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

# A MODAL PUSHOVER ANALYSIS PROCEDURE TO ESTIMATE SEISMIC DEMANDS FOR BUILDINGS: THEORY AND PRELIMINARY EVALUATION ANIL K.

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 .

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

iv .

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

vi .

.................................................2 Response History Analysis ......................................................33 4........4.........4......................................................................2 FEMA-273 Pushover Analysis ...........3 2..........................................................................2 Modal Pushover Analysis .............................4 Modal Response History Analysis..36 4................... Comparison of Modal and FEMA Pushover Analyses.....................................37 Comparative Evaluation of Analysis Procedures ...........38 4..............................55 Comparative Evaluation......................................2...........................................................27 Uncoupled Modal Response History Analysis ...12 Comparative Evaluation of Analysis Procedures ...........................................................12 Modal Pushover Analysis ....................1 Underlying Assumptions and Accuracy ....................................................................................................................................1 Uncoupled Modal Response History Analysis ..................34 4...............................................................................................................................................................................47 4.......................................13 3..2 3..22 4.3 System and Excitation Considered ...............9 Modal Response Spectrum Analysis ........9 3...............................................................................................3 Summary..............4 3...........................................................................................................6 Pushover Analysis ....3 3.6 Elastic Multistory Buildings ........................... iii Acknowledgment .....................4.................................................................. Inelastic Multistory Buildings ...................17 3.......................38 4.....................................................4 5...........................................................................3 2.........................1 3...................................................................................3 Modal Pushover Analysis .....................v Table of Contents...................................................1 System and Excitation Considered .................4...........3 Modal Pushover Analysis with Gravity Loads .........................................................................................................55 5......27 4.........2.........2 Response History Analysis ...........................................................................................28 4...........55 vii ...............................................41 4.............................5 Response History Analysis ...............................................3.......1 Summary.....................CONTENTS Abstract ........................................... 2............................1 5........................................................................1 2.........13 3...............................................4....................................................................................................... Equation of Motion ...................3 4........36 Modal Pushover Analysis ..................................................... Introduction..............................1 One-Story Systems ..........................................................................2 Properties of nth-“mode” Inelastic SDF System ...................................................................................................................................2 2........1 4.....................................2.......................... vii 1.............4.............................

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

1998. such satisfactory predictions of seismic demands are mostly restricted to low. 2000]. Gupta and Krawinkler. current structural engineering practice uses the nonlinear static procedure (NSP) or pushover analysis in FEMA-273 [Building Seismic Safety Council. Miranda. requires explicit consideration of inelastic behavior of the structure. 1998. 1991.1 Introduction Estimating seismic demands at low performance levels. such as life safety and collapse prevention. but investigations [Saiidi and Sozen. 1999. 1988. Maison and Bonowitz. 1994. 2000. 1981. they are conceptually complicated and computationally demanding for routine application in structural engineering practice.. None of the invariant force distributions can account for the contributions of higher modes to response. While nonlinear response history analysis (RHA) is the most rigorous procedure to compute seismic demands. 1999. 1999. after the structure yields both assumptions are approximate. 1997]. 2000] have led to good estimates of seismic demands. While these adaptive force distributions may provide better estimates of seismic demands [Gupta and Kunnath. Skokan and Hart. Kim and D’Amore. Fajfar and Fischinger. 1997. Gupta and Kunnath. 1999]. 1988. Attempts 1 . Krawinkler and Seneviratna. Gupta and Krawinkler. several researchers have proposed adaptive force distributions that attempt to follow more closely the time-variant distributions of inertia forces [Fajfar and Fischinger. 2000]. 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. or for a redistribution of inertia forces because of structural yielding and the associated changes in the vibration properties of the structure. Obviously. Lawson et al. To overcome these limitations. However.. 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. Bracci et al.and medium-rise structures in which inelastic action is distributed throughout the height of the structure [Krawinkler and Seneviratna.

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

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

in addition to the form of the force-deformation relation. z . u £ u y ). 2001. m depends on three system parameters: w n . 2. Tn = 2p w n is the natural vibration period. and u y ..1.3) ωn = k m ζ = c 2mω n f fs = s fy (2.fS m fs c ut u fy 1 αk k 1 uy ug -fy k k um u 1 1 (a) (b) Fig. or absolute (without regard to algebraic sign) maximum.ug t n where bg bg b g bg (2. sign u = .e.4) and w n is the natural vibration frequency. and R y (Chopra . bg 4 . and the ductility factor is u m= m uy (2. (a) Idealized one-story structure. The peak. and z is the damping ratio of the system vibrating within its linear elastic range (i. the deformation u t depends on three systems parameters: w n .1) is divided by m to obtain ~ u + 2zw n u + w 2 u y f s u. This becomes evident if Eq. Section 7. and (b) bilinear hysteretic forcedeformation relation For a given excitation u g t .5) For a given u g t . (2.3). z y . deformation is denoted by um .

4 in Chopra. α = 0.2311 .826 kips) for w = 169.2.04 40 Base Shear.3.26 kN (8. m h = 3.1. is f y / w = ( f o / w ) ÷ 8 = 0. 2. 2. One-story. f o w = 184 . this system and excitation. one-bay frame 50 uy = 1.32 m • Fig. 2001) and scaled up by a factor of 2. For . Vb (kN) 30 20 10 0 0 2 4 6 Roof Displacement.2. 6. The yield strength of the inelastic system. and f y = 39.66 m • EIb EIc EIc • • L = 7.2 5 . based on Ry = 8 .376 cm. 2.2: the dimensions and flexural rigidity of the structural elements are noted. Pushover curve for structure shown in Fig.5 sec and z = 5% subjected to the north-south component of the El Centro (1940) ground motion (Fig.2 SYSTEM AND EXCITATION CONSIDERED Consider the one-story system in Fig.9 kN (38. Vby = 39.26 kN. u (cm) 8 Fig. 2.2 kips) . with Tn = 0.

. Observe that the pushover curve matches the initial loading path of the hysteretic system. the energy dissipated in 6 . and E = 2 × 108 kPa (29×103 ksi) gives yield moments of 21.36 cm .3 in.0217 rad . wherein the hysteretic force-deformation history of Fig. as apparent in Fig. A static pushover analysis of this one-story system leads to the force-displacement relationship shown in Fig.4 ).65 kN-m (191.3. (d) shows the rotation q p bt g of the plastic hinges at the bg bg beam ends. 2.1 kip-in.4edetermined from RHAis superimposed. the ductility factor m = 5. the joint rotation and beam hinge rotation are identical to values q m and q pm determined from RHA.) and 50. The yielding stiffness of each structural element is defined as 3% of its initial stiffness. and (e) shows the force-deformation relation. 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 . respectively.4 shows the earthquake response of the system described in the preceding section determined by response history analysis (RHA). Implementing this analysis with I c = 6. 2.35.4 ). the peak value determined from RHA.4f.36 cm. 2. and q pm = 0. However.18 kN-m (444.077 × 107 mm 4 (146 in.3 RESPONSE HISTORY ANALYSIS Figure 2. The system is excited well beyond the yield deformation. The peak values of the various response quantities are as follows: um = 7. q m = 0. 2. 2.4e.134 × 107 mm 4 (75. pushover analysis cannot provide any cumulative measure of response.017 rad . 2.The yield moments in the beam and columns are defined as the bending moments due to the lateral force f y .6 kip-in.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. Determined from pushover analysis at the exact peak deformation. e.) for the beam and columns. I b = 3. (b) shows the lateral force f s t or base shear Vb t normalized relative to the bg weight. (c) shows the joint rotation q bt g .g. It is organized in five parts: (a) shows the deformation u t . The resulting pushover curve is shown in Fig.

(d) plastic hinge rotation.04 0.4 0.0217 • θ (rad) 0 (c) −0.04 0. 15 u (cm) 0 • 7. or the cumulative rotation at a plastic hinge.04 0. (e) force-deformation relation.017 • 0 (d) θp (rad) −0.2 Vb / w 0. This represents an inherent limitation of pushover analyses.5 0.2 −0.2311 (a) Vb / w 0 f / w = −0. Response of one-story system to El Centro ground motion: (a) deformation.5 fy / w = 0.4 −10 −5 0 u (cm) 5 10 −0.2 0 (e) 0 (f) −0.36 −15 0.4 0. 2.2311 y (b) −0.2 −0.04 0 5 10 15 Time (sec) 20 25 30 0. (c) joint rotation. and (f) pushover curve 7 . (b) base shear.yielding during the ground motion.4 −10 −5 0 u (cm) 5 10 Fig.4.

8 .

(3.3.m ι u g t bg bg (3.2) 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 .1) can be interpreted as effective earthquake forces: p eff t = . and p eff t = Â peff.4) 9 .mι ug t bg bg bg (3. and k are the mass. This force distribution can be expanded as a summation of modal inertia force distributions s n (Chopra.1) where u is the vector of N lateral floor displacements relative to the ground.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 = . classical damping. c.12): mι = n =1 Â sn = N n =1 Â Γ nmφ n N (3.n t = Â .snug t n= n =1 The effective earthquake forces can then be expressed as af N af N af (3. Elastic Multistory Buildings 3.3) where φ n is the nth natural vibration mode of the structure. each element of the influence vector ι is equal to unity. and lateral stiffness matrices of the system. m. 2001: Section 13. The right side of Eq.

6) Next. The equations governing the response of the system are mu + cu + ku = .peff (t ) = n =1 ∑ peff . p eff . we will outline that the response of the MDF system to p eff .9) to the equation of motion for the nth-mode elastic SDF system. (3. (3.n t = . it can be demonstrated that none of the modes other than the nth mode contribute to the response.sn ug t bg (3.10) bg bg bg 10 (3.9) and (3. with no contribution from other modes.Gn ug t (3.u g t n Comparing Eqs.9) in which w n is the natural vibration frequency and z n is the damping ratio for the nth mode.11) and substituting in Eq. Then the floor displacements are u n t = φ n qn t bg bg bg bg (3.10) gives qn t = Gn Dn t bg bg bg (3.sn ug t bg bg bg bg (3.7) By utilizing the orthogonality property of modes.n (t ) = N n =1 ∑ −sn u g (t ) N (3. (3.5) The contribution of the nth mode to s and to p eff t are: sn = Gn mφ n respectively. subjected to u g t : Dn + 2z nw n Dn + w 2 Dn = .8) gives the floor displacements u n t = Gnφ n Dn t bg (3. The solution qn t can readily be obtained by comparing Eq.n t is entirely in the nth-mode.12) .8) where the modal coordinate qn t is governed by 2 qn + 2z nw n qn + w n qn = . an SDF system with vibration propertiesnatural frequency w n and damping ratio z n of the nth-mode of the MDF system.

However. and Eqs. and An t = w 2 Dn t n bg (3. Therefore.9) is the standard modal equation governing qn t .13) represent the response of the MDF system to p eff. (3. 2001.Forces sn An(t ) st rn ωn. etc.15) rt = bg n =1 Â rn t = bg n =1 Â rnst An bt g N (3. In contrast to the classical 11 bg .13) st where rn denotes the modal static response.1). (3. 3. the static value of r due to external forces sn . Conceptual explanation of modal response history analysis of elastic MDF systems Any response quantity r t story drifts.12) and (3.15) and (3.n t .14) is the pseudo-acceleration response of the nth-mode SDF system (Chopra. internal element forces.12) and (3. st The two analyses that lead to rn and An t are shown schematically in Fig. the response of the system to the total excitation p eff t is ut = bg bg bg n =1 N Â un bt g = Â Gnf n Dn bt g n =1 N N (3.13) define the contribution of the nth-mode to the response. ζn ¨ g(t ) u (a) Static Analysis of Structure (b) Dynamic Analysis of SDF System Fig. these standard equations have been derived in an unconventional way.can be expressed by st rn t = rn An t bg bg bg bg bg (3. Eqs.1. Equations (3.16) reflect combining the response contributions of all modes.1.16) This is the classical modal RHA procedure wherein Eq. 3. (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. z n of the pseudo-acceleration response (or design) spectrum for the nth-mode SDF system.17) where An is the ordinate A Tn .8.derivation found in textbooks (e. This concept provides a rational basis for the modal pushover analysis procedure developed later.g. the peak nth-mode response as in Eq. 3. we note that static analysis of the structure subjected to lateral forces f no = Γ n mφ n An (3. we used the modal expansion of the spatial distribution of the effective earthquake forces. (3. provides an estimate of the peak value of the total response: b g F N r 2 I 1/ 2 ro ª G Â no J H n =1 K 3. 2001. Chopra.17) (Chopra.1.3).16). Alternatively. Section 13.2 MODAL RESPONSE SPECTRUM ANALYSIS The peak value ro of the total response r t can be estimated directly from the response spectrum for the ground motion without carrying out the response history analysis (RHA) implied in Eqs. 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. The SRSS rule.. 2001. and Tn = 2p w n is the natural vibration period of the nth-mode of the MDF system.1).9) .4 and 13. (3.18) To develop a pushover analysis procedure consistent with RSA.19) will provide the same value of rno . In such a response spectrum analysis (RSA). Sections 12. which is valid for structures with well-separated natural frequencies such as multistory buildings with symmetric plan.(3.3 MODAL PUSHOVER ANALYSIS (3. this response value can be obtained by static analysis of the structure subjected to lateral forces distributed over the building height according to 12 .

