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 .

ro . 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 .… according to an appropriate modal combination rule. and nonlinear RHA. and identifies locations of most plastic hinges. p eff t = . Comparing the earthquake-induced demands for the selected 9-story building determined by pushover analysis using three force distributions in FEMA-273. but provides superior accuracy in estimating seismic demands on buildings. in the modal expansion of the effective earthquake forces. which retains the conceptual simplicity and computational attractiveness of current procedures with invariant force distribution. it is demonstrated that the FEMA force distributions greatly underestimate the story drift demands.25 to 3. a pushover analysis determines the peak response rno of the inelastic MDF system to individual modal terms. 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. plastic hinge rotations are less accurate.sn ug t . 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. Second. The peak deformation of this SDF system is used to determine the roof displacement. the total demand. 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. MPA. at which the seismic response.mι ug t . The standard response spectrum analysis (RSA) for elastic buildings is reformulated as a Modal Pushover Analysis (MPA). The base shear-roof displacement Vbn . 2. 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.urn curve is developed from a pushover analysis for force distribution s* . The MPA procedure is extended to estimate the seismic demands for inelastic systems: First.ABSTRACT The principal objective of this investigation is to develop a pushover analysis procedure based on structural dynamics theory.n t = . such as hinge plastic rotations. This pushover curve is idealized as bilinear n and converted to the force-deformation relation for the nth-“mode” inelastic SDF system. 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. where m is the n mass matrix and φ n its nth-mode. However.0. Combining these peak modal responses by modal combination rule leads to the MPA procedure. all pushover analysis procedures considered do not seem to compute to acceptable accuracy local response quantities. rno . Instead. and the structure is pushed to the roof displacement determined from the peak deformation Dn of the nth-mode elastic SDF system. is determined by pushover analysis. iii . p eff . is determined by combining the rno n = 1. Thus the MPA procedure is accurate enough for practical application in building evaluation and design.

iv .

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

vi .

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

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

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

the seismic demands determined by pushover analysis using three force distributions in FEMA-273 are compared against the MPA and nonlinear RHA procedures. 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. The MPA procedure is then extended to inelastic buildings.. 2000. First.have also been made to consider more than the fundamental vibration mode in pushover analysis [Paret et al. 2 . Matsumori et al. the underlying assumptions and approximations are identified. 2000. Sasaki et al. and the errors in the procedure relative to a rigorous nonlinear RHA are documented. Gupta and Kunnath.. Kunnath and Gupta. 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.. 1996. Finally. but provides superior accuracy in estimating seismic demands on buildings. we show that pushover analysis of a one-story system predicts perfectly peak seismic demands. 1998.

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

and (b) bilinear hysteretic forcedeformation relation For a given excitation u g t .fS m fs c ut u fy 1 αk k 1 uy ug -fy k k um u 1 1 (a) (b) Fig. Section 7. m depends on three system parameters: w n .e. This becomes evident if Eq.4) and w n is the natural vibration frequency. and z is the damping ratio of the system vibrating within its linear elastic range (i.. (a) Idealized one-story structure. the deformation u t depends on three systems parameters: w n . sign u = .5) For a given u g t .1) is divided by m to obtain ~ u + 2zw n u + w 2 u y f s u. and R y (Chopra . z . 2001. and the ductility factor is u m= m uy (2.1.ug t n where bg bg b g bg (2. deformation is denoted by um . or absolute (without regard to algebraic sign) maximum. z y . u £ u y ). 2. Tn = 2p w n is the natural vibration period. bg 4 . in addition to the form of the force-deformation relation.3) ωn = k m ζ = c 2mω n f fs = s fy (2. The peak. (2. and u y .3).

and f y = 39. 2001) and scaled up by a factor of 2.2: the dimensions and flexural rigidity of the structural elements are noted. Vb (kN) 30 20 10 0 0 2 4 6 Roof Displacement.4 in Chopra.376 cm. 2. with Tn = 0.2 5 .2311 . one-bay frame 50 uy = 1.5 sec and z = 5% subjected to the north-south component of the El Centro (1940) ground motion (Fig.826 kips) for w = 169.9 kN (38. 6.66 m • EIb EIc EIc • • L = 7.26 kN. based on Ry = 8 .2. m h = 3.2 kips) . Vby = 39. u (cm) 8 Fig. One-story.1. The yield strength of the inelastic system. Pushover curve for structure shown in Fig.3. α = 0.04 40 Base Shear. this system and excitation.2.2 SYSTEM AND EXCITATION CONSIDERED Consider the one-story system in Fig.26 kN (8. 2. For . is f y / w = ( f o / w ) ÷ 8 = 0. f o w = 184 . 2. 2.32 m • Fig.

36 cm.3 in. the peak value determined from RHA. The yielding stiffness of each structural element is defined as 3% of its initial stiffness. e.077 × 107 mm 4 (146 in. It is organized in five parts: (a) shows the deformation u t .1 kip-in.3.) for the beam and columns. the energy dissipated in 6 .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. the joint rotation and beam hinge rotation are identical to values q m and q pm determined from RHA.6 kip-in. The peak values of the various response quantities are as follows: um = 7. (b) shows the lateral force f s t or base shear Vb t normalized relative to the bg weight.36 cm .The yield moments in the beam and columns are defined as the bending moments due to the lateral force f y .4 shows the earthquake response of the system described in the preceding section determined by response history analysis (RHA). (d) shows the rotation q p bt g of the plastic hinges at the bg bg beam ends.4f. The system is excited well beyond the yield deformation.g.0217 rad . pushover analysis cannot provide any cumulative measure of response. q m = 0. Implementing this analysis with I c = 6.35.18 kN-m (444. I b = 3.65 kN-m (191.3 RESPONSE HISTORY ANALYSIS Figure 2..134 × 107 mm 4 (75. wherein the hysteretic force-deformation history of Fig. The resulting pushover curve is shown in Fig. A static pushover analysis of this one-story system leads to the force-displacement relationship shown in Fig. as apparent in Fig. 2. 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 . Determined from pushover analysis at the exact peak deformation. 2. However. and q pm = 0. and E = 2 × 108 kPa (29×103 ksi) gives yield moments of 21. 2. 2.017 rad .) and 50. (c) shows the joint rotation q bt g . 2.4 ).4edetermined from RHAis superimposed. the ductility factor m = 5. 2.4e. and (e) shows the force-deformation relation. respectively.4 ). Observe that the pushover curve matches the initial loading path of the hysteretic system.

2311 (a) Vb / w 0 f / w = −0. (d) plastic hinge rotation.4.2311 y (b) −0.04 0. (c) joint rotation. and (f) pushover curve 7 . (b) base shear.36 −15 0. Response of one-story system to El Centro ground motion: (a) deformation.2 Vb / w 0.2 −0.4 0. 15 u (cm) 0 • 7.0217 • θ (rad) 0 (c) −0.04 0.2 0 (e) 0 (f) −0.4 0.4 −10 −5 0 u (cm) 5 10 −0.5 0.04 0. (e) force-deformation relation.4 −10 −5 0 u (cm) 5 10 Fig.5 fy / w = 0. 2.04 0 5 10 15 Time (sec) 20 25 30 0.017 • 0 (d) θp (rad) −0. or the cumulative rotation at a plastic hinge.2 −0. This represents an inherent limitation of pushover analyses.yielding during the ground motion.

8 .

c. Elastic Multistory Buildings 3.n t = Â .1) where u is the vector of N lateral floor displacements relative to the ground.3) where φ n is the nth natural vibration mode of the structure.4) 9 .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 . m. 2001: Section 13. This force distribution can be expanded as a summation of modal inertia force distributions s n (Chopra.mι ug t bg bg bg (3.m ι u g t bg bg (3. and lateral stiffness matrices of the system. (3.3.12): mι = n =1 Â sn = N n =1 Â Γ nmφ n N (3. and p eff t = Â peff.snug t n= n =1 The effective earthquake forces can then be expressed as af N af N af (3. and k are the mass.1) can be interpreted as effective earthquake forces: p eff t = . The right side of Eq.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 = . classical damping. each element of the influence vector ι is equal to unity.

n (t ) = N n =1 ∑ −sn u g (t ) N (3. subjected to u g t : Dn + 2z nw n Dn + w 2 Dn = .n t = .9) and (3.10) gives qn t = Gn Dn t bg bg bg (3.10) bg bg bg 10 (3. The equations governing the response of the system are mu + cu + ku = .sn ug t bg bg bg bg (3.8) gives the floor displacements u n t = Gnφ n Dn t 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. (3.6) Next.12) .sn ug t bg (3. we will outline that the response of the MDF system to p eff . Then the floor displacements are u n t = φ n qn t bg bg bg bg (3.peff (t ) = n =1 ∑ peff .9) in which w n is the natural vibration frequency and z n is the damping ratio for the nth mode. The solution qn t can readily be obtained by comparing Eq.11) and substituting in Eq. with no contribution from other modes.Gn ug t (3.8) where the modal coordinate qn t is governed by 2 qn + 2z nw n qn + w n qn = .5) The contribution of the nth mode to s and to p eff t are: sn = Gn mφ n respectively. p eff .u g t n Comparing Eqs. it can be demonstrated that none of the modes other than the nth mode contribute to the response.9) to the equation of motion for the nth-mode elastic SDF system. (3.7) By utilizing the orthogonality property of modes. (3.n t is entirely in the nth-mode.

1. In contrast to the classical 11 bg . internal element forces. (3. Section 13.12) and (3.1). Equations (3.can be expressed by st rn t = rn An t bg bg bg bg bg (3.9) is the standard modal equation governing qn t .13) define the contribution of the nth-mode to the response.13) represent the response of the MDF system to p eff. etc. 2001. Conceptual explanation of modal response history analysis of elastic MDF systems Any response quantity r t story drifts.16) reflect combining the response contributions of all modes. Therefore. the static value of r due to external forces sn .1.n t .14) is the pseudo-acceleration response of the nth-mode SDF system (Chopra.Forces sn An(t ) st rn ωn.16) This is the classical modal RHA procedure wherein Eq. 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. 3. (3. ζn ¨ g(t ) u (a) Static Analysis of Structure (b) Dynamic Analysis of SDF System Fig. st The two analyses that lead to rn and An t are shown schematically in Fig.15) and (3.12) and (3. (3.15) rt = bg n =1 Â rn t = bg n =1 Â rnst An bt g N (3. However.13) st where rn denotes the modal static response. these standard equations have been derived in an unconventional way. 3. and Eqs. Eqs. and An t = w 2 Dn t n bg (3.

(3. and Tn = 2p w n is the natural vibration period of the nth-mode of the MDF system.9) . 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. this response value can be obtained by static analysis of the structure subjected to lateral forces distributed over the building height according to 12 .17) where An is the ordinate A Tn . The SRSS rule. the peak nth-mode response as in Eq. z n of the pseudo-acceleration response (or design) spectrum for the nth-mode SDF system. 2001.1. This concept provides a rational basis for the modal pushover analysis procedure developed later.3). we used the modal expansion of the spatial distribution of the effective earthquake forces.17) (Chopra.derivation found in textbooks (e. 3. In such a response spectrum analysis (RSA).g. Chopra.1).16). The peak modal responses are combined according to the Square-Root-of-Sum-ofSquares (SRSS) or the Complete Quadratic Combination (CQC) rules.19) will provide the same value of rno ..(3. 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. Alternatively.3 MODAL PUSHOVER ANALYSIS (3. Sections 12. which is valid for structures with well-separated natural frequencies such as multistory buildings with symmetric plan. Section 13.18) To develop a pushover analysis procedure consistent with RSA. 2001. (3.4 and 13. we note that static analysis of the structure subjected to lateral forces f no = Γ n mφ n An (3.8.

