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 .

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

iv .

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

vi .

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

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

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

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

a massless frame that provides stiffness to the system.1a. reloading from a minimum deformation takes place along a path parallel to the initial elastic branch. This lateral-force displacement relation is idealized as shown in Fig. where 0 < a << 1. 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. sign u = . It consists of a mass m (or weight w) concentrated at the roof level. Similarly. and a linear viscous damper with damping coefficient c.1 One-Story Systems EQUATION OF MOTION Consider the idealized one-story structure shown in Fig. the yield deformation. The yield strength is related to f o .1b. the yield strength. The yield strength is the same in the two directions of deformation. It is the familiar bilinear hysteretic relationship. sign u . this system is linearly elastic with stiffness k as long as the force does not exceed f y .2 2. 2.2) . Yielding begins when the force reaches f y and the deformation reaches u y . On initial loading.1) b g ug t is mu + cu + f s u. through the yield strength reduction factor. the strength required for the structure to remain elastic during the ground motion.mug t bg The governing equation for this inelastic system subjected to horizontal ground acceleration b g bg 3 (2. Unloading from a maximum deformation takes place along a path parallel to the initial elastic branch. During yielding the stiffness of the frame is a k . defined by f Ry = o fy (2. R y . 2.

1) is divided by m to obtain ~ u + 2zw n u + w 2 u y f s u. u £ u y ).e. (2. (a) Idealized one-story structure. in addition to the form of the force-deformation relation. sign u = . and (b) bilinear hysteretic forcedeformation relation For a given excitation u g t . 2. the deformation u t depends on three systems parameters: w n . or absolute (without regard to algebraic sign) maximum. and z is the damping ratio of the system vibrating within its linear elastic range (i.3).1.5) For a given u g t . and the ductility factor is u m= m uy (2. Section 7. and R y (Chopra . bg 4 . The peak. Tn = 2p w n is the natural vibration period. and u y . m depends on three system parameters: w n . z y .fS m fs c ut u fy 1 αk k 1 uy ug -fy k k um u 1 1 (a) (b) Fig.ug t n where bg bg b g bg (2. This becomes evident if Eq.4) and w n is the natural vibration frequency. 2001.3) ωn = k m ζ = c 2mω n f fs = s fy (2. z . deformation is denoted by um ..

3. f o w = 184 . α = 0. One-story.4 in Chopra.2. m h = 3.66 m • EIb EIc EIc • • L = 7. this system and excitation. and f y = 39. 2. is f y / w = ( f o / w ) ÷ 8 = 0. The yield strength of the inelastic system.2 5 . 2001) and scaled up by a factor of 2.26 kN. based on Ry = 8 . u (cm) 8 Fig.26 kN (8. one-bay frame 50 uy = 1. 2.826 kips) for w = 169.2 SYSTEM AND EXCITATION CONSIDERED Consider the one-story system in Fig.2. Vb (kN) 30 20 10 0 0 2 4 6 Roof Displacement.2311 .1. 2.32 m • Fig.9 kN (38. with Tn = 0.2 kips) . 2.5 sec and z = 5% subjected to the north-south component of the El Centro (1940) ground motion (Fig. Pushover curve for structure shown in Fig. 6. Vby = 39.376 cm. For .2: the dimensions and flexural rigidity of the structural elements are noted.04 40 Base Shear.

wherein the hysteretic force-deformation history of Fig.18 kN-m (444.4f. the joint rotation and beam hinge rotation are identical to values q m and q pm determined from RHA. and q pm = 0.g.35.4e.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. 2. The yielding stiffness of each structural element is defined as 3% of its initial stiffness.017 rad .4 ). 2. 2. 2.) for the beam and columns.3. Observe that the pushover curve matches the initial loading path of the hysteretic system.4edetermined from RHAis superimposed.1 kip-in.4 shows the earthquake response of the system described in the preceding section determined by response history analysis (RHA). Implementing this analysis with I c = 6. It is organized in five parts: (a) shows the deformation u t .134 × 107 mm 4 (75. pushover analysis cannot provide any cumulative measure of response. the ductility factor m = 5. the energy dissipated in 6 . The system is excited well beyond the yield deformation. and E = 2 × 108 kPa (29×103 ksi) gives yield moments of 21. 2.3 RESPONSE HISTORY ANALYSIS Figure 2. The peak values of the various response quantities are as follows: um = 7. the peak value determined from RHA.65 kN-m (191.0217 rad . (b) shows the lateral force f s t or base shear Vb t normalized relative to the bg weight. (d) shows the rotation q p bt g of the plastic hinges at the bg bg beam ends. 2. (c) shows the joint rotation q bt g .. as apparent in Fig. The resulting pushover curve is shown in Fig. 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.) and 50. Determined from pushover analysis at the exact peak deformation. and (e) shows the force-deformation relation. A static pushover analysis of this one-story system leads to the force-displacement relationship shown in Fig.36 cm . However. respectively.077 × 107 mm 4 (146 in.4 ).The yield moments in the beam and columns are defined as the bending moments due to the lateral force f y .3 in. I b = 3.6 kip-in.36 cm. e.

(d) plastic hinge rotation.017 • 0 (d) θp (rad) −0.04 0.4 0.5 fy / w = 0.04 0 5 10 15 Time (sec) 20 25 30 0.2 −0.0217 • θ (rad) 0 (c) −0.4 −10 −5 0 u (cm) 5 10 −0.36 −15 0.5 0. (b) base shear.2 Vb / w 0.04 0. This represents an inherent limitation of pushover analyses.yielding during the ground motion. (c) joint rotation.2 0 (e) 0 (f) −0. and (f) pushover curve 7 .2311 y (b) −0. Response of one-story system to El Centro ground motion: (a) deformation.4.2311 (a) Vb / w 0 f / w = −0. 15 u (cm) 0 • 7. or the cumulative rotation at a plastic hinge.04 0. 2.4 −10 −5 0 u (cm) 5 10 Fig. (e) force-deformation relation.2 −0.4 0.

8 .

n t = Â . and k are the mass. 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 .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. m.1) can be interpreted as effective earthquake forces: p eff t = . 2001: Section 13.mι ug t bg bg bg (3.m ι u g t bg bg (3.12): mι = n =1 Â sn = N n =1 Â Γ nmφ n N (3. each element of the influence vector ι is equal to unity.snug t n= n =1 The effective earthquake forces can then be expressed as af N af N af (3. The right side of Eq.4) 9 . Elastic Multistory Buildings 3.1) where u is the vector of N lateral floor displacements relative to the ground. classical damping.3. c. This force distribution can be expanded as a summation of modal inertia force distributions s n (Chopra.3) where φ n is the nth natural vibration mode of the structure. and p eff t = Â peff.

The equations governing the response of the system are mu + cu + ku = .10) bg bg bg 10 (3.Gn ug t (3. p eff . Then the floor displacements are u n t = φ n qn t bg bg bg bg (3. it can be demonstrated that none of the modes other than the nth mode contribute to the response. we will outline that the response of the MDF system to p eff .12) .8) gives the floor displacements u n t = Gnφ n Dn t bg (3.10) gives qn t = Gn Dn t bg bg bg (3.9) to the equation of motion for the nth-mode elastic SDF system. an SDF system with vibration propertiesnatural frequency w n and damping ratio z n of the nth-mode of the MDF system.11) and substituting in Eq.6) Next.n t = .sn ug t bg bg bg bg (3.n (t ) = N n =1 ∑ −sn u g (t ) N (3. (3.n t is entirely in the nth-mode.9) in which w n is the natural vibration frequency and z n is the damping ratio for the nth mode. subjected to u g t : Dn + 2z nw n Dn + w 2 Dn = .u g t n Comparing Eqs. (3.sn ug t bg (3.9) and (3.5) The contribution of the nth mode to s and to p eff t are: sn = Gn mφ n respectively. The solution qn t can readily be obtained by comparing Eq.7) By utilizing the orthogonality property of modes. (3.8) where the modal coordinate qn t is governed by 2 qn + 2z nw n qn + w n qn = .peff (t ) = n =1 ∑ peff . with no contribution from other modes.

15) rt = bg n =1 Â rn t = bg n =1 Â rnst An bt g N (3. etc.12) and (3.Forces sn An(t ) st rn ωn.13) st where rn denotes the modal static response.1.16) This is the classical modal RHA procedure wherein Eq.14) is the pseudo-acceleration response of the nth-mode SDF system (Chopra. Therefore. the static value of r due to external forces sn . However. 3. Conceptual explanation of modal response history analysis of elastic MDF systems Any response quantity r t story drifts. In contrast to the classical 11 bg .15) and (3.9) is the standard modal equation governing qn t .n t . internal element forces. 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. and Eqs.can be expressed by st rn t = rn An t bg bg bg bg bg (3. Equations (3.1). and An t = w 2 Dn t n bg (3.16) reflect combining the response contributions of all modes. ζn ¨ g(t ) u (a) Static Analysis of Structure (b) Dynamic Analysis of SDF System Fig. (3. 3.12) and (3.13) represent the response of the MDF system to p eff. (3. Eqs. st The two analyses that lead to rn and An t are shown schematically in Fig. (3.1. Section 13. 2001. these standard equations have been derived in an unconventional way.13) define the contribution of the nth-mode to the response.

8.derivation found in textbooks (e.1. 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. The peak modal responses are combined according to the Square-Root-of-Sum-ofSquares (SRSS) or the Complete Quadratic Combination (CQC) rules. (3.16). z n of the pseudo-acceleration response (or design) spectrum for the nth-mode SDF system.3 MODAL PUSHOVER ANALYSIS (3. 2001.3).4 and 13.(3.1). we note that static analysis of the structure subjected to lateral forces f no = Γ n mφ n An (3. Sections 12. Section 13. this response value can be obtained by static analysis of the structure subjected to lateral forces distributed over the building height according to 12 .17) (Chopra. 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. 2001. This concept provides a rational basis for the modal pushover analysis procedure developed later. (3. Alternatively. The SRSS rule. In such a response spectrum analysis (RSA). we used the modal expansion of the spatial distribution of the effective earthquake forces.9) .18) To develop a pushover analysis procedure consistent with RSA. the peak nth-mode response as in Eq.17) where An is the ordinate A Tn .19) will provide the same value of rno . which is valid for structures with well-separated natural frequencies such as multistory buildings with symmetric plan.g. 3. and Tn = 2p w n is the natural vibration period of the nth-mode of the MDF system. Chopra.

can be combined according to Eq. each determined by one pushover analysis.19 m (122 ft) in elevation. the Applied Technology Council (ATC). and California Universities for Research in Earthquake Engineering (CUREE). The building has a basement level. A benchmark structure for the SAC project. The levels of the 9-story building are numbered with respect to the ground level (see Fig. was designed by Brandow & Johnston Associates1 for the SAC2 Phase II Steel Project. Richmond. and 37.20) and the structure is pushed to the roof displacement. CA 94804-4698. with five bays each in the north-south (N-S) and east-west (E-W) directions. Obviously Dn and An are available from the response (or design) spectrum.2. Typical floor-to-floor heights (for analysis purposes measured from center-of-beam to center1 2 Brandow & Johnston Associates. shown in Fig. denoted B1. this building is 45. in both directions. 3. The interior bays of the structure contain simple framing with composite floors. The columns are 345 MPa (50 ksi) steel wideflange sections.2). SAC is a joint venture of three non-profit organizations: The Structural Engineers Association of California (SEAOC). 3. This modal pushover analysis (MPA) for linearly elastic systems is equivalent to the well-known RSA procedure (Section 3. this structure meets seismic code and represents typical medium-rise buildings designed for the Los Angeles. 1660 W. Consulting Structural Engineers. The peak modal responses. Third St.15 m (30 ft) on center. urno . 1301 S.12) is urno = Gnf rn Dn (3.73 m (150 ft) by 45. 3. which from Eq. (3. Although not actually constructed.1 System and Excitation Considered The 9-story structure.18) to obtain an estimate of the peak value ro of the total response. with the ninth level being the roof. (3. California. CA 90017. 13 . SAC Steel Project Technical Office. 46th Street. Los Angeles. The bays are 9.4 COMPARATIVE EVALUATION OF ANALYSIS PROCEDURES 3.. the peak value of the roof displacement due to the nth-mode.21) 2 where Dn = An ω n .s* = mφ n n (3.4. 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. region.2). rno .73 m (150 ft) in plan.

