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 .

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

iv .

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

vi .

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

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

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

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

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

(a) Idealized one-story structure. 2. 2001. This becomes evident if Eq.4) and w n is the natural vibration frequency. (2.1.5) For a given u g t . m depends on three system parameters: w n . sign u = .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. Tn = 2p w n is the natural vibration period.e.. z y . the deformation u t depends on three systems parameters: w n .3).3) ωn = k m ζ = c 2mω n f fs = s fy (2. and z is the damping ratio of the system vibrating within its linear elastic range (i. in addition to the form of the force-deformation relation. and R y (Chopra . bg 4 . The peak. z .1) is divided by m to obtain ~ u + 2zw n u + w 2 u y f s u. and (b) bilinear hysteretic forcedeformation relation For a given excitation u g t . u £ u y ).ug t n where bg bg b g bg (2. Section 7. and the ductility factor is u m= m uy (2. and u y . deformation is denoted by um .

2. u (cm) 8 Fig. f o w = 184 . 2001) and scaled up by a factor of 2. and f y = 39. m h = 3.04 40 Base Shear. Vby = 39.4 in Chopra. one-bay frame 50 uy = 1. 2. α = 0. The yield strength of the inelastic system. is f y / w = ( f o / w ) ÷ 8 = 0.26 kN (8.826 kips) for w = 169.66 m • EIb EIc EIc • • L = 7.2311 . 2.26 kN.5 sec and z = 5% subjected to the north-south component of the El Centro (1940) ground motion (Fig.2. based on Ry = 8 . Pushover curve for structure shown in Fig.376 cm. One-story. 2. with Tn = 0.1.3.2 kips) . this system and excitation.2 SYSTEM AND EXCITATION CONSIDERED Consider the one-story system in Fig.2 5 .2: the dimensions and flexural rigidity of the structural elements are noted. 6.32 m • Fig. For . Vb (kN) 30 20 10 0 0 2 4 6 Roof Displacement. 2.9 kN (38.

e. The yielding stiffness of each structural element is defined as 3% of its initial stiffness.017 rad . and q pm = 0. the peak value determined from RHA.1 kip-in. 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 . as apparent in Fig. and E = 2 × 108 kPa (29×103 ksi) gives yield moments of 21.4 ). I b = 3. Determined from pushover analysis at the exact peak deformation. (c) shows the joint rotation q bt g . the ductility factor m = 5. 2. the energy dissipated in 6 .g.4 shows the earthquake response of the system described in the preceding section determined by response history analysis (RHA).4e. A static pushover analysis of this one-story system leads to the force-displacement relationship shown in Fig. The peak values of the various response quantities are as follows: um = 7. (d) shows the rotation q p bt g of the plastic hinges at the bg bg beam ends. 2. (b) shows the lateral force f s t or base shear Vb t normalized relative to the bg weight.6 kip-in. 2. 2. pushover analysis cannot provide any cumulative measure of response.) for the beam and columns. It is organized in five parts: (a) shows the deformation u t . wherein the hysteretic force-deformation history of Fig.4f. Observe that the pushover curve matches the initial loading path of the hysteretic system.65 kN-m (191. The system is excited well beyond the yield deformation.134 × 107 mm 4 (75.) and 50.36 cm.3 in.3 RESPONSE HISTORY ANALYSIS Figure 2. Implementing this analysis with I c = 6.077 × 107 mm 4 (146 in. 2. the joint rotation and beam hinge rotation are identical to values q m and q pm determined from RHA. respectively.The yield moments in the beam and columns are defined as the bending moments due to the lateral force f y .0217 rad . 2. However.35.3. The resulting pushover curve is shown in Fig.18 kN-m (444.4 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 = 7.4edetermined from RHAis superimposed. q m = 0..36 cm .4 ). and (e) shows the force-deformation relation.

or the cumulative rotation at a plastic hinge.2 −0.4 −10 −5 0 u (cm) 5 10 Fig.04 0.017 • 0 (d) θp (rad) −0. (c) joint rotation. This represents an inherent limitation of pushover analyses.04 0 5 10 15 Time (sec) 20 25 30 0. (b) base shear.2 −0.2311 (a) Vb / w 0 f / w = −0. (d) plastic hinge rotation.04 0. (e) force-deformation relation.yielding during the ground motion.2 0 (e) 0 (f) −0.4. 2.4 0. and (f) pushover curve 7 . Response of one-story system to El Centro ground motion: (a) deformation.4 −10 −5 0 u (cm) 5 10 −0.04 0.36 −15 0.2 Vb / w 0.4 0.5 fy / w = 0. 15 u (cm) 0 • 7.5 0.2311 y (b) −0.0217 • θ (rad) 0 (c) −0.

8 .

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 . The right side of Eq. each element of the influence vector ι is equal to unity.1) can be interpreted as effective earthquake forces: p eff t = . and p eff t = Â peff.3. c. This force distribution can be expanded as a summation of modal inertia force distributions s n (Chopra.n t = Â . classical damping.mι ug t bg bg bg (3. and k are the mass.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 (3.1 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. m.12): mι = n =1 Â sn = N n =1 Â Γ nmφ n N (3. Elastic Multistory Buildings 3. and lateral stiffness matrices of the system.1) where u is the vector of N lateral floor displacements relative to the ground. (3.3) where φ n is the nth natural vibration mode of the structure.4) 9 .

10) gives qn t = Gn Dn t bg bg bg (3.7) By utilizing the orthogonality property of modes.n (t ) = N n =1 ∑ −sn u g (t ) N (3.n t is entirely in the nth-mode.9) and (3.11) and substituting in Eq.10) bg bg bg 10 (3. p eff . we will outline that the response of the MDF system to p eff . (3. (3. with no contribution from other modes. The solution qn t can readily be obtained by comparing Eq.u g t n Comparing Eqs.sn ug t bg bg bg bg (3.8) where the modal coordinate qn t is governed by 2 qn + 2z nw n qn + w n qn = .5) The contribution of the nth mode to s and to p eff t are: sn = Gn mφ n respectively.6) Next.9) to the equation of motion for the nth-mode elastic SDF system.12) .n t = .9) in which w n is the natural vibration frequency and z n is the damping ratio for the nth mode.Gn ug t (3. Then the floor displacements are u n t = φ n qn t bg bg bg bg (3. it can be demonstrated that none of the modes other than the nth mode contribute to the response. The equations governing the response of the system are mu + cu + ku = .sn ug t bg (3. (3. an SDF system with vibration propertiesnatural frequency w n and damping ratio z n of the nth-mode of the MDF system.8) gives the floor displacements u n t = Gnφ n Dn t bg (3.peff (t ) = n =1 ∑ peff . subjected to u g t : Dn + 2z nw n Dn + w 2 Dn = .

14) is the pseudo-acceleration response of the nth-mode SDF system (Chopra. Equations (3. 2001. ζn ¨ g(t ) u (a) Static Analysis of Structure (b) Dynamic Analysis of SDF System Fig.13) represent the response of the MDF system to p eff. these standard equations have been derived in an unconventional way. internal element forces.13) st where rn denotes the modal static response. the static value of r due to external forces sn . Section 13. However.can be expressed by st rn t = rn An t bg bg bg bg bg (3. In contrast to the classical 11 bg .15) and (3.12) and (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.Forces sn An(t ) st rn ωn.1. Therefore. 3. and An t = w 2 Dn t n bg (3.1). 3.16) This is the classical modal RHA procedure wherein Eq. etc.1.12) and (3. st The two analyses that lead to rn and An t are shown schematically in Fig.13) define the contribution of the nth-mode to the response. (3.9) is the standard modal equation governing qn t .15) rt = bg n =1 Â rn t = bg n =1 Â rnst An bt g N (3. Eqs. Conceptual explanation of modal response history analysis of elastic MDF systems Any response quantity r t story drifts. and Eqs.n t . (3.16) reflect combining the response contributions of all modes. (3.

3 MODAL PUSHOVER ANALYSIS (3. (3.(3. In such a response spectrum analysis (RSA). Sections 12. 2001. The peak modal responses are combined according to the Square-Root-of-Sum-ofSquares (SRSS) or the Complete Quadratic Combination (CQC) rules.1.8.derivation found in textbooks (e. This concept provides a rational basis for the modal pushover analysis procedure developed later.17) (Chopra. the peak nth-mode response as in Eq. and Tn = 2p w n is the natural vibration period of the nth-mode of the MDF system.g.3). which is valid for structures with well-separated natural frequencies such as multistory buildings with symmetric plan. The SRSS rule. 3.9) . this response value can be obtained by static analysis of the structure subjected to lateral forces distributed over the building height according to 12 . Chopra. z n of the pseudo-acceleration response (or design) spectrum for the nth-mode SDF system. 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 (3.19) 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. Alternatively.16).4 and 13.2 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. 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. (3. 2001. Section 13. we used the modal expansion of the spatial distribution of the effective earthquake forces.17) where An is the ordinate A Tn ..1).18) To develop a pushover analysis procedure consistent with RSA.

3. was designed by Brandow & Johnston Associates1 for the SAC2 Phase II Steel Project. This modal pushover analysis (MPA) for linearly elastic systems is equivalent to the well-known RSA procedure (Section 3. 3. 1301 S. SAC is a joint venture of three non-profit organizations: The Structural Engineers Association of California (SEAOC). in both directions. Richmond. Typical floor-to-floor heights (for analysis purposes measured from center-of-beam to center1 2 Brandow & Johnston Associates. 46th Street. CA 90017. Consulting Structural Engineers. urno .18) to obtain an estimate of the peak value ro of the total response. SAC Steel Project Technical Office.73 m (150 ft) by 45.21) 2 where Dn = An ω n .s* = mφ n n (3. denoted B1. Obviously Dn and An are available from the response (or design) spectrum. The columns are 345 MPa (50 ksi) steel wideflange sections. California. this building is 45.2). The levels of the 9-story building are numbered with respect to the ground level (see Fig. rno . (3. and 37. each determined by one pushover analysis. The bays are 9. A benchmark structure for the SAC project. 1660 W. the Applied Technology Council (ATC). shown in Fig. (3. region.4 COMPARATIVE EVALUATION OF ANALYSIS PROCEDURES 3. The building has a basement level. 13 .15 m (30 ft) on center.. Although not actually constructed.2).4. with the ninth level being the roof. Third St. with five bays each in the north-south (N-S) and east-west (E-W) directions. this structure meets seismic code and represents typical medium-rise buildings designed for the Los Angeles. The interior bays of the structure contain simple framing with composite floors.20) and the structure is pushed to the roof displacement. The peak modal responses. 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.12) is urno = Gnf rn Dn (3. and California Universities for Research in Earthquake Engineering (CUREE).73 m (150 ft) in plan. can be combined according to Eq.1 System and Excitation Considered The 9-story structure.19 m (122 ft) in elevation. CA 94804-4698.2. Los Angeles. 3. which from Eq. the peak value of the roof displacement due to the nth-mode.

2.07×106 kg (73.0 kips-sec2/ft).65×105 kg (66. This model is based on centerline dimensions of the bare frame in which beams and columns extend from centerline to centerline. The floor-to-floor height of the basement level is 3. ceiling/flooring.. third. The force distributions.00×106 kg (616 kips. fifth.0 kips-sec2/ft).3. 3. The strength. The building is modeled in DRAIN-2DX (Allahabadi and Powell. The seismic mass of the above ground levels of the entire structure is 9.85.2 kips-sec2/ft). Column splices. The column lines employ two-tier construction. floor slabs. The first three vibration modes and periods of the building for linearly elastic vibration are shown in Fig. the vibration periods are 2.65 m (12 ft) and for the first floor is 5. The floor system is composed of 248 MPa (36 ksi) steel wide-flange beams in acting composite action with the floor slab.sec2/ft). Each frame resists one half of the seismic mass associated with the entire structure. The seismic mass of the ground level is 9. 1988) using the M1 model developed by Krawinkler and Gupta (1998).01×106 kg (69.27. 3.20). and for the ninth level is 1. and 0. To ensure that this structure remains elastic. monolithic column pieces are connected every two levels beginning with the first level.49 m (18 ft). which are seismic (tension) splices to carry bending and uplift forces. 3.83 m (6 ft) above the center-line of the beam to column joint. and shear distortion of panel zones are neglected but large deformation (P∆) effects are included. for the second through eighth levels is 9.of-beam) are 3. mechanical/electrical. These force n distributions will be used in the pushover analysis to be presented later. and seventh levels at 1. roofing and a penthouse located on the roof. for the first level is 1.49 sec. The 9-story N-S MRF is depicted in Fig. s* (Eq. such as those described in Gupta and Krawinkler (1999) can be used. including the steel framing.7 kips-sec2/ft). respectively. we select a weak ground motion: the northsouth component of the El Centro (1940) ground motion scaled down by a factor of 4.89×105 kg (67. 0. 3. dimension.e. are located on the first.96 m (13 ft). Concrete foundation walls and surrounding soil are assumed to restrain the structure at the ground level from horizontal displacement. The column bases are modeled as pinned and secured to the ground (at the B-1 level). if desired more complex models. 14 . for the first three modes are shown in Fig. i. The seismic mass of the structure is due to various components of the structure.4. partitions. The simple model adopted here is sufficient for the objectives of this study.

