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

CHOPRA • RAKESH K. GOEL PEER 2001/03 JAN. 2001

Pacific Earthquake Engineering Research Center

**A Modal Pushover Analysis Procedure to Estimate Seismic Demands for Buildings:
**

Theory and Preliminary Evaluation

Anil K. Chopra University of California Berkeley Rakesh K. Goel California Polytechnic State University San Luis Obispo

A report on research conducted under grant no. CMS-9812531 from the National Science Foundation: U.S.-Japan Cooperative Research in Urban Earthquake Disaster Mitigation

PEER 2001/03

A Modal Pushover Analysis Procedure to Estimate Seismic Demands for Buildings: Theory and Preliminary Evaluation

Anil K. Chopra University of California Berkeley

Rakesh K. Goel California Polytechnic State University San Luis Obispo

A report on research conducted under Grant No. CMS-9812531 from the National Science Foundation: U.S. Japan Cooperative Research in Urban Earthquake Disaster Mitigation

PEER Report 2001/03 Pacific Earthquake Engineering Research Center College of Engineering University of California Berkeley January 2001 i

ii .

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

iv .

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

vi .

............................................................22 4.... Introduction.............................................................3 4...............27 4..3 Modal Pushover Analysis ............................6 Pushover Analysis ...........................1 3...............................................41 4...............36 Modal Pushover Analysis ............................4............ vii 1.............................................................5 Response History Analysis ....................55 vii ....................................................38 4...........................................28 4...............1 System and Excitation Considered . iii Acknowledgment ................... Equation of Motion .....55 Comparative Evaluation..2 Response History Analysis .....3 2.......................................................................1 4.........17 3..........................................4.....................33 4................................v Table of Contents.............6 Elastic Multistory Buildings ..............................................12 Modal Pushover Analysis .............................4 3.....................................3 2........................................................3 Summary............................4 Modal Response History Analysis.......................................................................................................................................................................................................................3 3..................................................47 4.............2 Response History Analysis .1 Summary.........................................2............................................................9 3..........................................4......................................................34 4......2.............................................1 One-Story Systems ............9 Modal Response Spectrum Analysis .................38 4........................................................................................................................................36 4................... Comparison of Modal and FEMA Pushover Analyses..12 Comparative Evaluation of Analysis Procedures ................................................................... Inelastic Multistory Buildings ....4.................................................................1 Underlying Assumptions and Accuracy .................................................................4..............................2 FEMA-273 Pushover Analysis .1 2.4....................................................................................................................1 5..............2 Modal Pushover Analysis .................................................2....................3 Modal Pushover Analysis with Gravity Loads ...................................4 5...................... 2..............................................................................................13 3...........................................55 5...............................................2 2...3 System and Excitation Considered ......................................37 Comparative Evaluation of Analysis Procedures ..............2 Properties of nth-“mode” Inelastic SDF System ...........................................................................................2 3...................................................................................................................1 Uncoupled Modal Response History Analysis .............................................................CONTENTS Abstract ...................................................3...............27 Uncoupled Modal Response History Analysis .......................................................................................................13 3................

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

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

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

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

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

this system and excitation.04 40 Base Shear.9 kN (38. m h = 3. 2. 6. α = 0.2.2 SYSTEM AND EXCITATION CONSIDERED Consider the one-story system in Fig.2 kips) .376 cm.2. Vb (kN) 30 20 10 0 0 2 4 6 Roof Displacement.3.2 5 .26 kN (8. For . 2. is f y / w = ( f o / w ) ÷ 8 = 0. Pushover curve for structure shown in Fig.5 sec and z = 5% subjected to the north-south component of the El Centro (1940) ground motion (Fig. 2.826 kips) for w = 169. one-bay frame 50 uy = 1. Vby = 39. The yield strength of the inelastic system. f o w = 184 . One-story.26 kN.2: the dimensions and flexural rigidity of the structural elements are noted.1.2311 . 2.32 m • Fig. u (cm) 8 Fig. and f y = 39.66 m • EIb EIc EIc • • L = 7. 2001) and scaled up by a factor of 2. based on Ry = 8 .4 in Chopra. with Tn = 0.

2. and E = 2 × 108 kPa (29×103 ksi) gives yield moments of 21.4 ). The system is excited well beyond the yield deformation.077 × 107 mm 4 (146 in. (b) shows the lateral force f s t or base shear Vb t normalized relative to the bg weight.The yield moments in the beam and columns are defined as the bending moments due to the lateral force f y . and q pm = 0.4edetermined from RHAis superimposed. the peak value determined from RHA.6 kip-in. Observe that the pushover curve matches the initial loading path of the hysteretic system. and (e) shows the force-deformation relation.017 rad . the joint rotation and beam hinge rotation are identical to values q m and q pm determined from RHA. e. However.36 cm .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 = 7.3 in. the ductility factor m = 5. It is organized in five parts: (a) shows the deformation u t . The peak values of the various response quantities are as follows: um = 7.134 × 107 mm 4 (75.g. (d) shows the rotation q p bt g of the plastic hinges at the bg bg beam ends. 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 .) and 50. respectively.4 ). The yielding stiffness of each structural element is defined as 3% of its initial stiffness.3 RESPONSE HISTORY ANALYSIS Figure 2.65 kN-m (191. 2. The resulting pushover curve is shown in Fig. A static pushover analysis of this one-story system leads to the force-displacement relationship shown in Fig. 2. (c) shows the joint rotation q bt g .. Implementing this analysis with I c = 6.0217 rad . 2. wherein the hysteretic force-deformation history of Fig.36 cm.4e.) for the beam and columns. Determined from pushover analysis at the exact peak deformation.18 kN-m (444.3.1 kip-in.35. as apparent in Fig. the energy dissipated in 6 . 2. I b = 3.4f.4 shows the earthquake response of the system described in the preceding section determined by response history analysis (RHA). q m = 0. 2. pushover analysis cannot provide any cumulative measure of response.

4 0.4 −10 −5 0 u (cm) 5 10 Fig.5 0.2 0 (e) 0 (f) −0.04 0.5 fy / w = 0.2 −0.4.4 −10 −5 0 u (cm) 5 10 −0.4 0.2311 (a) Vb / w 0 f / w = −0. (e) force-deformation relation. Response of one-story system to El Centro ground motion: (a) deformation.36 −15 0. 15 u (cm) 0 • 7. 2. This represents an inherent limitation of pushover analyses.2 −0.2311 y (b) −0. and (f) pushover curve 7 . or the cumulative rotation at a plastic hinge. (c) joint rotation.017 • 0 (d) θp (rad) −0.yielding during the ground motion.04 0.2 Vb / w 0. (d) plastic hinge rotation. (b) base shear.0217 • θ (rad) 0 (c) −0.04 0 5 10 15 Time (sec) 20 25 30 0.04 0.

8 .

and lateral stiffness matrices of the system. Elastic Multistory Buildings 3.mι ug t bg bg bg (3.1 MODAL RESPONSE HISTORY ANALYSIS The differential equations governing the response of a multistory building to horizontal earthquake ground motion u g t are as follows: mu + cu + ku = .3) where φ n is the nth natural vibration mode of the structure.4) 9 .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 . (3. m. This force distribution can be expanded as a summation of modal inertia force distributions s n (Chopra.n t = Â . classical damping.12): mι = n =1 Â sn = N n =1 Â Γ nmφ n N (3. c. and p eff t = Â peff.snug t n= n =1 The effective earthquake forces can then be expressed as af N af N af (3. and k are the mass.m ι u g t bg bg (3. The right side of Eq. 2001: Section 13.3.1) can be interpreted as effective earthquake forces: p eff t = .1) where u is the vector of N lateral floor displacements relative to the ground. each element of the influence vector ι is equal to unity.

Then the floor displacements are u n t = φ n qn t bg bg bg bg (3.5) The contribution of the nth mode to s and to p eff t are: sn = Gn mφ n respectively. (3.9) and (3. (3. (3. The equations governing the response of the system are mu + cu + ku = . an SDF system with vibration propertiesnatural frequency w n and damping ratio z n of the nth-mode of the MDF system.sn ug t bg bg bg bg (3.u g t n Comparing Eqs.n (t ) = N n =1 ∑ −sn u g (t ) N (3.Gn ug t (3.11) and substituting in Eq. it can be demonstrated that none of the modes other than the nth mode contribute to the response. The solution qn t can readily be obtained by comparing Eq.10) bg bg bg 10 (3.8) where the modal coordinate qn t is governed by 2 qn + 2z nw n qn + w n qn = .12) .8) gives the floor displacements u n t = Gnφ n Dn t bg (3.peff (t ) = n =1 ∑ peff .n t is entirely in the nth-mode. p eff .9) to the equation of motion for the nth-mode elastic SDF system.sn ug t bg (3. with no contribution from other modes.n t = .9) in which w n is the natural vibration frequency and z n is the damping ratio for the nth mode. we will outline that the response of the MDF system to p eff .7) By utilizing the orthogonality property of modes.10) gives qn t = Gn Dn t bg bg bg (3.6) Next. subjected to u g t : Dn + 2z nw n Dn + w 2 Dn = .

and An t = w 2 Dn t n bg (3. 3. However.16) This is the classical modal RHA procedure wherein Eq.13) st where rn denotes the modal static response. 2001.can be expressed by st rn t = rn An t bg bg bg bg bg (3. these standard equations have been derived in an unconventional way.9) is the standard modal equation governing qn t .14) is the pseudo-acceleration response of the nth-mode SDF system (Chopra. Conceptual explanation of modal response history analysis of elastic MDF systems Any response quantity r t story drifts.12) and (3. (3.13) define the contribution of the nth-mode to the response. Section 13. (3. st The two analyses that lead to rn and An t are shown schematically in Fig. Eqs.1. internal element forces. ζn ¨ g(t ) u (a) Static Analysis of Structure (b) Dynamic Analysis of SDF System Fig. (3.Forces sn An(t ) st rn ωn.1). and Eqs. etc. 3.15) rt = bg n =1 Â rn t = bg n =1 Â rnst An bt g N (3.12) and (3. In contrast to the classical 11 bg . the static value of r due to external forces sn . Therefore. the response of the system to the total excitation p eff t is ut = bg bg bg n =1 N Â un bt g = Â Gnf n Dn bt g n =1 N N (3.n t .15) and (3.1.16) reflect combining the response contributions of all modes.13) represent the response of the MDF system to p eff. Equations (3.

4 and 13. The SRSS rule.9) . The peak modal responses are combined according to the Square-Root-of-Sum-ofSquares (SRSS) or the Complete Quadratic Combination (CQC) rules.17) where An is the ordinate A Tn .. In such a response spectrum analysis (RSA). 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 (3.8. 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 3.1). 2001. (3. (3.(3. Sections 12.18) To develop a pushover analysis procedure consistent with RSA. Alternatively. which is valid for structures with well-separated natural frequencies such as multistory buildings with symmetric plan. we note that static analysis of the structure subjected to lateral forces f no = Γ n mφ n An (3. 3. Chopra. we used the modal expansion of the spatial distribution of the effective earthquake forces. This concept provides a rational basis for the modal pushover analysis procedure developed later.g.1. and Tn = 2p w n is the natural vibration period of the nth-mode of the MDF system.16). Section 13.17) (Chopra. this response value can be obtained by static analysis of the structure subjected to lateral forces distributed over the building height according to 12 .derivation found in textbooks (e. z n of the pseudo-acceleration response (or design) spectrum for the nth-mode SDF system.3).19) will provide the same value of rno . 2001.3 MODAL PUSHOVER ANALYSIS (3. the peak nth-mode response as in Eq.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.

73 m (150 ft) by 45. this structure meets seismic code and represents typical medium-rise buildings designed for the Los Angeles.2. Typical floor-to-floor heights (for analysis purposes measured from center-of-beam to center1 2 Brandow & Johnston Associates. shown in Fig. region. 3.12) is urno = Gnf rn Dn (3. Although not actually constructed. A benchmark structure for the SAC project. and 37. the Applied Technology Council (ATC).4.21) 2 where Dn = An ω n . The peak modal responses. this building is 45. CA 90017. The interior bays of the structure contain simple framing with composite floors. rno . SAC Steel Project Technical Office. Los Angeles. was designed by Brandow & Johnston Associates1 for the SAC2 Phase II Steel Project. California. (3. denoted B1. Richmond.19 m (122 ft) in elevation. 46th Street.1 System and Excitation Considered The 9-story structure. with the ninth level being the roof.s* = mφ n n (3. 3. and California Universities for Research in Earthquake Engineering (CUREE). The building has a basement level. The levels of the 9-story building are numbered with respect to the ground level (see Fig. This modal pushover analysis (MPA) for linearly elastic systems is equivalent to the well-known RSA procedure (Section 3. 1301 S. The bays are 9. which from Eq. Obviously Dn and An are available from the response (or design) spectrum. CA 94804-4698. with five bays each in the north-south (N-S) and east-west (E-W) directions. in both directions. (3.2).4 COMPARATIVE EVALUATION OF ANALYSIS PROCEDURES 3. 3.15 m (30 ft) on center. can be combined according to Eq. the peak value of the roof displacement due to the nth-mode.18) to obtain an estimate of the peak value ro of the total response. urno . 1660 W. Consulting Structural Engineers. 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. Third St.. 13 . The columns are 345 MPa (50 ksi) steel wideflange sections.73 m (150 ft) in plan. SAC is a joint venture of three non-profit organizations: The Structural Engineers Association of California (SEAOC). each determined by one pushover analysis.2).20) and the structure is pushed to the roof displacement.

