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 .

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

iv .

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

vi .

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

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

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

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

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

deformation is denoted by um .3). the deformation u t depends on three systems parameters: w n .ug t n where bg bg b g bg (2. or absolute (without regard to algebraic sign) maximum. and the ductility factor is u m= m uy (2.3) ωn = k m ζ = c 2mω n f fs = s fy (2. z . This becomes evident if Eq.4) and w n is the natural vibration frequency.e. Section 7. 2001..1. z y . and (b) bilinear hysteretic forcedeformation relation For a given excitation u g t . (2. in addition to the form of the force-deformation relation. m depends on three system parameters: w n . sign u = . bg 4 . and u y . Tn = 2p w n is the natural vibration period. u £ u y ). and R y (Chopra . The peak.5) For a given u g t .1) is divided by m to obtain ~ u + 2zw n u + w 2 u y f s u. and z is the damping ratio of the system vibrating within its linear elastic range (i. (a) Idealized one-story structure.fS m fs c ut u fy 1 αk k 1 uy ug -fy k k um u 1 1 (a) (b) Fig. 2.

One-story.2. f o w = 184 .9 kN (38. this system and excitation. Vb (kN) 30 20 10 0 0 2 4 6 Roof Displacement.32 m • Fig. 2001) and scaled up by a factor of 2.66 m • EIb EIc EIc • • L = 7.826 kips) for w = 169.26 kN. 2.376 cm. For .26 kN (8. m h = 3. with Tn = 0.3. Pushover curve for structure shown in Fig. 2.2 5 .2311 . 2.4 in Chopra. The yield strength of the inelastic system.04 40 Base Shear. u (cm) 8 Fig. 2. 6. one-bay frame 50 uy = 1.2 SYSTEM AND EXCITATION CONSIDERED Consider the one-story system in Fig.2 kips) .2: the dimensions and flexural rigidity of the structural elements are noted.2. is f y / w = ( f o / w ) ÷ 8 = 0.1.5 sec and z = 5% subjected to the north-south component of the El Centro (1940) ground motion (Fig. based on Ry = 8 . α = 0. Vby = 39. and f y = 39.

The yielding stiffness of each structural element is defined as 3% of its initial stiffness. 2.0217 rad .077 × 107 mm 4 (146 in.4e.6 kip-in.4 ). the joint rotation and beam hinge rotation are identical to values q m and q pm determined from RHA. 2. 2.017 rad .4 shows the earthquake response of the system described in the preceding section determined by response history analysis (RHA).4edetermined from RHAis superimposed. The resulting pushover curve is shown in Fig.3 RESPONSE HISTORY ANALYSIS Figure 2.4 PUSHOVER ANALYSIS Static analysis of the one-story nonlinear system subjected to lateral force that increases in small increments is implemented until the lateral displacement reaches um = 7.3 in. as apparent in Fig. wherein the hysteretic force-deformation history of Fig. 2. 2.18 kN-m (444. respectively.. and q pm = 0.) and 50.1 kip-in.36 cm .134 × 107 mm 4 (75. (b) shows the lateral force f s t or base shear Vb t normalized relative to the bg weight. the ductility factor m = 5. (d) shows the rotation q p bt g of the plastic hinges at the bg bg beam ends. the energy dissipated in 6 . the peak value determined from RHA. The peak values of the various response quantities are as follows: um = 7.4 ).The yield moments in the beam and columns are defined as the bending moments due to the lateral force f y . 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 . pushover analysis cannot provide any cumulative measure of response. However. e.36 cm. I b = 3.65 kN-m (191.g. and (e) shows the force-deformation relation. It is organized in five parts: (a) shows the deformation u t .35. A static pushover analysis of this one-story system leads to the force-displacement relationship shown in Fig.) for the beam and columns. The system is excited well beyond the yield deformation. (c) shows the joint rotation q bt g . 2. and E = 2 × 108 kPa (29×103 ksi) gives yield moments of 21. Implementing this analysis with I c = 6. Observe that the pushover curve matches the initial loading path of the hysteretic system. q m = 0.3.4f. Determined from pushover analysis at the exact peak deformation.

2 −0.2311 y (b) −0.2 −0.4. and (f) pushover curve 7 .04 0.yielding during the ground motion.04 0 5 10 15 Time (sec) 20 25 30 0. (c) joint rotation. or the cumulative rotation at a plastic hinge.2311 (a) Vb / w 0 f / w = −0.4 0.2 Vb / w 0. 15 u (cm) 0 • 7.5 fy / w = 0.36 −15 0.5 0.04 0.04 0. 2.0217 • θ (rad) 0 (c) −0. Response of one-story system to El Centro ground motion: (a) deformation.4 −10 −5 0 u (cm) 5 10 −0.4 0. This represents an inherent limitation of pushover analyses. (b) base shear.017 • 0 (d) θp (rad) −0.2 0 (e) 0 (f) −0. (d) plastic hinge rotation. (e) force-deformation relation.4 −10 −5 0 u (cm) 5 10 Fig.

8 .

2001: Section 13.1 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 = .4) 9 .n t = Â .12): mι = n =1 Â sn = N n =1 Â Γ nmφ n N (3.snug t n= n =1 The effective earthquake forces can then be expressed as af N af N af (3. each element of the influence vector ι is equal to unity. and p eff t = Â peff. This force distribution can be expanded as a summation of modal inertia force distributions s n (Chopra. m.1) where u is the vector of N lateral floor displacements relative to the ground. and k are the mass. Elastic Multistory Buildings 3.3) where φ n is the nth natural vibration mode of the structure. (3.m ι u g t bg bg (3.3. c.1) can be interpreted as effective earthquake forces: p eff t = .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 . and lateral stiffness matrices of the system.mι ug t bg bg bg (3. The right side of Eq. classical damping.

subjected to u g t : Dn + 2z nw n Dn + w 2 Dn = .9) to the equation of motion for the nth-mode elastic SDF system. an SDF system with vibration propertiesnatural frequency w n and damping ratio z n of the nth-mode of the MDF system.8) where the modal coordinate qn t is governed by 2 qn + 2z nw n qn + w n qn = . with no contribution from other modes. (3.n (t ) = N n =1 ∑ −sn u g (t ) N (3. we will outline that the response of the MDF system to p eff . (3.6) Next.sn ug t bg bg bg bg (3.10) bg bg bg 10 (3. it can be demonstrated that none of the modes other than the nth mode contribute to the response. The solution qn t can readily be obtained by comparing Eq.8) gives the floor displacements u n t = Gnφ n Dn t bg (3. (3.12) .n t = . p eff .11) and substituting in Eq.n t is entirely in the nth-mode.sn ug t bg (3. The equations governing the response of the system are mu + cu + ku = .peff (t ) = n =1 ∑ peff .9) in which w n is the natural vibration frequency and z n is the damping ratio for the nth mode.5) The contribution of the nth mode to s and to p eff t are: sn = Gn mφ n respectively.u g t n Comparing Eqs.Gn ug t (3. Then the floor displacements are u n t = φ n qn t bg bg bg bg (3.7) By utilizing the orthogonality property of modes.10) gives qn t = Gn Dn t bg bg bg (3.9) and (3.

16) reflect combining the response contributions of all modes.15) rt = bg n =1 Â rn t = bg n =1 Â rnst An bt g N (3. 3. In contrast to the classical 11 bg .13) represent the response of the MDF system to p eff. Conceptual explanation of modal response history analysis of elastic MDF systems Any response quantity r t story drifts.13) define the contribution of the nth-mode to the response. and Eqs. 2001.Forces sn An(t ) st rn ωn. st The two analyses that lead to rn and An t are shown schematically in Fig.1.1). (3.15) and (3.14) is the pseudo-acceleration response of the nth-mode SDF system (Chopra. these standard equations have been derived in an unconventional way. Equations (3. (3.16) This is the classical modal RHA procedure wherein Eq.n t .12) and (3. the response of the system to the total excitation p eff t is ut = bg bg bg n =1 N Â un bt g = Â Gnf n Dn bt g n =1 N N (3. and An t = w 2 Dn t n bg (3. internal element forces. (3.1.9) is the standard modal equation governing qn t . Therefore. However. Section 13.13) st where rn denotes the modal static response. etc. ζn ¨ g(t ) u (a) Static Analysis of Structure (b) Dynamic Analysis of SDF System Fig. 3. Eqs.12) and (3. the static value of r due to external forces sn .can be expressed by st rn t = rn An t bg bg bg bg bg (3.

2001. this response value can be obtained by static analysis of the structure subjected to lateral forces distributed over the building height according to 12 . we used the modal expansion of the spatial distribution of the effective earthquake forces.4 and 13.16). Alternatively. 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. and Tn = 2p w n is the natural vibration period of the nth-mode of the MDF system.3 MODAL PUSHOVER ANALYSIS (3. 3. we note that static analysis of the structure subjected to lateral forces f no = Γ n mφ n An (3.3). the peak nth-mode response as in Eq.1).19) will provide the same value of rno . which is valid for structures with well-separated natural frequencies such as multistory buildings with symmetric plan. The SRSS rule.2 MODAL RESPONSE SPECTRUM ANALYSIS The peak value ro of the total response r t can be estimated directly from the response spectrum for the ground motion without carrying out the response history analysis (RHA) implied in Eqs. provides an estimate of the peak value of the total response: b g F N r 2 I 1/ 2 ro ª G Â no J H n =1 K 3. (3.17) where An is the ordinate A Tn . In such a response spectrum analysis (RSA).8. z n of the pseudo-acceleration response (or design) spectrum for the nth-mode SDF system. (3. 2001.1. Chopra..17) (Chopra. Sections 12.9) .18) To develop a pushover analysis procedure consistent with RSA.derivation found in textbooks (e. This concept provides a rational basis for the modal pushover analysis procedure developed later.(3. Section 13. The peak modal responses are combined according to the Square-Root-of-Sum-ofSquares (SRSS) or the Complete Quadratic Combination (CQC) rules.g.

4 COMPARATIVE EVALUATION OF ANALYSIS PROCEDURES 3. can be combined according to Eq. (3. The peak modal responses. The building has a basement level.2.21) 2 where Dn = An ω n . in both directions. region. The levels of the 9-story building are numbered with respect to the ground level (see Fig. Consulting Structural Engineers. 46th Street. 3.s* = mφ n n (3. CA 90017. Although not actually constructed. The bays are 9.73 m (150 ft) by 45. with five bays each in the north-south (N-S) and east-west (E-W) directions. with the ninth level being the roof.20) and the structure is pushed to the roof displacement. SAC is a joint venture of three non-profit organizations: The Structural Engineers Association of California (SEAOC). (3. shown in Fig.18) to obtain an estimate of the peak value ro of the total response. 3. CA 94804-4698. and California Universities for Research in Earthquake Engineering (CUREE). Typical floor-to-floor heights (for analysis purposes measured from center-of-beam to center1 2 Brandow & Johnston Associates.2). Richmond. 1660 W.4. The interior bays of the structure contain simple framing with composite floors. denoted B1. Los Angeles.19 m (122 ft) in elevation.2). each determined by one pushover analysis. California.15 m (30 ft) on center. SAC Steel Project Technical Office. rno . Third St.. 3. A benchmark structure for the SAC project. the peak value of the roof displacement due to the nth-mode. this building is 45. this structure meets seismic code and represents typical medium-rise buildings designed for the Los Angeles. 1301 S. 13 .73 m (150 ft) in plan. This modal pushover analysis (MPA) for linearly elastic systems is equivalent to the well-known RSA procedure (Section 3.1 System and Excitation Considered The 9-story structure. The columns are 345 MPa (50 ksi) steel wideflange sections.12) is urno = Gnf rn Dn (3. urno . 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. and 37. Obviously Dn and An are available from the response (or design) spectrum. was designed by Brandow & Johnston Associates1 for the SAC2 Phase II Steel Project. which from Eq. the Applied Technology Council (ATC).

and for the ninth level is 1. monolithic column pieces are connected every two levels beginning with the first level.85.65×105 kg (66. including the steel framing.2.89×105 kg (67. 14 . The 9-story N-S MRF is depicted in Fig. Each frame resists one half of the seismic mass associated with the entire structure.96 m (13 ft).7 kips-sec2/ft).4. such as those described in Gupta and Krawinkler (1999) can be used. floor slabs.of-beam) are 3. To ensure that this structure remains elastic.e. This model is based on centerline dimensions of the bare frame in which beams and columns extend from centerline to centerline. The seismic mass of the structure is due to various components of the structure. The column lines employ two-tier construction.65 m (12 ft) and for the first floor is 5. for the second through eighth levels is 9. 1988) using the M1 model developed by Krawinkler and Gupta (1998).27.20). roofing and a penthouse located on the roof.0 kips-sec2/ft). The floor-to-floor height of the basement level is 3. 0. i. The simple model adopted here is sufficient for the objectives of this study.0 kips-sec2/ft). 3. 3. and 0.49 sec. we select a weak ground motion: the northsouth component of the El Centro (1940) ground motion scaled down by a factor of 4. s* (Eq. and shear distortion of panel zones are neglected but large deformation (P∆) effects are included. The strength. respectively..01×106 kg (69. The first three vibration modes and periods of the building for linearly elastic vibration are shown in Fig.83 m (6 ft) above the center-line of the beam to column joint.07×106 kg (73. are located on the first. 3. The seismic mass of the above ground levels of the entire structure is 9. 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. for the first three modes are shown in Fig. which are seismic (tension) splices to carry bending and uplift forces. partitions. fifth. Concrete foundation walls and surrounding soil are assumed to restrain the structure at the ground level from horizontal displacement.49 m (18 ft). for the first level is 1. 3.2 kips-sec2/ft). The seismic mass of the ground level is 9. and seventh levels at 1.00×106 kg (616 kips. third. Column splices. The force distributions. The column bases are modeled as pinned and secured to the ground (at the B-1 level). if desired more complex models. ceiling/flooring. mechanical/electrical. dimension.3. The building is modeled in DRAIN-2DX (Allahabadi and Powell.sec2/ft). These force n distributions will be used in the pushover analysis to be presented later.

