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 .

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

iv .

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

vi .

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

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

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

have also been made to consider more than the fundamental vibration mode in pushover analysis [Paret et al. The MPA procedure is then extended to inelastic buildings. the seismic demands determined by pushover analysis using three force distributions in FEMA-273 are compared against the MPA and nonlinear RHA procedures... Finally. 2000]. Matsumori et al. Gupta and Kunnath. 1998. 2000. 2 . Kunnath and Gupta. First. but provides superior accuracy in estimating seismic demands on buildings. 1996. 2000.. we show that pushover analysis of a one-story system predicts perfectly peak seismic demands. 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. Sasaki et al. the underlying assumptions and approximations are identified. and the errors in the procedure relative to a rigorous nonlinear RHA are documented. 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 yield deformation. defined by f Ry = o fy (2. 2.2) . 2.1) b g ug t is mu + cu + f s u. through the yield strength reduction factor.1 One-Story Systems EQUATION OF MOTION Consider the idealized one-story structure shown in Fig. The yield strength is related to f o . It is the familiar bilinear hysteretic relationship. It consists of a mass m (or weight w) concentrated at the roof level. R y . sign u . this system is linearly elastic with stiffness k as long as the force does not exceed f y . the yield strength. During yielding the stiffness of the frame is a k . Unloading from a maximum deformation takes place along a path parallel to the initial elastic branch. where 0 < a << 1. Yielding begins when the force reaches f y and the deformation reaches u y . reloading from a minimum deformation takes place along a path parallel to the initial elastic branch. a massless frame that provides stiffness to the system. sign u = . The yield strength is the same in the two directions of deformation.1a. Similarly.2 2. the strength required for the structure to remain elastic during the ground motion. The hysteretic relation between the lateral force f s and lateral displacement u of the mass relative to the base of the frame is denoted by f s u.mug t bg The governing equation for this inelastic system subjected to horizontal ground acceleration b g bg 3 (2. This lateral-force displacement relation is idealized as shown in Fig. On initial loading.1b. and a linear viscous damper with damping coefficient c.

. and u y . This becomes evident if Eq. Tn = 2p w n is the natural vibration period.fS m fs c ut u fy 1 αk k 1 uy ug -fy k k um u 1 1 (a) (b) Fig. (2. or absolute (without regard to algebraic sign) maximum. and (b) bilinear hysteretic forcedeformation relation For a given excitation u g t .3) ωn = k m ζ = c 2mω n f fs = s fy (2. z . m depends on three system parameters: w n . the deformation u t depends on three systems parameters: w n . u £ u y ).3). Section 7. (a) Idealized one-story structure. 2001.e. 2.5) For a given u g t . The peak. bg 4 .4) and w n is the natural vibration frequency. in addition to the form of the force-deformation relation. deformation is denoted by um . sign u = .1. and z is the damping ratio of the system vibrating within its linear elastic range (i.1) is divided by m to obtain ~ u + 2zw n u + w 2 u y f s u. and the ductility factor is u m= m uy (2. and R y (Chopra .ug t n where bg bg b g bg (2. z y .

m h = 3. α = 0.26 kN (8.1. 2.9 kN (38. is f y / w = ( f o / w ) ÷ 8 = 0.2 SYSTEM AND EXCITATION CONSIDERED Consider the one-story system in Fig.2.66 m • EIb EIc EIc • • L = 7. For . 2. this system and excitation. One-story.376 cm. Vby = 39.26 kN. f o w = 184 .2 5 . The yield strength of the inelastic system. with Tn = 0. 2. u (cm) 8 Fig.2 kips) .3. 2.2.5 sec and z = 5% subjected to the north-south component of the El Centro (1940) ground motion (Fig.4 in Chopra. 2001) and scaled up by a factor of 2. 6.2: the dimensions and flexural rigidity of the structural elements are noted.826 kips) for w = 169.2311 . based on Ry = 8 . Vb (kN) 30 20 10 0 0 2 4 6 Roof Displacement. one-bay frame 50 uy = 1.04 40 Base Shear. and f y = 39.32 m • Fig. Pushover curve for structure shown in Fig.

and q pm = 0.36 cm. pushover analysis cannot provide any cumulative measure of response. the energy dissipated in 6 . e. 2.4edetermined from RHAis superimposed. as apparent in Fig.0217 rad . A static pushover analysis of this one-story system leads to the force-displacement relationship shown in Fig. Implementing this analysis with I c = 6.4e.35. 2. respectively.4 shows the earthquake response of the system described in the preceding section determined by response history analysis (RHA)..6 kip-in.4 ). Determined from pushover analysis at the exact peak deformation. (b) shows the lateral force f s t or base shear Vb t normalized relative to the bg weight. It is organized in five parts: (a) shows the deformation u t .65 kN-m (191. and E = 2 × 108 kPa (29×103 ksi) gives yield moments of 21.18 kN-m (444. the joint rotation and beam hinge rotation are identical to values q m and q pm determined from RHA. The peak values of the various response quantities are as follows: um = 7.077 × 107 mm 4 (146 in.134 × 107 mm 4 (75.1 kip-in. 2. 2. However. (d) shows the rotation q p bt g of the plastic hinges at the bg bg beam ends.3 RESPONSE HISTORY ANALYSIS Figure 2. 2. wherein the hysteretic force-deformation history of Fig.017 rad . the ductility factor m = 5.) and 50.36 cm .) for the beam and columns. The system is excited well beyond the yield deformation. The resulting pushover curve is shown in Fig.4f.3.4 ).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. I b = 3. (c) shows the joint rotation q bt g . 2. The yielding stiffness of each structural element is defined as 3% of its initial stiffness. Observe that the pushover curve matches the initial loading path of the hysteretic system. 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 . q m = 0. the peak value determined from RHA.g. and (e) shows the force-deformation relation.3 in.The yield moments in the beam and columns are defined as the bending moments due to the lateral force f y .

and (f) pushover curve 7 .2 Vb / w 0.4 0.36 −15 0. 2.2 −0.4 −10 −5 0 u (cm) 5 10 −0.04 0.4 −10 −5 0 u (cm) 5 10 Fig.0217 • θ (rad) 0 (c) −0. This represents an inherent limitation of pushover analyses.04 0 5 10 15 Time (sec) 20 25 30 0.2311 (a) Vb / w 0 f / w = −0.017 • 0 (d) θp (rad) −0.4.04 0. (b) base shear.2 −0.2 0 (e) 0 (f) −0.yielding during the ground motion. (d) plastic hinge rotation.2311 y (b) −0.04 0. (c) joint rotation.5 0.4 0. 15 u (cm) 0 • 7. Response of one-story system to El Centro ground motion: (a) deformation.5 fy / w = 0. or the cumulative rotation at a plastic hinge. (e) force-deformation relation.

8 .

mι ug t bg bg bg (3. each element of the influence vector ι is equal to unity.3) where φ n is the nth natural vibration mode of the structure. m. (3. The right side of Eq. c. and p eff t = Â peff. 2001: Section 13.12): mι = n =1 Â sn = N n =1 Â Γ nmφ n N (3.1 MODAL RESPONSE HISTORY ANALYSIS The differential equations governing the response of a multistory building to horizontal earthquake ground motion u g t are as follows: mu + cu + ku = .3.4) 9 . Elastic Multistory Buildings 3. and lateral stiffness matrices of the system.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 . This force distribution can be expanded as a summation of modal inertia force distributions s n (Chopra.snug t n= n =1 The effective earthquake forces can then be expressed as af N af N af (3.m ι u g t bg bg (3. classical damping.n t = Â .1) where u is the vector of N lateral floor displacements relative to the ground. and k are the mass.1) can be interpreted as effective earthquake forces: p eff t = .

with no contribution from other modes. (3.12) . it can be demonstrated that none of the modes other than the nth mode contribute to the response. an SDF system with vibration propertiesnatural frequency w n and damping ratio z n of the nth-mode of the MDF system.8) where the modal coordinate qn t is governed by 2 qn + 2z nw n qn + w n qn = . The solution qn t can readily be obtained by comparing Eq.9) and (3.u g t n Comparing Eqs.n t = .7) By utilizing the orthogonality property of modes. p eff .peff (t ) = n =1 ∑ peff .n t is entirely in the nth-mode.5) The contribution of the nth mode to s and to p eff t are: sn = Gn mφ n respectively. we will outline that the response of the MDF system to p eff .n (t ) = N n =1 ∑ −sn u g (t ) N (3. subjected to u g t : Dn + 2z nw n Dn + w 2 Dn = . (3.Gn ug t (3.10) gives qn t = Gn Dn t bg bg bg (3.8) gives the floor displacements u n t = Gnφ n Dn t bg (3. Then the floor displacements are u n t = φ n qn t bg bg bg bg (3.6) Next.9) to the equation of motion for the nth-mode elastic SDF system. (3.sn ug t bg (3.10) bg bg bg 10 (3. The equations governing the response of the system are mu + cu + ku = .11) and substituting in Eq.sn ug t bg bg bg bg (3.9) in which w n is the natural vibration frequency and z n is the damping ratio for the nth mode.

16) This is the classical modal RHA procedure wherein Eq.1.15) and (3.12) and (3. internal element forces. (3.14) is the pseudo-acceleration response of the nth-mode SDF system (Chopra. st The two analyses that lead to rn and An t are shown schematically in Fig.1. However. 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. Conceptual explanation of modal response history analysis of elastic MDF systems Any response quantity r t story drifts.can be expressed by st rn t = rn An t bg bg bg bg bg (3.Forces sn An(t ) st rn ωn.12) and (3. Equations (3.13) define the contribution of the nth-mode to the response. (3. ζn ¨ g(t ) u (a) Static Analysis of Structure (b) Dynamic Analysis of SDF System Fig.1). In contrast to the classical 11 bg . and Eqs. the static value of r due to external forces sn .9) is the standard modal equation governing qn t .16) reflect combining the response contributions of all modes.n t . these standard equations have been derived in an unconventional way. etc. 3.15) rt = bg n =1 Â rn t = bg n =1 Â rnst An bt g N (3. Therefore. and An t = w 2 Dn t n bg (3. 2001.13) st where rn denotes the modal static response. Eqs. Section 13.13) represent the response of the MDF system to p eff. (3. 3.

Chopra. we note that static analysis of the structure subjected to lateral forces f no = Γ n mφ n An (3.3 MODAL PUSHOVER ANALYSIS (3. Sections 12. The SRSS rule. which is valid for structures with well-separated natural frequencies such as multistory buildings with symmetric plan.16).. and Tn = 2p w n is the natural vibration period of the nth-mode of the MDF system. we used the modal expansion of the spatial distribution of the effective earthquake forces. this response value can be obtained by static analysis of the structure subjected to lateral forces distributed over the building height according to 12 . 3. (3. Alternatively.1).1. 2001.g.3). In such a response spectrum analysis (RSA). 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 3.2 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.18) To develop a pushover analysis procedure consistent with RSA. The peak modal responses are combined according to the Square-Root-of-Sum-ofSquares (SRSS) or the Complete Quadratic Combination (CQC) rules.17) where An is the ordinate A Tn . (3.derivation found in textbooks (e.(3.4 and 13. 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. Section 13. This concept provides a rational basis for the modal pushover analysis procedure developed later. 2001. the peak nth-mode response as in Eq. z n of the pseudo-acceleration response (or design) spectrum for the nth-mode SDF system.19) will provide the same value of rno .9) .17) (Chopra.8.

the Applied Technology Council (ATC). was designed by Brandow & Johnston Associates1 for the SAC2 Phase II Steel Project. The building has a basement level. 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. Although not actually constructed. California. denoted B1.15 m (30 ft) on center.1 System and Excitation Considered The 9-story structure. The interior bays of the structure contain simple framing with composite floors.. Third St. (3. 3.s* = mφ n n (3. The bays are 9. Typical floor-to-floor heights (for analysis purposes measured from center-of-beam to center1 2 Brandow & Johnston Associates.18) to obtain an estimate of the peak value ro of the total response. Consulting Structural Engineers. the peak value of the roof displacement due to the nth-mode. this building is 45. in both directions.2). region. and California Universities for Research in Earthquake Engineering (CUREE). Obviously Dn and An are available from the response (or design) spectrum.21) 2 where Dn = An ω n . The peak modal responses.4 COMPARATIVE EVALUATION OF ANALYSIS PROCEDURES 3.12) is urno = Gnf rn Dn (3. SAC Steel Project Technical Office. CA 94804-4698. urno . A benchmark structure for the SAC project. (3. 3. shown in Fig.73 m (150 ft) in plan.73 m (150 ft) by 45. 3. 1301 S. 1660 W.20) and the structure is pushed to the roof displacement. This modal pushover analysis (MPA) for linearly elastic systems is equivalent to the well-known RSA procedure (Section 3. 46th Street. The columns are 345 MPa (50 ksi) steel wideflange sections.2). with the ninth level being the roof. this structure meets seismic code and represents typical medium-rise buildings designed for the Los Angeles. Los Angeles. and 37. with five bays each in the north-south (N-S) and east-west (E-W) directions. The levels of the 9-story building are numbered with respect to the ground level (see Fig. CA 90017. each determined by one pushover analysis.4. 13 . SAC is a joint venture of three non-profit organizations: The Structural Engineers Association of California (SEAOC).2. which from Eq. can be combined according to Eq. Richmond.19 m (122 ft) in elevation. rno .

