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

Welcome to Scribd! Start your free trial and access books, documents and more.Find out more

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

iv .

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

vi .

.... Comparison of Modal and FEMA Pushover Analyses......17 3.36 Modal Pushover Analysis ...4......................................................................................................................................3 Summary............................................................27 4..13 3..............................................................................................................................................................1 5......................................................................................................................13 3............................................................. iii Acknowledgment .......................................................3 2.....................2 2............1 Uncoupled Modal Response History Analysis .................................. 2..2...........................................................28 4...............v Table of Contents..............................2.............2.......55 Comparative Evaluation.1 Summary... Inelastic Multistory Buildings .............1 One-Story Systems ...............................................................................................................................3 Modal Pushover Analysis with Gravity Loads ....................................................47 4...........................................3 System and Excitation Considered ..................................................................4.............................................41 4........................ Introduction.......................................4 3...........................1 2.............................................38 4...........................33 4.............36 4.........9 3...12 Comparative Evaluation of Analysis Procedures .............. vii 1.1 Underlying Assumptions and Accuracy ..................................37 Comparative Evaluation of Analysis Procedures ....4.................................................................................................................................5 Response History Analysis .......................................................3 4..........................4 5.........................3 2......................................1 System and Excitation Considered ...27 Uncoupled Modal Response History Analysis .......................................................................22 4.............................6 Elastic Multistory Buildings ...........................2 FEMA-273 Pushover Analysis ....................................................................................................4...............................CONTENTS Abstract .....................................1 3.............................55 vii ............................4..............................12 Modal Pushover Analysis ...................2 Response History Analysis ................2 Response History Analysis .....................................4.........................................................................................................2 3............4 Modal Response History Analysis....................................................................3 3....................................................................9 Modal Response Spectrum Analysis .................................................................................................................1 4...................55 5..................................... Equation of Motion ................................................................................3.............2 Modal Pushover Analysis ..........34 4................38 4..............................................3 Modal Pushover Analysis ..........................................................6 Pushover Analysis .........................2 Properties of nth-“mode” Inelastic SDF System ..............

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

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

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

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

and R y (Chopra . sign u = . This becomes evident if Eq. m depends on three system parameters: w n .e. 2.ug t n where bg bg b g bg (2.3) ωn = k m ζ = c 2mω n f fs = s fy (2. the deformation u t depends on three systems parameters: w n . u £ u y ). in addition to the form of the force-deformation relation. and u y .4) and w n is the natural vibration frequency.5) For a given u g t . (2. z . Section 7. Tn = 2p w n is the natural vibration period. (a) Idealized one-story structure.. bg 4 . or absolute (without regard to algebraic sign) maximum. The peak.fS m fs c ut u fy 1 αk k 1 uy ug -fy k k um u 1 1 (a) (b) Fig. and z is the damping ratio of the system vibrating within its linear elastic range (i. and (b) bilinear hysteretic forcedeformation relation For a given excitation u g t . 2001. and the ductility factor is u m= m uy (2.1) is divided by m to obtain ~ u + 2zw n u + w 2 u y f s u.1.3). z y . deformation is denoted by um .

9 kN (38.2: the dimensions and flexural rigidity of the structural elements are noted.1.26 kN (8.5 sec and z = 5% subjected to the north-south component of the El Centro (1940) ground motion (Fig.2311 . Vb (kN) 30 20 10 0 0 2 4 6 Roof Displacement.04 40 Base Shear.2. based on Ry = 8 . α = 0.2 SYSTEM AND EXCITATION CONSIDERED Consider the one-story system in Fig. 2. one-bay frame 50 uy = 1. 6.3. 2.66 m • EIb EIc EIc • • L = 7. 2. with Tn = 0.4 in Chopra. f o w = 184 . 2. and f y = 39.376 cm. u (cm) 8 Fig. For . Vby = 39. is f y / w = ( f o / w ) ÷ 8 = 0.2.2 5 . this system and excitation. m h = 3. One-story.2 kips) .32 m • Fig. Pushover curve for structure shown in Fig.26 kN. The yield strength of the inelastic system.826 kips) for w = 169. 2001) and scaled up by a factor of 2.

