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 .

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

iv .

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

vi .

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

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

While nonlinear response history analysis (RHA) is the most rigorous procedure to compute seismic demands. 1981. 1999. 1999]. Gupta and Krawinkler. 2000]. 1998. Gupta and Kunnath. Bracci et al.. Miranda. they are conceptually complicated and computationally demanding for routine application in structural engineering practice. such as life safety and collapse prevention.and medium-rise structures in which inelastic action is distributed throughout the height of the structure [Krawinkler and Seneviratna. Krawinkler and Seneviratna. but investigations [Saiidi and Sozen. 2000] have led to good estimates of seismic demands. such satisfactory predictions of seismic demands are mostly restricted to low. 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. 1997]. 1988. 2000].. 1997. Lawson et al. 1998. Skokan and Hart. While these adaptive force distributions may provide better estimates of seismic demands [Gupta and Kunnath. 2000. Gupta and Krawinkler. 1991. Maison and Bonowitz. after the structure yields both assumptions are approximate. Attempts 1 . None of the invariant force distributions can account for the contributions of higher modes to response. 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. 1988. 1999.1 Introduction Estimating seismic demands at low performance levels. Obviously. several researchers have proposed adaptive force distributions that attempt to follow more closely the time-variant distributions of inertia forces [Fajfar and Fischinger. Kim and D’Amore. requires explicit consideration of inelastic behavior of the structure. Fajfar and Fischinger. However. 1994. 1999. To overcome these limitations. or for a redistribution of inertia forces because of structural yielding and the associated changes in the vibration properties of the structure. current structural engineering practice uses the nonlinear static procedure (NSP) or pushover analysis in FEMA-273 [Building Seismic Safety Council.

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

1 One-Story Systems EQUATION OF MOTION Consider the idealized one-story structure shown in Fig. R y . During yielding the stiffness of the frame is a k . where 0 < a << 1. defined by f Ry = o fy (2. the strength required for the structure to remain elastic during the ground motion. The yield strength is related to f o . through the yield strength reduction factor. On initial loading. It consists of a mass m (or weight w) concentrated at the roof level. This lateral-force displacement relation is idealized as shown in Fig. Unloading from a maximum deformation takes place along a path parallel to the initial elastic branch.1b. Similarly. sign u .2) . the yield strength. the yield deformation. 2. a massless frame that provides stiffness to the system. and a linear viscous damper with damping coefficient c. 2.2 2.1) b g ug t is mu + cu + f s u.mug t bg The governing equation for this inelastic system subjected to horizontal ground acceleration b g bg 3 (2. this system is linearly elastic with stiffness k as long as the force does not exceed f y .1a. It is the familiar bilinear hysteretic relationship. Yielding begins when the force reaches f y and the deformation reaches u y . The yield strength is the same in the two directions of deformation. reloading from a minimum deformation takes place along a path parallel to the initial elastic branch. sign u = . 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.

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

2 5 .5 sec and z = 5% subjected to the north-south component of the El Centro (1940) ground motion (Fig. 2001) and scaled up by a factor of 2. Pushover curve for structure shown in Fig. α = 0. 2.376 cm. u (cm) 8 Fig.2. Vby = 39.26 kN. based on Ry = 8 . is f y / w = ( f o / w ) ÷ 8 = 0. 6. with Tn = 0.26 kN (8. 2. For . this system and excitation.2311 .9 kN (38. one-bay frame 50 uy = 1.1. The yield strength of the inelastic system.3. m h = 3.32 m • Fig.2 kips) . f o w = 184 .04 40 Base Shear. 2. Vb (kN) 30 20 10 0 0 2 4 6 Roof Displacement.4 in Chopra. 2.826 kips) for w = 169.2: the dimensions and flexural rigidity of the structural elements are noted.66 m • EIb EIc EIc • • L = 7.2 SYSTEM AND EXCITATION CONSIDERED Consider the one-story system in Fig. One-story. and f y = 39.2.

. and E = 2 × 108 kPa (29×103 ksi) gives yield moments of 21.4 ).65 kN-m (191.4 shows the earthquake response of the system described in the preceding section determined by response history analysis (RHA).35.g.4edetermined from RHAis superimposed. Observe that the pushover curve matches the initial loading path of the hysteretic system. the peak value determined from RHA. e. the joint rotation and beam hinge rotation are identical to values q m and q pm determined from RHA.0217 rad . The system is excited well beyond the yield deformation.3.4e.18 kN-m (444.The yield moments in the beam and columns are defined as the bending moments due to the lateral force f y .36 cm . However. and q pm = 0. 2. This pushover curve turns out to be bilinear because the beam and the columns are designed to yield simultaneously when f s reaches f y . Determined from pushover analysis at the exact peak deformation. respectively.3 RESPONSE HISTORY ANALYSIS Figure 2. 2.134 × 107 mm 4 (75. and (e) shows the force-deformation relation. (d) shows the rotation q p bt g of the plastic hinges at the bg bg beam ends. It is organized in five parts: (a) shows the deformation u t . The resulting pushover curve is shown in Fig. 2.) and 50. The yielding stiffness of each structural element is defined as 3% of its initial stiffness. 2.017 rad .4 PUSHOVER ANALYSIS Static analysis of the one-story nonlinear system subjected to lateral force that increases in small increments is implemented until the lateral displacement reaches um = 7. The peak values of the various response quantities are as follows: um = 7. the energy dissipated in 6 . 2.077 × 107 mm 4 (146 in. as apparent in Fig. wherein the hysteretic force-deformation history of Fig. (c) shows the joint rotation q bt g . q m = 0.) for the beam and columns. I b = 3.36 cm.4f.4 ). A static pushover analysis of this one-story system leads to the force-displacement relationship shown in Fig.3 in. (b) shows the lateral force f s t or base shear Vb t normalized relative to the bg weight. 2. pushover analysis cannot provide any cumulative measure of response.1 kip-in. the ductility factor m = 5.6 kip-in. Implementing this analysis with I c = 6.

(e) force-deformation relation.017 • 0 (d) θp (rad) −0.5 0.04 0. (b) base shear.yielding during the ground motion. (d) plastic hinge rotation.5 fy / w = 0.4.04 0 5 10 15 Time (sec) 20 25 30 0.4 −10 −5 0 u (cm) 5 10 −0.2311 y (b) −0.04 0. (c) joint rotation.2 −0.4 0.04 0.2 0 (e) 0 (f) −0.2311 (a) Vb / w 0 f / w = −0.2 −0.2 Vb / w 0. and (f) pushover curve 7 .4 0. or the cumulative rotation at a plastic hinge.0217 • θ (rad) 0 (c) −0. 2. 15 u (cm) 0 • 7.4 −10 −5 0 u (cm) 5 10 Fig. Response of one-story system to El Centro ground motion: (a) deformation. This represents an inherent limitation of pushover analyses.36 −15 0.

8 .

3.1) can be interpreted as effective earthquake forces: p eff t = . and k are the mass.4) 9 . This force distribution can be expanded as a summation of modal inertia force distributions s n (Chopra. The right side of Eq. c. and p eff t = Â peff.snug t n= n =1 The effective earthquake forces can then be expressed as af N af N af (3.m ι u g t bg bg (3. m.mι ug t bg bg bg (3. and lateral stiffness matrices of the system. 2001: Section 13.3) where φ n is the nth natural vibration mode of the structure.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 . each element of the influence vector ι is equal to unity. (3.1) where u is the vector of N lateral floor displacements relative to the ground.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 = .n t = Â . Elastic Multistory Buildings 3.12): mι = n =1 Â sn = N n =1 Â Γ nmφ n N (3. classical damping.

The equations governing the response of the system are mu + cu + ku = .5) The contribution of the nth mode to s and to p eff t are: sn = Gn mφ n respectively.12) . we will outline that the response of the MDF system to p eff .sn ug t bg bg bg bg (3.n t = .7) By utilizing the orthogonality property of modes. The solution qn t can readily be obtained by comparing Eq. p eff . it can be demonstrated that none of the modes other than the nth mode contribute to the response.10) gives qn t = Gn Dn t bg bg bg (3. subjected to u g t : Dn + 2z nw n Dn + w 2 Dn = . an SDF system with vibration propertiesnatural frequency w n and damping ratio z n of the nth-mode of the MDF system.9) to the equation of motion for the nth-mode elastic SDF system.9) in which w n is the natural vibration frequency and z n is the damping ratio for the nth mode.Gn ug t (3.peff (t ) = n =1 ∑ peff . with no contribution from other modes.9) and (3.6) Next.8) where the modal coordinate qn t is governed by 2 qn + 2z nw n qn + w n qn = .8) gives the floor displacements u n t = Gnφ n Dn t bg (3.10) bg bg bg 10 (3. (3. (3.n t is entirely in the nth-mode. (3.sn ug t bg (3. Then the floor displacements are u n t = φ n qn t bg bg bg bg (3.n (t ) = N n =1 ∑ −sn u g (t ) N (3.11) and substituting in Eq.u g t n Comparing Eqs.

13) define the contribution of the nth-mode to the response. the static value of r due to external forces sn .can be expressed by st rn t = rn An t bg bg bg bg bg (3. etc. In contrast to the classical 11 bg . Conceptual explanation of modal response history analysis of elastic MDF systems Any response quantity r t story drifts. and Eqs. Equations (3.1.12) and (3. st The two analyses that lead to rn and An t are shown schematically in Fig. (3.14) is the pseudo-acceleration response of the nth-mode SDF system (Chopra.15) and (3. (3. 3.9) is the standard modal equation governing qn t . (3.Forces sn An(t ) st rn ωn. ζn ¨ g(t ) u (a) Static Analysis of Structure (b) Dynamic Analysis of SDF System Fig. and An t = w 2 Dn t n bg (3.13) st where rn denotes the modal static response.15) rt = bg n =1 Â rn t = bg n =1 Â rnst An bt g N (3. However. Eqs. 2001. 3. these standard equations have been derived in an unconventional way. Section 13.16) This is the classical modal RHA procedure wherein Eq. internal element forces.13) represent the response of the MDF system to p eff.1.12) and (3.n t . Therefore. 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 (3.16) reflect combining the response contributions of all modes.1).

18) To develop a pushover analysis procedure consistent with RSA. This concept provides a rational basis for the modal pushover analysis procedure developed later. we used the modal expansion of the spatial distribution of the effective earthquake forces. Sections 12.17) where An is the ordinate A Tn .3 MODAL PUSHOVER ANALYSIS (3.8.17) (Chopra. In such a response spectrum analysis (RSA). (3.1). 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. z n of the pseudo-acceleration response (or design) spectrum for the nth-mode SDF system. Section 13. which is valid for structures with well-separated natural frequencies such as multistory buildings with symmetric plan.19) will provide the same value of rno .3). The peak modal responses are combined according to the Square-Root-of-Sum-ofSquares (SRSS) or the Complete Quadratic Combination (CQC) rules. this response value can be obtained by static analysis of the structure subjected to lateral forces distributed over the building height according to 12 .g.(3. Alternatively. (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 SRSS rule. and Tn = 2p w n is the natural vibration period of the nth-mode of the MDF system.4 and 13. 2001. we note that static analysis of the structure subjected to lateral forces f no = Γ n mφ n An (3..16).1. Chopra. 3.derivation found in textbooks (e. 2001.9) . provides an estimate of the peak value of the total response: b g F N r 2 I 1/ 2 ro ª G Â no J H n =1 K 3. the peak nth-mode response as in Eq.

can be combined according to Eq. and California Universities for Research in Earthquake Engineering (CUREE). (3. which from Eq. Obviously Dn and An are available from the response (or design) spectrum. The levels of the 9-story building are numbered with respect to the ground level (see Fig.12) is urno = Gnf rn Dn (3. the peak value of the roof displacement due to the nth-mode. (3. The building’s lateral load-resisting system is composed of steel perimeter moment-resisting frames (MRFS) with simple framing on the farthest south E-W frame.2).19 m (122 ft) in elevation. Consulting Structural Engineers. 1660 W. 13 . 3. region. 46th Street. 1301 S. with the ninth level being the roof.20) and the structure is pushed to the roof displacement. The columns are 345 MPa (50 ksi) steel wideflange sections.21) 2 where Dn = An ω n . The interior bays of the structure contain simple framing with composite floors. each determined by one pushover analysis. shown in Fig. and 37.4.4 COMPARATIVE EVALUATION OF ANALYSIS PROCEDURES 3..73 m (150 ft) in plan. SAC Steel Project Technical Office. Although not actually constructed. denoted B1. 3. A benchmark structure for the SAC project. 3. California. SAC is a joint venture of three non-profit organizations: The Structural Engineers Association of California (SEAOC). The bays are 9. the Applied Technology Council (ATC). Richmond. Typical floor-to-floor heights (for analysis purposes measured from center-of-beam to center1 2 Brandow & Johnston Associates. Los Angeles. was designed by Brandow & Johnston Associates1 for the SAC2 Phase II Steel Project.18) to obtain an estimate of the peak value ro of the total response. this structure meets seismic code and represents typical medium-rise buildings designed for the Los Angeles. rno . CA 90017. this building is 45. Third St.2. This modal pushover analysis (MPA) for linearly elastic systems is equivalent to the well-known RSA procedure (Section 3. in both directions.1 System and Excitation Considered The 9-story structure. urno . The building has a basement level.2).73 m (150 ft) by 45. with five bays each in the north-south (N-S) and east-west (E-W) directions.s* = mφ n n (3.15 m (30 ft) on center. CA 94804-4698. The peak modal responses.