for the first level is 1. the vibration periods are 2. monolithic column pieces are connected every two levels beginning with the first level. and seventh levels at 1. mechanical/electrical.65×105 kg (66.0 kips-sec2/ft). The column lines employ two-tier construction. partitions.01×106 kg (69. The column bases are modeled as pinned and secured to the ground (at the B-1 level). This model is based on centerline dimensions of the bare frame in which beams and columns extend from centerline to centerline.4.3.of-beam) are 3. The simple model adopted here is sufficient for the objectives of this study. 0. are located on the first. we select a weak ground motion: the northsouth component of the El Centro (1940) ground motion scaled down by a factor of 4.27. The 9-story N-S MRF is depicted in Fig. and 0. for the second through eighth levels is 9. 3.65 m (12 ft) and for the first floor is 5. 3.2 kips-sec2/ft). 1988) using the M1 model developed by Krawinkler and Gupta (1998). roofing and a penthouse located on the roof.7 kips-sec2/ft).e. To ensure that this structure remains elastic. The seismic mass of the ground level is 9. 3. The strength. Column splices.0 kips-sec2/ft). The floor-to-floor height of the basement level is 3. The force distributions.49 m (18 ft). dimension.2. 3. which are seismic (tension) splices to carry bending and uplift forces. and shear distortion of panel zones are neglected but large deformation (P∆) effects are included. floor slabs.85. i. The floor system is composed of 248 MPa (36 ksi) steel wide-flange beams in acting composite action with the floor slab..00×106 kg (616 kips. third. if desired more complex models. such as those described in Gupta and Krawinkler (1999) can be used.49 sec.sec2/ft). including the steel framing. ceiling/flooring. These force n distributions will be used in the pushover analysis to be presented later. The first three vibration modes and periods of the building for linearly elastic vibration are shown in Fig. fifth. and for the ninth level is 1. respectively.83 m (6 ft) above the center-line of the beam to column joint.20). The seismic mass of the structure is due to various components of the structure.07×106 kg (73. s* (Eq. for the first three modes are shown in Fig. Each frame resists one half of the seismic mass associated with the entire structure.89×105 kg (67. The seismic mass of the above ground levels of the entire structure is 9. Concrete foundation walls and surrounding soil are assumed to restrain the structure at the ground level from horizontal displacement. 14 . The building is modeled in DRAIN-2DX (Allahabadi and Powell.96 m (13 ft).

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

31 −0.8 −2. 2.728 2.67 −1.04 1. 3.1 −2.93 −1.13 −1. 2.03 −1. n = 1.5 Mode Shape Component 1 1.9th 8th 7th 6th T = 0.12 0.39 3. (b) shows the base shear Vbn t normalized relative to the weight W of the building.61 2. First three natural-vibration periods and modes of the 9story building 3.4.33 2. and 3.27 sec Ground −1.05 1. Each figure is organized in four parts: (a) shows the roof displacement urn t .6.05 s * 1 s * 2 s * 3 Fig.7. 3. n = 1.49 sec 3 T = 0.38 0. is shown in Figs.75 1. 3.3.05 2.796 0.94 2.2 Response History Analysis The structural response due to individual vibration modes. and 3 n 3.13)].51 0. (c) shows the joint rotation q n t of an external joint at 16 bg bg bg . Force distributions s* = mφn .85 sec 2 Floor 5th 4th 3rd 2nd 1st T1 = 2. respectively.0272 −2.5 −1 −0.72 −2.5 Fig.37 2.4.12) and (3.5 0 0.44 1.487 −1.1 3. (3.5. and 3 . determined by RHA [Eqs. 3.

8.422 cm. As expected. (3. For a fixed number of modes included. ur 2o = 2. and 3. including one.2.1.1. and 3. but three modes—perhaps even two modes—are sufficient. 3. This is illustrated in Fig. three. which are listed in Tables 3. especially in estimating the story drifts.16)]. two. The linear relationship between the base shear and roof displacement for each mode implies that the structure did not yield. larger in story drifts. or all modes.3.the roof level. and three vibration modes. the exact response considering all modes. errors are smallest in floor displacements. where the second and third modal responses are a larger percentage of the top story drift compared to roof displacement. errors generally decrease as response contributions of more modes are included. drifts in all stories.3. and rotations of external joints with moment connections are presented in Tables 3. 17 . in particular. Combining the modal response histories for all modes gives the total response [Eqs. and (d) shows the Vbn . the results for the roof displacement and top-story drift are shown in Fig.urn relation. respectively. 3. and ur 3o = 0. The same method was used to determine the peak values of many response quantities. and the percentage errors due to truncation of higher modal contributions.15) and (3. with the first three modes provide essentially the exact response. 3. two. It is apparent that the first mode alone is inadequate. Also included are the combined response due to one. are presented in Fig.9. The peak values of the various response quantities are noted in these figures. The peak values of floor displacements and story drifts determined by RHA. consistent with the increasing significance of the higher mode response among these three sets of response quantities.12 cm. The errors due to truncating the contributions of vibration modes beyond the third mode are negligible. The peak values of displacements of all floors. and even larger in joint rotations. respectively. 3. 3.23 cm.2. the peak roof displacement due to each of three modes is ur1o = 9.8.

5 0.090 0.008 -0.080 0.152 0.009 -0.0 -0.44E-03 1.1 -2.9 2.307 0.202 0.9 -16.25 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 3 Modal Response Mode 1 2.5 -1.76E-03 1.03E-03 1.88E-03 2.25 × 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 0.7 4.253 0.413 RHA (all modes) 0.227 0.22E-03 2.192 0.253 0.237 0.310 0.1 Building height is measured from the ground floor to the 9th floor.097 0.055 0.9 8.231 -0.062 -0.260 0.3 -3.3 -33.258 0.78E-04 -3.124 0.4 -22.1 -11.152 0.156 0.159 0.229 0.9 -24.3 -0.2 -2.99E-03 Mode 3 3.8 1.89E-03 1.121 0.4 -3.5 18.6 4.060 0.4 0.259 0.4 -6.177 0.275 0.9 3.062 0.045 0.042 0.006 0.060 -0.266 0.0 -46.65E-03 RHA (all modes) 2.023 0.311 0.Table 3.237 0.8 1.4 -1.91E-04 1.333 0.26E-04 -5.011 0.263 0.069 0.035 0.2 1.325 0.89E-03 1.45E-03 3.203 0.453 0.14E-03 2.24E-03 2.069 0.399 0.31E-03 Mode 2 -1.370 0.407 0.264 Error (%) 1 Mode 2 Modes 3 Modes -23.09E-03 1.4 -41.00E-03 1.226 0.11E-04 -5.002 -0.259 0.336 0.72E-03 3.01E-04 -2.300 0.01E-04 3.001 -0.44E-03 3.227 0.058 -0.1 Peak values of floor displacements (as % of building height = 37.205 0.226 0.50E-03 2.28E-04 1.13E-03 2.76E-03 1.010 -0.0 7.4 -7.202 0.9 9.0 -2.3 Peak values of joint rotations (radians) from RHA for 0.378 0.097 0.2 -57.1 3.156 0.74E-04 6.6 0.73E-03 3.3 -8.00E-03 2.042 0.9 -15.6 -1.117 0.8 -10.406 0.63E-03 2.038 0.1 -2.9 -22.8 -15.097 0.15E-03 4.3 -0.012 -0.38E-03 2.74E-03 1.14 m3) from RHA for 0.071 0.03E-03 -6.4 -1.13E-04 9.3 Table 3.401 Error (%) 1 Mode 2 Modes 3 Modes -23.6 11.260 0.09E-03 Error (%) 1 Mode -23.7 2.199 0.6 -0.282 0.2 -1.466 0.303 0.33E-03 2.321 0.7 2 Modes -3.088 -0.7 -19.2 -20.99E-03 2.63E-03 2.295 -0.1 -2.157 0.2 0.6 -1.225 0.183 0.229 0.266 0.2 Peak values of story drift ratios (as % of story height) from RHA for 0.29E-03 2.054 0.0 -2.235 0.6 0.7 3. 18 .350 0.202 0.125 0.173 0.8 -1.88E-03 2.311 0.130 0.38E-04 2 Modes 2.026 0.94E-03 2.50E-03 4.56E-03 2.1 4.65E-03 2.42E-04 1.00E-03 1.4 -10.475 0.266 0.08E-03 2.012 0.177 0.179 0.11E-03 1.1 -14.74E-03 1.9 1.173 0.364 0.64E-03 3.1 -0.133 -0.022 0.7 -50.8 -56.3 19.265 RHA (all modes) 0.125 0.9 -23.03E-03 1.181 0.6 1.245 0.197 0.032 -0.1 -19.089 0.0 -10.235 0.261 -0.2 9.6 9.003 0.09E-03 1.47E-03 1.4 -53.5 -2.4 -0.66E-05 -3.015 -0.1 Table 3.282 0.7 7.8 -5.25 × 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 0.400 0.09E-03 2.31E-03 Joint Rotation (rad) Combined (RHA) 1 Mode 2.0 3 Modes -5.011 0.03E-03 3 Modes 2.90E-03 3.265 0.85E-03 3.74E-04 9.124 0.024 -0.060 -0.2 -4.245 0.043 0.

Response due to first mode: (a) roof displacement.5 15 −0.12 −0.0435 • 0 (b) V b1 −0. (c) joint rotation.12 • 0 (a) r1 −15 0. and (e) pushover curve.1 0.0435 9.00174 • 0 (c) r1 −5 0 5 10 15 Time (sec) 20 25 30 0.1 0.25 × El Centro ground motion 19 .1 5 x 10 −3 θ (rad) 0. 3.1 /W 0.1 −15 −7. (b) base shear.15 u (cm) 9. Excitation is 0.5 0 u (cm) r1 7.5.12 0 (d) 0 (e) 15 Fig. (d) force-deformation history.5 0 u (cm) r1 7.5 9.0435 Vb1 / W 0.1 −15 −7.

1 0.1 −5 −2.5 5 Fig. (c) joint rotation.1 −5 −2. (d) force-deformation history. 3.5 5 −0. and (e) pushover curve.0252 −2. Response due to second mode: (a) roof displacement.1 Vb2 / W 0.23 −5 (a) 0.6.05 5 x 10 −3 b2 θ (rad) 0.05 /W 0.23 0.23 −2. Excitation is 0.5 0 u (cm) r2 2.5 0 u (cm) r2 2. (b) base shear.00244 • 0 (c) r2 −5 0 5 10 15 Time (sec) 20 25 30 0.5 u (cm) r2 0 • 2.0252 • 0 (b) V −0.25 × El Centro ground motion 20 .0252 (d) 0 0 (e) −0.

0108 −0.02 −0.5 1 Fig.02 −1 −0.02 /W V b3 0 • 0. Excitation is 0.000913 • 0 (c) r3 −2 0 5 10 15 Time (sec) 20 25 30 0.422 (e) −0.5 0 u (cm) r3 0.02 2 θ (rad) 0. (d) force-deformation history. (c) joint rotation.02 −1 −0.5 1 −0.0108 −0. Response due to third mode: (a) roof displacement.422 −1 (a) 0.25 × El Centro ground motion 21 . and (e) pushover curve. (b) base shear.02 0. 3.5 0 u (cm) r2 0.7.422 /W V b3 0 (d) 0 −0.1 u (cm) r3 0 • 0.0108 x 10 −3 (b) −0.