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

8 −2.4. (3.13)]. and 3 n 3.7.27 sec Ground −1.5 Fig.12) and (3.4.5.05 s * 1 s * 2 s * 3 Fig.2 Response History Analysis The structural response due to individual vibration modes.67 −1.0272 −2.6.44 1.487 −1. 3.61 2. n = 1.12 0.51 0. Force distributions s* = mφn .5 0 0.1 3.5 −1 −0. First three natural-vibration periods and modes of the 9story building 3. respectively.13 −1. 3.796 0.03 −1.75 1. (c) shows the joint rotation q n t of an external joint at 16 bg bg bg .37 2.33 2. determined by RHA [Eqs. 3.9th 8th 7th 6th T = 0.94 2. Each figure is organized in four parts: (a) shows the roof displacement urn t .93 −1. is shown in Figs. 2.3. 2. n = 1.728 2.05 2.72 −2.49 sec 3 T = 0. and 3 .85 sec 2 Floor 5th 4th 3rd 2nd 1st T1 = 2.31 −0. 3.1 −2. and 3. (b) shows the base shear Vbn t normalized relative to the weight W of the building.39 3.5 Mode Shape Component 1 1.04 1.38 0.05 1.

the roof level. and ur 3o = 0. 3.23 cm. the exact response considering all modes. 17 . or all modes. The peak values of the various response quantities are noted in these figures. drifts in all stories. and three vibration modes. and (d) shows the Vbn . consistent with the increasing significance of the higher mode response among these three sets of response quantities. and rotations of external joints with moment connections are presented in Tables 3. Also included are the combined response due to one. The linear relationship between the base shear and roof displacement for each mode implies that the structure did not yield. 3. The peak values of floor displacements and story drifts determined by RHA. For a fixed number of modes included. 3. in particular.urn relation. The peak values of displacements of all floors.8. 3.2.8.1. and the percentage errors due to truncation of higher modal contributions. but three modes—perhaps even two modes—are sufficient. the results for the roof displacement and top-story drift are shown in Fig.9. and 3. Combining the modal response histories for all modes gives the total response [Eqs.15) and (3. including one. errors generally decrease as response contributions of more modes are included. As expected. with the first three modes provide essentially the exact response. the peak roof displacement due to each of three modes is ur1o = 9. larger in story drifts. which are listed in Tables 3. respectively. ur 2o = 2.3. and 3. The errors due to truncating the contributions of vibration modes beyond the third mode are negligible. It is apparent that the first mode alone is inadequate. respectively. where the second and third modal responses are a larger percentage of the top story drift compared to roof displacement. two. and even larger in joint rotations. especially in estimating the story drifts.422 cm.3. 3. (3.1. two. three.2.12 cm. This is illustrated in Fig. The same method was used to determine the peak values of many response quantities. are presented in Fig.16)]. errors are smallest in floor displacements.

89E-03 1.466 0.1 Building height is measured from the ground floor to the 9th floor.069 0.7 2.205 0.042 0.74E-03 1.183 0.015 -0.72E-03 3.295 -0.364 0.8 -1.0 -10.179 0.008 -0.0 3 Modes -5.0 -0.336 0.002 -0.0 -2.399 0.64E-03 3.229 0.130 0.91E-04 1.097 0.006 0.3 19.058 -0.66E-05 -3.09E-03 Error (%) 1 Mode -23.003 0.2 -57.73E-03 3.4 -53.226 0.6 -1.09E-03 1.325 0.173 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 0.2 1.9 -16.25 × 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 0.035 0.7 7.09E-03 1. 18 .4 -3.121 0.9 9.4 -6.8 1.011 0.74E-04 6.9 -23.192 0.0 -46.4 -1.11E-03 1.263 0.237 0.042 0.261 -0.303 0.054 0.6 0.3 -3.282 0.5 0.38E-03 2.9 3.9 -24.159 0.01E-04 -2.1 -2.060 0.63E-03 2.47E-03 1.4 -1.1 4.060 -0.266 0.56E-03 2.199 0.38E-04 2 Modes 2.6 -1.65E-03 2.475 0.2 9.062 0.1 Peak values of floor displacements (as % of building height = 37.1 -2.7 3.258 0.177 0.5 -1.2 -1.235 0.266 0.50E-03 4.265 0.043 0.229 0.264 Error (%) 1 Mode 2 Modes 3 Modes -23.2 -20.8 -15.453 0.265 RHA (all modes) 0.9 2.3 -8.1 -19.1 -0.203 0.202 0.4 -22.300 0.259 0.133 -0.4 -41.038 0.0 7.401 Error (%) 1 Mode 2 Modes 3 Modes -23.026 0.63E-03 2.310 0.03E-03 1.09E-03 2.253 0.350 0.9 8.2 -2.76E-03 1.333 0.124 0.152 0.227 0.307 0.00E-03 1.2 Peak values of story drift ratios (as % of story height) from RHA for 0.3 -33.156 0.097 0.245 0.3 -0.15E-03 4.282 0.024 -0.001 -0.125 0.226 0.311 0.2 -4.74E-03 1.125 0.7 -50.99E-03 Mode 3 3.6 11.90E-03 3.259 0.010 -0.29E-03 2.3 Table 3.50E-03 2.31E-03 Joint Rotation (rad) Combined (RHA) 1 Mode 2.011 0.152 0.00E-03 1.8 -10.25 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 3 Modal Response Mode 1 2.060 -0.124 0.14 m3) from RHA for 0.3 Peak values of joint rotations (radians) from RHA for 0.7 -19.235 0.266 0.6 4.88E-03 2.009 -0.1 Table 3.088 -0.03E-03 1.42E-04 1.245 0.6 0.4 0.097 0.413 RHA (all modes) 0.5 18.78E-04 -3.231 -0.94E-03 2.406 0.03E-03 3 Modes 2.055 0.08E-03 2.181 0.0 -2.062 -0.88E-03 2.2 0.260 0.33E-03 2.4 -0.8 -5.156 0.1 -11.99E-03 2.237 0.370 0.9 -15.197 0.227 0.7 2 Modes -3.44E-03 1.378 0.089 0.012 0.4 -10.069 0.5 -2.400 0.01E-04 3.Table 3.65E-03 RHA (all modes) 2.8 1.275 0.1 -2.9 -22.22E-03 2.260 0.8 -56.321 0.24E-03 2.9 1.6 -0.89E-03 1.13E-04 9.045 0.4 -7.1 3.6 9.012 -0.202 0.071 0.13E-03 2.090 0.1 -14.28E-04 1.3 -0.202 0.177 0.76E-03 1.44E-03 3.85E-03 3.26E-04 -5.117 0.74E-04 9.31E-03 Mode 2 -1.225 0.407 0.253 0.6 1.173 0.00E-03 2.032 -0.023 0.157 0.45E-03 3.11E-04 -5.022 0.03E-03 -6.080 0.7 4.14E-03 2.311 0.

1 −15 −7.5 0 u (cm) r1 7.5 15 −0.1 /W 0. Excitation is 0.1 −15 −7. and (e) pushover curve.0435 9.12 • 0 (a) r1 −15 0. (d) force-deformation history. (c) joint rotation.15 u (cm) 9.1 5 x 10 −3 θ (rad) 0.5.25 × El Centro ground motion 19 .1 0.12 0 (d) 0 (e) 15 Fig.00174 • 0 (c) r1 −5 0 5 10 15 Time (sec) 20 25 30 0. Response due to first mode: (a) roof displacement.0435 • 0 (b) V b1 −0. (b) base shear.5 0 u (cm) r1 7.1 0.5 9. 3.12 −0.0435 Vb1 / W 0.

1 −5 −2. Response due to second mode: (a) roof displacement.23 0.5 0 u (cm) r2 2.23 −2.5 u (cm) r2 0 • 2.0252 −2. (b) base shear.25 × El Centro ground motion 20 . and (e) pushover curve.1 −5 −2.5 5 Fig.1 0.1 Vb2 / W 0.6. Excitation is 0.5 0 u (cm) r2 2. 3.23 −5 (a) 0. (c) joint rotation.05 /W 0.5 5 −0.00244 • 0 (c) r2 −5 0 5 10 15 Time (sec) 20 25 30 0.05 5 x 10 −3 b2 θ (rad) 0. (d) force-deformation history.0252 (d) 0 0 (e) −0.0252 • 0 (b) V −0.

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

8) leads to the pushover curve shown in Fig.5.7d).2.6 present estimates of the combined response according to Eq.e.6.5e. 3.9 and Tables 3. This pushover curve is consistent with the relationship between the base shear and roof displacement determined by RHA (Fig.4. respectively.3 Modal Pushover Analysis Implementing MPA for the fundamental vibration mode. (3.7e and to the floor displacements.5.20) with n = 1 (Fig. pushing the structure using the force distribution of Eq. (3.6d and 3. or three vibration modes. the floor displacements are proportional to the mode shape φ 1 because the structure remains elastic. two.1 through 3.1. 3.3). 3. In practical application.12 cm. 3. implying that the modal truncation errors are small if at 22 . leads to the pushover curves shown in Figs. These values of the response quantities are identical to the peak response values determined from RHA (Tables 3. and 3. and 3. 3. (3.4. this value would be determined directly from the response (or design) spectrum. and it is apparent in the example considered that three modes provide most of the response (Fig. For a fixed number of modes included considered. 3.5. story drifts.23 cm .1. although both analyses led to identical peak values of the individual modal responses.3). Implementing pushover analysis for the second and third modes.18).5d).. using the force distribution of Eq.ur relations determined by RHA (Figs. and 3. and external joint rotations computed by pushover analysis are presented in Tables 3. which would provide the Dn value to be substituted in Eq.e.4.2. As for the first mode. the value determined by RHA (Fig.21). (3.20) with n = 2 and 3 up to roof displacements ur 2 o = 2.3. In RHA the errors arise only from truncating the responses due to higher modes. 3. i. i.. these pushover curves are consistent with the Vb .422 cm . 3. pushing the structure. considering one. respectively.4) to roof displacement ur1o = 9. (3. respectively. and the computed response values are identical to the peak response values determined from RHA (Tables 3.6.9). the errors in the MPA results are generally larger than in RHA (Fig.12).3). 3. As suggested by Eq. except for the algebraic sign associated with Γ1 and minor round-off errors.6e and 3. and external joint rotations in Tables 3. and 3. confirming that MPA gives the exact values of the individual modal responses. story drifts. 3.10 and Tables 3. Observe that the target roof displacement in each pushover analysis is identical to its exact value determined by RHA. 3.4. and ur 3o = 0. Figure 3. 3. and the errors in these estimates relative to the exact response from RHA considering all modes. and 3. 3. The floor displacements.

25 × El Centro ground motion 23 . Heightwise variation of floor displacements and story drifts from RHA for 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. 3. Response histories of roof displacement and top-story drift from RHA for 0. sec 25 30 r 5 10 15 20 Time.1 0.12 • r1 −15 15 0 r1 0 0 0.eps fig3_9b.4 Story Drift Ratio (%) 0.eps r 5 10 15 20 Time. sec 25 30 Fig.3 0.6 Fig.422 r2 0 • 1. (a) Roof Displacement 15 (b) Top Story Drift 3 Mode 1 ∆ (cm) u (cm) 9.eps Ground 0 0.5 Ground 0 0.48 −3 3 1.9.8.83 • r3 0 • 0.03 −3 3 Mode 3 ∆ (cm) u (cm) r3 −15 15 9.59 • Total (All Modes) 0 −15 0 ∆ (cm) u (cm) 0 −3 0 fig3_8.685 • −3 3 Mode 2 ∆ (cm) u (cm) • 2.least three modes are included. 3.2 0. Additional errors are introduced in pushover analysis due to the approximation inherent in modal combination rules.2 0.3 0.1 0.23 r2 −15 15 0 • 0.4 Displacement/Height (%) 0.5 0.

