effect of the girder-supported steel plate on stiffened and unstiffened spsws seismic behavior
TRANSCRIPT
Proceedings of the 4th International Conference on Seismic Retrofitting, Tabriz, Iran, 2-4 May 2012
Paper Code. 4116
1
EFFECT OF THE GIRDER-SUPPORTED STEEL PLATE ON STIFFENED
AND UNSTIFFENED SPSWs SEISMIC BEHAVIOR
M. Sartipi, J. Asgari Marnani, A. Ghafooripour
Islamic Azad University, Central Tehran Branch
ABSTRACT
Steel Plate Shear Wall (SPSW) has been known as one of the effective lateral resisting systems in the steel
structures. The classic types of SPSW consist of plate and stiffeners which connected to the surrounding frame, while
using its unstiffened type is more common due to its economical aspects. Optimized usage of strength capacity, high
ductility and energy dissipation in steel plate during the post buckling phase make the SPSW as an effective type. The
separation of steel plate from columns is a useful idea developed in order to reduce column's and foundation's lateral
transferred overturning moments; however it may increase the ductility versus the shear strength.
This paper discussed an investigation trend on the stiffened and unstiffened girder-supported panels subjected to
monotonic in-plan shear and to understand and to avoid plate's edges tearing, the use of two vertical stiffeners at the
ends of plate has been investigated.
Finite element method for mathematical modeling to use the pushover method has been the main mechanism for
simulation the behavior.
The most important parameters such as: Frame aspect ratio, plate width to thickness ratio, covering plate area
and the dimension of stiffeners in these systems has been investigated.
As conclusion, behavior of these types of SPSW were comprised with the classic type to find and publish the
different aspects, and therefore, appropriate behavior of this system has been studied, specially for rehabilitation and
retrofitting the existing buildings and damaged structures.
Key Words: Steel Plate Shear Wall, Girder-supported panels, Surrounding frame, Seismic behavior
1 INTRODUCTION
Steel plate shear walls (SPSW) have been used in buildings as the lateral force resisting
system in last four decades. There have been numerous SPSW research programs to help foster a
better understanding of this system’s behavior, particularly as it relates to earthquake-resistant
design.
The typical SPSW system is comprise of unstiffened or stiffened steel plate connected to both
vertical and horizontal boundary elements. The structural elements usually are designed to carry out
the gravity loading without considering the steel plate's effects in this system. Elements around the
panels consist of SPSW, also should be designed based on the capacity of the steel plate wall due to
the development of its tension-field action. This demand is based on the panel’s aspect ratio, the
steel plate’s thickness and the steel plate’s expected strength [1]. It makes necessary to design
strong columns and large foundations due to their high seismic loadings.
The separation of plate from columns is a useful idea developed in order to reduce the number
of main members in lateral resisting structural system. In 1994, Xue and Lu [2], at Lehigh
University examined analytically the behavior of SPSW system with different connection
configurations. The used models were consisted of three spans and twelve stories frames, with
2
moment resisting beam to column connections in the exterior spans and filled with steel plates in
the middle span. They found that, in the specimen which plate was only connected to beams with
shear connections, the story shear was distributed symmetrically between four columns. Because of
no connection between plate and columns, the girder bending moments was essentially controlled
by the tension field action of the plate. Based on these results, the peak bending moment for this
system moved away from the beam to column connections and forcing the plastic hinges to occur
within the beams. As such, this system was considered more favorable for enhancing the overall
ductility of SPSW while reducing deformation demand on the beam to column connections.
Later, in the second paper, they reported a numerical parametric study conducted on the one
span, one story girder only-supported SPSW model. They found that the variation of width to
thickness ratio has no significant effect on the overall load-displacement response of the structure
whereas the aspect ratio of the panels has a significant effect on the behavior of this system. They
also presented an empirical equation to predict the yield strength, yield displacement and post-yield
stiffness in SPSW system [3].