3 Modal Pushover Analysis Implementing MPA for the fundamental vibration mode.1.10 and Tables 3.18).8) leads to the pushover curve shown in Fig. Figure 3. 3.6e and 3.3).6. story drifts.3). This pushover curve is consistent with the relationship between the base shear and roof displacement determined by RHA (Fig. 3. 3. and external joint rotations in Tables 3.3. respectively. and 3. two.7d). respectively. Observe that the target roof displacement in each pushover analysis is identical to its exact value determined by RHA. The floor displacements.20) with n = 2 and 3 up to roof displacements ur 2 o = 2.2.6 present estimates of the combined response according to Eq. implying that the modal truncation errors are small if at 22 . 3. and external joint rotations computed by pushover analysis are presented in Tables 3.4) to roof displacement ur1o = 9. using the force distribution of Eq.12).9). In practical application. story drifts. Implementing pushover analysis for the second and third modes. and 3. pushing the structure. (3. or three vibration modes. pushing the structure using the force distribution of Eq. 3.9 and Tables 3. and the errors in these estimates relative to the exact response from RHA considering all modes.4.5.1. the floor displacements are proportional to the mode shape φ 1 because the structure remains elastic. 3.7e and to the floor displacements. although both analyses led to identical peak values of the individual modal responses. 3. respectively. (3. the value determined by RHA (Fig.12 cm. except for the algebraic sign associated with Γ1 and minor round-off errors. As for the first mode.5. and ur 3o = 0. These values of the response quantities are identical to the peak response values determined from RHA (Tables 3. In RHA the errors arise only from truncating the responses due to higher modes.422 cm . 3. 3. confirming that MPA gives the exact values of the individual modal responses.. considering one.. As suggested by Eq. (3. and 3.4. and it is apparent in the example considered that three modes provide most of the response (Fig.e. leads to the pushover curves shown in Figs.6. i. 3. 3.1 through 3.2.e. i. and the computed response values are identical to the peak response values determined from RHA (Tables 3. and 3. the errors in the MPA results are generally larger than in RHA (Fig.ur relations determined by RHA (Figs.6d and 3. and 3. which would provide the Dn value to be substituted in Eq.5.5e. (3. this value would be determined directly from the response (or design) spectrum.20) with n = 1 (Fig. 3. these pushover curves are consistent with the Vb .4.5d). (3.21).4. 3.3). For a fixed number of modes included considered.23 cm .

Heightwise variation of floor displacements and story drifts from RHA for 0.eps Ground 0 0. 3.422 r2 0 • 1.2 0.12 • r1 −15 15 0 r1 0 0 0.eps r 5 10 15 20 Time.5 Ground 0 0. sec 25 30 Fig.5 0.685 • −3 3 Mode 2 ∆ (cm) u (cm) • 2.least three modes are included.4 Story Drift Ratio (%) 0. Additional errors are introduced in pushover analysis due to the approximation inherent in modal combination rules. (a) Roof Displacement 15 (b) Top Story Drift 3 Mode 1 ∆ (cm) u (cm) 9.9.2 0.48 −3 3 1.eps fig3_9b.3 0.83 • r3 0 • 0.1 0. Response histories of roof displacement and top-story drift from RHA for 0.4 Displacement/Height (%) 0.59 • Total (All Modes) 0 −15 0 ∆ (cm) u (cm) 0 −3 0 fig3_8.25 × 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.3 0.23 r2 −15 15 0 • 0.8.25 × El Centro ground motion 23 .1 0.03 −3 3 Mode 3 ∆ (cm) u (cm) r3 −15 15 9. sec 25 30 r 5 10 15 20 Time. 3.6 Fig.

229 0.2 Table 3.042 0.89E-03 1.7 -21.9 -15.276 0.53E-04 -9.203 0.267 0.05E-03 3.00E-03 2.286 0.1 -11.Table 3.9 -14.259 -0.74E-04 -6.261 -0.3 -41.235 0.09E-04 -3.002 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 -0.28E-03 2.9 -13.09E-03 -1.4 -19.8 -15.133 0.058 -0.09E-03 Mode 2 1.31E-03 2.1 -11.2 -16.65E-03 2.9 3 Modes -12.048 0.04E-03 3.8 -14.048 0.321 0.0 -2.385 0.43E-04 -1.00E-03 1.80E-04 3.63E-03 -2.0 -46.060 -0.117 0.006 -0.080 0.03E-03 1.4 -11.22E-04 2.008 -0.235 -0.071 0.106 0.156 0.73E-05 3.227 0.9 -15.9 -15.4 -7.024 0.3 -13.2 1.260 0.74E-03 -1.9 -13.125 0.08E-03 2.237 -0.407 0.270 0.036 -0.062 -0.088 -0.203 0.03E-03 3.9 -12.03E-03 -1.2 -20.274 0.106 0.33E-04 5.4 -4.023 -0.12E-03 1.015 -0.003 -0.181 0.63E-03 2.011 -0.6 -15.89E-03 2.226 0.09E-03 1.133 0.9 -18.92E-04 -1.229 0.121 0.00E-03 3.245 0.042 0.156 0.89E-03 -2.177 0.9 -8.31E-03 Error (%) RHA (all modes) 2.079 0.00E-03 2.4 -14.3 Table 3.270 0.9 -15.89E-03 2.157 0.2 -57.267 -0.022 0.1 -0.090 0.40E-04 5.179 0.282 0.152 0.94E-03 2.44E-03 3.1 -18.9 -24.331 0.3 -19.069 0.203 0.055 0.001 0.270 0.260 0.2 -0.9 -11.012 0.322 0.285 0.157 0.097 0.332 0.57E-03 1 Mode -23.7 2.14 m) from MPA for 0.8 -56.125 0.00E-03 -1.9 -13.253 -0.22E-03 -2.038 -0.00E-03 2.177 0.2 -11.15E-03 1.7 -15.03E-03 6.74E-03 1.90E-03 1.078 0.272 0.173 0.203 0.6 -17.7 24 .4 -9.253 RHA (All Modes) 0.42E-04 -1.011 0.72E-03 3.4 -4.374 0.8 -23.73E-03 3.3 -9.267 0.4 1.237 0.4 -9.313 0.3 -33.264 Error (%) Modal Response 1 Mode 2 Modes 3 Modes -23.401 Error (%) 1 Mode 2 Modes 3 Modes -23.0 -10.4 -2.203 0.010 0.097 0.2 -12.069 0.179 0.266 0.336 0.9 -14.9 -16.336 RHA (all modes) 0.043 -0.4 0.0 -16.133 -0.009 0.032 0.012 -0.231 -0.124 0.76E-03 1.99E-03 Mode 3 -3.96E-03 2.026 0.230 0.197 0.4 -22.38E-03 2.31E-03 Joint Rotation (rad) Combined (MPA) 1 Mode 2.062 0.6 Peak values of joint rotation (radians) from MPA for 0.7 -19.3 -14.38E-03 3 Modes 2.152 0.045 -0.282 -0.7 -50.3 -12.25 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Modal Response Mode 1 -2.5 -16.9 -14.97E-03 1.259 0.3 -2.4 Peak values of floor displacements (as % of building height = 37.296 -0.7 2 Modes -13.89E-03 -1.5 Peak values of story drift ratios (as % of story height) from MPA for 0.8 -22.253 0.4 -5.300 0.0 -18.0 -0.328 0.466 0.44E-03 -1.227 0.173 0.370 0.6 -19.3 1.310 0.9 -13.060 0.230 -0.24E-03 -2.253 0.4 -22.15E-03 2.90E-04 -9.4 -53.245 -0.6 -16.1 -19.2 -4.47E-03 1.78E-04 2 Modes 2.25 × 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 0.76E-03 -1.8 -15.260 -0.

Heightwise variation of floor displacements and story drifts from MPA for 0.5 Ground 0 0. 3.5 0.4 Displacement/Height (%) 0.4 Story Drift Ratio (%) 0.(a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th (b) Story Drift Ratios Floor 4th 3rd 2nd 1st Floor 5th RHA All Modes MPA 1 Mode 2 Modes 3 Modes 5th 4th 3rd 2nd 1st RHA All Modes MPA 1 Mode 2 Modes 3 Modes fig3_10a.1 0.eps fig3_10b.1 0.3 0.25 × El Centro ground motion.2 0.10.2 0.eps Ground 0 0. shading indicates modal combination error 25 .3 0.6 Fig.

26 .

1) becomes mu + cu + f s u.2) to the modal coordinates of the corresponding linear system.3) . 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 (4.1) With this generalization for inelastic systems. Thus.1) is not valid for inelastic systems.mι ug t b (4.2) The standard approach is to solve directly these coupled equations.1 Inelastic Multistory Buildings RESPONSE HISTORY ANALYSIS For each structural element of a building. Eq. (4. the natural vibration periods and modes of the corresponding linear system are the same as the vibration properties of the inelastic system undergoing small oscillations (within the linear range).4 4. (3. sign u b g g bg (4. Each structural element of this elastic system is defined to have the same stiffness as the initial stiffness of the structural element of the inelastic system. Although classical modal analysis (Section 3. Therefore. but depend on the history of the displacements: f s = f s u. the initial loading curve is idealized as bilinear. Both systems have the same mass and damping. leading to the “exact” nonlinear response history analysis (RHA). sign u = . the relations between lateral forces f s at the N floor levels and the lateral displacements u are not single valued. it is useful for later reference to transform Eq. and the unloading and reloading curves differ from the initial loading branch.

” implying that other “modes” will also contribute to the solution.2).3).3) will. 2.and classical n damping-orthogonality property of modes gives F qn + 2z nw n qn + sn = .2 UNCOUPLED MODAL RESPONSE HISTORY ANALYSIS bg Neglecting the coupling of the N equations in modal coordinates [Eq. sign qn = φ n f s u n . sign u = .4)] leads to the uncoupled modal response history analysis (UMRHA) procedure. 4. (4. however. qr t = 0 for all 28 bg . The equations governing the response of the inelastic system to p eff.Gn ug t Mn bg n = 1. (4. (4.9) for linearly elastic systems. (3. The spatial distribution s of the effective earthquake forces is expanded into the modal contributions sn according to Eq.n t given by Eq. Eq. This approximate RHA procedure is the preliminary step in developing a modal pushover analysis procedure for inelastic systems.2).6) for inelastic systems will no longer be described by Eq. in principle.4) is rarely solved because it offers no particular advantage over Eq. sign u n b g b g bg (4. (4.6) The solution of Eq.… N (4.4) represents N equations in the modal coordinates qn .5) This resisting force depends on all modal coordinates qn t .sn ug t bg b g bg bg (4. (4.9) involves T Fsn = Fsn qn . (4.Substituting Eq.3) in Eq. (3. (3. and using the mass. where φ n are now the modes of the corresponding linear system. Simultaneously solving these coupled equations and using Eq. premultiplying by φ T . Equation (4.6b) are mu + cu + f s u. these equations are coupled for inelastic systems. (4.8) because qr t will generally be nonzero for “modes” other than the nth “mode. implying coupling of modal coordinates because of yielding of the structure. However.4) where the only term that differs from Eq. give the same results for u t as obtained directly from Eq. (3. (4. (3. For linear systems. Unlike Eq.2).

(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 .4. Fsn / Ln ¨ g(t ) u (a) Static Analysis of Structure (b) Dynamic Analysis of Inelastic SDF System Fig.3. 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. ζn.89 50 r1 0 u (cm) 0 • 78. 2. n = 1. 4.n t = .02 −150 n=2 −150 50 n=2 14.0 ¥ El Centro ground motion: (a) exact solution by NL-RHA.s n u g t .Forces sn Unit mass An(t ) st rn ωn.575 5 10 15 20 Time (sec) 25 30 ur3 (cm) 0 • 5.112 5 10 15 20 Time (sec) 25 30 −10 0 −10 0 Fig. and 3. 4.9 • 0 −50 10 −50 10 n=3 ur3 (cm) 0 • 5.51 • ur2 (cm) 0 ur2 (cm) n=3 14. where u g t = 3. Roof displacement due to p eff.

008 • 6.5)] is neglected. and (b) approximate solution by UMRHA af bg bg 4.2… N ) according to Eq.n t (n = 1. (4.4 and 4. 4. 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. and (3) the Fsn Ln . this intense excitation was chosen to ensure that the structure is excited well beyond its linear elastic limit. 2.956 5 10 15 20 Time (sec) 25 30 ∆r3 (cm) 0 • 5.6) by UMRHA is compared with the “exact” solution by nonlinear RHA.(a) Nonlinear RHA 20 n=1 20 (b) UMRHA n=1 • 5. the errors in either response quantity are only a few percent.855 ∆r1 (cm) 0 ∆r1 (cm) n=2 • 6.817 5 10 15 20 Time (sec) 25 30 −10 0 −10 0 Fig.15) is strictly valid only for linearly elastic systems.s n u g t .0 ¥ El Centro ground motion.Dn relation is 33 bg bg .1 Underlying Assumptions and Accuracy The approximate solution of Eq. 4. Top story drift due to p eff. (2) the superposition of responses to p eff.2. Such comparison for roof displacement and top-story drift is presented in Figs.5.0 ¥ El Centro ground motion: (a) exact solution by NL-RHA.5. respectively. (3.33 0 −20 20 −20 20 n=2 ∆r2 (cm) 0 ∆r2 (cm) 7. (4.4) and (4. The errors are slightly larger in drift than in displacement. n = 1. both for 3. where u g t = 3. but even for this very intense excitation.n t = .744 • 0 −20 10 n=3 −20 10 n=3 ∆r3 (cm) 0 • 5. and 3.