3. Concrete foundation walls and surrounding soil are assumed to restrain the structure at the ground level from horizontal displacement. Each frame resists one half of the seismic mass associated with the entire structure.3.7 kips-sec2/ft). fifth. The column bases are modeled as pinned and secured to the ground (at the B-1 level). which are seismic (tension) splices to carry bending and uplift forces. if desired more complex models. mechanical/electrical.2. Column splices. The first three vibration modes and periods of the building for linearly elastic vibration are shown in Fig.01×106 kg (69.4.00×106 kg (616 kips. ceiling/flooring. The seismic mass of the structure is due to various components of the structure. and 0.e. monolithic column pieces are connected every two levels beginning with the first level. The strength. such as those described in Gupta and Krawinkler (1999) can be used.49 m (18 ft). The seismic mass of the above ground levels of the entire structure is 9.83 m (6 ft) above the center-line of the beam to column joint. The floor-to-floor height of the basement level is 3.0 kips-sec2/ft). respectively.. for the first level is 1. 3. The column lines employ two-tier construction.2 kips-sec2/ft). The simple model adopted here is sufficient for the objectives of this study. These force n distributions will be used in the pushover analysis to be presented later. 0. for the first three modes are shown in Fig. dimension.89×105 kg (67. To ensure that this structure remains elastic.65 m (12 ft) and for the first floor is 5. partitions. floor slabs. third. This model is based on centerline dimensions of the bare frame in which beams and columns extend from centerline to centerline. including the steel framing. The 9-story N-S MRF is depicted in Fig.85.49 sec.of-beam) are 3.96 m (13 ft).27.07×106 kg (73.20).0 kips-sec2/ft). the vibration periods are 2. 14 .65×105 kg (66. roofing and a penthouse located on the roof. i. for the second through eighth levels is 9. s* (Eq. are located on the first.sec2/ft). and for the ninth level is 1. 3. 1988) using the M1 model developed by Krawinkler and Gupta (1998). and seventh levels at 1. The building is modeled in DRAIN-2DX (Allahabadi and Powell. The floor system is composed of 248 MPa (36 ksi) steel wide-flange beams in acting composite action with the floor slab. and shear distortion of panel zones are neglected but large deformation (P∆) effects are included. 3. The seismic mass of the ground level is 9. The force distributions. we select a weak ground motion: the northsouth component of the El Centro (1940) ground motion scaled down by a factor of 4.

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

is shown in Figs.487 −1.5 0 0.51 0.728 2.12 0.12) and (3.85 sec 2 Floor 5th 4th 3rd 2nd 1st T1 = 2. Force distributions s* = mφn . and 3 n 3.75 1.03 −1.33 2.27 sec Ground −1.2 Response History Analysis The structural response due to individual vibration modes.5.13)].61 2.4.796 0.0272 −2.31 −0. 3.1 −2.1 3.38 0.44 1.39 3. Each figure is organized in four parts: (a) shows the roof displacement urn t . 3.49 sec 3 T = 0. (c) shows the joint rotation q n t of an external joint at 16 bg bg bg . 2.72 −2.5 Fig. (3. and 3 .05 1. First three natural-vibration periods and modes of the 9story building 3. 3.67 −1.05 s * 1 s * 2 s * 3 Fig.05 2.8 −2. n = 1. respectively.5 −1 −0.3.94 2. determined by RHA [Eqs. and 3.6. n = 1.04 1. 2. (b) shows the base shear Vbn t normalized relative to the weight W of the building.9th 8th 7th 6th T = 0.37 2. 3.93 −1.13 −1.7.4.5 Mode Shape Component 1 1.

consistent with the increasing significance of the higher mode response among these three sets of response quantities.422 cm. two. or all modes.15) and (3.3. errors are smallest in floor displacements. and 3. where the second and third modal responses are a larger percentage of the top story drift compared to roof displacement. Combining the modal response histories for all modes gives the total response [Eqs. and ur 3o = 0. drifts in all stories. which are listed in Tables 3. Also included are the combined response due to one. larger in story drifts.2. and the percentage errors due to truncation of higher modal contributions. For a fixed number of modes included. the results for the roof displacement and top-story drift are shown in Fig.16)]. 3. two.9. are presented in Fig. The same method was used to determine the peak values of many response quantities. (3. 3. with the first three modes provide essentially the exact response. and 3.23 cm. and (d) shows the Vbn . The peak values of displacements of all floors. respectively. three. The linear relationship between the base shear and roof displacement for each mode implies that the structure did not yield. in particular. 3. 3. errors generally decrease as response contributions of more modes are included.3.8.12 cm. 17 . the exact response considering all modes. and three vibration modes. including one. ur 2o = 2. and rotations of external joints with moment connections are presented in Tables 3.8. 3.the roof level. The peak values of floor displacements and story drifts determined by RHA. The errors due to truncating the contributions of vibration modes beyond the third mode are negligible. This is illustrated in Fig. As expected. It is apparent that the first mode alone is inadequate. The peak values of the various response quantities are noted in these figures.1.urn relation. the peak roof displacement due to each of three modes is ur1o = 9. and even larger in joint rotations.1. especially in estimating the story drifts. respectively. but three modes—perhaps even two modes—are sufficient.2.

003 0.11E-03 1.13E-03 2.1 -2.65E-03 RHA (all modes) 2.6 4.035 0.65E-03 2.259 0.72E-03 3.5 -2.4 -10.300 0.042 0.7 2 Modes -3.124 0.13E-04 9.407 0.26E-04 -5.01E-04 3.152 0.097 0.130 0.63E-03 2.11E-04 -5.28E-04 1.060 -0.7 7.9 3.225 0.333 0.6 -1.25 × 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 0.44E-03 1.5 -1.85E-03 3.133 -0.99E-03 Mode 3 3.1 -2.3 -0.009 -0.311 0.4 -53.6 -0.5 18.266 0.09E-03 2.74E-04 6.56E-03 2.2 -4.069 0.173 0.7 -19.310 0.25 × El Centro ground motion Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Modal Response Mode 1 Mode 2 Drift Ratio (%) Combined (RHA) 1 Mode 2 Modes 3 Modes Mode 3 0.303 0.9 9.9 1.475 0.325 0.229 0.227 0.156 0.157 0.0 -10.089 0.47E-03 1.266 0.73E-03 3.9 -22.90E-03 3.31E-03 Mode 2 -1.4 -1.062 0.9 -16.069 0.183 0.03E-03 1.3 -33.38E-03 2.453 0.181 0.259 0.38E-04 2 Modes 2.0 -2.043 0.263 0.9 -15.3 Peak values of joint rotations (radians) from RHA for 0.282 0.058 -0.088 -0.237 0.4 -1.265 RHA (all modes) 0.08E-03 2.045 0.012 -0.5 0.88E-03 2.378 0.023 0.14E-03 2.024 -0.400 0.226 0.192 0.202 0.50E-03 4.0 -2.265 0.1 Table 3.7 -50.1 -11.1 Building height is measured from the ground floor to the 9th floor. 18 .370 0.09E-03 1.203 0.261 -0.466 0.74E-03 1.3 -0.4 -22.6 11.66E-05 -3.117 0.1 4.8 -56.03E-03 1.202 0.295 -0.002 -0.152 0.8 -5.235 0.275 0.00E-03 2.88E-03 2.94E-03 2.173 0.9 2.00E-03 1.125 0.4 -3.7 3.001 -0.2 -1.258 0.74E-04 9.264 Error (%) 1 Mode 2 Modes 3 Modes -23.121 0.76E-03 1.22E-03 2.026 0.74E-03 1.2 9.8 1.399 0.406 0.080 0.1 Peak values of floor displacements (as % of building height = 37.89E-03 1.1 -0.3 Table 3.33E-03 2.282 0.253 0.227 0.011 0.4 -0.Table 3.01E-04 -2.205 0.2 0.179 0.197 0.8 -10.4 -41.008 -0.2 1.45E-03 3.6 9.321 0.364 0.78E-04 -3.14 m3) from RHA for 0.038 0.4 -7.124 0.03E-03 -6.1 -19.042 0.31E-03 Joint Rotation (rad) Combined (RHA) 1 Mode 2.3 -3.245 0.9 -23.011 0.25 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 3 Modal Response Mode 1 2.156 0.99E-03 2.8 -15.237 0.3 19.4 -6.76E-03 1.9 -24.253 0.2 -2.062 -0.010 -0.245 0.8 -1.0 -0.6 0.235 0.2 -20.060 -0.29E-03 2.4 0.260 0.09E-03 Error (%) 1 Mode -23.91E-04 1.6 0.401 Error (%) 1 Mode 2 Modes 3 Modes -23.229 0.413 RHA (all modes) 0.202 0.7 4.15E-03 4.015 -0.055 0.2 Peak values of story drift ratios (as % of story height) from RHA for 0.260 0.006 0.266 0.097 0.1 -14.50E-03 2.060 0.1 -2.307 0.311 0.226 0.64E-03 3.090 0.0 -46.071 0.7 2.44E-03 3.0 3 Modes -5.097 0.350 0.63E-03 2.159 0.3 -8.42E-04 1.012 0.199 0.9 8.125 0.00E-03 1.6 1.24E-03 2.032 -0.0 7.022 0.2 -57.1 3.054 0.177 0.336 0.03E-03 3 Modes 2.89E-03 1.231 -0.177 0.8 1.09E-03 1.6 -1.

(b) base shear.1 0.5 0 u (cm) r1 7. (c) joint rotation.1 /W 0. 3.0435 Vb1 / W 0. (d) force-deformation history.12 0 (d) 0 (e) 15 Fig.0435 • 0 (b) V b1 −0.1 −15 −7. and (e) pushover curve.5.25 × El Centro ground motion 19 .1 0.5 9.5 0 u (cm) r1 7.12 −0. Response due to first mode: (a) roof displacement.12 • 0 (a) r1 −15 0.1 5 x 10 −3 θ (rad) 0.0435 9. Excitation is 0.15 u (cm) 9.5 15 −0.1 −15 −7.00174 • 0 (c) r1 −5 0 5 10 15 Time (sec) 20 25 30 0.

5 u (cm) r2 0 • 2. Response due to second mode: (a) roof displacement.00244 • 0 (c) r2 −5 0 5 10 15 Time (sec) 20 25 30 0.05 5 x 10 −3 b2 θ (rad) 0. (b) base shear.5 5 −0.1 −5 −2.23 −2. (d) force-deformation history.25 × El Centro ground motion 20 .1 Vb2 / W 0. 3.05 /W 0.1 0.0252 (d) 0 0 (e) −0.0252 −2.23 0.5 0 u (cm) r2 2. and (e) pushover curve. (c) joint rotation.6.5 0 u (cm) r2 2.5 5 Fig.1 −5 −2.0252 • 0 (b) V −0.23 −5 (a) 0. Excitation is 0.

422 (e) −0. (c) joint rotation.02 −0.0108 x 10 −3 (b) −0.000913 • 0 (c) r3 −2 0 5 10 15 Time (sec) 20 25 30 0.422 −1 (a) 0.25 × El Centro ground motion 21 . Response due to third mode: (a) roof displacement.5 1 −0.5 1 Fig. (d) force-deformation history.1 u (cm) r3 0 • 0.02 −1 −0.0108 −0.422 /W V b3 0 (d) 0 −0. 3.0108 −0.02 2 θ (rad) 0. Excitation is 0.7. (b) base shear.5 0 u (cm) r3 0.02 −1 −0.02 /W V b3 0 • 0.02 0. and (e) pushover curve.5 0 u (cm) r2 0.