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

33 2. First three natural-vibration periods and modes of the 9story building 3. is shown in Figs.728 2.05 1.1 3.1 −2.2 Response History Analysis The structural response due to individual vibration modes. n = 1.67 −1.93 −1. and 3 n 3. 3. 2.04 1.7. 3.38 0.05 s * 1 s * 2 s * 3 Fig.75 1. n = 1. and 3 .5 −1 −0.31 −0.12) and (3.72 −2.05 2. Force distributions s* = mφn . (c) shows the joint rotation q n t of an external joint at 16 bg bg bg .39 3.0272 −2.6.85 sec 2 Floor 5th 4th 3rd 2nd 1st T1 = 2.12 0.37 2.4. 3.13)].3.49 sec 3 T = 0.796 0. 2.8 −2.5.5 Mode Shape Component 1 1. and 3.44 1.27 sec Ground −1.13 −1. Each figure is organized in four parts: (a) shows the roof displacement urn t .51 0. determined by RHA [Eqs.61 2.5 Fig. respectively.487 −1.5 0 0.94 2. 3.9th 8th 7th 6th T = 0. (b) shows the base shear Vbn t normalized relative to the weight W of the building. (3.4.03 −1.

especially in estimating the story drifts.1.urn relation. but three modes—perhaps even two modes—are sufficient. and ur 3o = 0. As expected. and rotations of external joints with moment connections are presented in Tables 3. The peak values of floor displacements and story drifts determined by RHA. and (d) shows the Vbn . and three vibration modes. errors generally decrease as response contributions of more modes are included. drifts in all stories. larger in story drifts. and the percentage errors due to truncation of higher modal contributions.422 cm. The same method was used to determine the peak values of many response quantities.9. are presented in Fig.15) and (3.1. in particular. The linear relationship between the base shear and roof displacement for each mode implies that the structure did not yield.8. which are listed in Tables 3. 3. ur 2o = 2. the results for the roof displacement and top-story drift are shown in Fig. 3. two. It is apparent that the first mode alone is inadequate. Also included are the combined response due to one. respectively. consistent with the increasing significance of the higher mode response among these three sets of response quantities.16)].3.2.23 cm. 3. The errors due to truncating the contributions of vibration modes beyond the third mode are negligible. two. with the first three modes provide essentially the exact response.12 cm. the peak roof displacement due to each of three modes is ur1o = 9. and 3. (3. For a fixed number of modes included. 3. 3. or all modes. three.the roof level. and even larger in joint rotations. errors are smallest in floor displacements. This is illustrated in Fig.3.8. including one. where the second and third modal responses are a larger percentage of the top story drift compared to roof displacement. Combining the modal response histories for all modes gives the total response [Eqs.2. The peak values of the various response quantities are noted in these figures. the exact response considering all modes. and 3. respectively. The peak values of displacements of all floors. 17 .

2 9.310 0.45E-03 3.177 0.012 -0.01E-04 3.13E-04 9.2 -2.09E-03 Error (%) 1 Mode -23.125 0.002 -0.350 0.0 -2.202 0.177 0.266 0.261 -0.9 8.7 4.65E-03 RHA (all modes) 2.001 -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 0.89E-03 1.9 -22.006 0.227 0.009 -0.85E-03 3.9 1.01E-04 -2.89E-03 1.226 0.5 -1.401 Error (%) 1 Mode 2 Modes 3 Modes -23.03E-03 1.054 0.74E-04 6.011 0.060 -0.2 Peak values of story drift ratios (as % of story height) from RHA for 0.275 0.042 0.124 0.78E-04 -3.121 0.097 0.227 0.295 -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 0.8 -56.325 0.9 -16.203 0.1 -2.91E-04 1.264 Error (%) 1 Mode 2 Modes 3 Modes -23.260 0.8 1.8 -10.307 0.237 0.2 -1.026 0.50E-03 4.64E-03 3.1 -2.65E-03 2.4 -10.09E-03 1.8 -1.010 -0.173 0.1 Building height is measured from the ground floor to the 9th floor.035 0.282 0.080 0.4 0.159 0.24E-03 2.2 0.237 0.2 -4.229 0.157 0.370 0.25 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 3 Modal Response Mode 1 2.060 0.152 0.26E-04 -5.4 -1.6 1.0 -0.3 -33.88E-03 2.263 0.199 0.060 -0.399 0.5 -2.205 0.2 -57.7 2.7 3.73E-03 3.253 0.089 0.2 -20.3 -0.3 Peak values of joint rotations (radians) from RHA for 0.400 0.0 -46.3 -0.9 -24.03E-03 3 Modes 2.156 0.50E-03 2.5 18.15E-03 4.99E-03 2.4 -41.8 -5.1 Peak values of floor displacements (as % of building height = 37.024 -0.364 0.3 -3.133 -0.045 0.72E-03 3.300 0.09E-03 2.6 0.6 -0.6 4.38E-03 2.56E-03 2.1 -2.4 -3.1 4.6 9.012 0.6 -1.008 -0.062 0.14 m3) from RHA for 0. 18 .00E-03 1.03E-03 -6.225 0.9 -23.130 0.76E-03 1.0 -2.7 -19.33E-03 2.311 0.245 0.260 0.032 -0.156 0.231 -0.124 0.003 0.14E-03 2.071 0.11E-04 -5.29E-03 2.8 -15.9 -15.259 0.235 0.11E-03 1.76E-03 1.90E-03 3.8 1.303 0.197 0.022 0.1 -0.7 7.3 19.266 0.94E-03 2.023 0.9 3.6 11.069 0.088 -0.4 -1.453 0.42E-04 1.173 0.5 0.202 0.1 3.2 1.069 0.282 0.183 0.4 -22.042 0.44E-03 1.0 7.117 0.63E-03 2.3 Table 3.011 0.125 0.1 -14.466 0.259 0.179 0.08E-03 2.192 0.1 -19.74E-03 1.00E-03 1.6 0.44E-03 3.03E-03 1.00E-03 2.181 0.6 -1.66E-05 -3.475 0.4 -53.311 0.055 0.9 2.74E-03 1.038 0.258 0.28E-04 1.097 0.9 9.235 0.058 -0.0 -10.47E-03 1.31E-03 Mode 2 -1.13E-03 2.043 0.88E-03 2.378 0.99E-03 Mode 3 3.097 0.31E-03 Joint Rotation (rad) Combined (RHA) 1 Mode 2.406 0.38E-04 2 Modes 2.63E-03 2.253 0.413 RHA (all modes) 0.7 -50.407 0.266 0.4 -6.062 -0.265 0.09E-03 1.22E-03 2.336 0.152 0.7 2 Modes -3.202 0.333 0.245 0.1 -11.015 -0.265 RHA (all modes) 0.74E-04 9.1 Table 3.090 0.229 0.226 0.321 0.3 -8.0 3 Modes -5.4 -7.4 -0.Table 3.

1 5 x 10 −3 θ (rad) 0.0435 9. (c) joint rotation.5 9. (b) base shear.5 0 u (cm) r1 7.0435 Vb1 / W 0.5 0 u (cm) r1 7.1 0.1 /W 0.5 15 −0.1 −15 −7.12 • 0 (a) r1 −15 0.5.1 0.12 0 (d) 0 (e) 15 Fig. and (e) pushover curve.1 −15 −7. Excitation is 0.25 × El Centro ground motion 19 . (d) force-deformation history. 3.00174 • 0 (c) r1 −5 0 5 10 15 Time (sec) 20 25 30 0.0435 • 0 (b) V b1 −0. Response due to first mode: (a) roof displacement.12 −0.15 u (cm) 9.

1 −5 −2.5 0 u (cm) r2 2. Response due to second mode: (a) roof displacement.1 −5 −2.23 0.0252 • 0 (b) V −0. Excitation is 0.23 −5 (a) 0.00244 • 0 (c) r2 −5 0 5 10 15 Time (sec) 20 25 30 0. (c) joint rotation. (d) force-deformation history.25 × El Centro ground motion 20 .6.5 u (cm) r2 0 • 2. 3.5 0 u (cm) r2 2.05 /W 0.1 0.23 −2.0252 −2.0252 (d) 0 0 (e) −0.5 5 −0. and (e) pushover curve.5 5 Fig.1 Vb2 / W 0. (b) base shear.05 5 x 10 −3 b2 θ (rad) 0.

422 (e) −0. and (e) pushover curve. (d) force-deformation history.0108 −0. Excitation is 0.0108 −0.000913 • 0 (c) r3 −2 0 5 10 15 Time (sec) 20 25 30 0.02 /W V b3 0 • 0.5 0 u (cm) r3 0.02 2 θ (rad) 0.422 −1 (a) 0.25 × El Centro ground motion 21 .5 1 Fig.02 −1 −0. (c) joint rotation.7. (b) base shear.02 0.02 −1 −0. 3.02 −0.422 /W V b3 0 (d) 0 −0. Response due to third mode: (a) roof displacement.5 1 −0.5 0 u (cm) r2 0.1 u (cm) r3 0 • 0.0108 x 10 −3 (b) −0.

i.8) leads to the pushover curve shown in Fig. and ur 3o = 0. and the computed response values are identical to the peak response values determined from RHA (Tables 3..3).2. These values of the response quantities are identical to the peak response values determined from RHA (Tables 3. which would provide the Dn value to be substituted in Eq. and 3. Figure 3. 3. and 3. and 3.4.6d and 3. story drifts.4.1 through 3.9 and Tables 3. 3. implying that the modal truncation errors are small if at 22 . confirming that MPA gives the exact values of the individual modal responses. and 3. (3. leads to the pushover curves shown in Figs. the floor displacements are proportional to the mode shape φ 1 because the structure remains elastic. respectively. this value would be determined directly from the response (or design) spectrum. The floor displacements.18).20) with n = 2 and 3 up to roof displacements ur 2 o = 2. and external joint rotations computed by pushover analysis are presented in Tables 3. Observe that the target roof displacement in each pushover analysis is identical to its exact value determined by RHA.422 cm . two.4. i.3 Modal Pushover Analysis Implementing MPA for the fundamental vibration mode. In RHA the errors arise only from truncating the responses due to higher modes.7d). respectively. As suggested by Eq.10 and Tables 3. (3. This pushover curve is consistent with the relationship between the base shear and roof displacement determined by RHA (Fig.5.7e and to the floor displacements.1.12 cm. respectively.5e. 3. As for the first mode.21). the errors in the MPA results are generally larger than in RHA (Fig.6.e. 3.ur relations determined by RHA (Figs. 3. 3.3.6.20) with n = 1 (Fig. although both analyses led to identical peak values of the individual modal responses.6 present estimates of the combined response according to Eq. 3.4) to roof displacement ur1o = 9. (3. 3. pushing the structure. and external joint rotations in Tables 3.9).5d). pushing the structure using the force distribution of Eq. considering one. Implementing pushover analysis for the second and third modes.3). 3. For a fixed number of modes included considered. 3. (3.6e and 3.12). except for the algebraic sign associated with Γ1 and minor round-off errors.3).2. using the force distribution of Eq. In practical application. 3. and it is apparent in the example considered that three modes provide most of the response (Fig.5.e. these pushover curves are consistent with the Vb . and 3.4. (3.5. the value determined by RHA (Fig. 3. 3. or three vibration modes. story drifts.1.23 cm . and the errors in these estimates relative to the exact response from RHA considering all modes..

eps Ground 0 0.3 0.685 • −3 3 Mode 2 ∆ (cm) u (cm) • 2.2 0.1 0.422 r2 0 • 1. 3.5 Ground 0 0.2 0.83 • r3 0 • 0.25 × El Centro ground motion 23 .25 × El Centro ground motion: first three modal responses and total (all modes) response (a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th (b) Story Drift Ratios Floor 4th 3rd 2nd 1st Floor 5th RHA All Modes 1 Mode 2 Modes 3 Modes 5th 4th 3rd 2nd 1st RHA All Modes 1 Mode 2 Modes 3 Modes fig3_9a.3 0.4 Story Drift Ratio (%) 0.12 • r1 −15 15 0 r1 0 0 0.48 −3 3 1.03 −3 3 Mode 3 ∆ (cm) u (cm) r3 −15 15 9. Heightwise variation of floor displacements and story drifts from RHA for 0. sec 25 30 r 5 10 15 20 Time.eps fig3_9b.23 r2 −15 15 0 • 0.8.9. (a) Roof Displacement 15 (b) Top Story Drift 3 Mode 1 ∆ (cm) u (cm) 9. sec 25 30 Fig.eps r 5 10 15 20 Time. 3.1 0.6 Fig.least three modes are included. Response histories of roof displacement and top-story drift from RHA for 0.4 Displacement/Height (%) 0.59 • Total (All Modes) 0 −15 0 ∆ (cm) u (cm) 0 −3 0 fig3_8. Additional errors are introduced in pushover analysis due to the approximation inherent in modal combination rules.5 0.