10) . when specialized for linearly elastic systems it is identical to the RHA procedure of Section 3.9) should be determined by nonlinear static analysis of the structure as the structure undergoes displacements u = Dnφ n with increasing Dn .8) governing Dn t is based on Eq.urn pushover curve to the Fsn Ln .1. and the base shear Vbn .2.12) for floor displacements. 4. (3. (3. Although several approximations are inherent in this UMRHA procedure. Therefore we impose this constraint in developing the UMRHA procedure in this section and modal pushover analysis in the next section. (3. within the linearly elastic range of the structure. (4. When implemented by commercially available software. 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.20) to some bg pre-determined roof displacement. The overall errors in the UMRHA procedure are documented in the examples presented in Section 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. 4.8) before it can be solved? Because Eq.6a. (4. However.2 Properties of the nth-“mode” Inelastic SDF System How is the Fsn Ln .20).4. At the yield point. the only force distribution that produces displacements proportional to f n is given by Eq. The structure is pushed using the force distribution of Eq. is plotted against roof displacement urn . How to convert this Vbn . it can conduct a force-controlled nonlinear static analysis with an invariant distribution of lateral forces. (4. (4. such nonlinear static analysis provides the so-called pushover curve. (4. the base shear is Vbny and roof displacement is urny .9).approximated by a bilinear curve to facilitate solution of Eq.8) in UMRHA. the relationship between lateral forces f s and Dn in Eq.Dn relation to be determined in Eq. Therefore. A bilinear idealization of this pushover curve for the nth-“mode” is shown in Fig. this distribution seems to be a rational choiceeven for inelastic systemsto determine Fsn in Eq. Although most commercially available software cannot implement such displacement-controlled analysis. which is different than the Fsn Ln .Dn curve.

4. which is not a meaningful quantity.13) This value of Tn . (4.11).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.6a is 2 kn = ω n Ln .6b. Properties of the nth-“mode” inelastic SDF system from the pushover curve Equation 4. the elastic vibration period Tn of the nth-mode SDF system is computed from F Ln Dny I 1/ 2 Tn = 2p G H Fsny JK (4.2. 4.6. (4. the initial slope of the pushover curve in Fig. which may differ from the period of the corresponding linear system.6b is w 2 .5). The two are related through Fsny Ln 2 = w n Dny (4. 4. Knowing Fsny Ln and Dny from n Eq.12) implying that the initial slope of the curve in Fig. 4. 2001. where the yield values of Fsn Ln and Dn are Fsny Ln = Vbny * Mn Dny = urny Γ nf rn (4.8). Section 13. should be used in Eq. 35 .10 enables conversion of the pushover curve to the desired Fsn Ln − Dn relation shown in Fig.11) * in which M n = Ln Γn is the effective modal mass (Chopra. In contrast.

details are available in Appendix A: 1. Typically.3b) with force-deformation relation of Fig. of the nth-“mode” inelastic SDF system (Fig. is now determined by 36 bg bg . For the nth-mode. 4.6b.13). This value of the roof displacement is given by Eq. Dn (t ) . 7. 4. ωn . 5. n (3.n t .2. 4.16) to determine the total response. (3.4. Compute the natural frequencies.15) and (3. φn . develop the base-shear – roof-displacement ( Vbn − urn ) pushover curve for the force distribution s* [Eq. Compute the deformation history. 4. (3. 4. 2. the peak value of Dn t . (3. Calculate histories of various responses by Eqs.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. of the total response r t obtained in Step 8. 4. 8. for linearly-elastic vibration of the building. Convert the idealized pushover curve to the Fsn Ln − Dn relation (Fig. (4.11). (3. which is summarized next as a sequence of steps. Combine the “modal” responses using Eqs. Repeat Steps 2 to 6 for as many modes as required for sufficient accuracy.6b) by utilizing Eq.3 MODAL PUSHOVER ANALYSIS bg bg A pushover analysis procedure is presented next to estimate the peak response rno of the inelastic MDF system to effective earthquake forces p eff . An t .20)]. with the structure is pushed to the roof displacement urno .12) and (3.6a). r o . n 3.20)].21) where Dn . Calculate the peak value. the first two or three modes will suffice. Idealize the pushover curve as a bilinear curve (Fig. 6. Consider a nonlinear static analysis of the structure subjected to lateral forces distributed over the building height according to s* [Eq. 9. and pseudo-acceleration history. and modes.

The peak “modal” responses rno . This application of modal combination rules to inelastic systems obviously lacks a theoretical basis. Sections 7. are combined using an appropriate modal combination rule.2 and 3. 6. joint rotations.8).18).2.3) because. 4. 37 . (4. story drifts. as described in Section 4.1 Summary The peak inelastic response of a building to earthquake excitation can be estimated by the MPA procedure just developed. each determined by one pushover analysis.12). or from the inelastic response (or design) spectrum.4. it can be determined from the inelastic response (or design) spectrum (Chopra. Dn . Compute the peak deformation.2.solving Eq. Calculate the peak roof displacement urno associated with the nth-“mode” inelastic SDF system from Eq. rno also represents the exact peak value of the nth-mode contribution rn t to response r t . alternatively.6). for elastic systems. It provides results for elastic buildings that are identical to the well-known RSA procedure (Section 3.3. seems reasonable. governed by Eq. etc. the lateral force distribution used possesses two properties: (1) it appears to be the most rational choice among all invariant distribution of forces. (3.21). 4. plastic hinge rotations. Steps 1 to 4 of the MPA are same as those for UMRHA. details are available in Appendix B.n t . 5. The response value rno is an estimate of the peak value of the response of the inelastic system to p eff. As shown in Sections 3. which is summarized next as a sequence of steps.g. e. and (2) it provides the exact modal response for elastic systems.8)..3.6b by solving Eq. Eq. At this roof displacement. although somewhat intuitive for inelastic buildings. of the nth-“mode” inelastic SDF system (Fig. the pushover analysis provides an estimate of the peak value rno of any response rn t : floor displacements. (4.6 and 7. bg bg bg bg Thus. This pushover analysis. as mentioned earlier. it seems reasonable because it provides results for elastic buildings that are identical to the well-known RSA procedure described in Section 3. we will refer to rno as the peak “modal” response even in the case of inelastic systems. However. to obtain an estimate of the peak value ro of the total response. (3. 2001. 4.3b) with force-deformation relation of Fig. (4.

Observe that errors tend to decrease as response contributions of more “modes” are included. the 38 . the combined response due to three “modes”.2.9. At urno . the UMRHA procedure lacks a rigorous theory. two. Determine the total response by combining the peak “modal” responses using the SRSS combination rule of Eq. although the trends are not systematic as when the system remained elastic (Section 3.0. say. From the total rotation of a plastic hinge.2).18). and ur 3o = 2. respectively. is presented next. the first two or three “modes” will suffice. (3.1). This deficiency also implies that. 9.5 ¥ the El Centro ground motion including the response contributions associated with three “modal” inelastic SDF systems. ur 2o = 11.4 COMPARATIVE EVALUATION OF ANALYSIS PROCEDURES The response of the 9-story building described earlier is determined by the two approximate methods: UMRHA and MPA.0 to 3. Figure 4.1 and 4. in particular. This is to be expected. three “modes” included.8. and the percentage errors in the approximate results. rno . 4. in contrast to modal analysis (Section 3.4. also included are the combined responses due to one. 4. 8.3 cm. The peak values of response are as noted.1 Uncoupled Modal Response History Analysis The structural response to 1. the “exact” results. and three “modes”.7 shows the individual “modal” responses.7 cm. 4. 4. The peak values of displacements of all floors and drifts in all stories are presented in Tables 4. and three modes are compared with the “exact” values in Fig. with. subtract the yield value of hinge rotation to determine the hinge plastic rotation. the peak roof displacement due to each of the three “modes” is ur1o = 48. Typically.53 cm. and the “exact” response from nonlinear RHA for the roof displacement and top-story drift. and compared with the results of a rigorous nonlinear RHA using the DRAIN-2DX computer program. determined by the UMRHA procedure. extract from the pushover database values of other desired responses. The peak values of floor displacements and story drifts including one. two. Repeat Steps 3 to 8 for as many “modes” as required for sufficient accuracy.7. and the errors in the approximate results are shown in Fig.4. 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.

1.75.2) if the system yields significantly versus if the system remains within the elastic range (Tables 3.62 5. 4.7.44 • "Mode" 2 r1 0 • 48.88 u (cm) "Mode" 3 −80 80 0 −80 80 0 −80 0 • 44. just as for elastic systems.0.1 and 3.2). for a fixed number of “modes” included. sec 25 30 r −12 0 Fig. 0. the errors in responses computed by UMRHA including three “modes” relative to the “exact” response were determined. Next.5 (Tables 4.0. Response histories of roof displacement and top-story drift due to 1.1 −12 12 ∆r (cm) u (cm) UMRHA 3 "Modes" 0 • 7.85. (a) Roof Displacement (b) Top Story Drift 80 12 0 • 3. and total response from NL-RHA 39 .5. 0.5.2) and 1. 0.25. we investigate how the errors in the UMRHA vary with the deformation demands imposed by the ground motion.2).24 • NL−RHA r −12 12 ∆r (cm) u (cm) 0 5 10 15 20 Time.25 (Tables 3. defined as the El Centro ground motion multiplied by 0.response is much less accurate (Tables 4.0.6 5 10 15 20 Time. the degree to which the system deforms beyond its elastic limit. the errors in story drifts are larger compared to floor displacements.1 and 4.53 ∆r3 (cm) 12 0 • 2. sec 25 30 • 48. For each excitation.7 • −80 80 ∆r1 (cm) u (cm) "Mode" 1 −12 r2 0 ∆r2 (cm) 12 0 u (cm) −80 80 −12 r3 0 • 2.38 6.3 11. in particular. and 3.5 × El Centro ground motion: individual “modal” responses and combined response from UMRHA.1 and 4. 1. For this purpose the UMRHA and exact analyses were repeated for ground motions of varying intensity.1 and 3. 2. However. recall that the computed errors have been presented earlier for ground motion multipliers 0.

8. For this purpose.20)] for the first three n modes of the system are shown in Fig.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. To interpret these results.5 ¥ El Centro ground motion Figure 4.5 1 1. in each story drift (Fig.10a).5 Displacement/Height (%) 2 Ground 0 0. and the error envelope for each case.11.(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. 4.5 Story Drift Ratio (%) 2 2. or three “modes” for 1.10b).5 Fig. Heightwise variation of error in floor displacements and story drifts estimated by UMRHA including one. Shown is the error in each floor displacement (Fig. with the peak displacement of each “modal” SDF system noted for each ground motion multiplier. Two versions of the pushover curve are 40 . Heightwise variation of floor displacements and story drift ratios from UMRHA and NL-RHA for 1. 4. two.5 1 1. the pushover curves using force distributions s* [Eq. indicated by a ground motion multiplier. 4. it will be useful to know the deformation of the system relative to its yield deformation.9.10 summarizes the error in UMRHA as a function of ground motion intensity. 4. (3. 4.

the values determined by RHA of the nth-mode inelastic SDF system (Fig. 2. (4.2 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 1.10 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. The location of plastic hinges and their rotations.20) with n = 1.11a). (3.10) for each “modal” inelastic SDF system (Figs 4.2. as mentioned in Section 4.53 cm. 4.7). The structure is pushed using the force distribution of Eq. First.3 cm. Perhaps this explains why the errors are large at this excitation intensity.4) to roof displacements urno = 48.1). and 3 (Fig. 4. UMRHA lacks a rigorous theory and is based on several approximations.7 cm. The system remains elastic up to ground motion multiplier 0. were noted but are not shown here. and plastic hinge rotations at the external beam end at each floor level (Table 4.6 and 4. 3.11). 4. the first reason mentioned above seems to be the primary source for the errors.75.3). 41 . Additional errors are introduced in UMRHA of systems responding beyond the linearly elastic limit for at least two reasons. For more intense excitations.4. and (4) are larger in story drift compared to floor displacements.5).4).included: the actual curve and its idealized bilinear version. Second. Each of these three pushover analyses provides the pushover curve (Fig.0.0. 4. respectively. 11. The idealized curve for the first “mode” deviates most from the actual curve near the peak displacement corresponding to ground motion multiplier 1. drifts in all stories (Table 4. the errors in truncating the higher mode contributions are negligible. determined from “exact” analyses. the peak values of displacements at all floors (Table 4. even though the system remains essentially elastic.75. the pushover curve for each “mode” is idealized by a bilinear curve in solving Eq.2. (3) maintain roughly similar values for more intense ground motions. and.01 (Fig. Figure 4.4.5 ¥ the El Centro ground motion. as mentioned in Section 3. the ductility factor for the first mode system is only 1. and 2. (2) increase rapidly as the ground motion multiplier increases to 1.