1660 W.18) to obtain an estimate of the peak value ro of the total response. 13 . SAC is a joint venture of three non-profit organizations: The Structural Engineers Association of California (SEAOC). Typical floor-to-floor heights (for analysis purposes measured from center-of-beam to center1 2 Brandow & Johnston Associates.. The building has a basement level. denoted B1.73 m (150 ft) by 45. in both directions. 3.2. Consulting Structural Engineers. this building is 45. each determined by one pushover analysis. (3. with the ninth level being the roof. with five bays each in the north-south (N-S) and east-west (E-W) directions. A benchmark structure for the SAC project. this structure meets seismic code and represents typical medium-rise buildings designed for the Los Angeles. was designed by Brandow & Johnston Associates1 for the SAC2 Phase II Steel Project. The levels of the 9-story building are numbered with respect to the ground level (see Fig. (3. Los Angeles.4.21) 2 where Dn = An ω n . Third St. and 37. The peak modal responses. urno .2).20) and the structure is pushed to the roof displacement. SAC Steel Project Technical Office. The interior bays of the structure contain simple framing with composite floors. region. which from Eq. 3.15 m (30 ft) on center.s* = mφ n n (3. the Applied Technology Council (ATC). 3. Although not actually constructed.19 m (122 ft) in elevation. and California Universities for Research in Earthquake Engineering (CUREE). CA 90017. can be combined according to Eq. rno .12) is urno = Gnf rn Dn (3.1 System and Excitation Considered The 9-story structure. 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. The columns are 345 MPa (50 ksi) steel wideflange sections. Obviously Dn and An are available from the response (or design) spectrum. The bays are 9. the peak value of the roof displacement due to the nth-mode.73 m (150 ft) in plan.2). California. 1301 S. shown in Fig. This modal pushover analysis (MPA) for linearly elastic systems is equivalent to the well-known RSA procedure (Section 3. CA 94804-4698. 46th Street. Richmond.4 COMPARATIVE EVALUATION OF ANALYSIS PROCEDURES 3.

The seismic mass of the structure is due to various components of the structure.07×106 kg (73. The floor-to-floor height of the basement level is 3.0 kips-sec2/ft). and shear distortion of panel zones are neglected but large deformation (P∆) effects are included. The simple model adopted here is sufficient for the objectives of this study. for the second through eighth levels is 9. 3. and seventh levels at 1. s* (Eq. fifth.65 m (12 ft) and for the first floor is 5. floor slabs. The column lines employ two-tier construction. 0.01×106 kg (69. 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 9.2.49 sec.85.96 m (13 ft). dimension. for the first level is 1. are located on the first.20). mechanical/electrical.65×105 kg (66. monolithic column pieces are connected every two levels beginning with the first level.4. The floor system is composed of 248 MPa (36 ksi) steel wide-flange beams in acting composite action with the floor slab.3.00×106 kg (616 kips. third. These force n distributions will be used in the pushover analysis to be presented later. To ensure that this structure remains elastic.e.49 m (18 ft). such as those described in Gupta and Krawinkler (1999) can be used. the vibration periods are 2. 1988) using the M1 model developed by Krawinkler and Gupta (1998). This model is based on centerline dimensions of the bare frame in which beams and columns extend from centerline to centerline. ceiling/flooring.2 kips-sec2/ft). The column bases are modeled as pinned and secured to the ground (at the B-1 level). The 9-story N-S MRF is depicted in Fig. 3. and 0. The force distributions.sec2/ft). 14 .. Concrete foundation walls and surrounding soil are assumed to restrain the structure at the ground level from horizontal displacement. for the first three modes are shown in Fig. The first three vibration modes and periods of the building for linearly elastic vibration are shown in Fig. we select a weak ground motion: the northsouth component of the El Centro (1940) ground motion scaled down by a factor of 4. Column splices. and for the ninth level is 1. 3.0 kips-sec2/ft). partitions. The building is modeled in DRAIN-2DX (Allahabadi and Powell. if desired more complex models. respectively.27. i.of-beam) are 3. The seismic mass of the ground level is 9.89×105 kg (67.7 kips-sec2/ft). roofing and a penthouse located on the roof. which are seismic (tension) splices to carry bending and uplift forces.83 m (6 ft) above the center-line of the beam to column joint. 3. The strength. including the steel framing.

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

5. and 3 .5 0 0. First three natural-vibration periods and modes of the 9story building 3.5 Fig. 2.67 −1.39 3. 3. (c) shows the joint rotation q n t of an external joint at 16 bg bg bg .1 −2.27 sec Ground −1.04 1. 3. respectively. 2.0272 −2.93 −1.72 −2.31 −0.4.85 sec 2 Floor 5th 4th 3rd 2nd 1st T1 = 2.12) and (3. n = 1.2 Response History Analysis The structural response due to individual vibration modes.44 1.94 2. and 3. (3.796 0. 3. is shown in Figs. determined by RHA [Eqs. n = 1.61 2.49 sec 3 T = 0.4. and 3 n 3. Force distributions s* = mφn .37 2.12 0.33 2.13)].8 −2.6.5 Mode Shape Component 1 1. Each figure is organized in four parts: (a) shows the roof displacement urn t . (b) shows the base shear Vbn t normalized relative to the weight W of the building.5 −1 −0.38 0.51 0.487 −1.3.75 1.1 3.05 2.13 −1.7. 3.728 2.05 s * 1 s * 2 s * 3 Fig.05 1.03 −1.9th 8th 7th 6th T = 0.

and 3.2. ur 2o = 2. and rotations of external joints with moment connections are presented in Tables 3. Combining the modal response histories for all modes gives the total response [Eqs. with the first three modes provide essentially the exact response. but three modes—perhaps even two modes—are sufficient.3. two.8. in particular. consistent with the increasing significance of the higher mode response among these three sets of response quantities. 17 . The peak values of displacements of all floors. The peak values of floor displacements and story drifts determined by RHA. and ur 3o = 0.1. 3.3. especially in estimating the story drifts. and even larger in joint rotations. The same method was used to determine the peak values of many response quantities. Also included are the combined response due to one.12 cm. 3. or all modes. the exact response considering all modes. errors generally decrease as response contributions of more modes are included. As expected. respectively. 3. errors are smallest in floor displacements. drifts in all stories.16)]. including one. and the percentage errors due to truncation of higher modal contributions. the results for the roof displacement and top-story drift are shown in Fig.422 cm.2. This is illustrated in Fig. The errors due to truncating the contributions of vibration modes beyond the third mode are negligible. are presented in Fig. and 3. larger in story drifts. which are listed in Tables 3. and three vibration modes.urn relation. For a fixed number of modes included. It is apparent that the first mode alone is inadequate.9. 3. two.1.8. (3. respectively.the roof level. three.23 cm. where the second and third modal responses are a larger percentage of the top story drift compared to roof displacement. The peak values of the various response quantities are noted in these figures. the peak roof displacement due to each of three modes is ur1o = 9. The linear relationship between the base shear and roof displacement for each mode implies that the structure did not yield. 3.15) and (3. and (d) shows the Vbn .

76E-03 1.3 -3.012 0.205 0.088 -0.202 0.325 0.4 -10.475 0.91E-04 1.1 -0.8 -56.2 Peak values of story drift ratios (as % of story height) from RHA for 0.3 Peak values of joint rotations (radians) from RHA for 0.260 0.09E-03 2.023 0.88E-03 2.090 0.94E-03 2.370 0.24E-03 2.229 0.264 Error (%) 1 Mode 2 Modes 3 Modes -23.31E-03 Mode 2 -1.72E-03 3.7 2.01E-04 3.235 0.453 0.173 0.50E-03 4.5 -1.0 7.1 -14.5 0.6 0.26E-04 -5.231 -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 0.133 -0.060 -0.09E-03 1.42E-04 1.29E-03 2.2 0.44E-03 1.63E-03 2.78E-04 -3.156 0.09E-03 1.055 0.3 -8.3 -33.3 -0.035 0.010 -0.6 -1.197 0.1 3.097 0.6 -1.0 -2.2 -57.88E-03 2.2 1.8 -5.0 -0.226 0.3 -0.3 Table 3.74E-03 1.9 3.13E-04 9.400 0.261 -0.9 9.097 0.1 -2.89E-03 1.042 0.130 0.043 0.080 0.7 -19.062 0.11E-04 -5.336 0.378 0.058 -0. 18 .253 0.125 0.Table 3.227 0.260 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 0.259 0.33E-03 2.2 -1.282 0.265 RHA (all modes) 0.8 1.245 0.0 -10.38E-03 2.199 0.5 18.008 -0.295 -0.74E-04 9.235 0.012 -0.125 0.237 0.311 0.265 0.124 0.152 0.6 4.11E-03 1.9 -24.406 0.202 0.45E-03 3.266 0.227 0.15E-03 4.275 0.4 -0.00E-03 1.8 -15.321 0.50E-03 2.202 0.121 0.74E-04 6.03E-03 1.14 m3) from RHA for 0.311 0.407 0.203 0.1 -11.44E-03 3.7 7.4 -3.054 0.001 -0.65E-03 RHA (all modes) 2.8 -1.303 0.401 Error (%) 1 Mode 2 Modes 3 Modes -23.2 -20.229 0.266 0.85E-03 3.9 -22.2 -2.192 0.225 0.09E-03 Error (%) 1 Mode -23.00E-03 2.006 0.6 0.14E-03 2.5 -2.66E-05 -3.333 0.7 3.9 -16.00E-03 1.1 -19.156 0.99E-03 Mode 3 3.4 -7.266 0.0 3 Modes -5.7 4.1 Peak values of floor displacements (as % of building height = 37.2 9.350 0.7 2 Modes -3.045 0.259 0.13E-03 2.071 0.9 -15.022 0.89E-03 1.015 -0.90E-03 3.31E-03 Joint Rotation (rad) Combined (RHA) 1 Mode 2.6 11.413 RHA (all modes) 0.032 -0.9 -23.258 0.74E-03 1.25 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 3 Modal Response Mode 1 2.4 -53.6 1.159 0.9 1.63E-03 2.062 -0.237 0.002 -0.060 -0.01E-04 -2.310 0.399 0.99E-03 2.76E-03 1.4 -1.3 19.157 0.300 0.179 0.263 0.8 1.011 0.282 0.28E-04 1.124 0.73E-03 3.03E-03 -6.038 0.08E-03 2.011 0.03E-03 3 Modes 2.003 0.466 0.307 0.097 0.1 Building height is measured from the ground floor to the 9th floor.177 0.183 0.009 -0.56E-03 2.64E-03 3.6 9.069 0.22E-03 2.9 8.1 Table 3.7 -50.026 0.8 -10.65E-03 2.6 -0.177 0.1 -2.1 4.024 -0.1 -2.173 0.03E-03 1.117 0.0 -46.38E-04 2 Modes 2.364 0.226 0.4 -6.4 -41.181 0.2 -4.9 2.0 -2.253 0.060 0.069 0.47E-03 1.042 0.245 0.4 -1.4 -22.152 0.4 0.089 0.

and (e) pushover curve.1 −15 −7.1 0.1 /W 0. Response due to first mode: (a) roof displacement.5 15 −0.12 0 (d) 0 (e) 15 Fig.00174 • 0 (c) r1 −5 0 5 10 15 Time (sec) 20 25 30 0.5 9.1 5 x 10 −3 θ (rad) 0.5.15 u (cm) 9.12 • 0 (a) r1 −15 0.1 −15 −7. 3.1 0. (d) force-deformation history.25 × El Centro ground motion 19 .0435 Vb1 / W 0.5 0 u (cm) r1 7.12 −0. Excitation is 0.0435 9.5 0 u (cm) r1 7. (c) joint rotation.0435 • 0 (b) V b1 −0. (b) base shear.

(d) force-deformation history.6.0252 • 0 (b) V −0. Response due to second mode: (a) roof displacement.0252 (d) 0 0 (e) −0. and (e) pushover curve.5 5 Fig.5 0 u (cm) r2 2.0252 −2.5 5 −0.1 −5 −2.1 0.1 −5 −2.23 −2.5 u (cm) r2 0 • 2. (b) base shear.05 /W 0. (c) joint rotation.23 −5 (a) 0. Excitation is 0.23 0. 3.5 0 u (cm) r2 2.05 5 x 10 −3 b2 θ (rad) 0.00244 • 0 (c) r2 −5 0 5 10 15 Time (sec) 20 25 30 0.25 × El Centro ground motion 20 .1 Vb2 / W 0.

