new required conditions in Eurocodes 3 and 4, practical tools for designers (Rotation capacities of profiles…) Ductility of plastic hinges in steel s...
Design and Optimization of Laminated Composite Materials - Z.Gurdal, R. Haftka, P.HaelaFull description
Design and Optimization of Laminated Composite Materials - Z.Gurdal, R. Haftka, P.HaelaDescription complète
steel design short cuts for Istructe
Design of Light Steel SectionsFull description
Full description
Cold Formed Thin Walled Sheet SteelFull description
Recently, the introduction of Pre Engineered Building PEB design of structures has been as an optimised alternative for the construction of steel sheds. The adoption of PEB design concept instead of use of traditional rolled section resulted in many
construction
:%\
European Commission Science Research Development
'ίϊ-ν'
technical steel research
Properties and in-service performance
Promotion of plastic design for steel and composite cross-sections: new required conditions in Eurocodes 3 and 4, practical tools for designers
Report EUR 18366 en
STEEL RESEARCH
EUROPEAN COMMISSION
Edith CRESSON, Member of the Commission responsible for research, innovation, education, training and youth DG XII/C.2 RTD actions: Industrial and materials technologies Materials and steel *V;
.
*'*
I::
Contact: Mr H. J.-L. Martin Address: European Commission, rue de la Loi 200 (MO 75 1/10), B-1 049 Brussels Tel. (32-2) 29-53453; fax (32-2) 29-65987
European Commission
ï%
Properties and in-service performance
Promotion of plastic design for steel and composite cross-sections: new required conditions in Eurocodes 3 and 4, practical tools for designers J. B. Schleich,
P.
Chantrain
ProfilARBED-Recherches 66, rue de Luxembourg L-4221 Esch-sur-Alzette
B. Chabrolin, Y. Galea, A. Bureau CTICM Domaine de St Paul BP1 F-78470 Saìnt-Rémy-lès-Chevreuse
J. Anza, E Espiga Ensidesa and Labein Cuesta de Olabeaga, 1 6 E-4801 3 Bilbao
Contract No 7210-SA/520/321/935 1
July 1993 to 30 June 1995
Final report
Directorate-General Science, Research and Development
1998
EUR 18366 en
LEGAL NOTICE Neither the European Commission nor any person acting on behalf of the Commission is responsible for the use which might be made of the following information.
A great deal of additional information on the European Union is available on the Internet. It can be accessed through the Europa server (http://europa.eu.int). Cataloguing data can be found at the end of this publication. Luxembourg: Office for Official Publications of the European Communities, 1 998 ISBN 92-828-4894-9
This Guide for plastic analysis has been elaborated in the scope of the ECSC research project entitled:
"Promotion of plastic design for steel and composite cross-sections: new required conditions in Eurocode 3 and 4, practical tools for designers (rotation capacities of profiles, ...)"
That international research project has been carried out
by the following partners:
ProfilARBED, Luxembourg: Mr. Philippe Chantrain, CTT.CM, France:
MM. Bruno Chabrolin, Yvan Galea, Alain Bureau,
ENSIDESA and LABEEN, Spain: MM. Juan Anza, Fernando Espiga, and with the following subcontractor:
RWTH - Aachen, Germany: Prof. Gerhard Sedlacek, Dr. Markus Feldmann. Duration of the project: 2 years (from 01.07.1993 to 30.06.1995).
This Guide for plastic analysis is the separate Annex aforementionned ECSC research project (Ref. 16).
13
or Part II of the final report Part I of the
Acknowledgements are addressed to the Commission of the European Community and to all members of the C.E.C. F6 Executive Committee "Steel Structures".
Many thanks are also due to all, who by any means may have contributed in this research and in the elaboration of this guide for plastic analysis, especially MM. Klòsak Maciej, Linh Cao Hoang, Conan Yves and Mauer Thierry.
Acknowledgements
This research project n° P3263 which has been sponsored by C.E.C., the Commission of the European Community, has been performed from 01.07.1993 to 30.06.1995 by the working group composed of : -
-
Profil ARBED (coordinator) CTICM (partner)
(C.E.C. Agreement 72 10-SA/520)
ENSIDESA-LABEIN (partner)
(C.E.C. Agreement 7210-SA/935)
(C.E.C. Agreement 7210-SA/321)
We want to acknowledge first all the financial support from the Commission of the European Community, as well as the moral support given during this research by all the members of the C.E.C. Executive Commitee F6 "Steel Structures".
Many thanks are also due to all, who by any means may have contributed in this research : ProfilARBED-Recherches RPS Department (Luxemburg):
MM. Chantrain Ph., Conan Y. and Mauer Th.,
MM. Klosak M. CTICM (France)
and
Linn Cao Hoang (as trainees),
:
MM. Chabrolin B., Galea Y. and Bureau Α., M. Mazuy F. (as trainee), -
ENSIDESA and LABEIN (Spain)
MM. Anza J. and Espiga F., -
RWTH- LfS (Germany)
:
M. Feldmann M., EPFL - ICOM (Switzerland)
M. Couchman G.
:
:
Table of Contents
5
List of Symbols
7
List of Figures
11
List of Annexes
13
References
15
1.
Introduction
17
1.1.
Obj ectives of the research proj ect
17
1.2.
Ways and means
17
1.3.
Final report
18
2.
Bibliography
19
3.
Generalities
19
4.
5.
3.1.
Ductility of plastic hinges: concept
19
3.2.
Influence of lateral restraint
22
3.3.
Guide for plastic analysis
25
3.4.
Safety evaluation for rotation capacities based on plastic hinge stability considerations
26
Available rotation capacity
33
4. 1 .
Generalities about available rotation capacity
33
4.2.
Test results
34
4.3.
Calculation models of available rotation capacity
34
4.4.
Numerical simulations of available rotation capacity
49
Required rotation capacity
67
5.1.
Introduction
67
5.2.
Influence of second order effects
67
5.3.
Influence of load level (design loads or collapse)
67
5.4.
Influences of loading path
68
5.5.
Continuous steel beams
70
5.6.
Frames
72
5.7.
Evaluation of required rotation capacity for composite beams
76
6.
Review of Eurocode 3 rules
79
7.
Conclusions
80
Annexes
83
Annex 2
91
Annex 3
101
Annex 4
121
Annex 5
147
Annex 6
195
Annex 7
207
Annex
8
215
Annex 9
219
Annex 10
229
Annex ll
265
Annex 12
313
Annex 13
379
Annex 14
381
Annex 15
395
Annex 16
407
LIST OF SYMBOLS Latin symbols A»
Atotal
A5.65
area of gross cross-section
percentage elongation after fracture on gauge length L0 = 5,65 JA¿ (where Aq is
the original cross-section area)
Ay
shear area of cross-section
Ayy
shear area of cross-section about y-y axis
A^
shear area of cross-section about z-z axis
AWeb
ttea. of the web
ASD
allowable stress design
b
flange width of H or I cross-section
CL
concentrated load
d
web depth of H or I cross-section
DL
distributed load
E
modulus of elasticity or Young Modulus
EGA
elastic global analysis
F
concentrated load
Fi
concentrated load (for which the 1st plastic hinge occurs)
Fu
ultimate concentrated load
fu
ultimate tensile strength
fy
yield strength
fy A
yield strength of the flange
fy w
yield strength of the web
h
overall depth of the cross-section
iz
radius
I
moment of inertia of cross-section
ly
moment of inertia of cross-section about y-y axis
Iz
moment of inertia of cross-section about z-z axis
kr
factor in formula
ku
factor in formula
L
system length
Lgq
equivalent length of member
Lj
distance between two adjacent lateral bracing
L0
gauge length of specimen for tensile coupon test
LTB
lateral-torsional buckling
M
bending moment
of H or I cross-section (= A - 2btf)
of gyration of cross-section about z-z axis
Mpi z
plastic moment capacity of cross-section about z-z axis (= Wpj z fy)
Mjd
design bending moment applied to the member
Ν
normal force; axial load
Nfl
specific term in formula
Npi
design plastic resistance of the gross cross-section (= A fy)
Nsd
design value of tensile force or compressive force applied to the member
Ρ
point load
Pc
specific term in formula
Ppl
plastic point load for which plastic hinge appears in 3 point bending beam
Pu
specific term in formula
PGA
rigid-plastic or elastic-plastic global analysis
q
distributed load
qι
distributed load for which the 1 st plastic hinge occurs
qu
ultimate distributed load
R
rotation capacity of plastic hinge
Rav
available rotation capacity of plastic hinge
Rreq
required rotation capacity of plastic hinge
rad
radian ( = unit for rotations;
1
radian =
180
degrees = 57,3 degrees)
π
tf
flange thickness of Η or I cross-section
tw
web thickness of Η or I cross-section
ULS
ultimate limit state
Vcr
total vertical load for elastic instability in a sway mode
Vp]
design shear plastic resistance of cross-section (=
Vga
design shear force applied to the member
Wei
elastic section modulus
We
external work done by the load
Wj
internal work absorbed by the structure
Wpi
plastic section modulus
Wpi y
plastic section modulus about y-y axis
Wpi z
plastic section modulus about z-z axis
y
major axis of H or I cross-section
ζ
minor axis of Η or I cross-section
Ay .fy λ/3 )
2.
Greek symbols
α
length factor (showing the position of point load)
Ym
partial safety factor
YmO
partial safety factor for bending moment resistance of cross-section
Ymr
partial safety factor for available rotation capacity Rav
7Μφ
partial safety factor for available inelastic rotation
δ
deflection of a beam
δρΐ
plastic deflection of a beam
δβι
elastic deflection of a beam
Δ
virtual displacement
ε
coefficient = j
£u
ultimate strain corresponding to fu
ey
yield strain corresponding to fy
η
length factor (of adjacent span in continuous beam)
φ
inelastic rotation of plastic hinge
characteristic value of available inelastic rotation of plastic hinge
design value of available inelastic rotation of plastic hinge
9b> 9c
rotation in perfect hinges obtained from an elastic analysis
9remaining
remaining inelastic rotation in plastic hinges with elastic return
required inelastic rotation of plastic hinge
9req.d
required inelastic rotation of plastic hinge at design loads
required inelastic rotation of plastic hinge at ultimate loads (related to structural collapse mechanism, instability,...)
maximal rotation of plastic hinge obtained by the intersection between decreasing part of experimental (Μ-φ) curves from 3-point bending tests and the plastic moment level Mpi of the profile
λ
load factor
λα-
critical load factor for elastic instability in a sway frame
λ"Γ
critical load factor for elastic instability in a sway frame with η hinges
X\jl
reduced slendemess of member according to lateral-torsional buckling
Arøax, Xy
maximal load factor (related to structural collapse : mechanism, instability,...)
λ λ^
collapse load factor for first order plastic global analysis
μ
factor in formula (=F/(qL))
ψ ω
bending moments ratio for a member (or parts of it) between lateral bracings
(with fy in N/mm^)
collapse load factor for second order plastic global analysis
ratio between distributed load on adjacent spans
:
3.
Drawing symbols
-O
perfect hinge
plastic hinge
t£7 y
fr
simple support (with vertical reaction)
simple support (with vertical and horizontal reactions)
fully fixed support
10
List of Figures Figure 1 : Moment rotation curve of 3-point bending beam Figure 2 : Moment rotation curve for different lateral restraints Figure 3 : Flow-chart for plastic global analysis Figure 4 : Parameters influencing the rotation capacity R of plastic hinges Figure 5 : Definition of critical slope Xoit in (Μ-φ) curve Figure 6 : (Μ-φ) curve for fully supported and uniformly loaded beam (A) Figure 7 : (Μ-φ) curve for fully supported and uniformly loaded beam (B) Figure 8 : (Μ-φ) curve for uniformly loaded continuous beam Figure 9 : Definition of required rotation
Xcrit in (Μ-φ) curve
Figure 12 : Plastic hinge stability related to negative slope in (Μ-φ) curves Figure 13 : Flow-chart : Safety evaluation for rotation capacity checking Figure 14 : Definition of parameters in (Μ - φ) curves Figure 15 : Influence of flange slendemess on aV (h = constant) (Feldmann's model) Figure 18 : Influence of steel grade on
11
Annexes "Complete set of distributed documents (10.95)"
(7 pages)
Annex 2:
Excerpts from Mr. Couchman's thesis (Ref. 14) "Design of continuous beams allowing for rotation capacity"
(9 pages)
Annex 3:
Document 3263-3-12 (LABEIN) "Conclusions from simulation results : Deformation buckling modes moment gradient & LTB restraints influence"
Document 3263-3-21 (LABEIN) (excerpts) "Feldmann's model checking within LTB limits"
(25 pages)
13
References Ref. 1:
Eurocode 3, ENV 1993-1-1, Design of Steel Structures, Part 1.1, General Rules and Rules for Buildings, CEN European pre-standard. ENV 1993-1-1 : 1992/A1, Annex D of Eurocode 3 (ENV 1993-1-1 [3]) : "The use of steel grades S460 and S420", CEN, December 1994.