49 sec.65 m (12 ft) and for the first floor is 5. for the second through eighth levels is 9. and seventh levels at 1. 3. Each frame resists one half of the seismic mass associated with the entire structure. including the steel framing.of-beam) are 3. The force distributions. 0. The strength. third. 1988) using the M1 model developed by Krawinkler and Gupta (1998).85. ceiling/flooring. for the first level is 1. fifth. These force n distributions will be used in the pushover analysis to be presented later.0 kips-sec2/ft). The floor system is composed of 248 MPa (36 ksi) steel wide-flange beams in acting composite action with the floor slab.96 m (13 ft). the vibration periods are 2. for the first three modes are shown in Fig. The building is modeled in DRAIN-2DX (Allahabadi and Powell. The first three vibration modes and periods of the building for linearly elastic vibration are shown in Fig.49 m (18 ft).sec2/ft). which are seismic (tension) splices to carry bending and uplift forces. To ensure that this structure remains elastic. i. The simple model adopted here is sufficient for the objectives of this study.2.20).0 kips-sec2/ft). partitions. The column bases are modeled as pinned and secured to the ground (at the B-1 level). if desired more complex models. The column lines employ two-tier construction. and shear distortion of panel zones are neglected but large deformation (P∆) effects are included. roofing and a penthouse located on the roof. mechanical/electrical.83 m (6 ft) above the center-line of the beam to column joint. The 9-story N-S MRF is depicted in Fig.00×106 kg (616 kips.27. 3.4. 3. 14 . are located on the first. Column splices. 3.65×105 kg (66.7 kips-sec2/ft). respectively.07×106 kg (73. and for the ninth level is 1. Concrete foundation walls and surrounding soil are assumed to restrain the structure at the ground level from horizontal displacement.2 kips-sec2/ft). and 0. monolithic column pieces are connected every two levels beginning with the first level.01×106 kg (69.. The floor-to-floor height of the basement level is 3. floor slabs. The seismic mass of the above ground levels of the entire structure is 9.3. dimension. such as those described in Gupta and Krawinkler (1999) can be used.e.89×105 kg (67. s* (Eq. we select a weak ground motion: the northsouth component of the El Centro (1940) ground motion scaled down by a factor of 4. The seismic mass of the ground level is 9. This model is based on centerline dimensions of the bare frame in which beams and columns extend from centerline to centerline. The seismic mass of the structure is due to various components of the structure.

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

44 1. is shown in Figs. 2.05 s * 1 s * 2 s * 3 Fig.5 Mode Shape Component 1 1.85 sec 2 Floor 5th 4th 3rd 2nd 1st T1 = 2. and 3 n 3. Each figure is organized in four parts: (a) shows the roof displacement urn t .5 0 0.2 Response History Analysis The structural response due to individual vibration modes.05 2. n = 1. determined by RHA [Eqs.67 −1.9th 8th 7th 6th T = 0.5 Fig.37 2.33 2.796 0.5 −1 −0. (c) shows the joint rotation q n t of an external joint at 16 bg bg bg .39 3. respectively.5.1 3.13 −1.0272 −2. First three natural-vibration periods and modes of the 9story building 3.4.49 sec 3 T = 0. 3.1 −2.72 −2. and 3 .487 −1.728 2.12 0.51 0.03 −1. Force distributions s* = mφn .7. 3.04 1. (3.6. 2.27 sec Ground −1.3.93 −1.12) and (3.13)]. 3. (b) shows the base shear Vbn t normalized relative to the weight W of the building. n = 1. 3. and 3.05 1.38 0.75 1.31 −0.61 2.8 −2.4.94 2.

2. errors are smallest in floor displacements. The peak values of the various response quantities are noted in these figures. especially in estimating the story drifts. or all modes. drifts in all stories. 3. which are listed in Tables 3. The peak values of displacements of all floors. two. Also included are the combined response due to one.23 cm.422 cm.9. Combining the modal response histories for all modes gives the total response [Eqs. The same method was used to determine the peak values of many response quantities. including one.urn relation. The peak values of floor displacements and story drifts determined by RHA. but three modes—perhaps even two modes—are sufficient. It is apparent that the first mode alone is inadequate. 3. consistent with the increasing significance of the higher mode response among these three sets of response quantities. in particular. and (d) shows the Vbn . 3. where the second and third modal responses are a larger percentage of the top story drift compared to roof displacement. two. and 3. (3. with the first three modes provide essentially the exact response. For a fixed number of modes included. the exact response considering all modes. three. and 3.3. As expected. and rotations of external joints with moment connections are presented in Tables 3. and the percentage errors due to truncation of higher modal contributions. 3. The linear relationship between the base shear and roof displacement for each mode implies that the structure did not yield. 3. This is illustrated in Fig. The errors due to truncating the contributions of vibration modes beyond the third mode are negligible.12 cm.16)].the roof level.1. are presented in Fig. and three vibration modes. respectively. and even larger in joint rotations. the results for the roof displacement and top-story drift are shown in Fig. errors generally decrease as response contributions of more modes are included.8. 17 .2. ur 2o = 2. larger in story drifts. and ur 3o = 0. respectively.8.15) and (3.1.3. the peak roof displacement due to each of three modes is ur1o = 9.

50E-03 2.11E-04 -5.264 Error (%) 1 Mode 2 Modes 3 Modes -23.3 Peak values of joint rotations (radians) from RHA for 0.8 1.226 0.006 0.44E-03 1.261 -0.1 Table 3.47E-03 1.0 -0.14 m3) from RHA for 0.7 3.080 0.258 0.74E-03 1.310 0.3 -8.226 0.76E-03 1.202 0.1 -11.38E-04 2 Modes 2.401 Error (%) 1 Mode 2 Modes 3 Modes -23.336 0.197 0.333 0.062 -0.45E-03 3.24E-03 2.65E-03 2.31E-03 Mode 2 -1.1 -2.0 -2.125 0.1 Peak values of floor displacements (as % of building height = 37.89E-03 1.012 -0.062 0.3 Table 3.300 0.8 1.0 -46.026 0.99E-03 Mode 3 3.089 0.13E-04 9.311 0.370 0.99E-03 2.7 2 Modes -3.2 -1.03E-03 1.058 -0.275 0.325 0.295 -0.9 1.121 0.466 0.227 0.00E-03 2.29E-03 2.3 -3.157 0.259 0.6 -1.032 -0.1 -0.2 -57.3 -33.231 -0.03E-03 1.263 0.235 0.2 Peak values of story drift ratios (as % of story height) from RHA for 0.060 -0.2 9.88E-03 2.15E-03 4.282 0.023 0.069 0.022 0.5 -2.74E-04 9.265 0.2 -2.8 -10.260 0.6 -1.307 0.2 -20.7 7.6 0.85E-03 3.65E-03 RHA (all modes) 2.009 -0.03E-03 3 Modes 2.4 -6.237 0.235 0.1 -14.9 9.117 0.177 0.266 0.3 -0.177 0.453 0. 18 .1 -2.4 -3.4 -41.2 1.5 18.09E-03 Error (%) 1 Mode -23.4 -22.1 Building height is measured from the ground floor to the 9th floor.202 0.192 0.245 0.03E-03 -6.265 RHA (all modes) 0.3 19.9 -15.253 0.038 0.097 0.6 9.024 -0.6 11.09E-03 1.399 0.01E-04 -2.303 0.205 0.071 0.76E-03 1.202 0.054 0.282 0.124 0.035 0.4 -0.1 -19.89E-03 1.63E-03 2.229 0.173 0.245 0.179 0.Table 3.042 0.152 0.90E-03 3.5 -1.4 -1.259 0.94E-03 2.159 0.001 -0.74E-04 6.14E-03 2.78E-04 -3.9 -16.8 -15.7 -19.66E-05 -3.311 0.91E-04 1.09E-03 2.73E-03 3.28E-04 1.8 -56.00E-03 1.56E-03 2.413 RHA (all modes) 0.13E-03 2.055 0.4 -10.1 -2.183 0.203 0.227 0.266 0.350 0.64E-03 3.253 0.045 0.1 4.097 0.173 0.0 7.9 3.31E-03 Joint Rotation (rad) Combined (RHA) 1 Mode 2.015 -0.9 8.7 2.08E-03 2.09E-03 1.4 -53.4 0.321 0.002 -0.9 -22.225 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 0.0 3 Modes -5.069 0.043 0.38E-03 2.33E-03 2.7 -50.88E-03 2.378 0.475 0.060 0.124 0.011 0.6 -0.5 0.133 -0.266 0.8 -1.25 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 3 Modal Response Mode 1 2.4 -1.181 0.8 -5.25 × El Centro ground motion Displacement /Height (%) Modal Response Combined (RHA) Mode 1 Mode 2 Mode 3 1 Mode 2 Modes 3 Modes Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 0.260 0.6 1.26E-04 -5.00E-03 1.406 0.4 -7.152 0.407 0.097 0.2 -4.44E-03 3.0 -2.042 0.50E-03 4.199 0.11E-03 1.6 0.01E-04 3.6 4.060 -0.74E-03 1.010 -0.088 -0.72E-03 3.22E-03 2.3 -0.090 0.7 4.9 -24.63E-03 2.130 0.2 0.237 0.1 3.156 0.9 -23.156 0.229 0.400 0.125 0.364 0.42E-04 1.9 2.011 0.003 0.0 -10.012 0.008 -0.

00174 • 0 (c) r1 −5 0 5 10 15 Time (sec) 20 25 30 0.12 • 0 (a) r1 −15 0.0435 9.1 −15 −7.1 0.5 0 u (cm) r1 7. and (e) pushover curve.1 −15 −7.1 5 x 10 −3 θ (rad) 0.5 15 −0. (c) joint rotation. Excitation is 0.5.12 0 (d) 0 (e) 15 Fig.25 × El Centro ground motion 19 .5 0 u (cm) r1 7.15 u (cm) 9.1 0.12 −0.0435 Vb1 / W 0.5 9. (b) base shear.0435 • 0 (b) V b1 −0.1 /W 0. (d) force-deformation history. Response due to first mode: (a) roof displacement. 3.

Excitation is 0.5 u (cm) r2 0 • 2.23 −5 (a) 0.0252 −2.1 0. (c) joint rotation.1 −5 −2.5 0 u (cm) r2 2.1 Vb2 / W 0.0252 (d) 0 0 (e) −0.5 0 u (cm) r2 2. (b) base shear. Response due to second mode: (a) roof displacement.05 5 x 10 −3 b2 θ (rad) 0.23 −2.0252 • 0 (b) V −0.5 5 −0. (d) force-deformation history.05 /W 0.00244 • 0 (c) r2 −5 0 5 10 15 Time (sec) 20 25 30 0. and (e) pushover curve.5 5 Fig.23 0.1 −5 −2.6. 3.25 × El Centro ground motion 20 .

3.25 × El Centro ground motion 21 .5 0 u (cm) r2 0.02 −1 −0.5 0 u (cm) r3 0. Response due to third mode: (a) roof displacement. (c) joint rotation.7.02 0.0108 x 10 −3 (b) −0. Excitation is 0.422 /W V b3 0 (d) 0 −0.02 2 θ (rad) 0.422 (e) −0. (b) base shear. and (e) pushover curve.0108 −0.422 −1 (a) 0. (d) force-deformation history.02 /W V b3 0 • 0.0108 −0.5 1 −0.02 −1 −0.1 u (cm) r3 0 • 0.02 −0.000913 • 0 (c) r3 −2 0 5 10 15 Time (sec) 20 25 30 0.5 1 Fig.

the value determined by RHA (Fig. For a fixed number of modes included considered.3 Modal Pushover Analysis Implementing MPA for the fundamental vibration mode. the floor displacements are proportional to the mode shape φ 1 because the structure remains elastic. leads to the pushover curves shown in Figs. i. these pushover curves are consistent with the Vb . story drifts.5. 3. implying that the modal truncation errors are small if at 22 . 3.1 through 3.4. 3.6. or three vibration modes. and 3. and 3. and 3.7d).18). 3. 3. 3. this value would be determined directly from the response (or design) spectrum.e.3).21).3). the errors in the MPA results are generally larger than in RHA (Fig..4.12).2.4) to roof displacement ur1o = 9.5d). respectively. In practical application. The floor displacements. As suggested by Eq.20) with n = 2 and 3 up to roof displacements ur 2 o = 2. and it is apparent in the example considered that three modes provide most of the response (Fig. two.5.5e. and external joint rotations computed by pushover analysis are presented in Tables 3.20) with n = 1 (Fig.9 and Tables 3.23 cm . (3.6 present estimates of the combined response according to Eq. Implementing pushover analysis for the second and third modes. 3. although both analyses led to identical peak values of the individual modal responses. In RHA the errors arise only from truncating the responses due to higher modes. 3. (3. pushing the structure. 3.422 cm . respectively.9).6e and 3. pushing the structure using the force distribution of Eq.3. and 3. respectively. As for the first mode. (3. story drifts.6d and 3.7e and to the floor displacements. Observe that the target roof displacement in each pushover analysis is identical to its exact value determined by RHA.ur relations determined by RHA (Figs..8) leads to the pushover curve shown in Fig. i.2.e. and the errors in these estimates relative to the exact response from RHA considering all modes. 3. considering one.12 cm. and the computed response values are identical to the peak response values determined from RHA (Tables 3. 3. which would provide the Dn value to be substituted in Eq. confirming that MPA gives the exact values of the individual modal responses. 3. and ur 3o = 0.1.3). and 3.5.4.10 and Tables 3. using the force distribution of Eq. This pushover curve is consistent with the relationship between the base shear and roof displacement determined by RHA (Fig. except for the algebraic sign associated with Γ1 and minor round-off errors. (3. (3. These values of the response quantities are identical to the peak response values determined from RHA (Tables 3. 3. Figure 3.6.1. and external joint rotations in Tables 3.4.

