P. 1
a modal pushover analysis procedure - chopra - goel

a modal pushover analysis procedure - chopra - goel

|Views: 120|Likes:
Published by gogus79
a research report modal pushover analysis
a research report modal pushover analysis

More info:

Categories:Types, Research
Published by: gogus79 on Jan 04, 2013
Copyright:Attribution Non-commercial

Availability:

Read on Scribd mobile: iPhone, iPad and Android.
download as PDF, TXT or read online from Scribd
See more
See less

02/25/2015

pdf

text

original

Sections

  • 1 Introduction
  • 2 One-Story Systems
  • 2.1 EQUATION OF MOTION
  • 2.2 SYSTEM AND EXCITATION CONSIDERED
  • 2.3 RESPONSE HISTORY ANALYSIS
  • 2.4 PUSHOVER ANALYSIS
  • 3. Elastic Multistory Buildings
  • 3.1 MODAL RESPONSE HISTORY ANALYSIS
  • 3.2 MODAL RESPONSE SPECTRUM ANALYSIS
  • 3.3 MODAL PUSHOVER ANALYSIS
  • 3.4 COMPARATIVE EVALUATION OF ANALYSIS PROCEDURES
  • 3.4.1 System and Excitation Considered
  • 3.4.3 Modal Pushover Analysis
  • 4 Inelastic Multistory Buildings
  • 4.1 RESPONSE HISTORY ANALYSIS
  • 4.2 UNCOUPLED MODAL RESPONSE HISTORY ANALYSIS
  • 4.2.1 Underlying Assumptions and Accuracy
  • 4.2.2 Properties of the nth-“mode” Inelastic SDF System
  • 4.2.3 Summary
  • 4.3 MODAL PUSHOVER ANALYSIS
  • 4.3.1 Summary
  • 4.4 COMPARATIVE EVALUATION OF ANALYSIS PROCEDURES
  • 4.4.1 Uncoupled Modal Response History Analysis
  • 4.4.2 Modal Pushover Analysis
  • 5. Comparison of Modal and FEMA Pushover Analyses
  • 5.1 FEMA-273 PUSHOVER ANALYSIS
  • 5.2 COMPARATIVE EVALUATION
  • 6 Conclusions
  • 7 References
  • Appendix A Uncoupled Modal Response History Analysis

A MODAL PUSHOVER ANALYSIS PROCEDURE TO ESTIMATE SEISMIC DEMANDS FOR BUILDINGS: THEORY AND PRELIMINARY EVALUATION ANIL K.

CHOPRA • RAKESH K. GOEL PEER 2001/03 JAN. 2001

Pacific Earthquake Engineering Research Center

A Modal Pushover Analysis Procedure to Estimate Seismic Demands for Buildings:
Theory and Preliminary Evaluation

Anil K. Chopra University of California Berkeley Rakesh K. Goel California Polytechnic State University San Luis Obispo

A report on research conducted under grant no. CMS-9812531 from the National Science Foundation: U.S.-Japan Cooperative Research in Urban Earthquake Disaster Mitigation

PEER 2001/03

A Modal Pushover Analysis Procedure to Estimate Seismic Demands for Buildings: Theory and Preliminary Evaluation

Anil K. Chopra University of California Berkeley

Rakesh K. Goel California Polytechnic State University San Luis Obispo

A report on research conducted under Grant No. CMS-9812531 from the National Science Foundation: U.S. Japan Cooperative Research in Urban Earthquake Disaster Mitigation

PEER Report 2001/03 Pacific Earthquake Engineering Research Center College of Engineering University of California Berkeley January 2001 i

ii .

25 to 3. and identifies locations of most plastic hinges. Thus the present trend of comparing computed hinge plastic rotations against rotation limits established in FEMA-273 to judge structural performance does not seem prudent. Second. and nonlinear RHA. The base shear-roof displacement Vbn . ro . The MPA procedure is extended to estimate the seismic demands for inelastic systems: First. is determined by pushover analysis. iii . MPA. However. show that MPA estimates the response of buildings responding well into the inelastic range to a similar degree of accuracy when compared to standard RSA for estimating peak response of elastic systems. p eff t = . bg bg bg bg b g b g Comparing the peak inelastic response of a 9-story SAC building determined by the approximate MPA procedure with rigorous nonlinear response history analysis (RHA) demonstrates that the approximate procedure provides good estimates of floor displacements and story drifts. in the modal expansion of the effective earthquake forces. Combining these peak modal responses by modal combination rule leads to the MPA procedure. such as hinge plastic rotations.ABSTRACT The principal objective of this investigation is to develop a pushover analysis procedure based on structural dynamics theory.sn ug t .urn curve is developed from a pushover analysis for force distribution s* . at which the seismic response. it is demonstrated that the FEMA force distributions greatly underestimate the story drift demands. structural performance evaluation should be based on story drifts known to be closely related to damage and can be estimated to a higher degree of accuracy by pushover analyses. Thus the MPA procedure is accurate enough for practical application in building evaluation and design.0. Comparing the earthquake-induced demands for the selected 9-story building determined by pushover analysis using three force distributions in FEMA-273. plastic hinge rotations are less accurate. The results presented for El Centro ground motion scaled by factors varying from 0. and the MPA procedure is more accurate than all the FEMA force distributions methods in estimating seismic demands. all pushover analysis procedures considered do not seem to compute to acceptable accuracy local response quantities. The standard response spectrum analysis (RSA) for elastic buildings is reformulated as a Modal Pushover Analysis (MPA). The peak deformation of this SDF system is used to determine the roof displacement. This pushover curve is idealized as bilinear n and converted to the force-deformation relation for the nth-“mode” inelastic SDF system. the total demand. a pushover analysis determines the peak response rno of the inelastic MDF system to individual modal terms. and the structure is pushed to the roof displacement determined from the peak deformation Dn of the nth-mode elastic SDF system. which retains the conceptual simplicity and computational attractiveness of current procedures with invariant force distribution. The peak response of the elastic structure due to its nth vibration mode can be exactly determined by pushover analysis of the structure subjected to lateral forces distributed over the height of the building according to s* = mφ n . but provides superior accuracy in estimating seismic demands on buildings. 2. Instead.mι ug t . rno . p eff .n t = . is determined by combining the rno n = 1.… according to an appropriate modal combination rule. where m is the n mass matrix and φ n its nth-mode.

iv .

v . This financial support is gratefully acknowledged. Japan Cooperative Research in Urban Earthquake Disaster Mitigation.ACKNOWLEDGMENT This research investigation is funded by the National Science Foundation under Grant CMS9812531.S. a part of the U.

vi .

..............................................................38 4......................3 System and Excitation Considered .......2 3..12 Modal Pushover Analysis ..................................................3........................................................................22 4...................................................9 Modal Response Spectrum Analysis ....... vii 1.....................................................................................................27 Uncoupled Modal Response History Analysis ......3 4..........................................................3 2.......55 Comparative Evaluation..............17 3.........3 Modal Pushover Analysis with Gravity Loads .....55 vii .......................4 5.......................................................................... Equation of Motion ..............................9 3...........................1 Summary..........................................................................................................................................4.2 2..........................................2 Response History Analysis ......................55 5........................................................................................................................1 3.....................28 4......................v Table of Contents.........................................2............................ 2......................5 Response History Analysis .......................13 3.........................................................................41 4....................................................13 3........................................................................................................................1 4.....................................................4 3..............3 Modal Pushover Analysis ........................................................2 Properties of nth-“mode” Inelastic SDF System ..............1 2.............................4.......................................6 Elastic Multistory Buildings .................. iii Acknowledgment ......................3 2.......1 5..........................1 System and Excitation Considered ...............................................3 3...2 Modal Pushover Analysis .......................12 Comparative Evaluation of Analysis Procedures ......................................................6 Pushover Analysis ..................................................................................................4......2...............................1 Uncoupled Modal Response History Analysis .............................36 4....................................................................................................4...............33 4........................34 4................................... Introduction............................................CONTENTS Abstract ..................................................................................................1 Underlying Assumptions and Accuracy .........................................................47 4....................4......2 Response History Analysis .............................................................................2 FEMA-273 Pushover Analysis .................................................1 One-Story Systems ............................................................ Comparison of Modal and FEMA Pushover Analyses...........4 Modal Response History Analysis..........................36 Modal Pushover Analysis ....37 Comparative Evaluation of Analysis Procedures ...................................................4.....................3 Summary.........................2................38 4..................................................................................... Inelastic Multistory Buildings .......................................27 4...............

..................................................71 B: Modal Pushover Analysis .................................................................................................. Conclusions......... 7....................65 References.......................................................69 Appendices A: Uncoupled Modal Response History Analysis .........................................................................................................................................85 viii ..................................................................6............................................................................81 C: FEMA Force Distribution Calculation.....

1997. 1981. Gupta and Krawinkler. or for a redistribution of inertia forces because of structural yielding and the associated changes in the vibration properties of the structure. Obviously. such satisfactory predictions of seismic demands are mostly restricted to low. 1999. after the structure yields both assumptions are approximate. The seismic demands are computed by nonlinear static analysis of the structure subjected to monotonically increasing lateral forces with an invariant height-wise distribution until a predetermined target displacement is reached.. Both the force distribution and target displacement are based on the assumption that the response is controlled by the fundamental mode and that the mode shape remains unchanged after the structure yields. 1999. 1998. 1991. 1999]. 1994. While these adaptive force distributions may provide better estimates of seismic demands [Gupta and Kunnath. While nonlinear response history analysis (RHA) is the most rigorous procedure to compute seismic demands. several researchers have proposed adaptive force distributions that attempt to follow more closely the time-variant distributions of inertia forces [Fajfar and Fischinger.and medium-rise structures in which inelastic action is distributed throughout the height of the structure [Krawinkler and Seneviratna. Skokan and Hart. 1999. requires explicit consideration of inelastic behavior of the structure. Gupta and Krawinkler. they are conceptually complicated and computationally demanding for routine application in structural engineering practice. 1988. Maison and Bonowitz. Attempts 1 . However.1 Introduction Estimating seismic demands at low performance levels. Kim and D’Amore. None of the invariant force distributions can account for the contributions of higher modes to response. 1997]. Bracci et al. current structural engineering practice uses the nonlinear static procedure (NSP) or pushover analysis in FEMA-273 [Building Seismic Safety Council. but investigations [Saiidi and Sozen. such as life safety and collapse prevention. 1998. 2000.. Lawson et al. To overcome these limitations. Fajfar and Fischinger. Miranda. 2000]. 1988. Krawinkler and Seneviratna. 2000]. Gupta and Kunnath. 2000] have led to good estimates of seismic demands.

2000. The principal objective of this investigation is to develop an improved pushover analysis procedure based on structural dynamics theory that retains the conceptual simplicity and computational attractiveness of the procedure with invariant force distribution.. Matsumori et al. 1998. 2 . 2000. we show that pushover analysis of a one-story system predicts perfectly peak seismic demands. the seismic demands determined by pushover analysis using three force distributions in FEMA-273 are compared against the MPA and nonlinear RHA procedures. Gupta and Kunnath. The MPA procedure is then extended to inelastic buildings. 1996. the underlying assumptions and approximations are identified. Sasaki et al. 2000]... but provides superior accuracy in estimating seismic demands on buildings. Next we develop a modal pushover analysis (MPA) procedure for linearly elastic buildings and demonstrate that it is equivalent to the well-known response spectrum analysis (RSA) procedure. Finally. First. Kunnath and Gupta.have also been made to consider more than the fundamental vibration mode in pushover analysis [Paret et al. and the errors in the procedure relative to a rigorous nonlinear RHA are documented.

It consists of a mass m (or weight w) concentrated at the roof level.2) . During yielding the stiffness of the frame is a k . the yield deformation. This lateral-force displacement relation is idealized as shown in Fig. through the yield strength reduction factor. It is the familiar bilinear hysteretic relationship.1b.1) b g ug t is mu + cu + f s u.1 One-Story Systems EQUATION OF MOTION Consider the idealized one-story structure shown in Fig. reloading from a minimum deformation takes place along a path parallel to the initial elastic branch. where 0 < a << 1. The yield strength is related to f o . defined by f Ry = o fy (2. the strength required for the structure to remain elastic during the ground motion. the yield strength. The yield strength is the same in the two directions of deformation. sign u = . a massless frame that provides stiffness to the system.1a. The hysteretic relation between the lateral force f s and lateral displacement u of the mass relative to the base of the frame is denoted by f s u.mug t bg The governing equation for this inelastic system subjected to horizontal ground acceleration b g bg 3 (2. R y . Similarly. Yielding begins when the force reaches f y and the deformation reaches u y . sign u . this system is linearly elastic with stiffness k as long as the force does not exceed f y . and a linear viscous damper with damping coefficient c.2 2. 2. Unloading from a maximum deformation takes place along a path parallel to the initial elastic branch. On initial loading. 2.

and (b) bilinear hysteretic forcedeformation relation For a given excitation u g t .4) and w n is the natural vibration frequency. or absolute (without regard to algebraic sign) maximum.1.5) For a given u g t . and the ductility factor is u m= m uy (2. in addition to the form of the force-deformation relation. Section 7.. and u y . and z is the damping ratio of the system vibrating within its linear elastic range (i. deformation is denoted by um . 2001. z y . z . u £ u y ).ug t n where bg bg b g bg (2. 2. (2. This becomes evident if Eq.fS m fs c ut u fy 1 αk k 1 uy ug -fy k k um u 1 1 (a) (b) Fig. bg 4 . (a) Idealized one-story structure.1) is divided by m to obtain ~ u + 2zw n u + w 2 u y f s u.e. the deformation u t depends on three systems parameters: w n . m depends on three system parameters: w n . The peak. and R y (Chopra .3). Tn = 2p w n is the natural vibration period.3) ωn = k m ζ = c 2mω n f fs = s fy (2. sign u = .

f o w = 184 . Vby = 39. this system and excitation. For .2.2: the dimensions and flexural rigidity of the structural elements are noted. 2.26 kN. 2.2. 2001) and scaled up by a factor of 2. is f y / w = ( f o / w ) ÷ 8 = 0. based on Ry = 8 .1.2 SYSTEM AND EXCITATION CONSIDERED Consider the one-story system in Fig.32 m • Fig.4 in Chopra.04 40 Base Shear. 2. 2.26 kN (8. The yield strength of the inelastic system. One-story. one-bay frame 50 uy = 1. and f y = 39. u (cm) 8 Fig.376 cm.2 kips) . Pushover curve for structure shown in Fig. 6. m h = 3.2311 .2 5 .3. with Tn = 0.5 sec and z = 5% subjected to the north-south component of the El Centro (1940) ground motion (Fig. α = 0.66 m • EIb EIc EIc • • L = 7.9 kN (38. Vb (kN) 30 20 10 0 0 2 4 6 Roof Displacement.826 kips) for w = 169.

pushover analysis cannot provide any cumulative measure of response.1 kip-in. q m = 0.134 × 107 mm 4 (75.4 PUSHOVER ANALYSIS Static analysis of the one-story nonlinear system subjected to lateral force that increases in small increments is implemented until the lateral displacement reaches um = 7.) for the beam and columns. 2.077 × 107 mm 4 (146 in.4e.4 ). as apparent in Fig. the joint rotation and beam hinge rotation are identical to values q m and q pm determined from RHA. the ductility factor m = 5. respectively. 2.3 RESPONSE HISTORY ANALYSIS Figure 2. The resulting pushover curve is shown in Fig. 2.017 rad . This pushover curve turns out to be bilinear because the beam and the columns are designed to yield simultaneously when f s reaches f y .3 in. the peak value determined from RHA. 2.18 kN-m (444. The peak values of the various response quantities are as follows: um = 7.3. and E = 2 × 108 kPa (29×103 ksi) gives yield moments of 21. and q pm = 0. However.0217 rad . (c) shows the joint rotation q bt g .6 kip-in. Implementing this analysis with I c = 6. the energy dissipated in 6 ..65 kN-m (191.36 cm . (d) shows the rotation q p bt g of the plastic hinges at the bg bg beam ends. e. (b) shows the lateral force f s t or base shear Vb t normalized relative to the bg weight. 2.) and 50.g.4edetermined from RHAis superimposed.35. wherein the hysteretic force-deformation history of Fig. A static pushover analysis of this one-story system leads to the force-displacement relationship shown in Fig. The yielding stiffness of each structural element is defined as 3% of its initial stiffness. Observe that the pushover curve matches the initial loading path of the hysteretic system. The system is excited well beyond the yield deformation. and (e) shows the force-deformation relation.4 shows the earthquake response of the system described in the preceding section determined by response history analysis (RHA).4f.36 cm. It is organized in five parts: (a) shows the deformation u t . 2.4 ). I b = 3. Determined from pushover analysis at the exact peak deformation.The yield moments in the beam and columns are defined as the bending moments due to the lateral force f y .

04 0 5 10 15 Time (sec) 20 25 30 0.4 0. Response of one-story system to El Centro ground motion: (a) deformation.2 0 (e) 0 (f) −0.2311 (a) Vb / w 0 f / w = −0. and (f) pushover curve 7 . (b) base shear.2311 y (b) −0.2 −0. 15 u (cm) 0 • 7.04 0.yielding during the ground motion.4 0. This represents an inherent limitation of pushover analyses.5 fy / w = 0. (e) force-deformation relation.2 Vb / w 0. (c) joint rotation.4 −10 −5 0 u (cm) 5 10 Fig.2 −0.4.04 0.0217 • θ (rad) 0 (c) −0. 2.4 −10 −5 0 u (cm) 5 10 −0.04 0. (d) plastic hinge rotation.36 −15 0.017 • 0 (d) θp (rad) −0.5 0. or the cumulative rotation at a plastic hinge.