. the vibration periods are 2.20). This model is based on centerline dimensions of the bare frame in which beams and columns extend from centerline to centerline. s* (Eq. To ensure that this structure remains elastic. Column splices. for the second through eighth levels is 9. ceiling/flooring.83 m (6 ft) above the center-line of the beam to column joint. The building is modeled in DRAIN-2DX (Allahabadi and Powell.27. 3. The strength. 14 .2. The column lines employ two-tier construction. including the steel framing. The seismic mass of the ground level is 9. These force n distributions will be used in the pushover analysis to be presented later.0 kips-sec2/ft). such as those described in Gupta and Krawinkler (1999) can be used. The first three vibration modes and periods of the building for linearly elastic vibration are shown in Fig. 3. for the first three modes are shown in Fig.00×106 kg (616 kips.0 kips-sec2/ft).sec2/ft). i. respectively.7 kips-sec2/ft). 3. roofing and a penthouse located on the roof.49 sec.49 m (18 ft). we select a weak ground motion: the northsouth component of the El Centro (1940) ground motion scaled down by a factor of 4. and shear distortion of panel zones are neglected but large deformation (P∆) effects are included. 1988) using the M1 model developed by Krawinkler and Gupta (1998). The floor system is composed of 248 MPa (36 ksi) steel wide-flange beams in acting composite action with the floor slab. partitions. third.07×106 kg (73. The column bases are modeled as pinned and secured to the ground (at the B-1 level). The 9-story N-S MRF is depicted in Fig. and for the ninth level is 1. The seismic mass of the above ground levels of the entire structure is 9. floor slabs. mechanical/electrical. dimension.01×106 kg (69. Each frame resists one half of the seismic mass associated with the entire structure. monolithic column pieces are connected every two levels beginning with the first level.89×105 kg (67.of-beam) are 3. if desired more complex models. The simple model adopted here is sufficient for the objectives of this study. The seismic mass of the structure is due to various components of the structure.4. 3. for the first level is 1. Concrete foundation walls and surrounding soil are assumed to restrain the structure at the ground level from horizontal displacement.85.3.e. The force distributions.65×105 kg (66. are located on the first. The floor-to-floor height of the basement level is 3. fifth. 0. which are seismic (tension) splices to carry bending and uplift forces.2 kips-sec2/ft). and seventh levels at 1. and 0.96 m (13 ft).65 m (12 ft) and for the first floor is 5.

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

38 0.13 −1.12) and (3. (b) shows the base shear Vbn t normalized relative to the weight W of the building.05 2.728 2.5 Fig.05 1. and 3 n 3.05 s * 1 s * 2 s * 3 Fig. Each figure is organized in four parts: (a) shows the roof displacement urn t .8 −2.49 sec 3 T = 0.93 −1. 3. and 3.72 −2.5. (c) shows the joint rotation q n t of an external joint at 16 bg bg bg .5 0 0. n = 1.1 −2. 2.1 3.67 −1.39 3.61 2. is shown in Figs.31 −0.37 2.12 0. and 3 .85 sec 2 Floor 5th 4th 3rd 2nd 1st T1 = 2.7.9th 8th 7th 6th T = 0. 2. 3.75 1.796 0. First three natural-vibration periods and modes of the 9story building 3. determined by RHA [Eqs.13)].0272 −2.6.5 Mode Shape Component 1 1.2 Response History Analysis The structural response due to individual vibration modes. 3. Force distributions s* = mφn . n = 1.487 −1.94 2. 3. respectively.04 1.33 2.5 −1 −0.4.51 0.44 1.4.03 −1.3.27 sec Ground −1. (3.

(3. respectively. and (d) shows the Vbn . larger in story drifts. 17 .15) and (3. For a fixed number of modes included. and rotations of external joints with moment connections are presented in Tables 3.2.urn relation. 3. especially in estimating the story drifts. consistent with the increasing significance of the higher mode response among these three sets of response quantities. respectively. and the percentage errors due to truncation of higher modal contributions. ur 2o = 2. The errors due to truncating the contributions of vibration modes beyond the third mode are negligible. Combining the modal response histories for all modes gives the total response [Eqs. are presented in Fig. It is apparent that the first mode alone is inadequate. where the second and third modal responses are a larger percentage of the top story drift compared to roof displacement. 3. Also included are the combined response due to one. the peak roof displacement due to each of three modes is ur1o = 9.8.8.1. in particular. two.1. The peak values of floor displacements and story drifts determined by RHA. The peak values of the various response quantities are noted in these figures.the roof level. two.23 cm.2. and 3. which are listed in Tables 3. The linear relationship between the base shear and roof displacement for each mode implies that the structure did not yield. drifts in all stories. The same method was used to determine the peak values of many response quantities. 3.9. and 3. The peak values of displacements of all floors. and even larger in joint rotations. 3. 3. As expected. errors generally decrease as response contributions of more modes are included.3.12 cm. This is illustrated in Fig. the results for the roof displacement and top-story drift are shown in Fig.422 cm. but three modes—perhaps even two modes—are sufficient. with the first three modes provide essentially the exact response.3. three. the exact response considering all modes. errors are smallest in floor displacements.16)]. and ur 3o = 0. including one. and three vibration modes. or all modes.

1 Peak values of floor displacements (as % of building height = 37.99E-03 Mode 3 3.265 0.5 18.261 -0.9 9.31E-03 Mode 2 -1.156 0.25 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 3 Modal Response Mode 1 2.156 0.73E-03 3.407 0.89E-03 1.65E-03 RHA (all modes) 2.85E-03 3.3 -33.336 0.88E-03 2.124 0.413 RHA (all modes) 0.011 0.071 0.097 0.022 0.13E-03 2.1 Table 3.466 0.038 0.6 0.29E-03 2.199 0.76E-03 1.7 7.023 0.99E-03 2.069 0.055 0.1 -2.4 -6.9 -16.1 -0.2 Peak values of story drift ratios (as % of story height) from RHA for 0.229 0.060 -0.4 -1.225 0.50E-03 2.024 -0.4 0.001 -0.74E-04 9.152 0.8 -1.406 0.33E-03 2.310 0.227 0.08E-03 2.197 0.237 0.260 0.125 0.00E-03 1.300 0.042 0.0 7.266 0.130 0.11E-03 1.38E-03 2.01E-04 -2.205 0.183 0.09E-03 1.7 2 Modes -3.44E-03 3.6 9.0 -0.089 0.097 0.400 0.88E-03 2.3 Peak values of joint rotations (radians) from RHA for 0.14E-03 2.2 0.295 -0.058 -0.133 -0.054 0.045 0.8 -15.350 0.3 -0.235 0.275 0.0 -2.229 0.09E-03 2.2 -4.76E-03 1.00E-03 2.264 Error (%) 1 Mode 2 Modes 3 Modes -23.203 0.4 -41.7 4.13E-04 9.1 4.260 0.202 0.179 0.177 0.121 0.152 0.9 8.9 -23.6 -0.42E-04 1.64E-03 3.03E-03 -6.266 0.66E-05 -3.74E-03 1.44E-03 1.03E-03 3 Modes 2.15E-03 4.1 -19.56E-03 2.1 3.011 0.008 -0.263 0.245 0.010 -0.0 3 Modes -5.173 0.65E-03 2.157 0.321 0.237 0.4 -7.090 0.5 -1.370 0.9 2.4 -0.032 -0.069 0.47E-03 1.259 0.7 -50.042 0.080 0.6 1.31E-03 Joint Rotation (rad) Combined (RHA) 1 Mode 2.63E-03 2.Table 3.333 0.0 -10.282 0.11E-04 -5.3 19.25 × 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 0.088 -0.2 -2.89E-03 1.253 0.006 0.159 0.227 0.1 Building height is measured from the ground floor to the 9th floor.9 -24.9 -15.9 1.062 -0.03E-03 1.9 3.78E-04 -3.4 -22.002 -0.8 -5.231 -0.012 0.1 -11.24E-03 2.303 0.003 0.1 -2.09E-03 Error (%) 1 Mode -23.060 0.311 0.2 9. 18 .7 2.38E-04 2 Modes 2.253 0.173 0.45E-03 3.4 -3.177 0.2 -57.5 0.0 -46.74E-04 6.259 0.14 m3) from RHA for 0.265 RHA (all modes) 0.202 0.060 -0.6 -1.202 0.2 -1.3 -8.364 0.03E-03 1.453 0.09E-03 1.4 -1.8 1.015 -0.26E-04 -5.311 0.062 0.399 0.266 0.7 -19.63E-03 2.4 -53.2 -20.325 0.3 -3.226 0.94E-03 2.2 1.72E-03 3.117 0.097 0.235 0.258 0.3 Table 3.6 0.475 0.8 1.192 0.01E-04 3.90E-03 3.6 -1.181 0.026 0.401 Error (%) 1 Mode 2 Modes 3 Modes -23.1 -2.378 0.9 -22.043 0.28E-04 1.012 -0.7 3.035 0.3 -0.5 -2.74E-03 1.6 4.50E-03 4.0 -2.22E-03 2.00E-03 1.91E-04 1.6 11.125 0.1 -14.4 -10.307 0.8 -10.226 0.25 × El Centro ground motion Displacement /Height (%) Modal Response Combined (RHA) Mode 1 Mode 2 Mode 3 1 Mode 2 Modes 3 Modes Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 0.124 0.245 0.009 -0.282 0.8 -56.

1 5 x 10 −3 θ (rad) 0.25 × El Centro ground motion 19 . Excitation is 0. (b) base shear.5 0 u (cm) r1 7.15 u (cm) 9.12 • 0 (a) r1 −15 0. (c) joint rotation.5 9.5 0 u (cm) r1 7.12 −0.5 15 −0.0435 • 0 (b) V b1 −0. and (e) pushover curve. (d) force-deformation history. 3.0435 Vb1 / W 0.1 0.1 /W 0.1 0.5.0435 9.1 −15 −7.00174 • 0 (c) r1 −5 0 5 10 15 Time (sec) 20 25 30 0. Response due to first mode: (a) roof displacement.12 0 (d) 0 (e) 15 Fig.1 −15 −7.

and (e) pushover curve. Response due to second mode: (a) roof displacement. 3.5 0 u (cm) r2 2.1 Vb2 / W 0.1 −5 −2. (b) base shear.6.0252 (d) 0 0 (e) −0. (d) force-deformation history. (c) joint rotation.5 0 u (cm) r2 2.5 5 Fig.05 /W 0. Excitation is 0.23 0.25 × El Centro ground motion 20 .1 0.1 −5 −2.05 5 x 10 −3 b2 θ (rad) 0.23 −2.0252 −2.5 u (cm) r2 0 • 2.00244 • 0 (c) r2 −5 0 5 10 15 Time (sec) 20 25 30 0.23 −5 (a) 0.0252 • 0 (b) V −0.5 5 −0.

7.5 0 u (cm) r3 0. 3.5 0 u (cm) r2 0. Response due to third mode: (a) roof displacement. Excitation is 0.02 −1 −0. (b) base shear. (c) joint rotation. (d) force-deformation history.02 −1 −0.02 2 θ (rad) 0.02 /W V b3 0 • 0. and (e) pushover curve.1 u (cm) r3 0 • 0.422 −1 (a) 0.422 /W V b3 0 (d) 0 −0.0108 −0.000913 • 0 (c) r3 −2 0 5 10 15 Time (sec) 20 25 30 0.02 0.25 × El Centro ground motion 21 .422 (e) −0.0108 x 10 −3 (b) −0.0108 −0.5 1 Fig.5 1 −0.02 −0.

and external joint rotations computed by pushover analysis are presented in Tables 3.6d and 3. the floor displacements are proportional to the mode shape φ 1 because the structure remains elastic. For a fixed number of modes included considered.3. and it is apparent in the example considered that three modes provide most of the response (Fig. although both analyses led to identical peak values of the individual modal responses.ur relations determined by RHA (Figs. these pushover curves are consistent with the Vb .4) to roof displacement ur1o = 9.6 present estimates of the combined response according to Eq.5. These values of the response quantities are identical to the peak response values determined from RHA (Tables 3. Observe that the target roof displacement in each pushover analysis is identical to its exact value determined by RHA. 3. two.3). story drifts. As for the first mode. (3. In practical application.5. 3.21). and external joint rotations in Tables 3. (3. respectively.6. and 3.2. and 3. considering one. 3. pushing the structure using the force distribution of Eq.6e and 3.5e. or three vibration modes. 3. and 3.. which would provide the Dn value to be substituted in Eq. the value determined by RHA (Fig.4.2.20) with n = 2 and 3 up to roof displacements ur 2 o = 2.5d). As suggested by Eq.9).4. and 3. 3.e.3). 3.9 and Tables 3.18).10 and Tables 3.1.8) leads to the pushover curve shown in Fig. i.1 through 3.23 cm . 3. using the force distribution of Eq.6. and ur 3o = 0. confirming that MPA gives the exact values of the individual modal responses.4. Figure 3.7d).. 3.12 cm. except for the algebraic sign associated with Γ1 and minor round-off errors. 3. pushing the structure.7e and to the floor displacements. (3.422 cm . implying that the modal truncation errors are small if at 22 . In RHA the errors arise only from truncating the responses due to higher modes. 3.3 Modal Pushover Analysis Implementing MPA for the fundamental vibration mode. and the computed response values are identical to the peak response values determined from RHA (Tables 3. story drifts. This pushover curve is consistent with the relationship between the base shear and roof displacement determined by RHA (Fig.5. Implementing pushover analysis for the second and third modes. The floor displacements. respectively. (3.1. (3.3). respectively. 3. this value would be determined directly from the response (or design) spectrum. 3. leads to the pushover curves shown in Figs. 3. and the errors in these estimates relative to the exact response from RHA considering all modes. i. the errors in the MPA results are generally larger than in RHA (Fig.4.20) with n = 1 (Fig. and 3.e.12).

