
"
o
:>'
'" ...o'"
:>'
=>
:>' ;(
~~n~~{-Ai~
:>'
sc
:>'
'"
:>'
w
:>'
i=
o 7 . G!--7f::lfV:t/~"
~'I;-' 0.5· . ' . DO I
.002.003005
Figure 3-67
11~~7' 7f;LV-II ¥/f"._.A-. A-,f--H-Yb.L,.LY'744L?h /,V·
5
7
10
TIME or YIELD TO 'LOAD DURA nON, p
'",
59 0
oo-
'"
.V'
,., '0
/
0
/
'.1'
'",
0 'J .
'" 0
0
o·
'&
0 p
.1'
0 ..
0
'S
0 v' 0
.
0
'",
0
'",
0
tclT 0
/
0-
0
>v'
0
0 /
0
'" 0
".J.
.
'0
'0
'0
0 .1' 0
0
'" '"'" '"
V'
0
0
J 0 rNI~:>A-~d."kfl-::dA!=f~8td='lo!+
Lon J
w
"E "z
z
d'
'" '/
10 0
..-- . . -
70
Ai7\.
;::: U LJ
L.
II'! I
W
V I V I V
a
1\
Cl W
a'"
0
...
-- ...
¥-
7J
S
'of
-
':6' ':0 .t
1<
~-J\-'J'
>.
1'-
I-
-J
).7
'~
L.
1.5 1.3 1.1 0,93 0.81 0.70 0.61 nlU 0.53 "Wtl4lq
W Q
:,; ::> :,;
><
"" ::f
~
...
.05 .07
.\
.2
.3
LOAD DURA nON TO rUNDAMENTAL PERIOD.
Figure 3-68
.5
':7
I
2
3
TIT N
Maximum response of elasto-plastic, one-degree-of-freedom system for bilinear-triangular pulse (Cl = 1.215, C2 = 1.7)
5
7
w ::f
f--
?Q
w
Z
.J. o
u
"'
03
-J
::J
0
t:
.02
'"c,w
s
?(i ?,I .
;j<1l"
.01
O.
f--
-J
,
."
6'
~
50
-J .
W
)
a
<,
0
f--
10
L.
0
Z
0
t: u w
...J
L.
w
0
::f
::> ::f
>< 4: ::f L.
0
w
:::f.
c:
•
•
TIME OF YIELD TO LOAD DURATlON, 0-
"r-J
..r
.v'
,., ,., .p
'0
"'.p "',.,
0
0
'0
'...>
'"0
0
0
.p 'p
•
0
0
'0-
.p
-::,
.,.,
0 v'
0
p
0 .~
0
0
"0'
•
'ciT 0
"v'
0
'0
c1'
O.
0 0' .p
0
0 v'
0
.p 0
0-
....
0
:.~t%tM*=*-fA--~J!.
.... 0....0
0
'"
.,.,
d'
z
f-
<,
w
--'
...E
6'
a
'/
><
<,
Cl
m~AJL-AJ ~
E ><
',,>
.s :.:1
,-:::
.
u
w
-'
, ",'\ \1 IVlvoS'.7\3W
,~
w 0
ca
,...• ...'"
..r, a
>--
z
.=
10
u w
'V 1\/ 1\ I
-'
L W
a
:>:
~l
X,I:Ki11 J
1,7 1·5 1.3
::J
:>:
>< :
01
.02C)505 .07
,/ ..r/LV--L"'Y ,I
. 2 ' ,3
.5'.7
I·
2·
3
LOAD DURATION TO FUNDAMENTAL PERIOD. T/T N
Figure 3-69
Cl
<·0
....~,~fh-i~~
1~7r/'
~O'
<'Ii <.j
I\A
Maximum response of elasto-plastic, one-degree-of-freedom system for bilinear-triangular pulse (C1 = 0.147, C2 = 1.7)
5'
7
10
Z
w
jo
<,
..J
'6'
J.s~
nwtR~~~iJf\
~
o
w
Q.
z a
w
a:
:>:
z
::J
L.
a fz a
t:
u
w
..J
L.
w
Cl
:>:
::>
I. " '
:>:
0.93 0.8' 0.70 0,61
L.
0.53
r:::
><
:>:
0
w
:>:
riME or YIELD TO LOAD DURATION,
..
v'
'",
soo
v'
'" '" '!'
/
'0
',p
"'",
0
W";T',/"--lK--' '''I. -'.\ ..J...
,cW'§( ", .. SOO, ''\")
.;(
) z o c"
u
w
if..'
'" .....
..
0
0
v:.
p
'&
0
0
0
,p
0 u'
0
,-v' .... ~ \~.' ..'-'f':l'~' \ /1\
'V
rxrx 1\
I' I·t;;
/1\1
v
'X'!' I~
I)(
\/
\
.li
100\J:~~\~- J\i\: ~-/ ~ ;, '/I~ so
w
'" '"
r'
o
i..
z o .... u
w W
',.,0
p
'0
0 /
'"
.~
0
'0
'0
rJJ
0
~tx
~I; I~II
~
.
'"
0
.....
V'
.....
0
'"
'0
..... 0
~'T')
~Y\"
rJJ
'"
.
-' /
~~
'"., .
.
'6'
10~:: , j-
1\ I f.
'A~
K-Mro
-':0
~l<
..
;)J\\[V
~I
<.f> <..1
)<
I.,
1.5 1.3
S
2 't'"':~:::-
o.s
!4-~~/1/
!~--,~
.001
.002.003 .OOS
.0,1
It
• , ..........-
p/r" = 0;J>q
tm
~.Y./"""
.0203 . . OS .07.1
.. 2 . 3
,5 .. 7
2
3
TIT N Maximum. response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (C1 = 0.100. C2 = 1.7)
. lOAD DURATiON TO rUNDAMENTAl PERIOD,
Figure 3-70
'"u,
W
w
....-c zw
~
5
7
0
z
....0
"
\/1\11\ f.
5
'I-.A-
0
0
.1. 0
><
::l'
E
J:,s
~.
::> ::l'
:>
<.0
'yrv
....Z
<,
..:
u. W
'0
0
0 ,p 0
'..f
1Y~~I1\
7
·'0
.t.
~/\1~
-'
o
0
'",
\l.'t-."''X=~~\~r:- ':/\ PI " / 1 / . '\~.f\. k-\ 1-"", .. f--\I \~X -~P-\;r\-f, -~ .' . \71D1\'7.q,A~1 ",II
70 50
c.,
o o
'"'
0
'...>
200 ,V\ _..J,,\\ '\ ¥I " .L /..',.1/
. ,.J W
\~'''t'!'
0
iciT
10
1.1 0.93 0.81 0.70 0.6\ 0.53
L..
z
0
i=
u w
W
L..
w
o. ~
:>
~
x ..: ~
L..
0
w
~
i=
•
e
,
TIME or YIELD ro lOAD DURA TlON. "r-J
'v'
·V'·O
'J'
c:'~
o
'?""
0.. t?v" OOV'V'-lO
P"v"r.:l~//
C?s>
't-l
']%=-W-W---V-fV1 ,'"... .. I ~
0
',.,.. ',.,0
o
0
''0'
•
.
'ciT .~
o
'0
"'0
~
"':7YJ:YlY I A\
'0.
~
'0
'0
'0
..o
'"o
'--V+/..1 /
or ,.,
o,..,
,..,
0
'"
0
~"'/:'\
F\\bt=t~ft~.f=¥cF#=f:*=jj I
z
>--
'"
<,
->'
VI\
~
'"E
'-.
~X
><
'/
a
V\J !'U IV
6'
',0
s.
o u
W -i
\I\l-}\h'YnTI \II\7l\'1~7\.. 1/
u. W
, ....
'"
'0
30
',f
.. ~~~
>-
w
~
'6'
-'
Z
w ::f
a
z
::>
_J
>-
'"
Q
I-
n
w
o W
z
o
...E
L-
20
-'¥fs;A'\=i'1~tlE'tt5~M~
,-o z o >--
u w
--'
o
. MJ>:"_A 'V \v\M..V
r.. w
~
1\1
\V 1 A A Ii\
'11\11/\/1\/1\
I/\j
o
I-
I~ 11 Y
Z
o u
i=
~
Y \ noV-I X"!.'/'('-J I WI/,
LW
a
::f
::> ::f
=>
><
::f
>< <(,
::f L-
::f
o
w ::f
i=
y
0,5
.oo i
/
/
1,/
1/ .002.003 ,005
-1tt7'7Hn%lL .01
,02
.03
.05,07
~ ,I
,2.3
,5.7
1
2
3
LOAD DURATION TO rUNDAMENTAl PERIOD, T/T N
Figure 3-71
Maximum response of elasto-plastic, one-degree-of-freedom system for bilinear-triangular polse (Cl = 0.056, C2 = 1.7)
5
7
10
TIME or YIELD TO LOAO DURATlON .. ldT.
',..,
500
2 aa
J-
o
0
. 50
>-'
2
a
\0
0
0
.....
'" 0
'.,&-
0
0
,0
It \ 1/
T
\1\ AX ~,1\ ~l1'
',..,
v" 0
~
1\
.
,x
)f
-.I
~ dl
\ /
I\,T
n:
-
'\
II
\ I 1\ X \
'\
,/-\ /,-,>
V
\'1
\
/
-V
V\ IY
1\
L 1 J ""
/
y 1/
v"
'0
0
..J
I~IYII
y.\
~~ V
1\
)
~.
.....
'0
.....
'0
..... 0
v-
f\l.J\:rIV1 'I
d'
.,..,
~ -I·
I)
'",
1\1
.,..,
l'Y~/lYoc-ru\
Ar~v
I
7
.t.
K A\.IX\ \K '''~
~r"Vkl< , VO'
\:
~
<'6
f6tjz---J:4-L~ .002.003 .005
.0'
.02
.03
.2
... 3
'.5
.7
\
2·
3·
I.J I. I
0.93 0.81 0.70 0.61 0.53
LOAD DURA TlON TO rUNDAMENTAL PERIOD, TIT N
Figure 3-72
-)
Maximum response of elasto-plastic, one-degree-of-freedom system for bilinear-triangular pulse {Cl = 0.032, C2 = 1.7}
7
L-
a fz a
;:::
t.7
1:5
5·
'"
::>
w
~y
.05.07.\
«
u w
1\
2
s
~
<.J <,0
v
~
..J
a
./.0
r-Ir'
~
5
1'-
'""-w
~o J
,¥ \'IV
o·
.,..•IS'
;;:
....E
« fzw
',r
1'\
~w
~
5
Z f-
0
IS' -
~lVfA~
~
::>
'"
0
<,
1\ 1 f':I N 1Ai7't--
L..
:;
~
0
"\f-I-.,{j 1 J
'0 v'
0
0
~
.I 'JI IV IV I If\:
f-
~
0
~~ . .. ~. "~i~~
1\11
r11\~%X=t~~\I¢'\~1r=K~l\;t'~c\~/~I\=X:1~V~"/~~~~J~lJ iA '1r V i ' X J\ II /\ I 7' ,~/
'"
~
'\-
»r
....
0
0 ~
l~
1/\ 1\7[\
/\1'< , '1*'0)(
0
V\ !l\ V,\ l' \V VhI
K-J\-'I 1'1:
IJ
~
\
o. ''t,
0
0
~
~
X.X.7\ 71
A If\:: A~,-
'v /\
f-
~
\
70 ;-f\,v 50" IV \
~
w
0
',..,
'",
,I;:W'\:;>/
'\/ \
o
,
.-
~
.\V.
100
::
t;;.
'0
'" ~T~-\~,"'r:'V-'~- rr'\ 1\7 IT T\: ~ 'f/1 a'X ,;-17 - 1\ / \1\; \V 1\
z
:::; o w:J
'1.1'
v"
s0Q.( ~ ~ <,
,.., ,..,
v"
~
10
..J
L-
a
~
::> :2'
X
«
:2' L-
0
w :2'
;:::
e
• "NIL p'vy.Jy.J .~ 'v'
,..,,,,,. 'tJ; 'y.J
''!''o
~oo
f\ I
VI
.\
or YIELD
TO
LOAD DURAnON.
.~
'"p'
o
,v
IX I V
'",
0
'"
. $I'"
J'
'v'
\r
J
0
.~. ~
><
I
.x V\I /\1 / [\
100
f-
u
W ..J L..
·1 \L
IV A
y ... 1\/
f-
u
I /
Y
It-
/1\ 1)\
r-'
VI ii'
lEV' ;\~]lQv
R'txff~-
.
0-,
L./ r\rrtj
~y
....
-'
'/
\1;
,.,
_Lw.
s-',j-
V" \.\, 1\ .
\
.
• 1':L A JI .1"~~" (\ 'K l'\j
I'X':XV ~'"V ~ D:
II,'
"
~V'~ I'U<
W
::f :0 ::f
>< ~
"'0
s
.~ 0.53
"f / y;,,v :.:.j~~l7f~::-:: :::Y}'lXI~_ .001
.002 003
.OO~
.01
.. 02
.03
.05.07.\
.2.3.5.7
2
3
LOAD DURAnON,TO rUNDAMENT AL 'PERIOD. T/T N '
Figure 3-73
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (Cl = 0.018, C2 = 1.7)
5
7
\0
..J
ci Z ::>
1.7 '.5 1.3 I. I 0.93 0.81 0.70 0.61
~ w
11.
f-
<·Ii <..1 <.0
'"7I\I-ttJ-~~*¥~'4\;
0
a
Z w ::f
,fo
..
1\1
..J L..
<,
...E
'6'
.f
.. \-
W
z
f-
0
6'.
J
v L.1~
J.
>-
a z a
rv- ~l.,7t 1,-\1l< /
T'j
0
7' / \ f"-.. , I/~ Y\"'/){ ~
-0. _lI-l
t-J
lJ\
j
llv ,x
W
~
t..J
VVN !\
r\
o·
,0
,V"
,0
y
IY
0
..J
I Y I
'0
..
o
_Vy
W 0
1\
'0
t.1'
r\1 IAlL .1\ I/tfKR"Ji\
o
~TArl
'0
'0
~. ~
1/\ ·\1/11
.A
f-
!I[
A >- f\
.
'0
1J
v'"
v
V\ /1\
z a
'"," .... .... '""
'?
..,
.~
7\"
\II IVb( 1\11 I'll
w
E
..,
0
...
200
"-
'£IT
0000000000
.l°tK ><
•
L.-
a fz a
>=
u
w
..J
L..
w 0
::f
::> ~
x
L..
a
w ::f
>=
TO LOAD DURA TiON, lEfT
TiME or YIELD p
,00
,., ,.,
v'
',J
'v'
'..,..
0
"'..J' "',.,
'0
0
'"
"'';?t. \ I,.'o/,_!'S>L
zoo '\l\ _lL~~/-Y <,
E
><
z a
10 [)
1\
!\IX j-,\-,I '\=ll"..--x--'
70
w
j\
I
t:
u
~\C
\
I,
v
,0
i.,
...,, .... ..., N
'-'
,"
JO
0
~
w
;;::
20
1\
10
w
7
u
\
,X
\'1
1/
v \
\11
0
.
I
:>: :J
:>:
J
><
'"
:>:
2
·1
x
t'l
t\ f
V
1\
11\ 1\~/r.i II
\
I
.
I
f
0.5 .00 I
/
,..,
V'
0
..J'
N'
0 N 0
T,,:-rJ,C'CI''''''
f /
~ 'f\:j- .•./ '\ f\-..J I ~ I' ~
,.,., d'
V "f..../ \ i "-.-',.'vl+-lI-fHHA-\I
Z 0-
<,
"\ ;\J> ,.,., /
E
-'
'"
6'
co
a
l>:
\
~I
/ 1/
I
K,.
0J. J. /\t / (-~ \ Ii Y X- ~ -,(1\
X 7'\
\
~N(
3\/
I
of 'J-
rv-: \ YI'-.j \,\ 1\
'9,\
1\7 "X/f-..Y
ltv'
\j::J .I Y\
I
\
:;--1.
/\
~If
fl
~IK
X.
17
~ X\I Y IX "II'
j
'6'
1'\
L\1'1 \/I\,d / 1')\ M '\ "
\5
.
,\, I'-
II
'I
.f
f
X
f
.
v
V·
11\ 1\ ,y
,f
IV hi /';10. D< ~[)
~,L.'7-
.002 DO.! .005
A f ,f
c-.
'Ai
X
. f\
--..
.'~
f
oJ IX' "
' XiYl
~
..J: o <'6 _
<·s
rg; ~r>
~./
J);,;"I><' ~.x
1'\
~ X p
/.5
",/
, I.j
r---\.< [);~~.\ 1\
~1xZ
~~
p / 'r
~
.\
I. !
0.95 0.8 I
~f-1.:I
D,:;l.
/
V / .01
/ / .02
/'.-; .05
.05.07
/ /. '/ '/y' / .1
.2
.3
LOAD DURATION TO rUNDAMENTAL PERIOD.
Figure 3-74
1'-1)
)/ ~
.5
,7
2
J
TIT N
Maximum response of el~sto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (Cl - 0.010, C2 - 1.7)
5
7
~ <[
k
1.7
-,
Z
W
..J: s
U
,\ .
~
-c 0-
':0
2·0 '
7"7
/
p
"':11:"
1,'/ V V'){I'\ I~~ ~ 0
V
\
v
\I \
IX
f
\1
,X\ X" V ,'V l'vf\~~hl ~--'o lX V\ [>( I>: V f\ \I{ ~/\I/) ,I[\,
0.7
IA
'0
0
0
7\Y...:y "A ~z.~
'\
A;" /
".J. ~
I
.
X 'IV\. \I./!
\0 ~ IX ~.
r,
0 ..J' 0
0
w
X I/'
,A,~Z[=:X y.
5
0
11\ 1/\ \ / II ~ A X 1\ j V / X
,1\ 1/\ j\j I
':
~
d'
1\.
L
W
0
"-
-JIYs
7\ ,
0
0
"
\
j\
j\
0
0
-I
17\ k;i' \; 'I.
,0 r-.
/ \ /./ oJ"
1\ ' X , r ' \ 'I'
a L
\
~r-\'
\l , 1\...>/
0
p
0
YY.7\ 7 .
<: V\ I--
~
',., .,.,
0 V'
1\ 1/\ 1\ J '
SOCK
f\
0
,'T:',r \.
\
11;'1 W "'Y.\I/
0
~.'~
.'~
0
~-xC', Till'\; - " . "
><'"
0
0
'...>
0
10
0.70 0.61
0.55
=> .
o0Z
a t:
u w
...J u, W
co :>: :J
:>:
>< <[ :>: k
a w :::E
;::.:
e
• TIME or YIELD fa LOAD DURA TlON, p
'",
'"'v- '"a
v' u'
/
a
a
/
'v-
a
.-'.
,j)
'",
'vv-
a
a
;00
"c;-',--r J\ A - /, ~,,;~ \~\ ( X \j\ JOD ., ~ II \ I \ J /\ / f-\-_",II \ 200 .'~ . -A,(Xi~ ~A -'x'l-~
w
x
f
-c
100
Z 0
70
t: u
..., • r..., ...,
w --' w
.0
--' '" w
r
;::-
20
10 7
w 0
0
L
:}'
" "
.l
><
2
0.7
\
~
lZ\ .
/\
'I
-,,* -I
•
-l-v-»-
~/
-1:\ I.,L
II
/\
/ y
_
/I.
,;;/
I
A
j
/\'
iV"
v
I x
a
0.0 .001
v'
'0
a>
'0
0
II'-.
./-"7 lLI-lr' 17
.002 ODS .005
0
a'
U'
'0 V' 0
r"
f
J
1\
i\i/
jl--lk
0
0
'"
/ r'.
~.
0
'" '" '" U'
~.
II
~
V\'I)\
0
Z
-
J.
'\
/\
\
<,
t---u
8·2
....E
tl
7.1
o
6·2
w n,
3.,!
--J
N
II
'"
-., t-J\ I ')rl\
~/ \~V~ ~~ 1\'
1\1);
I ~ ~ 'f...'1 ~ <: IV -s: X y 'Nif~~'\ .~
ri
I'-,(.
~'6
1':---...
~ ~ ~I
:..:
' . -.-
-----.:
.
17'"
r-...;:
"
.
_~
r-;
v-
I)VI')( rJ( r"'V
_ .0 I
.03
~
~ .6
I--'
~~~? ~/(
///// .02
'X: () ') \ /(t--J
N) ~ ,,~ ,X'...: ~:.L 1\ . 'X \)'-.. V vl,(
RSit><:t><:;~I}
f\ 1')<
X
r-
,.
t-.~
I. I 0.93
_\ 1\
Y
'\
.
KS~ pjr u = ()~."Ft#d
I//'
.05 .07
.t
.2 . . l
.5.7
2.,!
1.3
\ I~ ~ - v
I\cx
2
.l
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (C1 = 0.681. C2 = 3.0)
5
7
\0
w
::E
::J L...
1.7 1.5
V
« .... z
..ts
2·0
....
'"
'"
a
<'6
-
~
Cl
~o
..t. o
, .~
V-..
f--
9·5
!',
LOAO DURA nON fa rUNDAMENf AL PERIOD, r If N
Figure 3-75
0
«. \\Il\.I I'" .Y-
I\V-.I\
'0/\ -A 171 rffl;>t---
0
~
T"I:j-!>..N
',,)(1' ----A
\1
<, I/1'\ -J;b?~K >
0 ~
.- -
::J.
iZ~
a
./IY
I)
0
'..,
l'l\
X /\
IXlL, 1\/ ) I=~f\ '\ '\ X II ~ - y, \.' V l\ f ..f.. 1./ ~ IrI t\I' r 7 rr 6 1\-'" r\ k-7' V\' rv \ 1\ I~ rX'IJ',< ):r-I'\ ./, 17 I X
a
'",
'\7' "'I-..J.
r;·f..,'
. IV·
.~IL
./\ x
a
/ \11 V·
g,j
,,\
k
a
1:'1 X-
.
, - I
.
'",
tclf
•.
1..1
'J ' J
a
v'
\
17
T 7\1 f
a
~
I
-X~L7~IVf\VI\ -- y~ ~>
V
J(
w --'
::J
:Y1~1:;xp..:17\
SO
f--
u
X)
kt: .l1S:.X K17. 00 It--\--lh,X· f\-!l\ . I-s..j: IF\V'\lf-I\lf
0 Z ·0
)III
a
.r
],17
Ix' ,JI~ ~I\ M~ _ 1/ _f/ IX / X
IX
<,
WI
..
a
•
z
o .... z
~ .... o
w ~ w Cl ::E ::J
::E
>< ~
0 -.8 I· 0.70' 0.6 ,
L...
0.53
;::
0
w ::E
, TIME or YIELD TO LOAD DURA nos, iciT .• ':"
. v"
r'
500
Q
v
12/
I
Or
.-',j' .-',-'
r'
r'
',j'
'
0
L\
IJ\
2 (J 0 W
><
i\-~
Iy
><
z 0
70
I,
U
w
50
.~
t.,
w 0
..... , ..... ..... ~
30
C\
-'
w -
20
l~
,. z 0
,.U w
. 7 5
-'
:l' ::>
::!:
>< <. ::i'
.l
2
.2
V\,
-Y?' I¥
¥
1/'\
'J
I\f
:7'"
~'
0.5 .001
.
1\
0
0
'r'
.'(1'
0
/1\\
II \
I'
x
1/ I
o
'
~
l/\. '''\: V
l I rN\X....., r"\
f\
'
)
II
II ,XI I
N.1'7
I'x'VX XI/ I\; V"vrvrx /rvl\
l<5
/ /
/
.0.1
/\t-0
. I r-£ IX\.-.
1\/ V' V 1/\
.01
0
0
0
0
I/\'~' IA\
/>
0
)
)1\. /"
I)(
0
U'
9· S
l---
J / J'-.
8:<
[)'-..
..:v 1\1 \') . .:::I-~ K'''K t-.
McI-"OI\-OO, .>-/
<,
t'-.,( 1'- \--.
."
6'
oS.
r-,
I
r-,
K. 1)1 r-,
~_
f-
'J'
or
F\:
-J.'ll \'... f\ .' ,,1/
~
V
f7'\:i< XNt':\ f--.)" k') ~ ~l/\ 1/ ~ "'\; f4. P< P( I;' ~ K:2lYKts
v.; "\
r-,
~
I'- ~
~ )?
.'l
D?
5:S~
.2
.. 1 . 5 . 7
s .f o <.J 1.7 1.5 1.3
t-
I.'
J
0'-.
='nO;;J"~~ •
2
.I
Maximum response of elasto-plastic. one'degree~of-freedom system for bilinear-triangular pulse lei· 0.464, (2 • 3.0)
"0 0
ir
W
0-
.,..-'
s
7
Z
w
0
r:/C;:' L>.....
.05.07.1
E
':0
<'6
rt««
:f
.1
\ . Y \
<,
'6'
J
~I 1:---' I"'--
z f-
f-
(7 ' "
LOAD DURA nON TO rUNDAMENTAL PERIOD, TIT N
Figure 3·76
0
'" '" '" '"
I
I
~
002.00.1 .005
..,.
0
..,.
lJ'
<.0
IIY
II
'0
0 dJ 0
1111"'\'- n
VII" ') _t\h . / V K1I1\
~\
I
11 - N hit-- VI rI..
I '"1 L\ 1/\7 --\,( IV IXl'> j;A; f'...
I Xi\'
'1'\
',..
XI/l\f'-
1)\ I ~ T
,\1
0
.-
v-
I /
{/
\
b'1- ' 1/
0
0
IX
/~
tx'
1/\
-
/1/177
_)(
1/
X
Y\
I
II~ ~~I'> 1~k:K2t>
1/
II /
U
1/\ \1/ \ II l/ f,( I~ IVI V IVI)
'
""
.
'r'
'XII'Jj 111\ f:}7Nl/\Jl-l'/1'\v I ~ I tv 1\7"--.. tv<.. '" IX' 11\1 \: I \ f.I r~f\I--...VI'V'l\ "/\ 1\ II \I r--Alo/·t'-/~b(\I/i'-
\11
.' " \
~
'IX
.~*~**~;j>:1, ,¢ IV V 1\/1\
..L
0.7
/'
\ 1
rv
0 v'
IX
r'\'L'\ 1/\ IX IV t \ \ ,\ 1
,j'
\
I
'. ~:IV
1:37 -\ 1l..L 1 'i . !J \ '1<-17 lA'-, I -XI-\; --,IX
10
\
IX I.
k~\~~~il-IX
1/\ If.;;
I. 1..1
0
I
I\
0
I
0
0 p
'IV 1/\ I / - , \
l/-\~\IAI~ -.JX.
1\
100
I
0
',j' ,j'
0
lJ.~~ , . 'II \; b( ~~\Y ,.!r\, IX _ k~fl-V
"<,
F
0
'J
"'0
'
·''\.,W\'xX V
JOO
0
0
10
0.93 0.8 I 0.70 0.61 0.53
<
z
...
::>
0
f-
z 0 t:
o
w -' L.w 0
:f
::>
:f
X .,.. :f L.-
0
w ::f t:
e
•
, v
,>
if
'",
'"
.r
v'
f1ME or YIELD ro LOAD DURA nON, ~
'.r
0
0
~
',~
0
0
0
.r .r
.~
'"0
0
..
0
~
.r
"
....
0
U'
0
•
0
,
tEif 0
0
-a-
'",
0
•
0 ~
o
0
0
o·
0
IJ'.U"~t-.l
IS'
o
0~'7'.,.-. fr" J..-.. ~/- ... -IT.\1.1/ 1-1 ) '\=' r\ /\ 1/ IX' rl\ \OQ' ~ :,CV\~7 y nVIAI / '~~~~fhP~N
o
0
<1'
'0
v'
OOt-JU\O
'"
;0
/I
'l_~t7
{ , 200 "
.Q .~
~
w
, ,.... w
'"
~CO
,~'
100
"'
U·
.::-;::
',,\,l\~)
J\.:>;:
2°;0-
(-".1./.\1101-/
'Sl.
..:\.-
-'
'"
'"
10 _
y
\
\
1/ .\ll A. A .'\; . E IY' r.; .. 0 ' . " 1/\ ' ' ' ' ..7, . .
,-/....
.
5 IXd'-" I' '\0 T\
.\
o
3.X J 2
o7
1:
~-~
- ...
J
K',,!\ A KI)<
.t/'StI1?<'
1\1
,
.
f\,
YI
:J
1
~
u.. '..LhJ..l,L.l::l
n .......
1
/
Qj/I'..4....I'YIIYY
1/
/
1--.'(" '/\
i'\l\i'\ ,Y,
X
1/
IN '\
,IV
. ~r\ V:l~v l/ .. X Ix" -./ / IX 1)1.
117'-. l/ .tl\- 11\
""X
.
'x1">}\i'4I_~
--.,<.,~ /
,)\1,(
J
,
J..
y
I ,f I. I"'""t'"...
• .01
/_!A
X
.. ;1. /' I
.02
V
0
.OJ
AI
AI
,
/ .
y~!X cltl\
~VIVV-~I). ,
V I
~
•• ,
.05 .07.1
LOAD'DURA'nON TO
Figure 3-77
fZ
;:~
r
~ .~ ~
...lJ.lhl...lL :::bllLlJ _
'"
,
.. :--
V
r--\-<
rUN~AM[NfAL PERIOD,
I'" J
.2
J 1
1.3 1.1 0.93
\
?) )?~u ~\
, •• , ,
.2.3.~.7
'"
I.S
"1
~:~6
"
~~j)~p{>~'(' txl:)rc2
0.;" X·: .001" c.002 OOJ -:005
o
.
YI
I'Z
L.
1
..
'
_ 'IVI
~
W 0..
w ::t:
\ /' , ,
o.
o
'"
~J:'
.AI/··"
1\..
' r. iY "';r
X
J
1\.
V
"""
.....E
.t.
.
X! 17 . . - -/if'I;'
Z I-
<,
."
IS'
__;:;..
.. \ I'
_·j\-X
"
'J
~',.-['!'Y.'II~~I\I~ JJ ~OC~ 00 A-I-\-- :·)E ;-1-1-\1/· ... ·i)\-r-·~ . A 1 !\ ;, ~''X.:V\- :1\ .;' / /y j( I } JOJ\~~IV\:'I\ i/ l\ 17\ ~ 1 . . \ I , m .
',:7\ -f
3
:"
I~~~~i~~,:.:)~~~
'j.
~.
r
I /
70
o " z o
,~.
~
I
T/T N
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (Cl = 0.316. C2 = 3.0)
J
.~
0.61
0.53
I J 7
'10
o
t:
u
w· ~
L. W
a
::t: :J.
::t:
~
::t:
L.
o
w
::t: i=
TIME ,....,...~rl
().~
C>
'..j.' ' •
~."
.U"
t.p·o
'''"''
or YIELD TO lOAD DURA TlON,
o' J
o oJ)
o
o
.0
0
!J'
o
0
'1"'.'v"
<1'
<1'
J
0
v"
o
,.
50~'Vc;r1.7\11<:1-'-r;·~~~
.~
o
""6-
0
0
o
.~
"
_J\. I XI AI /
., '''''' X I XI / \1 / •.
I/\I~.
0
.~
tilT 00 01 G'
ol..f\
'00'0'00 ~ VI t-J t--l
.l,J'
OOt-JlJ'O
---
VI:! )f..\. /r~J /
.'"
.I VI xI-¥H{~
v
IV
-'
1/\
'/
'"
'Z
o
-'f"J'~'h-¥--p,'\-OOl\,! 1, JAIl!
>--
u
<\'(-=-~\A~J"
W ...J
.... .., '"
,-u.r
.s.
.
o
l>: W
c, ..J
'"
.-t
fZ
'0'
W
,. '"
a
,:':
:z:
·n ...J ,.,
::>
L
'"
>--
~
o
>--
'z
o
,u
-
W ...J L W
/
rv
~ -r,
a
v
::f
::> ::f
:Ii
>< .,2
K \J3; ~p
0.5 .00 I
1---..t'--j /1 /
7'
.002.00.5 .005
1/
LlL /1/ l /
.01
.02
/1/
.03
~
.05.07.1
.2
.3.
~ / r. u = O£1"!H Jl ~ ~
y
.5',.7
2
'>.5
lOAD DURATION TO rUNDAMENTAl PERIOD. T/T N
Figure 3-78
Z
o .;= u
,.
,-lb'1f'<--1f"'f+-lI++"~
::>
-~-y'=tt
of-
W ..J L W
v
o
:f
...E o
6' .
..,,
Z f-
<,
'Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (C1 = 1,215~ C2 = 3.0)
,.
s,. ,.
w
;:=
e
• o
0
0
0
0
0
'J' 'po 'v" "v" oO\J'lf';...J,O
".p
"-I.
o
0
.~
...
0
'~
0
> d"
0
.",
v"
o ''0
'0
._._...UJA I\
/\1
~rl\~
20a
'0
'"o '"'"
o
'0-'0
~ 00
'"
V'
...
'0
0
...
N
'"
0
I-L \V ~ A\'1..I fl lt4Y 1\
I I),) X I \ \/1\1 " , 1/1\1 I /\ 11\1 I'll '--\-,<,--~tl='I¥l='"~., fj I
500
•
tdT
TIME cr YIELD TO LOAD DURA TlON,
1m
IX
d'
."
<,
...;
a
'i
o;.J
-. S . v'
f~ \.)
""
Z
-. V
'6'
w ::F
..<:
a
n
':0
0
J
s
L-
.f
o
f..-
-'
'.i
·oJ
--' f..-
,-,.,'"-'"'
'-' W
0;
L.. O.
z
,
0
I"'·
100
0
W
~
'/
>(
F
<,
E
--'
'"
z
1-
,..
I~'\ji/
0
r-.
'L
'Z
0
;:: U
...c --'
L.. cJ
a
<'6'
<.J <.0
.:it
7
'"""
1 1.1
S
1·5 I.J 1. ,
~
::> ~
x
!kf\hf-N''-¥I-oAl ~!f?~~WE~!W~I\
~
-fr''v4~-.f.
1/
o5 .001
.002.00.> .005
.0'
'7t7 .02
.03
/1/
~iru='~
~
.05.07.1
.2.3.5.7
LOAD DURA TlON TO rUNDAMENTAL PERIOD,
Figure 3-79
2
3
TIT N
Maximum response of ·elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (Cl = 0.147. C2 = 3.0)
5
7
"10
z
" a
z
0
;::
u w --' Lw
.,.::>a
:t: ;(
0.93
<:
O.B 1
...0
0.70 0.6: 0.53
~
,.
w
;:::
rlM( p',-J
...... .....,.
~ ~ 'oJ' '0
~
'J'
o
~
.~
0
....\.
~
0
or YI(LD 0 '
TO LOAD DURA TiON, '(IT
0 .
o
0
.
~":t~
'v
o
0
0
~.'~
..
0
o
~
o
0 .~
-;,
o
o
Cl'
o
'"
500
'0
'0.
'"
o
0
..
00
0
....
V'
....
N
'"
0
z
>-
<,
E
~
~
><
<,
E 100 >< 'z
o
>-
u
W _J
1\
70
~ ,A,\
1
r
50
o
1
.,..,
6'
)1
p,
/\
";:-1,J.1(
..J
<:: >-
z
II
w ::f -0:
,~
W
w
,
o
o
()
:::J
Z
.~
.....
w >-
00
o
w
'"w Q.
I/~
A
o
L-
a
>-
z
e·
a
Z
t:
o
;::: u W
u w
7
..J LW
o
__ \ II \I 1\ 1/\
-'
....,
A 11\ I Y
L-
W
o
5
::f :::J
~
:::J ~
::f
3
~
2
L-
><
..: :>:
::f
a
w ::f
;::: 0.7
"II/IV~
1.1
0.5
.001
.002.003 .005
.01
.02
.03
1/ .05.. 07.1
~ .2.3.5.7
, I
,
I
2
J
LOAD DURATiON TO rUN·DAM(NTAL PERIOD, TIT N
Figure 3-80
Maximum response of elasto-plastic, one-degree-of-freedom system for bilinear-triangular pulse (C1 = 0.100, C2 = 3.0)
4
,
·5
,
,
7
, , I
'0
•
•
• tdT
T1t"EOr YIELOTO LOAD DURATION, 'r'v"yJ~
"t-l
500
'1.1'
'v'
'J\
"
/
/
't->
'<..1'
'0
I~
r\ II
\
0.0000 '09 .~ -If' '.. .~ 00 lJ' .. 1.1'
,~()~ ! -, lli 500' ~ \1 .\ "/\ W 1\ 1\ 200
IV..'l :' \
w
x
<,
E x 100 'z
a
c: u w
70
L~
, .... w W
'C
Cl
--' w >-
z
t:
u
w
--Vt\- 1\"'1 kf~ I I '.1\
50
I
[I\: f,r
IA 10
\
::E
5 .!
x
4:
::E
.r.
l/
IXII
2
.......
I \
..
V::S Y J
v
III
17
\III
fx' \
V
\1
\.
x
X I I\,~I
V 11\ 1\ II. I.
\/IY IY rv
1\
x·
'v
)
II 1\ .
II
m "\
V V\ .'X -. I ~ I{) J/ ll~J r--, ~ IAlJ l/lyt-· IA/j.,V P\- hl/l\ 1\ ~J'I i1\
-r
~7' D"- ',(lJ .
"
IV
\ I
.
/
J. \
.... 0....
V'
o
.....
'"
0
U l/· 1'--..1\ t\-)i l\711
II
'
K.
',., d'
li"-\i ~ 1"1\' ,. ~ll ~\\ v
-'
."
.'"
6'
X X Y41iVIYIJ f\
'\
\fo~) .Jb(I'>
IX IX V./ J<1Il\ txtx~Rb
' nl~
I'"
/ '"
V I;\~i-.)(
rv/i~1\(
~I)
JI
Ii· \\ (\.
1\01 \J l/ M'M'
\l . 1/
D
ll(
1'\
V .
.'-. Y ij; "\
k)<~~~~~rs
/ .0.
V
/ .02
.03
/1/.1/ .05 ,07.1
'/
~/I"/./V
,7
cc W
c,
--' 4:
w
9: 0
a z
<'6
::£ 4:
::> u,
a fz o
i= u W
<,
1.7
w
"!-.. V
/.5 I,J 1.1 0,9.! 0,81 0,70 0.61
t--.\.
.\.:j;::.l'.
.
K-s~ p/r u = ~:N4.I 2
a
o
<·s <·0
•
.2.3.5
E
Z
.J: o
j
-
'6'
.J: s
i'..". \l
Z f-.
<,
f-
I
1)(
Ix'\ 1/.
~
.
1\
f(l/. ~ ~.~ .\
O
LOAD DURA TlON TO rUNDAMENTAL PERIOD,
Figure 3-81
'J'
'\\ 1\
' / 1/
/
.002.00.! .005
s.
X"" ~I/
-/1,\
Y 1X--lX. "".
\
.....
V
F
I i'J"-Il('
IX b( 1)<1/1\ 1~k:1X~t><81/
00 I
IV
''\:1-'1'
I
IL\
o
'0
~
o
1"Y--
1\
r/\
'0
<1'
"""
1\ 1/ J~ ~
[J
Y
'0
'0 Cf'
""l
v« I\: 1\
V
IX
~;;
0.5
V'I
'\
IIX
"em '\ 1\ 7\ 11 /\ X ~<,r\ r, r\ 1)\
/
dJ 0
1\
..
0.7
'0
.0
/ ~I
J \ 1/
II
'''!...
-!J\ 1/\ 1/
1\
l/
.
CS'
I\ 'v
V
V
II \ \;A
l6.1)\ 1/
1~1"
7
::E :::J
V\ II
o >
0
.".
o.
IV
[\ I"
V\
v
0
',.J
po
1\1.
20 ~d \1/\1/\1\
--'
0
0
',.J
I\ I \ Y. \ 1"/ 11\ I I I J 1\ 1\ I
I~
0
a
o
I
u,
w
'-11 ~
h\
.
\1/ \1/\
1\
1\ 1\
,-0 0
X -~~1-
\ I
'v"
50
--'
W '0
,I
IX
o
3
TIT N
Maximum response of elasto-plastic, one-degree-oT-freeaom system for bilinear-triangular pulse (C1 = 0.056, C2 = 3.0)
5
7
\0
0.5J
..J
u,
a
::£ ::> ::E
x
4:
::£ u,
o
w
::£
s:
TIME or YIELD TO 'LOAD DURA TlON,t£!T
.-,.,
v' v'
,"<1'
,.,
0
0 '-' 0
-- --',., '" '<1'
'0
0
500
0
0
".-
0
0
.~ <1'
'S
0
'r'
v' 0
0
'r' 0
..,
0
0
-:;.
e-
<1'
<1'
0
'0
0
0 <1' 0
0
0
''0
.-0
0
'0
....
v'
....
'"0
-
,, ilo>,ll4;7'''l=r~Rt='t=¥'
Z f-
"t+f--'<;++-'\f--l,"",~..fo.A-.l.Q
:" -:
100
u
w
70 50
~
l __
w Cl w
Cl
s0
~
".
'"
o
,~
>.
20
Q 1- .•
1.
o
10
1-
u
~:,j
u
c: ::>:
=>
"
>< ..;
s
\/1
,rt 'tl,-r.t'D: .~6~ 1\l\!I-';;~f '-: ,.
'[\-\
§ . ,-
E
'11:-\-t'lf:V1
<,
A\ ie-I I H
V"J:Z:sl-X-
;,fj-f'.J.. .
u'
\
'-r::::iAtt'l-bt--t---\cH--'\A-\f-hPifiR
,.
~ W-X:
."
~
,n
-
'-!j
l-m~I,3:-~.
: '~1'S7\'iX --"','4-:::;" h '~':~-, -x v Y
::> u.
~'.J.\/ \OIfe--Y I\t
I \
Y
I
_
H/~...l.Jal...ffi..LlL U.
rJ'-o.".l
~ _ IV 1"OlL'l I W
/
'
,'\.y
V '1;---1\
Vif\rx--.x
D.S 'r .00 I
'Y/iT 1/ Y
--I,L.
r#r/..L -V--y
.002.003 .005
717
•
JI'
.01
.02
.J3
..ll.LU.Jl.::bLJLl..J
. • , , •IXV7)?;~ i '" I'
j
.05.07:'
I
,
,
• , j
1'1
.:1 ..:j.
.
j"
4
2
5
TIT N
Maximum response of elasto-plastic, one-degree-of-rreedom system for bilinear-triangular pulse (CI = 0.032, C2 = 3.0)
,
7
, ,
I
10
..J
u, W
Cl
::f
::>
II
'"~
053
,
;::: Ll
w
13
~:~~
~
3
;:~
g:7
\UJ
-\.
.2.3.5.7
·<·0
t<-f+-
~~ p/r u ~ ~
--.L..L '.LLL
<.J
"rr- :/
y
17!71:;.r-1~
LOAD DU.RA nON TO rUNDAMENTAL PERIOD,
Figure 3-82
r
Z
o
I \ ""
~
t-,
f-YI/0/>«.\
--.
/ 1/ vv -+rf7!I'J...., , ,
/
w;
. [r/"t
~~~~K, M><~2~Q~ 0<~?2)~~~
I /
\
~!lx)~~\, ~ ~ jS
\
l::tX.~ ~~ ,iJX b('/\ )
I
of-
~
.
1\17"'>'-.
I'f N';
<::
<::
1\/
\
...J
o z
F\ I n-n+-i-.I....:.. \ I I
.I \
Q.
::f
rvl
\ ~-~4t¥~-1 '/,"YIXV4Io/,;11 . K*X\ {\ G'<~~ .IiV\. ~ f~ h(ll 11\.IX-'1f
J
W
"J
-1--"
-'
'" z
f'siI~~t~\V\I)I:it: . .fN-i~~--Pt'-fl-...j,C ~I/ \/fz~~
L6.".L
o
f-
.,.ijL ~
-I'
07·f
'"."
-
;'Yrtsu~~~
'.~.
Cl
\1>
.
/
~
1\
ii\=" . -4+¥-+\i-A i\..J\f;!~~~ It - IX ~ I'\J I v ,-,,,'''1.~7,(1)11\ III ~ r1\-lYr \'\1-' '/ j)' .~(f]ll:SJZ: f 'V YI A ~/ .::U~
<,
::f
l..
o
w ::f t:
e
e
..
00000000 '\$1 .~ """ 'po 'v' ....,.. OOi.J't.1'~o
.~
..
r
e,
I 1\ f l ( \
X <,
I IIX \ II
.~
o
oz
u
W
'0
.'0
.~
'0
1\ J
1'\hT7\ ....
11\
II
f
IY I 1\ I 1\1
\
1\1
\If
~,
o
\17
o
.. 0
o
o
'0 V'
·0
'"
t.l
0 .... 0
""",'
r'
...
./
."
;
70
-t,r
50 V"",--j"l
w ,
......,. ....
30
~~
20
>-
II
o
,"L
o
t:
J..
~
~IV
'\/1\71\ 1\
c:w
~IM;I, W'\-l U" \'L IV'
II X 'V \V\U},I
o
::i :>
-...,
of
::i
.....
~
Z 7 1_. . - . _-- ./'_L. l~dzj;fj I YI I/l V /17l7 o.5 .!.' --t-""t-"'"t...,....,...-,. 07-·,
. . 001
_.
-
--
-
.002.003 .005
.0'
.~
.-----.---.---,.
:02 ·03
.....
~ I
.05.07.1
- . ~. I . 1""' •
.23.5.7
LOAD DURA nON TO rUNOAMENTAL PERIOD, r
Fi gure 3-83
"
i'
•
2
3
ITN
Maximum response of elasto-plastic. one-dellr~e-of-freedGm system for bilinear-triangular pulse (Cl = 0.018. C2 = 3.0)
I
I
7
«
0 f-
Z 0
;:::
u w
1.1
o·
~
5
::f
<..J <·0
.7177 i
S
1.3 I. I 0.93 0.81 0.70 0.61 053
'"
«
a z => "-
1.5
::i
--J
~o
<'6
U
.w
W
W
,J~ 0
10 tr"4~='1Y·"""·
'" 0-
·0'
of
v'~_~
J
0
>·z
~
_I
o c.
0
6'
_ .
L_
W
>.,-
E
-'
I-f'H-Ij>,f-
.. 1
z
11\1'1 IT
x'=-~~~~g~~i~M:~~~~ ~
E . ~ ~ . .... .>( 100
,-
... ~J .....
o
0
~
o
"
tclT
TIME or YIELD TO LOAD DURA nON, .. ·.r~rl/" 'r-l 'v' 'U'" 0 \1' '-rl
e
10
-J c,
w
::f
=> ::f
x-c ::f
"0
w :2 rr
TIME or YIELD TO LOAD DURA TlON, 1i"v"~~""'''' ',.,) ',» 'V"o '\1' ',-J
k#~~\1Jf
000,·0,0000 'J) 'J '.,j' '.. 'v" 'v" .~ .~
o
tdT 0"
0
'0
'0 IS'
o 01' o
'0 v'
'0
0 ....
~~~UuslJill~~'00
iJ't1'J
0
...
0
.",
Cf'
'""
v"
. >0
d'
o..I'
~
0
....
01'
0 ....
0
'A I !\IF==l'FI'IIr.=;f\:#
,., ...
d'
200 w
><
1/\
\ /1 \
'/
.x V\I /\1 / 1\
<,
E
sc
1\1~'f II'
0
S
y. I /\17
'
Q:
w
-' ..:
I-
u w
Z
-+--'1~r--I+f-t<-l"
-' LW
':6'
':0
Q
-'
..t:
...;
,. ....
tf
0
z 0
j::
u
w
I"
'\/1\/1\ AI
-' r,
~"
W Q
::>
sc
.J. O
I-
<'6
0
<.,J ~·O
1\
::E
~~P/ru
~
,002.003 .005
Figure 3-84
/1/17/1/
..~LOAD
,01
.02 .03
=
z
::>
L-
-Ndf~i<:
..
::E
::E
Cl
S
v
::E
w
'..:
a
~
0 Q.
f=
...•
)
Cl
.,.,
6'
1\
z
....
z
I-
O;f'tlH~
1.7 1.5 1.3 I. 1 0.93 0.81 0.70 0.61
0.53
0
z f= u w· -' L-
w
Cl
::E
::> ::E
s :f
L-
0
w
:::Ii
;::
V/Yff// .05 .07.1
.2
.3·
DURA TlON TO rUNDAMENTAL PERIOD,
_I
.5.7
2
j
5
7
10
TIT N
Maximum response of elasto-plastic, one-degree-of-freedom system for bilinear-triangular pulse (C1 = 0.010, C2 = 3.0)
_I
•
•
•
TIME OF YIELD TO LOAD DURA TlON, ~
',..,
v' 'v'
,.., '",
,..,
~.
o
~
',..,
'",
'0
''''0
o
....),
o
0
'oJ'
.J<
0
•
0
~
•
~
tclT
C? vi. o
f? '?
C?
~o
.~
C?
.... O"IJ'I
~~;
·0
0
-CJDCfl 00Ul
o
0
0
U'l
..
lJoI,
0
N
0
N
OONUla
._~~-
.~._~. _1-4._L..U _ _
l-111f=t\t=l=~~=14""l~
w
E x 100
z
o
PI
,u W
-.J u, W
v
o
...... w
Z
f-
<,
...E
tvmi
<,
,
~
A\
VI
X
w
1\
Cl
o
...
Ct:
II
II \
ll. ...J
". -
1/1
«
f-
z
j\d,j
W ~
D'-
«
Cl
W
~ L..
>-
o
C1 -.J
f-
o ,-
~ K1<\~~
z
o
f=
Z
o
~
s: u
W ...J u, W
u
W
-.J
"W
C!>
o
::::>
.
::f ::::> :::f
~~
~
~
~
x
~
~cl~
:::f
2
L..
o
w
:::f
t= 0.7
j,Lc¢=:L1::;:+-
0.5'
.001
71~
.002.003 .005
.01
~ .02
.03
.05 :07.1
---.2.3.5.7
,
"
2
3
LOAD DURATION TO FUNDAMENTAL PERIOD, T/T N
Fi gure 3-85
Maximum response of elasto-plastfc, one-degree-of-freedom system for bilinear-triangular pulse (C1 = 0.750, C2 = 5.5)
5
7
10
TIME or YIELD C)
~
'P"
'r>
v" 'v'
rJ
to'
'0.1'
'0
/
o
/
'oJ'
o
u>
.~
TO LOAD DURA TlON, lEfT
o
0
.~
.~
0
0
.~
.~
o
0 .~
't--! Ii"'
0
0
.~
0
0
°oocoooo -UlO'lU'l""lJolNN o OU'l OONV'lO
......... C1'
ul
l'lIl'X
t
~IIIH
Z
f-
1--- .1/
w
L0 0
i.--P!~f--'1'"'\i'.t
X
<,
E x
17 100
z
o
,.Jl-:
>-'
W
o
.....,...,. o
-/I
'"
f'\71~L
w
L,
>-
20.~
f=
u
~I/
Q:
W
«. ~
1/
z:
..J
o r-, z o
a
o
,.
'6'
v
I""--'
o W
E
~
II
'J..
u
..,
/1 Y. I Y.
<,
~
'0
1/
.J. S
/ r---.1\7'
o :>:
5
:>:
3
:oJ
!~~i~I~~~;~
x q ':>:
2
07
0,5
tg--7!7v
.00 I
,002.003 ,005
~A/ .0.
.02
.Q3
2
3
TO rUNDAMENTAL PERIOD, T/T N Maximum response of e l es to-p l as t i c , one-degree-of-freedom system for h;l;n~,~-t~;.nQular pulse (CI = 0.562. C2 = 5.5)
LOAD DURATION
Fi gure 3-86
.2.3.5.7'1
4:
Q
Z
:oJ L..
Z
t:
u
W
..J
L..
W
Q
:>:
::> :>:
~
0.93 0.81 0.70 1.1 0.61
a w
~~=N~ 0.53
s:
f7F .05 .07.1
:>:
o
1.7 1.5 1.3
W
w
~'6
7
..J
4:
f-
Z
of-
<·0
W
..J
.J. O :?,]
i .,
Q.
5
7
10
:>:
L..
:>:
••
•
• tdT
TIME or YIELD TO LOAD DURA TlDN, r
',.,
,..,
v' v'
,..,
'01'
50
30 20 <,
E 100 x 0
70
f-
U
L" -i
LW Q
, .... ..,. W
'"
50
SO
.-1
:-0
20
f-
Z
0
w
7
W
5
-i u,
0 ~
::J ~
X
lL LX.
['S', 'i
3
2
0.7
~
0
'01' '01'
0
. '" 01'
IS6, ~ IIV'r V--' C;Z,,,'-': b'~ ~
-,,",
\ .. , ,
, 1/ "
.i2.L
~""t\ 11>." . 0
, -"
.00 ,
~
K.
0
'0
';
'/
'" '"
..
'0
0
o
'"0 '"'" '"
0
0
0
1'\7
."-.I~.f- --.l ,--'.i-' LA.!l \0,
--"-'" DZ.
L"'''
• 0."
~
.002 003 .005
'r .L L .z,
.01
L L,JL..
.02
I'\,
e"" ""
1~.I":
.03
.05.07.1
<:;;.
L
'A
Uk'
Ai ~~ ~ ~ _V
~
lL
~
~
~"
.2.3
,5.7
"
"0
0
J
<·0
w :f
5
7
L-
0 f-
Z
0 f-
......i
w
a
IS
::J
~ ~
><
0.93 0.81 0.70 0.61
L-
, 0.53
;::
.
,
z·
::J
1.7 '/.5 '
I. I
.
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (Cl ; 0.422. C2 ; 5.5)
s
<'. <.$
".".
TIT",
Z
.J. O
~
3
...i
,
2
0 cc
~12< ~ ~~
...E
.,
'0'
j l\ .
?
.',.>
'J
I'
d
0'
J.
"
e:::: e::::
LOAD DURATION TO rUNDAMENTAL PERIOD,
Fi gure 3-87
.». !
z f-
<,
W Q.
~ ~.' '\ ,_
l():/\J~~ PS j?r"
CO><
8;
a
' c "'" .
'/(1..\
9.5
,
"-
0.
!.l
v
A. "-
'> """'" ~,~~,
~-".' ., 'I'>' l2':lX:x --l
N 0
'"
_
,
~.
o
N
Li
~ , L v:::~,d.\-:~~I\: ,. \0,ri
~
o
'" N
I
, """
L!. -
o
.bV' \ rf\. L.J . U ~. ' t.L > t\., • ,_ v
c6:1\ .1-11, 1
~hUi..l&Jl) 1£\~jI-
,
.'J(
' -P". .l0''£'"
0.
"
, ~.....\L
't'1Xl1\ 1\,,, i Lx. ::-K V ..2 L ~ >D" 1>,,< L "'I~)
.;=;
/
0
.
f.-'
"
I-'> r
f.;i '0.A
o·
0
'Ir
--'J..LY
LL' . l.:'.
.LtQ!'ett,- •
'.
..
0
".. d", ,
' of• lfC'c\ '
.... 0....
0
v'
!
"~ I~:.'.ES:LSlI\ ut:L1\
."
o.
~
~ I-~ .' ~u ~ . ". "i. ~~., .LL.lAtAL'-, ~L1 \ A , , , D ~ 12, ,/I j A ./.. ~~ L
LX:
.L
0.5
.
.p 0
.
(.:i,0- .~ c: . " ..J.. \ '\ "-.. '.. _. L' L'1 . II 'j'l 11 . 1'1;::(';; t!\ IlCll'; ""
-'
O'
0
0
',..,
, , ..tY..J D',~ .I- L~'v ' " to. {':;
i",' " !II "Y '" I's"
10
f-
L
'01'
'"
0
OJ
0.
.5Y'f
w X
z
E '
/
/
0
10
:f 0 w ~
t(lT
TiME or YIELD TO LOAD DURATiON,
..',.., "'"
~
v' v'
,.., ,.., '",
,~
.-'", .',.., .
0
t1i lL.lI:' '.A oil
l.)(
~
0
....
V)leO
v '"
f...:.-",-
p
'" '"
''S
0'
.
0, 'v' 0
0
0
0
.-
',.., ',..,0
ll, ="J
0
".-
co
~ YI Ll l'
IA
,L
"'.i: ..>1-'
~ ~ -\~"~
,L
~
'0
~
v'
'"
ix
L>" ,
D
6
6
U'
'" '" 0
U'
0
"~i". & v 0D 1'..1
tyl"I(.I'
1)('/ ~~L",)«."P'
0
Jl.J e...:::t Q
6
'0 N
6
N 0
'".... uJ III
,
U'
9· S
J'
4..,
~~-'.f..j
"....
...E
, 6'
o
.
./
I \,
I...:::¥.. u.
~I'"'
iz,:
,~ ~
~
¥
~
~~.
Ll);
1\
~.
~ ~~
>C
,J
·s .J. o
~
.L.j:::... I£.,
.02 '.03-
.05.07
»:
, ..,
I
""
-t
~~ .1·
.2
.3
.5
.7,
2
3
TIT N
Maximum response ofelasto-plastic. one-degrce-of-freedom system for bilinear-triangular pulse (C1 = 0.316. C2 = 5.5)
5
4:
,
<.J <·0
f'....
L
~
<'Ii
11 .I" ~ 521::' 'I'"'
....
W
'6'
I. ':0
X
z, s: .<. • ' KD >
4:
Z
'I'
ld: ,'\-;>
,
..J J-
,L",
c2\ L ~H y,~
J..L~
.01
V
C> " x
'
1 I.(~~~ " Ie> . .L.. JL L
u
.A
Q.
'J-
4
'1)(
W '"
, . , .-s
U
.L
o
",0
I'
-I Gt'
Z 0-
<,
,_
" ".1
LOAD DURA TiON TO rUNDAMENTAL PERIOD,
Fi gure 3-88
~
""ro VII J' v"['~ ~'A..x,L 1\" V:~l\UJ(\'~A~.lV" ~-x- -X.:;.. /' 1:'\,p.II\.,~';J w'1.l ~I,l. I( P',J,,~" "-',~~ ~ ~ ~. 'q'"~...~
0.7 I;C.L -r-0.5 .002.003 ,005· .00 I
6
h6 I ....l.l l:\ '-,L '" .,! u V "t.l
c,f ,\6' -"
"'-c> -' -" ' -'v v '" " '. II'L b L . I) "1l o
'0
0
'",
P'\..
~1L
P&~ /.-~ ... ~, c
'
"0
.,
13,10 c/--:t. ,'Ap .
L
0
L'J
•
c± .zxrz '''-''''-'. .is;
0
'"
IT, '" ..,
I~:
.
.L~;'" ll~" , I'I'". " 1-'1L' ~/'"~'r, o' . -\.' I'.. 11::,,, A iI'I .E)I ,,"
.L1 .. J c,
'iii "i.>?L I!C til..A ..~ a c>'" ,I.> f1t'rC::Ni~·,/;tt,-\X1 '"i"\Z,. .
0
7
10
o z ::> "-
o
0Z
o
t: u W
..J
1.7
"W
/.5 1.3
::>
1.1 0.93 0.81 0'.70 0.61 0.53'
0, ~ ~
X 4: 20
"0
~ t:
e
e
•
TIME OF YIELD TO LOAD DURATION. I[/T po
.~
v'-
"..r
~
'I.fo
t-1
~
'oJ'
'0
50
o
~
o..... o
'J)
't-J
o
000 .~
"-:,
.~
000000 'v' .~ ~ .~ '", ·.... 0 .0
~
0
u·
'00 (III CJ'
0
v'
'0'0000 U' .. ....... to-J !'oJ
0
U'
0
N
0
U'
IV ,
w
1/
<,
Jv..IYIVI \11
E sc
....
I/~"""
'/
,v
J Y ITJlI
...
--,
-
u W
...J.
L. W
o
IV 1\ l\-l'\1A
0' ...J W
~~
>-
o
.-z o
;= u
I~V
11\
~
I~l<
0
0
~"" '6'
'" ...z'" w
<:0
0
J'.s
:::>
~"{V
~.[I;i
.to ~. Ii
~'.J
2.0 <,
,eJ ...J
1.7 t5 tj
'0
'"
....E
e-
yo
/rI~
W
:r :::> :r x :r
...z
<,
W
Q.
z o .,=
..., .... .......
~
\
I'P
>(
...
~
II
5'
-k=1~~nv't'~t
!j
2
~ 07
12
0.5 .001
./ VII .002 003 .005
I .01
IV'" V .02
.05
~k'
.05.07.1
.2.3.5.7
I
2
3
LOAD DURA TlON TO FUNDAMENTAL PERIOD, TIT N
Figure 3-89
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (Cl = 0.237. C2 = 5.5)
5
7
10
..J
:r
'"
z
....
...0z 0
t:
u
w
......J w a
:E :::> :E
tl
;;:
0.93 0.81 0.70 0.61
....
0.55
;=
'"
:E
0
w
:E
TIME or Y-lELO TO LOAD DURA nON. trJf
.-
',.>
v' v'
,., .J'
,.> 0
,
,
'.J'
',.>
0
0,
'"
0
0
'J
'{,.
0
0
0
0
.-.J'
0
0
',.>
v'
'S
. /1\ /1\Vl:If
,
c-
v
II
.
,1T~~1l1\ ->
j , " •. \ X 1\1:1" r _ "" ' "'-, Jl ." \ "" I -.,I hi
,
d'''~T A '
W'",'57'
z
o
,u
.
w
~
0>
a c,
-
A
m
I X'
,o..
'z
o
,7:''.1"" \ !
\0
1/
U
w
¥
z;
.I
W L, W
IY """'
d-
a :>
,oU"\IX
::f
>< -a:
.::f
I
/ .> 7I7J7L' -
0.' ,00 I
,002.005 .005
,01
7'F\
If
.02
7
'0
~
IJ'
'V"1 7'
rr
~
"'V'v
~
.,..." " ' ' ' ," , , '
" .[ "1 I;"r ' 'I". J'
I
k
r-,
, / ' j 's
"
•
.'
" .II \TIT
i<
717 >,'
:52;>::0"'''
,,"', "
~
.
,2.J.5.7
I
"
,
~ l
W
s
5
~ \1.' D? ~T~
I,
C
II
>
n.s ,
o.s ,
,
,,,
T '/'
.'
,OJ" '"
UT
u
2
W
~ ~
'05
L
3
Maximum response of elasto-plastic, one-degree-of-freedom for bilinear-triangular pulse ICI = 0.178, Cz = 5.5}
~ystem
.~i 00 -
L " "~,, ~ ,,~ " ho ,
'. ' ''7'
""",,
"
',e ' .. s k "
"r "
Iz X lI"'j( "
if"""''''
'l
0
LOAD DURATION ro rUNOAMENfAL PERIOD. fjf N
Figure 3-90
•
"
".M""~
.t
~ 0'
K"'>
-r; ,,(
k If: XJ1 , _ ""'" J
.05.07
I
""/~",,,,
I
L
0
"I
V
(;2
U'
~::;:: f.I-v"
I
.
'ilG ,l:" _
~ ~
'0
'0
'" '" '"
0
~ "
' , LI/L!:ru6: . . :::'dL..? 7J7
,05
'0
0
.
"
I If.;! ';< x:lfi ;:x i>
:oJ
~
0 U'
U'
-' I
I
~"" ~ ~. !'il[,;" \
'"
t-~ "0I/~ 'jf:fB,L: .w.! I
W
0.)
t:
'
.
,Lt!\ J-or'"I II ,
w -' c.,
...'",
. -'i
I' l'
'0
O'
0
11(1"I x 1\ T\7Fi'i . +=: T\/ IT I XI!\. V\I l\TlA\ " \,
'0
0
v'
'"'
0
0
0
0
',.>
.-
O.
T
s
7
"" \0
>
W
>
,
e
•
TlME·or YIELD TO LOAD DURATlON, p
',.,
v' v'
,., ,., '",
'0
.- .-
'" ',.,
0
.,.,
0
.
0
0
0
0
.~
0
p
'"
0 0
tdT
0 0 0 0 ',., .,.,0 '.., .V' '", II 'V 1\ /t\~ 0
v'
'S
•
.
'0
;
'"
oJ'
0
..
'0 oJ' 0
0
<1'
~
'0
0
1\
,., ,., ,.,
'0
'0
V'
'0 0
oJ'
11\ JI"I-{j
...z
<,
. I
AI
...E
=t\lz+st#;~(4¥&-Vil=tA
c"
.,.,
IS'
o
cc W
..
0..
-
_.'~
...J
...z..:
w ::IE
..:
c z
::> u.
., '~';:::I;t'j'l\t1~F~~~V1~)\
...oz
1
~
=:::c::::x
'
o
>=
u ..L WI
rw \
'b
1\
w ~ w
">-
C
>i~~JiR v l
'~/;-,--
::lE ::>
::IE
)(
-c
Vi. '-.ia-,.n ~
::lE u.
o
w
::lE
t:
0.7
~
os.1TJ,L.::LL14 .001
.002003 .005
.01
.02
.03
~ .05.07
.1.2
.3
LOAD DURATlON TO rUNDAMENTAL PERIOD,
Fi gure 3-91
.5.7
I
2
3
TIT N
Maximum response of elasto-plastic, one-degree-of-freedom system r or bilinear-triangular pulse (C1 = 0.133, Cz = 5.5)
5
7
.\0
TiME or YIELD TO LOAD DURATiON, lEfT o
o·
0
",p
0
'J'
.....1
0 ..
'\1"
0 'v'
o
o
0
',.>
OO\.1',J'~O
~
',.>
0
.~
0
'0'0'000
'"",
'"
If'~Vlt'''t..l
OOt-JIJ'O
v'
d' ',,>
-'
w
'"
X
<,
bX IVI X ILl/I)
E ><
«,.>
....z
...) a
a
'"w
0-
z 0 c:
S
n
u
w
'oj
.,'6'
~
--'
i., W
0
0: 0
W
.....;: z w :>:
.;:
a
..... U>
w
--'
J: s
...5
0
0 >-
"<.(>
z a
'",
a
,..
z
0 >-
10
,W
7
u
w w
u, 0
1X5~~#.j.,bb /
\ II \
a ....
<."
i= u
2·0 1.7
V\
,
s
'0
:f
:>
:f
>< -c :>:
2
0.5 ,001
~ /-~I/ ,002005 ,005
~
~ piru-;;"~
~ ,01
,0205
,05,07
. / A.//I'.// .t
,2
,3
,5.7
I
2
3
LOAD DURA nON TO rUNDAMENTAL PERIOD, TIT N
Figure 3-92
Maximum response of elasto-plastic. one-degree-of-freedom system for bil ine3r-triangular pulse (C1 = 0.100. C2 = 5.5)
5
7
10
I.S I.; 1.1 0,93 0.81 0.70 0.6\ 0.53
w
...w
~
a
. ::>:
=>
:>:
~
....
::f
0
w :>: i=
•
e
•
t[/T
TIME 0,' YIELD TO LOAD DURATION, o
0 .~
0 .~
"'0
.l~
0 .~
0 .~
0 .~
0 0 .~'" .~
0 .~
'0
.~
o
'0
.....
'0 Cf' of'
dJ
a
a
0'0'0'00 IJ'''~~N
OOy.)tJlO
....z
<,
flY
1\
E
I
'U'
Q
o
#=M=¥=t"7ldi'"M1/
cc w
n,
z o
.1'.
/1..1
c-
'of
z
U
w
W
....J
.'"
L W
w, .....
..... '""
«
o
o o
:>
....J
L
W
o
>-
~
Z
10
j ••
u W
....J
7
\.
L W
o
)£
(\
-
....
R
oco
....J
« .....
z o u
i= W
....J
L W O.
II 'vi \
'"
'A\NtrL.~ IAI'\l
::J
VI
\ .....
:>
~~ \~
'"
X
~
",.
L
.'"
'"
o
W
'"
i=
tzr~~-IA--~.I 1/ m -
07 0.5 .001
IA--Yf
.002 003
.005
.01
'
--JZ;Y---7'4Z I"'"
I'
/1
.0203
/1/1
A
IA
.05 .07
'/I~/A/A//
.1
.2
.3
i i i
.5.7
2
iii
3
LOAD DURATlON TO rUNDAMENTAL PERIOD. l/T N
Figure 3-93
Maximum response of elasto-plastic, one-degree-of-freedom system for bilinear-triangular pulse (CI= 0.068, C2 = 5.5)
5
iii i
7
\0
TlM[ or YI[LD TO LOAD DURA TlON,
tclT
. . . . . . l1'~~~~tQ~~
0000000000
. ~
»'v"~~"'''' .~ 'u' '..1"0 '\l' ',..J
.~
~.
0
o.
.0
(,
~
/1\ Y VI \ l'f'II"X'IXI /\1
',,1
~I~I\
J
'0
Cl'
tS'
o
.
'0
'"
'0
'"
o,...
00,...
/\ 1/
0
,...
'"
0
vr-::..J\ 1)< Z
t<,
I ~
-
><
<,
E
><
z o
i= u
100~~ 70~
II\.
/
XI
...E
\//
CI
tf\
1\
.'"
6'
C;V \17
.t.
1\
'v'
,
... V>
W
11. ...J
::>:
-' w
w
Q:;
Z W
w
r.,
....
o
<> z
a o
-'
:> u,
>-
o
o
Z
w
t-
...
z o t: o
o o
;::
10
W
w
7
w
5
7\
-' c-
a
::>:
3
>< ::>:
2
r¥-
<>
'\ /IV
::>: :>
~+VJl::..L
::>:
:::>
v\
...J .... w
:::f
~ :::f .... o w :::f ;::
0.7~V:b1~ 0.5 . .001
Zl2ttt .002.003 .005
~
.01
.02
.03
z ~ .05 .07.1
.2.3.5.7
2
3
LOAD DURATlON TO rUNDAM[NT AL P[RIOD, TIT N
Figure 3-94
f1aximum response of e le s tn-p l as t i c , one-degree-of-freedom system for bilinear-triangular pulse (C1 = 0.046. C? = 5.5)
5
7
10
e
e
• tdT
TIME or YIELD TO LOAD DURA TlON,
..',.., ·500
,.., ,..,
c-
'0'
v'
191 0-.1
'oJ'
100 X
<,
E x 100 0
t: u w
70
l~
W
w
,
,... IJ> W
--'
w
>-
\
30
w
a
'" '" X "
:::J
::I
I
10
1
0.7
\
"
, /'\ 7
7
3
-
:>..-.
-i/-
\
iV
V
V
~-:].
G
I
0.5 .001
'I'
\ I
1/\17
/' /
1\
'X
I
'"
1\
\-
1\
\
/
I JA
1"1 II
~\L
-\l
AI/I\ v<
K.·
~
,.,.., d'
'A II
J.
.;
r-L.\I/ K (\ . ~ ~ .... 1\ ,y.
IV
X y X~../
1/",/\/
V\
1/'\ ' ";
Irv,.....lI~
I
~I
\" r....... /\ 1\17'
:'j-./.. J
k:. ~ r\ x 1'\ ». ~ V /
'\
' ; "-
v
IAI)( -,
I X'
V
Y J
lA rx 'X .
/
1,,)(
,/
.....
IX·R V /V iJ b
/1/
//
.01
.02
//1/
YJiP0~~~v -XV k")L11X'~V bC~ Q<~IXES~
.03
l'-'!-
J: s
:::J
J: o
V
t--.h.\.
~~ r/«; - ~~ 0,53
LOAD DURA TlON TO rUN'DAMENTAL PERIOD,
TIT N
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (CI = 0.032. C2 = 5.5) .
7
10
o
.... z
o ;:: o u ..... L.. U
a
::£ .. ;;:
f)('
5
z
L..
1.1
0.70 0.61
J
<{
:::J
0,81
-t
::I
1.5 1.3
0.93
2
Z
a
~~ '9
'.5.7. I
.....
<{ I-
~o
. 1/
·.2.3
'"
0..
u
/' ,,'/' v/'...-:r//v
,05.07.1
a
o
'6'
'."'"
/
1/
'~V
~
...E
~
~'6 ~'J <·0 1.7
X "/.;
1\
1,)\ 171\
f\;!'o ~VKI>
r
Y lJ(-- k"\ X I/'d7)( ~ ~ I)
pc:
/1\
/'
....2:
<,
U
i\
1'\
,.,
6' .
-s';r
Y .~ ' \ \
'f. I I
"
v."'" .P/\ K
IX 'V
0
..,...
11\ I LV /
0
oJ'
'0 0 ,.., ,..,oJ' ,..,0
'0 v'
I--..l 1/ ,/ f...-V\ [') I/ \ r\)( IV ~
J
/1\. cr t-1. .
.
'0
'0 . oJ'
0
d' 0
/
AI/ 1\"
/
Figure 3-95
0
''0
1\ II
(\
x rxr x ',;(
},
.002.003 .005
IV
II
.
/ / // 1/ I
'"
0
>v'
•
A
\'t' 1\ \ / 1\ " 7\ 7- /\} f\:'v'f\tf\A\ Y\l/ 1,( \ rx k 1X1lf\ ~k:jxJ>~kD~
IJ:
~
V 1)\ X
'0
'"0-
'''''0
1/\ f\ I J II Y. \ X lJ III ;/ IV '\
IV 7'
-'X
.
0
',..,
If,.
::J..
1/\.
V\ !A!! 'J
v' 0
~ Iv
'x
;I},
1\/'
7\
K,
0
0
~
oJ'
V\ .
1-
0
v "
\-;/" -\ -\ \iJ\7 ,
5
V\
'J
.
'\\ " I
;
J \
Ii
--'
L.. W
, \
-t \ 'It
10
',&
V\; II \II
'V" -I'[-'tl\
r-
u
0
I):
I
. '1\
.i-,~
--" '-~"I
1-
0
0
1\1','
e/-"
0
0
'..>
~~~I~? !\ 'I! \ \ . ..:...y. :'0N9'
0
Z
0 .J)
\b1\~IJ}\i
V
50
--'
a a
0
')1' PI
w
z
',..,
\ I , II
"-~~'j:>f\ 1>..1 .
300
~
~
'0
::£
....
::E
L..
o
u ::I
t:
TIME or YIELD IO LOAD DURA TlON, .. .~
u'"
:v<
t.:I
'1.1'
~
"
'0
,.
o
000
",-J
'oJ'
'"p
'-1.
o
0
"1.1' oJ'
~
0 .~
0 .~
o
...,.. 0',.,0
o
'''6-
tdT 0 .~
o .
-0
.
0
~
.
~ \I'
'0 '0 '0 oJ' .. V' 00,..,
'0
0
'"
,..,lJ'
0._
.,.,
d'
w
>< E ><
1\ I I V I
<,
z
o
100~~ 70
~':f-\j:
V I
)lJ
->
IA
'/
1/\1 'V 11 IXI!
.,.,
6'
J'
II \'It/\ I I
"
f-
U W
'J'
I /\
50
~
0..W
w
,
.... '" ~
"
a a
i\ I
)-
10
w
7 5
a
X I X 1\1' I
z
o
t:
u
W ....J
cW
1'_"'::
o
:E
:>
:E
~
3
><
:f
-c
>Z w :::t:
>-
~
::> :f
w
0... ....J
~ o
~
~w
'\v '{A IAII 1"'1"
0..W
~
Q:
c-
o ,u
a
o
a
~
c:
...E
~
w
z o
z
>-
<,
:::t: c-
2
o
w :::t:
f=.
- - -
14~
./L-LLL
.002.003 .005
. / V Y /I/VV
.01
.02 .. 03
. / Y /V/./1/
V/Y/A'//V
.05 .07.\
.2 . . 3 . 5 . 7
LOAD DURA TlON TO rUNDAMENTAL PERIOD.
Fi gure 3-96
2
3
TIT N
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (Cl = 0.022. C2 = 50S)
5
7
\0
e TIME ~..,..~~,.",.
"r>
v"
·.,j'·0
0
...
z
. \
--
.... u
w ,
.....
'" '"
30
'"
20
--'
?3~Ai~~W
0
't-=l
.•~
,,:,0
0
0,
.~
.~
0
0
.~~",;
0 0 '0 dJ -' Cf'
Pr"O
'ttj ~~ 1\'
'\ I AI
~J1:II
'0
'0
'0
oJ' • v4 00000,..,
"0
....OJ'
0 ....
0
-A-~
1\
II
\
:1-
/\
-~.
_
--'
..>
....E
-1'....
z
N: \7o'f/\-,l/
'~"Y1l1.
'(0
0
J: s J: o
"-'~MIX
<.f; <.,J <·0 /.7
1.5 I.J I. I 0.9J 0.81 0.7.0 0.61
~
~
..
2
0.5J
0.7 0.5
~-.- -
~~ .
.
I
.001
. .
'w
=>
x«
a: w
'(IS'
V
s
~
0
0
'f-
1<-
~
f-
-J
/
7
u w
<,
',. IS'.,.,
\
z
.,.,
d'
..,
10
,-
o
.' '.
ll.
/
8 ~I} z o
0
/ t7r'X
p!\
V
a a
0
~
oJ'
I)(,
/ 1\/
~O
w --' Co. w
0
..'.v".....,..
j
70
o
oJ'
1\ I 1\ Y \I /\ f\ / I
><
E >< 100
0
'oJ'
0
1\
<,
YIELD TO LO ....D OUR .... TlON, tdT
X
300~-s: : a
.J
0
~
200 -
or o
0
'"
.~
'oJ'
~OO~
w
•
fit
.002.003 .005
.01
.02
.OJ . . 05.07.1
.2.J.5.7
2
J
LO ....D.DUR.... TlON TOrUN'O ....t.lENT....L PERIOD, TIT N
Fi gure 3-97
.Maximum response.of eJasto-pJastic. one-degree-of-freedom system for btl inear-tria~gular.pulse (Cl =-O.Ol5 •. CZ ~ .6)
5
7
to
~
S u, 0
f-
Z 0
t:
o
w oJ ".w '0
::£
=>
~
~
::£
"0
w
::£
;:::
TIM¥: or YIELD TO LOAD DURATlON, p
u' u'
,.., ,..,
~
',..,
0
0
0
.p
'J
0
0
o.
0 p
0.r . 1
'?rJ
0
0
.~
.~
0
...
0
0
'0
0
" " ,00,l~1?fWl~ , '~;'ftNf'\r1fl .,~y "1", _ ;00 ',..,
'1.1'
'0
'"
~
td T
C>
w
><
200
<,
E
><
z 0
w
-' u.
w
C\
....
'"
a-
C1 W
p
0
0
~
d'
0
~
ri~l\Izt~Ij\~\
I
II
0
J
'"0 · 0 "
0
~
0
'0
0
,..,
~
v'
0
'0
,..,
0
....0
'"
.....
~
.
',..,
IV
1/
~
'"
.J" 'J'
« ...,.
.
'6'
<:0 J: s
r
0
."
6'
~rx
z
0
....E Cl
~'-'tf-\JL.~IV,
>-
,-
I-
<,
.~
w
0
z
d'
~·IY 7(~ p,( I A
0
-'
-~~~IV
u
,
'-S
..
100~~~m
,~
W
.s-
W
Q.
...J
z
w ::£
«. a
z
:> "0
'.1. 0
I-
~'G
0
~•.J
z
;= u
u
;·0
:J u,
-'
I.; 1.5 I.J 1.1 0,9J O.B 1 0,70 0.61 0,,3
a
w
L. W
0
:>:
::> ::£
><
'"
::£
0.7~1/
o. s
.00 I
.002.003 .00,
.01
.02
.03
.O~
.07.1
.2.3.,.7
2 . 3
~
LOAD DURATION TO rUNDAMENTAL PERIOD, TIT N
Fi gure 3-98
Maximum response of elasto-plastic, one-degree-of-freedom system for bilinear-triangular pulse (Cl = 0.010. C2 = 6)
s
7
10
w
::£ :> ::£
s
::£ "0
w ::£
;=
..
•
•
nME OF YIELD TO LOAD DURA nON, tciT
.-'\1'
•
.-'.t-J
o
'",
o
.
o
000000 '", '.. 'v' 'V" t-J N
'.,i
o
\1'1.1'-10
...
0
o
0
'" '"
°000000 -0
Cd Q'I CJ\ ,.. ..... N oVtOONIJI
....z
<,
-
E
w
V
>< <, E
Il~
, ~tt~---mT\I/
sc
o
o
Mil M
z
o·
I A I VI
t:
:.VI
~/\..l/\ --_.f-\t-A--Ft-M+-I'ik-N'r--W--
u
W
,...J
~
W
o o
...J W
, .... ...'" w
<>: w Q.
...J
....z-0: w ~
«
o
z
"o
lo.
....
>--
JE==:r-,~'-k-J'k ~/I
o
e· Z
o
j::::
u
10
W ...J
7
W
5
o_
o
z o
t:
l~EFx=J=='N~
~
IV
~
~
~
w o
Iv
~
:J
::t:
.~
3
;(
«
><
-0:
...J
lo.
L7I"
~
::>
u w
..Yr" I I
2
I<-
::t:
lo.
o
W
::t:
j::::
07 .00 I
,~ .002.003
.005
.01
.02
.03
.05 .07
.1
·.2
.J
.5.7
2
J
LOAD DURATION TO FUNDAMENTAL PERIOD, T/T N
Figure 3-99
Maximum response of elasto-plastic. one-degrce-of-freedom system for bilinear-triangular pulse (Cl = 0.750. C2 = 10)
5
7
10
TIME or VIELO TO LO~O DUR~ TlON,
..
e-
',.0
0
500
J57~T\ .-:\ ::?7\ I \
JOO 200 w
><
IJ
(\- _~ \-1 I -\If-
><
100
z
70
0 t:
u
JO
0
.... '" V>
a-J W
20
>0
>-
z
10
0 t:
u
lJ
t
\
\
\
\
/
I
/
I
v \
2
"1/
1'\
1\1/ Y ·X 1~~ .j 'v \; I'\-~~
/
V
I"
o
0
'" a> 0'"
v-
tt---: hZl'\A
r-,
l)
k
,\V
'"0
0
v-
...
J
b(
j
f'(
.
0 0 0
0
0
0
N
N
'"
0
16
\; 'j.
'"
\\ 9.~
II
V\
D<
8.2
.J
I'-
7~ \
\
-.' I--
/
/ \
1\
\.
/.\1"
VI
II'" ~
II'\.
IV
r\Y
<,
-t.. /
-
'* ""
1./
~ (I>" ",'>X w ~ 11'1..
.p<\'
/>;
~ll
-I--.
/
5.3
/
/ 1/
1/ 4·6
/
J: s
r7
/
17
1/
J. o
,<.~
'l,~ ~
/
.I
V
1\ \
Y.
1/,._
I\.
IX IX
0 -. 7 0.5 .001
/
~
/1/ 1/
\><
1/
.002003 .005
\7
I\-.
Pi
x
\) ~V
/
-.::z
"'1. -/:. .I
1/
fAil>( !';II'> P'--b<
.\
rv
<'6
/
<.J
1/
\>. v . /
~~ ~[X/)k
~ -lK
Figure 3-100
It
~y
9
.~
~
lA"
AK: ,,~LS p,p
'1
f'«:~ p / r u = ....~l:*I--l-d
. . .0.
.02
.03
.05 .07.\
.2.3.5.7
LO~D DUR~ nON TO rUNDAMENTAL PERIOD,
•
I.S
')(
~ ~~
I><2KJ)/ i>X ~ ~
1.1
/ .
I'XK:~Jb<~r:x
....z
<,
...E a
o a: w
ll. ...J
....41: z
w
:E
·'0
';.0
IV\\\I/I\JI
a>
0
I~
l'
~f--tIA~\
}'.
Y):!:!I""W
f'\
I \lj J Ivl\lll\
3
><
\
0.0 ., .,
-y
:E :E
....
7-..L 11K II : \ \ I\L ~ I~ ~ / 1/
\ I
1\,,\
IV: X 1)\ D\ II
5
::>
r7
II.
)\ /
\.IVI
Po'o
a
'" ,\1-\ '"
.....
o o
0
0 0 .V' . .V' --' 0
1A I i'\, V
/ " \ f7
/
...
6.2
'! / !'\-
\
1\
\
I\"il\y .-\$ --
V \,
1\
0
'",
1\ 71\
IX IA [7\ j
II
0
0
0
Y \
/
'--,
'"
',.0
jl..-
j
-,
7 1 /
W -J L. W
0
s::T --\/--\/-- j fA- I "V I \
-J L.
w
'",
'0
0
~
~
<:>
~
50
W
w ,
\ hi
,.0
\
I
I\;'i\
<,
E
,.0 '",
v' v'
ttiT
2'
3
TiT N
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (C1 = 0.648. C2 = 10)
5
7
\0
1.3 1./ 0.93 0.81 0.70 0.61 0.53
-c
a
·Z
...::> o ....
z o ;:: <.)
....:J w
O'
:E
::>
:E
~
:E
...o w
:E
;::
-
•
T1M[ or YI[LO TO LO~D DUR~T10N.
.
R
'0
0
3 o, ,,1---"
w X
x
100
z
o
70
U
so
t:
.
,., ,., \I'
'0
o
""
"\..
0
".0
'-l
00
tclf
0 0 0 0 't/' .~ 'v' .V' !J'U'-1o
o
0
'j...i
~
0
l-J ...... 0
0'
. .)
,
0
~
'"
0
o 0 co ao l1'
~
7.\
6.2
s..!
J/
1/
o o -' w
1/
~
>-
o
fZ
1/
....z
..."-E a
o a: w
Q.
...J
«
f-
til'
W
'" '"
\~
8.2
l..
...
....l1' 0....0
....
".
-'
....,
00
V\ 1/\
1(0
o
000 U' ~ '"
v
t\=fcl~~~!J1\
LA·
. .-
o
~ ~
2001~~*~~
<,
E
.,.
v' 'v'
• ~
. '6'
z
w ;>
« o 2:
:> l..
....o z
\7\1\7
§
o
o
w
I-
U W
/1\
-'
...J
.....
/
l.. W
"t'KtiA:ti'\t--nr~J5<
a
;> :OJ ;>
/
l.. W
o
::f
,~IV
:> ::f
x«
~~
><
« ;>
:Ii
l..
o
w ::f
t: -J
0.7
Y"A~J-:
0.5
.001
.002 .005
.OO~
.01
.02
.05
l=" l--
.05.07.1
.2.5.5.7
2
3
LOAD DURAnON TO rllND~M[NTAL P[RIOD. TIT N
Fi gure 3-101
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (C1 = 0.562. C2 = 10)
5
7
\0
TIME
..',.,
'"0 500 30 U 20 w
<,
E
'"
100
z
70
:'>
-
u
50 '",
..... W
0
'",
.... 0 '"
_\~
•
r
20
t·
.i
-'
~
JA
IA.L. +
z
-"
-I-.....l
0
10
i="
u
W -J
7
W
5
u. 0
~
::::> ~
3
><
4: ~
2
k-
(;
~-' -\' I
~
-,
...,.. --'
+
I
Ib\
Ll
~L L L - : L -,L
.00.
~
~
-'_ '"
'>! l'
~
.L
.02
.03.
0
'"
0
0
N
N
'"
0
1\
"~ ~
9.5 8.]
?/
Lt.
IL
II
I)
~A" ~
,
"ll
'" ..t L.\
4:
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (C1 = 0.422. C2 = 10)
Z
W
.~
4:
,
c
.t. s .t. o
C
<..1 <·0 1.7 /·5
'.J
u
::...:::.
3
I-
~,o
4
2
-J
z
o
I-
z
<'62
.
.2.3.5.7
o
:or
Ll-
~N"'/','_ ~'"_ '"
(
r r; rr
~ ~ r. ,
.E
w '" a..
."
\' 1
rbl'
-
'"
'0'
o\>.4 J '.<. d "- /
Z I-
<,
o
0'
of
LOAD DURATION TO fUNDAMENTAL PERIOD, T/T N
Fi gure 3-102
'" ....
~t::IL11 I'
iz,
1\
r>
.05.07.,
6 6
'" 0
o '"
I-
a
,¥C ' "
'+
.
0
"-
-¥
'"
"j<"
0>
.\,lL-l
~ "~
-¥
D!
6 6
0
D
•
~
0
~. ,'J" I'>' -1-, ILl 'f1:
,-,",U If, oil>
L::[
'" \
I'.
.0'
'"
-'"
0
•
ll'I l ' """ Ll>lJI., it l:o
YI)'-
~ ~ ~ L.. 4-<~
.002.003 .005
:..
0
IV"
~~ -"'- ",j}; " '~.v-
II
v
:'> -'
V
:.\
'"
-' -"
0
'r->
;t lA: :t- &
"f'
~ I'> M A.-<: <-.. 2 i- i- DU' ~~ --I.i) O"-X'v,w £'-. "'" ~ I:s:: '" I~,-, '" ~ " C!" f'!<" Ki K~I(: < ~K~~ ~ ~~ '
0
-..'l -\ :)'.. DI l.LW..,4:: ~
1Q,p.,
0.5
'0
.'v' 0
1:1';"
0.7
..,.
o.
0
VI
~
h_
.L
.
oJ'
LA k'\
-'I
tll
t'
h
L
'&
0
0
0
0
LX LY 'I " ,
+'
0
,-
....
0
"'
.....l
0
-s
tdT
~ ~ ~" Ic:;~ "/'5;14-," • ~" ~
.L
L
A
30
0
-J W
-,
0
TO LOAD DURA TlON,
.A~ ~ 'L",.L:.'
I!.\
.L
"-
i=" W -J
0
...::
>< 0
-- --,., 'oJ'
,p~ ~ l'\,,"~ ~- ' "c, .L~~ C'.
><
'oJ'
~ ~~
~ .c
il;
,., ,.,
v' v'
or YIELD
5
7
10
I-
u w
~ ~
.~ ::::>
~
x
1.1 0.9J O.B I 0.70 0.61
....
0.53
;::
~ 0
w
~
e
-
TIME or YIELD TO LOAD DURA TlON,
,..,
~
~
v' if
,..,
,'0.1' "'r-J
,..,
'of
'0
•
l[/T o
00000000 ..p ......lo '1.1' ... 'v' v' f..> tJ 00oJ'l1'-10 ... 0
0 0 0 0 ..... ..... ..... ttl (J'l C1'1JI.°OU'l
o
0
0
0
OONlJIO
~iL±l
\1/
0
\JI~'-"'1"-J'N
1/
f\ftz~nr ~~ii717lIILhUI ~/ 6' ~
z o
r~\ "/I' >"1
~~
\
u -' u. u
o co
....'"
...""
u :~
1'-
.v, iVt~17'\-t-N~
~~
.s'cJ'
~
u
,.>
/ l/
z
;-
<,
...E o
o cc W Q.
. -'
;-
r
z
'6'
..
w
:>'
'
It><
o
z
~
1/
'-
o
.
;-
o
;-
......
z
o
,~~I/ ~lt'l
z o
jN!\-IV
t:
u
w
;-
u
...J
'-
w
u
-'
,~
~
u
o :>'
=>
:>'
x
I
-c
:>'
o :>'
v
::>
.
:>'
sc ::t:
w :>'
;::::
u
LL~
0.5 .00 I
~002
.003 .005
.01
.02
.03
*
~
.05.07.1
.2.3.5.7
LOAD DURA nON TO rUNDAMENTAL PERIOD,
Fi gure 3-103
235
TIT N
Maximum response of elasto-plastic. one-degree-of-freedom system for' bilinear-triangular pulse (C1 = 0.316, C2 = 10)
7
10
TIME or YIELD TO LOAD DURA nON,
..',..,
'"0
.,..,.
,.., ,.., '",
"'", "',..,
0
0
0
0
oJ)
'J
0
0
0
.~
500~,'~/I\II/~I\ IV
... '"
0
0
0'0 .~
0
0
. '" '" ~~
.~
30
l\
I
tdT
....0
.,..,
.~
.~
-
Z f-
......
JY
w
><
<,
-'
'"
J.
E
><
."
6'
z
0
cY
t:
1\/
u
w
..., ....a-
""
a a
".~
>0 f-
Z
0
t: u w
W
a
v
'6'
1/
<[
~ a
L.-
~v
>-
z a
i= o w
...J
L.W
1'.,
\/
a ::E ::>
'5
:1: ::J ~
;;: -c
:1:
w ::E
a
II
7
-'
L.-
~
w
-' <[
I
-' w
Q: Q.
>z
\=;zjid\71'\tJ~(\
-'
L.-
a
a
J:;r I
w
...)
::E
x
3 ::l
,{\
2
«
q C\l~\I\
:1:
L.-
a
w ::E
i=
07 0.5 .001
r
~
/ .002.003
.005
.01
.02
.03
~.
v .05 .07
,I
.2'.3.5.7
LOAD DURA TlON TO rUNDAMENTAL PERIOD,
Fi gure 3-104
I'
2
3
TIT N
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (Cl : 0.237. C2 = 10)
5
7
\0
e
e
• tdT
TIME or YIELD TO LOAO DURATION, 0'
v·
O'
',::,
0
500
h'l' p)(, &' 30Q.· ._~
200 w
><
I-A
E
><
100
z
70
0
t-.
U W
)( I T
'::' 0
1\
1\1 V_,' 11\
W...> \11 ~ ~ TT 1\"-~'-'" "'\I!\--rr
\.
'\
)(
1\1 1/ 1\
,
....'" <. '"
W
20
>0
...J
-.'"
v "Iv'
X IV
\
1\ I
I rJ> ll(
1\T \
~
I\INI/\I
(\
Z
10
U
-' .... w
c
:>' ~
::E
x
::E
1,l ..).IX-- '" . Y:' v x 1:)-
V . 1/ !\ It' I~I.)<)( IAi( il1/
1/ \1/ 3 2
0
U'
'" '"
'0 U' 0
jj
II
\
II
VN., 1\ l--. LL.)(I fiN I J. II
~
~ c\1
IVJr-
\
1\)(
'0
....0
.U'
11[\ II
II
'lr-
.,.,
r-,
1
VI\i~
~
Z l--
....
...E
'"
1\
1\
Cr.,.,
l~.
',r
~I
.
'6'
~ t:< 1\ . \ 1I' 1'\1) I.J r,.",U I'\.. ~
/~"
1\
1"'- IX' ~\ ) ~/ f...\ I}; . 17 I ~ 0) fl(....
1\
')
~
'"
~.
'0
['7
"i<
,./'1'><
tJ\ W
If"..
\
\1 -
f\ J 1'>'..)(
7~'..·\
[)<
I'
b<1\
Illl\Y[\
1/ '\ _
I\7 rX;;~[;)~V 1\
f\./"f'S(lX \V ~kZ-K) ")~r) i~K&,~[)
1r7 0.7
0.5 .001
/
'/
-
.
-
1/
/./
/1/1/
.002.003 .005
/
/
.02
.03
.5
rY'1) /\ IA ~h
"X IV IV v~l\
~~~ /"
J
.05.07.\
.~
~es~~ p ir u = "--0;;.~1\;~~
r.//l/
'.2 .. 3
LOAD DURATION TO rUNDAMENTAL PERIOD,
Figure 3-105
~
/
/ 1 / -" L - - / ' l '7v
/ .01
1/\ . IV
' .5. . 7 ..
I
2
3
TIT N
Maximum 'nispo'nse of e l asto-p l as t f c , one-degt'ee-of-freedom system for bilinear-triangular pulse (C1 = 0.178. C2 = 10)
5
7
\0
~
..: c z
~
l--
1.7 1.5
r-..... t> r--.~ J.
W
,fa <.J
Y I'
~~
..: z
l--
u,
2·0
~'\.~
1
V
/
Y\
/
0:
,fs
<·Ii.
,Ii<
1/
c
o n,
So
~
~
<,
W
., A
"l
rr-.
'"
....
0
1
II\-~l
'0
'0 v'
~
\ .
1\7~b'( ~v
1/
'0
r-,
7 5
'0
~v:: )'t--li
t-:,('
'A.. II
V\ I X
V \ 'y \ b'-1Jl '1/\1/./\/1t-1\ 1\ I'i-
-
'0
-Y. ..
f--
w
I,,, "
I
i'lllX'
V
0
-c.
\1 W
\/ \11 IV \V 1\I X
IV
IJ\
0
V \ r-V
1\
II \I
,!f--tv . 11\
\1'\
0
'"0'.
'",
"-IA-J
I,
f--
0
~I
0
0
.
',::,
s
V\
c.,
30
II
0
'.,
SO
a a...J
I
II V
~
II
ltV
_J
W
oJ'
II '.
X
0 'v' 0
0 O'
oJ'
J\ I/J./ II 11\ 1\
I}\ IV
'bl
0
'",
0
I \ II \
1\ I
Ij
I
0
0
,g 0
dlll\\ '\PI I
\11
....
0
--'oJ' . --.,.,
'1/
\1, \ II
j\-/o
'<,
'::' oJ'
or
o Z
o ;:: u
~
cw· c· ~
1.3
::>.
I. I 0.93 0.81 0.70 0.61
>< -c
.....
0.53
;::
~
~
o
w
~
,.
. ':
TIME or YlELD/ro LOAD DUR~ TlON, t[/r P'v'~t.l
't-l
'u'
'..J'
'0
........
't1'
't.l
O' '..9
o
0
.~
0
O' 0 Q. 'oJ' .... 'vI' oJ' t.1' :....1
o : "v" 0
0
0
"t-l
..
't-J,
..
0
0
0,
0
'-::"':"'... u·
-0
"'-
'0
e:tI 0
'0 CI'
'0
L1'
0
lfI
0
0
0
0
VI.
r-J
N
~
0
,~
N
0
\J'o
llCl~
'r'
,~
x E
E
II \
),
,
~
o
<,
I"'."
>< ·z
.L
~IW1A-T\
o
;::: u
of 'J ~
Iv
'0
o
o z o
z
W
'"
::>
~
..J
~
'"
_i .«
o
o
....
w '" c.
« z
w
W
o
I-
"' ~
w ,
Z I-
<,
>~
f::
l..
.~
.1. 0
o
I-
Z
o
t:
u
W
U
w
, I),
..J u,
ViA r 7'1'"7'., 1I r:::I'
I
! \ /!
l-;iT
;~,Il.!!T(
"
--~~~
W
o
'"
::>
'"
..J L..
w
o
-11~-
'"
'::>
'"
·x
«
><
'"
.<[
l..
'"
o
W
'"
;:::
,.
·/v./V/v
/1/
/
.002 .003
.oos
.ot
.02
.03
.O~
.07 .. 1
.2.3.5.7
2
3
LOAD DURATION TO rUNDAMENTAL PERIOD. T/T N
Figure 3-106
Maximum response of etasto-plastic, one-degree-of-freedom system for bilinear-triangular pulse (C1 = 0.133, C2 = 10)
5
7
10
e
e TIME
or
YIELD
ro LOAD
DURA TlON,
o
0
V"
o:
"y-J'~
• telr 0 ....
0 .....
o~oCJ''''''.
o 'o
·0
cJJ 0
'0
'0'0'000 l1' ~ ...... t-J t-J. o 0 f'.J 1.1' 0
-',-,,---u--,r--,
., Z 0-
~
-'
'"F.
6'
'"
<,
.'"
x
.:z
o
J
/
>-
.
u
o
'0
'" W
Q.
...J
Z
w ::>:
W
t: o
o z
..J
-tv
.J
, .... '" '"
E
. ",
lI'--:,-¥--JI,,~N-,4r-I'\
W
,.. w
'.J-
-
a o
'J r
...::::> o
0-
,o..
Z
~L~
.\ I
:z o
,-u
o
~
u
·w
... ..J
~
I~
~
u, W
o ::>:
::::> ::f.
x
W
o
.::>: ::::>
::>:
><
~
...o::t:
L~I\
::>:
v .~ .00 ,
002003 .005
.01
.02
03
~
.05.07.1
.2.3.5.7
2
3
LOAD DURA TlON TO fUNDAMENTAL PERIOD, TIT N
Figure 3-107
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (Cl = 0.100. C2 = 10)
5
7
10
TlMC v
',.,
u'
v'
,., ,., 'J'
0
/
/
'J'
'.'
0
'"
0
or YICLD 0 ..... 0.~ 0
TO LOAD DURATION, 0
0
0
v
'"
~
0
'v' 0
0
',., ',., ~
0
tdT 0
'..,
0 "; u'
0 ';
0
'0
'0 t1'
'0
'0
'"0 '" '"
~
0
0
... ...'" "...
'0 u'
'0
0
, _, v
w
><
<,
E
L'
sc
z
0
;::: u w -'
...,
.... '" '"
~\tTI\I-1
50~pl*
~
IZVlXIM:! \/
<,
....
.....E
'/
rv--r~IV
.,.,
1t'C~~-1Lf\
1-
.[\f\ ./
\l
/\1
l\
:::f
'0
z·
x
~
r> 5
Figure 3-108
7
\0
::> L-
a
f-
Z
a u
t:
w .s
I.,
"w
1.5 I. .3
:::f
1.1 0.93 0.81 0.70 0.61 0.53
p/r'
~
•
<.0
~~~I\
«:
0
<.J
~~~~~~
~
z w
•IS'
<.f>
v
~
::>
«:
./.0
.'
u
-'
. .
·s
~
0
LW 0
-' f-
.J
;::
w
-s
'.1'
30
o
0 0 ~ w a,
IS'
n -' w
Z 0 i:
f-
'r'
.A
I. W
(")
IV
z
d'
0
::>
:::f
>< «: :::f
"0
w
:::f
;::
e
•
•
TIME OF YIELD TO LOAD DURA TlON, tEiT
o
0
0
0
0
"-1 '0.1' 1iI'" 'v" OOl1'''''~
...0
o
0'
v'
o
.,.,
.s-
o.
0 '""
.~
0
0
~
o
'0 OJ
'....
o
0 Ol
0
'"
o
0 ...
l1'
'0
v4
,.., 0,..,
'0
~
00,..,
0
,."Ji=W=>#9i.1"==1o~ 2
IX
\+--hl;+-t,-J-A
>-
<, ..>
,x
001:;.,.,
IVI X I J.JI
,
I '07
,.-
I~
os:.oJ'
/1 \ / I X 1/
.
u
'-,
'cS'
~.
t.. w
,
~
....'"
-' <: >z w :::f
«
z
...
::>
-'
w
~
>-
~
f:' w
<>:
w
o
o a
u
o
,0
u,
z o
w
.....E
7
-'
L..
f#
x'
>-
z o i= o
W.
...
..J.
"
W
...
w
o
'0
:::f ::> :::f
a
~
::>
X
~
-c
>< <:
:::f
...o
~
w
:::f
i=
0.7
/v
/
I / •.
~
v
.002 003 .005
.01
.. 02
.03,
~ .05.07.1
.2
.3 ..,
.5
.7
1
2
3
LOAD DURA TlONTO FUNDAMENTAL PERIOD, T/T N
Fi gure 3-109
'Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (C1 = 0.046. C2 = 10)
5
7 .
TIME or YIELD TO LOAD DURA TlON, /
/
'",
0
',.,
0
'"
-0
'..>
0
0
0
'oJ' . p_. oJ' r oJ'
0
0
~
.~
0
',..,"
0 ''''0
0
'£IT
..,
0
>OJ'
0 /
'0
0
rJ>
0
0
''""
.
'0
0
'"
'0
0
'0
.... 0....0
V'
....
0
/\ 1/ \Fl/--...~
'"
I)
,.,
rJ>
..>
~
'"
/
',6'
'-'
~:
S.
~1'Jl~1
f-
'J'
.,
U
c, .. 1 c..
'0'
CJ
~o
CJ
'.:, I.~
'" '"
Q
0
cr W
D.-
2: 0
,...,
...E C
E
'J'
Z I-
<,
.t s
--'
:-
j j
,0
<.(;
:r-
<·s
f-
U W
,if
--'
c..
I\J...J:
<·0
/1"1\7
\
w C
"
::J
"><
~
~ pir;;-'Q;,.~t tHcl
0.7~ro .002.003 .005
~/ .01
~
1.
.02 .03
.05.07.1
.2.3.5
.7·
2
3
LOAD DURATION TO rUNDAMENTAL PERIOD, T/TN
Figure 3-110
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (CI = 0.032. C2 = 10)
5
7
10
Z
w ::f
. Q
::J
"0
'0
~
z
'"
I-
1.7 I.S 1.3 I. 1 0.93 0.81 0.70 0.61
0.53
I-
z 0
i= 0
~
u,
w
c
::f' ::J
::f
X
. "0
w ::f
i=
e
•
•
TIME or YIELD TO LOAD DURA TlON,tdT p
',..>
v' v'
500 ~-fT'
,..> ,..> OJ' '0
,.. ,..
0
',..>
'oj'
--:X'-,'-"
0
0
,p
'...1
0
0
0
'~
0 p
0
~
oJ'
v'
.
0 ',..>
0
... 1"\71\"V--'"
0 ',..> 0
0
...
0,
,;' v'
0
0
0
'0
0
oJ' '0 0 "J. 0 . . -.--.-,
'"
'0
'0
'0
.... ....OJ' ....0
~
V'
0
"'N'i',V' \7' ., 'X' AI r M l~ .'\ J. ,.. , . '\If I.'.
300 200
IX
E
x z
0 f=
o
w -'-' u, w
\00 7a
./, V ;,p., '\.
w
,
.... '" '"
-i
W
-i
VO: r\7'
\
7
\
VJ 1\
5
7
x
1\
Y
'\1'1'\ \. 71\
3
~
2-1-\--'"·
J
'fV-l-f .x
1\
',X
/ \
\ /
V IV
A =~V\_~I~ ~
...J
.
J: o ~'6 ~'J 2,0
;l«
In:
,\.LV.
1,7
1, S
"
II
YiI'Y \! \/\ II)( i.
d·' -j-'V / ,,), 1'-.:. -, 1\ /1\
4(~~~ 0,7
p/ru:=~
;L1A'j:/~~;Z'I.A'IYII.r"V
Y
/1/1/I.A'VIVl/.A'/V/A/""'-Y1I111
'T~:?¥JL/VVV .""7 7' -%'[;1-1.
0,5 " ,001
,002,003 ,005
,01
,02
,03
,05.07
I
I
I
,
,1
.2
,3
,5
,7
2
5
LOAD DURATION TO rUNDAMENTAL PERIOD. T/T N
Figure 3-111
1111111
177f:7P'77V
Maximum response of elasto-plastic, one-degree-of-freedom system for bilinear-triangular pulse (Cl = 0.022, C2 = 10)
5
7
w
t-
J: s
,~
a
-!':;.
':0
1111
J 1\
0 O;
':
f\1'
\
.s-
c,
I~
Y II
/
E
'"
'''':>
.If+--¥--j\¥-JI'I+Ik-I'\
)I"
t<,
-'
',..
~ V'~tillLbHja=tl\.J'h\-ttr,-~'·--
IT
'1.
'r'\: / rv
Y
7\17' /\/ '
::t ><
'"
'IX '
-'\ - ,,/,,,,,,,-'·/\-11
~
::J
't1\ 7
~"",. \ \ IA· ~l~ ~VV '\' -;. T -7,' T ' .,
10
u, W
a
·1-\, l\i~
\ I I 'x' .\ _\ I' ./..
I-
W
.f., ~
"vR"l\" X' \
30 2a
0
U
/I'
.+-,.; II, " \ l~
\f-\Ii- ."'/\ 'y'\
1-
0
I" ,
-\7 3V,.\.,
50
>Z
/
r<"IVL 11' I-A,I/\ /
X_\, ". . '\
Z
',""
j~~:~X Vf\ ~ ~~, , ~lrv~ x'
0
a
V · 1\
-'-\i1\I-A 'Y.
X
<,
:::" \ 07
II \ t-I?!Wr<--° \ . \ -
w
(p
"
10
1.3 1,1 0,95 0.81. 0,70 0,61 0,55
'"
Z
w
~
'"
a
z
::J
u,
0' t-
z
0
;:: o w
...J .... w
a
~
::J
::t:
X
« ::t:
.... 0
w ::t: ;::
triT
TIME or YIELD TO LOAD DURA TlON, ..
.~
v,~~,;,;
'v'
'\1"0
'J-
o
't-l
0 '-::'
".9
0
0 0 _.~
':r
0
'0
.~
.~
o·
o ',-.)
...
.~
0
0
0
',...
';
IS'
C?
'0
'0
'0
~0d.lc1'
vl.
0\1'
'0
'0
,",,~1.J4
OO~
'0
0
....
....
'"
0
:r-;N:"""""'".lml
e:tb6l.....Al
~---'=l1~\f- f*'HI-'d.~~d'
.....
..> '/
\1)
1/\
....=+u~F11\t
'
'rl\~
W
,W .J
a a
...J
w
.. , r-
>-
o
/
~;;.
II
i= u W
-x
~~~
~
10 7
u, W
::f
J
X
-c
::f
J. s
::J
<.f> <.J <·0 t.r
2
p/r u = ~
...
.002.003 .005
.01
.02
/1'
.03
l . r ' / I . . . . , ; I ......
.05 .07.1
.2.3.5.7
2
5
LOAD DURATION TO rUNDAMENTAL PERIOD, T/T N
Figure 3-112
Maximum response of elasto-plastic. one-degree-of-freedom syste~ for bilinear-triangular ~ulse (C1 = 0.015. C2 = 10)
7
10
Z
l..
0 ....
z
0
i=
u
~
l.. W
a
::f
1.3 I. I 0,93 0.81 0.70 0.61
::J
0.~3
i=
·~ ~?f2~-i~'-lzl~Ltj-i1t~7'~~~~z12';1Z;~,~~2Z?:f22f~P~7-L7-t4-~~ s
0.5 .G. .00 I
w
::f
a
I. S
e
...J
e-
":0
J. o
">
~
5
::f ::J
1\'- II/
AlA
...J
a
IX W
Z
":6'
e-
z o
....
2 Q..
'o" , .... .... o
...E a
6'
z
o
co
Z
e-
<,
~
~
~ l..
0
w ::f
_.
e
e,
•
"
TIME OF YIELO TO LOAD DURA TlON, p .~
50
v'" 'v"
.... .... ~..r
'0
"'J' .-.....
0
0
'...> 0
,p 0
0
''ft
0 p J'
o' 'S
0
..... 0.~
0 v' 0
O.PJlk~"l"~'r···WId':lldM'I=A~~>¥#~
p
0
telT 0
.~
.-v'
0
0
.0
MdlvIAIJ1
0
'X
0
CI' J'
'0
0
'0 v'
p
J'
....
0
0
.... "0 ....0
'0
'"
-l\~I~
.....
...>
'"
.,.,
6'
-.,.l;:-'VN;>q..>.dA"t--1r1'<1f-t"<1It I~,}
o
t:
u
-d\+--N-'IY-t"""'lw-+-~HllI-l
W ...J
IX
w
...J .<[
'0'
.2
w :>:
.
-c
W
a
.....
Q
o
f-.
..
u.
, ..... ....
..."-E c,
z
w
Z I-
Q
a
::>
...J
L.:
w
O.
>-
1-"
of-
2
o ;=
z o
u
t:
':1 L.:
u
W ...J
W
u,
Q'
W
:>:
a :>::
::>
:>:
:::>
~
:>:
><
:>:
:>:
L.:
«
o
w :>: i=
.
07
.
0.5
~
.001
-
-
. --
-
._--I.L
.002003005
747'V7f7f7Dl ~~ -.01
'.02 .. 03
.05 .07.\
,2.3.5.7
LOAD DURATION TO FUNDAMENTAL PERIOD,
Figure 3-113
2
3
TIT N
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (Cl = 0.010. C2 = 10)
5
7:' 10
telT
TIME Of YIELD TO lOAD DURA TION, OJ ·0
'0
'..,.
°Ei' Y 7\ .3/'",:F· (;7 I "e
·50
so
20 w
sc
'E>< z
0 t:
u
w -' u. w
0
, '" ....
.... N
W 7
5 3 2
r
Z
0
W
0
:>'
<:
.If'
A"
X"
X
.. ...
y
...
~ ·01-'
~
0
C?~??~?~~?oooooo ""V4lJ4NN--
V\
0
t./'
"v
'
U\"'-IoV'oQ\.l.N
IJ\
IV
A•
V
V.·
'\A."
~~II "r:FJ.'I.
it" '-'-". r~
1'Ji' i><1X I')
1-' tz' 1_/ ..,,{.
'/J1M i'>< .!/,Y/
./1"7' .
...L /
P7J
-
.
"'v,,,
. ~iI'"
1/ . .
~ ,/.. "
'
!I
I)< I I[>< " 1/
C,·
W
1/
... , rs,
V
,/I'-.
1/
.01
~ ~~.
rV'
I;'
~~ /9
/7 /5
~
13
"
•
.9.5 8.2 7.1 6.2 5.3
4.6 4.0 3.5 3.0 2.6 2.3 2,0 t.7
.
i.s
U
K
KV
\.1
'
v-
'"'
1(": ......
'
,
.', ,-
.
"I' . ,
j
.02
.03
.05 .07.\
.2.3.5.7
2
3·.
. Maximum response of elasto~plastic. one-degree-of-freedom system for bilinear-triangular pulse (C1 = 0.909. C2 = 30)
5
7
....z
<,
k,1/ - -
lOAD DURA TlON TO fUNDAMENTAL PERIOD, T/T N
Figure 3-114
~G
I'
'"
,
- '2 /v
.002.003 .005
=,.. ,1Iff
•
V '•
r-..
O. DO'
l..HN CUI
D'
'\'-. ,''' r-.
'
.r -
1)<
'"
.
~v ~ ~
r-.4 »\
'
'"
.
"~"')~'/ '~I/ IA '> I.!' rv v wtx 'X'S: ~ U
i'Y VIVH I
.
'0,'
F-
IX-n X I /
~.I"
'v'" > f'-: II
\i V ""
tx . V
II
iJ
y ,v " , , , "\
VI"" 00
-D)Q'l
00Ul
,
\ ""
IX
~
A
N) J
~i 7'I-I-J <'~'"
IT 1/
b'J
o.
hi
I\-,l
"\....... "
lIll"L~fj:;f\~,-+X-\-l~u'I-' • I'x'-->
"
y
J
,_.lI
I-,.C
V
1-'1<
.
5
2
.... ~
-t.
~
/\ !tt\ /\ . - I'-' Ii.
\
/
~ A !I ~\"V V 1/
, A
A
s
~
-:-
',;
"""',.
'\0lc
><
:>'
.
~
T
7
.:>, ::J
'0
~",~x- ,J YI~~'~~, ~~VV X~' ,. r=t 't-'" r> I~
t: U
~t-I~
v" .6'
'p.
7'-:""-1'
12t.,\ bX._'\. .
I-L-'
0
w -' u.
..
rzcrz
0
-' w >-
t.1' .....,..
d'
10
...E o
o
'"w Q.
-'
4: .... Z
L.J
::f 4:
o Z
...
~
o ....
·z o ;= u
...'j
L.J
o
::f ~
::f
X
<:
0.93 0.81 0.70 0.61
"-
0.53
;:::
::f
o w ::f
e
•
•
'. ) ;;4
TIME or YIELD TO LOAD DURA TiON,
'0'
d' '0
'"
'J'
U' 'v"
.. "..
vi.
~
v" '0
(J'
'If'
t-J
....
~
•...1
~
'..
tEiT
"". '0 0'" ~. '~'.."J, t-J 0 U' 0
'?~ ~ ~ ~ ~ ~ ? ~ l1"
.':!'
..
lJ\
lJI,VIoNN .............. " O'VI 0 (7) VI 0
.
0
0
o
0
IJ'
00
ooO\
'"
0
1'\
~ ~I~
'" '" 0'"
'
v
II
\I'
,_~ .. i\=ltt:~U-~
00
Z l-
...)
nt
n
o
11
""
l..J Q.
z
o
IN..l
t: 'I
U
'."..... ,
""'-I'IX\II"
IVlI'V
-J
'''~I % Ni(
'\
f.,JX"
Z
,..1
l..J
-J u,
l'
l..J
w
• I-'
.... w
'" n -J
l>
~
>-
o ,-
z
r
o
tLJlf\lJ A }N YllvCllig l\."'l\ l<. tv'i.1Y1.J Y I ¥J V 1\/ ~ )(\ V't-IU'Il 0 ,
:>:
~
a z
~
':)
u,
o lZ
)1\;
o
.1:).--
i= u
.-= U
l..J
-J
l..J
u,
-J
l..J
c,
a :>:
l..J
n
M
:>:
:::>
:>:
. ir"'>l:71~
-
><
:>:
I
M"~'/}
':)
:>:
~
::>: l..
o
\
.A~/.
W ::>: i=
VA/;/
.01
.02
.03
.05.07
,I
.2.3.5.7
2
3
LOAD DURATION TO rUNDAMENTAL PERIOD, T/T N
Figure 3-115
Maximum' response of elasto-plastic, one-degree-of-freedom system for bilinear-triangular pulse (CI = 0.866. C2 = 30)
5
7
10
nME or YIELD TO LOAD DURA nON,
.
,
,
Cl I\ClI
500"· 300 200 w
»c
z
0
~
u
w
17 ~
100 70
lis
.........
~
>-
30
IV \
20
0
,z 0
w
\,
~.I
a
\
::f
L
3
><
« ::f
/ 2
\
~
--\
Y.
.001
It-
II
-iL'11
fl
\I IVr\~I\ ./\1/ ~D' I\)
1'\/ lV1\~
l7
1\
~
,
.
~
.
.
\ t'\ \\(
fJ 1:' \1
0
I"
0
'"
'" o
. 0
0
00
'" ..... 0'"
I'
~
r-;
r-,
o
(J\
.I, ~
1'\
PC I)
00
o
I"
1'\
0 0
:...
:'\.
')(
X
0 A\ 1\
~
I
i\
~~
.,
26
/
22
rr-
"I',
19 17 15
r-
J
/
13
-
..
7\ t-i .J\. M kI tv \!I\. 0-
1/
~~~
fj i!tI
~ )< k><
9. 5
/
IV
. ')
IX I!\ }Cv f\'1K~ . W
.)
\)< r-, ' I" - 1'-11, r-, ,
I" I
r-
IJ'
r-
KI'- r-
~
t'---
-
,,'
'"/
]\, ,0\
II
v
2.6 2.3
V\c-, \ K ,/ v 7-
~ I~ VX X
~~ W~
IVY k>
~ ~
1,1 I,S 1.3 1.1 0.93
~ ./
':::- ....
,.t
.... :-0
.:.l:'EP 'h-'!
~,.-:
.l.
I-"
,\ I\
~\
.x
r-;.,
\
-,
1\
•
' P /r u
/.
/
VI
,.002 003
005
.. 02
.03
,OS .07, .1'
.2
.3
LOAD DURA nON TO rUNDAMENTAL PERIOD,
Fi gure 3-116
.5.7
2
TIT N
Maximum response of e;asto-plastic, one-degree-of-freedom system for bilinear-triangular pulse (C1 = 0.825, C2 = 30)
7
....0 z
0
~
u w ..J "w
a
:::E
=>
:::E
~
0.53
~
'"
5
-c
u
=
3
:::E
0,,61
'//1' //1'
.01
u
:::E "-,
-
/
ll. ...J
O.'SI 0.70
17' 7
/
'"w
a z => I.-
4. 0 3.5 3.0
•...1
l';>(: 1\
a
a
5·;
I'itI
~
....E
....z
4. 6
v
....z
<,
Il~ 1.\ 6.l
2.0
f'\
-'v /
IX"
--
tx- IX
' 1\ 17" tT -l" '-\N .. . }1'1rJ\ ~ q,V'
1\
'\.,\1 \ '\ 11/\ I /
~ )I.. I. OJ Ir' / D,S
/
~ ....
X V
V 1\
1\' -Y,- 1-1\
X
5
',,;
/"'-1 \
I
/I.
~ •...J,
t-J
telT
0000000000 '000 -.t. lJI .. l.J'lt.-"N"" .... \/'0 lJIlJ\ ....... oU\o(Jl.(Jo.l
\I
1\
xo
::f
oc,
J
.1
7
Iv
y
r\
\
/ II 1\
10
1/\
\
1\7' 1\/ 1\/
yp 1\
...J
I.W
"V
r7\
'\ !IV IA
'oj'
'-\
~
u
V
-I- lo;;, v
'x
~
" f.
.",>l-
50
...J
w
'(1'
.lJ\
IX~ I-\-If\-
W
,
/
W'
vi' '0
v'"
....
\
I' :J 1\/. Y
u.
w
\, 'f.. I I
IV ¥v
•...J
a a
'oJ'
'0
..
'c-
ff'
><
<,
E
I~
oJ'
'" '"
'0
C)
.
10
a
:::E
-
•
• ldT
TIME or YIELD TO LOAD'DURATlON, cP6'I.1'P"~v'"t-.l~
'0 500
sc
1
~\-:
rs:
W
SO
_I
0
50
0
w
,
.... -.J
...
--' w >-
20
Z t:
u
w
a
::f
>< -c
::f
lY
'" ...\.
\-.J ~
10
.'1l. \
7
.
5
::f
::>
A
r '\ :.;
--' u, w
-:i
1"1 r
5
2
0.7 0.5 .001
-~
LV
lD.
~
~
0
f\
0
-?~~O~?~??oooo U' ~V4VlNN"""" -Ql)0'l U1~ U' lJI -a 0"'" 0 (J'l Vl 0 0 fJI 0 Q
11
j"- \ (\)
~
)
Y\
~
I'!..11 PI<. ~ 11' r-
00 lo04N
0
UI
26
v
'J 'f..
22
l)tolf r'V I"" IL
1\
~ [)(
I 1\ 1\
I'-
v 1)(
/ "'-' -'-.u>, )('-- ' \
/\
J,
b
\
1L It'
Eo .'1.
"1/
xl" _
[')
1-;'
'"
~
/
v
v
,( .
:l"
4.Eo 4.0 3.5
t-
t':
12 l::'r-- k" r-
I'l( I ,
",.;'
v t-
,~,~ r- K' I/'\,Y r-
>\It
.L'. ,
~
J.O
f-
I'
It-
IX
~j lV v
&6
~
1"'
~~K-
v. ~
;K
r'i:
-t:. . \.
or
ur.J
::I
C"
L
~
r--
.
~
"
n
~rp/ru=
.
.01
.02
.03
.05.. 07.\
.2.3.5.7
2
3
LOAD DURA TlON TO rUNDAM(NTAL PERIOD, T/T N
Fi gure 3-117
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (Cl = 0.750. Cz = 30)
7
::f <[
0
z
::> "0
J-Z
/.5 l.J
::>
'nU;q!~'t'M
S
w
1.7·
L~
':./'.I/ l"c/ V 1:::2
...J
<[ J-Z
w
11
.' -
'"
w
lL
;:: o w
1\
~.
0
l./i l ..] l.Q
V
,/ I-:::
....
~ ?V~~ IX(
J'
I'X
D-'
..h' 'rl.,
-I"X1l I\t'x
xr~~
.002003 .005
~
a
1,\
. """-
.1\
\\ ~.~
...:>
1/
..fj
11 ~JX:'.:.I"
...E
\~
lL
z
J--
<,
\;
r/
!"
~ I'
\9
\1
'6.'1.
\1"--,.'
1/'';;D
~
lJ(-
-"L IY Ii( ~ ~ 1'\
,
"-.l
/'
:..}!Z'
D~
,--.j
.
0
'-"
.Jj,
v
~)XX:"~~ ~I">:""J< L >
lJ\
'>~, Lt. . .!"'" 1\ \£\ :l\..
~ 1\ V LY. \ c: 1/\1 [:WI Pi
o
-:,.~
r- x...
..i.i L\
'£. IAv ~ j< ts, \v. .i
':' -:' ~ t--J
..l
j
II
.
I~ Jl.. t\ 'i I\
0 ....
0
IJ
..D..
~
."
~ f v. X X::L II
1\
JL
Ll
'""
l f r~ '\("\1/' \ IL\ " k::\ I ~
\
.L 1YJ..\ f !",.I~
70
u, W
'0
"G'
IX
~L
A 'iT
t:
u
....
-/
x\
/
E >< III0 0
_'.Ii. -"
r.2...\.
<,
z
.v'
.~ -1 X .L 1/ -.'£f --i
500 200
'oJ'
Q
~ilI
w
'0
10
l. I 0.93 0.8\ 0.70 0.61 0.53
0
...J
L-
a
::f
:%
>< <[ If L-
0
w ::f
;::
riME or YIELD TO LOf>.D DURA n'ON,
~oo
,,"II
.~
300
w
><
E >< \00
z
0
t:
u
w
~o
I~
W
'"
a a
30
w
10
-'
>-
I-Y'
-'
a
10
7
:l'
3
:l'
2
1\
\1
il'l
~()IX ')f')4i"V
o,~
/
/ 1 V L /
.00\
/
/
/
\,
1/ 1\ ,II\.
1\)
I\
l/\
/1"'--
""''''
K I"Nr"--r--. [)( 1\ !-1Il' \)< t'---l) r--- "17/'t-tit;>'11l \) [\ 1/:,,"t--- j I'
1"- II
'.
i'-ll'1'- ' i'-
I"
lJ<,
I/'
,/'
V-~IX;~~
I'.. / [',
I-;
f.-
I')
)
oJ:--
.l<', ~
I'
)' r-
....
L
III' \ IN)(
V
r-
'\)<,,1)\ M ~ V~..c PZ:Y)l) l/ t¥L\:V V f\
~7 ~~
~
k-'ttT
II)&O~ '?1Z1-'I:" ~
V
~
_'
1/
,~ ~PI pir u ="Qi,f!I'J~ffi
Q.
.002,003 ,005
Figure 3-118
4.6
:f
.
w
a
;Z
::> I..-
.1. 0
...
<'6
0
0
z
s:
o
w
2.0
.-1
t.»
w
1.5 I.J 1.1 0.93 0.81 0.70 0.61 0.53
~
Maximum response of elasto~plastic. one-degree-of-freedom system for bilinear-triangular pulse (Cl = 0.715, Cz = 30)
-1
":;
~'l~'/QL~+--/.--J---J-.-I-I-+-4'----l--J---c--I--I---+--I-I-++-i---'---+--i--+-+'-+-H--H .01 .02 .03 .05.07.1 ..2 . 3 . 5 . 7 2 3 5 7 10 LOAD DURA nON TO rUNllAMENTAL PERIOD, T/T N
0
w
...Z-<:
/
/ 1/ -Z. / 1'/
a
1·\
<.,J
•
...E IX
"7
~
z
l-
<,
9·'.> '/).l
J.5
.. '""1-..
\\
4.0
1--'2!-1-W'17H
V
.
I'> \;
<--"" ~ .L.
r- t:><
/ r-,
22 19 11
~.l
,....
. I'-
I\~V Vi'\- J "--i ~ / 1\ v: V IV
_'X lX \1 ')(~K lX.If).x~ ~~
1~",~l));:C2l) 0,7
IY j 1\ J<
l\.lI
K I\. 1/1\
""'I \
VI\tWrv
"
\. ,)(
""~ J,\ l\
K V I\-
\It-I.L '\N 1/
\if
V
1\.6 I'X
0
_v -x
'I'll"
)
><
-<:
-w
'.
\0 ..\ f\
00'00000000°000000 \1"~Vl~NN:"":"" -OOC7\ VI. v.N tJo 0 V" U'...a oV" o
-:'""r;IJ"....i.
1\1 1\ V\ . f\. 1\ 17' 1\""9 N
II \ 11\
IV"
r-
V
-:'" ~
,.A-
\.l7I\1 I\I
..
l\.'/A]')
>\
rv
\
lA
..,
/\
r;
"/'
-I.
IX IV IV J I\-~t---~ - \ \ /\ 1r--:V / f\ Rt-M \ A'
1\
\
~\AX
,5
:£ :::J
\
'i to
.
I..-
w
. \1
K7"'. lX
Z
w
' \ '\
I\
,'I'
It'\
I-
0
II'
J' 11\-./ \//1\1.< r,\. .1'>. . -I--IX
I I \' ,/I
0
i= u
J
1/ I~
r-
70 t- :J,
-'
w , .... ....
'\ 1\
'V'I ?1 ~I ' \ 200 I,>\-y \ I ' \~I
<,
~
cPcf'o.J'''v''-..rt-l~ '0 'oJ' "..r "p'IS' '0 "I$' .....
'0
tclT
I..-
a
.:f
::>
.::f
i<
-c
~
...a w ::f
i=
e
•
I; tdT .....0 0 . 0.l0 .0.0.0. 0. 0 .0 . 1I"'VlO/oNN--
TIME or YIELD TO lOAD DURA TlON, <1'
'0
ti"<.1'''v'-v'-,-)~
o
'\1'
."v"
.....
'd'
'0
"\I'
....
"'-'
"
0-10'
...
",.
\-.J·o~
IJl
-l
0
U\lJ'-IOlJ\OO\Vi
>< <, , E >< 100
,J " 11'\... :,L..J I\.
z
T=r2\",--
a
;:: --' I, W
o
....
a
--'
7
I.W
a
:1" ::> :1"
..L
5
11\
o
0
'"
000
"'....
0"
....0
~
Z f-
<,
...E
1!'NJt¥tl
M"p,
c
a
5Q.
~
1---lo(d-'~£>,7*'i"-l'\-I
fJ
--' .", f-
l/
~
~}\ ",illf\~V
>-
a ,z 'w
' III
JLM\1I
'w:3
i= u
1"
~
w
I
11\'
. ,1/I
u
.......""
m
~I\f\\.
o
0
00
~7
······JW~ ~ w
.. .
°00 o :.. ()l) en
IV
~
l/tw
.i-
~
~ "
z· w
.
~
:1"
c
Z
:>
I.-
a fz a t: u w
• ..J
IV_'·
I.W
c
t><
~
~
L..-
><
'" :1"
:f
i-
:01/
:> :1"
I-
x
~
I.-
'"
:::if
a
w :::if
i= .1::
~ .0.
,02
.03 .. 05 .07.1
. 2 . 3 '.
lOAD DURA TlON TO rUNDAMENTAl PERIOD,
Figure 3-119
.5
.7
TIT N
I _
2
3 ~l
Maximum 'response of elasto~plastic. one-degree-of-freedom system for bilinear-triang~lar pulse (C1 = 0;681. C2 = 30)
5· },
~
10
nME or YIELD TO LOAD DURAnO'N, le/T
500
'" '" '"
'. "o
'\
300 200 w
<,
sc
z
0
t: Ll
w
100
0
, ....'"
'" CD
30 20
r
lA \
10
7
AO
Z
~=
""""
1\
N}
1-11
'1"
r'\
11'
1\1'--
1Il
-&
0
IJl
\
V\
0
I"J .I
\'1. rx
0
l,.Il
(J'l
r-
It(
_I
I".
0
VI
0
0
.0 VI
0
,q ,1 ,'>
i'..L I'--Ar"__-JI4l
'>
I
Ii'-.
r-,
V
N
PI
r\
u;
I'l'- IV
~
1\
1\
1/
"ik"--l\1'1\ . :;I;j: f\ I"\~\'./W'· 1 \
AhA J\ tAl)!
V
IV
f<,
-"/I-.... [\~l II'\. (\/N V I\X V
,I r-, 1/
. ""
lJr--K~
r--
..
W ...J
l.. W
0
5
~
::J
:20
3
sc
:20
2
_\
)(
\
II
M
V \ /\~c-
i/:S/
IV~}
~ I\. v
1/,/i
-,
./. /
~/~/~~/~.IL§7'~l75i~-i
0.:5 1/ 001
/
' /,
.002.003 .005
0
V
ll"
F--
4.6
1-0
IJ
1-'
r-
'l.
/.:2
~ \.
r\
~Q --~ "" ; . -""
o
<.~
<·0 1.7
\
k-
\?
~
.~ 0-:
~ ~~ p/ru = ~Ft++J - "
/.5 I.J 1.1 0.93 0.81 0.70 0.6\ 0.53
A
~:jqL'/':L'/.J:V~""'-~·.j;.'--l---I-~I-++-.l..--I---l--l-+-+-++'H---'-...j.-...j--+-If-+-+-H-i .01 .02.03 .05.07.1 . 2 . 3 _ .5 .7_ 2 3 5 7 10 LOAD DuRA·nON ro rUNDAMENf.AL·P[RIOD. T/T H
Fi gure 3-120
l..
/
tc
z w :::r
I J: s
I\. ~ ~ -r
...J
~
It
<.J
n
w
Q.
I-
_
AK
iii
4·0
~ 1--.
,
0
:..-17' 'J-
1/
J ["
,,- \
~~ )0~
I~Kt>
0.7
r ~v
I I'r'-, IX Y /' ~I\M~
/
!\.f
'O~
5'.3
l-
Ll
...E
~
V
1/ f':;l1Z
I'...
tI"' /
,V"
z
I-
<,
"q.'> ,.\
HI-
-,
vr-YV4.~r.\ ~ lJ
f\
1\/
\II
&.2
I . 1\
\ II
1\
~,~o~~~oo ° 0 0 0 0 0 0 0 ""'-lfol.V
'X '\It' J \ll [ \ / \ 1X--f1f\;~~"Y Xr\,1 \'\111*
/1\
r\
rv 1\ j '\0 '\ Y.
I-
1\
;\r-V\
r\~ '".'. IV
0
II \ 1/_\ ---
-"J
Y
o
"~"'~'~'-a ·....l.·... ·~°tJ'O
~ ',.-
IX V 1\ 17 e.-L\ \i. II \ ~t- i\
~
,-.;;/: , 1\. /{
'\I'
l\
11\
fV-V --
,..J
1\/1\/1\.
I
,\
1\
/\'
0
0
I
..r "'"'0 '",
\
1\
;t'J
"
...J
,.
\ -
50 LL
0
w
¥-f?(-
70
...J l..
W.
J, o
...
/ \ VI\
J
,] to
.
II .
\ - \ -)
sc
E
X
'o-
'\I'
,.0
Maximum response of elasto-plastic. one-degree-of-freedom syste~ for bilinear-triangular pulse (C1 = 0.648. C2 = 30)
0 ~ "::>
0
Z 0
;:: o w
...J
l..
w 0
:::r :::r
::J •
><
:::r l..
0
w
:::r
;::
e
e
TlM£ or YI£LD TO LOAD DURA TlON,
-0
cP
CS'
o
'oJ'
..r
U' '..,..
'4'
V''0
~ ,..> '\1""
.~
"!'
...1
~
,
•
~
0
Ie/T
C?~.C??~?~~?
.... 0
•
'01 -lo l.JI 0
!J'
t,I\
~ X
E
x
100
z o
70
u
50
f-~ \
;:: W
rX'..-IJ\
1\
~
IX
W
'",
........
'"
~
w
>-
o
f-
~v
o
t:
u
'-W
a
0
'!'
N 0
-
-
01 !.It
v
0,0
0 . 0 VI
00
co
-
~
II'
•
~oo
II'
(tl
..
1ft-.z .... <,
...E <:>
o
Q:
w
c,
'/\I~
~
4.6
..J
4( ....z·
w
::E'
«
<:>
z
:>
WI
Z
W ..J
..,J,
'"" N
503
..J
..J
Vt
".I
c,
a a
... U\
~J])<
w <,
•
L..
o ....
I'll ~
b!
z
o
r7 i? I"-
;:: u
j
7
L.. W
p
<:>
5
v
,
::f ~
v
,/
.
::E :>' ::E
\'<
::f
>< «
V'~
L\--
::f
>< «
~\f\
::E
L..
~
o
w ::E ;::
l:2
0.5 ! / .001
X /Ih ,. 'Y / . y.
/..r..«
l?Y.I'Y/'....-r.....-:f __
.002003 .005
'.01
.02' '.0'3"
,.
.05.07
.,,'
,
., .2.3
.5
.7
2
J
LOAD DURA TlON TO rUNDAM£NT AL P£RIOP, TjT N
Fi gure 3-121
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (Cl = 0.619; C2 = 30)
5
7
10
TiME Of YIELD TO LOAD'DURA TiON, lEfT ,.
cfJ '0
o
1\ C>/I
4" "J'
50 0 :-:-71-;-V~ \ /
oJ'
~.
po
'v'
0'
'""
<-;'
0
~
~
J'''''
~~
-:' -:' iii"
"",,-V,I
.\Ii. \J .
/1 ___UA \ .I-".Li.i:Jt\ l\i'-~~ 1\"tA-.
,
11J\IAIA
~
OD 01 '-4 U' 0
-:"
0
opoooooo
iI' '. I~ l)\
\11-.1
;....
N N :.. :..
QUI
0
l1'Vi
1\7'\.1/\\(. X_\~)6J 1\ _
Ill'--
N
I"
>f'\\V\\ IX' I\c! I
?
o
0 0 ~ ~ ~
0
r-,
0
000
6
... N
o
1ft
N
0
_JI1-:
z
>-
y
<,
l7
~~~~~rr
/1~_k~..Iv
....'E a
o
1--
'"w c,
z o
5.S
~
t:
u
~fi
W
---J
L.
w ::>: o
z
:::>
---J
.... o '"
Z
W
o o
'",
---J
>-
L.
,~
o
>-
>-
o ,-
z
z o
o
;:: u
()
u '""
W
...J
w
L.
---J L.
W
~
W
o
J;
:::> J;
><
.v
::>:
.002.00.l .005
.01
.02
.03
.05.07.1
t
a
~
:f
:::>
:f
><
1M ,2.3.5.7
::>: L.
o
w :f
;::
2.
3
LOAD DURATiON TO fUNDAMENTAL PERIOD, T/T N
Figure 3-122
Maximum'response of elasto-plastic. one-degree-of-freedom system 'for bilinear-triangular pulse (C1 = 0.562. C2 = 30)
5
7
',0
• ,,;
o
dl '0
c1'
'U'
"!'
..r
..
!
..
TIME
or
........
r:J
.......
0
(J'
•
oJ'
":'
~ ""
":"
-l
..-
0
.
~ ~
...
0
'00 IJI
0
... 'v\ IX. V\ V- I\J1-I
...... 0
'¥.
•
ldr
YIELD TO LOAD DUR:ATlON. C?
0
\1" t,}\
\J\ \
0
..
IJ'l.
l}\
-I
?
~? Vt.
0
~?
N·N U'I 0
hi Xi\l
)(J
?
.- C1l Vi
0
0 u\'" 0 0
00 0' en 0 lJ\
000
CI
VI N 0 U\
N 0
1-1
1/\ '-It/I
_--_.~ ~ nztrllL ~ "LA" /L AI If ~\ x ~++-A
...
\AI \?i \
If JJ-
I'l(
z o
-
/'
u
-......L \
..J
»r
..
I V"'. A ' ..... 1 A'
I~
/
...en, r-
... '"
o
~
__
V1 I
l/
~~..
..........._ 11 jW" ~
.1-.........11
_
_ Ir 1---17 r U
K
. "",'V1<+*Idt'-J'l,.j.hl*:--V'w'11--ll-4-1l'H¥\-:;:;>l\-''''II6r-1¥-::
W
:>-
I'-..
f\ IA I Y\ II 't'l II I't: I\R--..u II"N.
o
>-
II
z o
" .I
\
1I "'11 r-, 1/11 . . . . ' · . u ' · I .I/1 1/ ~:~: 1.1 V·'Tf'i...-AJ. l-/I;r1<" I T1l1\l 'VII1-lVnM II Y"W.n1:JrLlV
....
\V\j
I~'~
,k,U'...........-....-.....,..,Jt--f7ft-t1' IJIo ..
"
,"
. IA
s-,
z
w :>
«
H rtl"-H--I.l..UI--Vr IA ~ t
o
..J
cr
w a.
>-
..u--. .
LLL.~~_~~.
I 'bI-J I AI IIVI \ '" . "'. . ~~/ --.,..........--..... ..... I
\
... ~:\,: ~ /\l
W
w
".
V
o
o
..J
-~
'11.
<,
...E «
>W
z
>-
",,-;-r--~
u
o
z
::>
"-
o >z o
t= o w
..J
,-,
--.....£.\1
-'
L W
/ \ ' / \ I \ V I \MI . - · _rH~ I - t _ -V . . - - » _ _ -.... __ -.-.,..l!':'1'":tt:T
. . . , ._. . .,
,~
.~,
'rv'-. ..Jo\~\b<-\~-) "'£1 \ ~\ A X )f.....~•.
o
'I'
"'"
~
,
"w
<,
"'IlJ~::wl::::j4:A-~
•
:> •
./ jx iX;.-::v);~\un ..r ":P'"n /\..I"'o-.Al
,-
...... ,
:>
::> ::E x
:::>
:> >< .[ :>
o
«
.1.
~
•
"-
o
w
=~
'!-j 1005005
Figure 3-123
I. 2
I 3
Maximum response of elasto-plastic. one-degree-of-freedom system' for bilinear-~riangular pulse (CI - 0.511. C2 - 30)
I
II1I1I 5 7 \0
::E
t:
tdT
liME or YIELD TO LOAD DURATION,
"0
~
0.1'
U'
'o-
v"v4~t-'
po
'6'
'Ii"
'0
'1.1'
",;
"',.,. .~. ' . ~
OOOQOOOOOO ~~'cP'-1
~
....
~
~N
N:":'"
\S"U'~OU'loO'I./o4
01$'0
,~~ ~\IX I); ,. ~1~~rtK~ Iff
r
I· \
w
<,
Ii
I Y
0
'"
o '" o
.
0
0
00
"'o ....'"
0
....0
."
-
._
o
V I 11A\1U--
G.<
a
C<
w Q.
J'
>u
1&
w -.i W
o o
-,
\1/".,
w
>-
th
a
Nd~
>-
z
a ;= u
L
ilL
I,~I
W
-oJ L.W
1"/
o
~
:J
·kW
< ~
f'"
~ ~Oi'i
.01
,02
,03
,05 ,07.1
,fJ'
,3
.5.7
2
4:
o
Z
:::J L..
o
<.j
o
~\l/
1.3 1./
Maximum response of elasto-plastic. one-degree-of-freedom syste~ for bilinear-triangular pulse (Cl = 0.464. C2 = 30)
~
<'6
1.7
3
.~
.$,0
<·0 /·5
z
f=
w
-oJ L.. W
Cl ~
:::J ~
;< 4:
0.8 ,
~
0,70 0.93 0.61
L.-
Qd!~~f.j 0,53
s:
Ii
.2
-oJ
....4:
a .-
"
LOAD DURATION TO rUNDAMENTAL PERIO~. T/T N
Fi gure 3-124
<:0
~]~i~;~~!~1
~ >(
., '6'
7
L.-
...~
"-
.II
z o
..,,
.-z
\1I]J....Y'
E
X\I~
»c
co '"
N. 11/
1/ ._ "LP..!,t...j.
E
o
00
~ ~
I
1\
>(
o -
5
'1. \0
a
w
~
•
tit
IdT
TIME or YIELD TO LOAD DURA TlON. .,.
o
rP '0
0'
'1.1'
'v"
500
.. 's-
v",-)~
..".
'0
"(I'
, VI V
'U"......
~ .:....lo
0000000000
":"~
~
r..J
":". .f:/J
•..,J,
\J'
0
0
XJ '¥ VI\17'if
\.JI·.v<~NN:-:'" 1./1 -l 0 lJ\ 0 0"1 vo
. 1.1'
'Xl~
\;\
~
300
>(
<,
E
-o
o
00
'" '" '" o 0 '"
. 0
0
00 '" N
o '"
0 N 0
R \ IIi LII-t1\l 11-1-1
1/
9,5
171IM,u..,~
ri<'JI
~~
w
o
• Z f-
<,
...E
FA----=*-4A~~L
a
a
>(
z o ;:: o
1\/
w --'
L.. W
..., ... ... 0>
'"w 0..
.D
70
A
a a
~
,.
,,..
ow. , .
'it'-
~
j
l-1'-
',j
.
'6'
>-
1\)(1 YN'\~
'0
-
f-
'z
,-, ,- _
o
::E
L..
.,. .I-....·"V
.... ''''''''''- AT"
.., JNl,(1N(1I VI V I ~.
w
Z ::J
r-1Tl\IN M \1'-
..••
Z
c
il.l"'IINI~ -.~ u,
--' w
-' f-
;::
afz a ;:: u w -'
u
w --' u.
L.. W
<,
.W
a
FTJ9*t-x::;,t7iq::;, ,
:>' ::J
:>:
.,.
>(
:>:
05r~ /~ .002.003 ,005
v
::E ::J
::E
'~ ~\ {/3F\r
;;:
a
~
0.7~ .00 I
a
1:1"
~..-r/ ,01
,02
,03
,05,07
,I
-
. ,2
,3
,5
,7
w
::E
t:
2
3
LOAD DURATION TO rUNDAMENTAL PERIOD, T/T N
Figure 3-125
Maximum ,response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (C1 = 0.383. C2 = 30)
5
7 . \0
TtM[ or YIELD TO LOAD DURATlON~ tciT ~
~ 0
o
'-!
~\ oJ'
vi.
II" II"
~
":
0
0'
500
'~ .100
.~
x
_X
200
><
~...l""'f" \
A ")1
If-
':L
><
z 0 eLl
w
\
100
50
...J I.e
W 0
,
W
.....
'"
~
.10
0 ...J ~
20
)-
Z
o
w
A
I'
10
5
:>: :£
>< >:
'-1
,JI
\ 1\
'lIY ,\ \
11\
,\
.I
2
0)
0
r<-
'"
0
00
0
0
'"o '"'" '" 0
--1
/
.rr. 1--; -'- ""-
9.5
V I-""" .'
8<
IL-[l :l':....
~r
L.
)
.
o '" o
0
->./
LJ~
f-' l::::: d::. 1
..
sz
.It V
rt t::
I'
It
l:>
1\\ / . L l If',IX
/
.,..,
'\ ~/
Ii.
1 .~
r-
t2!
.¥~v
.
-1
¥-I.Y
~ "' ..,, ,~" \(
11\ IV -> '"
L,It1b(
I~..l
~tl,._,
..LI\<
l/\ I~.)( V
~t'2's[Q(
~j)<[?92,:;;~v ~0;rx
1'.L~~jLJL·"'2P C.--/. /-, ~ . ~ t:L'. v
t:
.01
iIt
I()I~,". ~t';1?)~ / ~. ~11i'1E d\. ~
,/\ IN
j/j V
ftX"1\
V\
'.'ll11l P<,
L~ 'L" v. 0.5 1<". / .002.005 .005 .001
x.
t-:
:02
.05
~~I:;;K
t-'nrt>S
'.... .05.07.1
Do
-.j
I 2'
":0-
2
o
<'6 <,.5
I::::::
f..d:~ 'c.. ''';~ LJc_ ~ I
5
7
W. ~
-<
o
fZ
o
;:: u
W ..J
t...
1. .1
OJ
I.'
X
0.5J
10
Z t...
W
0.70 0.61
_
u
Maximum response of elasto-plastic, one-degree-of-freedom system for bilinear-triangular pulse (CI = 0.316, C2 = 30)
-<
Z
1.7 I·S
0.9.1 0,8 I
__
.J
...J
OJ
.....
.2.J.5:7
W 0..
.J: s
IX--:Ict: .~
~~ p/r
'"
"':0
<,0
~
o
o
.f.
":,,.
'\.
...E f-
Ak'..l 1'X I" -L I"., ~ V />
LOAD DURATION TO rUNDAMENTAL PERIOD, T/T N
Figure 3-126
.s'',f
'
/
Z
f.....
c 6' .J
.L!
-~X I'X" . ~ -S "'~ Y: . '\ .1. ~tt2. . ,?A j -:rh1
rv i\1olX
co '"
I'
X,
•
.L..x 1. .... 1 \
U
-
o
I'> 1/
/
o
b::: Y
c.1y,
.v
)
/\
0
r: "" '\
,->
1:7\
0
o"'fJoI.
Ll
'/.
..xx~'
l41~ )~l.L';lllLl . \L~\~ 1\ i'L Il' Liu \ Ii. 7~ll ~ .\1 \ ~-I- fnX J\ r'\
1=:1;
0
.J
...
Ii.".
OUI
.....
.c: N. ..lI. "U\
LL
IT
r \'0
0
i,,- \,t N N ., :..
lJ'
_.
-.6 V\ I.\Y hL r:. -1 C\ II I V .x LV.
0
Q
....
lJ'
~y,,-
,L \
0
'""
0
.Y ...~i.L
...J
OJ
IV
Y
J'1
If
7
u.
oJ 0
~
o.
0
':'" ':'" 'fIJ t-J otJ'
I--";
-:J'i 7-
I:X=V' I!'s:
J\
0-
;::
f~
I± k V
/\
..
1
0 0
)1' li\"", 1-+ L ' II- )(.~
70
':'"
,\!I..YL\V!l',\ £':"IL
-'l;
/\.L ,1
':"'-1
'"
II
<,
E
o
t;J'
~XLV
\::,/ \V \ \if ]~ Of "'!.. .1 +-"#1~-:1,:Xc 1
~
~ J'
C ~ ~
-<
~
t...
o W
~
i=
e
•
e
. T:!ME O~ YIELD TO LOAD OURA nON. ,.
o
rJI '0
6"
'..,.
U' v'
po
po
v"v"~,.-J '6' ·0'.·U' ',;
~~."
.... C?'? cP~
•••• ...1
..
~
0
l1'
.'
t[/T
,??,?o~~ooooooo tJ"
0
\f'
-03(J'lIJl~ 0 0 V'i 0 0
"\)I.\,rlNN-U\ ..... 0 .lJ' 0 (J\ VI
o o
"',.., ,.., 0
0
0
V'
9,S
30
~" ..>
."
~
><
-=-'i' .
<,
E
~
><
z
0
I--
~ 17~
.s. .,J-
W
0
/
0
--' w
>::
F:1f V-+&1i
yo
Z
0
7
'6'
a
'"a.w ;i Z
w
':l'
a
z
P<
::> "-
a. z a
~1\...IV
~
I--
w --'
Lt
t'::::
0
;:: u
E
1--.
I 1lL.l\/\
--' u, w
'" '"
a
[\[~
U
...'",
Z I--
<,
I--
;:::
u w
..J
l.·w
~<1l
a
l.W
'"':l'
':l' ::>
.. _IV"
':l'
::>
><
':l'
><
':l'_ L.-
':l'
a
w _n1l.t
OJ
. A
/
~.
.002.003 .005
~~ .01
.02
.,03
~ .05 .07.1 ,
.2.3.5 . .7. I
LOAD DURA nON TO rUNDAMENTAL PERIOD.
Figure 3-127
':l'
;::
2
3 .
TIT",
Maximum response of elasto-plastic, one-degree-of-freedom system for bilinear-triangular pulse (Cl = 0.261, C2 = 30)
5
7
10
_
TIME, or YIELD TO LOAD DURA TlON,
0°
rP '0
(J'
'oJ'
v" v" t.l ~ "Cf' '0 ',J> >
V'
"o'
-:
. ...1
-: -: ..
~
-:
'0
0
'(9
oJ'
o
.~ 0
0
0
'IJ" \1'
'., ""
ttiT
C? lJ4
__
a
~
0
VI
lJ4 N
~ tool 0
0 (J'I
0 V<
0 0
0 -
00 0
0
01 U'
0
0
U'l 0
.. 0
o
'0
'"o '"
0
.....0
U'
- ... - .
_.
~='\(.'tN7'-\1P<~:;lt:::JI
z
>-
<,
:XW~ v
~-'I'h
Q
o
JZJU\'1X
w
~J
-
-E
s',J-
/
w
~~ 1~7fU~ \)
::f
« z Q
~""J7\t{ WII'
:>
l..
VI"U~
t:NXL~
o
>-
z o
~~IV
17
« z
>-
1/
LL
<>: w c, ...J
t: u
w
...J
\;,..,.\
l.. W
y\ <,
Q
::f
:> ::f
x « ::f
l..
o
w ::f t:
'1//1/
~
y,/{./
.002005 .. 005
.01
.02
.05
~
.05.07 . . 1
.2.5.5.7
I
2
J
LOAD DURATION TO rUNDAMENTAL PERIO~ T/T N
Figure 3-128
Maximum response of.elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (Cl = 0.215. C~ = 30)
5
7
10
e
e TiM[ or YIELD TO LOAD OURATlON. "
dJ
o
IS'
'0 '1'\1'
t1'
..
'v"
....
u" v'- ~ ~ '11' '0 '\/'-""
soo
~ _/ /\ ,fI oK 'Y y 300 \1' \ \, \ )
1\
w
0 0U
W
..J l... W
Cl
w ,
.... ....CD
."
>-
70
'I
0
u
W ..J u, W
Cl
::F :::> ::F
>< -c
::F
'\1' IJ'
.... ..,.
r\
0
"'. 0
\II
j
-.1
' \/
'\ 1'\
(.:..
-ll i\.. ~ )f
rvr\1V' j""'.
(J\
...,.,
o
"0"0
0
a
~ ~
~
x., . . / V' / ~ '"
"0
"0 ....o
t
~rli 1/
rI I
X_.
~ ~
0
0
'" '"0 lJ'
~ 1"\' 1111'-
1-11' '\ i
r.--
'f
\
'"
"'"'1"-\(\ \
Va. \,l N N :... :..
"I
[--.
.. 'I-A.
I
~o
'1
1\ '\
30
/\
/
20
1
'\ '\
Il(
v
'\
/\ \
l/
/ I /\
\II
./1\
10
I' I XI ~XV ~
7
"h
0-
i=
0
~~
I
II"
0 Z
'"
:..--: .J- 1. I'
It---
1rf"1 ' L
"ll
....z
-"
....E
' .,..,
<,
'/
c
~
O'
.,..,
Il' ,
,
Cl
..J W
f'\
"cP'~
I'
[7-
tdT
0000000000
0
. X \ f\
A V 1 n . . I \ '\ IV 1/\ " I ,....,.
1/\
'y \
E >< 100 z
x
~
..
X X
Y '\
I~~I
0
1/\
><
<,
1\ . ~
~
":' ":'
...l
/ \"
rv
200
-t
•
\ I \
s 3
2
\
V 1\1 J\ 1-\
,J'\ '7 ,\-~\ ~f\; t( '\ IA ~V ."\v '\ 1'\-...1\. fJ
A
A. ~
.\
VX I \.I'- /
\,lX
_I/- -I '1 ~ '--~ l' ~ I It"
1'\
~
~
1.:.r: I
~D)l<'S( ~ ~ X
"V)'" /. x '\ ~ ~'
txY~~1
~
~~0~~~
.r\
XlttX
::\-\~
\.
.A
.
II
D'J v x r-, 1\ )<
'Il'
~ '" K I
r-,
1/
'I
9;0
.,
IX\
~
~ v ~C:>
l-1r:--t< l>b '.~ '"
~Kv
I'
/
~~~ "'V~
rx
//.
.002 ,003
.cos. .
-:'
.01
.02
.03
LOAD OURAn.ON
Figure 3-129
/'/z
.05.07.1
o ....
r-.
'<,~ <;.1
i= u
1.5
v
".3 1.1
r
0.93
.7'
1
2
3
TO rUNDAMENTAL PERIOD, TIT N
Maximum response of elasto-plastic. one-degree,of-freedom system for bilinear-triangular pulse (Cl = 0.178. C2 = 30)
5
7
10
z o
~
u
c
::F :::> ::F
s ::F
0.81 0.70 0.61
u
0.~3
i=
Qi~p;~u='~~
'/
,2.J.5
« c z.
.1,0
/'
o.~
Z
u ::F
I'.
1.7
.... -
..J
0-
:::> l....
<,
Y.1V'
Q.
J:,s
<.0·
II, if /
Kx
1..:
'.
IL k1
~
II
II U
L l--'
~
1I ~
,)I"
!'}\'V 1\ ~~ t-,.
'0.7 .001
Lt1
h
I,
V ,"-1/\
X 1\ "
1\1 ~
LVII
IVI~~I~II I~
'~.f
1'"-.! \ rv-~
0
'"
u
L.-
o
::f
, r:P
C>
500.:0/
o
o~ "oJ'
oJ'.... .",r
'!P'
w
vi. '(f'I
v" ,-) ~ '0 'J' >
'"
.....1
M \ll \ V\ I X1~~ :1 IX 1 ~ 111\ /\1
... ,; '..
)(J
".-
C!
't-l '0 c:P... ...
l?
~o
X
200f~~~#~~
w
70
0, 0
i-oJ
'Ul
0
0
N :.... :..
j'x'
IY
9 5
~
~
ee '"
of-
0
0
0
o
0
0
V' o
'" '"
'" 0
nA
IX17fi1\
x
l
D
'"
.'"
IS'
.
• IS'
<,
...E a
o a: w
0...J
Z
w
::r
z
IV
2
~
v
j
.\--
Z f-
a
I
Z
.....
f-Cl>f--k+
>-
o
.'"
So .
x
~~ttn---'t--}fT';l'l
50
d'
-1\
~1<1\'
:--
LW
...'"
0
~~
.W-
-I..
a o
...J W
0 0 o U\
-OOlC7'lU',",
fI-\tAlL._ I"!
...J
o
o
oU'oO'lUl.
J.-rI7
E x
z:
0
1'\'1...\ I'X
"'o j:: o
C? o. '""lJ'1JI ... '"",......
0
I)(
300
X
telT
TIME or YIELD TO LOAD DURA TlON,
;::
o
w
I
...J
LW
a
5
::r
:::>
::r
><
::r
3 . t'1't117~~j) ~
2
L-
o w
::r
;::
v;r~ 0.7·~'"
¥/
0.5 '. .001
.002003 .005
.01
I··......r x .02 ,.03
~
.05.07.1
.2.3.5.7
lOAD DURA TlON TO rUNDAMENTAl PERIOD,
Figure 3-130 .
2
3
TIT N
Maximum response of elasto-plastic, one-degr~e-of-freedom system for bilinear-triangular pulse (Cl = 0.147, C2 = 30)
5
7
10
-
e TIME or YIELD TO lOAD DURA TlON, rfJ
'0
0' ',j'
U'
"vi:
1l" 'J>'
v'" v'- ~ ~ 'If' '0 '11'.....
--: ..1
o
':' ':' ~. p t-l 0 ~
o
0 '-1
0
'1.1" \l'
0 '..
\1'
0 'ut ....l
•
tdT
0
0
0
"vt0 'N N \II 0
500
o
-'" -'" 0
0
0"0
o
~ ~
0 ~
0
'0
'0
'"
~_.
Q.
... ... '0
,0
\lit
U\/l '1'.1 / w
><
<;
-'
....E 0
E
0 t:
>-
'/
<,
>< z
z
.,.,
~,
''.
70
,.
~
-s•oJ-
u w
...J
«
..
l •.
W
'"
W Il.
z>w
.~
-'
a
:f
«
W
0
<:0
CI
..... CD
0
..t~
::>
W
,
..,
-' >-
0 f-
Z
0 t:
10
W
7
L.-' W
5
u
...J
0
--
::>
::t
>< « ::f
2
p Ir'
-
07,.LM 05 .00 I
'~.,
,002003 ,005
.01
,02
,03
.05 ,07
.t
.2"
,3
,5 ,.7
lOAD DURATION TO rUNDAMENTAl PERIOD, T/T N
Figure 3-131
<.(1 <.J 2·0
L.-
1.7
0
1.5 I.J 1.1 6.93 0.81 0,70 0.61 0.53
::t
1
2
3
Maximum 'response of elasto-plastic, one-degree-of-freedom system for bilinear-triangular pulse (C1 = 0.121, C2 = 30)
5
7
\0
L.-
a >z a c: u
..to
I~
z.
w
...J
w
:f
::>
:f
x< :f
L.-
a
w
:f
;::
TIME or YIELD TO LOAD DURA TlON, ttlT c:P '0
e-
U'
'oJ'
..
'v'
'v"
'"
~..-J
vi.
~
«" ":' ~ ~ .. or.:. 0
'a" '0 'V'....
500 1 P/I \ A\ Ii. ...
C?
~
...l 0
000 .""
,n".". " " . . . . . " ....
\j\
'.. . '"
,
'lJ'lo :.J
o
0
0
N· t.o '" 0
'Ol
~
9·9 ~
'"
~
'" o
..
~
0
'0'0
'0
'0
'0
o
t,I\
0
'0
0
'"o ....'" ....0
CJl)O'''''''
0
...
'Z
bPr+AJMII.~...:>1 XI\
I/L~
\ fA! z o t: o w
~
...,E
.,
0-.
a
o
Q:
w
n,
.t.
",18
50
-' u.
w o
...,, .... o '"
>-
<,
a
'J'
w
":£ 4:
a
Z
.
:::>
~
L.-
1'\!J7wrv
>-
a >z
~\Jf
o
4:
Z
-'
'J
-' f-
~~
.r
~~IV
vL.\
a fz a
r-, r-,
c: u
w
t:
u
-' c..
- ...
W
..J
w
a
u,
w
~+7'?-...t.---;p.'{-.ljL-\"l'I'-'
a
":£ :::> ":£
":£
s
:::> ":£
><
...o":£
4:
J'
w
":£
;::
07l9BrdZ 0.5 . .001
.002.003 .005
....1/1.......-1/'f t.01
.02
.03
.05 .07.\
~ .2.3.5.7
\
2
3
LOAD DURA nON TO rUNDAMENTAL PERIOD, TIT N
Figure 3-132
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (e1 = 0.100. C2 = 30)
5
7
\0
e
•
•
TIME or YIELD TO LOAD DURATlON, t(/T r:P
'0
~
J'
'-!'
v"
~
po
~ ~ d'
t:>J' ~-"
0
':"'J ~ '-:'''; 'P
~
0
00000 "fP·....1 'If' .~ .U' \I' 0 1.1' \l' ....1
'?4? V' 0
~
\JI
c:??O t-J 0
-
(J'l
-
VI
c:?'oo'o'o -
0
rP 0
(J'l
l1'
""
'0'00
t-J
U'4
..
0
0
0
N
lJ'I
0
~oo
f-l
;00
w
2 00
IA~W\=1~~
X
"-
E x
z 0 >-
o
w
\1
70 ~o
--'
L~
w
,
.... .... '"
-M-;:---Y-HNI--'k>H1't:1
\I
.,
-j
L.J
Cl
T7
U1
fA
30
0
--'
'" ,r
2 0 j /'i->T :h'--'lI
Z 0
10 IV'
t-
U W
7
<,
" . Vh..\I\ .1'-..\
--'
L-
W
i>
Cl ~
=>
:1:
3
x
:1:
~P
,./ v
0
~
2
p / r ~ ';;11.. 0.7
~]±,
0,5
--V
I
.001
.
.002.003 .005
~.
.01
'--:02
.03
~
.0-5 -.07.1
.2.3.5.7
LOAD DURAJION TO rUNDAMtNTAL PERIOD, TIT N
Fi gure 3-133
I
2
3
•
Maximum response of elasto-plastic. one-degree-of-freedom 'system for' bilinear-triangular·pulse (C1 = 0.075. C2 = 30)
,s
7
10
TIME Or YIELD TO LOAD DURA TlON, tf'
'J'
V'
v
v"
'.,r'1'"
v' ,..J rl
'0' '0 'J' '-"
-: J,
-: po
-:
rl
-: 0
C!
~
~ 0
tdT
00-00000 "..,. ... 'vi' 'v' ;... i-.J :IJ' iJ' -' 0 \JI 0 CT'
o
0
:... '"'"
:0
'0
cP 0
'0 CJl \JI
'0
o U' o
'0
<1'
o
~
0
'0
0
...
...
<1'
0
~.
"~\M~~m.~r
so
Ii
.,..,
d'
~
w
Cl
.,.,
E
6'
.><:
.j '0/'
0 t:
~'A
u
w --' u. w
w
a a
.....
>-
\Y Iii
\
P
2
o ....
~
0 , ... 1\1/\
z
If>
Z
--'
::f
--'
0
'"
w
w
v
w
I-
o Q.
rrr-rr-w-
Z
'" '"
....,E
'/
X
<,
,
....z
<,
~ ~~"l
z o
;::: u
W ..J
I-
U W
."'-"
--'
'''''~.,
u,
<,
y
W
a
l~
W
r>
a
:f
::f ::> ::f
::>
;<
:f
x -c
'\1 \
::>:
t.-
~
o
w ::f t:
~ !.005 .005
Figure 3-134
2
3
5
7
10
e
•
TIME OF YIELD TO LOAD DURAnON, If'
.r
1'"
....,. 'c- 'po
v" v" t..l,-.) '~'o '\.1'....
500 l-c-J.--
~
J
~
..
~
--:
r.l
0
o'rP
tclT
000000'00 'l1' po 'v' '"", ~ 'r-J ''... l1' ~ ~ 0 IJ' 0. 0' VI
0
.~
q
11'
• o ...... o
'0
cP 0
0 (f'
oJ'
'0 oJ'
o
'0
'0
V'
~
o
0
'0
0
,..,
,..,
oJ'
0
-~
1'1\1\1'1 V IV d' 'r'
200 -'
w
'"
X
<,
E x z~
o
100 70
~~Ijt~~rn
'r'
'7;
.t
~/.--lr'A
"I'
u,
w
50
-y
A
'J
v
~
...'6'
1\7
-'
....
w
,
..,
00
Cl
o
cc w 'l-'
... 4:
Z
w
::E
o
o
o
)
«
LW
w
...z
;z
...
OJ
-'
...o
:-
or-.
z o
Z
t:
,-
w
o
u
w
o
.... u,
"7
7
w
-'
Cl
L-
W
~
o
.:>
~
::f
OJ
x «
"
::E
X
...
4:
~
0,
w
::E
;:: -;7 y, VI
VII
,00200,005
,01
/,01/ ,02
./
,03
,05,07".1
'//V/J'// .2.3
LOAD DURATION TO FUNDAMENTAL PERIOD,
Figure 3-135
,5 '~T
T
3
TIT,.,
Maximum response of e)asto-plastic, one-degtee-of-freedom system for bilinear-triangular pulse (CI ; 0.042, C2 ; 30)
5 '7
10
TIME \J" .. v' 'oJ' 'v"'" 'Cf'
(1:'
,., '<1'
y.)
~ « > ~'?
>~
..
x
<,
z i=
u
'"
'<1'
~o~o
'"
0 ~
<1'
0
~
0
0
\-oJ • f..,J o '" '"
'v' .
'"
.~
'"
0
0
.~
'0
.~
V'
'0
'0
~
0
0
0
'0 V'
'0
'" '"'" '"
0
0
... '0...
'0
'"
0
....
<1'
" :~ ~
200
0
0
0
vl\"
w
x
'?~
~\
.10
'E
or YIELD TO LOAD DURATlON, telT
100
7°l~ 50
..> '/
',,>
~
'
.10
0
..J
w
'"
0 >-
~
;:
Z
0 t:
u
w
..J u, W
0
:::E ::J :::E
x
20
MJJVItV tl~\P\IY
10
7
..J
.,
f-
5
-':0
~-
O.7j~ 0.5 , . - -- .-- ; --"-----;;r ;r 717[7 .001
.002.003 .005
.01
.02
.0.1
.05.07.1
~~~~?:Qi{j;l.~
P/
~
ru = Q;;R
.7
2
3
LOAD DURA TlON TO fUNDAMENTAL PERIOD, T/T N
Figure 3-136
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (CI = 0.032. C2 = 30)
5
.J '0
f-
e. e
0
H1~
7
Z
::J
~ .2.3 . .5
:::f
J: s
V"tI7I-'t1Pr7r [)\.
\
Z
w
.I
, .....-.-'
'6'
x
~
Q:
W
f.:t
<.J <·0
~
2
0
"-
W
, "" ,....
...) 0
6'
..J u,
c
z
f-
10
1.7 1.5 1.3 1.1 0.93 0.81 0.70 0.61 0.53
"0
Z
i=
u
j
"w
c
:::f ::J :::f
x
«
:::f
"0
w :::f
t:
e
• triT
TlM[ or YIELD TO LOAD DURA TlON,
()
1.1' .. v" ~ 'v' '"" '0' 'd'
50~o Cl,/
\1
V
v 17'
300 w
ZOO
X
<,
E x 100
z
0 eU w
~r"v
Y
C\ \ ,
·~·~ll\
y.)
'".;
~
":' p
J
\
...
~
0
0
~.
~
0
0
0
1\ /
Y I
X\t \ lX,
.\
,/1\
'"
0
'..r
-"
0 'v'
0
\
\/1\
_
r
0
~
';
0
!J'
Cf\
Il 1\
1\
'; 0
....
"
IV
'0'0 rP tJ'
0
';
I
I I \/
I!\/A '\
1\ \I
0
V
II
X /\
..'
0
't-J
0
Y V \ I \
V\
/!JI !A II.... \
1'\_I /1\ ~o\,~
'..
If'
J'.1 "
II
0
'\I'
....l
~
• ?
v :
0
IV
"I
\l \ ' .
~! jW I" IX K / / 1\ I" III Y\
'0 IJ'
'0 ..
0
11':17\~
1IV IT 7
[I
'0'0'0 V' f"J N 0 IJ' 0
0
7
v-.. V 'j
It
Y f\ ~ ~\'\~I\ V .A ~I/ tJ ./ \ ~ i\ /'l.:::tb.l\ lLi \ 1\
.,.,
d'
..,.
.,.,
6'
lQ\~~I\T\lA 1\
"'0 [':.
S
1\II.H
"of
-c.
":6'
'"
L.'~~~~
0
>0
20
,-
l~1
z
0 r-.
u
W ...J
7
X'~IYM'
•
o·
_
•••
_
•
~~ ~
l/~
• A I'" 1\1' \ 11
\
.. .'
~I
A
0
~1t--
V \ ....../ T-oVI \111"'\.11 / \ , 1\/
\
L..
W
...J
z
w :lI
0
.....,
'"w I-
W
...J W
0
a
(L
70
...J
,
E
'/
L..
W
Z l-
<,
:lI
sc.
:lI
J: s
L..
3
I-
~
~~~ /l"rrV
IT .01
V/f/ .02 ,.03
'/#/X//V .05.07.1
.2.3.5.7
1
,2
3
LOAD DURATION TO rUNDAMENTAl PERIOD, T/T N
Figure 3-137
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (C1 = 0.026. C2 = 30)
5
7
0
~'6 ~·3
N...lV'
~
::> :lI
0
l"\...L
MI~
5
":0
10
~·o 1.7 1.5 I.J 1.1 0.93 O.B I 0.70 0.61 0.53
~
a z a
;:: u w
..J
L..
w 0
:lI ::> :lI
i<
:lI L..
a
w
:f
t:
·'.4
T1t.1C ,or .YICLD TOLOAD DURA TlON, J'''v'~
.v- .... .o-
"dJ
o
0
t'"' ~ "':' ":' ":"'r!' >..J,P'~0J'
0
....lo 0
",
0
'I/' l1'
0
'..
'V"
J'-Io
o
'v'
·0
.0 t-1 . I"J\ 0
f.,I
O.
,hiT 0,. OJ ..... ,....;. l1' v4.
0 .....
0 rP
0
0
CP
n
L1'
o o
'0
'"
'0 u'
~
o
0 ,..,
o
0 ,..,
'"
0
z
;-
<,
-' '/
o
'" lJ\ 'V\IL\L: . z o
'if
f':
.,.,
6'
1"/\1 V A
Ii
;-
Z
w
..J
...., .... '" '"
'"c. ..J
:Itb ,
W--l.I.X-I-\.
W
o W
-.'rPf. ,
u
...E
L.W
::t'
C
C
~.~
C ..J W
.~\~
>-
o ,z o
L.-
o
;-
J\'~
'k-f,;
;-
Z :>
z
'0
t:
u
w
u
..J
W W
C
I"
_\LL>'
..J
,~
L.-
W C
5
::t' :J
J;
::t'
:>
><
J;
x -< J;
::r . -~A~oV1''it''
I.-
o
w :f
;::
-...x.U:I.L-..Y-Y-A.L vI J.1 /ILlL.A'..A_Y...t...01
.02
.03
.~
.05 .07.1
.2.3.5;1
1
2
J
LOAD DURATION TO rUNDAMENTAL PERIOD, TITN
Figure 3-138
. Maximum response of elasto-plastic, one-degree-of-freedom system for bilinear-triangular pulse (CI = 0.018, C2 = 30)
5
7
10
e
• TIME or YIELD'TO LO't,D DURA TlON, ,. , p v" ~ 'v' '.. '6" 'rjJ
oJ'
,-)
~
~
~
",
o
0' 'cP' '-i 0
>~"'"t-J01.1'
o
0
.~
.~
0
0
.~
.~
•
·tdT
00000
0 .~
-to . .~ 'op, ,'~
'0
''0
c:P
- 0
0
'0
-I.
0
(J'
0
'0 tJ'
0
'0
'0
'0
V'
~
o
o
,..,
'"
0
_ ... ~
I.,.,
···ctJ·~~~
<,
<,
~
E >< 0
70~
u w
50
w
30l~~¥
-' u,
w ,
....
... 'CO
'"
",'
'" IS' ..,.,
l~
11.'
.'"
-I\fit1 \1/ AI"};
,.
..
z
10
t:
u w
" w
I~IVI\I)I'
~
\Y\ X'I \ Y\
u.
",.
~
5
:::I'
::> :::I'
,AroY 111'
3
:::1'"
2
IN*"d-T"'t/''d7t-'
,"-
::.~
~.4.l(1\
.00\
.002 003 .005
'~ ~ .0\
.. 02 . . 03
1/
.l.s J
<'6
t.7
1.5 I. J . /., 0.93 0.81 0.70 0.61 0.53
p/ru=~"
~
.05.07.1
.2'.3 -
.5
.7
\ 2 -
'0
., ;·0
3
LOAD DUf/ATlONIO rUNDAMENTAL PERIOD, T/T N
Figure 3-139
'0
~
1\ 1/\
'.
<..1
.'"
IA
~.
1--
(\J(\)
,~~
><
«
v.~
Ii
-'
.-'6' •
~
IX
>-
....
<>:
S.
-'
0
c;
0
w
w
0
....E
-'
w
....
....z
d'
><
z
0
,..,
Maximum response ot: elasto~plastic, one-degt;ee-of-freedom sys tem 'for bilinear-triangular pulse (C1 = 0.013, C2 = 30)
5' 7
:,0
....z« w
::f
« z
'"
::> "-
a .... z. a
;:: u w -' u,
w.
o
::f
::> ::f
;<
« ::f
"-
a
w ::f
f::
tciT
TiMC or YICLD TO LOAD DURATiON,
o l..J
0
~
~ ,>(
><
0
W . ...J L.W
·n
..."", 00 '"
>0
fA d'
."" ',. .,..,
6'
\.
\V\
x
" "\
I
It:d~~~~
0
l/~I)
~~., VI' I
.
,_.
<:>
,
~
;:: u
w
7
c. w n
5
-"
}/
3
.~
>< "
~6'
·~o
~
f
b<
L\,.
-
\
P/r\.l~
7
as
,001
-,jLl-~~1
r'l, ,002.003 ,005
.0.
/ILLL.-
",',-1/17Yl7 .02
.03
,2
,3
.5.7
'I
2
,3
LOAD DURATION TO rUNDAMCNTAL PCRIOD, T/rN
Fi gure 3-140
Maxi~um
'"c,w ...J
response of elasto-plastic. one-degree-of-freedom system'for bilinear-triangular'pulse (Cl'= 0.010. C2 = 30)
,
Z
W ~
.
z
I-
0.93 0.81 0.70 0.6\ 0,53
fT.7/Yff//v
.05 .07.\
0
.J. o
1.1
\ '\
a
n
::>
1.7 /.5 I.J
2
~
) ....
.J. s
<'·s <'·0
l~
~
::>
s·',J-
z
t-
I-
'f~
IX:
I
II ]I
7\17\/\1/
3 0 '
tZ
'0'0 0 vtr..)N
ooooooU"o
~
50
20
'0 '0 '0 '0 '0 t;p...1Cf'!J''''
..~
70
Cl t...J
w
'"",
.J''YJ\ V f\17
E >< 100 t: u
0
\ I Y \ 1/\ I I\>
<,
z
0 ~ .~.
~'"'t.J"
o
w
0
'", '",
' 5 . ; 7, : 10
L.-
0
Z 0
;::
u
w
...J
L.W
0 :f ::J
:f
x
....
0 w :f
;::
..
e or
TIME
,.., '0
-0-. 'i-
oOO~Ai 71\
,~
'0
o
1\17-J IX
~
'!'
V'
Y1ELO TO LOAD 'DURA TlON, t[/T !'
vi-
•
..
v(J\
':' oJ)
~
..,.
\71--\,{ V II\'
':"'
':""
DO"
o
0
0
- 0
~;,g.
(\ U\ \ I K
0
0
lJI
loot 1./1, -....l 0 lJI 0
0
:..., i..n ;.. ~
0
0
0
0' 0
~oooooo
N N :...:.. (Jl
IJ<..J 'V'"
-I»OlUl.u..N °OlJlOONUI
CJ.t
[)(.
I)
J) 2
.....
<, -
....E
-_..
o
a
'"w Q.
70
..J
i~
1'\
'1~
~
w
,
...
~
D'
r-
..: toz
t-
W
~
..:
~
O'
z
...
::>
. £1.-'11'
a z a
.....
'" '"
;:: Ll w
...
..J
~
I v
W
o
~
::>
r-
~
i-
~
I\
~
0.5 .001
~
...o. w
~
;::
'-'"
.0'....
.02 .03
',05. ,07.\
,2·,.J
. 'lOAD DURATIO'N TO rUNDAMCNTAL PCRIOD,
Fi gure 3-141
-
..:
J'
" '~I .002.00.1 .. 005
~
)(
,5.7'
2'3
TIT,.,
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (Cl = 0.909. C2 = 100.)
5
7
10
TlM[ or YI[LD TO LOAD DURA TlON,
"'0
~.'-;'
~
01'
'0
(J'
o
'"
\f' ul
~
:..
--';'NCUJ
~
vl.
r-J
(J'
"..0
IJ4
vr»
~
':"" Ol
...
~x
~..,.N=t~-t'k+~i:::-XI·W-.- ~ y
?~?OO??OOOOOOOOO ...... U1.lAlANN _ .... - ( X l ( T l U ' l " VI'" a lJl l/I ..... OU'lO (f)lA 0 a t.1l 0 0 N l/I
_? :..
1.D
o
11\1 ~
\)
r\.Yl
~ B! /l:i..f\ 1'\1/ \l.
IA
,,'
w --'
'L W
..
...
..
It
...~
a.-l
~
~
o
bt
o
'"
w
. . . . . . _.I~
>-
o >z o
,. 7
~IAI~
--'
u,
w
a
,J~
::f :J
::f
><
-
"
o z
:>
L-
o ....
V"
~ ~
..
~~. rW///A/
w ::E
z o
~
i= u w -' LW
>< -c
. ....
::E
• •..l.-H"
1\
0.5 .001
~
o ::E :> ::E
_~ ~ _~.
0.7
....Z
I..--
w
u w
E
Jl'>
Nk
a
'"'
>-
<,
-'
t: u
'"o
z
t--
Q.
z o
w ,
tj
l>r-.
P\
l)l~1J
'k
-,
/
00
I-xTVIIX
/'f'-j'1 '¥I
"cU~
tdT
L.-
o
~
w
::E s:
Y
.002003 .005
.01
.02
.03
.05 .07.1
.2.3.5
LOAD DURATlON TO rUNDAM[NT AL P[RIOD,
Figure 3-142
'.7
L
2
j
TIT N
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (Cl = 0.866. C2 = 100.)
5
7
10
..
e
•
TIME OF YIELD TO LOAD DURATION. lEiT . . ~
~,,,,~
L
I \'
1J
cP
'0
"!'
q'
\I'
200 u
sc
<,
E
>(
z
0
100
J'
50
-'
1W
0
... ...'" 1
0
.., 0
-'
>0
.... z
0 t:
W
5
3
>(
"Ii'-
U\ V
2
II
I-/..\II ./ ~ 1'>(1)11' ?5 I\.dl2\. /I It: v. xv, :~~ .K --),1 />
~ VA
~~
0.7
Y;l()I ~
~
ID 0
Ul'" tA t.M IV IV Ul Ln ...... 0 tI' 0
I\. I\.
1'\ 1/ ['\
Ill.
"I'\.. "l( r>
P' 00
~~~.~!='P'P'~
.." 0
"IA. / I)(
.t'~. r-,
1"\
IX
Ii' lJ"WI1
--
at
--
0
t."t
011 en 0 Ul
J
V [)IE.
i'.
_
"
h
~
,;.
I""
a
CJII ~ eM IV 0 0 N UI
36
~
_
50 26 22
;::
\g
\1 \'i \\ ~.'i
r; 1·\ ~""~ffi~=ftfl~~ 3 l-
fl·'1 0''1
r-,
'i.~
4. 0 4.0
~K VI~_ ~ "..rP ./ .
I-- ""
~
3.5 3.0 2.6 2.3 2.0
~~
/.7
V
. L
/.5
b--'r-
~ ~V
_
r--
r
'P"
~~\tro~FfJ ~ r-
~
v: -\- H
' ' - ' P I r;;
r-
1'0"1"
>6
000
r-
V'
~ ~ t(
x,
~l=M:'H
v
:L /'
0.5 .001
~
\~
X' l)(1~ ill"
'>(
~
.\
::f ::f
~
IX ~ II\\ til. \\ f\
~
IXc:P
:..
1/VI, I\.
_r
t\V
I
•
DI'"
.r: ..)
I \,
IV
7
:::>
X ~ 1\'/\
__
•
LL
-~~~, 20 \. !/, ~ IX c& V\ f-ry_V:~~~~txim~'-L L'x I r---lX (1 10
u w -' u, 0
t\0l. V_\ .1L~~\
.-
II 1\
II
...l
30
VI
~ -'-,\
h(.
_'V.
':"
~
.x' , \ tt r-/.. ~ ~
-e-, -
Y
':"
IIXIIv
IX-~ \I " 'ii -A ''\
V-\,,j
_U
70
t:
u w
,1\ "A
-Y~ r\ fl. .1,,~\ nV k \ ,7\ .r>X I~
"f\
300
cr
...
500
11\
"!"
!""
vi-
.",.
"/1/
.002.003 .005
.01
.02
.03
I
.05.07.\
.2.5.5.7
I
2
5
'LOAD DURA TlON TO FUNDAMENTAL PERIOD. T/T'"
Figure 3-143
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (el = 0.825, C2 • 100.)
5
7
10
I.j
1.1 0.93 0.81 0.70 0.61 0.55
z
I-
<,
...E a 0
a: w Q.
""'
I-
Z
w ::f
a
z
...:::> 0
I-
z
0
t: 0
w
...w
...J
a
::f
:::>
::f
s ::f
...0 w
::f
t:
-"0',
triT .
nM,E OF YIELD TO LOAD DURA nON,
"'0 -a. .-;. 500
~
dJ
"0
o
'J'
IJ' .V'"
.. ....
V" (J'
~
....0
~
"'"
~
..._.. ..: ~-~.:f.J~ - .-,,-V~r\-ll
_ .... ?
~
.~
'00
:"!oD
? ??
0
"-I
Lot LoNN
0
0
IJI'"
IJI IJI ~
?? ? ? ? ? OUI
N"" 0 (J\
~
fI<.
00
00
o·
C» (fl. 0 IJI
lI'
lit N
.... LH 0
0
....
0
0 N V\
t1b
Z I-
II
<,
-
E
w
><
<,
,\
E ><
z 0
1\
I /~ /\
~
.~
...J
'~~i~
u
W -J <..
0 '" '"
« IzW
/
o
y
-J W
>0 Z
0 t:
10
w
7
Z
~
~
fI
~.
f.-
"'"
I>"
W 0 0
u
'"a.w
'\ I 1\
t:
....,
a o
k{
:::J
"-
o
I-
Z
o ;::: o
~
~
w
-J <..
a
w
a
" "~
~\~
::I: :::J
::I:
:::J
"
><
«
::I:
-\ I \
o"-
\I \
w
"
t:
I'-: r/h ~
0.7 0 -. 5 .00'
~~/ .. 002.003 .005
.0'
.02 .03
.05.07.'
.2.3.5.7
LOAD DURA nON TO FUNDAMENTAL PERIOD,
Fi gure 3-144
\
2
3
TIT N
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (C1 = 0.787, C2 = 100.)
5
7
\0
•
e
e
TIME Of YIELD TO LOAD DURA TlON, lE/T'
-a. -;.
~ 500
w
?
-0
,c/,\1TlX -\II \1 DC
....
cS'
tI'
"oJ'
'0 I
1.1' .U"
~
.p
'7
U" 'Cf'
0000000000
~oooooo -:-~~:"'\1';..i...tl..tNN:":'" -0 00 CTl Ul"'lJ..iN ... 0 0 UI VI:...... O.VI 0 en VI
':"':'"
,J)
......
. 1711\tt--\J' \ VI "'-Ahl{ xt,;
00
QUIOONUI
IX \
IIX' \1'\1
1)(\
~
,
~
1/11"-..
111*-"
><
E >< 100 0
-J'r-->,jl,-......i;
70
~
I/'
a a.-! f-
w
--'
10 7
~
3
~
>< ~
2
~
v
~
~
0,5 .001
1\ \
~"
I'-.~P/ru= ~
~"""-'--' ,002.003 ,005
.02
,01
.
,OJ'
,05 ,07
,I
.2
,3
"
,5
,7
I
2. 3
TIT,.. . M~xi~um response or elasto-plasti~, one-degree-of-freedom
LOAD DURATION TO fUNDAMENTAL PERIOD,
.
Figure 3-145
system for bilinear-triangular pulse (C1
= 0;750,
C2
= 100.)
.5
7
10
Q.
--'
'"
I-
':>',
w
.,6
«
.,0 5,5
::>
3,0
f-
2,6
0
1,7 /,5
I4b
l>:
w
6,1
2. .1 2,0
,;q
~ 0,7
~
,/
I 11
a
'"
I'><
K.
I •. W
::> :::;:
~,':>
.:-?
t'S'MV~
0
u
0
0
,/
\L~
~
0
)' '. 1\\
[ALH
w
j::
....E
r-;
r~,
--'
"w
Z
\'>
1 _\
u
'" w
<,
\1
_ .. f7 _--l,_,f.jI."I' --+t1
z
f-
\~
'0,1
w
0
II" ~
r Fl.
j::
w ,
,0
16 11
200W--
<,
z
l--,
1.3 1,1 0,93 0,8 I 0,70 0,61 0,53
z
~
0
z
u,
0
z
tz u
w --' "w 0
:::;: ::> :::;:
x
«
:::;: u, 0
w :::;: j::
.,
tdT
TiME or YIELO TO LOAO DURA TiON,
~
'4> ....
~
-0
(J'
'11'
'0
~ •po
'"
"vi-
V"t.Jr-J-'(f' "\D .~ .~
OJ
0000000000
?oo
o
-
~ ~
.... \Jl ... i:. i,. N·.N :. :.. 0
VI UI "oJ.9U1
0
en
000
'" VI o co 0 '" o
Vi
......... 0 IV
0 N VI
,.VI'\Jt-.\J/\ 1JNj\" ........•• _ . _...... _...._ .. _.. 1'\I?1 I!'l I/II--U.
Lll_rkliA--\AHHM1 ~
idI1~h~-PI-+*Nrl\H1'\I
w X
I \.
'"
E
a
I
o
f-;
t:
V\ll,l\
u
"/1 "l7'
1/1" /'.:/
W
I'-- 1)<
...J
~K
o a
""c,w
><:
t> r-,l/ r-.
>-
oez
o
~
IL T
Z
w
:E
v
r-t ~
~
--' .,; e-
7"?
b(
Ef ~ v
w
~
lot"'"
W ...J I~
....
E
~
rVr'r
z o
'", ., o
z
e-
<,
11~~'1111l dL
.~+'I\+~""'-N'-+
<,
x
---.
V
«
a z
~.~ ~
r
:J
'-
o ....
~
z
o
~
t:
i= u
w
u
W ...J
--' w
'-
"'-
'W
a
Cl
v
k'"
::f :J ::f
:E :J
:E X
..,;
X
:E
IV
.,; ::f
o
~ .002.003005
.01
.02
w
:E
c:
.'-.I'\..
.03
.05 .07.1
.2.3.5.7
LOAD DURA TiON TO rUNDAMENT AL PERIOD,
Figure 3-146
u,
2
3
TIT N
Maximum response of elasto-plastic, one-degree-of-freedom system for bilinear-triangular pulse (Cl = 0.715. C2 = 100.)
5
7
10
e
• T1I.lE or YIELD
~
-a.'P-.~1J
r:P '0
6" 'o.JI
\.I' ........
~
.~
TO t.J
v" 'CfI
-
'oj)
• 'ciT
LOAD DURA TlON, N •
"'"
-
_
~ ? ~ ~? ? .....a U\ .... Ut \.H N N
_? ?'
~".,:... . U)
0
0
IJI
VI "'-I 0
(Jl
0
?
0
-.(J'l
'LH
? 0 0 -
00
a
0
en (Jl
0
0 ....
,.,.
0
N
=T'
500
I"
2 0 0 tf--\,'--IV-I,~~
";]t-1
>< z
0
70
y
>-
11:> 11
It-
12
IV
IVlI\
w
0
'"
9''>
!l.1
0
E
.
1·\
--' '--
..,'",
I-
<,
-
""
t'l'
u w
z
,'>
w
E
0 N t.n
~
\
.100
>< <,
0
(Jl
0
a a
"J
'l>
0
,z
0 >-
\0
--' '--
7 !L,:__ d' 5
u w
w
a
" lj
~
::>i
r-
w a. --'
5 . .3
>-
4.6
w
4.0
a
0
1..1 I. I
2
~\l\
~
..
I
2
~
1\
}-'"
0.5 .001
.002.003 .005
.0\
..02
.03
.05 .07.1
.2.3.5.7
3
LOAD DURA TION TO rUNDAMENTAL PERIOD, TIT N
Figure 3-147
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (C1 = 0.681, Cz = 100.)
5
7·' 10
<:
z
I-
0.93 0.8\ lr 0.70 0.6\ p/ru='~ p.53
9
'1.
~
J. o
'.7 1.5
p
.I
z
::>
<·0 L>
«
J.5
<.J
17
x
-.f.
v
_ /\11\ 'I"
::F
::>
v
'Q:
II.2
W
>-
a
L..
0
z.
i= u w --' '-w
a
::F
::>
~
;;: <: ~
L..
0
w
~
c:
tclT
TIME Of YIELD TO LOAD. DURA TlON, ,0
~ ~
o
rPO'\I''''uolo~~ '0 '\1' 'u'o .... 'Cf' 'tJ) .~
-0
~
500
/ \ /1\/\ 300
t: V\ ">1\
) _'1. \ 200 w
_L /I L ,-
D
><
<,
E >< 100 2:
o
70
o
50
t: W ...J u, W
....o N
o en
a
30
a
...J
w
>-
20
o
\
X :\,
IL\
0\.\
~ 1\ C'i.-
r::t
,.Y
j\
bL N.~ 'II IX
.zr, r--\
-:!X- L
lX'..:::J
t-\
'-\
j 10
V I\
-A
.1\1 r"
"':L
en
0
o
~
X,
~
I"
UI
0
"I" f--.
1/ I'Jc:I T\.I< V I" 'l' r-, I" IL ~
"
0
N
U'l:0
'l-'l-
V . :I. i<--
II.
(.'
f-;
r-
.~.. ~ .
[J.>:<' [/' . . ' t'>
~
I'>~.
I-"
~
,,>
p- V, II;..
I'>(
r---
¥ .~
[)
0
Ul"'~NN
JII-/..a , ' i') f'-'
1'\
i""-
'r--
,\1 J lL ~ J L<) C\I/\"
X
"
. 1')
\
1\'\.\ ~
I,/i'f '1\ rIf\bX _\. 'r\ IT r-
UI.V'o -..1"0 U'I
0000000
-caD'!
..
'I
~'\ Ij\'\ lV
\1/
\. 1\
0
.
.
N.:"·:'"
-l'
.fl~"'( ~ Ie:"-
/
I\
\.<1
.; "
r--,tl
t=f!'/I~~:w.~ \~Mr:o 'v
IXI
f~ ~ I/\:' r... ~, ---.l.
Rl)'
:.... "" .... i..H iA N
....
it ~L\/~ ,\
_.
o
0600000000
'CJQ
r x ' \
I'st "'"~J'il/
.L /
l-
2: 0'
\
'~I\;,\f\ . -::\ ' '1/ S~ 1/
- - !. tDo
,'!>
V' ;-
~~ V
1/ -1
~£ ~
~
d
'1--1
W ...J
"W
a
:::f
::>
:::f
3
><
<
:::f
2
·1
0.5
l--\.L ~
l)( /
"'v'
"'"
.I ~
:-x ,,\,
•\
IY,...,l\ ~
'\-;>
I)'"
rl:V ~[).:
k%
4.6
.001
.L
-J--
..
V.
""
F--' __
J:o
l-
.01
~, f'll.o\ ' _ ~ ~ P/ r u : "~:1"tt1
..-t'
F
_
.02
.03
~ .~-"1..n\ 1\
. ,
t'I
.05 .07.1
I'
•
.2.3.5.7
I
2
3
. 5
LOAD DURA i1o.N TO rUNDAMENTAL PERIOD. T/T N
Fi gure 3-148
w
::> u,
f-':;~~ 'n1.3
1-\-11
b-
.".......
V
.002.003 .005
~ :.-\
..J
< .... 2:
J: s ,
1.7 /.5
\..
~
w
< a
<.0
t-
6- VA:
a
a
4·0
<'6
Xl::': t>
E
:::f
2-
>< / '
~ f.~~ ~
1I.. :~Yo t"\I ~ ~ ~~ .. VV
c,
..
f-...,6 /
0.7
1·\ 6. 1
11
~
(\0' \I
_'l..
-
<'.1
7 j- , 5
\\ 'I''> 'b .'l-
5.3
l-
o
z
l-
.....
Maximum response of elasto-plastic. one-degree-of-freedom syst~m for bilinear-triangular pulse (Cl = 0.648. C2 = 100.)
7
10
1.1 0.93 0.81 0.70 0.61 0.53
z
a
2:
0
;= 0
w
...J
"w
a
:::f
::>
::t: )(
<.
:::f
"0
w
::t:
;=.
e
e iclT
TIME or YIELD TO LOAD DURA TlON,
~
'6- -p.
~
'0
~
o
~
~
~
~
~
~
~
~
~
~
~
~
~:~
-
...
~????!='?????oo
000
00
a
o
til
N N Ul 0·
:... co
~
~
a
""''''UtLHN N - ClJO' til""' .... 0'" 00' lI
~ <,
[\
><
z
0
70
t: U
-'
"
...
Lot
0
N
M;
...z
~
w
><
E
• <,
...E o
o
'"D. W
..J
nR;QTTT~
...z« w
::f
«
a
0
'"',
'" ....0
z
0
:::>
-'
L..
...oz
r-
>0
,z
o u
~,1V I/'
0
~ U
Y11
A""L A
11""'o.J. ..... I 1.1
.r
I
I I
.If
~
~~~ ...~~
-'
u,
0
::f
:::> ::f
i=
~
w Cl
v
::f
:::>
'7
s
AIVI~
>< «
~
~
::f
0.7 0.5· . .00 I
m
,002.003 .005
~
.01
.02
1\
1\' \
L..
o
k
\
w
::f
i=
.
.03
.05 .07.\
,2 .
,3
'
,5' ,7..
I
2
3
TITN Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (Cl : 0.619, C2 : 100.)
LOAD DURATION TO rUNDAMENTAL PERIOD,
Figure 3-149
::f
I'""
5'· 7
10
[lME or YIELD TO LOM DURATION,
~
"0
'"
'0
z
o
~v
\
WI!.
'H f
~
0U
-
W ..J u, W
..
W
o
>-
CD
...
o
..J
:"\0
~lft"'VtVtNN--
00
~ flIt-.
lr\.1 II .JJI I
U'lfJI-..lOl,/\ 00"
Vt
VI""Vo
-ctIO" 0 lI\ 0
00
N
?
NV
1711. H 1L.i Z 0-
j(lIJ
. -g
<,
-
E
o
IV
IX
f7l1.tY"V/
o
'" w
0..
..J
«
i-
0-
z
-ti7t
•,ttk<1l....t.11l
Cl
...,,
_~~~?O~~?O~OOOOOO
O\\J'~V"t-.I~-- 1.1' 'u' .. 'If' ',p '\14 '01 ....
.
lEfT
w ~
~~~ffi
« o z
:::> L.-
o
j..-
o 0z
0Z
o
r-,
o
i=
i=
u w
W
L.-
u
..J
W
..J L.W Cl
o
N-
~
~
IV
::>
~
:::>
x «
~
x
«
-,-
~
J-\'
~~ ~"-,., "'" t,
0.7
.001
~:T////'1/
--·-t
\
,.,~...
q~y~~---_.
r::;
05!7%
"~.
>
~
L.-
,
Lei,
o
W~
U ~
~
i=
/1/
.002.00j .005
.0.
.02
.03
.05 .07.1
.2.3.5.7
LOAD DURA nON TO rUNDAMENTAL P£'RIOD,
Figure 3-150
I
.2
3
TIT N
Maximum response of elasto-plastic, one-degree-of-freedom syst~m for bilinear-triangular pulse (C1 = 0.590, C2 = 100.)
5
7
10
e
e
•
TiME or YIELD TO lOAD DURA TiON, .~ ~o!1P
....~~
500 '
b/\- I
Y. \1/ VI'
1\
w
200
x
50
--' u,
w
w , I'>
0
'"
a a
--' ,..,
>0
V
\J(
. .
If'\ 7\' 1(,\ I
0
>= u
\j .
.~
7 ,Vol / 1\ I\. V "
20 17\--W. 1\ /.
8
__ • •
oa
..
1)1
-
~
.'\
X \ II
IJ x.
1/
'\lJ
I" !'.
~
1/ k 1/ ~U
IX I\. ~. I/~
v-, IJ'
. lLlJ.
I'J(
r-
,-I-
/
I-.....
10<-
P'17
Ii
00
NN
"'0
\1
\"
I'J
\;
~
\\
~rn
'I,!>
l.,t- V V
11.2.
t:~~
7. \
,\
I\~\
r-;
a
::f
t.I
)(
Y..
~ lV
J
.:>
·2
~
I
J
r-, ~
'> b( V l>' I'>' ~
r;
::-,Ic; ;0(
:r:x
~
-.. lot:
")10
,-
N I/'lJ
~
"~
--'
..... I~W
_-
u
L•• W
,
"A 'I' ',\ :.y" 11. ' -- _..(\._...\ _l'fl\,Y' .._...
r-. '\
.
.)(
.1\
t\
\
w
::f
.\ 1\' .x
.... •
'Z
0-
<,
...E a
0
1 6.2
Q:
w
e,
t:
><
~
'oS).'"",
tclT
00000000000" . . . . . • . . . . . ooooo \0 .... \11 .... '-" LH N N co en UI .. "" 0 0 1ft V' .... 0 Ul 0 en "" 0 0 VI a 0 N
'
.
0
,
'(J'
' tJbll)) '\~Y\
'~7'l/\ iV N. Z\' ·"dlJl-.\ b( \ /
0Z
.~
, './\ X'f'. 1/ \ j~~K lX y A-I--X ~lV 1·
,J l' IX
~
'vi-
-Y.
,
jp '1~'iIMI\ >/
100 .
z
w
OY W \V\II'
\ , 'OX-It";
<,
E x
/1\
lo1'
CJ' .'l/'
0"0.
.
~I\V 300
'"
-0
. V"t.lt-J
/
D'~ b1 ~ r
iI
.t--I
....- /
i/.
r1 A
~ --1 ,....,
1""0
rz.
"'" Z-1
p;:;; ~ ....-'7 I
K
l,tV
v,
,.--
r-
~ ~
:Y
~
,01
,02
,03
,OS .07.1
,2
..3 . 5 . 7
LOAD DURA nON TO rUNDAMENTAl ('ERIOD,
Figure 3-151
a
',7 1.5 I.J 1./
Ifcr!
3
TIT N
Maximum response of elasto-plastic. one,degree-of-freedom system for bi1inear~triangular pulse (Cr= 0.562. C2 = 100.)
5
7
::f
'Z
::> '0 0-
z
'0
t: u W
--J
"'-
w
a
::f
::>
::f
'"::f
0.53
;::
p/r u =
2
w
0.93 0.81 '0.'70 0.61
~
.002'.003 .005
0: 0
<.J <,0
.~
-
Z
J. O
I'
~ ..)
~'II
J. S
"
k
--J
0-
r-
I
~
.,,(,.~
6.J
IlI--:i'
10
0
w ::f
,
.~
-
"f'
TIME or YIELD TO LOADDURATlON,
"',0-
'6-~
~
cP '0
'0
\l'
0'
'\S'
..
'...,.
....
~
t-1
vi '6'
'",
....,.
--
,'0)
". .
.··I~
°00
o
00'"
~iD~U. 0
. _~'iJ)<
l<1M
~~~
IX
:ti::--lly)}1'\l:\{ .\'1\
I)(
.... ~i...tNN:-:a.
'"" VI ...., CUI 0
00
-~I
z o
'ciT
0000000000 ll'l
:..
CD
000
0
0
".!
N
N
U\" 0 (...
en
NLI I 11\9
r..c:
Z f-
~
~
<,
...E
~ftV
~~~
"o
./
'"a.w
S'J
II'
1/
>-
;"'(,110
u
z
w
w
,w --'
,
w
.... '" o
...J
« f-
o a
~
k'\~ X\ ~ ~~ttrN:n)K, \p<,
, '-
«
o
z
...
:::>
--'
w
o
>-
o
f-
n'KYbruJtf')\(JY.
fZ
o
\1
\
'r\
,,'\./t
,-::
Z
o
WVb<[~
;:: u W
...
...J w
U
w ...J
~
u, W
a
~
...-
:::>
~
?'-.
a
~
:::> ~
>< .«
><
...o:::E
<:
~
w ...
,)
~
07 0.5
.001
",~:,
~y
.002003 .005
__"l'r...
I\._--'_' ,
I
.. , .
',-+<#~'". - ."- ...
,,
:::E
B.r
;::
~
~
.01
.02 .03
.05 .07.1
.2.5.5.7
LOAD DURA nON TO rUND.AMENTAL PERIOD.
Figure 3-152
,
I
2
5
TIT N
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (Cl = 0.511. Cz = 100.)
5
7
10
e
e nME or YIELD TO LOAD DURA nON, ....0
'0- .... ?
'0
dJ
lJ'
(f\
'0
'\1'
v'-
'v'-
.,..,a
'Cf""\O
to>
~
0'
'ut
tdT.
:...lD-&c.ft ... lJotVtNN .... o OUllJl....aoU\OO'1~
7f w
".E X
x
100·
2:
a
,-u w
70
~M I)
~~~t~~~ "YT'tj=;\rlJIJ
50
--' L.-. CJ
, .... '" : .... ,W
a a
--' w
>-
a· f2:
a
i= u
V\I
r-..\J'
.,,,
50 20
Iv f\Y l<\1.
\\
. . . . _,.
9. 5 8.2
I'
~J--
t. t7 H1J
t7
~
v-
~
s':f
11/
.
IV~I..L
":0 of
S
'/'0
rn,~
<.• <.J <·0
--'
-\-:;
::f
::> ::f
X
~/r~~
m
.002.003 .005
~'::l'<'T
po .01
.02 .p5
.05.07'.1
.2 . . 3 . 5 . 7
LOAD DURA nON TO rUNDAMENTAL PERIOD,
Figure 3-153
1
2
3
TIT N
Maximum response of elasto-plastic, one-degree-of-freedom system for bilinear-triangular pulse (Cl = 0.464, C2 = 100.)
5
7
10
a
'"'''' c,
--' 4:
2: w
::f 4:
0
:a:
::> L.-
a
I--
2:
a
i= u w
...J L.-
1.7 1.5 I.J I. I 0.95
o.J
0.70 0.61
a
0.55
i=
o.s I
4: ::f
'('
I--
'6'
w
L.W 0
<,
0
e. ;
~~ V\l7:Il.LI IV
•
z
I--
E
7./
'jz YW
IO~
-ODlo\lII"'l.ooiNN alJlOONUlO
'/
f
" ........ ,""'-1
~ooooooo
1/1('.,
...m m
11.",\, ,,,' .... y, ..
r.
000 ..-.0 0 . 0 .. .0 ..0 0 0
--:..
•
0
::f
::> ::f
x
4:
::E L.-
w
::E
TIME or YIELD TO LOAD DURA TlON, ie/T
""0
~l
-;.
~
~
~
o
0
~
~
.~
!'
~
..,.
~
~
~
'7
~.p
~
.....
00-00000000
?ooooo o
o o
°OU\OON
~:. ~ ..
i.I' .... ~ i...tN N :...:..
UlVl ..... OUlO(J\LH
-
0lI
en
lI\
..
lH
~ Nv
~
't\ 500
~
111~~-A4--!l.il
'~~v
N '"
r>
,"~
_w
o
..,
., ...
.... 'w
o
1\'Wq,\of~l~
Xli'll _. . . . _.- _._._ . . - . . -_
_
...... "
......
""~
•• -... ........ ,
~'" J: 'J-
~
~,
a f-
y1
a f= u w
.
.~
Z
~
v
/l~
'6'
a
cc w
11. ..J
4:
:z
w
~
4:
o
;z ::::>
~ ~
w
-;:-
Z 0-
<,
0-
i'::2
-_ . . . . - - --_ .. - ~
-'
o
8 2
o
a
W _J
0
E
t:
L.-
N
1.J!...lWllj
:z
u
0
L.-
a
'f
0-
\7-[L'~ /
t:
z
a
u
W
..J
\
7
L.-
I'
-'
L.-
w
o
W
o
~
~
:f
::::>
:f
-c
4: :f
-~~ _
0.5
'
,00 I
::::> ~
~ ~
L.-
a
w J;'<'l t:
::t: f=
t:N'i.~~~~~~~
~ ,002,005 ,005
,01
,02
,05
,05.07
,1
.2.5.5
LOAD DURA TlON TO rUNDAIoIENTAL PE.RIOD;
Figure 3-154
,7
I
2
5
TIT,..
Maximum response of e lasto-p last tc , one-deqree-of-freedon system for bilinear-triangular pulse (CI = 0.422. C2 = lOa.)
5
7
10
e
•
TIME or YIELD TO LOAD DURA TlON, tP
"0,...... -;., -0 suo <:>
~f,i-\
200 w
><
0
u
W -oJ
50
W
o
,
W
...'" W
30
o...J W
;:: Z
0 t:
10
W ...J
7
W
5
... Cl ~
:>
~
~
'\
\1"
II· r --.l.
f4
X"
1
\ Q,
J
/'
'fI.. ~ . j,
X /\ /
V f\7Mr..d\
f.X \
" "\~
"r-:--f\Kf\ " \.
I\ ~
,1\ 17'
j
IX'V\
!\.,( IX V)w 1~~~N~n ~ '\ ./~~ fj;
77
~'
'.
17
.00 I
'/1"/
.
~
\
j,-
. \
.02
.03
~
1.-
:Jr:;
"1. ~-y\ X
v~ ' J X "'"
..
.05,07.1
17
)<
v
"i. '
I,
I ,J. S
I .J. O <'6
2
/.5
= ';,~iUt=N-N
Maximum respOnse of elasto-plastic, one-degree-of-freedom system for bilinear-triangular pulse (Cl = 0.365. C2 = 100.)
5
7
-J
w
~
o z
:>
u,
o .... z o
u,
'\
3
W 0..
<·0
~
U
Cl:
<.J
V
~
Cl
o
t: u
I.,
r<.. t--~
E
":0
"-
LOAD DURA nON TO rUNDAMENTAl:' PERIOD, T/TN
Figure 3-155
'6'
V):~ ~ "' I' \U
I
~
~'"
~
.2.3.5.7
>./
I"
r-,
II
k'\~ p/ r ~,
z .... <,
.'"
~ 1-,.,.( 'I r-, ~" .~t-'~
I\--
:::.k1(: " "
i>
4
8<
I,6'
14' V
~." ~
~ .01
II
I-..- ....
.~~
. ' ts~. v:s.. i'\ k-'<: '- \.
.002.003 .005
It r.- '
~~
:::: __
~
~
b~ft v~
.~".
~4:v~V I/.K V
V
'[)
f1'
9.5
I'V"t.---
.A1 -" rr
II
f,(
v
rI [I p I- V
I
~ 1/ i/
V 1/
C'ONlJIO
I./,.A
7 It..I
~
"\
l%
-'V/V
r>..
I"....c. r-
'-~
0.7
~ J r-- be
I'.
IX V Wrv p V
II.-.. l!( ~ I\~l~
~ :I'J< lZ',;&:;
')
000 0 a Ultlo"""I'JN
-c»Ch
/I
N
Ij II
~IX.I.
\I I
00Ul
1/ : . . . . "
J'l ~.1
'\. 'II "1\
•\
.
,r-
r-;
?oo
Ol1'UI ..... OU'lOCflOl
\I 'L ""- /
"'~ ~r"\~
-s:'
~~~~~~?~?
o
LY
•
tclT
....aU'.v.Ul.NN--
:"l.O
il
-r-:i 1\7' J
.
,,\ -<0,'
V\ --tl,l.
1\;.; I
/6'
o,~
_~
t-J'_'.'~ '(lIto '. .
K V \rr-'lfl\k NIl\! K
'X 7\
IL'\
2
\0
1/\~bVIV' V
3 V~h
><
X I
X
\ 7r--1.
iJ
\
I\vo
.... u
"\
20
0
~ '"",_
IV"I\(I\'A . 1\ I) \y \ J\ X \ M r-\~~)\\JI\ ''...>\IJf,lI-\-. '
70
...
X \'
V <:>'
E >< 100
....
/
'/'-'~/
<,
z
u" ,'Cf'-
\JI 'v'
.",
"
~
500
tr'
'0
•
10
~.
w
Cl ~
I.J
:>
I. I
><
0.93 0.81 0,70 0.61 0.53
~
~
...o
w ~ ;;:
TIME Of YIELD TO lOAD DURATlON."lc/T
'6' ~ 500
"
~
" 0
dJ '0
~.
0'
'IJ'
N i l \1\/1 \
'..;.
vi ocT'
~
'19
~
'!J4
.~
OJ
~/\I)tV\\1
~~
~
...
? ??
_ ~ ~ O'? 0 ~ ~ :"'\D"-.IUI-'l..HtJ04NN-o o UlUl-"'OUlOO"l..I'&
0
0
000
'"
I "IIN/I'4ll 'til lX1I ;1'N/II l'J(
E
><
0
9.5
w
~.< .>
~7
><
<,
'"
i/1
~
w
.
o 0 ... .... '" 0 0 '" ...
0 0
-o '" '"
./
'".'" .1'.
0
--' u,
..
w
..., ....... !'>
n
.. ......I Al'WY .,-l">..... XI '\11/ , _.,
-~
)!7Il
,-,1 }J.L riot:--
a_J
,'"
W
>-
a s-.
~
z 0 t:
u
w
'or
1'1
u
w
7
v
!I
a
"
P=:1!
~
"AI
~
....::::l
r-
.J. o
.....
1'--"1'
<.~
~ ,"1
~~~ -I I I I I III-=::v=r:.cr I I 1,1 III. .03
.05.07.1
.2.3.S.7
1
I
I
I I I
2
3
lOAD DURA nON TO fUNDAMENTAL PERIOD. TIT N
Figure 3-156
Maximum response of elasto-plastic. one-degree-of-freedom syst~m for bilinear-triangular pulse (Cl = 0.316. C2 = 100.)
5
7
z
a
w:
::::l
.02
a
1.7
~
<:
z
<.,j <·0
a
><
~
i= u
L.. W
.01
z
J'.s
I'-
.002.003 .005
--' <: >--
<: a
v
~
~ w c,
0: 0
NV
--'
a
w
i'\.tK
~}bC
-E
'ti'
'm'
fjI
a
ltV'
z
,.
z
>--
<,
10
/.5 /. J 1.1 0.93 0.81 0.70 0.6\ 0,S3
w
--'
L..
a
~
::> :::f
>< <: ~
L..
a
w ~ ;::
..
e
•
TlM£ Or YI£L~ TO LOAD DURA TlON. l[/T
. '0' -;. ~
500 .
-5J.S..
soo -c
0
c: o w
.
w ,
..... '"
'"
a
-'
w
,.
20
0
v-:
z
0
;.:: u w
10
7
-'
If'
a
"
::J ~
5
J
><
«
"
2
«:
-..A
I~ I-X
I'it'
\1'\
.,j
v~
j
:~
II 1/
..... 0
l/I
r-: ,I
.
r-
. ~I":()
'V
I)<,
V
v./ ~ v,
l7 .
..1
~
~
i:"V';
r 1:7\1~ [l:
II(
.
-l
0.5
.L.
r;.
.001
I-'Z I.e.
:L:k:.
=
I-""
.002.003 .005
.0.
J.
<9..".
Z f-
-'
...)
IS'
o
'/
a
',.,
L
~
IZ. -r: /
Do
.s-' ',J-
.
r-
'6'
I '\
¥-
I'0:0 J.
·s.
¥
V
.t. o
'1/:\
<.f>
,L
~
<.J 2·0
....\
~ k" ~.
1.7
-
V
~ ~ I~ .' IX' __ ES). ~ p/r y = ~m:MI' ~
IX
i;':'
....\
,;.02 ,OJ .05.07..
.2.3.5.7
I
2
3
Maximum response of elasto-plastic, one-degree-of-freedom system for bilinear-triangular pulse (Cl = 0.274. C2 = 100.)
5
/.5 1.3
I.,
I 0.93
.... t::::. .<:::..
LOAD DURATION TO FUNDAM£NTAL P£RIOD, TIT N
Figure 3-157
9· S
' iii Y
M 11'
c::::::tz
V;
0.7
'L
..... ti1 -,
f"'
It!
V'
. .
K..
0
v:z
P"
f""
.
D
o 0 ..... ..... '" ..... ""0
...
~ .I )6L2()l1 LIZ..( M: ~~ '\/'h:XI}v1( ~ .d V ..-- kR.t t;:: r-:'.'kF\ f /,\ lCSO .r ~lL::1(L .::¥ ~I\ <' 1'. ~ l2S ~J8G~ P. -'t::" ~f-'" ~ v: y- cfS l"\:,. . \S5~ ~ ~:-'" -: ')l:lc'\
,L /.
11
.
r'/12
'" .. 00
000
00'"
p:-17 I
\
h(
0CD (, C7I
l:i:r /'
::-, L.l1I I" r- V , \ If IL
:-
VI
'T II
/~..k IY\
. 1ft". . P.41X-~ 'v'. , 'v 1: --i.."i. 16,\ I
en
0
,I
j~W.~.
4_
17..:!
.x.I.
~
~\l 13'11 "f.,.. i':L Dr- t( t'.-.x ::L ~ L\J )f(.~. _ ~v J v''.Ili L\. • j{
I
0.01 U"
f'{.
V \" n '-l'. 1\
tn'.BII\I\/I"'_\ \t7'1LX..,,···E'~'::;Q;J
~
0
•••••••••
0
t , . \ J..L:i:~
rv v.. \")J: II,V' . IJ
J.
2·1.J..
t,
w
III ....
:-1.D"-.&U'''l.r4l1tNN--
..
. Y-
1.\/'- '"
-&'
I \
~
"
_yf\['l tL~\~
oc
0000000000
.-0
c.L 1\
'-4
~V-i IA.\' -I'll
1.6. SO
L \
J
.'L
:'V1V
50
-' Cl
1:::.
70
~ ~\I .
tll.
yt.::"
100
k
W
.be -.X "
I~X ~,JXIYX
<,
z
c1'lo1''''''''''''~t.J-:'J' 'v' ... 'CJ' 'lSI '1J4 OJ
~ - D. \ \. 0.
20u
>(
. 1 r l '.
dJ "0
o
I.'.
w
E
;'
~
7
10
0.81 0.70 0.61
0.53
ec w 0..
.J
f-
Z W
~
«
~
2
o
f-
Z
o
t:
o
W
.J
k
w
a
~
::J ~
s ~
L.-
a
w ~
;=.
TIME or YIELD TO LOAD DURA TlON, ~
..
~
-0
rP
lJ' ."'"
G"
·0
.~
~ '..
v'o 'Cf'
"'" ',p
tool
V.
-:" t:P
or
..
oe
~
:- to 00
,\I
tdT
oC/oooooo ,
. . .N .- -
o
00
Ul"'lrI",N U'lU''"'''OlJl.
:...
CJ)
CJ'l
0
o
0
'"
0\","
:--,1
X
..-
>< E
<,
x
z a >-
u
W
-'
100~.'-
,
....'" '"
a a
--J
W
,..
0
30
20~
f-
Z
0 t:
u
w
~
X .« ~
~'" ".\
'" tf'
.."
A
1L.'MI· Ll_ .l-
r
s.'-f
1"1' \ J K.
J ii' r--. V '\
II
,f--\j81 ~ :::V~
or
.~
I. II
V V
1"-
V
'6'
~
,005
.0.
.02
.03
.05 .07.1
<,6
0
IX 17 M
.1
.2.3.5.7.
I
e. ,J
I
I
2
3
I
1.7 1.5 1.3 1.1 0.93 O.B I 0.70 0.61
LOAD DURATION TO rUNDAMENTAL PERIOD, T/T N
Figure 3-158
response of elasto-plastic, one-degree-of-freedom system for bilinear-triangular pulse (CI = 0.261, C2 = 100.)
Maxlm~m
J
5
I I JJI 7
~
«
....0
0.53
•. 002.003
w
.l,o
2
,DO'
....z«
~
<,0
~
--J
-f.s
k'\1~
0.5
a:
w
a
)<
t5l~,..~
a
0
:::>
3
0.7'
...E
'!"-o
7 5
'Z .... <,
Q.
..-
~
:::>
~·s
t- v·'lJL.
II
~ y~
VII
7
fuA"V"/l~
~ ~
l.-
a
~
11'
r-,
~'" NXN?}\I"
-'
W
Y
~hL
SOj~
1_ W
W
~I\
1/ I)<
0
o
'1,-11 f.-7 11, r " ' , / ~Zl
1'" V
-,
w
0
U'-.",NN 0 N I.fI 0
v'1//\
tz- /
"\ ,,'/ ..... 1"\ 'If.J l,( ,/
o 0 0
10
z
z
t: u W
...J
~
w
a
~
:::> ::f
x
«
::f ~
0
w ::f
i=
e
e
•
: TlM( OF YIELD TO LOAD DUR~ TlON, 6_
"".? ....o
rIA
d' 'J'
'0
lJ'
..
'V"
'4\
500
,:1".\
VI\ r\.Ll\
~
V"
....
-'
\
~
t-J
cIJ
'v'
'I.D.
I,
I'X
\1'lY
f'
~
.... :"'!J)
o
'iUl.IJI,v.,NN-0 0 U'U' .... OUlOO\..,.
:..
00 00 0\
o
0
l~
)I.
I)
30
ll?1 ~Q)
'-'
><
fV\_\ .li~'f;..
<,
E
><
100
z
10
0
>-
w w
50
W
"",
.... .... N
50
Cl
--'
w
20
>-
-\
l'wl J ,/,
Vi.
6 \ \-
fJ
>-
,.• W
W --'
,W o
'a
f·
5
\I
::f
« ::f
3
1
t17 0.5 00
~ ~/\
V\~
vi
t>o:
r-
Y
-
1&
&
0./
X'r--, v I~ h
TI~
f-'
0
0
U'
-
..
1\
IliA a,s
.:.I
~" ,
t2-
-"
-La.,
'"
11
II
.'"
6'
'.1'
/
~l2;: ~ r\ ~ .
\
1/
,il
V
e---
":0-
yN.; I'\, \< =" ~/\. \/
":0
~~
l.'
v11:
Y-=:-
~
r-,
x_
~
r-s-
('\ vi'
kY
VlX:
es
',7
v
<,
~" ~ ~I II
~:)., f".. r\ :"> ~
'<,J
<:0
~." ~
\.
<,(;
,
........ ~LJt r1.-\K /\
fA;. ,
.1: 0
1/..l .
lV\'rI
,
I
.l: s
J ',
IIJ11/
P- LA; /
r.
X
r7
. . /-'
r-
Vi. \.
"1'\
,I
~~ k-
r-, ,
.(!>
~
t'
II
.k" ",n
~~\ ~:l -\ {\
... ...
o
...
.t
I,.,
1/
I-l ~ /0 l>' ~xI: K<~ IS(; I~
I
I"
[0)1) :~ t1*. 1& (\ 'I ,v ¥v ~ V
11\)(
.'E"'" ,,0(
l'~/ IL 1///,:10
I " l l . II
'
0
-II ';
V
./',,,
1\
/\
V
.
'A,J
II'\,.
U'
•
":"-;b-'j .A
_. -''d+INItr-
~
1i..L )(I~""" rv I~
(',
,..
~'~J ,L.,l,l/
"'-;.L
0
'"
o
'"
Y,
~ ,
1 /I
0(:,
V
Ir
I"\li .t lli Ii
}fr,;_->t 7 . .~A
::f
'"><
..L /
'A
.
-J
.
.1:.
~ 1\'y. ~ \t~
V 1)\
.
r I / IX 1/\'1\\ X II\FX
0
Z 0
..
~.A~. I¥--' -
t'iloTl.1
V
--'
L-
Cl
1/
. . . . . .
0000000000
J1hI~
~!i\".. AI1\, I". ) \0y\ f\of\ '! _ x)~-[7:".\'1 20
. .
trlT
I.s 1.3
I., 0,93 0:81 0,70 0,61
'!'
~~
, _ ,':' P/ r u
=.u.t;J!3t=N~ 0,53
~ L
.
V/
,002,003
.005,
..
r ........ /'
t"'"
,01
' .02
,03,.
.cs
,07,1
' .2,': .,3
LOAD DURA TlON TO FUNDAMENTAL :PERIOD,
Figure 3-159
,
,5.7
I.
.2
3
TIT N
Maximum response of elasto-plastic, one-degree_of-freedom system for bilinear-triangular pulse (C1 = 0.237, C2 = 100.)
s
7
10
...z
'-.'
E
c. . .
c
a a:: w Q.
--'
...z« w
::t:
« Cl z
'"a ...z
L.-
a
t:
u
~
LW
C
'::t:
'"
::t: ;(
«
::t:
L.-
o
w ::t: ;:::
tclT
ru.. [ or YI[LD TO LOAD DURAnON, ~,."",.. \l" t-l 0
rfJ '0
c1"V',""v"~t-.l..-'oJ' .r '", '41' ',.g '"" .~ ....
_~~~~O~~~~?
.... ·lJ)-.ltJI ................ N N ...... OOtJ'ltJ'l"-lOU\OCFIv..
- ~ 17\ii\-..lY '+' "111\1J
..
__ __ 1/
\\ M"\
°00
o
o
UI
- '" '" 0
o
U'
0 ... 0
_ _ _ 11~. tl....-1 rr
'0 "'" t""
0 t""
0 N 0
tJ'I
1.1 ~
'"
Z I-
<,
...E
7
Cl
o
'l\'[\
'"w ll.
z
o
fP¥
I-
U
u
W
..J L_ LI
o
...,, ,.... '" 0>
oJ' ',r
Z
~m~~~~1'\
w :I'
'"
Cl
z
::>
....
..J
W
o
I<
>-
o
'"
I-
, hi. 'Y1lo,V
o
..J
I-
Z
I-
o
Z
i=
t: u
W
o
o
W
l---I
..J
'-
Cb1
W
....W..J
I~
o
:I' ::> :I'
o
" " ".'" ::>
><
x
'"
:I'
..n,~~~ IA1 I j..---
.002 00.5
_005.
.01
w :1'.
i=
-r----il I ~.....-I'"/f""""'" .v .. 020.5
.05 .07
_I
.2
.. 5 . 5 . z
LOAD DURA nON TO rUNDAMCNTAl P[RIOD,
Fi gure 3-160
.... o
I
2
3'
TIT N
Maximum. response of e l as to-p las t i c , one-degree-of-freedom system for bilinear-triangular pulse ICt = 0.215. C2 = 100.)
5
7
10
e
e .'
TIME Or ~IELD TO LOAD DURA TlON, trtT
6',,~ '0
<,
E x z
0
t:
u
w -' w
w
20
'"
I-
..J
0
Z
0 t:
u
10
7
W
5
a
::E :J ::E
>< -c
::E
....
-
~
"",.., 'ff' '!J)
V'
3
l\V\~ ,-.., --L
V
I..\.
J
cP
...
-
0
0
'\. " 1\
0
V'
~
0
Cf\
\It
I '
I" I "b( hi:
~
I"
0 ,0 .... 0.... •0 '" .... II' ,0
0
'
~n iLl.
1/ vf d
-1!
1!::'1 ~ r\ j.,D
"-",
V
"
II' 0
lJ'l
-: I--'
1/ \1
I"
a
0
Il'"
.
,,1 I'
UI til ......
r
IJ
-Q
~
• 4 Y..d..::, LC¥-: l..'>
''7 D.L" .. X
~~;~t!J. / "" '
~li~' ~ /',K ~~~~~K
I0P
? " . ""
..-'
~
Fi gure 3-161
i-"1J
f"
~ It\
e:::.
,•
L ""
zi
~ 1\" .z~ ~.t W d
-z
J:,s .J.
o
<.,]
'''' "'" ~ 1~~~\i4m IQi ~'"
41'.
~ ~lll ~ ~ .~~ ~~ 0; ~ <::: ESP t':5:).c-'P Ir ~ ~ ~ , '" u \ .,.; A
.a .'L >" r
;(LI
="
'" '" .
u
1.7
/.5 I.J 1.1 0.93 0.81 0.70 0.61 0.~3
.L.
G
:t=
V
])
. : : :~ : t: ,;', . L t::. ,02
,OJ
,05.07.\
.2
'.J
.5
'.7
\
2
3
LOAD DURATION TO rU/
s
..J
0: 0
<'5
1<11-'"
~'"
4 Y
a: w
v•IS'
'"
..A --\.. >..1-
n.-c , -Y I V"L.
a
o
f-
.,
'"
'-.E
~
« zw
'J'
C
Il"-',
1-
Z I-
a,
.t
" ,
. Lt "' "'" .,~
,01
."
6'
~,
0,7
1/.,4 ,4 ~ J:LL 0,5 ,I
.-'
<,0
:.a. ",.
.;
t::.::::::
!'l >l<' ~ ~"
Il'
•
'''" .
.
"
r
Ll.I
'''' .I K. > \ -,
A
2
..- ....... ~?~~~?Pp?~?oo 0 .• '!J) ~ UI"~lr4NN-- _ OJ(fl
'~ -i' --X K\' ~ .Jl .is: ."" L.¥- • ".f V "r 1M' ,Ll.
~-"
-i \ LY
W
..J u,
'vi'
.Y'~.. - ~r ~ 1'.. Ii' • u: ~LI" ,l' jI'j ~ r < r: r~ ok: X l J...\ >l../1'.'- 110 IJ<,I ~" " C ~"J" V'1..::I' . "
N
;::
='X '¥.
V..
30
....
iJ'''
,~- U [\A t7f. L u IL..>'~ ~ fA ~ ~<)" .Jh1' L\Ll~~ 100 JL.A .'x' -w·\ 70 4 rJ1 50
a a
,
'\I'
~ ~ ~~, " " '" '1'>.
u,
w
d'
/ \ 'V\ \ " IY\ -'L "' ) "."
20 w X
'0
c
50'
30
fP
•
.:.
7
10.
::E
-c
a
:J
'--
ofZ
o c:
u w
..J
u,
,W
a -- ::E
::> ::E
~ .... ::f
o
w ::E ;::
, TIME pr YIELD TO LOAD DURATION:
.
6
~~.
'0
dI
o
'!'
If
V'
v"
~
~
"t
~
a'
"
-:"
~
oJ)
~
v4
tclT ,
0000000000
":" {,::,.~ :..:a i.l' .... ~
~
0
~
;". N N :.... :...
U'lI''''-lOVl
0
(J\lJot
o ·0 .
.0
-o
'" _VI
DO 0
o
U'
o. ....
00
o
'0
'0
....11'0 ....
l,r4
....
500
w
3 0 0 f\'I)¥-',I=I',,",'l9t-+->.!+.loI!= 200
....z
<,
~
w
><
<,
E ><
.J.Io'--f'hliI~~III'
z
o
,,~
w
..., '" o '"
....w-'
. >",'
Q
••
Q
....o z o
;::c u w
~I
10
7
1\,
~
-- -_. - - - ----
~v
,V\'\V\,;'J
._-\ /
~
."
'" -~~
o
'" W
0..
...J
'..r
« J-
0:",
3::f «
C z ::> •
M7l<
u,
oJz·
~ v"
o
~
u
j
1/\
-'
w
"Il..U-Jl">LIl'.11I1 - -
....
I\1\1\ lI\ \ I'xI \ !A'
5
u, w
~
r-,
u, Q
I
'7
IlL
E
Q
cl'
oS '11 .
~~Wlil¥t~~~cA:J.J:lII-H'II~
...J
w
>-
/
17
;::c u
"
Q
::f
::>
::f
::f )(
::> ::f
-c
>< «
::f
.....o
::f
w ::f ~
~
IS
t::
;....-""':'.