8 .

snug t n= n =1 The effective earthquake forces can then be expressed as af N af N af (3.m ι u g t bg bg (3. and p eff t = Â peff. (3. classical damping.3.1 MODAL RESPONSE HISTORY ANALYSIS The differential equations governing the response of a multistory building to horizontal earthquake ground motion u g t are as follows: mu + cu + ku = .1) where u is the vector of N lateral floor displacements relative to the ground. each element of the influence vector ι is equal to unity. and lateral stiffness matrices of the system. and k are the mass. c. The right side of Eq.12): mι = n =1 Â sn = N n =1 Â Γ nmφ n N (3. This force distribution can be expanded as a summation of modal inertia force distributions s n (Chopra.3) where φ n is the nth natural vibration mode of the structure. 2001: Section 13. m.4) 9 . Elastic Multistory Buildings 3.n t = Â .2) The spatial distribution of these “forces” over the height of the building is defined by the vector s = m ι and their time variation by u g t .1) can be interpreted as effective earthquake forces: p eff t = .mι ug t bg bg bg (3.

9) to the equation of motion for the nth-mode elastic SDF system. with no contribution from other modes.10) gives qn t = Gn Dn t bg bg bg (3.8) where the modal coordinate qn t is governed by 2 qn + 2z nw n qn + w n qn = .9) in which w n is the natural vibration frequency and z n is the damping ratio for the nth mode. (3.11) and substituting in Eq.sn ug t bg (3.Gn ug t (3. p eff .9) and (3.n t = .n t is entirely in the nth-mode.5) The contribution of the nth mode to s and to p eff t are: sn = Gn mφ n respectively.u g t n Comparing Eqs. we will outline that the response of the MDF system to p eff .6) Next.sn ug t bg bg bg bg (3. it can be demonstrated that none of the modes other than the nth mode contribute to the response.7) By utilizing the orthogonality property of modes.8) gives the floor displacements u n t = Gnφ n Dn t bg (3.peff (t ) = n =1 ∑ peff .10) bg bg bg 10 (3. subjected to u g t : Dn + 2z nw n Dn + w 2 Dn = . (3.12) . The solution qn t can readily be obtained by comparing Eq. (3. The equations governing the response of the system are mu + cu + ku = .n (t ) = N n =1 ∑ −sn u g (t ) N (3. Then the floor displacements are u n t = φ n qn t bg bg bg bg (3. an SDF system with vibration propertiesnatural frequency w n and damping ratio z n of the nth-mode of the MDF system.

the static value of r due to external forces sn . However.1.15) and (3. Equations (3.14) is the pseudo-acceleration response of the nth-mode SDF system (Chopra. In contrast to the classical 11 bg .1.13) define the contribution of the nth-mode to the response.9) is the standard modal equation governing qn t . Eqs. and An t = w 2 Dn t n bg (3. Therefore.13) st where rn denotes the modal static response. 3.1). Conceptual explanation of modal response history analysis of elastic MDF systems Any response quantity r t story drifts. ζn ¨ g(t ) u (a) Static Analysis of Structure (b) Dynamic Analysis of SDF System Fig.n t . (3.16) reflect combining the response contributions of all modes. (3. the response of the system to the total excitation p eff t is ut = bg bg bg n =1 N Â un bt g = Â Gnf n Dn bt g n =1 N N (3.12) and (3.12) and (3. these standard equations have been derived in an unconventional way.can be expressed by st rn t = rn An t bg bg bg bg bg (3. 3.Forces sn An(t ) st rn ωn. (3.16) This is the classical modal RHA procedure wherein Eq.13) represent the response of the MDF system to p eff. 2001. st The two analyses that lead to rn and An t are shown schematically in Fig. internal element forces. Section 13.15) rt = bg n =1 Â rn t = bg n =1 Â rnst An bt g N (3. and Eqs. etc.

(3. this response value can be obtained by static analysis of the structure subjected to lateral forces distributed over the building height according to 12 . 2001. (3. (3. z n of the pseudo-acceleration response (or design) spectrum for the nth-mode SDF system. This concept provides a rational basis for the modal pushover analysis procedure developed later. we used the modal expansion of the spatial distribution of the effective earthquake forces. the peak value rno of the nth-mode contribution rn t to response r t is determined from st rno = rn An bg bg bg (3.2 MODAL RESPONSE SPECTRUM ANALYSIS The peak value ro of the total response r t can be estimated directly from the response spectrum for the ground motion without carrying out the response history analysis (RHA) implied in Eqs. Sections 12.19) will provide the same value of rno . Section 13.3 MODAL PUSHOVER ANALYSIS (3. the peak nth-mode response as in Eq. 3.1). In such a response spectrum analysis (RSA). which is valid for structures with well-separated natural frequencies such as multistory buildings with symmetric plan.16). Chopra. provides an estimate of the peak value of the total response: b g F N r 2 I 1/ 2 ro ª G Â no J H n =1 K 3.18) To develop a pushover analysis procedure consistent with RSA.1.3). The SRSS rule.derivation found in textbooks (e. 2001. Alternatively.4 and 13. The peak modal responses are combined according to the Square-Root-of-Sum-ofSquares (SRSS) or the Complete Quadratic Combination (CQC) rules. we note that static analysis of the structure subjected to lateral forces f no = Γ n mφ n An (3.9) . and Tn = 2p w n is the natural vibration period of the nth-mode of the MDF system.17) (Chopra..g.8.17) where An is the ordinate A Tn .

3. The building’s lateral load-resisting system is composed of steel perimeter moment-resisting frames (MRFS) with simple framing on the farthest south E-W frame. SAC Steel Project Technical Office. The interior bays of the structure contain simple framing with composite floors. California. denoted B1. Third St.2. CA 90017.s* = mφ n n (3. the peak value of the roof displacement due to the nth-mode.12) is urno = Gnf rn Dn (3. The peak modal responses. Richmond. The building has a basement level.18) to obtain an estimate of the peak value ro of the total response. shown in Fig.20) and the structure is pushed to the roof displacement. A benchmark structure for the SAC project. this structure meets seismic code and represents typical medium-rise buildings designed for the Los Angeles.19 m (122 ft) in elevation.. Obviously Dn and An are available from the response (or design) spectrum.73 m (150 ft) in plan. 3. 13 . which from Eq.1 System and Excitation Considered The 9-story structure.2). in both directions. the Applied Technology Council (ATC). each determined by one pushover analysis.4 COMPARATIVE EVALUATION OF ANALYSIS PROCEDURES 3. Los Angeles. this building is 45. rno . urno . 1660 W. region. was designed by Brandow & Johnston Associates1 for the SAC2 Phase II Steel Project.73 m (150 ft) by 45. 1301 S. 46th Street. can be combined according to Eq. CA 94804-4698. Typical floor-to-floor heights (for analysis purposes measured from center-of-beam to center1 2 Brandow & Johnston Associates. (3. The columns are 345 MPa (50 ksi) steel wideflange sections. This modal pushover analysis (MPA) for linearly elastic systems is equivalent to the well-known RSA procedure (Section 3. with the ninth level being the roof.21) 2 where Dn = An ω n . The bays are 9. and California Universities for Research in Earthquake Engineering (CUREE). with five bays each in the north-south (N-S) and east-west (E-W) directions. The levels of the 9-story building are numbered with respect to the ground level (see Fig. Although not actually constructed.15 m (30 ft) on center. Consulting Structural Engineers. (3. SAC is a joint venture of three non-profit organizations: The Structural Engineers Association of California (SEAOC). 3.2). and 37.4.

The 9-story N-S MRF is depicted in Fig. fifth. The floor system is composed of 248 MPa (36 ksi) steel wide-flange beams in acting composite action with the floor slab.sec2/ft).07×106 kg (73. for the first three modes are shown in Fig. monolithic column pieces are connected every two levels beginning with the first level. 1988) using the M1 model developed by Krawinkler and Gupta (1998). respectively.01×106 kg (69. i. The column lines employ two-tier construction. The seismic mass of the structure is due to various components of the structure. 3.0 kips-sec2/ft). The strength. The building is modeled in DRAIN-2DX (Allahabadi and Powell. roofing and a penthouse located on the roof.49 m (18 ft). The simple model adopted here is sufficient for the objectives of this study. for the first level is 1. This model is based on centerline dimensions of the bare frame in which beams and columns extend from centerline to centerline.4.89×105 kg (67.00×106 kg (616 kips. and for the ninth level is 1. Column splices.20)..83 m (6 ft) above the center-line of the beam to column joint. floor slabs.65×105 kg (66. such as those described in Gupta and Krawinkler (1999) can be used. dimension. which are seismic (tension) splices to carry bending and uplift forces. and 0. 0. These force n distributions will be used in the pushover analysis to be presented later. To ensure that this structure remains elastic. if desired more complex models. partitions.85.0 kips-sec2/ft). 3. 14 . mechanical/electrical. The column bases are modeled as pinned and secured to the ground (at the B-1 level). ceiling/flooring.27. third. 3. Each frame resists one half of the seismic mass associated with the entire structure. The seismic mass of the above ground levels of the entire structure is 9.3. The seismic mass of the ground level is 9. Concrete foundation walls and surrounding soil are assumed to restrain the structure at the ground level from horizontal displacement. including the steel framing. for the second through eighth levels is 9.e. The force distributions. The first three vibration modes and periods of the building for linearly elastic vibration are shown in Fig. and seventh levels at 1.2. 3.2 kips-sec2/ft). and shear distortion of panel zones are neglected but large deformation (P∆) effects are included.7 kips-sec2/ft). s* (Eq. the vibration periods are 2.96 m (13 ft). are located on the first.65 m (12 ft) and for the first floor is 5. The floor-to-floor height of the basement level is 3. we select a weak ground motion: the northsouth component of the El Centro (1940) ground motion scaled down by a factor of 4.49 sec.of-beam) are 3.

Fig. 2000] 15 .2.. 3. Nine-story building [adapted from Ohtori et al.

n = 1. (b) shows the base shear Vbn t normalized relative to the weight W of the building. n = 1.7.49 sec 3 T = 0.75 1.12) and (3.94 2.3.93 −1.5 Mode Shape Component 1 1.05 1.4.6.0272 −2. Each figure is organized in four parts: (a) shows the roof displacement urn t . 2.5 Fig.38 0.05 s * 1 s * 2 s * 3 Fig.67 −1.51 0.5 −1 −0.03 −1.05 2.12 0.13)].13 −1.33 2.85 sec 2 Floor 5th 4th 3rd 2nd 1st T1 = 2.487 −1. and 3 n 3. 3. 3. Force distributions s* = mφn .1 −2. 2. 3. determined by RHA [Eqs.04 1.8 −2.37 2. respectively.2 Response History Analysis The structural response due to individual vibration modes. First three natural-vibration periods and modes of the 9story building 3.9th 8th 7th 6th T = 0.4.27 sec Ground −1. and 3 .796 0.5.61 2. is shown in Figs.72 −2.39 3.5 0 0. (3.1 3.44 1.728 2. 3.31 −0. and 3. (c) shows the joint rotation q n t of an external joint at 16 bg bg bg .

3.23 cm. The peak values of the various response quantities are noted in these figures.422 cm.the roof level.16)]. This is illustrated in Fig. are presented in Fig. the peak roof displacement due to each of three modes is ur1o = 9. two. The peak values of displacements of all floors. which are listed in Tables 3. ur 2o = 2. and 3. 3. larger in story drifts. Combining the modal response histories for all modes gives the total response [Eqs.8. and ur 3o = 0. 3. For a fixed number of modes included. or all modes. but three modes—perhaps even two modes—are sufficient. and 3.2. It is apparent that the first mode alone is inadequate. consistent with the increasing significance of the higher mode response among these three sets of response quantities.12 cm. 3. including one. respectively. three. The linear relationship between the base shear and roof displacement for each mode implies that the structure did not yield. two. The errors due to truncating the contributions of vibration modes beyond the third mode are negligible. with the first three modes provide essentially the exact response.urn relation.8.3. 3.1. especially in estimating the story drifts.9. 17 . and rotations of external joints with moment connections are presented in Tables 3. As expected.2.15) and (3. drifts in all stories. respectively. and the percentage errors due to truncation of higher modal contributions. (3.3. errors generally decrease as response contributions of more modes are included. and even larger in joint rotations. and (d) shows the Vbn . and three vibration modes. errors are smallest in floor displacements. The same method was used to determine the peak values of many response quantities. in particular. the exact response considering all modes. The peak values of floor displacements and story drifts determined by RHA. the results for the roof displacement and top-story drift are shown in Fig.1. Also included are the combined response due to one. where the second and third modal responses are a larger percentage of the top story drift compared to roof displacement.

245 0.3 19.91E-04 1.31E-03 Mode 2 -1.026 0.6 0.8 -56.50E-03 2.1 Table 3.152 0.225 0.2 0.3 -0.7 4.4 -22.8 -15.245 0.38E-04 2 Modes 2.325 0.6 -0.31E-03 Joint Rotation (rad) Combined (RHA) 1 Mode 2.466 0.63E-03 2.65E-03 2.235 0.0 -2.0 7.260 0.009 -0.14 m3) from RHA for 0.1 -2.00E-03 2.9 -22.307 0.76E-03 1.09E-03 1.5 0.015 -0.2 Peak values of story drift ratios (as % of story height) from RHA for 0.9 8.227 0.4 0.192 0.22E-03 2.235 0.177 0.229 0.253 0.9 9.378 0.011 0.177 0.090 0.258 0.25 × El Centro ground motion Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Modal Response Mode 1 Mode 2 Drift Ratio (%) Combined (RHA) 1 Mode 2 Modes 3 Modes Mode 3 0.054 0.259 0.15E-03 4.407 0.156 0.023 0.44E-03 3.253 0.226 0.5 -1.9 2.003 0.159 0.173 0.475 0.9 -23.0 -2.1 Peak values of floor displacements (as % of building height = 37.282 0.74E-03 1.321 0.2 -20.7 3.9 -15.042 0.8 1.1 4.152 0.3 -33.202 0.8 -5.9 -24.29E-03 2.4 -53.2 -2.197 0.8 -10.4 -41.74E-04 9.8 -1.259 0.181 0.56E-03 2.310 0.44E-03 1.99E-03 2.5 18.097 0.76E-03 1.63E-03 2.4 -3.13E-03 2.38E-03 2.157 0.01E-04 -2.6 -1.399 0.124 0.14E-03 2.0 -10.64E-03 3.89E-03 1.Table 3.03E-03 1.006 0.88E-03 2.00E-03 1.24E-03 2.74E-04 6.4 -10.03E-03 1.7 -19.038 0.035 0.1 -14.2 -57.024 -0.0 -0.1 -2.09E-03 2.124 0.25 × El Centro ground motion Displacement /Height (%) Modal Response Combined (RHA) Mode 1 Mode 2 Mode 3 1 Mode 2 Modes 3 Modes Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 0. 18 .4 -1.265 0.300 0.89E-03 1.043 0.008 -0.264 Error (%) 1 Mode 2 Modes 3 Modes -23.002 -0.74E-03 1.09E-03 1.3 -8.6 0.011 0.09E-03 Error (%) 1 Mode -23.0 -46.26E-04 -5.173 0.6 4.03E-03 3 Modes 2.226 0.156 0.227 0.2 1.295 -0.060 0.25 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 3 Modal Response Mode 1 2.72E-03 3.94E-03 2.069 0.266 0.130 0.03E-03 -6.8 1.001 -0.401 Error (%) 1 Mode 2 Modes 3 Modes -23.4 -7.350 0.7 -50.055 0.3 -3.2 -4.229 0.1 -0.060 -0.266 0.303 0.263 0.7 2.6 9.069 0.097 0.042 0.117 0.133 -0.90E-03 3.282 0.2 9.0 3 Modes -5.4 -0.9 1.13E-04 9.9 -16.2 -1.11E-04 -5.3 Table 3.311 0.370 0.125 0.4 -1.406 0.99E-03 Mode 3 3.062 -0.65E-03 RHA (all modes) 2.089 0.9 3.08E-03 2.3 -0.265 RHA (all modes) 0.205 0.1 -11.231 -0.453 0.1 -2.012 0.364 0.47E-03 1.7 7.7 2 Modes -3.1 -19.032 -0.5 -2.058 -0.088 -0.202 0.203 0.202 0.125 0.199 0.33E-03 2.071 0.4 -6.42E-04 1.260 0.1 Building height is measured from the ground floor to the 9th floor.336 0.183 0.78E-04 -3.3 Peak values of joint rotations (radians) from RHA for 0.121 0.045 0.1 3.413 RHA (all modes) 0.00E-03 1.6 -1.400 0.012 -0.73E-03 3.179 0.062 0.6 1.237 0.11E-03 1.45E-03 3.237 0.060 -0.266 0.010 -0.311 0.333 0.022 0.66E-05 -3.28E-04 1.6 11.261 -0.85E-03 3.01E-04 3.88E-03 2.275 0.097 0.50E-03 4.080 0.

5 15 −0.12 −0. 3. (d) force-deformation history.1 5 x 10 −3 θ (rad) 0. and (e) pushover curve.5 0 u (cm) r1 7. Excitation is 0.5 9.1 −15 −7.1 0.1 /W 0.12 • 0 (a) r1 −15 0.15 u (cm) 9.0435 Vb1 / W 0.00174 • 0 (c) r1 −5 0 5 10 15 Time (sec) 20 25 30 0.5 0 u (cm) r1 7.1 0.0435 9.1 −15 −7.25 × El Centro ground motion 19 .0435 • 0 (b) V b1 −0. (b) base shear. Response due to first mode: (a) roof displacement.5.12 0 (d) 0 (e) 15 Fig. (c) joint rotation.