44E-03 -1.045 -0.90E-03 1.197 0.328 0.9 -24.270 0.22E-03 -2.253 0.008 -0.267 0.062 0.267 -0.76E-03 -1.31E-03 2.08E-03 2.9 -15.270 0.5 -16.00E-03 2.89E-03 -1.080 0.0 -16.2 -0.97E-03 1.53E-04 -9.9 -11.42E-04 -1.4 -9.3 -33.229 0.203 0.322 0.062 -0.9 -18.179 0.43E-04 -1.25 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Displacement /Height (%) Modal Response Combined (MPA) Mode 1 Mode 2 Mode 3 1 Mode 2 Modes 3 Modes 0.73E-03 3.4 -22.237 0.44E-03 3.74E-03 -1.121 0.270 0.152 0.313 0.370 0.76E-03 1.179 0.89E-03 1.09E-03 Mode 2 1.152 0.9 -15.4 -22.8 -23.229 0.173 0.069 0.9 -12.Table 3.235 -0.097 0.012 0.03E-03 6.260 0.259 0.296 -0.253 0.9 -16.0 -18.157 0.63E-03 2.3 -12.282 0.4 1.7 -50.9 3 Modes -12.253 -0.3 Table 3.300 0.6 Peak values of joint rotation (radians) from MPA for 0.4 -4.09E-03 -1.374 0.2 Table 3.7 24 .032 0.65E-03 2.4 Peak values of floor displacements (as % of building height = 37.058 -0.156 0.47E-03 1.7 -19.3 -13.231 -0.9 -15.3 -2.8 -22.1 -19.9 -15.264 Error (%) Modal Response 1 Mode 2 Modes 3 Modes -23.332 0.4 -5.226 0.4 -14.227 0.117 0.9 -13.011 -0.282 -0.157 0.0 -46.2 1.05E-03 3.2 -12.125 0.7 -21.9 -14.9 -13.156 0.227 0.177 0.038 -0.024 0.237 -0.245 -0.0 -0.203 0.272 0.22E-04 2.276 0.001 0.060 0.00E-03 2.3 -41.022 0.03E-03 -1.2 -57.80E-04 3.31E-03 Joint Rotation (rad) Combined (MPA) 1 Mode 2.2 -4.25 × 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 -0.15E-03 1.230 0.321 0.042 0.060 -0.00E-03 2.203 0.048 0.036 -0.38E-03 2.310 0.00E-03 3.03E-03 3.042 0.003 -0.336 0.57E-03 1 Mode -23.401 Error (%) 1 Mode 2 Modes 3 Modes -23.2 -20.009 0.336 RHA (all modes) 0.010 0.00E-03 -1.133 -0.286 0.385 0.043 -0.3 -14.4 -19.173 0.260 0.1 -11.92E-04 -1.5 Peak values of story drift ratios (as % of story height) from MPA for 0.9 -14.048 0.235 0.1 -11.1 -18.6 -19.181 0.9 -8.4 -2.94E-03 2.078 0.071 0.4 -53.72E-03 3.090 0.2 -11.124 0.63E-03 -2.8 -15.2 -16.89E-03 2.3 -9.266 0.203 0.011 0.133 0.097 0.0 -10.96E-03 2.73E-05 3.12E-03 1.04E-03 3.99E-03 Mode 3 -3.40E-04 5.38E-03 3 Modes 2.002 0.15E-03 2.00E-03 1.1 -0.106 0.33E-04 5.133 0.8 -15.4 -9.09E-03 1.006 -0.055 0.4 -4.9 -14.125 0.90E-04 -9.89E-03 2.203 0.260 -0.274 0.6 -16.4 0.9 -13.267 0.253 RHA (All Modes) 0.285 0.088 -0.9 -13.106 0.31E-03 Error (%) RHA (all modes) 2.78E-04 2 Modes 2.331 0.6 -17.177 0.4 -7.261 -0.14 m) from MPA for 0.3 -19.0 -2.015 -0.03E-03 1.7 2.466 0.74E-03 1.24E-03 -2.09E-04 -3.7 2 Modes -13.8 -56.28E-03 2.74E-04 -6.25 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Modal Response Mode 1 -2.230 -0.069 0.7 -15.4 -11.245 0.407 0.3 1.8 -14.023 -0.079 0.6 -15.026 0.89E-03 -2.012 -0.259 -0.

eps fig3_10b.1 0. 3. Heightwise variation of floor displacements and story drifts from MPA for 0. shading indicates modal combination error 25 .2 0.5 0.eps Ground 0 0.4 Story Drift Ratio (%) 0.4 Displacement/Height (%) 0.6 Fig.1 0.2 0.3 0.5 Ground 0 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.25 × El Centro ground motion.10.

26 .

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

sign qn = φ n f s u n . (3. sign u n b g b g bg (4.n t given by Eq. Unlike Eq.2 UNCOUPLED MODAL RESPONSE HISTORY ANALYSIS bg Neglecting the coupling of the N equations in modal coordinates [Eq. give the same results for u t as obtained directly from Eq. in principle. 4.” implying that other “modes” will also contribute to the solution.4) is rarely solved because it offers no particular advantage over Eq. these equations are coupled for inelastic systems. premultiplying by φ T . This approximate RHA procedure is the preliminary step in developing a modal pushover analysis procedure for inelastic systems.2).8) because qr t will generally be nonzero for “modes” other than the nth “mode. The spatial distribution s of the effective earthquake forces is expanded into the modal contributions sn according to Eq.4) represents N equations in the modal coordinates qn . qr t = 0 for all 28 bg .6) The solution of Eq.… N (4. however.4) where the only term that differs from Eq.Gn ug t Mn bg n = 1. where φ n are now the modes of the corresponding linear system.2). (4. 2. (4.9) for linearly elastic systems. (3.5) This resisting force depends on all modal coordinates qn t .6) for inelastic systems will no longer be described by Eq.sn ug t bg b g bg bg (4. (4. (3. However. (4. implying coupling of modal coordinates because of yielding of the structure.3) in Eq.6b) are mu + cu + f s u. (4. Simultaneously solving these coupled equations and using Eq. Eq.3) will.2). and using the mass.9) involves T Fsn = Fsn qn . sign u = .and classical n damping-orthogonality property of modes gives F qn + 2z nw n qn + sn = . The equations governing the response of the inelastic system to p eff.3). (4.4)] leads to the uncoupled modal response history analysis (UMRHA) procedure. (3. (4. (3. Equation (4.Substituting Eq. (4. For linear systems.

(a) p 15

eff,1

= −s × 0.25 × El Centro

1

(b) p 5

eff,2

= −s × 0.25 × El Centro

2

ur1 (cm)

0

ur1 (cm)

9.112 •

Mode 1

Mode 1

0

−15 15 Mode 2

−5 5 Mode 2

ur2 (cm)

0

ur2 (cm)

0 • 2.226 5 10 15 20 25 30

−15 0 15

5

10

15

20

25

30

Mode 3

−5 0 5

Mode 3

ur3 (cm)

0

ur3 (cm)

5 10 15 20 Time (sec) 25 30

0

−15 0

−5 0

5

10 15 20 Time (sec)

25

30

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

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

motion

af

af

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

bg

bg

(a) p 80

eff,1

= −s × 1.5 × El Centro

1

(b) p 20

eff,2

= −s × 1.5 × El Centro

2

Mode 1

Mode 1 • 1.437

u (cm)

r1

0 • 48.24

ur1 (cm)

Mode 2

0

−80 80

−20 20 Mode 2

u (cm)

r2

0

3.37 •

ur2 (cm)

0 • 11.62 5 10 15 20 25 30

−80 0 80

5

10

15

20

25

30

Mode 3 0.4931 •

−20 0 20

Mode 3 • 0.7783

u (cm)

r3

0

ur3 (cm)

25 30

0

−80 0

5

10 15 20 Time (sec)

−20 0

5

10 15 20 Time (sec)

25

30

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

af

af

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

bg

bg

by Eq. (3.8),

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

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

b

g

b

g

(4.7)

30

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

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

bg

bg c h

(4.8)

c

h

(4.9)

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

b

g

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

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

bg

bg

bg

bg

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

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

4

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

31

n t = . 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) • 75.02 −150 n=2 −150 50 n=2 14.112 5 10 15 20 Time (sec) 25 30 −10 0 −10 0 Fig. 4.4. where u g t = 3. 4.51 • 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. 2.89 50 r1 0 u (cm) 0 • 78.Forces sn Unit mass An(t ) st rn ωn.575 5 10 15 20 Time (sec) 25 30 ur3 (cm) 0 • 5.3. Roof displacement due to p eff. n = 1.9 • 0 −50 10 −50 10 n=3 ur3 (cm) 0 • 5.0 ¥ El Centro ground motion: (a) exact solution by NL-RHA. ζn. and 3. and (b) approximate solution by UMRHA bg af bg 32 .s n u g t .

4 and 4.008 • 6. respectively.33 0 −20 20 −20 20 n=2 ∆r2 (cm) 0 ∆r2 (cm) 7. and (3) the Fsn Ln . Top story drift due to p eff.s n u g t . both for 3. 2.956 5 10 15 20 Time (sec) 25 30 ∆r3 (cm) 0 • 5. 4.n t = . Such comparison for roof displacement and top-story drift is presented in Figs.855 ∆r1 (cm) 0 ∆r1 (cm) n=2 • 6. (4.15) is strictly valid only for linearly elastic systems. The errors are slightly larger in drift than in displacement.744 • 0 −20 10 n=3 −20 10 n=3 ∆r3 (cm) 0 • 5. and 3.(a) Nonlinear RHA 20 n=1 20 (b) UMRHA n=1 • 5.n t (n = 1.5. (4. (3.Dn relation is 33 bg bg .1 Underlying Assumptions and Accuracy The approximate solution of Eq.4) and (4. n = 1. (2) the superposition of responses to p eff. the errors in either response quantity are only a few percent.5)] is neglected.6) by UMRHA is compared with the “exact” solution by nonlinear RHA. where u g t = 3.2… N ) according to Eq. and (b) approximate solution by UMRHA af bg bg 4.5.0 ¥ El Centro ground motion. but even for this very intense excitation. 4.0 ¥ El Centro ground motion: (a) exact solution by NL-RHA.2. 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. this intense excitation was chosen to ensure that the structure is excited well beyond its linear elastic limit.817 5 10 15 20 Time (sec) 25 30 −10 0 −10 0 Fig.

the relationship between lateral forces f s and Dn in Eq.1.Dn relation? The two sets of forces and displacements are related as follows: V Fsn = bn Γn Dn = urn Γ nf rn 34 (4.8) before it can be solved? Because Eq. At the yield point. Although most commercially available software cannot implement such displacement-controlled analysis.20).8) governing Dn t is based on Eq. this distribution seems to be a rational choiceeven for inelastic systemsto determine Fsn in Eq. (3. Although several approximations are inherent in this UMRHA procedure.10) . (4. (4. What is an appropriate invariant distribution of lateral forces to determine Fsn ? For an inelastic system no invariant distribution of forces can produce displacements proportional to f n at all displacements or force levels. is plotted against roof displacement urn .6a. When implemented by commercially available software. How to convert this Vbn .Dn curve. which is different than the Fsn Ln .9) should be determined by nonlinear static analysis of the structure as the structure undergoes displacements u = Dnφ n with increasing Dn .Dn relation to be determined in Eq. However. (4. the only force distribution that produces displacements proportional to f n is given by Eq. (3. the base shear is Vbny and roof displacement is urny .20) to some bg pre-determined roof displacement. within the linearly elastic range of the structure.approximated by a bilinear curve to facilitate solution of Eq. it can conduct a force-controlled nonlinear static analysis with an invariant distribution of lateral forces. The overall errors in the UMRHA procedure are documented in the examples presented in Section 4. when specialized for linearly elastic systems it is identical to the RHA procedure of Section 3. (4. such nonlinear static analysis provides the so-called pushover curve. (4.urn pushover curve to the Fsn Ln .12) for floor displacements. 4. (3. 4.9).2 Properties of the nth-“mode” Inelastic SDF System How is the Fsn Ln .4.8) in UMRHA. and the base shear Vbn . A bilinear idealization of this pushover curve for the nth-“mode” is shown in Fig. Therefore. Therefore we impose this constraint in developing the UMRHA procedure in this section and modal pushover analysis in the next section.2. The structure is pushed using the force distribution of Eq.