Implementing this analysis with I c = 6. However. (d) shows the rotation q p bt g of the plastic hinges at the bg bg beam ends. 2. and (e) shows the force-deformation relation.35.6 kip-in.65 kN-m (191. the energy dissipated in 6 . 2. 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. The yielding stiffness of each structural element is defined as 3% of its initial stiffness. (b) shows the lateral force f s t or base shear Vb t normalized relative to the bg weight. Observe that the pushover curve matches the initial loading path of the hysteretic system.1 kip-in. the joint rotation and beam hinge rotation are identical to values q m and q pm determined from RHA.0217 rad . and E = 2 × 108 kPa (29×103 ksi) gives yield moments of 21.) for the beam and columns.36 cm . the ductility factor m = 5.077 × 107 mm 4 (146 in. the peak value determined from RHA. Determined from pushover analysis at the exact peak deformation. 2. wherein the hysteretic force-deformation history of Fig. pushover analysis cannot provide any cumulative measure of response. as apparent in Fig. The resulting pushover curve is shown in Fig.4edetermined from RHAis superimposed.3 RESPONSE HISTORY ANALYSIS Figure 2.4 ).017 rad .3.g.3 in. (c) shows the joint rotation q bt g .4f.36 cm. The peak values of the various response quantities are as follows: um = 7.4 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 = 7.4 ). 2.18 kN-m (444.) and 50.4 shows the earthquake response of the system described in the preceding section determined by response history analysis (RHA). The system is excited well beyond the yield deformation.The yield moments in the beam and columns are defined as the bending moments due to the lateral force f y . 2. and q pm = 0. I b = 3. e. 2. q m = 0.134 × 107 mm 4 (75. A static pushover analysis of this one-story system leads to the force-displacement relationship shown in Fig. It is organized in five parts: (a) shows the deformation u t .4e.

017 • 0 (d) θp (rad) −0.04 0.yielding during the ground motion. (b) base shear. and (f) pushover curve 7 .5 0. 15 u (cm) 0 • 7.4 −10 −5 0 u (cm) 5 10 −0.04 0.4 0.2 0 (e) 0 (f) −0. 2. (c) joint rotation.2 −0.04 0.5 fy / w = 0.4 0. Response of one-story system to El Centro ground motion: (a) deformation.2311 (a) Vb / w 0 f / w = −0.4 −10 −5 0 u (cm) 5 10 Fig.4.0217 • θ (rad) 0 (c) −0. (e) force-deformation relation.2 −0. or the cumulative rotation at a plastic hinge. This represents an inherent limitation of pushover analyses. (d) plastic hinge rotation.2311 y (b) −0.36 −15 0.2 Vb / w 0.04 0 5 10 15 Time (sec) 20 25 30 0.

8 .

12): mι = n =1 Â sn = N n =1 Â Γ nmφ n N (3. classical damping.1 MODAL RESPONSE HISTORY ANALYSIS The differential equations governing the response of a multistory building to horizontal earthquake ground motion u g t are as follows: mu + cu + ku = . (3.4) 9 .n t = Â . c.3. The right side of Eq. each element of the influence vector ι is equal to unity. m.1) where u is the vector of N lateral floor displacements relative to the ground.2) The spatial distribution of these “forces” over the height of the building is defined by the vector s = m ι and their time variation by u g t .3) where φ n is the nth natural vibration mode of the structure. and lateral stiffness matrices of the system. and k are the mass.mι ug t bg bg bg (3.m ι u g t bg bg (3.1) can be interpreted as effective earthquake forces: p eff t = .snug t n= n =1 The effective earthquake forces can then be expressed as af N af N af (3. 2001: Section 13. This force distribution can be expanded as a summation of modal inertia force distributions s n (Chopra. and p eff t = Â peff. Elastic Multistory Buildings 3.