0\
·.02
.03
v./x.../.V/v
.05 .07.\
.2.J.5.7
LOAP DURA nON TO rUNDAMENTAL PERIOD,
Figure 3-162
I
2
. J
TIT N
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (el = 0.147. C2 = 100.)
5
7
\0
e
e
•
• TlM[ or YI[LD' TO LOAD DURATlON. ~
~ p
"" ~
~
0
·fP
"0
(f'
'\I'
'"
.V'
~
'po:
~ to' '19 'U'
v''0'
---
,.
,..o
...
'Ol'
:.,.I.D
...
o
....
I~
'ciT
0 ..... ." ' '. ' ' ' \ I0.oat. U0.l o0.N N . 0 ---
0
0
O"U'
0' Ul
a.
~
01.1'
0
~
..,\ I Y ___
oz c:
... ·!PI
.~ 1V\ 70
-~-+V'tI-fJI
w
-'
W
....
co '"
o
'"
'0 ".. l/'
"0
'0
0
'V' t-J N OONl/'O
~
....z
<,
....E
~
a
o w '" a..
.S"
r-AJ 1\1 I \ 1'1. _.",\ I
L
w ,
lMlV
o o
0
::3jz\.r-,Xr-:
u
N N
I)<
,f
/1tft\/
lI4
~
- -
'.l\\
Q\
a a
..,
~
~
..J
~-
~
o
I(\~
f-
Z
o
"-._~_~
r-
l/
• •_ _
. ~ . _ ~ ~
_
~"
•••
',J'
nn
..J
w
~
o
z
r--
:::> L
....o
\7
~~
~'tt1 \ 'iJIl
z
o
"'
i=
u
fU W
W
..J
L' W
1'-
..J
L W
a
L
a
~
:::> ~
><
:£ :::> ~
K
~
L
o w
:£
i=
l--'"
0.7
0.5 .001
·'......-'1...--.11
.002.003 .005
01
~
.02 .03
p
.05.07
'//.VA// .1.
.2.3
LOAD DURAnON TO rUNDAM[NTALP[R.IOD.
Figure 3-163
.5.7
I
2
3
TIT N
Maximum response of elasto-plastic; one-degree-of-freedom system for bilinear-triangular pulse (Cl ; 0.121. C2 ; 100.)
'5
7
to
TIME.or YIELO TO LOAD DURA nON, ~
:..1
""
.....
t.J
..
0
rP '0
t1'd'~v"~~ 'oJ'
'v" '.,.
'0'
'v'
'.p
e-cP
tciT
00000'00000
':'"-:'"·.,b:...&"" ~ ... 0 0 U' ~
0
NN:-:V\
0
0\
~
°00'0'0 -
O'
(II:)
0
01
lJ\
~
U'I IJI
U"
"0
'0 '0 '0 vovtoy-,)N lJ' 0 lJ' 0
d'
',.,
""i:
1\
·~
-'
'"
"
<,
F >(
.,.,.
6'
t:
70 I'{
\I / Kf
\I~
o
w
,f
~t\%I,1\iTh 1/
"oj
1/ 1.A-r V .r: \ It-
/V'-.. ':6'
LW
..,..,
a a.J
30
,..
20
W
E
a 0
"" W
1~~"'N~8*~
':0
.J. s
.t. o
0 f-
<'5
Z
-l-I-A~,lLJ.l.~.""-.A.'-'<-.&...L..l""""-'A.,
0
;::: o
"
w
<.J
<·0
\
..J
L-
/.7 /.5 1.3 1.1 0.93 0.81 0.70 0.61 0.53
~~ ~l.L\.r~-t:tf\Pt~tf:\4--lr-/X
W
a
"
I"l.-H1I'"k--r
R
:> ::f
>< ..: ::f
UI
..>t
~
::j~ .001
.002.003 .005
--P .. 01
'1---'
.02' .03
~ .05 .07 .. 1
.2.3'.5.7
1
2
3
LOAD DURATION TO rUNDAMENTAL PERIOD. T/T N
figure j-164
..J
.« l-
..J
..,"",
.... 0..
z
0
z
l-
<,
Maximum response of elasto-plastic, one-deqree-or-fraedon system for bilinear-triangular pulse (CI = 0.100, C2 = 100.)
5
7
10'
Z w
~
-c
a ·z ::>
... a z a
I-
I-
u
w
..J
L...
w
a
~
::> ~
x
«
....a
~
w ::t: .~
11
•
e
e
TiME or YIELD TO l.OAO DURAnON. """ ... ~ 0
500
J/\
200
><
<,
F >< 100 z
0
t: u
70
W
'",
'"
'" '"
n ...J W
>0
0 rU
w ...J r.,
::J
"
><
:>
0'11
7 L
J
5
~
v'
.;0
~
~~OO~?OO
~o·o
-0
lJI"l,rILANN-\J\ lJ'I ;.I 0 tJI 0 (f\...,.
0
~"\I)..J
":'
OJ
..
~
0
..
\-{\J{'\N.
\.1.1 y.l.Ii.~ Il-\nJ1IVT~
i~lX.
'. ' .. 1\11 - ~\
0
'0 !J'
CJ'I
0
U"
~11\
"IJ'lVlot-.> O!J'
0
\I'
0 N
• 0
h- \( ~
bf111 X"1Vl7
d'
nJ1\1\~
',,>
'
~
.-
1\/r~lY\
~~~~.
''\\-
._
/\
H+
I-~ ~\'1iLl ~
., ,.
.
' A IA:...I\"'/Ut-\"'Tl\ LH'"T7 rr
~,.
v
'Ci'
t: o
"
J. s ~.(i
1'-
~'J
.
2·0
L"
.-I
-
IV'
~~J
I
~~~0~
'-
V//lI"""~~
O,51~~~/ .--4'.L''-'''-l-'<'/+-+'''I--I-hK''--'---f-''''I--IL-l--;'''' .001
.1. 0
... ;
I~'
.002.005
.005
.01
.02
.05
.05.07.1
//.Y/....r/,/V .2.5.5.7
I
2
5
Maximum response of elasto-plastic, one-degree-of-freedom system for bilinear·triangular pulse (Cl = 0.075, C2 = 100.) !"
<
1.7 1:5 1.5 1.1 0.95 0.81 0.70 0.61
~~ .. 0.55 vO'Ciii
lOAD DURATiON TO fUNDAMENTAL PERIOD, T/T N
Figure 3-165
'" ...J
f-
tJVf~n~~~
....~X1~iJ' '.. ,_ . .~. '''-
~I~ _'. . . ,
0 W
'J'
" WI....~",,~.
r::ll::=lf-l+
~} r\
, . c . ~I:'i'
--
.7
IV
,~*~j(\fj,
...E 0-
-r '~R"7'\'r\~U-;:
'".:--
z
r<;
0
6' ',,>
J'
2~~~ 0.7
'0'0'0'0
1\
_..-~-.AlHW' - ~ rx'l " 11 -J ~i\. .LX "'I ,\ :\\... /7. rt . ~/ 1)1.. X
Il.lf;t; ":"~I\
co
(:;t ~ 1\ 11\ _f\. rx~'. ~.~ ~L.~--1I1;H.W\l.Jf--"\/L\-M, te, Ll fv\ IAI') i-v ~)< lY!. IVf n
1 0 - r.
W
":>
~~ - . _ -
..1-.
... z
...
50 20
~
"'J\II~LYJ--I,i
,-\i··IY ~'-AI-/" . . 'V-'
'''''w 1/\ FT'j:s2 \
c,
-'ri"
.~"'~\~~ '.\14.'
50
W ...J
',6'
""
f¥~ ~I~)-~ \ 17\1" J\ >;Ji~W
-- :~ . -:vIIV
,..
v'
./ Y Iv' 1\ \; '\ lXA \I!~-
JOO""
w
o'V'
1\ 'I
\
(",!!v"~
rP
tdT
5
7
10
'2!
w
::E
4: 0
z => u, 0
fZ
0
r::
u w
...
...J
w
0
::E
=>
::E
'i(
::E
... 0
w
::E i=
TIME or YIELD TO LOAD DURA TlON, ",
o
v'
",r;JJt1'I.1' .. v-~ 0 ·0 'I./' '0..1' ' . . . t1' dJ
tdT
00000000000"0
t-l-!'~':'~'~'-l ",,'"~VtNN-:''' >...1 .. ~°U'o l I ' U ' I - I o t l l 0 CfIU"
~ 0
~ 0
'0 '0 IS' IS' IS'
'0'0'000 ... ~
~
"'"
N
v'o~o~O
IS'
~
w
....z
...E
6'
o
'"
><
<,
E ><
z 0
'r
70
t: u
or'of
'l:
\'11\
..J
1\1
W
"",
'" e'"
..J
>0
~'/I \1 mJ V ,
1-
"
Z
0
~V"
u, W
a
N M"b< _1'1' . JI' I
'~I
t: ..J
..J
-< .... z ,-< z
~
:>
Ii
, ",,;::l1\t-ttn~7Ii\:tr\"j
W
W
.... '"a,
a
a a
u
a
....::IE
W
'-
<,
...
o ....
x
X'","l ,.._
'-
z
NIf.t~\ P<
o ;:: u
.
7 VJ 1\-V'Itt*",-;J-'-t-t-J\'-~t¥,p"
~
'-
~
Q '"
::IE
:>
~
::IE
s
:::> :>'
X
::IE
-0: :f
.... o
.... ::IE
;::
-/lVfI~
.01
.02 .03
,05 .07.1
~ ;2.3.5.7
LOAD DURATION TO rUNDAMENTAL PERIOD,
Figure 3-166
I
2
3
TIT,.,
Maximum response of e lastc-p last tc , one-degree-of-freedom system for bilinear-triangular pulse (e1 = 0.056. C2 = 100.)
5
7'
10
• n'ME or ...
o-
... 0
OlCf'oJ'~vl' ·0
'\1'
'o- ....
'6'
•
YIE~D TO LOAD DURA nON, telT 00
~-:""~~~cp'~ ." -" ~ t.J 0 U' 0
t-l
'dJ
•
o\
00000
""'·... ""
U'
\,I.'-"N ""'" 0 lJl.
t-)
.
0000 ' " -
lJ'l'"
o
__
t-.l
0
'0 '0'0 ~ -.1 lJ'
'0 \J\
'0 ..
'0 '0 V" V"
'0 t-J
0 N
OOOO~IJ'OlJ'O
500 .100
d'
.,..,
Z t-
<,
oJ
....
...E
6'
Cl 0
'"
»c
<,
f
sc
100
z
70
0
.""
oS
;:: 0 ,~
Jk
-
5U
A"1-..)\J
v ~\
0
w
,
'" '" '"
0
30
-'
N"¥~:t1\,Ti\7f!1.ji
v
>-
xv
0
t-.
z 0
;:: '0
w
10
~
,I/t'-U
-'
KWD
~
W
0
::t:
~XIX"
« ::t:
.J. o
I-
~'6
.
~ .]
~~
>< 2
1.7 1·5 1.3 1.1
0.93
I:
0.81 0.70
'{';:l;~: :'p I ru = ~Zl J q'lJ
07·~'3-VYlY 17/f7Y7 0.5 .001
.002.003 .005
~I/ .01
'.02 .:.03
f;+ .05 .07
~ .1.
.
..2
.3
.71
2
3 ,. '5
T/TH Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (C1 = 0.04l. C2 = 100.)
LOAD DURATION TO FUNDAMENTAL PERIOD,
Fi gure 3-167
.5
7
<:
Cl
z
::>
IJl.-
'\/~110
3
:f
J.,s
2.0
71\15"
L.-
:::t ::>
m
... Z
":0
w
-' <:
I-
":6'
-' L.W
. ',r
'"l1.w
10
L.-
0
Z 0
;:: o w ..J u,
w
Cl
:f ::> :f ')(
<:
....
:f
'0
0.61
...
0.53
i=
:f
tdT,
TIME OF YIELD TO LOAD DURATION, ".
o
r;P
·0
er..,.. '!'v'"'!".. ~.
(1"'
~
fIJ
~
,,;
~ ~.1' -lo
..
~
l'
0
o.C? 'O!' \J'
~ 0
C? \J' U'
~
~
U'
'? '? '? '? '-" ...,J
""
0
N
\II
N
0
0 0 - -
(J\'"
'? '? -
to)
'0'0'0
-
QlI - l
0
0
0
l7'
0
'0 \f"
o.
'0 ...
t.J
'0 '0 '0 "'" ~.t-.J I,J' 0 \1'
0 N
0
Z I-
-"
E
o
Cl'
',.>
-s
a
';,r
.
~
u
,
'"
'" '"
'"w Q.
z
w
a
W -0 L.W
'6'
o
b"I-\4tf+'od"
o
..J
«
I-
z.
w
::r
«
/
a
z
:::>
-0
L.-
W
a
>-
I-
a.
z
z
o ~ u
u
c! w
I-
.-a
w
W -0 L.-
a
r-,
W
::r
o
:::>
::r :::> ::r ;< -a; :r
::r
s ::r
L.-
a
w
::r ;:::
. nU 0.7 r....4iLY~
0,5'~ ,00'
,002,003 .005
Figure 3-168
~ ,01
,02 ',03
'/#/A'// ,05 ,07,'
.2
,3
,5
,7
LOAD DURATION TO FUNDAMENTAL PERIOD, T/T N
1
2
3
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (C1 = O.03!. C2 = 100.)
5
7
'0
e
•
• '.
.
TIME or YIELD TO LOAD DURATlON, Ie/T
a
'0
dJ
CJ' ~
..
v"
'0 "I.$' .•v"'.... 'Cf'
t-l
'fP
~~".-",.oo ..., .~ .....~ '0 '(j! '..J, os- 0
t? lI" \II
~ ..
VI
~
~
lr'-
'"" 0
-l
C?
~
l/"
~.
N' 0
t? t? '? '? -
....
d'\ ~
.-1'
'0
~
'"
t-J
0
'0
'0
~
CJ'
\I'
0
'0
~
V"
V"
V' .. IJ'
200
sc F.
<,
0
0
Z f-
<,
~~$!J±t@~B~~=r~.Ml'\fJl:''''"
sc
z
lJ'I
I Y1\
, \If \ IX'\;{J\
lA=tI~~
0 t..I 0
~
V"
~H~M\itl~ffii\;:::ltrr~lJJ':ru: I. w
'0
'0'0
',,>
s·
70
'of
;:: W
.J
:::f <('
W
...
w '" o,
W
.J L.-
..,..,,
Q
o
fZ
u
W
...E
'"
Cl
Z
Cl .J
::> L.-
W
o
>-
~
0 f-
Z
.._
0
-
~
I JI
I
fZ
IT'
o
.
t:
u -
fU
W
W W
0
I .,
:::f
::>
.J L.-
»<
~
.J L.-.
W
v
Q
:::f
•.
::>
:::f
X
:::f
><
:::f
L.-
0 .. _
n
w'.
J:id\
:::f
;::
./V./
0.5 .001
/
/-17'1 Y .002.003 .005
/
./ .01.'
.02
v
.03 . . 05.:07.1
'/~// .
'.2
.3
,5.7
I
.';
2
3~
LOAD GU!lA nON TO rUHDA!'IEN! ~L PERIOD, TIT N
Figure 3-169
~Maximum re~ponse of'elasto-plastic, one-degree-of-freedom system for bilinear-triangular pulse (Cl = 0.026. C2 = 100.) ,
,
5
7
10
.'
\
!lrLO
lHoil,:O[
., .
r:P
0' V' '" v' t-' '0 ":101' 'u""" '0' '($I
~
/
,~.
.',
,
~.
0 ':"
.-;
_\ '"'t->
0
":"
"r;j)
0
lJ'
'It0\ 7t.. ' ~'1) \ . \ \ 1\' rs:?r.... JOO ~i'1\ \;~\j"IX / ,'X\-¥'l§';n' ,'w ·.1\, ,-XI\-i\" ~f~ . • ~ X\JA,
0
0
\.1',
0
0
do
t.,. i-J ''r-J
lJl -I
....
~
0
...;..
0
a\
0
...
'?
...a
~
7
~OO .'0;' \~ X\ ·'il\'.\
,
0
O_.ll'
/I,
1/ ' ' ' ' ' ' I
'.'
triT
10 .LOAOPURA nON, ,0 .~
•....1
.,
!
'0
."
.; .
'0
'0
'0
0'.,.."" ... o· ~~ u:' _ t.1'
'0 .1,14
'0 "0 V' f'..I
.t.1' ,0
0 N
.....
0
,v~, ." d'
,..,
200 w
><:
<,
F.
:-< z a .;= u
W
100
so
,,,J
n
JO'l'\.
1-
L-
a
t: u
w
-.J
,u,.... 0
>: :1 >: sc
-c :>
'J'
.
X'·X~" . --}1
'0'
!\/,",~~~ .. ~"
':0
~.\tIY\'t\il'-· ~U~Y1'lJ~--t*l-\
.J
a
'.
V ~ :;~~ Et:\\J."~ .. ~ .~. IL'{,~"'Y\;~> / ,, ~
.,j
:.
. . ,
'''f'''{ A
.J
C)
0'.,.., /:'~'\ IV IL }. \ 1) -;""(\~ ,'\,(::V't'\lIt--n:r~LX:l::t;d"E1t:'l1L1AAL::n::ttrt:lf\t\t:tlutttN1 \.. 1\ \. ,.1' - '.,~, '~. .. \f'I..... . . . hJl .r
j\
u
10 I
I.,
...., '" 00 '"
"
--' '-"
20
J
IO~Vc~¥...\~,~r~~-fV\i,A--v~.-~-~.~~.~~ . ~~~- , .~. ", . ..} ......-1\. '-V",··rt·,.l: ' -if'-\1 L\,l,~ '7'_
~
1A·~
: ~\J~\~A~
.1:- .......... _ >1, -'
\~t-7M:'~ Y
l\
I'"
I
Y
VI
•
;"' }''''I..\ VI-...vI
1'1.1." ........
1
ltv
,
~
'I
\l""
1'\.\<1\'\
2
'''/
IX 1\
~:>
~
.~ - '~r--7 ~ 17 17" ,7'1-01-'
,-,-
0.1
o.s
1/-:
.001
/"
V
7721£: V 1/ 7-'
.002.003 .005
V
-X- Xl""
11E..\ X '
'1/
I/' y\ If\
j<
~, ~
. ~l<:~~~~~ i-L-,..... .
X P\7k:[) "
'
.
p
~~ p/r" = o...-w.rlloH
'
./ /' L .01.02
.05
V /A77.YV Vl/ .05,07.1
.2.5.5.7
2
J
TIT N
Maximum response of elasto-plastic, one-deglree-of-freedom system for bilinear-triangular pulse (Cl = 0.018. C2 = 100.)
5
1.7
1.5 I.J 1.1
, - 0.93 0.81 0.70 0.61 0.53
~~ ~ .~ '=' ' \ - - . \I~\I \
-/17~ V
LOAD DURA nON TO rUNDAMENT AL PERIOD.
Figure 3-170
2.0
v
'-I
; ~~~~~~~ ~\ 0~~r<;~: \i \ -'x~v~r v~~:: ~ 7 '" ~~V'lr~~ t¥,ot.. '::s;' ~ .. -1'\ 1\
<.(; <.J
7
~~~\~-X-v '~""""'''T,n,,,
s
"-0
.. r
7
10
....:z
<,
E
a
a
0: w
a.
:....J
....z...-
..
w
::E
a
;z
::> "-
0 ....
z a
....o
w
....-' ·w a
::E
::> ::E
><
...-
::E "0
w
::E
;=
-
•
.:
TIME or YIELO TO LOAO OURA TlON, dJ '0
d'..r'0·0
'"
'6'
,..,'dJ
t-.I
."
-="-l. --: ":' -('
0000000 "(jJ '-l. , "l1' .... 'v' 'v ~ 0 "'+.1' ..... 0\1'
Pt..10",
•
, tdT
c:' C?' 't? C? '?
'0
~~""~'O.~
o
'0
'" ..,. ..,.
oj'
'0 ~ oj'
'0 '0 u' V' \f' 0
'0 r.)
lJI
0 N
O.
tv
',..,
-'
c, X
F
<, ><
z
o
;:=
u
w
-'
u,
,
'" '" '"
-\1'4\/1'I7CIl 1\
Q
..
70~ 50
Fvl\\'
1\
'F\f---\-f
., M-h-~tA-J~'* 30~~~JnU1
w
20
-'
~6'
":0
.J'.s
1
..t. o <'6
z o
<.J
f=
W
7
W
5
2.0 1.7
..J u,
~
OJ ::t:
x
::t:
2
1.5
"M~~
~~~_n p/r;=·~
0.7
0.5 .00.
AIYII~:Z .002.003 .005
.01
.02
.03
r
E
a 0
'"c,w -'
.... Z
w
::t:
.05.07. . 1
.2.3.5.7
"I
,2
3
" 7 10' 5'"
TIT N
Maximum response of elasto~plastlc. one-degree-of-freedom system for bilinear-triangular pulse (Cl = 0.013. C2 = 100.)
a
.z
OJ
"0 f-
·z 0
t: u w -J
b a
1.3
::t: OJ
I.'
X
0.93 0.81 0.70 0.61 0.53
/..1//.1"//1/
LOAO DURATioN TO fUNDAMENTAL PERIOD,
Figure 3-171
....o
a
.
~.r
k 7 \17 k71 \
>-
u
<,
'"."
y
.\ I
a a
w w
'/
100 ~ . _ -0 '
....z
dJ
::t:
::t: "0
w
::t:
f=
TIME Of YIELD TO LOAD uURATlON,
~~ 'f'v" !~ "ta-' ~c9 ':""".. . '-, '''''to> -~
500
~
"
1/\1
".,;
o
"..
0
...,j,
'0 ~. 0
I\VI\IIV,,\
telT
00000000 '\I' '10" , '"" '""" 't-J ~ ..... \l' \l' ...1 0 \II 0 (1\ '"'"
>
~\I ~ /
Il'W-l\I
'?
,;
o
'0·0'0'0 fJJ - l G' t1'
0
a
0
·0
'0 ...
0
V'
o
0
'0
'0
.... U'
....
0
~--_
.. Z l--
<,
w
\
~
J
\00
...E
~
1\
I~\ ycnA\.
.x.v
a
1\:
.s-
1\
',>
o
-,
;::::. L)
oS'
\ V\
',J-
/\1/
... \! \ /
w
-'
L... W
I~~~~V\
a a
-' w
.y
.~
1/
K
o
l-Z
o
~
z
'6'
w
~o
a
J:,t
::>
.J. o
I-
~'fi
o
~.,J
;:::: L)
-' L.. W
a
~
:::> ~
><
-c
~
7
2.0
-
1"-.
:E
,
.Y ~ '\
k1 X--
t.r
"'i
1.5
1/
1.3
""-L' \ \
2
1.1
,5
;:::: L) w
r:!
w a
:::£
::> ::f
X
«
0.70
o
0 ,6 1 0.53
w
\ .
,7'\
I
I
I
2
3
LOAD DURA nON TO fUNDAMENTAL PERIOD, TIT Ii
Figure 3-172
Z
).?I~
VI..ff1 f(%jW4/JV I I I IIII ',2·.3
o
0,81
~:-' p/r u = -,,',~
,I
z
L..
~ : :-- ~l-lT '<"'\ \:
+-:;/,-+T-i-7'lH""I-
::f
«
::f
I><
... , .y A " O.5 -:~'---'-""-+"--+---+£..j-,I"-I--l-+'if-L.L.--+~'l-4.001 .002.003 .005 .01 .02 .03 .05 .07
-' «
l--
Iv·
20
>-
w
'"
w
Q.
z
..,"", ""o
a
Maximum response of elasto-plastic, one-degree-of-freedom system for bilinear-triangular pulse (C1 = 0.010, C2 = 100.)
5
7
10
L..
::f
;::::
e
•
•
TIME or YIELD TO LOAD DURA nON, i,. tiT '.~'
':.,
.,.o ,., ,., ~,
~
0
.....
~
....
f$J
o · l..J
t-J
,(1\
0'
"!' V".
~ N
\.1<1
'.,
.
~
!"'I
....t
.....
~
~
~
~
~ .0
VI.
o
0
?OO??O? ?oo 00
00
o.
0
"'l.NVINN-\II ' " 0 VI _0 VI
0'"
m
CD
~
1/1'\ Ul'..L~+7I-~It-*""'+-
~I~
... ,,~ .::JT----' V
I
~
~
'"
.. - ... I
U
-'
L W
co
~
lR
o
'" .....
",,,,
0
r-
M-=
Z
I-
<,
E
...>
vt+:JVI
o
o <>:
w
'0
t-
'z W :>: « o
~
I·m
W
:/
,.o
«
~
-'
>-
..J
1);-
I~
W
,
0
Q.
r-.
co
\II"'"
"LI~1/ ~-JP-
'0:::1;
."o
-OOO"l 0 0 \II
en
'Z
:> L
o
,f-
Z
o
~
"o
i=
,W W ..J
I
s-,
U W
.-' L
L W
:0
~
W
o :>:
:>: :>
:>:
J
:>:
'x «
-t
><
:>:
.<[
'L
:>:
?
M 0.7'1/
~
o
~ II .l:.\
\I \ 1\
P Ir
'
=
w :>:
.>.
i=
_
O. ~ -:..L-'--I--J--I~-J-++-++_....L-l;,.--I---+--+-I-4.J.+-""'--+--I---+-J--+-H-++:--...L.-J--+--J--I-+-I-I-H .001
.002.003 .005
.01
.02 .03 ',05 ,07
..1"
.2
,3
,5
.7'
I
.. 2
3
L9.ADDURA no~ TO rUN~AME~T,~,L PERIOD, TIT N
Figure 3-173
'Maximum response of elasto~plastic. one-degree-of-freedom system for bilinear-triangular pulse (C1 = 0.909. C2 = 300.)
5
7 • 10
/
TIME or YI[lD TO lOAD DURA TlON. L ciT u'
o
~ ..
rJ
.....
_ ........
-' fSJ 0",..>
0=-.1.~
V'
(]'I
~
lJoI.
O l l J < f ' - I ·•
~
!'J
....&
...
o
':""
~
':"" .
""'lI'IO~
~
~OO~~ 0 0
000
0000
o
"Uov.t,,)N l/I"OVIO
-000'1 0 0 l.Jl
Ul'" 00
en
-CJ\v.
--~'"
~
11-.IA 1\
• I il
~~
--~
z·
o
70~~~g
u
50
t:
w --' ,~
'" N '" I
N
ci
~
~rtI4lll/
--'
a
o
'"w Q.
oJ
«
~\
>w
z
-c
a
z
.•.
::>' u,
''''>.....
~
z o ..... w
--'
t~
o
i= u
3
2
0.1
05
ffi?"tt . 1~
.001
I J J .002.003 .005
1\
l~
~ ,t
II
,01
I
J .02
I .03
I
,t
J
,t
.05 .. 07
~ J II .1.
I
I I .2'.3
1\ p/r =
I \
I, 1 I I I J I .5.7 1
lOAD DURATl9N TO rUNDA~M[NTAl P[RIOD:
Figure 3-174
::l'
::>
Ii 1
Vi
'~.~
"w a
I
~
x
Of
y
~~
a
«
o >z ~
7 S
W
Of ::> Of
E
Of
'0
u
z
>-
<, ..J
~1/
Z.
\~.: ~f~\--'
I.' a
\l\Vil'-,
k:-'-I.~f-+7'!-<:""
~
JI!-''l'+-\-If
\.olN 0"'.
\
I
I 2
I 3
I J 1 J I: I 5 7 .. 10
TITN Maximum'response of elasto-plastic;one-degree-of-freedom system for bilinear-triangular pulse (Cl = 0.866. C2 = 300.)
::l'
~
::l'
"-
o
w ::l' >:=.
e
• TIME
~
,..,
..
,..,
0
6, ......
~
~ 0
\oJ
or Y1E'i.D tIJ
TO LOAD DURA TlON, L ciT "Vlt'-JN
(J\IJ'
~
U4,.N
"CJ'
•
...
. \1 i'l J'\-I .I."l \XI I VI M -~ I l-li(' ~~X\
:...
O
:....:"'"
~ 0 00 o
O
~ 0
?~~!=>???OOOOO "'""",c...NN
--
-
OJ
fJI ...... OVlO
atlH
0
0
NX N,.-INI
.1>'. . _1/1 N 1ft( 11><
NII~II/
Vi'll rNlI1NV NIA'IJ $"V/I-
~ K l---IJ/
I n I\rrl(j.' Y,
-:;\"Vi V1N 1\/1\\ ~ W'NV "-JA\! WJ\J K I ,f)(r-IJA ~ YI\\ LXI A [X , l'llHN \X~ '\1 NT'-
00
lI\ ... lHN 0 0 ' 0 ...
C7'l VI
lHJ I f'Y-i
t'lNU'td 1114'
Z 0-
<,
E .J
a,
-L
l>
o
1-7
e>: w Q.
IX! \hI!
!l
NJikl-I1 1 H':!"I'7f+"'\,f-+;>h:--j WLJI,vi!l fl+1 I)' •
" '"
~,1'1<1
'« 0z
W ~
-c
a
i!:
/rl~
w , w w
..J
m
'-"'-
1M: ~'
....=> o0-
~H1
,/
f-
U
'J -' r.. W
a
7 5
::>
:>:
,
><
«
:>:
a
~.
I--
=>
j----\JJ-H\~
2
~
~
.~
\
~
p Irl
=-n.:
:::E
I
><
«
~
.... o
w ~
i=
"%
07 o.~
'~'-i
.00 I
I I I I Hi
.002.003 .005
.01
I
i
I
I I I I III
.02
.03
.05.07.1
I. I.
1'1'1 I I III
.2.3.5.7
LOAD DURATION TO rUNDAMENTAL PERIOD,
Figure 3-175
u w
......J'
~
3
o
w
1'--.' /I\V V, LX--...\)\'vf\XlU'\N~J1?
:>'
Z
t:
T/T N
I
I
I
I
I I I I III
2
3
5 .7
Maximum response of elasto-pl~stic. one-deg~ee-of-freedom system for bilinear-triangular pulse (C1 = 0.825. C2 = 300.)
10
TIME OF YIELO TO LOAO OURA TlON, L
'"
~ ,-J~~"'~o~ .. 0
v'
o
50
20 .w X
o
<,
E
sc
z
a
s-:
u
W
--'
L. W
w ,
'" ..,. w
a a
--'
w
;:::
"\1 'I"v ' 'YO, ~:t ~ l/ \lL "- 1/\
30
.~. 70
100
50
20
0 tU W
10
W
5
:>: ::> ::f
3
c-c
::f
2
/
.,.,
'i.
111' LlI:'
ux Ci
b,Y
1/,...-
r-
~
""
I.
u
\)<
L
~?~?~????oooo
?co
-
i.No
o
_'
V
I'
II
r-
I' V
V:-'
. D< 12
~ [)< 1" IZ
l::lL j(
P'
I'-
I'
~
~«
r
k';
~
.\
1"'-0
.
I:::::
;0
"
1-" 1-1-
f'
r'llf-;
-t
r
I'
:::: .:::::
;'>
r
~_
!:6~
~~r- r.
;,
O~
v
..L P
"
~
i2
",N
'UI"'040VlOlJ'l
r-
V
00
"L04ViNN---OOCJIU'l" LHO OUl 0 0
(fI
o
I
~.<./
,)<
V
-
. r': Ie <::' ,/ 10<::
r::: r-
>-
P
t::. Y-
I<
V
~t:St;
I-':
w
.1·\ ...1-
I-
,>.
::f
tc
I1
:L..J
1':.\
P r
.'.\, .02
a
~. ~\
_
C'
.03
.05 --07.1
.2 ,.3
.5.7.1
2' 3
LOAO DURATION TO FUNDAMENTAL PERIOD. T/T N
Fi gure 3-176
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse ([1 = 0.787. [2 = 300.)
5
7
2.6
~
_01
z w
I-
l:"-
.002.0'03 .005
--' .::
'3.0
l-
1'/ 0.5 -r.001
Q.
•. 0 3.5
/. _
-
'"
'Z
2.J
i.
Cl
"''/,.1\.'>
2·0
E': I:::::
E ...J
a
....s
~r
Z
t-
<,
,;
r--
W
.
~
.l\I
1)~ { I '~f'
. r-
ill
-
D
'. \ )
V\~ \ r:-,~M 4):1< I ? , \ : : " I'AlL'(..2,< L ~~~t, .l::: ...l' ~~ r- 1:::1 H" ~~l'0jfj. .:::I: l...Ll U'~ . ~-" ~ /'
N
:.....:.~
""
'\\'
..t- Ll>L
~'
0.7
~
' .. I'
Jj
....
\1(
I
''''''
ffitM .~~r'\V
--'
a
'Ii
l
t-.J
\
\ '
OC \
rf ~ .x .A v \K \ IXJL "" f1i"-lX V "
7
L-
u
:4..
Ii' .so
a r-. z
~
VI' N
•
"01·....
JL IX n\he-!0 ~ ~ h
t1
\J'
(J\
ciT
10
::> L.-
a
z
i= u w
r:'w
1.6
Cl
I.J 1.1 0.93 0.81 0.70 0.6\ 0.53
::>
::f ::f
;(
-c
::f
L-
0
~
i=
e
e
• £IT
TlM£ or .YI£LD TO LOAD DURA nON, l ~
v' ".. 0
"t,
~--;.
~
~
rP
~l,oI.N
V'
CfI
N~
O'~·(J'l~N"':...a:..--..
500
.o
200 w
><
<,
o
E >< \00 z 0
r= u w
70 50
-'
L..
w
..., ... N
'"
a a
30
w
20
-'
;::
0
f-
Z
0
\0
f-
U W
7
W
5
-' L..
a
~
::>
~
3
><
2
0.7
A +
"
X-\ 8....' ./....' . "C":"
~+
~.
~1-,f .~ 1-4 p
0 +. :, 1"..'tJ(,
I~~,~V~ \"VI1 ~~~~
I
C
~~
~ '"
~ 14~~c ~~ '-'
-
L'I~. I.\C ¥ t' c 1'1', ~ nM •'
liS" '".r~'k M '" ~ '" '>" 1>+ t~V 1/11:;
,J &
a ~
H.y..
I) I)
~~'"
..
;\<::l;
rJIr
~
I
-LH 0
t
L2
DO 0
00
&I c.n ..
c.n
>'-
~ ~
I
0
;" ;0
I'!'" . II' l--
•. ,
",'
1(1.
"'''' 0"
0
"L
,
'"
=rl-,;\. ~ ~
. •
~.~ "" '>
l(
~"-'
en
. 'I".'~1Il'l'
,
f,.ob
. .
~. ~
,",u
'-'
--
1.10 1.1.
i? ~
<
\; \
'
V
N 0
""
~ ~
e<.
... Lot tJotN Ul .... QUI
0
d
~t::
;.....a:: u
:s:
pI
_~
.01
. . 02
.03.05 .07:\
:_
2
3
D:
w
,'.\
-'
0''1.
f-
.,.;
w
4. 0 4.0 3.5 3.0
Z
~
a
Z
::> L..
0
f-
z
0
t:
<.>
w ..J w
L..
/.6
a
1.3
::>
I. I 0.93 0.81 0.70 0.6\
)(
0.53
TITN Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (Cl = 0.750. C2 = 300.)
LOAD DURA TlON TO rUNDAM£NTAL P£RIOD,
Figure 3-177
'\
a
0
~
.
.2.3.5.7
·E :.,
a.
1..] 1,0
~
0.5 f0:::.. "'.00\ .002.003 .005
z
f-
<,
'b.t
1,6
'" -\ u
'" "
I,
"T ~
':\"
...
.....,\
..
.
"'...
-'~~ ~~
1:~m:r~tT1
r: '
7::;.
J;';"":" "-::
.,.
\.
~.
'I;
V'
Il'
~
T
~:;
~J<, .l,,-,),
.
o
~ "cfIlt
~,,, r 1'J"'v
L'
??????O?oooo
Vto(GCJ'l
. JI...II ~ .4 I- ' '"Il I 'J.-: ,... ~ 11 I
--'" -\' ' ,,,,,.... J 'j:)" IXil "
~ \A . ~" ~
300',Y
-: ~ ~
:""
5
7
\0
~
L..
0
w
~
r=
TI~E Of YIELD TO LOAD.DURATION. ldT v'
~'
t-J.,...l
o
..
0
-;.
~
••••
- .....
'"
'0
U'
Gl
't-J
...
~
"a>
li-l
N
. I'JI'U \'\ J\ \1 X\J VNII .I 0 0
1
'i;)r~
,
r \.
t'\
I~ I
r
I
\.~
~~ .~~ \I
'{V
HI I 'Ii II
E
.
w
-
--'
~
W
a
-
.-l
w
>-
a
....
llA'\.IlIF\.
", .....
Ir
'-I
~
f--
y
.<)
7 17'
l/
'V
J\-v1 \X iI\
L-
W
~
a
o6 ~ v.JJ 1/
OCJ'lVl
0
~
6
.'4b1tt
o
0
'" o N..
Z f--
<,
E .-l
J-;
~ v
1'- .~
-
0
a
~ r-
~II"-
a
a
'"w
~
<,
k::,
V
~ ./ vb-
r::YV
..
«
I¥
r-, V, V
/
./'i(vV·
k
--' f--
,... ..et
<:'
k,f-~
V
V
o,
i/""
--=
t
Ja
"/."
2
{,.7
.00.
n
,.-
.:::>
,.-
\
<; 'I.- "L-,
>< «
vm
1m'
~
~ ~
L-
a
w :::I
p/r =ll.,
>=
./
.002003 .005
.01
.02
.03
.05 .07.\
.2..3.5.7
\
2
3
LOAD. DURA TlON T9 rUNDAMENT AL PERIOD. TIT N
Fi gure 3-178
a
f--
>=
0.7 0.5
L-
'"w--'
i>
1\ v :
~
:::>
W
1\ 'I
::E
« a z
a
t.-
~
w
z
~TI7
t-
L-,--.
......,"::'
z
L-
II--"
. .1
Y:
II"
\<1.dr .J...-
,.-
,.-
? 00
N--
'?11 <
'Y
v
a
>< .«
-- - -
1--.11
z
,.-
IJ
h
"'VI~N
V\~OUl
VlfM
l\I
II l)([b(rY ~" \';ttl 1',
'0
f--
::>
"'III J'I I
--
:0:~
a
w --'
???~ ?~?
n,
70
L-
N
I
f\
)1
-
>= u
'", '" '"
N I
J\jJK\MIX J>(f>K.- 'K
<,
~ ~ O'l o o
-
J\ I 11.1 U\ I I'll I I
.... "m 'W \\l'\ \II
><
-
V\L
A . . ~ .
--+-VI \I III NIl I X '-I 1\
w
>< z a
l\
N" -
N·,,:...a:-:...a~003
Maximum ~esponse of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (Cl = 0.715. C2 = 30D.)
5
7
\0
..
e
-,
•
.,
,
TIME OF YIELD TO LOAD OURA TlON, L ~/T _ ~
•
500 300
-
100
z
70
t:
50
...J
... ...'".... I
W
a a
30
w
20
'---
f
f=
~
u
W ...J
-7.
w
5
u,
a
:>; :::> :>;
>< .;
.
0
~ ~
3
2
0.7
'\'---'
·o!
N
"'" _ ,N
(\
.
1
N
-
:..
:..,
[\ "
,\ J\" '
'
. .
,~
•
0
? ? ?
~
' 1'- V f\J'
I
J r-, " " "
}
!>~?
O?
"l.#llNNN-VI -.l 0 VI 0 (fl.lN
,
".~~~'
~~.1 ~ \.
~ k!:>
'\.
'" K '"
V' r-,
.v"
~
J\ )' ~ /',
~ ~~ •
~~
.tIl
0
0
./
'" z
~
0
'v "" r, "'he '
~h
7"
1(.
0
0
'
-J
~.
~
,'>
t-\; \\
"
~.'>
~
~~
'"
"'<-' . \
\
'"w
5. 3 4.6 4.0 .I. 5
"
'7-
.!.o ~'6 ~'.I
\
,,' -t
.0.
.02
.03
.05.07.1
.2.3.5.7
2
3
LOAD DURATION TO FUNDAMENTAL PERIOD, T/T N
Figure 3-179
....Z
w
:>;
a
:::>
"-
o
>-
z
o
f=
u.
w
.:w
1·6
a
1.,5
:::>
I. I
X
0,93 0.81 0.70 0.61 0.53
"
0.5 .002.003 .005
...J
::E
,/. =.,.
I/' I'OJY
0..
'Z
,U
\
a
1·\
v
b0c
.,.,E a
).0
~\:;, ....,
z
><,
'/>.1...1-
?:
.00 I
1-1> 1-1-
•v
/
/
o 0 '" o IV ..
'"
.; '
,,_
0
0
-OttUlUl.
/
.....
oj
? 0
V
,~" 'r " ,
IV \
0
i.-.~OoCn .00
1'..
",,,,, V\ 1\ i/' '\ ,
0 /
~
~)().
-tao
U'I
'k" I I" J
J:
%-- Y
:~.
(J\.
f\. IX ,r-;IJ i\
II""
_II:---.. <\,,,
(lit.
-r.. II
~
-
JI ,.:"
'l.
10
0
' '"'
D\7" .Ji , V'
>-
0
~ J ~
~
...
. II ¥-"1<
I
_...
'
lD'
0
z
0'
....
.s< : '
...J
>-
,-
L
sc
'" c-
"
l' OC
<,
u
~
)'\ ,/:sf f-J' , J)" or --.:U
II
, 200 1-1/
0
0
11 1'1\' ~ V\
w
>< E
,-.)
~
u'
Maximum response of elasto-plastic. one-degree-of-freedom syste~ for bilinear-triangular pulse (C1 = 0.681. C2 = 300.}
5
7
10
::E
::E
cO
w
::E
i=
fIME or YIELD TO lOAD DURA fION, L ~/T _
~
.~~ ~ ~
500 r-r-r-r-r-r-t-r-e-ecr-
... ~
'0 t'fr-J J C!'01
.'
~
If'
N
l.JoI.
Y" t"
...........
..J\YI~ I\J 1i.:'I/\ r, \/'~ 'd'~ \X I/W'...I
.. I
..... am:... 300 1~~=J,'l.-=-:I!-\,-f1'1:l
\ N
\
'Xmt1 \N/V\ 200 1+:~$~UW~W4~\-Ai>rA).J~~~
w
,)(
E
o
-
100 70
0
....,
~ ~.~
[) I)
~
r*
~
<,
z)(
.~-~
o
r;.
~~~~~~~~ 11\....... 11\..............
W
o
'
\)<
\~1I
o-J W
>-
~
ntm~
o l--
A'" ,,,,,,,n,v
Z
o l--
o
~
...x ~
00 0 VI ..... 1I4 00 0
000l 0 VI
00 NN ....,0
r-r:tt:tlht:J;m;oo ~ \'\'> " ,~Qdf; M 1-1-
:;r:~~~lltm
'"
'I' >
~ '
"-
.
'
~~
B'k Ii Q-
E J
o
t> .~ 1·\
o
&.1
Q.
5.3
4.6
<>: w
... -J
l--
Z
w
...
:::f
P Z ::>
L-
o
l--
Z
v
o
i= u W
L-
w
'G
5
I--
3
rt
2
~
v
v
r-
~
UYkl~
~
~ p/r
"\..
:ll.
0.5~1·'!111/ I
.002.003 .005
I
I
I
.01
! II .02
IIII1III .03
.05 .07.1
'I
111/1111 . 2 . 3 . 5 '.7
'I 2
1111/111 3. 5 7 10
lOAD DURA TiON TO rUNDAMENTAl PERIOD, TIT N
Fi gure 3-180
o
:::f
::> ~
>< ...
:::f L-
o
W ~
;::
0.71~
.001
Z l--
<,
\\
~/
:2C
::>
0
-J
7
L-
W
-O'l ~
.
~
-J C.
W -J
.... """f",N1V N 1II"'-J oU",o
.'
w
U
o
!b;
u
'", co '"
O'l
MIWIlI'
i=
N
~~?~?????oo
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (Cl = 0.648. C2 = 300.)
e
•
.
~
•
nME or YIELD TO LO"O OUR" nON, t[/T o0
,., ,., p
0
...
<:>
~
'"...
'0
~
~
a'
.~
1<\\
..
-c
<,
-VI
-_ ... "
>(
.
0
0
0
0
0
co
t...NJA~\N
.
.
0 0000 . .0 .. . 0
...
~u..Vl.NN--
V\ ' " 0
VI
0
(J\
VI
o'
.
00
o
'"
0
co '" ~ o 0 '" o 0
-
Z /-
<,
... ,.. ,... J'X ..4"'-:L.....=
"N)!'i
, ,j l'i.XI J}I'>l;\d
F.
~
~
rt'lII:H-A"l....,'I-Jl(-'H'
w
-
'. , .... .... .'f' N '" :... :.... :.... '" '"
....E
.__ .__ ... _.
~
NA l>i1 ' Ilt--
a
o
'"w Q.
z 0 /-
U
x
W
-' cw a
.~
a
w ,
-' w
w '" '"
0
~
>/-
vi
z·
0
i= u w -' cw a
iL-\: 7
I
5.3
~
4.6
/ I:l\ / IV I/'
5
J: S
J
>(
2 >F~~
~1.J....+411l II
1I.A-1r1 II V1
LI-11"" I A. V.oI1I
I'''''''
I
I
\
---
.002,005 .005
,01
III I
,03
.05,07
,1
\/\ 1\
.2:3
,5
,7.
I
2
J
LO"O OUR" nON TO rUNO"MENT "L PERIOD, TIT N ,
Figure 3-181
.
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear,triangular pulse (C1 = 0.619. C2 = 300.)
5
Z
j::'
<'J
o
~.O
u,
w
'"
w
\
',02
O
o
0.7~ .00 I
c-
<'5
,-/10
0,5
-c,
~ ":> " /-
'0
'~
~\
W
J'
:1: ":> ::f
Z
~
~·O
~-
',(~
-'
/-
7
10
1,6
a.
1.3 I. I 0.93 0.81 0,70 0,61
":>
0,53
t:
~
If
;(
~
cO
w If
TIME or YIELD TO LOAD DURA TlON, L ciT v'
o
500
~
..
..-J~
'0
,.~OJ
....
0
_.~.~UlJJ '\,(J
:.....:..:.....Vl :-"00 °0
(flU',,"Il.""NN-
~o·~(fl~~ ....
..
-
-- -
0
000000
0
0
U' ..... OUl
lH
~
.. c....OcNN:...:00'1
-
~ 00 00 -OCldCfltJI'" 0(,1100
000
Vo"'''' 0 .. 0
t:b
\X Y\I V\I
,':>
/ w
l-\:~~i--"'t;!::rv1\00
>(
<,
E
>(
z 0 t:
u
w
70
~~
"q.'>
~
\I
,
W
'" " 0
z
f-
<,
E
l!l.l
."
1.\
a
6.2
a
'"
w
0.-
5.3
--' -c
f-
50
--' I. W
~,'>
Z
w :::f
«
a a
30
t: o
.0
20
J:S
:>
'w
--'
>-
z
L.-
a
0
.1. 0
0
~·o ~'.J 2.0
i= u
1. 0
0
1.3 1.1 0.93 0.81 0.70 0.6\
:>
0.53
i=
fZ
,uw
7
--'
L.W
a
5
::f
:> ::f
>< « ::f
3
-r
2
f':1-J "'~
0.7 0.5 .00 I
~
.002.003 .005
.01
.02 .. 03
.05 .07.\
.2.3.5.7
I
2
3
LOAD DURA TlON TO rUNDAMENT AL PERIOD, TIT N • . , f
Fi gure 3-182
Maximum.response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse .(CI = 0.590. C2 = 300.)
5
7
\0
f-
Z
a
w -'
"w
:::f :::f
>< «
:::f
"-
a
w
:::f
e
e TIME v" vi.
~
.....
~
....
~
0
,.;
6'" ... ~
... 0
tJ
or
YIELD TO LO~O DUR~ TlON,
e tdT
0000 . 0 ..0 0 0. c:P~IJ'~~~~~:"",,:,,,,,!=' .en .. ""v.ViNN .... -.. ~ N ................. C,.ooI0b o V1 ..... 0UlOOll.o< (t\
.....
~ooooo -oaOlUl'" .... 0 ... 0
Ul
0
0
0
------------,.-...",TI<,• -ecr-rsr-r-cr-r-r-p-r-s-w-' .r 2 f--
<,
--·~=-M .. ::l-v"dl\\6\lL 1'it\;KT'tH~\N
R
,,,
z
o
.~.
w
,
......'"
,.
Q.
Cl
...J
-~.
W
>'0
,z o >-
Ll W
w
5.J
-;
...J
,._ I
~~
...J
"W
Q:
c,
.
>-
u
Cl
o
'SJITFJ IV &111 OJD
1>1 ......
W
....E
~'6
i
Z
W
~
-c
C Z
::>
"-
v
of-Z
o ;::: o
k
w
...J
7
. ...J
I
"w
64/1/
F:-..
u, W
Cl
" >"<
::> •
~
~
1\
~
r\
0. ~
'1
.01
.02
.03 . . 05.07.1
.2.3.5.7
~PAD DURA TlON TOrUND~MEN~ AL PERIOD.
Figure 3-183
~ p/r I'
~ ~
t>
c
I. 3
::>
i
~
.
='11.
,
J
5
,r
7
TIT N
Maximum response of elasto-plastic, one-degree-of-freedom system for bilinear-triangular pulse (eJ = 0.562, [2 = 300.)
.,
.
~ ~
I.'
'x
0.93 0:8 I' 0.70 0.61
'i
}.53
'1
2
1.6
~ "0
i:
t
TIME OF YIELD TO LOAD DURATION, ttiT v' v'
,., '"
,.,
~
,.,o. -::.
....
...
",,-
~~~
tr' '!'
r:P
l.JO'->G""""N
....
-..l
-
....:..:....e
0 ·0 ·· ·0 · _0 . ·0 0 0 0 0 0 0 0 0 0 0 UI.
0
CD
o
O'l
0
V. ·N N U'I " 0 " " 0
- CJ\ UI
0
01 0
en
U'l
VI 0
.. ..... O' 0
\ .
"f-\-tf'IH~H'r"1f\1--7\h
Z 0-
....)
~
w X
\\J\W t-,
<,
E x
Cl
n
a
0::
w
c,
z
a
;= u
~
W
...J L.. W Q
..,'",
.., ~
I~~ t>
a
...J
w
>-
a 0z
o
I.<::
...J
« zw
,u~:~
0-
::::f
~.
v
"
r-,
VII~
10~1)/'
Cl.
U W
...J L.. W
a
7
I:
5
:::f
=>
:::f.
3
J
~
sc
-c
:::f I
2
r-
~
•
~'J
.002.003 .005
.01
~
\1 \
.02
.03
W ...J
L..
.O~
.07.1
.2.J.~.7
1·6
Cl
!.J I. \
=>
0.93 0.81 0.70 0.6\ 0.53
2
J
LOAD DURATION TO FUNDAMENTAL PERIOD. TIT N
Figure 3-184
a o
;::
::::f
0.7r~ .001
=>
a Iz
W
.\
0.5'~
L-
<'6
i>
\'
'Z
.J. s
<.0
1\
0.
'0
.J. o
0-
«
a
Maximum response of elasto-plastic. one-degrce-of-freedom system for bilinear-triangUlar pulse (Cl =0.536. C2 = 300.)
~
7
\0
:::f
~
::::f L..
a
w ::::f ;::
e
•
.,
•
molc'or YICLDTO LOAD DURATION. telT v'
v'
~
rP'
~
...
~.~ 0
~.'O
,.., ";,'"
n,
><
:1.
X,
c
\l
W
..J ,,-. W
30
20
t:f-......,-jA--H\~IA-·l'<\HI'd\
j""lf-jt-l*--1".II+-W'l\-'i~
'0
.... o
\I'
0
/
,I
.~.
~
)
.\0
1/
. ,;1
r-...
~ 1')1
~
k
.....
V.12b< :--.. .1)
1/
b<
LH 0
NN ..
p
\; \\ ~.~
82 7.1
~
6·2
1/
rv
1/
;-
/
I?
:", I- V
kI-
S. .s-
P. I/l7f: "1': ~ .f.7
•IJ
1/
1£
~
0:0
0.00
~
1~lltv·
iT I'i/ .
'lIE V.
V
(JI'.
....
-1-1 ? 01'1--'
P.o:
I...
t- if
.,.
~
'6'
if '71
. 1/ .t'-'1'-
~P ~ :'--~l1' ..,11
,....
f?
r-
r-...
t:
u
W ..J
._-
~
><
«
:f
2
~
!\
\
I\,· \
1\.
e
.002.00.5 .005
.01
.02 .03
.05 .07.1
.2.3.5.7
;;=
1.3 1.1 0.93 0.81 0.70 0:61 0.53
.p /(1 ';"'Q.
1
2
3
LOAD DURAnON TO rUNDAMCNTAL PCRIOD, TIT N
Fi gure 3-185
Maximum response. of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (Cl = 0.511. C2 = 300.)
5
7
«
z
,<'6 <..s-
0.7~~ 0.5·.~ .001
:f
0 ....
.,~
H
w
=> u,
1.7 1.5
v
..J
....z«
of'
5
:f
0
exw
a
2.0
._-~~- -,~-
Cl
...'0
·7
:f
....E
·s oJ.()
"-
~
z .... <,
"-
IX I'X
VIANI'\>
(ft
11
K)
'.'
00
en
VI LA1'J
M'\KI'i.11 ~I
.~
-
0'00
•
.
I\V..:tt
.... u.
z·
~?'~~~'?O
..,.C,.NNN· ........ ' .... 00 O. VI ..... O.Ul.O m,(,H. 9 0
0\
~\XiIX1~1
.~
".
-".
0'00
I\.
-
z
-'0 ..J W.
"'-I
_ _ ~?
~;iH
.
.~
~ttt\l'~iiI~r
.E
Cl
ltt,;..N . "
N-
I
>< ,.
.... .....,., ....
toJ :.
N
.",
W9l2~M~¥:iJ~~
w
/
.- ''''''
"1(1 N V\IIA
I/~ I lei
300
CJl-U'-·~ .. ·,,"
~
10
'Z
0
u w
..J
"w 0
:f
=>
:f
;;:
«
:f
"0
w :f t:
... ,.
.,
tt;T'
TIME or YIELD TOLOADDURA TlON, '..r
~
.v" ,dJ
..-J ...
..-
~ ~ ,0 ,~''''-;' .
c:P
(J\
o. ~
'dl
""
N
t.,.
'\-.LI \Ii.. \ XII \
tN!\~I)( tvlM I) w X
x
~
z
o
c: u W
•...J u,
" ~\J ~~ f5f
100 . _. "
:;~
>:
,-cS z a
W
~ 0. ~ o 0
:
~·II\,
~ ~
. . VI. I,JI ......
?
~ ~ ~!='
lIl. N OUI
~
•
W
~'
5
::f
...x ::f
0'00 .
0
~
00
OUI
LH N.N
o
'.:0
t(
I;
M tRJ) 17 17
~
'1:i 1--
~.»L
'"
,VIl7~
vtr1tlm
":II""'-"
..-y:..-r'"A
1T"'l'
::.l-
.
"tf
3
J\'::!lJ::rn: J" :JI ..l lII.
2
~
::f
0 UI
0
,I
i'\
7
::>
0
-'DIJ G\
Jt1
b.
...J
u,
a
?·o
N o~cn VI
17
i~
9 w
;= u
1m
l.A
o. ~
~.
.
W
...'", ....
I
\ I-XfW"fl
<,
0'0
NN-
~'~' ~ ~
..
t.U
~
"
I
-
J\. \
I
"11"
IT
.
I'>: ~
U
"1''''
,.&'
1/
I'"
1/"
17m::
/;/:f-. \
t-
s...c~.
.
",\1/
17 t-,V tS1~
U\ II
=\. ::l '\ ~ ~,
p/r
-r
-r
07~
O.5·"~ . 001
.002.003 .005
0\
.02
.03
,05.07.\
.2' .3
' .5 ',7
,.
2
3
LOAD DURA TlON TO'rUNDAMENTAl PERIOD, 111 N
Pigure 3-186
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (Cl = 0.487. C2 = 300.)
5
7
10·
..
e
•
TIME or YIELD TO IcOAD DURA TlON, tdT
~
~ ~~
~
~
-;.
..-cPCJ'lIJl~V
....
t-J0t-J·""IJ4~·"~
...... ~
i..N
_~O~~??5>000aooo 0 0 0 • ClD °0
"'l.oIlHNN-Ul ....... OU'!OOl~
Q'l
0
-
000'l
Ul'" 00
00Ul
'"
o
N
N
.0
\;
G X
\\\P'\~\K
<,
E
>(
9. 5
It-
tt
(\,.1 JfItJII
7"
1/
V1"v
6.<
s,,j
~,
u
W ..J u, W
a a
~
r
'" '"
..J
20
"
l~~~{-}~M(L~Q ~R
z
r
~
0
~
~
1..J u, W
a
r
..-- ,-.
::f
Iv;
'"
'0
.t. o
~m
<'6 <.,]
<.0
7
t.7 1.5 1.3 1.1 0.93 0.81 0.70 0.61
5 3
>(
::f
0 <>:
2
~
u
1\
~~~
~
1'..--1--'
0.53
0.7 0.5 .001
,
.002.003 .005
.0.
.02
.03
.05.07.1
.2.3
..5
~7
2
3
LOAD DURATION TO rUNDAMENTAL PERIOD, TfT Il
Fi gure 3-187
.J
'"
Z
'6'
J. s
:>'
::>
a
f-'
1/
0 .....,.-.
U ,.j
....E w
0
,-
,~
<,
a,
z
'",
., I-
8.2
- ~-/I\:l*KJ'P'ff-~~
w
\\
Maximum response of elasto-plastic. one-degree-d-freedom system for bilinear-triangular pulse (C1 = 0.464. C2 = 300,)
,
5
I
I
7
10
w
::f
.~
z
::>
"0
f-
z
0
t: u w .J
"w
a
::f
::>
::f.
><
'"
::f
"0
w ::f ~
tdT
TIME or YIELO TO LOAD DURA TlON, v' v'
,., '"
,.,
~
~
o
o
~.,;,.,. ~ ~ 11' !"'~~~-:-:-:"'". ... \.JOt-JCfllJO.1'J' .. ..,J-...,VtO~
?
?~??~OO
o
I.I\
en "l.otf.HNN--
..... O\llOO"U4
00,0
00
o
o~~~~ '" o
I t
00 N N ~ 0
V£l4U1 8.2
7.,
Z 0-
<,
...,E a
~<
z o
.J u,
C1
'"
~
v-tt¥--IVtIl~
W
~
I:¥
':.~JhIlf\
W
'"
', 1/1 I I)qJI
~
~
~ l~~~~~~~~~ / ......
.U
o
~VIIV
a
.J
~-
w >-
o· 0-
l>'
z
o
~
v.
~{j
.J
'J-
« 0z
w
7
w
N
::f
« C1
z
.=>
I"
L.-
o0~7
v
Z
.0
i=
u
W
L.-
<,
... ., .~.. w._1"'-
u, W
C1
W
a
/
v
J/
::f
""rl.r--....V'f V
=> ::f
X
::f
I
s
.J
W ..J
«
'"w
Q..
70
i=
""
o
-"+'I:\::::lI=:i::t:ili:~~H"f\:;~
~l\l\
2
I'
I.t
~~
~
=> ::f
>< «
~
~
L.-
o
w ::f
i=
~
.002.003 .005
.01
.02 .03
.05.07.1
.2·.3.5.7
LOAD DURATION TO rUNDAMENTAL PERIOD,
Fi gure 3-188
235
TIT N
Maximum response of elasto-plastic. one-degrce-of-freedom system for bilinear-tri~ar pulse (C1 = 0.422. C2 = 300.)
7
\0
..
e
TIME or YIELD TO LOAO DURA TlON.
"
~~...
,..,o
....
.
.
.
riJ '" U' ;' ~~O~(Jl""
~
~
. '" ....
• t[/T
N _? ~ !=» ~ ~ ? ~ ~ ~ ..? 0 e o co O'l ..... IN" l.N N N - CD en V' N '.. :...... .- :.... i". 0, 0 0 Cft ...... oc.nocnOl·OOlft o
0
000
..
VI.
o
0
N
..
N.
0
"
9.5
z
8·2
200
7.,
w
><
><
0 6'"
100
.s
0
W
...
'J
50
'6'
-J
1_
W
,
W
N
....'"
-:0
0 0
J: s
-J
w
>0
v
>Z
0
'"
U
W
-J
u,
~
M
J: o <.(;
<..1
~
7
'I:
v
2·0
»; /\,r
<,
W
1.7 1.5 1.3 1.1 0.93 0.81 0.70 0.61 0.53
i-
0
::;: :c> :>' ~
><.
l;::::
~ ~'1d1-J
--- ---~
1'-
.002.003 .005
.01
•
.02
.03
~~~-\V
_.
~
.2.3.5.7
LOAD DURA TlON TO rUND'AMENTAL PERIOD.
Figure 3-189
~,
~ p /rl ,;","'ll.
I
.~
.05.07.1
'" W 0-
,
z
t: u
....E 0
"<,
E
f--
<,
2
3
TIT N
Maximum response of eldsto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (C1 = 0.383. C2 = 300.)
5
7
\0
..J
-
z
w ::1:
a
z
::> "0 f--
Z 0
;= u w ..J
"-
w 0
::1: ::> ::1: )(.
::£
"0
w :f
;=
TIME OF YIELD TO LOAD DURA TlON, ~ o
~ ,-.ld»'~
500~
~
..... ,.
-
....
~
~
tP
~
O'l
tdT
If" ~ _~ !"" ':-':'" :"" :"" ~ ~ ? ? ? ? ? ,lit
0
0 0 N :.. OOCJl. ...... OlJ'loO'lVt Vt Vt N
---~W1m~~~ 0
VI
N
..
......
_
\ i1~:1_
-~...Ltt---¥1\'k1'\HH£-lXc\-+--*~
300 1=lf'1 200
-~;;\ "
)f
• • • n,
....
QCI
(fl
~ n
on • • • _ ,
,_
0
0 0
-o '" '" 0
'"
.
00
'" o
0- 0
a
'"
N
o
0
N
..
~ ~
Z f--
\2JM ~ VI 1/
<,
....E
~--M
In
.0
;= u W
"
..J
I
L..
0
--
i'--
\)(LI~ z
..
."
6'
s
'J
v
~
~
::F
tzItm-tVUV
>-
17IJ..l-.-7, V,
- ~ Y IMIM IJ'-
w
o
f--
...
Z
o
w
7 5
::F
::>
::F X
::F
~~~ ",,_ ... -- ' .....
"
a
'2:
rl
v
::>
Jio'X h(Jl'. .....,
L..
o
f--
2:
~
,...... ~ ,.,:,/.-. IV:.
o
;=
u
W
L.. W
o
..J
w
-'
o
-'
n,
o
;= u
cc w
Z
1111'\0,'
W
..,.,.'", '"
a o
.19"
~
...-
W
a
::£
i>
3~ 2
..J L..
::F
-h
rwM
0' .?\J;~~.",><_, :;./A,y~:;-,...].->oc....
M
ffi I I
~
.002.003
.005
.01
.02
.03
.05 .07.1
....o
w ::£
s:
----
.'
.2.3.5.7
2
3
LOAD DURA TlON TO FUNDAMENTAL PERIOD, T/T N
Figure 3-190
x
\f.\
~
~"
0.7~ 0.5-, .00'
::>
Maximum response of elasto-plastic. one-degree-of-freedom system for bt l inear-t.r tenqul ar pulse (C1 = 0.365. C2 = 300.)
5
7
10
e
e
e td T
TIME or YIELD TO LOAD DURA TlON,
.
~ ~ 500
,.,
20 ~
x
<,
E: x 2:
a
70
;:: u
W
50
--J
u. W
, '" N
n n
30
w
20
--J
~
>::
'"
>-
a
2:
a
10
l) W
7
;:: --J
k W
n
OJ ~
X « :1
'"
'"
,II; 1\IX'1)\
e'
.
;.~
cst l&r.l)(W" ., " " ~ ,,~ '-\~ M
l7t1t r Ie -y.
"'hi\}"
""
"o~"I~ " ,
-1..' '-'
;h
'"
Y
I
~ W'
,)
J VI "
o
~~~???O?oo
0
.... ""NN:-:-~· . N·... :....:.~""'o~~
'"
- ' , ''\
p\II'.
. l/~
.
~~
r-,
II
~,,"
r-,
',(
"
I~,,I,'~J' Ilt
~ ~'/ rJ, '
'"
,
~
~' ,.c.
'.'
,15" 'I{ '- k:~./ '. .
>i
V
'"
b< K
.
, V
0.5 .001
:z
8.<
»
'-C-11-1
:1
rl
./
. IS'
~
',",
...t'J: ~.6'
0: 0
>",V:"
vob'" " /P
"
'"
J:,s
..t. o
v : :Jt't " " ',L' ; '" "
~.
L-
1/
,/
"
J.
_,
"',,<00
07
9.5
-r,,".
'
"
2
~
......
'
3
7,
.. 0
'" 0
<'8
<.J .<.0 1.7 •1.5 1.3 1.1 0.93
r-,
" " ',,",-\-).1/\[/ >'0/.""
1"//
~ Vci
." . [\
0 00
0
'/" -I'
V "
V ,,'- /
'"
.
0
VI~
-z
N ''" b,
1'." 1/r-, f-.... hI
II
o
'" o
.lAl",iNN---goCJ') U\....aOUl 0 en V4 0 0 til
T.I ./'
.... ' ' - . /
5
~
rJ> -0 i-> '"
'"vN VI I./\1 :---. I.
,
100
~
,f \ ) Yln~(...·· ~ f-...,L.I\
(\ 'Q.. V\ 300
.....
~
0
0
"
~
.002.003 .005
r\. rx"'\I'X." T . .
'\
v"/
,,~
r:>-a- ~
0.81
0.70 0.61 0.53
IJ...
.~, 'I' ' b.
><,,\l) '
'
......, .01
.02
.03
.05.07.1
.2.3
LOAD DURA TlON TO rUNDAMENT AL PERIOD,
Figure 3-191
~v
., ,/
.5'.7'1
2
3
TIT N
Maximum response of elasto-plastic. one-degree-of-freedom system fOr bilinear-triangular pulse (C1 = 0.348. C2 = 300.)
5
7
10
l-
"-
....E a
0
'"
W II. --J
«
l-
z
W ~
« ~
:> <.. 0
l-
Z 0
;:: u w oJ
u,
w
a
::s
:> ~
.~
::s ".0 w ~ ;::
rH.iE or YIELD TO LUAU DURA TlON, trlT v' v'
soo
w
<,
~
"
~
..... -f.
\
"
IV \
~ 1/
V \
0000000000" . ' ... ~NN---·· . . . . . -_ . .- 0 000000 .t-J G .CJ'I ' V. . . . . :...a' • ell O'l ... lH t....N N co 0'\ t.n ... UI N N ...,. N ........ lH 0 O. 0 U' ...... 0 U\ 0 a, Ui 0 0 UI 0 0 0 ... 0
c:P
,.,... .,l-J ....0
:7 -\
X \: Atf 200 I" I X ~ -\ ~A .Y \: \;; ~j
E 100
"
,.,
J
300 \
sc
,.,
./-c.
»: 1\/
II
" /' J
r-, P'\ 11; 1\ r\
kl\ I'v rx~
XI
II'\\
~[~
t:><
r-,
I\. /
~ 1\ V
"
1'\
.
"
/ "/ I'.
f.. /
''7- I;>(
rz- -IJ V ,./, V 11. .4
I,
l.-
V
I
:t-v ~ ~'-
j)'
:--
I'v
I'- /
IJ
~v: y 1-:: l'
~
z
0 >-
u
W
s 0 t-,..lo;kN--A1~b'-'l-).'M
IX
IihAL"'I 'i:.:.<:l ex.: II i ; }t-- A" . )
...J
c,
...,, I'.>
0 '"
W
0
a...J W
30
£-:::-rI-V't-'\"-,!;t-\--14 \-J
L 'Jt, .'1 ~ IVlKll A"'ft
1<:~~
Z
10
u
W
..J l.. W
0
..-
..--0'
_ . . . _A ~'-i1'-~---Ilt~k\-I,.~, IA~M\Ml-,+::iJ!"tH- IL 'l"::tfl I LV H ~
0 1-
>=
.
. ')K \7 i'..
20
)--
0
Pr IV'
7
:.L
stz~~~
P'
s
~ -'
l-'~
iLl
'AKI
7
N IN!
1-
'" . . v:.-: II p .
J.
7/
l;'
/
v
/
",~('..;V
N'
~
' / ...,
i"1.Ll
"'ll
7 / ~ I\. I'
z-
Y1
~
rn:r
,
.~
~
[/
:>:
:>
. :>:
3
x
:>:
0.7 o.~
.00 I
~
.~
2
~-= .002 .003
.oos
.01
.02
.03
.05.07..
P/ r I ';'n.
.2.3.5.7
I
2
3
LOAD DURATION TO rUNDAMENTAL PERIOD, T/T N
Figure 3-192
Maximum response of elasto-plastic, one-degrfe-of-freedom system for bilinear-triangular pulse (Cl = 0.316, C2 = 300.)
5
7
.0
e
e
e ~
.. -r-:
. -.l
TIME or ~IELD TO LOAD DURATION, lclT
~ ~""
~
o
~
", :...l
t-J
""
~
0
CJ'
~
t1'.
N
000000000 . , . '. . . . fo. ~NN-0 0 U' 0 U'I 0 cn Uo
"",Nto-J-:.a:--"· ' DO en ..
-.I
-.I
lH
0
~--------
0' , 00 ,
-0
co '" 0
'"
o 0000 (:, U'I.V1NNN o 0 ~ 01 -'_0
-----
~,f\ -~ «" ~ \f\ l/lV-ll4m ~\I/
9,S Z f-
'11.01\.
I)<,
[')<
Yk
:;~;';;~!I~~\)
II --------- --
<,
....E
.. _...
1.lVI1~
I
o
o
'" o,
w
z
o
..J
I' /
f-
U
.111\
W
-'
L.-
~""-I'
_
v
II
_
~\N/i\'
_.. _
f-
7w ::f.
W
..,
.., o
.... '"
o o -' w
p
IlL
Z
:::J L.-
....o
~
o
~
f-
Z
o
t:
u
W
,~Pftt/~
~
"Z
o
i= u
w
jf\,
7
..J
L.-
...J
W
o
L.W
o
Of :::J
:f
5
ri
3
;<' 4: Of
Of :::J ::£
2
~
\-
~
i.I'\. \ " \. J-. -,
0.7
J?
0.5 .001
v
p [c
,
~ .002.003 .005
.01
.02
.03
o
W
Of
I 'fll..,
i=
~ ,OS .07.1
,2.
,5
.5.7
1
2
5
LOAD DURATlON TO rUNDAM~NTAL PERIOD, TIT N
Fjqure 3-193
Of
L.-
Maximum response of elasto-plastlc. one-de9r~e-of-freedom system .for:bi.11near~trian9ular·pulse ((;1 =.0.287. C2= 300.)
5
7
\0
TIME or YIELq TO LOAO D~RA TION, telT ,.>
',.>
~
d'
500
r:P
tJI
(7\.
~~",~-of..,l'O'~ o .
'J\I/'l A ~
'\J'-....~
~
Vt
N
,N _
_
"_
0
0
N·..:~:.~.~oo,m o 0
.v.r. . . .
til ...... 0
NN U\
0
....
-
C7l
lot
?ooooo
-OOO'lV'.v.r. ,0 0 lJ'I V' lJ\ 0'
0 lit
0
o
0
N
N
..
0
I\J..II "lJ11',
~
fi \tI
?O??~O?
9 •.5
I
6~ w
>< f: ><
z
0
;:: u w
100. /\ 70 50
... ..,,
..,'"
-' w ~
',><- '..A
30
-- -
::::J-
.
IV I'l-
I~ '\ I i ~ i<\ /
u.
w
V\IJI
_
,
-'
a a
.:m~rw)K
.~ [\Ill( tv-~~vI
<,
-
I"
t - } ) I i"'-V
,
1\
IX
~
..
r-. . r1rni
~ l/ IJ Y"\)1'-"V Vf'. V '"
20 ,1""'V-hf-li7'
IS'
'r
~I
i'Vl~~:HEWM~RlM 1'\.11
nq
~'"
10M,!
"IN'· ill Il'Il-Irll /I V l-VT L%
I
f-
1.L
Z 0
t:
u w
-'
5
~
:> ~
><
..: ~
2
:". 'I. \.l.. \ 0.7
J.
.../
05 ,!~.d2 .001 ,002.003 .01l5
.01
.02
.03
,05 ,07
.1.
.2.3
LOAD DURA TlON TO rUNDAMENTAL PERIOD,
Figure 3-194
Maximumrespons~
.5
,7
2
3.
TIT N
of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (Cl • 0.274. C2 • 300.)
5
7
\0
~
..:
...
:> 0
.f. o
I-
<'6' <.,]
0
0.53
p / r ';ll..
z
a 'z
1. .1 1. .1 1.1 0.93 0.81 0.70 0.61
~
3
-' ..:
f-
0: 0
1.7
~.
w
a
'"c,w t:J
<·0
7
0
":6'
J: s
~
0
....E Q
NUdl7 _. - -
_ ... --- -
,1%'... ,,·
I/"~)
17~1Y1V ~~n ~~ \N V\7~ nI) I 'I)<
....
~
z f-
<,
z ;:: u w
..I
u,
w
Q
~
:> ,~
;;:
-c ~
"0
w ~
i:
e
e
•
,
TIME or YIELD TO 'LOAD DURA TlON, t[/T
~
t..I
~ t..I~ t..I~
o
,.
...
,. t-J
".. 0
'f:!J
t..J
'!'. ¥' ~ IS'
VI.
lJoI.
N
..
~
-a
po.)
:...
~ r: ~? ? ?? ?? ? 00 (fl -...I Lot 0 l.JoI, ..... N N o 0 ...." QUI 0
50Q
I",-/' "/ .
';_\.>.0",," '1',/_\ (\ y '/\ I\ V ~';' 200 :i -, D.. 1"-r>:1'i.... (I /, ITI\-~' :\\ ~
500
w
><
<,
E >< 100
z
0
70 50
-' CI
'" '" ""
-'
W
>-
30 20
0-
Z
O'
10
w
7
w "
5
-'
CI
::f
::> ::f
3
><
'"
::f
I"
IJ
\
L
2
I'
L'l1"N
IY---IX'l I
"-\~ 1" . ~~ r-,
1\
;!;;" r1'aLl-X:, [', It>--.:i\..J
1/
vt
v
/
Dv . r; 1/
\L ~ .r-
V
1/
I~
'.
lI'
, 1/
'"Y--
I""
.00 I
IJ
e<
K
I-'
J..L
F-'
_\
."
,.'" J
~~~-.lrHl'tH1t'ftj III 17 I
I''J'
\)C
.~
1 "-
r
'1/ 17
H:
Id::"'
,~
./
,;,- po
I
L'
I
1...""
~
":>~I"<
I-'
,___ ~
LA'"
Y
I'
~_ \'"
r.0 I
.02.03
.05.07
.\
.2
5
.7
3
LOAD DURA TlON TO rUNDAMENTAL PERIOD, T/T '"
Figure 3-IQS
Maximum response of e l as to-p l a s t ic., one-degree-of-freedom system for bilinear-triangular pulse (Cl = 0.261, C2 = 300.)
5
~
\0
Q
1Z
o Z
o
s: u
w ..J
w·
I.J
::>
1.1
s
0.53
7
W
<·0 1.7 1.5
0.70
'.
2
-c
0-
0
0.93
... .3
-' 0-
.1
o.a 1
L..
e: I-": L. l:::::..
0-
::> "-
. 0.61
,~p/( . _ •
es
J: s
<6
I ':: '.n..
o
..:
I
"'-'"
~\
'6'
<.J
l:" X' K\
\k' W ~K \~~ ~\ . ~ i:">. ~~~
c-:::
.002003 .005
A:
e-"v;Y1!l
Ll.
.
J(
.. e<.
~ l'l~n
""
z
':0 I'
L > ' ''':
v: 17,-z. /' -;
z
.....
<,
E
I"-''".
' j
~ ~ r-- /Y~'" r-
V': t-
~M.J:l~" ~.r ,-<; '-'l< r V.>It:'" "0 <:> -'", . _to
I..... L'L L. c. V L l--'L J-::. r
9·S
..l
"1<
I~~~ ~ F(')
~
v /
V
-r
ilC:>. hL H~~r\.0..V\
..
CI
Y
V
"0 "0 o
0
/'
1~
/
I.LI.
L\
Xl'D='
0.5
V
L
,oX 1-.IIli
... . .V\X
07
:~" 1/
')j
r
1/
0000
-OOCToU""'UovtNN °OUlUlUlOOO.,.O
v
:L"
~
lJo4
k'1",1 -t:.J V Iii V !,kt::/'" (/ r- 111/
,~
d
o
0 -
-,
". .
lLJ. ID r--."
-\
0
,U
V
'-\.
..Y
CI
"",
WI \
,
(fl
1 - . . . . , . '"
,,1-\>
W
"w
-';( .If'. .xr\,~ \Y\ A' " " ~\ ~ 1\
0
--r-'N
1-
U
~
v:
1
Ii
X'
I~, f'l' .
"l
\f-. liLt'- 1\\ vlL
'
0 -
u, CI
::f ::f
::f u,
o
w ::f
;=
::0
TIME or VIELD TO LOAD DURA TlON, tdT ~
..D
t..l
..
t..l 0
r:P
. t..l
~,.""...
...
~
0
(J'
.
lJI'
....
V4
N
'
(J>
.....
N
N -
'.' ....
....,J
-
o
0
0'000
0
00
0
"
0
0
VI ...... 0
0
(f\
l.H
0
0
~_.
. ·OOD' 0
-
.
. . . . .. 0 0 0 .. " " ' L H N N - - - c o m l J l VI
0
U'
..
.
0 0 0 .... ..,. ..... OU' 0
.
0 0 N!'J ..
0
1J
1 1 Y / 1 . - 1 .....
9· S 'Z
n. n''lI AJ ""
~
"'"
.Y
,."
. - ·tlflt\T.I~U' \\ J \I 111"'1 M R){J
><
<,
E
....
..
.....
......-
"'" .lIIMI
><
z a
. In..
j..: :
II:£...
XNJi .,,"w"
I~"
X\l~
v
I -:RI c
<9"
17
.~
VII --'-"--11
111~
--'
."
(AI ...1:11.....,1"""'- '.......,/1 \
llyn .K 111 LKIIL/ ~-~~.~~--
. . . . _ . _ .• ~ ~ ~. •
u
.-
J
."
.s
c+ti-l-M-tl.-
.-.J
0:.-
o
l~
W
0
"",
'" '" ~
0 ...J
'-'-""0~
~
W
>0
""-I
AI
'V'
rv" .........
I-
I
A''T'L.. f
r
I I
r
I
.., I
1..-.
I"'~
~x-
Z
a
c: w
0
::f
S
<..] ;·0 1.7 1.5
~
u,
.• I. J
1.1 0.93 O.B I 0.70 0.61
::J
::f
><'
..s:
::f
"
.00 I
0.53
p/rl';Q.. ::.-1........t>
o5 .002.003
.005
.01
.02
.03
.05 .07.1
.2.3.5.7
2
3.
LOAD DURA TlON TO rUNDAMENTAL PERIOD, TIT N
Figure 3-196
.. Max,imUlIi response of elasto-plastic, one-degree-of-freedom system for bilinear-triangular pulse (C1 = 0.237, C2 = 300.)
5
7
..:
I-
Z. w ~
..: ::J
.Ii"
...J
W
Q.
of
---
u
w
cr
Q
<'IS
'j...
0
0
0: 0
.1. 0
JV\l)<
....E
...J
'of
:::IZ'"
w
I-
<,
10
:z l..
a Iz a t: u w
...J
l..
w
Q' ~
::J ~
x ..: ~
l..
a
w ~
t:
e
• tdT
TIME or YIELO TO LOAD DURA TlON, ~
t-ll»'
'0
6
! ...
....
~
cP
'r-»
o
f!" '!' IJ4
(f>
N !"' ~ ~." . N
.... -1_-.1
. .
'""'°
•
?~O?~!=,?O!>OOO
0000
00 'G'l
.... CA
CllI (J'l
I)\
0
l/I-...&QUI
Olll
OOlllO
0
l;l
N
N
-
0
O"l tl'i
••
-
a
..
::
~~~~~Mmg~~A ')4_ ~n
~.
.
_.
r·A1
...,....,..
o
... ..
0 N 0
,9.
s 2
><,
.
....E
~"i\-...'
~. I~~ ~I ~ ~
z
............
'\X".
~''''$,£
4\-,I~H\,)H
o
w
~..:
u
w
II'
-'
u,
w
n
'",
w
'"
o
'" '"
o -' >-
>-
• r
z· o
'i1;~ lJ\"lk1 pt.
:~ .. _..
s'
.oJ"
~.6'
~
j...,
~
i=
u w
-'
7
UoIrI .k" \A
r.. W
a
.x A \
~I
/I '1.<\
1.V'--"
5
.. t+\~~=t-t::f\f.&:-~~=-r.t=::!J;I?\
:>C.
'":>C ><
.~.
4:
~
llU+T""
• ~
_~I\'"
o5 .00 I
7r4
p irl
Tn.,
~~/yYI'
V'"""' A"".A"/ ¥ ., .
.....-='J -j:;;
..,002.003 ,005
,01
.02
.03
.05.07
,I'
,2
,3
,5.7
2
3
LOAD DURA nON TO rUN DAM ENTAL PERIOD, T/TN"
Fi gure 3-197
Maximum response of elasto-plastic, one-degree-of-freedom system for bilinear-triangular pulse (Cl = 0.215, C2 = 300.)
5
,·7 . \0
TIME or YIELD TO LOAD DURA TlON, (ciT
""."" "' ..
500
~ ~ .~ 1".) '0 ~r.J ~O"'v< "!'
~VlNN7-':"'~O N ... -..I_ ...... l;lOCOO1 o 0
0000000
.. v.
VlN-N fJI ...... O U l 0
-
6 :.'" 0 0
0
-
O'll;l
LfI. . . . . . . . . .
000",-'00 (fl
U" lJI (Xl
0
lJ\ 0
I
SOD 200
o
0
N
N
..
0
, ...
."
dO
r-
w
-'
xc
<,
·E
><
z 0
A I
100
W ..J
-
~
70
'/
I it
I
A=:J::j::;=
0 I
:u f A;t
U7nt1:;;t\,) ::P" \ I. J,
~~~ ~rru \( \I KI ,
F-
U
,1M I\i' '{)<.! '"It,;I1 \J T\ ~ V\\V ~ f{i,-n-no'j\"a--1~iI a:: ,
50
,..
'", '" '" '"
30
r
wo
20
III I
."
6'
.,
'6'
:'·0 J
.~.~
LA' II Ll!~ ltt-m" 1\ 1A1. X I Ak>~'
10
f=
7
~.(; ~. .J
, 1/l1""}
-'
<·0
IVlII 1¥LHfVL--
1.7 1.5 I.J
L-
W
a
'"
::>
'"
5
S
V " \ I I_ _
'"
_ .•-
I I ~
~
><
«
2
~s..:31\l \1 \ ,\~
o5 .00 I
s
.f. o
,;:~
Z
·w
r
pl(1
....E •
a 0
'"
W 11.
..J
<{
f-. Z
w
::E
.<{
~~FftpTt-,t-Ibf-,irf:*f'1C7~'iEi~frlF1'IF:II~'l-.d,f-;d"i*,'ll
,u
.,,~_,
\*1'11
X.III/.VIIA~~"'--VI
0 0
.•,
ll, 1\',,"\
.s',J-
~
~-7IJ,f-H"-d+-NI\-j'W+lNf\-?'oI
0
.ssw
r
:f-\I7
W
Cl
1:1
2 f-
<,
~1l.
1.1 0.93 0.81 0.70 0.61 0.53
a
::> L-
0 fZ
'0
;=
o
w
....w
..J
a
::E
::> ::E ;( <{
::E
.... 0
w ::E
;=
~'
E .. 002.00300~·
.01
.
.02.03
.05,07
,1
.. 2·.3
.5' ':7
I
2'
3
5 . 7 : 10
LOAD DURATION TO rUN?AMENTAL PERIOD, TIT N
Figure 3-198
Maximum response of elasto-plastic, one-degree-of-freedom system for bilinear-triangular pulse (C1 = 0.198, C2 = 300.)
e,
e
e
•
TlM£ OF YI£LD TO LOAD DURATION, telT ,-.l
,-.l
.,
..
t..J
0
~
~~
cJj
.....
o
~
~
0"
•
0"
•
V'
.' ..
0
"vt~N---'
•
~
,
..
" • ':....i • • -a VI. 0
V'\.V\'lX \VWlN
~
.X
\/I)(~. '
)
~I-
x
.Z ,0
0
70
.
-
-.
1X\~
en.
.VoINN Olt 0 ..... (D",
0
I'--
ow (,,,
000 t»
o
Cf\ LI' .....
..,.
l/I U' -.1-
U' 0
... 0...0
'0
Ul
~
,t\-It1
A1
m Q~ l---tY\
~
17'
.
----~ No
\~
i'A
J1
w
00000
0 00000 .. -
aa CJ)'"
-'
,.,
If'.
IX
J'
- -----'
'J"
o W
wtr:~
W
'" .... '"
yo
I ..
...J L..
,
a a
. If'
...
~o
rrk'~
...J
w >-
o ....
J'
-z
f'-
~
17
o o
....
J: s J: o
11
~'G ~'J
w -' u.
~.o
A '\( \ ,".17'-.
1.7 1.5 I.J I. I 0.93 0.81 0.70 0.61
W
a
:>: ::J
:l' X
1\\
~,.
.< :>:
L O.7'T~
O.5/~~ .001
.002.00.1 .005
~'\I\
,~
0.53
~ .01
.02 ,.0.1 ,.05.07.1
.2, ,3.
.5
,7
I
2'
3
LOAD DURATION TO FUNDAM£NTAL P£RIOD, 'T/T N
Figure 3-199
....E 0 0 l>:
W Q.
';::
w
....z
<,
'",
r-,
wtr/
It
.,.,
Maximum response of elasto-plastic, one-degree-of-freedom system for bilinear-triangular pulse (Cl = 0.178. C2 = 300.)
5
7
10
...J
....z« w
~
« ·z
·0
:::> L.-
'0
.... z
0
i= u w
...J
L..W '0 ~
:::>
~
«>< ~
L.-
0
w
~
i=
TiME or YIELD TO LOAD DURA TiON, triT ~,-J~6
oS>
p
0
....
~ ~
rP
.~
o
d' '6'
\J"
v..
I'>
VtNN-
~.~
:....:..:....
__
~
0
~oQll o 0, 0
00000000·······
..... .... Ulv.N N U' .... OlllO .. ·l.ot
=. -
. -0
..
0000000
00 N N OlllLl\-JOLJIO"'O
QlI
CfllJl
..
I'>U'V'
Z
f-
"'-
E
·oJ
w
c-c
o
<,
o·
E
><
'"w
0..
z
J' ..J'
0
eU
z
...J
c-
::f
n. n
n
w
w
«
w
"",
'" CD '"
...J
«
f-
z
::>
c-
...J
W
o
>0
f-
Z
e-
o
t:
Z
o
0 ~.
u W
W ...J
7
v
...J u,
w
n
5
cW
o
LV
::f
::>
::f
::> ::f
::f
sc
3
«
><
«
::f
::f
....
2
a
w· ::f
": p 07 ~ 0.5 -: .00 I
E::t:~
,002.003 .005
Ir
t:
=
. .01
.02
.03
.05 .07.\
.2.3.5.7
I
2
3
LOAD DURA nON TO rUNDAMENTAL PERIOD, TIT N
Fi gure 3-200
Maximum response'of elasto-plastic. one-degrce-of-freedom system for bilinear-triangular pulse ((1 = 0.162. (2 = 300.)
5
7
10
e
e
•
TiME OF YIELD TO LOAD DURA TiON, tdT ~~. t-:I0 U": ..
~
'"~.~
,.
o
cP
~
(J\
...
;.;
":'"
•
"CJI
500~47\l7\17;~
V'
)Il
.~
~
~
..
N
N
.... -
0000000000···· • •• • • • 0 0 0 0
_. CIlI
en
•••••
"'-I
~
-
..... 0
0
0
~,."~~ IV' 'A ""'IIII)Ik \/1
"t,.ooIlI'IN
'VI ~ O. U\
N
-
-
-
Cd
0
(tl
Vol'
0
0
V I r,J Y u-
x n
200 w
<,
'"
100
..... 0 0 0 0 0 "V'v& l'J N 0, lJ'I 0
0
:1IIX ~7 <$'
',>
8 v
JI
H71V1~
~~~~S~~
..
'1/1\ /1
~
JOO
'"E
(J'IV'-' lJI V' ~
.:>
'"
.,..
/S"
....2
<,
....E o o
'"
Q..
.s-
'.,j
t7
...J' 4( ~-~
Z
.. - ..
..,....""
,,~ -fOO--k1th1'.V+7f.d-w:l~~ff<>j!
4(
a z
·~I.LI'b.l-b'ffi";\f-~:l-I+MW<~41,t:l ~ :"-:;;:';~~~ "" , ..rr--4...- Y1 I '
r
I I I 1AJ....r
~I
v
..Y/ 1/7 •
\I1'lli
.......A1
::>
\')<
f .
I
'"
::f
"1~
L..
o .... z o
~!'~
t: U
"
~~:t
v ~\ ~J.
~
::f
::>
::f
l1-,¥
.(p.D5
.~7
.1.
~-+. '
.2
,.3
.
'~.'
.5 . . 7
'LOAD 'DURATION TO FUNDAMENTAL PERIOD, T/T H
Figure 3-201
::f
:'\.I\, ,
_ ..~2
.~
~
k.'
"
.0\
a
I--
.• .,...."'-+ 2 3
-'--~,..
I
'
Maximum·response of elasto-plastic, one-degree-of-freedom system for bilinear-triangular pulse (C1 = 0.147, C2 = 300.)
5
7
\0
• - -
'.'
I.
'
TlM(O'(VIELD TOJOAD DURATION,
,., ,., '" ~
dl
t't:,~
500'~
\1'
(J\
'~
-;.~"O
.... _ VI _f'-J
N
0
j...) .......... :.."-a
'a>'"""
.~
r.
x-,
I'
~
t~';T
~ 0
~. 0
1IV IJ)<1
~
z a t:
u
-'
l'>JJ.L
W
'",
...
n
'"
a
~
~
-'
>-
~
>-
z a >-
....... ,~
000 (TI
UI."
VI I.J'
(1J
o
'0'0'0 ........ Vo 0 lJI 0
r->
0
r-> 0
~
"1/
'r1
1JV11/Y1~ ~\~ M; ~
2:
>-
J
tdl~1\Mi'
a a
0000
I'\i-.L" .r
U\ll~·n/
L..
0
~ooo en .\.HVlN N :..:..:... .• ~ O.C .(J\-....l CLII 0 O"Ifo.NO 0
II
<, .-> '",
.,.,
6'
Y1
'.r
\\~
-
.lI ~'YIN
a
a w '" ll.
.1"
J
....E
~
'6'
...J
4:
I-
Z
w
~
4:
a
." OJ L..
a >z a
\n{~ --,
i= u w
u
...J
w
L.. W
a
L..
.Ii"-
~ ..~~~~ /I Z
-'
W
a
~
:-I
~
Y' '
OJ
'1 '"
~
X
4:
~
~
-r'
-
.
~
f'f~ ~-'(
\
{'l\f'{ \ \
.02
~
vE:,.. 1~~
X
4:
~
.03
"'I \
b'
..,. .\,~~
l--.01
OJ
~ .05.07 .1. ' .2 .3
p Irl ';;l).:'l
.5.7
I
2
3
~.
lOAD DURATION TO rUNDAMENTAl PERIOD, TIT N
Figure 3-20l
Maximum response of elasto-plastic, one-degree-of-fl'eedom system for bilinear-triangular pulSe (C1 = 0.133. C2 = 300.)
5
L..
a
w
~
i=
7
\0
e
e tdT
·nM( or. YI(LD,TO LOAO DURA nON, ·0
,-l ,,$)
~
1iI'"
~. 6"
0
r:PCflU''''vtt..Jf'Jc:''"'=-'i' • -'."-a' -. I"N 0 1XI
.'" ~, .... 0· ..1 ","...-, t--!"'"
t-I
V'
~
--
o·
0
0
0
(J'l
...
0
VI.
Vi N
U' -.
0
.
•
0
0 0
0
,
• .
a
UI
N ~ -;. 0
0
0
.
.
"
0
V'
~fi:i z o
L.W
o
o --' w , "N
o z o
U'
0
0
...
0
~
z
f-
<, III~
'/
, .,., tl'
l/{'U I)
__.. _r "IVIIN/.\.Yli(..J
q II
\..IV
r-1
....- r
1n-1'"t?
1/ ur TI A
... _.
It--17Ih ~.A7r ..v:-t\71\ Y\" . _.rl\lA-
....E o
o <>:.
w
Q.
--' «
>-
z
~
--'
....'"
U"
'.j- .
w
>-
.
Kill)'
o
>-
.
oJ'
i=
..,
.
\!~
~I/If'-
w ,
.
l Y
,~
~
U'...co
000000000 OJ 0"1 ~ I.J" I.J" t..J N
N -0
w :>
«
a z
::> L.-
\jt~k~
r\
i= u
);
o .... z o
11--
t:
u --' L.-
w
w
W
--'
a
L.W
:>
, o
::>
:>
::J
:>
sc « :>
-\-m
>< «
.,,~~
:>
..... --"
\.
o
\/\
"
L-.0\
,02
.03
P'4.rj.
C
r r C
I I I I IJI
.2.3.5.7
,.
I
I
J
2
3
'.'
LOAD DURA'nON TO rUNDAM£iHAL P£RIOD, T/T N
Figure 3-203
Maximum response of elasto-plastic, one-degree-of-freedom system for bilinear-triangular pulse (Cl • 0.121, C2 • 300.)
5
, "
7
.\0
TlME'or YIELD TO LOAD DURATION, '"
'J.~~~
\f' ~ '0'00.
....,),.
o
VI
(J\
'dJ
'0
N •
__ • •
o
VI
"0
0
0
··· 00 (flo IlL
-
0
0
VI
t[/T 00000000 •..• Ut I.JoI, !"oJ N
,-
" -
~ ...
\II O' "" -
. ~:. . 0000
-
-
00
(J\
N
0
0
UI
!J"" VI
'0'0'00 V"\"l'l~
...
N
0
G'o!J\
v '\I JI,. IIVI
/
..... "
r-,
~ ~~: ~
---~-
.
t-
<,
/11) v-.-r::v:r-rb\""r,..~l'
I I ii£ i
----"~~~
Z
I""'"
-
....E
·~tI
!7JltWXlY Y1Vl--
o
ttl JJ.&- lIl"
o
\\
'" W
Q.
z
".....
o
;= u w
,
l'>
o
'" l'>
~ - --~A<-::I-:f'+7'ffilP6i!>I"F;ALi.f'<'H!--'
-'
w
)0-
A'
VIlYlH •
,
-
,m
~ ,_
,
."
I~
tZ
. rY\1 \-'1)/\ I___________ kl..V •. _
o
.....
r
.Lt ,
111A11 .r:
•• _.
t?
z
v
~
W
w
«
t-
L.
o o
-'
I'/'
II
"
-'
. ...... ... .....
w
::F
«
o
Z
...o
:J
u
1-X1.l---Y1\/
l-l
li
tZ
~~\I)<
o
;= o w -' "w o
tU W
7
-'
"w
-
o
.
::F
.17'
.It-
II
~
:J
::F ;.(
«
::F
~
~~=-O.
v
::F :J
::F )(
«
::F "'0
~
w
1\
::F
;=
p/r -
0.7
f-71;
0.5 .00 I
+-
.002.003 .005
.01
.02
,03
.05.07.1
.2 '·.3
LOAD DURA TlON TO rUNDAMENTAL- PERIOD,
Figure 3-204
.5.7
I:
2
3
TIT N
Maximum.response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (C1 = 0.110. C2 = 300.)
5
7
\0
e
e TIME ~ ..:f
~"",,.,,..:,,,,
ocf'v'
or
YIELD TO LOAD DURATION,
C!'IJ'!"'
o
oOJoooooV'l'o
~ ~ iJ·
YNf N
1\
E
><
z a
"""
. . l'-.l oN
lDNo"
Ii I II II1.l
....
...
"
_
... NO
00·0'0'0'000 DD Cf'I IJ' .. V'" V' .-J t.J oV' ..... V' .... oV' 0
~~
It-fflIM-!
."
~tIV\i.-V
'i. 1'\(1 Mr'XIJl
·cw
~
o o
..J
....E o
II
w
r<,
'6'
<:
U
z
1'1 Y1~~
\/1 \
r-
'" '"
VI
'~i)
<,
w
CJ'l
O
1'\1
><'"
w ,
.
ldT'
0000000000
t;-J~~(JI
......
• s·
'J'
~
. tv<.! \N ITI4I1I'h
Z
a
;:: u W
r
I
••
Q.
....J
«: .... z ~
« z
:>
L.-
....a
--.--,....,.,., r" II ,.....,
vV1..A1".J..f"'1\J -_M _~
r,v",v.--r.~....-~~
z o o
~\l<
t: W
....J
u, w
7
..J L.W
w
o
>r-
'" W
W
a
a
o
5 . M.:7'l'\-m
:f :>
:>c ><
3h>(.l;t/lXAPm
>< «:
:>c
2
o
:>c OJ
«
~
'>,""'""'-
:f "-
o
'W ~
1\
::'~
,001,002':003- .005
;::
___,- I.-,01
,02
.03
V.//V/f././V
.05,07-:1
,2.3.5
,7
2
3
LOAD DURA TlON TO ~U~DAMENTAL PERIOD. TIT N
Figure 3-205
Maximum respons~. of'el~sto-plastfc. one-degree-of-freedom system'for bilinear-tr'iarigular pulse (Cl = 0.100. C2 = 300.)
.5
7' 10
TIt.. r Of YIELD TO LOAD DURATI'ON, tciT
\r.
t-lo
--;. ~
'0
• .1
"cJl
, ,
(J\V''' '0'0'0
~
.. _
N
__
0
o
i.I' :-".
OJ
VI.
0
>< F ><
z
0
1-
U W
-' I~
.
.;
,
'" '"
>0
w
~
w
I"
Z
o
0 ..... NN---. 0 . '0 . 0 0 0 ooa
o
....
UlO....,,·t--)O
•
v
..
o
'0 '0 C1' IJI \JI Vl
'0 '0
'0
...... • V" ~
'0
V'
o
0
...
...
<1'
0
]
111\
•
,.,
<9
.. '1'0
~O
V to..
r
"it1 \""f
~'\I W \ I'IJ/
~
,~M-hfi.-~
-
n-.
"-~-~m
~..
1",v
',j-
~AM
-
Cl
~
0
ir
w
s
1.'3
1.1 0.93 0.81 0.70 0.6\ 0.53
3
,
>< <[
p/rl ';'0. 0.7
uF~ .001
ez:
---
.002.003 .005
.0 I
J .02
k .03
,
V/A/"/V/ .05.07.1
.2
.3
.5
.7
I·
2
3
LOAO DURATiON TO fUNDAMENTAL PERIOD, T/T N
Figure 3-206
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (C1 = 0.091. C2 = 300.)
5
7
10
<[
,-c
Z
w
a
"-
~·o 1.7 1·5·
.1:"-
-J
t-
J: s
~.,j
~
"
0
'Z
J' G
U
W
....E
~«?
J: o
~\l<
W ..;
.
.tf'
1-
IL
z
t<,
"-
'.1
20
0
."
...
• \1'\ ~ J \ll
10
J
tf'
Cl
a.;
'0 "0 '" 0..
~ 1\ AI" ~, ,_,~
w
<,
000
=>
0 I-
z
0
t:
u
w. -J "w
,=>a , , ,
)( <[
"0
w· ~
e .
•
•
.
liME or YIELD TO LOAD DURAnON, telT t-J
l.1'
~
0
.... (I'
,.
v"
.... 0
-l
(J\
.
•
0
OJ
IJ""" 0
"
0
Vl
N
,-
0
0
IJI
.'
-
:.
00000600000'" •. -, . ... 0 0 co oJl IJI .. tv N'" - 0' a>
0
0
Cl
0
VI
0""" .. N
0
0
U'
0
IJ'
0
V'
~
IJl
..
0
V'
,
0
(1'\
0
~
...,.
0
..,.
0
N
0
'Z f-
.......
,..,E ~.
0'
',",
z
....I1b-..~t'-\'td:-N-I.~~
o f-
U
.s',j'
...'0'
~
-'
L-
- '~j>t~llUt\
U
a
,
w N
'" '"
,.,yl'>d~
c.
-'
w
.,..~
>-
o
r-.
z
a
'.
,
.w, ",I
\0
~
'tlli-t7 I
a
'"u, LJ
..J
....'Z
:f
'2
...
:J
~
\~~~
c'
~
U
u ~
L.
W
.If
o :>
I 1~\ot-*f'H~¥;rn-¥li'+:!J.'~ot=:'iPa;I;~
'\
:J
:> >(
...a
LJ
:f
;::
0,7 0.5
JZ2r.Lr'.:-J:::::TI~_.__~-'-,-
.001
,,002.003 .005
.01
,02
.03
.05,07
~ , 2 . 3 . 5 ,7
.i
I
2' 3
LOAD DURA nON TO rUNDAMENT AL PERIOD, TIT N
Figure 3-207
Maximum response of elasto-plastic. one-degre~-of-freedom system for bilinear-triangular pulse (Cl = 0,083. C2 = 300,)
'5
,
7' '\0
. .
" ~
~o
-0
~
"0'
,I .
nMEOr ·YIElD TO,lOAD DURA nON, CfI IJ' .. '0'0'0
-1
'cP
V'
o
,..,'0
....
lI'
t(/T
o
ooooopoooo'
ClD
OlU'fo.lJoINN----." OOON tJlo"""' ... N O 0
o
'"
.....
v I~
0
.. 0
CfI \JI
'0'0'0'00
0
lJ'I
"vtvo.~N VI. CfI 0 IJ' 0
V"
~
~
1\
d>.,.., ..>
w
>:
'"
<,
E
\\
><
.,..,
6'
.t
0
;=
'J'
I~ ....
u
w
i.,
W
a
IIIYf
0 -oJ W
0
D::
w
•
>-
l\\
I~M'1\tt\7"/51::lctrDI1[
0: 0
a
'S
J'.o
k
f-
Z
0 f-
7
-'
i-
0
::I' :::> ::I'
i
sc
" ::~,-, .001
.002.003 .005
• P /r ,=
~ .01
.02 .03
.05.07.1
.2.3.5.7
lOAD DURA nON TO rUNDAMENTAl PERIOD,
Figure 3-208
2
3
TIT N
Maximum response of elasto-plastic. one-degr~e"of-freedom system for bilinear-trianqular pulse (C1 = 0.075. C2= 300.)
5
7
"0
-c z
...
::> 0
t-
Z 0 t:
<.>
<·0
l.-
1.1 0.93 0.81 0.70 0.61 0.53
1\
o:
::f
<,.1
I.J
~
w
1.7 1.5
l.W
-oJ
Z
or .
0
W
a
..
'6'
-oJ
U
,.,E e,
z
..,, '" '" '"
·z
t-
<,
w
..J
w
a
::f
::> ::f
><
<{
::f u,
0
w
::I"
;=
e
Cit . ,-
·1
0,;
•
, ,,-
TIME OF YIELDTO LO~D DURATION.• ldT ~
~. '0'
,.
•.
v'
a\l1''''
....l
,.
.. r:P
0
"
0
0
0
~ t-.I r.. 0
0
U'
_
:..
~ o:a
!=?
•
?~? ~~~ o~~t?'o O'IUI
..
l"lNt-J,----
0.0
0
N
0\ -
..... ... N
0
~
0
'0 '" 0.1'
'0
'0'0'0'00
"'\,I4V't"'N
0.1' V'
V'
a
CJl
a
lJ\
v. z
>-
<, .>
'"
.'"
6'
z o
70
u
w ..J
l.. W
..,
...
. '" o
......, \?Iv
g O;
t:
"
o
,~ ""
,
..J
W
".
.
~..I~ IwMW ill \ II' "
o
.
>-
I •
. ...
.
. . . . ..
-
~
>-
'';-
~~ _---~
~K
z o t:
u
r
'6'
_. .
._ __~ _. .
'1I,ANH\z1'l-¥.JJt1
,0
of'
IP1\7
.....E o o
'" w c, ..J
« z
>-
w ::Ii
-c
~
~-t~J\1l>< ~/l'
::> l..
o ·z o t: u
>-
W
... ..J
·w
.-Vr\l \ / n
-'
u,
W
o
W
',~ v
.0
'::Ii
::>
~
::f
><
::f.
::> ::f
;<
«
:t:
...o
1-\
f;
w
:t:
"" 0.7
0,5 ,001
/v
I
J--
,002.00.! .005
,01
.02
~14vt11{7~¥/Y1 .03
.05.07
,I
I
,2.3.5.7
i :
2
3
LOAD, DURATION TO FUNDAMENTAL PERIOD. T/T N
Figure 3-209
Maximum response of elasto-plastic, one-degree-of-freedom system for bilinear-triangular pulse (e1 = 0.068, (2 = 300.)
5.
7
10
liME O~ YIELD TO LOAD DURA nON, tciT ~
soa
~; lX
., ...
'"t,
o
200 w
sc
'"
z
0
w
100
'" ex> '"
a a
-'
w >-
~o
20
0 f--
Z
0
10
f--
U W
'- fl.",
7
i.,
w
a
5
::f
=> ::f
J
sc <{ ~
\"'\i
2
M-JH:N'il-lv-'--., ,~ ~~ , r '
'i<
r-,.
_
'X.I .
,
" '\oJ)<.; 1/ r-; )("lAJ'.'
/»
.
~ ~r--
~A r /vN(.~
~
\~
.
-
At
...
I
0
.
•
'
>1,
--..J
~
..;;
~
.
,
I; .
•
'~
,
. 1 - . ,"
_."'t
1\ (\ (1-' \1 \. I""lll
/\
~
~I
I
.02
y
/"
.05.07
.
II
I' /
" 0' '~
'..I
.I -,
!' \.
v'
:.> .~'"
\j
X
7 K-r-..
>Q -2 (~/. .: ~ //.'// »:
'l/. 'l': ,/./,/
.t
~. . •..
~,IS. ~V
,2
,J
,V
• . r--'
L---
-
'W:V
-v ;
1
'
..:Y
'\
--"::
.
~
1\
p/r :- -~ '
.'
'.' .5'.7
I
2
J'
TIT N'
'Maximum response of e Ia s to-p l as t i c , cne-deqree-uf-freedom system for bilinear-triangular pulse (Cl = 0.056. C2 = 30U.)
5
7
\0
4:
-f
L..
s
::f 4:
z
=>
o
f--
Z
o
t:
<'J
u
<·0
u,
w
..J
/.,
W
1.5 I.J I. I
=>
0.70
\
-' f--
-0: 0
o.e I
.
'"c,w
n
O,9J
'
a
o
w
~'6
, . ...",
....E
v
.1. 0
N/
'1.1 '•
.
/
x:: .?~
X~ 'i
/\I/,:y'.'
1,/
~ , ~ J\ /~b~ 1/1'-.. '
'
V,j
-
f--
<,
Z
'0'
~ I\) \ "\..
1'\ y
7\.
LO"D DURATION TO rU'NDAMENTAL PERIOD,
Fi gure 3-210
\,I'
,-\,,f
.
'\
'V-
,/ ,OJ
'.
,/ v
../
--
--' '
\
1J-
A '\/
~
20 '(\~~ ~:x
' ',.,
S
.
.-:1\7
Z
d'
1\
-A 1/1
\ , /
"
I
"
I'~"
v0 7:v .~.
V ,4 e)F-'\V
II \
.0\
"
.,'
*",v:..\_ ~ ~ ). --':,
V
1~'J1i-
Y
~~
0
. ,
TI
',;-1 .
V"c1'OI.J'
I
-<;f';
?
ljvt
' . . • .~; .;J U f 1 '
. ~
V"
11\
',.
T.
<,
,\0,
---
,002005 .605
'I ell,'
lJ'I
Ii /' " II
L---
.
'0'0'000 ... ..... vo."t-J N
000 o:J (J\ lJ'I
0
N
(/1 A . '
~
j
I
'
o,~
.001
II
. -
\,
1'-/5l'A..I'&/"~ /"" '\ -:/~ '-, , /./ ' / ' ,/
'" 00 (f\~"~N""'-0_.0 00 0 ...... ...a N co \.J\
-
'(
-
/'r-~
/7 0.7
,/
'f..
','
If'-At, -y Y... )< r--1i ~ \ /.
~ ~
lfI
I \
',I"
, _, "
I\jt"+'--~'"
_
o
, .
Z'i.A . ~ •' J-l. ':)( -I \ ! '
-'
00000000000
N.... .•.
......
'
'" ,/
Z1~I*/I\ ~.
~o
-'
w ,
.
...
7\--::,- :x, '/1i. K (
r--='J~...t~ 1
70
r.,
w
oJ'
).kY .''f' :~/~..l-..j~ : x: :'Y.-4,>I,~ 1., ~;:. I " , ' I~:~~;~~~~~ .' '. I)
;:.:
o
.
(f'
'IT
<,
E
.
-l
\I ,-/ \ ,,-'
o·~r
500
,.
If'V'OrSJoooo
a
::f ::f
>< 4:
::f u,
o
0.61_
~
O,5J
i:
e
e '-.\;
i-
~
,,':':
tJ
.. ,
•
"
·,1
TIME
~ ~ ~ ~ ''0
'.
or
hiT
YIELD TO. LOAD DURA TlON, O
...1Cf\"'V"f"Jf"J-_ 'd!o.. '0 '"" ~:...t
,
.
0000000000
;" U. ... ~ N N·~ :"0 i.". :.. ca .. o OOONUIO
500~
~
"_
~o
0'0'0
'0'0'0'00
~G'\J'I O. IJ' V'
~
... ...,..
~V"t-JN
0
l5'
0
IJ'
~
.,.,
dO
=¥j.¥l--\,tj:,W:--,1 Y1\
~
'" 6'
',.,
z o
t:
~
u
w
-'
L. W
a
1t1/V W
I y".....:7i
1\
-s..t
I "L:JKI"'_~"""*~V
,-'
'"
a >z a
>-
t:
u
w
-'
L. W
a
7 ·5
:J ~
3
>(
~ ~.
~
<'Ii <.J
.
\YJIIXIJr
-\/'1¥
~
IX
r~1$'~~~
p/rl
IL
00 I
,..
I
'r
I
1.5 1.3 1.1 0.93 0.8 I 0.70 0.61 0,53
\t\I\'
2
0.7 0.5 ,"/ ,
1.7
~ v
"'F ,-; I t;:ao
.002003 .005
Figure 3-211
~ :
:
;
,
i
;
,
;
~
i
:
:
;
;
4 I I
';""0.
I
,
4
"
.' 5
i
7 , 10
w
-'
« z
2·0 ,
cc c,
0: 0
.1: 0
~ N~\lK
a
a
r
.1: S
l!'-~
w
....E
>Z w
'6'
a
...'", '"
n I JtL
Z
>-
<,
~
a
::> L.
a
>-
z
a
i=
u w -' L. w a ~
::> 2'
X
«
2'
L.
a
w ~
i=
. ;
TIME or YIELD. TO LOAD DURA TlON, tdT 00000000000'
\,. ~o
". ... d'v'
... 0
....1..n.~vlr-J~,--·······.·· •. . "!. . . . . . :.. Oll l1' U' .. l.N N N .... - . . . -
dJ
0
\J'
~o'"" ...
..
_
500
k.7T7rN \71-,kY~ 'I.,><::::::r ,~ Tilvn7B VI I/~
0
OOO~1JI0~~NO
..
VII t.l\11\Y\1l-11\ J\.-
-~~ ~ h~~f~J; . tIJ:tl¥XJ ~
><
-~"t7.1rJ'~
E >< 100 .
z a
""" ...., ,."
w
--'
u,
-J
T~\l\
o z a
>-. >-
u
w
--' c,
5
~
~
><
.....E
.
~"
~
."
II 1
3
~
W ~
a
l7
z
jt-
:J
!).
4'<"" ~~ ~ ~"" A". • ',,, ,
.. WHJ~~
- -- .. _.. _.
_
-
.
---
f0-
.........
V
IN".
yj
.I?"
i=
1.7
w -' c.. w
a
1\
:p Ir l ~Q. t-fft
.0.
f' . ..02 .03 c
1/
r v, r .05.07
----
I
l'f_~_..L....
.7
I'
2
l·
~O~D yURA TlON TO rUNDAMENTAL PERIOD, TIT N
Fi gure 3-212
Maximum response of elasto-plastic. one-degree-of-freedom system for bt l ineare tr tanqul ar pulse (Cl " 0.042. C2 = 300.)
5
7
10
:2:
1.1
:J
0.: 0.81
0.70
.005.
o u
'., 1.5 1.3
J...
0.7~ IIIII~I~ O.5'~'1 .,r:.
o z
l-
rvJ '"".
lLV
.002.003
u,
Z
2~~~
.001
'"w l-
... ..,.
~~~~~11 ~\l v
a
o --'
- . .~~;,~
>-
=>
Z l-
<,
·~·-c-~~~a
~~tI'\/t'~j j>'lJ~~
w
a
.. --.1_,
~~IV
w a a
w
V4O"OU'O
" WI 1Ar..,rt. 1>1 I I I"UI I /
i= u
...., .... o'"
'-U'vlNN
YW ~
II
"
'0'0'0'00
\l7 \Y -lJlI:::I
~
1)
-.l 'n!'c'l'1l·.vI-1\-CA+MI-
w
.. QOO cP (Jl IJ' 0 \JI vl
:2:.
><
:2: L-
a
0.61
w
0.53
t:
:2:
e
-
•
TIME or YIELD TO LOAD DURATION, triT ~
v'-
....1
" ,, ; ,0 . .\J' ~ v'-
~
IJ'
~
':'7
00"'"
"..
.- l
.:
.N
~
'?
~.,~ ~ ~. ~ ~ ~. ~o 0
'0
l.D..,JlJI"VI~""N--""_fSJ OOlJlc.l',,-,o\1l-
...... ~
01.1"
'0·0
~G'
00
'0 lJ' 0
'0 ..
t-J
'0
'0
'0
V4V"t-J IJ'O lJl
(, N 0
Z f--
<,
-'
'" .,.,
t z o c: u
W
..J t.. W
..., • ..... '" .....
a a
6'
70
-s',r
.\t~
v
'6'
'I~I"J~
....E a
o
'" W
0.. ..J
....-< z
w :If
a
z
..J
::>
>-
o
l..
w
f--
o
~
j-,
Z
o f--
Z
o
t:
u w
'-oJ l..
U W
w a
..J u, W
:If
a
::>
::>: ::>
:If
;<
::f
-<
~
:If
::f
w :If
;::
I~II~ .01
.02 .03
~ .05.07
.1··
.2'.3
.5.7
I
.2
3
LOAD DURAliON TO rUNDAM(NTI>.L PERIOD, TIT N
Figure 3-213
Maximum response of elasto,plastlc. one~degree-of-freedom system for bilinear-triangular pulse (Cl = 0.032. C2 = 300.)
5
7
10
,
,,
'.
TIME or YIELD TO LOAD DURA TlON, trlT
". .....l.
". ""0 ;v'
~
oJ'
~
t.1'
~
.,-'
r.0
~ ~. 0
Vl.
-l
~
_
.N
?~~.~c;>,??
~
0
sao J aa
~
\D
~
IJI
0
l,II
Ut
IJl O'l 0
lJl
N
N
~c:'
-
U' -
-,.
C?'o
~ ~
-'
-
0
01
'"
'0'0
~ 0
Cf'
0
'0 \I' 0
'0'0·0
~ 0
V' t.1'
~ 0
'0 l'..J l.I\
---
\ I
1~f-jJ~l\Dj~t'l
-a
zoo I-tIT~·~mw
\till
.,.,
w
g
u
w
--1 u,
w
...., .... '" '"
o 6
f-
->
....E
100
g~~~~~~~
.,.,
6'
~31~~~~~~\:~,
70 50
o
•
• ....,
If)\J a l-lIf.L1"'~'!\-'J>-..A-I'-j\,'!G
~
W
20
10
a
::£
J
s-c
..:
::t.
':0
0
2
l~
~
:;: o.
:;7"" '.
;:,:'1~ .001
l--
J-
,002.003. ,005
-
.0.1
• . 02
.O~
.
. ~
~1/
..'
...
'
. .05.. 07.1. .'
.2.3
.5.7
2
3
~.'
LOAD DURATION. TO,rUNDAMENT AL PERIOD.- TIT N
Figure 3-214
"Maximumresponse ~f elasto-plastic, one-degree-of-freedom system for bilinear-triangular pulse (C1 = 0.026. C2= 300.)
5
7 . \0
'Z
::> c-
~o
o fz
<'Ii <.J
i= u
1.5
5
::£
::>
....Z::£
<.0 1.7
J;;
7
-oJ
..:
f-
':6'
~s
-oJ u,
w
'"w ..:
r-.
u
0 Q.
S 'J"
)-
,-o z
"0
17
-oJ
~
z
d'
';
-c
)
0 !'J 0
0
......J
.......
o· ::£
I.J
::>
1.1 0.93 0.81 0.70 0.61 0.53
>< ..:
::£
::£
.;: 0
....
.::£ t:
e
e
•
TIME OF Y1£LD, TO'LOAD DURA!IO'N, t[/T '.
6
~
'0
-'
oJ'
'oJ'
!"o '-:. t7. :, 0 IJl
'\f'
~
o
ce0
N
.
0 '-.I
0
.1'
0 0 0 0 0 0 'VI ..... 'vo ~ N N Vl'O\
lJI
o'U'\"-
0
0
0
0
:-'"_ l-l .. -~ 0
'0
cP
oJ'
o -\
'0 (J'I
00
' 0 ' 0 ' 0 ' 0 '0 l1'~" V' VI
Ol1'O
l1'
'0
'0
~ \II
0
'"o
,)'-
u
><
<,
' n ........' "
_....,
• .. "\"V\ .JJ--Yi'. I I\I~, ~
"U
E
•
>< z 0
/\
f'"
.~~_.
u w
-' u.
w
a a
'", ..., '"
'"
-' w
;;::
0 r-.
)lJ
,,\V
Z
0
IY1'U\I' V I L
~)1.,lIXI V
f"
u w -' '"w
.J-17I.
/\/.1.-"1A
I\vr-,J\.
I AV
r/~
a :i
~
::::>
"
>
r-----l: 0,5 .00 I
.'
~--V ,00200'\
,00'5
>'
\ - I ..
~ ,0\'
,02''',03
,05 :O}"
~ ,I"~;
{,2
,3'
,5 ''',7, "r-l '
.!",2
- '-te. ,
,3'~"
,LOAD pURATlON TO FUNDAMENTAL PERIOD, TIT N"
Fi gure 3-215
Maximum response of elasto-plastic. one-degree-of-freedom system'for'bilinear'-triarigular pulse (Cl = 0.022, C2 = 300.)
5
F
\0
'-"',",.
U",-
"
TlM(
.-
/'
:...£
/'
V"
/'
0
~
tJ'
'-!'
1.1'
!'
0
~,~ 0
vi.
C>
500
~I
'or ,Y,[LD TO LOAD DURAnON, ;0
"...
.- - . -l.
>
~~~~~ C?~~C?'?~ ~·o 0 0 ' 0 ' 0 · 0 ' 0 ' 0 ' 0 0 I.D -& \II ..,. \,rl toJ ,toJ ~ ~ -....._ ... "... ('/J -' G' 0.1' • '"' V' V' ~ N 0 U" lJl -I I.D ... o. ~ .. ~ o· lJ' .0 0 ~ ~ !J" 0 !J" 0 lJ' 0
N 0 ----'------_
\ 11\\
'-I>oJ I
t(/T
_.
....
.1
1
...
"
_
1#.J...A:'1'~
l /\17 \ k'\.Y XI}
11
IINI' - ,,-
Xl\ : "'.:....:.:-::••: '1'0.':': . •:~ t
~:
:/",
~Jr
d' ',,>
....E
,'.
o
o a: w 0..
z o
•. ,., -
,.~
... _.
'..J
. .. "'1
«
t:
I-
u
z ,w
.:.~·'l~N
W '...J u,
~
W
o
, '" N -.. ...
'Z I-
<,
-:0
II
c ....J
J:,s
w
'TT
;..
o
I-
Il-
\7
Z
l
•
o
j~
J: o
<'8
<.J
I-
U W
2,0
x
...J u, W
Ii
o
.~
::> ~
><
.
~ n
,-p~
.\
,......-r I,y
Figure 3-216
v
'1/
'/1 "
~
'1.7 1.5 l.3 l. I 0.93 0.81
0.70
1\
0.6t 0,53'
"
'2-
'3
5
7
10
«
o z =>
...o
'I-
Z
o
i=
,U
-w
...
..J
W
o
~
=> ~
;<
«
~
...o
~
i=
-
e TlM[ Or.YI[LO TO LOAO OURAnON, t(/T , :..J,
,;
V'
"
0
-1
'\f'
l.1' "1.1'
.. '0
V' '0
.... - ~o.D l? ?V' ~...... ? C??? C? ?' . '?...... '? ~ ~ ••• -.1 ""'Nt"" .. _ 'v"
01·...
-.
500.~ \17\ )I( AI \A .
0
0
CJI lJl
-.I
0
0\
~
0'
~
•
.,
~
0
"0
dJ
Lf\
'0'0 -' Cf" 0 0
0'0'0'0'0 "' ...... V' V' 0 lJ' 0 "'" 0
"0 t.J
0
N 0
lJl
il"ll I )
Z
....
d' 'r'
w
'-
....E
~
',.,
i<
.'-
E
~
(;-
><
'r'
0
~
~
'J'
. .
u
w -'
.If'
a
...'" '"
'0
ci' ...J W
J: s
~
~
0
....
J: o
..·4-\l'ttrlfMJi:k llThI
0( 1\ lWJI
z
<.& <.J
0
c: U W ...J
7
W
5
u,
a
::>
~
4: ~
.3
1·0
~-
""
~
;(
""
w
.n.
z
w ,
c 0
l*~~){f\/~
1.7
~
1.5 I.J I. I
0.93 0.81 0.70 '0.61 0.53
2
. "/
~IY
.002.003 .005
J(;v
.;/
.01
·A),2 .03
~ .
.05 .07.1
.2.3.5.7
-. '\
I I 2
3
LOAD OUR!"nON TO rUNDAM[NTA!- P[RIOD, liT II
Fi gure 3-217
Maximum·response of ela~to-plastic. one-degree-of-freedom system for bilinear-triangular pulse (C1 = 0.015. C2 = 300.)
5
7
10
...J 4:
....Z
w: ~
4: 0
t< ::>
L-
0 ....
z
0
~
u w
...J
::> :2:
><
4: :2: L-
.0
w
:2:
s:
TIME or YIELD TO LOAD DURA nON, tell ~
-0
~
'0.1'
.r 'l1'
!'
.... '",
.,.. 0
'!" ~ ..... 'so- ".,.." co
C? -.I.
~? 0 VI
.. -V"
«?
C;:>.
VI
c:' 7' '? C?
.~ ~ -~
....
~~
'0 '0'0'0'0 '0 "0 0 ..J. G' If' ~ .. V' lJl t.J t..l tS'OOO\1'Olf'OlJ'O
'?"o,'o
OOU'l,II-.&O"'O~"-
OOv'V-
"'0
cP
IIJ-17 H1V1\
W1KLJrAi1\! X
300.
/J.-,.A-1o'tHltrt\L....:l..
V,
200 w
><
E
>;
100 .
=
z
'1
0
f= u
'
'
~
W
-J u, W
0,
, N .... aW
30
..IT
0
-J
w
20
'"
0
~
>< -c ~
I7lr'Y~"
10
....E
....
r«: '" ,\
,s
..)-
A
y~
I}\
I\'N X / . ~Cl :v ~ io'\ ~ \\i l\
"I,
v
'6'
1)<
.
J:,s
'/'
V ~ ,,,\
~
5
.1 0
t'-,V
<·s <,.1
.v "'lin-..
<.0
~yy
·INJ VI-.....u l:nt'l 1.7 1.5
1.3 \.1
3 2
~ P Ir
I
. 0.5 .00 I
/'
....--r
IY
.002.003 .005
~ .01
.02
-,03
-.
.,
v .05 .07.\
O.
:0
'
~x.//V .2.3.5.7
\
2
3
LOAD DURA nON TO rUNDAMENTAL PERIOD, r/T N
Figure 3-218
Maximum response of elasto-plastic. one-degree-of-freedom . system for bilinear-triangular pulse (C1 = 0.013. C2 = 300.)
5
>-
c.
0
'"a,w -J
>z w :::f
':0
. -1\
7
~
:::>
~..."
-' 6'
-VI
..v
z· f= u w -J "-w
",1.'11:7 '1111 \J7
,"-"
0 r0
~ \
<,
'",
~
<,
z
d'.,..,
7
10
c
z
;:).
"0
>-
z 0
f=
u
W. -J .
'"cw
:::f :::> :::f
><
-c
0.93 0.81 0.70 0.61
...
0.53
f=
:::f
0
w :::f
e
•
T1t-li: Of Y,IClD. TO lOAD 'DURA TlON, ti:.I.T ~
"0
~ '1J'
..ro
'If'
p
...,..
'0
'0
~
~
". c:' t?-.I
.;.
'f"J
0'
IJ)'
o'
t.J
0
~ ~
~
\I' U'
U'
,
«?VI t?..,.. t?~ t?t..J t?... C?, C?, -./,
0
..
0
~
'"
•
~
0
'"
o·
'0
fP
\1'
0'0 ~
-1
00
00'0'("1'0"0 !J''' ~ v" lJl t-I O\1'OlJl 0 IJ'
"0 N
0
l-
...
~-
."
.--
·w X
~
<,
E x
z
0
tz u U
~T'-
V
w
-s.
50 P!'-j;at!
u, W
a w ,
IV
"'"
"
50
~~~:litit"Xt)l'
...J
W
>0 0-
Z
0 t:
u
W
."
6'
70
...J
0
'"
W\(
20
tt-*t::Yt-k'i\'
v
.~,~~r1tJ'9A
",mt~H~~~
~
,"~'rIkl1l
'or
1\/ V \ ,{\ r;..
y
~
\1~I\
":0
J:,s
,\AAtoy
.1. 0
I~V
J.t> J'J
2·0
7
1.7 /.5 /.3 1./ 0.93 0.81 0.70 0.61 0.53
...J
l.-
W Q
5~~fr~M
:'f :'f
x
v
~~
::J
f\
1\.Vl:
2
~ ~ ~~AX~ "
0.7
~
0.5 -: ' .00 I
'\
p/rl ~O.
.
-
,
v
/'
.002.003 .005'
.01
.02
1/ .03
.05 '.07.1
~ .2'.3.5.7
1
2
3
lOAD DURATION TO fUNDAMENTAL PERIOD. T/T N
figure 3-219
) 0
2Q: W Q.
...J
Z
":6'
.x
z
Maximum response of el e s to-p l as t ic , one-oeqree-nt-rreecom system' for' bilinear-triangular pulse (CI = 0.010. C2 = 300.)
5
7
10
w
:'f
Z
::J
l..
0
0-
z
0
;:: u
w
...J
l.-
w 0
:'f ::J
:'f.
~
:'f l.-
a
w
:'f
j::
TIME 'or YIELD TO LOAD DURA TlON, tr/T'
~
..
'b
vl r-J 0'-"
o
~OO'
;00 200
"'
"E " z
<,
0
i= u W
100
J''l!r:>l.~
~,7\\:A
70
.."
'"
20
0 >Z
10
W
7
;::: u ....J
"-
W
a
5
:t
>< -c
:t
3
2
~
11.', ?'AG, a2
~
~
L:' '"
.:-
~·???????OO?Oooo
:"'"
¢l
~LLL:
l::ll' I'
.t-
L: K
KI"V~
.
7'
.I'
r
t>~
-
v
'1
-
'1
r:
>Sot:
IL
,
s..
~
:S;::
~
,
!W
t:9.-
~ x:=
~
,
f'/
_-
8. t:L.l.
~
tJ UII
,.l.
L
....l,
~ ~ I;:S
l.l,
...l-\....l,
~....l,
....lj
Ql.
~/r = r8
'-' ..>
0.53
~.....,
::::::.. L
0.7
-
.-
0.5 .001
. ,
.002.003 .005
.01
.02
.03
.05.07.\
.2.3
.5'.7.
1
2'
3
LOAD DURA TlON TO rUNDAMENTAL PERIOD, TIT III
Fi9ure 3-220
Maximum response of elasto-plastic. one-degree~of-freedom system for bilinear-triangular pulse (Cl = 0.909. C2 = 1000.)
S'
7
\0
cc w a. t-
7. \
_ ff:d.
a
0
....J
6.2 5. , 4 .6 4.0 3.5 ,.0 2.6 2,3 2.0 \.1 \.5 1.3 \.1 0.93 0.81 0.70 0.61
if
....,E
\8
"
I 1_'·
2 I-
<,
\5 \3 9. 5 8.2
E:
F"I="-
_
65 56 48 .\ 35 30 25 2\
.~
1:.r
uz ~ "1m
j2~i{:. ~
'" o '" ..
w
L:.J
LL l..=: I" ,...::
00
1.0 ....... l1'''' .lH ...... N N -0 - (lD Q\ lI' .... OOU'lV'lU\O.,.OOllH OUIOO
et-.t.-A...l
/r
[: -j(
1"'>K.:t-
K
OUlN
'
(V ( I"::: f/ L2'
,
~
.....
AlA
~".~ t< ~
:t :>
r-.x
~
.L
Il'c
1(,
~\ ~ ~
- .
~
It
.' f - - "
~
,-.;{,
II
0
i
[' _V ~ I ' ~IJI X ~ \(
30
....J
>-
' '.
f-) It},, ~ "
50
....J
W
.-~
,;.\ ./-' -r-
W
...,,
~
l'I , ..f'. 1) ~V I
6.
'/...
~
U'UI
~~~jY,bIC. L 1", I....:::' ~~ ~~ l" . ~ 1..,2 . ' l~ -IfAr'"" - ~ . l( 'I; r - ~~.'. ~ ,..., ~ t~
"-
a a
4>
.-¥---::li.~ , .A
o
__ ~ 0\ 0-10
N _' _ OO'Vl
"
w
:t
a
z
:>
u,
0 t-
z
0
i= u w
....J
"w
a
:t
:>
:t
><
:t "0
w ::f
t:
;
e
e
•
TIME OF YIELD TO LOAD DURA TlON. t(/T l1' 0
a'
o
...
o;-J>
vi
1.1'
V"
0
N .....
N 0
(J\
-,lcn
VI
-. 0 - -.....
0,
tv
4_
500
j
0O;r;,=t~~5ft~~nI
."
E
sc
u
w
--'
70 50
L.W
__
0
..., '"
'.J
,
N
~Wif>\t:Zi:~ lQ..!
N/ 1/ ff:>o
,. Z 0
c::
u
w
--'
~
10
0
:>"
Q:
W Q.
..J
•
w
:>"
o z
I
:l2
::>
L.-
o
tZ
o ;::
Ie
u
5
.I
..J
L.-
w
:>"
><
o
o
W
7
::> :>"
·E
·oJ
l[;:;
L. W
t-
~.
~
--'
r 0
Z
<,
IS'
0
w
~ ' '"'
t-
iii
VI
00 '" N
][;
VIH<
<,
z
°OUlOO
r-....VINJ
w
0 t:
~ 00 0 0 -oaO'lUl"
OQOOOOOOOO I.D ...... t.n .. l . H U i N N - 0 0 l.n U'I ...... 0 VI 0 en .....
........ N N - - · · · · · · · · · ·
•...• VI lI' ...... 0 Ul
~
o
,~
=>
:>"
~
><
n
2m;' .
rrr
A-
I .6:
:>"
!!"
r-,
~
:>"
L.JT'i
L.-
o
1\ p/r =cr.
\ I \ 1\
w
1\ \
:>"
t:
0.7
0.5.L .001
I
I
I
I I I I I II
.002.00.1 .005
.01
I
.1 .02
I
I I I I III
.0.1
.05 .07.1
1
J
I
I I " III'
.2.3.5.7
1
I
I
I
2
3
I I I I III
LOAD DURA TlON TO FUNDAMENTAL PERIOD. TIT N
Figure 3-221
Maximum response of elasto-plastic, one-degree-of-freedom system for bilinear-triangular pulse (C1 = 0.866, C2 = 1000.)
5
7
10
TlM£
'"o '" 0
~
V'
V'
o '"
~
""
0
N
-
0
Cf'I
~ 'Xi. \11-
-
~
or YI£LD TO LOAD OURA TlON. tclT ...
-
:-t.CJ'I
.
0
""""
UI
0
NN ~. ." UI· """"
-.IN I/'" (\\)
~/~''j[I~ v " ......... ~I\" '"
~c\l \'clA 1\
- ~lli'tHIN+---M---J!~PI
"~I~I~
r, >(
<, >(
z
o LW
Cl
c-
CD
VII \)qj
YNJ V
N~~ II "1-.11
VI .. 0 0
'" o ...
00 ~
v
[)(.
b(.
~
1--1 V iL
~
k;:
CD en 0 UI
_ 0
z
>-
<,
E
'oJ
~
o
a
'"w
~
--'
'" o
lH.
UI ... c.... '"'" N N O"U1 UI " 0 UI 0"
0
--'
u
a
_
-
l>.
w
,
• '" ....,
70
>-
w
O~~~OO???O?oooo
'" ..
~
'AlIl!l~
E
0
W IVII"
--' w
~
o
>-
z o
>Z
I-..
~
w ::f
~~
.. "f',
Ii:::>
... a
~
F
>-
z
a
i=
u
t:
W -J
u w
--'
,~
w
o
::f :::> ::f
IS<:
x",'
3~
::f
2
I
0,7
~
~., -2:."
L-
W
c
0,5 ,00 I
~
B
~~
.~ '----=:
::f :::> ::f
%
\ I\ I \
~
" .002.003 .005
,01
,02
,03
,05.07,
~ ,I'
:2
"'l"-
,3
,5.7
>(
...a
.\ 1\ l;2.
,w ::f
~~, I
2
i=
3
LOAO DURATION TO FUNDAM£NTAL P£RIOD, T/T N
Figure 3-222
::f
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (Cl = 0.825. C2 = 1000.)
5
7
10
e
..
e TIME or YIELD TO LOAD DURA TlON, tt/T d'
o
.s'
... ............ r..J ot.Jl.1'OIJ'
N
0
.....
en
"'LolNN--
'" '" r\~
w
200 '<
~
D'
<,
><~ z
o
100
70
~~~~~~~~~
~ ~
~
W
...J ~.
W
o
,
N
....ec
.
.
0000
.
.
NN-
!=>~!=>~? -
-
00 00000 CG O'l
Ul.\Io{t.I4N
VlOal
(.H
0
OUl
00,",,0,"'_
~I/
k
~Y>
NIA1~
II
~,
rM
,,'>
.~ j,f~ . /
').'>
~
o
...J
w >
,Y> ,'>
4."
~~~8mJ4.0'
~~tl4if1lj 3.~
><
2.
:::----: 1~
_.
,
1.3 I. 1 0.93 0.81 0.70 0.61
~ \ I '"
p/r~ = 0..
0.53
~._--
071~-1 0.5 .001
1.7 /.5
p
~
~
I
=
.00200J· .005
.01
.02
.OJ
.05.07.1
.2.J -
.5
.7
I
)
J
LOAD DURA TlON TO rUNDAMENTAL PERIOD, TIT II
Fi gure 3-223
'"wZ ::f
< c ·z
::>
....
...oz o
....w..J
1/
.I
...J
<
<·0
v
__ _";;1 . ~ .~
w
;::
o
~
2.6
l>:
J..-tr/lloo ~-t-tli 2. J
I;l;'tlIt'
c.--1
W
.•::f
3.0
1I'9o-1f~l:;flii
w ...J u.
<:
q
.').
o c: ·u
c
0..
" ?!':3~~H<'MI1bf19 s ~
Z
....E
,"
1·\
t5
.:
<,
o
"~.')..'>
e·
::f
').,
IV
o
::>
~()
}'
f= u
w
~
~ ~
w
c
-o
o~ou.lJI:""~:""'''' ~\D""-IU'I"'t.oI o 0 11I UI lJ\
'Maximum response of e I asto-p I as ti c , one-degree-of-freedom system for bilinear-triangular pulse (Cl = 0.787, C2 = 1000.)
5
7
10
u w
c
::f
::> ::f
>< -c ::f
.... o
w ::f t:
.~.~
tdT
mole or YI(lD TO lOAD DURA TlON, ...",
o.J' 0
o
~
....... 0
0
I",>
0'
IJI
-
ao
-
-
Vt
0
0\
..
lA
N
N
~
0
U\
Ul
.....
0
500
I
If ...... tl ,
300- I
200
><
<,
E >< 100 z 0
,-u W ..J
L.-
w'
'",
N
'" N
a a
..J
W
>-
50
I'" ' •__, '
,\ ) l ",1\21, \iX K ,,'y1\\J1 .11 f'I' r-
~(~
7
W
5
r.,
a
"
::> ::f
3
sc
::f
)<\V -_. --- ,,"r ~W7 -
W ..J
2
•-
...) "y<-, ""
,~~"I-': -~', '" ~
,~ ",.
d
~ I~
~' 1"L'< I') V' / . V I" _ I': 6-.'''I'> V
-
.
IIIN. I'
0
.L
L"
1"""--
~~
c: r-
I)'
I'"
f<:
~ I'< l> ~ B -,/ K
~
f'
~ k' f?: ;.;:;
\;:=-
~~
->
\
_
,
'"
V
{' .~".
Jf
_
............
0
CD 0 "
""
'j." 'j.\
l-'
;::
-« 12
....
~r'"
r-
r:
\" \\
c o
~."
'b.1.
c,
w
I-
A. EJ
A.O
2::' I'
3.5
L.-
'9F
2,6
2.0 1,7 1,5 1,3
"-'"u, <" ,,,,, I I
t'::'1L
\ \
~
..1 l 1 , ..}. \
\ \ -"'-
"
I. I 0.93 O,BI 0.70 0,61 0,53
~. ~
~P, ~ Ir
=
';' 'tl..:.!!
,
Z
:::E
~
0' f-
Z
o
t: 2,,]
1/
~
~
it:
EJ .1.
-r..,....ylrh:>t<. 3.0
m:_
w
.,)
~~~ I"~
t:: t:
Q:
f-
It
~ I-'::
I:-
•
....E
,.\
...
Z f-
<,
\'b
,"
,
..,.
rs »: k"
~
0
VI.
I.'
I-' I
_. \
0
~\
_I
\
0 VI 0
V "r:J 11lI\li'ID7'i+h#=#tt1j1~r1
-'
~
0 0 ClIO'l 0'"
l--' I-'
t:
~ !>: ~ k:-:: 'f" , __ :t:; p
0 0
~,
.",<~ ~ , "t:", V
0 0 0 N0 en',"""
~
c)II
'l'
?
N UI
i><:::=: ><
. IX
E.E r-..J,
W;;; 1-'<'"
0.7
f'\
0.0 O? 1ft ... lH 0 1ft 1ft UI
U) ......
N
"'.
1I [.",. ,.', 1\/
20
.~
_
III
- - \ . ,
IJ':: :+~ iLl' 30 th ~ V l\L!'" 10
1/1"
"I~" I'l J (1/, II' 'V h(~v\~IC 1\
f=
u
"
-I\f::: .t\J~ --It"YH, !-:i 70
I-
Z
X. \' X. \"x" - -1\ XI)<
,I <:I!
0 0
- '/\
1\
l;'tj.;)( J;~t)< ~\I\
,1"w
" . \1
V
_
••••••••
u
w ..... w ~
c
:::E ~
:::E
><
::f
.... a
w
:::E t:
,
0.5 .00 I
.002003 ,.005 .
.0\
.02
,03
.05.07.\
_2
,3
lOAD DURATION TO rUNDAM(NTAl P(RIOD,
Figure 3-224
.5.7
I
2
3
5
7
to
T/r",
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (C1 = 0.750. C2 = 1000.)
e
-
-
•
.'
nMe or YieLD ro LOAD DURAnON, tEir '" o
.s0
.,. ,..,
p
o
o
N
-
-
--l •
-
0\
....
.'
'LM
0·0000
N' N
,"
'.
o(1\VlO~OUlUl"""O
IJ\~
ry •
~:""Ul;"'~
a
?OO?oooooo
N
N---OO(J'l U\~ ViN OUlNOOUlOOO"""
OOUlUlU\(h
if
8
l;><
. 'IV 'l\ 11\IVN/]' r-,
..
-
--
.. -
_...
~
~
-
....2
<,
...,E
II:"
v
.
o
o <>:
w
00-
z
o o
~17
'" w
...J I..W
...'",
'"
o o...J W
>-
30
t\'~~~~,<:0lD
20
r-,
o
...J
II
~
o
'" w
...J
o
~
~
«
o
;z
::> I..-
o .... o
k-Il7'
'" lJ W
...J I..-
7
)l'
5
w
'P:S
o
~
3
~
2
~I\
><
~
z
~
::>
II
et
..ex
,~
W
c: ~
;+
~
z
lJ
....
·I~
~
::>
~
><
«
~ I..-
o
W
".1 \
~
d(L~
'\".10..."1-.,
I 0.7
0,5 .001
.002.003 ,005
0\
.02
.03
,05.07
.1
.2
,3
.5
LOAD DURAnON ro rUNDAMENr AL PERIOD, r
figure 3-225
.7
I
2
3
Ir N
'Maximum.;response.of elasto-pl ast tc , one-degree-of-freedom system for bilinear-triangular pulse (Cl = 0.715, G2 = 1000.)
5
7
,0
'"
nME or 'Y![LD TO LOAD DURA TlON.
'" '"'
V'
0
0
0
500 0
~
1 -> 1---,]
,.., ,..,
0
<.1'
O'\~ ~( ';t _~~
E x 100 0
...u w
70 50
.J
"W
..., '" CIO ~
0 .J
u
w
30
. 20 ' "
0
::f ::> ::f ::f
I\. /v"
I~ ~ I
.. '" \0 7 5
~ I,
3
.
0
2
0.7
" ,,0
'"
'" 0
lJI
0
~
'" '"
o '" '"
0
00
N
.....
o.
~ .~,:;"J,,~
00'
. .00 o 0 --0001V\'"
NO o
r-,
>
',I) 1,/"'
K
f>
~
,e-,
v
1<:"
O
O
~" ~ tz',
,
L
0
.....
0
R
1,'-'
'\., ,'0
V H ,'>
. ,',I" .
"
"c-
c
L
'
<
t-
~ '\ ~
L:
.002.003 .005 .~
.01
\
"
.
.. .02
.03
w
Q.
..../
4.6
w
4.0
a
::f
-c
::>
v
.1. 0
...z
<'6
;:::
.
17.
. p)r
Q:
J.5
l.O
. ,
\
0
"
./
--\
a
'/>.'1. 1·\ !>.'1.
<.J
.
....E
'Z
r,
.
...z
<,
,\."
5. 0
.
.
..,<::>
'\."
I::
.
0
'" •
t-
.
~
UI
V"'-"
~,')""~~. • -
R1:>'''-
-
0'"
' " II' "",'
.~~ ~
I
0.5' .001
000 .~
I)< 0<
b%> l!'
~t -
~
/.7
' /·5
I.J
.
1.1 0.93 0.8\ 0.70 0.61
'W
~ r--' I~
"
0
i.II
b ( ' ,II r-, "
~.,- ~ ~;
'\75 .: -
l/
t-,
.
'/'
~ ,
x
«
~. .
1\
.. _"
y::<>"
.J L,
0-'
.....
,~ ~
et
~i
...
z ;=
.....
'" '"
'"
.
,,X" '" " 1\ f\ ,C;'I>(J ,0" 1.1 , i7 -
0 0
11
" "'-'to
0
w >-
'" '"
if~i\ljZ .V II"
<, \
;..._ ,
><
<,
z
0
... "" • 0
0
\ Ii K D'.- " IX ' " -- '. k -, 1\ I.";:;' 1\. 1 200 /: ):::-; I) f\" l/ " JOO
w
~~
, /
a, <;.
t[/T-
=n.G=
0.53
"0
0
u
w '-oJ "w a ::£ ::> ::£
x -c
::£
"0
w ::£
;:::
.
-
..
w·
.O~ .07.1
.2.3.5.7
.
I
2
3
5
7
\0
LOAD DURA nON TO FUNDAMENTAL PERIOD, TIT N
Figure 3-226
Maximum response of elasto-plastic, one-degree-of-freedom system for bilioear-triangular pulse ICI = 0;681, C2 = 1000.)
'e
e
e
•
nME or YIELD TO;LOAD DURA nON, t(/T
o
oJ' 0
.
o
V'
o
....
....
oJ'
0
a.
~
0
~
~.
...
0
~ ~ ~!=' ~ ~ ~ ~ • (.lot N·N - '... u . . N N ~ln~oi..nN tD-..&UI o 0 VIOl VI (J'l 0 G\
00
0
0
0
VI
Oll' O'l
••
~o
v ~,.
"'..
00
0
00
0 '"
1/
:z
0-
<,
.....E
w >(
<,
-I!"''fb<'L1 II" I R L i 10
E
n
o
I l
a: w
>(
e,
z o
t:
j,
u
w
-' LW
'", CD '" '"
n co
17'
i-
~
IrrW
4.6
o
W ~
::J
L-
o
s-
>-
z
« o z
l.."'J'tv
v
-'
w
-' «
0-
0Z
>-
o·
z
i= u
o
i= u
j
lS
w
-'
l.. W
L-
W
n
k:::
n
IV"
L.-
Co
~
~
::J ~
::J ~
k
>(
«
~
l~
~
-,
>< «
in
:::f
L-
o
w :::f
=11.
p/r
i=
0.7 o.~
!
.00 ,
' I -_-I .002 .003
I " _" I I .01
.oo~.
I
J .02
I .03
I
"
.os
1 II I .07.1
I
1
I
1 JIII'I I
.2.3.~.7
I
J 2
1 3
I
LOAD DURA'nON TO rUNDAMENTAL PERIOD, TIT",
Figure 3-227
'Maximum response of elasto-plastic. one-degree,of-freedom system for bilinear-triangular pulse (C1 = 0.648. C2 = 1000.)
J 1 Jill 7 10
s
TIME OF YIELD TO LOAD DURA. TlON. ttiT ..
o
..o
<1' 0
~
V'
o
.<1'
0(7\'-"'°
500
.....
0
Ilc "7 !\T'll)l (\ \
300
4oJ>I
~~ ~
w
x
~~
<,
10 0 1"f:-1-'r''k-'t--l'\--!'1
z o
70
u
50 hl\t-Afl--'dj.·-',\l
r\l/
.
. .
U)
......
lIlu, . . . O '" '"
~~
.
\H
. 1)<
<,
~
?
~
? ?
N
N
-
-
0\
0
... v D
-;
-
u.. 0
(7\
\-.V
1)< <,
I'..
"'-
..
I')(:
I'.... V l'><
t><
l1\
OOVlU\l.Jl
7'
V'\ V
.
00000
·N"N
f\17~I)< I)< I '" . - IIl\f-JI.-VnV-t"lt\-'lIH'K Ill'r,:!.n If. .. '
70 0
E x
;--l~
...
,.,
00
000
VI
"''''''' 0 .. 0
3rY' :« ¥
;'/
..
00
'\., ,'t>
..IVJ 1-11/
4
3~
k b< 1"-I K II .~~~':'~
00 000\
OUI
V
v .-1
V
~
"" V
~~
,'>
,;
V
"f V: /'
"q.'>
......E
'b .1-
a
. vfZv~EL~[;?
1,\
6.1. 5.3
t: W ...J
4.6
W
'"'. CD '" c,.
...J W
>-
W
2 0 Y-'~t.A-IAfi
:1
\)'wf#7'1
a
J: s J: o
IZ
N \IN <'6 10 -tl/I--Aila.
ID'
.-
"
7
iV
...J
o
::£ :::> ::£
x «
5
3
~
::£
~
j.-'
~J\F'::: ...
~
'~ 0.7
0
<.J <,0
...w
1.3 1.1 0.93 0,81 0,70 0.61 0,53
/
t-,
0
i= u
1,7 1.5
u, W
«
::J
c-
U
...J
« Izw Z
o Z
'"w
11.
4'0
30
I-
o
0
::£
u,
o o
z
I-
<,
w
...J
a
::£ ::J
::£
x
«
::£
... 0
w
.::£ t:
-",
j~'--
.0.5 .001
.002.00l .005
.01
.02
.03
.05 ,07
,I
,2 . . 3 . 5
,7
r
2
3
5
7
10
LOAD DURA"TlON TO FUNDAMENTAL PERIOD. T/TII
Figure' 3-228
Maximum response of elasto-plastic. one-degree~of-freedom system for bilinear-triangular pulse (Cl = 0.619. ,C2 = 1000.)
e
e . '
0
•
TIME OF YIELD TO LOAD DURA TlON, .r
p
0
0
V'
o.
....
'"
... a. 0
.
N '" '" '" 0 "'-' 0 u. u. '-'
'"0
•
td T 0000
'" N
l.D"-'U1"
OOUlt,l\
0
'."
000 N
"'-
'"
0
'"
0
'"
O'
.
0
.
0
00
000
Ul .. °00'1.,00
"'''''''
0Cl' O.
-
0 ....
0",-
IA
~ /NYI"'i'
JOO 200 w
»c <,
N<1lX'h
E sc . ~
,c:
u u
--' u
100
Jl
U,,"IIIINn I I
N
Y1:xw
Z I-
<,
...,E a
o II J,."l..L1.LLIU...Il........i:t'/Nf,A Y
H7 .........,.1I-.Yl~ b.l.
- -_.
70 I"''k·f-'»g·
..J
cr:
rHA13"..J)'1t.JrPlf4IJJ,.,......:K11 ""M
~O
4·6
r,
Cl
co ,
...'" cc
l>:
w a.. I~
w :f
cr:
a
jO
2:
a
::>
~
o I-
--' u
L..
~
o
f-
z o
Z
o
...
t: u
e:
u u
7
--' 'u
s
o
W ..J L.. W
a
:f ::> :f
:>:
::> ~
j
x cr:
2
L..
:f
><
"" ~
o
w :f
i=
07 O.~
.
g_
. ....
,
l7,
.001
.002 003
.OO~
.01
.02
.03
.05.07.1
.2.3.5
.7.
I
2
3
LOAD DURAllON T~ FUNDAMENTAL peRIOD,. T/T H
Figure 3-229
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (Cl = 0.590. C2 = 1000.)
5·7
10
TiME or YIELD TO LOAD DURA TiON, lEfT
- 000000
.s-
0'
o
~
lI'
o
o
0
~
l1'
t...J 0
C1'
-
'"-lcn""t..otNN--· . • ." CD"lJ' ... V4N ....... 0 lJ\ VI" 0 IJIN O O U l l / l . (J'l
-. 0
Vt
~.~M, '''-r-'Ft**~~rA~IT B ..... .. -.-.. _.. ..
~
'iJi
"'J Al'-Jll) .. . V" JV JlJ lJt'....I M LYI [')( II n . •V'l.IIMJrNrAF , .... ~' 'H.fl\)( IIJ-tl ----.-.--~
,)( 11 ),.,/ "'"A I
E W ><
l\.l-1
<,
t
><
z
o
~
u
W
.-l
u,
...,,
'" 0> 0>
'-l Cl
o
--'
,..
~J
")Q.
tf\-.. liN Ll'1'ilY1~ .01
""'
I'> I I ""1'111
~~~~~~~~~~~~~~~~~~~~~~'t'~<~.
In-.
'1
I"X..
J I
I
.I.f
lui,11
I
II
A
000
N--
0"''''
0
0
0
... co en
0
0
VI
...
COllloe
o.
~ }Z::uY
~
U
w
--'
L. W
o
:>'
::>
:>'
>< -<
:>'
'1'/ I I K:: II I;; I )(: I
....E
IA.o4I/
J9
Ill' I ,.,
I
:AI
Cl
o ~1~'1
I O.L
cc W
z
w
:>'
« z
o
....::> o'
,J
i-
~
-~
I-
Z
o
t:
u w ....w--' Cl
t-
~ r\
~'Ii
"""rl;l?l~A~./
l:O'""1Ffl' I II.<
1\
--' «
I-
lS"?f'hIl'!l'\;..4..'7I'-
~
5
I?
Z I-
<,
~~~_J(.~c._
/
~
0
Q.
':.2
..
mf!
11'IIV111"'~ ~ A'
I I b I 12 Ibf Ibf:>F
o Z
o
_.~_._~~~
I-
o
000 '" N N
~
p / r I ';"'ii.
:>'
~
::>
:>'
X
«
:>'
.... o
w :>'
i=
0.7 0.5
i====
.00'
.002.005 .005
.01
.02 ·.05
.05 .07.1
.2.5.5.7
I
2
3
5
71,
10
LOAD OURATiON Tq rUNDAMENTAL peRIOD, TfT N·
Figure 3-230
Maximum'response of elasto-plastlc. one-degree-of-freedom system for bilinear-triangular pulse ICl = 0.562. 'C2 = 1000.)
e
e
_ tdT
TIME or YIELD TO LOAD DURATION: 6'
o
500
w
\1' ;0
... 0
~
v< 0
j,/<--.lVI"f-.:'-=1Ii"-k f'\
N
iJ'
-
J'"\ V\ (\ 'j.
'" ,\
"('f)(
'\
\--/1\-
200
l4~ ~~\)\ -K \1'\ Y '~,xil\
70
.'"
~
50
l'
to '"
0>;::
20
1-
~ .,u
-,
.'
\:
0" \
X' f-
Y, /
. -
[)<'
. 1>- eX
l>
~
Cl
;~
3
t
2
V
If\...
VI VI
0 .... tH
?
??
VI VI
N (J'I
N 0
O? --.O"l VI
0 0
0 0
0
0 .. 0
V
0
/" ~
t
~
;;~ V J" If 11.
I'
lA'Vj'/ fiY /
t2
.I.
l' ~
V
\~
W
I "I
00 N N .. 0
tH
f/ IV
"
r-, V
0 UI 0
0Cl 0'1 0 VI
D/ J
\'
'~
b<
V I'>( [)< P'-Vi> k/ P'v t--.. V-I/o c.<
<::? v
1,( i>. 17" k'v'
vb
~[)'
K
\\
V.
~.~)
V
&1 7. \
....
62
'"
S. 3
V
'1
V
v:: -
v
v-
I-
Ai
"...-; .---:
I-
~j/ lI' I'
I-
I--' /
7
V ,....V
,,-, r--"'"v........ ~
b'
v
I-~ 8: I-'
b"r::: ~
A"'-
":6'
1/
..
'0
L :-
.J . .1
.J.
1-
y
V 1/
1.1-
X
..l.1 . '
A
v
~~'\10
'::. ' t
p/r -,.
o .53
j::- ;
. ~..
-J-
.002.G03 .005·
.0'
.02
.03
.05.07.1
.2.3.5.7
lOAD DURATION TO rUNDAMENTAl PERIOD,
Figure 3-231
Z
e
1·
2
3
TjT N
Maximum response of elasto-plastic, one-degree-of-freedom system for bilinear-triangular pulse ICI • 0.536, C2 • 1000.)
j
~
~
~
0.93
.'0
X
V\,
1.1
\. \ 1\
1::-
.OU \
a ...
::>
I\:;
,
l.-
' 1 . 5 . rj/ I. 3
v
P<- 1\
~ ~
~
""
~
'i
1.7
F'"
..\.
;;
G
r-
1~_
0:)
0
•
~l----
~
<.6' e. 3 :t 0
•
1;::-,:; /'--['--"1
g !z
VI/... 7
HIM1i-'~
.
'~~ '"""..<0
Ij.f
z
V
P'
c-\
1;;:::.i--'" i~f.-'
...... 0
Ie
i?\.~\g~ ~
II
"
0
:;: :l'
I
i'><
1:>'
'/-. X· l2 -,' ,lV!IV / I> ~ lX"d (X~.,
!j
U t-, VD<
, _,,,,-
, ~ -- "V. V - " ' i L ",./ /'t~-_.
10 .: Y 7
'IJ<
'I\.
ui\i \ V I'"
a
,..J N W \D
'--:\~
JO
V'
J
lD 0
~
IX.
V
,\ 'f,/_
j
~
? ' '?
_
N
/ ~
1'\
/
,- "ii
1","\\
'j
)
1'\)\
)I--....!L\ ,~, '\ V . l' »c 100 k /IV \t'\~ \ I\. "" "iI--1\;! f _- - ,7. .~
a
0
Vo
JOO
-c
-...l (f\ .... V< N N _ ••••••• ....... 0 VI (Jl -...l 0'1
-
o.?,
•
5
7
10
~
r::
e0
500
'" 0
,.., ,..,
V'
~
0
0
'"
0
200 w
sc
F
><
z 0
;:.: u w -' t., w 0
...,
0
a
0
..,..,,
-'
w
>-
IAI%l4:~f;;0/ t7. r-,
I----L
<,
100 70 50
30 20
0
~/I\'i
.\/1----0 L,~
W
,-
-' w 0
:>'
IJ
/X
~
1/
N
It.
~
I"
,,' L, I r-,
17
<.u
t-,
"
o-
"""
.
...0o . ..
00
"' 00
NN o~
\;
v.. . Ij
~~ ~
"
1/
1/
\\
9. 5 B.2
if
7.1
1.-1
~ ' i c:: 12 j;
'v
III
..
00
DO ' " 0 U'
Ii
If'. ......
.
00
6.2
K
1>
k. J) ,I><
r-,
K 7 i)(L/" ~
17
1../
~~
,c,
I-I--
17
i/ V D'V (><,.< .
IP
'
v V
r~ l7
t-- I-l
t:I
~v
L--
hft:vv~
.
II·
'6'
....
,7
1/
":0
r-,
J: s
~
J'.o
;1,&; ~-'7~~M~~
l.t
b~~ r-t>_ h'l---
'r VI U Ir
?- V.
V
I--
t...
t...
<'6
<.J' 2.0
"
~
1.7 I.s I:J 1.1 0.93 0.81 0.70 0.61
. "'t::-
V V
i-
\
-:...,
~
. . . \ - ' 1 _ - HI\-
:::::----1.---
I\ 1\
,
PI: Il
V
\ ...
\
..x
..
. L.-'
\.
1';I"'- ~
.....E o
a
cc Q.
S:J'
:15
....z
<,
w
''')
/
_
c::::1
t~ CJ..: 0.7
'
"' ... v
o'
.. .
])
I/(/}/ ._: .
2
I"
II)\.
I><
/ I\(;hif+,.rl:>'" / 0, 5 6 ~.
r.:Y:::
'"
00
'"
"fir>' K' r.x l'X /
I\\-II ..
J '\\[~ / -I j )<
...
V
f~~ r£- "tv r-
LX' 'lX<,!Y'1 l/ 11'\ r\ 1'\1/ X I>
7
3
1\"
\)
A IV
><
f\.
10
:>' ::J
:/1\
Ii\'~ Y.. . y \ .f\ bl\'x .7' t/D< '
f--
U
I\
!V:~~kX.: .
r-,
z
'"
l!"hl-~ -/ 11' \I!' 1/
~ lX'J 11 JOO
tci
TlM( or YI(LD TO LOAD DURA TlON, T ,'t . ~.~ ~~ ~ 0 0 .... ... N N ;.., ;.., C;; <; :.. 0 v- v- :.. 0 lAt...a clO-...aUl"c..... o t.l'UI UI 0 '"
11\
V"..._
0.53
p/r =:n'.
-'
....Z
w :>'
o
Z ::J L-
a.... z o
i= u
w ..J
LW
o
:>' ::J
:>'
;<
:>'
...a w
:>'
i=
!= •
0.5 .00 I
.002.003
,005
01
,02
.03
.05.07.1
.2
,3
LOAD DURATION 'TO rU,N.DAM(NTA~e(RIOD,
Figure 3-232
.5.7
",
2
3
5
7
10
T/T N
Maximum. response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (C1 = 0,511. C2 = 1000,)
e
o TIME
'"o
'" 0
~
o
OJ'
,.,
o
'"
~
o
-
CfI
-
"""
or
--
•
YIELD TO LOAD DURA TlON: t[/T L • -r 00000 'Icn.lolNN--· ....
-
~' • • ' UI 11I.......
.. 0
Q........
." UI
a
N
U)~U\.lH 0 0 VI IJI l/l
N _ ... 000'1 O'IoO'lVt°oU\
~olLLLR'.l ~ji~ I~'
"'~
. ":J W\\V \l1l't1JNJ 11 '
·_~tt?tf-tJrXI:
w
sc
.,1)<
<,
k'
E ><
z o
OO~
0-
,-'rl
U
t
,.,
--'
k W
1/
~
k:'
o
o
--'
'" ..... '"
w
>-
o o
;= w ..J,
o
,.
,. ,.sc-c
::>
... 0
Z 0-
~
II
<,
....E
~
o
6.;
o· Q:
w·
0..J <{
0-. Z
.
<{
,
,.
w
z
k
1:tlt~~~\7
a
0-
A
z o
n-
;=
o
W
..J
k'
u·
17
.I
~
~
l/
,.
~IV \
::>
.
x<{
~~'
~
2,
,.
o
~
5
,.
r\
k
a
,.
w
P I r ~ ';;"1l,
t:
,.
o7 0.5
.t:::
.00'
.0,'
.002'.00.1 .•005, ~
Figure 3-233
.02
.05.01" '.1
.0.1
.2
.. 1 . 5 . 7 . 1
LOAD DURATlON TO RUNDAMENTAL PERIOD.- TIT", ,
'
.
~
'j ~.
::>
7
I•. W
00 NN
o
~
"
0-
.Z
u
~
0<
I1W
3111l IlJ
r-t...
a
00.0
r-, ~~~tt7'~~~f'-iIS*4~~m~4fIJ
.,I/I'll 'J.fNJ I
W
"",
U'.
~~???oo 0 0
N
2
3
-
Moximum response of elasto-plastic, one-degree-of-freedom system'for,bilinear-triangular'pulse (Cl = 0.4&7. C2 ~ 1000.)
5
7
,0
'.~'
...
TiM( or YI(LD TO LOAD DURA TiON", t[/T cf'
o
lJ"
...
0
0
500
I"
...,.. . t'-l
0
D·ll'
",.
;;. "0 ... :-.... '" 0
............ \V"{XI\VllVIG I'I·IIAI
f{
.. '11I1\J7IINY " '" 0'V: v-
N
... "' . a-a '"
-0000
N
~
,
N
l1' l1'I
0
0
0
v-
N
N
'" '"
0
00
0
"'v-
0
00
0
0
0
0 0• "'''' 0 ' " '"
v-
0
00 NN
.. 0
\>
x'"
<,
F
»<
z sz:
"...Iil fl»{ I ..~,
0
u
W
50
rI"'~>,i-~~-".jf'"
.3 0
1f11'i1:'P\~LtJ<~~\[;~Ull
1/1V1 Lr JU--1II U/
Y-
J\,}'
f:;H'"M>:b-lf'::Jf1+-
-.J LU
a
'", '" '" '"
0 -.J U
>0 f-
Z
0 ~
U
L-
20~
10~~~,
f',
v
~
~
ETI
. 7
a
~
5
::>:
..
oi'
2'
'\I
\
J \
~-\
-Lll-\
1\ .
II
l
V \:' ~
~
-
3
x
::>:
fqJV1
-l-:
::>: OJ
.~
['7
<=
-.J L, W
V
il-U-lllL ~-,
"
r-,
'\ \
~
,;'
i?1.-"
.-t'
7't: .t' "'.
:v
\~
IT" 1\ """"l::::l \
--,,\'~
. p(rl::b..
d:
T
.. 3 . 5 . 7
1
, ,
,,
0.5
.. 001
._.002 :'00.3 .005·
.01
.02
.0.3.05.07 -.1
.2
2·.3
5
7
to
LOAD DU(lATiON roruNDAM(NTAL P(RIOD. T/TN
Figure 3-234
Maximum response of elasto-plastic; one-degree-of-freeaom system for bilinear-triangular pulse (C1 = 0.464. C2 = 1000.)
e
e
•
•
TiME or YIELD fD'LOAO DURATiON, t('/f ,
r
'" 0
500
U' 0
r'_.
E >< \00
l
z
II:: .1 "
70
1\ "-
r-s
50
~~ii
I
20
f--
0
\0
u w
7
;=
,w ~
><
e>
2
~
J
J<:a
. /'
>5@
_,(
I~o
,
o, 00 \
-
0000
:~ • ~~lA" 0 0 1IIU1
0'000
N' N :- -
'" '"
~ . ,,~ ~~ ~ D<
~
-
1/
II
r-;
0
O'IoO\u.
2
I II
)("-.1 II ...... /
,,y
U
'( 1'-
V.
....
'.
IV Y
0
o
0
NN
.. 0
II
9,0 Z
B.2
-
1"
,,",
YO" P .I' V:;i//z"
II:-
~
~
)<../
'oJ'
~
'~'- v ~
"-.IK
1£
V
t>
\
iY
\ i \
'
-,
vr--Y-a-vy.
....,l.-'"
,
.
/
.J,o I---.
~'6 ;'.1
I--
.
;·0 1,7
~> '<:\--.
1.5
./
\.
~
\
.
,0 r
,02
,0'3
'LOAD DURATiON
Fi gure 3- 235
L.J
:O~
,07
,I
.2
,3
ro rUNDAMENfAL PERIOD,
.,~
'.. 7
I '"
2
3
f/f N
Maximum response of elasto-plastic. one~degree-of-freedom system for bilinear-.triangularpulse (C1 = 0.422. C2 = 1000.)
~
7
Iq
...J
-c
f--
zW ~
<>:
CI
z
::>
L..
o
f--
Z
o o
t:
w
-'
L.. L.J
a
~
1.1
)(
0.53
- ,
.OO~·
l>:
::>
0.70 0.6\
-
.002003
o
1.3 0.93 O.B 1
\ .
~
">. x:xr'>:'P I ( ~
'-
J: s
/ "-
V /
r' '0
"k
f;,
~I~-
'6'
rX
"" ,; """'I '" ,. :..a: -ff
~'
r
l/'
,C
,/
f2 '
\ \ \ 1\ \
"-.
....E c,
oS'
<0
1-
a ~<
,
: / '
'k
0
f./
I" '7
00
'" 0 '" '" 0 0 '"
Cd
o
..
0
0
-
r; ~ / : ~ r,t ~ " ) ~~" {/ L-f 1/ 1A' /I " / IIJ,,~lffi I?
1/
o
-;;
I I
1/
k' / 1>;.... ~h-< "v 1'--7>/1><"" " ~ y"" _ V ,,~~:.. ' L ~
,-r
Lc::' 7
0
./) ,)'",~".
11/
i';.
r-,
N ':'" .",
D<
K.' r-, ' .:-.
'" 1/
~_ is ~/-:?
M '
_
\'7'l /"
,-/J
, 3
'r" 1/ J
x
I'--
~/' ~~
.-,
N
'"
l/
I'.
'\. : I Y.
r-;
_;Lr/ A.../ .I ,V "" JI. . ~Q
0
«
.' ... . '0 :.; '" 0 '"
..,
r-,
:l'
::>
'
'X .
f-/
-'
CI
-,'"
,,_:Xl
7
1~~'iX <' \ <1/ 7
30
CI
'" ,. ..,'" 0 '" Z
'"
t\
. 'j'
W
-' w
a.
0
\ / "j,'\.;" . 'Y--. V'M'. Y ~
<,
CI
0
rA--\~bl ~~ /
w ><
'-' w -' L
0
N
N
J:.l. IV' v. IV Y 1£7~ : ~ I~~') \V
200
f-
v'
'lL~
;00
0
l-
~
-c
"
.L..
o
W
:::f ~
TiME or YIELO fO LOAD DURA TiON, tdT 0;
..
...,.
o
v'r-J ......... Oli'O Q'V"
o
0
'W'l L
500
~t'~- fi [{I~ ---4.l JII
,
300
200
'-E
'" z.-
0 ,_.
u w -' u. w 0·
w
,
...'"cc
\00 70
50 30
0
-' w
>-
20
0 f--
Z C'
\0
f--
U
w
)
w
0 ~
"
>< 4: ::i
~
.
-
Vl
.••• -oJ 0 V\N
.
.L'J.
[7< 1-
-IX-V\-,
i~
~ r\
I\
9
.
'.
t:
.
""
_
k:
U
I"
~
<
I~
~ t1--
. 0.1:\:
'" l-\
00
000
a
0
VI . •
V.
o
°oUtOO
)
Lf::.A e:
').
N.N ~ 0
r: It::: f"1
1\
LL
T"
j"
9.5
z
I-"I-":VV(",...V L>:t 'a. r/ t' r- y iI'
8.2
:>.,
f-
)
." a
V ~~tf t:~ _
a
L'-
0:"
'" W
0..
."
c.
..
J'
e
'or
,-'-~ P
' - ' . ~. k:::
'7
f6,k
v...r:: V·
<"'. rM
~ _I' ~
.
~
~
.
~
t:: :; rr "
. "'1. ~
,1
~
..;;
~ I<:';
~
.i!l.
f1]
-
u
\)
0:: 0
'2
J
\.
:l;;rl';ll
s
a
'"a L.-
.J. o
I'
<.6;
a
<.J
u w
2·0 1.1 '.5
!:::::
~
~ If l!2:::\ ..I.6\, . C-.'
k:--
::i
j~ I'
tl t1
/1l21'
";:::
~
w
I.L.
X
.
':6'
-c
I
~..l.
..J 4: f--
2
. LL l ~ I"
D
-1d: ..,l
1= ~ R::: ~
r""" IT;:.c\
1.3 1.1
1\
.
~
0.93 0.81 0.70 0.61 0.53
.
f-
2
f--
..J
L.-
w
a
::i :0
:::f
><
« ::i
L.-
a
w :::f ~.
=:r:::.
:t== .00.2003 .'005
IL
.
ot.::
p. t'S::~
~
...... OQ 0\
. V111 .
1;,( IL
j~
.00"·
N -·0 0'1 VI
1I
0.1 ~
c. s I::::::"
01
ILI.j
.:1
~I\Y"~c -~ ~ f-:'
000
[")I
c_.e:r-
.~~.
.
N V
0 .
N
'-"" V\
'. liI,JI\/··v
,\,- >1
ki
~_c \~
...... 0\ ... Oll\U\
'K tlI
IVUii'W I')
100 0
P
0'1;" lA" " _'" r- _,~ , Li'\; " IP l0 VIX JI\V' - ,," l" ') r- ,.. ",t;. 7" e t.1 1(,. k',.'i-L i- .~ I.' ~v .
3 2
a
~~l¥ '. - cj'r,'
:0
::i
....,J
' VI
0
_J
L
}J - Il ' -
.•
~L~ . ./-I\ J~\' ~v ~ "~V\ [j
w
x:
.
I!\"_\~f\&' _'i.
00000 -lOl"'\..HNN--····
0
.0 \
.02
.03
.05.01
.1
.2
.3
.5
.1
2'
3
5
1
10
LOAD DURA TiON TO rUNDAMENTAL PERIOD, . TIT N
Figure 3-236
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (Cj = 0.383. C2 = 1000.)
e
e
e
I;
TIME or YIELO TO LO"O OUR" nON, t((.T e-
o
.J'
0
..o
v'
o
r-J
c.J'
N
0
_
...
_
:--&.?'
(J\
V'
0
-.l
I
0
~
u.
l1\1Jl
~ -.l
~ 0
00000
0
00
(Jl
~:- ,,-,to
U\N
:.... i..n ;.. i.... N (ftu\
""
.
oo~~oo 0 0
;"':"'_-(»17\ o(7\l"I,
00V'
'" o
0
o
00
'" '"
o .. '"
0
~
",**H-4f+!-J1.I+-H'lM-A-lFI.oA--lf""l<
9.5
z
' . \Vll 1"1 /1\ 1/\.1 A 1'\
'"E '" z
0
>-
u
w --' w 0 0 --' w >0 >-
if '\J Vil.1tl1'\..V1 \)'
,-
N
<.n '"
>:,
J...j
'"
w
<,
...,,
8.~
~-M
~
~
w
..... 'of
.1'-,
r.,6'
1/'1)
w --'
7
W
5
L-.,
L.
0,_
::::> :):
'"
~·o 1.1
/.5 1.3 I. I '.
v
i ~.93
.~ '-J',.
0.7
...
05 .00 I
.002.003 .005
.0·,
.02
.os
.05.07.\
.t5 .2.j.5.7
1
2
3
LO"O OUR" TlON TO fUNDAMENTAL PERIOD. TIT N
Figure 3-237
0.8\ 0.70 0.61 0.53
p';r I ';11..
-,
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (el = 0.365. C2 = 1000.)
zw :):
~
~.,J
3
< :):
>-
I.-
<.;'
:):
...I
<.
J: s
J: o
""
~
.....
'" c,
W
<
>-
U
0
":0
~
0
"-
.....E 0
6'
S'
~I\'\
z
>-
5"7
10
::> 0 >- • Z
0
;:: u
w
...I
I.-
w
0
::E ::> ::E
x
-<
::E 1.-'
o. w
::E
t:
0'
o
-s: 0
..
o
THolE' or YIELD TO LOAD DURA nON,
v ,
v o
f'-l IJ'
f'-l
_
~.
0
Q'I
VI
_ 0,
;:--i ~
en a
~
U\
.~.~ ~
Il.,
....1'
NN---oom
OU\NOOUlV4Ut
. _... _.._.
0d\X"f1 D II
??O?Ooo o
0 0. 00 0 N _ •. ···\D'lUl-'\.H.
..... ~ c..n'"
UlOO'l(.H°OUl
..._. ..- .- .. . .-r-- . . . -
-- .. . '\
tdT
I
I 1"1.1 ,...
1'\.1
I
A
I I
~
Wlv
0 lJoI.
0
00 N N .. 0
/f/,
._ . ____. v - _ _ .VAI: ~
- - -
9.5 2:
. .
f-
I--
i"JI
l\
0 Ul" o 0
<,
~~:Itzf::+~
litAR
...
~I~ .
~
./
:,0
o
I Y~'a. JIh.
t:
v
u
,I'll AI Nl
.J
" 11'l.ll W II lIlV
o
'" a.. '"
\..'-I/.~I
IYI"IIl../1
PJ)
l~
'"
o
W
>-
o ,-
ILll/~
z o
:I
t:
u
W
7
.J .L. W
o
-A,,~. "
:£
...J
w
::f
o
Z
:;)
L..
ofZ
o
f= u W ...J L.. W CI
.19'--
5
::f
i-
:;)
::f
3
sc :£
""W Z
~f
:£
=>
'J-
~
~III/
.J
.s
JI
11.v
\)[1 i1 ~
W I. W
,
o
o 0-
z
w
....E
2
>< «
18
\~
,1\
\
::f
L..
o
w ::f
p/rl ';ll.. 0.7
o 5 .j::::; .001
'.
.002.003 .005
.01
.
.02 .03
.05.07.1
'
LOAD DURA nON TO rUNDAMENTAL PERIOD,
Figure 3-238
4'
.2.3.5.7
, i
I
i=
i
4
2
3
•
,
5
7
10
TIT II
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (CI = 0.348. C2 = 1000.)
e
e
..
e riME or YIELD TO LOAD DURA nON, tdT 0'
o
oJ' 0
..o
v'
o
....
....
oJ'
0
z,
~
x
~
-
(Jl
..
N
Loo
N
-
• • . • • •
0",
0
Ull/I"'"
-
?
··y)
~ ~ ??
XI\ It'll _ IIX II AN.IIAlf,U I All
JJI
A'\.lfl
I\lI'
III"
IIIII~
? ? ? ??
~,~~l~~~-~ ._._-. . __
D
z o ~
W
..J L W
...,, ...."" '"
'"o
0
IV'
~..t
~I
2:
I-
<,
-HL
....E n
a
...
30
_. . .
-fftbmt
n
_-
...'" Q..
...
-- -_.
- ._. . _. --
... _-_.
10
"'
b¥
... ::E
~
z
r-
...::> a
I-
z
.... _.
.-
. _. __...
a u
~\/f~
t:
........:J o
U
[:c~~*t1
.. J,...of'""I.1 ---t-"'lV---It'""".I........--
W
n :>
<
Z
eW ..J L
....J I-
J
~
Z
'oj
< o
'-l Yl 'j.j,( Ub:
,..
J'
bi'
20
,)
o
...... ..
00(:)
0
!.,o
. -
u
..J U
.
0
, ff..:-'\(J I A '\..1 r lI"!WAf V-'; .' I'''''' 1 Id
,..
O.
o
'" o "tr
. ___... J-: _---- VI\~
__ . . . .
_ ...
..
0
/
.. ... ... .. . . ... .-Ill N'-t+7'IM'I
l N)( 1l'iJ ~.. IXJAI~
0
..... N t - J - - - Q l J O \ (,HUlOo\t.M00Ul
...... lII~
OUlNOOll'l"oi
V ~A
I\IA\I
/ \ "l-..!/t...l1""-
::E
::> ::E
::::J
)(
:>
'"..
<
"
.
:>
~ 0./
..-
l
<,
I \I
,
I
-, .....
'"
1'-
\
,,
::E
i=
-c-r:::~:t-
k-:t'4 I IIIII-l .002.003
.OOS
.0\
.02.03.05 .07.\
.2.3.5.7
\
I
I
I W-J
2
3
LOAD DURA lION.TO rUNDAMENTAL PERIOD, TIT N
Figure 3-239
...
'\1'\1\
o. ~ .00 I
~
\
""\
...o::E
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (CI = 0.316. C2 = 1000.)
5
7
\0
TIME O'
o
.J' 0
.. . . o
Of
YIELD TO l.OAD DURA TlON,
..
~r.J--_
O\J'OlJ'lV40~
,00
01
...
l.1t
N
o
lJI
l/l
"'-I
N
__
0
..•
a
.lJIN
~
tdT ? ?
~ ~.?
0
00
U)~1II'"
I.A
N
N
-
0
VI
'"
0
VI
~o
a
lIIU1
, ... ~
0
0 o --OJ.,.. '" 0 III o
..
o
0
0
.._
_
( IIX I /1I'lIlIl'NL.I' INI 1111 I V 1/ .-~"fI-1+f'1h::H'-H---,J,h-ll-!M"l,fI,..A~ll
!\" JoI ~/I '\If .lK.".!1'Nf liP'"" . " " .to.....ll~
j(JQ
no
0
'"o ...... ....0
0
I
z
;,r;:
t IV'iLlNiI Ilfl
>-
<,
.....E
w
sc
<,
oz u
~
e' ..J
II
,0
".f\r-Y
w
,
'" '"
0>
».
o >-z o i= u ,oJ
....W..J a
l'nJ'r"o..-l
~ 1lW-¥'.
1"1
IA'I
I
I"~
VAJ,.../,
'-H ~ 1.J:1tJ 'P%1'P/1. .I. .% Wi' JV' VJ.I1j ~ II VII....". 6 . l .... I , .. I
,
~~
.\0
V-
20
•
od"'"
~
B
_
N/I'\
M
------..
s
.....-::::I::::i
I
II
I-
~
z
... o
I-
Z
o
ll~
- . .1
i=
o j
...w a
r=r=J.
"
v
~
::l
.
:E
::l
><
.\
~ ~ p/r
\1\
1'-
.'-
,
[ .0 I
.02.05
.05 .07
.'
~ .2
.5
lOAD DURA nON TO fUNDAMENTAL PERIOD,
Figure 3-240
.5
.7
I
~
lJ~
._'.~ -l~ \.~ __: :~_1l'_
2
o7
...J
::l
,[
I --A--1l
Of
"-c "'"
..
l. 0 pt::;..
.. .
a
7 A A't-..c:.1 A"""-
Q:
Ii: Z W
O!
1-'"
I-,. \0
-s',J-
.i.
f--
I
e,
...J ~
...
10
e.
a a
a
o'
II
_..NIK NYN:I Lfi-.ll,11111A-1- 11
E >< 10 0
.... o
w
~
='~
2
c:
3
TIT N
Maximum response of elasto-plastic, one-degree-df-freedom system for bilinear-triangular pulse (Cl = 0.287. C2 = 1000.)
5 .1
\0
e
-
•
f1ME or YIELD fO LOAD·DURAftON. t d f 0'
o
<1'
0
~
o
v
'"
... r-JO;~o:...ao o
'"
0'<1'
'0-0
00
0
!"'~~!".l:""':-lD:""Ut"'",,"
fJI
tJl
-....1·0
V\N
0
0
VIOl
lI'
o
0
o
to
N
....
"'.0
00 :"':"00 Ol u \ N O O Ift
0 0
o lJ' o
o "0 (:) '"o ...... ....0
0 ..
0
_. 9'5
\ 17
rp"",n, }'l-f~Ic-,fh.-f lJi'<-jfP!H-bHlll"
SOD
""
,no
-"
...J
'"E
~
~.
'"
L
"
>(
z
0
.,.,
6'
\00 70
,-'u
•
W ...J
r., W
'", '" '" '"
a a
.\ 0
w
}O
-'
>-
IN.t11
r .
z c:
u
l°ftt~~~
~
v
~
'6'
~
'0
~s .t.
'0
~
~'fi <.J
-l.
.f-
...J L.
w 0
. .
rr+_IlL'"
<·0
7
W
~
~~-'Q;
0
0
.'m
oS 'J"
1.7
~
1.5
1.3 I. I 0.93 0.81 0.70 0.61
L
::>
.5
L
'"
«
2
~
I\
I \ 1\ \
,I\."I -,
~
v'
V'('
1\
p/r =rn-.
0.53
I-~....-r-=~I--
---_ ..
0.7
o
~
==
$
,.--r:::-t== ,
00'
.00200.5 .005
.01
.02
.0.1
.05.07.1
.2.3.5.7
I
2
3
LOAD DURATION fO rUNDAMENTAL PERIOD. f/T N
Figure 3-241
Maximum response of elasto-plastic. one-degree-bf-freedom system for bilinear-triangular pulse (Cl = 0.274. C2 = 1000.)
5
7
10
...z
<,
....E a 0
'"
W 0..
-'
...« Z
w
::E-
«
a z
::>
.... 0
f--
z o·
t:
u
w
....w...J '? ::E
::>
::E
>< «
::E
.... 0
w ::f
t:
td(
nME or YIELD TO. lOAD DURA nON,
'"o '" 0
__ :--"(J\ OIJ'OO'VlO""O
V"~t..J-
~ "0
500
'bl\t-..l \ .,,~ .100 1\--<)1\ J ~/ K
_
I~SJ~~'~ zoo
1~--1):~,)(!, ~
w ><
<,
F.
>< z
0
f=
u
w
100 70
-'
w
I"
oj
W 0 0
a
-'
>-
z 0 i=
10
w
7
u
:::> ::f
...
.I
><
::f
2
VI.
rII X ~
I
i7
.
VI',
'000 _ .0 0 0 •. . · · L D ..... U' ... lJ,iN
-
U'N
a
0
U1l1\
,..
en
_
It'
,/.
I~ ~l!
ViNI'..l
°OtftOO
0
'I-
...
0
LV
[t ~
L-<
liY.. ~ t:I:,--
II l>
000
-OOO"llJl.
IY
--- [t
Y.
°0000
cn,lJ,i
i/IL
Ii
1)1'
--
000
....a
1/.11'
I' K (',
K
9·S
t1 1/
V I
-
,
~
\_
w:
I--V
t-.'
, .,
~
1--:7
~rs:~tf ~
r-,"'" k
Y
-::
c<,
.>
....E
I
'" r7
rI
r;
•
r> I-
-
~
17
I--
V
~I--Z
f?l~ 7"~ ~'F\V 1I
.
.v
I\
t-, 1'\
,
-,
-,
-, \. I'\.
'". \
.....
~ i'\.. K-.X k-:::
~ K) I::)..G ~
I'
I'\,.
l.f,
t/ I",
1>\ 1\ ~ lK"
> :p I r V'.
-J
.
0: 0
o
...
::£ Z
L.
I'--
.f.
o
c-
<'Ii
1.7 I.s 1..1
'
~.
...
cZ w
<·0
'r ki I> r~ ~"t-
De
"J n,
s
/
-I--:
J" ',r
<.:5
"--
V-V 1/
---:
',.,
Cl
a
.f
1/ . '"
:.-0 V
e-
"
,r-.
-cv
'"
' (;'
~ 'tTvl" I)
~~
V
1/1'\
If
•. "\( It
rJ
/.
f/;t
f--
:::;:
\
I. I 0,93 0,8 I
"I
0.70
1\
0,61
0.53
;: o,
~
a z o
t:
u
W
..J
L. W Cl
::£
:::>
::£
...::£>< L.
a
w
.::f
c:
-
~
,DO I
::J "".17'" ~ It
_
.z
I,<9"
~I
.) I b
"i
~ p.<
I> '
><-
~ ~ ',-
0,5
1-]1,
_ . r"-
/
0,7
~/
r-,.
V' / x· t:5:
/-'..-'
~
I
-'V
5
::f
,
\A)< A
-'
Cl
;t'-~
J
L.
w
tv
IJ
.'X 1\ I/~ ~ I)< rx) I"-
-§(\y V [)])
"
20
c-
If
....
~ ~ o ..
I'\.
I'
J 1/
_.
)(\ I \\ ')c ~ It,( K\ ~ ~ 1)<
Cl
,
W
30
",
It~~(\~~'!..~\I·~I'\~\i)iVl!",~V~~MII·I~II~J."~~I}~~lla~v~IJ~·I/IJIh!I~ V-:I'J&-~ I\](
50 Ii \ I "
u,
o
\\
.u,
-'
/-
V ,002,003 .005
.0 I
,02
,03
.05 ,07
.t
,2
,3
.5
.7
2
3
5
7
\0
lOAD DURA nON TO rUNDAMENTAl PERIOD, TIT N
Figure 3-242
Maximum response of elasto-plastic. olle-degree-of-freedo(l1 system for bilinear-triangular pulse (Cl = 0.261. C2 = 1000.)
e
e
e
•
"
TIME OF YIELD TO LOAD DURA nos, tEfT
'"o
~ .0
U'
0
.... ....0 '"
,- '"
U'
,
'>"
0
'" :""0
~VtNN~~
...
t.n :....: (:,
~
lI\N
0000 . ... t,p"'-lUl'" OOU\Vl
0
0
000
N N :..:.. o '" U'
'"'" '"
1/11 VJb~
500
2 0 0 ~+-ill-'\l>(..\
~
w
x
<,
E
z o
70
e: u ...J
...
t4;
~
. /\1'~ /\V Ill\,
w
"w
w , w o
~ I/I~
j~~AA:A--~
0
00
o
'" 0 0 ' '" " '"
0
o o
o
0 ~
U'
0
....
~
0
N 0
iPfA,ff-:;.f'-;;.oWH /
...<,z
~LY
~
'" 6'
..-'
~
7/
'J-
v
'6'
o o
...J
w
>-
o 1Z
._V.U __ A
o
,_~v
.-
.
__ ._-------
t:
u
W
.-o
...J I
W
"
::>
• ...-r-ll
" "
>->:<
.-.
-..
Lb
.. ••
~
I
I
v·
.,...
\~
I.
.
.~
,"-. OJ
;:.,::::;
1~_"=
0.5 ,00 I
,002,005 .005
.\ I ...."
/'
,0\
. ,02' .05
.05,07,1
.2
,3
".5
.7
2
3 "
LOAD OURA TlON TO FUNDAMENTAL PERIOD, TIT N
Fi gure 3-243
Maximum response of el;sto-plastic, one-degree-of-freedom system for bilinear-triangular pulse (Cl = 0.237. C2 = 1000.)
5 7 ,\0
....E
m"c or YICLD TO LOAD DURATION. tclT
...o
..
o
~
V'
~
r.a
loP
AI /'I. V\
N 0
\it.
::
011
en
00
..
Uf,N-
NO..
IV l>{ IX 1)<
><
E >< z 0
~
bl:..
r:..-
~
<,
NN-
~
I-')(J\ l{V.V
w
UIi
OOUli""""'l»v. 1lD
~~\~--Au1~n~'"
I/MNf'A·l(
oob'oooooooo ·",··· . 000000
_
N'O
:...a
v
::-=.
~
-0010"'\1'
aU''''''''' .....
_.
.~
00
...... ....0
Vl
0
s
.9·
I
... •I
~~H.II ~ IY'T ,u.-
-'
~~
i'-.
u
'"
0
6' .
a:
.,
oJ'
'J-
50
W
.... ....0 0
...
0
u
...J
Z
':oS'
L.-
0 0
....E Q.
70
;:: W ...J
~'"
z
I-
<,
30
u
::f
':0
'z
.l: s
'2
0
.1: 0
I-
Z
<.~
0
<.j
...J W
>I-
c:
u
W ...J
7
W
5
<.0 /.7 /.5 1.3 1.1 0.93 0.81 0.70 0.6\ 0.53
......
L.-
0
::f
::> ::f
><
3
2
0.7
1\
0
Z 0
t: u u
...J
L.-
u 0
::f
::>
::f
><
'« ::f L.-
0
u ::f
;::
l===-
0.5 ., .001
.002.003 .005
.~I
.02. . 03
.
.. 05.07.\ .
.2.J
.
..5 . 7 · I
2. . 3
5
7. \0
lOAD DURAnON TO rUNDAMeNTAl peRIOD. TIT Il
Fi gure 3-244
Maxi~umresponse of eiasto~plastic. one-degree-of-freedom system for bilinear-triangular pulse (C1 = 0.215, C2 = 1000.)
e
e
e
--
TlM( or YI(LD TO LOAD DURATION, 01'
~
o
o
v'
....
....01' ....0
__ V' -
./~
-
'" .... ON:'"
CIOcn~
.
o
. 0
• '"
NtD ..... "" .. 0 0 " - " \It
d\--t\lI-Hr'Ir--'ll!--i'V--fuJ'.ch'
17\rk1l'Xl bt' IX IN)
~,
x
z 0
""'........ ~
70
--.oMM\,j-J,I"""A-~wf--\.F"
-oJ
w
,..
~
0 f-
Z 0 1-
U W
-oJ L. W
0
OOlH tD
--
:E
...... 0....0
',>
lI.1.? \~
-'
'/
6"
',>
v
I/l~~
i-
'6" ~
'0
~y
5
VI
~ k\ d·~n\ "
~
[)(
~
~
P'"'I
p./r
1\
-'~
---I
.1-
.001
~~llfliJI :'
.002.003 .005
.0'
.02
.03
.05 .07.1
.2.3.5.7
•
• ,t
I
II 2
I II J
·LOAD DURi\TIONTO rUNOMIOUAL PERIOD, T/T M
Figure 3-245
0
w '"
Maximum response of elasto-plastic. one-degree-of-freedom system for bi1inear-tri'angular pulse (C1 = 0.198. C2 = 1000.)
5
7
'10
-oJ
....4'
·3 ::f
«
a
·2
....0
1.5 1.3 1.1 0.93 0.81 0.70 0.61 0.53
[J\'- Ik' .-:-
3
___
a
.J. O
1'X"-ll .I1'NU \I nt'l 1.7
[Y"17'~
0.7
...E
...:>
2·0 / \'Il( \
2
....z
<,
J.,s
<'6 <·'05
7
05
o
V4...,. ..... 0
0\\1\'" VI"" U\
~
x
«
,::f
00
VI~
~
:>
0000 0
-01
lI'N
IIr
-oJ L.
'", 0 '" '"
....
Q.
W
a a
h£.
UiNN-
.J' 'J'
t: u W
~
v
<,
E x
t[/T
_~~~?'~~~~oo ~o
l/
AI'\i1 "J
e
2 0
i= u \oJ
... ..J
\oJ
'0
::f ::l
::f
>< « ~
...
,0
w ::f
i=
,'.
'I
•. ' ' ' '
"."
TIME or YIELD. TO LOAD OUR .... nON. tEfT
'"
o
500 300
w
:
<,
U
20
>< 'E ><
z 0 ;:::
u
."
,-' u,
'w
...,
.....,. 0
70
50
30
w >0 r-
20
z 0 f-
U W
-'
u,
w
a
:>:
sc -c :f
t-.l
y-.J
l.1'
N
0
:,/~;:,I~ ~f:\oA"'" 0" .
~LL. 1&-,
_,011
IJI
-
.
2
0.7
Jx Il 12
~12
I•
"V •
I.tl'
.
.LJhlJ
-; iLl
.r:
- : 17
.
'..,.g ,
I:ol~
r-", "",. ",po' w-:
Di; , ..,," I;'c'
>"K
Ht<- r, r 'W.
"
'A - -
,
,
..
'~
'c::,
I=1 = - , F=-
• I ", .
eo r
J,
e
" .
.01
.02
:03
r;
:"
6'
'" -s
"•
'J
~'
'6'
~b:
.~~" s:.
+,05,07.1
rt
.
..
~
a: w
11.
...J
-< z
>-
w
:E
-<
Cl
"-
<'IS
<,.1 <,0 I.]
1.5
~"
<'< ._ Co'. pl'L
Cl
o
.J. s
" " ' - ' . C'
"".D
....E
%
.J. o
L
z >-
<,
W: O
.
."c
[l, C
,
IX: y tl
.1.
:>
o z a t: u
>-
w ...J "w
a
:E
1.3
:>
1.1 0,93 0.81
~
0,70
0,61 0,53
:E :E
"-
O.
w :E
;:::
<::"4-
,2.3:5
LOAD DURATION TO rUNDAI.4ENTALPERIOD.
Figure 3-246
..
..-' B: '" -" •
~
.002003 .005
",e
~4 ~ ~ ~ ~.~,~""'~~ ~ v~ . -'
'X C ' , " '
r uzr
-'
'-"
~ '" 0/ " ... u c1 " -""
."
<9
~ ~
"'.
"',
~
u.,Q
4- "' . .
~ ;j.J:. •
,
r,
~r:sl's'
,..tILL\. 'i=-
v.' j ""
e;
.
!'oJ 0
N ...
~" r, ~
¥:~ i-.Ll
e,
7
0 0
~I' Ji . ::::::.. ~ z ~ "~ ~'A' ~ ~.. ~ ~ ~. ~" ~"'->u
IX'.
~
V
"",,
V.,.[
. "".
0.5 .001
~
rJ
W. "'-'>4 L: ,
~'
7
3
IL
? ?
?
1\'" ' ' " :xuu...
10
5
~!,,>.
N O N ......
'.Jl
0: 0
0 ~ ~ 0 ~ ~? 0 0 0 0 -0 • lD"'" 1Il ... lH N N.... -- CllI 0\ VI .. VIlJl NOOlllv.""COlHWlIlNOOU'lV"U\-.l 0
..... _
~
en
00
.
-'lL'.p '
F'.Ji',
-
"V~ "", ,~.
-
I I"
:f
::>
o
"e~ il, .. I'¥Lj" I~ISv~ LiJ!.-'/"1L "t[, f\~ l I"; V , I1\v,", . ¥ " i>I:' ~<. " 4 kilt V/", ~ ':::
100
a a
-'
~ ~ "
u'
po
i
,7 -
T/T N '
'2 " 3
5
7
10
-
Maximum response of elasto-plastic, one-degrec-of-freedom system for bilinear-triangular pulse (CI = 0.178, C2 = 1000.)
e
e
.-
e TlM[ OrYI[LD TO LOAD. DURA nON, o
"o
sbo
o
1.J
r--.l
N
0
""
",
, 1~\T'1J:: Y~
U'
0
0
0
.... Vt N _ _ . .. • • .' .••.• \D ...... UI ... 00 N O N ...... NOOUl~
to·en
......
tdT
-
~VWI ')<
1/
-0 0 0 0 0 0 0," ... ... .. 000 0 0 0 v.NN-- - 02CJ'U''' vt. \" "'CllI\I
iNJII IYII
. 0 N
. 0 N
..
0
LI1JK llt--tY \ II'
.. 'm,n"
I
1/1
-l« IN'.1I x\I IN '
w
'"E '" z
-~
....
~
IX
w
-'
L W
o a
'~1~
-'
w >-
.,'~
o
f-
Z
o
.s.
i-
~ 'R-\N
. .
---
-
i=
.__
....l
....z
w :20
a
z
l)
y --
---
'J-
~\\
~
U W
'"
~i4V\l\
..... ,. _. IV1'IIA'
f-
o
a
o 0..
o
...,, ...,
....E
____________________ 3
-~~~
<,
....z
<,
~~ ._--~\~ I~
:oJ L..
....o z
o
t: u
w
u
...J
w
-' L
- ----
W
-~
-
-~
a
" !
.-nrJ.....-
y
L
.....
"
r"
a
:20 :oJ
:20
:oJ
'Y
.."""
I\
J\. \I\. 1\ \ c,
-1-'"
.00 I
,002 003
,00,
01
.02
..x
Maxim~m
:20
L..
o
w :20
c:
,03
,0' ''-07.1
.2
,3
LOAD DURATION TO rUNDAM,[NTAL p!=nlOD,
Figure 3-247
1M
~,
0,7
o.s
rt~
~
,5.7
1
2
3
TIT N
response of elasto-plastic, one-degree-of-freedom system' for bilinear-triangular pulse (CI = 0.162, C2 = 1000.)
5
7
10
. ',';'r i
"
"1','
~"~'4"". ~"·'·'."·I:t"'~W~·-'''''·''''·1:n·'¥ -t;t.~i·.'"
I \ '"
•. ,.
TIME or YielD TO lOAD DURATION, telT oJ'
o
u.V'
.!'DIIPbI--->,j\-f-ll,.<-+N
... --;;:;;;.!'lIl! I
. .z a tz
\q-
17 'v....
~ t'f«tl7'~
~"""J- P>.4~-]f--
~
,n '.JII"<'-+'.4'''''A-+I:;K.~
u
W
mI
.~
~,
..-
M
Z I-
.¥
<,
,E
.... o
I
J.--n\/
a
t--.lA '/\
It:
w
e,
-s•oJ'
v
w
::E
..J L-
'< a
W
,
w W
o
'"
o o
W
;;
L
a Iz
'"
1'1. ....
u
a
i=
u
z
~
..J
W ..J LW
7
~
I.-
a z a
I-
..,
t:
W
...J I.-
'"
W
a
.fC:
o
:2" ~
:2"
::!:
s
=>
::!:
>< 4:
::!:
...J
« z
I-
:2"
fIT
2
8'
I.-
:::\1-\1\
a
w
:2"
;:::
1'\ \I\. 0:7 0.5 .001
.f-"
.0'02.003 .005 .
.01'
.02
.03
~
.05 :07.1
..,~
.2'.3:5
,7
I
. 2
.3
lOAD DURATlON TO rUNDAMf.NTAl PER10D. TIT N
Figure 3-248
Maximum response ~f elastv-plasti~, one-degree"of~freedom system for bilinear-triansular pulse (C1 = 0.147, C2 = 1000.)
5
7
10
e
e TIM[ ~
0.
..
0
V'
t-.l
t..J
lJ'
t..J
0
lJI
-
Or YI[lD TO lOM DURA TlON. trlT" ODCJl
•• 00
...
00000'000000'
l#IoN--······
.....
.I 00
.- .:-'P'rl~~--,\\,...--l~f1 MH~~'i;-}\jJ
200
!l=::\f-l----\:N
II. 1/
___
,,,,tt
, ¥-'1rlb'ttlA-li"<1QHM--I
><
<,
E
V
><
00000
, ." "" .... ua ... l,If.NN- - - CD O'll,ll'" Ul NQ""'''''''OOUlllo<",CDl.ItloDlftNOOUlUltJ!.''''''''
h'-
w
• tool
Ul 0
----
if III
to-)
...
~m
z o
0
~
z >-
....)
~V\
o
o
'"
W
ll.
.s.
;:". u
',j-
I~~
w
....w-'
...., .... ....o
.
000
w
:2'
IL
a
CI
2!
CI
w
....::>
ft
1/
-'
o ....
>-
o
>-
-' u,
.;J .... w
rxt··
o
.9
.::f
::>
:2'
::>
><
F u w
1
w
:2'
z o
/
z o
F u ,w
-'
.I !
2~,
1\
r-
"_ II
.002.003 .005
.01
.02
.:03
o
w ::f
F
l..--1'"" I
.05 .01.1
.2.3.5.1
I ..
2
3
UlAD DURA nON TO rUNDAM[NTAl P[RIOD, TIT N
Figure 3-249
::f u,
~I\"
I,"\j~
0.7 J~
0.5 .001
::f
X
Maximum' response of e l asto-p last ic , one-degree-of-freedom system for bilinear-triangular pulse (C1 = 0.133. C2 = 1000.)
5
1
10
nME OF YIELD TO LOAD DURA nON. tEiT ·00000-000
'"
o
po
o
V'
....
~
l1'
f"oJ 0
U'
a
00
DltOl' ... lJo4N--····· ..•.... •••••• I.D "l.n", \.MNN_ 0 0 N O N " N 0 0.01. l.H ... 00 IN U) tJ'I NO
-
III"
n
...,
I I
.....V
II "W"L VI
0000'0'0'0'0 CJ)C1'lUl~~,",,~N
OU'U\U\~'O
..-0
~
III
-z r-
~ __J~ J
,r :~~~ ._- ---- ,n..I,.i 'roliiiii'iir"ii7
_
VII 1..·1 111"/1 ";t." I. 1/1 'N:J II V IIIAlI3L
<,
....E o
o
'" W
0..
o
-'
~.
-c
r-
t\
U
Z
w
w
-'
:::t:
L-
W
a
·0
..., ...o
0>
z => a
o
-'
.~
w
r-
>-
,o..
,,~
Z
o
--
ru w
..
L-
f-
z
a >:.
-_.v -_.
u
w
>'<-.Yf I I·A!' lOY I
...J
II~
"
......-..-. '+q:>4\'f~""':'-'~44c~4loo!=Fl"A.d'~!:;K~
-' LW
0'
.
~
i
=>
:£
~.
><
LW
a
:::t:
=>
-
I\~\I
,
:::t:
n.'"' , ___
1\ I \ 1 \...1-1-
~
:::t:
\ Ir
:2i
L-
a
Fil\
"'" p/r
~N",\I\ - - --
.
w :::t:
=' n-.
>:.
., 0.5 .00 t-
.002.903
.005
.QI
.02
.03'
.05 .07
.1
.2
.3
.5
.7
2
3
s
7
10
LOAD DURA nON TO FUNDAMENTAL PERIOD.T ITN
Figure 3-250
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (C1 = 0.121. C2 = 1000.)
e
-
e ~
t...lN-IJ' 0 V' -
~
po
0
00000000000"'" .. , 00000000 .. "'"NN----ClDCTiI,1\~V4~t_.lN N O N " ' - & N O O U l L H .... OD(,rI,!,DUlNO Oll'lJI "'" '-' 0 .. 0
caCJ\
-··· . · , · · ·N I o-D ..... Ul
00
500
11. 'f; '"
()
)
300
1-l--'lLV,K
~I~~,"
200
><
<,
><
100
z
70
0
c: u w
...., .... 0 '"
-' >0
30 20
I-
Z
0
\0
'I--
U W
-'
LW
0
7 5
~
::J ,~
~
.""
Vl
:-YC1l
., .
~
I~
~b1"-'
l~ ~ 1/ r-
~ / .1'
/ 1 7
I~n
r-
'll
r-
li: f'1 v Vi
V
2'" . 111""': I/,~p,( I-'" VIP' If ~ I"·
I II . fI.,/ c-....)'
~IT
I_I
/
Ii
!"
1/
i: l'"'[
~
k::l
~
v.
r': fA '.
"" ~
-«
l,.!) I-;
If
'.-'li
:J:-1
'\.
~V It VlIAIAIII L:.Lj,.
,
-=
0.7 0.5 .00 I
l'
'.J'
~,\
":",
~/ IX \V n I.... ,/
cI.
.•
t::
~
V .L'< K
1/
1::::.
V IX /11"..i
,.....
I--
--
-\" =.ill f-'
\ '
,.'I
~ It'S; '..\i ~ IX122'
v.1--L r7 11..
\--::
t:
u- -r- :::l: fr.l-,~ ?-'I' ~41-" ;f"
:~f'e'><~ ""'~""L> l'-:
,002.003 .005
.o r "
,02
,03
IL
f.,..\
l(4Q
, '.J'/';x I)
~V' 0~.Lt;
. 0:::::::
.s
1.
":0
V
J.s
I'W
~10 V~
.f.
1)<" ILl.:: u.
r
.05.07.1
~v~
»:/ X/ V/ v
.2.3.5.7
(\V
l/i': ~ .
~ 4:
''1
':.v'I
'l:
Z\
<.J
Q.
J.ll
;' p
, 3
lOAD DURA nON 'TO fUNDAMENTAL 'PERIOD: TIT N
Figure 3-251
Maximum response of elasto-plastic, one-degree-of-freedom system for bilinear-triangular pulse = O.llO. Cz = 1000.)
«.
.E
..... a
o Of. W
..
5
7
10
-'
I-
Z
.
W
~
o
~ o IL-
Z
o
i=
o
w
<·0 1.7 /.5
....-'
1.3
::>
.1.1 0.93 0.81 0.70 0.61
x
O,~3
_
2
o
e. e
L\
7/f
Z I-
<,
Q.
.s.
..::0
~ ,~ V F:?A-rl- .....~ 1'~I\ 22 'S' t:'(,
.,..
6'
[.,...., ,
17~ v ~
.....
-' ",.
6!" :~A. ()
_
-s
.L
.?,-
k' r- .
I"X--
',.,
l"
[[' /
y VI)(
d'
r-
'rT f'"
bill ''
J1
,. - '
~
V --:-,/
.k" 0'
17 i"
...!J It-
J
~
1\
L( L. y:
2
}< t':
D(
-(J .'\
~k\f ~ rvil)l/Kj2
ki
J)
3
><
.«
'j"p<-
1:'1.
~,N/III
-I'" -- \""\ t'l'
50
0
w
~\ '-.:\I I
~
Ii, 'tIILYI
1)(
N ~l'fo' )(1 LL !-l.
-'
LW Cl
rlY 7'(;.. '~
'
w
E
•
TIME Of YIELD TO lOAD DURATION, t(/T
w a
~
. ~
~
.... o
w ::f
i=
TiME or YIELD TO LOAD DURA TiDN.
t.~/T
00000000000'
v-
~
o
o
.....
~ lJ'
t-J 0
., .t \
l/. 1\
w
-
-
U'
-
U'I'"
(lllIa""~N--""""
. • • . • - •• lD ...... u.NN00 N O N ...... NOOU'll.N ... (lllI~lD
-
~
E
x
.
00000000 00 (fI ~ . . V4 U4 ~ tool 0 U'U" U"..I 0 ... 0
L.-\'1
~-lJJI
~.
I--MXr\Zl~J
~I
-'
u, W
o o
-'
w
W
5
.
~
~
.1" .01'
. . ..J
I-
Z
o
M
2:
1:
...oz o
;:: u
W
w. o
?>
~
~
~
~
2
",I -, '.1 \,
---J--
0.5 ,
.002.003 .005
.01
.. 02
.03
.05 .07
.1
.2 ' .. 3
LOAD DURATION TO rUNDAMENTAL PERIOD,
Figure 3-252
-c ~
~
t..
o
W
~
;::
~./v·
1_
·.001
;;;
~_E
./
0.7
w n.
t..
3
x
~
Q:
..J
~ ~
o
:::E
z
o
0'
w
o>-
:J. u,
<,
l'\
>-
7
...z ....E
N
-
70
w
o t: u
\
~\
u
...., .... ....o
-
,""NO
~~
~]D(
X
o ;::.
•.
V I VVfl M/1U1I
IX\ IXl YI'vI1 JtI/
<,
z
-
.5.7
2
.I
5
7.
\0
TIT",
Maximum response of elasto-plastic. one-degree-of-freedom system for bi 1inear-triangular pul se (C1 = 0.100. C2 = 1000.)
e
e
e
f,< .
e
,.
TIME 'Or yiCLD TO LOAD DURATION, t~/T ~.
.... ,'O .. ~ ....<1'.0
~
~
... en
",'lI
00 N
N~
0
_?~???~?OO.OO '~""',",,"',""NN---
""'OOUI
"'CDI,Itl.QJ
Ut
I11'VlV
I N(JII/
200 w
v:
\
<,
100
z
70
0 t:
u
w -'
l.W
9
...,, ..., .... ....
VI'"
1/ r-,
50
·w 1- IV r-, rI I') I~ II IY . . . b
30
~ b<
'" >-
20
r-. ...L
I
17
D" t--
f-
10
f:.
L,..;IJ
V
'1/
rJ\h VI \ 7t.~JXJ\
xr.t·1\l\
,
)'"
II
1t:.I'l..-1 l-'
,
~
1/
~J\~
.u;
II
;-
r\
l.--/l:.Ii 1'1\17'
.f"
~~~~
...
oS"
',f
rx
~cr 1(0
~J~
j'
'S
./.0
r-
~'6 ~:J 2·0 1.7 I.S 1.3 1.1 0.93 0.81 0.70 0.61 0.53
i=
u
w
7
W
0
-' c, Cl
~~
•.t.
~
::> ~
>< « ~
5
'AA
2
I
0.7
-
"
IS
?
0.5 .00 I
.002.003 .005
.01
.02
.03
.~ .05.07.1
.2.3.5.7
I
2.5
LOAD DURA TlON TO rUNDAMCNTAL'PCRIOD, TIT N
Figure 3-253
',., ... Cl'
tV \ \
~
d'
Maximum response of eiasto~plastic. one-degree-bf-freedom system for bilinear-triangular pulse (C1 = 0.091. C2 = 1000.)
5
2 I-
<,
....E
'/
1\
r;; rl'-:7
I..,;·'?
r-
v
~ I~
[:?
f
~
10.~L.'
h(-
Z
Vl'1 i:
~V'.
0 0
t')I
L\-1\ XI\"
t1'
r/fAJ...-,
I ~.1
Ii ~ V
a
-'
0'0'00'0'00'0 0:1 (J'l lJ\ .. '" U4 tJ N ·0 ..,. IJ' \I' -.l o· .. 0
r-t
t1I..1l1
~
XC
NO
1111' rWIt I1"rflAA' II
500
'"E
VI
7
10
Cl
0
w '" Cl. .:J
«
I-
Z
w
~
«'
Cl
z
::>
L..
a I-
z
0
;:: u w
:.J
L..
w
Cl ~".
::>
:::::E .. ~-
X
'« ~ t, 0
w
~
;::
tdT
TIME OF YIELD TO LOAD DURA TlON,
o 0"0 UtN--········ 0
oJ'
,P
o
500 300 200
I
U W
50
--' L, W
"", ....""
.
0 .0
--' .W
>:-
30 20
0 0
i=
u
10
w
7
w
5
c'. 0
- ¥\ s-: --;j~
::E
3
X
«
::l'
2
,)
...
.•••••
00
'\ -
•
ID
..... U'I
r--.
ff<·
'0 0
0
o·
0
.
J-
I
V vl1l/~~
.!IW'
. ,/
II1.A'I
'\
x
'J ~ I' /
I?I:<:::
,/ /
r/ kj V'
0:5
'!-
,00 \
"»
v7
t~'~fi v
/v.
it""
t/ lY rf . :;
.
--
rJ r-..
I.
I-.l 1'\'
K
I ":-"
-I..
.~~
II
~
1/
I>
../
;-
A~~ /~
~..I
7 .., 11 . '\ 'j- 11' r--y-'y , 17 .
IV
~ .....
-~
b<:~;; ;"-~fj '1-
I .vI-
\\
L.f
I
'I<
'"
E
'"
IS'
.,..,
/R< 7>/
p/r
::..-'<
~
\..-
\
I
'i:~'~
'.002.003 ,905
, Figure 3-254
,.0\
.02 .03
,-
V/,VA/'/V
.05 ,07.\
.2
,.
,3 -,....5
LOAD DURATION TO FUNDAMENTAL PERIOD. ., " ....., ... " , ' .' ~
.7
\",
2'
3
TIT,.. -
Maximum 'response'of elasto"plastic, one-degree-of-freedom system for bilinear-triangular pulse (Cl = 0.083, C2 = 1000.)
. 5 ·'7 -'10
....z ~
« z
J: s
:::J
'" o
o
.... z
<.{;
o i=
J
1.7 1.5 1.3
'
....« o
<.0
.......
IX .... 11.
t: o
<..1
."
o ...J'
'6'
-. v
.
~
'-/1'-.
,
c
.
'oi'
I/~ K
~~
,X4 AVO/'$:
"'--~ kS<:
.!"
.?:'-v.
~/
&. Jlvl)~
~ \:1--'1
I
1\
t1 / P / ,0'1\
..>
II \
_1)\ X,~'
V
I?
<,
.s
/\:"-
1
1-;; ~>
IPfI
,'\.,\
1\
t- .
'of
....z
,d'.,..,
/. r:. I"0 x .;;; to-... 1\
/v
t> /
"-.,r/.
~;>;><: .....
~I-:::v
<,
');-....!L ' V
1/;><-
~\. -
.
11·
~rJ"V / /
VI
~/ t..... <,
-I 1<.
r/,7
l~~
0.7
-I
~~G: :V / IX I>
I~
.
t-
A;'
"
.
-)I
ik-
II'
'x
.
T
,I(
l\i
.
--
I~.
IIY
'j/
""NN--
...
l..-;
-,
'X
0
.00000000 aD Cl"tJ' .. "'" ~""!"oJ N O N ..... N O O \ l l l . H ... OIIJI,IDU'NO 0 U'tJ' \II....t. 0 ... 0
(lltcn
-
.~ ~
II "
-
::l'
::>
0
~~~'-J~IX'f
1/
f-, Z
,.,
IJ\
\
I~I/\
E >< 100 70
~
f'-l IJ'
v- II' V \ 1 ' "X! .Jt\ 1\ II 1 K\ lrt...
~l\-I-V..
>< -..,
z
0
,~.
w
0
V'
0
I. r 0.93 0.81 0.70 0.61 0.53·
~
t:
.... c ~
::> ~
~ ~
'" 0
!i >:'
--
e TlM[ '-.I'
po-
00
cr-l
r-l 1J'
..... 0
-:= LI'
or YI[LD
(.OOl 00
•
IO LOAD DURATlON, tdT 00000000000,"
... v - . N - - · · · · . . ~. o.O"""V1,. NqN
NOOVlv..
'.'
. . . . .
. -00000000 __ CJl (J\lf'''' Vl u" ~~ ClILNU)U\NO 0 iJ'1J' IJt -.10 0 ... 0
.... NN--
W'rX1 Yl\ l'\t'l<1 \
i'-"'A-ttrl VYlV1l./
d' 'r'
~/t11::{V~-vj¥t'i2
"-
E
'"
r ...
~
'0
o
a: w c,
z 0
.s
I-'
'J'
~
u
«
w
'-"'~~
L
U C)
a
-' ' >0
:f
« z
o
u
z 0
1="
~
N
l.-
o
~v
~
,..
I-
z o
t: u 'w -J
LJ
u
I.-
<,
-'
_._-
l~
U
.. - _.
W
o
--
:f
a
;:>
:>: :>:
-J I-
z
u -'
J
<,
....E
'-" xC
'", ..... '" '"
Z
I-
:f
.I
x
~~~M~~~'\
-
4:
.:>:
OJ 1--.
o, .00 I
.002.00.5005
.0'
"-
.02
.0.5
:f
w :f
i=
V /VA/./.
v /A/"/ l/'/ V .05.07.1
.2.3.5
',7
1
2
3
LOAD DURATION TO FUNDAM[NTAL PCRIOD, TIT N
Figure 3-255
X «
Maxfmum response of elasto-plastic. one-degree-of-freeaom system for bilinear~triangular pulse (CI = 0.075. C2 = 1000.)
5
7
10
.,
TIME or YIELD TO LOAi;> DURIITION -. U'
p
00
5 aa
vl
t-J
~t-J
__
Ol(tl~
""'0
U\-
OON
~g,::'TVl\Y'-r<
'" o
t(/T
_~??????OOOO'o
..,N:...a
•
N
\D~V'I"'VlNN-
O
O
UI
v.. ...
c:alJol,loD
u\
--
NO
o
·0 '0 OJ
0
'"'" '"'"
'0
'0
'0
...
u'
V'
'"
...
0
'0
....• ....0
J O o l \ l, \ . . ¥\f>1t'H'l~rffi\;l--,f\~
_.,
'DO
d'
'7t--\II1l4
[\fl'l
u
E
".
100 <:
. I
17IlrI~
~
u
~
u
....-' w
l7
. .V
'¥~ .
0 0
..,......
....E
-' w 0>-
I·
0 ~
~
1/11'0-.1.-1 1/
-s',r
I:i
'f'
LA
'6'
0
""
...J
«
.... z
w ::f
«
z·
1/
::> l..
o ....
VI I
Z
o
;:: u W
t-
W -' u, w
Q:
W
o
1'>r ill1.Jl./111 I W-UV
z U
o
o o,
z
w ,
-'
.-.,..,
0
W
<,
'/
'" -- .
....z
.,..,
7
- -_.
_. .
- -
..
--
. -
_.- - .
....W
...J
-'"
o
::f
... _..
0
::>
~
::f
::J
i<
-c
:::E
"::>'"
::f
\~
.... o
w ::f
;::
po-0.7
~
0.51-00 \ "
.002.003 .OO?
.0\
.02
.OJ
.. 05.07.1
LOAD DURA nos TO
Fi gure 3-256
~ax;mum
~ ,2 . :3'-'-.5
rUND~MENTAL
P£RIOD.
.7: "
2
3 -.
5
7,
10
ilT N
response cif elasto-Plasiic, one-deg;ee-of-freed~m' system for bilinear-triangular pulse (Cl = 0.068. C2 = 1000.)
• . •
,I.
.
.
,.
e
e
e
•
,
TiME or YIELD ro LOAD DURATION. t[/f t
-so
•
0
...,.
~
t-J IJ'
~ 0
... \I'
= c:o 0
i!'
a
~
~~:-":o'~"
N
•
\
,
o
0- 0
0"·00 0
0
0
LH.
~:""lJI"
N
U'I
0
lHNN...
CJI \.H
f.D
00° --:lJl N o.
--~H·1lb#1'W,o¥-bV''thf-lH
0'0.0'0'0 0' 0
CflIJ''' U" IJ' IJ'
"'" -1
'0
'0
o
'0
..
U'
'0"
'"
-4\l:=~~~~\ z
>-
<,
E .."
w X <,
x
o
o
E
<>: w
:z
0..
>=
<:: >:z w·
...J-
0
u
W
...J
::l;
L
W
<::
c...J
'i :J
a ,
'", .... '" '"..
~l:1-"\.~\F'II--IP'5l>
W
>0 0
>=
o s:
10
u
.:;,
(j W ...J L
7
W
5
C
....J
l.t-r/
::l;
::l;
2
"
.
0.7 .002.003 .005
.01
......
"-'""
.02
v .03
~
.. 05 :07.1
LOAD DURATION
Figure 3-257
::l;
x· <::
::l; L
o w ::1:.
pjr1;;'n..
J--- p
1~
0.5 .001
C ::l;
._1\
3
f--=:::='---
w
:J
X
<::
L
"
~ ~:l>"f"-l\>;.-r:::
::l;
:J
>-
:z
t-.
:z
L
o
to
. 2 . 3 · , 5 - ,7
'c
2·
i=
I I I I III' 3
rUNDAt.lENf AL PCRIOD,' r If N
Maximum responsecof elasto-plastic. one-degree-of~freedom systemfor-bilinear-tr.iangular:pulse (tl= 0.056,.C2 = :1000..)
5
7 . 10
TIME or YIELD TO LOAD DURA TlON;
.
o
,.., '"o ,..,
~
C1'
-
~ ~
-
00
__
~ ~ !'J.
NON.....
.
,f'.,)
~????
0 \D
0
tclT
? ? ?'?? C? '?
.... lJ' . . . . Vt NN
0
(I)
CD 0
..
CJ)
'""
U)
-- -
0\" ~
'0
0
'0
CP 0
(J\
\I'
'0
'0
tJt If'
'0
v'
.. !J'
_
'0
'0 N ....
U"
0
0 N 0
~ I
z
I
>-
<,
....E Cl
o
'"wa,
z
oS
o
'of
I-
U '.J ..J L.W
\Yti
Cl
...., .......
'"
Cl
o
~ o
1>
.hI
L.-
>-
o
i= u W
>-
z
,
"1~IJl
,0
.1/
\0
i= '-' w
7
L.-
..J
W
..J
L.-
w
Cl
Cl
5
::f :J
::f
""
5
><
::f
>Z w ::f
L.-
..J W
o >z
'..J
I\.~
2
::f
~
:I'
~I\
.1\"'
L.-
o
w
:I'
=n.
-: 0.7
~," ~
0.5 .001
:I'
=>
.002.005' .005
.0\
.02
.05
'//V/"r//
.05.07 . . \'
.2.3
LOAD DURA TlON TO rUNDAMENTAL PERIOD,
Fi gure 3-258
i='
'.5, ,7
i.
2
3
'
5
7
10
TIT N
Maximum response of elasto-plastic. one-degree-of-freedom system for bilinear-triangular pulse (C1 = 0.046. C2 = 1000.)
e
..
e
It
nME OF YIELD TO LOAD DURA nON, t(/T
.
o
t V'
o
f'-ol
,..-,","
t.>
t1'
OJ
... . -
0
O'
. a
...
~
.'.
~
N
-
: ; ,:..: . -
. :.....' N
NON
,
.
','
. , . t"
~ ~ ~ ~ .f? ~ ~.
? ~ ? ~ C? '? N
-".- - . ; - . . .
00011000..,011
Vt
I.OCJl"N
U): ..... l,Il' ....... "'"
N
0
~
'0 '0 .'0 '0 G" l1' ~ "V' l1" &S' \Jl ~
'0 01 0
'0
V'
0
·0
"0
t.>
.... 0
~
I
IJ
p
Z
f-
<;
....E
\H--\--,,,*,,,'I-I-*\
MAI-:~+,l;"
a
II
6'
.~
a
'"
w
Q.
.Z ·'0
;::
o
U ...J W
a a
"'.
0' 0-
lit
'z o ;::
10
w
7
.....w
...J
a
::E :> ::E
::E
«
z
::l L.,
l.--K
3
::E
;:: (j
W ...J
.... W o
g
~.~~~~A.t:tlh VI
><
«
a fz a
.~
1/
5
::E
.::l
'::E
~
::E
2
0.7 ,... _-0.5 .001
P;;I \
.'11
=
-
.002'.003
L----r~. .005
_01
.02
.03
.05 .07
~ .1
.2
L-
a
w ::E ;::
.,
.3 -. _5
LOAD DURA nON TO FUNDAMENTAL PERIOD.
Figure 3-259
...J
« ~ w
'0
~U.-J
JIll
...J
w
o
I ~ ,......
u,
'", ........'"
-s'J-
I~
(7
.7
TIT N - . .:
1
2
5
3 >
"
Maximum response of elasto-plastic. one-degree-of-freedom system for bili~ear-triangula~ pulse (CI • 0.042. C2 • 1000.)
7
\0
s-
TIME or YIELD TO LOAD DURATlON, t(/T
•o
u'o
o
,-.)
......
t.J
0'\
01 0
-
(fl.
• . 0
~ ,.
Vt
.' •. N·. 0
N • N
• .....
-
N,
00000000'000'0'0 . . . . . • •. . .. lO -..&. l1' ... • Lot N !'oJ - _ _ ... 0 0 DI OCt 0 .... 011 V' \D t-) 0
(J\"
,0
'0 fIJ'
••
o
'0 G'
'0
'0
'0
...,. '..- . V'
LJ'IIJI
'0 U'
o
lJ'...l
'0
0
...... ~
0
lR
.TI711:4/V 1/ BfllJ
. ·IHIM+
w
~iib
>< E
<,
A/I/
--
><
~1Vi
. NV\JAJ\
z o
II
>-
u
w· ...J l.. W Q
...., .... .... 00
,
[\
~,~ ~l>
Q ...J W
,..
o
y\-rV\h'~
>-
z o
1'-'
t:
u
~
W ...J l~
'w
a
'"
5
. /'( L~ X I
\i'Kl{
"
f\1"tYl M
.~
:f.
::> :f,
3
\17
v
')("
<.
'. :f
0,7
0.5 .001"
f'/r l =
""J'-
.t
--
---rJ
',002003'
III~ .... ci I
,9°'5 '
.. 02
.03,.05.07".1
.2'.3
.s
.7
'2
3
'5
7. \0
LOAD DURATION TO rUNDAMENTAI: PERIOD, T/T N"
fi'gu're' 3-260· . Maximum response ofelasto-plastic. one-deqree-of'-f'reedom ' system for bilinear-triangular pulse (C1 : 0.032. C2 : 1000.)
e
"
e
TIME or YIELD TO LOAD DURATlON. "",N• . • :.... NON
(IOCtl" P"
o
v' 0
t-.l ~
0'
-
. 0
• 0
, _
N
e
tdT
0 0 ' 0 0 0 0 0 0 '0 0 0 0 0 . . . . "." tD "'-J U' . . . . . VI' N N ..... a 0 02 m 0... 00 ~ lD C1' .... t-J 0 '"0
••
_.,
• •
. . . 000 ". G" !J' 0 IJ' IJ'
'0
'0
..
.ro
'0
V'
...
V'
o
"
• M I Iv' iNI A N£..l VilLI
"0
0
...
N
..
0
'~ Z I-
.......
....E
~
'"E
sc
~~}.'\(wV\
100
2:
o
>-
u
-.J
u. o
o-.J
,
• •
•
I
>-
.2:
~
u
1
« z
-XIV
m
N~t
2
o
I-
Z
o .,: <.)
~
L..
W
W
-.J
a
'"W
a
::f
.JV
.I
>(
«
"
...J
«
I-
a'
u
"
'"e,w ::f
n
.
>-
~
o
zw
,
o o
',.>
'J:
... -11-"Cff1I'=l¥dp.j.d~<\4j;~
W
w .... '"
'ii' ....L I .
.'11
W
d'
oS
-~'
W
w
a
J.l-t\ AI 1\
<,
,v .fir
::f
1\/IVI
~
::f
-I1A
~R
;(
«
:f
\~
2 ' t:?'\*~~=fI::
L..
o
w ::f t:
-17
0.7
~
1/1%1.-' ,01
,02
,0.1
,OS ,07
,I
.2
.
,.I
.5'.7 ,
LOAD DURA TlON TO rUNDAMENTAL PERIOD.
Figure 3-261
, .2
.I
TIT N
Maximum response of elasto-plastic, one-degree-of-freedom system for bilinear-triangular pulse (C1 = 0.026, C2 = 1000.1
5 ,'7
10
.
TiME or YIELD TO LOAD DURA TiON~ tdT p.
V"
00
r.l t-.l
~ (J'
C!'~h~-l"rI.~~
_
OONON-t
-
_? ~~~~?C? 9 t? C?,? «? 7' N~~~:;'6~~~~'~-;'~ '0
'0
~
o
0
'0
'"'" '"'" '"
5001'>1=,"'-
..
'0
'0 v'
....
'0
v' o
'0
'" ..
'" 0
Z 0-
~~~~~+tI~~4.h~flrl
Y1\
It
.zu-
<,
....E
'"....
z o
oS.
.--
I
fl .....' I f
U
'J'
If'"ll
w
~
--' c, W
a
a
w , W N
o
--' w
~A
In! Yft-.,
>-
o
0-
Z
o 0-
U W
7
"'\
..J
u, W
a
.~
::£
:::>
::£
><
~
'"
::£
w %
i=
0.7 0.5 .001
~ .002.003 .005
.01
.02
.03 .. 05.07.1
~ .2.3.5.7
I
2'
3,
5
7
10
LOAD DURATiON T~ rUNDAt.4ENTAL PERIOD, T/T N .
figure 3-262
Maximum response of elasto-plastic. one-degree-of-freedom for bilinear-triangular pulse (C1 : 0.022. C2 = 1000.)
~ystem
e
e
-
It TiME OF YIELD TO LOAD DURA TiON, tdT p.
o
500
v"
0
,.-)
t..l
00
..-
-
0'
-
0
Cf\ ~ lA N~ .••. -4
o.~
I\. y ~,-
N
U)
'0
C? 9-
lJI . . . . . VlNN OOOClOO.,. oa VI
....