Response histories of roof displacement and top-story drift from RHA for 0.1 0.25 × El Centro ground motion 23 .48 −3 3 1.3 0. Additional errors are introduced in pushover analysis due to the approximation inherent in modal combination rules.eps fig3_9b.2 0.8.23 r2 −15 15 0 • 0.03 −3 3 Mode 3 ∆ (cm) u (cm) r3 −15 15 9.59 • Total (All Modes) 0 −15 0 ∆ (cm) u (cm) 0 −3 0 fig3_8.6 Fig.5 0.4 Story Drift Ratio (%) 0.422 r2 0 • 1.eps r 5 10 15 20 Time.12 • r1 −15 15 0 r1 0 0 0.25 × 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.685 • −3 3 Mode 2 ∆ (cm) u (cm) • 2.eps Ground 0 0. Heightwise variation of floor displacements and story drifts from RHA for 0. 3. (a) Roof Displacement 15 (b) Top Story Drift 3 Mode 1 ∆ (cm) u (cm) 9. 3.9. sec 25 30 Fig.5 Ground 0 0.3 0.83 • r3 0 • 0.4 Displacement/Height (%) 0. sec 25 30 r 5 10 15 20 Time.least three modes are included.2 0.1 0.

9 -18.65E-03 2.124 0.33E-04 5.181 0.Table 3.25 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Displacement /Height (%) Modal Response Combined (MPA) Mode 1 Mode 2 Mode 3 1 Mode 2 Modes 3 Modes 0.203 0.4 -9.125 0.042 0.267 0.264 Error (%) Modal Response 1 Mode 2 Modes 3 Modes -23.331 0.3 -33.1 -18.157 0.048 0.203 0.260 0.1 -11.069 0.4 -4.92E-04 -1.9 -13.058 -0.8 -22.2 -12.7 2.3 -2.006 -0.00E-03 3.42E-04 -1.38E-03 3 Modes 2.6 -16.63E-03 2.8 -56.03E-03 -1.63E-03 -2.40E-04 5.310 0.088 -0.9 -13.321 0.00E-03 2.002 0.9 -15.045 -0.173 0.003 -0.062 -0.9 -15.4 Peak values of floor displacements (as % of building height = 37.74E-03 -1.152 0.6 Peak values of joint rotation (radians) from MPA for 0.89E-03 2.097 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 -0.0 -46.31E-03 2.2 Table 3.96E-03 2.078 0.15E-03 2.4 -11.08E-03 2.274 0.203 0.253 -0.048 0.2 -57.012 -0.4 -7.260 -0.282 -0.023 -0.04E-03 3.4 -53.179 0.2 -11.106 0.090 0.055 0.7 -50.03E-03 1.285 0.89E-03 -1.062 0.117 0.370 0.260 0.57E-03 1 Mode -23.8 -15.9 -13.00E-03 2.060 -0.466 0.313 0.2 -4.6 -15.8 -14.332 0.00E-03 2.286 0.157 0.4 -19.300 0.227 0.89E-03 1.032 0.8 -23.177 0.5 Peak values of story drift ratios (as % of story height) from MPA for 0.09E-04 -3.230 -0.245 -0.253 RHA (All Modes) 0.15E-03 1.036 -0.011 -0.31E-03 Joint Rotation (rad) Combined (MPA) 1 Mode 2.010 0.177 0.245 0.015 -0.266 0.4 -2.12E-03 1.00E-03 -1.069 0.259 -0.73E-05 3.203 0.2 -20.156 0.267 -0.012 0.097 0.133 0.401 Error (%) 1 Mode 2 Modes 3 Modes -23.226 0.152 0.022 0.90E-04 -9.272 0.011 0.22E-04 2.038 -0.024 0.9 -14.336 0.2 -16.9 -8.385 0.1 -0.106 0.22E-03 -2.229 0.9 -16.253 0.197 0.7 -19.235 -0.89E-03 2.03E-03 6.94E-03 2.78E-04 2 Modes 2.237 0.060 0.080 0.44E-03 3.4 -14.270 0.7 -15.9 -12.72E-03 3.229 0.9 -15.328 0.125 0.043 -0.0 -0.7 2 Modes -13.4 -22.009 0.9 -14.179 0.1 -11.31E-03 Error (%) RHA (all modes) 2.270 0.43E-04 -1.336 RHA (all modes) 0.74E-03 1.05E-03 3.322 0.173 0.235 0.374 0.90E-03 1.276 0.071 0.3 -12.7 -21.231 -0.80E-04 3.0 -10.25 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Modal Response Mode 1 -2.121 0.026 0.14 m) from MPA for 0.4 -5.7 24 .53E-04 -9.407 0.0 -16.9 -13.4 -22.38E-03 2.3 -9.079 0.0 -2.9 3 Modes -12.9 -15.1 -19.3 -14.008 -0.09E-03 Mode 2 1.6 -19.00E-03 1.156 0.89E-03 -2.237 -0.3 1.28E-03 2.2 1.267 0.24E-03 -2.296 -0.09E-03 1.282 0.97E-03 1.73E-03 3.2 -0.3 Table 3.76E-03 -1.270 0.44E-03 -1.042 0.227 0.03E-03 3.4 1.133 0.6 -17.3 -41.99E-03 Mode 3 -3.4 -4.47E-03 1.9 -11.203 0.3 -19.230 0.001 0.4 -9.133 -0.9 -14.5 -16.0 -18.259 0.3 -13.74E-04 -6.09E-03 -1.253 0.9 -24.4 0.76E-03 1.8 -15.261 -0.

25 × El Centro ground motion.1 0. Heightwise variation of floor displacements and story drifts from MPA for 0.2 0.5 0.10. shading indicates modal combination error 25 .5 Ground 0 0.4 Story Drift Ratio (%) 0.eps Ground 0 0.3 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.2 0.1 0.6 Fig.eps fig3_10b. 3.4 Displacement/Height (%) 0.3 0.

26 .

the natural vibration periods and modes of the corresponding linear system are the same as the vibration properties of the inelastic system undergoing small oscillations (within the linear range).2) The standard approach is to solve directly these coupled equations. but depend on the history of the displacements: f s = f s u. sign u b g g bg (4. 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).3) . Eq. Although classical modal analysis (Section 3.1) is not valid for inelastic systems.1) With this generalization for inelastic systems. 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. and the unloading and reloading curves differ from the initial loading branch. Thus. the relations between lateral forces f s at the N floor levels and the lateral displacements u are not single valued. sign u = .1 Inelastic Multistory Buildings RESPONSE HISTORY ANALYSIS For each structural element of a building.2) to the modal coordinates of the corresponding linear system. Each structural element of this elastic system is defined to have the same stiffness as the initial stiffness of the structural element of the inelastic system.1) becomes mu + cu + f s u. it is useful for later reference to transform Eq. Therefore. (4. (3.4 4.mι ug t b (4.

8) because qr t will generally be nonzero for “modes” other than the nth “mode. (4. give the same results for u t as obtained directly from Eq.3) will.2). (4.6) for inelastic systems will no longer be described by Eq. in principle. For linear systems. (3.” implying that other “modes” will also contribute to the solution.sn ug t bg b g bg bg (4.4) represents N equations in the modal coordinates qn . where φ n are now the modes of the corresponding linear system. sign u n b g b g bg (4. However. Eq. premultiplying by φ T .6b) are mu + cu + f s u. 2. qr t = 0 for all 28 bg . and using the mass.n t given by Eq.9) for linearly elastic systems.3) in Eq. Equation (4. (4. (3. (3.6) The solution of Eq. Unlike Eq.4)] leads to the uncoupled modal response history analysis (UMRHA) procedure. (3.… N (4.2). implying coupling of modal coordinates because of yielding of the structure.Gn ug t Mn bg n = 1. sign qn = φ n f s u n .4) is rarely solved because it offers no particular advantage over Eq. (4.3). sign u = . Simultaneously solving these coupled equations and using Eq. these equations are coupled for inelastic systems. (3.2).4) where the only term that differs from Eq. however. The equations governing the response of the inelastic system to p eff. (4. (4.2 UNCOUPLED MODAL RESPONSE HISTORY ANALYSIS bg Neglecting the coupling of the N equations in modal coordinates [Eq. (4. This approximate RHA procedure is the preliminary step in developing a modal pushover analysis procedure for inelastic systems.9) involves T Fsn = Fsn qn . (4. 4. The spatial distribution s of the effective earthquake forces is expanded into the modal contributions sn according to Eq.Substituting Eq.and classical n damping-orthogonality property of modes gives F qn + 2z nw n qn + sn = .5) This resisting force depends on all modal coordinates qn t .

(a) p 15

eff,1

= −s × 0.25 × El Centro
1

(b) p 5

eff,2

= −s × 0.25 × El Centro
2

ur1 (cm)

0

ur1 (cm)

9.112 •

Mode 1

Mode 1

0

−15 15 Mode 2

−5 5 Mode 2

ur2 (cm)

0

ur2 (cm)

0 • 2.226 5 10 15 20 25 30

−15 0 15

5

10

15

20

25

30

Mode 3

−5 0 5

Mode 3

ur3 (cm)

0

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

0

−15 0

−5 0

5

10 15 20 Time (sec)

25

30

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

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

af

af

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

bg

bg

(a) p 80

eff,1

= −s × 1.5 × El Centro
1

(b) p 20

eff,2

= −s × 1.5 × El Centro
2

Mode 1

Mode 1 • 1.437

u (cm)

r1

0 • 48.24

ur1 (cm)
Mode 2

0

−80 80

−20 20 Mode 2

u (cm)

r2

0

3.37 •

ur2 (cm)

0 • 11.62 5 10 15 20 25 30

−80 0 80

5

10

15

20

25

30

Mode 3 0.4931 •

−20 0 20

Mode 3 • 0.7783

u (cm)

r3

0

ur3 (cm)
25 30

0

−80 0

5

10 15 20 Time (sec)

−20 0

5

10 15 20 Time (sec)

25

30

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

af

af

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

bg

bg

by Eq. (3.8),

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

b

g

b

g

(4.7)

30

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

bg

bg c h

(4.8)

c

h

(4.9)

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

b

g

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

bg

bg

bg

bg

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

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

31

112 5 10 15 20 Time (sec) 25 30 −10 0 −10 0 Fig. 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. Fsn / Ln ¨ g(t ) u (a) Static Analysis of Structure (b) Dynamic Analysis of Inelastic SDF System Fig.3. 4. and (b) approximate solution by UMRHA bg af bg 32 .51 • ur2 (cm) 0 ur2 (cm) n=3 14.4.9 • 0 −50 10 −50 10 n=3 ur3 (cm) 0 • 5.Forces sn Unit mass An(t ) st rn ωn.0 ¥ El Centro ground motion: (a) exact solution by NL-RHA. and 3.89 50 r1 0 u (cm) 0 • 78. n = 1. 4.n t = .575 5 10 15 20 Time (sec) 25 30 ur3 (cm) 0 • 5. ζn. 2. where u g t = 3.02 −150 n=2 −150 50 n=2 14. Roof displacement due to p eff.s n u g t .