In RHA the errors arise only from truncating the responses due to higher modes. although both analyses led to identical peak values of the individual modal responses. i. except for the algebraic sign associated with Γ1 and minor round-off errors.5e. story drifts.9 and Tables 3. respectively.3).4. two.20) with n = 1 (Fig. 3. As suggested by Eq. i. or three vibration modes. respectively.1 through 3. The floor displacements. 3.3). This pushover curve is consistent with the relationship between the base shear and roof displacement determined by RHA (Fig.5.e. respectively.4) to roof displacement ur1o = 9. These values of the response quantities are identical to the peak response values determined from RHA (Tables 3. and 3. 3. (3. and external joint rotations computed by pushover analysis are presented in Tables 3. 3. considering one.3).18).6. and 3. (3.422 cm .6e and 3. and the errors in these estimates relative to the exact response from RHA considering all modes. (3.e. 3..20) with n = 2 and 3 up to roof displacements ur 2 o = 2.4. Figure 3. which would provide the Dn value to be substituted in Eq.23 cm .5d). pushing the structure.3. the floor displacements are proportional to the mode shape φ 1 because the structure remains elastic. Implementing pushover analysis for the second and third modes.1. implying that the modal truncation errors are small if at 22 .2. this value would be determined directly from the response (or design) spectrum. and ur 3o = 0. Observe that the target roof displacement in each pushover analysis is identical to its exact value determined by RHA. confirming that MPA gives the exact values of the individual modal responses.4. pushing the structure using the force distribution of Eq. 3. As for the first mode. the value determined by RHA (Fig.5.1.6d and 3.10 and Tables 3. the errors in the MPA results are generally larger than in RHA (Fig.6. story drifts.9).8) leads to the pushover curve shown in Fig.5.21). 3. 3.12). these pushover curves are consistent with the Vb . For a fixed number of modes included considered. 3.12 cm. In practical application.4. 3.ur relations determined by RHA (Figs.7e and to the floor displacements. and the computed response values are identical to the peak response values determined from RHA (Tables 3.3 Modal Pushover Analysis Implementing MPA for the fundamental vibration mode. (3. using the force distribution of Eq. leads to the pushover curves shown in Figs.2. 3. 3. 3. and 3.6 present estimates of the combined response according to Eq.7d). and external joint rotations in Tables 3. and it is apparent in the example considered that three modes provide most of the response (Fig.. and 3. (3. and 3.

8. (a) Roof Displacement 15 (b) Top Story Drift 3 Mode 1 ∆ (cm) u (cm) 9. Response histories of roof displacement and top-story drift from RHA for 0.422 r2 0 • 1.12 • r1 −15 15 0 r1 0 0 0.2 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.6 Fig.5 Ground 0 0. sec 25 30 r 5 10 15 20 Time.03 −3 3 Mode 3 ∆ (cm) u (cm) r3 −15 15 9.eps r 5 10 15 20 Time.83 • r3 0 • 0.48 −3 3 1.1 0. 3.685 • −3 3 Mode 2 ∆ (cm) u (cm) • 2.25 × El Centro ground motion 23 .3 0. sec 25 30 Fig. Heightwise variation of floor displacements and story drifts from RHA for 0. 3.5 0.least three modes are included.59 • Total (All Modes) 0 −15 0 ∆ (cm) u (cm) 0 −3 0 fig3_8.9. Additional errors are introduced in pushover analysis due to the approximation inherent in modal combination rules.2 0.3 0.eps fig3_9b.eps Ground 0 0.1 0.4 Story Drift Ratio (%) 0.23 r2 −15 15 0 • 0.4 Displacement/Height (%) 0.

133 0.229 0.33E-04 5.53E-04 -9.259 -0.026 0.31E-03 2.001 0.235 0.63E-03 2.09E-04 -3.106 0.03E-03 1.9 -16.04E-03 3.5 Peak values of story drift ratios (as % of story height) from MPA for 0.097 0.2 -12.9 -11.227 0.4 -22.177 0.03E-03 6.058 -0.72E-03 3.6 -15.0 -2.385 0.1 -18.203 0.235 -0.3 -14.272 0.080 0.270 0.045 -0.3 -41.43E-04 -1.011 -0.4 -4.6 -19.152 0.466 0.7 -50.253 0.97E-03 1.38E-03 2.328 0.89E-03 2.097 0.253 -0.0 -18.7 -15.76E-03 1.125 0.088 -0.1 -11.179 0.313 0.6 -16.4 -2.7 2 Modes -13.253 0.00E-03 1.3 -12.261 -0.15E-03 1.270 0.023 -0.230 0.237 -0.267 0.181 0.226 0.260 0.012 -0.4 -53.0 -0.3 -2.8 -15.133 -0.2 -57.22E-03 -2.03E-03 3.4 -9.9 -8.042 0.76E-03 -1.121 0.245 0.286 0.9 -15.227 0.006 -0.060 0.25 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Modal Response Mode 1 -2.002 0.009 0.0 -10.6 Peak values of joint rotation (radians) from MPA for 0.310 0.062 0.1 -19.062 -0.9 -15.203 0.74E-03 1.331 0.336 RHA (all modes) 0.133 0.125 0.285 0.8 -22.74E-03 -1.0 -46.270 0.2 1.253 RHA (All Modes) 0.00E-03 2.Table 3.015 -0.124 0.9 -14.003 -0.7 -21.2 -20.090 0.44E-03 3.4 -7.4 -19.9 -13.3 -13.407 0.229 0.036 -0.259 0.078 0.03E-03 -1.5 -16.57E-03 1 Mode -23.99E-03 Mode 3 -3.00E-03 -1.9 -13.3 -33.9 -24.9 -14.25 × 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 0.00E-03 2.80E-04 3.322 0.40E-04 5.282 0.260 -0.9 -13.89E-03 -2.2 -4.276 0.4 0.060 -0.4 Peak values of floor displacements (as % of building height = 37.1 -11.038 -0.237 0.230 -0.8 -15.90E-04 -9.9 -18.010 0.048 0.011 0.152 0.3 -19.31E-03 Error (%) RHA (all modes) 2.73E-03 3.28E-03 2.203 0.9 -15.8 -14.08E-03 2.173 0.47E-03 1.069 0.179 0.370 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 -0.9 -14.231 -0.12E-03 1.374 0.9 3 Modes -12.156 0.117 0.3 -9.069 0.7 24 .203 0.44E-03 -1.321 0.332 0.7 2.14 m) from MPA for 0.055 0.3 1.0 -16.024 0.012 0.42E-04 -1.05E-03 3.09E-03 1.6 -17.336 0.177 0.282 -0.89E-03 2.3 Table 3.15E-03 2.266 0.7 -19.94E-03 2.267 -0.008 -0.96E-03 2.8 -23.300 0.032 0.173 0.00E-03 2.264 Error (%) Modal Response 1 Mode 2 Modes 3 Modes -23.274 0.9 -12.22E-04 2.197 0.1 -0.2 -11.260 0.079 0.156 0.4 -4.90E-03 1.157 0.2 -0.4 -11.401 Error (%) 1 Mode 2 Modes 3 Modes -23.9 -15.09E-03 -1.38E-03 3 Modes 2.4 -9.4 1.24E-03 -2.89E-03 -1.31E-03 Joint Rotation (rad) Combined (MPA) 1 Mode 2.4 -14.022 0.63E-03 -2.106 0.78E-04 2 Modes 2.2 Table 3.00E-03 3.296 -0.4 -5.043 -0.048 0.73E-05 3.2 -16.65E-03 2.042 0.071 0.92E-04 -1.267 0.09E-03 Mode 2 1.4 -22.74E-04 -6.245 -0.203 0.8 -56.89E-03 1.9 -13.157 0.

1 0. shading indicates modal combination error 25 .1 0.eps fig3_10b.25 × El Centro ground motion. 3.5 Ground 0 0.10.eps Ground 0 0.4 Displacement/Height (%) 0.6 Fig.4 Story Drift Ratio (%) 0.3 0. Heightwise variation of floor displacements and story drifts from MPA for 0.3 0.2 0.5 0.2 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.

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). sign u b g g bg (4. the initial loading curve is idealized as bilinear. (4.2) The standard approach is to solve directly these coupled equations.mι ug t b (4. Expanding the displacements of the inelastic system in terms of the natural vibration modes of the corresponding linear system we get ut = bg n =1 Â φ nqn bt g 27 N (4. leading to the “exact” nonlinear response history analysis (RHA).1) With this generalization for inelastic systems. Both systems have the same mass and damping.4 4. Although classical modal analysis (Section 3. but depend on the history of the displacements: f s = f s u.1 Inelastic Multistory Buildings RESPONSE HISTORY ANALYSIS For each structural element of a building. Therefore.2) to the modal coordinates of the corresponding linear system. sign u = .3) .1) becomes mu + cu + f s u.1) is not valid for inelastic systems. Each structural element of this elastic system is defined to have the same stiffness as the initial stiffness of the structural element of the inelastic system. (3. and the unloading and reloading curves differ from the initial loading branch. Thus. it is useful for later reference to transform Eq. the relations between lateral forces f s at the N floor levels and the lateral displacements u are not single valued. Eq.

… N (4. (4. (4.2).6) for inelastic systems will no longer be described by Eq. (4. implying coupling of modal coordinates because of yielding of the structure. The equations governing the response of the inelastic system to p eff.” implying that other “modes” will also contribute to the solution.Gn ug t Mn bg n = 1. sign u = . 4.3).4)] leads to the uncoupled modal response history analysis (UMRHA) procedure. 2. (3.9) for linearly elastic systems.4) is rarely solved because it offers no particular advantage over Eq. however. This approximate RHA procedure is the preliminary step in developing a modal pushover analysis procedure for inelastic systems.and classical n damping-orthogonality property of modes gives F qn + 2z nw n qn + sn = .Substituting Eq. (3. Unlike Eq.6) The solution of Eq. these equations are coupled for inelastic systems. (4. in principle. sign qn = φ n f s u n .n t given by Eq.3) in Eq.2).9) involves T Fsn = Fsn qn .5) This resisting force depends on all modal coordinates qn t . (4. and using the mass. However. Simultaneously solving these coupled equations and using Eq. where φ n are now the modes of the corresponding linear system. The spatial distribution s of the effective earthquake forces is expanded into the modal contributions sn according to Eq. (4.sn ug t bg b g bg bg (4. give the same results for u t as obtained directly from Eq.8) because qr t will generally be nonzero for “modes” other than the nth “mode. For linear systems.4) represents N equations in the modal coordinates qn . Equation (4. (3. (4. qr t = 0 for all 28 bg .2). premultiplying by φ T . Eq. sign u n b g b g bg (4. (3. (4.6b) are mu + cu + f s u.4) where the only term that differs from Eq.2 UNCOUPLED MODAL RESPONSE HISTORY ANALYSIS bg Neglecting the coupling of the N equations in modal coordinates [Eq. (3.3) will.

(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

89 50 r1 0 u (cm) 0 • 78. Roof displacement due to p eff. ζn. 4. n = 1.575 5 10 15 20 Time (sec) 25 30 ur3 (cm) 0 • 5. 2.02 −150 n=2 −150 50 n=2 14. Fsn / Ln ¨ g(t ) u (a) Static Analysis of Structure (b) Dynamic Analysis of Inelastic SDF System Fig.n t = .Forces sn Unit mass An(t ) st rn ωn. and (b) approximate solution by UMRHA bg af bg 32 .9 • 0 −50 10 −50 10 n=3 ur3 (cm) 0 • 5. 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. 4. and 3.112 5 10 15 20 Time (sec) 25 30 −10 0 −10 0 Fig.51 • ur2 (cm) 0 ur2 (cm) n=3 14.4.s n u g t . where u g t = 3.3.0 ¥ El Centro ground motion: (a) exact solution by NL-RHA.

744 • 0 −20 10 n=3 −20 10 n=3 ∆r3 (cm) 0 • 5.Dn relation is 33 bg bg .(a) Nonlinear RHA 20 n=1 20 (b) UMRHA n=1 • 5.0 ¥ El Centro ground motion.6) by UMRHA is compared with the “exact” solution by nonlinear RHA. n = 1.008 • 6.0 ¥ El Centro ground motion: (a) exact solution by NL-RHA.n t (n = 1.2… N ) according to Eq.2. The errors are slightly larger in drift than in displacement.956 5 10 15 20 Time (sec) 25 30 ∆r3 (cm) 0 • 5. (2) the superposition of responses to p eff.5. 4. These errors arise from the following assumptions and approximations: (1) the coupling among modal coordinates qn t arising from yielding of the system [recall Eqs. and 3.s n u g t . this intense excitation was chosen to ensure that the structure is excited well beyond its linear elastic limit. Such comparison for roof displacement and top-story drift is presented in Figs.n t = . 4. respectively.15) is strictly valid only for linearly elastic systems.4 and 4. (4. and (b) approximate solution by UMRHA af bg bg 4. the errors in either response quantity are only a few percent. Top story drift due to p eff.1 Underlying Assumptions and Accuracy The approximate solution of Eq.4) and (4. both for 3. (4. 2.855 ∆r1 (cm) 0 ∆r1 (cm) n=2 • 6.33 0 −20 20 −20 20 n=2 ∆r2 (cm) 0 ∆r2 (cm) 7.5. and (3) the Fsn Ln .817 5 10 15 20 Time (sec) 25 30 −10 0 −10 0 Fig.5)] is neglected. where u g t = 3. but even for this very intense excitation. (3.

approximated by a bilinear curve to facilitate solution of Eq. Therefore.12) for floor displacements. 4.1. The structure is pushed using the force distribution of Eq. (4. Although most commercially available software cannot implement such displacement-controlled analysis. Therefore we impose this constraint in developing the UMRHA procedure in this section and modal pushover analysis in the next section. However. How to convert this Vbn . (3.6a. The overall errors in the UMRHA procedure are documented in the examples presented in Section 4. the only force distribution that produces displacements proportional to f n is given by Eq. within the linearly elastic range of the structure.Dn relation to be determined in Eq. (4.10) .2. (3.4. and the base shear Vbn . Although several approximations are inherent in this UMRHA procedure. the relationship between lateral forces f s and Dn in Eq. At the yield point.20). when specialized for linearly elastic systems it is identical to the RHA procedure of Section 3.20) to some bg pre-determined roof displacement. When implemented by commercially available software. (3.9) should be determined by nonlinear static analysis of the structure as the structure undergoes displacements u = Dnφ n with increasing Dn . (4.2 Properties of the nth-“mode” Inelastic SDF System How is the Fsn Ln . this distribution seems to be a rational choiceeven for inelastic systemsto determine Fsn in Eq. such nonlinear static analysis provides the so-called pushover curve.Dn relation? The two sets of forces and displacements are related as follows: V Fsn = bn Γn Dn = urn Γ nf rn 34 (4. it can conduct a force-controlled nonlinear static analysis with an invariant distribution of lateral forces.9). the base shear is Vbny and roof displacement is urny .8) before it can be solved? Because Eq. A bilinear idealization of this pushover curve for the nth-“mode” is shown in Fig. is plotted against roof displacement urn . 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.Dn curve. (4. which is different than the Fsn Ln .8) governing Dn t is based on Eq. (4.urn pushover curve to the Fsn Ln . 4.8) in UMRHA.