1 0.eps r 5 10 15 20 Time.3 0.2 0.5 0.83 • r3 0 • 0. 3. 3.9.1 0.422 r2 0 • 1.48 −3 3 1.4 Story Drift Ratio (%) 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_8.25 × 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.eps Ground 0 0. (a) Roof Displacement 15 (b) Top Story Drift 3 Mode 1 ∆ (cm) u (cm) 9.5 Ground 0 0. sec 25 30 Fig.3 0.685 • −3 3 Mode 2 ∆ (cm) u (cm) • 2. Additional errors are introduced in pushover analysis due to the approximation inherent in modal combination rules.eps fig3_9b.8.4 Displacement/Height (%) 0. sec 25 30 r 5 10 15 20 Time. Response histories of roof displacement and top-story drift from RHA for 0.2 0.least three modes are included.23 r2 −15 15 0 • 0.6 Fig.12 • r1 −15 15 0 r1 0 0 0.

322 0.9 -14.53E-04 -9.069 0.2 -11.57E-03 1 Mode -23.80E-04 3.374 0.2 -12.024 0.229 0.203 0.6 -16.4 -22.09E-04 -3.008 -0.203 0.300 0.4 -4.156 0.89E-03 -2.090 0.177 0.6 -19.012 0.245 0.062 0.235 0.203 0.25 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Modal Response Mode 1 -2.0 -46.274 0.99E-03 Mode 3 -3.Table 3.9 -12.44E-03 3.197 0.4 0.060 -0.270 0.78E-04 2 Modes 2.043 -0.2 -57.00E-03 2.89E-03 2.8 -14.177 0.24E-03 -2.1 -11.4 -9.253 RHA (All Modes) 0.332 0.088 -0.43E-04 -1.080 0.4 -2.00E-03 1.272 0.227 0.331 0.011 0.9 -15.9 -13.0 -0.1 -19.336 0.8 -23.010 0.267 0.121 0.1 -11.00E-03 2.245 -0.3 -19.125 0.04E-03 3.328 0.63E-03 2.310 0.4 -7.152 0.003 -0.133 0.3 Table 3.7 24 .106 0.157 0.2 -4.03E-03 3.6 -15.267 0.4 -53.253 0.133 -0.12E-03 1.173 0.106 0.9 -15.0 -10.2 -16.9 -13.14 m) from MPA for 0.74E-03 -1.63E-03 -2.38E-03 3 Modes 2.2 1.03E-03 1.078 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 -0.92E-04 -1.7 -19.124 0.270 0.015 -0.226 0.9 -24.385 0.253 0.42E-04 -1.9 -11.058 -0.026 0.7 2 Modes -13.8 -56.038 -0.76E-03 -1.0 -2.47E-03 1.4 -19.133 0.401 Error (%) 1 Mode 2 Modes 3 Modes -23.7 2.055 0.048 0.03E-03 -1.370 0.73E-05 3.22E-03 -2.89E-03 1.9 -13.042 0.179 0.045 -0.3 -13.22E-04 2.1 -0.237 0.022 0.89E-03 2.313 0.9 -16.74E-04 -6.286 0.069 0.009 0.229 0.152 0.7 -50.012 -0.285 0.260 0.157 0.2 -0.267 -0.7 -15.05E-03 3.15E-03 2.125 0.097 0.036 -0.3 -12.276 0.0 -16.76E-03 1.40E-04 5.5 Peak values of story drift ratios (as % of story height) from MPA for 0.260 0.179 0.4 Peak values of floor displacements (as % of building height = 37.336 RHA (all modes) 0.002 0.231 -0.2 Table 3.38E-03 2.4 -11.96E-03 2.0 -18.3 -2.33E-04 5.7 -21.173 0.8 -15.74E-03 1.15E-03 1.296 -0.4 -5.097 0.28E-03 2.407 0.97E-03 1.4 -4.181 0.282 0.071 0.270 0.90E-04 -9.237 -0.230 -0.00E-03 -1.25 × 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 0.3 -41.9 -13.8 -15.4 -22.466 0.9 3 Modes -12.4 -14.011 -0.94E-03 2.203 0.117 0.260 -0.9 -14.73E-03 3.8 -22.08E-03 2.89E-03 -1.09E-03 -1.3 -9.282 -0.266 0.230 0.00E-03 3.079 0.261 -0.2 -20.253 -0.44E-03 -1.6 Peak values of joint rotation (radians) from MPA for 0.9 -18.9 -14.048 0.062 -0.65E-03 2.042 0.1 -18.5 -16.001 0.09E-03 Mode 2 1.31E-03 Error (%) RHA (all modes) 2.4 1.259 -0.90E-03 1.72E-03 3.09E-03 1.060 0.032 0.264 Error (%) Modal Response 1 Mode 2 Modes 3 Modes -23.321 0.4 -9.156 0.9 -8.3 -33.3 -14.203 0.006 -0.6 -17.9 -15.31E-03 2.9 -15.227 0.00E-03 2.023 -0.235 -0.03E-03 6.259 0.3 1.31E-03 Joint Rotation (rad) Combined (MPA) 1 Mode 2.

4 Story Drift Ratio (%) 0.(a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th (b) Story Drift Ratios Floor 4th 3rd 2nd 1st Floor 5th RHA All Modes MPA 1 Mode 2 Modes 3 Modes 5th 4th 3rd 2nd 1st RHA All Modes MPA 1 Mode 2 Modes 3 Modes fig3_10a.5 Ground 0 0.2 0.3 0.eps Ground 0 0.5 0.10.2 0.6 Fig.3 0. shading indicates modal combination error 25 .1 0.1 0. 3.4 Displacement/Height (%) 0. Heightwise variation of floor displacements and story drifts from MPA for 0.25 × El Centro ground motion.eps fig3_10b.

26 .

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

Unlike Eq. This approximate RHA procedure is the preliminary step in developing a modal pushover analysis procedure for inelastic systems. premultiplying by φ T .4) is rarely solved because it offers no particular advantage over Eq. qr t = 0 for all 28 bg .4)] leads to the uncoupled modal response history analysis (UMRHA) procedure.and classical n damping-orthogonality property of modes gives F qn + 2z nw n qn + sn = .Gn ug t Mn bg n = 1.8) because qr t will generally be nonzero for “modes” other than the nth “mode.… N (4. (4. 4.n t given by Eq.4) where the only term that differs from Eq. implying coupling of modal coordinates because of yielding of the structure. Equation (4.2). (3. however.9) involves T Fsn = Fsn qn . 2.6) for inelastic systems will no longer be described by Eq. The spatial distribution s of the effective earthquake forces is expanded into the modal contributions sn according to Eq. give the same results for u t as obtained directly from Eq. (4.Substituting Eq. (4. (3. and using the mass. these equations are coupled for inelastic systems. (4. (4.6) The solution of Eq.3).2). Simultaneously solving these coupled equations and using Eq. For linear systems. (4. (3.2 UNCOUPLED MODAL RESPONSE HISTORY ANALYSIS bg Neglecting the coupling of the N equations in modal coordinates [Eq.6b) are mu + cu + f s u.3) in Eq. (3. The equations governing the response of the inelastic system to p eff.sn ug t bg b g bg bg (4. Eq. (3.5) This resisting force depends on all modal coordinates qn t .9) for linearly elastic systems.” implying that other “modes” will also contribute to the solution.2). (4. in principle. (4. sign u n b g b g bg (4. sign qn = φ n f s u n . However. sign u = .4) represents N equations in the modal coordinates qn . where φ n are now the modes of the corresponding linear system.3) will.

(a) p 15

eff,1

= −s × 0.25 × El Centro

1

(b) p 5

eff,2

= −s × 0.25 × El Centro

2

ur1 (cm)

0

ur1 (cm)

9.112 •

Mode 1

Mode 1

0

−15 15 Mode 2

−5 5 Mode 2

ur2 (cm)

0

ur2 (cm)

0 • 2.226 5 10 15 20 25 30

−15 0 15

5

10

15

20

25

30

Mode 3

−5 0 5

Mode 3

ur3 (cm)

0

ur3 (cm)

5 10 15 20 Time (sec) 25 30

0

−15 0

−5 0

5

10 15 20 Time (sec)

25

30

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

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

motion

af

af

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

bg

bg

(a) p 80

eff,1

= −s × 1.5 × El Centro

1

(b) p 20

eff,2

= −s × 1.5 × El Centro

2

Mode 1

Mode 1 • 1.437

u (cm)

r1

0 • 48.24

ur1 (cm)

Mode 2

0

−80 80

−20 20 Mode 2

u (cm)

r2

0

3.37 •

ur2 (cm)

0 • 11.62 5 10 15 20 25 30

−80 0 80

5

10

15

20

25

30

Mode 3 0.4931 •

−20 0 20

Mode 3 • 0.7783

u (cm)

r3

0

ur3 (cm)

25 30

0

−80 0

5

10 15 20 Time (sec)

−20 0

5

10 15 20 Time (sec)

25

30

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

af

af

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

bg

bg

by Eq. (3.8),

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

T Fsn = Fsn qn , sign qn = φ n f s qn , sign qn

b

g

b

g

(4.7)

30

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

T Fsn = Fsn Dn , sign Dn = φ n f s Dn , sign Dn

bg

bg c h

(4.8)

c

h

(4.9)

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

b

g

**w n and damping ratio z n —of the nth-mode of the corresponding linear MDF system; (2) unit
**

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

bg

bg

bg

bg

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

and An t (3.16). This is the UMRHA procedure.

4

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

31

n t = .112 5 10 15 20 Time (sec) 25 30 −10 0 −10 0 Fig.9 • 0 −50 10 −50 10 n=3 ur3 (cm) 0 • 5.51 • ur2 (cm) 0 ur2 (cm) n=3 14. n = 1.0 ¥ El Centro ground motion: (a) exact solution by NL-RHA.s n u g t .Forces sn Unit mass An(t ) st rn ωn.3.89 50 r1 0 u (cm) 0 • 78. and (b) approximate solution by UMRHA bg af bg 32 . and 3.4.575 5 10 15 20 Time (sec) 25 30 ur3 (cm) 0 • 5. 4. Roof displacement due to p eff. ζn. Fsn / Ln ¨ g(t ) u (a) Static Analysis of Structure (b) Dynamic Analysis of Inelastic SDF System Fig. Conceptual explanation of uncoupled modal response history analysis of inelastic MDF systems (a) Nonlinear RHA 150 n=1 150 (b) UMRHA n=1 ur1 (cm) • 75. 2. where u g t = 3.02 −150 n=2 −150 50 n=2 14. 4.

855 ∆r1 (cm) 0 ∆r1 (cm) n=2 • 6.1 Underlying Assumptions and Accuracy The approximate solution of Eq. (2) the superposition of responses to p eff.4 and 4.6) by UMRHA is compared with the “exact” solution by nonlinear RHA.817 5 10 15 20 Time (sec) 25 30 −10 0 −10 0 Fig.33 0 −20 20 −20 20 n=2 ∆r2 (cm) 0 ∆r2 (cm) 7. 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. n = 1. (4.956 5 10 15 20 Time (sec) 25 30 ∆r3 (cm) 0 • 5. where u g t = 3. and 3. Top story drift due to p eff.5. (4.744 • 0 −20 10 n=3 −20 10 n=3 ∆r3 (cm) 0 • 5.2… N ) according to Eq. and (3) the Fsn Ln . (3.(a) Nonlinear RHA 20 n=1 20 (b) UMRHA n=1 • 5.0 ¥ El Centro ground motion. Such comparison for roof displacement and top-story drift is presented in Figs.5)] is neglected.5.n t (n = 1. but even for this very intense excitation. and (b) approximate solution by UMRHA af bg bg 4.n t = .0 ¥ El Centro ground motion: (a) exact solution by NL-RHA. 4. 4.008 • 6. respectively.15) is strictly valid only for linearly elastic systems. both for 3. this intense excitation was chosen to ensure that the structure is excited well beyond its linear elastic limit.Dn relation is 33 bg bg . The errors are slightly larger in drift than in displacement.2.s n u g t .4) and (4. the errors in either response quantity are only a few percent. 2.