Ref. 2:
Eurocode 4, ENV 1994-1-1, Design of Composite Steel and Concrete Structures, Part 1.1, General Rules and Rules for Buildings, CEN European pre-standard.
Ref. 3:
"Elasto-plastic behaviour of steel frameworks"; by J.C. Gérardy & J.B. Schleich, E.C.S.C. agreements 7210-SA/508; Draft of final report, ProfilARBED-Recherches, Luxembourg, August 1992.
Ref. 4:
"Elasto-plastic behaviour of metallic frameworks- Interaction between strength and ductility" ; by D'Haeyer R., Delooz M., Defoumy J.; ECSC agreement 7210-SA/204; Draft of final report 1992.
Ref. 5:
"Elastish-Plastisches verhalten von Stahlkonstruktionen, Anforderungen und Werkstoffkennwerte"; SedlacekG., Spangemacher R., Dahl W. und Langenberg P/.EGKS-F6 Projekt 7210-SA/l 13; Abschußbericht 1992.
Ref. 6:
"Promotion of plastic design for steel and composite cross-sections: new required conditions in Eurocodes 3 and 4, practical tools for designers (rotation capacities of profiles,...)", E.C.S.C. agreements 72 10-S A/520/32 1/935, Technical report n° 1, ProfilARBEDRecherches, Luxembourg, March 1994.
Ref. 7:
"Promotion of plastic design for steel and composite cross-sections: new required conditions in Eurocodes 3 and 4, practical tools for designers (rotation capacities of profiles,...)", E.C.S.C. agreements 72 10-SA/520/32 1/935, Technical report n° 2, ProfilARBEDRecherches, Luxembourg, September 1994.
Ref. 8
"Promotion of plastic design for steel and composite cross-sections: new required conditions in Eurocodes 3 and 4, practical tools for designers (rotation capacities of profiles,...)", E.C.S.C. agreements 72 10-S A/520/32 1/935, Technical report n° 3, ProfilARBED-
:
Recherches, Luxembourg, April 1995.
Ref. 9
"Promotion of plastic design for steel and composite cross-sections: new required conditions in Eurocodes 3 and 4, practical tools for designers (rotation capacities of profiles,...)", E.C.S.C. agreements 72 10-SA/520/32 1/935, Technical report n° 4, ProfilARBED-
:
Recherches, Luxembourg, September 1995.
Ref. 10
:
"Rotation Capacity of wide-flange beams under moment gradient", by Lukey A.F. and Adams P.R., Journal of the Structural Division, ASCE Vol. 95, n° ST 6, pp. 1173-1188, June 1969.
Ref.
11 :
RWTH Thesis of Mr. Feldmann M.
: "Zur Rotationskapazität von I-Profilen statisch und dynamisch belasteten Träger" (Aachen; Heft 30; 1994; ISSN 0722-1037).
"Experimentelle ermitûung der Rotationskapazität biegebeanspruchte I-Profile", by Roik K. and Kuhlmann U., Stahlbau 56, n° 12, December 1987, pp. 353-358.
Ref. 12
:
Ref. 13
:
Ref. 14
:
Ref. 15
:
"Available rotation capacity in steel and composite beams" by Kemp A.R. & Deckker N.W., The structural Engineer, volume 69, n° 5/5, March 1991.
Ref. 16
:
PEP-Micro Version 2.01, Plastic Analysis Computer Program, User's manual by Y. Galea, A. Bureau, CTICM, France.
Ref. 17
:
"Improved classification of steel and composite cross-sections: new rules for local buckling in Eurocodes 3 and 4", E.C.S.C. agreements 7210-SA/519/319/934, Draft of final report, ProfilARBED Recherches, Luxembourg, January 1996.
Ref. 18
:
Projekt P169 "Untersuchung der Auswirkungen unterschiedlicher Streckgrenzen -Verhältnisse auf das Rotations- und Bruchverhalten von I-Trägem"; von Sedlacek G., Spangemacher R., Dahl W., Hubo R. und Langenberg P.; Studiengesellschaft Stahlanwendung e.V-Forschung für die Praxis; 1992. EPFL Thesis n°1308 (1994) of Mr. Couchman G., Lausanne, EPFL, 1995 : "Design of continuous composite beams allowing for rotation capacity."
CM 66 - Additif 80 - DPU P22-701 (French code) : "Règles de calculs des constructions en acier". 15
Ref. 19
:
ECCS european recommendations on plastic global analysis of steel structures, R 7, 1976. Excerpt from Chapitre 13 - Calcul plastique des constructions - Volume I Structures dépendant d'un paramètre, 3eme édition, Massonnet Ch., Save M., 1976.
Ref. 20:
"Lateral stability of steel beams and columns - common cases of restraint", D.A. Nethercot, R.M. Lawson, The steel Construction Institute 1992.
Ref. 21
EN 10025 + Al : "Hot-rolled products of non-alloy structural steels - Technical delivery conditions (includes amendment Al : 1993)", CEN, March 1990 (EN 10095), August 1993.
Ref. 22
EN 10113 : "Hot-rolled products in weldable fine grain structural steels", Part 1,2,3, CEN, March 1993.
Ref. 23
"Capacité de rotation d'une section plastifiée : Extension du calcul plastique", Frédéric Mazuy, Mémoire de fin d'études. CUST - CTICM 1994.
Ref. 24
"Safety considerations of Annex J of Eurocode 3", M. Feldmann, G. Sedlacek, Third International Workshop on Connections in Steel Structures, AISC, ECCS, Trento, 1995.
16
1.
1.1
Introduction Objectives of the research project Presently in Eurocode 3 (design of steel structures) (Ref. 1) and in Eurocode 4 (design of composite steel and concrete structures) ÇRef. 2), the plastic analysis is governed by two criteria. The first one refers to mechanical characteristics of steel and the second one to the geometry of the used profiles. Previous researches on the rotation capacity for plastic analysis, performed by ARBED (Ref. 3), CRM (Ref. 4) and RWTH Aachen (Ref. 5) have allowed to understand the behaviour of plastic hinge formation for conventional steel. The results of these researches have shown that the requirements of Eurocodes 3 and 4 (b / 1 - ratios and its yield stress dependence, fu / fy > 1.2, £u > 20. £y, Aj· > 15 %) are very safe sided and could be substantially reviewed especially for high strength steels {S 460).
A new single criterion called rotation capacity allows to quantify the capacity of a profile to develop a plastic hinge without discrimination for high strength steels. This rotation capacity of plastic hinges includes a lot of parameters: geometrical and mechanical characteristics of the cross-sections, ductility, deformation range, local buckling problems, influence of welding procedures, influence of defects and, the analysed structure (geometry, loads). Nowadays the more competitive plastic design is not often used on the market because guides and practical tools, which evaluate available rotation capacities of cross-sections and required rotation capacities for different structures, are missing. The first aim of this research is to determine the available rotation capacities for composite crosssections (steel beam collaborating with a concrete slab) and for all the steel shapes included in sales programme and in function of different steel grades. The second aim is to determine the required rotation capacities for different types of structures. These both practical tools ensure a real promotion of plastic analysis because the designer will be able to know easily and quickly if a plastic analysis is possible by comparing the required rotation capacity fornis structure with the available rotation capacities of the used cross-sections.
The final aim of this research is to introduce these new rules of plastic analysis for steel and composite cross-sections in Eurocode 3 and Eurocode 4, with the support of expertises.
In such
a way the competitivity of steel and composite cross-sections will be improved and with this advantage their market share will increase substantially.
1.2
Ways and means
(1) The following financially independent partners participated in the research project:
ProfilARBED - Recherches, Luxembourg : Mr. Chantrain Ph. CTICM, France : MM. Chabrolin B., Galea Y., Bureau A.
LABEIN and ENSIDESA, Spain : MM. Anza J., Espiga F. (2) The technical co-ordination was handled by ProfilARBED - Recherches Department "Recherches et Promotion technique Structure (RPS)". It was decided that only one common ECSC report had to be written by ProfilARBED for each period. Each report included the contributions done by different partners during the different four research periods (Ref. 6, Ref. 7, Ref. 8, Ref. 9).
(3) During this research project, the main works were distributed between partners as follows:
-ProfilARBED:
. .
management of the project, comparison of existing design models evaluating available rotation capacities of plastic hinges (see chapter 4.3),
17
development and safety evaluation of models about available inelastic rotations of plastic hinges (with RWTH Aachen University as subcontractor) (see Annex 12), development of chapter 1 (Generalities) and chapter 3 (Available inelastic rotations of plastic hinges) of the "Guide for plastic analysis" (see chapters 3.1, 3.2, 4.1, 4.2 and Annex 13), numerical simulations of continuous beams and frames to evaluate required rotation capacity of plastic hinges (see chapters 5.5 and 5.6 ), development of chapter 2 O^equired inelastic rotations of plastic hinges) and chapter 4 (Design examples) of the "Guide for plastic analysis" (see chapter 5 and Annex 13),
CTICM:
- LABEIN
:
safety evaluation of rotation capacities of plastic hinges (see chapter 3.4), numerical simulations of tests results (see chapter 4.4.1), numerical simulations about nominal cases (see chapter 4.4.2), numerical simulations about beams submitted to My - Ν loading (see chapter 4.4.3), numerical simulations of LTB limits for Feldmann's model (see chapter 4.4.4), numerical simulations of double clamped beams (see chapter 4.4.5), numerical simulations about influence of lateral restraints (see chapter 4.4.6), numerical simulations about equivalent length evaluation for continuous beams (see chapter 4.4.7).
1.3
Final report The present final report compiles all results of works done in the scope of this research project. This final report presents :
in chapter 2, bibliography, the concept of plastic hinges ductility, the influence of lateral restraints, . the guide of plastic analysis, . the safety evaluation for rotation capacities based on plastic hinge stability considerations, in chapter 4, available rotation capacity of plastic hinge (generalities, comparison of design models, Feldmann's model, numerical simulations) ,
in chapter 3, generalities about
.
.
in chapter 5, required rotation capacity of plastic hinge (introduction, influence effects, load level and loading path, continuous beams, fiâmes),
in chapter 6, review of Eurocode 3 rules, and in chapter 7, conclusions and proposal for future researches.
18
of second order
Bibliography
2.
(1) Collection of information according to Eurocode 3 (Ref. 1) and Eurocode 4 (Ref. 2) has been performed by all partners : bibliography, technical reports, papers, results from tests, statistical evaluations, conclusions of previous or in progress researches, existing rules or new proposals of rules, development of calculation models,....
(2) For convenience a specific numbering has been introduced for the documents distributed in the scope of this research project. The list of the numbered documents distributed up to December 95 is given in Annex 1. The convention ofthat numbering is proposed as follows (for example 3263-1-10) :
"number of the project" ("3263"),
"number of the partner" ("1", "2" or "3" respectively related to ProfilARBED, CTICM or LABEIN), "number of the paper in the chronological order of distribution".
Generalities
3.
Ductility of plastic hinges: concept
3.1
(1) The plastic hinge method may be used for the ultimate limit state design of steel structures subject to static loading.
By this method plastic zones and zones with local buckling are modelled by plastic hinges which exhibit a simplified bilinear moment-rotation-characteristic with unlimited rotation capacities on the level Mpi (plastic moment resistance of cross-section). Hence it is necessary to verify by a rotation assessment that the rotation requirement resulting from the moment redistribution at ultimate limit state does not exceed the actual available rotation capacity. (2) Using plastic analysis, required inelastic rotation
9req.d where
γΜφ
-
9av.d
,
9av
9av.d=JLaï-
with
is a partial safety factor to allow for the uncertainties.
