Jordan Journal of Civil Engineering, Volume 6, No.

4, 2012

Pushover Analysis for Cold Formed Storage Rack Structures

Sreedhar Kalavagunta 1), Sivakumar Naganathan 2) and Kamal Nasharuddin Bin Mustapha 2)
1)
Bentley Systems Limited, Singapore,
Corresponding Author. E-Mail: sreedhar.kalavagunta@bentley.com
2)
Department of Civil Engineering, Universiti Tenaga Nasional, Jalan IKRAM-UNITEN, 43000,
Kajang, Selangor, Malaysia

ABSTRACT
This study investigated the progressive collapse of cold formed storage rack structures subjected to seismic
loading, using pushover analysis. A simple storage rack cold formed steel structure was analyzed with static,
non-linear procedure in accordance with FEMA 356 specifications and progressive collapse recorded such as
occupancy, immediate occupancy, life safety and collapse prevention. Pushover analysis was found to be a
useful analysis tool for the conventional storage racking systems giving good estimates of the overall
displacement demands, base shears and plastic hinge formation.

KEYWORDS: Cold formed steel, Pushover analysis, Plastic hinge, FEMA, ATC, Capacity.

INTRODUCTION 2000) is now used by the structural engineering
profession as a standard tool for estimating seismic
While conventional limit-states design is typically a demands for buildings. In the past, several researchers
two-level design approach having concern in the have discussed the underlying assumptions and
service operational and ultimate-strength limit states limitations of pushover analysis (Krawinkler and
for a building, performance-based design can be Seneviratna, 1998; Elnashai, 2001) and proposed
viewed as a multi-level design approach that adaptive force distributions that attempt to follow the
additionally has explicit concern in the performance of time-variant distributions of inertia forces (Bracci et
a building at intermediate limit states related to such al., 1997; Gupta and Kunnath, 2000). They considered
issues as occupancy and life-safety standards. With the more than the fundamental vibration mode (Sasaki et
emergence of the performance-based approach in al., 1998; Kunnath and Gupta, 2000; Matsumori et al.,
design, there is a need to develop corresponding 1999). Anil K. Chopra reviewed the above papers and
analysis tools. Nonlinear static (pushover) analysis is proposed a modal pushover analysis procedure to
often an attractive choice in this regard because of its estimate seismic demands for asymmetric-plan
simplicity and ability in identify component and buildings (Chopra and Goel, 2004).
system-level deformation demands with accuracy
comparable to dynamic analysis (Hasan et al., 2002; Fu DEFINITION
and Alayed, 2006). The non-linear static procedure
(NSP) or pushover analysis, as described in FEMA-273 Pushover Analysis: is a static, nonlinear procedure
(BSSC, 1997) and its successor FEMA-356 (ASCE, in which the magnitude of the structural loading is
incrementally increased in accordance with a certain
Accepted for Publication on 13/8/2012. predefined pattern. In effect, the structure is pushed

- 489 - © 2012 JUST. All Rights Reserved.