4. 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. (3.4. Therefore we impose this constraint in developing the UMRHA procedure in this section and modal pushover analysis in the next section. When implemented by commercially available software. is plotted against roof displacement urn .8) in UMRHA. it can conduct a force-controlled nonlinear static analysis with an invariant distribution of lateral forces.urn pushover curve to the Fsn Ln .approximated by a bilinear curve to facilitate solution of Eq.8) governing Dn t is based on Eq. (4. within the linearly elastic range of the structure.2 Properties of the nth-“mode” Inelastic SDF System How is the Fsn Ln . (4.Dn curve.Dn relation? The two sets of forces and displacements are related as follows: V Fsn = bn Γn Dn = urn Γ nf rn 34 (4. Therefore. How to convert this Vbn .9).9) should be determined by nonlinear static analysis of the structure as the structure undergoes displacements u = Dnφ n with increasing Dn . and the base shear Vbn .12) for floor displacements.20). (4. this distribution seems to be a rational choiceeven for inelastic systemsto determine Fsn in Eq. The overall errors in the UMRHA procedure are documented in the examples presented in Section 4. At the yield point. the only force distribution that produces displacements proportional to f n is given by Eq. which is different than the Fsn Ln . such nonlinear static analysis provides the so-called pushover curve. However.10) . when specialized for linearly elastic systems it is identical to the RHA procedure of Section 3.8) before it can be solved? Because Eq. the relationship between lateral forces f s and Dn in Eq. Although most commercially available software cannot implement such displacement-controlled analysis.20) to some bg pre-determined roof displacement. (3.Dn relation to be determined in Eq. (4. The structure is pushed using the force distribution of Eq.6a.2. the base shear is Vbny and roof displacement is urny . (3. 4. A bilinear idealization of this pushover curve for the nth-“mode” is shown in Fig.1. Although several approximations are inherent in this UMRHA procedure. (4.

(4. where the yield values of Fsn Ln and Dn are Fsny Ln = Vbny * Mn Dny = urny Γ nf rn (4.12) implying that the initial slope of the curve in Fig.8). 2001. the elastic vibration period Tn of the nth-mode SDF system is computed from F Ln Dny I 1/ 2 Tn = 2p G H Fsny JK (4.2.6b. 35 . the initial slope of the pushover curve in Fig.5). should be used in Eq.13) This value of Tn . (4. Section 13. 4. Properties of the nth-“mode” inelastic SDF system from the pushover curve Equation 4.6. Knowing Fsny Ln and Dny from n Eq.6a is 2 kn = ω n Ln . In contrast. The two are related through Fsny Ln 2 = w n Dny (4. 4.6b is w 2 . which may differ from the period of the corresponding linear system. 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.10 enables conversion of the pushover curve to the desired Fsn Ln − Dn relation shown in Fig.11) * in which M n = Ln Γn is the effective modal mass (Chopra. 4. which is not a meaningful quantity.11).

4. (3. of the total response r t obtained in Step 8.16) to determine the total response. 4. 9. and pseudo-acceleration history.n t . 7. 4. ωn . Compute the natural frequencies. Calculate histories of various responses by Eqs. Convert the idealized pushover curve to the Fsn Ln − Dn relation (Fig. the first two or three modes will suffice. r o . Repeat Steps 2 to 6 for as many modes as required for sufficient accuracy. for linearly-elastic vibration of the building. Idealize the pushover curve as a bilinear curve (Fig.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. the peak value of Dn t .6b) by utilizing Eq. Dn (t ) .21) where Dn . develop the base-shear – roof-displacement ( Vbn − urn ) pushover curve for the force distribution s* [Eq. An t .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. n (3.20)].2.13). 8. 5. 4. 6. is now determined by 36 bg bg . 4.11). Consider a nonlinear static analysis of the structure subjected to lateral forces distributed over the building height according to s* [Eq. details are available in Appendix A: 1. (3. Combine the “modal” responses using Eqs. Typically. Calculate the peak value.20)]. n 3.12) and (3. (4. (3. and modes. Compute the deformation history. For the nth-mode.3b) with force-deformation relation of Fig. with the structure is pushed to the roof displacement urno . 2. φn . which is summarized next as a sequence of steps.4. (3. of the nth-“mode” inelastic SDF system (Fig.6b. This value of the roof displacement is given by Eq.6a).15) and (3.

as mentioned earlier. or from the inelastic response (or design) spectrum. which is summarized next as a sequence of steps. joint rotations. Compute the peak deformation. to obtain an estimate of the peak value ro of the total response. 4.6 and 7. 4.18).4. e. details are available in Appendix B. although somewhat intuitive for inelastic buildings. (4. (4.g. Dn . plastic hinge rotations.3. As shown in Sections 3.2. as described in Section 4. Calculate the peak roof displacement urno associated with the nth-“mode” inelastic SDF system from Eq.8).n t . Eq. However. The peak “modal” responses rno .. of the nth-“mode” inelastic SDF system (Fig.6).3b) with force-deformation relation of Fig. This pushover analysis. alternatively. 4.8). we will refer to rno as the peak “modal” response even in the case of inelastic systems. (4.12). This application of modal combination rules to inelastic systems obviously lacks a theoretical basis. 6. it seems reasonable because it provides results for elastic buildings that are identical to the well-known RSA procedure described in Section 3. are combined using an appropriate modal combination rule. Steps 1 to 4 of the MPA are same as those for UMRHA. and (2) it provides the exact modal response for elastic systems. 2001. the lateral force distribution used possesses two properties: (1) it appears to be the most rational choice among all invariant distribution of forces. Sections 7. 5. (3. (3. bg bg bg bg Thus.1 Summary The peak inelastic response of a building to earthquake excitation can be estimated by the MPA procedure just developed. 37 . At this roof displacement.21). each determined by one pushover analysis. rno also represents the exact peak value of the nth-mode contribution rn t to response r t .solving Eq.2.3. etc. seems reasonable. The response value rno is an estimate of the peak value of the response of the inelastic system to p eff.6b by solving Eq. it can be determined from the inelastic response (or design) spectrum (Chopra. for elastic systems.2 and 3. It provides results for elastic buildings that are identical to the well-known RSA procedure (Section 3. story drifts. the pushover analysis provides an estimate of the peak value rno of any response rn t : floor displacements. governed by Eq.3) because.

subtract the yield value of hinge rotation to determine the hinge plastic rotation. Figure 4.3 cm. 9.8. and three “modes”.1). 4. ur 2o = 11. (3. in particular. 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. Determine the total response by combining the peak “modal” responses using the SRSS combination rule of Eq. is presented next. rno . with. the 38 .7. although the trends are not systematic as when the system remained elastic (Section 3. and the “exact” response from nonlinear RHA for the roof displacement and top-story drift. At urno .18). respectively. and compared with the results of a rigorous nonlinear RHA using the DRAIN-2DX computer program. the peak roof displacement due to each of the three “modes” is ur1o = 48. the combined response due to three “modes”. in contrast to modal analysis (Section 3. say. This is to be expected. and three modes are compared with the “exact” values in Fig.53 cm. This deficiency also implies that. the first two or three “modes” will suffice. Repeat Steps 3 to 8 for as many “modes” as required for sufficient accuracy. The peak values of floor displacements and story drifts including one.7 shows the individual “modal” responses.0.4. From the total rotation of a plastic hinge.5 ¥ the El Centro ground motion including the response contributions associated with three “modal” inelastic SDF systems. and the errors in the approximate results are shown in Fig.4. Observe that errors tend to decrease as response contributions of more “modes” are included. 4. two. 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.9.1 Uncoupled Modal Response History Analysis The structural response to 1. The peak values of response are as noted. and ur 3o = 2. The peak values of displacements of all floors and drifts in all stories are presented in Tables 4. and the percentage errors in the approximate results. determined by the UMRHA procedure. the “exact” results. also included are the combined responses due to one. three “modes” included.7 cm. Typically. extract from the pushover database values of other desired responses.2).1 and 4. 8. two.2.0 to 3. the UMRHA procedure lacks a rigorous theory.

1 and 4. 0. Response histories of roof displacement and top-story drift due to 1. sec 25 30 • 48.44 • "Mode" 2 r1 0 • 48.1 −12 12 ∆r (cm) u (cm) UMRHA 3 "Modes" 0 • 7.1 and 3.2) if the system yields significantly versus if the system remains within the elastic range (Tables 3.2).3 11. and 3.7 • −80 80 ∆r1 (cm) u (cm) "Mode" 1 −12 r2 0 ∆r2 (cm) 12 0 u (cm) −80 80 −12 r3 0 • 2. the errors in responses computed by UMRHA including three “modes” relative to the “exact” response were determined.0. in particular.0.75.6 5 10 15 20 Time. the errors in story drifts are larger compared to floor displacements.25. just as for elastic systems.25 (Tables 3. 0.88 u (cm) "Mode" 3 −80 80 0 −80 80 0 −80 0 • 44.62 5.5 (Tables 4.1 and 4. 1.0.24 • NL−RHA r −12 12 ∆r (cm) u (cm) 0 5 10 15 20 Time. 1. 2. defined as the El Centro ground motion multiplied by 0. for a fixed number of “modes” included.2) and 1.5. Next. 0. and total response from NL-RHA 39 . the degree to which the system deforms beyond its elastic limit. For this purpose the UMRHA and exact analyses were repeated for ground motions of varying intensity. recall that the computed errors have been presented earlier for ground motion multipliers 0. we investigate how the errors in the UMRHA vary with the deformation demands imposed by the ground motion.85.53 ∆r3 (cm) 12 0 • 2.7. For each excitation. sec 25 30 r −12 0 Fig.2).1 and 3.38 6. 4. (a) Roof Displacement (b) Top Story Drift 80 12 0 • 3.response is much less accurate (Tables 4.5. However.5 × El Centro ground motion: individual “modal” responses and combined response from UMRHA.

For this purpose.5 Fig.9. or three “modes” for 1.8.5 ¥ El Centro ground motion Figure 4.5 × El Centro ground motion (a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th (b) Story Drift Ratios Floor 4th 3rd 2nd 1st UMRHA 1 "Mode" 2 "Modes" 3 "Modes" Floor 5th 5th 4th 3rd 2nd 1st UMRHA 1 "Mode" 2 "Modes" 3 "Modes" Ground −60 −40 −20 0 Error (%) 20 40 60 Ground −60 −40 −20 0 Error (%) 20 40 60 Fig. it will be useful to know the deformation of the system relative to its yield deformation.(a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th NL−RHA UMRHA 1 "Mode" 2 "Modes" 3 "Modes" (b) Story Drift Ratios Floor Floor 5th 4th 3rd 2nd 1st 5th 4th 3rd 2nd 1st NL−RHA UMRHA 1 "Mode" 2 "Modes" 3 "Modes" Ground 0 0. the pushover curves using force distributions s* [Eq. Shown is the error in each floor displacement (Fig.5 Displacement/Height (%) 2 Ground 0 0. (3. 4. 4.5 1 1.10b).11. To interpret these results. Heightwise variation of floor displacements and story drift ratios from UMRHA and NL-RHA for 1. Heightwise variation of error in floor displacements and story drifts estimated by UMRHA including one.10a). 4. 4. in each story drift (Fig.5 1 1. Two versions of the pushover curve are 40 .10 summarizes the error in UMRHA as a function of ground motion intensity. two.20)] for the first three n modes of the system are shown in Fig. indicated by a ground motion multiplier. and the error envelope for each case. 4.5 Story Drift Ratio (%) 2 2. with the peak displacement of each “modal” SDF system noted for each ground motion multiplier.

(3) maintain roughly similar values for more intense ground motions. the values determined by RHA of the nth-mode inelastic SDF system (Fig.3). and 2.20) with n = 1. (3. Additional errors are introduced in UMRHA of systems responding beyond the linearly elastic limit for at least two reasons.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. and plastic hinge rotations at the external beam end at each floor level (Table 4. and 3 (Fig. Second. drifts in all stories (Table 4.7 cm. 41 . 2. the peak values of displacements at all floors (Table 4. even though the system remains essentially elastic.4.75. Figure 4. as mentioned in Section 4. the errors in truncating the higher mode contributions are negligible.3 cm. (4. and (4) are larger in story drift compared to floor displacements. 11.75. as mentioned in Section 3. First.4) to roof displacements urno = 48. respectively.10) for each “modal” inelastic SDF system (Figs 4. the first reason mentioned above seems to be the primary source for the errors. 4. UMRHA lacks a rigorous theory and is based on several approximations.2. The location of plastic hinges and their rotations. 4. and.included: the actual curve and its idealized bilinear version. the pushover curve for each “mode” is idealized by a bilinear curve in solving Eq. Each of these three pushover analyses provides the pushover curve (Fig. The system remains elastic up to ground motion multiplier 0.5). were noted but are not shown here.5 ¥ the El Centro ground motion.7). Perhaps this explains why the errors are large at this excitation intensity.53 cm. 3.0.11a). 4.10 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. determined from “exact” analyses.11).0. 4. (2) increase rapidly as the ground motion multiplier increases to 1. The structure is pushed using the force distribution of Eq.4). The idealized curve for the first “mode” deviates most from the actual curve near the peak displacement corresponding to ground motion multiplier 1.2.1). the ductility factor for the first mode system is only 1.6 and 4.4. For more intense excitations.01 (Fig.