(d) force-deformation history.02 −1 −0.0108 x 10 −3 (b) −0.000913 • 0 (c) r3 −2 0 5 10 15 Time (sec) 20 25 30 0.0108 −0.02 −1 −0.422 −1 (a) 0.02 2 θ (rad) 0.02 /W V b3 0 • 0. Excitation is 0.5 1 Fig. 3.02 −0. Response due to third mode: (a) roof displacement.5 0 u (cm) r3 0.25 × El Centro ground motion 21 .5 0 u (cm) r2 0.422 /W V b3 0 (d) 0 −0. (b) base shear.1 u (cm) r3 0 • 0.0108 −0. and (e) pushover curve.7.02 0.422 (e) −0.5 1 −0. (c) joint rotation.

i. these pushover curves are consistent with the Vb .12 cm. respectively. although both analyses led to identical peak values of the individual modal responses. 3.3. two. 3. Implementing pushover analysis for the second and third modes. (3. and external joint rotations computed by pushover analysis are presented in Tables 3.6d and 3. and 3.23 cm . pushing the structure using the force distribution of Eq..5. and 3.4.18).422 cm . 3.2. except for the algebraic sign associated with Γ1 and minor round-off errors.5d).20) with n = 2 and 3 up to roof displacements ur 2 o = 2. this value would be determined directly from the response (or design) spectrum. and ur 3o = 0.4.6e and 3.4. The floor displacements.21). and the errors in these estimates relative to the exact response from RHA considering all modes.e. respectively. 3. or three vibration modes. Observe that the target roof displacement in each pushover analysis is identical to its exact value determined by RHA. These values of the response quantities are identical to the peak response values determined from RHA (Tables 3. 3.5.4) to roof displacement ur1o = 9. 3. confirming that MPA gives the exact values of the individual modal responses. 3. using the force distribution of Eq. the errors in the MPA results are generally larger than in RHA (Fig. story drifts.1. (3. 3. In RHA the errors arise only from truncating the responses due to higher modes.2.3). 3. and 3. (3. and 3.. and the computed response values are identical to the peak response values determined from RHA (Tables 3. pushing the structure. and it is apparent in the example considered that three modes provide most of the response (Fig.7d).1.3).5.20) with n = 1 (Fig.10 and Tables 3. and 3. 3.9).e. For a fixed number of modes included considered. This pushover curve is consistent with the relationship between the base shear and roof displacement determined by RHA (Fig.6 present estimates of the combined response according to Eq.1 through 3. 3. As for the first mode.ur relations determined by RHA (Figs.5e. (3.7e and to the floor displacements. considering one.3).4.8) leads to the pushover curve shown in Fig. (3.6. i. which would provide the Dn value to be substituted in Eq. 3. Figure 3.3 Modal Pushover Analysis Implementing MPA for the fundamental vibration mode. the floor displacements are proportional to the mode shape φ 1 because the structure remains elastic.9 and Tables 3. story drifts. As suggested by Eq.12). In practical application. 3. the value determined by RHA (Fig. respectively. leads to the pushover curves shown in Figs. and external joint rotations in Tables 3.6. implying that the modal truncation errors are small if at 22 .

4 Displacement/Height (%) 0.2 0.least three modes are included.6 Fig. Response histories of roof displacement and top-story drift from RHA for 0.83 • r3 0 • 0.5 0.422 r2 0 • 1.25 × El Centro ground motion 23 .3 0.5 Ground 0 0.8. 3.4 Story Drift Ratio (%) 0.685 • −3 3 Mode 2 ∆ (cm) u (cm) • 2.03 −3 3 Mode 3 ∆ (cm) u (cm) r3 −15 15 9.eps fig3_9b.59 • Total (All Modes) 0 −15 0 ∆ (cm) u (cm) 0 −3 0 fig3_8.1 0. (a) Roof Displacement 15 (b) Top Story Drift 3 Mode 1 ∆ (cm) u (cm) 9.eps r 5 10 15 20 Time.48 −3 3 1. 3. sec 25 30 r 5 10 15 20 Time.3 0. Heightwise variation of floor displacements and story drifts from RHA for 0.1 0.23 r2 −15 15 0 • 0.12 • r1 −15 15 0 r1 0 0 0.2 0. Additional errors are introduced in pushover analysis due to the approximation inherent in modal combination rules.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. sec 25 30 Fig.eps Ground 0 0.9.

043 -0.3 -2.31E-03 Error (%) RHA (all modes) 2.3 Table 3.25 × 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 0.8 -56.4 -9.008 -0.9 -14.1 -0.266 0.00E-03 2.7 -21.203 0.310 0.261 -0.4 -5.121 0.286 0.231 -0.203 0.99E-03 Mode 3 -3.253 RHA (All Modes) 0.9 -15.00E-03 2.062 0.203 0.267 0.9 -13.385 0.012 -0.63E-03 -2.229 0.328 0.270 0.270 0.4 -11.65E-03 2.7 -19.156 0.6 Peak values of joint rotation (radians) from MPA for 0.285 0.7 -50.4 Peak values of floor displacements (as % of building height = 37.332 0.9 -13.4 -22.53E-04 -9.31E-03 2.3 1.0 -16.2 -16.088 -0.9 3 Modes -12.060 0.322 0.022 0.0 -10.267 0.229 0.2 -57.374 0.230 -0.282 -0.09E-04 -3.8 -15.6 -16.05E-03 3.4 -4.5 -16.156 0.038 -0.259 0.117 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 (MPA) 1 Mode 2 Modes 3 Modes Mode 3 -0.47E-03 1.9 -18.38E-03 3 Modes 2.260 -0.173 0.267 -0.89E-03 -1.023 -0.125 0.106 0.Table 3.0 -18.042 0.230 0.78E-04 2 Modes 2.15E-03 1.011 0.0 -46.058 -0.89E-03 2.336 0.22E-04 2.15E-03 2.078 0.321 0.8 -22.00E-03 3.177 0.74E-04 -6.76E-03 -1.197 0.235 -0.4 0.2 -20.274 0.6 -17.09E-03 1.401 Error (%) 1 Mode 2 Modes 3 Modes -23.313 0.12E-03 1.2 -12.03E-03 1.4 -22.2 -0.133 0.4 -2.9 -14.237 0.96E-03 2.44E-03 3.177 0.04E-03 3.466 0.73E-05 3.300 0.3 -14.270 0.7 -15.8 -14.1 -19.4 -7.203 0.4 -4.010 0.92E-04 -1.032 0.7 2.079 0.43E-04 -1.09E-03 Mode 2 1.042 0.090 0.89E-03 2.226 0.125 0.40E-04 5.4 1.0 -0.3 -19.264 Error (%) Modal Response 1 Mode 2 Modes 3 Modes -23.260 0.03E-03 -1.9 -14.048 0.097 0.03E-03 6.002 0.89E-03 -2.237 -0.012 0.8 -23.152 0.2 -4.7 2 Modes -13.080 0.1 -11.9 -11.3 -9.73E-03 3.253 0.44E-03 -1.9 -13.9 -12.036 -0.90E-04 -9.4 -19.3 -41.38E-03 2.28E-03 2.9 -24.9 -15.011 -0.253 0.157 0.1 -11.097 0.22E-03 -2.3 -13.235 0.08E-03 2.6 -19.001 0.336 RHA (all modes) 0.7 24 .9 -16.259 -0.14 m) from MPA for 0.2 1.1 -18.009 0.09E-03 -1.6 -15.0 -2.80E-04 3.106 0.173 0.2 Table 3.055 0.9 -8.069 0.4 -14.133 0.370 0.276 0.045 -0.133 -0.227 0.331 0.071 0.003 -0.00E-03 -1.90E-03 1.015 -0.006 -0.179 0.9 -15.026 0.048 0.203 0.25 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Modal Response Mode 1 -2.89E-03 1.179 0.24E-03 -2.407 0.024 0.5 Peak values of story drift ratios (as % of story height) from MPA for 0.2 -11.152 0.227 0.245 -0.74E-03 -1.124 0.157 0.4 -53.253 -0.245 0.00E-03 2.03E-03 3.3 -33.74E-03 1.181 0.9 -13.272 0.97E-03 1.3 -12.00E-03 1.33E-04 5.8 -15.9 -15.260 0.94E-03 2.4 -9.282 0.72E-03 3.31E-03 Joint Rotation (rad) Combined (MPA) 1 Mode 2.63E-03 2.069 0.42E-04 -1.57E-03 1 Mode -23.76E-03 1.062 -0.296 -0.060 -0.

4 Story Drift Ratio (%) 0. Heightwise variation of floor displacements and story drifts from MPA for 0.3 0. shading indicates modal combination error 25 .(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.25 × El Centro ground motion.5 Ground 0 0.10.2 0.1 0.6 Fig. 3.eps Ground 0 0.4 Displacement/Height (%) 0.1 0.eps fig3_10b.5 0.3 0.2 0.

26 .

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

(4.3).n t given by Eq.4) is rarely solved because it offers no particular advantage over Eq. (3.… N (4.6) for inelastic systems will no longer be described by Eq. Unlike Eq.4)] leads to the uncoupled modal response history analysis (UMRHA) procedure. (4.Substituting Eq. The equations governing the response of the inelastic system to p eff.sn ug t bg b g bg bg (4.5) This resisting force depends on all modal coordinates qn t . give the same results for u t as obtained directly from Eq. in principle. The spatial distribution s of the effective earthquake forces is expanded into the modal contributions sn according to Eq.” implying that other “modes” will also contribute to the solution. (4. sign qn = φ n f s u n . however. (3.Gn ug t Mn bg n = 1. However. sign u = .6) The solution of Eq. 4.9) for linearly elastic systems. (4. (4. Simultaneously solving these coupled equations and using Eq. This approximate RHA procedure is the preliminary step in developing a modal pushover analysis procedure for inelastic systems.6b) are mu + cu + f s u.2 UNCOUPLED MODAL RESPONSE HISTORY ANALYSIS bg Neglecting the coupling of the N equations in modal coordinates [Eq.3) will.9) involves T Fsn = Fsn qn .4) where the only term that differs from Eq. (4.8) because qr t will generally be nonzero for “modes” other than the nth “mode.and classical n damping-orthogonality property of modes gives F qn + 2z nw n qn + sn = . (3.4) represents N equations in the modal coordinates qn . implying coupling of modal coordinates because of yielding of the structure. (3. Eq. Equation (4. and using the mass. where φ n are now the modes of the corresponding linear system. (4.2). (4.2). these equations are coupled for inelastic systems. qr t = 0 for all 28 bg . For linear systems. (3.3) in Eq. sign u n b g b g bg (4.2). 2. premultiplying by φ T .

(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

and (b) approximate solution by UMRHA bg af bg 32 . Fsn / Ln ¨ g(t ) u (a) Static Analysis of Structure (b) Dynamic Analysis of Inelastic SDF System Fig. ζn.n t = .02 −150 n=2 −150 50 n=2 14.575 5 10 15 20 Time (sec) 25 30 ur3 (cm) 0 • 5.51 • ur2 (cm) 0 ur2 (cm) n=3 14. 4.0 ¥ El Centro ground motion: (a) exact solution by NL-RHA.9 • 0 −50 10 −50 10 n=3 ur3 (cm) 0 • 5. and 3.112 5 10 15 20 Time (sec) 25 30 −10 0 −10 0 Fig.4. 4. where u g t = 3.89 50 r1 0 u (cm) 0 • 78. n = 1.3.s n u g t .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. 2. Roof displacement due to p eff.

4) and (4. (2) the superposition of responses to p eff. Such comparison for roof displacement and top-story drift is presented in Figs.1 Underlying Assumptions and Accuracy The approximate solution of Eq.n t (n = 1. and 3. and (b) approximate solution by UMRHA af bg bg 4. (3. respectively.Dn relation is 33 bg bg . The errors are slightly larger in drift than in displacement. 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. n = 1.5. where u g t = 3. and (3) the Fsn Ln .n t = .6) by UMRHA is compared with the “exact” solution by nonlinear RHA. but even for this very intense excitation.(a) Nonlinear RHA 20 n=1 20 (b) UMRHA n=1 • 5.855 ∆r1 (cm) 0 ∆r1 (cm) n=2 • 6.4 and 4.0 ¥ El Centro ground motion: (a) exact solution by NL-RHA. 2.5. Top story drift due to p eff.2.0 ¥ El Centro ground motion.s n u g t . this intense excitation was chosen to ensure that the structure is excited well beyond its linear elastic limit.5)] is neglected.956 5 10 15 20 Time (sec) 25 30 ∆r3 (cm) 0 • 5. 4. both for 3.2… N ) according to Eq. the errors in either response quantity are only a few percent.15) is strictly valid only for linearly elastic systems.008 • 6.744 • 0 −20 10 n=3 −20 10 n=3 ∆r3 (cm) 0 • 5. (4.33 0 −20 20 −20 20 n=2 ∆r2 (cm) 0 ∆r2 (cm) 7. 4. (4.817 5 10 15 20 Time (sec) 25 30 −10 0 −10 0 Fig.

such nonlinear static analysis provides the so-called pushover curve. 4.9) should be determined by nonlinear static analysis of the structure as the structure undergoes displacements u = Dnφ n with increasing Dn . Therefore. (4.Dn curve.2 Properties of the nth-“mode” Inelastic SDF System How is the Fsn Ln . The overall errors in the UMRHA procedure are documented in the examples presented in Section 4.9). (4.8) before it can be solved? Because Eq. it can conduct a force-controlled nonlinear static analysis with an invariant distribution of lateral forces.8) governing Dn t is based on Eq. (4. Although several approximations are inherent in this UMRHA procedure.20) to some bg pre-determined roof displacement. When implemented by commercially available software. The structure is pushed using the force distribution of Eq. the base shear is Vbny and roof displacement is urny . How to convert this Vbn . within the linearly elastic range of the structure.10) .8) in UMRHA. (3.approximated by a bilinear curve to facilitate solution of Eq. the relationship between lateral forces f s and Dn in Eq. (4.20). (3. which is different than the Fsn Ln . A bilinear idealization of this pushover curve for the nth-“mode” is shown in Fig. Therefore we impose this constraint in developing the UMRHA procedure in this section and modal pushover analysis in the next section.Dn relation to be determined in Eq. (3. 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.2.4. this distribution seems to be a rational choiceeven for inelastic systemsto determine Fsn in Eq.urn pushover curve to the Fsn Ln . Although most commercially available software cannot implement such displacement-controlled analysis.1. 4.12) for floor displacements. (4.Dn relation? The two sets of forces and displacements are related as follows: V Fsn = bn Γn Dn = urn Γ nf rn 34 (4. is plotted against roof displacement urn . However.6a. At the yield point. and the base shear Vbn . when specialized for linearly elastic systems it is identical to the RHA procedure of Section 3. the only force distribution that produces displacements proportional to f n is given by Eq.

