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 .

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

iv .

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

vi .

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

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

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

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

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

5) For a given u g t .4) and w n is the natural vibration frequency. Tn = 2p w n is the natural vibration period.3) ωn = k m ζ = c 2mω n f fs = s fy (2. u £ u y ). or absolute (without regard to algebraic sign) maximum.ug t n where bg bg b g bg (2. in addition to the form of the force-deformation relation. z y .fS m fs c ut u fy 1 αk k 1 uy ug -fy k k um u 1 1 (a) (b) Fig. and R y (Chopra .. z . the deformation u t depends on three systems parameters: w n . 2001. This becomes evident if Eq. and (b) bilinear hysteretic forcedeformation relation For a given excitation u g t . (a) Idealized one-story structure. and z is the damping ratio of the system vibrating within its linear elastic range (i. and the ductility factor is u m= m uy (2. deformation is denoted by um . Section 7.1. The peak.e.3). 2. (2. m depends on three system parameters: w n . bg 4 . sign u = . and u y .1) is divided by m to obtain ~ u + 2zw n u + w 2 u y f s u.

2311 . One-story. 2. 2001) and scaled up by a factor of 2.66 m • EIb EIc EIc • • L = 7.2. this system and excitation. m h = 3. f o w = 184 . For .4 in Chopra. Vby = 39.1. 2. α = 0.2.2: the dimensions and flexural rigidity of the structural elements are noted.2 5 .04 40 Base Shear. u (cm) 8 Fig.826 kips) for w = 169. 6.3.26 kN (8. 2. based on Ry = 8 .376 cm.9 kN (38. is f y / w = ( f o / w ) ÷ 8 = 0.32 m • Fig. Vb (kN) 30 20 10 0 0 2 4 6 Roof Displacement.26 kN.5 sec and z = 5% subjected to the north-south component of the El Centro (1940) ground motion (Fig. Pushover curve for structure shown in Fig. and f y = 39. one-bay frame 50 uy = 1.2 SYSTEM AND EXCITATION CONSIDERED Consider the one-story system in Fig. The yield strength of the inelastic system. 2. with Tn = 0.2 kips) .

(b) shows the lateral force f s t or base shear Vb t normalized relative to the bg weight.1 kip-in.3 in. and E = 2 × 108 kPa (29×103 ksi) gives yield moments of 21. and q pm = 0.36 cm .) for the beam and columns.. I b = 3.017 rad .The yield moments in the beam and columns are defined as the bending moments due to the lateral force f y .) and 50. 2. However. The resulting pushover curve is shown in Fig.4f. the joint rotation and beam hinge rotation are identical to values q m and q pm determined from RHA. pushover analysis cannot provide any cumulative measure of response.4edetermined from RHAis superimposed. wherein the hysteretic force-deformation history of Fig. respectively. 2.35. the ductility factor m = 5.g. The system is excited well beyond the yield deformation.4 ). the peak value determined from RHA. Observe that the pushover curve matches the initial loading path of the hysteretic system.077 × 107 mm 4 (146 in.18 kN-m (444. 2. the energy dissipated in 6 . 2. 2. It is organized in five parts: (a) shows the deformation u t . The peak values of the various response quantities are as follows: um = 7.6 kip-in. Implementing this analysis with I c = 6. Determined from pushover analysis at the exact peak deformation.65 kN-m (191.36 cm. q m = 0. 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 . e. (c) shows the joint rotation q bt g .4e. 2. as apparent in Fig. (d) shows the rotation q p bt g of the plastic hinges at the bg bg beam ends.4 shows the earthquake response of the system described in the preceding section determined by response history analysis (RHA).0217 rad .3 RESPONSE HISTORY ANALYSIS Figure 2. and (e) shows the force-deformation relation.4 ).4 PUSHOVER ANALYSIS Static analysis of the one-story nonlinear system subjected to lateral force that increases in small increments is implemented until the lateral displacement reaches um = 7. The yielding stiffness of each structural element is defined as 3% of its initial stiffness. A static pushover analysis of this one-story system leads to the force-displacement relationship shown in Fig.3.134 × 107 mm 4 (75.

2311 y (b) −0. (c) joint rotation.2 −0.4 0.0217 • θ (rad) 0 (c) −0.04 0 5 10 15 Time (sec) 20 25 30 0. (e) force-deformation relation. (b) base shear.5 0.5 fy / w = 0.4. (d) plastic hinge rotation.2 Vb / w 0.04 0.2311 (a) Vb / w 0 f / w = −0.4 0.yielding during the ground motion. 15 u (cm) 0 • 7.017 • 0 (d) θp (rad) −0. Response of one-story system to El Centro ground motion: (a) deformation.4 −10 −5 0 u (cm) 5 10 Fig. and (f) pushover curve 7 .4 −10 −5 0 u (cm) 5 10 −0.2 0 (e) 0 (f) −0.04 0. This represents an inherent limitation of pushover analyses. or the cumulative rotation at a plastic hinge.2 −0. 2.36 −15 0.04 0.

8 .

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

an SDF system with vibration propertiesnatural frequency w n and damping ratio z n of the nth-mode of the MDF system. (3.11) and substituting in Eq.10) bg bg bg 10 (3. (3. The equations governing the response of the system are mu + cu + ku = .5) The contribution of the nth mode to s and to p eff t are: sn = Gn mφ n respectively.8) where the modal coordinate qn t is governed by 2 qn + 2z nw n qn + w n qn = . p eff . (3. we will outline that the response of the MDF system to p eff .n t is entirely in the nth-mode.6) Next.sn ug t bg (3.8) gives the floor displacements u n t = Gnφ n Dn t bg (3. it can be demonstrated that none of the modes other than the nth mode contribute to the response.9) and (3. subjected to u g t : Dn + 2z nw n Dn + w 2 Dn = .9) in which w n is the natural vibration frequency and z n is the damping ratio for the nth mode.12) .7) By utilizing the orthogonality property of modes.u g t n Comparing Eqs.peff (t ) = n =1 ∑ peff . The solution qn t can readily be obtained by comparing Eq.Gn ug t (3.n t = .10) gives qn t = Gn Dn t bg bg bg (3.sn ug t bg bg bg bg (3.n (t ) = N n =1 ∑ −sn u g (t ) N (3. with no contribution from other modes.9) to the equation of motion for the nth-mode elastic SDF system. Then the floor displacements are u n t = φ n qn t bg bg bg bg (3.

(3. (3.Forces sn An(t ) st rn ωn. Eqs.13) st where rn denotes the modal static response.15) and (3.13) represent the response of the MDF system to p eff.1). In contrast to the classical 11 bg . (3. etc. 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 (3.12) and (3.16) reflect combining the response contributions of all modes. However.13) define the contribution of the nth-mode to the response. internal element forces. and An t = w 2 Dn t n bg (3. Equations (3.15) rt = bg n =1 Â rn t = bg n =1 Â rnst An bt g N (3. and Eqs. Conceptual explanation of modal response history analysis of elastic MDF systems Any response quantity r t story drifts. 2001. st The two analyses that lead to rn and An t are shown schematically in Fig.n t .1. the static value of r due to external forces sn . Section 13. 3.9) is the standard modal equation governing qn t . 3.1. Therefore.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.14) is the pseudo-acceleration response of the nth-mode SDF system (Chopra.12) and (3. ζn ¨ g(t ) u (a) Static Analysis of Structure (b) Dynamic Analysis of SDF System Fig.16) This is the classical modal RHA procedure wherein Eq.

the peak nth-mode response as in Eq.8. The peak modal responses are combined according to the Square-Root-of-Sum-ofSquares (SRSS) or the Complete Quadratic Combination (CQC) rules.(3. (3.. 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 . z n of the pseudo-acceleration response (or design) spectrum for the nth-mode SDF system. In such a response spectrum analysis (RSA). Alternatively.derivation found in textbooks (e.3).g. 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 (3.4 and 13.18) To develop a pushover analysis procedure consistent with RSA. The SRSS rule. This concept provides a rational basis for the modal pushover analysis procedure developed later.1. and Tn = 2p w n is the natural vibration period of the nth-mode of the MDF system. Sections 12. we used the modal expansion of the spatial distribution of the effective earthquake forces. Section 13.9) . 2001.17) where An is the ordinate A Tn . we note that static analysis of the structure subjected to lateral forces f no = Γ n mφ n An (3.3 MODAL PUSHOVER ANALYSIS (3.16). 3.1).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. which is valid for structures with well-separated natural frequencies such as multistory buildings with symmetric plan. 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. 2001.17) (Chopra.19) will provide the same value of rno . (3.

region. The interior bays of the structure contain simple framing with composite floors. Obviously Dn and An are available from the response (or design) spectrum.2). with five bays each in the north-south (N-S) and east-west (E-W) directions. The levels of the 9-story building are numbered with respect to the ground level (see Fig. and California Universities for Research in Earthquake Engineering (CUREE).s* = mφ n n (3. 1660 W.18) to obtain an estimate of the peak value ro of the total response. can be combined according to Eq. 1301 S. SAC is a joint venture of three non-profit organizations: The Structural Engineers Association of California (SEAOC). 13 . SAC Steel Project Technical Office. The peak modal responses. shown in Fig.2). with the ninth level being the roof. The building has a basement level. Los Angeles.20) and the structure is pushed to the roof displacement. Typical floor-to-floor heights (for analysis purposes measured from center-of-beam to center1 2 Brandow & Johnston Associates. was designed by Brandow & Johnston Associates1 for the SAC2 Phase II Steel Project.73 m (150 ft) by 45. and 37. California. each determined by one pushover analysis.4.21) 2 where Dn = An ω n . The columns are 345 MPa (50 ksi) steel wideflange sections. Although not actually constructed. (3. Richmond. this structure meets seismic code and represents typical medium-rise buildings designed for the Los Angeles. rno . (3. Third St. 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. 3.19 m (122 ft) in elevation. which from Eq. Consulting Structural Engineers. This modal pushover analysis (MPA) for linearly elastic systems is equivalent to the well-known RSA procedure (Section 3.12) is urno = Gnf rn Dn (3. The bays are 9.4 COMPARATIVE EVALUATION OF ANALYSIS PROCEDURES 3.2. this building is 45.1 System and Excitation Considered The 9-story structure. in both directions.73 m (150 ft) in plan. 3. the Applied Technology Council (ATC). A benchmark structure for the SAC project.. denoted B1. CA 90017. 46th Street. 3.15 m (30 ft) on center. CA 94804-4698. urno . the peak value of the roof displacement due to the nth-mode.

The seismic mass of the ground level is 9. third. The 9-story N-S MRF is depicted in Fig. Column splices.00×106 kg (616 kips. such as those described in Gupta and Krawinkler (1999) can be used. The building is modeled in DRAIN-2DX (Allahabadi and Powell. partitions. i. dimension. These force n distributions will be used in the pushover analysis to be presented later.65×105 kg (66.of-beam) are 3.65 m (12 ft) and for the first floor is 5. The strength.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. the vibration periods are 2.07×106 kg (73. To ensure that this structure remains elastic. The first three vibration modes and periods of the building for linearly elastic vibration are shown in Fig.3. for the first level is 1.49 m (18 ft).85.83 m (6 ft) above the center-line of the beam to column joint. The seismic mass of the above ground levels of the entire structure is 9.89×105 kg (67. 14 . The floor-to-floor height of the basement level is 3. s* (Eq. if desired more complex models.7 kips-sec2/ft). roofing and a penthouse located on the roof.. The simple model adopted here is sufficient for the objectives of this study. 1988) using the M1 model developed by Krawinkler and Gupta (1998). The column lines employ two-tier construction.4. respectively. Concrete foundation walls and surrounding soil are assumed to restrain the structure at the ground level from horizontal displacement. This model is based on centerline dimensions of the bare frame in which beams and columns extend from centerline to centerline. for the first three modes are shown in Fig.2. 0. fifth. The seismic mass of the structure is due to various components of the structure.27. 3.2 kips-sec2/ft).e.96 m (13 ft). 3. The column bases are modeled as pinned and secured to the ground (at the B-1 level). including the steel framing. and for the ninth level is 1. monolithic column pieces are connected every two levels beginning with the first level. we select a weak ground motion: the northsouth component of the El Centro (1940) ground motion scaled down by a factor of 4. floor slabs.0 kips-sec2/ft).sec2/ft). mechanical/electrical. are located on the first. 3. for the second through eighth levels is 9. and shear distortion of panel zones are neglected but large deformation (P∆) effects are included. and 0. The force distributions.20). which are seismic (tension) splices to carry bending and uplift forces. Each frame resists one half of the seismic mass associated with the entire structure. and seventh levels at 1.01×106 kg (69.49 sec. 3. ceiling/flooring.