4. SAC is a joint venture of three non-profit organizations: The Structural Engineers Association of California (SEAOC). (3.21) 2 where Dn = An ω n . urno . region. can be combined according to Eq. which from Eq.73 m (150 ft) in plan. 46th Street. shown in Fig.2. 3. The columns are 345 MPa (50 ksi) steel wideflange sections. and California Universities for Research in Earthquake Engineering (CUREE). with the ninth level being the roof. Third St. This modal pushover analysis (MPA) for linearly elastic systems is equivalent to the well-known RSA procedure (Section 3.s* = mφ n n (3.2).20) and the structure is pushed to the roof displacement. this building is 45. 13 . in both directions.18) to obtain an estimate of the peak value ro of the total response. 1660 W.19 m (122 ft) in elevation. The building has a basement level. The bays are 9.4 COMPARATIVE EVALUATION OF ANALYSIS PROCEDURES 3. Obviously Dn and An are available from the response (or design) spectrum. SAC Steel Project Technical Office. the Applied Technology Council (ATC). The peak modal responses. A benchmark structure for the SAC project. Although not actually constructed. each determined by one pushover analysis. (3. Los Angeles. CA 90017. rno . CA 94804-4698. with five bays each in the north-south (N-S) and east-west (E-W) directions.2). Consulting Structural Engineers. 1301 S. Richmond.1 System and Excitation Considered The 9-story structure. The interior bays of the structure contain simple framing with composite floors.73 m (150 ft) by 45. California.12) is urno = Gnf rn Dn (3. 3. denoted B1. the peak value of the roof displacement due to the nth-mode. and 37. this structure meets seismic code and represents typical medium-rise buildings designed for the Los Angeles. was designed by Brandow & Johnston Associates1 for the SAC2 Phase II Steel Project. Typical floor-to-floor heights (for analysis purposes measured from center-of-beam to center1 2 Brandow & Johnston Associates.15 m (30 ft) on center.. 3. 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. The levels of the 9-story building are numbered with respect to the ground level (see Fig.

00×106 kg (616 kips. The building is modeled in DRAIN-2DX (Allahabadi and Powell. The column lines employ two-tier construction. 1988) using the M1 model developed by Krawinkler and Gupta (1998). This model is based on centerline dimensions of the bare frame in which beams and columns extend from centerline to centerline.65×105 kg (66.01×106 kg (69. The strength.83 m (6 ft) above the center-line of the beam to column joint. The floor-to-floor height of the basement level is 3. fifth. 3.96 m (13 ft). The seismic mass of the structure is due to various components of the structure. The first three vibration modes and periods of the building for linearly elastic vibration are shown in Fig.20). for the first level is 1. third. and shear distortion of panel zones are neglected but large deformation (P∆) effects are included. s* (Eq. which are seismic (tension) splices to carry bending and uplift forces.sec2/ft). respectively.2. To ensure that this structure remains elastic. Concrete foundation walls and surrounding soil are assumed to restrain the structure at the ground level from horizontal displacement. the vibration periods are 2.7 kips-sec2/ft). including the steel framing. The floor system is composed of 248 MPa (36 ksi) steel wide-flange beams in acting composite action with the floor slab. ceiling/flooring. mechanical/electrical. 0. The column bases are modeled as pinned and secured to the ground (at the B-1 level). and 0. Column splices. if desired more complex models. Each frame resists one half of the seismic mass associated with the entire structure.49 sec. such as those described in Gupta and Krawinkler (1999) can be used. These force n distributions will be used in the pushover analysis to be presented later. The simple model adopted here is sufficient for the objectives of this study. and seventh levels at 1.2 kips-sec2/ft).0 kips-sec2/ft). we select a weak ground motion: the northsouth component of the El Centro (1940) ground motion scaled down by a factor of 4.3. dimension.0 kips-sec2/ft). The 9-story N-S MRF is depicted in Fig. 3. The force distributions. 3.27. for the first three modes are shown in Fig.89×105 kg (67. floor slabs. partitions.85.4. roofing and a penthouse located on the roof. 3.49 m (18 ft). The seismic mass of the above ground levels of the entire structure is 9. monolithic column pieces are connected every two levels beginning with the first level. for the second through eighth levels is 9.07×106 kg (73. 14 . are located on the first. The seismic mass of the ground level is 9..65 m (12 ft) and for the first floor is 5. i.of-beam) are 3.e. and for the ninth level is 1.

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

44 1. respectively.05 1.12) and (3.9th 8th 7th 6th T = 0.5 0 0.85 sec 2 Floor 5th 4th 3rd 2nd 1st T1 = 2.6.03 −1. determined by RHA [Eqs.13)].4.1 −2. 3.0272 −2.2 Response History Analysis The structural response due to individual vibration modes.12 0. is shown in Figs. n = 1.4.7. 2.5 Mode Shape Component 1 1.93 −1.5 −1 −0.27 sec Ground −1.3.39 3. and 3.72 −2.13 −1.67 −1. (b) shows the base shear Vbn t normalized relative to the weight W of the building.5.796 0.04 1.94 2.8 −2. n = 1.05 2.33 2.49 sec 3 T = 0. 3.31 −0.51 0.61 2.05 s * 1 s * 2 s * 3 Fig.728 2. and 3 n 3. Each figure is organized in four parts: (a) shows the roof displacement urn t . 3. (3.75 1.37 2.487 −1. (c) shows the joint rotation q n t of an external joint at 16 bg bg bg .38 0. First three natural-vibration periods and modes of the 9story building 3. 2. and 3 . 3.5 Fig.1 3. Force distributions s* = mφn .

and 3. The linear relationship between the base shear and roof displacement for each mode implies that the structure did not yield.16)]. It is apparent that the first mode alone is inadequate. 3.3. The same method was used to determine the peak values of many response quantities. The peak values of floor displacements and story drifts determined by RHA. respectively.422 cm. in particular. and ur 3o = 0.23 cm. This is illustrated in Fig. and three vibration modes. errors generally decrease as response contributions of more modes are included. two.1. three. The errors due to truncating the contributions of vibration modes beyond the third mode are negligible. respectively.12 cm.1.3.9. 3. The peak values of displacements of all floors. and 3. larger in story drifts. with the first three modes provide essentially the exact response.2. For a fixed number of modes included. 3. 3. 3. consistent with the increasing significance of the higher mode response among these three sets of response quantities. and rotations of external joints with moment connections are presented in Tables 3. As expected. which are listed in Tables 3. The peak values of the various response quantities are noted in these figures. the exact response considering all modes.2. two. the peak roof displacement due to each of three modes is ur1o = 9. and even larger in joint rotations. Combining the modal response histories for all modes gives the total response [Eqs. the results for the roof displacement and top-story drift are shown in Fig. drifts in all stories. are presented in Fig. but three modes—perhaps even two modes—are sufficient. (3. and the percentage errors due to truncation of higher modal contributions. including one.15) and (3.8. where the second and third modal responses are a larger percentage of the top story drift compared to roof displacement. and (d) shows the Vbn . errors are smallest in floor displacements. 17 .urn relation. especially in estimating the story drifts. or all modes.8. ur 2o = 2. Also included are the combined response due to one.the roof level.

413 RHA (all modes) 0.229 0.264 Error (%) 1 Mode 2 Modes 3 Modes -23.2 -20.5 -2.307 0.097 0.260 0.2 -57.401 Error (%) 1 Mode 2 Modes 3 Modes -23.01E-04 -2.64E-03 3.043 0.9 -15.225 0.0 7.3 -0. 18 .333 0.088 -0.202 0.202 0.097 0.4 -22.09E-03 1.1 -2.1 Peak values of floor displacements (as % of building height = 37.0 -46.4 -1.400 0.6 4.192 0.203 0.1 -2.045 0.062 0.3 Table 3.8 -56.24E-03 2.235 0.071 0.5 0.00E-03 2.7 -50.22E-03 2.9 2.9 8.311 0.63E-03 2.9 -16.173 0.2 0.042 0.38E-03 2.44E-03 3.3 -8.91E-04 1.63E-03 2.03E-03 1.11E-04 -5.364 0.023 0.7 3.65E-03 RHA (all modes) 2.0 -2.179 0.4 -7.124 0.89E-03 1.226 0.266 0.069 0.8 -5.7 2.181 0.2 -4.14 m3) from RHA for 0.00E-03 1.183 0.0 -0.038 0.09E-03 2.7 2 Modes -3.011 0.069 0.47E-03 1.42E-04 1.026 0.14E-03 2.227 0.130 0.001 -0.6 0.058 -0.245 0.1 -14.235 0.65E-03 2.125 0.226 0.74E-04 6.08E-03 2.9 1.378 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 0.282 0.2 Peak values of story drift ratios (as % of story height) from RHA for 0.237 0.4 -3.4 -0.152 0.054 0.3 -3.50E-03 4.74E-03 1.8 1.9 -24.66E-05 -3.76E-03 1.157 0.56E-03 2.258 0.227 0.310 0.2 -2.15E-03 4.197 0.33E-03 2.09E-03 Error (%) 1 Mode -23.006 0.062 -0.1 -2.300 0.99E-03 Mode 3 3.72E-03 3.4 -53.8 1.1 3.090 0.9 -22.080 0.406 0.13E-03 2.032 -0.78E-04 -3.024 -0.124 0.03E-03 -6.295 -0.9 3.29E-03 2.1 Table 3.009 -0.042 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 0.259 0.9 9.31E-03 Joint Rotation (rad) Combined (RHA) 1 Mode 2.3 19.321 0.89E-03 1.199 0.475 0.6 -1.156 0.1 Building height is measured from the ground floor to the 9th floor.03E-03 1.6 9.336 0.008 -0.325 0.0 -10.350 0.8 -10.3 -0.022 0.060 -0.0 -2.8 -15.260 0.44E-03 1.73E-03 3.407 0.88E-03 2.25 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 3 Modal Response Mode 1 2.275 0.266 0.055 0.03E-03 3 Modes 2.11E-03 1.7 -19.4 -10.003 0.010 -0.177 0.74E-04 9.453 0.012 0.31E-03 Mode 2 -1.097 0.202 0.45E-03 3.089 0.159 0.50E-03 2.133 -0.011 0.370 0.237 0.311 0.74E-03 1.229 0.266 0.90E-03 3.4 -1.6 -1.1 -19.2 1.035 0.253 0.173 0.1 -0.00E-03 1.9 -23.Table 3.76E-03 1.002 -0.399 0.5 18.26E-04 -5.0 3 Modes -5.94E-03 2.6 -0.6 11.060 0.253 0.125 0.259 0.5 -1.2 9.466 0.4 -6.177 0.88E-03 2.1 -11.117 0.38E-04 2 Modes 2.99E-03 2.261 -0.7 7.85E-03 3.156 0.265 RHA (all modes) 0.8 -1.152 0.28E-04 1.4 0.2 -1.060 -0.4 -41.1 4.121 0.265 0.245 0.09E-03 1.3 -33.13E-04 9.7 4.263 0.303 0.015 -0.282 0.012 -0.231 -0.205 0.01E-04 3.6 0.6 1.3 Peak values of joint rotations (radians) from RHA for 0.

0435 9.12 −0.5 15 −0. (b) base shear.25 × El Centro ground motion 19 .1 /W 0. (c) joint rotation.0435 Vb1 / W 0.12 0 (d) 0 (e) 15 Fig.5 0 u (cm) r1 7. (d) force-deformation history.1 −15 −7.1 0. Response due to first mode: (a) roof displacement.1 5 x 10 −3 θ (rad) 0.12 • 0 (a) r1 −15 0.0435 • 0 (b) V b1 −0.1 0. Excitation is 0. 3.00174 • 0 (c) r1 −5 0 5 10 15 Time (sec) 20 25 30 0. and (e) pushover curve.5 9.1 −15 −7.15 u (cm) 9.5 0 u (cm) r1 7.5.

1 0.23 −5 (a) 0.5 0 u (cm) r2 2.5 0 u (cm) r2 2.23 −2.00244 • 0 (c) r2 −5 0 5 10 15 Time (sec) 20 25 30 0.5 5 Fig.0252 • 0 (b) V −0. (c) joint rotation.0252 −2. and (e) pushover curve.1 Vb2 / W 0.0252 (d) 0 0 (e) −0.5 5 −0. 3.6.1 −5 −2.05 /W 0.1 −5 −2.5 u (cm) r2 0 • 2. Response due to second mode: (a) roof displacement. Excitation is 0.23 0. (b) base shear.05 5 x 10 −3 b2 θ (rad) 0. (d) force-deformation history.25 × El Centro ground motion 20 .