698 1.877 0.072 -1.668 -23.490 -1.5 28.856 2.256 1.938 1.3 8.5 10.484 0.0 -9.410 1.220 -0.863 1.049 -0.200 8.9 12.982 9.914 2.071 -0.900 -10.540 0.938 -1.9 16.003 -31.495 1.5 × El Centro ground motion Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Modal Response “Mode” 1 “Mode” 2 Drift Ratio (%) Combined (UMRHA) 1 “Mode” 2 “Modes” “Mode” 3 -1.0 2.2 1.154 0.5 2 GM Multiplier 2.379 1.0 7.983 1.852 1.6 1.130 0.9 12.2 12.5 × 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” -0.033 0.1 1.2 1.430 1.3 0.010 0.5 9.293 1.088 10.298 -0.5 18.9 31.(a) Floor Displacements 60 50 40 Error Envelope Error for Floor No.291 0.126 0.0 1.3 1.071 0.4 1.1 8. Noted 60 50 40 (b) Story Drifts Error Envelope Error for Story No.298 0.4 4.371 -0.0 11.410 -1.3 6. 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 0.820 -19.370 -0.772 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 1.6 4.009 -0.472 1.133 1. 4.057 -0.7 14.201 -0.10.8 1.256 -1.914 -0.055 -0.5 1 1.806 -0.4 -1.3 42 .065 0.616 -0.201 -1.942 -0.5 2 GM Multiplier 2.315 -0.214 0.241 1.676 0.8 1.707 1.072 1.5 -3.018 0.8 14.373 -0.6 2.727 1.214 -0.8 1.811 1.1 Peak values of floor displacements (as % of building height = 37.121 -0.333 0.8 0.366 0.104 0.372 -1.1 0.763 -15.202 11.169 0.722 0.317 0.2 6.5 9.096 0.5 3 Fig.5 28.120 1.2 Peak values of story drift ratios (as % of story height) from UMRHA for 1.070 1.7 31.473 -22.003 0.478 0.820 -0.751 1.5 1 1.513 0.8 1.372 1.407 -10.260 -15.945 -37.3 25.241 -1.942 1.135 9.575 -41.376 -1.216 1.7 Table 4.663 -0.819 2.5 3 0 0 0.863 0.4 -7.844 -25.806 0.8 1.235 -0.554 1.201 1.248 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.9 31.338 -1.079 0.0 9.366 -0.663 0.350 -0.426 -1. Errors in UMRHA as a function of ground motion intensity: (a) floor displacements.513 -0.220 0.526 -0.338 1.1 1.044 1.0 11.187 -0.226 -0.971 1.14 m) from UMRHA for 1.376 1.1 3.068 0.256 1.136 1.371 0.2 22.283 1.138 1. and (b) story drifts Table 4.490 1.2 4.9 5.

14. although the error is recorded as 100% if MPA estimates zero rotation when the nonlinear RHA computes a non-zero value.” two “modes. 4. Obviously.3 through 4. 4.11c).12 and 4.1 and 4. For example.5). significantly smaller than in UMRHA (compare Fig. The locations of plastic hinges shown in Fig.13 and Tables 4. The drifts are underestimated by up to 13% in the lower stories. but the third “mode” contributions do not seem especially important (Fig. 4.13 with Fig. the hinges identified in beams at the 6th floor were at variance with the “exact” results. The errors are especially large in the hinge plastic rotations estimated by the MPA procedure. the results were not always accurate.9 and Tables 4. 4. 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%. significant improvement is achieved by including response contributions due to the second “mode”.2). and are within a few percent of the exact values for the upper stories. two. 4.3 and 4. overestimated by up to 18% in the middle stories. 4. 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.4 with Tables 4. and the errors in these estimates relative to the exact response from nonlinear RHA. the third “mode” does not contribute because this SDF system remains elastic (Fig. MPA failed to identify the plastic hinges at the column bases in Fig. Observe that the primary contributor to plastic hinge rotations in the lower stories is the first “mode” and the second “mode” in the upper stories.” and three “modes. in general. 4. 4. (3.14.Figures 4. Furthermore.3 and 4.18).” and nonlinear RHA. the errors in the modal pushover results are.4. but was more successful when the excitation was more intense (results are not included). were determined by four analyses: MPA considering one “mode.13 and Table 4.13 and Tables 4.3 and 4. 4. Pushover analysis seems to be inherently limited in computing accurately hinge plastic rotations.12 and Tables 4.4).5 present estimates of the combined response according to Eq. even if three “modes” are included (Fig. The second “mode” was necessary to identify hinges in the upper stories. this error is not especially significant because the hinge plastic rotation is very small. considering one. 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.3 and 4. 43 . Fortuitously.4). The first “mode” alone is inadequate. especially in estimating the story drifts (Fig. However. As shown in Figs.12 and Tables 4. and three “modes.” respectively. however. for two or three modes included.

5. 2.13 y by Base Shear (kN) 8000 6000 1 1.25 1 2 4 6 Roof Displacement (cm) 8 10 Fig.9 cm. 4.0 × El Centro ground motion 44 .5 0.5 1 0.19 y by 3 2 1.5 2 3 Actual Idealized 2000 0 0 0. 1.5 0.14 y by Base Shear (kN) 8000 6000 4000 1. 0.85 0. α = 0.85 Base Shear (kN) 8000 6000 4000 2000 0 0 0.85 0.25 0.0.5. V = 7616 kN.75 0. “Modal” pushover curves with peak roof displacements identified for 0.5 0. V = 4952 kN.6 cm. α = 0. and 3. V = 5210 kN.75 Actual Idealized 20 40 60 Roof Displacement (cm) (b) "Mode" 2 Pushover Curve 80 12000 10000 u = 9.25 5 10 15 20 Roof Displacement (cm) 25 (c) "Mode" 3 Pushover Curve 12000 10000 u = 4.2 cm. α = 0.5 2 3 4000 2000 0 0 Actual Idealized 0.11.0.75 0.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36. 0. 1.25.75.

the MPA procedure is equivalent to the RSA procedure.5 Displacement/Height (%) 2 Ground 0 0. an observation with practical implications. 4.2 and 3. Shown is the error in each floor displacement (Fig. identified in Section 3.75.3 for elastic systems (or weak ground motions).5 Fig.4.10 and 4.15).0.12 Heightwise variation of floor displacements and story drift ratios from MPA and NL-RHA for 1.15b). 45 . As mentioned in Section 3.5 × El Centro ground motion. whereas MPA contains errors inherent in modal combination rules. and the error envelope for each case. shading indicates errors in MPA including three “modes” Figure 4. While various sources of errors in UMRHA. UMRHA is essentially exact. In this case. as discussed in Sections 3.15) for ground multipliers larger than 1. the errors in MPA were larger for ground motion multipliers less than 0. each story drift (Fig.15 summarizes the error in MPA considering three “modes” as a function of ground motion intensity. implying excitations intense enough to cause significant yielding of the structure. 4.3.5 1 1.15a). now standard in engineering practice.5 Story Drift Ratio (%) 2 2. 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. 4. The errors are only weakly dependent on ground motion intensity (Fig.5 1 1. 4. However. implying excitations weak enough to limit the response in the elastic range of the structure.4. implying that the modal combination errors contained in these procedures are acceptable. the errors in MPA were fortuitously smaller than in UMRHA (compare Figs.4 also apply to MPA. 4. indicated by a ground motion multiplier.(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.

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

338 1.498 1.10E-02 9.36E-03 6.2 -3.071 -0.980 -0.820 -7.2 -100.8 -6.942 6.5 × El Centro ground motion Hinge Plastic Rotation (rad) “Modal” Response Combined (MPA) “Mode” 1 7.0 15.101 -0.168 -0.687 0.6 1.9 2.8 17.8 “Mode” 3 -1.116 1.0 3 “Modes” -32.614 0.652 1.517 1.018 -0.1 18.666 Table 4.503 1.641 1.640 -1.154 0.5 2.066 -0.668 -13.2 0.00E+00 0.473 -15.7 -12.667 -1.116 1.76E-03 4.756 0.8 1.200 8.336 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.895 1.1 62.752 1.2 6.18E-03 7.5 -6.37E-04 Error (%) “Modal” Response 1 “Mode” -32.37E-03 1.4 -8.244 0.8 7.895 1.76E-03 4.00E+00 2 “Modes” 7.105 0.018 0.728 1.1 13.003 -16.1 46.331 1.514 -1.0 1.5 10.5 1.02E-03 3.37E-03 1.00E+00 0.1 62.407 -27.298 0.00E+00 3.76E-03 4.9 -4.197 -0.00E+00 “Mode” 3 0.426 15.135 -7.1 62.705 1.60E-04 7.581 0.8 1.02E-03 0.429 0.220 1.88E-03 0.118 0.5 Peak values of hinge plastic rotations (radians) from MPA for 1.5 × El Centro ground motion Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th “Modal” Response “Mode” 1 “Mode” 2 Drift Ratio (%) Combined (MPA) 1 “Mode” 2 “Modes” 3 “Modes” Error (%) NL 1 2 3 RHA “Mode” “Modes” “Modes” 1.311 0.015 0.304 -1.804 1.36E-03 6.266 -0.8 -6.00E+00 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th “Mode” 2 0.202 8.8 0.259 1.518 1.683 1.781 0.372 0.3 -3.4 Peak values of story drift ratios (as % of story height) from MPA for 1.305 -0.8 -29.527 -0.22E-10 NL RHA 1.00E+00 0.8 -29.00E+00 0.910 1.315 0.88E-03 0.7 1.399 0.5 2.00E+00 0.9 1.667 1.233 1.1 -8.00E+00 0.9 1.0 -50.781 0.088 12.733 1.304 1.37E-03 1.6 13.6 -44.9 0.260 -14.8 -12.125 -1.55E-03 3.8 -29.53E-03 7.9 -100.478 0.72E-03 7.233 1.250 0.581 0.298 -0.00E+00 0.053 1.435 0.6 -9.3 11.0 -100.705 -1.6 -7.00E+00 1 “Mode” 7.7 1.611 0.55E-03 3.76E-03 4.3 -3.911 0.0 -100.5 × 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 0.00E+00 3 “Modes” 7.763 -14.1 1.071 0.2 -100.00E+00 0.02E-03 3.0 47 .19E-10 3.50E-10 3.012 1.5 7.6 7.176 0.503 -1.594 -1.6 -8.88E-03 1.9 5.049 -0.516 0.0 -50.0 2 “Modes” -32.156 -0.209 1.26E-04 9.033 -0.351 -0.222 0.00E+00 0.14 m) from MPA for 1.72E-03 7.399 0.00E+00 0.99E-03 6.60E-03 2.00E+00 0.1 46.5 10.116 1.37E-03 1.2 1.00E+00 0.371 -0.5 2.00E+00 0.0 -5.756 0.55E-03 3.190 -0.9 -100.130 0.015 0.6 13.00E+00 0.9 7.737 1.5 Table 4.066 -0.055 0.945 -49.900 -0.2 0.009 0.72E-03 7.36E-03 6.018 -0.00E+00 0.007 1.2 -4.745 1.879 1.694 1.0 -100.2 -100.00E+00 0.844 -7.8 -4.220 1.00E+00 0.737 0.068 0.3 13.575 -53.6 -44.414 1.222 0.3 Peak values of floor displacements (as % of building height = 37.980 0.Table 4.36E-03 6.00E+00 0.02E-03 0.057 0.2 11.982 13.738 1.05E-03 2.00E+00 0.4 1.72E-03 7.3 1.053 -1.1 46.640 1.007 1.