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

4.75 1.61 2.9th 8th 7th 6th T = 0.85 sec 2 Floor 5th 4th 3rd 2nd 1st T1 = 2.33 2.5 Fig. (3.04 1. and 3 n 3. Each figure is organized in four parts: (a) shows the roof displacement urn t . respectively. 3.5 0 0.05 1.27 sec Ground −1.67 −1.03 −1.7. Force distributions s* = mφn .728 2. and 3 .5.44 1.93 −1. is shown in Figs. First three natural-vibration periods and modes of the 9story building 3.37 2. and 3.5 −1 −0. (b) shows the base shear Vbn t normalized relative to the weight W of the building.05 2. determined by RHA [Eqs. 3. n = 1.94 2.1 3. 3.487 −1.13)]. n = 1.3.6.12 0. 2.31 −0.5 Mode Shape Component 1 1. (c) shows the joint rotation q n t of an external joint at 16 bg bg bg .39 3.51 0.2 Response History Analysis The structural response due to individual vibration modes.72 −2.49 sec 3 T = 0.12) and (3.13 −1.796 0.1 −2.4. 3.0272 −2.38 0.05 s * 1 s * 2 s * 3 Fig. 2.8 −2.

ur 2o = 2. but three modes—perhaps even two modes—are sufficient. 3. 3. and ur 3o = 0. For a fixed number of modes included. (3. including one.3.422 cm. are presented in Fig. It is apparent that the first mode alone is inadequate. larger in story drifts. which are listed in Tables 3. 3. the exact response considering all modes. and even larger in joint rotations. or all modes.1. and (d) shows the Vbn .3. As expected. This is illustrated in Fig. consistent with the increasing significance of the higher mode response among these three sets of response quantities.9. and rotations of external joints with moment connections are presented in Tables 3. drifts in all stories. in particular. 3. the results for the roof displacement and top-story drift are shown in Fig. and 3. the peak roof displacement due to each of three modes is ur1o = 9. and 3.2. errors are smallest in floor displacements. The peak values of displacements of all floors. The linear relationship between the base shear and roof displacement for each mode implies that the structure did not yield. The same method was used to determine the peak values of many response quantities. respectively. The peak values of floor displacements and story drifts determined by RHA. Combining the modal response histories for all modes gives the total response [Eqs. with the first three modes provide essentially the exact response. The peak values of the various response quantities are noted in these figures. Also included are the combined response due to one. two.8. where the second and third modal responses are a larger percentage of the top story drift compared to roof displacement.8.the roof level.urn relation.12 cm.15) and (3. and the percentage errors due to truncation of higher modal contributions.1. two. especially in estimating the story drifts.2. The errors due to truncating the contributions of vibration modes beyond the third mode are negligible. and three vibration modes. 3. 17 . errors generally decrease as response contributions of more modes are included.16)].23 cm. three. respectively.

225 0.1 Table 3.09E-03 1.94E-03 2.260 0.060 -0.336 0.45E-03 3.069 0.1 -19.7 4.11E-04 -5.407 0.197 0.008 -0.054 0.080 0.7 2.062 0.192 0.63E-03 2.2 -2.5 -1.282 0.401 Error (%) 1 Mode 2 Modes 3 Modes -23.042 0.1 4.76E-03 1.78E-04 -3.4 -1.4 -6.50E-03 2.0 -2. 18 .28E-04 1.183 0.4 -1.117 0.032 -0.85E-03 3.8 1.226 0.133 -0.069 0.11E-03 1.071 0.8 -56.060 -0.76E-03 1.6 0.4 -10.50E-03 4.99E-03 2.1 Peak values of floor displacements (as % of building height = 37.009 -0.012 -0.253 0.8 1.89E-03 1.2 -57.88E-03 2.060 0.156 0.001 -0.156 0.3 -3.9 9.045 0.0 -2.227 0.062 -0.5 -2.5 18.91E-04 1.66E-05 -3.259 0.310 0.157 0.3 Table 3.265 RHA (all modes) 0.38E-04 2 Modes 2.159 0.124 0.6 -1.399 0.245 0.31E-03 Joint Rotation (rad) Combined (RHA) 1 Mode 2.024 -0.9 1.14 m3) from RHA for 0.026 0.8 -15.29E-03 2.44E-03 3.090 0.15E-03 4.01E-04 3.260 0.00E-03 1.1 3.202 0.9 3.311 0.125 0.235 0.9 -15.6 -1.9 8.333 0.6 4.199 0.2 -4.350 0.130 0.74E-03 1.125 0.181 0.303 0.88E-03 2.89E-03 1.089 0.13E-04 9.25 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 3 Modal Response Mode 1 2.177 0.2 0.321 0.300 0.1 -14.7 2 Modes -3.202 0.74E-03 1.010 -0.012 0.9 -24.4 0.2 1.42E-04 1.8 -5.3 -8.011 0.Table 3.152 0.2 Peak values of story drift ratios (as % of story height) from RHA for 0.6 0.205 0.26E-04 -5.47E-03 1.058 -0.202 0.275 0.6 -0.4 -53.3 Peak values of joint rotations (radians) from RHA for 0.235 0.370 0.055 0.006 0.9 -23.226 0.4 -3.266 0.7 -19.9 -22.2 -20.413 RHA (all modes) 0.99E-03 Mode 3 3.8 -1.3 -33.7 7.74E-04 6.1 Building height is measured from the ground floor to the 9th floor.1 -11.09E-03 2.90E-03 3.264 Error (%) 1 Mode 2 Modes 3 Modes -23.0 -46.231 -0.72E-03 3.307 0.3 -0.002 -0.33E-03 2.266 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 0.325 0.043 0.0 -10.03E-03 1.63E-03 2.03E-03 -6.7 3.229 0.011 0.245 0.364 0.0 7.253 0.261 -0.124 0.08E-03 2.0 -0.09E-03 1.4 -0.0 3 Modes -5.44E-03 1.2 9.9 2.266 0.22E-03 2.14E-03 2.00E-03 1.097 0.13E-03 2.015 -0.7 -50.5 0.475 0.00E-03 2.038 0.38E-03 2.023 0.03E-03 3 Modes 2.042 0.3 -0.121 0.453 0.259 0.173 0.263 0.01E-04 -2.65E-03 2.088 -0.152 0.097 0.311 0.09E-03 Error (%) 1 Mode -23.3 19.64E-03 3.74E-04 9.466 0.8 -10.24E-03 2.227 0.31E-03 Mode 2 -1.022 0.003 0.73E-03 3.6 1.203 0.4 -22.179 0.378 0.177 0.4 -7.6 11.237 0.2 -1.56E-03 2.400 0.1 -2.295 -0.282 0.229 0.258 0.03E-03 1.237 0.097 0.1 -2.65E-03 RHA (all modes) 2.1 -2.1 -0.4 -41.265 0.9 -16.035 0.6 9.406 0.173 0.25 × El Centro ground motion Displacement /Height (%) Modal Response Combined (RHA) Mode 1 Mode 2 Mode 3 1 Mode 2 Modes 3 Modes Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 0.

1 −15 −7.5. Response due to first mode: (a) roof displacement.1 0.1 5 x 10 −3 θ (rad) 0.5 9.25 × El Centro ground motion 19 . and (e) pushover curve. 3.0435 9.00174 • 0 (c) r1 −5 0 5 10 15 Time (sec) 20 25 30 0.1 −15 −7.5 15 −0. (c) joint rotation.12 −0. (d) force-deformation history.1 0.0435 • 0 (b) V b1 −0.0435 Vb1 / W 0.5 0 u (cm) r1 7.5 0 u (cm) r1 7.12 • 0 (a) r1 −15 0. Excitation is 0.12 0 (d) 0 (e) 15 Fig. (b) base shear.1 /W 0.15 u (cm) 9.

6.00244 • 0 (c) r2 −5 0 5 10 15 Time (sec) 20 25 30 0.25 × El Centro ground motion 20 .05 5 x 10 −3 b2 θ (rad) 0.1 −5 −2. and (e) pushover curve. (b) base shear.5 0 u (cm) r2 2.05 /W 0.23 0.0252 (d) 0 0 (e) −0.23 −2.1 −5 −2.23 −5 (a) 0.0252 • 0 (b) V −0.5 5 −0. Response due to second mode: (a) roof displacement. Excitation is 0.1 0.1 Vb2 / W 0.5 0 u (cm) r2 2.5 5 Fig. (c) joint rotation. 3. (d) force-deformation history.0252 −2.5 u (cm) r2 0 • 2.

0108 −0.02 −0.0108 x 10 −3 (b) −0.02 2 θ (rad) 0.5 1 −0. (d) force-deformation history.02 −1 −0.02 /W V b3 0 • 0.422 /W V b3 0 (d) 0 −0.5 0 u (cm) r3 0. (c) joint rotation. Response due to third mode: (a) roof displacement. and (e) pushover curve.0108 −0.25 × El Centro ground motion 21 . Excitation is 0.02 −1 −0. 3.000913 • 0 (c) r3 −2 0 5 10 15 Time (sec) 20 25 30 0.5 1 Fig.422 −1 (a) 0.422 (e) −0.1 u (cm) r3 0 • 0. (b) base shear.02 0.7.5 0 u (cm) r2 0.

3. leads to the pushover curves shown in Figs. In practical application. and 3. story drifts.18). these pushover curves are consistent with the Vb . and 3. 3. 3.4.e. As for the first mode.9).6.5e.3 Modal Pushover Analysis Implementing MPA for the fundamental vibration mode. and 3.5..1. i. pushing the structure using the force distribution of Eq. As suggested by Eq.4. For a fixed number of modes included considered.422 cm . and the computed response values are identical to the peak response values determined from RHA (Tables 3. implying that the modal truncation errors are small if at 22 .2. and it is apparent in the example considered that three modes provide most of the response (Fig.3). 3. (3.12). 3. 3.5d). respectively. and the errors in these estimates relative to the exact response from RHA considering all modes.4.7d).12 cm.6e and 3.. 3. Figure 3. (3.1. 3. two. and external joint rotations computed by pushover analysis are presented in Tables 3. This pushover curve is consistent with the relationship between the base shear and roof displacement determined by RHA (Fig.23 cm . the errors in the MPA results are generally larger than in RHA (Fig. 3. this value would be determined directly from the response (or design) spectrum. and 3.5.3. and 3.6d and 3. pushing the structure.20) with n = 2 and 3 up to roof displacements ur 2 o = 2. 3.3). 3. (3. story drifts. which would provide the Dn value to be substituted in Eq.6. considering one. 3.1 through 3.2.3). or three vibration modes. The floor displacements. the floor displacements are proportional to the mode shape φ 1 because the structure remains elastic. and ur 3o = 0.21).e. (3.9 and Tables 3. respectively. although both analyses led to identical peak values of the individual modal responses. (3. using the force distribution of Eq. respectively. except for the algebraic sign associated with Γ1 and minor round-off errors.6 present estimates of the combined response according to Eq.ur relations determined by RHA (Figs.10 and Tables 3. Implementing pushover analysis for the second and third modes. and external joint rotations in Tables 3. 3.8) leads to the pushover curve shown in Fig. Observe that the target roof displacement in each pushover analysis is identical to its exact value determined by RHA.5. confirming that MPA gives the exact values of the individual modal responses.4.7e and to the floor displacements.4) to roof displacement ur1o = 9. In RHA the errors arise only from truncating the responses due to higher modes. i. These values of the response quantities are identical to the peak response values determined from RHA (Tables 3. the value determined by RHA (Fig.20) with n = 1 (Fig.