1 Vb2 / W 0.5 0 u (cm) r2 2.0252 (d) 0 0 (e) −0.5 5 Fig.0252 • 0 (b) V −0.1 0. and (e) pushover curve. Excitation is 0. 3.1 −5 −2. (b) base shear. Response due to second mode: (a) roof displacement.1 −5 −2.25 × El Centro ground motion 20 .5 u (cm) r2 0 • 2. (d) force-deformation history. (c) joint rotation.05 5 x 10 −3 b2 θ (rad) 0.5 0 u (cm) r2 2.05 /W 0.6.23 −2.5 5 −0.23 −5 (a) 0.0252 −2.23 0.00244 • 0 (c) r2 −5 0 5 10 15 Time (sec) 20 25 30 0.

7.422 /W V b3 0 (d) 0 −0. Excitation is 0.422 (e) −0.02 −0.000913 • 0 (c) r3 −2 0 5 10 15 Time (sec) 20 25 30 0.02 0. Response due to third mode: (a) roof displacement.25 × El Centro ground motion 21 .0108 x 10 −3 (b) −0.422 −1 (a) 0. 3.5 1 −0.02 2 θ (rad) 0.0108 −0.1 u (cm) r3 0 • 0.0108 −0.5 0 u (cm) r3 0.5 1 Fig. (b) base shear. and (e) pushover curve.02 /W V b3 0 • 0.02 −1 −0.5 0 u (cm) r2 0.02 −1 −0. (d) force-deformation history. (c) joint rotation.

5. i. and 3. These values of the response quantities are identical to the peak response values determined from RHA (Tables 3. 3. In practical application. 3. 3. respectively.20) with n = 2 and 3 up to roof displacements ur 2 o = 2. two.e. i.1.5d). and it is apparent in the example considered that three modes provide most of the response (Fig. implying that the modal truncation errors are small if at 22 .12 cm. 3.4) to roof displacement ur1o = 9. considering one. As suggested by Eq.7d). using the force distribution of Eq. Implementing pushover analysis for the second and third modes.3 Modal Pushover Analysis Implementing MPA for the fundamental vibration mode. and 3.3). the errors in the MPA results are generally larger than in RHA (Fig. In RHA the errors arise only from truncating the responses due to higher modes. 3. pushing the structure using the force distribution of Eq. or three vibration modes.7e and to the floor displacements. except for the algebraic sign associated with Γ1 and minor round-off errors.3).12).6e and 3. this value would be determined directly from the response (or design) spectrum. Observe that the target roof displacement in each pushover analysis is identical to its exact value determined by RHA.6 present estimates of the combined response according to Eq. 3. story drifts. story drifts. these pushover curves are consistent with the Vb . 3. which would provide the Dn value to be substituted in Eq.23 cm . and external joint rotations in Tables 3. and 3. and the computed response values are identical to the peak response values determined from RHA (Tables 3..4. 3. confirming that MPA gives the exact values of the individual modal responses. (3.5. (3. the value determined by RHA (Fig.20) with n = 1 (Fig.3). Figure 3. pushing the structure. 3.9 and Tables 3.422 cm .18).4. and 3. This pushover curve is consistent with the relationship between the base shear and roof displacement determined by RHA (Fig. and the errors in these estimates relative to the exact response from RHA considering all modes.1.9). 3.6. (3.ur relations determined by RHA (Figs. and external joint rotations computed by pushover analysis are presented in Tables 3.2.21). respectively. leads to the pushover curves shown in Figs.6.1 through 3. 3.2..5. (3. 3. 3. The floor displacements.8) leads to the pushover curve shown in Fig.6d and 3. the floor displacements are proportional to the mode shape φ 1 because the structure remains elastic. (3.4. and ur 3o = 0. respectively.e.10 and Tables 3. and 3.4. For a fixed number of modes included considered. although both analyses led to identical peak values of the individual modal responses.3. As for the first mode.5e.

least three modes are included. (a) Roof Displacement 15 (b) Top Story Drift 3 Mode 1 ∆ (cm) u (cm) 9.5 Ground 0 0. Additional errors are introduced in pushover analysis due to the approximation inherent in modal combination rules. 3.2 0. 3.3 0.4 Story Drift Ratio (%) 0.eps fig3_9b.3 0. sec 25 30 r 5 10 15 20 Time.25 × El Centro ground motion: first three modal responses and total (all modes) response (a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th (b) Story Drift Ratios Floor 4th 3rd 2nd 1st Floor 5th RHA All Modes 1 Mode 2 Modes 3 Modes 5th 4th 3rd 2nd 1st RHA All Modes 1 Mode 2 Modes 3 Modes fig3_9a.23 r2 −15 15 0 • 0.9.1 0. Response histories of roof displacement and top-story drift from RHA for 0.685 • −3 3 Mode 2 ∆ (cm) u (cm) • 2.48 −3 3 1.2 0.422 r2 0 • 1.1 0.25 × El Centro ground motion 23 . sec 25 30 Fig.12 • r1 −15 15 0 r1 0 0 0.4 Displacement/Height (%) 0.8.6 Fig.eps Ground 0 0.83 • r3 0 • 0.59 • Total (All Modes) 0 −15 0 ∆ (cm) u (cm) 0 −3 0 fig3_8.eps r 5 10 15 20 Time.03 −3 3 Mode 3 ∆ (cm) u (cm) r3 −15 15 9.5 0. Heightwise variation of floor displacements and story drifts from RHA for 0.

229 0.267 0.44E-03 -1.9 -8.276 0.73E-03 3.9 -13.22E-03 -2.43E-04 -1.310 0.4 Peak values of floor displacements (as % of building height = 37.181 0.253 0.097 0.9 -14.63E-03 -2.6 -17.04E-03 3.25 × El Centro ground motion Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Modal Response Mode 1 Mode 2 Drift Ratio (%) Combined (MPA) 1 Mode 2 Modes 3 Modes Mode 3 -0.336 RHA (all modes) 0.76E-03 -1.152 0.40E-04 5.1 -19.012 0.069 0.31E-03 2.048 0.8 -56.38E-03 2.4 -22.071 0.203 0.370 0.090 0.42E-04 -1.7 -50.03E-03 1.2 -0.3 -19.328 0.3 -33.270 0.3 -14.124 0.2 Table 3.125 0.97E-03 1.31E-03 Error (%) RHA (all modes) 2.038 -0.89E-03 2.9 -14.231 -0.125 0.272 0.015 -0.8 -22.74E-04 -6.72E-03 3.9 -24.2 1.9 -12.4 -14.133 0.374 0.0 -16.74E-03 -1.3 -9.282 -0.2 -12.300 0.229 0.4 0.1 -11.048 0.5 -16.7 2.53E-04 -9.060 -0.385 0.00E-03 2.203 0.157 0.253 -0.088 -0.157 0.24E-03 -2.036 -0.3 -2.6 Peak values of joint rotation (radians) from MPA for 0.4 -9.15E-03 1.177 0.4 -11.401 Error (%) 1 Mode 2 Modes 3 Modes -23.042 0.079 0.0 -10.Table 3.0 -0.00E-03 1.89E-03 -1.25 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Modal Response Mode 1 -2.203 0.133 0.245 -0.285 0.96E-03 2.9 -15.227 0.322 0.25 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Displacement /Height (%) Modal Response Combined (MPA) Mode 1 Mode 2 Mode 3 1 Mode 2 Modes 3 Modes 0.270 0.00E-03 -1.069 0.9 -13.230 -0.09E-03 Mode 2 1.156 0.4 -9.00E-03 2.31E-03 Joint Rotation (rad) Combined (MPA) 1 Mode 2.282 0.024 0.3 -41.90E-04 -9.9 -18.260 0.4 -4.92E-04 -1.99E-03 Mode 3 -3.260 0.1 -18.33E-04 5.261 -0.2 -20.6 -15.177 0.179 0.023 -0.78E-04 2 Modes 2.8 -15.6 -16.274 0.173 0.4 -19.94E-03 2.00E-03 3.7 -21.173 0.237 -0.008 -0.8 -14.12E-03 1.156 0.9 -13.57E-03 1 Mode -23.264 Error (%) Modal Response 1 Mode 2 Modes 3 Modes -23.003 -0.011 0.28E-03 2.4 -7.9 -16.0 -2.466 0.097 0.235 -0.006 -0.321 0.058 -0.2 -16.8 -15.226 0.63E-03 2.010 0.89E-03 1.03E-03 6.7 -15.032 0.009 0.0 -46.296 -0.09E-03 1.203 0.05E-03 3.3 Table 3.0 -18.9 3 Modes -12.062 0.7 24 .9 -15.89E-03 2.9 -15.73E-05 3.260 -0.026 0.4 -5.235 0.44E-03 3.060 0.4 -22.152 0.253 RHA (All Modes) 0.253 0.2 -57.38E-03 3 Modes 2.080 0.08E-03 2.1 -0.2 -4.9 -11.012 -0.106 0.203 0.042 0.5 Peak values of story drift ratios (as % of story height) from MPA for 0.332 0.03E-03 -1.179 0.055 0.230 0.001 0.14 m) from MPA for 0.197 0.4 -53.8 -23.267 -0.00E-03 2.6 -19.7 2 Modes -13.47E-03 1.106 0.3 1.1 -11.117 0.286 0.266 0.4 -4.2 -11.270 0.74E-03 1.227 0.121 0.3 -12.336 0.331 0.237 0.15E-03 2.03E-03 3.245 0.078 0.267 0.4 -2.4 1.9 -13.022 0.9 -15.09E-04 -3.259 -0.76E-03 1.062 -0.09E-03 -1.90E-03 1.89E-03 -2.043 -0.22E-04 2.045 -0.9 -14.313 0.259 0.3 -13.011 -0.80E-04 3.407 0.65E-03 2.7 -19.002 0.133 -0.

5 Ground 0 0.3 0.10.2 0.eps Ground 0 0.2 0.1 0.eps fig3_10b.5 0.(a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th (b) Story Drift Ratios Floor 4th 3rd 2nd 1st Floor 5th RHA All Modes MPA 1 Mode 2 Modes 3 Modes 5th 4th 3rd 2nd 1st RHA All Modes MPA 1 Mode 2 Modes 3 Modes fig3_10a.1 0.4 Displacement/Height (%) 0.3 0. 3.4 Story Drift Ratio (%) 0.6 Fig. shading indicates modal combination error 25 .25 × El Centro ground motion. Heightwise variation of floor displacements and story drifts from MPA for 0.

26 .

1) With this generalization for inelastic systems. and the unloading and reloading curves differ from the initial loading branch.1 Inelastic Multistory Buildings RESPONSE HISTORY ANALYSIS For each structural element of a building.1) becomes mu + cu + f s u. the natural vibration periods and modes of the corresponding linear system are the same as the vibration properties of the inelastic system undergoing small oscillations (within the linear range).4 4. sign u = . the relations between lateral forces f s at the N floor levels and the lateral displacements u are not single valued.2) to the modal coordinates of the corresponding linear system.mι ug t b (4. sign u b g g bg (4. Both systems have the same mass and damping. leading to the “exact” nonlinear response history analysis (RHA). Therefore. it is useful for later reference to transform Eq. Expanding the displacements of the inelastic system in terms of the natural vibration modes of the corresponding linear system we get ut = bg n =1 Â φ nqn bt g 27 N (4. Each structural element of this elastic system is defined to have the same stiffness as the initial stiffness of the structural element of the inelastic system.3) . Thus.2) The standard approach is to solve directly these coupled equations. the initial loading curve is idealized as bilinear. Although classical modal analysis (Section 3.1) is not valid for inelastic systems. (4. (3. but depend on the history of the displacements: f s = f s u. Eq.

Equation (4. sign u n b g b g bg (4.4)] leads to the uncoupled modal response history analysis (UMRHA) procedure. (3. (4. (4. (3.6b) are mu + cu + f s u. give the same results for u t as obtained directly from Eq.n t given by Eq.Gn ug t Mn bg n = 1.and classical n damping-orthogonality property of modes gives F qn + 2z nw n qn + sn = . qr t = 0 for all 28 bg . where φ n are now the modes of the corresponding linear system. premultiplying by φ T .sn ug t bg b g bg bg (4.2 UNCOUPLED MODAL RESPONSE HISTORY ANALYSIS bg Neglecting the coupling of the N equations in modal coordinates [Eq. The equations governing the response of the inelastic system to p eff.” implying that other “modes” will also contribute to the solution.… N (4. (4. (4. (3. these equations are coupled for inelastic systems. (3. 2. For linear systems. implying coupling of modal coordinates because of yielding of the structure. and using the mass. The spatial distribution s of the effective earthquake forces is expanded into the modal contributions sn according to Eq.5) This resisting force depends on all modal coordinates qn t .3) will.9) for linearly elastic systems. (4.2). (4. (3.6) The solution of Eq. However.Substituting Eq. (4.4) represents N equations in the modal coordinates qn .2).9) involves T Fsn = Fsn qn . (4. sign u = . 4. Simultaneously solving these coupled equations and using Eq. Unlike Eq. sign qn = φ n f s u n .4) is rarely solved because it offers no particular advantage over Eq.8) because qr t will generally be nonzero for “modes” other than the nth “mode. however.3) in Eq. in principle.2).6) for inelastic systems will no longer be described by Eq.4) where the only term that differs from Eq.3). Eq. This approximate RHA procedure is the preliminary step in developing a modal pushover analysis procedure for inelastic systems.

(a) p 15

eff,1

= −s × 0.25 × El Centro
1

(b) p 5

eff,2

= −s × 0.25 × El Centro
2

ur1 (cm)

0

ur1 (cm)

9.112 •

Mode 1

Mode 1

0

−15 15 Mode 2

−5 5 Mode 2

ur2 (cm)

0

ur2 (cm)

0 • 2.226 5 10 15 20 25 30

−15 0 15

5

10

15

20

25

30

Mode 3

−5 0 5

Mode 3

ur3 (cm)

0

ur3 (cm)
5 10 15 20 Time (sec) 25 30

0

−15 0

−5 0

5

10 15 20 Time (sec)

25

30

Fig. 4.1. Modal decomposition of the roof displacement due to (a) p eff,1 t = - s1 ¥ 0.25

¥ El Centro ground motion; and (b) p eff,2 t = - s 2 ¥ 0.25 ¥ El Centro ground
motion

af

af

modes other than the nth-mode; therefore, it is reasonable to expect that the nth “mode” should be dominant even for inelastic systems. These assertions are illustrated numerically in Figs. 4.1 and 4.2 for the selected 9-story building. Equation (4.6) was solved by nonlinear RHA, and the resulting roof displacement history was decomposed into its “modal” components. The modal decomposition of the roof displacement for the first three modes due to 0.25 ¥ El Centro ground motion demonstrates that because the building does not yield during this weak ground motion, the response to excitation p eff,n t is all in the nth-mode (Fig. 4.1). The structure yields when subjected to the strong excitation of 1.5 ¥ El Centro ground motion, and the modes other than the nth-mode contribute to the response. The second and third modes start responding to excitation p eff,1 t at about 29

bg

bg

(a) p 80

eff,1

= −s × 1.5 × El Centro
1

(b) p 20

eff,2

= −s × 1.5 × El Centro
2

Mode 1

Mode 1 • 1.437

u (cm)

r1

0 • 48.24

ur1 (cm)
Mode 2

0

−80 80

−20 20 Mode 2

u (cm)

r2

0

3.37 •

ur2 (cm)

0 • 11.62 5 10 15 20 25 30

−80 0 80

5

10

15

20

25

30

Mode 3 0.4931 •

−20 0 20

Mode 3 • 0.7783

u (cm)

r3

0

ur3 (cm)
25 30

0

−80 0

5

10 15 20 Time (sec)

−20 0

5

10 15 20 Time (sec)

25

30

Fig. 4.2. Modal decomposition of the roof displacement due to: (a) p eff,1 t = - s1 ¥ 15 ¥ . El Centro ground motion; and (b) p eff,2 t = - s2 ¥ 15 ¥ El Centro ground . motion

af

af

5.2 sec, the instant the structure first yields; however, their contributions to the roof displacement are only 7% and 1%, respectively, of the first mode response (Fig. 4.2a). The first and third modes start responding to excitation p eff,2 t at about 4.2 sec, the instant the structure first yields; however, their contributions to the roof displacement are 12% and 7%, respectively, of the second mode response (Fig. 4.2b). Approximating the response of the structure to excitation p eff,n t

bg

bg

by Eq. (3.8),

substituting Eq. (3.8) in Eq. (4.6) and premultiplying by φ T gives Eq. (4.4), except for the n important approximation that Fsn now depends only on one modal coordinate, qn :
T Fsn = Fsn qn , sign qn = φ n f s qn , sign qn

b

g

b

g

(4.7)

30

With this approximation, solution of Eq. (4.4) can be expressed by Eq. (3.11) where Dn t is governed by F Dn + 2z nw n Dn + sn = - ug t Ln and
T Fsn = Fsn Dn , sign Dn = φ n f s Dn , sign Dn

bg

bg c h

(4.8)

c

h

(4.9)

is related to Fsn qn , sign qn because of Eq. (3.11). Equation (4.8) may be interpreted as the governing equation for the nth-“mode” inelastic SDF system, an SDF system with (1) small amplitude vibration properties—natural frequency

b

g

w n and damping ratio z n —of the nth-mode of the corresponding linear MDF system; (2) unit
mass; and (3) Fsn Ln - Dn relation between resisting force Fsn Ln and modal coordinate Dn defined by Eq. (4.9). Although Eq. (4.4) can be solved in its original form, Eq. (4.8) can be solved conveniently by standard software because it is of the same form as the SDF system [Eq. (2.3)], and the peak value of Dn t can be estimated from the inelastic response (or design) spectrum (Chopra, 2001; Sections 7.6 and 7.12.1). Introducing the nth-“mode” inelastic SDF system also permitted extension of the well-established concepts for elastic systems to inelastic systems. Compare Eqs. (4.4) and (4.8) to Eqs. (3.9) and (3.10): note that Eq. (3.11) applies to both systems.4 Solution of the nonlinear Eq. (4.8) formulated in this manner provides Dn t , which substituted into Eq. (3.12) gives the floor displacements of the structure associated with the nth“mode” inelastic SDF system. Any floor displacement, story drift, or another deformation response quantity r t is given by Eqs. (3.13) and (3.14), where An t is now the pseudost acceleration response of the nth-“mode” inelastic SDF system. The two analyses that lead to rn

bg

bg

bg

bg

b g are shown schematically in Fig. 4.3. Equations (3.13) and (3.14) represent the response of the inelastic MDF system to p eff,n bt g , the nth-mode contribution to p eff bt g . Therefore the response of the system to the total excitation p eff bt g is given by Eqs. (3.15) and
and An t (3.16). This is the UMRHA procedure.
4

Equivalent inelastic SDF systems have been defined differently by other researchers (Villaverde, 1996; Han and Wen, 1997).