286 0.313 0.89E-03 2.33E-04 5.00E-03 2.90E-03 1.6 -16.73E-05 3.060 -0.006 -0.12E-03 1.179 0.9 -11.2 1.2 -11.09E-03 -1.124 0.96E-03 2.197 0.73E-03 3.001 0.3 -19.401 Error (%) 1 Mode 2 Modes 3 Modes -23.009 0.078 0.9 -15.4 -5.227 0.89E-03 -2.071 0.023 -0.76E-03 -1.276 0.22E-03 -2.63E-03 -2.321 0.4 -53.9 -14.4 1.9 -15.76E-03 1.062 -0.42E-04 -1.012 -0.44E-03 3.74E-03 1.253 RHA (All Modes) 0.231 -0.230 0.133 0.156 0.245 0.062 0.0 -10.2 -20.133 -0.4 -4.9 3 Modes -12.125 0.3 -14.15E-03 2.6 -17.2 Table 3.152 0.24E-03 -2.106 0.74E-04 -6.80E-04 3.012 0.181 0.38E-03 3 Modes 2.7 2.9 -18.47E-03 1.9 -13.0 -0.230 -0.90E-04 -9.032 0.04E-03 3.229 0.173 0.89E-03 2.156 0.2 -16.003 -0.1 -19.3 1.385 0.4 -11.011 0.336 RHA (all modes) 0.008 -0.9 -24.2 -4.022 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 -0.14 m) from MPA for 0.03E-03 6.03E-03 3.2 -12.6 Peak values of joint rotation (radians) from MPA for 0.2 -0.94E-03 2.7 -15.069 0.4 -9.173 0.4 Peak values of floor displacements (as % of building height = 37.203 0.300 0.03E-03 1.203 0.125 0.106 0.0 -16.267 -0.045 -0.72E-03 3.8 -23.058 -0.00E-03 2.8 -14.322 0.260 0.079 0.133 0.274 0.270 0.055 0.245 -0.4 -2.227 0.22E-04 2.237 0.3 -41.08E-03 2.261 -0.9 -16.00E-03 -1.336 0.97E-03 1.9 -15.9 -13.048 0.407 0.370 0.8 -22.78E-04 2 Modes 2.74E-03 -1.09E-03 Mode 2 1.260 -0.4 -22.331 0.282 -0.3 -9.31E-03 2.097 0.7 -21.0 -46.042 0.080 0.9 -14.267 0.31E-03 Joint Rotation (rad) Combined (MPA) 1 Mode 2.043 -0.9 -15.Table 3.05E-03 3.4 -19.2 -57.7 24 .097 0.00E-03 1.024 0.8 -56.44E-03 -1.6 -19.9 -13.038 -0.270 0.89E-03 1.229 0.267 0.4 -22.5 -16.177 0.466 0.4 -7.99E-03 Mode 3 -3.57E-03 1 Mode -23.8 -15.253 -0.296 -0.15E-03 1.40E-04 5.3 Table 3.042 0.203 0.5 Peak values of story drift ratios (as % of story height) from MPA for 0.203 0.235 -0.28E-03 2.89E-03 -1.011 -0.121 0.226 0.6 -15.4 -4.00E-03 2.253 0.235 0.3 -13.272 0.1 -11.00E-03 3.285 0.03E-03 -1.036 -0.31E-03 Error (%) RHA (all modes) 2.253 0.015 -0.328 0.259 0.7 -19.157 0.63E-03 2.088 -0.25 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Modal Response Mode 1 -2.026 0.002 0.260 0.0 -2.259 -0.8 -15.25 × 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 0.270 0.177 0.266 0.9 -13.069 0.1 -11.7 2 Modes -13.010 0.43E-04 -1.0 -18.9 -14.332 0.3 -2.179 0.152 0.7 -50.3 -33.4 -9.65E-03 2.4 -14.264 Error (%) Modal Response 1 Mode 2 Modes 3 Modes -23.374 0.060 0.237 -0.048 0.9 -12.157 0.090 0.92E-04 -1.1 -18.53E-04 -9.4 0.9 -8.282 0.203 0.09E-04 -3.310 0.1 -0.09E-03 1.38E-03 2.3 -12.117 0.

2 0.6 Fig.4 Story Drift Ratio (%) 0.1 0. 3.eps fig3_10b. shading indicates modal combination error 25 .10.5 Ground 0 0.3 0. Heightwise variation of floor displacements and story drifts from MPA for 0.eps Ground 0 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.25 × El Centro ground motion.4 Displacement/Height (%) 0.2 0.3 0.1 0.5 0.

26 .

mι ug t b (4.2) to the modal coordinates of the corresponding linear system. sign u b g g bg (4.3) . Therefore.4 4.1) With this generalization for inelastic systems.1) becomes mu + cu + f s u. the initial loading curve is idealized as bilinear. and the unloading and reloading curves differ from the initial loading branch. 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). Eq.2) The standard approach is to solve directly these coupled equations. but depend on the history of the displacements: f s = f s u. Both systems have the same mass and damping. (4. sign u = . Thus. Each structural element of this elastic system is defined to have the same stiffness as the initial stiffness of the structural element of the inelastic system. Although classical modal analysis (Section 3. the relations between lateral forces f s at the N floor levels and the lateral displacements u are not single valued. Expanding the displacements of the inelastic system in terms of the natural vibration modes of the corresponding linear system we get ut = bg n =1 Â φ nqn bt g 27 N (4. leading to the “exact” nonlinear response history analysis (RHA). (3.1) is not valid for inelastic systems.1 Inelastic Multistory Buildings RESPONSE HISTORY ANALYSIS For each structural element of a building. it is useful for later reference to transform Eq.

qr t = 0 for all 28 bg .2). (3.4) is rarely solved because it offers no particular advantage over Eq.Gn ug t Mn bg n = 1. Simultaneously solving these coupled equations and using Eq. 4. (3. (4.sn ug t bg b g bg bg (4. (4.2 UNCOUPLED MODAL RESPONSE HISTORY ANALYSIS bg Neglecting the coupling of the N equations in modal coordinates [Eq.… N (4. Unlike Eq. sign u = . (4. (3.4)] leads to the uncoupled modal response history analysis (UMRHA) procedure.5) This resisting force depends on all modal coordinates qn t . (4.” implying that other “modes” will also contribute to the solution. where φ n are now the modes of the corresponding linear system. (4.9) involves T Fsn = Fsn qn . (3.Substituting Eq.3) in Eq. (4. implying coupling of modal coordinates because of yielding of the structure.3) will. sign qn = φ n f s u n . sign u n b g b g bg (4. give the same results for u t as obtained directly from Eq.2). Equation (4.9) for linearly elastic systems. However.6) for inelastic systems will no longer be described by Eq.3).4) represents N equations in the modal coordinates qn . (4.n t given by Eq.6) The solution of Eq. these equations are coupled for inelastic systems. however.6b) are mu + cu + f s u.8) because qr t will generally be nonzero for “modes” other than the nth “mode. The spatial distribution s of the effective earthquake forces is expanded into the modal contributions sn according to Eq. The equations governing the response of the inelastic system to p eff.2). This approximate RHA procedure is the preliminary step in developing a modal pushover analysis procedure for inelastic systems. 2. (3. For linear systems.4) where the only term that differs from Eq. Eq.and classical n damping-orthogonality property of modes gives F qn + 2z nw n qn + sn = . and using the mass. premultiplying by φ T . (4. in principle.

(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

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. Fsn / Ln ¨ g(t ) u (a) Static Analysis of Structure (b) Dynamic Analysis of Inelastic SDF System Fig.Forces sn Unit mass An(t ) st rn ωn.575 5 10 15 20 Time (sec) 25 30 ur3 (cm) 0 • 5.3.02 −150 n=2 −150 50 n=2 14.0 ¥ El Centro ground motion: (a) exact solution by NL-RHA. Roof displacement due to p eff.4. and 3. and (b) approximate solution by UMRHA bg af bg 32 . where u g t = 3.89 50 r1 0 u (cm) 0 • 78.51 • ur2 (cm) 0 ur2 (cm) n=3 14.n t = . n = 1. 4. 4.s n u g t .9 • 0 −50 10 −50 10 n=3 ur3 (cm) 0 • 5. ζn. 2.

2.0 ¥ El Centro ground motion: (a) exact solution by NL-RHA.6) by UMRHA is compared with the “exact” solution by nonlinear RHA.2. (4. both for 3. and (b) approximate solution by UMRHA af bg bg 4.n t = . and 3.2… N ) according to Eq.(a) Nonlinear RHA 20 n=1 20 (b) UMRHA n=1 • 5. the errors in either response quantity are only a few percent.15) is strictly valid only for linearly elastic systems. and (3) the Fsn Ln . (4. 4.008 • 6. Such comparison for roof displacement and top-story drift is presented in Figs.817 5 10 15 20 Time (sec) 25 30 −10 0 −10 0 Fig.5.0 ¥ El Centro ground motion.1 Underlying Assumptions and Accuracy The approximate solution of Eq.33 0 −20 20 −20 20 n=2 ∆r2 (cm) 0 ∆r2 (cm) 7.5. (3.4) and (4.Dn relation is 33 bg bg . where u g t = 3.744 • 0 −20 10 n=3 −20 10 n=3 ∆r3 (cm) 0 • 5. but even for this very intense excitation. Top story drift due to p eff. respectively.s n u g t . The errors are slightly larger in drift than in displacement. n = 1. 4.855 ∆r1 (cm) 0 ∆r1 (cm) n=2 • 6.5)] is neglected. this intense excitation was chosen to ensure that the structure is excited well beyond its linear elastic limit. (2) the superposition of responses to p eff.4 and 4.956 5 10 15 20 Time (sec) 25 30 ∆r3 (cm) 0 • 5.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.

4. Therefore.20).8) before it can be solved? Because Eq. (3. such nonlinear static analysis provides the so-called pushover curve. this distribution seems to be a rational choiceeven for inelastic systemsto determine Fsn in Eq. Therefore we impose this constraint in developing the UMRHA procedure in this section and modal pushover analysis in the next section. the relationship between lateral forces f s and Dn in Eq.2. However.9).8) in UMRHA. The overall errors in the UMRHA procedure are documented in the examples presented in Section 4. When implemented by commercially available software. (4.9) should be determined by nonlinear static analysis of the structure as the structure undergoes displacements u = Dnφ n with increasing Dn . (4. within the linearly elastic range of the structure.12) for floor displacements. it can conduct a force-controlled nonlinear static analysis with an invariant distribution of lateral forces. (4.Dn relation? The two sets of forces and displacements are related as follows: V Fsn = bn Γn Dn = urn Γ nf rn 34 (4.4. is plotted against roof displacement urn .approximated by a bilinear curve to facilitate solution of Eq.Dn curve.20) to some bg pre-determined roof displacement. (4. A bilinear idealization of this pushover curve for the nth-“mode” is shown in Fig. 4. (4. which is different than the Fsn Ln .urn pushover curve to the Fsn Ln .6a. The structure is pushed using the force distribution of Eq.Dn relation to be determined in Eq. when specialized for linearly elastic systems it is identical to the RHA procedure of Section 3.1. Although several approximations are inherent in this UMRHA procedure. 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. (3. the only force distribution that produces displacements proportional to f n is given by Eq. How to convert this Vbn .2 Properties of the nth-“mode” Inelastic SDF System How is the Fsn Ln . the base shear is Vbny and roof displacement is urny . (3.8) governing Dn t is based on Eq. and the base shear Vbn .10) . At the yield point. Although most commercially available software cannot implement such displacement-controlled analysis.

13) This value of Tn . 4.V bn (a) Idealized Pushover Curve F sn /L n (b) F sn / L − D Relationship n n Idealized Vbny Actual 1 αnkn V bny /M * n 1 α ω2 n n kn 1 1 ωn 2 u u rny rn Dny = ur n y / Γn φ r n D n Fig. In contrast.6. Knowing Fsny Ln and Dny from n Eq. which may differ from the period of the corresponding linear system. (4. should be used in Eq. the initial slope of the pushover curve in Fig.12) implying that the initial slope of the curve in Fig. 2001.10 enables conversion of the pushover curve to the desired Fsn Ln − Dn relation shown in Fig. which is not a meaningful quantity.6b is w 2 . (4. 4.6b.11). 4. where the yield values of Fsn Ln and Dn are Fsny Ln = Vbny * Mn Dny = urny Γ nf rn (4. The two are related through Fsny Ln 2 = w n Dny (4. 4. Properties of the nth-“mode” inelastic SDF system from the pushover curve Equation 4.5).2. 35 . 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 (4.6a is 2 kn = ω n Ln .11) * in which M n = Ln Γn is the effective modal mass (Chopra.8). Section 13.