12) . we will outline that the response of the MDF system to p eff .7) By utilizing the orthogonality property of modes. p eff .6) Next. subjected to u g t : Dn + 2z nw n Dn + w 2 Dn = .sn ug t bg bg bg bg (3.peff (t ) = n =1 ∑ peff .9) and (3. The equations governing the response of the system are mu + cu + ku = .5) The contribution of the nth mode to s and to p eff t are: sn = Gn mφ n respectively. The solution qn t can readily be obtained by comparing Eq.8) gives the floor displacements u n t = Gnφ n Dn t bg (3. (3.8) where the modal coordinate qn t is governed by 2 qn + 2z nw n qn + w n qn = .sn ug t bg (3. (3. Then the floor displacements are u n t = φ n qn t bg bg bg bg (3. (3.n t = .u g t n Comparing Eqs.n (t ) = N n =1 ∑ −sn u g (t ) N (3. an SDF system with vibration propertiesnatural frequency w n and damping ratio z n of the nth-mode of the MDF system.n t is entirely in the nth-mode. it can be demonstrated that none of the modes other than the nth mode contribute to the response.10) bg bg bg 10 (3.11) and substituting in Eq.10) gives qn t = Gn Dn t bg bg bg (3.9) to the equation of motion for the nth-mode elastic SDF system.9) in which w n is the natural vibration frequency and z n is the damping ratio for the nth mode. with no contribution from other modes.Gn ug t (3.

15) rt = bg n =1 Â rn t = bg n =1 Â rnst An bt g N (3.Forces sn An(t ) st rn ωn. (3.16) This is the classical modal RHA procedure wherein Eq. (3. In contrast to the classical 11 bg . internal element forces.14) is the pseudo-acceleration response of the nth-mode SDF system (Chopra. st The two analyses that lead to rn and An t are shown schematically in Fig.13) define the contribution of the nth-mode to the response.16) reflect combining the response contributions of all modes. Equations (3.can be expressed by st rn t = rn An t bg bg bg bg bg (3.13) st where rn denotes the modal static response.1. etc. (3. the static value of r due to external forces sn . However.1. Eqs. Conceptual explanation of modal response history analysis of elastic MDF systems Any response quantity r t story drifts.15) and (3. 3.12) and (3. Section 13.13) represent the response of the MDF system to p eff. 2001. 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 (3.9) is the standard modal equation governing qn t .n t .1). ζn ¨ g(t ) u (a) Static Analysis of Structure (b) Dynamic Analysis of SDF System Fig.12) and (3. Therefore. 3. these standard equations have been derived in an unconventional way. and An t = w 2 Dn t n bg (3. and Eqs.

. and Tn = 2p w n is the natural vibration period of the nth-mode of the MDF system. Chopra. In such a response spectrum analysis (RSA).17) (Chopra.19) will provide the same value of rno . z n of the pseudo-acceleration response (or design) spectrum for the nth-mode SDF system.g. Alternatively.3).derivation found in textbooks (e. we used the modal expansion of the spatial distribution of the effective earthquake forces.(3. This concept provides a rational basis for the modal pushover analysis procedure developed later.1.8. Sections 12. which is valid for structures with well-separated natural frequencies such as multistory buildings with symmetric plan. 2001. The SRSS rule. this response value can be obtained by static analysis of the structure subjected to lateral forces distributed over the building height according to 12 . 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 (3.18) To develop a pushover analysis procedure consistent with RSA. (3. (3. 2001. The peak modal responses are combined according to the Square-Root-of-Sum-ofSquares (SRSS) or the Complete Quadratic Combination (CQC) rules.1). 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 3.2 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.17) where An is the ordinate A Tn . 3.16). the peak nth-mode response as in Eq.9) .3 MODAL PUSHOVER ANALYSIS (3. we note that static analysis of the structure subjected to lateral forces f no = Γ n mφ n An (3.4 and 13. Section 13.