In 2001, Driver et al. [4] proposed the idea of separation of SPSWs from the moment resisting
frame by inclusion of supplementary columns.
In 2005, Moharrami et al. [5] has been introduced the semi-supported steel shear wall
(SSSW). In this type of shear wall, steel plate does not connect to the boundary columns of the
frame and instead, it is connected to the secondary columns of the structure with no gravity loads.
Results showed that SSSW system has considerable shear capacity; therefore, reduce the size of
main columns. On the other hand, development of plastic hinges in columns not only does not
threaten the stability of the structure, but also increases ductility, energy dissipation and ultimate
deformability of the structure. The secondary columns were used to enable the plate to enter into its
post-buckling behavior and develop tension field action. Other experimental and numerical studies
on SSSW [6],[7] showed that tension field action in the plate can be developed by secondary
columns as well as the traditional type of SPSWs and also, developed a new method for evaluating
the ultimate shear capacity of SSSW.
In this paper, the results of pushover analyses on stiffened and unstiffened SSSW have been
presented, using FEM. First for verifying, the analytical model’s results are compared with the
corresponding experimental data. Then make several models of SSSW system in order to
investigate effect of some parameters such as: frame aspect ratio, plate width to thickness ratio,
covering plate ratio and the dimension of stiffeners. Finally, compare these types of SPSWs with
classic one and identify advantages and disadvantages of them.
2 VERIFICATION
To verify the finite element approach and pushover analyses, the solid steel plate shear wall
which has been tested by Bruneau and Vian [8], modeled by ABAQUS. The experimental specimen
consisted of a steel frame (with 2000 mm height and 4000 mm width) and a steel wall panel as
shown in Figure 1. Beam with reduced beam sections (RBS) details, W18x65, was connected to the
columns, W18x70, such as moment resistance type and also the low yield steel (LYS) plate (with
2.6 mm thickness, 165Mpa yield stress and 300Mpa ultimate strength), fully covered frame span
with continuous connections. The initial yield stress for steel boundary elements had considered
345Mpa to ensure their elastic behavior during test.
3
Figure 1: Solid specimen used by Bruneau and Vian [8]
In numerical model, all of boundary elements and plate are simulated with 4-node shell
elements so that its integration points are reduced to one (S4R). There is 6 degrees of freedom in
each node and large deformation effects, on element formation are considered. The boundary
elements and the LYS plate are modeled with the inclusion of material yielding with bi-linear
material behavior by using the isotropic hardening model available in ABAQUS. Elasticity
modulus and Poisson ratio for all elements considered 210Gpa and 0.3 respectively. The Von-Mises
criterion is used to detect material yielding, and to ensure of statically behavior in analysis, the
kinetic energy value controlled to remain negligible during the analysis.
As shown in Figure 2, there is a good compatibility between of ABQUS/Explicit results with
those of the test shown. ABQUS/Explicit is capable of predicting the monotonic behavior of the
frame-wall panel but there are little differences between ultimate strength of the test and analytical
results due to dispensing with degradation of strength in cyclic loading in analytical models.
Figure 2: Comparison of experimental and pushover analysis results
3 ANALYTICAL MODELS
In order to investigate on the girder-supported SPSWs, several FE models have been
considered. All of analytical models are included a single span and single story steel frame. Girder
plates 550x530x35x20 and 600x430x35x23 are used for beams and columns, respectively [9], and
BOX-100x5 has been chosen for plate's edges stiffeners. Connections are considered rigid and all
members have been selected structural steel property of St37 as shown in Table 1.
4
Table 1 Stress-Strain Values of St37
Stress (Mpa) 240.00 240.00 360.00 370.00 370.00
Plastic Strain 0.001143 0.020 0.180 0.200 0.350
To simulate monotonic loading, a gradual increasing displacement equal to 1.5 times more
than the FEMA’s target displacement suggestion (0.02d) [10], at top of the frame has been induced.