817 5 10 15 20 Time (sec) 25 30 −10 0 −10 0 Fig. but even for this very intense excitation. and 3.956 5 10 15 20 Time (sec) 25 30 ∆r3 (cm) 0 • 5.4) and (4.6) by UMRHA is compared with the “exact” solution by nonlinear RHA.744 • 0 −20 10 n=3 −20 10 n=3 ∆r3 (cm) 0 • 5. (4.5)] is neglected. 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.s n u g t .2… N ) according to Eq.n t (n = 1.33 0 −20 20 −20 20 n=2 ∆r2 (cm) 0 ∆r2 (cm) 7.Dn relation is 33 bg bg . and (b) approximate solution by UMRHA af bg bg 4. respectively.008 • 6. the errors in either response quantity are only a few percent.15) is strictly valid only for linearly elastic systems.2.0 ¥ El Centro ground motion: (a) exact solution by NL-RHA. The errors are slightly larger in drift than in displacement.0 ¥ El Centro ground motion. this intense excitation was chosen to ensure that the structure is excited well beyond its linear elastic limit. where u g t = 3.4 and 4.(a) Nonlinear RHA 20 n=1 20 (b) UMRHA n=1 • 5.1 Underlying Assumptions and Accuracy The approximate solution of Eq. (4.n t = . n = 1. Such comparison for roof displacement and top-story drift is presented in Figs. 4. Top story drift due to p eff. (2) the superposition of responses to p eff. (3.855 ∆r1 (cm) 0 ∆r1 (cm) n=2 • 6. both for 3. and (3) the Fsn Ln .5. 2.5.

approximated by a bilinear curve to facilitate solution of Eq.1. However. when specialized for linearly elastic systems it is identical to the RHA procedure of Section 3.Dn relation to be determined in Eq. (3.6a. (3. and the base shear Vbn . it can conduct a force-controlled nonlinear static analysis with an invariant distribution of lateral forces.8) in UMRHA. 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. which is different than the Fsn Ln . When implemented by commercially available software. Although most commercially available software cannot implement such displacement-controlled analysis. Although several approximations are inherent in this UMRHA procedure. A bilinear idealization of this pushover curve for the nth-“mode” is shown in Fig. (4. (4.9).2.8) before it can be solved? Because Eq.12) for floor displacements. How to convert this Vbn . this distribution seems to be a rational choiceeven for inelastic systemsto determine Fsn in Eq. (4.8) governing Dn t is based on Eq. (4.Dn curve. the relationship between lateral forces f s and Dn in Eq. (4.20) to some bg pre-determined roof displacement.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. 4.10) . is plotted against roof displacement urn . within the linearly elastic range of the structure. The overall errors in the UMRHA procedure are documented in the examples presented in Section 4. Therefore we impose this constraint in developing the UMRHA procedure in this section and modal pushover analysis in the next section.urn pushover curve to the Fsn Ln .9) should be determined by nonlinear static analysis of the structure as the structure undergoes displacements u = Dnφ n with increasing Dn .2 Properties of the nth-“mode” Inelastic SDF System How is the Fsn Ln . (3.20). Therefore. the only force distribution that produces displacements proportional to f n is given by Eq. 4. such nonlinear static analysis provides the so-called pushover curve. The structure is pushed using the force distribution of Eq. the base shear is Vbny and roof displacement is urny . At the yield point.

2. 2001. Section 13. The two are related through Fsny Ln 2 = w n Dny (4.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.11).11) * in which M n = Ln Γn is the effective modal mass (Chopra. 4.13) This value of Tn . 4. where the yield values of Fsn Ln and Dn are Fsny Ln = Vbny * Mn Dny = urny Γ nf rn (4. the elastic vibration period Tn of the nth-mode SDF system is computed from F Ln Dny I 1/ 2 Tn = 2p G H Fsny JK (4. which is not a meaningful quantity. Knowing Fsny Ln and Dny from n Eq. the initial slope of the pushover curve in Fig.10 enables conversion of the pushover curve to the desired Fsn Ln − Dn relation shown in Fig. Properties of the nth-“mode” inelastic SDF system from the pushover curve Equation 4. 35 .6b. 4.6.6a is 2 kn = ω n Ln .5). (4.12) implying that the initial slope of the curve in Fig. (4.8). In contrast. 4.6b is w 2 . which may differ from the period of the corresponding linear system. should be used in Eq.

develop the base-shear – roof-displacement ( Vbn − urn ) pushover curve for the force distribution s* [Eq.20)].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. (3.6a). Idealize the pushover curve as a bilinear curve (Fig. ωn . for linearly-elastic vibration of the building. Calculate histories of various responses by Eqs.15) and (3. details are available in Appendix A: 1.n t .11). Typically. (4. 4. 2. with the structure is pushed to the roof displacement urno .3b) with force-deformation relation of Fig. (3. and modes.21) where Dn . 6. n 3. For the nth-mode. An t . the peak value of Dn t . of the total response r t obtained in Step 8. r o .4. Convert the idealized pushover curve to the Fsn Ln − Dn relation (Fig. the first two or three modes will suffice. 7.12) and (3.20)]. 8.6b. (3. 5. Combine the “modal” responses using Eqs. which is summarized next as a sequence of steps. of the nth-“mode” inelastic SDF system (Fig. and pseudo-acceleration history. Compute the natural frequencies.2.3 MODAL PUSHOVER ANALYSIS bg bg A pushover analysis procedure is presented next to estimate the peak response rno of the inelastic MDF system to effective earthquake forces p eff . 4. Calculate the peak value. φn . This value of the roof displacement is given by Eq. 4. (3. Consider a nonlinear static analysis of the structure subjected to lateral forces distributed over the building height according to s* [Eq. 4.6b) by utilizing Eq.13). n (3. 9. 4. Dn (t ) .16) to determine the total response. 4. Compute the deformation history. is now determined by 36 bg bg . Repeat Steps 2 to 6 for as many modes as required for sufficient accuracy.

4. details are available in Appendix B. etc.2. governed by Eq. 4. Eq. However. This pushover analysis. story drifts. which is summarized next as a sequence of steps. (4.solving Eq. it seems reasonable because it provides results for elastic buildings that are identical to the well-known RSA procedure described in Section 3.18).n t . 6. the pushover analysis provides an estimate of the peak value rno of any response rn t : floor displacements. 2001. rno also represents the exact peak value of the nth-mode contribution rn t to response r t . are combined using an appropriate modal combination rule. (4. Compute the peak deformation. or from the inelastic response (or design) spectrum. as mentioned earlier..2. as described in Section 4. Dn . alternatively. of the nth-“mode” inelastic SDF system (Fig.3) because.2 and 3. It provides results for elastic buildings that are identical to the well-known RSA procedure (Section 3.3.12). The peak “modal” responses rno . 5.6b by solving Eq. bg bg bg bg Thus. e. Calculate the peak roof displacement urno associated with the nth-“mode” inelastic SDF system from Eq.8). to obtain an estimate of the peak value ro of the total response.1 Summary The peak inelastic response of a building to earthquake excitation can be estimated by the MPA procedure just developed. 4. As shown in Sections 3. (4. it can be determined from the inelastic response (or design) spectrum (Chopra. we will refer to rno as the peak “modal” response even in the case of inelastic systems.4. This application of modal combination rules to inelastic systems obviously lacks a theoretical basis. and (2) it provides the exact modal response for elastic systems. plastic hinge rotations. joint rotations. although somewhat intuitive for inelastic buildings. Steps 1 to 4 of the MPA are same as those for UMRHA. seems reasonable. each determined by one pushover analysis. the lateral force distribution used possesses two properties: (1) it appears to be the most rational choice among all invariant distribution of forces.6 and 7. At this roof displacement.3b) with force-deformation relation of Fig.3. (3. for elastic systems.21). The response value rno is an estimate of the peak value of the response of the inelastic system to p eff. (3. Sections 7.8).6).g. 37 .

4. and ur 3o = 2. and three modes are compared with the “exact” values in Fig. in particular. subtract the yield value of hinge rotation to determine the hinge plastic rotation.9. Observe that errors tend to decrease as response contributions of more “modes” are included. the peak roof displacement due to each of the three “modes” is ur1o = 48. Figure 4.18). Repeat Steps 3 to 8 for as many “modes” as required for sufficient accuracy.8. respectively. This is to be expected.4. 4.4 COMPARATIVE EVALUATION OF ANALYSIS PROCEDURES The response of the 9-story building described earlier is determined by the two approximate methods: UMRHA and MPA. This deficiency also implies that. and three “modes”. 8. the first two or three “modes” will suffice. Typically. (3. two. determined by the UMRHA procedure.2. 4.2).7. three “modes” included. and compared with the results of a rigorous nonlinear RHA using the DRAIN-2DX computer program.7 shows the individual “modal” responses.7 cm. two.0 to 3.4.1 Uncoupled Modal Response History Analysis The structural response to 1. the “exact” results.3 cm.5 ¥ the El Centro ground motion including the response contributions associated with three “modal” inelastic SDF systems. 4. the 38 . rno . Determine the total response by combining the peak “modal” responses using the SRSS combination rule of Eq. ur 2o = 11. with.1 and 4. although the trends are not systematic as when the system remained elastic (Section 3. also included are the combined responses due to one. At urno . and the “exact” response from nonlinear RHA for the roof displacement and top-story drift. is presented next. say. and the errors in the approximate results are shown in Fig. The peak values of response are as noted. the combined response due to three “modes”. in contrast to modal analysis (Section 3. The peak values of displacements of all floors and drifts in all stories are presented in Tables 4. extract from the pushover database values of other desired responses.1). and the percentage errors in the approximate results.0. 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. 9. From the total rotation of a plastic hinge. The peak values of floor displacements and story drifts including one. the UMRHA procedure lacks a rigorous theory.53 cm.

5. sec 25 30 • 48.25. Response histories of roof displacement and top-story drift due to 1.1 and 4.88 u (cm) "Mode" 3 −80 80 0 −80 80 0 −80 0 • 44.7 • −80 80 ∆r1 (cm) u (cm) "Mode" 1 −12 r2 0 ∆r2 (cm) 12 0 u (cm) −80 80 −12 r3 0 • 2.6 5 10 15 20 Time. 2.0.2) if the system yields significantly versus if the system remains within the elastic range (Tables 3. 4. However. the errors in story drifts are larger compared to floor displacements.62 5.1 and 4. Next. we investigate how the errors in the UMRHA vary with the deformation demands imposed by the ground motion.response is much less accurate (Tables 4. 1.25 (Tables 3.75. and total response from NL-RHA 39 .2) and 1. in particular.38 6.0. defined as the El Centro ground motion multiplied by 0. For this purpose the UMRHA and exact analyses were repeated for ground motions of varying intensity.5 × El Centro ground motion: individual “modal” responses and combined response from UMRHA.3 11. 0.2). and 3.5. recall that the computed errors have been presented earlier for ground motion multipliers 0. the errors in responses computed by UMRHA including three “modes” relative to the “exact” response were determined. the degree to which the system deforms beyond its elastic limit.2).24 • NL−RHA r −12 12 ∆r (cm) u (cm) 0 5 10 15 20 Time.85. 1.53 ∆r3 (cm) 12 0 • 2. 0.0. For each excitation.7. sec 25 30 r −12 0 Fig. just as for elastic systems. (a) Roof Displacement (b) Top Story Drift 80 12 0 • 3. for a fixed number of “modes” included.1 and 3.44 • "Mode" 2 r1 0 • 48. 0.1 and 3.1 −12 12 ∆r (cm) u (cm) UMRHA 3 "Modes" 0 • 7.5 (Tables 4.

4. it will be useful to know the deformation of the system relative to its yield deformation.9. in each story drift (Fig.5 ¥ El Centro ground motion Figure 4. To interpret these results.(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 0.5 Fig. or three “modes” for 1.11. Two versions of the pushover curve are 40 . For this purpose. and the error envelope for each case. the pushover curves using force distributions s* [Eq. Heightwise variation of floor displacements and story drift ratios from UMRHA and NL-RHA for 1. 4.5 Story Drift Ratio (%) 2 2. 4.5 × El Centro ground motion (a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th (b) Story Drift Ratios Floor 4th 3rd 2nd 1st UMRHA 1 "Mode" 2 "Modes" 3 "Modes" Floor 5th 5th 4th 3rd 2nd 1st UMRHA 1 "Mode" 2 "Modes" 3 "Modes" Ground −60 −40 −20 0 Error (%) 20 40 60 Ground −60 −40 −20 0 Error (%) 20 40 60 Fig. indicated by a ground motion multiplier.20)] for the first three n modes of the system are shown in Fig.5 Displacement/Height (%) 2 Ground 0 0. Heightwise variation of error in floor displacements and story drifts estimated by UMRHA including one. Shown is the error in each floor displacement (Fig.10 summarizes the error in UMRHA as a function of ground motion intensity. 4.5 1 1.5 1 1. 4.10a). with the peak displacement of each “modal” SDF system noted for each ground motion multiplier.10b). two.8. (3.