7). As shown in Fig.3 . Significant improvement is achieved by including response contributions due to the second “mode.7). Figures 4. MPA including two “modes” underestimates the displacements of lower floors by up to 6% and overestimates the upper floor displacements by up to 22%. 4.18) considering one. The response contributions of the second and third modes are the same as before (Tables 4. The first “mode” alone provides adequate estimates of floor displacements.6 and 4.” and the errors in these estimates relative to the exact response from nonlinear RHA. but it is inadequate in estimating the story drifts (Fig.0 cm.5 × El Centro ground motion are presented below.5).17 and Tables 4. The story drifts are underestimated by up to 7% in the lower stories. (3. The errors are especially large in the plastic hinge rotations estimated by the MPA procedure even if three “modes” are included (Fig. drifts in all stories (Table 4.18 and Tables 4. 4.8 present estimates of the combined response according to Eq.7 cm and ur 3o = 2. and overestimated by up to 29% in the middle stories and by up to 13% in the upper stories. two. 3.6 and 4. (3.20) with n = 1 (Fig. 4. Results of such analyses for the selected building subject to 1.” but not the contributions of higher “modes. 4. and the structure is pushed using the force distribution of Eq.” however. Static analysis of the structure for gravity loads provides the initial state—forces and deformations—of the structure.7). Also presented at these roof displacements are the displacements at all floors (Table 4.20) with n = 1 to a target roof displacement. resulting in the pushover curve shown in Fig.18 and 48 . as are the roof displacements ur 2o = 11.6).3 Modal Pushover Analysis with Gravity Loads To evaluate the accuracy of the dynamic response of the system. 4.17 and Tables 4.4. the third “mode” contributions do not seem especially important (Fig.4.4) to roof displacement ur1o = 52. 4.” these effects will modify the combined modal response as well as the results of nonlinear RHA.18 and Tables 4.53 cm. (3.16a—which is slightly different than the one excluding gravity loads (Fig. Obviously gravity load effects will influence the seismic demands due to the first “mode.11a).17 and 4. the results presented so far did not include gravity load effects.6 through 4.4. the structure is pushed using the force distribution of Eq.7.8). The pushover curves for the second and third modes included in Fig. and three “modes. 4. Starting with its initial state under gravity loads.16 are unchanged.6 and 4. and plastic hinge rotations at selected external beam end at each floor level (Table 4. They are now included in the pushover analysis of the structure for the first “mode” only.

Locations of plastic hinges determined by MPA considering one.8). Errors in MPA as a function of ground motion intensity: (a) floor displacements. (a) MPA.5 3 0. Most pushover analysis procedures do not seem to compute to acceptable accuracy plastic hinge rotations. and (b) story drifts 49 . Noted 60 50 40 (b) Story Drifts Error Envelope Error for Story No. 2−"Modes" •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• • •• (c) 3−"Modes" • • • • • • • •• •• • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• • •• • • • • • • • • • (d) Nonlinear RHA •• •• • •• •• •• •• • •• •• •• •• •• •• •• • •• •• • •• •• •• •• • • • • • • •• •• •• Fig. Noted Error (%) 30 20 10 0 0 2 8 3 4 1 9 Error (%) 30 20 5 4 6 5 7 2 9 10 0 0 1 8 7 3 6 0. 4.5 2 GM Multiplier 2.5 × El Centro ground motion (a) Floor Displacements 60 50 40 Error Envelope Error for Floor No.5 1 1. and three “modes” and by NL-RHA for 1.5 2 GM Multiplier 2.5 1 1.15. two. 1−"Mode" • • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • •• • •• • • • • • •• •• • •• •• •• •• (b) MPA.14. 4.Table 4.5 3 Fig.

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

(a) "Mode" 1 Pushover Curve 12000 10000 u = 36.19 y by 3 2 1. V = 7433 kN. 0. α = 0. 1.85.5 0.25 0.16.5 0. α = 0.14 y by Base Shear (kN) 8000 6000 4000 1.75 0.5 2 3 4000 2000 0 0 Actual Idealized 0.0.85 0.25.75 0. V = 5210 kN.5 2 3 Actual Idealized 2000 0 0 0. 0.85 0. 2.25 1 2 4 6 Roof Displacement (cm) 8 10 Fig.13 y by Base Shear (kN) 8000 6000 1 1.9 cm. noted are peak values of roof displacement for 0.25 5 10 15 20 Roof Displacement (cm) 25 (c) "Mode" 3 Pushover Curve 12000 10000 u = 4.0. and 3.0 × El Centro ground motion 51 . “Modal” pushover curves with gravity loads included.85 Base Shear (kN) 8000 6000 4000 2000 0 0 0. V = 4952 kN. 4.50. α = 0. 0.6 cm.5 0.75.5 1 0.3 cm.75 Actual Idealized 20 40 60 Roof Displacement (cm) 80 (b) "Mode" 2 Pushover Curve 12000 10000 u = 9.

18. story drifts.5 Story Drift Ratio (%) 2 2. 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.5 1 1. two.17. and hinge plastic rotations estimated by MPA including one.5 × El Centro ground motion. 4. Errors in floor displacements. and 52 .(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 Displacement/Height (%) 2 Ground 0 0. 4.5 1 1. gravity loads included. Heightwise variation of floor displacements and story drift ratios from MPA and NL-RHA for 1.5 Fig.

5 1 1.(a) MPA. Locations of plastic hinges determined by MPA considering one.5 3 Fig. two.5 2 GM Multiplier 2.5 3 0. and (b) story drifts.5 1 1. gravity loads included 53 . 4.20. Errors in MPA as a function of ground motion intensity: (a) floor displacements. 3−"Modes" (d) Nonlinear RHA • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• •• •• • • • • • • • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • • • • • •• •• •• Fig. 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. 4.5 × El Centro ground motion.5 2 GM Multiplier 2. Noted 60 50 40 (a) Story Drifts Error Envelope Error for Story No.19. 2−"Modes" •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• •• •• (c) MPA. and three “modes” and by NL-RHA for 1. 1−"Mode" • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • •• •• •• • • • • • •• •• •• •• •• •• •• (b) MPA. gravity loads included (a) Floor Displacements 60 50 40 Error Envelope Error for Floor No.

00E-03 5.466 0.0 37.7 4.921 1.603 1.00E+00 0.037 0.594 -1.78E-03 1.88E-03 1.128 -1.018 -0.00E+00 3 “Modes” 8.00E+00 0.7 16.5 1.507 1.00E-03 5.068 0.11E-03 9.7 1.311 0.26E-03 3.00E+00 0.11E-03 9.6 19.854 0.5 × El Centro ground motion.75E-03 0.263 0.13E-03 5.813 1.6 1.23E-03 3.213 1.35E-03 8.4 20.687 0.667 0.998 21.310 1.5 54 . 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” 0.213 1.996 0.11E-03 9.00E+00 0.850 -1.2 -0.429 -1.8 0.783 1.00E+00 0.00E+00 “Mode” 2 0.049 -0.168 -0.8 0.434 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.7 19.744 1.5 10.2 0.888 1.8 1.9 -6.17E-03 9.996 -0.3 -22.00E+00 0.114 1.114 -1.3 9.207 18.821 1.637 0.35E-03 8.5 21.2 4.371 -0.933 1.197 -0.00E+00 0.19E-03 1.Table 4.603 -1.00E+00 0.00E+00 0.35E-10 3.2 1.066 -0.4 1.0 -100.2 12.858 2.066 -0.0 -30.105 0.927 1.1 13.00E+00 0.490 -11.14 m) from MPA for 1.156 -0.2 9.0 37.3 -100.257 0.00E+00 2 “Mode” 8.2 16.461 0.270 -12.3 -22.5 “Mode” 3 -1.4 26.8 -32.6 2.319 1.00E+00 0.822 1.064 -10.101 -0.55E-03 3.330 1.938 1.860 1.88E-03 0.2 2.998 1.4 -4.983 1.436 1.00E+00 0.071 0.8 -32.530 1.55E-03 3.530 1.9 -4.351 -0.3 -100.3 -22.125 -1.176 0.429 1.00E+00 3.7 Peak values of story drift ratios (as % of story height) from MPA for 1.00E+00 0.9 31.3 13. 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” 1.2 9.830 -12.516 0.055 0.686 -7.237 0.00E+00 0.7 -2.673 Table 4.6 2.0 16.4 -6.5 21.5 0.4 1.913 7.199 16.102 1.0 2 “Modes” -32.2 -2.330 1.998 0.23E-03 3.154 0.55E-03 3.00E-03 5.836 -0.214 0.213 1.23E-02 1.2 21.754 1.00E+00 0.19E-03 1.190 -0.399 -0.19E-03 1.7 -2.8 Table 4.1 4.00E+00 0.071 -0.454 1.130 0.850 1.098 20.00E+00 Error (%) “Modal” Response 1 “Mode” -32.353 -23.305 -0.5 -5.8 Peak values of hinge plastic rotations (radians) from MPA for 1.3 -100.00E-03 5.009 0.04E-02 8.00E+00 0.0 -30.908 1.8 28.821 -1.35E-03 8.033 -0.19E-04 5.057 0.04E-10 3. gravity loads included Hinge Plastic Rotation (rad) Floor “Modal” Response “Mode” 1 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 8.728 1.4 -4.037 -0.00E+00 “Mode” 3 0.414 28.2 9.88E-03 0.2 4.8 9.102 1.9 -3.854 0.3 1.831 0.983 1.478 0.1 1.11E-03 9.637 0.5 3 “Modes” -32.5 × El Centro ground motion.6 Peak values of floor displacements (as % of building height = 37.953 15.319 1.5 × El Centro ground motion.0 37.2 1.19E-03 1.315 0.665 0.514 -1.530 1.23E-03 0.877 -46.35E-03 8.908 -1.831 0.00E-10 NL RHA 1.527 -0.434 0.434 0.754 0.1 21.266 -0.515 -50.6 0.2 4.00E+00 Combined (MPA) 1 “Mode” 8.23E-03 0.263 1.5 1.109 0.00E+00 0.399 0.2 -3.0 -100.015 0.372 0.00E+00 0.237 0.107 1.

5. story drifts. computed at the target displacement represent the earthquakeinduced demands on the structure. j 2. a plot of base shear versus roof displacement. 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. 5. MPA considering three “modes.2 COMPARATIVE EVALUATION Compared in this section are the earthquake-induced demands for the selected building determined by five analyses: pushover analysis using the three force distributions in FEMA-273.5 sec . 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. gravity load effects were included in all 55 . and varies linearly in between. SRSS distribution: s* is defined by the lateral forces back-calculated from the story shears determined by response spectrum analysis of the structure. 2… N ).5 sec .” and nonlinear RHA. and 3. plastic hinge rotations. assumed to be linearly elastic. Specified in FEMA-273 are three distributions for lateral forces: 1. “Uniform” distribution: s* = m j (where the floor number j = 1. etc..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. Comparison of Modal and FEMA Pushover Analyses 5. joint rotations. k = 2 for T1 ≥ 2. 1997] The pushover curve. The floor displacements. and the exponent k = 1 for fundament period T1 £ 0.

The “uniform” distribution overestimates all floor displacements by 17-28%.3.042 0.3. 5. 5.3a and 5.11 0. the value determined from RHA of the first-mode inelastic SDF system for 1.1. The pushover curves are given in Fig. both presented in Section 4. and the displacements of upper floors are overestimated by 9-22%. 5.00719 0.5 times the El Centro ground motion. with the errors being larger for the SRSS distribution.11 0. the story drift demands in Fig.4. 5. and Table 5.11 0. Using each of these force distributions. Figures 5.0654 0. 5.112 0.0 cm.165 0. The errors in the FEMA and MPA estimates of seismic demands relative to the “exact” demands are presented in Fig. pushover analyses are implemented for a target roof displacement of 52. the floor displacement demands in Fig.3. Force distributions in FEMA-273: (a) “uniform”.1.2.1 through 5. 5.4. The three FEMA force distributions are presented in Fig.21 0.11 0.1.367 (a) Uniform (b) ELF (c) SRSS Fig. wherein the first two are obvious and the third is determined from response spectrum analysis of the building (Appendix C). and (c) SRSS 56 .0913 0. (b) ELF.126 0.281 0.0981 0.1 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.0381 0. 0.0896 0.analyses. The MPA procedure is more accurate than all the FEMA force distributions.4 and Tables 5.11 0.4a. Also included in these presentations are the MPA results considering three “modes” and the “exact” demands from nonlinear RHA.062 0. The first four floor displacements are within 5% of the “exact” value.0197 0.11 0. 5. plastic hinge rotation demands in Table 5.0446 0.119 0.11 0.0702 0.3b and Table 5. and the locations of all plastic hinges in Fig.2.177 0.0466 0.3a and Table 5.

gravity loads are included 57 . α = 0.24 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized 10 20 30 40 Roof Displacement (cm) 50 60 Fig. 5.2. Pushover curves using three force distributions in FEMA-273: (a) “uniform”. α = 0.8 cm. α = 0.18 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 = 39.14 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. V = 7456 kN. V = 6897 kN.(a) Uniform Distribution 12000 10000 u = 33 cm. (b) ELF. V = 8530 kN. and (c) SRSS.3 cm.

errors are largest in the upper stories. (The 100% error in a 6th floor hinge is ignored as it simply represents that the MPA estimated zero rotation.5 Story Drift Ratio (%) 2 2.3 demonstrate that the hinge plastic rotations estimated by all three FEMA force distributions contain unacceptably large errors. reaching 64%. For the SRSS distribution. larger errors are noted in the upper and lower stories.5 Ground 0 Ground 0 Fig. This seems to be an inherent limitation of pushover analysis. For the uniform distribution. For the ELF distribution. reaching 35%. 5. gravity loads included 58 .5 1 1. 7%.4b.” and NLRHA. whereas nonlinear RHA computed an insignificantly small value.) The pushover analysis procedures considered seem incapable of computing accurately local response quantities. Modal pushover analysis procedure gives estimates better than all the FEMA force distributions. with errors reaching 37% in this example.3. and Table 5. but it is still inaccurate. reaching 31%.3b and 5. such as hinge plastic rotations.5 Displacement/Height (%) 2 Floor 5th 5th 4th 3rd 2nd 1st NL−RHA MPA. (a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th (b) Story Drift Ratios Floor 4th 3rd 2nd 1st NL−RHA MPA. the errors are largest in the lower stories. and overestimated by no more than 32%. Figure 5.5 1 1. MPA including three “modes.2 demonstrate that the story drift demands are greatly underestimated by all the FEMA force distributions. the MPA procedure is more accurate than all the FEMA force distributions. In contrast.Figures 5. with story drifts under estimated by. 3 "Modes" FEMA Uniform ELF SRSS 0.4c and Table 5. Heightwise variation of floor displacements and story drift ratios estimated using FEMA-273 force distributions. at most. 3 "Modes" FEMA Uniform ELF SRSS 0.