In Table 2, the name of each specimen with its some specification and in Figure 3, several FE
models are shown. For denominating, it's used abbreviated characters: B for box cross section of
plate’s boundary stiffeners, S for stiffened plate, F for steel moment frame and FP for full covered
panel and the first, second and third digits refers to specimen number, covering ratio of the plate
rather than full panel (in percent) and thickness of plate (mm), respectively.
Table 2 Specification of Samples
Plate
Weight(kg)
Plate’s
Stiffener Story Height(m) Span Width(m)
Plate Width (m)
Plate’s Boundary
Stiffeners Specimen
- - 3.00 4.50 - - F1
- - 3.00 6.50 - - F2
- - 3.00 9.50 - - F3
- - 3.00 15.50 - - F4
527.52 - 3.00 4.50 4.00 - FP1
791.28 - 3.00 6.50 6.00 - FP2
1186.92 - 3.00 9.50 9.00 - FP3
1978.20 - 3.00 15.50 15.00 - FP4
527.52 PL-100x10 3.00 4.50 4.00 - SFP1
791.28 PL-100x10 3.00 6.50 6.00 - SFP2
1186.92 PL-100x10 3.00 9.50 9.00 - SFP3
1978.20 PL-100x10 3.00 15.50 15.00 - SFP4
263.76 - 3.00 4.50 2.00 BOX-100x5 B1-50-7
395.64 - 3.00 6.50 3.00 BOX-100x5 B2-50-7
593.46 - 3.00 9.50 4.50 BOX-100x5 B3-50-7
989.10 - 3.00 15.50 7.50 BOX-100x5 B4-50-7
131.88 - 3.00 4.50 1.00 BOX-100x5 B1-25-7
395.64 - 3.00 4.50 3.00 BOX-100x5 B1-75-7
263.76 PL-50x10 3.00 4.50 2.00 BOX-100x5 SB1-50-7
395.64 PL-50x10 3.00 6.50 3.00 BOX-100x5 SB2-50-7
593.46 PL-50x10 3.00 9.50 4.50 BOX-100x5 SB3-50-7
989.10 PL-50x10 3.00 15.50 7.50 BOX-100x5 SB4-50-7
131.88 PL-50x10 3.00 4.50 1.00 BOX-100x5 SB1-25-7
395.64 PL-50x10 3.00 4.50 3.00 BOX-100x5 SB1-75-7
94.20 - 3.00 4.50 2.00 BOX-100x5 B1-50-2.5
188.40 - 3.00 4.50 2.00 BOX-100x5 B1-50-5
263.76 - 3.00 4.50 2.00 BOX-100x5 B1-50-7
376.80 - 3.00 4.50 2.00 BOX-100x5 B1-50-10
828.96 - 3.00 4.50 2.00 BOX-100x5 B1-50-22
94.20 PL-50x10 3.00 4.50 2.00 BOX-100x5 SB1-50-2.5
188.40 PL-50x10 3.00 4.50 2.00 BOX-100x5 SB1-50-5
263.76 PL-50x10 3.00 4.50 2.00 BOX-100x5 SB1-50-7
376.80 PL-50x10 3.00 4.50 2.00 BOX-100x5 SB1-50-10
828.96 PL-50x10 3.00 4.50 2.00 BOX-100x5 SB1-50-22
263.76 PL-50x15 3.00 4.50 2.00 BOX-100x5 *SB1-50-7
263.76 - 3.00 4.50 2.00 BOX-200x10 B5-50-7
263.76 - 3.00 4.50 2.00 BOX-150x5 B6-50-7
263.76 - 3.00 4.50 2.00 BOX-80x5 B7-50-7
5
(a) (b) (c)
(d) (e)
Figure 3: Some Specimens Consist of SPSW in FE Models
(a) F1, (b) FP1, (c) B1-25-7, (d) SFP1, (e) SB1-50-10
4 DISCUSSION AND EVALUATION OF RESULTS
Shear force - displacement response curves, result of finite element analysis, have been shown
and compared in Figures 3 to 8.