It can be shown that the available inelastic rotation parameters
(3.1)
ΎΜφ
of plastic hinge depends only on local
:
-
material properties (yield strength, ultimate strength, ... )
-
shape and dimensions of the cross-section
-
internal forces at the location of the plastic hinge
This criterion (3. 1) is sufficient in so far as the available inelastic rotation
19
(3) A great number of authors have preferred to talk about "rotation capacity" (R) instead of "inelastic rotation" (
Rav _9; av
9pl
where
_ Mpl.L 2.E.I
with
9pl =
L
is the length of the beam
is the sum of the elastic rotations (determined at Mpi level) at the ends of the beam.
Mpl
is the plastic resistance moment of the cross-section
E
is the modulus of elasticity of steel
I
is the moment of inertia of the cross-section
M
Φ/2
lF
i
i
*'2
i
Mpi-
9av /
Tav
L
j '
!
9rot
9pl
Figure
1 :
9
Moment rotation curve of 3-point bending beam
So, the validity and the consistency of a plastic analysis can also be checked in all cross-sections by the following limit states criterion for ductility in bending which is equivalent to criterion (3.1):
ρ
Kreq
^
<üay_
s
(3.2)
YMR
(4) The concept of rotation capacity R introduces a supplementary parameter such as the length L of a reference beam. It must be noted that this parameter is a structural and not local parameter. Rreq is to be calculated from frame plastic analysis as explained in chapter 3 of Annex 10.
The available rotation capacity Rav can also be given in tables but it must refer to a particular beam length because (ppi depends on the length. Because the verification has to be consistent, the required rotation capacity Rreq must be computed by considering the same length of reference which will be used to determine Rav.
As explained in chapter 4.4.7.1, difficulties may appear to clearly identify the length of reference in order to calculate
(
(5) The proposed method which compares required and available inelastic rotations (φ) or rotation capacities (R) for each relevant plastic hinges, is an alternative to the use of width / thickness limits (rules for classification of cross-sections) existing in Eurocodes 3 and 4 (Ref. 1 and 2) , for the verification of sufficient ductility of plastic hinges. Eurocode 3 Q^ef. 1) provides general rules fnceming rotation requirements of plastic hinges: " 5.3.3 Cross-section requirements for plastic global analysis (...) (2) At plastic hinge locations, the cross-section of the member which contains the plastic hinge shall have a rotation capacity of not less than the required rotation at that plastic hinge location.
(3)
To satisfy the above requirement, the required rotations should be determined from a rotation analysis.
(4)
For building structures in which the required rotations are not calculated, all members containing plastic hinges shall have class
1
cross-sections at the plastic hinge location."
(6) A general flow-chart is proposed in Figure 3 defining the scope of the research project which concerns the evaluation of the ductility of plastic hinges. (7) A table of parameters influencing the rotation capacities of plastic hinges, is provided in Figure 4: required rotation capacity ^required (concerning structures) and available rotation capacity ^available (concerning cross-sections). More details will be provided in respective chapters 5 and 4. (8) A specific design method has been proposed by Mr. Couchman G. (Ref. 14) for continuous composite beams (see Annex 2 (1/9 and 2/9)).
That design method based on the idea of rotation capacity
:
includes the influences of all relevant parameters, is applicable to beams with plastic (Class 1), compact (Class 2) or semi-compact (Class 3) sections,
gives an uniform margin of safety for all cases, is suitable for everyday use by practising engineers.
That design method allows considerable increases in beam load capacity for beams with compact or semi-compact critical sections. Details concerning available and required rotation capacities of composite beams, are given respectively in chapters 4.3.5 and 5.7.
At present state, more developments (tests results, numerical simulations, statistical evaluations; ...) are necessary to exploit that method for continuous composite beams in order to elaborate design aids or charts which will help designers in their daily works.
21
3.2
Influence of lateral restraint
(1) In order to realize the necessary inelastic rotations at plastic hinge locations a member must have sufficient lateral restraints to ensure that the plastic moment at these hinges locations is not reduced by lateral-torsional buckling (LTB) before a mechanism has formed. Local buckling is controlled by limiting the width/thickness ratios of the flanges and the web; lateral-torsional buckling should be controlled by limiting the unbraced length of the member. (2) The rotation capacity depends largely upon its unbraced length on either side of a plastic hinge. Based on tests results (Ref. 10) and on numerical simulations (see chapter 4.4.6), Figure 2 shows, qualitatively, for different L/iz ratios, the moment-rotation relationship of a beam under uniform moment As L/iz decreases, the rotation capacity increases. In order to maintain the plastic moment and provide adequate rotation capacity, the L/iz ratio must be controlled.
Figure 2 : Moment rotation curve for different lateral restraints where
L
is the member span,
iz
is the radius of gyration about minor axis of the member.
In general, lateral-torsional buckling shall be avoided in plastic analysis. (3) On the basis of numerical simulations chapter 4.4.6 presents some conclusions about the combined influence of the moment gradient and the lateral restraints on the resulting available rotation capacity .
22
Flow-chart for Plastic Global Analysis Type of frame (geometry, boundary conditions, loading, yield strength) rigid
J
R required (structure)
4
i
/
assumption :\
semi-rigid
X^type of joints^
Incase of simple frames => tables to evaluate R required
I plastic hinges in columns: (M, N) interaction ι (M, N, V) interaction
J
ι
±
>
ω
Plastic hinges in joints
I
R required is deduced from tables (or results of plastic global analysis) and, R available is issued from tables or formulas: i) ! R required < ( R available / γ£) \
τ
Τ
Plastic global analysis of the frame considering the global and local stability (geometry of the frame, boundary conditions, loading, yield strength) => maximum load multiplicator of design loads: Xmax
u
s
ce (Λ
'ig a
I
«fa
O JS
Sufficient bearing capacity of the frame if: >1
e£
.2 o J3
"rt >
yes^ R required is deduced from results of plastic global analysis and, R available is issued from tables or formulas: not fuelled R required < ( R available / γ *)
Notes : R = rotation capacity of plastic hinges / M = bending moment/ Ν = axial load /V = shear force 1)
= tables of Ravailable
: . .
ok if no lateral-torsional buckling (LTB) (if rules of maximum spacing of lateral supports; ...) difficult if influence of LTB
Figure 3 : Flow-chart for plastic global analysis 23
2.2. cross-section slendemess: for web and flange of steel profile: d/tw and cAf 2.3. concrete slab reinforcement (amount of rebars): ratio of support to span plastic moment resistance, Mp/Mp
X?
2.4. degree of steel-concrete shear connection, N/Nf
Figure 4 : Parameters influencing the rotation capacity R of plastic hinges
24
no no
(4) Presently lateral-torsional buckling is not explicitly forbidden by Eurocode 3 (Ref. 1) but it is related to strength and limited as follows: if the relative slendemess λτ^χ > 0,4, then the ultimate bending moment resistance of the beam (= XLT-Mpi) is lower than the plastic moment resistance of class 1 and 2 cross-sections (= Mpi) because the reduction factor %lt is lower than 1,0. The factor Xjjr includes the unbraced length of the member (ή_τ). In order to respect the process of plastic analysis which excludes lateral-torsional buckling as regards
strength of frames and rotation capacity of plastic hinges, the Eurocode 3 condition
Xlt ^0,4
shall be
fulfilled in all cases. (5) On the other hand Eurocode 3 (Ref. 1) give some general remarks on lateral restraint
:
"5.2. 1 .4 (3)
When plastic global analysis is used, lateral restraint shall be provided at all plastic hinge locations at which plastic hinge rotation may occur under any load case.
5.2. 1 .4 (4)
The restraint should be provided within a distance along the member from the theoretical plastic hinge location not exceeding half the depth of the member."
(6) In alternative to Eurocode 3 rule (Xlt), other existing rules defining maximum distances between lateral restraints should be provided to designers : for instance rules from CM66-Additif 80 (French code) (Ref. 18), from ECCS recommendations (Ref. 19) or from SCI publication (Ref. 20).
3.3
Guide for plastic analysis As expected in the scope of this research project, a practical tool has been elaborated to help designers concerned by plastic global analysis of steel structures. Softwares already exist to determine the resistance of steel structures according to plastic global analysis, but there is a missing link for complete designs at ultimate limit states : a simple tool for the evaluation of ductility of steel plastic hinges. This gap is filled by the document entitled Ductility of plastic hinges in steel structures - Guide for plastic analysis", (see Annex 13) which has been developed by ProfilARBED for chapters 1 and 3 and by CTICM for chapters 2 and 4. This document focus on inelastic rotation φ of steel plastic hinge and not on rotation capacity R, as explained in chapter 4.4.7.1.
This stand-alone document of Annex 13 ("Ductility of plastic hinges in steel structures - Guide for plastic analysis") contains the following chapters : Chapter 1 : "Generalities" ; this chapter 1 presents : the concept of plastic analysis, the concept of inelastic rotation of plastic hinges, the influence of lateral-restraint, the parameters influencing the inelastic rotation, the design rules for ductility of plastic hinge , the concept of plastic analysis based on inelastic rotation.
Chapter 2 : "Required inelastic rotation of plastic hinges in structures" ; this chapter 2 details the influence of parameters on required inelastic rotation, the cases of continuous beams, the cases of frames.
Chapter 3 presents
:
"Available inelastic rotation of plastic hinges in cross-sections"
;
:
this chapter
3
:
Feldmann's model : assumptions, limitations, formulas for different load cases, extension to other load cases, the influence of parameters on inelastic available rotation
:
step-by-step method with the help of an elastic analysis program.
Appendix 1: "Plastic resistance of I and H cross-sections"; this Appendix gives tables with plastic resistance values for I and H hot-rolled sections (IPE, IPEA, IPEO, HEAA, HEA, HEB, HEM, UB and UC) and for different steel grades (S 235, S 275, S 355, S 420 and S 460).
3.4
Safety evaluation for rotation capacities based on plastic hinge stability considerations
3.4.1
Introduction
This section presents a safety evaluation approach for rotation capacities based on considering the stability requirements of the plastic hinge expressed in terms of the required rotation. This safety evaluation method presented in this chapter 3.4 is provided only for information because our working group selected Feldmann's model to determine
An assessment of the rotation capacity by checking the fulfilment of the following condition is assumed:
«*% where
7c
(3.3)
is the partial safety factor evaluated by the standardized statistical procedure for test results outlined on Annex Ζ of Eurocode 3 (see Annex 12).
An improved safety factor is proposed on the basis of calculating an additional value for the factor taking into account the criteria for the plastic hinge stability:
<9av
9pi^
(3.4)
y = max(yc,ys)
(3.5)
where:
and Ys is the partial safety factor to be evaluated from considerations about the requirements stability of the plastic hinge.
for the
The rotation capacity approach used in plastic design assumes implicitly that a given required rotation can be reached by the plastic hinge provided that, in the moment-rotation curve, the value of the moment remains over the plastic moment level (available rotation definition). However, this condition does not guarantee the plastic hinge stability in the structure since this stability depends not only on the moment level corresponding to the required rotation but also on the descendant branch slope of the moment-rotation curve at that rotation.
A value for the critical slope of the descendant branch, Xcnt, will be derived and expressed in terms of the required rotation and the load redistribution factor. On the other hand, the actual value of the slope for the moment rotation curve, at the point where the available rotation is reached, will be expressed, assuming certain simplifications, as a function of the available rotation. This will allow the checking of the plastic hinge stability (X < XCrit> to be integrated as part of the rotation capacity checking by means of the introduction of a modified safety factor as shown in (3.4) and (3.5).
26
3.4.2
Stability requirements
3.4.2.1 Critical slope in the moment-rotation curve The stability of the plastic hinge when reaching a given value of the rotation is not lost provided that the following condition is fulfilled:
X
(3.6)
where X is the actual value of the slope of the moment rotation curve and Xcrit is the critical slope to be evaluated. Xcrit is mainly dependent on the structure and the load conditions and will be expressed in terms of the required rotation capacity and the load redistribution factor.