4. 2. (3.n t . of the total response r t obtained in Step 8. (3.12) and (3.6b. This value of the roof displacement is given by Eq. Compute the natural frequencies. develop the base-shear – roof-displacement ( Vbn − urn ) pushover curve for the force distribution s* [Eq. is now determined by 36 bg bg .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. 4. Dn (t ) . Calculate the peak value. (3.2. and pseudo-acceleration history. the first two or three modes will suffice. with the structure is pushed to the roof displacement urno . details are available in Appendix A: 1. which is summarized next as a sequence of steps.4. for linearly-elastic vibration of the building. Convert the idealized pushover curve to the Fsn Ln − Dn relation (Fig. 4. 8. 4. Calculate histories of various responses by Eqs. n 3. 5.6a).6b) by utilizing Eq. of the nth-“mode” inelastic SDF system (Fig. φn . Idealize the pushover curve as a bilinear curve (Fig. r o . For the nth-mode. Combine the “modal” responses using Eqs. (3.16) to determine the total response.11). Consider a nonlinear static analysis of the structure subjected to lateral forces distributed over the building height according to s* [Eq.21) where Dn .20)]. 4. Typically.13).15) and (3. the peak value of Dn t . 4. and modes. 7. Repeat Steps 2 to 6 for as many modes as required for sufficient accuracy.20)].3b) with force-deformation relation of Fig. ωn . 9. 6. Compute the deformation history. n (3. An t .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.

2. for elastic systems.2 and 3. details are available in Appendix B. or from the inelastic response (or design) spectrum. 6.3.12). to obtain an estimate of the peak value ro of the total response. each determined by one pushover analysis.6 and 7. alternatively. governed by Eq. of the nth-“mode” inelastic SDF system (Fig. and (2) it provides the exact modal response for elastic systems.1 Summary The peak inelastic response of a building to earthquake excitation can be estimated by the MPA procedure just developed. rno also represents the exact peak value of the nth-mode contribution rn t to response r t . 5.18). Steps 1 to 4 of the MPA are same as those for UMRHA. as described in Section 4. etc. bg bg bg bg Thus. plastic hinge rotations. Eq.2. Dn . are combined using an appropriate modal combination rule. 4. Calculate the peak roof displacement urno associated with the nth-“mode” inelastic SDF system from Eq. Compute the peak deformation. However. (3. it seems reasonable because it provides results for elastic buildings that are identical to the well-known RSA procedure described in Section 3.3b) with force-deformation relation of Fig. which is summarized next as a sequence of steps.6). Sections 7. 4. e.8). we will refer to rno as the peak “modal” response even in the case of inelastic systems.g. as mentioned earlier. (3.n t . 37 .3) because. story drifts.8). 2001. joint rotations. It provides results for elastic buildings that are identical to the well-known RSA procedure (Section 3. seems reasonable.solving Eq. 4. the pushover analysis provides an estimate of the peak value rno of any response rn t : floor displacements.3. the lateral force distribution used possesses two properties: (1) it appears to be the most rational choice among all invariant distribution of forces. although somewhat intuitive for inelastic buildings. (4.6b by solving Eq. The response value rno is an estimate of the peak value of the response of the inelastic system to p eff.4.21).. it can be determined from the inelastic response (or design) spectrum (Chopra. This pushover analysis. The peak “modal” responses rno . At this roof displacement. (4. This application of modal combination rules to inelastic systems obviously lacks a theoretical basis. As shown in Sections 3. (4.

the peak roof displacement due to each of the three “modes” is ur1o = 48.18).3 cm. Determine the total response by combining the peak “modal” responses using the SRSS combination rule of Eq. the “exact” results. in particular. From the total rotation of a plastic hinge.1). and ur 3o = 2. 4. 4.4 COMPARATIVE EVALUATION OF ANALYSIS PROCEDURES The response of the 9-story building described earlier is determined by the two approximate methods: UMRHA and MPA. (3. ur 2o = 11. the 38 . three “modes” included. At urno .5 ¥ the El Centro ground motion including the response contributions associated with three “modal” inelastic SDF systems.7 shows the individual “modal” responses. and the percentage errors in the approximate results. determined by the UMRHA procedure.2). and the “exact” response from nonlinear RHA for the roof displacement and top-story drift. 8.2. is presented next. rno . Repeat Steps 3 to 8 for as many “modes” as required for sufficient accuracy.1 Uncoupled Modal Response History Analysis The structural response to 1. the first two or three “modes” will suffice. with.53 cm.7 cm. although the trends are not systematic as when the system remained elastic (Section 3. say. To ensure that this structure responds well into the inelastic range the El Centro ground motion is scaled up by a factor varying from 1.4. 9.7. in contrast to modal analysis (Section 3. and the errors in the approximate results are shown in Fig. the combined response due to three “modes”. and three modes are compared with the “exact” values in Fig. 4. two. and three “modes”. and compared with the results of a rigorous nonlinear RHA using the DRAIN-2DX computer program. Typically.4.0 to 3. respectively. also included are the combined responses due to one. subtract the yield value of hinge rotation to determine the hinge plastic rotation. The peak values of displacements of all floors and drifts in all stories are presented in Tables 4.0.9. This deficiency also implies that. Observe that errors tend to decrease as response contributions of more “modes” are included. 4. Figure 4.8.1 and 4. This is to be expected. The peak values of response are as noted. extract from the pushover database values of other desired responses. The peak values of floor displacements and story drifts including one. the UMRHA procedure lacks a rigorous theory. two.

44 • "Mode" 2 r1 0 • 48.25. the degree to which the system deforms beyond its elastic limit.7.75. and total response from NL-RHA 39 .5. 2.5 × El Centro ground motion: individual “modal” responses and combined response from UMRHA.7 • −80 80 ∆r1 (cm) u (cm) "Mode" 1 −12 r2 0 ∆r2 (cm) 12 0 u (cm) −80 80 −12 r3 0 • 2. 1.88 u (cm) "Mode" 3 −80 80 0 −80 80 0 −80 0 • 44.1 and 3. 4.2). defined as the El Centro ground motion multiplied by 0. However. 0. and 3. in particular. For each excitation. sec 25 30 r −12 0 Fig.2) if the system yields significantly versus if the system remains within the elastic range (Tables 3.1 −12 12 ∆r (cm) u (cm) UMRHA 3 "Modes" 0 • 7. Next. 0.6 5 10 15 20 Time. the errors in story drifts are larger compared to floor displacements.response is much less accurate (Tables 4.5.0.1 and 4. we investigate how the errors in the UMRHA vary with the deformation demands imposed by the ground motion.5 (Tables 4. recall that the computed errors have been presented earlier for ground motion multipliers 0. Response histories of roof displacement and top-story drift due to 1.25 (Tables 3. 1.24 • NL−RHA r −12 12 ∆r (cm) u (cm) 0 5 10 15 20 Time.62 5.0.3 11.0.2) and 1. For this purpose the UMRHA and exact analyses were repeated for ground motions of varying intensity.2). 0. for a fixed number of “modes” included.53 ∆r3 (cm) 12 0 • 2. the errors in responses computed by UMRHA including three “modes” relative to the “exact” response were determined.85. (a) Roof Displacement (b) Top Story Drift 80 12 0 • 3.38 6.1 and 3.1 and 4. just as for elastic systems. sec 25 30 • 48.

4. and the error envelope for each case.10b). the pushover curves using force distributions s* [Eq.5 ¥ El Centro ground motion Figure 4. indicated by a ground motion multiplier. Heightwise variation of error in floor displacements and story drifts estimated by UMRHA including one.8. (3. 4. Two versions of the pushover curve are 40 . 4.9. with the peak displacement of each “modal” SDF system noted for each ground motion multiplier. For this purpose. Heightwise variation of floor displacements and story drift ratios from UMRHA and NL-RHA for 1.10a). or three “modes” for 1.20)] for the first three n modes of the system are shown in Fig. two. 4.5 Displacement/Height (%) 2 Ground 0 0.5 1 1. To interpret these results.11.5 Story Drift Ratio (%) 2 2.5 × El Centro ground motion (a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th (b) Story Drift Ratios Floor 4th 3rd 2nd 1st UMRHA 1 "Mode" 2 "Modes" 3 "Modes" Floor 5th 5th 4th 3rd 2nd 1st UMRHA 1 "Mode" 2 "Modes" 3 "Modes" Ground −60 −40 −20 0 Error (%) 20 40 60 Ground −60 −40 −20 0 Error (%) 20 40 60 Fig.5 1 1.10 summarizes the error in UMRHA as a function of ground motion intensity.(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. it will be useful to know the deformation of the system relative to its yield deformation.5 Fig. in each story drift (Fig. 4. Shown is the error in each floor displacement (Fig.

were noted but are not shown here. and. UMRHA lacks a rigorous theory and is based on several approximations. The structure is pushed using the force distribution of Eq. For more intense excitations. the peak values of displacements at all floors (Table 4. 3. The idealized curve for the first “mode” deviates most from the actual curve near the peak displacement corresponding to ground motion multiplier 1. respectively.0. the ductility factor for the first mode system is only 1. the values determined by RHA of the nth-mode inelastic SDF system (Fig. the first reason mentioned above seems to be the primary source for the errors. (4.53 cm. the pushover curve for each “mode” is idealized by a bilinear curve in solving Eq. 4. and plastic hinge rotations at the external beam end at each floor level (Table 4. 11. even though the system remains essentially elastic.75.75. Perhaps this explains why the errors are large at this excitation intensity. as mentioned in Section 3. 41 .1). Each of these three pushover analyses provides the pushover curve (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 1.2. The location of plastic hinges and their rotations. Additional errors are introduced in UMRHA of systems responding beyond the linearly elastic limit for at least two reasons. Second.5). First.6 and 4.4) to roof displacements urno = 48. 2.7 cm.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 0.4).11a).2.7). (2) increase rapidly as the ground motion multiplier increases to 1. and (4) are larger in story drift compared to floor displacements. (3.10) for each “modal” inelastic SDF system (Figs 4.01 (Fig.3 cm. and 3 (Fig. (3) maintain roughly similar values for more intense ground motions. determined from “exact” analyses. the errors in truncating the higher mode contributions are negligible.included: the actual curve and its idealized bilinear version. and 2.11).4.3). 4.4.20) with n = 1. drifts in all stories (Table 4. 4. The system remains elastic up to ground motion multiplier 0.0. 4. Figure 4. as mentioned in Section 4.5 ¥ the El Centro ground motion.

009 -0.133 1.763 -15.256 1.806 0.10. 4. 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 0.169 0.5 9.200 8.104 0.0 1.5 2 GM Multiplier 2.2 1.088 10.317 0.478 0.298 0.9 12.856 2.6 1.490 1.430 1.202 11.8 0.942 -0. Errors in UMRHA as a function of ground motion intensity: (a) floor displacements.298 -0.057 -0.366 -0.5 18.214 -0.8 1.5 28.201 -0.722 0.241 -1.7 14.811 1.071 0.248 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.844 -25.216 1.2 6.513 0.6 4.201 -1.283 1.495 1.0 7.333 0.426 -1.14 m) from UMRHA for 1.065 0.698 1.575 -41.3 25.4 4.072 -1.372 -1.235 -0.6 2.472 1.407 -10.0 -9.8 1.8 1.4 -1.5 × El Centro ground motion Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Modal Response “Mode” 1 “Mode” 2 Drift Ratio (%) Combined (UMRHA) 1 “Mode” 2 “Modes” “Mode” 3 -1.2 22.7 Table 4.018 0.806 -0.5 10.727 1.1 8.0 2.7 31.372 1.120 1.668 -23.379 1.410 -1.5 3 0 0 0.1 Peak values of floor displacements (as % of building height = 37.526 -0.540 0.5 3 Fig.5 -3.9 16.914 -0.1 1.5 2 GM Multiplier 2.3 6.291 0.370 -0.2 12.863 1.135 9.315 -0.820 -19.877 0.5 9.350 -0.4 1. and (b) story drifts Table 4.220 0.9 31.663 0.410 1.1 0.226 -0.5 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Displacement /Height (%) “Modal” Response Combined (UMRHA) “Mode” 1 “Mode” 2 “Mode” 3 1 “Mode” 2 “Modes” -0.121 -0.201 1.5 1 1.663 -0.(a) Floor Displacements 60 50 40 Error Envelope Error for Floor No.187 -0.8 14.376 -1.900 -10.2 1.126 0.130 0.5 28.938 -1.366 0.3 42 .3 8.371 0.2 Peak values of story drift ratios (as % of story height) from UMRHA for 1.484 0.010 0.071 -0.055 -0.513 -0.044 1.376 1.072 1.819 2.338 -1.772 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 1.9 5.138 1. Noted 60 50 40 (b) Story Drifts Error Envelope Error for Story No.5 1 1.241 1.220 -0.136 1.0 11.0 9.096 0.707 1.8 1.554 1.293 1.1 3.8 1.068 0.863 0.1 1.3 1.373 -0.982 9.2 4.820 -0.033 0.751 1.938 1.971 1.4 -7.676 0.079 0.256 1.9 12.214 0.473 -22.260 -15.003 0.3 0.256 -1.616 -0.945 -37.070 1.983 1.9 31.371 -0.852 1.914 2.490 -1.0 11.049 -0.338 1.003 -31.154 0.942 1.