3.48 −3 3 1.8.2 0.03 −3 3 Mode 3 ∆ (cm) u (cm) r3 −15 15 9.83 • r3 0 • 0.4 Displacement/Height (%) 0.685 • −3 3 Mode 2 ∆ (cm) u (cm) • 2.422 r2 0 • 1.least three modes are included.1 0.25 × El Centro ground motion 23 .5 Ground 0 0.2 0.59 • Total (All Modes) 0 −15 0 ∆ (cm) u (cm) 0 −3 0 fig3_8. Heightwise variation of floor displacements and story drifts from RHA for 0. (a) Roof Displacement 15 (b) Top Story Drift 3 Mode 1 ∆ (cm) u (cm) 9.1 0.9.eps fig3_9b.3 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.4 Story Drift Ratio (%) 0.3 0. Additional errors are introduced in pushover analysis due to the approximation inherent in modal combination rules.6 Fig.23 r2 −15 15 0 • 0.5 0.12 • r1 −15 15 0 r1 0 0 0. sec 25 30 r 5 10 15 20 Time.eps r 5 10 15 20 Time.eps Ground 0 0. sec 25 30 Fig. Response histories of roof displacement and top-story drift from RHA for 0. 3.

9 -13.3 -9.133 -0.80E-04 3.010 0.1 -11.282 -0.7 -21.6 -15.22E-04 2.00E-03 2.002 0.0 -2.24E-03 -2.253 RHA (All Modes) 0.3 -12.9 -24.4 -53.9 -15.97E-03 1.152 0.9 -11.2 -12.267 0.9 -14.09E-03 1.78E-04 2 Modes 2.94E-03 2.385 0.313 0.203 0.038 -0.72E-03 3.9 -14.00E-03 1.227 0.2 -0.6 -19.055 0.76E-03 1.001 0.259 0.088 -0.42E-04 -1.00E-03 2.060 0.4 0.106 0.285 0.237 0.9 -13.227 0.296 -0.47E-03 1.1 -11.4 -5.2 1.043 -0.062 -0.3 1.43E-04 -1.7 -50.73E-03 3.8 -15.9 -15.9 -14.121 0.197 0.63E-03 2.03E-03 1.12E-03 1.33E-04 5.4 -2.89E-03 -1.048 0.237 -0.3 Table 3.0 -16.282 0.74E-03 1.253 -0.331 0.401 Error (%) 1 Mode 2 Modes 3 Modes -23.181 0.011 0.03E-03 6.74E-04 -6.235 0.270 0.023 -0.036 -0.00E-03 2.008 -0.2 -16.229 0.090 0.310 0.76E-03 -1.3 -19.09E-04 -3.2 Table 3.6 -17.25 × 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 0.260 0.062 0.8 -15.203 0.04E-03 3.6 -16.203 0.4 -14.73E-05 3.44E-03 -1.253 0.44E-03 3.28E-03 2.260 0.253 0.125 0.3 -33.133 0.336 RHA (all modes) 0.4 -11.259 -0.9 -18.4 Peak values of floor displacements (as % of building height = 37.65E-03 2.097 0.53E-04 -9.0 -18.045 -0.272 0.40E-04 5.156 0.4 -9.9 3 Modes -12.1 -19.3 -2.226 0.270 0.4 -9.1 -18.097 0.229 0.4 1.5 Peak values of story drift ratios (as % of story height) from MPA for 0.96E-03 2.177 0.179 0.012 0.0 -46.231 -0.006 -0.230 -0.106 0.015 -0.8 -14.63E-03 -2.274 0.336 0.9 -8.261 -0.2 -20.15E-03 2.078 0.4 -22.7 -15.03E-03 3.270 0.7 2 Modes -13.2 -4.069 0.25 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Modal Response Mode 1 -2.5 -16.00E-03 3.235 -0.009 0.179 0.003 -0.9 -15.3 -13.374 0.57E-03 1 Mode -23.321 0.7 24 .3 -41.466 0.177 0.31E-03 Joint Rotation (rad) Combined (MPA) 1 Mode 2.3 -14.032 0.060 -0.7 -19.22E-03 -2.2 -57.9 -12.012 -0.0 -0.133 0.92E-04 -1.026 0.173 0.328 0.300 0.173 0.9 -13.15E-03 1.0 -10.31E-03 Error (%) RHA (all modes) 2.7 2.8 -22.245 -0.4 -4.058 -0.260 -0.264 Error (%) Modal Response 1 Mode 2 Modes 3 Modes -23.022 0.Table 3.079 0.266 0.069 0.90E-03 1.048 0.31E-03 2.9 -13.4 -7.230 0.09E-03 -1.14 m) from MPA for 0.071 0.09E-03 Mode 2 1.332 0.370 0.152 0.157 0.4 -22.38E-03 3 Modes 2.042 0.99E-03 Mode 3 -3.03E-03 -1.024 0.1 -0.74E-03 -1.117 0.8 -56.157 0.89E-03 1.89E-03 2.9 -15.2 -11.322 0.08E-03 2.4 -19.267 -0.276 0.38E-03 2.4 -4.267 0.89E-03 2.203 0.8 -23.080 0.89E-03 -2.00E-03 -1.25 × 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 -0.6 Peak values of joint rotation (radians) from MPA for 0.125 0.124 0.286 0.042 0.011 -0.203 0.156 0.245 0.05E-03 3.9 -16.90E-04 -9.407 0.

6 Fig.4 Displacement/Height (%) 0.3 0.(a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th (b) Story Drift Ratios Floor 4th 3rd 2nd 1st Floor 5th RHA All Modes MPA 1 Mode 2 Modes 3 Modes 5th 4th 3rd 2nd 1st RHA All Modes MPA 1 Mode 2 Modes 3 Modes fig3_10a.5 Ground 0 0.10.2 0. 3.25 × El Centro ground motion.1 0.2 0.eps Ground 0 0.4 Story Drift Ratio (%) 0.eps fig3_10b.5 0. shading indicates modal combination error 25 . Heightwise variation of floor displacements and story drifts from MPA for 0.1 0.3 0.

26 .

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

2). Simultaneously solving these coupled equations and using Eq. (3. where φ n are now the modes of the corresponding linear system. sign qn = φ n f s u n .4)] leads to the uncoupled modal response history analysis (UMRHA) procedure.5) This resisting force depends on all modal coordinates qn t . (4.Substituting Eq.n t given by Eq. Eq. 2. give the same results for u t as obtained directly from Eq. premultiplying by φ T .Gn ug t Mn bg n = 1. these equations are coupled for inelastic systems.and classical n damping-orthogonality property of modes gives F qn + 2z nw n qn + sn = .9) for linearly elastic systems. qr t = 0 for all 28 bg .3) will. Unlike Eq.4) represents N equations in the modal coordinates qn . (4.6) The solution of Eq. implying coupling of modal coordinates because of yielding of the structure.sn ug t bg b g bg bg (4.4) is rarely solved because it offers no particular advantage over Eq.2).6) for inelastic systems will no longer be described by Eq. in principle. The equations governing the response of the inelastic system to p eff.9) involves T Fsn = Fsn qn . (4. (4. (4. 4. however. This approximate RHA procedure is the preliminary step in developing a modal pushover analysis procedure for inelastic systems. (3.8) because qr t will generally be nonzero for “modes” other than the nth “mode. (4. (4.4) where the only term that differs from Eq.2).2 UNCOUPLED MODAL RESPONSE HISTORY ANALYSIS bg Neglecting the coupling of the N equations in modal coordinates [Eq. Equation (4. (3.3). (3.… N (4. However.6b) are mu + cu + f s u.3) in Eq. (3. The spatial distribution s of the effective earthquake forces is expanded into the modal contributions sn according to Eq. sign u = . For linear systems. (4. sign u n b g b g bg (4.” implying that other “modes” will also contribute to the solution. and using the mass.

(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

ζn.9 • 0 −50 10 −50 10 n=3 ur3 (cm) 0 • 5. n = 1.3.89 50 r1 0 u (cm) 0 • 78. and (b) approximate solution by UMRHA bg af bg 32 . Roof displacement due to p eff. 4. 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.n t = .575 5 10 15 20 Time (sec) 25 30 ur3 (cm) 0 • 5.Forces sn Unit mass An(t ) st rn ωn. 2.51 • ur2 (cm) 0 ur2 (cm) n=3 14. Fsn / Ln ¨ g(t ) u (a) Static Analysis of Structure (b) Dynamic Analysis of Inelastic SDF System Fig.02 −150 n=2 −150 50 n=2 14.0 ¥ El Centro ground motion: (a) exact solution by NL-RHA.4.112 5 10 15 20 Time (sec) 25 30 −10 0 −10 0 Fig. where u g t = 3. 4.s n u g t . and 3.

respectively. this intense excitation was chosen to ensure that the structure is excited well beyond its linear elastic limit. 4.5)] is neglected.5. Top story drift due to p eff. 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.0 ¥ El Centro ground motion.855 ∆r1 (cm) 0 ∆r1 (cm) n=2 • 6.1 Underlying Assumptions and Accuracy The approximate solution of Eq.744 • 0 −20 10 n=3 −20 10 n=3 ∆r3 (cm) 0 • 5.5.s n u g t . The errors are slightly larger in drift than in displacement.6) by UMRHA is compared with the “exact” solution by nonlinear RHA.4 and 4.956 5 10 15 20 Time (sec) 25 30 ∆r3 (cm) 0 • 5.15) is strictly valid only for linearly elastic systems. (4.817 5 10 15 20 Time (sec) 25 30 −10 0 −10 0 Fig.4) and (4.2. and (b) approximate solution by UMRHA af bg bg 4. the errors in either response quantity are only a few percent. n = 1. 2.n t (n = 1. both for 3.Dn relation is 33 bg bg . but even for this very intense excitation. and 3. (3.33 0 −20 20 −20 20 n=2 ∆r2 (cm) 0 ∆r2 (cm) 7. Such comparison for roof displacement and top-story drift is presented in Figs.008 • 6. (4.n t = .0 ¥ El Centro ground motion: (a) exact solution by NL-RHA. where u g t = 3.(a) Nonlinear RHA 20 n=1 20 (b) UMRHA n=1 • 5. 4. (2) the superposition of responses to p eff. and (3) the Fsn Ln .2… N ) according to Eq.

(3. such nonlinear static analysis provides the so-called pushover curve. 4.1.9) should be determined by nonlinear static analysis of the structure as the structure undergoes displacements u = Dnφ n with increasing Dn .Dn curve. (3.4. which is different than the Fsn Ln .8) before it can be solved? Because Eq. 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. when specialized for linearly elastic systems it is identical to the RHA procedure of Section 3. the base shear is Vbny and roof displacement is urny .9).urn pushover curve to the Fsn Ln . this distribution seems to be a rational choiceeven for inelastic systemsto determine Fsn in Eq.2. (4.Dn relation? The two sets of forces and displacements are related as follows: V Fsn = bn Γn Dn = urn Γ nf rn 34 (4. The overall errors in the UMRHA procedure are documented in the examples presented in Section 4. However. How to convert this Vbn . the only force distribution that produces displacements proportional to f n is given by Eq. At the yield point.2 Properties of the nth-“mode” Inelastic SDF System How is the Fsn Ln . What is an appropriate invariant distribution of lateral forces to determine Fsn ? For an inelastic system no invariant distribution of forces can produce displacements proportional to f n at all displacements or force levels. The structure is pushed using the force distribution of Eq. Although most commercially available software cannot implement such displacement-controlled analysis. (3. A bilinear idealization of this pushover curve for the nth-“mode” is shown in Fig. is plotted against roof displacement urn . Therefore.20) to some bg pre-determined roof displacement.8) governing Dn t is based on Eq.12) for floor displacements.approximated by a bilinear curve to facilitate solution of Eq. 4.20). the relationship between lateral forces f s and Dn in Eq.8) in UMRHA. it can conduct a force-controlled nonlinear static analysis with an invariant distribution of lateral forces. (4.6a. (4. and the base shear Vbn . within the linearly elastic range of the structure.Dn relation to be determined in Eq. Although several approximations are inherent in this UMRHA procedure. (4.10) . (4.