31

ζn. Roof displacement due to p eff. 4. 4.112 5 10 15 20 Time (sec) 25 30 −10 0 −10 0 Fig.89 50 r1 0 u (cm) 0 • 78.0 ¥ El Centro ground motion: (a) exact solution by NL-RHA. 2. and 3.02 −150 n=2 −150 50 n=2 14.s n u g t . where u g t = 3.n t = . n = 1. and (b) approximate solution by UMRHA bg af bg 32 .575 5 10 15 20 Time (sec) 25 30 ur3 (cm) 0 • 5. Fsn / Ln ¨ g(t ) u (a) Static Analysis of Structure (b) Dynamic Analysis of Inelastic SDF System Fig.4.Forces sn Unit mass An(t ) st rn ωn. Conceptual explanation of uncoupled modal response history analysis of inelastic MDF systems (a) Nonlinear RHA 150 n=1 150 (b) UMRHA n=1 ur1 (cm) • 75.9 • 0 −50 10 −50 10 n=3 ur3 (cm) 0 • 5.51 • ur2 (cm) 0 ur2 (cm) n=3 14.3.

4.956 5 10 15 20 Time (sec) 25 30 ∆r3 (cm) 0 • 5. respectively. These errors arise from the following assumptions and approximations: (1) the coupling among modal coordinates qn t arising from yielding of the system [recall Eqs. but even for this very intense excitation. The errors are slightly larger in drift than in displacement. the errors in either response quantity are only a few percent. where u g t = 3.0 ¥ El Centro ground motion.2… N ) according to Eq. and (b) approximate solution by UMRHA af bg bg 4.n t = .n t (n = 1. Top story drift due to p eff.s n u g t . (4.817 5 10 15 20 Time (sec) 25 30 −10 0 −10 0 Fig.4 and 4.2. (3.33 0 −20 20 −20 20 n=2 ∆r2 (cm) 0 ∆r2 (cm) 7.5. both for 3. 2.4) and (4.(a) Nonlinear RHA 20 n=1 20 (b) UMRHA n=1 • 5.1 Underlying Assumptions and Accuracy The approximate solution of Eq.5)] is neglected. (2) the superposition of responses to p eff. (4. this intense excitation was chosen to ensure that the structure is excited well beyond its linear elastic limit.0 ¥ El Centro ground motion: (a) exact solution by NL-RHA. Such comparison for roof displacement and top-story drift is presented in Figs.15) is strictly valid only for linearly elastic systems.008 • 6.6) by UMRHA is compared with the “exact” solution by nonlinear RHA. n = 1. and 3.744 • 0 −20 10 n=3 −20 10 n=3 ∆r3 (cm) 0 • 5.855 ∆r1 (cm) 0 ∆r1 (cm) n=2 • 6. 4. and (3) the Fsn Ln .5.Dn relation is 33 bg bg .

The structure is pushed using the force distribution of Eq.approximated by a bilinear curve to facilitate solution of Eq.8) governing Dn t is based on Eq. the base shear is Vbny and roof displacement is urny . Therefore we impose this constraint in developing the UMRHA procedure in this section and modal pushover analysis in the next section.20) to some bg pre-determined roof displacement. (3. The overall errors in the UMRHA procedure are documented in the examples presented in Section 4. within the linearly elastic range of the structure. A bilinear idealization of this pushover curve for the nth-“mode” is shown in Fig.2 Properties of the nth-“mode” Inelastic SDF System How is the Fsn Ln . (4. the only force distribution that produces displacements proportional to f n is given by Eq.10) . it can conduct a force-controlled nonlinear static analysis with an invariant distribution of lateral forces. and the base shear Vbn . (4. (4.9) should be determined by nonlinear static analysis of the structure as the structure undergoes displacements u = Dnφ n with increasing Dn .12) for floor displacements. such nonlinear static analysis provides the so-called pushover curve. Therefore.Dn curve. When implemented by commercially available software.4. when specialized for linearly elastic systems it is identical to the RHA procedure of Section 3. (3. this distribution seems to be a rational choiceeven for inelastic systemsto determine Fsn in Eq. is plotted against roof displacement urn . However. (4. 4.8) in UMRHA.9). (3.Dn relation to be determined in Eq.Dn relation? The two sets of forces and displacements are related as follows: V Fsn = bn Γn Dn = urn Γ nf rn 34 (4.6a.8) before it can be solved? Because Eq.1.urn pushover curve to the Fsn Ln . (4.20). What is an appropriate invariant distribution of lateral forces to determine Fsn ? For an inelastic system no invariant distribution of forces can produce displacements proportional to f n at all displacements or force levels. the relationship between lateral forces f s and Dn in Eq. 4. At the yield point. which is different than the Fsn Ln . Although several approximations are inherent in this UMRHA procedure.2. Although most commercially available software cannot implement such displacement-controlled analysis. How to convert this Vbn .

the initial slope of the pushover curve in Fig. (4. 4.8).13) This value of Tn . The two are related through Fsny Ln 2 = w n Dny (4. the elastic vibration period Tn of the nth-mode SDF system is computed from F Ln Dny I 1/ 2 Tn = 2p G H Fsny JK (4.2. 4. Knowing Fsny Ln and Dny from n Eq. 2001. Properties of the nth-“mode” inelastic SDF system from the pushover curve Equation 4.6a is 2 kn = ω n Ln .6b.11) * in which M n = Ln Γn is the effective modal mass (Chopra. Section 13. which is not a meaningful quantity. where the yield values of Fsn Ln and Dn are Fsny Ln = Vbny * Mn Dny = urny Γ nf rn (4.11). (4.10 enables conversion of the pushover curve to the desired Fsn Ln − Dn relation shown in Fig.6b is w 2 . 4. 35 .V bn (a) Idealized Pushover Curve F sn /L n (b) F sn / L − D Relationship n n Idealized Vbny Actual 1 αnkn V bny /M * n 1 α ω2 n n kn 1 1 ωn 2 u u rny rn Dny = ur n y / Γn φ r n D n Fig. which may differ from the period of the corresponding linear system.6.5).12) implying that the initial slope of the curve in Fig. should be used in Eq. In contrast. 4.

For the nth-mode. 2. 4. φn . Calculate histories of various responses by Eqs.6b) by utilizing Eq. 7. 4. develop the base-shear – roof-displacement ( Vbn − urn ) pushover curve for the force distribution s* [Eq. Convert the idealized pushover curve to the Fsn Ln − Dn relation (Fig. Idealize the pushover curve as a bilinear curve (Fig. Typically. the first two or three modes will suffice. ωn . n (3. 4. Consider a nonlinear static analysis of the structure subjected to lateral forces distributed over the building height according to s* [Eq. 4. 8. r o . (3. for linearly-elastic vibration of the building. 4.6b. Calculate the peak value. and pseudo-acceleration history. 6.13). of the total response r t obtained in Step 8. Dn (t ) . An t .3 Summary The inelastic response of an N-story building with plan symmetric about two orthogonal axes to earthquake ground motion along an axis of symmetry can be estimated as a function of time by the UMRHA procedure just developed.6a). with the structure is pushed to the roof displacement urno .2. 9. This value of the roof displacement is given by Eq.3 MODAL PUSHOVER ANALYSIS bg bg A pushover analysis procedure is presented next to estimate the peak response rno of the inelastic MDF system to effective earthquake forces p eff . of the nth-“mode” inelastic SDF system (Fig. n 3.4.15) and (3. (3.20)]. the peak value of Dn t . (3.12) and (3. Repeat Steps 2 to 6 for as many modes as required for sufficient accuracy. (3.n t . is now determined by 36 bg bg . and modes. (4.16) to determine the total response. 4. details are available in Appendix A: 1.11). Combine the “modal” responses using Eqs.3b) with force-deformation relation of Fig. which is summarized next as a sequence of steps.21) where Dn .20)]. Compute the deformation history. Compute the natural frequencies. 5.

(3. The peak “modal” responses rno . are combined using an appropriate modal combination rule. (4. which is summarized next as a sequence of steps. At this roof displacement. (4. e.3b) with force-deformation relation of Fig. 37 .6). it can be determined from the inelastic response (or design) spectrum (Chopra. The response value rno is an estimate of the peak value of the response of the inelastic system to p eff.solving Eq.g. (3. Calculate the peak roof displacement urno associated with the nth-“mode” inelastic SDF system from Eq. alternatively. It provides results for elastic buildings that are identical to the well-known RSA procedure (Section 3.3. story drifts. the lateral force distribution used possesses two properties: (1) it appears to be the most rational choice among all invariant distribution of forces.4. it seems reasonable because it provides results for elastic buildings that are identical to the well-known RSA procedure described in Section 3.8). joint rotations.1 Summary The peak inelastic response of a building to earthquake excitation can be estimated by the MPA procedure just developed. as mentioned earlier.6 and 7. Eq.. Sections 7. we will refer to rno as the peak “modal” response even in the case of inelastic systems. although somewhat intuitive for inelastic buildings.2 and 3. 4. Dn .8). details are available in Appendix B. However. as described in Section 4. for elastic systems. bg bg bg bg Thus.2. rno also represents the exact peak value of the nth-mode contribution rn t to response r t . Compute the peak deformation.3. 2001. 5.2. the pushover analysis provides an estimate of the peak value rno of any response rn t : floor displacements. plastic hinge rotations. As shown in Sections 3.3) because. each determined by one pushover analysis. (4. seems reasonable. 6. 4. etc. This pushover analysis.18). to obtain an estimate of the peak value ro of the total response.12). of the nth-“mode” inelastic SDF system (Fig. or from the inelastic response (or design) spectrum. Steps 1 to 4 of the MPA are same as those for UMRHA.21).n t .6b by solving Eq. 4. governed by Eq. This application of modal combination rules to inelastic systems obviously lacks a theoretical basis. and (2) it provides the exact modal response for elastic systems.

2). and three modes are compared with the “exact” values in Fig.18). Determine the total response by combining the peak “modal” responses using the SRSS combination rule of Eq.1). rno . extract from the pushover database values of other desired responses.8.3 cm.0 to 3.7 cm. 9. and ur 3o = 2. Figure 4. The peak values of displacements of all floors and drifts in all stories are presented in Tables 4.4. the UMRHA procedure lacks a rigorous theory. Observe that errors tend to decrease as response contributions of more “modes” are included. ur 2o = 11. three “modes” included. and three “modes”. At urno . also included are the combined responses due to one. say. in contrast to modal analysis (Section 3. 4. and the percentage errors in the approximate results. The peak values of response are as noted. and the errors in the approximate results are shown in Fig. is presented next.9. determined by the UMRHA procedure.7. 8. This is to be expected.4. 4. the peak roof displacement due to each of the three “modes” is ur1o = 48.7 shows the individual “modal” responses. To ensure that this structure responds well into the inelastic range the El Centro ground motion is scaled up by a factor varying from 1. in particular.53 cm. Typically. although the trends are not systematic as when the system remained elastic (Section 3. 4. The peak values of floor displacements and story drifts including one. the “exact” results. Repeat Steps 3 to 8 for as many “modes” as required for sufficient accuracy.4 COMPARATIVE EVALUATION OF ANALYSIS PROCEDURES The response of the 9-story building described earlier is determined by the two approximate methods: UMRHA and MPA. This deficiency also implies that. the 38 .1 Uncoupled Modal Response History Analysis The structural response to 1.2. two.1 and 4. 4. and the “exact” response from nonlinear RHA for the roof displacement and top-story drift. From the total rotation of a plastic hinge.0. respectively. and compared with the results of a rigorous nonlinear RHA using the DRAIN-2DX computer program.5 ¥ the El Centro ground motion including the response contributions associated with three “modal” inelastic SDF systems. two. (3. subtract the yield value of hinge rotation to determine the hinge plastic rotation. the combined response due to three “modes”. with. the first two or three “modes” will suffice.

the errors in story drifts are larger compared to floor displacements.2).25 (Tables 3.2) and 1.0.44 • "Mode" 2 r1 0 • 48. (a) Roof Displacement (b) Top Story Drift 80 12 0 • 3. 2. Response histories of roof displacement and top-story drift due to 1. just as for elastic systems. sec 25 30 r −12 0 Fig.5.75. 1. sec 25 30 • 48.3 11. Next. in particular.1 and 4. 4. and total response from NL-RHA 39 .7 • −80 80 ∆r1 (cm) u (cm) "Mode" 1 −12 r2 0 ∆r2 (cm) 12 0 u (cm) −80 80 −12 r3 0 • 2. However. 0.1 and 3.38 6. and 3. 1.85.25. the degree to which the system deforms beyond its elastic limit.53 ∆r3 (cm) 12 0 • 2. we investigate how the errors in the UMRHA vary with the deformation demands imposed by the ground motion.response is much less accurate (Tables 4.24 • NL−RHA r −12 12 ∆r (cm) u (cm) 0 5 10 15 20 Time.5.6 5 10 15 20 Time. For each excitation.1 −12 12 ∆r (cm) u (cm) UMRHA 3 "Modes" 0 • 7.0.1 and 3. the errors in responses computed by UMRHA including three “modes” relative to the “exact” response were determined.62 5.5 (Tables 4.5 × El Centro ground motion: individual “modal” responses and combined response from UMRHA.88 u (cm) "Mode" 3 −80 80 0 −80 80 0 −80 0 • 44.2) if the system yields significantly versus if the system remains within the elastic range (Tables 3.1 and 4. 0.7. 0. for a fixed number of “modes” included.0. recall that the computed errors have been presented earlier for ground motion multipliers 0. For this purpose the UMRHA and exact analyses were repeated for ground motions of varying intensity. defined as the El Centro ground motion multiplied by 0.2).

Heightwise variation of floor displacements and story drift ratios from UMRHA and NL-RHA for 1. For this purpose. (3. the pushover curves using force distributions s* [Eq.20)] for the first three n modes of the system are shown in Fig.5 1 1.5 × El Centro ground motion (a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th (b) Story Drift Ratios Floor 4th 3rd 2nd 1st UMRHA 1 "Mode" 2 "Modes" 3 "Modes" Floor 5th 5th 4th 3rd 2nd 1st UMRHA 1 "Mode" 2 "Modes" 3 "Modes" Ground −60 −40 −20 0 Error (%) 20 40 60 Ground −60 −40 −20 0 Error (%) 20 40 60 Fig. indicated by a ground motion multiplier.10 summarizes the error in UMRHA as a function of ground motion intensity.5 1 1.11. Heightwise variation of error in floor displacements and story drifts estimated by UMRHA including one. 4.8.5 Story Drift Ratio (%) 2 2. Shown is the error in each floor displacement (Fig. with the peak displacement of each “modal” SDF system noted for each ground motion multiplier.(a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th NL−RHA UMRHA 1 "Mode" 2 "Modes" 3 "Modes" (b) Story Drift Ratios Floor Floor 5th 4th 3rd 2nd 1st 5th 4th 3rd 2nd 1st NL−RHA UMRHA 1 "Mode" 2 "Modes" 3 "Modes" Ground 0 0. 4.10b).5 Fig.5 ¥ El Centro ground motion Figure 4.10a). 4. it will be useful to know the deformation of the system relative to its yield deformation. To interpret these results. 4.5 Displacement/Height (%) 2 Ground 0 0. two.9. in each story drift (Fig. Two versions of the pushover curve are 40 . 4. or three “modes” for 1. and the error envelope for each case.

4. and plastic hinge rotations at the external beam end at each floor level (Table 4.7 cm. the peak values of displacements at all floors (Table 4. the first reason mentioned above seems to be the primary source for the errors. as mentioned in Section 4. For more intense excitations.7).5 ¥ the El Centro ground motion. determined from “exact” analyses. the ductility factor for the first mode system is only 1.2. and (4) are larger in story drift compared to floor displacements. the errors in truncating the higher mode contributions are negligible.4) to roof displacements urno = 48.01 (Fig. even though the system remains essentially elastic.11).53 cm. (3. 3. drifts in all stories (Table 4.10 permits the following observations regarding the accuracy of the UMRHA procedure: the errors (1) are small (less than 5%) for ground motion multipliers up to 0. UMRHA lacks a rigorous theory and is based on several approximations.75.5).2 Modal Pushover Analysis The results of modal pushover analysis procedure considering the response due to the first three “modes” was implemented for the selected building subjected to 1.2. Figure 4. Perhaps this explains why the errors are large at this excitation intensity.10) for each “modal” inelastic SDF system (Figs 4. and 3 (Fig. First.4.3 cm.6 and 4. 4. respectively.included: the actual curve and its idealized bilinear version.0. Each of these three pushover analyses provides the pushover curve (Fig. (4.75. The idealized curve for the first “mode” deviates most from the actual curve near the peak displacement corresponding to ground motion multiplier 1. (3) maintain roughly similar values for more intense ground motions. Additional errors are introduced in UMRHA of systems responding beyond the linearly elastic limit for at least two reasons. The structure is pushed using the force distribution of Eq.0. 4. 4.4). 41 . 2. (2) increase rapidly as the ground motion multiplier increases to 1. were noted but are not shown here.3). as mentioned in Section 3. Second. and 2. The system remains elastic up to ground motion multiplier 0.1). the values determined by RHA of the nth-mode inelastic SDF system (Fig.20) with n = 1. 11. the pushover curve for each “mode” is idealized by a bilinear curve in solving Eq. The location of plastic hinges and their rotations. and.11a). 4.