1/
IV'
tJ)
t?~ C?
_ ........ (J'I"".~
C?-
'0
tP 0
0
r1Jr\!.-\-1\
1
~
j,
~/ rnJ.
w -c <, xC
~ ~ ??~.???
,I'
200
F
N
'0
oJ'
r---.:
toO
~
~V"V'"""N
J'
lJI...J,o~O
1-. I x: 1/\ If\--
Ir \Y
~I\
'0·0'0'00
'0
(f\
-'-'-'"""VT---'-'TOrnr-n'CrTT"7"17T7r~,.,..,.-"""r-r-r-
~-AJ,{~IA
SOO
0
.xJi'\-
M-c
I\~
~JI\
..>
'/
~~
.'"
6'
z
o
~;r
,u w
':6'
-'
k W
..
a
w
,
w
>-
....
J. S
-,~rtn~~1M
J. o
o
,-
<'G <..1
L
o
,-
10
u
W
)(1 )(1/\
-J
u, W
a
5
" x"
::J
__ •
_
n'~
R~ ~ OJ O. 5 .!_-,-'.l,_~
.001
.00200S .005
_, I.J 1:1
\ If
f""
~
2.0 1.7 1:5
y ""_,,,
~ '/0 .01
l/rf=';;;'"~l;4~m~~~~ .02
.03
.05, .07.1
.2.3.5.7
2
3
'LOAD DURATiON TO FUNDAMENTAL PERIOD, TjT N
Figure 3-263
"'-
....E a
0
'" '"z" W 0..
..j
w
::f.
0: 0
COl ..J
W N
f'-..
....z
Maximum response of elasto~plastic, one-degree-of-freedom system for bilinear-triangular pulse (Cl = 0.018, C2 = 1000.)
5
7
10
a
~.
k
0 ....
z
0
1= o w
.......J
w a
::f ::J'
::f
~
0.93 0.81 0.70 0.61
..... 0
0.53
1=
::f.
w ::f
TIME or YIELD TO LOAD DURA TlON, V'"
o
t-l
t...l
'
0'
-::::
00
(J\
0
0
r: .- ~J ~ ~.~ ~ c;:> ~ C? '? '? '? C?
....
VI
N
O. N
N
..•.•
.. '
tdT
Nl.D"""t.I\ ...... lr&NN- ........ ' 0 0 co oa 0, ~ 0' VI. ,J) CJl .. ~
-l
.... 0
·0
dI 0
. _ __ . _ ..
'0 a:r~
·0
'0 ..
I,JI
·0 vi.
'0 V'
...l
'0 '0 toJ t-J
0
... 0
--- ---
,
>-Y~m+7-lM;l-Jfl:J.ilyfi!~
w
."scE ><
z
o e-
U
w -'
.
'oo~~~~~MMg 70~
.•
PlI.nl
w , w
'"'"
n
56
'"
70
.-' w
>-
.......... ...,. •
'\fA
o w
7
.~
v
::>
sc
1\
17\V\~
10
. 5
. -Yl-V
L7
A
~\-W u,
~v "
.~
-'
::F
.V I \
1\
I ~ 1\1
J :1t'1
J
L~
"
k=ft'Th'Y.IX -,..." ""
~
eU
\71\71\:
A _ V_~
I.<'VVlA.A7 I IVJ.-\I-r\
oJ ......,..- I 111 t-*'''d'I.-l'H~I~I4~'*F~ihJ.4::I: "--t1"IYV .Y-7 U-Jjf I/~
o 1Z
. .. .Y'"nlf~
50
u,
w
.....-,
.:I"-
~~
1-~
::F
~
n 1 .........1 I
. ..J-.--0.5 .!~ .001
r
.......
Y
~_.
J-"
Lk.:l.vr--
.002.005 '.005
III/' 1/1/
/'
..01.
.02
.05
" /"v/Y//1/ .05.07.1
.
.2.5.5.7
..
1
2
J
5
7' 10
LOAD DURATION TO rUNDAMENTAL PERIOD, T/T N
Figure 3-264
Maximum response of elasto-plastic. one-degree-of-freeaom system for bilinear-triangular. pulse (Cl =0.015. C2 = 1000.)
e
e
e
e ) ...,.
~
....
• tdT
.rII;4E or YIELD TO l.OAD DURA TlON,
..-
OJ c:"
o~(I'-oo
_1"_ , 0 ~~,':' ~C?,? !"' ~,~.""'" ~N'\O . . " \J\ .. to vVN NON
0
0
til 00
0 .....
01
C?N c;>........ ~ t?? C? ..- " , . ; V"
""
"'
...
l'..!
'0 dJ
0
0
o G"'
iJ'
'0'"'l.1' \S'
'0
too iJ'
'0 v4
~
o
'0
0
'" '"
V"
0
~
0
'Z l-
do
',.,
w
-c
~
-1oYI-,HlH-'.,J·-)-.o;i
<,
'",
.
-----_
. >(
z
'0
.....
'V-A-'7I~-
u w
,. , I II '\l
--r-I \ n
\
....
0"
....
.
J'
:,j'
11\- !I\
a
N1
1\/ 171lV
":0
0
0
J: o
w ;:
,IX
" ,,~'" ." ./~.rX1~
0
w
7
,,~,.
--'
'"-
w
0
5
:>' J
:>'
~.
:>'
lX'~\1<
~
'3
.~
2
l)..C-.P<7't"
~ 07 _O. 5
'. . 001 '
--
,,~
,...01
.002.003 .005 G·
...
Figure 3-265
.02 ..03'
v
~'6 ~'.J ~·o
~
sc
«
'« .... w
J: s
....
.--'
":",
--'
z
W
:>'
«
"rY
.0
s-:
'" z
--'
w
0 Q.
u,
u
....E 0
~
...,, ..., ..., '"
<,
~l~: p/r -
~h
V/Qff/.7l7
.05 .07.1
.2'.3 -
.5
.7
.2
3
LOAD DURATIgN TO rUN~AM[NTAL PERIOD. T/T~
Max.imum response of e las to-p le s t'ic , one-degree-of-freedom system for bilinear-triangular pulse (C1 = 0.013, C2 = 1000.)
5
7.
10
"Z J
L..
....0 z
0
;:: u
w
--' -L.. W
1.7 /.5. 1.3 1.1 0.93 0.81 0.70 0.61
'0
0.53
;::
:>' J
:>'
~
::r
L..
0
w
.::r
~
:'
~
.
v'.
o
t-J ~
G\ ... V4 N • • • • '"-..
>p
O'
#~
0
~ •
J
0;--1" •
,. ~,
~"1
~-i
TlMLOr YIELD'TO LOAD DURATION. "t[/T
'. -
? ~?? 0'0 C? C? '? '??'. t? .... lJ\ . . . l." N t..J .... -
j.."j-lD
0
NON
0
CJ)
co 0...
011.
"'"
o.D
t?
~
(fI
0 rJl
0
0
0'0 (J\
tJ'
1.1' tJ'
0
~
'0
'0 vol
u"
\1',
...:
'0
0 N
~
~
0
0
-'~"'------
"
;.('"'
._'.".~
<,
E sc 100 z 0
c: l>
- -
I.
n'lQ
ItI)A 1:'-... 1A
"-"~I"'~""~"
I lL-.
r-,
70
VI?
50
IA
-
,/\ IA
LX
c.J
-'
L••
W
a a
W ,
'" '" '"
-' w
30 2 O·
>-
I~~lttt~
0
r"£".
0
r·
l> L~
I;
-'
\ LJ.Ir , 1
1'1
I'"
I ~~,.,A..l-.A
.. ,,"\---j;f-~. '. "
L..
--E--M_~ _ _R
01 II Im\L
"
....J
/
W
II..
D·' ~
::J ~
X
<:: ~
... 0.5.~ ~ :00 I
~-
/1/1 /V Y 1/
.002.00J .005
.01
.,
Figure 3-266
'.02
.OJ
.05.07
~
.1
.2
.J
LOAD DURA nON TO rUNDAMENTAL.PERIOD.
.5
,7
2
3.
Tit II
Maximum response of elasto-plastic~ one-degree-of-freedom system for bilinear-triangular pulse IC] = 0.010. C2 = 1000,)
5
7
10
,
,C
,.
"' .-
r'
,
r-r-
<
-,
·
..
,
,
' '
,
,
,
'.
~
0..
•
I.' . ,>~ .~~
;:. ..
I
:
~
Cl
c
,
'
.
,
... .. '
.
~ :: .....
"
"
,
0
,
,
"
pI'l5:>"• Of
•
"
'2"
.,"." ,
IF ;)
re
o<=
....
~
t-. ~ • ',I I.'
.e
"
..
.
•
,
,
I~•
,
;-
.
"
,
. ,
'
"
: ,
,
!
,
,
I
I
,
-;'1 I,
,
,
i
,
•
.
,
"
,
,I'
,
I
i !
I
, i ;;
,
3-325
,
.
.
c.
,~
<=
',~
~
'"
u
J, >
2
"
,
'
' •
. !.;
Q
'"o
~
·v
..
j .
",
"
,~,
.
.;
;;
.
..
~
r-,
•
,
...'
,U
.
:8Q
.,
•
',~
Cl
'.
0N
~,
N
It s! ,
j
,
i'
N
N
! - ~-
..
'-
. r-:~.,"\
,
,
U-
Cl
.
"
-
'.
•
~
r--
.
:-
'r- ~=
,
,
..
,
"
.
,,
.:
r-'
"
<
~
.s:::
i- .'"
, ,e
•"
c. '
CJ
'
.
?7I!J ..
.~
P~T
...(;....
-~
-e-e--t-
-c
""::r.....'
".''irltlt ~
Pressure'- Time
"0 L XE
Resistance
10
~HI., ~ I~ . .. '.
'-'~
'f~
r.
.' ~
"-
,
.,
20
~~~1io~
='= -~-
O''
!.. ~'*!r
I
I~~i+tl r:
,~~
ji
-Def,eetion~l~'g~~i~~ Xm
oj ~
-
I
'1'.' 1I".II/OAi.LLL
-I,Ii::!I'~1I
'II :
il.
l •
'fI11 ~BlIiIi=.=:-." .,
'ii-r'/if,tifl<:f7f:rrrtT:T""'T:TT1 tIItrTTTT '~I __ "!'!~~1111111; .1'·1·-:-:--1-11·' i
" IlolTTTT:fT1I;-inTTTTTTT z ..
...,,
..., '"
-Xni XE T'4
N
5
-,;r..;1=T:_....-
C1t=~"±:IOl:c
-I---:!.c.
:.+;..~"f""':'
v-
·o~(~
J "," 1'1....
---.....--t-+
rA
'0--,-
_• .::r '--"-,"- .__. ......
:-+
t-I$-~ ::j::;. ," - _..~.. '.
I:P=~-
......,.
.2' .'
~~;
I'
!
~.I . '
.....
;
,
...1-.......,
"""":'"
~
.. , tI;
•
iT
~_._-
!i+~~_
_-
.~~-:..
Ii:
~",,,--",,:S::
i~::"-=.,...........-:::"""'..,...I.-:
-..-f.-
'~'"
~-
-
. . -+
._:t=1.7..I....
jf-HWf:;it! I
I ;:
•
- ,,"
•. I
II
I..
I
'i
I.
"-'-11+ I
;
I
""
I~.
I ! _
~~;I1-,*j.;.;~·1..~
I' li'~' I;
, It,
I,
rii:
I- .
2
::
Ll.!.
5
,
.! illll :11 i
..
'~ i±llilli±tttH
II .:.
.'
y_
I I!·11.... • ~;· j
:++;
r rr t
;r;,
; ,.....:. , : I I ",I:
i·IIII;I·:ili.~q;~
0.2-
9:},,7.:........,.
:j::f.t::.I=4-_...
' i l ::.
.;., .":
~;'
..··_=c, _......=..:.. '~:':~x.:.t:=:_.
o' '" I '---:
E,
I' I;
'-;'+*
-t
II) "
,"
10
20
50
T/TN
Figure 3-268
Elastic rebound of simple spring-mass system
e
100
P
.,.
.1
<
B,LAST ,IMPULSE i b
,
A
/1
.
B
I
,
:
TIME
Figure 3-269
Pressure-time and resistance-time curves for elements which respond to impulse
TK 5-1300/NAVFAC P-397/AFR 88-22
Table 3-14
Details of computation by acceleration impulse extrapolation method .t.
n
0
0
Pn
I<"
Pn - I<"
Po
RO
Po- Ro
"nO
(P n
I<"J/m
~(4t)2
zx,.
",,-, ",,+1
'ao(·6t;)~.' .. 0':
'0
. xl .
0'
,
i
. At i
Z
2(4t
P" R, Pz ' RZ
1)
P,- R, P2- RZ
"
'I
81 (At})2
2X,
0
Xz
'2
aZ(4t 1)2
2X Z
X,
X3
"
, j-l
.... j '.
(j-l)4t
J4t 1.,...r
Pj
1 .,'
-,
R
j
_,
Pj _,-R j
_,
R, PL ~j .. , Pj - ;.~J. "-
j+l j(t.t 1 ) + 4t Z Pj + 1 ~j+l P j +1 - R j + l
.'
,
.aj_l
...
'J
.' -
6j+l .. _
6j-l (4~1')'
~j-:
X
_
a j (4t 1)2
,2Xj
X
_,
Z
aj+l(4tl~2
2X j + 1
j j
Z Xj
Xj - 1
,
Xj +1 ~j+2 _
, I,
. ;" ,
; :Yr"- " . f s :,..>-;.' ! . ' I
3-328'" i'
.'
i
,
.~
,
t'·
.. ;."
'"
j
~
TK 5-1300/NAVFAC P-397/AFR 88-22
Table 3-15
Figure numbers corresponding to various combinations of C1 and Cz
i
i
i
\
C2
i
i
i
i
1.00
1.70
3.00
5.50
10.0
3-64
3-64
3-SIt
3-64
3-64
C1 \ 1.000
I
i
i
i
i
30.0
100.
300
1000
3-64
3-64
3-64
3-64
3·1U 3-115 3-116
3-141
3-142
3-173 3-174
3-221
3-143
3-175
3-222
3-144
3-176
3-223
3-177 3-178
3-225
1
0.909 0.866 0.825
1
3-220
1
0.787
0.150
3-85
3-99
0.715
0.681
3-64
3-65
3-75
,:0 -,64S" 0.619
3-100
0.590 0.562
--
- -
3-117 3-118
3-145 3-146
3-119 3-120"
3-147
3-121
3-149
3-86
3-101
. O.Sl~·r: ·0:487·
0;422
.
3.,64
3-66
3-76
3-181
",3-228
3-122
I 3-182 3-151 3-183 .I· 3-184
3-229 3-230
3-123
3-152
3-102
0.383
·3-125
.' O~3~S flo 348.:' 0.316 0.281 0.2104 0.261 0-:231 0.215
.
. 3-61t
3-61
3-11
3-88
3-103
3-69
3-104
3-126
3-121 3-64
3-68
3-18
3-126
0~·'t98
0.118 ·0.162 0.141 0.133 0.121 0.110 0.100 0.091 0.083 0.015 0.066 0.056 0.046 0.042 0.032 0.026 0.022 0.018 0.015 0.013 0.010
.3-64
3-69
3-90
3-105
3-91
3-106
3-19
3-129 3-130 3-131
3-64
3-10
3-60
3-92
3-101
3-93
3-106
3-94
3-109
3-95
3-110
3-96
3-111
3-91
3-112
3-98
3-113
3-132
3-133 3-64
3-71
3-61
3-64
3-12
3-82
3-64
3-13
3-83
3-64
3-14
3-84
3-134 3-135 3-136 3-131 3-138
3-329
3-139 3-140
3-231
,3-185 3-232 I" F186 ';3-233 3-153 3-187 3-234 '3-154 1 3-188 3-235 3-189 3-236 3-155 I 3-190 . 3-237 3-191 3:-238 - -3-i56 3-239 3-193 3-240 3-151 3-194 3-2041 3-158 3-195· 3-242. 3:1'96 3-159 3-243 3,:160 3-191 3-244 ' 3-196 3-245 3~161 3-199 3-2463-200 3-247 3-162 3-201 3-248 3-202 3-249 3-163 3-203 3-250 3-251 3-204 3-164 3-205 3-252 3-253 3-206 3-201 3-254 3-165 3-206 3-255 3-256 3-209 3-166 3-251 3-210 3-211 3-258 3-161 3-259 3-212 3-166 3-260 3-213 3-169 3-214 3-261 3-262 3-215 3-110 3-216 3-263 3-264 3-217 3-111 3-218 3-265 3-112 3-266 3-219
..
3-124 3-87
3-224
3-226 3-179 3_148; 1 ·3"180 . : 3':227"
3-150
'0;536
0.464
-
I: ;-~9~
TK 5-1300jNAVFAC
P-3~7/AFR ~8-~2
, n
i ;
f.
-, .
, :,.
" I.
i ' , c "Response Chart Interpolation'
Table 3-16 Figure Number "
.C 1
C2
Desired Parameter
C2l
Cll 1
\'1
Cl
C2l
• •
"
. (;12
Cn
Cl
. C2~'
2
C· , ,13
3
,
- C2l ,"
Cl 4 '
C24
3-330.
Y2 Y
C23
"
·a
!
Cl:
"
.
y
'?~
,. ,.C23 ~
4
'.
'.
•"
'Ch
..
0/"
."
::J
Yb 'JI
u
APPENDIX 3A ILLUSTRATIVE EXAMPLES
TH 5-l300iN.\VFAC P-397/AFR'SS-22
.,
~.'. ~
-'. ~
" u
-,
(
,
3A-l(A) . Ultimate Unit Resistimce ~ ~
"
J
: .~
,,'
- . ~I .
..
'
" -' !:, 1
Determine the ulHmate··unit·'·resi~tanceof' a'twi>"way structural element ~sing (1) general solution and (2) charts.
Problem:
C)
Procedure:
N)"J~ '.1;',
Part (a) - General Solution
Step 1. Step 2. "
'i~-
:;.'
Establish design t.:, ', ..
~
,I..
,d- :-. .(1J:' , , " ' : dc . .
param~ters. ~~.:.,:
_
-~
fll.··
r
r
.~
r
Assume yield line locations in terms of x and/or y considering support conditions, presence of openings, etc. '-"Jl;
~ 4
I'
'
•
~
.
. ;:.'
~._
•
Step 3.
Deteroiine' negative 'and' p~sitive-moment capaciti'es' of sections crossed by assumed yield lines,
Step 4.
Establish distribution of moments across negative and assumed yield lines, cons~dering corner effects and those of openings.
Step 5.
Determine the ultimate unit,resistance for each sector in te~s of x and/or y considering free bi>dy diagram of the sectors (fig. 3-3). Summation of the moments about the axis of rotatio~ (support) of the sector yields equation 3-3,
.
Step 6.
;,..,~",
Equate the,:ultimate unit resistance of." the sectors and solve for 'the yi!!ld line location x and/or y., .:
Step
7.
~
With known· yield line location, solve for ultimate unit resi'stance'of the element, using equations obtained in Step 6.
,.--.~'"
'
\.<>;;"-""':','-""
.. ,", •• ;',
'
Note: For complex problems (three or more different sectors) the solution for the. ultimate unit resistance is most easily accomplished through a trial-and-erro; procedure by.determiningru for each sector for a given '(assumed) yield line location and adjust-. ing the yield lines until the several values of r u agree to within a few percent. ,
Procedure:
--'
'...
Part (b) - Chart Solution
Step 1.
.'
11<11 ..
4-
,>,. "
Same as in step 1 of , part a.
.
,
..r~
"
Step 2.
' Same as 'in step 2 o"f part a.
Step 3.
Determine' the ':ri~gatIve and positive 'ultimate moment capacities in vertical and horizontal directions . • : 's> ..., .
-..
l~ i
J'
..
: .r '
".
! . .'.' .
3A-l
,.r: .
- ):
TK
5-l300/NAVFACP-397/AFR88~22
Step 4,
For' given support conditions (and value of x2/xl in the case of an element with three. edges supported and fourth free), use the appropriate'~hart (figs, 3-4 through 3-20) to. obtain yield'line location'ratios x/L or y/H for value of quantity obtained in step 4,. Then calculate x or y', ~ .'
Step 5.
Using the appropriate 'equation from table 3-2, determine the ultimate unit 'resistance of the ele~ent,
.
";-
~
',;.'
:J.: .:
, , ,'
Example 3A-l
Uitimate Resistance ,
,
Ultimate unit resistance of two-way structural'steel element shown below;~sing (1) general sol~tion and (2).char~s.
Required:
-'./
'.
;
,
"
.
,
Figure 3A~l -:
(Solution: Part (a) - General Solution) Step 1. (a) (b) ~.
Given:
H' - 168 'in L - 240 in free' at the fourth Fixed on three sides and > •
,
,
Step 2.
Assume y,ield line. lo~.ati~n.(fig. 3A - 2)
Step 3.
The negative and positive moment capacities in both the horizontal and vertical. directions are determined from the prop~rties of the material. For this example; it will be assumed that the moment: capacities are equal to M - 20,000 inlbs/in,
3A-2 '
TlI 5-1300/NAVFAC P-397/,AFR 88-22
240
120
-
~
I
~ ~
II:
... /
,"
,,,
.,
eo
,
,
"
"
""!N
..
.,
CD
"
""!N "
eo
120
"
.
III
III
~!i
~!i
ti: >
ti: >
> III
I CD
, ,)'" ", ...' ... , ,,' I
CD
CD
,"'
= -
"'M~
> III
119III
f9III
jMHN"
jMHP M
"
ri III
1111
AT NEGATIVE ,YIELD LINE
\,2 I..,..,I""M'" AT POSITIVE YIELD LINE
al ASSUMED YIELD LINES AND DISTRIBUTION OF MOMENTS
...~
,(Q
120
.-R.:. ).~
CI
.W
RII
liMVN
120
'bl FREE-BODY DIAGRAMS FOR INDIVIDUAL SECTORS
Figure 3A-2
3A-3
....
.;
88~22
TM 5-1300/NAVFAC P-397/AFR
'MHN - MHP - MVN - MVp - 20,OOO,in 1bs/in. Step 4.
For distribution of moments across negative and assumed positive lines, see f~gure 3A -2 (a). ,
Step 5.
The ultimate unit resistance of each sector 'is obtained by taking the summation of the moments about its axis of rotation (supports) so that I:MN +I:Mp - Rc - ruAc
a. " Sector I
(Ug. 5A - 2)
D{VN; + I:MVP - 120(20,000) + 2(2/3)(20,000)(60) -I' , :' 120(20~000) + 2(2/3)(20,000)(60). 8.0x10 in-1bs. 240(y) 2
therefore, b.
Sector II
I:MHN + I:MHP -
(fig. 3A - 2)
(168-y/2)(20,OOO) + 2/3(20,OOO)(y/2) + (20,OOO)(168-y/2) '+ '2/3(20,OOO)(y/2)
- 336(20,000) - y/3(20,OOO) 120[168+2(168-y)
120(168+168-y)
[ -----'---] / 3
2
(168+~68-y)
therefore, .336(20,000) - y/3(20,OOO) :-
:.
r u-
Step 6.
4800 (252-y) Equate the ultimate unit resistance of the sectors.
'.
r "
:;
'.
"
10(20,000) , 2
336(20,000) - y/3(20,OOO)
Y
4800(252-y)
Simplifying: y3 _ 1008y2 _ 144000y + 36288000 - 0 and the desired root is:
y - 137.6 ins. 3A-4
1M 5-l300/NAVFAC'P C397iAFR 88-22
Step 7.
The ultimate unit resistance is obtained by substituting the value of y into either equation obtained in step 5, both of which yield: r u - (20,000)/(137.6)2 - 10.6 psi
Solution:
Part (b) - Chart Solution
Note: Element conforms' to the requirements of section 3-,8 since it. is fixed on three sides and free on the remaining side and has uniform thickness in the horizontal and vertical directions .. Step 1. Step 2.
,Same as step 1 in part a. For illustrative purposes, a different 3A-3) will be assumed.
'
~ield
,. !
..
\,
! .
240·
p-
X1
\
I
CD/ I
pattern (fig.
~ ~
\
I
®
I
\
~
\®~ ~ \ ~
\
I ,
,
io
... CD
~ \~
Figure 3A-3 Step 3. Step
4.
For ultimate moment capacities, see step
~
of part a
.... ,
,For three sides fixed and the fourth' f r ee; calculate cheparameter.
From figure 3-11, (X2/Xl
=
1.0) calculate the parameters:
L/H[MVp/(MHNl + MHP)] - 240/168[20,OOO/(2)(20,OOO)J l/2 - 1.01 ., ,; 3A-5
TK 5-l300/NAVFAC 'P-397/AFR 88-22 and 20,000
HVp
1.0 20,000
Read yield.line location
.
I,
-,
. Xl/L exceeas the maximum possible value of 0.5 therefore, assumed yield line pattern is wrong. Assume alternate yield' line, pattern ~s shown in . figure 3A-2 .. From figure 3-16 calculate the following parameters: -,
L· H ,,'
240
(20,000+20,000) 1/2
] - 0.71
168
, [ (20,000+20,000)1/2 + (20,000+20,000)1/2
and,
[40,000/40,OOO]1/~,/
1 +[40,000/40,000]1/2 - 1/2
from figure 3-16 read of yield line location: I
y/H - 0.82;
Y -,0.8(1~8) - 137.6 in
X/L - 0.50;
X - 0.5(240) - 120.0 in
From table '3-2
Step 5.
NOTE: Both equations given in the table for each edge condition and yield line location, will provide identical values of r u:
vp)
5.,(Mw" H
"yZ' . .. \
.',
\_
-. '. \
5(20,000+20,000) 137.6 2
Required:
,.
"1
Example 3A-l(B)
(0.000?5)20,OOO - 10.5 psi
Ultimate Unit Resistance
Ultimate unit resistance of the element considered' in example 3Al(A) except there is an opening as shown in,figure 3A-4. 3A-6
.'
TIl: 5-l300/NAVFAC P'-39i/AFR,'S-S-22
".J .1~'
.~
:./ ";
.. .: r
240'
,
;t;
'.
!
.'.
..
~
.
l~
/-
I"
,...
.... , . i
f
1-
-:..,
'':.:.
,
~- ~.~.
'. t":~'
",f'-
J
,
';\
~ Ij
, "
i
"
figure 3A-4
."
.
'
"
'.
j"
Solution: Step 1.
Given:
Two additio'nal free edges are formed 240" H - 168" to due to the presence of the opening, (a) L -
j
.',
Step 2.
Assumed ii~ld line location is shown in figrue 3A-5 (three different sectors ar~ formed),
Step 3.
S&ue as step 3 of Example
1A- l ( A) ,
p~rt
a, "
• .e ,
. ,'~ ~'0 ". ~ . (.) " i0' •
,
/
,
,,/
I
.'
t-, ' ,,' ,"-
"
-
xJ2
.r . -
...
\
I
'
,
"
0'
,
.
,
t
•
....
,
180-xJ2
',..
~
I
;
I I111I11 I
, .
-,
"
!@ITITI[[] II I I I il(f . I
a" -
o~ .~
-
m '0 "" '. c'] . ·
;.~ .:~ tr I \
,I
I .
.
..f
;
00 24
Figure 3k-5
,3A-7 '.
I
,
TH 5-1300/NA.VFAC P~ 397/AJ1R88 ~.22 Step '4'.
For distribution positively yield located at lower capacity in this
Step 5.
The ultimate unit re~iitance is obtained from: EMN +
a.
~
of moments across negative and assumed lines ,see' figure 3A,5. (Since opening is right corner, there is no reduced moment area.)
- Rc'- ruAc
Sector 1 (fig. 3A-5), - (20,OOO)(168-y/2),+ 2/3(20,000) (y/2) + (20,000) (168-y/2)' +. 2/3 (20, 000)(y/2) - 336(20,000) ~ y/3(20,000) -,(336-y/3)(20,OOO)
ruAc - r u[x(168+l68-y)/2j [x(168+2(168-y»/3(168+l68-y)] r ux 2(252-y)/3 therefore,
.
,
'(1008-y)(20,OOO) , ru
-
b.
Sector II
x 2(252-y), (fig. 5A-5)
,
EMHN + EMHP - (20,000)(80) + (20,000)(80) - l60(2q,<:>,OO) Note: I
The sector is divided into 'parts a, b, and c so that the centriod may be obtained (see table 'below). Portion of
Sector ,
.
Area (A' )
Distance from Centroid to axis of rotation' (c')
'3
2 (y-88)(60)
,
6q '
2
(168-y)(240.x)
"
(y-88)(180-x)(360-x)
(y-88)(60)2
... .
.' -:
A'e'
-
. c
..
6
3
, b
- 360-x,
(180-x) , +,60
(y-88)(180-x) a
,
2 ,
-
(240-x)
,
2 3A-8
., "
,
.
'
,,'
(168-y) (240-x)2 2
TK 5 e1300/NAVFAC P-397/AFR 88-22 (ye88) (l80-x) (360-~)
+_(y-88)(60)?+ (168-y) (240-x)2
Ac - 1:A'c' .6
- 1/2
2
-
(y-SS)c\ I [(lSO-x)(360:x) +'lOSOO]'+ (16S-y)(240-x)2 , I 3 '... :-
..
.
(
ru -
2
'
Ac 6,400,000. (y-SS) ---- [(lSO-x)(360-x) + 10,800] + (168-y)(240-x)f
.
3
c.
." .' "
Sector III ~
.+
(fig. 3A-5)
DI"p.- (20,,000) (lSO-x/2) +,2/3(2.0,000)(x/2) + (20,000)(lSO-x/2) +.2/3(20,000)(x/2) - 360(20,000) .~ . (20 , 000),x/3 , , , ' .-'
Area (A' )
Portion of Sector
. '.
Distance from Centroid to axis of rotation (c' )
A'c' ,
(y-SS) + SS - Y + 176
riM.-x) (y-SS)
--
a 3
2 .
(lSO-x)(SS)
b
c . ~I
: (lSO,x) (y-SS) (y+176)
3
v r-
6
:
l'
.yr
xy
2., . , y
2
3
----
.
(lSO-x)(SS)2
SS
.
"
, '
.'
.
'
.
'
..
2 xy2 6
'.
Ac - 1:A'c' - (y-SS)(lSO-x)(y+176)/6 + (lSO:~)(SS)2/2 + xy2/6 , - 1/6 (lSO-x) [(y-SS)(y.+176)+23.,232] + xy2 c;
."1
ru -
Ac
......
••.
, t;t
(2l60-2x)(20,000)
Step 6.
(lS'O-x) [ (y-S8 ).(n176 )+23; 232] +xy2,' ., Due to the ~omplexity of obtaining a direct solution,for ultimate unit resistance, a trial-and-error solution will be used,( see table below): 3A-9 - ,
TH 5-1300/NAVFAC P-397/AFR 88-22 ' ,
x 125 125 125 125 125, 130 130 131
y
,
,
130 135 140 145 .150 130 135 135
,
-
rI 9.21 9.55 9.92: . 10.32 10.77 8.52 8.83 8.70
rIl
rIll
7.67 7.92 8.19 8.4,8 8.79 8.29 8.55 8.68
9.33 8.77 8.25 7-.78 7.35 9.50
..
"
~.92
..
.
8.85
Therefore: "
x - 131. iris y - 135 ins r u - 8.68 psi
Problem 3A-2 Problem:
,
..
Resistance - Deflection Function
Determine the actual 'and equivalent r e's Ls t ance deflection function in the elasto-plastic region for a two-way structural element.
Procedure:
Step 1. Establish design parameters a. Geometry of element. b. Support conditions Step 2.
Determine ultimate positiye and negative moment capacities.
Step 3. Determine static· properties: a. Modules' of elasticity for the element. b. Moment of inertia of the element. Step 4.
Establish points of interest and their ultimate moment· capacities (fig. 3-23)
Step 5.
Compute properties at first yield. Location of first yield ' . b. Resistance at first-yield r e c. Moments at remaining points consistent with r e d. Maximum deflection at first yield. a.
Step 6 .
Compute properties at second Remaining moment capacity Location of second yield. Change in unit resistance d. Unit resistance at second e. Moment at remaining point
a. b. c.
.r
3A-io
yield at other points Ar between first and second yield. yield rep" consistent with
TK S-l300/NAVFAC P-397/AFR 88-22
f. g.
Change in maximum deflection. Total maximum deflection.
Note: An element with unsymmetrical support coriditions may exhibit three or four support yields; Therefore, repeat Step 6 as many times as necessary to obtain 'properties at the various yield points. Step 7. a. b. c. d.
Compute properties at final yield (ultimate unit resistance) Ultimate unit'resistance. Change in resistance between ultimate unit resistance and resistance at prior yield. Change in maximum deflection (for elements supported on two, three, or four sides, use stiffness obtained from figure 3-26, 3-30 and 3-36, respectively). Total maximum deflection.
Step 8.
Draw the actual resistance-deflection curve (fig. 3-39).
Step 9.
Calculate equivalent maximum elastic deflection of the element.
.
" Example: Required:
3A-2
Resistance-Deflection Function
The actual and equivalent resistance-deflection function (curve) in the elasto-plastic region for the two-way structural steel element.
Solution: Step l. a. b. Step 2. Step 3.
Given: L - 240 in; 'H - 168.in. Fixed on three sides and free at 'the fourth. Same as step 3 of example 3A-l(A), part a. Static properties.
a.
Modulus of elasticity, Es for steel Es - 29 x 10 6 psi
b .'
Considering a I-inch strip' (b - 1 inch) Assume I - 144 in4
Step 4.
For points of interest, see figure 3A-6.,
Step 5.
Properties at first yield. 3A-ll
TH 5-1300/NAVFAC'P-397/AFR 88-22
'
. , "
POINT 2
MHN ~n' '". ,
:: 1., ,,'1
.
.'I' ..'.
..-
,
....,
-j
--!
,
"
~
-~
.....
I'(POINT 3
IJMVN
,.
,,
.
I
"
'.
~"
.,
.t.
Figure 3A-6
c· q'
"
From figure 3-Z7 for H/L - 0.7 \ ~.~! • -
a.
. ' t·
B1 - 0.077
BZ - 0.160
Y1 - 0.012
v - 0.3
,,
f
p'
,
MHP - MHN - MVp - MVN - ZO,OOO in-1bs/in Mp
-_BrH~.
r '"1 ' ..
_.., ~_
,.' -,
....,
1::_
~.
M/BH Z ..J
••
.. ':.
l
.-.....
...
".
r1 ... 20,000/[(0.077)(168)2]."'·,9:Z0 psi
..
rZ - 20,000/[(0.160)(168)2] - 4.43 psi
,.
r3 - 20,000/[(0.115)(168)2] - 6.16 psi ! -,
First yield at point 2 ,(smallest. r) .'-> .,
+
...t
b. 'r e - 4.43 psi c .. Mp (Point
1) -
,
,.
"
"
(0.077)(4.43)(168)2
-
9,627 in'lbs/in
MN (Point 3).-'(0.115)(4.43)(16,8)2,,.. ~14.379. in-1bs/in d.
D - EI/b(1-v 2)
-,
- 29 x 10 6 . x 144/1[1::.(0.3)2]-
45.9 x 10 8 in-1bs
Xe - Y1reH4/D - (0,'0120)(4:43)(168)4'/43(10 8)-.:0.0092 in
Step 6.
Properties at second yield.
3A-12· ,
",I
TK 5-1300fNAVFAC
P-397/AF~
88-22
After first yield e~ement assumes a simp1e-simp1e-fixed-free stiffness, therefore from figure 3-29 for H/L - 0.7.
B1 - 0.120
B3 - 0.220 v -
Y1 - 0.045 a.
0.3
(Point 1) - MHP - Mp (at r e)
Mp
- 20,000 -, 9,627 MN
10373 in-1bs/in,
(Point 3) - MVN -' ,Mp (at r e) -.20,000 - 14,379
b.
"
5621 in-1bs/in
Mp (Point 1) - 10373 in-lbs/in - B1 Ar H2 Ar -,103~3/(0.120)(168)2 - 3.06 psi
~ (Point 3) - 5,621 in-lbs/in - B3Ar H2 Ar - 5,621/[(0.220)(168)2] - 0.90 psi Second yield at Point 3 (smaller Ar) c. Ar - 0.90 ,psi d. rep - r e + Ar
-
4.43 + 0.90
-
5.33 psi
e. Mp (Point 1) - 0.120(0.90)(168)2 1
- 3.048 in-1bs/in f. D - EI/b(1-v 2) -
(29)(10 6)(144)/1[1-(0.3)2]
45.9 x 10 8 in-1bs/in Ax - Y1ArH4/D - 0.030(0.90)(168)4/45.9(10)8 - 0.0047 in g. Xep Step 7.
Xe + AX
-
0.0092 + 0.0047
-
0.014 in
Properties at final yield (ultimate unit resistance). After second yield element assumes a simp1e-simp1e-simp1e-free stiffness, therefore from figure 3·30 for H/L'- 0.7. Y1 - 0.045
v - 0.3
a. r u - 10.6 psi (part a, example 3A-1(A» b. Ar - r u - rep - 10.6 - 5.33' - 5.27 psi
3A-13
..
c. D
,
"
- 45.9 x 10 8 in-1bs
Ax -
Y1rH4 /D
(0.045)(5.27)(168)4/45.9 x 10 8;
.'
.',
0.041 in d. ~ - X ep + AX'! _. '0.014 + '0.041 - 0.055 in Step 8.
For actual resistance-deflection curve .:see .f Lgure 3A-7.
Step 9. 3-35 XE - 0.0092(5.33/10.6) + 0.014 [1-(4:43/10.6») + 0.055[1-(5.33/10.6)J . 0.00463'+ 0.0081 + 0:0273 0.040 in
'.
1~ • • '
The equivalent resistance-deflection curve·is 'shown in figure 3A7.
. Problem 3A-3
'.'
'
-
-r,
Dynamic Design Factors For A One, Way Element i
".' I
Problem:
Determine the plastic load. mass and load-mass factors for a oneway ,element.
Procedure:
. . ,.
Step 1. '
:
'
...
J
'1
Step 2.
•
Establish design parameters'. t' ,
Determine deflected shape."
a. geometry of element b. '" support' condf t Lons c.' type of load'a~d mass
' • . l.
.'
Step 3.
Determine maximum deflection \
Step 4. a.
Determlne deflection function For distributed ~oad and/or continuous mass determine the deflection at any po Lrrt ."
3A-14 :'
TH
5-1300/~AVFAC
P-39?/AFR
88-~2
12
ruJ-----
10
/
XE
Xp
/ / H
Ul
8
·EQUIVALENT~
!
~
/ r--ACTUAL
I
L
W U
Z
"">--
6
Ul H
Ul W
a: >--
H
Z
4
:oJ
2
O-f----+---+----+---t---+----j----t----t0.0
1.0E-2
2.0E-2
3.0E-2
4.0E-2
5.0E-2
DEFLECTION X (INCHES)
c.
F.igure. 3A- 7
3A-15
6.0E-2
7.0E-2
8.0E-2
b. Step 5.
·For concentrated loads and concentrated mass determine the deflection at the load. Calculate the shape function For distributed equation 3-43.
b.
For concentrated load and concentrated mass calculate ¢r' equation 3-46.
Step 6.
arid/or continuous mass calculate ¢(x),
Calculate the load factor, KL.
.a.
Use equations 3-1>1 and 3-42 for a distributed load .
b.
Use equations 3-41 and 3-45 for a concentrated load
Step 7.
Calculate the mass factor,
~.
a.
Use equations 3-47 and 3-48 for a continuous mass
b.
Use equations 3-44 and 3-49 for concentrated mass
Step 8.
Calculate the load-mass factor KLM, from equation 3-53.
Example 3A-3(A) Required:
lo~d
a.
Dynamic Design Factors For A One-Way Element
The load, mass and load-mass factors for a structural steel beam in the elastic range, with a distributed load.
Solution: Step 1:
Given structural steel beam shown in figure 3A-8
L
x
Figure 3A-8 a. b. c.
L - 120 in. Simply supported on both edges p(x) 2,000 Ib/in m(x) - 0.0055(lb-s 2/in4)/in
3A-16
TN 5-1300/NAVFAC P-397/AFR 88 022
,.
Figure 3A-9 Step 2:
Assumed deflected shape for elastic range is shown in figure 3A-9
Step 3:: .The maximum deflection at the center is 5p(x)L4 384 EI Step 4:
Determine deflection function p(x) o(x) _ ---- (L 3 _ 2Lx2 + ~3) , . 24EI
Step 5:
Calculate the'shape function using equation 3-43 ......o(x) p(x)x 384EI >. __' (L3 _ 2Lx2 + x 3 ) 0max
24EI
5p(x)L4
.,
,..--Step 6: L
FE -
f
o
Using equation 3-42, determine equivalent force 120 ". 16 (L3X •. - 2Lx3 + x 4)dx p(x)>(x)dx (2,000 lb/in) 4 o 5L a.
-f
2Lx4
.+
<
4
x5 5
I"
- 1, 280L
- 153,600 lb. b.
From equation 3-41, find the load factor FE
KL - -F
KL
-
153,600
lb.
(2,000 lb/in x 120 in. )
0.64 in th!, eLas t Lc range
3A-17
.'
.
TK 5-l300/NAVFAC P-397/AFR 8S i 22
ME
Find the equivalent mass from equation 3-48
a.
Step 7:
-,J
256
L 2 m(x)~(x)dx
.:0055
120 (L 3x
2Lx3 + x 4)2
J
25L8
dx
0
0 1.408 25L8
., 120
+ x9
1.408 25L8
5
6
7
8
9
]
o
.00277L .:
.
0.3325 lb 2_ s3/ i n b.
From equation 3-47, calculate the mass factor
-
. ·0. 3325.,lb - s2 '/in 3
',; (0.0055 lb - s2 /in4' x 120 in) KM -
Step 8:
0.50 in the elastic range
Calculate the load-mass factor as defined by equation 3-51 KLM - KM/KL
- 0.50/0.64 KLM -
Example 3A-3(B) Required:
0.78 in' the elas tic range Dynamic Design Factors For A One-Way,Element
The load, mass and load-mass factors for a structural steel beam in the .plastic range with a 'distributed load . .,
Solution: Step 1. a. b. c. d. Step 2.
Given the structural steel beam shown in figure 3A-8 L - 120 in. Simply-supported on both ends p(x) - 2,000 lb/in m(x) - 0.0055 (lb - s2/ i n4)/in Assume deflected shape for the plastic range is shown in figure 3A-lO 3A-18
LIZ
LIZ
x
Figure 3A-10
Step 3.
Determine maximum deflection 0max - (L/2)tane
Step 4.
Determine the deflection at any point. o(x) - xtane. x < L/2
Step 5.
Calculate the shape function,; equation 3-43 o(x) xtane
x < L/2
Step 6: a.
Find FE using equation 3-42. FE
J
60
L
-J p(x)
2
o
2,000 lb/in) (2x/L)dx
o
_ 4,000 lb/in [ x2 /L ]:0 - 120,000 lb b.
From equation 3-41 FE 120,000 lb. KL - -
F
(2,000 lb/in) 120 in
KL - 0.5 in the plastic range Step 7: a.
Use equation 3-48 to find the equivalent mass
J
L
ME -
m(x)
o - 00'4 [
J
60
::,
r
0
(0.0055)(4 x 2/L2)dx
TM 5 c1300/NAVFAC P~397/AFR.'88~22 " , '
b.
As defined by equation 3'47 ME KM - - M :., , KM
Step 8.
..
- s2
0.22 1b
/in 3:
s2 /in4)120 in
(0.0055 1b:
0.33 in the plastic range
Calculate KLM using equation 3-53 KLM - ~/KL
0.33/0.5 KLM - 0.66 in the plastic range Example 3A-3(C) Required:
Dynamic Design Factors For A One-Way Element
.
,
The load, mass and load-mass factors for a structural steel beam in the elastic range, with a·concentrated load.
Solution: Step 1:
'.'
I
'
Given structural' steel beam shown in figure 3A-l1
r L/2 {
r. ,'.
.•
',J:.
Figure·3A-ll \.
a. b. c.
;,
L - 120 in. 'j Simply supported on both sides F - 240 kips
m(x) - 0.0055(lb - s2/i n4)/in. Step 2:
Assume deflected shape for elastic range is shown in figure 3A-12
TK 5-1300/NAVFAC P-397/AFR 88"22
Figure 3A-12 Step 3:
Determine maximum deflection
°max - - - 48EI Step 4: 8.
Determine deflection functions for continuous mass,
O(x) 48EI b.
for concentrated load
48EI Step 5: a.
b.
Calculate shape functions for continuous mass use equati~n 3-43 o(x) Px(3L - 4x 2) t/J(x) 48EI °max _ (3L 2x -.' 4x 3)/L3 for concentrated load, use equation 3-46 PL3 48EI t/J r 48EI PL3 - 1.0
Step 6: a.
Find equivalent force from equation 3-45 i FE -
L
Frt/Jr - Pxl - 240 kips
3A-21
48EI PL3
..
TM 5"1300/NAVFAC·P-397/AFR ., _.,. ," ".. ...'88-22 '.~
b.
Using equation 3-41, calculate the load factor 240 kips 240 kips KL - 1.0 for the' elastic"gange
Step 7:
Equation 3-48 gives the equivalent mass.
a.
<
120 (0.0055)
L
ME -
Jm(x)¢2(x)
J
dx -
o
o 0.0055 L6
!'
b.
From equation 3-47, cel cu'l at.e rche mass .f'ac tio'r 0.321b - s2/ i n3 (0.0055lb - s2/ i n4 x 120in) ";
.I',
KM - 0.49, in,.th.e elastic range "
Step 8:
.'
Calculate the load,mass factor, frof,equation 3-53
KLM - KM/KL ':- 0.49/1.0 KLM - 0.49 for the elastic range F
Example 3A-3(D) Requ!.red:
Dynamic Design Factors For A One-Way Element
Determine the load, mass and the load-mass factors for a structural steel beam, in the plastic range, with a concentrated load.
Solution: Step 1. a.
b.
Given structural steel beams shown in figure 3A-ll. L - 120 in Simply-supported at both edges 3A-22'
r-,
TK 5-1300/NAVFAC
c.
P~397/AFR
88-22
F - 240 kips· m(x) - 0.0055 (lb - s2/ i n4)/in
Step 2.
Assumed deflected shape for the plastic range is shown in figure 3A-13 .1 .<
L/2
L/2
.x
Figure 3A-13 Step 3:
Determine maximum deflection 0max - (L/2)tan9
Step 4: a.
Determine deflection function .. for continuous mass
x< L/2
O(x) - xtan9
b.
for a concentrated load Or - (L/2)tan9
Step 5: a.
Calculate shape factors using equation 3-43 for continuous mass o(x) ¢(x) -
-°max
-
xtan9 (L/2)tant'
2x/L b.
x < L/2
equation 3-46 for concentrated load (L/2)tan9 (L/2)tan9 1.0
3A-23
TM 5-1300/NAVFAC P-397/AFR 88-22
Step 6:
a.
The equivalent force is found using equation 3-45 i ~
FE -
Fr'I'r - PXl -
240 kips ..
r-l -,
b.
Equation 3-41 gives. the load factor
KL -
a.
240 kips
F
240 kips
1.0 for plastic range
KL -
Step 7:
FE
The equivalent mass is found using equation 3-48
J L
ME ;.
m(x)¢2(x)dx - 2
J
o
o
b.
60 (0.0055)(4 x 2/ L2)dx
Solve for KM using equation 3-47 .~-
(0.0055 lb
KM - 0.33 in the plastic range Step 8:
From equation 3-53, calculate KLM
- 0.33 in the plastic range Problem 3A-4 Problem:
Plastic Load-Mass Factor
Determine the plastic load~mass factor KLM for a two-way element using (1) general solution and (2) chart solution. Note: The determination of the plastic load-mass factor follows the calculations for the ultimate resistance, hence the structural configuration and the location of the plastic yield lines will be known.
Procedure:
Part (a) - General Solution
Step 1.
See part a, problem 3A-l for the 'structural configuration and location of plastic yieJ.d lines. Denote sectors formed by yield lines. 3A-24 .
TM 5-l300/NAVFAC·P-397/AFR 88-22
Step 2.
Determine the load-mass factors properties I, c, and L' for
Step 3. Step 4. Step 5.
Determine the factor I/cL' for all sectors ... Calculate the total area of the element. With values obtai~ed above, calculate the plastic load-mass factor for the element using equation 3-57·.
~
all sectors.
;..
I
Note: In the above problem, an element of uniform thickness was considered. For non-uniform elements, the load-mass factor is c a l.cul at edlus tng equation 3-53 where the ma·58 of the individual sectors must be considered.
Procedure:
Part (b) - qhart Solution ,;
Step 1.
See part·b, problem 3A-l for structural configuration and location of plastic yield lines in terms ·of x/L or y/H.
Step 2.
For known value of X/L or y/H and support condition, determine the load-mass factor for the element from figure 3-44.
Note: Chart solution may be used only:if.the element conforms to the requirements listed in section 3,17.3 Example 3A-4 Required: Solution:
Plastic load-mass factor for the eleme~t considered in example 3Al(A) using (1) general solution and (2) chart solution. Part (a) - General Solution
Step 1. L H
X
Plastic Load-Hass Factor
\
.
Given structural configuration and location of yield lines shown below (see part a, exa~ple 3A-l(A)) in figure 3A-14.
240 in 168 in 120 in
Y - 137.6 in constant
Tc
L
x
I I
I I
n
Fixed .........
/
/ /
/
/""
"" .
" m
I
/
"
.,
Figure 3A-l:4
3A-25 .
TM S-1300/NAVFAC P:397/AFR 88-22
Step 2. Load-mass factor properties. a. Sector 1.. L' - Y - 137.6 in c - y/3 -137;6/3 I - L(L,3) /12 - 240(137.6)3/12. '.
~L
Sector II.
b.
e.g.
AXIS OF
+
ROT~mN7
.
:z:
rZ??,,~ti ILl .
I-AXIS OF
I
Figure 3A-15
ROTA11ON
Figure 3A-16
L' - x - 120 in H - Y - 168 - 137.6 - 30.4 in L' [H + 2 (H-y)] c -
120 [168 + 2 (30.4) J
3 [H + (H-y)]
3 (168 + 30.4)
c - 120 (0.384) (H-y)(L' ) 3 +
y(L'P
I 3
12
30.4 (120)3 +
137.6(120)3
3
12
-
21. 60(120) 3
Step 3. Calculate factor I/cL' for each sector: I
240(137.6)3/ 12
Sector I. cL' I
(133.4/3)(133.4) 21.60(120)3 6,646 in2
Sector II. cL' I Sector III.
8,256 in 2
(0.390 x.120)(120) 6,646 in 2
cL' Step 4. Area of panel A - LH - 240 (168) - 40,320 in2
3A-26
TIl 5-1300/NAVFAC P-397/AFR 88-22
Step 5.
Load-mass factor l/cL' (eq. 6-14) A
'8,256 K
2(6,646) 0.534
lJol
Solution:
+
"
40,320
Part (b) ,,- Chart Solution
Step 1.
Given: Panel fixed on 3 edges, 1 free and y/H - 0.803 part, b, example 3A-l(A».
Step 2.
From'figure 3-44, read load-mass factor
._'
I'
(see
.
KlJol -0.,543
Problem 3A-5 Problem:
Response of a Single-Degree-of Freedom System subject to Dynamic Load Determine the maximum response and the corresponding time it
occurs of a single-degree-of-freedom system subjected to dynamic load using (a) numerical methods and (b) design charts. ,
Procedure:
Part (a) - Numerical Methods
Step 1.
Establish dimensional parameters of the system.
Step 2.
Determine the natural period of vibration and integration time interval.
Step 3.
Construct a table similar to table 3-14 of section 3-19.2. Note': For the first interval n-l, Equation 3- 59 is used and subsequent intervals, the recurrence formula (eqn. 3-56) is used.
Procedure:
Part (b) - Chart solution
Step 1.
Same as step 1 of example 3A-5, part a.
Step 2.
Determine the non-dimensional parameters.
Step' 3.
Example 3A-5 Required:
,Determine. the ratio of the maximum displacement to the elastic displacement ~/XE and the ratio of the time at which this maximum displacement occur's to the duration of the blast load. Maximum Response of Single-Degree-of-Freedom System Subjected to a Triangular Load.
The maximum response and the time it occurs,' of a single-degreeof-freedom system subjected to blast loads, using (a) numerical methods and (b) design charts. 3A-27
TH 5-1JOOVNAVFAC P-397/AFR 88-22
Solution: Part (a) - Numer1cal Methods Step 1. ::
Given: p
.' ,'u
K
XE ,1"'
TIME
(a) Single-Degree-ofFreedom System
'.
DISPLACEMENT
(b) External Load ,
"
Figure 3A-17 '
"
r-
(c) Resistance Function
I., -' ~"
2.5 Kips-sec 2/ft - 9,860 - 750 KipsKips/ft XE - 0.076 ft T - 0.10 sec 1000 Kips P
m K ru J ..
"'·1 ...
Step',2 '., Natural period .of vibration and integration' time interval . • I
2" [m/ Kjl/2''':'
'~',
2" [2.5/9,860]1/2
t - TN/10 - 0.01 sec Step 3,
Construct table as shown below,
",r,. -,'to
3A-28 '
- 0.10 sec
TM 5-1300/NAVFAC P-397/AFR 88-22
1
2
3
n
t
Pn
(sec)
Kip.
0 1 2 3 4 5
0 0.01 0.02 0.03 0.04 0.05
6 0.06 7 0.01 8 0.08
4
(Xips) (Kips)
0
1000
900 800 100 600 500 40 300 200
Pn-R"
R"
"n-(Pn -R,,/m (it/ne Z) 400
1000
197.200 702.800 671.68. 128.426 -50.0 150
150 750 750 750 750
1
6
5
51.366 -20.0
-150.0
-60.0 -100.0
zx"
"..-1
"..+1
(ttl
(ttl
(ttl
(ttl
0.040
0.0
0.0
-0.00600 0.34522 -0.0100 0.43673
-140.0 -180.0 -220.0
10
"n(4< l 2
0.028112 0.0400 0.0 0.05131 0.13622 0.020 -0,00200 0.242114 0.06811
281.120
-250.0 -350.0 -450.0 -550.0
9
8
-0.0140 -0.0180 -0.0220
0.020 0.068112
0.121351 0.17261
0.121351 0.21186 0.17261
0.25312
0.050623 0.21786 O.S48H 0.25312 0.55525 0.27431
0.271t37 0.27762 0.25880
9 0.09
Note:
Pn - f(t n)
- 1000 [l-n(At/T)] ~-
Note: For n-O, ~+l (Column 10)
- XO+l - Xl - (1/2)aO(At)2
- (1/2)(0.040) - 0.02 ft For n-l,
2~ (Column 8)- 2X l - 2[(1/2)aO(At)2 j 2(0.02) - 0.04 ft
•
~-l
(Column 9) - Xo
~+l
(Column 10)
-
0.0
X2 - 2X l-XO + al(At)2 - 2(0.02) - 0 + 0.02811 - 0.06811 ft
For n-2,
2~
(Column 8)- 2X 2 - 2(0.06811) - 0.13622 ft.
3A-29
TH 5·l300/NAVFAC P-397/AFR 88-22
~-l
(Column 9)'
- Xl
- 0.02 ft.
~+l (Column 10)
- X3
- 2X 2 - 'Xl + a2 (At)2
- (2)(0,06811) - 0.121357 ft.
"
:. "
For n-3. 4 •...• repeat the above procedure. ,
" "
:'
0.02 +'0.005137
Solution:
,
'
Part (b) - Design Charts
Step 1.
Same as step 1 of example 3A-5, part a
Step 2.
Non-dimensional parameters
a.
Natural period of vibration, Tn Tn - 2rr[m/Kjl/2 -
b.
2rr[2.5/9.860J l/2 -
0.10 sec
Ratio of duration of blast load T to natural period TN T/T N - 0.10/0,10 - 1.0
c.
Ratio of peak resistance r u to peak load P ru/P - 750/1000 - 0.75
Step 3.
Using the ratios calculated in step 2 and figures 3·54 and 3-55, determine 'the value of ~/XE and tm/T N. For T/TN - 1
and from,figure 3- 54 from figure 3-55
Step 4. Determine
~
and t m
~/XE
- 3.7
~
- (3.7)X E - (3.7(r u/K E) - (3.7)(750/9,860) - 0.28144 ft,
tm/T - 0.77 tm
- (0.77)T - 0.77(0.10) - 0.077 sec
3A-30
TK 5~1300/NAVFAC P-397/AFR 88-22 Problem 3A-6
Maximum Response of a Single-Degree-of-Freedom System to Bilinear Blast Loads
Problem:
Determine Xm/X E• tm/T N and tE/T (when applicable) for a singledegree-of-freedom system subject to various bilinear blast loads.
Procedure:
Part (a) - Solution in Region D
Step 1.
Establish normalized parameters
Step 2.
Enter table 3-15 with the given C parameters and determine which figures have to be used.
Step 3.
Enter each of the figures determined in step 2, with the given values of the other two parameters and determine the region where the intersection points are located.
Step 4.
Based on the region where the intersection points are located. enter the appropriate figure and f~nd Xm/XE• tm/T N and tE/T.
Procedure:
Part (b) - Solution in Region C - Graphical Interpolation.
Step 1.
Same as step 1 in part a.
Step 2.
Same as step 2 in part a.
Step 3.
Same as step 3 in part a.
Step 4 ..
Set up a table as shown in table 3A-l .. Post each figure number and the corresponding values of Cl and C2• leaving a space between each line of information. Post in the spaces the appropriate values of Cl and C2 "needed for interpolation. Enter each of the figures determined in Step 2 with the given parameters and find the values of Xm/X E. Post these values in table 3A-l.
Step 5.
Use log-log graph paper to plot the points obtained in Step 4. Post these values in table 3A-l. using linear interpolation where necessary.
Step 6.
Procedure:
Plot on log-log graph paper the points which represent (Xm/XE' C2) for the given value of C. Use linear interpolation to find Xm/X E for given value of C2.
Part (c) - Solution in Region C - Mathematical Interpolation
Step 1.
Same as step 1 in part a.
Step 2.
Same as step 2 in part a.
Step 3.
Same as step 3 in part a.
3A-31
TM 5-l300/NAVFAC P-397/AFR 88-22 Step 4. Same as step 4 in part b. Step 5. Solve for lnY a and lnY b using equations 3-83 and 3-84. Step 6. Solve for lnY using equation 3-85. Step 7. Solve for Y using equation 3-86. Example 3A-6 Required: Solution:
Maximum Response of a Single-Degree-of-Freedom System to Bilinear Blast Loads Determine Xm/XE • tm/T N and tE/T (when applicable) for a singledegree-of-freedom system subject to various bilinear blast loads.
Part (a) - Solution in Region D
Step 1.
Given:
P/r u
-
1.0
T/TN - 3.0 Cl - 0.66 C2 - 50 Step 2.
Enter table 3-15 with Cl, - 0.66 and C2 - 50. 3-119, 3-120, 3-147, and 3-148 apply.
Step 3.
Enter each of the figures determined in step 2, with P/r u 1.0 and T/T N'- 3.0. Note that the intersection point is located to the right of the line of solid squares, defined as region D. In region D, the maximum dynamic response depends only on the shock load described by P/r u and T/TN; the gas,load described byClP/r u and C2T/Tn does not influence the maximum dynamic response. Consequently, figures 364a and 3-64b for a single triangular load pulse apply. Enter figure 3-64a with P/r u - 1.0 and T/T N - 3.0 and find Xm/X E - 3.55, tm/T N - 0.98, tE/T - 0.086.
Solution:
Note figures
Part (b) - Solution in Region C - Graphical Interpolation
Step 1.
,-
Given: T/T N - 0.10 - 0.06
-;
Step 2.
Enter table 3-15 with Cl - 0.06 and C2 - 20. 3- 108, 3-109, 3-133 and 3-134 apply.
3A-32
Note figures
TK 5-1300/NAVFAC P-397/AFR 88-22
Step 3.
Enter each of these figures with P/ru ~ 32 and T/Tn - 0.10. Note that the intersection point is not located in regions A, B or D. Therefore the intersection'points lie in region C and interpolation between charts is required to obtain a solution.
Step 4.
Set up table as shown in table lA-l below. ·Post each chart number and the corresponding values of Cl and C2 leaving a space between each line of information. Post in the spaces the appropriate values of Cl and C2 needed for interpolation. Enter figure 3-108 with P/r u - 32 and T/TN - 0.10 and find Xm/X E - 112. Post this value in the table. Enter figure 3-109 with P/r u - 32 and T/TN - 0.10 and find Xm/X E 86. Post this value in the table. Repeat this process for figures 3-133 and 3-134, and post values for Xm/X E in the table.
Step 5.
Use log-log graph paper to plot the points (112,0.068) and (86,0.046) which represent (Xm/XE, Cl) for C2 - 10 as shown in figure 3-267. Use "straight-line interpolation to find Xm/XE - 103 for Cl - 0.060. Post this value in the table. Repeat this process for C2 - 30, and find Xm/X E - 251 for Cl - 0.06 as shown in figure 3-267.
Step 6.
Plot on log-log graph paper the points (103,10) and (251,30) which represent Xm/XE' C2) for Cl' - 0.060. Use straightline interpolation for finding ~/XE - 182 for C2 - 50 as shown in figure 3-267. Thus the solution is Xm/XE - 182. Table 3A-l
Figure No.
Cl
C2
Xm/X E
3-108
3-109
0.068
112
0.060
10
--
0.046
10
86
0.060
20
--
f-
103
-
-
3-133
3-134
0.075
30
340
0.060
30
--
0.056
30
230
-
3A-33
I-
251
-
182
TH 5-l300/NAVFAC P-397/AFB.88-22 "
Solution:
Part (c) -
S~lution
in region C - Mathematical Interpolation
Step l.
Same as step -l.-of part (b) .
Step 2.
Same as step· 2 of part (b).
Step 3.
Same as step 3 of part (b).
Step 4.
Same as step 4 of part (b) .
Step 5.
using equation 3-83 and 3-84, find lnY a and lnYb. 1nYa - 1nY1
In[Y2/ Y1] 1n[C1/ Cll]
+
1n [ C12 / Cll] 1n(86/112)ln(0.060/0.068)
-
1nll2 + 1n(0.046/0.068)
1nYa
- 4.6339
In[ C14 / . C13 1
-
1n(230/340)ln(0.06/0.075) 1n340 + - - - - . . . : . - - - - - - 1n(0.056/0.075)
1nYb - 5.5304 Step 6.
Find 1nY from equation 3-85 1nY - 1nYa +
(lnYb - 1nYa)ln(C2/C 2l)
1n(C 23/C n
)
(5.5304 - 4.639)ln(20/10) - 4.6339 + 1n(30/10) 1nY - 5.1995 Step 7.
Solve for Y using equation 3-86 Y _ _ Y -
e 1nY e5.1995 181
3A-34
TM 5-1300/NAVFAC P-397/AFR 88-22
APPENDIX 3B LIST OF SYMBOLS
1M 5-l300jNAVFAC P-397/AFR 88-22
(1) (2)
a
acceleration (in./ms 2) depth of equivalent rectangular stress block (in.)
area (in. 2 )
A
area of diagonal bars at the support within a width b (in. 2 ) area of openings (ft 2 ) area of tension reinforcement within a width b (in. 2 )
A' s
area of compression reinforcement within a width b (in. 2 ) Grea of flexural reinforcement within a width b in the horizontal direction on each face (in. 2 )* area of flexural reinforcement wIthin a width b in the vertical direction on each face (in. 2 )* total area of stirrups or lacing reinforcement in tension within a distance, ss or sl and a width b s or b l (in. 2 ) . area of sector 1 and II, respectively (in. 2 )
(1) (2)
width of compression face of flexural member (in.) width of concrete strip in which the direct shear stresses at the supports are resisted by diagonal bars (in. )
width of concrete strip in which the diagonal tension stresses are resisted by stirrups of area (in.)
Av
width of concrete strip in which the diagonal tension stresses are resisted by lacing of area (in.)
Av
B
constant defined in pargraph
c
(1) (2)
distance from the resultant applied load to the axis of rotation (in.) damping coefficient
distance from the resultant applied load to the. axis of rotation for sectors I and II, respectively (in.) dilatational velocity of concrete (ft/sec)
c
shear coefficient critical damping
*
See note at end of symbols
38-1
TH 5-l300/NAVFAC P-39i/AFR 88-22 shear coefficient for ultimate shear stress of
one~way
ele-
ments
post-failure fragment coefficient (lb 2_ms4/in. 8) peak reflected pressure coefficient at angle of incidence a shear coefficient for ultimate support shear for one-way elements
shear coefficient for ultimate support shear in horizontal direction for two-way elements* shear coefficient for ultimate support shear in vertical direction for two-way elements* drag coefficient drag pressure (psi) ,~t·
peak drag pressure (psi) equivalent load factor shear coefficient for ultimate shear stress in horizontal direction for two-way elements* leakage pressure coefficient maximum shear coefficient impulse coefficient at deflection ~ (psi_ms 2/in. 2) Cu '
impulse coeffici~nt at deflection ~ (psi-ms 2/in. 2) shear coefficient for ultimate shear stress in vertical direction for two-way elements* I
(1) (2) (3)
impulse coefficient at deflection Xl (psi-ms 2/in. 2) parameter defined in figure ratio of gas load to shock load
impulse' coefficient at deflection'~ (psi-ms 2/in. 2) ratio of gas load duration to shock load duration distance from extreme compression fiber to centroid of tension reinforcement (in.)
d'
distance from extreme compression fiber to centroid of compression reinforcement (in.) distance between the centroids of the compression and tension reinforcement (in.)
* See note at end of symbols
38-2'
TK 5-l300/NAVFAC P-397/AFR 88-22
e
~
distance from support and equal to distance d or de (in.)
di
inside diameter of cylindrical explosive container (in.)
dl
distance between center lines of adjacent lacing bends measured normal to flexural reinforcement (in.)
dc o
diameter of steel core (in.)
dl
diameter of cylindrical portion of primary fragment (in.)
D
(1)
(2) (3)
unit flexural rigidity (lb-in.) location of shock front for maximum stress (ft) minimum magazine separation distance (ft)
nominal diameter of reinforcing bar (in.) equivalent loaded width of structure for non-planar wave front (ft)
DIF
dynamic increase factor
DLF
dynamic load factor
e
base of natural logarithms and equal to 2.71828 ...
(2E·)1/2
Gurney Energy Constant (ft/sec)
E
modulus of elasticity modulus of elasticity of concrete (psi) modulus of elasticity of reinforcement (psi)
f
unit external force (psi)
f' c
static ultimate compressive strength of concrete at 28 days (psi) dynamic ultimate compressive strength of concrete (psi) dynamic design stress for reinforcement (psi) dynamic ultimate stress of reinforcement (psi) dynamic yield stress of reinforcement (psi) static design stress for reinforcement (a function of f y ' i u
and 8) (psi)
static ultimate stress of reinforcement (psi) static yield stress of reinforcement (psi)
*
See note at end of symbols
38-3
TH 5-1300/NAVFAC P-397/AFR 88-22
(1) (2)
F
(3)
total external force (lbs) coefficient for moment of inertia of cracked section function of C2 and CI for bilinear triangular load
force in the reinforcing bars (lbs) equivalent external force (lbs) g
variable defined in table 4-3
h
charge location parameter (ft)
H
(1) (2)
span height (in.) distance between reflecting surface(s) and/or free edge(s) in vertical direction (ft)
height of charge above ground (ft) scaled height of charge above ground (ft/lb l/ 3) height of structure (ft) scaled height of triple point (ft/lb l/ 3) unit positive impulse (psi-ms) unit negative impulse (psi-ms) sum of scaled unit blast impulse capacity of receiver panel and scaled unit blast impulse attenuated through concrete and sand in a composite element (psi-ms/lb l/3) unit blast impulse (psi-ms) scaled unit blast impulse (psi-ms/lb l/ 3) total scaled unit blast impulse capacity of composite element (psi-ms/lb l/3) scaled unit blast impulse ca~acity of receiver panel of composite element (psi-ms/lb /3) scaled unit blast impulse capacity of donor panel of composite element (psi-ms/lb l/3) unit excess blast impulse (psi-ms) unit positive normal reflected impulse (psi-ms) i
r
-
unit negative normal reflected impulse (psi-ms) unit positive incident impulse (psi-ms)
i
s
*
-
unit negative incident impulse (psi-ms)
See note at end of symbols
38-4
TM 5-1300/NAVFAC P-397/AFR 88-22 moment of inertia (in. 4 )
I
average of gross and cracked moments of inertia of width b (in. 4 ) moment of inertia of cracked concrete section of width b (in. 4 ) moment of inertia of gross concrete section of width b (in. 4 ) mass moment of inertia (lb-ms 2-in.) j
ratio of distance between centroids of compression and tension forces to the depth d
k
constant defined in paragraph
K
(1) (2)
unit stiffness (psi-in for slabs) (lb/in/in for beams) constant defined in paragraph
elastic unit stiffness (psi/in for slabs) (lb/in/in for beams) elasto-plastic unit stiffness (psi-in for slabs) (psi for beams) equivalent elastic unit stiffness (psi-in for slabs) (psi for beams) load factor
KLM
load-mass factor
(KLM)u
load-mass factor in the ultimate range
(KLM ) up
load-mass factor in the post-ultimate range
KM
mass factor
KR
resistance factor
Kl
factor defined in paragraph
KE
kinetic energy
1
charge location parameter (ft) spacing of same type of lacing bar (in.) (1)
(2)
span length (in.) except in chapter 4 (ft)* distance between reflecting surface(s) and/or free edge(s) in horizontal direction (ft)
length of lacing bar required in distance sl (in.)
*
See note at end of symbols
3B-5
TM 5-l300/NAVFAC P-397/AFR 88-22
embedment length of reinforcing bars (in.) wave length of positive pressure phase (ft) wave length of negative pressure phase (ft) wave length of positive pressure phase at points band d, respectively (ft) total length of sector of element normal to axis of rotation (in. ) m
unit mass (psi-ms 2/in.) average of the effective elastic and plastic unit masses (psims2/in.) effective unit mass (psi-ms 2/in.) effective unit mass in the ultimate range (psi-ms 2/in.)
~p
effective unit mass in the post-ultimate range (psi-ms 2/in.)
M
(1)
(2)
unit bending moment (in-lbs/in.) total mass (lb-ms 2/in.)
effective total mass (lb-ms 2/in.) ultimate unit resisting moment (in-lbs/in.) moment of concentrated loads about line of rotation of sector (in. -lbs) fragment distribution parameter equivalent total mass (lb-ms 2/in.) ultimate unit negative moment capacity in horizontal direction (in. -lbs/in.)* ultimate unit positive moment capacity in horizontal direction (in. -lbs/in.)* ultimate unit negative moment capacity at supports (in.1bs/in. ) ultimate unit positive moment capacity at midspan (in.1bs/in. ) ultimate unit negative moment capacity in vertical direction (in. -lbs/in.)* ultimate unit positive moment capacity in vertical direction (in. -lbs/in.)* * See note at end of symbols
3B-6
1M 5-1300/NAVFAC P-397/AFR 88-22
n
(1) (2)
N
number of adjacent reflecting surfaces
modular ratio number of time intervals
number of primary fragments larger than Wf p
reinforcement ratio equal to
p'
reinforcement ratio equal to
or A' s
or
bd
A' s
bd c
reinforcement ratio producing balanced conditions at ultimate
strength Pm
mean pressure in a partially vented chamber (psi)
Pmo
Peak mean pressure in a partially vented chamber (psi) reinforcement ratio in horizontal direction on each face*
Pr
reinforcement ratio equal to PH + Pv
Pv
reinforcement ratio in vertical direction on each face*
p(x)
distributed load per unit length
P
(1) (2)
pressure (psi) concentrated load (lbs)
negative pressure (psi) interior pressure within structure (psi).
interior pressure increment (psi) fictitious peak pressure (psi) peak pressure (psi) peak positive normal reflected pressure (psi) p r
peak negative normal reflected pressure (psi) peak reflected pressure at angle of incidence a (psi) positive incident pressure (psi) positive incident pressure at points band e, respectively (psi)
*
See note at end of symbols
3B-7
1M 5-1300/NAVFAC P-397/AFR 88-22
Pso
peak positive incident pressure (psi)
Ps o -
peak negative incident pressure
Psob,Psod,Psoe
peak positive incident pressure at points b , d, and e, respectively (psi)
q
dynamic pressure (psi) dynamic pressure at points band e, respectively (psi) peak dynamic pressure (psi) peak dynamic pressure at points band e, respectively (psi)
(1)
r
(2)
unit resistance (psi) radius of spherical TNT (density equals 95 lb/ft 3 charge (ft))
r
unit rebound resistance (psi)
Ar
change in unit resistance (psi) elastic unit resistance
elasto-plastic unit resistance (psi) ultimate unit resistance (psi, for slabs) (lb/in for beams) post-ultimate unit resistant (psi) radius of hemispherical portion of. primary fragment (in.) R
(1) (2)
total internal resistance (lbs) slant distance (ft)
distance traveled by primary fragment (ft) radius of lacing bend (in.) normal distance (ft) equivalent total internal resistance (lbs) ground distance (ft) total ultimate resistance total internal resistance of sectors I and II, respectively (lbs) spacing of stirrups in the direction parallel to the longitudinal reinforcement (in.)
*
See note at end of symbols
3B-8
TK 5-l300/NAVFAC P-397/AFR 88-22
spacing of lacing in the direction parallel to the longitudinal reinforcement (in.) S
height of front wall or one-half its width, whichever is smaller (ft)
SE
strain energy
t
time (ms) time increment (ms) any time (ms) time of arrival of blast wave at points b, e, and f, respectively (ms)
(1) (2)
clearing time for reflected pressures (ms) container thickness of explosive charges (in.)
rise time (ms) time to reach maximum elastic deflection time at which maximum deflection occurs (ms) duration of positive phase of blast pressure (ms) duration of negative phase of blast pressure (ms) fictitious positive phase pressure duration (ms) fictitious negative phase pressure duration (ms) fictitious reflected pressure duration (ms) time at which ultimate deflection occurs (ms) time to reach yield (ms) time of arrival of blast wave (ms) time at which partial failure occurs (ms) T
duration of equivalent triangular loading function (ms) thickness of concrete section (in.) scaled thickness of concrete section (ft/lb l/ 3 ) effective natural period of vibration (ms) rise time (ms) thickness of sand fill (in.)
*
See note at end of symbols
3B-9
TK 5-l300/NAVFAC P-397/AFR 88-22 scaled thickness of sand fill (ft/lb l/ 3) u
particle velocity (ft/ms) ultimate flexural or anchorage bond stress (psi)
u
shock front velocity (ft/ms)
v
velocity (in./ms) instantaneous velocity at any time (in./ms) boundary velocity for primary fragments (ft/sec) ultimate shear stress permitted on an unreinforced web (psi) maximum post-failure fragment velocity (in./ms) average post-failure fragment velocity (in./ms) velocity at incipient failure deflection (in./ms) initial velocity of primary fragment (ft/sec) residual velocity of primary fragment after perforation (ft/sec) striking velocity of primary fragment (ft/sec) ultimate shear stress (psi) ultimate shear stress at distance de from the horizontal support (psi)* ultimate shear stress at distance de from the vertical support (psi)*
v
volume of partially vented chamber (ft 3) ultimate direct shear capacity of the concrete of width b (lbs) shear at distance de from the vertical support on a unit width (lbs ./in.)* shear at distance de from the horizontal support on a unit width (lbs/in.)* volume of structure (ft 3) shear at the support on a unit width (lbs/in)* shear at the vertical support on a unit width (lbs/in.)*
* See note at end of symbols
38-10
TH 5-l300/NAVFAC P-397/AFR 88-22
Vs v
shear at the horizontal support on a unit width (lbs/in.)*
Vu
total shear on a width b (lbs)
w
weight density of concrete (lbs/ft 3 )
W
weight density of sand (lbs/ft 3 )
W
charge weight (lbs)
Wc
total weight of explosive containers (lbs)
Wf
weight of primary fragment (oz)
Wco
total weight of steel core (lbs)
Wc l ' Wc 2
total weight of plates 1 and 2, respectively (lbs)
Ws
width of structure (ft)
WD
work done
x
yield line location in horizontal direction (in.)*
X
deflection (in.)
Xa
any deflection (in.)
Xe
elastic deflection (in.)
Xep
elasto-plastic deflection (in.)
Xf
maximum penetration into concre ce of armor-piercing fragments (in. )
Xf '
maximum penetration into concrete of fragments other than
lSn
maximum transient deflection (in.)
~
plastic deflection (in.)
Xs
(1)
s
armor-piercing (in.)
maximum penetration into sand of armor-piercing fragments
(2)
(in. ) static deflection
ultimate deflection (in.) equivalent elastic deflection (in.) partial failure deflection (in.) y
yield line location in vertical direction (in.)*
* See note at end of symbols
38-11
TN S-1300/NAVFAC P-397/AFR 88-22
z
scaled slant distance (ft/lb l/3) scaled normal distance (ft/lb l/3) scaled ground distance (ft/lb l/3)
B
(1)
angle formed by the plane of stirrups, lacing, or diagonal reinforcement and the plane of the longitudinal reinforcement (deg)
(2)
angle of incidence of the pressure front (deg)
(1)
coefficient for determining elastic and elasto-plastic resistances
(2) y
particular support rotation angle. (deg)
coefficient for determining elastic and elasto-plastic deflections
y
deflections increase in support rotation angle after partial failure (deg)
e
support rotation angle (deg)
e
angular acceleration (rad/ms 2) maximum support rotation angle (deg) horizontal rotation angle (deg)* vertical rotation angle (deg)*
Eo
effective perimeter of reinforcing bars (in.) summation of moments
(in.~lbs)
sum of the ultimate unit resisting moments acting along the negative yield lines (in.-lbs) sum of the ultimate unit resisting.moments acting along the positive yield lines (in.-lbs) ductility factor v
Poisson's ratio
(1) (2)
capacity reduction factor bar diameter (in.)
assumed shape function for concentrated loads
"'(x)
assumed shape function for distributed loads
* See note at end of symbols
38-12
e
1M 5-l300/NAVFAC P-397/AFR 88-22 free edge simple support
IIIIIII
fixed support
xxxxxxx
either fixed, restrained, or simple support
*
Note.
walls.
This symbol was developed for two-way elements which are used as When roof slabs or other horizontal elements are under consideration,
this symbol will also be applicable if the element is treated as being rotated into a vertical position.
3B-13
TH 5-1300/NAVFAC P-397/AFR 88-22
APPENDIX 3C BIBLIOGRAPHY
TM 5-1300/NAVFAC P-397/AFR 88-22
1.
Blast Resistant Design, NAVFAC Design Manual 2.8, Department of the Navy, Naval Facilities Engineering Command, Alexandria, VA 22332, April 1982.
2.
Design of Structures to Resist Nuclear Weapons Effects, ASCE Manual of Engineering Practice, No. 42, American Society of Civil Engineers, New York, NY, 1961. (1983 Edition under preparation)
3.
Designing Facilities to Resist Nuclear Weapons Effects, Structures, TM5858-3, Headquarters, Department of the Army, Washington, DC.
4.
Structures to Resist the Effects of Accidental Explosions, Technical Manual TM5-l300, Navy Publication NAVFAC 1'-397, Air Force Manual AFM 8822, Department of the Army, the Navy, and the Air Force, Washington, DC, June 1969.
5.
Crawford, R. E., et aI, The Air Force Manual for Design and Analysis of Hardened Structures, AFWL-TR-74-l02, Air Force Weapons Laboratory, Kirtland Air Force Base, New Mexico, 87117.
6.
Healey, J., et at, Design of Steel Structures to Resist the Effects of HE Explosions, by Ammann and Whitney, Consulting Engineers, New York, NY, Technical Report 4837, Picatinny Arsenal, Dover, NJ, August 1975.
7.
Hopkins, J., Charts for Predicting Response of a Simple Spring-Mass System to a Bilinear Blast Load, Technical Note N-1669, Naval Civil Engineering Laboratory, Port Hueneme, CA, April 1983.
8.
Stea, W., et aI, Nonlinear Analysis of Frame Structures Subjected to Blast Overpressures, by Ammann and Whitney, Consulting Engineers, New York, NY,
Contractor report ARLCD-CR-77008, U.S. Army Armament Research and Development Command, Large Caliber Weapon Systems Laboratory, Dover, NJ, May
1977 . 9.
Stea, W. Weissman, S., and Dobbs, N., Overturning and Sliding Analysis of Reinforced Concrete Protective Structures, by Ammann and Whitney, consulting Engineers, New York, NY, Technical Report 4921, Picatinny Arsenal, Dover, NJ, February 1976.
10. Javornicky, J. and Van Amerongen, C., Tables for the Analysis of Plates, Slabs and Diaphragms Based on the Elastic Theory, Bauverlag GmbH. , Wiesbaden/Germany (2nd Edition 1971). 11. Roark, R.J. and Young, W. C., Formulas for Stress and Strain, McGraw Hill, New York (5th Edition 1975). 12. Cohen, E. and Dobbs, N., Design Procedures and Details for Reinforced Concrete Structures Utilized in Explosive Storage and Manufacturing Facilities, Ammann and Whitney, Consulting Engineers, New York, NY, Annals
of the New York Academy of Sciences, Conference on Prevention of and Protection against Accidental Explosion of Munitions, Fuels and Other Hazardous Mixtures, Volume 152, Art. 1, October 1968.
3C-l
•
"STRUCfURES TO RESIST THE EFFEcrs OF ACCIDENTAL EXPLOSIONS"
r
CHAPTER 4. REINFORCED CONCRETE DESIGN
TK
5-1300/~AVFAC
P-397/AFR 88-22
CHAPTER 4
REINFORCED CONCRETE DESIGN' I~TRODlicTION
.
P
.,
!
•
r.
4-1. Purpose
• The purpose of this' manual ·i;.to present methods of'design'fot protective construction u~ed in facilities'for development, testing, productio~,.storage, maintenance, modification, inspection, demilitarization, and disposal of' explosive materials. . . . . '" .' • 'I'
4-2. Objective The primary objectives are to establish design procedures and construction techniques whereby propagation of explosion (from 'one structure or part ofa structure to another)'or mass detonation can be pr~vented and to provide protection for personnel and v;'luable equip;;'';n!:,' ': " The secondary objectives are to: (1)
e, "
( 2) (3)
(4)
Establish the blast load parameter~ ~equired for design of p~ote~ i:ive s truc t;"r~s. '. Provide methods for calculating the'dynamic respon;e'of 'structural • .' r .'. , elements including reinforced'concrete, and structural steel: Establish construction details and procedures.necessary to afford the requi~ed strength resist the applied blast 'loads. ". ',; . ',' Establish'guidelines,for siting expl!,sive facilitie~,to.obtain maximum cost effectiveness in, both' the p Lannd.ngvand structural •. arrangements, providi~g' ~losure~, 'and preventing' damage to interior portions ,of structures because of. 'j..structural motion, shock, and " ., , fragment perforation. " ,..."
to ~
","'
.
" ,
"
.,,~
-r ,
4-3. Background
:-.'
For the,first 60 years of . the 20th century; criteria' and met~ods ';(based upon . , . ' . results of catastrophic , events were used, for the design of explosive facilities. The criteria and methods did, not include a detailed or reliable,quantitative basis for assessing the degree ,of protection afforded by the. protective faCility. In the late.1960's· quantita~ive procedure~ were set,forth in the first edition of' the present.manual,'''Structures ~o R~sist'the Effects ~f Accidental Explosions". This manual was based 'on extensive research"and development programs which permitted ..a.inore,reliable approach to current and future'design·requirements. ',Sin~e the original 'publication of this manual, more extensive' testi')g and d~ve~opmen~ pr;olp:;ams 'have t~lken place: This , additional research included ,work with materials other 'than reinforced concrete which was the principal con~t~uction ma~erial referenced in the initial version of the manual. . ,
~
I .
,..
,
_
~
"
• .
,
•
,
Modern methods for the manufacture and storage of explosive materials, which include many exotic chemicals: fuels, and prop'ellants, require-less,space for a given qt1,mtityof explosive materiar'than'w~s prevfous Iy needed. Such concentration of explosives increases the possibility of the propagation of accidental explosions. (One accidental explosion causillg the detonation of
4-1
-.~
TM
5-l30~/NAVFAC
P-397/AFR 88-22
other explosive materials.) It is evident that a requirement for more,accurat~ design techniques is essential. This'manual describes rational design methods' to provide the required structural prot,:ction. " These design methods a~count for the ~lose-in effects of a 'detonation including the high pressures and the nonuniformity of blast loading on protective structures or barriers. These methods also account for intermediate and farrange effects for the design of structures located away from the explosion. The dynamic response of struct~res, constructed of v~r~~us m~terials, or combination of materials, can be calculated, and details are given to provide the strength and ductility required by the design. Tge design approach is directed, primarily 'toward protective structures'subjected to the ,effects of a high explosive detonation. However, this approach is general, and it is applic~ble to the design of other explosive environments as well as other explosive materials as mentioned above. The design techniques set forth in this manual'are based upon the re~ults of numerous full- and small-scale structural response and explosive effects tests of various materials conducted in conjunction with' the d~velopment of this manual and/or related projects. 4-4. Scope It is not the intent 'of this manual to establish ,safety criteria., Appiicable documents should,be consulted for this purpose. Response predictions for personnel and equip~ent ,are included for information. " ' ' In this ~~nuaicaneffort is However, sufficient general been included in order 'that situations other th~n those
made to:cover th~ more probable design'situations. information on protectiye desig~ techniques has application of the basic theory can be made to which' wet;e fully cons Lder'ed , "
.
.
.
"
,',.
.
,
This manual is applicable to the' design of prot~ctive structures subjected to the effects associated with high explosive detonations. For these 'design situations, the manual will apply for explosive quantities less than 2s,QOO pounds for close-in effects. However, this manual is also applicable to other situations such as far~ or intermediate-range effects. For these latter cases the design procedures are applicable for explosive quantities iri the order of 500,000 pounds which, is the maximum quantity of high explosive approved for aboveground storage 'facilities in 'the Department of Defense manual, "Ammunition and Explosives Safety Standards", DOD 60ss.9-STD. Since tests were pr~marily directed toward the response of structural steel 'and reinforced concrete elements to blast overpressures, this manual concentrates on design procedures and techniques for these materials. However; this does'not imply that concrete and steel are the only useful materials for protective'construction. Tests to establish the response of wood, brick blocks, 'and plastics, as well as the blast at cenuat Lng andjnas s e ff'ec cs tof soil are contemplated. ' .The result" of these tests may require, at a later)date~' the supplementation of these design methods for these and other materials~~ • Other manuals are,available to design protective structures 'against, the effects of high explosive or nuclear detonations. The procedures' in these manuals will quite often 'complement this manual and 'should be consulted'for specific applications'. ( , ,
.
4-2
"
•
TK 5-l300/NAVFACP-397/AFR 88-22
Computer programs, which are' consLs cent; with procedures and techniques contained in the manual, have been approved by the appropriate representative of the US Army, the US Navy, the US Air Force and the Department of Defense Explosives Safety Board (DDESB). These programs are available'through the following repositorie~:" ," ,
,',
'.
t
Dep.i.£tment of the'A~niy , Comminder and"Director U.S. Army Engineer
(1)
",
•• "
'.',
I,
... .:. ..... "'.t
,".t
~~t~rways Experi,~ent.Station
Post Office Box '631 ~icksburg,'Missi~sippi 39180-0631 Att~: :WESKA ' Depart~ent of the Navy
(2)
..
Commanding Officer Naval Civil 'Engineering ,~borat~ry Port Hueneme, California 93043 Attn: . , i Code . L51 ,.
(3)
~
l
'
Department of the Air Forc~ Aeio~pace' Structures Information and Analysis Center Wright Patterson Air 'Force Base Ohio 45433. " Attn: . . .AFFDL/FBR '-
.~.
,~,'
t'
" ,'
,-
~
',"
,
If any modifications to these programs are required, they will be submitted for review by. DDESB and 'the above services.' 'Up~n concu~rence of the revisions, the necessary changes'will be made and notification of the changes,will be made by the individual repositories. 4-5. Format
!. . -,
This manu~l is ~ubdivided into six spe~ific' chapter~ dealing with various aspects' of des'ign'., The titles of these 'chapters ",,-,'" '.: .: . , are as,, follows': '
Chapte; ,1 Chapter' 2 Chapter 3 Chapter'4' Chapter 5 Chapter 6
In'troduction· ~
'j
;.
.~
)llas t, Frag~ent,: anfShock ,Lo~ds' /," Principles of Dynamic Analysis Reinforced Concrete,Design Structural Steel D~~ign' " . " Special Considerations in ,Explosive Fa~i1ity' Design •
_'"
. L . , . .
When applicable, illustrative examples are included in the ,Appendices . .
'.
"
.'.
'.".
,.
"
,
".
' : ) . "
Commonly accep~ed symbols are, used as much as possible. Howevo;r, protective design invo1ve~ many different scientific and' engineering fields', and, therefore, no attempt is made to standardize completely all the symbols used: . Each symbol is defined where it is first used; and in the lIst of symbols at the end of each chapter.
4-3
TM 5-1300/NAVFAC P-397/AFR ,88-22
1,1'
CBAP'l'ER , ,CONTENTS '
4-6. General
,
.~.
.' ,
, ,~
'.
-, - . '
~.,
I
.,,,~
,.
This chapter is concerned with the design of above ground blast . resi~tant ' . \.:-, , concrete structures. Procedures are presented to obtain the dynamic strength of the various structural components of concrete structures. Except for ,the• I ' .' '. • . ... ..ll particu~ar case of the design of laced reinforced ~opc,r~te.elem~~t~" 7he dynamic analysis of the structural components is presept~d,in Chapter;~, ,
"
•
. ,'...
•
.'
,t'
The dynamic strengths of both the concrete and reinfor~ement'under yarious stress conditions are given for the appl~~abledes~gn.range and t~e;allowable deflection range. Using these strengths, the ultimate dynamic' capacity of various concrete elements are given. These capacities include the ~ltimate moment capacity for various possible cross-section types, ultimate shear
.'
capacity as a measure of diagonal tension as well ~sultimate .direct shear'and punching shear, torsion capacity of beams,.and.the development of 'the reinforcement through bond with the concrete.". ,,' •
.'
'I'
"
This chapter contains procedures for the design of non-laced (conventional reinforcement) and laced concrete slabs and walls as'well as procedure~ for,., the design of flat slabs, beams and columns," Procedur~s ate presented for the design of laced and non-laced slabs and beams, for close-in effects,whereas procedures for'the design of'non-laced and flat~slab~, bea~~' and columns are given for far range effects. It is not ecoriomical to use laced slabs for far range effects. Design procedures are given for the flexural response' of oneand two-way non-laced slabs, beams and flat slabs which undergo limited deflections. Procedures are also given for large deflections of these elements when they' undergo tensile membran';,actio':'. 'Laced reinforced'slabs' are designed for fle~ural action for'both li~ited and large deflections. Lastly', the design of co Iumns is presented for elastic or, at best',' slight ' plastic action. ,C The above design procedures are concerned with the ductile response of structural elements. Procedures.' , are also given-'" for the '.brittle mode response ,",¥,' . ' , ,",1 " of concrete elements; The occurrence of both'spalling arid scabbing of the concrete as well as pr6tection against their ~ffects is treated. In addition~ procedures are presented for post-failure fragment- design- of laced concrete ,. . walls and slabs. The resistance of concrete elements to primary fragment impact is consid';red. For the primary ~bigme~ts determined in Chapter' 2', ' methods are presented to determine if fragment is'embedded in or perforate~ a concrete wall. If embedment occurs,' the depth of penetration is determined and the occurrence ,of spalling of the far face can. be evaluated. If perfora-, tion 'occurs; the resid.ual velocity of tHe fragment is determined. c. h. 1 "
>
•
a
Required cons t ruc t Lon "details' and procedures for' conventionally reinforced arid laced reinforced concrete structures is the last item discussed in the chapter.' 'Conformance to' these details', will'Insure' a' due tile response 'of'· the , structure co the applied dynamic loads ;,!;" ..... I. '
.
...'
•
....
I
"";~:~!
.',
...
,~
4-4
')'
,
;". ,
t
•
TM 5-l300/NAVFAC P-397/AFR 88-22 BASIS FOB. STRUCTURAL'DESIGN
4-7. General Explosive storage 'and opet;"at1ng facilities are' designed to p'rcvfde a predetermined level of protection against the hazards o~ accidental explosions. The type of protective structure depends upon both 1;he donor and acc~pt~r systems. The donor system (amount , type and loc_~tion of the' pot.errt La'l Ly detonating explosives) produces the' damaging output: while the accept:or system (personnel, equipment, and "accepcor" explosives)' requir~s'"a level. of protection. The' protective 'structure or s t ructura'l. elements are de s Lgned to shield against or attenuate the hai~rd9us effects to levels which are t:ol~rable to the'accept9r system. v
Protectiv~ concrete strtictures are classified ~s'eitheI'cshelters or barri~rs. '. t . . " " • -. ; Shelters,enclose the rece~ver system and are generally located far from a potential explosion. Barriers, on itihe other' hand, gene,rally enclose the'donor system arid; consequently, are Locaced close to the potemtial explosion:' A shelter is a fully enclosed structure which is designed to prevent its contents (acceptor system) from being subjected to the direct effects of blast pressures and fragments. A barrier. may be" either a fully enclosed structure (containment structure) r or an, open 'structure '(b~rricadE! "or cubicle tyPe ' structure ,in ~hich one or m~re surfaces 'are frangible ~r open to the atmosphere). Barriersar;, generally designed to resist close7in detonations. Their purpose is to prevent acceptor explosives, and to'a'lesser extent, personnel and equipment from being subjected to primar)' fragment impact and to attenuate blast pressures in acco~dance with,the structural configuration of the barrier. ' ,
4-8. Modesqf'St,ructu~al'Beha~ior "!~
' . !
'
The response of a;' concrete element can be expressed in terms of two modes of structural behavior; ductile and ,brittle. In the ductIle mpde of response the element may attain large inelastic deflections without.complete collapse. While, in the brittle mode, partial failure or total co l Laps e of the element occurs. The selected behavior of an element for a par t Lcul ar design is governed by: ,(1), the niagni~ude and duratl,on of thebillst output;" (2) the' occurrence .of primary fragmehts, ,and (3) the function of the .p:ro!:"cti'!'e structur~, i.e., shelter or barrier depending,upon'the'protection level required. -, . ,
I'
,
..i · ·
,
.
•
'
4-9. Structural ... Behavior of ,Reinforced Coricrete 4-9. L 'General.. When a reinforced concr
4.5 .
TM 5-l300/NAVFAC P"397/AFR 88-22 The resistance-deflection curve ,shown in Figure 4-1 demonstrates the flexural action of a reinforced concrete 'element. When the element is first loaded, the resistance ideally increases linearly with deflection until yielding of the reinforcement is first initiated. As the element continues to deflect,' all the reinforcing steel yields and the resistance is constant with,increasing deftection., Within this yieldra~ge at a deflection corr~spondirigto 2 degrees sUp'port rotation,_ the compression co~.crete crushes.
For e Lemeritis
without shear reinforcement, this crushing'of the concrete results in failure of the element. For elements with shear reinforcement (single leg stirrup; 'shown in Fi'guie 4-2 'or laci~g shown in Figure 4-3) which properly ti~ the' flexural reinforcement, the crushing 'of the concre ce results in'a slight loss of capacity since the compressive 'force is transferred to the, co':pression ' reinforcement. As the element is further'deflected, the reinf~rcement enters in~o its strain hardening region, and the resistance increases with increasing deflection. Single leg stirrups will restrain the compression reinforcement for a short time into its strain hard~ningregion. At four (4) degrees support rotation, the e Lemerit; loses Its struct~ral integrity and f~ils.' 'On the other hand, lacing thro~gh its truss action will restrain the'reinforcement through its entire strain hardening region'until tension failure of the" reinforcement occurs at 12 degrees support rotation. . Sufficient shear capacity must b~ affo~ded by the concrete alone,or iri ' ' combination with shear reinforcement in order to develo.p'the flexural capacity of an element (Figure 4-1). An abrupt shear failure can occur ,at any time " during the flexural response if,the,flexural 'capacity exceeds the shear' capacity of the element, ' 4-9.2. Ductile Mode of Behavior in the Far D~sig~ Range In the far design range, the distribution of the applied blast load is fairly uniform and the deflections required to absorb the loading are'cornp~ratively small. Conventionally reinforced (i.e., non-laced) concrete elements with comparatively minor changes to standard reinforcing details are peif~ctly adequate to resist such loads. While laced reinforcement could be used, 'it' would be extremely uneconomical ' .'to. do so. "" . ' . The flexural response of non-laced reinforced concrete elements is demon~ strated through the resistance-deflection"curveof Figure 4-1. Fo~ elements' without shear reinforcement: the ultimate deflection is limited to deflections corresponding to 2' degrees support rotation whereas 'elements with shear ' reinforcement are capable of attaining 4 degrees support rotation. For ease of constructiori, single leg stirrups (Figure 4-2) are used as shear reinforcement in slabs and walls, This type of reinforcement is capable of pfoviding shear resistance as well as ,the necessary restraint of the flexural reinforcement to enable the slab to achieve this increased deflection.
~. ~
A conventionally reinforced slab may attain subs t ant t a l.l.y Lar ge r deflections than those corresponding to 4-degrees support rotations . These increased deflections are possible only i f the ,element has suf f Lci'ent; rateral re,itraint to develop in-plane forces. The resistance-deflection' curve 'of Figure'4 C4 ' illustrates' the structural response' of an 'element having lateral restraint:' Initially, the element 'behaves essentially as a flexural member. If the ",' t
lateral restraint prevents small motions, in~plane compressive "forces are
developed. Under flexural action, the capacity is constant with increasing deflection until the compression concrete crushes. As the deflection in4-6
:' I
TH 5-l300/NAVFAC P-397/AFR',88-22
creases 'further and ,the load'carried by the slab decreases, membrane action in the slab is developed. The slab carries load by the reinforcement net acting as a plastic t~nsile membrane. The capacity ~f the element increases with increasing deflection'until the reinforcement fails in tension. 4-9.3. Ductile Mode of"Behavior in the Close-in Design Range' , . ,
Close-in det(;nati~ns produce nonuniform, high intensity blast load. Extremely high-pressure concentrations are developed wpich, ~n turn, can'produce local (punching) failure of an element. To maintain the structural integrity of elements subjected to these loads and to permit the large deflections necessary. to balance the kihetic energy produced, lacing reinforcement has been ", developed. ' ,
,
Lacing reinforcement is shown in figur~'4-3 while' a typical laced wall is shown in Figure 4-5. A laced element is reinforced syffimetrically, i.e., the compression'reinforcement is the same as the tension reinforcement. The straight flexural reinforcing bars on each face of the element and the intervening concrete are tied together by the truss action of continuous bent diagonal bars. ~ This system of lacing contributes to the'int~grity of the protective elemen~ in the following ways:
(1)
Ductility of the flexural reinforcement, including the strain region, is fully developed.' , Integrity of the concrete between the.two layers of flexural
ha~dening
(2)
reinforcement is ma Lntia Lned despite
maSSiVE!
c r ackl.ng ,
(3)
Compression reinforcement is restrained from bucklfng.
(4)
High she a r stresses at the suppo r t s' .are. resisted.
(5)
Local shear failure produced by the high intensity of the peak blast pressures is prevented. , Quantity and velocity of post~failuie fragments produced during the brittle mode of behavior are reduced.
(6)
l
'
The flexural response of a laced reinforced concrete element is illustrated by the entire resistance-deflection curve shown in Figure 4-1. The lacing permits the element to attain large deflections and fully develop the reinforcement through its strain hardening region. The maximum deflection of a laced element corresponds to 12 'degrees support rotation. Single leg stirrups contribute to the integrity of a protective element in much the'same way as lacing, however, the stirrups are less effective at the closer explosive separation distances. The explosive charge must be located further away from an element containing stirrups than a laced element. In addition, the maximum deflection of an element with single leg stirrups is limited to 4 degrees support rotation under flexural action or 8 degrees under tension membrane action. If the charge l,?cation permits and reduced support rotations are required, elements with single leg stirrups may prove more economical than laced elements. 4-9.4. Brittle Mode of Behavior The brittle behavior of reinforced concrete is composed of three types of concrete failure: direct spalling, scabbing and post-failure fragments. Direct spalling consists of the dynamic disengagement of the concrete cover over the flexural reinforcement due to high intensity blast pressures. 4-7
TM 5"l300/NAVFAC P-397/AFR88-22 Scabbing ,also consists of the disengagement of the concrete cover over the flexural reinforcement, howev~r, scabbing is due to the element attaining large deflections ... Finally post-failure fragments are the r e suLt; of the, collapse of an element and are'usually the more serious. Post-failure fragments are generally large in number and/or size with substantial velocities which can result in propagation of explosion, Spalling and scabbing are usually only of concern in those protective structures where personn~l, equipment, or sensitive explosives require protection. Controlled postfailure fragmen,s are 'only permitted where the acceptor system' consists' of relatively insensitive' explosives . . ' • The two tyPes of spalling, direct spalling and sc~bbing; occur during the ' . ductile mode of behavior. Because direct spaLl.Lng is dependent upon the transmission of shock pressures, fragments formed from this type of spalling are produced immediately after t~e blast pressures strike the wall., Scabbing, on the other hand,'occurs during the later stages of the flexural (ductile mode) action of the 'element. Both'types of spalls affect the capacity of the element to resist the applied blast load. ' Post-failure fragments are the 'result 'of a flexur~l failure 'of an element. The 'failure characteristics of laced and unlaced elements differ significantly, The size of failed ,sections of laced element is fixed by the location of the yield lines. The element fails at the yield lines and the section between yield lines remain intact. Consequently; failure of a laced element consists of a few large sections (Figure 4-6). On 'the other hand, failure of an unlaced element is a result of a los~ of structural integrity and the fragments take the form of 'concrete rubble (Figure 4-'7). The velocity of the post failure fragments from both laced 'and unlaced elements is a:function of the amount of blast overload~ However, tests have 'indicated that the fragment velocities of laced elements are'as low as 30 percent of the maximum velocity of the rubble formed from similarly loaded unlaced elements '.
•
4-8
•
'o.,
YIELD
z
;'! (I)
..,iii
a:
-,
,:
,.
q
.1
"t
I' ';
~
........... ~.
"
~'
q.
I
,.. NO SHEAR REINF. MAXIMUM DEFLECTION
SINGLE LEG STIRRUPS
LACING
"
Figure 4-1
."
.~.
. Typical resist'ance-deflection:,"'curve for flexural response of concre~e.elements
.. SINGLE LEG STIRRUPS .
. ...
.
'
· ... .
'
' "
'
. ·, , .'. , • 4'
•
'0, " •
\ :. Ii:,
..
• .:a '",
•
,
"
" ~
., " •• ,'0 ,
'
. .. .
•
.
" "
. '.' " "
....
'0-
"
. ., .
• 0.
00
.'
,
......"
1>,
'0
...
•, , ,
..
'0' ";..
I
FLEXURAL
"
Figure 4-2
.:
-;
REINF. -
..
Element reinforced with 'single leg stirrups
"
j,',~,
,. "-~'!
4-10'
,
!.
»,
LONGITUDINAL FLEXURAL REINF. LACIN(; REINF.
TRANSVERSE FLEXURAL
Figure 4-3
REIN~
Lacing reinforcement
4-11.
'"oz ~
Ul
iii a:
'"
.. ,
~
'
.
:
DEFLECTION
FLEXURAL ,ACTION
Figure 4-4
TENSION MEMBRANE ACTION
Typical "resistance-deflection curve for tension membrane response of concrete
4·12
TYPE
MaM
HORIZ. LACING
YERT. LACING
3/4" ....~_CL_._. TYPE "b"
...o N
,
(DIAGONAL BARS TYPE
FLOOI~
. SLAB
A
3" tONSTRUCTION JOINT
CL.
PAD
Figure 4-5
Typical laced wall
4-13
~.
Figure 4-6
Failure of a laced element
4·-14
Figure 4-7
Failure of an unlaced element
TK 5-l300/NAVFAC P-397/AFR 88-22 DYNAMIC STRENGTH OF MATERIALS
4-10. Introduction A structural element subjected to a blast loading exhibits a higher strength than a similar element subjected to a static loading. This increase in strength for both the concrete and reinforcement is attributed to the rapid rates of strain that occur in dynamically loaded members. These increased stresses or dynamic strengths are used to calculate the element's dynamic resistance to the applied blast load. Thus, the 'dynamic ultimate resistance of an element subjected, to a blast load is greater than its static ultimate resistance.:'
."'
, . . , -
Both the concrete and reinforcing steel exhibit greater strength under rapid strain rates. The higher the strain rate, the higher the'compressive strength of conc re cetand the higher the yield and ultimate strength of the reinforce-, ment.
This phenomenon is accounted for in the design of a blast resistant structure by using dynamic stresses to calculate the oynamic ultimate resistance of the reinforced
~oncrete
members.
4-11. Stress-Strain Curve Typical stress'-strain curves for concrete and reinforcing st~el are shown in Figure 4-8. The' solid curves represent the stress-strain relationship for the materials when tested at ,the strain and loading rates specified in ASTM Standards.. At a , higher strain'rate, their strength is greater, as illustrated by the dashed, curves. Definitions of the symbols used in Figure 4-8 are as follows: • "
f'c - static' ultimate compressive strength of concrete f'dc - dyn.~i~ultimate co~pressive strength of concrete f y - static yield stress of reinforcing steel f dy - dynamic' yield stress of reinforcing steel f u ..; static ultimate stress of reinforcing,ste'el f du - dynamic,ultimate stress ofreinforcin~:steel •
~
.
I
_
Es
modulus of, elasticity for'reinforcing steel
Ec
secant modulus of elasticity of concrete
eu
rupture strain
..
From the standpoint of structural behavior and design, the most important effect of strain rate is the increased yield and ultimate strengths of the reinforcement and the compressive strength of the concrete. For typical strain rates encountered in reinforced concrete elements subjected to blast loads, the increase in the yield strength ~fthe-steel and the compressive strength of the concrete is substantial. The ultimate strength of the
TIl ;5~130CI~AVFAC P-397/AFR 88"22
reinforcement is much less sensitive to the strain rate .. The increase in the ultimate strength is slight and the strain at which this stress ?ccurs i~ slightly reduced. There is essentially no change with strain rate in the modulus of elasticity and rupture strain of the steel. In the case of concrete, as the strain rate increases the scant modulus of elasticity increases slightly, and the. strain at maximum stress and rupture remain nearly constant.
'I,
''''
"
4-12 ;'Allowable Material Strengths'
"
,-. ,
4-12.1. General .-/
..
~.
-,
The behavior of a structural element ~ubjected·to a bl,,~t loading depends upon the ultimate strength and ductility of 'the materials from'which it is constructed. The required strength' of /i'ductile element is; considerably 'less' than that necessary for a brittle element to resist th" same applled loading. A ductile element maintains its peak or near:peak'streIlgth through'large plastic strains whereas a brittle element fai1s,'abruptly with"llttle'energy absorbed in' the plastic range. Reinforced concrete wIth well tied and ' anchored ductile reinforcement can be classified as a ductile material. , .
.
•
4-12.2. Reinforcement
I
Reinforcing steel, designated by the·American Society for Testing and Materials (ASTM) as A 615, Grade 60, is c~nsid~red to have adequate ductil"it};, in . sizes up to No. 11 bars. The large No. 14 bars also have the desired duc: . tility, but their usage is somewhat restricted due to !:heir special require'ments of spacing and anchorage. No. 18 bars are not r ecommended for use in blast resistant' structures. "For all reinforcement, duc:tility is reduced at bends, lapped splices, mechanical"splices';' e cc . , and location ,of·,the'se ' anchorages near points of maximum stress is'undesirabl" and should'lie avoided. ,
;.'
,..,,'.,'
.
",
Reinforcing steel ha~ing a minimum yield of 75,000 psl can'be produced'having chemical properties' similar to'ASTM A 615,' Grade 60. However, production of thiS steel requires Ii'special order to be 'placed in whIch iarge quantities of individual bar sizes (in' the order' 'of' 200 tons per bar Size) must be' ordered. It is recommended that for these high strength bars only straight lengths of bars be utilized, splicing of oars be avoided'and application'of this reinforcement be limited to members designated to attain an elastic respons~ or a slightly plastic' resporise' (~/XE}e'ss' than .~r,.equal. to. 3).\.... I t is desirable to' know the stress-strain'relationship'for the reinforcement being utilized in order to calculate the ultimate resistaric~ of· an element. This information is not usually available; however, minimUm values of the yield stress f y and'the ultim~te tensile stress f u are,~equired'by ASTM Standards, For ASTM A 615; Grade' 60 reinforcement, the 'minimum yield and ultimate stresses' are 60,000 psi and 90,000 psi, re'spec:tl.velY-:" Review of num~rous mill test reports for this steel indicate yiel.d stresses at least 10 percent greater than the ASTM minimum, and ultimate stresse's 'at least' equal to but not much greater than,the ASTM minimum. Therefore i.t·is.~eco~en~ed that for des Lgn purposes, ,the minimum ASTM yield stress be Lricreased ,by 10 percent while the minimum' ASTM ultimate stress be used' without any increase, 'So that, the recommended design values for ASTM A 615, .,Grade'60'reinforcement, are: . , - l. - • '. .~
f
~
y·.. 66,000 'psi
'"
4-17
.
TK 5-1300/NAVFAC
P-39~/AFR 88~22'
and
'"
.
f u - 90,000 psi 4-12.3. Concrete Even though the magnitude of the concrete strength i~ on,~y> significant in the calculation of the ultimate strength of elements with 'support rotations less than 2 degrees, its effects on the behavior of 'elements with both small and, large support rotations are of equal importance. The shear capacity of an element is dependent upon the magnitude of the concrete strength. For elements with ,small suppprt rotations (less than 2 degrees)', the use of higher strength, concrete may eliminate.,theneed,for.shear reinforcement; while for elements requiring shear reinforcement, the amount of reinforcement is reduced as th!" concret;e ,strength is 'i~creased. .For elements with large support ' rotations (2 to 12 degree~),. the cracking,an~ crushing of the concrete associated with the 1arger,rotations is less severe when higher strength concrete'is, employed. Therefore, t~e strength, of the concrete used in a blast resistant structure may be selected to ,suit the particular design requirements of the structure. ' However, under'no circumstances should the concrete ' strength f' c be less than 3,000 psi. It is recommended that 4,000 psi' strength concrete be used in all blast resistant structures regard1e~s ,of the magnitude of the blast load and ,deflection criteria. 4-13. Dynamic,Design Stresses for Reinforced Concrete 4:13.1.
Gener~l
Ductility is a significant'paramete~ influencing the dynamic response and behavior of,reinforced'concrete members subjected to blast loadings. The, importance.of.ductility ipcreases as the durati~n of the blast load decrease~ relative to the natural period of the member. In general, to safely withstand a blast load, ,the required ultimate resistance ,,!Iecreases wi.th increasing, ductility of the msmber., In.fact, the ultimate resistance required of ductile members is 'considerably less than that required for brittie members which fail abruptly with littl~ energy absorbed in, the plastic i'ange .of behavior. ' A ductile' member is one that develops plastic hinges in regions of maximum moment by first yielding of the tension reinforcement followed by, crushing of the concrete. This behavior is typical of under-reinforced concrete sections. A section can be designed to be very ductile by maintaining an under-rein, forced section, adding compression reinforcement, and utilizing lacing bars to' prevent buckling o(,tpe compression reinforcement. For a laced section, the . reinforcement i~ stressed through its entire strain-hardening region, that is, the stee~ reaches its ultimate stress f du and fails at its rupture strain Eu ' In a flexural ,member, 'l'he straining of, the, reinforcrement, and consequerit Ly its stress, is.expressed,in,te~s of its angular ,support ~otations. , ~.
.
f.~
I .
4-13.2. Dynamic. :Increase Factor . .' "
• >
i'
.. ,
. •
.
The- dynamfc im,rea"e factor, DlF, is equal. to the nitio'of the dynamic str!,'i~
to the static stre~'i', e. g., fdy/f y' fdu/f u ,,;nd f' dc/f~c" The DIF depends, upon the rate of strain ,of, the element, increasing as the strain rate increases. The design curves for' the DIF for the unconfined compressive strength of concrete and for the yield stress of ASTM A 615, Grade 60, reinforcing steel, are 4-18
TM 5-l300/NAVFAC P-397/AFR 88-22,.
given in Figure 4-9 and 4-10, respectively. The curves were derived from test data having a maximum strain rate of 10 x 10- 3 in./in,/msec, for concrete and 2.1 x 10- 3 in./in./msec. for steel. Values taken from these design curves are conservative estimates of OIF and safe for desigri purposes. Values of OIF have been established' for design of members in the far design range as well as for members in the close-i~ design range. These design values of OIF are given in Table 4-1. Because of the increased magnitude 'of the blast loads and subsequent increase in the strain rate, the dynamic increase factors, for elements subjected to,'a close-in detonation are higher than 'those for elements subjected to an explosion located far from the element. '"
'
-
"
-j
.
•
•
"
The, design values of OIF presented 'in Table 4-1 vary not only for the design ranges and type of material but also 'with the state of stress (bending, diagonal tension, direct shear, bond, and compression) in the material. The values for fdy/f y and f'dc/f'c for reinforced concrete members in bending , assume the strain rates in the reinforcement and concrete are 0,0001 in./in./ msec. for the far design range and 0.0003 in./in,/ msec, in the close-in design ran~e. For members in compression (columns). these strain rates are 0.0002 in./in./msec.and 0.0095 in./in./msec. The lower strain rates in compression (compared to bending) account for'the'fact that siabs, beams' and girders "filter" the,dynamic effects of the blast load, Thus, the dynamic loan reaching columns is typically a fast "static" load (long rise time of load) which z-eau Lc s 'in lower strain rates in co'lumns" Thes e strain rates and the
corresponding values of OIF in Table 4-1 are considered safe values for design purposes.
Available data is not sufficient to permit the construction'of a design curve for the DIF for the ultimate' stress of ASTM A 615 Grade 60 reinforcing steel. However, it is known that the increase in the ultimate strength of the steel is small and; therefore, not a significant factor in the design of reinforced concrete members, A nominal: va Lue of th~ DIF is given 'in Table 4'-'1. The listed values of O~F for shear' (diagonal. tension and direct shear) and bond are more conservative than~ for' bending ·compression. This conservatism
or
is justified by the need to prevent brittle shear'imd bond failure and to account for unce r t afnt'Le ad.n t,he design . process for' shear and .bond'. . ' A more accurate estimate of the OIF may be 'obtained utilizing the DIF design curve for concrete and steel given in Figure 4-'9 and 4 .. 10, respectively. The increase in capacity of flexural elements is primarily a function of the rate of strain or the reinforcement, in particular, the tim" to reach yield, t E , of the, reinforcing ,steel~ The average rate of'strain for both the concrete' and steel may be ,obtained considering the strain in the materials at yield and the tIin~ to reach' yield, The .membe r is 'first designed' (p rccedures given in subsequent ~ections ) using the D~F values given in Table 4-1, The time' to reach yield, t E', is then calculated using the response chart s presented in Chapt-er 2.' For the value t E , rhe average s t r a Ln rate :In the materials can be, obtained. The 'average, strain rate in the concrete (based 'on f' de be Lng reached at €c - 9,992 in./in.) is: " 4-1 "
while the average strain rate' in the reinforcement is:
TH5~1300/NAVFAC P-397iAFR 88-22 f' s
t t
where for~concrete
E'C
average strain rate
f'S
average strain rate for reinforcement
,t E.- time'to yield the~reinforcement For th~ strain rates obtain~d from Equations 4-1 and 4-2, the-actual DIF is . . obtained for the concrete and reinforcement fr~m Figure 4-9 and.4-l0, respectively. If the difference between the calculated DIF values and the design, values of Table 4-1 are small, then the correct values of DIF are those calculated. If the difference is larg~, the calculated values of DIF are used as new estimates and the process is repeated until the differences between the "estimated" and, "calculated", values of DIF are small, The process'converges' "
.
very rapidly and. in most cases, the ae cond iteration of, the process corrve r'ge s
on the proper values of DIF. In most cases,the values of DIF' obtained from Table 4-1 are satisfactory for design and the determination of the' actual DIF values is' unwarranted~ " However, the DIF values can significantly effect the final design of certain members" and the extra calculations required to obtain the actual DIF values are fully warranted. These include dee p members, members ,subjec,ted to ' impulse-type blast loads, and members' designed to sustain large deflections. The actual DIF values (usually:higher than the desig~ value~,of Table 4-1) result in a more realistic estimate of the ultimate flexural resistance and, therefore, the maximum shear and bond stresses which must be resisted by the member .
. '
For'the elasto-plastic or plastic design, of concrete elements, an equivalent elas~ic curve is considered rather than the actual elasto-plastic resistancedeflection function, The time to reach yield t E is C9mputed'based on this curve using 'the equivalent elastic.deflection XE ana stiffness KE , Actually, the reinforcement along the supports yield in less time than t E whereas the r~inf?rcement at mid-span yields at a time greater ,than 'tEo These differences are compensati~g errors, ,There,fore, - the time to reach yield, t E for the equivalent curve when used in Equations 4-1 and' 4-2 produces an accurate average DIF for the concrete and 'reinforcement at' the critical sections throughout a reinforced concrete element. ,
,.
, .'
4-13'.3. Dynamic Design Stress~s,
The magnitude of stresses ,produced in the reinforcement of an element responding'in the elastic range, can be related, directly to the strains. However, in the plastic range'the stresses. cannot be related ~irectly to the'strains., An estimate of the average "stress over portions of the plastic range can be made by relating this average stress to the deflection of the element.' The deflection is defined in'terms of the angular rotation 'at the supports. The average, dynamic stress is expressed as a function of the dynamic yield stress f dy and the dynamic ultimate s t re's s f du' Criteria for the dynamic stresses to be used in, the plastic design of ductile reinforced concrete elements are presented in Table 4-2. The dynamic design stress is expressed in, terms o~ f dy' f du' and f'dc' The value of these terms 4-20
"
TK 5-l300/NAVFAC P-397/AFR 88-22 is determined by multiplying the appropriate static design stress by the appropriate value of the DIF (Table 4~l). so that: ;
4-3
f(dynamic) - DIF x f(static)
..
4-21
, fdc ~
~
f
--------------,/~I--<,
'
·
.
/
....
I/) I/)
/
.5f c'
--A
/~~ ;;/~
I
<, <,
I
I I I - - - ASTM
/
I/)
..
I
//
W
II: I-
//"
c-·----7--
I - - - RAPID
Ec
I 0.002S
i~
I
f
......
"<,
l
STRAIN 'RATE STRAIN RATE
(a) STRESS-STRAIN. CURVE FOR CONCRETE
f du --------::------f u --------,."...-----,/
z--, .......
",,'/
I
~
I/) I/)
~-_-/
I
...
fy
W II: lI/)
II I I I
I I
I
- - - ASTM STRAIN RATE - - - RAPID STRAIN RATE
0.07 ~Eu < 0.23 in.!in. APPROX. 0.01 to 0.02 in.!in. APPROX.
STRAIN .e (in.! in.) (b) STRESS-STRAIN CURVE FOR STEEL
Figure 4-8
Typical stress-strain curves for concrete and reinforcing steel
4-22
I
I
0.002 approx.
.;
I
u <0.005 in.!in. APPROX.
STRAIN,E (in.!in.)
f dy
•
....
'u
e
e
e
-
_ 0 <,
o
_"0
w
IW
_0:: lL U
-z 00
~U
O::lL
00 IU I iiI-
wz
UlW 'd:0::
..,, '" ~
-i-t -r :" , ,,,i "'D"'" f- +:, "j
wIO::Ul
U
zW ->
uU5 -Ul
-r
~W 'd:o::
Za..
>-~
00
H
U
w
l'd:
1.10
+1+1:
~
I-
-l
::::>
STRAIN RATE. E(in.{i~:/rrsec.l
Figure 4-9
De s Lgn . curve
~for·DIFfor .ultimate
compressive' strength' of concrete
1.70
1.6o ~
1.1. 0_>' ~,
a::
>.
1.50
0."
~
U
it
...
-
,rJ) WrJ) (/)W
..., ...""
1.40
<.
1.30
::!:
1.2 o
zc -..J u!!:! -)-
zo )-1.1.
C
1.1 o
I. 00 10'6 STRAIN RATE I t {in.lin.lmsec'>
Figure 4-10
e
Design curve for DIF for yield stress of ASTM A 615 grade 60 reinforcing steel
e
e
TK 5-l300/NAVFAC P-397/AFR 88-22,
.. ' Table 4-1
Dynamic Increase Factor (DIF) for Elements
Des~gn
of Reinforced Concrete t.
f.
, FAR DESIGN'RANGE
CLOSE-IN DESIGN RANGE
TYPE OF STRESS Reinforcing Bars
Concrete
R~inforcing
Bars
Concrete
f du If';
f'dc/f'c
1.05
1.25
. f dy /fy
f dy /fy
f du /f u
f' dc/f',c
Bending
1.17
1.05
1.19
1. 23
Diagonal Tension
1.00
--
LOa
Direct Shear
1.10
1.00
1.17 1.10
.
Bond Compression
..
1.10
_. 1.00 -
1.00
1..10
1.10
1.00
1.10
1.05
1.00
1. 23_.
1.05
1.00
--
1.12
1.13
--
1.16
.-. ...
"
4-25
r
..
TK 5-l300/NAVFAC 'P-397/AFR 88-22
•• Table 4-2
Dynamic Design Stresses for Design of Reinforced Concrete Elements TYPE OF STRESS
.
TYPE
OF
REINFORCEMENT
Diag~naf
Tension
and
2
f dy +,
Compression
S<8mS12
Stirrups
O;;8m:S Z 2 < ~ s 5
(1) (2) (3) (4)
,
Diagonal
Bars Column
(f'dv+ fdu)/2
f'de· (2) (2)
fdy.
t'de ,
f dy
l'de
f dv'
f'de
O<8m:S Z
f dy
f'de
2<8m:5S
f dy + C(du- tdy)/4
f'de
5
(fdy + f du ) / 2
f'de
0
f dy
f'de
5
Tension
Compression
REINFORCEMENT, fda CONCRETE, f d e f dy (1)
Lacing
Shear
DYNAMIC DESIGN STREss
0.< ~:!l 2
Diagonal
Direct
(DEGREES)
Tension 0
Bend~n8
MAXIMUM SUPPORT ROTATION. -8m
< 8
m :5 12
2<8m:5S
S
<
8m :s
~2
(4 )
f~y
+ (~du-fdy)/4
(fdv
(3 )
+ f du ) / 2
(3)
f dy
['de
Tension reinforcement only, Concrete crushed and not effective in resisting moment, Concrete is considered not effective and shear is resisted by the reinforcement only. Capacity is not a function of support rotation.
4-26
TH 5-1300/NAVFAC P-397/AFR 88-22
STATIC PROPERTIES' 4-14. Modulus of Elasticity 4-14.1. Concrete The modulus of elasticity of concrete Ec is equal to: E" - w 1.5 33 (f~ )1/2 c c c
4-4
for values of Wc between 90 and 155 lbs/ft 3 where concrete and normally equal to 150 lbs/ft 3 .
c is the unit weight of
W
4-14.2. Reinforcing Steel The modulus of elasticity of reinforcing steel'E s is:' E - 29 x 10 6 psi s
4-5
4-14.3. Modular Ratio The modular ratio n is: 4-6 and may. be taken as the nearest whole number.
4-15. Moment of Inertia The deter~ination of the'deflection of a reinforced'concrete member in the elastic and elasto-piastic ranges is complicated by the fact that the effective moment of inertia of the cross section along the element changes continually as cracking progresses. It is'further complicated by the fact that the modulus of elasticity of the concrete changes as the stress increases. It is recommended that the computation,of defle~~ions throughout this volume be based upon empirical relations . de t e.rndned froin test' data. . . ..:"f..
.'
The ayerage moment of inertia I a should be used in all deflection calculations and is given by:
Ia -
4-7 2
,
'
For the design of beams, the' entire cross-section is consider~d, so that , bTc 3 ,
'"
4-8a
12 and 4-8b 4-27,.
TM 5-1300/NAVFAC
P~397/AFR
88-22
For the design of slabs, a unit width of the cross-section is considered, so that T 3 c
, 4-9a
12 and: I
,- Fd 3
c
where:
4-9b
.' I
a
average moment of inertia of concrete cross section
Ig
moment of inertia of the gross concrete cross section (neglecting all reinforcing steel)
I
moment of inertia of cracked concrete cross section
c
b
width of beam
Tc
thickness of gross concrete cross section
F
coefficient given in Figures 4-11 and 4-12
d
distance from extreme compression fiber to centroid of tension reinforcement
considers
The moment of inertia of the cracked concrete section" the compression concrete area and steel areas transformed into equivalent concrete areas
and is computed about the centroid, of the transformed section. The coefficient F varies as the modular rati~ n and the amount of,reinforcement in the section. For sections ,with tension reinforcement only, the coefficient F is given in Fdgur'e 4-11' .wh~'le for sections with. equal r~lnforcement on opp~sit'~ faces, the coefficient F is given in Figure'4-l2. ' The variation in the cracked moment of inertia obtained from Figure 4-ll tand 4-12 is insignificant ~or-low reinforcement ratios." The variation incre~ses for the larger ratios. Consequehtly, for the comparatively low reinforcement ratios normally used in, slab elements either chart,maY,be,t,lsed with,negligible error. For the higher reinforcement 'ratios normally used for beams'; Figure' 412 must be used for equally reinforced sections whereas a weighted average' from Figure 4-11 and 4-12 may be used for sections where the compression steel is 'less than the tension steel. For one-way members the reinforcement ratio p used to obtain the factor F should be ,an average of the tension steel at the supports and midspan'. Also, the effective depth d used to compute the cracked moment of inertia I c should be an average of the effective depth at the supports and midspan. However, for two-way members, the aspect ratio must be considered in the calculation of the cracked moment of inertia. Average values for the reinforcement ratio p' and effective depth d should be used to obtain the cracked moment of inertia in each direction and cracked moment of the member is then obtained from:
4-28
~"
TK 5-l300jNAVFAC P-397/AFR 88-22
., where: IcV -. cracked moment'of inertia in, vertical direction IcH.~·crackedmoment of inertia in horizontal direction
L
span length
H
- span height
•
.,
4-29.
4-10
11
;;'-, ,C'
1
0.060
0.050
... ~
z
,r •
0.040
t
'
1
l&l
Q
......
'
~
u
D..
••
0.030
.
t
•
tr
l'
','-. =, '++ 0.020
,
5:1': ' "
-
,l +
'8 . til
t
.~,
"
...
• i
.; r .. Il~~l·t
0'
0.010
,
..
......... r- .. ;~~ t
·do
o
o
Figure 4-11
0.004
0.008
-r'r
0.012 REINFORCEMENT RATIO p: As / bd
~tt1 . ~-~ t
0.016
Coefficient for moment of inertia of cracked sections with tension reinforcement only
4-30
_
.
0.020
" j
1 I . .;.
.070
~
~-+-
t1
ttr
·-t
. d±
r +I
T _ j.
- oj;". .• ... ·t ..;
:.
It'i; •
++::"1
11
T
:Jtt 1
t ..
+
-+
. , .
•
f
Jim f
I,
e>
';'j-++
.060
itT
t
-H
.i
fZ UJ U
... u,
,. +
r.040 -:-
.•. ;
++.
:t
.1.'
~
UJ
ou
e·
.,
",,~
'.030 -r 'r'
• T
~t
n ="Es/E c rc:=Fbd 3
.020 ...j
+
',> •
.~
-I
H'"
f .010
,
•
." ~ -t'
-,-
.I 1.,.
,
o .'
, 'hf.::l$$ 1 c++- 1. .1-,
o
..004
.008
,"
,"1
.012
REINFORCEMEt{T, RATIO
Figure 4-12
.016 p.7As/bd (
Coefficient for moment of inertia of cracked sections with' equal reinforcement on opposite faces
4-31
.020
TN 5-1300/NAVFAC P-397/AFR 88-22
'.
ULTIMATE DYNAMIC STRENGTH OF SLABS 4-16. Introduction Depending upon th~ magnitudes of the' blast output and permissible deformations, one of three types of reinforced'concrete cross sections (Figure 4-13) can be utilized'in the design or analysis of blast resistant concrete slabs: a.
type I
The COncrete is effective in resisting moment.
The
concrete cover over the reinforcement on both surfaces
of the element remains intact. b.
type II'
The concrete is crushed and not effective in resisting moment. Compression reiriforcement equal to the tension reinforcement is required to resist moment. The Concrete cover over the reinforcement on both surfaces of the element remains intact.
c.
type III -
The concrete cover over the reinforcement on both
surfaces of the element is completely disengaged. Equal tension and compression reinforcem~nt which is properly tied together is required to resist moment. Elements designed using 'the full cross section (type I) are usually,encountered in those structures or portions of structures 'designed to resist the blast output at the far design range. This type of cross section is utilized in elements with maximum deflections corresponding to support rotations less than 2 degrees.' Maximum strength of an element is obtained from a type I cross section. Type I elemen~s may be reinforced on either one or:both faces. However, due to' rebound forces,' reinforcement is required on both faces of an element. " Crushing of the concrete cover over the compression reinforcement is exhibited in elements which undergo support rotations greater than 2 degrees. This failure results in a transfer of the compression stresses from the concrete to the compression reinforcement which, in ~urn, results in a loss of strength. Sufficient compression reinforcement must be available to fully develop the tension steel (tension and ,compression reinforcement must be equal), Elements which sustain crushing of the concrete wi,thout any disengagement of the concrete cover are encountered in structures at the far design range when the maximum deflection conforms to support rotations greater than 2 degrees but less than ~ 'degrees. . ' Although the ultimate strength ,of elements with type III cross sections is no less than that of elements with type II cross sections, the overall capacity to resist the blast output is reduced. The spalling ,of the Concrete cover over both layers of reinforcement, caused by either the direction transmission of high pressures through the element at the close-in range or large deflections at the far range, produces a loss of capacity due to the reduction in the concrete mass. A'moredetailed treatment of the phenomena of crushing and spalling is presented in subsequent sections of, this, .chapter'.
4-32
TM 5-l300/NAVFAC
88"22
P-39~/AFR
The ultimate dynamic strength of reinforc~d concrete section~'may b~calcu~' lated in accordance with the ultimate'strength design methods'of the American Concrete Institute Standard Building Code Requirements for 'Reinforced Concrete (hereafter referred to as the ACI Building Code). The capacity reduction factor 0 which has been established for convent Lone l, st.•tic load condd t.Lons vLs omitted for the determination of ult,imate' dynamic scrength. Safety or reliability of the protective, ,structure is inherent in the establishment of the magnitude of the blast output for the donor charge, and in the ,criteria specified for deflection, support rotation, 'or fragment velocity. 'Other permissible departures from the criteria for stattc or gas'pressure loadings are described below. , ., _
.
c
..
f
'.
_.
'.
0.',.
1
,
Although certain formulae for elements constructed with conventi.onal weight ' " concrete are given in the following paragraphs of this chapter, more.d~tailed information and design aids are given in the bibliography. '·Because tests have not yet been conducted to determine the response of lightweight concrete elements designed for close-in and far design ranges, the"pertinent f6rmulae for this type of concrete are not included in the manual. However, lightweight concrete may be utilized for structures designed for low pressures (less than 10 psi); but the reduction in mass from conventional weight concrete must be accounted for in the design to maintain the blast resistant capacity of the
~ I!'
structure.
4-17. Ultimate Moment Capacity
t' ..
..
"
'.-
4-17.1. Cross Section Type I The ultimate unit resisting moment ~.of a rectangular tension' reinforcement only is given by:
~ectJon
of
wi~th
~
b with
-.
in which:
4-12
"
."
where: As
area of tension reinforcement ,within thewidth.b .....
dynamic design ~tress for ,reinforcement d
.. ,
distance from extreme compression fiber to centroid of tension reinforcement '
J;;
••• ~
. ~.~ .
•
, J
,
'
.
,
width of compression face
f' dc - dynamic ultimate compressive strength of <:oncrete .
"
, .' .e '
The reinforcement ratio p is defined as:, • '.'.i:..
4-33
,
.. ..
a - deptll of equivalent. rectangular stress blo<:k ... _ ".'l.' .... .. ... b
\
4-13
.,
'
TH 5-1300/NAVFAC P-397/AFR 88-22 To insure against. sudden compression failures, the .reinforcement ratio p must not exceed 0 :'75 of the ratio Ph which produces balanced conditions at ultimate strength and is given by:. 4-14 where:
"'"
.'.,.
•
, f
0.8.5 for f'dc up .to 4,000 .psi~and is reduced 0.05 for each'1;000 ,. . \. psi in excess of 4,000 psi. For a rectangular section of width b with compression reinforcement, the ultimate unit resisting moment .is: '~'
~ -; (A's)fds!.b][d - (a/2) J . , .
+ (A!;sfds
;b)
.
I '
(d- di)
in which: a
4-16
".
where: A's
area of compression reinforcement within the width b
d'
distance from extreme compression fiber to centroid.of compression reinforcement
..
depth of'~quiva1ent rectangular stress block The reinforcement ratio p' is:
~-17
p' - A'slb
Equation 4-15 is valid only when the compression steel reaches the. value f ds at ultimate strength, and this condition is satisfied when: p - p'~ 0.85K1(f'dcd'/fdsd) [87,000/(87,000 - f ds)]
4-18
If p - p' is less than' the v~lue given'by Equation 4-18 or when compression steel is neglected, the calculated ultimate unit resisting moment should not exceed that given by Equation 4-11. The quantity p - p' must not exceed 0.75 of the value of Pb given in Equation 4-14. 4-17.2. Cross Section Types II and III The ultimate unit resisting moment of type'II and type III rectangular sections'~f width b is:
4-19 where: I
As
'area of tension or compression reinforcement within the. width b 4-34
TH 5-1300/NAVFAC P-397/AFR"'SS"-22 dc -
distance i>etween'th'e centroids of'the compression and the tension ,~
reinforcement
'p'
The reinforcement ratios p a~d
are equal to:
t', •
'" .c - A' /bd ' ' ,Ps, - , ,P
\-20
r:
The above momerit" caplicity:can oriri be "obtaine'd' whenthl! ar';iis of the tension and compression r~inforcemei1t a,r" e,qu'al. , in'~dditi~n; th,is"reinfcir:cemeI),t ,must be properly restrained so as to m~in~ain the integrity of 'the erement'wh"n large deflections are encounte'red.· ... " ' , '0' .' " ,
.,
~
4-17.3. Minimum Fiexu~al'Reinf~fcement, , ~
~'
,
....
-
To insure proper structural behavior under both conventional and blast loadings, Ii mi!,lm\im amount of fle'xural :rel.nf~r:ce'ment Ls req·uired.. this quantity• of reinforcement Lnsures that' the' moment capac Lty ' of the reinforced t . section is ,greater than that corr,esponding, to 'the plain concrete .sec t Lon computed from l,ts modulus of rupture .. Failure, of, a .plain concrete section is quite sudden. 'ilso, this minimum'reinforc"ment prevents exces~ive cracking and deformations under conventional loadings'. . .. ' . " -, F
;
..
•
'I
-.
. '
, .
..
l
'. . '
.
•
_
'
,".
:
'
.'!
.'
,
"
, J
The minimum reinforcement required for slabs is somewhat less than that requirea for' beams, since an overload would be distributed laterally and sudden failure would be less likely. Th~'ml.nimum' reinforcement ratio for dynamic design .of slabs is given in, Table ,4-3. However ~ this quantity of reinforcement must also'~atisfy static' design requirements: Except for blast loads in the order of magnitude of static loads, the minimum requirements for c, dynamic loads will control. In cases where minimum requirements ,for static cond1tions control, the quantity pf'reinforcement,m~st be at least 1.33 times the quantity required by static .analysis or ,0.0018 times the gross concrete area, whichever is less.. '.'
.
."
.
~
. .<
PI'
•
,',,
..
r..
Concrete '~ections ~ithc~e~sion i~f~r~ ..ment only are not 'permitted. For type I sections, compressipn reinforcement equal to at least'one half the required tension reinforcement must be provided. This reinforcement is required to resist the ever present rebound forces. Depending upon the magnitude of these reb~und forces, the compression reinforcement required may be greater than one half the tension reinforcement and substantially greater , than the minimum. quantity given in Table 4-3. For type II and'III cross sections, the compression,reinforcement is always equal to the tension, reinforcement. 4-lS. Ultimate Shear (Diagonal Tension) Capacity 4-lS.l. Ultimate Shear Stress The ultimate shear stress v u' as a measure for type I sections from:
~f
_diagonal tension, is
compute~
"
4-21 and for type II and III sections frpm: t
4-35
4-22
TK5-1300tNAVFAC P-397/AFR 88-22 where Vu is the, total shear, at the section a distance d the support. For the latter the support and the' section
on a width b at either the face of the support, or (type I) or'dc (type; II oz: ill) fr~m the ,face of case, the shear at sections between the' face of d or d c away nee?not be consiqered critica~; ,
.
For'l~ced elements, -the shear stress is always calculated at 4c fr~m the face of the support (or haunch) since the lacing and required diagonal bars provide sufficient corner reinforcement. For'unlaced elements, the shear stress is caLcu'l.aced at d' .f'~om ,the face of tpe. support'~for ~!"\ose ,me.mb~~s ,'that c8:4se :,:' comp re ss Lon in their supports (Fig: 4~14a). This provision should not be ' applied for 'tHose members that c'aus''; tension i',,' their supports' (Fig. 4-14b)':' For this case, the ultimate shear stress should be' calculated at the face of the support. In 'addition, the shear within the connection should be investigated and special corner reinforcement should be provided.
The ultimate' shearstJ;ess 'vu must not' "exceed ~IO Xf' dc),\ in sections' using' stirrups. The thickness of such' sections must "tie increased and/or the ~ . ~ . '. '" .' ", quantity. of flexu,ial re,infor!,ement', reduced in .ord~r'~to bring the value of vti 'within tolerable limits. In sections using lacing, "there Is n,o restric,tion on' V u because of ,the continuity provided by this .type of shear ,reinforcement, However, for large shear stresses the' area of ,the lacing bars required may' '. . " become impractical. "
"
.
,
"
4-18.2. Shear Capa.city of Unreinforced Cclnc'rete .. The shear stress permitted on' an unreinfo':ced w~b ~f a member subj ect;'d to .. , flexure only'is limited to: • ' " . . "1-.1. V
-,
.
dc)I~2 + 2500p] <-. "3.5(f'd~)1/2 ~ c - [l.9(f· .. , ' '. '. '
.
•
•
. ' ,-
-j
~
,
• ,-
where p is tne reinforcement ratio of the tension requirement at the support. For the computation of the reinforcement ratio, d is used for type I sections and dc for type II and III. sections, For m~mber.s ,subjected to significant axial tension, the shear stress 'permitted on 'an urireinforced web is limited to: , )1'
g)
v c,- 2(1 + Nu / 500A
,
(f'dc)I/2 ~~.O
.
while for, significant 'axial comp res s Icn ;" V
,
,
'
) 1/2 .' . , g )( f' dc-": c - 2 (1 + Nu/2000A
, 4-25
where: axial load normal to the cross section A'g -' gross area of'the cross section The axial load Nu must occur simultaneously with the total shear Vu on the section in order to apply Equations 4- 24 and 4- 25. ' The value of Nu shall be taken as positive for compression and negative for tensi9n. The simplified dynamic analysis normally performed is not sufficient to accurately determine the time variations between'the desired forces'and , 4-36
,
TK 5:l300/NAVFAC
p C 397/ AFR
88-22
moments from which Nu 'and Vu are obtained. Unless 'Ii' tim~-hisi:.
by
t,
4-18.3. Design of Shear Reinforcement
Whenever the ultimate shear stress Vu exceeds the shear capacity V c of the concrete, shear reinforcement must, be provided to carry the excess. This shear re-inforcement can be either stirrups or lacing, dependdrig uporithe magnitudes of the blast loading and support rotation permitted'., Stirrups can be used only for elements designed to attain small deflections ',under flexural behavior. Lacing can, also be used for elements des Lgned to attain'small deflections; however, lacing must be ,used for eiements designed,to attain large deflections. Therefore, stirrups may be used for elements with a type I, II, or III cross section as long as the element is designed to attain small deflections. An exception "to this deflection criteria is for the particular case of slabs subjected 'to 'tension membr';ne"action. He:re" the slab can attain large deflections. Lacing may be used for type II and III,crosssecti~ns designed for either small 'or large deflections. It wo,.ld.be grossly unwarranted to use lacing for a'type I: cross section. ", ' The required area of stirrups is calculat~d from:
"
.
, A.; .: [(vu '. vc>,b'~s'slh f ds
4-26
while the required area of lacing reinforcement
dna
{s: 4-27
+ cosc )
where: ,
,,'
"
~
total, area of s t Lrrups or,: 'lacing r'e Lnfo rceraent; in tension: within a width' bsor b l,' and a "dfstance Ss 'or si .excess shear stress
wi'dth' of concrete strip in which resisted by, stirrups of area
Av
t~e:.diagonaf,t.ens Lon. st~e.~~es
are'
width of concrete strip in:whic~. ,the, df.agoual, te'nsion,st'resses are resIs ted- by lacing of area P..vss -
spacing of stIrrups in the direction,paranel ,rein,forcement ,', .
t~'
the longitudinal
spacing of lacing in the direction parallel to the longitudinal reinforcement " "' capacity reduction 'factor, equal.to O.8~ , ,
'
4-37"
TK
5~1300/NAVFAC
,a
-"1
pc 397/AFR ' 88-22: t~e
angle f.orme,d by the plane, of th,e lacing, and reinforcement .
:~oI)gitudinal ". '.
.
;.
.,
.". ~
-2B(1-B) ±
.
),
_
4
~
i [2B(1-B)]2- 4[(1-B)2+
A2][B2_ ';'2Jil/2
<
,:
'j
B
<
.
'j
y
"
" " :~ I
.
.t:,
_.;
".
j-' .'
P
..."
~.
t.
...
:
4,-29b
'.
..
,
1.
I,
"
s,
,
""
where
" 4-29a
":4. '.'
; I
~
. ,d l
.'
,-
-s.
.
,
4-28
in which
,~
th~
A2 ]
2[(1-B)2 +
~.,
plane of "
~ 'v
_'
between centerlines of , lacing bends measured normal '.. ',cf1,,,iural 'reinforcemept '" • ' distan~e
.
,.,.J-
•
, "ra'drus '1,f bend in lacing bars (min i~ nominal diameter of
.... baF
rei~for~ing
.
'fdb) :
..~
'1'
:
"
,
A typical section of a lacing bar illustrating the terms used. in the above I. . equation, is shown in,Figure 4~15. To facilitate the design ,of lacing'bars, the angle a can be determined from Figure 4~15, .~ ~'1'"'l ! :) -
,,'- -
•
.l
4-18.4. Minimum Shear Reinforcement
,.
.-.
.
In order to develop the full flexural capacity of a, slab, a premature shear failure must be prevented. Shear reinforcement must be provided to resist shear stresses in excess of the capacity of the concrete. However, except for slabs having a type I cross section' or subjected to tension-membrane action in the fa~ design range, minimum she~r ~einforceme~t must always be provided to insure' the full development'of the flexural reinforcement arid enable the slab to attain large deflections. , , " . "." , Stirrups or lacing must conform to the following li~itatio;s to insure a proper distribution of shear reinforcement throug~out the element and, in specific'case;;; fo provide a minimum quantity ... , of shear'reinforcement:" . . ". 1.
'f
2'~
,'Th'e. minimum design '!tress (excess shear stress V u .- v c) used to , caLcuLace the required amount 'of shear reinforcement, must conform to the limitations of Table 4-4." " When' stirrups or lac'in'g reinforcem';nt is" required, the, area ~ should not be less than 0.0015 b s Ss for stirrups or' 0'.0015 b l -i , for lacing. t. . . .
3.
,
~
j
..
!
',.1.
• •
When stirrups or lacing are provided, the required area ~ is determined at the critical section and this quantity of reinforcement must be unifor~lY"di~tr'ibuted throughout, the element'.
"
TM 5-l300/NAVFAC
4.
P~397/AFR
88-22·
Single leg stirrups should be used for slabs. At least one stirrup must be, located at .aach- bar. interse,ction. .. -.'
5.
A lacing bar is bent f'r om a single; .reInforcing bar. In one direction, the lacing must be continuous across the slab between opposite supports. In all cases, the lacing must be carried past the face of the support and securely anchored within the support.
6.
" The maximum spacing of stirrups Ss is limited to d/2 for Type I cr'os sc.s ec t Lons and dc/2 for Type II and III sections, .but not· greater than 24 inches.
,
7,
,
" The maximum· spacing of lacing sl is limited to d c or 24 inches, whichever is smaller, . - , ." '"
The spacing of stirrups. arid' lacing is 'a' function .o f the -f Lexura L bar spacing. Consequently, the above limitations .for, shear ,reinforcement should be considered in selecting the flexural bar spacing. Once selected, the flexural bar spacing may have to be altered to suit the above limitations'. "
.
4-19, Direct Shear Capacity
.
4-19.1, General. ." :.
,
..
:.; J
,,. '
',
'
.'
.
'
'., '"
Direct shear failure of a member is characterized by the rapid propagation of: a vertical crack through the depth of the member, ThIs crack is usually located at the supports where the maximum shear stres sescoqcur . .Failure of this type is possible even in members reinforced for diagonal tension.. Diagonal bars are required at slab supports to prevent: direct shear failure: when the design support rotation exceeds 2'(unless che.vs Lab is simply \, supported as given in section 4-19.2); when the desi~~ support rotation is ~ 2' but the' direct' shear capacity of the concrete -Ls insufficient; ','or when the section is in tension (as in containment cells), Diagonal.reinforcement consists of inclined bars which extend from the support into the slab element. 4-19.2, Direct Shear Capacity of Concrete If the design support rotation, e, is greater than '2! ;(e > '2!)', or if a section (with any support rotation) is in net tension, then the ultimate direct shear capacity of the concrete, Vdl -is zero and diagonal bars are required to take all direct shear. If the design support rotation, e, is less than or equal to 2' (e ~ 2'), or if the section; with any rotation ,e, is . simply suppo r t.ed (total moment capacity .: of adjoining 'elements at the support ritust be significantly ·less thari the moment capacity of 'the section' being 'checked for 'direct 'shear), tnen the ultimate direct shear force, Vd, that carr-be r e s'Lst.ed by ,the"concrete in a' slab is: .,
:
~,
"
-.: 4-'39'
'
.
.
,
4-30
"t.:
TH 5-1300/NAVFAC P-397/AFR 88-22 4-19.3. Design of Diagonal Bars The required area of diagonal bars ,is determined from:' 4-31 where:
'
Vd or
Vd
and
Ad Vs 'Q
0.18 f'dc bd.
o
(9
..
~
~
2' or simple supports),
(9) 2' or section.in tension).
total area of diagonal bars at the support within a width b '.' shear at the support of unit width.b '. " angle formed by the plane' of .the' diagonal reinforcement and the longitudinal, reinforcement.' ~
.
• •f
4-20. Punching Shear 4-20.1. Ultimate Punching Shear Stress When a flat slab is supported on a column or a column rests on a two-way slab, failure occurs around and against the concentrated load, punching out a pyramid of concrete from the slab. The ultimate shear stress v u' as a measure of punching shear, is computed from: ~ ",
4-32,
,':
where: ,Nu
the total concentrated axial load or. reaction distan~e
bo
,failure perimeter .located at a load or reaction area
d/2 from, the concentrated
de -
,either d or de depending on the type of cross section
4-20.2. Punching Shear, Capacity of Concrete The shear stress permitted for punching shear is, limited to: 4-33 Equation 4-33 applies to circular col~ns and tO,rectangular col~~s with aspect ratios no·greater than 2, For rec~angular areas with aspect,ratios greater than. 2 , the allowable value of V c .shcul.d be reduced.,according.to the ACI provisions (not I'Ls t.ed , in thi~ text). Shear reinforcement is not ,permitted to increase the punching shear capacity of a slab. If the ultimate shear stress V u is' greater than .the .s t res s permitted for punching shear v c' the slab thickness must be increased., In flat slab design, the use of a drop panel to increase slab thickness, and/or a 4-40
TH 5-1300/NAVFAC P-397/AFR,88-22
column capital to increase failure perimeter.' may be employed to prevent punching shear failure, If a'drop pane'l, -Ls used; punching shear mus trbe checked at the perimeter of the drop panel, as' well as at the top of the 1 column .... . r.
4-21, Development of
Reinfoicem~nt I.' •
.
4-21.1, General' "
.
,
In order ·to 'fully develop the' 'flexural and/or' axial load 'capacity of a concrete slab or .wall. the full strength of the reinforcement must be realized,"At'.'ariy section along the length' of 'a 'member, the tensile or compr e s s Lve force in the reinforcement must be developed on each side of the section by proper embedment length. splices (lapped or mechanical). end anchorage,' or' for tension only, hooks. At a point of peak stress. this development length or anchorage is nece~sary on both sides of the point; on one side to transfer stress into and on the other side to transfer stress out of the reinforcing bu,
" ;,...
."', -
.
.
'.'
I',
.
,
'
The tYpes.and locations of reinforcement anchorages are severely restricted for blast;resistant structures', These restrictions are necessary to insure that the'·"stiu~t,jre acts' in a duc t.LLe manne'r .: Typical details' for 'both. .
i'.'.",· .-,.....
.:
. . laced '. reiriforced
convention~11yreinfoicedand
~oncrete
.
.'
..
elements are given in'
latter sections dealing with construction detaiis~ Conformance to ~hese' details greatly simplifies the 'calculation of' development lengths:' :The required development lengths to be used in conjunction with the 'required typical'details are 'given below, • •... . '. . .
.-:
-I
I
•
"
.,
.
0"
"
-, .
. - ,
\
~
.
While conformance to the typical details given is mandatory. certain coridit Lonsjaay "preclude' their use, For these"unusual conditions .. the required anchorages'~':e calculated according to the 'procedures g~ven in the1CI Building Code, The basic development length is first calculated and then modif~ed ?ased O? the constructi,on detail~ employed to obtain,the required end anchorage or. splice length: This procedure is out.Lf.ned in Sections 4-21.4 through ~-21:7, Ho~~v~r, it must be'~epeated that the typi~al details given must. _be. foIlciwed and any' ,. ' deviation , ' . from ' - these. restrictions r ., and requirements must be carefully 'considered to insure pioper structural'behavior. ~.
-
•
4-21. 2
.
•
',I
I
-
-
.'
'.
_.'
'Provisions for Conventionally Reinforced Concrete Elements J •
•
'
I'
~
'.'
~'.,.
.
... ~
.
I _
.
Typical d~tails for conventionally reinforced (non-laced) concrete elements are given,i~ subsequent sections, These'details locate splices ~n reinfor~e ment at points of low stress, This permits the minimum length of lap splices, as wellc~s,th~ d~~elopmerit length for end anchorages, to be given by
..
. '
,
Id , 40 db
'.
4-34 .
~
where lE' .
.
Id .- development length "" i _ diameter of reinforCing bar db
. "I'
-.
. ..
The value of Id shall not be less than 24 inches, Equation 4-34 applies for end anchorage of #14 bars and smaller. and for lap splices of ~ll bars and smaller since lap splices of #14 bars are not permitted, 4-41 '
TH'S-1300/NAVFAC P-397/AFR 88-22 Lap splices of reinforcement must not be located at critical sections. Rather, ,they must be located in regions -of low stress (inflection points) where the .area of reinforcement provided is more ,than twice the area of .reinforcement required by analysis. In addition, not more than one-half of the reinforcement may be spliced at one location. The splice of adjacent bars must be staggered at least the required lap length of the bars sinc~ ove,lap of splices of adjacent bars is not desirable. Under these conditions, the required minimum length of lap splices is given in Equation 4-34. If it is impractical to locate splices at the inflection points, then the length of the splice must be .calculated according to the provisions ofSec.tion 4-21.4., Typical details for intersecting. walls and slab/wall ,intersections ,avoid the use of end anchorage' 'of the primary reinforcement:,. Rather, the reinforcement is .anchored by continuing it through the support and.bending it into the'. intersecting wall or slab. This reinforcement is then lap spliced with the reinforc~ment in the intersecting wal~. . .,'.
4-2l.'3
'Provisions for Laced Reinforced, Con~rete'Elements
Tests.of.laced elements have indicated that i~ continuity.of th~lla~\ngand flexural reinforcement is maintained throughout the element, the required development.Je?gth for end anchorage as ,:well '"s ..t,h.e minimUm i~ngth,: o(lap , . ' splices is given by Equation 4-34, but not Less than 24 inches .• Thi~ equation applies for end anchorage of #14 bars and ,smaller., and for lap lerigths :of #11 bars and smaller since Lap spli~es of #14 ·.bars. a re no c ' permi tted... ' . ':_ Required cons~~~~~ion details and procedures for laced'p;inforced c;ncret~ .' " .elements are given in subsequent sections. These details must be f'o l Lowedico insure the full development of both the concrete and reinforcement well into the range o f plastic action of t.he materials. , The use of Equation.4-'34 t,o obtain the required development lengths of the reinforcement is predicated on' the use of these d e t a i l s . ' .,. '., .... ,' The t~pical de t a LLs fo r laced r e Lnf'orced concret~;elements' i~qu'ire that' the reinforcement (flexural as well as lacing bars) must not be spliced,at . critical' sections but'rather must be spliced in regions ,of low stress ,(infle-," ction points) where the' area of reinforcemeritprovided is more than'twice the area of reinforcement required by analysis. 'In addition, not ~ore than one:' quarter of ~he' reinforcement may be sp~iced at on~.location. The ~plice of adjacent bars' must be staggered at least the required'lap length of the bars since overlap of. splices of adjacent,~aq is not, de s Lr'abLe , . Specific end ancno~age.details ~re r~quired for. laced reirifqrced cohcrete walls and slabs to enable the reinforcement to attain its ultimate strength. The preferred method of end anchorage is through ,the use of' wall extensions sinc~ this method presents the least cons~ruction problems. If architectural requirements do not permit the use of wall extensions, the reinforcement is anchored by continuing it through the support and bending it into the intersecting wall or slab~ In this latter case, the reinforcement is developed by a combination of anchorage and lap splic~ . . In either case, the lacing .extending into the supports provides the necessary co~finem~nt which' pe,mits'the use of Equation 4~34." "
4-42
til 5-l300;lNAVFAC' P,-397;lAFR 88-22 4-21. 4. Development' Length for Reinforcement in Tension The basic development length for //11 b~rs' and ~maller 'which are ,in' tension is given by:
The basic development length for //14 bars in tensl'on is given by: "
ld The
~se
- 0.085
f ds/;I(f' ' de ) 1;12
of f/18 bars is not
~ermnted
4-36 ,"
by this ma';ulii.
"
"•
.I~
.1. •
'':''
For top rei~forcement,' is, horiz·o·ntal J reinforceme~t: 'so.~ pla'c~d that ni~re than 12 inehes of eoncre~e is cast in the slab below the reinforcement, the
t;hat
basic development Leng th must 'be' multiplied by 1.40. This p r ov Ls Lon applies, to ho r Lzorrta L slabs only. Walls ,with mul t.LpLe runs o f iho r Lzorrt.aL bars plus vertical bars are not e f f'ec t ed 'by this proyision. ,.',In addition, the basie development length of all bars may be multiplied by 0.80 if the bars are' spaced laterally at least six (6) inches on center. In no case shall the development length be less than 24 inches nor 40 tim~s the di~m~ter ,of'tne reinforcing, bar. 4-21.5
Development 'for Reinforcement in Compression . ' Length .' . .~
.
'
;
The basic development length,for bars in compression is 'given by: 4-37a but not less than:
.
~
,,'
4-37b The development length for compression is not modified for top bars nor lateral bar spacing. In no case:shall' the dev,:lopment length , used be less rhan F",elve,'<12) inehes., :' Under dynamic load conditions, members are subject to load reversal or rebound, Reinforcement ,subject to comp~~ssiv~ forces unde"the primary load. may be subject to tensile forces ,under rebound. Consideration must be given to this st~ess. r~versal: sinc~ the d~v~lopme~t,~e~g~h of.bar~ :ih co~pression. is less than the 'development 'length 'of bars in tension:'" ,,, 4-43 .
T!I,'5-1300/NAVFAC P-397/AFR BB~22
•
4-21.6. Development Length of Hooked Bars
"I
The basic development length for .bars in, tension which terminate in a standard gO-degree or lBO-degree hook is given by: ldh - 0: 02 db fds/(f' de) 1/2 "
4-38 ...
j
-.'
•
but not'less than 8d b or 6 inches, whichever is greater. The development length for a hooked bar ldh is measured from the critical section·al~ni the' length of the bar to the end of the hook. That is, the length ldh includes the straight length of the bar between the critical section and point of tangency of the hook, the bend radius and one bar diameter. The required hook geometry as specified in the ACI Building Code is given in Figure 4-16. In the development of hooked bars, no'distinction is made between top bars and other bars. However, since hooked bars are .especially susceptible to concrete splitting failure if the concrete cover is small, 'the above' equation takes into account the effect of minimum concrete cover. For #11 bar and smaller with the cover not less than 2-~ inches, ind for a gO-degree hook, with cover on the bar extension beyond the hook not less than 2 inches, the development length Idh may b~ muLt tpl.Led 'by 0.7. .' . , :". ' Hooks are not to be considered effective in developfng bars in compression. However, in the design of membe;s subjected to dynamic loads, rebound or load reversal must be considered. That is, under the primary loading, reinforcement is subjected to tensile forces and anchored utilizing a standard hook, but this same hooked 'reinforcement may be s~bjected to compre~sive forces unde r rebound. Therefore, the .straight portion of the hooked bar must be sufficient to develop this compressive force. For'th;se ca'ses where· 100 percent ~ebound is e~~o~ntered, the straight portion of a'hooked bar "must be equal to, the d",velopment length for bars in compression,. .
,
4-21.7. Lap Splices.of Reinforcement , '
ln blast resistant structures, reinforcing bars may be lap-spliced using only contact lap splices; noncontact lap splices are.not permitted. Lap splices shall not be used for reinforcing bars larger than #11 bars. If #14 bars are used, they must be co')tinuous -. Lap splices of adjacent parallel reinforcing bars must be staggered by at least the length of the lap, The minimum length 'of lap for tension lap splices depends upon the location of the splice. For blast resistant structures, it is strongly recommended that splices be located in regions of low stress, where the area of reinforcement provided is at least twice that required. In such cases, the length of the lap is equal to the basic development length ld for bars in tension as given by Equation 4-35 or 4-36 and modified, if applicable, for top bars and%orlateral spacing of the bars. In other design situations where the lap splice is not located in regions of low stress, the lap length is equal to 1.3 times the modified 'development length, ld' ) .... ) ." " The minimum'length of lap for·compression lap spiices'is equal to the basic· development length ld for bars in compr'e s s Lontas given by Equation 4;37. However;· due to the occurrence of load'revetsal, it is recommended' that the .,....
4-44
.
~',
,
,
TN 5-l300/NAVFAC P-397/AFR 88-22 , length of lap splices be based in tension unl~ss it can be shown that the reinforcement will a~ways be in compressi~n~ 4-21.8. Mechanical Splices of 'Reinforcement Mechanical devices may be·used for end anchorage and splices in reinforcement. These devices must be capable of developing the· ultimate dynamic tensile strength of the reinforcement without reducing its ductility. Tests showing the adequacy of such devices under dynamic conditions must be performed before these devices are deemed acceptable for us~ in hardened structures. 4-2l~9.
Welding of Reinforcement
Welding of reinforcement is to be avoided in blast resistant structures since it results in a reduction of the ultimate 'strength and ductility of the reinforcing steel. In those cases where welding is.absolutely essential, it may be necessary to obtain special reinforcement manufactured with controlled chemical properties. Tests showing the adequacy of the combination of weld and reinforcing steel under dynamic conditions must be ,performed to demonstrate that this weldirig does .not reduce the'ulti~ate strength and ductility of the reinforcing steel. In lieu of these tests,. welding is permitted if the stress in the reinforcement is maintained at a level less .thim 90 percent of the yield stress. 4-21.10. Bundled Reinforcing Bars The use of bundled bars may be required for unusual conditions. However., their use is not desirable and should be avoided where possible. -A 3 bar bundle should be the maximum bundle employed. The development length and lap splice length of ' individual' bars within:a'bundle shall be that of the individual bar increased 20 percent for a 2 bar bundle and 33'percent for a 3
bar bundle. In addition, splices of the individual bars within a bundle should be staggered .. That is, only one bar of the bundle should be spliced at a given location.
4-45
A'
d'
.Sl··
-
,
I-
d Te I I I
I I I I I I
.,
L As OR .NO CRUSHING ,
.,
SPALLING
.'
:, ,
-",
'\
TYPE I :'
. ",
'"
'.
.>
~s --As
'. f---
- 7
d'
-
.
,
.,
tr I"
I
'.
~.
de Te I
I' I I I
I
'LA s ..
i "
CRUSHING
TYPE
.
\.
n
-l-
----'~
".'
--
d'
,
", .
,."
'.
de Te
--t-· -.---As
SPALLING
TYPE
m LEGEND: III I I!! CRACKING - - - CRUSHING - - - - DISENGAGEMENT ;
, t .
r.: Figure 4-13
Typical reinforced concrete cross sections
4-46
dOl' de Vu
POTENTIAL DIAGONAL . TENSION CRACK SUPPORT
a. CRITICAL
SECTION AT d or de FROM FACE OF SUPPORT.
POTENTIAL DIAGONAL TEl'SION CRACK
SUPPORT
Vu
b, CRITICAL SECTION AT FACE OF SUPPORT
Figure 4-14
Lncatipn of critical sections for d.iagonal tension
4-47
"
..
'
90
""s»:,
80
a>
• 70 Ul 1LI 1LI
a:
60
a
50
CO 1LI
I 11: ! I ' , , , I ' ,,,, 0.70 "" I II ill i i II' II . , , , iii,lIilllllllllllI11111i"'''1 i i i i Frlill I
I
I
I
,
,
I
I
I
,
,
,
,
,
, '"
I
I
I
1
I
I
I
I
I
,
'L'
'lQ.8O I i I . ' , , , , , , , , , , , , , , , , , "0.90 1.00
'l{
o
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
(2R t +dbl dt
Figure 4-15
e
Angle of inclination of lacing bars
e
e
db
I
A
.. .Critical sectian
12db db'
#3 thraugh 2)12' min.
#8
#9..#\0 and #11 #14
Figure 4-16
Geometry for standard hooked reinforcement
e· 4·49
TM
5~1300/NAVFAC
P-397/AFR 88-22
Table 4-3 Cross-Section
Minimum. Area of Flexural Reinforcement
Reinforcement Main' Direction
As - O.OO15bd A' s - O.OOlObd
Secondary , Direction
As - O.OOlObd A' s - O.OOlObd
Type I
,
Type II and Type III
One-Yay Slabs
Main Dir'ection
, ..
Secondary Direction'
Two-Yay Slabs
. I
As - A' s .- O.OO15bdc. 'As - A' s -
O.OO15b~c*
As - O.OO15bd A's - O.OOlObd As - O.OO15bd A' s - O.OOlObd As - A's- O.OO15bdc
As - A's- O.OO15bdc*
* but not less than As/4 used in the main direction
4-50
TH 5"1300/NAVFACP-397/AFR,88-22
,
-,
-,
'Table 4-4
Design Rang~
FAR
.
Excess Shear Type, of ,Type of , ,Stress Vu - V c Type of Structural Shear Reinforcement vuSvc vc
Flexure
Stirrups
Type 'II
Flexure
Stirrups
o:8S:;'c
Type II & Type III
Tension Membrane
Stirrups
0
Type II & Type III
Flexure or Tension Membrane
. CLOSEIN
Minimum Design Shear Stresses for Slabs
0
- Vc
Vu
- Vc
O,8Sv c
Vu
- Vc
- Vc
Vu
v' - ,c
Vu
'vu
.
, Stirrups or Lacing
O,8,Svc
.
0: 8Svc
p... ;'. 'J
v,;,
"
"
"
.' ....
4-51
.
,
,
Vc
TK 5-l300/NAVFAC P-397/AFR 88-22 DESIGN OF NON-LACED REINFORCED SLABS
4-22. Introduction Conventional reinforced concrete elements are for the purpose of this manual, members without lacing. These non-laced elements make up the bulk of protective concrete construction. They are generally used to withstand the blast and fragment effects associated with the far design range but, may also be designed to resist the effects associated with the close-in design range. In the latter case. the distance between the center of the explosive charge and the element must not be less than that corresponding to a scaled distance Z equal to 1.0. Laced elements are required for scaled distances less than 1.0. Non-laced elements may be designed to attain small or large deflections depending upon the protection requirements of' the acceptor system. A non-laced element designed for far range effects may attain deflections corresponding to support rotations up to 2 degrees under flexural action. Single leg stirrups are not required to attain this deflection. However, shear reinforcement is required if the shear capacity of the concrete is not sufficient tO,develop the ultimate flexural strength. A type I cross-section provides the ultimate moment capacity.' The flexural action of a non-laced element may be increased to 4 degrees support rotation if 'single leg stirrups are provided to restrain' the .compre s s Lon reinforcement. "In this deflection range, a .type II cross~section provides the . ultimate moment and mass . , capacity . to resist motion. Single leg stirrups must be provided when a non"laced element is designed to resist close-in effects. The shear reinforcement must be provided to prevent local punching shear failure. When the explosive charge is located at scaled distances less than 1. 0, a laced rather than a non-Laced element mu'st b~ employed. For scaled distances greater than 1.0 but less than 3.0, single leg stirrups must be provided, whil~ for scaled distances greater than 3:0,shear reinforcement should be used only,if required by analysis.' With single leg. stirrups, the member may attain deflections corresponding to support rotati~ns up to 4 degrees under flexural action. A type I and II c ros svaec t.Lon provides the ultimate moment capacity and mass to resist motion for elements designed for 2 and 4 degrees support rotation, respectively. If spalling occurs,then a type III cross-section would be available.' In addition, a non-laced element designed for small deflections in the close-in design range is not reusable and, therefore, cannot sustain multiple incidents. A non-laced reinforced element may be designed to attain large deflections, that is, deflections corresponding to incipient failure. ,These increased deflections are possible only if the element, has sufficient lateral restraint to develop in-plane forces. The element may be designed for both the close-in and far design range. A type III cross-section provides the ultimate moment capacity and mass to resist motion. The design of non-laced reinforced elemerits subjected to a dynamic load involves an Lt.erat.Lve (trial and error) design procedure. An element is assumed and its adequacy is verified through a dynamic analysis. ,The basic data required to determine the ultimate strength of the reinforced concrete section has been presented in previous sections. Procedures to determine the resistance-deflection function used for design, the dynamic properties of the
4-52
TH 5-1300/NAVFAC P-39.7/AFR 88-22
section, and the'dynamic analysis required to determine an element's response is presented in Chapter 3. This section contains additional data needed to establish the resistance-deflection curve for design as well as procedures to design the element for shear. The interrelationship between'the various parameters involved in the design of non-laced elements is readily described with the use of the idealized'resistance-deflection curve shown, in Figure 4-17. 4-23. Distribution of Flexural 'Reinforcement 4-23,1. General A prime factor in the design of any facility is cons eruccf.on economy. Construction costs are divided' between labor; and material costs. Labor cost is further di~ided into shop and field work, with field labor being generally the more costly. Labor cost can account for as mush as 70X of the cost of blast resistant reinforced concrete. Proper selection of concrete thickness and reinforcement steel will result in a design which optimizes .the structural resistance and minimizes required construction materials. Subsequent sections will 'discuss procedures to optimize the required materials. The designer must then evaluate the issue of constructability. Factors such as standardization of rebar
slzes~
spacing, congestion, ·steel erection and concrete placement
difficulties, These considerations may require that the initial design, optimized for material quantities, be modified.' Such a modification may actually increase the cost of materials while reducing the overall construction cost by reducing labor intensive' activities. In addition, improved constructability greatly reduces the risk of quality control problems during construction, 4-23,2. optimum Reinforcement, Distribution Fora given total amount of flexural ,reinforcement and a given concrete thickness, 'the dynamic capacity of an element varies with the amount of reinforcement placed in the vertical'direction to the amount in the horizontal direction, For a given support condition and aspect ratio L/H, the ideal distribution of the reinforcement'will result in the most efficient use of the reinforcement by producing the greatest blast capacity . .
.".
The optimum distribution of the reinforcement for a two-way element is that distribution which'results in'positive yield'line~ that bisect the 90 degree angle at the corners of the element (45 degree yield lines). For two-way elements there are numerous combinations of support con,ditions with various' moment capacities due either to quantity ofreiriforcement provided or degree of edge restraint as well as various positive moment capacities 'again due to variations in the quantity of reinforcement provided, Due to these variations in possible moment' capacities, the ratio of the'vertica,l to horizontal ' reinforcement cannot be expressed as 'a function of the aspect ratio L/H for different support conditions. Figures 3-'4 .through 3-20 'must be used to determine the moment 'capacities which will result in a "45'degree yield line", The reinforcement ,can then be selected from these moment capacities. In some 'design case;, it may not be possible to furnish the optimumdistribu- ' tion of reinforcement in aparticular'element. One such case would be where the optimum distribution violates the maximum' or minimum ratio of the vertical 4-53
TN 5~l300/NAVFAC P-397/AFR 88-22 to horizontal reinforcement AsV/As H of 4.0. and 0.25, respectively.. A second situation would arise when the optimum distribution requires less than minimum reinforcement in one direction.' The most common case would result from the structural configuration of the building in which support moments may not be fully developed (restrained support rather than fully fixed) or from the need of maintaining continuous reinforcement from adjacent elements; In these and other situations where. the optimum distribution of the reinforcement cannot be provided, the reinforcement should be furnished to give a distribution as close as the situation allows to the optimum distribution to maintain an economical design. ", J 4-23.3. Optimum Total Percentage of Reinforcement The relationship between ther; quantity. of reinforcement to the .quantity of. concrete which results in the minimum cost of an element may be:expressed as a total percentage of reinforcement. This·total percentage of reinforcement PT is defined as
,
4-39
'Pr .,. Pv + PH
where Pv and PH are the average percentages of reinforcement on·one face· of the element in the vertical and horizontal .directions, respectively. Based ~n the average costs of concrete· and steel, the optimum percentage for ~ nonlaced reinforced element using single leg stirrups ·has been determined to be between 0.6 and 0.8 percent with 0.7 a reasonable design value. For elements wh~ch do not contain shear reinforcement, the' optimum percentage would be s~m~what higher. For large projects, a detailed cost analysis should be performed. to obtain a more economical design. In some design cases, it maybe desirable to reduce the concrete thickness below low the optimum thickness. A smal L. increase in, cost (10 percent) wo,,:ld be incurred by increasing the value of Pr to one percent. Beyond one percent, the cost increase would be more rapid.- .However, except; fot; very thin elements, it may be impractical t~ furnish 'such. large quantities of reinforcement. In fact, in thick walls it may be Jmpractical to even furnish the optimum percentage. ,
.•
• t
Unless single leg stirrups are required for other than shear capa~ity such as . for close-in effects or to extend fle~ural action in the f~r design range from 2 to 4 degrees support rotation, it is more economical to design non-laced elements without she.ar reinforcement. In.thi';1 case, the' total percentage of reinforcement must be limited so th.atthe ultim.at~.resistance of th~ element does not produce shear stresses in excess of.,.the concrete cepac Lty .. . . ;
.~.
.~
1 .' , ::"
~
4-24. Flexural Design for Small
.»
Deflec~~ons
:
:,
.(
..,
. .
,
.
.'
The design range for small deflections maybe divided into two regions; elements ,with support.rotations less'than'2 degrees (limited deflections) and support ~otations between 2 and 4 ~degrees: .Except for stirr~p.requirements and the type of cross-section available to. resist moment, .the design procedtire is the same. . . ' " In the flexural. design of a non-laced reinforced concrete slab, the optimum distribution of. the flexural reinforcement must first be determined. A 45 degree yield line pattern is assumed and, based on the support 'conditions and 4-54
TH 5-l300/NAVFAC;P-397/AFR 88-22
aspect ratio; the ratio of the vertical to horizontal moment capacities are determined from the yield ,line location figures of Chapter 3. Reinforcing bars and a concrete thickness are next chosen such that the distribution of reinforcement is as close as possible to that determined above and such that the total reinforcement ratio PT is approxImately 0.7 for elements ,utilizing,'stirrups. For those elements not utilizing shearreinfor- " cement PT is minimized.so that the shear capacity "fthe concrete ~s'not exceeded.' '.Using the, equations of preyious sections (eq. ,4·11 for, type I cross sections, eq , 4-19 for ,type II or III cross -aec t Lons ) rhe moment capacities are computed. 'The moment capacities are required to calc:ulate the ultimate unit ze s Lst.ance r u and the equivalent elastic deflection X E . These parameters along with the·natural period of vibration'TN define the equivalent singledegree-of-freedom system of the slab, and are discussed in detail'in Chapter 3.
..' '
...
A dynamic analysis (see Section~4-26)'is then performed to check that the slab meets the response' c'r Lt e r La . Lastly, thi":shear capacity is' checked (Section 4-27). If the slaD does not meet the response criteria or fails in shear (or is greatly overdesigned) a new concrete section is assUID"d and the entire design procedure is repeated. \. '
3.
-
_.
t
'. , •
4-25. Design for Large Deflec'i:ion.s "
-.
,
,t.
r
•f
4-25.1. Introduction
Design of non-laced reinforced concrete elements without shear reinforcement (single leg stirrups) for support rotations greater than 2 'degrees' or elements with single leg stirrups for support rotations greater than 4 degrees depends on their capacity to act as a tensile membrane. Lateral restraint of the ~lement must be~provided to achieve this action. Thus, if lateral restraint does not'exist, tensile membrane action is n9t developed and th~ element reaches incipient f'a'l Lur'e at 2' degrees (4 'degrees' if adequate singte leg' stirrups are provided) support rotation, How~ver': if lateral restraint exists, deflection of the element induces membrarieaction and Ln-p Lane forces. These in-plane f'orc e s provide 'the means for the' element co continue 'to develop substantial re's'istance up to maxim\IID support' 'r"hitions of approximately 12 degrees. .' j ' .. 4-25.2. Lateral Restraint Adequate' later~l ie~traint' of the reinfor~e~ent is mandatory in order for the element to develop 'and the' designer to utilize the benefi~s of tensile membrane behavtor" '-Sufficient lateral restraint 'is prov Lded 'if 'the reinforcement is adequately anchored into adjacent supporting members capable of resisting the lateral force induced by tensile membrane action. Tensile membrane behavior should not be considered in the design'process unless full external lateral restraint is provided in che span directions shown in Table 4-5. Full external lateral restraint means that adjacent members can effectively resist a total lateral force equivalent to the ultimate strength of all continuous reinforcement in the element passing the boundary identified by the arrows in Table 4-5. External lateral restraint is not required for elements supported on four edges provided the aspect ratio L/H is not less than one-half nor greater than 2. Within this range of L/H, 4-55' '
TN 5-l300/NAVFAC P-397/AFR 88-22 the inherent lateral restraint provided by the element's own compression ring around its boundary is sufficient lateral restraint to develop tensile membrane behavior. 4-25.3. Resistance-Deflection Curve A typical resistance-deflection curve for laterally restrained elements,is shown in Figure 4-18. The initial portion of the curve is due primarily to flexural action. If the lateral restraint prevents small motions" in-plane compressive forces are developed. The increased capacity due to these. forces is neglected and is not shown in Figure 4-18. The ultimate flexural resistance is ,maintained until 2 degrees support rotation is produced. At this support rotation, the concrete begins to crush and the element ,loses flexural capacity. If adequate single leg stirrup~.were provided, ,the flexural action would be extended to 4 degrees. However, due to the presence of continuous reinforcement and adequate lateral restraint, tensile membrane action is developed. The resistance due to this action'increases with increasing deflection up to incipient failure at approximately 12 degrees support rotation. The tensile membrane resistance is shown as the dashed line in Figure 4-18. In order to simplify the design calculations, the resistance is assumed to be' due to flexural action throughout the entire range of behavior. To'approxi~ mate the energy absorbed under the actual resistance-deflection curve, the deflection of the idealized curve is'limited to,8 degrees support rotation. Design for this maximum deflection would produce incipient failure conditions. Using this ~quivalent design curve, deflections between 2 degrees (or 4 degrees if single leg stirrups are provided) and incipient failure cannot be accurately predicted. For the design of a non-laced laterally restrained'. element for ,8 degrees support rotation, a type III cross-section is used to compute the ultimate moment capacity of the section as well as to provide the mass to resist motion. The .stress in the reinforcement f ds would be equal to that cor~ responding to. support, rotations 5 ~ ,8m .~ l? glven in. Table 4-2 -. '. At .every. section throughout the element, the tension and compression reinforcemen~ must be continuous in the restrained direction(s) in order to develop the tensile membrane action which is discussed in d~tail below. 4-25.4. Ultimate Tensile Membrane Capacity As can be ,seen·in Figure 4-.18, the tensile membrane resistance of an e Lemerit; is a function of the element's deflection. It is also a function of the span length and the amount' of continuous reinforcement. The tensile membrane resistance, rT of a ,laterally restr!'ineci element a't a, deflection,. X, 'is expressed as: For one-way elements 8Ty [--]
"
Ly2
4-56
4-40
TIl 5-1300/NAVFAC P-397/A.FR88-22
For two-way elements 4-41 1
1
_(_1)(n-1)/2 ,
,
n 3, ; "
4 :E
1 -
"
cosh' [,
n-1,3',5
,
n7TL~ 2r..y'
T
1/2
:'[ r:l]
in which
and "
;
where rT
tensile membrane resistance
X
deflection of element
ly
clear span in short direction
Lx
clear span in long direction< force in the continuous'reiriforcement in the'short direction ,
;
force in'thecontinuous reinforcement in the ,long direction "
,
corrt Lnuoirs 're"inforcem~~t in the short direction 'contirnlous re1nforcement,:"in the long direction '. .. Even though the capacity of ~. laterally restrained' ele~ent is based on flexural' action, 'adequate tensile membrarie ~apacity must be provided" That is, sufficient continuous' 'teinforcement 'must 'be"provided so that the" tensile membr~ne resistanc~ rr l:~~r~spon~ing't~:8d~gr~e~suppcirt rotatloj.,,:must be, greater than the flexural res'tstance r u . Th!!. def Lec t Lon is compu t ed as a' function of the yield 'line 'locations (shortest'sector length).' The force in 'the corrt Lnuous reinforcem~nt is calctil~ted using'the dyn':'mic design stress f ds corresponding to 8 degrees support rotation (Table 4-2), ' 4-25.5.1 Flexural Design ' J
•
• .
.
.
, '
•
.~.
~..,
1
j
Since ~he actual tens~le membrane resistance deflection curve is replaced with an equivale~t flexural curve, the'design of a non-laced ele~ent for'large deflecti6ns, is greatly, simplified', "The, design is, performed in a similar ,,' manner as for small deflections. However, sufficient continuous reinforcement must b~"provided to develop the" required t-;;nsile membra,;e resistance': Where 4-57
TH 5-1300/NAVFAC P-397iAFR 88-22, '"
external restrain is required, the support must withstand ,the lateral forces Ty and Tx as ,given in Equations 4-42, .and 4-1,3" respectively, 4-26. Dynamic Analysis 4-26.1. Design for Shock Load The dynami~ ,analysis of a slab is accomplished by first representing it 'as a single-degree-of-freedom system and then lfinding the response of that system
when subject to a blast l~~d. The equiv~lent ,single,deg~ee-~f-freedom system ,is defined·in terms of its ultimate resistance ru"equivalentelastic deflection XE and natural ,period of vibration TN' The ultimate unit resistance,is calculated 'from the equations of Chapter ,3 for the moment capacities determined according to the previous sections. The procedures and parameters necessary to 'obtain the equivalent elastic deflection and natural period of vibration can also be found in Chapter 3. For elements subjected to dead loads in the same direction as the blast loads, , ' (for example a roof or retaining wall exposed to an exterior explosion) the resi~tance available to withstand, the blast load is reduced, An ~pproximation of the resistance available is 4-44
. where dynamic resistance available
•
uniform dead load , Chap!'er 2 describes procedures for determining the dynamic load which i~ ,defined by its peak value P and duration T. For the ratios P/r u and T/T N, the ductility, ratio ~/XE and tmlT can be .obtafned ~r()m J:he respopse charts of Chapter,3. These values, ~ which'is the maximum deflection, and t m; the time to reach the maximum ,deflection, define ,the dynamic .. . ,' . --. response,of ,. .. the element. The effective mas~ ,and the stiffness used in computing the natural p~riod of vibration TN depends on the type of cross'sJ!ction and loaamassfactor used" both of which ~~pend on the range of the maximum deflection. When the. , deflections are small (less than 4 degrees) a type I or type II cross section, is used". The mass is calculated u~ing the eritire'tllickness of the conc re ce ' ~' element Tc': The spalling that,occurs when an ,element a~t~ under tensile ' membrane behavior" or which may occur ,"due to ,close - in, effects requires the'use of a type III cross section to resist moment." Since: the concrete cover ,Qv!'r the flexural reinforcement is completely disengaged, the mass' is calculated based on the distance between the .centrroIds 'of the, compression' and tension" reinforcement'. " . '..When designing for completely elastic behavior, the elastic stiffness is used while, in ot~er cases, the ,equivalj'nt elasto~plastic stiffness KE is used, The elastic value of the load-mass factor KLM is used for the elastic range while, in the e Las t.o -p Las t Lc range', the load-mass factor~s the average of'the elastic,and elasto-plastic values. For small plastic deformations, the value of KLM is,equai to, the a~erag:,of. the, equi~~len~ elastic value a~~ th~,plastic 4-58
TK 5-1300/NAVFAC .P-397/AFR88."22 - (,
value. The plastic value of KLM is used for slabs designed for large plastic deformations. Due to the large .number of variables involved in the design of non-Tac ed.. " reinforced elements, design equations have .not; been devel.oped.. Howeve r , design ~quations have been developed for laced elements subJected to impulse loads and are presented in subsequent sections of this chapter. U~e of these procedures fo, the des~gn of:non-laced elements subje~ted.to,impul~e.load,will r e suLt in, a variety of errors depepd Ing upon support conditions, t.h i.ckne s s of the concrete section,' quantity and distribution .of . the flexural reinforcement. etc. However, these procedures may be used.to ,obtain a trial section which then may be analyzed as described above. 4-26.2. Design for Rebound
Elements must be designed to resist rebound, that is, the damped elastic or elasto-plastic harmonic motion which occurs after the maximum positive di5placern~nt ~
has been attained.
When an element reaches
Xm,
the resistance
is at a maximum, the velocity is zero, .and its deceleration is a maximum, The element,will vibrate about the blast load curve (positive and/or negative. phase) and/or the zero line (dead load' for roofs) depending on the time to reach maximum deflection t m and the duration of the blast load T. Usually only those elements with a type I cross-section will require additional reinforcement to resist rebound. Additional reinforcement is not required for type II and III cross-sections since these sections have equal. reinforcement on,~pposite faces and t~~ maximum possible rebound resistance is equal to the ultimate (positive) resistance. However, the supports for all types of cross sections • including the anchorage of the reinforcement (compression reinforcement under positive phase loading is subjected to tension forces under rebound conditions ) must be investigated for rebound (negative) reactions. . Also. it .should be noted that. the support conditions for rebound are not always ~he same as for the positive load. The negative resistance r-" attained by an element when subjected to a triangular pressure - time· load, is obtained . from figure 3 - 268 of Chapte r 3. Entering. the figure with the ratios of ~/XE.and T/TN, previously determined for the positive phase of.desig!l" the ratio of the required rebound, resistance to the ultimate resistance r-/r u is obtained .. The element must be reinforced to withstand this rebound resistance r- to,insure that· the slab will remain elastic during rebound. However, in some cases, negative plastic deformations are permissible. The tension reinforcement provided to withstand rebound forces is added to what is the compression zone during the initial loading phase. To obtain this reinforcement, the element is essentially designed for a negative load. equal . '.
to the calculated value of r-. However, in no case shall the rebound reinforcement be less than one-half of the positive. phase reinforcement., The moment capacities and the rebound resistance capacity are ca LcuLatied using the same e~uations previously presented. Note that while.dead load reduces the available resistance for the dynamic loading, this load increases the available resistance for rebound.
4-59
TM 5-1300/NAVFAC P-397/AFR 88-22 4-27. Design for Shear .'
4-27.1. General The ultimate shear V~ at any section of a flexural element is a function of its geometry, yield line location and unit resistance r. For one-way or twoway elements the ultimate shear is developed when the resistance reaches. the ultimate unit v~lue r u' In the design of 'a concrete element, there are two crit~cal locations' where shear must be considered . . The ultimate shear stress V u is calculated at a distance'd ordc from the supports to check the diagonal tension stress 'and to provide shear reinforcement (stirrups) is necessary. The direct shear force or the ultimate support shear Vs is calculated at the ,face of the support to determine the required quantity of diagonal bars, 4-27.2. Ultimate Shear Stress at d. from the Support 4-27.2.1. One-Way Elements The ultimate shear stresses V u at a distance de from the support are given in Table 4-6 for one-way elements. Depending upon the cross section type being considered, de can represent either d or d c'
For those cases where an element
does not reach its ultimate resistance r~ is replaced by the actual resistance r attained by the element. For those members whose loading causes tension in their supports, the ultimate shear stress is calculated at the face of the support. For those cases, the ultimate support shear V is calculated as explained in the next section. This shear is' then divided by the effective cross-sectional area (bd or bd c) of the element to obtain the ultimate shear stress.
4-27.2.2. Two-Way Elements For two-way elements the ultimate shear stress must'be calculated at each support. The shears acting at each section are calculated using the yield line procedure outlined in Chapter 3 for the determination of the ultimate resistance r u" Because of the higher stiffness at the corners, the shear along any section parallel to'the support varies. The full shear stress V acts along the' supports except in the co rner s where only 2/3 of the shear stress is used (Fig. 4-19). Since the shear is zero along the yield lines, the total shear at any section of the sector is equal to the resistance r u times the area between the section being considered and the positive yield lines . .
To illustrate this procedure, consider a two-way, element, fixed on three sides and free on the fourth, with the yield line pattern as shown in Figure 4-19; For the triangular sector I, the shear VdV and shear stress vuV at distance de from the, support is
...
2 3
L' Vdv ( 4
-
deL 2y
L "deL + --, .4 ·2y
- ru
4-60
)
(
L + Vd"v(-) 2 y-d e· 2
4-45a deL' ,
) (L
-
)
'y
TK 5-l300/SAVFAC P-397/AFR 88-22 3ru y (1 -'d~/Y )2
4-45b
(5 - 4 de/Y ).
and since the shear stress
V
u is equal to Vfbde and b equals one. inch
(1 - de/Y )2
3ru
4-46 Y
For the trapezoidal sector II 2 Y -V (dH
3
L
2
L
2
)
2
2d e y
1
r u ( - - de)
Y
) + VdH (H -
(H - Y + H -
2
)
4-47a
L
3r u(L - 2 de) [2H - Y -
2d ey L
] 4-47b
Vdh 2
3ru (1 -
[6H
2d e
- Y -
8d ey L y
) (2 -
L
] -
H
2d ey
--)
LH
4-48
v uH 2
(6 -
L
)
H
LH
Values of the ultimate shear stresses v uH and vuV at a distance d J from the· support for several two-way elements are given in Table 4-7. As stated above, de represents either d or dc' depending upon the type of cross section being considered. The ultimate shear stress is calculated at the face of the support for those members whose loading condition causes tension in their supports. For these cases, the ultimate support shear V is calculated as explained in the next section. This shear is then divided by the effective cross-sectional area (bd or bd c) of the element to obtain the ultimate shear stress v.
For· the situations where the ultimate resistance of an element is not attained, the maximum shear stress is less than the ultimate value. However, the distribution of the shear stresses is assumed to be the same and, there-
4-61
TN 5-l300/NAVFAC P-397/AFR 88-22 fore, the shear stresses can be calculated from the equation of Table 4-7 by replacing r u with the actual resistance attained (re• rep' etc.), 4-27.3. Ultimate Support Shear See Chapter 3,'for procedures used to calculate the ultimate shears of both one-way and two-way elements.
4-62
w
u
UL TrflATE
Z
(/)
W CI::::
X( 8 = 1')
X(
X( 8= 2')
e = 4')
X(8= 8')
X(8= 5')
DEFLECTION <]
ot" I
<]
2
Not
®
PROTECTION CATEGORY
@ &
G:
SHELTER PROTECTIVE STRUCTURE
BARRIER PRESSURE - TIME· DR IMPULSE
STRUCTURE-LOAO SENSITIVITY
FAR OR CLOS -IN DESIGN RANGE RESPONSE CHAR S, ITERATION PRoC DUR£, 1M ULSE (QUA TIONS DESIGN METHOD LIMITE D
LARGE
DEFLECTION CRITERIA FLEXURAL DUCTILE MODE OF RESPONSE FAR RANGE
FLEXURA,L ACTlON
CTloN
S IRRUPS IF !;>EOUIRED
STIRRUPS
"'EO'D
TENSION MEMBRANE ACTION
-
STIRRU S IF RE lJIRED
FLEXURAL ACTIO S IRRUPS REQUrRE D CLOSE-IN TENSION MEMBRANE
ACTION
STIR~ UPS REQl HED
BRITTILE MODE OF RESPONSE NONE
CRUSHING
SCABBING
FAR RANGE
'SPALLlNG
CLOSE-IN CRUS:HING
SCABBING
POST FAILURE FRAGMENTS
CROSS-SECTION TYPE
I
-
FAR RANGE I OR
I
CLOSE-IN
fely
::
-
II
III
II OR III
III
fely +l/4( ·~cfu- ely)
--
l/2(fely+fdu l
Nate 11 Nei-ther shear reinforceMent ·nor tensile r'leMbrane- action Note 21 Either shear reitlforceMent or +.ensite nenbr-cne action
Figure 4-17
Relo. i;ionship between deSign pc r-o.ne t er-s for unlaced eleMents
4-63
.,
r
r:
t
ACTUAL RESISTANCE DEFLECTION FUNCTION IDEALIZED RESISTANCE DEFLECTION FUNCTION
....,
1
,/
./
,/
./
CONCRETE CRUSHES, LOSS OF FLEXURAL RESISTANCE
-,
,
'--
_./
/'
/'
../
./
./
/'
./
.
Figure 4-18
Idealized resistance-deflection curve for large deflections
4-64
AT SUPPORT
AT de FROM SUPPORT
"'1'" I :r
I
I
"VSH
e·
s>
II i1111
..
> ;ff'
II1II1
AT SUPPORT
"'I'" . > ."
>
I
.
:ffill1
TrITl
> ,
~
del
"'I'"
I- .2!
.
-
I-
AT. de FROM SUPPORT deL 2y
,
Figure 4-19.
Determination of ultimate shears
4-65
,,
Table 4-5
Restraint and Aspect Ratio Requirements for Tension
Hembrane Behavior Edge Support 'Conditions L
I
.'
~ ....
..... ~
L·
I
External Lateral Restraint Requirements
Opposite edges
...
..
I
..
I
...
.,
,
~<.! H
....
:I:
•
,
-
..
Opposite edQes
;~,
~
t-
.
r:EJ3 I . I L
.
.
Opposite edQes short direction
-
-
I
L
1-" 'I /
.-
h<.!. H •
.
I
L
/
,
..
....
:I:
,
.
.
opposiie edQes short djre~tion ,
.,
I
I
None required
.!.
:I:
Indicates direction of external laterol restraint which must be resisted by adjacent member
4-66
..
--
Table 4-6
Ultimate Shear Stress at Distance de from Face of Support for One-Way Elements ULTIMATE SHEAR STRESS Vu
Edge Conditions and Loodil)g Diagrams
,
~. •
t
f
r
L
~.
f
,
L/2
•
I
L/2
L/2
.1
...&.
2d e
t
L/2
5L
ruts-de
RI,GHT
SUPPORT
ru(s--de)/d e
LEFT
SUPPORT
RIGHT
SUPPORT
3L
IIR u 16de 5R u 16de
de
•
,
"
Ru 2d e
·1
r u (L-del
,I
L
.
:'
-.f. i
I
L
L/3< L/3
de
Ru de j
,, .. I
f/2 L/3
t
4-67
Ide
SUPPORT
ru (~--de)
,
)
LEFT
•
/2
,
,
t
·1
"
de
~
t
t -.
L/2
L
I·
•I
t
L
L/2
ru(~-de)
.
Ru 2d e.
Table 4-7
Ultimate Shear Stress at Distance de from Face of Support
for Two-Way Elements Ed-. cOQdiwna
tntimUI borUorItal.ta.r . . . . . .
Umla
Yield liM locatioD
u...
Ultimate verticltJ ehear ,k_ ... V
-
Two adjac:ent edgee fixed and two edges free
OS;d,,/zSl
h
,
D I
I
L
OSd./HSl
.
r.(l-d.j:)
tSd./zSl
%[[;;;:.::)3
3r.(I-4.../%)' dJz(3 f#J../%)
O:Sd.,/L::~a
2(d./::I:)
ISd.IHSl
3r.(I-d./L) (2-1I/H -d.,,/LH) do/L{8 '118 -4d.rIUl)
OSd./"St
r.(I':'d./H){2-%/L-d..zIHL)
3r" (1..."d,,/,,)'
do11l(5 4d..1,,)
.
,..(I-d./,)
..../
tSd./vSI
2dJL(1 ....ILH)
-
2d./H.(I-d.z/HL)
i
r. (I-dolL) (2-,!H -d.,,/L8)
t,S;d.IL:S:I
3r.(I-d./H) (2 -zIL-d..zIHL) d./H(6-%/L fd.z/HL)
.
, Three edges fixed and one edge Iree ~
,
a-
b % I
00
,
',-
/'
ISd./HSl '
2(clJz)
.
>'j
d..1;(5-o&d./:)
d./H(3-:/L 4/4%jHL) r,,(I':"'d./HHl-:/L-d.zIHL) d.jH(1 2d,;z/HL)
3T.(l-d./,}' 4d,/r)
, d.11I(fJ
OSdJH51
,
.
.
r.(I-d./II)
..../.
lSd./,:S1
3J'.(l-d../J:)t
°Sd./J:St
3r.(I-d,./H)(I-:/L-d.,z/HL)
.
OSd./.SI
r.(I-2d./L) rJ-,/H -2d.,/LH) fd./L(I- .....ILH)
iSd./LSt
t-!--
,
3,...(1 ~UJL)(2-,IH -2d.I1/LH) 2d./L(6 r/H-Sd.rjLH)
O:S;d,,/LSt
<;
..
r.(I-d..b)
.,-.
I
....J<,
.O:Sd./HSt
d.b:(S-t4./:)
tSd./zSI
\
L
Four edges fixed
ar.(l-dJ:)1
°Sd./zst
a-,(t -d..IH) (1-J:IL-2d..J:IHL) dolH (3 -;/L -Bd.zIHL)
I'-,
" ,," --<"........ '-
%! --y--.. ..
r.(I-d./J:) /.
....
tSd.lzSI
.~~" " "
~
__ i
,"" .
__
L
tSd.IHSt
r.
-, OSd./LSI
-
3r.
3r.{I-d,11I)1 -d.I,,(5-td.,III)
OSd./IISt
: ISd.ILSt
r.
lSd./rSl
,
r.(I-d.I,)
..../.
-
e
e
e,
TH 5-1300/NAVFAC P-397/AFR 8-8-22
DESIGN OF FLAT SLABS 4-28. Introduction The typical unhardened flat 'slab structure consists of a two-way slab supported by columns. Except for edge beams which may be used at the exterior edge of the slab, beams and girders are not used to transfer the loads into the columns. In this case, the columns tend to punch upward through the slab. There are several methods that can be used to prevent this; the upper end of the column can be enlarged creating a column capital, a drop panel can be added by thickening the slab in the vicinity of the column, or both a column capital and a drop panel may be used. Hardened flat slab structures may be designed to withstand the effects associated with a far range J explosion.' The flat slab of a hardened structure . ' is similar to an unhardened slab but for a hardened.flat slab structure, the exterior supports must be shear walls which are monolithic with the roof. The shear walls transmit the lateral loads to the foundations. Due to the stiffness of the walls, there is negligiBle sidesway in the columns and hence no induced moments due to lateral. loads. 'Shear walls ,may also replace a row of interior columns if additional stiffness is required. Earth cover mayor may not be used for hardened flat slab ·~,tructures. • ' A portion of a typical hardened flat slab structure is shown in Figure 4-20. As depicted, there are generally four differ~nt panels to be considered, interior, corner and two exterior, each of 'which has a different s t Lf fneas-, The exterior pan~ls,are designated as short and long span panels which Fefers to the length of the span between the columns and the exterior wall. Inthe typical flat slab, the reinforcement would be ,distributed according to elastic theory. The elastic distribution of th~ flexural'stresses is approximated'by the methods presented in the ACI Building Code, The static design must meet all of the criteria of the code as well a~ of.all applicable ,local code~. However, for blast resistant structures, certain design criteria are more restrictive than those giveri in the ACI Building Code. To ensure two-way action in the slab, the aspect ratio'L/H of each panel must be greater than 1 but less than 2. While the ACI code permits 'unequal span lengths and offset ,columns, it is str~ngly re~ommended that offset columns not be used and the variation in span lengths be li~itedto 10 percent. Columns and column capitals may have either a round or square cross section, but round columns and capitals are preferred to avoid shear,stress concentrations. It is also recommended that haunches be provided at the shear walls. Flat slabs may be·designed to attain limited or large deflections depending upon the magnitude and duration of the applied blast load and.the level of protection required by the acceptor system. Under flexural action alone, the slab may attain deflecti~ns corresponding to 2 degrees support rotation. The flexural action may be extended to 4 degrees ,rotation If single leg stirrups are added to restrain the flexural reinforcement .. 'If ,s~ufficient continuous flexural reinforcement is provided, the slab, may ~ttain'8 degrees support rotation through tension membrane actio!". Unless" requIre.d for shear, single leg stirrups are not required for the slab to: achieve.support rotations less than 2 degrees nor ,tensIon membrane acti?n., The 'stress in the reinforcement as well as the type of cross section used to determine thE! ultimate moment capacity of the reinforced concrete is a function of tt,e maximum deflection. 4-69
TM 5-l300/NAVFAC P-397/AFR 88-22 The basic data required for determining the ultimate strength of ,the reinforced concrete, including the ultimate moment capacity and the ultimate shear capacity, have been presented in previous sections. Procedures for performing the dynamic analysis are presented in Chapter 3. Only modifications and additions relating to flat slabs are presented in this section. The interrelatiopship between the various parameters involved in the design of flat slabs is readily described with the use of the idealized resistance-deflection curve shown in Figure 4-21. 4-29. Distribution of Flexural Reinforcement 4-29.1. General For a two-way slab continuously supported on its edges, the flexural stresses are distributed uniformly across the entire slab (except for the reduced stresses at the corners). The flexural stresses in a flat slab supported,by walls and columns are distributed from one panel to'the n'ext depending on: the relative stiffness of the supports and the spans of the panels. Flat slabs also dis~ribute the flexural stresses 'transversely, concentrating the' stresses in the vicinlty of the column. A uniform distribution of reinforcement would result in a failure due to local "fan" yield lines around 'the columns at a relatively low resistance. By concentrating the reinforcement over the columns, a higher ultimate resistance is obtained'. " " An elastic distribution of reinforcement is required in the design procedure , presented in this Manual. This distribution will insure the formation of a predictable' collapse mechanism. Local failures around the' columns, and one-or two-way folding (local one-way action) will 'be prevented. With an elastic~ distribution of reinforcement, the yield lines form simultaneously across the entire ·~lab. In addition, the design will be more economical 'and cracking under service loads will be minimized. ' 4-29.2
Elastic' Distribution of Moments According to the ACI Building Code
Procedures outlined in the ACI BUilding Code are employed to determine the elastic distribution of the reinforcement (and hence of the moments): The Code presents two design methods, namely the Direct Design Method and the' Equivalent Frame Method. The Equivalent Frame Method may be used for all flat slab configurations whereas the Direct Design Method can only be used for three or more spans. Since the Direct Design Method 'requires fewer calculations, it is the preferred method and is discussed in detail in'this section. For the typical flat slabs with continuous exterior 'walls and L/H > 1, the column strips are H/2 in width in each direction. A wall strip is parallel and adjacent to an exterior wall and its width is H/4. The remaining portions of the slab are 7alled middle strips. ' Using the Direct Design Method as given in Chapter 13 of ACI 318-77 the moments are distributed taking into account the relative flexural and torsional stiffnesses of the wall, slab and beams. 'Assuming there are'no beams or interior shear walls, the ratio of the flexural stiffness of the beam section to the slab section a , is zero. The torsional' resistance of a concrete wall monolithic w~th the slab is very large and, therefore, the torsional stiffness ratio of the wall to the slab'B t may be assumed to be greater than 2.5. 4-70
TN
5~1300/NAVFAC
,.
P-397/AFR 88-22
.
The ratio of the flexural stiffness of the exterior wall and the flexural stiffness of the slab is defined as: (4E c I w) / H.,
4-49
(4E c I s ) / ls where a ec - ratio of the flexural stiffness of the exterior wall to slab I w - gross moment of inertia of wall Is
- gross moment of inertia of slab
H., - height of wall ls
- span of flat slab panel
In direction H, Equation 4-49 becomes Tw 3 H a ecH - - - - -
,,
4-50
and in direction L
a ecL - ---,---
4-51
-.
where
Tw
thickness of wall ,
H
- short span of flat slab panel
Ts
- thickness ~f fl~Cslab
L
- long span of flat' slab panel
The unit column and midstrip moments' a~e proportioned from the total span moments. The distributions percentages for a flat. (slab with equal spans in each direction is as follows (see Fig. 4-22): For Direction H: 4-52
0.65 a'ecH MoH /L
e
m2
+ - 0.40 (0.63
m3
0.25 (0.75
m4
0.25 (0.65) MoH /(L-H/2)
0.28 a'ecH) MoH /(L - H/2)
4-53
0.10 a'ecH) MoH !(L-H/2)
4-54
4-71
4-55
TN 5-1300/NAVFAC P-397/AFR 88-22 m5+ - 0.40 (0.35) MoH./(L-H/2)
4-56 '.
m6+ - 0.60 (0.63 - 0.28 a'ecH) MoH /(H/2)
4-57
m7- - 0.75 (0.75
4-58
0.10 a'ecH) MoH /(H/2)
m8- - 0.75 (0.65) MoH /(H/2)
4-59
m9+ - 0.60 (0.35) MoH /(H/2)
4-60
in which wL(H-c)2 4-61
8 1
and
a' ecH - - - - - 1 + (l/aecH)
4-62
MoH - total panel moment for direction H w
- applied uniform load
c
-. width of column capital
For Direction L: mlO- - 0.65 a'ecLMoL /H mll+ - 0.40 (0.63
0.28 a'ecL) MoL /(H/2)'
m12
0.10 a'ecL) MoL /(H/2)
- 0.25 (0.75
4-63
4-65
m13- - 0.25 (0.65) MoL /(H/2) m14+ - 0.40 (0.35) MoL /(H/2)
..
4-67
m15+ - 0.60 (0.63 - 0.28 a'ecL) MoL /(H/2)
4-68
m16- - 0.75 (0~75
4.-69
m17- - 0.75 (0.65)
0.10 a'eeL) MoL /(H/2)
...
~~
/(H/2)
4-70
m18+ - 0.60 (0.35) MoL /(H/2) .
4-71
wH (L_c)2 MoL----
4-72
in which
8
and 1
a' eeL - - - - - -
4-73
1 + (l/aecL)
4-72
TIl 5-l300/NAVFAC'P-397/AFR 88-22
where MoL' 'is' :the total' panel moment) for; direc'tion' L. '" For the wall strips in both di.rections the same reinforcement as in the adj acent middle strips' is used .. At"the co lumn; the targer' or' the"hici:'negative moments is' chosen and the posLt Lve moment 'can then .be a,djusted tip to '10' , percent so' that C the' total, panel' mome'nts remain unchanged,": '" " ," ';"
4-29.3
"
:
".~
<,
•
-
. ,
\
_.'
r'
r
,Design for Small Deflections 'T'
.....
l-
t.. "
The resistance-'deflectlon function 'fo~ flat slabs' with sm:all'~'deflections- is shown in' Figure 4~23a.' With an elas't·rc·'di~trl.bliti.6n of r,einfor'cetnent aU the ' . . . . " .' . I • ,. • , " yield lines form simultaneously and the slab remains elastic until it reaches its ultimate' resis'hmce. Since"'for 'smail 'deflections' the conc re t;e r.!tnains effictive in' .esisting s tr'e as , a type I' cross ""ec,tion i Ls ,Used,to compute the ultimate moment c":pacities.' 'The slab may undergo a'maximum support rotation of 2 degrees' at which point the ccnc re t.e: crushes.' Shear' reinforcem'krit is' gene r a.LLy not required for fla,t slabs, but must be provided if r~"qtiirE"d by analysis. If properly designed, single leg s t Lrrups are provided, the, flexural action of the slab may be extended to 4 degre,,,,' supp6rt rotation:' i,' While stirrups may be furnished to ~esist shear or to "xterid flexural action, it is 'usually more cost 'effective 'to design flat slabs' without shear'reinforcement.-' ~,'F--' t'\' .. ·· "'~".-' ''':'..It,· ,". ~,'
~
: .. :
4- 29.4. De-S'lgn"fbr 'Large·D~fi.ections .
..;
_..
. - -: .
.;.';'-.
-
.
'.
.
. . ., . ~.. -, ~ .,'., ." t . ~, Due to the geometric limitations' (aspect"ratio L/H of each parie L must be greater 'than .r' !'but 'less tlian '2) imposed on flat slabs des'igne'd' for blast .' " .',.'" ,. ' . 1 .• loads, sufficient' lateral restraint'·is available ,to develop "Lri-p Lane forces.and induce' ten'S'ion inembra'ne 'action.' 'This tension 'membrane action provide's the ,me';ns for the sl',lb to: attairi"defl~ctions corresponding to'. a"';aximi.un support' , . '-' .. . rotation in excess of 12 degrees. Continuous reinforcement must be provided to r_~s",ist I thr~~~'..~n-p,la~e !~.~~sion,. ~orc~s. , '"'. l' ,.~
.
~'(
A typical resi"t'1in~e-deflection'~urve :f~r th~' flat slabs up to incipient failure is shown in Figure4-23b'. ,.The initial po r t Lon of the, curve is due primarily' to f\ex\frai acdon .. : 'AV ,2"degreeS' 'support rocactcn, 'the 'concre ce begins to crush,~nd the s Lab 'lqse's' flexural capacity, However, due to 'the" presence 'of continuous 'refnforceinent; tension membr ane action is -mobil1~ed. " .i".'i"' "'. ."" .,', .' ., " -" " , " , The re'!'ristan!'ed,ue' t~ t;his ,action,':increases with increasing deflection up, )00" incipient faIlure at 'approximately"12 'degrees support rotation" The t ensLon ' membrane action is shown' as the dashed line in Figure 1.-23b. ~"
,~,
~.
';/
' ,
~,
"
,"
In order to simplify the design calculations, the resistance is assumed to be equal to the flexural action throughout the entire range of behavior.' To approximate the energy absorbed under the actual resistance~defl~ction curve, the deflection of" the equivalent curve is limite,d to 8 degrees support " rotati~~: This deflectioiirw~uid pz oduce inc'ip'ient failure conditions. Us Lng thi~equivalent design-'curve, deflectioris, between 2 'deg;re,esand, incipient faiiuie ,cannot, b,e 'accurately predicted. " -- , " ",. ' , .'
,-
.r.'
,
.
' "'
-.
-
"
"', ".
A type II or ,Ip,cro~'s sec:ti~rri,s,used,~~' compute :the ul,timate moment ~,apacity of a 'flat ,sfao"designed for large d<;fleftions, At. every section ,throughout. the slab, tension and compression reinfo'rtement must bra continuous. in order to develop the tension membrane action (tension membrane capac icy fs 'di'scussed in detail below). It should be noted that in addition to the above requirements
4-73
\
TH 5-1300/NAVFAC P-397/AFR
88-~2
for the reinforcement, an elastic distribution of reinforcement must, still be maintained, Shear reinforcem~nt is.~nlyprovid~d,whe~ -iequired' by', analysis', I f tlJ'\' concrete' can resist the shear stress,es, shear reinforcement is not required for flexural'action (deflections le~~, than 2, degrees), nor ,for tension membrane: action (deflections between 2 and 8 degrees). 'However', shear reinforcement, " in the fom of single leg stirrups, does, ,allow, the slab to rotate "!p to 4 . ' .. degrees under flexural action. There are two design situations where single leg stirFups are desirable when designing for rotations between 2,and 4 degrees,: ' '(I) the dab, is ,incapable of: deveLop i ng adequate' tens,ion'm,embrane action and (2) 'the maximum deflection, mus't' be"accurateiy predicted, (which ,: "" cannot be done utilizing tension membrane ~~tion)'.•', In 'all' 'other design ' situ~tions, it is usually mor~ ,economical ~~eliminate,single leg stirrups,by increasing .cbe slab thickne,ss (to increase shear capaci,ty). and/or, by Lnc'reas -".. ing the amount; of .contInuous reinforcement (!=o ,develop 'adequate rens'Lon , membrane capacity). ' '.•
'"
.~;
~
. . "
"
n
'
~.
••
-,
~
.•
"
4-29.5, MinimUm Reinforcement . 'fe' , •
..'
•
j
To ensure proper strUCtural behavior under, dynamic loads ,and also, to minimize' excessive'deformations under conventional loads, the"mlnimum area of r;i~for cement m~st.be at least equal to that specified in Table 4-3. With an elastic , distribution of'reinforcement in a flat ,slab; the ml,,!imum'reinforcement generally will occur only in the center of the' inidstr1pandior 'in 'the wall strip. It is important to also check ,the static requirement~ for minimum reinforcement. Where static conditions contro L, ,the,area 'or' reinforcement, must. be,at least equal to 0.0018 times the gross :0';' 1.33 ,~ " " . 'are,{of,. concrete ". . , times the area required ,by static loading conditions, whichever Ls less ',::' . Unles,s ,the blast load,s are in the!,~~ order .of ;"~gnftude as th<;..stafic l-;'adl','. this crlteria does not contro L. " .,,_ , .. i . ,-
,
,
.
, .
~
~,
Although the spacing of the flexural reinforcement m~st not exceid'two times the slab thickness nor 18 inches, the preferred, spacing is 12 inches or Less., •
.
"r
There fs no minimum shear reinf;;~ce;"eritrequirementfor' flat slJbs. ,shear reinf';rcement is only"provided when required~i.' analysis.' 'Howe~er, .:whe;'a slab is" de s Lgned to undergo ,flex;ural response with support r0t;atlons .be tween 2 and 4 degrees (i.e.. , where t.ens Lon membrane action is not consider.ed). " ' stirrups are required. The mini;"um area of the".s·tirrups"i.s' given Ln Table 4-" 4. . ' " ~. . " 4-30; Dynamic Analysis "' ' . \ . ' . "
4- 30,1. 'General
.r: '
,
:" '.:
"
by
The dynamic analysiS of a 'structural ele;'e';"!: isahc'o;'pli~hed fi,rst' represeriting the ,struct~ral' element ,a!, a 'single-degre~'of-freedo~ system a~d'then finding the response of that system when subject to a, bla~t loa.d. ,Chaptef 3 presents procedures, figures and response charts for 'determining the dynami~ cally equivalent system and its response. However, certain parameters of ~ flat slab, such as the 'ulti~ate' resistance and the elastic deflections, cannot be calculated'using the methods' of Chapter 3', Methods for calculating those paranieters ,,;re pre'seneed below'. . '.,.. '-' ," , , •
,-
" I
"
...
"
-
,-
.'
4-7,4
'..
~
.
. . -
,':
~
TH 5-l300/NAVFAC P".397/AFR" 88-22
4-30.2. Ultimate Flexural Resistance 4-30.2.1. General The ultimate resistance r u of a flat slab is a function of the strength, amount and distribution of the reinforcement, the thickness and strength of the concrete and the aspect ratios of the panels. The ~ltimate resistance is obtained using a yield Lfne analysis .. Since Ln-p Lane compre ss Lon 'forces'and' tension membrane forces 'arel'lio-t considered;':;the u'l t Lmatievres Ls t.ance determined
from a yield line 'analysis will.generally be lower than the actual resistance. .
f) -, ,
.
",'
The first step in finding', tne ultimj,.te'resistance is 'to assume a yield'line ' pattern consistent with'the support conditions and' the distribution of/,the' reinforcement. The pattern will contain one or more ul1known dimensions: which locate the yield lines. The correct solution is the one which gives the lowest value of the ultimate resistance, Figure 4-24 shows the yield line pattern that will form in a multi-panel flat slab with an elastic distribution of reinforcement. The roof-slab interactions must be designed to insure that the perimeter yield lines form in the roof s Lab and not in the wall. The yield lines at the columns are assumed to form at the face of the column capitals. The ultimate' resistance can be .. found from the y Le Ld line" pattern.'\ising either
the equilibrium method or the virtual work metl\od;' ,both of which nave been discussed in Chapter 3. The equilibrium method is one 'that has been employed in previous examples but, in the case of flat slabs, r,equires the introduction of nodal forces which ar'e n:ot always readily determine,ti'. The'virtual' work method though more difficult to solve algebraically, does not'require the' calculations of the nodal forces. Consequently, the virtual,work method is the easier method to applY'to flat slabs and'is the method detailed below. The virtual work method does not predict the correct yield line pattern but rather gives the minimum resistance of an assumed yield line pattern, If the distribution of reinforcement is not elastic and/or'the span lengths are not approximately equal , the minimum, resistance found by the virtual. work method 'may not be the ei tiin:a:te r e s Lscance , "In, these cases, 'loc';l failures are '.' possible. It is strongly recomm!"nded that these design' s i tua'tLons be a";o'ided. In ,the rare instances where they cannot be avoided, die, nodal' forces must, be ... calculated and the equilibrium method used to predict the correct. yield line' pattern., .I 4-30'.2.2. Virtual Work Method In the virtual work method" equations for the external and internal work are written,in••terms' of theuU:it resistance r~~,the mqment~apacities and the' geometry. "The expression for the ext.erna l, work is set equal to that fo~ tpe" internal work, end the resulting equation is solved for the minimum value of" r u an~, the ass.ociated, failure meclianism. '", •
'.
'
:"".
I .
.!
r
A point within the 'slab boundaries i~given.a small displacement in the direction of the load. The resulting deflections and rotations of all of the slab segments are determined in terms of the displacement and the slab segment dimensions. Work will be done by ,the ,external loads and,by the internal reactions along the yield lines. 4-75' ,
~
TK'S-1300/NAVFAC P-397/AFR 88"22 c",·
,
'
,
"
The external work done by r u is: W- E
r.J>A
,
where:
, ,
4-74
'
external work' A- area of the sector A,- deflection of the sector's centroid
,
."
The :i.~tern;'l work done by the reactions'at'the yi~ld H;'e~ i~ due only to the bending,moments since the support reactions do not undergo any, displacement and the, work done by the shear forces'(n~da~ for~es)' is'zero when s~~d ~~er the,entire, slab. The internal :work is: E - E
,
mel
,,-.
.
4-75.
'
where E m
-.
.
,9
I
internal work ultimate unit moment relative' rotation' about yiel? line, length' of the yield line,
'
••....of
.
reinf~rcement direcIn terni~_of the moments ~rid r~~a~ions in the, principal .L tions X, 'and y: 4-76
,E Equat~ng
the external and internal
w~rk,
W- E 4-77
Particular ~tte'1tion must be paid to the negative .J!lOment capacit~es of the. yield lines' radiating from the column capitals/when determining E, Top bar' cut-offs, i f present, will r'educe tthe moment/c;"pacity on the part of the yield line' furthest from the coi.:.mn. 'In addition;' corrier effe~ts' must be cons Idered wherE,"'the two walls intersect. That is, a'result, ~f 'the increased stiffness at the corners, the ultimate moment/of the reinforcement is reduced 'to 2/3'of its capacity over a length "equal' to Ii the length .of ,t;he positive yield line. "
as
To illustrate the application of Equation 4-77; consider the fl~t 'slaD shown in Figure'4-25. Thi~ flat' slab is the roof';ofa, square structure with one central column, and is symmetrical about the x and y axes.' '" '.
,
Note that Sectors I and III, 'and Sectors, ILand' IV are identical because' of symmetry. To simplify calculations, each sector has been resolved into a ' rne rectangle and.a 'triangle. .. ' . 'external '. - work . • f,k ..' each "ectdr is: c ". .;
..
'.,
4-78'
4-76,
TH 5-1300/NAVFAC P-397/AFR 88-22
VII - VI V - ruc(L - x - c)(A/2) + r u (L
-'X
-
c)[(L - x -'c)/2](2A/3)
4:- 79
Substituting L - 41 and summing
~ V - 2 (ruA/6)(32 1 2_ 4lx + cx - 4cl
4·80
where c L x 1 A-
width of column capital' length 'of panel , horizontal location of the yield line width of ~ of the column strip maximum deflection of slab
,'~
The internal work for each sector is: < ,
x
2 -
3
+ [ - (2m) 2
3
m (L -
2
x
2
x
+
m -
)
] 9A
2
x + 2m(3l - - ) + 3ml]9A 2
4-81
Ell - EI V - [3ml + 2m(L - x - 1)]9B + [4.5ml + 1.5ml (L
x
4-82
1)]9B
Substituting L - 41, :9A - 'A/x and 9 B ~ A/(L - x - c) ~
E
18 1 - 3.5x 13 1 1 2Am'[-,-, • - - + 41'- x - c x 2
,
Equating V - E and solving for r u 131 6m x r .u -
4-83
]
'
"
iai 3. 5x 1 - -+ -----2 41 - "X - C 4-84
---"7""~---'--------:-
321 2 - '4lx + cx - 4cl _c2
with x as the' only'unknown~' The minimum value of·ru breadily determined by trial and error. A complete design example is presented in Appendix 4A. In general, the ~irtual'work equation wiil, contain morEl than' one unknown, and it will be correspondingly more difficult to obtain the minimum ultimate resistance. However, a trial and error,process rapidly converges on the correct soluti~n. "
4-77
TIl:
5-l300/NAVFAC P-397/AFR 88-22- '
A trial and error procedure to solve for the minimum value of the resistance function (ru/Mn for a preliminary design and r u for a final design) with two unknown yield line location~, x and y, can be accomplished as,follows: L
Start with both yield lines located close to the centerline of the respective middle strips.
2.
Vary x, holding y constant, in the direction which minimizes the resistance function until it begins t~, increas~.
3.
Hold x constant~nd vary y in, the, minimum direction until the resistance function begins ,to increase. :j' .
4.
Once this minimum point is achiev~d', shift each yield line to either side of the minimum location to check that a,further refinement of the yield'line is ~ot necessary to minimize the resistance function:
'-j
.'
It should be noted that if the yield li~e' should shift out of the middle strip, a new resistance function equation must be written and the procedure then repeated since the magnitude of the unit moments acting on the yield lines would change. 4-30.2.3. Effect of Column Capitals and Drop Panels r. ~ .
Although column, capitals and di~p panels are primarily used to prevent shear failures, they have a significa.nt effect o~,the ult~mate'resistance. The addition of a column capital or revision of the size of the capital changes the clear span of a flat slab and req~ires, the,re-evafuation of a slab's ultimate resistance. ' Drop "panels increase the ultimate resj.~.t.ance by incr!,.a,\ing the depth of the section and thus the moment capacity in the vicinity of the column. This effect can be countered by decreasing the amount of reinforcement to maintain the same moment capacity. If the drop panel is used to increase the negative moment capacity, it must extend at least 1/6 of the center-til-center span length in each direction. The width of ,the drop panel may be up to 20 percent larger than the column strip. When'the Arop panel.is larger than the column strip, the percentage of reinforcement calculated for the column strip shall b~ provided throughout the drop panel. ,Additiona.l r~inforc~ment must be provided in the bottom of the drop panel to prevent it from scabbing and becoming hazardous debris. For a typ~ II cross section, the reinforcement in the drop panel is the same as the negative reinforcement over the column. Only ~ the.amount,of,the negative reinforcement is required in a drop panel for a type' I' cross se~tion. ,,' , .' ".. " 4-30.3. Ultimate Tension Membrane
~apa~lt~
,
"
'
When the support rotation of a flat ,slab reaches 2 degrees" the.concrete begins to crush and flexural ~ction ~s nO,longer possible. ,Ho!eve!, the,slab is capable of sustaining large rotations due to tension membrane action. As previou~ly ~xpiained;"the' actu~l re~i~tance-deflection curve describing the tension membrane action has been replaced with an equivalent curVe which considers flexural action only (Fig. 4-23b). Using this idealized curve,' , ,
" ,
•
" , j
4'C78
,
•
TN S;1300/NAVFAC P-397/AFR88-22 incipient failure is taken to occur at 8 degrees which corresponds to'an actual support rotation of approximately 12 degrees. It can be seen from Figure 4-23b, that .the tension membrane resistance is a function of the deflection. It is'siso a'functiori of the'span length and the amount of the continuous reinforcement. Data is riot p:resently available to . . . , . obtain the tension membrane capacity of a flat slab. However; an approximation may be made using the equation developed for two-way ·slabs. Therefore, the tension membrane capacity, rT' of a flat slab is given' hy':" ' , .~,
.-~'
"
1
. '[" ~Lx.
of
cosh
:2~
[~H] 1/2]' . ~L,
"
J,
_/
where
,
.\
•.
rT
ten~ion membrane r~sistante
X
deflection of slab
La
clear span in short direction
,
Lx.
'..1
4-85
. •'!:of
'.
-
TH
,
clear span in long direction
I'
:-:..
force in the coi.i:inuous reinforcement in short span direction
".
Ti Altho~g~,~h~,~~p'a~i~yof.a,~~at,.~~ab 1~ b~~e~,on ~}exural:acti~p:,adeq~~te
tel).~ion JIlembrane capacity must be provided~ '. That, ~s, rT corr~sponding to 8 degrees support'rotation must be grea~~rthan tlie'Flexura} re~istance r u '~hen designing for large deflections, The deflection' is computed as'a function of the'yield line locations (shortest sector length): The force in. the continuous'reinforcement is calculated using the dynamic design stress cor- . responding to 8 degrees (Table 4-2), . The clear span 41 and Lx. are calculated as the clear distance between the faces of the support" (face of'the column if no column capital, is used, ~ace of the column capital, face of the wall if no haunch is used or' the 'face of hau";ch),
.".
"::'
•
-".
j
•
>.
4-30.4 .. Elastic'Deflections'
'~.
"'.' .
,.
The elas~ic'd~fi~ction of variou~ ~oint~)m.an . b\t~r1or panel of a flat Slab are given,'~y .t~e::'geI!eral eq~a~loh "
.;
~
4-86
Xe -
4-79'
TN 5-l300/NAVFAC P-397/AFR 88-22 where elastic deflection from'~able
C
deflection coefficient , long span of panel
4-8
v
poisson's, ration - 0.167.
Ia -
average of the cracked and gross moment of inertia of the concrete ' slab
The deflection coefficient varies with the panel aspect ratio L/H, the ratio of the support size to the span C/L and the location within the panei. The values of the deflection coefficient given in Table 4-8 are based on a finite difference method.an~ are gi~en for the c~nter ~f the pane~ Cc and the midpoint~ of the long and short sides, CL and C S' respectively. The deflection for the interior panel is determined by using Cc in the above expression. For the long and short span panels and the corner panel (Fig. 420). No. simplified' solution for the center 'deflections are currently available. Generally, the deflections for these panels will be smaller than the deflection of the interior .panel because of the restraining effects of the exterior walls. These deflections can be approximated by using the following expressions: Long Span Panel
C - Cc -, CS/2
4-87
Short Span Panel
C
Cc - CL!2
4-88
Corner Panel
C
Cc • CS/2 - CL!2
4-89
where the valu,!s 4-8.
~f
~C' Cs and CL are, those for the .interior panel from Table
The.dyn~ic response of a flat slab is more sensitive to the elastic s~iffness when the maximum' allowable deflection is small. The possible error diminishes with increasing allowa~le maximum deflecti~n.
4-30.5. , Load-Mass Factors "
4-30.5.i. Elastic Range
tK .
No Oata, is 'currently availabie to determine .the loadmass factor, K of a flat slab in the elastic range of behavior. It is, therefore, recommended, that the values listed in the table of the load-mass factors for two-way elements'be used (Chapter 3). The slab should be considered as fixed on all four edges with the appropriate L/H ratio. Since an average value of th~ elastic and plastic load-mass factor is used in determining the natural period of vibration, the possible error incurred will diminish with increasing allowable maximum deflection.
TK
5-130~VFAC
P-397/AFR 88-22
4-30.5.2. Plastic Range The load-mass factor on the plastic range 'is determined'uslng' ~he procedure outlined' for two-way elements in Chapter 3. The '~upports for the individual sectors are at the face of the exterior walls (or haunches, if present) or at the face of the column capitals. Flat slabs without drop panels have a uniform thickness and the equation for determining the load-mass factor may be expressed in terms of the area moment of inertia and the area of the individual sectors. For flat slabs with drop panels" the' equations must be expressed in terms of the mass moment of inertia and th.. non-uniform mass of the individual sectors to account for the norr-rrndfo'rm sl.ab 'thickness. ' 4-30.6. Dynamic Responsa
J' ",
r
The equivalent single-degree-of-freedo~system'of'the' f l.at'slab is defined in terms of its ultimate resistance r u' elastic deflectIon XE and its natural period of vibration TN' The procedure for determining the value of TN has been presented in Chapter ~ while the calcula~ion of r u and XE h~s been presented above. The .resi~tance deflection curve used in the dynamic analysis is shown in, Figure 4-26. The resistance available to withstand the ~last' ' loads must be reduced by the dead loads. An approximation of tne resistance available is . f
ds
4-90
rDL (--) fy
wher~
dynamic resistance available uniform dead load The total deflection of the flat slab includes deflections due to dead load arid blast Xm. so t;hat' the' ;"aximum 'support rotation Elm is given by
X DL
.em· -
tan~l (
Xm
+
XDL ,
Ls
4-91
)
.' i
where Ls is the length of the shortest sector. The blast load is.defined in terms of its peak pres~ure P and.its du~ation T which are determined from Chapter 2. Chapter 3 contiadns .the pxocedures to determine the dynamic' r espcnse of,..a slab "~hicll LncIude , the maximum dynamic deflectionXj" and the tim~ 'to reach th~t' de f Lec t Lon t m. ' I t ' must 'be r'emembezed that using" the equivalent'resistance-deflec:tion curve 'to include tension' membra~e action, deflections between 2 degrees and incipient failure cannot be accurately predi,cted: ' , '
"
The , required rebound resistance of the flat ~lab':i.s calc:ulilted in accordance with 'Chapter 3 and the reinforcement "necessary to attain th:i.scapacity must be provided. Note that while the dead load reduces the 'av'dlable resistance for' the dynamic loading. this load increases the available resistance for rebound. ,
,
4-31. Dynamic Design 4-81
,
TN 5-l300/NAVFAC P-397/AFR 88-22 4-31.1. Flexural Capacity The ultimate moment capacity of a flat slab'is usually based upon a type I or, type, IIi cross section depending the magnitude of the maximum aliowable' deflection., Th~ distribution of reinforcement is critical in flat slab design. The actual moment capacity provided must be as close as possible to the unit moments required for an elastic distribution of stresses. The quantity of flexural reinforcement which is mad~; continuous provides the
on
tension membrane. resistance.
If the amount of' continuous' reinfor'cement provided is inadequate for .t.ens Lon membrane action, care must be taken'in furnishing additional reinforcement. Any additional reinforcement must be placed to maintain the elastic distribution of reinforcement and the new moment capacities and ultimate resistance must be re-evaluated. The ultimate moment capacity will not be altered if the additional reinforcement is provided by increasing the compression reinforcement.
4"31.2. Shear CapaCity'
.. Unlike continuously'supported two-way slabs where'shear stresses are '"checked" .