in compression each of its principle directions to stimulate the loading and bending. performance by estimating the strength and • Verification of the adequacy of the load path. both and comparing these capacities with the demands at the structural and nonstructural. controlled. • Identification of the strength irregularities in plan Purpose and Methodology: The purpose of or elevation that cause changes in the dynamic pushover analysis is to assess the structural characteristics in the inelastic range. Ungureanu and Dubina. The basic procedure of this method is to perform a sequence of elastic static Development of Plastic Hinges: Short members of analysis under monotonically increasing lateral loads in thin-walled cold formed steel sections. 2000). strength and stiffness discontinuities. 2011.. In this way. the collapse of slender decade and it has found its way into seismic guidelines members. • Capacity Spectrum Method (CSM). analysis (Zeynalian and Ronagh. Poursha et al. deformation capacities using static. A proper Analysis Highlights: The pushover is expected to identification of the geometry of a local plastic provide information on many response characteristics mechanism is crucial for correct evaluation of the post that can’t be obtained from an elastic static or dynamic buckling rigid-plastic curve (failure curve). local plastic mechanism failure mode (Kotelko et al. hence. A collection of plastic moment demands on beam to column connections mechanisms of failure for thin-walled cold formed steel or shear demands on short. CEN overall buckling modes. nonlinear analysis considering all the elements of the system. The following are examples of such response subsequently results in a correct estimation of the load- characteristics (Bajoria and Sangle. 1995. PUSHOVER ANALYSIS • Estimates of the deformation demands on elements that have to deform in elastically. . corresponding performance levels. shear-dominated members subjected to compression and bending are elements. 2000. is always characterized by 1995]. Dubina elements. 2008): carrying capacity of the member. 2000.. due to the interaction between local and [FEMA 356. Sivakumar Naganathan and Kamal Nasharuddin Bin Mustapha sideways well into the inelastic range till total failure or progress of the overall capacity curve of the collapse occurs. 1981). fail by forming local plastic mechanisms history of the structure during collapse. There are two non-linear procedures using pushover damage on nonstructural elements can be methods. which analysis. • Sequence of members yielding and failure and the • Displacement Coefficient Method (DCM). 1997. SEAOC. ATC. Inel and Ozmen. where the inelastic deformations are expected to be high. 2001). in order to Theory: Nonlinear static (pushover) analysis is dissipate energy. 2004.490 - . often an attractive choice in this regard because of its • Consequences of strength deterioration of simplicity and ability to identify component and particular elements on the overall structural system-level deformation demands with an accuracy stability. • Identification of the critical regions. tabulated [Table 1]. as well as of the • Realistic force demands on potentially brittle energy absorption at collapse (Kotelko. comparable to dynamic analysis (Bentley Reference • Estimates of inter-storey drifts. 2006).Pushover Analysis… Sreedhar Kalavagunta. such as axial demands on columns. The potential (Murray and Khoo. Taking into account the of the pushover analysis has been recognized in the last localized buckling pattern. and Ungureanu. this method is called pushover structure. 2004). accounting for Manual.

Energy method app plied . 2012 2 Table 1. Strip 1 plaane) equilibrium m method applied Kotelko (2004).Jordan Journ nal of Civil Eng ngineering. No.. 2010) No Beeam cross-con nnection Mech hanism References Murray an nd Khoo Chhannel (bending in the flange f (1981). 3 section Energy method applied Kotelko (1996). Kotelko (1996). 6 Trriangular crosss-section Energy method applied Elchalakan ni et al.491 - . Reectangular orr trapezoidal box- 4 Energy method section applied Kotelko (1996). (2002 2). Chhannel (bendding in the web 2 Energy method plaane) applied Kecman (1983). 1 Plastic mecchanisms (Un ngureanu et al. Reectangular orr trapezoidal box.. (2002). 4. Elcchalakani 7 Ciircular hollow w section et al. Vollume 6. 5 Trrapezoidal boxx-section Energy method applied Kotelko and Krolak (1993).

