Stessa 2009 – Mazzolani, Ricles & Sause (eds) © 2009 Taylor & Francis Group, London, ISBN 978-0-415-56326-0
Shear lugs for column bases C. Aguirre Departmentt of Civil Engineering Departmen Engineering,, Universidad Universidad Técnica Federico Federico Santa Santa María, María, Valpar alparaíso aíso,, Chile
I. Palma SKM Minmetal, Santiago, Chile
ABSTRACT: Column base plates of steel structures ABSTRACT: structures in seismic zones zones have have usually to transfer large large shear forces to foundations, and, and, in some cases shear lugs have to be used. It is normally assumed that the controlling limits states are the bearing strength of concrete and the yielding of the steel lug working as a cantileve cantileverr beam. In order to review the present design of shear lugs, an experimental program was conducted. This paper presents the results of the testing of shear lugs made of cross section steel shapes. It was found that the steel shear yielding mode becomes a more ductile and reliable failure than the concrete bearing failure mode, which, in some cases, exhibits an explosive failure. It is also demonstrated that the hypothesis of a cantilever beam behavior behavior of the shear lug is not appropriate; so as a result, the analytical model should be modified.
1
2
INTRODUCTION
Shear lugs are frequently used in structures placed in seismic areas, where the lateral seismic loads require transferring large horizontal loads to foundations. Figure 1 shows a cross type shear lug, they were thought to provide strength to any direction of the horizontal loads first. Nevertheless, there are only a few studies related to this matter, and the present design is based on methodologies coming from the engineering experience, and some basic theoretical principles adapted for the design of these shear lugs devices. This research work, points to evaluate the usual design conditions for cross type shear lugs. It also intends to improve the design method so it should guarantee a better seismic behavior of these elements. In order to validate the proposal, an experimental testing program was prepared, including different cross type sections shear lugs. They were selected to produce both type of failure modes, either in concrete or in the steel lug.
BACKGROUND
The previous work about experimental testing of shear lugs, are not directly related to columns bases design, but they are based on studies about behavior of steel plates embedded in concrete able to support and fix steel s teel elements. (Figure 2) Anchor bolts of embedded plates are welded to the plate, and the concrete concrete pouring is done once the the plate is already installed in the proper position; therefore, in a difference with column bases they do not need a grout. In spite of these differences regarding column bases, the above mentioned research work is the most relevant information reported, that can be found about shear lugs behavior. Rotz and Reifshnelder (Rotz, 1989) performed a series of tests of embedded steel plates with different dimensions. They studied the behavior of the plates under a combination of shear and axial load, in tension and compression.
M P
Figure 1.
Cross type shear lug geometry.
Figure 2.
247
V
Embedded plate in concrete.
Table 1.
The results of the tests presented a prime and first failure mode denominated “bearing mode”, related to the concrete compression bearing capacity. The “bearing mode” is associated to the formation of shallow fracture plane on the top surface of the specimen, a fast increase of horizontal and vertical displacement, and a fast reduction of the shear load strength. From this study, the following relationship for the shear capacity (Vconcrete) was stated. Vconcrete = K b f 'c Alg + K c (Py − Pa)
Required bearing area (AISC, 2003).