0 11.663 -0.5 1 1.373 -0.5 3 0 0 0.490 -1.1 1.763 -15.2 4.315 -0.7 Table 4.126 0.256 1.121 -0.914 2.811 1.057 -0.372 -1.945 -37.003 -31.3 1.214 -0.169 0.490 1.676 0.3 42 .202 11. and (b) story drifts Table 4.495 1.291 0.2 1.942 -0.5 18.298 -0.187 -0.513 -0.4 -7.616 -0.0 9.478 0.1 Peak values of floor displacements (as % of building height = 37.096 0.104 0.698 1.376 -1.8 1.8 1.220 -0.071 -0. 4.819 2.9 31.070 1.226 -0. Noted 60 50 40 (b) Story Drifts Error Envelope Error for Story No.9 12.554 1.068 0.668 -23.938 1.293 1.900 -10.200 8.472 1.727 1.982 9.298 0.430 1.201 1.751 1.220 0.0 7.003 0.707 1.371 -0.0 1.877 0.3 6.5 10.333 0.3 25.4 1.0 -9.2 Peak values of story drift ratios (as % of story height) from UMRHA for 1.010 0.338 -1.5 1 1.079 0.856 2.379 1.214 0.5 × El Centro ground motion Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Modal Response “Mode” 1 “Mode” 2 Drift Ratio (%) Combined (UMRHA) 1 “Mode” 2 “Modes” “Mode” 3 -1.2 1.526 -0.044 1.410 1.065 0.9 16.938 -1.201 -1.7 31.407 -10.663 0.0 11.473 -22.256 -1.338 1.513 0.5 -3.8 0.7 14.6 4.366 -0.806 0.426 -1.235 -0.2 6.4 4.366 0.722 0.9 12.376 1.049 -0.10.5 28.371 0.136 1.5 3 Fig.018 0. Noted 7 6 Error (%) 30 1 Error (%) 30 5 8 20 6 5 2 20 2 4 9 9 10 0 0 10 3 8 7 3 1 4 0.410 -1.372 1.8 1. Errors in UMRHA as a function of ground motion intensity: (a) floor displacements.241 -1.863 1.2 12.2 22.248 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.14 m) from UMRHA for 1.1 8.1 0.852 1.1 3.317 0.5 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Displacement /Height (%) “Modal” Response Combined (UMRHA) “Mode” 1 “Mode” 2 “Mode” 3 1 “Mode” 2 “Modes” -0.820 -19.863 0.241 1.350 -0.033 0.5 2 GM Multiplier 2.9 31.201 -0.942 1.844 -25.071 0.971 1.6 2.133 1.5 9.983 1.072 1.370 -0.130 0.772 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 1.3 8.(a) Floor Displacements 60 50 40 Error Envelope Error for Floor No.009 -0.820 -0.8 14.575 -41.9 5.088 10.3 0.154 0.138 1.540 0.5 2 GM Multiplier 2.484 0.256 1.914 -0.806 -0.8 1.120 1.5 28.072 -1.216 1.283 1.055 -0.6 1.5 9.4 -1.0 2.260 -15.135 9.8 1.1 1.

significant improvement is achieved by including response contributions due to the second “mode”. The locations of plastic hinges shown in Fig.12 and Tables 4.4). but was more successful when the excitation was more intense (results are not included).13 and Tables 4.5 present estimates of the combined response according to Eq. (3.14.3 and 4. One “mode” pushover analysis was unable to identify the plastic hinges in the upper stories where higher mode contributions to response are known to be more significant.14.12 and Tables 4. however. but the third “mode” contributions do not seem especially important (Fig. 4.12 and 4.13 and Tables 4.4.3 and 4. Observe that the primary contributor to plastic hinge rotations in the lower stories is the first “mode” and the second “mode” in the upper stories. and three “modes. 4.18).3 through 4. Pushover analysis seems to be inherently limited in computing accurately hinge plastic rotations. the errors in the modal pushover results are.9 and Tables 4. and are within a few percent of the exact values for the upper stories. the third “mode” does not contribute because this SDF system remains elastic (Fig. especially in estimating the story drifts (Fig. MPA including three “modes” underestimates the displacements of the lower floors by up to 8% and overestimates the upper floor displacements by up to 14%. the hinges identified in beams at the 6th floor were at variance with the “exact” results. Obviously.4). considering one.1 and 4. 4. overestimated by up to 18% in the middle stories. although the error is recorded as 100% if MPA estimates zero rotation when the nonlinear RHA computes a non-zero value.11c). significantly smaller than in UMRHA (compare Fig. the results were not always accurate. for two or three modes included. MPA failed to identify the plastic hinges at the column bases in Fig. The first “mode” alone is inadequate. 4.5).3 and 4. 4. Furthermore.” and nonlinear RHA. were determined by four analyses: MPA considering one “mode. even if three “modes” are included (Fig. and the errors in these estimates relative to the exact response from nonlinear RHA. The errors are especially large in the hinge plastic rotations estimated by the MPA procedure. The second “mode” was necessary to identify hinges in the upper stories.13 with Fig. 4. The drifts are underestimated by up to 13% in the lower stories. in general. Fortuitously. However. this error is not especially significant because the hinge plastic rotation is very small.4 with Tables 4.” respectively. the additional errors due to the approximation inherent in modal combination rules tend to cancel out the errors due to the various approximation contained in the UMRHA. For example. 4. 4.2). two.3 and 4.” two “modes.” and three “modes.Figures 4. As shown in Figs.13 and Table 4. 4. 43 .

19 y by 3 2 1.75 Actual Idealized 20 40 60 Roof Displacement (cm) (b) "Mode" 2 Pushover Curve 80 12000 10000 u = 9. 4.75.5 1 0. “Modal” pushover curves with peak roof displacements identified for 0.85 0. and 3.5 2 3 4000 2000 0 0 Actual Idealized 0.5.25 0.13 y by Base Shear (kN) 8000 6000 1 1.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36.5 2 3 Actual Idealized 2000 0 0 0. V = 5210 kN. 0.14 y by Base Shear (kN) 8000 6000 4000 1. α = 0.0.25 5 10 15 20 Roof Displacement (cm) 25 (c) "Mode" 3 Pushover Curve 12000 10000 u = 4.2 cm. 1.5 0.25 1 2 4 6 Roof Displacement (cm) 8 10 Fig.75 0.85 0.0 × El Centro ground motion 44 .5 0.75 0.5 0. α = 0. V = 4952 kN.25. 1.5. 0.85 Base Shear (kN) 8000 6000 4000 2000 0 0 0.11. V = 7616 kN.9 cm. α = 0.6 cm. 2.0.

implying excitations intense enough to cause significant yielding of the structure. implying that the modal combination errors contained in these procedures are acceptable. an observation with practical implications. as discussed in Sections 3.4 also apply to MPA. 4. each story drift (Fig.15a).0. the errors in MPA were fortuitously smaller than in UMRHA (compare Figs.5 Displacement/Height (%) 2 Ground 0 0.4.(a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th NL−RHA MPA 1 "Mode" 2 "Modes" 3 "Modes" (b) Story Drift Ratios Floor Floor 5th 4th 3rd 2nd 1st 5th 4th 3rd 2nd 1st NL−RHA MPA 1 "Mode" 2 "Modes" 3 "Modes" Ground 0 0.15) for ground multipliers larger than 1. 45 . In this case.75. Shown is the error in each floor displacement (Fig. While various sources of errors in UMRHA. now standard in engineering practice.5 Story Drift Ratio (%) 2 2.15b).4.3. and the error envelope for each case.15 summarizes the error in MPA considering three “modes” as a function of ground motion intensity.2 and 3.5 Fig.5 × El Centro ground motion. whereas MPA contains errors inherent in modal combination rules.10 and 4. However. shading indicates errors in MPA including three “modes” Figure 4. 4. The errors are only weakly dependent on ground motion intensity (Fig.15). The fact that MPA is able to estimate the response of buildings responding well into the inelastic range to a similar degree of accuracy indicates that this procedure is accurate enough for practical application in building retrofit and design. As mentioned in Section 3. identified in Section 3.12 Heightwise variation of floor displacements and story drift ratios from MPA and NL-RHA for 1. implying excitations weak enough to limit the response in the elastic range of the structure. 4.5 1 1. 4. indicated by a ground motion multiplier.5 1 1. the errors in MPA were larger for ground motion multipliers less than 0. the MPA procedure is equivalent to the RSA procedure. 4. UMRHA is essentially exact.3 for elastic systems (or weak ground motions).

and hinge plastic rotations estimated by MPA including one. story drifts. 4.13. two. Errors in floor displacements.(a) Floor Displacements 9th 8th 7th 6th Floor 5th 4th 3rd 2nd 1st MPA 1 "Mode" 2 "Modes" 3 "Modes" Ground −60 −40 −20 0 Error (%) 20 40 60 (b) Story Drift Ratios 9th 8th 7th 6th Floor 5th 4th 3rd 2nd 1st MPA 1 "Mode" 2 "Modes" 3 "Modes" Ground −60 −40 −20 0 Error (%) 20 40 60 (c) Hinge Plastic Rotations 9th 8th 7th 6th Floor 5th 4th 3rd 2nd 1st MPA 1 "Mode" 2 "Modes" 3 "Modes" Ground −120 −80 −40 0 Error (%) 40 80 120 Fig. and three “modes” for 1.5 ¥ El Centro ground motion 46 .

1 46.737 1.222 0.9 -4.00E+00 3 “Modes” 7.705 1.5 2.55E-03 3.0 2 “Modes” -32.00E+00 0.737 0.176 0.426 15.6 13.414 1.1 -8.00E+00 0.066 -0.37E-03 1.1 62.116 1.260 -14.473 -15.4 Peak values of story drift ratios (as % of story height) from MPA for 1.298 -0.5 2.2 -100.88E-03 0.053 1.879 1.640 1.6 -44.202 8.1 62.1 46.233 1.304 1.9 -100.00E+00 0.895 1.200 8.5 × El Centro ground motion Displacement /Height (%) “Modal” Response Combined (MPA) “Mode” 1 “Mode” 2 “Mode” 3 1 “Mode” 2 “Modes” Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 0.10E-02 9.76E-03 4.250 0.1 18.8 -6.60E-03 2.8 -29.763 -14.0 -50.6 -8.1 1.8 0.018 -0.26E-04 9.00E+00 1 “Mode” 7.37E-03 1.116 1.7 1.518 1.018 0.305 -0.0 -100.18E-03 7.0 -100.130 0.517 1.9 5.88E-03 0.8 7.36E-03 6.02E-03 0.0 -5.331 1.781 0.88E-03 1.02E-03 3.945 -49.00E+00 0.8 -12.2 -3.00E+00 “Mode” 3 0.745 1.049 -0.37E-04 Error (%) “Modal” Response 1 “Mode” -32.0 1.9 0.611 0.8 -29.053 -1.37E-03 1.088 12.982 13.5 Peak values of hinge plastic rotations (radians) from MPA for 1.351 -0.668 -13.8 -29.00E+00 0.2 6.1 62.0 47 .9 -100.5 10.007 1.068 0.738 1.2 11.19E-10 3.804 1.00E+00 0.683 1.8 1.5 × El Centro ground motion Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th “Modal” Response “Mode” 1 “Mode” 2 Drift Ratio (%) Combined (MPA) 1 “Mode” 2 “Modes” 3 “Modes” Error (%) NL 1 2 3 RHA “Mode” “Modes” “Modes” 1.55E-03 3.018 -0.101 -0.6 7.1 46.3 1.9 1.594 -1.00E+00 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th “Mode” 2 0.76E-03 4.652 1.516 0.156 -0.733 1.429 0.781 0.5 -6.498 1.6 13.14 m) from MPA for 1.209 1.980 -0.012 1.00E+00 2 “Modes” 7.2 0.76E-03 4.168 -0.057 0.00E+00 0.372 0.220 1.00E+00 0.6 1.527 -0.581 0.72E-03 7.154 0.0 -50.00E+00 0.00E+00 0.Table 4.1 13.667 -1.00E+00 0.6 -7.5 × El Centro ground motion Hinge Plastic Rotation (rad) “Modal” Response Combined (MPA) “Mode” 1 7.105 0.3 -3.514 -1.009 0.244 0.135 -7.233 1.298 0.22E-10 NL RHA 1.687 0.071 -0.003 -16.00E+00 0.900 -0.259 1.3 13.116 1.911 0.503 -1.311 0.5 Table 4.015 0.37E-03 1.00E+00 0.76E-03 4.399 0.478 0.36E-03 6.910 1.756 0.033 -0.00E+00 0.72E-03 7.6 -44.222 0.266 -0.00E+00 3.066 -0.304 -1.371 -0.00E+00 0.942 6.007 1.0 -100.4 -8.00E+00 0.641 1.5 10.9 7.581 0.666 Table 4.844 -7.820 -7.2 -100.2 1.728 1.60E-04 7.00E+00 0.8 17.72E-03 7.8 “Mode” 3 -1.6 -9.2 0.2 -4.02E-03 3.0 3 “Modes” -32.220 1.99E-03 6.315 0.071 0.00E+00 0.694 1.055 0.36E-03 6.00E+00 0.667 1.752 1.02E-03 0.338 1.72E-03 7.05E-03 2.0 15.50E-10 3.9 2.407 -27.2 -100.980 0.3 -3.7 -12.197 -0.3 Peak values of floor displacements (as % of building height = 37.640 -1.190 -0.55E-03 3.503 1.5 7.125 -1.8 -4.118 0.8 1.756 0.336 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.705 -1.435 0.5 1.015 0.614 0.895 1.3 11.5 2.36E-03 6.399 0.7 1.8 -6.4 1.575 -53.53E-03 7.9 1.

4. and plastic hinge rotations at selected external beam end at each floor level (Table 4.17 and Tables 4. 4.7 cm and ur 3o = 2. The errors are especially large in the plastic hinge rotations estimated by the MPA procedure even if three “modes” are included (Fig.8 present estimates of the combined response according to Eq.53 cm. The pushover curves for the second and third modes included in Fig. (3.6).4.7). As shown in Fig.18 and 48 .” and the errors in these estimates relative to the exact response from nonlinear RHA. and three “modes.16a—which is slightly different than the one excluding gravity loads (Fig. 3.6 and 4.” these effects will modify the combined modal response as well as the results of nonlinear RHA. They are now included in the pushover analysis of the structure for the first “mode” only.4.16 are unchanged. (3. Also presented at these roof displacements are the displacements at all floors (Table 4. two.6 and 4.4) to roof displacement ur1o = 52. Starting with its initial state under gravity loads.17 and 4.7). The story drifts are underestimated by up to 7% in the lower stories. 4.3 .5 × El Centro ground motion are presented below.6 through 4. 4. The response contributions of the second and third modes are the same as before (Tables 4. 4.” however.3 Modal Pushover Analysis with Gravity Loads To evaluate the accuracy of the dynamic response of the system. The first “mode” alone provides adequate estimates of floor displacements.20) with n = 1 (Fig.5).” but not the contributions of higher “modes.4. the results presented so far did not include gravity load effects.18 and Tables 4. Figures 4. 4. but it is inadequate in estimating the story drifts (Fig.7.20) with n = 1 to a target roof displacement.7). 4. the structure is pushed using the force distribution of Eq. as are the roof displacements ur 2o = 11. Obviously gravity load effects will influence the seismic demands due to the first “mode.17 and Tables 4. drifts in all stories (Table 4. Static analysis of the structure for gravity loads provides the initial state—forces and deformations—of the structure. the third “mode” contributions do not seem especially important (Fig. (3.18) considering one.0 cm. MPA including two “modes” underestimates the displacements of lower floors by up to 6% and overestimates the upper floor displacements by up to 22%. and overestimated by up to 29% in the middle stories and by up to 13% in the upper stories.11a). and the structure is pushed using the force distribution of Eq.8).6 and 4. Results of such analyses for the selected building subject to 1.18 and Tables 4. Significant improvement is achieved by including response contributions due to the second “mode. resulting in the pushover curve shown in Fig.

14.5 3 Fig. Noted Error (%) 30 20 10 0 0 2 8 3 4 1 9 Error (%) 30 20 5 4 6 5 7 2 9 10 0 0 1 8 7 3 6 0. and (b) story drifts 49 . 4. and three “modes” and by NL-RHA for 1. 1−"Mode" • • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • •• • •• • • • • • •• •• • •• •• •• •• (b) MPA. Noted 60 50 40 (b) Story Drifts Error Envelope Error for Story No.15. 4. Errors in MPA as a function of ground motion intensity: (a) floor displacements.Table 4.5 1 1.5 1 1.5 2 GM Multiplier 2.5 × El Centro ground motion (a) Floor Displacements 60 50 40 Error Envelope Error for Floor No.5 2 GM Multiplier 2. 2−"Modes" •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• • •• (c) 3−"Modes" • • • • • • • •• •• • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• • •• • • • • • • • • • (d) Nonlinear RHA •• •• • •• •• •• •• • •• •• •• •• •• •• •• • •• •• • •• •• •• •• • • • • • • •• •• •• Fig.5 3 0. two. (a) MPA. Locations of plastic hinges determined by MPA considering one. Most pushover analysis procedures do not seem to compute to acceptable accuracy plastic hinge rotations.8).