(4. (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. the initial slope of the pushover curve in Fig.2. 4.6b.8). 4. 35 .12) implying that the initial slope of the curve in Fig. where the yield values of Fsn Ln and Dn are Fsny Ln = Vbny * Mn Dny = urny Γ nf rn (4.6a is 2 kn = ω n Ln . which may differ from the period of the corresponding linear system. The two are related through Fsny Ln 2 = w n Dny (4.11) * in which M n = Ln Γn is the effective modal mass (Chopra. 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. Section 13. should be used in Eq.6b is w 2 . Properties of the nth-“mode” inelastic SDF system from the pushover curve Equation 4.6.11). In contrast.13) This value of Tn .10 enables conversion of the pushover curve to the desired Fsn Ln − Dn relation shown in Fig.5). 4. 2001. which is not a meaningful quantity. 4. Knowing Fsny Ln and Dny from n Eq.

6b. ωn . with the structure is pushed to the roof displacement urno . Combine the “modal” responses using Eqs.20)]. and modes.13).6b) by utilizing Eq. 8. 5. Calculate the peak value. develop the base-shear – roof-displacement ( Vbn − urn ) pushover curve for the force distribution s* [Eq. the peak value of Dn t . Typically.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. r o . Idealize the pushover curve as a bilinear curve (Fig.4. 4.3b) with force-deformation relation of Fig. 4. of the total response r t obtained in Step 8. the first two or three modes will suffice.6a). Calculate histories of various responses by Eqs. (3.21) where Dn . and pseudo-acceleration history. n (3.15) and (3. 4. For the nth-mode. Dn (t ) .3 MODAL PUSHOVER ANALYSIS bg bg A pushover analysis procedure is presented next to estimate the peak response rno of the inelastic MDF system to effective earthquake forces p eff .2. 4. 4. 6. 4. This value of the roof displacement is given by Eq. (4. (3. Convert the idealized pushover curve to the Fsn Ln − Dn relation (Fig. φn . An t . (3. of the nth-“mode” inelastic SDF system (Fig. is now determined by 36 bg bg . for linearly-elastic vibration of the building. Compute the deformation history. 7. which is summarized next as a sequence of steps. Repeat Steps 2 to 6 for as many modes as required for sufficient accuracy.20)]. 2. (3. Consider a nonlinear static analysis of the structure subjected to lateral forces distributed over the building height according to s* [Eq.n t .12) and (3. 9. n 3.16) to determine the total response. details are available in Appendix A: 1.11). Compute the natural frequencies.

It provides results for elastic buildings that are identical to the well-known RSA procedure (Section 3. bg bg bg bg Thus. 5.2 and 3. 4. the pushover analysis provides an estimate of the peak value rno of any response rn t : floor displacements.8). e. Dn .1 Summary The peak inelastic response of a building to earthquake excitation can be estimated by the MPA procedure just developed. plastic hinge rotations. Eq. governed by Eq. At this roof displacement. As shown in Sections 3. Sections 7. 2001.2. However. of the nth-“mode” inelastic SDF system (Fig. 37 .n t . 4. 4. (4. joint rotations. (3. alternatively. etc. Compute the peak deformation. to obtain an estimate of the peak value ro of the total response. (3. and (2) it provides the exact modal response for elastic systems.12).3. The response value rno is an estimate of the peak value of the response of the inelastic system to p eff.21). which is summarized next as a sequence of steps. each determined by one pushover analysis. Steps 1 to 4 of the MPA are same as those for UMRHA. as mentioned earlier.3) because. the lateral force distribution used possesses two properties: (1) it appears to be the most rational choice among all invariant distribution of forces. for elastic systems.3b) with force-deformation relation of Fig. This application of modal combination rules to inelastic systems obviously lacks a theoretical basis. 6. although somewhat intuitive for inelastic buildings. it can be determined from the inelastic response (or design) spectrum (Chopra.g.6). Calculate the peak roof displacement urno associated with the nth-“mode” inelastic SDF system from Eq. seems reasonable. rno also represents the exact peak value of the nth-mode contribution rn t to response r t . This pushover analysis. (4.solving Eq. it seems reasonable because it provides results for elastic buildings that are identical to the well-known RSA procedure described in Section 3. details are available in Appendix B.6 and 7. as described in Section 4. story drifts. are combined using an appropriate modal combination rule.18). we will refer to rno as the peak “modal” response even in the case of inelastic systems. The peak “modal” responses rno ..8).6b by solving Eq.2. (4. or from the inelastic response (or design) spectrum.4.3.

4.0 to 3.1 and 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.4.8. the combined response due to three “modes”. Typically. and the errors in the approximate results are shown in Fig. two. the peak roof displacement due to each of the three “modes” is ur1o = 48.1). and three “modes”.2. This is to be expected.7. Repeat Steps 3 to 8 for as many “modes” as required for sufficient accuracy. the UMRHA procedure lacks a rigorous theory. with. The peak values of response are as noted.3 cm. ur 2o = 11. the “exact” results. two.0. 4. 9.7 shows the individual “modal” responses. although the trends are not systematic as when the system remained elastic (Section 3. To ensure that this structure responds well into the inelastic range the El Centro ground motion is scaled up by a factor varying from 1. Figure 4. and the “exact” response from nonlinear RHA for the roof displacement and top-story drift. (3. The peak values of displacements of all floors and drifts in all stories are presented in Tables 4.1 Uncoupled Modal Response History Analysis The structural response to 1. the first two or three “modes” will suffice. Observe that errors tend to decrease as response contributions of more “modes” are included. Determine the total response by combining the peak “modal” responses using the SRSS combination rule of Eq. 4. in particular. This deficiency also implies that. also included are the combined responses due to one. From the total rotation of a plastic hinge. 8. extract from the pushover database values of other desired responses. and the percentage errors in the approximate results. subtract the yield value of hinge rotation to determine the hinge plastic rotation.2). rno . and ur 3o = 2.4. respectively. the 38 . The peak values of floor displacements and story drifts including one. is presented next. and compared with the results of a rigorous nonlinear RHA using the DRAIN-2DX computer program. and three modes are compared with the “exact” values in Fig. three “modes” included.18). say.53 cm. At urno .9.5 ¥ the El Centro ground motion including the response contributions associated with three “modal” inelastic SDF systems. determined by the UMRHA procedure. 4.7 cm. in contrast to modal analysis (Section 3.

3 11. Response histories of roof displacement and top-story drift due to 1.1 and 4.7.7 • −80 80 ∆r1 (cm) u (cm) "Mode" 1 −12 r2 0 ∆r2 (cm) 12 0 u (cm) −80 80 −12 r3 0 • 2. for a fixed number of “modes” included.2). 1. 2.44 • "Mode" 2 r1 0 • 48. we investigate how the errors in the UMRHA vary with the deformation demands imposed by the ground motion. For this purpose the UMRHA and exact analyses were repeated for ground motions of varying intensity. 1.5. recall that the computed errors have been presented earlier for ground motion multipliers 0. and 3.85. 0. 4. sec 25 30 • 48. the errors in responses computed by UMRHA including three “modes” relative to the “exact” response were determined.response is much less accurate (Tables 4.2) if the system yields significantly versus if the system remains within the elastic range (Tables 3.0.1 and 3. just as for elastic systems.5 × El Centro ground motion: individual “modal” responses and combined response from UMRHA.1 −12 12 ∆r (cm) u (cm) UMRHA 3 "Modes" 0 • 7.5 (Tables 4. the errors in story drifts are larger compared to floor displacements. For each excitation. defined as the El Centro ground motion multiplied by 0. the degree to which the system deforms beyond its elastic limit. However. Next.0.88 u (cm) "Mode" 3 −80 80 0 −80 80 0 −80 0 • 44. sec 25 30 r −12 0 Fig. 0.2) and 1.0.24 • NL−RHA r −12 12 ∆r (cm) u (cm) 0 5 10 15 20 Time. and total response from NL-RHA 39 .53 ∆r3 (cm) 12 0 • 2. in particular.38 6. 0.1 and 3.1 and 4.5.25.25 (Tables 3. (a) Roof Displacement (b) Top Story Drift 80 12 0 • 3.6 5 10 15 20 Time.2).75.62 5.

(3. two.5 × El Centro ground motion (a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th (b) Story Drift Ratios Floor 4th 3rd 2nd 1st UMRHA 1 "Mode" 2 "Modes" 3 "Modes" Floor 5th 5th 4th 3rd 2nd 1st UMRHA 1 "Mode" 2 "Modes" 3 "Modes" Ground −60 −40 −20 0 Error (%) 20 40 60 Ground −60 −40 −20 0 Error (%) 20 40 60 Fig.5 Displacement/Height (%) 2 Ground 0 0. or three “modes” for 1.5 1 1. 4.9.10a). Heightwise variation of floor displacements and story drift ratios from UMRHA and NL-RHA for 1.11.5 Fig. the pushover curves using force distributions s* [Eq. with the peak displacement of each “modal” SDF system noted for each ground motion multiplier. 4.20)] for the first three n modes of the system are shown in Fig.10 summarizes the error in UMRHA as a function of ground motion intensity. in each story drift (Fig. For this purpose.5 ¥ El Centro ground motion Figure 4.(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. and the error envelope for each case. Heightwise variation of error in floor displacements and story drifts estimated by UMRHA including one. Shown is the error in each floor displacement (Fig.8. it will be useful to know the deformation of the system relative to its yield deformation. Two versions of the pushover curve are 40 . 4.5 Story Drift Ratio (%) 2 2. 4. 4.10b).5 1 1. To interpret these results. indicated by a ground motion multiplier.

the peak values of displacements at all floors (Table 4. the values determined by RHA of the nth-mode inelastic SDF system (Fig.4). the pushover curve for each “mode” is idealized by a bilinear curve in solving Eq.3 cm. the errors in truncating the higher mode contributions are negligible. The system remains elastic up to ground motion multiplier 0.1). The idealized curve for the first “mode” deviates most from the actual curve near the peak displacement corresponding to ground motion multiplier 1.0. Each of these three pushover analyses provides the pushover curve (Fig.0. as mentioned in Section 4. the first reason mentioned above seems to be the primary source for the errors. (2) increase rapidly as the ground motion multiplier increases to 1. 4.2.2 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 1.01 (Fig. determined from “exact” analyses. 2. 4.2. (3) maintain roughly similar values for more intense ground motions.4. the ductility factor for the first mode system is only 1. even though the system remains essentially elastic.7). (3. 4. and (4) are larger in story drift compared to floor displacements. 4. Second. and plastic hinge rotations at the external beam end at each floor level (Table 4.10) for each “modal” inelastic SDF system (Figs 4. and. For more intense excitations. Figure 4.3). and 3 (Fig. The structure is pushed using the force distribution of Eq.11a). UMRHA lacks a rigorous theory and is based on several approximations.75.20) with n = 1. respectively. (4.4) to roof displacements urno = 48. First. The location of plastic hinges and their rotations.11).5).4.53 cm. Additional errors are introduced in UMRHA of systems responding beyond the linearly elastic limit for at least two reasons. as mentioned in Section 3. 3. 41 . and 2.6 and 4.5 ¥ the El Centro ground motion.included: the actual curve and its idealized bilinear version. 11. Perhaps this explains why the errors are large at this excitation intensity. drifts in all stories (Table 4.7 cm. were noted but are not shown here.75.10 permits the following observations regarding the accuracy of the UMRHA procedure: the errors (1) are small (less than 5%) for ground motion multipliers up to 0.