(4.13) This value of Tn .6.12) implying that the initial slope of the curve in Fig. The two are related through Fsny Ln 2 = w n Dny (4. In contrast. should be used in Eq.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.6b is w 2 .10 enables conversion of the pushover curve to the desired Fsn Ln − Dn relation shown in Fig. 4.8). 2001. 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 (4.11). 4. where the yield values of Fsn Ln and Dn are Fsny Ln = Vbny * Mn Dny = urny Γ nf rn (4. the initial slope of the pushover curve in Fig. Properties of the nth-“mode” inelastic SDF system from the pushover curve Equation 4. 35 . which is not a meaningful quantity. Section 13.6b. Knowing Fsny Ln and Dny from n Eq. which may differ from the period of the corresponding linear system.2. (4.5). 4.6a is 2 kn = ω n Ln . 4.11) * in which M n = Ln Γn is the effective modal mass (Chopra.

Calculate the peak value.20)]. the peak value of Dn t . Calculate histories of various responses by Eqs. for linearly-elastic vibration of the building. Dn (t ) . (3. the first two or three modes will suffice. n 3. and pseudo-acceleration history. Combine the “modal” responses using Eqs.6b. 6. An t .12) and (3. Convert the idealized pushover curve to the Fsn Ln − Dn relation (Fig. (3.6a). 8. 4.2.20)]. which is summarized next as a sequence of steps. φn . and modes. 2. n (3. Typically. 4. of the nth-“mode” inelastic SDF system (Fig. of the total response r t obtained in Step 8. For the nth-mode. Idealize the pushover curve as a bilinear curve (Fig.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. develop the base-shear – roof-displacement ( Vbn − urn ) pushover curve for the force distribution s* [Eq.16) to determine the total response. ωn . (4.13).3b) with force-deformation relation of Fig. 5. 4. with the structure is pushed to the roof displacement urno .11).21) where Dn . 4.4.15) and (3.3 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 . (3. Compute the natural frequencies. details are available in Appendix A: 1. 9. 4. Repeat Steps 2 to 6 for as many modes as required for sufficient accuracy. 7. 4. This value of the roof displacement is given by Eq. Compute the deformation history. Consider a nonlinear static analysis of the structure subjected to lateral forces distributed over the building height according to s* [Eq.6b) by utilizing Eq. (3. r o .n t . is now determined by 36 bg bg .

Steps 1 to 4 of the MPA are same as those for UMRHA.n t . This application of modal combination rules to inelastic systems obviously lacks a theoretical basis.2. rno also represents the exact peak value of the nth-mode contribution rn t to response r t .6b by solving Eq. 37 . as described in Section 4. (4. This pushover analysis.g. we will refer to rno as the peak “modal” response even in the case of inelastic systems. for elastic systems. At this roof displacement. each determined by one pushover analysis. Calculate the peak roof displacement urno associated with the nth-“mode” inelastic SDF system from Eq. (4.8).solving Eq. Eq. As shown in Sections 3.12). it seems reasonable because it provides results for elastic buildings that are identical to the well-known RSA procedure described in Section 3.3) because. 6.18). 5. it can be determined from the inelastic response (or design) spectrum (Chopra. or from the inelastic response (or design) spectrum.3. etc. joint rotations. (3. bg bg bg bg Thus. alternatively. 2001.3b) with force-deformation relation of Fig. are combined using an appropriate modal combination rule. However. plastic hinge rotations.. The peak “modal” responses rno . which is summarized next as a sequence of steps. 4. Dn . 4. story drifts.3. e. (3. of the nth-“mode” inelastic SDF system (Fig. Compute the peak deformation.2. 4. (4. the pushover analysis provides an estimate of the peak value rno of any response rn t : floor displacements. Sections 7. governed by Eq. to obtain an estimate of the peak value ro of the total response.8). seems reasonable. details are available in Appendix B. and (2) it provides the exact modal response for elastic systems.2 and 3.6).21). The response value rno is an estimate of the peak value of the response of the inelastic system to p eff.4. the lateral force distribution used possesses two properties: (1) it appears to be the most rational choice among all invariant distribution of forces. although somewhat intuitive for inelastic buildings.6 and 7. as mentioned earlier.1 Summary The peak inelastic response of a building to earthquake excitation can be estimated by the MPA procedure just developed. It provides results for elastic buildings that are identical to the well-known RSA procedure (Section 3.

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. three “modes” included. The peak values of displacements of all floors and drifts in all stories are presented in Tables 4. determined by the UMRHA procedure. and ur 3o = 2.18).8. Repeat Steps 3 to 8 for as many “modes” as required for sufficient accuracy. Figure 4. (3.1 and 4. also included are the combined responses due to one. 4. The peak values of response are as noted. Determine the total response by combining the peak “modal” responses using the SRSS combination rule of Eq. 4. the peak roof displacement due to each of the three “modes” is ur1o = 48. is presented next.4. The peak values of floor displacements and story drifts including one. 4. 9. and the errors in the approximate results are shown in Fig. the 38 .0.0 to 3. 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.1).53 cm. say. and the “exact” response from nonlinear RHA for the roof displacement and top-story drift. respectively. the UMRHA procedure lacks a rigorous theory. This is to be expected. with. and three modes are compared with the “exact” values in Fig. and three “modes”.5 ¥ the El Centro ground motion including the response contributions associated with three “modal” inelastic SDF systems. 8.2). in contrast to modal analysis (Section 3.1 Uncoupled Modal Response History Analysis The structural response to 1. and the percentage errors in the approximate results. rno . two. two. the combined response due to three “modes”. This deficiency also implies that.7 cm. and compared with the results of a rigorous nonlinear RHA using the DRAIN-2DX computer program. Observe that errors tend to decrease as response contributions of more “modes” are included.7.4. From the total rotation of a plastic hinge. subtract the yield value of hinge rotation to determine the hinge plastic rotation. Typically. in particular.2. the first two or three “modes” will suffice. ur 2o = 11. extract from the pushover database values of other desired responses.7 shows the individual “modal” responses. At urno . the “exact” results.3 cm. although the trends are not systematic as when the system remained elastic (Section 3.

44 • "Mode" 2 r1 0 • 48.3 11.0.6 5 10 15 20 Time. For this purpose the UMRHA and exact analyses were repeated for ground motions of varying intensity. sec 25 30 r −12 0 Fig.7.1 and 4.85.2). 2.5.24 • NL−RHA r −12 12 ∆r (cm) u (cm) 0 5 10 15 20 Time. and 3.1 −12 12 ∆r (cm) u (cm) UMRHA 3 "Modes" 0 • 7.25 (Tables 3. defined as the El Centro ground motion multiplied by 0. For each excitation. (a) Roof Displacement (b) Top Story Drift 80 12 0 • 3. the errors in story drifts are larger compared to floor displacements.25. Response histories of roof displacement and top-story drift due to 1. 1.0.75.7 • −80 80 ∆r1 (cm) u (cm) "Mode" 1 −12 r2 0 ∆r2 (cm) 12 0 u (cm) −80 80 −12 r3 0 • 2.response is much less accurate (Tables 4. just as for elastic systems.2) if the system yields significantly versus if the system remains within the elastic range (Tables 3. 0.5. the errors in responses computed by UMRHA including three “modes” relative to the “exact” response were determined. 4.88 u (cm) "Mode" 3 −80 80 0 −80 80 0 −80 0 • 44. in particular.2).38 6. for a fixed number of “modes” included.1 and 4.1 and 3. 1. sec 25 30 • 48.1 and 3. 0.53 ∆r3 (cm) 12 0 • 2.62 5. 0. we investigate how the errors in the UMRHA vary with the deformation demands imposed by the ground motion. the degree to which the system deforms beyond its elastic limit. and total response from NL-RHA 39 .0. However.5 × El Centro ground motion: individual “modal” responses and combined response from UMRHA.2) and 1.5 (Tables 4. recall that the computed errors have been presented earlier for ground motion multipliers 0. Next.

5 Displacement/Height (%) 2 Ground 0 0.10b).5 1 1.10 summarizes the error in UMRHA as a function of ground motion intensity. (3.5 ¥ El Centro ground motion Figure 4. 4. it will be useful to know the deformation of the system relative to its yield deformation. Heightwise variation of error in floor displacements and story drifts estimated by UMRHA including one.10a). or three “modes” for 1. Shown is the error in each floor displacement (Fig. 4.5 1 1. 4. Two versions of the pushover curve are 40 .8.9.5 Fig. 4. with the peak displacement of each “modal” SDF system noted for each ground motion multiplier. the pushover curves using force distributions s* [Eq. 4. indicated by a ground motion multiplier. two. and the error envelope for each case.5 Story Drift Ratio (%) 2 2.11.(a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th NL−RHA UMRHA 1 "Mode" 2 "Modes" 3 "Modes" (b) Story Drift Ratios Floor Floor 5th 4th 3rd 2nd 1st 5th 4th 3rd 2nd 1st NL−RHA UMRHA 1 "Mode" 2 "Modes" 3 "Modes" Ground 0 0. To interpret these results. For this purpose. Heightwise variation of floor displacements and story drift ratios from UMRHA and NL-RHA for 1. in each story drift (Fig.5 × El Centro ground motion (a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th (b) Story Drift Ratios Floor 4th 3rd 2nd 1st UMRHA 1 "Mode" 2 "Modes" 3 "Modes" Floor 5th 5th 4th 3rd 2nd 1st UMRHA 1 "Mode" 2 "Modes" 3 "Modes" Ground −60 −40 −20 0 Error (%) 20 40 60 Ground −60 −40 −20 0 Error (%) 20 40 60 Fig.20)] for the first three n modes of the system are shown in Fig.

3.11). The idealized curve for the first “mode” deviates most from the actual curve near the peak displacement corresponding to ground motion multiplier 1. Additional errors are introduced in UMRHA of systems responding beyond the linearly elastic limit for at least two reasons.7). the values determined by RHA of the nth-mode inelastic SDF system (Fig. as mentioned in Section 3.4) to roof displacements urno = 48. the pushover curve for each “mode” is idealized by a bilinear curve in solving Eq.6 and 4. (2) increase rapidly as the ground motion multiplier increases to 1.75.10) for each “modal” inelastic SDF system (Figs 4.4.11a).75.2 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. drifts in all stories (Table 4. First. UMRHA lacks a rigorous theory and is based on several approximations.4). the peak values of displacements at all floors (Table 4. For more intense excitations. (3.20) with n = 1. even though the system remains essentially elastic. the first reason mentioned above seems to be the primary source for the errors. 41 . Second. Each of these three pushover analyses provides the pushover curve (Fig. and plastic hinge rotations at the external beam end at each floor level (Table 4.3). and 3 (Fig.5). 4. The location of plastic hinges and their rotations. Figure 4.2.2.01 (Fig.0. and (4) are larger in story drift compared to floor displacements.3 cm. respectively. the errors in truncating the higher mode contributions are negligible. 11.4. 4. The structure is pushed using the force distribution of Eq. Perhaps this explains why the errors are large at this excitation intensity. the ductility factor for the first mode system is only 1.included: the actual curve and its idealized bilinear version.5 ¥ the El Centro ground motion. were noted but are not shown here. 2. 4. and 2. and.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. (3) maintain roughly similar values for more intense ground motions.7 cm.0. 4.53 cm. determined from “exact” analyses. (4. The system remains elastic up to ground motion multiplier 0. as mentioned in Section 4.1).

