This action might not be possible to undo. Are you sure you want to continue?

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 .

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

iv .

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

vi .

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

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

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

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

the yield deformation. It is the familiar bilinear hysteretic relationship. 2. 2. sign u = . reloading from a minimum deformation takes place along a path parallel to the initial elastic branch. Yielding begins when the force reaches f y and the deformation reaches u y . Unloading from a maximum deformation takes place along a path parallel to the initial elastic branch.1b. this system is linearly elastic with stiffness k as long as the force does not exceed f y .2) . The yield strength is the same in the two directions of deformation. It consists of a mass m (or weight w) concentrated at the roof level. through the yield strength reduction factor.1a. defined by f Ry = o fy (2. 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 a linear viscous damper with damping coefficient c.mug t bg The governing equation for this inelastic system subjected to horizontal ground acceleration b g bg 3 (2. This lateral-force displacement relation is idealized as shown in Fig. Similarly. the yield strength. R y . a massless frame that provides stiffness to the system.1 One-Story Systems EQUATION OF MOTION Consider the idealized one-story structure shown in Fig. where 0 < a << 1.1) b g ug t is mu + cu + f s u.2 2. The yield strength is related to f o . During yielding the stiffness of the frame is a k . On initial loading. sign u . the strength required for the structure to remain elastic during the ground motion.

the deformation u t depends on three systems parameters: w n . in addition to the form of the force-deformation relation. sign u = .. This becomes evident if Eq. and the ductility factor is u m= m uy (2. and (b) bilinear hysteretic forcedeformation relation For a given excitation u g t .3). 2.e. The peak. (2.3) ωn = k m ζ = c 2mω n f fs = s fy (2.1. u £ u y ). z . z y .fS m fs c ut u fy 1 αk k 1 uy ug -fy k k um u 1 1 (a) (b) Fig. or absolute (without regard to algebraic sign) maximum. (a) Idealized one-story structure. Tn = 2p w n is the natural vibration period.ug t n where bg bg b g bg (2.5) For a given u g t . and R y (Chopra . deformation is denoted by um . bg 4 . and z is the damping ratio of the system vibrating within its linear elastic range (i. m depends on three system parameters: w n . 2001.4) and w n is the natural vibration frequency.1) is divided by m to obtain ~ u + 2zw n u + w 2 u y f s u. Section 7. and u y .

The yield strength of the inelastic system.4 in Chopra. m h = 3. one-bay frame 50 uy = 1. Vb (kN) 30 20 10 0 0 2 4 6 Roof Displacement.9 kN (38.376 cm. with Tn = 0. Vby = 39.2. 2. f o w = 184 .826 kips) for w = 169.04 40 Base Shear.2 kips) .26 kN.2 SYSTEM AND EXCITATION CONSIDERED Consider the one-story system in Fig.32 m • Fig.5 sec and z = 5% subjected to the north-south component of the El Centro (1940) ground motion (Fig. For . 2001) and scaled up by a factor of 2. α = 0.2: the dimensions and flexural rigidity of the structural elements are noted.26 kN (8. One-story.1. is f y / w = ( f o / w ) ÷ 8 = 0. this system and excitation. 2.66 m • EIb EIc EIc • • L = 7. and f y = 39. 2.2311 . 2.2 5 . based on Ry = 8 .2. 6. u (cm) 8 Fig.3. Pushover curve for structure shown in Fig.

1 kip-in.The yield moments in the beam and columns are defined as the bending moments due to the lateral force f y .017 rad . q m = 0. Implementing this analysis with I c = 6. the energy dissipated in 6 .3 RESPONSE HISTORY ANALYSIS Figure 2.36 cm. 2. I b = 3.g.3 in. the ductility factor m = 5. the joint rotation and beam hinge rotation are identical to values q m and q pm determined from RHA. It is organized in five parts: (a) shows the deformation u t . as apparent in Fig. (b) shows the lateral force f s t or base shear Vb t normalized relative to the bg weight. The yielding stiffness of each structural element is defined as 3% of its initial stiffness. A static pushover analysis of this one-story system leads to the force-displacement relationship shown in Fig.4 shows the earthquake response of the system described in the preceding section determined by response history analysis (RHA).4e. The peak values of the various response quantities are as follows: um = 7.4f. Determined from pushover analysis at the exact peak deformation. and E = 2 × 108 kPa (29×103 ksi) gives yield moments of 21.18 kN-m (444.65 kN-m (191.) for the beam and columns. 2. Observe that the pushover curve matches the initial loading path of the hysteretic system. and (e) shows the force-deformation relation.35.6 kip-in.3.36 cm . e. wherein the hysteretic force-deformation history of Fig. 2. This pushover curve turns out to be bilinear because the beam and the columns are designed to yield simultaneously when f s reaches f y .0217 rad . pushover analysis cannot provide any cumulative measure of response. the peak value determined from RHA.) and 50.4edetermined from RHAis superimposed. However.077 × 107 mm 4 (146 in.134 × 107 mm 4 (75.4 ). The resulting pushover curve is shown in Fig. 2. (c) shows the joint rotation q bt g .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 = 7.4 ). and q pm = 0. respectively. 2.. The system is excited well beyond the yield deformation. (d) shows the rotation q p bt g of the plastic hinges at the bg bg beam ends. 2.

04 0. 15 u (cm) 0 • 7. This represents an inherent limitation of pushover analyses.36 −15 0. (c) joint rotation.017 • 0 (d) θp (rad) −0.0217 • θ (rad) 0 (c) −0.4 −10 −5 0 u (cm) 5 10 Fig.04 0.2 −0.04 0. or the cumulative rotation at a plastic hinge.4. and (f) pushover curve 7 .yielding during the ground motion. (b) base shear.2 0 (e) 0 (f) −0.5 0.5 fy / w = 0.4 −10 −5 0 u (cm) 5 10 −0. 2.4 0.4 0. (d) plastic hinge rotation.2311 (a) Vb / w 0 f / w = −0.2 −0.2311 y (b) −0. (e) force-deformation relation.2 Vb / w 0.04 0 5 10 15 Time (sec) 20 25 30 0. Response of one-story system to El Centro ground motion: (a) deformation.

8 .

m ι u g t bg bg (3.snug t n= n =1 The effective earthquake forces can then be expressed as af N af N af (3. (3.1 MODAL RESPONSE HISTORY ANALYSIS The differential equations governing the response of a multistory building to horizontal earthquake ground motion u g t are as follows: mu + cu + ku = . 2001: Section 13. and p eff t = Â peff. and lateral stiffness matrices of the system.2) The spatial distribution of these “forces” over the height of the building is defined by the vector s = m ι and their time variation by u g t .mι ug t bg bg bg (3.n t = Â .1) where u is the vector of N lateral floor displacements relative to the ground. This force distribution can be expanded as a summation of modal inertia force distributions s n (Chopra. Elastic Multistory Buildings 3.4) 9 . each element of the influence vector ι is equal to unity. and k are the mass.12): mι = n =1 Â sn = N n =1 Â Γ nmφ n N (3. The right side of Eq.1) can be interpreted as effective earthquake forces: p eff t = .3. c.3) where φ n is the nth natural vibration mode of the structure. classical damping. m.

(3.9) and (3. subjected to u g t : Dn + 2z nw n Dn + w 2 Dn = .10) bg bg bg 10 (3.n (t ) = N n =1 ∑ −sn u g (t ) N (3.7) By utilizing the orthogonality property of modes. an SDF system with vibration propertiesnatural 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. p eff .n t = .sn ug t bg (3.Gn ug t (3.n t is entirely in the nth-mode.u g t n Comparing Eqs. with no contribution from other modes.12) . we will outline that the response of the MDF system to p eff . it can be demonstrated that none of the modes other than the nth mode contribute to the response. Then the floor displacements are u n t = φ n qn t bg bg bg bg (3.11) and substituting in Eq. The equations governing the response of the system are mu + cu + ku = .8) gives the floor displacements u n t = Gnφ n Dn t bg (3.10) gives qn t = Gn Dn t bg bg bg (3. The solution qn t can readily be obtained by comparing Eq. (3.8) where the modal coordinate qn t is governed by 2 qn + 2z nw n qn + w n qn = .9) in which w n is the natural vibration frequency and z n is the damping ratio for the nth mode.6) Next.peff (t ) = n =1 ∑ peff . (3.5) The contribution of the nth mode to s and to p eff t are: sn = Gn mφ n respectively.sn ug t bg bg bg bg (3.

3.1).15) and (3.12) and (3. (3. the response of the system to the total excitation p eff t is ut = bg bg bg n =1 N Â un bt g = Â Gnf n Dn bt g n =1 N N (3.13) represent the response of the MDF system to p eff.13) define the contribution of the nth-mode to the response. and Eqs. (3. Section 13. 3. internal element forces. st The two analyses that lead to rn and An t are shown schematically in Fig.16) reflect combining the response contributions of all modes. Equations (3. In contrast to the classical 11 bg . these standard equations have been derived in an unconventional way.12) and (3. Therefore.Forces sn An(t ) st rn ωn. etc. However. 2001.15) rt = bg n =1 Â rn t = bg n =1 Â rnst An bt g N (3.13) st where rn denotes the modal static response.16) This is the classical modal RHA procedure wherein Eq.n t . ζn ¨ g(t ) u (a) Static Analysis of Structure (b) Dynamic Analysis of SDF System Fig. the static value of r due to external forces sn .1. (3.1. and An t = w 2 Dn t n bg (3. Conceptual explanation of modal response history analysis of elastic MDF systems Any response quantity r t story drifts.14) is the pseudo-acceleration response of the nth-mode SDF system (Chopra. Eqs.can be expressed by st rn t = rn An t bg bg bg bg bg (3.9) is the standard modal equation governing qn t .

19) will provide the same value of rno . we note that static analysis of the structure subjected to lateral forces f no = Γ n mφ n An (3. This concept provides a rational basis for the modal pushover analysis procedure developed later.3).17) (Chopra. The SRSS rule. 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 value rno of the nth-mode contribution rn t to response r t is determined from st rno = rn An bg bg bg (3. which is valid for structures with well-separated natural frequencies such as multistory buildings with symmetric plan. 2001.derivation found in textbooks (e.1. we used the modal expansion of the spatial distribution of the effective earthquake forces. 2001.3 MODAL PUSHOVER ANALYSIS (3.4 and 13. z n of the pseudo-acceleration response (or design) spectrum for the nth-mode SDF system.g. this response value can be obtained by static analysis of the structure subjected to lateral forces distributed over the building height according to 12 . Sections 12.17) where An is the ordinate A Tn . and Tn = 2p w n is the natural vibration period of the nth-mode of the MDF system.16). The peak modal responses are combined according to the Square-Root-of-Sum-ofSquares (SRSS) or the Complete Quadratic Combination (CQC) rules. Chopra. the peak nth-mode response as in Eq. (3.18) To develop a pushover analysis procedure consistent with RSA.9) . In such a response spectrum analysis (RSA).(3.1). (3.8. Section 13. Alternatively.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. 3.

the Applied Technology Council (ATC). region. denoted B1. SAC Steel Project Technical Office. Consulting Structural Engineers. can be combined according to Eq.12) is urno = Gnf rn Dn (3. rno . (3. CA 94804-4698. This modal pushover analysis (MPA) for linearly elastic systems is equivalent to the well-known RSA procedure (Section 3. The building has a basement level. this building is 45. and California Universities for Research in Earthquake Engineering (CUREE).2. this structure meets seismic code and represents typical medium-rise buildings designed for the Los Angeles. Third St.15 m (30 ft) on center.4 COMPARATIVE EVALUATION OF ANALYSIS PROCEDURES 3. 1660 W. CA 90017. 3. the peak value of the roof displacement due to the nth-mode. The levels of the 9-story building are numbered with respect to the ground level (see Fig. 1301 S.4.1 System and Excitation Considered The 9-story structure. which from Eq. 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.s* = mφ n n (3. The peak modal responses. Richmond. and 37.20) and the structure is pushed to the roof displacement. Typical floor-to-floor heights (for analysis purposes measured from center-of-beam to center1 2 Brandow & Johnston Associates. 13 . Although not actually constructed. The interior bays of the structure contain simple framing with composite floors. The bays are 9. 46th Street. 3. A benchmark structure for the SAC project.19 m (122 ft) in elevation. 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. Obviously Dn and An are available from the response (or design) spectrum. Los Angeles. in both directions. urno .2). SAC is a joint venture of three non-profit organizations: The Structural Engineers Association of California (SEAOC).21) 2 where Dn = An ω n . California.73 m (150 ft) in plan.73 m (150 ft) by 45. each determined by one pushover analysis. (3. with the ninth level being the roof. shown in Fig. with five bays each in the north-south (N-S) and east-west (E-W) directions.. 3.2). The columns are 345 MPa (50 ksi) steel wideflange sections.