(4.6a is 2 kn = ω n Ln . 4. Properties of the nth-“mode” inelastic SDF system from the pushover curve Equation 4. In contrast. the elastic vibration period Tn of the nth-mode SDF system is computed from F Ln Dny I 1/ 2 Tn = 2p G H Fsny JK (4.10 enables conversion of the pushover curve to the desired Fsn Ln − Dn relation shown in Fig.2.8). The two are related through Fsny Ln 2 = w n Dny (4. the initial slope of the pushover curve in Fig. 4.12) implying that the initial slope of the curve in Fig. 2001.6b. Knowing Fsny Ln and Dny from n Eq.6b is w 2 . (4. 35 . 4.6.13) This value of Tn . 4. where the yield values of Fsn Ln and Dn are Fsny Ln = Vbny * Mn Dny = urny Γ nf rn (4. which is not a meaningful quantity. which may differ from the period of the corresponding linear system. should be used in Eq.11). Section 13.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.5).11) * in which M n = Ln Γn is the effective modal mass (Chopra.

and pseudo-acceleration history. 5. for linearly-elastic vibration of the building. Dn (t ) . 4. Convert the idealized pushover curve to the Fsn Ln − Dn relation (Fig. 4. (4. r o . 4. This value of the roof displacement is given by Eq. Combine the “modal” responses using Eqs. ωn . 7. which is summarized next as a sequence of steps. n (3. (3. 8. Repeat Steps 2 to 6 for as many modes as required for sufficient accuracy. Consider a nonlinear static analysis of the structure subjected to lateral forces distributed over the building height according to s* [Eq.20)]. and modes. 4. 4. For the nth-mode. Compute the natural frequencies. φn .20)].6a). 4. 2. of the nth-“mode” inelastic SDF system (Fig.16) to determine the total response.6b.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 . n 3.13). Calculate the peak value.4. details are available in Appendix A: 1. the first two or three modes will suffice.21) where Dn .12) and (3.3b) with force-deformation relation of Fig.3 Summary The inelastic response of an N-story building with plan symmetric about two orthogonal axes to earthquake ground motion along an axis of symmetry can be estimated as a function of time by the UMRHA procedure just developed. with the structure is pushed to the roof displacement urno . Idealize the pushover curve as a bilinear curve (Fig. 9. (3.11). An t . Typically. develop the base-shear – roof-displacement ( Vbn − urn ) pushover curve for the force distribution s* [Eq. Calculate histories of various responses by Eqs. is now determined by 36 bg bg . the peak value of Dn t .2.6b) by utilizing Eq. (3. Compute the deformation history.15) and (3.n t . of the total response r t obtained in Step 8. 6. (3.

each determined by one pushover analysis. although somewhat intuitive for inelastic buildings.2 and 3. (4. alternatively.1 Summary The peak inelastic response of a building to earthquake excitation can be estimated by the MPA procedure just developed.2..8). At this roof displacement. bg bg bg bg Thus. or from the inelastic response (or design) spectrum. are combined using an appropriate modal combination rule. As shown in Sections 3.6).n t .6b by solving Eq.12). 4. it can be determined from the inelastic response (or design) spectrum (Chopra. and (2) it provides the exact modal response for elastic systems. (3. Sections 7. 6. plastic hinge rotations. governed by Eq.3. 4. It provides results for elastic buildings that are identical to the well-known RSA procedure (Section 3. Compute the peak deformation. it seems reasonable because it provides results for elastic buildings that are identical to the well-known RSA procedure described in Section 3. the lateral force distribution used possesses two properties: (1) it appears to be the most rational choice among all invariant distribution of forces.21). rno also represents the exact peak value of the nth-mode contribution rn t to response r t .3. as described in Section 4. of the nth-“mode” inelastic SDF system (Fig. (4. for elastic systems. the pushover analysis provides an estimate of the peak value rno of any response rn t : floor displacements. However. The peak “modal” responses rno . This pushover analysis.3) because. which is summarized next as a sequence of steps. seems reasonable. Steps 1 to 4 of the MPA are same as those for UMRHA. story drifts. Calculate the peak roof displacement urno associated with the nth-“mode” inelastic SDF system from Eq. etc.solving Eq.18). 4. we will refer to rno as the peak “modal” response even in the case of inelastic systems.3b) with force-deformation relation of Fig.6 and 7. Dn . 37 . as mentioned earlier. joint rotations. e. (4. The response value rno is an estimate of the peak value of the response of the inelastic system to p eff. 2001. details are available in Appendix B. This application of modal combination rules to inelastic systems obviously lacks a theoretical basis. 5. to obtain an estimate of the peak value ro of the total response.2. (3.g. Eq.8).4.

This is to be expected.8.1).1 and 4. rno . 4.0 to 3.0. Typically. and ur 3o = 2. (3. From the total rotation of a plastic hinge. the 38 . the first two or three “modes” will suffice. 4. 9. the “exact” results.7 cm. This deficiency also implies that. and three modes are compared with the “exact” values in Fig. in particular. Determine the total response by combining the peak “modal” responses using the SRSS combination rule of Eq.1 Uncoupled Modal Response History Analysis The structural response to 1. Observe that errors tend to decrease as response contributions of more “modes” are included.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. say. ur 2o = 11. subtract the yield value of hinge rotation to determine the hinge plastic rotation.2. is presented next. 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. two. The peak values of response are as noted. the combined response due to three “modes”. 4. determined by the UMRHA procedure. three “modes” included. Figure 4. The peak values of displacements of all floors and drifts in all stories are presented in Tables 4. 4.5 ¥ the El Centro ground motion including the response contributions associated with three “modal” inelastic SDF systems. with. and compared with the results of a rigorous nonlinear RHA using the DRAIN-2DX computer program.2). respectively. the peak roof displacement due to each of the three “modes” is ur1o = 48.53 cm. and the percentage errors in the approximate results. the UMRHA procedure lacks a rigorous theory. and the errors in the approximate results are shown in Fig.7. in contrast to modal analysis (Section 3.9. extract from the pushover database values of other desired responses. 8. At urno .3 cm.4. and the “exact” response from nonlinear RHA for the roof displacement and top-story drift. two. Repeat Steps 3 to 8 for as many “modes” as required for sufficient accuracy. also included are the combined responses due to one.7 shows the individual “modal” responses. The peak values of floor displacements and story drifts including one. although the trends are not systematic as when the system remained elastic (Section 3.18). and three “modes”.4.

85. 0.53 ∆r3 (cm) 12 0 • 2. just as for elastic systems.25 (Tables 3. and 3. and total response from NL-RHA 39 . sec 25 30 • 48.25.6 5 10 15 20 Time. in particular. recall that the computed errors have been presented earlier for ground motion multipliers 0.0. (a) Roof Displacement (b) Top Story Drift 80 12 0 • 3.88 u (cm) "Mode" 3 −80 80 0 −80 80 0 −80 0 • 44.75.0.1 −12 12 ∆r (cm) u (cm) UMRHA 3 "Modes" 0 • 7.3 11.5 (Tables 4. For each excitation. For this purpose the UMRHA and exact analyses were repeated for ground motions of varying intensity.1 and 3.24 • NL−RHA r −12 12 ∆r (cm) u (cm) 0 5 10 15 20 Time. Next.62 5. 4.2). for a fixed number of “modes” included. sec 25 30 r −12 0 Fig.5 × El Centro ground motion: individual “modal” responses and combined response from UMRHA. Response histories of roof displacement and top-story drift due to 1. we investigate how the errors in the UMRHA vary with the deformation demands imposed by the ground motion. However. defined as the El Centro ground motion multiplied by 0.5.7 • −80 80 ∆r1 (cm) u (cm) "Mode" 1 −12 r2 0 ∆r2 (cm) 12 0 u (cm) −80 80 −12 r3 0 • 2.2) and 1.1 and 4.0.5. 2. 0.1 and 3. the degree to which the system deforms beyond its elastic limit.38 6. 1.7.2). 1.response is much less accurate (Tables 4.44 • "Mode" 2 r1 0 • 48. the errors in story drifts are larger compared to floor displacements.2) if the system yields significantly versus if the system remains within the elastic range (Tables 3.1 and 4. the errors in responses computed by UMRHA including three “modes” relative to the “exact” response were determined. 0.

two.5 Fig. Shown is the error in each floor displacement (Fig. To interpret these results. it will be useful to know the deformation of the system relative to its yield deformation.8.20)] 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. For this purpose.(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 0.11. 4.9. Heightwise variation of floor displacements and story drift ratios from UMRHA and NL-RHA for 1. 4.5 1 1. Two versions of the pushover curve are 40 .5 1 1. 4.5 × El Centro ground motion (a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th (b) Story Drift Ratios Floor 4th 3rd 2nd 1st UMRHA 1 "Mode" 2 "Modes" 3 "Modes" Floor 5th 5th 4th 3rd 2nd 1st UMRHA 1 "Mode" 2 "Modes" 3 "Modes" Ground −60 −40 −20 0 Error (%) 20 40 60 Ground −60 −40 −20 0 Error (%) 20 40 60 Fig. indicated by a ground motion multiplier.5 Displacement/Height (%) 2 Ground 0 0. (3. the pushover curves using force distributions s* [Eq. in each story drift (Fig.5 ¥ El Centro ground motion Figure 4. 4.10b). 4. with the peak displacement of each “modal” SDF system noted for each ground motion multiplier. and the error envelope for each case. Heightwise variation of error in floor displacements and story drifts estimated by UMRHA including one.10a).5 Story Drift Ratio (%) 2 2. or three “modes” for 1.

4) to roof displacements urno = 48. The location of plastic hinges and their rotations. 3. First.6 and 4.10) for each “modal” inelastic SDF system (Figs 4. and. Second.5 ¥ the El Centro ground motion. For more intense excitations.75. and 3 (Fig. as mentioned in Section 3. Figure 4. were noted but are not shown here. UMRHA lacks a rigorous theory and is based on several approximations. drifts in all stories (Table 4. (2) increase rapidly as the ground motion multiplier increases to 1. the peak values of displacements at all floors (Table 4. the values determined by RHA of the nth-mode inelastic SDF system (Fig.4. and (4) are larger in story drift compared to floor displacements.01 (Fig.75. Perhaps this explains why the errors are large at this excitation intensity. the errors in truncating the higher mode contributions are negligible.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. even though the system remains essentially elastic. 4. 4. the pushover curve for each “mode” is idealized by a bilinear curve in solving Eq. the ductility factor for the first mode system is only 1. (4. determined from “exact” analyses. as mentioned in Section 4.3 cm.1). 2. The structure is pushed using the force distribution of Eq. 4.10 permits the following observations regarding the accuracy of the UMRHA procedure: the errors (1) are small (less than 5%) for ground motion multipliers up to 0.4. and plastic hinge rotations at the external beam end at each floor level (Table 4. respectively. The idealized curve for the first “mode” deviates most from the actual curve near the peak displacement corresponding to ground motion multiplier 1.5). (3) maintain roughly similar values for more intense ground motions. (3.7 cm. and 2. 11.2.3). 4.7). Each of these three pushover analyses provides the pushover curve (Fig. the first reason mentioned above seems to be the primary source for the errors.11a).0. The system remains elastic up to ground motion multiplier 0.4).53 cm.20) with n = 1.included: the actual curve and its idealized bilinear version.11).2. Additional errors are introduced in UMRHA of systems responding beyond the linearly elastic limit for at least two reasons.0. 41 .