Errors in UMRHA as a function of ground motion intensity: (a) floor displacements.806 0.2 1.0 11.5 -3.003 -31.3 25.763 -15.370 -0.10.201 1.8 14.5 1 1.154 0.256 1.018 0.7 Table 4.0 1.071 0.727 1.214 0.938 -1.616 -0.945 -37.820 -0.(a) Floor Displacements 60 50 40 Error Envelope Error for Floor No.044 1.484 0.3 6.5 × 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 -1.072 1.057 -0.372 -1.3 1.983 1.0 2.216 1.187 -0.9 12.4 1.088 10.201 -0.663 -0.376 1.3 0.9 31.130 0.241 1.055 -0.104 0.1 Peak values of floor displacements (as % of building height = 37.8 1.513 -0.877 0.6 1.0 -9.9 16.410 1.1 1.4 -7.5 × 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” -0.5 9.8 1.3 8.8 1.201 -1.003 0.2 6.863 1.298 -0.350 -0.133 1.811 1.5 28.722 0.942 -0.379 1.698 1.410 -1.068 0.9 31.490 -1.707 1.126 0.472 1.338 -1.072 -1.138 1.260 -15.819 2.806 -0.490 1.844 -25.338 1.033 0.513 0.315 -0.4 -1.298 0.317 0.863 0.5 2 GM Multiplier 2.2 22.376 -1.256 1.6 2.079 0.071 -0.495 1. and (b) story drifts Table 4.120 1.226 -0.820 -19.4 4.430 1.540 0.065 0.096 0.293 1.010 0.372 1.136 1.6 4.241 -1.473 -22.900 -10.5 9.049 -0.121 -0.478 0.2 12.214 -0.14 m) from UMRHA for 1.0 11.248 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.220 0.8 0.5 2 GM Multiplier 2. Noted 60 50 40 (b) Story Drifts Error Envelope Error for Story No.554 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 0.1 1.0 7.333 0.371 -0.202 11.1 3.1 0.7 31.942 1.426 -1.526 -0.366 -0.291 0.220 -0.5 10.1 8.2 1.0 9.371 0.5 3 0 0 0.938 1.5 1 1.2 Peak values of story drift ratios (as % of story height) from UMRHA for 1.9 5.676 0.772 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 1.407 -10.283 1.575 -41.914 -0.3 42 .5 28.373 -0.914 2.200 8.256 -1.8 1.982 9.7 14.852 1.8 1.971 1.135 9.663 0.2 4.5 3 Fig.009 -0.5 18.9 12.668 -23.856 2.751 1.070 1.169 0.366 0. 4.235 -0.

especially in estimating the story drifts (Fig. although the error is recorded as 100% if MPA estimates zero rotation when the nonlinear RHA computes a non-zero value. 4. 4.11c).13 and Tables 4. The first “mode” alone is inadequate. for two or three modes included.13 and Table 4. However. 43 . MPA failed to identify the plastic hinges at the column bases in Fig. 4.4. The drifts are underestimated by up to 13% in the lower stories.4).” respectively. the hinges identified in beams at the 6th floor were at variance with the “exact” results. 4.3 through 4. even if three “modes” are included (Fig.3 and 4. Furthermore. 4.3 and 4. the results were not always accurate. two.9 and Tables 4. considering one.12 and 4.” two “modes. this error is not especially significant because the hinge plastic rotation is very small.5).” and nonlinear RHA. The locations of plastic hinges shown in Fig.12 and Tables 4. For example.13 with Fig.14. overestimated by up to 18% in the middle stories. and three “modes. Obviously. 4.3 and 4.3 and 4.1 and 4.5 present estimates of the combined response according to Eq. 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%. in general.4 with Tables 4. One “mode” pushover analysis was unable to identify the plastic hinges in the upper stories where higher mode contributions to response are known to be more significant. (3.13 and Tables 4. but the third “mode” contributions do not seem especially important (Fig. the errors in the modal pushover results are. 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. significant improvement is achieved by including response contributions due to the second “mode”.18). but was more successful when the excitation was more intense (results are not included). The second “mode” was necessary to identify hinges in the upper stories.4). however. and are within a few percent of the exact values for the upper stories. As shown in Figs. and the errors in these estimates relative to the exact response from nonlinear RHA.Figures 4. 4. were determined by four analyses: MPA considering one “mode. 4. The errors are especially large in the hinge plastic rotations estimated by the MPA procedure. 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. Pushover analysis seems to be inherently limited in computing accurately hinge plastic rotations. the third “mode” does not contribute because this SDF system remains elastic (Fig. significantly smaller than in UMRHA (compare Fig.” and three “modes. 4.14.2). Fortuitously.12 and Tables 4.

25 0.5 0.5 1 0.19 y by 3 2 1.75.5 0.5. 0.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36.0 × El Centro ground motion 44 . V = 7616 kN.9 cm.13 y by Base Shear (kN) 8000 6000 1 1. α = 0.25 1 2 4 6 Roof Displacement (cm) 8 10 Fig.5 2 3 Actual Idealized 2000 0 0 0. V = 5210 kN.75 0.2 cm.85 0.5.5 0. “Modal” pushover curves with peak roof displacements identified for 0.6 cm. 1.11. and 3.85 0.85 Base Shear (kN) 8000 6000 4000 2000 0 0 0. 4.5 2 3 4000 2000 0 0 Actual Idealized 0.25 5 10 15 20 Roof Displacement (cm) 25 (c) "Mode" 3 Pushover Curve 12000 10000 u = 4. α = 0. V = 4952 kN. α = 0.0. 1. 0.25. 2.0.75 0.14 y by Base Shear (kN) 8000 6000 4000 1.75 Actual Idealized 20 40 60 Roof Displacement (cm) (b) "Mode" 2 Pushover Curve 80 12000 10000 u = 9.

The errors are only weakly dependent on ground motion intensity (Fig. and the error envelope for each case. 4. As mentioned in Section 3.5 Story Drift Ratio (%) 2 2.2 and 3.5 × El Centro ground motion. shading indicates errors in MPA including three “modes” Figure 4. the errors in MPA were fortuitously smaller than in UMRHA (compare Figs.4 also apply to MPA. 4. indicated by a ground motion multiplier.10 and 4.5 1 1.3 for elastic systems (or weak ground motions). UMRHA is essentially exact. identified in Section 3. While various sources of errors in UMRHA. implying that the modal combination errors contained in these procedures are acceptable.15). implying excitations intense enough to cause significant yielding of the structure. 4.3. 4. implying excitations weak enough to limit the response in the elastic range of the structure.75. now standard in engineering practice. each story drift (Fig. as discussed in Sections 3. The fact that MPA is able to estimate the response of buildings responding well into the inelastic range to a similar degree of accuracy indicates that this procedure is accurate enough for practical application in building retrofit and design. However.15) for ground multipliers larger than 1.5 Fig. 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 0.4. whereas MPA contains errors inherent in modal combination rules. In this case. the MPA procedure is equivalent to the RSA procedure.5 1 1. 4.15b).12 Heightwise variation of floor displacements and story drift ratios from MPA and NL-RHA for 1.5 Displacement/Height (%) 2 Ground 0 0. Shown is the error in each floor displacement (Fig. the errors in MPA were larger for ground motion multipliers less than 0.0.15 summarizes the error in MPA considering three “modes” as a function of ground motion intensity.15a).4. 45 .

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

763 -14.756 0.435 0.6 7.125 -1.2 -3.049 -0.5 Peak values of hinge plastic rotations (radians) from MPA for 1.6 -44.527 -0.5 1.5 × El Centro ground motion Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th “Modal” Response “Mode” 1 “Mode” 2 Drift Ratio (%) Combined (MPA) 1 “Mode” 2 “Modes” 3 “Modes” Error (%) NL 1 2 3 RHA “Mode” “Modes” “Modes” 1.5 Table 4.0 -50.399 0.3 1.3 -3.351 -0.76E-03 4.980 -0.737 0.2 -100.2 6.4 Peak values of story drift ratios (as % of story height) from MPA for 1.2 1.304 1.53E-03 7.222 0.667 -1.611 0.8 0.5 10.071 -0.00E+00 0.667 1.00E+00 0.50E-10 3.745 1.640 1.00E+00 0.371 -0.105 0.007 1.8 -4.76E-03 4.6 13.728 1.066 -0.641 1.1 13.053 1.00E+00 “Mode” 3 0.980 0.012 1.8 “Mode” 3 -1.00E+00 2 “Modes” 7.233 1.135 -7.9 5.05E-03 2.068 0.9 1.72E-03 7.8 1.190 -0.26E-04 9.6 -7.2 11.Table 4.5 2.1 62.2 0.911 0.154 0.498 1.02E-03 3.503 1.478 0.399 0.895 1.00E+00 0.910 1.756 0.057 0.0 -100.168 -0.407 -27.575 -53.752 1.503 -1.0 -100.00E+00 0.101 -0.00E+00 0.00E+00 0.694 1.02E-03 0.19E-10 3.37E-03 1.1 46.336 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.652 1.259 1.338 1.2 -100.00E+00 0.233 1.5 -6.72E-03 7.781 0.705 1.614 0.8 -29.5 2.088 12.00E+00 0.55E-03 3.60E-04 7.00E+00 0.8 -12.00E+00 0.015 0.00E+00 0.36E-03 6.6 1.8 -29.0 2 “Modes” -32.055 0.00E+00 0.8 7.37E-03 1.14 m) from MPA for 1.2 -4.7 -12.820 -7.197 -0.5 2.9 1.473 -15.066 -0.00E+00 0.071 0.37E-03 1.018 -0.8 -6.55E-03 3.311 0.6 -8.6 -9.02E-03 0.156 -0.9 -4.804 1.200 8.76E-03 4.250 0.705 -1.99E-03 6.683 1.9 2.4 1.00E+00 1 “Mode” 7.37E-04 Error (%) “Modal” Response 1 “Mode” -32.88E-03 0.015 0.36E-03 6.00E+00 0.298 -0.315 0.0 1.7 1.1 46.372 0.982 13.2 -100.668 -13.1 62.1 1.116 1.581 0.9 7.895 1.8 -29.202 8.1 -8.0 -5.1 62.942 6.3 Peak values of floor displacements (as % of building height = 37.518 1.260 -14.414 1.36E-03 6.0 15.00E+00 0.10E-02 9.00E+00 0.426 15.116 1.5 × El Centro ground motion Hinge Plastic Rotation (rad) “Modal” Response Combined (MPA) “Mode” 1 7.018 -0.220 1.517 1.516 0.00E+00 0.8 17.733 1.5 7.00E+00 3.76E-03 4.00E+00 0.220 1.3 -3.3 13.304 -1.176 0.4 -8.737 1.666 Table 4.0 47 .945 -49.00E+00 3 “Modes” 7.018 0.3 11.581 0.36E-03 6.72E-03 7.781 0.594 -1.72E-03 7.305 -0.222 0.8 1.003 -16.00E+00 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th “Mode” 2 0.007 1.298 0.844 -7.6 -44.209 1.331 1.9 0.55E-03 3.02E-03 3.429 0.244 0.118 0.033 -0.8 -6.1 46.053 -1.60E-03 2.266 -0.640 -1.2 0.22E-10 NL RHA 1.88E-03 0.0 -100.009 0.130 0.879 1.9 -100.116 1.5 10.514 -1.687 0.6 13.9 -100.18E-03 7.0 -50.738 1.37E-03 1.900 -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 0.1 18.88E-03 1.0 3 “Modes” -32.7 1.

6 and 4. 4.18) considering one.7). The response contributions of the second and third modes are the same as before (Tables 4. (3. The pushover curves for the second and third modes included in Fig.3 Modal Pushover Analysis with Gravity Loads To evaluate the accuracy of the dynamic response of the system.6 and 4. 4.” however.16 are unchanged. Results of such analyses for the selected building subject to 1.4.5 × El Centro ground motion are presented below. Also presented at these roof displacements are the displacements at all floors (Table 4.4.4. MPA including two “modes” underestimates the displacements of lower floors by up to 6% and overestimates the upper floor displacements by up to 22%. 4.20) with n = 1 (Fig.7 cm and ur 3o = 2. (3. 4.7).20) with n = 1 to a target roof displacement. resulting in the pushover curve shown in Fig.7. They are now included in the pushover analysis of the structure for the first “mode” only.5). Static analysis of the structure for gravity loads provides the initial state—forces and deformations—of the structure. The errors are especially large in the plastic hinge rotations estimated by the MPA procedure even if three “modes” are included (Fig.7).” and the errors in these estimates relative to the exact response from nonlinear RHA. 4. and overestimated by up to 29% in the middle stories and by up to 13% in the upper stories. Significant improvement is achieved by including response contributions due to the second “mode. 4. but it is inadequate in estimating the story drifts (Fig.” these effects will modify the combined modal response as well as the results of nonlinear RHA. Starting with its initial state under gravity loads. the third “mode” contributions do not seem especially important (Fig. the structure is pushed using the force distribution of Eq.3 . 3.6).18 and 48 .17 and 4.8 present estimates of the combined response according to Eq. as are the roof displacements ur 2o = 11.18 and Tables 4.4) to roof displacement ur1o = 52.16a—which is slightly different than the one excluding gravity loads (Fig. and the structure is pushed using the force distribution of Eq. As shown in Fig. (3. the results presented so far did not include gravity load effects.8). Figures 4. and three “modes.17 and Tables 4.6 and 4. two.6 through 4. and plastic hinge rotations at selected external beam end at each floor level (Table 4. 4.11a).53 cm.17 and Tables 4.0 cm.” but not the contributions of higher “modes.18 and Tables 4. drifts in all stories (Table 4. The story drifts are underestimated by up to 7% in the lower stories. Obviously gravity load effects will influence the seismic demands due to the first “mode. The first “mode” alone provides adequate estimates of floor displacements.