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

confirming that MPA gives the exact values of the individual modal responses. and ur 3o = 0.6. considering one.4. i.3). respectively. i.3.e.20) with n = 2 and 3 up to roof displacements ur 2 o = 2. 3. pushing the structure. respectively. and external joint rotations computed by pushover analysis are presented in Tables 3. In practical application. 3.3). and it is apparent in the example considered that three modes provide most of the response (Fig. 3. 3. 3. 3. In RHA the errors arise only from truncating the responses due to higher modes. (3.4. (3.3). The floor displacements. the value determined by RHA (Fig.5.4) to roof displacement ur1o = 9. and 3. which would provide the Dn value to be substituted in Eq.12 cm. Figure 3. and 3.21). except for the algebraic sign associated with Γ1 and minor round-off errors. 3. (3.422 cm . and the computed response values are identical to the peak response values determined from RHA (Tables 3.2.5.ur relations determined by RHA (Figs.. the errors in the MPA results are generally larger than in RHA (Fig. and 3.4. As suggested by Eq.4. and external joint rotations in Tables 3.18). 3. pushing the structure using the force distribution of Eq. 3. respectively. using the force distribution of Eq. 3.1.e. This pushover curve is consistent with the relationship between the base shear and roof displacement determined by RHA (Fig.2.12).5.3 Modal Pushover Analysis Implementing MPA for the fundamental vibration mode. two. For a fixed number of modes included considered.6e and 3. Observe that the target roof displacement in each pushover analysis is identical to its exact value determined by RHA.7e and to the floor displacements. These values of the response quantities are identical to the peak response values determined from RHA (Tables 3.5d). or three vibration modes.7d). these pushover curves are consistent with the Vb .1 through 3.6 present estimates of the combined response according to Eq.. Implementing pushover analysis for the second and third modes. implying that the modal truncation errors are small if at 22 . this value would be determined directly from the response (or design) spectrum. story drifts.6. story drifts.6d and 3. although both analyses led to identical peak values of the individual modal responses.9).20) with n = 1 (Fig.9 and Tables 3.1. 3. and 3. and 3.5e.10 and Tables 3. and the errors in these estimates relative to the exact response from RHA considering all modes. As for the first mode. leads to the pushover curves shown in Figs.23 cm . (3. 3. 3. (3.8) leads to the pushover curve shown in Fig. the floor displacements are proportional to the mode shape φ 1 because the structure remains elastic.

83 • r3 0 • 0.5 0.1 0.12 • r1 −15 15 0 r1 0 0 0.3 0. 3.least three modes are included. Additional errors are introduced in pushover analysis due to the approximation inherent in modal combination rules. Heightwise variation of floor displacements and story drifts from RHA for 0. Response histories of roof displacement and top-story drift from RHA for 0.59 • Total (All Modes) 0 −15 0 ∆ (cm) u (cm) 0 −3 0 fig3_8.03 −3 3 Mode 3 ∆ (cm) u (cm) r3 −15 15 9. sec 25 30 Fig.eps r 5 10 15 20 Time.6 Fig.3 0.4 Displacement/Height (%) 0.9.eps Ground 0 0.1 0.2 0.25 × El Centro ground motion: first three modal responses and total (all modes) response (a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th (b) Story Drift Ratios Floor 4th 3rd 2nd 1st Floor 5th RHA All Modes 1 Mode 2 Modes 3 Modes 5th 4th 3rd 2nd 1st RHA All Modes 1 Mode 2 Modes 3 Modes fig3_9a. sec 25 30 r 5 10 15 20 Time. (a) Roof Displacement 15 (b) Top Story Drift 3 Mode 1 ∆ (cm) u (cm) 9.685 • −3 3 Mode 2 ∆ (cm) u (cm) • 2.2 0. 3.48 −3 3 1.25 × El Centro ground motion 23 .8.5 Ground 0 0.23 r2 −15 15 0 • 0.eps fig3_9b.4 Story Drift Ratio (%) 0.422 r2 0 • 1.

173 0.226 0.2 -11.259 -0.179 0.3 -19.96E-03 2.Table 3.088 -0.9 -12.4 -22.012 -0.321 0.229 0.2 -16.9 -15.97E-03 1.370 0.8 -23.260 0.8 -14.00E-03 1.44E-03 3.179 0.011 0.231 -0.4 -19.6 -16.2 1.5 Peak values of story drift ratios (as % of story height) from MPA for 0.74E-03 1.261 -0.15E-03 2.001 0.264 Error (%) Modal Response 1 Mode 2 Modes 3 Modes -23.25 × 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 0.76E-03 -1.8 -15.038 -0.53E-04 -9.33E-04 5.8 -22.40E-04 5.300 0.89E-03 -1.4 -2.237 0.09E-03 -1.9 -24.9 -14.0 -2.3 1.57E-03 1 Mode -23.03E-03 6.9 -14.270 0.1 -11.267 0.15E-03 1.90E-04 -9.22E-03 -2.05E-03 3.313 0.042 0.055 0.31E-03 2.4 -5.25 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Modal Response Mode 1 -2.012 0.3 -2.4 0.002 0.230 -0.024 0.080 0.328 0.203 0.3 -41.9 -18.2 -0.9 3 Modes -12.106 0.65E-03 2.2 -57.026 0.310 0.245 0.125 0.99E-03 Mode 3 -3.097 0.276 0.090 0.9 -16.47E-03 1.253 0.156 0.124 0.1 -0.92E-04 -1.04E-03 3.0 -10.7 -15.9 -13.401 Error (%) 1 Mode 2 Modes 3 Modes -23.73E-05 3.31E-03 Joint Rotation (rad) Combined (MPA) 1 Mode 2.156 0.9 -13.253 -0.7 -19.336 0.2 -20.008 -0.227 0.6 -17.14 m) from MPA for 0.121 0.230 0.09E-03 1.253 0.0 -46.197 0.7 2 Modes -13.133 0.058 -0.069 0.266 0.2 Table 3.048 0.72E-03 3.043 -0.003 -0.133 -0.3 Table 3.9 -15.042 0.181 0.4 -53.173 0.2 -4.270 0.285 0.9 -11.31E-03 Error (%) RHA (all modes) 2.331 0.060 -0.4 -7.80E-04 3.2 -12.3 -13.44E-03 -1.4 -9.00E-03 2.336 RHA (all modes) 0.011 -0.9 -15.260 -0.78E-04 2 Modes 2.0 -18.7 -21.38E-03 2.015 -0.8 -15.8 -56.235 -0.036 -0.24E-03 -2.407 0.071 0.28E-03 2.048 0.203 0.177 0.060 0.286 0.7 2.42E-04 -1.76E-03 1.274 0.097 0.03E-03 -1.6 -19.010 0.63E-03 2.7 24 .43E-04 -1.3 -33.00E-03 3.0 -16.260 0.63E-03 -2.9 -14.4 Peak values of floor displacements (as % of building height = 37.00E-03 2.253 RHA (All Modes) 0.9 -13.4 -14.1 -18.245 -0.6 -15.079 0.03E-03 3.9 -8.272 0.203 0.22E-04 2.152 0.94E-03 2.062 0.9 -13.9 -15.022 0.235 0.74E-04 -6.237 -0.203 0.74E-03 -1.332 0.1 -19.296 -0.157 0.3 -9.157 0.89E-03 2.08E-03 2.4 1.009 0.0 -0.282 -0.4 -9.4 -11.385 0.270 0.90E-03 1.09E-04 -3.89E-03 2.259 0.00E-03 -1.227 0.89E-03 1.267 -0.106 0.152 0.203 0.3 -14.73E-03 3.006 -0.282 0.25 × El Centro ground motion Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Modal Response Mode 1 Mode 2 Drift Ratio (%) Combined (MPA) 1 Mode 2 Modes 3 Modes Mode 3 -0.4 -22.03E-03 1.09E-03 Mode 2 1.117 0.38E-03 3 Modes 2.4 -4.177 0.3 -12.4 -4.7 -50.062 -0.032 0.045 -0.078 0.6 Peak values of joint rotation (radians) from MPA for 0.5 -16.069 0.133 0.466 0.267 0.00E-03 2.12E-03 1.322 0.89E-03 -2.023 -0.229 0.374 0.1 -11.125 0.

5 Ground 0 0.eps Ground 0 0.(a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th (b) Story Drift Ratios Floor 4th 3rd 2nd 1st Floor 5th RHA All Modes MPA 1 Mode 2 Modes 3 Modes 5th 4th 3rd 2nd 1st RHA All Modes MPA 1 Mode 2 Modes 3 Modes fig3_10a.25 × El Centro ground motion. shading indicates modal combination error 25 . 3.6 Fig.1 0. Heightwise variation of floor displacements and story drifts from MPA for 0.5 0.3 0.4 Story Drift Ratio (%) 0.eps fig3_10b.2 0.4 Displacement/Height (%) 0.2 0.10.3 0.1 0.

26 .

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

Eq.3) in Eq.9) for linearly elastic systems.2 UNCOUPLED MODAL RESPONSE HISTORY ANALYSIS bg Neglecting the coupling of the N equations in modal coordinates [Eq. (4. premultiplying by φ T . However.and classical n damping-orthogonality property of modes gives F qn + 2z nw n qn + sn = . and using the mass. where φ n are now the modes of the corresponding linear system. 4.4) represents N equations in the modal coordinates qn .… N (4. (4.4) where the only term that differs from Eq.2). (4. For linear systems.sn ug t bg b g bg bg (4. these equations are coupled for inelastic systems.Gn ug t Mn bg n = 1. give the same results for u t as obtained directly from Eq. This approximate RHA procedure is the preliminary step in developing a modal pushover analysis procedure for inelastic systems. (3.5) This resisting force depends on all modal coordinates qn t . (4.9) involves T Fsn = Fsn qn .6) for inelastic systems will no longer be described by Eq. (4.2). in principle. implying coupling of modal coordinates because of yielding of the structure.” implying that other “modes” will also contribute to the solution. sign u = . qr t = 0 for all 28 bg . (4. (4. (3. (3.6b) are mu + cu + f s u.4)] leads to the uncoupled modal response history analysis (UMRHA) procedure. sign qn = φ n f s u n . (3. sign u n b g b g bg (4. 2.6) The solution of Eq. (3.n t given by Eq. however.8) because qr t will generally be nonzero for “modes” other than the nth “mode. (4.4) is rarely solved because it offers no particular advantage over Eq. The spatial distribution s of the effective earthquake forces is expanded into the modal contributions sn according to Eq.2). Equation (4.3) will. Unlike Eq.3).Substituting Eq. Simultaneously solving these coupled equations and using Eq. The equations governing the response of the inelastic system to p eff.

(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

4.02 −150 n=2 −150 50 n=2 14.4.89 50 r1 0 u (cm) 0 • 78.9 • 0 −50 10 −50 10 n=3 ur3 (cm) 0 • 5.3. and 3. where u g t = 3. 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) • 75.575 5 10 15 20 Time (sec) 25 30 ur3 (cm) 0 • 5. 4. n = 1. ζn.n t = .51 • ur2 (cm) 0 ur2 (cm) n=3 14. 2. Roof displacement due to p eff.112 5 10 15 20 Time (sec) 25 30 −10 0 −10 0 Fig.Forces sn Unit mass An(t ) st rn ωn. Fsn / Ln ¨ g(t ) u (a) Static Analysis of Structure (b) Dynamic Analysis of Inelastic SDF System Fig. and (b) approximate solution by UMRHA bg af bg 32 .s n u g t .0 ¥ El Centro ground motion: (a) exact solution by NL-RHA.

(3. the errors in either response quantity are only a few percent. and 3.n t (n = 1.Dn relation is 33 bg bg . both for 3.0 ¥ El Centro ground motion: (a) exact solution by NL-RHA.817 5 10 15 20 Time (sec) 25 30 −10 0 −10 0 Fig. Such comparison for roof displacement and top-story drift is presented in Figs. (4.1 Underlying Assumptions and Accuracy The approximate solution of Eq. Top story drift due to p eff.5.008 • 6.4) and (4. but even for this very intense excitation. respectively.5)] is neglected.744 • 0 −20 10 n=3 −20 10 n=3 ∆r3 (cm) 0 • 5.5. The errors are slightly larger in drift than in displacement. 4. where u g t = 3.0 ¥ El Centro ground motion.855 ∆r1 (cm) 0 ∆r1 (cm) n=2 • 6. (2) the superposition of responses to p eff. and (3) the Fsn Ln .n t = .2.15) is strictly valid only for linearly elastic systems.956 5 10 15 20 Time (sec) 25 30 ∆r3 (cm) 0 • 5.4 and 4. this intense excitation was chosen to ensure that the structure is excited well beyond its linear elastic limit. (4.6) by UMRHA is compared with the “exact” solution by nonlinear RHA. 4. and (b) approximate solution by UMRHA af bg bg 4. 2.2… N ) according to Eq. n = 1.s n u g t . 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.(a) Nonlinear RHA 20 n=1 20 (b) UMRHA n=1 • 5.33 0 −20 20 −20 20 n=2 ∆r2 (cm) 0 ∆r2 (cm) 7.