Section 13. Properties of the nth-“mode” inelastic SDF system from the pushover curve Equation 4.12) implying that the initial slope of the curve in Fig.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. The two are related through Fsny Ln 2 = w n Dny (4.6b is w 2 . the initial slope of the pushover curve in Fig. 4.6. 4.8).5). which may differ from the period of the corresponding linear system. (4. 4. (4. In contrast. Knowing Fsny Ln and Dny from n Eq.10 enables conversion of the pushover curve to the desired Fsn Ln − Dn relation shown in Fig. where the yield values of Fsn Ln and Dn are Fsny Ln = Vbny * Mn Dny = urny Γ nf rn (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. which is not a meaningful quantity.2. should be used in Eq.13) This value of Tn . 4.11) * in which M n = Ln Γn is the effective modal mass (Chopra. 35 .11).6a is 2 kn = ω n Ln . 2001.6b.

Consider a nonlinear static analysis of the structure subjected to lateral forces distributed over the building height according to s* [Eq. the first two or three modes will suffice.3b) with force-deformation relation of Fig. 4. Convert the idealized pushover curve to the Fsn Ln − Dn relation (Fig. details are available in Appendix A: 1. the peak value of Dn t .20)]. Combine the “modal” responses using Eqs. For the nth-mode. Calculate the peak value. Compute the deformation history. 4. which is summarized next as a sequence of steps.20)].n t . n 3.12) and (3.13). n (3. φn . is now determined by 36 bg bg . Calculate histories of various responses by Eqs. (3.6b) by utilizing Eq. 4. (3.16) to determine the total response.2. Dn (t ) . 7. 8.3 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 . (4. r o .4. An t . Typically.11). Repeat Steps 2 to 6 for as many modes as required for sufficient accuracy.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. ωn . and pseudo-acceleration history. 4. This value of the roof displacement is given by Eq. of the nth-“mode” inelastic SDF system (Fig. 4. Idealize the pushover curve as a bilinear curve (Fig. (3. and modes. with the structure is pushed to the roof displacement urno . (3. 5. for linearly-elastic vibration of the building. 9. 4.15) and (3.6a).21) where Dn . of the total response r t obtained in Step 8.6b. 2. develop the base-shear – roof-displacement ( Vbn − urn ) pushover curve for the force distribution s* [Eq. 6. Compute the natural frequencies.

4. This pushover analysis. details are available in Appendix B. Steps 1 to 4 of the MPA are same as those for UMRHA.8). The peak “modal” responses rno . Compute the peak deformation. the pushover analysis provides an estimate of the peak value rno of any response rn t : floor displacements. it can be determined from the inelastic response (or design) spectrum (Chopra.6). etc. each determined by one pushover analysis. 4. The response value rno is an estimate of the peak value of the response of the inelastic system to p eff. as described in Section 4. However. as mentioned earlier.g.2. (4. 2001. plastic hinge rotations. for elastic systems.. alternatively. or from the inelastic response (or design) spectrum. governed by Eq. As shown in Sections 3.8). 6. although somewhat intuitive for inelastic buildings. bg bg bg bg Thus. 4.18). Calculate the peak roof displacement urno associated with the nth-“mode” inelastic SDF system from Eq.2 and 3.3. 5.3) because. (4. Eq. This application of modal combination rules to inelastic systems obviously lacks a theoretical basis. e. (3. At this roof displacement.1 Summary The peak inelastic response of a building to earthquake excitation can be estimated by the MPA procedure just developed. Dn . (4. rno also represents the exact peak value of the nth-mode contribution rn t to response r t .4. Sections 7.6 and 7. we will refer to rno as the peak “modal” response even in the case of inelastic systems. (3. to obtain an estimate of the peak value ro of the total response.3.6b by solving Eq. story drifts.n t . 37 . which is summarized next as a sequence of steps.solving Eq. it seems reasonable because it provides results for elastic buildings that are identical to the well-known RSA procedure described in Section 3. It provides results for elastic buildings that are identical to the well-known RSA procedure (Section 3. joint rotations.12).2.3b) with force-deformation relation of Fig. are combined using an appropriate modal combination rule. of the nth-“mode” inelastic SDF system (Fig.21). seems reasonable. and (2) it provides the exact modal response for elastic systems. the lateral force distribution used possesses two properties: (1) it appears to be the most rational choice among all invariant distribution of forces.

in particular. and the errors in the approximate results are shown in Fig. and ur 3o = 2. 4.2). ur 2o = 11.18). the 38 . subtract the yield value of hinge rotation to determine the hinge plastic rotation. Figure 4.4. and the “exact” response from nonlinear RHA for the roof displacement and top-story drift. (3. and the percentage errors in the approximate results. Typically. and compared with the results of a rigorous nonlinear RHA using the DRAIN-2DX computer program. 9. respectively.7. 4. the UMRHA procedure lacks a rigorous theory. although the trends are not systematic as when the system remained elastic (Section 3. three “modes” included.0 to 3. extract from the pushover database values of other desired responses. is presented next.4. determined by the UMRHA procedure. 4.8. The peak values of floor displacements and story drifts including one. the first two or three “modes” will suffice. also included are the combined responses due to one. At urno . Determine the total response by combining the peak “modal” responses using the SRSS combination rule of Eq. two.1 Uncoupled Modal Response History Analysis The structural response to 1.1 and 4.2. 8.53 cm. This is to be expected.9. 4. and three “modes”.0.7 shows the individual “modal” responses. From the total rotation of a plastic hinge.1). in contrast to modal analysis (Section 3. the peak roof displacement due to each of the three “modes” is ur1o = 48. and three modes are compared with the “exact” values in Fig. To ensure that this structure responds well into the inelastic range the El Centro ground motion is scaled up by a factor varying from 1.7 cm. The peak values of displacements of all floors and drifts in all stories are presented in Tables 4. rno . Observe that errors tend to decrease as response contributions of more “modes” are included. This deficiency also implies that.3 cm.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. the combined response due to three “modes”. the “exact” results.5 ¥ the El Centro ground motion including the response contributions associated with three “modal” inelastic SDF systems. say. with. The peak values of response are as noted. two. Repeat Steps 3 to 8 for as many “modes” as required for sufficient accuracy.

1 and 3.0. the errors in responses computed by UMRHA including three “modes” relative to the “exact” response were determined. 2.1 −12 12 ∆r (cm) u (cm) UMRHA 3 "Modes" 0 • 7. for a fixed number of “modes” included.5 × El Centro ground motion: individual “modal” responses and combined response from UMRHA.2). just as for elastic systems.2) and 1. For this purpose the UMRHA and exact analyses were repeated for ground motions of varying intensity. recall that the computed errors have been presented earlier for ground motion multipliers 0.response is much less accurate (Tables 4.5.24 • NL−RHA r −12 12 ∆r (cm) u (cm) 0 5 10 15 20 Time. the degree to which the system deforms beyond its elastic limit. and 3.85.25 (Tables 3. 0. 1.0. 0. in particular.88 u (cm) "Mode" 3 −80 80 0 −80 80 0 −80 0 • 44. However.5 (Tables 4.53 ∆r3 (cm) 12 0 • 2. For each excitation.25.1 and 4. and total response from NL-RHA 39 . we investigate how the errors in the UMRHA vary with the deformation demands imposed by the ground motion. 1.38 6.2) if the system yields significantly versus if the system remains within the elastic range (Tables 3. Next.3 11. defined as the El Centro ground motion multiplied by 0.5.44 • "Mode" 2 r1 0 • 48.2).7 • −80 80 ∆r1 (cm) u (cm) "Mode" 1 −12 r2 0 ∆r2 (cm) 12 0 u (cm) −80 80 −12 r3 0 • 2. 0.75.1 and 4.62 5.7. (a) Roof Displacement (b) Top Story Drift 80 12 0 • 3. sec 25 30 r −12 0 Fig.1 and 3.0.6 5 10 15 20 Time. 4. the errors in story drifts are larger compared to floor displacements. sec 25 30 • 48. Response histories of roof displacement and top-story drift due to 1.

Heightwise variation of error in floor displacements and story drifts estimated by UMRHA including one. with the peak displacement of each “modal” SDF system noted for each ground motion multiplier.9.8.20)] for the first three n modes of the system are shown in Fig. 4. Heightwise variation of floor displacements and story drift ratios from UMRHA and NL-RHA for 1. it will be useful to know the deformation of the system relative to its yield deformation. two.5 ¥ El Centro ground motion Figure 4.5 Fig. in each story drift (Fig.11.5 × 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.5 Displacement/Height (%) 2 Ground 0 0. and the error envelope for each case.10a).5 Story Drift Ratio (%) 2 2. the pushover curves using force distributions s* [Eq. 4. or three “modes” for 1. (3.(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). 4.5 1 1.5 1 1. To interpret these results. indicated by a ground motion multiplier.10 summarizes the error in UMRHA as a function of ground motion intensity. For this purpose. 4. Shown is the error in each floor displacement (Fig. Two versions of the pushover curve are 40 .

The structure is pushed using the force distribution of Eq.0.75. and 3 (Fig. the pushover curve for each “mode” is idealized by a bilinear curve in solving Eq. 11. Second. For more intense excitations. Additional errors are introduced in UMRHA of systems responding beyond the linearly elastic limit for at least two reasons.included: the actual curve and its idealized bilinear version. 41 . First.5).11). 3.5 ¥ the El Centro ground motion.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. The idealized curve for the first “mode” deviates most from the actual curve near the peak displacement corresponding to ground motion multiplier 1. and (4) are larger in story drift compared to floor displacements. the errors in truncating the higher mode contributions are negligible. Perhaps this explains why the errors are large at this excitation intensity. (3) maintain roughly similar values for more intense ground motions. even though the system remains essentially elastic. (3. respectively.53 cm.75. 4.4. the values determined by RHA of the nth-mode inelastic SDF system (Fig.4). as mentioned in Section 4. 4. and. (4. and 2. the peak values of displacements at all floors (Table 4.10) for each “modal” inelastic SDF system (Figs 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. Each of these three pushover analyses provides the pushover curve (Fig. and plastic hinge rotations at the external beam end at each floor level (Table 4.20) with n = 1. UMRHA lacks a rigorous theory and is based on several approximations.11a).7 cm.0.01 (Fig.4.6 and 4. the ductility factor for the first mode system is only 1. were noted but are not shown here. 2.7). determined from “exact” analyses. The location of plastic hinges and their rotations. as mentioned in Section 3.1). 4.2.3 cm. drifts in all stories (Table 4.2.4) to roof displacements urno = 48. Figure 4. The system remains elastic up to ground motion multiplier 0. the first reason mentioned above seems to be the primary source for the errors.3). (2) increase rapidly as the ground motion multiplier increases to 1. 4.