even if three “modes” are included (Fig. in general. The second “mode” was necessary to identify hinges in the upper stories.5).14.4). significant improvement is achieved by including response contributions due to the second “mode”. 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. As shown in Figs. The errors are especially large in the hinge plastic rotations estimated by the MPA procedure.14.3 and 4. two. Furthermore. However. overestimated by up to 18% in the middle stories. although the error is recorded as 100% if MPA estimates zero rotation when the nonlinear RHA computes a non-zero value. 4. Fortuitously. and three “modes. especially in estimating the story drifts (Fig.3 through 4.Figures 4.5 present estimates of the combined response according to Eq. the results were not always accurate. and are within a few percent of the exact values for the upper stories.3 and 4.4). the additional errors due to the approximation inherent in modal combination rules tend to cancel out the errors due to the various approximation contained in the UMRHA. 4. Pushover analysis seems to be inherently limited in computing accurately hinge plastic rotations.9 and Tables 4. the errors in the modal pushover results are.13 and Tables 4.11c).12 and Tables 4. 4.3 and 4. MPA including three “modes” underestimates the displacements of the lower floors by up to 8% and overestimates the upper floor displacements by up to 14%.13 with Fig. (3. The drifts are underestimated by up to 13% in the lower stories. MPA failed to identify the plastic hinges at the column bases in Fig.13 and Table 4. For example.12 and Tables 4.18).2). for two or three modes included.13 and Tables 4. One “mode” pushover analysis was unable to identify the plastic hinges in the upper stories where higher mode contributions to response are known to be more significant. but was more successful when the excitation was more intense (results are not included). 43 . 4.3 and 4. and the errors in these estimates relative to the exact response from nonlinear RHA. 4.4 with Tables 4. significantly smaller than in UMRHA (compare Fig.” respectively.4. the hinges identified in beams at the 6th floor were at variance with the “exact” results. 4. Obviously. but the third “mode” contributions do not seem especially important (Fig.” and three “modes. this error is not especially significant because the hinge plastic rotation is very small. 4. however. were determined by four analyses: MPA considering one “mode. The first “mode” alone is inadequate. considering one.12 and 4.” two “modes. 4. The locations of plastic hinges shown in Fig. 4.1 and 4.” and nonlinear RHA. the third “mode” does not contribute because this SDF system remains elastic (Fig.

5. α = 0. 0.11.13 y by Base Shear (kN) 8000 6000 1 1.5 0. V = 5210 kN. 2.85 0. α = 0.75 Actual Idealized 20 40 60 Roof Displacement (cm) (b) "Mode" 2 Pushover Curve 80 12000 10000 u = 9.75. and 3.2 cm.25 5 10 15 20 Roof Displacement (cm) 25 (c) "Mode" 3 Pushover Curve 12000 10000 u = 4.75 0. 0. α = 0.85 0.25 1 2 4 6 Roof Displacement (cm) 8 10 Fig.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36.5 2 3 Actual Idealized 2000 0 0 0.9 cm.0.5 1 0.5. 1.5 0. V = 7616 kN. “Modal” pushover curves with peak roof displacements identified for 0.25 0.6 cm.75 0.14 y by Base Shear (kN) 8000 6000 4000 1. V = 4952 kN.85 Base Shear (kN) 8000 6000 4000 2000 0 0 0.19 y by 3 2 1. 4.0. 1.5 0.25.5 2 3 4000 2000 0 0 Actual Idealized 0.0 × El Centro ground motion 44 .

4. Shown is the error in each floor displacement (Fig. the MPA procedure is equivalent to the RSA procedure. each story drift (Fig. an observation with practical implications.4.5 Displacement/Height (%) 2 Ground 0 0.3 for elastic systems (or weak ground motions). implying that the modal combination errors contained in these procedures are acceptable. the errors in MPA were fortuitously smaller than in UMRHA (compare Figs. 4. implying excitations intense enough to cause significant yielding of the structure. as discussed in Sections 3. 4.75.5 Fig. the errors in MPA were larger for ground motion multipliers less than 0.15a). now standard in engineering practice.5 1 1. 4. However.12 Heightwise variation of floor displacements and story drift ratios from MPA and NL-RHA for 1.15) for ground multipliers larger than 1. In this case. identified in Section 3. 45 .15).2 and 3. UMRHA is essentially exact.4 also apply to MPA.15b).0. indicated by a ground motion multiplier.5 1 1. whereas MPA contains errors inherent in modal combination rules.10 and 4. While various sources of errors in UMRHA.4. As mentioned in Section 3. 4.5 × El Centro ground motion. implying excitations weak enough to limit the response in the elastic range of the structure.(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. shading indicates errors in MPA including three “modes” Figure 4.3. and the error envelope for each case. The errors are only weakly dependent on ground motion intensity (Fig. 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 Story Drift Ratio (%) 2 2.15 summarizes the error in MPA considering three “modes” as a function of ground motion intensity.

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

9 1.911 0.007 1.8 7.8 -29.3 11.8 “Mode” 3 -1.00E+00 0.516 0.02E-03 3.737 1.910 1.407 -27.116 1.351 -0.37E-03 1.399 0.00E+00 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th “Mode” 2 0.36E-03 6.5 1.02E-03 3.053 -1.00E+00 3.518 1.6 -9.8 1.003 -16.00E+00 0.5 -6.72E-03 7.304 1.209 1.5 2.895 1.220 1.0 47 .101 -0.222 0.233 1.18E-03 7.0 3 “Modes” -32.7 1.071 -0.55E-03 3.00E+00 0.844 -7.311 0.5 Table 4.8 -29.2 0.222 0.738 1.945 -49.1 46.168 -0.88E-03 0.9 7.9 -100.781 0.473 -15.575 -53.0 -50.372 0.72E-03 7.399 0.88E-03 0.36E-03 6.980 -0.00E+00 0.14 m) from MPA for 1.0 -100.50E-10 3.009 0.066 -0.0 15.8 17.00E+00 0.2 6.336 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.7 1.581 0.820 -7.071 0.980 0.305 -0.266 -0.9 -4.426 15.3 Peak values of floor displacements (as % of building height = 37.19E-10 3.8 -4.00E+00 0.116 1.00E+00 1 “Mode” 7.8 -6.6 13.37E-03 1.55E-03 3.250 0.02E-03 0.8 1.0 -100.527 -0.176 0.5 × El Centro ground motion Hinge Plastic Rotation (rad) “Modal” Response Combined (MPA) “Mode” 1 7.37E-03 1.2 -3.2 0.018 -0.2 -100.99E-03 6.498 1.00E+00 0.1 1.514 -1.233 1.37E-03 1.3 1.331 1.9 5.259 1.298 0.76E-03 4.752 1.652 1.7 -12.614 0.1 62.068 0.0 -50.053 1.5 2.2 -100.125 -1.1 62.088 12.5 × 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 0.4 -8.220 1.781 0.6 7.728 1.057 0.154 0.5 10.304 -1.00E+00 0.066 -0.049 -0.156 -0.202 8.22E-10 NL RHA 1.1 62.3 -3.581 0.2 11.190 -0.3 -3.982 13.37E-04 Error (%) “Modal” Response 1 “Mode” -32.00E+00 0.640 -1.260 -14.705 -1.942 6.00E+00 0.10E-02 9.015 0.338 1.694 1.667 1.00E+00 0.00E+00 0.00E+00 0.517 1.00E+00 0.8 -12.53E-03 7.00E+00 0.118 0.135 -7.2 -100.02E-03 0.763 -14.895 1.9 0.36E-03 6.007 1.298 -0.Table 4.5 Peak values of hinge plastic rotations (radians) from MPA for 1.197 -0.130 0.6 -44.5 10.015 0.756 0.033 -0.756 0.60E-03 2.668 -13.687 0.244 0.72E-03 7.1 46.5 × 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” 1.1 46.9 1.4 1.8 -29.429 0.8 -6.018 -0.2 1.478 0.1 18.879 1.6 -7.116 1.012 1.72E-03 7.00E+00 3 “Modes” 7.76E-03 4.2 -4.0 -5.00E+00 “Mode” 3 0.733 1.503 1.00E+00 0.36E-03 6.0 1.76E-03 4.6 -8.0 -100.00E+00 0.705 1.055 0.60E-04 7.05E-03 2.683 1.641 1.503 -1.00E+00 0.018 0.4 Peak values of story drift ratios (as % of story height) from MPA for 1.200 8.8 0.76E-03 4.1 -8.315 0.737 0.55E-03 3.1 13.371 -0.611 0.804 1.6 13.00E+00 0.0 2 “Modes” -32.5 2.6 1.9 2.640 1.00E+00 2 “Modes” 7.666 Table 4.3 13.745 1.88E-03 1.5 7.594 -1.435 0.900 -0.667 -1.105 0.26E-04 9.6 -44.414 1.9 -100.

7 cm and ur 3o = 2. but it is inadequate in estimating the story drifts (Fig.8 present estimates of the combined response according to Eq. 4.16 are unchanged.18 and Tables 4. and plastic hinge rotations at selected external beam end at each floor level (Table 4. and three “modes. MPA including two “modes” underestimates the displacements of lower floors by up to 6% and overestimates the upper floor displacements by up to 22%. The story drifts are underestimated by up to 7% in the lower stories. 4.11a). as are the roof displacements ur 2o = 11.6 and 4.6 and 4. Also presented at these roof displacements are the displacements at all floors (Table 4.20) with n = 1 to a target roof displacement.17 and Tables 4.5 × El Centro ground motion are presented below. The first “mode” alone provides adequate estimates of floor displacements. Figures 4. 4. The pushover curves for the second and third modes included in Fig. 4. and the structure is pushed using the force distribution of Eq.16a—which is slightly different than the one excluding gravity loads (Fig.” and the errors in these estimates relative to the exact response from nonlinear RHA.” however.4. and overestimated by up to 29% in the middle stories and by up to 13% in the upper stories. (3.6). drifts in all stories (Table 4.17 and Tables 4.6 and 4.3 Modal Pushover Analysis with Gravity Loads To evaluate the accuracy of the dynamic response of the system. the structure is pushed using the force distribution of Eq.5).7).8). 4. They are now included in the pushover analysis of the structure for the first “mode” only. the third “mode” contributions do not seem especially important (Fig.0 cm. (3.18 and Tables 4. Static analysis of the structure for gravity loads provides the initial state—forces and deformations—of the structure.” these effects will modify the combined modal response as well as the results of nonlinear RHA.18 and 48 . (3.3 .4. Results of such analyses for the selected building subject to 1. Obviously gravity load effects will influence the seismic demands due to the first “mode. The response contributions of the second and third modes are the same as before (Tables 4. the results presented so far did not include gravity load effects.7).18) considering one. Starting with its initial state under gravity loads.53 cm. 3. The errors are especially large in the plastic hinge rotations estimated by the MPA procedure even if three “modes” are included (Fig.4) to roof displacement ur1o = 52.7.20) with n = 1 (Fig.7). Significant improvement is achieved by including response contributions due to the second “mode. resulting in the pushover curve shown in Fig. two. 4.4.” but not the contributions of higher “modes. 4.6 through 4. As shown in Fig.17 and 4.

5 3 0.5 2 GM Multiplier 2. and (b) story drifts 49 . 1−"Mode" • • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • •• • •• • • • • • •• •• • •• •• •• •• (b) MPA.5 2 GM Multiplier 2.15. (a) MPA. Noted 60 50 40 (b) Story Drifts Error Envelope Error for Story No. and three “modes” and by NL-RHA for 1. Locations of plastic hinges determined by MPA considering one.8).5 3 Fig.14. 4. 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. 2−"Modes" •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• • •• (c) 3−"Modes" • • • • • • • •• •• • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• • •• • • • • • • • • • (d) Nonlinear RHA •• •• • •• •• •• •• • •• •• •• •• •• •• •• • •• •• • •• •• •• •• • • • • • • •• •• •• Fig. 4. two.5 × El Centro ground motion (a) Floor Displacements 60 50 40 Error Envelope Error for Floor No. Errors in MPA as a function of ground motion intensity: (a) floor displacements.Table 4.5 1 1.5 1 1. Most pushover analysis procedures do not seem to compute to acceptable accuracy plastic hinge rotations.