and three “modes” and by NL-RHA for 1.5 3 Fig. 4.5 2 GM Multiplier 2. and (b) story drifts 49 . 2−"Modes" •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• • •• (c) 3−"Modes" • • • • • • • •• •• • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• • •• • • • • • • • • • (d) Nonlinear RHA •• •• • •• •• •• •• • •• •• •• •• •• •• •• • •• •• • •• •• •• •• • • • • • • •• •• •• Fig. 1−"Mode" • • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • •• • •• • • • • • •• •• • •• •• •• •• (b) MPA. Errors in MPA as a function of ground motion intensity: (a) floor displacements.8).5 1 1. 4. (a) MPA. Noted 60 50 40 (b) Story Drifts Error Envelope Error for Story No.5 2 GM Multiplier 2.14. Locations of plastic hinges determined by MPA considering one. Most pushover analysis procedures do not seem to compute to acceptable accuracy plastic hinge rotations.Table 4.5 1 1.5 3 0.15. two.5 × El Centro ground motion (a) Floor Displacements 60 50 40 Error Envelope Error for Floor No. Noted Error (%) 30 20 10 0 0 2 8 3 4 1 9 Error (%) 30 20 5 4 6 5 7 2 9 10 0 0 1 8 7 3 6 0.

and the error envelope for each case. These errors are only slightly larger than those in Fig. 4. Figure 4. indicated by a ground motion multiplier.The locations of plastic hinges shown in Fig. 4. 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.20 summarizes the error in MPA considering three “modes” as a function of ground motion intensity. With two modes included in MPA.20b). 4. 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. each story drift (Fig.15. 50 .” two “modes”. and nonlinear RHA. excluding gravity load effects. 4. MPA provides response values accurate enough for practical application in building retrofit or design.20a). and the errors are only weakly dependent on ground motion intensity.19 were determined by four analyses: MPA considering one “mode.

0.6 cm. “Modal” pushover curves with gravity loads included.5 0.14 y by Base Shear (kN) 8000 6000 4000 1.5 1 0. and 3. 4.75 0.75.5 2 3 4000 2000 0 0 Actual Idealized 0. 0.5 0. noted are peak values of roof displacement for 0.13 y by Base Shear (kN) 8000 6000 1 1. α = 0.16.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36.25 5 10 15 20 Roof Displacement (cm) 25 (c) "Mode" 3 Pushover Curve 12000 10000 u = 4. 0.85 Base Shear (kN) 8000 6000 4000 2000 0 0 0.9 cm. α = 0.0 × El Centro ground motion 51 .75 0.85 0. 0. V = 7433 kN.3 cm.85 0. V = 5210 kN. V = 4952 kN.85. 2.25 0.50.25 1 2 4 6 Roof Displacement (cm) 8 10 Fig. α = 0.75 Actual Idealized 20 40 60 Roof Displacement (cm) 80 (b) "Mode" 2 Pushover Curve 12000 10000 u = 9.25.19 y by 3 2 1.0.5 0. 1.5 2 3 Actual Idealized 2000 0 0 0.

4. 4. Heightwise variation of floor displacements and story drift ratios from MPA and NL-RHA for 1. and 52 .17. gravity loads included.5 1 1.5 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. story drifts.5 Displacement/Height (%) 2 Ground 0 0. Errors in floor displacements.5 × El Centro ground motion. and hinge plastic rotations estimated by MPA including one.5 Story Drift Ratio (%) 2 2.18.5 1 1. two. 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 2 GM Multiplier 2. two. Locations of plastic hinges determined by MPA considering one.5 × El Centro ground motion.19. 4.(a) MPA. and three “modes” and by NL-RHA for 1.5 1 1.5 1 1.5 3 Fig. gravity loads included 53 . 1−"Mode" • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • •• •• •• • • • • • •• •• •• •• •• •• •• (b) MPA. 3−"Modes" (d) Nonlinear RHA • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• •• •• • • • • • • • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • • • • • •• •• •• Fig. 2−"Modes" •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• •• •• (c) MPA.20. Noted Error (%) 30 20 6 8 1 5 2 4 3 9 7 Error (%) 30 4 5 20 6 9 7 1 10 0 0 10 0 0 8 2 3 0. 4. gravity loads included (a) Floor Displacements 60 50 40 Error Envelope Error for Floor No. Errors in MPA as a function of ground motion intensity: (a) floor displacements.5 2 GM Multiplier 2. Noted 60 50 40 (a) Story Drifts Error Envelope Error for Story No. and (b) story drifts.5 3 0.

13E-03 5.5 × El Centro ground motion.5 21.2 4.938 1.353 -23.049 -0.673 Table 4.7 16.88E-03 0.8 Peak values of hinge plastic rotations (radians) from MPA for 1.00E+00 0.018 -0.55E-03 3.454 1.11E-03 9.35E-10 3.154 0.1 4.515 -50.998 21.908 -1.101 -0.8 1.00E+00 0.0 37.00E+00 3.754 1.4 20.35E-03 8.266 -0.263 0.4 26.75E-03 0.6 2.4 1.490 -11.429 -1.214 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” 0.434 0.744 1.00E+00 0.00E+00 0.3 -22.2 4.996 -0.860 1.850 1.00E+00 0.213 1.125 -1.00E+00 “Mode” 2 0.26E-03 3.35E-03 8.5 1.921 1.00E+00 0.319 1.105 0.055 0.754 0.461 0.830 -12.6 0.88E-03 0.071 0.3 -100.2 16.098 20.00E+00 0.527 -0.Table 4.6 19.176 0.822 1.0 -100.908 1.2 2. 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” 1.1 21.2 21.207 18.9 -4.6 2.436 1.371 -0.5 21.1 1.2 -3.5 3 “Modes” -32.066 -0. gravity loads included Hinge Plastic Rotation (rad) Floor “Modal” Response “Mode” 1 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 8.813 1.858 2.603 1.237 0.19E-03 1.0 -30.9 -3.00E+00 0.7 4.3 -22.5 “Mode” 3 -1.35E-03 8.305 -0.033 -0.6 1.466 0.00E+00 2 “Mode” 8.594 -1.319 1.530 1.9 31.00E-10 NL RHA 1.237 0.23E-03 0.5 54 .514 -1.665 0.071 -0.783 1.00E+00 0.19E-03 1.156 -0.19E-04 5.998 0.927 1.821 1.00E+00 0.197 -0.4 -6.399 0.11E-03 9.3 1.603 -1.19E-03 1.2 1.066 -0.877 -46.3 -100.23E-03 3.00E-03 5.23E-02 1.199 16.00E-03 5.728 1.5 × El Centro ground motion.00E+00 0.998 1.2 -2.8 28.168 -0.637 0.434 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.821 -1.996 0.11E-03 9.270 -12.0 37.00E+00 0.11E-03 9.434 0.00E+00 3 “Modes” 8.4 1.107 1.23E-03 0.128 -1.7 -2.637 0.037 0.14 m) from MPA for 1.311 0.530 1.0 37.372 0.7 Peak values of story drift ratios (as % of story height) from MPA for 1.190 -0.00E+00 0.064 -10.913 7.257 0.2 1.478 0.3 9.330 1.667 0.3 -22.0 2 “Modes” -32.015 0.8 9.35E-03 8.516 0.5 10.00E+00 Error (%) “Modal” Response 1 “Mode” -32.00E+00 0.5 -5.55E-03 3.5 0.2 0.530 1.114 -1.1 13.00E+00 0.2 4.5 1.009 0.310 1.2 9.831 0.17E-03 9.315 0.8 -32.2 12.037 -0.687 0.399 -0.00E+00 0.854 0.686 -7.4 -4.00E-03 5.78E-03 1.00E+00 Combined (MPA) 1 “Mode” 8.7 19.04E-10 3.850 -1.04E-02 8.836 -0.114 1.19E-03 1.5 × El Centro ground motion.88E-03 1.213 1.068 0.9 -6.00E-03 5.2 9.507 1.854 0.6 Peak values of floor displacements (as % of building height = 37.414 28.831 0.7 -2.8 0.0 -100.4 -4.983 1.953 15.3 13.933 1.8 -32.263 1.102 1.130 0.00E+00 0.00E+00 0.330 1.7 1.109 0.351 -0.0 16.102 1.8 Table 4.00E+00 “Mode” 3 0.213 1.0 -30.429 1.00E+00 0.2 9.3 -100.55E-03 3.057 0.983 1.888 1.23E-03 3.8 0.2 -0.

SRSS distribution: s* is defined by the lateral forces back-calculated from the story shears determined by response spectrum analysis of the structure. 5. k = 2 for T1 ≥ 2.2 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. plastic hinge rotations.5 sec . 2… N ). MPA considering three “modes. etc. is determined by nonlinear static analysis of the structure subjected to lateral forces with invariant distribution over height but gradually increasing values until a target value of roof displacement is reached.5 sec . Specified in FEMA-273 are three distributions for lateral forces: 1. and the exponent k = 1 for fundament period T1 £ 0. and 3.5. story drifts. “Uniform” distribution: s* = m j (where the floor number j = 1. joint rotations.. j 2. Comparison of Modal and FEMA Pushover Analyses 5.1 FEMA-273 PUSHOVER ANALYSIS In this investigation we focus on one step in the nonlinear static procedure in the FEMA-273 document [Building Seismic Safety Council. a plot of base shear versus roof displacement. assumed to be linearly elastic.” and nonlinear RHA. 1997] The pushover curve. computed at the target displacement represent the earthquakeinduced demands on the structure. The floor displacements. gravity load effects were included in all 55 . 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. and varies linearly in between.

0981 0. The MPA procedure is more accurate than all the FEMA force distributions. and the locations of all plastic hinges in Fig.11 0.11 0.119 0.042 0.0702 0. Also included in these presentations are the MPA results considering three “modes” and the “exact” demands from nonlinear RHA.0896 0.0913 0.3. The “uniform” distribution overestimates all floor displacements by 17-28%. and Table 5. and (c) SRSS 56 . wherein the first two are obvious and the third is determined from response spectrum analysis of the building (Appendix C).367 (a) Uniform (b) ELF (c) SRSS Fig. and the displacements of upper floors are overestimated by 9-22%.281 0. 5. 5.analyses.177 0.1 demonstrate that the displacement demands are underestimated by the ELF and SRSS force distributions by 12 to 30 % at the lower six floors of the building.11 0.11 0.126 0. Figures 5.0 cm.1 through 5.0466 0. The pushover curves are given in Fig.21 0. 5.11 0. plastic hinge rotation demands in Table 5. The errors in the FEMA and MPA estimates of seismic demands relative to the “exact” demands are presented in Fig.5 times the El Centro ground motion. both presented in Section 4.0446 0. 0.3. The three FEMA force distributions are presented in Fig. 5.0381 0.1.1.4a. (b) ELF.2.4 and Tables 5.4.3.0654 0.165 0. 5.00719 0. Using each of these force distributions.112 0. 5. the value determined from RHA of the first-mode inelastic SDF system for 1. Force distributions in FEMA-273: (a) “uniform”.11 0.3a and Table 5. pushover analyses are implemented for a target roof displacement of 52.062 0.4. the story drift demands in Fig.0197 0. with the errors being larger for the SRSS distribution.3a and 5.1. The first four floor displacements are within 5% of the “exact” value. 5.3b and Table 5. the floor displacement demands in Fig.2.11 0.

α = 0. V = 7456 kN.24 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized 10 20 30 40 Roof Displacement (cm) 50 60 Fig. Pushover curves using three force distributions in FEMA-273: (a) “uniform”. gravity loads are included 57 . α = 0.14 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 = 38.2.3 cm. and (c) SRSS. α = 0.18 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized 10 20 30 40 Roof Displacement (cm) (c) SRSS Distribution 50 60 12000 10000 u = 39. V = 6897 kN.8 cm. V = 8530 kN. (b) ELF. 5.(a) Uniform Distribution 12000 10000 u = 33 cm.

For the uniform distribution. 3 "Modes" FEMA Uniform ELF SRSS 0. Modal pushover analysis procedure gives estimates better than all the FEMA force distributions. and overestimated by no more than 32%. reaching 35%. 7%. (The 100% error in a 6th floor hinge is ignored as it simply represents that the MPA estimated zero rotation.5 Displacement/Height (%) 2 Floor 5th 5th 4th 3rd 2nd 1st NL−RHA MPA. larger errors are noted in the upper and lower stories.5 Ground 0 Ground 0 Fig.Figures 5. with errors reaching 37% in this example. 3 "Modes" FEMA Uniform ELF SRSS 0.3b and 5. gravity loads included 58 . For the ELF distribution. MPA including three “modes.4b. Heightwise variation of floor displacements and story drift ratios estimated using FEMA-273 force distributions. at most.3 demonstrate that the hinge plastic rotations estimated by all three FEMA force distributions contain unacceptably large errors. Figure 5. such as hinge plastic rotations. 5.5 1 1. with story drifts under estimated by. In contrast.4c and Table 5.5 1 1. reaching 64%. the errors are largest in the lower stories. This seems to be an inherent limitation of pushover analysis. errors are largest in the upper stories.2 demonstrate that the story drift demands are greatly underestimated by all the FEMA force distributions. and Table 5. (a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th (b) Story Drift Ratios Floor 4th 3rd 2nd 1st NL−RHA MPA.3.” and NLRHA. the MPA procedure is more accurate than all the FEMA force distributions. whereas nonlinear RHA computed an insignificantly small value. reaching 31%. For the SRSS distribution.) The pushover analysis procedures considered seem incapable of computing accurately local response quantities. but it is still inaccurate.5 Story Drift Ratio (%) 2 2.