sec2/ft).20). The floor-to-floor height of the basement level is 3.2 kips-sec2/ft). The first three vibration modes and periods of the building for linearly elastic vibration are shown in Fig.0 kips-sec2/ft). partitions. The seismic mass of the ground level is 9. and seventh levels at 1. which are seismic (tension) splices to carry bending and uplift forces. roofing and a penthouse located on the roof. 3.49 m (18 ft). fifth. The force distributions. To ensure that this structure remains elastic. Column splices. respectively. and 0.7 kips-sec2/ft). The simple model adopted here is sufficient for the objectives of this study.65×105 kg (66. for the first level is 1.83 m (6 ft) above the center-line of the beam to column joint. The building is modeled in DRAIN-2DX (Allahabadi and Powell. and for the ninth level is 1.49 sec. the vibration periods are 2. i. The column bases are modeled as pinned and secured to the ground (at the B-1 level). monolithic column pieces are connected every two levels beginning with the first level. Concrete foundation walls and surrounding soil are assumed to restrain the structure at the ground level from horizontal displacement. The 9-story N-S MRF is depicted in Fig. Each frame resists one half of the seismic mass associated with the entire structure.4. This model is based on centerline dimensions of the bare frame in which beams and columns extend from centerline to centerline. 1988) using the M1 model developed by Krawinkler and Gupta (1998). mechanical/electrical. and shear distortion of panel zones are neglected but large deformation (P∆) effects are included. we select a weak ground motion: the northsouth component of the El Centro (1940) ground motion scaled down by a factor of 4. The seismic mass of the above ground levels of the entire structure is 9. including the steel framing. The floor system is composed of 248 MPa (36 ksi) steel wide-flange beams in acting composite action with the floor slab. dimension. These force n distributions will be used in the pushover analysis to be presented later.e. for the second through eighth levels is 9. s* (Eq. The column lines employ two-tier construction.65 m (12 ft) and for the first floor is 5. 3. third. are located on the first. 0.2. if desired more complex models. 3.. The strength.85. such as those described in Gupta and Krawinkler (1999) can be used.89×105 kg (67. 14 . 3. The seismic mass of the structure is due to various components of the structure.96 m (13 ft).3.0 kips-sec2/ft).01×106 kg (69. ceiling/flooring. for the first three modes are shown in Fig.27.07×106 kg (73.00×106 kg (616 kips. floor slabs.of-beam) are 3.

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

7. 2. First three natural-vibration periods and modes of the 9story building 3.13)].796 0. 3. (b) shows the base shear Vbn t normalized relative to the weight W of the building. Each figure is organized in four parts: (a) shows the roof displacement urn t .487 −1.0272 −2.12 0. n = 1. determined by RHA [Eqs. 3.37 2.728 2. (3. Force distributions s* = mφn .05 2.85 sec 2 Floor 5th 4th 3rd 2nd 1st T1 = 2.72 −2.03 −1.33 2.05 1.31 −0.04 1. is shown in Figs. and 3 .2 Response History Analysis The structural response due to individual vibration modes. 3.75 1.38 0.05 s * 1 s * 2 s * 3 Fig.1 3.4.5 0 0.3. and 3 n 3.67 −1. respectively.5 Mode Shape Component 1 1.8 −2.5 −1 −0. 3.93 −1.4.6.44 1. (c) shows the joint rotation q n t of an external joint at 16 bg bg bg . and 3.39 3.13 −1. n = 1.5 Fig.51 0.94 2.49 sec 3 T = 0. 2.5.1 −2.12) and (3.61 2.9th 8th 7th 6th T = 0.27 sec Ground −1.

and ur 3o = 0. and 3. The peak values of floor displacements and story drifts determined by RHA. and rotations of external joints with moment connections are presented in Tables 3. the results for the roof displacement and top-story drift are shown in Fig.3. This is illustrated in Fig. and even larger in joint rotations. which are listed in Tables 3.urn relation. For a fixed number of modes included.12 cm. 3. are presented in Fig.422 cm. and (d) shows the Vbn . the exact response considering all modes.23 cm. two.2. respectively. especially in estimating the story drifts.9. The errors due to truncating the contributions of vibration modes beyond the third mode are negligible.15) and (3.2. (3.3. Combining the modal response histories for all modes gives the total response [Eqs. in particular. The same method was used to determine the peak values of many response quantities. where the second and third modal responses are a larger percentage of the top story drift compared to roof displacement. and the percentage errors due to truncation of higher modal contributions. 3.16)]. respectively. drifts in all stories. It is apparent that the first mode alone is inadequate. The linear relationship between the base shear and roof displacement for each mode implies that the structure did not yield.1. The peak values of the various response quantities are noted in these figures. or all modes. the peak roof displacement due to each of three modes is ur1o = 9. 3. As expected.8. 3.the roof level. larger in story drifts. ur 2o = 2. The peak values of displacements of all floors. errors are smallest in floor displacements.8. including one. and 3.1. two. and three vibration modes. Also included are the combined response due to one. but three modes—perhaps even two modes—are sufficient. with the first three modes provide essentially the exact response. 3. errors generally decrease as response contributions of more modes are included. three. consistent with the increasing significance of the higher mode response among these three sets of response quantities. 17 .

038 0.63E-03 2.25 × 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 0.088 -0.31E-03 Mode 2 -1.26E-04 -5.01E-04 -2.263 0.13E-04 9.0 -2.4 -41.097 0.7 7.38E-03 2.045 0.50E-03 4.94E-03 2.011 0.058 -0.22E-03 2.2 1.152 0.0 3 Modes -5.1 -2.133 -0.023 0.64E-03 3.227 0.91E-04 1.177 0.307 0.090 0.74E-03 1.73E-03 3.202 0.50E-03 2.3 Peak values of joint rotations (radians) from RHA for 0.192 0.069 0.08E-03 2.275 0.321 0.124 0.78E-04 -3.03E-03 1.2 -57.9 -23.125 0.6 9.261 -0.282 0.8 -10.47E-03 1.00E-03 1.6 4.4 0.266 0.311 0. 18 .15E-03 4.173 0.042 0.56E-03 2.325 0.203 0.01E-04 3.9 1.72E-03 3.350 0.097 0.85E-03 3.25 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 3 Modal Response Mode 1 2.245 0.407 0.4 -10.0 -46.2 -1.2 Peak values of story drift ratios (as % of story height) from RHA for 0.3 Table 3.9 3.282 0.9 -22.022 0.76E-03 1.260 0.0 -10.229 0.130 0.011 0.226 0.237 0.258 0.7 -50.156 0.008 -0.062 0.7 -19.006 0.231 -0.29E-03 2.264 Error (%) 1 Mode 2 Modes 3 Modes -23.4 -6.080 0.333 0.010 -0.3 -3.1 -14.1 Table 3.00E-03 1.336 0.226 0.33E-03 2.6 -1.225 0.8 -5.012 0.4 -1.237 0.1 -2.177 0.156 0.159 0.7 3.002 -0.65E-03 2.1 Building height is measured from the ground floor to the 9th floor.1 -19.5 -1.475 0.8 -1.401 Error (%) 1 Mode 2 Modes 3 Modes -23.2 -4.4 -53.032 -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 0.6 0.4 -0.0 7.44E-03 3.1 -11.7 2.1 -2.062 -0.2 -20.13E-03 2.124 0.88E-03 2.265 0.0 -2.9 8.266 0.24E-03 2.03E-03 1.042 0.413 RHA (all modes) 0.253 0.303 0.44E-03 1.117 0.015 -0.055 0.259 0.09E-03 Error (%) 1 Mode -23.6 -1.060 0.6 11.0 -0.03E-03 -6.202 0.3 -8.76E-03 1.2 0.8 -56.9 9.125 0.466 0.99E-03 2.2 9.173 0.6 -0.259 0.311 0.6 1.071 0.11E-04 -5.157 0.197 0.9 -16.181 0.8 1.65E-03 RHA (all modes) 2.009 -0.370 0.400 0.097 0.66E-05 -3.1 Peak values of floor displacements (as % of building height = 37.74E-03 1.89E-03 1.035 0.205 0.8 1.089 0.38E-04 2 Modes 2.3 -33.09E-03 1.31E-03 Joint Rotation (rad) Combined (RHA) 1 Mode 2.001 -0.2 -2.235 0.060 -0.265 RHA (all modes) 0.300 0.012 -0.378 0.235 0.179 0.89E-03 1.266 0.6 0.90E-03 3.74E-04 9.5 18.024 -0.3 19.152 0.4 -3.9 -15.364 0.45E-03 3.9 2.310 0.5 0.28E-04 1.4 -7.183 0.253 0.1 -0.8 -15.199 0.1 3.245 0.Table 3.3 -0.7 2 Modes -3.453 0.1 4.003 0.229 0.5 -2.202 0.406 0.63E-03 2.054 0.00E-03 2.42E-04 1.74E-04 6.4 -1.295 -0.09E-03 1.121 0.060 -0.9 -24.99E-03 Mode 3 3.88E-03 2.03E-03 3 Modes 2.260 0.043 0.026 0.4 -22.227 0.14 m3) from RHA for 0.069 0.11E-03 1.399 0.3 -0.7 4.14E-03 2.09E-03 2.

Response due to first mode: (a) roof displacement.00174 • 0 (c) r1 −5 0 5 10 15 Time (sec) 20 25 30 0. (c) joint rotation.1 0.1 −15 −7. Excitation is 0.1 0.15 u (cm) 9.25 × El Centro ground motion 19 .5. and (e) pushover curve.1 −15 −7. 3.0435 • 0 (b) V b1 −0.5 9.5 15 −0.1 5 x 10 −3 θ (rad) 0.1 /W 0. (d) force-deformation history.12 0 (d) 0 (e) 15 Fig.5 0 u (cm) r1 7.5 0 u (cm) r1 7.0435 9.12 • 0 (a) r1 −15 0.0435 Vb1 / W 0. (b) base shear.12 −0.

1 −5 −2.0252 • 0 (b) V −0. Response due to second mode: (a) roof displacement.5 0 u (cm) r2 2.00244 • 0 (c) r2 −5 0 5 10 15 Time (sec) 20 25 30 0.5 5 Fig.05 /W 0.5 u (cm) r2 0 • 2.0252 −2.23 −5 (a) 0.1 Vb2 / W 0.23 −2.0252 (d) 0 0 (e) −0.23 0. Excitation is 0. (b) base shear.1 −5 −2. 3.25 × El Centro ground motion 20 . (c) joint rotation.1 0.05 5 x 10 −3 b2 θ (rad) 0.5 0 u (cm) r2 2. (d) force-deformation history.5 5 −0. and (e) pushover curve.6.

and (e) pushover curve.000913 • 0 (c) r3 −2 0 5 10 15 Time (sec) 20 25 30 0.02 −1 −0.422 (e) −0.02 −1 −0. (d) force-deformation history.1 u (cm) r3 0 • 0.0108 −0. 3.02 0.0108 x 10 −3 (b) −0. (c) joint rotation.02 2 θ (rad) 0.02 /W V b3 0 • 0.02 −0. Response due to third mode: (a) roof displacement.5 0 u (cm) r2 0.422 −1 (a) 0. (b) base shear.7.5 1 −0.5 1 Fig.422 /W V b3 0 (d) 0 −0.5 0 u (cm) r3 0. Excitation is 0.25 × El Centro ground motion 21 .0108 −0.

12). and 3. and the computed response values are identical to the peak response values determined from RHA (Tables 3. story drifts.4. This pushover curve is consistent with the relationship between the base shear and roof displacement determined by RHA (Fig. and 3.2. respectively. pushing the structure. (3. and 3. Figure 3. considering one.4. leads to the pushover curves shown in Figs. 3. 3.5e. this value would be determined directly from the response (or design) spectrum.3).. 3. using the force distribution of Eq. and the errors in these estimates relative to the exact response from RHA considering all modes.422 cm .1. respectively.23 cm .7d).8) leads to the pushover curve shown in Fig. In RHA the errors arise only from truncating the responses due to higher modes. except for the algebraic sign associated with Γ1 and minor round-off errors. As suggested by Eq.9).6e and 3. the errors in the MPA results are generally larger than in RHA (Fig.ur relations determined by RHA (Figs.3). and ur 3o = 0.20) with n = 2 and 3 up to roof displacements ur 2 o = 2. story drifts. and external joint rotations in Tables 3.9 and Tables 3. 3. For a fixed number of modes included considered. and 3.18). 3.21). 3. confirming that MPA gives the exact values of the individual modal responses.6. The floor displacements. Implementing pushover analysis for the second and third modes. These values of the response quantities are identical to the peak response values determined from RHA (Tables 3. i. respectively. although both analyses led to identical peak values of the individual modal responses.2.6. 3.5.e. implying that the modal truncation errors are small if at 22 . (3. which would provide the Dn value to be substituted in Eq.10 and Tables 3. these pushover curves are consistent with the Vb .1 through 3. (3. two.12 cm. 3.6 present estimates of the combined response according to Eq.4) to roof displacement ur1o = 9. the floor displacements are proportional to the mode shape φ 1 because the structure remains elastic. 3. or three vibration modes.4.5. and it is apparent in the example considered that three modes provide most of the response (Fig. (3. and external joint rotations computed by pushover analysis are presented in Tables 3.20) with n = 1 (Fig. i. 3.6d and 3..3 Modal Pushover Analysis Implementing MPA for the fundamental vibration mode. (3. As for the first mode.1. pushing the structure using the force distribution of Eq. the value determined by RHA (Fig. Observe that the target roof displacement in each pushover analysis is identical to its exact value determined by RHA.3). 3.e.5d).5.3. and 3. 3. In practical application.4.7e and to the floor displacements. 3.