20 summarizes the error in MPA considering three “modes” as a function of ground motion intensity. this procedure is able to predict plastic hinge locations essentially consistent with nonlinear RHA. These errors are only slightly larger than those in Fig. Figure 4. One “mode” pushover analysis is unable to identify the plastic hinges in the upper stories where higher mode contributions to response are known to be more significant. 4. The second “mode” is necessary to identify hinges in the upper stories. 4. 4. excluding gravity load effects.19 were determined by four analyses: MPA considering one “mode. and the error envelope for each case. Shown is the error in each floor displacement (Fig. indicated by a ground motion multiplier. each story drift (Fig. MPA provides response values accurate enough for practical application in building retrofit or design. With two modes included in MPA.” two “modes”. and the errors are only weakly dependent on ground motion intensity. and nonlinear RHA. 50 .The locations of plastic hinges shown in Fig.20a). 4.15.20b).

0.19 y by 3 2 1. noted are peak values of roof displacement for 0.5 2 3 Actual Idealized 2000 0 0 0.75 0.25 5 10 15 20 Roof Displacement (cm) 25 (c) "Mode" 3 Pushover Curve 12000 10000 u = 4.25.5 1 0.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36. α = 0. 0.5 0.25 0.3 cm.9 cm. α = 0.5 0.0 × El Centro ground motion 51 . V = 4952 kN.6 cm. “Modal” pushover curves with gravity loads included. 0.85 Base Shear (kN) 8000 6000 4000 2000 0 0 0.75 Actual Idealized 20 40 60 Roof Displacement (cm) 80 (b) "Mode" 2 Pushover Curve 12000 10000 u = 9.85 0.14 y by Base Shear (kN) 8000 6000 4000 1. V = 7433 kN.85 0.75.13 y by Base Shear (kN) 8000 6000 1 1.5 2 3 4000 2000 0 0 Actual Idealized 0. 4. 0.0. α = 0.50.16.25 1 2 4 6 Roof Displacement (cm) 8 10 Fig.75 0.5 0.85. and 3. 1. 2. V = 5210 kN.

5 1 1. 4. Heightwise variation of floor displacements and story drift ratios from MPA and NL-RHA for 1.5 Story Drift Ratio (%) 2 2. two.(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 0.18. Errors in floor displacements. gravity loads included. story drifts.5 1 1.5 Displacement/Height (%) 2 Ground 0 0. and hinge plastic rotations estimated by MPA including one. shading indicates errors in MPA including three “modes” (a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th (b) Story Drift Ratios Floor 4th 3rd 2nd 1st MPA 1 "Mode" 2 "Modes" 3 "Modes" Floor 5th 5th 4th 3rd 2nd 1st MPA 1 "Mode" 2 "Modes" 3 "Modes" Ground −60 −40 −20 0 Error (%) 9th 8th 7th 6th 20 40 60 Ground −60 −40 −20 0 Error (%) 20 40 60 (c) Hinge Plastic Rotations Floor 5th 4th 3rd 2nd 1st MPA 1 "Mode" 2 "Modes" 3 "Modes" Ground −120 −80 −40 0 Error (%) 40 80 120 Fig.5 Fig.17.5 × El Centro ground motion. and 52 . 4.

two. and (b) story drifts.5 3 0. 1−"Mode" • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • •• •• •• • • • • • •• •• •• •• •• •• •• (b) MPA.(a) MPA. gravity loads included 53 .5 2 GM Multiplier 2. 3−"Modes" (d) Nonlinear RHA • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• •• •• • • • • • • • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • • • • • •• •• •• Fig. 4. and three “modes” and by NL-RHA for 1.5 1 1.5 3 Fig. Noted 60 50 40 (a) Story Drifts Error Envelope Error for Story 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. 4.5 2 GM Multiplier 2. Locations of plastic hinges determined by MPA considering one. gravity loads included (a) Floor Displacements 60 50 40 Error Envelope Error for Floor No.5 1 1.20.19. Errors in MPA as a function of ground motion intensity: (a) floor displacements. 2−"Modes" •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• •• •• (c) MPA.5 × El Centro ground motion.

071 -0.744 1.105 0.399 0.00E-03 5.19E-04 5.0 -30.6 2.5 21.5 10.55E-03 3.156 -0.821 -1.35E-03 8.7 Peak values of story drift ratios (as % of story height) from MPA for 1.603 1.7 4.2 -0.1 13.037 -0.353 -23.2 -2.101 -0.11E-03 9.049 -0.434 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” 0.199 16.4 26.00E+00 0.2 9.3 -100.5 54 .19E-03 1.9 -4.516 0.55E-03 3.068 0.434 0.88E-03 0.5 “Mode” 3 -1.850 1.102 1.88E-03 1.6 0.00E+00 3.530 1.55E-03 3.78E-03 1.854 0.7 19.2 9.311 0.26E-03 3.478 0.850 -1.310 1.6 2.933 1.998 1.0 -100.8 Peak values of hinge plastic rotations (radians) from MPA for 1.033 -0.998 0.2 4.071 0.7 16.2 0.037 0.860 1.754 0.686 -7.263 0.6 19.00E+00 0.858 2.7 -2.00E+00 0. gravity loads included Hinge Plastic Rotation (rad) Floor “Modal” Response “Mode” 1 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 8.00E-03 5.514 -1.04E-10 3.0 -100.854 0.130 0.19E-03 1.319 1.3 -100.2 16.637 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” 1.8 28.0 -30.414 28.371 -0.5 1.00E+00 Combined (MPA) 1 “Mode” 8.4 1.35E-03 8.1 21.19E-03 1.00E+00 0.490 -11.00E-03 5.00E+00 0.0 37.594 -1.3 -22.783 1.2 -3.754 1.00E+00 0.3 13.23E-03 3.527 -0.927 1.266 -0.114 1.399 -0.8 0.461 0.0 37.00E+00 0.466 0.064 -10.00E+00 0.23E-02 1.913 7.109 0.673 Table 4.04E-02 8.168 -0.11E-03 9.5 -5.429 1.213 1.066 -0.2 1.8 Table 4.830 -12.436 1.953 15.2 9.4 20.687 0.00E+00 0.00E+00 0.507 1.4 -4.11E-03 9.00E+00 0.237 0.3 -22.11E-03 9.5 × El Centro ground motion.098 20.35E-10 3.2 4.9 -6.237 0.7 -2.75E-03 0.00E+00 0.214 0.4 -6.213 1.515 -50.5 3 “Modes” -32.066 -0.270 -12.9 -3.6 Peak values of floor displacements (as % of building height = 37.822 1.821 1.213 1.0 37.6 1.530 1.637 0.263 1.125 -1.998 21.983 1.330 1.728 1.603 -1.429 -1.00E+00 0.4 -4.434 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.2 2.00E+00 2 “Mode” 8.107 1.813 1.3 -22.00E+00 0.018 -0.983 1.00E+00 0.35E-03 8.2 21.319 1.00E+00 3 “Modes” 8.305 -0.908 1.176 0.207 18.17E-03 9.5 0.3 -100.8 0.8 -32.330 1.00E+00 0.908 -1.102 1.665 0.19E-03 1.8 -32.2 1.00E+00 0.009 0.877 -46.057 0.3 1.23E-03 0.831 0.35E-03 8.921 1.2 4.00E+00 “Mode” 2 0.128 -1.315 0.2 12.114 -1.667 0.0 2 “Modes” -32.00E+00 Error (%) “Modal” Response 1 “Mode” -32.996 0.00E-10 NL RHA 1.00E+00 0.938 1.0 16.8 9.23E-03 3.197 -0.Table 4.88E-03 0.454 1.015 0.14 m) from MPA for 1.00E+00 0.4 1.888 1.13E-03 5.8 1.23E-03 0.351 -0.996 -0.530 1.7 1.831 0.836 -0.5 21.154 0.5 1.00E+00 “Mode” 3 0.1 4.372 0.00E-03 5.9 31.190 -0.3 9.257 0.5 × El Centro ground motion.1 1.5 × El Centro ground motion.055 0.

joint rotations. gravity load effects were included in all 55 . etc. k = 2 for T1 ≥ 2. and varies linearly in between. and the exponent k = 1 for fundament period T1 £ 0.5 sec .. 5.5. 2… N ). j 2. “Uniform” distribution: s* = m j (where the floor number j = 1. Specified in FEMA-273 are three distributions for lateral forces: 1.2 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. Comparison of Modal and FEMA Pushover Analyses 5. assumed to be linearly elastic. The floor displacements. Equivalent lateral force (ELF) distribution: s* = m j h k where h j is the height of the j j jth floor above the base. and 3.1 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. computed at the target displacement represent the earthquakeinduced demands on the structure. MPA considering three “modes. story drifts.5 sec . 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. a plot of base shear versus roof displacement. 1997] The pushover curve. SRSS distribution: s* is defined by the lateral forces back-calculated from the story shears determined by response spectrum analysis of the structure.” and nonlinear RHA. plastic hinge rotations.

0466 0.0197 0.119 0.0913 0.11 0.4.1.0702 0. Figures 5.1 demonstrate that the displacement demands are underestimated by the ELF and SRSS force distributions by 12 to 30 % at the lower six floors of the building. 5.11 0.21 0. The “uniform” distribution overestimates all floor displacements by 17-28%. Using each of these force distributions. 5.3. 5. and the displacements of upper floors are overestimated by 9-22%.11 0.177 0.0381 0.126 0.11 0.1 through 5. and the locations of all plastic hinges in Fig. and (c) SRSS 56 .281 0.0 cm. 5.2. the floor displacement demands in Fig. The pushover curves are given in Fig.3a and Table 5.3.042 0. 0. 5.11 0. wherein the first two are obvious and the third is determined from response spectrum analysis of the building (Appendix C).4. The MPA procedure is more accurate than all the FEMA force distributions. with the errors being larger for the SRSS distribution. the value determined from RHA of the first-mode inelastic SDF system for 1.112 0.1.0896 0. The errors in the FEMA and MPA estimates of seismic demands relative to the “exact” demands are presented in Fig.3.2. (b) ELF.0446 0.11 0. The three FEMA force distributions are presented in Fig.165 0.0654 0. both presented in Section 4. pushover analyses are implemented for a target roof displacement of 52.3b and Table 5. and Table 5. The first four floor displacements are within 5% of the “exact” value. 5. 5.062 0.1. Force distributions in FEMA-273: (a) “uniform”.4a.analyses.367 (a) Uniform (b) ELF (c) SRSS Fig.00719 0. Also included in these presentations are the MPA results considering three “modes” and the “exact” demands from nonlinear RHA.4 and Tables 5.0981 0. the story drift demands in Fig.3a and 5.11 0. plastic hinge rotation demands in Table 5.5 times the El Centro ground motion.

α = 0. gravity loads are included 57 . α = 0. and (c) SRSS. V = 6897 kN.3 cm.8 cm.24 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized 10 20 30 40 Roof Displacement (cm) 50 60 Fig.18 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized 10 20 30 40 Roof Displacement (cm) (c) SRSS Distribution 50 60 12000 10000 u = 39.2. Pushover curves using three force distributions in FEMA-273: (a) “uniform”.(a) Uniform Distribution 12000 10000 u = 33 cm. (b) ELF. V = 8530 kN. α = 0. 5.14 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized 10 20 30 40 Roof Displacement (cm) (b) ELF Distribution 50 60 12000 10000 u = 38. V = 7456 kN.

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

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.4. 5. and hinge plastic rotations estimated using FEMA-273 force distributions and MPA (including three “modes”).(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.” nonlinear RHA (“exact”). the SRSS distribution fails to identify yielding of beams in the middle floors. the “uniform” distribution fails to identify yielding of the beams above the fourth floor. and the three FEMA analyses.5 were determined by five analyses: MPA considering the three “modes. gravity loads included The structural engineering profession is now comparing these hinge plastic rotations against rotation limits established in FEMA-273 to judge structural component performance. The locations of the plastic hinges are not identified correctly by the FEMA force distributions. story drifts. they are not good indicators of damage. and the ELF distribution fails 59 . Based on the results presented here. 5. Errors in floor displacements. The locations of plastic hinges shown in Fig.