4.10) for each “modal” inelastic SDF system (Figs 4.2. 4. respectively. and.7 cm.included: the actual curve and its idealized bilinear version. and 2. (3.01 (Fig. The location of plastic hinges and their rotations. (4. as mentioned in Section 3. The system remains elastic up to ground motion multiplier 0. even though the system remains essentially elastic. (2) increase rapidly as the ground motion multiplier increases to 1. determined from “exact” analyses.75.7). UMRHA lacks a rigorous theory and is based on several approximations. and (4) are larger in story drift compared to floor displacements.3). Additional errors are introduced in UMRHA of systems responding beyond the linearly elastic limit for at least two reasons.11a). the values determined by RHA of the nth-mode inelastic SDF system (Fig. The idealized curve for the first “mode” deviates most from the actual curve near the peak displacement corresponding to ground motion multiplier 1. Perhaps this explains why the errors are large at this excitation intensity. 41 . 2.5 ¥ the El Centro ground motion. the pushover curve for each “mode” is idealized by a bilinear curve in solving Eq.6 and 4.1).3 cm.11). the ductility factor for the first mode system is only 1.2.53 cm. and 3 (Fig. 3. The structure is pushed using the force distribution of Eq.4) to roof displacements urno = 48.5). 4. 4.4.2 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 1.75. were noted but are not shown here.20) with n = 1. Figure 4.0. Each of these three pushover analyses provides the pushover curve (Fig. Second. the peak values of displacements at all floors (Table 4. the errors in truncating the higher mode contributions are negligible. as mentioned in Section 4. 11. drifts in all stories (Table 4.4).0.10 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. For more intense excitations. and plastic hinge rotations at the external beam end at each floor level (Table 4. the first reason mentioned above seems to be the primary source for the errors. 4. (3) maintain roughly similar values for more intense ground motions. First.

138 1.202 11.5 28.201 -1.293 1.0 11.0 2.1 3.5 2 GM Multiplier 2.126 0.0 9.0 -9.473 -22.410 -1.983 1.7 31.338 -1.371 -0.220 0.863 1.3 6.133 1.256 -1.256 1.370 -0.220 -0. 4.291 0.811 1.317 0.033 0.14 m) from UMRHA for 1.863 0.1 1.8 14.8 1.2 12.120 1.298 0.575 -41.9 12.616 -0.490 -1.350 -0.2 6.478 0.333 0.057 -0.938 1.554 1.088 10.8 0.727 1.121 -0.9 12. Noted 60 50 40 (b) Story Drifts Error Envelope Error for Story No.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 -1.068 0.8 1.490 1.772 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 1.8 1.806 0.6 1.072 -1.484 0.201 -0.820 -19.044 1.2 1.5 3 Fig.938 -1.9 5.135 9.4 1.136 1.154 0.1 Peak values of floor displacements (as % of building height = 37.1 0.819 2.2 4.315 -0.5 1 1.018 0.5 10.0 11.003 0.495 1.982 9.049 -0.8 1.856 2.3 1.5 9.338 1.216 1.806 -0.3 42 .9 16.1 1.4 -1.055 -0.877 0.844 -25.698 1.3 8.9 31.070 1.169 0.663 0.6 2.079 0.376 1.5 18.072 1.900 -10.372 -1.407 -10.260 -15.9 31.942 -0.2 22.373 -0.009 -0.945 -37.526 -0.676 0.8 1.668 -23.430 1.372 1.298 -0.366 0.248 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.214 -0.371 0.5 -3.722 0.096 0.971 1.426 -1.2 1.540 0.065 0.472 1.226 -0.0 7.241 -1.707 1.200 8.5 9.256 1.5 2 GM Multiplier 2.130 0.7 14.376 -1.852 1.6 4.104 0.5 1 1.751 1.366 -0.2 Peak values of story drift ratios (as % of story height) from UMRHA for 1. and (b) story drifts Table 4.214 0.5 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Displacement /Height (%) “Modal” Response Combined (UMRHA) “Mode” 1 “Mode” 2 “Mode” 3 1 “Mode” 2 “Modes” -0.241 1.410 1.(a) Floor Displacements 60 50 40 Error Envelope Error for Floor No.3 0.379 1.10.7 Table 4.5 3 0 0 0.235 -0.187 -0.513 0.0 1.820 -0.942 1.1 8.071 -0. Errors in UMRHA as a function of ground motion intensity: (a) floor displacements.283 1. Noted 7 6 Error (%) 30 1 Error (%) 30 5 8 20 6 5 2 20 2 4 9 9 10 0 0 10 3 8 7 3 1 4 0.763 -15.513 -0.010 0.071 0.4 4.3 25.201 1.003 -31.914 -0.4 -7.663 -0.5 28.914 2.

3 and 4.3 through 4.11c). 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. 4. and three “modes. the errors in the modal pushover results are. 4. The second “mode” was necessary to identify hinges in the upper stories. 4.13 and Table 4.14. in general. For example.14.4). the hinges identified in beams at the 6th floor were at variance with the “exact” results. significantly smaller than in UMRHA (compare Fig. the results were not always accurate. Furthermore. significant improvement is achieved by including response contributions due to the second “mode”.” two “modes. However. As shown in Figs. The first “mode” alone is inadequate.5 present estimates of the combined response according to Eq. for two or three modes included. 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%.” respectively. The errors are especially large in the hinge plastic rotations estimated by the MPA procedure. especially in estimating the story drifts (Fig. were determined by four analyses: MPA considering one “mode. 4.5). and are within a few percent of the exact values for the upper stories. the third “mode” does not contribute because this SDF system remains elastic (Fig.” and three “modes.12 and Tables 4. MPA failed to identify the plastic hinges at the column bases in Fig. Obviously.13 with Fig.18).3 and 4.4).Figures 4. however.1 and 4. but was more successful when the excitation was more intense (results are not included).3 and 4. but the third “mode” contributions do not seem especially important (Fig.13 and Tables 4.3 and 4. Pushover analysis seems to be inherently limited in computing accurately hinge plastic rotations. 4.12 and Tables 4. 4. 43 . 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. this error is not especially significant because the hinge plastic rotation is very small. overestimated by up to 18% in the middle stories.12 and 4.4 with Tables 4.9 and Tables 4. two. even if three “modes” are included (Fig. The locations of plastic hinges shown in Fig. 4. (3. 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 nonlinear RHA.13 and Tables 4. The drifts are underestimated by up to 13% in the lower stories.2). and the errors in these estimates relative to the exact response from nonlinear RHA.4. Fortuitously. 4. although the error is recorded as 100% if MPA estimates zero rotation when the nonlinear RHA computes a non-zero value. 4.

2 cm. 0. 0. V = 4952 kN.5 2 3 4000 2000 0 0 Actual Idealized 0. 4.13 y by Base Shear (kN) 8000 6000 1 1.5 0. α = 0. V = 7616 kN.25 5 10 15 20 Roof Displacement (cm) 25 (c) "Mode" 3 Pushover Curve 12000 10000 u = 4.9 cm. α = 0.0. 1.14 y by Base Shear (kN) 8000 6000 4000 1.19 y by 3 2 1.5 2 3 Actual Idealized 2000 0 0 0. V = 5210 kN.25 0.0 × El Centro ground motion 44 .75 0.0. α = 0. and 3.25 1 2 4 6 Roof Displacement (cm) 8 10 Fig.5.85 0.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36.5. “Modal” pushover curves with peak roof displacements identified for 0.5 0.75 Actual Idealized 20 40 60 Roof Displacement (cm) (b) "Mode" 2 Pushover Curve 80 12000 10000 u = 9.11.6 cm.25.75. 2.5 0.5 1 0.75 0. 1.85 Base Shear (kN) 8000 6000 4000 2000 0 0 0.85 0.

45 . implying excitations intense enough to cause significant yielding of the structure. each story drift (Fig.5 1 1. UMRHA is essentially exact. 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.15b).12 Heightwise variation of floor displacements and story drift ratios from MPA and NL-RHA for 1.15 summarizes the error in MPA considering three “modes” as a function of ground motion intensity. However. now standard in engineering practice.5 Displacement/Height (%) 2 Ground 0 0.0. as discussed in Sections 3. an observation with practical implications.(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. and the error envelope for each case. 4. implying that the modal combination errors contained in these procedures are acceptable. the errors in MPA were larger for ground motion multipliers less than 0.5 Story Drift Ratio (%) 2 2. The errors are only weakly dependent on ground motion intensity (Fig.5 Fig. the MPA procedure is equivalent to the RSA procedure.3. While various sources of errors in UMRHA. shading indicates errors in MPA including three “modes” Figure 4. In this case.2 and 3. 4.4 also apply to MPA.4.15a).5 × El Centro ground motion. whereas MPA contains errors inherent in modal combination rules.15). Shown is the error in each floor displacement (Fig. indicated by a ground motion multiplier. As mentioned in Section 3. implying excitations weak enough to limit the response in the elastic range of the structure. 4. 4.4. identified in Section 3.75.5 1 1. 4. the errors in MPA were fortuitously smaller than in UMRHA (compare Figs.10 and 4.15) for ground multipliers larger than 1.3 for elastic systems (or weak ground motions).

story drifts. Errors in floor displacements. two. 4.13.(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. and hinge plastic rotations estimated by MPA including one. and three “modes” for 1.5 ¥ El Centro ground motion 46 .

118 0.015 0.3 13.18E-03 7.00E+00 0.00E+00 0.640 -1.581 0.6 1.6 -44.498 1.728 1.99E-03 6.37E-04 Error (%) “Modal” Response 1 “Mode” -32.76E-03 4.667 1.407 -27.5 Peak values of hinge plastic rotations (radians) from MPA for 1.1 -8.72E-03 7.331 1.53E-03 7.705 1.088 12.5 10.1 13.02E-03 0.00E+00 0.1 1.7 1.7 -12.8 -4.4 1.36E-03 6.116 1.00E+00 3 “Modes” 7.910 1.6 -44.980 0.666 Table 4.00E+00 0.641 1.00E+00 0.259 1.517 1.05E-03 2.168 -0.033 -0.37E-03 1.763 -14.435 0.371 -0.015 0.298 -0.9 5.575 -53.8 -6.429 0.756 0.176 0.19E-10 3.88E-03 0.0 -100.8 17.687 0.0 3 “Modes” -32.9 0.9 1.2 -4.5 × 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” 1.8 7.37E-03 1.018 -0.614 0.298 0.00E+00 0.116 1.8 -29.2 0.581 0.88E-03 1.0 15.00E+00 0.8 1.6 -8.101 -0.5 1.130 0.36E-03 6.6 7.018 0.222 0.414 1.2 -100.071 0.8 -29.209 1.4 Peak values of story drift ratios (as % of story height) from MPA for 1.1 62.737 1.6 -9.007 1.00E+00 0.0 1.00E+00 0.4 -8.8 1.804 1.3 Peak values of floor displacements (as % of building height = 37.9 7.37E-03 1.00E+00 0.5 2.2 0.009 0.36E-03 6.1 46.1 62.0 47 .1 18.304 1.879 1.244 0.9 1.781 0.125 -1.781 0.00E+00 0.053 -1.5 10.514 -1.02E-03 3.0 -50.049 -0.10E-02 9.190 -0.9 -100.8 -12.14 m) from MPA for 1.60E-04 7.250 0.36E-03 6.012 1.200 8.72E-03 7.8 -6.0 -100.5 Table 4.00E+00 2 “Modes” 7.76E-03 4.611 0.351 -0.88E-03 0.76E-03 4.2 -100.737 0.003 -16.6 13.76E-03 4.233 1.338 1.220 1.0 2 “Modes” -32.980 -0.5 × 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 0.9 -100.2 -3.399 0.02E-03 0.2 6.1 62.2 -100.007 1.5 7.305 -0.895 1.00E+00 “Mode” 3 0.315 0.50E-10 3.018 -0.1 46.055 0.756 0.233 1.705 -1.135 -7.8 0.00E+00 0.336 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.00E+00 0.820 -7.37E-03 1.5 2.116 1.72E-03 7.3 -3.0 -50.00E+00 0.105 0.3 11.8 “Mode” 3 -1.9 -4.733 1.066 -0.220 1.7 1.518 1.426 15.068 0.745 1.60E-03 2.057 0.9 2.55E-03 3.652 1.00E+00 0.154 0.738 1.683 1.6 13.2 1.55E-03 3.00E+00 1 “Mode” 7.311 0.5 × El Centro ground motion Hinge Plastic Rotation (rad) “Modal” Response Combined (MPA) “Mode” 1 7.00E+00 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th “Mode” 2 0.900 -0.304 -1.222 0.2 11.Table 4.844 -7.55E-03 3.895 1.8 -29.1 46.399 0.22E-10 NL RHA 1.594 -1.053 1.5 2.982 13.5 -6.6 -7.372 0.473 -15.752 1.0 -5.668 -13.942 6.911 0.516 0.202 8.667 -1.066 -0.071 -0.503 1.640 1.00E+00 0.00E+00 0.00E+00 0.0 -100.00E+00 3.72E-03 7.197 -0.156 -0.945 -49.266 -0.00E+00 0.260 -14.02E-03 3.3 -3.527 -0.3 1.478 0.26E-04 9.503 -1.694 1.