5. they are not good indicators of damage. Errors in floor displacements.(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. the SRSS distribution fails to identify yielding of beams in the middle floors. and the three FEMA analyses. 5.4. story drifts. and the ELF distribution fails 59 . The locations of the plastic hinges are not identified correctly by the FEMA force distributions. The locations of plastic hinges shown in Fig.” nonlinear RHA (“exact”). the “uniform” distribution fails to identify yielding of the beams above the fourth floor. and hinge plastic rotations estimated using FEMA-273 force distributions and MPA (including three “modes”). Based on the results presented here. 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. it appears that structural performance evaluation should be based on story drifts that are known to be closely related to damage and can be estimated to a higher degree of accuracy by pushover analyses.5 were determined by five analyses: MPA considering the three “modes. While pushover estimates for floor displacements are more accurate.

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

061 1.789 0.8 -35.466 0.7 -60.0 37.199 27.306 1.673 Error (%) NL RHA 1.992 1.355 0.263 0.19E-03 1.00E+00 0.11E-03 9.2 10.2 -100.45E-03 3.00E+00 2.6 4.927 1.3 23.7 7.00E+00 0.1 0.0 -71.4 -23.335 2.686 0.351 0.372 1.17E-03 9.26E-03 3.4 -50.998 1.6 -11.5 -32.178 1.730 1.067 0.399 -27.4 0.5 -3.4 21.00E+00 0.75E-03 0.0 -63.672 1.3 -22.6 22.9 0.168 1.00E-10 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 1.839 0.560 1.462 1.23E-02 1.9 -70.7 -29.154 1.2 9.2 -4.3 29.8 -32.65E-03 7.4 -12.3 -14.736 0.6 14.597 0.6 -6.913 0.2 -12.78E-03 1.6 -4.50E-03 0.4 16.566 1.9 16.984 1.314 1.23E-03 3.00E+00 0.53E-02 1.7 MPA -2. gravity loads included Displacement /Height (%) FEMA ELF SRSS MPA 1.262 1.94E-03 2.35E-10 3.5 -33.8 2.830 2.2 6.9 15.9 16.8 44.6 21.4 MPA -2.78E-03 0.611 0.6 -41.953 0.854 0.214 1.13E-03 5.234 1.1.2 4.5 15.5 -27. gravity loads included Displacement /Height (%) FEMA ELF SRSS MPA 0.083 1.59E-03 5.209 0.323 1.58E-04 6.10E-02 7.667 0.9 28.1 -3.03E-03 5.8 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 0.93E-03 1.7 Table 5.3 Peak values of hinge plastic rotations (radians) from FEMA force distributions and MPA Hinge Plastic Rotation (rad) MPA FEMA ELF SRSS 4.007 1.8 -100.7 19.5 -29.341 1.34E-03 2.5 22.109 1.1 10.55E-03 3.414 1.270 0.1 -26.2 16.5 Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 2.128 1.9 SRSS -22.783 1.84 1.04E-02 8.62E-03 0.52E-04 1.708 0.515 FEMA Uniform ELF 27.7 -28.3 -4.098 1.4 16.1 163.35E-03 8.998 1.19E-04 5.2 16.00E+00 8.858 1.09E-03 4.0 -30.530 1.0 Table 5.221 1.8 -63.88E-03 1.00E-03 5.015 0.975 1.0 -57.5 61 .2 Peak values of story drift ratios (as % of story height) from FEMA force distributions and MPA.3 -11.5 26.4 -55.16E-03 0.875 0.9 -20.0 -100.6 17.860 1.547 -27.436 Error (%) NL Uniform RHA 0.623 1.00E+00 0.2 5.2 19.530 1.5 10.7 26.2 9.399 1.2 13.089 1.330 1.8 7.011 1.6 -17.8 16.00E+00 0.064 1.207 1.367 1.00E+00 Error (%) FEMA Uniform ELF SRSS MPA 24.5 -33.7 21.14 m) from FEMA force distributions and MPA.00E+00 4.399 1.8 -100.487 0.195 0.6 -73.0 -100.724 0.3 31.490 0.294 1.524 0.310 1.836 0.809 0.7 -15.888 0.0 -59.6 -25.877 1.Table 5.353 1.938 1.51E-03 4.7 FEMA ELF SRSS -22.344 0.318 2.3 -100.0 -100.8 17.562 1.5 -29.00E+00 NL RHA 1.9 11. Peak values of floor displacements (as % of building height = 37.0 -100.8 -2.417 1.9 -77.

5 × El Centro ground motion. gravity loads included 62 . 5. MPA including three “modes” and NL-RHA for 1.(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 1 1.5 2 GM Multiplier 2.5 2 GM Multiplier 2.(a) FEMA: Uniform 60 50 40 60 50 40 3 (b) FEMA: ELF Error Envelope Error for Floor No.5 2 GM Multiplier 2.5 1 1.6.5 3 Fig.5 1 1.5 3 0 0 0. 5.5 3 (c) FEMA: SRSS 60 50 40 60 50 40 2 3 1 5 4 (d) MPA (3 "Modes") Error (%) 30 20 10 0 0 8 Error (%) 30 20 6 8 1 5 2 4 3 9 7 6 7 9 10 0 0 0.5 3 0.5 2 GM Multiplier 2. Errors in floor displacements from three force distributions in FEMA-273 and from MPA including three “modes”. gravity loads included 63 .

5 1 1.5 1 1.5 1 1. gravity loads included 64 .5 2 GM Multiplier 2.5 3 Fig.7.5 1 1.5 3 0 0 0.5 2 GM Multiplier 2. 5.5 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 3 0 0 0.5 2 GM Multiplier 2. 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 2 GM Multiplier 2. Error in story drifts from three force distributions in FEMA-273 and from MPA including three “modes”.(a) FEMA: Uniform 100 60 50 9 8 (b) FEMA: ELF Error Envelope Error for Story No.

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

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

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

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

Engng. A.C. Struct. Krawinkler.. M. G. Y. Blume Earthquake Engineering Center. 123:256-265. Earthq. J. Earthq. N2—a method for nonlinear seismic analysis of regular structures.K. Gupta.. Washington.. Struct.M. Pros and cons of a pushover analysis of seismic performance evaluation. Krawinkler.. (2000). Engrg. UCB/EERC-88/06. 20(4-6):452-464. 9th World Conf.4. D. Story drift demands for steel moment frame structures in different seismic regions. H. A. S. Seismic performance and retrofit evaluation for reinforced concrete structures. University of California. (1999). and Krawinkler. Engrg. Adaptive spectra-based pushover procedure for seismic evaluation of structures. P.K. Report No. Stanford. H. Kim. and Reinhorn. Conf.D. H. Calif.K. S. 16(2):367-392 Han. G. 132. 69 . A. Nat. (1997)..H. Fajfar. ASCE 123(1):3-10. I: Equivalent nonlinear system.. and Wen. Berkeley. S. Seismic demands for performance evaluation of steel moment resisting frame structures (SAC Task 5. Spectra. Japan. Federal Emergency Management Agency. Proc. A. (1998). (1988). Tokyo-Kyoto.. Struc. NEHRP Guidelines for the Seismic Rehabilitation of Buildings. Earthq. B. Estimation of seismic drift demands for frame structures. and Gupta. Calif. (1997). Kunnath. and Krawinkler. Method of reliability-based seismic design. and Seneviratna. Pushover analysis procedure in earthquake engineering. H. Dynamics of Structures: Theory and Applications to Earthquake Engineering. Earthquake Engineering Research Center. FEMA-273. A. (2001). Report No. 15:417-434. S. Earthq.M. Chopra.S. Engrg. and Fischinger. 6th U. Dyn. on Earthq. Building Seismic Safety Council (1997). Prentice Hall: New Jersey.3). 5:111-116.7 References Allahabadi. (1998). (1999).W.K.. Bracci. Seattle. Engrg.. Engrg. Struct. Proc. ASCE.P. R. J.. J. Gupta. and Kunnath. J. John A. Spectra. (1988). 29:1287-1305. and Powell. E. Stanford University. Gupta. DRAIN-2DX user guide. (2000). Washington.K. and D’Amore.

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

g. ω n .2. Define the anchor point.. 3. using any numerical integration method.Appendix A Uncoupled Modal Response History Analysis A. Compute natural frequencies. 3. B. Idealize the pushover curve as a bilinear curve (Fig. 1988).1. This step can be conveniently implemented in any commercially available software. Calculate the area under the actual pushover curve. iterations may be necessary.urn ) pushover curve for the force distribution s* : n 2. 71 . e.1) using the FEMA-273 procedure (Building Seismic Safety Council. DRAIN2DX (Allahabadi and Powell.20): s* = mφn n n 2. The structure should be pushed just beyond the target (or expected) roof displacement in the selected mode.. For the nth-“mode”. Let the roof displacement and base shear at the anchor point be urno and Vbno . φn . e. 1977). 2. Apply force distribution of Step 2. Define the force distribution s* from Eq.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. trapezoidal rule. A.1 incrementally and record the base shears and associated roof displacements. of the bilinear curve at the target roof displacement. (3. Apn . and modes. develop the base-shear – roof-displacement (Vbn .1.2. Since the target roof displacement may not be known at the start of the procedure. 1. for linear-elastic vibration of the building. respectively.g. 3.

corresponding to the estimated yield i i i base shear. From the pushover data. Calculate the slope. Draw the curve OAB by connecting the three points O.6. Calculate Vbny = Vbny × Apn Abn . and B with straight-line segments to obtain the idealized bilinear curve.5. Calculate the post-yielding strain-hardening ratio. Calculate the error = 100 × Abn − Apn ( ) ( ) ( ) ) Apn . Let the point with base shear = Vbny and roof displacement = urny be denoted as A.8. Replace i+1 with i and repeat Steps 3.6 × Vbny . i 3.9.2. If desired. Calculate initial slope of the idealized bilinear curve. Vbny . 3.9. A. 3. 3.6 . and a point on the actual pushover curve with base shear equal to i 0. obtained by judgment.i 3. Vbny . determine the roof displacement. Calculate area under the bilinear curve OAB.6 .4. 72 . i 3. urny = Vbny kn .4 to 3.0. will be refined by an iterative procedure that seeks to equate areas under the actual and the idealized pushover curves.7.3.1. i i 3. kn = 0. i 3.0. kn . at base shear i equal to 0. Estimate the yield base shear.2. i +1 i i 3.9. Calculate the yield displacement. Abn .1.4. urn. i i i 3. iterations are necessary. other appropriate methods can be ( used.4.8.6 × Vbny ( ) i urn . This value. O. i i i αn =  Vbno Vbny − 1  urno urny − 1         i 3.6 × Vbny . by connecting a straight line between origin. If the error exceeds some pre-specified tolerance. This step gives the secant stiffness at a base shear equal to 60% of the yield base shear.

8. * 4. r o .3b) with unit-mass and force-deformation relation of Fig.15) and (3. of the combined responses obtained in Step 8.3. (3.11b). 4. Calculate histories of various responses using Eqs.2.2.0.1 Idealization of nth-“mode” pushover curve 73 .4.10b and 4. (3. 5. Calculate peak values.13). Develop the Fsn Ln − Dn relation (Fig.1. 4. Compute the Ln and Γ n from Eq. A. Repeat Steps 3 to 6 for as many modes as required for sufficient accuracy. 7. and pseudo-acceleration history. A. Dn (t ) .12) and (3. In general first two or three modes will suffice. 4. Scale the horizontal axis by Γnφrn to obtain Dno = urno Γ nφrn and Dny = urny Γ n φrn (Eqs. (3. 9. An (t ) . Combine the “modal” responses using Eqs. Vbn Idealization of Pushover Curve Vbno Vbny 0. 4. of the nth“mode” inelastic SDF system (Fig.eps ur n. A.2). Scale the vertical axis with * Mn to obtain * Fsno Ln = Vbno M n and * Fsny Ln = Vbny M n (Eqs.4) and effective modal mass from M n = Ln Γn . 4.6 ur n y ur n o ur n Fig. 6.6 × Vbny A • Idealized αnkn B • 1 Actual k 1 • O n figA_1. Compute deformation history.11a).16).10a and 4.