Mi k Mpi.
kf
^^^\ Β k ^ Χ<Χ~»STABILITY
^p X>Xc~»INSTABILITY «-3
**
>*
φ
Kp
Figure 5 : Definition of critical slope Xcrit m (Μ-φ) curve 3.4.2.2 Stability requirements: built-in beam uniformly loaded
In the first stage (A), the load Ρ is incremented until the moment at the end locations, Mi, reaches the plastic moment level, Mpi .
Figure 6 : (Μ-φ) curve for fully supported and uniformly loaded beam (A) The load Ρ and the end moment Mi at the plastic hinge formation moment are: 12. M pi
From this point on, the beam ends can be assumed to behave as elastic supports with a variable stiff¬ ness, K(
27
λΡ
Krø
.W
αΜ,Α
Λ
AR
ν
*?
Μ>χ
+
AR (B-g
δΜ,
aP=aR+/P, αμ,= Κ Δ*
is
(t
Figure 7 : (Μ-φ) curve for fully supported and uniformly loaded beam (B) Thus, an expression for the angle Δφ ι in function of the total load ΔΡ can be derived:
(βΤΓ) δΦι =
ϋ
(
24. E. I
f¿\
(βΤπ) δμι =
.ΔΡ,
Δφ,= .fl + - ^-.ΚίφΟί Ψ1 24.Ε.Ι I 2.Ε.Ι m;J
.ΔΡ2
.ΔΡ
κ12;
The plastic hinge will start to loose the stability at the moment in which an increment in the angle may occur without applying any external force:
Δφι <0 ΔΡ
-* K<
-2.E.I L
X>
E.I
3.4.2.3 Stability requirements: continuous beams The plastic hinge instability condition derived in the previous paragraph can be generalised for the continuous beam case following an analogue procedure.
When the plastic moment level (Mpi) is reached, the plastic hinge can be assumed to behave as a tor¬ sional elastic spring with a variable stiffness, Κ(φι) determined by the slope of the moment-rotation curve. The new case can be solved by superposition of two cases as indicated: the first one introducing a simple support at the plastic hinge location (B-I) and the second keeping the original boundary condi¬ tions (B-II). In addition, the compatibility relationship between the angle Δφι and the moment ΔΜχ needs to be applied. Thus, an expression for the angle Δφι in function of the total load ΔΡ can be derived:
The plastic hinge will start to loose the stability at the moment in which an increment in the angle may occur without applying any external force:
ΔΡ
Δφι
<0
x>
c9
Cm 2.C,
Next, this critical value for the slope in the moment-rotation curve will be expressed as a function of the required rotation. Let (preq the required rotation at the plastic hinge to be checked. As the required rotation is calculated under the hypothesis of elastic perfect-plastic global analysis, the moment incre¬ ment between the points A and Β is assumed to be null.
k.
M1
Β
Mp,
[
ì ψ
*
-*
φ »to Kp
Figure 9 : Definition of required rotation (freq in function of slope in (Μ-φ) curve
AM¿ = 0 -4 ΔΡ2|Α = 0 -> APJJ = ΔΡ)» φΓβς = and the increment
following way:
,Β
2.Δφ1|Α =2.<:φ.ΔΡ
of the external load can be ΔΡ| A = Kp . Ρ
-> Δφι|* = εφ.ΔΡ||
Β
expressed in terms
of a load redistribution factor in the
where Kp = load redistribution factor
being Ρ the value of the external load at the moment of reaching first Mpi (A):
® 9req
Mpl = CM.P
2.C
~~
Kp·"
1
CM
Mpl=Rreq-9pl 29
Therefore, the critical slope becomes:
γ
Amt Lcnt
_ CM _ KP (Pjgq
2.C
Rreq 9pl
P
3.4.2.4 Plastic hinge stability condition: example The built-in beam case is taken for evaluating the critical slope value to be considered in the checking of the stability of the plastic hinge.
Figure 10 : Example of fully supported and uniformly loaded beam The calculation of the required rotation capacity Rreq and the load redistribution factor Kp follows:
RU
Mi =
12
M2 =
= MD,-»P =
1
12.M pi
pl
and evaluating the critical slope value
Lz
.ΔΡ =
ΔΡ =
MDl
Sí-
2
4.Μηι
^->Κη=
~~
1»58
Kp = 0,33
1
->X£i,= Lcnt
ΔΡ
= 0,33
1,58
Rreq
'Lpl = 1-
^
9Ρ1
V3 for the moment-rotation curve as previously described:
Figure 11: Definition of critical slope Xcrit 'n (Μ-φ) curve Rreq
->
Mpl.L
Mpl.Lpl Ε. Ι
ΔΜ2 =
8
P.L2 _ M mpi 24 1
_ 9pl =
·
-2- = 0,21 Teq
30
3.4.3
Plastic hinge stability checking by the safety factor approach
An expression of the actual slope of the moment-rotation curve at the point of crossing down the Mpi level as a function of the available rotation, X = X(
***? y = max(yc,Ys) where Yc is the partial safety factor derived from statistical evaluations on test results and ys is the partial safety factor to be evaluated from the stability criteria
A number of assumptions will be considered about the moment-rotation curve in order to derive the required relationship for the slope: 1.
The ultimate moment is taken to be the plastic moment multiplied by a constant β.
2.
The available rotation is dependent on the rotation value at the ultimate moment level through the constant Kav.
3.
The slope of the curve at (fret is taken to be two times the slope corresponding to a descendant straight line between (
Figure 12 : Plastic hinge stability related to negative slope in (Μ-φ) curves
M
9av~Kav9u
(1+β)
pl
Taking into account these relations the plastic hinge stability condition can be expressed in the follow¬ ing way:
L- (Kav-l).
χ
.
-r-rit Lcnt
_ CM _ KP .JV1 M 9req
2-C
x
ent
2.Kav
(Kav-!)
pl
IL ^< K "P£ 9av
where: y
2.K a*av
^
(Kav-l) Kp 31
9req
>
9av > 9req Ys
Taking the following realistic values for the parameters involved in the formula:
Kav =2
β Kp a safety
3.4.4
=0,2 =1/3
factor yc = 2,4 is obtained from the stability requirement
Flow chart: safety evaluation for rotation capacity checking based on the plastic hinge stability approach
Available rotation capacity Generalities about available rotation capacity Definition of rotation capacity: The "three-point bending beam" model, which is investigated, consists of a simply supported beam with a concentrated load at mid-span.
As the load increases, a plastic hinge appears under the load; the plastic hinge is usually accompanied by local buckling of the compressed parts of the cross-section. The bending moment M in the hinge is plotted versus the sum of the two rotations at supports. Then the available rotation capacity ^available CSJ1 be deduced.
Referring to the following graph, the commonly accepted definition of available rotation capacity ^available (Rav) is given (with different designations) as :
M.max
Test curve
Mpi Elasto-plastic bilinear model
= θε+θ; inelastic
=
φ θ
Figure 14 : Definition of parameters in (Μ - φ) curves R
_9av
Kav -
9pl
or
p
_ 9rot
vav
9pl _ 9rot _ elastic
9pl
9pl
öe
where
M
is the bending moment in the plastic hinge,
Mpi
is the theoretical plastic resistance bending moment of the cross-section, 33
φρί (= θβ = eelâstic) is defined as the particular elastic rotation φ (= level (upwards intersection), (Prot
is defined as the largest rotation (downwards intersection) and,
θ) related to the theoretical Mpi
φ (= θ) related to the theoretical Mpi
level
9av (= 9rot = ^inelastic = θρ = öav) is the available inelastic rotation of the plastic hinge,
φ (= θ)
is a rotation defined as the sum of the rotations of both extremities of the beam; φ = φι + φ2 (see Figure 14), in the case of three-point bending tests .
Tests results
4.2
(1) The following 90 available 3 point bending tests results according to major axis yy are provided in Annex 1 1. Those tests were used to compare Feldmann's and Kemp's models (see chapter 4.3) : 15 tests
from Lukey and Adams (USA) (Ref. 10),
20 tests from Roik and Kuhlmann (Bochum, Germany) (Ref. 12), 26 teste from Sedlacek (RWTH Aachen, Germany) (Ref. 5 and 13), 29 tests from CRM (Liège, Belgium) (Ref. 4),
hi the tables of Annex steel grade)
9rot}
11
the models values are calculated with measured characteristics (geometry,
(2) In Annex 12 a more complete database of tests results is presented and used in order to evaluate partial safety factors γΜφ for Feldmann's formulas related to different load cases (My, Mz, N-My).
4.3
Calculation models of available rotation capacity
4.3.1
Feldmann's model for steel sections
4.3.1.1 Generalities about Feldmann's model (1) As presented in chapter 3.1, the validity and the consistency of a plastic analysis of a frame under a given loading can be checked in all cross-sections by evaluating the ductility of plastic hinges either (3.1) with plastic rotation, or (3.2) with rotation capacity : 9req.d
or,
r>
^ 9av.c[
^
Rav
,
,
with
where
and
in reference to the 3-point bending system. 34
9av.d
-
(3.1)
ΪΜφ
_9av
p av
~ m 9pl MpiL
φ* = 2EI
(3.2)
(2) Such a rotation check so far could not be carried out because the rotation capacity could only be determined from tests or sophisticated numerical simulations. The background document to Eurocode 3 which is enclosed in Annex 12 presents a simple procedure developed by Feldmann M. (Ref. 11) for calculating the available inelastic rotation
(4) ProfilARBED applied Feldmann's formulae for I and H hot-rolled cross-sections submitted to bending about major axis My, to bending about minor axis Mz and to combined axial compression and bending about major axis N-My . The "Guide for plastic analysis" included in separate Annex 13 provides practical tables and graphs with characteristic values of available inelastic rotations
(5) The values of inelastic rotations of plastic hinges (tøv are mainly influenced by the dimensions of the cross-sections as well as by the steel grades of web and flange (see Figures 15 to 18). Rotation capacity Rav is dependent on (j)av but is also function of φΡι related to the span length Lspän (see Figures 19 and 20).
0,2
τ
0,18
,;
[rad]
0,16 -·
0,14 -. 0,12 -0,1
-.
0,08 -.
fy = 320 MPa h = 027 m
..
b = 0,135 m tw = 0,0066 m
0,06
0,04 .. 0,02 ..
bAf
0 --
-4-
8
10
12
14
16
18
20
22
24
26
Figure 15 : Influence of flange slendemess on φ3ν (b = constant) (Feldmann's model)
35
fy = 320 MPa h = 0,27 m b = 0,102 m tw = 0,0066 m
0,13
0,12 -.
0,11
--
0,1
12
24
20
16
28
32
Figure 16 : Influence of flange slendemess on
φ av[rad] 0,18
-,
0,16
-
0,14
-
0,12 0,1 -
0,08
-
0,06
-
0,04 0,02
-
0
15
fy = 320 MPa h = 0,27 m b = 0,135 m tw = 0,0102 m
dAw 20
25
30
Figure 17 : Influence of web slendemess on
36
(pav (h =
35
40
45
constant) (Feldmann's model)
0,2 Tav 0,18
-
0,16
-
0,14
-
IPE 270
0,12 . 0,1
-
0,08 0,06
0,04
-
0,02
-
0
-
235
fy [MPa] i
275
315
355
Figure 18 : Influence of steel grade on
395
q>av
435
(Feldmann's model)
Figure 19 : Influence of steel grade on Rav (Feldmann's model)
37
475
20 τ
16
--
ΓΡΕ270 12
--
8
-.
4 -.
L span [m] 0
-i
1
0
I
I 1
1
I
1
1
1
Η
lili
I
I
I
Η
I
I
I
I
I
I
2
Figure 20 : Influence of span length on Rav (Feldmann's model)
(6) Feldmann's model (see Annex 11) delivers always positive values of «fev and Rav for all considered profiles and steel grades, because no limits are imposed to the model. In the "Guide for plastic analysis" (chapter 3 of separate Annex 13) tables and graphs which provide available inelastic rotation
-
comparison with tests results (see chapter 4.4. 1), simulation of normal cases without tests results (see chapter 4.4.2), check of formulas (4.3) and (4.4) for beams submitted to combined My - Ν loading
-
check of lateral-torsional buckling limits for Feldmann's model (see chapter 4.4.4).