844 -25.554 1. Errors in UMRHA as a function of ground motion intensity: (a) floor displacements.376 1.350 -0.010 0.772 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 1.663 0.4 4.315 -0.298 0.(a) Floor Displacements 60 50 40 Error Envelope Error for Floor No.9 31.8 1.5 18.003 -31.0 11.9 16.338 -1.3 42 .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 -1.820 -19.513 -0.033 0.126 0.202 11.5 × 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. 4.1 Peak values of floor displacements (as % of building height = 37.220 0.2 Peak values of story drift ratios (as % of story height) from UMRHA for 1.942 1.495 1.490 1.3 1.0 2.201 -0.727 1.472 1.071 0.938 -1.820 -0.6 1.096 0.1 1.945 -37.5 3 Fig.070 1.138 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 0.366 -0.072 -1.5 -3.214 -0.8 1.241 1.2 22.10.079 0.9 12.0 7.722 0.2 1.018 0.856 2.426 -1.065 0. Noted 60 50 40 (b) Story Drifts Error Envelope Error for Story No.877 0.811 1.009 -0.806 0.490 -1.5 1 1.9 5.983 1.7 31.7 14.291 0.900 -10.863 1.763 -15.0 -9.3 0.187 -0.2 1.410 1.120 1.201 -1.214 0.676 0.283 1.668 -23.256 1.068 0.135 9.376 -1.7 Table 4.003 0.4 1.914 -0.1 8.201 1.5 9.8 0.104 0.751 1.4 -7.3 8.410 -1.5 10.5 9.130 0.256 -1.663 -0.513 0.971 1.1 1.707 1.407 -10.1 0.5 2 GM Multiplier 2.8 1.2 12.982 9.293 1.248 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.1 3.372 1.938 1.14 m) from UMRHA for 1.057 -0.2 6.0 1.8 1.5 3 0 0 0.575 -41.333 0.044 1.133 1.055 -0.169 0.371 -0.8 14.049 -0.942 -0.154 0.6 2.379 1.6 4.200 8.298 -0.819 2.914 2.366 0.5 1 1.478 0.3 25.370 -0.072 1.5 28.9 31.8 1.226 -0.260 -15.0 11.071 -0.088 10.136 1.0 9.698 1.338 1.4 -1.852 1.430 1.473 -22.235 -0.3 6.216 1.121 -0.256 1.616 -0.806 -0.373 -0.484 0.5 28.241 -1.2 4.5 2 GM Multiplier 2. and (b) story drifts Table 4.540 0.526 -0.863 0.9 12.372 -1.317 0.371 0.220 -0.

the results were not always accurate. Furthermore. 4.3 and 4. and the errors in these estimates relative to the exact response from nonlinear RHA. 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.13 and Table 4.3 and 4. and three “modes. Pushover analysis seems to be inherently limited in computing accurately hinge plastic rotations.” and three “modes.2).5). The locations of plastic hinges shown in Fig.5 present estimates of the combined response according to Eq.13 and Tables 4. were determined by four analyses: MPA considering one “mode. MPA failed to identify the plastic hinges at the column bases in Fig.1 and 4. The drifts are underestimated by up to 13% in the lower stories. However. 4. the hinges identified in beams at the 6th floor were at variance with the “exact” results.13 and Tables 4. For example. significant improvement is achieved by including response contributions due to the second “mode”. the third “mode” does not contribute because this SDF system remains elastic (Fig. overestimated by up to 18% in the middle stories. the errors in the modal pushover results are. The first “mode” alone is inadequate.3 and 4. 4. significantly smaller than in UMRHA (compare Fig. but was more successful when the excitation was more intense (results are not included). Obviously. As shown in Figs.4). One “mode” pushover analysis was unable to identify the plastic hinges in the upper stories where higher mode contributions to response are known to be more significant. but the third “mode” contributions do not seem especially important (Fig. even if three “modes” are included (Fig. 43 . The errors are especially large in the hinge plastic rotations estimated by the MPA procedure. for two or three modes included.3 and 4.” two “modes. this error is not especially significant because the hinge plastic rotation is very small.” and nonlinear RHA.11c).13 with Fig.12 and Tables 4.14.14. 4. however. 4. considering one. The second “mode” was necessary to identify hinges in the upper stories. (3. 4. 4.12 and Tables 4. the additional errors due to the approximation inherent in modal combination rules tend to cancel out the errors due to the various approximation contained in the UMRHA. 4. Fortuitously. 4.9 and Tables 4.4 with Tables 4.4).3 through 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%.12 and 4.Figures 4.18). and are within a few percent of the exact values for the upper stories. in general. although the error is recorded as 100% if MPA estimates zero rotation when the nonlinear RHA computes a non-zero value. especially in estimating the story drifts (Fig. two.4.” respectively.

α = 0.9 cm. and 3.75. V = 5210 kN.75 Actual Idealized 20 40 60 Roof Displacement (cm) (b) "Mode" 2 Pushover Curve 80 12000 10000 u = 9. 1.11. 0.13 y by Base Shear (kN) 8000 6000 1 1.25 5 10 15 20 Roof Displacement (cm) 25 (c) "Mode" 3 Pushover Curve 12000 10000 u = 4. 1.0.5 0. α = 0.25 0. 2.5 1 0.85 0.25. α = 0.5 2 3 4000 2000 0 0 Actual Idealized 0.0.85 0.2 cm.5 0. V = 7616 kN.5.6 cm.5 0.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36. 4.19 y by 3 2 1. V = 4952 kN. “Modal” pushover curves with peak roof displacements identified for 0.5.75 0.0 × El Centro ground motion 44 .5 2 3 Actual Idealized 2000 0 0 0.75 0.14 y by Base Shear (kN) 8000 6000 4000 1.85 Base Shear (kN) 8000 6000 4000 2000 0 0 0. 0.25 1 2 4 6 Roof Displacement (cm) 8 10 Fig.

3 for elastic systems (or weak ground motions). implying excitations weak enough to limit the response in the elastic range of the structure. implying that the modal combination errors contained in these procedures are acceptable. In this case.0. The errors are only weakly dependent on ground motion intensity (Fig. whereas MPA contains errors inherent in modal combination rules.5 Story Drift Ratio (%) 2 2.5 1 1. As mentioned in Section 3.15 summarizes the error in MPA considering three “modes” as a function of ground motion intensity. However.15) for ground multipliers larger than 1. as discussed in Sections 3.4 also apply to MPA.12 Heightwise variation of floor displacements and story drift ratios from MPA and NL-RHA for 1. 4. 4. 4.4. 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. implying excitations intense enough to cause significant yielding of the structure. an observation with practical implications. and the error envelope for each case. each story drift (Fig.(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 0.75.15). identified in Section 3. 4. UMRHA is essentially exact.2 and 3.5 × El Centro ground motion. the errors in MPA were larger for ground motion multipliers less than 0.15a).5 1 1. While various sources of errors in UMRHA. shading indicates errors in MPA including three “modes” Figure 4. now standard in engineering practice. the errors in MPA were fortuitously smaller than in UMRHA (compare Figs. Shown is the error in each floor displacement (Fig.5 Fig.4. 4.10 and 4. the MPA procedure is equivalent to the RSA procedure.15b).3. 45 . indicated by a ground motion multiplier.5 Displacement/Height (%) 2 Ground 0 0.

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

8 17.728 1.8 0.473 -15.694 1.652 1.498 1.266 -0.176 0.1 62.581 0.641 1.7 -12.298 0.705 -1.057 0.2 -100.18E-03 7.514 -1.911 0.7 1.1 62.36E-03 6.575 -53.5 × 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 0.8 -4.37E-03 1.00E+00 0.752 1.9 1.00E+00 0.336 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.8 -29.053 -1.611 0.0 15.02E-03 0.5 10.2 -100.9 5.76E-03 4.02E-03 3.7 1.305 -0.72E-03 7.156 -0.88E-03 0.5 10.2 11.980 -0.00E+00 0.2 -100.2 1.304 1.763 -14.049 -0.371 -0.414 1.8 -12.116 1.55E-03 3.667 -1.9 2.594 -1.37E-03 1.018 0.351 -0.233 1.19E-10 3.4 Peak values of story drift ratios (as % of story height) from MPA for 1.55E-03 3.0 -100.1 18.503 1.2 0.00E+00 0.012 1.844 -7.066 -0.50E-10 3.2 6.3 1.00E+00 0.088 12.9 7.683 1.5 × El Centro ground motion Hinge Plastic Rotation (rad) “Modal” Response Combined (MPA) “Mode” 1 7.1 46.781 0.3 11.435 0.00E+00 3.0 47 .99E-03 6.8 -6.311 0.76E-03 4.687 0.9 0.02E-03 0.8 1.007 1.4 1.Table 4.00E+00 0.804 1.0 -50.304 -1.781 0.018 -0.250 0.37E-03 1.101 -0.4 -8.5 × 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” 1.05E-03 2.00E+00 2 “Modes” 7.879 1.1 1.0 -100.116 1.705 1.55E-03 3.895 1.018 -0.2 -4.007 1.00E+00 1 “Mode” 7.003 -16.190 -0.72E-03 7.8 1.667 1.015 0.37E-03 1.36E-03 6.135 -7.527 -0.72E-03 7.00E+00 3 “Modes” 7.9 -100.5 2.399 0.8 “Mode” 3 -1.517 1.372 0.125 -1.8 -29.071 -0.0 -5.8 -29.315 0.033 -0.00E+00 0.053 1.910 1.5 2.5 Table 4.116 1.738 1.581 0.009 0.5 7.478 0.071 0.5 1.00E+00 0.0 -100.055 0.733 1.220 1.6 13.980 0.36E-03 6.338 1.1 -8.518 1.298 -0.6 -44.154 0.1 13.209 1.76E-03 4.60E-04 7.5 2.3 Peak values of floor displacements (as % of building height = 37.222 0.222 0.60E-03 2.8 -6.737 0.756 0.02E-03 3.982 13.6 -9.745 1.0 -50.233 1.260 -14.9 -4.1 62.00E+00 0.895 1.015 0.5 -6.26E-04 9.200 8.942 6.516 0.00E+00 0.00E+00 0.130 0.168 -0.3 -3.37E-04 Error (%) “Modal” Response 1 “Mode” -32.00E+00 0.8 7.640 1.426 15.503 -1.220 1.10E-02 9.3 -3.2 -3.3 13.00E+00 0.6 -44.666 Table 4.6 7.756 0.066 -0.00E+00 “Mode” 3 0.614 0.6 13.6 1.5 Peak values of hinge plastic rotations (radians) from MPA for 1.00E+00 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th “Mode” 2 0.76E-03 4.737 1.9 -100.429 0.00E+00 0.00E+00 0.118 0.900 -0.399 0.0 2 “Modes” -32.259 1.105 0.068 0.88E-03 1.9 1.244 0.407 -27.00E+00 0.72E-03 7.202 8.00E+00 0.197 -0.945 -49.0 3 “Modes” -32.22E-10 NL RHA 1.53E-03 7.640 -1.820 -7.331 1.668 -13.0 1.00E+00 0.36E-03 6.6 -7.14 m) from MPA for 1.6 -8.2 0.1 46.00E+00 0.1 46.88E-03 0.

” and the errors in these estimates relative to the exact response from nonlinear RHA.16 are unchanged. 4. and the structure is pushed using the force distribution of Eq.” however. the third “mode” contributions do not seem especially important (Fig. as are the roof displacements ur 2o = 11.7). and plastic hinge rotations at selected external beam end at each floor level (Table 4.18 and 48 . 4. 4. (3. the results presented so far did not include gravity load effects.0 cm.16a—which is slightly different than the one excluding gravity loads (Fig. resulting in the pushover curve shown in Fig.3 Modal Pushover Analysis with Gravity Loads To evaluate the accuracy of the dynamic response of the system.18 and Tables 4.7. Static analysis of the structure for gravity loads provides the initial state—forces and deformations—of the structure.20) with n = 1 (Fig. (3.8).6 through 4. Significant improvement is achieved by including response contributions due to the second “mode. 4.11a). 3. Also presented at these roof displacements are the displacements at all floors (Table 4.5 × El Centro ground motion are presented below.5).” these effects will modify the combined modal response as well as the results of nonlinear RHA. MPA including two “modes” underestimates the displacements of lower floors by up to 6% and overestimates the upper floor displacements by up to 22%.17 and 4.17 and Tables 4. 4.7 cm and ur 3o = 2.4) to roof displacement ur1o = 52. and overestimated by up to 29% in the middle stories and by up to 13% in the upper stories.17 and Tables 4.4. 4.4.6 and 4. The errors are especially large in the plastic hinge rotations estimated by the MPA procedure even if three “modes” are included (Fig.3 . the structure is pushed using the force distribution of Eq. The first “mode” alone provides adequate estimates of floor displacements.53 cm. (3. The pushover curves for the second and third modes included in Fig. Results of such analyses for the selected building subject to 1.” but not the contributions of higher “modes. Starting with its initial state under gravity loads. and three “modes. The story drifts are underestimated by up to 7% in the lower stories. Figures 4. drifts in all stories (Table 4.20) with n = 1 to a target roof displacement.18 and Tables 4. Obviously gravity load effects will influence the seismic demands due to the first “mode.7).18) considering one.6 and 4. They are now included in the pushover analysis of the structure for the first “mode” only.6 and 4.6).7). but it is inadequate in estimating the story drifts (Fig. two. 4.8 present estimates of the combined response according to Eq.4. As shown in Fig. The response contributions of the second and third modes are the same as before (Tables 4.

Noted Error (%) 30 20 10 0 0 2 8 3 4 1 9 Error (%) 30 20 5 4 6 5 7 2 9 10 0 0 1 8 7 3 6 0. Noted 60 50 40 (b) Story Drifts Error Envelope Error for Story No. Locations of plastic hinges determined by MPA considering one.8).5 2 GM Multiplier 2. Errors in MPA as a function of ground motion intensity: (a) floor displacements. and three “modes” and by NL-RHA for 1.5 1 1.14. Most pushover analysis procedures do not seem to compute to acceptable accuracy plastic hinge rotations.Table 4.5 2 GM Multiplier 2. 1−"Mode" • • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • •• • •• • • • • • •• •• • •• •• •• •• (b) MPA. 2−"Modes" •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• • •• (c) 3−"Modes" • • • • • • • •• •• • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• • •• • • • • • • • • • (d) Nonlinear RHA •• •• • •• •• •• •• • •• •• •• •• •• •• •• • •• •• • •• •• •• •• • • • • • • •• •• •• Fig.15.5 1 1.5 × El Centro ground motion (a) Floor Displacements 60 50 40 Error Envelope Error for Floor No. (a) MPA. 4. and (b) story drifts 49 .5 3 Fig.5 3 0. two. 4.