or coefficients. stiffness degradation to building roof displacement.3. Sa = Response spectrum acceleration.5 for T e < 0. For larger elements or entire displacement which is the maximum displacement that structural systems composed of many the structure is likely to experience during the design components. Ki = Elastic lateral stiffness of the building.3. curve is by tracking the base shear and roof Te = Effective fundamental time period = Ti(Ki/Ke)1/2.2 of FEMA 356: building (to be calculated from demand 2000 spectrum). It provides a numerical process for the point at which a sufficient number of estimating the displacement demand on the structure.0 for T e ≥ T s appropriate demand and capacity spectra for the = [ 1. Ti = Elastic fundamental period. During this process. Alternatively. component is also evaluated. δt = C0C1C2C3Sa[Te2/(4π2)]g (1) Cm = Effective mass factor as determined by Table 3-1 of FEMA 356. taken as equal to the secant stiffness calculated at a base shear force equal to 60% of the effective yield strength of the structure obtained from bilinear representation of capacity curve. Value of C1 should not be less than 1. to inelastic deformation. = 1. taken from Table 3-3 of C1 = Modification factor to relate expected maximum FEMA 356 for different framing systems and inelastic displacements to displacements structural performance levels.0. structure and to determine their intersection point. as determined by and strength deterioration on maximum Table 3-2 of FEMA 356. performance of each structural Ts = Characteristic period of the response spectrum.0 for a nonlinear procedure.492 - . The structure capacity is defined as the period associated with transition represented by a pushover curve. The spectrum (to be calculated from demand most convenient way to plot the force displacement spectrum). calculate a target displacement. displacement response. often termed as from constant acceleration segment of the capacity curve. where: C2 = Modification factor to represent the effect of C0 = Modification factor to relate spectral displacement pinched hysteretic shape. Ke = Effective lateral stiffness of the building. C2 calculated for linear elastic response may be taken as 1.Pushover Analysis… Sreedhar Kalavagunta. individual components or elements have yielded by using a bilinear representation of capacity curve and and the global structure begins to experience a series of modification factors.1 sec C3 = Modification factor to represent increased .3.1) T s / T e ] / R for T e < T s. displacement (Figure 1).0 + (R . This represents the lateral displacement spectrum to the constant velocity segment of the as a function of the force applied to the structure. Vy = Effective yield strength calculated using the Displacement Coefficient Method is to find target capacity curve. base shear-V strength co-efficient = Sa/(Vy/W)Cm. Sivakumar Naganathan and Kamal Nasharuddin Bin Mustapha Capacity Spectrum Method is to develop = 1. the effective yield point represents earthquake. W = Effective seismic weight . at the effective The target displacement is calculated as per fundamental period and damping ratio of the procedure described in Section 3. R = Ratio of elastic strength demand to calculated yield Figure 1: Roof deflection-∆roof plotted vs.

For flexural actions of beams and columns. For columns: θy = Z·FyeLc /(6·EIc) (1 . stiffness. plastic hinge. • Point B represents yielding. QCE refers to the plastic moment capacity. ∆ is the total relation shall be characterized by a bilinear elastic and plastic displacement and ∆y is the yield relation. The parameters Q and QCE (Qy) are generalized = 1. No deformation occurs For columns: in the hinge up to point B.Jordan Journal of Civil Engineering. the plastic hinge will drop = 1.yield stiffness to effective elastic total elastic and plastic rotation of the beam or column. displacements at points C. For beams and columns. the component Q = Generalized component load. θ is the α = Ratio of post .18·Z·Fye (1 . Only plastic Fye = Expected yield strength of the material. P = Axial force in the member. strength. (2) (Figure 2). Qy. 2012 displacement due to dynamic P-D effects greater than point E. At deformation statistical mean value of yield strengths. the hinge. respectively. For braces.2. Effective expected strength.P/Pye) (5) deformation value specified for point B. responds with substantial substantial strength to QCE = Generalized component expected strength = point E. where the non-linear force-displacement θy is the rotation at yield. (4) • Point A is the origin.2. which shall be Figure 2: Generalized force-deformation calculated using equations (4) and (5): relationship for components For beams: QCE = MCE = Z·Fye. • Point D represents the residual strength of the Pye = Expected axial yield force of the member = AgFye.493 - .P/Pye). • Point C represents the ultimate capacity of the Lc = Column length. Lb = Beam length. plastic hinge.1)3/2/Te for buildings with component load and generalized component expected negative post . which is defined as the • Point E represents total failure. (3) Q and QCE are the generalized component load and generalized component expected strength. regardless of the QCE = MCE = 1. until it reaches point D. respectively.yield stiffness. degradation begins (hinge starts shedding load) kip-in. 4. deformation beyond point B will be exhibited by I = Moment of inertia. for a . At this point hinge strength MCE = Expected flexural strength of a member or joint. D and E.2 θy = Z·FyeLb /(6·EIb).5. displacement. The displacement (rotation or axial elongation as the where: case may be) will be subtracted from the E = Modulus of elasticity. θy shall be calculated based on equations (2) and (3). No.yield load to zero. Volume 6. stiffness.0 for buildings with positive post . Frame Element Hinge Properties Discrete hinge properties for frame elements are For beams: based on FEMA-356 criteria as per Section 5.0 + |α|(R . Beyond point D.