-
:
(see chapter 4.4.3),
(9) Partial safety factor:
In the scope of Feldmann's model, the values of partial safety factor γΜφ have been determined by statistical evaluation of available tests results (see Annex 12). The values of partial safety factor ytø
the design value of plastic bending moment for the resistance of cross-sections : MplJld = Mpi/TMo . with partial safety factor ymo according to Eurocode 3 (Ref. 1) ( Ymo = 1,1) or to NAD's values, and, the design value of available inelastic rotation for ductility of formed plastic hinges : 9av.d = 9av / ΤΜφ . with the proposed design model in chapter chapter 4.3.1.2 for
Formulas
Load cases
Ύίνΐφ
of chapter 4.3.1.2
1,52
j My
:
Bending about yy major axis
(4.1) & (4.3)
1,73
j Mz
:
Bending about zz minor axis
(4.2)
Bending about yy major axis combined with axial compressive force
(4.4)
^-y
2,02
My- Ν
:
(10) Steel grades:
All steel grades available in Eurocode 3 Part 1.1
can be applied in the formulas
:
235 MPa < fy < 460 MPa. Indeed the reliability of the model has been demonstrated by tests results with steel grades m that range (see Annex 12).
Eurocode 3 (Ref. 1) allows for plastic global analysis with steel materials characterized according to EN 10025 (Ref. 21) and EN 101 13 (Ref. 22) as given in following table :
Thickness t (mm) *)
Nominal Steel grade
t<40mm
40mm
EN 10025
fy
fu
fu/fy
A5.65
standard
[Mpa]
[Mpa]
[-]
[%]
235 275 355
360 430 510
1.53 1.56 1.44
26 22 22
390
1.42 1.38 1.19 1.15
24 22
S 235 S 275 S 355
£y
fu/fy A5.65
fy
fu
[Mpa]
[Mpa]
[-]
[%]
[%]
0.11 0.13 0.17
215 255 335
340 410 490
1.58 1.61 1.46
24 20 20
0.10 0.12 0.16
0.13 0.17 0.20 0.22
255 335 390
370 470 500 530
1.45 1.40 1.28 1.23
24 22
0.12 0.16 0.19 0.20
ey
EN 10113-3 Standard S 275 M S 355 M S 420 M S 460 M
275 355
420 460
490 500 530
19 17
Note:
*) t is the nominal thickness of the element : - of the flange of rolled sections (t = tf ), in general, or - of the particular elements of welded sections
39
430
19 17
^7^ * 1
J,
t
1
t
^ -» Λ
1
1
"t
Remark 1: about nominal values of yield strength fy
:
Eurocode 3 Part 1.1 (Ref. 1) gives a simplified variation of fy in function of the material thickness (thickness ranges: lower than 40 mm, from 40 mm to 100 mm), with values of fy - for certain thickness ranges - greater than the values specified by EN 10025 (Ref. 21) and EN 10113 (Ref. 22) (thickness ranges: lower than 16 mm, from 16 mm to 40 mm, from 40 mm to 63 mm, from 63 mm to 80 mm, from 80 mm to 100 mm, from 100 mm to 150 mm) (see following figure). As Eurocode 3 always provides yield strength f y equal or greater than values guaranteed by the specifications of delivery conditions (EN 10025, EN 10113), the available inelastic rotation av resulting from fy of EN 10025 or
EN 10113.
Minimum guaranteed yield strength RcH (or Rp0,2) m function of nominal thickness t of material Steel grades' RcH(orRpO^) [Mpa] 4 60
S 460
460
430
440 430
-42ΰ
S 420
420
410
-
400
390
400 390
370
355
S 355
360
355
335
345
340 335
325 315
-ZU
S 275
^
1
275
1
255
265 255
235
S 235
245
1
235
235
2/5
225
225
215
t {mm]
195
0
40
]6
80
63
100
150
Legend: .·
-EN 10113 ( for S 275, S 355. S 420 & S 460 steel grades). Euroc ode 3 specifications (for S 235, S 2?5,S3SS,S420& S 460 steel grades).
Remark 2: about overstrengthening of steel material
:
The real yield stresses of the steels delivered by mills are always greater than the nominal values of the steels indicated in the specifications EN 10025 (Ref. 21) and EN 10113 (Ref. 22). As presented in this chapter 4.3.1.1 clause (9), a statistical procedure has been used to evaluate partial safety factors γ^φ related to available inelastic rotation
40
Decreasing of the rotation requirement in case of actual yield stress which are greater than the
nominal yield stress
1
iHiiimmimnfTTTTrmr -·»-
i^reqWi
φ
Ψγ«Γ
4.3.1.2 Feldmann's formulas for values of
A. type of load: concentrated load with bending about major axis y-y of cross-section (My), B. type of load: concentrated load with bending about minor axis z-z of cross-section (M£), C. type of load: concentrated load with combined axial compressive force, bending about major axis y-y of cross-section (My - N).
In all formulas
:
Δο
=150 MPa,
fy.fi
= yield strength in flange must be introduced in MPa,
fy.w
= yield strength in web must be introduced in MPa,
E
= modulus of elasticity (= 210 000 MPa),
Npi
= plastic axial resistance of cross-section (=Afy),
Mpi
= plastic resistance moment of cross-section (=Wpi fy),
A = sectional area of cross-section. Available inelastic rotations
41
Α..
type of load:
concentrated load with bending about major axis y-y of cross-section
Plastic folding mechanism in compression flange of Η beams submitted to bending about major axis yy (Ref. 11):
5,2
-
f ,
^ M
+ 0,25 (
Y
3
4Et>4
17"
- ^.whtw + -J(^.whtw )
+ 4fy-wbtf twMa
V400
15.
type of load:
concentrated load with bending about minor axis z-z of cross-section
42
[rad]
(4.1)
Plastic folding mechanism in H beams submitted to bending about minor axis zz (Ref. 11)
This formula is only valid when stiffeners are applied at the loading point.
-
Wpi.zfy
(h-2tf)twfy-w
2tf
where
= Φκ,ρΐ +
[rad]
}|ιγ.Α+Δσ)
φκ?ρ1 =
0.2
('-f)
;
(4.2)
i = length of the beam.
In practical cases the ratio 2b/l are rather small. One can therefore assume the safesided value of Φκ,ρΐ = °<2·
v.".
type of load:
concentrated load with bending about major axis y-y of cross-section combined with axial compressive force
43
Plastic folding mechanism in H beams submitted to bending about major axis yy combined to axial compressive force (Ref. 11) :
Limit of application of these formulas is set to N/Npi < 0,5. To apply the appropriate formula, the following criteria should be considered
then the neutral axis lies in flange and φ3ν is calculated
from formula (4.4):
f* <-N v.
JMy
neutral axis >
[rad] where:
(4.4)
Ν is the applied axial load
MN.y = WpLyfy
1-
i-JL a_^, A-2btf
with the limit MNy
(4.5)
2A (see Eurocode 3 [3]
44
or Appendix
1
of present Annex
13)
Nfl=(tf-tM)bfy.fl XM
-' 2b
1-
(4.6)
N
(4.7)
Af,
yj
ρ _ 4EbtS
Pc 51?" Pu
(4.8)
= btf(fy.fl+Äa)
(4.9)
The following value of Niimit defines the border between formulas (4.3) and (4.4) for q>av- Niimit is related to the case where neutral axis hes in the border between the web and the flange of the crosssection.
h-tf
<-N
JMy
neutral axis
Niimit=(A-2btf)fy.w 4.3.2
(4.10)
Kemp's model for steel sections
(1) Kemp and Dekker have developed two models for predicting the rotation capacity of steel beams (Ref. 15). The first proposed model is based on a theoretical consideration of the strains in the section and of the length of plastic region of the flange for which local buckling is assumed to occur. Complicated formulae of the first model are given for information in Annex 11. The second model is a semi-empirical approach, which is based on comparisons with tests results and results produced using the first theoretical model. More details are given about the second simplified model in Annex 1 1 .
(2) ProfilARBED has applied formulae of Kemp's simplified model for I- and H-cross-sections submitted to bending according to the major axis (3 point bending beams), in order to evaluate available rotation capacities (Rav) (see Annex 11).
4.3.3
CRM's model for steel sections
(1) The available inelastic rotation
4.3.4
Comparison between models of
(1) Comparison has been made between 26 experimental results on steel section and calculation models determining available rotation capacity, allowing local buckling: the Kemp & Dekker theoretical and simple models (K & D simp.), the Johnson & Chen best fit (J & Β best) and lower bound (J & Β lower) models, and the Spangemacher & Sedlacek (S & S) model (see Figure 21a), issued from document 3263-1-19: "Design rules for continuous composite beams using Class 1 and 2 steel sections Applicability of EC 4"). 45
Kemp & Dekker simple model appears to be more appropriate, because : the calculated rotation capacity is nearly always less than the tests results (meaning that the model is conservative),
the spread of errors is less than for the other models (better scatter of values).
Kemp model and others give relative precise values of the ultimate load of the tested beams (differences of ± 5 %) but provide less precise values for φ>0ι, the largest rotations which are decisive to evaluate the available rotation capacities.
(2) On the other hand, Feldmann's model which has been recently developed (Ref. 11), intends to determine as accurately as possible the values of
(3) ProfilARBED compared the efficiency of both models on the basis of a databank of tests reults (see chapter 4.2 and Annex 1 1). Therefore rotations capacities are compared with tests results in Annex
11
as
follows
:
R experiments are plotted versus R models (Feldmann's model, Kemp's simplified model and CRM's model) (Figure 21 b)) and Annex 1 1) experimental results with >60, meaning !zc*e occurrence of lateral-torsional buckling (see clause 3.2 (5)), are excluded, assumptions for Feldmann's model
assumptions
for
Kemp's
:
simplified
model
^-<20οτ^-> 100 are excluded (see Annex 1zc·^
11
:
experimental
results
with
or Ref. 15).
ïzc·^
in (Ravaiiable - Lspan(= span length)) graphs (Annex 11), the curves are calculated with nominal dimensions of profiles and measured yield points in using Feldmann's model, Kemp's simplified model and Kemp's theoretical model; experimental results are given and local values of models with measured geometry are sometimes evaluated. (4) Conclusions
:
In order to elaborate practical tables or charts for designers, the working group of this research project has preferred Feldmann's model to Kemp's model and CRM's model for the following reasons, on basis of presented databank of tests results (see chapter 4.2) : more results from Kemp's model are on the unsafe side, even if extreme cases of tests are excluded (see Figure 21b) and Annex 11),
with Feldmann's model the evaluation of ductility of plastic hinges can be carried out by comparisons with either inelastic rotations (pav or rotations capacities Rav; and, on the other hand as explained in chapter 4.4.7.1, the criterion (3.1) comparing inelastic rotations
46
av. experiments
20 -ι
SAFE
15
M
>o
·
y
w · 10
-
5
-
^ /
0
χ
°
K&D simp. J&C best J&C lower
"VM** fl* η
·
S&S
UNSAFE|
fill (
1
)
10
5
15
20
R av. model (Spangemacher and Sedlacek values (S&S) given for fu / fy = 1,4) a) : Comparison
of rotation capacity from tests results and from various models
R av. experiments 24 23 22
X CRM's model
21
Δ Kemp's simplified model
20 19 18 17
Feldmann's model
16 15
14 13 12 11
10 9 8
7 6 5
4 3
2 1
0 0
1
2
3
4
5
6 7
8
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
R av. model b) : Comparison of rotation capacity from tests results and from models of CRM, Feldmann and Kemp
Figure 21 : Comparison of rotation capacity 47
Author Test N° dl6b3m dl9a4m d20b4m d01a4m d02b4m d09a3m
«« ¿5 ¿"i ¿o J,o ¿o m »ι ÍS s- ο <"ì S <". «η «ξ. S «ξ. -*-1 _·«^ cí
Rav/Rm
Figure 22 : Ratio of Rav/Rm for experimental data
48
o o
Λ
4.3.5
Couchman's model for composite cross-sections
(1) The purpose of the thesis of Mr. Couchman (Ref. 14) is to investigate the behaviour of continuous composite beams with plastic (Class 1), compact (Class 2) or semi-compact (Class 3) critical sections. (2) The software Compcal developed at EPFL has been used to compute the available rotation capacities of beams with plastic and compact sections by introducing Kemp's model (Ref. 15). A wide variety of parameters have been chosen to study the influential factors to the available rotation capacities. These factors were : slendemess of the cross-section,
reinforcement at the support, represented by the ratio of hogging to sagging plastic resistance moment (the plastic moment ratio), structural steel characteristics,
slip between the steel and concrete, degree
of shear connection between the steel and the concrete,
ductility of shear connectors, span lengths,
number of spans,
ratio of adjacent span lengths, type and arrangement of loading,
propping of the beam during construction. The influence of parameters which are related to the composite section, or length of beam in hogging, on available rotation capacity has been shown, and a single variable can be used to represent all such individual parameters (see Annex 2 (3/9) and (4/9)). All parameters which affect available rotation capacity should therefore be allowed for in a design model, which is not the case for simplified methods of analysis such as those proposed in Eurocode 4 (Ref. 2). (3) For semi-compact composite sections, the available rotation capacity depends on cross-section properties but also on the arrangement of spans and loads (see Annex 2 (5/9)). The Kubo and Galambos model is used (see Annex 2 (6/9)).