Richmond. 1660 W. each determined by one pushover analysis. Typical floor-to-floor heights (for analysis purposes measured from center-of-beam to center1 2 Brandow & Johnston Associates.12) is urno = Gnf rn Dn (3. the Applied Technology Council (ATC). (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. shown in Fig. The building has a basement level. CA 94804-4698.20) and the structure is pushed to the roof displacement. and 37. which from Eq. SAC is a joint venture of three non-profit organizations: The Structural Engineers Association of California (SEAOC). 13 .73 m (150 ft) by 45. SAC Steel Project Technical Office.1 System and Excitation Considered The 9-story structure. The peak modal responses.4.4 COMPARATIVE EVALUATION OF ANALYSIS PROCEDURES 3.. 3. Consulting Structural Engineers.s* = mφ n n (3. The levels of the 9-story building are numbered with respect to the ground level (see Fig. Third St.2).21) 2 where Dn = An ω n . 3. was designed by Brandow & Johnston Associates1 for the SAC2 Phase II Steel Project. CA 90017. rno . can be combined according to Eq. Obviously Dn and An are available from the response (or design) spectrum. The columns are 345 MPa (50 ksi) steel wideflange sections. in both directions. denoted B1. The bays are 9. 46th Street. this building is 45. The interior bays of the structure contain simple framing with composite floors. this structure meets seismic code and represents typical medium-rise buildings designed for the Los Angeles. the peak value of the roof displacement due to the nth-mode. (3. with five bays each in the north-south (N-S) and east-west (E-W) directions.19 m (122 ft) in elevation. region.15 m (30 ft) on center. Although not actually constructed. This modal pushover analysis (MPA) for linearly elastic systems is equivalent to the well-known RSA procedure (Section 3. urno . Los Angeles. A benchmark structure for the SAC project. 1301 S.2). with the ninth level being the roof. 3.2. California. and California Universities for Research in Earthquake Engineering (CUREE).18) to obtain an estimate of the peak value ro of the total response.73 m (150 ft) in plan.

The seismic mass of the structure is due to various components of the structure. floor slabs. The column bases are modeled as pinned and secured to the ground (at the B-1 level).27. Column splices. respectively. The force distributions. The floor-to-floor height of the basement level is 3. and 0.7 kips-sec2/ft). ceiling/flooring. The strength.96 m (13 ft)..07×106 kg (73. we select a weak ground motion: the northsouth component of the El Centro (1940) ground motion scaled down by a factor of 4. including the steel framing. The 9-story N-S MRF is depicted in Fig. Each frame resists one half of the seismic mass associated with the entire structure. The seismic mass of the ground level is 9. This model is based on centerline dimensions of the bare frame in which beams and columns extend from centerline to centerline. and for the ninth level is 1. These force n distributions will be used in the pushover analysis to be presented later. s* (Eq. 3. The seismic mass of the above ground levels of the entire structure is 9. if desired more complex models.sec2/ft). for the second through eighth levels is 9. The building is modeled in DRAIN-2DX (Allahabadi and Powell. The simple model adopted here is sufficient for the objectives of this study.of-beam) are 3. are located on the first.0 kips-sec2/ft).2 kips-sec2/ft).65×105 kg (66. third.3. and seventh levels at 1.20).e.49 m (18 ft). for the first three modes are shown in Fig. To ensure that this structure remains elastic. for the first level is 1. 3.89×105 kg (67. and shear distortion of panel zones are neglected but large deformation (P∆) effects are included. the vibration periods are 2. such as those described in Gupta and Krawinkler (1999) can be used. The floor system is composed of 248 MPa (36 ksi) steel wide-flange beams in acting composite action with the floor slab. 0.49 sec. roofing and a penthouse located on the roof.00×106 kg (616 kips. 14 . monolithic column pieces are connected every two levels beginning with the first level.83 m (6 ft) above the center-line of the beam to column joint. dimension. partitions. The column lines employ two-tier construction. The first three vibration modes and periods of the building for linearly elastic vibration are shown in Fig.01×106 kg (69. mechanical/electrical. i. fifth.2. which are seismic (tension) splices to carry bending and uplift forces. Concrete foundation walls and surrounding soil are assumed to restrain the structure at the ground level from horizontal displacement.4. 3. 3.0 kips-sec2/ft).65 m (12 ft) and for the first floor is 5.85. 1988) using the M1 model developed by Krawinkler and Gupta (1998).

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

2.9th 8th 7th 6th T = 0.31 −0. First three natural-vibration periods and modes of the 9story building 3. is shown in Figs.3.1 3.7.12) and (3. n = 1.487 −1.5 Fig. (3. 3.44 1.67 −1.8 −2.0272 −2. respectively. n = 1.5 0 0.37 2.1 −2.4.38 0.2 Response History Analysis The structural response due to individual vibration modes.85 sec 2 Floor 5th 4th 3rd 2nd 1st T1 = 2.12 0.13 −1.75 1.93 −1.61 2.13)].33 2.5.27 sec Ground −1. 2. (c) shows the joint rotation q n t of an external joint at 16 bg bg bg .51 0.03 −1.5 Mode Shape Component 1 1. and 3. 3.05 1.05 2.5 −1 −0. Force distributions s* = mφn .4.728 2. 3.796 0. and 3 .72 −2.39 3. Each figure is organized in four parts: (a) shows the roof displacement urn t . 3. determined by RHA [Eqs.6.94 2.04 1. (b) shows the base shear Vbn t normalized relative to the weight W of the building. and 3 n 3.05 s * 1 s * 2 s * 3 Fig.49 sec 3 T = 0.