575 -41.130 0.049 -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 -1.003 0.2 4.371 0.6 4.256 1.663 -0.371 -0.298 0.372 -1.9 12.214 0.3 1.806 0.4 4.372 1.8 14.410 1.133 1.366 0.033 0.283 1.2 Peak values of story drift ratios (as % of story height) from UMRHA for 1.0 7.079 0.863 1.490 -1.065 0.1 1.942 -0.055 -0.241 -1.856 2.8 1.338 1.554 1.727 1.9 12.1 3.4 1.513 0.201 -0.472 1.430 1.1 0.5 10.200 8.370 -0.068 0.513 -0.216 1.484 0.1 1.376 -1. 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 0.241 1.226 -0.4 -7.971 1.2 12.6 2.4 -1.426 -1.10.2 1.942 1.407 -10.410 -1.2 6.5 -3.844 -25.14 m) from UMRHA for 1.009 -0.495 1.248 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.088 10.0 -9.7 31.1 8.5 1 1.366 -0.5 2 GM Multiplier 2.668 -23.315 -0.5 18.2 22.0 2.983 1.187 -0.8 1.490 1.235 -0.071 -0. 4.8 0.811 1.9 5.8 1.003 -31.5 28.3 0.057 -0. Noted 60 50 40 (b) Story Drifts Error Envelope Error for Story No.(a) Floor Displacements 60 50 40 Error Envelope Error for Floor No.938 1.914 2.202 11.0 11.8 1.0 1.676 0.900 -10.120 1.707 1.9 31.5 3 0 0 0.0 11.044 1.3 25.072 1.338 -1.806 -0.5 28.293 1.478 0.820 -19.698 1.763 -15.135 9.071 0.9 16.5 9.5 2 GM Multiplier 2.3 6.526 -0.8 1.138 1.2 1.6 1.877 0.7 Table 4.722 0.333 0. Errors in UMRHA as a function of ground motion intensity: (a) floor displacements.938 -1.863 0.126 0.350 -0.214 -0.317 0.201 1.291 0.220 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” -0.072 -1.1 Peak values of floor displacements (as % of building height = 37. and (b) story drifts Table 4.169 0.9 31.663 0.945 -37.5 3 Fig.7 14.819 2.914 -0.201 -1.751 1.104 0.096 0.540 0.772 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 1.3 42 .220 -0.5 9.373 -0.852 1.136 1.820 -0.616 -0.070 1.256 1.3 8.379 1.121 -0.376 1.298 -0.010 0.260 -15.0 9.5 1 1.256 -1.982 9.473 -22.018 0.154 0.

4). Obviously.4 with Tables 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. significantly smaller than in UMRHA (compare Fig. and are within a few percent of the exact values for the upper stories. but the third “mode” contributions do not seem especially important (Fig.18).13 and Tables 4. The errors are especially large in the hinge plastic rotations estimated by the MPA procedure. considering one. but was more successful when the excitation was more intense (results are not included).” and three “modes. 4. MPA failed to identify the plastic hinges at the column bases in Fig. especially in estimating the story drifts (Fig. The first “mode” alone is inadequate. in general. the third “mode” does not contribute because this SDF system remains elastic (Fig. were determined by four analyses: MPA considering one “mode. 4. 4. The locations of plastic hinges shown in Fig.11c). 4.3 and 4.4. even if three “modes” are included (Fig. 4.12 and Tables 4. this error is not especially significant because the hinge plastic rotation is very small. 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%.14. 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.12 and 4. Furthermore. (3.13 with Fig. 4.14.5).Figures 4. although the error is recorded as 100% if MPA estimates zero rotation when the nonlinear RHA computes a non-zero value. overestimated by up to 18% in the middle stories. significant improvement is achieved by including response contributions due to the second “mode”.3 and 4. As shown in Figs.” and nonlinear RHA. 4. for two or three modes included.13 and Tables 4.3 through 4.2). Pushover analysis seems to be inherently limited in computing accurately hinge plastic rotations. the hinges identified in beams at the 6th floor were at variance with the “exact” results. the errors in the modal pushover results are.13 and Table 4. The drifts are underestimated by up to 13% in the lower stories.12 and Tables 4. the results were not always accurate. Fortuitously. The second “mode” was necessary to identify hinges in the upper stories.5 present estimates of the combined response according to Eq. and three “modes.3 and 4.3 and 4. two. For example. However. however. and the errors in these estimates relative to the exact response from nonlinear RHA.” respectively. 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. 4. 4.” two “modes.9 and Tables 4.4).1 and 4. 43 .

5 0. V = 4952 kN. α = 0. α = 0.85 0.85 Base Shear (kN) 8000 6000 4000 2000 0 0 0.0.5 2 3 4000 2000 0 0 Actual Idealized 0.25.75.5.25 1 2 4 6 Roof Displacement (cm) 8 10 Fig. α = 0.5 2 3 Actual Idealized 2000 0 0 0. 1.75 Actual Idealized 20 40 60 Roof Displacement (cm) (b) "Mode" 2 Pushover Curve 80 12000 10000 u = 9. V = 7616 kN.11. “Modal” pushover curves with peak roof displacements identified for 0. 4.5 0.13 y by Base Shear (kN) 8000 6000 1 1. V = 5210 kN. 2.19 y by 3 2 1.75 0.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36.14 y by Base Shear (kN) 8000 6000 4000 1. and 3.75 0.0.5.25 0. 1.2 cm. 0.5 1 0.9 cm.5 0.25 5 10 15 20 Roof Displacement (cm) 25 (c) "Mode" 3 Pushover Curve 12000 10000 u = 4. 0.6 cm.85 0.0 × El Centro ground motion 44 .

an observation with practical implications. and the error envelope for each case. each story drift (Fig.5 Fig.(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 0.4.10 and 4.4 also apply to MPA. the errors in MPA were fortuitously smaller than in UMRHA (compare Figs.5 1 1. indicated by a ground motion multiplier.15) for ground multipliers larger than 1. 4. 4. Shown is the error in each floor displacement (Fig.15a).3.5 × El Centro ground motion.5 Story Drift Ratio (%) 2 2. The errors are only weakly dependent on ground motion intensity (Fig.15b). implying excitations weak enough to limit the response in the elastic range of the structure. 4.75. As mentioned in Section 3. the errors in MPA were larger for ground motion multipliers less than 0. However. whereas MPA contains errors inherent in modal combination rules. UMRHA is essentially exact.3 for elastic systems (or weak ground motions). now standard in engineering practice.15 summarizes the error in MPA considering three “modes” as a function of ground motion intensity.4. While various sources of errors in UMRHA. implying excitations intense enough to cause significant yielding of the structure. 4. the MPA procedure is equivalent to the RSA procedure. 4. identified in Section 3. 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.15).12 Heightwise variation of floor displacements and story drift ratios from MPA and NL-RHA for 1. In this case. shading indicates errors in MPA including three “modes” Figure 4. as discussed in Sections 3.5 Displacement/Height (%) 2 Ground 0 0. 45 .2 and 3. implying that the modal combination errors contained in these procedures are acceptable.5 1 1.0.

and three “modes” for 1. 4.(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. two.5 ¥ El Centro ground motion 46 . story drifts.13. and hinge plastic rotations estimated by MPA including one. Errors in floor displacements.

00E+00 0.00E+00 1 “Mode” 7.304 -1.820 -7.36E-03 6.00E+00 0.1 62.9 -4.980 0.18E-03 7.426 15.00E+00 0.00E+00 0.00E+00 0.50E-10 3.687 0.55E-03 3.705 -1.5 1.478 0.00E+00 0.298 -0.414 1.0 1.007 1.399 0.594 -1.26E-04 9.982 13.60E-04 7.19E-10 3.5 × 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 0.429 0.009 0.8 0.945 -49.1 62.00E+00 0.5 × El Centro ground motion Hinge Plastic Rotation (rad) “Modal” Response Combined (MPA) “Mode” 1 7.733 1.176 0.154 0.8 -4.00E+00 2 “Modes” 7.068 0.066 -0.00E+00 0.88E-03 1.190 -0.033 -0.053 1.517 1.0 -50.9 5.55E-03 3.76E-03 4.00E+00 3.00E+00 0.844 -7.37E-03 1.99E-03 6.518 1.37E-03 1.00E+00 0.36E-03 6.60E-03 2.6 -44.1 -8.6 1.4 -8.399 0.8 17.260 -14.5 2.36E-03 6.00E+00 0.503 1.197 -0.233 1.737 0.738 1.652 1.8 -6.581 0.05E-03 2.088 12.503 -1.5 10.049 -0.220 1.737 1.015 0.756 0.575 -53.305 -0.9 0.895 1.066 -0.071 -0.018 -0.667 1.1 13.527 -0.76E-03 4.2 1.0 -100.012 1.135 -7.00E+00 0.156 -0.372 0.8 1.250 0.315 0.02E-03 0.331 1.2 0.00E+00 3 “Modes” 7.55E-03 3.00E+00 0.2 -100.351 -0.728 1.222 0.1 1.5 7.7 1.371 -0.5 Table 4.209 1.202 8.9 1.6 -7.6 13.200 8.683 1.007 1.7 1.057 0.003 -16.6 13.018 -0.5 Peak values of hinge plastic rotations (radians) from MPA for 1.640 1.3 Peak values of floor displacements (as % of building height = 37.0 47 .338 1.473 -15.015 0.640 -1.8 1.1 62.752 1.1 18.781 0.498 1.14 m) from MPA for 1.942 6.611 0.88E-03 0.3 -3.8 7.666 Table 4.2 6.72E-03 7.88E-03 0.2 11.304 1.900 -0.02E-03 3.8 -12.516 0.36E-03 6.581 0.641 1.336 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.22E-10 NL RHA 1.0 -50.0 -100.0 -100.4 Peak values of story drift ratios (as % of story height) from MPA for 1.5 2.9 1.756 0.9 -100.910 1.0 15.055 0.9 -100.101 -0.2 0.879 1.105 0.76E-03 4.3 -3.220 1.76E-03 4.37E-03 1.266 -0.053 -1.2 -4.514 -1.244 0.6 -44.3 11.5 × 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” 1.2 -100.72E-03 7.53E-03 7.00E+00 “Mode” 3 0.9 7.2 -100.895 1.02E-03 0.1 46.233 1.018 0.668 -13.125 -1.804 1.763 -14.3 13.311 0.37E-04 Error (%) “Modal” Response 1 “Mode” -32.0 3 “Modes” -32.130 0.02E-03 3.259 1.1 46.10E-02 9.5 2.8 -6.222 0.8 -29.37E-03 1.3 1.9 2.00E+00 0.00E+00 0.00E+00 0.6 7.694 1.745 1.72E-03 7.911 0.705 1.407 -27.00E+00 0.116 1.781 0.6 -8.980 -0.6 -9.8 “Mode” 3 -1.7 -12.118 0.0 -5.4 1.5 -6.071 0.168 -0.116 1.667 -1.2 -3.5 10.1 46.72E-03 7.298 0.8 -29.Table 4.00E+00 0.00E+00 0.116 1.0 2 “Modes” -32.614 0.00E+00 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th “Mode” 2 0.435 0.8 -29.

” but not the contributions of higher “modes.16 are unchanged. Static analysis of the structure for gravity loads provides the initial state—forces and deformations—of the structure.5 × El Centro ground motion are presented below. MPA including two “modes” underestimates the displacements of lower floors by up to 6% and overestimates the upper floor displacements by up to 22%.18 and 48 . Obviously gravity load effects will influence the seismic demands due to the first “mode. and three “modes.6 and 4. Figures 4. 4. As shown in Fig.7). Starting with its initial state under gravity loads.4.7).20) with n = 1 (Fig. The first “mode” alone provides adequate estimates of floor displacements. The story drifts are underestimated by up to 7% in the lower stories.18 and Tables 4.” these effects will modify the combined modal response as well as the results of nonlinear RHA. and the structure is pushed using the force distribution of Eq.3 Modal Pushover Analysis with Gravity Loads To evaluate the accuracy of the dynamic response of the system.” however. the structure is pushed using the force distribution of Eq. (3. the results presented so far did not include gravity load effects. and plastic hinge rotations at selected external beam end at each floor level (Table 4. 4. The response contributions of the second and third modes are the same as before (Tables 4.11a).53 cm.17 and Tables 4.5).18) considering one.7 cm and ur 3o = 2.7). the third “mode” contributions do not seem especially important (Fig. 4.0 cm.17 and Tables 4.16a—which is slightly different than the one excluding gravity loads (Fig. 4. drifts in all stories (Table 4.6). as are the roof displacements ur 2o = 11.8).4) to roof displacement ur1o = 52. Also presented at these roof displacements are the displacements at all floors (Table 4. 4.8 present estimates of the combined response according to Eq. but it is inadequate in estimating the story drifts (Fig. The pushover curves for the second and third modes included in Fig.4.6 through 4. (3.3 .” and the errors in these estimates relative to the exact response from nonlinear RHA.20) with n = 1 to a target roof displacement.17 and 4. Significant improvement is achieved by including response contributions due to the second “mode. resulting in the pushover curve shown in Fig. Results of such analyses for the selected building subject to 1. and overestimated by up to 29% in the middle stories and by up to 13% in the upper stories.7.18 and Tables 4. They are now included in the pushover analysis of the structure for the first “mode” only. 4.6 and 4. two. (3. The errors are especially large in the plastic hinge rotations estimated by the MPA procedure even if three “modes” are included (Fig. 3.6 and 4. 4.4.