base shear Table 2.067 0.displacement at control joint vs.057 1.… etc.Pro.837 0. IDARC.837 kip. Member length is 24 inch of wide flange of most of the general purpose finite element analysis W16X77 standard steel section with an expected yield software like STAAD. manual calculation Deflection (free end) in Percent Load step Force (free end) kips STAAD. I = 138. and capacity curve created with base shear Vs the proposed pushover analysis for three-dimensional displacements (Figure 3). storage rack structure used STAAD. and includes the 356: 2000 and ATC 40 codes. Software results vs. a transverse load P is applied to a section development. strength of 36 ksi. ETABS. l = 24 in.32 1.94% non-linear At load step 3 δz = PL3/3EI +PL/GAv Equation of elastic deflection at cantilever tip: P = 57.Pro results Hand calculations difference 3rd step 57.494 - .0 . STAAD. Sivakumar Naganathan and Kamal Nasharuddin Bin Mustapha population of similar components.Pro results are validated interval of 0. SAP2000. The analysis results are saved at an NISA-CIVIL. cantilever member and increased until the member Validations: Pushover analysis is implemented in fails.078 0% linear 14th step 79.078 0.1 inch deflection of the cantilever tip and with hand calculations based on the above theory. consideration of strain hardening and plastic For validation. E = 29000 ksi.Pro as per FEMA Figure 3: Capacity curve .Pushover Analysis… Sreedhar Kalavagunta.

δz elastic = [57.122 inch. base shear vs.837x 243 / (3 x 29000 x 138)] + Figure 4: Typical storage rack cold formed steel structure with loading Figure 5: Capacity curve.012 = Thus. Volume 6. Av = 10. G = 11154 ksi. roof displacement At load step 14 δz = PL3/3EI +PL/GAv Equation of elastic deflection at cantilever tip: P = 79.Jordan Journal of Civil Engineering.495 - .32 kip.0 in4.784 x24 / (11154 x 10.1097 + 0.4323 in2. l = 24 inch. E = 29000 ksi. [48. No. 4. elastic deformation: 0. 2012 in4.4323)] = 0. . I = 138.

δz elastic= [79. Sivakumar Naganathan and Kamal Nasharuddin Bin Mustapha G = 11154 ksi.107 Thus.924. total collapse of structure (CP) at 29.9 kN . elastic deformation at cantilever tip: inch. 34/ (11154 x 10.Pushover Analysis… Sreedhar Kalavagunta.4323)] = 0.32 x 303 / (3 x 29000 x 138)] + [79.091 + 0.32 x Figure 6: Plastic hinge -IO at base shear of 20.4323 in2.496 - .67 kN and LS at 22.07 kN Figure 7: Plastic hinge formation of LS-CP at base shear of 27. Av = 10.016 = 0.

DESCRIPTION OF STRUCTURE Since STAAD.Jordan Journal of Civil Engineering.Pro software results are C350X125X3 used for columns. static seismic method UBC Support reactions at node 1 Pushover analysis UBC 1997 Difference FX (kip) 0. structure and dead load of 2 kN/m are considered as permanent gravity loads (Figure 4).143 1.96 inch. Lysaght cold formed channel 1. width.497 - .178 1.96 = is taken as example.316 0% Plastic rotation = 0. section C230X76X1.067 inch. 4.073 0.107 + 0.073 0% FY(kip) 2. Volume 6. plastic deformation at cantilever tip A typical four storey storage steel rack of 3 meters δz plastic = l x θ = 24 x 0.5% MY (kNm) 0. 2012 Figure 8: Analysis of four storey storage steel rack as per UBC 1997 Table 3: Pushover analysis vs.5 is used for beams and Comparison: STAAD.7% FZ(kip) 0. 6 meters length and 3 meter height of each floor Total deflection = δz elastic + δz plastic = 0.141 0. No.316 0.000 0.161 1.00 0% MZ (kNm) 0.4% MX (kNm) 1.473 1.Pro assumes small displacements (sinθ = θ).515 2.04 = 0. Input is considered . Self-weight of the close to manual calculations (Table 2).04 radian.