8) governing Dn t is based on Eq. How to convert this Vbn . (4. this distribution seems to be a rational choiceeven for inelastic systemsto determine Fsn in Eq. Therefore.2 Properties of the nth-“mode” Inelastic SDF System How is the Fsn Ln .20). Although several approximations are inherent in this UMRHA procedure. such nonlinear static analysis provides the so-called pushover curve. is plotted against roof displacement urn . Although most commercially available software cannot implement such displacement-controlled analysis.12) for floor displacements.6a. The structure is pushed using the force distribution of Eq. within the linearly elastic range of the structure.Dn curve.4. 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. the base shear is Vbny and roof displacement is urny . A bilinear idealization of this pushover curve for the nth-“mode” is shown in Fig.Dn relation? The two sets of forces and displacements are related as follows: V Fsn = bn Γn Dn = urn Γ nf rn 34 (4. and the base shear Vbn . 4.approximated by a bilinear curve to facilitate solution of Eq.8) in UMRHA. when specialized for linearly elastic systems it is identical to the RHA procedure of Section 3. (3. When implemented by commercially available software. 4. (3. (3.1.20) to some bg pre-determined roof displacement. Therefore we impose this constraint in developing the UMRHA procedure in this section and modal pushover analysis in the next section.8) before it can be solved? Because Eq. (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. At the yield point.Dn relation to be determined in Eq. However.urn pushover curve to the Fsn Ln . which is different than the Fsn Ln .10) . (4.2.9).9) should be determined by nonlinear static analysis of the structure as the structure undergoes displacements u = Dnφ n with increasing Dn . (4. the relationship between lateral forces f s and Dn in Eq. the only force distribution that produces displacements proportional to f n is given by Eq. (4.

V bn (a) Idealized Pushover Curve F sn /L n (b) F sn / L − D Relationship n n Idealized Vbny Actual 1 αnkn V bny /M * n 1 α ω2 n n kn 1 1 ωn 2 u u rny rn Dny = ur n y / Γn φ r n D n Fig.10 enables conversion of the pushover curve to the desired Fsn Ln − Dn relation shown in Fig. Knowing Fsny Ln and Dny from n Eq.6. 4.6b is w 2 . the initial slope of the pushover curve in Fig.12) implying that the initial slope of the curve in Fig.5). 2001.11). 35 . (4.8). where the yield values of Fsn Ln and Dn are Fsny Ln = Vbny * Mn Dny = urny Γ nf rn (4.13) This value of Tn . should be used in Eq. The two are related through Fsny Ln 2 = w n Dny (4. Properties of the nth-“mode” inelastic SDF system from the pushover curve Equation 4. 4.6b. the elastic vibration period Tn of the nth-mode SDF system is computed from F Ln Dny I 1/ 2 Tn = 2p G H Fsny JK (4.6a is 2 kn = ω n Ln . (4. which may differ from the period of the corresponding linear system. which is not a meaningful quantity.2. 4. In contrast.11) * in which M n = Ln Γn is the effective modal mass (Chopra. 4. Section 13.

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 .6a). Compute the deformation history. 4. Typically. (3. (3. Convert the idealized pushover curve to the Fsn Ln − Dn relation (Fig. Calculate histories of various responses by Eqs. An t . 4. details are available in Appendix A: 1. of the total response r t obtained in Step 8.4. 5.20)]. Compute the natural frequencies. and modes. n (3. 7.6b. develop the base-shear – roof-displacement ( Vbn − urn ) pushover curve for the force distribution s* [Eq. Combine the “modal” responses using Eqs.n t . and pseudo-acceleration history. (3. 4. Calculate the peak value. 8. which is summarized next as a sequence of steps. 6. (3. ωn . Repeat Steps 2 to 6 for as many modes as required for sufficient accuracy.2. the first two or three modes will suffice.3b) with force-deformation relation of Fig. For the nth-mode.20)]. Consider a nonlinear static analysis of the structure subjected to lateral forces distributed over the building height according to s* [Eq.16) to determine the total response. is now determined by 36 bg bg .13). 2. (4. n 3. This value of the roof displacement is given by Eq. with the structure is pushed to the roof displacement urno .3 Summary The inelastic response of an N-story building with plan symmetric about two orthogonal axes to earthquake ground motion along an axis of symmetry can be estimated as a function of time by the UMRHA procedure just developed. 4. r o . for linearly-elastic vibration of the building. 4. the peak value of Dn t .15) and (3. Dn (t ) .21) where Dn .6b) by utilizing Eq. 9. Idealize the pushover curve as a bilinear curve (Fig.11). φn . 4. of the nth-“mode” inelastic SDF system (Fig.12) and (3.

we will refer to rno as the peak “modal” response even in the case of inelastic systems. (4.6 and 7.21). governed by Eq. The peak “modal” responses rno . alternatively. although somewhat intuitive for inelastic buildings. At this roof displacement.18). and (2) it provides the exact modal response for elastic systems.solving Eq. the pushover analysis provides an estimate of the peak value rno of any response rn t : floor displacements.3b) with force-deformation relation of Fig. e. plastic hinge rotations. (3. to obtain an estimate of the peak value ro of the total response. (3.8). As shown in Sections 3.1 Summary The peak inelastic response of a building to earthquake excitation can be estimated by the MPA procedure just developed.3. It provides results for elastic buildings that are identical to the well-known RSA procedure (Section 3.4.2 and 3. 4. as described in Section 4.3.. Calculate the peak roof displacement urno associated with the nth-“mode” inelastic SDF system from Eq.12). Dn .n t .6). etc. Steps 1 to 4 of the MPA are same as those for UMRHA. Compute the peak deformation. of the nth-“mode” inelastic SDF system (Fig. 6. However. Sections 7. which is summarized next as a sequence of steps. or from the inelastic response (or design) spectrum. 37 . 2001. for elastic systems. rno also represents the exact peak value of the nth-mode contribution rn t to response r t . bg bg bg bg Thus. are combined using an appropriate modal combination rule. This application of modal combination rules to inelastic systems obviously lacks a theoretical basis.2. story drifts. 4. Eq. seems reasonable. 5.2. joint rotations. (4. The response value rno is an estimate of the peak value of the response of the inelastic system to p eff. the lateral force distribution used possesses two properties: (1) it appears to be the most rational choice among all invariant distribution of forces. 4. it can be determined from the inelastic response (or design) spectrum (Chopra. as mentioned earlier.3) because. each determined by one pushover analysis.6b by solving Eq. (4. This pushover analysis.g. details are available in Appendix B. it seems reasonable because it provides results for elastic buildings that are identical to the well-known RSA procedure described in Section 3.8).

From the total rotation of a plastic hinge.0 to 3. determined by the UMRHA procedure. 4. two. in contrast to modal analysis (Section 3. (3. extract from the pushover database values of other desired responses. 4. Determine the total response by combining the peak “modal” responses using the SRSS combination rule of Eq. the first two or three “modes” will suffice. and three “modes”. Figure 4. At urno .1 Uncoupled Modal Response History Analysis The structural response to 1. subtract the yield value of hinge rotation to determine the hinge plastic rotation.3 cm. 8. The peak values of displacements of all floors and drifts in all stories are presented in Tables 4.4. Typically. Observe that errors tend to decrease as response contributions of more “modes” are included. The peak values of response are as noted.18).53 cm.2. and the “exact” response from nonlinear RHA for the roof displacement and top-story drift. 9. two. and the errors in the approximate results are shown in Fig.1). in particular.5 ¥ the El Centro ground motion including the response contributions associated with three “modal” inelastic SDF systems. three “modes” included. and ur 3o = 2.7.4 COMPARATIVE EVALUATION OF ANALYSIS PROCEDURES The response of the 9-story building described earlier is determined by the two approximate methods: UMRHA and MPA. also included are the combined responses due to one. the peak roof displacement due to each of the three “modes” is ur1o = 48. rno . although the trends are not systematic as when the system remained elastic (Section 3.1 and 4. the combined response due to three “modes”. the UMRHA procedure lacks a rigorous theory. the 38 .7 shows the individual “modal” responses. 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. and compared with the results of a rigorous nonlinear RHA using the DRAIN-2DX computer program.0. Repeat Steps 3 to 8 for as many “modes” as required for sufficient accuracy. with. 4. the “exact” results. and three modes are compared with the “exact” values in Fig.8. and the percentage errors in the approximate results.4.7 cm. 4. say. respectively. This deficiency also implies that. This is to be expected. The peak values of floor displacements and story drifts including one. ur 2o = 11.2). is presented next.9.

1 and 4.53 ∆r3 (cm) 12 0 • 2. just as for elastic systems. 2.2) if the system yields significantly versus if the system remains within the elastic range (Tables 3. 0. (a) Roof Displacement (b) Top Story Drift 80 12 0 • 3.1 −12 12 ∆r (cm) u (cm) UMRHA 3 "Modes" 0 • 7. and total response from NL-RHA 39 .38 6. Response histories of roof displacement and top-story drift due to 1. 0.0.5 (Tables 4.3 11. 1. the degree to which the system deforms beyond its elastic limit.2) and 1.0.25.62 5. 1.1 and 3. For this purpose the UMRHA and exact analyses were repeated for ground motions of varying intensity. the errors in responses computed by UMRHA including three “modes” relative to the “exact” response were determined.7 • −80 80 ∆r1 (cm) u (cm) "Mode" 1 −12 r2 0 ∆r2 (cm) 12 0 u (cm) −80 80 −12 r3 0 • 2.88 u (cm) "Mode" 3 −80 80 0 −80 80 0 −80 0 • 44.2).0.1 and 4.5 × El Centro ground motion: individual “modal” responses and combined response from UMRHA.25 (Tables 3.2). the errors in story drifts are larger compared to floor displacements. Next. 4.response is much less accurate (Tables 4. and 3. For each excitation.7.5.75. sec 25 30 r −12 0 Fig. However.1 and 3. in particular. recall that the computed errors have been presented earlier for ground motion multipliers 0. 0.85. sec 25 30 • 48. we investigate how the errors in the UMRHA vary with the deformation demands imposed by the ground motion.5. for a fixed number of “modes” included. defined as the El Centro ground motion multiplied by 0.6 5 10 15 20 Time.44 • "Mode" 2 r1 0 • 48.24 • NL−RHA r −12 12 ∆r (cm) u (cm) 0 5 10 15 20 Time.

(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 0.20)] for the first three n modes of the system are shown in Fig. 4. with the peak displacement of each “modal” SDF system noted for each ground motion multiplier. 4. Heightwise variation of floor displacements and story drift ratios from UMRHA and NL-RHA for 1.5 Story Drift Ratio (%) 2 2. two. it will be useful to know the deformation of the system relative to its yield deformation.5 Fig.10 summarizes the error in UMRHA as a function of ground motion intensity.10b). Two versions of the pushover curve are 40 . Heightwise variation of error in floor displacements and story drifts estimated by UMRHA including one.11. in each story drift (Fig.8. 4.5 1 1.9. Shown is the error in each floor displacement (Fig.5 Displacement/Height (%) 2 Ground 0 0. and the error envelope for each case. (3.5 × 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. For this purpose. 4. the pushover curves using force distributions s* [Eq. indicated by a ground motion multiplier.5 1 1. or three “modes” for 1. 4.5 ¥ El Centro ground motion Figure 4. To interpret these results.10a).

4. 3. and 2. drifts in all stories (Table 4. The system remains elastic up to ground motion multiplier 0. The location of plastic hinges and their rotations. the peak values of displacements at all floors (Table 4. respectively.75.0. (2) increase rapidly as the ground motion multiplier increases to 1.4. and 3 (Fig. (3. the pushover curve for each “mode” is idealized by a bilinear curve in solving Eq.20) with n = 1. the ductility factor for the first mode system is only 1.4) to roof displacements urno = 48.5). 4.7 cm. and plastic hinge rotations at the external beam end at each floor level (Table 4.0.75. as mentioned in Section 4. the errors in truncating the higher mode contributions are negligible. For more intense excitations. Figure 4. 41 . were noted but are not shown here. First. (3) maintain roughly similar values for more intense ground motions. 11.2.10) for each “modal” inelastic SDF system (Figs 4.53 cm. 2.included: the actual curve and its idealized bilinear version. 4.2. and.7). the values determined by RHA of the nth-mode inelastic SDF system (Fig. The idealized curve for the first “mode” deviates most from the actual curve near the peak displacement corresponding to ground motion multiplier 1.10 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 0.01 (Fig. Second. even though the system remains essentially elastic. determined from “exact” analyses. Additional errors are introduced in UMRHA of systems responding beyond the linearly elastic limit for at least two reasons. (4.6 and 4.4). Each of these three pushover analyses provides the pushover curve (Fig.11a). and (4) are larger in story drift compared to floor displacements. the first reason mentioned above seems to be the primary source for the errors. The structure is pushed using the force distribution of Eq.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 1.5 ¥ the El Centro ground motion.3). 4.1).3 cm. UMRHA lacks a rigorous theory and is based on several approximations.11). as mentioned in Section 3. Perhaps this explains why the errors are large at this excitation intensity. 4.