3.5 Ground 0 0.4 Displacement/Height (%) 0.03 −3 3 Mode 3 ∆ (cm) u (cm) r3 −15 15 9.5 0.12 • r1 −15 15 0 r1 0 0 0.9.1 0.8.25 × El Centro ground motion 23 .eps r 5 10 15 20 Time. Additional errors are introduced in pushover analysis due to the approximation inherent in modal combination rules.83 • r3 0 • 0.3 0.23 r2 −15 15 0 • 0. sec 25 30 Fig.eps fig3_9b. (a) Roof Displacement 15 (b) Top Story Drift 3 Mode 1 ∆ (cm) u (cm) 9.2 0.eps Ground 0 0.1 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.least three modes are included.422 r2 0 • 1. Response histories of roof displacement and top-story drift from RHA for 0.685 • −3 3 Mode 2 ∆ (cm) u (cm) • 2.3 0.2 0. 3.6 Fig. sec 25 30 r 5 10 15 20 Time.59 • Total (All Modes) 0 −15 0 ∆ (cm) u (cm) 0 −3 0 fig3_8.48 −3 3 1. Heightwise variation of floor displacements and story drifts from RHA for 0.4 Story Drift Ratio (%) 0.

267 -0.261 -0.89E-03 -2.9 -15.231 -0.253 RHA (All Modes) 0.3 -13.4 -4.296 -0.3 -19.3 -2.0 -46.2 -16.088 -0.9 3 Modes -12.74E-03 -1.6 -16.227 0.04E-03 3.173 0.267 0.133 0.31E-03 Joint Rotation (rad) Combined (MPA) 1 Mode 2.3 Table 3.47E-03 1.4 -53.99E-03 Mode 3 -3.4 0.270 0.090 0.282 0.0 -10.260 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 -0.062 -0.331 0.156 0.9 -13.179 0.22E-03 -2.9 -13.012 0.466 0.9 -14.002 0.229 0.9 -24.89E-03 -1.177 0.7 2 Modes -13.7 24 .036 -0.010 0.266 0.12E-03 1.079 0.042 0.009 0.2 -0.25 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Displacement /Height (%) Modal Response Combined (MPA) Mode 1 Mode 2 Mode 3 1 Mode 2 Modes 3 Modes 0.03E-03 3.00E-03 2.0 -2.203 0.038 -0.9 -16.328 0.008 -0.078 0.2 1.2 -12.6 Peak values of joint rotation (radians) from MPA for 0.9 -12.78E-04 2 Modes 2.080 0.89E-03 1.267 0.023 -0.253 0.042 0.05E-03 3.310 0.44E-03 3.03E-03 6.321 0.43E-04 -1.245 0.022 0.9 -8.8 -14.5 Peak values of story drift ratios (as % of story height) from MPA for 0.3 -12.407 0.015 -0.7 -21.069 0.097 0.259 0.92E-04 -1.133 -0.3 1.94E-03 2.024 0.4 -14.7 -19.2 -4.048 0.0 -16.3 -41.370 0.197 0.125 0.011 -0.227 0.74E-04 -6.73E-03 3.1 -19.22E-04 2.011 0.53E-04 -9.374 0.124 0.89E-03 2.09E-03 Mode 2 1.259 -0.071 0.3 -14.5 -16.177 0.9 -15.9 -15.73E-05 3.226 0.230 -0.6 -19.24E-03 -2.157 0.97E-03 1.2 -57.026 0.060 -0.14 m) from MPA for 0.38E-03 2.152 0.2 -11.1 -0.90E-04 -9.300 0.045 -0.8 -23.63E-03 -2.89E-03 2.4 Peak values of floor displacements (as % of building height = 37.133 0.006 -0.9 -11.4 -2.336 0.062 0.032 0.253 -0.42E-04 -1.44E-03 -1.332 0.33E-04 5.90E-03 1.74E-03 1.4 -9.3 -9.069 0.3 -33.4 -9.313 0.055 0.76E-03 -1.264 Error (%) Modal Response 1 Mode 2 Modes 3 Modes -23.57E-03 1 Mode -23.40E-04 5.282 -0.2 -20.28E-03 2.4 -22.048 0.276 0.260 0.31E-03 Error (%) RHA (all modes) 2.097 0.9 -18.270 0.179 0.6 -17.15E-03 2.65E-03 2.125 0.9 -14.08E-03 2.0 -0.8 -22.63E-03 2.181 0.03E-03 -1.003 -0.058 -0.7 -50.235 -0.15E-03 1.0 -18.1 -11.4 -22.237 -0.385 0.9 -14.6 -15.7 2.106 0.1 -18.09E-03 1.9 -13.152 0.09E-03 -1.8 -15.38E-03 3 Modes 2.203 0.270 0.173 0.237 0.203 0.00E-03 2.00E-03 2.4 -7.4 -4.203 0.235 0.272 0.80E-04 3.274 0.203 0.106 0.25 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Modal Response Mode 1 -2.336 RHA (all modes) 0.76E-03 1.Table 3.286 0.245 -0.260 -0.72E-03 3.253 0.00E-03 -1.012 -0.4 -19.96E-03 2.8 -56.156 0.00E-03 3.31E-03 2.229 0.157 0.060 0.8 -15.322 0.401 Error (%) 1 Mode 2 Modes 3 Modes -23.117 0.4 -5.03E-03 1.00E-03 1.1 -11.043 -0.285 0.230 0.4 -11.09E-04 -3.7 -15.4 1.9 -15.121 0.9 -13.2 Table 3.001 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.5 0. shading indicates modal combination error 25 .3 0. Heightwise variation of floor displacements and story drifts from MPA for 0.2 0.4 Story Drift Ratio (%) 0.eps Ground 0 0.6 Fig.2 0. 3.4 Displacement/Height (%) 0.3 0.1 0.5 Ground 0 0.25 × El Centro ground motion.10.eps fig3_10b.1 0.

26 .

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

Simultaneously solving these coupled equations and using Eq. (3. qr t = 0 for all 28 bg . (4. premultiplying by φ T . (3.3) in Eq. However. (4. (4. however.5) This resisting force depends on all modal coordinates qn t .6) The solution of Eq. implying coupling of modal coordinates because of yielding of the structure. in principle. For linear systems.2).Substituting Eq. (4.2). where φ n are now the modes of the corresponding linear system. (4.3) will. 2. This approximate RHA procedure is the preliminary step in developing a modal pushover analysis procedure for inelastic systems.9) involves T Fsn = Fsn qn .4) where the only term that differs from Eq. Equation (4. sign u n b g b g bg (4.… N (4. (4.n t given by Eq. sign u = .3). 4.9) for linearly elastic systems.4) represents N equations in the modal coordinates qn .4)] leads to the uncoupled modal response history analysis (UMRHA) procedure.6) for inelastic systems will no longer be described by Eq. (3.sn ug t bg b g bg bg (4. (3. (4.” implying that other “modes” will also contribute to the solution. (4.2 UNCOUPLED MODAL RESPONSE HISTORY ANALYSIS bg Neglecting the coupling of the N equations in modal coordinates [Eq.6b) are mu + cu + f s u. (3.8) because qr t will generally be nonzero for “modes” other than the nth “mode. Eq. The spatial distribution s of the effective earthquake forces is expanded into the modal contributions sn according to Eq. these equations are coupled for inelastic systems.Gn ug t Mn bg n = 1. and using the mass.2). sign qn = φ n f s u n . The equations governing the response of the inelastic system to p eff.and classical n damping-orthogonality property of modes gives F qn + 2z nw n qn + sn = . give the same results for u t as obtained directly from Eq. Unlike Eq.4) is rarely solved because it offers no particular advantage over Eq.

(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

s n u g t .0 ¥ El Centro ground motion: (a) exact solution by NL-RHA. 4.89 50 r1 0 u (cm) 0 • 78. Fsn / Ln ¨ g(t ) u (a) Static Analysis of Structure (b) Dynamic Analysis of Inelastic SDF System Fig.n t = . where u g t = 3.Forces sn Unit mass An(t ) st rn ωn.51 • ur2 (cm) 0 ur2 (cm) n=3 14. n = 1.4. Roof displacement due to p eff.02 −150 n=2 −150 50 n=2 14. 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) • 75.112 5 10 15 20 Time (sec) 25 30 −10 0 −10 0 Fig. and 3. 4. 2.575 5 10 15 20 Time (sec) 25 30 ur3 (cm) 0 • 5. ζn.3.

and (3) the Fsn Ln . Such comparison for roof displacement and top-story drift is presented in Figs. (2) the superposition of responses to p eff.4) and (4.4 and 4.956 5 10 15 20 Time (sec) 25 30 ∆r3 (cm) 0 • 5. n = 1.855 ∆r1 (cm) 0 ∆r1 (cm) n=2 • 6.1 Underlying Assumptions and Accuracy The approximate solution of Eq.5)] is neglected.817 5 10 15 20 Time (sec) 25 30 −10 0 −10 0 Fig. (4.2… N ) according to Eq. 4. and 3. The errors are slightly larger in drift than in displacement. and (b) approximate solution by UMRHA af bg bg 4.5.0 ¥ El Centro ground motion: (a) exact solution by NL-RHA.(a) Nonlinear RHA 20 n=1 20 (b) UMRHA n=1 • 5. Top story drift due to p eff.n t = . 4.2.s n u g t . respectively.008 • 6. (4.n t (n = 1. 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. where u g t = 3.15) is strictly valid only for linearly elastic systems.6) by UMRHA is compared with the “exact” solution by nonlinear RHA.744 • 0 −20 10 n=3 −20 10 n=3 ∆r3 (cm) 0 • 5.0 ¥ El Centro ground motion. the errors in either response quantity are only a few percent. 2. but even for this very intense excitation.33 0 −20 20 −20 20 n=2 ∆r2 (cm) 0 ∆r2 (cm) 7.5.Dn relation is 33 bg bg . both for 3. this intense excitation was chosen to ensure that the structure is excited well beyond its linear elastic limit. (3.

A bilinear idealization of this pushover curve for the nth-“mode” is shown in Fig.urn pushover curve to the Fsn Ln .4. (4.20). 4. (3.20) to some bg pre-determined roof displacement.10) . Although most commercially available software cannot implement such displacement-controlled analysis. the only force distribution that produces displacements proportional to f n is given by Eq. it can conduct a force-controlled nonlinear static analysis with an invariant distribution of lateral forces.1. and the base shear Vbn . (3. within the linearly elastic range of the structure. (4. which is different than the Fsn Ln .2.9) should be determined by nonlinear static analysis of the structure as the structure undergoes displacements u = Dnφ n with increasing Dn . when specialized for linearly elastic systems it is identical to the RHA procedure of Section 3. When implemented by commercially available software. At the yield point.9). (4. How to convert this Vbn . Therefore. However.8) in UMRHA.12) for floor displacements. (4.2 Properties of the nth-“mode” Inelastic SDF System How is the Fsn Ln . (3.approximated by a bilinear curve to facilitate solution of Eq.8) before it can be solved? Because Eq.8) governing Dn t is based on Eq. (4. 4.Dn curve.6a.Dn relation to be determined in Eq.Dn relation? The two sets of forces and displacements are related as follows: V Fsn = bn Γn Dn = urn Γ nf rn 34 (4. Although several approximations are inherent in this UMRHA procedure. such nonlinear static analysis provides the so-called pushover curve. the relationship between lateral forces f s and Dn in Eq. What is an appropriate invariant distribution of lateral forces to determine Fsn ? For an inelastic system no invariant distribution of forces can produce displacements proportional to f n at all displacements or force levels. The structure is pushed using the force distribution of Eq. is plotted against roof displacement urn . this distribution seems to be a rational choiceeven for inelastic systemsto determine Fsn in Eq. the base shear is Vbny and roof displacement is urny . Therefore we impose this constraint in developing the UMRHA procedure in this section and modal pushover analysis in the next section. The overall errors in the UMRHA procedure are documented in the examples presented in Section 4.