Design method ASD Alg =
V lg 0.35 ⋅ f ′c
Design method LRFD Alg =
V lgu 0.85 ⋅ ϕ c ⋅ f ′c
(1)
The first term of the equation: (K b f 'c A b) is the typical compression strength of the concrete; the second term: “(K c (Py − Pa))” is the concrete confinement effect on the base plate produced by the anchor bolts, it increases the strength capacity. K b and K c parameters were empirically obtained. The Chilean Code NCh2369-2002 “Seismic Design of Structures and Industrial Facilities” (INN, 2002) requires that: “Column base plates and equipments must have shear lugs or seismic split ends calculated to transfer one hundred percent of the basic shear load, excepting those cases when the shear load to be transferred is less than five tons in each support”. There are no standard procedures, detailing or engineering practice for the design of shear lugs in the Chilean codes. Current Shear Lug design in Chile is based mainly on AISC Steel Design Guide N°1 “Column Base Plates” (AISC, 2003). This guide gives some design requirements for a basic shear lug, which is a plate welded perpendicular to the column base plate (see Figure 3). AISC Steel Design Guide N°1, accepts both ASD and LRFD design methods. It accepts for the bearing stress in concrete a safe value of 0.35 ⋅ f 'c (unconfined concrete) for the ASD design method. For LRFD design method it applies 0.85 ⋅ ϕc ⋅ f 'c as a nominal bearing stress. Afterward, the dimensions of shear lugs must be such that the required bearing area between the shear lug and concrete (without con-
Figure 4.
Load and deflection of the Shear lug.
Table 2.
Shear lug thickness calculation.
Design method ASD
t lg =
6 ⋅ M l g 0.75 ⋅ F y
Design method LRFD
t lg =
4 ⋅ M g l u 0.90 ⋅ F y
sidering the grout) would be as it is shown on Table 1. The steel plate will behave as a cantilever beam; the design is based on the maximum moment (Mlg) at the base, acting on a unit length (Figure 4). The shear lug thickness is obtained considering the steel plate moment strength, according to Table 2.
3
DESIGN APPROACH
3.1 Limit state based on concrete bearing strength
Figure 3.
Typical shear lug.
ACI 349-01 Code (ACI, 2001), based on Rotz results, generated the B.4.5.2 indication. It states that design compression stress either for concrete or grout, placed on shear lugs shall not exceed 1.3 ⋅ φ ⋅ f 'c (LRFD method), and φ reduction factor shall be 0.70. A summary of ACI and AISC provisions is provided in Table 3. It can be seen that ACI nominal stress is 78% larger than AISC nominal stress, that means ACI pro posal for the concrete bearing pressure provides a larger strength than AISC-Guide N°1. As a consequence, it’s seems an appropriate and reliable approach to estimate the Concrete Bearing Strength for the design of future
248
Table 3.
ACI-349 and AISC-Guide N°1 comparison.
Nominal stress (LRFD) ACI 349-01
Not with standing steel shear yielding failure is another potential limit state, so the experimental program was designed to compare both type of failures modes and the conditions for their occurrences.
Nominal stress (LRFD) Guide N°1, AISC
τnom = 1.3 ⋅ φ ⋅ f 'c = 0.91 f 'c τnom = 0.85 ⋅ ϕc ⋅ f 'c = 0.51 ⋅ f 'c ⋅
4
EXPERIMENTAL PROGRAM
The first phase of the program includes the experimental testing of cross section type shear lugs; such shape was chosen because it is currently used in Chilean projects. Figure 6 shows the General Scheme of test program. Due to the features of the testing machine, the manufacture of concrete blocks with two equal shear lugs in a symmetrical fashion were projected, this geometry reduces the eccentricities
Figure 5.
Rotz tests results.
shear lugs according to ACI-349 recommendations, based on the experimental tests results. 3.2 Limit state based on a steel failure In order to understand whether the shear lug steel plates fail in flexural yielding or in shear yielding, Rotz tests results are compared with the calculated strengths, considering a cantilever beam in bending as well as the shear strength of the beam. The chart included in Figure 5 shows the shear force applied when the concrete bearing failure took place (V real), the shear force determined from analysis assuming a cantilever beam that yield in bending of the steel plate (V bend ) and also the shear yielding strength (V shear ) calculated with Equation 2 (The average yielding stress was assumed 1.25 times the minimum required yielding stress for A36 steel). Vsteel = 0.6 ⋅ Fy ⋅ Ashear
Figure 6.