575 -41.6 1.379 1.2 22.003 -31.8 14.0 7.3 25.3 42 .407 -10.371 0.136 1.293 1.317 0.7 Table 4.003 0.820 -19.8 0.187 -0.6 4.722 0. Noted 60 50 40 (b) Story Drifts Error Envelope Error for Story No.1 0.616 -0.235 -0.7 31.201 -1.1 1.071 0.2 12.707 1.133 1.5 2 GM Multiplier 2.806 -0.763 -15.478 0.0 1.373 -0.8 1.6 2.938 1.104 0.154 0.5 9.376 1.5 9.220 0.130 0.298 -0.5 28.14 m) from UMRHA for 1.856 2.9 5.138 1.071 -0.983 1. Noted 7 6 Error (%) 30 1 Error (%) 30 5 8 20 6 5 2 20 2 4 9 9 10 0 0 10 3 8 7 3 1 4 0.8 1.3 6.971 1.120 1.900 -10.088 10.490 -1.4 1.371 -0.202 11.214 -0.5 2 GM Multiplier 2.315 -0.291 0.2 Peak values of story drift ratios (as % of story height) from UMRHA for 1.010 0.121 -0.135 9.033 0.914 -0.2 6.5 28.2 4.2 1.676 0.5 10.484 0.376 -1.(a) Floor Displacements 60 50 40 Error Envelope Error for Floor No.772 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 1.3 1.260 -15.490 1.220 -0.3 8.863 0. 4.366 -0.079 0.472 1.057 -0.241 1.298 0.5 1 1.540 0.1 Peak values of floor displacements (as % of building height = 37.4 4.372 1.554 1.698 1.751 1.065 0.9 31.216 1.942 1.430 1.8 1.3 0.0 -9.1 1.9 16.9 31.1 3.9 12.7 14. Errors in UMRHA as a function of ground motion intensity: (a) floor displacements.5 -3.495 1.256 1.5 3 0 0 0.072 -1.283 1.4 -1.333 0.226 -0.201 1.372 -1.072 1.126 0.410 -1.338 1.4 -7.5 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Displacement /Height (%) “Modal” Response Combined (UMRHA) “Mode” 1 “Mode” 2 “Mode” 3 1 “Mode” 2 “Modes” -0.370 -0.169 0.9 12.806 0.096 0.877 0.241 -1.0 11.663 0.513 0.044 1.914 2.214 0.2 1. and (b) story drifts Table 4.8 1.982 9.070 1.256 -1.018 0.727 1.055 -0.5 1 1.338 -1.526 -0.811 1.366 0.5 18.068 0.513 -0.5 × 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 -1.200 8.0 11.0 2.938 -1.201 -0.863 1.852 1.426 -1.5 3 Fig.256 1.820 -0.410 1.473 -22.942 -0.350 -0.248 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.819 2.049 -0.668 -23.844 -25.663 -0.10.8 1.009 -0.945 -37.1 8.0 9.

4). Pushover analysis seems to be inherently limited in computing accurately hinge plastic rotations. and the errors in these estimates relative to the exact response from nonlinear RHA. but was more successful when the excitation was more intense (results are not included).4. Furthermore.13 with Fig. Obviously.3 and 4.” respectively. As shown in Figs. 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%.2). but the third “mode” contributions do not seem especially important (Fig. and three “modes. Fortuitously.5 present estimates of the combined response according to Eq.12 and 4. the errors in the modal pushover results are. this error is not especially significant because the hinge plastic rotation is very small. 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.12 and Tables 4. However. considering one.13 and Tables 4. MPA failed to identify the plastic hinges at the column bases in Fig. 4. the hinges identified in beams at the 6th floor were at variance with the “exact” results.Figures 4. 4. (3.13 and Tables 4.3 and 4.” two “modes. significant improvement is achieved by including response contributions due to the second “mode”.4 with Tables 4. The first “mode” alone is inadequate. overestimated by up to 18% in the middle stories.” and three “modes. significantly smaller than in UMRHA (compare Fig. The locations of plastic hinges shown in Fig. for two or three modes included. 4.11c). The errors are especially large in the hinge plastic rotations estimated by the MPA procedure. although the error is recorded as 100% if MPA estimates zero rotation when the nonlinear RHA computes a non-zero value.4). the third “mode” does not contribute because this SDF system remains elastic (Fig. 4.13 and Table 4.3 through 4. two.3 and 4.” and nonlinear RHA. One “mode” pushover analysis was unable to identify the plastic hinges in the upper stories where higher mode contributions to response are known to be more significant. 4.14.5). in general.1 and 4. 4. 4.3 and 4. the results were not always accurate.18). 43 . 4. and are within a few percent of the exact values for the upper stories. however.14. The drifts are underestimated by up to 13% in the lower stories. 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. For example. especially in estimating the story drifts (Fig.12 and Tables 4.9 and Tables 4. were determined by four analyses: MPA considering one “mode. even if three “modes” are included (Fig. The second “mode” was necessary to identify hinges in the upper stories.

(a) "Mode" 1 Pushover Curve 12000 10000 u = 36.5 0.5.25.75 Actual Idealized 20 40 60 Roof Displacement (cm) (b) "Mode" 2 Pushover Curve 80 12000 10000 u = 9. V = 5210 kN.14 y by Base Shear (kN) 8000 6000 4000 1.5 0.6 cm. “Modal” pushover curves with peak roof displacements identified for 0.0 × El Centro ground motion 44 .75.75 0.75 0.0. 4.85 0. 0. α = 0. α = 0.25 1 2 4 6 Roof Displacement (cm) 8 10 Fig.11.9 cm.25 0. V = 4952 kN.85 0. α = 0.2 cm.5 2 3 Actual Idealized 2000 0 0 0. 0.85 Base Shear (kN) 8000 6000 4000 2000 0 0 0. V = 7616 kN. 2.25 5 10 15 20 Roof Displacement (cm) 25 (c) "Mode" 3 Pushover Curve 12000 10000 u = 4.5 1 0.13 y by Base Shear (kN) 8000 6000 1 1.19 y by 3 2 1.0. 1.5 0.5. and 3.5 2 3 4000 2000 0 0 Actual Idealized 0. 1.

now standard in engineering practice.15 summarizes the error in MPA considering three “modes” as a function of ground motion intensity.15).4.4 also apply to MPA.5 1 1. as discussed in Sections 3.3. Shown is the error in each floor displacement (Fig. UMRHA is essentially exact. 4. the MPA procedure is equivalent to the RSA procedure. 4. identified in Section 3. implying excitations weak enough to limit the response in the elastic range of the structure. whereas MPA contains errors inherent in modal combination rules. 4.15) for ground multipliers larger than 1. 45 .2 and 3.5 Story Drift Ratio (%) 2 2. The fact that MPA is able to estimate the response of buildings responding well into the inelastic range to a similar degree of accuracy indicates that this procedure is accurate enough for practical application in building retrofit and design.5 × El Centro ground motion.15b).5 Fig. each story drift (Fig. As mentioned in Section 3. implying excitations intense enough to cause significant yielding of the structure. implying that the modal combination errors contained in these procedures are acceptable. indicated by a ground motion multiplier.10 and 4.3 for elastic systems (or weak ground motions).0. In this case. 4. 4. While various sources of errors in UMRHA.75. an observation with practical implications. the errors in MPA were fortuitously smaller than in UMRHA (compare Figs. the errors in MPA were larger for ground motion multipliers less than 0.5 Displacement/Height (%) 2 Ground 0 0. shading indicates errors in MPA including three “modes” Figure 4. However.4.(a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th NL−RHA MPA 1 "Mode" 2 "Modes" 3 "Modes" (b) Story Drift Ratios Floor Floor 5th 4th 3rd 2nd 1st 5th 4th 3rd 2nd 1st NL−RHA MPA 1 "Mode" 2 "Modes" 3 "Modes" Ground 0 0. The errors are only weakly dependent on ground motion intensity (Fig. and the error envelope for each case.5 1 1.12 Heightwise variation of floor displacements and story drift ratios from MPA and NL-RHA for 1.15a).

(a) Floor Displacements 9th 8th 7th 6th Floor 5th 4th 3rd 2nd 1st MPA 1 "Mode" 2 "Modes" 3 "Modes" Ground −60 −40 −20 0 Error (%) 20 40 60 (b) Story Drift Ratios 9th 8th 7th 6th Floor 5th 4th 3rd 2nd 1st MPA 1 "Mode" 2 "Modes" 3 "Modes" Ground −60 −40 −20 0 Error (%) 20 40 60 (c) Hinge Plastic Rotations 9th 8th 7th 6th Floor 5th 4th 3rd 2nd 1st MPA 1 "Mode" 2 "Modes" 3 "Modes" Ground −120 −80 −40 0 Error (%) 40 80 120 Fig. and hinge plastic rotations estimated by MPA including one.5 ¥ El Centro ground motion 46 . story drifts. two. and three “modes” for 1. 4. Errors in floor displacements.13.

176 0.8 -6.14 m) from MPA for 1.72E-03 7.2 -3.6 -7.018 -0.844 -7.2 6.611 0.36E-03 6.8 -29.168 -0.60E-03 2.37E-03 1.9 5.066 -0.6 -44.055 0.581 0.76E-03 4.00E+00 0.53E-03 7.895 1.015 0.0 3 “Modes” -32.804 1.55E-03 3.88E-03 1.1 62.130 0.1 62.641 1.305 -0.752 1.222 0.1 18.414 1.640 1.4 1.581 0.00E+00 1 “Mode” 7.1 62.9 0.0 2 “Modes” -32.652 1.10E-02 9.9 1.2 1.781 0.0 15.02E-03 0.5 × El Centro ground motion Hinge Plastic Rotation (rad) “Modal” Response Combined (MPA) “Mode” 1 7.00E+00 3 “Modes” 7.0 1.666 Table 4.012 1.503 -1.900 -0.738 1.8 1.2 -100.910 1.37E-03 1.1 46.00E+00 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th “Mode” 2 0.071 -0.5 Peak values of hinge plastic rotations (radians) from MPA for 1.4 Peak values of story drift ratios (as % of story height) from MPA for 1.0 -50.304 -1.135 -7.8 7.00E+00 0.8 -29.233 1.05E-03 2.2 -4.Table 4.879 1.5 2.00E+00 0.18E-03 7.37E-03 1.00E+00 0.125 -1.5 1.745 1.945 -49.5 -6.22E-10 NL RHA 1.895 1.1 -8.00E+00 0.2 0.527 -0.5 Table 4.3 -3.202 8.244 0.37E-04 Error (%) “Modal” Response 1 “Mode” -32.7 -12.259 1.668 -13.733 1.980 0.9 -100.7 1.6 1.5 7.057 0.0 47 .9 2.0 -100.72E-03 7.3 13.517 1.209 1.911 0.00E+00 0.068 0.55E-03 3.399 0.00E+00 0.088 12.8 0.116 1.426 15.756 0.72E-03 7.2 0.5 × 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” 1.76E-03 4.3 1.575 -53.36E-03 6.514 -1.00E+00 0.007 1.00E+00 0.116 1.00E+00 0.1 46.9 -100.250 0.407 -27.331 1.8 -4.9 7.399 0.8 “Mode” 3 -1.071 0.311 0.02E-03 3.190 -0.942 6.266 -0.9 -4.018 -0.5 2.36E-03 6.049 -0.7 1.200 8.8 17.066 -0.982 13.705 1.667 1.018 0.0 -50.50E-10 3.640 -1.033 -0.5 10.667 -1.0 -5.503 1.737 0.55E-03 3.8 1.9 1.371 -0.00E+00 “Mode” 3 0.154 0.705 -1.4 -8.516 0.6 13.338 1.37E-03 1.728 1.737 1.6 -8.220 1.614 0.6 13.233 1.02E-03 3.8 -29.88E-03 0.473 -15.26E-04 9.00E+00 0.00E+00 3.00E+00 0.72E-03 7.60E-04 7.781 0.00E+00 0.2 -100.3 Peak values of floor displacements (as % of building height = 37.00E+00 0.222 0.00E+00 0.304 1.6 7.336 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.00E+00 2 “Modes” 7.6 -44.756 0.8 -12.0 -100.00E+00 0.116 1.518 1.007 1.015 0.315 0.372 0.260 -14.3 11.003 -16.478 0.156 -0.1 46.351 -0.2 11.00E+00 0.8 -6.6 -9.687 0.76E-03 4.1 13.009 0.435 0.197 -0.3 -3.498 1.694 1.02E-03 0.00E+00 0.118 0.36E-03 6.00E+00 0.105 0.5 2.19E-10 3.298 -0.053 1.429 0.88E-03 0.99E-03 6.594 -1.0 -100.298 0.76E-03 4.101 -0.980 -0.2 -100.1 1.5 10.220 1.683 1.820 -7.5 × 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 0.053 -1.763 -14.