Errors in floor displacements. 5.4.” nonlinear RHA (“exact”). 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. the SRSS distribution fails to identify yielding of beams in the middle floors. and the three FEMA analyses. While pushover estimates for floor displacements are more accurate. gravity loads included The structural engineering profession is now comparing these hinge plastic rotations against rotation limits established in FEMA-273 to judge structural component performance. and the ELF distribution fails 59 . The locations of plastic hinges shown in Fig. they are not good indicators of damage. story drifts. the “uniform” distribution fails to identify yielding of the beams above the fourth floor. and hinge plastic rotations estimated using FEMA-273 force distributions and MPA (including three “modes”). 5.(a) Floor Displacements 9th 8th 7th 6th MPA FEMA Uniform ELF SRSS 9th 8th 7th 6th (b) Story Drift Ratios MPA FEMA Uniform ELF SRSS Floor 4th 3rd 2nd 1st Floor −40 −20 0 Error(%) 9th 8th 7th 6th 20 40 60 5th 5th 4th 3rd 2nd 1st Ground −60 Ground −80 −60 −40 −20 0 20 Error (%) 40 60 80 (c) Hinge Plastic Rotations MPA FEMA Uniform ELF SRSS Floor 5th 4th 3rd 2nd 1st Ground −200 −150 −100 −50 0 50 Error (%) 100 150 200 Fig.5 were determined by five analyses: MPA considering the three “modes. Based on the results presented here. The locations of the plastic hinges are not identified correctly by the FEMA force distributions.

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

597 0.6 -17.836 0.168 1. Peak values of floor displacements (as % of building height = 37.7 7.8 -100.8 17.3 Peak values of hinge plastic rotations (radians) from FEMA force distributions and MPA Hinge Plastic Rotation (rad) MPA FEMA ELF SRSS 4.007 1.6 -6.061 1.262 1.436 Error (%) NL Uniform RHA 0.7 MPA -2.414 1.109 1.938 1.11E-03 9.2 -100.975 1.0 -100.399 1.9 -77.0 -100.00E+00 0.10E-02 7.011 1.270 0.724 0.4 21.8 -32.547 -27.09E-03 4.03E-03 5.673 Error (%) NL RHA 1.3 -14.314 1.13E-03 5.84 1.2 -12.372 1.875 0.344 0.00E+00 0.830 2.0 -63.667 0.19E-04 5.58E-04 6.4 16.3 -100.2 4.35E-10 3.7 -15.877 1.00E+00 4.789 0.5 61 .335 2.5 -29.367 1.2 5.00E+00 2.8 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 0.34E-03 2.19E-03 1.00E-03 5.5 26.0 37.839 0.560 1.686 0.7 26.888 0.8 7.9 16.7 FEMA ELF SRSS -22.7 Table 5.2 19.78E-03 0.998 1.55E-03 3.2 13.9 SRSS -22.51E-03 4.809 0.00E+00 NL RHA 1.23E-02 1.8 44.16E-03 0.234 1.0 Table 5.7 -60.209 0.860 1.611 0.52E-04 1.93E-03 1.3 -11.50E-03 0.2 -4.6 21.00E+00 0.015 0.5 -3.04E-02 8.3 -22.88E-03 1.6 -73.098 1.8 -63.953 0.0 -100.65E-03 7.7 19.1 10.59E-03 5.462 1.927 1.4 -23.562 1.783 1.318 2.330 1.2 16.214 1.14 m) from FEMA force distributions and MPA.9 28.623 1.490 0.263 0.487 0.2 9. gravity loads included Displacement /Height (%) FEMA ELF SRSS MPA 1.0 -59.6 4.1 163.53E-02 1.7 21.913 0.2 9.2 6.2 Peak values of story drift ratios (as % of story height) from FEMA force distributions and MPA.306 1.6 14.4 0.672 1.75E-03 0.Table 5.9 0.524 0.128 1.341 1.0 -71.399 -27.294 1.984 1.353 1.7 -28.466 0.26E-03 3.3 31.310 1.9 16.3 29.1 -3.17E-03 9.9 11.5 -33.00E-10 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 1.5 -33.8 2.3 23.566 1.2 16.854 0.530 1.4 -50.2 10.083 1.23E-03 3.355 0.6 -4.417 1.4 16.730 1.7 -29.199 27.8 -100.858 1.4 -12.00E+00 0.067 0.515 FEMA Uniform ELF 27.089 1.0 -30.8 16.351 0.9 15.5 -27.1 -26.0 -57.5 15.00E+00 0.8 -2.78E-03 1.00E+00 0.221 1.6 17.708 0. gravity loads included Displacement /Height (%) FEMA ELF SRSS MPA 0.35E-03 8.178 1.6 -11.8 -35.5 22.6 22.62E-03 0.736 0.5 -32.0 -100.998 1.6 -41.4 MPA -2.94E-03 2.9 -20.323 1.064 1.00E+00 8.1 0.6 -25.992 1.207 1.4 -55.154 1.9 -70.1.5 Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 2.399 1.45E-03 3.5 -29.3 -4.00E+00 Error (%) FEMA Uniform ELF SRSS MPA 24.5 10.195 0.530 1.

5 × El Centro ground motion. gravity loads included 62 . MPA including three “modes” and NL-RHA for 1. Locations of plastic hinges determined from three force distributions in FEMA-273.(a) FEMA: Uniform • • • • • • • • •• •• •• •• •• •• (b) FEMA: ELF • •• •• •• •• •• •• • •• • •• •• •• •• •• • • • • • •• • • • • • • • •• •• •• •• • •• •• •• •• • •• •• •• •• • • •• •• •• • • • • • •• •• (c) FEMA: SRSS • • •• •• •• •• • • •• •• •• • • •• •• •• • • • • •• •• •• • • • • • • • • • (d) MPA with 3 "Modes" • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• •• •• (e) Nonlinear RHA • • • • • • • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • • • • • •• •• •• Fig. 5.5.

5 2 GM Multiplier 2.5 1 1.5 2 GM Multiplier 2.5 3 Fig. gravity loads included 63 .5 1 1.(a) FEMA: Uniform 60 50 40 60 50 40 3 (b) FEMA: ELF Error Envelope Error for Floor No.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 0.5 3 0 0 0.5 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 0.5 2 GM Multiplier 2.6. 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 1 1. 5.

7.5 1 1. 5.5 1 1.5 2 GM Multiplier 2.5 1 1.5 2 GM Multiplier 2. 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.(a) FEMA: Uniform 100 60 50 9 8 (b) FEMA: ELF Error Envelope Error for Story No.5 3 Fig.5 1 1.5 3 0 0 0.5 2 GM Multiplier 2. Error in story drifts from three force distributions in FEMA-273 and from MPA including three “modes”.5 3 0 0 0.5 2 GM Multiplier 2.5 3 (c) FEMA: SRSS 60 50 40 60 50 40 3 2 1 (d) MPA (3 "Modes") Error (%) 30 20 Error (%) 30 5 4 8 9 20 9 6 7 7 10 0 0 4 10 2 3 1 8 5 6 0.

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

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 . p eff .mι ug t . 2. (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. 5.4.sn ug t —n denotes mode number—in the modal expansion of the effective earthquake forces. (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. and (2) superposing responses of the inelastic MDF system to individual terms.g. an uncoupled modal response history analysis (UMRHA) is developed by (1) neglecting the coupling among modal coordinates arising from yielding of the system. To enable systematic extension of the MPA procedure to inelastic systems.. 6. The response value rno is determined by implementing the following steps: (i) Develop the base-shearroof-displacement Vbn .n t = . n where φ n is now the nth natural mode for small-amplitude linear vibration. and (ii) the total response ro is determined by combining the rno (n = 1. …) according to an appropriate combination rule (e.n t is determined by pushover analysis. 66 . p eff (t ) = . with vibration properties in the linear range same as those of the nth-mode elastic SDF system. These underlying assumptions and approximations are evaluated and the errors in the UMRHA procedure relative to the “exact” nonlinear response history analysis are documented.urn bg bg bg bg b g curve from a pushover analysis of the structure for the force distribution s* = mφ n . the SRSS rule).

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

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

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

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

A. 3.. 1988). For the nth-“mode”. trapezoidal rule. Define the force distribution s* from Eq.2. 1.1.. This step can be conveniently implemented in any commercially available software. φn . ω n .g. DRAIN2DX (Allahabadi and Powell.1) using the FEMA-273 procedure (Building Seismic Safety Council. Since the target roof displacement may not be known at the start of the procedure.Appendix A Uncoupled Modal Response History Analysis A. 3. 71 . respectively. Let the roof displacement and base shear at the anchor point be urno and Vbno . and modes. 2. e.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. B. Idealize the pushover curve as a bilinear curve (Fig. Apply force distribution of Step 2.urn ) pushover curve for the force distribution s* : n 2. The structure should be pushed just beyond the target (or expected) roof displacement in the selected mode. for linear-elastic vibration of the building. using any numerical integration method. 1977).g. of the bilinear curve at the target roof displacement. Define the anchor point. (3. e.20): s* = mφn n n 2. Calculate the area under the actual pushover curve.1 incrementally and record the base shears and associated roof displacements. Compute natural frequencies. iterations may be necessary.2.1. 3. develop the base-shear – roof-displacement (Vbn . Apn .

Calculate Vbny = Vbny × Apn Abn .4. From the pushover data. 3.i 3. Calculate area under the bilinear curve OAB. kn = 0. Calculate the slope. If desired.6 . i 3. Calculate initial slope of the idealized bilinear curve. 3.8. i +1 i i 3. This value.8. urn.5. Replace i+1 with i and repeat Steps 3.0. i i i αn =  Vbno Vbny − 1  urno urny − 1         i 3. will be refined by an iterative procedure that seeks to equate areas under the actual and the idealized pushover curves. 72 . at base shear i equal to 0. Calculate the error = 100 × Abn − Apn ( ) ( ) ( ) ) Apn . Calculate the yield displacement. and B with straight-line segments to obtain the idealized bilinear curve. other appropriate methods can be ( used. urny = Vbny kn . Let the point with base shear = Vbny and roof displacement = urny be denoted as A.6 × Vbny ( ) i urn . Draw the curve OAB by connecting the three points O.1. 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. corresponding to the estimated yield i i i base shear.2. Estimate the yield base shear.4. i i 3.3.9. Abn . by connecting a straight line between origin. A. kn . Vbny .1.2.9.6 . iterations are necessary.9. i 3. and a point on the actual pushover curve with base shear equal to i 0.4.7. obtained by judgment. determine the roof displacement. i 3.4 to 3.6 × Vbny .6. 3. Vbny . O.0. i i i 3.6 × Vbny . If the error exceeds some pre-specified tolerance.

10b and 4. Calculate histories of various responses using Eqs. Repeat Steps 3 to 6 for as many modes as required for sufficient accuracy.12) and (3.15) and (3. and pseudo-acceleration history. Scale the vertical axis with * Mn to obtain * Fsno Ln = Vbno M n and * Fsny Ln = Vbny M n (Eqs.6 × Vbny A • Idealized αnkn B • 1 Actual k 1 • O n figA_1. Scale the horizontal axis by Γnφrn to obtain Dno = urno Γ nφrn and Dny = urny Γ n φrn (Eqs. 4.4) and effective modal mass from M n = Ln Γn .2.16). (3. 4. of the combined responses obtained in Step 8.11b).eps ur n. A. of the nth“mode” inelastic SDF system (Fig.4. Vbn Idealization of Pushover Curve Vbno Vbny 0.6 ur n y ur n o ur n Fig. Develop the Fsn Ln − Dn relation (Fig.3. * 4. 9. Compute the Ln and Γ n from Eq.0.3b) with unit-mass and force-deformation relation of Fig.2. A.1.11a). 7. Combine the “modal” responses using Eqs. r o . 4. Calculate peak values. 5. Dn (t ) . (3.13).1 Idealization of nth-“mode” pushover curve 73 . In general first two or three modes will suffice. A. 8. An (t ) . 4. Compute deformation history.2). (3.10a and 4. 4. 6.