Locations of plastic hinges determined by MPA considering one. 1−"Mode" • • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • •• • •• • • • • • •• •• • •• •• •• •• (b) MPA. (a) MPA.5 1 1.5 × El Centro ground motion (a) Floor Displacements 60 50 40 Error Envelope Error for Floor No. 2−"Modes" •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• • •• (c) 3−"Modes" • • • • • • • •• •• • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• • •• • • • • • • • • • (d) Nonlinear RHA •• •• • •• •• •• •• • •• •• •• •• •• •• •• • •• •• • •• •• •• •• • • • • • • •• •• •• Fig. 4. Most pushover analysis procedures do not seem to compute to acceptable accuracy plastic hinge rotations. and (b) story drifts 49 .5 2 GM Multiplier 2. two.5 3 Fig.5 2 GM Multiplier 2. Noted 60 50 40 (b) Story Drifts Error Envelope Error for Story No. 4. and three “modes” and by NL-RHA for 1. Errors in MPA as a function of ground motion intensity: (a) floor displacements.5 3 0.15.8).14.Table 4.5 1 1. 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.

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.20a). These errors are only slightly larger than those in Fig. 4. each story drift (Fig.20b).20 summarizes the error in MPA considering three “modes” as a function of ground motion intensity. Figure 4. 50 . and the errors are only weakly dependent on ground motion intensity.The locations of plastic hinges shown in Fig. indicated by a ground motion multiplier. With two modes included in MPA. The second “mode” is necessary to identify hinges in the upper stories.19 were determined by four analyses: MPA considering one “mode.” two “modes”. excluding gravity load effects. 4. Shown is the error in each floor displacement (Fig. MPA provides response values accurate enough for practical application in building retrofit or design. and the error envelope for each case. this procedure is able to predict plastic hinge locations essentially consistent with nonlinear RHA. 4.15. 4. and nonlinear RHA.

0. 0.19 y by 3 2 1.25.6 cm.9 cm.25 0. noted are peak values of roof displacement for 0.5 1 0. α = 0. V = 4952 kN.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36.5 2 3 Actual Idealized 2000 0 0 0.75.5 0.14 y by Base Shear (kN) 8000 6000 4000 1.3 cm. 2.0 × El Centro ground motion 51 . α = 0.5 0. V = 5210 kN.25 5 10 15 20 Roof Displacement (cm) 25 (c) "Mode" 3 Pushover Curve 12000 10000 u = 4.13 y by Base Shear (kN) 8000 6000 1 1. “Modal” pushover curves with gravity loads included.75 0.25 1 2 4 6 Roof Displacement (cm) 8 10 Fig. and 3. 1. 4.85 0. 0.5 0. α = 0.50.75 0.85 Base Shear (kN) 8000 6000 4000 2000 0 0 0. 0.75 Actual Idealized 20 40 60 Roof Displacement (cm) 80 (b) "Mode" 2 Pushover Curve 12000 10000 u = 9.85.16.0.85 0.5 2 3 4000 2000 0 0 Actual Idealized 0. V = 7433 kN.

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. and 52 .5 Displacement/Height (%) 2 Ground 0 0.(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 0.5 Fig. gravity loads included. 4.5 1 1. two.5 Story Drift Ratio (%) 2 2.5 1 1. Errors in floor displacements. 4.18. Heightwise variation of floor displacements and story drift ratios from MPA and NL-RHA for 1.5 × El Centro ground motion. and hinge plastic rotations estimated by MPA including one.17. story drifts.

gravity loads included 53 . gravity loads included (a) Floor Displacements 60 50 40 Error Envelope Error for Floor No. Errors in MPA as a function of ground motion intensity: (a) floor displacements.20. 4. 1−"Mode" • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • •• •• •• • • • • • •• •• •• •• •• •• •• (b) MPA.19. two.(a) MPA. and three “modes” and by NL-RHA for 1.5 2 GM Multiplier 2. Noted 60 50 40 (a) Story Drifts Error Envelope Error for Story No. 4. 3−"Modes" (d) Nonlinear RHA • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• •• •• • • • • • • • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • • • • • •• •• •• Fig. 2−"Modes" •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• •• •• (c) MPA.5 3 Fig.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 0.5 3 0.5 2 GM Multiplier 2.5 1 1. Locations of plastic hinges determined by MPA considering one. and (b) story drifts.5 × El Centro ground motion.

037 -0.1 1.098 20.821 1.319 1.454 1.107 1.908 -1.665 0.00E+00 0.2 0.351 -0.0 37.19E-03 1.850 -1.8 Table 4.3 -22.00E+00 0.516 0.19E-03 1.35E-03 8.2 9.5 × El Centro ground motion.530 1.156 -0.037 0.26E-03 3.176 0.527 -0.877 -46.35E-10 3.2 4.6 Peak values of floor displacements (as % of building height = 37.00E+00 2 “Mode” 8.9 31.88E-03 0.00E+00 3.101 -0.998 21.23E-03 0.603 -1.88E-03 0.237 0.11E-03 9.5 21. gravity loads included Hinge Plastic Rotation (rad) Floor “Modal” Response “Mode” 1 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 8.00E+00 0.5 “Mode” 3 -1.854 0.Table 4.6 2.5 3 “Modes” -32.068 0.197 -0.2 -2.414 28.8 9.594 -1.310 1.0 37.371 -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” 0.00E+00 0.102 1.7 1.5 10.4 -4.55E-03 3.00E+00 0.434 0.190 -0.4 -6.466 0.938 1.1 13.436 1.5 54 .00E+00 0.19E-03 1.102 1.0 37.213 1.3 -100.1 4.530 1.009 0.821 -1.7 -2.2 -0.8 Peak values of hinge plastic rotations (radians) from MPA for 1.687 0.3 -22.2 9.2 1.00E-03 5.330 1.888 1.00E+00 0.128 -1.2 -3.8 -32.6 0.330 1.263 1.00E+00 0.066 -0.6 1.114 1.983 1.0 -100.2 1.3 -100.2 9.00E+00 “Mode” 2 0.5 1.728 1.7 -2.2 12.00E+00 0.130 0.3 -22.530 1.00E+00 0.00E+00 0.00E+00 Error (%) “Modal” Response 1 “Mode” -32.14 m) from MPA for 1.00E+00 0.996 0.75E-03 0.319 1.8 28.5 21.154 0.0 2 “Modes” -32.831 0.071 -0.19E-04 5.13E-03 5.2 21.399 -0.11E-03 9.3 -100.263 0.4 20.921 1.23E-03 3.429 1.490 -11.23E-03 3.353 -23.033 -0.983 1.04E-02 8.00E+00 3 “Modes” 8.9 -6.998 1.5 1.927 1.603 1.813 1.00E+00 “Mode” 3 0.55E-03 3.0 -100.207 18.996 -0.429 -1.399 0.00E+00 0.2 2.2 4.17E-03 9.315 0.11E-03 9.9 -4.015 0.4 1.637 0.00E+00 0.7 19.2 4.822 1.00E+00 0.00E+00 0.3 1.831 0.0 16.850 1.434 0.213 1.168 -0.8 -32.7 4.933 1.913 7.23E-03 0.00E+00 0.4 -4.1 21.461 0.434 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.213 1.55E-03 3.8 0.19E-03 1.00E+00 Combined (MPA) 1 “Mode” 8.057 0.836 -0. 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” 1.953 15.5 -5.783 1.667 0.071 0.00E-03 5.00E-03 5.018 -0.270 -12.908 1.35E-03 8.6 19.266 -0.514 -1.125 -1.11E-03 9.105 0.6 2.998 0.305 -0.8 0.066 -0.055 0.7 16.114 -1.00E-03 5.515 -50.00E+00 0.3 13.754 0.00E+00 0.5 × El Centro ground motion.860 1.78E-03 1.673 Table 4.744 1.0 -30.257 0.754 1.858 2.199 16.7 Peak values of story drift ratios (as % of story height) from MPA for 1.4 26.478 0.9 -3.23E-02 1.854 0.8 1.5 × El Centro ground motion.064 -10.04E-10 3.88E-03 1.507 1.0 -30.686 -7.2 16.3 9.637 0.4 1.00E-10 NL RHA 1.311 0.35E-03 8.214 0.5 0.109 0.35E-03 8.372 0.237 0.830 -12.049 -0.

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.” and nonlinear RHA.5. 2… N ).5 sec . assumed to be linearly elastic. 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.. Specified in FEMA-273 are three distributions for lateral forces: 1. gravity load effects were included in all 55 . plastic hinge rotations. and the exponent k = 1 for fundament period T1 £ 0. Comparison of Modal and FEMA Pushover Analyses 5. 1997] The pushover curve. computed at the target displacement represent the earthquakeinduced demands on the structure. “Uniform” distribution: s* = m j (where the floor number j = 1. k = 2 for T1 ≥ 2. 5. and varies linearly in between. a plot of base shear versus roof displacement. joint rotations.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. story drifts. The floor displacements. MPA considering three “modes. etc. and 3. j 2.5 sec . SRSS distribution: s* is defined by the lateral forces back-calculated from the story shears determined by response spectrum analysis of the structure.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.

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.042 0.2.165 0. the floor displacement demands in Fig.11 0.0446 0. (b) ELF. Also included in these presentations are the MPA results considering three “modes” and the “exact” demands from nonlinear RHA.4a.11 0. Figures 5.1.177 0. and the locations of all plastic hinges in Fig. 0.0702 0.11 0. 5. 5.4 and Tables 5. Force distributions in FEMA-273: (a) “uniform”. The three FEMA force distributions are presented in Fig.4.11 0. pushover analyses are implemented for a target roof displacement of 52.3a and 5. 5. The first four floor displacements are within 5% of the “exact” value. both presented in Section 4.analyses.3.3.5 times the El Centro ground motion. the value determined from RHA of the first-mode inelastic SDF system for 1. the story drift demands in Fig. 5.0896 0.1 through 5.0197 0. and Table 5. 5.3.4. and (c) SRSS 56 .0981 0.0466 0.00719 0.3a and Table 5. wherein the first two are obvious and the third is determined from response spectrum analysis of the building (Appendix C).3b and Table 5. 5. plastic hinge rotation demands in Table 5. 5.281 0.112 0.126 0.062 0.2. and the displacements of upper floors are overestimated by 9-22%. The “uniform” distribution overestimates all floor displacements by 17-28%.11 0. The errors in the FEMA and MPA estimates of seismic demands relative to the “exact” demands are presented in Fig.11 0.11 0. The pushover curves are given in Fig.367 (a) Uniform (b) ELF (c) SRSS Fig.0381 0. with the errors being larger for the SRSS distribution.119 0.21 0.1.0913 0.1. The MPA procedure is more accurate than all the FEMA force distributions.0654 0.0 cm. Using each of these force distributions.

5.2.3 cm.24 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized 10 20 30 40 Roof Displacement (cm) 50 60 Fig.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. gravity loads are included 57 . Pushover curves using three force distributions in FEMA-273: (a) “uniform”. α = 0.8 cm. 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. and (c) SRSS.(a) Uniform Distribution 12000 10000 u = 33 cm. V = 8530 kN. V = 6897 kN. α = 0. (b) ELF.