Plastic hinge formation starts with (Collapse Prevention (CP) . Washington. but possibly not useful until repair) is formed at a load typically for each step in the pushover analysis as per of 20. amount of damage in this direction will be limited. FEMA-356.1 inch. base shear vs. number of pushover loads is set as 1000 subjected to equivalent static earthquake loading. The capacity curve. K. next stage of plastic hinge (Life Safety FEMA 356 and ATC 40. The capacity curve (Figure 5) associated with each modal progressive collapse of storage rack initial plastic hinge pushover is determined based on the work conducted in (Immediate Occupancy (IO).4. conventional pushover analysis exhibits (Figures 6 and 7). Bajoria.3.2.07 kN. respectively.1 is class 1. the 29. The being a static method.67 kN.67 kN and 4.building is safe to occupy the analysis. then propagates to upper stories and continues load of 29. Federal Emergency 14th World Conference on Earthquake Engineering. Sivakumar Naganathan and Kamal Nasharuddin Bin Mustapha as moment frame.system is on the verge of beam ends and later proceeds to base columns of lower experiencing partial or total collapse) is formed at a stories. But. Capacity-based American Society of Civil Engineers. the demonstrated by a number of researchers. vertical distribution of base The paper has presented a simple computer-based shear is due to method 1 as per section 3. the support reactions of the four storey response. (LS) . REFERENCES Management Agency. plastic hinge formation of the structure may collapse partially or totally) is formed at building mechanisms has been obtained at different a load of 22. next stage of plastic hinge displacement levels. but the In the Z direction. provides valuable information for the performance- mapped spectral acceleration at short period is 0.1. . site category as per section frame structure. Expected yield stress is 36 kip/in2.25.3.6. with the yielding of interior intermediate beams However.9 kN at a displacement of 10. as CP (collapse prevention-Figure 7). and be considered as a valid alternative to nonlinear time the results are close to pushover analysis (Table 3). energy dissipation that are associated with dynamic However. and Sangle. displacement graph. Prestandard design of cold formed storage rack structures under and commentary for the seismic rehabilitation of seismic load for rigid and semi-rigid connections. The buildings.9 kN. neglecting other sources of at 20.498 - . base shear incremental value is 5 kip. CONCLUSIONS displacement incremental value is 0. The work is computed incrementally.81 inch.Pushover Analysis… Sreedhar Kalavagunta. such as kinetic energy and viscous damping. pushover analysis reproduces first hinge formation of base shear and displacement is material straining only. K. 1. storage steel rack is compared with static seismic as well as duration effects.3 of method for pushover analysis of steel storage rack FEMA 356.structure may be safe in the event. 2008.67 kN. since yielding occurs at also shortcomings and limitations that confine its range designing events B (yielding).M. DC. A with defined base shear of 50 kip.K. clearly indicates a collapse load of In the energy-based pushover approach. 1997 (Figure further research work is required before pushover can 8) at first hinge formation base shear of 20. This clearly indicates that method Uniform Building Code (UBC).1. Spectrum typical example illustrated that pushover analysis parameters: critical damping (%) of 1st spectrum is 5. history analysis. 2000. at based seismic analysis of steel storage rack moment- one-second period is 0. IO (Immediate of application and raise doubts about its effectiveness occupancy). LS (Life safety-Figure 6) and last hinge to accurately estimate structural seismic demand.1 inch. Moreover.