3 8.490 1.945 -37.513 0.260 -15.914 -0.9 16.010 0.0 1.8 0.526 -0.852 1.376 -1.1 8.5 2 GM Multiplier 2.256 1.1 1.2 Peak values of story drift ratios (as % of story height) from UMRHA for 1.379 1.5 10.214 0.942 1.5 28.5 × 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 -1.575 -41.049 -0.6 4.096 0.8 1.820 -19.4 4.0 9.844 -25.072 -1. Noted 60 50 40 (b) Story Drifts Error Envelope Error for Story No.169 0.373 -0.513 -0.8 1.0 -9.1 0.0 7.104 0.820 -0.698 1.071 0.044 1.856 2.126 0.982 9.983 1.220 0.291 0.072 1.070 1.863 0.338 1.9 12.018 0.071 -0.298 -0.1 1.201 -0.9 5.5 3 0 0 0.068 0.009 -0.003 0.202 11.707 1.2 1.283 1.811 1.490 -1.0 2.5 18.121 -0.033 0.478 0.5 2 GM Multiplier 2.5 28.3 25.5 9.1 3.3 1.187 -0.2 6.317 0.407 -10.426 -1.2 12.241 -1.214 -0.216 1.256 -1.130 0.120 1.8 14.10.135 9.877 0.2 1.772 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 1.201 -1.333 0. 4.0 11.410 -1.942 -0.366 0.003 -31.971 1.9 12.5 1 1.863 1.484 0.3 42 .722 0. Errors in UMRHA as a function of ground motion intensity: (a) floor displacements. and (b) story drifts Table 4.5 9.7 14.372 -1.154 0.315 -0.226 -0.7 31.6 2.370 -0.473 -22.3 6.8 1.727 1.663 0.201 1.065 0.138 1.298 0.371 -0.079 0.663 -0.5 1 1.914 2.293 1.338 -1.7 Table 4.372 1.220 -0.248 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.350 -0.676 0.6 1.495 1.5 -3.1 Peak values of floor displacements (as % of building height = 37.4 -7.371 0.806 0.938 1.4 -1.616 -0.8 1.057 -0.900 -10.136 1.9 31.055 -0.668 -23.2 4.133 1.472 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 0.200 8.5 3 Fig.376 1.938 -1.410 1.8 1.554 1.14 m) from UMRHA for 1.806 -0.3 0.088 10.819 2.241 1.(a) Floor Displacements 60 50 40 Error Envelope Error for Floor No.763 -15.430 1.540 0.9 31.366 -0.4 1.235 -0.5 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Displacement /Height (%) “Modal” Response Combined (UMRHA) “Mode” 1 “Mode” 2 “Mode” 3 1 “Mode” 2 “Modes” -0.0 11.256 1.751 1.2 22.

4.5). were determined by four analyses: MPA considering one “mode. The drifts are underestimated by up to 13% in the lower stories. 4. For example.3 through 4. (3.13 with Fig. but the third “mode” contributions do not seem especially important (Fig.Figures 4.13 and Tables 4. the results were not always accurate.5 present estimates of the combined response according to Eq. two. Obviously.18). in general.4 with Tables 4. overestimated by up to 18% in the middle stories.4). this error is not especially significant because the hinge plastic rotation is very small. 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%. Fortuitously. considering one. even if three “modes” are included (Fig.” respectively.3 and 4. The locations of plastic hinges shown in Fig.4. significant improvement is achieved by including response contributions due to the second “mode”.11c). the additional errors due to the approximation inherent in modal combination rules tend to cancel out the errors due to the various approximation contained in the UMRHA. 4. and are within a few percent of the exact values for the upper stories. for two or three modes included.1 and 4.12 and Tables 4. and the errors in these estimates relative to the exact response from nonlinear RHA. the errors in the modal pushover results are. MPA failed to identify the plastic hinges at the column bases in Fig.13 and Tables 4.14. but was more successful when the excitation was more intense (results are not included).4). and three “modes.” two “modes.12 and Tables 4.13 and Table 4. Furthermore.” and nonlinear RHA. However. 4. 4.2).3 and 4. the third “mode” does not contribute because this SDF system remains elastic (Fig.” and three “modes. 4.12 and 4.14. The second “mode” was necessary to identify hinges in the upper stories. The first “mode” alone is inadequate. 4. significantly smaller than in UMRHA (compare Fig. 43 . Observe that the primary contributor to plastic hinge rotations in the lower stories is the first “mode” and the second “mode” in the upper stories. Pushover analysis seems to be inherently limited in computing accurately hinge plastic rotations. especially in estimating the story drifts (Fig. The errors are especially large in the hinge plastic rotations estimated by the MPA procedure. the hinges identified in beams at the 6th floor were at variance with the “exact” results.3 and 4. however. although the error is recorded as 100% if MPA estimates zero rotation when the nonlinear RHA computes a non-zero value. As shown in Figs.3 and 4.9 and Tables 4. One “mode” pushover analysis was unable to identify the plastic hinges in the upper stories where higher mode contributions to response are known to be more significant. 4.

1.25 5 10 15 20 Roof Displacement (cm) 25 (c) "Mode" 3 Pushover Curve 12000 10000 u = 4. 2.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36.9 cm. 4. α = 0.5 0.5 2 3 Actual Idealized 2000 0 0 0.5. 0.11.85 0.5 2 3 4000 2000 0 0 Actual Idealized 0.0.19 y by 3 2 1.0. and 3.5 1 0.13 y by Base Shear (kN) 8000 6000 1 1.25. 1.5 0. α = 0.85 0. “Modal” pushover curves with peak roof displacements identified for 0.0 × El Centro ground motion 44 .25 1 2 4 6 Roof Displacement (cm) 8 10 Fig.6 cm.75.75 Actual Idealized 20 40 60 Roof Displacement (cm) (b) "Mode" 2 Pushover Curve 80 12000 10000 u = 9.2 cm. V = 4952 kN.85 Base Shear (kN) 8000 6000 4000 2000 0 0 0.75 0. 0. V = 7616 kN. α = 0. V = 5210 kN.75 0.5 0.5.25 0.14 y by Base Shear (kN) 8000 6000 4000 1.

10 and 4. the errors in MPA were larger for ground motion multipliers less than 0. As mentioned in Section 3.15a). now standard in engineering practice.(a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th NL−RHA MPA 1 "Mode" 2 "Modes" 3 "Modes" (b) Story Drift Ratios Floor Floor 5th 4th 3rd 2nd 1st 5th 4th 3rd 2nd 1st NL−RHA MPA 1 "Mode" 2 "Modes" 3 "Modes" Ground 0 0.4 also apply to MPA.15b). 4.75. 4.0.15) for ground multipliers larger than 1.15).5 × El Centro ground motion. The errors are only weakly dependent on ground motion intensity (Fig.4. implying that the modal combination errors contained in these procedures are acceptable. implying excitations weak enough to limit the response in the elastic range of the structure. identified in Section 3. 4. indicated by a ground motion multiplier. shading indicates errors in MPA including three “modes” Figure 4. However.12 Heightwise variation of floor displacements and story drift ratios from MPA and NL-RHA for 1. the errors in MPA were fortuitously smaller than in UMRHA (compare Figs. 4.3 for elastic systems (or weak ground motions).5 1 1. Shown is the error in each floor displacement (Fig. In this case. 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.15 summarizes the error in MPA considering three “modes” as a function of ground motion intensity. 45 .4.5 Story Drift Ratio (%) 2 2. whereas MPA contains errors inherent in modal combination rules.5 Displacement/Height (%) 2 Ground 0 0.5 Fig. implying excitations intense enough to cause significant yielding of the structure. as discussed in Sections 3. an observation with practical implications. UMRHA is essentially exact. each story drift (Fig.2 and 3.3. While various sources of errors in UMRHA. and the error envelope for each case.5 1 1. the MPA procedure is equivalent to the RSA procedure. 4.

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

687 0.0 -5.37E-03 1.9 7.473 -15.00E+00 0.176 0.3 -3.895 1.37E-04 Error (%) “Modal” Response 1 “Mode” -32.00E+00 0.00E+00 0.053 1.60E-04 7.0 1.407 -27.7 1.055 0.9 5.8 -29.2 -100.594 -1.895 1.130 0.088 12.942 6.980 -0.982 13.9 -4.5 2.737 0.0 15.910 1.752 1.60E-03 2.2 1.012 1.05E-03 2.1 18.168 -0.581 0.5 -6.55E-03 3.7 -12.371 -0.4 -8.8 7.5 10.503 1.209 1.8 -6.007 1.315 0.2 -100.666 Table 4.8 -12.414 1.26E-04 9.071 -0.244 0.00E+00 0.00E+00 0.5 7.8 “Mode” 3 -1.02E-03 3.057 0.00E+00 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th “Mode” 2 0.6 -9.015 0.1 13.5 × 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 0.694 1.5 2.033 -0.116 1.641 1.8 1.6 13.8 -6.00E+00 3.5 1.190 -0.8 1.945 -49.125 -1.8 0.00E+00 0.1 62.2 -100.36E-03 6.88E-03 0.00E+00 0.00E+00 3 “Modes” 7.18E-03 7.527 -0.399 0.498 1.00E+00 2 “Modes” 7.611 0.88E-03 0.049 -0.311 0.372 0.763 -14.154 0.900 -0.517 1.1 62.00E+00 0.053 -1.9 -100.738 1.399 0.336 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.00E+00 0.756 0.4 Peak values of story drift ratios (as % of story height) from MPA for 1.683 1.10E-02 9.705 1.018 -0.640 -1.5 10.1 -8.3 11.9 -100.36E-03 6.250 0.19E-10 3.55E-03 3.745 1.652 1.0 -100.00E+00 0.76E-03 4.575 -53.118 0.581 0.6 7.00E+00 0.00E+00 1 “Mode” 7.338 1.259 1.233 1.018 -0.705 -1.728 1.14 m) from MPA for 1.668 -13.298 -0.003 -16.37E-03 1.429 0.2 -3.5 2.02E-03 0.071 0.0 -100.00E+00 0.640 1.6 -7.6 -44.5 × El Centro ground motion Hinge Plastic Rotation (rad) “Modal” Response Combined (MPA) “Mode” 1 7.804 1.1 46.00E+00 0.2 6.879 1.135 -7.00E+00 0.781 0.6 -8.3 1.53E-03 7.2 0.980 0.37E-03 1.8 -29.36E-03 6.9 0.72E-03 7.76E-03 4.7 1.50E-10 3.1 62.00E+00 0.9 2.2 11.781 0.0 -50.331 1.351 -0.733 1.88E-03 1.72E-03 7.015 0.197 -0.0 -100.02E-03 0.222 0.1 46.0 47 .478 0.260 -14.72E-03 7.22E-10 NL RHA 1.02E-03 3.6 -44.068 0.8 -29.00E+00 0.8 -4.76E-03 4.0 3 “Modes” -32.202 8.3 Peak values of floor displacements (as % of building height = 37.4 1.1 1.304 1.Table 4.667 1.3 -3.5 Table 4.514 -1.116 1.37E-03 1.00E+00 “Mode” 3 0.0 -50.9 1.1 46.009 0.5 Peak values of hinge plastic rotations (radians) from MPA for 1.101 -0.156 -0.503 -1.2 -4.614 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” 1.2 0.222 0.55E-03 3.6 13.298 0.0 2 “Modes” -32.518 1.220 1.233 1.435 0.066 -0.018 0.6 1.3 13.911 0.667 -1.8 17.820 -7.99E-03 6.266 -0.756 0.105 0.00E+00 0.76E-03 4.00E+00 0.116 1.9 1.72E-03 7.066 -0.00E+00 0.200 8.737 1.36E-03 6.844 -7.305 -0.516 0.304 -1.426 15.220 1.007 1.

two. (3.0 cm. the third “mode” contributions do not seem especially important (Fig. As shown in Fig.4. 3.4. The story drifts are underestimated by up to 7% in the lower stories.20) with n = 1 to a target roof displacement. Static analysis of the structure for gravity loads provides the initial state—forces and deformations—of the structure. The pushover curves for the second and third modes included in Fig. 4.4.” and the errors in these estimates relative to the exact response from nonlinear RHA.7).8 present estimates of the combined response according to Eq. 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. The first “mode” alone provides adequate estimates of floor displacements.3 . and plastic hinge rotations at selected external beam end at each floor level (Table 4.18 and Tables 4.7. drifts in all stories (Table 4.11a). MPA including two “modes” underestimates the displacements of lower floors by up to 6% and overestimates the upper floor displacements by up to 22%. and the structure is pushed using the force distribution of Eq. They are now included in the pushover analysis of the structure for the first “mode” only.” however.6 and 4. Starting with its initial state under gravity loads.6 and 4.53 cm. The response contributions of the second and third modes are the same as before (Tables 4.17 and Tables 4.6 through 4.16a—which is slightly different than the one excluding gravity loads (Fig. 4.4) to roof displacement ur1o = 52.6).18 and 48 .16 are unchanged.7). 4.” but not the contributions of higher “modes.5).7). (3.18 and Tables 4.8). 4. Significant improvement is achieved by including response contributions due to the second “mode. as are the roof displacements ur 2o = 11. the results presented so far did not include gravity load effects. resulting in the pushover curve shown in Fig. 4. but it is inadequate in estimating the story drifts (Fig.20) with n = 1 (Fig.” these effects will modify the combined modal response as well as the results of nonlinear RHA. (3.6 and 4. Also presented at these roof displacements are the displacements at all floors (Table 4. 4.5 × El Centro ground motion are presented below.7 cm and ur 3o = 2.3 Modal Pushover Analysis with Gravity Loads To evaluate the accuracy of the dynamic response of the system. 4.17 and Tables 4.18) considering one. and three “modes.17 and 4. Figures 4. Obviously gravity load effects will influence the seismic demands due to the first “mode. Results of such analyses for the selected building subject to 1. The errors are especially large in the plastic hinge rotations estimated by the MPA procedure even if three “modes” are included (Fig.