Figure 3 Figure 4
Figure 3: Comparison of Specimen's behaviour, in Different Full SPSW’s Aspect Ratio (left)
Figure 4: Comparison of Specimen’s behaviour, in Different Steel Moment Frame’s Aspect Ratio
(right)
6
Figure 5 Figure 6
Figure 5: Comparison of Specimen’s behaviour in Different Half-Covered SPSW’s Aspect Ratio
(left)
Figure 6: Comparison of Specimen’s behaviour in Different Plate Thickness (right)
Figure 7 Figure 8
Figure 7: Comparison of Stiffened Specimen’s behaviour in Different Half-Covered SPSW’s
Aspect Ratio (left)
Figure 8: Comparison of Specimen’s behaviour in Different Covering Area Ratio (right)
In order to avoid of complicating of curves, values of ultimate strength, initial stiffness and
other specification, have been arranged and shown in Tables 3 to 10.
Table 3 Comparison of Steel Moment Frame Specimens in View Point of Aspect Ratio
Rational Increase of Ultimate
Strength )%(
Ultimate Strength
(KN)
Rational Increase of Initial
Stiffness )%(
Initial Stiffness
(KN/mm) specimen
0.00 3383 0.00 149.052 F1
-4.17 3242 -2.20 145.766 F2
-6.30 3170 -14.13 127.995 F3
-17.17 2802 -36.73 094.300 F4
As shown in Tables 3 and 4, in case of increase frame's span length, both ultimate strength
and initial stiffness values have been decreased. While, in systems consist of full steel plate,
increasing in aspect ratio (w/h) and plate width to thickness ratio (slender ratio), improves their
behaviours. It has been shown more appropriate behaviour for stiffened types.
Forc
e (K
N)
Forc
e (K
N)
7
Table 4 Comparison of Classic Steel Plate Shear Wall Specimens in View Point of Aspect Ratio
Rational Increase of
Ultimate Strength )%(
Ultimate
Strength (KN)
Rational Increase of
Initial Stiffness )%(
Initial Stiffness
(KN/mm)
Width to plate
thickness ratio specimen
0.00 6371 0.00 283.976 571.43 FP1
26.04 8030 27.92 363.258 857.14 FP2
64.04 10451 70.19 483.309 1285.70 FP3
139.82 15279 153.45 719.748 2142.85 FP4
6.29 6772 9.98 312.326 571.43 SFP1
30.98 8345 38.99 394.698 857.14 SFP2
72.50 10990 83.04 519.789 1285.70 SFP3
152.82 16107 168.55 762.627 2142.85 SFP4
Utility of half-covered steel plate, also improves the seismic behaviour of systems (Tables 5
and 6). In addition, utility of two vertical and horizontal plate on both side of panel (with thickness
equal or more than steel shear wall thickness) increases both ultimate strength and initial stiffness
values of SPSW.
Table 5 Comparison of Half – Covered Specimens in View Point of Aspect Ratio
Rational Increase of
Ultimate Strength )%(
Ultimate
Strength (KN)
Rational Increase of
Initial Stiffness )%(
Initial Stiffness
(KN/mm)
Width to plate
thickness ratio specimen
0.00 4544 0.00 205.737 285.71 B1-50-7
18.05 5364 19.12 245.075 428.57 B2-50-7
41.59 6434 39.35 286.689 642.86 B3-50-7
97.89 8992 104.81 421.360 1071.43 B4-50-7
7.50 4885 8.85 223.945 285.71 SB1-50-7
21.43 5518 24.94 257.055 428.57 SB2-50-7
46.57 6660 46.44 301.281 642.86 SB3-50-7
105.45 9336 131.67 439.598 1071.43 SB4-50-7
Table 6 Comparison of Half-Covered and Classic SPSW Specimens with Different Aspect Ratios
Rational degradation of Ultimate Strength
corresponding full plate (%)
Rational degradation of Initial Stiffness
corresponding full plate (%) specimen
28.68 27.55 B1-50-7
33.20 32.53 B2-50-7
38.44 40.68 B3-50-7
41.14 41.46 B4-50-7
23.32 21.14 SB1-50-7
29.24 31.28 SB2-50-7
36.27 37.66 SB3-50-7
38.90 38.92 SB4-50-7
On the other hand, as seen in Table 7, dimensions of plate stiffener, specially its thickness,
have no significant effect on improving seismic behaviour of the system. This result indicates that
unstiffened SSSW is an economic and well behaviour selection. Because of limitation in boundary
stiffener’s size (BOX-100x5), usage of wide plate following equations of choice stiffener’s
dimension of classic SPSW [9] have been impossible.