,
"
"
-
after the flexural design is completed, the ,design for shear of a flat slab must be considered during the flexural design. Due to the nature of the support system, flat slabs will usually generate large shear stresses. Flat slabs with high percentages of flexural reinforcement and/or long spans should be avoided. The shear forces acting at a support are a function of the tributary area of the sectors formed by the yield lines. The shears at the colUmns should be cliecked first, since design for these 'forces can drastically effect the flexural design of the slab. Two types of shear action must be considered; punching shear along a truncated cone around the column and'beam shear across the width of the'yield lines. These c~ndiiions are. illustrated in Figure 4-27. Shears at the columns may require the use of column capitals and/or drop panels. Punching shear can occur around the periphery of the columns or column capitals and drop panels. The critical section is,taken at d e/2 from the face of the support. The total load is calculated based on the area enclosed by the positive yield lines and is then distributed uniformly along the critical perimeter. Figure 4-27a illustrates the critical sections for punching shear. Beam shear, as a measure of diagonal tension, is taken as one-way action between supports where the width of the beam is taken as the spacing between the positive yield. lines. The critical section is taken as de away from the f'ace of the column or 'column capital and from the face "of the drop panel (Fig. 4-27b). The total load is uniformly distributed along the critical section. ' The slab at the e~terior walls must be' evaluated for diagonal tension capacity. Due to the assumed uniform distribution of load at the exterior walls, a unit width of loaded area may be considered between the positive yield line and the critical section. The cr!tical section is taken at 'de from the face of the e~terior wall or, if a haunch 'is used, from the face of the, haunch. ,
,
The ultimate ,shear capacity of slabs has been previously presented. Using! these procedures, the capacity of the slab is evaluated at the locations
4-82
TH 5-l300/NAVFAC P-397/AFR 88-22 described above. If required; stirrups' may be furnish"d . . However, it is more cost effective to revise the design to incorporate the use of column capitals, drop panels and/or .increased slab thickness to reduce shear stresses. As preyiously stated, the use of stirrups'is mandatory in the flexural design of flat slabs between 2 and 4 degrees support rotation. ' Diagonal bars must be provided at the face of all supports due to the cracking caused by the plastic moments formed. For'slabs designed for small·support rotations, minimum diagonal bars must be furnished. However, for slabs designed for large support rotations where:the cracking at the supports is severe, diagonal bars must be designed to resist the total support shear but not less than the minimum required. The· diagonal bars furnished at the column supports should extend from the slab into ·the column. In slabs where shear stresses are high, it may be impractical to place the required diagonal bars. r
If column capitals were not initially used, their addition would reduce the requiredquantity'of diagonal bars. In the case where column capitals are furnished, at least one-half' of the diagonal bars should extend into the column with the remainder cut-off in the column capital. Procedures for the design of diagonal bars have been previously presented while the required construction ,details are illustrated in subsequent sections. ,
,.
4-31. 3. Columns
The interior 'columns o'f a flat slab/shear' wall structure are not s'ubj ected to lateral loads· nor the moments they induce . . These colwnns are designed to resist the axial loads and unbalanced shears generated by the ultimate resistance of the nat slab. The axial load and monenxs 'at. the' top of the column are obtained from, the flat slab shear forces ac,ting on the perimeter of the column capital plus the load 'on' the tributary area of the column capital. As can be seen from Figure 4-28, the axial load is:
4-92 and the unbalanced moments are:
V2) (c/2)
4-93
(c/2)
4-94
V 3)
.The;procedures for the design of columns is presented in Section 4-49. When using these procedures, the unsupported length of the column is ,from the top. of the, floor to the bottom of the column capital.
4-83
,
,
,
'
L
\'
L
.' ."1"'
. -.
,----I
,
.
'
.
~
•
I"":
---+t==------+-__ ..,...--.....---:--f--:----,--::... , ----------
I
~'"
SHORT. SPAN
CORNER
I
1--' I r ,"l I,
I
I L cJ', L __ .J
.
".~
,. ',;
•
'.,
INTERIOR
LONG SPAN 1
f.~
-
.,' Irr -l l '
' - . --I
I r ',.
L
J
t
l r-t-t L "..J
I
L
_-.J
c'
"
,;
,
'1' . ,"
_..J
,L
t'
J
'.J
NOTE,
!:. ~ H
,
,
_.
'}
I.
., .,t.
PLAN
:. '
,
,-
,
c ~
~o
-
"
, ,
Tw,
"-
/ \
.
'
,
.
"'--'S:.,. RO~ND
.
'"
,.
SECTION
Figure 4-20
Typical flat slab structure 4-.84·
or SQUARE
.CAPITAL "
a
COLUMN
w u z
t1LTIHATE
ELASTIC
<[
fVl ~
Vl
W ~
X('~")
X( •• t")
X( '.8')
X('.5')
DEFLECTION
CD
otp 1
Not
2
PROTECTION CATEGORY
® @,LG) SH(L TER
PROTECTIVE STRUCTURE
BARRIER P ESSURE-TtME
STRUCTURE-LOAD SENSlTIV1TY .
" FAR-OUT
.
DESIGN RANGE ESPONSE
CHAR
S OR ITERATION Pit: CEDURE
DESIGN IIlETHOD
LIMIT
LARGE
D
DEFUC110N CRITERIA FLEXURAL S IRRUPS IF
FLEXURAL ACTION
CTION
~T1RRUPS REQ'D.
EQUJRED
DUcnLE, IwlODE OF RESPONSE
-
,
TENSION
EHBRMIE ACTION
STIRRU S IF RE UIRED
.
I
CROSS-SEcnON TYPE
(fdy
Not~
Note
Figure 4'-21
, II
,<
fdy-+1I4
Fdy
DESIGN STRESS
SCABBING
CRUSHING
NONE
aarrnr "lODE Of RESPONSE
"
III
1/2
I. Nplthe-I" she-G.I" re-lnforce-I"Ie-nt nor te-nsll,e- "~"bl"o.ne- ec-ncn Either shear relnforcel"lE'nt 01" tensIle nel'lbrane ec-ncn
e.
Retationship between design paroMeters for flo.t slo.bs
4-85
NOTE:
!:.H >- I L
•
m+ 2 12 E
+=
E
m+ 2
~
E
i'
IQ
m+6
1!2 E
o I~ E
,
- ....... -
+!!! -E
+!: -E
1!2 E
. ,.'
+!!!
"
E
m+9 12 E
+=
___
~
I
+!:
E -
0"-
E
IN
"
8-
· ·· :toI
,I
% Col.
:.F;
PLAN
Figure 4-22
Unit
'~~~ents
~
It: E
~o :I: U
ru +-+-----, ....
.
»: ~ t,
'
..
a) Small' Deflection
-_
....
-: "
:
ACTUAL 'RESISTANCE DEFLECTION FUNCTION
IDEAL.IZED RESISTANCE' DEFLECTION FUNCTION
j
r
.
'
-,
....
;'"[r- .'
XE
:
..
I~ • ,
-, ... _ ...
.
. ,/"" ,/
/ , '
CONCRETE CRUSHES, ...LOSS OF FLEXURAL . ...i RESISTANCE '.. .. ."'. . /
<,
.
:
1 ..
'..
-
•
"/"
,//
'
,'
..
X (6'=2°)
b) Large Deflection
']l~
Figure 4-23
1.
J ..
J'.:" •.
l : ..
l-:.~'
Typical resistance-deflection functions for flat slabs
4-87J'-
o·
,---1
L
L
T
•• •FACE OF WALL
..,....-------_r_--~_r__I_-----'--
W'
x
II
II
PLAN
Figure 4-24
Yield line pattern for multi-panel flat slab
4-88
"O~d
~L-~~[
I-x .l
L-X-C
.1
.
SECTION A - A
NEGATIVE YIELD LINES .-I .-I
I
@
~.
--
I
POSITIVE YIELI) LINES
I
.-I
CD I.L
0
"'C\l" ~
I
z
~
""(
:::l .-I
0
At.
I
I
I @'
. o+"':-_~""1-NEGATIVE ~ YIELD LINES
__1_- - I I
U
FACE OF WALL
I - - - - _I . -
"" «
.,
"
@I
I
~~
COLUMN CAPITAL
~A
1_ _-
SYMMETRICAL ABOUT ¢,'.
C
x
L-X
PLAN
UNIT MOMENT CAPACITIES: EXT, NEG, WALL STRIP MID STRIP
m* m*
COLUMN STRIP
m
*' Figure 4-25
POS.
2m· 2m· 3m
INT. NEG.
1.5m 1.5m 4.5m
. ,''2/3 OF MOMENT CAPACITY EFFECTIVE AT CORNERS
Quarter panel of flat slab 4-89 ' ,
ru
'KEI
., .
I
..
."> ..."
L-
..
xDL
I XE Ir
Xm
.,
, . Figure 4-26
Dynamic resistance-deflection curve
4-90
·,
dQYCj.
- 2
JL JL
FACE OF DROP PANEL
d QVlJ.
2
FACE OF COLUMN CAPITAL
o. PUNCHING
SHE.~R
:z:•
L ~F D~OP FACE
PANE-L
FACE OF EQUIVALENT COLUMLN/:APlr:r L
-x
b.
Figure 4-27
BEAM
SHEAR
Critical locations for shear stresses
4-91·
AREA I
AREA
4 . V:fl -:: 1:3; .
.
4
I
COLUMN
I
L_.J
¢>! I
U
I ~v,. I
CAPITAL
YIELD LINES
~
AREA'3
v = ruA ROOF PLAN
P
U-J_V
4 _ _ _ _ _ _V_
2.."....-
COLUMN. ELEVATION
Figure 4-28
Typical column loads
4-92 .'
_
e
e Table 4-8
Deflection Coefficients for Interior Panels of Flat, Slabs
center or Panel
-
C•
"'"
•
•• 1
•••
0.3·
•
•• 1
•••
D.]·
0.00289
0.00200
a.o0ll3s
0.003011
0.00173
0.00100
0.001135
0.0030"
O.OOt73
0.00100
0.0301
0.00189
0.00120
0.0378
0.00262
0.00155
0.00085
0.00230
" 0.00131
0.00057
0.00020
0.00327.
0.002311
0.001113
0.00080
0.00321
0.0022.8
0.00137
0.00015
. 0.00Q99
0.00040
0.00008
0.00005
, 0.0030'-
0.00216
0.00129
0.00071
0.00299
0.0021_
0.00121
0.00069
0.00058
0:,00018
0.000011
0.,'
1.00
0.00581 .
0.004.111
1.25
0.00420
1.61
2..00
LlB'
Midspan or Short Side
0.3·
•
oiL
.'
or Loa8 Siele
K1dapan
.co
.....
e
;
0.2 .
.'
..
- ..
..
'
'
• , Valu•• for ell = 0.3 ore •• tra~al~i'd,
.
,
,
.j:-
,
...'"
.'
,~
<
.,.
,
,.. . [Q] ' .... I
I
Cs
+
.~
c
.!:.. ~I H
+
Cl
.1 ,_, ~,.
COLUMN CAPITAL
rL
u
.
. . . . . . -.' '.-~
1-
I
- , ',1'. '
.
+ Cc
-
::t:
:'
-, [2] I I
[2]-' "
I
'I
.... ,
,~
i.
I L
..
J,-'
I
'
•
TN 5-l300/NAVFAC P~397/AFR 88-22 DESIGN OF LACED ELEMENTS 4-32. Introduction r
1
The detonation of an explosive charge c Lose to a barrier produces 'a nonuniform, h~gh intensity blast load which ~cts on the ,barrier for a,comparatively short period of time. The concept of lac,ing reinforcement (Fig:-4-4, and 4-5) has been developed for use ,in protective struc~ur~ssubjected t9 such loads. Lacing maintains the structural integrity ofa~bar~ier' and:permits it to attain large plastic deflec~ions. ' " Extremely'high·pressure concentrations are caused'by~close~'in detonations. These concentrations cari produce local (punching) failure, of ,an element. However, ,with ,the use of lacing" die, high shears, produced- Ln. the vic'inity of' these pressure concentrations are - q'arisferred to, other .areas of the element where the applied blast loads are, less seyere. In effect ,"the, lacing tends' to spread out the effects of the non-uniformity of the loading and permits the, use of an average blast load over th<; entire surface are a .o f the e Lemerrt .: In addition, lacing is required in those .e Lemant s where ,large,'deflections are desirable. 'In these cases" the lacing not only resists the high shears" produced but also maintains the integrity of the severely cracked concre~e between the tension and compression reinforcement during the latter stages of deflection. The primary use of laced elements i~ to resist the effects of explosive charges located close to 'barriers. The minimum separation;distance between the charge and the laced element i~ given in Section 2~14.i.l of Chapter 2. It should be emphasized that these separation distances are the minimum clear distance from the surface of the charge to the surface of the laced element. The n~rmal scaled distances RA (center of charge to surfac~ of barrier), corresponding to these minimum clear s~paration distances are equal to approximately 0.25 ft/lb l/3. ' A laced element may be designed for limited de f Lec t Lons (less than' 5 degrees support rotation), large deflections (up to 12 degrees' support rotation) "or controlled post-failure fragments depending upon·the protection requirements, of the receiver system. 'The stresses developed !n the 'reinforcement' is, 't function, of the deflection attained by the element. ,TIle type of cross-section which determines ~he ultimate moment'capacity of'the 'r~inforced section is also a function of the deflection' but, more importantly, is a function of the elements brittle mode response. High intensity blast pressures ca~se direct spalling during the initial phase of an element's response: Therefore, type III cross-section will usually be available to provide,moment capacity as well as the available mass to resist motion. ' I :. "
a
Single leg stirrups may be somewhat more economical than ,lacing as'shear~ reinforcement. However, in many design situations, the use of lacing reinforcement is mandatory. When explosives are located at scaled distances less than 1.0: lacing must be used; single leg ~tirrups are not effective:f2r such close charge locations. Also, the blast capacity of 'laced elements are-greater than corresponding (same concrete thickness and quantity of reinforcement) elements with single leg stirrups. Laced elements may attain deflections corresponding to 12 degrees support rotation whereas elements with single leg stirrups are designed for 'a maximum rotation of 8 degrees. These non-laced elements must develop tension membrane action in order to develop 4-94,
"
TN 5-l300/NAVFAC ,P-397/AFR ,88-22 this large support rotation. If support con1itions do not permit tension membrane action, lacing rein~orcement must be used to achieve large deflections. The design of concrete elements subjected to blast loads involves an iterative (trial and error) design procedure in which the element is assumed and then its adequacy is verified through a dynamic analysis (Chapter 3). The design of'laced elements for limited deflections· is performed in much the same manner. However, the design of laced elements for large deflections has ," unique features which permit the formulation of design equations. Since a laced element is subjected to very'short duration blast loads, the actual pressure-time relationship of'the load need not be consl-dered. In fact, the actual duration of the load need not be considered at an: The load may be taken as an impulse (area under the pressure-time curve), that is, the' entire load ia applied instantaneously'to the element.' This aasumpt Lon results in an insignificant error since' the time for the element to rl!ach the maximum deflection is large in comparison 'to the actual duratiol\ 'of the load, Secondly, the elastic portion'of the element'a resistanc:e-deflection'curve need not be 'considered. This assumption will s l so result in a ne'gligible error since the plastic portion'of the curve is many times that of the elastic portion. Lastly, laced elements must be symmetrically reinforced which greatly simplifies the expressions for an element's capacity. 'These feature's permit the formu~ation of design equations and design 'charts which are used to design laced elements for large deflections and for the preliminary design of laced elements for limited deflections. This section includes the design of laced elements for ductile mode response. The brittle mode of response including 'the occurrence of spalling and the design for controlled post-failure fragments are presented in subsequent sections'; The interrelationship of the parameters involved in ,the design of laced elements is illustrated in the'idealized resistance-deflection curve shown in Figure 4-29. 4-33. Flexural Design for Large Deflections 4-33.1. General The basic equations' for the analysis 'of the impulse capacity of an element were derived in Chapter 3. For a two-way element-which exhibits a post: ultimate resistance range and is designed for large deflections, the 'response' is: i 2 b 2'"u
- r u Xl +
'"u
r up
Xl)
4-95
'"up,
The response equation for a one-way element, or a two-way element which does not exhibit a post-ultimate resistance range is: 4-96
applied blast impulse load 4-95
TH S-1300/NAVFAC P-397/AFR88-22 effective unit mass in the ultimate and post-ultimate rang. . es, respectively ~
unit resistances in the ultimate and post"ultlmate'ranges, respectively deflection at partial failure maximum deflection The above equations give the impulse capacity "of a given structural element. Use of'such equations for design purposes is not practica~ since the procedure would involve a tedious trial ~nd error,design .. 4-33.2: "Impulse Coefficients
,
.
Equations suitable for design are o~tained by, substituting the general expressions from Chapter 3 for .t~e effective masses (~ and ~p)' the ultimate resistances (ru and r up) and maximum deflections (Xl' and~) into Equations, 4-95 and 4-96 The. reSUlting ,equations take the form:, "
..
,
4~97,
- C
height of the element
H
horizontal reinforcement ratio distance from the centroid of the compression reinforcement to the centroid of the tension reinforcement dynamic design strength 'of the steel impulse coefficient
C
To illustrate the method used to obtain the impulse coefficients, consider a two-way element (roof slab or wall) fixed on two adjacent edges and free on the other two. The yield line Locat.Ion is defined by y and' L < Y < H. The solution desired is for incipient ,failure (deflection ~) of a spalled section (cross section type III). From Chapter 3 the equations for the resistances, deflections and effective masses for this two-way element are as follows:
1. Ultimate unit resistance (Table 3-2)
ru -
where (eqs. 4-18 and 4-19) b
4-96
TH 5~PO(I/NAVFAC' P-39UAFB. 88-22"
and Pv is defined as the vertical reinforcement.ratiojon each face. 2. Post-ultimate unit resistance (Table 3-4) '.'
3, Partial failure deflection (Table 3-6)
Xl - L tan 12° 4. Ultimate deflection where
. . , '-!-.
Xu -
(Table 3-6)
ytan 12° + (H - y) tan y -
,(
y _ 12°' -
tan ~l [
tan 12 ° -'~]
.,'
y/L
5. Effective unit mass in the ultimate range ~
- (Kut)u m I
(Kut)u is from figure 3-44 m
de
---
[
1728
150 l-'·225dc,. 386, (10- 6? .';
6. Effective unit mass in the post-ultimate ,range ~p
- (Kut)up m - (2/3) m .
the units used are: inches psi-ms .MVN• MVp'
,
,
.
in, -lbs/in ..
.
f ds' r u' r up
psi'
As
in. 2
m.
p~i.-ins2/i~. '
r
~. ~p
Substituting into Equation 4-95 i 2 - 2 X 225d (Kut)u b c
[ 4-97
..
TN 5-1300/NAVFAC P-397/AFR 88-22
12° + (H - y) tan 1 -
Lta~120]
4-98a
Factoring
(H - y) tan 1 (y-L +
4-98b tan 12°
Dividing each side by PH the horizontal reinforcement ratio, and rearranging
.-
Pv
3(Klli)u 2
[
y
L - -
H
Pv
- 287(
Y
tan 1
H
tan 12°
]
'3. 33(Klli)u(L/H) )
+
4-98c
.
(y/H) 2
PH y
4.70 (Klli)u
+
+ (1 -
H
, ib 2H ' 3 PH d c f ds
) tan 12°
L --- + 4.70 (1H
[--H
Y ) tan 1 J
4-99
H "
where , . 'Y -
12 ° - tan - 1 (
0.2126
) .
4-100
y/L' The solution for partial failure (deflection Xl.) for,the above two-way element is obtained in a similar manner. Substituting the general expressions for partial failure into Equation 4-96 yields: ib 2H PHdc 3fds
Pv 957 (---) PH
( L/H
)
, 4-101
(Klli)u ( y/H )2
Equations 4-99 and 4-101 can be rewritten as: ib 2H . 3 PH dc f ds
- .C1,
4-102
4-98
TK
5-t300/NAVFAC'P~397/AFR
88-22
,:
4-103
PH d c 3fds
where the right-hand side' of' the equat Lon.d s 'design~ted as: the impulse coefficient: The impulse coefficient Cu h:used for incipient failure design (maximum deflection equals Xu) whereas Ci is,for partial failure design (maximum deflection equals Xl)' 'These impulse, coefficients are a proportional measure of the impulse capa~ity under the' resistance-deflection curve up to the maximum deflection. " " , ,
"
Expressions for the impulse 'coefficients of elements with various support conditions and yield lihe locations have been derived as above. Equations for Cl and Cu for two-way elem~nts" are given ih,'Table' 4-9 and Table '4-10,' respectively. For one-way elements which do not exhibit the, secondary resistance range (Xl - Xu), the coefficient Cl is equal to'C u' In addition, for a given support condition, Cu for ~ one-way, e!ement./s a c~nstallt vaLue . Table 4-11 gives the values of.'C u for one-way elemen~s.. ' " "; '.,
."
.....: - . : .• , .
. ;
, f" :
4-33.3. Design Equations for", Deflections 'Xl and -:
. "
,
,
,"
.
,.,
. ' •.
"
'"
~
,
,
For design purposes, Equations 4-102 and 4-103 can be, rewritten as:
, For a tW6-way element Cl and Cu are,functions;ofslJ,pport,cohditions, aspect ratio, ' yield line location" r'e Lnfo rcement; ratios and rhe.: Load-mas s factors: It was shown in Chapter 3 the'(KLM)u for a two-way element varies with the yield line location ratio y/H or x/L. further'!'ore, it was shown that yield line location ratio is a function of the span ratib L/H and the moment'ratio [(MVN + MVp)/(MHN + MHP)]' Since the cross sectiohs used'for large'deflection design are equally reinforced "on each face, the moment ratio is " in effect, the ratio of the reinfbrcement'r~ti~ PV/PH' Thus it can be'se~n that the impulse coefficient~ are solely, functions of L/~ and PV/PH for a given support condition, ' To. fac.l.litate the design procedure, charts,have been constructed for the impulse'coeffi~ients Cl and' Cu for t~~-way,eiem~nt;a~ a function' of PV/PH and L/H. These curves,for various suPPo~t cond.l.tions are given in Figures 4-30 through 4-32 for Cl and Figures 4-3) throlJ,gh 4-3S'~or Cu'. ,For' one-way elements Cu Ls a cons cant; (see: Taple' 4-11): ' 4-33.4:,Optimum Reinforcement
.
,
, A prime factor in,the des~gn of' any facility is construction economy. 'Proper selection of section si~es and reinf9rcing steel will result in a'design having optimum ?ap'acity and minimum cost. ' See discussion in paragrayh 4-23.1,. ,
4-99' ,-
"
TK S-1300/NAVFAC
P~397/AFR'88-22
'
To determine the optimum design of any particular two-way structural element, cons~deration must be given to the following: 1.
There is an ideal distribution of flexural reinforcement, defined by the reinf~rcement ratioPy~PH' which ,is independent of ,section depth. This ratio will yield the.maximum,blast impulse capacity for a given total ~oUIii:: of: fle.;{ural rei~forcement PT' •
2.
There is,an ideal relationship,between:the quantity of reinforcement to the quantity of'concrete which' will result in the ' minimum cost of an'element. This relationship is defined by the total percentage of reinforcement,in one face of an element. this total percentage PT is the ,,~of the yerti'i'a~ and ,horizontal , reinforcement r~tios, ,Py, an~ PH'. ~espective~y.
•
. ' I;f
•
•
.
oJ.i'
_
~,
.
'
, ' ••
'
'-.
H.
.
•
.
4-33.4.1. ' Opt'imum Reinforcement Distribution ,
r
'1
f
.
•
"
, •
The blast impulse capacity of an element varies with the,distribution of the reinforcement even though the total amount'of reinforc~ment and the concrete thickness remains the same. This optimum reinforcement 'ratio varies for different support conditions as a function of the aspect, ratio L/H. In' addition, the optimum ratio is different'-for' partial' faiiure 'and inc'ipient failure design. J' , ... ', . I
'/
To illustrate the determination of the' optimum reinforcement distribution ratio PV/PH' consider a two·way panel fixed on three sides. The panel has an aspect ratio L/H equal to 3 and a total percentage of rei~forcement'PT equal to 1 percent. For various values of PV/PH' the impulse capacity can be determined for both partial and incipient failure design from Figures 4-31 and 434, respectively. _ If (ib2H)/(dc3fds is plotted versus PVPH the resulting curves are shown in Figure !,-36:, The ideal PV/PH occur's at the maxiJ,num value of ib2H/dc3fds and is indicated,o~ ,the illustration as 1:58 ro~' incipient failure ana 1,93 for partial failure d~~ign: ~ncreasing or decre~sing the ,S~tal amount of steel PT' will shift-th~ curyes'up 9r down but not eff~ct the optimum Py/PH ratio. This optimum.PV/PH ratio for oth~r L/H ratios and support conditions are ' determined, from similarly const.ructied curves. . '.. '. '
,
The optimum ~al?es'of PV/PH for vario~s sup~oftcJoRditions are plot~~d- as'a, function of the'aspect ratio. Figure 4-37 gives the optimum reinforcement for partial' "failure design, while Figure 4· 38 'giv~s the'· optimum :ratio for' in~ cipient failure design. ',-
,.-
.
-'.
'.
'
The optimum reinforcement ratio for partial failure'design always results in , positive yield lines which bisect the 90 deg.~e angle. f, at the' corners of the .. . ..,. .. element (45 degree yield' lines). for all support;conditions. Consequently; all supports reach the maximum rotation of 12. 'degrees' simuitaneously and . they are. ' , 1· _ all on the verge of failure. Therefore, the' optimum condition for partial failure is a particular case of incipient failure. T9is condition is evident from the common point on Figure 4-36. It can also be seen from this 'figure that atpV/PH r~ti'?s other than the common ~oint, par t La l failure design is • more conservative than incipient failu~e design which includes the post ultimate range,' The optimum PV/PH ratio for partial' fail\l~e design maxLmf.z es the impulse capacity up to Xl leaving'no reserve capacity (post ultimate . ~
4-100, '
TM
5-1301~/NAVFAC
P-397/AFR 88-22
rang'e) . Therefore, at this ratio, the capacity is ~um,eric'ally equal to 'that for incipient failure design. . Whilei there is. no_ quantitative advantage to , . - " . ' . -. . optimum partial failure design over incipient:fa~lure design, there is a, qualitative advant age, The elements rema,tn J,.ritact' s Lnc'e, all suppo,~s are on the verge of failure as opposed' to ,optimUm incipi,,'nt failure where some supports have failed and the remaining supports are on the verge of failure. In this Lat t er case '. t;he,re ,is unlmown sec9lJ:dl\ry cracki'~g which is n~t accounted for in th!, ~esign., ' , ' , " '. '0
•
As previously expl8i~ed, 'incipient::failure design includes' the dlp~city fro{,t two:w~y action of an element up topartral 'fallure'Xl, an~ the capacity of one-way action up to LncLp Lent; failur,e ~. 'Except" as exp,lained below, the optimum reinforcement ratio' 'f.or it1cipient failure design results from' maximizing the capacity due to one-way action after partial failure (post-ultimate range). The, resulting optimum rei~forceDierit rat'ios for incipient failure' design produce various yield ~ine.configurations depending.upon the support condi~ions., For !~~r'edged'fixed, the,optimum reinforce~erit'ratio is 0.25 and 4.0 for aspect ratios less than and greater than one, respectively. This distribution. maximfzes the post ultimate one-way ac t Lon inthe shorter. directfon. 'Fo'r two edges fixed, rhe increase, in '.capacity due to, cantiiever action in thlji pcst; ultiIila~e ,range is less than th~ ,Aecrease in capac Ley 'of the ultimate range. Thus , for ,these elements, the capacity cannot; be Lncreased above that for partial failure, and the optimum ratio for 'incipient failure design is the same as for partial failure design (45 degree yield lines). For three edges fixed, the post ultimate range capacity is due to either cantilever a~tion in the,yertical direction or fixed~fixed'beam action in the horizontal direction. ' In regions where the,post ultimate range consists,of cantilever' action (L/H ratio, in the immedfate vic,inity of'"2 and L/H ratio , greater than 4) the optimum ratio is the same'as' for pa'rtial : failure . For L/H ratios 'less than 1. 5, the post 'ultima,te range, consists of fixed-fixed beam action and, therefore; the optimum ratio is'equal to 0.25. Between theseL/H regions, neither behavior dominates and'the resuiting optimum PV/PH ratios maximizes the combination of ultimate and post:ultimate rang'!' capacities" ~
."1"
,
. '
....
,
'"
I ,
4-33.4.2. Optimum Total Percentage of Reinfof~ement The optimum total percentage of reinforcement PT gives the relationship between the quantity of r~inforcement to the quantity of concrete which results in the minimum ,cost of an element. ,The total percentage of reinforcement in one face of the, element is defined as: "
4-106 The o~timum percentage of reinf~rcement depends upon the relative costs of ,the concrete and reinforcing steel. Based on' the average costs of concrete and steel, the optimum percentage of reinforcement PT has been determined to be betwe~nO.6 and 0.8 percent, with 0.7 being a reasonable value to be used for design. However, for'large projects, a detailed cost analysis may'result in a ' more economical design. In the usual design situation, the optim~ PV/PH ratio is first determined based 'on the support conditions and aspect ratio. Knowing this ratio, Clor Cu is determined and along with the given values'of i b H, f ds' Equation 4-104 or 4-105 results in: 4-101,
TK 5-l300/NAVFAC P-397/AFR 88-22 4-107
With the known, values of PV/PH.and the optimum total percentage of reinforcement equal to 0.7, the required quantity of hoxizontal,reinforcement'PHis calculated. The required thick~ess,~f'the element is then calculated from ,Equation 4 - 1 0 7 . " ' " "
"
In some design cases, it' may'be desirable to "reduc:e the c'oncrete thickness below the optimum thickness. The quantity cif reinforcement"in eX:cess of the optimUIII PT must be provided to obt;ain the ne,cessary imp~lse capacity. The , cost increase is small for total percentages of ,steal in the vicinity of ' the . optimum value of PT' In,fact, the 'use of PT '~~ual to 1 perce~t will result .~n, a cost increase of less than 10, percent.' Beyond 1 percent reinforcement, th~ cost increase is more rapid. However, except for very thin,elements, the'use of reinforcement in excess of 1 percent is impractical since the required r " , details cannot be' maintained ,with 'such"large 'quantities' of reinforcing ~tee~.~. For thick walls providing even the optimum PT of 0:7 percent maybe impr.acti'" cal and PT ~ay have to be redue,ed to as low as 0: 3', percent (mi.nimum rei~fOJ;ce - ", ment of 0.15 percent in each direction) in order to permit placement of the reinforcing steel. The total reinforcement PT m.ay also be less than optimum' if a minimum concrete 'thickness is required to preve~t fragment penetration. When the minimUm quantity of reinfor"ement ,is,pro~ided whether for strength, or to satisfy minimum requirements, the res~lting cost may b~ far in excess of optimum. ' . , ' ' , In some cases of incipient faiiure des Ign, the optimum reinforcement ratio Pv/PH is equal'to 0.25 or 4.0. However, in most cases, it is'impractical to provide four times as much reinforcement 'in one direction as in the other direction.' Since the minimum required percentage of reinforcement in a'given direction is 0.15, the 'orthogonal direction would require 0.6 percent'for a . total percentage of 6.75. ' Although this percentage is 'approximatelY,equal to the optimum percentage of 0.7. it may still be impractical in all 'but thin walls.: Consequently, Lri such design situations" a trade of'f between optimUm rein%ofceoient ratio Pv/PH and tne optimum total percentage reinforcement 'PT ' must be made for an economical design. , ' •
4-33.5
, Design
Equation for Deflections Less than Xl or
Xu
For certain conditions, it is sometimes desired,to design a structural element for 'maximum deflections other than p~rtial failure def'Lec't Lon Xl or incipient failure deflection For those cases, the impulse coefficients can be scaled relative to the deflections.
Xu.
For a maximum deflection 107 becomes
Xm
in the deflection range Xl <
Xm
<.~,
Equa~ion
4-
' ,
'4-108
'4-109
4-102
TK 5-1300/NAVFAC P-397/AFR 88-22
.
For a maximum deflection corresponding to a support rotation greater than 5 degrees, but less than Xl' Equation 4-108 becomes ,
PH dc
i 2H b
3
,
4-110
C'lfds
whe're C'l . ~.
Xm
4-111
( - - ) Cl
Xl .
The ~ptimum Pv/PH ~atio for ~ given elemen~ is a constant for any deflection less than partial failure deflection~l'. and 'is determi?~d from Figure 4-37. In the deflection range Xl < Xm< ~ the optimum,PV/PH ratio varies with the maximum deflection. However, for design purposes, the values from Figure 4-38 for incipient.failure may b~ used. 4-33.6. Design Equations for' Unspalled Cross Sections The impulse coefficients derived,above may.also be used 'for type II or unspalled cross sections. However, the'general form of the equation is' slightly modified to account f~r ~he'ch~nge in the physical properties of the cross section. For a tyPe II cross section, the full thickness, of concrete element is included in calculating the effective mass. Thus, the design equations, for the impulse coefficients of unspalled sections take the form:
.'
..
4-112
4-113 where Tc is the total thickness of the concrete section. The optimum reinforcement ratios and the impulse coefficients are the same for spalled and unspalled cross sections . The design procedure for unspalled cross sections is ,very similar to the procedure described in Section 4-33.4.2. The total thickness of ,concrete Tc caribe expre~sed in terms, of d c by approximati.ng the value of. d'. The. value of d'.. can be estimated by determining the required concrete cover and assuming the reinforcing bar sizes. 4-34. Flexu~ai Design for Limited Deflections In the design of elements for large deflections, only the plastic range behavior of~he element was consider~d, since. the ,capacity due to elastoplastic behavior is re~atively small. For elements where support rotations are limited 'to 5 degrees or less,the elasto-plastic range is a significant portion of the el~ment's total capacity as well as of its deflected shape. Therefore, it.mus~ be ~ncluded in the determination of the response of such elements. ..
4-103
TM 5-l300/NAVFAC P-397/AFR 88-22 The blast impulse capacity of an element whose maximum deflection is less. than or equal to 5 degrees was given in Chapter 3 as i 2H b
r u XE
2ma
where
2
+
ma
ru
(Xm -. XE)
4-114
~
ma
average of the effective elastic and plastic unit masses
XE
equivalent elastic deflection
This is an equation which is suitable for analysis rather than design. Impulse coefficients could theoretically be derived in a similar manner as that for large deflections'. However, 'the equivalent elastic deflection cannot be defined by a mathematical expression making the determination of impulse coefficients' for the various support conditions impractical.' ' ","
The design of an element subjected to an impul~e load' (short duration pressure-time load) for limited deflections is, accomplished using a trial and error procedure. An element would be'
Design for Shear
4- 35.1. General After the flexural design of an element has been completed, the required quantity of shear reinforcement must be determined. -: This 'shear' reinforcement insures that the desired flexural behavior in the ductile mode will' be attained. The design of the lacing reinforcement has been discussed in previous sections. This section is concerned with ,the 'determination of the shear stresses' and forces to be used i~ the design equations. Shear coefficients can be. derived in a manner similar to that used to derive the impulse coefficients above. The equations ' for support shear given in Chapter 3 and for the ultimate shear'stress given in Section 4~27 show that the shear reinforcement is'a function of the resistance of the element, and not of the applied load. The shear forces'and stresses vary as the ultimate unit resistance, the geometry and yield line locations of the element, and the section depth. If r u is evaluated and substituted into these shear expres,4-104
TK 5-l300/NAVFAC
P-397/AFR88~22
aLons , it cap be shown 'that the ultim":te auppor t; shear Vs car{'be represented as an equatifon in the, 'general fOrm ". .,
•
,-,
I'
4~115
L
and tpe ultimate shear 'st;J;essa1; dist~~~e .:dc fr:.oin the support~s
.
" , "
where C
.p
,
, .,
•
shear co~fficient'
'"
,, 4-116
.
~
" !
j ••
,
,
"
f~exural rein~orcement ratio
,.
J
,
•
I
f d/-, ;dynami'c' de~'ign stress 'of the flexural reinforcement The shear;coefficient is differ~nt for each 'case and also different for oneway and two-way eleme~ts, Specific values ~re indicated in ~he following paragraphs of iliis section: ' ',. " y
•
0:
,
4-35.2. Ultimate Shear Stress
.
','
..
., ",
4-35.2.1. One-Way Elements
The ultimate shear stress v u at distance d c from the support for element is
a one-way 4-117
where Cd is the shear coefficient and a function of the ratio of dc/L. Values of Cd are shown in Table 4-12. 4-35.2.2. Two;Way Elements , . ,
",
:' ...
~.
The' ultimate shear .. stress v,J;" in the horizontal' direction (along side H) at a distance d c from the support for a two-way element is given as " 4-118 vuH - CHPHfds c
and in the vertical direction (along side L)" 'a~ 4-119 where., CH and Cv are the horizontal and vertical shear coefficients, respectively. The shear coeffici~nts"given in Table.4-l3, vary a~,dc/x or dc/Y for the triangular sectors and as x/L and'dc/H or y/H and dc/L for the trapezoidal sectors. The solution for the shear coefficients is presented graphically in Figures 4-39 through 4-52. The shear coefficients for the triangular sectors, can be read directly from either Figure 4-39 or 4-40, since the yield line location is the only variable involved .. ' Plotting the- .'shear, coeffi.~i'llits. for the trapezoidal seCtors for a ' . • t . • • • particular support condition yields a family of curves. That is, the shear '~....'
4-105
TN 5-l300/NAVFAC
P~397/AFR,88-22
coefficient is plotted versus dc/L for various values of y/H, (or dc/H for various values of x/L). The maximum value ,of the shea~ coefficient is, ' different for each curve of y/H or x/L' ~nd occurs at various values of dcIL or dc/H. Therefore, these family of curves overlap and accurate interpolation between curves is diffic~lt. ' Using a method of coordinate transformation, the family of curves has been reduced to a set of curves with a common maximum point defined (using the horizontal shear coefficient as an example) by CH/CM - 1 and (dc/L)/(dc/L)M 1. The quantides CH and (dc/L)H represent the co'ordinates of the maximum point, on the original family of curves for y/H and x/L. The left-hand portions of the curves become identical, and accurate interpolation in the right-hand portion iS'now possible. This transformation results in two figures to define'the shear coefficient for a particular support condition and yield line pattern. . ' The above sets of curves are presented in Figures 4-41 through 4-52. When using these curves, the shear parameter curve, for ,the applicable support conditio? is entered first'with the value of'x/L'or y/H to determine CM and (dc/H)M or (dc/L)M". The second curve is then used to determine ,CH or Cv. f '
.
:
It should be noted that when designing two-way panels for incipient failure;: the shear stresses in the post-ultimate range must also be checked using the equations for one-way elements. " 4-35.3. Ultimate Support Shears 4-35,3.1. ,One-Way element The
~ltimate
support shear Vs for a one-way element is 4-120
where Cs is the shear coefficient and is a constant for a given support condition. Values of Cs for several one-way elements are given in Table 4-14. 4-35.3.2. Two-Way Elements For a two-way element, the ultimate support shear VsH' in the horizontal direction (along side H) is repres~nted as PH d c 2 f ds
4-121
L
and VsV' in the vertical direction' (along,side L) is • Pv d c 2 f.ds'" H
where CsH..a nd Csv are the horizontal and vertical shear coefficients', ,respec-' tively. For a given support condition! these. coefficients vary" as the yield 4-106
TH S-13ClO/NAVFAC P-397/AFR 88-22 line location ratios x/L or Y/H. The shear coefficients are listed in Table 4-15 and for the trapezoidal 'sectors only are plotted in Figures 4-53 through 4-56 for various support conditions.
,.,",
.;
-;
": -,
>---,-,-
"
4-107
\
...
'.
e
.'
w
U
Z
'I
<[
l-
(/)
(/)
w
~
POSTI ULTIHATE:
LA.. TlPIfAT£
, ELASTICI
1
II
I
I
I
I
1
1
'I
I
1
1
I
I«- INCIPIENT FAILURE
!
I
,X( 9= 2-)
XE
TCTAl
FAILURE
X( •• 5")
X,
Xu
DEFLECTION CD
e
PROTECTION CATEGORY
®
<9
$HE:LTER
.~ RJER
PROTECnVESTRUCTURE
PRESSURE - TI I/!E OR IMPULSE:
STRUCTVRE-LOAD SENSln\1TY
cec (-IN
DESIGN RANGE RESP
SE CHARTS DR HERAnON PROC(DUR
DESIGN METHOD
IMP
SE (QUA TIONS
IMPULSE
CR C
LARGE
LlHIT C
DEFLECllON CRITERIA
QUAlIONS
ARTS TOTAL FAILURE
BENDING'S ~AA
DUCTILE MODE Of RESPONSE
POSTCRUSHING
BRITTlE MODE OF
[RIll
CROSS-SECllON TYPE
SHEAR REINfORCEMENT
Figure 4-29
'
U OR
m
f'dy+1I4Cf'c1u-fdy>
FAILURE FRAGMENTS
DJRECT SPAL ING
RESPONSE
DESIGN STRESS
SCABBING
1Il
1/2Cf'ctu+ f'rI>
,
LACING
Rele.tlonship between deSign pe.re.Meters for le.ced etenerrt s
4-108
~
2000
,~.,...
,,' I
,
;;
I I'
I
rrrr I'
I
1000." '""
,I" I"
1 II
i!III'
"
I
I
I
rrt
I,
'" II,
'II'
,
". ,,',
I
"
I
Ii:
I
I"
:li'
II'
,
-:-++,
I
'," "
"i' ",:
>iii lrt!
I
"
I
".
'"11"
I I
, , ,,. 'I
I
UI
I
--'~
, t '
.,
"1"1
I I ' i II
II
'"
I!
1
I,'
,
"
1
," "i'lo" '1Iil':O
Iii
I
II
l-T '
" I: !
I
"
,
0.;
,
, , , ,
1""1
i
I
':
I,
, I
, II1I 'Iii
I
" I
I
'1,1
I
700
500 N.
•5O
<,
400
N
•e
•...
300
200
.. "
;"
,.j Ii
"
."' i;L;1 I
I
'!lId
Ilii'
100
.""
I
••
'1,'1
.
:0 "';>.
..
·~0 .
0"
, ,",
," , I'"
I I!
I'
011 I 'I
'I
"
1
.
'
I
I " I ".
",.
, I
"
III,
,
"
I I I
, , ,'
,
"
,
,, :
o "
I
III
III I I' I I II ,
,.,
"
,".'
I ' " """ "
""",,
.. I'" "",.,,, '1"1"11 , ;I 11\111'11 1'1'11' 1':1 III I 11"1'
'I"
"
I I
0
0"" ," "" i
'" I
"" ,'I "
::H-* I
I
111I
I
'0"
-=
70
50
.2
.4
.3
.5
.7
1.0
2.0
RATIO. OF VERTICAL TO HORIZONTAL flEINF.
Figure 4-30
Impulse coefficient Cl for an element' with two adjacent edges fixed and two edges free
4-109
3.0
4.0
5.0
3000
N. c <, N
400
co
E co
Q,
.. - -- -,._--
. -. - --"""1"
.. _-. _.
__
0 •• "._-
..
-:~::.::r=:.·,-~ . _...t -_._--,- : - - _...
- - - - _.•._ ..• ,-_... _ ...•••.•. -
•
! .• - • . • . • . . .
-'_··_·_·,_··:'::-:~::::::1:::::·:.:::::..._..:j-=- -- . . - :'1::. _ . .. ~- "I' .'~._4 0 •••• - . • • • •
'_~':~"".:.:::."._ .. - - '.
.l:-
.."
::1 :: :": :::.. i
.::;_:':~:':':' . ;.._
-_..
.:"1:
,.
-1. .
:~·~~=-l·:~·_. ~ ~ ! "_ _,-,~; =.:: .".• ".-.-.- -'j .- .. -~.- - ··-l-~-'·'·-·'· . t- ••••• _~: r·:~~:~..::;,~::.,~= ~:::.; 0
•• '
••
....
... ".-.
..,
~-~ ~ I ~~.~-~! ~~~ I ~~.:..:
.2
.4
.3
.5
;~~ ~~===::~::~:.::i:~~~;-:.:=j~~~'~:~~~~i~~~~~ :~~~i·~~~l~~~~~,~?:~
.7
1.0
2.0
3.0
RATIO OF VERTICAL TO HORIZONTAL REfNF. Figure 4-31
Impulse coefficient Cl for an element with three edges fixed and one edge free 4-110
4.0
5.0
N •
.E
......
...E
N
e
...<>. U
,
. .. ,
.. .. "'":." ,....... ........ ::' I.:.:::::.: c i...... . .. ", I." ..., ......... ,,--: :"'.C·.' .: ... : ...... ,
.......
. ""
'
.... ,
.2
.5
.3
.7
1.0
RATIO OF VERTICAL. TO Figure 1,-32
HO~IZONTAL
2.0
1---'
3.0
REINF.. Py I PH
Impulse coefficient Cl for an element with four edges fixed 4-111
..
.--,:: . : : :": :::':,":: ::l:':. .::· ::: .. ,..... I, ..·.., . . .... ...,::::::-. :·:T::: .... :::
4.0
5.0
-:s.... • I •
.2
.3
.4
.5
1.0
.7
2.0
3.0
4.0
RATIO OF VERTICAL TO HORIZONTAL.REINE. Py'/PH Figure 4-33
Impulse coefficient eu for fixed and two edges free
4-112
~n
element with two adjacent edges
tlO
2000
1000
.
100
..s
.....I
e
.-
•...
500 400
300
2
U
200
..
, ." .j,
"
., , ;
100
,
,
I
II
I
I
" , "
,
f
I I
II I
I I
'"''
1111111 I I I
, , I
I
I
70
50
.2
.3
.4
.5
.7
1.0
2.0
3.0
RATIO OF VERTICAL TO HORIZONTAL REINF.. Pv /PH Figure 4-34
Impulse coefficient e u for an element with three edges fixed and one edge free
4-113
4.0
5.0
2000
,
II' ,.
,
I
1000
I
I
I
,,
I
I,
i
700
500 .5
...... N
400
•E
""
rz-!J
. 'b!J ""
I
•Go
\ .'=>
~
N .
300
.
~
o
--
200
'" ", ,I"
'1"
," ""
"
" , ,
I"
'"
I'!I
'"
,,'
+:
.,
r •,
, ,,
II' I'
'"
I
I
I
,
,
"
100
70 .~-
50 .2
.3
.5
.7
2.0
1.0
3.0
RATIO OF VERTICAL TO HORIZONTAL REINF.. Py /PH Figure 4-35
Impulse coefficient four edges fixed
4-114
cu
for an element with
4.0
5.0
e
e
e THREE EDGES FIXED
L
- H "- 3.0.
P,- " 1%
2.8
-s -.,..,.
.~
':'
...,
..-
'"
.
Pv II;'H"=
1.93
2.6
~
.
%~ 0I.a
-
0
"
2.4
DEFL'ECTION
I
XI
2.2
2.0 0.25
0.5
0.7
IrO
2.0
3.0
Pyl PH Figure 4-36
Determination of optimum ratio of p /p for maximum impulse . V H capac1.ty
4.0
5.0
PH
'm'77">7r.n "7'7'
:JH
I.
.1
L
TWO EDGES FIXED
L
THREE EDGES FIXED
FOUR EDGES FIXED
TWO EDGES FIXED
3.0 :1.'
,
, , , , , ,, ,, , , ,, , "
2.0
, , , ,, , , , ,,, ,. , , , ," ,,
'" ,
,
• "
. ,• ,"• , '" , ," ", ,
'.' '.
III
li'l
'i'
"
iii'
% Q.
......
",,
> Q.
,
IL
~
C(
II:
2
::;)
2 ~
IL
0
,"'!
;,.
j',
"I"
, " " , "
'll
'" ,
I'i
'11
0 0
"" "i
," ,
. '"
,
Ii_
t r
"
.. '" ,
""
'
,
IIII
"
•
(I,'
f.'1
;,1,
I I I",
,
FOUR EDGES FIXED
r :
,
'" ,.,
, ,
, ,, ,, , , , , ,, , , ,, , ,,, . , , ,, ,, , , , , ,, , ,
,
,, ,
,
"
I· t,
".,
I);
,I'
, ,
I:.
.'
.'
.,
It:1
'i'
",
"
. ,, "
Ii-
, ,
""
"
"
. . .' r . .,
,
1.0 0.9
0.8 0.7 0.6 0.5
0.3
.
", , ,
t
0.2
,I "
" t
, t
"
,
I·'i ,
!
, ';1'
"
"
IIII
0
I
"
!I
,
t
r
,
,
,
, ,
III
I; ,
" ,
, I II "
,
r
,
'"
"
,
I 'I
"
,
'Ii II, I
,
2
. ", , p
·1, '
"
I' "
3
,
,
..
,J'
,
, " " " ,"'" I'
,,
,
,
"4
"
,"
,
"
I
5
RATIO OF ELEMENT LENGTH TO HEIGHT. L/H Figure 4-37
Optimum ratio of pv/p for maximum capacity at partial H failure deflection,x 1
4-116
, ,
'"
",, 6
H
L
L
L
TWO EDGES FIXED
THREE EDGES FIXED
FOUR EDGES FIXED
FOUR EDGES FIXED
3.0 TWO EDGES FIXED
,,
,",
, , ,,
,
,
"
2.0
,
z
~
.....
THREE EDGES FIXED·
~
e
0 j:
c
II:
:. :) :. ~
Q,.
0
, , ,, f
.... 0
I
"
I
'
•
1.0 0.9 0.8 0.7
0.6 0.5
0.4
0.2
O~,
3
2
6
RATIO OF ELEMENT LENGTH TO HE:IGHT t LlH Figure 4-38
Optimum ratio of P /P for H incipient failure,V x u
4-117-
maximu~
capacity at
l-,,:~~'/j3
!
:r "
,'- .... ...........
»1
--
" ....-T............ ........
_
........
,.,1
L
"
1.4
. 1.2
1.0 .'
0.8
0.6
0.4
0.2
o Figure 4-39
0.2
0.4
0.6
0.8
1.0
Vertical shear coefficients for ultimat~ shear stress at distance d c from the support (cross ~ection type II and III) 4~1l8
r-,
1\
1 .,JI'",JI'
:z:
/'/
I
I
I
.!.
"
\
\
'~- --<, , ;'
\
\
,'"
-
I
1.4
1.2
1.0
0.8
0.6
-
0.4
"
-
0.2
o
0.2
0.4
0.6
0.8
1.0
de/x" Figure 4-40
Horizontal shear coefficients for ultimate shear stress at distance d c from the support (cross section type II and III)
4-119
1.3
x
~[
0.56
I
/'"
'"'"
'"
1.2 L
1.1 1/2 ~/HS:I
1/2~ dc/HS: I
0.52
1.0
0.9
0.50 O
0.48
0.8
0:46
0.7
0:44
0.6
0.42
o
Figure 4-41
0.2
0.6'
04
0.8
X/L Vertical shear parameters for ultimate shear stress
at distance d c from the support (cross section type II and III) 4-120
0.5 1.0
e
e
e
x
I . 1
.
:J:~/ ./'
L
1.0
0.8 '/L=OIHO
,
0.6
~
.... .... '"
Cv CII
'/L =0.4 '/L=0.2 ._
0.4
'/L=0.6666 0.8
0.2
o
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
dc/H (dc/H).. ·
Figure 4-42
Vertical shear coefficient ratios for ultimate shear stress
at distance d
c
from support (cross section type II and III)
1.8
2.0
1.3
--
:z:
,- -'
>-1
l
0.56
,
,
1.2
,
0.54
1.1 1/2 ~ dell
~ 1/2~
0.52
dell <1
,
~
, O~de/l51/2
0.50
1.0 ,
0.9
O~dell:S;1/2
,
,. 0.48
,
, ...
,
0.8
,
0.7
.0.46
.'
0.44
0.6 ,
0.42
o
Figure 4-43
0.2
0.4
0.6
Y/H
0.8
,
0.5 1.0
Horizontal shear parameters for ultimate shear stress at distance de from the support (cross section type II and III)
4-122
e
e ,,-
x
" " "
e
R
I .
1.0
• .
III
0.8I
...,
.... '" '"
Y/H=O 81.0
0.6 ..
CH CM
.
.
Y/H=0.4
lif'l
0.4
Y/HaO.2
.' .
.r
Y/H=O.666 8 OS
0.2
o
,
0:2
0.4'
0.6
,0.8
1.0
dell
1.2
-
1.4
1.. 6
-,
(clc/llM Figure 4-44
Horizontal shear coefficient ratios for ultimate shear stress at distance d c from the support (cross section type II and III)
.1.8
2.0
1.3
r1 :I:
0.56
I
I
\
I
I
\
1.2
L
0.54 1/2 < dC/H~ I
0.52
1.0
0.50
0.9
( dC/H)M 0.48
0.8
0.46
0.7
0.44
0.6
0.42
o
0.1
0.2
X
0.3
0.5 0.5
0.4
~L Figure 4-45
Vertical shear parameters for ultimate shear stress de from the support (cross section type II and III)
4-124
;t
distance
e
e
e
e
,
xrgr;\ I
\
I
•
\
L
0.8 '/L.=O 60.5
,
~
t-'
'" '"
0.6
~
eM
X/L=0.2 X/L=O.l
0.4 ,"'t
X/L=0.3336:0..4
0.2
o
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
dclH (dc/HI.. Figure 4-46
Vertical shear coefficient ratios for ultimate shear Gtress at distance de from the support (cross section type II and III)
1.8
2.0
I
'::1:
,."
.A, ... ,."
<,
>-I
L
0.28 "
~
..., .
0.27
1.30
1;4:$dc/L:$ il2
0,26
1.25
. V4
0.25
1.20
0.24
II
1.15 O
0.23
I .10
0.22
I.05
0.21 Figure 4-47
o
0.8
0.2
Horizontal shear parameters for ultimate shear stress at distance de from the support (cross section type
II and III)
. 4-126
I.00 1.0
e
e
e
I:·
:I
--
"""',....... -i, ,
,"
.
L
1.0)
0.8l
Y/H= OIHO .Y/H= 0.6
..., ...'"
....
0.6
~*!YIH=O.4
S!-
ee.
III
,
0.4~
.
.Y/H=O.2
,
0.22 .Y/H·0.8
.. '0
0.2.
Y/~=0.9 0.4
0.6
.0.8
1.0 ,.
1.2 .,
1.4
1:6
dell (delL)., t-
Figure 4-48
Horizontal shear coeff"icient ratios for ultimate shear stress at distance d c from the support (cross section type II and III)
G' 1.8
2.0.
~
-,
•
.. 7
I,/~---';'-<
J:
0.28
"
.
.
I
0.27
1.30 1/1~dc/H~1/2
,
0.26
'. .
1.25
1/4S'delt+.;' 1/2
0.25
1.20
OS"dc/H~I/4
CM 0.24
1.15
...
tim:
O~dc/~J/4
0.23
1.10 ~
0.22
0.21
1.05
O.
0.1
0.2
0.3
0.4
1.00 0.5
X/L Figure 4-49
Vertical shear parameters for ultimate shear stress at distance d c from the support (cross section type II and III
r
.4-128
e
e
e
t-!-
"
:z:
/
/
)C--
.
-.,
,
L
..
1.0
~ 0.8 '/L=0800.5
.
....,
'/L=O.3
-
0.6
.... '" '"
EY e
'/L=O.2
lll
-
..
0.4 '/L=O.l ..
.
,.
~.
0.2
. . ,'/L=O.4 "/L#o.451
o
,
0.2
0.4
0:6
0.8
1.0 .,
dc/H
1.4
1.2 "
1.6
~
-(dc/Hlw Figure 4-50
Vertical shear coefficient ratios for 'ultimate shear stress at distance d c from the support (cross section type II and III)
1.8
2.0
,.
......-
:r
.....
--
-- _ ....
".-
....."
>f
............
L
0.28
..
•
1.30
0.27 1/4!dc/L~1/2 I
0.26
L25·
1/4~ delL S' 1/2
, I
..
r.20
0.25 O!d e/L ...1/4
I
.. I-~
I
,
.
0.24 , ,
I
I
;,
,
~
I
,
I
-r-t-
, ,
,
I
0.23
,
I
I
+-t I
,
,
I
,
,
I.15
0 ...d e/LSI(4 I
,
..
.. I ,
,
..
,
,
I.10
~
,
r+
rt+:
t ,
I
'
.
, I
0.22
' -!-_tm ~.'
'.'
'J
.
,
"
I
"
, ,
I ,
I
,
.4-I
I
,
.,
,
I
++-
•
I
I
,
.. ,
,
, ,
I.05
,, I
I
I
0.21
o
0.1
02
03
0.4
Y/H Figure 4-51
Horizontal shear parameters for ultimate shear £tress
at distance de from the support (cross section type II and III) 4-130
I.00 0.5
e
e .
1-
-i -;
'
-
--
e
,
.i .
-
....-"":..:.T.--:I:
--. -L
--
. '
,
--
.
~.
-
1.0 --
~
0.8
!l'H=OaO,5
*H-O,3
....,
-
.....
w .....
0.6
-
-
•
•
~
•'III
-
...
-
·0.4 -- .-
~
~
-
'H=O.l
-
.
-- - - . .
. .L
~
---
0.2,
Y/H=O.2
--
-
-
Y/H=OA
Y/H=O.45
o
0.2
0.4
0.6
08
1.0
1.2
1.4
1.8
1.6
dC/l ( dC/Lly Figure 4-52
Horizontal shear coefficient ratios for ultimate shear stress distance de from the support (cross section type II and III)
at
2.0
,
"
"
x
, "
.' :,fl;;;,::,j] . .I' , .I '
"
..
"
.
L
.
"
.
C IV
.
:z: 5 u'
tI 1:'
u
4
2
o
0.2
'..
0.4
0.6 It
0.8
Y
T'O( H Figure 4-53
Shear coefficients for u~timate support shear (cross section type II and III)
4-132 .
1.0
[.~ I
I
I
I
I
", rr,,'77
L
.5
..
> U
4
3
0.1
, 0.3
0.2
0.4
..!.. L Figure 4-54
shea r coeff Lci.erit.s for ultirnatesupport shear .(cross section t"ype II and III)
4-133
0.5
L
! .
,.1... ".
".
.,
".'"
'.'
' ....
.
....
,
, ,
,.1
L
II
i (,,).
10
9
8
o
0.2
Figure 4-55
0.4
y H
0.6
0.8
. 1.0
Shear coefficients fo} ultimate support shear (cross section type II and III)
4-134
..
. ..
....- ...... ...., ... '
!
.".""
-
......
>.
"""" L
%
~ ~
•
r
II
to
o
~
~
o
10
9
0.1
0.2
0.3
x
L Figure 4-56
0.4
Y
Of'tr
Shear coefficients for ultimate support shear (cross section type I I and III)
4-135
0.5
Table 4-9 Edge Conditions
Impulse Coefficient Cl for Two-Way Elements Yield Line Locations
.
~
(Kull ).
'=It,;;;,;,:.:J] I
,
(sIH) . (Ku l I, '(:r/B)I
957 (K..,,).(p,lpH) (.IH) 957 (K..,,).(pvlpH)(LIH) '. (.IH)'
OSz/H;Sl
(K", ). 957 (.IH)
H5,.%'5.L/2
(K..,,). 957 (x/H)'
.,
..
,
.
'=IL;';~::?J~ L
~
..
91:
..
H
8
)----<,
.
H5,.z$L.
L5,1I5H
0 I
957
O$y/L$1
.1-
L
~
Three -edges supported and one edge free
Impu!se Coefficient t C.
~
OS,z/HSl
EJ
Two adjacent edges supported and two edges free
. . Four edges. supported
Limits
I
9
L/2"."H
478 (K..,,).(p,lpH)(LIH) . (.IH)'
0$%/H54
(K..,,). 957 (.IH)
8/25.%$.L/2
471 (x/H)'·
3
~
{ -..........-"
51 (KLJI);,(pvi'PS)
O".IL"I
O".IL"I
L
3 LIZ"."HIZ
4-136
(XLII i,
S5
__ ..J... __ .
,
"'IH)
.
7· (KLJI ).(PV/'PH) (II/H)'.
• (KL"),,(py/Ps)(L/H) 47_ (II/H)' 1
e Table 4-10 Edge Conditions
Yield Line Locations
,(x/H)'
287
+ (Ku,).
[3.333 (K LII ) . (pV/PH)
xIL;;,;,:J3 L
0 I
Three edges supported and one edge free
.1
elY -1
1
(L/H)I+
4.705(L/H -x/H) ;
(L/B)1
tan~)
]
Where, A._120 _ la n- 1
(0.2126) xlB
L
I
~
8 -. Ie£] /
)----<,
L5.y$H
287 CPr/PH)
,
o5.%IH 5.1
", 287
H5x$L/2
1148
.
ylH
+
(Ku, }.'(L/B -y/H) ] (LIB)' .
LAI [3.333 (K ) . (LIB) (yIH)!
+ (K
LJ1
" ] Where, ),,_12° +4.705(1-y/H) }.'(yIH -t.tu
tanA.)
(02126) gIL
-tan-\
[3.333(KLJ,.)~+(KLJI) ..t(PY/PH)(1_X/d)J rlH
[0.833 (KL.JI).. (riB)'
+2(Kl.AI).
r
(XIB -1 4..705(L/2H -:LIB) tan>.)] (L/HP+ (L/B)I
Where,).,= 120 _ t.a.n- 1 (0.2126) xlB
__ ..J...-.-.
ILl
'
[1.333(Kud,,(PV/PH)+32(K
)
1
(L12H-YIHJ]
05yjLSt
717
LI2$y$H
(LIH) (K LN )~f(yIH lanA) 287 (pVIPH)[L667 (Kud" (y IH ) !-LI2H+t.i05(l-YIH) .
.O$xIH$!
1112$x$LI2
O$yIL-st
ylH
.'
LM..
+
• 1148
3
~
:I:
287.
,
H
I.
O$.y/L:51
,
{L;,;~;:;J3
Four edges supported
«rn
-,
:-
"
287 [3.333 (Ku). + (Kul ).'(pV/PH) (l-xIH) ]
[3.333(Kud. H5:x$.L
t2--t
w
Impulse Coefficient I Cu
o'5:x/H5, 1
EJ 1
" >-'
Impulse Coefficient Cu for Two-Way Elements
Limits
H
Two adjacent edges supported and two edges free
e
e
.. 1148
1148
(LIB)!
,
[0,&13(K'·)·+(K _jp )(1-2~)] xIH·· LN )'( ~ pi" H
[O.4li(KLM)~ (xIH)1
ylH
-'
)]
'-
1148 (pdpH)
[0.417(KLN)~(LIH) (1IIH)'
.
Where, A_120_tan-1
,
) I (LIII-2 YI H LM .. (LIB)1
'
(02126) 2 11 IL
(02126) ~ _, 2xlll
l]
3 LI2$.y$.HI2
Where, >._12°_t.a.n-
-,,
, lelH-t 4./05(LI2H-xIH) tanx +2 (Kutl .. (LIII)I+ (LIB)!
[O:S33(KU,)~(PdPH) +(K
]
+ (K LM. ) !(2 YIH -LIB +4.705(1-2YIH) tenx)
J
Where, A-12°-t&n- 1
(02126) 2YIL
TN 5-1300/NAVFAC P-397/AFR 88-22
Table 4-11
Impulse Coefficient Cu for One-Way Elements IMPULSE COEFFICIENTS
EDGE CONDITIONS
Cu
CANTILEVER
127
FIXED SUPPORTS
510 L
4-138
Table 4-12
Shear Coefficients for Ultimate Shear Stress at Distance dc from the Support for One-Way Elements (Cross Section Type II and III) S,ULTIMATE SHEAR TRESS COEFFICIENTS Cd
EDGE CONDITIONS
CANTILEVER
~
FIXED SUPPORTS
~
I.
L
~I
2(dC)(I_ dc) L L
. dc I dc .16(-)(---) L 2 L
"
4·139
Shear Coefficients for Ultimate Shear Stress at Distance de from the Support for Two-Way Elements (Cross Section Type -II and III)
Table 4-13
- .Edae c::onditionll
Two adjacent edges fixed and two· edges free
Limita
Yield line loeatioD
H
O:S:d./%:~:a
0 %[t·;-;;,::8 I
" ....
-,
"o
Three edges fixed and one edge free
0 % I
I
L
I
•
""",.
\
\
L
,t~
f5J % ,
:•
-,
-
--..,....--__
__ k
L
e·
,
IA ...............
","
3O(d.!z)(1 (5
at~
coefficient
eN
.
5 (d./z) (1-40/%)
OSd./LSt
6(d-IL) (3 +2y/H) (i-d./ L) (2-,1/8 -d.VILH) (3-2,,/H)(6-IIIH dollfLH) (d./L)(3 +2~/H)(1-d./L) ('1.-IIIH (3 2V/H){1 d.,,/LH)
' r.. ..
O$d./r:S:1
(5-4d./%l
t:Sdo/:9
5(4,/%)(I-do/:)
Vertical ultimate mear atre!lll coefficient CV
O:S:d./H:~a
6(4./H) (3+2%/L) (I-(MH) (2 -z/L-doz/HL) (3-'1.%/L)(6 -z/L-4d.J:/HL)
l:Sd./H$l
tSd./z$l
tSd./L:Sl
UfI\i",
.-
d./Z)I
U,(%)
-do",LH)
3O(d.j%)(I-do/z)t
rr r
Four edges fixed
Horiaontal ultimatol ahear
(d./H)(3+2z/L )(1-d./H) (2 -zlL -tLzIHL) (3-2%IL) (l-doz/HL)
3O(d"lv)(I-d./r)1
, OSd./rSt
(5 4d./V)
5(d"lv) (i-do/r)
t:S:c4/rS l
O:S:do/HSt
6(do/H) C3+h:/L) (I-do/H) (1-%/L-tLzIHL) (3
u/L)(3-~/L
~/HL)
tSd"IHSI
2(4./H)(3+4:/L) (I-do/H) (I-:/L -~/HL) (3-u/L)(I 'ld.%/HL)
05dolLSt
12(doIL)(6 I/IH)(I~2d./L)(2-I//H 24.1IILH) (3-21//H)(6 -11/ H -StklllLH)
05d,,/1ISt"
3O(d,,/1I) (I -4.111)' (5 ·4d./11)
tSdo/LSt
2(d.IL) (6-1IIH) (1-24.1L) (2 -IIIH -2d.IIILH) (3- 211/H)(I-2d.IIILH)
tS:d./lISI
6(14/1/)(l-dol'l)
05~/HSt
24(4./H)(3-%/L)(1 -2d./H)(I-.;JL 'ld.%/HL) (3-4%/L) (3 -.;IL-8d.zIH L)
OSd./~SI
30(d./.;)(I-d.I.;)'
(5-4d./z)
8(do/HH3-z/LHI -2d.IH)(I-.;fL-24.zIH L)
ISdol.;SI
5(tI.1.;)(l-do/z)
ISd-/HSt
OSdolLSt
24(doIL)(3 '1IH)(I-24./L)(I-I//H 2d.lIILH) (3-4.,,/H)(3 1I/H-&1.'1ILH)
05d.hSI
3O(do/l/)(I -11.1'1)1 (5 4d./I/)
ISd-ILSI
8(d-IL) (3-lIIH) (l-2d.IL) (l-lIIH -2d.'11 LH) (3-411IH) (i-4d-'1/LH)
I S:d.hSI
5 (do/v) (t -tI./1I)
e
(3 4%IL)(1 4d.z/HL)
e
i
Table 4-14
Shear Coefficients for Ultimate Support Shear for One-Way Elements (Cross Section Type II and III)
,
ULTIMATE SUPPORT SHEAR COEFFICIENTS Cs
EDGE CONDITIONS
,
CANTILEVER
FIXED SUPPORTS
~
..
~ ~"
I.
L
4-141
·2
8
.1
Table 4-15
Ed.-
Shear Coefficients for Ultimate Support Shear for Two-Way Elements (Cross Section Type II and III) Horiao:t&al uhima\e npport ahear coemciut Coli
Yield liM loea\ioll
eoDdltio ..
V.=~~~~
H
0
TwO adjaeeot edsee fixed IIDd two edges
. ..
/"
b..
~[1;;;,~:J3 I,
,
. Throe od.,. fixed and one edp Iree
\\
\
""" ~
"
'
6(2-./H)(3+2./H) (6 ./H)(3-2./H)
6 ./H
,
I
~
6(2-./L)(3+2./L) (6-./L)(3-"/L)
I
L
nj
"
6 ./L
,
I
J.<...
J''''''
<,
:, ,.,t
6 ./L
6(1-./L)(3+U/L) (3-./L) (3- ../L)
12(2-./H)
6 ./H
(3-2./H)
-
,,~
L
.
.
..!,
,
Four qCll .fixed
~
<, -<,," , ,.,~- ... " , "', , , , ....---- ..... -:
--
-
~~
,
~,.,
:,
L
4-142 .
6 ./L 24(1-./H) (3-4./H)
24(1-./L) (3-"/L)
6 ./H
TK 5-l300fNAVFAC P-397/AFR 88-22 COMPOSITE CONSTRUCTION 4-36. Composite Construction 4-36.1. General Composite elements, are composed of two concrete panels (donor and acceptor) separated by a sand-filled cavity. They have characteristics which are useful in the blast resistant design of structures located close-in to a detonation. For a large quantity of explosives, replacing a single concrete pan~l with a composite element can result in a considerable cost sa~·ings. It is not usually cost effective to use a composite element for smal Ler quantities of explosives where a single concrete panel would be three, feet thick or less. Where a single concrete panel would be between three and five ,feet thick, a detailed cost analysis is required to dete~ine whether or not'a composite element would be more cost effective. ' Composite walls are generally used as barricades to prevent propagation of explosion between large quantit~es,.of explosives. The"" s~ructures are usually designe,dtf?,r, '~,P\cipie.nt fail~r~. Composite e Lementis may be designed to provide higher degre~s of protection, but the massive ~,alls (greater than 5 feet thick), that m
4-36.2. Blast Attenuation Ability 'of Sand Fill The method for calculating the impulse capacity of composite elements is similar, to that: for'single laced concrete elements exc..pt that the blast attenuating ability of the 'sand m;;st' be included in the ·calculation. The blast wave attenuation is partly due to the increased lDass of the slab. When computing the impulse capacity of each co~crete' p~nel, the total effective mass includes both the mass of the concrete and the mass of one-half of the sand. This increased mass is taken into account'by multiplying the impulse coefficients Jor spalle~ sections, by,
4-143
TN
5~1300/NAVFAC
P-397/AFR 88-22
[ , T.
,
+ '\" +
(
2
~s
) (
Ts
2
Wc
,]
Idc
4-123
or for unspalled ,!ections by
[
'!s
Tc + -'Wc
(
' T's 2
)
] fTc
where w~
'wc
Ts
weight density of sand 'weight density of concrete thickness of sand fill
The attenuating ability of the sand due 'to'blast'wave'dlspersion and energy absorption is a function of the thickness and density of the sand, the impulse c~pacity of the concrete panels and the quantity of explosive. Figures 4-57 and' 4-58 have been developed to predict the impulse capacity of the concrete ' element for a sand density equal to'85 pcf'and lOO'pcf; respectively. These' figure,s' are based on identical donor and acceptor panels. The effect of the' " quantity 'of 'explosive is taken into account through the use of "scaled" parameters which are defined as follows: ' 4-125
4-126
4-127
4-128
Wl / 3 where scaled thickness of concrete panel weight of 'explosive charge scaled thickness of sand
scaled blast impulse which can be resisted by acceptor panel i ,a b -
blast impulse capacity of acceptor panel 4-144
TK~ 5-l300/NAVFAC 'P-397/AFR 88-22 ;1
r
sum of scaled blast impulse resisted by the acceptor panel and the sc~ied blast impulse absorbed"by the' sand sum of blast impulse capacity of the accept~r'panel and the blast impulse absorbed by the .sand ,
ia -
.
,
,'
.
~." ~ '.~' ..
Explosion response slab tests have indicated that the density of the sand fill affects the amount of blast energy absorbed by the sandi dfsplacement', 1. e , , the higher the initial sand "density ,the smaller' amount of: blas t; energy absorbed, Also, it was observed in~he above response 'tests' that 'for a unit weight of sand equal to 100 pcf, the deflection of the donor panel is approximately equal in magnitude to the-, .deflection of the acceptor panel. On ,,1· • ,.. ' r the other hand, with a unit weight ofrsand fill equal to 85 pcf, it was observed that the deflection of the donor panel usually,was significantly larger than that of the acceptor panel. This iatter phenomenon was caused by " sand . had ~the fact that, with the lower density, - 'the more voids and, therefore, more room for movement of the sand particles. This sand movement in turn permitted larger displacements "of "the donor panel before the near . t . --,;' "., . . .: '. ( • !.' .. < . solid state, of, the sand occurred. ' , , " ' . " .. . " ' J." r' . t' ". . "..... ~ 'J,." • ,.~ " ~
~".)..."'"
I
~
~'
~'
',
Based on the ,above inforn"ition,·.it can be ;seen that' i f nea£ equaL displaceme~t of' the' donor and receiver panels,.'!,re .desired, then a unit weight of sand fill equal to 100 pcf"shoul,d -be used. ' A variation of the dIsplacement of donor and receiver'panels can be achieved,using 'a unit weight of ,sa~dequ~l'to 85 pcf,' but the actual: variation cannot be predict;ed." , .
•
t
._ ..1
'. •
'
Since the• impulse capacity of composite elements is a function of the density -. -' . . -•.;. . r • of the sand, it is important to prevent 'the sand from compacting due to Its own weight and/or water drainage. Several possible methods fo'r matncatnt.ng the proper sand density are discussed in subsequent sections concerned with construction details of composite elements. I
.
,-~.
.'
..
',."
."
f'
~
..' ,
~.
•
'
,
.~,
4-36,3, Pro~edure' for Design of' 'Composite Elements "," J
t.
- . ;1.
"-'.
I
.' '
,"
C'
~
r,
The design of composite elements is a' trial and' error procedure. 'By using Figures 4-56 and 4-57 and the impulse coefficients of previous ~ections, the calculations are greatly simplified. The donor and acceptor slabs are identical making it necessary to design only one wall. The depth of the sand fill is usually equal to the total thickness of the two concrete panels. Using the procedures in the previous sections, each panel is designed to have a blast impulse capacity slightly less than half the required. This includes the increase in capacity due ,to the additional mass of the sand (Equations 4123 and 4-124). It should be noted that the design is based on the assumption that both panels will attain the same deflection. If the density of the sand fill is 85 pcf, this will not,be.true. The donor panel will probably have a larger deflection than the acceptor 'panel. Since the accuaI deflection of each panel cannot be predicted, it must be assumed that the design deflection is an average of the two. With the blast impulse capacity of the two concrete pan.els, Figure 4-56 or 457 is used ,to d!'termine,the, t;:otal b~ast' capacity,of the composite element. The following procedure illustrates the use of these'figures.
TN 5-l300/NAVFAC P-397/AFR ,88-22 1.
Using the given charge weight calculate the scaled thickness of the concrete panel and the sand, Tc and ,T s' respectively.
2.
Calculate the scaled
Impuls~
capacity which can be resisted by the
donor panel i b d and the acceptor panel i ba· 3.
.
Using either Figure 4-56 or 4-57 a)
"
" Enter
the ordinate at value of. i ba·
b)
c)
d)
4.,
Proceed horizontally from Tc to i a (the sum of ' the scaled unit blast impulse resisted by the acceptor panel and the scaled unit'impulse absorbed by the s~nd).
" ,Calculate the summations of .ibd .and. i a to get 'thetot;al unit
.
,
impulse which can be resisted by the composite wall i bt.' 5.
Determine the scaled unit blast impulse i b acting on the composite , element. ." ,
" 6.
Compare i b t and i b. If the blast impulse which can be resisted by the composite element is not greater than,or equal to the impul~e produced by the blast then the impulse capacity of the walls shou~d be increased and/or the thickness of the sand increased . ., ~
,
,
..
...." '
',
,.
1000'
I
JT" l~.
(
Figure 4-57
Attenuation of blast impulse in sand and concrete,w s = 85pcf
'-4-147
"-
!S.s AIL J.";'O
d5
.:
J.-
20 .. _.-
Figure 4-58
Attenuation of blast impulse" in sand and concrete, w s = 100 pc f
4-148
TK 5-l300/NAVFAC P-397/AFR 88-22 ,
~
...
J1LTIHATE. DYNAMIC STRENGTH OF REINFORCED. COlIC~TE BEAMS 4-37
Introduction
Blast resistant.~oncrete b~ildings subjected to external blast pressures are generally shear wall structures rather than rigid frame structures. Shear wall. structures respond to lateral loads in' a.somewhat differ~nt manner than rigid frame structures; the basic difference being the manner in which the lateral loads are transferred to the founda~ion. In rigid frame struc~ures the lateral loads are transmitted to'the foundation through bending of the columns. Whereas, in shear wall structures, the lateral forces are transmitted to·the foundation through both. bending and shearing action of the shear walls. Shear walls are inherently strong and will resist large lateral forces. Consequently, shear wall structures are inherently capable of.resisting blast loads and can be designed to resist substantially large blast loads whereas rigid' frame structures cannot be.econ~mically·designedto resist significant blast loads. .. . In shear wall structures ,. beams and columns are usually provided between shear walls to carry the vertical loads including blast load" on the roof and not to transmit lateral loads to the foundation. For example, blast. loads applied to the front wall o f- a two-story shear wall s t ructwreiare transmitted through the roof and intermediate floor slabs to the shear.walls (]perpendicular walls) and thus to the ·foundation. The front wall spans vertically between the foundation, the floor, and the roof slab. The'upper floor and r~of.~labs· act as deep beams, and in turn, transmit the front wall reactions to the shear walls. The roof and floor beams are not subjected to sign'ific,ant axLaL loads due to the diaphragm action of the slabs. The design of beams as presented in the following sections applies to beams in shear wall type structures rather than rigid frame structures. The design procedure presented is for transverse loads only; axial loads are not considered. However, the procedure includes the design for torsion. The design of beams is similar to the design of slabs as described i? sections 4-13 through 4-18. The most significant and yet not'very important· difference in the design procedure is that in the case of a slab the calculations are based on a unit area, whereas, for a beam, they are based on a unit length of beam. Beams may be designed to attain limited or large deflections in the same manner as non-laced slabs. However, unlike non-laced slabs which in some cases do not require shear reinforcement (single leg stirrups), shear reinforcement in the form of closed ties must always b~ provided in beams. Under flexural action, a beam may attain deflections corresponding to 2 degrees support rotation with a type I cross-section to provide the ultimate moment capacity. The flexural action may be extended to 4 degrees support rotation if equal tension and compression reinforcement is furnishe~. A type II or III cross-section provides the ultimate moment capacity and the required closed ties restrain the compression reinforcement., If sufficient ' lateral restraint is provided, the beam may attain 8 degrees support rotation under tension membrane action. The above support rotations are incipient failure conditions for the structural configurations described. Beams are primary support members and, as such, are generally not permitted to attain large plastic deformations. For personnel protection, the maximum
4-149
: 'TN 5-1300/NAVFAC P-397/AFR'88-22 deflection is limited to a support rotati?n of 2.0 degrees. Structures intended to protect equipment and/or explosives may be designed for deflections up to incipient failure conditions. Beams are generally employed in structures designed to resist the effects associated with far range explosions.' IIi: these structures, beams are usually' used in the roof as primary support members and as secondary support members such as pilasters around door openings. To a far lesser extent, beams are designed to resist the effects of close-in detonations in containment type structures. In these cases, they ~re'general1y'used secondary' support members such as pilasters around door openings. Large" tensile forces are induced in containment type structures and, therefore;, these structures lend themselves to tension membrane'sction when the applicable design criteria' permits large deformations. :,'
as
The interrelationship between the various parameters involved in the design of beams' is readily described with the use of the idealized resistance"deflection curve shoWn in Figure 4- 59. ' . ' 4-38. Ultimate Moment Capacity
,. ..
4-38.1. Tension Reinforcement Only
,
.
"
./~
The ultimate dynamic resisting moment Mu:of'a'rectangular'beam section of width b with tension reinforcement only (type I) is given by: •. a/2)
4-129
and: 1
, ' 4-130
a
0.85b f'dc where:
Mu
ultimate' moment capacity ~
As
total area'of tension reinforcement within the beam
fds
dynamic design stress of reinforcement
d
-'
distance from extreme compression fiber to centroid of tension reinforcement " eq~ivalent
a
depth of
rectangular stress block
b
width of beam •
f'dc"
dynamic ultimate compressive 'strength of concrete
The reinforcement ratio p is aefined'as: p
..
bd ,I
',.r
4-150
4-131
". .
..
(
'\"
~
and to insure against sudden,compression failuF~s" the reinforcement ratio p must nqt exceed 0.75 of. the ratio Pb which pro,duces balanced conditions ~t ultimate strength and is given by: l,.
" -[
0.85K~
.... f'dC]
4-132
f ds
where:
• It '''l '-~
"
0.85 for f'dc up to 4,000 psi and is reduced by 0.05 for each l,OOO.psi in excess of ~,OOO p~i 4-38.2. Tension and Compression Reinforcement
: "
The ultimate dYnamic resisting moment ~ ·o.f. a rectangular'.b,e,am width b with compression re.inforcem~n~. Ls. given ]>Y:,
,~ect~on of .
T.,." . .
. ';"
:J.
•
~.,
"
--i'
and: "
'.
: '.;, ...... , .
...
0.85 b f'dc
'.,1'1, ' •. .
where:
",,~~
.
, 'a ,.,."
"
J"
,~.
'.
.
..
. !
..!">
.••
.•'
total area of compression' reinforcement within th~,beam
d'
distance from extreme compression fiber compression reinforcement'
.
,to,~entroid
of
"
.defi~e~
as:
A' s
4-135
bd ,
,
Equation 4-133 is valid.only when the compression rein~orcement yields ultimate s.trength. This condition is satisfied ~hen:.
p-p' S
..
I
A' s
The compression reinforcement ratiop' is
0.85 Kl
.-
[
f'dc f ds
:' ]
[ - '7,000
87,000,.-
.
'd,] .:
·a~
4-136 <: ~
• .r
In addition,' the quantity p-p' must not exceed 0.75 of the value of Pb given in Equation 4-132 in order to insure against sudden compression failures. If p-p' is less than the value given by Equation 4-136, l:he ultimate resisting moment should not exceed the value given by Equation 1,-129. For the design of concrete beams subjected to far range blast loads which are to attain support rotations of 2 degrees or less, it Is recommended that the ultimate .resisting moment be computed using Equation 1,-129 even though a 4-151
~
.
. .-.
TM: 5-l300/NAVFAC P-397/AFR 88-22' considerable amount of compression reinforcement is required to resist rebound loads. It should be noted that large amount of compression steel that does 'not yield due to the linear strain'variation across' the depth of t~e section, has a negligible effect on the total capacity.
a
For type II or III cross-sections, the ultimate resisting moment rectangular beam section of width b is given by:
Mu
of a
where
area of tension or,compression reinforcement within the width b distance between the centroids of the compression and the tension reinforcement The above moment capacity 'can only be obtained when the areas of the tension and compression reinforcement are equar". In addition,thesupport rotation must be greater than 2 degrees except for close-in designs where direct spalling may occur and result in a type III. ",,' •• ". ' 4-38.3. Minimum Flexural Reinforcement To insure proper structural behavior under both conventional and blast loadings, a minimum amount of flexural reinforcement is required. ' The minimum reinforcement required for beams is somewhat greater than that required for slabs since an overload load in a slab'would be distributed laterally and a sudden' failure will be less likely. The minimum required quantity of reinfor-" cement is gfven by: '
,
,
p ;.. 200/fy
4-138
which', for 60,000 psi yield strength steel, is equal to a reinforcement ratio of 0.0033. This minimum reinforcement ratio applies to the tension steel at mid-span of simply supported beams and to the tension steel at the supports and mid-span of fixed-end beams. Concrete beams with tension reinforcement only are not permitted. Compression reinforcement, at, least equal to one-half the 'required tension reinforcement, must be provided., This reinforcement is required to resist the ever present rebound forces. Depending upon the magnitude 'of these rebound forces, the required compression reinforcement may equal the tension reinforcement. 4-39. Ultimate Shear (Diagonal Tension) Capacity' 4-39.1 .. Ultimate Shear Stress The nominal shear stress vu, as a measure of diagonal tension, is from:
co~puted
e
bd where: 4-152,
4-139
TK 5-1300/NAVFAC P-397/AFR 88-22 ~"
•
j . .
•
nominal shear stress total shear at ~ritical section' The critical section is taken at a distance d from the face of the support'for those members that ~aus~ compression in their supports. The ~hear' at sections between the face ,of the support and the section d therefrom nee,d, not be' considered critical. For those' members' that cause tension .Ln ,'their supports, the critical section is 'at tlje, face of th~ supports. ' 4-39.2. Shear Capacity of Unreinforced, Concrete The shear st,ess permitted on a~ unreinfo~ced web of a beam ,subjected to flexure only is limited to: Vc
where:
- [1.9 f'dc l/2 + 2,500 P] S ..
, '
3.5 f'dc l/2
.
' ) ; .,..
maximum shear capac Lty of an unre fnf'o rc ed web
,..·t'
reinforcement ratio, of the tension rei'}forc'';~~rit 'at the support
p
"
4-39.3., Design of Shear Reinforcement
"
,
Whenever the nominal shear stress '''':I ~,!"ceeds, the"she~r ,capacity vc of the concrete, ' shear, r eLnforceraent; must be provided to ..carry the excess.', Closed ties ,placed perpendicular to the flexural reinforcement must be used to furnish the additional shear capacity. Open stirrups, either single or double leg, are not permitted. The required area of shear reinforcement ,is calculated using:
[ (vu - v c ) b ] 5g
Av where:
,
Av
4-141
¢ f dy '
total area of stirrups excess shear stress
¢ -
4-39.4. Minimum Shear Reinforcement
.
,
In order to insure the full development of, the f'Lextrre l, reinforcement in a beam, a premature shear failure must be prevented. ~he followi~g limitations must be considered in the design of closed ties: 1.
The design shear stress (excess shear s t r'e s s V u - v c) used in Equation 4-140 shall be equal to or greater than the shear
·
,
'
TN 5-1300/NAVFAC P-397/AFR 88-22 capacity of unreinforced concrete,vc as qbtained from equation
4-139.
2.
The nominal shear stress
3~
The are~
4.
'The r~~uired area Av of closed ties shall be determined at 'the critical section and this quantity and spacing of'reinforcement shall be used throughout the entire member.
5.
The maximum spacing of closed ties is limited to d/2 when v u'- Vc ~s less than 4 ~ f'dc)1/2 or 24 inches! whichever is smaller. When V u - Vc is greater than 4 ( f'dc) /2 the maximum spacing is limited to d/4.
Vu
must not exceed 10 ( f'dc)1/2.
of closed tie~ should,natbe' less than 0:0015 bs s'
Av
r
4-40. Direct Shear Direct shear failure of a member is characterized by the rapid propagation of a vertical crack through the depth of the member. 'This crack is usually located at the supports where the maximum shear stresses occur. Failure of this type is' possible eveh in members reinforced< for diagonal tension. ' Diagonal bars are'required at supports to prevent direct shear failure: when the design support rotation exceeds 2° (unless.. the beam is simply supported), when the design support rotation is s 2° but the direct shear capacity of the concrete is insufficient', or when' the sec tLond.s in tension. Diagonal . reinforcement consists of inclined bars wnich'extend from the support into the beam. , Diagonal bars are not typically recommendea in beams. Therefore, beams should be designed for small rotations and with an adequate cross-sectional area for' the direct shear capacity of the concrete, Vd' to exceed the ultimate direct shear force, Vs' If the design support rotation, 9, -Ls less than or equal to 2° (9 s 2°), or if the section, with any rotation 9, is simply supported (total moment capacity of adjoining elements at the support must'be significantly less than the moment capacity of the section being checked for direct shear), then the ultimate direct shear force, Vd' that can be resisted by the concrete in a slab is given by Equation 4-30. If the aesign support rotation, 9, is greater than 2° (9) 2°), or if a section (with any support rotation) is in net'tension, then the ultimate direct shear capacity of the concrete, Vd,'is zero and diagonal bars are required to take all direct shear. " ', ••. If diagonal bars must be used, the required cross-sectional area is: ,
.
..
Vd)/(f~ssin(~)r'"
,
.
where: -'Vd' -
0.18 f'dc bd.
(9 S 2° or, simple supports),
L
4-154
4-142
TM 5-1300/NAVFAC P-~97/AFR 88-22 " section in tension). (8) 2' or
or
o
and
total area of diagonal bars at the support,within a width-b shear at the support of unit width b
a
'.
angle formed by the plane of the diagonal reinforcement and ,the longitudinal reinforcement.,' "
4-41. Ultimate Torsion Capacity '
4-41.1. General
... ,
In addition to the flexural effects considered above, concrete beams may be subjected to torsional moments. Torsion rarely occurs alone in reinforced concrete beams. It is present more often in combination with transverse shear and bending. Torsion may be a primary influence but more frequently it is a secondary effect. If neglected, torsional: stresses can cause distress or failure. Torsion is encountered in beams that are unsymmetrically 10a~ed. Beams are subject to twist if the slabs on each side' are not the same span or if they have different loads. Severe torsion will result on beams that are essentially loaded from,one side., This condition exists for beams around'an opening in a roof slab and for pilasters around a door opening. The design for torsion presented in this Section is limited to rectangular sections. For a beam-slab system subjected to conventional loading conditions, a portion of the slab will assist the beam in resisting torsional moments. However, in blast resistant design, a plastic hinge is usually formed in the slab at the beam and, consequently, the slab is not effective in resis'ting torsional moments. 4-41.2. Ultimate Torsional Stress The nominal torsional stress in a rectangu1ar.beam in the vertical direction (along h) is given by: ',.' 4-143 and the nominal torsIonal stress in the, horizontal direction ,(along b) is given by: 4-144
"
where: vt u
nominal torsional stress
"Tu
total torsional moment at critical section
b
width of be am 4-155
TIl:
5-l300/NAVFAC P-397/AFR~88-22 h
overall depth of beam
The critical section for torsion is taken at the same location as diagonal tension. It should be noted that the torsion stress in the vertical face of the beam (along h) is maximum when b is less than h whereas the torsion stress along the horizontal face of the beam (along b) is maximum when b is" greater than h. Capacity of Unreinforced Concr$te for Combined Shear and Torsion
4-41.3
For a beam subjected to combined shear (diagonal tension) and torsion, the' shear stress and the torsion stress permitted on an unreinforced section are reduced by the presence of the other. The shear stress permitted on an . unreinforced web is limited to:
,
4-145' '
, ,
while· the torsion stress taken by the concrete of the same section' is limited to:
J
.
.
.
"-
4-146
-'
.'
~: '
where:
-'
.,~
maximum shear capacity of an unreinforced web Vtc Vu
Vtu
••
~
1
maximum torsion capacity of an unreinforced web nominal shear stress
nominal torsion stress in the direction of Vu
It should be noted that the shear stress permitted on an unreinforced web of a beam subjected to shear only is given by Equation-4-l39, Whereas, the torsion stress permitted on an unreinforced web of a beam subjected to torsion only'is given by:
Whenever the nominal shear stress vu exceeds.vche' shear capacity vc of. the concrete, shear reinforcement must be provided to carry the excess. This quantity of shear reinforcement is calculated using Equation 4-140 except the 4-156
TM 5-1300/NAVFAC P-397/AFR88-22
value of vc shall be obtained from Equation 4-144 which includes the effects of torsion. 4-41.4. Design of Torsion Reinforcement j.
4-41.4.1. Design;of Closed Ties Whenever the nominal torsion stress"vtu exceeds the maximum torsion capacity of the concrete, torsion reinforcement in'·the shape of. closed ties,' shall be provided to carry the excess. The required area of the vertical leg of the closed ties is given by: ; ( v(tu)V - v t c)
b 2hs 4-148
'3
( v(tu)H - v t c ) bh 2 s
4-149
3
in which: 0.66 + 0.33 (ht/bt) S
1.50 for h t
0.66 + 0.33 (bt/ht) S
1.50 for h t ? b t
~
b t'
4-150a ~-150b
where: area of one leg of a closed stirrup resisting torsion within a distance s ,
s -
spacing of torsion reinforcement in the longitudinal reinforcement
II
direction parallel to
capacity reduction factor equal to 0.85 center-to-center dimension of a closed rectangular tie along b
center-to-center dimension of a closed rectangular tie along h
The size of the closed tie provided to resist torsion Jnust be the greater of that required for the vertical (along h) and horizontal (along b) directions. For the case of'b less than h ." the torsion stress' in the vertical direction is maximum and the horizontal direction need not be considered. However, for b greater than h, the torsion stress -Ln the horizontal direction is maximum. In this case the required At for the vertical and horizontal directions must be obtained and the greater value used to select the closed stirrup. It should be noted that in the horizontal direction a beam, in shear wall type structures, is not subjected to lateral shear (slab resists lateral loads) and the 4-157
TK 5-l300jNAVFAC P-397/AFR 88-22 value of v t c used in Equation 4-148 is calculated from Equation 4-146 which does not include the effect of shear. When torsion reinforcement is required, it. must be provided in addition to reinforcement required to resist shear. The closed ties required for torsion may be combined with those required for shear. However, the area furnished must be the sum of the individually required areas and the·most restrictive requirements for spacing and placement must be met. Figure 4-60 shows seve,al ways to arrange web reinforcement ..
~or
low torsion and shear, it is con-
venient to combine shear and torsional web reinforcement in the form of a single closed stirrup whose' area is equal to' At + Av/2 . For high torsion and shear, it would be economical to provide torsional and shear reinforcement separately. Torsional web reinforcement consists of closed stirrups along the, periphery, while the shear web reinforcement is in the form of closed stirrups distributed along the width of the member. For ve~ high torsion, two closed stirrups along the periphery may be used. The combined area of the stirrups must equal At and they must be located as close as possible to each other, i.e., the 'minimum separation of the flexural reinforcement. In computing the required area of stirrups using -Equ.at.Lon 4-147 , , the .v'a Lue of bt should be equal to the average center-to-center dimension of the closed stirrups as shown in Figure 4-60. 4-41.4.2. Design of Longitudinal Reinforcement In addition to closed stirrups, longitudinal reinforcement must be provided to resist the longitudinal tension caused by the torsion. The required area of longitudinal bars Al shall be computed by: t Al - 2A t [ b
+ s
h
t
1
.4-l5la
or by:
Al -
["Ob. f dy
Vt u Vtu + Vu
•
,.,] [b'
+ ht ·s
]
. 4-l5lb
"
whichever is greater. When using Equation 4-l5lb, the value of 2A t shall be greater than or equal to 50 bs/fdy' It should be noted that Equation 4-l5la requires the volume of longitudinal reinforcement to be equal to the volume of the web reinforcement required by Equation 4-147 or 4-148 unless a greater amount of longitudinal reinforcement is required to satisfy the minimUm requirements of Equation 4-l5lb. Longitudinal bars should be uniformly distributed around the perimeter of the cross section with a spacing not exceeding 12 inches. At least one longitudinal bar should. be placed in each corner of ,the closed stirrups. A· typical arrangement of longitudinal bars ~s shown in Figure 4-60 wh~re , torsional longitudinal bars th'\t are located.. in the flexural tension zone and flexural compression zone may be combineq with,the flexural steel.
4-158 '
r
TN 5-l300/NAVFAC P-397/AFR 88-22 The addition of torsional and flexural longitudinal reinforcement in the flexural compression Zone is not reasonable. It is illogical to add torsional steel that is in tension to the flexural steel that is in compression~ This method of adding torsional steel to flexural steel regardless, of whether the latter is in tension or in compression is adopted purely for simplicity. For blast resistant design, flexural reinforcement added but not included in the calculation of the ultimate resistance could cause a shear failure. The actual ultimate resistance could be significantly gzeace r than the calculated ultimate resistance for which the shear reinforcement is provided. Therefore, torsional longitudinal reinforcement cannot be indiscriminately ,placed but rather 'must be placed only where required. In the design of a beam subjected to both flexure and torsion, torsional longitudinal reinforcement is first assumed to be uniformly distributed around the perimeter of the beam. The reinforcement required along the vertical face of the beam will always be provided. However, in the flexural compression zone, the reinforcement that should be used is the gr..ater of the flexural compression steel (rebound'reinforcement) 'or the torsional steel. In terms of the typical arrangement:of reinforcement in Figure 4-60, either A's or Al is used, whichever is greater, as the design steel area in the flexural compression zone. For the tension zone at the mid span of a uniformly loaded'beam the torsional stress is zero and torsional Long Lcud.LnaL reinforcement is not added. Conversely, the tension zone at the supports is the location of peak torsional stresses and longitudinal torsional reinforcement must be added to the flexural steel. 4-41.5. MinimUm Torsion Reinforcement In the design of,closed ties for beams subjected to b<>th shear and torsion, the 'following limitations must-be considered: 1.
The minimum quantity of closed ties provided 'in 'a beam subjected to both shear and torsion shall 'not be less than, that required for a beam subjected to shear alone,
2.
The maximum nominal shear stress
3.
The maximum nominal torsion stress v t u shall not exceed 12
4.
5.,
(f'
Vu
must not exceed 10 (f'dc)1/2 ,
) 1/2 dc,
The required spacing of closed stirrups shall not exceed (b t + . h t)/4 or 12 inches nor the maximum spacing required for closed ties in beams'subjected to shear only.
,
Av
The required areas and At shall be determined at the critical section and this quantity and spacing of reinforcement shall be used throughout the entire beam. 4-159
TN 5-l300/NAVFAC P-397/AFR 88-22 6.
To insure the full development of the ties, they shall be closed " using 135-degree hooks.
4-42. Flexural Design 4-42.1. Introduction The flexural design of beams is very similar to the design of non-laced concrete slabs. The main difference is that in the case of a slab the calculations are performed based on a unit area, whereas for a beam, they are based on a unit length of beam. In addition, since beams are one-way members, the distribution of mutually perpendicular reinforcement does not have to be considered. 4-42.2. Small Deflections The design range for small deflections may be divided into two regions; beams with support rotations less than 2 degrees (limited deflections) and support rotations between 2 and 4 degrees. Except for the type of cross-section ava Ll eb Le to' resist moment" the de s Lgn procedure is the same. A concrete section and reinforcement are assumed. Using the equations of section 4-38 (Equation 4-129 for type I cross-sections, Equation 4-137 for type II and III cross-sections) the moment capacities of the trial section is computed. The moment capacities are,require4 to calculate the ultimate unit resistance r u and the equivalent elastic deflection XE. These parameters, along with the natural period of vibration Tn' define the equivalent sing1edegree-of-freedom system of the beam, and are discussed in detail in Chapter 3. A dynamic analysis (see section 4-43) is performed to check if the'beam meets the allowable deflection criteria. Finally, the assumed section is designed for shear and torsion, if applicable. If the beam d~es not meet the allowable response criteria, the required shear reinforcement is excessive, or the beam is overdesigned, a new concrete section is selected and the entire design procedure is repeated. 4-42.3. Large Deflections 4-42.3.1. Introduction Design of reinforced concrete beams for support rotations greater than 4 degrees depends on their ability to act as a tensile membrane. Lateral restraint of the beam must be provided to achieve this action. Thus, if lateral restraint does not exist, tensile, membrane action is not developed and the beam reaches incipient failure at 4 degrees support rotation. However, if lateral restraint exists, deflection of the beam induces membrane action and axial forces. These axial tension forces provide the means for the beam to continue to develop substantial resista~ce up to maximum support rotations of approximately 12 degrees. -'4-42.3.2. Lateral Restraint Adequate lateral restraint of the reinforcement is mandatory in order for the beam to develop and the designer to utilize the benefits 9f tensile membrane 4-,160
'qi 5,-1300/NAVFAC P-397/AFR 88-22, •.t,'
.~
.,
behavior. Sufficient lateral restraint is provided if the reinforcement is adequately anchored into, adjacent supporting ~embers capa1?,l,!, of res is ting the axial forces induced by tensile membrane action,
.,. Tensile membrane behavior should not be considered in the design process unless full external lateral restraint is provided . . Full lateral restraint means that adjacent members can ef'fectively resist 'a tot~i lateral force equivalent to the ultimate strength of all continuous reinforcement in the beam. This external resistan~e is more difficult 'to' realize fo~ beams than' for slabs due to th~ concentration of the. end ., reactions:..
'- ,"
'
4-42.3.3. Resi~tance ",Deflection qurve
'.'
. r '
The resistance-deflection curve for a .beam is the. same as that for a slab which is shown in Figure' 4-18. The initial portion of the curve is primarily due to flexural action (increased capacity due to possible comp~ession forces is not shown). ,At 4 degrees support rotation, the beam, loses flexural capacity. However, due to the presence of continuous re'inforcement and adequate lateral, restraint, ten.sile'membrane action deve Lope d . The resistance due to this action increases with increasing deflection up to incipient failure'~t approximately rotation. ' " . . . . 12 -,degrees support . .~ ~
"
• f
••
\.;
"',
•
In .o r de r. to' simplify the design' c a Lcu'l e t.Lcns , the res:i.st~nce is asswned to be
due to flexural action throughout ,the entire range of behavior (same procedure for siab calculations). To approximate the energy absorbed under the actual resistance-deflection curve, the maximUm support of the ,idealized is limited to 8 degrees. Design for this deflection would produce incipient failure conditions. For the design of a laterally restrained beam for 8 degrees support rotation, a type III cross-section is used to compute the ultimate moment capacity of the section as well as to provide the mass to resist mot Lon . The stress in the ,reinforcement,fds .would,be, equal to that correspo:~ding to support rotations 5 S em S 12 given in Table 4~2,. ,At every ' s ec t Lon ,throughout .tihe ,beam, the tension and compression reinforcement must; be cont.Lnuous in order. to develop the tensile'membrane action di~cussed below. 4-42.3.4. Ultimate Tensile Membrane Capacity, As can be seen in Figure 4-18, tensile membrane ·.resistance, is a function of de f Lec t Lon , It is also a function .,of the span length 'and th~ amount, of " ' continuous reinforcement. I The tensile membrane, resistance rt of ~ later~lly restrained beam at a deflect~~n X is expressed as: ,
.
"
rT - X in which'
[
8T L2
.
~,
I
.
-, , 4-152
"
.,
"
4-153
" :.
"
"
where r
c
tensile membrane resistance
4-161
,',
TH 5-l300/NAVFAC P~397/AFR 88-22 X
deflection of the beam
T
force in the contfnuous 'reinforcement
L
clear span
"
A sc
total area of continuo~s reinforcement
Even though the cepac Ltiy of a Lat e r a l.Ly restrained beam is based on flexural action, adequate' tensile membrane capacity must be provided, that is, sufficient continuous reinforcement must be'provided so that the tensile membrane resistance rt corresponding to 8 degrees support rotation must be greater than the flexural resistance r u' The deflection is computed as a function of the plastic hinge locations. The force in the coptinuous reinforcement is calculated using the dynamic design stress fds, corresponding to 8 degrees s~pport 'rotation (Table 4-2). " '
...
4-42.3.5. FlexUral Design'
i •
Since the actual tensile membrane' resistance-deflection curve is replaced 'with an equivalent flexural curve, the design of a beam 'for large deflections is greatly simplified. The design' is performed in a similar manner as for small deflections. However, sufficient continuous reinforcement must be provided to develop the required' tensile membrane resistance.' Thisreinforce'ment must be fully anchored in the lateral supports. 'Care must. be taken to ensure that the lateral supports are capable of resisting the lateral force T as given in Equation 4' 153. ' 4-43. Dynamic Analysis 4-43.1. Design for Shock'Load When a concrete slab supported' by beams is subjected to a blast load, the slab and beams act togetheito resist the 'load. The'beam-slab system is actually a two-mass system' and should be t re'aced as such. However, a reasonable design can be'achieved by considering the slab and'beams separately. That is, the slab and beams are transformed into single-degree-of-freedom systemscooipletely independent of each other and are analyzed separately, The dynamic analysis of slabs is treated extensively in previous sections. The equivalent single-degree-of-freedom system of ani structural element is defined'in terms of its ultimate unit resistance, r u' equivalent elastic' deflection XE and natural period of vibration TN' The ultimate unit resis-' tance is obtained from the table for one-way elements in Chapter 3 for the moment capacity given above, The procedures and parameters necessary to obtain the equivalent elastic deflection and natural period are also obtained from Chapter 3. Chapter 2 describes procedures for determining the dynamic load which is defined by its peak value P and duration T. For the ratios P/r u and T/T N the ductility ratio Xm/XE and tm/T can be obtained from the response charts of Chapter 3. These values Xm, which is the maximum deflection, and t m, the time to reach the maximum deflection define the dynamic response of the beam.
4-162
TIl 5-l300/NAVFAC P-397/AFR 88-22
A beam is designed to resist the blast load acting OVE!r the tributary area supported by the beam. Therefore, the peak value of l:he blast load P is the product of the unit peak blast pressurE! times the spaeing of the beams. and has the unit' of pounds per inch. In addition to the short term effect of the blast load. a beam must be able to withstand the long term effect of ~he resistance of tI,e element(s)'being supported by·:the beam when, the response time of theelement(s)"is equal to or greater than :the duration of the blast load. To insure against,'premature failure, 'the 'ultimate resistance of the beam must be gr-eace r than the reaction of the supported element (slab. wall, blast door. etc,), applied to .rhe beam as a static load. . In the case of a supported slab. the slab does. in fac t , act with the beam; a portion of the mass of the slab acts with the mass of the beam to resist the dynamic ,lo·ad." I t is. therefore.' recoriunended thii:t 20 percent' of the mass of the slab (or blast door. wall. etc.) on each side of che beam be added to the actual mass of the beam. This increased mass is then used to compute the natural period of vibration TN 'of the beam, I t should b.!'.,noted that in the calculation of TN the va Iues used for the effective mass and stiffness of the, beam depends upon the allowable 'maximum deflection. lJhen-designirig'for completely elastic behavior. the elastic stiffness is used while. in other
E is
cases, the equivalent eLas covpLas edc stiffness K
used. ~ The elastic value
of the effective mass is used for. the 'elastic range while ....in the elastoplastic range, the effective mass is the average of the: elastic and elastoplastic values. For small plastic deformations, the '~alue 'of the effective mas s Ls equai to the average oftlie':~quivalent e Las tLc value: arid the plastic value while for large plastic de fo rmatiLons , the effective mass is equal to' the plastic value. ....,; ,. i
4-43.2. Design for Rebound The beam must be designed to'. resist the negative de f'Lec t Lon. or rebound :which occurs after the maximum positive deflection has been reached. The n~g~tive resistance r-. attained by the beam when subjected to a triangular pressuretime load. is obtained from figu~e' 3-268 .in Chapter 3. Entering the figure with the ratios of x.,/X E and: T/TN'; previously determined for the positive' phase of design. the ratio of ,the required rebound resistance to the ultimate resistance r " /r u is obtained ... Th~ oeam'must be rei;'f,orced to withstand this rebound resistance ,r- to insure that the beam will remain elastic during rebound. • ," -'
,
.'~";
The ten'!ion reinforcemerit provided to withstand rebound'forces is added to what is' needed for the'compression zone • during the initial loading, phase. To obtain this reinforcement', ',the 'beam is essentially designed 'Eo'r 'a negative load equal to the, calculated value of r;. However. in no case shall the rebound reinforcement be less 'than'one-nalf'of,the positive phase reinforcement. The moment capac Lt Les and t.he rebound resistance capacity are calculated using the same equations previously presented. "il.-
4-163
w U
TOTAL FAILURE
ULTIMATE
Z
<[ INCIPIENT FAILURE
t-
(/)
...... (/)
I I I I
w
~
I
I
I I I I I
I
X( 11= 2°)
XE
I
I I I I X( 6= 4°)
X( 6 = 5°)
X(
e=
8°)
DEFLECTION
kD
Not.e 1
-,
®
PROTECTION CATEGORY
@ ~I G: SHELTER
PROTECTIVE STRUCTURE
BARRIER PRESSURE-
STRUCTURE-LOAD SENSITIVITY
TJ~IE O. IMPULSE
FAR OR CLOS -IN DESIGN RANGE RESPONSE CHAR S, ITERATION I'ROC DURE.
1M
UL.SE EQUATIONS
DESIGN METHOD LARGE
LIMITE D DEFLECllON CRITERIA FLEXURAL ACTIO
DUCllLE MODE OF RESPONSE
CLO ED TIES REQUI ED
TENSID
MEMBRANE ACTION
CLOS D TIES REQUIRED BRITTLE MODE OF RESPONSE CRUSHING
NONE FAR RANGE
::
SCABBING
SPALLING
CLOSE,...IN CRUSHING
CROSS-SECllON.TYPE
II
I
FAR RANGE I DR
I
::
POST FAILURE
SCABBING
FRAGMENTS
!II
1I DR III
III
rely ·1I4( fdu- dy)
Jl2( fel/ felu)
-
CLOSE-IN
fdy
DESIGN STRESS
Note i. Stirrups (closed ties) o.lwo.ys reqUired.
Figure 4-59
Relo. tionship between deSign pc r-cne t.er-s for beOMS
4-164
bt
.
~'
1"
~'
'"
"
-
,
..
~
2 ,
•
a SHEAR
Arrangement
(0)
. Av
.
-;
LOW TORSION
HiGH TORSION
Ai
3"
•
3"
•
Ai .3
Al
TORSION
(b)
Figure 4-60
a SHEAR
Arrangement
•
Web. Reinforcement
A's
+
•
•
FLEXURE
2
VERY HIGH TORSION
of Shear and Torsional
•
At
,>-
As
•
.: A~ A S +T
•
3"
•
TORSION
AI.
•
A~ , A S,+-3
+ FLEXURE
of Torsional and Flexural Longitudinal Steel
Arrangement of reinforcement for combined flexure and torsion
4-165
TK 5-1300/NAVFAC P-397/AFR 88-22 DYNAMIC DESIGN OF INTERIOR COLUMNS '4-44. Introduption The de~ign of columns is limited to those in shear wall type structures where the lateral loads' are transmitted through ,the floor and roof slabs to the exterior (and interior, if required) shear walls. 'Due to the extreme stiffness of the shear walls, there is negligible sidesway in the interior columns and, hence, no induced moments due to lateral loads. Therefore, interior columns are axially loaded members not subjected to the effects of lateral load. However, sig~ificant moments can resul~ f!om unsymmetrical loading conditions. 4-45. Strength of Compression Members (P-M Curve) 4-45.i.General The capacity of a short compression member is based primarily on the strength of its cross section. The behavior of the member ,encompasses that of both a beam and a column. The degree to which either behavior, predominates 'depends upon the relative magnitudes of the axial load and moment. The capacity of be the column can be,determined by constructing 'an interaction diagram as shown in Figure 4-61. This curve is a plot of the column axial load capacity versus the moment it can simultaneously withstand. Points on this diagram are calculated to satisfy both stress and strain compatibility. A single curve would be constructed for a given cross section with a specified quantity of reinforcement. The plot of a given loading condition that falls within the area represents a loading combination,that the column can support, whereas, a :.plotcthat falls outside the interaction curve represents a failure'combination: Three points of the interaction' diagram are used to define t~e behav~or of compression members under combined ,axial and flexural loads. These points 'are: (1) pure compression (Po' M - 0), (2) pure flexure (P, - 0, Mo)' and, (3) balanced conditions (Pb' Mb)' 'The eccentricity of the design axial. load for the condition of pure compression is zero. ,However, under actual conditions, pure axial loads will rarely, if ever, exist. Therefore, the maximum axial ~oad is limited by a minimum eccentricity, emin' At'balanced conditions, the r eccentricity is defined as eb while the eccentricity at pure flexure is infinity. The strength of a section is controlled by compression when the design eccentricity e - Mu/Pu' is smaller than the eccentricity under balan?ed conditions. The strength of the section is controlled by tension when the design eccentricity is greater than that for balanced conditions. 4-45.2. Pure Compression 'j
The ultimate dynamic strength of a short reinforced concrete column subjected to pure axial load (no bending moments) is given by: Po where:
0,85 f'dc(A g - As t) + As t fdy
Po
maximum axial load
Ag
gross area of section
As t
total area of reinforcing steel
4-166
4-154
TH 5-l300/NAVFACP-397/AFR 88-22
A member subjected to pure axial compression is a hypothetical situation since all columns are subjected to some moment due to actual load conditions. All • must be designed for a minimum load,eccentricity. tied and spiral columns This minimUm design situation is presented in a subsequent section. 4~45.3.
Pure Flexure
An interior column of a shear wall type structure cannot be subjected to pure fiexure under normal design conditions. For the purpose of plotting a P-M curve, the criteria presented for beams is used. 4-45.4. Balanced Conditions A balanced strain condition fo~ a column'subjected to a dynamic load is achieved when the concrete reaches 'its-limitirig strain of 0.003 in/in simul-, taneously'with the tension steel reaching its dynamic yield stress. f d. This condition occurs under .the ac t.Lonvof the balanced load Pb and the cor~espond Lng balanced 'moment 'Mb. 'At balanced conditions,' -the eccentricity of the load is defined .as eb' .and is given by:r,' . ", ,
.
~.
\!.
4-155 The actual values of,the balanced load arid'corresponding balanced moment are generally 'not required~ The balanced eccentricity is the important parameter since a comparison of the actual eccentricity,to the balanced eccentricity distinguishes whether the strength of the section is controlled by tension or compression. The co~parison of the actual eccentricity to the balanced eccentricity dictates the ,choice :o':f the appropriate equation for calculating the ultimate axial load capacity 'Pu' . --, . ~
.
'.
,
'
Approximate expressi~ti.s·have been derived for the balanced eccentricity for both rectangular and .circular members. These expressions, are sufficiently accurate for design purposes. For a rectangular tied column with equal, reinforcement on opposite faces (Fig. 4-62a). the balanced eccentricity is given by: (
r
eb - 0.'20h·+ (l.54mA s )!b
and: where:
As
,
4-157
-
.
'
depth of rectangular 'section
h b
4-156
.
balanced eccentricity
eb
,','. ,,'
(C,
•
,width of rectangular section area of reinforcement on one face of the section !
For'a circular section with spiral reinforcement (Fig. 4-62b), the balanced eccentricity is given by: .~ eb - (0.24 + 0,39' Prm) D and: 4-167
4-158
TN
5~1300/NAVFAC
P-397/AFR 88-22 4-159
where:
"
total percentage of reinforcement, Ast Ag D
total area of reinforcement gross area of circular section overall diameter of circular section
4-45.5. Compression Controls When the ultimate eccentric load Pu exceeds the balanced value Pb' or when the eccentricity e is less than the balanced value eb' the member acts more as a column than as a beam;, Failure 'of the section is initiated by crushing of the concrete. When the concrete reaches its ultimate strain, the tension'steel . has'not, reached its yield point and may actually be in compression rather'than tension. The ultimate eccentric load at' a given eccentricity e less than eb maybe obtained by considering the actual strain variation as the unknowri and using the principles of statics. However, equations have been developed which approximate the capacity of the column. These approximate' procedures are adequate for design purposes. For a rectangular tied column with equal reinforcement on opposite'faces (Fig, 4-62a), the ultimate axial load capacity at a given eccentricitY,is,approximated by: bh f'dc'
+ [e/(2d-h)] + 0.5
,4-160
(3he/d2) + 1.18
where: ultimate axial load at actual eccentricity e . e -
actual eccentricity of applied load area of reinforcement on one face of the section
d
distance from extreme compression fiber to centroid of tension reinforcement
h
depth of rectangular section
b
width of rectangular section.
For a circular section with spiral reinforcement,.the ultimate axial load capacity at a given eccentricity is approximated by:
'. - [
As t 3e
fdy
+
Ag
f' dc'
9.6D e . ,2 (0.8D + 0.67 Ds)
,+ 1.0
Ds
4-168
4-161
+ 1.18
]
TH 5-l300/NAVFAC P-397jAFR-.88-22
where:
As t
total area of uniformly distributed longitudinal reinforcement
Ag
gross area of, circular section -r ,
D
overall diameter of circular section
Ds
diameter of the circle through centers of reinforcement arranged in a circular pattern
4-45.6;· Tension Controls When the ultimate eccentric load Pu is less than the balance value Pb or when the eccentricity e is greater than the balanced value" eb' the member acts more as a beam than as a column._ Failure of the section is J.nitiated by yielding of the tension steel; The ultimate. eccentric load at a given eccentricity e greater than eb may be obtained by considering the actual strain variation as the unknown and using the principles of statics. However, again, equations have been developed to approximate the capacity of the column. It should be pointed out that while tension controls are a possible design situatio~ it is not an usual condition for interior columns of a shear wall· type s truc ture. ,..'
.
.
For a recta.ngular tied column with equal reinforcement cm opposite faces (Fig. 4-62a). the ultimate axial loa_d capacity at a given eccEmtricity is approxi- .. mated by:_ .
Ps -
where:~
.-;'
A
4-163
j bd
e' - e + d - (hj2)
4-164
m - f dy j (O.85f' dc)
4-165
I.:
p'percentage of'reinforcement on one face·of section e'
eccentricity of axial load at the ..end of member measured from the centroid of the tension reinforcement 4-169
TH 5-1300/NAVFAC P-397/AFR 88-22 For a circular section with spiral reinforcement (Fig. 4-62b), the ultimate axial lo~d capacity at a given eccentricity is approximated by:
.. Pu - 0.85 f'dc D2
[[
PTmDs (/2
. O.85e 0.38)2
(
D ""
'I-
.,
2.5 D
0.S5e ( "
D
where:
.
PT -
.
,
total percentage of reinforcement and is defined in Equation 4-159
4-46. Slenderness Effects 4-46'.1. General The preceding section discussed the capacity of short compression members. The strength of these members is based primarily on their cross section .. The effects of buckling and lateral deflection on the strength of these short members are small enough to be neglected.' Such members are not in danger of buckling prior to" achieVing their ultimate strength based on the properties of . the cross section. Further, the lateral deflections of short compression members subjected to bending moments are small; thus contributing little secondary bending moment (axial load P multiplied by lateral deflection). These buckling and deflection effects reduce the ultimate strength of a compression member below the value given in the preceding, section for short columns. In the design of columns for blast resistant buildings, the use of short columns is preferred. The cross section is selected for the given height and support conditions of the column in accordance with criteria presented below for short columns. If the short column cross section results in a capacity much greater than required, ,the dimensions may' be reduced to achieve an economical design. However, slenderness effects must be evaluated to insure an adequate design. It should be noted that for shear wall type structures; the interior columns are not subjected to sidesway deflections since lateral loads are resisted by the stiff shear walls. Consequently,' slenderness effects due to buckling and secondary bending moments (Pu) are the only effects that must be considered. ' 4-46.2. Slenderness Ratio The unsupported length Lu of a compression member is taken as the clear distance between floor slabs, beams, or other members capable of providing .,' lateral support for the compression member. Where column capitals or haunches are present, the unsupported length is measured to the lower extremity of capital or haunch in the plane considered. 4-170
TIl S-1301)/NAVFAC P-397/AFR 88-22 .,
The effective length of a column kLu is actually the equi~alent length of a pin ended column. For a column with pin ends the effective length is equal to the actual unsupported length (k - 1.0). Where translation of the column at both ends is· adequately prevented (braced column), the effective length of the column is the distance between points of inflection (k less. than 1.0). It is recommended that for the design of columns in shear wall type structures the effective length factor k may be taken as 0.9 for colwnns that are definitely restrained by beams and girders at the top and bottom. For all other cases k shall be taken as .1.0 unless. analysis shows that a lowe r value may be used. s s may For columns braced against sidesway, the effects of s Lenderne . . .be n!,glected when: ~,
kLu
.
< 34 - 12
4-167 H2
r
where: effective, length
k
~actor·
unsupported length of column radius of. gyration of cross section of co Lumn (r - 0.3h for tied. columns and 0.25D for circular <:olwnns)
r -
"value of smaller· end moment on column,positive if.memb",r is bent in single. curvature and negative in double curvature "\
..
value of larger end moment on column'
H2 .-
"
In lieu of a more accurate analysis, the value ot' 'HI IM2 may conservatively be taken equal to 1.0. Therefore, in the design of colwnns! the effect of slenderness may be neglected when: :5 22
4-168
r
The use of slender ,colwnns is not permitted in. order .t;o. avoid stability problems. Consequently, -the slenderness ratio must be limited to a maximum value of 50. 4-46.3. Moment Magnification Slenderness effects due to buckling and secondary bending moments must be considered in the design ofcolwnns whose slenderness ratio is greater than that given by. Equation 4-167. The r~duction in the ultimate. strength of.a slender colwnn is accounted for in the design procedure by increasing ,the design moment. The cross section and/or reinforcement is thereby increased above that required for a short colwnn.
A column braced against sidesway is designed for the applied axial load P and a magnified moment H defined by: 4-171
TK 5~1300/NAVFAC P-397/AFR 88-22 4,169 in which:", 6
-~----
P
1
,1
where: design'mome!1 t
moment magnifier value of larger end moment on column equivalent moment correction factor defined by equation 4-171 value of smaller end moment on colUmn'
design axial load critical axial load causing buckling defined by equation '4-172
..
The value of the moment magnifier 6 shall riot be' taken less than 1.0. For columns braced against sidesway and not subjected to transverse loads between supports, i.e. interior colUmns of shear wall type structures, the equivalent moment factor Cm may be taken as: 4-171
The value of Cm may not under any circumstances be taken less than 0.4. In lieu of a more accurate analysis, the value of Ml/M2 may conservatively be taken equal to 1.0. Therefore, in the design of interior columns, Cm may' be taken as 1. a.
,.
The critical axial load that causes a column :to buckle is given by: ".2 EI
Pc -
4-172
In order to .appIy Equation 4 -172, a realistic value; of EI must be obtained for the section at buckling. An approximate expression for EI at the time of buckling is 'given by: "
EI
4-173 1.5,
4-172 •
~
5-l3C10/NAVFAC,P-397/AFR 88-22
4-174
4-175 where: aver~ge moment of inertia of section f'
,:,
'
•
moment of inertia of gross concrete section axis, neglecting reinforcement
abo~t
centroidal
moment of inertia of ,cracked concrete section with equal reinforcement on oppo~ite faces F -
coefficient given in Figure 4-5
4-47. Dynamic Analysis l
'
Columns are. not. subjected to the blast loading directly, Rather, the load that a column must resist is transmitted through the roof slab, beams and girders. These members "filter" the dynamic effects of the blast load. Thus, in buildings 'designed to obtain plastic deformations, the dynamic load reaching the columns is typically a fast "static" load, that is, a flat top pressure time load with a relatively long rise time. The roof members and columns act together to resist the applied blast load. However, reasonable design can be achieved by considering the column separately from the roof members. The response (resistance-time function) of the roof members to the blast load is taken as the applied dynamic load acting on the columns.
a
Columns are
SUbje~ted to
an actual axial load (with
~ssociated eccentricity~
equal ,to the ultimate resistance of the appropriate roof members acting over the tributary area supported by ~the column. It· is re"ommended for design of columns the ultimate axial toad be equal to 1. 2 times the actual. axia!' load. This increase' insur:es that the maximum response of th.. column will be limited to a duct~lity ratio (~/Xe) of· 3.0 or less. I f the r Lse t~me of the load (time to reach yield for the appropriate roof J!lembers) divided by the natural ,period of the column is small (approximately. 0.1), th.. maximum ductiJity is, limited to 3.0. Whereas, if the time ratio is equal to 1.0 ,or greater, the column will remain elastic. For the usual design cases, the ratio .of.the rise time to the natural period will be in the vicinity of 1:0. Therefore', the columns will remain elastic or, at best, sustain slight plastic action . .
'l.,
In some instances: buildings may be designed to remain cOlllpletely ela';tic. In these cases, 'the actual axi~l load that the column must resist is equal to the maximum actual response of the roof mem~ers framing into the colUllln. This response and, therefore, the maximum load on the column, can be,.no more than two times the ,blast load. , , ' 4-173
TK 5-l300/NAVFAC P-397/AFR' 88-22 4-48. Design of Tied Columns 4-48.1. General Interior columns are not usually subjected to excessive bending moments since sidesway is eliminated by the shear walls. However, significant moments about both axes can result from unsymmetrical loading conditions. These moments ,may be due to unequal spacing between columns or to time phasing of the applied loads. As a result of the complex load conditions, the columns must be proportioned considering bending about both the x and y axes simultaneously. One method of analysis is to use the basic principles of equilibrium with the acceptable ultimate strength assumptions. This'methodessentially involves a trial and error process for obtaining the position of an inclined neutrai axis. This method is' sufficiently complex so that no formula may be developed for practical use. '. ~ An approximate design' method has been, developed 'which gives satisfactory results for biaxial bending. 'The equation is in the form of an interaction formula which for design purposes can be written in the form: 1
1
1
1
Py
Po
+ Pu
Px
,.
where:
I ';
Pu
ultimate load for' biaxial bending with eccentricities ex and ey
Px
ultimate load when
~
ultimate 'load when eccentricity e' y is
~
ultimate' load for a'concentrically loaded column (e x- e y- 0)
"
ecce~tricity
ex is present (e y pre5~nt
0)
(ex - 0)
,
Equation 4-176 is valid provided Pu is equal to or greater than 0.10 Po' The usual design cases for 'interior colUmns 'satisfy this limitation. The equation is not reliable where biaxial bending is prevalent and is accompanied by an axial force smaller than'O.IO Po' In the case of ~trongly prevalerit~bending, failure is"initiated by yielding of the steel (te~sion,controls region of P~M' curve). In 'this range it is safe and satisfactorily accurate to neglect the axial force' entirely. and to calculate the - section for, biaxial bending' , . only. This procedure is conservative since the 'addition ~faxial load in the tension controls ~egion increases the moment capacity. 'It should be mentioned that the tension controls case would be unusual and, if possible, should be avoided • in the design. Reinforcement must be provided on all four faces of a tied column with the reinforcement on~opposite faces of the col~n equal. In applying Equation 4176 to the design' of tied columns, the values of Px and Py are obtained from Equation 4"160 and 4-162 for the 'regions where'compression and tension control the design,respectively. The equations are for rectangular columns with equal reinforcement on the faces of the column parallel, to the' axis of
TH 5-l300/NAVFAC P-397/AFR 88-22
L··.
.
,
bending. Consequently, in the calculation of Px and.Py' the r~inforcement perpendicular to the axis of bending is neglected. Conversely, the total quantity of reinforcement provided on all. four faces of the column is used to calculate Po' from Equation 4-154. Calcu.!ation of Px' Py.and,P o in the:manner described will yield a co~s~rv,:,tive value,pf Pu from Eq'uation 4-176. : 4-48.2. Minimum Eccentricity Due to the possible complex load conditions that can result in blast design, all tied columns shall be designed for biaxial bending. If computations show that there are no moments at the ends of the column or that the computed eccentricity of the axial load is less than O.lh, the column must be designed for a minimum eccentricity equal to O.lh .• The va~~e of his the, depth of the column in the bending direct~on considered. The minimum, eccentricity shall apply to bending in both the x and,y dire~tions, simultaneous!y., 4-48.3: LongitudInal Reinforcement Requir~ment's
',
To insure proper behavior of a tied colum~. the longitudinal re~nforcement must meet certain restrictions. The area of longitudinal reinforcement shall not be less than 0.01 nor more than 0.04, times' the gross area, of the section. A minimum of 4 reinforcing bars shall be provided. The size of the longitudinal reinforcing bars sha~l not.be less thap#6 nor ,larger than #11. The use or: #14 'and .f118 bars .as welJ. as the y.se,,!>f bundLed bars, are not recommended due to p~oblems associated with the,developme~t and anchorage ,o~ ,such bars. To p~rmit proper ,placement of the.concret~. th~ mini~um clear distance between longitudinal bars shall not be, less thaIJ·'l. 5, times. the nominal, diameter of the .,. longitudinal bar-s nor 1.5 inches. ,,,:: , ,'.' ',_ ~.
4-48.4. Ci~sed Ties Requirements
,
Lateral ties must enclose all longitudinal ~ars in compre~sion to;-ins1J.re their full development, These ties must conform'to the following: '.
F
•
1.
Th~
ties shall be at least.#3,bars for longitudinal bars #8 or smaller and ,at least .#4 bars for #9 longitudinal bars.or greater.
2.
To insure the full development of the ties they shall, be closed ~ using l35-degree hooks. The use of 90-degree bends is not recommended.
3.
The vertical spacing of the ties shall not exceed 16 longitudinal , bar diameters, 48 tie diameters or .~ of ,~he,'least dimension of the column· section.
4.
The ties shall be located verti~ally not more than ~ the tie spacing above the top of footing or slab and not more than \ the I tie spacing below the lowest horizontal reinforcement in a slab or drop panel. Where beams frame into', a column', the ties! may be terminated not more than 3 inches below the lowest reinforcement in the shallowest of the beams.
5.
The ties shall be arranged such that every corner and alternate longitudinal bar 'shall have lateral support provided by. the corner of a tie with an included angle of not more than 135 degrees and
.
,
,
4-175
.. '
".,
TN 5-l300/NAVFAC P-397/AFR 88-22' no bar shall be farther than 6 inches clear on each side along the 'tie from such a laterally supported bar. The above requirements for the lateral ties is to'insure against buckling of the longitudinal'reinforcement in compression. However, if the section is subjected to. large, ,shear or torsional stresses, the closed ties mUst be increased in accordance with the provisions established for 'beams (see section 4-39). 4-49. Design of Spiral Columns 4-49.1. General Spiral columns may be subjected to significant bending mom~nts about both,axes and should, therefore, be designed for biaxial bending, However, due'to the uniform distribution of the longitudinal reinforcement'in the form of a circle, the bending moment (or eccentricities) in each direction can be resolved into a resultant bending moment (or eccentricity). The column can then be designed for uniaxial bending using Equations 4~16l and 4-166 for the regions where compression a~d tension control the design, respectively. 4-49.2.'Hinimum Eccentricity Since spiral columns show greater toughness than t,ied columns, particularly when eccentricities are small,"the minimum eccentricity for spiral columns is given as 0: 05D in each direction rather than O.lh in each direction foro,tied columns. 'The resultant minimum eccentricity for a spiral column is then equal to O.0707D. Therefore, if computations show that there are no moments at the' ends of a column or that the computed resultant eccentricity of the axial load is less than O.0707D, the column must be designed for a resultant minimum eccentricity of O.0707D. 4-49.3. Longitudinal Reinforcement Requirements To insure 'proper behavior of a spiral reinforced column, the longitudinal reinforcement must meet the same restrictions given for tied columns concerning minimum and maximum area of reinforcement, smallest arid largest reinforcing bars permissible and the minimum clear spacing between bars. The only difference is,that for spiral columns the minimum number of longitudinal bars shall not be less than 6 bars. 4-49.4. Spiral Reinforcing Requirements Continuous spiral reinforcing must enclose 'all longitudinal bars in compression to insure their full development. The required area of spiral reinforcement Asp is given by: ' f'dc 4-177
where:
-
,
area of spiral reinforcement' 4-176
.'
TK 5-l300/NAVFAC P-397/AFR 88-22
s -
pitch of spiral
D
overall diameter of circular section ,
,
diameter of the spiral measured through the centerline of the spiral bar The spiral reinforcement must conform to the following: 1.
Spiral column reinforcement shall consist of evenly spaced continuous spirals composed of 'continuous #3 bars or larger. Circular bars are not permitted.
2.
The clear spacing between spiral shall not exceed 3 inches nor be less than 1 inch.
3.
Anchorage of spiral reinforcement shall ,be provided by l-~ extra turns of spiral bar at each end.
4.
Splices in spiral reinforcement shall be lap splices equal to l-~ turns of spiral' bar.,
5.
Spirals shall extend from top of footing or slab to level of lowest horizontal reinforcement in members supported above.
6.
In columns with capitals, spirals shall extend to a level at which the diameter or width of capital is two times that of the column.
!
,. '
.
...
p.
-,
MAXIMUM AXIAL COMPRESSION STRENGTH
-,
" "
l1.
....., ,
P ( MAX.)
Z
e-n 0
I I
en
l&l
'.
e,
;I ~
:::E 0 c.J ..J
Cl:
x Cl:
COMPRESSION CONTFIOLS
I
0::
.
I
....
fIJI Pb
- I-I - - - - - - - - - - .-/
I I / /
//
./
BALANCED CONDITION
.,/
0'0 .,/ / ...... ./
/
......
.,/
/.
......
TENSION CONTROLS "I.
BENDI NG
Figure 4-61
Mb
MOMENT Mu
Column interaction diagram
4-178
. .'I ;;.
"
AS~ .. ,CLOSED TIE
d
h
.. ,. ,
..'.A~ s , .
,
\.
\
"
'
: '. •• 14 ••• :-.
., REO'D COVER .f (TVP.)
-a) RECTANGULAR· SECTION WITH EOUALREINFORCEMENT ON OPPOSITE· FACES "
As!
Os
Ds
SPIRAL REINF Asp
REO'D
COVER----'
b) CIRCULAR
SECTION WITH UNIFORMLY REINFORCEMENT
Figure 4-62
Typi~al
DISTRIBUTED
interior column sections
4-179
D
.
TN 5-l300/NAVFAC P-397/AFR 88-22 ' DYNAMIC DESIGN OF ExTERIOR 'COLUMNS 4-50. Introduction Exterior columns may be required for severe'loading conditions. These columns could be monolithic with the exterior walls and',as such would be subjected to .boch axial and transverse loading'. The axial load results ft~m the direct transfer of, floor and roof beam reactions while'the transverse load is due to the direct impact: of the blast, load. ' ,
,
,
The use of exterior columns would'normally be restricted to use in framed structures to transfer roof and floor be~ reactions to the foundation, Normally, only tied 'columns woJld be used since they are compatible with,the, placement of walland beam reinforcement. Exterior columns are not normally required for flat slab structures since roof and floor loads are uniformly transmitted to the exterior walls.' 4-51. Design of Exterior Columns Exterior columns are generally designed as beam elements. 'The,axial load on these columns may be significant, but usually the effect of the transverse load is greater~ The coiumn,will usually be in'the tension controls' region (e greater than eb) of the P-M curve (Fig. 4-61) where the addition of axial load increases the moment capacity of the member. Consequently, the design of an exterior column as a'beam, where the axial load is neglected, is conservative. Since an exterior column is a primary member which is subjected to an axial load, it is not permitted to attain large plastic deformations. Therefore, the lateral deflection of exterior columns must'be ,limited to a maximum ductility (Xm/XE) 'of 3.
'! ',I.
.
"
;
, "
.,
TM 5-l300/NAVFAC P-397/AFR 88-22
STRUCTURAL ANALYSIS AND DESIGN FOR BRITTLE MODE RESPONSE 4-52. Introduction The response of a structural element in the, brittle mode consists of ". that structural behavior which is associate4 with either partial or total failure of the .element and is characterized by two types of concrete fragmentation: (1) spalling (either direct spalling or scabbing) which is the dynamic disengagement of the surface of ,the element, and (2) post-failure fragmentationwhich is associated.with structural collapse. Spalling ·is usually of concern only for those acceptor systems' where personnel, valuable equipment and/or extremely sensitive .explosives require prqtection. Where the acceptor system consists of relatively insensitiye explosives so that fragment impact will not result in propagation of explosion or mass detonation, then post-failure fragmentation can be.considered in the design. For this latter case, even though the velocity of the spalls can be greater than the velocity of the post-failure fragments, the effects of spalling can be neglected because of the smaller masses involved. Post-failure fragmentation cannot be permitted when personnel are being protected. 4-53. Direct Spalling Direct spalling of a concrete eleme~t (Fig. 4-63) is the result of a tension failure in the concrete normal tG its free surface and is caused by the shock pressures of an impinging blast wave being transmitted through the element. When a shock front strikes the donor'surface of a concrete element, compression stresses are transmitted from the air to the element. This stress. disturbance propagates through the element in the form of a compression wave, and upon reaching the r!'ar (acceptor)' free surface, is reflected as a tension wave identical in shape and magnitude to the compressiorl wave. ·During the return passage, if the tension stresses in the reflected! wave'exceed the stresses in the compression wave plus the tensile capacity of' the concrete, the material will fracture with that part of the element: between the rear free surface and the plane of failure being displaced from the remainder of the element. A portion of the stress wave is trapped in the failed section and contributes to. its velocity. The part of the stress wave which remains within the main section continues to propagate with additional reflections and concrete fractures until its magnitude is reduced to that level below which spalling does not occur. Direct spalling generally results in the formation of small concrete fragments. The size of the fragments is attributed to the ncmuniformity of the shock wave (close-in effects) and the further distortions of the wave during its propsgation through the element (nonhomogeneous mat..rial, nonelastic effects, etc.). Localized .failures occur under the ac t Lori. of both flexural and shear stresses resulting. in the'rupture of the mortar binding the ·stone aggregate together. The failure zone propagates across the concrete surface forming a large number of comparatively small concrete fragments. The thickness of concrete between the rear (acceptor) surfaee of the element and the centroid of the rear face reinforcement is the usual depth of. concrete dynamically disengaged from the element. Although the concrete between the layers of reinforcement may be cracked to some extent, it is confined by the flexural and lacing reinforcement, thus preventing its disengagement. 4-181
TK 5-l300/NAVFAC P-397iAFR 88-22 The size of,the surface area which spalls depends upon the magnitude and duration of the applied blast' loads striking ,and subsequently being transmitted through the element, in addition to the size and shape of the element itself. For long cantilever-type barricades, only a portion of the wall will usually spall, since the magnitudes of the applied blast pressures decrease rapidly along its length, while for cubicle-type structures, the entire wall surface will usually spall because of the amplification of the blast pressures due to their multiple reflections within the structures. ' A wide range of velocities exists for spalled fragments. The'initial velocity at which spalled fragments leave a structural element has been found to be low' (50 feet per second or less). However, concrete elements subjected to the close-in effects 'of"a detonation are generally accelerating before or soon after spalling takes place. This accelerated motion of 'the element in turn' accelerates spall~d fragments. The fragment velocities produced by these acceleration effects may be as high as several hundred feet per second. For analytical purposes, an upper 'limit for the velocities of direct spalled fragments from elements sensitive to impulse may be taken as the initial velocity of the element which is also assumed to be the maximum velocity. However, for elements which respond to the pressure only or pressure-time relationship, an evaluation of the resistance-time and pressure-time curves
must be performed to obtain the maximum fragment velocity. The procedures and equations that are necessary to determine the above velocities are contained in Section 4-58. 4-54. Scabbing Scabbing of reinforced concrete elements (Fig. 4-64) is the end result of a tension failure in the concrete normal to its free surface and is asso,ciated with large deflections. In the later'stages 'of the ductile response mode of a reinforced concrete element, extremely large'deflections are developed producing large strains in the flexural reinforcement and, consequently, severe cracking'and/or crushing of the concrete perpendicular to the free surfaces. 'Because the tension,and compression strains are highest at the surface and 'since the lacing reinforcement in the later stages of deflection confines the concrete between the layers of flexural reinforcement, damage to the concrete is more severe at the exterior of the reinforcement than between the layers. The applied loads having· long since passed, the element is in a stage of deceleration at these'large deflections. Therefore, the velocities of scabbed fragments, which are equal to the yelocity of the element at 8 _ 5° (start of scabbing)" are lower than. the velocities of accelerated direct spalled fragments. However, the velocities of scabbed fragments also may be in the order of several hundred feet per second. Refer to Section 4-58 to determine the velocity of the element at a support rotation of five degrees. 4-55.
Predictio~
of Concrete Spalling ,
.
As previously explained, direct spalling is due,to,a.compression wave: traveling through a concrete element, reaching .the back face and being reflected as, a tension wave. Spalling occurs when the tension is greater than the tensile strength of the concrete. Spalling will occur: for: , s.l.O
when
~
4-182
1.0
4-178
TM
5-1300/NAV~AC
P-397/AFR 88-22
or for:
where: au V
-
-.
"
,0.1 f' c
"
h.
f.'
4-18<)
rl
(E c/p)1/2
4-181
and: re(lected pressure
Pr
peakno~a1
au
tensile strength of concrete
•
velocity of c,?mpression wave,through concrete.
V
:
ir
normal .ref1ected impulse
Tc
thickness of concrete element
f' c Ec P
'
,.
static compressive strength of concrete
_..
..
modulus of elasticity of concrete mass density of concrete ;~
.. ' ;
Equations 4-178 and, 4-.179 are shown 'graphically' in Figure 4-65 .:·Use of this figure . predicts the incidence of spalling, that .Ls " whether or not spalling' will occur a~ the point on the element which is subjec:ted to the peak normal reflected pressure Pr and impulse i r. .~,
J.
'r'
" _ ,
If spallingis predicted, the spalled area ,cannot be c:alculated ,from the above data. However, the,spa1Ied,area may be 'qualitatively estimated.by considering the distance'of the .p10tted point above the line on Figure 4-65. The greater the distance above. the ,line, the larger, the spalled area is likely to be.·' ,In addition, if the average reflected pressure Pbsnd impulse i b are substituted ,. for Pr and i r respectively, andspa11ing is still predicted, tnen the spalled. area will likely include the,major portion, if .not al1,.of the element~s surface.
j •
...
r'
..... '-'. ,
If a concrete element is subjected to.side-on.pressurcas, only, Figure.4"65 may still. be used ,to predict the occurrence .of : spalling;, In this case, 'the peak reflected pressure Pr and impulse ir'are ,replaced with peak side-on'pressure Ps o and impulse is, respectively.
.
\'.
4-56. Minimization of Effects of Spalling and
Scabbinl~ .s ,
~.
I ••
If it is determined that concrete spall.due to blast loading will- occur (using the procedures outlined, in the.preceding Section), or that scabbing will occur 4-183
.•
TK 5-l300/NAVFAC P-397/AFR 88-22 due to large deflections (>5°), then there are several procedures which can be utilized to minimize its effects. 4- 56.1. Design parameters The occurrence of direct spalling can be eliminated by an adjustment of the charge location, i.e~. if an explosive charge is placed at a sufficient distance away from the surface of an element, the magnitude of the blast pressures striking the element will be less than those which will cause tension failure of the ~oncrete. Large adjustments of the donor charge location in a design for the sole purpose of preventing direct spalling is usually not economically feasible. Although sufficient separation distance between a detonation and an element prevents direct spalling, its effect in reducing scabbing is negiigible. A reduction of scabbing is accomplished by limiting the magnitude of the maximum deflection of the element. By reducing this deflection, the strains.in·the concrete and reinforcement are lowered to a level where tensile failure of the concrete and subsequent scabbed fragment formation is prevented. Scaled·tests have indicated that scabbing does not occur when deflections are limited to values less than those corresponding to support rotations of no larger' than ' five degrees. To maintain the same response of the element, the resistancemass product of the element must be increased proportionally to ·the decrease in deflection, ;and this capacity (resistance and/or mass) increase results in ' an increased construction cost. 4-56.2. Composite Construction Direct spalling and scabbing can be eliminated through the use of composite elements composed of two concrete panels (donor and acceptor) separated by a sand-filled cavity. Spalling of the donor panel is not generally of'coricern since resulting fragments enter and are trapped in the sand fill. On the other hand, spalling of the receiver panel will endanger the acceptor system. By maintaining certain design parameters, both. direct spalling and scabbing of the receiver panel· can 'be prevented. To prevent the occurrence of direct spalling, the,high peak blast pressures applied to the donor panel of a composite element· must be attenuated by the sand fill ... This attenuating capability of the sand is attained by providing: (1) a thickness of the sand fill at least equal to twice the thickness of the 'donor panel where the panel thickness is predicated upon the required strength to resist the applied loads, (2) for one-way elements a ratio of the cavity thickness to span~length not less than 0.25 for cantilevers and O,05,for elements fixed on' two opposite sides shall be used, and for two-way elements each direction (span) shall be considered separately as shown above and the larger cavity thickness used, and (3) the sand density shall not be greater than 85 pounds per'cubic foot. Since scabbing is eliminated by limiting the element's deflection, scabbing of composite elements is prevented by limiting the deflection of the acceptor panel to the support rotation previously cited. Figure 4-~6 illustrates the use of composite construction to prevent spalling. The magnitude of the donor panel deflection.was such that the panel was near incipient failure, the panel experiencing the effects of both direct spallirig· and scabbing. On the other hand, the deflection of the acceptor panel has been limited and; ·therefore, had only minor cracking. Directspalled or 4-184
TK 5-l300/NAVFAC P-397/AFR 88-22
scabbed fragments were not formed on the exterior surface, of the acceptor panel. The use of composite barriers for the'sole purpose of 'eliminating spalling is not usually economically feasible. However, if the magnitude of,the:applied blast Loada warrant, the use of comp"osite, construction, tl)en the:eli.'llination of spalling can, be acha eved at asligl!t increase; in .cost by, conforming ;.to the, previously stated,element configur~tions and response. <' 4-56.3. Fragment Shields 4-56.3.1. General
,, Fragment shields are composed of stee~ ,plates or other structura~ material which can be attached to, or placed a short distance from, a, protective barrier (fig. 4·67). Unlike the other me,thods, the use, of shields does not reduce Clr deter the ,formation of ,spalled fragments but, rather .confLnes and prevents them from striking the acceptor system.
,
where:
.
a -
, .ru ,
I
4 c182 deceleration of the structural, element
,
.u'l t Imat;e unit resistance of conc ret;e element .
'.
,
IIlu -.
effective unf.ta.mas s of .the.concreto. element -Ln -che plastic ,,' ."::' range
The force acting on the shield is due to the inertial force of the fragments. Thus, the required resistance of the fragment shield must be equal to this inertial force,or:
4-183 4-185
TM '5-l300/NAVFAC P-397/AFR 88 c22 where: ultimate unit resistance of fragment shield mass of the spalled fragments When calculating'the mass of'the spalled fragments ,the actual mass of the' disengaged concrete is used, It is assumed that all the concrete from the rear face of the 'element to the centroid of the 'rear face reinforcement disengages. Thus, the spall thickness for a laced'section"is: 4-184 where: ds p
depth of spalled concrete thickness of the concrete element ,! '
average ,distance between the centroids of the compression and tension reinforcement An attached fragment shield usually consists of a flat or corrugated 'steel' plate spanning between angles or channels. The angles or channels act as beams spanning between anchor bolts. The' anchor bolts connect the~plate and beams to the,unspalled portion of the concrete. To insure that they do not fail due to concrete pull-out, the 'anchor bolts are hooked around the flexural reinforcement as shown in Figure 4-68. If the required resistance of the " , fragment shield is small, the fragment shield maybe' considered a two-way" spanning member without the supporting beams. In such cases, the shield is designed as a flat slab with the anchor bolts supporting it at multiples of the flexural bar intersections. For larger resistances, the plate thickness would become 'excessive and the design uneconomicaL 'J'
4-56.3"3. Separated Fragment Shield
,
'
If the barrier is' permitted 'to attain large deflections (greater than fivedegree support rotation) then the shields must be 'separated from the barrier. This separation distance should be sufficient to'eliminate the. possibility of impact between the deflecting barrier and shield. Fragment shields, separated , from the main protective elements and affording protection against spalled fragments from the walls and/or roof, are shown in Figure 4-67b. The shield may consist of structural steel, reinforced concrete, wood, etc., and must be designed to resist the impact and penetration of the spalled fragments as well as the overall motion of the main protective structure and any leakage pressure which may occur. ~ To design a separated 'shield; the mass and velocity of the spal1ed fragments must be determined. As an approximation" an average velocity for all spalled fragments can be utilized. The average velocity of the spalled fragments is taken as equal to the maximum velocity of the element (single-degree-offreedom system) so that: 4-185 where: v -
average velocity of spalled 4-186 '
frag~ents
TH 5-1300/NAVFAC P~397/AFR 88-22 ,. ib
blast impulse
Mu
effective unit mass of the concrete element in the plastic.' range
The impulse imparted to the fragment shield by the spalled fragments is equal to their momentum or: '
: 4-186 where:
"
~fs -
required impulse capacity of the fragment shield. mass of the spalled fragments (Section 4-56.3.2)
The shield is designed to resist this impulse, i f s' The cost of separated shields may be somewhat more expensive than shields attached directly to the barrier. However, the cost reduction achieved. by permitting the larger barrier deflections may offset the increased cost of the separated. shield. "
.
4-57. Post-Failure Concrete Fragments When a reinforced concrete element is substantially overloaded by the blast output; the element fails and concrete fragments (post failure) are formed and displaced at high velocities. The type of failure as·well as the size and number of the fragments depends upon whether the element has laced or unlaced reinforcement. Failure of an unlaced element (Fig. 4-69)' is characterized by the dispersal of concrete fragments formed by the cracking and displacement of the concrete between the donor and acceptor layers. of the reinforcement. With increased deflections,·these compression forces tend· tOe buckle the'reinforcement outward thereby initiating the rapid disintegration of· the element. Laced concrete elements exhibit a different type of failure from unlaced, elements, the failure being characterized by reinforcement failures occurring at points of maximum flexural stress (plastic hinges) with the sections of the element between the'points of failure.remaining essentially intact. When fracture due to excessive straining of.the tension rel.nforcement occurs at the positive yield lines, some small concrete fragments wl.ll be formed· at the' acceptor side of the barrier. Quite often, if the overload is not· too severe. the' compression reinforcement at'the hinge points does not'failsandthereby prevents total disengagement of the sections' between the hinges', (Fig. 4-70). In cubicle type structures where contl.nuous laced and flexural reinforcement is used throughout, failure I.ssometimes initl.ated at the positive yield lines where flexural and lacing reinforcement fa 1.1 , whl.1e at the supports; only the. tension reinforcement fails. The intact sections bet"een the 'failure points rotate with the compression reinforcement at the supports, acting:as the mechanical hinges of an analogous swinging door (Fig. 4-71). The compression reinforcement at the supports, serving as hinges, produces rotational rather than translational motion'of the failed sections; and energy which would ordinarily result· in translational velocities is transferred' to sections of the structure adjacent to the failed element. In other situations, where
4-187
TH'5-l300/NAVFAC'P-397/AFR 88-22 there is a larger overloading of the element, the failed sections of the laced element are completely disengaged and displaced from the structure. The translational velocities of these sections are usually less than the maximum velocity of the element at incipient failure .. 4-58. Post-Failure Impulse Capacity 4-58.1. General Elements which protect non-sensitive explosives may be designed for controlled post-failure fragments with a substantial cost savings. These elements fail completely, but detonation is prevented by limiting the mass and velocity of the fragments. Barriers and'shelters which will protect personnel, equipment and/or sensitive explosives cannot be designed for post-failure criteria. Procedures are presented below for determining the post-failure impulse capacity of laced elements. 4-58.2. Laced Elements 4-58.2.i. General The idealized curves of Figure 4-72 illustrate the response of an impulse-' • sensitive two-way element when the applied blast impulse load is larger than it flexural impulse capacity (area under the resistance-time curve, Fig. 4-72a). The assumptions made in these curves are the same as those for impulse-sensitive systems whose response is less than or equal to incipient failure (Vol. III) namely: (1) the element prior to being loaded is at rest, and (2) the duration of the applied blast load and the time to reach to yield are small in comparison to the time to reach the ultimate deflection. In Figure 4-72 the duration of the applied blast load and time to reach yield have been taken as equal to zero so that the element will respond and reach its maximum velocity instantaneously (i.e.,. at to - 0, Vo - ib/mu). When the element is designed to remain'intact (equal to or less than incipient failure conditions), its velocity at time tu (deflection Xu) is equal to zero .. However, if the element is overloaded, then the velocity just prior, to failure is a finite value, the magnitude of which depends upon: (1) the magnitude of the overload, (2) the magnitude of the flexural capacity, and (3) mass of the . element ... When laced elements are overloaded, failure occurs at the hinge lines and the I element breaks into a small number of large sections. The magnitude of the velocity of each sector at failure varies' from a maximum at the point of maximum deflection to zero at the supports. The variation of the fragment velocity'across.a section produces tumbling .. This tumbling action may result in an acceptor charge which is located close to the barrier being stuck by that portion pf a failed sector traveling at the highest velocity. If the acceptor charge is, located at a considerable distance from the barrier, the velocity of the fragments'should be taken.as the translational velocity. The •. translational velocity is approximately equal to the average velocity of the sector before.failure (the average momentum pf the element before· failure is equal to the ~verage momentum after failure).
.. .
The analytical relationship which describes the response of a laced .element,. both in its flexural' and post-failure ranges of action, is obtained through' a
..
4-188
TK 5-130Ci/NAVFAC P-397/AFR 88-22 "
semigraphical solution of Newton's equation of motion similar to that described for incipient failure design in Chapter 3. If the. areas under the pressure-time and resistance-time curves' (Fig. 4-72a) are considered to be positive and negative, respectivel.y, and the velocity of the system before the onset of the load is zero, the slwmation of the areas at any time divided by the 'appropriate effective mass for each range is equal to the instantaneous velocity at that .time. The velocity vi at the incipient failure deflection ~ (time, ~) may be expressed as: 4-187 roup where the values of i b, r u' tl and t u are defined in Figure 4-72 and rou and roup are the effective masses of the single-degree-of-freedom system in the various flexural ranges (ultimate and post-ultimate) of the two-way element. The acceptor charge is assumed to be close to the barrier, so that the maximum velocity of the fragment after failure vf is equal to the maximum velocity of the element at incipient failure vi' The expression for the deflection at any time may be found by multiplying each differential area (between the time to and the time in question), divided by the appropriate effective mass, by the time which is defined by the distance between the centroid of the area and the time in question,' and adding these values algebraically. Using this procedure and the expression for the deflection at partial failure (initial failure of a two-way element) the deflection at.time t l is: 2 i b 'tl r u t1 " 4-188 while the equation for-the deflection at incipient failure at t u is given by:
,
~
,"
ib tu
mu
,r u tl
--"'U
[
tl t ,u -
4-58.2:2. Post-Failure Impulse Capacity
- 2
]
r up (t~ -- tl) 2
4-189
2~p"
•
The expression for'·the' blast overload impulse capacity of a two-way element includes both the flexural capacity and the post-failure fragment momentum portion of the element's response. Solving Equations 1.-l~7through 4-189 simultaneously gives the expression for the blast overload impulse capacity. For a two-way element the resulting equation is: i
b
2 4-190 2
For a one-way element or a two-way element which does not exhibit a postultimate range, the blast overload impulse capacity is: 4-189
TN 5-1300/NAVFAC P-397/AFR 88-22 ib 2
IDu
vf2 2
4-58~2.3.
Response Time
The response time is the time at which the elements reach the·ultimate deflection Xu, and failure occurs. 'The expression for the response time tu is found by solving Equations 4-187 and 4-188 simultaneously. The response time for a two-way element is:
IDup
4-192
The response time of a one-way element or for a two-way element which does not exhibit a post-ultimate range is:
4-58.2.4. Design Equations The basic equations for the analysis of the blast impulse capacity of an element are given above. However, the form of these equations is not suitable for design purposes. The use of these equations would require a tedious trial-and-error solution. Design equations can be derived in the same manner as for elements designed for incipient failure or less (see Sect. 4-33). If Equations 4-190 and 4-191 are compared with Equations 4-95 and 4-96, it may be seen that except for"the right-hand term in each of the above equations, the corresponding analytical expressions are ,the same if Xu is substituted for Xm The additional term in the above equations is the kinetic energy of the fragments after failure. Because of·this similarity, Equations 4-190 and 4-191 may be expressed in a form which will be a function of the impulse coefficients (section 4-33), the geometry of the element, applied blast impulse and a post-failure fragment coefficient. The resulting equation is: .
+
".4-194
.
"
where ib
applied unit blast impulse
PH
reinforcement
dc
distance between centroids of the compression and tension reinforcement
rati~
in the horizontal direction"
"
4-190
TM 5-1300jNAVFAC P-397/AFR 88-22
f ds H
dynamic design stress for the reinforcement span height
vf
maximum velocity of the post-failure fragments
Cu
'impulse coefficient for ultimate deflection
Cf
post-failure fragment coefficient
x.,
Equation 4-194 is applicable to both one-way and two-way elements which are uniformly loaded. Values of Cu can be found in Section 4-33. The postfailure fragment coefficient takes into account the variation in ,the effective mass. It is a function of the element's horizontal and vertical reinforcement ratios, aspect ratio and boundary conditions'. To facilitate the design procedure, values of the post-failure fragment coefficient Cf have been plotted in Figures 4-73, 4-74 and 4-75 for two-way elements supported on two adjacent edges, three edges and four edges respectively. For one-way elements, the value of coefficient Cf,is a constant (Cf - 22,500). 4-58.2.5. Optimum Reinforcement The optimum arrangement of the flexural reinforcement in two-way elements
designed for post-failure fragments will not necessarily be the same as that for similar elements which are designed for incipient failure, damage or less. The optimum ra~io of the vertical to horizontal reinforcemen~ Pv/PH will be a function of the ,amount of the blast, impulse absorbed through the flexural action of the element in,comparison to that which contributes to the momentum of post, failure fragments. Unlike incipient failure design, the optimum ratio will vary with a variation in the depth of the' element. As discussed earlier, the optimum depth of an e1ement,an~ total amount of reinforcement PT is a function of' the relative costs of concrete and reinforcing steel. In a given situation, the designer must establish, through a 'trial and error 'design " procedure, and; a cost analysis, the" optimum depth, reinforcement ratio Pv/PH " ,and total, amount of reinforcement PT", For very large projects. this type of detailed analysis may result significant cost benefits, but for most projects using the values of Pv/PH and PT recommended for incIpient failure design will yield an' economical design.
4-191
(
f· \
4-1.92
(
4-193
100
50 I II I 'I! II tIl1l! 30
10
5 ~
,..
.t5'
0
3
<,
'"
~
~
a,
.5 .3
.111
•.01
.03
.05
.10
:3
.5
3
5
10
30
50
100
irv/P.Tc -\
Figure 4-65
e
Spall threshold for blast waves loading walls
e
e
(
(
(
Figure 4-66
Unspalled acceptor panE!l of composite panel
'.
,
PROTECTIVE BARRIER
~
....
'" '"
...
F I
MAXIMUM BARRIER DEFLECTION
t
f
\
II"-SHIELD
PROTECTIVE BARRIER--"""
I
'-SHELTER,
I ~. I ,
>
\DON IR SYSTEM)
'-SHELTER,
//////////
e
PROTE :T1VE BARRI R--'"
\1g..,.SHIELD
\PONOR SYSTEM'""'\
"
\ \
\
-
M4XIMUM ROOF DEFLECTION !'-SHIELD
,,I
. .I I ~
'"SHELTER,
//////////
b 1 LARGE DEFLECTIONS
LIMITED DEFLECTIONS
Figure 4-67
":""'----
\
\..DDNOR SYSTEM"')
0)
MAXIMUM BARRIER DEFLECTION
Sheilding systems for protection against concrete fragments
e
e
I
J, ' 4
.. •
.
SHIELD
.
•
'
~ ....
It
·
I (
LACING/REINF. FLEXURAL REINF.
V
V~ ..
~ ;...
.
• -
Ii 4·
II
Figure 4-68
.
.
-.
-,.
.
-
K!cl lD
.J cl
cr :>
'
,
x
.
!oJ
.J
lL lL
•
.'"
!-'-+'
C)
z
~
.
.
~
V>
.
.
.
1-,
0
~
u
. ...· ...·. ... . ·, . .. .
•
:z
,.
. ~
'
.::-. '"
-.
t
'
cl
~
.
V
..
.
.. x
~;
.
~
I"~'~
II III
,..-.c''''
,
'
rANCH~R BOLT- ,
'-
.
-.
•
•
,...
.
~,
..
..
4
•
..
'.
t.
,
II
..
-
Rigid attachment of fragment shield to barrier
4-197
Figure 4-69
Failure of an unlaced element
4-198
(
.
~,'
4-199-
''t'-\''
.pI ~
o o
Figure
4-71Bac~wall
fail\lreOf
r>.
(
\
-
e '.
-.p
-
l~.
.,~,
•
.-
:•.n. 0 ... ... (,) 0: 2
i
:::>' ",t;" II en_ ... en
«w
.. 0:
...o
II
b
'u
i
'u
,•
~"I-_------"';""'-..;;...-"t.. , - . -:-__ I
,
'0 I
'
'u '
'u
. I
'up
X,
IINCIPIENT FAILURE
I
'-"
~.'-
Xu
TIME
"",
0) IDEALIZED PRESSURE - TIME AND RESISTANCE - TIME CURVES
o ....'"
Cl IDEALIZED RESISTANCE - DEFLECTION CURVE " ~
'. ;..'
-.
·0
i& -, unit blast impulae r • ultimate unit .resistance u
>' !: (,) o
,INCIPIENT FAILURE
...> ..J
y:_y
'0
t
'u
TIME bl IDEALIZED VELOCITY - TIME CURVE
Figure 4-72
~-.
t'up - p08t-ult~te 'unit resistance _ :" to ·.duration of po~itive phase of bl8Bt'pres8u~e t ... time at which partial failure occurs :--. 1 t • time at which ultimate deflection occurs vf • post-failure fragment velocity Vi • velocity of element at incipient failure Vo • initial velocity of element
Xl - partial failure deflection XU --ultimate deflection
Idealized curves for determination of post-failure fragment velocities
-
...EICl
-.
.5
I
.,:,,'d.::I:::I:': .r.. J 2. 201' ··I·'··! . . ,. "'1: .
"1,"
. .., .1 .
"""..,.. ,.. ·,· . ,j··,.. IjLL\_:_~:·~l:":i:t:::;lL.tLl:unl::{~U~ . .. '1'
-,
2.00
,
,.,
·!
,
,
I.I I
; I,
2.10 .---.-.-.
N.o
TO
'I', '·I.;··"·"l .1.,1 .!
-t-~.
··1 ' .,
,·,·, , ,..
....
U
IZ
I&l
U ii:
I&. I&l
,
~
8
t,)
0
l-
t,)
z
I&l
2
C!I
of
a::
I&.
I&l
a::
:::> -J
~, ,
IUI
e
1.20' .
I
i
1
0.25
1 ., .. I
, I·'
0.5
1
1
1.0
2.0
1
I'
I
4.0
RATIO· OF VERTICAL TO HORIZONTAL REINFORCEMENT. PV/P H. Figure 4 .. 73
e
Post-failure coefficient C for an element fixed on two f adjacent edges and two edges free
e
e
e
• -
.•
EI'".s
N 1,
~ -...-
.
.
'Q '
U IZ I&l
U
k:
,
.0N 0
...
IL
8 IZ
I&l
2
(l)
4(
If I&l II:: ::;)
....
:,
lll)
,,0' 0..
•
2.20 ,!:;, i t.: I;: I; .,,;;: .... • • :; r : 1'1:, ::" '1>,';' In!'' it;; :;:I'Ii.: '::. :it:' . .' : ;,; , ,.:' '" I, ,:.:: :: ,'. Ii;: . ;I!::::::,;::O:,,,,.,,.,,,:!:," '. 210 . " ... ,,,,,1'.,,, 1;".'.1': ''',''" :::: , ,.'"..,., ,··...' ··1" , ...: · . .' .... .,.I;:''1:''''''T'' ....
';::: : "I; ': ." .;: . ';:'
.. .. ;: I·'ii··'·'···" r: I: " ! . ; ·"''-'I!';:!!·.'.' : " ,., Ii.,:: "",::.,,:1., ::: I; 'TfTTTT ., .. ..•,.::r: i::! :if'!::::: " . •. .,',iT' .j.:.'.+i"i. ::'. I' , •.' ,-;' i 'i; I iii:l::l "T.,
r;±rIHHiL:j.:. : : . t UHf :::'i;::n::,::;:':., ;: :,': 2.00 : : .: ,:1 ,.. ;1';1': : {'IEer,miff:
IF I ; : .;
1 I ... ;
,!,: i: ;: II: .: F" :'! 11:' 1.90 tiT'..... : .. i".':.I,'. " •. ' i: 1 ; ;:.1
1.80
rn1 I:
j'.
i
"'..,
f f Ii!"
U Il;,!Ji" .•
-,
.
:I,·,:+7J ..
IT!;,
,,:'T! !'; i IT. i ! j ITGlnE ::., I : i
hi" .
.' : u:; ' 1::!Iil" '. : ';:.: .':'.1': ['! :..,:, ; : i ;iii . ".il"::.;:!:.: ~,,- tn+:"" H:'!, ~ ll,
.; , ; ,:
it::
i :
:i~" 1' " .
t :
~
~n-rir: n::'';-:i: Fi"" .':••• I::::: Iii I: F'...,; 1ii)'.... ::;......'..... f 1~M:.:.I.:~ .7 ; .~.l ;c, ~~... '; =......rrrr.:.' i:. ;~,.:. pm}',· :
I:,: ,:.::.::: I
:~.
I T ' : .:.:::::1' :':; ".; .".1.,
~60
,,:: H,' I::! :.:: I':!,":;'!!:LI. i , :' ':' ' [ ' ; ' \-\ " ...".0 .
" : ...::..il"!:i\.... ,,.!I'......' . . '.• I:'
H-t'T:I: ..'.'
u..... :.; ;
:;1'.
~'c
,:Ii, .,,,,.,' .,'. .. ..::::'1"::,,::., ....... , , , , ' l : I . : I , . ,:: " . ".... ,,,. ,..,"'"lii.."... ......-:.: ' " , " '.". 1701"" :"'I!q C+ .ii .:i:: ' . ;:, k..::I;;:: :: :Ii"::::' ..,'il,::;:1""··:':·: iffl \..
. 0
.
..
; i I 1i::1 .' :', ..... 'm'T
' l . '. 1\-\~''"LrF Lili';" ~.,:-;; '. l f- fiU ,~:!!t: " ,,·.rr:;:
,,~'"
'Ii 'Ii" l': ,.,.•." :: . . . i : I~'.,: ' " ,:",,,,,' "P' , "" ,",.""" '"::~,:,,, ' , , '" "", /\-\'- \.. ,[""k::;":;';!":"'::' -~:: :,:: Ii,' · ..•.... : ,. \.../.~;'.....,..,:: ti::... "i' l 1.:..J..!. I::~ ." .... ,.,,' ".",.' r ': ,, i..l " ' ,": : I::.,: j ,'f, .!.:~ 160 ., '!iN Ji II ,'iHbl?,i;: ;' j' : i ' ' ' : A T ' : f; i i ,.. .. :i Ii:i ' ;: : :" I \.6..-. ..'.• .' l.o'":.1 I 50 !, yL(;.. K 1:.:i',· ... !",;b:;:li'f:. ::::10'1."......n .: .:." : .""'1": .• .. D." · 1/ • . ' " . ".,,:::':" I':,!",':::';':::::'''b :: "TIC H. ::I''N: ,;,
; ,
:>.:).;~! .... ,I"~ilf.'H,::'y.'i::,'::::::'
~~ 1':,:1:;;;;; !~;H,~lo""""" .', ,,;;,\,:;:'; ii' I~~ '. ::,~i-:+ ; TT F: !iT! •'.', ;;],;ll,' jj:',js>'.H", " Iii: 1i'f!: \.. . !: :: i ,:: 1.40 K:: .:I,.,:Iu..... ~.;rs.:.:SL :1::r:.:::::::'.'::::..· ..• 'i.::".'. :.:.",;.10"." jl, .,[-:T1".>·~;"· :; 'H.;,i:Pf ::!' '::1/::; ~. •:i' i;; I' :JX(: . -:-,-t O·./. "'r! I·.;,,· tYN-: TC FC·.;.;!!U!,; ~j:i . .:. :X'. 1/ ,':: •. \)~ ... / : : ·1 1.30 :,;.IX: . ;iili~: '...:. 'iii '.' :_;..,,~r:":::: : I: :A. [f!'! :::1 V:" . 'K.'i,;,;"c, , .' ., ... ::v:,~I;;··"':,:, i':"""""
x·
j
:!.....:
:;'L!iY;:::t,' I,.:"';:::::::'''''· .. " ' ' ' ' ,":':' z" I! i1::'" 1•:. ' "I· i f; :i i! . . ", ,::j'" ;U ::i' .... "",. X' ::; :1 ,'K:D !" :., ".: .200.25..... j" ... .•• ,:':."".,.",·.1... ,'·....... . ' ''1':. , '.' .:.. :'N':.' i '' : ;:.i::· ..'. ... ::,J!r::""'" .: . l . ,:,.:! ':.'!::!: f.ii'!···· ' .., : ,.:,,;-:1: j.:; :'. " :: ; Ii : !: • :i: '.' .. :::
I
1
0.5
1.0
2.0
4.0
RATIO OF VERTICAL TO HORIZONTAL REINF.. P /P v H Figure 4-74
Post-failure coefficient C for an element fixed on three f edges and one edge free
-
••
EllDC
-
N
I
.-
-,
,I
2.20
I
.•
2.10
.0
'
,
2.00 " .. ,
T Q
>C
....
u
IZ
1&1
u
...,,
~
0
1.80
It 1&1
1.70
'IZ
1.60
8
..,
1.90
1&1
2
(!)
« II:: u,
1.50
1&1 II:: :::l .J
1.40
-e' IL •
1.30
I-
fI)'
e
1.20 0.25
,
0.5,
1.0
2.0
4.0
,RATIO, OF VERTICAL TO HORIZONTAL ~EINF:, PV/P H Figure 4-75
~ost-failure
coefficient C
edges
e
e
f
for an element fixed on four
e
TH 5-l300/NAVFAC P-397/AFR
88~22
,
STRUCTtJ!lALBEHAVIOR TO PRIMARY FRAGMENT IMPACT 4-59. Introduction 4-59.1. Fragment Characteristics Detonation of cased explosives results in. the formation of primary fragmen,s due· to the shattering of the casing. These fragments are us~ally small in size and initially travel at velocities, in the order of thousands of feet per second. Upon contact with a barrier, the fragments will, either pas~ through (perforate), be, embedded. in (penetrate with or' without spalling), or be deflected by the barrier. The resulting effect is dependent; on the interac-, tion of the following 'factors: ' (1) the magnitu~e of the initial velocity, . (2) the distance between the explosion and the barrier, (3) the angle·at which the fragment strikes the'barrier (angle of obliquity), a~d (4) th~,physical properties of the fragment (mass, shape and material strength) and the barrier (concrete strength and ,thickness). 4-59.2. Velocity and Impact Limitations This section deals with ~he situation. where a.fragment with given properties strikes a barrier element with a known velocity and where the angle of obliquity between the traje6tory path ,of the fragment and a normal to the surface of the barrier is zero degrees (normal impact). The striking velocity of the fragment is assumed equal to its initial velocity for a detonation in close proximity to. the element or is determined according to the procedure in Chapter 2 for cases where the fragment travels a distance greater than 20 feet. . 4-60. Fragment Impact on Concrete 4-60.1 .. General When a primary fragment strikes a concrete barrier, penetration resisting pressures, in the order of thousands ~f psi, act on' the cross-sectional area of the fragment. For a given mass of the fragment, as the striking velocity is increased, the resisting pressures also increase, while for an increase in the cross-sectional area of the fragment, the resisting pressures decrease. If the fragment can withstand these' pressures acting on its,frontal surface, then the amount of penetration will be governed 'by its mass, shape, and striking velocity. On the other hand, if the fragment deforms under the applied loads, then the resisting pressures in the c9ncrete become effective over an increased cross-sectional area, thereby reducing,the possible penetration for a given available kinetic energy of the fragment. Generally, larger penetrations may be expected with less ductile"metals, such as fragments from armor piercing casings. As,afragment ·impinges on a wall surface, a section of the wall adjacent to the point of impact spalls, forming a crater around .the impact area (Fig. 476). 'This crater is conical in shape but irregular. As the striking velocity of the fragment increases,· the size of the crater also increases. At small velocities, the increase in the crater size for a given velocity increment is more rapid than at high~r velocities in the order of several thousand feet per second. At a striking velocity of approximately 1,000 feet per second or less, the fragment does not usually penetrate beyond the depth of the crater, 4-205
TN 5-l300/NAVFAC P~397/AFR 88-22 .whi~e
for larger velocities, the fragrnentpenetrates beyond the front of the crater and either lodges within or passes through the barrier. A crater similar to that formed on the front face of the barrier is also produced on the back side if the kinetic energy of the fragment upon impact is sufficient to produce excessive tensile stresses in the concrete.
As in the
case of the front-face crater the size of the crater'on the back'£ace in- . creases with an increase in striking velocity of a particular fragment. The back-Face crater'is generally wider and shallower t.harr-che front-face crater, though again the surface of the crater is irregular. Quite often, the kinetic energy afforded by' the striking fragment is orrl.y sufficient to dislodge the concrete on the exterior side of the rear face reinforcement. In this case,, •. -" f , . ';. only spallingwill occur. As the striking velocity is increased beyond th~, limit to cause_spalling, the penetration of the fragment into the slab" increases more and'moreruntil perforation is·attained~ - 'J'
4-60.2. Penetration by Armor-Piercing Fragments A certain amount of experimental data, which is analogous to primary fragment' penetration, has been accumulated in connection with projects to determine the effects of bombs and projectile impact on concrete structures. This data has been analyzed in order to develop relationships for the amount of fragment penetration into concrete elements in terms of the physical properties"of both the metal fragment and the concrete. A general expression for the maximum penetration into a massive concrete slab (i.e.'a slab'with infinite thickness) by an armor-piercing fragment ,has been obtained as follows:: Xf - 4.0x 10. 3 (KND)0.5 d l. l v s 0.9
for Xf
X~ 4.0 x 10- 6 KND d l. 2 v s l. 8 + d
for Xf > 2d
:5
4-195
2d,
4-196
and K - 12.91/(f· c)1/2
4-197
where:
Xf
penetration by armor-piercing steel fragments penetrability constant '. .
K
,
N'-
.' nose shape factor as 'de f Lnad in Figure 4- 77
:' D
cali~er
d
fragment diameter'
vs
striking velocity
density as defined'in Figure 4-77
For the standard primary fragment and concrete strength f'e equal to 4,000 psi, Equations 4-195 and 4-196: reduce to the following equations in ternfs of, fragment diameter (in) and ,the striking velocity s (fps): •. , ,
v
Xf - 2.86 x 1O-3· d 1. l v s 0.9
for Xf
2d
'4-198
Xf -' 2,04 x 10- 6 d1. 2 v s 1. 8 + d
for' Xf > 2d
4-199
..
:5
. 4-206
,r..
TK
5-l300/~AVFAC
P-397/AFR 88-22
Equations 4-198 and 4-199 can also, be expressed in. terms of fragment weight (standard shape) as: . -,.-
.,
0,37 v 0.9 for Xf 1.92 x 10- 3 101 fi s 101 0 . 4 v 1.8 + 0.695,W 0.33 Xf-, 1. 32 x 10-6 .f· , s f Xf
s
2d
.jt
! ~
•
4-201
for Xf >. 2d
••
II " - '.
4-200
,..-
I
'"
Figure 4-78 is ..: p Lot; .of the'maximum pen:etr'ation of a s1:and;'id fragment through 4,000 psi concrete for various fragment weights arid"str1kihg velocities. Maximum penetration of fragments in concrete of strength~_.other than .4,000 psi, may be calculated using the '\(alue~. of Xf·f.rom. Equadon 4- fqo ,,4-201 or Figure 4-78 and the following equation:
;
• >'.
.4- 202
.,
where: Xf X'f
maximum penetration into (>,000. psi concrete maximum penetration into concrete with compressive strength
;. equal to f' c .
't:
In addition to the weight and striking velocity of a primary fragment, its shape will also .atfect the resulti!.'g penetration. The sharpen the leading edge of a fragment; the greater the distance traveled through the concrete. The shape indicated in Figure .4-77 is not necessarily the' most critical. When the container,of'an explosive. shatters, it is statistically probable that some of the resulting .,fragment'! will have a 'sharper shape .chan the st~ndard bullet shape assumed· in, this manual. lJowever, the ..numbercof these fragments is usually very small in comparison to the tota! number formed, .and the.probability that these sharper fragments will have normal penetrations though the concrete is low. In most instances, the majority.of the primarY'f;~gments will have a more blunt shape than that shown. Therefore, for design purposes, the normal penetrations defined for a;bullet-shaped fra.gment can usually be assumed as critical.
4-60.3. Penetration of Fragments Other than Armor-Piercing
.'
.
d'
To estimate the concrete penetration of metal fragments other than armorpiercing, a procedure has been developed where the concrete 'penetrating capab Ll Lt Les of armor-pierCing fragments haye been r e l ated. to those of other metal fragments. This relationship is expressed in terms of reiative metal' hardness (the ability of the metal to resist deformation) and density,' and is represented, bythei.c~nstant, in Equat.Lcn 4-2<)3. ..' ._ 4-203
4-207
TK S-1300/NAVFAC P-397/AFR 88-22' where: X'f
'maximum penetration in concrete of metal'fragments other than armor-piercing fragments
k
constant depending on 'the casing' metal,' from Table 4-16:, r
-.(.."
,
maximum penetr~tton of armor-piercing fragment'
Xf
It should be noted that Xf is calculated from'Equation 4-200, 4-201 or Figure 4-78 i~ f'c -,4~000,psi, and f~om Equation 4-202 when Xf is modified for concrete strengths odler than 4,000 psi. ,",' ' . :
,-':
.
.... ..~
4-60.4. Perforation of Concrete ' .
_
. ' .
"
_
-
.
. •
.
l
•
Quite often the'magnitude of the' initial kinetic energy primary: fragments , ' ;" . , .of . . . . will be'large. enough to produce perforation of the concrete.' The depth of' penetration Xf of a fragment into massive concrete is less than into a wall of finite thickness'due to the high resisting stresses afforded by the massive concrete. Consequently, the concrete thickness required to prevent perforation is always greater than the depth of penetration Xf into massive concrete. The minimum thickness of concrete required to prevent"perforation can be expressed in terms of the equivalent depth of penetration into massive concrete and the fragment size according to the following relationship: ~
".
.
-;
4-204 where: minimum thickness of concrete to prevent'perforation by a ,given ,fragment,' 'depth'of penetration' corrected for concrete 'strength,and • , .. , "fragment mater'ial
...
'.-
"
Fragments which perforate a 'concrete element will have 'a residual/velocity which may endanger the receiver system. 'The magnitude of this velocity may be approximated ,from ,the expression which defines the velocity of the' fragment at ,!ny time as it penecraees the concrete. ., t:' .'
".
A
"
i
'"
, For cases where Xf' is less than 2d: . '
~
'[
,4-205
1 ',0
and for cases where' Xf, is greater, ~han 2d:'
.
,
-
c / Tpf)]
(T
where:
0.555 . 4-206 ."- , J'
thickness of the concrete, less than'or vr
,
"
equar't~ Tpf
residual velocity of fragment as it leaves the,element
4-208
.,
~
Plo'ts' of v r' /
V
5-l300/NAVFACP-397/AFR 88-22
s against: Tc / Tp f a,!c"rding to Equations 4-79 and 4-80, re~pectively.
4~205,and
4-206 are
pres~~ted in~Figures
4-60.5: ~s.palling due to F~agment Imp~~t
"
When a primary fragment traveling at a high velocity strikes the donor surface of a concrete barrier, large compression stresses are produced in the vicinity of the point of impact. These stresses form a compression wave which travels from the impact point,expanding spherically 'until it reaches the back face e l.ement, At this fr~e surface, the compression 'wave is refl'ecte'd, (reverseg in direction and changed from a compression wave to a tension wave). Whe~ the stresses in the resulting tension'wave exceed the tensile capacity of the concrete, spalling of the c,oncrete at:the receiver,surface occ~:r~. The spalling forms a crater on the recei.ver surface.' This crate~ does not usually penetrate beyond ,the reinforcement, at the receiver surface, The occur renee of spalling Is a function of the fragment pene t r at.Lon ; '1. e. , the fragment must penetrate a barrier element to ,such a depth that suffi' cien~ly large stress~s,are formed at the receiv~r surface to produce spalling. If the thickness of the, element is increased above the critical thickness at which spalling occurs, for a particular t:rag';'';nt, rhen t:he spaiiing is eliminated'. ':On 'the: other ,hand, concrete .spa Ll.Lng always occurs ,with j'ragment ' perforation. The minimum.'thickness of con<~rete barrier required, to p r event; spalling 'due to penetration of a given fragment can be expressed in terms of thefragme?t penetration , into massive concrete~nd the fragment diameter: Ts p -1.2l5Xf dO. l
+ 2.l2d
4-207
,
whe re ;
minimum concrete thickness to prevent spalling depth of
.pene~~ation"corrected.jforconcrete stren~th
fragment materi~l
and
'
The' secondary fragment velocities ass~ciated with spalling r~sulting from fragment impact are usually small (less than 5 fps). "However, when blast pressures are also involved, the magnitude of the resulting velocities can be quite large. The secondary concrete fragments are accelerated by the motion of the ~arrier resulting in,possible fragment velocities up to several hundred feet per second, "
Because of tJ:1e potentially large secondary ,fragment velocities associated with primary fragment impact" .full protection .Ls usually, required for personnel, valuable equipment, and sensitive explosives. This protection may be ac-, complished either by·providing sufficient concrete thickness to eliminate spalling or by othe~ m~chanical means used to minimize the effects of spalling resulting ,from blast pressures as described· in section 4-56. The required concrete thickness to eliminate spalling caused by primary fragment impact may be obtained from Equation 4-207. ,
4-61.
,
\
.
Fragment Impact on Composite Construction
4-61.1. General To evaluate the effect of primary fragments on composite (concrete-sand-concrete) barriers, the penetration of the fragment through both the concrete and sand must be considered. For damage to be sustained at 4-209
TN
5~1300/NAVFAC
P-397/AFR 88-22
the rear, of a composite barrier a fragment must first perforate both the donor concrete pane l and the sand , and then penetrate or perforate the receiver " panel. If the fragment penetrates only part of the way through the receiver panel, then spalling may'or may not occur, ,depending on the panel thickness. Obviously, fragment perforation' of the receiver panel' indicates perforation of the entire barrier. ' '
.
4-61.2. Penetration of Composite Barriers ,
To determine the degree of damage at the receiver' side, t~~ penetration of the fragment in each section of the barrier must be investigated in sequence. Starting with the striking velocity and the weight of the primary fragment, the donor concrete sectibn is first analyzed to determine whether or not ' perforation of that section occurs. If the'calculations indicate that the fragment will stop within this section, then no damage will be sustained by the remainder of the barrier. On the other hand, if perforation does occur in the forward (donor) section, then the fragment penetration through the sand must be investigated. • The amount of the penetration through the sand depends upon the magnitude of the residual velocity as the fragment leaves the rear of the donor panel. This residual velocity is determined from'Figures 4-79'and 4-80 utilizing ,the striking velocity, the thickness of the donor panel; and the theoretical maximum fragment penetration obtained from'Figure 4-78.
.
'
The maximum penetration through massive sand is obtained from Equation 4-208, using the residual velocity calculated above as the , striking velocity of the fragment on the surface of the sand. Xs - 1.188 Dd In (1 + 2.16 x 10- 3 v s 2 )
4-208
where: Xs
penetration of 'the fragment "into 'the sand
Substituting the caliber density D for a standard shape (Fig. 4-77), 4-208 becomes:
E~uation
4-209 A plot of ' Equation 4-209 for a range of fragment weights and striking properties is shown in Figure 4-81. If the penetration in the sand is found to be less than the thickness of the sand layer, no damage is sustained at the rear surface of the barrier. In case of perforation, the penetration of the frag~ ment into the rear section (receiver) of, the ,barrier is governed'by the residual velocity as the fragment leaves the 'sand. This residual velocity is . calculated in a manner similar to that used for computing the residual velocity for the donor panel, except that the 'fragment penetration and striking velocity are those associated· with movement of the fragment through the 'sand, that is: 4-210 where:
actual thickness of the sand layer
4-210
;., .
.....
TK5-l300/NAVFAC P-397/AFR 88-22 Figure 4-80 can be used to calculate the residual velocIty of a fragment perforating the sand layer. Similar to the fragment penetration through the donor panel and sand, penetration of the fragment through the receiver panel is a function of the magnitude of the fragment velocity as the fragment strikes the forward surface of the panel. This velocity is equal to the residual velocity as the fragment leaves the sand. Once the penetration in the receiver panel ill known, then the dam~ age sustained at the rear of the composite barrier can be defined in terms of either spalling or fragment perforation.
<:. , 'J
.. . <
r '.'
•
.j
.. .,,')
:: '.0'· DONOR SURFACE
· I ".
.. :. ~~:
•
r~XIS OF PENETRATION ~F PRIMARY FRAGMENT
FRONT CRATER
...;.: .. ~
..
~..
-··...." ': ..
.
" .~.
"0°.
•
.:." ~'.:
.... ~.
_,
:?: .... '.=:. . '. ,
'.
· . ','
Figure 4-76
REAR CRATER
.
RECEIVER SURFACE
','
.: t',
Perforation and spalling of concrete due to
primary fragments
4-212'
.
. ..
..
/" y
d = DIAMETER OF CYLINDRICAL PORTION OF FRAGMENT r = RADIUS OF HEMISPHERICAL PORTION OF FRAGMENT
r =
d
2"
r
r
d
D= CALIBER DENSITY. Wf Id 3 - 2.976 oz.lin.3
Wf = FRAGMENT WEIGHT N = NOSE SHAPE FACTOR = 0.72 + 0.25
.
In, -0.25
n = CALIBER RADIUS OF THE TANGENT OGIVE OF THE FRAGMENT NOSE = rid
..
Figure 4-77
Shape of- standard primary fragments
4-213
= 0.B45
NUMBERS NEXT TO CURVES INDICATE ,. , FRAGMENT WEIGHT, WI (oz.ll···~ '~
50 L
,
1
I
·P-'>'h"7.''>1
C~.o.r.~:.I.e..=.~,o.?o.~.si ..
10.0 c' . 8.0
L. 30i-··-:.-r···i·····:·······_ , I
6.0
!
4.0
,!. 20,'-: :.. _; _..:._. ,-.. :-.:.
2.0
LO :I: ~
0..
W
0.5
o
0.3
0.1
....
X
1 "1
, 2 r-
.:.··.1
;. ..-... ~ ..-_._._- _.-:- .._-_._-.----_! . .
•• J
..!
I ': j ..,.._....: i...__ _L. ....1...__.j 2000
4000
6000
8000
vs , STRIKING VELOCITY, Ips Figure 4-78
Concrete penetration chart for armor-piercing steel fragments
,. 4-214
1.2
1.0
0.8
...
~
.....
0.6
,
~
.,
0.4
0.2
..
,.
,'
o
o
',0.2
0.4
0.6
0,8
1.0
1.2
RESIDUAL VELOCITY,j STRIKING .VELOCITY I Vr IVs
.... Figure 4-79
Residual fragment velocity upon perforation of concrete barriers (for cases where x < 2d)
4-215
1.0
0.8
X
..... ~ Ct:.
0
0.6
....
1-0. <,
~
0.4
0.2
o
o
0.2
0.4
0.6
0.8
RESIDUAL VELOCITY I STRIKING VELOCITY.
Figure 4-80
1.0
v; IVs
Residual fragment velocity upon perforation of (1) Concrete barriers (for cases where x > 2d) (2) sand layers
4-216
1.2
e
e 10,00011
0.5 ':11' ITT , ' . II
i!
Ii i t !
'j"
, it'
: 1 '1 ; i :
, i
, I
,._ I I
-ti-r,'
-"';--'-4-
r:
•• I i
1.0
2.0
3.0
5.0
e 7.5
10
20
15
I'
25
30
I. I ! ,
t i,
I •
L
"i-:-"'~
-1_
i J 1 _I' i . iL11 !I j"I 1 I'
.i
I
8,000
I.I
I
I
I
I
i
! t·!"j
I·!l
Ul
...a. >-
h+:
~
.,., ,.,;
.... .....
6,000
-
-I' -i-,I" 1. t
1\ : II! . -i I ;
__:... ,
~+_+-;
'j' TT .
I
:
:
i -I-LI '1' '.
"!"!" "'. -'It ,- " l .-,,-, -r-: I, I
I
I
I,
,-
I-
._1 ..
(J
0
--l W
>
(.\I
Z
:><: II::
I1Il
- - _. -
20
30
40
50
MAXIMUM
"'.
-
". .'"
--
60 PENETRATION
- - .. ,
-
-
-
-",.
70 I
_.
80
90
(In.)
NOTE: NUMBERS NEXT TO CURVES INDICATE FRAGMENT WEIGHT, ( ca.) Figure 4-81
Depth of penetration into sand by standard primary fragments
100
TH 5-l300/NAVFAC P-397/AFR 88-22
Table 4-16
Relative.Penetrability-Coefficients for Various Missile Materials
. ..
Type of- Metal
Constant k
Armor-piercing steel
1.00
Mild steel
0.70
Lead
0.50.
Aluminum
0.15
-,
4-2~8
TM 5-l300/NAVFAC P~397/AFR 88-22 V.'
CONSTRUCTION DETAILS·AND,PROCEDURES, 4-62. Introduction A major portion of the detailing, and construction"procedures required for structures designed to resist blast pressures is the same'as required for structures designed conventionally. However, some differences do exist and " neglecting them would result in an unsafe situation, since,the structure 'would not act as assumed in the des Lgn, ,·These sections descrIbe -the differences in construction. Particular 'attention is directed towards: the construction . of, structures subjected to close,-in blast effects (elements reinforced with lacing or single leg stirrups) but the construction of conventionally reinforced elements, flat slabs and composite elements is "Iso discussed. Although the,construction of blast resistant structures is'similar- to conventional structures, some changes in the fabrication and construction procedures are required to insure full development of both the 'concrete and the reinforcement well ,into the range of plastic action of the'various elements. Since these changes primarily affect the reinforcement rather than the concrete, the major portion of the following discussion pertains to reinforcing steel , details. Typical details are presented to illustrate detailing procedures and design considerations. These details "may not be applicable to all design situations, and may-have'to'be modified by the engineers, within the guidelines given below.
,
4-63. Concrete
, .'
'
I
; ••
The dynamic characteristics and high magnitude-of' the applied blast load require the strength of the conc rat.a-uaed in blast resistant' construction to be higher than, that . required for -convent Ional, construction . ' -Because' of the flexural ' action of blast resistant elements while ' large deflections are required and the high pressures associated with blast loadings, it is recommended that a minimum concrete strength ,of 4,000'psi be·used . . ! .:
The properties and testing of the concrete materials (cement, aggregate, water) used in blast resistant concrete construction are the--same as those normally used and should conform' to the'standards specified'in the ACI Building Code . High early-strength Portland cement; (type IIll'may be used. To minimize the ,effects of spalling, 'it is recommended that the size of the' aggregate used be not ,greater than 1 inch.' This limitation of the aggregate size also facilitates the' placement of the concrete, particularly-where the cover over the reinforcement is held to a minimum. In all cases, the minimum conr.rete cover should;conform to that, specified in the ACI Building Code and, wh~rever possible,;should also be the maximum thickness,of tne concrete' cover. B""ause of the large amounts of',flexural reinforcement and, in laced elements, the presence'of lacing reinforcement, the'concrete slump used iS'usually larger than that permitted for conventional construction: A'concrete slump of 4 to 6 inches is recommended for laced elements to insure that concrete voids do not occur.
..
Wherever possible, both horizontal and vertical construction-joints' should be avoided. A wall'whose height is equal to or less' than ,10 feet usually can be poured without a horizontal joint. On the other hand, good construction 4-219
TII,5-l300/NAVFAC Pc397/AFR 88,-22 techniques and economy may require the use of horizontal joints for higher walls. These joints should be located at points of, low str~ss intensity. A more detailed discussion' of joint locations is given later in sections. 4-64. Flexural Reinforcement The flexural reinforcement, used ,in blast resistant'constructio~'shouldbe designateo as. ASTM A 615, Grade 60. Slabs must,be reinforced in two mutually perpendicular directions. 'In all elements, the'reinforcement should be cont~nuous
in. any direction.
AII.flexural
reinforc~ment
should consist of
straight bars, and bends in, the reinforcing within the' spanof'an element should, be, avoided. However, the reinforcement may be bent well within the eLemerrt ' s supports when additional anchorage' is required.) . .
~
,
The spacing of flexural bars is 'governed by the required area of th~ reinforcing steel, the selected bar size and, as discussed later, the spacing required to achieve a working and economical arrangement of lacing reinforcement and , single leg stirrups. In general, flexural bars should lie' 'spaced fairly close together to insure that the cracked concrete between the layers of reinforcement,will not be dislodged from the element. Tests have indicated that a maximum spacing of approximately 15 inches will insure confinement of the, concrete.
Because of their reduced ductility, reinforcing'steel larger than No. 14 bars (No. 18) should not be ,used as flexural,'reinforcement in blast resistant elements. Also, th~ 'size of the flexural steel should be at least equal to a' No. 4 bar. Where necessary, the area of steel normally furni~hed by the special large bars should be provided by bundling smaller bars. However, the use of these bundled bars should be limited to 'one direction only," for any laced element. If bundled bars are used i'\ an, element Whose',main 'span is between two opposite supports: then all bars of each bundle should be con- • t Lnuous -across the full, span. On che co ther ihand , if the main span is' between a support and a free edge, then bundled bars may be cut off at points of ' reduced stress. At least, one bar in each bundle should be continuous across the full span. For two-way elements these 'cutoffs must, be located beyond 'the positive, yield line with sufficient anchorage to dev~lop the,bars. ' Continuous reinforcement ,should be used, in, blast' resistant elements. b~t,in many cases, this, is .:.. physical impossibility. Therefore, splicing of the reinforcement is necessary. Splices should be,located in regions of reduced stress and their number held to a minimum by using tne longest reinforcing, " bars practical (bars 60 feet in length are generklly available throughout the country). Tests have indicated that the preferable method of splicing flexural reinforcement in laced.concrete elements is' by lapping the reinforc-, ing bars. The length of' each lap should be at least equal ,tp 40 diameters of the larger of the two bars spliced, but not less·than 2 feet ,(usually the same size bars will be spliced). In addition, splices of adjacent parallel bars should be staggered,toprevent the formation of a plane of weakness. Figur~ 4-82 illustrates typical splicing. patterns for -bo rh single, and bundled bars used in the close-in . . .design range. . ., " .... ' In conventionally reinforced' (non-laced) concrete elements used in the far. design range, .a two' bar splice pattern is used. The length. ~f the lap splices must be calculated to ensure full development of the reinforcement. If the" splices occur in regions of low stress,.the length of the lap 'is taken ,as 40 4-220
A .,
TH 5-l300/NAVFAC P-397/AFR 88-22
•
bar diameters. Lap splices of No. l4'bars are not permitted in either laced or non-laced concrete elements. Mechanical splices may also be used, but they must be capab Le of developing the ultimate .s t rengch of the reinforcement without reducing ,its ductility,. If the bar deformations have to be ,removed in the 'preparation of these splices, grinding rather than heat should be employed since heat can alter the chemical properties thereby changing 'the physical properties of the steel and possibly reducing the capacity of ,the element., Welding of 'the re~nforcement should, be prohibited unless it can be determined that the combination of weld and ' reinforcement steel will not result in' a ,reduction' of tne Mltimate strength and ductility. In'those cases where 'welding is absolutely essential, it may be necessary to obtain ,special reinforcement manufactured with controlled chemical properties. '
4-65. Construction Details' for "Far'Range" Design 4-65.1. General Unlaced reinforced concrete elements arergene rel l.y used in those facilities designed'to resist explosive output at the far design range. The facilities generally, consist of shelter-type structures:', Blast resistant structures utilizing unlaced 'reinforced concrete construction will only differ from conventional construction insofar as the increased magnitude and reversibility 'of applied loads 'are concerned.'·These differences are reflected in the details of ,the blast' resistant "s t r'uc trirr e-, An increase in the amount of reinforcement is'required to resist the'iarge dynamic loads: Also, the reinforcement at both 'surfaces of ,an element must be detailed to' prevent failure due to ~rebound stresses. .'
"
In general, conventionally reinforced elements are limited to deflections corresponding to support rotations of 2 degrees or less., However, elements designed for far range effects are capable of attainingdeflections'up,to 4 degrees through the use of single leg'stirrups and up 'to 8 degrees by developing tension membrane ,action. Elements 'subject to close,in effects'require stirrups and are limited to 2 degrees support rotations. Details ar'e presented below for elements designed for -limited deflections', elements acting under tension membrane action and columns., For elements 'subjected to low and intermediate pressure and reinforced with stirrups, the details given below may be used without modification. However, details for elements with stirrups and subject' to high pres'sur'e s are dt scussedtn the following, sec t Lon..
"
~
:
• <..
. . "
4-65.2. Elements Designed for Limited D~ftect1ons • Construction and detailing of'unlaced, blast resistant structures is very similar to conventional structures. The major difference iS,the method of anchoring the reinforcement. , A typical section through a n~n-laced wall is shown in Figure 4-83. At the roof-wall intersection, the exterior wall reinforcement is continued through the 'regions of high s~ress 'and lap spliced' with the roof reinforcement in the vicinlty
4,221
TH 5-1300/NAVFAC
P-397/AFR'~8-22
therefore, the straight portion of the bar must be sufficient to develop the reinforcement in compression .• The bottom floor slab reinforcement is extended through the floor-wall intersection into the wall in the,same manner as the .roof-wall intersection. Again the reinforcement is developed, into the wall .. The vertical wall reinforcement is supported on the floor slab rather than supported above the floor slab on the·reinforcement. Figure 4~83 illustrates a building ,with a 'slab-on-grade foundation. Figure. 4-84 illustrates several alternative
,
.
arrangements.
A horizontal section.through the intersection through two discontinuous. walls (Fig. 4.-85) would reveal details very similar to those shown for the roof,wall intersection. The number of splices used would. depend on the length of,each wall. • Wherever possible, continuous reinforcement should be used. Lap, slices' ma~ be used when necessary but their number should be held to a minimum and they must be located in regions of reduced stress. To prevent the formation of a plane of weakness, splices on adjacent parallel bars must be staggered. In addition, the splice pattern for the. reinforcement on -,opposite faces of an element should not be in the same location. Figures 4,86 and 4-87 illustrate preferred splice locations for a two-way slab' and ,a multi,span slab respectively. 4-65.3. Elements Designed for Large Deflections
..
'
Unlaced concre ce.vs Labs are capable of attaining deflections up to 8 degrees support rotation through the development of tension.membrane action. Construction details are basically the same whether the .slab attains large or small . deflections, however, there are two important exceptions. At large deflec-' tions, tension membrane action produces large tensile strains in both tension and "compression" reinforcement. Therefore, all anchorage lengths and lap splice lengths must be calculated for the design stress fds and all the reinforcement is in tension. The second difference,. in detailing concerns the location of splices. Although 'splices should still be located in regions of low flexural stress, they will· be, located in regions of high tensile stress when the element attains its full tension,membrane capacity. The length of, the lap splice must, therefore, be increased to 1. 3 times the modified . development ·length given in Section 4-21.4 of. a reinforcing bar Ln-j.ens Lon,
,
-
4-65.4. Column Details •
.. 0;'1
Columns are, generally r'equ l red in blast r e's Lstant , .shear wall structures. r, Their details differ little from those used in conventional structures. Both round, spiral reinforced c?lumns.and, ~ect~ngular, tied columns can be used, but one or the other is preferred for a given design situation. Details of the reinforcement and formwork of rectangular cqlumns are compatible with beam details and, therefore, are reco~ended fqr beam-slab. roof systems. Round· columns should be ,used for flat sla~ roofs. .
."
.\
Figure 4-88 illustrates a typical se~tion through a circular column with.a " capita\. supporting a flat slab. A round' column prevents stress concentrations that may cause. local, failures. ,The column capital .. altho,!gh not required" is preferred:since it,simplifies the. placement ,of the diagonal bars as well as I decreasing shear stresses. All diagonal bar svshouLd extend into the column. 4-222 .
TK 5-1300/NAVFAC P-397/AFR 88-22 , " If, however,. the number and/or size of the diagonal bars do not permit all the bars to extend into the column, up to half the bars may be cut off in the capital as shown. Lateral reinforcement of a circular column consists of spiral reinforcement beginning at the top of the floor slab and extending to the underside of the drop panel or the underside of rhe roof. slab if no drop panel is used. Within the column capital, No.4 hoop reinforcement at 6 inches on centeromust enclose the diagonal.bars. In a.beam-slab system, beam reinforcement does not permit the addition of diagonal bars at the top of the column. Therefore, th" beams themselves must be able to provide adequate shear strength. Closed t i e s provide lateral c, support of the longitudinal reinforcement in rectangular columns. The ties must.,start not more than '5 tie spacing above the top of the floor slab and' end not more than 3 inches below the lowest reinforcement in the shallowest beam. The ties must.be arranged so that, every corner and alternate longitudinal bar has the lateral support provided.by the corner. of· a· til> having an included angle 'no more than.135 d~grees. In addition, no 'longitudinal bar shall be farther than 6 inches clear on either side from such a laterally supported· bar. The column footing illustrated. in' Figure 4-88 is the same fo, both rectangular and circular columns. Dowels anchor the column into its footing; .Since there is little or no moment- at the bottom of the column, splices need not -be
staggered. as, they are in a 'wal'L . The splice should 'be able to fully develop the reinforcement in" tension in order to resist~rebound tension forces. Within.the column, no splices of the.longitudinal reinforcement are permitted, Continuous reinforcement should not be,a problem as blast resistant structures
are' limited to one and two-story buildings. ,. 4-66. Construction Details for "Close~In Design" 4-66. L General· Laced reinforced concrete elements are usually used in those facilities which are designed to resist the explosive output of close-in detonations (highintensity pressures.with short duratlonsj. The functional requirements of these facilities (storage and/or manufacture of explosives). normally dictate the use of one-story concrete buildings wi'th austere architecture. Basically, these structures consist of'a series of inte~connecting structural elements (walls, floor slabs, and/or roofs) forming several 'compartments or cubicles. Because of this cubicle arrangement, the walls separating the individual areas are the' predominant 'element used in .Laced construction and are the most' critical component in the design. However, in soinecases, the roof and/or floor slab can be of equal importance . '.
Single leg stirrups may replace lacing bars when the scaled normal distance between' the charge and the element is equal to or'greater .than 1.0 ft/lb s l I.3. However, the maximum allowable support rotation for el.ements with slngle leg stirrups is less than for laced elements. Single leg stirrups are more economical to 'fabrlcate and slightly more economical to place during construction: Unlike laced reinforcement which requires that the position of the flexural reinforcement be changed to suit the horizontal and vertical laclng, the position' of the flexural reinforcement remai'ns constant for single leg: stirrups. The stirrups are. tied around the 'outer' bars whether"they 'are
4-223
TN 5-1300jNAVFAC " . . P-397/AFR 88-22 . horizontal, or ",v~r'ti~aL ,'Details for' flexural reinforcement, splice location and length;etc., are 'the same for single leg stirrups and laced reinforcement. 4-66.2. Laced Ele~ent8. 4-66.2.1. Lacing Reinforcement All flexural reinforcing bars must be'tied with continuous diagonal lacing bars (Fig. 4-89). At any particular section of an element, the longitudinal or main tension and compression reinforcement is placed to the interior of the transverse or secondary flexural reinforcing steel around which the diagonal lac ing bars are. bent. ' ., . Cacing reinforcement must be fabricated without the formation of excessive stress concentrations at the bends. The bending should be performed without the use of heat, and in no case should the','radius of the pin used to bend the lacing be less than four times·the diameter of the bar. Figure 4-90 il-' lustrates, the typical lacing bends used with both single and bundled flexural bars. The amount of lacing reinforcement required iri an element depends upon the element's capacity (quantity and distribution of flexural reinforcement, . thickness of the element, and the type' and number of supports). while the' size of the lacing bars is'a function of the required area and spacing. The . maximum and minimum size of the"lacingbars should be No.8 and No.3 bars, respectively. However, the.preferred maximum size of lacing'bar is No .. 6. Several lacing schemes have been developed (Fig. 4-91) which avoid excessive stress con~entrations .as .a result of large angular bends, provide adequate restraint against buckling of the compression reinforcement, and·make use· of the most efficient arrangement'of,the;laclng reinforcement (lacing making an angle of 45. degrees with the longitudinal reinforcement is most efficient). In these scheme s, the transverse flexural.reinforcement mayor may not be tied at every intersection with a longitudinal bar. However, a grid system is established.~hereby,bar intersections are tied within a distance sl or 2 feet, whichever is less. The choice of the scheme to be used depends upon the flexural bar spacing and the thickness of the element.so that the angle a is approximately equal to, but never les~ than, 45 degrees. Although not tying every transverse flexural bar results in a large lacing bar ';'ize, the total cost of the lacing may'be reduced since the·size of the lacing bend associated with an increase :in the spacing sl reduces ~he.overall length of the bar and the number of bends required to cover a given longitudinal dd st.ance .. Anadditiorial cost' saving in the fabrication of the lacing may be realized by utilizing the equipment that is used to bend the steel bars for open web steel joists. However, when detailing the flexural and lacing reinforcement and,the concrete wall thickhess,' consideration must be given to the physical capabilities (size of bend ~nd bar, depth of lacing, etc.) of the equipment. Alteration of'vbar' joist' equipment to meet the requirements of a d~sign is not usually practical with; respect to both. time and cost. The placement of the lacing depends upon the distribution of the flexural reinforcement and the number and type of supports. Lacing is always placed 'perpendicular to the element's suppprts to resist diagonal tension stresses. Rec~use .of the'nonuniformity,of the blast loads associated with close,in detonations, continuous lacing across the span. length should be used to 4-224
TK 5-l300/NAVFAC P-397/AFR 88-22 distribute.the loads. Exc~pt for ,ca~tileveF' elements, lacing in one-way elements is placed in, the direction 9,f a~d s~ntinuous across the span, Canti~ever ,elements require lacing in two directions., Discontinuous, ,lacing is located perpendicular and adjacent ,to the support ,while continuous lacing is placed across the ,full width of the ' element in direction parallel a~d adjacent to t~e free edge located opposite the element's 'support. For two-way elements, diagonal tension stresses must be resisted in two. directions. Because of interference, lacing can orily be continuous in one direction, which in general is in the direction of the longest span. Figure 4-92 illustrates the location of the lacing used in a cant~lever wall as well as in several tw~-way elements. For two-way elements ,.th.. .LocatiLori of .the .LacLng is 'not affected by, the tyPe of supports., Therefore, the suppor~s indicated, in Figure 4-92 can be simple, restrained; or fully fixed. Similar to the flexural reinforcement, lacing will usually require.spIJces. Tests have indicated'that the'preferable'way of splicing lacing is by lapping the.bars., The lap length which is measured along the bars should be at least equal to that required for a full tension. slice (40~bar diameters). ,However, the lacing should also be bent around. a minimum of t.hree flexural bars. :' The splices of, adjacent lacing bars should b~ staggered,to avoid forming, a' plane of weakness 'in the element,. and the slices should be iocated in regions of l~w stress (away'from the supports and,positive yield lines). Typical details for the splicing 'of lacing' bars" are' presented in' Figure .4- 93. ,Wherever -po s s LbLe-, welding of lacing bars should be avoided and is only permitted while the full development of, the ultimate' strength can be assured without 'any reduction in strength, or ductility. - '.' " '0
The location of the splices is determined from the distance along,the length of the element which can be covered by a given length of lacing bar,(usually 60 feet) '. The express Ion for the actual length o~' lacing bar LI' .r equdr e d to cover the length s1. is a func t.Lon of, the' flexural bar, spacing, and the geometry of the lacing and is given in' Figur~ 4-94.
-.
4-66.2.2. Corner Details' ,
.
....
. ,
I
.
Because of the magnitude of the blast loads associated'with close-in detonations'and their amplification at c6rners, the 'use of concrete haunches has been found' to', be a satisfactory method, of maintaining the integrity, of, these sectfons of a structure. All corner s should be feinforced. with. dd agona L'b ars to transfer the high shear f~rces from the' element to its support ,and to as- . sist in maintaining the integrity of the intersection. Diagonal 'bars should be used in eiements both with, and without haunches. 'Reinforcement, details for corners are shown' in conjunction with wall intersectfons,"in the fo'llowing sect ten,
' "
,
4-66,2.3,' Walls Figure 4-95 illustrates the detailing procedure for a typical laced wall, The wall is free at the top, supported at its sides by other walls (not shown), and at its base by the floor slab. It has vertical lacing which is continuous from the bottom. of the wall .to approximately·midheight . , In the upper 'half 'of the'wali, the horizontal lacing reinforcement is' continuous over the.~ull length of the wall and anchored 'in the side suppor t s, , It should be noted that the horizontal flexural steel' in the lower half of the wall is .located at·, the' . exterior of the vertical reinforcement, while in the upper half of the wall it 4-225
TM'5-l300/NAVFAC P-397/AFR 88-22 is at ,the interior, This arrangement is necessary for the placement of the lacing so that the lacing can provide 'full confinement of both the flexural steel and the concrete separating th~ two layers of reinforcement, If' spalling is critical', its 'effects may be"minimized by-reducing ,the concrete cover over the'reinforcement in the upper'haleof the wall. The addition of U-bars at the top of the free'edge also minimiies'the formation of concrete fragments from this area. Reducing the concrete cover is not cost effective and, therefore, if spalling is not critical; the wall 'should have a constant thickness. "
A portion of the wall should extend below the ,floor' slab a'distance no less than that required to anchor the flexural reinforcement (40,bar diameters) and in no case less than I' foot; - 6 inches'. 'In addition ,'to providing"anchorage for the vertical reinforcement, the portion of the wail,below the floor slab assists in resisting sliding of the structure by developing the passive pressure of the soil' adjacent to the wall. ' The working pad at the base of the wall provides the support'required for the erection of the wall reinforcement'and also affords protection for the reinforcement after 'construction is completed. ,It should be noted that in the example illustrated in Figure 4-95 ,the cover over the reinforcement in the portion of the wall above the floor slab is specified as 3/4 inch (minimum reinforcement cover'required by the ACI'Building Code for concrete not exposed to the weather). while below the floor slab it is specified as 2 inches (minimum, cover required by ACI BUilding Code for concrete in contact with the ground after- removal of the forms). The, increased· cover below the' floor slab is achieved by increasing the wall thickness rather than bending ,the reinforcement. l'
Diagonal bars are provided at the intersection of the floor slab and,wall. These bars transfer the high shear forces from the base of the wall,through the haunch and into the floor slab. Section,C-C, Figure 4-95 indicates the location of the diagonal bars relative to the lacing reinforcement. ' Details of the reinforcement at wall intersections are similar to those at· the intersection of the wall and floor s Lab. Figures 4-96 through 4-98 illustrate these details for .the intersection of two continuous walls,' one continuous and one discontinuous wall, and two discontinuous walls, respectively. Lnva.Ll, cases, both the'flexural and lacing reinforcement'are fully anchored by being made continuous through the intersection, (a distance of at least 40 bar' di~eters), In discontinuous walls, the wall must be extended a distance sufficient to anchor all reinforcement but not less· than 1 foot-6 inches. This extension (or extensions) aids in resisting both the overturning of the structure and.the tension force, produced in the walls (discussed further in subsequent sections). Diagonal bars have been provided at all intersections to transfer the support shears and to maintain the integrity of the section, Where discontinuous walls are encountered, only one diagonal bar is. effective, for the continuous wall of Figure 4-97 ,and for both discontinuous walls of Figure 4-98, For these cases, bundled diagonal' bars (2, bars maximum) may have to be used. If wall ,extensions are not permitted due to architectural or other criteria, the reinforcement is anchored by bending it within the corner (figs, 4-99 and 4-100).· However, the distance from the face of the support to the end of the hook must be at least 20 bar diameters. A standard hook may be used if;. (1)
4-226
•
TK'5-l300/NAVFAC
P~397/AFR
88-22
the total distance from the face of the support to the end of the bar (including the hook) is 40 bar diameters and (2) the length of the hook is at, least ,12 inches. The use of hooks may cause problems in the placement of the lacing reinforcement and is discussed in conjunction, with the sequence of construction below. ', In addition to flexural reinforcement, the walls of containment structures may require the addition of tension reinforcement. This reinforcement is'placed at the mid-depth of the wall and has the same spacing as the flexural reinforcement. It may be required in the vertical 'and/or horizontal directions. The reinforcement'is anchored at wall intersections in the same'manner as the flexural reinforcement. Vertical tension reinforcement usually does not require. splices: However, if the horizontal ,tension ndnforcement requires splicing; the splice, (lap, mechanical, weld) and pattern should be' same as the flexural reinforcement. Of course; the tension steel should never be" spliced at the same location. as the flexural reinforcement.' ,
4-66.2.4. Floors Slabs
•
Floor slabs on'grade'must provide sufficient capacity to fully develop the' wall reinforcement. " If sufficient resistance is provided by the soil" slabs' poured on grade usually do, not require lacing' nor other shear reinforc'ement; although lacing r'e Lnf'o r ceraent; is always :required in slabs exposed to multiple' detonations. Soil strata having enough bearing capacity to support the'dead load of the structure ,can be considered to provide the support required by the slab. Before placement of the slab, the top 6 inches of the subgrade should' be compacted to 95 percent of maximum density in accordance with ASTM Standard 01557. Piles are used to support a structure where the,bearing capacity of the soil is inadequate. ,The piles are placed under the walls and the' floor must span between thein. Lacing or single leg'stirrups'must be provided. The reactions of the slab are transferred to the portion of the wall below the floor 'slab which··acts 'as a beam' spanning between piles. In addition to flexural reinforcement, floor slabs may require,tension reinforcement located at mid-depth of the slab. Tension forces are'discussed in conjunction with single and multi-cubicle structures.
4-66.2'.5. Roof Slabs Roof slabs are similar to walls since they' are usua Ll.y supported only at their periphery and require the addition of'lacing to distrIbute and'resist'the app Ld'e d blast loads. In those facilities"where the explosion 'occurs within a structure, the blast pressures acting on the'interior surface of a nonfrangible roof causes tension stresses in the walls which require the addition of tension reinforcement above that needed for bending. Tension forces are discussed in conjunction with single and multi-cubicle structures.
4-66.3. Elements Reinforced with Single Leg Stirrups :
4-66.3,1. Single Leg' Stirrups "
,
.
A single ,leg stirrup consists of a straight bar with a hook of at least' 135 degrees at each end. Hooks shall conform to the ACI Building Code. At any
4-227
TK 5~1300/NAVFAC P-397/AFR 88-22 particular section of an element, the longitudinal flexural reinforcement is placed .to the interior of the transverse reinforcement and the stirrups are bent around the transverse reinforcement (Fig. 4-101). The required quantity of single leg stirrups is calculated·in the same,manner as lacing. They are a function of the element's flexural capacity while size rebar used is a function of the required area and spacing of the stirrups. The maximum and minimum size of stirrup bars are··.No. 8 and No.3, respectively, while the spacing between stirrups is limited to a·maximum of d/2 or d c/2 for type I and type II or III, ·respectively. The preferable placement of Single-leg stirrups is at every flexural. bar intersection. However, the transverse flexural reinforcement does. not have to be tied'at every intersection with a longitudinal bar. A grid system may be established whereby alternate bar intersections in one or both directions are tied within a distance not greater than 2 feet . . The choice of the three possible schemes depends upon the quantity of flexural reinforcement, the spacing of the. flexural bars and the thickness of the concrete element. For thick, lightly reinforced elements, stirrups may be furnished at alternate bar intersections, whereas for thin and/or heavily reinforced elements, stirrups will be required at every bar intersection. For those cites where large stirrups are required.at every flexural bar intersection, the bar size used may be reduced by furnishing two stirrups at. each flexural bar intersection. In .this situation, a stirrup is .provided at each side of longitudinal bar. .
.;
Single ~eg stirrups must be distributed through~ut an element. Unlike shear reinforcement in conventionally loaded elements, the stirrups cannot be reduced in regions. of low shear stress. The size of the stirrups is determined for the high stress areas and, because of the non-uniformity of the blast loads associated with close-in detonations, this size stirrup is placed across the span length to distribute the loads. For two-way ,elements , diagonal tension stresses must be resisted in two directions . . The. size.of stirrup~determined for each direction is placed to the same extent as - the lacing shown in Figure 4-92. However, t~e distribution does not apply for cantilever elements since they are one-way elements requiring only one stirrup size which is uniformly distributed throughout. 4-66.3.2. Corner Details Corner details for elements with single leg stirrups are the same as for laced elements. Concrete haunches, reinforced with diagonal bars, should be used at all corners. For those cases where compelling operational requirements prohibit the:.use of haunches , diagonal bars must still .be placed at the.se· " corners.' In addition to diagonal bars. closed ties must be placed at all corners' (fig. 4-102) to assist in maintaining the integrity of the intersec, tion. The tie should be the same size as the stirrups but not les~ than a ~o. 4 bar. The spacing of the ties should be the same as the flexural reinforcement. 4-66.3.3. Walls The detailing procedure for a wall ,with single leg stirrups is similar toa laced wall. Figure 4-103 illustrates the detailing procedure for a typical wall with single leg stirrups. This wall is the same as the wall shown in Figure 4-95 except that stirrups fire used rather, than lacing. These are only. 4-228
:
TN 5-l300fNAVFAC P-397/AFR 88-22 two differences. between ,the two walls. First, there is no need to .alter· the position of the horizontal flexural reinforcement,for the placement of stirrups. The horizontal reinforcement is in the main, direction (assumed for wall illustrated) and, therefore, this steel is placed exterior of the vertical reinforcement for the entire height of the wall. Second, closed ties are 'placed at the wall and floor slab intersection to a~~ist i~ maintaining the integrity of the section. The common requirements for both walls include the addition of U-bars, diagonal bars, concrete'haunches, increased cover over the reinforcement below the floor slab by increasing the wall thickness, shear reinforcement (stirrups or lacing) in the wall extension below the floor slab, and the preferred use of a working pad. Details of the reinforcement at wall intersections are similar to those at the intersection of the wall and floor slab. The requirements' for anchorage,of the flexural reinforcement and diagonal bars at wall inters~ctions are exactly the same as laced walls (fig., 4-96 through 4-100). The placement of the.. single leg stirrups and the required closed ties are shown ,in Figure 4-102, Similar to laced. walls , the use of wall extensions is the preferred method of reinforcement anchorage at discontinuous walls. 4-66.3.4. Floor Slabs The floor slab must provide sufficient capacity to Eu'lLy-vdeve Lop the wall reinforcement. The requirements are the, same as a floor slab for laced walls. 4-66.3.5. Roof Slabs ' Roof slabs are similar to walls since they are usually supported only at their periphery and require the ' addition of single leg stirrups ~o distribute and resist, the applied blast loads. For interior explosions, the roof causes tension forces in the walls; Tension reinforcement i" discussed in conjunctiop with. single and multi-cubicle structures. 4-67. Composite Construction Composite construction is .primarily used for barricades'and consists of two concrete panels separated'by sand fill (Fig. 4-104). Details of each panel are similar to those described for single laced walls (or walls reinforced with single leg stirrups). The 'concrete panels may be supported:' at the base e Lther by the floor -s Labs or a concrete pedestal. When the pedestal is used, reinforcement across the base of bo~h panels terminates in the floor slab and provides a monolithic connection between the two panels. The floor reinforcement serves as the monolithic connection when pedestals are not used. The upper portion of the wall may either be open or solid.' .Open sections are usually used when the upper edge of the wall is, unsupported; the solid section is used when.anexternal tie system is used to restrain. the motion and provide support for the top of the wall. The solid section m\lst be reinforced to resist torsion and bending induced by the ties 'and rhe panels.' The impulse capacity of composite walls is a function of the density of' the sand fill. The sand will be compacted after construction due to its own weight and/or by water drainage when the wall is exposed to the weather. The 4-229
\ TN 5-l300/NAVFAC P-397/AFR 88-22 sand fill must be continuously maintained at the level stipulated in the design by mechanical means which will "allow periodic' rearrangement of the sand fill. Clay pipe or· other similar material may be placed in the wall cavity with the sand so that when the wall is loaded the clay pipe will be crushed by the impact of the donor panel, thereby providing space within the wall cavity into which the compacted sand"may fiow, hence reducing its density. If possible; the sand should be protected from the elements by sealing the top of the cavity. 4-68. Single and Hulticubicle Structures I
In single-cell structures (Fig. 4-105) unbalanced force (support reactions)" exist at all element intersections (walls, and floor and wall intersections) and must be resisted by tension force produced-in the support elements~' In addition to·the r~inforcement"requiredto resist flexural and shear stresses, tension reinforcement, distributed along the centerline of ' the elements, is required. Horizontal tension reinforcement in the side wall and floor slab' (parallel to the side walls) is required to resist the vertical and horizontal rea~tions of the back wall, while horizontal steel in the back "wall and floor' slab (parallel to the back wall) resists the tension force produced by the side wall reactions. These unbalanced forces are transmitted to the structure's foundation and,
\
depending upon their magnitude, the size and configuration of the structure and the subgrade conditions, the structure may be subject to both"translational and vertical rotational motions. Translation of the structure is resisted byrthe ext.ens Lon of the walls below the floor slab (shear key) and the friction developed between the floor slab and subgrade , whereas rotation Is resisted by" the mass of the structure with ass Ls t ance from the blast load acting on the floor slab of the donor cell. 'The stability of the structure can be:substantially increased by the extension of the walls and floor slab as illustrated" in Figure 4-105b. "" This "extension of the walls and floor slab' (1) increases the resistance of the structure to overturning (rotation), (2) increases the rigidity of the structure, (3) reduces the effects of the unbalanced wall moments which cause twisting of the corners, (4) reduces the required thickness and/or reinforcement in the floor slab (moment capacity of the' floor slab extension must· be developed" by bearing on the subgrade). and (5) eliminates the need to anchor the'reinforcement by bending at the corners which would ordinarily hinder" the placement of the concrete. End cells of multicubicle arrangements also require the addition of tension reinforcement to resist unbalanced blast roads acting on the end walls., The interior cells do not require this additional reinforcement since the mass and basecfriction of adjoining, cells provide the restraint to resist the lateral forces. Two possible multicubicle" arrangements are shown in Figure.4-106. In both arrangements the back walls of the cells 'are continuous, whereas the side walls between the adjoining cells are either continuous or discontinuous. The type of cell arrangement (either one of those shown in Figure 4-106 or a combination of both arrangements) used in"a particular design depends primar-" 11y. upon the functional. requirements of the facility and the economy involved. However, there are certain structural features which should be considered in the final selection of either structural arrangement. The horizontal reinforcement (flexural and lacing) in the side walls may be placed continuous across the back wall of scheme" a, whereas with scheme b, the side wall horizontal reinforcement must be bent and anchored in the back wall. ,This latter "4-230
.TK S-1300/NAVFAC P-397/AFR 88-22 arrangement ·can·result· in'congestion of the horizontal reinforcement at the wall intersections., On the other hand, by offsetting the side walls at each side of the back wall,. the span.length of the back.wa'l l, between adjacent side walis is reduced thereby reducing the required strength (concrete thickness and/or reinforcement) 'of the back wall. In general, continuous'walls usually require constant concrete thickness and horizontal reinforcement. However, where economy can be achieved, it may be desiraDle to reduce the thickness and reinforcement of the continuous wall of one cell in comparison to those of adjoining cells. This reduction should only be made between the supports in order that a constant moment ,capacity can be maintained across the length .of the reduced element'. This capacity reduction requires the'horizontal reinforcement to be sll.ced at the supports, and extreme caution should be exercised in the detailing of the splices. ,' 4-69. Sequence of Construction Although the construction procedure for all blast r'es Ls't.ant; .ccncre ce elements is similar; each structure must be evaluated to determine. the specific sequence of construction which is most appropriate for the particular situation. This evaluation should consider: (1) type and location of shear reinforcement (single leg stirrups, horizontal and vertical lacing). (2) location of reinforcement splices, (3) erection sequence of the reinforcement (flexural and.shear), (4) location of'horizontal construction joints' and (5) pouring sequence of concrete. To illustrate the construction of a laced concrete' structure, consider the recommended construction procedure for the cubicle 'structureshown in Figure 4-105 b. A vertical section through any wall is similar to the wall described in Figure 4-95. Figure 4-107 illustrates the pouring sequence for the following procedure. 1.
Fabricate the reinforcement as indicated on the drawings.
2.
Pour a working pad.
3.
Erect vertical flexural reinforcement, vertical lacing and vertical diagonal bars in all walls. Thread horizontal flexural bars between vertical lacing ,and vertical flexural bars up to the top of the floor slab. Also place reinforcement for the floor slab. "
4.
Adjust reinforcement to··required -pos LtLons and complete, second pour to the top of the floor slab. As an alt~rnative, place sufficient horizontal lacing (as described' in step 7) to insure ~,': proper positioning of the vertical flexural.reinforcement and then complete second pour .. 'Additional horizontal flexural bars may be placed beyond' the limit.of the pour to help stabilize .the reinfor-' . cement.
5.
Thread horizontal flexural bars between the v~rtical lacing and vertical flexural bars beyond the limit.of the third pour. Adjust reinforcement and complete the .third pour . .
6.
Thread the remainder of the horizontal flexural bars up.to the top of the vertical lacing.
4-231
TN 5-l300/HAVFAC
P-3~7/APR·88-22
7.
Place horizontal. flexural and lacing reinforcement. and diagonal , ' bara between the.. top of the. verticai lacing, and, the top, of the' wall. Placement of the horizontal flexural and horizontal lacing reinforcement is accomplished by lowering this reinforcement over the vertical reinforcement. At wall intersections the proper sequence is to first lower diagonal bars, type a. Then in the north-south walls lower horizontal lacing bar 'type b,. place horizontal flexural reinforcement and 19wer opposing lacing bar type .a , Repeat this.,sequence with the reinfo'rcement in the east.west walls and .complete this individual layer of reinforcement by placing'the horizontal-diagonal bars type b. The entire sequence, is repeated for the remaining reinforcement.
8.
Add U bars at the top of the wall and adjust reinforcement to required positions. Pour remainder of the wall.
1
-
The above procedure is for the cubicle structure of Figure 4-105 b, where wall extensions are provided at the corners. For the case where 'wall 'extensions' are not used (Fig. 4-105 ali-the horizontal reinforcement must utilize a 90 degree-hook for 'anchorage (Fig. 4-10P). The horizontal flexural reinforcement for the side walls requires a 90 degree hook at one end.- Therefore, the reinforcement must ,be threaded between the vertical lacing and vertical flexural bars' from behind the back wall. -, If the back wall was close to an existing structure, the horizontal reinforcement could not be threaded. The horizontal flexural reinforcement in the back wall requires a 90 degree hook at each end which would prohibit threading the bars. To place this steel, the bars would have to be spliced so that they could be .threaded through the back wall from each side wall.. The use of splices is not desirable and should be avoided, making the use of wall extensions _preferable. ' The construction procedure for an element reinforced with single leg_stirrups is similar, but not quite as complex as laced elements. The single leg stirrups should be lowered into position if the vertical, flexural reinforcing bars are exterior of the horizontal bars. However, if the horizontal bars are exterior of the vertical bars (Fig. 4-103). the horizo~tal bars should be threaded between the stirrups and vertical bars. Again, as for laced elements, the reinforcement of intersecting walls and ·the diagonal bars must be placed in sequence. The use of construction joints. (both vertical and horizontal) should be avoided wherever possible since all joints are a potential plane of weakness. However; joints in large structures· cannot be.avoided because goo~ practice for placement of concrete and/or_economy 'requires their ·use. All'joints should be. located in regions of .low stress intensity, and, if possible, for laced elements,' vertical joints should be'situated in areas where horizontal lacing is located, and·horizontal joints should be situated in areas where vertical 'lacing is ,located. However, vertical joints are difficult to form in laced construction. In most cases, vertical joints are not used, 'and a certain height of all walls is poured simultaneously. In addition, concrete surfaces should be roughened at all joints. ,"
'.'
'.
The above construction procedure required the use of two horizontal construction joints.- The joint located at the floor slab is generally used in all blast resistant structures while the second joint in the upper section of the wall should-only be used if the height of the wall warrants it.- The use of 4-232
TH 5-1300/NAVFAC P-397/AFR 88-22 vertical construction joints is generally required for multicubicle arrangements. Walls (intersecting walls must be poured simultaneously) and corresponding floor slabs should be poured in checkerboard fashion to guard against joint separation due to shrinkage and temperature variations. To maintain a minimum rate of pour, multiple pouring crews may have to be used, and pumping of concrete; rather than the use of tremies, may be required for high walls. joints are generally not required for laced concrete elements due to the presence of relatively large amounts 'of reinforcement. However, their use should be considered for long buildings and/or structures subjected to extreme temperature c h a n g e s . ' .,
Expans~on
"
.. ,
.'
.
.
..
'
4-233
, ;'
,.
,.
CD I
Ii Of SPLICE PATTERN
~ !
=+=-+-1
-1::;[
w '" ~
cp cp
It
OF SPUCE PATTERN
i'~
===IW ===1 +t1
~
81
L~· .. · =r '=f=1=- B- -~- -' ' ' ;
BUNDLED BARS
e
Tc
i
,J
·1 I
~~
.
A..J
SINGLE BARS
,
~
I
I' I
Tc
J
,!,
SECTION B-B
.
,
e·;
I
It .oF
SPLICE PATTERN
, 4011 MIN. SEE NOTE ,
I
/.