018 0.8 0.763 -15.256 1.065 0.201 -0.2 6.914 -0.5 2 GM Multiplier 2.003 0.2 1.2 Peak values of story drift ratios (as % of story height) from UMRHA for 1.4 -7.722 0.727 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 0.293 1.(a) Floor Displacements 60 50 40 Error Envelope Error for Floor No.2 22.472 1.938 1.5 2 GM Multiplier 2.663 -0.9 31.407 -10.8 1.366 0.3 42 .126 0.201 -1.1 3.009 -0.104 0.2 4.379 1.676 0.154 0.3 8.6 4.3 1.248 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.376 -1.982 9.136 1.298 0.663 0.942 -0.371 0.8 1.495 1.540 0.120 1.806 0.241 1.5 3 Fig.220 -0.072 -1.819 2.315 -0.0 9.4 -1.938 -1.806 -0.096 0.071 -0.3 25.9 31.6 2.0 11.0 2.1 8. 4.072 1.698 1.4 4.9 12.291 0.033 0.410 1. Errors in UMRHA as a function of ground motion intensity: (a) floor displacements.707 1.338 -1.8 1.5 9.333 0.055 -0.5 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Displacement /Height (%) “Modal” Response Combined (UMRHA) “Mode” 1 “Mode” 2 “Mode” 3 1 “Mode” 2 “Modes” -0.138 1.376 1.370 -0.8 1.7 31.372 1.5 1 1.200 8.5 28.856 2.473 -22.256 1.372 -1.971 1.4 1.5 3 0 0 0.135 9.226 -0.366 -0.7 14.5 × 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 -1.256 -1.14 m) from UMRHA for 1.478 0.057 -0.220 0. Noted 60 50 40 (b) Story Drifts Error Envelope Error for Story No.088 10.5 1 1.373 -0.616 -0.426 -1.9 5.484 0.668 -23.914 2.003 -31.070 1.350 -0.371 -0.9 16.9 12. and (b) story drifts Table 4.490 -1.820 -0.283 1.1 1.121 -0.202 11.3 0.298 -0.338 1.844 -25.214 0.513 -0.513 0.490 1.820 -19.0 7.5 10.772 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 1.2 1.044 1.945 -37.10.0 1.241 -1.5 9.7 Table 4.5 18.5 -3.010 0.235 -0.169 0.8 14.526 -0.317 0.751 1.8 1.216 1.187 -0.133 1.1 Peak values of floor displacements (as % of building height = 37.1 1.1 0.2 12.3 6.201 1.900 -10.575 -41.130 0.877 0.942 1.5 28.983 1.811 1.071 0.079 0.214 -0.049 -0.068 0.852 1.410 -1.863 1.0 11.863 0.554 1.260 -15.0 -9.430 1.6 1.

The errors are especially large in the hinge plastic rotations estimated by the MPA procedure. two.4).18).4.12 and 4. 4. 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%.14. and three “modes. this error is not especially significant because the hinge plastic rotation is very small. (3.9 and Tables 4.2).4).” two “modes. considering one. 4. Observe that the primary contributor to plastic hinge rotations in the lower stories is the first “mode” and the second “mode” in the upper stories. Obviously.” and nonlinear RHA.13 and Tables 4. MPA failed to identify the plastic hinges at the column bases in Fig. the errors in the modal pushover results are.3 and 4. However.3 through 4. and are within a few percent of the exact values for the upper stories. Pushover analysis seems to be inherently limited in computing accurately hinge plastic rotations. and the errors in these estimates relative to the exact response from nonlinear RHA. For example. significantly smaller than in UMRHA (compare Fig. Fortuitously. but the third “mode” contributions do not seem especially important (Fig. Furthermore. The drifts are underestimated by up to 13% in the lower stories. in general.14.Figures 4. 4.13 and Tables 4. even if three “modes” are included (Fig. for two or three modes included. 4. the hinges identified in beams at the 6th floor were at variance with the “exact” results. the third “mode” does not contribute because this SDF system remains elastic (Fig.5 present estimates of the combined response according to Eq.3 and 4. 4.4 with Tables 4. especially in estimating the story drifts (Fig. As shown in Figs. 4. The locations of plastic hinges shown in Fig. 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.” and three “modes. significant improvement is achieved by including response contributions due to the second “mode”. The first “mode” alone is inadequate. The second “mode” was necessary to identify hinges in the upper stories.12 and Tables 4.3 and 4.5). however.1 and 4. 43 . although the error is recorded as 100% if MPA estimates zero rotation when the nonlinear RHA computes a non-zero value. the results were not always accurate. were determined by four analyses: MPA considering one “mode.13 and Table 4.11c).3 and 4.12 and Tables 4. 4. overestimated by up to 18% in the middle stories. but was more successful when the excitation was more intense (results are not included).13 with Fig. 4.” respectively. 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.

75 0. 2.5 2 3 Actual Idealized 2000 0 0 0.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36.85 0. 4. α = 0.0.13 y by Base Shear (kN) 8000 6000 1 1. “Modal” pushover curves with peak roof displacements identified for 0.0.25 0.25 1 2 4 6 Roof Displacement (cm) 8 10 Fig. 1. 0. α = 0.5 1 0.19 y by 3 2 1.25.75 0.5 0.11. α = 0.14 y by Base Shear (kN) 8000 6000 4000 1.9 cm. V = 5210 kN.85 0.85 Base Shear (kN) 8000 6000 4000 2000 0 0 0.75 Actual Idealized 20 40 60 Roof Displacement (cm) (b) "Mode" 2 Pushover Curve 80 12000 10000 u = 9.5 0. V = 7616 kN. 0.5. V = 4952 kN.75. 1.25 5 10 15 20 Roof Displacement (cm) 25 (c) "Mode" 3 Pushover Curve 12000 10000 u = 4.2 cm.5.6 cm.5 0.5 2 3 4000 2000 0 0 Actual Idealized 0. and 3.0 × El Centro ground motion 44 .

5 Story Drift Ratio (%) 2 2.15b).5 Fig.12 Heightwise variation of floor displacements and story drift ratios from MPA and NL-RHA for 1. whereas MPA contains errors inherent in modal combination rules. 4.3. As mentioned in Section 3. In this case. 4. the MPA procedure is equivalent to the RSA procedure. implying that the modal combination errors contained in these procedures are acceptable.4.5 × El Centro ground motion. UMRHA is essentially exact. as discussed in Sections 3. implying excitations intense enough to cause significant yielding of the structure. an observation with practical implications. and the error envelope for each case.3 for elastic systems (or weak ground motions). 4. Shown is the error in each floor displacement (Fig. implying excitations weak enough to limit the response in the elastic range of the structure.5 1 1.(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. now standard in engineering practice. each story drift (Fig. 4. identified in Section 3.15a).2 and 3.5 1 1. 4. the errors in MPA were fortuitously smaller than in UMRHA (compare Figs. 45 . The errors are only weakly dependent on ground motion intensity (Fig. While various sources of errors in UMRHA. The fact that MPA is able to estimate the response of buildings responding well into the inelastic range to a similar degree of accuracy indicates that this procedure is accurate enough for practical application in building retrofit and design.5 Displacement/Height (%) 2 Ground 0 0.10 and 4.15) for ground multipliers larger than 1.75. However. shading indicates errors in MPA including three “modes” Figure 4.15). indicated by a ground motion multiplier.4 also apply to MPA.4.15 summarizes the error in MPA considering three “modes” as a function of ground motion intensity. the errors in MPA were larger for ground motion multipliers less than 0.0.

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

76E-03 4.2 1.8 -29.503 -1.02E-03 0.266 -0.8 -6.222 0.018 -0.9 -100.00E+00 0.116 1.1 46.6 13.125 -1.473 -15.00E+00 0.667 -1.55E-03 3.667 1.351 -0.233 1.0 2 “Modes” -32.72E-03 7.336 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.118 0.190 -0.752 1.2 0.00E+00 0.745 1.88E-03 0.36E-03 6.311 0.6 7.3 -3.00E+00 0.00E+00 0.00E+00 0.00E+00 0.5 10.00E+00 0.820 -7.298 -0.2 -100.435 0.0 47 .88E-03 1.503 1.9 1.220 1.00E+00 0.0 -100.910 1.2 0.55E-03 3.517 1.0 -50.705 1.371 -0.728 1.733 1.304 -1.3 1.304 1.2 -3.6 -8.2 -4.426 15.99E-03 6.130 0.982 13.0 15.581 0.5 × El Centro ground motion Hinge Plastic Rotation (rad) “Modal” Response Combined (MPA) “Mode” 1 7.02E-03 3.154 0.033 -0.298 0.1 62.8 1.331 1.478 0.756 0.763 -14.00E+00 3 “Modes” 7.88E-03 0.007 1.1 62.8 17.5 × El Centro ground motion Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th “Modal” Response “Mode” 1 “Mode” 2 Drift Ratio (%) Combined (MPA) 1 “Mode” 2 “Modes” 3 “Modes” Error (%) NL 1 2 3 RHA “Mode” “Modes” “Modes” 1.5 10.012 1.399 0.429 0.015 0.53E-03 7.209 1.071 -0.7 -12.00E+00 0.694 1.197 -0.9 7.02E-03 3.156 -0.053 1.407 -27.36E-03 6.705 -1.7 1.10E-02 9.315 0.9 5.980 -0.60E-04 7.055 0.581 0.0 -5.911 0.066 -0.5 2.068 0.9 -100.5 2.3 13.37E-04 Error (%) “Modal” Response 1 “Mode” -32.053 -1.8 -6.00E+00 0.00E+00 0.945 -49.088 12.00E+00 0.514 -1.4 Peak values of story drift ratios (as % of story height) from MPA for 1.36E-03 6.0 -100.641 1.879 1.4 -8.305 -0.220 1.116 1.6 -44.0 3 “Modes” -32.55E-03 3.4 1.003 -16.5 -6.738 1.19E-10 3.22E-10 NL RHA 1.9 2.015 0.72E-03 7.5 Table 4.071 0.8 0.5 7.5 2.76E-03 4.00E+00 1 “Mode” 7.3 Peak values of floor displacements (as % of building height = 37.8 7.18E-03 7.527 -0.594 -1.00E+00 0.2 11.00E+00 0.5 1.259 1.2 6.009 0.498 1.2 -100.26E-04 9.640 1.6 13.101 -0.756 0.37E-03 1.1 1.76E-03 4.00E+00 2 “Modes” 7.0 1.666 Table 4.7 1.76E-03 4.6 -44.50E-10 3.8 -12.9 -4.02E-03 0.687 0.9 1.668 -13.1 46.372 0.200 8.5 Peak values of hinge plastic rotations (radians) from MPA for 1.176 0.0 -100.895 1.250 0.614 0.1 -8.737 1.1 18.00E+00 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th “Mode” 2 0.00E+00 3.222 0.1 62.168 -0.8 -29.8 -4.804 1.049 -0.36E-03 6.202 8.37E-03 1.518 1.018 -0.72E-03 7.942 6.516 0.6 -7.980 0.00E+00 “Mode” 3 0.338 1.3 11.737 0.37E-03 1.399 0.6 1.844 -7.3 -3.5 × 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 0.72E-03 7.05E-03 2.640 -1.900 -0.0 -50.1 13.057 0.244 0.575 -53.781 0.066 -0.6 -9.00E+00 0.1 46.007 1.116 1.260 -14.135 -7.018 0.8 “Mode” 3 -1.683 1.8 -29.Table 4.14 m) from MPA for 1.233 1.00E+00 0.105 0.414 1.2 -100.8 1.37E-03 1.60E-03 2.00E+00 0.781 0.611 0.652 1.9 0.895 1.00E+00 0.