6 and 4. 4. They are now included in the pushover analysis of the structure for the first “mode” only. the results presented so far did not include gravity load effects. the structure is pushed using the force distribution of Eq.6 and 4. but it is inadequate in estimating the story drifts (Fig. Results of such analyses for the selected building subject to 1. Also presented at these roof displacements are the displacements at all floors (Table 4. The pushover curves for the second and third modes included in Fig.17 and Tables 4.3 Modal Pushover Analysis with Gravity Loads To evaluate the accuracy of the dynamic response of the system. (3. Static analysis of the structure for gravity loads provides the initial state—forces and deformations—of the structure.5 × El Centro ground motion are presented below.18 and 48 .5).8 present estimates of the combined response according to Eq.” but not the contributions of higher “modes.16a—which is slightly different than the one excluding gravity loads (Fig. 4. The first “mode” alone provides adequate estimates of floor displacements. Figures 4.” and the errors in these estimates relative to the exact response from nonlinear RHA.4.6 through 4.6 and 4. and the structure is pushed using the force distribution of Eq.7).18 and Tables 4. resulting in the pushover curve shown in Fig. 4.18 and Tables 4.0 cm. As shown in Fig. MPA including two “modes” underestimates the displacements of lower floors by up to 6% and overestimates the upper floor displacements by up to 22%.7 cm and ur 3o = 2. 4.17 and 4.4) to roof displacement ur1o = 52.3 .11a). (3. Significant improvement is achieved by including response contributions due to the second “mode.20) with n = 1 to a target roof displacement.6). The errors are especially large in the plastic hinge rotations estimated by the MPA procedure even if three “modes” are included (Fig.” however. Starting with its initial state under gravity loads. Obviously gravity load effects will influence the seismic demands due to the first “mode. The story drifts are underestimated by up to 7% in the lower stories. (3. The response contributions of the second and third modes are the same as before (Tables 4. and three “modes. 3.4. the third “mode” contributions do not seem especially important (Fig.18) considering one. drifts in all stories (Table 4.8).20) with n = 1 (Fig. and plastic hinge rotations at selected external beam end at each floor level (Table 4.7).7). and overestimated by up to 29% in the middle stories and by up to 13% in the upper stories. 4.4.17 and Tables 4.” these effects will modify the combined modal response as well as the results of nonlinear RHA. 4.7. two. as are the roof displacements ur 2o = 11. 4.53 cm.16 are unchanged.

5 1 1. and (b) story drifts 49 . (a) MPA.Table 4.8). 4. and three “modes” and by NL-RHA for 1.15. 1−"Mode" • • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • •• • •• • • • • • •• •• • •• •• •• •• (b) MPA.5 × El Centro ground motion (a) Floor Displacements 60 50 40 Error Envelope Error for Floor No. Noted 60 50 40 (b) Story Drifts Error Envelope Error for Story No.5 3 0. Errors in MPA as a function of ground motion intensity: (a) floor displacements. 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. Most pushover analysis procedures do not seem to compute to acceptable accuracy plastic hinge rotations.5 1 1. Locations of plastic hinges determined by MPA considering one. two.5 2 GM Multiplier 2.5 3 Fig. 4.14. 2−"Modes" •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• • •• (c) 3−"Modes" • • • • • • • •• •• • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• • •• • • • • • • • • • (d) Nonlinear RHA •• •• • •• •• •• •• • •• •• •• •• •• •• •• • •• •• • •• •• •• •• • • • • • • •• •• •• Fig.5 2 GM Multiplier 2.

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

α = 0. α = 0.25.6 cm.75 0. V = 7433 kN.14 y by Base Shear (kN) 8000 6000 4000 1. 0.75.3 cm.9 cm. α = 0.85 Base Shear (kN) 8000 6000 4000 2000 0 0 0.16. V = 4952 kN. V = 5210 kN. and 3.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36.13 y by Base Shear (kN) 8000 6000 1 1.0.25 1 2 4 6 Roof Displacement (cm) 8 10 Fig. “Modal” pushover curves with gravity loads included. 2.5 2 3 4000 2000 0 0 Actual Idealized 0. 1.25 5 10 15 20 Roof Displacement (cm) 25 (c) "Mode" 3 Pushover Curve 12000 10000 u = 4.5 0.75 Actual Idealized 20 40 60 Roof Displacement (cm) 80 (b) "Mode" 2 Pushover Curve 12000 10000 u = 9.5 0.5 0.85.0.5 2 3 Actual Idealized 2000 0 0 0.25 0.5 1 0. 0.50.19 y by 3 2 1. 4. 0.75 0.85 0. noted are peak values of roof displacement for 0.0 × El Centro ground motion 51 .85 0.

5 × El Centro ground motion. story drifts.17.5 Displacement/Height (%) 2 Ground 0 0. Heightwise variation of floor displacements and story drift ratios from MPA and NL-RHA for 1. and hinge plastic rotations estimated by MPA including one. two.5 1 1. gravity loads included. 4.5 Story Drift Ratio (%) 2 2. Errors in floor displacements. 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 Fig.(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 1 1.18. and 52 . 4.

20.5 2 GM Multiplier 2.5 1 1. 4. Noted 60 50 40 (a) Story Drifts Error Envelope Error for Story No. and (b) story drifts. Errors in MPA as a function of ground motion intensity: (a) floor displacements.5 1 1. 1−"Mode" • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • •• •• •• • • • • • •• •• •• •• •• •• •• (b) MPA. Locations of plastic hinges determined by MPA considering one.5 2 GM Multiplier 2. and three “modes” and by NL-RHA for 1.5 3 Fig. gravity loads included 53 . 3−"Modes" (d) Nonlinear RHA • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• •• •• • • • • • • • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • • • • • •• •• •• Fig. 4. 2−"Modes" •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• •• •• (c) MPA.5 3 0.5 × El Centro ground motion. gravity loads included (a) Floor Displacements 60 50 40 Error Envelope Error for Floor No. two.(a) MPA. 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 0.19.

4 20.429 1.9 -3.983 1.998 0.728 1.2 -3.466 0.530 1.00E-10 NL RHA 1.263 0.130 0.107 1.353 -23.2 21.515 -50.5 × El Centro ground motion.516 0.04E-10 3.35E-03 8.860 1.009 0.35E-10 3.190 -0.2 2.836 -0.8 28.454 1.00E+00 0.5 3 “Modes” -32.015 0.908 1.00E+00 0.665 0.372 0.850 -1.071 0.213 1.199 16.6 2.9 -6.7 1.11E-03 9.55E-03 3.055 0.311 0.037 -0.319 1.114 -1.6 1.102 1.35E-03 8.00E+00 0.00E+00 Combined (MPA) 1 “Mode” 8.7 16.530 1.530 1.414 28.00E+00 0.4 1.102 1.507 1.88E-03 0.5 -5.35E-03 8.154 0.00E+00 0.00E+00 0.2 -2.037 0.933 1.888 1.5 54 .00E+00 0.822 1.00E-03 5.371 -0.109 0.13E-03 5.0 16.7 -2.071 -0.00E+00 0.19E-03 1.921 1.197 -0.35E-03 8.7 Peak values of story drift ratios (as % of story height) from MPA for 1.237 0.0 37.399 0.3 -100.0 -30.2 1.4 -4.1 4.214 0.821 1.4 26.049 -0.850 1.270 -12.101 -0.436 1.877 -46.983 1.19E-03 1.2 4.6 2.55E-03 3.478 0.913 7. 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” 1.213 1.854 0.8 9.176 0.603 -1.351 -0.831 0.00E+00 0.00E+00 2 “Mode” 8.033 -0.6 19.7 -2.821 -1.11E-03 9.637 0.527 -0.00E+00 0. gravity loads included Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Displacement /Height (%) “Modal” Response Combined (MPA) “Mode” 1 “Mode” 2 “Mode” 3 1 “Mode” 2 “Modes” 0.00E+00 “Mode” 3 0.00E+00 0.018 -0.813 1.213 1.831 0.754 1.266 -0.434 0.399 -0.5 “Mode” 3 -1.0 -100.4 -6.88E-03 0.098 20.9 31.5 × El Centro ground motion.1 21.429 -1.207 18.310 1.4 -4.1 13.11E-03 9.057 0.686 -7.8 0.996 -0.75E-03 0.5 1.490 -11.687 0.7 4.330 1.594 -1.830 -12.23E-03 3.00E+00 3 “Modes” 8.603 1.00E+00 “Mode” 2 0.26E-03 3.8 Peak values of hinge plastic rotations (radians) from MPA for 1.2 9.5 10.3 -22.00E+00 0.9 -4.5 21.0 -100.064 -10.927 1.673 Table 4.8 -32.744 1.6 0.23E-03 3.00E-03 5.783 1.00E+00 0.19E-03 1.00E+00 3.00E+00 0.908 -1.854 0.4 1.0 37.125 -1.7 19.514 -1.996 0.04E-02 8.55E-03 3.3 -22.5 0.Table 4.3 -100.8 0.330 1.00E+00 Error (%) “Modal” Response 1 “Mode” -32.066 -0.3 13.938 1.2 1.2 9.1 1.156 -0.00E+00 0.237 0.19E-04 5.858 2.434 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.8 -32.2 4.00E-03 5.066 -0.2 9.315 0.2 -0.0 -30.257 0.0 37.23E-02 1.23E-03 0.11E-03 9.128 -1.00E-03 5.00E+00 0.00E+00 0.998 21.263 1.00E+00 0.00E+00 0.114 1.88E-03 1.0 2 “Modes” -32.3 -22.14 m) from MPA for 1.434 0.2 4.3 -100.2 12.168 -0.3 1.17E-03 9.78E-03 1.319 1.2 0.23E-03 0.19E-03 1.461 0.2 16.5 21.667 0.105 0.3 9.5 × El Centro ground motion.637 0.068 0.5 1.8 1.6 Peak values of floor displacements (as % of building height = 37.754 0. gravity loads included Hinge Plastic Rotation (rad) Floor “Modal” Response “Mode” 1 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 8.998 1.305 -0.8 Table 4.953 15.

5. gravity load effects were included in all 55 . 1997] The pushover curve. 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. etc. story drifts.” and nonlinear RHA. The floor displacements. and varies linearly in between. computed at the target displacement represent the earthquakeinduced demands on the structure.5.5 sec . j 2.. joint rotations.2 COMPARATIVE EVALUATION Compared in this section are the earthquake-induced demands for the selected building determined by five analyses: pushover analysis using the three force distributions in FEMA-273.1 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. “Uniform” distribution: s* = m j (where the floor number j = 1. a plot of base shear versus roof displacement. 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. assumed to be linearly elastic. plastic hinge rotations. 2… N ). k = 2 for T1 ≥ 2. and 3. SRSS distribution: s* is defined by the lateral forces back-calculated from the story shears determined by response spectrum analysis of the structure. MPA considering three “modes.5 sec . and the exponent k = 1 for fundament period T1 £ 0. Specified in FEMA-273 are three distributions for lateral forces: 1.

0381 0.2. with the errors being larger for the SRSS distribution.11 0.1. The MPA procedure is more accurate than all the FEMA force distributions.21 0.0 cm. 0.2.analyses. Figures 5. pushover analyses are implemented for a target roof displacement of 52. Using each of these force distributions. Force distributions in FEMA-273: (a) “uniform”.367 (a) Uniform (b) ELF (c) SRSS Fig. the value determined from RHA of the first-mode inelastic SDF system for 1.1.0913 0.3. plastic hinge rotation demands in Table 5.3. and Table 5.0981 0. The errors in the FEMA and MPA estimates of seismic demands relative to the “exact” demands are presented in Fig. 5. wherein the first two are obvious and the third is determined from response spectrum analysis of the building (Appendix C).042 0. and the locations of all plastic hinges in Fig. 5. 5.165 0. The three FEMA force distributions are presented in Fig.5 times the El Centro ground motion. (b) ELF.3.11 0.4.3a and 5. The first four floor displacements are within 5% of the “exact” value.112 0. both presented in Section 4.0654 0.11 0. The “uniform” distribution overestimates all floor displacements by 17-28%.11 0.119 0.4a.0702 0.11 0.0466 0. and the displacements of upper floors are overestimated by 9-22%.11 0.177 0.0197 0.126 0.281 0. 5. The pushover curves are given in Fig.1. and (c) SRSS 56 . 5.11 0. 5. Also included in these presentations are the MPA results considering three “modes” and the “exact” demands from nonlinear RHA.4. the story drift demands in Fig.4 and Tables 5. 5. the floor displacement demands in Fig.0446 0.0896 0.1 through 5.00719 0.3b and Table 5.062 0.3a and Table 5.1 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.

and (c) SRSS. α = 0. V = 7456 kN. 5.24 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized 10 20 30 40 Roof Displacement (cm) 50 60 Fig.18 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized 10 20 30 40 Roof Displacement (cm) (c) SRSS Distribution 50 60 12000 10000 u = 39. V = 6897 kN. Pushover curves using three force distributions in FEMA-273: (a) “uniform”. α = 0. (b) ELF.3 cm. α = 0. V = 8530 kN.2. gravity loads are included 57 .(a) Uniform Distribution 12000 10000 u = 33 cm.8 cm.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.