H. Proceedings of the U. Kunnath. and Goel. D. N. M. Kotelko. 79-94.B. 123: 3-10. buildings. Shiohara. Earthquake member deformation demands in Elnashai. and Structural Dynamics. Engineering (ASCE).E. and Zhao. 2417–2434. H. Engineering Structures. Load-capacity estimation and collapse DC. structures. Structural the U. 194. L. subjected to large deformation pure bending. 15 (2): 127-141.W.H.. X.. M. X. Beijing. T. Some basic Engineering Conference. . F. and Rhodes.walled steel Dead Sea. Kecman. Thin-walled Structures. China. 23 (12): 703-713.S.K. and Alayed. and Ungureanu.L. S. Thin-walled Structures. Post-failure interactive buckling of thin-walled steel compression of box sections beams under pure bending (an members. using pushover analysis approach. A. Federal Emergency Management Agency. S. and Ozmen. Validity of deformation in Structural Engineering—An International Journal. 2000. B. V.. Collapse behavior of unsymmetric-plan buildings. 2006. 1983. bending.S. T. and Reinhorn. 1998. 4th Jordanian Civil Murray. and Krolak. R.. 5 demand estimates using nonlinear static procedures. 2002. Bending collapse of rectangular and Seismic performance and retrofit evaluation for square section tubes. 1981. Performance-based Engineering for R/C Building Plastic mechanism analysis of circular tubes under pure Structures.S. T.499 - .section beams under pure bending.S. Zhao. and Moghadamc. M. April 10-13. FEMA-273. Advances Kunnath.–Japan Workshop on Performance-based Engineering and Mechanics. Otani.C. 28: 1494-1502. Matsumori. M. 223-237. 2000.. pushover procedure for seismic evaluation of Poursha. 12 (1): 51-69. 38: 179- conference on cold formed steel structures.K. R. 44 (2): 229-251. A consecutive modal pushover procedure for Hasan. 20: 452-464.K. 2004. St.–Japan Workshop on Elchalakani.. structures. Inel. Seismic analysis of bridges Structures. Lim. 2006. (4): 241-257. and Paret. Bracci. S. Hokkaido. H. No. 1997. K. properties in nonlinear analysis of reinforced concrete Bentley Reference Manual on STAAD. Thin-walled Structures. S. Krawinkler. analysis of thin-walled beams and columns—recent Chopra. Louis. 80: 2483-2493. M. cons of a pushover analysis of seismic performance Plastic collapse analysis of slender circular tubes evaluation. 1993. 33: Computers and Structures. tall building structures. Xu. 44 (6): 1117-1143.M. Grzebieta. V8i. Dubina.L. Earthquake Engineering triangular cross-section girders subject to pure bending. Elastic–plastic Kotelko. and Seneviratna. for the seismic rehabilitation of buildings. 25 (9-10): 623-636. Pros and Elchalakani. 4. Engineering Structures. Engineering Structures. 16: 367-392. International Journal of Mechanical Science. and Kunnath. 2000. 1997. Japan. S.A. Pushover nonlinear static analysis of one-way unsymmetric-plan analysis for performance-based seismic design. Amman and the mechanisms in the local buckling of thin. J. Proceedings of analysis for earthquake applications. 1998.K. B. 2004. 2002.D. EngineeringTransactions. A modal pushover advances.F. 33: 903-927. USA. Effects of plastic hinge Sasaki. D. and Grzebieta. 1999. Maui. T. 2012 October 12-17. Volume 6.K. Ultimate load and post-failure behavior Building Seismic Safety Council. M. NEHRP guidelines of box.S. M. and Khoo.and Grierson. 42 (2): 153-175. Hawaii. Advanced inelastic static (pushover) reinforced concrete frame structures. Washington. International Journal of reinforced concrete structures.. C. D. Kotelko. M. P. R.H.. Khoshnoudianb. and Gupta. Sapporo. A. Jordan. M.M. International Journal of Mechanical Gupta.K.K. and Kabeyasawa. A. Earthquake Engineering Methodology for R/C Building Fu. G.Pro. R.Jordan Journal of Civil Engineering.P. analysis procedure to estimate seismic demands for Kotelko. 2011. 2001. Journal of Structural Mechanical Sciences. J. H. H. A. 2002.. Earthquake Spectra. Proceedings of the 15th international experimental study). 2000. 1996. Adaptive spectra-based Sciences. Freeman.

1525. H. 48: 818-826. 2011. Part I: Plastic–elastic steel shear walls. V..500 - . Thin-Walled Structures. 2004. Recent research on seismic characteristics of knee-braced cold formed advances on ECBL approach. V. Zeynalian. Proceedings of the 6th U. Mania. Thin-walled Structures.S.R. M. Kotelko. Thin-Walled Structures. Seattle. Sivakumar Naganathan and Kamal Nasharuddin Bin Mustapha Multimode pushover procedure (MMP): a method to Ungureanu. . 49: 1517- interactive buckling of cold-formed steel sections. bending. 42 (2): 177-194. A numerical study Ungureanu. D.J. National cold-formed steel members in compression and Conference on Earthquake Engineering. and Dubina. M.Pushover Analysis… Sreedhar Kalavagunta.. and Ronagh. Plastic mechanisms database for thin-walled analysis. R. identify the effects of higher modes in a pushover 2010. and Dubina. Washington. D.