20) with n = 1 (Fig. 4.6 through 4.16a—which is slightly different than the one excluding gravity loads (Fig.7. 4. the results presented so far did not include gravity load effects.3 Modal Pushover Analysis with Gravity Loads To evaluate the accuracy of the dynamic response of the system.4.53 cm.8 present estimates of the combined response according to Eq.7). As shown in Fig.4. 4. 4. They are now included in the pushover analysis of the structure for the first “mode” only.” and the errors in these estimates relative to the exact response from nonlinear RHA. (3.6 and 4.17 and Tables 4.5). Starting with its initial state under gravity loads. (3. 3. Obviously gravity load effects will influence the seismic demands due to the first “mode. and plastic hinge rotations at selected external beam end at each floor level (Table 4. two.20) with n = 1 to a target roof displacement.” however. Significant improvement is achieved by including response contributions due to the second “mode. Results of such analyses for the selected building subject to 1. Figures 4.8). 4.6).5 × El Centro ground motion are presented below.6 and 4. and three “modes. and overestimated by up to 29% in the middle stories and by up to 13% in the upper stories. the structure is pushed using the force distribution of Eq.7).7 cm and ur 3o = 2.7). 4.11a). The first “mode” alone provides adequate estimates of floor displacements. The response contributions of the second and third modes are the same as before (Tables 4.16 are unchanged. MPA including two “modes” underestimates the displacements of lower floors by up to 6% and overestimates the upper floor displacements by up to 22%.17 and 4. drifts in all stories (Table 4.4. 4. but it is inadequate in estimating the story drifts (Fig. Static analysis of the structure for gravity loads provides the initial state—forces and deformations—of the structure. resulting in the pushover curve shown in Fig. (3. as are the roof displacements ur 2o = 11.” but not the contributions of higher “modes. the third “mode” contributions do not seem especially important (Fig.17 and Tables 4.18 and 48 . The story drifts are underestimated by up to 7% in the lower stories.” these effects will modify the combined modal response as well as the results of nonlinear RHA.3 .4) to roof displacement ur1o = 52. Also presented at these roof displacements are the displacements at all floors (Table 4.18 and Tables 4.0 cm. The pushover curves for the second and third modes included in Fig. and the structure is pushed using the force distribution of Eq.18) considering one. The errors are especially large in the plastic hinge rotations estimated by the MPA procedure even if three “modes” are included (Fig.6 and 4.18 and Tables 4.

Noted 60 50 40 (b) Story Drifts Error Envelope Error for Story No. two. Most pushover analysis procedures do not seem to compute to acceptable accuracy plastic hinge rotations.5 1 1. 4.5 3 Fig.Table 4.15.5 3 0. 2−"Modes" •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• • •• (c) 3−"Modes" • • • • • • • •• •• • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• • •• • • • • • • • • • (d) Nonlinear RHA •• •• • •• •• •• •• • •• •• •• •• •• •• •• • •• •• • •• •• •• •• • • • • • • •• •• •• Fig. 1−"Mode" • • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • •• • •• • • • • • •• •• • •• •• •• •• (b) MPA.5 2 GM Multiplier 2.8). (a) MPA.5 1 1. and (b) story drifts 49 . 4. Errors in MPA as a function of ground motion intensity: (a) floor displacements.5 2 GM Multiplier 2. and three “modes” and by NL-RHA for 1. Noted Error (%) 30 20 10 0 0 2 8 3 4 1 9 Error (%) 30 20 5 4 6 5 7 2 9 10 0 0 1 8 7 3 6 0. Locations of plastic hinges determined by MPA considering one.5 × El Centro ground motion (a) Floor Displacements 60 50 40 Error Envelope Error for Floor No.14.

and the errors are only weakly dependent on ground motion intensity.20 summarizes the error in MPA considering three “modes” as a function of ground motion intensity. indicated by a ground motion multiplier. excluding gravity load effects. 4. 4. These errors are only slightly larger than those in Fig.15.19 were determined by four analyses: MPA considering one “mode.” two “modes”.20a). With two modes included in MPA. each story drift (Fig.The locations of plastic hinges shown in Fig. Figure 4. 4. 50 . Shown is the error in each floor displacement (Fig.20b). MPA provides response values accurate enough for practical application in building retrofit or design. The second “mode” is necessary to identify hinges in the upper stories. this procedure is able to predict plastic hinge locations essentially consistent with nonlinear RHA. 4. and nonlinear RHA. One “mode” pushover analysis is unable to identify the plastic hinges in the upper stories where higher mode contributions to response are known to be more significant. and the error envelope for each case.

α = 0. 2. 0.25.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36.9 cm.14 y by Base Shear (kN) 8000 6000 4000 1.25 0. α = 0. 4. noted are peak values of roof displacement for 0.85 0. V = 4952 kN.19 y by 3 2 1. and 3.0.75.85 0.50.25 1 2 4 6 Roof Displacement (cm) 8 10 Fig.5 1 0.3 cm.75 0.16.5 0.0 × El Centro ground motion 51 .5 0.5 2 3 Actual Idealized 2000 0 0 0. 0. V = 7433 kN. V = 5210 kN.5 0. α = 0. 0.13 y by Base Shear (kN) 8000 6000 1 1.0.85.25 5 10 15 20 Roof Displacement (cm) 25 (c) "Mode" 3 Pushover Curve 12000 10000 u = 4. “Modal” pushover curves with gravity loads included. 1.85 Base Shear (kN) 8000 6000 4000 2000 0 0 0.75 0.6 cm.5 2 3 4000 2000 0 0 Actual Idealized 0.75 Actual Idealized 20 40 60 Roof Displacement (cm) 80 (b) "Mode" 2 Pushover Curve 12000 10000 u = 9.

and hinge plastic rotations estimated by MPA including one. gravity loads included. and 52 .5 × El Centro ground motion.17. 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. 4.5 1 1. Heightwise variation of floor displacements and story drift ratios from MPA and NL-RHA for 1. 4.5 Displacement/Height (%) 2 Ground 0 0. two.5 1 1. story drifts.(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. Errors in floor displacements.5 Story Drift Ratio (%) 2 2.18.

5 2 GM Multiplier 2.5 1 1. 4. 1−"Mode" • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • •• •• •• • • • • • •• •• •• •• •• •• •• (b) MPA. and (b) story drifts. gravity loads included (a) Floor Displacements 60 50 40 Error Envelope Error for Floor No. two. 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 0.20.5 1 1. and three “modes” and by NL-RHA for 1.19. 2−"Modes" •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• •• •• (c) MPA.5 3 Fig. Locations of plastic hinges determined by MPA considering one.5 3 0.(a) MPA. gravity loads included 53 .5 2 GM Multiplier 2. 4.5 × El Centro ground motion. Errors in MPA as a function of ground motion intensity: (a) floor displacements. Noted 60 50 40 (a) Story Drifts Error Envelope Error for Story No. 3−"Modes" (d) Nonlinear RHA • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• •• •• • • • • • • • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • • • • • •• •• •• Fig.

13E-03 5.068 0.0 -100.353 -23.5 × El Centro ground motion.2 16.8 28.7 1.190 -0.6 19.033 -0.530 1.2 9.728 1.066 -0.434 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.125 -1.2 -2.057 0.9 31.105 0.2 4.266 -0.4 -4.00E+00 0.399 -0.454 1.015 0.2 4.00E-03 5.3 13.257 0.927 1.5 × El Centro ground motion.850 1.603 -1.4 -6.754 1.23E-03 0.7 19.854 0.00E+00 0.19E-04 5.414 28.4 1.102 1.237 0.665 0.6 1.88E-03 0.Table 4.434 0.998 1.071 -0.7 4.1 1.5 10.434 0.466 0.1 4.311 0.372 0.263 0.00E+00 “Mode” 2 0.00E+00 0.637 0.8 1.305 -0.2 -3.530 1.2 2.515 -50.102 1.436 1.813 1.107 1.00E+00 0.018 -0.19E-03 1.429 -1.00E+00 “Mode” 3 0.14 m) from MPA for 1.6 2.55E-03 3.5 1.461 0.26E-03 3.00E+00 0.128 -1.04E-02 8. 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” 0.11E-03 9.908 -1.821 1.996 -0.23E-03 3.330 1.009 0.998 21.3 1.8 -32.00E+00 0.3 -22.913 7.3 -100.667 0.5 1.11E-03 9.176 0.214 0.0 -100.858 2.2 0.055 0.637 0.130 0.5 21.9 -6.7 -2.933 1.00E+00 2 “Mode” 8.35E-03 8.921 1.821 -1.00E-03 5.11E-03 9.75E-03 0.783 1.673 Table 4.00E+00 0.516 0.00E+00 0.310 1.6 2.098 20.88E-03 1.3 -22.00E+00 0.888 1. 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” 1.836 -0.7 Peak values of story drift ratios (as % of story height) from MPA for 1.2 21.860 1.371 -0.830 -12.23E-02 1.854 0.049 -0.8 0.351 -0.7 16.594 -1.686 -7.8 Table 4.2 1.831 0.04E-10 3.00E+00 Combined (MPA) 1 “Mode” 8.55E-03 3.00E-03 5.00E-10 NL RHA 1.109 0.213 1.23E-03 0.4 26.8 -32.5 3 “Modes” -32.3 -100.270 -12.0 16.315 0.6 0.399 0.4 -4.00E+00 0.5 0.8 0.9 -3.822 1.2 4.3 -22.78E-03 1.831 0.35E-03 8.490 -11.6 Peak values of floor displacements (as % of building height = 37.744 1.037 0.983 1.5 -5.319 1.168 -0.0 37.4 1.00E+00 0.00E+00 0.850 -1.8 Peak values of hinge plastic rotations (radians) from MPA for 1.1 21.0 2 “Modes” -32.037 -0.938 1.199 16.88E-03 0.996 0.207 18.00E-03 5.5 21.319 1.0 37.17E-03 9.00E+00 0.507 1.9 -4.114 1.00E+00 0.197 -0.19E-03 1.687 0.754 0.066 -0.953 15.983 1.2 -0.35E-10 3.0 37.00E+00 0.877 -46.8 9.00E+00 Error (%) “Modal” Response 1 “Mode” -32.5 “Mode” 3 -1.2 9.3 9.114 -1.11E-03 9.156 -0.429 1.071 0.7 -2.064 -10.0 -30.00E+00 3.00E+00 3 “Modes” 8.5 54 .527 -0.1 13.530 1.19E-03 1.2 9.00E+00 0.19E-03 1.213 1.23E-03 3.213 1.55E-03 3.4 20.00E+00 0.5 × El Centro ground motion.35E-03 8.154 0.330 1.00E+00 0.3 -100.478 0.998 0. gravity loads included Hinge Plastic Rotation (rad) Floor “Modal” Response “Mode” 1 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 8.00E+00 0.35E-03 8.908 1.237 0.263 1.514 -1.2 12.101 -0.603 1.0 -30.2 1.

j 2.1 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. joint rotations. Comparison of Modal and FEMA Pushover Analyses 5.2 COMPARATIVE EVALUATION Compared in this section are the earthquake-induced demands for the selected building determined by five analyses: pushover analysis using the three force distributions in FEMA-273.5 sec . and the exponent k = 1 for fundament period T1 £ 0. plastic hinge rotations. 2… N ). 1997] The pushover curve.” and nonlinear RHA. Specified in FEMA-273 are three distributions for lateral forces: 1. etc. Equivalent lateral force (ELF) distribution: s* = m j h k where h j is the height of the j j jth floor above the base. assumed to be linearly elastic. “Uniform” distribution: s* = m j (where the floor number j = 1. gravity load effects were included in all 55 . 5. k = 2 for T1 ≥ 2. computed at the target displacement represent the earthquakeinduced demands on the structure. and 3.5.5 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. SRSS distribution: s* is defined by the lateral forces back-calculated from the story shears determined by response spectrum analysis of the structure. The floor displacements. and varies linearly in between.. MPA considering three “modes. a plot of base shear versus roof displacement. story drifts.