Locations of plastic hinges determined by MPA considering one. Noted 60 50 40 (b) Story Drifts Error Envelope Error for Story No. (a) MPA.14. 2−"Modes" •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• • •• (c) 3−"Modes" • • • • • • • •• •• • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• • •• • • • • • • • • • (d) Nonlinear RHA •• •• • •• •• •• •• • •• •• •• •• •• •• •• • •• •• • •• •• •• •• • • • • • • •• •• •• Fig. 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 0.15.5 3 0.5 × El Centro ground motion (a) Floor Displacements 60 50 40 Error Envelope Error for Floor No. 4.5 3 Fig.8). 1−"Mode" • • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • •• • •• • • • • • •• •• • •• •• •• •• (b) MPA. and three “modes” and by NL-RHA for 1.5 1 1. 4. two. and (b) story drifts 49 . Most pushover analysis procedures do not seem to compute to acceptable accuracy plastic hinge rotations.Table 4. Errors in MPA as a function of ground motion intensity: (a) floor displacements.5 2 GM Multiplier 2.5 1 1.5 2 GM Multiplier 2.

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

V = 4952 kN. 0. 0.50.25 5 10 15 20 Roof Displacement (cm) 25 (c) "Mode" 3 Pushover Curve 12000 10000 u = 4.85.85 Base Shear (kN) 8000 6000 4000 2000 0 0 0. 1.75 0.9 cm.0.25 1 2 4 6 Roof Displacement (cm) 8 10 Fig. V = 5210 kN.85 0. noted are peak values of roof displacement for 0.14 y by Base Shear (kN) 8000 6000 4000 1. 0. α = 0.75.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36.5 0.5 0.75 Actual Idealized 20 40 60 Roof Displacement (cm) 80 (b) "Mode" 2 Pushover Curve 12000 10000 u = 9. α = 0.0.19 y by 3 2 1.16. “Modal” pushover curves with gravity loads included.85 0.5 2 3 Actual Idealized 2000 0 0 0. 2.0 × El Centro ground motion 51 .75 0. V = 7433 kN.5 0. α = 0. 4.13 y by Base Shear (kN) 8000 6000 1 1.25.5 2 3 4000 2000 0 0 Actual Idealized 0.5 1 0. and 3.3 cm.25 0.6 cm.

5 1 1.5 Story Drift Ratio (%) 2 2. Heightwise variation of floor displacements and story drift ratios from MPA and NL-RHA for 1. two. story drifts.5 1 1. 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 × El Centro ground motion. and hinge plastic rotations estimated by MPA including one.17.18. 4.5 Fig.(a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th NL−RHA MPA 1 "Mode" 2 "Modes" 3 "Modes" (b) Story Drifts Floor Floor 5th 4th 3rd 2nd 1st 5th 4th 3rd 2nd 1st NL−RHA MPA 1 "Mode" 2 "Modes" 3 "Modes" Ground 0 0. 4. and 52 .5 Displacement/Height (%) 2 Ground 0 0. gravity loads included. Errors in floor displacements.

gravity loads included (a) Floor Displacements 60 50 40 Error Envelope Error for Floor No. Errors in MPA as a function of ground motion intensity: (a) floor displacements. and three “modes” and by NL-RHA for 1.5 2 GM Multiplier 2. gravity loads included 53 . Locations of plastic hinges determined by MPA considering one.5 1 1. 4.5 × El Centro ground motion.20. 1−"Mode" • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • •• •• •• • • • • • •• •• •• •• •• •• •• (b) MPA. 3−"Modes" (d) Nonlinear RHA • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• •• •• • • • • • • • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • • • • • •• •• •• Fig. 2−"Modes" •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• •• •• (c) MPA. Noted Error (%) 30 20 6 8 1 5 2 4 3 9 7 Error (%) 30 4 5 20 6 9 7 1 10 0 0 10 0 0 8 2 3 0.19. 4. two.5 3 Fig.5 1 1.(a) MPA. Noted 60 50 40 (a) Story Drifts Error Envelope Error for Story No. and (b) story drifts.5 2 GM Multiplier 2.5 3 0.

1 4.9 -4.754 0.5 1.813 1.429 -1.305 -0.2 -0.1 21.35E-03 8.9 31.3 9.049 -0.429 1.515 -50.0 37.2 1.854 0.507 1.00E+00 0.066 -0.197 -0.037 -0.55E-03 3.5 × El Centro ground motion.19E-03 1.673 Table 4.00E-03 5.213 1.461 0.8 0.190 -0.00E+00 0.5 21.04E-02 8.130 0.19E-03 1.5 -5.2 -3.637 0.728 1.098 20.19E-04 5.199 16.821 1.1 1.998 21.530 1.5 10.4 -6.908 -1.0 -30.6 19.114 -1.23E-03 0.033 -0.057 0.11E-03 9.2 21.436 1.858 2.665 0.00E+00 0.7 -2.594 -1.921 1.399 0.877 -46.2 4.434 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.927 1.14 m) from MPA for 1.00E+00 0.983 1.00E+00 0.530 1.018 -0.4 -4.5 3 “Modes” -32.996 0.105 0.237 0.831 0.7 Peak values of story drift ratios (as % of story height) from MPA for 1.5 1.237 0.850 1.00E+00 0.667 0.04E-10 3.23E-03 3.2 16.2 4.00E-03 5.6 2.2 2.055 0.071 -0.88E-03 0.214 0.5 “Mode” 3 -1.490 -11.998 1.101 -0.434 0.0 16.854 0.330 1.125 -1.168 -0.527 -0.933 1.066 -0.466 0.514 -1.9 -6.11E-03 9.3 -100.00E+00 0.8 1.266 -0.88E-03 0.00E+00 “Mode” 3 0.371 -0.00E-03 5.8 Peak values of hinge plastic rotations (radians) from MPA for 1.998 0.00E+00 0.00E+00 3.5 54 .270 -12.850 -1.5 × El Centro ground motion.5 0.037 0.00E+00 2 “Mode” 8.2 0.263 0.351 -0.3 13.213 1.434 0.7 19.23E-03 0.00E+00 0.311 0.00E+00 0.315 0.35E-03 8.071 0.454 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” 1.8 -32.8 0.913 7.888 1.3 -100.2 9.78E-03 1.3 -22.23E-02 1.2 1.7 4.908 1.310 1.064 -10.75E-03 0.3 -100.744 1.19E-03 1.00E+00 0.55E-03 3.11E-03 9.0 -30.00E+00 0.3 1.2 -2.516 0.0 37.35E-03 8.530 1.2 9.353 -23.13E-03 5.953 15.637 0.783 1.Table 4.830 -12.068 0.00E+00 0.00E-10 NL RHA 1.9 -3.4 1.2 9.8 Table 4.257 0.860 1. gravity loads included Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Displacement /Height (%) “Modal” Response Combined (MPA) “Mode” 1 “Mode” 2 “Mode” 3 1 “Mode” 2 “Modes” 0.00E+00 0.207 18.55E-03 3.603 1.6 1.0 -100.7 16.00E-03 5.5 21.19E-03 1.00E+00 Combined (MPA) 1 “Mode” 8.00E+00 0.3 -22.754 1.4 1.5 × El Centro ground motion.8 -32.2 4.6 0.00E+00 0.102 1.319 1.88E-03 1.603 -1.109 0.114 1.821 -1.687 0.414 28.2 12.176 0.831 0.009 0.11E-03 9.156 -0.00E+00 0.822 1.00E+00 “Mode” 2 0.17E-03 9.938 1.0 2 “Modes” -32.7 1.015 0.8 28.686 -7. gravity loads included Hinge Plastic Rotation (rad) Floor “Modal” Response “Mode” 1 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 8.4 20.263 1.00E+00 0.3 -22.836 -0.35E-03 8.213 1.399 -0.00E+00 Error (%) “Modal” Response 1 “Mode” -32.4 26.1 13.00E+00 3 “Modes” 8.7 -2.8 9.4 -4.996 -0.0 -100.319 1.128 -1.23E-03 3.00E+00 0.107 1.0 37.983 1.154 0.330 1.26E-03 3.478 0.102 1.6 Peak values of floor displacements (as % of building height = 37.6 2.35E-10 3.372 0.

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

3. The first four floor displacements are within 5% of the “exact” value. with the errors being larger for the SRSS distribution. The errors in the FEMA and MPA estimates of seismic demands relative to the “exact” demands are presented in Fig.3a and Table 5.4 and Tables 5.0702 0.1.3. wherein the first two are obvious and the third is determined from response spectrum analysis of the building (Appendix C). the value determined from RHA of the first-mode inelastic SDF system for 1. Also included in these presentations are the MPA results considering three “modes” and the “exact” demands from nonlinear RHA. 0. The MPA procedure is more accurate than all the FEMA force distributions.062 0.119 0.367 (a) Uniform (b) ELF (c) SRSS Fig. pushover analyses are implemented for a target roof displacement of 52.177 0.11 0.3.11 0. Force distributions in FEMA-273: (a) “uniform”.11 0. 5.0913 0.4a. plastic hinge rotation demands in Table 5.1 through 5.0446 0.21 0.5 times the El Centro ground motion.11 0.11 0.1 demonstrate that the displacement demands are underestimated by the ELF and SRSS force distributions by 12 to 30 % at the lower six floors of the building.281 0.0654 0. 5. (b) ELF.165 0.11 0.0896 0.4.analyses. 5.112 0. 5. The three FEMA force distributions are presented in Fig.0466 0.00719 0. 5.1.2.0381 0.4.1. the floor displacement demands in Fig. and Table 5. The pushover curves are given in Fig. and (c) SRSS 56 . and the locations of all plastic hinges in Fig.3a and 5. and the displacements of upper floors are overestimated by 9-22%.0197 0. The “uniform” distribution overestimates all floor displacements by 17-28%.2.126 0. Figures 5. 5.042 0. 5. Using each of these force distributions. the story drift demands in Fig. both presented in Section 4.11 0.0981 0.0 cm.3b and Table 5.