23 cm. For a fixed number of modes included. (3. 3. 3. two. and the percentage errors due to truncation of higher modal contributions. respectively. which are listed in Tables 3. The linear relationship between the base shear and roof displacement for each mode implies that the structure did not yield. The peak values of floor displacements and story drifts determined by RHA. three. 3. The peak values of displacements of all floors. where the second and third modal responses are a larger percentage of the top story drift compared to roof displacement. consistent with the increasing significance of the higher mode response among these three sets of response quantities.8. The peak values of the various response quantities are noted in these figures. are presented in Fig. errors generally decrease as response contributions of more modes are included. drifts in all stories.urn relation.422 cm. and three vibration modes. and (d) shows the Vbn . Combining the modal response histories for all modes gives the total response [Eqs. and rotations of external joints with moment connections are presented in Tables 3. The errors due to truncating the contributions of vibration modes beyond the third mode are negligible. especially in estimating the story drifts. and 3.12 cm.8.3. larger in story drifts. Also included are the combined response due to one. 17 .15) and (3. with the first three modes provide essentially the exact response. It is apparent that the first mode alone is inadequate. and ur 3o = 0.2. and 3. two. or all modes. errors are smallest in floor displacements. This is illustrated in Fig. the results for the roof displacement and top-story drift are shown in Fig. the exact response considering all modes.the roof level. The same method was used to determine the peak values of many response quantities. 3.1. but three modes—perhaps even two modes—are sufficient.1.2.16)].9. ur 2o = 2. in particular. the peak roof displacement due to each of three modes is ur1o = 9.3. As expected. respectively. including one. 3. and even larger in joint rotations.