7%. larger errors are noted in the upper and lower stories. the errors are largest in the lower stories.2 demonstrate that the story drift demands are greatly underestimated by all the FEMA force distributions. and Table 5. and overestimated by no more than 32%.4c and Table 5. with errors reaching 37% in this example. the MPA procedure is more accurate than all the FEMA 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.4b. gravity loads included 58 .3b and 5. MPA including three “modes. reaching 64%. with story drifts under estimated by. reaching 35%.Figures 5.5 Ground 0 Ground 0 Fig. such as hinge plastic rotations.5 1 1. 3 "Modes" FEMA Uniform ELF SRSS 0.3 demonstrate that the hinge plastic rotations estimated by all three FEMA force distributions contain unacceptably large errors. In contrast.5 Story Drift Ratio (%) 2 2.) The pushover analysis procedures considered seem incapable of computing accurately local response quantities. at most. errors are largest in the upper stories. For the SRSS distribution. but it is still inaccurate. 3 "Modes" FEMA Uniform ELF SRSS 0.5 1 1. Figure 5. For the ELF distribution.3. whereas nonlinear RHA computed an insignificantly small value. This seems to be an inherent limitation of pushover analysis. reaching 31%. Heightwise variation of floor displacements and story drift ratios estimated using FEMA-273 force distributions. (The 100% error in a 6th floor hinge is ignored as it simply represents that the MPA estimated zero rotation.” and NLRHA. Modal pushover analysis procedure gives estimates better than all the FEMA force distributions. 5.5 Displacement/Height (%) 2 Floor 5th 5th 4th 3rd 2nd 1st NL−RHA MPA. For the uniform distribution.

and the ELF distribution fails 59 . Errors in floor displacements. The locations of plastic hinges shown in Fig. Based on the results presented here. 5. 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. the SRSS distribution fails to identify yielding of beams in the middle floors. 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. 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. and the three FEMA analyses.4. the “uniform” distribution fails to identify yielding of the beams above the fourth floor.” nonlinear RHA (“exact”). they are not good indicators of damage. story drifts. The locations of the plastic hinges are not identified correctly by the FEMA force distributions. and hinge plastic rotations estimated using FEMA-273 force distributions and MPA (including three “modes”).5 were determined by five analyses: MPA considering the three “modes. While pushover estimates for floor displacements are more accurate.

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

195 0.5 -32.00E+00 0.310 1.8 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 0.8 2.8 -100.314 1.1 10.854 0.5 -33.9 -70.7 7.7 26.0 -63.984 1.2 9.667 0.09E-03 4.221 1.6 22.5 -27.19E-03 1.0 -100.4 16.03E-03 5.6 -17.3 -4.399 1.10E-02 7.2 -100.55E-03 3.128 1.51E-03 4.270 0.490 0.015 0.730 1.530 1.487 0.414 1.5 -3.2 16.6 4.836 0.23E-02 1.8 44.306 1.16E-03 0.560 1.4 -55.7 -29.263 0.3 Peak values of hinge plastic rotations (radians) from FEMA force distributions and MPA Hinge Plastic Rotation (rad) MPA FEMA ELF SRSS 4.3 -100.168 1.6 -6.52E-04 1.7 MPA -2.35E-10 3.067 0.877 1.7 Table 5.708 0.3 -14.672 1. gravity loads included Displacement /Height (%) FEMA ELF SRSS MPA 0.436 Error (%) NL Uniform RHA 0.23E-03 3.5 -29.083 1.199 27.58E-04 6.65E-03 7.341 1.323 1.6 -41.462 1.686 0.88E-03 1.94E-03 2.6 14.3 31.789 0.7 21.11E-03 9.234 1.953 0.913 0.9 -77.724 0.524 0.860 1.00E-03 5.6 17.0 -100.2 4.59E-03 5.372 1.2 Peak values of story drift ratios (as % of story height) from FEMA force distributions and MPA.9 16.19E-04 5.209 0.8 -100.7 -28.00E+00 0.783 1.318 2.830 2.5 -33.53E-02 1.351 0.9 28.262 1.17E-03 9.6 21.0 -100.00E+00 2.34E-03 2.0 -59.344 0.7 -15.975 1.00E+00 0.8 -2.00E+00 4.335 2.0 -71.611 0.178 1. Peak values of floor displacements (as % of building height = 37. gravity loads included Displacement /Height (%) FEMA ELF SRSS MPA 1.3 -22.62E-03 0.530 1.214 1.9 -20.2 13.399 1.00E+00 0.089 1.1 0.00E+00 0.566 1.13E-03 5.7 19.736 0.2 6.2 -12.14 m) from FEMA force distributions and MPA.998 1.5 22.50E-03 0.0 -30.809 0.7 FEMA ELF SRSS -22.9 0.294 1.0 Table 5.9 16.5 10.5 26.623 1.562 1.154 1.8 7.2 16.1 -26.353 1.6 -11.5 Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 2.5 15.6 -4.466 0.673 Error (%) NL RHA 1.Table 5.4 -50.2 5.1 -3.3 29.2 -4.2 19.1 163.00E-10 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 1.417 1.399 -27.927 1.00E+00 Error (%) FEMA Uniform ELF SRSS MPA 24.9 11.4 -12.2 9.4 -23.061 1.93E-03 1.597 0.4 MPA -2.5 61 .7 -60.8 -32.00E+00 8.75E-03 0.888 0.0 37.007 1.0 -100.875 0.858 1.547 -27.998 1.4 16.330 1.011 1.938 1.45E-03 3.4 0.355 0.992 1.1.04E-02 8.064 1.8 16.9 SRSS -22.515 FEMA Uniform ELF 27.78E-03 1.367 1.26E-03 3.00E+00 NL RHA 1.35E-03 8.3 23.207 1.5 -29.6 -73.3 -11.84 1.109 1.8 -35.6 -25.9 15.00E+00 0.2 10.839 0.8 -63.8 17.78E-03 0.4 21.0 -57.098 1.

MPA including three “modes” and NL-RHA for 1. 5.5 × El Centro ground motion.(a) FEMA: Uniform • • • • • • • • •• •• •• •• •• •• (b) FEMA: ELF • •• •• •• •• •• •• • •• • •• •• •• •• •• • • • • • •• • • • • • • • •• •• •• •• • •• •• •• •• • •• •• •• •• • • •• •• •• • • • • • •• •• (c) FEMA: SRSS • • •• •• •• •• • • •• •• •• • • •• •• •• • • • • •• •• •• • • • • • • • • • (d) MPA with 3 "Modes" • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• •• •• (e) Nonlinear RHA • • • • • • • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • • • • • •• •• •• Fig. gravity loads included 62 . Locations of plastic hinges determined from three force distributions in FEMA-273.5.

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 3 0.5 1 1.5 2 GM Multiplier 2. Errors in floor displacements from three force distributions in FEMA-273 and from MPA including three “modes”.5 1 1.5 3 0 0 0.(a) FEMA: Uniform 60 50 40 60 50 40 3 (b) FEMA: ELF Error Envelope Error for Floor No. gravity loads included 63 .5 1 1. 5.5 2 GM Multiplier 2.5 3 Fig.5 1 1.5 2 GM Multiplier 2.6.5 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.

gravity loads included 64 . 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 0.5 1 1.5 3 0 0 0.5 2 GM Multiplier 2.5 2 GM Multiplier 2. Error in story drifts from three force distributions in FEMA-273 and from MPA including three “modes”.5 1 1.5 3 Fig.5 3 0 0 0.5 2 GM Multiplier 2.5 1 1.5 1 1.7.5 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.(a) FEMA: Uniform 100 60 50 9 8 (b) FEMA: ELF Error Envelope Error for Story No.5 2 GM Multiplier 2. 5.

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

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

Pushover analyses seem to be inherently limited in computing accurately hinge plastic rotations.(iv) At the roof displacement determined from Dn .” 10. Comparing the earthquake-induced demands for the selected 9-story building determined by pushover analysis using three force distributions in FEMA-273. story drifts. it fails to compute with acceptable accuracy plastic rotations of the hinges. Based on results presented for El Centro ground motion scaled by factors varying from 0. and hinge plastic rotations. 8. all pushover analysis procedures do not seem to compute with acceptable accuracy local response quantities. the errors in the MPA procedure are shown to be only weakly dependent on ground motion intensity.25 to 3. MPA. and identifying locations of most plastic hinges. story drifts. demonstrates that the FEMA force distributions greatly underestimate the story drift demands. such as hinge plastic rotations. 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. and lead to unacceptably large errors in the hinge plastic 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.0. However. 9. 7. 67 . etc. joint rotations. The MPA procedure is more accurate than all the FEMA force distributions in estimating floor displacements. plastic hinge 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. and nonlinear RHA. the pushover analysis provides the peak value rno of any response quantity: floor displacements. Thus the MPA procedure is accurate enough for practical application in building evaluation and design.

such as life safety and collapse prevention. they are not good indicators of damage. 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.11. 68 . with the excitation scaled to cover a wide range of ground motion intensities and building response. While pushover estimates for floor displacements are even more accurate. should obviously be evaluated for a wide range of buildings and ground motion ensembles. This new method for estimating seismic demands at low performance levels. Instead. 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. 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.

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

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

The structure should be pushed just beyond the target (or expected) roof displacement in the selected mode.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. Define the force distribution s* from Eq.g.g. 1988)..2. Apn . Since the target roof displacement may not be known at the start of the procedure. iterations may be necessary. e. 2. Idealize the pushover curve as a bilinear curve (Fig. Apply force distribution of Step 2. trapezoidal rule. of the bilinear curve at the target roof displacement.Appendix A Uncoupled Modal Response History Analysis A. using any numerical integration method. 3.2. A. and modes. ω n . 3. develop the base-shear – roof-displacement (Vbn .1 incrementally and record the base shears and associated roof displacements. 71 . for linear-elastic vibration of the building. Let the roof displacement and base shear at the anchor point be urno and Vbno . respectively. This step can be conveniently implemented in any commercially available software.urn ) pushover curve for the force distribution s* : n 2. Compute natural frequencies. φn . For the nth-“mode”. 3.1. Define the anchor point.. e.1) using the FEMA-273 procedure (Building Seismic Safety Council. DRAIN2DX (Allahabadi and Powell.1. (3. B. Calculate the area under the actual pushover curve. 1977).20): s* = mφn n n 2. 1.

4. Calculate initial slope of the idealized bilinear curve.2.3. Abn . i i i 3. kn .8. Vbny . and B with straight-line segments to obtain the idealized bilinear curve.6 . Draw the curve OAB by connecting the three points O.4.9. corresponding to the estimated yield i i i base shear. i 3. From the pushover data.2. Calculate Vbny = Vbny × Apn Abn .5.0. i 3. If the error exceeds some pre-specified tolerance.1. Replace i+1 with i and repeat Steps 3. This step gives the secant stiffness at a base shear equal to 60% of the yield base shear. Calculate the post-yielding strain-hardening ratio. Calculate the yield displacement. If desired. kn = 0. 3. A. Calculate the slope. O. i 3.6. i i i αn =  Vbno Vbny − 1  urno urny − 1         i 3. iterations are necessary.9. obtained by judgment. i +1 i i 3. by connecting a straight line between origin.6 × Vbny ( ) i urn . and a point on the actual pushover curve with base shear equal to i 0. other appropriate methods can be ( used.0. at base shear i equal to 0.7.6 × Vbny .4 to 3. 3.4. 72 . urn. Vbny . Calculate area under the bilinear curve OAB. will be refined by an iterative procedure that seeks to equate areas under the actual and the idealized pushover curves. Let the point with base shear = Vbny and roof displacement = urny be denoted as A. Calculate the error = 100 × Abn − Apn ( ) ( ) ( ) ) Apn .9.6 × Vbny . This value. urny = Vbny kn .1.i 3.8. determine the roof displacement.6 . 3. i i 3. Estimate the yield base shear.

6. (3. A. * 4. 4. 4. 5. Compute the Ln and Γ n from Eq.4.2. (3. r o . Scale the vertical axis with * Mn to obtain * Fsno Ln = Vbno M n and * Fsny Ln = Vbny M n (Eqs.3b) with unit-mass and force-deformation relation of Fig. 4. Dn (t ) . Scale the horizontal axis by Γnφrn to obtain Dno = urno Γ nφrn and Dny = urny Γ n φrn (Eqs. and pseudo-acceleration history.3.4) and effective modal mass from M n = Ln Γn .6 ur n y ur n o ur n Fig.12) and (3. Combine the “modal” responses using Eqs.1 Idealization of nth-“mode” pushover curve 73 . (3.2. An (t ) .16). Compute deformation history. Calculate peak values. of the nth“mode” inelastic SDF system (Fig.1. Vbn Idealization of Pushover Curve Vbno Vbny 0.6 × Vbny A • Idealized αnkn B • 1 Actual k 1 • O n figA_1.10b and 4. of the combined responses obtained in Step 8.11b). 7. Develop the Fsn Ln − Dn relation (Fig.0. Calculate histories of various responses using Eqs. 4. A. 4.10a and 4.eps ur n. In general first two or three modes will suffice.11a).2).15) and (3. Repeat Steps 3 to 6 for as many modes as required for sufficient accuracy. 9. 8.13). A.