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

330 1.0 -100.318 2.530 1.4 -55.530 1.9 -20.673 Error (%) NL RHA 1.35E-03 8.7 -28.3 -14.00E+00 2.52E-04 1.45E-03 3.2 Peak values of story drift ratios (as % of story height) from FEMA force distributions and MPA.6 -4.061 1.4 -23.547 -27.0 -100.0 -100.3 -22.736 0.234 1.3 29.515 FEMA Uniform ELF 27.9 -77.724 0.9 -70.809 0.78E-03 1.877 1.5 -33.5 Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 2.399 1.8 17.2 -12.667 0.927 1.0 -71.8 -100.00E+00 4.221 1.953 0.62E-03 0.84 1.417 1.5 26.938 1.414 1.888 0.199 27.9 SRSS -22.5 15.5 -29.8 -100.7 19.2 -100.011 1.34E-03 2.341 1.11E-03 9.399 -27.367 1.3 -11.708 0.015 0.6 -11.128 1.8 44.2 5.490 0.4 MPA -2.5 -33.344 0.335 2.3 -100.5 -29.7 7.263 0.2 19.207 1.4 0.6 -6.836 0.214 1.13E-03 5.6 21.466 0.310 1.562 1.4 21.00E+00 0.00E+00 0.858 1.7 21.098 1.154 1.2 16.7 26.00E+00 0.6 -25.623 1.19E-03 1. gravity loads included Displacement /Height (%) FEMA ELF SRSS MPA 1.53E-02 1.854 0.88E-03 1.5 61 .789 0.611 0.9 16.78E-03 0.03E-03 5.65E-03 7.09E-03 4. Peak values of floor displacements (as % of building height = 37.860 1.00E+00 0.8 16.8 -63.00E-10 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 1.50E-03 0.2 10.067 0.372 1.462 1.59E-03 5.51E-03 4.8 -2.306 1.2 9.00E-03 5.1 -3.7 MPA -2.524 0.6 -17.9 15.1.23E-02 1.10E-02 7.23E-03 3.5 -27.9 16.353 1.8 -32.984 1.9 28.00E+00 0.6 -41.686 0.783 1.6 -73.487 0.8 2.270 0.314 1.4 -12.975 1.3 -4.00E+00 Error (%) FEMA Uniform ELF SRSS MPA 24.351 0.560 1.7 Table 5.2 4.8 -35.6 22.294 1.2 6.0 -57.566 1.355 0.064 1.839 0.2 9.5 10.5 -32.4 16.2 16.1 163.262 1.830 2.0 -30.00E+00 8.0 -100.Table 5.6 4.0 -63.3 Peak values of hinge plastic rotations (radians) from FEMA force distributions and MPA Hinge Plastic Rotation (rad) MPA FEMA ELF SRSS 4.0 -59.0 37.7 FEMA ELF SRSS -22.597 0.913 0.7 -15.5 -3.998 1.209 0.195 0.19E-04 5.58E-04 6.730 1.6 14.672 1.399 1.93E-03 1.2 -4.998 1.7 -29.3 31.00E+00 NL RHA 1.75E-03 0.4 -50.178 1.4 16.0 Table 5.14 m) from FEMA force distributions and MPA.17E-03 9.2 13.94E-03 2.8 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 0.9 11.083 1.1 10.9 0.875 0.007 1.8 7.04E-02 8.323 1.55E-03 3.7 -60.3 23.436 Error (%) NL Uniform RHA 0.992 1.1 -26.089 1.1 0. gravity loads included Displacement /Height (%) FEMA ELF SRSS MPA 0.16E-03 0.26E-03 3.35E-10 3.5 22.6 17.00E+00 0.109 1.168 1.

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

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

5 1 1.7.5 1 1.5 1 1. 5.5 2 GM Multiplier 2.5 2 GM Multiplier 2. gravity loads included 64 .(a) FEMA: Uniform 100 60 50 9 8 (b) FEMA: ELF Error Envelope Error for Story No.5 3 Fig.5 3 (c) FEMA: SRSS 60 50 40 60 50 40 3 2 1 (d) MPA (3 "Modes") Error (%) 30 20 Error (%) 30 5 4 8 9 20 9 6 7 7 10 0 0 4 10 2 3 1 8 5 6 0.5 2 GM Multiplier 2. 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 3 0 0 0.5 3 0 0 0.5 1 1. Error in story drifts from three force distributions in FEMA-273 and from MPA including three “modes”.5 2 GM Multiplier 2.

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

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

Thus the MPA procedure is accurate enough for practical application in building evaluation and design. However. 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.0.(iv) At the roof displacement determined from Dn . and hinge plastic rotations. Based on results presented for El Centro ground motion scaled by factors varying from 0. Comparing the peak inelastic response of a 9-story SAC building determined by the approximate MPA procedure—including only the first two or three rno terms—with rigorous nonlinear RHA demonstrates that the approximate procedure while providing good estimates of floor displacements and story drifts. such as hinge plastic rotations. the pushover analysis provides the peak value rno of any response quantity: floor displacements. Comparing the earthquake-induced demands for the selected 9-story building determined by pushover analysis using three force distributions in FEMA-273. 9.25 to 3. 67 . the errors in the MPA procedure are shown to be only weakly dependent on ground motion intensity. joint rotations. The MPA procedure is more accurate than all the FEMA force distributions in estimating floor displacements. 8. and identifying locations of most plastic hinges. Pushover analyses seem to be inherently limited in computing accurately hinge plastic rotations.” 10. The initial state—forces and deformations—of the structure can be considered in the MPA procedure by including these gravity load effects in pushover analysis of the structure only for the first “mode. and lead to unacceptably large errors in the hinge plastic rotations. plastic hinge rotations. etc. all pushover analysis procedures do not seem to compute with acceptable accuracy local response quantities. and nonlinear RHA. MPA. 7. story drifts. demonstrates that the FEMA force distributions greatly underestimate the story drift demands. it fails to compute with acceptable accuracy plastic rotations of the hinges. story drifts.

should obviously be evaluated for a wide range of buildings and ground motion ensembles. with the excitation scaled to cover a wide range of ground motion intensities and building response. such as life safety and collapse prevention. The present trend in the structural engineering profession of comparing computed hinge plastic rotations against rotation limits established in FEMA-273 to judge structural performance does not seem prudent. This report has focused on development of the MPA procedure and its initial evaluation in estimating the seismic demands on a building imposed by a selected ground motion. Instead.11. 68 . This new method for estimating seismic demands at low performance levels. While pushover estimates for floor displacements are even more accurate. they are not good indicators of damage. structural performance evaluation should be based on story drifts that are known to be closely related to damage and can be estimated to a higher degree of accuracy by pushover analyses.

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

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

develop the base-shear – roof-displacement (Vbn .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. Since the target roof displacement may not be known at the start of the procedure.20): s* = mφn n n 2. 71 . and modes. e.. trapezoidal rule. Apply force distribution of Step 2. respectively. e. Define the anchor point. ω n . Define the force distribution s* from Eq. Apn . Idealize the pushover curve as a bilinear curve (Fig. using any numerical integration method. A. B. 1977). iterations may be necessary. 2. Compute natural frequencies. The structure should be pushed just beyond the target (or expected) roof displacement in the selected mode. 1988). (3.1.urn ) pushover curve for the force distribution s* : n 2.1 incrementally and record the base shears and associated roof displacements..1) using the FEMA-273 procedure (Building Seismic Safety Council.2.1. φn .g. This step can be conveniently implemented in any commercially available software. DRAIN2DX (Allahabadi and Powell. 1.g. 3. 3. Let the roof displacement and base shear at the anchor point be urno and Vbno . For the nth-“mode”. Calculate the area under the actual pushover curve. of the bilinear curve at the target roof displacement.Appendix A Uncoupled Modal Response History Analysis A. for linear-elastic vibration of the building. 3.2.

Calculate the slope. kn = 0. Calculate initial slope of the idealized bilinear curve. 3. Calculate the yield displacement.6 × Vbny .6 × Vbny . by connecting a straight line between origin.8. 3. Calculate Vbny = Vbny × Apn Abn . i i i αn = Vbno Vbny − 1 urno urny − 1 i 3. 3.2.1. at base shear i equal to 0.7. Vbny . obtained by judgment. O. If desired. i +1 i i 3. Calculate area under the bilinear curve OAB. Abn . urn.4. iterations are necessary. i i i 3.8. i 3.5. i 3. 72 . A.2.i 3. Calculate the error = 100 × Abn − Apn ( ) ( ) ( ) ) Apn .4.6 . will be refined by an iterative procedure that seeks to equate areas under the actual and the idealized pushover curves. i i 3. From the pushover data.6 . If the error exceeds some pre-specified tolerance.1.0. corresponding to the estimated yield i i i base shear. This step gives the secant stiffness at a base shear equal to 60% of the yield base shear.4. Let the point with base shear = Vbny and roof displacement = urny be denoted as A. Calculate the post-yielding strain-hardening ratio. Replace i+1 with i and repeat Steps 3. kn . Draw the curve OAB by connecting the three points O. other appropriate methods can be ( used. and a point on the actual pushover curve with base shear equal to i 0. i 3. Vbny .0.9. Estimate the yield base shear.9. urny = Vbny kn .4 to 3. and B with straight-line segments to obtain the idealized bilinear curve.6 × Vbny ( ) i urn .6.3. This value.9. determine the roof displacement.

Dn (t ) . Calculate histories of various responses using Eqs. 4.6 ur n y ur n o ur n Fig.3b) with unit-mass and force-deformation relation of Fig. 6. of the nth“mode” inelastic SDF system (Fig. (3. A.1. 8. 4. A. (3.10a and 4. (3. Combine the “modal” responses using Eqs.13). Compute deformation history.16).2). 7.2. In general first two or three modes will suffice. 5. Scale the horizontal axis by Γnφrn to obtain Dno = urno Γ nφrn and Dny = urny Γ n φrn (Eqs.4) and effective modal mass from M n = Ln Γn . of the combined responses obtained in Step 8.15) and (3.1 Idealization of nth-“mode” pushover curve 73 .12) and (3.10b and 4. Vbn Idealization of Pushover Curve Vbno Vbny 0. Repeat Steps 3 to 6 for as many modes as required for sufficient accuracy. and pseudo-acceleration history.eps ur n.11a). 9. Compute the Ln and Γ n from Eq. r o . * 4. 4.3.0. Calculate peak values.6 × Vbny A • Idealized αnkn B • 1 Actual k 1 • O n figA_1. 4.11b). Scale the vertical axis with * Mn to obtain * Fsno Ln = Vbno M n and * Fsny Ln = Vbny M n (Eqs. A.4. An (t ) . Develop the Fsn Ln − Dn relation (Fig.2. 4.

1.5. 3. At this point. for the first.2. Ap1 = 360777 kN-cm. computed for the first three modes are shown in Fig. A. 25. First three mode shapes and frequencies of the selected building were computed and are shown in Fig.). A. are shown in Fig. B. 1. Following is step-by-step implementation of the procedure described in Section A.4 cm (10 in. 3. The force distributions. and 12. 3.urn ) pushover curve for the force distribution s* : n 2.Fsn / Ln Fsn / Ln − Dn Relationship Vbno / M* n Vbny / Mn * 1 α ω2 n n ω2 1 n figA_2. ur1o = 63.3. is defined at the target roof displacement. s* . The pushover curves for the first three modes.5 cm (25 in.2. 74 . Target displacements used to generate these pushover curves are 63.4.). The following steps illustrate the procedure to develop the idealize curve for the first “mode”.3.6 kN.7 cm (5 in. respectively. The base-shear – roof-displacement (Vbn .2 Properties of the nth-“mode” inelastic SDF system A. Area under the actual pushover curve.1. 2. generated using DRAIN-2DX.3.4.2 EXAMPLE The UMRHA procedure is implemented to calculate the response of the 9-story building described in Section 3. A.1.eps Dny = ur n y / Γnφr n Dno = ur n o / Γnφr n Dn Fig.5 cm and Vb1o = 8729. n 2. The anchor point. to the north-south component of the El Centro (1940) ground motion scaled up by a factor of 1. second. 3. 3. and third mode. Idealized bilinear curves for each of the three modes are included in Fig.).