062 0.0197 0.11 0. Force distributions in FEMA-273: (a) “uniform”.2.21 0.4 and Tables 5.0466 0.165 0.11 0.281 0.3. The pushover curves are given in Fig.3. Using each of these force distributions. and Table 5. 5.11 0. 0.11 0.3a and 5.3.2.1. The “uniform” distribution overestimates all floor displacements by 17-28%.1 through 5. with the errors being larger for the SRSS distribution.5 times the El Centro ground motion. The first four floor displacements are within 5% of the “exact” value. 5. the value determined from RHA of the first-mode inelastic SDF system for 1. (b) ELF. 5.0702 0.0981 0.0654 0.126 0. The errors in the FEMA and MPA estimates of seismic demands relative to the “exact” demands are presented in Fig. 5. pushover analyses are implemented for a target roof displacement of 52.0 cm.3b and Table 5. the story drift demands in Fig. wherein the first two are obvious and the third is determined from response spectrum analysis of the building (Appendix C).112 0. The MPA procedure is more accurate than all the FEMA force distributions.177 0.analyses.11 0. Figures 5. and the locations of all plastic hinges in Fig.4. 5. 5.3a and Table 5.1.00719 0.11 0.11 0. both presented in Section 4. the floor displacement demands in Fig.0381 0. and the displacements of upper floors are overestimated by 9-22%.0913 0.4.367 (a) Uniform (b) ELF (c) SRSS Fig.0896 0.042 0.4a.119 0. The three FEMA force distributions are presented in Fig. 5. Also included in these presentations are the MPA results considering three “modes” and the “exact” demands from nonlinear RHA.1.0446 0. and (c) SRSS 56 . plastic hinge rotation demands in Table 5.1 demonstrate that the displacement demands are underestimated by the ELF and SRSS force distributions by 12 to 30 % at the lower six floors of the building.

(a) Uniform Distribution 12000 10000 u = 33 cm. V = 7456 kN. α = 0. V = 6897 kN. and (c) SRSS. gravity loads are included 57 . V = 8530 kN. 5. (b) ELF.8 cm.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. Pushover curves using three force distributions in FEMA-273: (a) “uniform”.2. α = 0.3 cm. α = 0.14 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized 10 20 30 40 Roof Displacement (cm) (b) ELF Distribution 50 60 12000 10000 u = 38.24 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized 10 20 30 40 Roof Displacement (cm) 50 60 Fig.

3b and 5.) The pushover analysis procedures considered seem incapable of computing accurately local response quantities. Figure 5. 3 "Modes" FEMA Uniform ELF SRSS 0. with errors reaching 37% in this example. For the SRSS distribution.5 Ground 0 Ground 0 Fig. and Table 5.3. Heightwise variation of floor displacements and story drift ratios estimated using FEMA-273 force distributions.5 1 1. with story drifts under estimated by. 5. reaching 64%. the MPA procedure is more accurate than all the FEMA force distributions.Figures 5. whereas nonlinear RHA computed an insignificantly small value. (a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th (b) Story Drift Ratios Floor 4th 3rd 2nd 1st NL−RHA MPA. 3 "Modes" FEMA Uniform ELF SRSS 0. such as hinge plastic rotations. larger errors are noted in the upper and lower stories. (The 100% error in a 6th floor hinge is ignored as it simply represents that the MPA estimated zero rotation. reaching 35%. gravity loads included 58 . 7%.2 demonstrate that the story drift demands are greatly underestimated by all the FEMA force distributions. and overestimated by no more than 32%. In contrast. This seems to be an inherent limitation of pushover analysis. errors are largest in the upper stories.5 Story Drift Ratio (%) 2 2. but it is still inaccurate.” and NLRHA. For the ELF distribution. reaching 31%. Modal pushover analysis procedure gives estimates better than all the FEMA force distributions. the errors are largest in the lower stories. at most.4b.3 demonstrate that the hinge plastic rotations estimated by all three FEMA force distributions contain unacceptably large errors.5 Displacement/Height (%) 2 Floor 5th 5th 4th 3rd 2nd 1st NL−RHA MPA. For the uniform distribution.5 1 1. MPA including three “modes.4c and Table 5.

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

The MPA procedure identifies yielding in most locations predicted by “exact” analysis. especially in estimating story drifts.6 and 5.to identify yielding in some locations. Included for comparison is the error in MPA with three “modes. and the error envelope for each case.” The MPA procedure provides estimates of earthquake demands that are significantly more accurate than all FEMA273 analyses. Figures 5. but fails to predict yielding in a few locations. Shown is the error in each floor displacement and each story drift. 60 .7 summarize the error in FEMA analyses relative to the “exact” demands as a function of ground motion intensity indicated by a ground motion multiplier. The MPA procedure is superior to FEMA-273 analyses over the entire range of ground motion intensities considered.

7 7.736 0.611 0.672 1.067 0.417 1.8 -63.839 0.547 -27.00E-10 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 1.436 Error (%) NL Uniform RHA 0.061 1.562 1.65E-03 7.234 1.2 4.168 1.836 0.9 16.00E-03 5.4 MPA -2.2 19.8 2.09E-03 4.0 -57.178 1.953 0.2 6.0 -100.109 1.00E+00 4.783 1.75E-03 0.207 1.5 -27.34E-03 2.6 -6.2 Peak values of story drift ratios (as % of story height) from FEMA force distributions and MPA.4 0.23E-03 3.4 -55.5 -29.306 1.4 -23.0 -100.462 1.314 1.16E-03 0.998 1.294 1.730 1.94E-03 2.84 1.2 13.1 163.9 -70.998 1.7 FEMA ELF SRSS -22.35E-03 8.098 1.3 -11.17E-03 9.351 0.011 1.8 -100.154 1.59E-03 5.9 11. Peak values of floor displacements (as % of building height = 37.9 SRSS -22.515 FEMA Uniform ELF 27.4 21.888 0.7 Table 5.5 -33.6 21.195 0.51E-03 4.3 31.3 Peak values of hinge plastic rotations (radians) from FEMA force distributions and MPA Hinge Plastic Rotation (rad) MPA FEMA ELF SRSS 4.860 1.2 16.530 1.9 0.7 -60.3 -22.5 -33.5 Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 2.875 0.686 0.0 Table 5.9 -20.0 -30.00E+00 Error (%) FEMA Uniform ELF SRSS MPA 24.270 0.13E-03 5.530 1.877 1.399 1.372 1.323 1.00E+00 2.9 -77.50E-03 0.78E-03 0.8 17.53E-02 1.913 0.490 0.623 1.52E-04 1.2 -4.263 0.221 1.5 -32.466 0.2 -12.7 19.262 1.3 -4.854 0.1 -26.597 0.00E+00 0.14 m) from FEMA force distributions and MPA.00E+00 0.35E-10 3.00E+00 0.975 1.10E-02 7.5 -29.00E+00 8.6 -73.330 1.089 1.62E-03 0.03E-03 5.88E-03 1.5 15.927 1.00E+00 0.064 1.353 1.0 -100.992 1.3 -14.8 44.015 0.3 -100. gravity loads included Displacement /Height (%) FEMA ELF SRSS MPA 0.8 16.7 -15.00E+00 0.2 9.2 5.19E-04 5.724 0.344 0.9 16.128 1.7 MPA -2.0 37.00E+00 0.199 27.4 -12.560 1.3 23.335 2.8 -32.5 22.8 7.7 21.2 16.8 -35.58E-04 6.1 0.8 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 0.809 0.6 4.6 22.984 1.78E-03 1.0 -59.6 -11.5 10.399 1.26E-03 3.5 -3.2 9.55E-03 3. gravity loads included Displacement /Height (%) FEMA ELF SRSS MPA 1.399 -27.1 10.00E+00 NL RHA 1.667 0.11E-03 9.5 26.1 -3.414 1.23E-02 1.2 10.708 0.367 1.341 1.858 1.6 -25.487 0.007 1.6 -41.2 -100.7 -29.04E-02 8.6 14.45E-03 3.6 -17.93E-03 1.9 15.355 0.789 0.0 -63.673 Error (%) NL RHA 1.1.524 0.6 17.566 1.Table 5.3 29.8 -2.0 -100.4 -50.214 1.9 28.4 16.7 -28.5 61 .830 2.083 1.4 16.318 2.19E-03 1.8 -100.6 -4.938 1.0 -71.310 1.209 0.7 26.

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

5 3 Fig.5 1 1.5 1 1.5 1 1. Noted 2 Error (%) Error (%) 1 2 30 20 10 0 0 1 4 6 7 30 20 10 8 3 5 8 4 9 7 9 6 5 0.5 3 (c) FEMA: SRSS 60 50 40 60 50 40 2 3 1 5 4 (d) MPA (3 "Modes") Error (%) 30 20 10 0 0 8 Error (%) 30 20 6 8 1 5 2 4 3 9 7 6 7 9 10 0 0 0. 5.6. Errors in floor displacements from three force distributions in FEMA-273 and from MPA including three “modes”.5 1 1.5 2 GM Multiplier 2.5 3 0. gravity loads included 63 .(a) FEMA: Uniform 60 50 40 60 50 40 3 (b) FEMA: ELF Error Envelope Error for Floor No.5 2 GM Multiplier 2.5 2 GM Multiplier 2.5 2 GM Multiplier 2.5 3 0 0 0.

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

pushover analysis is inherently limited in the sense that it cannot provide any cumulative measure of response. 65 . Pushover analysis of a one-story inelastic system predicts perfectly peak seismic demands: deformation. now common in structural engineering practice.. which retains the conceptual simplicity and computational attractiveness of the procedure with invariant force distribution. hinge plastic rotation. where m is the mass matrix n of the building and φ n its nth mode. Dn is available from the elastic response (or design) spectrum. SRSS) leads to the modal pushover analysis (MPA) procedure. Combining these peak modal responses by an appropriate modal combination rule (e. 3. e. etc.. joint rotations. This system has vibration properties—natural frequency w n and damping ratio. 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 . where the nature and magnitude of errors arising from approximate modal combination rules are well understood. and the structure is pushed to the roof displacement determined from the peak deformation Dn of the nth-mode elastic SDF system. ζ n —of the nth-mode of the MDF system. It has led to the following conclusions: 1.6 Conclusions This investigation was aimed toward developing an improved pushover analysis procedure based on structural dynamics theory. 2. the energy dissipated in yielding or the cumulative rotation of a plastic hinge. For this system. This MPA procedure for elastic buildings is shown to be equivalent to the standard response spectrum analysis (RSA) procedure.g. However.g.

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

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

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

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

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

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

i i 3.7.4. i i i αn = Vbno Vbny − 1 urno urny − 1 i 3. urny = Vbny kn . Replace i+1 with i and repeat Steps 3.6.3. Abn . From the pushover data.6 . Vbny . i i i 3.8.9.6 × Vbny . kn = 0. Draw the curve OAB by connecting the three points O. Vbny .0. kn .i 3. 3. O. urn.6 . Calculate the slope. and a point on the actual pushover curve with base shear equal to i 0.4 to 3. 3.0. and B with straight-line segments to obtain the idealized bilinear curve. i 3. other appropriate methods can be ( used. Calculate initial slope of the idealized bilinear curve. obtained by judgment.6 × Vbny ( ) i urn . This step gives the secant stiffness at a base shear equal to 60% of the yield base shear.4. i +1 i i 3. Calculate the yield displacement.2. determine the roof displacement. Estimate the yield base shear.5. 72 .8.1. 3. at base shear i equal to 0.9.9. by connecting a straight line between origin.6 × Vbny . Calculate Vbny = Vbny × Apn Abn . Calculate area under the bilinear curve OAB. If the error exceeds some pre-specified tolerance.2. i 3. Calculate the error = 100 × Abn − Apn ( ) ( ) ( ) ) Apn .1. will be refined by an iterative procedure that seeks to equate areas under the actual and the idealized pushover curves. Calculate the post-yielding strain-hardening ratio. This value. corresponding to the estimated yield i i i base shear. iterations are necessary. If desired.4. A. i 3. Let the point with base shear = Vbny and roof displacement = urny be denoted as A.

7. 4. Scale the vertical axis with * Mn to obtain * Fsno Ln = Vbno M n and * Fsny Ln = Vbny M n (Eqs.11a).10a and 4. Vbn Idealization of Pushover Curve Vbno Vbny 0. 8.2.15) and (3.1.0. A.6 ur n y ur n o ur n Fig. (3. Compute the Ln and Γ n from Eq. 4.3. 4. 4. Scale the horizontal axis by Γnφrn to obtain Dno = urno Γ nφrn and Dny = urny Γ n φrn (Eqs. A. Combine the “modal” responses using Eqs. 9. An (t ) .16).4.12) and (3. Calculate histories of various responses using Eqs. r o .6 × Vbny A • Idealized αnkn B • 1 Actual k 1 • O n figA_1.eps ur n.4) and effective modal mass from M n = Ln Γn . Compute deformation history. (3. 6. 4.1 Idealization of nth-“mode” pushover curve 73 . (3. In general first two or three modes will suffice.10b and 4. Calculate peak values. Dn (t ) .11b). * 4. A. Develop the Fsn Ln − Dn relation (Fig. and pseudo-acceleration history.13). of the combined responses obtained in Step 8.2). 5. of the nth“mode” inelastic SDF system (Fig.2.3b) with unit-mass and force-deformation relation of Fig. Repeat Steps 3 to 6 for as many modes as required for sufficient accuracy.