4b. Figure 5.4c and Table 5. Heightwise variation of floor displacements and story drift ratios estimated using FEMA-273 force distributions. the errors are largest in the lower stories. 5.5 1 1. reaching 31%. For the uniform distribution. MPA including three “modes. reaching 35%. with story drifts under estimated by.3.” and NLRHA. but it is still inaccurate. Modal pushover analysis procedure gives estimates better than all the FEMA force distributions. whereas nonlinear RHA computed an insignificantly small value. with errors reaching 37% in this example. gravity loads included 58 .5 Displacement/Height (%) 2 Floor 5th 5th 4th 3rd 2nd 1st NL−RHA MPA. 3 "Modes" FEMA Uniform ELF SRSS 0. 3 "Modes" FEMA Uniform ELF SRSS 0. larger errors are noted in the upper and lower stories. the MPA procedure is more accurate than all the FEMA force distributions. such as hinge plastic rotations.5 1 1.5 Story Drift Ratio (%) 2 2.3 demonstrate that the hinge plastic rotations estimated by all three FEMA force distributions contain unacceptably large errors. at most. (a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th (b) Story Drift Ratios Floor 4th 3rd 2nd 1st NL−RHA MPA.Figures 5. For the SRSS distribution. reaching 64%. 7%. and overestimated by no more than 32%. In contrast.3b and 5.2 demonstrate that the story drift demands are greatly underestimated by all the FEMA force distributions.) The pushover analysis procedures considered seem incapable of computing accurately local response quantities. For the ELF distribution. and Table 5. (The 100% error in a 6th floor hinge is ignored as it simply represents that the MPA estimated zero rotation. This seems to be an inherent limitation of pushover analysis.5 Ground 0 Ground 0 Fig. errors are largest in the upper stories.

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

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

00E+00 Error (%) FEMA Uniform ELF SRSS MPA 24.62E-03 0.00E+00 NL RHA 1.26E-03 3.341 1.2 Peak values of story drift ratios (as % of story height) from FEMA force distributions and MPA.367 1.530 1.7 FEMA ELF SRSS -22.88E-03 1.93E-03 1.8 2.992 1.672 1.0 -100.00E+00 0.1 10.00E+00 2.7 -29.94E-03 2.0 -30.098 1.1 0.209 0.854 0.2 19.6 -17.667 0.128 1.0 -100.9 -20.3 -22.8 44.00E+00 0.17E-03 9.9 -70.168 1.809 0.1 -26.4 -50.8 -63.03E-03 5.0 -57.195 0.00E+00 8.19E-04 5.5 22.09E-03 4.Table 5.314 1.836 0.984 1.3 -14.5 15.061 1.6 22.5 -29.462 1.015 0.50E-03 0.8 16. Peak values of floor displacements (as % of building height = 37.2 6.00E-03 5.00E+00 0.178 1.1.35E-10 3.58E-04 6.310 1.78E-03 1.839 0.0 -63.5 Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 2.2 -100.3 Peak values of hinge plastic rotations (radians) from FEMA force distributions and MPA Hinge Plastic Rotation (rad) MPA FEMA ELF SRSS 4.6 21. gravity loads included Displacement /Height (%) FEMA ELF SRSS MPA 0.597 0.7 21.0 -71.78E-03 0.6 14.953 0.263 0.8 -2.234 1.23E-03 3.089 1.7 -28.1 -3.515 FEMA Uniform ELF 27.7 26.0 -100.4 -55.2 9.560 1.083 1.318 2.6 -73.3 23.998 1.9 0.3 31.789 0.13E-03 5.686 0.45E-03 3.4 -12.998 1.84 1.623 1.3 -100.270 0.8 -32.109 1.2 10.566 1.5 -3.399 -27.04E-02 8.067 0.5 -29.221 1.436 Error (%) NL Uniform RHA 0.2 13.154 1.8 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 0.6 -25.3 -4.913 0.562 1.5 -33.7 7.2 4.55E-03 3.1 163.724 0.9 15.7 -60.323 1.199 27.5 -33.00E+00 4.0 37.214 1.399 1.830 2.547 -27.927 1.064 1.4 21.53E-02 1.8 17.9 11.19E-03 1.730 1.888 0.399 1.2 16.6 -11.35E-03 8.330 1.938 1.2 16.353 1.487 0.011 1.7 MPA -2.0 -100.858 1.207 1.736 0.877 1.8 7.6 17.355 0.59E-03 5.8 -35.708 0.5 -27.5 -32.23E-02 1.466 0.875 0.490 0.9 16.4 MPA -2.783 1.9 28.6 -6.344 0.7 19.0 Table 5.34E-03 2.6 4.294 1.860 1.3 -11. gravity loads included Displacement /Height (%) FEMA ELF SRSS MPA 1.5 61 .417 1.4 -23.524 0.00E+00 0.975 1.51E-03 4.262 1.2 9.00E+00 0.52E-04 1.9 -77.11E-03 9.611 0.00E-10 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 1.4 16.3 29.6 -4.306 1.6 -41.10E-02 7.65E-03 7.8 -100.2 -12.414 1.7 Table 5.9 16.00E+00 0.75E-03 0.351 0.0 -59.2 -4.5 26.7 -15.8 -100.335 2.5 10.530 1.4 0.372 1.14 m) from FEMA force distributions and MPA.673 Error (%) NL RHA 1.007 1.4 16.2 5.9 SRSS -22.16E-03 0.

MPA including three “modes” and NL-RHA for 1.5 × El Centro ground motion. gravity loads included 62 . 5.(a) FEMA: Uniform • • • • • • • • •• •• •• •• •• •• (b) FEMA: ELF • •• •• •• •• •• •• • •• • •• •• •• •• •• • • • • • •• • • • • • • • •• •• •• •• • •• •• •• •• • •• •• •• •• • • •• •• •• • • • • • •• •• (c) FEMA: SRSS • • •• •• •• •• • • •• •• •• • • •• •• •• • • • • •• •• •• • • • • • • • • • (d) MPA with 3 "Modes" • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• •• •• (e) Nonlinear RHA • • • • • • • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • • • • • •• •• •• Fig. Locations of plastic hinges determined from three force distributions in FEMA-273.5.

5 1 1. Noted 2 Error (%) Error (%) 1 2 30 20 10 0 0 1 4 6 7 30 20 10 8 3 5 8 4 9 7 9 6 5 0.5 3 0.5 1 1.5 2 GM Multiplier 2. Errors in floor displacements from three force distributions in FEMA-273 and from MPA including three “modes”.(a) FEMA: Uniform 60 50 40 60 50 40 3 (b) FEMA: ELF Error Envelope Error for Floor No.5 3 0 0 0.6.5 1 1.5 3 Fig.5 2 GM Multiplier 2. gravity loads included 63 . 5.5 2 GM Multiplier 2.5 1 1.5 3 (c) FEMA: SRSS 60 50 40 60 50 40 2 3 1 5 4 (d) MPA (3 "Modes") Error (%) 30 20 10 0 0 8 Error (%) 30 20 6 8 1 5 2 4 3 9 7 6 7 9 10 0 0 0.5 2 GM Multiplier 2.

5 1 1.5 3 0 0 0.5 2 GM Multiplier 2.5 2 GM Multiplier 2. gravity loads included 64 . 5.5 1 1. Error in story drifts from three force distributions in FEMA-273 and from MPA including three “modes”.5 1 1.5 3 (c) FEMA: SRSS 60 50 40 60 50 40 3 2 1 (d) MPA (3 "Modes") Error (%) 30 20 Error (%) 30 5 4 8 9 20 9 6 7 7 10 0 0 4 10 2 3 1 8 5 6 0.5 2 GM Multiplier 2.7.5 2 GM Multiplier 2.5 3 0 0 0.5 3 Fig. Noted 80 40 9 1 Error (%) 7 Error (%) 60 30 20 10 2 5 6 4 8 40 1 6 20 2 5 3 4 3 7 0 0 0.5 1 1.(a) FEMA: Uniform 100 60 50 9 8 (b) FEMA: ELF Error Envelope Error for Story No.

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

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

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

should obviously be evaluated for a wide range of buildings and ground motion ensembles. The present trend in the structural engineering profession of comparing computed hinge plastic rotations against rotation limits established in FEMA-273 to judge structural performance does not seem prudent. This new method for estimating seismic demands at low performance levels.11. 68 . 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. Instead. they are not good indicators of damage. 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. While pushover estimates for floor displacements are even more accurate. with the excitation scaled to cover a wide range of ground motion intensities and building response.

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

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

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

72 .4. kn = 0.6 . will be refined by an iterative procedure that seeks to equate areas under the actual and the idealized pushover curves.6. Abn . by connecting a straight line between origin. Calculate initial slope of the idealized bilinear curve.6 . i 3. kn .4. corresponding to the estimated yield i i i base shear. Replace i+1 with i and repeat Steps 3. Calculate Vbny = Vbny × Apn Abn .7.6 × Vbny . other appropriate methods can be ( used. and B with straight-line segments to obtain the idealized bilinear curve. Draw the curve OAB by connecting the three points O. urn. determine the roof displacement.5. 3. and a point on the actual pushover curve with base shear equal to i 0. i i i 3.i 3. obtained by judgment. i +1 i i 3. i 3.8.2.1. Calculate the error = 100 × Abn − Apn ( ) ( ) ( ) ) Apn . i i i αn =  Vbno Vbny − 1  urno urny − 1         i 3.9.4. This step gives the secant stiffness at a base shear equal to 60% of the yield base shear. iterations are necessary. i 3. Estimate the yield base shear. If the error exceeds some pre-specified tolerance.2. From the pushover data.9. Calculate the post-yielding strain-hardening ratio.6 × Vbny ( ) i urn .6 × Vbny . Calculate the slope. Vbny .1.0. This value.0. O.3. Calculate area under the bilinear curve OAB. 3.9. at base shear i equal to 0. Calculate the yield displacement. A. Vbny . urny = Vbny kn . If desired. Let the point with base shear = Vbny and roof displacement = urny be denoted as A.8.4 to 3. 3. i i 3.

Develop the Fsn Ln − Dn relation (Fig. of the nth“mode” inelastic SDF system (Fig.3. Calculate histories of various responses using Eqs. 4.13).10a and 4.6 ur n y ur n o ur n Fig. 6.4) and effective modal mass from M n = Ln Γn .3b) with unit-mass and force-deformation relation of Fig.1 Idealization of nth-“mode” pushover curve 73 . 8. * 4. 4. 4.eps ur n. 4. (3. of the combined responses obtained in Step 8. An (t ) . 9.1. and pseudo-acceleration history.4. Compute the Ln and Γ n from Eq. Vbn Idealization of Pushover Curve Vbno Vbny 0. Dn (t ) .0. Combine the “modal” responses using Eqs.6 × Vbny A • Idealized αnkn B • 1 Actual k 1 • O n figA_1. Calculate peak values. Repeat Steps 3 to 6 for as many modes as required for sufficient accuracy. A.2). Scale the vertical axis with * Mn to obtain * Fsno Ln = Vbno M n and * Fsny Ln = Vbny M n (Eqs. 4. 7. In general first two or three modes will suffice.2. Compute deformation history. A.15) and (3. r o .2. (3.10b and 4.11a). 5.12) and (3.16). Scale the horizontal axis by Γnφrn to obtain Dno = urno Γ nφrn and Dny = urny Γ n φrn (Eqs. (3. A.11b).