4 cm (10 in. are shown in Fig.eps Dny = ur n y / Γnφr n Dno = ur n o / Γnφr n Dn Fig. Idealized bilinear curves for each of the three modes are included in Fig. generated using DRAIN-2DX. Target displacements used to generate these pushover curves are 63. The anchor point.7 cm (5 in.).2 EXAMPLE The UMRHA procedure is implemented to calculate the response of the 9-story building described in Section 3.5. First three mode shapes and frequencies of the selected building were computed and are shown in Fig. B. and 12.5 cm and Vb1o = 8729.2 Properties of the nth-“mode” inelastic SDF system A.4.3.5 cm (25 in. 3.3.1.2.6 kN. The base-shear – roof-displacement (Vbn .). 3. 2.urn ) pushover curve for the force distribution s* : n 2. A. 1. ur1o = 63. computed for the first three modes are shown in Fig. The following steps illustrate the procedure to develop the idealize curve for the first “mode”.Fsn / Ln Fsn / Ln − Dn Relationship Vbno / M* n Vbny / Mn * 1 α ω2 n n ω2 1 n figA_2. s* . 3.3. 3. and third mode. 3. for the first. A.1.1. respectively. A. second. Following is step-by-step implementation of the procedure described in Section A.4. The force distributions. is defined at the target roof displacement. Ap1 = 360777 kN-cm.2. to the north-south component of the El Centro (1940) ground motion scaled up by a factor of 1. Area under the actual pushover curve.). 74 . At this point. The pushover curves for the first three modes. 25. n 2.

4 ) − 1 (63.4 kN. L1 = 2736789 kg. and M1 = 2736789 × 1.3666 = 3740189 kg. i i 3. The first estimate of the yield base shear Vb1 y = 8006.135.5 38.09 cm.i 3. Also included are the modal damping ratios and the periods calculated from Eq. Determined from the pushover database.1.1.2. and α1 = 0. iterations are necessary.18 kN/cm. k1 = 0. The results of the iterative procedure are summarized in Table A.8 22.6 = 4803.9. The next estimate of the yield shear is Vbny = 8006.6 kN. α1 =  Vb1o Vb1 y − 1  ur1o ur1 y − 1 =         ( ) ( ) (8729. i i 3. The results for other modes are summarized in Table A.0.194.6 × Vb1 y = 4803. 3.9.3.5.18 = 38. Vb1 y = 7615.8.01%.     i 3.13). ur1 y = Vb1 y k1 = 8006.6 × Vb1 y ( ) i ur1. * 4.4. 3.6. The initial slope of the idealized bilinear curve.1. Therefore. i +1 3. 75 .3666. The Fs1 L1 − D1 relation for the first “mode” is developed as follows.86 cm at 0.23 cm. Γ1 = 1. This value exceeds the prespecified tolerance of 0.4. i i i 3. The yield displacement.09 ) − 1 = 0. Area under the bilinear curve OAB.0.1. ur1. i i i 3. and B with straight-line segments gives the idealized bilinear curve.2. (4.6 = 22.86 = 210. 3.8 kN.7. Ab1 = 365100 kN-cm. Error = 100 × (365100 − 360777 ) 360777 = 1.4. A.4 × (360777 365100 ) = 7911. 4. The post-yielding strain-hardening ratio.09 cm and Vb1 y = 8006.198%. The point A on the i i bilinear curve is defined by ur1 y = 38. i 3.9.4 210.2. The procedure converged after nineteen cycles to give ur1 y = 36. is calculated as follows. k1 .6 8006. The curve OAB obtained by connecting the three points O.4 kN.9 kN.

2.4. 7.7. 4.8.7. A. 4.4. The peak values are also plotted in Fig. 9. 4.62 (cm/sec2).7. * M1 gives Fs1o L1 = 233. Histories of roof displacement and top story drifts for the first “mode” are computed and presented in Fig. 6. 4. 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.3. The results were generated for first three “modes’ and are included in Fig. 4. Scaling the vertical axis by Fs1 y L1 = 203.46 cm and D1y = 26. Scaling the horizontal axis by Γ1φr1 gives D1o = 46.40 (cm/sec2) and 76 . The combined modal responses are presented in Fig. The peak values are computed and are summarized in Tables 4.2. 8.1 and 4.51 cm. 5.

eps 0 0 2 4 6 8 10 Roof Displacement (cm) 12 Fig.13 6000 Base Shear (kN) 4000 2000 Actual Idealized figA_3b. α = 0.14 y by 6000 Base Shear (kN) 4000 2000 Actual Idealized figA_3c.9 cm. α = 0. α = 0.6 cm. V = 7616 kN.19 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized figA_3a. A.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36. V y by = 4952 kN.3. V = 5210 kN.eps 0 0 5 10 15 20 Roof Displacement (cm) (c) "Mode" 3 Pushover Curve 25 8000 u = 4. “Modal” pushover curves for the example building 77 .2 cm.eps 10 20 30 40 50 Roof Displacement (cm) (b) "Mode" 2 Pushover Curve 60 8000 u = 9.

747 5 10 15 20 Time (sec) figa_4. 78 .(a) Deformation 80 n=1 3 (b) Pseuso−acceleration n=1 • 0.4.52 5 10 15 20 Time (sec) 25 30 0 • 1.2769 D1 (cm) 0 • 35.5 × El Centro ground motion for the first “mode.” and third “mode” inelastic SDF systems.” second “mode. Histories of deformation and pseudo-acceleration due to 1. A.m −80 0 −3 0 25 30 Fig.06 0 • 1.223 −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.

25 36.77 21.6 i k1 i ur1 y i α1 i Ab1 Error (%) 1.18 210.3 7745.4 7647.46 2.162 0.948 ζ n (%) 79 .037 0.56 19.082 0.18 210.0.79 21.194 0.40 46.18 (cm) 38.Table A.09 18.3 4603.75 21.2406 167531.5 7633.693 0.95 21.18 210.18 210.81 21.50 36.56 47.18 210.1 4569.0 4704.186 0.194 0.85 36.1 1013.5 3109.79 0.5 7624.18 210.176 0.0 4577.190 0.2 4671.135 0. Properties of “modal” inelastic SDF systems Properties Ln (kg) Γn * M n (kg) “Mode” 1 2736789 1.18 210.4927 1.05 36.7 4580. No.25 36.193 0.2 4628.8 4647.64 37.59 22.404 0.85 0.193 0.74 21.4 7911.2 4571.309 0. Results of iterative procedure to develop the idealized bilinear curve for the first “mode” inelastic SDF system Itr.194 0.5 (cm) 22.529 0.9 4570.65 1226.11 22.048 0.010 (kN) 8006.70 36.180 0.4 7672.9 4573.26 36.182 0.23 0.2671 1.107 0.6 × Vb1 y i ur1.12 3876.05 52.74 (kN/cm) 210.18 210.18 210.8525 1.6 4583.5 4614.029 0.6 7840.192 0.39 36.9 (kN) 4803.4 7714.4 4595.1 4574.022 0.86 21.1 7616.75 21.910 0.1.29 36.18 210.32 36.136 Fsny Ln (cm/sec2) Dny (cm) Fsno Ln (cm/sec ) Dno (cm) Tn (sec) 233.30 37.3 7658.194 (kN-cm) 365100 364060 363276 362684 362235 361892 361631 361432 361279 361162 361073 361004 360951 360911 360880 360856 360838 360824 360813 Table A.9 4570.28 36.59 36.3 7786.9 7615.0 4588.063 0.18 210.193 0.18 210.18 210.184 0.3666 3740189 203.139 0.38 22.83 21.8 7622.198 0.3 7628.18 210.191 0.18 210.23 22.5309 488839.2 7690.170 0.09 37.0 7619.76 21.18 210. i 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 i Vb1 y i 0.35 36.24 36.62 26.103 “Mode” 3 696400 0.90 21.44 36.151 0.013 0.51 2 “Mode” 2 -920860 -0.18 210.237 0.78 21.188 0.2.02 21.18 210.017 0.86 22.7 7639.8 4747.8 7618.

80 .

3b) with unit mass and force-deformation relation of Fig. of the nth-“mode” inelastic SDF system (Fig. 13.Appendix B Modal Pushover Analysis B.4. 4.33 cm. B.6b by solving Eq. extract from the pushover database values of other desired responses.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 = 35.5. Steps 1 to 4 of the MPA are the same as those for UMRHA presented in Appendix A.21). 10. 6. 4.1 to the north-south component of the El Centro (1940) ground motion scaled up by a factor of 1. This is the same example as solved in Appendix A. Repeat Steps 3 to 7 for as many “modes” as required for sufficient accuracy. (4. (3. Determine the total response by combining the peak “modal” responses using the SRSS combination rule of Eq. rno . Dn . (4. Calculate the peak roof displacement urno associated with the nth-“mode” inelastic SDF system from Eq. From the total hinge rotation. the first two or three “modes” will suffice. 11.1.8). 81 . Following is step-by-step implementation of the procedure described in Section B. At urno . Compute the peak deformation. (3. subtract the yield hinge rotation to determine the plastic hinge rotation. 12.18). Solving Eq. Typically. 10. or from the inelastic response (or design) spectrum.2 EXAMPLE The MPA procedure is implemented to calculate the response of the 9-story building described in Section 3.1 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.

At ur1o = 48. The total response computed by combining the peak “modal” responses using the SRSS combination rule of Eq. 12. 8.4. Peak roof displacement ur1o = Γ1 × φr1 × D1 = 1.28 cm.3 and 4.33 = 48.18) are also included in Tables 4. where results for other ground motion intensities are also included. Steps 3 to 7 are repeated for first three “modes. Results of Steps 5 and 6 are summarized in Table B.1.3 and 4.5 are the plastic hinge rotations computed by subtracting the yield hinge rotation from the total hinge rotations.366 × 1 × 35. The values are summarized in Table 4.4 for the story drifts normalized by the story height.3 for the floor displacements normalized by the building height and in Table 4.4. (3.11.28 cm. 9. Also included in Table 4.” and the results are included in Tables 4. 82 . values of floor displacements and story drifts are extracted.

52 0.126 13.766 7.913 22. Calculation of roof displacements urno from peak deformation of inelastic SDF systems Ground Motion Multiplier 0.28 46.735 3.755 27.03 0.901 8.25 20.27 0.969 0.577 16.467 14.71 1.513 0.37 57.450 4.267 5.268 0.117 5.75 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.275 1.139 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.4222 3.70 0.85 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 1.37 1.8451 5.252 9.691 0.07 “Mode” 2 4.13 2.367 1.676 6.36 1.551 2.38 22.007 36.5 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.225 2.535 14.457 12.03 26.154 78.229 8.25 Quantity Dn (cm) “Mode” 1 6.59 0.82 1.312 1.1.023 0.05 0.18 27.332 13.436 7.33 1.184 0.0 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 3.73 24.678 0.856 31.185 11.395 0.690 10.332 48.660 14.35 0.0 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 1.52 “Mode” 3 1.755 0.Table B.748 63.200 0.5 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 2.06 1.79 0.50 35.504 18.379 21.0 µ = Dn Dny urno = Γnφrn Dn (cm) 83 .35 1.

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

9 -153.1 Story Shears V j2 (kN) (6) 1114.0 1476.5 -973.2 285.1 -967.0 1231.8 1381.3.4 250.6 366.7 694.0 381.7 -46.9 1683.8 -326.9 880.9 366.6 286. 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) 1917.065 0.1 -646.090 0.5 320.7 1578.9 832.7 (10) 0.0 136.8 430.2 200.045 0.2 277.4 1759.3 -6.098 0.1 -525.367 87 .7 1622.7 525.7 234.8 374.2 105.6 319.4 1842.1 87.9 1446.047 0.1 -438.042 0.0 -5.9 -166.3 240.2 148.177 0.7 95.1 832.4 400.For convenience.6 -359.7 2065.2 97.0 980.6 -732.7 Vj (kN) (8) 2268.0 176.2 fj = ∑ fi i fj (2) (3) (4) 59.3 -646.3 222.7 355.6 1233.5 159.2 95.0 354.5 -350.7 101. the lateral forces are normalized by the base shear to obtain column 10 in Table C.070 0.6 -352. Table C.6 V j3 (kN) (7) 478.1 1857.9 446.3.7 374.5 215.2 Lateral Force fj (kN) (9) 203.5 -320.

Sign up to vote on this title
UsefulNot useful