20b).15. 4.The locations of plastic hinges shown in Fig. and the error envelope for each case. 4. and nonlinear RHA. With two modes included in MPA. Shown is the error in each floor displacement (Fig. Figure 4.20 summarizes the error in MPA considering three “modes” as a function of ground motion intensity. 4.20a). These errors are only slightly larger than those in Fig.19 were determined by four analyses: MPA considering one “mode. The second “mode” is necessary to identify hinges in the upper stories. MPA provides response values accurate enough for practical application in building retrofit or design. indicated by a ground motion multiplier. this procedure is able to predict plastic hinge locations essentially consistent with nonlinear RHA. excluding gravity load effects. each story drift (Fig. 50 . One “mode” pushover analysis is unable to identify the plastic hinges in the upper stories where higher mode contributions to response are known to be more significant. and the errors are only weakly dependent on ground motion intensity.” two “modes”. 4.

16. 1. V = 5210 kN.0. 0.25 0. 2.5 0.25.3 cm. V = 4952 kN.5 1 0. 0. α = 0. 4. “Modal” pushover curves with gravity loads included.75.5 0.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36.5 2 3 Actual Idealized 2000 0 0 0.75 0.75 0.75 Actual Idealized 20 40 60 Roof Displacement (cm) 80 (b) "Mode" 2 Pushover Curve 12000 10000 u = 9. noted are peak values of roof displacement for 0. α = 0.19 y by 3 2 1.25 5 10 15 20 Roof Displacement (cm) 25 (c) "Mode" 3 Pushover Curve 12000 10000 u = 4.14 y by Base Shear (kN) 8000 6000 4000 1.0 × El Centro ground motion 51 . V = 7433 kN.5 0.5 2 3 4000 2000 0 0 Actual Idealized 0.50.85 0. 0.85 Base Shear (kN) 8000 6000 4000 2000 0 0 0.13 y by Base Shear (kN) 8000 6000 1 1.9 cm.6 cm. α = 0.25 1 2 4 6 Roof Displacement (cm) 8 10 Fig.85. and 3.0.85 0.

5 1 1. Heightwise variation of floor displacements and story drift ratios from MPA and NL-RHA for 1.(a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th NL−RHA MPA 1 "Mode" 2 "Modes" 3 "Modes" (b) Story Drifts Floor Floor 5th 4th 3rd 2nd 1st 5th 4th 3rd 2nd 1st NL−RHA MPA 1 "Mode" 2 "Modes" 3 "Modes" Ground 0 0. gravity loads included. Errors in floor displacements. two.5 Displacement/Height (%) 2 Ground 0 0. shading indicates errors in MPA including three “modes” (a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th (b) Story Drift Ratios Floor 4th 3rd 2nd 1st MPA 1 "Mode" 2 "Modes" 3 "Modes" Floor 5th 5th 4th 3rd 2nd 1st MPA 1 "Mode" 2 "Modes" 3 "Modes" Ground −60 −40 −20 0 Error (%) 9th 8th 7th 6th 20 40 60 Ground −60 −40 −20 0 Error (%) 20 40 60 (c) Hinge Plastic Rotations Floor 5th 4th 3rd 2nd 1st MPA 1 "Mode" 2 "Modes" 3 "Modes" Ground −120 −80 −40 0 Error (%) 40 80 120 Fig.18. story drifts.5 × El Centro ground motion.5 1 1. and 52 .17.5 Fig. 4. 4.5 Story Drift Ratio (%) 2 2. and hinge plastic rotations estimated by MPA including one.

and three “modes” and by NL-RHA for 1.(a) MPA.19. 3−"Modes" (d) Nonlinear RHA • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• •• •• • • • • • • • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • • • • • •• •• •• Fig. Errors in MPA as a function of ground motion intensity: (a) floor displacements. gravity loads included 53 .5 3 0. two.5 × El Centro ground motion.5 1 1. Noted 60 50 40 (a) Story Drifts Error Envelope Error for Story No.5 2 GM Multiplier 2.5 1 1. Noted Error (%) 30 20 6 8 1 5 2 4 3 9 7 Error (%) 30 4 5 20 6 9 7 1 10 0 0 10 0 0 8 2 3 0. 2−"Modes" •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• •• •• (c) MPA. gravity loads included (a) Floor Displacements 60 50 40 Error Envelope Error for Floor No.5 2 GM Multiplier 2.20. 1−"Mode" • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • •• •• •• • • • • • •• •• •• •• •• •• •• (b) MPA. Locations of plastic hinges determined by MPA considering one. 4.5 3 Fig. 4. and (b) story drifts.

478 0.6 2.19E-03 1.35E-03 8.371 -0.5 10.5 1.821 -1.00E+00 0.7 -2.921 1.0 37.128 -1.00E-03 5.190 -0.213 1.687 0.0 16.04E-02 8. gravity loads included Hinge Plastic Rotation (rad) Floor “Modal” Response “Mode” 1 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 8.037 0.7 16.983 1.854 0.850 -1.55E-03 3.00E+00 0.066 -0.033 -0.068 0.938 1.114 1.199 16.26E-03 3.2 4.998 21.5 0.1 21.04E-10 3.2 21.213 1.8 1.908 -1.1 1.55E-03 3.5 54 .00E+00 0.19E-04 5.821 1.515 -50.9 -3.17E-03 9.877 -46.9 -4.7 1.8 0.156 -0.2 4.4 1.176 0.109 0.102 1.594 -1.927 1.018 -0.530 1.530 1.530 1.00E+00 2 “Mode” 8.2 -3.6 19.434 0.8 Peak values of hinge plastic rotations (radians) from MPA for 1.00E+00 “Mode” 3 0.6 0.353 -23.2 16.00E+00 3 “Modes” 8.319 1.831 0.88E-03 0.0 -30.00E+00 0.466 0.23E-02 1.399 0.00E+00 0.19E-03 1.8 -32.064 -10.00E+00 0.11E-03 9.13E-03 5.114 -1. gravity loads included Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th “Modal” Response “Mode” 1 “Mode” 2 Drift Ratio (%) Combined (MPA) 1 “Mode” 2 “Modes” 3 “Modes” Error (%) NL 1 2 3 RHA “Mode” “Modes” “Modes” 1.098 20.207 18.00E+00 0.2 4.049 -0.830 -12.00E+00 0.11E-03 9.071 -0.7 19.330 1.310 1.00E+00 Combined (MPA) 1 “Mode” 8.168 -0.00E+00 0.00E+00 Error (%) “Modal” Response 1 “Mode” -32.197 -0.7 4.88E-03 1.3 -100.3 13.372 0.14 m) from MPA for 1.3 -22.461 0.Table 4.3 -100.434 0.4 -6.3 -22.603 1.983 1.454 1.3 9.102 1.831 0.00E-03 5.673 Table 4.8 28.8 -32.4 20.266 -0.858 2.5 -5.11E-03 9.23E-03 3.2 -0.75E-03 0.00E+00 0. gravity loads included Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Displacement /Height (%) “Modal” Response Combined (MPA) “Mode” 1 “Mode” 2 “Mode” 3 1 “Mode” 2 “Modes” 0.23E-03 0.4 -4.754 0.2 9.908 1.5 “Mode” 3 -1.071 0.686 -7.00E-03 5.107 1.257 0.507 1.4 1.3 1.00E+00 0.11E-03 9.783 1.603 -1.037 -0.154 0.754 1.429 -1.00E-10 NL RHA 1.996 -0.315 0.8 0.2 1.305 -0.35E-03 8.00E+00 “Mode” 2 0.7 Peak values of story drift ratios (as % of story height) from MPA for 1.2 2.0 37.5 × El Centro ground motion.214 0.23E-03 3.00E+00 3.6 1.744 1.78E-03 1.0 2 “Modes” -32.057 0.998 0.015 0.436 1.7 -2.2 -2.953 15.933 1.311 0.23E-03 0.237 0.00E+00 0.5 1.105 0.836 -0.101 -0.9 31.0 37.9 -6.490 -11.055 0.35E-03 8.637 0.998 1.130 0.996 0.00E-03 5.2 1.2 0.5 × El Centro ground motion.19E-03 1.4 26.399 -0.4 -4.1 13.263 0.00E+00 0.5 × El Centro ground motion.0 -30.35E-03 8.0 -100.0 -100.00E+00 0.270 -12.066 -0.2 9.00E+00 0.667 0.6 Peak values of floor displacements (as % of building height = 37.665 0.55E-03 3.516 0.728 1.813 1.527 -0.414 28.00E+00 0.00E+00 0.263 1.5 3 “Modes” -32.6 2.8 Table 4.330 1.213 1.00E+00 0.514 -1.822 1.009 0.888 1.854 0.3 -22.434 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.429 1.125 -1.860 1.351 -0.00E+00 0.3 -100.35E-10 3.913 7.2 12.5 21.319 1.2 9.19E-03 1.88E-03 0.5 21.8 9.1 4.850 1.637 0.237 0.

5.1 FEMA-273 PUSHOVER ANALYSIS In this investigation we focus on one step in the nonlinear static procedure in the FEMA-273 document [Building Seismic Safety Council. 2… N ). computed at the target displacement represent the earthquakeinduced demands on the structure. MPA considering three “modes. j 2. a plot of base shear versus roof displacement. Specified in FEMA-273 are three distributions for lateral forces: 1. and the exponent k = 1 for fundament period T1 £ 0. and 3. gravity load effects were included in all 55 .5 sec . is determined by nonlinear static analysis of the structure subjected to lateral forces with invariant distribution over height but gradually increasing values until a target value of roof displacement is reached. Equivalent lateral force (ELF) distribution: s* = m j h k where h j is the height of the j j jth floor above the base. joint rotations. etc.2 COMPARATIVE EVALUATION Compared in this section are the earthquake-induced demands for the selected building determined by five analyses: pushover analysis using the three force distributions in FEMA-273.5. k = 2 for T1 ≥ 2. plastic hinge rotations. SRSS distribution: s* is defined by the lateral forces back-calculated from the story shears determined by response spectrum analysis of the structure.” and nonlinear RHA. and varies linearly in between. 1997] The pushover curve. “Uniform” distribution: s* = m j (where the floor number j = 1.5 sec . Comparison of Modal and FEMA Pushover Analyses 5.. assumed to be linearly elastic. story drifts. The floor displacements.

3. and the displacements of upper floors are overestimated by 9-22%. The first four floor displacements are within 5% of the “exact” value.1. (b) ELF.11 0. 0.0913 0.1.5 times the El Centro ground motion. and Table 5.1. The MPA procedure is more accurate than all the FEMA force distributions.177 0.2. 5.119 0.00719 0.0466 0.0654 0. 5. and the locations of all plastic hinges in Fig.042 0. 5.367 (a) Uniform (b) ELF (c) SRSS Fig.126 0.112 0.3.3. 5. the story drift demands in Fig.281 0.11 0.0197 0. Figures 5.1 through 5. wherein the first two are obvious and the third is determined from response spectrum analysis of the building (Appendix C).0896 0. the value determined from RHA of the first-mode inelastic SDF system for 1. Force distributions in FEMA-273: (a) “uniform”.11 0.11 0.3a and Table 5.11 0. 5.3b and Table 5.11 0.21 0.165 0. and (c) SRSS 56 . The three FEMA force distributions are presented in Fig.3a and 5. pushover analyses are implemented for a target roof displacement of 52.1 demonstrate that the displacement demands are underestimated by the ELF and SRSS force distributions by 12 to 30 % at the lower six floors of the building. 5.11 0.0702 0. The “uniform” distribution overestimates all floor displacements by 17-28%.analyses.0 cm.2. the floor displacement demands in Fig.062 0.4 and Tables 5. The errors in the FEMA and MPA estimates of seismic demands relative to the “exact” demands are presented in Fig.0981 0. The pushover curves are given in Fig. both presented in Section 4. 5.4.4a. plastic hinge rotation demands in Table 5.0381 0. Also included in these presentations are the MPA results considering three “modes” and the “exact” demands from nonlinear RHA.0446 0. Using each of these force distributions.4. with the errors being larger for the SRSS distribution.

V = 7456 kN.14 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized 10 20 30 40 Roof Displacement (cm) (b) ELF Distribution 50 60 12000 10000 u = 38. α = 0.(a) Uniform Distribution 12000 10000 u = 33 cm. α = 0. gravity loads are included 57 . Pushover curves using three force distributions in FEMA-273: (a) “uniform”.18 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized 10 20 30 40 Roof Displacement (cm) (c) SRSS Distribution 50 60 12000 10000 u = 39.8 cm. and (c) SRSS.3 cm. (b) ELF.2. V = 8530 kN. V = 6897 kN.24 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized 10 20 30 40 Roof Displacement (cm) 50 60 Fig. 5. α = 0.

5 1 1.4c and Table 5.2 demonstrate that the story drift demands are greatly underestimated by all the FEMA force distributions. For the ELF distribution.) The pushover analysis procedures considered seem incapable of computing accurately local response quantities. and overestimated by no more than 32%. 3 "Modes" FEMA Uniform ELF SRSS 0. reaching 35%. For the uniform distribution. 5.Figures 5. MPA including three “modes.3b and 5. the MPA procedure is more accurate than all the FEMA force distributions. whereas nonlinear RHA computed an insignificantly small value.4b.5 Story Drift Ratio (%) 2 2. This seems to be an inherent limitation of pushover analysis.5 Displacement/Height (%) 2 Floor 5th 5th 4th 3rd 2nd 1st NL−RHA MPA.5 1 1.5 Ground 0 Ground 0 Fig. such as hinge plastic rotations.3 demonstrate that the hinge plastic rotations estimated by all three FEMA force distributions contain unacceptably large errors. the errors are largest in the lower stories. In contrast. and Table 5. with story drifts under estimated by. reaching 64%.” and NLRHA. For the SRSS distribution. 7%. but it is still inaccurate. gravity loads included 58 . Figure 5. 3 "Modes" FEMA Uniform ELF SRSS 0. larger errors are noted in the upper and lower stories. Heightwise variation of floor displacements and story drift ratios estimated using FEMA-273 force distributions. (a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th (b) Story Drift Ratios Floor 4th 3rd 2nd 1st NL−RHA MPA. at most. errors are largest in the upper stories. reaching 31%. (The 100% error in a 6th floor hinge is ignored as it simply represents that the MPA estimated zero rotation. Modal pushover analysis procedure gives estimates better than all the FEMA force distributions. with errors reaching 37% in this example.3.

and the ELF distribution fails 59 . the “uniform” distribution fails to identify yielding of the beams above the fourth floor. Based on the results presented here. story drifts.5 were determined by five analyses: MPA considering the three “modes.” nonlinear RHA (“exact”). While pushover estimates for floor displacements are more accurate. and the three FEMA analyses. and hinge plastic rotations estimated using FEMA-273 force distributions and MPA (including three “modes”). 5.(a) Floor Displacements 9th 8th 7th 6th MPA FEMA Uniform ELF SRSS 9th 8th 7th 6th (b) Story Drift Ratios MPA FEMA Uniform ELF SRSS Floor 4th 3rd 2nd 1st Floor −40 −20 0 Error(%) 9th 8th 7th 6th 20 40 60 5th 5th 4th 3rd 2nd 1st Ground −60 Ground −80 −60 −40 −20 0 20 Error (%) 40 60 80 (c) Hinge Plastic Rotations MPA FEMA Uniform ELF SRSS Floor 5th 4th 3rd 2nd 1st Ground −200 −150 −100 −50 0 50 Error (%) 100 150 200 Fig. Errors in floor displacements. The locations of plastic hinges shown in Fig. it appears that structural performance evaluation should be based on story drifts that are known to be closely related to damage and can be estimated to a higher degree of accuracy by pushover analyses.4. they are not good indicators of damage. 5. the SRSS distribution fails to identify yielding of beams in the middle floors. The locations of the plastic hinges are not identified correctly by the FEMA force distributions. gravity loads included The structural engineering profession is now comparing these hinge plastic rotations against rotation limits established in FEMA-273 to judge structural component performance.

and the error envelope for each case.” The MPA procedure provides estimates of earthquake demands that are significantly more accurate than all FEMA273 analyses.6 and 5. The MPA procedure identifies yielding in most locations predicted by “exact” analysis. The MPA procedure is superior to FEMA-273 analyses over the entire range of ground motion intensities considered. Shown is the error in each floor displacement and each story drift. especially in estimating story drifts. Included for comparison is the error in MPA with three “modes. but fails to predict yielding in a few locations. 60 . Figures 5.to identify yielding in some locations.7 summarize the error in FEMA analyses relative to the “exact” demands as a function of ground motion intensity indicated by a ground motion multiplier.