α = 0. V = 8530 kN. V = 6897 kN.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. α = 0. α = 0. 5.8 cm. Pushover curves using three force distributions in FEMA-273: (a) “uniform”.3 cm.(a) Uniform Distribution 12000 10000 u = 33 cm. (b) ELF.24 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized 10 20 30 40 Roof Displacement (cm) 50 60 Fig. and (c) SRSS.2. gravity loads are included 57 . V = 7456 kN.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.

the MPA procedure is more accurate than all the FEMA force distributions. In contrast. For the ELF distribution. For the SRSS distribution.Figures 5.” and NLRHA.5 1 1. Modal pushover analysis procedure gives estimates better than all the FEMA force distributions. 7%. with errors reaching 37% in this example.) The pushover analysis procedures considered seem incapable of computing accurately local response quantities. This seems to be an inherent limitation of pushover analysis.5 Displacement/Height (%) 2 Floor 5th 5th 4th 3rd 2nd 1st NL−RHA MPA.2 demonstrate that the story drift demands are greatly underestimated by all the FEMA force distributions. 3 "Modes" FEMA Uniform ELF SRSS 0. Figure 5.4c and Table 5.3b and 5. MPA including three “modes. For the uniform distribution.5 1 1. (The 100% error in a 6th floor hinge is ignored as it simply represents that the MPA estimated zero rotation. and overestimated by no more than 32%. such as hinge plastic rotations. reaching 64%. reaching 35%. gravity loads included 58 . at most. larger errors are noted in the upper and lower stories. the errors are largest in the lower stories. 5. but it is still inaccurate. (a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th (b) Story Drift Ratios Floor 4th 3rd 2nd 1st NL−RHA MPA. Heightwise variation of floor displacements and story drift ratios estimated using FEMA-273 force distributions.3. with story drifts under estimated by. and Table 5. reaching 31%.5 Story Drift Ratio (%) 2 2.3 demonstrate that the hinge plastic rotations estimated by all three FEMA force distributions contain unacceptably large errors. 3 "Modes" FEMA Uniform ELF SRSS 0.4b. whereas nonlinear RHA computed an insignificantly small value.5 Ground 0 Ground 0 Fig. errors are largest in the upper stories.

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

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

839 0.8 -63.1 -26.355 0.5 -33.3 Peak values of hinge plastic rotations (radians) from FEMA force distributions and MPA Hinge Plastic Rotation (rad) MPA FEMA ELF SRSS 4.8 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 0.3 -4.207 1.1.547 -27.466 0.8 -32.5 -27.566 1.9 15.436 Error (%) NL Uniform RHA 0.875 0.45E-03 3.9 28.417 1.927 1.00E+00 8.2 -4.3 31.4 -55.6 -25.6 -41.089 1.7 Table 5.854 0.263 0.9 16.65E-03 7.8 2.724 0.4 -23.367 1.860 1.998 1.414 1.35E-10 3.098 1.34E-03 2.2 9.6 -73.195 0.178 1.730 1.0 -71.998 1.9 11.597 0.0 -30.9 16.7 FEMA ELF SRSS -22.938 1.1 163.00E+00 0.2 10.686 0.836 0.487 0.5 22.8 -100.0 -63.3 29.3 -14.310 1.462 1.4 -12.4 -50.6 17.975 1.2 16.5 -32.2 4.5 10.262 1.736 0.9 SRSS -22.7 -28.35E-03 8.3 23.7 7.992 1.5 -29.23E-02 1.306 1.1 10.830 2.344 0.00E+00 0.00E+00 0.26E-03 3.19E-04 5.011 1.0 37.789 0.88E-03 1.7 MPA -2.007 1.877 1.490 0.5 -3.953 0.Table 5.84 1.399 1.10E-02 7.353 1.13E-03 5.16E-03 0.667 0.5 15.04E-02 8.199 27.23E-03 3.9 -77.888 0.00E+00 0.11E-03 9.58E-04 6.530 1.2 -100.62E-03 0.6 -4.4 21.0 -57.168 1.858 1.0 -100.2 9.9 0. gravity loads included Displacement /Height (%) FEMA ELF SRSS MPA 0.2 6.5 26.524 0.323 1.5 61 .913 0.0 -100.00E-03 5.3 -11.4 16.2 19.128 1.0 -59.8 7.6 -11.623 1.4 16.234 1.2 13.209 0.0 -100.673 Error (%) NL RHA 1.52E-04 1.8 17.6 4.214 1.9 -20.7 21.611 0.783 1.221 1.5 Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 2.3 -22.59E-03 5.5 -33. gravity loads included Displacement /Height (%) FEMA ELF SRSS MPA 1.9 -70.2 Peak values of story drift ratios (as % of story height) from FEMA force distributions and MPA.351 0.015 0.09E-03 4.318 2.560 1.7 19.984 1.335 2.1 0.7 -29.154 1.6 21.0 Table 5.330 1.515 FEMA Uniform ELF 27.067 0.8 -35.399 1.809 0.53E-02 1.55E-03 3.530 1.2 5.00E+00 2.03E-03 5.50E-03 0.6 22.7 26.061 1.6 -17.672 1.4 0.8 -2.51E-03 4.562 1.94E-03 2.372 1.064 1.2 16.78E-03 0.4 MPA -2.8 44.00E+00 0.00E-10 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 1.7 -15.00E+00 0.341 1.75E-03 0.6 -6.3 -100.109 1.7 -60.93E-03 1.0 -100.00E+00 Error (%) FEMA Uniform ELF SRSS MPA 24.2 -12.270 0.8 16.78E-03 1.083 1.00E+00 NL RHA 1. Peak values of floor displacements (as % of building height = 37.17E-03 9.294 1.5 -29.14 m) from FEMA force distributions and MPA.708 0.314 1.1 -3.399 -27.19E-03 1.6 14.00E+00 4.8 -100.

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

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

5 1 1. Noted 80 40 9 1 Error (%) 7 Error (%) 60 30 20 10 2 5 6 4 8 40 1 6 20 2 5 3 4 3 7 0 0 0.5 2 GM Multiplier 2.5 1 1.5 3 Fig.5 2 GM Multiplier 2.5 2 GM Multiplier 2.5 1 1.7. Error in story drifts from three force distributions in FEMA-273 and from MPA including three “modes”.(a) FEMA: Uniform 100 60 50 9 8 (b) FEMA: ELF Error Envelope Error for Story No.5 1 1.5 3 0 0 0.5 2 GM Multiplier 2. 5.5 3 0 0 0.5 3 (c) FEMA: SRSS 60 50 40 60 50 40 3 2 1 (d) MPA (3 "Modes") Error (%) 30 20 Error (%) 30 5 4 8 9 20 9 6 7 7 10 0 0 4 10 2 3 1 8 5 6 0. gravity loads included 64 .

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

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

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

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

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

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

1. Define the anchor point. 1988). The structure should be pushed just beyond the target (or expected) roof displacement in the selected mode. 3. ω n .Appendix A Uncoupled Modal Response History Analysis A. Let the roof displacement and base shear at the anchor point be urno and Vbno .20): s* = mφn n n 2. e. Calculate the area under the actual pushover curve. trapezoidal rule.g. 1977). 1. 3. 2. for linear-elastic vibration of the building. 71 . Define the force distribution s* from Eq. Idealize the pushover curve as a bilinear curve (Fig. develop the base-shear – roof-displacement (Vbn . iterations may be necessary. using any numerical integration method. and modes. This step can be conveniently implemented in any commercially available software..g.. Since the target roof displacement may not be known at the start of the procedure. e. Apn . (3.2.1 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. Apply force distribution of Step 2.2.urn ) pushover curve for the force distribution s* : n 2. Compute natural frequencies.1 incrementally and record the base shears and associated roof displacements. respectively. of the bilinear curve at the target roof displacement. 3. φn . B.1) using the FEMA-273 procedure (Building Seismic Safety Council.1. For the nth-“mode”. DRAIN2DX (Allahabadi and Powell. A.

0.4. This step gives the secant stiffness at a base shear equal to 60% of the yield base shear.3. O. Calculate Vbny = Vbny × Apn Abn .1. by connecting a straight line between origin. Abn .8. kn .4.6.1. Let the point with base shear = Vbny and roof displacement = urny be denoted as A.7. obtained by judgment. at base shear i equal to 0. Draw the curve OAB by connecting the three points O. corresponding to the estimated yield i i i base shear. A. 3. i i i αn = Vbno Vbny − 1 urno urny − 1 i 3. Vbny .6 .4. If desired. will be refined by an iterative procedure that seeks to equate areas under the actual and the idealized pushover curves. 72 .9. and B with straight-line segments to obtain the idealized bilinear curve. Estimate the yield base shear. i 3. iterations are necessary. Vbny .8. urn.6 × Vbny .4 to 3.2. kn = 0. i i 3.i 3. This value.6 × Vbny ( ) i urn .6 × Vbny . Calculate initial slope of the idealized bilinear curve. other appropriate methods can be ( used. If the error exceeds some pre-specified tolerance.9.6 . i 3. and a point on the actual pushover curve with base shear equal to i 0.2. i +1 i i 3. Calculate the error = 100 × Abn − Apn ( ) ( ) ( ) ) Apn . urny = Vbny kn . i i i 3. Calculate the slope. From the pushover data.9.5. Calculate the yield displacement. 3. i 3. determine the roof displacement. Replace i+1 with i and repeat Steps 3. Calculate area under the bilinear curve OAB.0. 3. Calculate the post-yielding strain-hardening ratio.

Compute deformation history.1. 4.6 ur n y ur n o ur n Fig. Scale the vertical axis with * Mn to obtain * Fsno Ln = Vbno M n and * Fsny Ln = Vbny M n (Eqs. Develop the Fsn Ln − Dn relation (Fig.0. A. * 4.11a). 4. Vbn Idealization of Pushover Curve Vbno Vbny 0. (3. In general first two or three modes will suffice.4. Scale the horizontal axis by Γnφrn to obtain Dno = urno Γ nφrn and Dny = urny Γ n φrn (Eqs.10a and 4. and pseudo-acceleration history. Combine the “modal” responses using Eqs. r o .16). Calculate peak values. 8. (3. 4.13).2. of the combined responses obtained in Step 8. (3.2). 4. Repeat Steps 3 to 6 for as many modes as required for sufficient accuracy. 6. Calculate histories of various responses using Eqs. Dn (t ) . 4.11b).3. 5.2.3b) with unit-mass and force-deformation relation of Fig. An (t ) . 9.4) and effective modal mass from M n = Ln Γn . Compute the Ln and Γ n from Eq.12) and (3.10b and 4. of the nth“mode” inelastic SDF system (Fig. A.15) and (3. 7. A.6 × Vbny A • Idealized αnkn B • 1 Actual k 1 • O n figA_1.1 Idealization of nth-“mode” pushover curve 73 .eps ur n.

5 cm (25 in.4. First three mode shapes and frequencies of the selected building were computed and are shown in Fig.5.6 kN. The pushover curves for the first three modes. 25.5 cm and Vb1o = 8729. generated using DRAIN-2DX. and third mode.1.2.2. s* . Following is step-by-step implementation of the procedure described in Section A. A. The force distributions.3.1. Ap1 = 360777 kN-cm. A. 3. 2.2 EXAMPLE The UMRHA procedure is implemented to calculate the response of the 9-story building described in Section 3. 3. The following steps illustrate the procedure to develop the idealize curve for the first “mode”. and 12.eps Dny = ur n y / Γnφr n Dno = ur n o / Γnφr n Dn Fig. second.). for the first.4 cm (10 in. to the north-south component of the El Centro (1940) ground motion scaled up by a factor of 1.7 cm (5 in. B.Fsn / Ln Fsn / Ln − Dn Relationship Vbno / M* n Vbny / Mn * 1 α ω2 n n ω2 1 n figA_2. 3. n 2. is defined at the target roof displacement. 3. ur1o = 63.). 3.1. Target displacements used to generate these pushover curves are 63. At this point.).3. are shown in Fig. The base-shear – roof-displacement (Vbn . 1.4. Idealized bilinear curves for each of the three modes are included in Fig. computed for the first three modes are shown in Fig.urn ) pushover curve for the force distribution s* : n 2. respectively.3. Area under the actual pushover curve. The anchor point. A. 74 .2 Properties of the nth-“mode” inelastic SDF system A.