4.6b.11) * in which M n = Ln Γn is the effective modal mass (Chopra.12) implying that the initial slope of the curve in Fig.5). 4.8).13) This value of Tn .2.6a is 2 kn = ω n Ln . (4. 4. Knowing Fsny Ln and Dny from n Eq. 35 .V bn (a) Idealized Pushover Curve F sn /L n (b) F sn / L − D Relationship n n Idealized Vbny Actual 1 αnkn V bny /M * n 1 α ω2 n n kn 1 1 ωn 2 u u rny rn Dny = ur n y / Γn φ r n D n Fig. the initial slope of the pushover curve in Fig.10 enables conversion of the pushover curve to the desired Fsn Ln − Dn relation shown in Fig. 4. which is not a meaningful quantity.6b is w 2 . Section 13. 2001.11). should be used in Eq. In contrast. 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 (4.6. The two are related through Fsny Ln 2 = w n Dny (4. where the yield values of Fsn Ln and Dn are Fsny Ln = Vbny * Mn Dny = urny Γ nf rn (4. Properties of the nth-“mode” inelastic SDF system from the pushover curve Equation 4. which may differ from the period of the corresponding linear system. (4.

6a).n t . 4. and modes.6b) by utilizing Eq. which is summarized next as a sequence of steps. (3. φn . Consider a nonlinear static analysis of the structure subjected to lateral forces distributed over the building height according to s* [Eq. the first two or three modes will suffice. Dn (t ) . n 3. Combine the “modal” responses using Eqs.3 Summary The inelastic response of an N-story building with plan symmetric about two orthogonal axes to earthquake ground motion along an axis of symmetry can be estimated as a function of time by the UMRHA procedure just developed. Calculate the peak value. r o .3b) with force-deformation relation of Fig. is now determined by 36 bg bg . of the nth-“mode” inelastic SDF system (Fig.6b. 4. For the nth-mode. (3. (3. 2. Idealize the pushover curve as a bilinear curve (Fig.16) to determine the total response.20)]. 4. Calculate histories of various responses by Eqs. Typically. the peak value of Dn t . ωn .11). Convert the idealized pushover curve to the Fsn Ln − Dn relation (Fig.15) and (3. Compute the natural frequencies. This value of the roof displacement is given by Eq.13). 4.12) and (3. and pseudo-acceleration history.3 MODAL PUSHOVER ANALYSIS bg bg A pushover analysis procedure is presented next to estimate the peak response rno of the inelastic MDF system to effective earthquake forces p eff . 4. 9. An t . (4. 8. with the structure is pushed to the roof displacement urno .21) where Dn . Repeat Steps 2 to 6 for as many modes as required for sufficient accuracy. details are available in Appendix A: 1. develop the base-shear – roof-displacement ( Vbn − urn ) pushover curve for the force distribution s* [Eq. (3.4. 5. Compute the deformation history.20)].2. 4. 6. for linearly-elastic vibration of the building. of the total response r t obtained in Step 8. 7. n (3.

Dn .g. Sections 7. etc. alternatively. to obtain an estimate of the peak value ro of the total response. It provides results for elastic buildings that are identical to the well-known RSA procedure (Section 3. (3. (4. As shown in Sections 3. it seems reasonable because it provides results for elastic buildings that are identical to the well-known RSA procedure described in Section 3. of the nth-“mode” inelastic SDF system (Fig. 5. 4. governed by Eq. Eq. each determined by one pushover analysis.6 and 7. as described in Section 4.12)..2 and 3. plastic hinge rotations. 2001. rno also represents the exact peak value of the nth-mode contribution rn t to response r t . (4.18). At this roof displacement. and (2) it provides the exact modal response for elastic systems.solving Eq. bg bg bg bg Thus.6b by solving Eq. 4.8). (3. story drifts.n t . it can be determined from the inelastic response (or design) spectrum (Chopra. Calculate the peak roof displacement urno associated with the nth-“mode” inelastic SDF system from Eq. are combined using an appropriate modal combination rule.3b) with force-deformation relation of Fig. This pushover analysis. However. 37 . seems reasonable.3. e. joint rotations. 4. Compute the peak deformation. (4. the pushover analysis provides an estimate of the peak value rno of any response rn t : floor displacements.3.2.6). Steps 1 to 4 of the MPA are same as those for UMRHA. The response value rno is an estimate of the peak value of the response of the inelastic system to p eff. The peak “modal” responses rno . which is summarized next as a sequence of steps. for elastic systems. as mentioned earlier. although somewhat intuitive for inelastic buildings. we will refer to rno as the peak “modal” response even in the case of inelastic systems.1 Summary The peak inelastic response of a building to earthquake excitation can be estimated by the MPA procedure just developed. details are available in Appendix B. This application of modal combination rules to inelastic systems obviously lacks a theoretical basis. the lateral force distribution used possesses two properties: (1) it appears to be the most rational choice among all invariant distribution of forces. or from the inelastic response (or design) spectrum.4.2.21).3) because. 6.8).

4. the peak roof displacement due to each of the three “modes” is ur1o = 48. three “modes” included.1 and 4. rno .4. and ur 3o = 2.7. The peak values of displacements of all floors and drifts in all stories are presented in Tables 4. and the errors in the approximate results are shown in Fig.0 to 3. the combined response due to three “modes”.18). 4.2). Observe that errors tend to decrease as response contributions of more “modes” are included.2.0. the 38 . (3. Repeat Steps 3 to 8 for as many “modes” as required for sufficient accuracy.4 COMPARATIVE EVALUATION OF ANALYSIS PROCEDURES The response of the 9-story building described earlier is determined by the two approximate methods: UMRHA and MPA.9. ur 2o = 11.4. This deficiency also implies that.1). Typically. Figure 4. also included are the combined responses due to one. although the trends are not systematic as when the system remained elastic (Section 3.7 cm. and three modes are compared with the “exact” values in Fig. The peak values of response are as noted. and the “exact” response from nonlinear RHA for the roof displacement and top-story drift. the UMRHA procedure lacks a rigorous theory.5 ¥ the El Centro ground motion including the response contributions associated with three “modal” inelastic SDF systems. This is to be expected. and compared with the results of a rigorous nonlinear RHA using the DRAIN-2DX computer program. say. is presented next. 4. in particular. two. in contrast to modal analysis (Section 3. 4. the first two or three “modes” will suffice. and the percentage errors in the approximate results. extract from the pushover database values of other desired responses. At urno . Determine the total response by combining the peak “modal” responses using the SRSS combination rule of Eq.7 shows the individual “modal” responses. subtract the yield value of hinge rotation to determine the hinge plastic rotation.8. 8. 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. respectively. two. From the total rotation of a plastic hinge. The peak values of floor displacements and story drifts including one. and three “modes”.53 cm. the “exact” results.3 cm. determined by the UMRHA procedure.1 Uncoupled Modal Response History Analysis The structural response to 1. 9. with.

62 5.5 (Tables 4. Response histories of roof displacement and top-story drift due to 1. (a) Roof Displacement (b) Top Story Drift 80 12 0 • 3. For each excitation.85.5.1 and 4.53 ∆r3 (cm) 12 0 • 2.5.0.1 and 3. 1. we investigate how the errors in the UMRHA vary with the deformation demands imposed by the ground motion.response is much less accurate (Tables 4. However. 0. for a fixed number of “modes” included. For this purpose the UMRHA and exact analyses were repeated for ground motions of varying intensity.24 • NL−RHA r −12 12 ∆r (cm) u (cm) 0 5 10 15 20 Time. recall that the computed errors have been presented earlier for ground motion multipliers 0.2) if the system yields significantly versus if the system remains within the elastic range (Tables 3.2) and 1.3 11. 0.6 5 10 15 20 Time.7 • −80 80 ∆r1 (cm) u (cm) "Mode" 1 −12 r2 0 ∆r2 (cm) 12 0 u (cm) −80 80 −12 r3 0 • 2. sec 25 30 r −12 0 Fig.0. defined as the El Centro ground motion multiplied by 0.1 −12 12 ∆r (cm) u (cm) UMRHA 3 "Modes" 0 • 7.25. 2. and 3.5 × El Centro ground motion: individual “modal” responses and combined response from UMRHA.25 (Tables 3. 4. in particular.38 6. 0.2).0.7.2). the errors in responses computed by UMRHA including three “modes” relative to the “exact” response were determined. Next. sec 25 30 • 48. 1. the degree to which the system deforms beyond its elastic limit.1 and 3.44 • "Mode" 2 r1 0 • 48. and total response from NL-RHA 39 .1 and 4. the errors in story drifts are larger compared to floor displacements.88 u (cm) "Mode" 3 −80 80 0 −80 80 0 −80 0 • 44. just as for elastic systems.75.

Heightwise variation of error in floor displacements and story drifts estimated by UMRHA including one.5 1 1. 4.5 1 1. Two versions of the pushover curve are 40 .8.5 Displacement/Height (%) 2 Ground 0 0.5 Fig.20)] for the first three n modes of the system are shown in Fig. 4.5 Story Drift Ratio (%) 2 2. or three “modes” for 1.9. 4. (3.10b).5 ¥ El Centro ground motion Figure 4. To interpret these results. 4.10 summarizes the error in UMRHA as a function of ground motion intensity. Heightwise variation of floor displacements and story drift ratios from UMRHA and NL-RHA for 1. it will be useful to know the deformation of the system relative to its yield deformation.(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. Shown is the error in each floor displacement (Fig. two.10a). For this purpose. the pushover curves using force distributions s* [Eq. indicated by a ground motion multiplier. with the peak displacement of each “modal” SDF system noted for each ground motion multiplier. in each story drift (Fig. and the error envelope for each case.11. 4.5 × El Centro ground motion (a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th (b) Story Drift Ratios Floor 4th 3rd 2nd 1st UMRHA 1 "Mode" 2 "Modes" 3 "Modes" Floor 5th 5th 4th 3rd 2nd 1st UMRHA 1 "Mode" 2 "Modes" 3 "Modes" Ground −60 −40 −20 0 Error (%) 20 40 60 Ground −60 −40 −20 0 Error (%) 20 40 60 Fig.

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. 41 . UMRHA lacks a rigorous theory and is based on several approximations. Perhaps this explains why the errors are large at this excitation intensity. and (4) are larger in story drift compared to floor displacements. determined from “exact” analyses. (3. the ductility factor for the first mode system is only 1.2. For more intense excitations. the first reason mentioned above seems to be the primary source for the errors. even though the system remains essentially elastic.11).20) with n = 1. the values determined by RHA of the nth-mode inelastic SDF system (Fig. and.7).53 cm.1). the errors in truncating the higher mode contributions are negligible. The structure is pushed using the force distribution of Eq.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. 11.included: the actual curve and its idealized bilinear version. and 2.11a). the peak values of displacements at all floors (Table 4.4). Figure 4.10) for each “modal” inelastic SDF system (Figs 4. (4.4) to roof displacements urno = 48. Additional errors are introduced in UMRHA of systems responding beyond the linearly elastic limit for at least two reasons. Each of these three pushover analyses provides the pushover curve (Fig. (3) maintain roughly similar values for more intense ground motions.0. and plastic hinge rotations at the external beam end at each floor level (Table 4. as mentioned in Section 4. First. and 3 (Fig. drifts in all stories (Table 4.75. The system remains elastic up to ground motion multiplier 0. 4.0. 4. The location of plastic hinges and their rotations.7 cm. were noted but are not shown here. 2.6 and 4. Second.75.3 cm. respectively. the pushover curve for each “mode” is idealized by a bilinear curve in solving Eq.5). as mentioned in Section 3.2.4.3).4. The idealized curve for the first “mode” deviates most from the actual curve near the peak displacement corresponding to ground motion multiplier 1. 3. (2) increase rapidly as the ground motion multiplier increases to 1. 4.5 ¥ the El Centro ground motion. 4.01 (Fig.