8.2.6 8006. iterations are necessary.2. A.2.4 210.6 × Vb1 y = 4803. The initial slope of the idealized bilinear curve. The results of the iterative procedure are summarized in Table A.1.198%.1. (4.09 cm. and α1 = 0.4 ) − 1 (63. 75 .i 3.0.4 kN. * 4. The procedure converged after nineteen cycles to give ur1 y = 36.4. is calculated as follows.9.3. This value exceeds the prespecified tolerance of 0. i i 3.5 38. i i i 3. k1 = 0.13).194. and M1 = 2736789 × 1. The next estimate of the yield shear is Vbny = 8006.8 22. The post-yielding strain-hardening ratio. L1 = 2736789 kg. 3. The Fs1 L1 − D1 relation for the first “mode” is developed as follows. The results for other modes are summarized in Table A. i i 3. The yield displacement. Determined from the pushover database.5.4 × (360777 365100 ) = 7911.3666 = 3740189 kg. 3.86 cm at 0.4 kN.6.9. The curve OAB obtained by connecting the three points O.09 cm and Vb1 y = 8006.18 kN/cm.01%. Also included are the modal damping ratios and the periods calculated from Eq. Vb1 y = 7615. α1 = Vb1o Vb1 y − 1 ur1o ur1 y − 1 = ( ) ( ) (8729. 4. Area under the bilinear curve OAB. Ab1 = 365100 kN-cm. 3.9 kN.09 ) − 1 = 0.18 = 38.0.7.9. The point A on the i i bilinear curve is defined by ur1 y = 38. ur1 y = Vb1 y k1 = 8006.3666.4.86 = 210. and B with straight-line segments gives the idealized bilinear curve.6 = 22. i i i 3.1.6 × Vb1 y ( ) i ur1. Error = 100 × (365100 − 360777 ) 360777 = 1.6 kN. i 3. k1 . i 3. ur1.6 = 4803. Γ1 = 1. Therefore. The first estimate of the yield base shear Vb1 y = 8006.1.135.4.23 cm. i +1 3.8 kN.

1 and 4. 7. The peak values are also plotted in Fig. A. 9.7.62 (cm/sec2). 6. Histories of roof displacement and top story drifts for the first “mode” are computed and presented in Fig.3.4.51 cm. 4. 4. The combined modal responses are presented in Fig. 5.2.8. The results were generated for first three “modes’ and are included in Fig. Scaling the horizontal axis by Γ1φr1 gives D1o = 46. Scaling the vertical axis by Fs1 y L1 = 203. * M1 gives Fs1o L1 = 233. 4.4.46 cm and D1y = 26. 4.2. 8. The peak values are computed and are summarized in Tables 4.7.40 (cm/sec2) and 76 .7. Deformation and pseudo-acceleration histories of the inelastic SDF systems for the first “mode” with unit mass and force-deformation relation developed in Step 4 are plotted in Fig. 4.

19 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized figA_3a. “Modal” pushover curves for the example building 77 . V = 5210 kN.eps 0 0 2 4 6 8 10 Roof Displacement (cm) 12 Fig.eps 0 0 5 10 15 20 Roof Displacement (cm) (c) "Mode" 3 Pushover Curve 25 8000 u = 4.eps 10 20 30 40 50 Roof Displacement (cm) (b) "Mode" 2 Pushover Curve 60 8000 u = 9. V = 7616 kN.6 cm. A.3. α = 0. V y by = 4952 kN.9 cm.2 cm. α = 0. α = 0.13 6000 Base Shear (kN) 4000 2000 Actual Idealized figA_3b.14 y by 6000 Base Shear (kN) 4000 2000 Actual Idealized figA_3c.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36.

” and third “mode” inelastic SDF systems. Histories of deformation and pseudo-acceleration due to 1.5 × El Centro ground motion for the first “mode.4.06 0 • 1. 78 .(a) Deformation 80 n=1 3 (b) Pseuso−acceleration n=1 • 0.52 5 10 15 20 Time (sec) 25 30 0 • 1.747 5 10 15 20 Time (sec) figa_4.2769 D1 (cm) 0 • 35.33 A1 (g) n=2 0 −80 80 −3 3 n=2 D2 (cm) 0 A1 (g) • 22.” second “mode.223 −80 80 n=3 −3 3 n=3 D3 (cm) 0 A3 (g) • 10. A.m −80 0 −3 0 25 30 Fig.

56 47.910 0.3 7786.3 7658.44 36.194 0.190 0.90 21.24 36.309 0.1 4574.063 0.18 210.95 21.107 0.198 0.237 0.18 210.38 22.4 7714.184 0.4 7911. Results of iterative procedure to develop the idealized bilinear curve for the first “mode” inelastic SDF system Itr.5309 488839.18 210.048 0.18 210.194 0.6 × Vb1 y i ur1.180 0.9 4570.5 (cm) 22.7 7639.18 210.3 4603.74 (kN/cm) 210.193 0.23 0.Table A.18 210.2.18 210.35 36.56 19.81 21.18 210.139 0.32 36.0 7619.18 210.8525 1.182 0.4 7647.2 4571.18 210.013 0.18 (cm) 38.59 36.83 21.79 0.022 0.77 21.3 7628.2671 1.26 36.12 3876.4 4595.0 4704.18 210.8 7622.2 7690.8 7618.79 21.09 18.40 46.74 21.6 i k1 i ur1 y i α1 i Ab1 Error (%) 1.1 4569.9 (kN) 4803.7 4580.78 21.0.103 “Mode” 3 696400 0.30 37.05 36.3666 3740189 203.192 0.75 21.1 7616.193 0. i 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 i Vb1 y i 0.082 0.9 7615.18 210.010 (kN) 8006.05 52.529 0.8 4647.136 Fsny Ln (cm/sec2) Dny (cm) Fsno Ln (cm/sec ) Dno (cm) Tn (sec) 233.39 36.0 4577.50 36.151 0.64 37.28 36.6 4583.2 4671.6 7840.186 0.188 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.18 210.85 0.4927 1. No.25 36.46 2.65 1226.1 1013.176 0.62 26.193 0.029 0.85 36.194 0.5 3109.25 36.59 22.86 21.18 210.76 21.18 210.29 36.75 21.02 21.5 7624.18 210.11 22.18 210.70 36.191 0.23 22.948 ζ n (%) 79 .51 2 “Mode” 2 -920860 -0.170 0.86 22.2 4628.3 7745.037 0.18 210.1.2406 167531.9 4570.0 4588.9 4573.5 7633.162 0.404 0.5 4614.135 0.4 7672.09 37.8 4747.693 0. Properties of “modal” inelastic SDF systems Properties Ln (kg) Γn * M n (kg) “Mode” 1 2736789 1.017 0.

80 .

6.4.3b) with unit mass and force-deformation relation of Fig.18). or from the inelastic response (or design) spectrum. From the total hinge rotation. (3. Calculate the peak roof displacement urno associated with the nth-“mode” inelastic SDF system from Eq. At urno . Compute the peak deformation. of the nth-“mode” inelastic SDF system (Fig. Determine the total response by combining the peak “modal” responses using the SRSS combination rule of Eq.6b by solving Eq. 11. Steps 1 to 4 of the MPA are the same as those for UMRHA presented in Appendix A.2 EXAMPLE The MPA procedure is implemented to calculate the response of the 9-story building described in Section 3.Appendix B Modal Pushover Analysis B. extract from the pushover database values of other desired responses.1 to the north-south component of the El Centro (1940) ground motion scaled up by a factor of 1. rno . Dn .5. subtract the yield hinge rotation to determine the plastic hinge rotation. 13. 4. (4. (3. 10. 4.1 STEP-BY-STEP PROCEDURE A detailed step-by-step implementation of the modal pushover analysis (MPA) procedure is presented in this section and illustrated by an example in the next section. 10.33 cm.8). (4.1. 81 . the first two or three “modes” will suffice.21). Following is step-by-step implementation of the procedure described in Section B. 12.8) 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. Solving Eq. B. Typically. This is the same example as solved in Appendix A. Repeat Steps 3 to 7 for as many “modes” as required for sufficient accuracy.

4. 8. (3. 9.” and the results are included in Tables 4. The values are summarized in Table 4.18) are also included in Tables 4. 82 . At ur1o = 48. Results of Steps 5 and 6 are summarized in Table B. Steps 3 to 7 are repeated for first three “modes. 12.3 for the floor displacements normalized by the building height and in Table 4. Peak roof displacement ur1o = Γ1 × φr1 × D1 = 1.11. The total response computed by combining the peak “modal” responses using the SRSS combination rule of Eq.28 cm.3 and 4.28 cm. values of floor displacements and story drifts are extracted.4.1.33 = 48.3 and 4. where results for other ground motion intensities are also included. Also included in Table 4.5 are the plastic hinge rotations computed by subtracting the yield hinge rotation from the total hinge rotations.4 for the story drifts normalized by the story height.366 × 1 × 35.

59 0.551 2.467 14.913 22.200 0.5 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 2.577 16.37 1.184 0.1.52 0.38 22.139 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.748 63.0 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 3.5 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.79 0.766 7.85 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 1.35 0.06 1.0 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 1.535 14.229 8.33 1.0 µ = Dn Dny urno = Γnφrn Dn (cm) 83 .691 0.4222 3.03 0.52 “Mode” 3 1.05 0.755 27.275 1.023 0.27 0.267 5. Calculation of roof displacements urno from peak deformation of inelastic SDF systems Ground Motion Multiplier 0.117 5.35 1.660 14.504 18.82 1.678 0.332 48.379 21.07 “Mode” 2 4.395 0.457 12.367 1.735 3.450 4.312 1.676 6.332 13.436 7.252 9.25 20.37 57.225 2.969 0.75 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.901 8.18 27.690 10.73 24.154 78.28 46.Table B.50 35.856 31.268 0.185 11.513 0.755 0.70 0.007 36.126 13.71 1.8451 5.25 Quantity Dn (cm) “Mode” 1 6.13 2.36 1.03 26.

84

**Appendix C FEMA Force Distribution Calculations
**

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

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

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

Floor, j

mj (10-3×kg)

s* = j

∑i mi

0.112 0.110 0.110 0.110 0.110 0.110 0.110 0.110 0.119

mj

1 2 3 4 5 6 7 8 9

503.5 494.7 494.7 494.7 494.7 494.7 494.7 494.7 534.1

C.2

EQUIVALENT LATERAL FORCE (ELF) DISTRIBUTION 85

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

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

Floor j 1 2 3 4 5 6 7 8 9

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

-3 k

s* j

=

m jhk j

∑i mihik

0.007 0.020 0.038 0.062 0.091 0.126 0.165 0.210 0.281

C.3

SRSS DISTRIBUTION

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

N

∑ n (V jn )

2

to get column 8 of

2 200.6 319.1 -525.5 -350.0 354.1 Story Shears V j2 (kN) (6) 1114.8 430.1 832.2 Lateral Force fj (kN) (9) 203.9 446.5 159.2 105.1 -646.7 694.7 234.0 176.6 286.7 (10) 0.070 0.4 1759.1 87. 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) 1917.3 240.3 -6.7 101.6 -352.177 0.3 222.1 1857.2 fj = ∑ fi i fj (2) (3) (4) 59.6 1233.9 880.2 277.090 0.5 215.5 -973.1 -438.047 0.2 148.For convenience.9 832.7 1622. the lateral forces are normalized by the base shear to obtain column 10 in Table C.0 1476.7 355.7 95.2 95.7 2065.7 -46.4 400.045 0.7 1578.3 -646.3.042 0.9 366.2 285.4 250.2 97.9 1446.0 -5.0 136.8 1381.9 1683. Table C.3.7 525.098 0.7 374.0 1231.6 366.367 87 .7 Vj (kN) (8) 2268.9 -153.0 381.5 320.9 -166.6 -359.5 -320.6 -732.0 980.4 1842.8 374.1 -967.8 -326.065 0.6 V j3 (kN) (7) 478.

- 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
- EVALUATION OF PUSHOVER ANALYSIS PROCEDURES
- Pushover Analysis Final
- Pushover Ppt
- Performance Based Design Using Nonlinear Analysis
- Non-Linear Static Push Over Analysis
- Pushover Analysis using ETABS and SAP2000
- Non Linear Analysis Pushover
- Nonlinear Pushover Analysis of Reinforced Concrete Structures
- SAP-Etabs-Course-360
- Performance Based Seismic Design
- 2013 Pushover Analysis Guide Final RC Building With Masonry Infill Walls EditedDraft January 2013
- A comparison of single-run pushover analysis techniques for
- Pushover Analysis Using Etabs
- Performing Pushover Analysis in Etabs
- Action Plan for Performance Based Seismic Design
- Thesis
- Etabs-step by Step
- Pushover Analysis
- Etabs Guide
- 15WCEE 1940 pushover analysis.pdf
- ETABS Pushover Analysis
- Pushover
- pushover
- ATC-40 Vol 2
- A Modal Pushover Analysis Procedure to Estimate
- FEMA 451 Compl -Struct Analysis Performance-based
- Building report
- Pushover
- Pushover Analysis Guide Final
- PUSHOVER Report Chopra[1]