I 40~: I
~
•
CONSTRUCT;ON JOINT
.
::E I
I
I
I
I
I
I
I
I
....
.,
~
r
,
/ ~
(
..
I
-
:
CONSTRUCTION JOINT
' ~
AS REQUIRED FOR , FULL ANCHORAGE (
I
,
~.
... ~
,
li W TD. 90:1 HOOK
..
i
,
Figure 4-83
~
"
NOTE:, CENTERLINE OF SPLICE PATTERN TO BE LOCATED IN REGION OF LOW STRESS (POINTS OF INFLE CTION PREFERRED)
olD
L...
40
'
• SPLICE .. ", SPLICE" LENGTH LENGTH'
,
Z
.
.
•
~
.)\
i
i.
-
".
MIN.
Typical sectio? through conventionally reinforced concrete wall I
4-235
.S.
",
~
" l'
,."'
.
.
. ' A·"
...
. . . . ~
'
',.
•
-' CONSTRUCTION 'JOINT
EXPANSION JOINT
,
.
..
'
~."
..,. '" '"
.'
'
•
o
.'
.. '.
".'
.
,.'
'. ::~.~.,.
..
'"
:.:,;>::.:
p
•
..'
=J r..-'-'::=::_=-_,~_ PAD <; W~~':,':I~NAL)
..
,'.
"
:".: • • I"
.,
: ;: ,"
.- .. I'
CONSTRUCTION JOINT
0
•
';. :4 ..'
'4.: .
"
CONTINUOUS WALL FOOTING
r:
"
Figure 4-84
e
,
Floor slab-wall intersections
e
e
•
-
• I--t ,
SEE NOTE
SPLICE
r
t
ALT, SPLICE
.
"
•
,"
..
",
,
-
'J'
9
' . .: .. '~
,
~
l
.
.
40'
,
:
J
,'~"',
': ~
'~:.'~:..
,",'
. - 40"
,
, "
,
-I
"",
.. ~ ,
t
.
.... t-
o
0
z
,
'.". N W
.....
..i
I .'r ",. ': I
~
,
"
"
'" '
'.\
,
-
•
•
..
•
•
401I
.... t-
.l. ALT, SPLICE
,
"
,
:...:.~'
.
L£ SPLICE
1-- ~t
... . I . .
,
•
O"
,lI)
-,
...~
.... ....
•
...o•
SPLICE
'SEE NOTE
-e-
O Z
,
.... ....
I
lI)
L
. I
• • ,
"
r
t
ALT. SPLICE
NOTE' 'SPLICES TO BE LOCATED IN REG IONS OF LOW STRESS (POINTS OF INFLECTION PREFERRED) BUT NO LESS THAN 40 1I FROM THE FACE OF SUPPORT,
0"
Figure 4-85
Typical horizontal corner details of conventionally reinforced concrete walls
,
,
~OF SPLICE PATTERN "
>. I
:I:
t
Figure 4-86
OF SPLICE PATTERN
Preferred location'of lap splices for a two-way element fixed on four 'edges
4-238
..
e
I- It.
.-
I
40~' .
I-t. SPLICE PATTERN
SPLICE PATTERN OF TOP BARS '
I MIN.
MIN. ;
'/
-
/'
"'
,
MIN"
..., '" '"
,
I.
'/
~
I><
'/
w
OF TOP BARS
4011'
..vP
..
r-
-
, r
rx t
SPLICE PATTERN OF BO'fTOM BARS .
NOTE: :t.'s OF SPLICE PATTERNS TO BE LOCATED , - - . IN REGION OF LOW STRESS (POINTS OF INFLECTION PREFERRED)
-.
..'.
.
Figure 4-87
Splice locations for multi-span slab
...
-"';
"
...... '
.
"
..
•
.
..
•
'
ADDITIONAL BOTTOM REINF.
DIAGONAL BARS
*4@6" HOOP BARS
CONSTRUCTION JOINT
U·---SPIRAL REINF.
'---COLUMN REINF.
:
CONSTRUCTION
'
JOINT - - _
'
•
,
.
.
.. ' .
Figure 4-B8
"
•
'
• .. '
'.
Section through column of flat slab
4-240
.'
,
"
,"
"
•
.. ..... , : ,,~,
.'
,~
,"
.
J "r-
LONGITUDINAL FLEXURAL REINF. -, ('
LACING REINF •
.~ on
....
. TRANSVERSE FLEXURAL REINF.
Laced reinforced elemant
Fiqure 4,-89 ,h
.
4-241
t;"
LONG FLEX. REINF.
TRANSV. FLEX. REINF. RA
t
t
OF FLEX. -tqo:~--I-''--- ,REINF.
OF FLEX. REINF.
WAll SURFACE LACING WALL, SURFACE
---1--
'7:
,',
,
i
NO:rE:, MIN. ,
.
SINGLE BARS
Figure 4-90
R.t
LACIN~
SHALL EQUAL, 4d';' ,
.
BUNDLED BARS
Typical lacing bend details
e
e
-3 ;x~
~
I
TYPE.
..J '
l'
~
-_.
TYPE •
~X8~,~,_,
~ ., '. ",. r" ~,
o
...,'" ~
'
I,',
LACING 'METHOD' NO.2
~
I'~
~
"'1 ;~
~,
TYPE b
'"
b",
~ ,
'~"" " ... /US-X_}-)-I':-'-""""4 '
~
r, r
;,
'-TYPE b
"",
TYPE b
"
l~
"
"",S d", .. S 2 ft . .!pSl.Sd",Dr,12 ft.
"'.,
,
LACING METHOD NO.1
!!!!.!!.:
l'
I
TYPE b
"
,i
1: '
J
TYPE •
I '
I
I
~'~
~ I
TYPEc TYPE d
'
LACING METHOD NO.3
:1 *1. Io
~
1 l~T,Y:.E. -p '"
I ,, .,.
J,
...
,.
"
TYPE d TYPE •
•
II
...ee
,LACING METHOD NO.4
LONGITUDINAL; SECTioN
';1,
.....
I 'I
o ,
;I
~
LJ.
.,
TYP E • TI PE d TYPE f
r
:-Et??TY~~p; ,
"
TRANSVERSE SECTION
',. ~"
Figure 4-91
b.
Typical methods of lacing,
c TYPE b
HORIZ.
.......
.. .
-.
I
.
.
,.
....
-
-
.A A A
L
-- -- -- ...-...--..
0) CANTILEVER
l
HORIZ •
,
'",. ~
.. .: :
- .. / " , /." .;/ VERT .1 I .
[;.00'
,-.',
: .. L
~'"
/'
~
.
-
~N.... ~ .
.... c)
i"
..
I
lL--- ._---
.... N
....
.
'" r
L
vEin •
V HDRIZ.
~
N .... z
.... z
N
~
VERT. ._--
:--
~
\
\ k.
--<~ -,
.I
N N
'"
a
VERT.
z
x
e
/.'
N
N .... z
THREE SIDES SUPPDRTEO AND ONE' SlOE FREE
">-
Figure 4-92
.
/
. HOllZ.
b) TMO ADJACENT SlOES SUPPDRTED AND TMD SIDES FREE
.. :-
.
~
..I
/-
.
'
I'
~
~
"
~
.'"
VERT.
',.
~
HORIZ •
'",
..
.,
d) FOUR SlOES SUPPORTED
I . I VERT. , ~
Typical location of continuous and discontinuous lacing
e
e
.
"
..
,. 1
-
ft bF' SPLICE :
(2
PATTERN
2
1
-"
TYPE A
"-- TYPE B
--"
"
TYPE C
..
"-'--.
"-- TYPE 0
.
•
TYPE A
TYPE B
..
a)
SPLICE PATTERN
~'.
..
, b) LACING SPLICE t,
Figure 4-93
Typic:"l detaii~
for: splicing of lacing bars
4-245
It
Q"
NOTE L.t IS MEASURED ALONG CENTER LINE OF LACING BAR BETWEEN 'POINTS "a"· AND b", .
, 2R.+db
A =s,t/d.t AND B =
Figure 4-94
..d.t .
Length of lacing bars
4-246
e
e
e TYPE"b"
TYPE "all
TYPE lib'" TYPE "all
D
-.
r:
HORIZ. LACING
VERT. LACING
3/4
I l:::.o(.
" IJ~TYPE"b" HORIZ. REINF.--r.;r"'"-'t.1 VERT. REIIlf.
A
A
TYPE "a"
'~
HORIZ. REINf.
VERT. REniF.7
HORIZ. LACING
11
CL. TYPE."b"
TYPEtla"
SECTION A-A
..,. ..,. '" I
.....
t
r~"< B 0'
.
TYPE B B
SECTION B-B
-
-ffi --
NI
~
TYPE B
C
VERT. LACING
,
h
~
DIAGONAL BARS· .
/FLOOR C
/
TYPE A
~LAB
.' 20
DIAGONAL
t:ONSTRUCTION JOINT
VERT.
° ..;0 ARS tlNF. -
~ ~ ~HOF J
• REINF.
r--
-<
'SECTION C-C
Figure 4-95
Reinforcement details of laced concrete walls
In ".
FLEXURAL, REINF. DIAGONAL
DIAGONAL BAR
TYPE B
TYPE A
• • e.
O"(TY,P,1
(TYP,)
DIAGONAL
BAR
TYPE A
..
Figure 4-96'
Typical detail at intersection of 'two continuous laced walls'
4-248
AS REQUI RED FOR FUL.L. ANCHORAGE
-I 0"
L'ACING MIN.
20d (TYP.)
.'.!
~
,
.
DIAGONAL CORNER , BAR (,TYP.)
,
t- ,
FLEXURAL REINF.
Figure 4-97
Typicai d'~tai·.l;. at intersection of continuous and d.i acontif nuous laced walls
4-249
,,
.
-,'
-<0 Z
..J ..J :::>
"
' - ::iE,
Ll..Q;
>-
Cl:1-
f2~
o
w
W
<.!l
« o.
Cl: Cl:
:::>
o w
::t:
U Cl: Z
en «
. : ':9".
«
201'1' (-TYP.) ,
" -L...J..L-I
',' ',
Figure 4-98
Typical~detail ~~
4-250
corner of'laced walls
.
/' ,,":/.;.
, ,
"
0"
,
.
-,
LACING' ~
,
"
(TYP.) ,',
* ". DIAGONAL CORNER BARI,TYP.)
"
i~~:
.
FLEXURAL RE INF. ~ii: 0)-
-
Nt-
"
,/
Figure 4-99
LEGEND: STANDARD HOOK IS USUALLY ADEQUATE
*
"
Intersection of continuous and disc~ntinuous laced walls without wall extensions'
4-251
,.
.~,
"LEgEND:
~.
*
"
~r
"·:1'
*
.-ExTmOR SURF:ACE
~t.
'
t~
,.
STAN'DARD HOOK ·IS USUALLY ADEQUATE
*
iii#! ,I ,!A
d::~) ~'j =f 'O~
.,., I~A'
f'o,)
.."
/
h.
~I'
o
.!!!
'01
i
b
i
',']
0:'
'"
cf'
\.30 I TVP.)
1-0 W_MIN,ITV")
,
INlERKlR SURFACE
.
a:
;
"
EXTERIOR LOAD
t--- om:RlOR SURfllC€ INTERIOR U:>AO
,bl FLEXURAL. REtNF.
01 LACING
,.., 4.
Figure 4-100
Corner· details fo~ laced walls without wall' extensions
,
.-'
lJ-F. . ,:
"'
e
e
e
"
,SINGLE LEG 'STIRRUPS
:· .
I,
....:. ::.;
-::'::~ .~
.. . ·.
·"
.'
','
'
,.'
.'
"
'.
,
'.
,
'.'
.
"
'
D,
"
"
....
FLEXURAL
Figure 4-101
. '.
..
,~.
"
•,
'
•..
'
REINF.
Element reinforced with single leg stirrups
4-253
AS REQUIRED FOR FULL ANCHORAGE
,SINGLE LEG STIRRUP
" -6"
20 0
:
MIN. t.~:
,.,
....
'.
v,
'"
",
,.
..
. ~,. '.
"\
:
:
... - .. ..
-
,1
,
,
Figur; 4~l02
--
-'
.,
.
-.
-
. ',.
, '
,
-,-;,;.,,~
"
~
,:
,
. '. .. ..:. . -- · .
.
",
~.
'-,
.;~
",
-"
.~
"
...
.: "" ~. : ....
.
Typical detail at corner for walls reinforced with single leg stirrups
4-254
DIAGONAL BAR
I
.. . , .. , '-
-BAR
.. ':'.,
.:.:':.fl: ': .:.;
.
LOSED TIE
"
.-~~,. ";'
..
U
.". .. W'
, ..... -. . -,. • . < 1 . ..
.- .. ..:. - .... '. .... ; .. -- .' : . ". .. .~". , .
>
-
.
~ ; ~•• _f ••'. ....
-",
AS REQUIRED:FOR _ 'FULL ANCHORAGE
~r.·
MI,~ l~
-
.••.•',: .0" •
l'
..
'j'
.'. "
HORIl. FLEXURAL,i .'. REINF. .• t • ,
~
>.'
.-..
FLEXU'i~;(R-EINF.-;:'
;.
.. ,.. '.',· ....-, .
.,:.;.
I'\.
ll'
;; •. v:.·~.·, . . , .
'"' VERT.
.
.
"
~: :-
\ ...... '-IIL'.' "'.' -. ~: -'. • t' -.- . ,.",,~~
MIN.
- .-
-
-,
I
SINGLE LEG STIRRUP
-' ,
~I
n BAR
'. '. J . "
"
. . 0'
. . ,•-c-• "
"I
- ,
.i. iJ'
VERTICAL REINF.
'
'
I-,
-,
"
.
r-HORIZ. REINF.
,
r
.,
I-
3;4'CL'
,
-
c,,
i
,r-
"
!""
.; ;
SINGLE LEG'STIRRUP ,
,
-
,
...
""
,
,
~-
'.
,-
t-
,
. "
-'';'
ii,
:........:; ....:-:
-
~
'~.'
.
c.
"
.
"
•. --
N
-,
I
-
/
C/'" ..
DIAGONAL BAR
~~, ,.~ ,. ,-r<.,',. -,
.
~
.
~.
- . - ..
",'
'
_
,
,
"
},.,
,
'
'
,
,
1(3" CL
.',- 1--'-',' -, t:f CL. ... .
.
"
! J
,
! ,
Figure 4-103
•
FLOOR SLAB,
"
'••....:.-.'
. ... t·"...... '.• { .:..- .. . .- -...... .,-.' , , ,-. . o' • .
- ..
"
0
:i
,;:.
':,
, , ,
S
CLOSED TIE
,.
,"
;,
~
~',;,"
-
"
~ . ~ 1-".
,
-~"
..
.. :..,.'.
WORKIN'G PAD '
..• "-v.-
"
Reinforcement detail of wall with single leg stirrups
4-255
.. :~i',' .'.. ;;;
TieS ystem
'j"
,
.~!:i :.~::~ ~.
II
::.::..: ..': F
f,.;
...
....
1:~llli
;';.(~..
t.- :.', f."
I
I- ":':'. ~. ; •••. f."
N
'" '"
,:.;-. ......:;jj
'1'"
~
7, A
I
.~ '~
=l" Open
, ~
C'Sand Fill
.
II
Sand, f ill
.' .
. .
,.
'
-
-.'
t-Pedes 01 , "-
"r" I·... ,
.... r Slob
" . ,•
,..
-
~
.,
,w
.....
i I
=.;
0' Top Figure 4~164
e
.....
l
Closed
0' Top
Typical composite wall details
e
e
e
e .'
:.
e
::-.
~
G.
~
, PLAN
SECTION A-A "
.
..... ,.
N.
-
'
. -
0) DISCONTINUOUS . WALL CONSTRUCTION.. •
'-
-
."
•
•
<
"
•
..
.... '"
.z," -:':
G..
-.!J .' ,
.'
,
'-'
.s
PLAN --
SECTION 8-8
b) CONTINUOUS ,
.
WALL
C'ONSTRUCTION '
"
Figure 4-105
Typical single-cell structures
, .'.,
...•.
"
',"
'".1
SCHEME
d ) CONTINUOUS SIDE WALL CONSTRUCTION
SCHEME b) DISCONTINUOUS SIDE WALL CONSTRUCTION
Figure 4-106
Typical multicubicle structures
4-258
LEGEND:
(!)
Z U
I
4th POUR
ISS:'J
3rd POUR,
Wffi3
2nd POUR
mm
I st POUR
I
IL
0
oJ I- oJ Z IlJ I- Z
~
SIDE WALL
X 0
IlJ N 0::
0
:t:
~_HORIZ.
CONSTR.
JOINT
'(!)
Z IL -
OU
l-oJ
z..J 1lJ
BACK WALL
I-
u X_
IlJI0::
IlJ
>
~~:--~+-- HORIZ. CONSTR. JOINT .......--FLOORSLAB
--~WALL
PEDESTAL
~~fj~~Jr---'WORKING
Figure 4-107
PAD
Pouring sequence
4-259
TK 5-1300/NAVFAC P-397/AFR 88-22
APPENDIX 4A ILLUSTRATIVE EXAMPLES
TH S-1300/NAVFAC 'P-397/AFR 88-22 Problem 4A-l. Problem:
.
Elements Designed for the Pressure~Time Relationship ,
"
,
,
,
.
.'
Design an element: which responds to the pressure'"time' 'relationship. •
f'"
'.
Note:' Steps S, 8,·10 through 18; 22 'and 23 are'specl!fic for two-way elements" "however, references' are givim. defining similar .prccedures for one-way elements. ,
Procedure: <
. a. b. c. d. ':-
: i'l
..
Establish design parameters:
Step 1.
,
r
Blast loads including 'pressurectime relationship,(Chapter 2). Deflection criterta;, . "' r,','_ ,.' '. J"" 'J:' Structural configuration including geometry andnsupport'conditions . . , . . :. .' J ' . i Tjpe of section available to·resistblast; Type I, ,II·or ,III I depending upon' the occurrence of spaLl.Lng and/or crushing of the concrete cover': J ! r~-:' ~.:~'~. . 0'
'.
•• j
Step 2.
Select cross section of element including thickness and concrete cover over reinforcement ..i.·'Also-'determine'static, stresses of . concrete and'reinforcing steel (Section 4-12)" . ',". 'n.
Step 3.
Determine dynamic increase factors for both concrete and reinforcement from table 4-1. Using the' abovejDIF"and' the 'static stresses of s.tep 2, calculate the dynamic strength" of:'materials.
f, ",
...
Step 4.
Step 6.
,,.
Determine the dynamic design stresses 'using, table 4-2·:and the," results from step 3. jo
Step 5.
• .f,l ;"
,
Assume vertical and horizontal reinforcement bars to yieid the optimum steel distribution', 'The- steel' distribution is optimum when the resulting' yield: Lf.rie s make an, angle of 4S"degrees 'with supports. . i . \. . a . ' .' Calculate de (d or dc, depending upon type of cross section available to res Is t blast) for both the pcs LtIve :and negative' moments in both vertical and horizontal.directions. Determine reinforcing ratios. Also check for minimum steel'ratios from table,4-3.
, ..
Step 7.
Using the area of, reinforcement'; ,t~e value "of' d,,',ifrom step 6, and the dynamic design" stresses of, step': 4 ;r calculate' the moment capacity (Section 4·17) 'of both the-positive and'negative reinforcement. '. -0. 4· Note: Steps 8, 10 through 18 are required to determine the actual and equivalent resistance-deflection curves for two-way elements. To obtain these curves for one-way elements, see problem 4A-6.
4A-l' .,
TK'5-1300/NAVFAC ·P-397/AFa 88-22 Step 8.
Using the equations of table 3-2 or table 3-3 and the. moment capacities of step 7. calculate the ultimate resistance in the plastic range.. '.'
Step 9.
Using equation 4-4, the static concrete stress of step 2,. and unit weight for concrete equal to 150 psf, calculate the modulus of elasticityf!,r concrete. With th~ above modulus for concrete and that for steel (eq. 4-5) and equation 4-6,. calculate the modular
.
=t~.
Step 10.
With the use of equation 4-9a and the assumed concrete thickness· .of step 2, calculate the gross moment of inertia of the concrete. Using the value of de for the negative and positive reinforcement of step'6, calculate an average value of de' Also calculate an average' percent of.positive and negative reinforcement using the above de and the area of reinforcement 'of step' 6 .Ln both vertical and ,horizontal· directions. With the.values of p(average) and figure 4-11 or 4-12, determine the values of the constants F and calculate the moment of inertia of a cracked. section w~th equation 4~b in both directions . .Calculate the average cracked. moment 'of inertia for the element using equation 4-19, and also, the average moment of inertia of the element from equation 4-7.
Step 11.
Using·equation 3- 33, and the modulus. of elasticity.. of step 9 and the moment of inertia of'step 10, calculate ·the unit flexural rigidity.
Step 12.
Establish'points of interest and their ultimate moment capacities (fig; 3-23).
Step 13;
C0!Dpute properties of first yield."
"
"
a.Location of first yield. Resistance at first y,ield·r e. " -c , Moments at remaining points consistent with r e.· d. . Maximum deflection at fir~t yield. b ,..
Step 14.
Compute properties at second yield.
" .. Remaining moment capacity at other points . Location of second yield. Change in unit resistance Ar, between first and second yield. Unit' resistance at ~econd yield. rep' Moment at remaining points consis~ent with rep' Change in:maxi~um deflection.' Total maximum deflection.
.
d, e.
f. g.
-.
4A-2
•
TH 5-l300/NAVFAC'P-397/AFR 88~22 ,: Note: 'An element, with unsymmetrical support condd t.Lons may exhibit three 'or four 'support' yields.' Therefore, 'repeat step 14 as many time,~ as necessary 'to obtain properties at various yield ,points. . ' ,-'
Step 15.
.
,
,
'.Compute properties at final yield (ultimate unit 'resistance). ,',' . \ ~"
a.
Ultimate unit resistance.
b. ~r:
c;
..
'
".
.Change' in :resistance . between': 'ultimate and resistance at prior yield,
"
.. '':
Step
iz.
unit resistance .. '
Change in maximum deflection (for elements ,s~pport~d on two, '. three or four 'edges" use stiffness ,obtairied".from· figure 3<26; 3-30 and 3-36 ,',respectively) . .,'
d. Step 16.
,
' . . . J:.
Total maximum 'deflection.
,'"
t
_ ,,_I:
Draw the actual resistance-deflection curve (fig. 3-39). Calculate equivalent-maximum elastic deflection of, the' element. r'
Step 18.
Calculate the equivalent elastic unit stiffness KE from equation 3-36. .j, .'
Step 19.
Determine I the
1-\
. •;.
Note: '
"
......
For one-way elements, use table 3-12 to determine the' average load mass factor. Step 20.
Using the effective mass of step 19 and the equivalent stiffness, calculate the natural period of vibration TN' from equation 3-60:,
Step 21.
Determine the response chart parameters: a.
Peak pressure P (step 1) .
b.
Peak resistance r u.. (step 8) .
c.
Duration of,load,T (step 1).
d.
Natural period of vibration TN (step 20).
4A-3 "
I"
,:
.
TH 5-l300jNAVFAC'P-397/AFR 88"22
Also calculate the rat~os of peak pressure P to peak resistance r u and duration T to period TN' Using these ratios and the response charts of Chapter 3, determine the .value of Xm/XE and tm/T N '. Compute the value of Xm and compare it 'to the maximum' permissible deflection of step 1, and if found satisfactory, proceed to step 22. If comparison is unsatisfactory, repeat steps 2 to step 21. In addition, compute the value of Cm/t o .from tm/T N and T/TN and' ,assuming that T - to' determine whether or . not correct procedure has been used; for elements to respond to the pressure- time relationship, 0.1 < tm/t o < 3';','''' Step 22.
•
Using the'ultimate resistance of step 8, ,the value of de of step 6 and equations of table ..4-7,calculate the- ultimate diagonal tension shear stresses at distance de from each support. Also 'calculate the shear capacity of the element from equation 4-23. If the capacity is greater than that produced by the load, shear reinforcement is not ,required. However, if the shear produced by the load is greater than the capacity, then shear reinforcement must be added to resist the excess. Note:
"
For one-way elements, use table 4-6·to establish diagonal tension shear stress. Step 23.
Using the equations of table 3-10 or 3-11 and the ultimate resistance of step 8, calculate the shear at the supports. Determine required area of diagonal bars using equation 4-30.· However, if section type I is used, then the' minimum diagonal bars must be provided (eq. 4-31). " ".' :" , . Note: For one-way elements, use table 3-9'to calculate the shear at the supports.
Example 4A-l, Elements Designed for the Pressure-Time Relationship Required:
Design a wall which spans in two directions and is fully restrained at all supports for a given blast load.
Solution: Step 1.
,
Given:
a.
Pressure-time loading (fig. 4A-l) ..
b.
Maximum deflection equal to 3 times, elastic, deflection.
c.
L - 180 in., H - 144 in. and fixed on four sides (fig, 4A=1).
d.
Type I cross section.
4A-4
•
TK 5-1300/NAVFAC P-397/AFR'88-22
1 35 iii
APPLIED LOAD
§
CIl CIl
w
...a: 10,5-
TIME,rna
BLAST LOAD
SECTION· "
• Figure 4A-l Step 2.
. Select cross .sec t Lon of element and static st r e s s of reinforcement and concrete (fig. 4A-2). ,
'
I
".
NEGATIVE REINF. _
CL.
"
.• ::. ':', '.:: . I--POSITIVE REINF. ~"":. -.::. .'. INTERIOR SURFACE
'. 1'12'
.."....
..
f'
c - 4,000 psi
\'
.fy - 66,000 psi
•
Assume .T~ - 12 in. and concrete cover as shown,
EXTERIOR SURFACE
'. :
, !,~~\~,~/,:~ 1
.,
v
Figure 4A-2 Step 3,
Determine dynamic stresses. a.
Dynamic increase factors - DIF (table 4-1). Concrete:
,
Bending Diagonal tension
1.19 1.00
'
4A-5
..
TH S·1300/NAVFAC P-397/AFR"88-22 Reinforcement: Bending Diagonal tension Direct shear b.
1.17 1.00 1.10
Dynamic strength of materials
.
Concrete (f'dc): . Compression DiagonaF'tension
1.19 (4,000) 4,760 psi 1..00 :(4,OOOr -4,000 psi
Reinforcement (f dv): Bending Diagonal tension Direct: shear ," Step 4.
- 1.17 (66,000) - 1.00 (66,000) - 1.10 (66',000)
77,200 psi 66,000 psi 72,600 psi
Dynamic design stresses from table 4-2. Concrete (f'dc): Compression Diagonal tension Reinforcement
4,760 psi 4,000 psi
(fds - f dy): ,
>
Step 5.
Bending " Diagonal ·tension Direct shear
.
77 ,200' psi· 66,000 psi 72,600 psi
...
In order to obtain optimum steel ratio PV/PH; set x'- H/2 to have 45 degrees yield lines.
x
H
L
2L
144 - - - - 0.40 2 x 180
From figure 3-17,
., ."
Therefore, 1. 43 x 144 ] 2 [
- 1. 31 180
Try No.4 bars at 10 in. o.c. in vertical direction, and No.4 bars at 12 in. o.c. in horizontal direction. 4A-6 .
TK 5-1300/NAVFAC Step 6.
P~39~/AFR 88~22
Calculate de and steel ratios for,each direction, minimum reinforcement ratio is equal to 0.15 percent. Vertical: Vertical reinforcement bars are No, 4 at 10 in. o.c. from step 5. AsV - .20 X 12/10
.
Z
-
.~
0.24 in /fe--
Negative moment
dv -
12 - 1. 5 - 0.25 - 10.25 in
Positive moment
dv -
12 - 0.75 - 0.25 - 11.0 in
0.24 ------------ 12 X 11.0
bdV
0.00182
> 0.0015
O.K.
.'
Horizontal: Horizontal reinforcement bars are No.4 at 12 in. o.c. from step 5. AsH - .20 X 12/12
.../ft~
-
0.20 in
Negative moment dH - 10.25 - 0.25 - 0.25 - 9.75 in Positive moment dH - 11.0 - 0.25 - 0.25 - 10.5 in ;'
.20 -
0.00158
>
0.0015 O.K.
12 X 10.5 .
Step 7.
'~.
Calculate moment, capacity of both positive and negative reinforcement in both directions. a.
Depth of equivalent rectangular stress blocks. a
Asfds
- .85 bf'dc .24
aV -
77,200
X
- .382 in .85
X
.20 aH -
(eq. 4-12)
12 X
X
4760
77 ,200
;
, .85
- .318 in X
12
X
4760
.,
4A-7
, "'b:' 'Homent capacity (eq. ,!-11). i ' .
i'
(d- a/2) ,
, 0.24 (77,200)(10.25 vcrt,'(A ( 1'1,,~~/,'o/~
•J
. HVN t7 . ....
,
vcr],',.
J
f" ,~j, ':/i
f)
HVp
t.
-
12,
HHN
., HHP
-
.
-
15531 in-1bs/in e.
.382/2) 16689 in-1bs/in
12
~1 ~.,
0.20 (77,200)' (9.75
-
.318/2) ,
12
o
'.~'.
-
:'382/2)
-r:
• 0.24 (7'7,200) '(11.0
,.
-
0.20 (77,200) (10.5
.";
, ~
.318/2)
12340 in-1bs/in '
,
•
i)
(',i'
13305 in,-lbs/in 12
,.
,
.
·t·.... 0"
Step 8.
Determine ultimate 'resistance of the element. '.
"
:
Vp [ HVN +,.H
H
HHN + HHP
From figure 3-17; . 'r
...
t
'
180 '"' 1. 43 (step -s)
12340 +'13305
144 :.~;'"
•
L
-;':
x
.405 L
x - .405 X 180 - 72.9 in
',,':
Ultimate resistance (table 3-2).
' ,':\.'
5(12,340 + 13;305)
~tep
9.
"
24:13 psi
ru -
(72.9)2"",
Determine modulus of elasticity and modular ratio: a.
Concrete (eq. 4-4)
4A-S'
TM 5-1300/NAVFAC P-397/AFa. 88~22
Ec - w1 . 5
33
(f' c)1/2 - (150)1.5 (33) (4000):l/2
- 3.83 X 10 6 psi b.
.
.
.1.
Steel (eq. 4-5) Es - 29 X 10 6 psi
c. lltep 10;
Es
n -Ec
-
«. ::
29 X 10 6 7.56
~:
3.83 X i0 6
(eq. 4-6)
.,~
Determine average moment of inertia for an inch strip. Gross moment of inertia (eq. 4-9a)
a.
3 T· .c I -g 12
..
12 3.
'd
- 144 in4/in.
--12
'i~
b.
-,'
Moment of inertia of cracked section (eq. '4-9b). iJ
Vertica1.direction: 10.25 + 11.0 d(avg)- - - - - - - - :10.625 in 2 , '
As
. fX(avg) - - - bd(avg)
0.24 ------------ - 0.00188 12(10.625) , ','
..'
.. F - 0.0102 (fig. 4·12)
IcV - Fd(avg)3 - .0102 X (10.625)3 - 12.2 in4/in Horizontal direction: 9.75 + 10.50 . t.,:.
4A-9
t •
TK 5-1300/NAVFAC.P-397/AFR 88-22
- .00165
p(avg) - 12 (10.125) .. F - 0.0092
I cH - .0092 X (10.125)3 - 9.5 in4/in C.
,
"Average moment of inertia of cracked section. (eq. 4-10) (180 X 12.2) + (144 X 9.5) I c - -----------
180 + 144 '
d.
Average moment of inertia (eq. 4-7).
" .
144.0 + 11.0 - -
-
-
-
' .., - 7 7. 5 'i n 4/ i n·
' i.
2
Step 11.
Calculate unit flexural rigidity. (eq. 3-33) Use
y
-
.167 for concrete
Therefore, (3.83
X
10 6 ) 77 .5
305.34 X 106 in-lbs
1- (.167)2 Step 12.
For points of
Step 13.
Properties at first yield.
interest,.see~figure 4A-3;
4A-10
TIl 5-1300/NAVFAC P~39j/,AFR'88-22' ' From figure 3-33 of Chapter 3 'fdr H/L,:' 0,80
.,.,
"
~,
r'
.:
r '.
I'
...
"
( , ',
c. ' . ( . ." POINT 3, My" '.!\, :'
)< .
'I
r:
,.- POINT 2, MHII
• -e J
~".-
"
, . "
POINT' 3, II YII
. ,.". . ,.
'.
..,
',',
I,
Figure 4A-,3.'
'
'
. i
.. ..,'
B1H - 0.023
B2
- 0.056
BlV - 0.031
B3
0.068
.
I;d
Y1 - 0.0018 .~ ~
a.
.1
'
MHp - 13.305 in-lbs/i.n .
MVp.-
• i-
16,689
;
,
;): 1
MHN - 12,340 in-1bs/in f· .. :..-, '. . '1 I
in-1bs/in
'.
MVN - 15,531 in-1bs/in
M
.. r -
BH 2
rlH - 13,305/[O.023(144)2 J - 27.90 psi. rlV - 16,689/[0.031 (144)2] - 25.96 psi r2
- 12,340/[0.056(144)2] - 10.63 psi
4A-ll
(eq. 3-25)
TH
~~1300/NAVFAC
P-397/AFR88~22
r3 ~
15,531/[0.068(144)2],- 11.01 psi ..
"
Fir~t
yield at point 2 (smallest r).
b.
r e - 10.63 psi'
c.
M1H - (0.023)(10.63)(144)2 ~
5,070 in-1bs/in'
M1V - (0.031)(10.63)(144)2 ~
6,833'in-1bs/in
.
.
M3 - (0.068)(10.63)(144)2 - 14,989 in-1bs/in d.
Xe -
4'
I
Y1reH /D
~."
(eq. 3-32)
Xe - (0.001.8)cl,O.63)('i44)4/305:34
Step 14.
.
Properties at
-
X 10 6 -0.0269 in
yield,
se~ond
,After first yield element assumes-a sinipie-simp1e-ftxed-fixed configuration, therefore, figure 3-34 for·H/L ~,O.80.
B1H - 0.020 ,
:
~
BlV -0.039-
11 a.
- 0.0022
M1H - MHP
M1H (at r e) - 13,305
Mpj -MVp
MlV (at ,r e) - 16,689
M3 -MVN
M3 (at r e) - 15,531
5,070- 8,235 in-1bs/in 6.,833 - 9,856 in-1bs/in
- 14,989 _. ,
542 in-1bs/in
.' .{
M
b.
Ar -
-
-BH2
Ar1H - 8,235/[0.020(144)2] - 19.86 psi .;
ArlV - '9,8561['6.039(144)2]',-
1;.1~
psi
•. ' '1
l ~:. '
l.-.
•
".;'
. .
4A-12
,',.
TM 5-1300/NAVFAC P-397/AFR 88-22 &r3 - 542/[0.076(144)2] -
0.34 psi
Second yield at point 3 (smaller &r) c.
&r - 0.34 psi
d.
rep - r e + &r - 10.63 + .34 -
e.
K1H - (.020)(0.34)(144)2 +
f.
K1V - (0.039)(0.34)(144)2 + &X'- Y1&rH4/D
10.~7
5,070 -
psi
(eq. 3-26)
5,211 in-1bs/in
6,833 - 7,108 in-1bs/in
&X - (0.0022)'(0.34)(144)4/305.34 X 10 6 - 0.0011 in ,. .028 in
g. Step 15.
Properties at· final yield (ultimate unit resJ.stance). After second yield e1ement:assumes a simp1e-simp1e ..simp1e-s;l.mp1e configuration, therefore, from figure 3.36 for H/L - 0.80. 11 - 0.0054 a.
r u - 24.13 psi .(from step 8)
b.
&r - r u - rep - 24.13 - 10.97 - 13.16 psi &X'''Yr&tH4/D .
c.'
.
'
d.
&X -(0.0054)(13.16)(144)4/305.34 X 10 6 - 0.100 in ~
- Xep.t &X - 0.028.+0.100
-
0.128 in
Step 16.
For actual resistance deflection curve, see figure 4A-4.
Step 17.
Equivalent elastic deflection from equation 3:35.
4A-13 .
,
, r,
.,
....
..
'.' . .
ru = 24.13
'/0'
EQUIVALENT - y
ILl U
z
b
<[
I-
'
, i. ,
,
I/) 1/)'
ILl
ll::
r.p =10,97
I-
r. =10.63
z
,.:..-,-
/ ACTUAL
,
:
::::l
><.=0,0269
X. P =0.028 ,
,.'
"
,
..:.
FIGURE
4A
~4
.
'.-
...,
TH 5-1300/NAVFAC P-397/AFR 88-22 XE - Xe(rep/ru) + ~ep[l- (re/ru)] + ~[1 - (rep/r~)] XE - 0.0269 (10.97/24.13) +
0~02~
[1 - (10.63/24.13]
+ 0.128 [1 - (10.97/24.13)] - 0.098 in
The equivalent resistance-deflection curve is shown in figure 4A·4. Step 18.
Calculate equivalent elastic stiffness. KE - ------ - ------
Step 19. a.
-
(eq. 3-36)
246.2 psi/in
Ca1culate'effective mass of element. Load mass factors (table 3-13 and fig. 3-44),'
x/L - .405
L/H - 1.25 Elastic range
- KLK - .61 + .16 (1.25-1) - .65
Elasto-plastic'range: two simple edges
- KLK - .62 + .16(1.25-1) - .66
four simple edges
- KLK - .63 + .16(1.25-1) - .67
Plastic range'
KLK -
KLK(avg. elastic and elasto-plastic) -
.54 .65 + .66 + .67 3
.66 + .54 KLK (avg. elastic and plastic) b.
- .60
Unit mass of element.
wT c m-g c.
. 2
-
150 X (1) X 10 6 - 2,700 psi -' iIIs2/in 32.2 (1728)
Effective unit mass of element.
4A-15
-
.66
TM 5-1300/NAVFAC P-397/AFR 88-22 ,
me'- KLMm- .60 (2,700) - 1,620 c-
Step 20.
Calculate natural period of vibration. .:
i
..
TN
Step 21.
pSi-ms~/i~
- 2(3.14)
[
i, 620
(eq. 3-60) ~
]~:...
'16.1 ms
246.2
Determine response chart parameters (fig. 3'64a). Peak pressure
oP -.35 psi {step 1)
Peak resistance
.ru- 24,13·psi (step 8)
Duration
T - 10.5 ms .(step 1)
Period of vibration
TN - 16.1
P/ru - 3?/24.13 - 1.45
T/T N - 10.5/16.1 - 0.65
ms
(step 20)
From figure 3-64a:.
~ Use assumed section
(fig. 3-64a)
The correct' procedure has been used since tm/t o - 0.923 is within the range, 0.1 < tm/t o < 3. Step 22. a.
Check diagonal tension at de distance from support .
.
.
Ultimate,..shear. stress (table 4-7). , VuH -
-where de
4A-16
~
dW.(of negative moment)
TN S21300/NAVFAC P-397/AFR 88-22 3 X,24.13 (1 - 9.75/72.9)2 VuH - ----~---------
(9.75/72.9) [5 - 4
- 91.0 psi
(9.75/72.9)]
3ru (0.5 - de/H)(1 - x/L - 2 dex/HL) de/H(3 - x/L where de -
dv
8 dex/HL)
(of negative moment)
3 * 24.13(.5 - 10.25/144)(1 - .~05 ~ 2*.405*10.25/144) VuV - - - - - - - - - - - - - - - - - - - - - - - - (10.25/144)(3 - .405 • 8*.405*10.25/144) -99.1 'psi b.
Allowab1e',shear,. stress (eq. 4-23). V
c -
[1:9 ...
wher~p
(f'dc)1/2 +
2,500 p] S 3.50 (£'c)1/2'_ 221.4 psi
is, the steel ratio at support
, , ' [ 2,500 (0.20) ] vcH- 1.9 (4,000)1/2,+ ----'------- - 124.4 psi> 91.6 psi 12 (9.75) (4,000)1/2 + [ .. iitep 23.
2 , 500 (0.24) ] -
125.0 psi,> 99.1 psi
12 (10.25)
tio stirrups required
Det~rmine
minimum area of
~he
diagonal bars (cross section,type
I) .
. (eq., 4-31) where d ,fa equal'to de at support. Using a-45°, AdH - 124:'4 (12 X 9.75)/72,600 (0.707) - .283 in2/ft ""
AdV -125.0 (12 X 10"25)/72,600 (0.707) -
.
.300 in 2/ft
..
4A-17
~
TH S-1300fNAVFAC P-397/AFR 88-22. use #5 diagonal bars @ 12" Problem 4A-2.
Preliminary Flat Slab Design for Large Deflection
Problem:
Design a flat slab for large deflections.
Procedure: Step 1.
Establish, design' parameters: a.
Blast loads including pressure-time relationship (Chapter
b. c.
Deflection criteria. Structural configuration including geometry and support conditions. Type of section available to resist blast, type I, II or III depending upon the occurrence of spalling and/or crushing of
2).
d.
the concrete cover.
Step 2:
Select cross section'of the slab and the col~ or column capital.' Include concrete cover over reinforcement and maximum size of the reinforcing bars in the flat slab. Also determine allowable static stresses of concrete and reinforcing steel (Section 4-12).
Step 3.
Determine dynamic increase factors for both concrete and reinforcement from t~ble 4-1.' Using the above DIF and the allowable static stresses. of step 2, calculate the dynamic strength,of materials'.
Step 4.
Determine the dynamic design stresses using table 4-2 and the results from step 3.
Step 5.
Determine the ratio of the flexural stiffness of the.wall to slab in both directions using equations 4-50, 4-51, 4·62 and 4-73.
Step 6.
Proportion total span moments to unit column and midstrip moments in both directions using equations 4-52 through 4-60 and 4-63 through,4-71. Note: Use equivalent. frame method for· the direction(s) with only two spans.
Step 7.
Adjust unbalanced negative unit moment at column and midstrip in both directions of the roof. Correct the corresponding positive moments to maintain the same total span moments.
Step 8.
Calculate total external work done by r u from equation 4-74 using yield line patterns similar to figure 4-24. Use uniform deflection (A) for all positive yield lines.
4A-18
TH 5-l300/NAVFAC P-397/AFR 88-22
Step 9.
Calculate, 'total internal work done using equation 4- 76 and the unit moments determined in steps 6 and 8.
Step 10.
Equate,the total external work to the"internal work (equation 477). Solve the resulting equation for the ratio of ru/HoH . Use equations ,4-61 and 4-72 to substitute HoL with HoH'
Step 11.
Determine the minimum value 'of ru/HoH'by trial 'and error proce., dure. 'Vary the assumed'value of one of the yield' location vari,. 'abIes while assuming a constant value for the rest to find the minimum' rl.l/HoH" Repeat untllall 'yieldl1ne locationvariables , are established (X, Y, Wand 'Z): The last step wllLyield the final minimum value of ru/HoH to be used in the following steps. , ,..
Step 12.
Calculate the load-mass factor for the flat slab 'using the procedure outlined in Chapter, 3, for two-way elements. Use equation , 3-59 'for the slab' sectors with. no drop panel and equation 3-58 for the,:slab sectors with drop panel.
Step 13,
Calculate effective unit mass of the slab using the larger de of the assumed slab 'section from step 2'and equation 3-54.'
Step 14.
Calculate ,the'maximum deflection of the flat slab using the shortest 'sector ,length (Ls)"
Step 15.
Determine the required unit resistance (r"vail in equation 4-90) \ ,to resist the given -Lmpu l se loading (Chapter 2) and the' values from steps 13 and 14, Check that 'the correct procedure was used.
Step 16.
Determine, the uniform ,dead 'load of the flat slab and calculate the ultimate resistance of the slab (r u) from equation 4-90.
Step 17;
Determine the required total panel moments in each ,direction using the 'ultimate resistance from step 16, the minimum value of ru/HoH from step 11 and "the ratio of HoL to HoH established in step 10.
Step 18.
Calculate the minimum required unit moments in each direction from step 6 or 7 using the v~lues of HoL and HoH from step 17.
~
T'·.
':.
.
\
~-
.',
Step 19.
Calculate' the minimum moment capacity ,of the 'slab section in each direction'by choosing reinforcing bars, ;rhese capacities should be equal to or slightly larger' than the corresponding moments from step 18. Also check for minimum reinforcing ratios from table 43.
Step 20.
Determine provided, resistance in each direction by using 'the ratios provided to,requibid unit moments from steps 18 arid '19. Find the average of these values to establish the unit resistance of the flat slab.
Step 21.
Determine ultimate tension membrane capacity of the flat slab using equation 4-85. Find the average of continuous steel ,in the
4A-19
TM 5-1300/NAVFAC P"397/AFR 88-22. mid and column strip using: unit ,moment ratios from step 6 or 7 and step 19. Use f ds for bending from,step 4 in calculating unit tension forces in the continuous reinforcement, in each span direction.. , ' Step 22.
Calculat~ diagonal tension stresses at de distance ·from the edge of wall supports according to Section 4-31.2 in both directions. Determine concrete capacity in diagonal tension from equation 4-23 ',.>.'
using the ratios of unit moments -from
steps.6~
or 7 and the rein-
forcing ratios from step 19. If the. diagonal. tension stresses are larger chan the concrete capacity, 'single leg stirrups should be used' or a drop panel be, added along the wall in lieu of a change in flat ·slab cross section. If drop panels are used, the diagonal shear ~tress at de distance from the edge of wall drop panel must also be checked. ,. •
'n
Step 23.
Check punch Lng shear d e/2 df.st ance out and around the column or column capital. Use the load area ,between positive yield lines minus the area supported by ~olumn or its capital. If the shear s~ress is larger than 4(f'c> , use a column drop panel and check the punching shear with the new thickness of the slab over.the column.
Step 24.
Determine the size of column drop panel ,by checking punching shear d e/2 distance out and around the drop panel.
Step 25.
Chec k one-way vdf.agona I ",hear
yield lines de distance out from the column drop pane~ in each direction. Use equation 4-23 to find concrete capacity. Increase column drop panel size, ,if. r'equf r ed or use 'single leg shear stirrups according to Section 4-18..3.
Step 26.
Check one-way diagonal. shear, stress'between positive ,yield lines for an average de distance out from the column capital similar to step 25. Average de is b~sed on the width-of the drop panel to the total width. Increase column drop panel width or.thickness if required'.
.,
,
.'.
stress~"between positive
,. 'r
Step 27.
Assume preliminary reinforcement for'the flat slab using unit moment ratios from step 6 ,and 7,-MoL'and.MoH from step 17 and equation, 4-19 with the slab thicknesses established throughout this procedure. Calculate all ,actual' unit. moment capacities. Note: 'Check the actual flat slab resistance using' unit moments from step 27 and the established sizes and thicknesses; .of drop. panels. Repeat steps 8 to 27 for the actual values' 'in each direction. Also provide diagonal bars at wall and column according to sections 4-19 and 4-31.2.
4A-20 .
Example 4A-2. .»
Required:
Preliminery Flat Slab Design for Large Deflection ,.:j,'" . t- .
,
H
Design of a flat slab with three equal spans, in each direction for ... • f._ . large deflections. ~:
Solution: Step 1.
Given: a.
P - 96'psi,
T - 15 ms and. triangular loading.
b.
Maximum support rotation of 8 degrees.
c.
L - H - 240 in., continuous walls all around 207 in. high
and 21 in. thick.' .
Step 2.
'fl,
Type III cross'section.
d.
, j'
Assume:
a.
Tc - 15 in. thickness of flat. slab.
b.
D - 45 in. diameter of colUmn capital.
c.
Concrete cover r-
outside 2 in.
i
~":.
inside 3/4 in. d - 3/4 in. largest bar diameter.
d. L
•t
.
. f'~'- 4,000 psi compressive strength of concrete.
e.·
r
f. g. Step 3.
'
..
" .
•
fy
66,000 psi yield stress of'reinforcing bars.
f u'
90,000 psi ultimate· stress' of reinforcing bars.
Determine dynamic stresses. a.
Dynamic incFease fac cors .
DIF (table '4-1).
Concrete: Diagonal tension
- 1.00
Reinforcement: Bending,yieid: stress
..
~
.
1.17 1'.05 ,,;.\
',Bending, 'ultimate,'. s t're s s . •
b.
•.1
Direct shear yield stress
1.10
Direct. shear ultimate stress
1.00
Dynamic strength of materials.
4A-2l
TH 5d300/NAVFAC P-397/AFR 88-22 Concrete:
Diagonal tension (f'c) - 1.00 (4,000) - 4,000 psi Reinforcement:
Step 4.
Bending (f dv)
1.17 X 66,000
77
Bending (f du)
1.05 X 90,000
94,500 psi
t220
psi
Direct shear (f dv)
1.10 X 66,000 - 72,600 psi
Direct shear (f du)
1. 00 X 90,000
90,000 psi
Dynamic design stresses from table 4-2. Goncrete (f' c): Diagonal tension
- 4,000 psi
Reinforcement (f ds):
Bending
85,860 psi
Direct shear
81,300 psi
Note: Since the structure is symmetrical in both directions, the caloulations will be done only in one direction in steps 5 through 9, 12, 17 through 22, 25 and 26. Step 5.
Determine the ratio of the flexural s t i f'fne ss of the wall to the roof slab. Tw3 H QecH -
21 3
Ts3~
* 3 15 *
240 207
- 3.18
,
1
J.
- 0.76
Q'ecH
l+l/Ct ecH Step 6.
(e q , 4-50)
(eq. 4-62)
1 + 1/3.18
Calculate unit moments using equations 4-52 figure 4A-5 for locations.
4A-22
4·60.
See
TM 5-1300/NAVFAC P-397/AFR 88-22 m+2 - 0.40 (0.63 - 0.28 a'ecH) MoH/(L - H/2) - 0.40(0.63 - 0.28 X 0.76) MoH/(240 - 240/2) - (0.334) MoH/240 m-3 - 3.25 (0.75 - 0.10 a'ecH) MoH/(L - H/2) - 0.25(0.75 - 0.10 X 0.76) MoH/(240 - 240/2) - (0.337) MoH/240 m-4 - 0.25 (0.65)MoH/(L - H/2) - 0.25 X 0.65 MoH/(240 - 240/2) - (0.325)MoH/240 m+5 - 0.40 (0.35)MoH/(L - H/2) - 0.40 X 0.35 MoH/(240 - 240/2) - (0.280)M oH/240 m+6 - 0.60 (0.63 - 0.28 a'ecH) MoH/(H/2) - 0.60 (0.63 - 0.28 X 0.76) MoH/(240/2) - (0.501)MoH/240 m-7 - 0.75 (0.75 - 0.10 a'ecH) MoH/(H/2) - 0.75 (0.75 - 0.10 X 0.76) MoH/(240/2) - (1.011)MoH/240 m-8 - 0.75(0.65)MoH/(H/2) - 0.75(O.65)M oH/(240/2) - (O.975)M oH/240
4A-23
If.
NOTE:
-LH
r- t
=1
OF SYMMETRY
I
,'" L= 240
11
"-
[=240"
f------=--=-..:.-=----+'----'-,....;--~---:,.;... om,
I-
E
II
+= E
2
2
IQ
E
I
m, m+
m+
(\J
I
.
,
I' 0 V
,
I~
6
I~
+~
'-
o
E
nl+2
m+
E
+=
I, I
E
E
IN
E
- ' ' - - . _ ,0
E
I!::::
+~
E
E
o v (\J mi5
II
1-j--- I~ W
:IE ~
-r=
E
m+ 1!:2 E
m+
5
9
i:::; E
.,..~
'-
E
0 IN
E-
1---r-------1~----l--------,--..:.~-----1--
me
)C/)
u,
o
~
PLAN
FIGURE 4A-5
"''4
~~
m-9 - 0.60 (0.35)MoH/(H/2) - 0.60 (0.35)MoH/240/2 - (O.420)M oH/240 Step 7.
a.
b.
Balance the negative unit moment over the column and midstrip. m-3 > m-4
.,
m-3 - m-4 - (0.337) MoH/240
m" 7 > m-8
..
m 7 - ra" 8 - (1.011) MoH/240
Adjust the corresponding positive unit moment in order to keep total panel moments equal. Adjusted m+5 - m+5 - 2(m-3 :.
m-4)
m\ - [0.280 - 2(0.337
Adjusted m+9 - m+9
0.325)]MoH/240 - (0.256) l1oH/240
2(m-7 - m-8)
m+9 - [0.420 - 2 (1.011 - 0.975)]MoH/240 - (0.348)l1oH/240 Step 8.
Calculate total external work for an assumed deflection of ~. Use one quarter of the roof slab due to symmetry in both directions. See figure 4A-6 for yield lines and sectors. Equivalent square column capital, C X C. lTD2 C -[ - 4 ]
IT(45)2
[--:- ]
WI - WIV - r u
[
(
3L
-
2
3 X 240 - r u [(
1/2
2
- 39.9 in.
X2 - X) (X)- + ( - ) 2 2 !J.
X2 - X) (X) - + ( - ) 2 2 !J.
4A-25
say 40 X 40 in.
-;-]
+J
TM 5-1300/NAVFAC P-397/AFR 88-22
[
"
WIn - WVI - r u[
- ru
r
,,]
40 4 1 3*240 X)- + -(----- - X - 40)(240 2 2 2 2
40(240 - -;-
- ru
X) (;-)
1 3L 4 H - C 24 (--- )(-) C)-+ -(-X C) (H : 2 2 2 2 3
]
40 4 1 [(3) (240) ;- (240 - 40)- + - - - - - X 2 2 2
L
From equation 4-74.
VI W -!: W·i. i-I
Step 9.
r u4 --(243200 - 320X) 3
Calculate total internal work for the assumed deflection of 4. Use one quarter of the roof slab due to symmetry in both directions. See figure 4A-6 for angles of rotation.
4
4
8 2H - 8 2H -
H
X
C/2
240
4
8 3H -
8 3V -
4
4
Assume 0 < X < 3H/4
4
(240 - 40)/2
(H - C)/2
220 - X
40/2 - X
100
- 180 in. 3L
X )
2
2
L
X
2
+ m+681H ( - ) + m+281H (L - - ) + 2
2
4A-26
3
X m+281H(-) 2
TIl: 5-l300/NAVFAC
(240) (2) (240) X
MoH
X/3]
P~397/AFR
88"22
+ 0.501 (240)
11
x----
(2) (240)
[<0.501
1. 011) 240
220 - X
+ 2 (0.334 + 0.337) (240 - X)] L
L EI I I
- EvI - m+963H(-) + m+563H(L-X) +
m-~3H(-)
2
+ m-363H(L-X)
2
11
- _ _ x - - [ (0.348 + 1.011) 240 100 (2)(240) + 2 (0.251) + 0.337) (240 - X)] From equation 4-75. E - l:
m 61
VI E - l: E i i-I Step 10.
MoH 11 [ 139972.8 + l33l.76X - 9.7832X 2 + 0.01186X 3 J
- ----------------------
240X (220 - X)
Set the external work equal to the internal work and solve the equation for the ratios of ru/MoH' W- E
(243200-320X) -
MoHI1[139972.8 + l33l.76X 240X (220
3
4A-27
9.7832X2 +O.01186X3] X)
t L=240"
OF SYMMETRY
L/2=120"
FACE OF WALL
x
1:
b
... N
"
. :r
r-i'-,.--;;-_J---_..l L-C
COLUMN CAPITAL
-~
"
~L
(l)
-Z
..
:r
x
(l)
]I[
J[it
lbh~
OF SYMMETRY
LEGEND'
= NEGATIVE
• • • = POSITIVE
PLAN
FIGUR~
4A-6
4A-28
YIELD LINES YIELD LINES
TK
5~1300/NAVFAC
. P-397/AFR88-22
139972.8 + 1331.76 X -9.7832X2 + 0.01186X3
-- - - - - - - - - - - ' - - - - - - - -
.. -
80X (220 - X)(243200 - 320X)
Step 11.
Minimize, value of, ru/MoH using the equation from step 10 by assuming'variou~ values for X to locate the yield line locations. X (Ln)
Step 12.
1
110
ru/MoH (psi/in-lbs) 913.212 X 10- 9
115
911. 686 X 10- 9
120
912':445 X 10- 9
116
911.654 X 10- 9
117
911. 713' X 10- 9
Calculate plastic load mass factor for the flat'slab. 4-30.5 and figure 4A-6.
1
2
( - ) I - ( - ) I V - (-)
eLl
1
eL2
3L X
(-
3
1
(-)11- ( - ) V -
eLl
eLl
'-
2
2 -C(H 3-
2 (3)(240) 1 116 2 X + - ( - ) - -(116)( - 116) + -(--) .. 2 28 2 2 3 2
C
3 X) + -' (H 4
2
C - - X) [3: _ C - X] /2 2
40 (40)(240 -
116)
3
~
2
3
40
+- (240
4 1
See Section
X2
1
2
1
*minimum
116)
[ 3 X 240
2 2
(-)111- (-)VI - - C (H eLl eLl 3
2
. '] C)/2 + - (H 4
"
•~: v:
4A-29
"
.
- 40 - 116) ] /2
-X] /4
TK 5-1300/NAVFAC P-397/AFR 88-22
3
[ 3 X 240
+ - (240 - 40) 4
VI 1:
i-I
I (--)i CLI
-
..
116
-.-
(3L - X) 2
AU - AV - (H -
(3 X 240 - 116)
2 C
- X)
[
2
3L --+ ' 2
40 - 116)
- (240 -
[ 3
2 AI U- AVI - (H - C)
1:
Ai
C - X]
X 240
[ 3L + C
2'
/2
+ 40 - 116] /4
2
- (240 - 40)[ VI
116]./4 '
2
86558.67 in2
X
AI - AI V - -
-' 40,
X]/4
3 X 240 ] 2 + 40 - 116/4
- 128,000.0 in 2
i-I 1:
I (-)i CLI
86558.67
Ai
128,000.0
KU{ 1:
Step 13.
- 0.676
(eq. 3-59)
Calculate the unit mass of the slab .. dc -
15 - 2 - .75 -
,W m - ,- g
(2 X .75)/2 - 11.5 in.
11.5 X 150 X 1000 2 - 2583.5
psi-ms 2 lin
12 X 32.2 X 1728
Determine effective unit mass from equation 3-54. me - KU{ X m - .676 X 2583.5, Step 14.
1746.4
psi-ms 2/in
Find maximum deflection using shortest sector length. 4A-30
TK 5-1300/NAVFAC P-397/AFR 88-22 240 - 40
H - C Ls ~
Step 15.
( --- ) 2
----
- 100.
in
2
- Ls tan 9 maK - 100 X tan 8 - 14.05 in.
Determine impulse load and required resistance for blast. ib -
- (96.0 X 15.0) I 2 -
PT/2
i b2
720.0 psi-ms (eq. 3-93)
r~
2m,.
. 720.0 2 .. r - - - - - - - - - - - 10.56 psi - ravail 2 X 1746.4 X 14.05 Check for correct procedure. 720
ib
tm - - - tm/T Step 16.
- 68.2
r
10.56
-
68.2 I 15
-
Iils
4.55 > 3
O.K. (section 3-20)
Calculate uniform dead load of slab and the required ultimate resistance. Assume 150 psf concrete. rDL -
15/12
X 150/12 2 -
1.30 psi (eq. 4-90)
Required
85,860 r u - 10.56 + ( - - - ) 1.30 - 12.25 psi 66,000
4A-31
TK,5-1300/NAVFAC P-397/AFR Step 17.
8~-22
Calculate required panel moment using the required r u from step 16 and the ru/MoRfrom step 11. 12.25 Required MoR - (ru)/(
Step 18.
- - - - - - - 13,437.,115 in-1bs
)
911.654 X 10- 9
Calculate the minimum required unit moments from step 7. .
..
Minimum unit moment - m+5
. .,:
-
." "., '.:-~.
.,
_."
.
....;'
.
. 256 MoR 240
.256 X 13,437,115
... m+ ,5-
- 14333 in-1bs/in 240
Step 19.
Calculate actual moment capacity at m+5 ' assume No.4 reinforcing bars 12 in. o.c. .5,
de in R direction - dR,:," 15 - 2 - .75 -.'2 X .75 * + 2 X'- - 11.25 in 2
de in L direction - dV - 15 - 2 _.. 75 - 2 X .75*
* Assumed . ~-
~14
.5 2 X - - 10.25 in 2
No. 6 reinforcing in step 2. (eq. 4-19) b
.2 X 85,860 X 11.25 - 16099 in-1b/in > 14333 O.K. 12 .2 X 85,860 X 10.25
M+ 5 -
- 14668 in-1bs/in > 14333 O.K. 12 (eq. 4-13)
p -
.20 - .0015 - .0015 12 X 11.25
4A-32
O.K.
TM 5-l300/NAVFAC P-397/AFR 88-22 .20 -----12 X 10.25
Step 20.
~
.0016 > .0015
O.K.
Calculate the nrOVloeo unit resistance of the flat slab. m (unit moment provided) r u (provided) -
X
m (unit moment required) 16099 13.76 psi
(r u ) m14 - 12.25 X ------
14333
(ru ) m5
-
14668 12.25 X -----14333
(ru)m14 + (
-
12.54 psi
13.76 + 12.54 - 13.15 psi> 12.25 o.k.
2
Step 21.
2
Estimate minimum area of continuous steel in column strip using unit moment ratios from step 7 .
. 348 240
X (0.20) - .27 in2/ft
240 Calculate the average unit tension force in continuous steel. (L -
.20
X
H/2)
240/2 + .27 X (240 - 240/2) 85 ,860 X 2 - 3.362 lbs lin
----------------------------- X
12 X 240 Calculate tension membrane resistance from equation 4-85. rT- --------------------------------------------------------------1 1 { _ (_l)(n-l 4 :E [ 1 n-l,3,5 r
cosh
l
2
41
TM 5-1300/NAVFAC P-397/AFR 88-22 HL - H - C - 240 - 40 - 200 in
"3 1.5 X 14.05 X 3,363 /200 2 rT- - - - - - - - - - - - - - - - - - - - - - - - - - - - - 1
4
1
{_ (_1)(n-l)/2
1:
n-l,3,5
[ 1 -
n3
nn(200) cosh
[ 2(200)
[~3
]J.:i]
3,363
"3 X 1.5 X 14.05 X 3,363 - 23.97 psi> 13.15 psi O.K.
rT -
4 X (200)2 (.6015-.0364+.0080) Step 22.
Calculate diagonal tension stresses at de distance from the edge of wall supports according to section 4-31.2 in both directions.
3L - X r u X Area (Sector I) - VV(
3 X 240 - 116 13.15 (
116
)
2 460670.8 ... Vv - - - - 340.67
)
2
-
2 + - Vv 3
3 X 240
Vv
X 2 116 )
(
2
116 2 + -VV3 2
1352.3 lbs/in
Total diagonal shear load in L direction.
VuV - Vv
(
3L - X ) + 2
- 1352.3
[,
X Vv (- - dv ) 3 2
2
X 240 - 116 2
2 + - X [ 3
1~6
Diagonal shear stress in L direction.
4A-34
- 10.25J] - '51.'" 10.
TH 5-1300jNAVFAC P-397/AFR 88-22 v uV vuV
451,443 125.9 psi 3 X 240
3L ( - - d) dV 2
(---
10.25)
10.25
2
Estimate reinforcing ratio at support using the ratio of unit moments.
MoH
.494 m1 (-) m5
r m1 - r m5
240
- .0016
(
- .0031
)
MoH
.256
240 Calculate diagonal shear capacity of concrete. V
c - 1.9 (f'c)1/2 +
2500 p
(eq. 4-23)
v cV - 1.9 (4000)1/2 + 2500 (0.0031)
127.9 psi> 125.9 psi
:. No stirrups or wall drop panel required. Note: Diagonal shear at de distance from the H direction wall will be less than the one in L direction due to symmetry and larger de in H direction. Calculation is not required. Step 23.
Check punching shear around column capital. Use average de' d + d davg - - - - -
10.25 + 11.25 - - - - - - - 10.75 in
2
2
Dp - D + dav g -
Diameter of punching.
45 + 10.75 - 55.75 in.
Find area between positive yield lines minus column capacity.
xf --?TD 2
Area
=
[
:L -
3 X 240
P
4
[
- 57095 in 2
4A-35
- 116 2
n X 55.75 2
]2 -
4
TN
5-1300/NAVFACP~3!17/AFR
88-22 r u X Area
v (punching)
13.15 X 57095 - 398.8 psi
"Dp dav g V
"X 55.75 X 10.75 253.0 psi < 398.8 psi
c - 4 (f'c)1/2 - 4 (4000)1/2 -
:. Need drop panel, assume 6 in. d av g (revised) - 10.75 + 6 - 16.75 in.
Dp (revised) - 45 + 16.75 - 61.75 in. Area (revised) - [
3 X 240
- 116
2
12
- 5641 in 2 4
"
13.15 X 56541
V (punching) -
-
228.8 psi < 253.0
O.K.
"X 61.75 X 16.75 Step 24.
Assume 63 X 63 in. drop panel.
Check punching shear.
Punching Length - I p - 1 + d av g - 63 + 10.75 - 73.75 in 3L Area - [-2- -
v (punching)
12
XJ
1 2_ P
240
X
[
2
1161 - 73.7522 - 54097 in 2
J
2 13.15 X 54097
r u Area
- 224.3 psi < 253.0 psi
4 I p davg Step 25.
3
4 X 73.75 X 10.75
Check one-way diagonal shear de distance away from column drop panel and between positive yield lines. 3L Width in L direction - --2
240
3 X
- X-
- 116
2 1
Area in L direction
-
-
Width (H - X 2
4A-36
~)
244 in.
TIl:
5-1300/NAVFAC P-397/AFR 88-22
63
- 244 (240 - 116 -
- 10.75) - 19947 in 2
2
13 .15 X 19947
v-
O.K.
d" Width
10.75 X 244
Note: Diagonal shear in H direction will be less than the one in L direction due to symmetry and larger de in H direction. Calculation is not required. Step 26.
Check one-way diagonal shear at an average d distance away from column capital and between positive yield lines. (Width)L - 244 in
(step 25)
1
d" + ( - - ) X drop panel depth width 63
d avg - 10.25 + --- X 6 - 11.80 in 244 C
Area in L direction - Width (H - X 2
40 - 244 (240 - 116 -
- 11.80) - 22496.8 in 2
2
v -
d avg Width
13 .15 X 22496.8 - - - - - - - - 102.7 psi < "c 11.80 X 244
O.K.
Note: Diagonal shear in H direction will be less than the one in L direction due to symmetry and larger de in H direction. Calculation is not required. Step 27.
Calculate all remaining required moments similar to step 18. Assume reinforcing bars for each, and determine actual provided unit moment capacities similar to step 19.
4A-37
TH 5-l300/NAVFAC P-397/AFR 88-22 Problem 4A-3. Elements Designed for Impulse-Large Deflections Problem:
Design an element subjected to an impulse load for a large deflection.
Procedure: Step 1.
Establish design parameters: a.
Impulse load and duration (Chapter 2).
b.
Deflection criteria.
c.
Geometry of element.
d.
Support conditions.
e.
Type of section available to resist blast. type II or III depending upon the occurrence of spalling.
f.
Materials to be used and corresponding strengths.
g.
Dynamic increase factors (table 4-1).
static
design
Step 2.
Determine dynamic yield strength and dynamic ultimate strength of reinforcement.
Step 3.
Determine dynamic design stress for the reinforcement according to the deflection range (support rotation) required by the desired protection level (table 4-2).
Step 4.
Determine optimum distribution of the reinforcement according to the deflection range considered (sect. 4-33.4 and figs. 4-37 and 4-38). Step not necessary for one-way elements.
Step 5.
Establish design equation for deflection range considered and type of section (type II or III) available.
Step 6.
Determine impulse coefficient Cl and/or Cu for optimum PV/PH ratio and L/H ratio. Note: If the desired deflection is not equal to Xl or ~. determine yield line location (figs. 3-4 through 3-20) for optimum PV/PH ratio and L/H ratio and calculate Xl' and, if necessary, ~ (table 3-5 or 3-6).
Xm
Xm.
4A-38
TN 5-1300/NAVFAC P-397/AFR 88-22 Step 7.
Substitute known parameters into equation of step 5 to obtain relationship between PH and d c'
Step 8.
Assume value of dc and calculate PH and from optimum PV/PH calculate pv. Select bar sizes and spacings necessary to furnish required reinforcement (see Sect. for limitations).
Step 9.
For actual distribution of flexural reinforcement, establish yield line location (figs. 3-4 through 3-20). Note: . ' ; [MVN + MVp ]
H
1/2
MHN + MHP
L
[~f/2_
H
PH
L
H
[ As V ] 1/2 AsH
since: MN - Mp - P dc 2 f ds -
As b
dc f ds
Step 10.
Determine the ultimate shear stress at distance dc from the support in both the vertical (vuV from eq. 4-119) and horizontal (vuH from eq. 4-118) directions where the coefficients Cv and CH are determined from figures 4-39 through 4-52 (see sect. 4-35.2 for an explanation of the figures and parameters involved).
Step 11.
Determine the shear capacity V c of the concrete in both the vertical and horizontal directions (use eq. 4-21).
Step 12.
Select lacing method to be used (fig. 4-91). (Note: Lacing making an angle of 45· with longitudinal reinforcement is most efficient. )
Step 13.
Determine the required lacing bar sizes for both the vertical and directions from equation 4-26 where the parameters b l and sl are determined from the lacing method used (fig. 4-91), and the angle of inclination of the lacing bars a is obtained from figure 4-15. The lacing bar size db must be assumed in order to compute d l and Rl .
horiz~ntal
(Note: See sect. 4-18.3 for limitations imposed upon the design of the lacing). Step 14.
Determine required thickness Tc for assumed, dc, selected flexural and lacing bar sizes, and required concrete cover. Adjust Tc to the nearest whole inch and calculate the actual d c'
4A-39
TIl: 5-1300/NAVFAC P-397/AFR 88
Step 15.
022
Check flexural capacity based on either impulse or deflection. Generally, lacing bar sizes do not have to be checked since they are not usually affected a small change in dc' a.
Check of impulse. Compute actual impulse capacity of the element using the equation determined in step 5 and compare with anticipated blast load, Repeat design (from step 8 on) if capacity is less than required.
b.
Check of deflection. Compute actual maximum deflection of the element using equation determined in step 5 and compare with deflection permitted by design criteria. Repeat design (from step 8 on) if actual deflection is greater than that permitted.
Step 16.
Determine whether correct procedure has been used by first computing the response time of the element ~ (time to reach maximum deflection) from equation 3-95 or 3·96, depending on the deflection range considered in the des • and then compare response time t m with duration of load to' For elements to be designed for impulse, t m > 3 to'
Step 17.
Determine the ultimate support shear in both the vertical (VsV from eq. 4-122) and horizontal (Vs H from eq. 4-121) directions. The coefficients Cs v and Gs H are determined from table 4-15 and figures 4-53 through 4·56 (see sect. 4-35.3 for an explanation of the figures and parameters involved).
Step 18.
Determine the required diagonal bar sizes for the vertical and horizontal (intersecting elements may control) directions from equation 4-30. Diagonal bars should have the same spacing as the flexural reinforcement (fig. 4-95). Note: To obtain the most economical repeat the above steps for several wall thicknesses and compare costs. Percentl1j>es of reinforcement may be used to reduce the amount of calculations. In determining the required quantities of reinforcement, lapping of the bars should be considered. Example 4A-3, Elements Designed for Impulse-Large Deflections
Required:
Design the back wall of the interior cell (fig. 4A-7) of a multicubicle structure for incipient failure.
4A-40
TH 5-l300/NAVFAC P"397/AFR'SS-22
28'·0·
I'
...
I'
.', r-,'
..
:
]
-0 .' 2
PLAN Figure 4A-7 Solution: Step 1. a.
Given: i b - 3,200 psi-ms and to - 5 ms
...., "
,~ .
,
b.
Incipient failure
c.
L - 336 in, H - 120 in
d.
Fixed on three edges and one edge free
e.
Type III cross section
f.
Reinforcement f y - 66,000 psi-and f u - 90,000 psi
"
,"
concrete f'c - 4,000 psi g.
For reinforcement DIF - 1.23 for dynamic yield stress ,
,
DIF - 1.05'for ultimate dynamic stress f Step 2.
Dynamic Strength of Materials -181,l~0
1.23 x 66,000
psi
1.05 x 90,000 - 94,500 psi Step 3.
"f,
Dynamic Design Stress, from table 4-2, (81,180 + 94,500) ---------------, - 87,840 psi 2
4A-4~
TK 5-1300/NAVFAC P-397/AFR 88-22'
,Step 4.
L 336 From figure 4-38 for --- - --- - 2.8 and 3 edges fixed, H 120
Optimum: Step 5.
'Since
Xm -
Pv/PH - 1.41
Xu
(incipient failure):
i b2 H - - - : : - - - - Cu 3f PH d c ds
Step 6.
(eq. 4-103)'
L/H - 2.8 is not plotted on figure 4-34, ,therefore must interpolate for optimum Pv/PH'
For Pv/PH - 1.41 ' L/H
Cu
1.5 2.0 3.0 ,4.0
613.0 544,0 444.0 387.0 ,
"
From figure 4A-8, Cu - 461.0. Step 7.
Cu f ds
(3,200)2 (120) - 30.3 (461.0) Step 8.
(87,~40)
Assume de - 21 in: '.
30.3 PH -, --3-
30.3 -
0.00327
de Pv -
(PH) - 1.41 (0.00327) ~ 0.00461
.
4A-42 '
"
TM 5-1300/NAVFAC P-397/AFR 88-22
Step 9.
Yield line location.
Actual: 1.20 , 2 x 0.86'
'.
~.
0.698
"1/2 From figure 3 ell,for. . ' ; [ ASVJ , : : 2.8 (0.698)1/2 ,. 2.34 HAsH' and ~/M"N2
1 . 00 ,
x/L - 0.385,' x;' 0.385(336)' ';'; 129 in
."12'"' . - ;'"
•• 12'"
t
...
&~.
o !l!
'":z:
I.
L • 558"
'Figure 4A-9 Step 10.
Ultimate shear stress at distance d c from support. a.
Vertical Direction (along L): For: -- - -----H 120
- 0,175 and x/L - 0.385
4A-44
TH 5-1300/NAVFAC P-397/AFR ,8,8"22' ,
From figure 4-45: de
(
)K -
0.49
eK -
and
H
1.07
For: 0.175 0.357
From figure 4-46:
Cv
--
-
0.58
Therefore:
'1
A~V
Pv - ---- -
,1.20
'"
b de
'" J. 0.00476
, .
Cv Pv
~
(eq , 4-119)
f ds
- 0.62 (0.00476) (87,840)
,J"
- 259 psi
b.
Horizontal· Direction (along H) , For: -----, -
x
> •••.••
12 (21)
so that:
VuV ,-
-
... ;
21 -----129
-
0.163
·'0.00341
4A-45
TN 5-1300/NAVFAC P-397/AFR'88-22 : i
}'.
so that: VuH -
CH PH f ds
(eq. 4-118)
- 0,81 (0.00341) (87,840) - 243 psi Step 11,
:;;.
Shear Capacity of Concrete a.
(eq. 4-23)
Vertical Direction Vc
- (1.9 (f' c)l/f .+2.500i'v)
"
•
- [1.9 (4,OOO)i/2 + 2,500 (0.00476)J ;~~
129 psi
b.
Horizontal Direction Vc
- (1.9 (f'c)1/2 + 2,590
.
~'.
PH) \
- [1.9 (4,000)1/2,+ ,2;500 (0.00341)] - 132 psi
Step 12,
Use lacing method No, 3 (see fig, 4-91).
Step 13,
Lacing bar sizes: a.
.~
.'
.\
'
Vertical Lacing:Bars b 1 - 10 in
'h -
22 in"
:
Assume No. 6 Bars, "
(eq. 4-28)
4A-46
TK 5-1300/NAVFAC P-397/AFR 88-22
9 (0.75) .;. 0.271
(eq .
4-29)
24.88
..
~".
lHORIL)
Figure 4A-10 From figure 4-15:
a - 53.0' For shear: f dy - 1.10 x 66,000 - 72;600 psi f du - 1.00 x 90,000 - 90,000 psi'
fds - (72,600 + 90,000)/2 - 81,300 psi' (VuV - v c) hI sl
A.v - - - - - - - ~
fds (sin a + cos a)
(259 - 132) (10) (22) 0.85 (81,300)(0.799·+'0.602) - 0.289 in2 Min
A.v -
0.0015 hI
,,
si
- 0.0015 '(10) (22) - 0.330 in 2 '
4A-47
(eq. 4-26)
TH
5~1300/NAVFAC
,
,
P-397/AFRl!8-22;
Use No. 6 bars: (A
b.
-"0.'44
in 2 ) ,'
Horizontal Lacing:
. b l - 11 in
sl - 20 in
.'
,'" , '.
Figure 4A-11 '~
"
Assume No. 6 Bars: db - 0.75 in:'
-v ,
.'
d 1 - 21.0 + 1.13 + 0.75 - 22.88 in Min. R1 - 4 db For:
-i ----
20
-
.. ,-.0.874
d1
(eq. 4-28)
22.88 i
2 RI + db d1
,9 db ---d1
'
.
9;,(0. 75) -
-
'.\ 0.295
."
..
a - 53.5'
(eq. 4-29)
22.88.
From figure 4-15: ,"
I:,
,
• rv
I •• '
t
f ds (sin a + cos a)
•
,.'.
"
(243 - 129) (11) ,(20),' 0.85 (81,300) (0.8P4>+:0.595), - 0.259 in2
4A-48._ }
(eq. 4-26)
. [I .
,',
TK Min.
Av -
5-l300/NAVFAC,P-3~7/AFR
88-22·
0.0015. b l sl
- 0.0015 (11) (20) - 0.330 in 2 Use No. 6 bars: ',.....,.....
Step 14.
"
Actual d c depends upon vertical·lacing. Cover 2 x 0.75 - 2 x 0.75 Lacing Horizontal 2 x 1. 00 Vertical 1.13
.-
.1.50 1.50 2.00 1.13
:',
"
6.13 in
.r" -
dc
+ 6.13
- 27.13 in Actual d c. - 27 Check capacity. a.
Use 27 in
6.13 - 20.87 in
.
,
Step 15.
..
21 + 6.13
Actual impulse capacity. For:
Pv
-.1.~0 ~
1.41,
Cu - 461.0
-, (fig. 4-34)
For:
0.86· d c - 20.87 in, ,
• I
~
PH 3 f
d PH.c, .\
ds
'12(20.87) C
u
- 0.00343
'
, ...
H
0.00343(20.87)3 (87,8~0) (461.0) (120) , -o
4A-49
TN 5-l300/NAVFAC P-397/AFR 88-22 i b,- 3244 psi-ms> i b - 3,200 psi-ms b.
Actual maximum deflection. For:
Pv -- - 1.40, Cl - 452.0 PH
(fig. 4-31)
Note: Interpolation for C1 not shown. Cu - 461.0
From table 3-6 for x' Xl - H tan 12· ~
-
:> H: 120 (0.2125) - 25.5 in .~
x ,tan 9 max + (. 2
_ (129) tan 12 + (168-129) tan[12'-tan-'1[ tan 12 ] ] . . 129/120 - 27.42 + 0.56 - 27.98 in. From Step 5:
(3200)2 (120) "
0.00343(20.87)3(87.840) ~ - 25.5
452.0 + (461.0 - 452.0) . From 'which:
~ - 24.58 in'
[
27.98
]
25.5 .
l
Note: Since the deflection ~ is less than Xl' the above solution (~ - 24.58) is incorrect because the equation used is for
4A-50 .
TN 5-l300/NAVFAC
P~397/AFR
88-22
xm
the deflection ra~ge XI'S S~. Therefore, an equation for the deflection range 0 S Xm S-Xl must be used to obtain the correct solution. From Section 4-33.5 for Type I I I cross sections and valid for deflection range: "
i b2 H
Cl (
3 PH d c f ds
Xm-
Xm
)
Xl
i b 2 H Xl PH d,c
3 f ds Cl
"
(3200)2 (120) (25.5) (0.00343) (20.87)3 (87,840) 452.0,
Xm -
25.3 in
Note_:The element is slightly over-designed. To obtain a-more economical design, the amount of flexural reinforcemen~ may be reduced. Step 16.
The response time of the element is obtained 'from:
,
~---
'. -
(eq. 3-96)
where: (table 3-2)
ru -
but:
.. 4A-5l
TK 5-1300jNAVFAC P-397/AFR 88-22 '0.86 (87,840) (20,,87)'-"
..
12 131,380 in-1bs/fn therefore: 5 (2) (131,380) ru -
- 78.9 psi
(129)2
so that: ib
tu - -
ru
-
tm
3,200
-
78.9
40.6 ms-
"
40.6
"
;,
The correct procedure has been used since:
Step 17.
Ultimate s~pport shear. ,,"
a.
,.
-
Vertical Direction (along L): From figure 4-54, for:
x/L - 0.385 CSV - 4.40 (eq. 4-122) H
4.40 (1.20) (20.87)2 (87,840) 12 (20.87). (120) - 6720 1bs/in b.
Horizontal Direction (along H): From table 4-15: 6
CsH
- --(x/L)
6
- 15.6 0.385 4A-52·.
'
TM S-i300/NAVFAC P-397/AFR 88-22 '
Vs H - - - - - - - - -
(eq. 4-121) . },.,.
<_.. l
L
15.6 (0.00343) (20:87)2 (87,840)
336 6,090 lbs/in i.
Step
is.
,"
Diagonal bar sizes. Note: Place' bars
on
sin
a 45- "angl'...
Q
-
0.707 'r ' '.
Vertical'Direction (at floor'siab}:
a.
,Ad - - - - ' - -
f ds sin
Q
.
. Required area of bar: . " ,
0.58 in2 2 -'..-.\
Use No.8 @ 10. Horizontal Direction' (at wall, inters"ctibns) ': u
' . " -. ;
.'
6;090 (11)' Ad - - - - -
f ds sin
. ",
Q
.
, '81,300 ~0.707) "
.
..
j'. '.'
Required area of bar:
. :J:.:.:,
- 0.58 in 2 Use Nil. 8 @
'
2 .,- L16 in
n. .......
4A-S3
-"
(eq. 4-30)
TH
5"~300/NAVFAC
P-397/AFR 88-22
Problem 4A-4. Elements Designed .for Impulse-Limited Deflections Problem:'
Design an element which responds, to the impulse loading of a close-in detonation. •
Procedure: Step 1.
Establish design
I"
-'
.. ~-'
parame~ers:\ '0
a.
Blast loads including pressure-time relationship (Chapter 2).
b.
Deflection criteria. ' .
c.
Structural configuration including tions.
d.
Type of cross'section available depe~ding upon the occurrence of spa11ing and/or crushing of;the concrete cove~ .. _,'
Step 2,
geome~ry.
and
. ....
"\--..
s~pport
condi-
Select cross section of element including thickness. and concrete cover over the reinforc~~e~t. Also determine. the static design stresses of concrete and reinforcing' steeL (Sectio'n, 4-12),. •
•
'"
•
"
"1
·r··
i,'
Step 3,
Determine dynamic increase factors for both.concrete and.reinforcement from table 4-1. Using the above DIF's and the static design stresses of step 2, calculate the dynamic s~rength of materials.
Step 4.
Determine the dynamic design stresses using. table 4-2 and the results from step 3. f'\J ~;-:~ }', . s,.. "',
Step 5.
Assume yertica~ and horizontal.rl"ipf~~cemen~.,bi'.rs.co yieM the optimum s t ee l, ratio. The steel, ratio is optimum whEm the'resu1ting yield lines make an, angle· of 45 degrees with the supports. . ',' -. ,. \. , . -. ..' . .. Calculate de (dor de ;depending.upon th..,. type, of cross section available to resist the blast load) for both ',the"positive and negative moments in both the vertical. and' horizonta1..directions. Determine the reinforcing, ratios. A1so'check 'fot· the'~inimum stee1'ratios from table 4'3. -;'
~_.
Step 7.
--~
'~.
~
~
\'~
Using the area of reinforcement' and th~ :ir~l~e' of de -from "step 6, and the dynamic desi'gn, stress of step 4,'calcu1ate the moment capacity (Sect. 4-17) of both the positive and negatiye ..r"inforcement. Also calculate the Pv/PH ratio and compare to ,the optimum steel ratio from step 5. -. .
Step 8.,
Establish values of KE , XE and r u similar to .the procedures of problem 4A-1, steps 8 to 18.
Step 9.
Determine the load-mass factor KLM, for elastic, e1asto-p1astic and plastic ranges from table 3-13 and figure 3-44. The average load mass factor is obtained by taking the average KLM for the elastic 4A-54 . "
TK S-1300fNAVFAC'P-397/AFR 88-22 and elasto-plastic ranges,and averaging' this valu~ with.the,KIM of the plastic range. In' addition, calculate the unit mass of the element (account for reduced'concrete thickness,if spalling is anticipated) and multiply ,this unit mass 'by KIM for the element, to obtain the effective unit mass of the element. "';
Step 10.
Using the effective mass ~f step 9 and the equivalent stiffness of step 8, calculate the natural period of vibration TN from equation 3-60.
Step 11.
Determine the response chart parameters: a.
Peak pressure P (step 1).
b.
Peak resistance r u (step 8). .
c.
Duration of load T (step 1).
d.
Natural period of vibration'TN (step .10).
..
Also,calculate the ratios of peak'pressure p,to peak resistance r u and duration T to period of vibration TN' Using these ratios and the, response charts of Chapter 3, determine the value of'~/XE' ,co;"pute th;' value of x",. Step 12.
Determine the support rotation corresponding to the value of x", from step 11 using the equations of table 3-6. Compare this value to maximum permissible support rotation of step 1, and if found to be satisfactory, proceed to step 13. If comparison is unsatisfactory, repeat steps 2 to 12.
Step 13.
Using'the ultimate resistance of step 8, the values of de of step 6 and the equations of table 4-6 or 4-7 (table 3-10 or 3-11 if shear at support is required),' calculate the ultimate diagonal tension shear stress at a distance d~ :from each support (or at each support). Also, calculate the shear capacity of the concrete -from equation '4-23: If, the capacity of the concrete"is greater than that produced by the load, minimum shear.,reinforcemimt must be used. However, if the shear produced by the load is greater than the capacity of the concrete, then shear, reinforcement in excess of the minimum required must be provided. Also yheck for maximum spacing of shear reinforcement.
Sten 14.
Using the equations of table 3-9.3-l0'or 3-11 and the ultimate resistance of step 8, calculate the shear at the supports. Determine the required area'of diagonal bars ~sing equation 4-30, However, if section type I is. used, then the minimum diagonal bars ,. must' be provided. . ' ,_
..
4A-SS'
TH 5-1300/NAVFACP-397/AFR 88~22
Example 4A-4, Elements Designed for "Impulse-Limited Deflections
..
":,
"t,;
,.j
Required:
-.
"
Design the side, wall of cubicle',with,lno roof or front wall and subject to the effects of:a detonation of ,an explosive.within the' cubicle. ~t~
"
Solution: Step I:
:
t;
,,)
".'
,Given: a.
Pressure-time loading (fig. 4A-12).
b.
Maximum support rotation equal to 2 degrees.
c.
L - 180 in., H - 144 in. and fixed on two sides (fig. 4A12) . I '
d.
Type Ll Lr c r os s section . • r-
[
" ,
...
.. 1485
•
,
,',
.
,.
-
PLAN
SECTION .. ' -
;
,
.
Step 2.
.
Sele~t,element
..... ,
..l.
r
.'
Figure
I,
~l'
~]
.Eo -
BLAST LOAD -:'
'.'
.. ---
1,2 TIME,ml
,
'15'·O~
T.'
....
,
.
,
r
~
.
4A-12.1~<
I'~
thickness anq,static'stress of, reinforcement and concrete· (fig 4A-13). .'
';
I
"rIAnY! MINt-
.. "
: ~ :.:' ;:.: :': o' •
, .'
1~'cL
,
,
.. :~POSITIVE
,• • • , ,
.' .', .. .. "
IlCTrlUOlt SUR'ACr
, .'
•
,
.;
IItIN. ,
"
.
INTrRIOR SURFACt:
f'c - 4,OQO psi.,
, ~;CL
.
::...
-'.'
'. :fy - 66,OOO,psi
. •
".: T
~
•
I
~,~
r
-;»
.J •
',"
Assume'T c - 22 .Ln.. and-conc re t e cover is as shown in figure 4A-13,
':<.:'., ;(::
Figure 4A-13 4A-56' .'
TH S:'1300/NAVFACP-397/AFR 88-22
,. Dynamic increase "factors, DIF (from table 4-f) .
Determine dynamic stresses.
Step 3.
a.
-v-.
Concrete: " Diagonal Tension . t,
-l.QO .··r·
.1'
..
Reinforcement:
""
-
Bending
.
Diagonal Tension
L10'
Direct Shear' •
1.10
''''!'
b,
1.23
'
H,
r
,.
.' ~tf"
'
Dynamic strength of materials.
. ,_.
. ',..~ . i· .
'.l.~
j
••
, ; -'
,
:~
:' 4,000 psi . <, .
")'1 .<
,
-
•
~
1.00 ( 4,000) r, '\ ;. .
Reinforcement (f dy): '\
,
Bending
=''''
..,
,.
'
.,
_~ ~,'l,
>
,Concrete (f'dc): Diagonal Tension
,-~
./'..
,":",
. . ..
y
psi - 81,180 ,. 1,10'(66,000) - 72',600 psi ,; ...c.... 1.10 (66,000) - 72,600 psi 1.23 (66,000)
.
Dlagonill Tension' Direct Shear
'
~ ;,:
,,;:;
'0.
"
Step·4.
:-: "" Dynamic design stress from table 4-2:'
Concrete '(f' de): . ri:i ,
',_e'
'.
L
.
~
,
.
'1
-
Diagonal Tension Reinforcement (fds
.~;
L "
,,-
.• -.1'/
4,000 psi
- f dy for 0 < 2) -
Bending Di'agonal Tension Direct Shear
81,180 psi • 72,600 psi 72,600 psi
fl.
Step 5.
Determine tlle optimlim steel ratio P;'/PH' degree yield lines, 'x
'H
,
144
- - - L L 180
0.80
From figure 3-4,
4A-57
Set x - H to obtain 45
TK 5-l300/NAVFAC,
P~397/AFR
88-22 "
:~.
"'~-'- ',
"
LOS
)
Therefore,
or ;
..
..,.
Try No.7 bars atS,in. o .c , .Ln the vertical direction and No.6 bars at Sin. e ;e , in the ' horizontal direction.' Step 6.
1
<.
:
,'j
~ t:
•
•
Ca1culate'dc and the steel ratios, for each direction. "
Assume No. 3 stirrups.
.
"
,.';.
.",
...
"
d cV' - 22- (2 X 0.375) " 0.75 - ,1.5-, (2.XO:875/2) - 18.125 in.
:
d cH- lS.125
(2 X 0.875/2) .' ,
)'.
"
(2 X 0.75/2) ..'16.50 in. -,
0.60
"
"
......
0.44 0.0033 > 0.0015 minimum' S X 16.50, ' Step 7.
"
~'."
.
"
.
Calculate the moment' capacity of both the posJtive':and negative' reinforcement in both directions (eq. 4-19).' . '!'
M,,-
--~
"
b
, _
W-
M
.~,
. '.'
"
0.60'isi,180)(li.125} MVp - --:~"--"---:-,"-~~~' ';",,110,354 iri-lbs/in '8
0.44 (Sl,lS0)(16.50)
,
8, .' MVN/MHN'- 110,354/73,,671 :::: 1.5
-:1.5 d' ..,
4A- 5 t
·
from step 5 o.k.
'
TM 5c1300/NAVFAC P-397/AFR.88-22 Step 8.
Using the procedure inc'exampie' 4A~1."steps 8 through 18 and the moment capacities from step 7. establish the .va1ues of KE, XE and ,. ": r u' -'...
KE - 36.7 psi/in .. "o'
E-
:: .. ,. ,PX
~
C'
..
,
< "
~
.
0.968 in.
r u - 35 ..53 psi' Step 9.
Calculate the effective' mass of the·e1ement.·a.
Load mass factors (table 3-·13 and'·fig. 3-44) x/L - 0.80.' elastic. range'
.
"
,
"
'.'
e1asto-ptastic range:
" KUi -
onesimp1e.edge·
0.66
" . Ktit ~ 0 -, 66'
: two simple. edge s plastic range
"'"
... ..
,
r,,'~
.
. ', '
KU{ (average elastic and
e1asto~p1astic
values)
0.65 + 0.66 + 0.66 - .' 0.66
.
.:'
3
KU{ (ayerage elastic. and· p1astic.va1ues)·'· 0.66 + '0.54 0.60 -c
2
b.
Unit mass of element: Using the1arge·r. d c as the. thickness. of the element.
Due to spa11ing (Type III cross section) available thickness equals Tc - d c - 18.125 in. wcdc 150 (18.125)' 10 6 m---g 32.2 X 12 X 1728
4A-59
4,072 psi-ms 2/1n
TKS-i300/NAVFAC P-397/AFR 88-22
1"/
'c ...)
Effective unit mass of element:
"
.: ",
me - KLMm - 0.60 (4,072) - 2,443 psi-ms 2/in
.
Step 10.
.
Calculate the natural period of vibration.
TN - 2 " (meIKE)1/2
(eq. 3-60)
TN ~ 2 (3.14) (2443/36.7) l/{ - 51. 2 ms , Step 11.
Determine maximum response pf element.
a.
Response chart parameters:.'
,
.
"
Peak pressure, P - 1485' psi (step 1). Peak resistance, r u - 35.53 psi (step 8):
"
Duration of blast load, T ~"J1. 2 ms .(step 1) Period of vibration TN - 51.2
m~,(step6)
P/ru - l485/35.5~_- 41.8,. " . T/TN - 1. 2/51. 2."~ 0.023
b.
. From figure 3-64a:
.,
. . '.
Xm/XE - 5.0 . Xm - 5.0 X 0.968 - 4.841n Step 12.
Check support rotation (table 3-6). c Since x - H - 144 in. and 0 < Xm < Xl Xm ..; x tan e.t tane.t - 4.84/144 e.t-. ,1. 93° < 2°
assumed section is O.K. ..
-. . ."
4A-60.
"
TH 5-l300/NAVFAC P-397/AFR 88-22" Step 13.
Check diagonal,tension at supports (internal loading). a.
Calculate ultimate shear stresses at support by dividing the values of the support shear from table 3-10 by their respec. tive dc'-" ". -t ' VsH - 3rtix/5 - 3 X·35.53." X' 144/5 - 3,070 1b/in X
3ru H (2
-)
L VSV - - - - - - 'X
(6 -
3 (35.53) 144 (2 0.80) - - - - - - . - - - - - 3,542 lb/in '(6'-, 0:80)
)
L,
vuH - VsH/dcH vuV - Vsv/dcV b.
~
3,070/16,50
- ·186 :1, psi
3,542/18.125 - 195.4 psi
Allowable shear stresses (eq. 4-23) Vc - 1.9 (f'c)1/2 + 2500 p 53.5 (f'c)1/2 - 221.4 psi ."
.
.J.
~.'
where p is the steel ratio at the support; vcH .. 1.9 (4000)1/2 + 2500 (0.0033) - 128.4 psi < 221.4 psi vcV - 1.9 (4000)1/2 + 2500 (0.0041) - 130.4 psi < 221.4 .ps I
c.
Required area of single 'leg stirrups.' (Vu - Vc)bs
(eq. 4-26)
0.85 (f' ds) vuH - vcH - 186.1 - 128.4 - 57.7 < 0.85vcH - 108.8 psi Use 0.85 vcH as minimum .
.,
4A-6l
TN 5-l300/NAVFAC P-397/AFR88-22 ,vuV - vcV - 195;4 • 130.4 - 65'.0 <·0.85vcV... - 110:8 ,psi minim~.
. Use 0.85 vcV as
" ,
,
Tie every reinforcing bar intersection, therefore, b - s - 8 in. < dc/2 (maximum spacing), O.K.' 108.8 X 8 X 8
AvH -
0.85
-
X~
0.11 in 2.'
72,600 '
minimum Av~. 0.0015 bs - 0.10 in2 < 0.11 O.K. 110.8 X 8X 8
Avv-----0.85 X 72,600
minimum
Av -
0.0015 bs - 0.10 in 2 < 0.11 O.K.
The area of No. 3 bar is 0.11 in2 ;so O.K. Step 14.
~a~ ~ssUm~d
in step 6 is
Determine ~equired area of diagonal bars using the values of the shear ,at the ,support from step 13. ' (eq. 4-30)
,-
f ds sin a
Assume diagonal bars are inclined at 45 degrees. 3.070 X 8
AdM
- 72,600
X
0.707
3,542 X 8 AdV -
- 0.48 in2 at8 in. o .c.
\
'
- 0.55 in 2 at 8 in. o.c.
72,600 X 0.707
Use No. 7 bars (Ab, - 0.60 in 2) .ae 8 in. o.c. at both supports
TM 5-1300/NAVFAC
~~397/AFR 88~22
Problem4A-5 .. Elements Designed for Impulse-Composite Construction • . v . · . . Problem: Procedure: Step 1.
Step 2.
Design a compo~it~ (concrete-sand-concrete) wall to resist a given blast output for inci.plent failure. Establish design parameters. a.
Stru~ture
b.
Charge weight.
c.
Blast, impulse load (Chapter 2).
d.
Thicknesses of concrete and sand portions of wall.
e.
Blast impulse resisted by concrete panels .
f.
Density of concrete and sand.
configuration.
",
Determine scaled thicknesses of concrete and sand using:
. Tc - Tc JWl/3 and Ts - Ts JWl/3 . Step 3.
Determine: scaled blast impulse resisted by each concrete panel using: -i
1/3 (donor panel) bd - ibdJW
I ba Step 4.
-'i baJW l/ 3 (acceptor panel)
. l •.
Correct scaled blast impulse resisted by concrete (Step. 3.) to account for the 'in~reased mass produced by the sand and the reduction of the concrete mass produced by spalling and scabbing of the concrete panels using: Tc + d c (ccrr , )
i bd -
i bd
[
2
.n
Tc + d c (Corr. )
,i ba -
i ba
Ts W s + (-) (-) Wc 2
[ 4A-63
2
dc
Ts W s + (-.-) ( - . ) 2' W c dc
r r
TK 5-l300/NAVFAC P-397/AFR 88-22
Step 5.
Step 6.
Determine scaled blast impulse attenuated By acceptor 'panel and the sand i a from figure pcf., respectively.
4-5~
Calculate total impulse
resi~ted
or
4-58"f~~
ws,
~qua~
~5
'and 100
by the wall using:
•
i b t - i a + i bd Step 7.
to
Compare blast impulse which is re~istedby wall to that of the applied blast loads.
,
'
Example 4A-5. Elements Designed for i:mpuls~-CompoSite":Construction Required:
Design the composite wall shown below for incipient failure conditions. '
,.
....
,"
·· '·. ,.
'. ' •• ~~~ ..~:'.'. ' ': ~~a~ .... :z:
.... ~~~:r;;;
'" . ~r,~~ :. i::~"::~
'. :,-"::.,. .;-: ,:-.:,,~~
,
"
-v.
." " . :a:~: .... .. . ..' 0"
I.
L
ELEVATION
SECTION
; . . 1\
Figure 4A-14 Step 1.
Step 2.
a.
Structural configuration as shown ~in figure 4K-14'.
b.
w
c.
'ib .. 4,800 psi-ms (Chapter 2).
- 1,000 Lbs ,
d.
1 ft, T; - 2 ft, and d c - 0,833 ft.
e.
i b a - 1,500 psi-ms (sect. 4-33).
f.
150 pcf and
s - 100 pcf.
W
Scaled thicknesses of concrete and sand: 4A-64
TK
5-1300/NA~FACP-397/AFR88-22
(eq. 4-125) (eq. 4-126)
...
Scaled blast impulse resisted by individual concrete panels. .
Step 3.
-i bd· - -i ba
1500 - -----.,.(1000)1/3
-150 psi-ms/lb c,.
(eq. 4-127)
Correction of scaled impulse resisted by concrete panel used in . composite walls.
Step 4.
(Corr. )
2]; + -100 (-) 150 2
.'
150 0.833 - 207 psi-ms/lb l/3 - i ha (corr.) Step 5,'
,
Scaled .blast impulse attenuated by scceptor panel and sand. i a -'280 psi-ms/lb
Step 6.
Total scaled blast impulse resisted by wall,
...) Step 7.
(fig. 4-58)
I bt
-
I a + I bd
- 280 + 207 - 487 psi-ms/lb l/3
Comparison of wall capacity and applied blast load. O.K. "
4A-65 ,(
TH 5-1300/NAVFAC'P-397/AFR.88-22 '. . -...... ..., Problem 4A-6, Design of a Beam in Flexure Problem:
'.Dload. esign an interior beam of a roof subjected'to an overhead blast
Solution: Step 1.
Establish design parameters: "
a.
Structu~al
configuration.
b.
Pressure-time loading.
c,'
Maximum allowable support rotation.,
d.
Material properties
'J
Step 2.
From table 4-1, determine the dynamic increase factors, DIF. For the deflection criteria given ,in Step lc, find the equation for the dynamic design stress from table 4-2. Using the DIF and the material properties from Step Id, calculate the dynamic design stresses.
Step
3.
Assuming reinforcing steel and concrete cover, calculate the distance from the extreme compression.fiper to the centroid of the tension reinforcement; d.
Step 4.
Calculate the reinfo,cement ratio of the steel assumed in Step 3. Check that this ratio is greater than the minimum reinforcement required by equation 4-137,but,less,than the 'maximum, reinforcement permitted by equation 4-132.
Step 5.
Using equations 4-129 and 4-130, the dynamic design stresses from Step 2, and the'value of d,from Step 3. calculate the 'ultimate moment capacity of the beam.
Step 6.
Compute the ultimate unit ,resistance of the beam ,using the moment capacity of Step 4 and an equation from table 3-1.
Step 7.
Calculate the modulus of elasticity 'of concrete Ec and steelEs (equations 4-4 and 4-5, respectively) and the modular ratio n (equation,4-6). Determine the average moment of inertia I a of the beam according to Section,4-15.
Step
From table 3-8, find the correct equation for the equivalent elastic stiffness KE. Evaluate this equation using the values of Ec and I a from Step 7.
8.
Step 9.
With the ultimate resistance from Step 6 and the stiffness KE from Step 8, use equation 3-36 to calculate the equivalent elastic deflection XE.
4A-66 \
TH 5-l300/NAVFACP:397/AFR88-22 Step 10.
Find the values for the load-mass factor KLM in the elastic, elasto-plastic and plastic ranges from table 3-12. Average these values according to Section 3-17.4 to determine the, value of KLM to be used in design.
Step 11.
Determine the natural period of vibration TN using equation 3-60, ' KLM from Step~O, KE from.Step 8 and. the mass.of.the beam. The mass includes 20 percent of the adjacent slabs.
Step 12.
Calculate the non-dimensional parameters T/l'N and ru/P' Using the appropriate r,esponse' chart by Chapter 3' determine :the ductility ratio, j.l.
Step 13.
Compute the max~mum deflection Xm using the ductility ratio from Step 12 and XE from Step 9. Calculate··the suppo.rt rotation corresponding to Xm using an equation from table 3-5. Compare this rotation with the maximum allowable rotation of Step lc .
Step 14.
Verify that the ultimate support shear.V s ·given·in table 3-9 does not exceed the maximum shear permitted by equation 4-142. If it does, the . size of the beam':must, be', Lnc reaaed-rand Steps 2 through 13: repeated. -e : • ", ,
Step 15.
Calculate the diagonal tension stress Vu fr~m,equation 4:139 and' check, that it.:does -not; exceed.lO(fdc l ) l~ 2. ' ,
'." r>,
3.
-"~_.:
'.
...
;
. ,,,
',1·
'-(
~,
.,.
.",
-~::.
Step 16.
Using the dynamic concrete 'strength f dc' from Step 2 arid equation 4-140, calculate the shear capacity of the unreinforced web, v c'
Step'17 .
Design the shear reinforcement using_equa~lon 4-140, ~nd the excess shear stress (vu - v c) or the shear capacity of concrete v c' whichever is;greater. , ~.
Step 18.
Check'that the shear reinforcement meets .the'minimum'area and of Section 4-39.4.
~axi~umspacing requirements
Step 19 . .,., With T/TN and Xm/.XE from Step 12, enter figure 3-268 and read the required resistance of the beam in rebound. Step 20.,,"
Repeat; ~ ......... tance
.
Steps' 3 through 6-to -sat Lafy the required .rebound resis'.J
.
. .' . - Example 4A-6. Design of· a Beam in Flexure \' "..
~.".'
.'
Required: . 'Design of an interior of a roof beam subjected to an overhead blast load.
., ,. ~
4A-67",
-'.
TN 5-1300/NAVFAC: P"397/AFR 88-22 '. ,
Solution: Step 1.
Given: . ~ .' :
f
a.
Structural configuration is shown in figure 4A-15a.
b . ',. Pressure- time. loading is' shown· figure 4A-15c. '~
"
c.
Maximum support rotation of one degree.
I.> Yield stress
of, reinforcirig steel, ' f
Y
- 66 000 :psi 'i
Concrete compressive strength, f'c - 4,000 psi Weight of concrete; w -
,.
Step 2.
a.
Dynamic increase factors from table 4-1 for intermediate and flow preasure., range. . •. "
"
.
Reinforcing steel
bending, DIF .~ - direct shear, DIF
Concrete·
1.19. direct shear,' 'DIF' ... L 10 - diagonal t~nsi~n, DIF - 1.00 compression,~DIF
'c,' _
b.
.,"
;
I,
...
Dynamic design stresses from'equation 4-3. ;ReinforCing. steel - 'bending"·
.~
..
1. 17 -x 66,000 - ·77 ,220 psi
diagonal tensi?n :~y
,. '.
"
From table 4-2, for em S 2°: . f'ds. -
c.
1.17 -' 1: 10
It'.....-
:Concrete ,"
1. 00 x 66,060 " 66,000 psi
- compression.
f'dc '- 1.19 it 4,000· 4,760 psi
- direct shear
f'dc
1.10 x 4,000 - 4,400 psi . ,
diagonal tension f dy' - 1.00 x 4,000 - 4,000 psi ·Step 3.
Assume 5 No. 6 bars
for
bending:
As - 5 x .44 - 2.20 in 2 For concrete cover and beam sections,see figure 4A-15b.
,4A-68
,
14-'~'------il1-"---==,,'--=-.~ 14'-0"
18"
!I • II
"I
18"
. ' . ,.
..
.I
20'- 0 " ·
:1' . '
-
'.
,"
.-, .•
18"
.
I'!, I
r--r--;--- , - -.-~ r-,--, -----.,-... --r-"-,-I--=-+I. I ' I I' I I <.
,~~...,
'
--~-I--;---' --t--i----·------r-i-~
. ' II I
I I
'
il"
I I
1- -
I I
" ,'\'
I I
-
I I
II II
" I "I
I I
I I
I I I
I I I I
.'
I
I I I I
.
1
I I I I
2
'
I l
1 , I' ,r--r-r -.- - --- r-i----· -. . -, --""':--j-r- =!!! i-"'-jf---;---,-, -. ~-r-:"----7"----'---~-i-',i It'-+.
.
.
~
I I I
I I
I
o
.'
,
~,
.
. ,
,
(a) .\;,r
;' -CZ)
'(\I (\I
--(\I'1,...1 U<'i
,',
n::;·~~~~.:~, ., •
.
I~
.
.~
"
'--
., -
-
CL, .I: ..
.
:
....:,
"
"
," •
I~"
"=,,-2
CL.
~4 fJ
, , i ";
z
."
.
= ,f
r_;
(b) "
,p~i
.'
,
, ,,7.2
• .: :
;;~.f,·
60.7 ( c)
FIGURE
4A-15
4A-69 .'
ms
TK 5-1300/NAVrAc'P':397/AFR 88-22 Cnesk reInforcement requireme~ts:
Step 4.'
a:·" (i;;lcu1~te d nega'ti';'~' '('support) '~;;d positive(mid-~:pan) for , "'checking bending "re1nfcircementratioii. '" - ,
..Ji
,
"
,
,
,
.
d - h - d' (cover)' -
,
f
..
';
I,
:
•
q, , (tie)
(Blmding Bar)
2
.
,
30-2-0.5,-p.75/2
b.
27.125in
30·- 1.5 - 0'05 J; 0:75/2 - 27.625 in. , f Calculate reinforcement, ratio: j
From equation 4'131.
PN'-' 2.2/(18
x' 27;:125)-"0:0045
,*
"
Pp - 2.2/(18 x 2~)615) c.
'--1'" ".. ... "... .
..--- -,
"
\-
0.0044
Maximum reinforcement: ' Maximum r~inforciTlg.,.rati,o·:, Pmax .~ 0~J5 X..Pb, ,
From equ,ation .4-132;
i
',.:'t,:-
K1 - 0.85
.,
~ .. i
'.
.t .
-. "l
r
, .. 0.85K1 dc I' .,' Pb • where:
• ...,":f.~', .:
,,87.. 000
• , 8y;OClO , .
- Jr ._~_
.~
+ f dy '-'
.... ".
..
-..
"
O. 05 ~t
0.85 x 0.812,x 4760
87,000
Pb - - - - ' - - - ' - - - - - - .... ( - - - - - - ) 77,220' . d7~006 + 77,220'
0.0225
. I .
Pmax - 0.75 x 0.0225 - 0.0.169> PN - 0.0045 Pp - 0.0044 O.K.
d.
and
" .
I
'
.Check for ininimumreinforcing ratio usi~g,equation 4-138 ".i
-..~
TH 5'-1300/NAVFAC 'P:397/AFR 88-22 Pmin - 200 1 f y i .
200 Pmin - --------- - 0.0033 < PN - 0.0045 O.K. 60,000
, < Pp - 0.0044 O.K. Step 5.
Moment capacity of the beam using equations 4-129 and 4-130 is:
where: As f dy 0.85bf'dc 2.20 x 77,220 a--.,----,-.,---. ..' ,0.85 x 18 x 4,760
-2.333 in -,
at support:
MN - 2.20-x·77,220.x (27.125 • 2.333/2) . -.4,409,934·in-lbs
.•
at mid-span: ~
- 2.20 x 77,220 x (27.625 - 2.333/2) - 4,494;876 in-lbs
Step 6.
From table 3-1, ultimate resistance of a uniformly loaded beam with fixed ends is: ru -
8-(4,409,934 + 4,494,934) r u - - - - - - - - - - - - - 1,236.79 Ibslin Step 7. a.
From Section 4-15, calculate average moment of inertia of the beam section. Concrete modulus of elasticity (eq. 4-4):
4A-71
TH ·5-1300/NAVFACP-397/AFR88-22 '.
Ec - wl . 5 x 33 x (f'c)1/2 Ec - 150 1. 5 x.33 x (4,000)1/2
,
- 3.8 x 10 6 psi
Steel modulus of elasticity (eq. 4-5):
b.
Es - 29 x 10 6 psi c.
,.
Modular ratio ·(eq. 4-6) :
-
"
n----
Ec 29 x 10 6 n -
3.8 x 10 6
~
7.,6'
..
From figure 4-11 ~nd havf ng n , 'PN and Pp' the coefficients for moment of inertia of cracked sections are:
d.
FN - 0.0235 at support Fp -·0.0230 at:mid-span'· Cracked moment of inertia ·from equation 4-8b is: ~
,
,'
lcN - 0.0235·x 18 x'27.125 3 lcp - 0.0230
'x
8,442 in4
18 x 27.625 3 - 8,728 in4
Average:' Ic -
(lcN + lcp) / 2 • 8,442 + 8,728-.;,
Ic -
e.
•
""
. ,.' ...
Gross moment of inertia
,
(eq.
...
.,;
8,585 in4 4-8):
'.::
- ..
4A-72
TH 5-1300/NAVFAC'
P~397/AFR
88'-22
bh 3 I
---
12
g
.18 x 30 3
- 40,500 in4
I g - ----
, .....
12
;" -,
f.
Average moment of inertia of the beams from equation 4-7: ,
40,500 + 8,585 Ia -
Step 8.
From table 3-8,
.
- 24,542.5' in4
2 ,KE
of a uniformly loaded beam '.with f,ixed ends is:
307 Ec I a L4
307 x 3.8 x 10 6 x 24,542.5 2404 - 8,629.70 lbs/in/in Step 9.
Step 10.
Equivalent elastic deflection from equation 3-36 is: ru
1,236.79
KE
8,629.70
Load-mass factor from table 3-12 for a plastic range of a uniformly loaded beam with fixed ends is: elastic - 0.77 elasto-plastic - 0.78 - plastic - 0.66 KI1f for plastic mode deflections; from Section 3-17.4 from Chapter 3:
'u<-
[[O.";O·"J.o.,,]
4A-73 '
/2
-0.77
TH5-1300/NAVFACP-397/AFR 88"22 Step 11.
Natural period of the beam from equation.3-60 is: TN - 2lr (Rut mIKE) 1/2 Where m is the mass of the beam plus 20% of the slabs span perpendicular to the beam: m - wig .
,
m - (30 x 18 + 2 x 8 x 102 x 0.20) 1,000 2
150
x-x 12 3
32.2 x 12
194,638.50 1\?s-ms 2/in/in
Tn Step 12.
2lr [ 0.72 x 194,638.50
t
-- 25.3' ms
8,629.70
Find J.l. ductility ratio from figure 3-54. From Step 1: .'
~
T/T N - 60.7/25.3 - 2.40 P - (18 + 84 + 120) x 7.2 - 1,598.40 1bs/in 1,236.79 ru/P -
- 0.77 1,598.40
J.l - 9.0
Step 13.
,.
From table 3-5 support rotation is: L tan
e
~----
2 ~ - J.l ~.
x XE
- 9.0 x 0.1433 - 1.29 in
4A-74
TM 5-1300/NAVFAC"P-397/AFR 88."22·', , .. '\
2'.x 1 ..29' tan 9 - ' - 0.01075 240 9 - 0.620
Step 14.
sr:
"
, ;.,
\
O'.K.
Direct Shear froln'tab1e 3-9 :ts: ' '.,-
"
-»
",
Vs -
,1236 .79x 240'·' : ':. " - 148,415 Lbs 2
!r~
Section capacity in direct shearfrom.equaticm 4-142: Vd - 0.18;,f"dc bd . Vd - 0.18 x 4,400 x 111' x 27.125 ,\:. - 386,694 lbs > Vs - 148,415 Step 15.
Diagonal tension stress from equation ".
O.K.
"
4~139: \
"
10 (f',c )1/2 Total shear d distance from the face of support: ' Vu -
"
(L/2 -d), r u
1
240 ' .. - ( - - - 27.125) 1236.79 - 114,867 lb 2
114867,,, Vu - -------------- - 235.2 psi 18 x 27.125 10 (f'dc)1/2 - 10 x (4,000)1/2
Step 16.
Unreinforced web shear capacity using equation 4-140 is: V
c - [1.9 (f' dc)1/2 + 2,500 p)
< 3.5 (f'dc)1/2
4A-75
, .
TH 5-1300/NAVFAC' ,P~397/AFR,088:"22
- [1. 9 (4,,000) 1/2 + 2,500 x: 0.0045]
"e
•
.,
;;, .131.4 psi 3.5 (f'dc)1/2
3.5 x (4,000)1/2 221.4 psi '> 131'.4. psL·O.K.,
Step 170
Area of web reinforcing'from equation 4-141:
Av Vu
-
Vc
[(vu - v c) ~:bx ssl /~
X~dy;
- 235.2- 131.4 -, 104 < "e use .'
V
c
,;
Assume: Ss .. 9"in _,.'
;0'
_.
Ay'-:131.4 X 18 x 9/(0.85,Xi~6.000)
0.38 in /,9· in2 .
-r :
Use No.4 tie: .
Step 18.
:
I
"'"''
•
.
Minimum tie reinforcing area,:
'
Av A
•.
(min) - 0.0015 bs s
'(min)
~
0.0015 x 18.x 9:::
""V
. ' , :,:-, "
- 0.24 in2 < 0.40 in 2. O,.K:- .
')
"
Maximum tie spacing:
• J
- 253 psi> ~c - 131 psi.,'" > v u." Vc - 104 psi smax - d/2 Smax - 27,125 ,
'1.2.-
13. 56'in > 9 in ' O.K.
,
,
"
4A-76.· "
J
0
0
7
TH 5-130~ftIAVFAC'P-3971AFR 88-22
Step 19.""
Determ~ne required .resistance for r,eboun,d r- :,from' figure ,3-268: .' ~~
,/,'
i
~!
.;r-/ru:-'",o:50for T/TN,-,2.40, and,Xm/ XE
9.0
".
Requf.red s r
0.50 x 1236.79
'.I'
Step 20.
.1
618.4 lbs/in
H:<.
Repeat Steps 3 to 6: .Assume:
:'
'
: ~.' -
,,
2 No: 7 + 1 No.6
-
,p N at support
-0.0033 eo
"
p -p at ·mid-span
,:t.
.;,..,....r "
N-
N at.support . '"
200/fy
,. - 0; 0034 > 20,o!:fy
,
,
- 3,388,275 in-lbs ~,
. : : ! ':
3,324, 954
M-P "at mid-'span
r
i~'~:lbs
- 932.4 ,1bs/in . , ., .. ,.
> 618.4 lbs/in :·O.K. ", ,
,
Prob1em·4A-J. Design,of a Beam Subject to Torsion. ,. .' Problem:
Design a beam for.
a uniformly
distributed torsional load. •
r-t-
Procedure: Step 1.
Design the beam and. adjacent slabs in .fTexure for' the. applied blast load. ' " , : ... ",
Step 2.
Calculate the unbalanced slab support shears, VT using the ultimate resistance' of:'the slabs, from Step 1. ,.
.. ;,
f
.' "i
Using the unbs.Lanced' slab support' shears, compute the torsional . load at. d' distance "from .the. face of ,the'~support :from: : L T~
b
- ( - , - d) 2
4A-77
Step 3.
"Using the' torsional' load ,from Step 2 ..' compuce the 'noininal'"torsion- .... al stress .Ln the vertical direction from, equation, 4-143 and in the horizontal direction'from' equacdon 4-144. 'Ver.ify that the torsional stresses do not exceed the maximum stress permitted in Sec. 4-41.5. -z.. (' ·,If, . Note: :. If the height of the beam is greater than width,' the horizontal torsional stresses will not be critical arid'maybe ignore&
Step 4.
Determine the shear and torsional capacity of anunreiriforced web, V c and v t c' from equations 4-145 and 4-,146 or 4-147" the torsional stress from Step 3 and the shear stress I from step 1.
Step
Find the excess ,shear. stres~(vu ~ vc)'where'ithe- nominal shear stress v u' is from Ste,P I, !'nd the shear capacity o~ ,the unreinforced web v c is ',from Step 4:,' U~ing the' excess'lshear stress and equation 4-141, determine the area of, web reinforcing for shear. . . ,
5.
" , "
: I .,:
•• ' '.
,0'
:- .,
~;
::.....
•
•
Step 6.
With torsional capacity of the concrete from Step 4, the torsional stresses from Step'3, 'and equations 4"148 a~d 4"149, calculate the area of web reinforcement for torsion, in the vertical and, if required, in'the horizontal ,directions.
Step 7.
Add the, area of "shear reinforcem'mt from Step 5 and the area of torsion reinforcement in the, vertical direction, and compare with the area of torsion reinforcement required in the horizontal direction.' The larger.of .the' two values will'control for,the design of the closed·ties. (If height of beam is greater than width, see note at Step 3.)
Step 8.
Check minimum area and maximum spacing requirements of ties according to section 4-41.5.
:::.;
(J:
..... ~
.!
.~
, I,
Example 4A-7.Design of Beam in Torsion _' -.1,,, ',t I,' •
Required: Solution: .
•
1.
1.l'~· t
_
.';
...--.
,'
Design of 'beam in 'example 4A-6 " 'for "torsional 'load due to"unequal spans' of adjacent slab. .... •
.
.' -r:
4A-78 "
>l'
TK 5-1300/NAVFAC Step 1.
P-397/AFR88~22
Given: a. Beam designed for flexure in example 4A-6 where: 240 in
L -
d - 27.125 in;
b - 18 in
,. V
b.
u - 235.2 psi
Slabs designed , for flexure : wh~re: - 14:ft
168 in
7.85;psi· L2 - 20 ft
240 in
r u1 - 15.0 psi ~
r u2 Step 2.
Calculate torsional load.
b.
Torsional load at d from the support: Tu -
L (-
- d)
2
b 2
240
- ( -2 Step 3.
L1
(VT)
T
18 27.125) x - x 320 - 267,480 in-lb 2
Maximum torsional stress: Since h > b, the torsional' stress in the vertical direction is not critical and will be ignored. a.
Torsional stress: (eq. 4-143) 3 x 267,480 . 82.5 psi
1'8 2 x 30
4A-79
,.
TH 5-1300/NAVFAC P-397/AFR 88-22
b.
Shear stress:
c.
Check maximum allowable torsional stress.
Vu
-235.2 psi
12(f' dc )1/2
12 x (4000)1/2 - - - - - - - - - - - 212.9 > 82.5 O.K. 1 + [1.2 x 235.2 [
,
82.5 .
Step 4.
r]~ ,
Find shear and torsional capacity of unreinforced web. a.
Shear capacity:, 2 (f' dc)1/2
(eq. 4-145)
- 121.4 psi
(eq. 4-146)
howdy,
4A-80
TK 5-1300jNAVFAC P-397/AFR: 88"22
V
tc -
2.4
x
(4.000)1/2
-----------~-
r1'
[1 [".::'"
- 42.6 psi (Vertical Face) Step 5.
Area of web reinforcing for shear using equation.' 4-141:
Av·-· (':'u'
v c) x b x
s/(
Assume s - 12 in: .."
Av r."
(235.2 - 121.4) x 18 x 12/(0.85 x 66,000)
- 0.438 in2/ ft Step 6.
Web reinforcing' for torsional' stress using equation 4-148: (v t u
- v t c ) 2b h s · , · Vertical
where: 1. 50 \
h t - 30.0 - 2.0
1. 5,"
(2 x 0.5/2) - 26 in
See figure 4A-15. b t - 18.0 - 1.5
- 1.5 - (2 x 0.5/2)
See figure 4A-15 .
.,
.,
14.5 in
TH 5-1300/NAVFAC P-397/AFR88,;22 '26 at - 0.66 + 0.33 x --" . 14.5
1.25 < 1.50
O.K.
(82e5 - 42.6) 18 2 x 30 x 12
Ai: - -,---'----'------"-'-------3 x 0.85 x 1.25 x 14.5 x 26 x 66,000 0.059 in 2/ft Step 7.
Total web reinforcement:, ' At + Av/2 - 0.059 + 0,438/2 -'0.278 in2/ft/Leg Use No.4 ties @ 8 in - 0.300 in 2/ft/Leg.
Step 8.
Minimum torsion reinforcement (sect. 4-41.5): a.
Minimum tie reinforcing area: . Av (min) - Av shear alone. from example 4A-6 Use No.4 ties @:9 In., 0.38
12
---·x--
Av, (min)
2
9
- 0.253 in 2(ft/Leg < 0.300 in 2/ft/Leg b.
O.K.
Maximum spacing:
"
Smax - - - - 4
26' + 14,.5 Smax - - - - - - 4
- 10.125 in> 7 in Step 9.
O.K.
Required area of longitudinal steel is the greater of the two values from equations 4-l5la or 4-l5lb.
4A-82
TM
'..
5-l300/NAVF~CP~397/AF!t
,88-2,2
bt + ht t , Al - 2At x - - - s
14.5 + 26.0 Al - 2 x 0.06 x - - - - 12 .' .
..
or: _[ <00
Al
x b x s . f dy .
vt u
.' Vtu' +!vu
.
,,;,]
:.b t +
x
ht
,
s
where: SObs 2A t -
f dy
! •
SObs
50 x 18 x 12
f dy
66,000.
---
- 0 . .16 in 2/ft 2A t
. [ 400 x 18 x 12 ·,··,·(82.5) Al -
,
66.000
82.5 + 235.2
14.5 + 26.0
x------
.
;
0.~2J
0.74 in 2 .
12
Step 10.
Distribui~ Ai. As. and As-
as follows (see fig . .J4A-16): '1
Distribute Ai equally between four corners of ~he beam arid one on each face.of depth, a total of six .locations to~~atisfy maxi~um sp'acing 12 i.nches .~., . .: "
cof
....... Vertical Face:
'.
,-;
'
'.
.
One (1) No. 4 bar "', - 0.20 in2 > 0:12 .
-Horizontal Face at Top:
4A-83
,
O.K.
1M 5-1300/NAVFAc"'P-397/AFR 88-22Support - 2.20 (bending) + 2 x 0.12 (torsion) 2.44 in 2 Two (2) No. 7 at corners + three (3) No. 6
,
'
_ 2.52 in 2
O.K.
Midspan - 1.64 (rebound) Two (2) No. 7 at
co rrier sr e-
- 1.64 in 2
.1
one (1) No.6'
O.K.
Horizontal Face at Bottom:
"
Support - Greater of rebound (1.64 in 2) or torsion (2 x 0.12), Two (2) No. 7 at corners + one (1) No. 6 _ 1.64 in 2
.'
O.K.
j .. "
Midspan,- 2.20 (bending) Two (2) No.7 at .co rne r s + .one (1)· No.6 + two (2) No.5 i
'.
:.. 2.26 fn 2 > 2.20 ,O.K.• ,
,-'.'
,.
,f' .
., ..
"
r
r, •
...!...."-'. ~~.: ": ..---- " "
',.
f.=-
.. ~: ..
_.-
'.
.
.
l2-#7
'
'f.
'4 /),
':.~':' '. '
'[-: -
"
.~.
I
I
.
7'
'1
"
'
•
-
," X'.. . , ,.. . . . - -·-:H·· . . - ------ --; -.," 0_:'· ..... 1. , T :: !.:. - ,.
.,~\
,4
r
,
.'
2-#5 7
+ 1-#6
t
#4(§8~'
,. •
20', 0"
Figure 4A-16 .
4A-84
-,
-,
I #4 E.F./ , ,
---~
.•
}".
.
'oC
.
....;....,..
v
i...
..
1'-6',1'
.
TK 5-130oiNAVFAC P-397/AFR 88-22 . Problem 4A-8. Column Design" Problem:
Design an interio~,col~n ofa one-story structure with shear walls; .
Procedure: Step 1.
-
.' ; i
; .', .
t
"
Establish design parameters: " ,
. '.~'
Step
2.
~
a.
End conditions of coiumn.
b.
Clear height
c.
Dynamic loads 'from roof.
d.·
Static materfal properties.
for'
'.
co Iumn,
Find equivalent static loads on the column by increasing the dynamic loads' 20 percent (sect. 4-47). '\-
:
~.
" '.'
Step 3.
From table 4-1, determin~ the dynamic increase factors, DIF. Using the DIF, the material propertfes'from Step 'ld and equation 4-3, calculate the dynamic design' strength of the concrete and the re inforcement. . .'
Step 4.
Assume a column section and'reinforcing'steel. <: " .~
Step
5.
Step 6.
l·~""
Calculate the slenderness ratio of the column section assumed in Step 4, using 'either -equatrLon 4-167 ~i'·4:168. I f the slenderness ratio 'is less than 22, slenderne~'s 'effects may be neglected. If it is greater than 22 and le~s" than 50; the moment magnifier must be calculated from equation 4-170 and the moments increased according to equation 4-169. The col~n section must be increased if the slenderness ratio is greater than 50. the moment by the axial load to ,obtain the design eccentricity in both' df.rec t.Lons .' Verify that the'cdesign eccentricities are greater than the minimum eccentricity of O.lh for a tied . co Lumn and 0.0707D for a spiral column. .: Divi~e
Step 7.
Compute the balanced eccentricity eB of the'column using equation 4-156 or 4-158 (for a rectangular and circular colUmn, respectively), the dynamic material properties' from Step 3, and the section properties from Step 4. Compare the balanced eccentricity with the design eccentricity from Step 6. Determine' if the column failure is controlled by compressive strength of the concrete (eb > e) or tensile strength of reinforcement (eb'< 'e) ~ '
Step 8.
Calculate the ultimate axial load capacity, at theiactual eccentricity, in both directions. If compression controls use equation 4-160 'or 4-161. If tension controls use 'equation' 4,162 or 4-166.
4A-85
TH 5-l300/NAVFACP-397/AFR88-22 . ,. . - .... ~...
~
Step 9.
Using equation 4-154, compute the pure section.
Step 10.
Compute the ultimate ,capacity load capacities at the actual axial load capacity from Step the ultimate load capacity is load from Step 2.
aXi~l,load
of. the column section, using the, eccentridities fr~m. Step 8, pure . the ,. 9, and equation 4-176. Verify that greater than the equiv~lent static .,. ..,.
Step 11.
capacity of the
,
.
,
'.
'Provide ties according to section 4-48.4 for a tied column, or section 4-49.4 for a spiral reinforced c~lumn.",:, '.; Example 4A-8. Column. Desigll.. .. ~
Required:
\,
'
Design of a rectangular, tied interior' column.
Solution:
.. }
.
Given:
Step 1.
:
"
~
Both ends of'col~n fixed
a. b.
., Axial ioad 491,066 Ibs
c.
Moment about. x-axis, .2,9.46 ,OOO,inclbs· No calculated. moment about y-axis ,
~
.••• _'~
..
\j..
l
-:.
..,:
:. ,
a. '
.....
":'
IbsN . ••
..
. 1. i •.
.
: '
",
,
n -
_
-'0 Ln-Lbs. • s .: 1·.'t·-,
'.,
L
. ~ tt: .' :
r : ,.. :J\'. :~:.~ or, 11'.;
.. ;
,,1
.'
Dynamic material s t rength.. : __ .. .
I
T' -:-
-2,946,OOO'x 1.2 .. 3,535,200 in-lbs'
.', ""Y
~
~
"
Moment about y-axis (
c.
Step 3.
.:
"
"
,.
-
,5~~,200
..
...
.
"..
Moment about x - a~i" ': ~
'
.' '. ~ " -,
\.
P- A9l,OOO x 1.2 -
"
"
,"
Equivalent static . loads '.~
b.
...',
.r .' :.'
Axial load
."
.""
.•
" ~?ncrete, f: c ;'.:4, OOp psi
a.
r
r;
Reinforcing steel,:~y:" ~6iOOo., p,,~, . "'t.~
Step 2.
't
:'
Reinforcing Steel.
," .: ..
"
4A-86
,~
,
.: ;.'
":
.
'.
.
TM 5-1300/NAVFAC P-397/AFR88-22
72,600 psi b.
Concrete. f'dc - f'c x DIF f'dc _: ·4·,000 x 1.12
Step 4.
-
4',480 psi
Use an 18." x lS" column section with 12 No~· 7 r e Infcrc Ingrbars (see fig. 4A-17).
Figure 4A-17 Step
5.
Radius of gyration for rectangular sectiori is equal to 0.3 of depth. ,
..
~.
r x - r y - 0.3 x 18 - 5.4 in From-section 4-46: k - 0.9 kl
'
O. 9x 120
20 < 22 r
.....
5.4
4A-87
(eq. 4-168)
...
TH 5-1300/NAVFAC, P-397/AFR 88-22 •
, I
,.'
"0' -
.
~
•
.'
Therefore slenderness. effects may be nagLacced ... Step 6.
Minimum eccentricity in both directions,
"".
emin - 0.1 x 18 ~ 1.8 in ex
- ~/P - 3,535, 200/5~9, 200 - 6 in
"s.«. Step 7.
Hy/P - 0/5S.~,200 -.()
..
.
>
0j:K.
1. 8,' in
:< 1.8!, use 1.8 in,
;'
Balanced eccentricity: a.
From figure 4A-17,
dx - dy -
18 - h5·- 0.5 - 0.875/2
-
15.56 in
As x - 4 x 0.6 - 2.40 in 2 Asy - 2 x 0.6.- 1.20 in 2
b.
Find value of m, -.'
m - fdy/(0.85,~· qc)' - 72,600/(0.85 x 4,480)
19.06
I
c.
Using equatiop 4'-156, -
"
-,
0.20 h- + 1.54 s r,
.',
A m/b ;
.
0.20 , x 1S. '+(1.54 x2 .40 x 19.06)/18 ' . "
'.
- 7.51 i~ > 6 in,. compression'controls eby
-0.20 x'18 +·(1.54 x 1:2·x 19.06)/18' - 5.56 in> 1.8 in, compression controls
Step 8.
Axial ,load from equation 4-160:
Pu a.
-
. ~/(2d- -, h). + 0.5. '.'
+
bhf'dc . :.3he/d 2 + ,1.18 "
When only eccentricity ex is present: r
2.4 x 72,600 Px - - - - - - - - - - - 6/(2 x 15.56 - 18) + 0.5 - 758,417 Ibs I,
4A-88
18 x 18'4,480
+ 3 x 18 6!ii5·. 56) 2'+ 1.18
TH b.
5-1300/NAVFAC~P-397/AFR
88-22
When 'only eccentricity e y is"present: '. PY -
~
1.8/(2 x 15.56
18) + 0.5'
,.
,18 x·U x'4'.480 + -----------,-----3 xU x 1.8/(15.56)2 + 11.18 Ii:
1.054.557 1bs
"',,
-',
;
Compute pure axial load capacity from equaU~.n 4-154. ., . •. "i • c '. " .~ 7"! • ~ ,1 .• 0.85 f'dc (Ag - As t) +A~t fc:iy Po 324 in 2 Ag 18 x 18 ., :. As t - 12 x 0.6 - 7.2 in 2
Step 9.
-
-
Po - 0.85·x 4480 (324
7!2) + 7.2 x-72.600
..
1;729,094 1bs
Step 10.
1
1 + Px
-
1 Py
:' ."
-r,'
.< •
1 ---l/lbs 1.054.557 ..···:1.729.094' .:·592.254
592.254 1bs
1
>. ·589.200··'O.K.'
".
,
1
.
" , . ' _' '-"I
Provide ties. according to section 4-48.4 . •~
;.
.:' '.
• .' -, I{"
For 1/7 longitudinal bars. use 113."Ues.', '.
, 1"
;
., 1
758.417
Step 11.
,
Po
+ Pu -
~
1
1
1
..
'
. Ultimate capacity of the column from equation 4-176: .'
,
;
,
.
,
..
4A-89 -,
.
-
,TH 5-l300/NAVFAC P-397/AFR 88-22"
Use two (2) #3 ties at 9 inches ,arranged as 'shown in figure 4A-17. 'Problem 4A-9, Brittle Mode, Post-Failure Fragments
......
, Problem:
Design an element, which responds to blast impulse, for ,controlled post-failure fragments.
Procedure: Step 1.
Step 2. Step,3:
Establish design parameters: a.
Impulse load and duration (Chapter 2) .:, ,.'
b.
Maximum average velocity of post-failure'fragments'vf as required by receiver sensitivity.
c.
Geometry of element.
d,
Support conditions:
e.
Materials to be used and 'corre;ponding static design strengths. '
f.
Dynamic increase factors.(tabl~,4-l).·
Determine dynamic yield'strength and ultimate strength of reinforcement from equation 4-3.
,..
"
.
.
.'
.
Determine the dynamic design stress for the flexural reinforcement . according to the deflection ,range ,(support rotation) from table 4-2.
, Step 4.
Substitute known quantities of i b (step la); f ds (step' 3), H (step lc), arid.vf (step lb) into equation 4-194,
Step 5.
'Obtain optimum ratio of vertical to horizontal reinforcement Pv/PH for a given wall thickness Tc: a.
Assume a value of dc and substitute it into the equation obtained in step 4.
b.
Read optimum: Pv/PH ratio from figure 4-38.
4A-90
TIl:
Note:
1
i
5-l300jNAVFAC P",397/AFR 88-22
~
For one-way elements, the ratio of the main to secondary rein, f'or'cemerrt is always '4 to -1. unless minimum conditions. govern (table" "43) . . Obtain Cu f rcm.icab Le 4-.11 .fo r the given support condition. Cf is always equal to 22,500 (Sect. 4-58) Step 6.
F~r the'optimUm PV/PH r at i.o , ,determine Cu '(from figure 4-33, 4-34 or 4-35). and Cf (from figure 4-73, 4-74 or,4-75). Calculate Pv and PH' Select bar" sizes, and spacings -naces aary to 'furnish the required reinforcement ratios.
Step 7.
Determine the required lacing and diagonal bars~ (Procedure is exactly the same as that for, elements designed for' the ductile mode for incipient failure or less. See problem 4A-3.)
Step 8.
Determine the required Tc for the assume~ d c select~d flexural and lacing ,bar sizes and required concrete' cover. Adjust Tc to the nearest whole inch and ~alculate the actual d c'
,"
-,·,,'t"
o
Step
9.
. . .
••
?
Checkflexu~al
capacity of the element based on either blast impulse of-, post~'fai1ure 'fragment' velocity. "'Generally, lacing bar sizes do not have to be'checked since they are' not usually affected,by,a, smal~;change in'dc' a.
Compute the actual impulse capacity of the element using equation 4-194 and compare with'the antiLc Lpat.ed. b l.as t; load. "Repeat design (from step 5 on) if the capacity is less than 'that required. .
b.
Compute the actual post-failure fragment'velocity-using equation'4-l94 and compare with the value permitted by the - acceptor sensitivity. Repeat design (from step '5 on) if the • actuai ve Ioc LtyLs greate:rthan that permitted. .' .. - . ,~.,,,,
Step 10.
Determine whether the correct design procedure has ,been utilized by first' coraput Lng the response time of the: e Lement; 'tu (time to reach ultimate deflection) from equation 4,·192' or 4-193. Then compare the response time ~u with the duration of the blast load to' For elements ~hat respond to ,impulse load~ng, tu/t o > 3. Note: To obtain the most economical des Lgn , repeat' steps 5 through 10 for several wall thicknesses.and compare, their costs. Percentages of reinforcement can be used to' reduce ,the amount of calculations. In determining the required quantities of reinforce~ent, the . length of the lap spLf cevshouLd b'e' considered. -,: '
(
,',I
.f '
4A-9l
•
TK 5-1300!NAVFAC 'P-397/AFR88-22
Example 4A-9, Brittle Mode, Post-Failure ,Fragments " . ,~,
Required:
Design the back wall of an interior cell (fig:. 4A-18) .o f a multicubicle str~cture·for ,controlled post-failure' fragments. ,
Solution: Step'l.
-:
"
Given: ,
0",'
a.
i<800psi-ms,' and to' -, 1'.0 ms.
b.
100 fps - 1.2 in/ms. :
,
,
H ... 120' in.
c,
L-
d.
Fixed on three' edges and one ,edge free.
e.
Reinforcing bars; f 66,000 psi and' f u'- 90,000 psi '.. . . ~ . . ',.. Concrete, f' c - 4,000 psi,
360 Ln . ,
y-
,
"",'
DIF'~-
For reinforcement',
.f ,
,
)
1'.23 for, yield stress
'''~''.'.
DIF -'1.05 fqr'ultimate stress,'
... ,
,
'.
,
'I
, ...u ,
,
•
. ,
•
l
,
,
...
.:
I
",
.ELEVATION
h;
~
J
:.
~,
Figure.4A -.18
Dynamic strength of materials. f dy - DIY f y - 1,23
. .
-:;-
~ ,'~
.'.
,
"
. ;.::
Step 2.
.
;
X 66,000
.. '
.r '."
~,81.180:psi
,~ •
'\ -.
fd~ ~ DIF f u .- 1.05 X 90,000 - 94,500· psi
Step 3.
Dynamic design stress from'table 4-2. 81,180 + 94,500 87,840 psi 2
4A-92 '
. i
"
,
TH 5-1300/NA,VFAC P-39.71A,FR,,88-.22 Substitute known quantities into equation 4-194.
Step 4.
120
Optimum reinforcement ratio.
Step 5.
-. t __
'
..
a.
• l.~
•
·.Assume de .,. 21 in. and substitute into equation ..4-194. '23.04 X 10 6 '. -;
~
..
732 Cu PH (21)3 + 1.44 Cf (21)2 , L
Therefore:
..
r
23.04 X 10 6 PH - - - - - - - - - .;; "
."
b.
Read optimum PV/PH value from figure 4;38 for 'L/H - 3 .. Pv/PH'" 1.58 ,!
Step 6.
For ,?ptimum PV/PH -
1. 58
From ·figure 4-34, Cu - 477 From figure 4-74, Cf - 1.583 X 10 4 [23.04 X 10 6 - 635 (1.583 X 10 4)]
PH Pv -
-
(6.78 X 10 6). (47.7-) . 1.58 PH - 1.58 (0.00402)
- 0:00635
AsH - 0.00402 (12) (21) - 1.013 in 21ft.,
·e 4A-93
0.00402 h
-
TM 5-1300/NAVFAC'P-397/AFR 88-22 use #8 @ 9 in. (As - 1.05.iri2)
-"
,
"
..
rj
.s
0.00635 (12) (21)- 1.60 in 21ft., use #10 @ 9 in. (As ~ 1.69 in 2) 1.69 - ---- - ----
Actual PH
AsH
-
1.610
r.05
-,
No. f vertical lacing ba r s va r'e renot shown since they are similar to 4A-3 for incipient failure design. design for shear will not be shown.) .; . , .. , , ", The actual d c depends upon'the details of the'base of the wall (region of vertical lacing).
Step 7.
Using lacing method No.3, quired. (Calculations are ,those ,presented 'in example Also, the remainder of the
Step 8.
- 2 X 0.75 3"
~CL.
(TYP.)
'-
-10 VERT. REINF.
-8 HORIZ. REINF. ~_~-7 LACING BARS
lacing horizontal bars
- 1.50
2 X 0:.75
1.75
- 2 X '1.00
2.00 "9",.,
vertical bars - 1.27
-l.....U '··t.
.,
Figure 4A·19 Tc - d c + 6.52 - 27.52 in., use,28"in. "
actual d c - 28
':
6:52.- 2l.48·in. .'
4A-94
6.52 in .
TIl 5:l300/NAVFACP~397/AFR 88-22 Step 9.
.:
Actual capacity of, element. a.
Actual impulse capacity. For Py/PH; - '1.610, Cu '-, 481 " and
Cf - 1.~88 X 104
d c - 21. 48 in.
PH -
"
1.05
i c2 -
- 0.00407
12 (21.48)
' (0.00407) (21.48)3 (87,840)'] 481 [ 120
+ (l!588 X 10 4) (21.48)2 (1.2)2
i 2 _ c
24.75 X 10 6 ,
'.
i c - 4,975 psi-ms > i b - 4,800 psi-ms b.
O.K:
Actual post-failure fragment velocity.
4,800 2 - 481
0.00407 (21.48)3 (87,840); 120':".
Vf - 1.10 in/ms - 92 fps
4A-95. '
<
100 fps
}
O.K.
.,
.l".- •
TH 5-1300/NAVFAC P-397/AFR 88-22 Step 10.
Response time of element t u:
.
.
,
'~
"
(1.05) I (87,840.) .(21.48) . . ' <
12 - 165,095 in-1bs/in (1.69) (87,840) (21.48)
, 12
"
- 265,725 in-1bs/in ~ie1d
x2/x1 - 1. 0 ,(symmetrical
360
L
120
H
lines)
26, 725. ]l..i, [-2---X 165,095 '
2.69
265,725
MVp
-MVN2
-
-
1.0
265,725
From figure 3-11,
.... :
,~,
x1/L - 0.358 xl - x2 -
0.358 (360)
-
128.88 ,,
i'
5(M HN1 +MHP)
ru -
(table 3-2)
x 2. 1
.11,.
5 (2 X 165,095) , ru
-
. '120 2 '
- 99.4 psi
4A-96
TH 5-1300/NAVFAC P-397/AFR 88-22 m - 225d c - 225 (21.48) - 4833' psi-ms 2/in' (fig. 3-44)
(KLM)u -,0.557
~ - 0.557 '(4833) -2692 psi:ms 2/in ' Xl - 128.88 > H - 120 ~ Vertical supports fail first and the post-u1timate"range resistance is:
8 (2 X 165,095) 360 2
~p Xl -
20.4 psi
0.66 X ~833 - 3,190 psi;ms 2/in H tanBmax -
12 X tan 12° - 25.51 in.
(table 3-6)
Use equation 4-192 for t u'
t
4, 800 [ 3190 + u - - 2,692 2692 X.20.4 2692 ( - - ) 1.1 20.4
1
tu ,
2 _1] [4800 -' 2 X 2692 X 99.4 X 25.51] 99.4
3.78 ms
-' 3.78 / 1.0 ;. 3.78 > 3"..
4A-97
correct procedure: has been used
l>z
TH
5-l300/NAVF~C,P-397/AFR
88-22
Problem 4A-10, Maximum Fragment Penetration .,
Problem:
•
.L.
Determine the maximum penetration of a primary metal fragment into a concrete wall and determine if perforation occurs.
Procedure: Step 1.
Establish design parameters: a.
Type of fragment
b.
Weight Wf and diameter d of ,fragment
c.
Fragment striking velocity
d.
Thickness T~ and the ultimate compressive strength f'c of concrete w.all
'.
V
s
Step 2.
Determine the maximum penetration Xf from ,figure 4-78 (or equation 4-200 or 4-201) for the values of Wf and V s in step 1. Xf is the maximum penetration of an armor-piercing steel fragment into 4,000 psi concrete.
Step 3.
To determine the depth of penetration into concrete with ultimate strength other than 4,000 psi, use Xf from step 2, the concrete strength from step ld and equation 4-202:
"
,
Step 4.
X'f ~ Xf (4000/f' c)1/2' To obtain the maximum penetration into concrete by metal fragments other than those of armor-piercing fragments use the penetration from step 3, the penetrability coefficient k from table 4-16 and equation 4-203. X'f-kXf
Step 5.
Calcula~e the limit~ng thickness of concrete at which' perforation will occur from· equation 4-204:
Tpf - 1.13 X d O. l + 1.311 d Where applicable, replace Xf with X'f from step 3 or 4. If T f is less than Tc embedment will occur; if Tpf is greater than Tc ~he fragment will perforate the wall.
TK 5-1300/NAVFAC P-397/AFR 88-22 Example 4A-l0. Maximum Fragment Penetration, Maximum penetration of a primary fragment into a concrete wall; determine if perforation occurs.
Required: Solution: Step 1.
c.
Striking velocity:
V
s - 3.500 fps , '18 inches '.,
d.
Ultimate concrete compressive strength: Step 2.',
C
f'c - 4,000 psi
Maximum penetration:
I
For: 30 oz and
Vs
- 3.500 fps "
,
14.5 inches from figure 4-78 or'from equation 4-199 ,
'.'.;'
Xf - 2.04 x 10- 6 ~1.'2 V s 1.8 + d ~ 2 d 2~04 x 10- 6 (2.15)1.2 (3500)1.8 + 2.15 ,'.,'
. - 14.4 inches 2 x 2.15
4.30
O.K.
"
"e: .'
.'
":
J•.
4A-99
TH 5-1300/NAVFAC P"397/AFR 88-22 '(eq. 4-202)
Concrete 'strength adjustment
Step 3.
X'f
Xf (4000/f'c)1/2 14. 5 x 1 -' 14. 5 in.
'Fragment material adjustment:
Step 4.
a.
b.
"
!'-
(table 4-16)
k - 0.70
(eq. 4-203)
kXf
0.7 x 14.5 - 10.15 in Calculate.minimum thickness to prevent perforation.
Step 5.
1.13 X'f dO. l + 1.311 d
(eq. 4-204)
1.13 (10.15) (2.15)°·1 + 1.311 (2.15)
.'
15.20 inches S 18 in Since Tpf is less than Tc the fragment does not perforate the slab.
Problem 4A-11, Determination of the Occurrence and Effects of Perforation Problem:
Determine ,the residual velocity of a primary 'fragment if it perforates a conc,ete wall.
Procedure:
Step 1.
Determine the type, weight Wf, and striking velocity V s of the primary fragment. Also, the thickness of the concrete wall Tc and the ultimate compressive stress f'c of the concrete must be known.
Step 2.
Proceed through steps 2, 3, 4, and 5 of Problem 4A-IO.If Tpf is greater than Tc' perforation will occur. '
Step 3.
If perforation is not indicated by the calculations in step 2 then discontinue the analysis.' If' perforation' does result. then compute the value of Tc/Tpf'
4A-100
TM 5-l300/NAVFAC P-397/AFR 88-22 Step 4.
Utilizing the value of Tc/Tpf' obtain vr/~s from figure 4-79 or 480. , With the values of V s and vrlvs of Steps 1 and 4, calculate the residual velocity v r. ' '
Step 5.
Example 4A-ll, Determination of the Occurrence and Effects of-Perforation Required:
Residual velocity of a primary fragment if it perforates a con, crete wall.
Solution: Step 1.
Given: a.
Type of metal:
b.
Primary fragment weight:
c.
Primary fragment diameter:
d.
Striking velocity:
Vs
e.
Thickness of wall:
Tc - 12 inches
. f. Step 2.
mild steel Wf
20 ounces d - 1.89 inches
- 4,700 fps'
Ultimate concrete' compressive stress:
f'c - 4,500 psi
For given conditions: Xf - 19.75 'inches
(fig. 4-78)
Actual maximum penetration:
x', f
. '. .
- 19.75(4000/4500) 1/2
(eq. 4-202) , ,;
'
- 18.6 inches ,
"
}'
, (table 4-16)
k - 0.70 X' f - 0.7' (18.6) - 13;0 inches
"
-\
(eq. 4-203)
Since Tpf is greater than X'f' and X'f (13.0 inches) already exceeds the wall thickness (12 inches), perforation will occur. It is necessary to determine Tpf (eq. 4"204) in order to calculate the residual fragment velocity. Hence:
4A-lOl
TH 5 c1300/NAVFAC P-397/AFR 88-22
.,
Tpf - 1.13 (13.0)· (1.89)°·1 + 1.311 (1.89) ( . 18.1 inches
- 15.6 + 2.5. Step 3.
Tc/Tpf -
Step .4.
.'
.
12.0/18.1
-.0.663·
.
.
Since the given conditions correspond ..to·a case
'.:
.
where:
. ~ .' . .)
determine: for:
0.663
from figure 4-80.
Obtain: Step 5. - 0.55 (4,700) - 2,585 fps Problem 4A-12. Determination. of the Occurrence of Spalling , !'
Problem:
To determine if spalling of a concrete wall occurs if there is no perforation by the primary..fragment.
Procedure: Step 1. r -•.
J
Determine the type. weight Wf and striking.velocity V s of the primary fragment. Arso,' the thickness of ;the concrete .wall and the ultimate compressive stress of the concrete f'c must be known. I
Step 2.
Proceed through Steps 2, 3, 4, and 5 of Problem 4A-lO.
Step 3.
If embedment of the fragment occurs, compute .the limiting concrete thickness at which spalling witl occur according to:
.
,. ..
,'
4A-102
TK 5-l300/NAVFAC P-397/AFR 88-22 Ts p - 1.215 Xf dO. l + 2.12 d
(eq. 4-207)
If Ts p, is greater than Tc' spalling ,will occur. Example 4A-12. Required:
Determination of the Occurrence of Spalling
Determine if spalling of a concrete wall'occurs due to penetration by a primary fragment.
Solution: Given:
Step 1.
Step 2.
a.
Type of metal:
armor-piercing steel
b.
Primary fragment weight:
c.
Primary fragment diameter:
d.
Striking velocity:
v s - 3,000 fps
e.
Thickness of wall:
Tc - 19 inches
f.
Ultimate concrete compressive stress: ,f' c - 5,000 psi
Wf - 40 ounces 2.38 inches
For given conditions: Xf
- 12.8 inches
X'f
- 12.8 (4000/5000)2
- 11.5
c
(fig. 4-78)
°i·,
"
(eq. 4-202) r
in.
k - 1.00
(table 4-16)
Then: ,'.
Tpf - 1.13 (11.5) (2.38)°·1
+ 1.311 (2.38) - 14.2 + 3.1.,.. 17.3. Lnche s-. Since Tc is greater than 17.3 inches, embedment occurs.
4A-l03
(eq. 4-204)
TM 5-1300!NAVFAC P-397/AFR 88-22
'Determine minimum concrete thickness to prevent spalling from equation 4-207:
.
,,
1.215 (11.5) (2,38)°·1 + 2.12 (2.38)
-
,
20.3 inches ,,' Since Ts p is'greater than 19 inches. spalling will occur. Problem 4A-13. Determination of the Effects , of a Primary Fragment on a Composite Wall Problem:
Determine the maximum penetration by a primary fragment into a composite wall and the resulting effects on the donor panel, sand layer and acceptor panel.
Procedure:
Step 1.
Determine the type, weight Wf• and striking velocity Vs of the primary fragment. Also, the thicknesses of the conqrete donor panel Tc (donor), the sand layer Ts' and the concrete receiver panel Tc (acceptor) and the ultimate compressive stress f'c of the concrete must be known. '
Step 2. r >
Proceed through Steps 2, ,3, 4, and. 5 of Problem 4A-lO. If embedment of the fragment in the donor panel occurs,'perform Step 3 of Problem 4A-12 to determine if spalling takes place. If perforation of the donor panel occurs,'compute:
Tc (donor) IT pf and perform Steps 4 and 5 of Problem 4A-ll to find the residual velocity v r. Step 3.
Utilizing v r (donor) as the from figure 4-81.
V
s of the sand layer and Wf, obtain Xs
If Xs is less than Ts' the fragment is embedded in the sand layer and the analysis is discohtinued." . ' If Xs is greater than Ts• perforation of the sand occurs. Step 4.
Compute Ts/Xs if perforation results. I
I,
4A-I04·
Use this value and figure , .
TK 5-1300/NAVFAC P-397/AFR'SS-22 4-S0 to obtain vr/v s of the sand layer. Calculate vr(sand). Step
5.
Utilizing vr(sand) as the V s of the acceptor wall and Wf• obtain Xf (acceptor) from figure 4-7S. Proceed through Steps 3 and 4 of Problem 4A-IO, if necessary. If embedment of the fragment in the acceptor panel occurs, perform Step 3 of Problem 4A-12 to determine if spalling takes place. If perforation of the ,acceptor wall occurs, compute Tc and perform Steps 4 and 5 of Problem 4A-ll to find v r.
(accept~r)
!
Example 4A-13. Determination of the Effects of a Primary Fragment on a Composite Wall Required:
'Maximum penetration by a primary fragment into a composite wall and the resulting effects on the donor panel. sand layer and the acceptor panel.
. ..
Solution: Step 1.
Given: a.
Type of metal:
armor-piercing steel
, b.
Primary fragment weight:
c.
Primary fragment diameter:
d.
Striking velocity:
e.
Thicknesses:
f .
Vs
d -,1.89 inches
- 4,200 fps"
Tc (donor)
12 inche;s
Ts
24 inches
Tc (acceptor)
12 inches
Ultimate concrete compressive stress: f'c - 5,000 psi
Step 2.
Wf - 20 ounces
For given conditions:
4A-I05 "
'
TN 5-1300/NAVFAC P-397/AFR 88 c22 Xf - 16.5 inches
(fig. 4-78)
X'f - 16.5 (4000/5000)2
(eq. 4-202)
14.7 inches
..
k - 1.00
(table 4-16)
Since X'f is greater than 12 inches, perforation of the donor panel will certainly occur. Tpf must be calculated in order to determine the residual velocity of ·the fragment. Tpf
- 1.13 (14.7) (1.89)°·1 + 1.311 (1.89) -·17.7 + 2.5 - 20.2 inches
,
,
Tc/Tpf - 12/20.2 -0.594 Since ~f is greater than2d,'vr/vs ~ 0.61 from figure. 4-80. , (donor) - 0.61 (4,200) - 2,562 fps:
Vr
Step 3.
For: Vs
(sand) -2,562 fps,
(fig. 4-81)
Xs - 64.0 inches Since Ts is less than 64.0 inches, perforation of the sand layer will occur. Step 4.
Ts/Xs - 24/64.0 - 0.375 from figure 4-80 . Vr/V s - 0.77
Step 5.
v r (sand) - ,0.77
V
s
v r - 0.77 (2,562) -1;970 fps " ~"
.,
4A-106 r.
"
TM 5-1300/NAVFAC P-397/AFR 88-22
Step 6.
For: Vs
(acceptor) - 1,970 fps
Xf (acceptor) - 5.6 inches 'X'f - 5.6 (4000/5000)1/2
(fig. 4-78)
(eq. 4-202)
- 5.0 inches (table 4-16)
k - 1.00 X' f - Xf Tpf - 1.13 (5.0) (i.89)0.1
(eq. 4-204)
+ 1. 311 (1. 89) - 6.0 + 2.5 - 8.5 inches Since Tpf is less than Tc (acceptor), the fragment is embedded in the receiver panel. Cal.cul.at;e Ts p to determIne -Lf spa11ing occurs:
Ts p - 1.215 (5.0) (1.89)0.1
(eq. 4-207)
+ 2.12 (1.89)
- 6.5 + 4.0 - 10.5 inches Since Ts p is less than Tc (acceptor), spa11ing will not occur.
4A-107
TM ,5··1300/NAVFACP-397/AFR. 88-22
"'.
APPENDIX 4B LIST OF SYMBOLS
TK 5-l300/NAVFAC P-397/AFR 88-22
e
a
(1) acceleration (in./ms 2) (2) depth of equivalent rectangular stress block (in.) (3), long span of a panel (in.)
ao
velocity of sold in air (ft./sec.) acceleration in x direction (in./ms 2)
~
acceleration in y direction (in./ms 2)
By
e
A
(1) area (in. 2) (2) explosive 'composition f actor (oz.1/2 i n. -3/2)
Aa
area of diagonal bars at the support within a width b (in. 2)
Ab
area of reinforcing bar (in. 2)
Ad
(1) door area (in. 2) (2) area of diagonal bars at the support within a width b (in. 2)
AD
drag area (in. 2)
Af
net area of wall excluding openings (ft. 2)
Ag
area of gross section ,( in. 2)
AH
maximum horizontal acceleration of the ground surface (g's)
Al
area of longitudinal
AL
lift area (in. 2)
~
(l)cnet area of section (in. 2) (2) area of individual wall subdivision (ft. 2)
Ao
area of openings (ft. 2)
~s
area of prestressed reinforcement (in. 2)
t~rsion reinforcement (in. 2)
,.
'
As
area of tension reinforcement within a width b (in. 2)'
A' s
area of compression reinforcement within a width b (in. 2)
A s
area of rebound reinforcement (i~.2)
AsH
area of flexural reinforcement within a width b in the horizontal direction on each face (in. 2 )*
Asp
area of spiral reinforcement (in. 2)
*
See note at end of symbols
4B-l
TIl 5-1300/NAVFAC P-397/AFR 88-22 As t
total area of reinforcing steel (in. 2) area of flexural reinforcement within a width b' in the vertical direction on each face (in. 2 )* area of one leg of a closed tie resisting torsion within a distance s (in. 2 ) total area of stirrups or lacing reinforcement in tension within a distance, Ss or sl and a width b s or b l (in. 2 )
AV
maximum vertical acceleration of the 'ground surface (g's) area of wall (ft. 2 )
area of sector I and II, respectively (in. 2 ) b
(1) (2) (3)
bf
width of compression face of flexural' member :(in.) width of concrete strip in which the direct~liear stresses at the supports are resisted by diagonal bars (in.) short span of a panel (in.) .
width of fragment (in.) width of concrete strip in which the diagonal tension-stresses are resisted by stirrups of area Ay (in.) width of concrete strip in which the diagonal tension stresses are resisted by lacing of area Ay (in.) failure perimeter for punching shear (in.)'
"
bt
center-to-center dimension of
B
explosive constant defined in table '2-7 (oz.1/ 2in·J/6)
c
(1)
(2) (3)
cI' c II
a-
closed rectangular tie
distance from the resultant applied load.to ·the. axis of'rotation .for sectors I and II, respectively (In.)
dilatational velocity of concrete (ft./sec.)
C
( 1)
Cc
*
ong-b (in:)'
distance from the resultant applied load to the axis of rotation (in. ) damping coefficient width of column capital (in.)'
cs
(2)
al
shear coefficient deflection coefficient for .flat slabs
deflection coefficient for the center of interior panel 'of flat slab
See note at end of symbols
4B-2 .
.-'
.
TM 5-l300/NAVFAC P-397/AFR
~8-22
critical damping shear coefficient for'ultimate shear stress of one-way elements CD
drag coefficient'
CDq
drag pressure (psi)
CDqo
peak drag pressure (psi)
CE
equivalent load factor
"
"
'
post-failure fragment coefficient (lb. 2 ms4/in: 8) shear coefficient for ultimate shear stress in horl.zontal direction for two-way elements*
(1) (2) (3)
e
leakage'pressure coefficient from figure 2-235 deflection coefficient for midpoint of· long side of interior flat slab panel lift ,coefficient .
CM
maximum shear coefficient
Cm
equivalent moment correction factor
Cp
compression wave seismic velocity in the soil from Table 2-10 (in./sec.)
Cr
·sound velocity in reflected region from figure 2-192 (ft./ms)
CR
force coefficient for shear at the corners of a window frame
Cra
Peak reflected pressure coefficient at angle of incidence a
Cs
shear coefficient for ultimate support shear for one-way elements
Cs H
shear coefficient for ultimate support shear in horizontal direction for two-way elements*
Cs V
shear coefficient for ultimate support shear in vertical direction for two-way elements*
Cs
deflection coefficient for-midpoint of short side of interior flat slab panel
Cu
impulse,coefficient at.deflection.Xu (psi-ms 2/in. 2)
C' u
impulse coefficient at deflection
Cv
shear coefficient for ultimate shear stress in vertical direction for two-way elements * .
.,
*
. .
See note at end of symbols
Xm
4B-3
(psi-ms 2/in. 2)
TIl 5-l300/NAVFAC P-397/AFR'SS-22,'
shear coefficient for the ultimate shear a.Long- ,the long .s Lde of window frame shear coefficient for the ultimate shear along the short side,of window frame -t
CL
confidence level (1) (2)
impulse coefficient. at,deflection Xl (psi-ms 2/in. 2 ) ratio of gas load to shock load
impulse
C2 d
"
coeffici'en~ at d~flection ~' (~si:ms2/in.2)
;~ I,
___
v..
ratio of gas load duration to shock load duration , ,
(1)
(2) (3)
distance from extreme compression fiber to cent ro Ld of' tension ',reinforcement (in.) ',Jdiameter (in . .), ,. 'fragment diameter (in.) ~'.,
. .'
d'
distance from extreme compression fiber to centroid of reinforcement (in.)
db
diameter of reinforcing bar (in.)
dc
distance between the centroids of the compression and tension reinforcement (in.)
d cH
distance between the centroids of the horizontal'compression and tension reinforcement (in.)
,
,compr~ssion
'
,"
dc o
diameter of steel core (in.) "
d cV
y .•••
:1'; ."
distance between the centroids of the vertical compression ,and tension reinforcement (in.) " '" -" distance from support and equal'to distance d'or.dd'(in.) , . -(
di
average inside diameter of explosive casing (in.)
d'i
adjusted inside diameter of casing
•
, .J.;,.' J~"
"
'. (in.)
,
.
'
distance- between.center lines of adjacent'lacing bends measured normal' to flexural reinforcement (in.) "j
distance from extreme compression' fiber' to centroid of,prestressed reinforcement (in.) ds p
depth of spalled concrete (in.) diameter of cylindrical portion of primary fragment (in.) .'
48-4 '"
.'
"(.
,
TK 5-l300/NAVFAC D
(1) (2) (3)
(4) (5) (6) (7)
P~397/AFR
88-22
unit flexural rigidity (lb-in.) location of shock front for maximum stress (ft.) minimum magazine separation distance (ft.) caliber density (lb/in. 3) overall diameter of circular section (in.) damping force (lb.) displacement of mass from shock load (in.)
equivalent loaded width of structure for non-planar wave front (ft.) maximum horizontal displacement of the ground surface (in.)
DIF
dynamic increase factor diameter of the circle through centers of reinforcement arranged in a circular pattern (in.) diameter of the spiral measured through the centerline of the spiral bar (in. )
DLF
dynamic load factor maximum vertical displacement of the ground surface (in.)
e
(1) (2) (3)
eb
base of natural logarithms and equal to 2.71828 ... distance from centroid of section to centroid of pre-stressed reinforcement (in.) actual eccentricity of load (in.)
balanced eccentricity (in.)
.., ,
(2E,)1/2
Gurney Energy Constant (ft./sec.)
E
modulus of elasticity internal work (in.-lbs.)
(1)
(2) Ec
modulus of elasticity of concrete (psi)
~
modulus of elasticity of masonry units (psi)
Es
modulus of elasticity of reinforcement (psi)
f
(1)
(2)
unit external force (psi) frequency of vibration (cps)
f' c
static ultimate compressive strength of concrete at 28 days (psi)
f'dc
dynamic ultimate compressive strength of concrete (pSi)
f'dm
dynamic ultimate compressive strength of masonry units (psi)
f ds
dynamic design stress for reinforcement (a function of f y• f u• and 6) (psi)
4B-5
•
TK 5-1300/NAVFAC P-397/AFR 88-22
dynamic ultimate ,stress of reinforcement (psi) dynamic yield stress of reinforcement ,(psi)
f' m
,
,
static ultimate compressive, .s t r sngch of mas,onry units (p s L) natural frequency of vibration,(cps) average stress in.the spe~ified
prestres~ed reinforceme~t
at ultimate load1(psi)
tensile strength of prestressing, tendon (psi)
yield stress of prestressing tendon corresponding to.a 1 elongation (psi) ., • reflection factor fs fse
fu
static design stress for
reinforce~ent
p~rcent
(ps,i)." -.
"effective stress in prestressed reinforcement after "allowances for all
prestress losses (psi) static ultimate stress of reiT)forc<;ment (psi),
k"
static yield stress of reinforcement (psi)"., " , .
(1) (2) (3)
'.
total external force (lbs.) I coefficient for moment of inertia,of cracked section function of C2 and Cl for bilinear triangular load
Fo
force in the reinforcing bars (lbs.)
FE
equivalent external force (lbs.)
FD
Drag force (lbs.) frictional force (lbs.) lift force (lbs.) vertical load supported by foundation, (lbs.)
g
acceleration due to gravity (.32.2 ft./sec. 2 )
G
shear modulus (psi)
h
(1)
(2)
charge location parameter (ft.) height of masonry wal~
~average
clearing d~stance for indiv~~~al ,are~~ .of op~n}ngs from Section
2-15.4.2 'I
ht
~
I,
center-to-center dimension of a closed rectangular tie along h (in.)
48-6
TH 5-l300/NAVFAC h'
clear height between floor slab and roof slab
H
(1)
(2)
(3)
e
88-22
span height (in.) * distance between reflecting surface(s) and/or free edge(s) in vertical direction (ft.) minimum transverse dimension of mean presented area of object (ft. )
Hc
height of charge above ground (ft.)
Hs
height of structure (ft.)
HT
height of triple point (ft.)
I\,
height of wall (ft.)
HC
heat of combustion (ft.-lb./lb.)
Hd
heat of detonation (ft.-lb./lb.)
i
unit positive-impulse (psi-ms)
ia
sum of blast impulse capacity of the receiver panel and the least impulse absorbed 'by the sand (psi-ms) "
i ba
blast impulse capacity of receiver panel (psi-ms)
i-
unit negative impulse (psi-ms)
ia
e
P~397/AFR
.":
.
-
'.
sum.of scaled unit blast impulse capacity of:receiver panel and scaled
unit blast impulse attenuated through concrete and sand in a composite element (psi-ms/lb. l/ 3)
ib
unit blast impulse (psi-ms)
ib
scaled unit blast impulse (psi-ms/lb. l/ 3)
i ba
scaled unit blast imp.ulse capacity of receiver panel of composite element (psi-ms/lb. 173) -
i bd
scaled unit blast impulse capacity of donor panel of composite element (psi-ms/lb. l/3)
i bt
total scaled unit blast impulse capacity of composite element (psimS/lb. l/ 3)
ic
impulse capacity of an element (psi-ms)
id
total drag and diffraction impulse (psi-ms)
ie
unit excess blast impulse (psi-ms)
ifs
required impulse capacity of fragment shield (ps Lvras )
*
See note at end of symbols
4B-7
c·
• ,-
:
TH 5-1300/NAVFAC P-397/AFR 88-22 ,
ig
-.
gas impulse (psi-ms) unit positive normal reflected impulse (psi-ms) , ',."
i
r
-
-,
unit negative normal reflected impulse (psi-ms) ' .
peak reflected impulse at angle of incidence a (psi-ms) unit positive incident impulse (psi-ms) i s-
unit negative incident impulse (psi-ms)
i st
impulse consumed by fragment support connection
I
(1) (2)
(psi-m~)
moment of inertia (in. 4/in. for siabs) (in. 4 for beams) total impulse applied to fragment ,
,
, 4
l'
average of gross and cracked moments of inertia (in. lin. for slabs) 4 (in. for beams) , " t , :,' moment of inertia of cracked concrete section,(in., 4/in;; for slabs) ,(in. 4 for beams) "
moment of inertia of cracked concrete section in horizontal (in. 4 /in.) *
direction,~,
moment of inertia of cracked concrete section in vertical direction (in. 4/in.)* '" moment of inertia of gross' .conc re t ecsec t Lon (in,4/ i n. f'or vs Labs ) (in. 4, for beams) t~' , s: 1m
I
mass moment of inertia (lb,-ms 2-in,)
moment of inertia of net section of masonry uni~
... '
Is
gross moment of inertia of slab (in. 4/in.)
1s t
impulse consumed by the fragment support connection (psi-ms) gross moment of Lner t La ro f waL], (in. 4/in.) '_ i
j
ratio of distance between centroids of compression and tension forces to the depth d .'
k
(1) (2)
constant depending on the casing metal effective length factor
velocity decay coefficient K
(1) (2)
*
, .v-
J
unit stiffness (psi/in. for slab~)'·(lb'.'/in./in. for beams) (lb./in. for springs) constant defined in, paragraphj2-1B ..2' ,.
See note at end of symbols
4B-8·. •
"
'
TM 5-l300/NAVFAC' P-397/AFR 88-22 elastic unit stiffness (psi/in. for slabs) (lb./in./in. for beams) elasto-plastic unit stiffness (psi/in. for slabs) (lb./in./in. for beams)
(1)
(2) KL
equivalent elastic unit stiffness (psi/in. for slabs) (lb./in./ in. for beams) equivalent spring constant (lb./in.)
load factor
KLM
'load-mass factor load-mass factor in the ultimate range load-mass factor in the post-ultimate range
e
KM
mass factor
KR
resistance factor
KE
kinetic energy
1
charge location parameter (ft.) (1) length of the yield line (in.) (2) width of \ of'the column strip (in.)
ld
basic development length of reinforcing bar (in.)
ldh
development length of 'hooked bar (in.)
lc
length of cylindrical explosive (in.)
lp
spacing of s ame t.ype of lacing bar (in.')
ls
span of flat slab panel (in.)
L
(1) (2)
i
span length (in.) * distance between reflecting surface(s) and/or free edge(s) in horizontal direct~on (ft.) ,
Lcyl
length of cylinder (in:)
Lf
length of fragment (in.)
~
clear span in short direction (in.)
Ll
length of lacing bar required in distance sl (in.)
~
clear span in long direction (in.) embedment length of reinforcing bars (in.)
*
See note at end of symbols
4B-9
,
TN 5-l300/NAVFAC P-397/AFR 8.8-22.
..
Ls
length of shaft (in.)
lu
unsupported length of column (in. )
Iv Iv-
wave .length of positive pressure phase (ft. )
Lx
clear span in long direction (in.)
l-y
clear span in short direction (in.)
,
.
wave length of negative pressure phase (ft. )
wave length of positive pressure phase at points band d, respectively (ft.) Ll
total length of sector
m
(1) (2) (3)
o~
element.normal to axis of rotation (in:)
unit mass (psi-ms 2/in. for slabs) [beams, (lb./in-ms 2)/in.] ultimate unit moment (in.-lbs./in.) mass of fragment (lbs.-ms 2/in.)
average of the effective elastic and plastic unit masses (psi-ms 2/in. for slabs) [beams, (lb./in-ms 2)/in] . effective unit mass (psi-ms 2/in. for slabs) [beabls, (lb/in-ms 2)/in] mass of spalled fragments (psi-ms 2/1n.) effective unit mass in the ultimate range (psi-ms 2/in. [beams, (lb/in-ms 2J/in.]
ffiup M
'fo~
."
slabs)
effective unit mass in the post-ultimate range (psi-ms 2/in.)
(1) (2) (3)
unit bending moment (in.-lbs./in. for slabs) total mass (lb.-ms 2/in.) design moment (in. -lbs.)
(i~.-lbs.
»,
.
,.
~
.
..
for beams) " .•
effective total mass (lb.-ms 2/in.)
~ ~-
ultimate unit' resisting beams)
mom~nt (i~.-lbs./in.2 for
slabs) .(in.-lbs. for
ultimate unit rebound moment (in.-lbs:/in. for slabs) (in.-lbs. for beams) moment of concentrated loads about line of rotation of sector (in .• lbs:) fragment distribution factor
i -
..
equivalent total mass (lb.-ms 2/in.)
", "
48-10
.
,,"
,
TM 5-1300jNAVFAC P-397/AFR 88-22
HHN
ultimate unit negative moment capacity in horizontal direction (in.1bs. lin.) *
HHP
ultimate unit positive moment capacity in horizontal direction (in.1bs ./in.) *
MoH' MoL
total panel moment ·for· direction H and ..L respectively· (in. -lbs.)
HN
ultimate unit negative moment capacity at supports (in,-lbs./in. for slabs) (in.-1bs. for beams)
Hp
ultimate unit positive moment capacity at ,midspan (in.-1bs./in.·for slabs) (in.-1bs. for beams)
...
ultimate unit negative moment capacity in vertical direction (in.1bs ./in.)
*
ultimate unit positive moment capacity in vertical direction (in.1bs ./in.) * value of smaller end moment on column value of larger end moment on column \t
(1) (2)
n
(3)
(4) (1)
N
(2)
';
modular ratio number of time intervalsnumber of glass pane tests caliber radius of the tangent ogive of fragment nose number of adjacent reflecting surfaces nose shape factor
Nf
number of primary fragments larger than Wf
Nu
axial load normal to the cross section
NT
total number of fragments '.
p
reinforcement ratio equal to As/bd or As/bdc
p'
reinforceaent ratio equal to A'sl bd or A'sl bd c
Po
reinforcement ratio producing-balanced conditions- at ultimate strength ~:'J :. t ambient atmospheric pressure (psi)
Pp
prestressed reinforcement ratio equal to ..
~s/b~
.,
~ ~
.:
mean pressure in a partially vented chamber (psi) Pmo
peak mean pressure in a partially vented-chamber (psi) .J'
*
See note at end of symbols
4B-11
\
..
-
TH 5-l300/NAVFAC P-397/AFR 88-22· '.
Pr
average.peak reflected pressure (psi) reinforcement ratio in horizontal direction on each face* "
total reinforcement ratio equal to PH + Pv
Pv
reinforcement ratio .in vertical direction. on each face*
p(x)
distributed load per unit length
P
(1) (2)
P-
negative pressure (psi)
Pc
critical axial load causing buckling (lbs.)
Pg
maximum gas pressure (psi)
Pi
interior pressure within structure (psi)
APi
interior pressure increment (psi)
Pf
fictitious peak pressure (psi)
Pmax
maximum average pressure acting on interior face of wall (psi) . (1)
(2) (3)
pressure (psi) ,concentrated load (lbs.)
peak pressure (psi) maximum axial load (lbs.). atmospheric pressure (psi)
Pr
peak positive normal reflected pressure (psi)
P r
peak negative normal reflected pressure (psi)
Pr et
peak reflected pressure at angle of incidence et (psi)
L
PR1B maximum average pressure on backwall (psi) Ps
positive incident pressure (psi) positive incident 'pressure at points band e, respectively (psi)
Ps o
peak positive incident pressure (psi)
Ps o
peak negative incident pressure (psi)
Psob,Psod,Psoe
peak positive incident pressure at pointsb, d, and e, respectively (psi)
Pu
ultimate axial load at actual ecceritr LcLty, e, (lbs.)
Px
ultimate load when eccentricity ex
* See note at end of symbols
4B-12
\
.
TK 5-l300/NAVFAC
P~397/AFR
88-22
ultimate load when eccentricity e y is present (lbs.) q
dynamic pressure (psi)
qb' qe
dynamic pressure at points band e , respectively'(psi)
.,
peak dynamic pressure (psi)
peak dynamic pressure at points ·b and e, respectively' (psi) r
(1) (2) (3) :
r
unit resistance (psi) . . . radius of spherical TNT [densitY'equals .95 lb./ft. 3 ] charge (ft.)' r!,diusof .gyration of cross' 'section of :co1umn (Ln..) , ;
unit rebound resistance (psi, for slabs) (lb./in., for beams) dynamic resistance available; .(psi)
Ar
-t
change in unit resistance (psi, for slabs) (lb./in. for beams) .'
f,
1..
'"
'1 !
.• -f· t
,. • '
radius from center of impulse load to center of door rotation .(in. ) rOL
uniform. dead load. (psi)
re
elastic unit resistance (psi, for slabs) (lb./in. for beams)
\.
.
"
rep
elasto-plastic unit resistance (psi, for slabs) (lb./in. for beams)
rfs
ultimate unit resistance of fragment shield (psi)
rs
radius of shaft (in.)
rT
tension membrane resistance (psi)
ru
ultimate unit resistance (psi, for slabs) (lb./in. for beams)
r up
post-ultimate unit resistance (psi)
r1
radius of hemispherical portion of primary fragment (in.)
R
(1) (2)
,
(3)
(4)
~
..
total internal resistance,;(lbs.) slant distance (ft.,). ratio of S/G standoff distance (ft.) . t
Re f f
., '~'-"'-
effective radius (ft.)
(1) (2)
..
distance traveled by primary fragment (ft.) distance from center of detonation (ft.) ..
Rg
uplift force at corners of window frame (lbs.)
48-13
~tl
TH 5-1300/NAVFAC P-397/AFR 88-22 Rl
radius of lacing bend (in. )
Rt
target radius (ft. )
RA
normal distance (ft. )
RE
equivalent total internal resistance (lbs. )
~
ground distance, (ft.)
Ru
total' ultimate resistance (lb. )
•
,
'
'.-.
. ".
I""
total internal resistance 'of .seccor s I and' II, resjlectively' (lbs.) s
(1)
(2) (3)
sample' standard deviation" -", spacing of torsion reinforcement in a direction parallel to the longitudinal reinforcement (Ln. ),. pitch of spiral (in.) .
'.
,"
spacing' of stirrups in the direction parallel to the longitudinal reinforcement (Ln .') ," ',' spacing of lacing in the direction parallel to the longitudinal'rein c forcement (in.) ,
S
'
S'
height of front wall or one-half its width, .whichever is smaller (ft.) .,):: .. : .. .' weighted average clearing distance with openings (ft.)
SE
strain energy
t
time (ms)
At
time increment (ms)
ta
any time (ms)
,
:<,
time of arrival of blast wave at points b, e, and f, respectively (ms) ,'" (1)
(2) t' c
(1) (2)
clearing time for reflected pressures' (ms) average casing thickness of explosive charges (in.) adjusted casing thickness (in.) Clearing time for reflected pressures adjusted for wall openings (ms) " .
rise time (ms) tE
"
time to reach maximum elastic deflection (ms) ".'
.
fictitious gas duration (ms)
4B-14 ,
\,
TM 5-l300/NAVFAC P-397/AFR 88-22 time at which maximum deflection occurs (ms)
.......... .
duration" of positive phase of blast pressure (ms)
:
duration of negative phase of blast pressure (ms)
"'
fictitious positive phase pressure duration '"(nis)" fictitious negative phase pressure duration '(ms) fictitious reflected pressure duration (ms) time at which ultimate deflection occurs (ms) •• j
time to reach yield (ms)
..,
time of arrival of blast wave (ms) time of arrival "of "ground shocK (ms) time at which partial failure occurs (ms)"'"
(1) (2) (3) Tc
duration of equivalent triangular loading function '(ros) thickness of masonry wall" (in,) toughness of material (psi:in,/in.)
thickness of concrete section (in,")
,""
scaled thickness of 'concrete section (ft./lb,l/3)
angular impulse load (lb.-ms-in.) force in the continuous reinforcement in the long span direction (lbs,) effective natural ,period of vibration (ms) "minimUm.thickness of concrete/to prevent:perforatiori by"a given fragment (in. ) ,
Tr
.... ;
rise time (ms) "
(1) (2)
thickness of sand fill (in.)" thickness of slab (in,) ""J "," r.
-,
-i .
Ts p
minimum concrete thickness 'to prevent spalling (in,)
Ts
scaled thickness of sand fill (ft./lb. l/ 3)
Tu
total torsional moment at critical section (in.-lbs.)
4B-15
..
TH 5-l300jNAVFAC P-397/AFR 88-22 Tw
thickness of wall (in.)
Ty
force of the continuous
u
particle velocity (ft./ms) .
reinfor~emept
dir~ction (~bs.)
in the short
...
.' .:
..
ultimate flexural or anchoragevbond stress. (psi) "'.,
u
shock front velocity (ft./ms)
,',
strain energy
.
I
\'1
c,
v
velocity (in.jms)
va
instantaneous velocity at any time (in./ms)
:
~,
~.,
.,"
..
'-.,' t
boundary velocity for primary fragments, .(ft./sec.) . ultimate shear stress permitted on ,an unreinforced web (psi), -t I
maximum post-failure fragment vf(avg.)
,
vel.o·~ity;.<~n.I.".'.s),
ave r age tpcs c-Ea Ll.ure {ragaent .ve Loc Lty (in./ms)
'
-v : .-.
, I
vi
velocity at incipient failure deflection t(in./ms) ,. ",:
-,
,"
-,
(t:~./sec.),
initial velocity of primary fragaent
residual velocity of primary fragment after pe r for e t t on (t:t .j~!,c. ) . striking velocity of primary fragment (ft./sec.)
,
}
.J
maximum t;ot;~ion,capacity of ap,;un.reinfo~ced~web (psPl' ._'J ." nominal torsion stress in the direction of ultimate v uH
:~hear
V
u (ps L) . _.
"
stress (psi)
"
ultimate shear stress at distance de ,from the horizontal suppor-t. (ps L) ". ultimate
she~r
stress at distance de' from the vertical support (psi)*
velocity in,x direction (in./ms.) "
velocity in y direction (in./ms.)
v
(1) (2) (3)
. ;J
,
.1'
volume of partially vented chamb~r (ft.:i )'~ .' velocity of compression wave through concrete (in.jsec.) velocity of mass under ' shock load (in./sec.):
"
»,
j-, t
•
>:
".
"t It;
• ,l
*
See note at end of symbols
4B-16
;~
....
;
... --:. .'
TM 5-1300/NAVFAC- P-397/AFR 88-22
e
e
Vd
ultimate direct shear capacity of the concrete of width b (lbs.)
VdH
shear at distance de from the vertical ',support on a unit width (lbs ./in.) *
VdV
shear at d1stance de from the horizontal support on a unit width (lbs./in. )
Vf
free volume (ft. 3)
VH
maximum horizontal velocity of the ground surface (in./sec.)
Vo
volume of structure (ft. 3)
Vs
shear at the support (lb./in., for panels) (lbs. for beam)
V sH
shear at the vertical support on a unit width (lbs .fin.) *
V sV
shear at the horizontal support on a unit width ,(lbs./in.) *
Vu
total shear on a width b (los.)
Vv
maximum vertical velocity of the ground surface (in./sec.)
Vx
unit shear along the long side of window frame (lb./in.)
Vy
unit shear along the short side of window frame, (lbs./in.)
w
applied uniform load (lbs.-in. 2)
Wc
(1)
W
weight density of sand (lbs./ft. 3)
,
(2)
s
( 1)
(2) (3)
,..
'
unit weight (psi, for panels) (lb./in. for beam) weight density of concrete (lbs./ft. 3)
des Lgri-charge weight' (lbs. ), external work (in.-lbs.) width of wall (ft.)
J
~.
.',
.:,;,.•.
WA
weight of fluid (lbs.)
WACT
actual quantity of explosives (lbs.)
Wc
total weight of explosive containers (lbs.)
WE
effective charge weight (lbs.)
WE' g
effective charge weight for gas pressure (lb. )
WEXP
weight of explosive in question (lbs.)
Wf
weight of primary fragment (oz.)
*
,
See note at end of symbols
,
48-17
"
",
..
, :,1'
TM 5-l300/NAVFAC
P-397/AF~
88-2.2.' .
,
,~
".
average fragment weight (oz,) weight of
f;a~gible element ·(l'b./f~. 2) .,
weight of .Lnner .qas Lng (lbs ..) ,;.
Wc o
total weight of steel core (lbs.)
,weD
weight of outer casing (lbs.) ;'
Wc l ' Wc 2
.
.'~
Ws
width of structure (ft.)
WD
work done
x
yield line location in horizontal direction (in.)
x
(1)
.,.1
Xa
,
,
'0-.
"
'.
"
r-~
(2)
r:."
...
total weight of plates 1 and 2', respectively (lbs.)
It
\",
,
-,I
t~'
-;
..;
'* '..
.'
deflection (in.) distance from front of object to Loca t Lon .of Larges tvc ros svsec t icn to plane of shock front (ft.)
any deflection (in.) "
lateral deflection to which a masonry·wall develops no resistance (in.) .'
'. I:
.-}
•.•
•
deflection due to dead load (in.) ,
Xe
elastic deflection (in.) 1
f
equfva l enc elastic deflection (.in.)
Xep
elasto-plastic deflection (in.) . maximum penetration into concrete of armorcpier!,ing fragments (in.) , I
maximum penetration into concrete of fragments other than armor-piercing (in. ) I .;
~
maximum transient deflection ·(in.)
~
plastic deflection (in.)
.._
~
(1) (2)
j I
,
,.,
maximum penetration into sand of armor-piercing fragments (in.) static deflection (in.)
ultimate deflection (in.).
','
,
,
., , ,
*
.'.1'
See' note at end of symbols
4B-18
TM 5-1300/NA1(FAC·P-397/AFR.88-22
(1) (2)
partial failure deflection (in.) deflection at maximum ultimate'resistance. of masonry wall (in.)
, yield line location in vertical direction (Ln, ) *."
y
distance from the top of section to c~ntroid (in.) Z
scaled slant distance (ft./lb. l/3)
ZA
scaled normal 'distance '(ft. /lb,}/3) . scaled ground distance (ft./lb. l/ 3)
~.'
..J'
1
:i
a
Q
(1)
ec
.
,,~
.'
.'
:
angle formed by the plane of stirrups, lacing, or diagonal reinforcement and the plane of the longitudinal reinforcement· ·(deg) angle of incidence of the pressure, front (deg)
(2) (3)
acceptance coefficient
(4)
trajectory angie, (deg'.)
, H":
, ",.:
... ,......
:>,
ratio of flexural stiffness of exterior wall' to .f La t . slab, ratio of flexural stiffness of exterior' wall to slab 'in direction Hand L respectively
.
'.
(1) (2)
8
(3)
(4)
.
J
••
!
•
coefficient for determining elastic and elasto-plastic resistances particular support, rotation angle (deg): .: rejection coefficient target shape factor from figure 2-212
factor equal to 0.85 for concrete' strengths' up to 4 ;000 ps i and is reduced by 0.05 for each 1,000 psi in excess of 4,000 psi
81
j
, ,
'.:.
coefficient for determining elastic and elasto-plastic deflections
y
.. , .
'.\.
Yp
factor'for type of prestressing tendon
6
moment· magnifier
6n
clearing factor
.J
, '.
"-
deflection at sector's displacement (in.)
e' c
average strain rate for concrete (in./in./ms) r' ,
Em
unit strain in mortar (in./in.)
e
average strain rate for reinforcement (in./in./ms)
~ ~
-,
s
.'
Eu
,
.
rupture strain (in./in./ms)
e *
See note at end of symbols
"
TH 5-l300/NAVFAC 'P-397/AFR ,88-22' (1) support rotation angle (deg) , (2) angular. acceleration (rad/ms 2) .,
9
maximum support rotation angle (deg) horizontal rotation angle (deg) * vertical rotation angle (deg) * increase in support rotation angle after, partial failure (deg)
(1) (2)
ductility factor, coefficient of friction "
v
Poisson's ratio
p
(1) (2)
)
mass density (lbs.-~~:2/in.4) density of air behind shock front (lbs/ft. 3'l "
'
,
fracture strength of concrete (psi)-.· Eo
'
effective perimeter ,of reinforcing bars (in.) !':'
. ~'.
summation of moments (in.-lbs.) .
".
~
.' '",
,
,-":10 •
:
sum of the ultimate unit resisting moments acting along the negative yield lines (in.-lbs.) r-. •.) sum of the ultimate unit resisting moments acting along the positive yield lines (in.-lbs.) maximum shear stress in the shaft (psi) (1) (2) (3)
capacity reduction factor bar diameter (in.) , TNT conversion factor
."'1,0
rfJ r
assumed shape function for concentrated loads
rfJ(x)
assumed shape function for distributed loads free edge
'"
angular velocity (rad./ms)
•
,.'
.
r
*
See note at end of symbols
48-20·
...;','
TN 5-l300/NAVFAC P-397/AFR 88-22
simple support
IIIIIII
fixed support either fixed, restrained, or simple ,support
* Note. This symbol was developed for two-way elements which are used as walls. When roof slabs or other horizontal elements are under consideration, this symbol will also be applicable if the element is treated as being rotated into a vertical position. 4B-2l
TM 5-1300/NAVFAC P-397/AFR 88-22
APPENDIX 4C BIBLIOGRAPHY
TK 5-l300/NAVFAC P-397/AFR 88-22
1.
2.
3. 4. 5.
6.
7.
8. 9.
10. 11. 12. 13. 14.
15.
American 'Concrete Institute. Design Handbook. Volume 1 - Beams. One-Way Slabs. Brackets. Footings. and Pile Caps, in accordance with the Strength·Design Method of ACI 318-83. Publication ~P-17(84).,ACI' Committee 340, American Conc reee Institute, Detroit, MI,1984. American Concrete Institute. Design Handbook: Volume 2 - Columns" in accordance with the Strength Design Method of ACI 318-83, Publication SP-17A(85),.ACI Committee 340, American Concrete Institute, Detroit, MI, 1985. l American Concrete Institute, Design Handbook, Vol\une2 - Columns, Publication SP-17A, ACI Committee 340, American.Concrete Institute'" Detroit, MI, 1978. '. American Concrete Institute Standard Building Code Requirements for Reinforced Concrete, ACICommittee 318-77. American Concrete Institute, Detroit" MI, August, ,1979. , , American Concrete Institute Standard Building Code Requirements for Reinforced Concrete, ACI Committee 318-83, American Concrete Institute, Detroit; ,MI, November 1983; . Design Criteria for Deflection Capacity of Conventionally Reinforced Concrete Slabs, an investigation conducted by Construction Technology Laboratories; Structural Analytical Section, 5420 Old Orchard Road, SkokLe-, Illinois 60077, for Civil Engineering Laboratory .. Naval ConstructionBattalion.Center, Port Hueneme, CA, sponsored by. Naval Faciliti~s .Engd.ne e r Lng Command, Phase I -, State-of-the-Art R';port, CR 80.026, October,1980, Phase-II - Design and Construction Requirements, CR 80.027. .Oc cobe r 1980, Phase III - Summary ,of Design Criteria and Design and Construction Details - Design Examples, OR 80.028, October 1980. Design Manual. AEC Test Structures ,: Volumes II and III,' by 'Holmes and Narver, ,Inc., under Contract AT (29-,2)-20, for U.S. Atomic Energy Commission, Albuquerque Operations Office, Albuquerque, NM, December 1961. Design ,of Structures to Resist Nuclear Weapons Effects, ASCE Manual: of Engineering'Practice No. 42, American Society of Civil Engineers, New York, NY, 1961. ; .:,.. Design of Structures to Resist the Effects of Nuclear Weapons, Department of the Army ,Technical Manual TM 5-.856-2, 3 and 4, Washington, D.C. Dynamic, Tests of Concrete Reinforcing Steels, Technical ,Report R,394, U. S. 'Naval Civil Engineering Laboratory.., Port, Hueneme, CA, September 1965. ' Effects of Impact and Explosion, yolume 1, Office of Scientific Research and Development, National Defense Research Committee', Washington, D.C., 1946 (ConfIdent.La'l ) . .' Fundamentals of Protective Design (Non-Nuclear), Department of the Army, Technical Manual TM 5-855,1, Wa~hington, D.C., July 1965. Impact Design Criteria 'for Blockhouses Subjected to Abortive Launch Environments, Technical Publication PMR-TR-70-2, Mechanics Research, Inc., for the' Pacific Missile Range, Point Mugu, CA, July 1970. Industrial Engineering Study to'Establish Safety Design Criteria for Use in Engineering of Explosive Facilities and Operations », by Ammann & Whitney Consulting Engineers, ,New York, .,NY, under Contract' DA-28-0l750l-0RO- 3889, for Process Engineering Branch, A. P.M. E;,D., Picatinny Arsenal, Dover, NJ, ,April 1963. Manual of Standard Practice for Detailing Reinforced'Concrete Structures, ACI,Committee 315-65, American Concrete Institute, Detroit, MI, 1965. 4C-l
"
TH 5-l300/NAVFAC P-397/AFR 88-22
16.
17.
18. 19. 20. 21.
Non-Nuclear Weapons Effects on Protective Structures (U>. Technical Report No. AFWL-TR-69-57, (SECRET), Mechanics .Re se arch , Inc. ".for the Air Force Weapons Laboratory, Kirtland Air Force Base, NM, 'September 1969. • Strength and Behavior of Laced Reinforced Concrete Slabs· Under Static and Dynamic 'Load, Technical Report R620, prepared by ,Naval· Civil). Engineering Laboratory, Port Hueneme; California, Sponsored by'Department of the Army, Picatinny Arsenal, April 1969. . Summary of One-Third Sc,lle Slab Tests and Scaled Bay Structure ·Tests. 'by Ammann & WlJitney, Consulting Engineers', New York, NY, under Contract DA28-0l7-AMC-423(A) for Picatinny Arsenal, Dover, NJ, March 1969'; Ultimate Strength Design Handbook, Volume No. I, .Special Publication ·No. 17. ACI Committee 340, American Concrete Institute, Detroit, MI,.1967.· Ultimate Strength Design of Reinforced Concrete Columns, Publication SP7, Interim Report of ACI Committee 340, American Concrete Institute,,' Detroit, MI, 1964. r • ' U.S. Army Engineer Waterways Experiment Station, Technical Report N-7210, Design and Testing of a Blast-Resistant Reinforced Concrete Slab ' System, Criswell, M.E;'; Vicksburg, MS, 1972. '"
22.
·
·•
" .'-.
Amirian, A... Dobbs, N. and Dede, M., Technical Approach' for' Structural Design Procedures for a'Family of Horseshoe-Shaped MTC {Phase I>" prepared for Naval Civil Engineering Laboratory; Port Hueneme, CAby Ammann and Whitney, Consulting Engineers, New York, N·. Y';, August 1985, 23. Ayvazian, H.' and DeDe, M., Development of Preliminary Designs for ,NAVFAC MTCs and Preparation of: Construction' Cost Estimates,' prepared for Naval Civil Engineering Laboratory, Port Hueneme, CA by Ammann and Whitney,' Consulting Engineers, New York, N.Y., Draft. ' . 24. Ball, J. W., Baylot, J.T. and Kiger, S.A., Ki'rtland Underground'Muni-' tions Storage Complex Model Designs, Construction. and Test' Data', Technical Report SL-84-l4, prepared by. Structures Laboratory, Department of the Army, Waterways Experiment Station, Corps of Engineers, P. o.. : Box 631', Vicksburg, Mississippi 39180-0631, for Defense Nuclear ,Agency, Washington', ·D.C. 20305, Under DNA Subtask B990AXRE, 'Work Unit 0004 and Subtask B99QAXRD, Work Unit 0013, "Advanced Design Explosive Tests'," . September,1984. 25. Baylot, J. T., Kiger, S. A. and Ball, J. Wi·..· Structural Design..ofdBlastContainment Facilities, Abstract for the Twenty-First Exp Los Ive-Bafe ty Seminar, prepared by U.S. Army·.Engineer Waterways·Experiment Station, Vicksburg, Mississippi, August 1984. 26 . . Baylot, J.T., Vulnerability of 'an Underground Weapon Storage Facility, Technical Report SL-84-l6, prepared by Structures Laboratory" Department of the Army,. Waterways Experiment Station, Corps. of Engineers .... P.O. Box 631, Vicksburg, Mississippi '39180,00631',. for Defense Nuclear ..Agency, Washington, DC 20305, Under Subtask A99QAXFC, Work Unit 00046 .. September 1984. . 27. Bergstrom, P., et al., Hard-Structures Munitions -Phase I; Warehead {U>, AFATL-TR-67-13, Volume III (AD-379, 468) .. '(SECRET), Chamberlain Cor-poration, ,February 1967. . .v : . . .' .. 28. Beth, R. .·A., Final Report on Concrete Penetration,. Report No. ;A"388, National Defense Research Committee, 'Office of Scientific Research and Development, March 1946. ". 29. Beth, R. A. and Stipe, J. G., Penetration and Explosion Tests on ' Concrete Slabs, Report I;· Data, Interim Report No, 20, Committee on . Passive Protection Against Bombing, National Research Council, January 1943. 4C-2
TK 5-l300/NAVFAC P-397/AFR,88-22 30. 31. 32.
33.
34.
35.
36.
37.
38.
39. 40.
41.
Branson, D., Deformation of Concrete Structures, McGraw-Hill, Inc., New York, NY, 1977. ' ' . Brown, J. W., Response of Selected Materials to High-Speed Fragment Im~,U:S. Army Engineer 'Waterways Experiment Station,Vi~ksbur~.. MS, 1970. Cohen: E. and Dobbs, N., Design Procedures and Details for Reinforced Concrete Structures Utilized in Explosive Storage and Manufacturing' Facilities, Ammann & Whitney, Consulting Engineers, New York, NY, Annals 'of the New York Academy of Sciences .. Conference on Prevention of and Protection Against Accidental Explosion of Munitions , FUels and Other Hazardous Mixtures, Volume 152, Art.-l, October 1968. Cohen, 'E. and Dobbs, N.,' Models for Determining the Respo~se of Rein' ,forced Concrete Structures to Blast Loads, Ammann & Whitney, Consulting Engineers, New York, NY, Annals of, the New York Academy of Sciences, Conference on Prevention of and Protection Against Accidental Explosion of Munitions, Fuels and Other Hazardous Mixtures, Volume 152, Art. I, October 1968. ' , Cohen,' E. and Dobbs, N., Supporting Studies to Establish Safety Design Cri'teria for Storage and Processing of Explosive Materials - Inte'rim Report No, I, Summary of One-Third Scale Reinforced Concrete Slab Tests, Technical Memorandum Col, by Ammann & Whitney, Consulting Engineers, New York, ,NY,' under Contract: DA-i8-017 -AMC-423 (A) for Picatinny Arsenal, Dover, NJ, June 1965. ' , Coltharp"D.R., Vitayaudom, K.P. ' andKfge r , S.A., NATO Semiha'rdened Facility Design Criteria Improvement, Final Report, prepared by Engineering and Services, Laboratory, Air Force Engineering and Services Center, Tyndall Air Force Base, Florida 32403. June 1985. Criswell, M.E., 'Design and Testing of Ii Blast-Resistant Reinforced' Concrete'Slab System, Technic~l Report N-72-l0, Conducted by U.S. Army Engineer Waterways Experiment Station, Weapons Effects Laboratory, Vicksburg, Mississippi, sponsored by Defense Civil' Preparedness Agency, Work Order No. DAH C20-68-W-0192, Work Unit l127E, November 1972. Dede; M. and Amirian, A., Test Plan to Verify Failure C'riteria for Conventionally Reinforced Concrete Slabs, prepared for Naval Civil Engineering Laboratory, ,Port Hueneme, California, by Ammann and Whitney, Consulting Engineers, New York, N. Y." April,'1985. ' Doyle, 'J. M., Klein, M. J. and Shah, H., Design of Missile Resistant Concrete Panels, Preprints of the'2nd International Conference on Structural Mechanics in Rea2tor Technology, Vol. 4, Commission of the European' Communities, Brussels,' 1973, Paper No. J 3/3.. ' " Fisher, E. M., Explosion Effects Data Sheets', Report 2986, U. S. Naval Ordnance Laboratory" Silver Spring, MD, 1955. ' , Getchell,"J.V. and Kiger, S.A., Vulnerability of Shallow-'Burried FlatRoof Structures, Technical Report SL-80-7, prepared by Structures Laboratory, ,U.S. Army Engineer Waterways Experiment Station, P.O. Box 631, Vicksburg, Mississippi 39180, for Defense Nuclear Agency, Washington, D.C. 20305 and Office, Chief of Engineers; U.S. Army, Washington, D.C. 20314, under DNA Subtask Y99QAXSC062, Work Unit '42. and Subtask H19lAXSX337, Work Unit 02, and OCE R&D Project 4A7627l9AT40, Task Ad, Work Unit 008, Report 2, Foam HEST 4, October ,1980, Report '3" Foam HEST 5, February 1981, Report 4, Foam HEST 3 and 6, December 1981 and Report 5, Foam HEST 7, February 1982. Gewaltney, R. C>, Missile Generation and' Protection in Light-WaterCooled Power Reactor Plants, Nuclear Safety, 10(4);, July-August 1969. 4C-3 .
TK 5-l300/NAVFAC P-397/AFR 88-22 42. 43.
44. 45.
46. 47. 48.
49.
50. 51.
52. 53. 54.
Giere,.A. C., Calculating Fragment Penetration and Velocity Data for Use in Vulnerability Studies, NAVORD Reoort 6621, U.S. Naval Nuclear Evaluation Unit, Albuquerque, NM, October 1959. . Hoffman, P. R., McMath: R.; R. and·Migotsky:·E., Projectile Penetration Studies AFWL Technical Report No. WL-TR-64-102, Avco Corporation 'for the Air Force Weapons Laboratory, Kirtland Air Force Base, NM, December 1964. .. . Ingraham, J. M. and Abbott, K. N. ,Feasibility Study of Laminated Metallic Armor (U), AMRA TR 63.-30, '"(CONFIDENTIAL), U: S. Army M,\~erials Research Agency, Watertown, MA, December 1)163. Iqbal, M. and Derecho, A.,; Design Criteria for Deflection Capacity of Conventionally Reinforced Concrete Slabs. Phase I - State-of-the-Art .Report , Report No. CR'80.026, Naval.Civil Engineering Laboratory,. port Hueneme, CA, October 1 9 8 . 0 . " . .' . ' .Johnson, C. .and Moseley, J .W., Preliminary Warhead Terminal Ballistic Handbook. Part 1 : Terminal Ballistic Effects, NAWEPS Report No. 7673, U.S. Naval Weapons Laboratory, Dahlgren, VA, March 1964. Keenan, W. A., Strength and Behavior of Laced Reinforced Concrete Slabs Under Static and Dynamic Load, Technical Report R-620, U.S. -Nava L ,Civil Enginee~ing Laboratory, Port Hueneme, CA, April 1969. Keenan, W.·, Timcreto, J., Meyers, G" Johnson, F., Hopkfris , J., ,", NickersoI!, H. and Armstrong, W., NCEL Products Supporting bob Revision of NAVFAC P397, Technical Memorandum 2591TM, prepared by Naval Civil. Engineering Laboratory, Port Hueneme, California 93043, Program No.: Y0995-01-003-201, Sponsored by Nav~l Facilities Engineering Command, Alexandria, Virginia 22332, March 1983. Kiger, S.A. and Albritton, G.E., Response of Burried Hardened Box Structures to the Effects of Localized Explosions, Technical Report SL80-1, prepared by Structures Laboratory,.U.S. Army Engineer'Waterways Experiment Station, P.O.' B~x 631, Vicksburg. Mississippi 39180; for. Defense Nuclear Agency, Washington, D.C. 20305 and Office, Chief of Engineers, U. S. Army, Washington,. D.C, 20314, under DNA Subtask Y99QAXSC062, Work Unit 04; and'OCE R&D ,Project 4A762719AT40, Task AI, Work,. Unit 027, March 1980.. ." , .' . Kymer, J. R.,.Penetration Performance 'of Arrow Type ProJectiiesi.Report R1814,U.S. Army Frankford .Ar senaL, Philadelphia, PA, May 1966.• . . . Liu, T.C., Strength Design of Reinforced Concrete Hydraulic Structures Preliminary Strength. Design Criteria, rechnical Report,SL-80-4, Report 1 of a series, prepared by Structures Labor atory '. U. S. .Army Eng Lnee r Waterways Experiment Station, P.O. Box 631, .Vicksburg, Mississippi 39180, for Office, Chief 'of Engineers, ,1!.S. Army, Washington, D.~. 20314, under CWIS 31623. July 1980: ' Mascianica, F. S., Summary of Terminal Ballistic Data on Lightweight Armor Materials (U), AHHRC TR 69-17: (CONFIDENTIAL), U.S. Army Materials and Mechanics Research Center, Watertown, MA, July 1969. Nara, H. R. and Denington, ,~. J. Ricochet and Penetration of Steel' Spheres in Clay and Sand Soils, TM-E7; AD-059 883, Case Institute of Technology, Project DOAN BROOK, October 1954. , Newmark; N. M. and Haltiwanger, ,J. D. ,Air Force Design ManuaL Princi~ pIes and Practices for'Design of Hardened Structures, Technical'Documentary Report No. AFSWC-TDR-~2-138, by Department of Civil Eng~neering. University of Illinois, Urbana, Ll.Li noLs , under pontract AF29 (601) -2390 for· Research Directorate, Air Force'Special Weapons Center, Air Force Systems .CQmmand, Ktrtland Air Force Base, NM, December 1962. 4C-4 '.
TK 5-l300/NAVFAC P-397/AFR 88 c22
55. 56.
57.
58. 59.
60. 61.
62. 63.
64.
65.
66.
67. 68. 69.
Park, R. and Gamble ," W. L'., Reinforced- Concrete Slabs, John Wiley & •. Sons', New York, 'NY; 1980. ,-',' . Pecone, G, ,.Design of Flat Slabs Subjected to' Blast Loads, prepared for -Northern Division, Naval Facilities Engineering Command, U.S. Naval Base, Philadelphia"PA, by Ammann and Whitney, Consulting Engineers,' New York,-N.Y., Contract No. N 62472-76-C-1l48: June 1982: . . Recht, ,R., et a1., Application of' Ballistic Perforation Mechanics to Target Vulnerability and Weapons Effectiveness Analysis (U), NWC TR 4333 (Confidential), Deriver Research'Institute forthe.Naval Weapons Center, 'China Lake, CA, "October 1967. -; . .. . . Richey, H." M., Armed Services Explosives Safety Board Dividing Wall'Program, Phase C, Tests 1 through 13. U.S, Naval Weapon Center, China Lake, CA, 1964 (Secret Restricted Data). Rindner, R'-,M." Establishment of Safety Design' Criteria' for Use in Engineering of Explosive Facilities'and'Operations, Report No, 2, Detonation by Fra'gment Impact, Technical Report DB-TR: 6'-59, Picatinny Arsenal, Dover, NJ', March 1959. Robertson, H.P., Terminal Ballistics, Preliminary Report, Committee on 'Passive Protection Against Bombing, National Research Council, 'January 1941. ,.ii" . Rohani, B., Fragment and Pro·J ecti·le Penetration Resistance of Soils: Report 2. High-Velocity Fragment, Penetratiorr into Laboratory-Prepared ;'Soil Targets,' Miscellaneous Paper S-'7l-l2,U.S. Army Engineer' Waterways Experiment Station, Vicksburg, MS, 'June 1973.. " -St Lpe , J'. G.,. e't a-l., .Ballistic, Tests on Small' Concrete Slabs, .Lnt er Im Report No. 28, Committee on Fortification Design, Nationa-l Research Council, June 1944. Schumacher, R.N., Kingery, C.N. and Ewing, Jr., W.O., Airblast and Structural Response Testing of a 1/4 Scale Categorv 1 Suuuressive Shield, BRL Memorandum Report No. 2623, by U.S.A. Ballistic Research Laboratories, Aberdeen Proving 'Ground, Maryland, May 1976. Schwartz, A., et. al., Establishment of Safety Design Criteria for Use in Engineering of Explosive Facilities and Operations, Report No.6. Test of One-Tenth Scale Bay Structure, Technical Report 3439, by Picatinny Arsenal, Dover, NJ, and Ammann & Whitney, Consulting Engineers, New York, NY, under Contract DA-28-0l7-AMC-423(A), July·1966. Takayahagi, T. and Derecho, A., Design Criteria for Deflection Capacity of Conventionally Reinforced Concrete Slabs, Phase III - Summary of Design Criteria and Construction Details - Design Examples Report No. CR 80.028, Naval Civil Engineering Laboratory, Port Hueneme, CA, October 1980. Takayanagi, T., Derecho, A. and Iqbal, M., Design Criteria for Deflection Capacity of Conventionally Reinforced Concrete Slabs, Phase II Design and Construction Requirements, Report No. CR 80.027, Naval Civil Engineering Laboratory, Port Hueneme, CA, October 1980. Tancreto, J., Design Criteria for Deflection Capacity of Conventionally Reinforced Concrete Slabs,: Naval Civil Engineering Laboratory, Special Project 1311 SP, Port Hueneme, CA, 1980. Thomas, L.H., Computing the Effect of Distance on Damage by Fragments, Report No. 468, Ballistic Research Lab9ratories, Aberdeen Proving. Ground, MD, May 1944.' Timoshenko, S. and Woinowsky-Krieger, S., Theory of Plates and Shells, McGraw-Hill Book Co., New York, NY, 1959.
4C-5 '..
TM 5-l300/NAVFAC P-397/AFR 88-22 70.
71.
72.
73. 74. 75. 76.
77.
Vanderbilt, 'H., Sozen, H. and.Be Lss., G., Deflections of Hultiple-Panel Reinforced Concrete Floor Slabs, Journal of the'Structural Division" Proceedings of the ,American Society of Civil Engineers, August 196?, ' Vargas's, L. H.,'Hokanson, J. C."and Rf.nde r ," R. H., Explosive Fragmentation of Dividin~ Walls, by Southwest ,R~search Institute, San Antonio, TX, Contractor Report ARLCD-CR-8l0l8, U:S. Army Armament Research and, Development Command, Large Caliber Weapons Systems Laboratory, ,.Dov"r,' NJ; July 1981. " , "" . Wachtell, 5., Comparison of. Blast Response-Scaled VB. Full-Size Concrete
Structures, Picatinny Arsenal, Dover, NJ, Annals of the New York Academy of Sciences, Conference-on Prevention ,of and Protection Against Accidental Explosi~ns of Hunitions, Fuels'an~ Other Hazardous Hixtur~s , Volume 152, Art. 1, October 1968. ' Wang, Chu-Kia and Salmon, Charles G., Reinforced Concrete Desi~n, Intext Educational Publishers, New York, NY, 1973. Whitney, C. S., Plastic Theory of 'Reinforced Concrete Desi~n, Tran~a~ tions of the American Society of Civil Engineers, Vol. 107, ASCE, New York, NY, 1942. . ' ", Whitney, C. S. and Cohen", E., Guide ,for Ultimate Stren~th Desi~n 'of Reinforced Concrete, Ammann & Whitney, Consulting Engineers, New York, ,NY, Journal of the American -Conc r'e tie Institute , November 1956'. Whitney,' C. S., e"t. a1. ,:Desi~n of Biast~Resistant' Construction for Atomic Explosions, Ammann & Whitney, Consulting Engineers, New York"NY, Journal of the American Concrete Institute, Harch 1955. Wood, R. W., ,Plastic and E1B.stic Design 'of Slabs and Plates .. The Ronald Press Company, New York, NY, 19,61-.
,
. .-
";
"
~.
,
','"
'\
,. -.
".:
'
4C-6
'''STRUcruRES TO RESIST THE EFFEcrs OF ACCIDENTAL EXPLOSIONS"
CHAPTER 5. STRUCTURAL STEEL DESIGN .
TK 5-l300/NAVFAC
P-3~7/AFR
88-22
. CHAPTER 5 STRUCTURAL STEEL DESIGN INTRorUCTION 5-1. Purpose.
The purpose of this manual is to present. methods of design for protective construction used in facilities for development, testing, production, storage, maintenance, -modification, inspection, demilitarization, and disposal of explosive materials. 5-2. Objective r.
The primary objectives are to establish design procedures and construction techniques. whereby propagation. of explosion (from one sl:ructure or part of a structure to another) or mass detonation can be prevented and to provide protection for personnel and valuable equipment. The secondary objectives are to: (1) (2)., (3) (4)
Establish the blast load parameters required for design of protective structures. Provide methods for calculating the dynamic response of structural elements including reinforced concrete, and structural steel. Establish construction details and procedures necessary to afford the .. required strength to resist the applied blast .Loads , Establish guidelines for siting explosive facilities to obtain maximum cost effectiveness in both the planning and structural arrangements, providing closures, and preventing damage to interior portions of structures'because of structural motion, shock, and fragment perforation.
5-3. Background For the first 60 years of the 20th century', criteria and methods based upon results of.catastrophic events were used for the design of explosive facilities . . The crfteria and methods d~d not include. a detaile".or.·reliable quantitative bas Ls for assessing the degree of protection afforded by' the protective facility. ,In the late 1960's quantitative procedures were set forth .in the first edition of the present manual, "Structures to Resist the Effects of Accidental Explosions". This manual was based on extensive research and development. programs which permitted a more' reliable approach to current and future design. requirements. Since the original publication of this manual, more extensive. testing and development programs have ~aken place .. This additional research included work with materials other than reinforced concrete which was the principal construction material referenced in the initial version of the manual. Modern mathods-Tor the manufacture. and storage of explosive materials, which include ~any exotic chemicals, fuels, and. propellants , require less space for a given quantity of explosive mace r LaI than .was . previously needed. Such concentration of explosives increases the possibility of the propagation of accidental explosions. (One accidental e~plosion causing the detonation of 5-1
TM 5-1300/NAVFAC'P-397/AFR 88-22
other explosive materials.) It'is evident that a requirement for more accurate design techniques is essential. This manual describes rational design methods to provide the required structural protection. These design methods account for the close-in effects of a detonation includ" ing the high pressures and the nonuniformity of blast loading on protective structures or barriers. These methods also account for intermediate and farrange effects for the design of structures located away from the, explosion. The dynamic response of structures, constructed of various materials. ,or
combination of materials, can be calculated, and details are given to provide the strength and ductility required by· the design. The design approach is directed primarily toward protective structures subjected to the effects of a high explosive detonation.
However, this approach is general, and it is
applicable to the design of other explosive environments as well as other explosive materials as mentioned above.
The design techniques set forth in this manual are 'based upon the results of numerous full- and small-scale: structural response and explosive effects tests " of various materials conducted in conjunction with the development of this, manual and/or related projects. 5-4. Scope It is not the intent of this manual to establish safety criteria. Applicable documents should be' consulted for this purpose. Response predictions for personnel and equipment are included for information. . .• In this manualan'effort is However, sufficient general been included in order that situations other than those
made to cover the more probable design situations. information on protective design techniques has application of the basic theory can be made to which were fully considered,
This manual is applicable to the design of protective structures subjected to the effects associated with high explosive detonations. For these design situations, the manual will apply for explosive quantities less than 25,000 pounds for close-in effects. However, this manual is also applicable to other situations such as far- or intermediate-range effects. -For these latter 'cases the design procedures are applicable for explosive quantities·in·the order of' 500,000 pounds which is the maximum' quantity of high explosive approved,for aboveground storage facilities in 'the Department of Defense manua l , "Ammunition and Explosives Safety Standards", DOD 6055·.9-STD. Since tests were primarily directed toward the response of structural steel and reinforced concrete elements to blast overpressuresi this manual concentrates· design procedures and techniques for these materials. However, this does not imply that concrete and steel are the.only useful materials for protective construction. Tests to establish the response of wood, brick blocks, and plastics, as well as the blast attenuating and' mass effects of soil are contemplated. The results of these tests may require, at a later date, the supplementation of these design methods for these and other materials. '
on
Other manuals are available· to design protective structures against the effects of high explosive or nuclear detonations. The procedures in these manuals will quite often complement this manual and should be consulted for' specific applications. 5-2
TM 5-1300/NAVFAC P-397/AFR 88-22
Computer programs, which are consistent with procedures and techniques contained in the manual, have been approved by the appropriate representative of the US Army, the US Navy, the US Air Force and the Department of Defense Explosives Safety Board (DDESB). These programs are available through the following repositories: (1)
" Commander and Director U.S.'Army Engineer Waterways Experiment Station Post Office Box 631 Vicksburg, Mississippi,39l80-063l Attn: WESKA
(2)
Department of the Navy Commanding Officer Naval Civil Engineering Laboratory Port Hueneme, California 93043 Attn: Code LSI.
( 3)
Depar cment, of the Army
.~,
Department of the Air Force Aerospace Structures
Information and Analysis Center Wright Patterson Air Force Base Ohio 45433 Attn: AFFDL/FBR If any modifications to these programs are required, they will be submitted for review by DDESB and the above service~. Upon concurrence of the revisions, the necessary changes will be made and notification of the changes will be made by the individual repositories. 5-5. Format This manual is subdivided into six specific chapters dealing with various aspects of design .. The titles of these chapters are as follows: Chapter Chapter Chapter Chapter Chapter Chapter
I 2 3 4 5 6
Introduction Blast, Fragment, and Shock Loads Principles of Dynamic Analysis Reinforced Concrete Design Structural Steel Design Special Considerations in Explosive Facility Design
When applicable, illustrative examples are included in the Appendices. Commonly accepted symbols are used as much as possible. However, protective design involves many different scientific and engineering fields, and, therefore, no attempt is made to standardize completely all the symbols used. Each symbot' is 'defined where it is first used, and in the list of symbols at the end of each chapter.
5-3
TM 5-l300/NAVFAC P-397/AFR 88-22 CHAPTER CONTENTS 5-6. General This chapter contains procedures and guidelines for the design of blastresistant steel structures and steel elements. Light construction and steel framed acceptor structures provide an adequate form of protection in .a. pressure range of 10 psi or less. However,if' fragments are present, lightgage construction may only be partially appropriate. The'use of structural steel frames in combination with precast concrete roof and wall panels (Chapter VI) will provide a measure of fragment protection at lower pressure ranges. Containment structures or steel elements of containment structures, such,as blast doors, ventilation closures, fragments shields, etc. can be designed for almost any pressure range. This chapter covers detailed procedures and design techniques for the blast-resistant design of steel elements and structures subjected to short-duration, high-intensity blast loading. Provisions for inelastic, blast-resistant design will be consistent with conventional static plastic design procedures. 'Steel elements such as beams, beam columns, open-web joists, plates and cold-formed steel panels are considered. In addition, the design of steel structures such as rigid' frames, and braced frames are presented as they relate to blast-resistant design. Special considerations for blast doors, penetration of fragments into'steel, and.unsymmetrical bending are also presented.
1
5-4
TH 5-1300/NAVFAC P-397/AFR88-22
, STEEL STRUCTURES IN PROTECTIVE' ,DESIGN.
, .'
'.;'~,
5-7. Differences Between Steel.and Concrete:Structures in Pr~tective Design Qualitative differences between steel and concrete protective structures are summarized below:
(];)"
~
I.
t"'
'1:
In close-in high-impulse, design' situations where a containment
• ~.-
'.
: " structure is. utilized, - 8· massive.. reinforced concrete, struct.ure ,ft, J.rather ,than a steel structure, is generally employed in o rde rvco
" •'limit"deflections and to offer protection. against the effects of . pr Imary.vand: secondary fragments. ; However , elements of containment 'structures such as blast 'doors', ventilation c Loau're s i v e t.c , , .are generally ,designed .us Lng structural st ee L. Fragment protection Ls.> ,,' -,;usually accomplished by increasing the element thickness to resist '. fragment'penetration or by providing'supplementary fragment 'protection._ In some cases, structural steel can'be·used in the' ,design of containment cells. However;oexplosive charge weights are generallY"low; f thereby preventing brittle, modes of' failure, (Section 5 -18.3) due to high pressure intensity.",
_ I.
Structural steeL shapes are, considerably .mo re slender ,bothin
(2)
terms 'of the overall structure and the components of a typical
member' cross section: As a result: the effect of overall and, local instability upon the ultimate capacity is an important consideration 'in steel des Lgn ,
Moreover
I
in most cases. 'plate
eiements1and structures wili:sustain large deformations'in comparison to those of more rigid 'concrete elements.
:,
..,,,
(3), The'amount of rebound in concrete structures is considerably " reduced by internal damping (cracking) and is essentially elimi· nated in cases where large deformations or incipient failure are .... .permitted.to occur: In structural steel,however,. a larger response in rebound, up to 100, percent, can- be obtained for a combination'of short duration:load and a relatively' flexible element .• As a' result, steel structures require that . spec'LaL provisions be made to account for extreme responses of comparable magnitude in both directions. ,t
•
-cc
The •treatment of stress interaction is:, more of a consideration' in ;.'steel shapes since each' element' of the cross section' must be considered' subject to a state of combined stresses. In reinforced concrete t~' ·the provision of· separate steel reinforcement for flexuie,i shear and torsion.enables the designer to consider these stresses as being carriedby'more or less independent systems .
(4).
.... l
(5)
-
.r
~.
,'Special care must be taken in steel design to provide for connec· tion 'integrity up to the point of maximum r e spons e ,. For example. in order to'avoid premature brittle fracture in welded connections, the welding characteristics, of the'particular grade of steel must be considered and the introduction of any stress concentrationsat'joints and notches in main elements must be -avo Lded ,>" . ,', ,J
"
.,
I
_',
5-5
..
,
TM 5-l300/NAVFAC P-397/AFR 88-22
(6)
If fragments are involved, special care must'be given to brittle modes of failure as they affect construction methods. For example, fragment penetration depth may govern the thickness·of·a· steel plate.
5-8. Economy of Design of Protective Structures in the Inelastic Range
......
.. ,.'
The economy of facility design generally .requt resvnhac blast-resistant- structures be designed·to perform in:the inelastic response range during an. accident. In order to ensure the structure~s integrity throughout such severe conditions, the facility designer must be cognizant of. the various possible failure modes: and-xhe l r inter-relationships. The limiting design values ·are dictated by the attainment of inelastic deflections and rotations/without complete.collapse . . The amount of inelastic deformation is dependent not only upon the ductility characteristics of the material, but also upon. the.intended use of the structure following an accident as well as the protection required. In order for the structure to maintain such large deformations, steps must be taken to'prevent premature failure by either brittle fracture or instability (local or overall). Cuidelines and criteria for dealing with these effects. are presented in the body of this chapter. 5-9. Applications of Steel Elements and Structures In Protective Design - ...... >
The design procedures and applications of this chapter are directed· toward steel acceptor-. and donor-type structures,'. ,. •',
,
I
Acceptor-type. struc.tures are removed' from the immediate vicinity of the detonation. These include typical frame structures with beams, columns and beam-columns composed of standard structural shapes, and built-up sections. In many cases, t4e relatively low blast pressures suggest the use of standard building components such as open-web joists,. prefabricated wall panels' and roof decking detailed as required to carry.the full magnitude of' the dynamic l~ads. Another economical application can be the use of entire pre. engineered buildings, strengthened locally, to adapt their designs to low blast pressures· (up to 2 psi) with· short duration. For guidelines on the blast evaluation of pre-engineered buildings " see. "Special Provisions for Pre-engineered Buildings", Chapter. VI. .. Donor-type structures, which are located in the immediate ~icinity of the detonation may include steel containment cells or steel. components'of reinforced concrete containment structures such as .blast doors or ventilation and electrical· closure plates. In some c~ses, the use of suppressive shielding to control or confine the hazardous blast, fragment, and flame.effects of detonations may be an economically feasible .alternative.·' A brief review of suppressive shield. design. and criteria is outlined in Section 6-23 to. 6-26 of Chapter VI. The high blast pressures encountered in these structures suggest the use of large plates or built-up sections with relatively high resistances. In some instances, fragment impact or pressure leakage is permitted . ... .
5-10. Application of Dynamic Analysis. The first step in a'dynamic design entails the development of a trial design considering facility requirements, available materials, and economy with' members sized by a simple preliminary procedure. The next step involves the performance of a dynamic analysis to determine the response of the trial 5-6
/.
TH 5-l300/NAVFAC P-397/AFR 88-22 design to the blast and the comparison of the maximum response with the 'deformation limits specified in this chapter. The ,final design is then determined by achieving an economical ,balance between sti:ffness and resistance such that the calculated response under the, blast loading lies within the limiting.,values dictated by:the operational requirements of the facility. The dynamic response calculation involves,either,a.single-degree-of-freedom analysis using the 'response, charts "in"Chapter. 3, or, in more complex structures, a multi-degree-of-freedom.analysi~·using,availabl" dynamic elastoplastic frame programs. .,~ . . . . <.
A single-degree-of-freedom analysis may be performed for the. design analysis of either a given structural element or of an element for which a preliminary design has been performed according to procedures given in. this chapter. Since this type. of.dynamic analysis is described fully with accompanying charts and tables in 'Chapter 3 i it .will not becdup Li.ca t ed herein . . In principle, the structure'or'structural element is characterized by.an ideat"ized,' bilinear, . elasto-plastic. resistance function and, the loading is' treated as' an. idealized. triangular (or. bilinear) 'pulse with zero, rise.,time (Chapter 3). Response charts are presented. in Ch~pter 3 for 'determining the. ratio of the maximum response to the elastic response and, the time·to reach maximum response·for the initial response; The equations presented for the dynamic reactions are also.applicable to this chapter. Multi-degree-of-freedom, nonlinear dynamic analyses' of braced,'and unbraced rigid frames can be performed'using programs available through.the repositories listed in Section.5-4 and i.tihr ough the reports listed in the bibliography at the end of this. chapter. " .
••
.,
;
t •~
, 1\;
5-7
s:
TM 5-r300/NAVFAC
P-397/AFR88~22
PROPERTIES OF STRUCTURAL STEEL 5-11. General
,
('
Structural steel is known-to be a strong and ductile building:material~ The significant engineering properties of steel are strength expressed in terms ,of yield stress and ultimate tensile strength, ductility expressed in terms of, percent elongationat'.rupture, and rigidity expressed in terms of modulus of
-,,~
elasticity._ This section covers the mechanical properties of structural steel
subjected to static loading and dynamic loading. stresses for bending and shear are then derived. admissible 'in plastic design ·arelisted. .' 5-12. Mechanical Properties -
" :~I.
..
,.~ ~.
;
5-12.1., Mechanical Properties Under Static Load Ing ; Static Design Stresses'
.
r
.
Structural steel generally can be considered as exhibiting a linear stressstrain relationship.up,to the proportional limit, which is either'close'to or identical to the yield point. ,Beyond the~yield'point, it can stretch substantially without.appreciable·increase in stress, the amount of elongation' 'reaching 10 to 15 times that needed to reach yield, 'a. range that is termed II
the yield plateau".
Beyond that .. range
I
'./
Recommended dynamic design' Structural steels that are
.
strain hardening. occurs. i. e .. "
additional elongation is associated with an increase in stress. reaching a.maximum nominal stress called ~the tensile strength lt
'After ,
a drop in. the'
nominal stress accompanies further elongation and precedes' fracture at an . elongation (at rupture) amounting to'20 to 30 percent.of the specimen's original length (see Figure 5-1). It is this ability of structural steel
...
TH 5-l300/NAVFAC P-397/AFR"SS-22
the increase factor (a), should not be applied to high strength steels since the average 'increase may be less than 5 percent. The minimum yield stress, f y; and the tensile stress, f u' (minimum) for structural steel shapes and plates which conform to the American Society for Testing and Materials (ASTM) Specification are listed in Table 5-1. All are admissible in plastic design"except for ASTM A5l4 which exhibits the smallest reserve in ductility since the minimum tensile stress is only 10 percent higher than the minimum y~eld stress"" However, elastic dynamic design may require "the" use of this "steel' or its boiler" plate equivalent, as in ASTM A5l7" 5-12.2. Mechanical Properties
un~er
Djnamic Loading, Dynamic Increase Factors
The effects of rapid "loading on the" mechanical behavior of structural steel have been observed and measured in uniaxial "tensile stress tests" Under rapidly applied loads, the rate of strain increases and this has a marked influence on the mechanicaf properties of structural steel. Considering the mechanical properties under static loading as a basis, the effects of increasing strain rates are illustrated in Figure 5-1 and can be summarized as follows: (1) (2)"
(3) (4)
The yield point increases substantially to the dynamic yield stress value. This effect is termed the dynamic increase factor for yield stress. The modulus "of elasticity "in general will remain insensitive to "the rate of loading.. ' The ultimate tensile strength increases slightly. However, the percentage increase is less than" "that for the yield stress. This effect is termed" the dynamic increase "factor for ultimate stress. The elongation at rupture either remains unchanged or is slightly reduced due to increased strain.rate.