4 -7. 4.938 1.5 9.241 -1.220 -0.942 -0.3 42 .410 -1.9 5.863 1.877 0.0 11.5 18.751 1.3 6.049 -0.663 0.068 0.5 28.317 0.003 -31.983 1.5 1 1.1 8.2 1.472 1.9 16.298 -0.1 3.676 0.707 1.2 6.698 1.8 1.241 1.540 0.033 0.407 -10.071 0.126 0.1 1.811 1.820 -19.426 -1. Noted 60 50 40 (b) Story Drifts Error Envelope Error for Story No.513 0.104 0.376 -1.819 2.806 0.8 14.214 0.10.5 2 GM Multiplier 2.018 0.5 3 Fig.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 -1.6 1.088 10.3 8.5 × 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” -0.668 -23.256 -1.484 0.3 25.4 4.575 -41.7 31.8 1.201 1.376 1.138 1.938 -1.130 0.3 1.1 Peak values of floor displacements (as % of building height = 37.121 -0.044 1.900 -10.513 -0.5 28.430 1.772 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 1.1 1.763 -15.350 -0.2 4.0 11.298 0.366 0.2 1.8 0.333 0.366 -0. Errors in UMRHA as a function of ground motion intensity: (a) floor displacements.260 -15.372 -1.0 -9.1 0.5 3 0 0 0.982 9.490 -1.315 -0.971 1.133 1.6 4.727 1.293 1.072 1.256 1.055 -0.338 -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 0.065 0.372 1.9 31.071 -0.554 1.200 8.009 -0.070 1.120 1.8 1.942 1.169 0.379 1.2 Peak values of story drift ratios (as % of story height) from UMRHA for 1.852 1.8 1.945 -37.4 -1.616 -0.8 1.5 -3.371 0.844 -25.806 -0.9 12.057 -0.283 1.5 2 GM Multiplier 2.478 0.820 -0.2 22.9 31.856 2.7 14.135 9.373 -0.526 -0.6 2.201 -1.863 0.136 1.0 1.220 0.187 -0.370 -0.096 0.3 0. and (b) story drifts Table 4.079 0.003 0.248 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.14 m) from UMRHA for 1.072 -1.216 1.2 12.201 -0.410 1.5 10.010 0.371 -0.7 Table 4.0 2.4 1.5 9.256 1.473 -22.914 -0.722 0.0 7.0 9.226 -0.202 11.214 -0.490 1.338 1.235 -0.154 0.9 12.663 -0.5 1 1.914 2.495 1.291 0.(a) Floor Displacements 60 50 40 Error Envelope Error for Floor No.

” and three “modes.4).3 through 4. Obviously.13 with Fig.3 and 4. 43 . 4. 4. Furthermore.5). 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%.18). For example. the results were not always accurate.4. The first “mode” alone is inadequate.” respectively. MPA failed to identify the plastic hinges at the column bases in Fig. 4. even if three “modes” are included (Fig. this error is not especially significant because the hinge plastic rotation is very small. However. and are within a few percent of the exact values for the upper stories.Figures 4.14. significant improvement is achieved by including response contributions due to the second “mode”. considering one. especially in estimating the story drifts (Fig. Pushover analysis seems to be inherently limited in computing accurately hinge plastic rotations. The locations of plastic hinges shown in Fig. 4.12 and Tables 4. As shown in Figs. Fortuitously. 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. the errors in the modal pushover results are. two. The errors are especially large in the hinge plastic rotations estimated by the MPA procedure. significantly smaller than in UMRHA (compare Fig. but the third “mode” contributions do not seem especially important (Fig.12 and 4. The drifts are underestimated by up to 13% in the lower stories. overestimated by up to 18% in the middle stories.4).13 and Tables 4.13 and Table 4.9 and Tables 4. and three “modes.” two “modes. were determined by four analyses: MPA considering one “mode. 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.1 and 4.14. and the errors in these estimates relative to the exact response from nonlinear RHA. 4. for two or three modes included. 4. the hinges identified in beams at the 6th floor were at variance with the “exact” results.12 and Tables 4. 4. the third “mode” does not contribute because this SDF system remains elastic (Fig. 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 with Tables 4. in general.3 and 4. Observe that the primary contributor to plastic hinge rotations in the lower stories is the first “mode” and the second “mode” in the upper stories.3 and 4. although the error is recorded as 100% if MPA estimates zero rotation when the nonlinear RHA computes a non-zero value.13 and Tables 4. however.” and nonlinear RHA.3 and 4.11c). 4. 4. (3.5 present estimates of the combined response according to Eq.2).

5 2 3 4000 2000 0 0 Actual Idealized 0.5.25 0.6 cm.14 y by Base Shear (kN) 8000 6000 4000 1.5 0. V = 7616 kN. 0. “Modal” pushover curves with peak roof displacements identified for 0. and 3.0. α = 0.5 1 0.2 cm.9 cm.25 1 2 4 6 Roof Displacement (cm) 8 10 Fig. V = 4952 kN. 1.0.75.5 2 3 Actual Idealized 2000 0 0 0.5 0.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36.5.75 0.85 0.25.11. V = 5210 kN.13 y by Base Shear (kN) 8000 6000 1 1.25 5 10 15 20 Roof Displacement (cm) 25 (c) "Mode" 3 Pushover Curve 12000 10000 u = 4. 1. α = 0.85 0.5 0. 0.0 × El Centro ground motion 44 .19 y by 3 2 1.75 Actual Idealized 20 40 60 Roof Displacement (cm) (b) "Mode" 2 Pushover Curve 80 12000 10000 u = 9.85 Base Shear (kN) 8000 6000 4000 2000 0 0 0. 2. α = 0. 4.75 0.

5 Story Drift Ratio (%) 2 2. UMRHA is essentially exact.15) for ground multipliers larger than 1.75. each story drift (Fig.12 Heightwise variation of floor displacements and story drift ratios from MPA and NL-RHA for 1. 4. implying that the modal combination errors contained in these procedures are acceptable.15 summarizes the error in MPA considering three “modes” as a function of ground motion intensity.3. an observation with practical implications.4. identified in Section 3. the errors in MPA were fortuitously smaller than in UMRHA (compare Figs. 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.5 Displacement/Height (%) 2 Ground 0 0. implying excitations intense enough to cause significant yielding of the structure. and the error envelope for each case. The errors are only weakly dependent on ground motion intensity (Fig.4 also apply to MPA.0. 4. now standard in engineering practice. whereas MPA contains errors inherent in modal combination rules.5 1 1.10 and 4.15a). While various sources of errors in UMRHA. shading indicates errors in MPA including three “modes” Figure 4.(a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th NL−RHA MPA 1 "Mode" 2 "Modes" 3 "Modes" (b) Story Drift Ratios Floor Floor 5th 4th 3rd 2nd 1st 5th 4th 3rd 2nd 1st NL−RHA MPA 1 "Mode" 2 "Modes" 3 "Modes" Ground 0 0.15). indicated by a ground motion multiplier. the errors in MPA were larger for ground motion multipliers less than 0.15b). 4.5 1 1. However.3 for elastic systems (or weak ground motions). As mentioned in Section 3. the MPA procedure is equivalent to the RSA procedure. Shown is the error in each floor displacement (Fig. 45 .2 and 3.4.5 Fig. implying excitations weak enough to limit the response in the elastic range of the structure. In this case. 4. as discussed in Sections 3.5 × El Centro ground motion. 4.

story drifts. and three “modes” for 1. Errors in floor displacements. two.(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. 4.13.5 ¥ El Centro ground motion 46 . and hinge plastic rotations estimated by MPA including one.

3 11.00E+00 0.02E-03 3.1 46.756 0.2 -3.640 1.00E+00 2 “Modes” 7.00E+00 0.233 1.844 -7.8 -29.00E+00 0.435 0.72E-03 7.5 -6.668 -13.4 -8.6 -9.614 0.895 1.879 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 Peak values of hinge plastic rotations (radians) from MPA for 1.804 1.315 0.9 2.60E-03 2.2 -100.55E-03 3.36E-03 6.8 0.2 -100.351 -0.116 1.8 -29.9 1.99E-03 6.9 -4.7 1.9 1.0 -100.503 -1.00E+00 0.372 0.331 1.756 0.60E-04 7.399 0.72E-03 7.2 6.414 1.2 11.473 -15.575 -53.36E-03 6.399 0.6 1.202 8.737 1.37E-04 Error (%) “Modal” Response 1 “Mode” -32.517 1.003 -16.76E-03 4.066 -0.429 0.705 -1.00E+00 0.6 -44.6 -7.53E-03 7.88E-03 1.012 1.4 Peak values of story drift ratios (as % of story height) from MPA for 1.018 0.910 1.5 2.8 1.260 -14.2 0.5 10.336 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.00E+00 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th “Mode” 2 0.116 1.942 6.00E+00 1 “Mode” 7.00E+00 “Mode” 3 0.478 0.6 7.7 -12.18E-03 7.101 -0.426 15.2 -100.015 0.00E+00 0.1 62.37E-03 1.5 7.76E-03 4.066 -0.116 1.304 -1.4 1.72E-03 7.5 2.00E+00 0.50E-10 3.1 18.00E+00 0.00E+00 0.705 1.8 “Mode” 3 -1.581 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.Table 4.2 0.498 1.00E+00 0.156 -0.683 1.9 5.5 2.0 -100.36E-03 6.009 0.76E-03 4.250 0.1 1.00E+00 0.88E-03 0.820 -7.738 1.338 1.527 -0.00E+00 0.55E-03 3.154 0.088 12.22E-10 NL RHA 1.8 1.311 0.3 13.752 1.371 -0.053 -1.259 1.8 17.02E-03 0.980 -0.00E+00 0.135 -7.071 -0.1 46.049 -0.130 0.667 -1.007 1.0 -5.737 0.666 Table 4.3 -3.02E-03 3.3 -3.982 13.220 1.00E+00 0.8 -29.19E-10 3.6 13.018 -0.05E-03 2.581 0.37E-03 1.7 1.5 1.5 10.0 47 .6 -44.1 -8.222 0.514 -1.0 1.8 -12.5 × El Centro ground motion Hinge Plastic Rotation (rad) “Modal” Response Combined (MPA) “Mode” 1 7.197 -0.728 1.667 1.611 0.200 8.640 -1.266 -0.298 0.298 -0.76E-03 4.00E+00 0.687 0.641 1.2 1.9 -100.518 1.980 0.1 46.9 -100.00E+00 0.3 Peak values of floor displacements (as % of building height = 37.9 7.945 -49.176 0.8 -6.0 -50.033 -0.781 0.6 -8.26E-04 9.0 -100.516 0.10E-02 9.1 62.118 0.305 -0.3 1.900 -0.053 1.015 0.00E+00 3.733 1.37E-03 1.125 -1.068 0.911 0.694 1.652 1.209 1.00E+00 0.244 0.594 -1.745 1.2 -4.763 -14.37E-03 1.0 2 “Modes” -32.00E+00 3 “Modes” 7.190 -0.02E-03 0.14 m) from MPA for 1.00E+00 0.8 -6.55E-03 3.00E+00 0.168 -0.72E-03 7.233 1.071 0.0 -50.304 1.8 7.057 0.1 62.222 0.105 0.007 1.018 -0.0 3 “Modes” -32.503 1.407 -27.6 13.8 -4.88E-03 0.781 0.895 1.0 15.9 0.055 0.220 1.1 13.5 Table 4.36E-03 6.

resulting in the pushover curve shown in Fig.53 cm. and the structure is pushed using the force distribution of Eq.17 and Tables 4.6 through 4. The first “mode” alone provides adequate estimates of floor displacements. The response contributions of the second and third modes are the same as before (Tables 4.4.4.20) with n = 1 to a target roof displacement. The errors are especially large in the plastic hinge rotations estimated by the MPA procedure even if three “modes” are included (Fig.8).3 Modal Pushover Analysis with Gravity Loads To evaluate the accuracy of the dynamic response of the system.” however.18 and 48 . Significant improvement is achieved by including response contributions due to the second “mode. and plastic hinge rotations at selected external beam end at each floor level (Table 4.7 cm and ur 3o = 2. two. Results of such analyses for the selected building subject to 1.5 × El Centro ground motion are presented below.5). The story drifts are underestimated by up to 7% in the lower stories. drifts in all stories (Table 4.6 and 4.7. and overestimated by up to 29% in the middle stories and by up to 13% in the upper stories. (3.7). the structure is pushed using the force distribution of Eq. They are now included in the pushover analysis of the structure for the first “mode” only. as are the roof displacements ur 2o = 11.18 and Tables 4. Obviously gravity load effects will influence the seismic demands due to the first “mode. As shown in Fig.7). 4.16 are unchanged. Starting with its initial state under gravity loads. The pushover curves for the second and third modes included in Fig.17 and Tables 4. the results presented so far did not include gravity load effects.17 and 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%.3 .20) with n = 1 (Fig.4) to roof displacement ur1o = 52. Static analysis of the structure for gravity loads provides the initial state—forces and deformations—of the structure. but it is inadequate in estimating the story drifts (Fig.6 and 4.4. Figures 4.18 and Tables 4. 4.7).18) considering one.” but not the contributions of higher “modes.6). the third “mode” contributions do not seem especially important (Fig.8 present estimates of the combined response according to Eq. and three “modes.6 and 4.” and the errors in these estimates relative to the exact response from nonlinear RHA.0 cm. Also presented at these roof displacements are the displacements at all floors (Table 4. 4. 4.” these effects will modify the combined modal response as well as the results of nonlinear RHA. 4. 3.11a). 4. (3. (3. 4.16a—which is slightly different than the one excluding gravity loads (Fig.