310 1.4 16.78E-03 1.490 0.462 1.4 -55.2 Peak values of story drift ratios (as % of story height) from FEMA force distributions and MPA.3 23.1 -3.2 -4. Peak values of floor displacements (as % of building height = 37.854 0.8 -100.7 19.5 10.399 -27.8 -63.10E-02 7.23E-03 3.1 0.611 0.4 MPA -2.7 -60.4 -23.2 -100.399 1.6 -73.5 26.00E+00 4.789 0.414 1.067 0.372 1.487 0.2 5.547 -27.09E-03 4.530 1.7 FEMA ELF SRSS -22.7 -29.59E-03 5.4 -12.109 1.221 1.953 0.207 1.3 Peak values of hinge plastic rotations (radians) from FEMA force distributions and MPA Hinge Plastic Rotation (rad) MPA FEMA ELF SRSS 4.3 -100.530 1.178 1.00E+00 8.2 9.783 1.7 -28.9 -20.11E-03 9.65E-03 7.84 1.5 -3.Table 5.927 1.2 -12.860 1.7 Table 5.34E-03 2.061 1.26E-03 3.00E+00 0.00E-03 5.007 1.0 37.7 21.00E+00 2.011 1.083 1.436 Error (%) NL Uniform RHA 0.417 1. gravity loads included Displacement /Height (%) FEMA ELF SRSS MPA 0.75E-03 0.0 -100.8 -100.93E-03 1.2 19.8 -35.35E-03 8.0 -100.353 1.667 0.270 0.1 10.51E-03 4.3 29.4 21.9 28.209 0.736 0.6 -41.998 1.94E-03 2.2 9.3 -22.730 1.341 1.0 -100.19E-03 1.686 0.6 -6.128 1.9 16.7 26.45E-03 3.5 22.23E-02 1.098 1.04E-02 8.14 m) from FEMA force distributions and MPA.9 SRSS -22.323 1.836 0.00E+00 0.262 1.50E-03 0.5 61 .355 0.877 1.597 0.2 6.13E-03 5.2 16.62E-03 0.0 -57.351 0.3 -14.0 -63.52E-04 1.00E+00 0.294 1.0 Table 5.1 163.306 1.00E+00 0.4 16.35E-10 3.998 1.195 0.19E-04 5.515 FEMA Uniform ELF 27.8 16.3 31.4 0.214 1.234 1.1.5 -32.00E+00 0.00E-10 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 1.6 21.5 15.58E-04 6.6 -11.0 -59.9 0.560 1.330 1.5 Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 2.2 4.830 2.5 -29.9 11.992 1.809 0.839 0.566 1.913 0.6 -4.938 1.8 17.673 Error (%) NL RHA 1.3 -4.7 7.6 4.55E-03 3.1 -26.8 2.2 16.17E-03 9.064 1.9 15.154 1.623 1.858 1.524 0.6 -25.0 -30.3 -11.672 1.5 -33.984 1.8 44.5 -27.0 -100.9 -77.6 22.335 2.4 -50.88E-03 1.5 -33.7 -15.8 -32.8 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 0.466 0.562 1.888 0.6 -17.5 -29.0 -71.2 13.344 0.6 17.53E-02 1.8 7.8 -2.263 0.6 14.9 16.00E+00 NL RHA 1.78E-03 0.399 1.9 -70. gravity loads included Displacement /Height (%) FEMA ELF SRSS MPA 1.367 1.708 0.015 0.168 1.2 10.314 1.7 MPA -2.089 1.00E+00 0.199 27.875 0.724 0.16E-03 0.975 1.318 2.03E-03 5.00E+00 Error (%) FEMA Uniform ELF SRSS MPA 24.

MPA including three “modes” and NL-RHA for 1. 5. gravity loads included 62 .5 × El Centro ground motion.5. Locations of plastic hinges determined from three force distributions in FEMA-273.(a) FEMA: Uniform • • • • • • • • •• •• •• •• •• •• (b) FEMA: ELF • •• •• •• •• •• •• • •• • •• •• •• •• •• • • • • • •• • • • • • • • •• •• •• •• • •• •• •• •• • •• •• •• •• • • •• •• •• • • • • • •• •• (c) FEMA: SRSS • • •• •• •• •• • • •• •• •• • • •• •• •• • • • • •• •• •• • • • • • • • • • (d) MPA with 3 "Modes" • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• •• •• (e) Nonlinear RHA • • • • • • • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • • • • • •• •• •• Fig.

5 3 Fig. 5. Errors in floor displacements from three force distributions in FEMA-273 and from MPA including three “modes”. gravity loads included 63 .5 2 GM Multiplier 2.5 1 1.5 3 (c) FEMA: SRSS 60 50 40 60 50 40 2 3 1 5 4 (d) MPA (3 "Modes") Error (%) 30 20 10 0 0 8 Error (%) 30 20 6 8 1 5 2 4 3 9 7 6 7 9 10 0 0 0.6.5 1 1.5 1 1. Noted 2 Error (%) Error (%) 1 2 30 20 10 0 0 1 4 6 7 30 20 10 8 3 5 8 4 9 7 9 6 5 0.5 2 GM Multiplier 2.5 2 GM Multiplier 2.5 2 GM Multiplier 2.5 1 1.(a) FEMA: Uniform 60 50 40 60 50 40 3 (b) FEMA: ELF Error Envelope Error for Floor No.5 3 0.5 3 0 0 0.

5 3 Fig.5 1 1.7.5 1 1.5 3 0 0 0.(a) FEMA: Uniform 100 60 50 9 8 (b) FEMA: ELF Error Envelope Error for Story No.5 1 1. Noted 80 40 9 1 Error (%) 7 Error (%) 60 30 20 10 2 5 6 4 8 40 1 6 20 2 5 3 4 3 7 0 0 0. Error in story drifts from three force distributions in FEMA-273 and from MPA including three “modes”.5 2 GM Multiplier 2.5 3 (c) FEMA: SRSS 60 50 40 60 50 40 3 2 1 (d) MPA (3 "Modes") Error (%) 30 20 Error (%) 30 5 4 8 9 20 9 6 7 7 10 0 0 4 10 2 3 1 8 5 6 0.5 2 GM Multiplier 2. 5.5 2 GM Multiplier 2.5 1 1.5 2 GM Multiplier 2. gravity loads included 64 .5 3 0 0 0.

6 Conclusions This investigation was aimed toward developing an improved pushover analysis procedure based on structural dynamics theory. now common in structural engineering practice. etc. 3. For this system. hinge plastic rotation. This MPA procedure for elastic buildings is shown to be equivalent to the standard response spectrum analysis (RSA) procedure. the energy dissipated in yielding or the cumulative rotation of a plastic hinge. joint rotations. pushover analysis is inherently limited in the sense that it cannot provide any cumulative measure of response. However. e. ζ n —of the nth-mode of the MDF system. where m is the mass matrix n of the building and φ n its nth mode. Pushover analysis of a one-story inelastic system predicts perfectly peak seismic demands: deformation.. 2. SRSS) leads to the modal pushover analysis (MPA) procedure. This system has vibration properties—natural frequency w n and damping ratio. Dn is available from the elastic response (or design) spectrum. 65 .. where the nature and magnitude of errors arising from approximate modal combination rules are well understood. The peak response of an elastic multistory building due to its nth vibration mode can be exactly determined by static analysis of the structure subjected to lateral forces distributed over the building height according to s* = mφ n .g. Combining these peak modal responses by an appropriate modal combination rule (e. and the structure is pushed to the roof displacement determined from the peak deformation Dn of the nth-mode elastic SDF system. which retains the conceptual simplicity and computational attractiveness of the procedure with invariant force distribution. It has led to the following conclusions: 1.g.

g.n t is determined by pushover analysis.n t = .4. n where φ n is now the nth natural mode for small-amplitude linear vibration. and (ii) the total response ro is determined by combining the rno (n = 1. p eff . the SRSS rule). with vibration properties in the linear range same as those of the nth-mode elastic SDF system. and (2) superposing responses of the inelastic MDF system to individual terms. 2. 5. p eff (t ) = . The response value rno is determined by implementing the following steps: (i) Develop the base-shearroof-displacement Vbn . The MPA procedure developed to estimate the seismic demands for inelastic buildings consists of two phases: (i) the peak response rno of the inelastic MDF system to effective earthquake forces p eff .urn bg bg bg bg b g curve from a pushover analysis of the structure for the force distribution s* = mφ n . To enable systematic extension of the MPA procedure to inelastic systems. (ii) Idealize the pushover curve as a bilinear curve and convert it to the bilinear force-deformation relation for the nth-“mode” inelastic SDF system.sn ug t —n denotes mode number—in the modal expansion of the effective earthquake forces.mι ug t . 6. 66 . …) according to an appropriate combination rule (e. (iii) Compute the peak deformation Dn of this system with unit mass by nonlinear response history analysis or from the inelastic response (or design) spectrum. an uncoupled modal response history analysis (UMRHA) is developed by (1) neglecting the coupling among modal coordinates arising from yielding of the system.. These underlying assumptions and approximations are evaluated and the errors in the UMRHA procedure relative to the “exact” nonlinear response history analysis are documented.

and identifying locations of most plastic hinges. 67 . MPA. Comparing the earthquake-induced demands for the selected 9-story building determined by pushover analysis using three force distributions in FEMA-273. This implies that MPA is able to estimate the response of buildings responding well into the inelastic range to a similar degree of accuracy when compared to standard RSA for estimating the peak response of elastic systems. The MPA procedure is more accurate than all the FEMA force distributions in estimating floor displacements. all pushover analysis procedures do not seem to compute with acceptable accuracy local response quantities.(iv) At the roof displacement determined from Dn . and hinge plastic rotations.0. it fails to compute with acceptable accuracy plastic rotations of the hinges. 8. Based on results presented for El Centro ground motion scaled by factors varying from 0. the errors in the MPA procedure are shown to be only weakly dependent on ground motion intensity.25 to 3. and nonlinear RHA. joint rotations. Comparing the peak inelastic response of a 9-story SAC building determined by the approximate MPA procedure—including only the first two or three rno terms—with rigorous nonlinear RHA demonstrates that the approximate procedure while providing good estimates of floor displacements and story drifts. Pushover analyses seem to be inherently limited in computing accurately hinge plastic rotations. 7. plastic hinge rotations. story drifts. story drifts. 9.” 10. such as hinge plastic rotations. The initial state—forces and deformations—of the structure can be considered in the MPA procedure by including these gravity load effects in pushover analysis of the structure only for the first “mode. However. demonstrates that the FEMA force distributions greatly underestimate the story drift demands. the pushover analysis provides the peak value rno of any response quantity: floor displacements. and lead to unacceptably large errors in the hinge plastic rotations. Thus the MPA procedure is accurate enough for practical application in building evaluation and design. etc.

This new method for estimating seismic demands at low performance levels. with the excitation scaled to cover a wide range of ground motion intensities and building response. structural performance evaluation should be based on story drifts that are known to be closely related to damage and can be estimated to a higher degree of accuracy by pushover analyses. should obviously be evaluated for a wide range of buildings and ground motion ensembles. Instead. such as life safety and collapse prevention.11. The present trend in the structural engineering profession of comparing computed hinge plastic rotations against rotation limits established in FEMA-273 to judge structural performance does not seem prudent. This report has focused on development of the MPA procedure and its initial evaluation in estimating the seismic demands on a building imposed by a selected ground motion. While pushover estimates for floor displacements are even more accurate. they are not good indicators of damage. 68 .

Struct. Seismic demands for performance evaluation of steel moment resisting frame structures (SAC Task 5. and Powell.. Engrg.. Earthq. 20(4-6):452-464.M. Gupta. H. Berkeley.. and Gupta. Japan. (2000). A.K. and Reinhorn. John A. A. R. M. Nat. Prentice Hall: New Jersey. Y. S.7 References Allahabadi. Report No. I: Equivalent nonlinear system. S.. (1988). B. Struct. Dyn. Washington.. Earthq. Bracci.K. H. 123:256-265.K.D. (2001). Report No. Calif. A. Pushover analysis procedure in earthquake engineering. S. 15:417-434. and Krawinkler. ASCE 123(1):3-10. E. Tokyo-Kyoto. J. G. Struct. ASCE. and Kunnath. (1997). Engrg.W.. Seismic performance and retrofit evaluation for reinforced concrete structures. 29:1287-1305. and Seneviratna.4. Stanford University.M. A. and Fischinger. and Krawinkler. (2000). Earthq. University of California. (1998). FEMA-273. Gupta. Earthq. Chopra. N2—a method for nonlinear seismic analysis of regular structures. Stanford. Krawinkler. P. Seattle. Pros and cons of a pushover analysis of seismic performance evaluation. Engrg. H. J.K.C.. Gupta. and D’Amore. NEHRP Guidelines for the Seismic Rehabilitation of Buildings.H. J. Fajfar. 6th U. Conf. Earthquake Engineering Research Center. (1997). Building Seismic Safety Council (1997). Engrg. 16(2):367-392 Han. J. Kim. Federal Emergency Management Agency. Calif. Story drift demands for steel moment frame structures in different seismic regions. G.3). Adaptive spectra-based pushover procedure for seismic evaluation of structures. 132. D.K. Spectra. A. and Wen. 69 . H. 5:111-116. (1988). Kunnath. Engng.. S. Proc. Method of reliability-based seismic design. Engrg. DRAIN-2DX user guide. (1999). (1998).P.. Estimation of seismic drift demands for frame structures. Washington.S. 9th World Conf. UCB/EERC-88/06. Proc. Spectra. (1999). Struc. Dynamics of Structures: Theory and Applications to Earthquake Engineering. on Earthq. Blume Earthquake Engineering Center.. Krawinkler.

Y..K. Saiidi. (1996). Villaverde... Earthquake member deformation demands in reinforced concrete frame structures. Paper No. T. Washington. 6th U. Christenson. E. 11th World Conf. Paret. K. Multimode pushover procedure (MMP)—A method to identify the effects of higher modes in a pushover analysis. Simple nonlinear seismic analysis of R/C structures. (1998). ASCE.A.K. Freeman. Earthq. Mexico... Jr. H. Indiana. 15(4):765-789. Calif. Validity of deformation demand estimates using nonlinear static procedures. T. and Gupta. Otani. T. Simplified response spectrum seismic analysis of nonlinear structures. Reliability of nonlinear static methods for the seismic performance prediction of steel frame buildings. Auckland. T. http://www. Common pitfalls in pushover analysis. Maison. U. B. Proc. 966. when and how?. (2000).F.S. of Calif. pp.. 1972.S.. on Earthq. Eilbekc.S. Engrg. 12th World Conf.nd. and Bonowitz. Vance. Methodology for R/C Bldg. Seismic evaluation and upgrading of existing buildings. Hokkaido. S.J.A. S. Earthq.J. D. and Dyke. Proc. SEAOC 1998 Convention. Engrg. Lawson.Kunnath.. of Civil Engrg..C. Earthq.F. (1996). Dept.M. M. R. D... K. V. G. and Lobo. Sasaki. Conf. Div. and Paret. 107(ST5):937-951. R..H. Proc.S. Benchmark Control Problems for Seismically Excited Nonlinear Buildings. Berkeley. Spencer. 5th U. Structures. S.K. ASCE. Naiem. Skokan.S. Shiohara. M. H. Sasaki. Ph. Struct. (1998). (2000). Engrg. and Hart.-Japan Workshop on Performance-Based Earthq.. Japan. 122:282-285. Engrg. Proc. Hawaii Miranda.. Seattle. Univ. and Kabeyasawa. M. (1991). and Krawinkler. Structural Engineers of California. Proc. (1999). (1981). 70 . J. B. Ohtori.. Maui. Acapulco. Proc. S. R. (1994).E. Nat. Matsumori. Proc. J.edu/~quake/. Approximate inelastic procedures to identify failure mechanisms from higher mode effects. Mech. Sapporo.. 1:283-292.. 79-94. Engrg... (2000).. S. Earthq. and Freeman.. Japan Workshop on Performance-Based Engineering for Reinforced Concrete Building Structures. R. Engrg.. Spectra. and Sozen.A. Conf. Nonlinear static pushover analysis—why. How safe are pre-Northridge WSMFs? A case study of the SAC Los Angeles Nine-Story Building. Dissertation. (1999). B. Notre Dame University. D. U.F.. Paper No. New Zealand. F..

The structure should be pushed just beyond the target (or expected) roof displacement in the selected mode. and modes. 1988). Apn ..20): s* = mφn n n 2. 71 . of the bilinear curve at the target roof displacement. Since the target roof displacement may not be known at the start of the procedure.1 incrementally and record the base shears and associated roof displacements. e.Appendix A Uncoupled Modal Response History Analysis A.1.urn ) pushover curve for the force distribution s* : n 2.g.1. using any numerical integration method. 1977). Calculate the area under the actual pushover curve. (3.2. iterations may be necessary. ω n . This step can be conveniently implemented in any commercially available software.2. DRAIN2DX (Allahabadi and Powell. 1. Define the force distribution s* from Eq.1) using the FEMA-273 procedure (Building Seismic Safety Council. A.1 STEP-BY-STEP PROCEDURE A detailed step-by-step implementation of the uncoupled modal response history analysis (UMRHA) procedure is presented in this section and illustrated by an example in the next section. φn . 3. For the nth-“mode”. 3. B. Idealize the pushover curve as a bilinear curve (Fig. 2. Define the anchor point.g. respectively. e. Apply force distribution of Step 2.. trapezoidal rule. Let the roof displacement and base shear at the anchor point be urno and Vbno . for linear-elastic vibration of the building. develop the base-shear – roof-displacement (Vbn . 3. Compute natural frequencies.

9. and a point on the actual pushover curve with base shear equal to i 0.i 3.5.6 . This value.2. Abn . 3.9. Vbny .3. Calculate area under the bilinear curve OAB. Calculate initial slope of the idealized bilinear curve. corresponding to the estimated yield i i i base shear.1.0. 3.1. i 3.6 . i +1 i i 3. Vbny . A.7. i i 3.4.2. 3. and B with straight-line segments to obtain the idealized bilinear curve. determine the roof displacement. Calculate Vbny = Vbny × Apn Abn . kn = 0. urny = Vbny kn . will be refined by an iterative procedure that seeks to equate areas under the actual and the idealized pushover curves. i 3. Calculate the slope.4.6 × Vbny . Replace i+1 with i and repeat Steps 3. by connecting a straight line between origin.6 × Vbny .9.4 to 3. Calculate the yield displacement. Calculate the error = 100 × Abn − Apn ( ) ( ) ( ) ) Apn . 72 . at base shear i equal to 0. Let the point with base shear = Vbny and roof displacement = urny be denoted as A. O.6.8.4. From the pushover data. Calculate the post-yielding strain-hardening ratio. kn . If the error exceeds some pre-specified tolerance. obtained by judgment.8. urn. If desired.6 × Vbny ( ) i urn . Estimate the yield base shear. iterations are necessary. This step gives the secant stiffness at a base shear equal to 60% of the yield base shear. Draw the curve OAB by connecting the three points O. i 3. i i i 3. i i i αn =  Vbno Vbny − 1  urno urny − 1         i 3.0. other appropriate methods can be ( used.