8
Table 7 Comparison of Half-Covered Specimens in View Point of Dimension of BOX Section
Ultimate Strength (KN) Initial Stiffness (KN/mm) specimen
4900 224.123 *SB1-50-7
4885 223.945 SB1-50-7
Seismic behaviour of the system, as shown in Table 8, improves in case of increase plate
thickness, increase covering plate area ratio and utility plate’s stiffeners. In addition, utilizing half-
covered and unstiffened system with 22mm plate thickness, cause to about 35% growth in ultimate
strength and to increase initial stiffness to 30% more than correspondent full panel with 7mm plate
thickness. It allows having large openings in plate near the columns.
On the other hand, stiffened specimen with 22mm thickness has been shown different behaviour.
Because of high thickness rather than plate stiffener, plate has no out-of-panel deformation (Figures
6, 9). It had reduced ultimate strength of specimen. In case of SB1-50-10, there is such behaviour.
Generally, in order to have general buckling in stiffened SPSW, thickness of stiffener should be
more than plate [9].
Table 8 Comparison Behaviour of Net and Stiffened Edge Plates with Steel Moment Frame
Rational Increase of Ultimate
Strength )%(
Ultimate
Strength (KN)
Rational Increase of Initial
Stiffness )%(
Initial Stiffness
(KN/mm) specimen
0.00 3383 0.00 149.052 F1
16.58 3944 16.93 174.283 B1-50-2.5
29.23 4372 30.98 195.222 B1-50-5
34.31 4544 38.03 205.737 B1-50-7
54.44 5225 63.40 243.548 B1-50-10
155.04 8628 148.15 369.867 B1-50-22
18.68 4015 19.80 178.559 SB1-50-2.5
33.52 4517 36.64 203.665 SB1-50-5
44.40 4885 50.25 223.945 SB1-50-7
57.85 5340 74.50 260.096 SB1-50-10
145.97 8321 142.25 361.082 SB1-50-22
Figure 9: General and Local Buckling in Stiffened Specimens with Different plate thickness
Utility of two horizontal stiffeners in SB1-25, has no significant effect on its behaviour,
Therefore, un-stiffened cases, are more favourable due to their simple and economic application.
Forc
e (K
N)
9
As seen in Table 9, dimensions of the box area section, as boundary stiffeners of semi-
covered plate, have no significant effect on improving seismic behaviour of the system. This result
allows designers to select economic sections.
Table 9 Comparison of Half-Covered Specimens in View Point of Dimension of BOX Section
Rational Increase of
Ultimate Strength )%(
Ultimate
Strength
(KN)
Rational Increase of Initial
Stiffness )%(
Initial
Stiffness
(KN/mm)
Stiffener
weigh(kg) specimen
0.00 4469 0.00 203.132 55.68 B7-50-7
1.68 4544 1.28 205.737 70.56 B1-50-7
4.92 4689 4.59 212.464 108.48 B6-50-7
15.35 5155 14.31 232.207 282.24 B5-50-7
Finally, as the most important results, so that shown in Table 10, separation plate from
columns, give us convenient benefits such as: make opening possibility, gain rather than 65%
ultimate strength of the full panel and save more than 75% in plate material consumption. This
values increase with appropriate stiffened plate, too.