(2)
In 92% of the cases, flexural yield was the smallest strength, according to these results, the shear lugs should have failed according to this first mode. The failure mode reported was the bearing of concrete; such a failure mode produced a larger strength than the flexural yielding mode in almost all of the tests, so, according to Rotz results the bending never hap pened, otherwise, the typical failure mode should have been flexural yielding. Thus, tests results indicate that this kind of failure is less probable for shear lugs. On the other hand, it is easy to realize that shear lug failure due to shear yielding exhibits the highest strength, consequently, Rotz specimens always failed due to bearing of concrete.
249
Test scheme.
Table 4.
Tests program details.
Test Nº
W (mm)
H (mm)
t (mm)
G (mm)
1 2 3 4 5 6 7 8 9 10 11
100 100 100 100 100 100 100 100 100 140 140
100 75 50 100 75 100 75 50 100 140 140
5 5 5 8 8 12 12 12 19 5 8
8 8 8 11 11 15 15 15 20 40 40
12
140
140
12
40
that could happen during the testing of the specimens. The machine load was distributed towards both shear lugs through a rigid beam piece. Table 4 presents a detail of a shear lugs series. The twelve specimens were selected by changing the dimensions and/or shear lugs thicknesses. The idea was to produce different types of failures in the resistant system. In order to evaluate influence of the grout thicknesses, the last three tests have higher grout thicknesses (see Figure 1).
5 5.1
TESTS RESULTS Steel failure
Figure 7 shows the shear lug curve when the failure is steel yielding shear type (test Nº1). An initial high rigid range is shown where displacements are less than 1.0 mm; it follows the plastic range with displacements up to 17.0 mm. The beginning of the system’s yielding was chosen as the point where the decreasing of the stiffness starts. Figure 8 is a picture of the specimen Nº1 after the test, it can be seen a typical shear deflected geometry of the yielding zone
Figure 7.
in the shear lug at the grout depth. The rest of the shear lug remains straight, with a small inclination respect to the vertical axis. The picture demonstrates that flexure deformation of the embedded part of the steel lug did not happen. Figure 9 shows a summary of the curve shear— displacement for all the specimens that failed due to steel yielding; the shear is presented as V/V max. All the specimens have a high initial stiffness (less than 1.0 mm deformation) and the plastic zone reaches a large ductility level. The failure normally exhibits a crack, it sometimes happens in the grout and sometimes in the concrete. The collapse happens with the rupture of anchor bolts. Table 5 shows the relationship of the theoretical capacity (Vtheo, calculated by Equation 2) and the actual failure load (Vtest, obtained from tests). The average relationship between experimental and theoretical values is 1.02, with a standard deviation of 8%, it can be seen the good correlation of the theoretical shear yielding approach (equation 2) and the actual shear strength. A more detailed analysis of results suggests that Equation 2 should be multiplied by F slen = 0.9 for shear lug aspect ratio “(H-G)/W” less than 0.7. This way, a better adjustment to the results is reached and the generation of greater steel length embedded in the concrete is privileged in the design.
Load v/s displacement, shear lug N°1. Figure 9.
Comparison—shear steel failure.
Table 5. Comparison between theoretical and experimental failure values—steel failure.
Figure 8.
Shear lug test Nº1.
Test Nº
Ashear (cm2)
Fy (kg/cm2)
Vtheo (ton)
Vtest (ton)
Vtest /Vtheo
1 2 3 4 5 6
9.75 9.75 9.75 15.36 15.36 22.56
3169 3169 3169 3169 3169 3169
18.5 18.5 18.5 29.2 29.2 42.9
20.5 16.7 17.3 30.2 32.4 45.8
1.11 0.90 0.93 1.04 1.11 1.07
7
22.56
3169
42.9
40.6
0.99
250
Vsteel = 0.6 ⋅ Fy ⋅ Ashear ⋅ Fslen
(3)
Table 6.