The following steps illustrate the procedure to develop the idealize curve for the first “mode”. The base-shear – roof-displacement (Vbn . s* .3. to the north-south component of the El Centro (1940) ground motion scaled up by a factor of 1. 3. Target displacements used to generate these pushover curves are 63. B. ur1o = 63.4. 74 .2.urn ) pushover curve for the force distribution s* : n 2. and third mode.7 cm (5 in.6 kN. is defined at the target roof displacement. Ap1 = 360777 kN-cm. second. 3. n 2.).3. The pushover curves for the first three modes.1. The force distributions.5.5 cm (25 in.2. are shown in Fig. and 12.4 cm (10 in. for the first. First three mode shapes and frequencies of the selected building were computed and are shown in Fig. 3.3. computed for the first three modes are shown in Fig. 1. Idealized bilinear curves for each of the three modes are included in Fig.Fsn / Ln Fsn / Ln − Dn Relationship Vbno / M* n Vbny / Mn * 1 α ω2 n n ω2 1 n figA_2.2 EXAMPLE The UMRHA procedure is implemented to calculate the response of the 9-story building described in Section 3. Following is step-by-step implementation of the procedure described in Section A. 25. The anchor point. A.5 cm and Vb1o = 8729.). respectively. Area under the actual pushover curve. 3. generated using DRAIN-2DX. 3. A. At this point. A.1.eps Dny = ur n y / Γnφr n Dno = ur n o / Γnφr n Dn Fig.). 2.4.1.2 Properties of the nth-“mode” inelastic SDF system A.

0.6 kN.9.13). The results of the iterative procedure are summarized in Table A. Γ1 = 1.4 210.6 × Vb1 y = 4803. Error = 100 × (365100 − 360777 ) 360777 = 1. and B with straight-line segments gives the idealized bilinear curve. Vb1 y = 7615. k1 .4.0.86 = 210.5.2. i +1 3.4 × (360777 365100 ) = 7911.i 3. and α1 = 0. 3.198%. k1 = 0.8. Ab1 = 365100 kN-cm.6. Area under the bilinear curve OAB. iterations are necessary. and M1 = 2736789 × 1.7.8 kN.3. i i 3.18 kN/cm.1.9 kN.09 cm and Vb1 y = 8006.194. is calculated as follows. The post-yielding strain-hardening ratio.135.3666 = 3740189 kg.     i 3.6 8006.2. Therefore. Also included are the modal damping ratios and the periods calculated from Eq. Determined from the pushover database. The results for other modes are summarized in Table A. The Fs1 L1 − D1 relation for the first “mode” is developed as follows.1.01%. ur1 y = Vb1 y k1 = 8006. ur1. i i 3. 4.23 cm.8 22.5 38.1.1. The procedure converged after nineteen cycles to give ur1 y = 36. The first estimate of the yield base shear Vb1 y = 8006. i i i 3.86 cm at 0.4. A.18 = 38. i i i 3.09 ) − 1 = 0.3666. The yield displacement. The curve OAB obtained by connecting the three points O. (4. L1 = 2736789 kg.4 kN. The initial slope of the idealized bilinear curve. * 4. 3. 75 .4 ) − 1 (63. α1 =  Vb1o Vb1 y − 1  ur1o ur1 y − 1 =         ( ) ( ) (8729. The next estimate of the yield shear is Vbny = 8006. i 3.6 × Vb1 y ( ) i ur1. 3.9.09 cm.6 = 22.2.6 = 4803.4.9. This value exceeds the prespecified tolerance of 0. The point A on the i i bilinear curve is defined by ur1 y = 38.4 kN.

9. A.62 (cm/sec2). * M1 gives Fs1o L1 = 233. The results were generated for first three “modes’ and are included in Fig. 4.3. 6.40 (cm/sec2) and 76 . 4. 7. Scaling the vertical axis by Fs1 y L1 = 203.1 and 4. The combined modal responses are presented in Fig.7. 5. 8. Scaling the horizontal axis by Γ1φr1 gives D1o = 46.7. The peak values are computed and are summarized in Tables 4. 4. 4.7.51 cm. The peak values are also plotted in Fig.46 cm and D1y = 26.2.4.2. Histories of roof displacement and top story drifts for the first “mode” are computed and presented in Fig. 4.4. 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.8.

(a) "Mode" 1 Pushover Curve 12000 10000 u = 36. V y by = 4952 kN.eps 0 0 5 10 15 20 Roof Displacement (cm) (c) "Mode" 3 Pushover Curve 25 8000 u = 4.14 y by 6000 Base Shear (kN) 4000 2000 Actual Idealized figA_3c.eps 10 20 30 40 50 Roof Displacement (cm) (b) "Mode" 2 Pushover Curve 60 8000 u = 9.3. V = 5210 kN. α = 0. α = 0.2 cm. α = 0.13 6000 Base Shear (kN) 4000 2000 Actual Idealized figA_3b. “Modal” pushover curves for the example building 77 .9 cm.eps 0 0 2 4 6 8 10 Roof Displacement (cm) 12 Fig. V = 7616 kN. A.19 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized figA_3a.6 cm.

06 0 • 1.4.” second “mode.223 −80 80 n=3 −3 3 n=3 D3 (cm) 0 A3 (g) • 10. A.33 A1 (g) n=2 0 −80 80 −3 3 n=2 D2 (cm) 0 A1 (g) • 22. 78 .m −80 0 −3 0 25 30 Fig. Histories of deformation and pseudo-acceleration due to 1.” and third “mode” inelastic SDF systems.5 × El Centro ground motion for the first “mode.747 5 10 15 20 Time (sec) figa_4.(a) Deformation 80 n=1 3 (b) Pseuso−acceleration n=1 • 0.2769 D1 (cm) 0 • 35.52 5 10 15 20 Time (sec) 25 30 0 • 1.

18 (cm) 38.46 2.70 36.29 36.9 (kN) 4803.44 36.3 7628.86 21.8 4747.170 0.4 7714.18 210.2 4571.5 7633.191 0.3 4603.18 210.1 4574.6 7840.151 0.81 21.2 4671.18 210.186 0.135 0.18 210. No.6 i k1 i ur1 y i α1 i Ab1 Error (%) 1.65 1226.182 0.4 7647.18 210.180 0.162 0.010 (kN) 8006.193 0. Results of iterative procedure to develop the idealized bilinear curve for the first “mode” inelastic SDF system Itr.09 37.4927 1.74 21.237 0.86 22.35 36.85 36.30 37.082 0.75 21.18 210.1.18 210.3 7658.18 210.9 4573.7 7639.11 22.18 210.6 × Vb1 y i ur1.8 4647.4 7672.59 36.40 46.910 0.28 36.193 0.83 21.139 0.188 0.9 4570.18 210.38 22.18 210.09 18.50 36.1 1013.017 0.198 0.24 36.4 4595.5309 488839.103 “Mode” 3 696400 0.79 0.0.063 0.8 7622.192 0.62 26.02 21.190 0.9 4570.693 0.18 210.2 7690.3 7745.309 0.5 (cm) 22.107 0.76 21.2671 1.404 0.56 19.0 4577.75 21.8525 1.64 37.74 (kN/cm) 210.39 36.194 0.56 47.029 0. Properties of “modal” inelastic SDF systems Properties Ln (kg) Γn * M n (kg) “Mode” 1 2736789 1.193 0.05 52.18 210.194 0.3666 3740189 203.12 3876.184 0.194 (kN-cm) 365100 364060 363276 362684 362235 361892 361631 361432 361279 361162 361073 361004 360951 360911 360880 360856 360838 360824 360813 Table A.948 ζ n (%) 79 .176 0.95 21.2.2406 167531.90 21.013 0.51 2 “Mode” 2 -920860 -0.7 4580.79 21.0 4704.77 21.59 22.037 0.8 7618.05 36.25 36.194 0.048 0.022 0.6 4583.Table A.5 7624.78 21.18 210.26 36.5 3109.25 36.0 4588.32 36.529 0.1 7616.18 210.2 4628.1 4569.0 7619.5 4614. i 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 i Vb1 y i 0.85 0.3 7786.136 Fsny Ln (cm/sec2) Dny (cm) Fsno Ln (cm/sec ) Dno (cm) Tn (sec) 233.4 7911.18 210.23 0.18 210.18 210.23 22.9 7615.

80 .

81 .21).8). (4. of the nth-“mode” inelastic SDF system (Fig.3b) with unit mass and force-deformation relation of Fig. Following is step-by-step implementation of the procedure described in Section B. 11.1 to the north-south component of the El Centro (1940) ground motion scaled up by a factor of 1. Repeat Steps 3 to 7 for as many “modes” as required for sufficient accuracy. 10.5. 10.4. Determine the total response by combining the peak “modal” responses using the SRSS combination rule of Eq. Dn . the first two or three “modes” will suffice. 4. At urno .6b by solving Eq. 13. 6. extract from the pushover database values of other desired responses. Calculate the peak roof displacement urno associated with the nth-“mode” inelastic SDF system from Eq.2 EXAMPLE The MPA procedure is implemented to calculate the response of the 9-story building described in Section 3. B. Solving Eq.33 cm. rno . 12. Compute the peak deformation. subtract the yield hinge rotation to determine the plastic hinge rotation. or from the inelastic response (or design) spectrum.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. 4.1 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. From the total hinge rotation.Appendix B Modal Pushover Analysis B. (4. Steps 1 to 4 of the MPA are the same as those for UMRHA presented in Appendix A. This is the same example as solved in Appendix A. (3. (3.18). Typically.1.

3 and 4. where results for other ground motion intensities are also included. 8.4.11. 9.” and the results are included in Tables 4.1. Peak roof displacement ur1o = Γ1 × φr1 × D1 = 1. 82 . Also included in Table 4. 12.28 cm.3 and 4.3 for the floor displacements normalized by the building height and in Table 4. The total response computed by combining the peak “modal” responses using the SRSS combination rule of Eq. At ur1o = 48.4. Steps 3 to 7 are repeated for first three “modes.18) are also included in Tables 4.28 cm. The values are summarized in Table 4.4 for the story drifts normalized by the story height.5 are the plastic hinge rotations computed by subtracting the yield hinge rotation from the total hinge rotations. Results of Steps 5 and 6 are summarized in Table B.33 = 48.366 × 1 × 35. values of floor displacements and story drifts are extracted. (3.

37 1.367 1.913 22.901 8.0 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 1.52 0.0 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 3.33 1.252 9.139 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.184 0.07 “Mode” 2 4.379 21.25 20.126 13.735 3.748 63.27 0.13 2.007 36.71 1.457 12.5 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 2.1. Calculation of roof displacements urno from peak deformation of inelastic SDF systems Ground Motion Multiplier 0.395 0.513 0.117 5.766 7.5 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.05 0.755 27.660 14.4222 3.28 46.50 35.35 0.38 22.691 0.450 4.Table B.969 0.467 14.023 0.59 0.436 7.268 0.85 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 1.577 16.70 0.267 5.676 6.36 1.856 31.73 24.79 0.312 1.154 78.551 2.0 µ = Dn Dny urno = Γnφrn Dn (cm) 83 .82 1.185 11.225 2.229 8.03 0.25 Quantity Dn (cm) “Mode” 1 6.200 0.52 “Mode” 3 1.332 13.35 1.06 1.18 27.504 18.03 26.755 0.37 57.678 0.8451 5.690 10.332 48.275 1.75 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.535 14.

84

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

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

Floor, j

mj (10-3×kg)

s* = j

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

mj

1 2 3 4 5 6 7 8 9

503.5 494.7 494.7 494.7 494.7 494.7 494.7 494.7 534.1

C.2

EQUIVALENT LATERAL FORCE (ELF) DISTRIBUTION 85

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

Floor j 1 2 3 4 5 6 7 8 9

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

s* j

=

m jhk j

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

C.3

SRSS DISTRIBUTION

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

∑ n (V jn )

2

to get column 8 of

2 277.065 0.0 136. Table C.5 215.367 87 .7 -46.9 1446.9 446.1 -646.090 0.5 -320.7 525.5 -350.For convenience.1 832.4 1842.5 -973.2 fj = ∑ fi i fj (2) (3) (4) 59. the lateral forces are normalized by the base shear to obtain column 10 in Table C.0 1476.1 Story Shears V j2 (kN) (6) 1114.042 0.1 1857.7 101.8 374.047 0.1 -438.7 95.6 319.098 0.6 -359.7 355.5 320.6 286.3 240.7 Vj (kN) (8) 2268.070 0.7 2065.2 200.4 250.2 95.8 -326.1 -967.0 381.2 105.9 880.6 V j3 (kN) (7) 478.9 -166.3 -646.7 1578.2 97.3.045 0.7 (10) 0. FEMA 273 SRSS force distribution Lateral Forces Floor j (1) 1 2 3 4 5 6 7 8 9 f j1 (kN) f j2 (kN) f j3 (kN) V j1 (kN) (5) 1917.2 Lateral Force fj (kN) (9) 203.7 694.5 159.0 -5.9 -153.3.8 430.4 400.9 1683.0 354.6 1233.0 1231.9 366.3 -6.1 87.6 366.8 1381.0 980.9 832.177 0.2 285.7 1622.3 222.1 -525.4 1759.6 -352.0 176.6 -732.7 234.7 374.2 148.

Master your semester with Scribd & The New York Times

Special offer for students: Only $4.99/month.

Master your semester with Scribd & The New York Times

Cancel anytime.