4.4.
Numerical simulations of available rotation capacity
4.4.1
Simulations of tests results
(1) In order to check the efficiency of the finite elements modelization LABEIN has carried out numerical simulations of different 3-point bending tests from RWTH (Aachen) (Ref. 5 and Ref. 13) (Tests documentation: Annex 5 (6/46)) : the tests n° 16 to 18 of table 2 (n° EA28412, EA2843, EA2844) concerning HEA 280 profiles, with yield strength of 504 Mpa, which are in class 4 according to Eurocode 3 and nevertheless which furnish available rotation capacities from 4,1 to 16,5 ! * other tests specimens as n° 4 to 6 of table 2 (n° EA22412, EA2243, EA2244) concerning HEA 220 profiles, with yield strength of 420,5 Mpa, which are in class 3 according to Eurocode 3 and neverheless which furnish available rotation capacities from 1,5 to 15,4 ! * here is the list of profiles related to the numbers of tests specimens: . HEB 220 for tests n° 1 to 8 of table 1 (class 1 for S235 and S460 steel grades), . HEA 220 for tests n° 1 to 6 of table 2 (class 1 for S235 and class 3 for S460), HEB 280 for tests n° 7 to 12 of table 2 (class 1 for S235 and class 2 for S460), HEA 280 for tests n° 13 to 18 of table 2 (class 2 for S235 and class 4 for S460). The tables and figures included in Annex 5 present a summary of these simulations :
*
49
a) Assumptions of
finite element modelization are given in Annex 5 (1/46 to 5/46)
b)
The tables of Annex 5 (7/46 and 10/46) shows a synthesis of the results presented in an appropriate way to be compared with the experimental results. Beside this, the available rotation capacity given by Kemp's and Feldmann's models have been also included together with the computed and tests results. It can be observed how the finite element results correlate reasonably well with test data. Feldmann's model always provides conservative results and Kemp's model is often unsafe.
c)
The graphical layouts of Annex 5 (9/46 and 1 1/46) present the correlation between the calculated and measured values of different parameters for every case: maximum moment (Mu), rotation at the maximum moment (Qmax = Φπμιχ), inelastic rotation (Qr0t = 9av)>
f rotation capacity defined with Qmax · Rmax
~ « _9max-9pl
"\
Φρί and rotation capacity as usually defined with Qrot : Rpi
=R
^Φκ^-Φρΐ
ν,
d)
J
The figures of Annex 5 (8/46 and 12/46 to 29/46) give the comparison between calculated and measured moment-rotation curves for the cases where experimental relations were available. The Figures of Annex 5 (30/46 to 46/46) also provide deformations of beams related to the most right hand intersection between respective moment-rotation curve and plastic moment level. Those simulated deformations can be compared to tested specimens.
(2) CTICM has performed 5 types of numerical simulations in order to evaluate the available rotation capacity of a I-profile. The ahn of these simulations was to identify the main parameters which govern the available plastic rotation. More information is given in Annex 6 (document 3263-2-9) : a)
Firstly, comparisons were made between experimental results and numerical simulations from LABEIN (ABAQUS Program) and from CTICM (ANSYS Program). These results were not different in the increasing part of the curves, but after the top, significant deviations have been noted.
b)
4.4.2
The influence of a global imperfection, the influence of an eccentric load (biased load) and the case of a cantilever beam have been investigated (see also chapter 4.4.6.2).
Simulations of nominal cases without tests results
(1) In order to check if Feldmann's model is valid for all I or Η steel cross-sections, simulations were performed for different nominal cases for which no tests results are available. As numerical simulations have been successfully calibrated to existing tests results (see chapter 4.4.1), they are considered as representing the reality to which Feldmann's model has been compared (see Annex 14).
(2) Labein simulated following 3-point bending beams without tests results using nominal values of dimensions and yield strengths :
*
providing Ravailable > Ö, in case of class 3 or 4 cross-sections
:
- IPE A 300, with fy = 460 Mpa (EC3: class 3f) and span = L = 2,80 m ((L/2)/(iz.e) = 58,6 < 60), [Feldmann's model:
50
providing very high values of Ravailable^ hi case of class 2 cross-sections
*
:
- HE A 280, with fy = 235 Mpa (EC3: class 2f) and span = L = 7,9 m ((L/2)/(iz£) = 56,4 < 60), [Feldmann's model:
*
providing very low values of Ravailable-· m case of class 1 cross-sections
:
- HE A 650, with fy = 460 Mpa and span = L = 5,9 m ((L/2)/(iz.e) = 59,2 < 60), [Feldmann's model: (pav = 0,07 rad = 4,01°; Ravailable = 3,3, for L = 5,9 m], - IPE A 500, with fy = 460 Mpa and span = L = 3,7 m ((L/2)/(iz.e) = 59,1 < 60), [Feldmann's model:
*
providing very high values of Ravailable^ hi case of class 1 cross-sections
-
:
HEM 100, with fy = 235 Mpa and span = L = 3,0 m ((L/2)/(iz.e) = 54,7 < 60), [Feldmann's
model:
*
in case of high cross-sections -
*
:
HE AA 1000, with fy = 460 Mpa (EC3: class 2w) and span = L = 4,9m ((L/2)/(iz.e) = 59,1 < 60), [Feldmann's model: (pav = 0,04 rad = 2,29°; Ravailable = 3,33, for L = 4,9 m].
for HE A 200 and Ή EAA 200 profiles, for spans of respectively L = 2 m & 4 m and L = 2 m, for fy = 235 MPa, classifying the sections respectively in class 1 and class 3 and, for fy = 460 MPa, classifying the sections respectively in class 3 and class 4, in order to quantify the influence of the yield strength fy on the rotation capacity and on the classification of cross-sections.
(3) The tables and figures included in Annex 14 present a summary of these simulations a)
:
the tables and graphical layouts of Annex 14 (3/13 and 10/13) provide a synthesis with comparison between Abaqus simulations and Feldmann's model results.
Following symbols are used : Phi pl. =
Phi rot. = %0{, theta rot =
of the
results
and Rpl = R.
Feldmann's model fits very well with simulations results and moreover if partial safety factor Tm
b)
4.4.3
the Figures of Annex 14 (from 4/13 to 9/13 and from 11/13 to 13/13) show moment-rotation curves and deformed shape of beams for several singular cases.
Simulations of beams submitted to My-N loading
(1) In order to check Feldmann's model about My-N interaction loading, simulations have been carried out to determine the available inelastic rotation for different (N/Npi) values especially when the neutral axis moves from web to flange providing big differences of available inelastic rotations
^1
-(max)
if
51
Npi
<^^A
(see formula (4.3)),
- (min) ^2SL ΎΜφ
A°2
Npl
A
(see formula (4.4)),
where Aweb = A - 2btf , A = area of total cross - section and
'N^
-Aweb=0j257-
VNPlAimit
A
of Annex 15 (7/10) shows the plot of
(3) Figure 3
transition between those 2
4.4.4
|
N/T
)
ll/^Aimit
= 0,257 J
Simulations of LTB limits for Feldmann's model
(1) An important topic concerns the rules delimiting the application field of Feldmann's model in regards to lateral-torsional buckling (LTB) of members. In the formulas of Feldmann's model the length of member is not considered. But it is well known that for different levels (L/iz), the collapse of bended members with plastic hinge(s) is influenced : - either, only by LTB of members without occurrence of plastic hinge(s), - or, by an interaction between LTB of members and local buckling in plastic hinge(s), - or, by local buckling in plastic hinge(s) followed by LTB of members. (2) In order to check if Feldmann's model is valid within lateral-torsional buckling limits (of Eurocode 3 (Ref. 1) or CM66 (Ref. 18)) and on the other side in order to determine which limits from Eurocode 3 or CM66 are more realistic, numerical simulations have been realized by Labein. (3) Concerning the results for HEA 200 cross-section (see Annex 16 (9/25)) available inelastic rotation
52
LTB rules
:
χ
-Ξ-
/777Γ
Yy * Lateral restraints \i
S-r-Zr îl
Β
ψ
ψ=0
ψ=0
CM6:
if ψ = 0 and Lj/iz < 60 ε, then lateral-torsional buckling is prevented.
Eurocode 3
if Xlt á 0,40 , then lateral-torsional buckling is prevented. LTB rules: Lj/iz £
Standards
HEA200 CM66
Steeel grades
IPE 200 60 S 235
Eurocode 3
63,3
61,4
CM66
42,9 S 460
44,2
Eurocode 3
42,4
In these cases, the maximum length Lj between lateral restraints is quite similar according to CM66 and Eurocode 3 rules. (5) The differences between prescriptions from both design standards CM 66 and Eurocode 3 are presented for all standard I and H profiles (ΓΡΕ, IPE A, TPE O, HE AA, HE A, HE Β and HE M) and for both steel grades S 235 and S 460 (see Annex 16 (24/25) and (25/25) : Figures with L/^ = ί(ψ)), for simple cases with linear moment gradient and with ratio between extreme bending moment ψ :
-1< ψ ν
M2
<1.
As first global approximation, it can be said that CM 66 seems to be -
safer than Eurocode 3 when
0 < ψ < +1,
-
similar than Eurocode
ψ = 0,
-
less safe than Eurocode 3, when
3
when
-1 <
ψ < 0.
53
:
4.4.5
Numerical simulation of double clamped beam case
(1) A 6 meters long beam clamped at both ends and subjected to distributed load has been modelled by LABEIN in order to evaluate how the behaviour of the plastic hinge in this configuration differs from the simply-supported case. The data of the test case "17-EA2843" included in document 3263-1-8 (Spangemacher & Sedlacek) have been taken for the cross-section (HE 280 A ; b = 281 mm; tf = 12,6 mm; h = 275,6 mm; tw = 9 mm).
In this case, a first hinge is developed at the end locations, while the central section is still subjected to half of the plastic moment From this moment, the end hinges have to rotate up to the point in which the plastic hinge appears at the central section.
The figures included in the working document 3263-3-10 (see Annex 7) show, on one hand, the deformed shape after the formation of the first plastic hinge just before the maximum load (P/Po = 1 .46) (Annex 7 (1/7)) and, on the other hand, the deformed shape after the maximum of the load when all plastic hinges (at supports and at mid-span) have been developed (P/Po. = 1-29) (Annex 7 (2/7)) . Po indicated the load at which the plastic moment is reached.
A slightly buckled lower flange can be appreciated at the end plastic hinge just before the maximum of the load (the figure is affected with a magnification factor of 5) although the maximum end moment Mi is not reached yet, indicating that the local buckling starts to happen before the maximum moment Nevertheless, the figure representing the evolution of both moments, Mi (end section) and M2 (central section) show clearly how the first plastic hinge is capable of rotating without falling into instability until the last plastic hinge starts to form at the central section. The rotation has been taken at the section of zero moment to be consistent with the curves coming from the 3-point bending tests. (2) The same 6 meters long beam described above has been subjected to central load, extracting the same kind of results. In this case, both the end section and the central one reach the plastic moment at the same time. Because of this, both plastic hinges are simultaneously developed as shown in the figure representing the deformed shape (Annex 7 (5/7)) . (3) The comparison between the moment (at end section) - rotation (at zero moment section) curves obtained for the two cases, distributed and concentrated load, can been appraised in the figure (Annex 7 (6/7)) : the resulting available rotation capacity is very close one to each other. It is worth to note that the equivalent span for the concentrated load case (Leq = 3 m) is a bit larger than that for the distributed case (Leq = 2,5 m), due to the different sections where the zero moment appears. In the last figure (Annex 7 (7/7)) , the two curves have been represented together with the moment rotation curve coming from the 3-point test simulation and with the experimental one. It is interesting to note the good correlation with the experimental curve obtained for the clamped concentrated load case simulation.