17 and Tables 4. 4. 4. 4. but it is inadequate in estimating the story drifts (Fig.18 and 48 .20) with n = 1 (Fig.16 are unchanged. as are the roof displacements ur 2o = 11. Also presented at these roof displacements are the displacements at all floors (Table 4. and overestimated by up to 29% in the middle stories and by up to 13% in the upper stories.16a—which is slightly different than the one excluding gravity loads (Fig. 4.6). 4. The first “mode” alone provides adequate estimates of floor displacements.8 present estimates of the combined response according to Eq.” but not the contributions of higher “modes. (3. They are now included in the pushover analysis of the structure for the first “mode” only. As shown in Fig. the structure is pushed using the force distribution of Eq.18 and Tables 4.18) considering one.4.7). The pushover curves for the second and third modes included in Fig.20) with n = 1 to a target roof displacement. and three “modes. MPA including two “modes” underestimates the displacements of lower floors by up to 6% and overestimates the upper floor displacements by up to 22%. resulting in the pushover curve shown in Fig.3 .7 cm and ur 3o = 2. Static analysis of the structure for gravity loads provides the initial state—forces and deformations—of the structure.6 through 4. Starting with its initial state under gravity loads.7).18 and Tables 4.53 cm. (3. two. 3.3 Modal Pushover Analysis with Gravity Loads To evaluate the accuracy of the dynamic response of the system. drifts in all stories (Table 4.6 and 4. the results presented so far did not include gravity load effects. The response contributions of the second and third modes are the same as before (Tables 4.7.17 and Tables 4. The story drifts are underestimated by up to 7% in the lower stories.4. The errors are especially large in the plastic hinge rotations estimated by the MPA procedure even if three “modes” are included (Fig. 4. 4.17 and 4.8).6 and 4.4. Figures 4. Results of such analyses for the selected building subject to 1.” these effects will modify the combined modal response as well as the results of nonlinear RHA.11a).6 and 4.” and the errors in these estimates relative to the exact response from nonlinear RHA. and the structure is pushed using the force distribution of Eq.0 cm.4) to roof displacement ur1o = 52. Obviously gravity load effects will influence the seismic demands due to the first “mode. and plastic hinge rotations at selected external beam end at each floor level (Table 4.” however.7). (3.5).5 × El Centro ground motion are presented below. the third “mode” contributions do not seem especially important (Fig. Significant improvement is achieved by including response contributions due to the second “mode.

and (b) story drifts 49 .Table 4.5 3 Fig.5 3 0.5 1 1.8). (a) MPA.14. Errors in MPA as a function of ground motion intensity: (a) floor displacements. Noted Error (%) 30 20 10 0 0 2 8 3 4 1 9 Error (%) 30 20 5 4 6 5 7 2 9 10 0 0 1 8 7 3 6 0. 2−"Modes" •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• • •• (c) 3−"Modes" • • • • • • • •• •• • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• • •• • • • • • • • • • (d) Nonlinear RHA •• •• • •• •• •• •• • •• •• •• •• •• •• •• • •• •• • •• •• •• •• • • • • • • •• •• •• Fig. Noted 60 50 40 (b) Story Drifts Error Envelope Error for Story No.5 × El Centro ground motion (a) Floor Displacements 60 50 40 Error Envelope Error for Floor No. 1−"Mode" • • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • •• • •• • • • • • •• •• • •• •• •• •• (b) MPA. 4.15.5 2 GM Multiplier 2. and three “modes” and by NL-RHA for 1. Most pushover analysis procedures do not seem to compute to acceptable accuracy plastic hinge rotations. 4. two.5 1 1. Locations of plastic hinges determined by MPA considering one.5 2 GM Multiplier 2.

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

V = 5210 kN. V = 4952 kN.50.25 1 2 4 6 Roof Displacement (cm) 8 10 Fig. 0. “Modal” pushover curves with gravity loads included.19 y by 3 2 1. 4.0.3 cm. and 3. 0.5 1 0.75 0.5 0. V = 7433 kN.0.5 2 3 Actual Idealized 2000 0 0 0. 2.16.14 y by Base Shear (kN) 8000 6000 4000 1.25 5 10 15 20 Roof Displacement (cm) 25 (c) "Mode" 3 Pushover Curve 12000 10000 u = 4.25 0.85 Base Shear (kN) 8000 6000 4000 2000 0 0 0.13 y by Base Shear (kN) 8000 6000 1 1.75. α = 0.85 0.85.9 cm.85 0.75 0.5 2 3 4000 2000 0 0 Actual Idealized 0.0 × El Centro ground motion 51 . α = 0.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36. noted are peak values of roof displacement for 0. α = 0. 1.25. 0.6 cm.75 Actual Idealized 20 40 60 Roof Displacement (cm) 80 (b) "Mode" 2 Pushover Curve 12000 10000 u = 9.5 0.5 0.

two.(a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th NL−RHA MPA 1 "Mode" 2 "Modes" 3 "Modes" (b) Story Drifts Floor Floor 5th 4th 3rd 2nd 1st 5th 4th 3rd 2nd 1st NL−RHA MPA 1 "Mode" 2 "Modes" 3 "Modes" Ground 0 0.5 Fig.17.18. and 52 . 4. and hinge plastic rotations estimated by MPA including one.5 × El Centro ground motion. 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. Errors in floor displacements.5 1 1. gravity loads included. Heightwise variation of floor displacements and story drift ratios from MPA and NL-RHA for 1.5 Displacement/Height (%) 2 Ground 0 0.5 Story Drift Ratio (%) 2 2. story drifts.5 1 1. 4.

Errors in MPA as a function of ground motion intensity: (a) floor displacements.5 2 GM Multiplier 2.20.5 1 1. 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. gravity loads included 53 .19. two. Locations of plastic hinges determined by MPA considering one. gravity loads included (a) Floor Displacements 60 50 40 Error Envelope Error for Floor No. 3−"Modes" (d) Nonlinear RHA • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• •• •• • • • • • • • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • • • • • •• •• •• Fig. Noted 60 50 40 (a) Story Drifts Error Envelope Error for Story No.5 × El Centro ground motion. 2−"Modes" •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• •• •• (c) MPA.5 3 0. and (b) story drifts.(a) MPA. 4. 1−"Mode" • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • •• •• •• • • • • • •• •• •• •• •• •• •• (b) MPA.5 1 1. and three “modes” and by NL-RHA for 1.5 2 GM Multiplier 2. 4.5 3 Fig.

13E-03 5.983 1.018 -0. gravity loads included Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Displacement /Height (%) “Modal” Response Combined (MPA) “Mode” 1 “Mode” 2 “Mode” 3 1 “Mode” 2 “Modes” 0.998 0.2 1.00E-03 5.35E-03 8.00E+00 “Mode” 2 0.831 0.603 -1.5 -5.168 -0.00E+00 0.998 1.0 -30.1 21.399 0.371 -0.2 9.0 37.2 -0.351 -0.530 1.0 2 “Modes” -32.23E-02 1.8 -32.821 -1.3 -100.00E+00 0.19E-03 1.78E-03 1.1 1.4 26.23E-03 0.55E-03 3.8 -32.0 16.434 0.5 54 .877 -46.783 1.414 28.071 0.2 12.11E-03 9.850 -1.527 -0.17E-03 9.996 0.821 1.0 -100.5 21.071 -0.5 × El Centro ground motion.687 0.5 1.00E+00 0.88E-03 1.4 -4.850 1.4 -6.434 0.5 1.938 1.257 0.114 -1.0 -30.1 13.319 1.057 0.00E-03 5.19E-03 1.213 1.6 Peak values of floor displacements (as % of building height = 37.908 -1.2 1.213 1.813 1.7 16.830 -12.197 -0.11E-03 9.728 1.037 0.00E+00 0.515 -50.14 m) from MPA for 1.3 -22.00E+00 0.00E-03 5.263 0.128 -1.5 3 “Modes” -32.55E-03 3.75E-03 0.667 0.114 1.207 18.105 0.8 1.330 1.5 21.454 1.101 -0.0 37.514 -1.4 1.049 -0.190 -0.4 -4.1 4.00E+00 “Mode” 3 0.530 1.3 1.066 -0.9 -3.5 × El Centro ground motion.033 -0.55E-03 3.436 1.6 0.8 9.7 Peak values of story drift ratios (as % of story height) from MPA for 1.6 1.8 Peak values of hinge plastic rotations (radians) from MPA for 1.04E-02 8.854 0.921 1.35E-03 8.319 1.064 -10.00E+00 0.637 0.7 19.00E+00 0.154 0.11E-03 9.4 20.055 0.237 0.2 4.00E+00 0. gravity loads included Hinge Plastic Rotation (rad) Floor “Modal” Response “Mode” 1 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 8.00E+00 0.8 28.107 1.5 “Mode” 3 -1.860 1.466 0.037 -0.490 -11.983 1.19E-03 1.35E-03 8.23E-03 3.015 0.858 2.2 16.270 -12.461 0.2 -3.603 1.5 × El Centro ground motion.888 1.831 0.429 -1.00E+00 0.6 2.00E-10 NL RHA 1.822 1.125 -1.305 -0.516 0.00E+00 0.8 Table 4.6 2.372 0.478 0.4 1.214 0.744 1.2 21.04E-10 3.00E+00 0.35E-10 3.176 0.836 -0.933 1.00E+00 0.7 -2.3 -22.068 0.3 13.00E+00 0.Table 4.88E-03 0.00E+00 3 “Modes” 8.23E-03 0.315 0.754 1.854 0.434 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.8 0.686 -7.2 0.00E+00 0.9 -4.156 -0.00E+00 Error (%) “Modal” Response 1 “Mode” -32.88E-03 0.310 1.673 Table 4.098 20.3 -100.00E+00 0.311 0.19E-03 1.2 9.7 -2.2 4.2 -2.530 1.109 0.266 -0.2 4.00E+00 0.3 -100.213 1.3 9.7 1.330 1.009 0. gravity loads included Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th “Modal” Response “Mode” 1 “Mode” 2 Drift Ratio (%) Combined (MPA) 1 “Mode” 2 “Modes” 3 “Modes” Error (%) NL 1 2 3 RHA “Mode” “Modes” “Modes” 1.11E-03 9.26E-03 3.00E+00 0.2 2.3 -22.754 0.00E+00 3.066 -0.913 7.996 -0.998 21.00E-03 5.199 16.2 9.665 0.00E+00 2 “Mode” 8.23E-03 3.19E-04 5.399 -0.00E+00 0.594 -1.00E+00 Combined (MPA) 1 “Mode” 8.130 0.237 0.908 1.7 4.429 1.507 1.102 1.5 10.353 -23.5 0.0 37.102 1.263 1.927 1.637 0.0 -100.6 19.9 31.9 -6.953 15.8 0.35E-03 8.

joint rotations. 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. k = 2 for T1 ≥ 2. story drifts. computed at the target displacement represent the earthquakeinduced demands on the structure. assumed to be linearly elastic. j 2.2 COMPARATIVE EVALUATION Compared in this section are the earthquake-induced demands for the selected building determined by five analyses: pushover analysis using the three force distributions in FEMA-273.5 sec .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. The floor displacements. 5. Equivalent lateral force (ELF) distribution: s* = m j h k where h j is the height of the j j jth floor above the base. 1997] The pushover curve.. 2… N ).5.” and nonlinear RHA. and varies linearly in between. and the exponent k = 1 for fundament period T1 £ 0. Comparison of Modal and FEMA Pushover Analyses 5. MPA considering three “modes. “Uniform” distribution: s* = m j (where the floor number j = 1.5 sec . plastic hinge rotations. gravity load effects were included in all 55 . a plot of base shear versus roof displacement. SRSS distribution: s* is defined by the lateral forces back-calculated from the story shears determined by response spectrum analysis of the structure. Specified in FEMA-273 are three distributions for lateral forces: 1. etc. and 3.

and the displacements of upper floors are overestimated by 9-22%. and the locations of all plastic hinges in Fig. The pushover curves are given in Fig.0 cm. the value determined from RHA of the first-mode inelastic SDF system for 1.00719 0.1. plastic hinge rotation demands in Table 5. Using each of these force distributions.11 0. 5.11 0.11 0. Force distributions in FEMA-273: (a) “uniform”.3b and Table 5. and (c) SRSS 56 . (b) ELF.0466 0. Figures 5.2.0197 0. The MPA procedure is more accurate than all the FEMA force distributions. 5.112 0.062 0. wherein the first two are obvious and the third is determined from response spectrum analysis of the building (Appendix C). 0.4a. The errors in the FEMA and MPA estimates of seismic demands relative to the “exact” demands are presented in Fig.165 0.11 0.21 0.367 (a) Uniform (b) ELF (c) SRSS Fig.281 0. 5.4. both presented in Section 4. 5.177 0. with the errors being larger for the SRSS distribution. The “uniform” distribution overestimates all floor displacements by 17-28%.4 and Tables 5.0446 0.4.1 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.1 through 5.3. the story drift demands in Fig.3.11 0.11 0. The first four floor displacements are within 5% of the “exact” value. 5.119 0. 5.042 0. and Table 5.3.126 0.0702 0.0896 0.0381 0. 5.0654 0. pushover analyses are implemented for a target roof displacement of 52.3a and 5.11 0.1. the floor displacement demands in Fig.0981 0.3a and Table 5. The three FEMA force distributions are presented in Fig. Also included in these presentations are the MPA results considering three “modes” and the “exact” demands from nonlinear RHA.5 times the El Centro ground motion.1.analyses.2.0913 0.

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. gravity loads are included 57 .14 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized 10 20 30 40 Roof Displacement (cm) (b) ELF Distribution 50 60 12000 10000 u = 38.24 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized 10 20 30 40 Roof Displacement (cm) 50 60 Fig. V = 6897 kN.2.3 cm. Pushover curves using three force distributions in FEMA-273: (a) “uniform”. 5. V = 8530 kN. α = 0. α = 0. (b) ELF.8 cm. V = 7456 kN. and (c) SRSS.(a) Uniform Distribution 12000 10000 u = 33 cm. α = 0.