In actual members subjected to blast loading, the dynamic effects resulting from the rapid strain rates may be expressed' as a function of the time" to reach yielding. In this case, the mechanical behavior depends on both the loading regime and the response of the system which determines the dynamic effect felt by the particular material. For members made of ASTM A36 and A5l4 steels, studies have been made "to determine the percentage increase in the yield stress as a function of strain rate, Design curves for the dynamic increase factors (DIF) for yield stresses "of A36 and A5l4 structural steel are illustrated in Figure 5-2. Even though ASTM A5l4 is not recommended for plastic design, the curve in Figure 5-2 may be used for dynamic elastic design. The strain rate, assumed to be a constant from zero strain to' yielding, may be determined according to. Equation 5-1:
where
5-9
TH 5-l300/NAVFAC P-397/AFR 88-22
E
tE f ds
average strain rate in the elastic range of the ste'el' (in/in/sec) time to yield
(sec~
dynamic design stress (Section 5-13)
Dynamic increase factors for yield stresses in var-Lous pressure levels Ln the
bending, tension, 'and compression modes are listed in . T able 5-2, 'The values for bending assume a strain rate of 0.10 in/in/sec in the'low design pressure range and 0.30 in/in/sec in the high pressure design range. For tension and compression members, the DIF values assume the strain rates are 0.02 in/in/sec in the low design pressure range and 0.05 in/in/sec in the high design pressure range. Lower strain rates are selected for the tension 'and compression members since they are likely to carry the reaction of a beam or girder which may exhibit a significant rise time, thereby increasing the time, to reach yield in the tension or compression mode.
On the basis of the above, the dynamic increase factors for yield stresses summarized in Table 5-2 are recommended for use in dynamic design. However, a more accurate representation may be derived using Figure 5-2 once the'strain rate has been determined. Steel protective structures and members are generally not designed for excessive deflections, that is, deflections associated with elongations well into the strain-hardening region (see Figure 5-1). However" situations arise where excessiye d~flections may be tolerated and will not lead to structural failure' or collapse. In this case, the ultimate stresses and associated dynamic increase factors for ultimate stresses must be considered. Table 5-3 lists the dynaml.ciLnc re aae factors' for ultimate stresses of steels. Unlike the dynamic increase factors for yield stress, these values ~re independent Of the pressure ranges. ' , 5-13. Recommended Dynamic Design Stresses 5-13.1. General The yield point of steel under uniaxial tensile stress is' generally used as a base to determine yield stresses under other loading states namely, bending, shear and compression, or tension. The design stresses are also functions of the ave!age stre~gth increase factor, &, and the dynamic increase factor, c.
5-13.2. Dynamic Design Stress for Ductility Ratio
~ ~
10
To determine the plastic strength of a section under dynamic loading, the appropriate dynamic y~eld stress, f dy' must be used. For a ductility ratio (see Section 5-16.3) ~ ~ 10, the dynamic design stress, f ds' is equal to the dynamic yield stress, f dy' In general terms, the dynamic yield stress, f dy' shall be equal to the product of the dynamic increase factor, c, the average yield strength increase factor, a, (see Section 5-12.1) and the specified minimum yield stress of the steel. The dynamic design stress, f ds' for bending, tension, and compression shall be:
5-2 5-10
TK 5-1300/NAVFAC P-397/AFR 88-22
where dynamic yield stress c -
dynamic increase factor on the yield stress (Figure 5-2 or Table 5-2)
a -
average strength increase factor (- 1.1 for steels with a specified minimum yield stress of 50 ksi or less; - 1.0 otherwise)
fY -
static yield stress from Table 5-1
5-13.3. Dynamic Design Stress for Ductility Ratio,
~
> 10
Where excessive deflections or ductility ratios may be tolerated,-the.dynamic design stress can be increased to account for deformations, in. the strainhardening region. In this case, for ~> 10, the dynamic design stress, f ds' becomes
5-3 where: dynamic yield stress from .Equation 5-21 dynamic ultimate stress equal to the product of f u from ,Table 5-1 and the value of c from Table 5-3 or Figure 5-2 it should be noted that the average strength increase factor, a, does not apply to f du' , ,-., ' ..
5-13.4. Dynamic Design Stress f9r ,Shear The dynamic design stress for
shea~
shall be:
5-4 where f ds is from Equation 5-2 or 5-3 .
.;
,
5-11 '
.'
fdu . fu
fdy ~
'iii
c.
:;:'
fy
/1----,,-
,,I
.,:Jl
-
,,-' ,,-
--
,.
,. ASTM STRAIN RATE - - - - RAPID STRAIN RATE
~
0.070
( J)
--
0.01 TO 0.02 in.lin. APPROX.
STRAIN,
Figure 5-1
E
(in./in.) .
Typical stress-strain curves tor steel 5-12
e
e
e
1.70; -f
.
1.60
()
OC
.-u 0
1.50
it lIJ
l/)
,
V'
t-'
'"
'1.40
1t41,. -t-t-
OC U
z
1.30
u ~
z >-
1.20
tE
0
1.1°11 1.00 10-3
"
'" ,IO~2.
'
,10- 1
19_' ._
10
STRAIN RATE e (in.lin./sec.)
Figure 5:-2
Dynamic increase factors for yield stresses at various strain rates for ASTM A-36 and A-514 steels
100
TH 5-1300/NAVFAC P-397/AFR 88-22 .
t ,
.,
.
.
Table 5-1 Static Design. Stresses for Materials' Material
f~.min
fa.min
"(ksi)
(ksi)
A36
36
58
A529
'42
60.
'(ASTM)
·40 . 42 , 46· 50
60 63 67 "'70
42 50 : . 60" 65
60 65 75 , 80
A242
42 46 50.
63 67 70
A588
42 46 50
63 67 70
A44l •
,
A572
J
,
, .,
A5l4
.
90 100
,'.'
.
100 lio ,
.
5-14 \
;
TH 5-l300,iNAVFAC P-:l97/AFR 88-22
'
....
., ..
'.".
Table 5-2
Material
,
A36 As88 As14
Dyn~mic
. (f
Increase Factor., c. for .'yield Stress of Structural Ste.els
..
, Tension or' Compression
Bendf.ng ,
-,
Low Pressure
-
High Pressu,re
.
0 ..10 in/in/sec)
. (f
1;-19~, .
L09
(f
- '0,. .
92).,j
-,
1.36 1.24*;, ,,' " ' ... 1.12
-.
, ·Low . Pressure
, 0.30), . "
"
1.29
-
'.'
, .,
1.19 1.12* , 1 . .0 5" , '.
.
'"
,. High Pressure
-
,
. (~.." 0.05) .. - 1.24 .~
....
.
..
,.1.15*: .1,.07
*Estimated
Table 5-3
Dynamic Increase Factor,. c , for,Ultimate Stress of Structural Steel,S. ~ .,. ';. '.
.
Material
,
A36
..
'
:
':
c "
1.10 )
As88
.. As14
1.05* "
'1.00
"I.
.
*Estimated
"
,
,
'. ,
5-15
.
.'
,
"
I
,.
TN 5-1300(NAVFAC P-397/AFR.88-22 DYNAMIC RESPONSE OF STEEL STRUCTURES IN THE PLASTIC RANGE 5-14. Plastic Behavior of Steel Structures Although plastic behavior is not generally permissible under service loading conditions, it is quite appropriate for design when the structure is subjected to a severe blast loading only once or at most a few t~mes during its existence. Under blast pressures, it will· usually be uneconomical to design a structure to remain elastic and, as a result, plastic behavior is normally
anticipated in order to utilize more fully the energy-absorbing capacity of blast-resistant structures. Plastic design for flexure is based on the assumption that the structure or member resistance is fully developed with the formation 'of near to'tally plastifikd sec t Lons- at' the most"lii'ghiY"sties'~ed" ,. <..;. locations. For economical design, the structure should' be proportioned to assure its . ductile behavior up to .the' limit of Its'load-carryi:ng capacity: • ... I The structure' or structural element can attain its: full .plastic capacity, , • provided that'premature impairment of' strength due to secondary effects, ~uch . ' as brittle fracture' or instability, dce s'
•
~,~
,
., .._
t
"
Structural resistance is determined on the basis of plasti~ d~s~gn concepts, taking into account'dynamic yield sttength v~lues,' The' desigh' proceeds with the basic obj ect Ivef chat; th-e comput'ecl'd7formatipns'of
j
••
,
,
5-15. Relationship Between Structure
~nction
."
and Deformations
5-15.1. General Deformation' i t e are ih: ideiail for two' 'categorie~ off str;'ctU:fes, • namely, acceptor-type structures in the low pressure range and structures in the high pressure'range which may either be acceptor- or donor-type .. A description of the two categories of structures foI16~~' cr
r'La
spec Lf'Le'd
<
_~
_
5-15.2. Acceptor-type Structures
in the I
'.
~
~J
.,
.,..
i
Low'Pressure-RAnges 1
'
;,
_
I
The maximum deformations to be specified in this category are consistent with maintaining structural integrity into the plastic range while providing safety for personnel and equipment. The type of.structure generally associated with this design category may be constructed of one or two .stories with braced or rigid frames. Main members consisting of columns'and main beams should be , . fabricated from hot rolled steel while secondary 'members, consisting of '; r : .t purlins or girts which span the frame members, can be hot-rolled I-shapes and channels or cold-formed Z-shapes and channels. The structure skin shall consist of cold-formed siding and decking spanning between the wall girts or roof purlins. 5-15.3. Acceptor- or Donor-type Structures in the High Pressure Range The deformation criteria specified·in this category cover the severe conditions associated with structures located close-in to blast. In cases where the design objective is the containment of an explosion the deformations should be limited. In other cases where prevention of explosion propagation or of missile generation is required, the structure may be allowed to approach
a
5-16 '1· .'
TH 5-1300/NAVFAC P-397/AFR 88-22
incipient failure, and deformations well into the strain hardening ,region may be permitted for energy absorption. In general, plate elements and curved plate-type structures fall under these categories.' 5-16. Deformation Criteria 5-16.1. General The deformation criteria presented. in'this chapter will be consistent with designing the structure for one accident. However, if it is desirable for a structure to sustain two or three "incidents· in its lifetime, the designer may limit design deformations so that, in its post accident condition, the structure is suitable for repair and reuse.
.
'
The deformation criteria for beams (including purlins, spandrels and girts) are presented in Section 5-16.5. The 'criteria for frames, including sidesway, are presented in Section 5-16.6 and that for plates are given in Section 516.7. Special consideration is given to the deformation criteria for open-web joists (Section 5-33) and- cold-formed metal decking (Section 5-34)., Deformation criteria are summarized in Section 5-35. 5-16.2. Structural Response Quantities In order to restrict damage to a structure or element which is subjected to
the effects of accidental explosion, limiting values must be assigned to appropriate response quantities. Generally speaking, two different 'types of values are specified, namely, limits' on the level of inelastic dynamic response and- limits on the maximum deflections 'and rotations.
For elements' which can be represented as single-degree-of-freedom systems such as beams, floor and wall panels, open-web joists, and plates, the appropriate quantities are taken as the maximum duc t Ll.Lty ratio and the maximum r orat Ion at an end support .. For systems 'such as frame structures which can be represented by multi-degreeof-freedom systems, the appropriate quantities are taken as the sidesway deflection and individual frame member rotations. 5-16.3. Ductility Ratio,
~
Following the development 'in Chapter 3 of this manual, the ductility ratio, ~,' is defined as the ratio' of the maximum deflection (~) to the equivalent elastic deflection (XE) corresponding to the 'development of the limiting resistance on the bilinear resistance diagram for the element. Thus, a ductility ratio of 3 corresponds to a maximum'dynamic response three times the equivalent elastic response. " In the case of individual beam elements, ductility,ratios as high as 20 can be achieved provdded that sufficient bracing: exists. Subsequent sections of this' " chapt.er vcover bracing requirements -for beam elements. in the case 'of plate elements, ductility ratios are important insomuch as'the higher ductility ratios permit the us e of higher design stresses. ' " '.,:'
Support rotations, as discussed in the next paragraph, provide the;ba~is'for beam and plate design. For a beam element, "the ductility ratio must be ' 5-17
;.
TH 5-1300/NAVFAC> P-397/AFR 88-22
checked to. determine whether the specified rotation can be reached without premature buckling of the member.· A similar. provision shall apply to. plates. even though they may undergo larger ductility ~atios in the absence of premature buckling. 5-16.4. Support Rotation,
e
The end rotation, e, and the associated maximum deflection, ~, for a beam are illustrated in Figure 5-3. As shown, e is the angle between the chord joining the supports and the point on the element where ~he deflection. is a maximum. 5-16.5. Limiting Deformations for .Beams A steel beam element may be designed to attain large deflections corresponding to 12 degrees support rotation. To·,assure the integrity. of the beam element, it must be adequately braced to permit this high level of ductile behavior. In no case, however, shall the ductility ratio exceed 20. A limiting support rotation of 2 degrees, and a limiting ductility ratio of 10 (whichever governs) are specified as reasonable estimates of the absolute magnitude of the beam deformation where safety for personnel and equipment is required. These deformations are consistent with maintaining structural integrity into the plastic range. Adequate bracing shall be present to assure the corresponding level of ductile behavior. The interrelationship between the various parameters involved in the design of beams is readily described in the idealized resistance-deflection curve shown in Figure 5-4. In the figure, the values shown for the ductility ratio, ~, and the support rotation e, are arbitrary. For example, the deflection corresponding.to a 2-degree support rotation can be greater than that cprresponding to a ductility ratio of 10. 5-16.6. Application of Deformation Criteria to a Frame Structure In the detailed analysis of a frame structure, representation of.the response by a single quantity is not possible. This fact combined with the wide range and time-varying nature of the end conditions of the individual f r ame ymembe r s makes the concept of ductility ratio intractable. Hence, for this case, the response quantities referred to in the criteria are the sidesway deflection .of each story and the end rotation, e, of the individual members with reference to a chord joining the member ends, as Illustrated Ln Figure 5 - 3. In addition' in lieu of a ductility ratio criterion, the amount of inelastic deforma-. tion is restricted by means of a ,limitation_on th~' individualme~ber rotation. For members which are not. loaded between their ends, such as an interior column, 0 is zero and only the sidesway,criteria must be considered. The maximum member end rotation, as shown in Figure 5-3, shall be 2 degrees. The maximum sidesway deflection is limited to 1/25 of the story height. These .responsequantities, sidesway deflection, and end rotation are part of the required output of various c~mputer programs which ,perform an inelastic, multi-degree-of-freedom analysis of frame structures. These programs are available through the repositories listed in Section_5-4 and several reports listed in the bibliography at the end of this chapter. The designer can use . the output of.these programs to check the sidesway deflection of each story and the maximum rotation of each member. 5-18
..
1M 5-l300jNAVFAC P-397/AFR 88-22
5-16,7. Limiting Deformations for Plates Plates and plate-type' structure's can undergo -Lar gd deformations 'with regard to '... support rotations and ductility ratios. The effect of overa~l and local instability upon the ultimate capacity is·considerably more important to structural 'steel shapes 'than to plates ...• Depending upon 'the' functional requirements'for a plate, 'the following-deformation criteria should be considered in the design;(;"f a'plate'~ '.• C ' , "c " '
.'.
.....
'0. :',
...
(1)
'Large deflections at·or close to'incipient failure.
(2)
Moderate deflections where the structure is designed to sustain two or three "incidents" before being nonreusable.
(3)
Limited deflections where performance of the structure is critical during the blast as, in the case of a blast door designed to,
.(4)
··.. contain pressure and/or ftre':lleakage. c
,'.
'
" . .
r
'\.
Elastic deflections'where the 'structure must not sustain permanent' deflections, as in the case of an explosives test chamber.
This is a partial list of d~sign'considerations for piaties.· I t can be seen that the designer must establish deformation criteria based on the function of
the plate or plate system. A plate or plate-type structure may undergo a support rotation, as illustrated in Figure 3-22 of Chapter 3, of 12 degrees. The corresponding allowable ductility ratio shall not exceed'20. It should be noted that higher design stresses can be utilized when the ductility ratio exceeds 10 (See Section 513.3) .
.
-'
. A limiting support rotation of 2 degrees is speclfied as 'a reasonable estimate' of the absolute magnitude of the plate support rotation where safety for, personnel and equipment -in an acceptor-type ,i):ructure is required. As in the deformation for. beanis';ratio shall . not exceed 10. ' .. - criteiia , . the ductility . Two edge conditions may :goverri'the deformation'6f'plates in the plastic region: The' first o~cuis;when~opposite edges are'riot built-in. In this case, elastic plat~'deflection: theory and yield-line theory apply: The secorid involves tensiori·membrane.action·which occ~is when at least'two opposite edge~' are clamped. In this case, tensile"membrane'actiori can occur before the plate element reaches a support rotation of 12 degrees. Tensile-membrane action of built-in p La t e s is not cov~red in' this chapter'. . However,' the. designer carr' u~ilize yiel~-line theory for. limited,plate defl~ction p~oblems. .' The interrelationship between 'the various parameters involved in the design 9f plates is readily described in the idealized ,resistance-deflection curve shown in Figure 5'-5. The figure shows the values for the ductility rati;'" u ,' and' the support rotation, e; are arbr'£rary. For example,' 'the deflection cor': responding to a 2-degree support rotation' can be greater than that' co r r e sponding to a ductility ratio of 10. ' , . ,- , ,
5-19'
"
TK 5-1300/NAVFAC P-397/AFR 88-22 5-17. Rebound Another aspect of dynamic design of steel structu~es subjected to blast loadings is the occurrence of. rebound: Unlike,the conditions prevailing in reinforced concrete structures where rebound. considerations, may not be of primary concern" steel structures, will be subjected to relatively large stress 'reversals caused by rebound and will require, lateral ,bracing of ,unstayed compression flanges which were formerly in t ens Lon , ··Rebound is mo~e critical. for elements supporting light dead loads and subjected to blast pressures, of short duration. It is also a primary ~oncern in the design of reversal bolts for'blast doors. 5-18. Secondary Hodes of Failure 5-18.1. General In the process of designing for the plastic 9r.ductile mode of failure, it .is important to follow certain provisions in order to avoid premature failure of the structure, i.e., to ensure that the strugture can develop its full plastic resistance. These secondaFY modes of failure can be grouped in two main categories:
,
(1)
Instability,modes of failure
(2)
Brittle modes. of failure '.
5-18,2. Instability Hodes of
Fail~re
In this category, the problem of structural instability at two levels. is of concern, namely, overall buckling of the structural system as a whole, and buckling of the component elements .
....
Overall buckling of framed structures can occur in two essentially different manners~ In the first case, the load and, the structure are symmetric; deformations remain also symmetric up to a critical value of the load for which a sudden change in configuration will produce instant anti-symmetry, large sidesway and displacement, and eventually a.faiiure by collapse' if'not by excessive deformations. This type of instability' can also occur in the . elastic domain before substantialdeformati~nor any plastification has taken p Lace , It is' called "instability by' bifurcation;' . " " In the second case, the loading or th~ s t ruc ture ·.or both are nonsymmetric. With the application of the. load, sidesway develops progressively. In. such cases, the vertical loads acting through the sidesway displacements, commonly called "the P-A effect", create second order bending moments that magnify the deformation. Because of rapidly increasing displacements, plastic hinges form, thereby decreasing the rigidity of the structure and causing more sidesway. This type of instability is related to a continuous dete~ioration of the stiffness leading to an early failure by. either a c911ap~e mechanism or excessive sidesway. ' Frame instability need not be explicitly considered in the plastic design of one- and two-story unbraced frames provided that the individual columns and girders are designed. according to the beam-column criteria of Sectio~ 5-37. 5-20
TIl 5 c1300/NAVFAC P-397/A,FR88 c 22" ,
'
For' fiamesn.greatet .than two stories l"'bracing is' normally ·,required "acc'ording to . the AISC provisions for plastic design in order to ensure the ove'r a Ll, .s t ab Ll.Lc; ty of the structure. However, if an inelastic dynamic 'frame analysis is performed;to.'determiriethe complete time,history_of'theJstructura1 'response to the blast loading, including the 'P-Aeffects, 'it may be established, in particular cases, that lateral bracing is not necessary i.n a frame greater than .two stories. ,.As ;mentioned previously, .compuce r iproguams. which perform an ine1astic,.imu1ti-degree-of'freedom analysis of frame! structures may be employed for such an analysis. ..L~" ,., :::'.1,., ••' ,}" .~ ;.~ 1-,
'
"v
_f", • • •
Buck1ingcof an element in 'the structure (e.g,:.', a'beam, girder; 'qr column) can occur under certain loading 'and r end :conditions. Instabi1ityvisof' tw,o types, name Lyo vbuck l'Lng co f. the member' as a whole (e . g. , 'lateral torsiona1'buck1ing) and local buckling at certain sections, including. flange buckl Ingvand web' crippling. Provisions for plastic design of beams and columns a r e presented in a separate
section of this chapter. 5-18.3. Brittle Hodes of Failure Under dynamic loading, there is an enhanced possibility that brittle ,fracture can
d~velop
under certain conditions.
Since this type of failure is sudden in
, ,nature and difficult to predict, it is very important to diminish the risk of such premature f a Ll.uxe . '
The cOl)lp1exity of the brittle fracture phenomena precludes a complete quantitative .definition. As a result" i t is' Lmpos s Lb Le to establish simple rules for'design. Brittle fracture will b~ associated with a loss in flexural resistance ..
Brittle fractures are caused by a combination of adverse circumstances that may include a few, some', or all of the folloWing:
,"
(1)
Local stress concentrations a~d residual stresses
(2)
Poor welding
(3)
The use of a notch se~sitive steel
(4)
Shock loading or rapid strain rate
(5) ,Low temperatures'
(6) (7)
,(8)
'.
Decreased ductility due to strain aging The existence 'of 'a plane strain condition causing a state of triaxial tension'stresses, especially in thick gusset plates, thick webs and in the vicinity of welds Inappropriate'use of some forms of connections,
..
The problem of brittle fracture is closely related to the detailing of "connections, a topic that will b<;> treated in a'separate section of this chapter..
TH,5~1300/NAVFACP~397/AFR'88-22
However, there:a~e'certain gener~l guLde H.ne s to .. follow in 'order to ,minimi:!:e
the danger of brittle fracture:
,".,
,.0-
: (1) .'
.~
"
"
,0
~ -'~
"
.
,,"".
.
,
~
Steel material must be' 'selected to conform with the '. condi tion ," ... ,. anticipated in service.
(2)
Fabrication,and'w,orkmanshlp should meet high standards, e i.g s , sheared edges,and.notches should be. avoided, and ~aterial that'has been severely cold-worked sn~~id'b~-~e~oved.
(3)
Proportioning and detailing of connections should be such that' free movement of the ba~e material is .permitted, stress con~entrations and t
r
L a x
L a
l . v s
t
r e
s
s
ductility is ,provicle'd'.
conditions
- a r
avoided" and
e
~f
c' -',
r.
. a d e q u a
.( t . e
:'
')',
r
.",
•
, 1
,.
". 1 ",l '-.' '".
., . .'.'
'.
'-
.. ',.-
·1.
. ,:;- .'
~.
, _., '.-.,
I
, ',".1
",-
"
",
;
•.
'
.~-.
~
L
'
:.' ~ 'J
,
"
.
'\
.c·
.,,
.
~'
J';","
:'0.
j !.. ',.
"
, - .'
, i' : ... ;.~
,1 ."
',"
'" -",
"
•. ,'
'\ '5.-22
. . ~
"
t.,,"
.s, '
•
,
. .',' '.
~.
,,
Jr----------:::;...-. . , . -;-.~ ~m
.!£
-
BEAM, SLAB OR PANEL
.,
8· 8
8BO
80B
0-
H
FRAME
, , Figure 5-3
Member end rotations for -beams and frames
.'.,
F
ILl
TOTAL FAILURE
(.)
Z
sen en
ILl
a::
ULTIMATE
ELAST
-, , ..
.
. .'
I
I
.
I
I
I
I
I I' I
..
'I
.
.
I'
,-INCIPIENT FAILURE "
.
"
XI9=2") OR IfL =10)
XlfL= 20)
or
DEFLECTION
X(9 =12°)
CD
®
PROTECTIO N CATEGORY
(3)8(4)
SHELTER PROTECTIVE STRUCTURE
BARRIER, PRE SURES
STRUCTURE - LOAD SENSITIVITY
PRESSURE - TIME
STRUCTURE CATEGORY
..
-
..
....
.
I
ACCEPTOR' OR
DONOR
,.,.'
ACCEP,;rOR
,~,
,
.
'
..
-
"
-
'or-
;
RESp'O SE CHARTS OR ITERATfON
DESIGN METHOD
'"
LIMI TED
RESPONSE CRITERIA
..
LARGE.
,
DUCTILE MODE OF RESPONSE
BENDrN ·8 SHEAR S
DESIGN SECTION MODULUS'
(S+Z)/2
<'dy'
DESIGN STRESS
..
,
Z
, 'dy+
'dy
Y4(' du-'dy)
•
Figure
5-~ ,
,
. ,
,
,
,,
Relatio!15h1P!'. between des t gn parameters for beams •
'"1':
t
5-24~
r. J
,
•
•
..
,
ELAST
,
TOTAL ' flIlLURE
' ULTIMATE,
"
"....INCIPIENT FAILURE
/1l
!
I
•
I
,
'.
I
.'
1
•
I,
.
.
'
X(9:2") OR (JJ- :10).
X(JJ-' 20)
DEFLECTION
X(8:12°)
-'
or
'CD .' ,.
®
PROTECTION' CATEGORY"
)
(3) a (4) SHELTER PROTECTIVE STRUCTURE
BARRIER
. .. PRESSURE'S' ,
e
:
STRUCTURE - LOAD SENSITIVITY
PRESSURE -TIME ACCEPTOR, ,
,.
,
STRUCTURE CATEGORY
.
,
ACCEPTOR OR
RESPO SE CHARTS OR ITERATION
DESI,GN METHOD LIMITED
RESPONSE CRITERIA DUCTILE MODE OF RESPONSE
,
a
SHEAR
(S+Z)/2
S
DESIGN STRESS
'
LARGE BENDrN
DESIGN SECTION MODULUS
DONOR
Y
I d y+ 4 ( I d.-Idy)
I dy
"
.'
Figure 5-5 'Relationships between dest-gn parameters for plates ~.
5-25
1M 5-l300fNAVFAC P-397/AFR 88-22
. DESIGN OF. SINGLE 'SPAN..AND CONTINUOUS BEAMS'
5-19';' General The emphasis in this section is on the-dynamic plastic design of structural steel beams. Design data have been derived from the static provisions'of the , AISC Specification with necessary modifications and additions for blast design. It should be noted that all provisions on plastic desig~ in the AISC Specificati~n apply, except as modified in this chapter. The calculation of the· dynamic flexural capacity of beams is described in detail. The necessary information is,pre~ented for determining the equivalent bilinear resistance-deflection functions ·used in evaluating the basic flexural response of beams:
Also
pres~nted
are the supplementary considerations of
adequate shear capacity and local and overall stability which are necessary for the process of hinge formation, momenf redistribution and inelastic hinge rotation to proceed to the development of,a full collapse mechanism .
.
.
...
5-20. 'Dynamic Flexural Capacity 5-20.1. General The ultimate dynamic moment resisting capacity of a steel beam is given by 5-5
where f ds is the dynamic design stress" (as' de scr Ibed in' Section' 5 -13) of the material and ~ is the plastic section modulus. The plastic section modulus can be calculated as the sum of the.static moments of the fully yielded elements of the equal cross sec t fon '·areas. above and below the neutral axis, i.e. :
5-6
Note: where
Acml - Atm2 for a doubly-symmetric section area of cross section in compression
area of cross section in tension distance from neutral axis to the centroid of the area in compression distance from neutral axis to the centroid of the area in tension For standard I-shaped sections (S, W, and M shapes), the plastic, section modulus is approximately 1.15 times the elastic modulus for strong axis bending and may be obtained from standard manuals on structural steel design. It is generally assumed that a fully plastic section offers no additional resistance to load. However, additional resistance due to strain hardening of the material is present as the deformation continues beyond the yield level of the beam. In the analysis of structural steel beams, it is assumed that the plastic hinge formation is concentrated at a section.. Actually" the plastic region extends over a certain length that depends on the type of loading 5-26
TK 5-l300/NAVFAC P-397/AFR 88-22
(concentrated or'distributed) on the magnitude of the deformation, -and on the shape factor of the cross section. The extent of.the plastic hinge has no substantial influence on the ultimate capacity; it has, however, an influence on the final magnitude of the deflection. For all practical purposes, the assumption of a concentrated plastic hinge is adequate. In blast design, although strains well into the straIn-hardening range may be to1erat~di the corresponding additional resistance is generally not sufficient to warrant analytical consideration since excessive support rotation and/or ductility ratios of beams, which are susceptible to 10cal,f1ange or lateral' torsional buckling, are not rec8mmended .
.
5-20.2. Moment-curvature Relationship for Beams Figure 5"6 ,shows the stress distribution'at various stages of deformation for a plastic hinge section. Theoretically, the beam bends elastically until the outer fiber stress reaches f ds and the yield moment'designated,by Mv is attained (Figure 5-6a). As the moment increases above My" the yie1a stress progresses inward from the outer fibers of the'section towards the neutral axis as shown in Figure 5-6b. As the moment approaches the fully plastic moment, a rectangular stress distribution as shown in Figure 5-6c is approached. 'The ratio between the fully plastic moment to the yield moment is the shape factor, f,' for 'the section, i.e., the ratio between the plastic>and
elastic' section modu1ii: A representative moment-curvature relationship for a simply-supported steel' beam is shown'in Figure 5~7. The behavior is elastic until the yield moment My is reached. With further increase in load, the curvature increases at a greater ra~e as the fully plastic moment value, M2, is approached. Following the attainment of M2, the curvature increases: significantly, with only a'sma11 increase in'moment'capacity. For design purposes, a bi1in~ar representation of the moment-curvature relationship is employed as shown by the dashed lines in Figure 5-7, For beams with a moderate design ductility ratio (~S 3), the,design moment ~ - ' M1. For beams with a larger design ductility ratio (~> 3), the design moment ~ - M2·
'.
5-20.3. Design Plastic Moment,
'
~
The equivalent plastic design moment shall be'computed as follows: For beams with ductility ratios less than or equal to 3: "'
~ -
..
..
f ds (8 + Z)/2
~
'
5-7
where 8 and Z are the elastic and plastic section modu1ii, respectively. For beams with ductility ratios greater than 3' and beam columns: 5-8 Equation 5-7 is consistent with test results for beams with moderate ductilities. For beams which are a l Iowed to undergo' large ductilities, Equation 5- 8" based upon full p1astification of the section, is considered reasonable for design purposes. 5-27
,
,
TM 5-1300/NAVFAC P-397/AFR 88-22 It is important to note that the above pertains to beams which are supported against buckling ... Design provisions for guarding against local and overall buckling of beams during plastic deformation are discussed in Sections 5-24, 5-25, and 5 - 2 6 . "
5-21. Resistance and Stiffness Functions 5 -21.1. General The single-degree-of-freedom analysis which serves. as the basis for the flexural response calculation requires that the equivalent stiffness and .. ultimate resistance be defined for both single-span beams and continuous beams. The ultimate resistance values correspond to developing a full collapse mechanism in each case. The equivalent stiffnesses correspond to load-deflection.relationships that have been idealized as bilinear functions with initial slopes so defined that the areas under the idealized load deflection diagrams are equal to the areas under. the actual'diagrams at the point of inception of fully plastic behavior of the beam. This concept is covered in Section 3-13 of Chapter 3.
5-21.2. Single-span Beams , ..
l.
Formulas for determining the stiffness and resistance for one-way steel qeam elements are presented in Tables 3-1 and 3-8 of Chapter 3. The values of M in Table 3-1 represents the plastic design moment,~. For example, the value of r u for the fixed-simple, uniformly loaded beam becomes r u - l2~/L~.
5-21.3. Multi-span Beams The beam relationships for defining the bilinear resistance function for multi-span continuous beams under uniform loading are summarized below. These expressions are predicated upon the formation of a three-hinge mechanism in each span. Maximum economy normally dictates that the span lengths and/or member sizes be adjusted such that a mechanism forms simultaneously in all spans. It must be noted that the development of a mechanism in a particular span of a continuous beam assUmes compatible stiffness properties at the end supports. If the ratio of the length of the adjacent spans to the span being considered is excessive (say, greater than three). it may not be possible to reach the limit load without the beam failing by excessive deflection. For uniformly distributed loading on equal spans or spans which do not differ in length by more than 20 percent, the following relationships can be used to define the bilinear resistance function: Two·span continuous beam:
rubL - 12 ~/L
5-9
163 EI/L 3
5-10.
Exterior span of continuous beams with three or more spans:. i
5-11
5-28
TH 5-l300/NAVFAC P-397/AFR 88-22 5-12 ' Interior span of continuous beam with three or more spans:
Ru-
rubL -'16,O~/L
5-13
KE
300EI/L3
5-14
For, design' situations 'which do not meet the required condf t Lons , 'the bilinear resistance function may be developed by the'application of the basic procedures of plastic analysis! ' L ,
,,
"
5-22. Design for Flexure' 5-22.1. General The design ofa structure to resist the blast 'of an accidental explosion consists ,essentially of the determination of'the structural resistance required to limit calculated deflections to within the prescribed maximum values as outlined in Section 5-35. In general, ,the resistance and'deflection may be 'computed on the basis of flexure provided that the shear capacity of the web is not exceeded. Elastic shearing"deformations of steel members are negligible as long as the depth to'span ratio is less than.about, 0.2 ~nd hence. a flexural analysis is normally sufficient for establishing maximum de f Lec t Lons t.
' ,,'
5-22.2. Response'Charts ,
,
1& .•
Dynamic response charts for one-degree-of-freedom systems in 'the elastic or elasto-plastic range under various dynamic loads are given .in',;Chapter 3. To use vche charts, the effective natural period of, vibration 'of a'-s t ruc tura I steel beam must be determined. The procedures for determining.tne natural period of vibration for one-way elements are outlined in Section 3-17.4 of Chapter 3. Equation 5-15 can be used to determine 'the' natural 'period of vibration for any system for which the total effective mass, Me. and equivalent elastic stiffness, KE are known: . ','
5-15 -o.
5-22.3. 'Preliminary Dynamic'Load Factors For preliminary flexural design of beams situated in low pressure range, it is suggested that an equivalent static uitimateresistance equal to'the peak blast' pressure be used for those beams designed for, '2 degrees support rotation, ' For large support. rotations , a preliminary dynamic· load factor 'of 0.5 is recommended. Since·the duration,of the loading for lo~,pressure range will generally be' the same or longer than the period of ,vibration of tne structure, revisions to this preliminary design from a dynamic 'analysis ,will usually not be substantial. However, for structures where the loading environment pressure is such that the load duration is short as compared with the period of vibration of the 'structure, this procedure may result in·a substantial overestimate of the required 'resistance.
5-29 '
TM 5-l300/NAVFAC P-397/AFR 88-22
5-22.4. Additional Considerations in Flexural Design Once a dynamic analysis is performed on the single span or continuous beam" the deformations must be checked with the limitations. set in the criteria. The provisions for local buckling, web crippling and.lateral bracing must be met. The deformation criteria for beam elements including purl ins , spandrels and girts are summarized in.Section 5-35. The -rebound behavior of the structure must not be overlooked.-
The information
required for calculating the elastic rebound of structures is contained in Figure 3-268 of Chapter 3. The provisions for local buckling.and lateral bracing, as outlined.in subsequent sections of this chapter, shall apply in the design for rebound. 5-23. Design for Shear Shearing forces .are of.-significance in plastic design primarily because of their possible influence on the plastic moment capacity of a steel member. At points where large. bending moments and shear forces exist, the assumption of an ideal elasto-plastic stress-strain relationship indicates that during the progressive formation of.a.plastic hinge, there is a reduction of the web area available for shear. . This reduced area could result -.in an initiation of shear. yielding and possibly reduce ... the imoment capacity.
.
.~-'
,;
'-,
However, it has been found experimentally that I-shaped sections achieve their fully plastic moment capacity provided that the average shear stress over the full web area is less than the yield stress in shear. This result can basically be attributed to the fact that I,shaped sections carry moment predominantly. through the flanges and shear predominantly through the.web. Other contributing factors include .the beneficial. effects of strain hardening and the fact that~combinations of high shear and high moment generally occur at locations where the moment gradient is steep. "I
,
The yield capacity of steel beams in shear is given by: _,'
l
5-16 where Vp is the shear capacity, f dv is the dynamic yield stress in shear of the steel (Equation 5-4), and is the area of the web. For I-shaped beams and similar flexural members with thin webs, only the.web area between flange plates should be used in calculating
Aw
Aw.
'.
For several particular load. and support conditions, equations for' the support shears, V, for one.way elements are given in Table· 3-9 of Chapter 3. As discussed above, as. long as the acting s~ear V does not exceed V , .I-shaped sections can be considered capable of achieving their full plast~c moment. If V is greater than Vp,.'the web .area of the chosen section Lsi Lnadequat.e and either the· web mu~t be strengthened or·a different section should be selected .
..
However, for, cases where the web is being relied upon to carry a significant portion of the moment capacity of the section, such as rectangular cross. section beams or built-up sections, the influence.of shear on the available moment capacity must be considered as treated in Section 5-31.
5-30
TK 5-l300/NAVFAC P-397/AFR 88-22
5-24., Local Buckling In order to ensure that a steel beam will attain fully plastic behavior and attain the des Lred ductility at plastic hinge locations, it is necessary that the elements of the beam section 'meet minimum ,thickness requirements sufficient to prevent a local buckling failure. Adopting the plastic design, requirements of the AISC Specification, the width-thickness ratio for flanges of rolled,I- and W-shapes and similar built-up' single web shapes that would be subjected to compression involving plastic hinge rotation shall not exceed the following values: f y (ksi)
b f /2tf
36 42 45 50 55 60 65
8.5 8.0 7.4 7.0 6.6 6.3 6.0
•f
where fyis. the specified minimum static yield stress for the steel (Table 51), b f is the, flange width, and tf is the flange thickness.' The width-thickness ratio of similarly compressed flange plates in box sections and cover plates shall not exceed 190/(f )1/2. For this purpose~ the width-of a cover plate shall be taken as the distance between longitudinal lines of connecting rivets, high-strength bolts, or welds. The depth-~hickness ratio· of webs of members subjected to plastic bending shall riot,exceed the value given by Equation 5-17 or 5-18 as 'applicable. d
412 [ . P ] ~l - 1;4,~
p
when
s
0.27
5-17
,. p
d
when
> 0.27
5-18
where ,p
the applied compressive load the plastic axial load equal to the cross' sectional area times the specified minimum static yield stress, .fy
These equations which are applicable to local buckling under dynamic loading have been adopted from the AISC provisions for static loading. However, since the actual process of buckling takes a finite period of time, the member must accelerate laterally and the mass of the member provides an inertial force retarding this acceleration. For this reason, loads that might otherwise cause failure may be applied to the members for very short durations if they 5-31
TM 5-l300/NAVFAC' P-397/AFR 88-22
are removed before the buckling has occurred. Hence, it is appropriate and conservative to apply the criteria developed for static loads to the case of dynamic ,loading of, relatively short duration. These requirements on. design of all,members necessary to evaluate member which does not
cross section' geometry should' be adhered, to in, .the for blast loading. However, in the event that it is the load-carrying capacity of an existing structural meet these provisions, the ultimate capacity should be
reduced in accordance with the recommendations made in the Commentary!and-·
Appendix C of the AISC Specification.
,
-.
5-25. Web Crippling
Since concentrated loads and reactions along a short length of flange are· carried by compressive stresses in the web of'the supporting member, local yielding may occur followed by crippling or crumpling of the web. Stiffeners bearing against the flanges at load points and fastened to the web are usually employed in such situations to provide a gradual transfer of these forces to the web. Provisions for web stiffeners, as given in Section 1.15.5 of the AISC Specification, should be used in dynamic ,design. In applying these provisions, f y " should be taken equal to the specified static yield strength of the steel. . " 5-26. Lateral Bracing 5-26.1. General
,
'.'
•
I
Lateral bracing support is often provided by floor beams, joists or purlins which frame into the member to be braced. The unbraced lengths _(lcr", as defined in Sections 5-26.2 and 5-26.3) are either fixed by the spacing of, the.. ,purlins and girts or by the spacing of supplementary bracing. When the compression flange is securely connected to steel decking or siding, this will constitute adequate lateral bracing in most cases. In addition, inflection points (points of contraflexure) can be considered as braced points. Members built into a masonry wall and having their web perpendicular to this wall can be assumed to be laterally supported with respect to .t.he Lr .weak axis of bending. In addition, points of contraflexure can be considered as braced points, if necessary. Members subjected to bending about their strong axis may be susceptible to' , , lateral-torsional buckling in the direction of, the weak axis if their compression 'flange is not laterally braced. Therefore, in order for a plastically designed member to reach its collapse mechanism, lateral supports must be provided at the plastic hinge locations and at a certain distance from'the hinge location: Rebound, which constitutes stress reversals, is an important consideration ,for lateral bracing support. •'
,
; "
5-32
TN 5-1300/NAVFAC P-397/AFR 88-22
5-26.2. Requirements for Members with
~ ~
3
Since members with the design ductility ratios less than or equal to three. undergo moderate amounts of plastic deformation, the bracing requirements ,are somewhat less stringent. For this case, the following relationship may be used: l/rT -
[(102 x
103~) /f d S]
1/2 5-19
where 1 -
distance between cross sections braced against twist or lateral displacement of the compression flange radius of gyration of a section comprising the compression . flange plus one..- third of the-,compression web area taken about an axis in the plane of the web bending coefficient defined in Section 1.5.1-.4.5 of, the AISC Specification
5-26.3. Requirements for Members,with
~
> 3
In order to develop the full' plastic momen~, ~ for members ~ith' design ductility ratios greater than three, the distance from the brace at the hinge location to the adjacent braced points should not be greater ~han lcr as . determined from either. Equation 5-20 or 5-21, as. applicable: 1375
.
M
+ 25 when 1.0
B
> -0.5
~
f ds
~
1375
M
when -0.5
B
f ds where
5-20
~
~
-1. 0
5-21
~
the radius of gyration of the member about its weak axis the lesser of the moments at the ends of the unbraced segment M/~ -
B
the end moment ratio. The moment ratio is positive when the segment is bent in reverse curvature and negative when bent in single curvature. critical length correction factor (See Figure 5-8)
The critical length correction factor, B, accounts for the fact that the required spacing of bracing, lcr' decreases with increased ductility ratio. For example, for a particular member with r y = 2 in. and f ds 51 ksi and using the equation for M/~ = 0, we get lcr - 71.7 in. for ~ - 6 and lcr 39.7 in. for ~ - 20. 5-33
TII,5-l300/NAVFAC P-397/AFR 88-22
5-26.4. Requirements for Elements Subjected-to ReboUnd The bracing requirements for nonyielded segments of members and the bracing requirements for members in rebound can be determined from the-following relationship:
5-22
where
f
the maximum bending stress in the member, and in no case greater than f ds
When f equals f ds' -this' equation reduces to'the l/rT requirement of Equation 5-19. " 5-26;5. Requirements for Bracing Members ln order to function adequately, the bracing member must meet certain minimum requirements on axial strength and axial stiffness. These requirements are quite minimal in relation to the properties of tyPical framing members. Lateral braces should be welded or securely bolted to the compression flange and, in addition, a vertical stiffener should generally be provided at bracing points where concentrated vertical loads are-also· being transferred. Plastic hinge locations within uniformly loaded spans do not generally require a stiffener. Examples of lateral bracing details are presented in Figure 5-9 .
.
,".
-
;
5-34
~
N.A.
MZ "7 (c I
Figure 5-6
Theoretical stress distribution for pure -l:iending'at various stages of d:;7Hlllli c loading 5-35
CURVE'b
ACT:;:~~_;;:;~R~:G~ DESIGN MOMENT FOR BEAMS WITH MODERATE DUCTILITY ( P. H)
Mp=M1 =fds . (.!.!.!..) 2 DESIGN MOMENT FOR BEAMS WITH LARGE DUCTILITY (P..>3)
Mp=MZ=fds l CURVATURE
Figure 5-7
Manent-;curvature diagram for s-imple-supported, dynamically loaded, . I-shaped beams 5-36
3.0
. '.
~
/
'.I
a::
~. ~ 2.5
17
>
,
V ~
u,
V
..
Z
o
V
~
o
/
ILl
a::
:5
V
2.0
17
o
", V
(!)
Z
-l
.v
1.5
o
.- 1/
1/
!:::.". a::
. .
.,
,
.- 1/
o
~
,
V,
,.
. . . ILl
;i
.
7
i=
17
,
'.'
:
1.0 3
4
8
6
10
DUCTILITY
12
14
RATIO,
16
'18
20
Xm/X E
"
:,
..
\'~
.
Figure 5-8 . Values of
'a
for use in equations 5-20 and 5-21 5-37
~--BOLTS
BOLTS,,..
BOLTED CONTINUOUS CONNECTION
. BOLTED DISCONTINUOUS CONNECTION
STIFFNER
WELDED CONTINUOUS CONNECTION
WELDED DISCONTINUOUS CONNECTION
STIFFENERS
TI E D BOTTOM
Figure 5-9
FLANGE CONNECTION
Typi cal lateral bracing details
5-38
TM 5-l300/NAVFAC P-397/AFR 88-22
DESIGN OF PLATES 5-27. General The emphasis in this section is on the dyn~mic plastic design of plates. As in the case for simply supported .and continuous beams, design data have been derived from the static provisions of the AISC Specification with necessary modifications .and additions ,for blast design. "
This section' covers the dynamic flexural' capacity:..of plates, .as well ';'s' the' necessary information for determining the equivalent bilinear resistancedeflection'functions used in evaluating the flexural response of plates. Also presented is' the supplementary consideration of adequate shear capacity at negative yield lines. 5-28. Dynamic Flexural Capacity As is the case'for standard.I'shaped'sections, the ultimat~,dynamic momentres Ls t ingvcapacd ty of a steel plate is a function of the elastic and plastic modulii and the dynamic design, stress. For :·plates or rectangular cross section beams, the plastic section'modulus is 1.5 times the elastic section modulus. A representative moment-curvature relationship for a simply-supported steel plate is shown in Figure 5-10. The behavior is elastic until a curvature corresponding to the yield moment'Ky is reached. With further increase in, ' load, the curvature increases at a greater rate as 'the fully plastic moment value, M2 , · i s · appr oach ed . Following the attainment ,of M2 , the curvature increases while the moment remains constant.
For plates and rectangular, cross section beams, M2 is 50 percent,greater than and the nature of the t r ans'Lt Lon . from yield to the fully ·plastic condition depends upon the plate geometry and end conditions. It is recommended that a capacity midway between'Mv'and M2' be used to define the plastic design moment, ~ (M l in Figure ,5-11), fbr ,plates and rectangular cross section beams. Therefore,' for· plates ,with any ductility. ratio, Equation 5-7 applies.
'Y'
5-29. Resistance and Stiffness Functions Procedures for .determining stiffness and resistance factors for one- and twoway plate elements are outlined in Chapter 3. These factor~ are based upon elastic deflection theory and the yield-line method, and are appropriate for defining,the stiffness and ultimate .load-carrying capacity of ductile structural steel plates. :In applying these factors to steel plates, the modulus of elasticity should be taken equal to ,29,600~000 psi. For two-way isotropic steel plates, the ultimate unit positive and negative moments are equal .Ln all directions; i.e.
where ~ is defined by Equation 5-7 and the remaining values are ultimate unit moment capacities as defined in Section 3-9.3 of Chapter 3. Since the stiffness factors were derived for plates with equal stiffness properties in each direction, they are not applicable to the case of orthotropic steel plates, 5-39
TIl 5-1300/NAVFAC P-397/AFR 88"22
I such as stiffened plates, which have different stiffness properties in each direction. 5-30. Design for Flexure The procedure for the flexural design of a steel.plate is essentially the same as the design of a.beam. As for beam~,. it is suggested that preliminary dynamic load factors listed in Table p-4 be used for plate structures. With the stiffness and resistance factors From Section 5-29 and taking into account the influence of shear on the available plate moment.capacity as defined in I c, Section 5-31, the dynamic response anp rebound for 'a given blast loading may be determined from the response charts in Chapter '3. It should be noted.,that for IJ. > 10, the dynamic design stress, .incorporating the· dynamic ultimate stress, f du' may be used (see Equation 5-2). 5-31. Design for Shear In the design of rectangular plates, the effect of simultaneous high·moment and high shear at negative yield lines upon the plastic strength of the plate may be significant. In such cases,· the following interaction formula describes the effect of the supportcshear, V, upon the available moment'capacity, M: 5- 23
, "
where ~ is the fully plastic moment capacity in the absence of shear calculated from Equation 5-7, and Vp is the ultimate shear capacity in the absence of bending determined from Equation 5-16 where the web area, ~i is taken equal to the total cross sectional area at the support . . For two-way elements, values for the ultimate support shears which are applicable to steel plates are presented in Table 3-10 of Chapter 3. I
_ r_ "
It should be noted that, due to the inter-relationship between the support shear, V, the unit ultimate flexural resistance, r u' of the two-way element, and the fully plastic moment resistance, ~,. the d.etermination of. the resistance of steel plates considering Equation 5-23, is not a simple calculation. Fortunately, the number of instances when negative yield lines with support shears are encountered for steel plates will be limited. Moreover, in most applications, the V;Vp ratio is such, that the available moment capacity is at least equal to the plastic design moment for plates (Equation 5-7) .... ' It is recommended that for a V;V . ratio on negative yield lines less' than 0.67, the plastic design moment ~or plates, as determined from Equation 5-7, should be used' in design. However, if V;Vp is greater than. 0.67, the influence of shear on the available moment capacity must be accounted for by means of Equation 5-23.
5-40
"
:r
CURVATURE NOTE: SEE FIGURE 5- 7 --FORMy,M 1 ANDM 2
Figure 5-·' 0 Manent-curvature diagram for dynamically loaded plates anur-ect angui ar-. cross-section beams,' ';,
5-41'
ylt r
I->
i'-TRANSVERSE LOAD
J
'j x
x ELASTIC AND PLASTIC NEUTRAL AXIS THROUGH THE CENTROID
~,
y
8 cot
L -_ _.....
Figure 5-11
=ELASTIC DEF.LECTION
=-/ 8x' + 8/
8
Biaxial bending or-a 5-42
doubly-s~etrie ""etlon
TN
5~1300fNAV~AC
P-397/AFR 88-22
. f .' i
,
, '
\- "
. ;
,
.
t ..
" .
"
"
,
,Table: 5-4
Preliminary
••
Dyn~mic ~ad
Factors for .. ":~
.
I
Pl~tes
,
-.
Deflection '" ,
"
e
Magnitua,; Small ' , . I ." Moderate Large . '- . '
r
-:"
..
~
,~.\.:
..
"5
'2
4 12,
"
"
*
..
,
"
,.'-
nLF .
,
. J.~
0.6
1 •
".
,
..
....
-
; r: -, '
"
.J' .::'
,
"
"
,
"
..
"
'
,-
, ,
1.0
0.8
'10 20
Whichever governs'
,
"
·Deformation* .'. max ~ max
I
TIl 5-l300/NAVFAC.P-397/AFR ,88"22', SPECIAL CONSIDERATIONS, BEAMS 5~32.
Unsymmetrical Bending
5-32.1. General In blast design, the number of situations where unsymmetrical bending occurs is limited and where encountered, it can be treated without serious economic' penalty. Due to the fact that blast overpressure loads act normal to the surfaces of a structure, the use of doubly-symmetric cross sections for purl ins and girts (e.g., hot-rolled'S- and W-sections or cold-formed channels used back-to-back) is generally recommended. In such cases, the deformation criteria for flexural members in Section 5-22 apply. Unsymmetrical bending occurs when flexural members are subjected to transverse loads acting in a plane other, than a pr LncIpa l, plane. With this type of loading, the following are applicable:
(1) (2)
(3)
, (4)
The member's neutral axis is not perpendicular to the plane of loading.
Stre~s~.s cannot in general be'calcul~teii by means of the' simple bending ,_formula (Mc/I). The bending deflection does not coincide with the plane of loading but is perpendicular to the inclined neutrai axis. If'the plane of'loads does not 'pass thro~gh the shear center of 'the cross section, bending is also accompanied by twisting. .. ~
Doubly-symmetric S, 'w, and box sections acting ~s individual beam elements and subjected to' biaxial bending, i.e'., unsymmefrical bending without torsion, can be treated using the procedures outlined in the following se ct Lons. 5-32.2. Elastic and Plastic Section Modulus The inclination of the elastic and plastic neutral axis through the centroid of the section can be calculated directly from the following relationship (see Figure 5-11): tan a where
5-24
a
angle between the horizontal 'principal plane ,and the neutral axis
~
angle between the plane of the load and the vertical prIncipal plane
, and x and y refer to the horizontal and vertIcal principal axes of the crosssection.
~
The equivalent elastic section modulus may'be evaluated from the following equation: S - '(SxSy)/(Sycos ~ + Sxsin ~) 5-44
• 5-25
. TM 5-l300/NA.YFAC ;P-397/¥R 88-22 .
elastic' section. modulus about. the x-axis
where
. ~ ','
"elastic section modulus about the y-axis ",
..
?~';
The plastic. section modulus carr-be ec a LcuLa t ed using Equation·5-~-, · ,With these values of.the. elastic and.plastic section'modulii~, the design p~~stic mom~nt capacity can be determined from Equation 5-7 . . " .i
,. . ,'l
5-32.3.
Equi~alent
Elastic Stiffness
.. .':'
In order to define the stiffness:and bilinear resistance function, .. it is necessary to' determine the'elastic deflection of the beam. ~is.deflection may be calculated by resolving the.. load into.components acting in,;the principal planes of the cross section. The elastic deflection, &, is;calculated as the resultant of the deflection~ determined by simple bending calculations in each direction (see Figure 5-11).· The equivalent elastic de f Lec t Lon on the, bilinear resistance functionX E, may then be determined by As~uming ~hat,the elastic stiffness is. valid up to the .. development of the des Lgn plastic moment capac i ty, Mp. , ". .., ,. ~ , . ..' , ..... , 5-32.4. Lateral Bracing,and Recommended ~esign.~riteria
"
\
"
r- ,;
The bracing requirements of Section 5-26 may not be totally adequate to permit a b Laxf.a Ll.y loaded section to deflect into the inelastic range' without..,. premature ..failure. However.; ·for. lack of data, the provisions ,of .Sec t Lon 5 -26, on lateral bracing. may be used' if the total' member end r o t at Lcn -co r r eapondfrig , to the t o t'a l: deflection due to the incliped ,load is limited to 2 degrees., In addition, the actual details of support conditions and/or bracing prOVided to' such members by the other primary and secondary members of the frame must be carefully checked t;o. ensure that the proper. conditions exist t.ov-perm l t; ' " deflections, in'the inelastic range. ' .:; '.~ •.... : -v,
I
,
.: " ,i ~"
, ,.
5-32.5. Torsion and Unsymmetrical Bending
.'
.J ••
The Lne Last t cibehavf.ox 'of sections subjected to unsymmetrical bending, .with. twisting; is not totally known at present .. Consequently, the·use of sections ... : with the resultant load 'not passing through the shear center ,is not'recom-' ",' mended in.plastic design of blast-resistant structures, unless ~orsional constraints are provided f'o r. the elements In actual .installations',· however, the torsional constraint offered to a pur lin or girt by the. "flexural .,rigidity of the floor. roof, or wall panels to which it.is attached may force the secondary member to deflect in the plane of loading with little or no torsional effects', -Unde r such conditions or 'when some other means of, bracing, is p r ovf.ded to prevent torsional, rotation in both' .the loading, and rebound phases of the response, such unsymmetrically loaded members may be', capable of, . performing well in the plastic range. Ho';ever, because of the limit'ed data presently' avaIlable, there is insufficient, .bas Ls ·for providing practical design guidelines .Ln this area. Hence.,: if a c as e involving unsymmetrical bending' with torsion:cannot be prevented in design; the maximum ductility,. ratio should be limited to 1. O. " .• 01
r
Furthermore, special precautions may have to·be taken to restrict the tor·'
sional-flexural distortions that can develop under unsymmetrical loading, thereby reducing the f Lexura Lcc apac Lt.y of, t.he member." .. __
,
.'j
5-45
'.:
"
.
TM 5-l300/NAVFAC P-397/AFR 88-22
5-33. Steel Joists and Joist Girders (Open-web'Steel Joists) Open-web joists are commonly used as load-carrying members for the direct support of roof and floor deck in buildings.' The design .of joists for conventional loads ·is covered by ·the "Standard 'Specifications, Load Tables and' Weight Tables for Steel Joists and Joist Girders!', .adopced ·by the- Steel Joist· Institute. For blast design, all the provisions of this Specification. are in force, except as modified herein. These joists are manufactured using either hot-rolled or cold-formed steel. HSeries joists are composed of 50-ksi steel in the chord members and either 36ksi or 50-ksi steel for the web sections. LH Series, DLH Series and joist girders are composed of joist chords or web sections with a yield strength of at least 36 ksi, but not greater· than 50 ksi. Standard load tables are available for simply supported, uniformly loaded joists supporting Ii deck and so constructed that .the top chord'is braced against lateral' buckling. These tables indicate 'that the capacity of'a particular joist may be governed by. either flexural or shear (maximum end,. reaction) considerations. As discussed previously,it is preferable in blast applications to select a member whose capac Lty. is controlled. by flexure rather than shear, which may cause abrupt failure. " The tabulated loads include a check on the bottom' chord as an axially' loaded tensile member and ·thedesign of the top chord as a column or beam column ..; The width-thickness ratios of the unstiffened or stiffened elements of ·the cross section are also limited to values specified in the Standard Specifications: The dynamic ultimate capacity of open-web joists may be taken equal to '1.7 a x c times the load given in the joist tables. This value represents the safety factor of 1.7 multiplied by a dynamic increase 'factor, c, and the' average strength increase factor, a (see Section 5-12). The adequacy o~ the section in reboundimust be evaluated. Upon'calculating the required 'resistance in rebound, r-/ru,"using the rebound chart in Chapter 3 (Figure 3-268), the.lower chord must· be checked as a column 'or beam column. If the bottom'chord of a standard joist is not adequate in rebound, the chord must be'strengthened either by reducing the' unbraced length or by increasing the chord area. The top chord must be checked as an axial tensile member, but in most circumstances; it will be adequate. "
The bridging members required by the joist specification should be checked for both the initial and rebound phase of ' the response to verify that they satisfy' the required.spacing.of compression flange' bracing for lateral buckling .. The joist tables indicate that the design of.some joists is governed by shear, that is, failure of the 'web bar members in tension or compression near the supports. In such cases, the ductility ratio for the joist should not exceed unity. In addition, the joist members near the support should be 'investigated for the worst combination of slenderness ratio and axial load under load reversal. , '
For hot-rolled members not limited,by'shear considerations; desrgn ductility. ratios up to ~he values specified in Section 5-35 can be used. The design 5-46
TM 5-l300/NAVFAC:P-397/AFR'SS-22 ductility ratio of Joists with light' gauge chord meinbers -should be-limited to
1.0. c,
The top and bottom chords should be symmetrical about a vertical axis. , If double angles or bars are used as chord members. the components of each chord should be fastened together so as to act as a single member. -'
","
..
:~
r';
: J..;':
-
,
.
~
.-.~.
'.1•
-,
:
.
'
.. OJ
,
,j ~
0
-,
.
r.
'~1 • ~."
r· ;
,', '
,.
..
t.
.;
"
j,
. l:
t"
a~_,_
.... ·t "
,
'v '
-",
"
"
5-47
TM 5-l'300/NAVFAC!'P-397/AFR' 88-22
..SPE.CIAL .CONSIDERATIONS, :.COLJlcF,ORMED STEEL, PANE:I
,
5-34. Blast Resistant Design of Cold-Formed Steel Panels :.
5 -34.1. .General
• • '.
'J
. -: .
"
..
,!
~.
.,' .(
Recent studies on cold-formed panels have shown that the effective width relationships for cold-formed light gauge elements under dynamic loading do not differ significantly from the static relationships. Consequently, the recommendations presently contained in the AISI Specifications are used as the basis for establishing the special provisions needed for the design of coldformed panels subjected to blast loads. Some of the formulas of the Specification have been extended to comply with ultimate load condit1ons and'to'
permit limited performance in the inelastic range. Two main modes of failure can be recognized, one governed by bending and the other by shear. In the case of continuous members , the interaction of the two influences plays a major role in determining the behavior .and the' ultimate capacity. Due to the relatively thin webs encountered in cold-formed members, special attention must also be paid to crippling,problems. Basically. the design will be dictated by the capacity in flexure but subject to the constraints imposed by shear resistance and local stability. 5-34.2. Resistance in Flexure The material properties of the steel used in the production of cold-formed steel panels conforms to ASTM A446. This ~candard covers three grades (a, b, and c) depending on the·yield point. Most commonly; panels are made of steel complying with the requirements of grade a, with a minimum yield point of 33 ksi and an elongation of rupture of 20 percent for a 2-inch gauge length. However, it is generally known that the yield stress of the material used in the manufacture of cold-formed panels generally exceeds the specified minimum yield stress by a significant margin; therefore, it is recommended that a design mini~um yield stress of 40 ksi (corresponding to an average strength increase factor of a - 1.21) be used unless th~ actual yield stress of the material is known. For grades band c 'which exhibit higher minimum yield points, an average strength increase .factor of 1.21 is also recommended. In calculating the dynamic yield stress of cold-formed steel panels, it is recommended that a dynamic increase f actor , c , of 1.1 be 'applied irrespective of actual strain rate and, consequently, the value of the dynamic design stress to be'used is a x c x f y - 1.21 x 1.1 x f y - 1.33 f y
5-26
and hence, f ds equals 44 ksi for the particular case of f y - 33 ksi. Ultimate design procedures, combined with the effective width concept, are used in evaluating the strength of cold-formed light gauge elements. Thus, a characteristic feature of cold-formed elements is the variation of their section properties with the intensity of the load. As the load increases beyond the level corresponoing to the occurrence of local buckling, the effective area of the compression flange is reduced; as a result, the neutral axis moves toward the tension flange with the effective properties of the cross section such as A (area), I (moment of inertia) and S (section modulus), 5-48
' "
, TK 5-1300/NAVFAC P-397/AFR 88-22 decreasing with load increase.' The properties of the panels, as' t';bulated, by the manufacturer, are related to different stress levels. The value' ~f S ' r~ferred to that of the effective s~c~ion,modulus at ultim~te and theyalue of I related to a service stress level of 20 ksi. Ln tihe case of panels fabricated from hat sect:lons and a flat'sh~et, two' sectio~ ';'odulii'are tabulated, s+ and S-, referring to the effective section modulus for positive and negative moments, respectively. Consequently, the following ultimate moment capacities are obtained: i
~p -
f
ds
S+ ,
5-27
!
•. ~ '.I
~n '- fdsS' where
'.
5-28
~p
ultimate posi'tive Dioment capacity for a I-foot width of panel
~n -,
ultimate negative moment capacIty for a I-foot width of panel
It should be noted that in cases where tabulated section properties are not available, the 'required properties may be calculated based upon the relationships in the AISI Design Specification. As for any single-span flexural element, the panel may be subjected to different end conditions, either si;'ply supported or fixe'd. The fixed- fixed condition is seldom found in practice since this situation is difficult to achieve in actual installat10ns. ' The simply fixed' condition is found because of symmetry in each span of a two-span continuous panel. For multi-span members (three or more), the 'response is governed by that of the first span which is generally characterized by a simply supported condition'at one support and a partial moment restraint at the other. Three typical ca~es can" therefore, be considered:' " (1)
Simply supported at both ends (single' span),
(2)
Simply supported at one end and fixed 'at 'the other (two equal span contin~ous member).
(3)
Simply supported at one end and partially fixed at the other (first span of an equally spaced multi-span element).
The resistance o£ the pan:l is a function of both the strength of the section and the maximum moment in the member. "
The ability of ,the panel to sustain yielding of its cross section produces significant moment re-distribution 'in the'contihuous member'which' results in an increase of the resist~nce of. the panel. The behavior of cold" formed steel sections' in flexure is nonlinear as shown in Figure 5-12. 'To s Imulat;e the bi.linea~ app roxfmat.t on to the resist,u;cedeflection curve, a factor of 0.9 is applied' to'the peak resistance. Therefore, for design purposes, the recommended 'resistance formula for a simply supported, single-span panel is given by: " '
0.9 x 8
~p/L
2'
- 7.2
2
~p/L
5-49
• i.
5-29
1M 5-l300/NAVFAC P-397/AFR 88-22
where r u is the resistance per, unit lengtp of, panel, and L is the clear or effective span"length . .'
t··
,
The recommended resi~t'ance' formula for a' simply-fixed, .single-span panel or fi!"st span of anI equally spaced, continuous panel i~ gfven 'b:r
;~:9 x.4 (Mun + 2~p)/L2 '". 3.6' (~n + 2~p)/L2
r u ;.
'5-30
5-34.3. Equivalent Elastic Deflection As previously mentioned, the behavior of cold-formed sections in flexure is nonlinear as shown in Figure 5-12. A bilinear approximation of the resistance-deflection curve is assumed for design. The equivalent elastic deflection XE ~s obtain~d by using the f~llowing equationi , 5-31 where
B
is a constant which depends on the support conditions and whose values
are as follows: B
0.0130 for a simply suppori:~ci element
B - 0.0062 for simply fixed or continuous elements 20 is defined es the effective moment; 'of, i,',ertia of the section at a service stress of 20 ksi. Th~ value of 1 20 is generally tabulated as a section property of the panel. The value 9f r u is obtained from Equation 5-29 or 530 ' 1
'.,
5-34.4. Design foi Flexure'
t-
!
analysis of the'panel's behavior, ,the be evaluated by using a load-mass value applicable to all support condifor' the equivalent single-degree' in~o Equation 5-15:
TN = 2" (0.74 mL/KE)1/2 where
m -
5-32
~/g is the unit m~ss of ~he panel, ruL/X E .is the equivalent elastic stiffness of the system ,
'
5-34.5. Recommended Ductility Ratios Figure 5-12 illust~ates the nonlinear character of the resistance-deflection, curve and the recommended bilinear approximation. The initial slope of the actual curve is fairly linear until it enters a range of marked nonlinearity and, finally, a point of instability, ,However, excessive def Lec t.Lons ,cause, the decking to act as a membrane in tension (solid' curve) and, consequently, a certain level of stability sets in. It should be noted that, in,order tu use, the procedure outlined ,in this section~ care must be taken to adequately connect the ends of the decking so' that it' can achieve the desired level of t
eric Lorr-memb r ane action.
ed in Section
5·~8.
A d~_~(".ussi._')n :,F c r nne ct.o r s at; er.. d panels is r·::esentWhen t:ensi.on-mer:,brar:,::, .act , I on .Ls not p re senr , increased
5-50,
TK' 5-l300fNAVFAC P-397/AFR 88-22
deflection will result in a significant dropoff.in resistance as illustrated by the dott~d curve in Figure 5-12. Two limits of deformation are assigned, depending on, end-anchorage condition of the panel'. For panels having nominal end anchorage, that is, where tension-membrane action is minimal, the maximum deflection of the pane L is X o ' as illustrated in Figure 5-12, and is defined by:
"
r
5-33 .. -, ,. For pane Ls with sufficient' end anchorage to 'permi't tension-membrane action, the 'maximum deflectioh of the panel is Xm,'as'illustrated·iri Figure 5-12, and" is defined by: ' ' ... .,
,
5-34.6. Recommended Support Rotations In order to restrict the magnitude of rotation at,the supports, limitations are placed on the maximum deflections X o . and Xm as follows,: For elements where,tension-membrane'act10n is not present: !
,~
.
"Xo ~ 1;/92 or 9 max' -,1. 25 'degrees'
For elements where' tension-membrane 'action is present: .:
-',~ ,- L/92 ... -,', or 5-34': 7, Rebound ~
9ma~
"
~. "
'
- 4'~egrees
5-36
~,
" :
.' .! .,
.
':"~
.' f
" ,' , r., ' ' Appropriate dynamic response charts for one-degree-of-freedom systems in the, elastic or elasto-plastic range under various dynamic'loads are given 'in Chapter 3. The problem of rebound should be considered in the design of decking due' to' the different section properties of the'panel, depending' on whether the hat section or the rlat sheet is'incompression. Figure' 5-13 presents the maximum elastic resistance in rebound as a function of T/T N: While the behavior of the panel in rebound does not often control, the' de s Lgne r should be aware of 'the' problem;', in any" event, there is a need for providing connections capable 'of resisting uplift' or pull-out forces due' to load reversal in rebound. ' " , .'
.
5-34.8, Resistance in Shear Webs with hit in excess of 60 are in common use among co Ldv f orme'd members' and the fabricatio~~process'makes it impra~tical to'use' stiffeners. The desfgn • web stresses must', therefore, be limited to ensure adequate stability without' the aid of stiffeners, thereby preventing premature local web failure and the' accompanying loss of load-c&rrying capacity. ,
,
The possibility of web buckling due to bending stresses exists and the critical bending s t re ss Ls ~~ve~: by Equation 5-'37: ' • fer ~. 640,OOO/(h/t)2 ~ '
..
!y
, 5-51
.
5-37
TM5-1300/NAVFAC P-397/AFR 88-22 By equating f c r to 32 ksi, which is a stress close to the yielding of the material, a value hit - 141 is obtained. Since it is known that.webs do not actually fail at these theoretical buckling stresses due to the development of post-buckling strength, it can be safely assumed that webs with hit ~ 150 will not be susceptible to flexural,buckling. Moreover, since the AISI recommendations prescribe a limit of hit - 150 for unstiffened webs,_ this type of web instability need not be considered in·the design. Panels are generally manufactured in geometrical proportions which preclude web-shear problems when used for recommended spans and minimum support-bearing lengths of 2 to,3 inches. In blast design, however,.becausE!'of the greater intensity of the. loading, ,t~eincrease in required flexural resistance of ~he panels requires shorter spans. In most cases, the shear capacity of a web is dictated by instability due to either (1) (2)
Simple shear stresses Combined bending,and shearing stresses
For the case of simple shear stresses, as encountered at e~d supports, it is
important to distinguish three ranges ,of behavior depending on the magnitude of hit. For large values of hit, the maximum shear stress is dictated by elastic buckling in shear and for intermediate hit values, the inelastic buckling of the web governs; whereas for very small values of hit, local buckling will not occur and failure will be caused by yielding produced by' shear stresses. This point is illustrated in Figure 5-14 for f ds - 44 ksi. The provisions of the AISI Specification in this area are based on:a safety factor ranging from 1.44 to 1.67 depending upon hit. For blast-resistant design, the recommended design stresses for simple shear are based on an extension of the' AISI provisions to comply with ultimate load conditions. The specific' equations for use ;n design for f ds - 44, 66 and 88 ksi are summarized in Tables 5-5a, 5-6a, and,5-7a, respectively. At the interior supports of continuous panels, high bending moments combined with large shear-forces are pres~nt, and webs must ,be checked for buckling due to these forces. The interaction formula presented in the AISI Specification is given in terms of the, allowable, stresses rather than critical stresses which produce buckling. In order to adapt this interaction formula to ultimate load conditions, the problem of inelastic buckling und~r combined . stresses has been considered in the development of the rec9mmended design data. In' order to minimize the amount and complexity of design,calculations, the allowable dynamic'des~gn shear stresses at the interior support of a continuo!'s member have ,been computed for different depth-thickness ratios for 'f ds - 44, 66', and 88 ksi',' and tabulated in Tables 5-5b, 5-6b, and S-7b, respec-· tively. . '. 5-34.9. Web Crippling In addi.tio~' to shear problems, concentrated loads" or reactions at panel supports, applied over relatively short lengths, can produce load intensities that can cripple unstiffened thin webs. The problem of web crippling'is rather complicated for theoretical analysis because it involves the following: 5-52
TM'5-1300/NAVFAC.P C397/AFR 88-22
e,
Nonuniform stress ,distribution under tlieapp1ied load and the adjacent portions of the web
(1)
i
.• ,
";.",
-
,"I'
",,::•.
3~.:",
_ ..0
....
.r,......
'_Jh-
".
•'-~',J.
J f'
-'(2)" . Elastic' and inelastic' stability. of;'the web' element' .'" (3)
Local yielding ,in the intermediate region of'load application
(4)
The bending produced by the eccentric load (or reaction) when it is applied on the bearing flange at a distance beyond the curved transition of the web
The AISI recommendations have been developed by relating extensive'experimental data to service loads with a safety factor of 2.2 which was established taking into account the scatter in the data. For blast design of cold-formed panels, it is recon~ended that the AISI values be multiplied by a factor of 1. 50 in order to re,late the crippling loads to ultimate conditions with sufficient provisions for scatter in test data. For those sections that.provide a high degree of restraint against rotation of their webs, the u1t:imate crippling loads are given as follows: Allowable ultimate end support reaction ' 0.558 (N/t) 1/2 ] Qu - 1.5 fdst 2 [4.44,+
5-38
Allowable ultimate interior support reaction Qu - 1. 5 f:ds t 2 [6.66 + 1.446 (N/t)1/2] where
Qu f ds
5-39
ultimate support reaction dynamic design stress
N-
bearing length (in. )
t
web thickness (in)
The charts in figures 5-15 and 5-16 present the variation of Qu as a function , of the web, thickness for bearing lengths from 1 to 5 inches for fds - 44 ksi for end and interior supports, respectively. It should be noted that the values reported in the charts relate to one web ~n1y, the total ultimate reaction being obtained by multiplying.Q u by the number of webs in the panel. For design, the maximum shear forces and dynamic reactions are computed as a function of the maximum resistance in, flexure. The ultimate load-carrying capacity of the webs of the panel must then be compared with these forces. As a general comment, the shear capacity is controlled for simply supported elements and by the allowable design shear stresses at the, interior supports for continuous panels. In addition, it can be shown that the resistance in shear governs only in cases of relatively very short spans, If a design is controlled by shear resistance, it is recommended that another panel be selected since a flexural failure mode is generally preferred.
TN 5-l300/NAVFAC P-397/AFR 88-22 •
5-35
Summary ,of Deformation Criteria for Structural ,Elements
Deformation criteria are summarized in Table 5-8 for frames, beams and other structural ele~ents including cold-formed steel panels,open-web joists,and plates,
.
('
"
-". ~
" ';;',
,
'
.)
:
'-, 'I'
.
-",
,. "
"-. ~
;"
"v •
. . '-:1:. ,. ' .;f
-
.~
.,
~•.'>
"
5-54
.\
.e
.
, . ~.
-::-
'.
.
;"
. ' , . . 0'
STIFFNESS . . CORRESPONDING 10 120 '
-,
,
,,
,',.
/ . . . , P.EAK RESISTANCE:
.'
"
.
WITH TENSIONMEMBRANE
ILl
o Z
,.'
I-
,
en en
,.
ILl
a:
"
WITHOUT "TENSION - MEMBRANE .'
'.................
"
.
-r-
,.,
DEFLECTION
:
-,
,.,.
,-.
Figure 5-12
Resistance-deflection curve for a typical cold-fonned section 5-55
h ..
lc
o
:.;,
I.OO~ t tt;,
,P~T
. IH- .:+l
z :::> o m
-:::
~
i= ~
...J ILl II::
fr
..., ...
'"
ILl U
.1 ; ! '.I. IT""" ,
':"I'~'
~
1Il
f3
ItH+#I uuuuu f
J'" ....
ILl
11:.
Xm
IH
f~rnt~, tlH
i-r
fl= 1.0
+
-t
.40
.1 .• i.I·.;;IL .••
"":::1:';:1:':1:::: ,;
·.·d··',I;·•
IJ-= 1.25
tttH-1 tli
~c~
··' •• ·,I·', •• ··,tl...·I.· ··I·,··j·,.,I... ,/..
., ".W~ 'illllilllJ.ill .
. .,.
j
,
'ViE, ;...,.. " . . ".
..· 1 .. ·1 '..·1 ..·'· ! I""""· ...., _; ..;·I· !t~~ .' :
····t.. 'lc·±··". ...... 1+ ...
'.
.
I;
,
IJ-= I. 75
,
.+l-l
d:
•. I· ..•li; .. I.".
1J-=3.0 ,.,.=6.0
,..l
.l;.
i.u .'....Lt"..=r:';· . ·W·...;.:.~·1Hlhti'.· "., ....., ::I::r:::I:~J:~2J8; +-r:-: tnt tin , -t A"'~ ":1"'r :~;+t~)i.fHtt
·l....
·.I·!'.:I.. .,,, ,···4:::r::: 1;;:I':f~!'1~tt'~. 0.4
-i--+-:.....j... - ~._ •...:.. .;..;..;.t ,
t
'.
t,
I
"
I'·
"
.•
1,,Ilili
0.6 0.8 1.0
2.0
4.0
20
6.0 8.0 10
T Tn
.
- . ~
'40
.- .
..
60
"
,
Figure 5-13
e
,
uu: . . ." Ii;]:
••. ·'.·.·I·.·JI··..... ;·
11:
o ~ :::> o
I I
RESISTANCE - DEFLECTION
,,
H
... I+: :1~;:~"'l~"~" ". ;~:--FEi::::I I·'··t··,·
.60
;
XE
,.
Z
y
, I
-t-
o
~
't>
:t
u
.
PRESSURE - TI~E
,
Elasti c rebound of single-degree-of-rreedom system
e
e
.. '
:3 5
.
, ,..
,
._.~
,
..
:3 ~ YIELDING IN SHEAR
i I
00
I,
-
2r
~
2
o
,
-
I •
;
i I I
-,
i
5
~;'O' hit I
" " " 1.07x"10
I
10 'COMBINED BENDING 6 SHEAR
I I
I I o
I I
•
Figure 5-14
40
60
~hlt)Z
I
I .
i
•
I
0
' .1
!
I
; I
20
5'
1--~ :
I
"
.
.1
o
I
5
.
I
,
I I
•
.
INELASTIC I ELASTIC BUCKLING SHEAR BUCKLING
I,
0.50 fds
....
80
! "
I 0
-------
. 120
Allowable dynamic (design) shear stresses for webs of cold-formed members (f ds - 44 ks r) 5-57
18
-N=5 -N=4 16
N=3 ,.
~.
VI Q.
.
.
14
..
-
:0: ' .
I
.' .
12
.
.
. ,
,
a:::
.... Q.
10
o
"
u
« .... a:::
8 ,
.
i
..
'& W W
,,
,
4
IlJ II llf)
II vii WI 'I
6
'/j
h 1/'
V !J 1// IPh V
. z· I-
I IjIJ
'I I N=2 'lII -N =I I.'
.
2 I
0
o
,04
.08 .12 SHEET THICKNESS (IN.)
.16
.20
Figure 5-15 • Haxim\ID end support reaction Cor cold-Col'lled steel sections (C os ~ 44 .ks L) . 5 -58·,
.
~
..
36 -N=5
.. . .t 'N=4 .\
32
II VI,
.
-
28
( /)
-"
.. '_ ,r. ~ ..'
:lO:
0
24
lD
,
~
II:: ILl l1.
.
20
.
"J;/ II hIf/ I
Z
0 I-
o
~
16
1$II
ILl II:: ~
!JVI /
~
~
x
12
~
~
~
VI
~ 'l"1
8
A
~V
~~
4
.
o
~ '1/ /
..
,
-N ::;3.
f vI, -N=2 II 1// N=I If! VI
"
e,
'
'l .04
V
';08
.12
.16
.20
SHEET THICKNESS (tN.) .\,'
'("~,
.,;.;.
t'
Figure 5-16 Maximum interiorsuppart reaetion for cold-formed steel sections (f ds • 44 kst)
TK 5-1300/NAVFAC P-397/AFR 88-22 Table 5-5
.
Dynamic Design Shear 44 ksi)
~tress
for Webs of Cold-formed Members (f ds'-
(a) Simple Shear (hit) S
57
f dv ~ 0.50 f ds S 22.0 ksi
57 «hit) S
83
f dv :- 1.26 x.10 31 (hIt)
83 <. (hit), S 150
f dv - 1. 07 x 10 51 (hit)
(b) Combined Bending and Shear f dv (ksi)
(hit)
10.94 10.84 10.72 10.57 . 10.42 10.22 9.94 9.62 9.00 8.25 7.43
20 30 40 50 60 70 . 80 90 100 110 120
.! .. <.
5-60
TN 5 21300!NAVFAC Po397/AFR.88~22
Table '5-6.
Dynamic Design' Shear Stress for \lebs of 66 ksi)
Co1d-f6rmed'~Members
(a) Simple Shear
.f
(h/t) S
47
47 < (hIt) S
67
67 < (hit) S 150
~
f dv - 0.50 f ds ,S 33 ksi f dv - 1.54 x 10 3/(h/t) f dv - 1.07 x 105/(h/t)
,".'
(b) Combined Bending and Shear
·rr. .~ fdv(kid)
(h/t) 20 30 40 50 60 70 80 90
100 110 120
,..
16.41 16.23 16.02 15.75 15.00 14.20 13.00 11.75 10.40 8.75 7.43
,.
, .
(f ds
=.
TH 5-1300/NAVFAC Table 5-7.
P-3~7/AFR
88-22
Dynamic Design.Shear Stress for Webs of Cold-formed Hembers' 88 ksi)
(fd~'
",0,
(a) Simple Shear (hit) S
41
f dv - 0.50 fds S 44-ksi
41 < (hit) S
58
f dv - 1.78 x
10 3/(h/t)
_f dv - 1. 07 x 105/(h/t)
58 < (hit) S 150 (b) Combined Bending and Shear (hit)·
fdv(ksi)
-20 30 40 50 60 70 80 90 100 110 120
21.60 21.00 - 20.00 18.80 17.50 16.00 14.30 12.50 10.75 8.84 7.43-
,
5-62
TK 5-l300I.NAVFAC.P-397/AFR 88-22
Table
~-8'
. Summary of Deformation ,Criteria
Highest level of Protection (Category No.) Additional .* Specifications
Element
Beams, purlins,
spandrels or girts
..
Deformation Maximum Type ** Deformation
a
:
1
2° 10
I!-'
.
...
2
a
12° 20
'I!
Frame
0
1
structures
Cold-formed steel floor and wall panels
1
..
Without tensionmembrane ,action
I!
With tension-
a
membrane action
I!
, .' 1
Open-web joists
H/25 2°
a t a
1. 25° 1. 75
S .,
a I!
a 1
I!
a 2
.a
*
-** t
6
2° 4
I!
Joists controlled by , maximum end reaction Plates
4°
,
I!
.
1° 1: 2° 10' 12° 20
maximum member end rotation (degrees) measured from the chord joining the member ends . . .
o-
relative sidesway deflection between stories.
H
story height.
I! -
ductility. ratio (Xm/XE)'
as defined in Chapter.l. whichever governs. individual frame member.
5-63
TH' 5-l300/NAVFAC P-397/AFR 88-22'
SPECIAL CONSIDERATIONS; BLAST DOORS 5-36. Blast Door Design 5-36.1. General
..
~;,
This section outlines procedures for the 'design of steel blast doors. Analytical procedures for.the design of the individual elements' of the blast door plate have been pre senced in earlier sections of this chapter.
·In
addition to the door plate, door frames and anchorage, reversal bolts; gaskets and door operators are discussed.
Blast doors are categorized by their .
functional requirements and method of opening. 5-36.2. Functions and Methods of Opening 5~36.2.l.
Functional Requirements
Blast doors may be designed to contain an accidental explosion from within a structure so as to prevent pressure and fireball leakage and fragment propaga, . tion. Blast doors may also be designed to protect personnel and/or equipment , from the effects of external blast-loads. In' this case, a limited amount of blast pressures may be permitted to leak into the protected area. In most cases, blast doors may be designed to protect the contents of a structure, thereby negating propagation when explosives are contained within the shelter. ! j
5-36.2.2. Method of'Opening Blast doors may be grouped based on their method of opening, such as: (a) single leaf, (b) double leaf, (c) vertical lift, ,and (d) horizontal-sliding. 5-36.3. Design Considerations 5-36.3.1. General The design of' a blast door is intrinsically related' to i-ts function during and/or after an explosion. Design considerations include whether or not the door should sustain permanent deflections, whether rebound mechanisms or fragment· protection is required, and whether pressure leakage be tolerated. Finally; the design pressure range may dictate the type of door construction that is to be used, including solid steel plate or built-up doors. 5-36.3.2: Deflections As stated in Section 5-16.7, plates can sustain a support rotation of 12 degrees without failing. This is applicable' to blast doors providing that the resulting plate deflection does not collapse the door by pushing it, through the opening. However, deflections may have t o z be 'limited if 'the mechanism used to open the door after an explosion is required to function." In addition, if a blast door is designed with a gasket so as to fully or nearly contain the pressure and fireball effects of an explosion, then deflections should be limited in order to ensure satisfactory performance of the gasket.
5-64
TH 5-l300/NAVFAC P-397/AFR 88-22.
5-36.3.3. Rebound Mechanisms Steel doors will be subjected to relatively large stress reversals caused. by rebound. Blast doors may have to transfer these reversal loads by means of retracting' pins or "reversal bolts." These heads can be, mounted on any edge ,; (sides ,top, 'or bottom) of a doo rp Lace> Reversal bolts can be designed as an integral part of the panic hardware assembly. or, if .tapered, they can' be, utilized in compressing the gasket around a periphery. The magnitude of the rebound force acting on the blast doors is.discussed later. 5-36.3.4. Fragment Protection A plate-type blast. door, or the plat'e(s) .of a built-up blast door, may be s Lzed to prevent fragment penetration. 'However, when the blast door is subjected to large blast loads and fragments, a supplementary fragment shield may be necessary since the combined effects of· the fragments and pressures may cause premature door failure due to the notching effects produced by the fragments. Procedures ,for predicting the characteristics of primary fragments such as impact, velocity, and s Lz e of fragment are presented in Chap t e r ·2. Methods for determining the depth of penetration of fragments into'steel are given in Section 5-49. .'. 5-36.3.5. Leakage Protection (gaskets) Blast doors may be designed to partly ,or fully contain the pressure and fireball effects of an explosion in which case gaskets may' have to be.utilized around the edge if a door or its opening. A sample of a glisket is illustrated in Figure 5-17. This gasket will have to be compressed by means of a hydraulic operator which is capable of overcoming a force of 125 pounds per linear inch of the gasket. This gasket is made of neoprene conforming to the material callouts in Note 2 of Figure 5-17. 5-36.3.6. Type of Construction Blast doors are formed from either solid steel plate or built-up steel construction.
Solid steel plate doors .are usually used for the high pressure ranges (50 psi or greater) and where the.'door span is relatively short. Depending on plate thicknesses, these doors may be used when fragment impact is critical. These plates can range in thickness of I-inch or greater. For thick plates, connections using high strength bolts or socket head cap screws are recommended in lieu of welding. However, the use of bolts or screws must preclude the passage of leakage pressures into or, out of the structure depending on its use.
Built-up doors are used usually for the low pressure range and where long spans are encountered.' ,A typical built-up blast door may consist of a peripheral frame made from steel channels with horizontal channels serving as intermediate supports for .the .interior and/or exterior s t e eL cover plates. The pressure loads must be transferred to the channels via the. plate. Concrete or some other material may be placed between the plates in order to add mass to the door or increase its fragment resistance capabilities.
5-65
TH 5-l300/NAVFAC P-397/AFR 88-22
5-36.4. Examples of Blast Door Designs 5-36.4.1. General In order to "Ll.Lus t ra t e the relationship between the function of a blast door and its design considerations, four examples are presented'in the following section.' Table 5-9 lists the design requirements of each of the above door examples. 5-36.4.2. Door Type A (Figure 5-18) This blast door is designed to protect personnel and equipment from external blast pressures resulting from an accidental explosion. The door opening measures 8-feet high by 8-feet wide. It is a built-up double-leaf door consisting primarily of an exterior plate and a thinner interior plate both welded to a grid formed by steel tubes. Support rotations of each element (plate, channel, tube) have been limited to 2 degrees in order to assure successful operation of the panic hardware at the door interior.' The direct blast 'load is transferred from the exterior plate to tubular members which form the door ,grid. The grid then transfers the loads to the'door frame at the center of the opening through a set of pins attached to the top and bottom' of the center mullions of the grid, At the exterior, the loads are transferred to the frame through the hinges which are attached to the exterior mullions and the frame. The reversal loads are also transferred by the pins and by the,built-up door hinges. The center pins are also operated by the panic hardware assembly. 5-36.4.3. Door Type'B (Figure 5-19) This blast door is designed to prevent propagation from an accidental explosion into an explosives storage area. It is a built-up, sliding door protecting an opening II-feet high by 16-feet wide, consisting of an exterior plate and a thinner interior plate. These plates are welded to'vertical Sshapes which are spaced at IS-inch intervals. This door is designed to act as a one-way member, spanning vertically. Since flange buckling of the S-shapes is prohibited in the presence of the outer and inner plates acting as braces, the composite beam-plate arrangement is designed for a support rotation of 12 degrees.. The yield capacity of the webs of theS-shapes in shear (EquatIon 516) ,as well as web crippling (Section 5-25), had to be considered in the design. This door has not been designed to resist reversal or rebound forces. 5,-36.4.4. Door Type C (Figure 5-20) This single-leaf blast door is designed as part of a containment cell which 'is used in the repeated testing of explosives. The door opening measures 4-feet 6-inches wide by 7-feet 6-inches high. It is the only door, in these samples, designed elastically since it is subjected to repeated blast loads. It consists primarily of a thick steel door plate protected from test fragments by a mild steel fragment shield. It is designed as a simply-supported (four sides) plate for direct internal loads and as a one-way element spanning the door width for rebound loads. It is'equipped with a neoprene gasket around the periphery (Figure 5-17) as well as a series of six 'reversal 'bolts designed to transfer the rebound-load into the door frame. The large thickness of ,the door plate warrants the use of high-strength, socket head cap screws in lieu 5-66
TK 5-l300jNAVFAC P-397/AFR 88-22
of welding to connect the plate to the reversal bolt housing as well.as to the fragment shield. • •.r ,
5-36.4.5. Door Type D (Figure 5-21) This single;-leaf blast door is designed as part of a containment structure which is used to protect nearby personnel and·structures in the event of. an , accidental explosion. The door opening measures·4-feet wide by 7-feet high. It is designed .as simply-supported on four sides for direct load and 'as a oneway element spanning the door width for rebound loads. It is equipped with a. neoprene gasket around the periphery and a series of six reversal bolts which transfer the rebound load to the door frame. The reversal bolt housing and bearing blocks are welded to the door plate. Excessive deflections of the door plate under blast loading would hamper the sealing capacity of the gasket. Consequently, the· door plate design rotation is limited' to:2 degrees'. 5-36.4;6. Other
Type~
of Doors
Another type of blast door design is a steel arch or "bow" door: The tension arch door requires'compression ties to ·develop the compression reactions from the arch. The compression arch door requires tension members to develop the tension reactions from the arch. These-doors are illustrated in Figure 5-42. 5-36.5. Blast Door Rebound Plate or element rebound can be determined for a single-degree-of-freedom system subjected to·a triangular pulse (see Figure 5-13). However, when a system is subjected to a bilinear load, only a rigorous, step-by-step dynamic analysis. can determine the percentage of elastic rebound. In lieu of.a rigorous analysis, a method of'determining the upper bound on the rebound force is presented here. Three possible rebound scenarios are discussed. Figure 5-22 is helpful in describing each case. (a) ;
..
Case I - Gas load not present (Pgas -0). In this .case, the" required rebound resistance is ootained from Figure 5-13. I·
(b)
Case II - t m S t i In this case, the required rebound resistance is again obtained from Figure 5-10. This procedure, however, can overestimate. the rebound load.
(c)
Case III - ~ > ti Figure 5-22 illustrates the case whereby the time t? reach the peak response, .tm, is greater than the point where the gas load begins to act (t i). Assuming that the gas pressure can be considered constant over a period of time, it will act to lower the required rebound resistance since the resistance time curve will oscillate about the gas pressure time. curve. In this case, the upper bound for the required rebound resistance is: 5-40
r - r u - Pga s
However, in all three cases, it is recommended that the required rebound tesistance be at least equal to 50 percent of the peak positive door response. 5-67
L
TM 5-l300/NAVFAC P-397/AFR 88-22
5-36.6. Methods of Design 5-36.6.1. General Techniques used for the design of two types of blast doors will be demonstrated. The first technique is used for the door illustrated in Figure 5-18 while the second is used for the door shown in Figure 5-21. Detailed procedures for the design of plate 'and beam elements, as well as the related design criteria, are presented in earlier section of this chapter and numerical examples are presented in Appendix A. 5-36.6.2. Built-up Door The built-up steel ,door shown in Figure 5-18 is constructed by welding the steel plates to the steel tubular grid (fillet welded to the exterior plate and plug-welded to the interior plate). The heavy exterior plate is designed as a continuous mem~er supported by the tubes. The horizontal tubes, in turn, are designed as simply supported members, transferring load to the vertical tubes. The interior tubes are also designed as simply supported elements which transfer the direct and rebound loads 'to the pins while the side tubes transfer the direct load to the door frame proper and rebound loads to the hinges. The exterior tubes are also designed as simply supported elements with the supports located at the hinges: 5-36.6.3. Solid Steel Plate Door The steel plate of the blast door shown in Figure 5-21 is initially 'sized for blast pressures since no high speed fragments will be generated in the facilit~The plate is sized for blast loading, considering the plate to be simply supported on four edges. The direct load is transferred to the four sides of the door frame. In rebound, the plate acts as a simple beam spanning the width of the door opening. The rebound force is transferred to the six reversal bolts and then into the door frame. The door frame, as illustrated in Figure 5-19, consists of two units; the first unit is imbedded into the concrete and the second unit is attached to the first one. This arrangement allows the first frame to be installed in the concrete wall prior to the fabrication of the door. After the door construction is completed, the sub frame is attached to the embedded frame and, thus, the door installation is completed.
,
,
..
5-68
•
•
·l
i
CONTINUOUS GASKET (SHOWN COMPRESSED)
e
0;
7/8 "
hit.. GASKET,a I, - DOOR FRI.\ME \
r-11t..
"
,
SURFACE OF DOOR GROOVE
" BLAST DOOR
.,/1
,
,~
=
.
.1 : •
'.,
DOOR OPENING
J...-'--DOOR FRAME
. GASKET>'
DOOR. GROOVE
I
'.
'"o '" '"
DETAIL (TYPICAL AROUND PERIPHERY OF DOOR)
I :Z'~= 1 ~OT
, !;
1"
NOTES FOR GROOVE a GASKET I. ALL SURFACES OF DOOR GROOVE SHALL BE MACHINED TO 125 MICROINCHES AND SHALL NOT BE PAINTED.
.
VULCANIZE
2.
GASKETS FOR ALL DOORS SHALL t-lEET THE SPECIFICATION.REQUIREMENTS AS PER ASTM 02000-770 AND AS INDICATED INTHE"'FOLLOWING LINE CALL OUT; 2BC 520 AI4 BI4 CI2 F17. .
3.
THE FORCE REQUIRED TO'CLOSEDOOR IS APPROXIMATELY, 125 P,OUNDS PER L'INEAR. INCH OF GASKET. . .,
'
"
GASKET SPLICE
,
,
,;
Figure 5-17
Gasket det atI for blast door
r-----
a'-d'
'.' ,
I
BLAST LOAD ..
• r-
-1 t I I ~-
HINGE
-l
1 r-
J-!wl~
I
I======.:J
iI
I
-
•
- ---r------
•
I I
1 1-
eli- DOOR
DETAIL
1====-'9 I
l-
@Y~-
I I
t=-===---=i 1=====1 L I.!---@ I - .J-@ I: ~
--j
-ll==-==d I I
--, -
I
I
1==_===1
l-
I
I
~ B
I
-<;> .--/
00
SECTION
1=====1 F=-==l lI _ I I - I _I .... B I Bf"" - --i11=-=== =j' I=-!:;==~ fT~I
HINGE
! I~
_.J
,_!+~
--.1__ i:'oj ;
I 1-
~
B-B
®
"1
'" ....qo
-
'~"
L_
, :;~""
..
:
.
--
--
,
INTERIOR DOOR ELEVATION
-
-,
-
-'
DETAIL
SECTION _A-A LEGEND:
®_
0-_·CHANNEL'
C9~ MECHANICAL
_,(~y ~
e
".
~ -~STRUCTU~AL,TUBE
"
EXTERIOR PLATE
0-
INTERIOR PLATE
(9-
PINS FOR LOAD TRANSFER AND PANIC HARDWARE
Figure 5-18
REINFORCING BAR
®-
o
STIFFENER PLATE
~ BENT DOOR F,RAME PLATE
CD -
e
ANCHOR
PANIC HARDWARE
&lilt-up double-leaf blast door with rr-aae buil t into concrete
e
1
•
e
• .'. ~
, ... ., ,:.« ,
;.'
"'~"'"
.:
<1
,II
""
,,",,---
,,
,'
'~
I~'
-::------ - - - - "
I '
'",
....-.J
~
'a
- :.'u'·-11
c
b. '/
/
-
-:
,.<,.'
L--., _:!~6'~'0'"
r
.'
II I
I ",.
""
0'
DETAIL
.. ·':1
""
,
LOAD
f5
i
1
f ' BL~S/
SECTION
I
,.
r.ll.J
J'
-I
I . '.
I
:I'
I
",
"
I
,I
,
<,
c
:.~~
>// .: ' .:,,
•.
i
.INTERIOR It
A""-.PLUG WELD
• \ EXTERIOR It-:l
,..
EXTERIOR FACE OF SLIDING DOOR . "
DETAIL
"
"
~igure
5-19
Horizontal sliding blast door
SIO
'.
4
1 _
61~
-1
I
BLAST .. LOAD
7D -,
, ,
I I
, I I
,-
,
cD
-r<-'
---..<=11=11-11 :
'",
...
....
, SECTION
,
,
LEGEND:
(~}
® CD· .•..
~-
ELEVATlON-'OF B-tAST DOO'R' ,- , '.4,
'r
~
~,,,
.'\
.:'~ t
••
Figure 5-20
e
-. .:
HINGE
REVERSAL' BOLT HOUSING
~
CROSS ... SECTION) ':
CONTINUOUS GASKET
fH\ _REVERSAL BOLT (CIRCULAR
SOCKET - HEAD CAP"SCREWS
G)-ROD TOmCKING
.c~Y-!"ILD
0, ®.
0-
STEEL FRAGMENT SHIELD
THICK BLAST-DOORP'LATE
MECHANI~M
0-GAGE METAL COVER,PLATE
® ~600R FRAME STIFFENER It ,
DOOR FRAME
r:·
., " '
Single-leaf blast door with Crapent shield (very high pressure)
e
e
e
•
4 1_0"
.u.. :- .,
)-
10'
:!
,...
0
01
~
I
)-
I
I I
'~il,
~
F'
' I~f@ G
(£
I
~7::'
'"
U
'I
I"
"
'%:
' "" !I
tI
I~
B'LAST LOAD
'SECTION
I I
I
I
-
&: k *'
I
I '!-:--
Ifl
--'
I
.....
,
~
II L
oilo '
,
.
LEGEND:
L.J
i"ir
J..
@-
STEEL FRAME'EMBEDDED IN CONCRETE,
® - CONTINUOUS GASKET
...!l
@@-
ELEVATION OF BLAST DOOR
BEARING'BLOCK BLAST DOORPLATE
® - DOOR HINGE ® - REVERSAL BOLT HOUSING ® - REVERSAL BOLT @-
BAR CONNECTED TO CLOSURE' MECHANISM
CD - STEEL FRAME EQUIPPED WITH ,eiU\ST DOOR 0- LIGHT GAGECOVER PLATE Figure 5-21
,'f,'
• <7•• "
,-
,
- "'=. I~
@
1
'",
-
I
I
'l
-I
~;,
,.....,
I
I'
:'
.u
I
} "
•
Single-leaf blast door (high pressure)
%..)II ~.. .6 .. ... v ', • •.
LOAD-TIME
P LOAD RESISTANCE - TIME
7" .J.
TIME ,~
,
,
,
P = PEAK LOAD PIlOS = GAS PRESSURE ru
r
ti tm
= = = =
PEAK PO,SITIVE RESPONSE REQUIRED REBOUND RESISTANCE TIME AT WHICH SHOCK AND GAS LOADS INTERSECT TIME TO REACH MAXIMUM RESPONSE
'.
.
~.
Figure 5-22
.'
Bilinear blast load and single-degree-of-freedoo response for determining rebound resistance 5-74'
.
TK 5-1300/NAVFAC P-397/AFR 88-22
"< ,,'
",
, Table 5-9
Design Requirements for Sample Blast Do~rs' ~1
Door Description :,
4
,"
.
A
5-18
Double-Leaf
B
5-19
Sliding
G
5-20
Single-Leaf
D
5-21
Sirigle~Le~f
.. '.
'
.- . ,
_._
Rebound Fragment Mechanisms Protection
Permanerrt;
Method of Op"ning
'
Design Requirements Deflections
Door Figure
.
None Limited Large
Yes
X
X
.- -
X
.'
'
""
X X
'X"' "
X
,
.
,
"
N.o '
Low
X
X
X
X
"
.1" ,
"
High
X
X
X
X
.,
,. ,
,
5-75
Yes "
-
"
Required
No
X
"
I
,
-.
Protection
.
)
Level of Leakage
TK 5-1300/NAVFAC'P-397/AFR,88-22 .
COLUMNS AND
BEAll COLUMNS
5-37. Plastic Design Criteria 5-37.1. General The design criteria for columns and beam columns must account for their behavior not only as individual members but also as members of the overall frame structure.
Depending on the nature of the loading, several design cases
may be encountered. Listed below are the necessary equations for the dynamic design of steel columns and beam columns. 5-37.2. In-plane Loads In the plane of bending of compression members which would develop a plastic hinge at ultimate loading, the slenderness ratio l/r shall not exceed the constant (Cc) defined: as: , Cc - (21r2E/fds)~/2 where,
modulus
E
~f
5-41 elasticity of steel (psi) I
dynamic design
s~ress
(see Section 5-13)
The u1timat~ strength of an axially loaded compression member shall be taken as:
5-42 where
A-
gross area of member,
(1 - (K1/r)2 'I Fa -
2C c 2) f ds
5/3 + '3 (Kl/r) IBC c - (Kl/r) 3 I BC c 3 '
5-43
, and
largest effective slenderness ratio listed in Table'5-10 or 5-11
Kl/r -
5-37.3. Combined Axia1'Loads and Biaxial Bending Members subject to combined axial load and biaxial bending moment should be proportioned so as to satisfy the following set of interaction formulas: P/Pu + Cmx~ 1(1 - P/Pex)Mmx + CmyMy 1(1 - P/Pey)~ S 1.0 P/Pp + or
~
IMpx
where
~,
~
~
11.IBMpx +
~
11.IBMpy:S
r.o for P/Pp
~
0.15
+ ~ IMpy S 1.0 for P/Pp < 0.15
- maximum applied moments about the x- and y-axes
P - applied axial load' Pex - 23/12AF'ex 5-76
5-44 5-45 5-46
TK
5-l300/NAV~AC
P-397/AFR,
88~22
23/l2Af'ey F'ex
l2W2E/[23(Kl b/rx)2]
F'ey
l2W2E/[23(Klb~ry)2~
lb'-actual unbraced length in the plane' of bending,,: r x' 'ry--~corresp~ri~ing radii of ' gyration :P
r
:
P - Af d s
. ,"-
Cmx' C ,-, \ .~Y
coefficients applied to bending term in' interaction formula and dependent upon column curVature 'caused by applied moments (AISC Specification, Section 1.6.1)
~x' ~y -
plastic bending capacities about x and y axes
moments that can be resisted by the member in the absence of axial load. For columns braced in the weak direction,
Kmx -
~x
and
~y
-
~Y'
When columns are unbraced in ,the weak direction:
Kmx -
[1.07
(l/r y) (f ds)1/2 /3160] ~x S ~x
5-47
~y
[1.07
(l/r x) (f ds)1/2 /3160] ~y S ~y
5-48
Subscripts x and y indicate the axis of bending 'about which a particular design property applies. Also, columns may be considered braced in the weak direction when the provisions of Section 5-26 are satisfied. In addition, beam columns should also satisfy the requirements of Section 5-23, 5-38. Effective Length Ratios for Beam-columns The basis for determining the effective lengths of beam columns for use in the calculation of Pu' Pex' Kmx, and ~y in plastic design is outlined below.' For plastically designed braced and unbraced planar frames which are supported against displacement normal to their planes, the effective length ratios in Tables 5·10 and 5-11 shall apply. Table 5-10 corresponds to bending about the strong axis of a member, while Table 5-11 corresponds to bending about the weak axis. Iri each case, 1 is the distance between points of lateral support corresponding to r x or r y, as applicable. The effective length factor, K, in the plane of bending shall be governed by the provisions of Section 5-40. For columns subjected to biaxial bending, the effective lengths given in Tables 5-10 and 5-11 apply for bending about the respective axes, except that Pu for unbraced frames shall be based on the larger of the ratios Kl/r x or 5-77
TH 5-1300/llAVFAC P-397/AFR 88-22·
KI/r y. In addition, the larger of the slenderness ratios, l/r not exceed Cc'
x or
~/ry'
shall
5-39. Effective Length Factor, K In plastic design, it is usually sufficiently accurate to use the Kfactors from Table CI.8.1 -of the AISC-Manual (reproduced here as Table 5-12) for the condition closest to that in question rather than to refer to the alignment chart (Figure C.l.8.2 of AISC Manual) . . It is permissible to -interpolate between different conditions in Table 5-12 using engineering judgment. In general, a design value of K equal to 1.5 is conservative for the columns of unbraced frames when the base of the column is assumed pinned, since conventional column base details will',usually provide partial rotational,restraint at the 'column base: For girders oLunbraced frames" a design K value of '0.75 is recommended . .1
r
')
- "
,
..
"
, ;
.
';
.
,
, \
"
.. .'
5-78
,,'
,
,
'.'
TH 5-l300/NAVFAC
Table
5-io, ,
,
.,
.
,
Braced Planar Frames*
,
One-and Two-Story Unbraced Planar Frames*
. '" , , , Use la~ger ratio , l/r or Kl/r x y
"
Use larger ratio, l/r y or l/rx:
Pex
Use lir x
limx
Us,," l/ry
Table 5-11
,
,"
-
"
,
"Use Kl/r x
-
Use l/r y
Effective "Length Ratios fo. Beam Co Lumns (Flanges of members in the plane of the f r'ame; 1. e , bending about the ,weak axis) -
Use larger ratio', l/r y or l/r x
Pey Use l/r y
,
Use larger, ratio,
,
,-
Use Kl/r y Use l/r x
~y Use l/r x
*
"
One- and Two-Story Unbraced . ' Planar Frarnes*
t
Braced P'l'anar Frarnes* Pu
88-22
"Effective Length Ratios' for Beam ~olumris (Webs of members in the plan~ of the frame"; Le" bending about the strong axis) ,
Pu
P~397/AFR
l/r y shall not exceed Cc'
5-79
l/r x or Kl/r y '"
"
.
Table 5-12
,
Effective Length Factors for
.
(a)
(b )
, ,,:,(.u
'~~
"
Buckled shape of column is shown
by dashed line.
.
I I
/
\
\
:
(e )
and Beam-columns
(d)
(e )
(f )
• ,, •• J", •+ •• 'J
I
~I
'
..
,
r
',~:a
\ ,\ \
,
9 I
.
,f
I \
Colum~
,
~::r
,f\
I
\
I
I I
,I
I
I /
,
'"1" "f
I
t
,
1-" l'
·91'
I. I I I
,,
I
l-
I
I.
I " iJl '.'
, It··
"'?'r
.
Theoretical K value.
0.5 '
0.7
1.0
1.0
2.0
2.0
Recommended design value when ideal conditions are approximated . . .
0.65. ,
0'.80
1.2
I. 0,
2.10
2.0
I"
Rotation fixed and translation fixed.
End condition code.
~,
~ 'VfI' 9.
..
"
Rotation free
.~
.
.
,
.
and ' trunslctton fixed.
Rotation fixed and translation free. Rotation free
.
5-80
and translation free. .' . .
'.'
.
TH 5-1300/NAVFAC P-397/AFR 88-22
FRAME DESIGN 5-40. General The dynamic plastic design of frames for blast resistant structures is' '. oriented toward industrial building applications common to ammunition manufacturing and storage facilities, 1. e. , 'relatively low, single-story',' multi-bay structures. This treatment applies principally to acceptor structures subjected'to relatively low 'blast overpressures.
.
'
The design of blast resistant frames is characterized by (a) simultaneous application 'of,vertical and horizontal pressure-time loadings with peak pressures considerably in excess of conventional loads, (b) design criteria permitting inelastic local and overall dynamic structural deformations (deflections and rotations), and (c) design requirements dictated by the operational needs of the facility and the need for reusability with minor repair work after an incident must be considered. ' Rigid frame construction is recommended in the 'design of blast resistant structures since this system provides open interior space 'comb i.ned with
substantial resistance to lateral forces.
In addition, this type of construc-
tion possesses inherent energy absorption capability due to
t~e
successive
development of plastic hinges up to the ultimate capacity of 'the structure. However, where the interior since and wall opening requirements permit. it may
be as effective to provide bracing. The particular objective in this section is to provide rational procedures for efficiently performing the preliminary design of blast resistant frames. Rigid frames 'as well as frames with s'!pplementary bracing "and with rigid-or nonrigid connections are consider~d.. In both cas~s, preliminary dynamic load factors are provided for establishing equivalent static loads for both the local and overall 'frame mechanism>' .Bas ed upon the mechanism merhod , as employed in static plastic design, e's t Imaees are made' for the 'required 'plastic bending capacities as well as approximate values for the axial loads and .,' shears in'the frame members~'
The dynamic: deflections 'and rotations iIl" the"
sidesway and local beam mechanism'modes are estimatedbased~upon single degree-of· freedom analyses. The design criteria and the procedures established for the design of individual members previously d i scus'sed apply for this preliminary design procedure.
5-81
TH 5-l300/NAVFAC
P-397/AF~
88-22
5-41. Trial Design of Single-Story Rigid Frames 5-41.1. Collapse Mechanisms General expressions for the possible collapse mechanism of si~gle-story rigid frames are. presented in Table 5-13 for pinned and fixed base frames subjected to combined vertical and horizontal blast loads. The objective of this trial design is to proportion the fram~ members such that the governing mechanism represents an economical solution. For a particular frame, the ratio of horizontal to vertical peak loading denoted .by . a is influenced by the horizontal frame plan of the structure and is det~r mined as follows: ,
where
peak vertical load
qv
on
blast overpressure
o~
,
.'
'
rigid frame
peak horizontal ,load on Pv
-
rig~d
frame
roof
..
reflected blast pressure on front wall tributary width for vertical loading bh
tributary width for horizontal loading
The orientation of the ~oof purl ins with respect to the blast load directions are shown in Figure 5-23. The value of a will usually lie in the range from about 1. 8 to 2.5 when the direction, of the ~last, load is perpendicular to ,the roof purlins. In this case, the roof purlins are supported by the frame and the t.ributary width is the same. for the ho r Lzorrta], .and vertical load. The value of a is much higher when the direction .o f . the blast load is parallel to the roof purlins. In this case, the 'roof,p";rlins 'are not.supported by the girder of the frame.and the tributary width,of the vertical loading (b v - " purl in spacing) is much smaller than the tr~butary width of the hori~ontal loading (b h - frame spacing). " It is assumed in this procedure that the plastic bending capacity of the roof girder, ~, is constant for all bays. The capacity of ,the exterior and interior columns are taken as. C~ and Cl~' .respectively. s ince the exterior column is gellerally subjected to reflectea pressures, it is recolJl11lended that a value of C greater than 1.0 be selected. In analyzing a given,frame with certain member proper~ies, the controlling mechanism is the one with the lowest resist.'ince . .In design, however,. the load is fixed and the required design plastic moment is the largest ~ value obtained from all possible mechanisms. For that purpose, C and Cl should be selected so·as to minimize the value of the maximum required ~ from among all possible mechanisms. After a few trials, it will become obvious which choice of C and Cl tends to minimize the largest value of ~.
5-82 ,
J
1M 5-l300/NAVFAC,P-397/AFR 88-22 ~
; -'1
'5-41.2. Dynamic Deflections and,Rotations
It will normally be more economical to proportion the, memlieis so :that' the controlling failure mechanism is:a combined mechanism· rather 'thari'a beam, mecbarri sm ,
The mechanism having' the least resistance constitutes' an accep-
table mode of failure provLded -ehac the .magni tudas of .'the, maximum deflections and rotations do not exceed the maximum values recomnlended in Table'0-8. " . ~ :."~'" 5-41.3. Dyr(amic Load Factors
. ,
To obtain initial estimates of the required 'mechanism resistance, the 'dynamic load factors listed 'in this section may be 'used to obtain equivalent' static' loads for the indicated mechanisms, These load factors are necessarily approximate and make no distinction for different end conditions. However, they are expected to result in reasonable estimates of·the required'resistance for a trial design. Once the trial member sizes are established, then the natural period and deflecti~n of th~ frame can lie calculated. ,'
It is, recommended that the DLF for a beam collapse mechanism be equal to 1.25 while that for a panel or combined collapse mechanism be equal to 0.625. ' The DLF for a frame is lower than that for a beam mechanism, since the natural period of vibration in the sidesway mode will normally be much greater than the natural periods of vibration of the individual elements. 5-41.4. Loads in Frame Hembers Estimates of the peak axial forces· in the girders and the peak she'ars 'in the columns are obtained from Figure 5-24. In applying the values of Figure 5-24, the equivalent horizontal static load shall be computed using the dynamic load factor for a panel or combined sidesway mechanism. Preliminary values of the peak axial loads in the columns and the peak,shears in the girders may be' computed by multiplying the equivalent vertical static load by the roof tributary area. Since the axial' loads in the columns are due to the reaction from the roof girders, the equivalent static vertical load should be computed using the dynamic load factor for the beam mechanism.
,
..,w.'
5-41.5, Sizing of Frame Hembers ..', ,
.~ ·f
.
Each member in a frame under the<~ction of horizontal and vertical blast loa~s is subjected to combined bending.moments and axial loads. However. the phasing between critical values of the axial force and bending moment cannot be established using a simplified analysis. Therefore, it is'recommended that the peak axial loads and moments obtained from Figure 5-24 be assumed to act concurrently 'for' the purpose of i:'iial'beam'-column design: The overall resistance of the frame depends 'upon' the ultimate strength of 'the members acting as beam-columns, When an exterior frame of a building is positioned such that the shock front is parallel to frame, the loadings on each end of the building are equal and sideways action will only occur in the direction of the shock wave propaga-t.Lon, Frame action will also be in 'one direction, in the 'direction of the" sidesway. If the -bLas t; wave impinges 0;'; a building from a' quartering direction, then the columns and girders in the' exterior frames are subj ecced to biaxial bending due to the simultaneous loads'acting on the'various faces' of 5-83
·TK 5-l300/NAVFAC P-397/AFR 88-22 the structure. This action will also cause sidesway in both directions of the structure. The interior girders will usually be subjected to bending in one direction only. However, interior columns may be subjected to either uniaxia~ or biaxial bending, depending upon the column connections to the girder system -. · Ln. such cases ,. the moments and forces can be calculated by analyzing the respo~se of. the frame in eac~ direction and superimposing the respective moments. and forces acting on the individual elements. This approach.is. generally conservative since it assumes that the peak values of the forces in one direction occur simultaneously' throughout the three-dimensional·structure. Having estimated the maximum values of the forces and, moments throughout the frame, the preliminary sizing of the members can be performed using the criteria previously, presented for beams and columns. 5-41.6. Stiffness and Deflection The stiffness factor K for single-story rectangular frames subjected to uniform horizontal loading is defined in Table 5-14. Considering an equivalent single degree-of-freedom system, the sidesway natural period of this frame is .'
5-50 where KL is a load factor that modifies K, the frame stiffness, due to a uniform load, so that the product KKL is the equivalent stiffness due to a unit load applied at the equivalent lumped mass me' The load factor is given by
'KL - 0.55 (1 - 0.258)
5-51
!
where 8 is the base fixity factor and and fixed base frames, respectively.
~s
equal to zero and one for 'pinned base
The equivalent mass me to be used in calculating the period of .a sidesway mode consists of the total roof mass plus one_third of the column and wall masses. Since all of these masses are considered to be concentrated at the roof level, the mass factor, KM• is equal to one. The limiting resistance
Ru -
Ru
is givenpy
5-52
awH
where w.is equal to the equivalent static uniform load based on the dynamic load factor for a panel or combined sid~sway mechanism. The equivalent elastic deflection XE corresponding to
Ru
is
5-53 Knowing .the sidesway resistance Ru and the sidesway natural period of vibra-· tion Tn. the .ductility ratio (~) for the sidesway deflection of the frame can be computed using the dynamic response charts (Chapter 3). The maximum deflection ~ is then calculated from ·5_84
TH' 5-1300/NAVFAC P-397/AFR 88-22'
5-54
~ - I-lXE where
~
\.'
f
J-'~
'J
•
,
.',
,
1~
_
r-"
ductility r a t Lo-vLn s Lde sway '"f
JL -
5-.42'; Trial Design of Single-Story Frames with, Supplementary Bracing 5 - 42.1. General Frames with supplementary bracing can consist 'of (a) rigid frames in one, direction' and bracing .Ln- the other, direction .. ,(b) braced frames in two directions with rigid .connections, and vf c ) braced .f r ame s , in two directions with pinned -cormec t Lons. .:Most braced . frames' utilize pinned connections '.' ':-,r\'."'~' ";,;,",,r-!'~'l.·· .. ,:j., '~'I,.. 5-42,.2. :Coll~pse Mechanisms I " ~
t,
.'
". J .
:
,5
The possible collapse mechanisms of single-story frames with diagonal tension bracing (X-bracing) are presented 'in Tables 5-15 and 5-16 for pinned-base frames with rigid and nonrigid girder-to-column connections, respectively. ,.,In these tables, the cross sectional ,area of the tension brace is denoted by Ab, the dynamic design stress for the b r'ac Lng member is f d s;' and the number of _ braced' bays is' denoted 'by .tbe parameter, m..: In each,' c as'e , the ultimate capacity of, the ..fraine !':ics~ expressed. in terms 'of the equivalent 'static load and the member ultimate strength '(<:,ither Mp",or Abf ds)' In developing,'these ' expressionSlin~the. tables," the_same~assumptions were made as for rigid frames, Le., M., for the, roof girder is constant for all bays" the .bay width L is constant, and .t.heico Lumn moment c apac Lcyj.coe f f Lc Lerrt; C is greater, than 1 .. 0. ,'; For rigid frames with tension bracing, it, is' necessary to vary C, Cl" and Ab in order to achieve an economical design. When nonrigid girder-to column connections are used, C and Cl drop out of the resistance functIon for the sidesway mechanism and the area of the bracing can be calculated directly. 5-42.3. Bracing Ductility Ratio
,"
Tension brace members are not e~ttected to.remain·elastic under the blast loading. Therefore, it is necessary to determine if this yielding will be excessive when, the system is permitted t~ deflect to the limits of the design criteria previously given. 1-
T.
....'.,.,
1.
\.';i
The ductility r a t Lo i as soc La t ed-wf ch t'ensionyielding of the bracing, is defined as the maximum strain in the,brace divided by its yield strain. Assuming sma Lk rde f Le c t Lons I'and,neglecting axial deformations in the girders and columns, the -ductility ratio is given by " 5-55 where . /J".'
." 0 ...-: .
.~~q
,f: ~
,
• j_t
'
ductility ratio '
r••
',.
.
y
l
c
.
,
'-,1-1
';-.
"'i. •
,
r
:
..
.. .. ~
",
,,
"
'sidesway de f l.ec t Lon ;.. .Lnche's _ 1"·:
.
t·
"J
.1'
vertical angle between the bracing and a horizontal plane
.
-t • . -," 5~85
TH.5-1300/NAVFAC P-397/AFR
L -
88~22"
bay width, inches
From the deflection criteria, the sidesway deflection is limited to H/25. ductility ratio can be expressed further as
The
5-42.4. Dynamic Load Factor The dynamic load factors' Li s ced in Section 5-41.3 may also be used as a rational starting point for a preliminary design of a braced frame.' In general, the sidesway stiffness of braced frames· is greater than.unbraced frames and the corresponding panel or sidesway dynamic 'load factor may also be. greater. However, since 'these dynamic loaa factors. are necessarily approxi- . mate and serve only as a starting point.for a preliminary 'design, refinements to these factors for frames with supplementary diagonal braces are not warranted.
,
5-42.5. Loads in Frame Members
.
Estimates .of-the peak axial·loads-in the girders' and the peak shears in the . columns of a braced rigid frame are obtained from Figure 5-25. It should be' noted that the shear in the blastward column and the axial load,in the eX:terior girder are the same as, the' rigid frame shown in Figure '5-,24: The shears in the interior columns V2 are.not affected by the braces while the axial loads in the interior girders P are reduced by the horizontal components of the force ~n the brace FH. If a bay is.not braced, then the value ofFH must be.set equal to zero when calculating the' axial load in the girder of the next braced bay. To avoid,an'error,' horizontal equilibriums~ould be'checked using the formula: _'
Ru where
Ru, n
5-57
VI +nV2 + mFH
~
m'-
Hare
VI, V2 and F
defined in Figure 5-25 .
number of bays number of braced bays
In addition. the value of ~ used in Figure 5-25 is simply the design plastic moment obtained from the, controlling 'panel or combined mechanism. ,'; An e s tLma t e of, the peak loads for braced frames with nonrigid girder to column connections may be obtained using Figure 5-25, .However. the.value of Mn must be set equal to zero. For such cases, the entire horizontal load is taken by the exterior column and bracing. There is no shear'force in the'interior columns.
Preliminary values of the peak axial loads in the columns and the peak' shears in the girders are obtained 'in the same manner as, rigid frames. However,in computing the axial loads ,in the columns, the vertical components of the forces in the tension braces must be added . to ·..the vertical" shear in the roof girders. The vertical component of the force in: the brace is given by 5-58 5-.86
TK
5-1300/N~VFAC
P-397/~FR
88-22
The reactions 'from the braces will also affect .the· 'load on the foundation of the frame; therefore, the design of the footings -must; include these Loads, .
5-42.6. Stiffness and Deflection
,
The,equations for determining the sidesway natural period of vibration and the deflection at yield for braced frames are similar to that of rigid frames. The primary difference is the inclusion of the horizontal stiffness (Kb) provided by tension bracing. The equations for the natural period and elastic deflection are as f'o l.Lows t. natural period of vibration TN - 2" [me IKK L +
K~)11/2
5-59
and the equivalent elastic deflection is 't-7J:'
5-60 The horizontal stiffness of the tension bracing is given by
5-61 and other values have been defined previously.
It may be noted that for
braced frames with nonrigid girder-to-column connections, the
va~ue
of the
frame stiffness (K)'is equal to zero.
5-42.7. Slenderness Requirements for Diagonal Braces The slenderness ratio of the bracing should be less than 300 to prevent vibration and "slapping". This design condition can be expressed as
5-62 where
minimum radius of gyration of the bracing member length between points of support
Even though a compression brace is not considered effective in providing resistance, the tension and. compression braces should be connected where they cross.. In this manner, ~ for each brace' :>lay be taken equal to half of its total length.
5-42.8. Sizing, of. Frame Members Estimating the maximum forces and moments in frames with supplementary bracing is similar to the procedures describ~d for rigid frames. However, the procedure is slightly more involved'since it is necessary to assume a value for the brace' area in addition to the assumptions for the coefficients C and Cl. For frames with nonrigid connections, C and Cl do not appear in the resistance formula for a sidesway mechanism and Ab'can be determi~ed directly. In selecting a trial value of Ab for frames with rigid connections,the minimum brace size is controlled by slenderness requirements. In addition, in each particular application, there will be a limiting value of Ab beyond which there will be no substantial weight savings in the frame'members since there
TM 5-l300/NAVFAC P-397/AFR 88-22 are-maximum slenderness requirements for the frame members. In general, values"of Ab of about 2 square inches will result in a substantial~increase in the overall resistance for frames with rigid connections. Hence, an assumed brace area in this range is recommended as a starting point. Thei'design ·of the beams and columns of the frames follow the procedures previously presented.
,
.,'
,~
I,
.
,
.'
"
.
.
'. "
,
"I
.. 5-88
,
.
,
.
,.
,
'.
,
.
ROOF PURLIN
FRAME
. •
/~
DIRECTION OF BLAST LOAD • PARALLEL TO ROOF PURL INS ( by> b h ) ,,'
K
WALL GIRT
DIRECTION OF. BLAST LOAD PERPENDICULAR TO ROOF PURLINS ( by = b h )
t
. . Figure'S-23
Orientation of roof purlins with'respect to blast load direction 'for frame bl as t loadl11g
-.
.. ,}
Ru/4
Ru
Ru/8
3/8Ru
~
n=2
.
. C' r
Ru / 4
3/4Ru
Rul2-Ru/2n
Ru / 2
n
. Ru / 2 n
Ru / 2n
Ru 12n
,
.
. Ru/2'
•,
Ru/8
Ru/4
5/8R u
.
,r
Ru/ 2n
Ru/5
3/10Ru·
Ru / 2n
3/20Ru
n =I
4/5R u
n=2
Ru/5
Ru/3
2/3Ru
7/10R u
Ru13 - Rul3n
Ru/ 3n
Rul3n
.
3/20Ru
3I2ORu
Ru/ 3n
Ru/3n
Ru/ 3n
. n = NUMBER OF BAYS Ru = awH = EQUIVALENT HORIZONTAL STATIC LOAD
Figure 5-24
Estimates of peak sneans .and axial loads in rigid frame", due to horizontal 'loads 5-90
. PI
Ru
~
'.
P2
..
,P3.,
....t:>.c,
H
~~
VI
V2 ,
, :,
..
'Ru ". VI, • PI " P2 "
"
• fH'
awH Ru /2.i' Mp / H
Ru /~. - Mp / H ~ PI - V2 - FH FH • Abfds cos r FV • Ab fds sin r V2 .• 'Ru/2n::- Mp/nH P3 " P2-V2- FH Pn ,.• P(n-I)'-V2 - FH'
, \
-.-
"
Figure 5-25
,"
Estimates of peak shears and axial loads in braced frames due to horizontal loads 5-91
Table 5-13
Collapse Mechanisms for Rigid Frames with Fixed and Pinned Bases PLASTIC
COLLAPSE MECHANISM
MOMENT M"
PINNED BASES
FIXED BASES
16
16
BEAM MECHANISM
BEAM MECHANISM
I
! ! 7 70- 7° )
awH·
awH 2
4(2C+1)
4(3C+I)
awH·
I
PANEL MECHANISM
3;;0
I
-2- 0 "'27+7: ( n"'-"il""C:-,-
I+(n-I)C,+C 'C,22'·
(C,
I .
--0
2
PANEL MECHAN ISM
4~
w -0
2
2(n+C)+{n-I)C, (C'~21* aH
2
n •
+2 l
2(2n+C)+(n-I)C,
COMBINED MECHANISM
50)
I I ?
3 awH. 8
COMBINED MECHANISM
.rrr: COMBINED MECHANISM
t
2
3 C+(2n-T)
COMBINED MECHANISM I , , I
awl
,
I
W
I 1-1-
Iii Ii' , Iii i
MP !CMp
IC'MP
I
I..
L
i- [3aH2 + (n- il LZ]
Z
[3aH +(n - I)L ]
L
, , , ,
'7.1:I: I:I ~ 1:
H
1_
L·
5 .( n-I ) -+ C, (2n-3) -C+ 222
.1
i :- -;;. .1,2,3······ NUMBER OF B AYS
, r
.'UNIFORM EQUIVELENT STATIC LOAD
.
111 FOR CI • 2
HINGES FORM IN THE GIRDERS AND COLUMNS AT iNTERIOR JOINTS.
5-92
.
~
".1 '
,
'
-
,,;;1
Table 5-14
-.4
,.,,-
--
.
,
H3
.
.n = NUMBER OF BAYS
D = lea
...
'L
,,: •
(0,75~6.~5{3) IH •
...
~
~
1
'.
P " 'ASE,FIXITY FA';""t
",
lco.c~ .rIT(0.7-0.1 o)(n-l)l'
STIFFNES'S FACTOR' K =' E ,
-
Stit't'ness Factors t'or Single' s.tory •.Multi-bay Rigid Frames Subjected' to Unit'orm Horizontal Loading'
I~II. ",1 • 'I. 1---
, .
I.' L '~:'
. lea =
AVERAGE COLl/MN MOMENT' OF INERTIA= LIe l(n+l)
..
C2 D
{3= I. 0, 26.7 0,25 , "0,50 3,2.0, "
1.00
37.3
,
.. .
{3~0.5~ {3=0 :14,9 3.06 4.65 17. 8 , ~,29·6
6.0~
.. VAI"UE,S OFC2 ARE APPROXIMATE FOR THIS {3 t , {3 =.1.0 FOR FIXED BASE, =0.0 FOR HINGED' BASE
..
',.
,,;
"
~'
WHERE:' . 'E='MODU!2U'S OF ELASTICITY(PSi5 ' lea. lillIe = MOMENT OF INERTIA(IN:4) H = HEIGHT (FEET) L= BAY LENGTH(FEE1)
5-93
.
\
Table 5-15
Collapse Mechanisms for Rigid Frames with Supplementary Bracing and Pinned Bases PLASTIC
COLLAPSE MECHANISM
MOMENT
1[)1)1>1
wL
Mp
2
16 BEAM MECHANISM
.
2VC7V<1' . ,
,
.awH
'
".
2 "
4(2C+I)
BEAM MECHANISM
"
PANEL MECHANISM
~
4awH2_ mAbfds Hcos '4n
Y
2n,
PANEL MECHANISM
4~ COMBINED MECHANISM
5avt/1/1 COMBINED MECHANISM
3
5~
2m
aawH -2'AbfdsHcosY , C+
COMBINED MECHANISM
t
6~,
(~-
t)
(Co>2)*
[3aH+(n-IlLj-l}A ,
COMBINED MECHANISM
w
bfd s
Hcos
, C+(2n-"il
,
,
'i 11';\' I "8' " " , , 1'1' , '.1:t:l:1::I::l:1: 'm =NUMBER
I
aw
p
~ ~.1· <3
(J
I
it FOR
~f§]'-T-
~ ~~o'" ~ ~~o'" L
L
~~o'"
HI'
Y
I
L
Y
I
I..
,I
OF
BRACEDBAVS n = NUMBER OF
.
BAYS = 1,2,3" • w' UNIFORM EQUIVAL~ ENT STATIC LOAD
Co = 2 HINGES FORM IN THE GIRDERS AND COLUMNS AT INTERIOR JOINTS,
5-94
.», '
Table 5-16
Collapse Mechanisms for.. Frames wi th SuPPlementary Bracing, Non'rigid 'Girde~to-Column Connections and Pinned Bases
COLLAPSE
MECHANISM
I:.~ BEAM MECHANISM EXTERIOR GIRDER
2.
-
.. FRAMING TYPE
.
ULTIMATE CAPACITY -.
Mp
. M~ Mp
..._' ~
~
= wI..: /s = wL2/1~ = wL." /16
CD M p = wL." 116 ®a0 3.:Vl/V1. Mp' awH awH 2 -il CDa@· 2 Mp = BEAM· MECHANISM 0. 4(2C+ 1) , BLASTWARD COLUMN .. . Abfds = a w H 12m cosY (Da® '~V[7J;7r' 2Mp aw H Abfds = ® 2mcosy mHccisy PANEL MECHANISM ,
M p = wL."
.... BEAM MECHANISM INTERIOR GIRDER
Is
~
'
5,VL7l77 .
.
.
,
.
.
AtJds
,
COMBINED MECHANISM
3 a wH _ (2C+I)Mp mHcosy
GIRDER FRAMING TYPE: .
I
~ 3
GIRDER CONTINUOUS OVER COLUMNS AND RIGIDLY CONNECTED TO;E)(TERIOR COLUMNS ONLY ~
.I
· •
.. Ii
i
,.
GIRDER SIMPLY SUPPORTED BETWEEN COLUMNS GIRDER CONTINUOUS OVER COLUMNS .
2
~
0.'
= 4mcosy
:
'
•.
'. '.W ..
,.
.
]-1-
Igl I" I. I I I I " " I 1.1 II)" 1,1 I" hZ7r%%%%
~MpP
.. ' .
.'
H
~fdl'
Abfdl, L
.~.
L
L
.1'"
m
= NUMBER
I I
.. =
OF BRACED BAYS
UNIFORM EQUIVALENT . STATIC LOAD..
TK 5-1300/NAVFAC P-397/AFR 88-22
'e
CONNECTIONS .. 5-43. General The connect~ons in a steel structure designed in accordance with plastic design concepts must fulfill their function up to the ultimate load capacity of the structure. In order to allow the members to reach their full plastic moments, the connections must be capable of transferring moments, shears and axial loads with sufficient strength, proper stiffness and adequate rotation capacity. Connections must be designed with consideration of economical fabrication and ease of erection.
~elds,
Connecting devices, may be rivets, bolts,
screws or
various combinations thereof. 5-44. Types of Connections The various connection types generally encountered in steel structures can be classified as primary member connections, secqndary member connections and
panel attachments. Primary member connections are corner frame, beam-tocolumn, beam-to-girder and £ol.\!lIlnbase connections as'well as splices. Secondary member connections are purlin-to-frame, girt-to-frame and bracing connections. Panel attachments are roof-to-floor panel and wall siding connections.
.
,
Primary member connections refer to those used in design and construction of the framing of primary members. They generally involve the attachment of hotrolled sections to one another, either to create specific support conditions or to achieve continuity of a member or the structure. In that respect, connections used in framing may be classified' into' three groups, namely, rigid, flexible (nonrigid) and semirigid, depending upon their degre~ of restraint which is the ratio of the actual end moment that may be developed to the end moment in a fully fixed-ended beam. Approximately, the degree of restraint is generally considered as over 90 percent for rigid connections, between 20 to 90 percent for semirigid connections and below 20 percent for flexible connections. ' It should be mentioned that the strength and'rotation characteristics of semirigid connections are
.
Secondary member' connections are used to fasten members such as purlins, girts or bracing members to the primary members of a frame, either directly or by means of auxiliary sections such a.s angles .~nd tees .... Basic requirements for primary and secondary member connections, as well as general guidelines for proper design, are presented in Sections 5-45 and 5-46. In addition, dynamic design stresses to be used in the selection and sizing of fastening devices are given in Section 5-47. Panel attachments are used to 'attach elements of the skin or outer shell of an installation as well as floor and wall panels to the supporting skeleton. 5-96 '
TK 5-l300/NAVFAC
P-~9~/AFR 8~-22
Connections' of this type are distinguished by the fact that they. fasten r.elatively thin sheet materIal to one ano cher br to hea'(i~rroll;'d sections:' Roof decks and wall siding,have to wi 7hstand d~ring their lifeti~7'~apart,from :, accidental blast loads) exposure.to weather; uplift forces, buffeting and", vibration due to winds: etc, For this reason, and'be~ause of their widespread use, special ca~e should be ,taken ~n design to ,ensure their adequate behavior" Some bas i,c requi'rements.'for pane l connections are :presented. in Sec t Lon 5-4~, ,
5-45 ..
".,
.
•
' _
, ,0
n ." ,. . . ,. Requirements.f,or'Hain:Fr~mingConnections
The design requirements for frame connections may be'illu;tratep byconsid~r ing the behavior of a typical corner connection as shown in Figure 5-26. Two members are joined together without stiffening,of.the corner web. Assuming that the web thi~kness is insufficient, the behavior ot the connection is . represented by Curve I' which shows that yielding due to' shear f~rce' starts' 'at a relatively low load. Even though the connection rotates, past the required hinge r ot.ar Lon.. the plastic moment Hp. is not reached. In ' addition, the elastic deformations are also larger than those assumed by the theoretical design curve. . , . . '", ,. A second and different. connection m~y behave as indicated'by Curve, 2. , Althoughth~elastic stiffness is satisfactory and the maximum capacity exceeds .~, the connection faIled before reaching'the required hinge rotation and thus, is unsatisfactory. ,
These considerations indicate that 'connections must be designed for strength, stiffness and'rotation capacity. They must transmit the raquired moment, shear and axial load, a~d develop the plastic moment Hp of 'the members. Normally, an examination of a connection t~ see 'if it meets the requirements of stiffness will not be necessary. Co~pared to ~he tot~llength of the member, the length'of the connection is small, and, if the connection is slightly more flexible than the member which it joins, the general effect on the structural. behavior is not great., Approximately, the ayerage unit rotation of the connecting zone should not exceed, that of an equivalent length of the. members being joined. •
,.
•
~i'
•
Of equal dmpor t ance with the strength of the 'connection "is an adequate re~~rve of ductility after the plastic moment ha~ been attained; Rotation,capacity at plastic hinge locations is essential to the development of the full ultimate load ,capacity of th~.structure. 5-46. De~ign of Co~nections
r»
It is not the intent of this section to present, procedures, and equations for the design of the various types of connections likely to be encountered in the blast-resistant design of a steel structure. Instead, the considerations necessary for a pr~per design' will be outlin~d.". . . After completion of the dynamic analysis of the structure", the members are sized for rhe given load~.ngs. ,The ,moments~ shears, and axial loads at the connections are known. Full recognition must be given to the consideration,of rebound or stress r'eve r s a L. ill; designing the' connections'. Additionally, in continuous structures, the maximum values of P, M, and
y
may not. occur
simultaneously and thus, several combinations may have to be considered. 5-97
.
With rigid connections' sucha~ .a'
~
Stiffeners will normally 'be required to'prevent web cripPfin~ and prese7Ve,' flang~ continuity wherever flange=to-flange connections occur at columns in a continuous frame. Web crippling must also be checked at points of load application such as beam-girder intersections.
,.
In these cases, the require-
ments of Section 5-25,of this chapter and Sections 1.10.5 and 1.10.10 of the AISc' Spec'ificatic;n must be considered. ,>
."
r
'-
. '
, ,.
Since 'bolted 'joints will develop yield stresses only after slippage' of the member s'
.,
~
5-47. Dyna~ic'Design Stresses for Connections In accordance with Section 2.8 of the welds shall be proportioned to resist to 1.7 times those given in Part 1 of stresses are increased by the dynamic 12.2; hence,
AISC Specification, bolts, rivets and the maximum forces using stresses equal the Specification. Additionally, these increase factor spe~ified i~ Section 55-63
where
fei -: the maximum dynamic de s'Lgn stress for connections f
!.
the dynamic increase factor Figure 5-2 or Table 5-2)
c fs' - , '
the aliowable equivalent' static design'stress of the bolt; rivet, or weld '
Rather than compiling new tables for maximum dynamic loads' for 'the various types of connections, the de~igrier will find it advantageous to divide toe', forces being considered by the factor 1.7c and then to refer to the'allowabl~ load tables in Part 1 of the AISC Specification. 5-48. Requirements for Fioor and W~ll Panel Connections Panel connections, in general, can be considered either'panel-to-panei connec_ tions, or panel-to-supporting-frame connections. The former type involves the attachment of relatively light-gage materials to each other such that they act together as an integral unit. The latter type is generally used to att~ch sheet panels to heavier 'cross sections. j~ -,
. . ..; ., .
. ;,.
The most common'type of fastener for decking and, steel wali' panels is the' self-tapping screw with or without washer. 'Even for conventional design' and regular wind loading, the screw fasteners have often been the source of local failure by 'tearing the ~heeting materiaL. It is evident that under blast loading and particularly on rebound, screw connectors ,will be even 'more vulnerable to this type of failure. ,Special care should be taken in design to reduce the , probability of failure by using oversized washers-and/or increased material thickness'at the connection itself. 5-98
~
TK 5-1300/NAVFAC P-397/AFR 88-22 Due to the magnitude of forces involved, special types of connectors, as shown in Figure 5-27, will usually be necessary. These may consist of self-piercing, self-tapping screws of larger diameters with oversized washers, puddle welds or washer plug welds, threaded connectors fired into the elements to be attached, or threaded studs, welded to the supporting members, which fasten the panel by means of a special arrangement of bushing and nut. Apart from fulfilling their function of cladding and load-resisting surfaces, by carrying loads perpendicular to their surface, floor, roof and wall, steel panels can, when adequately' connected, develop substantial resistance to inplane forces, acting as diaphragms contributing a great deal to the overall stiffness and stability of the structure. As a result, decking connections are, in many cases, subjected to a combination of shearing forces and pull-out forces. It is to be remembered also that after the panel has deflected under blast loading, the catenary action sustained by the flat sheet of the decking represents an important reserve capacity against total collapse. To allow for such catenary action to take place, connectors and especially end connectors should be made strong enough to withstand the membrane forces that develop.
,
\
. 5-99
.
·"
M /
® UNSTABLE -
/REQUIRED M-
--T·.,L...;.-. .
-
"
II-I
e
CURVE
.
CD UNDERSTRENGTH PERMISSIBLE HINGE ROTATION
e ROTATION
v~L)
ll===ll=====
M
-~M v
Figure 5-26
Corner connection behavior 5-100
, .'
,
'
OVERSIZE WASHER
....
.-",'
. LIGHT ,GAGE '";I.fANE.L.
,,
'. SUPPORT MEMBER '.'
RAMSET TYPE FASTENER
SELF - TAPPING SCREW' ~,
, '.... .. - .
.1 '
..
CIRCULAR·
.,
. !'
WASHER .
j
1
.
SECTION A·A
PUDDLE WELDS
LIGHT GAGE PANEL
:
·~·T·
SECTION B-B
WASHER PLUG WELD
I
~
ATTACHMENT
STUD '. SPACER BLOCK
OVERSIZE WASHE
DETAIL.
SUPPORT MEMBER
"'-'
Figure 5-27
THREADED NELSON TYPE STUDS
....
Typical connections for cold-formed $teel panels 5-101 .
TIl:
5-l300/NAVFAC P-397/AFR 88-22 FRAGMENT PENETRATION
5-49. Penetration of Fragments into Steel
. ..,
5-49.1. Failure' Mechanisms
In deriving a prediction equation for the penetration and perforation of steel plates, it '1s important to recognize the failure mechanisms. The failure mode of primary concern in mild to medium hard homogeneous steel plates subjected to normal impact is ..d uctile failure. In this mode, as the missile penetrates the plate, .pLas td ca.LLy deformed material is pushed 'aside and petals or lips are formed on· both the front and back faces with no material being ejected from the plate. For plates with Brinell hardness values above 300, failure by "plugging" is a strong possibility: In this br Lt t l.e' mode of ~ailure" a plug of material is formed ahead of the penetrating missile and is ejected ·from the back side of the plate. A third mode of failure is dI sk Lng or flaking; in which circular disks or irregular flakes are thrown from ·the back face. This type of failure is mainly a concern with plates of inferior quality steel and. should not, therefore, be a commdn problem in the design of protective' structures.
5..49.2. Primary Fragment Penetration Equations In protective design involving primary fragments, a penetration equation. is required which yields reliable estimates corresponding to t~e st~ndard primary fragment illustrated in figure 4-77 of Chapter 4. These design equations consider only normal penetration which is critical for ·the design of protective structures. These equations apply to penetration into mild steel and are conservative for plates with a Brinell hardness value above 150 .... Steel penetration equations in design for primary fragment.impact are expressed in
the following forms: For AP steel fragments penetrating mild steel plates, x -
0.30 W 0:33 V 1.22 . f·
5-64
s
and for mild steel fragments penetrating mild steel plates, 5-65 where x
depth of penetration (in.) fragment weight (oz.)
Vs
striking velocity of fragment (kfps)
Charts for steel penetration by'primary'fragmentsaccording to these equations are presented in figures 5-28 and 5-29. • To estimate the penetration of metal fragments other than armor piercing, the procedures outlined in Section 4-60.3 of Chapter 4 are entirely. applicable to steel plates.
5-102
TM 5-1300/NAVFAC P-397/AFR 88-22
5-49.3. Residual Velocity After Perforation of
Ste~l
Plate
The'pe~etration equations presented in Sectio~:5:49.2 may be used for predicting the occurrence of perforation of metallic barriers and for calculating the residual fragment ,velocity .after perforation. . For normal impact of a steel fragment, with. the· shape illustrated in figure 477 of Chapter 4, the equation for .residual ve'locityis ,:,. ,. 5-66 residual velocity
where Vs
striking' velocity critical perforation velocity for the fragment.~f impacting the plate of thickness t (see explanation below)
d -
diameter of cylindrical portion of f r agment; (in.), as ·illus. trated. in figure' 4-77 of 'Chapter 4'
The value of Vx is determined from Figure.S-28 or 5-29 by substit~ting the, plate thickness t for the penetration depth x and reading the corresponding value of striking, velocity , Vs " This striking velocity becomes the critical perforation velocity, Vx ' A plot of the residual velocity equation for a. range of t/d ratios is presented in Figure 5-30. Multiple.plate penetration problems may be analyzed by.the successive application of Equations 5~64 or 5-65 for predicting the depth of penetration and Equation 5-66 for- calculating the residual velocity upon perforation of the plate. In addition, composite construction, consisting of concrete walls with attached spall plates, can be analyzed for fragment impact by tracing the motion of the fragment through each successive layer. The striking velocity of the fragment upon each intermeqiate layer is the residual fragment velocity after perforation of the previous layer. The conservative assumptions are made that the fragment remains intact during the penetration and that it does not'deviate from a straight line path as it crosses the interface between' the different media.
.'.
5-103
.'
~ 1
10.0
7.0._~
"
5.0.~~ ••
2.0 :r
l-
e.
w
o
z
o
1.0
I-
~
0::
I-
0.7
W Z W
e. >C ,
,
- .1:4-:t-
0.3 '- '
~Ct±-+--r--t--
NUMBERS NEXT TO CURVES INDICATE FRAGMENT: WEIGHT (OZ.)
0.2
0.1
o
4
6
Vs • STRIKING
Figure 5-28
8
10
12
14
VELOCITY (FT. x 103/SEC.l
Steel penetration design chart - AP steel fragments penetrating mild steel plate9 5-104-
.
,
10.0
;.
hf:ttttt:ttm-tti:t:rt1 t-
,.,'.': -WI =20.0
--
WI = 8.0 - Wf
=
+.
- -I'
WI=40.0 WI =30.0
...
6.0 r--'-"-r~r-'--'-i'-'-'--i-'-l
_..,
- WI = .4.0 '-'--~S-'c'-r.-'-'-'-f-':'-;-.:-j .
Wf = 60.0
~:=:J:=':"l"---
2.0
-Wf =
1.0
3.0
z
2.0 -r-rt-rt-t
:I:
~Wf=
0.5
-Wf =
0.3
·i.'.
--"i"::
:';':::F::-:: - +. -
-Wf=
:c:'c- Fe::.... .cc.c::· - - .. .
l-
... ,.-
..,-- _.':::: t.::.::::-
.. : t : - . -
c,
UJ
0
z
0 l-
1.0
.'
e(
0:::
I-
e
UJ
z
UJ
e, K
..
~-_.
.
-~
-.-,-------'.......
0.2
..
~.~
.... - .. -.--. . .
--
-
~
NUMBERS NEXT TO CURVES INDICATE +-- .-...' .. ,-_. FRAGMENT WEIGHT (OZ.)
~~.
-_. ---~ ..
~-_
--_.
:::;~~~:~~,::~::;:, ~~:~':~ ':::~: ~
--_._. - .-.- '.-.. _...•. -_..
.• t · - . '.
- .• -
.:;;I;'{ . . .. . ...
-+'-'"
. __ ._._.- --'.---
-~
_~
. ':.::,.-
"'-""",
::·x
o
2
4
6
8
10
Vs , STRIKING VELOCITY (FT.
14
103/SEC• 1
, Steel penetration design chart - ., mild steel fragments . penetratini mild steel plates J
Figure 5-29
X
12
5-105
•
'~
1.0
!
fJtW.1
!
I ;
~
0.8
1111I I" l" II! ! iii ! iI ! I' .I1'1I:~ . II !I,I IIIt I, : I, I :I I! I 1,1, I• I( iH-,
11
f/cJ "':1
I
I
II t
i
i
If1'l
illflTf
0.6
-
,
"'H
,.!/cJ"'.75 ...L.Ll. .
-!/d;/.5'
-
-
t/d"'2 t/d~3
-
-
.
-
-
-
-
-
-
0.2
-
,
tid; 5
0.2
.
I'
, t
t/d"'/.O
..
0..4
I
1
f/cJ"'.50 ,~j
l
,I,
-
,
-
..
,
-
-
l-U
-
-
0.4
.
,
-
0.6
0.8
I
1.0
. VX/VS
Figur~
5-30
ResIdual fralllDent velocity upon perforation of steel barriers . : '", , 5-106
TK 5-l300/NAVFAC ,P-397/AFR 88-22 "
TYPICAL DETAILS FOR BLAST-RESISTANT STEEL STRUCTURES , , ' 5-50. General
• I.
•
'I
This section presents several examples of typical framing connections, structural details and blast doors used in industrial installations designed to resist accidental blast loadings. This section is intended to augme~t ' those details present~d i~ prior sections of t~is :hapter. (
5-51. Steel Framed BUildings Such buildings are often 'rectangular' in plan, \wo or three bays wide and four' or more bays long. Figure 5-'31 shows, an example of a typ'ical' 'fr~i~g plan for a s Lng'Lee s to ry bud Ldf.ng designed 'to~ re'sist ,a pre,ssure-time 'blast loading impinging on,the,s'tr~cture at,an angle with respect to its .main axes. The structural system consists of an orthogonalpetwork of rigid frames. The girders of the frame~ running parallel to the building length serve also as purlins and are pLaced , for o'f erection, on top of the, . .ease . . 'frames spanning across the structure's width.
...
"
Figures 5-32 to ~,35 pre~e?t 'typical f~aming details rel~ted 'to,the g~neral layout of Figure 5-31. .. As a rule, the 'columns' are fabric';'ted without splices,
the plate.,cov~rs ~and . 'conn~ct'ionplates are shop w~lded to - the . columns, . .' '. -. . a~d all .
,~ .
' . )
"
girder to.co+um~ connections are field ~olte~. A channel,is, welded on top of the frame girders to cover the bolted connections and prevent (avoid) interference with the roof decking. All of the framing connec t i.ons are designed to minimize'stress concentrations and to avoid triaxial strains, 'They combine ductility with ease 'of 'fabrication. ' 5-52. Cold-formed, Light Gage Steel Panels
Figure 5-36 sh~ws typical cross sections of cold-formed, light gage steel panels commonly used in industrial installations. The closed sections, which are composed of a corrugated hat section and a flat sheet, are used to resist blast pressures in the low pressure range, whereas the open hat section is recommended only for very low pressure situations as siding or roofing material. A typical vertical section illustrates the attachment of the steel paneling to the supporting members. Of particular interest is the detail at the corner between the exterior wall and the roof; which is designed to prevent peeling of the decking that may be caused by negative pressures at the roof edge. Figure 5-37 gives some typical arrangements of welded connections for attaching cold-formed steel panels to their supporting elements. Type A refers to an intermediate support whereas Type B refers to an end support. , It is recommended that the diameter of puddle welds be 3/4 of an inch minimum and should not exceed 1-1/2 inches because of space limitations in the panel valleys. For deeper panels, it is often necessary to provide two rows of puddle welds at the intermediate supports in order to resist the uplift forces in rebound. It should be noted that welds close to the hooked edge of the panel are recommended to prevent lifting of adjacent panels. Figure 5-38 shows an arrangement of bolted connections for the attachment of cold-formed steel panels to the structural framing. The bolted connection consists of: a threaded stud resistance welded to the supporting member, a 5-107 .. '
TH 5-l300/NAVFAC P-397/AFR 88-22
square steel block with a concentrrLc hold used' as"a spacer, and a washer and nut for fastening. Figure 5-39 presents a cross section of that connection with all the relevant details along with informat~onpertaining to puddle' welds. 5-53. Blast Doors Figures 5-40 and 5-41 show details of single~leaf and double-leaf blast doors, respectively. Figure 5-40 presents a single-leaf door installed in a steel structure. The design is typical of doors intended to resist relatively low pressure levels. It is interesting to note that the door is furnished with its tubing frame to ensure proper fabrication and to pzovdde adequate stiffness during erection. In Figure 5-41, the double-leaf door with its frame is installed in place Rnd attached to the co~crete structure. in both figures details of hinges; latches, anchors, and panic hardware are illustrated. It should be noted that the pins at,the panic latch ends are made of aluminum in order to eliminate the danger of sparking, a hazard ,in alnmunition facilities which might arise from steel-on-steel striking;' Figure 5-42 shows details of compression arch and tension arch doors. The tension arch door requires compression ties to develop the compression reactions for the arch and to prevent the door from being blo~ through the opening. 'The compression arch door requi~es a ,tension 'tie plate'to develop the reactions'and to prevent large distortions in the 'door that may bind it in place. ' r
'
~
, ,
,, "
,
.
.
"
,"
'", ,'-.J
,
,~
5-108
,
-
,r;;;
~
....3_);ACES ",_._--
~'-O"--- ------
@17'---- 0" -- -
'.
~,
0 --
;'.,
,' ! {,.W8XS8;-WI12.79 , • :C"LUMN "-'-', - ' - H -""":'--:"--H
f?'i..A -_' . . ''':: o , i~ O. lD'
"
+
otl
~!?
I 10
~ 0
o"
'" ~I.; ~ -+.
o
-~:
~@. 01/)' -
N
')Ie:
~
~
U
IO~
"'otl
iH ,... .. _..:...._-:_- H
..
,W
u
W 10.25 "
!!l
~--'-----I:.( . -_..
on": ",on
on
"
" " ." 0..0
:t
:t
:t
0
:t
H
!!!
!!!
". 2" 0" ..:tu ,+
I rW12.58 I J. C,OL.UMN
,.
,,3fp, INt COL..
W 10. 25
:t~
2"
•
g}
II", , 10)
K
K'
0_ 'lD 3; !~
. H -"":""'_-':""'-H
I
STE EL' .PA NEL TYF 2P!20
Q.
I/)
ton
..-
W 10 x 25
E~
~E
,
0)l_tc:-:-,,-===J:~~~===~=,=====1:f,_·G .,
COLUMN
ROOF'
PLAN
".
Figure 5-31
Typical. framing plan for a single-story blas.t-resistant steel'structure ", "
5-109
B
,~ C7
STEEL PANEL ROOF DECK
x 9.B (CUT).
,
2- 3/4"; BOLTS (TYP.)
I.
I
WI2 COL.
,'-
'3/ 8 "t 's
"
--'L--\I~-J.L ~ 4 • 7/B"p BOLTS
-
:.J '" B
3/16
(TYP.)
'
SECTION A-A
,
3/8"
t
STEEL PANEL ROOF DECK
SHIM-
-C4 31116
3/8"t
I!
'-~-2-3/4"; BOLTS
(TYP. )
4-7/8"; BOLTS-+-! (TYP.)
'"u,
WI2 COL.
I
SECTION
Figure '5-32
B-B
Typical framing detail at 5-110',
interio~,Column 2-C
STEEL PANEL ROOF OECK
C7
3/8" Il
1/4 W10 ~II-..-lIt
3/16
3/8" It.
C-C
SECTION
STEEL PANEL ROOF DECK
C7l9',8(CUT)
.
'
. ;,
3/8" 'I. SHIM
, WIO
WIO
?- 3/4" III BOLTS
3/B" It.
(TYP.):
W 12 COL,
SECTION
0-0
FigUre 5-33' Typical framing detail at end Col,umn l-C 5-111
F
~
STEEL PANEL
ROOF DECK C7. 9.1 I CUT)
1/4
4-7/8" ~ 101,i-S--l-l 3/8"t .'
.', SHIM
.I
318"t WIO--+-~
2-3/4" _ BOLTS
.1 •
3/8" t
~F SECTION
E- E
'\ .
, ... ',
STEEL PANEL
ROOF DECK
C4
SHIM-=j~~~~~::!:*!T~T~I!t~~r
3/1"i,-
WIO--.j....
2-3/4"_
.'J--WIO
I
!
..
3/I"t J ~
2.7/1"~~ BOLTS
.
..
.
~.:
..
~
I
Ii . I'
SECTION
F.-F
FigUre 5-34: Typical framing detail at side ooll.llll1 2-D' . 5-112
' =-3/'' 160-''V',- STEEL PANEL ROOF DECK
• t
C7x9.B
(CUT)~
1/4
'
~ I
4-7/B"~ BOLTS ('typ')
--3/B"1l
.
SHIM
:1-
I~r==~~:+=·~
3/16
3/B"Il'S
3/B"Il-
'w 12 COL. SECTION
2-3/4"
~
BOLTS
lJ
G-G
STEEL PANEL ROOF DECK -----, C4
I I F==r'"'==a~;g, o!;;~ii"'~! 2·3I'l~·~
3/16
BOLTS 4-7ft!" ~ BOl TS;-+.,
( TYP.)
II
II
SECTION.
Figure 5-35
H-H
Typical framing detail at corner Column I:-D 5-113
STEEL PANEL CONNECTION TYPE B
CONNECTION TYPE A
~
CLOSURE L
TUBE CONNECT. TYPE B
J.
CHANNEL-
CHANNEL
STEEL PANEL
CONNECT. TYPE A
~ . .w&i
~~
. ,'ta" DEEP
PANEL (OPEN SECTION)
,
3~;!
.
;i',"·
~.
a
~
~~ I~" DEEP PANEL (CLOSED SECTION)
. 2~.·
, 57."
~~
~--+ ~ 3" DEEP PANEL (!=LOSED SECTION)
~~~rt-CONNEC~
..
3+-t-f~
TYPE A
~ ~l
ANGLE ~~
~~r-'CONNEC~
.... '. ','
TYPE B
24
11
4'/2" DEEP PANEL (CLOSED SECTION)
EXTERIOR VERTICAL SECTION
Figure 5-36
Typical. details for
TYPICAL STEEL PANELS
cold~formed.
5-114
light· gage steel paneling
-, 3/4' iJ PUDDLE, WELDS ._.
-
--".. ----
~_.:.
, -
- --
,--
-
--,
--
< ,,"
..
.. . ,
-
-
_- . - f- \-
,
'~
.,
--+r-f'.. ' .. I:-Q , • ., <0. •
f--
- ..
1-- -
"
~- ..
PLAN, - WELDED CONNECTION. TYPES A a B
l...:..l::::J.._.l
C":"J 'iI-1:=' '.r=:l :,' ,I.
•
24" PANEL WIDTH' :--"
--~ ~t-
SECTION ., 1V2" STEEL PANEL
.'., 3/4" iJ PUDDLE WELDS FOR 20 GA~ I" iJ PUDDLE WELDS FOR 18 GA.
or ----
---
I .
~Ul-
. :.. -_.~
+----H&+------
---f.l"+--+----f~+_-'-
-
-
~
'g;-
- -
--+--o-I---''---+-+-1f- i
-
-
;
.
-I' t.
•
' , '.
.
~
•
3xWELD
0 IA. ~MIN.
3/4" MIN.
..v>......J.-L.J.-+
PLAN -WELDED CONNECTION TYPE A
----
- ._--_._-- L
I[
~
Ii
•
v
.
I~
---- -
,
1-
,-
~
. .. - - -- y
. -. .'
v
~-
---- I- - -
-e - - ,
!
,>
~
--- -.. -A
,
PLAN - WELDED CONNECTION TYPE B
~-'W_U_U_~4L13 I I---l 24" PANEL WIDTH
SECTION - 3" STEEL PANEL
Figure 5-37
Typical welded connections for attachlni coid-rormed steel panels to supporting members .' . -; 5-115
".
,"" 1
THREADED WELDING STUDS
li;f - --
IL '
PLAN - aOLTED CONNECTION -TYPES Alia
~
'\t
~ I.
IS"
8"
0
~ ~ 24" PANEL WIDTH
SECTION
Flgure 5-38
j-i
I}:i STEEL PANEL
BOLTED CONNECTION TYPES A&B
PLAN -
I.
12:~
~', THREADED WELDING STUDS, lOU. ~'1 T,HREADED WELDING STUDS, III GA.
,
'W
\1i
'24" PANEL WIDTH'
SECTION -
f
,12",
18"
.
"1
W ~~ .1
3" STEEL PANEL
Typloal ,bolted oonneotlons tor attaohln. oold-tol'llled ete.l panele to s upportl ns 11811!bers 5-116
-+--r--\---r+ ,
J I i
'" 'II ,~.
. ,
., ,
1
h~
(
.'
1
PLAN
.
lit· •
- SPACER I I~" • \fa" WITH HOLE ~'.' I~I." •. HOLE It! PANEL . ~: I
WELDED THREADED STUD ~VY HEX. NUT .
WASHER
3:' STEEL PANEL
1~::::i:::t:i'!:L (OVERSIZED)
SUPPORTING MEMBER
SUPPORTING MEMBER r -,
SECTION
. '"
TYPICAL
SECTION BOLTED CONNECTION
.
.
,
.
.,,'
.
"
,
.
.
-:.. '
.
,
. ,"
.+--+--~~--+
PUDDLE'WELD
sf; -," PUD~~E 'WE~D STEEL FLAT PANEL WITH SHEET
PLAN . .
,
~.
~
..
.; PLAN
PUDDLE WELD
~.
. •.
I·
r. -
-~~",.~"lpllf~
SUPPORTING MEMBER _
," •
~4
PUDDLE WELD STEEL' PANEL WITHOUT FLAT SHEET
- SUPPORTING MEMBER
TYPICAL WELDED CONNECTION
Figure 5-39
Det~hs of t'yJICal f'asteners for col'd-fof'llled steel panels
,.
tl :;..
I
".
,
II ~
I
.
'..r
~
--'
---_ .....
'\ ~
li4'CL~'~~~;;~r
1/2" STIFF. IL.
u
-===
3 16
TS5.3•.375 . 1/4,
-N
L
HARDWARE . PANIC . . I.
4.3.3/B . -'-'-v'...L"'"lt'-
.I,
LATCH I -RELEASE U - '11 I.
r-;
-
. I" ~ ALUM. PIN
3/8" R.-
~
BLDG. FRAMING
.HOLE DIA. = . PIN DIA. + 1/64" •
y
~ .:-=.
Il
8'1~ .1/2"
. ·.·A
,. EXTERIOR
ELEVATION
SECTION
X-X
BLDG. FRAMING TS 5.3 ••375
3/16
1/4
. HINGE 1/4
SEC'TlON
y~y
SECTION
Z- Z
nO
Figure 5-40
Single-leaf blast door il)sta11.ed.. in a flt.eel struoture 5-118 ,
,
-l!I' _
2 ANCHOR BARS AT 12" nc,
~ CL.
i
"".,,," /2. . "\A
x " ., I
I •
•
!
'0
,
'
-...- :
'1;'STlFF.t I • ~ 1<.."
, ~2
~
'.
-. , . I LATCH. , , RELEASE , I
I
. -I
_u
~
:'.f!
::s"j __
',',
FIN. FL.
....
"'j-'
-
u·
1'2II ALUM. PIN ~,--'-"~- PANIC HARDWARE
Ye" Jl
L 4"", 3"x 1;2" I', ."
'
."
'~
:,. •
I I
I I INACTIVE I ACTIVE LEAF. 1 L;~~.F fJ, I .." " .
.
-~r~~~'
'21~ I
I
.J
HOLE OIA.' PIN OIA. + 1/64"
J "u:..~
~~~ .... ..... .... ......
+1.' X
' B"X 4"" 2 I" , .LX
+ANCHORS
EXTERIOR ELEVATION
'~"" , ,
.
."
".
,-
-
.'.
,"
AOO'L 't4' _ x IB" LG, ANCHOR AT BLOCK
SECTION X - X
'I; _ANCHOR' BARS AT 12"O.C.
L4"X3"X't;~
l'
L~IIX3IlX~· . I .... ;fa I<.
;' "
.\
SECTION
I
Y- Y
,
"..
FigUJ"e 5-41
Double-leaf blast door installed in a concrete structure 5-119
I
BLAST ••
,
,
I""
ABUTMENT LOCATION
COMPRESSION ARCH
:::?CE~~~D'ToO=R'F'T;='===9=:t=::s::::::l= ij~!:t:<=-':" "'t":- :HINGE
r----I
.' .. .'
.
BEARING, HOLDDOWN BqLT
. .'
..
ANCHOR--/ .' .t
-..
'.
.~
HORIZONTAL SECTION' COMPRESSION ARCH
COMPRESSION TIE BAR
DOOR
BLAST
'J
, ••.•
• ""_',' '," ""'.'
:. .
• 6"
;
,--
HORIZONTAL SECTION TENSION ARCH DOOR . ., ,
Figure 5-42
.I..O~.
'ANCHOR, fb~I:::::::o=~=?~'~BEARING '~, " " "~o' , '. : :. . .... 1),,- . TENSION ARCH:,,, ",.,:' HOLDDOWN BOLT .'c, ,:: .,,'
,
-'
.
"
Compression-arch ,and tension-arch blast doors , ' 5-120
TN 5-1300/NAVlAC P-397/AFR88-22
APPENDIX 5A· ILLUSTRATIVE EXAHPLES
TH 5-1300/NAVFACP-397/AFR 88-22 " This appendix presents detailed, design procedures and numerical 'examples on the following topics: 1.
Flexural elements subjected to pressure-time loading,
2.
Lateral bracing requirements
3.
Cold-formed steel panels
4.
Columns and beam-columns
5.
.,
. Open-web joists
6.
Single-story 'rigid frames
7.
Blast doors
8.
Unsymmetrical bending
References are' made to the appropriate sections of this chapter and to charts, tables, and equations.from Chapter 3 "Principles of Dynamic Analysis". Problem SA-l 'Deaign of'Beama for Preaaure-Time Loeding I
Problem: Procedure: Step 1.
Step 2.
Design of a purl in or girt as a flexural member which responds to a pressure-time loading. Establish the design parameters: a.
Pressure-time load
b.
Design criteria: Maximum' support rotation, e, depending on protection category.
c.
Span length, L, beam spacing, -b, and support conditions.
d:,
P,roperties and type of steel used, ':". f y alld E.
Determine the equivalent static load, w, using the following preliminary dynamic load factors as discussed in Section 5-22.3. 1. 0 for e - 2' DLF -
',0.5 fore - :12'
.•
Step 3.
Using the appropriate resistance formula from Table 3-1·and the equivalent static load derived in Step 2,. determine. ~.
Step 4.
Select a member'size using equation 5-7 or 5-8. Check the local buckling criteria of Section 5-24 for the member chosen.
"
5A-1
TN 5-1300/NAVFAC P-397/AFR 88-22 Step 5.
Determine the mass, m, includi~g the weight of the decking over,a distance center-to-center of purlins or girts, and. the weight of· the members.
Step 6.
Calculate the equivalent mass Me using Table 3-12 (Chapter 3).
Step 7. '
Determine the equivalent elastic' stiffness KE from Table 3.1.
Step 8.
Calculate the natural period ·of vibration. TN' using equation 515.
Step 9.
Determine the total resistance.Ru. and peak pressure ,load, P. Enter appropriate chart in Section 3-19.3 with the ratios T/T N and P/Ru and the values of Cl, and C2 in order to establish the ductility ratio ~.
Step 10.
Check the assumed DIF used in Step 4. Enter the, response charts with the ratio T/T N and ~ and to determinetE.Using,equation 51, determine the strain rate. Using figure 5-2, determine the DIF and C. If there isa significant difference from t~at assumed. repeat Steps 4 through 9.
Step 11.
Calculate the equivalent elastic deflection,XE as given by the equation
and establish the maximum
deflection.~
,given·.by
Compute the corresponding member end rotation. with the criteria summarized in Section.5-35. tan
e-
Compare
e
'.
Xm/(L/2)
Step 12.
Check for shear using equation 5-16 and Table 3-9.
Step 13.
If a different member size is required, repeat Steps 2 through 12 by selecting a new dynamic load factor. ·Ezample SA-l
Required: Step L,
Design of a Beam for Pressure-Time Loading
Design a simply supported beam for shear and flexure in a low pressure range where personnel protection is required. '-Given:". a.
Pressure-time loading (Figure SA-l)
SA-2
b.
Criteria:
c.
Structural configuration (Figure 5A-l).
d.
f y - 36 ~si, -E - 30 x 10 3 ks~,. ,A36 steel
Personnel protection required. limi ted to 2' .
Support rotation
. ,
•
.
..
>';"1.l
~J
.•
.,
e . ' Compression flanged. braced.
v:
,
1. ,
17'
. ,spacing
2.
;:)
6.5 c,psi . ~,
en 11.I
a: a.
~
40ms
'TIME
Beam configuration and/loading, Example.SA-l
the equivalent static ,load (i.e., required resistance). For this pressure range, the equivalent static load is assumed equal to ,the peak pressure'(Section '5-22.3)., The running load becomes: w - 1.0 x 6.5 x 4.5 x 144/1000 - 4.21 k/ft"
Determine required wL2 ~--
~;
-.
4.21 x 172
- - - - - 152.1 k-ft
(Table 3-1)
8
8
Step 4.
. a:en
" 'Determine
,. Step 3.
1
b·4.S'
Figure 5A-l Step
11.I
,,
Select a member. (S + Z)
- -- - - - - - - - - 2~
2 x 152.1 x 12
t,
(Equation 5-7) "
where fds - a x c x f y - 1.1 x 1.29 x 36 - 51.1 ksi
(Equation 5-2)
TK 5-1300/NAVFAC P-397/AFR 88-22' 'where' a - 1.1 from Section 5-13.2 c - 1.29 corresponding to a DIF in the low pressure range (see Table 5~2) Select W12 x 26,
S
z S+ Z ~
- 33.4 in3' " - 37.2 in3
-
I' - 204 in4
70.6 in 3
- (70:6 x 51.1)/(2 x 12) - 150.3 k-ft
Check .Lccal. buckling cd teda. (Equation 5-17)
d/tw - 53.1 < 412/(fy)1/2 - 68.7 O.K. b f/2tf - 8.5 Step 5.
(Section 5-24)
O.K.
Calculate H. wL '. [(4.5 x 4.8)'"+ 26] (17 x 10 6 )
- 25,130 (k-ms 2)/ft
H - --' - - - - - - - - - - - - -
g
32.2,
x 1000
~,
Step 6.
KLH' -
Step 8.
(Table 3-12)
(0.78 + 0.66)/2 - e.72 ..
He - 0.72 x 25,130 - 18,100 k-ms 2/ft Step 7.
-
Calculate the effective mass,' He; for a response ,in the elastoplastic range.
-:
Determine KE . 384 EI
384 x 30 x 10 3 x 204
5L3
5 x 17 3 x 144
(Table 3-8) ,
Calculate TN' TN - 2"(He/KE)1/2 - 2"(18,100/664)1/2
5A-4
(Equation 5-15)
TM.5-1300/NAVFAC P-397/AFR 88-22 Step 9.
Establish the ductility ratio
~
and compare with the criteria.
T/TN - 40/32.8 - 1.22 .. 6.5 x 17 x 4.5 x 144 .P - P x ~ x b - ---------,-~
71.6 kf.ps :
1000 ~
-
P~
8~/L
-
- (8 x 150;3)/17 - 70.7 kips
71.6/70.7 - 1.01
From figure 3-64a, ~
- XmIXE- 2. 1
At this point, the designer would check lateral bracing requirements. Sample problem '5A-2 outlines this procedure. Step 10.
Check the assumed DIF. From Table 3-64a, for PfRu - 1.01 and.T/T N - 1.22.
tE - 0.24 x 40 - 9.6 ms Find
E
:
E - fds/Est E - 51.1/30 x 10 3 x .0096 -·0.177 in/in/sec
(Equation 5-1)
From figure 5-2 O.K .:
DIF - 1.31 - 1.29. Step 11.
Determine XE: . XE -
~/KE
- (70.7 x 12)/664 - 1.28 inch
.,
5A-5
TH 5-l300/NAVFAC P-397/AFR 88-22 . , Find Xm: Xm -
~E
- 2.1 x 1.28 - 2.69 inches
Find end rotation, 9. 2.69/(8.5 x·12) - 0.0264
tan 9 - Xm/(L/2)
e Step 12.
(Table 3-5)
."
1.52" < 2"
O.K.
Check shear. Dynamic yield stress in shear "
f dv - 0.55 f ds - 0.55 x 51.1 - 28.1 ksi
'
(Equation 5-4)
Ultimate shear capacity Vp - f dv x
Aw -
'.
(Equation 5-16)
28.1 x 0.23 x 12,"" 77.6 kips
Maximum support shear
",
Vs - r u x L/2 -,Ru/2 - 70.7 /2
~35.4
kips
.""
.: (Table 3-9)
.'.t'<
Problem 5A-2 Problem:
Spacing of Lateral Bracing
Investigate the adequacy of the lateral bracing specified for a flexural member. '; , The design procedure for determining the maximum permissible spacing of lateral bracing is essentially a trial, and error procedure if the unbraced length is determined by the consideration of lateral torsional buckling only. However. in practical design, the unbraced length is usually fixed by the spacing of purl ins and girts and then must be investigated forelateral ,; .. torsional buckling.
Procedure: Step 1.
Establish design parameters. a.
Bending moment diagram obtained from a design analysis.
5A-6
TM 5-l300/NAVFAC P-397/AFR 88-22
Step
Step
~.
3.
Step 4.
b.
Unbreced length, l. and radius of gyration of the member, r y• about its weak axis,
c.
Dynamic design.strength, f ds'
d.
Design ductility ratio,
~.
(Section 5-l3)
from a design analysis.
From the momant diagram, find.tl\e end moment ratio, H,tM", for each segment of the beam between points of bracing. (Note that the end moment ratio is positive when the segment is bent in reverse curvature and negstive when bent in single curvature). Compute the maximum permissible unbraced length, lcr. using equation 5·20 or 5-21, as applicable. Since the spacing of .purlins and girts is usually uniform, the particular unbraced length that must be investigated in a design will be the one with the largest moment ratio. The spacing'of bracing.in.nonyielded segments of a member should be checked against the requirements of Section 1.5.1..4.5a of the AISC Specification (see Section 5-26.3). The actual length of a segment being investigated should be less than or equal ·to lcr' Example5A c2
Spacing of Lateral Bracing
Required:
Investigate the unbraced lengths shown for the W10 x 39 beam in figure 5A-2.
Step 1.
Given: a. Bending moment diagram shown in. figure 5A-2. b.
Unbraced length (each segment) ry
c.
Dynamic design stress - 51.1 ksi
d.
Design .ductility
ration,·~
- 36 inches - 1. 98 inches
- 5.
Step 2.
The moment ratio is -0.5 for segments BC and CD (single curvature) and 0.5 for segments AB and DE (double curvature) .
Step 3.
Determine the maximum permissible unbraced length. equation 5-21 results in ·the lower value of lcr'
. '
."
5A-7
By inspection,
TM 5-1300/NAVFAC P-397/AFR 88-22
'.
36"
36"
36"
36"
-M p
-M p
A,E SUPPORTS
W10X39 A
B,C,D BRACING LOCATIONS
E
C
Mp
Figure 5A-2
Bending moment diagram, Example 5A-2
From figure 5-9 for
Xm/xE - 5,
.
B-
, 1375 x '1. 98 lcr - - - - - - 1.36 x 51.1 Step 4.
Problem:
Design a Roof Deck as a Flexural Member which Responds to Pressure-Time Loading Design of cold-formed, light gauge steel panels. subjected to pressure-time loading.
Establish the design parameters: a. b.
Pressure-time loading Design criteria: whether
Step 2.
39.2"
Since the actual unbraced length is les~ than 39.2 inches, the spacing of the bracing is adequate.
Problem 5A-3
Step 1.
1.36
Specify values of
tension~membrane
~
and
e
depending upon
action is present or not.
c.
Span length and support conditions
d.
Mechanical properties of steel
Determine an equivalent uniformly distributed static load for a 1ft width of panel, using the following preliminary. dynamic load· factors.
5A-8
-t.
TK 5-1300/NAVFAC P-397/AFR 88-22'
Tension-membrane action not present
Tension-membrane action present
1.33
DLF
1.00
These load factors are based on an average value of T/TN - 10.0 the recommended design ductility ratios. They are derived 'using figure 3-64 of Chapter 3. Equivalent static load
w b
DLF
x
P
x
b
1 ft,
Step 3.
Using the equivalent load derived in step 2, determine the ultimate moment capacity using equation 5-29 or 5-30 (assume positive'and negative.are the same).
Step 4.
Determine required section moduli using equation 5-27 or 5-28. Select a panel.
Step' 5.
Determine actual section properties of the panel:
Step 6.
Compute r u' the maximum unit resistance per loft width of panel using equation 5-29 or 5-30.
Step 7.
Determine the equivalent elastic stiffness, KE- ruL/X E• using equation 5 - 31.
'·S
,:
" ••
-.
I
Step 8.
Compute the natural period of vibration. (Equation 5-32)
Step 9.
Calculate P/r u and T/T N. Enter figure 3-64 with the ratios P/r u .and T/T N to establish the actual ductility ratio ~ . Compare ~ with the criteria of step 1. If criteria value, repeat steps 4 to 9. -,
Step 10.
~
is larger than the
Compute the equivalent elastic deflection,XE using XE = ruL/K E. Evaluate-the maximum, deflection, ~ = ~XE' Determine the maximum panel end r ot.at Lon ..
e - _tan -1
[~/(L/2) 1
Compare e with the criteria of step 1. If e is larger than specified in the criteria, select another panel and repeat steps 5 to 10.
5A-9 .
TH 5-1300/NAVFAC P-397/AFR 88-22 Step 11.
Check resistance in rebound using figure 5·13.
Step 12.
Check panel for maximum resistance in shear by applying the criteria relative to: a.
Simple shear, Table 5·5a, 5-6a or 5-7a.
b,'
Combined bending and shear, Table 5-5b, 5-6b or 5·7b.
c;
Yeb crippling, figures 5-15 or 5-16. If the panel is inadequate in shear, select a new member and repeat steps 4 to 12.
Example 5A-3 Design a Roof Deck as a Flexural Kember which Responds to Pressure-Time Loading Required:
Design a continuous cold-formed steel panel in a low pressure range.
Step 1.
Given: a. Pressure-t,ime loading (figure 5A-3),. b.
Criteria:
(Tension-membrane action present)
maximum ductility ratio
/.I max - 6
maximum rotation
9 max - 4·
c.
Structural configuration figure 5A-3.
d.
Steel A446, grade a E
30 x 10 6 psi
f ds - a x c x'fy- 1.21 x1.1 x Step 2.
33,OO~
- 44,000 psi (Equation 5-26)
Determine the equivalent static load. Say DLF - 1.0 w - DLF x P x b - 1.0 x 5.0,x 12 x 12 -'·720'lb/ft
Step 3.
Determine required ultimate moment capacities. selection, assume
For preliminary
Mup - Mun - wL2/10.8 - 720 x (4.5)2/ 10. 8 - 1,350 lb-ft , (Equation 5-30)
5A-IO
5-1300/NAVFAC,P-397/AFR~8~;22
TlI
'L
,
..
'.
:~
,'.t:
'~.
f;
."
w
~
S.Opsi'
fa
I. '.
u: Do
.'
2'-0', , .
40ms
TIME ,
"
..... '
.. I"
.'
t----I~I----=--=--1 4'-6"
4'-6'
••. 1'
Figure 5A-3
'4'-6"
•
j
,
,r
,
Roof decking configuration and loading, Example SA-3
~ ':'t ,-_ ~
":~'~f
,;.
SA-ll.
'
,
I'
",
TM 5-1300/NAVFACP-397/AFR 88-22 Step 4.
Determine required section moduli. S+ - S- - (1350 x 12)/44,000 - 0.368 in 3 (Select Sec. 3-18,
Step 5.
1~1/2
inches deep)
Determine actual section properties. For manufacturer's guide: S+ - 0.3.98 in3 S- - 0.380 in 3 4
1 20 - 0.337 in4
w - 2.9. psf Step 6.
Compute maximum unit resistance r u' ~n
- (44,000 x 0.398)/12 - 1,459 Ib-ft
M
-
(44,000 x 0.380)/12 - l,3931b-ft .
;(Equation 5-27): (Equation 5-28)
3.6 4.5 2
(1,393 + 2 x 1,459) - 766 Ib/ft (Equation 5-30)
Step 7.
Determine equivalent static stiffness.
0.0062 L3
- 124,260 Ib/ft
0.0062 (4.5 3) x 144 Step 8.
(Equation '5-31)
Compute the natural period of vibration for the I-ft width of panel. mL - wig -' (2.9 x 10 6
~
.-
4.5)/32.2 - 4.05 x 105 Ib;ms 2/ft
5A-12
TK S-1300/NAVFAC'P"397/AFR 88-22' TN - 2x[(0.74 x 4.0S x 105)/124,260]1/2, - 9.7S'msec Step 9.
Calculate P/ru and T/T N P - P x b - 5.0 x 12 x 12, - 720 lb/ft P/r u - 720/766 - 0.94. T/T N - 40/9.75 - 4.·10 Entering figure 3-64a with ,these values. , ,
Xm/XE - 3.5 < 6 Step 10.
O.K.
Check maximum deflection and rotation. ,
XE -
r~L/KE
- 766 x 4.5/124,260 - 0.028 ft
Xm - 3.5 XE - 0.098 ft
eStep 11.
tan- l [x;.i(L/2)]
~ t~n-l[0.098/2.25·] -
2.5 < 4"
O.K.
Check resistance in rebound. From figure 5-13, r /r' - 0.33; ,O.K. ,since available maximum elastic resistance in rebound is approximately equal to that under direct loading. "
Step 12.
Check resistance in 'shear. a.
Interior support (combined shear and bending). Determine dynamic shear capacity of a loft width of panel: h - (1.500 - 2t) inches, t - 0.048 inch - 1.500 - 0.096 - 1.404 inches "
hit - 1.404/0.048 - 29.25 - 30
SA-13
"
,1
TN 5-1300/NAVFAC P-397/AFR 88-22 f dv -,10.84 ksi -
"
.
.',
~
Total web area for loft width of' panel: '
(Table 5-5)
...... ,
~
"
(8 x h x t)/2 - 4' x 1'.404, x 0:048 ;.' 0:270 in 2 Vu - 0~270 x 10.84 - 2.92 k - 2,922,~b' Determine maximum dynamic shear force: The maximum shear at, an interior ,support: of panel using limit design is:
a ,continuous
Vmax - 0~55 ruL - 0.55 x 766 x 4.5 - 1;896 lb ~
'
b.
, . '; "
- 1,896 lb < 2,922 lb ,, .. ... ' End support (simple shear)
.
11
O.K, ','
Determine dynamic shear capacity of a l·ft width of panel: .,. , For h/t S 57,
f dv - 0:50 f ds -.0.5 x 44.0 -,22.0 ksi, , . ...' . ' '- '",,'. ' (Table 5- 5a)
'Vu' - 0.270 x 22,000 -5 ..940
., .,
~,
'
'
Determine maximum dynamic shear force: The maximum shear at an end support,of'a continuous panel using limit design is ,V!Dax'- 0.45 x ruxL -0,45 x 766 x 4.'5 - 1,551 lb < 5,940 lb c,
O.K.
Web crippling (4 webs per foot) End, support (N -
2~~
inches)
Qu - 1,200 x 4 - 4,800 lb> 1,551 ,"O.K.
(figure 5-15)
TK Interior support (N - 5 inches)
Qu
P-397/AFR 88-22
S~1300/NAVFAC
j ..'
~
(2,400 x ,4)/2 - 4,800 Ib > 1,896 O.K. ... • Problem 5A-4 Design of Columns and Beam-Columns Problem:
Design a columno,r beam-column for axial' load combined with bending ,about the strong axis. "
"
Procedure: Step 1.
.
" Establish design parameters. , ' Bending moment ,M, axial load P, and sheai',V are obtained from either a preliminary design analysis or a computer 'analysis.
,
a.
-,
(figure 5-16)
"
'
b.
Span length 1 ,and'unbraced lengths Ix and ly.
c.
Properties of structural steel: Minimum yield strength f y ,
,
'
Dynamic increase factor c
(Table 5-2)
Dynamic, design 'strength' f ds,
,
(Equation 5-2)
Step 2.
Select a preliminary member size with a section modulus S such that S'~,M/fds and ~f/2tf complies with the structural steel being used (Section 5-24).
Step 3.
Calculate" P (Section 5-,24) and the ratio P/Py. ,Using either 'equation '5.17 or 5-18, determine the maximum allowable 'd/tw ratio and 'compare it .ro thar. of the section chosen. ':,If :the allowable d/tw ratio is less than that of the trial section, choose a new trial section.
Step 4.
Check the shear capacity of the web.
Determine the web area
Aw
(Section 5-23) and the allowable dynamic shear stress f dv ,(equation 5-4). Calculate the web shear capacity Vp (equation 516) and compare to the design shear V. If inadequate, choose a ' ' , new trial section and return to Step 3. Step,S.
Determine the radii of gyration, r x and'ry, and"plastic section modulus, Z, of the trial section from the AISC Handbook .
. '
f ~l.:
SA-1S
1M 5-l300/NAVFAC P-397/AFR 88c22 Step 6.
'.
Calculate the following quantities, using the ,various design parameters:
a.
... .'
"' Equivalent plastic resisting moment
,.
.
.
(Equation 5-8)
b.
Ef~ective slenderness ~atios Klx/r x and Kly/~y' For the effective length factor K, see Section 1.8 or the Commentary on the AISC Specification and Section 5-38.
c.
Allowable axial stress Fa corresponding to the larger value 'of K1:/r. ,,' •• ':" "., '" .
d.
Allowable moment
Hm
from equation 5-47 or 5-48.
e . F ' eand "EuLer " buckling ,load 'Pe (Section 5-37.3).
Step 7.
f.
Plastic axial load (Section. 5:37 .3) and ultimate axial load Pu (equation 5-42).
g.
Coefficient Cm (Section 1.6.1 AISC Specification).
Using the quantities 9btained in Step 6 and the applied moment M and axial load P, check the interaction formulas (equations 5-44 and 5-45). Both formulas must be satisfied for the trial section to be adequate. " .', Example 5A-4 (a)
Design of a Roof Girder as a Beam-Column .
.
J'
'"
i .
Required:
Design a fixed-ended roof girder' in a' framed structure for combined bending and axial load in a·,low pressure, range.'
Step 1.
Given: a. Preliminary computer analysis gives the folloWing values for . , design: ..
0'
"
.'~
b.
Mx
115 ft-kips
,My
0
-,
..
P
53.5 kips.
,V
15.1 kips,
Span length Ix
"
' ' 1'.
"
-'
17' -0"
..
Unbraced lengths Ix - 17'-0" and ly - 17'-0" c.
A36 structural
~teel
5A-16
TM5-1300/NAVFAC P-397/AFR 88-22 f y - 36 ksi r ,(Table 5-2)
c - 1. 29
(Section 5-12.1)
a - 1.1 fds - c x a x f y
~
(Equation 5-2)
1.29 x 1.1 x 36 - 51.1 ksi
Step '2. " S - Mx/f ds - 115 (12)/51.1 -:27.0 in 3 Try W 12 x 30 (S - '38.6 in 3) A-8.79in2
.
"
d/t,. -'47.5,
b f / 2t f ~ 7.4 < 8.5
..
,-
O.K.
(Section 5- 24)
',
Step 3. Py - Afy P/Py d/t,. -
,
,
8.79 x 36,-
53.5/316 -
0.169
[412/(f~)1/2] [1-
\
316 kips
(Section 5-24)
< 0.27 1.4 (P/Py )]
(Equation 5-17)
- (412/(36)1/2)' [1 - 1.4 (0.169)J ~' 52.4
>'47.5
O.K.
Step 4. (Equation 5-16) -,
f dv -'0.55 fds -
, Aw -
t,.(d - 2tf) -
0.55 (51.1) -
. f
'" (Equation 5-4)
28.1 ksi
0.260 [12.34 - 2 (0.440)]
- 2.98 in 2 Vp -
28;1(2.98) ..
(Section 5-23) 83.7 kips
> 15.1 kips
5A-l7 e..
O.K..,
TH 5-1300/NAVFAC P-397/AFR 88"22 Step 5.
r x - 5.2J. in. r y - 1. 52 in.
, (AISC Manual)
Z - 43.1 in 3 Step 6. a.
~x
b.
K - 0.75
c..
- fds x Zx -
51.1 x 43.1 x 1/12 -
183.5·ft-kips (Equation 5-8)
(Section 5-39)
K1 x/rx -
[0.75(17)12J /5.21 - 29
Kl y/ry -
[0.75(17)12) (.1.52 - 101
Fa - 12.85 ksi for Kl y/ry - 101 and f y - 36 ks L, (Appendix A, AISC Specification) -1.42(12.85) - 18.25 ksi for fds - 51:1 ksi
Hmx -
d.
[ 1.07
-
(l/ry) (f ds ) 1/2
J'
., 3,160
~x:S
J\,x
: .
(Equation 5-47)
(204/1. 52) (~1.1) 1/21 . - (1.07 - ----------~----J183.5 - 140.6 < 183.5 ft-kips 3,160 ,I'·
127T2E e.
F' ex -
23 (K1 b/r x) 2
23AF.'ex Pe x
f.
-
PP - f ds A -
----;-~:---
177.6 ksi .
23(29)2
(Section 5-37.3)
23(8.79)177.6
~
12
127T2(29.000)
12
2,992 kips: . (Section 5-37.3)
51.1(8.79) - 449 kips
5A-18
(Section 5-37.3)
TK 5-1300/NAVFAC
P~397/AFR.88~22
(Equation 5-42)
Pu - 1. 7AFa - 1. 7(8. 79)18.25. - 273 kips ..1
g. Step 7.
C
(Section 1.6.1,'AISC Specification)
- 0.85
P
(Equation 5-44)
".....,·+L~.,.:...,---
Pu
(1 .,.;p ' •"
-
• , o.
53.5 0.85(115) -- + 273 (1- 53.5/2992)140.6
f
0;196 +'0.708 -' 0.904 < 1 '. O.K.
,.
" 5-45) (Equation
'", •
J
,'"
53.5
115 + - . 0 ' . 1 1 9 + 0.531\- .0.650 449 1.18(183.5)
.. ~...-!'-.-
. .< 1
O.K.
Trial section meets the requirements of Section 5-37.3.
j
...
." ..
Example 5A-4 (b) .• " ••' . J
~
•
Design of Column
1..';
.,.
Required:
Design of an exterior fixed-pinned column in a framed structure for biaxial bending plus axial loads' in a low pressure range.
Step 1.
Given: a. Preliminary design'~nalysis .~f a particular column gives the following values at a critical section:
Hx -
311 ft-kips
My - 34 ft-kips P - 76 kips V-54 kips b.
Span length 1 - 17'-3" Unbraced lengths lx - 17'-3" and ly - 4'-0" (laterally supported by wall girts).
5A-19'
"
.
TM
5-1300/NAVFAC'P~397/AFR
c.
88-22
A36 structural ·stee1·. fy
-
c
- 1. 29
36 ksi .(Table 5-2)
a , - 1.1 Step 2.
f ds -
(Section 5-12.1)
x c x f y - 1.1·x '1.29 x 36 - 51.1 ksi (Equation 5-2) S - ~/fds - 311(12)/51.1 - 73.0 in 3 Try
a
W 14 x 68 (S - 103 in 3 ) ,
A - 20.0 in 2
d/~
- 33 ..8
O.K.
. (Section 5-24)
Step 3. Py - Afy - 20.0(36) - 720 kips P/Py - 76/720 - 0.106
(Section 5-24)
0.27
<
d/~ - [412/(fy)1/2 I [1 -·1.4(P/Py ) l -[412/(36)1/2] ,[1
(Equation 5-17) 58.5 > 32.9'
1.4(0.106)]< -
O.K.
Step 4.. (Equation 5-16) f dv - 0.55 fds -
Aw -
~
0.55(51.1)
28.1 ksi
..
(Equation 5-4)
~(d : 2t f)- 0.415 [14.04 .,2(0.720')] - 5.23 in 2 ,
(Section 5-23)
Vp'- 28.1(5.23) - 147 kips
~
...
54 kips .
O.K.
, !'
I" .
~
5A-20'
.. '
-c '
TH 5 -1300/NAVFAC P- 397IAn. ,'88,-22 Step 5.
r x - 6.01 inches
f· ...
r y - 2.46 inches
~ - 115 in3
Zy
,
.
(AISC Manual)
-'36'.9 in 3
Step 6.
~-,
a.
"
~ - fdsZ
,
.
(Equation 5-8) '
~x
- 51.1 x 115 x 1/12 - 490 ft-kips
~y
- 51.1 x 36.9 x 1/12 - 157 ft-kips
-'.1
b.
UseK - 1.5 'Kl
-
x
rx K1 y
_
", (Section 5-39)
1.5(17.25)12, 6.01
c.
-
52
0,
1.5(4.00)12 2.46
ry
:
Fa - 18.17 ksi for
r
-
K1~
29
Ir x -
'.
52 and f y - '36 ksi
'
1.42(18.17) - 25.79 ksi for fds - 51,1 ksi '.. ..t
Kmx - ~~ -
.. ~.:- .
'I ,
~,
•., ..:.
490 ft-klps " c;
d.
Hmy -
~y
- 157 ft-kips
,"
(Section 5-37.3) "
.'
'
12"2E 12"2(29,000) F' ex' ,.. ---'-....,....-...,-2 . 23 (K1b
irx)
12"2 E
"
'123(29)2
178 kSi (Section 537.3) ',' .
"
..
I' "
,
5A-21
,'.
TH 5-1300/NAVFACP-397/AFR.88-22 23(20.0)55.2
. '. - 2,116 kips 1 2 . , ' 'r,
23AF'ex Pe x -
,
.
12
23(20.0)178
23AF'ey Pey -
;;. :
1
~
- 6,823 kips 12
P.p - fdsA -
12 51.1(20)
I:',
Pu - 1. 7AF a -
(Section ... 5-37.3)
- 1,022 kips
1.7(20)25.79 4·
Cmx - Cmy - 0.85
- -,
..L'
':
-877 kips ~
.',
(Section 1.6.1, AISC Specification) ... •
'
Step 7. P
.
. CmxHx
-+ Pu (1 - P/Pex)Kmx
:i
+
(Equation 5-44)
1
76 0.85(311) --+ 877 (1 - 76/2116)490
0.85(34)
"
+
{I - 76/6,823)157
0.087 + 0.560 +,/0.186 '.,.
0.,83~
<.1 ",
O.K.
,
. .
-"
P/Pp + MX/(1.18~x) + ~/(1.18~y)
S
'1..i. •
1
(Equation 5-45)
(
76/1022 + 311/[1.18(490)] + 34/[1.18(157)] ~J
0.074 + 0.538 + 0.183 - 0.795 J ',(; •. ,""
<
~
I
"
1
,,
Trial section meets the- ,requirements, of··Section .5·37.3 \ M<
Problem 5A-5
•
i ~' ,.' ;
. " . \,
Desigrr of: Open-Web
.,
.
Stee1,~oists
Pzob Lem:;" .' Analysis or design ot an open-web joist subjected to a pressuretime loading. Procedure: Step 1. Establish design parameters. a.
Pressure-time curve
b.
Clear span length and joist spacing
5A-22 J
. . . .~
"
TK 5-1300/NAVFAC P-397iAF& 88-22: c.
Minimum yield stress f y for'chord and web membars Dynamic increase factor, c. ,
~
,
(Table 5-2)
and maximum end rotation,
e.
d.
Design ductility ratio
e.
Determine whether 'joist design is controlled by maximum end reaction.
Step 2.
Select a preliminary joist size as follows:
a. , ·b.
Assume a dYnamic load factor
, (Section 5-22.3)
Compute equivalent static load on joist due to blast overpressuy;e
wl - DLF x P x b , (Dead load of j oLs t; and decking not included) c.
d.'
Equivalent service live load on joist w2 - wl/l.7 x a xc
"'(Section 5-33)
From "Standard Load Tables" adop ced by' the Steel Joist Institute, select a joist for the' given span'and the structural steel being used, with a safe service load (dead load of joist. and decking excluded)'equalto or greater than w2' Check whether ultimate capacity ot joist 'is controlled by flexure or by shear.' ',;.'.'
Step 3.
Find the resistance of the joist by multiplying thci'safe service load by 1.7 x a x c (Section 5-33).
Step 4.
Calculate the stiffne~s of~he' joist; KE,,:usl'ng Table 3-8. Determine the equivalent elastic deflection XE given by
Step 5."
Determine the effective'mass using the weight of the joist with its tributary area' of' deck'Lng , and, the corr e spondfng load-mass factor given in Table 3-12 of Chapter 3.' \
Calculate the Step 6.
'
'1
natura~ .~eriod of,vibrati~n,
TN;
Follow procedure outlined in step 6a or 6b'dependlng on whether the joist capacity ~s controlled by fle~ure or by shear .
.
5A-23
TH S-1300/NAVFAC P-397/AFR 88-22, Step 6a.
Joist design controlled by flexure. a,
Find ductility ratio ~ - Xm/X E from the ~esponse charts in Chapter 3, using the values of T/T N and P/ru'
b.
Check if the ductility ratio and maximum end rotation meet , r the criteria requirements outlined in Section 5-35.
!
, If the above requirements are not satisfied, select another dynamic load,factor and repeat Steps,2,to 5. c. ' "
, Check the selection of the, dynamic ipc~ease factor used in Step 2c. Using the response charts, find t E to determine ,the strain rat~, £,in equation 5-1. Using figure 5-2, '. determine DIF. (If elastic response, use T/T N and appropriate response charts to check DIF).
d. Step6b.
Check if the top chord meets the requi,ements for a beamcolumn (Section 5-37.3).
Joist design controlled by shear. ,~.
b.
Find ductility ratio~ - Xm/X E from the response charts in Chapter 3, ~sing the values of T/T N and,P/~u'
.If, ..
'
.'
~ S 1,,0, d~sign~sO.K., . ,
! ",.',
If ~ > 1.0, assume a highe~ dynami~ load factor and repeat Steps 2 to 5. ContiLnue until ~ S I.O.Check end rotation, e, agaLns t de sLgn criteria. . '. c
c.
Check the selection of the dynamic increase factor used in St.ep 2c, usirg the value of T/TN and the, appropriate elastic response cha,t in ?ection 3-19.3.,
d.
Since the capacity is controlled by maximum end reaction, it will generally not be necessary to check the top chord as a beam-col,umn. .1:I0fev~r, wh!,n such a check is ..w arranted, the' procedure'in ~tep 6a can,be followed. "', . '.' ..
..
./
Step 7.
"
Check the bottom chord for rebound. a. b.
c.
.
. ( ."
.
•
!
.
Determine the required resistance, r, for eiastic behavior in ,rebound. "'"
'comp~te
-
the' bending moment, 'H, and'find'the axial forces in top and bottom chords using P - H/d where d is taken as the distance between the centroids of the top and bottom chord sections. Determine the ultimate axial load capacity of th~ bottom chord considering the actual slenderness ratio of its elements.
SA-24
TIl 5-1300/NAVFAC P-397jAFR88"22
!!~.- 1.7A~a
where Fa l~ defined in Section 5-37.3 . .'.
The value of Fa can be obtained by using either equation 5-43 or the tables in' the AlSC SpeCification which give allowable .s trre s ses for compression members. When using these tables. the, yield',. stress should, be taken equal to f ds' "'" Check i f Pu > P. ,
d.
Determine bracing: requirements:, • I'
Example 5A-5 (a) Required:
D~sign
Solution: Step 1.
Given:
-
•
Design of an Open-Web Steel Joist
a simply-supported open-web steel joist whose capacity is controlled by'flexure. '
..... '
(a)
\
.
I
a.
Pressure-time loading
b:
Clear span Spacing of joists Weight of decking
[figure 5A-5
1 50':0"~
7' _0" ,.' r
4 psf
Structurat steel propertfes
..c.
f;
"
Chords Web
il
,
",
'~a ~
"';
T
':.
50;000 ~si
Y
Dynamic -Lncrease ' factor' (chor ds oni y) . (Table 5-2,' for A588)
c - 1.19
,.
..
'," ..d>
Chords
f ds
·';~1.19
.
.. '
c x a x 'fy
f ds
Dynamic design stressj
,
x 1.1 x 50 ;'000 -r
(Equation 5-2) 65,450 psi
! -
Design 'criteria
(Section 5-35)
, Maximwn ductility ratio: Maximwn end rotation:
5A-25
,.
. "'max
4.0 t
2
0
~ •• , i
v.
TM 5-1300/NAVFAC P-397/AFll 88-22
~ ,.
.-
w.
6.76" Ii""--:-"'.:..o.--
-1--
a:
::::). II)
2.0psl
f3a:
1:1.
40ms
TIME
.N
N
V32LH11
N g
CO)
.-',
-f--_:_'---.-.-__
M[
..m.. o
~.
Figure SA-Sea) Step 2.
.
Joist cross-section and loading, Example SA-Sea)
Selection of joist size a.
Assume a dynamic load factor. For preliminary design, a DLF 1.0 is generally recommended. However, since the span is quite long in tbis'case, a QLF of 0.62 is selected.
=
b.
Equivalent static load on joist: .. .. v w1 -
c.
~ 14~.x
7.0
1,250 lb/ft
Service live load on joist: w2
d.
0.62 x 2.0
.
- w1/1.7 x 1.19x 1,! -.1,250/2,23 - 561 lb/ft
Using the "Standard Specifications, Load .Tab1es and Weight Tables" of the Steel Joist Institute, for a span of 50'-0 11 ,
try 32LH11. by flexure.
Joist tables show that capacity is controlled r .
~
;
Total load-carrying c':p~city (including dead load - 602 1b/ft). Approximate weight of joist and decking - 28 + (4 x 7) - 56 1b/ft 5A-26.
TH·,5-1300/NAVFAC P-397/AFR 88-22 '
capacity (excluding' dead load - 602 - 56
Total - 546
.
"
.,.
:
The following section properties refer to the selected joist 32L1l1L [Fl:gtire 5A-5. (a) }: . . . ' ., >, Top Chord: Two 3 x 3 x 5/16 angles A - 3.56 in 2 r x - 0.92 -Ln ,
r y - 1.54 in. Ix - ·3.02 in4
, .r.
Bottom' Chord: Two 3 x 2~1/2 x 1/4 angles
- 2.62 in2 r x - 0.945 in. , r 1. 28 in.
A
.,
• 'J
Ix - 2.35
::
...
<
y
in4
I xx for joist - 1;.383.0,··in4: . .Panel length - 51-.inches Resistance per·unit·length
Step 3.
l"
r u - 1.7 x 1.19 x 1.1 x 546 - 1215 Ib/ft Step 4. ••..
.. f
384·.1< 29 x 10 6 x'l,383
i.
J
(Section 5-33)
3 KE - 384 EI/5L
5(12 x 50)3
- 14,260 Ib/in (Table 3-8)
1,215 x 50 - 4.26 inches 14,260 Step 5.
Total mass of joist'plus decking 56 x 50 x 10 6
M-----386
Total effective mass Me - KLMM KLM - 0.5(0.78 + 0.66) - 0.72
5A-27 .
(Table 3-12)
TM 5-l300/NAVFACP-397/AFR 88-22.· Me-, 0.72 (7.25 x 10 6) - 5.22· x 10 6 lb-ms 2/in Natural period
T~
2n(M e / KE)1/2 -
2n(5.220.000/14:2~0)1/2.-
120.2 ms
Behavior controlled by flexure . . Use Step 6a. Step 6a.
a. T/T N - 40/120.2 - ·0.332 2.0 x 144 x 7 P/ru
----~-
-
- 1. 82·
1,105
From figure 3-64a.
'(
~.
O.K. b. Xm - 2.3 x 3.87 - 8.9 inches tan
e -
e -
Xm/(L/2) - 8.9/(25 x 12) - 0.0297
1.7 0
< 20
O.K.
Check selection of DIF.
c.
From Table 3- 64a. for /I. - 2.3 and T/T N - 0.33. tE/T - 0.55, Find
,
.
.'
t E - 0.55 x 40 - 22ms
~
~ - fds/Est E·- 65.45/30 x 10 3 x .022 - 0.099 in/in/sec .'.
.....
5A-28 :
(Equation 5-1)
.,
· TK S-1300/NAVFAC:P-397/AFR 88"22" From figure'S-2 (average of A36 and A514) DIF - 1.18 d.
'~
1:19 assumed
O.K.
Check top chord as a beam' column. Maximum moment at mid-span l,10S x (50)~ x 12 4,144 in-kips 8x 1",000 , Maximum axial load in chords
P - M/d d -
distance between centroids of top arid bottom chords [see figure 5A-5 (a)J 30.22 Lnche s
P - 4,144/30.22 - 137.1 kips 1
panel length
~
51 inches
Slenderness ratio, l/rx - 51/0.92 - 55.4 < Cc " where Cc -,(2n 2E/f ds)1/2, - 95 x (Equation 5-41) Fa - 23.5 ksi for f y - 50 ksi (Table 3-50, AISC Specification) 1.31 (23.5) - 30.8 ksi for f ds - 65,450 psi Pu - 1.7AFa - 1.7 x 3.56 x 30.8 - 186.4 kips (Equation 5-42) "
Considering the first panel as a fixed, simply supported beam, the maximum moment in'the panel is ,_. 1,105 (51)2 M -'r u L2/12 -.~~---~---~-- - 19.96 in-kips 12 x 12 x 1,000
.' . The effective slenderness ratio of the top chord in the first panel is
SA~29
TIl S-1300/NAVFACP-397/AFR 88-22
- 48.7 ksi Pex - (23/12)AF' ex' - 231.12 x. 3.
~6
x 48..7. - 333 kips
(1 - P/P ex) - (1.0 - 126 .. 6/3)3) - 0.62 .,
~
To determine ~, the plastic moment'~ is needed and the value of Zx has to be computed. The neutral axis for a fully plastic section is located. at a.distance x from the flange.
.
3x
-~
- (3
5/16) 5/16 + 3 (5/16 - x) ."
- (43) 5/(16 x 16) + 15/1$' ~3x x - 455/(6 x 256) - 0.296 inch . The plastic section ·modulus s , :Z", is found to be
..
'
(0.2 Zx - 2[ --------- x 3 296)2' ,
+
(3.0·~.0.3125)
(0.3125
0.296)2 + 2
.
·'0.263 + 0.0007.+ 2.285 - 2.549 in 3
~x
- fdsZ x - 65.45 x 2.549 - 166.8 in-kips
(Equation 5-8) (Equation 5-37)
where ry.is least radius of gyration - 0.92 - [1.07 - (55 ..4/391)J 166.8 - 154.8 in-kips
.
SA-30
,
TM S-1300/NAVFAC'P:'397/AFR'88-22 Cm - 0.85 . P/Pti + CmM/ [ ( l " .
(Section 1.6.1, AISC) P/Pex)~xJ
,.'
S
~Equation
1.10
137.1 + 0:85 (19.96) ", S 1.0 - 0.736 + 0.176 186.4 0.62 (154.8)
0.91~'~
5-44)
1.0 O.K.
"
Step 7.
Check bottom chord for rebound. a.
,
Calculate required resistance in rebound .. .
.,.
From figure 5-13, 100% rebound r/r u - 1.0 r b.
•
- r u - 1,105 lb/ft
Moment and axial forces in rebound M-
(r x L2)/8 -
4,144 in-kips
Maximum axial force in bottom 'chord
r
P - Mid - 137.1 kips (compression) c.
Ultimate axial load capacity Stability in vertical direction (about x-axis) 1
51 inches
r x - 0.945
l/r - 51/0.945 - 54.0 < Cc where Cc - [(27T2E)/fdsJ.1/2 -
95
Fa - 23.72 ksi for f y - 50 ksi (Table 3- 50, 'AISC Specification) 1.31(23.72) - 31.1 ksi 'for'f ds - 65,450 psi Pu - 1.7AFa - 1.7 x 2.62·x 31.1,- 138.5 'kips d.
Check bracing requirements.' '.
SA-31
TK.5-l300/NAVFAC P-39J/AFR 88-22 P
~
137:1
<
Pu - 138.5
O.K.
Adding-a vertical member between panel_joints of bottom chord would have been required nad P > Pu' This additional bracing would have been needed in mid-span but may be spared at the joist ends. Stability in the lateral direction (about y-axis) Pu - 137.1 kips
A - 2.62 in 2
r y - 1.28 inches,
Fa - Pu / (1 . 7 x 1.31A) - 137.1/(1.7 x-l.31-x 2.62) ~'23.5 ksi For a given Fa - 23.5, the corresponding slenderness ratio is l/r := 55
(Table 3-5'0, AISC Specification)
Therefore, the maximum unbraced length in mid-span is Ib - 55 x 1.28 - 70.4 inches s,
Use lateral bracing at panel points, i.e., 51 inches at midspan. The unbraced length may be increased at joist ends, but not greater than specified for bridging requirements in the joist specification. . Example 5A-5 (b) Required: Solution: Step 1.
Analysis of Existing Open-Web Steel Joist
Analyze'a simply supp~rted, open-web steel joist whose capacity is controlled by shear. Given: a.
Pressure-time loading [figure 5A-5 (b)] ,Joist 22Hll.
b.
Clear span - 32'-0" Spacing of joists ,- 6' -,0". Weight of decking - 4 psf
5A-32
TM 5-l300/NAVFAC P-397/AFR 88.-22
ttc:i
6.35"
I"
"
'.
•
I
~... lot
I/)
w
".
le' 1.0 ~ pal
!'
.. I,l
'
III III
in Go N
w
le
22H11
N ~ N
Q,
,
N
25ms
.,
TIME
~-- .~
-L-....I--' -
bI/) c:i Figure 5A-5(b)
'l
~:}
,
Joist cross-section and loading, Example 5A-5(b) r. '
c.
Properties of structural steel:
Chords
50,000 psi
Web
36,000 psi
Dynamic increase factor (chords only)
(Tabl~ '5 - 2 for A588)
c - 1.19 • J.
"
Dynamic design stress, f ds Chords d.
-c x a
1.0 1
-
65,450 psi
(Section 5-35)
For members controlled by ·shear:.
Step 2.
.r-
f ds - 1.19 'x 1.1' x 50,000
Design criteria
amax
x fy .
,
a.
Assume the DLF
b.
Overpressure load on joist
1. 25
5A-33
'
TK 5-1300/NAVPACP-397/AP& 88-22
wl - 1.25 x 1.0 x 144 x 6 - 1,080 lb/ft c.
Equivalent service load w2 - wl/l.7 x a x c - 1,080/2.23 - 485 lb/ft
d.
From the ·Standard Specifications and Load Tables" of the Steel Joist Institute: , ., Total load-carrying capacity (including dead load) - 506 lb/ft. Approximate weight of joist plus decking ·17 + (6 x 4) - 41 lb/ft Total load-carrying capacity (excluding dead load) 506' - 41 - 465 From the steel joist catalog', the following are the section properties of Joist 22HII [figure 5A-5 (b)]: Panel length - 24 inches Top Chord: A - 1. 935 in 2 4 Ix - 0.455 in r x - 0.485 in. r y - 1.701 in. Bottom Chord: A - .1. 575 in2 Ix
0.388 in4'
r x - 0.497 in. r y - 1.469 in. I xx for joist - 396.0 in4
•
TM 5-1300/NAVFAC P-397/AFR 88-22 Step 3.
Resistance per unit length r u - 2.23 x 465 - 1,035 1b/ft 384 x 29 x 10 6 x 396
384EI
Step 4.
- 15,580 Ib/in
(Table 3-8)
5 x (12 x 32)3 1,035 x 32 2.13 inches 15,580 Step 5.
Mass of joist plus decking 6
41 x 32 x 10 6 M-~----
386 Effective mass Me - KLMM - 0.78 x 3.4 x 106 - 2.65 x 10 6 1b-ms 2in Natural period TN - 2rr(Me/KE)1/2 - 2rr(2,650,000/15,580)1/2 - 81.8 ms Behavior controlled by shear.
Use Step 6b of the procedure.
Step 6b. a.
T/T N - 25/81.8 - 0.305 6 x 144 x 1.0 - - - - - - - 0.835 1,035 From Figure 3-64a of Chapter 3: ~
-
tan
~/XE
e -
< 1.0; elastic, O.K.
~/(L/2)
- 2.13/(16.0 x 12) - 0.0111
ec.
0.64° < 1°
O.K.
Check selection of DIF from figure 3-49 of Chapter 3, for T/T N - 0.305, tm/T - 1.12; t m - 1.12 x 25 - 28 ms
tH 5-1300!NAVFAC P-397/AFR 88-22
Find €
€
- fds/Est e
(t E -
~)
(Equation 5-1)
- 65.45/30 x 10 3 x 0.028 - 0.078 in/in/sec From figure 5-2, (average of A36 and A514) DIF - 1.18 - 1.19 assumed, O.K. d. Step 7.
Check of top chord as a beam-column is not ne<;:essary.
Check bottom chord in rebound. ~
a.
For
- 1 and T/T
- 0.305, rebound is 100%
b.
Determine axial load in bottom chord, P - M/d. ~
For an elastic response, DLF - 0.87
(Figure 5-13)
< 1.0, where T/T N - 0.305, the (Figure 3-49)
Equivalent static load,w w - DLF x b x P - 0.87 x 12 x 12 x 6 x 1.0 - 751 1b/ft Maximum moment in rebound, M - wL2/8 M - [751 x (32)2/ 8 J 12
-
1,155,000 in-1b
P - M/d - 1,155,000/21.28 - 54,300 1b - 54.3 kips c.
Check bracing requirements. (1)
Vertical bracing of bottom chord: Panel length - 24 inches r x - 0.497,
r y - 1 . 469 ,
l/rx - 24/0.497 - 48.3
5A-36
A - 1.575 in 2
TN 5-1300/NAVFAC P-397/AFR 88-22 Allowable P - 1.7 x a x c x A x Fa - 1.7 x 1.1 x 1.19 x 1.575 x 24.6 - 86.2 kips> 54.3 kips (Table 3-50, AlSC Specification) No extra bracing required. (2)
Lateral bracing of bottom chord: P - 54.3 kips, A - 1.575 in 2 Fa - P/l.7A - 54.3/(1.7 x 1.575 x 1.1 x 1.19) - 15.5 ksi For f y - 50 ksi and Fa - 15.5 ksi l/r - 96 1 - 96 x 1.469 - 141 inches (Table 3-50, AlSC Specification)
Therefore, use lateral bracing at every fifth panel point close to mid-span. The unbraced length may be increased at joist ends, but not greater than specified for bridging requirements in the joist specification.
Problem 5A-6
Problem: Procedure: Step 1. Step
2.
Design of Single-Story Rigid Frames for Pressure-Time Loading
Design a single-story, multi-bay rigid frame subjected to a pressure-time loading. Establish the ratio a between the design values of the horizontal and vertical blast loads. Using the recommended dynamic load factors presented in Section 541.3 establish the magnitude of the equivalent static load w for:
5A-37
TM 5-l300/NAVFAC P-397/AFR 88-22
Step 3.
a.
local mechanisms of the roof and blastward column
b.
panel or combined mechanisms for the frame as a whole.
Using the general expressions for the possible collapse mechanisms from Table 5-13 and the loads from Step 2, assume values of the moment capacity ratios C and Cl and proceed to establish the required design plastic moment ~ considering all possible mechanisms.
In order to obtain a reasonably economical design, it
is desirable to select C and Cl so that the least resistance (or the r~quired value of ~) corresponds to a combined mechanism. This will normally requlre several trials with assumed values of C and Cl. Step 4.
Calculate the axial loads and shears in all members using the approximate method of Section 5-41.4.
Step 5.
Design each member as a beam-column using the ultimate strength design criteria of Sections 5-37.3, 38, and 39. A numerical example is presented in Section 5A-4.
Step 6.
Using the moments of inertia from Step 5, calculate the sidesway natural period using Table 5-14 and Equations 5-50 and 5-51. Enter the response charts in Chapter 3 with the ratios of T/TN and P/~' In this case, P/~ is the reciprocal of the panel or sidesway mechanism dynamic load factor used in the trial design. Multiply the ductility ratio by the elastic deflection given by equation 5-53 and establish the peak deflection ~ from equation 5-54. Compare the maximum sidesway deflection ~ with the criteria of Section 5-35. Note that the sidesway deflection 0 in Table 5-8 is ~'
Step 7.
Repeat the procedure of Step 6 for the local mechanisms of the roof and blastward column. The stiffness and natural period may be obtained from Table 3-8 of Chapter 3 and equation 5-15, respectively. The resistance of the roof girder and the blastward column may be obtained from Table 5-13 using the values of ~ and C~ determined in Step 3. Compare the ductility ratio and rotation with the criteria of Section 5-35.
Step 8.
a.
If the deflection criteria for both sidesway and beam mechanisms are satisfied, then the member sizes from Step 5 constitute the results of this preliminary design. These members would then be used in a more rigorous dynamic frame
analysis.
Several computer programs are available through
the repositories listed in Section
5A-38
5~4.
TN 5-1300/NAVPAC P-397iAPR 88~22 b.
If the deflection criterion for a sidesway mechanism is exceeded, then the resistance of all or most of the members should be increased.
c.
If the deflection criterion for a beam mechanism of the front wall or roof girder is exceeded, then the resistance of the member in.question should be increased. The member sizes to be used in a final analysis should be the greater of those determined from Steps 8b and 8c.
Example 5A-6 Required:
Design of a Rigid Frame for Pressure-Time Loading
Design a four-bay, single-story, reusable, pinned-base rigid frame subjected to a pressure-time loading in its plane.
Given: a.
Pressure-time
b.
Design criteria: It is required to design the frame structure for more than one incident. The deformation limits shall be half that permitted for personnel protection, that is:
load~ng (~igure
5A-6)
& - H/50 and 9 max - 1· for individual members
Step 1.
c.
Structural configuration (figure 5A-6)
d.
A36 steel
e.
Roof purlins spanning perpendicular to frame (b v- b n, Figure 5-26)
f.
Frame spacing, b - 17 ft
g.
Uniform dead load of deck, excluding frame
.
'.
. Determine a:
. (Section 5-41.1)
b h - b v - 17 ft
qh - 5.8 x 17 x 12 - 1,183 lb/in qv - 2.5 x 17 x 12 - 510 lb/in a - qh/qv - 2.32
5A-39
TM 5-l300/NAVFAC P-397/AfR 88-22
I
..
'
.
, .
, 4 @ 18'-8" = 88'-0"
Root Dead Load = 13.5 pat
3 '
2
. Wall Dead Load = 18.5 pat
MEMBER REFERENCE NUMBER
~
5.8 psi
w
a:
,
::;)
~a:
III 2.5
fa
a: D.
D.
psi
78ms
78ma
TIME
TIME
BLASTWARD EXTERIOR WALL Figure SA-6.
ROOF (Average)
Preliminary design' of four-bay, single-story rigid frame, Example 5A-6
5A-40
TM 5-l300I.NAVfAC P-397/AFR ,88-22
Step' 2.
Establish equivalent a.
sta~ic
loads
(Section 5-41.3)
Local beam mechanism, w - DLF x qv. 1.25 x 510 x 12 7.65 k/ft
w-----~--
1,000 b.
Panel or combined mechanism, w - DLFx qh 0.625 x 510 x 12
w--------
3.83 k/ft
1,000
Step 3.
The required plastic moment capacities for the frame members are determined from Table 5-13 based upon rational assumptions for the moment capacity ratios Cl and C. In general, the recommended starting values are Cl equal to 2 and C greater than 2. From Table 5-13. for n - 4, a - 2.32, H - 15.167 ft, L - 16.5 ft and pinned bases, values of Cl and C were substituted and after a few trials, the. following solution is obtained: ~
- 130 kip-ft,
Cl - 2.0
and C - 3.5.
The various collapse mechanisms and the associated values of are listed below: Collapse Mechanism 1 2 3a, 3b 4 Sa, 5b 6
w
(k~t)
( k/f t )
_ 1'1
7.65 7.65 3.83 3.83 3.83 3.83
130 128 128 129 110 116
The plastic design moments for the frame members as follows: ,, Girder,' Mp - 130 k-ft , " 260 k- ft Interior column, Cl~ f •. ' Exterior column,
~
- 455 k-ft
..
5A-4l
~
ar~established
,.
TH 5-l300/NAVFAC P-397/AFR 88-22 Step 4.
a.-
Axial loads and shears due to horizontal blast pressure. w - 3.83 k/ft From figure 5-27, R - awH - 2.32 x 3.83 x 15.167 - 135 kips 1.
Member 1, axial load PI - R/2 - 67 kips'
2.
, .:7
Member 2, shear force V2 -R/2(4) - 135/8 - 16.8 kips
3. -
.,
Member 3, shear force V3 - R/2 - 67 kips
b.
Axial loads and shears due to vertical blast pressure, w - 7.65 k/ft 1.
Member 1, shear force VI - w x L/2 - 7.65 x 16.5/2 - 63.1 kips.
2.
Member 2, axial load P2 - w x L - 7.65 x 16.5 - 126.2 kips
3.
Member 3 ,axial load P3 - w x L/2 - 63;1 kips The dead loads are small compared to the blast loads and are neglected in this step.
Step 5.
The members are designed using the criteria of Sections 5-37.3, 538, and 5-39 with the following ~esults:
~
P
V
Member
(k-ft)
ill
ill
Use
(in4)
1
130
67.0
63.1
w12x35
285
2
260
126.2
16.8
w14x6l
640
3
455
63.1
67.0
w14x74
796
5A-42
Ix
TM 5-1300/NAVFAC P-391/AFR
8~-22
Determine the frame stiffness and sway deflection.
Step 6.
>.• . (3 x 640) + (2 x 796) lea - - - - - - - - - - - 702 in4
(Table 5-14)
5 '
19
~ 285 in4
:"
.......' r
B -
°
•. e '.
:',\ .,
285/16.8
I /L g
-~-~-~-- -,0.498
0.75I ca/H
(Table 5-14)
(0.75) d62;15.167)
. C2 - 4.65 ',..\.
t,
~-
El ca C2 H3
'
, (Table 5-14)
[1 + (0.7 -O.lB) (n'l)]
(30) (10)3 (702) (4.65) [1 + 0.7(3)] - 50.2 k/in KL,- 0.55 (1 - 0.25B) - 0,55
(Equation 5-51)
."
.'. '
Calculate dead weight, W:
.,
,
W - b[(4Lwdr) + (1/3) (Hwdw)] + (35 x 66) . + 1/3 (15.167) [(3 x 61) + (2 x 74)] - 20,548 1b
me - Wig - 20,548/32.2
~
638 1b-sec 2/ft - 638 x 10 6 1b-ms 2/ft
TN - 2n[me/KKL]1/2 (Equation 5-50) - 2rr[,(638'x 1-9 6 ) / ( 50 . 2 x_12 ~.103 x 0.55»)1/2 - 276 ms T/T N - 78/276 - 0.283 '
5A-43
..
TN 5-l300/NAVFAC P-397/AFR 88-22
,P/Ru - 1.6 ~
- Xm/XE -
1.40
(Figure 3-64a)
2.32 x 3.83 x 15.167 XE - R,,/KE - awH/KE - - - - - - - - - - 2.68 inches 50.2 (Equation 5-52 and 5-53)
Xm-
5 - 1.40 x 2.68 - 3.75 inches .;
,
,
,
(Equation 5-54)
5 - 3.75/(15.167) (12)H,- 0.0206H - H/48.5 Step 7.
Check deflection of possible local mechanisms. a.
Roof girder mechanism (investigate W12 x 35 from,Step 5) (Equation 5-15)
For an elasto-plastic response, take the average-load-mass factor for the plastic and,elastic response, or:, KLM - (0.77 - 0.66)/2 - 0.715
(Table 3-12)
. . m - 0.715 x Wig
w - (13.5 x 17) + 36 - 265 lb/ft W - w x L - 265 X 16.5 - 4372 lb. mE - 0.715 x 4372/368 ~ 8.1 1b-sec 2/in, KE - 307 EI/L 3
(Table 3-8)
307 (30) (10 6) (285) .' . K
- ---------'---
332,000 1b/in. '
(16.5 x 12)3 TN - 2" (8.1/332,000)1/2 x 1,000 - 31.0 ms
5A-44 .
TH 5-1300/NAVF"C,P-397/AFR 88-22 T/T N - 78/31.0 - 2.52 Ru - 16~/L - (16 x 130)/16,5,- 126 kips
(Table 5-13)
P - pbL - (2.5) (17) (144) (16.5)/1,000 - 101 kips P/Ru - 101/126.-'0.80 ~
-
(Figure 3-64a)
Xm/XE - 1.80 :; ; 1.
..• .s
Check end rotation of girder. XE - Ru/KE - 126/332 - 0.380 inch Xm - 1.80 x 0.380 - 0.69 inch Xm/(L/2) - 0.69/(8.25) (12) - 0.069-tim
e b.
-0.40·' < 1·
..
O.K.
e.
Exterior column mechanism (investigate W14 x 74 from Step 5). \(Equation 5 -15)
TN - 21r (me/KE) 1/2 ".' 0.78 "
+' 0.66
w
2
s .,:
.(Tab1e 3-12)
- 0.72 wig
w - (16.5 x 17) + 74 - 354 1b/ft W - 354 x 15.167 - 5369 1b Me - 0.72 (5369)/386'- 10.0 1b-sec 2/in. 'KE - 160 EI/L4
(Table 3-8)
(160) (30) (10) (796) ", (1.82)3
5A-45
632,000 1b/in
-.
TM 5-l300/NAVFAC P-397/AFR 88-22 TN - 2ff(lO.0/632.000)1/2 x .1000 -.25:0·ms' T/TN -,78/25.0 -3.12 4~(2C
. .4 (130) [.(2 x 3.5) + I']
+1)
R,,-
275 kips H
-' -
15.167'
.', . (Table 5 -13)
-',".. r" -•
. ~J
",
P - (2.32) (7.65/1.25) (15.167) - 215 kips ?
•
,.'
.'l
.'~
.
P/R" - 215/275 - 0.78.. ~
- x.,/X , E-
1.80
(Figure 3-64a)
. '~ .
Check end rotation of columns. ,-
,
~E
.
e,
- R,,/K - 275/632 - 0.435 inch -\
x., -
'.';"<
~tiL;o!
1.80 x 0.435 - 0.78 inch
• :_""~/(L/2) - 0.78/(7.58) (12) "'0'.0086 .. tan
Step 8. Summary:
.
. ,.'
e
0.49'" <·'!l'
e
O.K.
.
.
)':
. The deflections of the local mechanisms are within the criteria. The sidesway defle~tion is acceptable. The member sizes to be used in a computer analysis are as, follows: . . Member ~
t;
1
•
Size ,W12 x 35
.. '-
.'
Problem 5A-7 Problem:
2
W14 x 61
3
W14 x 74 -
.,
.\
, '.
Design of Doors for Pressure-Time Loading
Design a steel-plate blast door subjected to a pressure-time loading.
5A-46
TH 5-1300/NAVFACP-397/AFR 88-22 Procedure: Step 1.
Establish the design parameters a.
Pressure-time load
b.
Design criteria: Establish support rotation, em~x' and whether seals and rebound mechanisms are ,req~ired.
c.
Structural configuration of the door including geometry and support conditions
d. '
Properties' of steel used: Minimum yield strength, f y' for door components (Table 5-1) Dynamic incte~se'factor, c(Table 5-2)
Step 2.
Select the thickness of the plate.
Step 3.
Calculate the elastic section modulus, S, and the plastic section modulus, Z, of the plate.
Step 4.
Calculate the design plastic moment, ~, of the plate (equation 5-7) .
Step 5.
Compute the ultimate dynamic shear, Vp (equation 5'·16).
Step 6.
Calculate maximum support shear, V, using a dynamic load factor of 1:'25 and determine VfVp' If VfVp. is Le'ss 'than 0.67, use the plastic design' moment as computea in Step 4 (Section 5-31). If VfVp is greater than 0.67, use Equation 5-23 to calculate the' effective ~.
Step 7.
Calculate the ultimate unit resistance of the section'(Table 3-1), using the equivalent plastic moment as obtained in ~tep 4 and a dynamic load. factor of 1.25.
Step 8.
Determine"the moment of inertia of the plate s ec t.Lon , 'J
.••
Step 9.