4.4.6
Moment gradient and LTB restraints influence on available rotation
4.4.6.1 Introduction This section presents some conclusions about the combined influence of the moment gradient and the lateral restraints on the resulting available rotation capacity on the basis of interpreting the results of a number of numerical simulations carried out by LABEIN.
4.4.6.2 Simulation of simple cases (1) The working document 3263-3-9 includes results concerning different simulations carried out by LABEIN to appraise the influence of some factors on the available rotation capacity: type of load, position of the load and lateral restraints. (2) Position of the load:
Four different models have been considered moving the concentrated load from the center towards one of the beam ends but keeping constant the total length. The results are in good agreement with other simulations carried out on the basis of changing the beam span at the same time that the load was moved (document 3263-3-7) : the moment-rotation response, the predicted rotation capacity and the ultimate moment do not change significantly when moving the position of the load from the beam center provided that the load is not placed very close to the beam end. In this case, the deformation 54
mode changes suddenly due to shear effects. In this case when the load is near the support (25% for example), high shear deformation dramatically diminishes the available rotation capacity if the reference level remains Mpi.
But in case of high shear the reference level to evaluate the rotation capacity should not be the plastic moment resistance Mpi but the assumed ultimate resistance defined by the reduced plastic resistance moment allowing for the shear force My. For example that value My could be obtained from Eurocode 3; then the interaction between shear force and bending moment should be considered if the applied shear force is greater than half the plastic shear resistance of the cross-section. In that case of lower reference level (My < Mp]) , if moment-rotation curves are similar, the available inelastic rotation
Elastic rotation
Total rotation
Inelastic rotation
Φρί
9rot
9av =
100 % (CENTER)
1,2046
29,99
28,79
75%
1,2048
33,90
32,69
50%
1,2049
31,81
30,60
25 %
|
1,2049
?heJ£n?lusions of CTICM simulations about biased concentrated loads (see Annex 6) are similar to
LABEIN results.
(3) Type of load and LTB restraints: The following table presents the moment-rotation curves associated to the same simulation case (S10) but considering different load cases and different lateral restraints (see Annex 8 (2/3)):
Calculation rotation capacity, Rn
1
Concentrated Load
Distributed Load
No lateral restraint at hinge
24
19
Lateral restraint at hinge
34
33
The simulation results corresponding to cases with distributed load allow to observe a general trend in their behaviour in contrast with the concentrated load cases. The distributed load increases the sensitivity of the beam to suffer from lateral-torsional buckling (LTB) simultaneously interacting with the local buckling. As a consequence of this, the combined inelastic buckling mode (global and local deformation) makes the instability to happen at lower levels of load and rotation providing that no provision is made to avoid the global LTB deformation mode. The results is a lower rotation capacity caused by a combination of both factors: a very flat moment diagram and an insufficient lateral restraint ot the beam.
This effect can be clearly appreciated in the calculated moment-rotation curves (see Annex 8 (3/3))· if we compare the cases with concentrated and distributed load, the rotation capacity keeps constant in case of restraining the lateral displacement at the plastic hinge location. However, if this lateral restraint is removed, the rotation capacity becomes lower for the distributed load than for the concentrated one. Thus, there seems to be an interaction between the moment gradient and the lateral torsional restraints in such a way that one of the factors strongly affects the influence of the other on the calculated rotation capacity. A critical minimum length (Lb) for providing lateral restraints may be assumed to exist in order to ensure the calculated rotation capacity to be unaffected by the type of loading or the moment gi duldlL.
It is worth to note the presence of the stiffener at central section in the cases of concentrated load provides an additional torsional restraint compared to the distributed load case. This effect has been proved by applying the distributed load to a model with central stiffener. The calculated rotation capacity (Rav = 23) is increased compared to the case without stiffener (RaV = 19) 55
4.4.6.3 Interaction between moment gradient and LTB (1) A general trend could be observed when comparing simulation results corresponding to cases with distributed load in contrast with the concentrated load cases: the distributed load increases the sensi¬ tivity of the beam to suffer from lateral-torsional buckling (LTB) simultaneously with the local buck¬ ling. As a consequence of this, the combined inelastic buckling mode (global and local deformation) makes the instability to happen at lower levels of load and rotation providing that no provision is made to avoid the global LTB deformation mode. The result is a lower rotation capacity caused by a combi¬ nation of both factors: a very flat moment diagram and an insufficient lateral restraint of the beam.
The interaction between both factors can be appraised by defining the following two parameters:
- Lltb- Lb:
I)
Π)
*-!=-<
length between lateral restraints. length of the region of significant yielding determined by the moment gradient (flatter moment diagrams give rise to higher Lb values).
When sufficient lateral restraints are provided to avoid the global lateral deformation to interact with the local buckling (Lltb « Lb)» a single local flange buckling occurs and the available rotation seems to be unaffected neither the restraint length (Lltb), nor the moment gradient (Lb).
IOC« BaMPPOIICWtNRflinrilwOTR)
ΰίψ"
ΙΠ) In the intermediate case with a restraint length Lltb partially avoiding the global lateral buckling, a combined dependence of the available rotation on Lltb and Lb seems to happen.
r
j
Σ
Σ
1
ftF*{U,Xt>)
locai n»NPP»qqm lairaai auno hot ιρ^ππ}
In case of high Lltb values in comparison to Lb (Lltb » Lb), the combined inelastic buckling mode (local flange and global lateral) appears and the available rotation seems not to be depending on the restraint length (Lltb)» being only influenced by the moment gradient (Lb).
1
U>U>
Χ
Σ-Χ
«.-OS
toc« π aure» pappan YPPSTPAiNFn p¡ (wat utfpm rup
¿ 11
χ χ-κ-χ τ Figure 23 : Buckling mode related to unbraced length Lltb
Figure 24 : Inelastic rotation θρ of plastic hinge in function of unbraced length Lltb 56
(2) If we compare the cases with concentrated and distributed load, the available rotation keeps nearly constant in case of providing sufficient lateral restraints at the proximity of the plastic hinge location. However, if the lateral restraint is relaxed, the available rotation becomes lower for the distributed load than for the concentrated one. Concentrated
Load
Distributed
Load
Lb'
m
lili.
Mmax
-i.
μιιιι A
,γττ
£
U.»Lb
Σ
S
*
-Χ
L,.
ϊ -X
U-,
"
+
Γ
τ
i^vi
Ν^
4
4-
i À
Δ L,,.-«Lb
K^
>
-E
K *
Ut,
X-
x-x-
Ή
x
Figure 25 : Unbraced lengths Lltb for different moment gradients The question that arises is the determination of the critical minimum length Lltb for providing lateral restraints in order to ensure the calculated rotation capacity to be unaffected by the type of loading or the moment gradient
(3) Deformation buckling modes
:
The simulation cases modelled with the lowest values of Lltb (providing lateral restraints along all the upper flange or restraining the rotation at the plastic hinge location) gave, unexpectedly, lower available rotations than those obtained for the higher Lltb values. The reason for this behaviour is a change in the buckling deformation mode. The additional lateral restraints have avoided the lateral global buckling to interact with the local one but, at the same time, it has introduced a different mode con¬ sisting in a coupling between the flange local deformation and a local web buckling (B).
Therefore two different deformation buckling modes have been identified:
57
A) Flange local + lateral global Ducklings
Figure 26 : Flange local + lateral global bucklings B) Flange local + web local bucklings
Figure 27 : Flange local + web local bucklings (4) Summary of simulation results
:
The following cases have been modelled combining different lateral restraint conditions (varying the length between lateral restraints) with different moment gradients (varying the length between vertical supports) in order to evaluate the interacting influence of both factors on the calculated available rota¬ tions. The total length of the beam has been adapted, for each case, to the maximum required distance between vertical or lateral supports. The table shows the calculated available rotation (
HE 280 A
Material:
fy = 504 N/mm2 , fu/ fy =
Load:
Concentrated central load
58
1,35
Available Rotation
Type of beam
1
f"
S m
Lateral restraints X
X
«-
4 m
Buckling Deformation Mode
Li /M
0,21
0,10
All
41,8
0,17
0,10
AH
31,4
0,20
0,10
AH
12,6
Χ
0,24
0,30
AI
X XXX X XXX Χ
0,13
0,18
ΒΠ
0,29
0,36
BI
0,30
0,38
AI
4 m
-X-
3 m
-x-
1.2 m
S
1
ι SS- S3
Σ XXX Χ XXX Χ ttíppor fiLis«)
S-
X
(upper
II* Aft·)
Buckling deformation modes:
*AII
"^
*AI
Χ
*ΒΠ
Ι)
r
_37
Local flange + limited global lateral
Γ&
^"""W^ ® r*er -
*BI
Local flange + free global lateral
-s-^
ζ
Local flange + free local web
,
Local flange + limited local web
A further detailed description of the results is found on working document 3263-3-12 (Annex 3). 4.4.7
4.4.7.1 Inelastic rotation φ and rotation capacity R of a steel cross-section (1) As presented in chapter 3.1, the validity and the consistency of a plastic analysis of a frame under a given loading can be checked in all cross-sections by evaluating the ductility of plastic hinges either (3.1) with plastic rotation, or (3.2) with rotation capacity :
9av ΥΜφ
59
(3.1)
<
R
or,
av
(3.2)
Ymr
in reference to the 3 points-bending system
:
φ
7^ rot
mnnrtpnt 3α moment Hi diagram
^^*^1^-*"^
with
-^av Kav -
1?
φρι=
and
Φρί
Mpl.L
E.I
Figure 28 : Moment diagram and moment rotation curve for 3 points-bending system (2) The concept of rotation capacity R introduces a supplementary parameter such as the length L of a reference beam. It must be noted that this parameter is a structural and not local parameter. Rreq is to be calculated from frame plastic analysis as explained hereafter. The available rotation capacity Rav can also be given in tables but length because (ppi depends on the length.
it must refer to a particular beam
Because the verification has to be consistent, the required rotation capacity Rreq must be computed by considering the same length of reference which will be used to determine Rav
(3) By analogy with the 3-point bending system, this length of reference (or equivalent length) L related to a plastic hinge in a frame can be defined as the distance between zero-moment points on each side of the hinge : Plastic hinge ν
A "^ujjiipf
.Üb.
<
"Δ
τ
^¡W Figure 29 : Equivalent length L to 3-point bending system (4) For a beam where a plastic hinge occurs under a single concentrated load, the zero-moment points are often easy to identify. For more complex configurations, it is not so easy to determine the reference length, such as the following common examples in Figure 30 :
60
l\
\
A\\
Plastic hinge
\
Plastic hinge ..
m I
Figure 30 : Plastic hinge in structures (5) In chapter 3 of Annex 10, a method is proposed to calculate the required rotation capacity if the shear force is known on each side of the plastic hinge. The case of a plastic hinge at the end of a beam in a portal frame is also investigated (Figure 31).
Method of evaluation for required φ arid R
Bending moment gradients
or Rreq(Ti,T2) < Ravail(Leq) 9p.avail or Ravail(D given by formulae or tables
*req(Tl,T2) = "A ^.f M J, Ti+T2 Leq = Mpl
f J_
«Pjueq
_1_
σι + τ2
or RreqCO ^ Ravail(Leq) 9p.avail or Ravail(L) given by formulae or tables
9p.req- 9p.avail
ρKTeq
- ELT " 9p.req
M Pi Leq=2.Mpl/T
-req -
Çp.avail or RreqCO ^ Ravail(Leq) 9p.avail or Ravail(L) given by formulae or tables
At supports, when the moment is linear in the vicinity of the plastic hinge, the problem is the same that for a concentrated load on a beam. TI /fff|Yiv T2
il iibl W
TL
.Hastic hinge
i Figure 32 : Equivalent length L for different moment gradients 61
When the moment diagram is not linear (due to distributed loads), the reference length is given by the tangent on each side of the plastic hinge as shown in Figure 32. Therefore it does not correspond to the distance between zero-moment points, but this seems to be more realistic since the required rotation capacity depends only on the shape of the moment diagram near the plastic hinge. (7) Conclusions : It is difficult to clearly identify the length of reference (or equivalent length) L by analogy of 3 point bending beam in order to calculate φρί. On the other hand, computer softwares usually provide required inelastic rotations (preq at each formed plastic hinge for each load level till structural collapse (mechanism,...). Thus the most simple way to check ductility of plastic hinges clearly consists in comparing directly required and available inelastic rotations φ of plastic hinges instead of converting those values in rotation capacities R.