(a) MPA.5 1 1.Table 4.5 2 GM Multiplier 2.5 2 GM Multiplier 2. Noted 60 50 40 (b) Story Drifts Error Envelope Error for Story No.14. and (b) story drifts 49 .8).15.5 3 0. Most pushover analysis procedures do not seem to compute to acceptable accuracy plastic hinge rotations. Errors in MPA as a function of ground motion intensity: (a) floor displacements. and three “modes” and by NL-RHA for 1. Noted Error (%) 30 20 10 0 0 2 8 3 4 1 9 Error (%) 30 20 5 4 6 5 7 2 9 10 0 0 1 8 7 3 6 0. Locations of plastic hinges determined by MPA considering one.5 × El Centro ground motion (a) Floor Displacements 60 50 40 Error Envelope Error for Floor No.5 1 1. two. 2−"Modes" •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• • •• (c) 3−"Modes" • • • • • • • •• •• • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• • •• • • • • • • • • • (d) Nonlinear RHA •• •• • •• •• •• •• • •• •• •• •• •• •• •• • •• •• • •• •• •• •• • • • • • • •• •• •• Fig.5 3 Fig. 1−"Mode" • • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • •• • •• • • • • • •• •• • •• •• •• •• (b) MPA. 4. 4.

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

19 y by 3 2 1.5 0. and 3.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36.6 cm. 0.75.25 5 10 15 20 Roof Displacement (cm) 25 (c) "Mode" 3 Pushover Curve 12000 10000 u = 4.5 1 0.75 0. 4.75 0.85. 1. α = 0.75 Actual Idealized 20 40 60 Roof Displacement (cm) 80 (b) "Mode" 2 Pushover Curve 12000 10000 u = 9.25 1 2 4 6 Roof Displacement (cm) 8 10 Fig.0.25.13 y by Base Shear (kN) 8000 6000 1 1.85 Base Shear (kN) 8000 6000 4000 2000 0 0 0.3 cm.14 y by Base Shear (kN) 8000 6000 4000 1.0.50. V = 4952 kN. 0.5 0. α = 0. noted are peak values of roof displacement for 0.0 × El Centro ground motion 51 .5 2 3 4000 2000 0 0 Actual Idealized 0.5 0.25 0.16.9 cm. V = 7433 kN. V = 5210 kN. 2.85 0. 0.5 2 3 Actual Idealized 2000 0 0 0.85 0. “Modal” pushover curves with gravity loads included. α = 0.

5 1 1.5 Story Drift Ratio (%) 2 2. and hinge plastic rotations estimated by MPA including one. story drifts. gravity loads included.5 Fig.(a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th NL−RHA MPA 1 "Mode" 2 "Modes" 3 "Modes" (b) Story Drifts Floor Floor 5th 4th 3rd 2nd 1st 5th 4th 3rd 2nd 1st NL−RHA MPA 1 "Mode" 2 "Modes" 3 "Modes" Ground 0 0.5 1 1. Heightwise variation of floor displacements and story drift ratios from MPA and NL-RHA for 1.17. 4.18. Errors in floor displacements. 4. 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. and 52 .5 Displacement/Height (%) 2 Ground 0 0.5 × El Centro ground motion.

4.5 2 GM Multiplier 2.5 × El Centro ground motion. Locations of plastic hinges determined by MPA considering one. Noted 60 50 40 (a) Story Drifts Error Envelope Error for Story No.19. 1−"Mode" • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • •• •• •• • • • • • •• •• •• •• •• •• •• (b) MPA. Errors in MPA as a function of ground motion intensity: (a) floor displacements. 4. 3−"Modes" (d) Nonlinear RHA • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• •• •• • • • • • • • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • • • • • •• •• •• Fig.(a) MPA.5 3 0. two.20.5 1 1. gravity loads included 53 . and (b) story drifts. and three “modes” and by NL-RHA for 1.5 1 1. 2−"Modes" •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• •• •• (c) MPA.5 3 Fig.5 2 GM Multiplier 2. gravity loads included (a) Floor Displacements 60 50 40 Error Envelope Error for Floor No. 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.

854 0.00E+00 0.19E-03 1.8 1.938 1.8 -32.156 -0.2 9.213 1.530 1.88E-03 0.009 0.399 -0.351 -0.7 Peak values of story drift ratios (as % of story height) from MPA for 1.7 -2.00E+00 2 “Mode” 8.8 -32.603 -1.399 0.00E+00 0.4 1.00E-10 NL RHA 1.2 12.6 2.2 -2.5 -5.101 -0.665 0.Table 4.0 16.429 -1.7 1.88E-03 1.353 -23.00E+00 Combined (MPA) 1 “Mode” 8.530 1.00E+00 0.04E-10 3.8 Peak values of hinge plastic rotations (radians) from MPA for 1.315 0.921 1.7 16.128 -1.23E-03 3.00E+00 0.821 -1.998 0.5 “Mode” 3 -1.637 0.527 -0.8 9.199 16.2 9.6 Peak values of floor displacements (as % of building height = 37.19E-03 1.00E+00 0.23E-03 3.00E+00 0.3 -100.2 2.429 1.114 1.996 -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” 0.114 -1.55E-03 3.4 26.6 1.049 -0.515 -50.237 0.3 -100.9 31.2 16.830 -12.372 0.78E-03 1.9 -4.197 -0.176 0.55E-03 3.466 0.850 1.2 4.434 0.4 1.9 -6.35E-03 8.6 19.109 0.998 21.213 1.19E-03 1.4 -4.033 -0.5 1.319 1.11E-03 9.461 0.237 0.214 0.11E-03 9.213 1.2 0.00E+00 0.754 1.3 -22.35E-03 8.064 -10.2 1.04E-02 8.3 9.3 -22.00E+00 0.35E-10 3.728 1.00E+00 “Mode” 3 0.913 7.6 0.00E+00 Error (%) “Modal” Response 1 “Mode” -32.5 × El Centro ground motion.5 1.927 1.17E-03 9.00E-03 5.35E-03 8.107 1.2 1.00E+00 0.3 -22.105 0.687 0.5 × El Centro ground motion.860 1.5 21.19E-03 1.015 0.0 37.0 -30.125 -1.888 1.637 0.037 -0.00E-03 5.75E-03 0.266 -0.207 18.4 20.23E-03 0. gravity loads included Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th “Modal” Response “Mode” 1 “Mode” 2 Drift Ratio (%) Combined (MPA) 1 “Mode” 2 “Modes” 3 “Modes” Error (%) NL 1 2 3 RHA “Mode” “Modes” “Modes” 1.11E-03 9.908 1.130 0.330 1.854 0.2 4.744 1.270 -12.00E+00 0.098 20.0 37.754 0.066 -0.055 0.998 1.858 2.154 0.071 0.686 -7.516 0.5 54 .7 19.13E-03 5.831 0.00E+00 0.26E-03 3.594 -1.168 -0.311 0.4 -6.603 1.1 21.319 1.9 -3.933 1.436 1.507 1.11E-03 9.55E-03 3.7 4.00E-03 5.263 0.8 0.414 28.00E+00 3 “Modes” 8.2 -0.19E-04 5.2 -3.5 21.0 37.4 -4.14 m) from MPA for 1.673 Table 4.257 0.434 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.00E-03 5.1 1.23E-02 1.2 9.490 -11.037 0.0 -30.3 1.00E+00 “Mode” 2 0.0 -100.00E+00 0.5 10.2 4.068 0.908 -1.983 1.831 0.00E+00 0.5 3 “Modes” -32.0 -100.877 -46.821 1.3 -100.00E+00 0.836 -0.88E-03 0.813 1.996 0.667 0.102 1.371 -0.1 4.454 1.850 -1. gravity loads included Hinge Plastic Rotation (rad) Floor “Modal” Response “Mode” 1 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 8.057 0.102 1.5 × El Centro ground motion.3 13.00E+00 0.263 1.00E+00 0.00E+00 3.2 21.305 -0.434 0.783 1.23E-03 0.1 13.530 1.8 Table 4.190 -0.514 -1.478 0.8 28.00E+00 0.018 -0.953 15.066 -0.7 -2.071 -0.00E+00 0.00E+00 0.330 1.5 0.0 2 “Modes” -32.310 1.983 1.35E-03 8.8 0.822 1.6 2.

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

11 0. (b) ELF.1 through 5.0381 0.119 0.112 0. The errors in the FEMA and MPA estimates of seismic demands relative to the “exact” demands are presented in Fig.0466 0. the story drift demands in Fig.4 and Tables 5. and the displacements of upper floors are overestimated by 9-22%.1. pushover analyses are implemented for a target roof displacement of 52.3.0702 0.0913 0.11 0.11 0.367 (a) Uniform (b) ELF (c) SRSS Fig. 5.0981 0.4.062 0. 0.1.4a. The three FEMA force distributions are presented in Fig.11 0.2.3a and Table 5.2. with the errors being larger for the SRSS distribution. The pushover curves are given in Fig. plastic hinge rotation demands in Table 5. Figures 5.3.281 0.0 cm.177 0. Also included in these presentations are the MPA results considering three “modes” and the “exact” demands from nonlinear RHA. The MPA procedure is more accurate than all the FEMA force distributions. the value determined from RHA of the first-mode inelastic SDF system for 1. 5. and the locations of all plastic hinges in Fig.21 0. The first four floor displacements are within 5% of the “exact” value.11 0.11 0.4. 5.0446 0. and Table 5.3b and Table 5.0654 0.042 0.00719 0. Force distributions in FEMA-273: (a) “uniform”. 5.5 times the El Centro ground motion.3.126 0. Using each of these force distributions. both presented in Section 4. The “uniform” distribution overestimates all floor displacements by 17-28%. 5.1. the floor displacement demands in Fig.0197 0. 5.11 0. and (c) SRSS 56 .3a and 5.165 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.analyses.0896 0. 5. wherein the first two are obvious and the third is determined from response spectrum analysis of the building (Appendix C).

V = 8530 kN. α = 0. 5. V = 6897 kN.24 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized 10 20 30 40 Roof Displacement (cm) 50 60 Fig.18 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized 10 20 30 40 Roof Displacement (cm) (c) SRSS Distribution 50 60 12000 10000 u = 39.8 cm. gravity loads are included 57 . V = 7456 kN. α = 0. and (c) SRSS.2.14 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 = 38.3 cm. (b) ELF. Pushover curves using three force distributions in FEMA-273: (a) “uniform”. α = 0.(a) Uniform Distribution 12000 10000 u = 33 cm.

5 1 1.3 demonstrate that the hinge plastic rotations estimated by all three FEMA force distributions contain unacceptably large errors. This seems to be an inherent limitation of pushover analysis.” and NLRHA. 3 "Modes" FEMA Uniform ELF SRSS 0.4b.5 Displacement/Height (%) 2 Floor 5th 5th 4th 3rd 2nd 1st NL−RHA MPA. 7%.3. larger errors are noted in the upper and lower stories. Heightwise variation of floor displacements and story drift ratios estimated using FEMA-273 force distributions. MPA including three “modes. the MPA procedure is more accurate than all the FEMA force distributions. reaching 64%. at most. reaching 35%. In contrast. (The 100% error in a 6th floor hinge is ignored as it simply represents that the MPA estimated zero rotation. and overestimated by no more than 32%. whereas nonlinear RHA computed an insignificantly small value. but it is still inaccurate.5 1 1. 3 "Modes" FEMA Uniform ELF SRSS 0.) The pushover analysis procedures considered seem incapable of computing accurately local response quantities.5 Story Drift Ratio (%) 2 2. such as hinge plastic rotations.4c and Table 5. Modal pushover analysis procedure gives estimates better than all the FEMA force distributions. with story drifts under estimated by. reaching 31%. For the ELF distribution.2 demonstrate that the story drift demands are greatly underestimated by all the FEMA force distributions. errors are largest in the upper stories.3b and 5.5 Ground 0 Ground 0 Fig.Figures 5. (a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th (b) Story Drift Ratios Floor 4th 3rd 2nd 1st NL−RHA MPA. 5. with errors reaching 37% in this example. and Table 5. the errors are largest in the lower stories. For the uniform distribution. For the SRSS distribution. Figure 5. gravity loads included 58 .