Table 10 comparison of stiffened edge plates with full panel in view point of covering plate’s area
Rational Ultimate Strength Ultimate
Strength (KN) Rational Initial Stiffness
Initial Stiffness
(KN/mm) specimen
1.00 6371 1.00 283.976 FP1
0.86 5499 0.91 257.481 SB1-75-7
0.83 5317 0.85 241.852 B1-75-7
0.77 4885 0.79 223.945 SB1-50-7
0.71 4544 0.72 205.737 B1-50-7
0.67 4260 0.68 192.719 SB1-25-7
0.62 3971 0.64 182.385 B1-25-7
0.53 3383 0.52 149.052 F1
5 CONCLUSIONS
1. In contrast to steel moment frame, both ultimate strength and initial stiffness values of a full
steel plate shear wall have been significantly improved with increase of aspect ratio (w/h)
and plate width to thickness ratio.
2. The behavior of the un-stiffened half-covered steel plate shear wall system, in view point of
aspect ratio (w/h) and plate width to thickness ratio, is similarly full panel but with less than
about 40% of degradation.
3. Increasing plate thickness of semi covered specimens has no significant effect on seismic
behavior improvement of this type of SPSW.
4. Thickness of the plate’s stiffener has no significant effect on seismic behavior improvement.
To have good behavior of stiffened system, thickness of stiffeners should be more than steel
plate.
5. Size of the box area section, as boundary stiffeners of semi-covered plate, has no significant
effect on seismic behavior improvement.
6. Seismic behavior of girder-supported system will improve by increasing of the plate
thickness, increasing covering plate area ratio and utility of boundary stiffeners on plate's
edges.
7. To have large openings on the panel with high performance for SPSW behavior, it can be
selected thicker semi- covered panel instead of full thin plate.
10
8. It can be useful to rehabilitation and retrofit the existing buildings and damaged structures
using girder–supported SPSW.
REFERENCES
[1] Seilie, L. and Hooper, J. [2005]. “Steel plate shear wall: practical design and construction.”
North American steel construction conference, Modern steel construction.
[2] Xue, M. and Lu, L-W. [1994]. “Influence of steel shear wall panels with surrounding frame
members.” Proc. SSRC annual technical session. pp. 339–54.
[3] Kharazi, H. K. M. [2005]. “Rational method for analysis and design of steel plate shear walls.”
Doctor of philosophy thesis, the University of British Columbia.
[4] Driver, R. G. Grondin, G. Y. Behbahanifard, M. Hussain, M. A. [2001]. “Recent
developments and future directions in steel plate shear wall research.” Proc. American steel
construction conference.
[5] Moharrami, H. Habibnejad, A. Mazrouei, A. Alizadeh, H. [2006]. “Semi-supported thin steel
shear walls.” Research report no. 1– 4679. The Building and Housing Research Centre.
[6] Habibneghad, A. [2004]. “Behavior of semi-supported thin steel shear walls under lateral
loads,” M.Sc. thesis. Tehran (Iran): Tarbiat Modares University.
[7] Jahanpour, A. Moharrami, H. Aghakoochak, A. [2011]. “Evaluation of ultimate capacity of
semi-supported steel shear walls.” Proc. Journal of Constructional Steel Research.
[8] Vian, D. and Bruneau, M. [2004]. “Testing of special LYS steel plate shear walls.” Proc. 13th
World conference on Earthquake Engineering, Vancouver, B.C., Canada, Paper No.978.
[9] Sabouri-Gomi, S. [2001]. “Lateral load resisting systems, an introduction to steel shear walls.”
Angizeh broadcast.
[10] Federal Emergency Management Agency (FEMA), [2000]. Prestandard and Commentary for
the Rehabilitation of buildings, FEMA-356.