The pushover curves for the first three modes. A. Idealized bilinear curves for each of the three modes are included in Fig.). is defined at the target roof displacement. 3. and 12.3.1. A. The force distributions. s* .6 kN.5. 3. First three mode shapes and frequencies of the selected building were computed and are shown in Fig. generated using DRAIN-2DX. second. The base-shear – roof-displacement (Vbn . 3. 74 .eps Dny = ur n y / Γnφr n Dno = ur n o / Γnφr n Dn Fig. computed for the first three modes are shown in Fig.4.urn ) pushover curve for the force distribution s* : n 2.5 cm (25 in. are shown in Fig.2.2 EXAMPLE The UMRHA procedure is implemented to calculate the response of the 9-story building described in Section 3. The anchor point. Area under the actual pushover curve. 3.4 cm (10 in. Target displacements used to generate these pushover curves are 63. 3.). and third mode.).2 Properties of the nth-“mode” inelastic SDF system A. respectively. 1. Following is step-by-step implementation of the procedure described in Section A. 2.Fsn / Ln Fsn / Ln − Dn Relationship Vbno / M* n Vbny / Mn * 1 α ω2 n n ω2 1 n figA_2.5 cm and Vb1o = 8729. B. The following steps illustrate the procedure to develop the idealize curve for the first “mode”.4. Ap1 = 360777 kN-cm. for the first.7 cm (5 in.3.1. to the north-south component of the El Centro (1940) ground motion scaled up by a factor of 1. 25.1. ur1o = 63.2. n 2. A. At this point.3.

The yield displacement. The results for other modes are summarized in Table A.86 = 210.     i 3.5 38. α1 =  Vb1o Vb1 y − 1  ur1o ur1 y − 1 =         ( ) ( ) (8729. L1 = 2736789 kg. i 3. and α1 = 0. and M1 = 2736789 × 1.0.09 cm.9.6 kN. The procedure converged after nineteen cycles to give ur1 y = 36.2.194.6 = 22.2. iterations are necessary.4 ) − 1 (63. and B with straight-line segments gives the idealized bilinear curve. k1 . i i 3. Determined from the pushover database. The results of the iterative procedure are summarized in Table A.4. k1 = 0.86 cm at 0.3. A.09 cm and Vb1 y = 8006.2.1.6 × Vb1 y = 4803. i i 3. i i i 3. The Fs1 L1 − D1 relation for the first “mode” is developed as follows.01%. i i i 3. is calculated as follows. The post-yielding strain-hardening ratio. Therefore.18 = 38.3666. Error = 100 × (365100 − 360777 ) 360777 = 1. The point A on the i i bilinear curve is defined by ur1 y = 38. 75 .5.09 ) − 1 = 0.18 kN/cm.8. (4.4 kN.9 kN.9.1.0. * 4.4 × (360777 365100 ) = 7911. The first estimate of the yield base shear Vb1 y = 8006.4 210.6 = 4803. Vb1 y = 7615.1. 3.13).4. Γ1 = 1. 3.7.23 cm. The next estimate of the yield shear is Vbny = 8006.8 22.4.135. ur1. 4. The curve OAB obtained by connecting the three points O.9. Ab1 = 365100 kN-cm. ur1 y = Vb1 y k1 = 8006.8 kN.1. This value exceeds the prespecified tolerance of 0. Area under the bilinear curve OAB. Also included are the modal damping ratios and the periods calculated from Eq.6.6 × Vb1 y ( ) i ur1.i 3.3666 = 3740189 kg. i +1 3.4 kN. The initial slope of the idealized bilinear curve. 3.198%.6 8006.

5.1 and 4. Histories of roof displacement and top story drifts for the first “mode” are computed and presented in Fig. Scaling the vertical axis by Fs1 y L1 = 203. 7. 8. 9.2.8. A. 4. 6.7. The combined modal responses are presented in Fig. 4.7. 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.4.3. The peak values are computed and are summarized in Tables 4.2.4. 4. Scaling the horizontal axis by Γ1φr1 gives D1o = 46. 4. The peak values are also plotted in Fig.46 cm and D1y = 26.51 cm. The results were generated for first three “modes’ and are included in Fig.7. 4.62 (cm/sec2).40 (cm/sec2) and 76 . * M1 gives Fs1o L1 = 233.

α = 0.2 cm.3. α = 0. V y by = 4952 kN.eps 0 0 5 10 15 20 Roof Displacement (cm) (c) "Mode" 3 Pushover Curve 25 8000 u = 4.eps 10 20 30 40 50 Roof Displacement (cm) (b) "Mode" 2 Pushover Curve 60 8000 u = 9. V = 5210 kN.eps 0 0 2 4 6 8 10 Roof Displacement (cm) 12 Fig.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36.6 cm.14 y by 6000 Base Shear (kN) 4000 2000 Actual Idealized figA_3c.9 cm. “Modal” pushover curves for the example building 77 .19 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized figA_3a. α = 0. V = 7616 kN. A.13 6000 Base Shear (kN) 4000 2000 Actual Idealized figA_3b.

m −80 0 −3 0 25 30 Fig.” and third “mode” inelastic SDF systems. 78 .223 −80 80 n=3 −3 3 n=3 D3 (cm) 0 A3 (g) • 10.33 A1 (g) n=2 0 −80 80 −3 3 n=2 D2 (cm) 0 A1 (g) • 22.06 0 • 1.5 × El Centro ground motion for the first “mode.52 5 10 15 20 Time (sec) 25 30 0 • 1.4.747 5 10 15 20 Time (sec) figa_4. A.2769 D1 (cm) 0 • 35.” second “mode. Histories of deformation and pseudo-acceleration due to 1.(a) Deformation 80 n=1 3 (b) Pseuso−acceleration n=1 • 0.

3 4603.2 4628.25 36.50 36.910 0.1 7616.193 0.70 36.75 21.11 22.063 0.6 × Vb1 y i ur1.7 7639.107 0.79 21.0 4588.9 4573.1 4574.048 0.5 7624.6 i k1 i ur1 y i α1 i Ab1 Error (%) 1.9 (kN) 4803.9 4570.3 7786.76 21.29 36.162 0.8 7622.35 36.170 0.85 36.2 4571.8525 1.194 0.190 0.51 2 “Mode” 2 -920860 -0.188 0.237 0.2.1 1013.0.62 26.44 36.18 210.192 0.09 37.77 21.24 36.18 210.3 7745.8 4647.18 210.0 4704.05 36.23 0.184 0.309 0.176 0.8 4747.136 Fsny Ln (cm/sec2) Dny (cm) Fsno Ln (cm/sec ) Dno (cm) Tn (sec) 233.23 22.05 52.151 0.037 0.9 4570.191 0.64 37.103 “Mode” 3 696400 0.2671 1.38 22.4 7714.18 210.18 210.4 7672.18 210.46 2.4 4595.182 0.5 (cm) 22.90 21.32 36.5309 488839.404 0.693 0.18 210.85 0.5 7633.022 0.56 19.86 21.81 21.18 210.948 ζ n (%) 79 .1.7 4580.2406 167531.26 36.02 21.59 36.18 210.010 (kN) 8006.029 0.4927 1.194 0.18 210.2 4671.1 4569.135 0.18 210.18 210.082 0.79 0.18 (cm) 38.39 36.65 1226.2 7690.017 0.0 4577.0 7619.28 36.139 0.59 22.6 7840.18 210.18 210.74 (kN/cm) 210.30 37.18 210. Properties of “modal” inelastic SDF systems Properties Ln (kg) Γn * M n (kg) “Mode” 1 2736789 1.186 0.013 0.95 21. No.74 21.3 7628. Results of iterative procedure to develop the idealized bilinear curve for the first “mode” inelastic SDF system Itr.40 46.Table A.194 0.198 0.8 7618.180 0.193 0.9 7615. i 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 i Vb1 y i 0.3666 3740189 203.86 22.09 18.56 47.194 (kN-cm) 365100 364060 363276 362684 362235 361892 361631 361432 361279 361162 361073 361004 360951 360911 360880 360856 360838 360824 360813 Table A.3 7658.75 21.78 21.18 210.25 36.6 4583.5 3109.193 0.529 0.4 7647.18 210.83 21.12 3876.5 4614.4 7911.18 210.

80 .

B. Typically. Following is step-by-step implementation of the procedure described in Section B. 11. 4.1. From the total hinge rotation. or from the inelastic response (or design) spectrum. (3. rno .2 EXAMPLE The MPA procedure is implemented to calculate the response of the 9-story building described in Section 3. (4. subtract the yield hinge rotation to determine the plastic hinge rotation. 12. Calculate the peak roof displacement urno associated with the nth-“mode” inelastic SDF system from Eq. 4.4. of the nth-“mode” inelastic SDF system (Fig. Compute the peak deformation.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. (3.6b by solving Eq. 81 .8).33 cm.1 to the north-south component of the El Centro (1940) ground motion scaled up by a factor of 1. Steps 1 to 4 of the MPA are the same as those for UMRHA presented in Appendix A. At urno .Appendix B Modal Pushover Analysis B. extract from the pushover database values of other desired responses.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. 10. Dn . 13.18). Repeat Steps 3 to 7 for as many “modes” as required for sufficient accuracy. Solving Eq.21). (4. This is the same example as solved in Appendix A. Determine the total response by combining the peak “modal” responses using the SRSS combination rule of Eq. 6. the first two or three “modes” will suffice.3b) with unit mass and force-deformation relation of Fig. 10.5.

The total response computed by combining the peak “modal” responses using the SRSS combination rule of Eq. values of floor displacements and story drifts are extracted.11. where results for other ground motion intensities are also included.28 cm.18) are also included in Tables 4. Peak roof displacement ur1o = Γ1 × φr1 × D1 = 1. Steps 3 to 7 are repeated for first three “modes. Results of Steps 5 and 6 are summarized in Table B. 12.4 for the story drifts normalized by the story height.5 are the plastic hinge rotations computed by subtracting the yield hinge rotation from the total hinge rotations. Also included in Table 4. At ur1o = 48.4. The values are summarized in Table 4.” and the results are included in Tables 4. 82 .366 × 1 × 35.3 for the floor displacements normalized by the building height and in Table 4.1.33 = 48.4.3 and 4. (3.28 cm. 9. 8.3 and 4.

8451 5.748 63.79 0.755 27.4222 3.37 57.33 1.267 5.395 0.229 8.913 22.332 48.38 22.71 1.126 13.332 13.577 16.75 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.73 24.06 1.70 0.766 7.856 31.52 0.28 46.1.467 14.03 0.535 14.312 1.551 2.03 26.185 11. Calculation of roof displacements urno from peak deformation of inelastic SDF systems Ground Motion Multiplier 0.252 9.436 7.82 1.735 3.0 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 3.007 36.35 1.275 1.023 0.450 4.35 0.27 0.154 78.5 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.18 27.268 0.225 2.504 18.25 20.901 8.660 14.50 35.05 0.13 2.25 Quantity Dn (cm) “Mode” 1 6.0 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 1.969 0.117 5.37 1.52 “Mode” 3 1.367 1.676 6.5 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 2.184 0.755 0.139 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.690 10.36 1.0 µ = Dn Dny urno = Γnφrn Dn (cm) 83 .379 21.200 0.07 “Mode” 2 4.85 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 1.678 0.457 12.Table B.691 0.513 0.59 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

7 -46.2 95.6 1233.3 222. Table C.9 1683.2 285.6 319.For convenience.047 0.6 366.5 -350.1 87.0 136.2 200.0 176.090 0.5 159.7 694.4 1759.6 -359.7 234.9 -153.9 -166.7 Vj (kN) (8) 2268.7 2065.2 148.7 1622.0 980.1 -646.1 832.6 286.8 430.5 320.2 105.9 880.0 1231.1 -967.5 215.9 832.070 0.7 (10) 0.1 Story Shears V j2 (kN) (6) 1114.9 1446.4 400. 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) 1917.098 0.6 -352.065 0.3 240.8 374.1 -438.367 87 .7 374.2 277.0 -5.7 1578.5 -973.042 0.7 95.0 354.2 97.7 525.2 fj = ∑ fi i fj (2) (3) (4) 59.5 -320.8 1381.9 446.6 -732.0 381.7 355.8 -326. the lateral forces are normalized by the base shear to obtain column 10 in Table C.3.177 0.4 1842.7 101.0 1476.9 366.6 V j3 (kN) (7) 478.3 -646.1 1857.3 -6.045 0.2 Lateral Force fj (kN) (9) 203.4 250.3.1 -525.

Sign up to vote on this title
UsefulNot useful