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

3 cm. α = 0.5 0. 0. V = 4952 kN. V = 7433 kN.5 2 3 Actual Idealized 2000 0 0 0.5 0. 1.9 cm.85 0. “Modal” pushover curves with gravity loads included.0.19 y by 3 2 1. and 3.75 0. 0. 4.13 y by Base Shear (kN) 8000 6000 1 1.14 y by Base Shear (kN) 8000 6000 4000 1. α = 0.5 1 0.85 0.25. noted are peak values of roof displacement for 0.5 2 3 4000 2000 0 0 Actual Idealized 0. 0.85 Base Shear (kN) 8000 6000 4000 2000 0 0 0.75 Actual Idealized 20 40 60 Roof Displacement (cm) 80 (b) "Mode" 2 Pushover Curve 12000 10000 u = 9.50.0.0 × El Centro ground motion 51 .16.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36. V = 5210 kN.5 0. α = 0. 2.85.25 1 2 4 6 Roof Displacement (cm) 8 10 Fig.6 cm.25 0.75 0.25 5 10 15 20 Roof Displacement (cm) 25 (c) "Mode" 3 Pushover Curve 12000 10000 u = 4.75.

Heightwise variation of floor displacements and story drift ratios from MPA and NL-RHA for 1. story drifts.(a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th NL−RHA MPA 1 "Mode" 2 "Modes" 3 "Modes" (b) Story Drifts Floor Floor 5th 4th 3rd 2nd 1st 5th 4th 3rd 2nd 1st NL−RHA MPA 1 "Mode" 2 "Modes" 3 "Modes" Ground 0 0.5 Fig.18.5 × El Centro ground motion. 4.5 1 1. Errors in floor displacements.5 1 1. 4.5 Story Drift Ratio (%) 2 2. two. shading indicates errors in MPA including three “modes” (a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th (b) Story Drift Ratios Floor 4th 3rd 2nd 1st MPA 1 "Mode" 2 "Modes" 3 "Modes" Floor 5th 5th 4th 3rd 2nd 1st MPA 1 "Mode" 2 "Modes" 3 "Modes" Ground −60 −40 −20 0 Error (%) 9th 8th 7th 6th 20 40 60 Ground −60 −40 −20 0 Error (%) 20 40 60 (c) Hinge Plastic Rotations Floor 5th 4th 3rd 2nd 1st MPA 1 "Mode" 2 "Modes" 3 "Modes" Ground −120 −80 −40 0 Error (%) 40 80 120 Fig. and hinge plastic rotations estimated by MPA including one.17.5 Displacement/Height (%) 2 Ground 0 0. gravity loads included. and 52 .

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. Noted 60 50 40 (a) Story Drifts Error Envelope Error for Story No. two.5 × El Centro ground motion. Locations of plastic hinges determined by MPA considering one. gravity loads included 53 .5 1 1. Errors in MPA as a function of ground motion intensity: (a) floor displacements. 2−"Modes" •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• •• •• (c) MPA.5 1 1. 4. 3−"Modes" (d) Nonlinear RHA • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• •• •• • • • • • • • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • • • • • •• •• •• Fig. and (b) story drifts.5 2 GM Multiplier 2.5 3 Fig.(a) MPA. and three “modes” and by NL-RHA for 1.19.20. 4. gravity loads included (a) Floor Displacements 60 50 40 Error Envelope Error for Floor No. 1−"Mode" • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • •• •• •• • • • • • •• •• •• •• •• •• •• (b) MPA.5 2 GM Multiplier 2.5 3 0.

530 1.00E+00 0.00E+00 0.5 × El Centro ground motion.315 0.2 4.8 Peak values of hinge plastic rotations (radians) from MPA for 1.109 0.850 -1.490 -11.0 37.00E-03 5.3 13.35E-03 8.429 -1.319 1.00E-03 5.8 9.2 -2.686 -7.461 0.00E-10 NL RHA 1.2 1.7 19.197 -0.237 0.00E+00 2 “Mode” 8.2 9.516 0.114 1.983 1.154 0.6 0.00E+00 0.4 1.5 “Mode” 3 -1.207 18.00E+00 0.2 16.4 -6.7 4.068 0.3 9.00E+00 0.311 0.933 1.5 × El Centro ground motion.257 0.996 0.754 1.854 0.066 -0.35E-03 8.1 13.998 21.372 0.2 12.0 -100.0 -30. gravity loads included Hinge Plastic Rotation (rad) Floor “Modal” Response “Mode” 1 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 8.009 0.071 -0.8 1.00E+00 0.049 -0.603 1.515 -50.5 21.3 1.436 1.114 -1.7 -2.23E-03 3.877 -46.330 1.2 4.2 9.00E+00 0.434 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.04E-02 8.00E+00 0.429 1.8 Table 4.2 21.176 0.2 9.26E-03 3.813 1.11E-03 9.888 1.434 0.3 -100.00E+00 0.1 1.00E+00 0.466 0.3 -100.0 37.055 0.5 21.88E-03 1.5 10.00E+00 Error (%) “Modal” Response 1 “Mode” -32.19E-03 1.7 Peak values of story drift ratios (as % of story height) from MPA for 1.2 -0.125 -1.5 3 “Modes” -32.0 16.687 0.754 0.88E-03 0.351 -0.9 31.5 1.6 1.35E-03 8.19E-03 1.00E+00 0.858 2.527 -0.637 0.105 0.Table 4.478 0.55E-03 3.783 1.00E+00 0.13E-03 5.015 0.7 -2.4 26.637 0.23E-03 0.454 1.9 -4.6 19.8 28.831 0.530 1.822 1.399 -0.237 0.102 1.00E+00 0.00E-03 5.908 -1.213 1.860 1.998 1.037 -0.213 1.00E-03 5.7 1.214 0. gravity loads included Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th “Modal” Response “Mode” 1 “Mode” 2 Drift Ratio (%) Combined (MPA) 1 “Mode” 2 “Modes” 3 “Modes” Error (%) NL 1 2 3 RHA “Mode” “Modes” “Modes” 1.6 2.998 0.00E+00 Combined (MPA) 1 “Mode” 8.190 -0.17E-03 9.23E-02 1.305 -0.2 4.098 20.00E+00 0.11E-03 9. 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” 0.913 7.8 -32.14 m) from MPA for 1.130 0.2 0.353 -23.2 -3.854 0.00E+00 “Mode” 3 0.4 -4.938 1.4 -4.5 1.850 1.1 4.927 1.3 -22.5 -5.996 -0.4 1.399 0.00E+00 3 “Modes” 8.8 -32.213 1.319 1.5 × El Centro ground motion.199 16.270 -12.673 Table 4.19E-04 5.071 0.836 -0.2 1.78E-03 1.9 -6.263 1.0 37.953 15.5 0.066 -0.35E-10 3.88E-03 0.00E+00 0.35E-03 8.04E-10 3.371 -0.0 -100.434 0.8 0.603 -1.128 -1.55E-03 3.7 16.19E-03 1.00E+00 0.2 2.3 -22.037 0.23E-03 3.821 -1.266 -0.6 Peak values of floor displacements (as % of building height = 37.3 -22.5 54 .156 -0.594 -1.921 1.744 1.830 -12.9 -3.310 1.101 -0.667 0.018 -0.728 1.19E-03 1.821 1.23E-03 0.057 0.00E+00 0.908 1.665 0.11E-03 9.11E-03 9.033 -0.6 2.0 2 “Modes” -32.3 -100.102 1.0 -30.55E-03 3.4 20.1 21.983 1.514 -1.507 1.330 1.00E+00 0.00E+00 “Mode” 2 0.00E+00 3.107 1.414 28.530 1.064 -10.263 0.75E-03 0.168 -0.831 0.8 0.00E+00 0.

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. 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. gravity load effects were included in all 55 .. story drifts.5 sec . 1997] The pushover curve.” and nonlinear RHA. “Uniform” distribution: s* = m j (where the floor number j = 1. k = 2 for T1 ≥ 2. computed at the target displacement represent the earthquakeinduced demands on the structure.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. SRSS distribution: s* is defined by the lateral forces back-calculated from the story shears determined by response spectrum analysis of the structure. 5. etc. joint rotations. 2… N ). Comparison of Modal and FEMA Pushover Analyses 5. j 2. Specified in FEMA-273 are three distributions for lateral forces: 1. and varies linearly in between. MPA considering three “modes. plastic hinge rotations.5. assumed to be linearly elastic. Equivalent lateral force (ELF) distribution: s* = m j h k where h j is the height of the j j jth floor above the base. and 3.5 sec . and the exponent k = 1 for fundament period T1 £ 0. a plot of base shear versus roof displacement. The floor displacements.

1. wherein the first two are obvious and the third is determined from response spectrum analysis of the building (Appendix C).1.00719 0. both presented in Section 4.281 0. and the displacements of upper floors are overestimated by 9-22%. 5.0 cm.3b and Table 5.4.11 0.0913 0.11 0.3.21 0.4 and Tables 5.3a and 5.126 0.2.2. Also included in these presentations are the MPA results considering three “modes” and the “exact” demands from nonlinear RHA.0197 0.5 times the El Centro ground motion.0981 0.0896 0.0654 0.11 0.0702 0.165 0.4.0381 0.0446 0. Figures 5.11 0.3. The “uniform” distribution overestimates all floor displacements by 17-28%.11 0. plastic hinge rotation demands in Table 5. Using each of these force distributions.11 0.1. 5. Force distributions in FEMA-273: (a) “uniform”.112 0.3a and Table 5. The first four floor displacements are within 5% of the “exact” value.367 (a) Uniform (b) ELF (c) SRSS Fig. with the errors being larger for the SRSS distribution. the story drift demands in Fig.042 0. 5. 5.1 demonstrate that the displacement demands are underestimated by the ELF and SRSS force distributions by 12 to 30 % at the lower six floors of the building.0466 0. 5.119 0. 0.1 through 5. 5. pushover analyses are implemented for a target roof displacement of 52.062 0. and Table 5. and the locations of all plastic hinges in Fig. The three FEMA force distributions are presented in Fig. The pushover curves are given in Fig. The errors in the FEMA and MPA estimates of seismic demands relative to the “exact” demands are presented in Fig. and (c) SRSS 56 .3. 5. the value determined from RHA of the first-mode inelastic SDF system for 1. The MPA procedure is more accurate than all the FEMA force distributions.4a. (b) ELF. the floor displacement demands in Fig.analyses.177 0.11 0.

V = 6897 kN. α = 0. and (c) SRSS. α = 0.2. V = 8530 kN. V = 7456 kN. gravity loads are included 57 .3 cm.(a) Uniform Distribution 12000 10000 u = 33 cm. (b) ELF.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 = 38.18 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized 10 20 30 40 Roof Displacement (cm) (c) SRSS Distribution 50 60 12000 10000 u = 39. Pushover curves using three force distributions in FEMA-273: (a) “uniform”.24 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized 10 20 30 40 Roof Displacement (cm) 50 60 Fig. α = 0.8 cm. 5.

reaching 35%.4c and Table 5. errors are largest in the upper stories.5 Ground 0 Ground 0 Fig.2 demonstrate that the story drift demands are greatly underestimated by all the FEMA force distributions. Heightwise variation of floor displacements and story drift ratios estimated using FEMA-273 force distributions. with errors reaching 37% in this example. Modal pushover analysis procedure gives estimates better than all the FEMA force distributions. For the uniform distribution. 5. For the ELF distribution.4b.3b and 5.) The pushover analysis procedures considered seem incapable of computing accurately local response quantities. reaching 31%.5 Story Drift Ratio (%) 2 2. For the SRSS distribution. reaching 64%.3. such as hinge plastic rotations. the MPA procedure is more accurate than all the FEMA force distributions. (a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th (b) Story Drift Ratios Floor 4th 3rd 2nd 1st NL−RHA MPA.Figures 5. gravity loads included 58 .3 demonstrate that the hinge plastic rotations estimated by all three FEMA force distributions contain unacceptably large errors. In contrast. MPA including three “modes. whereas nonlinear RHA computed an insignificantly small value.” and NLRHA. the errors are largest in the lower stories.5 Displacement/Height (%) 2 Floor 5th 5th 4th 3rd 2nd 1st NL−RHA MPA. This seems to be an inherent limitation of pushover analysis. 3 "Modes" FEMA Uniform ELF SRSS 0. with story drifts under estimated by.5 1 1. Figure 5. 7%. larger errors are noted in the upper and lower stories. at most. and Table 5. (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%. but it is still inaccurate. 3 "Modes" FEMA Uniform ELF SRSS 0.5 1 1.

Errors in floor displacements. While pushover estimates for floor displacements are more accurate. they are not good indicators of damage. and the three FEMA analyses. and hinge plastic rotations estimated using FEMA-273 force distributions and MPA (including three “modes”). and the ELF distribution fails 59 .” nonlinear RHA (“exact”). 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. 5. story drifts. the SRSS distribution fails to identify yielding of beams in the middle floors. 5. The locations of plastic hinges shown in Fig. 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 the plastic hinges are not identified correctly by the FEMA force distributions.5 were determined by five analyses: MPA considering the three “modes.(a) Floor Displacements 9th 8th 7th 6th MPA FEMA Uniform ELF SRSS 9th 8th 7th 6th (b) Story Drift Ratios MPA FEMA Uniform ELF SRSS Floor 4th 3rd 2nd 1st Floor −40 −20 0 Error(%) 9th 8th 7th 6th 20 40 60 5th 5th 4th 3rd 2nd 1st Ground −60 Ground −80 −60 −40 −20 0 20 Error (%) 40 60 80 (c) Hinge Plastic Rotations MPA FEMA Uniform ELF SRSS Floor 5th 4th 3rd 2nd 1st Ground −200 −150 −100 −50 0 50 Error (%) 100 150 200 Fig. Based on the results presented here.4. the “uniform” distribution fails to identify yielding of the beams above the fourth floor.