Fslen = 0.9 if [(H-G)/W] ≤ 0.7
(4)
Test Nº
f c (kg/cm2)
A b (cm2)
Vtheo (ton)
Vtest (ton)
Vtest /Vtheo
8 9
304,2 304,2
35 75
34,5
32,85
0,95
50,5
43,7
0,86
Fslen = 1.0 if [(H-G)/W] > 0.7 5.2
Concrete failure
There were two tests where failure in the concrete occurred, Figure 10 shows such behavior. It was observed that anchor bolts broke in an explosive fashion and there was no a well defined plastic range. There were not external fissures or cracks appearances at the concrete block or grout before the failure, so it was difficult to foresee the concrete failure. The lower ductility suggests avoiding this failure mode for seismic resistant structures. Figure 11 shows the sample with a zone of detached concrete in a triangular fashion still bonded to the shear lug, similar to Rotz’s tests results. Table 6 shows the calculated failure load (V theo) according to Rotz equation 1 (in this case P y = 11.0 ton and Pa = 0.0 ton), they are compared with the actual failure load obtained from the tests (V test).
Table 7.
Failure strength values for concrete bearing.
Shear yielding strength for grout of 40 mm.
Test Nº
Ashear (cm2)
Fy (kg/cm2)
Vtheo (ton)
Vtest (ton)
Vtest /Vtheo
10 11 12
9,75 15,36 22,56
3169 3169 3169
18,5 29,2 42,9
21,9 31,5 33,8
1,18 1,08 0,79
Experimental to theoretical average ratio is 0.91, with a standard deviation of 6%. 5.3
Grout influence on the results
Test Nº 10, 11 and 12 were specially designed to evaluate the impact of grout on the results, table 7 is a summary of the theoretical failure load (V theo) by Equation 2 and the experimental failure load values (Vtest). The failure mode was the steel shear yielding, and it was expected a reduction of system strength due to the thicker grout. However, in two of the three cases the failure was at a higher load than it was expected. The average of (V test /Vtheo) was 1.02, with a standard deviation of 20%. These results indicate that grout would not have an important negative effect on the system strength.
6 Figure 10.
DESIGN CONSIDERATIONS
Comparison—concrete failure.
It is assumed that design is controlled by steel shear yielding and the concrete bearing failure is avoided. 6.1
Steel failure
Table 8, based on equation 3 shows how to determine the cross area of the shear lug.
Table 8.
Steel shear yielding mode for cross shear lug.
Design method ASD A shear ≥ Figure 11.
Concrete bonded to the shear lug.
251
V l g 1/ FS ⋅ (0.6 F y ) ⋅ F slen
Design method LRFD A shear ≥
V l g u φ v ⋅ ( 0.6 Fy ) ⋅ F sle
6.2
Concrete failure
It is necessary to calculate the maximum shear yielding load (equation 3), in order to guarantee that concrete failure will never occur. R y and 1.1 factors, according to the Seismic Provisions for Structural Steel Buildings (AISC, 2002) have been included. They allow estimating the maximum shear load that it is potentially generated in the shear lug when the steel is yielding in shear, so the maximum shear load will be given according to the following formula: Vmax = [1.1 ⋅ R y ⋅ (0.6Fy) ⋅ Ashear ] ⋅ Ωexp
(5)
In order to consider that tests exhibit larger strengths than the initial yielding loads, an over strength factor “Ωexp” has been included in Equation 5. For instance, in the case of figure 7 (test N°1), the initial yielding load was 20.5 ton and it increased as far as 24.3 ton. So for test Nº1: Ωexp = 24.3/20.5 = 1.19. This factor was calculated for all tests, the maximum value was Ωexp = 1.5 (it includes the factor F slen defined in Equation 3). Concrete bearing strength (without considering the grout) is taken from ACI-349 recommendations based on Rotz’s test results. It has to be larger than the strength of the steel shear in yielding, that means that shear lugs geometry, must be selected in a way that the required bearing area between the shear lug and concrete would be as follows: Table 9.
Required bearing area.
Design method ASD Alg ≥
V max 1/ FS ⋅ (1.3 ⋅ f c′ )
Design method LRFD Alg ≥
V max
This project has been sponsored by SKM Minmetal Company in Chile, subsidiary of SKM (Australia), under the Technology and Investigation Program.