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

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

78E-03 0.52E-04 1.93E-03 1.7 -15.17E-03 9.623 1.344 0.50E-03 0.3 -100.5 -3.199 27.3 23.064 1.306 1.0 -57.6 -4.84 1.23E-03 3.62E-03 0.515 FEMA Uniform ELF 27.355 0.00E+00 0.16E-03 0.09E-03 4.547 -27.367 1.53E-02 1.667 0.7 FEMA ELF SRSS -22.7 MPA -2.221 1.7 19.168 1.8 -35.530 1.00E+00 4.11E-03 9.975 1.858 1.353 1.399 1.75E-03 0.0 -63.7 21.00E-03 5.59E-03 5.8 44.5 -29.58E-04 6.0 -100.35E-10 3.487 0.8 7.5 22.015 0.3 Peak values of hinge plastic rotations (radians) from FEMA force distributions and MPA Hinge Plastic Rotation (rad) MPA FEMA ELF SRSS 4.00E-10 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 1.0 -100.35E-03 8.234 1.8 -2.4 16.414 1.0 -30.611 0.209 0.00E+00 8.9 -70.998 1.1.6 22.2 13.3 -11.5 -29.2 -4.4 16.00E+00 Error (%) FEMA Uniform ELF SRSS MPA 24.26E-03 3.5 61 .953 0.836 0.3 -14.0 Table 5.4 -12.2 6.6 17.2 16.938 1.1 163.1 10.45E-03 3.597 0.067 0.351 0.5 26.417 1.8 -100.178 1.686 0.462 1.399 1.708 0.00E+00 0.6 21.6 14.2 -100.736 0.927 1.2 4.6 -73.6 -25.3 31.007 1.562 1. gravity loads included Displacement /Height (%) FEMA ELF SRSS MPA 0.888 0.00E+00 0.9 SRSS -22.00E+00 0.490 0.5 10.Table 5.530 1.5 -27.0 -71.4 -23.399 -27.673 Error (%) NL RHA 1.2 -12.270 0.998 1.5 -33.154 1.672 1.00E+00 0.789 0.318 2.65E-03 7.214 1.809 0.913 0.2 10.6 -11.6 -6.854 0.560 1.5 15.4 -50.19E-03 1.0 -59.04E-02 8.3 29.2 9.3 -22.8 2.7 -28.9 0.4 MPA -2.3 -4.03E-03 5.8 17.9 16.4 0.2 Peak values of story drift ratios (as % of story height) from FEMA force distributions and MPA.00E+00 0.8 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 0.875 0.2 9.2 5.0 -100.466 0.19E-04 5.5 Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 2.061 1.992 1.524 0.6 4.263 0.23E-02 1.9 15.0 37.00E+00 2.128 1.083 1.7 26.13E-03 5.0 -100.089 1.314 1.6 -17.1 -3.7 -60.7 -29.9 11.1 0.5 -33.011 1.372 1.94E-03 2. gravity loads included Displacement /Height (%) FEMA ELF SRSS MPA 1.14 m) from FEMA force distributions and MPA.34E-03 2.7 7.839 0.098 1.9 -20.724 0.109 1.783 1.6 -41.2 16.1 -26.9 16.88E-03 1.566 1.341 1.2 19.436 Error (%) NL Uniform RHA 0.8 -100.195 0.310 1.51E-03 4.323 1. Peak values of floor displacements (as % of building height = 37.877 1.730 1.262 1.294 1.8 -32.8 16.9 28.10E-02 7.860 1.00E+00 NL RHA 1.4 21.8 -63.207 1.5 -32.7 Table 5.330 1.78E-03 1.335 2.830 2.4 -55.9 -77.55E-03 3.984 1.

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

5 1 1. Errors in floor displacements from three force distributions in FEMA-273 and from MPA including three “modes”.5 2 GM Multiplier 2.5 1 1.5 2 GM Multiplier 2.5 1 1.5 1 1. 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 0.6.5 2 GM Multiplier 2.5 3 0.5 3 0 0 0.5 2 GM Multiplier 2. 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. gravity loads included 63 .5 3 Fig.(a) FEMA: Uniform 60 50 40 60 50 40 3 (b) FEMA: ELF Error Envelope Error for Floor No.

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

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

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

MPA. story drifts. and nonlinear RHA.25 to 3.0. Based on results presented for El Centro ground motion scaled by factors varying from 0. all pushover analysis procedures do not seem to compute with acceptable accuracy local response quantities. The MPA procedure is more accurate than all the FEMA force distributions in estimating floor displacements. joint rotations. such as hinge plastic rotations. and identifying locations of most plastic hinges. 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. 67 . demonstrates that the FEMA force distributions greatly underestimate the story drift demands. plastic hinge rotations.(iv) At the roof displacement determined from Dn . and hinge plastic rotations. Comparing the earthquake-induced demands for the selected 9-story building determined by pushover analysis using three force distributions in FEMA-273. Thus the MPA procedure is accurate enough for practical application in building evaluation and design. 9. etc. 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. 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. 8.” 10. story drifts. the pushover analysis provides the peak value rno of any response quantity: floor displacements. 7. Pushover analyses seem to be inherently limited in computing accurately hinge plastic rotations. the errors in the MPA procedure are shown to be only weakly dependent on ground motion intensity. However. and lead to unacceptably large errors in the hinge plastic rotations. it fails to compute with acceptable accuracy plastic rotations of the hinges.

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

. Earthq. Adaptive spectra-based pushover procedure for seismic evaluation of structures. H. Conf. DRAIN-2DX user guide. (2000). (1998). Stanford. 132. J. Engrg. Earthq. Report No. Engrg. Y.. and Krawinkler. (1997). 15:417-434. 9th World Conf. and Seneviratna. H. A. Estimation of seismic drift demands for frame structures. ASCE. and Krawinkler. Earthquake Engineering Research Center. Prentice Hall: New Jersey. (1997). A. Federal Emergency Management Agency.K. Seismic demands for performance evaluation of steel moment resisting frame structures (SAC Task 5. Proc.M. Engrg. A.K. and Powell. Spectra. Struct. J. Stanford University. Dynamics of Structures: Theory and Applications to Earthquake Engineering. A. N2—a method for nonlinear seismic analysis of regular structures. E.P. Pushover analysis procedure in earthquake engineering.C.7 References Allahabadi. Struc. Calif. Proc. (1999). 20(4-6):452-464.3). and Kunnath. Engrg.. S. and D’Amore. P.H. (2001). and Reinhorn.. (1988). 6th U. Washington. NEHRP Guidelines for the Seismic Rehabilitation of Buildings. A. and Fischinger. Story drift demands for steel moment frame structures in different seismic regions. H. G. John A. Kunnath. M. Spectra. Krawinkler. Fajfar. J. on Earthq.K. (2000). Earthq. 29:1287-1305. (1988). Seismic performance and retrofit evaluation for reinforced concrete structures. Blume Earthquake Engineering Center. Earthq. Nat.M. Chopra. Washington. Method of reliability-based seismic design. Calif. 69 . B. Report No. S. Engng. Seattle.K. Berkeley. UCB/EERC-88/06. I: Equivalent nonlinear system. Tokyo-Kyoto. 16(2):367-392 Han. H.. University of California. Struct. S. Building Seismic Safety Council (1997).4. G. 5:111-116. (1999). Gupta. R. ASCE 123(1):3-10. Dyn.S. FEMA-273. and Wen. J. Gupta. Kim. Gupta..K... Pros and cons of a pushover analysis of seismic performance evaluation. Krawinkler. D. S. 123:256-265..D.W. Japan. (1998). and Gupta. Struct. Bracci.. Engrg.

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

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

4. by connecting a straight line between origin.6 × Vbny . 3. Replace i+1 with i and repeat Steps 3.6 . kn .5.3. Calculate the slope. O.6 × Vbny ( ) i urn . i 3. kn = 0. A. If desired. and B with straight-line segments to obtain the idealized bilinear curve. Calculate the post-yielding strain-hardening ratio.i 3.4 to 3.8.6.9.0. will be refined by an iterative procedure that seeks to equate areas under the actual and the idealized pushover curves. Abn .2. 3. This value.6 × Vbny .6 . and a point on the actual pushover curve with base shear equal to i 0. Calculate initial slope of the idealized bilinear curve. 72 .7. i i i αn = Vbno Vbny − 1 urno urny − 1 i 3. i +1 i i 3. Calculate the yield displacement.0. urn. i i i 3. From the pushover data.1. i i 3. 3. Calculate area under the bilinear curve OAB.9.9.8. urny = Vbny kn . other appropriate methods can be ( used. at base shear i equal to 0. i 3. iterations are necessary. Vbny . corresponding to the estimated yield i i i base shear. Calculate the error = 100 × Abn − Apn ( ) ( ) ( ) ) Apn . Let the point with base shear = Vbny and roof displacement = urny be denoted as A.2. Estimate the yield base shear. obtained by judgment. i 3. This step gives the secant stiffness at a base shear equal to 60% of the yield base shear.1. determine the roof displacement. Draw the curve OAB by connecting the three points O. If the error exceeds some pre-specified tolerance.4. Vbny . Calculate Vbny = Vbny × Apn Abn .4.

4. 7. 9.0.16). A. 4.1 Idealization of nth-“mode” pushover curve 73 . and pseudo-acceleration history. of the combined responses obtained in Step 8. A.4) and effective modal mass from M n = Ln Γn . Calculate histories of various responses using Eqs.2).2.15) and (3.12) and (3. Compute the Ln and Γ n from Eq.eps ur n. Combine the “modal” responses using Eqs. An (t ) . Repeat Steps 3 to 6 for as many modes as required for sufficient accuracy.2. 4. In general first two or three modes will suffice.10b and 4. (3.3b) with unit-mass and force-deformation relation of Fig. Dn (t ) .10a and 4. * 4. r o . 5. 4.6 ur n y ur n o ur n Fig. (3. Compute deformation history. of the nth“mode” inelastic SDF system (Fig.11b). Develop the Fsn Ln − Dn relation (Fig.1. 8. Vbn Idealization of Pushover Curve Vbno Vbny 0.11a).13).3. Scale the horizontal axis by Γnφrn to obtain Dno = urno Γ nφrn and Dny = urny Γ n φrn (Eqs.6 × Vbny A • Idealized αnkn B • 1 Actual k 1 • O n figA_1. (3. 4.4. Calculate peak values. Scale the vertical axis with * Mn to obtain * Fsno Ln = Vbno M n and * Fsny Ln = Vbny M n (Eqs. 6. A.

The force distributions. Following is step-by-step implementation of the procedure described in Section A. 1. 3.4. A.). First three mode shapes and frequencies of the selected building were computed and are shown in Fig. n 2. 3.eps Dny = ur n y / Γnφr n Dno = ur n o / Γnφr n Dn Fig. to the north-south component of the El Centro (1940) ground motion scaled up by a factor of 1. A. Target displacements used to generate these pushover curves are 63. respectively.3.2 EXAMPLE The UMRHA procedure is implemented to calculate the response of the 9-story building described in Section 3. for the first.5 cm (25 in.1. At this point. 25.urn ) pushover curve for the force distribution s* : n 2. Idealized bilinear curves for each of the three modes are included in Fig. and 12. and third mode. Area under the actual pushover curve. second. 3.7 cm (5 in. computed for the first three modes are shown in Fig. The base-shear – roof-displacement (Vbn .2 Properties of the nth-“mode” inelastic SDF system A.6 kN.3.4. generated using DRAIN-2DX. 3. The pushover curves for the first three modes.). B. 2. Ap1 = 360777 kN-cm.3. ur1o = 63.5 cm and Vb1o = 8729.1.2. 74 .Fsn / Ln Fsn / Ln − Dn Relationship Vbno / M* n Vbny / Mn * 1 α ω2 n n ω2 1 n figA_2. 3. is defined at the target roof displacement.).4 cm (10 in. s* .5.1. The anchor point.2. are shown in Fig. A. The following steps illustrate the procedure to develop the idealize curve for the first “mode”.

The Fs1 L1 − D1 relation for the first “mode” is developed as follows. ur1 y = Vb1 y k1 = 8006. i i i 3. k1 = 0. The procedure converged after nineteen cycles to give ur1 y = 36.23 cm. ur1.0. α1 = Vb1o Vb1 y − 1 ur1o ur1 y − 1 = ( ) ( ) (8729. i i 3.1.8 kN.4 210.9 kN.6 × Vb1 y = 4803.6 8006.8 22.0. and B with straight-line segments gives the idealized bilinear curve.i 3. The results for other modes are summarized in Table A. The initial slope of the idealized bilinear curve.01%. 3.4 kN. iterations are necessary.4. 3. i 3. 3. Also included are the modal damping ratios and the periods calculated from Eq. Area under the bilinear curve OAB. i +1 3.1. The post-yielding strain-hardening ratio. The next estimate of the yield shear is Vbny = 8006. The point A on the i i bilinear curve is defined by ur1 y = 38.4 ) − 1 (63. and M1 = 2736789 × 1.4 × (360777 365100 ) = 7911.9.5 38.86 cm at 0. i 3.6 × Vb1 y ( ) i ur1.6 = 22. is calculated as follows.6. Ab1 = 365100 kN-cm. Therefore.1.7.4.9.135.198%.18 kN/cm. A.09 ) − 1 = 0. The curve OAB obtained by connecting the three points O.6 kN.09 cm.86 = 210. The results of the iterative procedure are summarized in Table A.194. L1 = 2736789 kg. The yield displacement.18 = 38.3666 = 3740189 kg. * 4.1.3666.6 = 4803.9. Determined from the pushover database. This value exceeds the prespecified tolerance of 0.4. k1 .8. i i i 3.2. 4. i i 3. 75 .2. Γ1 = 1.13). Error = 100 × (365100 − 360777 ) 360777 = 1. Vb1 y = 7615. and α1 = 0.09 cm and Vb1 y = 8006.2.3.4 kN. The first estimate of the yield base shear Vb1 y = 8006. (4.5.