Therefore it would be advisable to refer to the criterion (3.1) dealing with inelastic rotations (Φ) and not with rotation capacities (R) (criterion (3.2)). This conclusion is highlighted by the results of LABEIN numerical simulations presented in chapter 4.4.7.4.
4.4.7.2 Introduction to numerical simulations
Two cases of 2-span continuous beams with concentrated and distributed load respectively have been modelled and simulated by LABEIN in order to evaluate the equivalent length to be considered for the checking of required rotation capacity in comparison with the available rotation capacity evaluated from the equivalent 3-point bending test in cases of different moment gradients. 4.4.7.3 Model description Geometry & cross-section:
Two span beam. IPE 270, Steel grade S 235
* Concentrated load case: L=10m λ = 0,5 α = 0,8
«L
* Distributed load case: L=10m λ = 0,4 γ=1
. "
A *
L
.1
t
m
XLÎ n
Figure 33 : Concentrated Load
* 3-point bending cases: 1) Li = 1 m L2 = 5 m 2) Li = 3 m 3) Li = 1,6 m 4) Li = 1,9 m
L2 = 3m L2 = 2,17m L2 = 4m
Figure 34 : Distributed Load
Material:
Elasto-plastic model with fy = 235 MPa and fu/fy = 1,58
Meshing:
Type of element: 8-node shell with parabolic shape functions (S8R Abaqus) A more refined mesh has been provided at the areas where local buckling behaviour is expected to happen.
62
Concentrated
Distributed
Number of elements
640
616
Number of nodes
1973
1913
Degrees of freedom
11.838
11.478
Fillet radius modelling: By means of web and flange overthickness calculated for keeping constant the total section area.
Initial imperfection: Pattern:
the first positive elastic buckling eigenmode is used as the initial imperfection shape.
Magnitude:
maximum of
1
%
of the flange thickness.
Load application:
*
Concentrated load:
Incremental-iterative solution procedure has been employed using displacement control. Appropriate constraints to degrees of freedom at upper flange of the loaded section have been applied to ensure the same vertical displacement of mese nodes.
*
Distributed load: Load control has been used in the load-displacement analysis together with an arc-length integration method (Riks) to cope with the post-buckling unstable response. The load has been applied by means of equivalent node forces at upper flange central line.
*
3-point bending cases: As described for concentrated load case.
Boundary conditions
Vertical supports have been introduced at every node of the lower flange located at the corresponding support sections.
Longitudinal support has been provided in the loaded section at the centre of the upper flange.
Lateral restraints have been introduced on the compressed flange of the support and load application sections and additional sections according to Additif 80 - DPU Ρ 22-701 (CM66) (Ref. 18):
*
concentrated load case: a moment ratio of 0 has been considered in all the length (leading to L = 1,81 m between lateral restraints) excepting the part between both plastic hinges, where the moment ratio is - 0,87 (L = 2,87 m).
*
distributed load case: a moment ratio of 0 (L = 1 ,8 1 m) has been taken for the right part of the beam, the first hinge influence area, while a conservative moment ratio of 1 has been considered for the left part (2nd hinge zone) leading to L = 1,05 m between lateral supports.
*
3-point bending cases: a moment ratio of 0 is assumed leading to L = 1,81 m between lateral supports.
63
The bending moment diagrams are as follows
:
Concentrated Load
Distributed Load
1
Γ
15
>^JL+
15
ν
--wáu*
^UJJILP^
Ά £_!_ y/ìW %l17 S
1
Î
N-U
fr-**L
1-L-5
/
2.17 J
1.6-L-2.1
"'
Ê
i
â
JL
3
ì
m
3-L-3
1.9-L-4
Figure 35 : Bending moment diagrams for concentrated and distributed loads The equivalent lengths selected for the 3-point bending tests have been determined according to two different criteria: the length between the two zero moment sections adjacent to the plastic hinge,
the length determined by the tangent to the bending moment diagram on each side hinge.
of the plastic
4.4.7.4 Results Comparison with 3-point bending case results
= Φτοί - Φρί
Φρί (deg)
Φτοί (deg)
1,7
8,3
6,6
3,9
1,65
8,7
7,1
4,3
1,65
8,5
6,9
4,2
Continuous beams with distributed load
14
15
13,9
12,6
3-point bending beam 1.9-L-4 Li = l,9;L2 = 4
1,62
8,7
7,1
4,4
3-point bending beam 1.6-L-2.1 1,6; L2 = 2,17
1,03
9,1
8,1
7,8
CASE Continuous beams with concentrated load
3-point bending beam l-L-5
Li = l;L2 = 5 3-point bending beam 3-L-3
Li = 3;L2 = 3
Li =
64
Rav
(deg)
Concerning the concentrated load case (Figure 38), a very good agreement can be observed between the moment-rotation curve of the continuous beam and that of the corresponding 3-point testing case (Li = 1 m; L2 = 5 m). Also the case of central load (L\ = 3 m; L2 = 3 m) gives the same result as expected.
In the distributed loads case (Figure 37), however, a clear disagreement is observed when comparing the moment-rotation curve of the continuous beam and the corresponding curves of the 3-point bending cases
In view of simulation results it seems better to evaluate rotation capacity R with equivalent lengths determined by tangent to the bending moment diagram on each side of the plastic hinge (case 1.6-L-2.1) The discrepancy seems to be caused by the interaction taking place between the lateral deformations induced by both hinges in the continuous beam and giving rise to a stiffening effect This interaction is caused by the fact that the continuous beam has got positive and negative moments along its length. The lower flange, compressed near the first hinge location (negative moment region), is subjected to tension stresses in the positive moment region. This tension, acting on the lower flange at second hinge location, is responsible for the suffer post-buckling response appreciated in the moment-rotation curve of the first hinge in comparison to the 3-point bending case. It can be seen as a sort of stress stiffening effect that produces a positive influence on the side of increasing the actual rotation capacity of the beam.
This effect is avoided in the concentrated load case since the lateral deformation at both hinges is mainly local and does not interact one to each other. Thus, the rotation capacity of continuous beams subjected to distributed loads (this load case is more likely to induce coupled lateral deformations) may be significantly higher than that given by the 3-point bending test due to the described stiffening effect
A further detailed description of the results is found in Annex 4 (working document 3263-3-17).
65
16
τ
^-,
8
10
12
14
16
13
Rotation (dag)
Figure 37 : Continuous beam with distributed loads
o
4-
ι
10
8
12
14
Rotation Weg)
Figure 38 : Continuous beam with concentrated loads
66
16
18
5.
5.1
Required rotation capacity
Introduction The work performed by CTICM has mainly consisted in developing practical tools in order to facilitate the verifications of the plastic rotations obtained from a plastic global analysis of a steel frame. For several simple cases of continuous beams, formulas and abacuses have been developed. For more complex cases, a general methodology has been established. All these results are explained in the "Guide for plastic analysis". CTICM has written the second chapter of that guide : "Required plastic rotation of plastic hinges in structures" (see chapter 2 of Annex 13). The content of that chapter is : 2.1
2.2 2.3 2.4
Introduction Influence of parameters on required inelastic rotation Continuous beams Frames
We give hereafter the main results of that chapter 2 of the "Guide of the plastic analysis" (see Annex 13).
5.2
Influence of second order effects Through the simple following example, it can be shown the influence of the second order effects on the required inelastic rotation (influence of parameter 10.1 of Figure 4). 1.25
Due to the second order effects, the collapse is reached before a full plastic mechanism is obtained. Therefore, only three plastic hinges occur instead of four and the required inelastic rotation is much lower, at the ultimate limit state, from a second order plastic analysis than it is from a first order plastic analysis. But for a given load factor (λ = 1 for instance), the second order effects generally increase the required inelastic rotations in plastic hinges.
5.3
Influence of load level (design loads or collapse)
(1) Influence of the parameter 10.2 of Figure 4 : practical methods give directly the plastic mechanism for frames, and also the plastic rotations in the hinges and the internal forces just before the last hinge occurs. So, an usual way is to check TJLS criteria just before collapse, comprising rotation capacity in hinges. However, design rules only require to check these criteria under the design loads; especially, the designer has to check that a plastic mechanism is not reached. Computer programs can calculate the plastic behaviour of frames under design loads by increasing applied loads proportionally to a common load factor, α for instance. So, it can be shown that the inelastic rotation in the first plastic hinge can be far greater at collapse than it is under the ULS design loads. For most cases, checking the plastic rotation at collapse can be too conservative if so far as the frame design is never accurately optimised for all the limit states (ultimate limit states, serviceability limit states, ...).Therefore it is sufficient and economical to check the rotation capacity criterion under the ULS design loads. 67
Load factor = λ Collapse
Design loading at ULS
Plastic rotation
Figure 41 : Evolution of the plastic rotation in a first plastic hinge (2) A remark arose from the F6 committee about the influence of the parameter 10.2 of Figure 4, which influences the required rotation capacity of plastic hinges: the ratio between design load level (k¿) and complete plastic mechanism load level (Xmax).
In view of that parameter, the required rotation capacity chosen by the designer (Rrequiredl; as an example related to λι = X¿) must be clearly specified in the calculation note. Indeed, afterwards the building could be used or refurbished with other conditions (higher loads, other boundary conditions,...) and, the required rotation capacity in that context (Rrequired2; as an example related to Λ.2 with λ<ι < %2 < %max) could be greater than the previous one (Rrequiredl) and so the available rotation capacity of the members should be checked (Ravailable ^ ^required^ )· On the other hand, at ultimate limit state, the design of a building would not be optimised if the difference between design load level (λ<0 and complete plastic mechanism load level (Amax), was too high. But mat big difference between load levels could also be explained by restrictions at serviceability limit state or other reasons (instability of the frame,...).
5.4
Influence of loading path
(1) On one hand Eurocode 3 (Ref. 1) states, about global analysis, that "It may be assumed to be sufficient, in the case of building structures, to adopt simultaneous proportional increases of all loads" (clause 5.2.1.1 (5) of Eurocode 3). (2) On the other hand, in case of plastic global analysis, for a given loading arrangement the mechanism and the ultimate load factor do not depend on the loading path, but the order of occurrence of the plastic hinges in the structure depends on the loading path. The required inelastic rotations also depend on the loading path under the design loading and at collapse.
(3) The examples of Figure 42 shows a continuous beam with different loading path
:
path A corresponds to a proportional increase of all loads as allowed by Eurocode 3 (Ref. 1) and, path Β defines a specific increase of loads which is not always proportional. The collapse mechanisms are reached for the same load level (Pi ; P2 )whatever the loading path. But the required inelastic rotations of plastic hinges (
(4) It can be concluded that whatever the loading path is, required inelastic rotations in plastic hinges
68
Different loading paths lead to the same φ at collapse if the same mechanism is reached and if no elastic return occurs in any plastic hinge during loading. The following examples highlight differences in (preq at collapse because of elastic returns :
Pl
P2
Qi = 8P1
ΓΡΕ160 S 235
*~~Z L
Path A : Path Β :
2m
=
4m
*H-
All loads increase proportionaly Pi and Qi increase together with P2 = 0, then P2 starts while Pi and Qi are also increasing
Euroabstracts The European Commission's periodical on research publications, issued every two months. For more information, contact: RTD help desk, European Commission, DG XIII, L-2920 Luxembourg Fax (352) 43 01-32084; e-mail: [email protected]
Price (excluding VAT) in Luxembourg: ECU
* "ÖÜV
ir
_
1 1
4
OFFICE FOR OFFICIAL PUBLICATIONS 0F THE EUR0PEAN COMMUNITIES