Alg Ashear FS Fslen Fy f 'c G H K b K c Mlg Mlgu
φ ⋅ (1.3 ⋅ f c′ )
(6)
This limit is based on the results gotten from experimental tests. A larger aspect ratio should be not appropriate because the theory of uniform bearing stress in concrete could not be valid. It is also important for easier engineering calculation of the concrete bearing capacity (Vconcrete, Equation 1) to use a uniform distribution of bearing stresses in concrete. 7
ACKNOWLEDGEMENT
NOMENCLATURE
From the standpoint of the bearing stress distribution, it is necessary to establish a limit for shear lug aspect ratio “(H-G)/W” (equation 6). [(H-G)/W] ≤ 1.0
1. Two types of failure modes were identified: shear yielding of the shear lug and bearing strength of the concrete. Tests results indicates that flexural yielding mode never happen, so it seems appropriate to discard that mode, as well as the current engineering design practice based on the steel plate acting as a cantilever. 2. Shear yielding mode of the steel exhibit adequate ductility for seismic resistance design. On the other side, bearing failure of concrete presents a less ductile behavior as well as potentially explosive failures. 3. AISC Steel Design Guide N°1 approach is extremely safe because it applies a low allowable bearing stress for concrete 4. The hypothesis of steel plate acting in a cantilever fashion produces larger thickness of the lug plates. It is not clear that this approach assure a ductile failure mode because it is possible to have a brittle failure of concrete first. 5. Design should be based on the shear yielding mode. Table 8 shows how to obtain the cross area of the shear lug under this approach; Table 9 shows the required area to avoid concrete failure.
Py Pa tlg t R y
CONCLUSIONS
There are several aspects evidenced during this research work, to be considered in the design of Shear Lugs.
Vconcrete Vlg
252
Required bearing area of shear lug Cross area of shear lug Security Factor (ASD) Factor for shear lug slenderness “(H-G)/W” Steel yield Stress Concrete compressive strength Thickness of grout Height of shear lug Concrete bearing strength factor (K b = 1.32) Confinement Factor (K c = 1,84 with shear lug) Cantilever end moment of shear lug (ASD) Factored cantilever end moment of shear lug (LRFD) Anchors bolt yield load Axial load applied Thickness of basic shear lug Thickness of shear lug in cross section Ratio of the expected yield Strength to minimum specified yield strength F y, according to table I-6-1 of “Seismic Provisions for Structural Steel Buildings”. R y = 1.5 for steel ASTM A36 Shear capacity of concrete Shear load resisted by shear lug (ASD)
Vlgu Vmax Vsteel W
ϕc τnom φ φv Ωexp
Shear load resisted by shear lug (LRFD) Maximum yield shear load in shear lug Shear capacity of steel shear lug Width of shear lug Resistance factor for bearing on concrete (ϕc = 0,6) Nominal stress in the concrete Concrete Reduction Factor ( φ = 0.70) (LRFD) Steel Reduction Factor (φv = 0.90) (LRFD) Experimental Overresistance Factor. (Ωexp = 1.5 for cross shear lug type)
AISC. 2003. Column Base Plate, AISC Steel Design Guide Series Nº1, EE.UU. INN. 2002. NCh2369-2002—Diseño Sísmico de Estructuras e Instalaciones Industriales, Instituto Nacional de Normalización Santiago, Chile. Rotz, J.V. & Reifschneider, M. 1989. Combined Axial and Shear Capacity of Embedments in Concrete , 10th International Conference, Structural Mechanics in Reactor Technology, Anaheim, Ca.
REFERENCES ACI Committee 349. 2001. Code Requirements for Nuclear Safety Related Concrete Structures, American Concrete Institute, EE.UU. AISC. 2002. Seismic Provisions for Structural Steel Buildings, American Institute of Steel Constr uction, EE.UU.
253