The procedure converged after nineteen cycles to give ur1 y = 36.8 22. i i i 3.8 kN. i i i 3.1. i i 3.6 8006. 3. Vb1 y = 7615.198%.8. Ab1 = 365100 kN-cm.4.09 cm.4 kN. The point A on the i i bilinear curve is defined by ur1 y = 38. 75 .23 cm. The Fs1 L1 − D1 relation for the first “mode” is developed as follows.01%. * 4. The results of the iterative procedure are summarized in Table A.2.3.4 210.i 3.6 × Vb1 y = 4803. (4. ur1. Γ1 = 1. 4. i 3. k1 = 0. is calculated as follows. The initial slope of the idealized bilinear curve. The yield displacement. A.4 kN. k1 .5.7. The results for other modes are summarized in Table A. The curve OAB obtained by connecting the three points O.6. 3.86 cm at 0.1.4 ) − 1 (63. α1 = Vb1o Vb1 y − 1 ur1o ur1 y − 1 = ( ) ( ) (8729.6 × Vb1 y ( ) i ur1. i i 3.6 kN. and B with straight-line segments gives the idealized bilinear curve.194.09 ) − 1 = 0. iterations are necessary.2.6 = 22. Determined from the pushover database.3666. Error = 100 × (365100 − 360777 ) 360777 = 1.1.18 = 38.9. Area under the bilinear curve OAB.0. 3.9 kN. ur1 y = Vb1 y k1 = 8006. i 3.86 = 210. and α1 = 0. This value exceeds the prespecified tolerance of 0.09 cm and Vb1 y = 8006.135.2.6 = 4803. Therefore.0.9.3666 = 3740189 kg.5 38.9.18 kN/cm. and M1 = 2736789 × 1. L1 = 2736789 kg. The next estimate of the yield shear is Vbny = 8006. The first estimate of the yield base shear Vb1 y = 8006.1.4 × (360777 365100 ) = 7911.4.13). i +1 3.4. Also included are the modal damping ratios and the periods calculated from Eq. The post-yielding strain-hardening ratio.

4. Scaling the vertical axis by Fs1 y L1 = 203.46 cm and D1y = 26. 7. 4.3.8. 8. The peak values are also plotted in Fig. 4.40 (cm/sec2) and 76 . Histories of roof displacement and top story drifts for the first “mode” are computed and presented in Fig. A. 5. * M1 gives Fs1o L1 = 233. 9. The combined modal responses are presented in Fig.51 cm. 6.2.7.62 (cm/sec2).7.1 and 4.2. 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. 4. Scaling the horizontal axis by Γ1φr1 gives D1o = 46. The results were generated for first three “modes’ and are included in Fig.4.4.7. The peak values are computed and are summarized in Tables 4. 4.

14 y by 6000 Base Shear (kN) 4000 2000 Actual Idealized figA_3c.eps 0 0 5 10 15 20 Roof Displacement (cm) (c) "Mode" 3 Pushover Curve 25 8000 u = 4.13 6000 Base Shear (kN) 4000 2000 Actual Idealized figA_3b.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36.9 cm. V = 5210 kN. α = 0. “Modal” pushover curves for the example building 77 . A.2 cm.3. α = 0. α = 0. V y by = 4952 kN.eps 0 0 2 4 6 8 10 Roof Displacement (cm) 12 Fig.19 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized figA_3a.6 cm. V = 7616 kN.eps 10 20 30 40 50 Roof Displacement (cm) (b) "Mode" 2 Pushover Curve 60 8000 u = 9.

52 5 10 15 20 Time (sec) 25 30 0 • 1.5 × El Centro ground motion for the first “mode.” and third “mode” inelastic SDF systems.2769 D1 (cm) 0 • 35.” second “mode.m −80 0 −3 0 25 30 Fig.06 0 • 1. 78 .33 A1 (g) n=2 0 −80 80 −3 3 n=2 D2 (cm) 0 A1 (g) • 22. A.(a) Deformation 80 n=1 3 (b) Pseuso−acceleration n=1 • 0. Histories of deformation and pseudo-acceleration due to 1.4.747 5 10 15 20 Time (sec) figa_4.223 −80 80 n=3 −3 3 n=3 D3 (cm) 0 A3 (g) • 10.

18 210.18 210.Table A.529 0.18 210.693 0.0 4588.237 0.56 47.162 0.9 7615.194 0.0 4704.048 0.2 4571.74 (kN/cm) 210.013 0.1 7616.86 22.79 21.18 210.74 21.18 210.3 7786.190 0.56 19.35 36.194 (kN-cm) 365100 364060 363276 362684 362235 361892 361631 361432 361279 361162 361073 361004 360951 360911 360880 360856 360838 360824 360813 Table A. i 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 i Vb1 y i 0.81 21.18 210.139 0.24 36.46 2.62 26.02 21.136 Fsny Ln (cm/sec2) Dny (cm) Fsno Ln (cm/sec ) Dno (cm) Tn (sec) 233.2. Properties of “modal” inelastic SDF systems Properties Ln (kg) Γn * M n (kg) “Mode” 1 2736789 1.2 4628.4 7911.182 0.1 4569.6 4583.4 4595.50 36.2 4671.44 36.30 37.5309 488839.0 7619.1.25 36.6 i k1 i ur1 y i α1 i Ab1 Error (%) 1.18 210.78 21.022 0.39 36.910 0.135 0.23 0.193 0.75 21.65 1226.5 7624.64 37.32 36.3 7628.05 36.170 0.192 0.8 7618. Results of iterative procedure to develop the idealized bilinear curve for the first “mode” inelastic SDF system Itr.40 46.186 0.38 22.948 ζ n (%) 79 .107 0.09 37.151 0.194 0.191 0.3666 3740189 203.8 4647.18 210.194 0.18 210.28 36.184 0.063 0.037 0. No.3 4603.4 7647.103 “Mode” 3 696400 0.59 36.010 (kN) 8006.86 21.18 (cm) 38.193 0.309 0.198 0.029 0.25 36.18 210.75 21.18 210.3 7658.2671 1.90 21.11 22.6 7840.082 0.8525 1.83 21.7 4580.76 21.7 7639.8 7622.09 18.188 0.4 7714.3 7745.1 4574.404 0.51 2 “Mode” 2 -920860 -0.5 4614.6 × Vb1 y i ur1.95 21.176 0.18 210.0.1 1013.18 210.18 210.2406 167531.59 22.2 7690.9 (kN) 4803.5 7633.70 36.26 36.05 52.0 4577.193 0.4 7672.12 3876.18 210.9 4573.4927 1.8 4747.5 3109.18 210.017 0.23 22.85 0.77 21.9 4570.29 36.18 210.18 210.5 (cm) 22.79 0.9 4570.85 36.180 0.

80 .

1. extract from the pushover database values of other desired responses. (4.8).3b) with unit mass and force-deformation relation of Fig.1 to the north-south component of the El Centro (1940) ground motion scaled up by a factor of 1. Compute the peak deformation.4.21). (4. 12. Calculate the peak roof displacement urno associated with the nth-“mode” inelastic SDF system from Eq. 11. the first two or three “modes” will suffice. of the nth-“mode” inelastic SDF system (Fig. or from the inelastic response (or design) spectrum.5. From the total hinge rotation. Following is step-by-step implementation of the procedure described in Section B. 13. Typically. Steps 1 to 4 of the MPA are the same as those for UMRHA presented in Appendix A.2 EXAMPLE The MPA procedure is implemented to calculate the response of the 9-story building described in Section 3. Solving Eq. Determine the total response by combining the peak “modal” responses using the SRSS combination rule of Eq. subtract the yield hinge rotation to determine the plastic hinge rotation. B. rno .6b by solving Eq.33 cm. Repeat Steps 3 to 7 for as many “modes” as required for sufficient accuracy. (3. This is the same example as solved in Appendix A. 4. 10. At urno .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. (3. 6.Appendix B Modal Pushover Analysis B. 4.18). 81 .1 STEP-BY-STEP PROCEDURE A detailed step-by-step implementation of the modal pushover analysis (MPA) procedure is presented in this section and illustrated by an example in the next section. 10. Dn .

where results for other ground motion intensities are also included. At ur1o = 48.18) are also included in Tables 4.3 and 4.5 are the plastic hinge rotations computed by subtracting the yield hinge rotation from the total hinge rotations.33 = 48. Peak roof displacement ur1o = Γ1 × φr1 × D1 = 1. The total response computed by combining the peak “modal” responses using the SRSS combination rule of Eq. 12. Results of Steps 5 and 6 are summarized in Table B. values of floor displacements and story drifts are extracted. 8.366 × 1 × 35.” and the results are included in Tables 4.28 cm. Also included in Table 4. 82 .1. Steps 3 to 7 are repeated for first three “modes.3 for the floor displacements normalized by the building height and in Table 4. 9.4 for the story drifts normalized by the story height. (3.3 and 4.28 cm. The values are summarized in Table 4.11.4.4.

275 1.457 12.184 0.154 78.37 1.59 0.535 14.755 0.856 31.913 22.690 10.4222 3.367 1.577 16.25 Quantity Dn (cm) “Mode” 1 6.691 0.Table B.901 8.35 0.73 24.70 0.395 0.25 20.13 2.467 14.379 21.551 2.748 63.678 0.03 26. Calculation of roof displacements urno from peak deformation of inelastic SDF systems Ground Motion Multiplier 0.225 2.82 1.023 0.267 5.75 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.52 “Mode” 3 1.766 7.332 48.268 0.5 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 2.332 13.5 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.312 1.126 13.450 4.660 14.007 36.71 1.35 1.28 46.85 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 1.07 “Mode” 2 4.79 0.755 27.36 1.504 18.1.229 8.139 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.50 35.27 0.03 0.52 0.8451 5.0 µ = Dn Dny urno = Γnφrn Dn (cm) 83 .252 9.18 27.735 3.38 22.513 0.200 0.117 5.0 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 3.969 0.436 7.185 11.676 6.33 1.0 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 1.05 0.37 57.06 1.

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

6 319.2 285.6 1233.3.1 87.6 -352.070 0.1 -438.7 374.042 0.9 832.0 354.098 0.2 95.5 215.2 105.5 -350.2 fj = ∑ fi i fj (2) (3) (4) 59.1 1857. 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) 1917.0 1231.090 0.9 880.2 200. Table C. the lateral forces are normalized by the base shear to obtain column 10 in Table C.3 222.7 (10) 0.4 1759.1 -525.9 1683.1 832.3 240.7 -46.5 -973.6 V j3 (kN) (7) 478.065 0.5 -320.8 -326.2 148.6 366.7 1578.047 0.9 -166.3 -6.1 Story Shears V j2 (kN) (6) 1114.3 -646.6 -732.7 Vj (kN) (8) 2268.For convenience.7 355.1 -967.8 374.1 -646.6 286.4 400.0 136.7 694.0 176.9 -153.6 -359.0 1476.9 366.5 159.2 97.2 Lateral Force fj (kN) (9) 203.8 1381.045 0.7 95.8 430.4 1842.367 87 .177 0.0 980.3.7 1622.0 -5.7 2065.7 101.4 250.5 320.7 525.9 446.7 234.0 381.2 277.9 1446.

- 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

a research report modal pushover analysis

a research report modal pushover analysis

- Nonlinear Pushover Analysis of Reinforced Concrete Structuresby betacivil
- Performance Based Seismic Designby ijump12
- 2013 Pushover Analysis Guide Final RC Building With Masonry Infill Walls EditedDraft January 2013by Juan Carlos Jiménez Pacheco
- Performance Based Design Using Nonlinear Analysisby ronald.imme

- Pushover analysis
- Pushover Analysis using ETABS and SAP2000
- EVALUATION OF PUSHOVER ANALYSIS PROCEDURES
- Pushover Ppt
- Pushover Analysis Final
- Non Linear Analysis Pushover
- A comparison of single-run pushover analysis techniques for
- Non-Linear Static Push Over Analysis
- Nonlinear Pushover Analysis of Reinforced Concrete Structures
- Performance Based Seismic Design
- 2013 Pushover Analysis Guide Final RC Building With Masonry Infill Walls EditedDraft January 2013
- Performance Based Design Using Nonlinear Analysis
- Performing Pushover Analysis in Etabs
- Action Plan for Performance Based Seismic Design
- SAP-Etabs-Course-360
- Etabs-step by Step
- Pushover
- Pushover Analysis Using Etabs
- Pushover Analysis
- ETABS Pushover Analysis
- A Modal Pushover Analysis Procedure to Estimate
- Thesis
- Pushover
- Building report
- Etabs Guide
- ATC-40
- 20120424 Gen Advanced Webinar Pushover Analysis
- 15WCEE 1940 pushover analysis.pdf
- Pushover
- PUSHOVER Report Chopra[1]

Are you sure?

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

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

Get the full title to continue

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

scribd