Compute the equivalent elastic unit stiffness, KE, of the plate section (Table 3-'8). "
Step 10.
Calculate the equivalent elastic deflection, XE; of the plate as given by XE - ru/KE' '
Step 11.
Determine the load-mass factor KLM and ,compute the effective unit mass, me'
5A-47
'j
TK
S-130q/NAVFAC,P~397/AF~
Step 12. Step 13.
88-22 .
Compute the natural period of vibration, TN'
.,
Determine the door plate response using the ,values of P/ru and Determine Xm/XE and
T/T N and the response charts of Chapter 3. TE· Step 14.
. Determine the 'support rotation, , Tan S - (Xm) I (LIZ) Comp are-B with the design
Step i
s,
crite:ri'ii;~f'step
lb.
Determine the strain rate, E, using"e'quation'S"l. Determine the dynamic increase factor using figure S-2 and compare with the DIF selected in Step ld. If the criteria of Step 1 is with a new plate thickness.
n~t
'I
satisfied, repeat Steps 2 to lS '
Step 16.
Design supporting flexural element considering composite action with the plate (if so constructed).
Step 17.
Calculate elastic and plastic section moduli of the co~bined section.
'v
Step 18.
Follow the design procedure for a flexural element as described in Section S A - l . " ..
Example'SA-7 (a)
Design of'~ Blast Door for Pressur~-T1me ,Loading
..
'
Required:
Design a do~ble-lea~, built-up door (6 ft by 8 ft) for the· given pressure-time loading.
Step l. _
Giver: a.
'Pressure- time loading (Figure SA-7)
b.
Design criteria: ,This door is to protect personnel from Leakage into the structure is permitted but the maximum end rotation of any member is limited to 2' since panic hardwar~ 'must'be operable after an accidental explosion. '
'e~terior loading.
c.
Structural configuration (Figure SA-i)
NOTE:. " •. This type of door configuration is suitable 'for' low-pressure range applications. d.
Steel used:
A36
SA-48
TIl 5-1300/NAVFAC,P-397/AFR88-22 --DETAIL, ' ,
,
SECTION;-,, -':' ,:", ,"
DETAIL
DOUBLE LEAF BL,AST DOOR
- '; l
14.8 psi
§ /I)
DOOR HEIGHT = 8'-0'
./1)
w
a::
lL
13ms
TIME
Figure 5A-7(a)
Door configuration and loading, Example 5A-7(a)
Yield strength, f y - 36 ksi ,
(Table 5-1)
Dynamic increase factor, c - 1.29
(Table 5-2)
Average yield strength increase factor, a - 1.1 (Se c t LonvS-12.1)
.
Hence, the dynamic design stress, "
f ds - 1.1 x 1.29 x 36
~
51.1 ksi
(Equation 5-2)
and the dynamic yield stress in shear, f dv - 0,55 f ds - 0.55 x 51.1 - 28.1 ksi
(Equat'ion 5-4)
Step 2.
Assume a plate ,thickness: ot'5/8 inch.
Step 3.
Determine the elastic and plastic section moduli (per unit width).
5A-49 '
,
'
TM 5-1300/NAVFAC
P~397/AFa
bd2
S
--6bd 2
88-22 1
(5/8)2
X
6
1 x (5/8)2
Z---4
Step 4.
4
Calculate the design plastic moment,
~:
~ - f ds (S + Z)/2 - 51.1 [(6.515 x "10- 2)
(Equation 5-7)
51.1 x8.14 x 10- 2 - 4.16 in:k/in Step 5.
Calculate the dynamic ultimate shear capacity, Vp ' for a I-inch width.
Vp - fdvAw Step 6.
28.1 x 1 x 5/8 - 17.56 kips/in
(Equation 5-16)
Evaluate the support shear and check the plate capacity. Assume DLF - 1.25
v-
1.25 x 14.8 x 36 x 1 DLF x P x L/2 - - - - - - - - - - - - 333 Ib/in - 0.333 kip/in 2
V/Vp - 0.333/13.61 - 0.0245 < 0.67
(Section 5-31)
No reduction in equivalent plastic moment is necessary. NOTE: When actual DLF is determined, reconsioer Step 6. Step 7.
Calculate the ultimate unit resistance, r u' (assuming the plate to be simply-supported at both ends). 8 x 4.16 x 10 3 25.7 psi
Step 8.
Compute the moment of inertia, I, for a I-inch width. bd 3
1 x (5/8)3
1- - - -
12
12
SA-50
(Table 3-1)
TIl 5-1300/NAVFAC P.-397/AFR'Ssc22
Step 9.
Calculate the equivalent elastic stiffness, KE.
, 384 x 29 x 106 x .02035
KE - 384EI/5bL4 -
27.0 psi/in (Table 3-8)
5 x 1 x (36)4,
Determine the equivalent elastic deflection, XE .
Step 10.
0.95: inch Calculate the effective mass of element.
Step 11.
a.
KLM (average.e1astic and plastic) - (0.78 + 0.66)/2 - 0.72
b.
Unit mass of element, m '",_ 5/8 x 'l'x "1 x .490 x 10 6. m - wig - 458.0 psi-ms 2/in ... .:. ,1,728 x 32:2 x 12,
'.
1" ,
c.
Effective unit of mass of element, me '"
'I, '
- KLMm - 0.72 x 458.0
.:
330 psi-ms 2/in Step 12.
Calculate the natural period of vibration, TN' t '
Step 13.
.,
-,
Determine the door response'. ' Peak overpressure', "! -\
•.
P - 14.8 'psi .. ~
..
Peak re s t s cancev '..:.'
Duration
.1
.... ..
•
T
-13.0ms
'
.
22:ms
Natural period of vibration
.
.,
P/r u - 14.8/25.7 - 0.58 13.0/22.0:~'0~59
..'
From Figure' 3:64a o f Chapter' 3, .
.''::'
':,/
...
5A-51~·,.
,",
.• '.f
TN 5-1300/NAVFAC P-397/AFR 88-22.
Xm/XE < 1 Since the response is elastic, determine theDLF from Figure 3-49 of Chapter 3. ,.,
DLF - 1.3 for T/T N - 0.59 Step 14.
.,
Determine the.support rotation. 1. 3 x 14. 8 x O.95 Xu. - - - - - - - - - - 0.713 inch, '" 25.7 tan
Step 15.
e - Xu./(L/2) e - 2.27' > 2'
0.713/(36/2) - 0.0396. ,
N.C.
Evaluate the selection of-the dynamic increase factor. Since this is an elastic response, use ,figure 3 -49(b) of Chapter 3 to determine t m. For T/T N - 0.59, tm/T - 0.7 and t m - 9.1 ms. The strain rate is: -, (Equation 5 -1) Since the response is elastic, 0:713
51.1
f ds - 51.1 x
x---
38.4 ksi
0.95
0.0091 sec.
Hence,
..
" - 38.4/29.6 ,x 10 3 x 0.0091 - .142 in/in/sec DIF - 1.31. 1.29 is acceptable.
Usingfigure\5~2,
.'
The preliminarY'selection of DIF -
Since the rotation criterion is not satisfied, change the thickness of the plate and repeat the. procedure. ~epeating these calculations, it can be shown that a 3/4-inch plate satisfies the requirements. Step 16.
Design of the supporting flexural element.
- ,
Assume an angle L4 x 3 x 1/2 and·attached to the plate as shown in Figure 5A-7(b) . Determine the effective width of plate which acts in conjunction with the angle
5A-52 - .. :::
TK 5-1300/NAVFAC~PC397/AFR 88-22
(Section 5-24)
... ~
where b f / 2 is the half width of the outstanding ,flange or overhang and tf is the'thickness of the, plate. With tf ... 3/4'.inch, ,bf/2 5 8.5.x 3/4"
i. e . , 6.38 inches
Use overhang of 6 inches:.'"", , .' Hence, the effective width - 6 + 2 - 8 ihches. with plate is shoWn in F1:gure' 5A-7(b)'.
The:angle together
,
Step 17.
Calculate the elastic and plastic .section '\noduli of the combined section.
..
,.' Let Y be. the distance of. c. g. of'the comb Ined-esec't fon from the outside edge of the plate as shown in Figure 5A-7(br.. therefore (8
x
3/10 x 3/8) + .(4,+ 3/4 -,1.33) x 3.25
"
.
1.445,cinches (8 x 3/4) + 3.25
-, Let yp be the distance to the N.A. of the full plasticity. . •. 1 yp - - -
.
section for
<
"
[(8 x 3/4) + 3.25J - 0.578 inch
'.
8 x 2 ..
,
"
8 x (3/4)3 I
combi~ed
,
.
+ 8 x 3/4 x (1.45 - 3/8)2
-
.'
, ','
12 ~.:t
1.33 _ '1.4:5)2 _
, , + 5.05 + 3.25 (4 + 3/4 Hence, Smin - 24.881/(4:·75
-~t881 i~4 \
1:445) - 7.54 in 3
Z - 8 (0.578)2;2+.8 (OJ5 + 3.25·(4.75 - 1.33 - 0.578)
SA-53, .
10.69 in 3
. ,.
,i
.'
TM 5-l300/NAVFAC P-397/AFR ,88-22',
)
,,'w Figure 5A-7(b)
:1.
I . ' ~,
• '}
'J':' '.' •
,
I.
j'
,t ',. '
,
'
Detarl' of composite angle/plate.supporting.eiement; Example SA-7(b) ..'
Step 18,
.
"'J
~
;,'
I,
•.
Follow the design procedure for the'composite element:~sing steps 4 .t.hrough 13, - Calculate the-des rgn- plastic moment' ~ of the supporting flexural element, . >" r - ' ~
- 51,1 (7,54 + .
~
10,6~)/2
- 465,8 in-kips
~."
"11
(Equation 5- 7)
~"
Calculate the ultimate dynamic shear capacity,;Vp;' Vp - fdA - 28,1 (4,0 - 1/2) 1/2 -49.2 kips
(Equation 5-16)
,'" ...
Calculate support shear and check shear .capac Lty . " ' L - 8'-0" - 96 inches t
• ".
Vp-, (14.8 x 36/2 x 96)/2 -
I'
•
1:
.:.
12,7901b
12.79 kips < V O.K. ( (Section 5-23)
~
Calculate the ultimate .ul1it resistance, r u . ..:~.
Assuming the angle to be simply supported at .,b'!th ends:' • c" .
5A-54,.,',
.
-
"
TK 5-1300/NAVFAC P-397/AFR 88-22 r u - 8~/L2 - (8 x 465.8 x 1,000)/(96),2 - 405 1b/in Calculate the unit elastic stiffness, KE . 384 x 29 x 10 6 x 24.881 4 K E - 384EI/5L - - - - - - - , . . . . - - - - - 5 x (96)4
(Table 3-1) '.
,652.5"'lb/in 2 (Table 3-8)
Determine the equivalent elastic deflection; XE .
XE - ru/KE - 405/652.5 - 0.620 inch Calculate the effective mass of the element. KIH
-
0.72
490 11.1 3 w - --- + -- x 18 x - - - (0.925 + 3.825) - 4.750 1b/in 12 1,728 4 Effective unit mass of element, 4.75 x 10 6 me - 0.72 x
- 0.89 x 104 1b-ms 2/in2
32.2 x 12 Calculate the natural period of vibration, TN' TN - 2n[(89 x 102)/652.5]1/2 -23.2 ms (Figure 3-64a)
Determine the response parameters,"
p'- 14.8 x 36/2'~ 266.5 1b/in
Peak overpressure Peak resistance
I.
r u - 405'lb/in
T -13.0 ma
Duration Natural period of vibration,
TN - 23.2 ma
P/ru - 266.5/405 - 0.658 ~' .T/T
N~
13/23: 2 - 0.56
,
5A-55
.'
TM 5-l300/NAVFAC P-397/AFR 88-22
,
From Figure 3-64a,
. From Figure 3-49 for. T/TN. - 0.56,. DLF - 1.28 1.28 x 14.8.x 36/2 Hence,
Xm -
-
0.522 'in
652.5 tan
ee -
Xm/(L/2) - 0.522/48 - 0.0109 0.69°
< 2°
O.K.
Check stresses at the connecting point.
er - My/I - 355 x 103' x (1.445 - 0,75)/24.881 - 9,900 ps i (M
x X
~
-
E
0.522, ------ x 405 - 341) 0.62 ,
. l
'I" -
12.79 x 103 x 8 x 3/4 x (1.445 - 0.75/2) VQ/lb - -------------.,....----------.....,.,,---------.,....------., 24.881 x 1/2
5,}2l psi
..
Effective stress at the section 7" «(72+ .. '1"2 )1/2
•
~
.!
~ 103· x (9.9 2 + 5.321 2 )1/2 _ 11,'239 p~i < 39.600 ~s~ . .O.K.
Example 5A-7 (b)
Design of a,Plate Blast Door for Pressure-Time Loading
Required:
Design a single-leaf,door. (4 ft by 7 ft) for the given 4 p r e s s u r e time loading.
Step 1.
Given: a. Pressure-time loading [figure·5A:7. (c)]
"...
b.
"
Design criteria: Door shall be designed to contain blast pressures from an internal explosion. Gasket and 'reversal
SA-56
TM 5-1300/NAVFAC p C 3 9 7/AFR88-22 '
;.>,
.'
1.5 Sym~'.abt. 't. ,."
••
','
.
•
- .. 'rI-....Lt ';
.-.
---,,
,
.,..
___ I
" ...
4'-0'
1 Bearing PLAN / SECTION
Legend: ~
,...'
- A B C D E F G
~
-
Steel -frame- embedded in concrete Steel sub-frame r Reversal bolt, housing Reversal' bolt ' Blast door plate Continuous gasket Continuous bearing block
W
/-
a:
:::J III 1100,Ill 'psi
I
~')
:
,<
' c
w
a:
ll. , .
"
I'
. "
100 psi I
TIME
1,Oms
Figure 5A-7(c)Door configuration and loading, Example 5A-7(b) SA-57
TK S-1300/NAVFAC P-397/AFR
88~22:
mechanisms shall limi ted to 3
be provided.
Support rotation shall be
0
Str~ctural configuration [see figure 5A·7 (c)l
c.
Note: This type of door is suitable for high pressure range applications. d.
Steel used:
ASTM A588
.Yield strength, f - 50 ksi·· Dynamic increase ~ctor,' c - 1.24
(Table ,5·1) (preliminary, Table 5·2)
Average strength increase factor, a -
1.~
.
(Section 5-12.1)
Hence, the dynamic design stress, f ds - 1.,1 x 1. 24 x 50 - '68.2 ksi
(Equation 5,2)
Note: It is assumed, for the limited design rotation of 3 'that < 10, and, therefore, that equation 5-3 does not govern . 0
~
,
.
The dynamic design stress in shear f dv - 0.55 f ds - 37.5 ksi
(Equation 5-4)
Step 2."
Assume a plate thickness of 2 inches, .
Step 3.
Determine the elastic and plastic section moduli (per unit width) bd 2
S
---
1 (2)2
6
0.667 in 3iiri
6 ,
bd 2
1 (2)2
4
4
Z -
Step 4.
"
,.'
3 .1.0 in lin d '
Calculate the design plastic moment, ~ - f ds (S +
Step 5.
,; t
~
Z)/2.- 68.2 (0.667.+ 1.0)/2 - 56.8 in-k/in (Equation 5-7)
Calculate the dynamic ultimate shear capacity, Vp' for a 1-inch width.
SA-58
TH 5-l300/NAVFAC P-397/AFR 88-22 75.0 kips/in Step 6.
(Equation 5-16)
Evaluate the support shear and check the plate shear capacity.
Example 5A-7(b) Assume DLF - 1.0
(Table 5-4)
For simplicity, assume the plate is a one-way member, hence:
v - DLF
x P x L/2 - 1.0 x 1100 x 48/2 - 26,400 Ib/in
- 26.4 kips/in 26.4/75. - 0.352 < 0.67
V/Vp -
(Section 5-31)
No reduction in equivalent plastic moment is necessary. Step 7.
Calculate the ultimate unit resistance, r u ' For a plate, simply-supported on four sides (direct load) (Table 3-2) where MHP -
~
and MHN - 0 and
X
L 0.35
for ---
L thus,
(Figure 3-17)
- 329 psi
Compute the moment of inertia, I, for a I-inch width bd 3 I - ----12
Step 9.
1.75
X - 0.35 x 12 x 7 - 29.4 in
r u - 5 x 56.8/(29.4)2 Step 8.
-
H
1 x 23 -
- 0.667 in4/in
12
Calculate the equivalent elastic stiffness, KE (Figure 3-36)
5A-59
where y - 0.0083 D - EI(b (1 -
(for H/L - 0.57) v 2)
(Equation 3-33)
D - 29.6 x 10 6 x 0.667/1 (1 - .3 2) - 2.17 x 10 7 KE - 2.17 x 10 7/0.0083 x 48 4 - 492.psi/in
Step 10.
Determine the equivalent elastic deflection, XE .
Step 11.
Calculate the effective mass of the element a.
KLM (average elastic and plastic) - (0.78 + 0.66)/2 - 0.72
b.
Unit mass of element, 2 x 1 x 1 x 490 x 10 6
psi-ms 2 - 1,468
m - wig 1,728 x 32.2 x 12 c.
in
Effective unit mass of element, me me - KLM x m - 0.72 x 1,468 - 1,057 ----in
Step 12.
Calculate the natural period of vibration, TN TN - 2n (1,057/492)1/2 - 9.2 ms
Step 13.
Determine the door plate response for: P/ru - 1100/329 - 3.34 T/T N - 1.0/9.2 - 0.109 (Figure 3-62, Region C)
Cl - 100/1100 - 0.091
SA-50
TM 5-l300/NAVFAC P-397/AFR 88-22 C2 > 1000 Using Figure 3-253, - 1.5
Xm Step 14.
Determine the support rotation.
eStep 15.
1.5 x 0.669 - 1.00 in
tan- l (1/24) - 2.39· < 3·
Evaluate the selection of the dynamic increase factor. (Equation 5-1)
fds/Est E tE/T - 1.8, € -
(Figure 3-253)
t E - 1.8 ms
68.2/29.6 x 10 3 x .0018 - 1.28 in./in./sec.
DIF - 1. 3
(figure 5-2, average of A36 and A5l4)
Initial selection of DIF - 1.24 is adequate. Since the support rotation criteria has been satisfied and the preliminary selection of the DIF is acceptable, a 2 inch thick plate is used in design. Steps 15 through 18
These steps are bypassed since the door plate has no stiffening elements.
Problem 5A-8 Problem:
Design of Doubly Symmetric Beams Subjected to Inclined Pressure-Time Loading
Design a purl in or girt as a flexural member which is subjected to a transverse pressure-time load acting in a plane other than a principal plane.
Procedure: Step 1.
Establish the design parameters.
5A-6l
TIl: 5-1300/NAVFAC P-397/AFR 88-22
a.
Pressure-time load
b.
Angle of inclination of the load with respect to the vertical axis of the section
c.
Design criteria:
d.
Member spacing, b
e.
Type and properties of steel used:
Maximum support rotation limited to 2·.
Minimum yield strength for the section (Table 5-1) Dynamic increase factor, c (Table 5-2) Step 2.
Preliminary sizing of the beam. a.
Determine the equivalent static load, w, using a preliminary dynamic load factor equal to 1.0. w-l.Oxpxb
b.
Using the approPriate resistance formula from Table 3-1 and the equivalent static load derived in Step 2a, determine the required ~.
c.
Determine the required section properties using equation 57. Select a larger section since the member is subjected to unsymmetrical bending. Note that for a load inclination of 10·, it is necessary to increase the required average section modulus, (1/2) (S + Z), by 40 percent.
Step 3.
Check local buckling of the member (Section 5-24).
Step 4.
Calculate the inclination of the neutral axis (equation 5-24).
Step 5.
Calculate the elastic and plastic section moduli of the section (equation 5-25).
Step 6.
Compute the design plastic moment,
Step 7.
Calculate ultimate unit resistance, r u' of the member.
Step 8.
Calculate elastic deflection, 0 (Section 5-32.3).
Step 9.
Determine the equivalent elastic unit stiffness, KE, of the beam section using 0 from Step 8.
Step 10.
Compute the equivalent elastic deflection, XE, of the member as given by XE - ru/KE.
5A-62
~'
(equation 5-6).
TM 5·l300/NAVFAC,P.397/AFR 88-22, Step 11.
Determine the load-mass factor, KLM, and obtain the effective unit
mass, ~~,' of the element.
Step 12.
Evaluate the natural period of v,ibration, TN'
Step 13.
Determine the dynamic response of the beam. Evaluate P/r u and T/TN, usin~ the respons~ charts of Chapter 3 to obtain ~/XE and e. Compare with, criteria.
Step 14.
Determine the ultimate ,dynamic shear capacity, ~p" (eq~ation 5-16) , and maximum support shear, V, using, Table 3-9· of Chapter 3 'and check adequacy.
Example 5A-8
Design an I-Shaped Beam for Unsymmetrical Bending Due to Inclined Pressure-Time Loading
Required:
Design a simply-supported I-shaped beam subjected to a pressuretime loading actipg at an angle of 10', with, respect to the principal vertical plane of the beam. This beam is part of a structure designed to protect personnel •.
Step 1.
Given: a. Pressure-time loading [figure 5A-8 (a)]'
b.
Design criteria:
The structure is to be designed for more
than one "shot."
A maximum end rotation - 1°, is therefore
"assigned; c.
Structural configuration [figure 5A-8(a)]
d.
Steel used:
A36
Yield strength, f y - 36 ksi
(Table 5-1)
Dynamic increase factor, c - 1.29
(Table 5-2)
. '-, Average yield strength increase factor, a - 1.1
(Section 5-
12.1) Dynamic design strength, f ds - 1.1 x 1.29 x 36 - 51.1 ksi . (equation 5-2) Dynamic yielding stress in shear, f dv 51.1 - 28.1 ksi
0.55 f ds - 0.55 x (equation 5-4)
Modulus of elasticity, E - 29,000,000 psi Step 2.
Preliminary sizing of the member. a.
Determine equivalent static, load.
5A-63
TH 5-1300/NAVFAC P-397/AFR 88-22 Se1ect··DLF - 1. 2
(Section 5-22.3)
w - 1.25 x 4.5 x 4.5 x 144/1,000
3.65 ·k/ft y
.
IP /TRANSVER,SE 17 LOAD .
19'-0'
,jE-x
w = 4.8 psf (excluding beam weight
: :
Spacing: b = 4.5'
~
.,
, 4.5 psi
;:)
~
II:
~:--.
II. L -
20ms TIME
Figure 5A-8(a) b.
.,'.
Beam configuration and loading,
Determine minimum
·:s•.: · ,•
165 k-ft'
·
(Table 3-1)
Selection of a member. For a load acting in the plane of the web, (S + Z)
2Mp/fds
(2 x 165 x 12)/51.1
(S + Z) - 77.5 in 3
,
Example 5A-8(a)
required-~
~ - (wL 2)/8 - (3.65 x 19 2)/8 c.
r
(S +'Z) required
1.4 x
n.5 - 109 in 3 .,
5A-64
(Equation 5-7)
TH 5:1300/NAVFAC P-397/AFR'88-22
(S + Z) - 116.1 in 3, Ix - 385 in4 I y - 26.7'in4 Check against local buckling ..
Step 3.
,
For W14 x 38,
.
d/~ - 45.5 < (412 / (36)1/2) (1
1.4 x P/Py) - 68.66 O.K . . (Equation 5-17)
O.K.
(Section 5-24)
Inclination of elastic and plastic neutral axes with respect to the x-axis.
Step 4.
','
"j
tan a ~ (Ix/l y) tan ¢ - (385/26~7) tan 10· a
- 2.546 '(Equation 5-24)
- 68.5·
Calculate the.equivalent elastic section modulus. S - (SxSy)/(SyCOS
+ Sxsin
¢
¢) ,
Sx - 54.6 in 3, Sy - 7.88 in 3, ¢'- 10· sin 10· -
0.174,
cos 10· - 0:985
S - (54.6) (7.881/(7.88 x 0.985 + 54.7 x 0.174) - 24.9 in 3 Calculate the plastic section modulus, oZ.
Step 5.
(Equation 5-6)
Z - Acml +,Atm2 Ac - At - A/2 - 11.2/2 - 5.6 in 2 .:
.,'" ,
" "
.
-.
5A-65 , '
,',
TK S-1300/NAVFACP-397/AFR 88,22·
y
Let be the distance of the e.g. of the area of cross section in compression from origin as.shown in Figure.5A-S (b). 1 y - --
[
14.10 . 6.770 x 0.5i5 x ( - - -
5.6
2
1 + - (14.10 2 1 14.10 x- (-2 2
0.515 2
2 x 0.515) x 0.310
5.42 inches
0.515)]
m1 - m2 - y sin a - 5.42 sin 6S· 30' -
.
5.05 inches
Z - 2Acm1- 11.2 x 5.05 - 56.5 ·in3
TRANSVERSE LOAD
,
. ., .
...
,
in
-r---L.x
---- Elastic anCl Plastic Neutral Axis Thro.ugh the Centroid
I/)
ci
~10· 6.770·
Figure 5A-S(b) Step 6.
Loading on beam section, Examp Le 5A-S(b)
Determine design plastic moment'.
'1> -
'1>:: .
fds(S + Z)/2 - 51.1(24.9 + 56.5)/2
. SA-66
(Equation 5-7)
TH 5-1300/NAVFAC,-P-397/AFR '88-22 ... 51.1 x 40.7 ,-,,2;080 in-kips ,-, Step·7.
,
'
;
.
Calculate u1 tim'!-te unit res Ls t.ance , 'ru''riJ - 8~/L2 - (8) (2;080) (1',000)/(19 x:12)2.- 320 1b/in (Table 3-1) ,
Compute elastic deflection, G. ., 2 G - [(Gx + G 2)J1/2
Step 8.
Y
."
(Section 5-32.3)
,
-
5w cost/) L4 Gy - -
-
-
- ,
,
-
384El x
. , -e : .' <)
5w sinrtl L4 , J'
",.,
Gx ' -,' ~~~~-
384El y
w -
equivalent static load + dead load (4.8 x 4.5)
+ 38
2.92 +
e: ,kips/ft"
.,..; 2.94 kips/ft
1,000 I,
G-
[,
Swcost/)L4 .]2
SwsinrtlL4 ]2
+ 384El x
384El y
5wL4
.,
i
r .
.'
'
'"
,
,
'
'-
.If,.;
',
-, : -
,.~
[(0.652)2 + (0.256)2 ]1/2-_ ,2..,08' 'inches '
38,400E Step 9.
Calculate the equivalent elastic unit stiffness, KE . 2.94 x 1,000 x 1 12 x 2.085 . fj': .
,1l7.8'lb/in 2
-,
,
(Get w from 'Step 8) Step 10.
Determine the equivalent elastic deflection, XE .
Xi -
ru/KE - 320/117.8 - 2.72 inches
SA'-·67·;
-r , I
'
TM 5-1300/NAVFAC P"397/AFR' 88-22 Step 11.
Calculate the effective mass of the.e1ement, me' Load-mass factor, KUC'
a.
1 •.1
(Table 3,12) ,
KLM (average elastic:and plastic) - (0.78 + 0.66)/2 b.
~
0.,72
Unit mass of element, m [(4.5 x 4.8) + 38J x 10 6 m - wig - - - - - - - - - - - 32.2 x 12 x 12
c.
Effective unit mass of element, me
me - KLMm - 0.72 x 1,286 x 10 4 - 0.93 ~'l04 lb-ms 2/in2 Step 12.
Calculate the natural period of vibration, TN' TN - 21r '[('93 x 10 2 ) / 117 . 8 ]1/ 2 - 55.8 ms
,
Step 13.
(Section 5-22.2)
Determine the beam,response. Peak overpressure
P - 4.5 x 4.5 x 12 - 243 lb/in
Peak resistance Duration
.
r u - 320'lb/in ,
T - 20 ms
'
Natural period of vibration,
TN - 55.8 ms
P/ru - 243/320 - 0.7'6: ' T/TN - 20/55.8 - 0.358 From figure 3-64a ~/XE
< 1 "
,
From figure 3-49, for T/T N - 0.358"
5A-68
TK 5-1300/NAVFAC P-397/AFR 88-22·
.
DLF - -0.97 Hence, Xm - 0.97 x 4.5 x 4.5 x 12/117.8 - 2.0 in Find end rotation, tan
Step 14.
e.
e-
Xm/(L/2) - 2.0/[(19 x 12)/2] - 0.0175
e -
1~0'
O.K.
Calculate the dynamic ultimate shear capacity,. Vp' and check for adequacy. Vp - fdvAw - 28.1 (14.10 - 2 x 0.515) (0.310) - 113.9 kips (Equation 5-16)
v-
DLF x Px bx L/2
.
- 0.97 x.4.5 x 4.5 x 19 x 144/(2 x 1,000) - 26.9 kips < 89.2 kips < Vp
O.K.
(Table 3-9)
TN 5-1300/NAVFAC P-397/AFR 88-22
APPENDIX 5B LIST OF SYMBOLS
TK 5-l300/NAVFAC,P-397/AFR 88-22
a
yie~d,stress
A
Area of cross section (in 2 )
Ab
Area of bracing member (in 2 )
increase factor
,.
Area of cross section in compression (in 2 ) Area of cross section in tension (in 2 )
•• j'
b
Width of tributary loaded area (ft)
bf
Flange width (in)
~
Tributary width for horizontal loading (ft)
bv
Tributary width for vertical io~ding (ft)
c,DIF (1) (2)
c
Dynamic increase factor Distance from neutral axis to extreme flexure (in) , .
fibe~
of cross-section in
C,Cl
Coefficients indicating relative column to girder moment capacity (Section 5-42.1)
Cb
Bending coefficient defined in. Section 1.5.1.4.5 of the AISC Specification
Cc
Column slenderness ratio indicating the transition from elastic to inelastic buckling Coefficients applied to the bending terms in interaction . . formula (AISC Specification Section 1.6.1)
C2
Coe'fficlent in approximate exp r as s Lon for sidesway stiffness factor (Table 5-14) - ". -. r
o
Coefficient indicating relative girder to columri stiffness (Table 5-14)
DLF
Dynamic load factor
d
(1) (2)
E ,f
Web depth (in) Diameter of cylindrical
"
port~on
Young's modulus of elasticity (psi) (1) (2)
Maximum bending stress (psi) Shape factor, S/Z
5B-l
of fragment, in.
,
,
TK 5-l300/NAVFAC P-397/AFR 88-22
Axial stress permitted in the absence of bending moment from Section 537.3 (psi) Bending stress permitted in the absence of axial force (psi) Web buckling stress (psi) Maximum dynamic design stress for connections (psi) Dynamic design stress for bending, tension and compression (psi) Dynamic yielding shear stress (psi) Dynamic ultimate stress (psi) Dynamic yield stress (psi) F'ex,F'ey
Euler buckling stresses
di~ided
by safety factor (psi)
FH
Horizontal component of force in bracing member. (lb)
Fs
Allowable static design stress f~r connections (psi)
fu
Ultimate tensile stress (psi)
fy
Minimum static yield stress (psi)
g
Acceleration due to gravity (386 in/sec 2)·'
H
Story. height (ft)
h
Web depth for cold-formed, light gauge steel panel sections (in)
1
Moment of inertia (in4) Average column. moment of inertia for single-story multi-bay frame(in 4) Effective moment of inertia for cold-formed section at a service stress of 20 ksi (in4 per foot width)_ Moment of inertia about the x-axis (in4)
-1
Y
K
Moment of inertia abo~t the y-axis (in4) .
(1)
(2)
.';
Effective length factor f~ra compression member Stiffness factor for rigid single-stor~, multi-bay frame from Table 5-14
Kb
Horizontal stiffness of diagonal bracing (lb/ft)
KE
Equivalent elastic stiffness (lb/in or psi/in)
5B-2
P-3~7/A~R
TH 5-l300/NAVFAC
e
K· L
Load factor .,
.~
..:.
'
88-ZZ
,,
,
,
KUI
Load-mass factor
KM
Mass factor
L
(1) (Z)
1
Distance between cross section braced against twist or lateral displacement of compression flange or distance between points of lateral support for beams or col~ns '
llr
Slenderness ratio
Ib
Actual unbraced length in the plane of bending (in)
lcr
Critical unbraced length (in)
M
Total effective mass (lb-msZ/in)
,.
Span length (ft, 01:", in), Frame bay width (ft)'
"
"
"
."
• I
"
Moments about the x- and y-axis that can be resisted by member in the absence of. .;>xial load, ,,, ~
Design plastic moment capacity
Ml,MZ Design plastic moment capacities, (f Lgure s ·5-6 and:S-lO) ~x'~y
Plastic bending moment capacities about .the x- and y-axes
~u
Ultimate dynamic moment capacity
~p
Ultimate positive moment capacity .f or :unit width of panel
~n
Ultimate ne gat.Lve-momerrtvcapac Lt.y for unit width of paneI:
"
Moment corresponding to first yield m
(1) (Z)
Unit mass of p'anel ,(psi'-msZ/-in),' Number of braced bays in multi-bay frame
Eff~~U';,e ~n'it~~~s
> "
: '.'
.
"
(psi-msZ/in)
Distance; from plastic neutral axis to the centro Ld of the .area in compression in a fully plastic section (in)
,mZ
Distance from plastic neutral axis to the centroid of the area in tension in a fully plastic section (in) ,
N
Bearing length at support for cold-formed steel panel (in)
n
Number of bays in multi-bay frame
p
(1)
Applied compressive load (lb) 5B-3 , '
.,
TH 5-l300/NAVFACP~397/AFR88'22"
(2)
Peak pressure of equivalent triangular loading function, (psi) [when used with rul. or peak total blast load (lb) [when used with
Rul,
Euler buckling loads about the x- and y-axes Ultimate capacity for dynamic axial load, Af
dy
(lb)
.
Ultimate axial compressive load (lb) .
. ; - ,.
Ultimate capacity for, static aXial'load',Af
"
y
Ph
Reflected blast pressure on front wall (psi)
Pv
Blast overpressure on roof (psi)
(Lb):'
Ultimate support capacity (lb) "
qh
Peak horizontal load on rigid frame (lb/ft)
qv
Peak vertical load on rigid frame (lb/ft)
R
Equivalent total horizontal static load on 'frame' (lb)
Ru
Ultimate total flexural resistance (lb) Radius of gyration of bracing member ,(in)
rT
Radius of gyration, equation 5-22 (in)
ru
Ultimate unit flexural resistance (psi)
rx,r y Radii of gyration' about 'the x- and y-axes"(in) r
Required resistance for elastic behavior in rebound (jJsi)
s
Elastic section modulus (in 3,) Elastic section modulus about the x- ,and y-axes (in 3 ) ,
.
Effective section modulus of cold-formed' section for positive moments (in 3 ) Effective 'section modulus of cold-formed ,section"for negative ,moments (in 3 ) T
Load duration (sec) Natural period of vibration (sec)
t
tE
(1) (2)
Thickness of plate element (in)' Thickness of panel section (in)
Time to yield (sec) 5B-4
","
TK 5-i300/NAVFAC, P-397/AFR 88 c22' Flange thickness (in)
"
Time to maximum response '(sec)
t,.
Web thickness (in)
V
Support shear (lb)
V
Ultimate shear capacity (lb)
Vr
Residual velocity of fragment (fps)
Vs
Striking velocity of fragments, (fps)
,
'
Critical perforation velocity of fragment (fps) II'
Total weight (lb)
We
Total concentrated load (lb)
WE
External work (lb-in)
Wf
Fragment weight (oz.)
WI
Internal work (lb-in)
(1) (2) (3 )
,
,.
Flat width of plate element (in) Load per unit area (psi) Load per unit length (lb/ft)
Xo
Deflection at design ductility ratio
XE
Equivalent elastic deflection (in)
Xm
Maximum deflection (in)
x
Depth of penetration of steel fragments (in)
Z
Plastic section modulus (in 3 )
(~igure
5-12)
Zx,Zy Plastic section moduli about the x- and y-axes (in 3 )
(1) (2)
B
(1) (2) (3)
Angle between the horizontal principal plane of the cross section and the neutral axis (deg) , Ratio of horizontal to vertical loading on a frame Base fixity factor (Table 5-14) Support condition coefficient (Section 5-34.3) Critical length for bracing correction factor (Section 5-26.3)
y
Angle between bracing member and a horizontal plane (deg)
s
(1) (2)
'Total transverse elastic deflection (in) Lateral (sidesway) deflection (in) 5B- 5 '
TK 5-l300/NAVFAC P-397/AFR 88'122
E
Strain (in/in)
E
Average strain rate (in/in/sec)'
8
( 1)
(2)
,
,
"
Member end rotation (deg) Plastic hinge rotation (deg)
,,,.,
8 max
Maximum permitted member end rotation
~
Ductility ratio
'"max
MaximJm permitted ductility ratio
Angle between the plane of the load and the vertical principal plane of the cross section '(deg) •• ,-
,
"l ,
It
,..
.'
.
,
';
.
"
,"
"
J,
"
...
.,
l~'
"
,
, -:
"
"t
'J
, "
,
.
,;'
,
,
.'
, ,
t,
'-
5B-6 :' '
-.;
"'":'
r"
.-
••
,0,
.'.
" t
f.
.' t' t
,
TK 5-l300/NAVFAC P-397/AFR 88-22
APPENDIX 5C BIBLIOGRAPHY
TH
5-l300/NA~FAC,P-397/AFR
88-22
1.
Healey, J .', eJ: a1., Design of Steel ,Structures to Resist the Effects of HE Expl~sions; by Ammann &'Whitney, Consulting Engineers, New York, New York, Technical Report 4837, Picatinny Arsenal, Dover, New Jersey, ""4· . ., . , August 1975. ", ." '., ' ,
2.
Keenan, ~tL, T~ncreto I J., Mey,:rs,. G.
3.
4.
5.
6.
7. 8. 9. 10. 11. 12.
13. 14. 15. 16.
I
Joh:nson,. F., Hopk Lns , J.,
~
Nickerson, H. and Armstrong, 'W." NCEL Products Supporting DOD Revision of NAVFAC 'P-397, ,Program No. ,Y0995-0l-003-20l, Technical Memorandum, 259lTM, sponsored by Naval Facilities Engineering Command,'~lexandria, Virginia, ,and. Naval Civil Engineering Laboratory, Port .Hueneme, Ca I Lf ornt.a-, March 1983.' ' .. Stea-: W., Ts~ng, G. and Kossover, D.. Nonlinear Analysis of Frame Structures Subjected to Blast Overpressures, by Ammann & Whitney, Consulting Engineers, New York, New York, Contractor Report ARLCD-CR77008, U.S. Army Research and Development Command, Large Caliber Weapons Systems Laboratory, Dover, New Jersey, May 1977. Stea, W., and Sock, F., Blast-resistant Capacities of Cold-formed Steel Panels, by Ammann & Whitney, Consulting Engineers, New York, New York, Contractor Report ARLCD-CR-8l00l, U.S. Army Research and Development Command, Large Caliber Weapon Systems Laboratory, Dover, New Jersey, May 1981. Stea, W., Dobbs, N. and Weissman, S., Blast Capacity Evaluation of Preengineered Building, by Ammann & Whitney, Consulting Engineers, New York, New York, Contractor Report ARLCD-CR-79004, U.S. Army Research and Development Command, Large Caliber Weapons Laboratory, Dover, New Jersey, March 1979. Healey, J., Weissman, S., Werner, H. and Dobbs, N., Primary Fragment Characteristics and Impact Effects on Protective Barriers, by Ammann & Whitney, Consulting Engineers, New York, New York, Technical Report 4903, Picatinny Arsenal, Dover, New Jersey, December 1975. Manual of Steel Construction, Eighth Edition, American Institute of Steel Construction, New York, New York, 1980. Specification for the Design, Fabrication and Erection of Structural Steel for BUildings and Commentary thereon, American Institute of Steel Construction, New York, New York, 1978. Specification for the Design of Cold-Formed Steel Structural Members and Commentary thereon, American Iron and Steel Institute, New York, New York 1968. Design of Structures to Resist the Effects of Atomic Weapons, Department of the Army Technical Manual TM 5-856-2, 3 and 4, Washington, D.C., August 1965. Bleich, F., Buckling Strength of Metal Structures, McGraw-Hill, New York, 1952. Newmark, N.M, and Haltiwanger, J.D., Principles and Practices for Design of Hardened Structures, Air Force Design Manual, Technical Documentary Report Number AFSWC, TDR-62-l38, Air Force Special Weapons Center, Kirtland Air Force Base, New Mexico, December 1962. Seely, F.B., and Smith, J.O., Advanced Mechanics of Materials, Second Edition, J. Wiley & Sons, New York, New York, '1961. Bresler, B., et al., Design of Steel Structures, J. Wiley & Sons, New York, New York, 1968. Johnston, B.G., ed., Guide to Design Criteria for Metal Compression Members, Second Edition, J. Wiley & Sons, New York, New York 1966. McGuire, W., Steel Structures, Prentice Hall, Englewood, New Jersey, 1968. 5C-l
TK 5-1300/NAVFAC P-397/AFR 88-22 17. 18. 19. 20. 21.
Tall, L., et al., Structural Steei Design, Ronald Press, "New York, N~w York, 1964. Beedle, L. S., Plastic Design of'Steel Frames, J. Wiley & Sons, New York, New York, 1958. . Hodge, P.G., Plastic Analysis of Structures, MCGraw-Hill, New York, 1959. Massonet, C.E., and' Save, M.A:, Plastic Analysis and Design, Blaisdell,' New York, New York, 1965. Plastic Design in Steel, A Guide and Commentary, Second Edition, Joint Committee of the Welding Research Council and the American Society of Civil Engineers, ASCE, 1971.
'.
5C-2
"STRUCTURES TO RESIST THE EFFECTS OF ACCIDENTAL EXPLOSIONS"
CHAPTER 6. SPECIAL CONSIDERATIONS IN EXPLOSIVE FACILITY DESIGN
TM 5-1300/NAVFAC' P- 397/AFR88':22,
"
, CHAPTER 6', 1 -;
'SPECIAL CONSIDERATIONS·IN EXPLOSIVE FACILITY DESIGN ".
-
...
INTRODUCTION
"
,
: 'J
6-1, Purpose
,,
"
'i
The
p~rpose
of this manual is to: present. methods of design for protective
";'
.. "
construction'used in facilities for development, .testing, production, storage, maintenance', modification, Inspec t i on.: demilitarization, and disposal of· " explosive materials. L
6-2. Objective',
,'."
""
The primary objectives are to establish design procedures and construction techniques whereby propagation of 'explosion (from one· structure' or part of a structure to another) .o r mass "detonation ,~an be -p r everrted and' to provide protection for personnel and valuable ':equipment .. The secondary objectives are to: (1)
Establish the blast load parameters required for design of protective' structures. e, ,,". (2) Provide methods .fo r calculating the dynamic response of structural elements including reinforcea concrete, and structural 'steel. (3) Establish construction details and procedures necessary to afford the required strength to z e s I's t; the applied blast' Lo'ads , (4)' Establish guidelines' for siting explosive' facilities to obtain '; maximum cost e f f'e'c t Lvene s s: in -hoth the planning 'and s t ruc t.ur a L arrangements. providing closures', 'and preventing' 'damage to interij'
or portions of structures because of structural motion, shock, and fragment .pe rfo r at.Lorr; ~ )'
6-3. Background
~.~
,
•, •
~
j' ..:
. '" .
"
For the first 60 years of the 20th ,century, criteria and methods based upon' results' of catastrophic' events', were' used for the de s I'grr-of 'explosive facilities. The.vc r I t e r La -and .methods 'did' no r 'include a detailed 'or reliable quant Ltative basis".for assessing, the'''(jegree,,'of'prot'ection 'afforded by the protectIve facility , I n ' the -La ce 1960' s -quantitative, procedures were set' forth in the first edition'of the present ,manual~ ~Stiuctures to Resist the Effects of Accidental "Explosions" .: This ·m"anual wa"s based on 'extensive -r e s e a r ch and development programs which permitted a more reliable approach,to current and' future design requirements,' Since the original publication, of' this manual" ,,' more extensive testing and development programs have taken p Lace . : This additional research included work with "materials other than reinforced con, crete which was" the p r Lnc i pal'
"
.
TM 5-l300/NAVFAC p c 39 7/ AFR, 88 - 22
other explosive materials.) It is evident that a requirement for more accurate design techniques is essential. This manual describes rational design methods to provide the'. required structural .p ro t.ec t Lorr. : " .,'
, 6-4. Scope
.'
"
It is not the intent of this manual to establish safety criteria, Applicable documents should be consulted for, this purpose. Response predictions for personnel and equipment are included for information. " ; In this manual' an effort is made ~p cover the more ,pt:obable design situations. However, sufficient general information, qn pro,tective .de s Lgn. techniques has,,' been included in order that application of , the basic theory can be made: to situations other than those ~hich were fully considered. ' , 'r, ., .' .~.
~
This manual is applicable to the design of protective -st.ruc tures subjected to the effects associated with high explosive detonations. For these design situations, the manual will apply for explosive quantities less than 25,000 pounds for close-in effects. However, this manual is also applicable to other situations such as far-' or Lnce rmed Lace-Yange effects'. For ,these latter cas e sf., the design procedures are 'applicable for explosive 'quantities in the order. o f.» 500,000 pounds wh~ch is the m~ximum quantity of. high ,explosive approved ,for. abovegr-ound s t.orage facilities .Ln the ,Oepa~tment,'of-De f'ens e manual, "Ammuri-rition and Explosives Safety Standards" ° ,DOD, 6055.9~S,TO. Since tests were primarily directed toward the response of structural steel and reinforced, concrete elements to blast ove rp re ssures., this manual concentrates -on des Lgn -, procedures and techniques for these ,materials. Howeve ri. this does, not imply. that concrete and steel are,.the only useful materials .f'o r protective construction. Te s't s to establish the re sponse iof wO,od, brick, blocks, and plastics,. as. well as the, blast attenuating ,and mass effects' .of soil are contemplated. .Theresults of these tests may require, at' 'a later date", the "supplementation, of' these design methods for these and other materials. ..' Other manual.s are available; to .,desigh:,protective. s t ructure s against .t.he effects of high explosive or nuclear de t.ona t Loris-.' The procedures .Ln these" manuals will quite often complement this ·manual ,and shoul.dibe 'consulted for specific app Lfc at Lons. ,!~, . , ','
6-2
'j
.
'
'. "
,
"
lfr
, :3"';
TK 5-l300/NAVFAC P-397/AFR 88-22 -
Computer programs, which are consistent with procedures and techniques contained in the manual, have been approved by the appropriate representative of the US Army, the US Navy, the US Air Force and the Department of Defense Explosives Safety Board (DDESB). These programs ~re available through the following repositories: (1)
Department of the Army Commander and Director U.S. -Army Engineer Waterways Experiment Station Post Office Box 631 Vicksburg, Mississippi 39180-0631 Attn: WESKA
(2)
Department of the Navy Commanding Officer Naval Civil Engineering Laboratory Port Hueneme, California 93043 Attn: Code LSI
(3)
Department of the Air Force Aerospace Structures
Information and Analysis Center Wright Patterson Air Force Base
Ohio 45433 Attn: AFFDL/FBR If any modifications to these programs are required, they will be submitted for review by DDESB and the above services. Upon concurrence of the revisions, the necessary changes will be made and notification of the changes will be made by the individual repositories. 6-5. Format This manual is subdivided into six specific chapters dealing with various aspects of design. The titles of these chapters are as follows: Chapter Chapter Chapter Chapter Chapter Chapter
I 2 3 4 5 6
Introduction Blast, Fragment, and Shock Loads Principles of Dynamic Analysis Reinforced Concrete Design Structural Steel Design Special Considerations in Explosive Facility Design
When applicable, illustrative examples are included in the Appendices. Commonly accepted symbols are used as 'much as possible. However, protective design involves many"different scientific and engineering fields, and, therefore, no attempt is made to standardize completely all the symbols used. Each symbol is defined where it is first used, and in the list of symbols at the end of each chapter.
6-3
1M 5-1300/NAVFAC P-397/AFR 88-22 CHAPTER CONTENJS 6-6. General
•
This chapter contains procedures for the design of blast resistant structures other than above ground, cast-in-place concrete or structural steel structures, as well as the design of other miscellaneous blas.t resistant components. Included herein is the design of reinforced and non-reinforced masonry walls, recast elements both prestressed and conventionally reinforced.,.preengineered buildings, suppressive shielding, blast resistant windows,. underground structures, earth~covered arch·type magazines, blast valves and. shock isolation systems.
','
.
...
.,
...
.
6-4
..
,
TK 5-1300/NAVFAC P-397/AFR 88-22
-MASONRY
;
-
.. j
.
6-7. Application Masonry units are used·primarily for.wall construction. These'units may. be used for both exterior walls subj ected to blast overpressures' 'and interior walls subjected to inertial effects due to building motions~,'Basicvariations in wall configurations may be related t~ the type of masonry unit such as brick, clay tile or solid and hollow concrete masonry'units (CM) , and.iche manner in which these units' are laid (running bond, stack bond, etc,), the number of wythes of units (single or double) ,
.
..
"',
In addition to their inherent advantages with respect to fire protection", acoustical and thermal insulation, structural mass and resistance to flying debris, ma~onry walls when properly' designed and detailed can provide economical resistance'to relatively low blast pressures, However, the limitation on their application includes a limited' capability for large deformations;';' reduced capacity in rebound due to tensile cracking in the primary phase of the response as well 'as, 'the limitations on ,the -amount; and type' of reinforcementwhich 'can be provided,' Because of 'these' limitations, masonry construc: tion in this Manual is ,limited to concrete ma-sonry unit '(CM) walls' placed in a running bond and with single or multiple wythes . . However, because of the
difficulty to achieve the required' Lrrte r ac t Lon between the individual"wythes,' the use of multiple wythes should..be avoided, ' ,; Except for small structures (such as tool sheds,' garage, etc,) where the 'floor area of the building is relatively small and interconnecting block walls can function as shear walls, masonry' walls will usua l.l.yYequdre supplementary framing to transmit the lateral forces produced by the blast forces to the buLLdLng foundation, Supplementary' framing is gene ra Ll.y classified into 'two categories (depending on the type ofvc orrs t r'uc t Lon used); name Ly': (1) flexible type supports such as structural steel framing, and (2) rigid supports including reinforced concrete frames 'or shear wall slao construction; , The use of masonry walls in combination with structural steel frames is usually limited to incident over-pressures of 2 psi or less while masonry walls when supported by'''rigid· supports may De:' designed 'to r'e sLs t; incident pressures as high as 10 psi, Figures 6-1 and 6-2 illustrate these masonry support systems, Depending on the type of 'construction used, masonry walls may,be classified into three categories: namely (1) cavity walls, (2) solid walls, and (3) a combination of cavity and solid walls, The cavity walls utilize 'hollow load bearing concrete masonry Jriits conforming to·ASTM C90, Solid walls' use solid load-bearing concrete masoriry units conforming to ASTM C145 'or' hollow units whose cells and voids are filled with grout, ,The combined cavity and solid walls utilize the combination of ho l l.oeeand- solid units',' Masonry walls may be subdivided further depending on the type of load-carrying mechanism desired: (1) joint reinforced masonry construction, (2) combined joint and cell reinforced masonry construction, and (3) non-reinforced masonry construction, •
' •. '"
; I
.J'
:
.1
l'
.'
•
Joint reinforced masonry construction consists of single or multiple wythes walls and utilizes either hollow or solid masonry units, rre joint reinforced wall construction 'utilizes comme rcLa l.fy available' cold drawn wire 'assemblies (see Figure 6-3), 'which 'are pl aced in the bed joints between the, rows of the masonry units, Two types'ofreinfoicement are available; truss and ladder 6-5
TK 5-1300/NAVFAC P-397/AFR 88-22 types. The truss reinforcement provides the more rigid system and, therefore, is recommended for use in blast resistance structures. In the event that double wythes are used, each wythe must be reinforced independently'. The wythes must also be tied together using wire ties. Joint reinforced masonry construction is generally used in combination with. flexible· type supports. The cells. of, the units located at the wall supports must be filled with grout. Typical joint reinforced masonry construction is illustrated in Figure 6·4. Combined joint and cell reinforcement masonry construction consists of single
wythe walls which utilize both horizontal and vertical reinforcement.
The
horizontal ,reinforcement may consist either of the. joint reinforcement
previously discussed or reinforcing bars. Where reinforcing bars are used, special masonry units are used which permit the reinforcement to sit below the j oint (Figure 6 - 5) .. The vertical reinforcement consists of reinforcing bars which are positioned in one or more .of the masonry units cells. All cells, which contain reinforcing bars, must be filled with grout. Depending.on the amount of, reinforcement used, this type of construction may be used with either the. flexible or rigid type support systems . . Non·reinforced masonry construction consists of. single wythe of hollow' or solid masonry units. This type of construction,does not utilize reinforcement for strength but solely relies on th~ arching.action of the masonry units formed by the wall deflection and support resistance (Figure 6·6). This form of construction is utilized with the rigid type. support system and, in particular, the shear wall and slab construction system.
6-8. Design Criteria for Reinforced Masonry Walls 6·8.1. Static Capacity of Reinforced Masonry Units. Figure 6·7 illustrates typical shapes and sizes of concrete masonry units " which are commercially available. Hollow masonry units shall conform to 'AS!M e90, Grade N... This grade is recommended for, use in exterior below and above grade and for interior walls. The minimum dimensions of the components of hollow masonry units are given in Table 6-1. .. . The spec Lf'Lc compressive strength (f'm)' for concrete masonry units may be" taken as:
Type of Unit
Ultimate Strength (f'm)'
.
'Hollow Units Hollow Units filled with.grout Solid Units while the modulus of elasticity (Ern) of
~asonry
Em - 1000 f' m
,1350 psi 15QO psi 1800 psi units is equal to:6-1
The specific compre~sive stre~gth and the ,modulus of elasticity of the mortar may be assume~ ,to be equal to that of the unit. , Joint' reinforceme";t shall conform to the requireme~ts of ASTM A82 and, therefore, it will have a minimum.ultimate (~un~,and yield (fm) stresses equal to 80 ksi ~nd 70 ksi respectively. Reinforcing ,bars shall conform to ASTM-,
6-6
TH 5-1300/NAVFAC P-397/AFR
88-22~
A615; (Grade 60) andhave minimum.·u1t1mate.stress (f;:m) of90.ksi and mi.m.mum yield stress (fm),'··of 60 ks L; ..The modulus 'of elasticity' ofrt.he r-e Lnf'or cemen t : is equalto.29,OOO,OOO psi. ,. -" ' .• '''. , "
6-8.2 Dynamic Strength of Material
....
..
-, .... , il.' Since design for blast resistant structures is based on ultimate strength; the actual yield stresses of the material. rather than conventional design. stresses or specific minimum yield stresses, are used for det~rmining the plastic strengths of members. Further, under the rapid rates of straining' that occur in structures loaded by blast forces, materials develop higher strengths than they do in the case of statically loaded members. In calculating the dynamic properties of concrete masonry construction· it is !ecommended that the dynamic increase factor be applied to the static yield strengths of the var t ous-tcomponent.s as f'o LLows : ' ... ,
,
11
\
Concrete
.
,
. Flexure ... r
Shear
·!rO,
t'l
.'
"
.JI .
.,~ -1. 00 - f" m.
,,
.',
~ .:.
Compression .",.
,.-
.,
,
Reinforcement
..
.. ~.
.
Flexure
.
1.17
6-8,3 Ultimate Strength of Reinforced Concrete Masonry Walls -,
The ultimate moment capacity of joint reinforced masonry construction may be conservatively estimated by utilizing the horizontal reinforcement only and neglecting the ,compressive strength afforded by the concrete', That is the reinforcement'in one. face will deve1op'the:tension forces while the steel in the opposite face resists the compression stresses, The ultimate moment relationship may be expressed for each horizontal joint of the wall as follows: . 6-2 '. ':
where:
, As
,
area of j oint reinforcement at one face
-, f
.,
'
.
dynamic yield strength of the joint reinforcement distance between the centroids of sion reinforcement
the;comp~ession
and ten-
ultimate moment capacity '
On the contrary, the ultimate moment capacity of the cell reinforcement (vertical reinforcement) .Ln a combined joint and cell reinforced masonry construction utilizes the concrete strength to resist the compression forces. 6-7
..
TM
5-l300/N~VFACP-397/AFR
88-22
The method c f vc a Lcu l at Ing ultimate. momeritrs.o f thee vertical. reinforcement 'is' the, same:as. that presented in Chapter' 4 of this:manual which is similar to that'. presented in the American Concrete Institute Standard Building Code' Require •. ments for Reinforced Concrete. . The ultimate shear stress in joint reinforced masonry walls is computed. by the , formula:
,
6-3
,. - '.
where:
'.
v u -.
unit shear stress-·
,
total applied design shear at d c/2 from the support ~
,J
,
'
net area. of section
In all cases, joint reinforced masonry walls, which are designed t~.resist blast pressures, shall utilize shear reinforcement which shall be d~signed to carry the total shear stress. Shear reinforcement.shall consist of; (1) bars or stirrups perpendicular to the longitudinal reinforcement, (2) longitudinal bars bent 'so that the axis or inclined portion of the bent. bar makes an angle of 45 degrees or more with the axis of the longitudinal part of the. bar; or (3) a combination of (1) and (2) above . . The area of .the shear reinforcement placed perpendicular to the flexural steel 'shall be. computed by the formuia:"
."
6-'4 -,
.' ,
. Av.
-
.area of shear- reinforcement. ' . '.1'
b
u~it
s -
spacing between stirrups ..
width of wall
~
j;r,
~~.'.
"
-.
fm
yield stress of the shear reinforcement
strength reduction factor equal to 0.85.'
. When bent or inclined bars are used, the area'of shear reinforcement shall be calculated using:. :,.,
.. :
-
.' 6-5
Q
+ cos
Q) r,
where:
.'
Q -.
angle between inclined stirrup. and longitudinal' axis of the .' \' mem1;>er, .... ,
6-8
-e
TK'
,
~'t'
"
5-l300/NAVFAC'P-397/A~R'88-22'
6-8.4; Dynamic Analysis
.
"
~
--';
..,
The principles for dynamic analysis of the response of structural elements to blast loads are presented in Chapter 3 of this manual. These' principles also apply to blast analyses of masonry walls. In order to.perform these analyses, certain dynamic properties must be established as follows: Load-mass .factors, for masonry walls spanning in 'either one' direction' (j oint reinforced masonry construction) or two directions (ccmbLne d"] oint and 'cell .:, reinforced masonry 'construction) 'are the same as those load-mass factors which are listed in Tables 3-12 and 3-13. The load-mass factors··are appli~d'to the "
.
'
~
,
1
'
•
,
actual mass of the wall.' The weights of masonry wall can-be"determined based
on the properties of hollow masonry units previously descr'ibed 'and utilizing Ii concrete unit weight of '150 pounds per cubic foot: The values of the loadmass factors KU!, will depend in part on the range of behavior, of the wall:;,:'Le., elastic, elasto-plastic, and plastic ranges. An average value of the elastic and elasto-plastic value of KU! is used for the elasto-plastic range while an average value of the average KLM for the elasto-plastic range and KLM of the plastic range is used for the wall behavior in the plastic range.' .
".,
,
,
The' resistance-'deflection 'function is 'illustrated in Figure, 3-'1 .. This 'figure illustrates the various ranges' of behavior'previously discussed and defines the relationship between the 'wall's resistarices and deflections as well as presents the stiffness K in each range 'of behavior. It 'may be noted'in Figure 3-1, that the' elastic and e Las covpLas t.Lc range's'of behavior hive been idealized forming a bilinear (or trilinear)' function. ' The equations for defining these functions are presented 'in Section 3-13. .
'j"
The ultimate resistance r u' of a wall varies; (1) as the distribution'of the applied load, (2) geometry,of the wall (length and width), (3) the amount and distribution of the reinforcement, and (4) the 'nUmber and type of' supports. The ultimate resistances of both one and two-way spanning walls are given in Section 3~9. . ! - '. "
••
Recommended maximum deflection criteria for masonry walls subjected to" blast
loads is presented in Table 6-2. This table includes criteria for both reusable and non-reusable conditions as well as criteria for 'both one.rand two-way spanning walls, ,
,.
When designing masonry 'walls for blast loads using response chart procedures of Chapter 3 the effective natural period of vibration is required. This effective period of vibration when related to the 'duration of the blast loading of given'intensity and a given resistance of the masonry wall'determines the maximum transient deflection ~ of the wall.· ·The expression ,for the 6-9
TM,S-1300/NAVFAC
P-397/AF~88-22
natural period of vibration is' presented in equation 3-60, where" the effective, unit mass'm e has been described prevIously and the equfval.ent; unit~.~tiffness K E is obtained from the resistance-deflection function. The equtval.ent . stiffness of one way beams is presented in Table 3-8. This table may be used for one w~y spanning walls except. that a unit width ,spall ,be used., ,Methods for, de'termining the stiffnesses and period of vibrations· for two-"!"8y'walls are presented in'Sections 3-11 through 3-13. Determining the stiffne~s,.in the e Las t Lc andelasto-plastic range .Ls complicated by, the fact that the moment of, inertia of the .·,c'ros.s sec,tion along the masonry wal~ changes cont~nual1y as
cracking progresses, and further by the fact that the modulus of elasticity changes as the stress increases.
It is recommended that computations for
deflections and therefore, stiffnesses be based on average, moments ,of. inertia I a as follows: ' - . .,• -. r,
2
"
f '
6-6
.
In Equatio~ ,6,-6, Iri is the moment of Lne r t La of the net. section and I c ~s the moment of inertia of the cracked section. For solid masonry units tpe value of In is rep~aced with the moment of, inertia of t~e gross sec t Lon ... The va Lues of In and I g for hollow and solid masonry units used in ,joint 'reinforced masonry construction are listed in Table 6-3. The values ,of I g for solid, units may. also be used for walls which utilize combined joint and cell masonry constructi;n. The values of I c for both hollow an~,solid masonry. co~;truction may be obtained using: .,'
I c - 0.005 bdc 3
;.
'::;
6-7
"
6-8.5: Rebound
.. : ;
t
Vibratory action of a .masonry wall.will result ,in n~gative deflection~ after the maximum positive deflection has. been ":ttained .. ,This negat.Lve . de f l ec t Ion is associated with negative forces.which will require t~nsion reinforcement to be po s Lt Lonedjat; the opposite, side. of the wall from the primary reinforcement ... In additio~, wall ties are requir~d to a~~ure that the wall.is~supporte~ by . the frame (Figure 6-8). The rebound forces are a function of the maximum' resi;tance of the wall as w~ll as the vibratory,properties'of the,w~il and the load duration. The maximum elastic rebound of a masonry wall may be obtained from Figur~, 3-268. _, 6-9. NoncReinforced Masonry.Walls
,.
.....
~
The resistance of non-reinforced masonry walls to lateral blast loads is a function of the wall deflection, mortar compression strength and the rigidity of the supports .
.
.,'
,
6-9.1. Rigid supports
If the supports are completely rigid and the mortar's strength is known, a resistance function can be constructed in. the follOWing manne r, Both supports are assumed to be completely rigid and lateral·motio~ of, the top and bottom of the wall is prevented. An incompletely filled joint is assumed to exist at the top as shown in Figure 6 - 9a. Under the action.of the blast 6-10
TK 5-l300/NAVFAC' P-397/AFR 88-22 load the wall is assumed to crack at the center. Each half then rotates as a rigfd body until the wall takes the position shown in Figure 6-9b. During the rotation the.midpoint m has .undergone a lateral motion Xc in which no resistance to motion will be developed in the wall, and the upper corner of the wall (point 0) will be just touching the· upper support. The magnitude of Xc can be found from the geometry of the wall in its deflected position: L2 -
[hi2 +" (h ,:- it) /2] 2
6-8
~2 _ (h'/2)2 where:
6-9 and , :
.. 6'10
L
All other symbo1s'are shown in Figure 6-9.
,.
For any further lateral motion of point m, compressive forces will occur at
points m and o. These compressive forces form a couple that produces a resistance to the lateral load equaL' to: .', 6-11
where 'a l L symbols have previously been defined. When point m deflects laterally to a line n-o (Figure 6-9c), ·th·e moment arm of .the resisting couple will be reduced to zero and the wall will become unstable with no further resistance to deflection. In this position the diagonals o-m and m-n will be shortened by an amount: 6-12
L - h'/2
The unit strain in the wall caused by the shortening will be: Em -
6-13
(L - h'/2)/L
where: Em '-
All the
unit strain.inthe mortar
s~ortening
is assumed to occur in the mortar joints and therefore: 6-14
where:
En -
modulus of elasticity of the'mortar 6-11
1M 5-l300/NAVFAC P-397/AFR,88 c22 •
f m'-.
compressive'stress corresponding to the strain
E
In most cases.fm will be' greater than the ultimate. compressive strength .. of the mortar f m, and therefore cannot; exist·. ' Since· for walls of normal height and thickness each half of the wall undergoes a small rotation to obtain. the '. position shown in Figure 6-9c, the shortening of the diagonals o-mcand m,ncan be considered a linear. function of, the lateral displacement of point m. The deflection at maximum resistance Xl' at:· which a'compressive str~ss f m exists at points m, nand 0 can therefore be found from the following: " . Xl
~c
f' m
f' m
T
Xc
fm
EnE m
6-15a
or Xl -
(T - Xc) f' m
+
6-l5b
Xc
(EnEm) The resisting moment that is caused by a lateral deflection Xl is found by assuming rectangular compression stress blocks to exist at the supports (points 0 and n) and at the center (point m) as shown in Figure 6~lOa. The bearing width a is chosen so that the moment Ku is a maximum, that is, by differentiating Ku with respect to a and setting the derivative equal to zero, which for a solid masonry unit will result,in:. a -
0.5 (T - Xl)
6-l~
"
and the corresponding ultimate moment and resistance (Figure 6-10b) ,are equal to:
, and
, - 0.25 f'm[T ~lV
Ku
.
.
1,
• ~
. ',.
J
ru
- (2jh2)
f'm[T
,
.
.
"
.',
v
6-17
~"'"
-
,< ,
.
. •
,
- XlP
',;
J "
6-18
When the mid span deflection is greater than. Xl ·the expression for the, resistance as a function of the displacement is:
s»
6-19
·•
As the deflection increases the resistance is :reduced until r is equal to zero and maximum deflection Xm is reached (Figure 6-10b). Similar expressions can be derived for hollow masonry, units. However, the, maximum value of, a can not exceed the thickness of the flange width. ..i
r
,
6-9.2. Non-rigid Supports For the case where the wall is supported by elastic supports at the top and/or bottom, the resistance curve cannot be constructed based on the value.of the compression, force (af' m) which is determined solely on geometry of the wall. 6-12..
-' .
TK5-1300/NAVFAC P-397/AFR 88-22 Instead the resistance curve' is a function of the stiffness of the supports. Once the magnitude of the compressi?n force is determined, equations similar to those derived for the case of the rigid supports can be used. !
• -.".'.
•
-
.. ... -
'"\.,
__ 6-9.3. Simply ,Supported Walls
.Lf the supports ,offer no resistance to vertical motion,. the compression in the 'wall will be limited by the wall weight above. the floor plus any roo.f load which may be carried by the .wall. If .the wall carries no vertical loads, then the wa,ll,must be analyzed as a simply ,supported beam, the maximum resisting moment being determined by the modulus of rupture of the mortar. .
,
:
.f.
"
•,
,-
,
I, .
,
".
"
~('
6-13
'.-
,
J"
"
•
I
-
.
I-
,
"
,
,
.
.L
.,
,
,.
. ,'.'
.".
~
::"
.
-
.1.
I. .
'.'
.
.. '"
•
-. ,-t
,
"
.L
•
,
,
,
.·
.
,.
.t- .
"
0, -0
..
-
'
.
•
.
•
.
or
~
~
-
.
.
....
·
.
4 Bavs (0)10'·0": 40'·0"
..
FLOOR PLAN
rMetal Roof
Column
Rigid
Frame
Cone. Floor, 'Slab
AT- ROOF
Figure 6-1
,ATFL00R
Masonry wall with flexible support
6-14 .
-0 I
·0
,
.- -
., ,
,
1:.:-" ..
;
"
--
.
-- .
•
, ,.,-'
..
,
.
""".
I
-
."'" ..,
,
.
.
,
~
,
.~
,
..-
,
.
I
-
,
.-
i
"
...
,
.
..
.
.. . ".- .
•'<
, ~
.'!
.~
--_ _. --..
-
~
... _--, ..
"
.
..
j
.
.-
..
•.1"
~
1-0" 4 Bays @ 10'-0"= 40
FLOOR PLAN Cone. Roof Siab
~:t:·'f.:~~~J·
''''-:-:
.-. . -,- .-..:~'.:::
.
.:-,,~
'
Cone:'·
'j
'Becrr...
.
.,
,
Cane.
Floor Slab
.. . i
AT ROOF
,
0
I .
'0
.,
AT FLOOR
,Figure 6-2 ;; Mas.o~r.ywan wi th rigid support
6-15,
r-
,
0
0
-
..
..
..
...
v
,"
V "'\. r-,
•
.:-:~=. ..
:
".
'.
-,
i/
-,
..
., .
-... .
... i:: ',': ",2jL::
....
'
'::""
/'. ...
v '" k
'
--
'
.1 •• _
'.
.'
"
.....
.
.";
··;V \~ . .. .......... . ... ...... ',': ,\.~"
'
~~
.. .. ,. ..
, ..
~':-:'
Truss Type . Reinforcement
,, . '
,, .
'.
. .....
,
..
~.
,
'·,11':.
. .... '
...
" '.
,
~ , "
. .' ........
"
.
,
"
, :;.:
..
...
........ :
• , : .n,,;; . :
\
'
:
,.
"
'.
,:~
-,~..
Ladder Type Reinforcement
..
I:.. :'.:
.. .. .. . ..' ... '.
, .. , ', •... .... ....
:
u.,
"
'
": ':"~
~"~'
.
'"
.',
SINGLE WYTHE CMU
:' . ~ ', •.:.:
......
..
';'::':.'.,,,:. v,
.. : ....
:'
.',
.' V ,,'K.'"". ;'-:'.:
)'.
,.~
/'
.,', . .. ~
. ..
........ ,
'
."
•••• "';';'.:: "
,:-.
'.
'. -" '..
. > '. -::.
t ".
.. ' . . .. .. , . .
,:'
t.'·
>...-Tie" "
,.',.
......
"
DOUBLE WYTHE CMU:
.
Figure 6-3
'.
Concrete masonry walls
6-16
r-
.7..~., '"
-r ..-.....
V;:..:.........:.·
"
"
"', .;,L 1- ,
.'
'.
t
.
.'
, .",
v~
""',
JOINT REINFQRCED -MASOII!RY CON.STRUCT ION
, I
i
COMBINED JOINT AND CEL"i.- REINFORCED MASONRY
CONSTRUCTION
Figure 6-4 Typical joint reinforced masonry construction
6-17 '
....
~
.:;,: -.
"."~ ~,
~
.
..
PR~VISIONS f~~'~'EINFORCEMENT
THE. SPECIAL PLACEMENT AS SHOWN ; ARE' AVAILABLE IN MANY OF THE BLOCK CONFIGURATIONS. ILLUSTRATED IN FIG, 6 -7, ';,:'
.,
.~
.'
"'"
,~
.. "....
. .-
.
";.
~-
'<,';
Figure 6-5 '
'. (,,-
...
~ ... '.~
I :
,-~
.
Special ,masonry unit'for ..lt§e with reinforcing bars .'1' .;-.
'. :_~
•
. T.I
1',,' " . ' , .~
"~'"~!.
. :'.,
II
.
'·;t·
"
6-18,
,
,j ' ,
.:
Fi gure 6-6
. -;:.
Arching action of non-reinforced masonry wall
6·19
Double eo...,e.. nr
Jemb
&ull Nose
Pie..
" .~.,, :~.
... Full Cut Header
Floo,.
Solid Top
St~h.r
.' r~
•
('l eo... ) (In ~ Jlrus the _bo". unit!\ arY available in 4-nolTtinal heights)
Soffit FlOOf'
.JelTtb
St,.e+cher
Solid
Corne,.
Stretch.-
Figure 6-7
Solid B..iclo.
, ,•
;1
,'J " ~
hem or L.l,,+.1
Fl"'OQgecI Brick
Trough
Corne"
. Chennel
Typical concrete masonry units
6-20
$'\
Stretcher (Module ..)
'~'.
12" CONC. BLOCK
,
.
.'
.
. .".
ANCHOR STRAP
WI2x31
.
.' • . . .t:1.
(0) Masonry
Anchor Straps at Corners
CI2 WT ~ x 13.5 FOR EACH STRAP· TS 2x 2
CONC. BLOCK
..
..
.'
- . .' I' • •
•
".--4L "
.
,
.
.:
It
0.1815 .
t
DOOR
MASONARY ANCHOR, STRAP
DOOR FRAME
(b) Masonry Anchor Strap Detail at Door.
Figure 6-8
Connection details fof rebound and/or negative overpressures
6-21
. T-xc
.
0
m
::}
• .<=
,
+
.
~
m
'. -'=
-11
N
+
n
n
~
-'=
W (0 l
Figure 6-9
(b)
(el
Deflection of non-reinforced masonry walls
-.
6-22
I
B
..
'
•
.
-I.
(\j'
.:;0
"
(a) ARCHING
BEHAVIOR 'j
,
.L.'
,
-4-l-Xc
Xm
DEFLE,CTION
x-
(b) RESISTANCE - DEFLECTION
FUNCTION
Figure 6-10 Structural behavior of non-reinforced solid masonry panel with rigid supports 6-23'[:
•
TM 5-1300;NAVFAC P-397/AFR 88-22
Pr0p.erties of Hollow Masonry Units
Table 6-1
Nominal Width of Units (in)
/Equiv. Web Thickness (in)
Face-Shell Thickness, (in)
"
. 3 and 4
0.75
1.625
6
1.00
2.25
8
1.25'
2.25
10
1.375
2.50
12
1.500
2.50
,
, I
Taple 6-2
Deflection Criteria for Masonry Walls
,
Wall Type Reusable
Non-Reusable
Support Type
Support Rotation
,
One-way
0.5'
Two-way
0.5' .
One-way
1.0' , ,
Two-way
2.0'
), ......
.
,
, i'
,
.
.'."
6-24-'
. '
TK 5-l300/NAVFAC ,P-397/AFR,88-22
, ,
,,",'
"
,
., 1
.'
'.
"',
,
.
t
......
"
TabLe ' 6-3
r,r
Moment of Inertia of Masonry Walls .. .r-
Type of Unit
Width of Unit (in)
:
, "
•
~ln~) Ii II
.
,
..
.
Hollow
Moment of Inertia
..
3 4
, .
2.0 4.0 12.7 28.8
,
,
6
.
8
,
i
,
10 12
.
"
51. 6
,
83.3
'.
:< .
Solid
,
.
.
.. '
,
3, '4 '
,
,.
"
6
a':
,
. "
10 12 "
"
, » , -,
, "
",.' ....
6-25
." .'
2.7 5.3
18.0 '42.7 8).0 144.0
.
, 1
TH 5-l300/NAVFAC'P-397/AFR 88-22 PRECAST CONCRETE 6-10. Applications Precast concrete construction can consist of either prestressed or conventionally reinforced members. Prestressing is advantageous in conventional construction, for members subjected to high flexural stresses such as long span or heavily loaded slabs and beams. Other advantages of precast concrete construction include: (1) completion time for precast construction will be significantly less than the required for cast-in-place concrete, (2) precast construction will provide protection against primary and secondary fragments
not usually afforded by steel construction and (3) precast work is generally more economical than cast-in-place concrete construction especially when standard precast shapes can be used. The overriding disadvantage of precast construction is that the use of precast members is limited to buildings located at relatively low pressure levels of 1 to 2 psi. For slightly higher pressure levels, cast-in-place cpncrete.or structural steel constru~tion becomes the more economical means of construction. However, for even higher pressures, cast-in-place concrete is the only means available to economically withstand the. appli,.ed Load . . ." _ . . -. . , Precast structures are of the shear wall type, rigid frame structures being economically impractical (see the-discussion of connections, Section 6-16 below). Conventionally designed precast structures may be multi-story, but for blast design it is recommended.that they be limited to single stor~ buildings. Some of the most common precast sections are shown in Figure 6'-11. The single tee and double tee sections are used for wall panels and roof panels. All the other sections are beam and girder elements. In addition ,. a modified flat slab section will be 'used as a wall panel· around door openings. All of the sections shown' can be prestressed conventionaily' reinforced. In general though, for blast design, beams and roof panels are prestressed, while t · • columns and.wall panels are not. For conventional design; prestressing wall . panels and columns is advantageous'in tall mu~ti-story building, and thus of' no benefit for blast,resistant design which uses only single story. buildings. In fact, in.the design of a wall panel, the blast load is·from the opposite direction of conventional loads and hence prestressing a wall panel decreases rather than increases the capacity ot' section. - - _. .
'
."
. 'j
•
'
.
.
or
6-11. Static Strength of Materials 6-11.1. Concrete Generally the minimum compressive strength of the concrete, f c' used in precast elements is 4000 to 5000 psi. High early-strength cement is usually used in prestressed elements to ensure adequate concrete strength is developed before the prestress is introduced. 6-11.2. Reinforcing Bars Steel reinforcing bars are used for rebound and shear reinforcement in prestressed members as well as for flexural reinforcement in non-prestressed members. For use in blast design, bars designated by the American Society for Testing and Materials (ASTM) asA. 615, grade 60, are recommended. As only small deflections are permitted· in precast members, the reinforcement is not 6-26
.'
TH 5-l300i'NAVFAC P-397/AFR"88-22
stressed into its strain hardening region and thus the static design strength of the reinforcement is equal to its yield stress (fm - 60,000 psi). 6-11.3. Welded Wire Fabric
"
Welded wire fabric, designated as A 185 by ASTM, is used'to reinforce the flanges of tee and double tee-sections. In conventional design welded wire fabric Is sometimes used as shear reinforcement, but it is not; used for blast design, which requires closed ties. The static design strength f m, of welded wire fabric is equal to its yield stress, 65,000 psi, 6-11.4. Prestressing Tendons There are several types of reinforcement that can be used in prestressing tendons. They are designated by ASTM as A 416, A 421 or A 722, with A 416, grade 250 or grade 270, being the most common. The high strength steel used in these types of reinforcement can only undergo a maximum elongation of 3.5 to 4 percent of the original length before the ultimate strength is reached, Furthermore, the high strength steel lacks a well defined yield point, but rather exhibits a slow continuous yielding with a curved stress-strain
relationship until ultimate strength is developed (see Figure' 6-12). ASTM specifies a fictitious yield stress fp'y' corresponding to a 1 percent elongaThe minimum value of fpydepenos on· the ASTM des"ignation, but it ranges
tion.
from 80 to 90 percent of the ultimate strength, f pu' ".
-
.
.'
'
.
.'
.'
6-12. Dynamic Strength of Materials Under the rapid rate of straining of blast loads, most materials' develop higher strengths than they do ,when statically, loaded. M:.excep t Lon , is the high strength steel used in prestressing tendons. Researchers have found that there was very little increase in the upper yield' stress 'and'ultimate tensile strengths of high strength steels under dynamic loading.
.
.
The dynamic design strength is obtained by multiplying the static design strength by the appropriate dynamic increase factor. DIF, which is as follows: (a)
Concrete:
Compression Diagonal tension Direct shear Bond'
DlF-1.l9' DIF - 1. 00 DIF - 1.10 ·DlF - 1.00 .
(b)
Non-prestressed Steel Reinforcement; Flexure DIF Shear DIF
1.17' 1.00
1';.
(c) ,Welded 'Wire, Fabric: (d) 6~13.
'"
.
~~
Prestressed Reinforcement
DIF
1.10
DIF .. 1'.00'
.-
'
Ultimate Strength of Precast Elements
The ultimate strength of non-prestressed precast members is exactly the same as cast-in-place concrete members and as such is not repeated here. For the 6-27
TH 5-l300/NAVFAC,P-397I.AFR 88-22
ultimate strength of non-prestressed precast elements, ,see Chapter 4 of this manual.
6-13.1
Ultimate Dynamic Moment Capacity of Prestressed Beams
The ultimate dynamic moment capacityMu of a prestressed rectangular beam (or of a flanged section where the thickness of the compression flange is greater than or,equal to the depth of the equivalent rectangular stress block, a) is as follows:
Mu - ~s f ps (~
- a/2) +
Asf dy (d-a/2)
•
6-20
and .!.,
a -
-.
(~sfps + Asf dy )
6-21
. 85f' dc b
where: ultimate moment capacity total area of prestress reinforcement average stress 'in the prestressed reinforcement at ultimate load distance from extreme compression fiber to the centroid of the.prestressed reinforcement a, -
As -
.'
depth ,of equivalent rectangular stress block total area of non-prestressed tension reinforcement
dynamic design strength of non-prestressed reinforcement distance from extreme compression fiber to the centroid of the non-prestressed reinforcement f'dc b
dynamic compressive strength of concrete width of the beam for a rectangular section or width of the compression flange for a flanged section
The average stress in the prestressed reinforcement at ultimate load f ps must be determined from a trial~and-error stress-strain compatibility analysis. This may be tedious and difficult especially if the specific stress-strain curve of the steel being used is unavailable. In lieu of such a detailed analysis, the following equations may be used to obtain an appropriate value of f ps : For members with bonded prestressing tendons:
6-28
TM 5-1300jNAVFAC P~397/AFR·88-22
f ps -
f,+ 2-]
f pu Pp
B1
.
+
f'dc
. ,
df(i .y
(p-p')'
6-22
~f'dc
"
and
.>'
.
,.
p
p'
,
~sfb~
Pp
-
. 6-23
As fbd
6-24
A' s fbd
6-25
where: f pu -
specified tensile 'strength of prestressing tendon factor for type of prestressing
yp ' - _
0.40 for f py /f pu
~
0.80
/fpu'~
0.90
.t endon
,
0.28 for f py
"fictitious" yield stress of prestressing tendon corresponding to a 1 percent· elongation 0.85 for f'dc up to 4000 psi and is reduced 0.05 for each '1000 psi -Ln excess of 4000.psi ,', prestressed 'reinforcement ratio'
Pp
ratio of non-prestressed. tension reinforcement
p
ratio of' compression reinforcement
p'
total area of compression reinforcement
,. If any compression reinforcement is taken into account when calculating f ps then the distance from the extreme compression fiber·to the centroid of the compression reinforcement must be less than 0.15~and .(p-p')
Pp f'dc
.!,
~f'dc·
~,0,17
6-26
.• j
If there is no comp~ession reinforcemerit and no non-prestressed tension reinforcement,Equation 6-22 becomes: . " 7'
6-29 '.'
TH 5-1300/NAVFAC P-397/AFR',88-22
6-27
Pp
For members with unbonded prestressing tendons and a span-to-depth ratio less than or equal to 35:
6-28a and f ps
~
6-28b
f s e + 60,000
where: effective stress in prestressed reinforcement after allowances for all prestress losses For members with unbonded prestressing tendons and a span-to-depth ratio greater than 35:
6-29a and f ps
~
6-29b
f s e + 30,000
To .Lnsure against, sudden compression failure the reinforcement ratios for a rectangular beam, or for a flanged section where the thickness of the compression flange is greater than or equal to the depth of the equivalent rectangular stress block will be such that: (p
f'dc
dpf'dc
p')
s
0, 36B 1
6-30
When the thickness of the compression flange of a flanged section is less than the depth of the equivalent rectangu1ar'stress'b1ock, the reinforcement ratios will be such that
,6-31 f'dc PpWI Pw' p'w -
~f'dc
reinforcement ratios for flanged, sections computed as for p , p and p' respectively except that b shall be the width o~ the web and the reinforcement area will be that required to develop the compressive strength of the web only.
6-30
TM 5-l300/NAVFAC P-397/AFR 88-22
6-13.2. Diagonal Tension and Direct,Shear of Prest'ressed Elements' ,.'. ,.,
'.~ "'t JjJ';
«,
' ..")
....
Under conventional 'service loads, prestressed elements remain almost entirely in compression,' ~nd;hence are'permitted·it higher concrete shear 'stress than non~pre·stressed·;·el'einents: -Howeve r at~'qltimate Lo ads the effect of prestress.
is lost and thus no increase in shear capacity is permitted: The shear capacity of'a precast beam may be calculated using the equations of Chapter 4 of this manual. The'loss 'of the effect of prestress also means that d- is the actual distance to the prestressing tendon and is not limited to O.8h as it is in the ACI code. It is obvious then that at the supports'of an element with draped tendons, d and thus' the shear capacity are .gre at Ly reduced. Draped tendons also make it difficult to properly anchor shear reinforcement at the supports, exactly where it' is needed most. ,Thus it'is recommended that only straight tendons be used for blast'design. 6-14. Dynamic Analysis The dynamic analysis of precast elements uses the procedures described in Chapter 3 of this manual, Since precast elements are simply supported;' the resistance-deflection curve is a one-step function (see Figure 3-39a), The ultimate unit resistance for various loading conditions is presented in Table 3-1. As precast structures are subject to low blast pressures, the dead load ,of the structures become significant; and must: be t aken'
'.
,.
~
'
...
d
t
.~,
e. .•.
"'.'"
The elastic stiffness of simply'supported beams with various loading'conditions is given in Table 3-7, In determining the stiffness, the effect of cracking is taken into account by using an average moment of 'inertia la' as follows: "I: - (I' +'1 )/2 a g. F .'
.,1
where: Ig
moment of inertia of the gross section'
Ic
moment of inertia·of the cracked section ,
,-
For non-p re st res sed elements', the cracked moment' of inertia' can be det'e rmi.ned from Chapter 4', For prestressed', elements ,the moment of inertia of the cracked section may be approximated by:
.... 6-33 where n is the ratio of the modulus of elasticity of steel to concrete. The load-mass factors, used' to convert the mass of the actual system to the equivalent mass, are given in Table '3-12 for prestressed elements thel,!admass factor in the elastic range is used, An average of the elastic and plastic range load-mass· factors is' used' in the design of norr-pr'e's t r e s s ed elements. "
6-31
TH 5-1300/NAVFAC ,P-397/AFR 88-22 The equivalent single-degree-of-freedom system,is defined in term~ of ,its; ultimate resistance r u' equivalent elastic deflection XE, and natural period of vibration T W The dynamic LoadTs de f Lned by. its peak pressure!, and duration T. The figures given in~hapter 3 maY,be used to determine the response of an element in terms of, its -raax Lmum deflection >SU. and" the time to, reach maximum. deflection t m, '0
.
'
;',:
'
Recommended maximum qeflection criteria,for precast,,!,lements is as follows: ,
<1-)
t'
1:.
For .pres cressed flexural mem!>er~::'
.'
'''','
9 max S 2 0 or'#max~S 1, whichever governs
(2)
For, non-prestresse~,flexural ~e~bers 9 max S 2' or ~max S 3, whicheve~,governs
(3)
For compression members
~max
where
..
'
S 1.3
9 max
maximum support ratio
~max
maximum ductility ratio
..
'_:--
6-15. Rebound v.'.
, " " Precast elements will vibrate under dynamic 'loads, causing negative deflections after the maximum deflection has been reached'. The negative forces associated with these negative deflections may be predicted using,Figure,.3 c 268. ,
"~'
"
"!
6-15.1. Non-prestressed elements The design of non-prestressed precast elements for the effects of rebound is the same as for cast-in-place members. See Chapter 4 for a discussion of rebound ,!,ffects in concrete elements.
,.
6-15.2. 'Prestressed elements
In pr e s t re s sed elements, ,non-prestressed reinforcement must be added to what is the compression zone during the loading phase to carry the tensile forces of the rebound phase. The rebound resistance will be determined from Figure, 3-268. but in"no case will it, be less than one-half, of the resistance available to resist the blast load. The moment capacity of a precast
e~ement
in rebound is as follows:
a-/2)
6-34
.,
where:
Mu-
ultimate moment capacity in rebound .. "
~
total area of rebound ,tension reinforcement dynamic design strength of reinforcement 6-32
TM 5-l300/NAVFAC P-397/AFR'88 l22
d'
distance, from-extreme compression fiber,to the centroid'of;' the rebound reinforcement
a
0
depth of: the' equivalent rectangular stress' block "
'.
It is important to take into accounti the compression in the' concrete' due to prestressing arid reduce the strength available for rebound. For, a conservative design, it may be assumed that the compression in the, concrete due to - prestressing is the maximum permitted by theACI code,i'.e'. 0;45"f'c' 'Thus the concrete strength available for rebound is' ,
:
.,..
;;,. '.'
,
.~
-,
0.85f' dc-' 0.45fi dc/DIF - 0.47f'·dc':' ",.rr . . .
. ".
-,
..
6-35 r.f"~.,
'~'
A more detailed analysis may be performed to determine the actual concrete compression due to prestress. In either case the maximum amount of rebound reinforcement added will be
[ As ' S
(O"'~~: " ,
f dy
-
l6]
.'
[
.
"
I~'"
(j7000 - nfl' bd(87000 .-: nf + f dy)
J
1
~
.
•
'
'-,;. "l _ .
" 6-37
....'
'.
6-16. Connections 6-16.1: General
,
:
One of ,the fundamentaL difference's 'betweeri 'a cast- in-place concrete s t ruc tur'e.'and one consisting of precast elements is .the nature' of -connections between' members.
For ·precast concrete structures ; as' in 'the c as eco f steel -structures;
connections' can be detailed 'to tra;"slIlit gravity loads only', gravity and ' lateral loads, or moments in addition to' these 'loads. In general though, connectors of precast'members should·be designed so that blast,ioads are transmitted to supporting members through simple beam action. Moment-resisting connections for blast resistant structures would have to be ,quite heavy and expensive because of the relatively large rotations, and,hence induced stresses, permitted in blast design. ' .•' ," .. ' . ,. . In the de s Lgn cof connections ·,the -c'apac Lty 'reduction factor't/>, .Eor shear and bearing stresses on concrete are asIp r e sc r Lb ed by ACI code, i:e.: 0.8'5 and' O.,7 respectively. No capacity reduction factor is used for moment calculations and no dynamic increase factors are used in determining the capacity 'of a' connector. Capacity of the conne~tion should be' at least 10 percent ,greater than rhe .reac t tcn 'of' .che member being connec t e d vt o account for cheibr Lt t Lenes s of the connection. ',In "addition che ,failure' mechanism' should be cont.rol Led by: tension or bending stress of, the steel, and, therefore the pullout st rengch of ~
6-33 "
;
" ,6-36;:
where f is the compression 'in the 'concrete due to .p res t re ss tng and 'all ,the other terms have been' defined preViously. If available concrete ·strength is assumed to be 0.47f,d c' equation 6-36 becomes:
.
'
"
TH 5-1300/NAVFAC P-397/AFR' 88"22 .
the concrete and the -strength of the welds should be greater than the··steel strength. The following connections are standard for use in· blast design but they are not intended to exclude other connection details. Other details are possible but they must be able to transmit gravity and blast loads, rebound Loadsvand lateral .loads without inducing moments".'
,"j
:J'"
.
~..
~',
.
.r. .•!.
,
6-16.2. Column-to-Foundation Connection.,
... ~ 'OoOF!
cr"
<:-'--
j
,.
",
; .
'.
The standard PCl column-to-foundation connection may be used for blast design without modification. However anchor bolts. must be checked for tension due ,to rebound in order to prevent concrete pullout . , -. . 'L . - • ·1 I r .. 6-16.3. Roof ~lab-tocGirder Connection .¥~ -.'
....
.t
.1 ~
, '. I
• 't',
'~"
, .
; 1
Figure 6-13 shows the connection detail of a roof panel (tee section) framing into a ledger, beam. The bearing pads transmit gravity loads while preventing the formation of moment couples. The bent prate welded to the plate embedded in the f~ange of the tee transmits lateral loads but is soft enough to aeform when the roof panel tries to rotate. The angle welded to the embedded plate in the web of the tee restricts' the panel, through shear action, from lifting off the girder during the rebound loading. The effects of dimensional changes due to creep, shrinkage and ,relaxation of prestress should be considered in : this type of .cormec t Lon . " ' ' "
6-16.4. Wall Panel-to-Roof Slab Connection
.
.
The b~sic concepts employed in the .roof slab-to-girdei connection apply ~o the wall panel-to-roof slab connection shown in Figure 6-14. The roof panel instead of bearing on the girder, bears on a corbel cast with the tee section. The angle that transmits lateral loads has been moved from the underside of the flange to the top of the flange to facilitate field welding. "...., :'...." " 6-16.5. Wall Panel-to-Foundation Connection The wal} "panel In Fdgur'e 6,-15 is attached to' .the 'fo,undation by 'means of, angles welded to plates c~st in both the wall panel and ,the foundation., ~t is essential· to provide.a method of at t.achmenc to the foundation ,that :is capable. c s of taking base shear inany"di-rection; -and also a method of -Leve Ll.Lng and ...... aligning the wall panel. Non-shrinking. grout is used to fill the 'gap between the panel and the -founda't Lon so as to transmit the loads to the foundation.' ",
6-16.6. Panel Splice
" ...
{
l
"1
•
"\
Since precast structures are of the shear wall type, all horizontal blast " loads are transferred by diaphragm action, through wall and roof slabs to the foundations. The typical panel splice shown in Figure 6-16 is used fo~:. transferring the horizontal ,loads ,between panels. , _.'
.".
6-16.7. Re Lnf'orcemerit; Around Door",Openings
. ~. rl
I
.
~ £";
•
v., ..
A standard double ,tee section cannot be. used around a door opening.,~.Instead a special panel~must be fabricated .co satisfy the .requirements for the door '.. . );l.
6-34
:' ,'. "1
TM 5-1300/NAVFAC P-397/AFR 88-22 opening. The design of the reinforcement around the door opening and the door frame is discussed in Chapter 4.
6-35
J I t..
..
':'
'j
a:: <[
...J
:::>
... ... Ul
~
1/:" ~ ..;.',\31 ... .. . ...... "
.J
II)
:>
o
'':,::~.: "
Z <[
~
...a::o ...
o
.,
C)
. ~:" ':,j
...,
II>
c: ClJ
E
ClJ ~
ClJ
..., '"
II>
.• ~ -,
.: : ••.. ~,
.
'.::
',.. :: : .-. . .
.".
:.:: .. :.:
U ClJ
S-
p','
:
a.
......
c:
o
.J
E' E
o
u
Ul Ul ~
Ul Ul
..
~
....; - - - -
o '. )..... i: .
Ul ~
a:: Ul
e
> Z
• u
6-36
2068.4
300
.,
,
.
t·,
-.
..
200
-
f
V ....
-
--,
,
w
I
•
. r.
-. ,
"
"
··c
-
,
'. 1379.0
.., .'
"
,
. ,
!
-
(J)
en
~
•
,!
en
.
--
~
1723.7
,,
"
,.
-'"
lr I-
,
~-
250
-
,
.
-
--
--
..
-,
, H
•
100 "
;
,;,),
,
\
:
,
.-. ..
..
. \
, ;
..
-- -,
, ,
.. I
,
en en
,
I
150
1034.2
~
Z
::::>
-
689.5
344.7
2
3
4
U NIT STRAIN (%)
Fi gure 6-12' 'TypiCal st ress-st re't n curve
6·37
for
lr II-
.
50
o
w en
,>
.
,
e
::;:
hi gh "strength wi re
Embedded Plate.
Precast Roof F\:Inel
\Ty,p.)
Embedded -:f"'----~:_ll.,.;.:, Plates
Loose Angle (Typ.)
.. .:. ..
. ..,. .
':
:,t'o"~'
recast Girder
'--~Precast
Column
Fi gU,re6-13 . Roof s 1a~-to-gi rder connecti on
6 - 38 c
LOOSE
PLATES CAST IN TEES
ANGLE
,
=== ====':iI== ,.1: • .,J
.....= =
t::=
~
-
= = ..=== .~.
.. '.
~
~
" LOOSE' ANGLE. '
\
"
PLATES CAST IN TEE!), ..
-.'
. .
...1'-"~
,Double Tee Wall Panel ~
,
Bearing Pad
Double Tee Roof Panel
' "..
"
,
.',
.",
Figure 6-14 Typical wall 'panel-to-roof slab connection
.....
6-'39
,
,
.
Tee'Section ~
...
Plate Cast. In Tee Section
Angle ,
.
Angle Cast in Foundation Stiffener Plate
41
o
o
. Q.
en
, Foundotlon .. ",
..'.. :~.·o::.. .: ,
.'
'
i., -: .:
0.
Anchor Bar for Angle
"/
Figure 6-15 Wall panel-to-fourdation connection
....
i' .•..
,
0
..
'
'0
,;.
I.
.:...
,
,
,.
,
•
:.' ;;. '\, "
. ""',,' ~.
~
,
"
0
,.,-
,
,,,
,'10 '10\ ,
"
'
:,At.."" .
.
'J'
,
'--
. .
,I'
r--
,
,
z
//
~
. 'I.~ '/,
.
. ~ .r
~,'Io
,
// ~,
~~
",,'.
Looie Plate-:-
,
.
.
~
III~ ~ II~; " :-..
I
'"
'q
I. //'
,
VI',
,
•
I'"l-- Angle .
"
"
""":-.. """~
0'
.
~-,
(,'/
'
" v
,Vo
" PLAN' .,
"
r .
,1
Angle Cast in .... ;Tee Section (Typo)
Of
;"
I
, Anchor Bar for Angle (Typo)
"
t
~.
,
Caulking with Back-up " Material
SECTION A-A'
,.
:.
,
.
~
"
;'
....
'0
, -
•
J
." '~,.'
Figure,,6~16,
.'
...
I- ,.' '
';
,Typical pane I, splice .
-. . •
6-41'
0
-'
...
'
TM 5-1300/NAVFAC'P-397/AFR 88-22
SPECIAL PROVISIONS FOR PRE-ENGINEERED BUILDING 6-17. General Standard pre-engineered buildings are usually designed for conventional loads (live, snow, wind and/or seismic). Blast resistant 'pre-engineered buildings are also designed in the same manner as standard structures. However, the conventional loadings', which are used for the latter designs, are quite large
to compensate for effects of blast loads. ings,
pre~engineered
Further, as with standard build-
structures, which are designed for blast ,"'are designed
elastically for the conventional loadings with the assumption that the structure will sustain plastic deformations due to the blast. The design approach,will require a multi-stage process, including:' preparation of general ,layouts and partial blast designs by the design engineer; preparation of the specifications, by the engineer including· certain features as recommended herein; design of the building a~d preparation of shop drawings by the,pre-engineered building manufacturers; and the final blast evaluation of the structure by design engineer utilizing the layouts on the previously mentioned shop drawings. At the completion of the' analysis some slight modifications in building design may be 'necessary. How~vei, if the following procedures are 'used, then the required modifications will be 'limited and in some cases eliminated for blast overpressures upward to'2 ?si. 6-18. General Layout The general layout of pre-engineered buildings is based on both operational and blast resistant requirements. Figure 6-17 illustrates a typical general layout of the pre-engineered building'. The general requirements for structural steel, concrete, wall and roof coverings and connections are given 'below. 6-18.1, Structural Steel In order for a pre-engineered building to sustain, the required blast loading, structural steel layout 'must conf'o rm'
.
,
I,
The maximum spacing between main 'transverse 'rigid frames (bay width) shall not exceed 20 feei:. ;~.
2.
The maximum spacing between column supports for rigid .. fr~mes shall not exceed 20 feet while .the overall height of frames shall be 30 feet or less.
3.
Slope of'the roof shall not exceed four ,horizontal ,to one vertical. ' However, the roof slope shall be as shallow as physically possible and be in compliance with the requirements of,the Metal Building Manufacturers' 'Association.
4.
Spacing between girts shall not exceed 4 feet while the space between purlins shall not be greater than 5 feet.
5.
Primary members, including frames and other main load carrying members, shall consist ..ofhot.rolle'd ,structural steel shapes. The shapes must be doubly-symmetrical'and have a constant depth., They may be wide· flange sections, I-sections, structural tubes, or 6-42,
TM 5-l300/NAVFAC
88-22'.
~-397/AFR
welded shapes built-up from hot r.olled,steel sheet, strips or plates. Secondary structural framing, such as girts, roof purl ins , bridging, eave struts and other miscellaneous secondary framing, may consist of either hot rolled or ~old-formed structural .steel. All main secondary members (purl1ns" gir,ts, etc.) shall be doubly-symmetrical. sections of constant deptQ;(e,g. wide flange ,,"V' -shaped, structural ,t~bing). ,,:'
,.
6.
Primary structural framing connections shall be either shop welded or bolted or field bolted assemblies. ASTM A 325,bolts·with appropriate nuts and washers shall be used for connesting of all primary members; whereas secondary members may ,use bolts conforming to ASTM A 307. A minimum of two bolts shall.·be used for each connection while bolts for primary and secondary members shall not be less than 3/4 and li2-inch in diameter, respectively,
7.
.Ba~e.plates
for columns shall be rolled and set,~n grout bed of t-inch minimum thickness. ASTM A 307 steel .bolts shall be used to anchor all columns.
6-18.2. Foundations Concrete floor and 'foundation slabs
shall~be
monolithic in
constr~ction
and
shall be designed,to transfer all horizontal and vertical loads from the preengineered superstructure to the foundation soil.
Minimum slab thickness
shall be 6 inches, with edge beams thickened to meet local frost conditions, 6-18.3. Roof and Walls , .. ,
l
',"
Roof and wall coverings must meet the following requirements: 1.
Roof and wall coverings shall conform to ASTM A. 446, G 90, have' a· minimum depth of 1-1/2 inches corrugation and have a material thickness of 22 gauge. ~'
2. '
Conventional side .laps are not, usually sufficient to resist the " ' effects of blast loads,' The construction details, required to strengthen'those joints .depend upon the type of decking employed, Chapter 5 gives the required panel-to-panel attachments for various types of decking.
3.
Insulation retainers or sub girts shall be designed, to' transmit· all 'external loads (listed. below) which act on the metal cover to the structural steel framing,"
4.
Roof' and wall liners shall be a minimum of 24 gauge and shall,be formed to prevent waviness " distortion. or f a LLure. as a result ·of the impact by external loads'.
6-18.4. Connections for Roof and Wall Coverings The connections used in a blast resistant s~ructure are especially critical. To ensure full development of structural steel and the roof and wall panels, connections must meet the following criteria: 6-43
TM 5-1300/NAVFAC
P~397/AFR
88-22
1.
Fasteners for connecting roof and wall coverings to structural steel supports shall be designed to support the external loads (listed below) and shall consist of 'either self-tapping screws, self-drilling and 'self-tapping screws, bolts 'and nuts, selflocKing rivets, self-locking bolts, end welded studs, or welds. " Fasteners of covering to structural steel 'shall 'be 'located at ,valleys of the covering and shall have a minimUm of one fastener per valley.
2.
.. Fasteners which do not provide'positive locking such as selftapping screws, etc. shall not be used at side-laps,and for , fastening accessories to panels. At least one fastener for side laps shall be located in each valley and at a 'maximum spacing 'along'the valley of 8 inches. J
, , '
3,
Self tapping screws shall not have a diameter smaller than a no, 14 screw while the minimum diameter of a self-drilling 'and selftapping type shall be equa Ll t o or greater -tihan a no, '12 screw. Automatic welded studs shall be shouldered type and'have a shank diameter of at least 3/16 inch, Fasteners for use with power actuated tools shall have a shank diameter of not h'ss!'than 1/2 inch. Blind rivets shall be stainless steel type and have a minimum diameter of 1/8 inch,' : Rivets shall be'threaded stem type if used for other' than fastening 'trim and' if of-the hollow type shall have closed ends, Bolts shall not be less than'l/4 inch in diameter' and will,be provided with suitable nuts' and washe r s ,' If suction andlor rebound loads dictate, provide oversized'washers with a maximum outside diameter of 2 inches or a 22 gauge thick metal strip along each valley.
6-19. Preparation of'Partial'Blast'Analysis ,
,
A partial blast analysis of a pre-engineered building shaI l.vbe performed by the design engineer, This analysis shall include the determination of the minimUm size of the roof and wall panels which is included in the' design' specifications and the design of the building foundation and floor slab. The foundation ,and floor slab shall be designed monolithically and have a minimum thickness,as previously stated. 'The slab shall be designed for a foundation load equal to either 1.3 times the yield capacity of the building roof equivalent blast load or the static roof and floor loads listed below, Quite often the'foundation below the'building columns must be thickened'to distribute the 'column loads. For the blast analysis of ,the building'foundation and floor s Lab , the dynamic capacity of the soil" below the .f'oundat Lorr. slab can conservatively be assumed to be equal to twice the static soil capacity, The resistance of the roof of the building can be determined in'accordance with the procedures given in Chapter 5. The front panel o~thebuilding is designed in the same manner as the roof panel. "The blast loads for determining the capacities of the roof and wall panels can be determined from Chapter 2,
6-20. Pre-Engineered Building Design ~,
-
~.
Design of the pre-engineered building: shall be performed by the pre-engineered, building manufacturer using static loads and conventional stresse~. 6-44
1M 5-l300/NAVFAC P-397/AFR 88-22
Conyentional stresses' are listed in "Specification for the Design. Fabrication and Erection of Structural Steel for Buildings with Commentary". Static design loads shall be as follows: 1.
Floor live loads shall be as specified in the report titled "American National Standard Building Code Requirements for Minimum Design Loads in Buildings and Other Structures" (hereafter referred to as ANSI) but not less than 150 pounds per square foot . .
\
2.
. 3. 4.
.
.
Roof live loads shall be as specified ANSI. Dead loads are b~s~d on the materials of construction . Wind pressure shall be as computed in accordance with ANSI for exposure ;·C·..· ami a wind speed of 100 miles per hour. .
~.
"
5.
Seismic loads will be calculated according to the Uniform Building Code for the given area. If this load is greater than die computed wind pressure. than the seismic' load will be substituted for wind load in all load combinations.
6.
Auxiliary and collateral loads are all design loads not listed above and include suspended ceilings,
insula~i?n,
electrical
systems. mechanical systems, etc. Combinations of design loads shall include 'the following (a)' dead loads plus live loads; (b) .dead loads plus wind loads. and (c) 75 percent of the sum of dead. live and Wind. loads. 6-21. Blast Evaluation of the, Structure Blast evaluation of the structure utilizing the shop drawings prepared in connection with the above design shall be performed by the design engineer. A dynamic analysis which describes the magnitude and direction of the elastoplastic stresses developed in the main frames and secondary·members as a result of the blast loads, shall be performed-using the 'methods described in Chapter 5. This evaluation should be made at the time of the shop drawing review stage. . 6-22. Recommended Specification for 'Pre-Engineered Buildings Specifications for pre-engineered buildings shall be consisterit with the recommended design changes set forth in the preceding Section. These example specifications are preserited using the Coristruction'Spe~ificationInstitute (CSI) format and shall contain as a minimum the following: . 1. APPLICABLE PUBLICATIONS. The following publica~ions of the issues listed below. but referred 'to thereafter by basic de s Lgna t I'on only. forma part of this specification to extent indicated by the reference thereto: , the . 1.1
American Society of Testing and .Materials (ASTM), A 36
Structural Steel
6-45
TK '-1300/NAVrAC P-397/Ara 88-22
..
A 307·80
Standard Specification for Carbon Steel Externally and Intarnally Threaded Standard Fasteners
A 325
Standard Specification for High Strength Bolts for Structural Steel Joint' Including Suitable Nuts and Plain Hardened Washers
A 446
Specification for Steel Sheet, Zinc Coated (Ga1vanized)'by the Hot-Dip Process, Physical (Structural) 9uantity
A ,501
Standa~d Specification for. Hot-Formed Welded and Seamress C~rbon St~el Structural Tubing
A 529
Standard Specification for Structural Steel with 42,000 psi Minimum Yield Point .
A 570
Standard Specification for Hot-Rolled Carbon Steel Sheet and Strip, Structural Quality
A 572 1.2
Specification of High-Strength Low-Allow Colum,bium-Vanadium Steels of Structural Quality
American Iron and Steel ,Institute (AISI) Specification for the Design of Cold-Formed Steel Structural Members and Commentary ,
1.3
American National Standards Institute (ANSI) A58.1 Minimum Design Loads for Buildings and Other Structures
B18.22.1 1.4
Plain Washers
American Institute of Steel Construction (AISC) Specification for the Design Fabrication and Erection of Structural Steel for Buildings with Commentary Research Council on Riveted and Bolted Structural Joints (RCRBSJ) Specification for Structural' Joints Using ASTM A 325 or A ,,90 ,Bolts,
• 1. 5
American Welding Society (AWS) ,~
J
01.1 Structural Welding Code 1.6
Metal Building Manufacturers' Association (MBMA) Metal Buildings Systems Manual
1.7
Uniform Building Code
6-.46
TK 5-1300jNAVFAC
2.
GENERAL.' ""(
-,
:,
','
1
P,397jAf~
88-22
"
. '.
l'~:'
2.1 This section covers the manufacture and erection of pre.-... engineered metal structures. The structure manufacturer shall be regularly engaged in the fabrication of metal structures. ',~ :2.2· ...The structure shall include ,th~ ,rigid Erami.ng , which are spaced at a. maximum o f- 20 feet .on .cerrte r- roof and 'wall cove r Lng , trim, ,closures " and accessories .as ·indicated,on the draw1ngs:.. Minor. a Lt.er a t i.ons in dimensions shown on the drawings will be considered. in .order: to.,comp1y with the manufacturer's standards building system, provided that all minimum clearances indicated on the' drawings are maintained. Su~h changes shall be submitted for review and acceptance prior to fabrication. 1
. 2.3 Drawings shall indicate. extent .and general as sembIy details of the metal roofing and sidings. Members and connections not indicated on the drawings shall be designed.by the Contractor·,in.;accordance with the manufacturer's standard details. The Contractor shall comply with the dimensions, profile limitations, gauges and fabrication. details shown on the drawings. Modification' of'detai1s will be permitted ,only when approved by the Owner. Should the modifications proposed by the Contractor be accepted by the Owner, the Contractor ,shall be fu11y·responsib1~ for ~ny re-design and redetailing of the building construction effected. -:', ,"
3.'
... J ':
DEFINITIONS. '
3.1 Low Rigid Frame. The building shall be single gable ,type with the roof slope not to exceed one on four. '.
3.2
Framing.
3.2.1 Primary Structural Framing ...,The primary structural framing includes the main transverse frames and other'primary load carrying members and their fasteners. "
.3.2.2 Secondary Structural. Framing. The secondary structura1 framing includes the girts, Foof pur1ins, bridging, eave struts, and other miscellaneous secondary framing members . and their fasteners... ' - '
.
.. , 3.2 ,3·.. Roof . and~Wall .covarIng , ',TI1e:roof and wall covering includes the exterior ribbed metal panel having a ..minimum·depth of one and one-half inches, neoprene closure, f~steners and s~alant. , "v
"
',',
" 3,3:
Building .Ceometry .
"", ,,",~'f'
'_'I
e } .• ~'
.'
"),
.3;3.1 Roof Slope.The~roof of .the bu Ll'dfng i shal L have a maximum slope not to exceed onev on ·four .. •. " '.t.r ,. '.,
';'-
3.3.2 Bay Spacing ..:. The.·bay·,spacing· shall -, not exceed 20 feet. t't'
', ..
-, '1.
:- ','
I
.'
, "
"""
r•
• 3.4 '~I CoLumnt.Shape .•. 'Main frame columns shall .:be doubly symmetrical members of constant, depth;" tapered .coLumns will not ,be permit,ted. ,
.........
,-
TH'S-1300/NAVFAC P-397/AFR'SS-22' 3.S Calculations. The Contractor shall submit for'review complete design calculations for all work, sealed by a registered professional engineer.·
:, .. .,
,,-"
,~-
".
STRUCTURAL DESIGN.
4.
'. "
4~1 Structural Analysis. The structural analysis of 'the primary and secondary framing' and covering shall be b'a sed on;'linear elastic behavior
and shall accurately z e f Lec t; the final 'configuration, o"f the structure" and a11-
tributary design' loadings"
-,'
,;" ,-
.',
Basic Design Loads.
-'-.'
"
4.2.1 Roof Live Load. Shall be applied to the horizontal roof p roj ec tLon . "Roof live loads shall be:
,-
Oto 200 'square feet ,trIbutary' area "
20 psf
,
linear variation, 20
'200 'to' 600 square feet tributary 'area c' , ps'f to 12 -p s f
.
, ..~
over 600 square 'feet tributary area
- 12
psf
4.2.2 WInd Pressure: Wind design loads shall be computed In accordance wIth ANSI' ,ASS. 1 for exposure "C" and a ba s Lc wInd'speed,of 100 mIles per hour. " t.;
,
.
4.2.2.1
Typical Wind Loading .
'As'shown on
drawings (Figure 6-18). 4.2.2.2 Wind Loadiqg at BuIlding Corners. shown on the drawings (Figure,:6-'18Y:· ,. ',.
j
4.2.2.3
;,
Wind Loading on Girts.: As'shown'on
.
drawings (Figure 6-18). Areas. ture.
As
'
,4.2:2.4 Wind Loading on Purlins and Roof Tributary As shown on 'drawings -'(Figure 6-18): As shown on
4:2.2;S,Wind'Loading"for Design of Overall Struc6,;18) ~ . .. -
drawing~ _(Fd gure
r
4.2.3 Auxiliary,arid Collateral Design Loads. Auxiliary and collateral design loads are those'loads other than,the:basic design live, dead, and wind loads'; which the building shall safely withstand, such as ceilings, insulation, "electrical,' mechanical, 'and "p LumhLng systems, and building equipment and supports: . " ; ,;' ': . 4.3 -App l i cat Lon- of Des ign Loads .. 4.3.l'Roof Live Load and Dead Load:
The ro~f live load (L),
andvde ad load (D), shall be cons Lde red- as . a uniformly distributed . loading
acting vertically on the horizontal ,projection of'the,roof. 4.3.2 Snow Loads.
appli'cation of 30 psf due to snow loads .. 6-4S:
TIl 5-1300/NAVFAC P-397/AFR 8,8-22 , 4.3.3 Win,1' LOAds (1.1):' Application of forces 'due to wind shall conform to the latest ANSI A58.1 '
' .• '
4.3.4 Combination of Loads .• The following combinations of loads shall be considered in the design of all.me~bers of the structure: 0'+ L 0+ 1.1
.75 (0 + L + 1.1)
, ,
4.4
.
Deflection Limitations. ,
4.4.1 Structural; Framing. cThe primary,and secondary framing members shall be so, proportioned that 'their maximum calculated roof live: load deflection does not exceed' 1/120 of the ,span; "
.-
5.
STRUCTURAL 5:1
FRAMING~
General ,
'.~
,
5 .1.1 'All hot rolled structural shapes ana structural. tubing shall·have a minimum yield point of 36,000 psi in conformance 'with ASTM A 36 or A 501.
All hot rolled steel plate, strip and'sheet used in the fabrication
of welded assemblies shall conform to the requirements of ASTM A 529, A 572, Grade 42 or A 570 Grade "E'!, 'as applicable. All hot. rolled 'sheet and strip used in the fabrication of 'cold-formed members shall conform 'to .the requirements of ASTM A 570, Grade"E" having a minimum yield strength of 50,000 psi. Design of cold-formed members shall .be ,in accordance with the' AISI specifications.
;-
'..;
5.1.2 The minimum thickness ,of 'framing members shall be: ,. Cold-~ormed seco~dary framing'members Pipe or tube columns Webs of welded built-up members Flanges 'of welded built-up members
Bracing rods v..
I,
,18 gauge 12 gauge, 1/8 inch 1/4 inch 1/4 inch
5.1.3 All framing members shall 'be .f'ab r Lc a t ed for bolted. field assembly. Bolt holes shall be punched or drilled only. No burning-in of holes will be allowed. The faying surfaces of, all bolted connections shall be, smooth'and free from burrs or distortions. Provide washers under head and nut of ' all bolts. Provide beveled washers to match sloping surfaces'as required. Bolts shall be of type specified below. Membeisshall be straight and dimensionally accurate.
5.l.4'All welded connections shall be in conformance with the STRUCTURAL WELDING CODE 01.1 ,of, the American Welding Society. The flangeto-web welds shall be one side'contin~ous submerged'arc fillet welds. Other. welds shall be by the shielded,arc process . .. '5.2. ' Primary Structural Framing; 5.2.1 The primary members shall be constructed of doublysymmetrical, hot rolled structural steel shapes or doubly-symmetrical built-up
TM 5-1300/NAVFAC' P-397/AFR'SS-22
mem~ers
of
const~nt
depth; welded from hot: rolled steel sheet, strip or
plates.
....
5.2.2 'Compression flanges shall be laterally braced to withstand any· combination of. loading. '.,5.2.3 Bracing system shall be provided to adequately transmit all' lateral forces on the building to the foundation. 5.2.4 All bolt connections of primary structural framing shall be made using high-strength zinc-plated (0.0003 bronze zinc plated) bolts, nuts, and washers conforming to ASTM A 325. Bolted connections shall have not less than two bolts . . Bolts shall not. be ..less.• than' 3/4 inch diameter. Shop. welds and field bolting are preferred. . A:ll field welds will· require. prior approval of the Owner. Installation of fasteners shall be by the..turn- d • of-nut or load-indicating washer method in accordance with the specifications for structural joints of the Research Council on Riveted and Bolt~d Structural Joints.
,.
5.3 Secondary members may be constructed of either hot rolled or cold-formed steel .. Purlins and girts shall be doubly 'symmetrical sections of constant depth.and they may be 'built-up, cold-formed. or hot rolled· structural shapes. '.
:.
" .
5.3.1 Maximum spacing ,of.. roof purlins and wall girts shall not· exceed 5 feet._ . 5.3.2 Compression flanges of purlins and girts shall be. laterally braced to withstand any combination of loading. .. 5.3.3 Supporting,lugs shall be used to ·connect the purlins and girts to the primary framing. The lugs shall be designed to restrain the light gauge sections from tipping or.warping at their supports. Each member shall be connected to each lug by a_minimum of two fasteners. 5,3.4 Vertical wall members not subjected to axial load, e.g. vertical members at door openings, shall be constant depth sections. They may consist of hot .rolled or cold-form steel. They shall be either built-up, cold-formed or hot rolled "C" .or "I" shapes. 5.3.5 Fasteners for all secondary framing shall bea minimum of 1/2. inch diameter (0.003 zinc plated) bolts conforming' to ASTM A 307.' The fasteners shall be tightened to' SNUG TIGHT. condition. Plain washers shall conform to ANSI standard B18. 22.1. ' . .' 6.
ANCHORAGE.
6.1 Anchorage. The building anchor bolts for both primary and secondary columns shall conform to ASTM A 307 steel and shall be'designed.to resist the column reactions produced by the. specified design loading. The quantity, size and location of anchor bolts shall be specified and furnished by the building manufacturer. A minimum of two anchor bolts shall' be used with each column.
.. , 6-50
TN
6.2. to
ASTM~
7.
Column Base Plates.,
5~1300/HAVPAC
P-397/APR 88-22
Baseplat.es for columns shall conform
36 and shall be set on a grout bed: of 1 inch
minimum,thic~ness.
ROOF AND WALL COVERING.
7,,1 Roof and wall panels shall conform to zinc-coated steel, ASTM A 446, G 90 cpating ,.designation. Minimum qepth of .each panel corrugation shall be 1-1/2 inches and shall have a minimum material thickness of 22 gauge. The minimum yield strength of panel material shall be 33,000 psi. Wall panels shall. be applied with the longitudinal ,configurations, in ,the ",ertical position. .Roof panels ,shall be applied with the longitudinal configuration in direction of the roof slope. Side laps of roof and wall panels shall be fastened as shown on drawings. End laps, if required shall occur at structural steel supports and have aminimu~ length of 12 inches.
7.2
Insulation.
siding shall be
suppli~d.and
,7.2.1 Semi-rigid insulation for, the',preformed roofing and installed by the preformed roofing and siding
manufacturer. 7.2.2 Insulation Retainers. Insulation retainers or sub girts shall be designed to transmit all external loads (wind, snow and live
loads) acting on the metal panels to the structural steel framing. retaine~s
The
shall be capable of transmitting both the direct and suction loads. 7.3
Wall and Roof Liners.
minimum of 24 gauge.
Wall and roof liners shall be a
All liners shall be formed
o~
patterned to prevent
waviness, distortion or failure as a·result of the impact by external loads.
7.4
Fasteners.
Fasteners for roof and wall panels shall be
zinc-coated steel or corrosion-resisting steel.
Exposed fasteners shall be
gasketed or have gasketed washers of a material compatible with the covering to waterproof
~he
fastener penetration.
Gasketed portion of fasteners or
washers shall be neoprene or other elastomeric material approximately 1/8 inch thick. 7.4.1 Type of Fasteners.
Fasteners for connecting roof or
wall panels to structural steel supports shall consist of self-tapping screws,
self-drilling and self-tapping screws, bolts, end welded studs, and welds. Fasteners for panels which connect to structural supports shall be located in
each valley of the panel and with a minimum of one fastener per valley while at end laps and plain ends, a minimum of two fasteners shall be used per
valley.
Fasteners shall not be located at panel crowns.
7.4.2 Fasteners which do not provide positive locking such as self-tapping screws or self-drilling and self-tapping screws shall not be used at side laps of panels and for fastening accessories to panels.
Fasten-
ers for side laps shall be located in each valley of the overlap and positioned a maximum of 8 inches on center.
7.4.3 Screws 'shall be not less than No. 14 diameter if selftapping type and not less than No. 12 diameter if self-drilling and selftapping type. 6-51
TN 5-1300/HAVFAC
P~397/AFR
88-22
7.4:4. Automatic end-welded studs shall'be shouldered type with a shank diameter of not less than 3/16 inch with cap and nut for holding' the covering against the shoulder. 7.4.5 Fasteners for use with power actuated tools shall have a shank diameter of not less than- 1/2 inch. Fasteners for securing wall panels shall have threaded studs for attaching approved nuts or caps.7.4.6 Blind rivets shall be stainless steel with l/S· inch nominal diameter shank.- Rivets shall be threaded stem type if used for other than fastening of trim. Rivets with hollow stems shall have closed ends. 7.4.7 Bolts shall not be less than 1/4 inch diameter, shoulders or plain shank as required with proper nuts. 7.4.8 Provide oversize washers with an outside diameter of 1 inch at each fastener or a 22 gauge thick metal. strip along. each valley of the panel to-negate pull-out of'the panel around the fasteners.
,,
,
"
6-52
.
- ,!
-, e
, 0)
I
,j~ ..
~'
1
".-
,
,,'::::J I, '
"
:./.i.~
I,
,
0
It
"
-,
':r
'.
I
~~
.... 1') ~"
,-'PI:..
,>q'l ~~ ''0..,(i,] 'i ~
, '" ~:; \ 111::",
~",
\.;
"
~~
>
"
,
- t
~
~ C)
f:
' 'if!3 ,;
". '7
...
'"
1 ' ,"
'
"'i
.
=F "f'
II
r; 01 .
.
'a!'I
>:
-
1
I
-.~.
;,11
{,
~
I
4-
o
..,::I
o
c-,
'"
.,
! '! ~
Z&
or ',; r ,
',; .
I' ji
8 ''',: f ~,'
I ~1
~
.
I, ; 0!'
~ ~; ;:,
, '., ..
11-
...
O"!_L.r-••I'
,
(,
,
6-53 ,
8····
,
\
1-
I~O
I. FLOOR LIVE LOAD: 2.ROOF LIVE LOAD
20
"-
.. 18 Do Q
~
...J
16
pII
"
""
14 12
200400600 SUPPORTED TRIBUTARY AREA,I," 3. ROOF SNOW LOAD: 30 pII 4. DEAD LOAD: AS PER MATERIALS USED 5. WIND LOADS: WIND PARALLEL OR PERPENDICULAR TO ROOF RIDGEIBASED ON 100mph WIND) A. WINDWARD/LEEWARD AND ROOF PRESSURES (pIll
.
e , WHEN TRIBUTARY SUPPORT AREAS> 200 ft' .
32.'
~ --
26
O
i
32
~
FOR DESIGN OF MAIN FRAMES AND OTHER INDIVIDUAL MEMBERS, USE VALUES SHOWN.
--
26
b.WHEN TRIB'UTARY SUPPORT AREA
:s 200 ft.
--43'0'143---
35
FOR DESIGN OF MAIN FRAMES, USE VALUES SHOWN FOR TRIBUTARY AREA . > 200 fl'. FOR DESIG N OF OTHER 43 INDIVIDUAL MEMBERS, USE VALUES SHOWN PRESSURES (psI)
35
.
43
B. SIDEWALL
P
~-::~~,.,.,.,...J.
FOR DESIGN OF INDIVUAL MEMBERS: WHEN TRIBUTARY SUPPORTED> 200 ft" USE P , 3 2 ' , WHEN TRIBUTARY SUPPORTED < 200 II", USE P'43 -
P --
~O
C.SIDESWAY PRESSURES (psI)
.
';
r1"l.'
FOR DESIGN OF MAIN FRAMES FOR TRIBUTARY SUPPORTED AREAS, TA, LESS THAN, EQUAL TO, OR GREATER THAN,200 ft',
~ .~ • WIIO I I .
D. LOCAL PRESSURES' SEE FIGURES IN 'B' AND OF APPLICATION REGION· '1 : 50psl SUCTION REGION 2: 105psl SUCTION REGION 3: 42 psI SUCTION.
"w" IS
'c'
FOR REGIONS
FOR DESIGN OF DECKING CONNECTIONS AT REGIONS CITED
LEAST WIDTH OF ENCLOSED AREA
NOTE: FIGURES DEPICT WIND PERPENDICULAR TO RIDGE. FOR WIND PARALLEL TO RIDGE, USE SAME VALUES.
6.
LOADING
COMBINATIONS:
A. 0 B. 0 l' L C. 0 t- W D.0.75(Dt-L+W)
Figure 6-18
WHERE 0' DEAD LOAD . L' LIVE LOAD W' WIND LOAD
Recommended pre-engineered building design loads ·6·54
TM 5-l300/NAVFAC P-397/AFR 88-22
SUPPRESSIVE SHIELDING 6-23. General This manual.presents methods for the design and construction of conventional reinforced concrete arid steel protective faciltties which provide adequate
..
safety for hazardous operations such as munitions loading, maintenance,
renovation, or demilitarization. Such safety considerations include the utilization of conventi~nal protective barrier~, total containment construction, or t.he use of separation.distances"or isolation of the, specific opera-, tion from other parts of" .the f ac i Li ry cusLng . appropriate quantity distance specifications .. Howeve r , an a l t.ernat Ive available to'. the designer of these. r facilities is the .use of suppres~ive shield~ng as outl~ned in HNDM 1110-J-2,. "Suppressive Shields Structural Design .an? Analysis Handbook,", 18. November 1 9 7 7 . , . ., A suppressive shield is a vented steel enclosure which controls or confines the hazardous blast, fragment, and flame effects of detonations. Suppressive shielding may provide cost or safety effective alternatives to conventional facilities, depending upl?n the.hazardous situation under study. HNDM 1110-1-2 presents . procedures for design, anaLys Ls; qua Li.t.y ..control., and economic analysis of .supp res s i.ve shields. .In .ch Ls. section, a br Le f review of che se .'. procedures is presented. The r e ade r shouLd r eferjt.o HNDM 1110-1-2 for details i
necessary. for design.
.'.
'..
'I"
.>.
6-24. Application Facility operations such as munitions loading, maintenance, modification,. renovation, or .demilitarization mus t .be analyzed to determine which operations involve potentially catasttophic (CAT I or II " MIL STD .~82~) hazards in the ,. event of an inadvertent ignition or detonation. Where the hazard analysis. shows such a potential, the facility design must provide adequate safety for those operations. The alternatives presented to the.design~r of the facility are varied and may include the utilization of conventional protective barricades with appropriate ?eparation!distances, reinforced concrete or steel structures, suppressive shields, or Lso La t i on of a particular operation from--,
the rest of the:facility by the·appropriate quantity-distance . . The decision as to. which alternative.system to use is based primarily on-economicf~ctors, provided all safety 90nsiderations are equal. The facility,. availability of real estate, ..and..equipment costs to include maintenance , .operation, usefuL life, replaceme~t; and modification or renovation must be analyzed for.,each alternative method of protection. Costs will be estimated and-compared over the facility life to determine the most economical mode of protection. .~
A major factor which is paramount in the determination of which form of protection to use is.:the requirement for approval of the .facility by the Depar-t-. ment of Def'ense Explqsives Safety Board . If ·the designer can ,based on economic factors, adapt suppressive shields in the design and support the-adaptation with proven accepted analytical techniques, he should begin development of a facility concept which employs suppressive shields using those shields which have been safety approved.
~,
i .
. • ,. r '
:
6-55
J
.'
.
TM 5-l300/NAVFAC P-397/AFR 88-22
6-24.1. Safety Approved Suppressive Shields There are eight suppressive shield design groups that have been developed to . various stages of definition. These shield groups are summarized in Table 6-4 and illustrated schematically in Figure 6-19; Of the design groups illustrated, five had been safety'approved by the Department of Defense Explosive' Safety Board in 1977. ' The five suppressive shield group designs approved' by the 000 Explosive Safety Board (Groups 3', 4, 5, '6, and 81 mm) have 'been de's Lgned to meet the requlrements for most' applications to ammunition load, as semb Ly , pack (LAP) in the Munitions'Production' Base Modernization and Expansion Program. However, specific shield requirements will vary with other applications and, even with LAP applications, design details will vary from'plant to planc'and between muni- . tions or different operations on the line. It will, therefore, frequently be necessary to modify the approved shields to adapt them to the operation under consideration.
Chapter II and Appendix A of HNDM 1110-1-2'describes the safety approved shield group designs, provides guidance concerning acceptable modifications, recommends procedures for securing safety approval of'new,~hield designs, arid provides swnmary information
on "overal1dimen.l"sions
of the shield structure,
.
charge capacity, rated overpressure, fragment stopping wall thickness, and' type of construction of the five approved basic shield' groups. ' 6-24.2. New Shield Design In exceptional, cases where a' safety approved shield cannot be made to fit 'a desired applicatiori, a new shield can be designed. The guidance needed to design a new shield and the procedures for obtaining:the~afety approvai for the new design are outlined in HNDM 1110-1-2.' 6-24.2.1. Hazardous Environments
• ,
p.
Considering that the hazardous environments' normally associated with suppressive shielding involves :explosives, and/or explosive ordnance, Chapter III of HNDM 1110"t-2 presents information relative to internal and external air blast, fragmentation, and fireball phenomena. This information can be used in support of blast and fragment methods in this manual ,to -de t e rmf.ne venting requirements, air blast loads on -tbe structure, and protection required to' defeat fragments. Some graphs and prediction methods in Chapter 2 of this manual are taken directly from the suppressive shields manual. -,
6-24.2.2. Structural Behavior Suppressive shields can be subjected to large, high pressure loads applied very rapidly. ' The allowance of inelastic behavior of the shield material structural,elements enables much more efficient use of the stru~tural material and does not impair the function of the shield provided, of course, 'that the inelastic behavior.is maintained within acceptable .limits.· The structural materials of primary interest in suppressive shielding are steel and reinforced concrete. Chapter IV of HNDM 1110-1-2 discusses the behavior and properties of these structural materials under static 'and dynamic loading. Additionally, ductility ratios as they apply to suppressive shield 6-56
,"
TH 5-l300/NAVFAC P-397/AFR 88-22 appli~ation are covered. Some.of the information provided will duplicate material in chapters of this manua'L, In case o f- a conflict·, this manual takes precedence.
6-24.2.3. Structural Design and.-Analysis Chapter V of HNDM 1110-1-2 describes techniques which are sufficiently accurate'for preliminary designs in alL.cases, and~in most cases, adequate ..for final designs. These methods deal primarily with the 'dynamic loadings imposed by internal explosions . .The design methods supplement material'presented in chapters .of this manual: Again, in case of conflict, .this manual takes precedence.
J
~.
6-24.2.4. Structural Details Each suppressive shield used for ammunition manufacturing and other hazardous operations will have specific.requirements forlutility penetrations,and doors for personnel, .equipment; and products. Guidance on the provision.of acceptable structural details such.as these:is presented·in Chapter VI of HNDM 11101-2 along with information on. structural details whi~h have been successfully proof-tested. 6-24.2.5. Economic Analysis The design of a facility entails the need to ascertain the. most co s tre f fiec t Lve configuration from among a set of workable design alternatives. All will be designed to provide the desired level of reliability and safety, and the~ selection'of one over another: will be ,based primarily on 'dollar costs. The . economic analysis of alternative facility design is a'complex process unique to each facility.' Chapter VII of HNDM 1110-1-2 illustrates the many factors that must be considered. 6-24.2.6. Assuring Structural Quality In the design of suppressive shields, specifications for.the quality of the basic material. is paramount:· The strength of welds and concrete components are also determining factors in the overall strength of the structure. ",-Chapter VIII of HNDM 1110-1-2 .provides the guidance.which outlines a quality. assurance program for .suppressive shield design.packages., .' Included in Appendix A of HNDM 1110-1-2 is a detailed: description of .the safety approved suppressive shields and guidance concerning acceptable. modifications. Copies of the fabrication drawings -f or each" approved shield design are included along with, direction' for ordering' full-size cop Les .... Appendix B of that manual includes response charts for use in preliminary design. The charts are based on the combined short duration -shock load' .and infinite duration quasi-static load, along with an undamped elastic-plastic responding structure.
• '
.
6-25. Design.Criteria Design criteria for use of suppressive shields,. or suppressive shielding panels, are very dependent on specific applications in protective structures. These criteria may include complete suppression of fragmentation effects, both primary and secondary; attenuation of blast overpressures and 'impulses' to 6-57
"
TIl: 5-1300/NAVFAC P-397/AFR 88-22
specified levels of specific distances from the shield or shield panels; attenuation of-fireball radiation; or even essentially complete suppression of' all of these effects. Suppressive shields mayor may not present reaspnable or 'cost effective_ solutions to specific design problems in protective structures. Generally, they have appeared attractive when fragment hazards are severe· arid'when' pocerrt La l Ly explosive sources are rather concent.rat ed.; The safety-app)'oved shields protect against effects as limited as.small trays of detonators, and as severe as a large melt kettlecin a HE melt-pour operation containing several thousand pounds of explosive. 'The designer should consider their use, and use the methods presented in HNDM 1110-1-2 to evaluate their efficacy, compared to other types of protective structures discussed in this manual.
..
No general design criteria can be given here because the criteria for different operations or plants, and available real. estate, differ too widely. In each specific protection design contract, the AE should be provided with quite. detailed design criteria, in addition to general regulations which fix safety, criteria 'such as AMCR 385-100. Both the specific and ,more general criteria must be evaluated 'when deciding.whether or not suppressive shields will be useful in the facility design. 6-26. Design Procedures 6-26.1. Space Requirements '.' Once the operation requiring _suppressive shields has been identified, consid-· eration must be given to the s~ie and shape of the equipment needed .to.perform the operation and the work'space required inside the shield. These factors' necessarily provide the designer with an estimate of the size and'shape of the shield required. Additionally, space available on the line or in·the building will place limitations on the overall shield base dimensions and height. 6-26.2. Charge Parameters .
-
<, -
,
A principal factor in the selection of a shield which will govern the shield requirements is the establishment of the charge parameters for any specific application. 'The charge parameters are: charge weight (W). shape, confinement, and composition; ratio of' charge weight to shield internal chapter. ,~ (W/V); and scaled distance (2) from the charge to the nearest wall or roof of the shield. (2 - R;Wl/3), where R is the distance from the center of the charge to the nearest wall. or roof in feet and'W is the charge weight· in pounds. ,These parameters for approved shield groups are summarized in Table 6-5. New shield designs can be rdeve Loped for individual needs. 6-26.3. Fragment Parameters -.:;
.
.:
• I
Another key factor in the procedure a designer follows in the selection:of.an·( approved design or the design of a new concept is the suppression of primary and secondary fragments generated by the detonation of explosives or munitions. Much of the material in HNDM 1110-1-2 for fragment perforation of spaced plates has been'adapted to Chapter 5 of this manual. :l J'
1
>
..
,{
6-58
• I
TM 5-l300/NAVFAC P-397/AFR 88-22 6-26.4. Structural Details: Suppressive shields used for,' ammunition manufacturing and other hazardous operations 'require provisions for gaining access to the operation being protected. Personnel must be able to' enter .the shield to accomplish .rcuttne and emergency' maintenance and clean-up and other essential operations. ,An opening of sufficient size must be provided to enable the installation or removal of equipment in realistically large subassemblies. Openings for conveyors and chutes.must also be provided and properly configured ,to prevent excessive pressure and fragments from escaping. Provisions must be made to provide all utilities and satisfy all environmental conditioning needs which may be essentrt a'l. to the operations inside the shield. Utility penetrations, ventilating and .air-conditioning ducts, and vacuum lines must not diminish the overall protective capability of tne shield. They must not alter the 'basic. mode o f vs t ruc tura I failure of the suppressive shield and should be small compared to the general size of the shield. . . Operations that produce explosive dust may require the use of liners both inside and outside the shield to prevent the accumulation of dust within shield panels. With configurations such as the Group 5 shield, which is primarily designed for use with propellants or pyrotechnic materials, liners must not, inhibit. the venting characteristics of the, shield. . , ,
Utility.lines passing through .suppressive shields are vulnerable to both air blast and fragment hazards. The air blast could push unprotected utility penetrations through. the walls of'the shield and create secondary fragments. Fragments from an accidental explosion could perforate the 'thin walls of an unprotected utility pipe and escape from the shield. To eliminate the threat of air blast and fragments. a protective box is used to cover the area where
the utility lines pass through the shield wall. The box is configured to rest on the inside' surface of the' shield and is welded to the shield. The size of the wall penetrations is limited to that required for the utilities . 'Each pipe is bent at a right angle inside the shield within the protective box. The penetrations "of' the' shield wall are reinforced .with a sleeve or box ",' .: section welded to the shield paner through which ,the utility line passes. The· penetration box is designed to maintain the structural integrity of the shield area penetrated. A typical ,protective box de's Lgn is shown in Figure 6-20. The cover plate thickness is selected to stop the'worst case fragment. Typical penetrations for approved safety design suppressive shields are illustrated in Figures 6-21 and 6-22, and a vacuum line penetration is • illustrated in Figure 6-23.
.,
6-26.5. Access Penetrations
In the munitions plant environment, suppressive shields are designed to· protect Category I or II hazardous operations as ,defined in MIL STD 882A~ Remote operation may be required so personnel will not be ,inside the shield during operations. However, personnel access is required to allow for maintenance,repair, and inspection. Further,:these doors must provide large openings to enable most equipment to be installed or removed in large subassemblies.
6-59
TM 5-l300/NAVFAC P-397/AFR 88-22
Access is also required for munitions components, explosives, and assembled· munitions to pass through the suppressive shield. In the case of conveyor transporting systems, consideration must be given to the proper pass-through of the conveyor. Requirements for this type of access depend on the configuration of the munitions product, transporting pallets, and conveyors, as well as production rates and other factors unique to each operation. For. these reasons, definition of specific design requirements is not possible. 6-26,5.1.
-Pe r
s orme L Door
Three different types of doors have been developed for use in .suppressive· shields: sliding, hinged, and double leaf. The hinged door was designed to swing inward. This feature reduces the usable space inside the shield. A sliding door is preferred for personnel access to munitions operations. Figure 6-24 illustrates a typical sliding door. This type door is used with the Group 4, 5, and Milan 81 mm shields. The sliding door consists of an entire shield panel suspended from a monorail system. The panel is inside the shield and is not rigidly attached to the column members. Special consideration was given to, the air gap between the door panel and the column to assure that excessive pressure leakage would not occur and that fragments could not
pass through the gap. The cylindrical Group 3 shield contains a two-leaf door, hinged at each·side. It swings inward as shown in Figure 6-25. The door is curved to match the shield wall contour and is fabricated from S5 x 10 I-beams. Pressure loading restraint is.provided by the door bearing on the external support rings of·the shield at the top and bottom of the door. An exte·rnal latch provides restraint during rebound of the door. ' .: 6 -26.5,2. ·Product Door Only one type of. product door has been developed conceptually for use in suppressive shields. I t. is the rotary, three lobed configuration shown in Figure 6-26 ... The design procedure for this door is described to illustrate the type of analysis required . . It can be used as a guide for analysis of similar alternate design concepts for product doors. . The air blast will most severely load the rotatfng.-product door when the munitions opening is coincident with the pocket in the rotary door. A nonoverriding clutch prevents the door from counter-rotating. The ·angular impuls i ve load Ls ; 6-38 where angular impulsive load reflected impulse door area radius from :center of impulse load·to the .center of door rotation Assuming the product door to be initially at rest, the rotational velocity imparted to. the door is given by: . .'. 6-39 6-60'
TK 5-1300/NAVFAC P-397/AFR 88-22 where angular velocity mass moment of inertia of the door about shaft axis
Im -
The kinetic energy imparted to the door is given by: , ,
KE
I m ~2
- --2
Ti 2 '
,
r' . ~
21'm
6-40
~ :'" ". ,
The strain energy absorbed ,by a circular shaft is given by: 6-41 "
where strain energy length of the shaft shear modulus of tlie shaft material radius of the shaft, '. maximum shear stress in the
shaft
Equating the kinetic energy'of th~, rotating'doo~ to the strain energy in the shaft and solving for the shear stress yields:
.,. s
-
[·::.. r
6-42
The computed shear stress in ,the shaft must be stress or the shaft material, i.e.,:
~ess,
than the dynamic shear
':»
s < 0.55 f dy , ,
6·43
"
,
,
; l "
6-61
.
~2LT. ~Itl
NOMINAL FRAGMENT ~X~;;~tr PATH
INSIDE_' "
SUPPRESSIVE SHIELD GROUP 3 ( GROUPS 1 & 2 ARE SIMILAR. aUT MUCH LARGER, AND "'AVE THREE EXTERNAL RINGS)
10.4 FT.
NOMINAL -1f-#-~~-'HI-"-,p>~'l- FRAGMENT PATH
INSIDE-
-OUTSIDE
SUPPRESSIVE SHIELD GROUP 4
NOMINAL
-mm4H- FRAGMENT PATH INSIDE_
-OUTSIDE
SUPPRESSIVE SHIELD GROUP 5
Figure· 6-19a.
General configuration of suppressive shield groups
6·62
..
o i~
i I
BAR
I i
REINFORCING , RING
.
1/4" MILD STEEL
/
STEEL BANDS
~ -
REVOLVING TRAY
.
SUPPRESSIVE SHIELD GROUP 6
'
" :'1 '.,.
-
.~
15.4 FT. INSIDE-
-OUTSIDE
SUPPRESSIVE SHIELD GROUP 81mm
Figure 6-19b 'General configuration of suppressive shield groups (continued)
6-63
,
.,
OUTSIDE
,
" " " INSIDE CAULKING SE AL
SUPPRESSIVE SHIELD WALL
d,~ r, ,
~
WELD
BOX STRUCTURE
i f
;
.i!~%t1.MI;:F ~r'l!i:"" ",;,p., ':'p'":;~:t~~:~;·~·; !·~:r:~i ....J''':' >t; :ilk
ELECTRICAL LINE
'"•
'"
'"
PNE UMA TIC LINE
-, ;~~~;,~t·~··I:·:'lf~ "--
GENERAL WATER LINE
·1", .. '
,.;'.:t.
~.'!
•.
'J'
"'I,
l
COVER PLATE
.... )
;::~:~ '
., ~
"j.
t'
;.\ • C',
l
. DELUGE WATER LINE
Fioure 6-20
-
I ' .. DRAIN
I~
Typical utility penetration
e
e
<.
VACUUM LINE PENETRATION ASSEMBLY PERSONNEL DOOR
, " .;.
,
:-~ 1.,
UTILITIES PENETRATION ASSEMBLY
:."
F';~ure 6-21
"
Typ ica I location cf tut t l t ty penetration i n vshi e l d groups 4,5, and 31 nm ,-,
UTILITIES PENETRATION ASSEMBLY (ALTERNATE LOCATION)
VACUUM LINE PENETRATION ASSEMBLY
" ,
",
...I-'
;
CD
en
•
I
"
UTILITIES PENETRATION ASSEMBLY
:
Figure 6-22
_ I '
i -,
I'
' .....' ,
<,
Typical location of utility penetrations inrshield group 3
6-66
.'
,
• "
OUTS1".-.
<,
U PP R E S S~VL1.
\
S~IELD
""
".
~
.' - INSIDE
w _
'",,",
'" SPLIT COLLAR
VACUUM. LINE
BUSHING
Figure 6-23 -Typical vacuum line penetr~tion
6-67
OOOR SUSPENSION BEAM ROLLER SYSTEM
>
I!I I 'i; I I
.!tt11 . ,
\
I·
~
.
I
• • • •, • • ,, • • • • • o • , • • '. •
:. I'
I ..
I
;
;) I I
II
•
, ,
o o
• o
•
, • , • • •
• •
• , • • •
o o
• o·• • • • • o
..
·
I
I
•
• 4
·. , SHIELD COLUMN
SLIDING DOOR
6-68
•
• •• • o • • • • o o• • o • . •• • • • o • •• • • o • •
I. I
Figure 6-24 Sliding
• • • • • ,• • • • • • •
personnel~door
e
e
e ::-t
'"
..
"
... " ~~
.
'J
"
II
SH)ELD SIDEWALL
II II II II II
I:
, STRUCTURAL TUBING (ALL, AR9UNP, OPENING)' " TOP HINGE
I
PLATE
INTERMEDIATE STEEL HOOP (RING BEAM)
,
C1'>
LATCH
,BOTTOM
...• ',. "
Figure 6-25 Door - group 3 shield
PLATE
~
LOWER EDGE (lESTRAINT o BASE SLAB
FRAGMENT SHIELD
SUPPRESSIVE SHIELD WALL
ROTATING CYLINDER
. WALL OF U:L
---- PRODUCT DOOR
~
81UM HE PROJECTILE
~~~~
BELT CONVEYOR
Figure 6-26
PROJECTILE HOLDING FIXTURE (NOT ATTACHED TO BELT)
Rotating product door
6·70
Table 6-4
Summary of· Suppressive Shield Groups
HAZARD PARAMETER SHIELD GROUP BLAST High.
1
High
2
,
REPRESENTATIVE APPLICATIONS
FRAGMENTATION
Porcupine Melter (2000 lbs) plus .two POUl; un i.t s ,2?0 lbs each
Severe
, " HE Bulk (750 lb) Minute melter
Severe'
LEVEL OF PROTECTION
*
Reduce blast pressure at intraline distance
by 50 perc,ept Reduce blast pressure at. intraline distance
by 50 percent
..
-"
3
.Hi gh~'. - Moderate.' _:. HE Bulk (37 1l~) Detonators, fuses
4
Medium
Severe
HE Bulk (91b) Processing rounds
5
Low
Light,
30 lb Illuminant
,
,
Category I hazard 6,2 feet from shield Category I hazard 19 feet from shield
.'
Category I hazard Igniter s~urry mix- 3.7 feet from shield ing HE processing
I I
(1. 84 1\»
, (
,6
, Very : High -'
Medium
7
Light
Laboratory, ' handling, and
f t,.
Mo4era~e
•
t ranspo r t.at Lon
F'Lamej/ f Lreb a l L
attenuation
81 mm
High
Moderate
Category I hazard at 5 feet from shield "
81 mm mortar drill- Category I hazard at 3 feet from shield and-face and/or cas t - finishing operation
*
Category I hazard at 'I foot from shield
All.shield groups contain all fragments
.
, P-397/AF~:
TM 5,-1300/NAVFAC
88,-22 '
"
"
.i
"
Table 6-5
ChargeParameiers for Safety Approved Shields
.
"
MINIMUM ,
ROOF
MAXIMUM W/V (lb/ft 3)
1. 63'
1..45
0.04157
' 2.23
'2.19
0.00762
4.14
6.79
,
..
3
"
4
..
.4'·
(ft/1b 1/3)
WALL
SHIELD GROUP
.'
Z
5
r
"
~,
, -,
6A
1.01
N/A
0.22970
6B
1.22
~/A
0.13200
3.62
3.21
0.00340
4.23
3;75
0.00280
Prototype 81 mm* "
, '
Milan 81 mm * •
"
,
,
See Figure 6·19 : .
,
~
ir.
*
0.00215
,.
~.
r :
6-72
,
TM 5-l300/NAVFAC P-397/AFR 88 c22
BLAST RESISTANT WINDOWS ,
6-27. Introduction'
.'
Historical records,of'explosion effects demonstrate that blast'propelled'glass fragments froin, failed windows', are a majo r- cause of injuries' froin,'accidental explosions.' Also" failed window glazing often leads to add l t Lona LtLnj ur Les as blast pressure can enter interior building'spaces' and' subject'personnel to high pressure jetting, incident overpressure, secondary debris impact, and thrown body ,'impact. 'These risks are heightened in modern facilities, which often have' large, areas of glass. Guidelines are presented for both the design,'evaluation, 'and certification' of windows to 'safely survive a prescribed blast 'environment'described by a 'triangular-shaped pressure-time curve, Window designs using monolithic (unlaminated) thermally tempered glass based on these guidelines can be' expected to provide a probability of failure equivalent to that provided by current safety. s't andar ds for safely' resisting wind loads.' J.
-; .
Guidelines are'present~d ~n the form of load criteria for the design of both the glass panes and framing system for the window. The criteria account for both the bending and membrane stresses and their effect on maximum principle stresses and the -nonlinear behavior of glass panes',
The criteria cover a
broad range of design parameters for rectangular-shaped glass panes. Design charts a.r erp r e s enned for monolithic· thermally tempered glazing'with blast overpressure capacity' up 'to 100 psi, an aspect ratio of 1.00 S alb S 4.00, pane area 1.0 S ab S 25. ft 2 . and nominal glass thickness 1/4 S e S 3/4 'inches. An alternate method'f~r blast capacity evaluation by calculati~n is also presented, This can be used to evaluate the blast capacity of glass when interpolation between charts is unadvisable, when design parameters are outside the limits of the chart, and to calculate rebound loads. Presently, the design criteria ·are for b Las t -xe s Ls t ant windows with'thermally t r ea t ed ;'
monolithic tempered glass. Further research is required to, extend these design criteri.a·to laminated tempere~ glass, 6-28'. Background'
.. "
The design 'c r Ltier La cover monolithic 'tempered glass meeting the requirements of Federal· Specifications DD-G-1403B and DDG-45ld. Additionally, thermally tempered glass must meet the' minimum fragment weight requirements' of ANSI 297.1-1984, Secdon 5.1:.3(2). ',' s:
".f • .
Annealed glass is the most common form of glass available today. Depending on manufacturing' techniques, ,it is also known as p l'ace , float or sheet·'glass. During' manufacture , it is cooled s Iovl.y .. 'The process r e'su l t s in very little; if any.,' residual' 'compre s s Lvevaur f ace s t r e s s.
Consequently, annealed "glass -Ls
of relatively .Low strength when compared to tempered 'glass -",It has 'large variations, in~trength and fractures into dagger-shaped, razor-sharp fragments. For, these reasons, annealed g l as s. is' not recommended for, use in blast resistant .windows., ~ ' ..
'. ,' " Thermally, tempered 'glass is the most readily available tempered glass on the market ., It .Ls manufactured' from annealed 'glass by ihea t Lng to' a' high uniform temperature and then, applying, controlled, rapid' cooling:', As the" internal temperature profile relaxes towards uniformity, internal stresses are created. 6 - 7,3
TM'S-1300/NAVFAC P-397/AFR 88-22
The outer layers, which cool and contract first, are set in compression, while internal layers are set in tension. As it is rare for flaws, which act as stress magnifiers, to exist· in the interior of tempered glass,sheets, the internal tensile stress is of relatively minimal consequence. As failure originates
from.te~sile
stresses
e~ci~ing
surface ,flaws in .the
g~ass,
precom-
pression permits. a larger load to be carried before the net tensile strength of the tempered glass pane is exceeded. Tempered glass is typically four to five times stronger than annealed glass. , .' I
The fracture characteristics of tempered glass are superior to ,those of annealed glass.
Due to the high strain energy stored
~y
the prestress,
tempered glass will eventually fracture into small cubical-shaped fragments instead of the razor-sharp, dagger-shaped fragments associated with,the. fracture of anDealed glass. Breakage patterns of side and rear'window~ in American automobiles are a,good example of.the.failure ~ode of , thermally or heat-treated tempered gl~ss. "
.
.
Semi-tempered glass is often marketed as ?afety or heat-treated glass. However, it exhibits neither the dicing characteristics upon breakage nor the higher tensile strength associated with fully tempered glass, and, therefore, it is not re~ornrne~d~d for blast resistant windows. Another common g~azing material,is wire-reinforced glass, annealed glass with. an embedded layer of wire mesh. Its only use is as a fire-resistant barrier. Wire' glass has the.fracture ?nd low strength characteristics of annealed glass and, although the wire binds 'fragments, it contributes metal fragments as an additional hazard. Wire glass is never recommended for blast resistant windows.
r
6-29. Design Criteria for Glazing
"
-'
.'
6-29.1. Specified Glazing
h.
.
,
The design of blast-resistant windows is currently restricted to~'hea~-tre~ted, fully-tempered glass in fixed or non-openable frames meeting both Federal Specification DD-G1403B and ANSI ,Z97.l-l984. Tempered glass meeting,only 00G-1403B .may possess a surface precompression of only 10,000 psi. At this leve~ of precompression, the fracture pattern is similar .to annealed and semitempered glass. Tempered glass meeting the minimum fragment specifications of ANSI Z97.l-l984 (Section 5.1.3(2» has a higher surface precompression level and tensile strength, which improves the capacity of blast-resistant 'wi~dows and results in smaller, cubical-shaped fragments on failure. Although therm~lly tempered glass exhibits the safest failure mode of any glass, fai~ure under blast loading still presents a significant health hazard, Results from blast, tests reveal that' on failure ,tempered glass' .fragments may be propelled ·in cohesive clumps' that only fragment on impact into smaller· rock-~alt'-shaped.fragments. Even. if the tempered glass .initially ·breaks into small fragments" sufficient velocities may still,be imparted by the ,blast ' loading to cause severe hazards to personnel. Because the expected geometry' of glass fragments .LnvoLve multiple, potentially sharp corners, a high probability of, injury would result from application of the 58 ft-Ib criterion for acceptable, kinetic energy. Because of -tihe s e 'hazards. to personnel, blastresistant glazing should be designed to survive e~pected loads. ~ : 6-74
TK 5-1300/NAVFAC P-397/AFR.88-22 6-29.2. Design Stresses The_design stress, the maximum' principal tensile stress allowed for the. glazing; fun" is set at l6000,psi. This correlates with a probability of failure equal to or less than 0.001. The, design 'stress for blast, resistant '. glazing is slightly higher. than that commonly used in the· design,' of one.-minute wind loads, 'but, it is justified b~ca~se of the expected, significantly shorter, duration of loading. f.
6-29.3. Dynamic Response to Blast. Load
".
.~
r ;
..s,
An analytical model was used to predict. the blast load capacity of annealed and tempered glazings. Characteristic parameters of the ·model .are illustrated in Figure 6-27. The glazing is a rectanguiar, fully thermally. tempered glass plate having a long dimension, a; a short dimension> b; a thickness, t;.a poisson ratio, v
0.22;. and an' elastic modulus ,: E'- 1 x 10 6 psi. The pl ate Ls simply supported along -al.L four edges ,'with noinplane and .ro.tational restraints .at the edges, The bending stiffness' of the 'support' elements ,is assumed, to,he' infinite, relative to the pane., Recent static and blast load tests indicate that the . allowable frame member deflections of 1/264-th of the span will not signifir
cantly reduce pane resistance from that predicted for an infinitely stiff
frame.
, .. The blast pressure loading·is.described.by a peak triangular-shaped pressuretime curve as.shown in Figure·6-27.. The blast pressure rises ~instantaneously to'a peak blast pressure, B, and then decays linearly with a blast pressure duration, T_' The pressure is uniformly distributed over t~e surface of the pane and applied normal to the 'pane. , The resistance function, r(X) (static uniform load, r as a function of center deflection, X) for the plate accounts for both bending and,membrane stresses, The effects of membrane stresses produce a nonlinear stif~ness of'the'resistance-deflection, function (Figure 6-27). The design deflection, Xu is defined as the center deflection where the-maximum principle tensile stress .at any point in the glass first reaches :the design stress, 'fun' of l6,OOO.psi. Typically, as the deflection·of the plate exceeds a third of its ,thickness, the points of maximum stress will migrate ,from the center and toward the corners of the plate. The model, illustrated in Figure 6-27, uses a single-degree-of-freedom system to simulate the ~ynamic response of the plate. To be conservative, no damping of the'window pane is assumed. The'model calculates the peak blast pressures required to exceed the prescribed probability of failure. ·The. model assumes that failure occurs when the maximum deflection exceeds ten times the glazing thickness_ This restricts solutions to the valid range of the Von Karmen plate equations while preventing edge disengagement of the plate. 6-29.4. Design Charts Charts are presented in Figures 6-28 to 6-48 and Tables 6-6 to 6-12 for the design and evaluation of glazing ,to. survive. a prescribed ,blast loading'with a probability of failure no greater than 0.001. The charts:relate the peak' blast overpressure capacity of thermally tempered glazing to all combinations 6-75
TH 5-1300/NAVFAC P-397/AFR 88-22
of the following design parameters: afb - 1.00, 1.25, 1.50. 1.75 2.00. 3.00, and 4.00; 1.00 S ab S 25 ft 2; 12 S b S 60 inches; 2 S T S 1.000 msec; and t 1/4. 5/16, 3/8, 1/2. 5/8, and 3/4 inches (nominal). Thermally t emper ed glass" up to 3/4 inch thick can easily be purchased in the United States ... Thickness greater than 3/4 inch can only be obtained by lamination, but. research and' blast load testing are required to'oevelop approved designs for laminated glass. ,"~ ... The charts are based on the minimum thickness of fabricated glass allowed by Federal Specification DD-G-45ld (see Table 6-7). They are ·created by numerically integrating the equations of motion of a single-degree-of-freedom system as'modeled in'Figure 6-27. A Wilson-Theta technique' was employed with a time step corresponding limited to 1/25th of each of the five increasing linear resistances used to model the resistance function. ' 6-29.5, Alternate Design Procedure ".
Design procedures in this ,section ean be used to evaluate the' blast capacity of monolithic tempered glass 'when' interpolation, between charts is unadvisable .. when design parameters are outside the limits"of the chart. or to calculate rebound loads," It is recommended that the design charts be used for initial guesses of required glass thickness. '.' Procedures to calculate resistance, r u' (Table 6-6) deflection, Xu' effective static resistance. reff' effective pane stiffness. Ke• and the period of vibration, TN follow. The response charts' can be used with these parameters to determine dynamic response. Tabre' 6-11 reports the fundamental period of vibration and Table 6-12 reports the effective elastic, static resistance for most dimensions of glass panes. In many cases these:values and the single equation in Step 10 below can be used to directly compute b Las t-vove rpxes sure capacity. t
-, '._>
•
•
Step 1, Determine i f the glass pane will' behave as a 'linearly ·responding'." plate under the design load. If the ratio of' the short side of the plate, b ; " to its actual' (not nominal) thickness is less than 'the maximum in the 'second; column of Table 6-8, then simple formulas can be used for 'parameters needed to use the response charts. Only glass panes .'above and to the left of the steeped line in Table 6-6 will qualify. If the glass 'pane has a b/t-ratio' less than that specified in Table 6-8. the glass will behave. in' a nonlinear manner with the membrane stresses induced by straining of.the'neutral plane or axis of the plate. Proceed to steps 2-9 to determine key parameters for this nonlinear plate behavior. . {-;
i' .
For glass plates that respond linearly, the-design static resistance and the" effective elastic resistance 'are : . '.. -:
,
, 6-44a .
"
The center deflection is:
..
..
6-44b
Coefficients for'computing the effective resistance, Cr' and the center deflection,. CD;' are listed in Table'·6-8. 1,
.I
6-76
I
TH 5-l300/NAVFAC P-397/AFR 88-22
The fundamental period of vibration is
Coefficient CT is reported in the last column of Table 6-8. 10 to determine blast capacity.
Proceed to Step
Step 2 For nonlinear behavior with the bit-ratio greater than specified in column 2 of Table 6-8, determine the nondimensional design stress, SSD' as: SND'- 0.0183 (b/t)2 where:
6-46
b -
short span of glass measured between center lines of the gaskets (inches), and
t -
actual thickness of glass,'from Table 6·7 (inches)
For values of alb greater than 4, use alb- 4. Stet," 3
Enter Figure 6-49 with the values of SND and alb to determine the design load, ~D'
nondim~nsional
Step 4
Compute static design resistance as:
r u - 876,000 (~D)(tlb)4 psi
, 6·47
Use this value for frame design calculations other than those 'for rebound. Use reff defined in Step 7 for the rebound phase. Step 5, Use alb and ~D in Figure 6·50 to obtain the nondfmens Lona l, deflection, X/to If X/t exceeds ~O, use:t~e value' of ~D corresponding 'to an X/t of 10 and recalculate r u using Figure 6-49. Step 6
Determine center deflection of the glass pane as: "
, ~ - (X/t) t, inches Step 7
.
6-48
Determine the effective elastic, static design resistance as: 6-49
where':
resistance at 0.2 rl re~istance at 0.4' r2 resistance at' 0.6 r3 res 1.8 tance at 0.8 r4 r u obtained in Step 4 ~ obtained in Step '6
u
~
Xu ~ ~
Figures 6-49 and 6·50 should be used for rl 'through r4' The equivalent static design load is the resistance that a linearly responding plate would exhibit at the' Same strain energy as the norrl.Lriear Ly responding plate at the design center deflection~, It is always less than r u' By this technique, the linear response charts can be used with reasonable accuracy.
6-77
TM 5-l300/NAVFAC P-397/AFR
88~22
Determine effective stiffness as:
Step 8
6-50 Determine the fundamental period of vibration as:
Step 9
6-51 KLM 0.63 + 0.16(alb ' 1) KLM - 0.79
where:
1 :S alb :S 2 alb ~ 2
The unit mass, m, of the glass is:
Step 10 Use Figure 3-49 of Chapter 3 with the ratio of load duration to period of vibration, T/TN, to obtain the dynamic load factor, DLF. The blast overpressure capacity of "the glass pane is: 6-52 For T/TN-ratios greater than 10, set DLF equal to 2. 0.05, set DLF equal to 0.3.
For ratios less than
6-30. Design Criteria for Frames 6-30.1. Sealants, Gaskets, and Bead~ All gaskets or beads must be at least 3/8 inch wide with a Shore "A" durometer hardness of 50 and conform"to ASTM Specification C509-84 (Cellular Elastomeric Preformed Gasket Sealing Material). The bead and sealant must form a weatherproof seal'. 6-30.2. Glazing Setting Minimum frame edge clearances, face clearance, and bite (Figure 6-51) .are specified in Table 6-7. 6-30.3. Frame Loads The window frame must develop the static design strength, r u of the glass pane (Table 6-6). Otherwise, failure will occur at less" than the predicted blast pressure capacity of the window pane. This results from the frame deflections which" induce higher principal tensile stresses in the pane, thus reducing the strain energy capacity available to resist the blast loading. In addition to the load transferred to the frame by the glass, frame"members must also resist the static design load, r u' applied to all exposed members. Maxim~ allowable limits for frame design are: 1.
Deflection:
Relative displacements of frame members shall be the smallest of 1/264th of its span or 118 inch.
6-78
.
TM 5-1300/NAVFAC P-397/AFR , 88-22 2.
3.
Stress:
Fasteners: . . 1 .'
•
~
The maximum stress in any member shall not exceed f~/1.65, where f m is the yield stress of the members material. ,
'
The maximum stress in any fastener shall'not exceed 1'.,. I', f m/ 2 . 00 . 'I, • .' . ,
The desLgn ioads :for the glazing are 'based on large deflection theory, but the resulting design ,loads transferred to the frame are based on'small deflectidn theory for normally loaded plates. Analysis indicates this approach to be considerably ~impler and more conservative than using the frame loading based exclusively on 'large deflection plate 'behavior: characteristic of window panes. The effect of the static design load, r u' applied· directly 'to the expo~~d frame 'members of w~dth, W, should also be consider~d. The designCroad, equal to:
-'~"
6-53 The des Lgri load, .ru ' produces a line shear, Vy' applied by the short side, b,. equal to: 6-54 The design load also produces a corner q6ncentrated load, R, tending to ,uplift
the corners of the window pane equal to: .'
, ,
~
6-55
Distribution of these forces are illustrated in Figure 6-52. Table 6-9 presents design coef~icients, Cx' Cy ' and CR ~~r practical aspe~t ratios. Linear interpolation can be used for aspect ratios not shown.
Altho~gh frames with mullions ,are 'included in the design cri~eria, it i;
recommended that single pane frames be used. Experience indicate~ that mullions complicate the design and reduce the reliability of blast-resistant frames. If mullions are used, the certification test must be conducted as the complexity of the mullion cross sections may cause some ~f'the assumptions to
be.unconservative for local shear and stress
concentratio~~.
Also, economic
analysis indicates that generally_thicker,glass will be more cost effective, than tqe more complex mullion fr~me. If mullions ar~ used, the loa~s f~om Equat~ons 6-55 to 6-57 should be used to check the frame mullions and fasteners for compliance with deflection and stress criteria. ;Note that the.design load for mullions is twice the load given by Equations 6-53 to 6-55 to.accou~t for the, effects of two panes supported by a common, mullion. . • . I
'
•
_
"
.
,
Special design considerations should be taken to assure that deflection of the building wall will not impose deflections on the frame greater than 1/264th of the length of the panes edge. When insuffi~i~nt wall rigidity is provided" it; is recommended that the frames be pinned at the corners to provide isolation from the walls rotation.
6-79
TM 5-1300/NAVFAC P-397/AFR 88-22
6-30.4. Rebound Response to the dynamic loading will cause the window to rebound (outward deflection) after its initial positive (inward) deflection. ,.The outward pane displacement and the stresses produced by the negative deflection must be safely resisted by the window while positive pressures act on the window: Otherwise, the window which safely resists stresses caused by inward deflec' tions may fa~l in rebound while the ,positive pressure still acts. This can ,',,> propel glass fragments int"o rhe structu're" However, if the window fails in rebound during ~he negative phase ,of blast loading, glass fragments will be drawn away from the structure. Rebound will occur during the~egative loading phase'if the effective blast duration is no greater than one half of the natural period of vibration, TN' of the glass,pane. For T ~ 10 TN' significant rebound does not occur during the positive blast loading, so, for this situation, design for rebound is not required. For 0.5 ~ TIT N ~lO, -the frame must be designed for the peak negative ,pressure acting during the positive ove.rp res sur e phase. Table 6-11 r'epo r t s TN for practical glass pane dimensions. As the rebound chart, Figure 3-268 of Chapter 3, can be unconservative for predicting maximum rebound (for glass panes), dynamic analysis using numerical integration or a more conservative simplified analysis is required. In lieu of a n~erical analysis, it is conservative to set the maximum rebound load, r ", to the static design load, r u'. The resistance. function for this analysis
can be generated by the Alternate Design Procedure. If the pane has a bit ratio less than specified in T~ble 6-~, the pane will behave as a linear plate and Equations 6-44 can be used to determ!ne rut Xu, and the resist~n~e , function. If the pane has a bit ratio larger than 'specified in TabLe.: 6-8, use Steps 2 through 7 to define the resistance function. The negative resistance -function is a mirror image of the positive resistance function. The portion of the frame outboard of the' glass fuust resist the r~boundload, r-. Use'Equations' 6-53 to 6-55 to apply: the rebound load to the frame members. In some design situations the resistance bull t ,~into the member outboard of the glass to resist the,corner concentrated force, R (Equation 655), during deflections of the pane inward will provide enough strength to resist r ebound',
6 ~31-. 'Acceptance 'Test Specification Certification tests ot the entire'window assembly are required' unless analysis demonstrates that the window design is consistent with the design criteria. All window assembly designs, usingmultions must be 'tested. The certification tests consist of applying static uniform loads on at least two sample window assemblies until failure occurs in either the glass or, frame. Although at least two static uniform load failure tests are required, the acceptance
criteria presented below encourages a larger nUmber of test samples. All testing should be performed by an independent testing laboratory certified by the c?ntr~cting officer. 6-.31.1. Test Procedure - lIindow Assembly Test The test windows (glass panes plus support frames) shall be identical in type, size, sealant, and construction to those furnished.
shall be secured to simulate the adjoining walls. 6c80
The test frame as~embly
Using either a vacuum or a
TIl '5-l300/NAVFAC P-397/AFR
88~22
liquid-fi'lled bladder .. an increasing uniform load shall be appli'ed to,the entire window assembly (glass and frame) until failure occurs' .Ln either the' glass or frame. Failure shall be defined as either breaking of'glass or loss of frame resistance. The failure load ,shall be recorded co. three significant figures. The load should be applied at a rate of 0.5 r u per minute which corresponds to approximately one minute of significant stress duration ~ntil failure. Table 6-6 presents the static ultimate resistance of old ,tempered glass, correlated with a probability of failure, equal to 0.001 and a load duration of.l second, whereas this criteria is for a load duration of '1 ,,', , _ minute. This longer duration will weaken the glass by ceramic fatigue ; but ,new glass should tend to be stronger than old glass. To account for these effects the certified static load capacity, r s' of ' a glass pane is to be rated as:
0.876 r u
'r _,s
,--,
6-56
6-31.2. Acceptance Criteria The window assembly (frame and glazing) are considered acceptable when the arithmetic mean, of all the samples tested, r, is such that: "r
~'r~ !
+ s a
6-57
.
where ru s a
~
static ultimate resistance of the glass pane sample standard deviation acceptance coefficient
For n test samples', r is defined as: n
1:: ri i-l.
6-58
r -. n
where ri is the recorded failure load of the i t h sample deviation, s, is defined as:
~est sa~ple.
The standard
1/2 6-59
s -
(n-l) The minimum value of the sample standard deviation, s, permitted to be employed in equation ,6-47 is: smin - 0.145 r u
6-60
This assures a sample standard deviation which is no better than the ideal tempered glass in ideal frames.
6-81
'
TM 5-1300/NAVFAC·P-397/AFR'88-22
The acceptance coefficient,' number" of samples tested.
Q,
is tabulated in Table 6-10 in' Ferms of the· "'?< ,-'"
. . ,:'-:
The ·following equation is presented to aid in determining if additional· test , samples are justified. If: , ' , ~' ,. -'
r
$;
.
d
',-'.:1
'. t .
,r u + s B
6-61 ,
then with 90% confidence, the design will not prove to, be, adequate with additional testing. The .rej ection coefficient, B, is from Table 6 -10 .: I .,
~
.
....
6-31.3. Certification for Rebound
<,
:..~."
..' .
Acceptance testing shall be ,performed for rebound unless analysis demonstrates that the frame meets rebound criteria. All frames with mullions mus~ be tested. Testing is performed with the load applied to the inboard surface of the window assembly. The equivalent static rebound load, r ", is subs.tituted: for the design load, r u' j
"
6-32. Installation Inspection A survey of glazing failures due to wind load indicates that improper·installation of setting blocks, gaskets, or lateral shims, or poor edge bite is a significant cause of the failures experienced.
.
\-
'.J '.
·e 6-82
I
'
-,
, &;
m (I)
..
Window pane aeomlay
B
'\1
-------------
li
•
.s•
~ 0'
..~ • ,;;
l,:
iii
Timc.T
Center Deflection. X
lbl Blue loadin.
(c) Resistance of glass,p,ane
r
x (d) Dynamic rnponw model
Figure 6-27
d1araetcristic parameters for gl:w pane. blast loading. resistance function and response model.
!,
•
6-83
100
I/O • 1.00
•
• 1/.. In•
r--...
b
10
II I
Un••
8'1..'
12.1:1
1.112
1•• 1'
,5.1'
,
c-
2'. 2"
3.IlZ
30.30
z.ol loa
31.31 42 _ 4:1
;:"'0
OM
"It•.
OM
M_M
.....
1.07 '
41"411
¥!
Po
a.
'OJ
'" .•..
... ~
0.1 .
....•.. "a
10
I
10
100
1000
a
'".a..
alb • 1.00 t • 1/1'1...
... u
.... 10 ,
II"in
12.12
1'"
1•• 1.
7.1lO
24.24
1.1'
30_30
:L48
31.38
:L48
4:,1.42 4lI.4lI
50&.50&
1.1:1 lA3 1.14
eo.eo
CI.M
,
, I
t"'--.
~ ..... ;..,; 10.
,
I
,
.
I
O. I', I
b II. (lnJ
10
100
1000
Duration of Blast Pressure. T (msee) Figure 6-28
Peak blast pressure eapaciry for tempered glass panes, alb. 1.00, r. 1/4 and 5/16 in.
6-84
. ,.
100
.no •
1.00 3IIlnw
t •
•
b •• Un.1
'~I
12.12
.
&1
1•• 1.
11.1
.
.....
\0
:M.:M ,
:-... n
~
I
..
I
...
100
.
--
MaM
•••
i
~
100
UI
:tai
".
I.
1000
DIll • 1.00' lIZ In,
1 •
,
.
UII
......
a
"" ..."
_S • •
_
10
1M LIZ
42.42
.....:
a. ......
. "" .."".
•
30.30
b •• lA!
.lpoll
12.12
qJI
1•• 1.
11.1
:M.:M
1°"
........ .....
...
.
"
I
,
10
30.30
, , .
, .
I I
I
"31.31
SA
Q.Q
4.5l1
.....
z.n
111.111
z.3O
1000
100
Duretion 0/ Blest Pressure. T (m s e c ): .
Figure 6-29
'
Peak blast pressure capacity for tempered glass pane>, alb = 1.00,
6-85
t
lA
54.54
•
10
po
= l/8 and 1/2 in.
tOO
~
•• UaJ
• \00II
12.12
70A
11. l' .
31.3
:M.:M
17.5
30.30
11.:z
"-
'. 10
31._
7.
'42.42
.....
t.aI
4lI.4lI
1I.:l1
111.10
AM
:ua
alii • 1.00
.. ...., 0.
t
I \
.r
\0
~ --;'~
100
1000
'" ...:J
I
. • 1000
... .•.. .a
.
0.11 .• 1.00 t ..
31......
0.
.
I
'"a •
•
-"
" •• IIllJ 12.12
0.
10 0
....
11.1• .,
m~ .~
,
,
i
:
I
§
10
,
30.30
1\l.3
31&31
11.3
~ 4lI.4lI 54-xM
I
I
10
I 1000
100
Duration of Blast Pressure. T (msec:) Figure &-30
Peak blast pressure capacity for tempered gl151 panes: alb
6-86
~
1.00, t
~
... 211.1
10.10
I 1
1a
:M.:M
q.q
.1
• fJoIl
S/8 and 3/4 in.
lIA3
\l",
5.110 II.1II
e
100
.•
.."
.
...
,
_.~.
,
,
.
,-'.
•
.
,
,
125
1/4 In.
t
"
,
I
.
"
, ~
10
b •• "
....
-
'"
~
".
......
. ...., ,
,1
~
•
,
....
'"
.".. :2
12. 15 ,
1.15
1', zz.a
4.IB
24.30
:LIII
30.31.1
1.
••••
1.1'
....
."..
.
.
.
0, t I
, ..
I
. .. •
a
..
'"
e,
.
"
.
I
81....1
.
CII.1ZJ1
'"
...'"a
-
.'
.
.
....... .......
-I
..
a.
e
.
un.i
oa oa
-
-
,
,
~
., ...,
I" -
10
..
100
.
.
"
.
..
.
..
1000
,
lL
100 ,
,
.
,
.
'
. .
i""'ooo. -
,
.
I, ,
.
.
I
I'
'
. b •• IIn.1
8 1....1
12. 15
12.7
-
... 1
t I
.
•
10
"
I
,
"
,
alb • 1.215 t • SIt. In..
I
""-..
~ .......
I
'I ,
II
r-.....
I I
.
I
I i
10
100
I, I
I
i I
1I'.2Z.5
S.-
24,'30
4.111
30.31.5
2.93
3111 4lI
2.oz
42. !S2.5
l.S2
•• ao
1.11
1000
Duration ot Blasl.Prlssure, T(msec) Figure 6-31
Peak blast pressure capacity for tempered glass panes, alb = 1.2S. t = 1/4 and S/16 in.
6-87 "
"
'
,
I.t
,
,
100
--
,
lib
-.
1.211 -
3111n. -
t
, ,
~
-
, ,
~
I. . . Un.1
,.
,
:
0
10
-,
-
...
-
0
0
0
1•• 22..B
..
~
:N.»
i""'oo...
~
,
1:"'000.
..........
"'-
"
.
..
'-
a.
"
-
".
..
-
10
~
-
I: .
~
I I
,
.
,
I
·.1·
Figure 6-32
, I
,
-
I
12.18
3Z.I
18. Z2.5
,."
..
:M.:III
8.13
:111.37.5 38.411 • 2.82.5
8.52 .02•
"Duration ot 81ast Pressure"r (msec)
,
,'
I '100
1000 -,
"'
. .. ~
6-88
.. ,
,
3.-
z.,
Peak blast pressure capacity for tempered gl~panes, .ib'= 1.25, t.= ll8 and 112 in.
""",-
,
i ... eo
.'
10
I
81l1li/
.
,
,,
~
,
b •• Un./
-
II
,
'
l
"
... .......... ,
, "
"
-,
...
z._ 1.1,
lib- l.2l1 t ~ lIZ In. -
,
-
42 _IIZ.I
I
,
'
.. I
UI
1000
I
.
."
a_ •
0'
100
t'
100
a.
8iD 4..lIZ
.
10
...
:111.37.1
•••
0
:,
'-
.. .
.
I
II
11.8
'0
,
.."
12·Jl 11
81. .'
100 .",. 1 •
1.21
!5IlI1";
,
,
m..
......
--.
-, ,.
.,
.
~
1
"
I
...
100
~
,. a :z:z.s
ZU
Z4a30
112
3Oa37.1
UII
3lIa'" 4Z a lIZ.5 4IalO
I.CII
.
,
, 10
100
.
,
1000
,
,
~ a.l/loW 12a 18
.
....
~ .
10
.
.
. ..
-
III ~
.. '
....
Es
, ,
, ,, L
,
...., .~
..
.~,
,
1
Figure 6-33
10"
100 ..
,.
:M.2
-..
Z4a30
11.3.
3Oa37.1
12.3
' '
:JIIa4B
lUO
4Z a 5ZJl "'allO
5.41 5.111
.
.;",'oi 1.21'1" i .' 3.'4
I, I,
..
I,
Ii
1In.
1000
Duration of 81a'st Pressure, T (msec) . Peak blast p:essure capacity for tempered glass panes, alb
= 1. 25, e = 5lS'md 314
in,
. '
6-89
..
18a ZZJI
,
,
II ' - ' 71.0
,
I
,
I'
4Ale
,
,
0
'"-'"a
....
,
...
Ql
.a
,
12 a 111
,
.
..
,
Q.
.•. . •.
.
..
~~
to
BlIooIl
.
,
1"--0..
I. . . Un.l
.
tOO alb • 1.50'· t •
114 1ft: •
..
..-. ,
.
.
-
.
· •
10 Is •• fln.1
8e.1
12 all'
1.70
lIa%7
1':1
:lh 31
:u3
:ilia •
,...,
.
""
.
........
....-.
.....
....a.
...., CD
.. .. .
':J
.,
.
31.....
,.11:1 .
GaA Clan
0.,.
. 08
..
'-
11.
.
-
..
a
CD
...a
;
....11.--
0,1 . 1
\00
.
10
'100
...
..
.....
"
" I
t •
•
...,
.
.
~ 10
-
. . 1000
,. '
511. in.
..
.-
.
.
-
.
-
..
b a .1....1
. 12 a la,
•
8lpol)
10.'
,
. I
· \ I
~..... 10
- .....
'"
'.
,
I
I
i
, \00
I
I
laa%7
4.18
24a38 JDa41
1'4 2.57
3Ial4
'.71
GaG
1.21
• • 72
1.02
1000
Duration at Blast Pressure. T (msec) Figure 6-34
Peak blas pressure capacity for tempered glass panes, alb = 1.50. e = 1/4 and 5/16 in.
6-90
100
.
..
.
alb· t •
.
'.
UG
.
311 In.
;
,
t..
b •• UnJ
B~l
t2. tl
ttlJl
10
-~.
'"•
.'". • '" -.. ~
""
tl.27
tl.8I
:M • •
4JllI
:sa •• a.M a •• .an
.....
,
3.tO ,2.14
,..,
tAl
t.
10
I
~",IOO
100
1000 .
..
..
..
a
'"a
alb •
-
~
•
~
,.eO
t · 1/2 ....
..
"-
b •• linJ
• (poll
t2. t.
211.1
ttl. 27
" ..
10
,
,
•
"
.
I
I I
I
~
I'
z.'3Il
as
:II1x 4&
1..12
~.M .a
!it 2.9
. . . 72
I 1
Figure 6-35
I 10
. 100
Ouration ot Blast Pressure,
r (msec)
Peak bias, pressure capacity for tempered glass panes, alb
6-91
1000
= 1.50, r = J/8 and 112 iii.
100 . .
.",
. 1.50
t • SIIln.
..
N' .... ~
.
~
10
.
III•• lin.)
8 (. .,
12.18
41..
18. %7
1...
:M • •
10.1
..
~ .. ....,
31.154
4.74 U7
•• n
.
Go
..n
a.a
.'
.....
30.41
So1•
CD
. ...
'2
'"
'...
-.".
I. 1
.. 10
1000
100
III •• (IRJ
,
12.1.
... .1A
CD
.... ...
.;<
". do
,
~
. 10
,
~i:'o... ~:M
I
I
I
...
-
..
1•• %7
.
"
,
:M • •
.
.. ,
I
,
,
,
,
, ,
,
,
:lOa.
I
,.
•
I !
!
8.14
42aa 41.72
B.oB 4.oB
alii • 1.50 . t •
I 10
Duration Figure 6-'-36
Peak blast pressure capacity for tempered glass panes, alb' 1.50.
6-92'
311, In..
I I I 1000
100
or 8Iast,Pressure, T (msec) t «
...
111A
38.154
I·
..
.
%7.3
r
518 and 3/~ in.
100
•.
.",
•
1.~
1'. In •
,
•
;
~
.
,
.
,
. j
.a. ....,
-
-
.-
It ••
• (II1II
11lLl
12.21
I.GII
.-.
-'
,.
~~
1
..
'.
e-,
",",'
. 1•• 31..
z.se
201.42
ZJIZ
311. u.s
1.11
3"" a
OM
-
ca.1I,
42.73JI
"'~
,
•
'-
..•."
-
.-
'-
lL
.
..
. ... - , ..-. .."
.
a
'"
,
,
---
,
~
-
•
10
,
..
.' .
O. 1
'"a • lL 100
, ,:.
..
r 10
.
"
,
- .
;
.
.
.
.1000
,
,
.
100
-
. '.75,'
.",
•
,
,
.
-
-
I
;, 11111n. ;
..
:
-
.!
. .
"
. -
-~
.-
.~
.
,
•
,0
,
-
,
,
. , r.n
~
, ,
,
,
I
, II 100
• {poll
12 _ 21
8.87
,
I
24_42 :II.IZ.I
z.zz
311_13
1.13
42.~
1.2l1
1000
Duration ot Blast Pressure, T (rn s e c) Figure 6-31
Peak bias. pressure capacity for tempered glass panes, alb = 1. 75. t = 114 and 5/16 in.
6-93.
~
"
I
I
10
,
•
I... Un.!
.1'_ 31 ..
-
~ ...... , ........ ....
-
-.
,
10
I
..
.. .
- -
..
.
,
. ;'.
. , - .. , . -
-
.
"
•
-
2.lIO
100
.", • 1.78 •• 318In. ,
-
"
~"'.(lnJ
II c.l
12121
11.1
10
.
--
~ ......:
,
"
'
Q.
,
,,
CD
.•. .•. .. " ..
Q.
.
.
"
11
-
UO
~Ica
20M
I"IU ,sa HI ,,,. ca .. 73.5
_.
"
10
'100 - , I
'1000
-.;.
\
100
tI I 31.1
,
on. •
a
t •
CD
'.78 112In.
""a •
~ I.
CL
12 I 2'
~ ,"'-
10
,
~
I
I
I I I I
I
,
I
:MIca
I
I'll
I
I !
I
:
I
Figure 6-38
Peak bias, pressure capacity for tempered glass ~,!,es, alb = 1. is.
<
~.
",
.;.
, 10,0.
•
3.82
1M
1000
••
6-94'
t
5.10
311113 42 I 73.5
,
..
10
10.3
'30 I 52.5
",
D,":Iration ot Slast Pressur,e', r. (msec)'
1
23.1
-, I
I III i ' ,
. 1
II l..u
1
tI I 31.1
I ,
(lnJ
= 3/8 andll/2 in.
2.311
, .
.
100
alb • 1.7& t • ;511 In.
1:.
b
,
,
--.a:
.
"
-
,
lL
100
" '"a.,'
18.5
.
,
.
~'4Z
• .21
3lh II2.ll
5.18
:ahB3
4.1.
Gx71ll
U7
.-
, 10
l-
.,",
'
.
....
. , .,.... . :>
I
"
....,.
.,...
-
--
..
.
CD
iB x 31.8
,
,}!i~ - ..
.-' "
37.1
..
,~ -
10
B (poil
12 x 21 'I
~....;
.
Un.)
100
1000
,
..
alb'. 1.7S
"
t
•
31....
CD
It x • lIftJ
• 'poi~
.12'x 21
IIoU
1. x 31.5
24.2
0-
1';,100,;
~~
lL
~
10
-
,
3D x II2.ll
."
,
... 0
. t 1
I
..
..
...
-
•
.
,
3IxB3
5Jll1
,
4Z x 72.5
• .47
0
L
..
..
"
.
100
.. 10
a0
Duration of 81ast Pressure, T (rns e c) Figure 6-39
Peak blast pressure capacity for tempered glass panes, alb; l. 75. r F 5/8 and ll4 in.
.. , 6·95
0
,
0
10
., .
'.
,
I
..
,
"
L72
..
I d.
13.a
~x.2
.
..
. ..
100
,......, ·2.00 1/4
1ft.
;.
,
~
"
,
'. .
to
,
, ~
b;'. (IoU
II (pol' 4.113
12x24 ,
......
a. ....,
,
.
'.
.
'
..-...
~
..... ~ ...
'J
1~
... 0.17
,
.
"
Go
CD
.xlO
1.12
41 . . .
..... .." .....
" '"".
2.Z7
•• 12
CD
..
lIx:ll 24• •
,
. 0.1.. 1
. .
,
J,
.
.
.
,,
-
to
.
... .
-
'
'"
,
-.
1.00
1000
r ,
Go
100
2.00 ,.....• •1r1181n.
.
.
,
.
I·
.,
"
10
-
~
,
,
..
bx•
..
1~
.
..
,
.
..
.
-
."'
(1ft.)
x 24
II lpoIl
.
8.011
;
I
~
I
i"-~
,
I
,
I
1
Figure 6-40
10
100
Duration ot Blast Pressure, T (msec) Peak blast pressure capacity for tempered glas. panes, alb 6-96
I I I
111 x 31
3.10
24x.
2.Z7
31x72
31h80
1.78 1.53
42.114
1.11 .
-
1000
=2.00. e =1/4 and 5/16 in. .....--.;
..
.
tOO
0
..
,
.
,
••••
,-
.~
.." -2.00 t • :WIn.
..
..
,
-
,
..
.
. :-
,
,
'.
:
.. . It'••
"
(1".1
12. :101
-
10
» s-
,
,
.•..
,
"
I-
"
1
"....
0.
,
.
,
.,
10
..
-
--
"a
.".,~
t
.
QI
""a.
.... I:D
a.-
..
,l.
",
10
~~~ ~ <,
"
•
,
2.00
.t/2 ....
.. •
. ..
.'
..
-
1
,
<
"
.. - ..
..
'\
..
<
.1. . . .·
I.zI
24 ••
~
31,. eo·
3.:11
».72 42 ....
ZJI4
-
, -.
I"
I
I
~
!
,
: ,
I I
10
100.
Peak blast pressure capacity for tempered glass panes, a/b = 2.00.
6-97
2JlZ
,
, 1000
_•
Duretion of Blast Pressure, T (msec) Figur-e 6-'11
2IIl.I
.'
.. ,
......
12.2It
,
.-- ...
rm
• CIIoII
,. . . lin.)
<
,
.
til
.
..
0.
Z.1lI
1000
100
:
100
~
:101 • •
~ QI
, •• 31' "
~,.
"
... , ".
t
= lI8 and liZ in.
100
.
,
."" 1•
-
, ...
~ .....
10
...
. .... .
1I
CL
100
--
.
»:;..
~
,
•
100
,
.... C&I
314111;'
,',
•
~,,.
,
.,
......,
.,
12824
'j
"
.
(
21.7
t •• 31
.
.
.
I .
i
,
. , .~
'I
-
I
..
• -
I
,
, .
"
Durc~ion ot 8lcs~ '"
j "
,
"
,
,
,
--.
-
,- ..
,
100
1000.
Pressure, T , (msec) .
Peak blast pressure capacity for tempered glass panes. a/b = 2.00, t = 5/8 and JI+in.
"' ..
-.
6-98
..
42884
I
.'-
. '7.82
30 • •
31·711,
I
.
I.
,
12,2
248.
I
I
.
.•
10
-
I
..
Figure 6-42
-
z.oo
I· ,
1'00..
,
I
1000
-
~ ....
10
•
"
.
-
1'-0...
2.11
•
--,
3.n
I
, 10
~
1.:14
•• n ca,,14...._.
.
..'
1
"
I.:M
24 • • ; ...
."",.
CL
14.11
.{.
a
'"a
-"'''31
,
-,
.."
11lioii 3U
j
,
.
,
b •• (1nJ
,12.24
~
,-
,
0-
•
!118m;'
,
,
~
.
• z.oo
..
~ ~~ ,
--..
.
,,
, ;
1.44 4.ll1
, 20
,
10
,
,
i
,
11Ao1.U t.
§
.... ,
'
1"li
!
'.,
,
b •• lin...
.
- .
.
.,
3.77 ,\
" ,-
t
• (psil
12.31
"
.'
e
1''- in.
'I'
,
\
l-
1.715 '
1'.&4
"2A.n
l.eM
0.780
30.90..
.'
..
!
,
,
~
0.\
.......
~ CD
I
W
.,
..
0:
~
V1
~
0,0\
"
\
1
.
100
\0
1000
"
0: Q.
I-
V1
15
'\\"" ...
::.:
\0
Q.
I'\.
I
I i
, i !
I
I I'\~
,.
<.::1 ;;.
1\
I I
I , I
I
I
1 \
,I
.1
I I ! >1
I:,
;
,. ,.
,
!.: I !
,';
[,
I
I
II
,
I ,I i;: I
..
I
I'
!
i Iii,
10
.Ipail
I
1,1,1 ii I " .
12.,31
6.70 ,
I~
I
,
'
I
I !
!,
"
~
i :{i
."
.
:I I -
,I
,
I
I,
.~, '~!
I
BLAST PRESSURE, T (rnsec)
n
I,, il
I
!
.,;
!
I, I, I
,
.. ,
I ,_
,
, I II
24",72
I·
!,
!i 1,1
;
30.,90
I
.
r- .•
. ,1000 '
Peak bl.., pres~ur~ c.p.~ity for t.mp~r.d gi.ss pon'.s: 'oJb'~3.cio. e • 1/4 >.nd'm6 in.
6-99
1.75
iii
i ! ! !-
.
, 2.9B
I ••• 50&
!
I,
.
'
I
100
,
_.
' 'l
i
I
I
, ,<
b.IJift.)
. !' 1 I'! ! .'.
,
I
~
"
.. ,
I
!
, ' I
.
.. .
,
' I
I I
I
,,
II '''~lli'il ki,l' I !I ' I ' " ·1 I " I III ' II
DURATION of Figure 6-43
,
0,.'
~k
i
,
"
I', 1:11' .............L . 'itl·! Iii ..
,,
..
.,
.
I
,
in.
T' ,
;
3.00, t. 511.
"
I
: 'j"
I
I IJ I ,lAo·
"
..
"
w
-
.
,
..J CD
1,18
.,
20
-,
\
\0
.!
, .......
........
,
li-3/lin.
.J
C
.
I·
..
,.
b.".
I,
..
.,
. ..
\'\.
v
"
.
I T I·
' a/b-100
i
BI..I
Un.)
.. t2x 3e
I
I
lUO
I
,,
,
I
I
!
,
[
<;
,
·
.
,
'0.-
...
r-
..
... ·
\0
I
35
"
\0
~
I
..
..
, , I ~
<,
I
,
I
I
, , 11 ,
,
..
I
-,
I .
:
t2"~ 3e
t7;3
l-
. .
I',
, :
I I
I
I
i
1, I I· , •
I
I
,
II x 5!t
7.69
:lh 72
4.33
30.90
2.71
f
:
J
:I I,
I
,
I
.
I
" 10.0 .,
10
1
DURATION of BLAST PRES~URE. T
Figure
~-41l
1000
(msec)
Peak blast pressure capacity for tempered glass panes. 'alb· 3.00, t = 3/8 and 1/2 in.
, "'"
•
-. . .
.
. .
.;.,
6-100 ", ,
'. 1 "
. 'i..
-
.. .
"
.
,
'.
BI. .I
..
,
·1
..... ....;""
I
,
..
,
!"
t ...
b •• (In.)
,
.
3l1'- 9O
.
I I I I I
..
I'
~ !'-o'
.
..
alb· 3.(0 t· 1/2 in.
•
:z.at
.
1000
..
..
.
,
I
,
:-..... ~ ....
I
,24x72
. I
.'. ..
4.40
..
•
I
.
.
.
.. .-e-
I .
I
I
,
1
II~IW
...
·
!
,
,
I
,
-,.. ~
I
•.
..
..
.
.. ,
,
60
I
.
....... .....: -~ ,
-
.
loIb'.i~
.
t .'5/1 in.
b X-I lin..1
-
,
M""
-
rr.7
18~ 54
12.3
....
:
.
.
81"';1
12 x 31
'
,
,
-
,
::::~
,
10
"-
24 '.72 ~ ~
..
..
....
.-
311.90
.
-
La.
. .
-
. 10
60
-. .
100
1000
. I •
TI
J
l....
.
"'-
.
;
'.
I
1
-
r
b •• 11...1
B 1".;1
12.38
••
11.54
,11.1
,Z4 ~ 72
10.2
30 ,x 90
" 8.51
~ t--.. jo... !-. I~~
I"
'.
'" .
, 10
,
.
.
..
,
"
I
,
I I
1
II ' I
;
I Iii ill
.
Ir ,I I
.
I
I
i
.:i I
I 1
,·t'
' 1. " i i ! III
.1
I 1
, ,
. I ,I
I
10
,I
'1·1
I .
I, ,.1.1' ,
-r
! I
I .
I
'i.
I I
, ,, , I
I I
1
I':
I
1
I
~
L
i '
II i
I I I. IoIb • 3.00 I 3/0. in. . I I I
1
t •
I I
"
100
1000
. DURATION of 8l:AST P.RES.SURE,T Jmse<;) Figure 6-45
Peak blast pressure capacity for tempered glass panes, a/b ./
"
.
6-101 '
= 3.00. t = 5/8 and
- "" ... 3/4- in,
20
"loIbU.~1 in.l t· 1/4
"
\0
..
I"~
i
,
..
,
1(' Un.1
b
I
IIli ,
,
!' . ,
.....
,
~
..
-
..
I
-, "
""
-.
:L51
12.41
: ' I
I'
t'--
,
B IIlIiI
11.72
1.51
2. . . .
OJICIO
I
,
,
.
'
..
,
:11,.,20
.
, 0.020
.. '
..
.'
0.1 I
20
\0
.,
\00
\000
. ,
\0
IaIb·4.00 I I II t·
.~
I
:
!I..
' I
,,
;" I';;
I
I'
I;
I,
<;
Ili~ "
". \
I
..
I
,
[
! ! I[ 10
~
I I
I'll
I
I
I
I
I
I I I!; , I I i·j Ii I ! ..... _,
I
I I
I
, ,
"
I
I
..
..
I
I
I I
II
I
I
i I '
b •• HR.)
r
I
,
I
!' I:. '~ I,
I
: "I I I: ! i !
"
'
I
:
-
I
.
I
: ill
" , I
! -: I'
II
! I I I !I ! ..
I
I
I
2.71
IIIII
i
I I
!I I I
ii' I
II ,I
..
1.51
:11,.,20
1.01
2•• 98
1,000
\00
DURATION ot BLAST PRESSURE, T (msec) , Figure 6-46
Peak bl.., pressure capadty for tempered gl..s panes, alb. 4.00. ,
6-102
B (Illil
12.41 .. 8.25
11.72
I
"", 1111
II~"
I
I
,
, I
"";'-
I
511. i....
D
1/4 and $/16 in.
20
..
~
to
I I 1I .
.
-,
...
I\\.
,
,
~
-
,
~
...
,.. ~'
.
,
W
, , ,
Q:
,
, ,1 \:'
I
24._
~
30.120
1.50
,
', .
..
-.
/"I"
to
I
r
4,11
.
,
.
:
, 18.72
,
.
a:I
,
'.
,
-,
,,
,
..
.- .
:--... i'-o
~
=:l
,
9.24 .j
,
. ;
.
Vi
V)
'
-.
-,
..
.
B (poil
.,
, '.
~ ..........
......
,bIlIHn.)
,
<;
I.' ;
e
..
12 x 048
,x
,.
~,.
,
.
r- r--.~
r.
alb· 4.00 t·3II1n.
..
100
1000
Q:
0lV)
« -J
35
,
a:I
.
::.:: .« w
<,,,,",
,.,
I~
,
\0
;
.'
.
I ,I!; , ,
,~ .'-I
0-
I I I I
,
"
I
....
,
,
I
I","!.
,
I
i
,
.
,.
"
:
...........
1',..
I' • I
! I
I
:
I
i
-II 'I, I'
,
!! i ~
i"
I
I'i: iii,r
,.. !
Ii! I
18
lit
72
30 x 120
I
"'.
,,1000
100'
DURATION of BLAST PRESSURE,T (msec). '. ~ . ' ._ ,; .. ~ '.
Peak blast pressure capacity for tempered glass panes. a/b = 4,00,
6-103
B (poil
15.21
I
I I I iT ; I ' : II
'
II :
;.
I
, i
,
(1... 1
12... .-
1;-:
-
, I . I '.
b ••
,' ,
,
.",
1
I
'I [, 111111'
•
Figure 6-47
I -,
'!o-
,. ," ·1
t ..
I
I I[ -
Illlll· . ,II ..
,
. I
I
I..
alb·· 4.00 -t· 1/2 in.:
t
e
l/8 and 1/2 in.
7.17
2.58
II i
!
I II
I:
B I••il
i ' II! ,1,1,
I'
12 . . .
III
1'1
,!';
,'\~
I
\. "I, "
I
r I", I, ~f~\ ,~~ c, i'I' 'I' ~',N-++~-W~~ I' , I;,' ,',' ".
~-
",
I
',,'"
'I
•
Ii' Vi
III .. i i
' I
~ ell
.s '0::
l Ili I
:J
V) V)
W 0:: e,
..... V) < ..J
I
,
100
ell
::.:: < w
...
l1.
~
I' , I
I
i
I
!
, ,
I
10
,
: 5
,
i
!
i
I
) .. 1
!
,I
i· I
I i
,
I
I
•
!
, ,
I, I
I : ! 'I! I
I
. 'I
"
' , ' ,
........ I I
!
i :'
, ,
I i
i
,
., i , 'I •
, j ,,
I
•
BI . .1
12 . . .
:17,.
, i
,
i
, i
i i •
,
,
'
!
i i' !
"B" 72
II! '
I ! i' I, .l I
,
, ::i
b • • Ii,,-)
., iii I ;! :, II
t. , I.' !-!
-
!
, ~.I
,
I
'
I ,
i
,
t
!
I
,I ,i I I ! i III ,, i ' I
I
!
,
in..,
I, I
i,
:! I'i
: 1.1 I " ,...... I !
,
,
111.9
.
lI
24'. ge
' 9.0.
i
: ,, ! ! I
,
' 'I ! ~
i
.. 5.08
30 x 120
I
10
1000
100
OU&ATlON of BLAST PRESSURE, r (msec)' Figure 6-48
'
' I 'I:
,
I i' iii !i I
I
t
I
1 '1I
!• . . . . . ,I
I
I
-,3/.
aIb·4.00"
I
,
,
, .
,
I : I , !
I
III
.
, I
'
.
I
100,0
ii , II I I; I I ! I ! : i .; I I, Ii' II! I
I
I 'j 1'1
I
100
,I
I
i
,
,
I
I~I I
,
1
~
"I'
, ,
~
,
I
311.120
I
, I·
,
"
n.' '
, I
I
10
'
, I
I' j
I
~
II
:
! II
......-...
"I I .......
.1
,I
Ji 72
j
~
Peak blast pressure capacity for tempered glass pan;';,; alb .'
6 -104 •
.. 't '.
".~;
~ 4,00,
t; 5'/8 and 3i4 in:
,
'
.. ~~tl;h
:
~r
llJlHlttHhl1
::ii•
i
!I;: r-r
.
, -;
I' I
I 1
I
. I
! :iil":
"1.:::'-:
,I
,
...
.
0
1000
z
a,
0
""
Z
0
'"
r-
r717:T.:;;rU'J":'I' :1:';, "'-I ,r-~XT1:'-n7";- ~+'+t::--r:1
,..-:p,
W
~
~-.
"'il! 1':'1'''1,''01 ~.;.
VI:7fXV
II: l..J
, ..
r71YV;
'~
,,-1.---
~--'I"j:::t:t: _ -~ =L-f~,I~-···
_.:-.,.
, _-'-0.---&'--'--'" - --"~
II"
WIA1'
II;
'1"
~ t/),
~
-~
l")
co
;'
,'" '';11
I II j'r. I •1 I l I [I '
dl!
N
M
if
, ~:·\+I-H-Ihj., ,--,:.:.:J-"hh
~
:a
', "
;
I
d
•
8
,:ll--·.. ·-
-+:.;j+-l': ~+
-.--
t/)
w
:e 0
'z 0 Z
-
- _ •• ..;...-I'--I-.:t--I.-~--.-•. -.,.-.
100 I~"~~~.~"" ot... - ---,-----n";-
·r
,
--I-
M
0
I
I
I
I
I
I
I
I"
"1
.,
""
Z
l/l
.... -V··.. ··...·-I..., ..• • -: 1-: 1 · 1
I ..
l: ~:'h::I~+44M+~+j+
.• _ J...1_'_'_',,"_
100
1000
, 10,000
~D= NO\IDI~IOIlAL LOAD=t 14X la6'rlWtl4 Figure 6-49
Nondil:lcnsional static
I~ad-s(rcss rcl~tions~i'~s'for ~~mpIY'~pporl~d'(c~pcrcd glass (af(~r'M~orc).
100,000
"'" -. '
"
'~--,
·,···.·,··n.,.··.. · "
',.' .• '"
,.,
.,~,',.-
,.",
",.
,
~q
l
-+,
(/)
.(/) uJ
z
:.c
u I
I-
+
v
;;
~
b
: :;
0
~ t •:
~ ,/
v
10Im~. ~r --'-' ., --1 .. I
~rl-
..
,
:. ..1 I .' .
JJ
16'
·1:1;
,..,-"
... ' L t -. .-
.. I . L:of':'L:"
Z
o
._ ..
I-
o
'-
'", .....
uJ
-'
U.
·-01
~
=I=t
,-.-,-,1--
."
'"
a:
,'I],/)'
I
"
... -..,-
:::.,._-
_1_t--t_I_J-~--I-I-
uJ
o
o
,
-
+++..H+1
~-'W-l-+l~-W-l. I+..c...+ -I+I-I..
_._1
uJ
IZ
,~
uJ
U
u
xl ..
It'v
'/
1/
'--:;fv
i'l
~
... :ill+L kb
~~.~~,,-
-,1-
I;¥
••
,.
·t:
", ..::, 1 ...,. , 1' ..;"'' 1 I::.1: ,.. __ ... I'~-... ,_,..1,.· •
.100 .
I
I
1000' ., . , .
'
10,000
•
lw = /IDIlDlMENSl
,'''p:' "I': rrr .. ;:.: -: .. !:'J
\0
,
,
:-1- 0
:::.
... ,
--
... -
. 100,000
.
,.' "
~
,.
Ncndimcnsional SIalic load-er ..er dcf1~ction rcl.tionshipsfor simply ..."supported tempered glass (.fter Moore). ~
"
e
e
e
A· edge clearance B· bite
e . face clearance , "
,'.
Figure '6-51
Edge. f3.ce. and bite requirements,
, .."
"',
~
,-
6-107
Figure 6-52
Distribution of lateral load rransrnitted by glass pane
6-108
to
the window frame.
TM 5-1300;NAVFAC P-397/AFR 88-22
Table 6-6
Static Design Strength r u (psi), for Tempered Glass* [a - long dimension of glass pane (in.); b short dimensio~.. of glass, pane (in.)] , ,. c
:
ASPECT RATIO - 1.00 Glasa
Static
o•• ign
Strength (pai) for,'. Window Thickness, t, of
Size, 3/4 in.
12><12
141 120 103 90.1 79.2
64.S
16%16:
20_ 17_ 151 132 11_
17x17 18.18 19x19
103 91.6 82.2
20x20
' '/8 in.
'11_ in.
3/S , i n .
49.3
32.2 28.3
.70.1 62.5 56.1
.43.7 39.0 35.0
25.5 23.1 21.0
15.1 14.1 13.5
50.7 46.0
. 31.6
19.2
12.9
11.0
17.7
12.7 12.6
25<25
74.2 67.3 61.3 56.1 51.5 .u .5
10.0 9.20 8.52 7.91 7.43
14z14
ISzl'
21x21 22z22 23x23
24x24
87.7
40.3
74.7
-42.8 36.9
1/4 in.
27.5 23.9 21.1 18.7 16.7
lax!3
56.1
28.6 26.4 24.4
41,9
38.3 35.2 32 .•
16. :3
15.1
22.7 ·21.2
14.3
13.8
26x26'
43.9
30.0
19.7
13.4
9.39
40.7 37.9
27.9 26.2
18.5 17 ,4
8.80 8.26
29x29 30:1:30
35.3
24.6
23.2 21.9
32x32
33.0 30.9 29.0
12.9 12.8 12.6 . 12.6 12.0
20.8
14.2
11.3
6.71
33x33 34%34
27,4 26.0
19.7
13.8
..10.6 .
18.7
13.5
10.0 .
35.35
24.8
.17.8
13.2
9.50
36:1:36
17 .0
12.8
9,OS
37s37 38x3a
23.6 22.S
. 16,2
21.5
15.4
12.7 12.7
8.63 8.24
39x39 40%40 41:1::41 42%42 43x.3 44x44 4Sx4S 46%46 47x47 48x48 49x49 SOx50 51%51 52][52 53x53 54x54 55x55 56x56 57][57 58%58 59%59 60x60
20.5 19.7 18.8 18.1 17 .3 16.7 16.0 .15.4 14.9 14.5 14.2 14.0 13.7 13.' 13.3 13.1 12.9 12.8 12.7 12.7 ,12.6 12.6
14.8 14.4 14.1 13.8 13.5 13.2 13.0 12.9 12.8 . 12.7 12._ 12.6 12.4' 11.9
,
11.'
11.1 10.7 10.3 9.99 '9.66 9.38· 9.11
. 16.4 15.4 14.6
12._ 12.' 11.9 .11.4 , 10.9 . 10.4 . 9.99 ' ,9.59 9.24 8.91 8.'9 8.30 8.02 ,7.76 7.54 7.33 .. '. 7.13 '6.94 6.76 6.59 6;40 6.22
.
7.88 7.57 .7.30 7.04
6.80 6.'6 6.32 6.08 '.86 5.65 5.45 < 5.27 5.09 4.92 •. 76 4.61 4.47 4.33 4.20 4.08 3.97 3.85
20.2 - 17.6
15.5 14.2 13.~
.
:
7.00
'6.22 5.86
7.39 7.04
5.53 5.22 4.94
, -
4.69 4.4'
4.23 4.04 3.86 3.69
5~24
5.01 4.79 4.58 4-.39 ; 4.21 4.05 3.90 ' . 3.75 3.62 3.49 , 3,.37 •. 3.25 3.15 3.04 2.95 2.85 2.77 2.68 2.60 2.53 2'-45 ·2.38 2.32
.
- 6.62
5.77 5.50
.
12.1
7,78
6-.39 . ,6.07
-
12.7 , 12.6
,
11.8 10.9 10.1
27<27 28.28 31:1:31
*
lIZ in.
b • • (in. )
--
3.53 3.39 3.2' 3.12 3.00 2.89 2.78 2.68 -2.'8 2.49 2.41 I 2.33 '2.25 2.18. !, 2.11 I 2.05 " 1.99 1. 93 , 1. 87 1.82 1.77 1.72
,
,
,
.'
"
Panes to the right and below the stepped dividing line behave according to large deflection plate theory. (continued)
6-109
'';1
TK
5-13~0/NAVFAC,P-397/A~R~8~22
Table 6-6
...
. Static Design Strerigth r u (psi); fo~ Tempered Glass* [a - long dimension of glass pane (in.); b - short dimension of glass pane (iri.)] (Continued) ASPECT RATIO - 1.25 Static Design Strength
Glasa
Size, 3/4 in.
b x • (in. )
,
s/a in.
~
(psi) for II
1/2 in.
Window thickness, t, of
3/8 in.
--
1/4 in.
5/16 in.
12x15 13x16.25
154 131
105 89.5
65.5
37.5 32.0
20.S
55.8
14x11.5
113 .98.5
,77 .2
48.1
27.6
41.9 36.8
24.0
15.8 14.2
12.3
67.2 59.1
13.0
10.2
.' 32.6
21.1 19,0
12.0
9.86
29.1
17 .2
11.1
9.18
10.3
9.72 9.54 8.69.
15x18.75 16x20 17x21.25 18x22.5
,
86.6 '76.1
-, 52.4
68.4
46.1
'.
-,
..
19x23.75 20x25 21x26.25
61.4
.41.9
26.1
55.4
37.8
50.3
.34.3
23.6 21.4
15.8 14 .6 13.6
22%27.5 .'
45.8
31.3
19.7
12.8
18.2 17.0
12.0
23x28.75
41.9
28.6
24:1:30
38.5
26.3
25x31.25
35.5
26x32.5 21x33.75
32.8 30 .•
24.2 .. 22,4 20.8
:
28.3
' 19.5
29x36.25 30x37.5 31x38.75 32x40 33x41.25 34x42.5 35x43.75 36x45 37x46.25
26.4 24.6 23.1 21.7 20.4 19.4
18.4 17 .4 16.4
28x35
38x47.5 39x48.75 40x50 41x51.25 42x52.5 43x53.75 44x55
,
15.9 14.9
10.2 9.98
9.79 9.77 9.71 9.65
14.9 14 .2 " 13.6
12.8 12.3 11.7 11.2 10.7 10.3 10.1
16.8
13.1 12.7
9.80
16.1
12.3
9.79
15.5 14.8 14.3 13.8 13.4 13.0
11.8 11.4 10.6 10.3 ' 10.2
12.6 12.3
10.0 9.87
18.5 17.6
15.6
11.0
9.80 •
7.95
9.70 9.49 8: 77
7.30 6.75 6.27
.8.14 7.57 .7.07
5.45 5.12
5.83
6.63
-
4.81 4.54 4.314.09 3.90 3.71 3.53
6,23 5.87 ' 5.54
9.22
5.25
8.71 8.24 7.81 7.41
4. 98 ~ 4.74 4.52 4.32
9".78 9.75 9.71 9.66 . 9.47 9.06 .
7.05 6.72 6.42 6.13 5.87 5.63 5.40
4.14 3.97
.8.68
9.93
11.1
9.96 9.79· 9.78
11.3 10.7
'14.1 13.4
15.4 13.5
11.9
-
3.36 3.20
3.66 3.51 3.37 3.24
3.06 2.92 2.80 2.68 ; 2.57 2.47, 2.37
5.19
3.11
2.28
8.32
5.00
3.00
2.19
7.99 7.67 7.37 7.11
4.81 4".64 4.49 4.34 4.21 4.08 3.96
2.89 2.78 2.68 2.59 2.50 2.42 2.34
2.11
3.82
"
45x56.25 46x57.5 47x58.75 48x60 49x61. 25 SOx62.5
11.9
. 9.80
S1:ll:63.7S
10.6
52x65 53x66.25
10.3
9.79 9.78 ' 9.77 9.74 9.70
10.2
9.67
11.5 11.2 10.9
6.85 6.61 6.38
I
2.03 1. 96 1. 89 1.83 1.77' 1.71
* Panes to the right and below 'the stepped dividing line behave according to large deflection plate theory. (continued)
6-110
TH 5-l300/NAVFAC P-397/AFR',88-22.,
Table 6-6
Static Design Strength r u (psi), for Tempered Glass* [a-- long dimension of glass pane (in.); b - short dimension of glass pane (in.i) 1 (Continued) ,' ' -ir .,'
'
ASPECT RATIO - 1.50 Gla••
"
Static D•• ign Strength (p,ai ) for • Window thicknDsa. t, of
--
Size,
b x-a
12x18
,
~
123 105
13][19 • .5 14x21 -
.
90.2
78.6
15x22.5 Ih24 17x25.5
61.2 54.6 49.0 .
-,
20x30 . 2lx31. 5 22%33
.
.
44.2 " 40.1 . 36.5
,
83.8
. 52.3 44.5
71.4 61. 6~
33.4 29.4
'
23.2
33.4
20.8
.
30.2 27.4
15.6
14.2
30.7
21.0
13.1
25%37,5
28.3
19.3 •
12.4
26:1:39
26.2
27x40.5
24.3 22.6 21.0 19.7
24:11:36
28x42
29:1:43.5. 30:1:45
,
."' .
31x46.5· 32:':48 33:1:49.5 34:1:51 35:1:52.5 36:1:54 37:1:55.5 , 38:1:57
..
!
,
17.9 11.7 16.6 ·11.0 15.4 : I 10.4 14.4 ••• -~ 9.89 13.4 9.42 \
,
12.8 • 12.2 11.6 ; 11.1 10.6 10.2 9.78 9.42. 9.21
-
39:1:58.5~
18.4 17.3 16.2 15.3 '-' ~ 14.4 13.6 ;. 13.0 12.5 12.0 -
40:1:60 , 41:1:61.5 42:1:63 43:1:64.5 44:1:66 45:1:67.5 46:1:69 .! 47:1:70.5"
11.6 11.2 " 10.8 . 10.4 ; 10.1 9.71 9.42 . 9.25
9.01 8.82 8.60 . 8.35 8.12 7.90 7.69 .. 7.62
.9. O~
7.5~
.'
..1 "
:/-
.
18.8 17.1
22.8
33.4
.-
.-.'
.
.
29.9 .
16.3
11.9
13.9 12.3
10.5 9.43
11.1
10.0
8.88· 8.26
>
9.31 8.85 8.84
8.14 8.02 7.90
·
7.83 7.81
7.78 7.62.
7.80 7.77
6.45
11.4
10.6 9.86 v . 8.98 8.611 8.24 1.86
,. -
7.85 7.85 7.84 7.83 7.82 7.72 7.62 7.28 6.90 6.55 6.22 5.92,
"
" . .'
7.91 7.88 '7.85.'
7.77 " 7.69 7.62 7.35 . 7.06
.
·
9.32
,
..
--
1/4 in .
5/16 in ..
25.5 22.0 16.8 14.9 13.3 12.3
.-
8.0S 8.02 -, 7.99 • 7.96 ' 7.93
6.78
.
'
9.16 8.91 8.65 8.34 "
,
.
19.2
26.0
41.8 37.3
24.9
23x34.5
.
.
3/8 in.
...
38.4 ,
,
,
","
53.6 47.2
69.1 -
18x27 19%28,5 __
1/2 in.
5/8- in.
3/4-1n.
(in. )
48:1:72
..
·
7.03
7.77
5.95-
7.63
5.50
7.19 6.69
5.10
6.24 5.83· 5.47
4.42 4.14 3.88
5.13 4.83 4.55 4.30 . 4.07 3.85 3.66 3.47 3.33
3.64 3.43 3.27 3.13 3.00 2.87 2.74 2.61 2,50
4.74
,
5.64 '5.38 5.13 4.91' 4'.70 4.50' 4.31 4.14
3.21 3.09 2.98 2.88 2.77 2.66 2.56 2.47
2.39 2.29 2.19 2.10 2.02 1. 94 1. 86 1. 79
3.97
2.38
1. 73
,
,
* Panes to the right and below the stepped dividing line behave' accOl;ding .lO0 large deflection plate theory.
-
..
(continued)
6-111.
TH
5~1300/NAVFAC'
Table 6-6
P-397/AFR 88-22
Static Design Strength r u (psi), for Tempered Class* [a long dimension of glass pane (in.) ;'" b.- short dimension of glass pane (in.) J (Continued)
.
ASPECT RATIO· 1.75 Glaaa
Static Desisn Strensth (pail tor
II
Size, b :II: II
3/_ in.
3/. in.
1/2 in,'
5/8 in.
Window thickness, t, of --
. .
:~
5/16 in.
'1/_ in.
(in. ) 12%21
10.
13x22.75 14x24.5 15%26.25
92.6 79.9 69.6
63.2 54.5 . 47.5
16x28
61.2
17x29.15 l8x31. 5' 19x33.25 20x35
54.2 48.3
41.7 37.0
14.2
,
33.0
46.3 39 .... 34.a 2•.•
2•. ~ 22 .•
14.4
10.2
12.3
8,91
19.5
10.6 9.46
8'.01 7.32
26.0 ( 23.0 20.6
14.9
18.5 1.,7
17.0
8.52
6.83
1.85
•. 3.
....
13.2 11.8 10 .•
~.'3
7.30
,
5.76
1t3.4
29.6
3',1
26.7
21::Jl:36.75
3~.~
22x3 •. ~ 23x40.25
32.4
2".2 22.1
29.6
20.2 .
24:1::42
27.2
18.6
25x43.75 26x45.5
25.1
11.1
10.7
7.11
5.70
23.2
15.8 -.
14.7 13.6 12.7 11.9 11.1 10.4
•.•1 •. ~2 6.24
5.69
21.~
'.'7 9.37 •.• 3 8.38
5.98
~.21
8.01 7.66 . 7.35
5.82 5.74 5.73 _
4.98 4.70 4.43
4.35 4,05 3.79 3.56 3.3. 3.17
7.11 6.89 •.• 7 6.45 , 6.24 6.04. 5.87 5.80. 5.74
5.71· . 5.70
...
4.17 3.94 3.73 3.54 3.37 3.21 3,0. 2,'3 2 .•2
3.00 2 .• ~ 2.72 2.60 2.49 2.37 2.2. 2.1. 2.06
5.72.'--: 5.70 5.69 5.70
4.71 4.50 4.30 4.12
2.7L 2 .•1
1.97 1.89 1.81 1.74
27:11:47.25
20,0 . 18.6 17.4 1•. 3 15.3
28x49
29x50.75 30%52.5 3b54.25 32%56
14.4
33%57.75 34%59.5 35%61.25 36%63 37%64.75 38x66.5· 39%68.25 40%70 4b71.75
12,. 12.1 11.4 10,. 10.3 9.91 9.52
42x73.5 43%75.25 44%77 45x78.75
9.15 8.81 8.50 8.24
13.~
•
'.'3 9.46 9.02 8.62 •. 30 8.00 7.73 . 7.47 7.26 . 7.07 6.90 . 6.73
..
~~
1~.1
13.8 . 12.6
11 .s
,
,
'.71
••••
.,3. 7.87 7.43
5.73
•. 12 5.8"
5.70
5.73
5.57 5.27
~
5.71
5.66 5.45
,
~
6.49
~ .•7 ~,.3
5.44 5.26 ~,O'
4.93
,
2:~1
2.42
.
...
5.00 4.67
* Panes to the right and below the stepped dividing line behave according to large deflection plate theory. (continued)'"
6-112;' "
TH 5-1300/NAVFAC P-397/AFR' 88-22 •
Table 6-6
Static Des~gn Strength r u (psi), for Tempered Glass* [a - long dimension of glass pane (in.); b short dimension of glass pane (in.)] (Continued) .•,
ASPECT RATIO • 2 00
Static D•• igD Strenath
GtalSS
Size, b x .i. (in. )
.
3/4 in.
..
13:1:26
lIIx28 lSx30 16x32
tor a Window thickn••s, .. , of ,--
.1/~ in.
'/8 in.
. 3/8 in .
1/4 in.
5/16 in.
"
".
12x24
(poi)
97:6 . 83.1, 71.7
41:'
66.6 56.1 48.9
35.4 30.' 26.6 23.4
23.8 20.3 17.' 15.2 13.3
-
13.0 11.0
. . 7.81
9.'2 8.31 7.43
6.87 6.29 '.81
6.72 6.26 5.86 '.'1 5.19
5.40 5.03
9.0S
62 •.4
_42.6
".9
37.'
17x34
48.6
11.9
43.4
33.2 29.6
20.7
18x36 19x38
10.6
38.9
26.6
18.'· 16.6
20x40
35.1.
2~.O
14.9
31.9
21.7
13.6
8." 7.85
22x44
29.0
19.8
12 .•
7.25
4.90
4.42
23x46
26.6
11.3
6.73
4.64
4.39
24x48 25x50
24.4 22.'
18.1 16.6
10.4
•. 55
4.37
1'.3
4.47
4.32
26x52
20.8.
9." 8.84
6.39 6.08
4.40
4.24
27x5'"
19.3
28x56
17 .9
r
Zix42
.
14.2
-
16 7'·- ~
29x58
0-
15.6 14.6' 13.7 12.9.
30z60
31x62
32x64 33x66
._
~
.. ,
-.
9.98 9.36 8.80.
'8.23 7.73 7.27 6.86 6.57 6.32 .6.08
13.2 12.2 . 11 .• 10.7
12.2 11.5 10.8 10.3
8.31 7.91 7.'3
5.87 .;: 5.66 5.47
1,18
5,29
38x76 39;1:78
9.73
6.86
5.12
9.24
6~62
40x80
8.78
6.42 6.23
34x68 35x70
36x72 37x74
4lx82 4231:84
.
8.37 8.03
.
'6.05
9. '"
S.79 .
4.71 4.56
'.53
4.39
4.01
4.38
3.141
'.07 4.86
4.37
4.67
'.'8 4.52
4.2' .' '.08 . 3.8'
3.'0 . 3.28 3.09 . 2.93 2.18
l
.
.
•.•7 4,41
3,64 3.44
2.6' 2.'1
3.26 3.11 2.97
2.39 2.28
4,96-
4.40 4.39 4.38 4.37
4.81
4.34
2.72
4.67 4.60
4.30
2.60 2.50
4.25
2.84 -
• I
4.46
5.29
4.31'
..
-
.
2.18 2.08 1,98 1.89
1'.80 , I
* Panes to the right and below the stepped dividing line behave according to large deflection plate theory. (continued)
6-113
TM 5-l300/NAVFAC P-397/AFR
88~22
,..-l . .
Table 6-6
Static Design Strength r u (psi), for Tempered Glass* [a - long dimension of glass pane'(in.);' b - short dimension of gla~s pane (in.j 1. (Continued) . AsPECT RATIO· 3.00 Glasa Size,
Static Design'Strength
bx
'/4 in:
II
5/8 in.
(psi) ,for
1/2 in.
a Window thickness,:
~,'
of .;,-
5/16 in.
1/4 in .. .
19.8
10.8
7.53
16,9 14.5 12.7
9.18 7.92
6.41 5.57 4,94 4.42
'/8 in.
(in. )
81.1 69.1 59.6
55.4 47.2 40.7
15:1:45 . 16x48
51.9
17ir:51 18x54
40.4 36.1
35.4t31.2 . 27,6
19x57
32.4 29.2 26.5
12x36
13x39 14x42
34.5 29.4 25.4 22.1 19.4 17.2
9.86 .
6.90 6.06 5.43
24.6
15.'
8.79
4.92
3.61
22.1
13,8
12,4 11.3
7.89 7.12 6.46
4.48 4.10 .
'.29
19.9 18.1
2.80
24.1
16.5
10.3
5.88
3.77 3.48
22.1 20.3
15.1
9.40
5.44
3.22
2.61 2.40'
13;8
~.63.
5.06
3.00
2.29
12.8 11.8
7.95 1.35 .'
>
18.7 17.3 16.0
29x87 .
14.9 13.9
30x90
13.0
8.86
31x93
12.2
8.'0
32x96
11 .•
7.79
33x99 34x102
10.7" 10.1
J.32 .
4.72
6.90.
4.48
20x60
21x63 22:11:66 23][69 ' 24x72
2SxlS 26x78 21x81
28z84
45,6
10.9 10.2
9.48
.
,
,
4.71 . 4.40
2.82 2.66'
3.98
3.01
2.15
6.82
4.13
2.51
2.08 2.01
6,34
3.88 3.6.5
2.'8 2.26
1.95 1. 89
- 3.44
2.14
1. 83
'.25
2.08
1.80
3.08
2.03
2.93 2.80
1.97
1. 78 1. 16 1.77
. 5.91
.
11.1
•
5':56 . 5.26 4.98
1.92
* Panes to the right and below the stepped dividing line behave according to large deflection plate. theory. ', -
6-114
,
, , ~
, .1:
,
'.
TN 5-1300/NAV,AC P-397/AFR.88-22
Table 6-6
Static Design Strength r u ·(psi), for Tempered Glass* [a - long dimension of glass pane (in.); b - short dimension of glass pane (in:) 1 (Continued)
,
ASPECT RATIO· 4.00 Glasa
Static Destan Stransth (pai) for a Window thickness,
Size, b x e.
3/4 in.
t,
of --
1 "
.5/8 in.
1/2,1n.
3/8 in.
5/16 in.
1/4 in.
(in. ) 12x48 13%52
75.7
51.7
32.2
64.5
44.0
27.5
Hz56
. 55.6
38.0 33.1 29.1
23.1 20.6. 18.1
16.1 14.3 12.9
15x60
48.5
16z64
42.6
17:1:68
37.7
18z72 19x76'
33.7
30.2
25.8 23.0. 20.6
18.5 15.7
10.1 &.57
7.02
13.6
5.16.
11.8
7.39 6.43
10.4
5.66
3.99 '
9.20 8.20
5.01
3.56
4.49
7.36
4.05
3,19 2.87 '
5.99
4.52 '
20x80
21.3
18.6
11.6
6.65
3.67
2.60
21%84
24.1
10.5
6.03
22.5'
23x92
20.6
14.1"
9.59 8.77
5.49 5.03
3.34 3.06
2.37
22><88
16.9 15.4
2.81
2.02
24x96'
18.9
12.9
8.05
17 .5
11.9
7.42
2.59 2.39
16.1
11.0
6.86
4.63 4.28 3:97
1,88
25x100 26xl04
2.23
1.66
27xl08 28x112 29x116
15.0 13.9 13.0 12.1
10.2 9.49 8.85
6.36
3.70
2.09
5.92 5.52
3.'5
1:96
3.22
8.21'
5.15
3,02
1,84 1,75
1.57 I.U I.U
30x120
,
2.18 1.16
1,34
* Panes to the right and·below,·the stepped-dividing line behave'according to large deflection plate theory.
6-115
TK 5-l300/NAVFAC' P-397/AFR 88-22
.. .
..t
Table 6-7 Minimum Delign Thicknell. Olelrencel. and Bite Requirement.
-
Glass Thickness (Nominal) in mm 5/32
-,
4.0
3/16
5.0
1/4
6.0
3/8
10.0
\ MA" Minimum Edge Clearance (in)
Actual Glals Thickness For Design, t (in)
0
1/2
1/8 '
0.180
3/16
1/2
1/8.
0.219
1/4
1/2
1/8
0.355
51.16,
.
"
.
1/2
12.0
5/8
16.0
3/4
19.0
, •
.
.
3/16
0·f49
,.
..
"CO Minimum Face 'Clearance (in)
"B" Nominal Bite (in)
0.469
3/8
0.59"
3/8
1/2
.
3/16,
1/2
.
1/4
.,
1/4
,
.
1/2
'
0.719
3/8
1/2 "
.
.'
6-116
.
5/16
TN , S-1300/NAVFAC P-397/AFR : .. 88-22
Table 6-8
Maximum (bIt) Ratio for Linear Plate Behavior Under Blast Load and Coefficients for Resistance and Deflection and Fundamental Period of Simply Supported Glass Platas Based on Small Deflection Theory (No Tensile ~embrane Behevior) C -
. .'