* M1 gives Fs1o L1 = 233. The peak values are computed and are summarized in Tables 4. The combined modal responses are presented in Fig. 9. 4.46 cm and D1y = 26. The results were generated for first three “modes’ and are included in Fig. A. Histories of roof displacement and top story drifts for the first “mode” are computed and presented in Fig.7.4.51 cm.8. Deformation and pseudo-acceleration histories of the inelastic SDF systems for the first “mode” with unit mass and force-deformation relation developed in Step 4 are plotted in Fig.4.2.1 and 4. 4. 4. 7.3. 5. Scaling the horizontal axis by Γ1φr1 gives D1o = 46. Scaling the vertical axis by Fs1 y L1 = 203. 4.40 (cm/sec2) and 76 .7. 6.62 (cm/sec2). The peak values are also plotted in Fig. 4.2. 8.7.

α = 0.eps 0 0 2 4 6 8 10 Roof Displacement (cm) 12 Fig.9 cm.eps 0 0 5 10 15 20 Roof Displacement (cm) (c) "Mode" 3 Pushover Curve 25 8000 u = 4.14 y by 6000 Base Shear (kN) 4000 2000 Actual Idealized figA_3c. A. α = 0.19 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized figA_3a.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36.3.13 6000 Base Shear (kN) 4000 2000 Actual Idealized figA_3b. V = 7616 kN. V y by = 4952 kN. V = 5210 kN.eps 10 20 30 40 50 Roof Displacement (cm) (b) "Mode" 2 Pushover Curve 60 8000 u = 9.2 cm. “Modal” pushover curves for the example building 77 .6 cm. α = 0.

A.” and third “mode” inelastic SDF systems.52 5 10 15 20 Time (sec) 25 30 0 • 1.(a) Deformation 80 n=1 3 (b) Pseuso−acceleration n=1 • 0.” second “mode.33 A1 (g) n=2 0 −80 80 −3 3 n=2 D2 (cm) 0 A1 (g) • 22. Histories of deformation and pseudo-acceleration due to 1. 78 .5 × El Centro ground motion for the first “mode.4.223 −80 80 n=3 −3 3 n=3 D3 (cm) 0 A3 (g) • 10.06 0 • 1.747 5 10 15 20 Time (sec) figa_4.2769 D1 (cm) 0 • 35.m −80 0 −3 0 25 30 Fig.

013 0.4 7714.5 3109.191 0.51 2 “Mode” 2 -920860 -0.2 4628.5 7624.12 3876.0 4577.193 0.30 37.107 0.02 21.56 19.186 0.25 36.192 0.4 4595.2 4571.18 210.6 × Vb1 y i ur1.170 0.81 21.74 (kN/cm) 210.65 1226.18 210.09 18.2671 1.3 7628.2406 167531. i 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 i Vb1 y i 0.1 4569.9 4573.063 0.188 0.048 0.59 36.017 0.0 4588.18 210.24 36.693 0.18 210.151 0. Results of iterative procedure to develop the idealized bilinear curve for the first “mode” inelastic SDF system Itr.8 7618.3666 3740189 203.39 36.77 21.9 7615.0 7619.75 21.9 (kN) 4803.6 7840.7 4580.75 21.4927 1.194 0.05 52.18 210.79 21.2.90 21.5309 488839.182 0.1 4574.5 4614.64 37. Properties of “modal” inelastic SDF systems Properties Ln (kg) Γn * M n (kg) “Mode” 1 2736789 1.176 0.83 21.23 22.56 47.76 21.18 210.193 0.4 7672.74 21.8 4747.9 4570. No.103 “Mode” 3 696400 0.8 7622.18 (cm) 38.1 1013.529 0.1 7616.18 210.18 210.309 0.4 7911.28 36.18 210.50 36.38 22.86 22.136 Fsny Ln (cm/sec2) Dny (cm) Fsno Ln (cm/sec ) Dno (cm) Tn (sec) 233.05 36.3 4603.29 36.70 36.037 0.135 0.082 0.3 7658.1.11 22.86 21.26 36.18 210.18 210.59 22.5 (cm) 22.948 ζ n (%) 79 .35 36.Table A.44 36.184 0.23 0.194 0.25 36.139 0.95 21.180 0.62 26.40 46.09 37.18 210.18 210.18 210.7 7639.79 0.2 7690.162 0.0 4704.029 0.8 4647.404 0.194 (kN-cm) 365100 364060 363276 362684 362235 361892 361631 361432 361279 361162 361073 361004 360951 360911 360880 360856 360838 360824 360813 Table A.18 210.2 4671.18 210.78 21.9 4570.0.198 0.193 0.022 0.3 7745.4 7647.237 0.190 0.8525 1.5 7633.46 2.6 4583.3 7786.194 0.18 210.910 0.85 36.18 210.6 i k1 i ur1 y i α1 i Ab1 Error (%) 1.85 0.010 (kN) 8006.32 36.

80 .

Determine the total response by combining the peak “modal” responses using the SRSS combination rule of Eq. rno .Appendix B Modal Pushover Analysis B. extract from the pushover database values of other desired responses.1 STEP-BY-STEP PROCEDURE A detailed step-by-step implementation of the modal pushover analysis (MPA) procedure is presented in this section and illustrated by an example in the next section. 4. 11.18).1 to the north-south component of the El Centro (1940) ground motion scaled up by a factor of 1. or from the inelastic response (or design) spectrum. Repeat Steps 3 to 7 for as many “modes” as required for sufficient accuracy. 12. 10. Dn . the first two or three “modes” will suffice. This is the same example as solved in Appendix A. subtract the yield hinge rotation to determine the plastic hinge rotation. 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. From the total hinge rotation. (3. (3. At urno . (4. B. (4. 6. Typically.33 cm. Compute the peak deformation.6b by solving Eq.1. 13. 10. Steps 1 to 4 of the MPA are the same as those for UMRHA presented in Appendix A. Following is step-by-step implementation of the procedure described in Section B. 81 .2 EXAMPLE The MPA procedure is implemented to calculate the response of the 9-story building described in Section 3.4.21).8).5.3b) with unit mass and force-deformation relation of Fig. Solving Eq. of the nth-“mode” inelastic SDF system (Fig. Calculate the peak roof displacement urno associated with the nth-“mode” inelastic SDF system from Eq.

where results for other ground motion intensities are also included. 9. 82 .11.3 for the floor displacements normalized by the building height and in Table 4. Peak roof displacement ur1o = Γ1 × φr1 × D1 = 1. 12. Also included in Table 4.4.33 = 48.18) are also included in Tables 4. At ur1o = 48.4 for the story drifts normalized by the story height. The values are summarized in Table 4.4.366 × 1 × 35.” and the results are included in Tables 4.1. 8. The total response computed by combining the peak “modal” responses using the SRSS combination rule of Eq. Results of Steps 5 and 6 are summarized in Table B.28 cm.3 and 4.5 are the plastic hinge rotations computed by subtracting the yield hinge rotation from the total hinge rotations. Steps 3 to 7 are repeated for first three “modes. values of floor displacements and story drifts are extracted.3 and 4. (3.28 cm.

117 5.03 0.901 8.332 48.0 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 1.8451 5.35 1.0 µ = Dn Dny urno = Γnφrn Dn (cm) 83 .37 57.268 0.275 1.577 16.50 35.395 0.766 7.007 36.535 14.82 1.457 12.25 20.52 “Mode” 3 1.75 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.52 0.5 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 2.139 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.748 63.71 1.200 0.913 22.504 18.367 1.0 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 3.18 27.28 46.551 2.59 0.27 0.450 4. Calculation of roof displacements urno from peak deformation of inelastic SDF systems Ground Motion Multiplier 0.229 8.691 0.37 1.4222 3.1.676 6.07 “Mode” 2 4.154 78.755 27.225 2.436 7.70 0.79 0.312 1.969 0.332 13.5 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.856 31.Table B.379 21.85 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 1.03 26.25 Quantity Dn (cm) “Mode” 1 6.513 0.023 0.38 22.35 0.185 11.252 9.690 10.33 1.735 3.126 13.678 0.13 2.467 14.36 1.73 24.660 14.06 1.05 0.267 5.755 0.184 0.

84

**Appendix C FEMA Force Distribution Calculations
**

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

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

Table C.1. FEMA273 “uniform” lateral force distribution

Floor, j

mj (10-3×kg)

s* = j

∑i mi

0.112 0.110 0.110 0.110 0.110 0.110 0.110 0.110 0.119

mj

1 2 3 4 5 6 7 8 9

503.5 494.7 494.7 494.7 494.7 494.7 494.7 494.7 534.1

C.2

EQUIVALENT LATERAL FORCE (ELF) DISTRIBUTION 85

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

Table C.2. FEMA273 equivalent lateral force (ELF) distribution

Floor j 1 2 3 4 5 6 7 8 9

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

-3 k

s* j

=

m jhk j

∑i mihik

0.007 0.020 0.038 0.062 0.091 0.126 0.165 0.210 0.281

C.3

SRSS DISTRIBUTION

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

N

∑ n (V jn )

2

to get column 8 of

0 381.1 832.7 95.0 176.2 Lateral Force fj (kN) (9) 203.3.070 0.1 -525.1 Story Shears V j2 (kN) (6) 1114.0 1231.7 694.047 0.5 -973.2 105.8 430.1 87.1 -438.7 101.177 0.367 87 .7 -46.9 -166.0 -5.9 832.4 250. 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) 1917.5 -320.9 366.7 1622.3 -646. the lateral forces are normalized by the base shear to obtain column 10 in Table C.8 374.9 1446.9 -153.2 200.3 222.1 1857.0 136.7 2065.6 -732.8 1381.098 0.5 320.9 1683.6 1233.9 880.7 (10) 0.7 355.6 319.2 148.7 Vj (kN) (8) 2268.042 0.7 525.7 1578.1 -967.8 -326.6 366.7 374.4 1842.2 285.9 446.2 97.045 0. Table C.3.3 -6.4 400.5 215.6 -359.4 1759.6 V j3 (kN) (7) 478.6 -352.1 -646.2 95.2 fj = ∑ fi i fj (2) (3) (4) 59.5 159.For convenience.3 240.0 980.2 277.065 0.0 1476.5 -350.7 234.0 354.090 0.6 286.

- Timoshenko Beam Theory
- Strength of Materials
- Simple Beam Theory
- FHWA Ground Anchor Circular
- FHWA Ground Anchor Circular
- Work Breakdown Structure
- Adam Saatler
- design of headed anchor bolts
- a fatigue primer for structural engineers
- emergency dam assessment
- soil structure interaction
- SOME ADVANCES IN THE ANALYSIS OF FLUID FLOWS
- Optimized Modeling and Design of Structures using SAP2000
- flood hydrology manual
- World Meteorological Organization Guide to Hydrological Practice Vol II
- guide to hydrological practices
- MayteRico_10
- 79127176 Hydraulics
- 15811820 Groundwater Resources of the World 2004
- 1411
- 363
- 1145
- Code of Practice for the Design and Installation of Anchors
- 2333205 Leri SAP2000 Yap Sistemlerinin Analizi Subat2007

Sign up to vote on this title

UsefulNot usefula research report modal pushover analysis

a research report modal pushover analysis

- Pushover analysis
- Pushover Analysis using ETABS and SAP2000
- EVALUATION OF PUSHOVER ANALYSIS PROCEDURES
- Pushover Analysis Final
- Pushover Ppt
- Performance Based Design Using Nonlinear Analysis
- Non Linear Analysis Pushover
- Nonlinear Pushover Analysis of Reinforced Concrete Structures
- 2013 Pushover Analysis Guide Final RC Building With Masonry Infill Walls EditedDraft January 2013
- SAP-Etabs-Course-360
- Pushover Analysis Using Etabs
- Performance Based Seismic Design
- Action Plan for Performance Based Seismic Design
- Non-Linear Static Push Over Analysis
- Thesis
- Performing Pushover Analysis in Etabs
- Etabs-step by Step
- ETABS Pushover Analysis
- ATC-40 Vol 2
- A comparison of single-run pushover analysis techniques for
- ATC-40
- pushover
- Pushover Analysis
- A Modal Pushover Analysis Procedure to Estimate
- Etabs Guide
- 15WCEE 1940 pushover analysis.pdf
- Analysis & Design Using ETABS
- Pushover
- Pushover
- Pushover Analysis Guide Final
- PUSHOVER Report Chopra[1]

Are you sure?

This action might not be possible to undo. Are you sure you want to continue?

We've moved you to where you read on your other device.

Get the full title to continue

Get the full title to continue reading from where you left off, or restart the preview.