The MPA procedure is superior to FEMA-273 analyses over the entire range of ground motion intensities considered. 60 . Included for comparison is the error in MPA with three “modes.6 and 5.to identify yielding in some locations. but fails to predict yielding in a few locations. especially in estimating story drifts. Figures 5.7 summarize the error in FEMA analyses relative to the “exact” demands as a function of ground motion intensity indicated by a ground motion multiplier. The MPA procedure identifies yielding in most locations predicted by “exact” analysis.” The MPA procedure provides estimates of earthquake demands that are significantly more accurate than all FEMA273 analyses. Shown is the error in each floor displacement and each story drift. and the error envelope for each case.

597 0.207 1.14 m) from FEMA force distributions and MPA.414 1. Peak values of floor displacements (as % of building height = 37.9 -70.8 -63.927 1.3 -11.672 1.8 -35.089 1.2 5.2 -100.6 -17.1 -3.5 -29.6 -25.3 -22.8 -32.294 1.1 163.5 -3.94E-03 2.00E+00 0.00E+00 8.2 16.1 10.0 -100.367 1.270 0.7 -29.2 6.35E-10 3.19E-03 1.7 19.466 0.067 0.50E-03 0.04E-02 8.098 1.724 0.436 Error (%) NL Uniform RHA 0.351 0.00E+00 0.214 1.6 -11.75E-03 0.3 -100.8 44.783 1.875 0.00E+00 0.0 37.52E-04 1.2 13.417 1.6 -41.3 -14.673 Error (%) NL RHA 1.2 9.Table 5.858 1.8 2.7 FEMA ELF SRSS -22.8 -2.2 4.7 21.4 21.10E-02 7.5 -33.16E-03 0.998 1.9 11.372 1.9 16. gravity loads included Displacement /Height (%) FEMA ELF SRSS MPA 0.2 Peak values of story drift ratios (as % of story height) from FEMA force distributions and MPA.9 0.23E-02 1.5 -32.109 1.809 0.0 -71.6 -6.562 1.5 26.6 17.330 1.462 1.93E-03 1.992 1.318 2.23E-03 3.4 16.13E-03 5.03E-03 5.4 -50.789 0.1 -26.9 16.6 22.0 -30.5 -33.4 -23.515 FEMA Uniform ELF 27.3 -4.6 -4.00E+00 2.8 17.355 0.00E+00 4.877 1.11E-03 9.8 7.938 1.611 0.5 10.3 23.4 -12.998 1.860 1.530 1.736 0.6 4.53E-02 1.00E+00 0.708 0.263 0.6 21.487 0.830 2.55E-03 3.17E-03 9.7 MPA -2.78E-03 1.00E+00 Error (%) FEMA Uniform ELF SRSS MPA 24.00E-03 5.58E-04 6.064 1.0 Table 5.839 0.667 0.888 0.4 0.209 0.335 2.128 1.9 SRSS -22.35E-03 8.34E-03 2.9 -77.0 -63.00E+00 NL RHA 1.09E-03 4.015 0.8 -100.2 9.399 -27.007 1.566 1.730 1.88E-03 1.5 -27.344 0.221 1.5 Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 2.8 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 0.4 -55.178 1.490 0.3 31.310 1.524 0.7 26.2 -12.306 1.984 1.5 -29.7 7.547 -27.7 -28.399 1.9 15.1.975 1.623 1.26E-03 3.399 1.913 0.7 -60.353 1.854 0.953 0.00E+00 0.59E-03 5.341 1.061 1.4 16.154 1.0 -57.7 Table 5.6 -73.4 MPA -2.011 1.083 1.0 -59.195 0.78E-03 0.323 1.262 1.7 -15.51E-03 4.0 -100.3 Peak values of hinge plastic rotations (radians) from FEMA force distributions and MPA Hinge Plastic Rotation (rad) MPA FEMA ELF SRSS 4.168 1.9 28. gravity loads included Displacement /Height (%) FEMA ELF SRSS MPA 1.836 0.530 1.2 16.234 1.199 27.560 1.5 22.45E-03 3.0 -100.2 -4.65E-03 7.8 16.2 10.1 0.2 19.3 29.6 14.84 1.00E+00 0.314 1.8 -100.62E-03 0.5 15.5 61 .686 0.19E-04 5.00E-10 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 1.0 -100.9 -20.

5. gravity loads included 62 . Locations of plastic hinges determined from three force distributions in FEMA-273.5 × El Centro ground motion.(a) FEMA: Uniform • • • • • • • • •• •• •• •• •• •• (b) FEMA: ELF • •• •• •• •• •• •• • •• • •• •• •• •• •• • • • • • •• • • • • • • • •• •• •• •• • •• •• •• •• • •• •• •• •• • • •• •• •• • • • • • •• •• (c) FEMA: SRSS • • •• •• •• •• • • •• •• •• • • •• •• •• • • • • •• •• •• • • • • • • • • • (d) MPA with 3 "Modes" • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• •• •• (e) Nonlinear RHA • • • • • • • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • • • • • •• •• •• Fig. 5. MPA including three “modes” and NL-RHA for 1.

(a) FEMA: Uniform 60 50 40 60 50 40 3 (b) FEMA: ELF Error Envelope Error for Floor No. 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. Errors in floor displacements from three force distributions in FEMA-273 and from MPA including three “modes”.5 3 0 0 0.5 1 1.5 3 0.5 1 1. gravity loads included 63 . 5.5 2 GM Multiplier 2.5 2 GM Multiplier 2.5 3 Fig.6.5 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 0.5 2 GM Multiplier 2.5 2 GM Multiplier 2.5 1 1.

Noted 80 40 9 1 Error (%) 7 Error (%) 60 30 20 10 2 5 6 4 8 40 1 6 20 2 5 3 4 3 7 0 0 0.5 3 (c) FEMA: SRSS 60 50 40 60 50 40 3 2 1 (d) MPA (3 "Modes") Error (%) 30 20 Error (%) 30 5 4 8 9 20 9 6 7 7 10 0 0 4 10 2 3 1 8 5 6 0.5 2 GM Multiplier 2.5 2 GM Multiplier 2.7.(a) FEMA: Uniform 100 60 50 9 8 (b) FEMA: ELF Error Envelope Error for Story No.5 2 GM Multiplier 2. 5.5 1 1.5 1 1.5 3 0 0 0. Error in story drifts from three force distributions in FEMA-273 and from MPA including three “modes”.5 3 Fig.5 1 1.5 2 GM Multiplier 2. gravity loads included 64 .5 3 0 0 0.5 1 1.

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

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

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

such as life safety and collapse prevention. they are not good indicators of damage. structural performance evaluation should be based on story drifts that are known to be closely related to damage and can be estimated to a higher degree of accuracy by pushover analyses. should obviously be evaluated for a wide range of buildings and ground motion ensembles. 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. with the excitation scaled to cover a wide range of ground motion intensities and building response.11. 68 . Instead. This new method for estimating seismic demands at low performance levels. 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. While pushover estimates for floor displacements are even more accurate.

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

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

2. Calculate the area under the actual pushover curve. The structure should be pushed just beyond the target (or expected) roof displacement in the selected mode.1 incrementally and record the base shears and associated roof displacements. B. ω n . 3. Define the anchor point. Apn . 1988). This step can be conveniently implemented in any commercially available software. 3. trapezoidal rule.urn ) pushover curve for the force distribution s* : n 2.1) using the FEMA-273 procedure (Building Seismic Safety Council. A. for linear-elastic vibration of the building. e. respectively. Idealize the pushover curve as a bilinear curve (Fig. φn . iterations may be necessary. develop the base-shear – roof-displacement (Vbn . using any numerical integration method.g. (3.1. For the nth-“mode”.. Apply force distribution of Step 2. DRAIN2DX (Allahabadi and Powell. Compute natural frequencies.2. Define the force distribution s* from Eq. 71 .20): s* = mφn n n 2. 1.1 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.g. of the bilinear curve at the target roof displacement.. 2. e. 3. Let the roof displacement and base shear at the anchor point be urno and Vbno .1. 1977). and modes. Since the target roof displacement may not be known at the start of the procedure.Appendix A Uncoupled Modal Response History Analysis A.

i +1 i i 3. Calculate the post-yielding strain-hardening ratio. and B with straight-line segments to obtain the idealized bilinear curve.6 × Vbny . at base shear i equal to 0. i i i αn = Vbno Vbny − 1 urno urny − 1 i 3.8. Draw the curve OAB by connecting the three points O. If the error exceeds some pre-specified tolerance. determine the roof displacement.4 to 3.6 .3. This value. Calculate initial slope of the idealized bilinear curve.6 × Vbny ( ) i urn . kn = 0.0. and a point on the actual pushover curve with base shear equal to i 0. From the pushover data. Calculate the error = 100 × Abn − Apn ( ) ( ) ( ) ) Apn .0. O.1. Calculate area under the bilinear curve OAB. This step gives the secant stiffness at a base shear equal to 60% of the yield base shear. Vbny .6.5. Calculate the yield displacement. i i i 3.6 . Vbny . If desired. Estimate the yield base shear. 3.4. will be refined by an iterative procedure that seeks to equate areas under the actual and the idealized pushover curves. Let the point with base shear = Vbny and roof displacement = urny be denoted as A. 72 .i 3. other appropriate methods can be ( used. i i 3. Abn . Calculate Vbny = Vbny × Apn Abn . i 3.4.4.2. corresponding to the estimated yield i i i base shear. urny = Vbny kn . 3.9. Calculate the slope.1. Replace i+1 with i and repeat Steps 3.7. 3.9. obtained by judgment.9.6 × Vbny .8. i 3. A. urn. iterations are necessary.2. kn . i 3. by connecting a straight line between origin.

(3. Calculate peak values. 8. Scale the vertical axis with * Mn to obtain * Fsno Ln = Vbno M n and * Fsny Ln = Vbny M n (Eqs.3b) with unit-mass and force-deformation relation of Fig. Dn (t ) .1. Combine the “modal” responses using Eqs. An (t ) . Compute the Ln and Γ n from Eq.11a). 4.1 Idealization of nth-“mode” pushover curve 73 .2. 4.6 ur n y ur n o ur n Fig.16).eps ur n.15) and (3.12) and (3. Scale the horizontal axis by Γnφrn to obtain Dno = urno Γ nφrn and Dny = urny Γ n φrn (Eqs. In general first two or three modes will suffice. 5. A. Repeat Steps 3 to 6 for as many modes as required for sufficient accuracy.10b and 4.13). of the combined responses obtained in Step 8. (3.10a and 4. and pseudo-acceleration history. (3. 4. * 4.4. A.6 × Vbny A • Idealized αnkn B • 1 Actual k 1 • O n figA_1. of the nth“mode” inelastic SDF system (Fig. r o . 9. Compute deformation history. 6. 4. 7. 4.3.4) and effective modal mass from M n = Ln Γn . A. Calculate histories of various responses using Eqs.11b).2). Develop the Fsn Ln − Dn relation (Fig.2. Vbn Idealization of Pushover Curve Vbno Vbny 0.0.

A. respectively.2. 3. computed for the first three modes are shown in Fig. At this point.1. 3.).).). generated using DRAIN-2DX.eps Dny = ur n y / Γnφr n Dno = ur n o / Γnφr n Dn Fig. The following steps illustrate the procedure to develop the idealize curve for the first “mode”. The base-shear – roof-displacement (Vbn . Following is step-by-step implementation of the procedure described in Section A. to the north-south component of the El Centro (1940) ground motion scaled up by a factor of 1. for the first.3.3. Idealized bilinear curves for each of the three modes are included in Fig.5 cm and Vb1o = 8729.2 Properties of the nth-“mode” inelastic SDF system A.3. Area under the actual pushover curve.urn ) pushover curve for the force distribution s* : n 2.6 kN.2 EXAMPLE The UMRHA procedure is implemented to calculate the response of the 9-story building described in Section 3.1.1. 74 .4 cm (10 in.4.5.Fsn / Ln Fsn / Ln − Dn Relationship Vbno / M* n Vbny / Mn * 1 α ω2 n n ω2 1 n figA_2. The pushover curves for the first three modes. are shown in Fig. 3. and 12. 2.5 cm (25 in. second. A. is defined at the target roof displacement. B. 25. and third mode. First three mode shapes and frequencies of the selected building were computed and are shown in Fig.4. n 2. A. 1. The anchor point. The force distributions. ur1o = 63.7 cm (5 in. s* .2. 3. Ap1 = 360777 kN-cm. 3. Target displacements used to generate these pushover curves are 63.