6 kN. second. The following steps illustrate the procedure to develop the idealize curve for the first “mode”. respectively. 25. Ap1 = 360777 kN-cm. for the first.eps Dny = ur n y / Γnφr n Dno = ur n o / Γnφr n Dn Fig. 3. ur1o = 63.2 EXAMPLE The UMRHA procedure is implemented to calculate the response of the 9-story building described in Section 3.3. n 2. Idealized bilinear curves for each of the three modes are included in Fig. The force distributions. computed for the first three modes are shown in Fig.3. The base-shear – roof-displacement (Vbn . and 12. A. The pushover curves for the first three modes.). Following is step-by-step implementation of the procedure described in Section A. and third mode. are shown in Fig. 2. First three mode shapes and frequencies of the selected building were computed and are shown in Fig.3. The anchor point.1. Area under the actual pushover curve. 3.1.2.4 cm (10 in.4.2 Properties of the nth-“mode” inelastic SDF system A. s* . 3. A. At this point.Fsn / Ln Fsn / Ln − Dn Relationship Vbno / M* n Vbny / Mn * 1 α ω2 n n ω2 1 n figA_2.urn ) pushover curve for the force distribution s* : n 2.4.2. is defined at the target roof displacement.7 cm (5 in.5 cm and Vb1o = 8729.).). B. 74 . 3.5.1. A. 3. 1.5 cm (25 in. to the north-south component of the El Centro (1940) ground motion scaled up by a factor of 1. Target displacements used to generate these pushover curves are 63. generated using DRAIN-2DX.

i i i 3.6 8006.4.6 × Vb1 y ( ) i ur1.2. α1 = Vb1o Vb1 y − 1 ur1o ur1 y − 1 = ( ) ( ) (8729. Also included are the modal damping ratios and the periods calculated from Eq. and M1 = 2736789 × 1. The results for other modes are summarized in Table A. is calculated as follows. The first estimate of the yield base shear Vb1 y = 8006.9 kN.9. Therefore.5 38. This value exceeds the prespecified tolerance of 0.5.194.198%.0. * 4. iterations are necessary. The curve OAB obtained by connecting the three points O.3666. 4. Error = 100 × (365100 − 360777 ) 360777 = 1. 3.86 = 210. Determined from the pushover database. The next estimate of the yield shear is Vbny = 8006.i 3. i +1 3.01%. i 3. i 3. Γ1 = 1. and B with straight-line segments gives the idealized bilinear curve. i i i 3.9.86 cm at 0. ur1.13).7.0.09 cm and Vb1 y = 8006.3666 = 3740189 kg. k1 .1. The Fs1 L1 − D1 relation for the first “mode” is developed as follows. The post-yielding strain-hardening ratio.6.18 kN/cm. The point A on the i i bilinear curve is defined by ur1 y = 38. i i 3.6 × Vb1 y = 4803. Vb1 y = 7615. (4. L1 = 2736789 kg. A. and α1 = 0.4.2.8 22.4 210.3. Area under the bilinear curve OAB.1.4 kN. The procedure converged after nineteen cycles to give ur1 y = 36. The yield displacement.135. i i 3.9. 3.09 cm. 75 . The results of the iterative procedure are summarized in Table A.1.18 = 38. k1 = 0.4 × (360777 365100 ) = 7911.09 ) − 1 = 0.8 kN. Ab1 = 365100 kN-cm.4 ) − 1 (63. 3.6 kN.4 kN. ur1 y = Vb1 y k1 = 8006.8.23 cm.6 = 4803.4. The initial slope of the idealized bilinear curve.2.1.6 = 22.

1 and 4.62 (cm/sec2).8.2. 7.7.4.46 cm and D1y = 26. Scaling the vertical axis by Fs1 y L1 = 203. 4.3. 4.4. 5. A. 4. Histories of roof displacement and top story drifts for the first “mode” are computed and presented in Fig. 6. * M1 gives Fs1o L1 = 233. Scaling the horizontal axis by Γ1φr1 gives D1o = 46. 4.7. The peak values are computed and are summarized in Tables 4. 4. 9.51 cm. The combined modal responses are presented in Fig. The peak values are also plotted in Fig. The results were generated for first three “modes’ and are included in Fig.7.40 (cm/sec2) and 76 . 8.2. 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.

14 y by 6000 Base Shear (kN) 4000 2000 Actual Idealized figA_3c.19 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized figA_3a. V y by = 4952 kN.13 6000 Base Shear (kN) 4000 2000 Actual Idealized figA_3b. α = 0. α = 0.eps 10 20 30 40 50 Roof Displacement (cm) (b) "Mode" 2 Pushover Curve 60 8000 u = 9.9 cm.3. “Modal” pushover curves for the example building 77 .2 cm.eps 0 0 5 10 15 20 Roof Displacement (cm) (c) "Mode" 3 Pushover Curve 25 8000 u = 4.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36.6 cm.eps 0 0 2 4 6 8 10 Roof Displacement (cm) 12 Fig. V = 5210 kN. α = 0. A. V = 7616 kN.

52 5 10 15 20 Time (sec) 25 30 0 • 1.2769 D1 (cm) 0 • 35.223 −80 80 n=3 −3 3 n=3 D3 (cm) 0 A3 (g) • 10.m −80 0 −3 0 25 30 Fig. 78 .4.06 0 • 1.(a) Deformation 80 n=1 3 (b) Pseuso−acceleration n=1 • 0. Histories of deformation and pseudo-acceleration due to 1.” and third “mode” inelastic SDF systems.33 A1 (g) n=2 0 −80 80 −3 3 n=2 D2 (cm) 0 A1 (g) • 22.” second “mode. A.5 × El Centro ground motion for the first “mode.747 5 10 15 20 Time (sec) figa_4.

9 7615.3666 3740189 203.12 3876.5309 488839.81 21.39 36.76 21.910 0.85 0.0.26 36.3 7628.9 4570.28 36.693 0.0 4588.1 1013. Results of iterative procedure to develop the idealized bilinear curve for the first “mode” inelastic SDF system Itr.2 4671.90 21.18 210.05 36.18 210.6 4583.186 0.529 0.2.4927 1.70 36.44 36.103 “Mode” 3 696400 0.Table A.18 210.029 0. No.2 4628.184 0.8 4647.9 (kN) 4803.4 7672.85 36.24 36.948 ζ n (%) 79 .59 22.194 (kN-cm) 365100 364060 363276 362684 362235 361892 361631 361432 361279 361162 361073 361004 360951 360911 360880 360856 360838 360824 360813 Table A.77 21.3 7745.75 21.18 210.02 21.4 4595.51 2 “Mode” 2 -920860 -0.139 0.5 4614.8 7622.40 46.38 22.198 0.170 0.18 210.18 210.136 Fsny Ln (cm/sec2) Dny (cm) Fsno Ln (cm/sec ) Dno (cm) Tn (sec) 233.190 0.1.2 7690.18 210.151 0.063 0.6 × Vb1 y i ur1.191 0.194 0.23 0.18 210.46 2.1 4569.62 26.18 210.3 7658.74 21.8 7618.7 4580.30 37.192 0.188 0.8 4747.0 7619.176 0.18 210.9 4570.50 36.18 210.082 0.237 0.56 19.404 0.5 3109.18 210.3 7786.162 0.194 0.25 36. Properties of “modal” inelastic SDF systems Properties Ln (kg) Γn * M n (kg) “Mode” 1 2736789 1.182 0.0 4704.23 22.18 210.32 36.037 0.135 0.05 52.9 4573.013 0.18 210.3 4603.5 7624.65 1226.86 22.59 36.56 47.1 4574.010 (kN) 8006.7 7639.8525 1.4 7714.35 36.5 7633.78 21.1 7616.022 0.2406 167531.18 210.017 0.75 21.048 0.79 0.193 0.18 210.18 (cm) 38.2671 1.194 0.11 22.0 4577.25 36.86 21.18 210.18 210.83 21.6 7840.2 4571.09 37.09 18.95 21.5 (cm) 22.4 7911.193 0.309 0.74 (kN/cm) 210.180 0.64 37.29 36.193 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 0.4 7647.79 21.6 i k1 i ur1 y i α1 i Ab1 Error (%) 1.107 0.

80 .

4. (4. rno .2 EXAMPLE The MPA procedure is implemented to calculate the response of the 9-story building described in Section 3.8) 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. of the nth-“mode” inelastic SDF system (Fig. (3. Solving Eq. Determine the total response by combining the peak “modal” responses using the SRSS combination rule of Eq.4.1 to the north-south component of the El Centro (1940) ground motion scaled up by a factor of 1. 4. 11. Typically.33 cm. Following is step-by-step implementation of the procedure described in Section B.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. or from the inelastic response (or design) spectrum. 10. This is the same example as solved in Appendix A. Steps 1 to 4 of the MPA are the same as those for UMRHA presented in Appendix A. Calculate the peak roof displacement urno associated with the nth-“mode” inelastic SDF system from Eq.21). Repeat Steps 3 to 7 for as many “modes” as required for sufficient accuracy. extract from the pushover database values of other desired responses.Appendix B Modal Pushover Analysis B.8). subtract the yield hinge rotation to determine the plastic hinge rotation.6b by solving Eq. 13.1.5. (4. 81 . 10. At urno . Compute the peak deformation.18). 6. From the total hinge rotation.3b) with unit mass and force-deformation relation of Fig. B. Dn . the first two or three “modes” will suffice. (3. 12.

At ur1o = 48.1. Peak roof displacement ur1o = Γ1 × φr1 × D1 = 1. Results of Steps 5 and 6 are summarized in Table B.3 and 4. 8. The total response computed by combining the peak “modal” responses using the SRSS combination rule of Eq. where results for other ground motion intensities are also included.33 = 48. Steps 3 to 7 are repeated for first three “modes. 82 .3 and 4.5 are the plastic hinge rotations computed by subtracting the yield hinge rotation from the total hinge rotations. 12.18) are also included in Tables 4.4 for the story drifts normalized by the story height.4. (3. values of floor displacements and story drifts are extracted.4. The values are summarized in Table 4.11.28 cm. Also included in Table 4.28 cm.” and the results are included in Tables 4. 9.3 for the floor displacements normalized by the building height and in Table 4.366 × 1 × 35.

5 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 2.1.755 0.436 7.200 0.0 µ = Dn Dny urno = Γnφrn Dn (cm) 83 .535 14.913 22.05 0.154 78.901 8.504 18.367 1.18 27.35 1.577 16.70 0.82 1.52 0.007 36.513 0.71 1.06 1.36 1.0 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 1.755 27.35 0.59 0.379 21.691 0.252 9.467 14.185 11.023 0.52 “Mode” 3 1.4222 3.25 20.79 0.Table B.85 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 1.332 13.551 2.37 1.332 48.28 46.03 0.748 63.07 “Mode” 2 4.25 Quantity Dn (cm) “Mode” 1 6.50 35.03 26.275 1.73 24.27 0.450 4.126 13.676 6.117 5.5 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.395 0.8451 5.225 2.184 0.267 5.33 1.13 2.268 0.457 12.766 7.312 1.139 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.660 14.0 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 3.75 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.735 3.690 10.229 8.38 22. Calculation of roof displacements urno from peak deformation of inelastic SDF systems Ground Motion Multiplier 0.969 0.37 57.856 31.678 0.

84

**Appendix C FEMA Force Distribution Calculations
**

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

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

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

Floor, j

mj (10-3×kg)

s* = j

∑i mi

0.112 0.110 0.110 0.110 0.110 0.110 0.110 0.110 0.119

mj

1 2 3 4 5 6 7 8 9

503.5 494.7 494.7 494.7 494.7 494.7 494.7 494.7 534.1

C.2

EQUIVALENT LATERAL FORCE (ELF) DISTRIBUTION 85

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

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

Floor j 1 2 3 4 5 6 7 8 9

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

-3 k

s* j

=

m jhk j

∑i mihik

0.007 0.020 0.038 0.062 0.091 0.126 0.165 0.210 0.281

C.3

SRSS DISTRIBUTION

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

N

∑ n (V jn )

2

to get column 8 of

9 880.9 366.5 320.7 (10) 0.2 95.7 374.3 222. the lateral forces are normalized by the base shear to obtain column 10 in Table C.2 fj = ∑ fi i fj (2) (3) (4) 59. FEMA 273 SRSS force distribution Lateral Forces Floor j (1) 1 2 3 4 5 6 7 8 9 f j1 (kN) f j2 (kN) f j3 (kN) V j1 (kN) (5) 1917.1 832.047 0.0 -5.3 -6.2 105.8 1381.1 Story Shears V j2 (kN) (6) 1114.For convenience.8 430.4 250.9 1683.7 101.0 176.7 Vj (kN) (8) 2268.7 694.0 1231.6 319. Table C.4 400.0 1476.0 136.1 -438.0 354.042 0.070 0.2 285.9 1446.6 286.5 -973.0 381.090 0.1 -646.3.6 V j3 (kN) (7) 478.2 200.2 97.7 234.2 148.9 -166.9 -153.5 -350.1 1857.065 0.7 355.7 95.2 277.1 -525.7 1578.7 525.5 159.5 215.177 0.367 87 .8 374.3 240.3.6 366.6 -352.4 1759.6 -732.0 980.7 2065.1 -967.045 0.5 -320.3 -646.7 -46.1 87.6 1233.9 446.7 1622.098 0.6 -359.2 Lateral Force fj (kN) (9) 203.8 -326.4 1842.9 832.

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

Sign up to vote on this title

UsefulNot usefula research report modal pushover analysis

a research report modal pushover analysis

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