5. n 2. Area under the actual pushover curve. The pushover curves for the first three modes. 74 . Ap1 = 360777 kN-cm. Target displacements used to generate these pushover curves are 63.eps Dny = ur n y / Γnφr n Dno = ur n o / Γnφr n Dn Fig. and third mode.2 Properties of the nth-“mode” inelastic SDF system A.2.5 cm (25 in. At this point. A. First three mode shapes and frequencies of the selected building were computed and are shown in Fig. 1.3.urn ) pushover curve for the force distribution s* : n 2.3.1. The base-shear – roof-displacement (Vbn .4 cm (10 in. second. computed for the first three modes are shown in Fig. The anchor point. 3. for the first. The force distributions.7 cm (5 in. The following steps illustrate the procedure to develop the idealize curve for the first “mode”.2 EXAMPLE The UMRHA procedure is implemented to calculate the response of the 9-story building described in Section 3.). is defined at the target roof displacement. 2.4.Fsn / Ln Fsn / Ln − Dn Relationship Vbno / M* n Vbny / Mn * 1 α ω2 n n ω2 1 n figA_2. A. A. Idealized bilinear curves for each of the three modes are included in Fig. 3. 25.).5 cm and Vb1o = 8729. 3. ur1o = 63.3. are shown in Fig. to the north-south component of the El Centro (1940) ground motion scaled up by a factor of 1.4. generated using DRAIN-2DX. and 12.2. 3. Following is step-by-step implementation of the procedure described in Section A. s* .1. B. respectively. 3.).6 kN.1.

3.6 kN. The procedure converged after nineteen cycles to give ur1 y = 36. L1 = 2736789 kg.0.6 = 22. The initial slope of the idealized bilinear curve.7. * 4. and B with straight-line segments gives the idealized bilinear curve. This value exceeds the prespecified tolerance of 0.4 kN. Determined from the pushover database.9 kN.0. Also included are the modal damping ratios and the periods calculated from Eq.01%.4 210.86 cm at 0.5 38.9.1. Therefore. The Fs1 L1 − D1 relation for the first “mode” is developed as follows.1.2.2. i i i 3.1. The post-yielding strain-hardening ratio. k1 = 0. The next estimate of the yield shear is Vbny = 8006.8. iterations are necessary.18 kN/cm. i i 3. The point A on the i i bilinear curve is defined by ur1 y = 38. i i i 3.6 × Vb1 y = 4803.09 cm and Vb1 y = 8006. i i 3. 3. The yield displacement.4. k1 .23 cm. i +1 3.5. (4. Ab1 = 365100 kN-cm.9. 4. i 3. and α1 = 0. The first estimate of the yield base shear Vb1 y = 8006.4 ) − 1 (63. and M1 = 2736789 × 1. 75 . Vb1 y = 7615. A.4.198%. The curve OAB obtained by connecting the three points O.4. ur1 y = Vb1 y k1 = 8006. α1 =  Vb1o Vb1 y − 1  ur1o ur1 y − 1 =         ( ) ( ) (8729.1. Area under the bilinear curve OAB. 3.i 3. is calculated as follows.135.9.86 = 210.4 × (360777 365100 ) = 7911. ur1.6 8006.09 ) − 1 = 0.194. Error = 100 × (365100 − 360777 ) 360777 = 1. The results of the iterative procedure are summarized in Table A.3666.4 kN.18 = 38. 3.6 = 4803. The results for other modes are summarized in Table A.8 kN.     i 3.6 × Vb1 y ( ) i ur1.8 22.09 cm. Γ1 = 1.2.13).3666 = 3740189 kg.6.

7.62 (cm/sec2). 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. Scaling the vertical axis by Fs1 y L1 = 203.4. Scaling the horizontal axis by Γ1φr1 gives D1o = 46. 4.3. A. 7.8. Histories of roof displacement and top story drifts for the first “mode” are computed and presented in Fig. 4. 9. The combined modal responses are presented in Fig.7. The results were generated for first three “modes’ and are included in Fig. The peak values are also plotted in Fig.1 and 4. 6. * M1 gives Fs1o L1 = 233.46 cm and D1y = 26.40 (cm/sec2) and 76 .4.2. 5. 4. The peak values are computed and are summarized in Tables 4. 4.2. 8.51 cm. 4.7.

6 cm. V = 5210 kN. α = 0.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36. α = 0. V = 7616 kN.eps 10 20 30 40 50 Roof Displacement (cm) (b) "Mode" 2 Pushover Curve 60 8000 u = 9. A.eps 0 0 5 10 15 20 Roof Displacement (cm) (c) "Mode" 3 Pushover Curve 25 8000 u = 4.13 6000 Base Shear (kN) 4000 2000 Actual Idealized figA_3b.3.9 cm.2 cm.19 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized figA_3a. α = 0.14 y by 6000 Base Shear (kN) 4000 2000 Actual Idealized figA_3c. V y by = 4952 kN. “Modal” pushover curves for the example building 77 .eps 0 0 2 4 6 8 10 Roof Displacement (cm) 12 Fig.

5 × El Centro ground motion for the first “mode.52 5 10 15 20 Time (sec) 25 30 0 • 1.2769 D1 (cm) 0 • 35.4.223 −80 80 n=3 −3 3 n=3 D3 (cm) 0 A3 (g) • 10.” and third “mode” inelastic SDF systems. Histories of deformation and pseudo-acceleration due to 1.06 0 • 1. 78 .747 5 10 15 20 Time (sec) figa_4.” second “mode.(a) Deformation 80 n=1 3 (b) Pseuso−acceleration n=1 • 0. A.m −80 0 −3 0 25 30 Fig.33 A1 (g) n=2 0 −80 80 −3 3 n=2 D2 (cm) 0 A1 (g) • 22.

38 22.192 0.12 3876.Table A.56 19.184 0.022 0.51 2 “Mode” 2 -920860 -0. i 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 i Vb1 y i 0.18 210.70 36.18 (cm) 38.3 7658.0 4704.3 7745.048 0.136 Fsny Ln (cm/sec2) Dny (cm) Fsno Ln (cm/sec ) Dno (cm) Tn (sec) 233.1 4574.8 7618.1 7616.76 21.2 7690.09 18.95 21.39 36.193 0.162 0.180 0.3 4603. Properties of “modal” inelastic SDF systems Properties Ln (kg) Γn * M n (kg) “Mode” 1 2736789 1.18 210.28 36.74 21.18 210.75 21.8525 1.151 0.4927 1.194 0.09 37.037 0.25 36.0 4577.693 0.18 210.910 0.193 0.9 7615.85 36.4 7911.18 210.6 4583.40 46.186 0.193 0.64 37.0.5 (cm) 22.59 36.18 210.2 4671.35 36.0 7619.2406 167531.029 0.79 0.6 × Vb1 y i ur1.063 0.237 0.4 7647.46 2.23 0.082 0.18 210.18 210.1. Results of iterative procedure to develop the idealized bilinear curve for the first “mode” inelastic SDF system Itr.50 36.6 7840.1 1013.3 7628.23 22.135 0.78 21. No.3666 3740189 203.65 1226.2671 1.18 210.30 37.26 36.194 (kN-cm) 365100 364060 363276 362684 362235 361892 361631 361432 361279 361162 361073 361004 360951 360911 360880 360856 360838 360824 360813 Table A.05 36.182 0.24 36.017 0.139 0.7 4580.18 210.013 0.18 210.62 26.0 4588.190 0.8 4647.2.191 0.2 4628.74 (kN/cm) 210.90 21.18 210.194 0.86 21.9 4573.309 0.86 22.8 7622.404 0.79 21.170 0.18 210.29 36.81 21.18 210.02 21.25 36.8 4747.83 21.9 (kN) 4803.188 0.5 7633.18 210.85 0.529 0.103 “Mode” 3 696400 0.010 (kN) 8006.1 4569.18 210.75 21.5 7624.5 3109.32 36.4 7714.5309 488839.4 7672.948 ζ n (%) 79 .77 21.3 7786.6 i k1 i ur1 y i α1 i Ab1 Error (%) 1.176 0.56 47.194 0.44 36.18 210.107 0.05 52.59 22.11 22.198 0.18 210.2 4571.9 4570.9 4570.4 4595.7 7639.5 4614.

80 .

(3. of the nth-“mode” inelastic SDF system (Fig.21). Repeat Steps 3 to 7 for as many “modes” as required for sufficient accuracy.1. 12. Dn . 6. Steps 1 to 4 of the MPA are the same as those for UMRHA presented in Appendix A. or from the inelastic response (or design) spectrum. (3. (4. 11. 81 . This is the same example as solved in Appendix A. 10. (4.18). At urno .4. Solving Eq. From the total hinge rotation. Calculate the peak roof displacement urno associated with the nth-“mode” inelastic SDF system from Eq. 10.8).5. extract from the pushover database values of other desired responses. Compute the peak deformation. Typically. B. Determine the total response by combining the peak “modal” responses using the SRSS combination rule of Eq.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.3b) with unit mass and force-deformation relation of Fig.Appendix B Modal Pushover Analysis B. subtract the yield hinge rotation to determine the plastic hinge rotation.1 to the north-south component of the El Centro (1940) ground motion scaled up by a factor of 1. the first two or three “modes” will suffice.33 cm. 4.8) 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. Following is step-by-step implementation of the procedure described in Section B. 13.2 EXAMPLE The MPA procedure is implemented to calculate the response of the 9-story building described in Section 3. 4.6b by solving Eq. rno .

4. (3. 82 . Results of Steps 5 and 6 are summarized in Table B. Peak roof displacement ur1o = Γ1 × φr1 × D1 = 1. 9. where results for other ground motion intensities are also included.4.3 and 4. 8. values of floor displacements and story drifts are extracted.11. Steps 3 to 7 are repeated for first three “modes.366 × 1 × 35.3 and 4.3 for the floor displacements normalized by the building height and in Table 4. The values are summarized in Table 4.28 cm.5 are the plastic hinge rotations computed by subtracting the yield hinge rotation from the total hinge rotations. 12.1.33 = 48. At ur1o = 48.28 cm.” and the results are included in Tables 4. Also included in Table 4.4 for the story drifts normalized by the story height.18) are also included in Tables 4. The total response computed by combining the peak “modal” responses using the SRSS combination rule of Eq.

969 0.901 8.755 0.577 16.913 22.18 27.37 1.184 0.85 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 1.267 5.36 1.225 2. Calculation of roof displacements urno from peak deformation of inelastic SDF systems Ground Motion Multiplier 0.52 “Mode” 3 1.73 24.71 1.126 13.1.395 0.79 0.332 48.504 18.367 1.457 12.690 10.35 0.25 Quantity Dn (cm) “Mode” 1 6.312 1.75 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.551 2.82 1.70 0.52 0.467 14.59 0.436 7.691 0.332 13.185 11.50 35.27 0.28 46.660 14.154 78.766 7.678 0.03 26.03 0.275 1.4222 3.535 14.0 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 1.06 1.05 0.748 63.33 1.Table B.0 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 3.5 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.856 31.38 22.5 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 2.023 0.35 1.0 µ = Dn Dny urno = Γnφrn Dn (cm) 83 .229 8.735 3.8451 5.139 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.268 0.513 0.676 6.13 2.450 4.200 0.37 57.379 21.252 9.07 “Mode” 2 4.007 36.755 27.117 5.25 20.

84

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

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

Floor, j

mj (10-3×kg)

s* = j

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

mj

1 2 3 4 5 6 7 8 9

503.5 494.7 494.7 494.7 494.7 494.7 494.7 494.7 534.1

C.2

EQUIVALENT LATERAL FORCE (ELF) DISTRIBUTION 85

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

Floor j 1 2 3 4 5 6 7 8 9

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

s* j

=

m jhk j

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

C.3

SRSS DISTRIBUTION

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

∑ n (V jn )

2

to get column 8 of

7 1578.045 0.6 -732.3 240.2 97.For convenience.4 1759.1 -525.5 159.7 101.090 0.9 -166.5 320.6 319.8 1381.7 355. 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) 1917.1 832.4 400.6 366.7 1622.0 1231.3 222.0 176.9 880.5 -973.3.177 0.9 -153.8 374.7 374.1 87.5 -320.2 285.6 -352.2 fj = ∑ fi i fj (2) (3) (4) 59.3 -6. Table C.9 446.7 95.367 87 .8 -326.2 105.9 1683.047 0.1 -967.7 (10) 0.6 -359.0 136.2 148.3 -646.1 1857.6 V j3 (kN) (7) 478.5 215.7 694.070 0.9 366.0 381.6 1233. the lateral forces are normalized by the base shear to obtain column 10 in Table C.0 -5.065 0.3.2 200.1 -438.042 0.0 1476.0 980.2 95.4 1842.6 286.1 Story Shears V j2 (kN) (6) 1114.7 -46.0 354.2 277.7 234.7 Vj (kN) (8) 2268.1 -646.9 832.098 0.5 -350.8 430.7 2065.2 Lateral Force fj (kN) (9) 203.4 250.9 1446.7 525.

Sign up to vote on this title
UsefulNot useful

Master Your Semester with Scribd & The New York Times

Special offer for students: Only $4.99/month.

Master Your Semester with a Special Offer from Scribd & The New York Times

Cancel anytime.