Error = 100 × (365100 − 360777 ) 360777 = 1.4 × (360777 365100 ) = 7911. and M1 = 2736789 × 1. iterations are necessary.135. i i 3. The results of the iterative procedure are summarized in Table A. i 3.6.6 × Vb1 y ( ) i ur1. i i i 3.4 kN.3666 = 3740189 kg. k1 .86 = 210.01%. i i i 3.6 × Vb1 y = 4803.9.5 38. The yield displacement.5. The results for other modes are summarized in Table A. The Fs1 L1 − D1 relation for the first “mode” is developed as follows. ur1 y = Vb1 y k1 = 8006.4 ) − 1 (63.9 kN. Also included are the modal damping ratios and the periods calculated from Eq. The initial slope of the idealized bilinear curve.3666.8 22. 3.86 cm at 0.6 = 4803.1.18 = 38.23 cm.9.8 kN.13).4 210.0. This value exceeds the prespecified tolerance of 0.6 kN.0. 75 .4. The first estimate of the yield base shear Vb1 y = 8006.09 ) − 1 = 0. Determined from the pushover database.2. 3.4. L1 = 2736789 kg. A.18 kN/cm.6 = 22. ur1.9. and B with straight-line segments gives the idealized bilinear curve.8.4.6 8006. i 3.3. The next estimate of the yield shear is Vbny = 8006.2. Therefore.194.09 cm and Vb1 y = 8006. i i 3.4 kN. i +1 3. (4. and α1 = 0.1. Ab1 = 365100 kN-cm. The post-yielding strain-hardening ratio. Γ1 = 1.1.1.09 cm.2. α1 = Vb1o Vb1 y − 1 ur1o ur1 y − 1 = ( ) ( ) (8729. 3. The point A on the i i bilinear curve is defined by ur1 y = 38.i 3.7. The curve OAB obtained by connecting the three points O. Area under the bilinear curve OAB.198%. is calculated as follows. k1 = 0. * 4. 4. Vb1 y = 7615. The procedure converged after nineteen cycles to give ur1 y = 36.

4. 9. Scaling the vertical axis by Fs1 y L1 = 203. The peak values are also plotted in Fig. The peak values are computed and are summarized in Tables 4.1 and 4.51 cm.2. 8.4. 4.7. 4. Scaling the horizontal axis by Γ1φr1 gives D1o = 46. 6.4.7. 4.62 (cm/sec2). * M1 gives Fs1o L1 = 233. Histories of roof displacement and top story drifts for the first “mode” are computed and presented in Fig. 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.3. 7. 5.7.8. The results were generated for first three “modes’ and are included in Fig. 4.2.46 cm and D1y = 26. A. The combined modal responses are presented in Fig.40 (cm/sec2) and 76 .

α = 0.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36. α = 0. α = 0.eps 0 0 2 4 6 8 10 Roof Displacement (cm) 12 Fig. A.eps 0 0 5 10 15 20 Roof Displacement (cm) (c) "Mode" 3 Pushover Curve 25 8000 u = 4.6 cm.2 cm. V = 7616 kN.13 6000 Base Shear (kN) 4000 2000 Actual Idealized figA_3b.19 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized figA_3a.eps 10 20 30 40 50 Roof Displacement (cm) (b) "Mode" 2 Pushover Curve 60 8000 u = 9. “Modal” pushover curves for the example building 77 .9 cm. V y by = 4952 kN.3. V = 5210 kN.14 y by 6000 Base Shear (kN) 4000 2000 Actual Idealized figA_3c.

5 × El Centro ground motion for the first “mode.2769 D1 (cm) 0 • 35. Histories of deformation and pseudo-acceleration due to 1.06 0 • 1.52 5 10 15 20 Time (sec) 25 30 0 • 1.4.33 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. 78 .747 5 10 15 20 Time (sec) figa_4.m −80 0 −3 0 25 30 Fig.” second “mode.223 −80 80 n=3 −3 3 n=3 D3 (cm) 0 A3 (g) • 10.” and third “mode” inelastic SDF systems.

18 210.18 210.56 47.56 19.1 7616.107 0.029 0.5309 488839.39 36.18 210.082 0.18 210.2.4 7714.79 21.81 21.948 ζ n (%) 79 .8 4747.192 0.40 46.18 210.2671 1.188 0.1 4574. No.6 × Vb1 y i ur1.4 7647.46 2.5 7624.4 7672.194 0.1 4569.194 (kN-cm) 365100 364060 363276 362684 362235 361892 361631 361432 361279 361162 361073 361004 360951 360911 360880 360856 360838 360824 360813 Table A.193 0.Table A.32 36.182 0.5 (cm) 22.02 21.51 2 “Mode” 2 -920860 -0.62 26.198 0.4927 1.2 4671.135 0.6 7840.9 4570.11 22.85 36.063 0.0 4588.18 210.2 4628.05 52.037 0.7 7639.162 0.193 0.191 0.4 7911.3 4603.194 0.59 22. Properties of “modal” inelastic SDF systems Properties Ln (kg) Γn * M n (kg) “Mode” 1 2736789 1.86 21.74 21.8525 1.64 37.5 3109.180 0.77 21.237 0.103 “Mode” 3 696400 0.18 210.010 (kN) 8006.12 3876.3 7745.75 21.910 0.3 7628.8 7618.136 Fsny Ln (cm/sec2) Dny (cm) Fsno Ln (cm/sec ) Dno (cm) Tn (sec) 233.30 37.50 36.18 210.18 (cm) 38.0.693 0. i 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 i Vb1 y i 0.90 21.6 4583.194 0.18 210.65 1226.017 0.18 210.6 i k1 i ur1 y i α1 i Ab1 Error (%) 1.0 4577.151 0.35 36. Results of iterative procedure to develop the idealized bilinear curve for the first “mode” inelastic SDF system Itr.18 210.18 210.28 36.8 4647.78 21.139 0.18 210.25 36.59 36.2406 167531.18 210.9 7615.24 36.23 0.83 21.9 (kN) 4803.29 36.18 210.86 22.70 36.3 7786.4 4595.9 4573.190 0.184 0.1.74 (kN/cm) 210.170 0.7 4580.05 36.25 36.0 4704.75 21.8 7622.193 0.26 36.529 0.38 22.18 210.18 210.09 37.18 210.5 7633.022 0.186 0.2 4571.309 0.09 18.44 36.85 0.5 4614.3666 3740189 203.176 0.0 7619.404 0.76 21.048 0.2 7690.79 0.23 22.1 1013.9 4570.013 0.95 21.3 7658.

80 .

81 . From the total hinge rotation.2 EXAMPLE The MPA procedure is implemented to calculate the response of the 9-story building described in Section 3. (4.21). 6. (4.6b by solving Eq. the first two or three “modes” will suffice. Calculate the peak roof displacement urno associated with the nth-“mode” inelastic SDF system from Eq. (3.18). Following is step-by-step implementation of the procedure described in Section B.33 cm.8).Appendix B Modal Pushover Analysis B. Repeat Steps 3 to 7 for as many “modes” as required for sufficient accuracy.1. Compute the peak deformation. This is the same example as solved in Appendix A. 11. Determine the total response by combining the peak “modal” responses using the SRSS combination rule of Eq.5. of the nth-“mode” inelastic SDF system (Fig. Typically. 10. Solving Eq. subtract the yield hinge rotation to determine the plastic hinge rotation.3b) with unit mass and force-deformation relation of Fig. Steps 1 to 4 of the MPA are the same as those for UMRHA presented in Appendix A. B. At urno . 10.4. 4. extract from the pushover database values of other desired responses. rno . or from the inelastic response (or design) spectrum. (3.1 to the north-south component of the El Centro (1940) ground motion scaled up by a factor of 1. 12. 13. Dn .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 = 35.1 STEP-BY-STEP PROCEDURE A detailed step-by-step implementation of the modal pushover analysis (MPA) procedure is presented in this section and illustrated by an example in the next section. 4.

values of floor displacements and story drifts are extracted. 82 . 9.3 and 4.” and the results are included in Tables 4. The total response computed by combining the peak “modal” responses using the SRSS combination rule of Eq. Also included in Table 4. 8.366 × 1 × 35.11. where results for other ground motion intensities are also included. (3.28 cm.18) are also included in Tables 4. The values are summarized in Table 4.3 and 4. At ur1o = 48. Steps 3 to 7 are repeated for first three “modes. Peak roof displacement ur1o = Γ1 × φr1 × D1 = 1.33 = 48. Results of Steps 5 and 6 are summarized in Table B.4 for the story drifts normalized by the story height.1.28 cm.5 are the plastic hinge rotations computed by subtracting the yield hinge rotation from the total hinge rotations.3 for the floor displacements normalized by the building height and in Table 4. 12.4.4.

Calculation of roof displacements urno from peak deformation of inelastic SDF systems Ground Motion Multiplier 0.52 “Mode” 3 1.457 12.267 5.007 36.467 14.395 0.225 2.03 26.901 8.70 0.660 14.52 0.200 0.755 27.436 7.748 63.Table B.27 0.513 0.913 22.73 24.690 10.312 1.184 0.185 11.25 Quantity Dn (cm) “Mode” 1 6.07 “Mode” 2 4.504 18.535 14.678 0.18 27.37 57.766 7.577 16.82 1.755 0.38 22.154 78.35 0.06 1.367 1.0 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 1.117 5.0 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 3.25 20.229 8.4222 3.0 µ = Dn Dny urno = Γnφrn Dn (cm) 83 .13 2.551 2.691 0.969 0.03 0.332 48.252 9.28 46.8451 5.33 1.59 0.856 31.1.023 0.71 1.676 6.735 3.379 21.332 13.37 1.85 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 1.450 4.5 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.36 1.75 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.35 1.268 0.139 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.5 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 2.79 0.275 1.05 0.50 35.126 13.

84

**Appendix C FEMA Force Distribution Calculations
**

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

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

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

Floor, j

mj (10-3×kg)

s* = j

∑i mi

0.112 0.110 0.110 0.110 0.110 0.110 0.110 0.110 0.119

mj

1 2 3 4 5 6 7 8 9

503.5 494.7 494.7 494.7 494.7 494.7 494.7 494.7 534.1

C.2

EQUIVALENT LATERAL FORCE (ELF) DISTRIBUTION 85

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

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

Floor j 1 2 3 4 5 6 7 8 9

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

-3 k

s* j

=

m jhk j

∑i mihik

0.007 0.020 0.038 0.062 0.091 0.126 0.165 0.210 0.281

C.3

SRSS DISTRIBUTION

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

N

∑ n (V jn )

2

to get column 8 of

9 -153.2 95.8 374.9 832.047 0.3 222.7 2065.0 1231. Table C.0 980.1 -525.6 1233.7 -46.9 1683.5 320.8 430.1 Story Shears V j2 (kN) (6) 1114.7 1622.1 832.7 101.070 0.7 Vj (kN) (8) 2268.5 215.1 -438.2 fj = ∑ fi i fj (2) (3) (4) 59.2 97.1 87.3.7 694.6 -732.0 1476.7 355.0 136.3 -646.0 -5.5 159.6 -352.7 525.0 381.6 366.2 277.7 95.7 1578.7 234.6 286.9 366.3 -6.4 400.1 1857.2 105.0 354.7 374.4 1842.045 0. 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) 1917.2 200.9 446.2 148.3.9 880.4 250.0 176.5 -973.7 (10) 0.5 -320.6 319.4 1759.1 -646.8 -326. the lateral forces are normalized by the base shear to obtain column 10 in Table C.8 1381.3 240.9 -166.177 0.042 0.2 Lateral Force fj (kN) (9) 203.6 -359.065 0.098 0.5 -350.090 0.6 V j3 (kN) (7) 478.For convenience.367 87 .2 285.9 1446.1 -967.

- 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
- Action Plan for Performance Based Seismic Designby Daniel Reynolds
- 2013 Pushover Analysis Guide Final RC Building With Masonry Infill Walls EditedDraft January 2013by Juan Carlos Jiménez Pacheco

- Pushover analysis
- Pushover Analysis using ETABS and SAP2000
- EVALUATION OF PUSHOVER ANALYSIS PROCEDURES
- Pushover Analysis Final
- Pushover Ppt
- Performance Based Design Using Nonlinear Analysis
- Non Linear Analysis Pushover
- Non-Linear Static Push Over Analysis
- Nonlinear Pushover Analysis of Reinforced Concrete Structures
- Performance Based Seismic Design
- Action Plan for Performance Based Seismic Design
- 2013 Pushover Analysis Guide Final RC Building With Masonry Infill Walls EditedDraft January 2013
- Pushover Analysis Using Etabs
- A comparison of single-run pushover analysis techniques for
- SAP-Etabs-Course-360
- Performing Pushover Analysis in Etabs
- Thesis
- Etabs-step by Step
- Pushover Analysis
- ATC-40 Vol 2
- ETABS Pushover Analysis
- A Modal Pushover Analysis Procedure to Estimate
- 15WCEE 1940 pushover analysis.pdf
- Pushover
- Pushover
- Etabs Guide
- pushover
- ATC-40
- About Pushover Analysis and SAP 2000 Method
- Seismic+Design+Aids+for+Nonlinear+Pushover+Analysis+of+Reinforced+Concrete+and+Steel+Bridges
- 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