,
Design Coefficients . ,
Aspect Ratici, alb
Maximum (bIt) Ratio ..
,
' Fundamental Period of Vibration, , CT
5.79 x 104
'Design Deflection, \C D 2,58 x 10. 4
59.0
4.42 x 104
2.72 x 10. 4
6.30 x 10. 3
1.4
63.9
'3.68 x' 104
'2.83 x 10- 4
7.21 x 10- 3
1.5
66.2
2.88 x 10- 4
7.60 x 10- 3
1.6
67.9
3.22 x 104
2.91 x 10- 4
7.99 x 10- 3
1.8
71.3
. 2.91 x 104
2:98 x 10- 4 .
8.65 x 10- 3
2.0
74.7
'2.72 x 104
3'.02 x 10- 4
9.23 x 10- 3
3.0
84.3
2.32x 104
3: 12 x' 10- 4
10.12 x 10- 3
4.0
89,4
2.24 x"10 4
3.15 x 10- 4 •
10.36 x 10- 3
89.4
2.24 x 104
3.15 x 10- 4
10.44 x 10- 3
1.0
53.6
1.2
!
~
-,
.'
..
Design Re~1st~nce,
Cr
3.36 x 104
•
;
I
6-117
5.21 X 10. 3
TH 5-1300/NAVFAC P-397/AFa 88-22
.'
'
Table 6-9
Coefficients 'for' Frame Loading' , ," <-
alb
CR
Cx
Cy
1.00
0.065
0.495
0.495
1.10
0.070,
0.516
,0.516
1.20
0.074
0.535
0.533 •
1.30
0.079
0.,554
0.551
.
1.40
0,.083
0,.570
0.562
1. 50
0.085
0.581
0.574
1.60
0.0~6
0.590
' 0.583
1. 70
0.088
0.600 , ',0.591
1.80
0.090
0.609
0.600
1. 90
0.091
0.616
0.607
2.00
0:092
0.623
0.614
3.00
0.093 '
0.644
0.655
4.00
0.094
0.687
0.685
, ,
TK S-1300/NAVFAC P-397/AFR 88,22
Table 6-10
Statisti~al
Acceptance
.a~dRejection
,
Coefficients
,
,
Number of Window Assemblies n. ,
.'
.
2 3 4 5 6 7 8 9 10
"'12
. . "
,
,. , '
,
13
,
14 15
16 ,17 ·18 19 , .20
,
.>
.
'.
.
"
...
Coefficient B
,
4.14 , 3.05 2.78 2.65
.546 .871 1.14 . 1.27·
2.56 2.50 2.46 2.42 2.39, '
1.36 . 1.42 1.48 1.49 1.52
.
,.
,
2.37 ' 2.35 2.33 2.32 2.31 ,. 2.30 2.28 2.27 2.27 2.26
22 23 , 24 25
2.25 2.24 2.24 2.23 2.22
30
2.19
40
2.17
50
2.14
21
,
a
"
,U
. ,.Rejection
Accept!,nce Coefficient
6-119
1. 54 1. 57,' 1. 58 1. 60 1. 61
,
1. 62 1.64, 1.65. 1.65, 1.66
.
-
,
,
1.67 1.68 1.68 1.69, 1. 70
; .
,
"
-
1.72
,
.. 1. 75
..
..
-
1.77
TH 5-1300/NAVFAC P-397/AFR 88-22
6-120-
TN 5-1300/NAVFAC P~397/AFR 88~22
6-121'
TH
5-1~OO/NAVFAC
P~397/AFR:88·22
r,
. ,
.
~
Fundamental Period of Vibration. Tn· for Monolithic Temp~re'd 'Glass (Continued)
Table 6-11
A.p.c~ R.ti~
Fund~nt.l
Gla.. Dimenaiona (in. )
•
b
12 13
18
14
21 22.5 24 25.5 27 28.5 30
15 16 17 18 19 20 21 22 23 24 25 26 ' 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43
"4546 47 46
19.5
t31.5
33 34.5
36 37.5
, 39
40.5
.42 43,5 45 46.5-
48 49.5 51 52.5 54 55.5 57 58.5
60 61.5 63 64.5 66 61.5 69 10.5 72
•
1.~O
Period of Vibration (ma.c) for Window Thickn •••• t. of ~-
"
3/4 In.
--1.59
5/8 in:
3/16 In. "-
3/8 in.
---'
1/4 in.
---
2.86
3,22 3.78
2.62
3.31
3.80
".38
5.84
3.00
5.03
6.56
2.82
3.4Z
3.19 3.57
".33 4.89 S.U'
5.12 6.46
7.32
8.70
3.86 4.33
8.09
7.23
9.37
3.98 :4.41
4.82 5.3.
6.10 6.76
1.93
10. 3
8.65
5.89
7.46
9.40 -
11.2 12.1
9.63 10.5 11.3 12.1 13.0
6.416
8.18 8.94
10.2 10.9
13.0
14.1.
13.8
15.1
1.86 2.16.
2.48
4.86 5.34 .:
1.92 2.26
1/2 in.
--2.43
5.83 6.35
7.06
6.89 7.46 8.04
8.34 9.02 9,73
11. 9
8.65
10.5 11.2
12.6
9.28 9.93
10.6 11.3 12.0' 12.8 13.5 14.3 15.0
15.7 16.4 17.2 17.9 18.6
19.4
20.2 21.0 21.1 22.4. 24.5
7.69
12.0
12.7 13.4 .
14.1 14.9 15.6 16.4
17 .2 18.0 18.7
20.5 21.4 22.3 23.2, 24.2 25.1 26.0 26.9 21.8,
9.11 10.4
11.1 ~
13.4 14.2
14.9 16.6 17.5 18.4 19.3.20.2
6.61 7.79
~
12.4
14.6
16.2
15.5
17.3
16.5
15.1 ;
17 .6
18.4 19.6
16.0 16.9
18.6
17.8 '.
20.8 21.9 23.1. 24.2 25.4
19.3
20.2 21.1 22.1 23.2' ,
24.3'
25.3'
20.7 21.9
19.7
18.5
22.1 22.9 23.8 24.5 26.2 27.1 28.0 29.0 30.0 31.1
5.09 5.83
13.3 H.2
21.2
25.3
4.36 5.12
-
,
26.4
27.5 28.6 29.7 30.8 32.0, 33.1 34.3 35.5 . 36.7
...
"
23.1
24.3
25:5 26:7 27.8 29.0 30.3
26.6
27.8 28.9 30.2
31.3 32.5 33.7 34.9 36.1. 31.4 38.7 40.0 41.3 42.6
31.5 32.8 34.1
.,
35,4 36.7 38.1 39:4 40.8 42.2 43.6 45.0 4~6 .4 .
6-122
.
-'
TN 5-1300/NAVFAC P-397/AFR·88-22'·
6-123 .
TK 5-l300/NAVFAC P-3'97/AFB.' 88-22
-,
Table 6-11
Fundamental Period of Vibration, Tn' for Monolithic Tempered Glass (Continued) Aap.c~ R.~io ~
Gla.,
Fund.m.nt.l Pariod ot Vibr.iion (ms.c) for Window thiekn.... t. of --
Dimension. (in. )
•
b
12 13 14 15 16 17 18 19 20 21 , 22: 23 24 25; 26' 27 28 29 30 31 32 ,33
34 35 36 37 38 39 40 '41 '42
24 26 ,28 30 32 • 34 ' 36 38 ' 40 ' 42 44 46 48 50 52 54 56 58 60 62 64
' 66 .68 70 72 74 76 78 80 82 84
2.00
3/4 In.
--1.78 2.09
2.43 2.78 3.17 • 3.58 4.01 4.47 " 4.95 5.46
5.99 6.55 7.13 7.73 8.37 9.02 9,70 . 10.4
., 11.1 11.9 12.7 13.5 U.3 15.2 16.0 16.9 17.9 18.8 19.8 20.8 ' 21.7
'/8 in. ' 2.16 2.53 2.94 3.37 3.83 4.33 4.85 5.41
1/2
~n.
2.73 3.21 3.72
,
... 27 ... 86-
6.61 7.25 7.92
5.48 6-:15 6.85 7.59 '8.37 9 :18 10,0
8.63
. 10.9
5.99
9.36 10.1 10.9 11,7
11.9 12.8
13.8 14.8
12.6-
15.8
13.5 . 14.4 15.3 16.3 17.3 18.3 19.2 20.3 21.3 22.3 23,4 24.5 ' 25.6
16.8 17.8 18.9
20.0 21.11
·22,1 24.0 25.2 26.5 27.7 29.0 30.3 31.5
6-124
3/8 in.
3.61 4.24 4.91 5.64
: 6.42 1.24 8.12 9.05 10.0 11.09
' 12.0
.:'1.3.0
14.1 15.2 16.2 . ~ 17.9 . 19.1 20.3 21.6 22.9 24.1 25.3 26.6-
27.8 28,9 . 30.1 31.2 32.4 33.5 34.7 35.8
S/16 in.
4.89 5.74 6,66 7.63 8:'61 9.64 10.7 11.8 . , 13.3 14,5 15.7 • 17.0
18,3 19.5 20.1
21,9 23.0 . 24.1
25,3 26.4 21,7 29.1 30.5 32.0 33.5 35.0 36.5 37.9 39.4 40.9 42.4
'~/4
in.
--S.85
6.82 .. 7.84
,
8.89 9.96 11.4 12.7 13.9 15,2 16.4 "17 .6 18.8 19.9
21.1
22.2 .
23.5 2".9· 26.4'
27.9 29.3 30.8 32.3 33.8 35.3 36,8 38,4 40.0 41.6 43.2 .... 9 46.6
e
TN
5~1300/NAVFAC
P-397/AFR.88-22
Table 6-11 Fundamental Period of Vibratior.., Tt:\' for Monolithic Tempered Class . .;:, ..:r.,:'''' '.,' (Continued)I, ' .. " "
.
Aapect RAtio· 3.00
rund~t.a1 Period of Vibrat.ion
. Gi.... DiiDensiOn.
t
; (in.)
•
b
12' 13" 14'
'15' ,16, .17. 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34
~/~
in.
36'
1"2.02"
,:: 424548
2.37 2.75 3'.16 3:60. 4.06
.~.~ 39
51 54 .57 60 63 66 6'9 72 75 78 81 84 .87 :90 93 96
. "99 102
4.55
5.07 5.,"Z, 6.19
6.80 7.43 8.09 8.78 9.49 10.2 11:0 11.'8 12.6 13.5 14.4
15.3 16.2
-
,./8 In.
2:4.
2.87,
3:33 3.83 4.35 4.91 5.51 6.14 6.80 7.50 8.23 8.99 9.79 10.6 11:5 12:4 13.3 H.3 13:3 16.3 17 .4
.. ··18.5· . 19.7
Thick~.~•• !
__ 1J:~
in.
!- •. of--
3/~ ~n. "
'3'.10 3.64 4.22 •. 84 5.51 6.22
.4;10
6.98,
9:2.2
7.77 8,61.
10.3 11.4 12.6 13.8 15:0 16.3 17.6
9.50
10.4 11.4
12.4 13:5 14.6 15.7 16.9 18.1 19.3 20.6 21.9 23.2 24.6
6-125·
(m-oc')
for Window
.'
.
, ~(~6
.~!' .;
5'.55'
~:81'
":51
5.58 6.40' 7.28 8.22
7.55' 8.67 9:87 . 11.1 12.4 13.7 15.1
19.0
20.4 21.8 23:'3 24.8 26.9 28.6 30.3 32.0
9'.02 . 10.3 11.6 13.1 .14.5 -16.4
18.1 -19.8
,8:, 20.1 2i.7 23.5 25.3 27.1 28.9 30.8 32.8 34. ,. 36 :5 40 .•
6.64 7,.79
16.6
~38.4
1/. In.
21.6 ,23.4 "25.3
...
27.2 29.1 31.0 32.9 34.9 36.8 '38.7 40.5 42.3 44.1
TH 5-1300/NAVFAC P-397/APR 88-22
Fundamental Period of Vibration. Tn' for Monolithic (Continued)
Table 6-11
T~mpered
Aapect Ratio'· 4.00 . Fund~nt.l Period'of Vibration ems.c) for Window ~lcltn•• ~. t, of -:-
Gla•• Dimension. (in. )
•
b
12 13 14 1S 16 17 18 1. 20 21 22 23 24 25 26 27 28 2. 30
·,t.
3/4 in.
5/8
in.
1/,4 in.
S. 7S"
6.87
S.77
6.14 7.82
8.07 9.36
6,'63
8.98
10.7
7.54
10.2 11.S·
12.2
12.9 14 ..3 15.9 17 .4
15.3
. 3/8 In; 4',24 4.98
2.09
""2:;T
3.21
2.46
2.97
3.77
2.8S
3.4S 3.96
4 . .51
4.37 S.02 S.H
S.09
6.~4
,_
'5/16 Jin~
1/2 in.
48 S2 S6 60 64 68
3.72 4.20
72
4.71
5.70
76 80 84 88 '2 .6 100 104 .108
5.25
6.35
8.05
10.6
5.82
1.04
8.92
11.8 .
6.41
7.76
9,83
7.04
8.52
10.8
7.69
9.31
11.8
8.37 '.09
12.8 13.'
'112
11.4 12.2
10.1 11.0 11.9 12.8 13.8
13.1 14.3 15.6 16.' 18.3 19.8 21.'3
35.8
18.8
22.9 24.5 26;2
30.7
14.8
32.8 35.0
38.2
116 120
3.27
9.83
10.6 13.1
lS.8
7.22,- .
15.1
16.3 17.5 20~
r
6-126 ' '
8.S1 9.S.
13.7, 17.0
'18.' '20.8
1'.1 .
22.8
20.8
24.8
22.8 24.7' 26.6 28.6
26.' 2'.1 31.3 33.S 40.6
cl
Glass
TH 5-l300/NA;VFAC P-397/AFR.8Sc22
Table 6-12
Effective Elastic Static Resistance, reff .,
A~p.ect
Glus
12 13
3/4
in.
5/8" 1~.
,1/2 in.
3/8 in. 50.3
175,
140. 119.
87.7
13
74.7._
42.8.
14
14
151.~
103.,
36.9
15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
15 16 17 18 1. .20 21 22 23 24 25 26, 27 28
131~
115.
90.1 79.2
102.
70.1
64.5 56.1 49.3 , 43.7 39.0 35.0
29',
35.3
24.6
33 34 35 36 37 38 39 40
·30 31 ,.32. '.33' ~" 34 35 36 37 ·38\ 39 40.
41
41
42 43
42 43
32
"
45 46 47 48 49 50 51 52 53 54 55 56 57 58 5' 60
"
45 46 47 48. 49. 50 51 52 53 54 55 56 57 58 59 60
.
Glaaa Thickness. t. of
(in; )
• -12
"
Effeclive Elaatio Static Resistance (psi) for
Dimenaiona b
Ratio - 1
- .'
206.
91.6 82.2 . 7.1t.2 :
.,
62.5 . 56.1 50.7
67.3 6L3 . 56.1 '.'
46.0 41.9 38.3
51.5 47.5
35.2
2'2.7
32.4.
21.2
40.7
23.0 27.9
18.6 17.3
37.9
26.2
43.9
33.0 30.9 :
of.
23,'2
.,.
29.0 ."1 27.4 26.0 24.8 23.6. 22.5 21.5 19.4 18.5 17.6. 16.8 16.1 15.4 14.8 14.2 13.6 , 13.2 . 12.8 12." . ~ 12.1 ' .. 11.7 11.4 11. 2 • 10.9 10.7 10.5 10.4' . 10.3 10.2
.
21.9. 20.8 18.5 17.5 16.6 15.7 14.9 : 14.2 13.5 13.0 12.6 12.1 11.7 .. 11.4 11.1 10.'8 10.5 10.4 10.3 10.2 9.96 9.56 9.19 8.84 8.52 8.21 .7.92 - 7 :64 7.40 7.10
,
-
27.• 5.
32.2 .
28.3 25.5
23.1 21.0
' 26.4
15.1 , '13.9
24.4
;,n:
.19.1
21.1_ 17.5 ..'
15.5 13.8 < 12.6., - .
16 .•
28.6 :~~'i
1/4
23. 9,~
18.0
.31.6
5/16 in.
16.3 14.2 12.1 11.6, 10.7 10.3 9.75
11'.7. ;. 11.0 10.4 10.2 9.51
8.76 7.93
.7.20" 6.57
12.9 12.2
8.71 8.01
5.56
11.5
7.39
16.2
11.0 10.5 .
6.83 : 6.33
5.1' 4.87
15.1
10.3
5.89..
4,25
14 ..2
10.2 9.61
5.53 5.22 ,. .95 : 4.68 ... 4.42 : 4.18 3.96 3.75 3.57 3.39 3.24 3.09 2.96 2.83 2.70 2.59 2.48 2.38 2.29 2.20 2.11 2.04 1. 96 1.90 1. 83 1.77 1. 72 1. 66 1.61 1.56 1.52
3.98
13.4,. 12.8 12.2 11.7 .r • 11.3 10.9 10.5 10.4 10.3 10.1 9.59 9.12 8.69 8.29 7.92 7.57 . 7.24· ... 6.93 6:64 6.37 6.11 5.87 5.67 5.48 5.30 5.14 4.99 4.85 4.70 4.55
.,
6-127
9.00 8.45 7.9.5 7.49 7.06 6.67 6.31 5.98 5.70 5.45 5.22 5.02 4.83 4.6,3 4.43 4.2.5 4.08 3.92 3.77 3.63 3.50 3.37 3.26 3.15 3.05 2.95 '2.85 2.76 2.67
"
6.01
4.55
..
3.74 3.51 3.31 3.14 2.98 2.82 2,.67 2;54 2.42 2.30 2.19 2.09 2.,00 1. 92 1. 84 1.77 1. 70 1.64 1. 58 1.52 1. 47 1.42 1.38 1. 33 1.29 1.25 1.21 1.17 1.14 1.11
P-397/AF~
TH 5-1300/NAVFAC
88-22
"
Table 6-12
Effective Elastic Static Resistance, reff (Continued) A.pact Ratio
Effective EI ••tic Static R.ii.tance of' Gl... Thickn... '.
Gl ... Dimenaions (ln~)
12 13 1. 15 16 17 18
••.5 46 47
'8
1/2 in'.
3/8 in.
154.
105.
65,S
37.5
131. 113.
89.5 77.2
55.8
32.0
17.5
27.6
18.15
98.5
67.2
48,1 41.9
20
86.6
59.1
36.8
2'.0 21.1
11.9
8.91
21.25
76.7 68.4
52.4
32.6
19.0
10.8·
8.36"
46.7 41.9 37.8 3'.3 31.3 28.6 26.3
29.1
17.2
9.79
26.1 23.6
15.8
8.99
14,6
21.4
13.6 11.6 10.8
8.50 8.19 . 8.07
8.12' 7.86 7.49 6.73 6.11
27.5
28.75
30 31.25
32.5
55.4
50.3 45.8 41.9 ~ 38.5 . ~ 35.5
24.2
33.75
22.' 20.8
35
28.3
36.25 37.5
26.'
U.S 18.4
24.6 23.1
19.7
18.2 ."
17 .0
15.9 . 14.9 14,1 13.4
11.7 . 11.0 10.'
5/16 in.
10.1 9.40
8.89 8.53 8.25 8.14
. 15.2
15.8 14,2
11.0;
5.57 5.09
5.79
4,00
5.37
3.73 3.49 3,28 3.09
4.99
4.65' ! . ".34 :
21. 7
9.91 9.41
42.5
14.9 14.2
8.99
6.15
'3.75 .5 46.25
18.5 17.6 16.8
13.6 12.0 11.5
8.70 8.'5 8.25
6.35
3.43
5.99
3.26 3.10
47.5 48.75
16.1
11.0 10.6
8.16 8.09
10.1 9.72
7.98 7.81 7.66
50 51.25 52.5 53.75 55 56.25 57.5 58.75
60
"5051
61.25 62.5 63.75
52 53
65 66.25
15.5 14.8 ".3 13.8 13.'
9.34
9.01 8.78 8.57 8.38
11.8
11.' 11.0 10.6 10.3 9.93 9.60
8.24
8.17 8.11 8.06 7.93
9.29 9.02 8.82
7.80 7.68
7.43 7.06
4.57 4.34 4.14 3.96
6.72 6.'2
3.79 3.63
5.63 1
3.36 3.23
6.14 5.88 5.39
5.18 4.97 4.78
.
6-128
3.49 3.12 3.01 2.90
2.80
,
1.41t 6.81 6.27
20.4 19.4
5.66 5.35 5.07 '.81
13.5 9,84
7.79
8.0' 7.82 7.62 . 7.21
41~2S
1/4 in.
20.5 17 .9
17.' 16.' 15.6
'0
,
. 61.'
32.8 30.'
38.7,5
32
33 3. 35 36 37 38 39 '0 41 .2 '3
5/8 in.
22.5 23.75 25 26.25
19
for
3/. in:
15 16.25
20 21 22 23 2. 25 26 27 28 29 30 31
(pd)
"
...
•
b
··~.25
4.07 3.84
3.62
2.96
..
2.82 2.69 2.57 2.45 ' 2.34
2.25 2.16 2.07 1.98 1.90 1.83 1.76 1.69 . 1. 63 : 1.57
4.68 4.31
2.92. 2.16 2.61 2.47 2.34 2.22 2.12 2.01
1.91 1.82 1.74 1.66
1.59· 1.52 1.46 1.40 1. 34 1.29 1.25 1.20 1.16
1.12
.. .'$
TH 5-1300/NAVFAC P-397/AFR 88-22
Effective Elastic Static
Table 6-12
Resista~ce,
",'';:'"
..
':eff (Continued)
,Aspect Ratio· 1. 50 Effective Elastic Static ReaistanJ. Glasa Thicknes., t, of
Gla.a Dimensions (in,;')
b
12 13. 14· 15 16 17
18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41
42 43 44 45 46 47 48
3/4 in.
. e.
'-.18 19.5 21. 22.5 24 25.5 27
5/8 in . .
123.
83.8
105.
71.4 61.6
90.2
112 in. 52.3 404.5 38.4 33.4
3/8
(psi)
25.5 22.0
1/4 in .
5/16 in.
in.~
29.9
for
... ,.'
16.3
13.9 12.3 11.1
.s
10.1
~.
11.9 10.5 9.43 7.86
69.1
53.6 47.2
61.2
41.8
54.6
37.3 33.4
23.2
49.0
20.8
14.9 13.3 12.3
30
44.2
30.2
18.8
11.4
6.71
.6.23
31.5
40.1-
21.4
17.1
10.6
6.37
5.91
24 ..9
9.86 9:32
5.36
22.8
15.6 14.23
6.16
6.35
30.7 28.3
21.0
13.1
7.99
6.21
19.3
7.56
5.94
39
26.2
17.9
12.4 11.7
7.13
5.51
40.5
24.3
16.6
11.0
6.75
5.06
3.:47'·
42
22.6.
15.4
10.4
6.46
4.68
3.21
21,0 19,1
14.4 13.4
9.89 9.42'
6.28
6.33
4.35 4.05 3.78 3.54
2.99
6.18
2.79 2.6l.
3.32
2.33
78.6
28.5.
33 34.5 36 37',5
-
-43. 5~
36.5 33.4
29.4 26.0
19,2
16.8
45 46.5 48
18.4
12.8
9.16
-11.3
12.2
7.90
6.32
49.5
16.2
7.57
51 52.5 54
15.3
11.6 11.1
14.4
10.6
13:6 13.0 12.5 '.' 12.0
10 :2 9.78
6.11 5.92 5.59 5.24
11:6
8.03
11.2 10.8
7.78 7.52
10.4
7.26 7".02 6.79 6.57 6.45 6.33
55.5
57 58'.5 60 61.5 63 64.5 66 67.5
69 . 70.5 72'
10.1 9.71 9.42
9:25 8.12
9.42
9.21 ,
7.24
,.
6.90\ 6.66 6.46 6.32 6.20 6.22 6.33 6.36 6.20
6.05
7;16 6.56,
7.83" 7.24
6.25 6.30,
4.86
.,
4.44 4,06 3.75
2~45
3.12"
2.21
2.93
2.10
2.77
2.00
2.62 2.49
1. 90 1.81
2.38
1.72
2.28
3.98 3.78 3.60 3.43
2.18 2.09
1. 64 1. 57 1. 49 1.42 1. 36 1. 30 1.25 1.20 1. 15
5.91
3.27
3.13 2.99 2.86
6-129
9.31
4.94 4.67 4.42 4.19
5.65 5.39 5:14
,.
2.01 1.92 1. 84 1.77 1. 70 1. 63
TM 5-1300/NAVFAC P-397/AFR 88-22
,,1
Table 6-12
'.
.
,~.
,~'"
Effective Elastic Static Resistance, reff (Continued) , 'Aspect Ratio - 1.75.
.
Dimensions (in. ),
•
b
12
21
13
22~75'
14 15 16 17 18 19 20 21 22 23 2. 25 26 27 28 29 30 31 32
24.5 26.25 28 29.75 .. 31.5
33 34 35 36 37
3. 39 '0 41
..
'2 43 45
33.25
35 36,75 ·38;5 40,25
.2 43.75
45;5 47:25 .9 50:75 52.5 54.25 ,56 ,57.75 ,59.5 61.25 63 64 . .7 5 66.5 68.25 70 '71.75. 73.5 75.25 77 : 78: 75
~.
" Effective Elastic Static- Resistance (psi) for Glasa Thickness, <. of
Glass
'I'
Ln r-
5/8 in.
109.
74.2
92.6 79.9
63.2
54.5 47.5
69.6 61.2 54.'2 48.3
41.7
37.0 33.0 29.6 26.7 24.2 22.1
43.4' 39.1
35.5 32.4' 29.6 27.2 25.1 23.2
21.5 20.0 18.6 17 .4 16.3 15.3 14.4 13.5 12.8 12.1 11.4 10.8 10.3 9.91 9.52 9.15 8.81 8.50 8.24
20.2
':
18.6 17.1 15.8 14.7 13.6 12.7 11.9 11.1 10 .• 9.93 9.46 9.02 8.62 8.30 8.00 7.73 7.47 7.26 6.51 6.29 6.07 5.86
1/2 fn;-----· 3/8 in ..
--_.
46.3 39.4 34.0 29.6
~
...
26.5 22,6
~i
19.5 17 .0
26.0.
14.9
23.1 20.6 18.5 16.7 15.1
13.2 11.7 10.67' .
..
9.11
:
13.6\
12.6 11.6 10.7 9.97 9.37 8.83 8.38' 8.01 7.66 7.35 7.11 6.27 6.00 5.74 "5.51 5.30 5.10 4.97 4.85 4.75 4.70 4.67 4.64
5/16 "in.
,
,.
8.96 6.36 7.87 7.43 7.11 6.17 5.82 5.52 5.23 . 5.02 '.86 '.73 4.68 4.64 4.56 4.46 4.36 4.18 4.01 3.86 3.72 3.53 3.36 3.20 3.05
8.91~~ 7.32+'
7.85 1.30 6.27 5.79
5.65 5:19 4.89
'"
~
'.
;
6':20'
4.70 4,64
...
,..50' ,.. air
. 4.84
,.;
.. ,
8.01.
5.05
.,
1/. ~~.
9.46 8.52
4.70 4.65 4.54 4.41 4.19 3.96 3.77 3.53 3.30 3.09 2.90 2.73 2.58 2.44 2.32 2.21 2.11 2.02 1.93 1.85 1.78 1.71
"1
10.2'
14.4 12.3 10.6
5.38
,'' '
,
"!' ./
.
4.02'; 3.79'::
)
3:5i 3.24 3.00 2:78' 2.60· 2.'3 2.29 2.16 2.05 1. 94. 1'.85{ 1. 76 1.67 1.59 1.51 1.44 1. 38 1.32' 1;26 1.2'0 ~
,.'i..
, ,-
,'-
6-130
.1
TN S-1300/NAVFAC P-397/AFR 88-22
Table 6-12
Effective Elastic Static Resistance. reff (Continued) Aspect Ratio - 2.00
..
."
..
Effective Elastie Static R•• istance (pili ) tor
Gla..
Gla••
Dimenstons
~lckn••••
t, ot
(in. )
-
b
12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42
0,
24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70
3/4 'in.
s/8.in::- t. 1/2 In.
.,3/8 in.
97.6 ..83.1
66.6 56.7
U.S 35.4
23.8 20.3
71-.7
48.9 42.6
30.S
26.6
17.5 15.2
.62 •.4
54.9
37.5
48.6
33.2
23.4 20.7
43.4 38.9 35.1 31.9 .29.0
29.6
18.5
26.6
16.6
5/16 in. 13.0
6.87 6.29
13.4
7.43
5.81
11.9 10.6
6.72 6.26
4.98
;5.86 5.12
4.19
4.12
3.74
24.0
,14,9
21.7 19.8
;'13.6
1.85
12.4
7.25
4.40
3.59 3.48
3.42
26.6
18.1
11.3
6.73
16.6 15.3 1'1.2 13.2 12.2
10.4
6.39 6.08 5.79 5.15
3.92 3.75
17.9
4.54
9.49 8.56
24.4 22.5 19.3
9.05 7.81
11,0 9.52 8.31
,. .11
20.6
1/4 ,In .
16.7 15.6
10.7
14.6
9.98
13.1 12.9 12.2 11. S
9.36 8.80
72
10.8
74 76 78 80 82 84
10.3 9.73 9.24 8.78 8.37 8.03
7.53 7.18 "6.86 6.62 6.42 6.23 6.05
11.4
8.31
7.91
9.56 8.84
8.23 7.13
4.85 4.58
7.27 6.86 ·6.57 .6.32 6.08
4.36 ,4.15 "3.99 3.86 3.74
5.87 5.66 5.07
.4.84 4.64 4.46 ·4:30 4.14 4.02
6-131
3.64
+
3.56 3.49 3.45 3.41 3;36 3.30 3.25
3.93
3.33 3.24
3.62
3.04
3.51 3.45 3.40
2.81 2.61
3:32
2.42
3.25 ·3.10
2.21'
2.14 2.02
.2.90
1.91
2.12 2,56
2.41
2.28 2.17 2.06 1.97 1. 88 1. 80
1.61
,
1.72 1. 63 1.56 1. 48 1. 41 1. 34 1.28
P~397/AFR
TM 5-1300/NAVFAC
88-22
,
"
Table 6-12
"
.
,';
,
,
Effective Elastic Static Resistance, reff (Continued) Aspect Ratio - 3.00
GlasB
Effective Elasiic'Static Resistance Glas's Thickn'ea., t' of
DimeI:lSions (in. )
b
•
12 13 14 15 1. 17 18 19 20 21 22 23 24 25 2. 27 28 29 30 31 32 33 34
3. '39 42 45 48 51 54 57 .0 .3
"
3/4 in. 81.1 69.1
59.' 51. 9 45.'
5/8 in.' .55 -,4 47 ..2 olI0.7
19.4
40:4
'27.6
3'.1
24.' 22.1
17.2 :'15.3 13.8·
19.9
"12.4
32.4 29.2
26,S
18.1
211. 1
"16,5
22.1
15.1
72
20.3
13.8
18.7 17.3, ".0 14.9 13.9 13.0 12.2 11.4 10.7 10.1
12.8 11.8 10.9 10.2 9.48 8.86 8.30 7.79 7.32 '.90
102
25.4 22:1
35.4
••
96 99
34" 5
29.4
31.2
.9 ' 75 78 81 84 87 90 93
1/2 in.
11,3 .
10.3 9.40 8.63
7.9S :7.3S ".82 '.34
3/8 in.
1/4 in.
---
10.8 9.18
14.5 12.7 ,11.1 9.8' '
7.92 '.90
5.57
6.06
4.42 3.98 3.61
8.79 1.89 1.12 6.46 5.88 5.4'1
5.0' 4.71"
4 .• 0
s.n
5." 5.2. 4.98 4.72 4.48
3.11 2.93 2.78 2,'4
6-132
5/1. in.
"19.8 16.9
'4.13 3.88 3.65
,
(ps~~ for
3.44
.,
5.43
4.92
7.53 6.41 4.94
4.48 4.10 3.77 3.48 3.08 '.2.84
2.64 2.45
2." 2.49, '2.34 2.18 2.04 '1.91 .1.83 1.7. 1.70 1.63
1.92 '1.82 1.74 1." 1.58 1.51 1.47 1.43 1.41 1.39
3.15 2.86 2.25 2,08
TK 5-l300/NAVFAC P-397/AFR 88-22
.' "
r-;
.,
.,
..
.,'
'/
,
"
,
,
t ,
"
"
',I .":"
Tabl~:6-l2
R~sistance.
Effective Elastic·Static .~
.
Aspect ~a~i~ - ~.OO
,. ,
reff (Continued)
,.
','
Glass Dimensions r , '( if!.. ) ."
b
, • ,. .~
12 13
48; 52 56 .60 64
14
15 16 . 17 18
., ;68
72 . 76'
19
20 21 22 23 24 25 26 27 28 29 30
80 84 BB 92 96 100 104 108 112 116 120
'I.
1
"
3/4
---.
---,' 75 , 7
51.7
64",5
44 .0
55,6 48,5
38,0 33,1
42.6 37.7 33.1
25,8, 23.0
30',2
27,3 . .2"..1
20,6 18,9 17,5 16,1 15,0 13,9 13,0 12,1
14 .3 <::~
14.1,
12.9
11,9 11,0 10,2 9,49 8,85 8,27
"
10.5 9.59
8,77. -8.05 7,42 6,86 6,36 5,92 5,52 5,15
j
.1.
7.02
8;57
5.99
7.. 39 6.43
5.16 4:52
J..
I,
<':
4,63
4,28 3,97' ; 3.70.•.
13,.'
;.
3,22 3,02
".: ~
5.66
3.99'
5.01
3.56
4.49 4.05 3.67 .
3.19 2.87
3.34
"2.54 2.30
'3.06
2.10
2.81 2.52 j 2,32' , 2,15 ,2,00. 1:88 1. 76 1,66
1. 94 1. 79
.,
•
"
.c ~ ,
., '
6-133"
~,.~
.
.
1. 67 1,56 1.46
1,38 1,30 1.23
I
..
~
,,'J'f
, ..
.
••. e•
"
e'· ..r
10.1
15.7
7.36 6.65 6.03 5.49 5.03
12.9 Ill. 6'
in.
1/4'" in:
18.5
10;"4 .9.20 8.20,
16.1
'5/16
'in.
13.6 11. 8
~~J. '~"
'20.6
15.4
"
27~5
23.7
18.'6: 16",9
-,
.; ;22-,5
3/B
in;
32,2 "
29,1 I!": .:
.i /2 l
5/if in.
in,
:! ': • ~I
~J.
Effective Eiasti~ Static Resistance " (pst> for Glasa Thickness, t, of ,--
. t , '. ' ~ 1~~~
e
TM 5-l300/NAVFAC
P-397/AFR.88·~2
UNDERGROUND STRUCTURES 6-33. Introduction Underground structures are not usually used for production and handling of explosives since access for both personnel and explosives is more difficult than for an aboveground structure. However, an explosion may result in severe hazards from which an aboveground structure can not provide adequate protection and a buried structure will be required. An example might be a manned control building at a test site which must be located very close to a highhazard operation involving a relatively large quantity of explosives. There is limited test data available to predict the pressures acting on an underground structure. What test data that is available was developed for· use in the design of protective structures to resist the effects of an attack with conventional weapons. The results of this data and the design procedures developed from it are given in the technical'manual, "Fundamentals of Protective Design for Conventional Weapons," TM 5·855·1. The data presented may be expanded to include the design of structures subjected to accidental explosions. The pertinent sections are briefly .summarized . . ':-..... .below. ,
.-;
A typical underground structure used to resist conventional weapons attack is' shown in Figure 6;.)3. The burste, slab pie,"~nts' a weapon 'from penetrating through the soil and detonating adjacent to the structure. A bur's cer slab' is not mandatory, but if it is not used the structure will have ,to be, buried much deeper. The burster slab must extend far'enough beyond the edge of the building to form at least a 45 degree angle with the bottom edge of the ' building. It may have to be extended further, though, if it is possible' for a bomb to penetrate at a very shallow angle, travel beneath th 7 burster slab' and detonate adjacent to the "structure .(see F;gure 6-53). Sand is used as backfill because materials with high volume.of air,fi1led voids and low relative densities are poo~ transmitters of ground shock. In addition, sand resists penetration better than soil. 6-34, Design Loads for Underground Structures 6- 34.1. General The pressure-time relationships for roof panels and exterior walls are determined separately. For the roof panel, an overhead burst produces the most critical loading while for an exterior wall a sideburst is critical. A general description of the procedure for determining the peak pressures and their durations is given below. For a more detailed description,. including the required equations, see TM 5-855-1. The magnitUde of the ground shock is affected by: 1.
The size of the explosive charge and its distance from the structure;
2.
The mechanical properties of the soil, rock, and/or concrete between the detonation point and the structure; and
3.
The depth of penetration at the time of detonation. 6-134
'I
,e
TH'5-i300/NAVFAC P-397/AFR 88-22 The ~tres~es and ground m~tjo~s'are gr~atly:enh~nc~d'as'the deptfi,of the explosion increases. To account, .for• \ this effect • -a coupling factor is used: <" ." • ., • . -' • . -,." . The,coupl~ng ,factor i~,defined as the ratio of ~he,grou~d shock magnitude from a near' surface' burs tl to that of a,fu'il y buried burst. A single coupling factor applies to"all ground shock p;'ramete~s and depends on the dep th of the ~
expio'sion' and whether de"tonation occurs in 'soil, Jconcrete or air.' , " , ',' 6-34.2: Roof Loads
.
A typical roof load (shown in Figure 6-54) consists of a free-field pressure Po and a reflected pressure Pro The r~flected pressure oc~~rs 'whed the free-field pressure impinges on the roof panel and' is instantly increased to a higher pressure. The amount of increase is a func~iori of the pressure 'in the free-field wave and the angle formed between the rigid surface and the plane of the pres~ure' fro~t. However, TM 5-855-1 "suggests that an average reflection' factor' of L 5 is reasonable. " : " ' The. p re ssure s on the roof 0'£ 'ari'u~derground~structure~ are"bot uniform across the 'pane L, 'especially if the' i1e'pth: of' the explosion' is sh;;llow ... However , in order' to use a sirigle-degfee'~of-freedom 'analysis", a uniform Lo ad is required an,:i"h'ence an average uriiform pressure 'must be de t e rmi.ne d . TM '5- 855 -1 presents'
figures that give,the ratio peak pressure; at the center of a roof panel to the average pres sure across chevent t re panel. This ratio' is a: function of the support conditions and aspect ratio of the panel and 'the height of the burst above tl)e roof. ., For 'the, inos,t severe roof load the expl.os Lve is po's Lt.i.oned directly over the center of the panel. The 'averag~ free:fleld and~average reflected pressures' are c;'ibulatedas described above.' The &iration of the pressure pulse als'o ' varie~ across th~ roof;panel ahd a' fictitfous averag~ duration to must be~ . determined: TM 5~855ct recommends calculating the "duration of the peak free-field pressure pul.se at 'a 'point located one quarter of the way' along the short span and at the center of the long span, This duration is then used as tne average duration of the entire panel. The peak free-field pressure and impuls~are calculated using equations given in TM 5:855-1. The duration is found by' as'sumfng a triangular pressur'e"time r e I'a t Lorish i.p ," The duration of the 'average reflected pre ssure t r is 'given ih TM '5-855-1 as a function of, eirher:the thickness of the structural element or the distance to the nearest free edge '0'[ the structure. Tlie smaller cif"tlie two' numbers should'be used in anal~sis. ! . ' ;'i .~.
• i
~
r
6-34':3, Wall: Loads
"
The design loads on an exterior wall are determined using the procedures described in Section 6-34.2 for roof loads. However, in addition to the pressure wave traveling directly from "the explosion, the wall may be subject to ""pressure wave' reflected' off the 'ground surface or burster, slab and/or"a pressure' wave,·r'eflected'offl 1,a lower rock-Layer 'or 'water table. •
,
I
.: • ~
_
...,.~
The parameters of'each wave (average reflected pressure, average free-field pres sur e., durations and -t Lme .of arrival~)I·are determined separately using pr ocedure s very similar to those de'sc r-LbediLn Section' 6-34.2.. The total
pres sur-e- t Ime history is 'equal, to the superposition of the three' waves as shown in Figure '6-55.'-" The' 'superposition 'results in a very complicated load shape. The response charts ,of Chapter 3 are 'not, applicable for such a shape, 6-135
~
TM 5-l300/NAVFAC P-397/AFR 88-22
therefore the, load must be idealized. The, actual load is transformed into a triangular load having the same total impulse (area under the' actual load curve equal to the sum of the areas of the, direct 'and reflected waves)., The maximum pressure of the idealil,iid Load is equal to the ~axiinum pres~ure,' of the' actual load neglecting the short reflected peaks. The duration is then established as a func t Loniof the total impulse and maximum pressure (Figure, 6 7 55). For an exact solution, the actual load curve is used ,in a single' degree-of-freedom computer program analysis or numerical analysis as, given in Section 3-19.2. 6-35. 'Structural D~sign 6-35.'1. Wall and Roof Slabs J •
•. " ~.
'I,
•
The struct~ral design' of underground sEructures is very similar to the design of aboveground structures as described in Chapter 4. , The effect of the' soil is to modify the response of the structural compon~nts. The dead load of the soil,reduces the ,resistance, available to resist blast. At the same time a portion of the soil acts with the structural elements to increase the natural period of ,vibration. ,'in the case of a wall, i t is' ass~ed ,that ,the ma~s !'If, .~' two" (2) "feet of, soil acts with the mass of the wall. Whereas for a roof, t.ha . . entire'mass of the soil' supported the' roof", or a depth of soil 'eq~ai,to, quarter of the roof span (shoFt span' for a two-,way'panel), whicheve.r i~, ,', smaller, . is. added to the mass,of the roof. ' ' .
;;pe
by
...
,
'
The dynamic response of underground structures must obviously be limited to comparatively small deformations to prevent collapse of the structure due to earth loads. A protective ~tructure,subje~ted to ~onventional weapons att~ck should be designed for a duc tLl.Lt.y ratio of 5.0, :a/ recommended by TM 5~-8,55:i. This ratio may be increased, to 10 if special provisions are taken. A maximum deflection ~orresp~':':ciing to' '~" s;'pport rotatio~ of one (1)' degree or 'a ductility, ratio of 10.0 is permitted 'f~r 'underground structures subjected,toaccidental explosions. , , Spalling is the ejection of material from the back face of, a slab or beam. It results from high-intensity, close-in explosions. ,.Fr~gment shields or backing plates, as shown in Figure 6-56, are of some value in protecting ,personnel and equipment. These steel plates must. be, securely' anchored to the inside face of the concrete member. Te s t s have sh~~ th":t the shock of a deep penetr at Lng ,:', detonation to be enough to cause inadequate welds to fail over a large area, adding the whole steel plate to the concrete spall. A strongly attached plate adds about 10 percent to the perforation resistance of a concrete.~lab .. For a further discussion of backing plates, see Chapter 4, "
6-35.2. Burster Slab
For protective structures, a burster slab, prevents a weapon from, penetrating through the soil and detonating adjacent .to the structure., Its, thickness and length may have to be determined by a trial and error procedure in order'to limit pressures on the structure to a given -vaLue .
However-, 'the minimum
dimensions ,are shown in Figure 6,-53. The minimum ,reinforcement is 0.1 percent in each face, in each direction or a total of 0.4 pe rcenr; .Ln the design-of structures subject to accidental exp10s~ons, the ground floor slab of the donor building serves a purpose similar to that of a burster slab. The floor, '.. slab helps tO,prevent fr",gment, penetration and to attenuate the load.
TH 5-1300/NAVFAC P-397/AFR 88-22
6-36. Structure Motions 6-36.1. Shock Spectra TM 5-855-1 gives equations for acceleration, velocity and displacements for underground structures. These simplified methods take into account the attenuation' of the pressure wave 'as. it transverses the structure.
For a
sideburst, the vertical acceieration, velocity and displacements are 20 percent of the horiz~ntal valu~s. The horizontal motions are'~niform over the entire floor while vertical motion~'at the leading edge are twice those at ~;'
m.i.dsp an .
,
,
Once the peak in-structure acceleration, velocity and displacements have been determined an in-structure shock spectra can be d~veloped using the principles of Chapter 2 of this manual. 6-36,2. Shock Isolation Systems' Chapter 1 presents the upper limits of the shock environment that personnel and equipment can tolerate. If. the:shock environment exceeds. human tolerances and/or equipment "fragility le;'els'), thEm "shock, isolation -sys tems are required to..'p'fotect personnel and sensitive equipment. Using cthe shock'spectra developed as described above. shock ~s'olation systems' are designed as outlined
in',_Sections 6-43 through 6-49.,
,"
-, '
'I ~
-, -, , ,
,
6-137
•
.-, -.
,
~,,~
"t.>
.,.'
<0
'
BURSTER SLAB
: :!~
~ ~
Of
I
...
. .
v,
. .it
'
.
"
.
" " . ' : to.
. '. ':,,::/::
:'
.
.". :" .', :.~',
...
- ".
.
.
,"
"
...
DIREC.T WAVE FOR :.:::: '.i.:<:-. ROOF LOADING " .. :
.: . AJ /7 '.::.'.:'::"'//
, :i -
~ '
".,. ::E
. 1-i "10
'10
.'.. :-.: .:<>// . .... 'X// .'
'.
. ~ :::",
.
'"
.'
-
•
,Z
,~
,
SAND FILL
'",
.... w
....
CD
~-.
'.~ .. (MIN.1
WATER TABLE OR. ROCK LAYER
.,
"" ,..:
~
., c-
" .' :.:
.. .I.
• Figure 6-53
.
or
..-
Geometr-y.of a buried structure "
e
e
'.
t,o TIME (ms), -
.....
".p. .
Figure 6-54
Typical roof panel load
6-139
, 'Tn'·'-):
'·:.Jfi_ -1;"tT~:t:o·t·~··"··V'T,·r-.'.""r~·~:"·1TI"
IT'"
1'··•.1 r'·I"".·':''':'~}'''' ,_., 'Tt,M·.....' "
I-Ii~.
" , .•_~"
;'/.'
..',' ' .:».~ '>
."
TOTAL PRESSURE - TIME LOAD ON WALL.
_. EQUIVALENT TRIANGULAR ; . LOAD
(Pra_v.<1~d DIRECT WAVE
PRESSURE WAVE REFLECTED FROM THE SURFACE.
~
~
.,
0-
0
.
'", ..... .l:0
-w
0-
0
I
(lbavg1d
."-
a: en en w a: a..
PRESSURE WAVE REFLECTED FROM LAYER .
(Pravg)s
.:J
'-'
(f'oavgls (IPoavgl,
~
Ird
~
Irs
~
..!.U IrM
~
t~e .1
TIME (ms)
Figure 6-55 Contribution of three pressure waves on a wall
e
e
e
[
I .. . • "
"
,
,~
J f!
. ,. ;~
"
.• , '
"
c
'
,
..
'G'
.. -
•
1'1,
.'
..
-, '
... ...
e ,
'
. ..• .,
•
~
, 4:..'
.
SCAB PLATE
'.
'G
'
;J
o·
SHEAR STUDS
..
• •
•
..,
"
'.'
.. RETENTION OF SCAB MATERIAL
"
Figure 6-56
6-141
Spall plate
.-
'f.·
.
'
TH 5-l300/NAVFAC P-397/AFR 88-22
EARTH-COVERED ARCH-TYPE MAGAZINES 6-37. General Certain types of earth-covered concrete-arch and steel-arch magazines have been approved and standardized for use by the Department of Defense Explosives Safety Board. These magazines provide definite advantages over other types of magazines.
1.
Among these advantages are':
Less real estate'per magazine 1s
req~ired·because
of the decreased
intermagazine separations permitted when approved magazines are used.
2. An almost infinite number of storage situations exists because magazines can be designed to any length. 3. Because of the reduced separation distances, less roads, fences, utilities, etc. I are required.
Unlike the other structures discussed elsewhere in this manual, an earthcovered magazine is not designed to resist the damaging effects of its exploding contents. It is accepted that the magazine will be demolished if an internal explosion occurs. During such an incident, the inside of a largespan arch might experience initial blast pressures considerably in excess of 10,000 psi. Less'than 100 psi could lift the arch completely out of the ground; therefore, the major portion of the protection is provided by the receiver magazines rather than the donor magazine. Earth-covered magazines are utilized primarily to prevent propagation of explosion. These structures may also be used for operating buildings and can provide personnel protection.
In such cases;" separation distances greater
than those required to prevent propagation ,of explosions will'be necessary. In' addition, a specfal evaluation of the' structure is requir~d. This evaluation must include the leakage of blast pressures into the protected area, the strength and attachment of easily damaged or lightly supported accessories which may become hazardous debris, the transmission of shock to personnel through the walls or floors, and overall movement of the magazines. 6-38. Description of Earth-Covered Arch-Type Magazines
A typical earth-covered arch-type magazine used for storing explosives has the following features: 1. A semicircular or oval arch constructed of reinforced concrete or corrugated steel used to form roof and sides. 2.
A reinforced concrete floor slab, sloped for drainage.
3.
A reinforced concrete rear wall.
4. A reinforced concrete headwall that extends at least 2-1/2 feet above the crown of the arch. 5. Reinforced concrete wingwalls on either side of the headwall. wingwalls may slope to the ground or may adjoin wingwalls from adjacent 6-142
The
5-l300/~AVFAC
TM
P-397/AFR 88-22
magazines. The wingwalls m/!-y be·either. monolithic or separated by expansion joints from the headwall. 6.. Heavy steel doors in the headwall (either manually·operated and/or motorized)" :
.-
7.
An optional gravity ventilation system.
8.
Earth cover over the top, sides and rear of the structure.
This
. cover must be at least 2 feet thick at the crown of the arch. The earth above. the structure (within the spring line of each arch and between the head and rear,walls) is sloped for drainage while .beyond the outline of the structure the· earth is sloped.2 horizontal to 1 vertical. 9.
.,
Its own built-in lightning protection and grounding systems . .,
A·typical earth-covered.steel arch magazine is illustrated in Figure 6-51 . . I.
6-39 .. Separation Distances·.of Standard Magazines" . -I
"',
Numerous full scale tests of standard magazines.have been performed over a period of several years. As a result magazine separation formulae have been established, which
will~·prevent
magazine-to-magazine propagation of exp Lo-
sions . . All possible .right angle arrangements have been'considered, i.e., side- to-side, rear· to ":' rear ,·tfront- to-rear:,! etc.
.The standard, magazines, which.
are at least equivalent in strength to those. tested, are listed in the 000 Standard,· "000 Ammunition and Explosives Safety Standards, 6055.:9-STD." The required separationdis~ances,.as a function of~the quantity of explosives stored in the structure are also given in' the DoD Standard. A possible magazine arrangement is shown in Figure 6-58. 6-40. Design
",
I
.•
Protection of magazines adjacent to a donor magazine is accomplisheq by combining the·following. fact~rs: 1. The intensity of the pressure front moving from the donor magazine to receiver magazines diminishes rapidly as the distance traveled increases.
..
"' .2. The earth cove rtove r-.andraround the'donor magazine provides
some confinement ~nd·tends.to'·directionalize·the explosive force both upward and outward. from the:door .end:of the magazine. 3. The earth around and over r ece tve r-magaaines resists fragment penetrations and provides mass to the arch to resist the bl'ast pressures.
4. The arch of receiver magazines is capable of resisting blast loads considerably in excess of the dead loads normally imposed on it . . Design of presently used magazines is essentially conventional except for two features, which are doors and arch. The doors are designed to withstand the dynamic forces from an explosion in a nearby magazine if the siting is in accordance with Figure 6-58. However, they provide almost no resistance to the effects of an· explosion within the magazine. Also, the capacity of the
TM 5-l300/NAVFAC P-397/AFR'SS-22
doors to resist elastic rebound and,negative phase pressures may be less than their capacity to resist positive phase pressures. Therefore, where personnel are concerned, all doors should be analyzed 'to determine their ultimate capacity to resist all the loadings involved. The arches used for the standard earth-covered magazines are the same as those used in the test structures to establish the required separation distances. These arches have not been dynamically designed, for the blast loads and' may be in excess of that required. 6-41. Construction
"
"
" ,
Effectiveness of earth'covered magazines is largely determined by the quality of construction. A few of the construction details that, could be sources for problems in this type of structure are discussed below. Moisture proofing of any earth-covered structure is usually difficult. This difficulty is increased with'a steel-arch structure because of the many lineal feet of joints available for introducing moisture. For example, a large 26by SO-foot magazine contains approximately 1,050 ,feet of edges, Asealant tape must be used that will not deteriorate or excessively deform under 'any an~icipated environmental or structural condition: ".", -
.:
-,
Earthfill material s'houLd be clean', cohesive, and free from large stones', A minimum earth cover of two feet must be maintained. Surface preparation of· the fill is usually required to prevent erosion'of the,2-foot'cover; ,. Restricting granular size of material'reduces,throw9ut'.of.~ragmentsin case of' an accidental explosion and creates a'more uniform energy absorbent over the'0' top of the magazine, ,
,"
C
•
Lightning protection is rather easily obtained in a steel-arch structure, sections of steel-arch plate must be interconnected so that they become
All
electrically continuous. In a concrete·arch magazine, the reinforcing steel must be interconnected. In effect, a ."cage" .Ls created about the magazine
contents, Probably the most critical point for lightning protection is the optional ventilator stack which projects above the.surrounding earth cover, .,; l
6-42. Non-Standard Magazines
,',
•
~l'
.,
Non-standard earth-covered magazines, that is magazines not listed in 000 Standard 6055.9-STO may also be used. for 'explosive storage. However, if a "nonvs t.anda rd" earth-covered magaz-ine o'r an aboveground magazine is used the'-' separation distances must be increased.' The OoO:Standard 6055,9-STO includes the increased separation distances, as well. as other criteria, for non standard" earth-covered and aboveground magazines. II
,
'
"
6-I44
,
12" •
·
,
,,,"
VARIABLE
·~k~F.
r:
CONCRETE W-.c; 'MAlL
'~
'-
RE~i
--
CONCRETE REAA WAi,.L
\.
n
-;0.
,
STEEL .00<-,
.. -
.'"
~
J
m
c
~
l•
. .. ....REINF. CONCRETE
..
"
HEAOWAU..
,
·
:±
®
fOElNF' """CRETE · WING WAlL
. PLAN AIR TERMINAL
VENTI LATOR
.,
2'~O" MIN. EARTH FILL
SLOPE EARTH FILL
STEEL~
,
,
. CoNcRETE
:~
SLAB
SLOPE ..• -'r-
SECTION
0
EAA;TH FILL
COftRUGATEO STfEL AR04
_ :_.: . •
Figure 6-57
f"~,
c· •.•
Typical earth-covered steel-arch magazine
,
-
-
-
.
1----. 1
1----"
rr---
II
II I- ___ ...JJ__
II U
..._-,_..11·
S
"
~
_
.
on
~
"
0
,
1----11-
1----.1' .
I-----U
f---- ...JJ
II il
II
----n
----11 . I'
..
II LL____
II
I D'2.0 Wl/· .
I
•
_
·rr----
----I
II ___ ...JJ
I.
rr---
I .
D. 1.25 WI!,
HEADWALL
-----
REAR WALL EDGE OF EARTH COVER
-----
EDGE OF ARCH .
=
CHARGE WEIGHT
W
Figure 6-58
Ninimum separation distances 'of standard magazines 6-146
TM 5-l300/NAVFAC P-397/AFR
~8-22
BLAST VALVES 6-43. General
.~,
.
6-43.1. Applications A prime concerp of blast resistant structures is to ,restrict the flow of high pressures into or out of a s cruc ture. For donor structures pres sur ea.Ye Leased may have t~ be restricted in order to limit pressures acting ,on adjacent structures to .coLerabl e levels. ,Also, pressures leaking into acceptor structures m;'st be ,limited' to' p~event pressure buildup beyond acceptable, leveis.' In' either case openings may have to be completely sealed to prevent the diffusion of' ccncemtnent;s . .' The simplest, most economical way of limiting leakage pressure into or out of a protective structure is to restrict the number·and size of air intake and exhaust openings. In a donor structure the leakage pressures may be further reduced by venting them through a stack. The stack increases the distance from the point of release of the pressure to ~he accepto~ struc~ure thereby attenuating the blast loading. Methods for predicting the pressures leaking out of' a building and 'the pressure buildup within a building are discussed in Chapter 2. If the leakage pressure can not be reduced to acceptable levels or if contaminants are released during an explosion, thE openings must be sealed with either blast valves or other protective closure device. , Blast valves may be either remote-actuated (closed mechanically by remote sensors) or, blast-actuated (c Los ed by the pressure wave itself).. Both types can be non-latching or latching. A non-latching valve will open under negative pr~ssures." A latching valve is one that can only be reopened manually. In addition, a blast-actuated valve can be double-acting. , A double-acting valve will seal against the positive blast pressure, move in the opposite direction to seal against negative pressures and then reopen when pressures return to normal.
6-43.2. Remote-Actuated valves Remote-actuated val~es are dependent on extern~l sensing devices which trigger the closure mechanism and close the, valve before arrival of the blast wave. Actuating devices have been developed that are sensitive to the blast pressure of an explosion and react electrically to trigger protective closure systems. Other actuating devices sensitive to flash and thermal radiation are also available .. The pressure sensing device is placed on a circumference at a predetermined radius from the valve (closer to ground zero) in order to compensate for time delays of electrical and mecha~ical control equipment and to permit valve closure before the blast arrives. Thermal sensors are designed to det~ct the characteristic pulse emitted by an explosion to prevent actuation by other sources such as lightning, fires, etc., wh1chmay occur with flash sensors. Remote-actuated valves present problems of protection against multibursts and button-up time for combustion-type equipment installed within the structure. In addition to problems of hardenability of the exposed sensor and suitability for multiburst operation, i t is often necessary for sensors to initiate reopening of the valves as soon as dangerous pressures have subsided. In general, remote-actuated valves-can not ,close fast enough to be effective during an H. E. explosion. 6-147
TK S-1300/NAVFAC P-397/AFR 88-22
6-43.3. Blast-Actuated Valves Self-acting blast-actuated valves, which close under the action of the blast' pressure, overcome some of the disadvantages described above and present other factors to be considered. Since the valves are closed by pressure, they are not dependent upon sensing devices for operation. They can be automatically reopened after passage of the positive phase or latched closed during the negative phase if this is required. Double-acting valves automatically seal the opening during the negative phase.-
Since blast-actuated valves'are closed
by the blast,' there is an inherent leakage problem to be considered due to the finite closing' time. Although this is in the order of milliseconds for most valves, sufficient'flow to cause damage may pass the port openings for certain valve designs and pressure levels. Effectiveness of closing at both' high- and low-pressure ranges must be checked. The amount of blast entering depends on the closing time of the valve which, in turn, depends on the mass of the moving parts, disk'diameter, and the distribution of pressures on both faces of the disk. Ideally, a blast~'actuated valve should possess the folloWing characteristics: instantaneous closure or no leakage beyond the valve during and after closure, no rebound of moving parts, equal efficiency at all pressures below the design pressure, operational and structural reliability, minimum of moving parts, low-pressure drop through the valve at normal ventilation or combustion air flows, multiple detonation capability, durability, and easy maintenance. Although instantaneous closure is not physically possible, the closing time can be reduced sufficiently to reduce the leakage to insignificant values. This may be accomplished by increasing the activating pressure-force-to-- ' moving-mass ratio, decreasing the length of travel, permitting no deceleration during closure; and other methods. 6-43.4. Plenums Blast valves, especially blast-actuated valves, will allow some pressure leakage. While the leakage pressure may not significantly increase the ambient pressure at some distance from the valve, there might be a "jetting effect" causing very high pressures in the immediate vicinity. A plenum may be used to protect against these high pressures. Two examples of plenUms are, a plenum chamber and a plenum constructed of hardened duct work. A plenum chamber is a room where pressures attenuate by expansion. A hardened duct work plenum reduces the'pressures by increasing the distance traveled (similar to the stack discussed in Section 6-43.1). For a donor structure, a plenum would only be necessary if contaminants are released during an explosion and the air must be filtered beforebeing,vented to the exterior. In that case, a plenum would be used to lower the pressures and, prevent damage to the filters. A plenum in an 'acceptor structure would be used to prevent high leakage pressures from directly entering the building's air duct system and possibly causing local failure of the system. Plenum'chambers should be designed to avoid a buildUp of interior pressure which would impede closIng of the valve. The ratio of the area of the chamber cross section to that of the valve outlet should be preferably greater than 4:1 so as to'diffuse leakage flow as quickly as possibYe.' A chamber which has the prescribed necessary volume but has little change in area would act like a 6-148
TK 5-l300/NAVFAC P-397/AFR 88-22 tunnel wherein entering pressures would encounter little, attenuation in the length provided. 6-43.5. Fragment Protection To ensure ,that blast valves function properly', they must be protected from fragments ,that may pe~forate.the valves or j~ them in an open position .. One of several methods of accomplishing the required protection is by' offsetting the opening from the blast valve by means of a blast-resistant duct or tunnel which would prevent the propagation of fragments to the blast valve. Another method is to enclose ~he blast valve in a concrete chimney. Other methods include using a debris pit or steel shields or debris cover attached directly to the blast valves. 6-44. TyPes of Blast Valves
.
6-44.1. General
~.
Various types of blast valves have been developed and many of them are available commercially from suppliers both here and overseas (see Table 6-13). For present designs, air flow rates from about 300 cfm to 35.060 cfm can be obtained. Some valve designs are available in more than one size and can be either blast or remote-actuated. The pressure loss across the valve at the rated air flow is, in most cases, less than one inch of water. The maximum incident pressure capability of available valves is above 50 psi and generally at least 100 psi. For ~helter purposes, these valves may be overdesigned since the protection level for many shelters will be less than 50 psi. Except for cost factors, using a 100-psi valve for a 10-psi shelter design should not necessarily present technical problems since a valve must operate at all pressures below the maximum design level. The best type of actuation (blast or remote) depends partly on the.design pressure as previously discussed with regard to reaction time and operational considerations. For long arrival times (low pressure). a remote-actuated device can close the valve before the blast arrival whereas leakage may occur for a blast-actuated valve. At a high pressure (short arrival time) ,the closing time for the remote-actuated valve may be. longer than the arrival time. 6-44.2. Blast Shield ~.
If it can be assumed that the ventilation system can be closed off during a hazardous operation and kept closed until there is no danger of further blast" a relatively simple structural closure (blast· shield) , such as a steel plate' can be utilized. ,This type of closure is especially useful for,an exterior opening which would only be opened periodically, such as maintenance facilities where the release of toxic fumes from within the structure is required. 6-44.3. Sand Filter, .1
In shelters where normal operational air requirements are small, sand filters are useful in the ·attenuation of leakage pressures. With this type of filter the pressures continue to increase throughout the positive phase. Thus this 6-149
TM 5-1300/NAVFAC P-397/AFR 88-22 filter is good only for loads with a relatively short duration. 300-cfm sand filter is shown in Figure 6-59.
A
~ketch
of a
6-44.4. Blast Resistant Louvers The blast resistant louver shown in Figure 6-60 is blast actuated and has a rated flow of 600 cfm at less than'l inch water (gage) pressure drop. If a larger volume of air is required the louvers can be set into a frame and used
in series (see Figure 6-61). Louvers can be used in acceptor structures' subject up to 50 psi. A major drawback of 'the louvers is that there'may be as much as 40:percent leakage across the valve, especially at lower pressures. 6-44.5. Poppet Valves 6-44.5.1. Applications A poppet valve has many advantages. It has f~w moving. parts which might need repair. It can be blast or remote-actuated, latching, non-latching, or double-acting and is available in sizes from 600 to 5.000 cfm. A blastactuated poppet valve has a very fast closing time. approximately 20 milliseconds, making it the only valve that reacts fast enough to be used in a containment c e l l . ' . A typical blast-actuated poppet valve is shown in Figure 6-62:' The valve' consists of an actuating plate, a valve seat, a backing plate that supports the actuating plate. and a spring which holds the valve open during normal operations. The normal air flow is around the actuating plate. A blast load will compress'the spring and move the actuating plate against the valve seat thereby sealing the opening. As the blast pressure moves the actuating plate, some pressure will flow around the plate while it is closing. The leakage pressures can be completely prevented by using a valve similar to the one schematically illustrated in Figure 6-63. In this valve the normal air flow is around the actuating plate through a length of duct. When the valve is subjected to a blast load, the pressure starts moving the actuating plate while at the same time flowing through the duct. The length of the duct must be long enough to ensure that the· time it takes the blast pressures to flow through the duct (delay path) will be longer than the time required for the actuating plate to seal the valve. As an alternate to the long'duct. an expansion chamber may be used to delay the blast. 6-44.5.2
Recommended Specification for Poppet Valves
Presented below is an example specification for the design, testing and construction of a poppet valve,'but it may be adapted for other types of blast valves. This example specification is presented using the Construction Specification Institute (CSI) format'and shall contain as a minimum the following: 1. APPLICABLE PUBLICATIONS: Except as otherwise stated herein all materials and work furnished in accordance with this specification shall comply with the following codes and standards.' 1.1
Federal Specification (Fed. Spec.)
6-150
TM 5-l300/NAVFAC P-39!/AFR 88-22
Paint, Alkyd Resin; Exterior Trim, Deep Colors
TT-P-37D
• TT-P-645A
Prime~, Paint Zinc Chromate , Alkyd Type
'. .l
~aint,
TT-P-86G
Red- Lead-Base,
R~ady
Mixed
, American Society for Testing and Materials (ASTM)
1.2 .'
,-
',~
'. .
1.3
Pipe, Steel Black, and HotDipped, Zinc-Coated, Welded • and Seamless
A 53-8la
.
D 2000-80
Rubber Products in Automotive Applications, Classification System For
"
Magnetic Particle Examination
E-709-80 ,
• American Institute of Steel Construction (AISC) Specification for the Design, Fabrication and Erection of ; Structural Steel for Build~ngs (Eighth.Edition) with commentary
.- .
~.4
American Iron and Steel 304 l7-7th
1.5
Americap Welding Society (AWS) D.l.l Structural Welding Code(latest edition)
I~stitute
(AISI)
2. SUBMITTALS: The f.oll~wing information shall be submitted for approval. Materials shall not be delivered to the site until approved shop drawings have been returned to the Contr~ctor. Partial submittals, or submittals for less than the whole of any system made up of interdependent components .will_not be accepted. -Submittals for manufactured items shall be manufacturer's descriptive literature, shop drawings, and catalog cuts that include the manufacturer's dimensions, capacity, specification reference, and all other info~ation necessary to establish contract, compliance . . ':l: i 2.,1 Qualifications: The. Contractor shall submit for approval, data to suppor,t the quali.fi"ations ,of .the manufacturer, and installer. A list of previously,successfully completed jobs of a similar nature, indicating the name and addr~ss'of the owner of the installation shall be included with the information . Manufacturer's Data: Before.executing any fabrication work, .2.2 a completely marked and coordinated package, of documents sufficient to assure full compliance with the drawings and .specifications shall be submitted. The submittal shall include a complete technical evaluation of the capacity of each valve as described below. The drawings shall include: detailed fabrication and equipment drawings; . assembly showing the complete installation, including methods for supporting the valves, subframes and frames; a listing 6-151
TH 5-1300/NAVFAC P-397/AFR 88-22
.'
of all materials and material specifications; surface finishes; fabrication, assembly and installation tolerances; locking devices and locking device release mechanisms; and a detailed sequence for installation of valves and frames in conjunction with other phases of construction. Structural fabrication drawings shall conform to the requirements of AISC and welding shall conform to AWS. Assembly and installation shall be based on field established conditions and shall be fully coordinated'with architectural, structural, and mechanical systems. All aspects of any work developed in connection with the development of these valves shall be fully documented and become the property of the Government. 1• 2.2.1 Standard Compliance: Where equipment or materials are specified to conform to requirements of the standards of organizations such as ANSI, NFPA, UL, etc., which use a label or listing as a method of indicating compliance, proof of such conformance shall be submitted for approval. The label or ~isting of the specified organization will be acceptable evidence. In lieu'of the label or listing, the Contractor shall submit a notarized certificate from a nationally recognized testing organization adequately equipped and competent to perform such services, and approved by ,the Contracting Officer stating that the ~tems have been tested with the specified organization's methods and that the item conforms to the specified organization's standards.
2.3 Preliminary Hydraulic Characteristics: Prior to Construction, submit with shop drawings an estimate of the hydraulic characteristics of each valve, under actual operating conditions. -Ra t Lngs shall be based on tests or test data. All necessary corrections and adjustments shall be clearly identified. Corrections shall be established for actual altitude and air flow directions as shown on the drawings as well as hydraulic effects produced by mounting and/or connection of the valve. 2.4 Tests and Test Reports: Except as noted otherwise, the testing requirements for materials stated herein or incorporated in referenced documents, will be waived, provided certified cop i.es ' of reports of tests from approved :laboratories performed on previously manufactured 'materials are submitted and approved. ' Test reports shall be accompanied by notarized certificates from the manufacturer certifying that the previously tested material is of ..the same type, quality and'manufacture as that furnished for this project. 2.5 Blast pressure analysis calculations and/or results of approved tests shall be submitted, for both the blast valve and subframe, for approval and shall conform to the requirement of the paragraph entitled, BLAST VALVE TESTS. The calculations 'arid/or test
TK 5-l300/NAVFAC P"397/AFR
88~22
tion that' each of these welders has passed the qualification test using procedures covered in AWS Standard Dl.l.
.
"
2,.7 . Operational and Maintenance Manual: Operation and maintenance manuals shall be furnished by the Contractor. Complete manuals shall be furnished prior to the time of installation. The manual shall have a table of contents and· shall be assembled·to conform to the table of ·contents with tab sheets placed before instructions covering the subj ect .. .',
.
2.8 Shop Test Reports: The Contractor shall furnish copies of. shOF inspection and test results. of fabrication welding. 3.
MATERIALS:
Structural. steel' pipe used for> blast valv~ construction 3.1 shall conform to ASTM A 53 seamless pipe. 3.2 All structural steel plate components of the valve shall consist of stainless steel and shall conform to AISI 304. 3.3 . Spring type' components shall consist of stainless steel and conform to AISI 17-7th, Condition C. 3.4
Blast Seal. Material: Seals for blast valves; shall conform to
ASTM D 2000. 4. BLAST VALVE'REQUIREMENTS: All blast valves shall be poppet type and shall have the following characteristics.
-.
4.1 Pressure.,capacity: Each valve shall be capable of withstanding a sustained blast pressure of 100 psi as well as the impact force produced by the closing of the valve. The ,valve shall be designed to.sustain elastic deformation when subjected to the above loads. The blast valve shall be capable of closing under a force of 15 pounds. 4.2 Temperature Capacity: Each valve shall be capable of satisfactory operation over a temperature range of 35" to 300" F. 4.3 Valve Actuation: Each valve shall be actuated by the blast overpressure. The valve shall be in the _closed position 20 milliseconds after the onset of the b l ast front. The blast pressures are given. on the drawings. ;
",
..
4.4 Valve,Parameters:A minimum.of 12-inch-diameter blast valve shall be used, After-the valve is closed by the blast overpressure, it'shall remain in the c Los edjpos Lt Lon until manually opened. This shall require that the valve be equipped-with aclocking,device which shall.be' i~cated at the exterior side of the :vaive. A release mechanism for ,the locking devices shall be provided which shall be operated from a position immediately adjacent to the interior of the valve. -Any ·penetration ..through the valve or the structure must be capable of being sealed against blast leakage through the penetration. The air flow capacity of the valves shall be 1500 SCFM (1710 ACFM) for the supply and return valves and 880 SCFM (1000 ACFM) for the exhaust valve. Total actual pressure drop across the valve with air movement in either direction shall not exceed one 'inch of water (gage). The valve and its operating.parts shall be designed for a 20-year life and shall have-an. operating frequency of 10,000. 6-153
TM 5-l300/NAVFAC' P-397/AFR 88-22
4.5 Blast Seals: Blast seals shall be provided between the face of each valve and sub frame and between the sub frame and the'frame to provide a pressure tight condition. Seals shall be adjustable. and easily replaceable. The seal shall be designed,to be leakproof with a pressure differential across the seal of 100 psi. 4.5.1 Blast ,Seal Material: Seal material, shall conform to ASTM D 2000. Four sets of blast seals sha~l be furnished with each valve, . Three sets of the seals shall be packaged for long term storage. 4.5.2 Adhesive: Adhesive for.blast,sealslshall be as recommended by the manufacturer of the seals. Sufficient adhesive shall be provided for installation of the packaged seals at a later date; " 4.6 Field Removal: Blast, valves shall have the capabilities of being completely field removed and disassembled. 5.
FABRICATION:
5.1 Qualification of Manufacturer: The manufacture and installation of blast valves and frames shall be performed by the blast valve manufacturer who' shall be fully responsible for valve operation. The manufacturer shall have complete f ac i Ld t Les , equipment and technical personnel forthe.·design, fabrication,
installation and
testi~g
of complete
b~ast
valve assemblies.
5.2 ,General: The drawings indicate the location of the blast valves in the structure. The manufacturer shall carefully investigate the drawings and finished conditions affecting his work and shall design the units to meet the job condition and the dynamic loads. The blast valves shall'be .complete with gaskets, fasteners, anchors, mechanical operators, and all other equipment and accessories 'as required for complete installation. ".', 5.3
Metalwork: Except as modified herein, fabrication .ahal I be at a
"
minimum, in accordance with the AISC Specification for the Design, Fabrication and Erection of Structural Steel for Buildings. Welding of steel shall be in accordance with the requirements for AWS Specification Dl.l.. ,A welding'" sequence to reduce distortion and locked-up stresses to a minimum shall be used. All welded units shall be stress relieved. All-welded -member s shall be post weld straightened free of twist and wind. Fabricated steel' shall be well formed to shape and size, with sharp lines and angles. 'Exposed welds' shall be ground smooth. Exposed surface of work, in place, shall have a smooth finish. Where tight fits are required joints shall be milled.to,a close fit. Corner joint shall'be coped or mitered, well formed; and in true alignment. .Pe rmanent connections 'foL all assemblies and.components, except those· requiring' removal for installation and maintenance, shall be welded. Each' valve and, subframe shall be removable from the· embedded 'frame. , .
:5.3.1 Machining: Pa r ts : and assemblies shall be machine finished' wherever necessary to insure proper fit of the parts and the'satisfactory performance of the valves. 5'.3.2 Weld Details: The types of edge preparation used for welds shall be chosen by the manufacturer to be the 'most suitable for the joint and position of welding. Where required, all groove welds shall be complete penetration welds with complete joint fusion. Groove weld edge preparations: 6-154
TK 5-1300/NAVFAC P-397/AFR 88-22
shall be accurately and neatly made. All full penetration groove joints shall be back-chipped and back welded where both sides are ac ce s s fb Le . Where both sides are not accessible, backing strips 'not'exposed to view may be left 'in place unless removal is required for clearance. Backing strip not removed shall be made con 7inuous by weldi?g ends and junctions.
" of welds . 5.3.3 Weld'Tests: Inspection and . tests shall be!,s specified in AWS Specification 01.1. All welding' shall be subjected to normal , I ,;' continuous inspection. ,,'
.
'
5.3.3.1 Nondestructive dye penetratortesting shall be performed for all welding'in accordance with, Method B of ASTM E lb5 or ASTM E 709, Allowable defects shall-conform to' AWS Specification 01.1. 5.3.3.2 Penetration Welds: AIr full or partial penetration corners, tees and inaccessible butt joints shall be subjected to 100 percent ultrasonic examination, All penetration joints shall be' considered to be tension joints. All tests shall be performed by a testing laboratory approved by the Contracting Officer. The testing laboratory shall be responsible for interpretation of the testing, which shall be certified and submitted in a written report for each test. In addition to the weld examination performed by the 'Corrt r ac t or ....the Contracting Officer reserves the right to perform independent examination of any welds at any time. The cost of all Government reexamination wi~l be borne by the Goyernment. 5.3.3.3 Correction of Defective Welds: Welds containing defects exceeding the allowable 'which have been 'revealed by the above testing shall be chipped'or ground'out'f6r full depth and rewelded. ',This correction of the defected weld area and'retest shall be at the Contractor's expense . . ' 5.4 Metal Cleaning and Painting: ,
5.4.1 Cleaning: Except as modified herein, surfaces shall be cleaned to bare metal by an approved blasting process. Any surface that may be damaged by blasting shall be cleaned to bare 'metal by powered wire brushing or other mechanical means. Cleaned surfaces which become contaminated with rust, dirt, oil, grease, or other contaminants shall be washed with solvents until thoroughly clean. , 5.4.2 Pretreatment: Except as modified herein, immediately after cleaning, steel surfaces shall be given a crystalline phosphate base coating; the phosphate base coating shall be'applied only to blast cleaned, ~are metal surfaces.
I
",
5.4.3 Priming: Treated surfaces'shall be primed as soon as practicable after the pretreatment coating has dried. Except as modified herein, the primer shall be a coat of zinc chromate primer conforming to Fed. Spec. TT-P-645, or a coat of red lead p~int, Type I or Type III conforming to Fed. Spec. TT-P-86G, applied to a minimum dry film thickness of 1.0 mil. Surfaces that"will be concealed after construction and will require no overpainting for appearance may be primed with a coat of asphalt varnish, applied to a minimum dry film thickness of L 0 mil. Damage to' primed surfaces shall be repaired with the primer.
6-155
TK 5-l300/NAVFAC P-397/AFR
~8-22
. 5.4.4 Painting: Shop painting shall be provided for all metalwork,
except for non-ferrous metals'and corros~on resistant metals and surfaces to
be embedded in concrete. Surfaces to be welded shall not be coated within three inches of the weld, prior to welding. All machined surfaces in contact with outer surfaces and bearing surfaces shall not be painted. These surfaces shall be corrosion protected by the application of a corrosion preventive compound. Surfaces to receive adhesives for.gaskets shall not.be painted. Surfaces shall' be thoroughly dry and·cl"an when the paint is applied. No painting shall
be
done
in
freezing
or
wet weather except under cover; the
temperature shall be above 45' F but not over 90' F. Paint shall be applied in a workmanlike manner and all joints and crevices shall be coated thoroughly. Surfaces which will be concealed or inaccessible after assembly shall be painted prior to assembly. ·Paint shal~ 9onform to Fed. Spec. TT-P37D. 6.
BlAST VALVE TESTS: .
6.1 Response Tests: The following static and dynamic response shall be performed to demonstrate the b l ast; resistant capabilities of the blast value design. These tests shall be witnessed by the Contracting Officer. 6.1.1 Closure Time Test: Prior to shipment to the site, the Contractor shall perform a.test to demonstrate that the closure of the blast valve will not exceed the 20 milliseconds specified. A suggested method for recording the valve closure is with the use of a high speed camera. 6.1,2 Static Pressure Test: Prior to shipment to the site, the Contractor shall perform .a pneumatic test to demonstrate the static capacity of the blast valve design,. The valve must sustain the pressure of 100 psi for a minimum of two hours. The total pressure'loss during that period shall not exceed 1 psi. 6.1.3 Dynamic Pressure Test: Prior to shipment to the site, the Contractor shall perform ~ test to demonstrate ,the dynamic capacity of the blast valve design. This test shall simulate the combined effects of impact forces produced by. the valve closure system and the blast load. This test may. be replaced by design analyses which demonstrate that the head and frame of the valve shall have the capability to resist the· stresses produced by the' above forces. The blast load as indicated on the drawings shall be used for this analysis. 6.1. 4 Blast Tests: ,'If the effects, of one or more of the above blast valve tests have been demonstrated by prior'blast valve tests on similar valves, then the results of ·these tests shall be submitted for review; and the above test performances may not be required. 6.2 Test: After installation, a trip test shall be performed and demonstrated to the operating personnel. 6.3 Hydraulic Characteristics: Prior to shipment to the site: the Contrac.tor shall perform a final test. to establish the hydraulic characteristics of each valve and provide the nece s s ary cor r ec t Lons and adjustments as stated previously.
6-156 .
-
e
e r
A
.' . "'''W/-W~~\\VJ/7f1i,~,
....•
2 FT. DIA. STEEL PIPE
STEEL ARCH.
•
....vo,
."
-..l
2 IN. THICK PEA GRAVEL· ..•.... '." , ..... " .-.• ] [ 3/16 IN.:MESH: :." ::;¢:SAN[).'~·~··>·:: WIRE SCREEN .. '" ",:' '.', .':. r
..
:9 '. ·
=0 o
1/2 IN. f6 STEEL PIPES, I FT. O.C. SECTION A-A
'0.f8~;1 ,..'
'(Ja'o t,
;,~:.~~,~~L::,~i.~~~l~L~~,.
,[(.JI ,UlJ,.'l..J!'·.'! ~
·"oto.,
3': •
L
GRAVEL SUMP
.'7
I 12JN.P STEEL PIPE
10'-0"
L. A
Figure 6-59
SHELTER
Sketch of 300 cfm sand filter
REINFORCED CONCRETE (FLOOR MUST SUPPORT II TONS OF SAND)
J'--- 9
I
01
~
CLOSED LATCH;CLOSED POSITION LIFT MANUALLY TO OPEN
O'
'", .... <.n ex>
,
.
. ."
".:-
o
°.
-[J
~
1':1 0
0
LS ."" Il-r»
0
o
•
°1
0
~
L-
~
.
rr"J
•
PROTECTED SIDE
I--
l-.
°
I-- BLAST 0
0
0
. FLOW
0 o
TWO LOUVER OPENINGS , I
I
15 in. x 3'2 in. , 52'2 sq. in. EACH
Figure 6-60
-
I '-
SEAT SPONGE RUBBER
L.~.
OPEN CLEAT. OPEN POSITION TILTS AUTOMATICALLY TO CLOSE
Blast-actuated louver
e
e
Figure 6-61
Arrangement of multiple foJvers for a large volume of air 6-159
.'
BACKING· . PLATE
.
• .. .
,
ACTUATING: PLATE.(CLOSED . . POSITION) '~--t-.:....c....-.:_-.:..
'
!r'---ACTUATING PLATE (OPEN POSITION) +-BLAST
SPRING NORMAL VENTI LATION :.:'~ :::.'. -,' :
-..
.', . '::,.:"> ;.... . ~".
" ~'::. <4.
Fi gure 6-62
VALVE SEAT
Typi ca 1, bla s t-e c tue ted poppet valve
6-160
rr:- - - VALVE
(OPEN POSITION)
ON-r__~~_
LENGTH DEPENDS CLOSING. TIME .'.:
l-
en
. ...
".
. Cl
.
:-'~~.':"'~ ";, ":"
.'
' ...
:3 lD
.: •
;.' " ",
.
:~ '.~'.~
EC~~~~~N
'..
.;
I;:~:·I (ALTERNATE TO USING
....,,'..".'. '. ~ ttl- [;l-n' .~:. .... . .. '.:.
"
.
I
. . ... ~
." '
'
LONG DUCT)
:
\
'
':. L..:-"':"""----
.,: ;.:', -:':'.'; : " ; -, . .. ",
:.
.'
~
.'
.
"
",
VALVE---(CLOSED POSITION) NORMAL VENTILATION
Figur~ ?~63
6-161
Tjme delay path
'.
)( (t) MASS
.,
r
I ,
OASHPOT---4-r~
I I
, :':--ISOLATOR
L
.J y (t) ~:
.
, ..
.. ,
,
Figure 6-64
Idealized model
6-162 "
of
shock isolated mass
TK 5-1300/NAVFAC P-397/AFR
Blast Valves
Table 6-13
Blast Characteristics
,
Air
Type Name of' Valve
(Actuated)
U.S. Army Chemical Corps
Cloaing Locking Flow Pressure Time Mech- Rate.
(ma'
(pai)
100
•
Ye.
300
Blast
50
•
Y..
600
100
•
Y..
600
50
50
Y..
1,200
50
•
Y.~.
2,50,0
•
20
Y..
5,000 Shock tube
E191Rl 12"0
Office of
Tested
(ofm)
Blast
M-l U.S. Army Chemical Corps
ani.ma
Field Shock tuba
end field
Blut
end
16"0
Civil Defense
"
Remote
24"0 Blast and Remote
Office of Corps of Enginee~~
. ,
Bureau of Yards and Docks ~
Remote
Field
•
500
2,200
Y..
to 30,000
•
2,250
Temet USA, Inc.
Blast
100
20
Y..
Shock tub.
to 3,100
4"0
150 0
Jearn end Plat
•
280
8"0
600
Field
1',750
14"0
.-.1
Technical Remote
Facilities 1'8-107 A2
•
•
;
American Machine Foundry Luwa
Ne. Naval Civil
Brechenridge
Engineer-
Bayles-Denny
Blast
100
•
Y••
Blast
15 to 160
2
No
Blast
4 to 20
•
100
•
No
600
• • •
No
180
No
125
No
'00
Blast
5 to 100
ins Lab
5 to 104
Stephenson Artes Machinery (Sand Filtered)
Blast
*
• "
• Suffield Experimental Stations
35,000
100
Unknown
6-163
,
80
•
450 Shock tube
,
•
•
• Compressed Air
Field
88~22
TH 5-l300/NAVFAC P-397/AFR 88-22
SHOCK ISOLATION SYSTEMS 6-45. Introduction Previous sections have presented methods for the prediction of blast and fragment effects associated with the detonation of explosives and the design or analysis of structures to withstand these .effects. .In the design of shelters, an important part of the design process is to insure the survival of personnel and equipment. It is possible that the structure could withstand the air blast and ground shock effects but the contents be so severely damaged by structure motions that the facility could not accomplish its lntended function. A similar problem is in the design of shelter type structures that houses sensitive explosives.
Th~se
explosives must be
protecte~
from struc-
ture motions since these motions could result in initiation of the explosive. This section deals with the protection of Vulnerable c9mpohents from structure motions due to air blast and ground shock. 6-46. Objectives The objective of shock isolation ..in protective design applications is to reduce the magnitude of motions 'transmitted by a Vibrating structure to personnel or shock sensitive equipment. These motions must be attenuated to levels tolerable ~o personnel and to be various pieces of equipment used in the facility. A second consideration in some cases is to reduce the magnitude of motions transmitted by Vibrating equipment to its supports. These latter motions can be significant for equipment mounted on shock· isolated platforms. The general functional objectives of a 1.
s~ock i~olation
system are:
Reduce input motions to acceptable levels.
2.
Minimize rattle space requirements ,consistent with system effectiveness and cost.
3.
Minimize coupling. of horizontal and vertical motions.
4.
Accommodate a
5.
Limit the number of cycles of motion of the isolated body.
6.
Support the system under normal 'operating conditions without obj ectionable motions ..
7.
Maintain constant attitude under normal operating conditions.
8.
Accommodate changes in load and'load distributio~.
9.
Maintain system vibration characteristics over long periods of time.
~pectrum
of inputs of uncertain waveforms.
10.
Interface properly with other components or parts.
11.
Require minimum maintenance
6-164
TM 5~1300/NAVFAC P-397/AFR'88-22
6-47. Structure Motions Ground shock results from the energy which is imparted'to the, ground by an' explosion. Some of this' energy is transmitted through the air in the form of air-blast-induced ground shock and some is t r ansmf t t e dvt.hrough the ground' as direct-induced ground shock. 'Both of these forms of ground'shock when, imparted to a structure will cause the structure to move' in both a vertical
and horizontal direction.
Movement' of the structure imparts' motions to items
attached to the structure's interior.'
Motion of interior items-is'obtained
from a response spectrum. This is a plot giving the maximum responses (in terms of displacement, velocity, and acceleration) of all possible linear single-degree-of-freedom systems which may be attached to the structure due ,to a given input motion. Therefore, having the spectra for the structure ,and , given input motion, the maximum response of any item within the structure is obtained based on' the natural' frequency of the item. 'Methods for preparing response' shock spectra are presented in Chapter 2 'of' this' manual.,' . t ·
"
I.
In addition to motion of the structure as a whole, the exterior walls and roof respond to the direct application ,of the' blast' load. Methods for calculating the response of these elements are ,given in Chapter 3 of this manual using the, parameters given in ,Chapter 4 and S'for concrete and steel, respectively. ' Maximum displacements, velocities, and accelerations of these elements 'can be determined in a straightforward manner.
These quantfties:can' be' used to
determine effects on items attached or located near walls or roofs. 6-48. Shock Tolerance of Personnel ana
Equipment'~
The requirement for shock isolation is based upon the shock tolerance of personnel and/or critical items of equipment, contained within the protective '" structure. If the predicted shock input exceeds the shock tolerance of personnel, a shock isolation system is required. If ,the shock input exceeds the shock tolerance of equipment"the,equipment can either be:ruggedized to increase its~shock tolerance or it can be shock isolated. There 'are practical limits to ruggedization and the costs 'may exceed those of an isolation system. If the. input does not exceed the shock tolerance of the equipment,it 'can be hard·mounted to the structure.
r
6-48.1. Personnel
The effects of structural motions on personnel depend on the magnitude, duration, frequency, and direction of the motion, as well as their position at the time of the loading. The shock tolerance of personnel is presented in Chapter I of this manual. 6-48.2. Equipment In many cases, the need for shock isolation of equipment must be established before detailed characteristics of the system components are established. Further, because of the constraints of procurement procedures, shock isolation
systems ment to minimum data is
must be 'designed and built prior to specific knowledge of the equipbe installed.L,In such instances, the choice lies between specifying acceptable shock tolerances for the new equipment or using whatever available' for, similar types .o f equipment. «.
6-165-
TM 5-l300/NAVFAC P-J97/AFR 88-22 . The most practical means of determining the shock tolerance of a particular item of equipment is by testing. However, even experimental data can be of questionable value if the test input motion characteristics differ greatly from those that would actually be experienced by the equipment. Since testing of equipment may not be practical in many cases due to the amount of time allotted from the inception of a project to its completion, procurement procedures, and cost limitations, it is often necessary to rely on data obtained from shock tests of similar items.' The shock capacity of various types of equipment is presented in Chapter 1 of this manual. 6-49,. Shock Isolation Principles 6-49.1., Gene,ralConcepts
,.
:
A full. ,treatment of the problem, of shock isolation systems is not possible in this manual. The following discussion p rovf.desven introduction to the subject and presents some of the important characteristics of shock isolation systems. In general, the analytical treatment of shock isolation systems is based upon the principles of dynamic analysis presented in Chapter 3. In most cases, the actual system can be represented by a,simplified mathematical model consisting of a rigid mass connected bya spring and dash pot as shown in Figure 6-64. The figure represents the simplest case, that of a single-degree-of-freedom system restrained to move in only one direction. Actually, an isolation system would have at least six degrees of freedom, i.e., three displacements and three rotations. Under certain conditions, these six modes can be uncoupled and the system analyzed as six single-degree-of-freedom systems. The single-degree-of-freedom system shown in Figure 6-64 can be used to illustrate,the importance of some of the parameters affecting the effectiveness .of shock isolation systems in general. The isolator is represented by the linear spring and viscous damping device enclosed within the dotted square. The suspended mass is taken to be a rigid body. It is assumed that the base of the .system is subjected to a periodic sinusoidal motion whose frequency is f. The undamped natural frequency of the system is f n_ and is given by: " ;.
386.4 K]
1/2
6-62
W
, where
fn
natural frequency of vibration
K
unit stiffness of spring
W
weight supported by spring
When the frequency of the disturbing motion f is 'small compared to the natural frequency f n of the single-degree-of-freedom system, the displacement of the mass is approximately equal to the displacement of the base. When the frequency of the base motion is several times that of the system, 'the motion. of the mass is a small fraction of the base motion. When the ratio 9f frequencies become large (20 to 30). the system can not respond to the base
6-166
TM 5-l300/NAVFAC P-397/AFR 88-22
motio~'to_any
significant' degree. At 'frequency ,ratios near ,one, large motions of the mass are possible and the magnitude is strongly affec eed by the amount of damping in the ,system: +, ~,
One ,obvious shock isolation:approach, is to use a low, frequency suspension system,so that the ratio of frequencies is ,always large .. ' However, low frequency ,(referred to as soft) systems possess 'the undesirable characteristic of larger static' and dynamic' displacements and greater probability 'of coupling between modes of vibration. ' Although soft systems may be acceptable 'under, , some condd t Lons-jr the .obvLous- constraint that will preclude their use is a. limit on the 'relative motion between the, suspended mass and,its supports or 'adjacent parts of the facility . .This relative motion 'determines the:amount of rattle space that must be provided to avoid impact between the mass and other' fixed or moving parts of the facility. .
- : . (>;'
! .
. -
r -: :
'.1
The acceleration .of 'the 'mass -Ls a func t t orrof the forces applied to the mass by the spring and damping ,devices. In the case of a linear undamped 'spring, the force is'a function of the relative displacement between the mass ,and 'its support: ; '-In '·viscous damping devices,' -rhe .damping force is a function of the percent.damping and the relative velocity between the mass and its 'supports. Acceleration limits for the' 'critical ':items will 'impose restraints' on jsprLng s t Lff'ne's s and' theaniount of, damping rLn the isolation 'system.' In practice, a ' compromise combination of spring'stiffness'and damping is
.necessary~to
minimize 'input,motions .to the mass for a ,specified "a I LowabLe rattle space or: to minimize the:rattle"space required .f'o r 'specified allowable motions of ,the,
......
",
mass .:
'.
-,
'
,
.
The need to -avo Ld resonance (ratio of the ,frequency of the' base motion to the nacura L'tf'requency of the isolation, system equal .t o one)" is obvious. The structural motions ,resulting-from an explosion 'are not steady-state sinusoidal in nature. .However , these motions ;are-of an oscillatory type and the dis- . placement-frequency relationships discussed above are applicable . . . -: A more'detailed'discussion'of'the effects of load duration, nonlinear springs, damping,' and system 'frequency on response can be obtained from publications listed inthe.bibliography . . :,"
r
:.
• ""1,
•
I':
The'basic,objective in'shock isolation is to select a combination'of isolation system properties which will reduce the input motions to the desired level. In design, it is a straightforward process. System properties are assumed and an analysis .Ls performed to determine Ltsvresponse to the'input motions. If the shock.tolerance and.rattle space criteria are not satisfied, the 'system, must be altered and the analysis repeated until the criteria"are satisfied. 6-49:2. Single-Mass Dynamic Systems 'r l
"
..
A single. mass ,system can have 'six degrees of freedom, -that is, ttanslation in three orthogonal axes ,and three rotations. The system can'also be classifi~d as coupled or uncoupled. -.'
'.
•. j .
fl
A, coupled system is' one in which forces o r-vd Lsp Lac emerrts in one mode will affect or 'cause, a response in another mode. ,For example, a vertical' displacement of a single rigid mass might also cause rotation of the mass, 'An uncoupled 'system; on the other hand; is one·where forces or displacements in one,mode do not:generate a'response,in anofher,mode. If ,the 'system is
TH 5-l300/NAVFAC P-397/AFR 88 c22 completely uncoupled, base translations in anyone of the three orthogonal directions will cause translations of the mass in that direction only. 'Similarly, a pure rotation of the base about anyone of the three orthogonal principal inertia axes with their origin through the mass center, will cause only pure rotations of the body about that axis. The principal inertia axes are those about.which the products of inertia vanish. The principal elastic axes of a resilient element (isolator) are those axes for which an unc~n"' strainedelement.will experience a displacement collinear with the direction of, the applied force. If the principal elastic axes and the principal inertia axes of the shock,isolation system' coincide with the origin or point of intersection of both sets of axes at the center of gravity 'of the mass, the modes of vibration are uncoupled. Such a system is a'l so vre fe r red to asa balanced system. In Figure 6-65, if all the springs have tlie same elastic stiffness, the elastic center will .be Loca t e'd at point A, ,which, in this case, is, at the center of. the ·individual springs. If the, suspended mass, -Ls of uniform density, its ,center o f gravity is also located at A, and the system is uncoupled for motion input through the springs: Some systems may ,be ,uncoupled' only -fo r motions in a par.t Lcul ar direction., "If point, B in Figure 6 - 65 is the -, , center of gravity·of the mass, a horizontal motion ,in the direction parallel to the X-axis of the structure would cause only a horizontal motion of the mass. A vertical motion of·the structure would'cause both ,a vertical and rotational ,motion of the mass .. In this case, the vertical and rotational modes are coupled. If the center of gravity were located cat point C"then. vertical, horizontal and rotational modes are coupled. If the characteristics; of the mass and shock isolation system are such that the modes of vibration can,pe uncoupled, the system can be analyzed aS,a,series of independent' single-degree-of - freedom, systems. The response of. each -of these systems .can : r be determined on. the basis of, input motions' and isolator proper t Ie s in a ' . direction parallel, to or about one of t.herpr LncIpal inertia axes .. ,The response in each one of these modes can be summeq in various ways.to obtain the total response of the system. The sum of the maximum responses would neglect differences in phasing and should represent an upper limit;of "the actual mO,tions.' It is, recommended that the square root, or'the sum of the, squares of the maximums (root mean square values) be used to represent,a,' realistic maximum response since it is unlikely that response will occur simultaneously in all modes. Superposition of modal response is appropriate for elastic systems' only. , . . , A dynand cal LyibaLance d shock :~olation system, offers advantages other than a simplification of the computation effort .. ' A balanced system results .Ln reduced motions during oscillation. As a result o~ the absence of coupling,of modes in a balanced system and the usually.small, if any, rotational inputs to the system in protective construction applications, rotational motions of the shock isolated mass will be minimized. This is particularly important for large masses where.small angles of rotation can result in large displacements at locations :far from the center of gravity. ' Because of the advantages of a dynamically balanced system, various,approaches are taken to minimize coupling of modes. One criterion is that frequencies in the six modes should be separated sufficiently to avoid resonance ,between modes. Because of the importance of minimizing rotational modes of response, it is suggeste~ that extremely,low stiffnesses in these modes should be avoided . . For .tihe . analysis of multiple degrees of,.freedom, .s LngIe -mas s sys cems 6-168
TH 5-l300/NAVFAC P-397/AFR 88-22
where the various modes of response are coupled, the modal method of analysis or numerical integration techniques can be utilized. The modal method of analysis requires solution of simultaneous equations of motion to determine characteristic shapes and frequencies of each mode and is limited to the elastic case. The numerical techniques dO'not require predictionof'mode shapes and frequencies and will handle both elastic and inelastic response. If the dynamic system is also a multiple mass system, the above methods can be utilized to an~lyze the system. While an in-depth discussion is beyond the scope~of ·this,manual, a complete discussion of these methods can be found in publications .listed in the bibliography .
.'
6-49.3 .. Shock Isolation Arrangements .' 6-49.3.1: Individual versus Group'Mounting'
-,
The two, basic approaches to shock isolation in protective construction are to provide individually tailored systems for each component and to group together two or:more items on a.common platforrn. i In the latter case, the" system is selected to satisfy the requirements'of the most critical·:item. In some cases,'where the shock tolerance of the various items differs greatly,a combination of the·,two approaches may"be the :most effective solution. : Although the relative location or size of some items may 'make individual mounts' the more practical approach in ·certain cases, group mounting will
generally be as reliable and the least costly solution. ' Where personnel must be protected, 'a platform is the most'practical solution. Except for extremely sensitive'equipment;.ihe shock tolerance of the 'personnel will govern the: .das i gn of, the sys t em- . The combination of personnel and equipment on the same platform will permit'the personnel'to move freely (however cautiously) between items of equipment. Where personnel are ,not required to be mobile, but rather may be able ,to remain seated while operating the'equipment during' hazardous periods, the. shock tolerance of the personnel are greatly increased. This ·increased tolerance will reduce ,the shock isolat"ion requirements while 'at the same time affording a higher degree of protection for personnel since they are protected from the unknown consequences .of falling. There are several advantages of 'group mounted systems. A group mounted system is less sensitive to variations in weights of individual items of equipment because of the larger combined weight of all items and the platform. With a number of items there is a greater flexibility of controlling the center of gravity of the total mass. Lnvfac t , ballast 'may be 'added t o . the' platform to align the center, of gravity with the princ"ipal axis to form a balanced system. A gzoupuaount.ed system generally requires -Les s "rattle space than several independently mounted··items. Also, the interconnections ·between·components is greatly simplified if they are all mounted on a single platfor~. Finally, an important advantage of group' systems is cost . . Individual mounts will require a large number of isolator units. Although larger. more costly, units are required for the group mounting system, fewer units are required and the cost per pound of supported load will be 'much lower,
,."
.
6-169
TM 5-l300/NAVFAC P-397/AFR 88 c22 6-49.3.2. Platform Characteristics A platform for group mounted systems offers great flexibility in controlling the center of gravity of the supported masses ,to produce a balanced system where modes ;of, vibration are uncoupled. Ballast may be securely anchored' to the platform at locations which would.inove the center of gravity of the total mass to coincide with the elastic center of the isolation system. The ,. determination of the weight and location of;this ballast can be' greatly s Lmp Lf.f Led vby uncoupling the effect of adding weight in the x and y directions of the principle elastic axes. This uncoupling can be accomplished b~. , locating the ballast symmetrically about the x axis when moving the location of the center of gravity in the y direction. In this manner, the location of the center of gravity may be altered independently about the elastic center in the x and y directions. If for practical·reasons the ballast cannot be' " located symmetrically about a principle axis, then the two directions must be considered simultaneously. Providing additional ballast in excess of that required to balance the platform provides .for future changes in equipment. or the addition of new equipment without actually changing the isolation system. The springs will not require replacement nor will the structural members of the platform ,need to be increased in size. Additional equipment is placed on the platform and ballast is removed and/or relocated to balance the new equipment arrangement. ·To provide for future equipment changes,it is suggested that additional ballast equal to 25 percent· of the weight of the equipment and the. required ballast be distributed on the platform. The location of this ballast must not change the center of gravity of the existing balanced..system. If future needs have been established, the platform and isolators would be. designed for. the future equipment. However, ballast would be provided to compensate for the weight of the future equipment and balance the. system for the, existing equipment. The stiffness of the platform must be. large enough to insure ·that the platform and associated group mounted equipment ..can be treated as a rigid body.. This criterion is usually satisfied if the lowest natural frequency of any member , of·the .platform is at least five (5) times the natural frequency of the spring mass system. When large, heavy items of equipment are involved, platforms ',' meeting this stiffness ,criteria may not be.practical. In such cases, the platform equipment configuration should be treated'as .a.multi-mass system. 6-49.3.3. Isolator Arrangements There are.many ways to support a shock isolated item .. Some desirable features have been discussed previously in connection with dynamically .bal anced systems. The isolators may be positioned in many ways. The more important factors affecting· the selection of an isolator arrangement are:. 1.
The size, weight,'shape and. location' of the' center of gravity·of. the
~uspended
mass; , 1.
2. 3.
The direction and magnitude of the input motions; Rotation of the lines of action of the devices should be small over the full range of displacements of the system to avoid system nonlinearities; 6-170 .
TH 5-i300/NAVFAC P-397/AFR 88-22 4.
Coupling of modes should be minimized;
5.
Static and dynamic instability must be prevented;
6.
It is desirable in most cases, and necessary in some, that the system return to its nominal position;
7.
Space available for the isolation ,system; and type of isolation devices used.
Some of the more common isolator arrangements are shown in Figures 6-67 and 668. The systems 'shown are assumed to have the same arrangements of isolators in a plane through the center of gravity (e.g. ).. and perpendicular to the surface of the page. The dynamically'balanced system (intersection 'of the elastic axes and the principal inertia axes located,at point A) shown in Figure 6-65, is probably the least common of 'all suspension systems. 6-49.3.4. Base-Mounted Isolation Systems In Figure 6-66a, the mass is supported on ',.four , (4) isolators. These isolators must provide horizontal, vertical 'and rotationalstiffnesses in order for the system to be stable under all possible motions. There will be coupling between horizontal displacements and rotations ·about horizontal axes.
This
arrangement and that shown in Figure 6-66b are appropriate' in those cases where there are no convenient supports for horizontal i~ol~tors .. The arrangement of Figure 6-66b is preferred since the line of action of the isolators can be directed towards the e.g. of the mass to allow decoupling of some modes. As in the case of Figure 6-66a, the isolators must possess' adequate stiffness·in axial and lateral directions to insure'stability under static and dynamic conditions. In Figure 6-66c, the'isolators are oriented parallel to the three orthogonal system axes. This arrangement provides system stability even when the isolators possess only axial stiffness. If the e.g. of the suspended mass is located as shown, decoupling of modes is possible. While. the lines.of action of the isolators pass through the e.g. under static conditions, response of the system to base motions will obviously alter its geometry. :When the line of action of the isolators is changed due tO'displacement ,of the mass ..relative to its supports, coupling of the modes of vibration will be introduced. The degree of coupling is affected by the'magnitude of the displacements and the length ,of the isolators. Consequently" isolator properties and arrangement should be selected so as to minimize the effects of displacements. 6 -49.3.'5. Overhead Pendulum Systems Using,Platforms Two arrangements of overhead,pendulum shock isolation devices using platforms to support the sensitive components are shown in Figure 6-67. In both cases, the center of gravity of the suspended maas Ls relatively Lov . These types 'of suspension systems have been used ,extensively in protective structures for various ,conditions including individual small and large items,.multiple items of various sizes as 'well as a combination of personnel and equipment supported on various sized platforms. The overhead pendulum system normally uses swivel joints at the points of attachment so that the system may swing freely. Horizontal input motions cause the pendulum to swing. Gravity provides the i
6-171
TM 5-1300/NAVFAC P-397/AFR 88-22. horizontal restoring force or stiffness. This force is a function of the total weight of the suspended mass. The natural frequency of vibration of the pendulum is a function of the length of the pendulum and is. given by: 1 fn -
386.4] _ _ 1/2 [
6-63
L
-.
where
natural frequency of vibration length of pendulum. 'Each pendulum arm includes an isolator which sstablishes the stiffness of the system in the vertical direction. These .isoiators can introduce nonlineari~ ties and coupling between the pendulum and·vertical spring modes. The system is linear for small angular displacements, that is, when the angular change e of the pendulum arm from the vertical position is approximately equal to the sine of the angle (9 - sin 9). The system can be considered uncoupled if the pendulum· frequency is not near 'one half. of' the vertical spring frequency. , If the pendulum frequency is in the vicinity of one half the vertical frequency, the interchange of energy between the modes can lead .to pendulum motions greatly exceeding those predicted by linear assumptions.. . In a shock spectra maximum displacements occur at low frequencies, maximum velocities at intermediate frequencies, and maximum accelerations at high frequencies ... Since most pendulum systems have low. natural .frequencies, they are displacement sensitive. These systems attain maximum displacements and minimum accelerations. Consequently, they will normally require greater rattle:space than other systems while at the same time providing.maximum protection against horizontal accelerations at minimum costs. It should be realized that for explosions, 'maximum displacements are comparatively small and can be accommodated. One of the main advantages of overhead pendulum systems is that they do not require horizontal stiffness elements. Their attractiveness is greatly diminished in those cases requiring horizontal damping because .of large motions. The swivel joint attaching the pendulum arm to the platform determines the location of the horizontal elastic axis of the system. Figure 6-58b illustrates two ways of varying the point of attachment of the pendulum arm.to the platform. The horizontal elastic axis'is raised to coincide with the center of gravity of, the suspended mass at the .equilibrium position and help minimize coupling between modes of response. At the left side of the platform the' isolator is contained in a housing rigidly attached to the platform. At the right side, a structural member is rigidly attached to the platform and the isolator is included in the pendulum arm. In addition to supporting personnel and equipment, overhead pendulum systems can.be used to shock-isolate building utilities. Individual utility runs may be isolated or several different utilities may be supported on a single platform. A single platform may cover an entire room and all building services may be supported. They would include a hung ceiling; lighting fixtures, utility piping, HVAC ducts, electrical cables·.and process piping. Of course, flexible ·connections. must be used when connecting the.services to the building or equipment.
6-172
TK 5"1300/NAVFAC P-397/AFR 88-22 6"50. Shock Isolation Devices 6-50.1. Introduction A fundamental element of every shock isolation system is some sort of energy storage or'energy dissipative device. These devices 'must 'be capable of supporting the ,items to be isolated under static arid dynamic conditions and, at the same time, prevent transmission of any harmful 'shock loads to the items. In most cases, the isolator must have elastic force-displacement characteristics so that the,system,will return to a nominal equilibrium'· position after the dynamic loads have been applied. The desirable features of these devices include: ' 1.,
. The dynamic force-displacement relationship of the isolator should be'predictable for all directions'in which it'is required to provide stiffness."
.
2. 3.
,
..
The isolator should have' low mass in order to'minimize transmission of high frequency motions to the'supported mass. The frequency of the isolator should'remain constant with changes' in load, that is, its, stiffness should vary in"direct proportion 'to the load 'it supports.' This allows 'the system, to' remain dynamically balanced throughout changes in the position of the supported mass.
.
4.
The static position of the isolator should be adjustable so that the' system can be, leveled and returned to its nominal position should the suspended load change.
5.
The isolator should have high low cost,
reliabili~y,
long service life and
The various types of isolators used in most protective construction applications possess these characteristics in-varying degrees. Any real' isolator has some mass, and in some applications, the mass can be quite large and must be considered in the final analysis.' Nonlinear force-displacement characteristics are often accepted to gain some other advantage. In energy dissipative systems, it, may be necessary to provide other means of , restoring the system to its original position. In general:.. most devices are some' compromise combination of' the desirable features which best suit the particular design situa, tion. .t ' The inclusion of energy dissipative (damping) devices in the isolation system offers .s eve raL significant advant.age s;" that is ,'damping can: 1.
Reduce the severity of output motion response; "
,2;
.
, Reduce the effect 'of coupling between modes, thus' reducing' rattle t ' space requirements; " '.
..
'3..
Restore, the system to an equilibrium position more quickly,;" "
,
,4. " Decrease' the, sensitivity of the system to var Let Lons in input motions. 6-173
TH 5-1300/NAVFAC P-397/AFR 88-22 Damping can be provided internally in some isolation devices such as in liquid springs, but must be added externally in others such as those systems using helical coil springs. Different types of damping offer advantages and disadvantages which must be evaluated in the design process. A damping device may be effective in attenuating low frequency components,of input motions but can increase the severity of high, frequency components. Also" a damping device could prevent the system from returning to its nominal equilibrium position., Thus, ca~e must be exercised in either designing a system employing isolator. devices possessing inherent damping charac~eristics or adding damping devices; 'if the isolation system is to perform properly. There are numerous types of isolators'~hich can be used to accomplish the shock isolating function. In the design ,of protective structures for explosions, the induced building motions are not.usu~lly severe and the maximum building displacements are relatively small. As a result, shock'isolation systems using helical coil springs (Figure 6-68) are by far the most common system employed. The reasons for "the extensive use of helical springs should be 'obvious from the discussion below. Other shock isolation devices which may also be used are presented, in less detail, ~elow. . It should be noted that the protective design engineer does not furnish the design for the shock isolator. The engineer designs the shock isolation system to be used but ,does not design the isolators,,(in most cases, a helical coil spring). Rather, specifications are furnished which define the desired characteristics of the isolator. For a helical spring, the specifications may include some or all of the following: maximum load, maximum static deflection, maximum dynamic ~eflection, spring stiffness, maximum height, maximum, diameter, and fac~ors of safety regarding allo~able stresses and bottoming of the spring. It must be realized that as the number of specified parameters increase, the options available to the spring manufacturer are decreased. 6-50.2. Helical Coil Springs A helical' coil spring is fabricated from bar stock or wire which is'coiled," into a helical form. Figure 6-68 illustrates several spring mounts . .
(.,
The helical coil spring has numerous advantages and comparatively few.disadvantages. The advantages are that the spring is not strain,rate sensitive, self, restoring after an applied load has been removed, resists both axial and lateral loads, linear spring rate and' requires little or no maintenance. For most applications, ,the coil spring usually ,requires a larger space compared to other available shock isolators; and the spring cannot be adjusted ,to compensate for changes in loading conditions. If the weight of the supported object is changed, it ~s,necessary to either change the spring or add additional' springs. For most purposes, the helical coil spring can be considered to have zero damping. ' If damping is required, it must be provided by external means. Helical coil springs may be used in either compression or extension. The extension ~prings' are not subject to buckling and may offer a more convenient attachment arrangement. However, extension spring attachments are usually more costly and cause large stress concentrations at the point of attachment. For shock isolation applications, coil springs are generally used in compression. Buckling which can be a problem with compression springs, can be overcome by,proper design or ,through the use of guides,which are added either 6-174,
TM 5-l300/NAVFAC, PC397/AFR,88-22,
internally or externally. to the coils. 'The discussion below will be concerned primarily. with compression springs unless otherwise stated. Helical coil springs may be mounted in two ways, the ends are either clamped or hinged. In most shock isolation applications, the spring ends are clamped since this·method greatly increases. the force required. to buckle the spring. If space is at a: premium, the.energy. storage capacity may be, increased.by. ne s t Lng. the springs'. (placing' one or more springs inside the.outermost spring) .. When nesting spr.ings;li.it is· advisable to alternate the direction of coils to' prevent the springs from becoming entangled.' Although permanent set may-be acceptable in some instances;' it is normally required that the system return to its original position after being-loaded. This can be accomplished in various ways, but the most common approach in the case' of helical .coil spr Lngs t Ls to prevent· .LneLas t Lc 'action' of the spring . .; : c :
.1.
Helical coil springs are capable of resisting lateral load: While it'is possible, to use ,springs ,in this application, carel should be ,exercised. There are possible arrangements which avoid subjecting the -sp r Lngs to this. type of' loading.;,", ''','. •. q-
While the actual
d~sign
j'
of the helical coil spring is done by the manufactur-'
e r , the: engineer must be .ce r caLn that the springs he is:l specifying can
actually be, obtained' and .the space he has ,allocated for, the springs are sufficient. Therefore, preliminary spring sizes must be obtained by the engineer ,to suit his intended application. I t is suggested ·that avaLLabLemanufacturer-.';s data be used for this purpose. .'
t.,
I
••••
"
,.
6-50.3. Torsion Springs Torsion springs provide resistance to torque applied to the spring~ ~In.shock isolation applications, the torque is usually the result of a load applied to a torsion lever. which is part of the torsion spring system. ' A typical .torsion spring 'shock isolation system is illustrated· in Figure 6- 69 .. ,. ; . ' Since the axis of a torsion spring
is.normar~to,the
direction of displacement,
it can be ,used advantageously when space in the'direction of displacement is limited. Torsion springs have linear spring rates, are not strain-rate sensitive, are self-restoring, and require little or no maintenance. ~orsion' springs can.not be'adjusted to compensate for changes in weight of shock isolation equipment, and damping must be ,provided by external means. The axial length· of some types may preclude their use when space is limited. .
,
,
.
There are three basic types of. torsion springs; (1) torsion bars, (2) helical torsion springs, and .(3) flat torsion·springs. The type to be used will depend upon the space available and the capacity required. The torsion bar is normally used for light to heavy loads, the helical torsion spring for light to moderate ·loads, and the flat' torsion .sp r Lng for 'light loads . The, torsion bar is ,the -rype most, commonly; found in protective structure' applications and is most. commonly used where large . loads ,must be supported. ,'.. ,
,
6-50.4. Pneumatic,Springs •
_I
.~
.,
..
"
"
.
. -: i'J
Pneumatic springs,' are springs' whose action·is due to the resiliency'of compressed air. They,.are used ina manner similar to coil.springs.. The two' 6-175
.. '.
TM.5-l300/NAVFAC P-397/AFR 88-22
basic types are the'pneumatic cylinder with single or compound air chambers, and the pneumatic bellows. The pneumatic cylinder is shown schematically in Figure 6-70. Pneumatic.springs have the advantage of being. adjustable to compensate for load changes. The spring rate can be made approximately linear over one range) of deflection but will be highly nonlinear over another. They' are quite versatile due to the variety of. system characteristics which:can be obtained '.' by regulation cif· the. air flow'. between the cylinder chamber and the reservoir tank. Some of the possible variations include: ' , , ." 1.· :.Velocity-sensitive damping by a variable orifice between chamber ,.l.'.
2. .. 3.
and reservoir;
.
..
J
~
-
,
Displacement-sensitive' damping by a variable orifice controlled by differential pressure between chamber and reservoir; . A nearly 'constant height· maintained under slowly changing static ' .. load by increasing or decreasing the system' aii· content using an external air supply and a displacement-sensitive servo-system controlling inlet and exhaust valves; , , -. ,, A constant· height under widely varying temperatures achieved"by . the same system described, for maintaining a.icons t ant; height. "
The disadvantages of pneumatic springs include higher cost and more 'fragile construction. They have a limited life span in comparison to mechanical springs and must be maintained. Also these springs provide resistance for axial loads only. 6-50.5. Liquid Springs A liquid spring consists of a cylinder, piston rod', and a .ht gh pressure seal around the piston rod. ,The cylinder is completely filled with a liquid, and as the piston is pushed into the cylinder, it compresses the liquid to very high pressures. The configurations of 'liquid springs' are divided into thre'e ma] or classes according to the method of loading. .. The' classes are simple compression, simple tension and compound compression-tension. Although they are loaded .in different .ways·, all three types function as a result, of compression of the liquid in the cylinders. Schematics of the tension and compression types are shown in Figure 6-71. The compound spring is merely a more complex mechanical 'combination of the two basic types. The tension type is the more common. in protective construction applications. The cylinders are often fitted with ported heads to guide the piston and provide damping. Damping can also be provided through the ,addition of drag plates to ,the .piston ·rods. Liquid springs 'are very compact devices with high,nearly linear, spring rates. They 'can be adjusted to·compensate:for.-load changes, areself"restoring and can absorb larger'amounts of energy. They are highly sensitive to ,'". changes in temperature and. fluid' volume changes. Because' liquid' springs normally operate at high pressures, high quality, close'tolerance seals are required around the'piston. Friction between the seal and piston provides appreciable damping and increases the spring rate f~om 2 to,5percent. Liquid springs are high pressure vessels requiring high 'quality materials and pre6-176
TH 5-l300/NAVFAC P-397/AFR·88-22
cisiim. machine work,. and as a result, they are expensive.' 'However, they are difficult to equal as' compact energy absorption devices. 6-50.6. Other Devices' 6-50.6.1. Introduction The helical, torsion, pneumatic and liquid springs are the more common types of isolators for larger masses. There are other devices especially suited for particular applications. and .smaller·loads. Some' of these isolators are discussed 'below. 6-50.6.2. Belleville Springs Be Ll.evf l.l,e. 'springs, also, called Belleville washers or coned-disc springs, are
essentially spring steel washers which have been formed into a slightly conical shape. A typical Belleville" spring is illustrated in Figure 6"72. The main advantage of Belleville springs over other types of springs is the ability to support'large loads at small',deflections with minimum space requirements in the direction of loading, They are useful in applications requiring limited shock attenuation and as back up systems to reduce shock in the event of bottoming'of coil springs. They are relatively inexpensive and readily avai Lab Ie-dn capacities up to 60,000' pounds .. Changes 'in 'loading conditions are accommodated by the addition or removal of units.. "
..
6-50.6.3. Flat Springs A flat spring is simply a steel beam or plate whose physical dimensions and support conditions are varied to' provide the desired force displacement relationship. The two basic configurations' are .rhe simple spring with one, element and leaf springs with multiple' elements. Flat springs normarly· require only a limited' amount of space' in' the direction of displacement and provide li~ear~ non-strain-rate sensitive and self-restoring,spring. They' require little .or no maintenance. Single',element flat springs, 'can be considered to have no damping while leaf springs will exhibit some damping due to the friction between individual' elements. 6-50.6.4. Solid Elastomer Springs Solid elastomer spr~ngs,are made from rubberlike materials. They are often ·called shock mounts because of their wide use in shock isolation applications. They are normally used in medium to light duty applications and represent an. economical .so Lut.Lon .co the isolation of small items of equipment. " However, these springs 'will allow only small displacements. These springs are' fabricated' from a wide variety of natural and synthetic -rubbers and compounds and·. in numerous sizes ,and shapes. to satisfy a wide range 'of applications. ;. Because of the range in capacity and·.characteristics of colIlIliercially available units;' only'in unusual cases is it necessary to design a unit. In most applications, the solid elastomer spring will require little space and exhibits good weight" to energy storage ratios.' Use of these spr Ings vrequt res consideration of the operating environment. The desirable properties of some elast9mers can be"significantly 'degraded'when exposed to low or high temperatures, sunlight; ozone, water.or petroleum pr'cduc t s . 6-177
rn ~-!3QO/NAVFAC P-397/AFR 88-22
The response of elastomeric springs ·.is· nonlinear in most applications because of the nonlinear stress-strain properties of elastomers'. ,.The springs are,' self-damping because of the viscoel~stic properties of the elastomers. They are almost always. in compression because of bonding limitations .' These springs will only. permit comparatively small displacements. ' '.,
6-51. Hardmounted Systems Some.item~ 'of equipment.do not require shock isolation .because· the predicted.
motions at their point of attachment to the supporting ,structure does ,not· , exceed their shock tolerance. Those items can normally be har dmount ed.rco the. supporting structure. A hardmount is a method of attachment which has not been specifically designed to provide a significant reduction in the input motions to the equipment. Since all methods of attachment exhibit some flexibility, there is no precise division between shock isolators and hardmounts. " Both types of devices will modify input motions. to some degree. , However,' the' modification of input· motions produced by hardmounts will generally be small while shock isolators can greatly affect these motions. In contrast "to shock isolation systems, hardmount.ed rsys cems will normally exhibit natural frequencies much higher than those corresponding to, .the .lower modes 'of vibration of 'the supporting structure. Although this characteristic offers .the advantage of' reduced rattlespace, it also provides for the more efficient ·transmission of higher. frequency: components' .of .che support, structure, ' motion to the attached item. Thus, it would appear that a more.exact structural analysis is required for hardmounted systems in order to include higher modes of vibration. In practice, the need for exact analyses 'is at least ' partially offset by higher factors of safety in mount design and equipment shock tolerance. However, such an approach can lead to unrealistic attachment' designs. A more practical approach is to ,choose,'or design, attachments'which limits the funda~enta~ frequency of the hardmounted system. A lower frequency system provides some attenuation of higher frequency'input moti9ns, .and reduces the possibility of resonance' with high frequency. motions ·resulting from stress wave reflections within structural' elements.- Although the choice of a natural frequency ~ill depend on the properties of the supporting'structure and the..har dmourrted 'equipment,' fundamental frequencies in the range of 10 to 1000 cycles per second are reasonable for most applications. The approach,~' chosen for hardmount design is normally a combination of higher safety factors and the use of lower frequency systems. The design will be based upon' considerations of cost, importance of the item supported, the size and weight of the item, ,and the consequence' of failure of the .attachment. system,c.·
.
.;.
The use of, shock spectra.to define the 'input motions of hardmount systems is considered adequate for final design of all' simple hardmount systems of a noncritical nature. ' It is also 'considered ·adequate for preliminary design. of critical systems 'and those whose representation as a single degree of freedom system is questionable. However, it is recommended that the final design be' performed using a more exact dynamic analysis wherever practical. 6-52. Attachments 6-52.10 Introduction
"
In a shelter type structure subjected.to air blast and ground shock effects, all interior contents must be firmly'attached to the·structure'. This attach6-178
,"
, r,
TK 5-1300/NAVFAC P-397/AFR 88-22 ment insures that the building contents will not be dislodged and become a source of injury to personnel or damage to critical equipment. The building contents would include not only equipment which is either shock isolated or hardmounted (attached directly to structure) but also the building utilities as well as interior partitions and hung ceilings. The building utilities would include all piping (such as process, potable water, sanitary, fire protection, etc.), HVAC ducts, electrical cables, light fixtures and electrical receptacles.
6-52.2. Design Loads An object subjected to a shock loading produces an inertial force which acts through its center of gravity. The m~gnitude of this force is given by: F - Wa
where F W a ~
inertial force Weight of object acceleration in g's
Accelerations. may be ,imparted to the object in one or more directions producing inertial forces.in the. respective directions. These inerti~l forces are resisted by the reactions developed at,the object's supports'. 'All inertial forces are assumed to be acting on'the object concurrently.! The support reactions are obtainedby'cons1dering the static equilibrium of the system.
6-179
y .
:
a
.
a
,.
1/ / , / 1/ :~ > : 1/ •: 1/ .... £.. .... .
,
L'
;.:,
/
/
/
/
"
J7
/
_,I
...
:
~ ~ ;.
,
1/ /
eC -
.1/
,
~'.".
/
,
. """YVy
A'S
1/ L..u.... .;
/
1/
........ 1/1/
-
1/
/
/
:= .
'
1/
.
, ., .
,
.'
'
.
: . :~ • ~
,
...... 1/ -""',,.,,- , .'
/
~ "
/
/
Figure 6-65
/
/
/
/
///
Shock isolation system
6-180
j
b
xP. :
"
EQUIPMENT
C.G.
HARD
ELASTIC ~ ELe:MENTS~
STRUCTURE
(0 )
EQUIPMENT
C.G.
I
( b) , r.
EQUIPMENT 1·
PIVOT;ED JOINTS
. HORIZONTAL
(TYp..)
ISOLAT;:
,.
C.G.
FLOOR
PLAT ORM b'?"'/.'l77.,L,.-7'7":""""'''7''''"''?'''7'''TT7''7';,..,J-;~7''7';''7" r; VERTICAL ( C) ISOLATOR
Figure 6-66
Bas e-rnoun ted isolation systems conf.t qur-a t ton
6-181
r
HARD STRUCTURE
"-'il~>...>...>-:....>...>,;w....>..:.....:....::..~~~»-'
SWIVEL JOINTS
ISOLATOR EQUIPMENT
e.G.
(a) , /
.
L1NKA GE---..
SWIVE L JOINT
"""'l
..
:
ISOLATOR
:• i"~
SWIVEL JOINT
V
:
.
....J .....
BASE FIXED ISOLA TOR
HARD STRUCTURE
C.G. "
STRUCTURAL EXTENSION
EQUIPMENT
....
.
FLOOR ~-..:i"--
PLATFORM
.-
.
,...
( b)
Figure 6-67 Overhead pendulum shock isolation systems using platforms 6·182
CENTER OF GRAVITY MOUNT PREVENTS ROCKING UNDER SHOCK
VERTICAL SHOCK • • MOUNT r
HORIZONTAL SHOCK MOUNT I'
"
'"
,~;"Figure
,.
"
,
6-68' Helical ccmp re s s i on spr i nq mounts t
6-183
"
BALL JOINTS
SUPPORT STRUT AND DAMPER ASSEMBLY.
PLATFORM
TORSION LEVER
TORSION SPRING
SIDE
VIEW
TORSION SPRING
FRONT VIEW
Figure 6-69 Typical torsion spring 'shock isolation system
6-184
v, V V
J
A/
SURGE TANK
Ps.V s
,
r
CYLINDER CHAMBER PC I Vc
H
I
x},'
..
, .'
.'
,,
,
...•
~ .Pi:'V2
';I /SURGE TANKL
/
~1,VS
r
,..
"
,
,
....
I
A.) .
..'1 AI./'
V
V
Figure 6-70
SURGE TANKr I PI,V S
~
PI ,VI
·Sc~~maticCof singlea~Jdouble actirig'pneumatic cylinders 6-185
"
SEAL (TYP.)
FLUID
FLUID .
PISTON ROD
~,
..,,
.
,
CYLINDER, ,
,
.Io )
CYLINDER-':
.... ..
,
"",
PISTON ROD
I
,
"
COMPRESSIQN
(b) TENSION
"
, .
~."
-.
.~ . >.:~~
."'
,
6-186
:
.. ,
Q
I' ty~ I
I ~ 0
(Q) SERIES
(b) PARALLEL
.I
(e) PARALLEL-SERIES
Figure 6-72 Belleville springs
6-187
TK 5-1300/NAVFAC P-397/AFR 88-22
APPENDIX 6A ILLUSTRATIVE EXAMPLES
TM 5-1300/NAVFAC.P-397/AFR'88 c22
Problem 6A-I
Masonry Wall Design
Problem: Design a reinforced masonry wall for an exterior blast load. Procedure: ;;
Step 1. .
Establish design parameters:'
'L'
\ ~a.
Pressure-time loading.
r;
b.
Structural configuration including geometry, support,condi-
,
tions and type of wall (i.e. reusable or non-reusable). \
"
'"
Step 2.
. Select masonry unit s Lzer and the size ",md type of reinforcement. Assume the distance between the tension and compression reinforcement (d c)' Also determine the static design stresses. for the masonry unit and the reinforcement (Section 6,8.1) . . The average yield stress of reinforcement is increased 10 percent.
Step 3.
Calculate the dynamic design·stress of the reinforcement, using the static stresses from Step 2 and· the dynamic increase factors from Section 6-8 ..2·,(For joint' reinforced· masonry construction the
~r
j"
compressive strength of the concrete may be ignored.
e.
6-8.3) ..
See Section
-,
Step 4.
leFor the' size and type'of reinforcement selected in Step 2, calculate the area of reinforcement per unit width of the·wall. Using the value of d c from Step 2, the dynamic design strength ; from Step 3 and. the area of reinforcement from above, ·determine ! the ultimate. moment capacity of the wall (Eq. 6-2) .
Step 5.
. , Determine the ultimate resistance of .the wall using the ultimate ,moment 'capacity of Step 4 and the equations of ~able 3-1 (if the wall is'a one~way spanning element) or Table 3-2 or 3-3 (if the wall spans two directions).
Step 6.
From Table 6-3. find the moment of inertia of the net section In' Calculate .t.he moment ·of inertia" of the cracked section I c' using Equation 6-7 and the value'of d c 'from Step 2.' Determine the average moment of inertia, using the values of I g and I c from c.above and Equation 6"6." ' ..
Step 7'.
Calculate the modulus of elasti'city of, the masonry unit Em from -, Equation 6 -1 and the masonry unit strength of Step" 2. ,', t
Step 8,
If the wall spans one direction only, use the average moment of inertia from Step 6, the modulus of elasticity from Step' 7 and the equations from Table 3-8, to find the equivalent elastic stiffness. For a two-way spanning wall, use the methods of Section 313 to calculate the equivalent elastic stiffness.
GA-l •.
TK 5-l300/NAVFACP-397/AFR 88-22
Step
9.
Determine the equivalent elastic. deflection using the ultimate resistance (Step 5), the equivalent stiffness (Step 8) and Equation 3-36. '..
Step 10.
Find the load-mass factor KLM, for the elastic, elasto-plastic and plastic ranges from Table 3-12 or 3-13. Average the values'of KLM for the elastic and elasto-plastic ranges .. Average· that value with the KLM for the plastic range to find the value of KLM to be used for the element. Calculate the unit mass of the masonry unit and multiply the unit mass by KLM to obtain the effective unit mass of the·wall.
Step 11.
Calculate the natural period of vibration from Equation 3-60, the effective mass from Step 10 and the equivalen~ stiffness from Step 8.
Step' 12.
Determine response .chart parameters: ., a. Peak dynamic loading P (Step 1). b. Ultimate resistance r u(Step.5). c. Duration of load'T· (Step 1)'. d. Natural period of vibration ~N (Step 11).
! '}. ,
Calculate the ratio of ultimate resistance to peak dynamic loading (rufP) and duration of load to natural period (T/T N). Using these r at Lo's, and the appropriate figures (Figs.-3-54 through 3-266) determine the ductility ratio .x,./X E.· .' Step 13.
Compute the maximum deflection x,. by multiplying. the,.ductility ratio by the elastic· deflection of Step 9." From 'Table 3-5. (for a one-way spanning wall. For a two-way spanning wall use Table 3-6) and the value of x,. , find the. maximum support rotation. ,Find the' maximum r ot at Lon. permitted from Table. 6-2 and compare with the rotation calculated above. If rotation·is larger than that permitted, repeat steps 1 through 13.
Step 14.
Determine the ultimate shear'stress ·at dc/2 from the support. Using Equation 6-4 or 6-5, compute the area of shear reinforcement required for the above shear stress.
Step 15.
Using x,./XE and T/T N (both values from step 12) find. the rebound resistance from Figure 3-268. With rebound resistance and the equations of Table 3-9, 3-10 or 3-11, 'calculate the rebound shear. Then compute the area of anchor reinforcing required using the rebound shear from above and the dynamic strength of 'the rein'forcement from step 3. .
.'
6A-2
TK 5-l300/NAVFAC P-397/AFR '88-22
Example GA-l Required:
Masonry Wall Design
Design a joint reinforced masonry wall supported by steel columns for an exterior blast load.;.'
Solution: Step 1.'
Given:
:
.
,
Pressur~"time
a.
I
loading (Fig. GA-l) .
,
Wa!l spans in 'one direction only, i~' rigidl~ s~pported at both ends, and has a clear span between ,colUmns of 150 inches. 'The wall is part of a reusable ~tructure.
b.
.".
Step 2.
Use 12" ,wide hollow concrete masonry units and ladder type reinforcing with No. 8 Gage side rods and No. 9 Gage cross rods 16" o .c . Ass.:mie: dc - 10" for this ,tYP.e o'f reinforcing. For hollow concset~ m~sonry units the static c~mpressive,stress (f'm) is 1350 P~~'~';,·F0F. joint~~einforcement the yield stress of '70,000 psi is 'incr'eased 10 percent to.'17, 000 psi.' ,,'
Step 3.
Calculate . dynamic design stresses ! • '.+ ,
So"
-
.'..
-i
'of
the reinforcement:
Dynamic 'Inc!~~se factors ~.' ' ','v
.r
,
b.
"
flex~r~: 1.17'
shear:
r
....
",
1. 00,,;
j
, .'
Dynamic strengths ftexure f dy shear
Step 4.
"
1.17 X 77;000
f dy -,l.O'X 77,000
90,090 psi 77, 000 psi
Determine the ultimate moment capacity of the wall. a.
Calculate the area of, reinforcement per unit width of the wall. Use one layer of reinforcement between every masonry unit joint, therefore 8 inches o.c. As - 0.0206/8 - 0.0026 in 2/in
b.
Ultimate moment capacity (Eq. 6-2).
,Mu -
As f dy dc - 0.0026 X 90,090 X 10 - 2342 in-lbs/in
6A-3
TN 5-1300/NAVFAC:P-397/AFR 88-22
,.,"
,
,
'-"1 1 .'~ ,
Figure 6A-l,.•
"
"
'
"
100,0
TIME
"
(MS)
',r 1
i
i
.if; ..
(0)
,
-, '
,
'
'
. . ' ,1, COMMERCIALL Y, AVAILABLE ' REBOUND ANCHOR '..'vIELDED TO .COLUMN EACH COURSE,'
SUPPORT LADDER ,TYPE REINFORCING
" "~ADD'L' SHEAR REINF,
FILL '\JITHCO'NCRETE
.L-~~~~~~~~rr=;~~~~CO~LU!:1N
'.t·
CMU
,. (e) t .
,
-r .
r
"
'I
\'
,
.'-; , '"
.
e-r
6A-4 ,. "
,
'
TH 5-1300/NAVFAC P-397/AFB. 88-22 Step 5.
Calculate.ultimate resistance (Table 3·1)
- 8(2342 +, 2342)
- 1.66 psi
Step 6.
Find the average moment of inertia. a.
Moment of inertia of net section (Table 6·3). For a 12 inch unit
b.
Moment of inertia of cracked section (Eq. 6-7). I c - 0.005 X' d c 3
~\
- 0.005 X 10 3)( 12
- 5.0
c.
in4/in
.
:. Average moment of inertia (Eq. 6·6).
- (83.3 + 5.0) / 2 - 44.2 in4/in Step 7.
Compute modulus of elasticity of the masonry unit (Eq. 6·1). ~
- 1000 f'm - 1000 X 1350 - 1.35 X 10 6 psi
Step 8.
Determine the equivalent elastic stiffness (Table 3-8).
KE - (307 ~Ia) / L4 (307 X 1.35 X 10 6 X 44.2) / 150 4 - 36.19 psi/in
TN 5-1300/NAVFAC, P-397/AFR;ft8-22' Step 9.
Calculate the equivalent elastic 'deflection. (Eq. 3-36).'
Step 10.
0.046 'in -,
"e
Calculate the effective unit mass of the wall. a.
Find the average load-mass factor,K LM (Table 3-12) Elastic KLM - 0.77 Elasto-plastic KLM- 0.78 I .
Plastic KLM
-
~
,
0.66 - , '1'
For limited plastic deflections
.
. ' :.
,
...
,
'
KLM - [(0.77 + 0.78)/2 + 0.66]/2 - 0.72 •
b.
:
I.
Determine the unit mass of wall. Using Table 6-1 :..
[(16 X 12) -2(4.25 X 9)J
w- - - - - - - -_ _
X 150 pcf
0.627 psi
16
-e
\.1:
w
c.
0.627 m- - ---...,.-------.,. - 1622.7 psi-ms 2 / in g 32.2 X 12 x 10- 6 , Calculate effective unit mass.
-0.72 X 1622.7 1168.3 psi-ms 2/in Step 11.
, Determine the natural period of vibration (Eq. 3-60).
.,
TN - 2"
(me / KE) 1/
2
- 2" (1168.3/36.19)1/2 - 35.7 ms "
'-
".
'
6A-6 ~.:
TK 5-l300/NAvFA:C P-397;1AFR 88-22 Step 12.
.
Determine the response-' of the wa11.a.
Calculate design chart parameter~; T/T N - 100.0/35.7 - 2.80 ru/P - 1.66/2.0 - 0.83
b.
From Figure 3-54 ,>,
Step 13.
•• '
Check support rotation. a.
Compute maximum deflection. Xm .." I£XE - 8.0 X 0.046
,
- 0.368 in b.
Calculate support rotation (Table 3-5) tan- l(2Xm/L)
e
-e
-
tan- l(2 X 0.368/150) 0.28·
·.. r,'_
.c
, c. ". 'Coinijare' rotation 'with c rLt e r'La..
.-
From Table 6-2
e Step 14.
,.
0.5·" ·>0.28·
O. K:
Design shear reinforcement. a.
Calculate shear force d c/2 from support.
-
ru (L-d c) 2 1. 66 (150-10) 2
- 116.2 lb/in
6A-7 :
-,
.
TN 5-l300/NAVFAC P-397/APa 88·22
b.
Find net area of section from Table 6-1, An
- 2b X Face thickness/b, -
c.
2 X 8 X 1. 5/8
Compute ultimate shear stress (Eq ; 6-3) V
u
- Vu/An
','
116.2
--3.0 d.
38.73 psi
"
•
,.
.Ju
t,' ,
Find area of shear reinforcement required (Eq. '6-4) Assume s - 4". ,
,
, tfy
38.73 X 8 X 4 0.85 X ,77,.000 - 0.0189 in 2
Use No. 8 Gage
Use 3 legs of No. 8 gage wire at Step 15.
~"
Design rebound anchor ties. a.
Find rebound resistance ,force. T/T N -
2.8
Xm/XE - 8.0 From Figure 3-268 r-/r u - 0.47 r
- 0.47 X 1.66 - 0.78 psi
:.
6A-8 .
o.c. between each cross rod.
TK'5"1300!NAVFAC'P-397/AFR!88-22 Calc~late rebo~nd shear at support:
b.
I ,
Co
r ' L.
Vr -
~
.'r,'r
J'.
:.! .
"
2
.'
0.78 X 150
,
.;'
- 58.5 lb/in 2
c.
Required' area of anchor reinforcing A~
'.
rV
58;5
f dy
90,090
0.00065 ' in 2/in l" ."", y
Use one anchor be tweerr-eve ry masonry unit joint, therefore 8" o .c.
....
0.00065 X 8,
"
, .
s.
.....
e...:
F
0.0052 in 2/anchor
-
.,
,"
"'.1
.~
Use
•:-
-.'i::
.'
'.;'~
diameter triangular ties,
3/l6~
.'
.,'
.•..
... . ~
'}
.. "
...
-
" •
,
:.'
i." _
.,~.
'.
,"
1..
'.
~ "
"
..:
t . -~
,
,
.
'IL"
.' .~,~?~~'
" I ,"
'i"
•• ... 'i
.. ,
~
5 ,"
.': ,
..
'.
r
6A-9
TIl 5-l300/NAVFAC'P-397/AFR':88-22,
Problem 6A-2 Design of Prestressed'Recast Element Problem:
Design a prestressed recast element for a given blast load.
Procedure:
Step 1.
Establish design parameters: a.
Pressure-time loading.
b.
Material strength.
c.
Span
d.
Static loads.
. ,e.
Step 2.
.,
J:
le~gth.
Deflection crit;eria;
Select a standard recast section from the'PCI Design Handbook or Also sele~t a ~tandard strand ~attern.
manu~a~~urer's catalogs ..
Step'3.
Determine dynamic increase factors for concrete and reinforcement from paragraph 6-12. Increase the static des'ign stress of the reinforcing bars and 'welded wire fabric 10 percent for the average yield stress. Calculate the dynamic design design stresses using the above DIF and the static stresses.
Step 4.
From Chapter 4 and a unit weight of concrete equal to 150 pcf, calculate the modulus of elasticity for concrete. With the above modulus for concrete and those for reiriforcing bars and prestressing tendons, calculate the modular ratio.
NOTE: If the section has a flange (e.g. a single or double tee section) it 'must be designed according to the principles of Chapter 4. The flange is.a one-way, non-prestressed slab spanning between webs. The·critical section is usually a cantilever. Step 5.
Calculate the properties of the section: 'a ,
Gross area.
b.
Gross moment of inertia.
c.
Unit weight.
d.
Distance from extreme compression fiber to the centroid of the prestressing tendon, ~.
e.
Area of prestressing tendons.
f.
Prestressed reinforcement ratio (Eq. 6-23).
6A-IO
TM 5 c1300/NAVFAC P-397/AFR 88022
Step 6.
Given the type-of prestressing tendon, determine Y p.
Using the
dynamic concrete stress from step 3, calculate B Using Equation 6-27, ,6-28 or 6-29 and the values of yp and from above, calculate the average stress in the prestressing tendon. .
Bi
Step 7.:
With the reinforcement ratio from step 5e, the average stress in the prestressing tendon and the value of Bl from· step 6, and the dynamic strength of materials from step 3, check that· 'reinforcement ratios are less than the maximum permitted by Equation 6-30 or 6-31.
Step 8.
,Using the area of reinforcement and the:val~e of,dpc from step 5, the dynamic stresses of step 3, and the average stress in the 'prestressing tendon'from step 6, calculate the,moment capacity of the element (eqs .. 6-20 and 6-21).
\.
Step 9.
}
:t,
',With the equatiLons co f I'ab Le 3-1 ,and the moment. capacity o'f step 8, i
calculate the ultimate unit resistance.
As recast buildings are
only subject to'low blast pressures, the st~tic loads become ' significant'. To determine ,the resistance available' to resist the blast load, subtract ,the static dead and live loads from the ultimate 'unit r e s i.s t ance ;, Step 10.
Calculate the moment of inertia of,the cracked section Ie' using Equation 6-33. the modular ratio from step 4 and the area of '... prestressing tendons, the value of d p and the prestressed r e i.nforcement ratio from step 5. Using this value of I c and,the gross moment ofinertia..of'step' 5b, find the average moment of inertia . from' Equation 6 - 32., .~.
Step 11.
Using the equations of Table 3-'8, the modulus of elasticity for concrete from step 4 and the average moment of inertia from step
10, calculate the'elastic stiffness of the section .
.-' Step 12.
Determine the lo~d-mass' factorK LM, in the elastic range for the 'appropriate Loadf.ng.rcond t tion from Table, 3-12. Also, calculate the unit mass of the section and multiply it by KLM to obtain the effective unit mass of the' 'element. '
Step 13.
With Equation 3-60,- the 'effective mass from step 12 and the e~astic
stiffness from step II, calculate the natural period of
, vibration TN' of the section.' Step 14.
Determine the response chart parameters:
a.
Duration of load T (step 1).
b.
Natural period of vibration TN (step 13).
~J
... !
( _
"
Calculate the ratio T/TN.andusing this ratio, determine'the dynamic load factor DLF from Figures ·3-49 through 3-53. Section must remain elastic, and hence the, actual resistance obtained.by the element (which is equal to the peak dynamic load from step 1 multipl~ed by the DLF) must be less than the resistance available. 6A-11
TIl 5-1300/NAVFAC P-397/AFR88-22 If the section does not r~main' elasti2 steps' i,through'~4'must be ,repeated. NOTE: For bilinear loads, calculate the ratio of the'peak dynamic load'p to the resistance available. Using P/r u and the value of T/T N calculated above, ente~ the appropriate response chart ,(Figs. 3-6~ .tihr ough 3-266) and f Lndvche ductility .r at Lo ~/XE' For the section to remain elastic the ductility ratio must be less than or equal to one.
,Step 15.
Calculate the deflection" of the element ~ by "adding' the dead and ,live loads ,(step '1), and the resistanc~ obtained by' the structure .. (step 14) and dividing by the elastic stiffness (step 11). Using the deflection and the' equations of. Table 3-5. determine the support rotation. Compare the. rotation with deflection 'criteria of step Ie. If comparison is satisfactory continue with step 16. If comparison is .no c satisfactory, repeat steps ,I through 15; ,
Step 16.
Calculate the elastic deflection Xm' ,from Equation 3-36, using the ultimate resistance of step 9 and the elastic stiffness .of step 11: Then calculate the ductility ratio ~/XE.using the'value of ~ from step 15. With the ductility ratio and the ratio:,T/T N from step 14, enter Figure 3-268 and find the percentage of rebound. Extrapolate if necessary. _
Step 17.
Find the required rebound resistance by multiplying the ultimate unit resistance by the ratio from step 16 and ,subtracting- the dead load. In no case should che required rebound resistance be less than half the resistance ava t LabLe during 'the loading phas e , With equations of Table 3-1 and the required rebound resistance, find the required -rebound moment capacity.
i
Step 18.
.. ;
Step 19.
·1
.'
Calculate an approximate value of d- and the.'amount of concrete strength available for rebound (Eq. 6-35). Assume a depth of the equivalent rectangular stress block and, by a trial and error method, using Equation 6-34 find the area of rebound reinforcement required. Check 'that,theamount'of 'reinforcement does 'not exceed the maximum permitted by Equation 6-36 or 6-37. Calculate the shear stress at ~ away from the support. Also '. calculate the allowable shear'stress on an unreinforced web. using the dynamic design stress, (step 16). Design the shear'reinforcement.
':
:
:.:.
Step 20.
Calculate the shear at the support Vd from the equations of Table 3-9 and the value of the total'load. Calculate the maximum allowable direct shear using the dynamic concrete strength. Compare the allowable shear with Yd' size of-the section must be increased.
If Vd is greater than the
,-
Step 21.
Check if ,section 'is adequate for service loads using the PCI Design Handbook'and·the ACI Code.
6A-12
• ?' .
"
'
TH 5-1300/NAVFAC P-31l7/AFB."88~22
Example 6A-2
Required:
Design of a Prestressed B.eeast B.oo! Pariel
Desigri a prestressed recast roof subject to an overhead blast load. Use a double tee section.
Solution: Step 1.
Given: a.
Pressure-time loading (Fig 6A-2).,
b.
Material strengths "
Concrete:
f' c - ' 5000 psi";
'Prestressing Steel:· f pu, -,270,000 psi fpy/f pu - 0.85 Welded Wire Fab,dc:"
f y .. 65,.000 P'sJ
,Reinforcing Bars c.
f y - 60,000 psi
Spa~
d.
length is 40 ft. - 480 in. , I " Live ,load iS,15 psf
e.
Maximum ductility ratio < 1.0
,
;
Maximum support rotation < 2· Step 2.
:
Select a double tee ,section and strand pattern. •
Try 8DT24 (Fig. 6A·2) Section properties from PCI Design Handbook: A - 401 in2 I
Yt - 6.85 in
- 20,,985 in2
..
e' - ~'-
\,"
'
:
--
W - 418 lb/ft. ~
Try strand pattern 48~S."two 1/2 incn diameter ;traight'str'ands in each tee.
"
'. , .' !
Area of each strand,~' 0
\S3
in 2.
e - 14.15 in
6A-13
TM S.1300/NAVFAC P·397/AFR 88·22
Figure 6A-2
"Vi
c,
1.1 ps:
v
TIME
(1'15)
43,9
(0.)
PRESSURE-TIME LOADING
8'-0' I
2' J
2 4'
I 5;: I
\oIELDED \oIIRE"\ FABRIC
r
*!
I
f-·_· e
.
~
""
. .
\--.- !-------- 1-- ... ~.\. .~
.
..'" tL
.tI.
,.
."
CENTROID OF PRESTRESSING STEEL
j
..
"
..
•
"
~
:- f-
I.-
4'-0#
DOUBLE TEE SECTION (b)
CENTROID OF SECTION
3 3//
TH 5-1300/NAVFAC P-397/AFR 88"22 Step 3.
Determine design stresses. a.
Dynamic increase factors: Concrete flexure:
1.19' ' r'
Concrete diagonal tension:
1.00
'Concrete direct shear:
1.10
J..
'1.00 '
Prestressing Steel: ·Welding Wire' Fabric:
1.10
-!
'):...
'J
~
1.17 1;'_
! .'_
•
Reinforcing Steel shear: b.
'J
:. I" '• .
. ~ ~ ,~
Reinforcing Steel flexure: l
....
1.00
Dynamic strengths. Concrete flexure:
f'dc - 1.19 X 5,000 - 5950 psi
- diagonal tension: f'dc - 1;0 oX 5,000
~'5000
psi
f'dc- 1.1 ~c5,OOO -'5500 1psi
• direct shear: Prestressing Steel:
f dy - 1.0 X 270,000 - 270,000 psi
Welded Wire Fabric:
f dy - L10 x.i :10 X 65,000 - 78,650 psi
Reinforcing Bars
Step 4.
- flexure:
f dy'- 1.17 X: 1.10 X' 60,000 '- 77,220 psi
shear:
f dy - 1. O·:X 1.10 X .60,000 - 66,000 psi
Calculate modulus of elasticity,'and modular ratio'. a.
Concrete _ 33w 1.5 (f' )1/2 c
c
_ 33(150)1.5(5000)1/2 i
4.29 X 10 6 psi
6A-15
.,
TH
5·1300~AVFAC
b.
P-397/AFR 88-22. Steel
,.
Es - 29.0 X 10 6 psi n -
Es
29.0 X 10 6
Ec
4.29 X 10 6
---
- 6.76
NOTE: The flange is designed in accordance with the. principles of Chapter 4. Considering the flange a one-way non-prestressed slab, the cantilever portion was found to be critical. In order to remain elastic the flange thickness must be increased to 3 inches. The reinforcement is. two layers of we~ded wire fabric; 6 X 6 W1.4 X W2.0 in the top and 6 X 6 - W1.4 X W1.4 in the bottom. "
Step 5.
;
.'
Calculate the section properties of the double tee with a 3 inch flange .. a.
new A - 497 in 2
b.
new Yt - 6.43 in new I g - 25180 in4
X 150,pcf/12~
c.
w,- 497
d.
~ -J.85 +14'.15 - 22.0 in
- 43.1 1b/in
,.
Aps - 4 X 0.153 - 0.612 in 2 f.
. Prestressed reinforcement ratio (Eq. 6-23) Pp -
Step 6.
~s/b~
- 0.612/(96 X 22.0) - 0.000290
Determine the average stress in the prestressing tendon. 0.85
so
0.40
a;
f py /f pu -
b.
8 1 - 0.85 - 0.05(5950-4000)/1000 _ 0.7525
c.
Average stress (Eq. 6-27)
y -
P
f,. -f,+ 0.40 (0.000290 X 270,000) - 270,000 [ 1 (.7525) (5950) fps
- 268,111 psi
6A-16.
]
TK 5-1300/NAVFAC p C397/AFR·88-22 Step 7.
Check maximum reinforcement ratio (Eq. 6-30), ppfps/f"dc~:
0;.36 Bl c
ppfps /f'dc
0.0131
0.000290 X 268,111/5950., . ~:0·.2709~> 0.0131
Step 8.
O.K.
Calculate moment capacity of beam. From Equation 6-21 ~sfps
0.34 in
0.85f'dC b.
0;.85 X 5950 X
c - a/B l ... 0.3{>/0.. 7525
-.
96!~
... :,0,.45 < 3 ..0 in thick flange
Hence the neutral axis is within the flange and 'the section can be analyzed as a rectangular section. If the neutral axis had extended into the web, a strain compatibility analysis would be required. . J
From Equation 6-20· ;
r ,.'
l\, - ~sfps(~ - a/2) - 0.612 X 268.111 (22.0 - 0.34/2) l\, - 3582 k- in
, Step 9.
Find the resistance available to resist blast load. roo
a.
Find the ultimate resistance (Table 3-1) . .J
r u - 8l\,/L 2 - 8 X 3582/480 2
- 0.124 k/in
- 124 lbs/in
b.
Resistance available for blast load '
.i."-;..
..
• "
.., I"
ravail - r u - DL - LL
' - 124 - 43,1 , (15 psf X 8 ft/12)
v....
"
vr
~:
- 124 - 43.1.- 10.0 - 70.9 lb/in :
,~
'.
"
TM'5-1300/NAVFAC Step 10,
~-397/AFR,
88-22
Determine the average moment of inertia, a,
Moment of inertia of cracked section (Eq, I c - ~s~2 [lC(p )1/2 - 6,76 X 0,612(22,0)2
b,
'6~33);
i [1-(0,00029)1/2]
1970 in4
Average moment of inertia '(Eq" 6-32) Ia -
{Ig + I c)/2
,'
,
(25180 + 1970)/2-'- 13575 in4 Step 11,
Using equations of Table 3-8, ,'calculate the e1asi:'ic·stiffness. 384 X 4,29 X 10 6 X 13575 ',' " 5
X
480 4
84,25 1b/in/in Step 12,
Calculate the effective mass a,
Load-mass factor (Table 3-12) In the elastic range KU{ -
b.
0.78
Unit mass m - wig - 43.1/(32.2 X 12) - 0.1116 1b_s 2/in2 - 11.16 X 10 4 1b-ms 2/in2
c.
Effective mass me -
KU{
m
0.78 X 11.16 X 10 4 >, ",
- 8.70 X 10 4 1b-ms 2/in2 Step 13,
Calculate the natural period of vibration (Eq. 3-60). TN - 2" (me / KE)1/2 - 2" (8.70 X 10 4/84.25)1/2
6A-18 ,
-, 201. 9 ms
TM 5-l300/NAVFAC P~397/AFR 88~22
Step 14.
Determine response of beam. T - duration of load· - 43.9 ms (step'lal TN - natural period - 201.9 ms (step 13)
.:r,
T/TN - 43.9/201.9 - 0.217 From Figure 3-49, DLF - 0.65 Actual resistance obtained r - DLF X P
S
rav~il' -
P - peak dynamic load 1.1 psi -(step la)
• 0;;.
r - 0.65-X '(1.1 X'96) . - 68.4 lb/in < 70.9 lb/in·'(raval.l' Step 15.
step 9)
Check rotation. a.
Total load on beam - (DLF X p) + DL + LL 68.4 + 43.1 + 10.0 - 121.5 lb/in
b.
Maximum deflection: ~
- (121.5 lb/in)/K E
.
121.5/84.25
-
- 1.44 in c.
Support rotation (Table 3- 5).....
e-
tan-l(2~/L) tan- l(2 X 1.44/480)
- 0.34° Step 16.
< 2°
O.K.
Calculate percent of rebound; a.'
Calculate. elastic deflection (Eq. 3-36) XE - ru/KE
." ,.124/84.25 - 1.47 in
6A:-19
O.K.
TH 88-22 . . 5-1300/NAVFAC P-397/AFR . . b.
Calculate ductility ratio. j.I
,
• ,I'
x,./XE
-
- 1.44/1.47 .: 0.98 c.
...
I'
Find percent of rebound from Figure 3-268. x,./XE - 0.98 and.T/T N - .0.216
_.
j,
~,
..
1
/0.
r-/r - 1.0 so. 100 percent rebound, Step 17.
Determine required rebound moment capacity. a.
.Required rebound resistance
.
I'
'
I.'
- r- - DL >
rreq -
6~.4
.. " ,-
.
ravai1/2
- 43.1
- 25.3 1b/in 35.4 > 25.3 1b/in
ravai1/2 - 70.9/2
Use 35.4 1b/in or considering a single tee - 17.7 1b/in b.
"
Required moment capacity (Table 3.-1).
Mu- -
rreq-
L2/8
- 17.7 X 480 2/8
MuStep 18.
~ 509,760 in-1b/stem
Determine rebound reinforcement. a.
Approximate value of
'.
- 25 - 0.625. - .0.135 - 0.375 '-0.5/2 - 23.62 in b.
Rebound concrete strength f - 0.47f'dc
0.47 X 5950.
6A-20 .
2796.5 psi
c.
Requ~red
reboun4 reinforcement
Assume a - 2.0 in As- - Mu-/[fdy(d- - a/2)
(Eq. 6-34)
- 509,760/[77;220(23.62 - 2.0/2»)
- 0.29 in 2
a -
(0.47f'dc)b 0.29 X 77,220 - 2.1 in
~
O.K.
2.0 in
2796.5 X 3.75 Use 2 No.4 bars in each stem As d.
- 0.40 in 2
Check maximum reinforcement (Eq. ,6-36).' 0.47f'dCB1.[ f dy
87,000 - 0..378. n f'dC]
. bd"
. (87.000 - 0.378nf'dc+ fdy)
87,000-0.378 X6.76 2796.5 XO.7525.,[ _ _-'--X -5950 ---77,220
(87.000-0.378.X 6.76 X 5950 + 77,220)
X (3.75 X 23.62) - 1.16 in 2 > 0.40 1n 2 Step 19.
O.K.
Design the shear reinforcement. Calculate shear at distance Vu
':"
~
from support.
r(L/2 ~ ~)/bw~
r - 121.5 lblin (total load on beam, step. '15a) Vu
b.
- 121.5 (480/2 - 22.0)/(2 X 4.75 X 22.0)
Maximum allowable shear stress. 10 (f c,)1/2
- 10 X (5000)1/ 2
- 707 psi > 127 psi
6A-21
O.K. '
- 127 psi
]
TN 5-1300/NAVFAC.P-397/AFR 88-22 c.
,
Allowable shear stress on unreinforced web. V
c - 1.9(f· c)1/2 + 2500p p S ·2.28(f· c)1/2 0.612 - 1.9(5000)1/2 + 2500
142 psi '(2 X 4.75 X 22)
2.28(f· c)1/2 - 2.28(5000)1/2
O.K.
161 psi> 142 psi d.
Excess shear. l
'.'
-15 psi use "c
so,
e.
Shear reinforcement. Assume i/3 closed ties
.j
Ay - 2 X 0.11 - 0.22 in 2
.22 X 0.85 X 66,000 \, Ss -
142 ·X4.75
,.
- 18.3 in f.
Check maximum spacing and minimum required reinforcement. s
S
~/2
- 22.0/2 - 11.0 in.
0.0015bs s
Ay S
- 0.0015 X 4.75 X 11.0
- 0.08 in 2
> 0.22 in 2
Shear reinforcement is thus #3 ties at 11 inches in both " ~tems. '. Step 20.
. Check direct shear. a.
Calculate shear at the support· (Table 3'9) Vd
-
rL/2
- 121. 5 X 480/2 - 29160 1bs
6A-22
TH 5-l300/NAVFAC"P-397/AFR 88-22 :. . " '.', . ~
,
Calculate allol"able direct shear.
b.
Vd S 0.18 f' c bd ~
- 0.18 X 5,500 X (2 X 4.75) X 22.0 - 206,910 lbs > 29,160 lbs Step 21.
,...
O.K. '.'
Check section for conventional loads.
This section as designed for blast loads is shoWn in Figure 6A-3. Using the PCl Design Handbook and the latest ACl code, the section must be checked to make sure it is adequate for service loads.
..
figure 6A-3
'vI'vIF 6x6 . . 'vilA x 'vI2,O
2 -
IIA ',i-.
3'
'vI'vIF ' 6x6 'vilA x 'vilA ~
, 113
..
..
0 @lJ'
22'
,.112', PRESTRESSING STRANDS
6A-23
TN 5-1300/NAVPAC P-397/AFR S8-22 Problem6A-3 Problem:
De81sn of \lindows
Determine the 'minimum thickness of glazing to resist a given blast load, and the design loads for' the framing.
Procedure: Step 1.
Establish design parameters: a.
Pressure-time loading.
b.
Dimensions of pane(s);
c.
Type of glazing.
Step 2.
Calculate aspect ratio of pane. '
Step 3.
With the parameters of Steps 1 and 2, enter Figures 6-28 to 6-42 to determine which one applies. Using the peak pressure of the dynamic load, its duration and the dimensions of the pane, determine the minimum,required glazing thickness.
NOTE: If given window geometry differs from chart parameters, .interpo1ation as outlined in Section, 6-28.4 may be required. Step 4.
Find the static ultimate resista~ce r u of the glazing from Table 6-6 for the given aspect ratio, the short dimension and the thickness of the glazing' (interpolate if required).
Step 5.
From Table 6-9 and the aspect ratio determine the design coefficients CR, Cx and Cy for the windo~ frame loading. With these coefficients, the dimensions of the pane from step Ib, the static ultimate resistance of step 4 and Equation 6'-55," calculate the uplift force in each corner., Then using Equations 6-53 and 6-54 calculate the design loads along the long 'and short spans of the pane. Finally from Table 6-11, determine the fundamental period, TN' and check for rebound requirements in accordance with Section 6·30.4
TH 5-1300/NAVFAC',P-397/AFR' 88~22
.
Example 6A-3 . Des ign of Windows'"
-'
..
Find the minimum glazing thickness and. the design loads on,the frame of a non-openab1e window consisting of four equal size panes • i" of glass.
Required:
Solution:
.
Given:
Step 1.
a.
·Pressure-time.1oading (Fig. 6A-4)
IJ.
! - ..
1, :,
b.
Each pane is 37.5 inches long,by 30'inches high ...
c.
The glazing is heat-treated tempered glass meeting Federal Specification DO-G-1403B and ANSI Z97.1-1984.
Figure 6A-4 , ,,.,,
1"',;'
J
j
,.
"V1 0.
'-'
w
5
Q::
:::>
V') V')
w
Ck:
c, .,
500 TIME Step 2.
'.'
Calculate. aspect ratio.
(f'\s)
'.
'
alb - 37.5/30 - 1.25 " "
l"
,.
.,
6A-25' .... '
...
TH 5-l300/NAVFAC P-397/AFR '88-22 . , Step 3.
,Determine minimum' glazing·thickness. a.
.. For .
..
alb - 1. 25 and tempered glass USe 'Figure 6- 33 .
'
b. Step 4.
From Figure 6-33, t - 3/4 inch.
Find static ultimate resistance r u• (Table 6-6). ru
- 24.6 psi
The window' frame must be designed 'to safely~ support', without undue deflections, a static uniform load of 24.6 psi applied normal to both the glazing and the .exposedframe members. Step 5.
Compute design loads on the window frame. Note: a.
The exposed surface width of the frame is 2 inches. Determine design coefficients from Table 6-9, interpolating for alb - 1.25. CR - 0.077
Cx - 0.545 Cy - 0.543 b.
Calculate the unit shear aiong the-iong span of the frame (Eq. 6-53).
0.545 X 24.6 X 30 sin ("x/37.5) + 24.6*2 - 402 sin("x/372.5) + 49.2 lb/in c.
Calculate the unit shear along the short span of the frame (Eq. 6-54).
Vy - Cyrub sin("y/b) - 0.543 X 24.6 X 30 sin("y/30) + 24.6*2 - 400 sin("y/l8) +49.2 lb/in
6A-26 .
TH 5-1300/NAVFAC P-397/AFR 88"22 d.
Calculate the uplift force at the corners of the panes (Eq. 6-55).
R - -CRr ub 2 0.077 X 24.6 X 30 2 1705 lbs. Distribution of the design load of the pane on the frame is sbowndn Figure 6-52. " e.
From Table 6-U, TN - 8.·74 msec, T/T N - 1000/8.74 -- U4 Since',T/T& ,- 'U4
~
'10, design for rebound is not necessary.
",
..
"
'
. ., .
, 6A-27
'
TIl S-1300/NAVFAC.P-397/AFR88722
Problem 6A-4.Design of Shock I!,olation Sy:stem. Problem:
Design an overhead pendulum shock isolation ,system using a platform for a given loading.
Step 1.
Establi~h
a.
.' .
r'
c.
Structural configuration. Magnitude and location of loads on platform. Shock spectra fo~ h~ri~ontal and:vertic~~·motion..
d.
Maximum. allowable motion.
b.
Step 2.
design parameters:
.' iCompute membe r sizes of ·platform.
-1;'
'.'
1,:..
,.• ,' '•
,.l.
Step 3.
Compute· cencer of.. gravity of the loadsr(live',anddead) on platform.
Step 4.
Compute the elastic center of the spring supporting system.
Step 5.
Determine the required weight and location of ballast to balance the system (i.e. to move the center of gravity of loads to coincide with the elastic center of the spring supporting system).
Step
Compute the total weight of the isolation system. Additional ballast equal to 25% of the weight of the equipment and ballast from Step 5 is added to provide for future changes in equipment. Also determine the equivalent uniform load equal to the total load divided by the area of the platform.
6.
Step 7.
Determine the natural frequency of the individual members of the platform (Using the equivalent uniform load computed in Step 6). The natural frequency is 9.87
EIg
27T
wL4
f -
for a simply supported beam with a uniform load. Step 8.
Using the shock spectrum for vertical motion, determine the required frequency of the system that will reduce the input accelerations to the maximum allowable. In addition, determine the displacement at this frequency.
Step 9.
Verify rigid body motion of the platform. The natural frequency of the individual members of the platform should be at least 5 times greater than the natural frequency of the system for rigid body motion of the platform to occur. To increase the frequency of the individual members, increase member sizes, and repeat Steps 3 through 7.
Step 10.
Determine the natural frequency for horizontal motion action) of the platform where 6A-28
(pendulum
TK 5-1300/NAVFAC P-397/AFR 88-22 1/21r
fhorizontal'motio~
(386.4/L)1/2
(6-56)
where L is the suspended length. Step 11.
Verify that dynamic coupling will not occur between vertical and horizontal motions. According to Section 6-47.3.2 dynamic coupling will not occur if fhorizontal motion < 1/2 fvertical motion
Step 12.
From the shock spectra for horizontal motion, determine the maximum dynamic displacement, velocity and acceleration using the frequency computed in St~P. 10. Verify that the maximum acceleration is less than the allowable.
Step 13.
Compute the load in each spring. Total Load Load in each spring - -----------------
Step 14.
Determine stiffness of springs to produce the required frequency of the system. Compute the static and maximum displacement. Using the static displacement andth~ load in.each spring from Step 13, calculate the required spring·s~iffness K from, load in each spring K----~----
static displacement Or using the vertical frequency from step 8, and the mass on each spring the stiffness can be calculated from' 2
K - [21r fvertl m
,.
~.
In addition, compute the travel of the spring according to Section 6-48.2, i.e. Maximum displacement travel 0.85
6A-29
'.
TH 5-l300/NAVFAC P-397/AFR 88-22 Example 6A-4' Design Shock
Isola~ion
System
Required:
Design an overhead pendulum shock. isolation system using a platform for a given loading.'
Step 1.
Given: a.
"
Structural configuration shown in Figure 6A-Sa and Figure 6A-Sb.
Figure ,6A-5
r
~S'vIIVEl CTYP,)
..
17.4'
L
JOINT
,-
SPRING CTYP,) ,
;
. (0.)
PLATFORM
,
ELEVATION
r
SECONDARY MEMBERSCTYP,) , . P,) , CTY •
10~8'
~
~
I
./I2,4' ~:~; )..,.-
I .
I ,
I~
I~ ~
/,
I~
I
I~ ~
I
I- 13.1' --l- 14.1' ---I--:- 16.1' ---1 (b) PLATFORM' (PLAN) ,
,"
6A-30, ,
2,4'
TM 5-l300/NAVFAC P-397/AFR 88-22
b.
Magnitude and location of loads on platform (Figure 6A-6) Equipment:
4960 Ibs . 4960 Ibs. 4960 Ibs.
El E2 E3
Floor panel -
5.5 psf (covers the whole area of platform)
Crating
7.2 psf
Live load
- 150 psf
Figure 6A-S
:C.8'
PLATFORM (PLAN)
FLOOR PANEL
~
GRATING FOR SUPPORT OF BALLAST
SECTION
c.
Shock spectra for hori~ontal and vertical motions are given in Figure 6A-7a and 6A-5b, respectively.
d.
Maximum
allowable acceleration
Vertical Motion Hori~ontal
0.50g
Motion - 0.75g
6A-31
TN S-1300/NAVPAC P:397/APa 88-22 . ".'
o o o o....
0
....
·0
r-,
en Q
U
I
«
to
>U C ill
ill
L :J
:J
CJ)
a-
on
u,
....
0
o
o o o
....
o o
....
....o
6A-32
00
ill
L
u,
Maximum Velocity,
inches/sec
~ ~
o o
00
~
o o o
., . ~
o o o
o
.
,~.
.
,'
'
ZZ-88 ~IAV/L6£~cI.:iVAAVN/00£1-~ H.L
o o o
TM 5-l300/NAVFAC p C397/AFR 88,:22 . Step 2.
Member sizes
Design Loads: Dead load - 20 psf (assume) Live Load - 150 psf Primary Member (a):
(See Fig. 6A-5a).
~
"
, w - i70 X (7.05,+ 8.05) Maximum Bending momept '- (wL2)/8
2567 lbs/ft.
(at center)
(2567 X 10.8 2) 1 8
.:""
- 37427 lb. -ft. Using allowable 'stress design, Allowable bending stress
- 0.66 y
(for compact shapes 'AISC) .
. Maxi~um Bending Moment
so, Required Sx ,-
Allowable Bending Stress' (37427 X 12) 1 (0.66 X 36,000) - 18.9 in 3 Try section'WlO X 21', Sx - 2l."in 3 > 18.9 in 3
O.K.
Check deflection: L'
Maximum allowable deflection _ 360 Maximum deflection'
.
'
.
-
10.8 X 12, I
360
(5WL4) 1 384EI
(at center)
.'
For WlO X 21,
L- 107 in 3
5 X (2567/12) X' (10.8,X 12)4 Maximum deflection 384 X 29,000 X 10 3 X 107 - 0.25 in < 0.36 in.O.K. so, Use WlO X 21 for all primary members.
6A-34 • ;
TH '-1300/NAVFAC
P~397/AFR.88-22
Secondary, Member, (b): See Fig. 6A-5a. w -'170'(1.0 + 1.2)
374 lb/ft:, ..
Maximum Bertding moment -' (wL2) / 8
'(at center)
. ,
(374 X 16.1 2)/8 - 12118 lb.-ft. Using allo~able stress d~sign,
,
.
Allowable bending stress" 0.66 F y (for 'compact shapes, AISC).' " 12118 X 12 so,
- 6.'12 in 3
Requited Sx , 0.66 X 36,000
TrY'~ection
W8 x 13, Sx - 9.90 in 3
>
6.12 in 3
O.K.
Check deflection: Maximum allowable deflection-
1360 16.1 X 12 360' .' 0.54'in.
For W8 X 13,
1.- 39.6 i~4
5WL4 Maximum deflection 384 EI. 5 X (374/12) X (16.1 X 12)4 384 X 29,000 X 10 3 X 39.6 . - 0.49 < 0.54 so, NOTE: Step 3.
Use W8 X 13 for all
s~condary
. O.K. members.
Members should be checked for concentrated equipment loads. Find center of gravity of the loads on the platform (see Figure 6A-8) .
6A-35 ".
TIl 5-1300/NAVFAC'P c397/AFR 88-22
Item
Weight· WI (lbs.)
A1A2 ... A6
B1B2 · · .B6
.'
X : (ft.)
W1X
Y (ft. ) .
,
.
."
W1Y (lb. ft.)
:(lb: ft.)
.
226.8
0.0
5.4
"'(.':{I
,226.8.__
13 .1
5.4
2971:1
,
0: 0_;.-
1224.7 1224.7
,
C1 C2 · . ,C6
226.8
27.2
5 ..4 :...
°1°2 , .. 06
226.8
43.3
5..4
A1B1C101
562.9
21.65
6169.0,: ,
t1
1224'.7
.. ,
,9820 ..4,
1224 ..7
10.8' '.,12187.0
6079.3
..
A2B2C 2D2
562.9
21.65
8'.4'
12187.0
A3 B3 C3 D3
562.9
21. 65
6.4
12187.0
A4 B4C4D4
562.9
21.65
4.4
12187.0
2476.8
A5B5C5D5
562.9
21. 65
2.4
12187.0,
1351.0
A6B6C6D6
562.9
21.65
0.0
12187.0
4728.4
.
"
. , 3602.6
0.0 ,
E1
4960.0
5.9
E2
'4960.0
E3
,
-,
"
, ,
-
1.0,
29264.0
4960.0
20.3
i.o
100688.0
4960.0
4960.0
34'.8,
1.0
172608.0
4960.0
A1D10 2A2
748.2
21.65
9.6
16199.0
7183.0
A5D5D6A6
748.2
21.65
1.2
16199;0
897:8
A2 Gl l G12A5
259.2
3.0
5.4
D2D5G14G15
259.2
4,0
5.':4
.
, Floor Panel
2572.0.
21. 65
, 23}51.4
:E
"
.-
.
,.
,
,
.
.
1400.0
10368.0
1400.0
55684.0
13889.0
493870.0
62786.7
>
5,4
..
.
:E'
I
.
777.6
.
i,
'1 '.'!"
1J
6A-36
e
e
e
Figure 6A-8 GRATING y, • , ,
'"""."",
I·
. ,, , cccc///ccccc~
',F»»»»»,
' " I ' "< f,CC/CC~ ,"57.,57 ~ .
t-
.1
,
c: / /
?I""'7nDl~ 2,0 2.4', ///«//fS,SSSS',>j4D 20' 7' ; } / "
2
,
, 2,' 0'
'"o > W
"'"
2.4'
x
:il
...'"w, o
t·
o
~... o
o W
'" "'"
$ co co ,
'" '"
TH 5-l3()O/NAVFAC P-397/AFR,,88-22
Center of gravity of loads (dead + live) - (Xl' 'Yl)
23751.4 -
20.8 ft.
;1'
.'
"
62786,.7'
,
,
'23751.4
, .-
2.64 ft. -'
l
Elastic center of the spring s'upport"~y.sfem:
Step 4.
,
,
,1
, .'
Figure 6A-9
.,
y
r-
S2
.
10,8'
L
S6! ,0 .
S4
,
Sa 1-:11--'
,
I~
S3
S5:'
Sl ~ 13.1' ~ 14.1 r ~'16,1' ~, ,
"
S7'
. 2.4'
1~;
2.4'
X
NOTE: All springs helve -the 'SOMe s'tiffness ..,'
,...'.
6A-38
I/~~. '.
TN 5-1300/NAVPAC P~397/APR 88-22
.
,
Spring No.
Force (W)
X
y
(ft; ),
(ft.' )
Sl
P
0
S2
P
0
S3 t ,
.'
p, ,p
S4
.
"
..
'
.
\
0
,
0
.
;
10. S '
0
0
S6
P
27.2
10.S
S7
P
43.3
•
.
10.SP
43.3P ~
10.S
"
"
SP
I:
0
27. 2P"
I j..'.
.
0
'.
43.3 ,
13.1P 27 .2P ,
27.2
I:
10.SP
,
P
P
.,
13 .1P
S5
Ss
0
':'
13.,1 '
.
0
'-' -'.
c.,
WY
,
0
10.S
.. 13.1'
, , "
WX
,
,\,
10.SP
,
0
43.3P
10.SP
l67.2P
43.2P
-to.
Elastic center of spring support system- '(X~,Y~ ) .
.'-' ,.
20.9 ft. , I
I .'
" 43.2P 5.4 ft. SP "
Step 5.
. " J.
t
Find the weight and loca:tfon'~f ballast to balance the system (relocate the e.g.' of the pi~tform t~ coiricid;'with the elastic center of the isolation sys t'ein) '. i ' a.
For the x-direction - Try placing ballast at x - 30.3 ft., i.e. 4 feet from the right edge of the platform. The ballast is placed symketrically about the x axis of the elastic center'so as-not to affect the location of the center of gravity in the y direction.
6A-39
TH 5-l300/NAVFAC P-397/AFR 88-22 Weight of ballast - WBX
'.
EWlX + 39. 3WBX - Xii [ EW l + WBX] 493870 + 39.3Wax - 20.9(23751.4 + WBX) Wax b.
-
2534.3/18.4
- 137.7 Ibs. '
For the y direction - Try placing ballast at y _ 9 ..6 feet. i.e. 1.2 feet from the top of the platform. The ballast is placed symmetrically ~bout the y axis of the,elastic center so as not to affect the location of the center of gravity in the x direction. Weight of ballast -WBY EWlY + :9.6W By -
Ys [
EW l + WBy] ,
62786.7 + 9.6WBy - :5.4(2375.. 1 + WBY) WBY - 65470.9/4.2 c.
- 15588.3 1bs.
Total ballast: WB - WBX + WBY - 137.7 + 15588.3 - 15726 1bs.
Step 6.
Total load on the platform and equivalent uniform load. a.
Total load. Dead Load: Floor panel Members Grating ~ L -
2572 Ibs. 4284.6 2014.8 8871.4 1bs.
Live Load: Equipment - 14880 Lbs . 'Ballast . -, 15726 Additional Ballast - 0.25 (30606) '7651. 6 Personnel (5 @ 150 lb.) r >. 750 ,
, Total
'
L
,_.
3~,007.
5 Ibs.
8871.4 +,)9007.5 -, 47878.91bs.
6A-40
TH5 c1300/NAVFAC. Equivalent uniform load
b.
P-397/AFR~'88-22
.,
i ,t
. 47878.9 10. 8X 43.3
.
.-
- 102.4 psf S './
;:
)
.170
psf ~,
O.K.
,
Preliminary design 'of platform members'.is O;K. However, members supporting ballastmust be checked as their actuaL,load"may be higher than the equivalent uniform load. , Step 7.
Natural frequency of the individual members of the platform., r; r
.
J
For a simply supported, member with a uniform load, 9.87 Natural frequency f n
211'
Primary Members (a): A portion of the adjacent slab acts with the beam. Add 20% of the msss of'the 'slsb on each side of the beam to 'the, actual mass of the beam. 10110 X 21 I -
107 in 3.
L
10.8 X 12
,
j ..
0'
b
- 6.04 feet -rg - 386:.:4" in/sec 2 ,
w:
"
. E••- 29000, X 10 3 psi. ~. ..,'.. .c (Use equivalent uniform load); ,
...
.'
;'
w - ·102.4 X.6.04 • \..
,.;' •
~
r
-, ("l
.: ..': - 51'.54) lb/in tn,,)"
',t·
:
I . . . . . ,j
618.5 Ib/ft
r
6A-41
. i
',0'
·c< ~', ,','
TN 5-1300/NAVFAC P-397/AFR 88·22
9.87 . [ 29000 X 10 3 X 107 X 386.4 ] fn -
2'1l'
~.
51.54 X (29.6)4
- 14.3 cps Secondary Members (b): Since the spacing of the secondary beams is less than 1/4 of the length of the beams the total mass of the slab aet with the beam. .,.
w
102.4 X 2.2 ' ..
-<
- 225.3: lb/ft 18.78 lb/in E
- 29000 X 10 3 psi
g
- 386.4 in/see 2 9.87 ['29,000 X 10 3 X 39.6 X 386.4
fn -
'18.78 X (193.2)~
2'1l'
- 6.5 cps Step 8.
r
Required frequency of system ·to l'imit motions. To limit the maximum acceleration of 'the system to 0.5g or less, choose the frequency of the system from Figure 6A-7b as f - 1 cps which' produces a maximum acceleration of 0.4g and a maximum dynamic displacement of 4.7 inches.
Step 9.
Verification of rigid body motion of, platform .. findividual member>
5 X fsystem
From Step 8., fsystem
-
1 cps
6A-42.
TM5-l300/NAVFAC:P~397/AFR88-22 .1
From Step 7 .•
14.2 cps> 5 cps
fprimary member
O.K..
".
fsecondary member'- 6.5 cps> 5 cps
O.K.
Therefore, platform is in 'rigid body motion. Step 10.
Natural frequency of platform for horizontal motion (i.e. pendulum type action). . =1
'
Assume the center of gravity of supported mass is located at the top of the platform so that the length of the pendu1~ arm is 17:4 feet. ,i
Frequency of platform for horizontal motion' f
1/2w (386.4/L)1/2 0.22 cps
Step 11.
(6-56)
,
Check for dynamic coupling of vertical and horizontal motion. From Step 8., fvertica1 motion From Step 10.,
1 cps, f " •
i:
fhorizonta1 motion - 0.22 cps
<
1/2 fvertica1 motion - 0.5 cps
Dynamic coupling will not occur. Step.12.
Maximum dynamic disp1acement"ve10city and acceleration for horizontal motion. From shock spectra for horizontal motion, (Fig.6A-7a) for f - 0.22 cps,
Maximum acceleration - 0.007g< 0.75g Maximum velocity - 0.9 in/sec.
.....
~
-
,.-
O.K. ,
-
, Maximum dynamf c displacement - 1.3 in. The maximum dynamic displacement is the required "rattle space" or the minimum horizontal clearance' between the platform and the structure or anything attached to the structure.
6A~43
TH5-1300/NAVFAC P-397/AFR 88-22 ' Step 13.
Load in each spring.
. :
Total Load, Load in each spring - - - - _ , Number of Springs
.'
47878.9
..- ,.,'
'
n-
8 ,
Step 14.
D~sign
a.
5985 Ibs.
,
the springs.
:. I
Static displacement From Step 8 .• For a maximum acceleration of O.48g the maximum dynamic displacement - 4.7 in. .se , Static
b.
d1splacem~rt
- •4.7/.0 .48
_ ':
Stiffness of spring - 4.7/0.48 Load in each spring
.' .'
K--------Static displacement 5985
" - 611.2 Ib/in
.'
9.79
,
or K -
Kc.
,(2fT X lcps)2(5985/386~.4)
611.5 lb/in
Maximum travel of spring. Maximum displacement - (static + dynamic) displacement - 9.79 + 4.7 - 14,49 in
6A-44', .
"
TK·5-l300/NAVFAC P-397/AFR 88-22 From Section 6-48.2 maximum displacement Travel of spring 0.85 14.49 0.85 17.0 in. Thus the vertical rattle space (clearance) is 17.0 inches.
6A-45
TH 5-1300/NAVFAC P-397/AFR 88-22
APPENDIX 6B LIST OF SYMBOLS
TH' S-1300/NAVFAC P-397/AFR 88-22 ' • acceleration' (in,/ms 2),
(1)'
(2)
depth of equivalent rectangular stress block (in,) long span of a panel (in:)
(3)
area (in. 2) area'of diagonal'bars at: the support· within 'a 'width b (in. 2): . t:;
"
door area (in. 2) area of gross section (in 2) t:
r,
~.
....
'.__~
.:
net' area 'of section (in. 2) area of openings (ft 2)
.
.
•
fO,.
-
',;;.:
~,
~
.
area of prestressed reinforcement (in. 2) area of tension reinforcement within a width b (in. 2) . " "
area of compression reinforcement within a width b (in. 2) area of rebound reinforcement (in. 2)
..
area of flexural reinforcement within a width b in the horl•. ' zontal ·di·rection on each, face (in 2 ) *
area of flexural reinforcement within a width b in the vertical direction on each face (in2,)~. ' total area of stirrups or lacing reinforcement in tension within a distance, Ss or sl and a width b s or b l (in. 2 ), J
J.
~
I"
_
area of sector I and II, respectively (in. 2 ) ,.
(1) r.
. (2)
(3)
'~.'
,;
. '.
width of compression face of flexural member (in.) 'width of concrete 'strip in which ·the direct shear stresses at the supports are resisted by diagonal bars (in.) ~, short span of a panel (in.)
width of concrete strip in which the diagonal,tension stresses are resisted by stirrups of area Av (in.) width of concrete strip in which the diagonal tension stresses are resisted by lacing of area Av(in.) 'e
B
(1) (2)
constant defined in paragraph. peak blast overpressure capacity ;':' ...: .- .
c
l .
shear coefficient f • .'
"
*
See note at end of symnbols
6B-l.
- d
,.
TIl 5-1300/NAVFAC P-397/AFR 88-22 c
(1)
(2) (3)
distance from the resultant applied load to the axis of I·'.,·, rotation (in.) damping coefficient distance from extreme compression fiber to neutral axis (in. )
distance"from,the resultant applied load to the axis of rotation for sectors I and II, respectively' (in.) critical damping shear coefficient fo rc.u'l t Imate shear stress of one-way elements
(1) (2)
drag coefficient coefficient for center deflection of glass "
drag pressure (psi) • peak drag pressure (psi) equivalent load factor "
post-failure fragment coefficient (lb2-ms4/in. 8) shear coefficient for ultimate shear stress in horizontal direction for two-way elements*
,
" leakage pressure' coefficient
, ;maximum shear coefficient ,. " force coefficient for shear at the corners of a window frame ,
, coefficient for effective, resistance of glass peak reflected pressure coefficient at angle of incidence a dilatational velocity of concrete (ft/sec) shear coefficient for ultimate support shear for one-way elements shear coefficient for ultimate support shear in horizontal . direction for two-way el'ements* shear coefficient for ultimate support shear in vertical ' direction for two-way elements* , coefficient for period of vibration for glass
C' u
*
impulse coefficient at deflection
'Xu
(psi-ms 2/in. 2 )
impulse coefficient at deflection
Xm
(psi-ms 2/in. 2 )
See note at end of symnbo1s
68-2
,
.
,
TH 5-l300/NAVFAC P-397/AFR 88-22
shear coefficient for ultimate. shear stress in vertical direction for two-way elements* shear coefficient for the ultimate shear along the long side of window frame shear coefficient for the ultimate shear along the short side of window frame
..
impulse coefficient at deflection Xl (psi-ms 2/in. 2) (2) '.parameter defined in figure, (3) ratio of gas load to shock load (1)
.
"
impulse coefficient at deflection
Xm
(psi-ms 2/in. 2)
ratio of gas load duration to shock load· duration distance from extreme compression fiber. to centroid of tension reinforcement (in.) 'distance from extreme compression fiber to centroid of compression reinforcement (in.)
d'
c • •,.
• !
~
distance between the centroids of the compression and tension reinforcement -(in.) .
I.-
diameter of steel core·(in,) distance from support and equal to distance d·or. d c (in.) inside diameter of-cylindrical explosive 'container (in.) distance between'center lines of adjacent lacing bends measured normal to flexural reinforcement (in.) j.l
! ..,'
.'~
.
'
,
distance' from extreme compression fiber to centroid of prestressed reinforcement (in.) diameter of cylindrical portion of primary fragment (in.)
(1) (2) (3)
unit flexural rigidity (lb-in.) location of shock front for maximum stress (ft) minimum magazine separation distance (ft)
nominal diameter of reinforcing bar (in.) . equivalent loaded width of structure for non-planar wave front (ft) ;
',I;'
DIF
dynamic increase factor
DLF
dynamic load factor
*
See note at end of symnbols
TK
5-~300/NAVFACP-397/AFR
88-22
base of natural logarithms and equal to·2.7l828." distance from centroid of section to centroid of prestressed reinforcement (in.)
e
(1) (2)
(2E·)1/2
Gurney Energy Constant (ft/sec),
E
modulus ofLelasticity
Ec
modulus of elasticity of concrete (psi)
~
modulus of elasticity of masonry units (psi)
Es
modulus of elasticity of reinforcement (psi)
f
(1) unit external force (psi) (2) frequency of vibration (cps)
. ,., ~
, .
.
f' c
,
.
static ultimate compressive strength of concrete at 28 days (psi) ,dynamic ultimate compressive strength of concrete (psi)
f'dc
dynamic ultimate compressive strength of masonry units (psi)
f' elm "
.
,'.'
"
f ds
dynamic design stress for reinforcement (psi)
f du
dynamic ultimate stress of reinforcement (psi):
f dy
dynamic yield stress of ,reinforcement (psi)
f' m
static ultimate compressive strength of
fn
natural frequency .of vibration (cps)
f ps
average stress in the prestressed reinforcement at ultimate load (psi) • '
f pu
specified tensile strength of prestressing tendon (psi)
f py
yield stress of prestressing tendon corresponding to a 1 percent elongation (psi)
fs
static design stress for reinforcement (a, function of f y' f u &
f se
effective stress in prestressed reinforcement after allowances for all prestress losses (psi)
masonry~units
(psi)
•
.I
fu
static ultimate stress of reinforcement (psi)
fy
static yield stress of reinforcement (psi)
F
(1) (2) (3)
total external Iorce (lbs) coefficient for moment of inertia of cracked section function of C2 & Cl.for bilinear triangular load
1M 5-l300/NAVFAC P-397/AFR 88-22
force in the
rein~orcing
bars (lbs)
equivalent external fo rc s- (lbs) g
(1) (2)
variable defined in table 4-3 , acceleration due to gravity (ft/sec 2)
G
shear .mo du Lus (psi)
h
(1) (2)
h'
clear height between floor slab and roof ,slab
H
(1) (2)
charge location parameter (ft) height of masonry wall
span ,height (in,) distance between reflecting surface(s) and/or free ,edge(s) in vertical direction (f t ).
Hc Hc Hs
e
e
ILr i
unit positive impulse (psi-ms)
i-
unit negative, i;'puls'e (psi-ms)
ia
sum of scaled unit blast impulse capacity of receiver panel and scaled unit blast impulse attenuated through ioncrete and sand in a composite element (psi-ms/lb l/ 3) ,
ib
unit blast impulse (psi-ms)
ib
scaled unit blast impulse (psi-ms/lb l/ 3)
i bt
total scaled unit blast impulse capacity of composite element (psi-ms/lb l/ 3)
i ba
scaled unit blast impulse ca~acity of receiver panel of composite element (psi-ms/lb 13 )
i bd
scaled unit blast impulse capacity of donor panel of composite element (psiCms/lb l / 3) , r
ie
unit excess blast impulse (psi-ms)
ir
unit positive normal reflected impulse (psi-ms)
i r
unit negative normal reflected impulse (psi-ms)
is
unit positive incident impulse (psi-ms)
6B-5
TK 5-1300/NAVFAC P-397/AFR 88-22 i
s
-
unit negative incident impulse (psi-ms) moment of inertia (in. 4)
I
average of gross and cracked moments of .inertia of width b (in. 4 ) . moment of inertia of cracked concrete section of width b (in. 4) moment of inertia of gross concrete section of width b (in. 4) mass moment 'of inertia (lb-ms 2-in.) moment of inertia of net section of masonry unit (in. 4 j
ratio of distance between centroids of compression'and tension forces to the depth d
k
constant defined in paragraph
K
(1) (2)
unit stiffness (psi-in for slabs) (lb/in/in for beams) (lb/in for springs) constant defined in paragraph
elastic unit stiffness (psi/in for slabs) (lb/in/in for beams) elasto-plastic unit stiffness (psi-in for slabs) (psi for beams) equivalent elastic unit stiffness (psi-in for slabs) (psi for beams) equivalent spring constant load factor - ;
,
load-mass factor load-mass factor in the ultimate range load-mass factor in the post-ultimate'range mass factor . resistance factor
factor defined in paragraph kinetic energy 1
charge location parameter (ft) spacing of same type of lacing bar' (in.)
68-6
TK
L
(1) (2)
5~1300/NAVFAC
P-397/AFR 88-22
span length (in ..) * ., distance between reflecting ~urface(~)and/or free edge(s) in horizontal direction (ft)
length of lacing bar required in distance sl (in.) ,embedment length of reinforcing bars (in. ) length of shaft (in.) i .
wave length of positive pressure phase (ft) wave length of negative pressure phase (ft) wave length of positive.pressure phase at points band d, respectively (ft) total length of sector of element normal to axis of rotation (in. ) m
unit mass (psi-ms 2/in.) average. of the effective elastic and plastic.unit masses (psims2/in. ) effective uni/mass (psi-ms 2/in.) effective unit mass in the ultimate range (psi'ms 2/in.) effective unit mass in the post-ultimate range (psi-ms 2/in.)
(1) (2)
unit.bending moment (in-lbs/in.)· total mass (lb-ms 2/in.)
effective total mass (lb-ms 2/in.) ultimate unit resisting moment (in-lbs/in.) ultimate unit rebound moment (in-lbs/in.) moment of concentrated loads about line of rotation of sector
(in.-lbs) fragment distribution parameter equivalent total mass (lb-ms 2/in.) ultimate unit negative moment capacity in horizontal direction (in.-lbs/in.)* , ' . ' ultimate unit positive moment capacity in horizontal direction (in. -lbs/in.) *
*
See note at end of symnbols
6B-7
.
... .
TM 5-1300/NAVFAC P-397/AFR 88-22 ultimate unit negative moment capacity at supports (in. . lbs/in.) ,
Mp
.'
ultimate unit positive moment capacity at midspan (in. lbs/in. ) ultimate unit negative moment capacity in vertical direction (in. -lbs/in.)* ultimate unit positive moment capacity in vertical direction (in. -lbs/in.)*
(1)
n
(2)
(3) N
modular-ratio number of time intervals number of glass pane tests
number of adjacent reflecting surfaces number of primary fragments larger than Wf
p
reinforcement ratio equal to (As/bd) or (As/bdc)
p'
reinforcement. ratio equal to (A"s/bd) or (A's/bdc) reinforcement ratio producing balanced conditions at ultimate strength prestressed reinforcement ratio equal to
~s/bdp
Pm
mean pressure in a partially vented chamber (psi)
Pmo
peak mean pressure in a partially vented chamber (psi) reinforcement ratio in horizontal direction on each face * reinforcement ratio equal to PH + Pv
Pv
reinforcement ratio in vertical direction on each face *
p(x)
distributed load per unit length
p
(1) (2)
pressure (psi) concentrated load (lbs)
negative pressure (psi) interior pressure within structure (psi) .. interior pressure increment (psi) fictitious peak pressure (psi) peak pressure (psi)
*
See note at end of symnbols
6B-8 '
TH 5-1300/NAVFAC P-397/AFR
e
peak positive normal reflected pressure (psi)
Pr Pr
88~22
-
peak negative normal reflected ,pressure (psi)
'r.
Pra
peak reflected pressure at angle of :incidence a' (psi)
Ps
positive incident pressure (psi),
Psb' Ps e
positive incident pressure at points, band e, respectively (psi)
,
,
,
Ps o
peak positive incident pressure (psi)
Ps o -
peak negative incident pressure , A"
'.
,,
, ,
Psob,Psod,Psoe
peak positive incident pressure at'points b, d, and e, respectively (psi)
P(F)
probabili ty of failure of glass pane 1
•• ..1. •
dynamic pressure (psi)
q ..: i; '.'
,.',' 'dynamic- pressure, at 'points band
'e ,
respectively (psi)
peak dynamic pressure (psi) .... ,
peak dynamic pressure at points band e, respectively (psi) r
(1)
(2)
e
unit resistance (psi) radius 'of spherical TNT (density equals 95 lb/ft 3 charge (ft»
r
unit rebound resistance (psi, for panels) '(lb/in for beams)
b.r
change in unit resistance (psi, for panels) (lb/in for beams)
rd
radius from center. of impulse load to,center of door rotation (in. )
re
elastic unit 'resistance (psi" for panels) (lb/in for beams)
rep
elasto-p1astic unit resistance (psi, for panels) (lb/in for beams)"
rs
radius of shaft .(,in.)
ru
ultimate 'unit resistance (psi, for panels) (lb/in for beams)
r up
post-ultimate unit resistant (psi)
r1
radius of hemispherrca1 portion of.primary fragment (in.)
R
(1) (2)
'.'.
'
total internal resistance (lbs) slant distance (ft) 68-9
.«
TH 5-l300/NAVFAC P-397/AFR 88-22
distance ·traveled by primary fragment (ft) uplift force at corners' of· window. frame
(lb~)
..,.'
radius of lacing bend (in.) normal distance (ft) equivalent total, .internal .resistance (lbs) ground distance (ft) total ultimate resistance total internal resistance of sectors I and II, respectively (lbs) '" s
sample standard deviation spacing of stirrups in the direction parallel to the longitudinal reinforcement (in.) spacing of lacing .in the direction parallel to the longitudinal reinforcement (in.)
s
height of front wall or one-half its width, whichever is smaller (ft) strain energy
t
time (ms)
At
time increment (m?) any .time (ms) time of arrival of blast wave. at points b, e , and f, respec. tively (ms) (1) clearing time for reflected pressures (ms) (2) container thickness of explosive charges (in.) rise time (ms) time to resch maximum elastic
defl~ction,
•
,time at which maximum deflection occurs (ms) duration of positive phase of ,blast pressure .(ms) duration of negative phase of blast pressure (ms) fictitious positive phase pressure duration (ms) fictitious negative phase pressure duration (ms) 68-10
,
TK 5-l300/NAVFAC P-397/AFR 88-22'
e
tr
fictitious reflected press~re' duration (ms)
tu
time at which ultimate dliflecti'on occurs (ms)
ty
time', to,; 'reach yield (ms)
tA
time of arrival of blast wave (ms) ~
;,. ~
'. ,.
.
tl
time at' which partial failure occurs (ms)
T
(1)
(2)
duration 'of equivalent triangular.' loading function (ms) thickness of masonry wall .'
e
"
Tc
thickness of concrete section (in.)
Tc
scaled thickness of concrete section (ft/lb l /.3 ) ,
Tg
thickness of glass (in.)
Ti
angular impulse load (lb-ms-in.)
TN
effective natural period of vibration (ms)
Tr
rise time (ms) ,
Ts
thickness of sand fil,l (in :),;
Ts
scaled thickness of sand fill (ft/lb l/3)
u
particle velocity (ft/ms)
Uu
ultimate flexural or anchorage bond stress .(psi)
U
shock front velocity, (ft/ms) -
Us
strain,energy
v
velocity (in./ms)
va
instantaneous velocity at .any-tLme (in.-/ms)
vb
boundary velocity for primary'fragments (ft/sec)
V
ultimate shear stress permitted on an unreinforced web (psi)
c
"
r
.'
',"'
- .
....
=j
vf
maximum post-failure fragment velocity (in./ms)
vf(avg. )
average post-failure fragment velocity (in./ms)
vi
velocity at incipient failure deflection (in./ms)
Vo
initial velocity of primary fragment (ft/sec)
vr
residual velocity of primary fragment after perforation (ft/sec)
•
e
.',
j'
,~
I
'~
v.' .
6B-11
TH 5-1300/NAVFAC' P,.397/AFB. 88-22
striking velocity of primary fragment (ft/sec) ultimate shear stress;(psi) ultimate shear stress at distance de from the. horizontal support(psi) *
.'. ultimate shear stress at distance de from the vertical support (psi) *
v
volume of pa~tia11y.vented.chamber·(ft3). ultimate direct shear capacity of the concrete of width b (lbs)
V dH
shear ati d!stance.de from the. ve r t Lca Lesuppor t; ona unit width (lbs ./in.) i
" . shear at distance de from the horizontal support on a unit width (lbs/in.) * , .'
volume of structure (ft 3).
:'"
.I
• J
'1.'.
",
shear at the support (lb/in, for panels). (lbs.for 'beam) shear at the vertical support· ~n a unit width (lbs/in.)* shear at the n~rizonta1 support on a:unitwidth (lbs/in.)* total shear on a width b (lbs)
..'
.'
unit shear. along the long side of window frame' (lb/in.) unit shear along the short side;of .window~ framec 01bs/in.) w
unit weight (psi, for panels) (lb/in for beam) .
Wc
weight density of concrete. (lbs/ft 3)
W
weight density:of.sand,(lbs/ft 3)
s
(1) (2)
charge weight (lbs) weight (lbs)
, I' .
total weight of explosive containers (lbs) .1
: [
•
"
weight of primary fragment (oz) total weight of steel core (lbs) :
tota1.weight of plates land 2·,. respectively (lbs). 1
width of structure (ft) -c',
* See note at end of symnbo1s
68-12
,,
... 1
TK S-1300/NAVFAC P:.397/AFR 88-22
work done x
yield'line location in horizontal direction (in.)*
x
deflection (in.) -
any deflection '(in.)
,
lateral deflection to which a masonry wall develops no resistance (in.) elastic deflection (in.) , elasto-plastic deflection (in.) maximum penetration into concrete of armor-piercing fragments (in. )
X'f
maximum penetration into concrete of fragments other than armor-piercing (in.) maximum transient deflection (in.) plastic deflection (in.)
(1) (2)
maximum penetration into sand of. armor-piercing fragments (in.) static deflection
ultimate deflection (in.) equivalent elastic deflection (in.).
(1) (2)
partial failure deflection (in.) deflection at maximum ultimate resistance of masonry wall (in.)
y
yield line location in vertical direction (in.)*
Yt
distance from the' top of-section to centroid (in.)
Z
scaled slant· distance (ft/lb l/3) scaled normal distance (ft/lb l/3) scaled ground distance (ft/lb l/ 3)
a
(1) (2) (3)
*
angle formed by the plane of stirrups; lacing, or diagonal reinforcement and the plane of· the longitudinal reinforcement (deg) angle of incidence of the pressure front (deg) acceptance coefficient
See note at end of symnbols
6B-13
.
TK 5-l300/NAVFAC P-397/AFR 88-22.
B
(1)
coefficient for determining elastic and elasto-plastic
(2)
particular support .r'o t a t Lon angle,(deg)rejection coefficient
resistances
(3)
factor equal to 0.85 for concrete strengths up to 4000 psi and is reduced by 0.05 for each 1,000 psi in excess-of 4,000 psi ~oefficient
y
for determining elastic and elasto-plastic deflec-
tions factor for type of prestressing tendon unit strain in mortar (in./in.)
e
. support rotation angle (deg). angular acceleration (rad/ms 2)
e
maximum support rotation angle (deg) horizontal rotation angle (deg) * vertical rotation angle (deg) * .increase in support rotation angle after partial failure (deg) ductility factor
- .'
......
v
Poisson's ratio
~o
effective perimeter of reinforcing. bars (in.) summation of moments-(in.-lbs) , . sum of the ultimate unit resisting. moments acting along the negative yield lines (in.-lbs) ~
.
.~.
sum of the ultimate unit resisting moments acting along the positive yield lines (in.-lbs) maximum shear stress in the shaft,(psi) ( 1)
(2)
capacity reduction bar diameter (in.)
fac~or
,.
assumed shape function for concentrated loads
,-
.~
"'(x)
assumed
shap~
,
function for distributed. loads free edge
'angular velocity (rad./ms) '.
simple support
*
See note at end of symnbols
6B-14
.-
~.
TK 5-l300/NAVFAC P-397/AFR 88-22
III/II
fixed support either fixed, restrained,; or simple support
*
Note.
walls.
This symbol was developed for two-way elements which are used as When roof slabs or other horizontal elements are under consideration,
this symbol will also be applicable if the element is treated as being rotated into a vertical position.
6B-15
TM 5-1300/NAVFAC P-397/AFR 88-22
APPENDIX 6C
BIBLIOGRAPHY
TK 5-l300/NAVFAC P-397/AFR'SS-22
Mas.onry:·
1. 2.
3.
4.
'. '
"
'.
.~
Design of Masonry Structures to Resist the Effects of HE Exp1osions~ prepared by Ammann (, W)1itney, , Consulting E..ngineers;. New. York,' ,NY, for. Picatinny Arsenal. Dover;' NJ, 1976.", . ". ',:' Gabrielsen. G..• Wilton, C., and Kaplan. K.• Response of Arching',Walls" " and Debris from Interior Walls Caused'by Blast LoadingiURS,'7030-n.:.URS Research Company, San,Mateo;CA: .. February'1976. , . McDowell, E., McKee, K.. , and Sevin, "E.,.• :'Arching Action Theory'of Masonry ~ , Journal. of the Structural Division. ASCE, Vol.'· 82, No. ST.2. March 1 9 5 6 . " <' • , , . . . . . .' Wilton. C.• and Gabrielsen, B.• Shock Tunnel Tests of'Pre-Loaded and Arched Wall Panels. URS 7030-10, URS Researg~. Company, San: Mateo , CA. June 1973. '. .'
,~".
Precast· Concrete _",
5. 6.
7. 8. 9. 10, 11,
12.
13. 14. 15,
"
J,
"\ . '
,
','
PCI Design Handbook Precast', Prestressed-~Concrete, Pres.tres~ed Concrete
Institute, :Chicago. IL, 1978.. ,..., . '... , , PCI Manual for Structural Design of Architectura1 Precast Concrete ..... Prestrel!se.d Concrete Institute. Chicago; I L . , , -. , , PCI Manual on Design of ConneCtions tor Precast Prestressed 'Concrete.: Prestressed Concrete Instit,:,te; Chicago'; ..IL. '."" ,.... .. Lin, T. 'Y~.,Design of Prestressed ConCrete.Structures, Johp W~ley (, Sons, New, York. NY •. July 1955.' ; , -". : . Pre ~ Engineered: Buildings .~ American National Standard Minimum Design Loads for Buildings and ~ . Structures;.ANSI.A58.1-1982, Ameri~an National Standards Institute. New York. NY. March 1982. , ,; .Specification for the Design of Cold-Formed Steel Structural Members. American, Iron and Steel 'Institute. New York •. NY. 1968 .. Specification for the Design, "Fabrication 'and Erection of Structural. Steel for Buildings. Manual of 'Stee1 Construction., American Lns t i tutie of. Steel. Construction, New York, NY. 1969. ' Uniform'Iitiilding Code <1979 Edition)'. International. Conference of ,,;', Building Officials, Whittier, .CA. 19}9. , : . ... ','. " Healey, J" J,. "et a1., Design of Steel Structures to Resist the Effects of HE Explosions, Technical Report 4837. Picatinny Arsenal, Dover •. -NJ, August ,1975. :'. " " .., .. ,.. , . Stie a , W.• ' et a1.:. B1a~t Capacity Evaluat-ion 'of Pr~'-Engineered- Buildt'ng, Contractor Report ARLCD·CR·79004, U. S. Army Armament Research and Development Command, Dover. NJ, March 1979.' '.::' " Stea, W.< et al., -Nonlinear Analysis of Frame Structures to -Blast Overpressures, by Ammann (, Whitney, Consulting Engineers, New Yo~k. ~~ Contractor Report ARLCD-CR-77.0.08, U. S., Army Armament. Research and Development Command. Dover, May 1977. . , Tseng, G., et.a1., Design Charts for· Cold-Formed Steel Panels and'WideFlange Beams Subjected to Blast Loads, Technica~ Report 4838, PicatinnyArsenal...· Dove r, NJ, August 1975. ".' ..
"
&i.
16.
"l
;:.'
Suppressive Shielding 17,
Final- Report -'Application of Suppressive Structure Concepts to Chemical~<
Agent -Munitions Demilitarization System (CAMDS), Report EA-FR-2B02,. Edgewood Arsenal, Aberdeen Proving Ground, MD, June 27, 1973. 6C-I
"
TM 5-1300/NAVFAC P-397/AFR'88-22 18. 19. 20.
21. 22.
23. 24. 25. 26.
27. 28. 29.
30.
31. 32.
Shields Operational for Ammunition Operations. Criteria for Design of. and Tests for Acceptance, MIL STD 398, U.. S. Government Printing Office, Washington', D. C., November 5 ,_~ 1976. ., .!_ Study of Suppressive Structures Applications to an' 81 mm Automated. Assembly Facility, Report EA 1002, Edgewood Arsenal~'Aberdeen Proving Ground; MD, ·April 16, 1973·. I,. Suppressive Shielding, AAI Corporation~ C6ckeysville~. MD, DAAA15-75-C c ' 0120, U. S. Army Armament Re seerchrand Development Command, Chemical Systems ·Laboratory, ,Aberdeen Proving 'Ground; MD, April 1977. ,. . Suppressive· Shields Structural Design' and:Analysis 'Handbook,' HNDM 11l0~ " 1-2, U. S. Army Co~ps of Engineers, Huntsville Division, Huntsville. AL,: November 1977.. . i ,". r: '. -, System Safety Program Reguirements~\ MIL, STD," 882A, U. S. Gover,nmep.t > '" Printing Office, Washington, D. C., June 28, 1977. ~,f: Baker, W. E., et al., Design Study of a Suppressive Structure for a Melt Loading Operation, EM-CR-76043, Report NO,'9; Edgewood Arsenal, Aberdeen Proving Ground, MD, December 1975, Baker, W.'E.'and Oldham, G. A,."Estimates of Blowdown of· Ouasi-Static Pressures in Vented Chambers, EM-CR-76029, Edgewood Arsenal;. ,Aberdeen Proving Ground, MD, November 1975,·,:' , Baker, W. E, and Westine, P. S:, Methods of Predicting Blast Loads" Inside and o'utside' Suppressive' Structures', 'EM,CR'- 76026, -Re po r.t; No..' 5',,:, ;'" Edgewood Arsenal, Aberdeen Proving Ground, ' MD; ·1975, " ' . -: Cox, P. A. 'et al. Analysis and,Evaluation of Suppressive Shields,t Edgewood Arsenal Contractor Report ARCFL-CR-'77028', Report No.. '10', . ' ; " , ' Contract No, DAAA15-75-C-0083, 'Edgewood Arsenal,' Aberdeen Proving Ground, MD; January 1978... " , 'J_ • : ,:1; Dutton, S. R.' and Katsanis, D. 'Computer-Aided Design of' Suppressive, Shields, EM-CR-76080, AAI Corporation, Baltimore, MD, Report· for'; . , " Edgewood Arsenal, Aberdeen,Proving.Ground,·MD, June 1976.' " Esparza, E. D., Estimating External Blast Loads from 'Suppressive c'" Structures,'. Edgewood Arsenal Contract Report EM-CR-76030, Report ,No~ 3:, Edgewood Arsenal .. Aberdeen Proving Ground"MD, November 1975. _. Esparza, E. D., Baker, W. E., and Oldham, 'G.' A., Blast Pressures ,Inside' and Outside' Suppressive Structures, Edgewood.Arsenal Contractor Report: EM-CR-76042, Report No.8, Edgewood Arsenal, ~berdeen Proving Ground,.' MD, December 1975. '" . . . :, Gregory,F: H., Blast Loading' Calculations and Structural. Response ..", Analyses of the 1/4-Scale Category I Suppressive Shield, BRL Report No., 2003, U. ,'S. Army Ballistic Research Laboratory, Aberdeen Proving Ground,. MD, August' 1977. ,. Hubich, H. O. and Kachinski, R. L., ExpYosive Waste Removal Systems'for Suppressive Shields, EdgewQodArsenal Contractor Report No. EM-CR76002, " Edgewood Arsenal, Aberdeen 'Proving Ground,' MD, August,..1975. Jezek, B. W.~ Suppressive Shielding for Hazardous Munitions Production Operations, Technical Report No. ARCSI.;-TR-77020,. U. S. Army Armament Research .and Development Command, Chemical Systems Laboratory, Aberdeen' Proving Ground, MD, April"1977. '._ , 1 _ .' Kachinski, R. L., et al., Technical Feasibl:-I'fty of Suppressive.:Shi'elds' for Improved Hawk Launch Sites, Edgewood Arsenal Contractor Report No. EM-CR-76057. Edgewood Arsenal, Aberdeen Proving'Ground, MD, February 1976. Katsanis, D. J .. Safety Approval of Suppressive Shields, Edgewood';' Arsenal Technical Report No.' EM-TR-76088, Edgewood Arsenal', Aberdeen Proving Ground, MD,' Augus t 1976. ." .. t
,r
I
i ,
33.
34.
6C-2
,
,
TM 5-l300/NAVFAC P-397/AFR 88-22
35.
Katsanis, D.·J. and Jezek,·B. W., Suppressive Shielding of 'Hazardous Ammunition Production.Operations, Technical Report No. EM-TR-760l5, Edgewood Arsenal, Abe rdeenvPr ovfng Ground, MD, December 1975.'.
36.
Kingery, C.
37.
38.
39. 40.
41. 42.
43. 44.
45.
46.
47.
48.
49.
N., Coulter, ·G'., and Pearson, R.', Venting of'·Pressure
through Perforated Plates. Technical Report No. ARBRL-TR-02105, U. sG Army Ballistic Research' Labor atory , Aberdeen Proving Ground, 'MD, September 1978. Kf.nge ry-, C. N., Pearson, R., and Coulter, G., Shock Wave Attenuation !u Perforated,Plates with Various Hole Sizes, BRL'Memorandum Report No. 2757, U. S Army Materiel Development and Readiness Command, Alexandria~ VA; June 1977. ' .. Kingery, C. N., Schumacher;.R., and Ewing, W., Internal Pressure from Explosions in Suppressive Shields, BRL Memorandum Report No.. ARBRL-MR02848, U. S. Army Armament Research'and Development Command, Aberdeen Proving Ground, MD, June . 1 9 7 8 ; ' , . Koger, D. M and McKown,G. ·L., Category 5 Suooressive Shield Test Report No. EM-TR-76001, Edgewood Arsenal, Aberdeen Proving Ground,.MD, October 1975. ' , Kusher, A. S., et al., Recommended Design and Analysis Procedures for Suppressive Shield Structures, Technical Report No. NSWCjWOL/TR76112, Naval Surface Weapons Center, White Oak Laboratory, White Oak,. Silver. Spring, MD, March 1977: Nelson, K. P. Spherical Shields for the Containment of Explosions, Edgewood Arsenal Technical Report No. EM-TR-76096, Edgewood Arsenal, Aberdeen Proving Ground, MD, March .1977. Nelson, K. P., The Economics of Applying Suppressive Shielding to the~ M483Al Improved ConventiomH Munitions Loading. Assembling, 'and Packing Facility, Technical Report No. EM-TR~76087, Edgewood Arsenal, Aberdeen Proving Ground, MD, January 1977. . Oertel, F. H., Evaluation'of Simple Models for the Attenuation of Shock Waves by Vented Plates ,. BRLReport No. 1906, U.. S. Army Ballistic Research Laboratory, Aberdeen Proving Ground, MD, August 1976. Pei, R., A Design Aid and Cost Estimate Model for Suppressive Shielding Structures, Department of.Safety Engineering USAMC Intern Training .. Center, Report No. YTC-02-'08- 76 -413, Red River Army Depot , Texarkana, TX, December ,1975. .., Ricchiazzi, A. J. and Barb, J .. C., Test Results for 608 Gram Fragments against Category I Suppressive Structures, BRL Memorandum Report No. 2592, U. S. Army.Ballistic Research Laboratory, Aberdeen' Proving 'Ground, MD, February 1976. Schroeder, F. J., et al., Engineering Design Guidelines, Drawings and Specifications for Support Engineering of Suppressive Shields, Edgewood Arsenal Contractor Report No. EM-CR-76097, Edgewood Arsenal, Aberdeen Proving Ground, MD, December 1976.··· Schumacher, R. N., 'Air Blast and Structural Response Testing of a Prototype Category III Suppressive'Shield, BRL Memorandum Report No. 2701; U. S., Army Ballistic' Research Laboratory, Aberdeen'Proving Ground, MD, November 1976. . Schumacher, R. N. and Ewing,'W. 0., Blast Attenuation Outside CubicalEnclosures Made Up of Selected Suppressive Structures Panel Configurations, BRLMemorandum Repor e. No ..2537·. U. S .. Army 'Ballistic Research Laboratory, Aberdeen Proving Ground, MD, September 1975. Schumacher,. R'. N., Kingery, C. N., and Ewing,' W. 0:, Air Blast and Structural Response' Testing of a 1/4~Scale Category I Suppressive' GC-3
TH 5-l300/NAVFAC P-397/AFR 88-22
50.
Shield, 'BRL Mem,?r.andum Report No. 2623. U. S. Army Ballistic Research Laboratory, Aberdeen Proving Ground, MD, May 1976., Spencer, A. F. and McKivrigan, J .. L., Preliminary Design Procedures for Suooressive Shields, Technical Report No. EM-TR-76089, Edgewood Arsenal, Aberdeen Proving Grpund, MD, December 1976~
51.
W'estine,. P.
52.
53.
S.'
and Baker, W..
K., Energy Solutions for Predicting
Deformations in Blast-Loaded Structures, Edgewood Arsenal Contractor Report No. EM-CR-76027, Report No.6, Edgewood Arsenal, Aberdeen Proving Ground, MD, November 1975. Westine, P. S. and Cox, P. A., Additional Energy Solutions for Predicting Structural Deformations, Edgewood Arsenal Contractor Report· No. EM-CR-76031, Report No.4, Edgewood Arsenal, Aberdeen Proving Ground, MD. November 1975. Westine, P. S. and Kineke, J .. H., "Prediction of Constrained Secondary Fragment Velocities," The Shock and Vibration Bulletin, Part 2, Isolation and Damping, Impact, Blast, Bulletin 48, September 1978. Blast Resistant Windows
54. 55. 56. 57. 58. 59.
60.
Federal Specification Glass., Plate. Sheet. Figured (Float, Flat. for Glazing. Corrugated. Mirrors and Other Uses). General Service Administration, Federal Specification DD-G-451d, Washington, D. C., 1977, Glass: Plate (Float). Sheet. Figured, and Spandrel (Heat Strengthened and Fully Tempered). General Service Administration, Federal Specification DD-G-1403B, Washington, D.. C., 1972. A Method for Improving the Shatter Resistance of Window Glass, U. S. Army Picatinny Arsenal, National Bomb Data. Center, General Info,mation Bulletin 73-9, Dover, NJ, November ..1973. , • PPG Glass Thickness Recommendations to Meet Architect's Specified 1.Minute Wind Load, PPG Lndus t r Les , Pi~tsburgh', PA, March 1981. Safety Performance Specifications and Methods of·Test for Safety Glazing Material Used in Buildings, American National Standards Institute, ANSI 297.1,-.1975, New York, NY, 1975. Structural Performance of Glass in Exterior Windows. Curtain Walls, and Doors under the Influence of Uniform Static'Loads by Destructive Method, American Society for Testing Materials, ASTM Standard (draft). Draft of Proposed Standard by ASTM Committee E06.51, Philadelphia, PA, October 1982. . _ Anians, D., Experimental Study of Edge Displacements of Laterally Loaded Window Glass Plates, Institute for Disaster Research, Texas Technical
61. 62. 63. 64. 65. 66.
University, Lubbock, TX, June 1980. Beason, W. L., A Failure Prediction Model for Window Glass, Texas Technical, University, NSF/RA 800231, Lubbock, TX, May 1980. Beason, W. L., TAMU Glass Failure Prediction Model. Preliminary Report, Texas.A!ili University, College 'Station, TX, March 1982. Beason, W. L... and 'Morgan, J. R., A Glass Failure Prediction Model, submitted·for publication in the:Journal of the Structural Division, American Society of Civil Engineers. Levy, S., Bending OD Rectangular· Plates with Large Deflections" NACA Technical Note No .. 845.1942. . . Meyers, G. E., Interim Design Procedure.for Blast-Hardened Window Panes, 567h Shock and Vibration Bulletin, Monterey, CA, October 1985. Meyers, G. E., A Review of Adaptable Methodology for Development of a Design Procedure for Blast Hardened Windows, Naval Civil Engineering. Laboratory, Special Report, Port Hueneme, CA, August 1982. 6C-4 "
•
TM 5-l300/NAVFAC P-397/AFR 88-22 67. 68. 69. 70.
71.
Moore, D:' M.,' Propo'sed Method for Determining the Thickness of Glass in Solar Collector Panels, Jet Propulsion Laboratory, Publication 8034', Pasadena, CA, March 1980. ' , Moore, D. M., Thickness Sizing of Glass Plates Subjected to 'Pressure Loads, FSA Task Report No. 5101-291. Pasadena, CA, August 1982. 'I'Lmo shenko , S. and Woinowsky-Kiiegei, S ... Theory of Pl'ates' and Shells, McGraw-Hill Book Company, New York, NY. 1959. . .J" Vallabha~, C. V. G. and Wang. B./Y., Nonlinear Analysi·t of Rectangular Glass Plates.by Finite Difference' Method, Texas Techni~al Un{versity, Institute for Disaster Research, Lubbock, TX, June ,1981. Weissman, S., et
a~ ..
'Blast Capacity Evaluation of 'Glass Windows and
Aluminum Window Frames', U. S. Army Armameiit 'Research and Development Command, ARLCO·CR-780~6, Dover, NJ, June 1978 .. Underground Structures' 72. 73.
Fundamentals of Protective Des"igr(for, Conventional Weapons, TM 5-8551;prepared by U. S~ 'Army Engineer'Waterways Experimental Station, Vicksburg, MS, November '1983' (Draft). .. ,.. Arya, et 'aI., Blast Capacity Evaluation of Bela'w-Ground Structures. "by' Ammann & Whitney, Consulting Engineers, New York, NY. Contractor Report ARLCO-CR-77006. U. S. Army Armament Research and Development'Command, Dover, NJ, May 1977. Eart~- Covered "Arch- Type .Hagaz ine
74. 75.
76. 77.
..,
,~
,
DoD Ammunition'and Explosives Safety Standards, Department of DefenseStandard, 6055.9-STD ' Flathou,'W. J .• et aI., Blast Loading and Response of Underground" Concrete-Arch Pr~tective Structures WT-1420. U. S. Waterways Experiment Statiori~,jackson, MS, Operation Plumbbob, Project 3.1, Chief, Defense' Atomic Support Agency', Washington; D. C. '. Sound, A: R., Summary Report of Earth-Covered, Steel'·Arch Magazine Tests, Technical Progress Report No. 401, U. S. Naval 'Weapon Center ... · China Lake, CA, July 1965. Weals: F. j H., Te'sts' to Deterihine' Separation Distances' of Earth- Covered Magazines, U. S. Nayal Weap0I! Cerrte r , China Lake, CA, Anna Lsj of' the New York Aca?emy of Sciences; Conference on Pr~vention of and Protection
against Acc'idental Explosion of Munitions; Fuels and Other Hazardous' Moistures, Volume 152, Art. '1, October 1968. '/
78.
79. 80.
Investigations Concerning' Feasibility of Vartous Designs for a Blast Closure Device, by American Machine and Foundry Co., Chicago, IL, under Contract No. NBy-13030, for U. S. Navy, Bureau of Yards and Docks, Washington, D. C. Shock Tube Test of Mosler Safe Company Blast Valve, BRL Information";· Memorandum No. 20, Explosion Kinetics Branch. Ballistic Research', Laboratories', 'Aberdeen Proving Ground, MD, August 1959',. Shock Tube Test of Mosler Safe Company Blast Valve, Phase II, BRL, Information Memorandum No. 25, Explosion Kinetics Branch, Ballistic
Research Laboratories, Aberdeen Proving Ground, MD, August 1959 . .
6C-5
1M 5-l300jNAVFAC P-397/AFR88-22
81. 82.
83. 84.
Study of Blast-Closure Devices, AFSWC-TDR-62-l0, by American Machine and Foundry Co; ,Chicago, IL for Air Force Special Weapons Center, Kirtland Air Force Base, NM, February 1962. ' Allen, F. C. et al., Test and Evaluation of Anti-Blast Valves for Protective Ventilating'Systems, WT'1460, Operation Plumbbob, Project 31.5, available from the Office of Technical Services, Department.of Commerce, Washington, D. C.' , , ' " Bayles, J., J., ,Development of the B-D Blast-Closure Valve, Technical Note No." N-546, U. S. Naval Civil Engineering Labo r a t ory , J::ort Hueneme, CA, 'December 1963. Bergman, S. and Staffors,' B., Blast Tests on Rapid-Closing Anti-Blast Valves, Royal Swedish Fortification Admiriistration, Stockholm, Sweden, November 1963.
85. 86. 87.
88.
. ,
•
-'
Breckenridge, T. A., Preliminary Development and Tests of a Blast-, Closure Valve, Technical Note No. N-460, U. S. Naval Civil Engineering Laboratory, Port Hueneme, CA, September 1962. Chapler, R: So., Evaluation of Four Blast Closure Valves, Technical Report R 347, DASA-13.l54, U. S. Naval Civil Engineering Laboratory, Port Hueneme, CA, January 1965 (Official Use Only). Cohen, E.,', Blast Vulnerability of Deep Underground Facilities as Affected by Access and Ventilation Openings, Ammann & Whitney, Consulting Engineers, New York, NY, Proceedings of the Second Protective ' Construction Symposium, R-34l, Volume I, The RAND Corporation, Santa Monica, CA, 1959. Cohen, E. and Weissman, S". Blast Closure Systems, Ammann & Whitney,
Consulting Engineers, New York, NY, Proceedings of the Symposium on Protective Structures for Civil Populations, Subcommittee on Protective Structures, Advisory Committee on Civil Defense, National Academy of Sciences and National Research Council, April 1965. '
"
M.
and Cohen, E., Review of Blast Closure Systems, ~ann
&
89.
Hassman,
90.
Whitney, Consulting Engineers, New. York, NY, 29th Symposium on Shock, Vibration and Associated Environments, Part III, Bulletin No. '29,. u. S. Naval Research Laboratory, Washington, D. July'196l, available from the Office of Technical Services, Department of, Commerce, Washington,'D. C. Hellberg, E. N., Performance of the Swedish Rapid-Closing Anti-Blast Valve, Technical Note No. 439, Task Y-F 008.10-11, U. S. Naval Civil Engineering Laboratory, Port Huenem~, CA, MaY,1962. Jones, W. A., e t; al., A Simple Blast Valve', Suffield Tecbnd c a I' Note No. 113, Suffield Experimental Station, Ralston, Alberta, Canada, Defense
91.
Research Board, Department of National Defense, Canada, DRB Project No.
D89-l6-0l-09, February 1963. 92.
.'
93. 94.
Ort, F. G. and Mears, M. D., Development of the Closure. Protective
Shelter, Anti-Blast, 600 cfm, E19Rl, Technical Report CWLR-2269, Chemical Warfare Laboratories, ~. S. Army Chemical Center, Edgewood, MD, January 1959. . Stephenson, J. M., Preliminary Tests of the Stephenson Valve, 2nd Report, Technical Note No. N-6l9. U. S. Naval Civil Engineering Laboratory, Port Hueneme, CA, July 1964. Stephenson, J, M., Test of German 'Sand-Type Filter, Technical Report R263, U. ·S. Naval Civil ~ngineering Laboratory, Port Hueneme, CA, November 1963. Shock Isolation Systems,
6C-6
-e
TK,5-1300/NAVFAC P-397/AFR 88-22 95. 96. 97. 98.
99. 100. 101. 102. 103. 104.
A Guide for the Design of Shock Isolation Systems for Underground Protective Structures AFSWC TDR 62-64, Air Force Special Weapons Center, Kirtland AFB, NM, December 1962. Handbook of Mechanical Spring Design, Associated Spring Corporation, Bristol, CT, 1964. Study of Shock Isolation for Hardened Structures, Department of the Army, Office of the Chief of Engineers, Washington, D. C. AD 639 303, June 1966. Crawford, R. E. Higgins, C~ J., and Bultmann; E.H., A Guide for the Design of Shock Isolation Systems for Ballistic Missile Defense Facilities, TR S-23, U.S.' Army Construction Engineering Research Laboratory, Champaign, IL; August 1973. Harris, C.M. and Crede, C.E., Shock and Vibration Handbook, McGraw-Hill Book Company, New York, NY, 1961. Hirsch, A.E., Man's Response to Shock Motions, David Taylor Model Basin Report 1797. Washington, D.C. AD 4 36809, January 1964. Platus, David L., et al., Investigation of Optimum Passive Shock Isolation Systems, AFWL-TR-72-l48, 'Air Force Weapons Laboratory, Kirtland Air Force Base, NM, November 1972. Saffell, H.R., Development of Standard Design Specifications and Techniques for Shock Isolation Systems, Document No. SAF-37, Vol. I, U.S. Army Engineer Division, Huntsville, AL, August 1971. Shigley, J.E., Mechanical Engineering Design, McGraw-Hill Book Company, New York, NY, 1963. . Veletnor, A.
I
Design Procedures for Shock Isolation System for Under-
ground Protective Structures, RTD-TDR-63-3096, Vol. III,.Air Force Weapons Laboratory, Kirtland Air Force Base, NM, January'1964.
6C-7
By Order of the Secretaries of the Army, the Navy and the Air Force:
Official:
CARL E. VUONO General, United States Army Chief of Staff
THOMAS F. SIKORA Brigadier General, United States Army The Adjutant General
Official:
D.E.BOTTORFF Rear Admiml, United States Navy .Commander, Naval Facilities Engineering Command
Official:
MERRILL A. McPEAK General, USAF Chief of Staff
EDWARD A. PARDINI, Colonel, USAF Director of Information Management
DISTRIBUTION:
e-
·U.S. Gowrnment Printing omce: 1993- 342-074172270