This seems to be an inherent limitation of pushover analysis. Heightwise variation of floor displacements and story drift ratios estimated using FEMA-273 force distributions.3 demonstrate that the hinge plastic rotations estimated by all three FEMA force distributions contain unacceptably large errors. 3 "Modes" FEMA Uniform ELF SRSS 0. reaching 64%.” and NLRHA.2 demonstrate that the story drift demands are greatly underestimated by all the FEMA force distributions.3. errors are largest in the upper stories. (a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th (b) Story Drift Ratios Floor 4th 3rd 2nd 1st NL−RHA MPA.5 Displacement/Height (%) 2 Floor 5th 5th 4th 3rd 2nd 1st NL−RHA MPA. 5.5 Story Drift Ratio (%) 2 2. Figure 5. but it is still inaccurate. 7%.4c and Table 5.5 1 1. For the SRSS distribution.5 1 1.5 Ground 0 Ground 0 Fig. In contrast. MPA including three “modes. the errors are largest in the lower stories. with story drifts under estimated by. reaching 35%. the MPA procedure is more accurate than all the FEMA force distributions. Modal pushover analysis procedure gives estimates better than all the FEMA force distributions. such as hinge plastic rotations. reaching 31%. gravity loads included 58 .Figures 5. whereas nonlinear RHA computed an insignificantly small value.3b and 5.) The pushover analysis procedures considered seem incapable of computing accurately local response quantities. with errors reaching 37% in this example. For the ELF distribution. and Table 5.4b. (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%. 3 "Modes" FEMA Uniform ELF SRSS 0. at most. For the uniform distribution. larger errors are noted in the upper and lower stories.

the “uniform” distribution fails to identify yielding of the beams above the fourth floor. 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. story drifts. and the ELF distribution fails 59 . While pushover estimates for floor displacements are more accurate. The locations of the plastic hinges are not identified correctly by the FEMA force distributions.(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. 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.” nonlinear RHA (“exact”).4. and hinge plastic rotations estimated using FEMA-273 force distributions and MPA (including three “modes”). they are not good indicators of damage. The locations of plastic hinges shown in Fig. the SRSS distribution fails to identify yielding of beams in the middle floors. 5. 5. and the three FEMA analyses. Errors in floor displacements.5 were determined by five analyses: MPA considering the three “modes.

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

5 -3.5 61 .623 1.673 Error (%) NL RHA 1.294 1.2 13.2 16.938 1.809 0.7 7.2 6.58E-04 6.3 Peak values of hinge plastic rotations (radians) from FEMA force distributions and MPA Hinge Plastic Rotation (rad) MPA FEMA ELF SRSS 4.35E-03 8.927 1.860 1.8 44.9 16.00E+00 0.562 1.984 1.736 0.16E-03 0.2 16.109 1.490 0.4 -23.061 1.0 -57.335 2.2 9.318 2.2 19.4 -50.3 -100.672 1.19E-04 5.0 -100. Peak values of floor displacements (as % of building height = 37.789 0.84 1.178 1.5 Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 2.6 -6.234 1.724 0.686 0.00E+00 0.00E-10 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 1.0 -100.0 -100.7 FEMA ELF SRSS -22.00E+00 0.6 14.344 0.93E-03 1.353 1.2 10.00E-03 5.372 1.00E+00 Error (%) FEMA Uniform ELF SRSS MPA 24.7 MPA -2.1 0.6 21.6 22.854 0.8 -35.1 -26.52E-04 1.8 -100.597 0.5 -33.351 0.45E-03 3.5 -32.51E-03 4.4 MPA -2.524 0.5 -33.888 0.10E-02 7.0 -71.323 1.3 -4.7 Table 5.3 29.875 0.4 0.8 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 0.2 4.00E+00 4.6 -4.65E-03 7.Table 5.00E+00 0.53E-02 1.26E-03 3.083 1.466 0.2 -4.6 -25.417 1.9 SRSS -22.8 -63.314 1.128 1.6 17.168 1.88E-03 1.730 1.50E-03 0.7 -28. gravity loads included Displacement /Height (%) FEMA ELF SRSS MPA 0.94E-03 2.2 -100.9 -77.462 1.263 0.011 1.78E-03 0.270 0.089 1.04E-02 8.5 -29.23E-02 1.35E-10 3.367 1.098 1.207 1.0 -30.11E-03 9.783 1.00E+00 2.998 1.560 1.913 0.262 1.8 -2.0 Table 5.55E-03 3.6 -73.708 0.530 1.17E-03 9.8 -32.7 -60.9 15.0 37.436 Error (%) NL Uniform RHA 0.8 7.6 -41.830 2.4 16.00E+00 8.9 28.399 -27.341 1.858 1.4 21.4 -55.78E-03 1.0 -59.975 1.4 16.221 1.5 15.330 1.6 -17.067 0.9 0.0 -63.5 22.214 1.4 -12.310 1.2 5.13E-03 5.7 19.5 -29.3 -11.306 1.00E+00 NL RHA 1.7 26.007 1.195 0.209 0.00E+00 0.414 1.154 1.015 0.1 10.199 27.75E-03 0.399 1.2 -12.877 1.1.34E-03 2.530 1.487 0.6 4.064 1.7 -15.0 -100.7 21.5 26.5 10.6 -11.566 1.2 9.611 0.9 16.8 16.19E-03 1.3 -22.953 0.09E-03 4.399 1.23E-03 3.1 -3.8 2.355 0.9 11.59E-03 5.547 -27.2 Peak values of story drift ratios (as % of story height) from FEMA force distributions and MPA.00E+00 0.8 -100.3 -14.836 0.3 23.1 163.9 -20.998 1. gravity loads included Displacement /Height (%) FEMA ELF SRSS MPA 1.5 -27.62E-03 0.14 m) from FEMA force distributions and MPA.3 31.515 FEMA Uniform ELF 27.7 -29.839 0.667 0.9 -70.8 17.992 1.03E-03 5.

5 × El Centro ground motion. gravity loads included 62 . MPA including three “modes” and NL-RHA for 1.5. Locations of plastic hinges determined from three force distributions in FEMA-273.(a) FEMA: Uniform • • • • • • • • •• •• •• •• •• •• (b) FEMA: ELF • •• •• •• •• •• •• • •• • •• •• •• •• •• • • • • • •• • • • • • • • •• •• •• •• • •• •• •• •• • •• •• •• •• • • •• •• •• • • • • • •• •• (c) FEMA: SRSS • • •• •• •• •• • • •• •• •• • • •• •• •• • • • • •• •• •• • • • • • • • • • (d) MPA with 3 "Modes" • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• •• •• (e) Nonlinear RHA • • • • • • • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • • • • • •• •• •• Fig. 5.

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

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

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

(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 . p eff (t ) = . 5. (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. The response value rno is determined by implementing the following steps: (i) Develop the base-shearroof-displacement Vbn .n t = . the SRSS rule). n where φ n is now the nth natural mode for small-amplitude linear vibration. To enable systematic extension of the MPA procedure to inelastic systems. p eff . an uncoupled modal response history analysis (UMRHA) is developed by (1) neglecting the coupling among modal coordinates arising from yielding of the system.. 6.n t is determined by pushover analysis.g. 2. …) according to an appropriate combination rule (e. 66 .mι ug t .sn ug t —n denotes mode number—in the modal expansion of the effective earthquake forces. These underlying assumptions and approximations are evaluated and the errors in the UMRHA procedure relative to the “exact” nonlinear response history analysis are documented. and (2) superposing responses of the inelastic MDF system to individual terms. 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 . and (ii) the total response ro is determined by combining the rno (n = 1.4. with vibration properties in the linear range same as those of the nth-mode elastic SDF system.

The initial state—forces and deformations—of the structure can be considered in the MPA procedure by including these gravity load effects in pushover analysis of the structure only for the first “mode. However. the errors in the MPA procedure are shown to be only weakly dependent on ground motion intensity. such as hinge plastic rotations. Based on results presented for El Centro ground motion scaled by factors varying from 0. all pushover analysis procedures do not seem to compute with acceptable accuracy local response quantities. demonstrates that the FEMA force distributions greatly underestimate the story drift demands. the pushover analysis provides the peak value rno of any response quantity: floor displacements. Pushover analyses seem to be inherently limited in computing accurately hinge plastic rotations. and nonlinear RHA. 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. etc. 9. 7. 8.25 to 3. story drifts. 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. Thus the MPA procedure is accurate enough for practical application in building evaluation and design.0. 67 . MPA. joint rotations. and lead to unacceptably large errors in the hinge plastic rotations. plastic hinge rotations. story drifts. and identifying locations of most plastic hinges. and hinge plastic rotations. Comparing the earthquake-induced demands for the selected 9-story building determined by pushover analysis using three force distributions in FEMA-273.” 10. it fails to compute with acceptable accuracy plastic rotations of the hinges.(iv) At the roof displacement determined from Dn . The MPA procedure is more accurate than all the FEMA force distributions in estimating floor displacements.

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

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

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

Since the target roof displacement may not be known at the start of the procedure.20): s* = mφn n n 2. of the bilinear curve at the target roof displacement. for linear-elastic vibration of the building.1) using the FEMA-273 procedure (Building Seismic Safety Council. develop the base-shear – roof-displacement (Vbn . 71 .2.1 incrementally and record the base shears and associated roof displacements. B.1. Calculate the area under the actual pushover curve. 1977). 1988). A. This step can be conveniently implemented in any commercially available software..g. respectively.1. trapezoidal rule. (3. 3. For the nth-“mode”. e.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.Appendix A Uncoupled Modal Response History Analysis A. Idealize the pushover curve as a bilinear curve (Fig. The structure should be pushed just beyond the target (or expected) roof displacement in the selected mode. φn . 3. DRAIN2DX (Allahabadi and Powell. Define the force distribution s* from Eq. Apply force distribution of Step 2. 3. iterations may be necessary. 2.. and modes. e. Let the roof displacement and base shear at the anchor point be urno and Vbno .g. 1.2.urn ) pushover curve for the force distribution s* : n 2. ω n . Define the anchor point. Compute natural frequencies. Apn . using any numerical integration method.

8. i 3. Calculate area under the bilinear curve OAB. A.4. Calculate initial slope of the idealized bilinear curve. Vbny . Calculate Vbny = Vbny × Apn Abn . corresponding to the estimated yield i i i base shear.9. Let the point with base shear = Vbny and roof displacement = urny be denoted as A.6 . This step gives the secant stiffness at a base shear equal to 60% of the yield base shear.1. Vbny .6 × Vbny . Calculate the error = 100 × Abn − Apn ( ) ( ) ( ) ) Apn .7. determine the roof displacement. If the error exceeds some pre-specified tolerance. and B with straight-line segments to obtain the idealized bilinear curve.i 3. kn = 0. obtained by judgment.3. i 3. will be refined by an iterative procedure that seeks to equate areas under the actual and the idealized pushover curves. 3.9. by connecting a straight line between origin. i 3.2. urny = Vbny kn . Draw the curve OAB by connecting the three points O.6 .0. Replace i+1 with i and repeat Steps 3. i i 3.6 × Vbny . kn . If desired.2. From the pushover data. 3. at base shear i equal to 0. 72 . Calculate the yield displacement. i +1 i i 3.6 × Vbny ( ) i urn . Abn . i i i αn = Vbno Vbny − 1 urno urny − 1 i 3. other appropriate methods can be ( used. This value. Estimate the yield base shear.4 to 3.1. urn.8.5. and a point on the actual pushover curve with base shear equal to i 0.6.4. Calculate the post-yielding strain-hardening ratio. i i i 3. iterations are necessary. O.9.4.0. 3. Calculate the slope.