4. (3.2.16). 5. and pseudo-acceleration history.12) and (3.10b and 4. 4. Scale the vertical axis with * Mn to obtain * Fsno Ln = Vbno M n and * Fsny Ln = Vbny M n (Eqs. Compute deformation history. Calculate peak values. Compute the Ln and Γ n from Eq.0. Repeat Steps 3 to 6 for as many modes as required for sufficient accuracy. 4.2). Scale the horizontal axis by Γnφrn to obtain Dno = urno Γ nφrn and Dny = urny Γ n φrn (Eqs. A. 7.15) and (3.4. A. A. 9. * 4.2. Develop the Fsn Ln − Dn relation (Fig.4) and effective modal mass from M n = Ln Γn .13). Combine the “modal” responses using Eqs. Dn (t ) . of the nth“mode” inelastic SDF system (Fig. (3. In general first two or three modes will suffice.eps ur n. 4. Calculate histories of various responses using Eqs. of the combined responses obtained in Step 8. (3. Vbn Idealization of Pushover Curve Vbno Vbny 0.3.6 ur n y ur n o ur n Fig.1 Idealization of nth-“mode” pushover curve 73 .11b).3b) with unit-mass and force-deformation relation of Fig.10a and 4. 8.6 × Vbny A • Idealized αnkn B • 1 Actual k 1 • O n figA_1. 6. 4. An (t ) . r o .1.11a).

The anchor point.5. At this point. B.7 cm (5 in.Fsn / Ln Fsn / Ln − Dn Relationship Vbno / M* n Vbny / Mn * 1 α ω2 n n ω2 1 n figA_2. Area under the actual pushover curve. Idealized bilinear curves for each of the three modes are included in Fig. second.5 cm (25 in. ur1o = 63.3. and 12. First three mode shapes and frequencies of the selected building were computed and are shown in Fig.). The following steps illustrate the procedure to develop the idealize curve for the first “mode”. and third mode. The base-shear – roof-displacement (Vbn . s* . generated using DRAIN-2DX. 3.1.2. Target displacements used to generate these pushover curves are 63.1.urn ) pushover curve for the force distribution s* : n 2. 74 .). A. respectively.3. is defined at the target roof displacement. Following is step-by-step implementation of the procedure described in Section A.5 cm and Vb1o = 8729. for the first.eps Dny = ur n y / Γnφr n Dno = ur n o / Γnφr n Dn Fig.4. 25.1. to the north-south component of the El Centro (1940) ground motion scaled up by a factor of 1. The pushover curves for the first three modes.4.4 cm (10 in.). are shown in Fig. 2. The force distributions.2 Properties of the nth-“mode” inelastic SDF system A. 3. 1.2. 3.6 kN. 3. Ap1 = 360777 kN-cm. n 2. A. A.3. 3.2 EXAMPLE The UMRHA procedure is implemented to calculate the response of the 9-story building described in Section 3. computed for the first three modes are shown in Fig.

iterations are necessary.2. and M1 = 2736789 × 1. 3.135.2. i +1 3. Determined from the pushover database.7.86 cm at 0. i 3.1.18 kN/cm.18 = 38.198%.4 × (360777 365100 ) = 7911.0.09 cm and Vb1 y = 8006. The results for other modes are summarized in Table A. ur1 y = Vb1 y k1 = 8006.6 8006.4 210.23 cm. * 4.6.4 ) − 1 (63.3.6 × Vb1 y = 4803. (4.4. Area under the bilinear curve OAB.86 = 210.8 22. The procedure converged after nineteen cycles to give ur1 y = 36. i i i 3. Also included are the modal damping ratios and the periods calculated from Eq. L1 = 2736789 kg. 4. ur1. The post-yielding strain-hardening ratio.0.9.9.2. i i 3. 75 .8 kN.5.13). The Fs1 L1 − D1 relation for the first “mode” is developed as follows. The first estimate of the yield base shear Vb1 y = 8006.01%. The yield displacement.4 kN. Vb1 y = 7615.1. This value exceeds the prespecified tolerance of 0. and α1 = 0. k1 .1. The results of the iterative procedure are summarized in Table A. 3. i i 3.     i 3. i i i 3. The initial slope of the idealized bilinear curve.09 cm.3666. and B with straight-line segments gives the idealized bilinear curve.4. Error = 100 × (365100 − 360777 ) 360777 = 1.1.9 kN. Γ1 = 1.194.6 = 4803.8.09 ) − 1 = 0. The curve OAB obtained by connecting the three points O.6 × Vb1 y ( ) i ur1. The next estimate of the yield shear is Vbny = 8006. α1 =  Vb1o Vb1 y − 1  ur1o ur1 y − 1 =         ( ) ( ) (8729.5 38.4. Therefore.6 = 22. A.9. Ab1 = 365100 kN-cm. The point A on the i i bilinear curve is defined by ur1 y = 38.i 3. 3. k1 = 0. is calculated as follows.4 kN.6 kN.3666 = 3740189 kg.

7. The peak values are computed and are summarized in Tables 4. 5. Scaling the vertical axis by Fs1 y L1 = 203.2. 4.46 cm and D1y = 26.40 (cm/sec2) and 76 . The combined modal responses are presented in Fig. 4. Scaling the horizontal axis by Γ1φr1 gives D1o = 46.4. A.4. 4. 4. 4. The results were generated for first three “modes’ and are included in Fig.2.7.8.62 (cm/sec2). 8. Histories of roof displacement and top story drifts for the first “mode” are computed and presented in Fig. 7.3. 9. The peak values are also plotted in Fig. Deformation and pseudo-acceleration histories of the inelastic SDF systems for the first “mode” with unit mass and force-deformation relation developed in Step 4 are plotted in Fig.1 and 4. 6. * M1 gives Fs1o L1 = 233.51 cm.7.

6 cm.14 y by 6000 Base Shear (kN) 4000 2000 Actual Idealized figA_3c.2 cm.eps 10 20 30 40 50 Roof Displacement (cm) (b) "Mode" 2 Pushover Curve 60 8000 u = 9. A. “Modal” pushover curves for the example building 77 .19 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized figA_3a.13 6000 Base Shear (kN) 4000 2000 Actual Idealized figA_3b.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36. V = 5210 kN.eps 0 0 2 4 6 8 10 Roof Displacement (cm) 12 Fig. V = 7616 kN. α = 0.3. V y by = 4952 kN.eps 0 0 5 10 15 20 Roof Displacement (cm) (c) "Mode" 3 Pushover Curve 25 8000 u = 4. α = 0. α = 0.9 cm.

” second “mode.5 × El Centro ground motion for the first “mode.33 A1 (g) n=2 0 −80 80 −3 3 n=2 D2 (cm) 0 A1 (g) • 22.(a) Deformation 80 n=1 3 (b) Pseuso−acceleration n=1 • 0. Histories of deformation and pseudo-acceleration due to 1. 78 .52 5 10 15 20 Time (sec) 25 30 0 • 1.” and third “mode” inelastic SDF systems.747 5 10 15 20 Time (sec) figa_4.223 −80 80 n=3 −3 3 n=3 D3 (cm) 0 A3 (g) • 10.06 0 • 1.2769 D1 (cm) 0 • 35. A.m −80 0 −3 0 25 30 Fig.4.

0 4704.18 210.51 2 “Mode” 2 -920860 -0.529 0.029 0.2671 1.18 210.3 4603.7 7639.Table A.188 0.2 7690.11 22.010 (kN) 8006.3 7786.190 0.32 36.18 210.35 36.23 22.4927 1.18 210.4 7672.5 3109.170 0.1.193 0.194 0.237 0.50 36.180 0.0 7619.18 210.59 36.3 7658.1 4574.39 36.76 21.44 36.2.25 36.198 0.151 0.948 ζ n (%) 79 .082 0.09 18.56 47.184 0.18 210.29 36.78 21.8525 1.8 7618.23 0.05 52.0 4588.162 0.4 7647.9 4570.65 1226.8 4747.30 37.5 (cm) 22. No.7 4580.1 4569.194 (kN-cm) 365100 364060 363276 362684 362235 361892 361631 361432 361279 361162 361073 361004 360951 360911 360880 360856 360838 360824 360813 Table A.18 210.4 4595.048 0.5 7633.3666 3740189 203.135 0.9 4573.18 210.194 0.64 37.3 7745.18 210.136 Fsny Ln (cm/sec2) Dny (cm) Fsno Ln (cm/sec ) Dno (cm) Tn (sec) 233.18 210.6 × Vb1 y i ur1.6 4583.037 0.063 0.9 7615.77 21.18 210.18 210.95 21.2 4628.186 0.18 (cm) 38.193 0.182 0.46 2.5 4614.6 7840.1 1013.38 22.0 4577.2 4671.05 36.86 22. i 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 i Vb1 y i 0.70 36.017 0.75 21. Properties of “modal” inelastic SDF systems Properties Ln (kg) Γn * M n (kg) “Mode” 1 2736789 1.5 7624.013 0.02 21.404 0.910 0.9 (kN) 4803.309 0.85 36.693 0.75 21.83 21.18 210.4 7714.59 22.12 3876.18 210.3 7628.6 i k1 i ur1 y i α1 i Ab1 Error (%) 1.9 4570.18 210. Results of iterative procedure to develop the idealized bilinear curve for the first “mode” inelastic SDF system Itr.193 0.86 21.191 0.09 37.192 0.18 210.18 210.022 0.176 0.26 36.90 21.74 (kN/cm) 210.79 0.8 4647.103 “Mode” 3 696400 0.8 7622.85 0.74 21.107 0.194 0.24 36.1 7616.4 7911.2406 167531.81 21.139 0.56 19.79 21.18 210.5309 488839.28 36.40 46.62 26.0.2 4571.25 36.

80 .

1 to the north-south component of the El Centro (1940) ground motion scaled up by a factor of 1.3b) with unit mass and force-deformation relation of Fig.8) for the peak deformation of the first-“mode” inelastic SDF system with unit mass and force-deformation relation developed in Step 4 gives D1 = 35. 13. extract from the pushover database values of other desired responses. rno .33 cm. 4. or from the inelastic response (or design) spectrum. Solving Eq. 4. subtract the yield hinge rotation to determine the plastic hinge rotation.21).18). 12. 6. Following is step-by-step implementation of the procedure described in Section B. (3.2 EXAMPLE The MPA procedure is implemented to calculate the response of the 9-story building described in Section 3. (3.Appendix B Modal Pushover Analysis B. This is the same example as solved in Appendix A.4. From the total hinge rotation.1.8). Compute the peak deformation. Repeat Steps 3 to 7 for as many “modes” as required for sufficient accuracy. Calculate the peak roof displacement urno associated with the nth-“mode” inelastic SDF system from Eq. B.5. (4. Typically. the first two or three “modes” will suffice.1 STEP-BY-STEP PROCEDURE A detailed step-by-step implementation of the modal pushover analysis (MPA) procedure is presented in this section and illustrated by an example in the next section.6b by solving Eq. (4. 10. 11. Dn . of the nth-“mode” inelastic SDF system (Fig. 10. Steps 1 to 4 of the MPA are the same as those for UMRHA presented in Appendix A. Determine the total response by combining the peak “modal” responses using the SRSS combination rule of Eq. 81 . At urno .

Results of Steps 5 and 6 are summarized in Table B.11. 8.33 = 48.4. 82 . 12.4 for the story drifts normalized by the story height.3 for the floor displacements normalized by the building height and in Table 4.3 and 4.5 are the plastic hinge rotations computed by subtracting the yield hinge rotation from the total hinge rotations.” and the results are included in Tables 4. values of floor displacements and story drifts are extracted.18) are also included in Tables 4.28 cm. The values are summarized in Table 4. where results for other ground motion intensities are also included.28 cm.1. (3. 9.366 × 1 × 35.3 and 4. Also included in Table 4. The total response computed by combining the peak “modal” responses using the SRSS combination rule of Eq. At ur1o = 48.4. Peak roof displacement ur1o = Γ1 × φr1 × D1 = 1. Steps 3 to 7 are repeated for first three “modes.

37 57. Calculation of roof displacements urno from peak deformation of inelastic SDF systems Ground Motion Multiplier 0.735 3.5 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.37 1.18 27.71 1.03 26.678 0.139 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.8451 5.660 14.28 46.07 “Mode” 2 4.450 4.225 2.35 1.25 20.535 14.03 0.268 0.33 1.267 5.82 1.Table B.007 36.85 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 1.457 12.551 2.332 13.5 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 2.70 0.577 16.332 48.52 “Mode” 3 1.766 7.4222 3.395 0.25 Quantity Dn (cm) “Mode” 1 6.59 0.023 0.312 1.252 9.901 8.229 8.748 63.185 11.06 1.379 21.504 18.467 14.50 35.0 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 1.969 0.38 22.0 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 3.275 1.27 0.755 27.913 22.35 0.154 78.690 10.755 0.691 0.184 0.126 13.79 0.1.367 1.117 5.200 0.436 7.513 0.05 0.13 2.676 6.75 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.0 µ = Dn Dny urno = Γnφrn Dn (cm) 83 .73 24.856 31.36 1.52 0.

84

Appendix C FEMA Force Distribution Calculations
Presented in this appendix are the calculations leading to the FEMA-273 force distributions (Fig. 5.1) used in developing the pushover curves (Fig. 5.2). These distributions were described in Section 5.1 C.1 “UNIFORM” DISTRIBUTION

The lateral force at a floor is equal to the mass at that floor, i.e., s* = m j . For convenience, the j floor forces are normalized with the base shear. The results are summarized in Table C.1.
Table C.1. FEMA273 “uniform” lateral force distribution

Floor, j

mj (10-3×kg)

s* = j

∑i mi
0.112 0.110 0.110 0.110 0.110 0.110 0.110 0.110 0.119

mj

1 2 3 4 5 6 7 8 9

503.5 494.7 494.7 494.7 494.7 494.7 494.7 494.7 534.1

C.2

EQUIVALENT LATERAL FORCE (ELF) DISTRIBUTION 85

The lateral force at a floor is computed from s* = m j h k where m j is the mass, h j is the height j j of the jth floor above the base, and the exponent k = 1 for fundamental period T1 ≤ 0.5 sec, k = 2 for fundamental period T1 > 2.5 sec; and varies linearly in between. For the selected building, T1 = 2.27 and k = 1.885. The resulting lateral forces are summarized in Table C.1. For convenience, the floor forces are normalized with the base shear.
Table C.2. FEMA273 equivalent lateral force (ELF) distribution

Floor j 1 2 3 4 5 6 7 8 9

m jhk j (10 ×kg-m ) 371.0 1015.0 1963.2 3196.8 4707.4 6488.3 8534.2 10840.6 14471.1
-3 k

s* j

=

m jhk j

∑i mihik
0.007 0.020 0.038 0.062 0.091 0.126 0.165 0.210 0.281

C.3

SRSS DISTRIBUTION

The calculation of the SRSS distribution is summarized as a series of steps as follows: 1. For the nth-mode calculate the lateral forces, f jn = Γn m jφ jn An in which j denotes the floor number and An is the pseudo-acceleration of the nth-mode SDF elastic system, leading to columns 2 to 4 of Table C.3. 2. Calculate the story shears, V jn = ∑ i = j fin where j is now the story number. Implementing this step gives columns 5 to 7 of Table C.3. 3. Combine the modal story shears using SRSS rule, V j = Table C.3. 4. Back calculate the lateral forces at the floor levels from the combined story shears V j to obtain column 9 of Table C.3. 86
N

∑ n (V jn )

2

to get column 8 of

2 fj = ∑ fi i fj (2) (3) (4) 59.7 101.065 0.3.2 285.0 1231.047 0.6 -352.8 1381.0 136.7 -46.1 Story Shears V j2 (kN) (6) 1114.090 0.1 -525.4 400.7 1578.7 525.0 354.9 832.070 0. FEMA 273 SRSS force distribution Lateral Forces Floor j (1) 1 2 3 4 5 6 7 8 9 f j1 (kN) f j2 (kN) f j3 (kN) V j1 (kN) (5) 1917.9 -166.2 105.1 -967.4 1842.For convenience.3.7 2065.7 95.3 222.7 694.7 (10) 0.0 -5.6 286.9 1683.5 215.3 -646.5 -350.177 0.0 980.1 -646.367 87 .8 -326. the lateral forces are normalized by the base shear to obtain column 10 in Table C.2 Lateral Force fj (kN) (9) 203.8 430.7 355.7 Vj (kN) (8) 2268.5 159.9 446.9 880.5 -320.2 277.1 832.098 0.7 234.4 1759.0 176.6 319.1 1857.6 -359.8 374.2 95.4 250.0 1476.1 -438.045 0.6 V j3 (kN) (7) 478.3 240.9 1446.9 366.0 381.5 -973.2 97.1 87.2 200.042 0.6 -732.9 -153.2 148.3 -6.7 1622.6 366.6 1233.5 320. Table C.7 374.

You're Reading a Free Preview

Download
scribd
/*********** DO NOT ALTER ANYTHING BELOW THIS LINE ! ************/ var s_code=s.t();if(s_code)document.write(s_code)//-->