12) and (3.2). (3.10a and 4. A. 4. Scale the vertical axis with * Mn to obtain * Fsno Ln = Vbno M n and * Fsny Ln = Vbny M n (Eqs. of the combined responses obtained in Step 8. Repeat Steps 3 to 6 for as many modes as required for sufficient accuracy. Combine the “modal” responses using Eqs. 7. A.6 ur n y ur n o ur n Fig.0.11b). 4.3b) with unit-mass and force-deformation relation of Fig.11a). 5.15) and (3. 4. Develop the Fsn Ln − Dn relation (Fig. Compute the Ln and Γ n from Eq. 8. (3.3. r o .2. Scale the horizontal axis by Γnφrn to obtain Dno = urno Γ nφrn and Dny = urny Γ n φrn (Eqs. An (t ) . 4. Vbn Idealization of Pushover Curve Vbno Vbny 0. A. 6.1.4) and effective modal mass from M n = Ln Γn .13). Calculate peak values.16).6 × Vbny A • Idealized αnkn B • 1 Actual k 1 • O n figA_1. (3. Calculate histories of various responses using Eqs.1 Idealization of nth-“mode” pushover curve 73 .2. 4.10b and 4. 9. of the nth“mode” inelastic SDF system (Fig.4. and pseudo-acceleration history. Dn (t ) . In general first two or three modes will suffice. * 4.eps ur n. Compute deformation history.

are shown in Fig.1.7 cm (5 in. Idealized bilinear curves for each of the three modes are included in Fig. The following steps illustrate the procedure to develop the idealize curve for the first “mode”.2 EXAMPLE The UMRHA procedure is implemented to calculate the response of the 9-story building described in Section 3. ur1o = 63.1. Following is step-by-step implementation of the procedure described in Section A.Fsn / Ln Fsn / Ln − Dn Relationship Vbno / M* n Vbny / Mn * 1 α ω2 n n ω2 1 n figA_2. s* . generated using DRAIN-2DX. 2.3. B.5 cm (25 in. 3. n 2.4. A.6 kN.4.3.2. is defined at the target roof displacement.5 cm and Vb1o = 8729. second. computed for the first three modes are shown in Fig.2. 3. respectively. The anchor point. Ap1 = 360777 kN-cm. 3.4 cm (10 in.3. The pushover curves for the first three modes. A. Target displacements used to generate these pushover curves are 63. A. The base-shear – roof-displacement (Vbn .). for the first. 1.1.).urn ) pushover curve for the force distribution s* : n 2. and 12. The force distributions. 25. 3.2 Properties of the nth-“mode” inelastic SDF system A. to the north-south component of the El Centro (1940) ground motion scaled up by a factor of 1. 74 . At this point. and third mode. First three mode shapes and frequencies of the selected building were computed and are shown in Fig.eps Dny = ur n y / Γnφr n Dno = ur n o / Γnφr n Dn Fig. Area under the actual pushover curve.).5. 3.

k1 = 0. (4.09 cm.09 ) − 1 = 0. Therefore.6. Vb1 y = 7615.i 3. k1 .4.18 = 38. and α1 = 0.9. 3. The point A on the i i bilinear curve is defined by ur1 y = 38. Error = 100 × (365100 − 360777 ) 360777 = 1. A.2. The initial slope of the idealized bilinear curve. Also included are the modal damping ratios and the periods calculated from Eq. The curve OAB obtained by connecting the three points O. i i i 3.1.23 cm. ur1 y = Vb1 y k1 = 8006. This value exceeds the prespecified tolerance of 0.6 × Vb1 y ( ) i ur1.2. i i i 3.4 210.6 = 22.4.1. 3. The procedure converged after nineteen cycles to give ur1 y = 36. Determined from the pushover database.6 kN. i i 3.198%. and M1 = 2736789 × 1.3.9. i 3. α1 = Vb1o Vb1 y − 1 ur1o ur1 y − 1 = ( ) ( ) (8729.18 kN/cm.8. 4. The first estimate of the yield base shear Vb1 y = 8006.7.2. The Fs1 L1 − D1 relation for the first “mode” is developed as follows.0.1.6 × Vb1 y = 4803.3666 = 3740189 kg. The yield displacement.6 8006.8 kN. * 4. The results for other modes are summarized in Table A.4 kN. Area under the bilinear curve OAB. L1 = 2736789 kg. ur1.09 cm and Vb1 y = 8006.4. Ab1 = 365100 kN-cm.4 kN.86 = 210.8 22.0. iterations are necessary. and B with straight-line segments gives the idealized bilinear curve.5.01%. Γ1 = 1.3666.9. i +1 3. The results of the iterative procedure are summarized in Table A.194.135. is calculated as follows.86 cm at 0.4 × (360777 365100 ) = 7911. The post-yielding strain-hardening ratio. i 3. 75 . 3.5 38.4 ) − 1 (63. The next estimate of the yield shear is Vbny = 8006. i i 3.1.6 = 4803.13).9 kN.

8.4. Histories of roof displacement and top story drifts for the first “mode” are computed and presented in Fig. Scaling the vertical axis by Fs1 y L1 = 203. 4. 4. 7. 4. A.51 cm. 8.2. 4.40 (cm/sec2) and 76 .7. 5.4. 9. 6. The peak values are computed and are summarized in Tables 4.62 (cm/sec2). 4.7. Scaling the horizontal axis by Γ1φr1 gives D1o = 46.1 and 4. The combined modal responses are presented in Fig.7.2.46 cm and D1y = 26. * M1 gives Fs1o L1 = 233.3. The results were generated for first three “modes’ and are included in Fig. The peak values are also plotted 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.

“Modal” pushover curves for the example building 77 .13 6000 Base Shear (kN) 4000 2000 Actual Idealized figA_3b. V = 5210 kN.eps 0 0 2 4 6 8 10 Roof Displacement (cm) 12 Fig.19 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized figA_3a.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36.eps 0 0 5 10 15 20 Roof Displacement (cm) (c) "Mode" 3 Pushover Curve 25 8000 u = 4. α = 0. V = 7616 kN.2 cm.14 y by 6000 Base Shear (kN) 4000 2000 Actual Idealized figA_3c. α = 0.3.eps 10 20 30 40 50 Roof Displacement (cm) (b) "Mode" 2 Pushover Curve 60 8000 u = 9. α = 0.6 cm. A.9 cm. V y by = 4952 kN.

78 . A.4.747 5 10 15 20 Time (sec) figa_4.(a) Deformation 80 n=1 3 (b) Pseuso−acceleration n=1 • 0.223 −80 80 n=3 −3 3 n=3 D3 (cm) 0 A3 (g) • 10.33 A1 (g) n=2 0 −80 80 −3 3 n=2 D2 (cm) 0 A1 (g) • 22. Histories of deformation and pseudo-acceleration due to 1.2769 D1 (cm) 0 • 35.52 5 10 15 20 Time (sec) 25 30 0 • 1.m −80 0 −3 0 25 30 Fig.5 × El Centro ground motion for the first “mode.” and third “mode” inelastic SDF systems.” second “mode.06 0 • 1.

037 0.35 36.8 4647.170 0.693 0.18 210.4927 1.18 210.79 0.3 7786.18 210.32 36.18 210.910 0.6 i k1 i ur1 y i α1 i Ab1 Error (%) 1.5 7624.74 (kN/cm) 210.194 (kN-cm) 365100 364060 363276 362684 362235 361892 361631 361432 361279 361162 361073 361004 360951 360911 360880 360856 360838 360824 360813 Table A.09 18.8525 1.28 36.9 7615.9 4573.05 52.18 210.1 4574.18 210.95 21.193 0.18 (cm) 38.18 210.8 7618.1 7616.18 210.5 (cm) 22.188 0.38 22.948 ζ n (%) 79 .176 0.3 7745.103 “Mode” 3 696400 0.2 4571.4 7714.09 37.4 7911.59 22.85 0.56 19.6 × Vb1 y i ur1.40 46.18 210.162 0.5309 488839.59 36.86 21.23 0.9 4570.18 210.151 0.192 0.3666 3740189 203.3 7658.191 0.75 21.02 21.85 36.44 36.12 3876.90 21.5 3109.30 37.2 7690.76 21.063 0.62 26.013 0.2406 167531.180 0.64 37.9 (kN) 4803.0 4588.2 4628.3 7628.4 4595.0 4577.18 210.193 0.048 0.184 0.3 4603.05 36.5 4614.186 0.029 0. No.010 (kN) 8006.Table A.75 21.50 36.194 0.26 36.39 36. Properties of “modal” inelastic SDF systems Properties Ln (kg) Γn * M n (kg) “Mode” 1 2736789 1.18 210.2.190 0.65 1226.194 0.11 22.4 7647.5 7633.78 21.7 4580.18 210.2671 1.017 0.18 210.136 Fsny Ln (cm/sec2) Dny (cm) Fsno Ln (cm/sec ) Dno (cm) Tn (sec) 233.6 7840.83 21.79 21.082 0.8 7622.18 210.23 22.193 0.1 4569.0.51 2 “Mode” 2 -920860 -0.404 0.18 210.25 36.9 4570.77 21.198 0.24 36.56 47.309 0.182 0.8 4747.529 0.139 0.18 210.022 0.18 210.25 36.70 36.1 1013.4 7672.86 22.1. i 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 i Vb1 y i 0.29 36.237 0.46 2.2 4671.135 0.7 7639.194 0.0 7619. Results of iterative procedure to develop the idealized bilinear curve for the first “mode” inelastic SDF system Itr.107 0.74 21.6 4583.81 21.0 4704.

80 .

Solving Eq.Appendix B Modal Pushover Analysis B. (4. 4.18). At urno . 10. 10. Typically. B. Repeat Steps 3 to 7 for as many “modes” as required for sufficient accuracy.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.1.4. or from the inelastic response (or design) spectrum.33 cm. Determine the total response by combining the peak “modal” responses using the SRSS combination rule of Eq. Steps 1 to 4 of the MPA are the same as those for UMRHA presented in Appendix A.3b) with unit mass and force-deformation relation of Fig. 13. the first two or three “modes” will suffice. Dn .1 to the north-south component of the El Centro (1940) ground motion scaled up by a factor of 1. (3. of the nth-“mode” inelastic SDF system (Fig.6b by solving Eq. rno . 11. This is the same example as solved in Appendix A. 6. Calculate the peak roof displacement urno associated with the nth-“mode” inelastic SDF system from Eq.8) for the peak deformation of the first-“mode” inelastic SDF system with unit mass and force-deformation relation developed in Step 4 gives D1 = 35. (3. Compute the peak deformation. Following is step-by-step implementation of the procedure described in Section B. 4. extract from the pushover database values of other desired responses. 81 .21). subtract the yield hinge rotation to determine the plastic hinge rotation.5.8). From the total hinge rotation. 12. (4.2 EXAMPLE The MPA procedure is implemented to calculate the response of the 9-story building described in Section 3.

Also included in Table 4. values of floor displacements and story drifts are extracted.4.3 and 4.18) are also included in Tables 4. 82 .33 = 48.5 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. where results for other ground motion intensities are also included. Peak roof displacement ur1o = Γ1 × φr1 × D1 = 1. At ur1o = 48. The values are summarized in Table 4. 9. (3.366 × 1 × 35.28 cm.3 and 4. 8. Results of Steps 5 and 6 are summarized in Table B. 12.4 for the story drifts normalized by the story height.11.28 cm. The total response computed by combining the peak “modal” responses using the SRSS combination rule of Eq. Steps 3 to 7 are repeated for first three “modes.” and the results are included in Tables 4.1.4.

25 20.225 2.268 0.467 14.535 14.023 0.379 21.367 1. Calculation of roof displacements urno from peak deformation of inelastic SDF systems Ground Motion Multiplier 0.755 0.139 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.50 35.79 0.73 24.36 1.690 10.332 13.5 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.185 11.436 7.901 8.0 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 3.03 0.38 22.229 8.252 9.007 36.856 31.969 0.577 16.06 1.85 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 1.395 0.748 63.37 57.07 “Mode” 2 4.25 Quantity Dn (cm) “Mode” 1 6.551 2.8451 5.154 78.117 5.35 1.Table B.13 2.312 1.913 22.200 0.126 13.35 0.267 5.755 27.59 0.504 18.33 1.0 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 1.05 0.735 3.275 1.676 6.27 0.70 0.457 12.184 0.513 0.660 14.0 µ = Dn Dny urno = Γnφrn Dn (cm) 83 .75 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.678 0.03 26.52 “Mode” 3 1.766 7.691 0.5 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 2.332 48.52 0.18 27.82 1.71 1.450 4.28 46.37 1.4222 3.1.

84

**Appendix C FEMA Force Distribution Calculations
**

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

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

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

Floor, j

mj (10-3×kg)

s* = j

∑i mi

0.112 0.110 0.110 0.110 0.110 0.110 0.110 0.110 0.119

mj

1 2 3 4 5 6 7 8 9

503.5 494.7 494.7 494.7 494.7 494.7 494.7 494.7 534.1

C.2

EQUIVALENT LATERAL FORCE (ELF) DISTRIBUTION 85

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

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

Floor j 1 2 3 4 5 6 7 8 9

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

-3 k

s* j

=

m jhk j

∑i mihik

0.007 0.020 0.038 0.062 0.091 0.126 0.165 0.210 0.281

C.3

SRSS DISTRIBUTION

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

N

∑ n (V jn )

2

to get column 8 of

2 200. FEMA 273 SRSS force distribution Lateral Forces Floor j (1) 1 2 3 4 5 6 7 8 9 f j1 (kN) f j2 (kN) f j3 (kN) V j1 (kN) (5) 1917.3 -6.4 1842.1 -967.6 286.6 1233. the lateral forces are normalized by the base shear to obtain column 10 in Table C.6 366.5 -973.2 97.7 2065.047 0.2 277.177 0.7 525.9 366.1 -438.7 101.065 0.9 -153.6 -732.2 95.3.090 0.9 832.7 -46.098 0.1 87.5 159.0 354.7 (10) 0.0 980.5 -320. Table C.6 319.0 136.6 V j3 (kN) (7) 478.3.9 1446.042 0.4 400.3 222.7 374.For convenience.6 -352.0 381.9 880.2 285.8 430.1 832.5 215.2 Lateral Force fj (kN) (9) 203.4 1759.4 250.6 -359.7 234.3 240.2 fj = ∑ fi i fj (2) (3) (4) 59.1 -646.045 0.8 374.9 1683.0 -5.2 105.1 -525.367 87 .9 -166.5 -350.5 320.7 1622.0 1476.7 694.3 -646.7 Vj (kN) (8) 2268.1 1857.7 95.9 446.070 0.2 148.8 -326.7 355.8 1381.0 1231.1 Story Shears V j2 (kN) (6) 1114.7 1578.0 176.