205 0.3 -0.325 0.50E-03 4.1 -14.130 0.1 -2.44E-03 3.0 -46.66E-05 -3.03E-03 1.453 0.042 0.133 -0.2 -4.062 0.035 0. 18 .060 -0.177 0.4 -1.01E-04 3.74E-04 9.5 -1.3 Table 3.7 7.310 0.25 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 3 Modal Response Mode 1 2.63E-03 2.9 1.336 0.6 -0.183 0.265 RHA (all modes) 0.006 0.72E-03 3.0 -2.65E-03 RHA (all modes) 2.6 0.152 0.42E-04 1.227 0.74E-03 1.76E-03 1.9 -16.1 -2.045 0.3 -33.400 0.88E-03 2.50E-03 2.259 0.117 0.054 0.227 0.2 1.91E-04 1.45E-03 3.237 0.152 0.011 0.89E-03 1.25 × El Centro ground motion Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Modal Response Mode 1 Mode 2 Drift Ratio (%) Combined (RHA) 1 Mode 2 Modes 3 Modes Mode 3 0.245 0.73E-03 3.253 0.071 0.7 3.202 0.2 Peak values of story drift ratios (as % of story height) from RHA for 0.253 0.2 9.055 0.13E-04 9.6 11.4 -22.9 -24.6 -1.043 0.1 Table 3.15E-03 4.300 0.002 -0.012 0.181 0.4 -0.266 0.09E-03 Error (%) 1 Mode -23.282 0.90E-03 3.012 -0.225 0.399 0.03E-03 -6.8 -56.350 0.2 0.401 Error (%) 1 Mode 2 Modes 3 Modes -23.264 Error (%) 1 Mode 2 Modes 3 Modes -23.260 0.258 0.2 -1.003 0.307 0.7 2.28E-04 1.13E-03 2.4 -7.226 0.29E-03 2.08E-03 2.2 -20.4 -53.097 0.3 Peak values of joint rotations (radians) from RHA for 0.00E-03 1.14E-03 2.042 0.038 0.406 0.64E-03 3.7 2 Modes -3.11E-04 -5.069 0.1 Building height is measured from the ground floor to the 9th floor.333 0.413 RHA (all modes) 0.266 0.069 0.125 0.78E-04 -3.202 0.125 0.370 0.199 0.8 1.089 0.008 -0.364 0.03E-03 1.9 9.475 0.44E-03 1.058 -0.173 0.124 0.5 0.023 0.173 0.202 0.63E-03 2.1 3.088 -0.85E-03 3.8 -5.31E-03 Mode 2 -1.3 19.097 0.235 0.89E-03 1.6 -1.7 -19.097 0.060 0.179 0.261 -0.1 -11.0 -10.7 4.4 -1.4 -41.09E-03 2.311 0.03E-03 3 Modes 2.4 -6.9 3.00E-03 2.9 -23.407 0.01E-04 -2.6 4.56E-03 2.65E-03 2.010 -0.321 0.1 -0.8 1.032 -0.157 0.466 0.09E-03 1.3 -8.245 0.080 0.303 0.74E-04 6.263 0.009 -0.024 -0.266 0.159 0.0 -0.8 -15.22E-03 2.237 0.9 8.4 -3.001 -0.1 -19.94E-03 2.121 0.156 0.260 0.015 -0.275 0.060 -0.022 0.99E-03 Mode 3 3.99E-03 2.88E-03 2.229 0.7 -50.203 0.192 0.090 0.265 0.38E-04 2 Modes 2.00E-03 1.9 -22.8 -1.156 0.282 0.4 0.24E-03 2.9 -15.4 -10.011 0.74E-03 1.378 0.Table 3.0 7.124 0.311 0.11E-03 1.76E-03 1.226 0.1 4.6 0.1 -2.259 0.1 Peak values of floor displacements (as % of building height = 37.6 9.47E-03 1.26E-04 -5.14 m3) from RHA for 0.0 -2.33E-03 2.2 -2.295 -0.0 3 Modes -5.231 -0.38E-03 2.235 0.9 2.5 -2.026 0.3 -3.229 0.6 1.31E-03 Joint Rotation (rad) Combined (RHA) 1 Mode 2.3 -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 0.5 18.09E-03 1.197 0.062 -0.177 0.8 -10.2 -57.

(b) base shear.5 9.00174 • 0 (c) r1 −5 0 5 10 15 Time (sec) 20 25 30 0.12 −0. Excitation is 0.15 u (cm) 9.1 5 x 10 −3 θ (rad) 0.1 −15 −7. Response due to first mode: (a) roof displacement.5 15 −0.12 • 0 (a) r1 −15 0. and (e) pushover curve. (c) joint rotation.5 0 u (cm) r1 7.5.0435 Vb1 / W 0.1 /W 0.0435 9.12 0 (d) 0 (e) 15 Fig.1 0. 3.0435 • 0 (b) V b1 −0. (d) force-deformation history.1 0.5 0 u (cm) r1 7.25 × El Centro ground motion 19 .1 −15 −7.

Excitation is 0.1 −5 −2. (d) force-deformation history.25 × El Centro ground motion 20 . (c) joint rotation.6.05 5 x 10 −3 b2 θ (rad) 0.0252 −2.23 −2.5 u (cm) r2 0 • 2.1 −5 −2.5 0 u (cm) r2 2.05 /W 0.00244 • 0 (c) r2 −5 0 5 10 15 Time (sec) 20 25 30 0.23 0. 3. Response due to second mode: (a) roof displacement.0252 (d) 0 0 (e) −0.1 0. (b) base shear.5 5 −0.5 5 Fig.23 −5 (a) 0.0252 • 0 (b) V −0.5 0 u (cm) r2 2. and (e) pushover curve.1 Vb2 / W 0.

0108 x 10 −3 (b) −0.1 u (cm) r3 0 • 0.422 (e) −0.02 2 θ (rad) 0.02 −0. Response due to third mode: (a) roof displacement.25 × El Centro ground motion 21 .422 −1 (a) 0.7. 3.0108 −0.422 /W V b3 0 (d) 0 −0.5 0 u (cm) r2 0.5 0 u (cm) r3 0.02 0.5 1 −0. and (e) pushover curve.0108 −0. (d) force-deformation history.000913 • 0 (c) r3 −2 0 5 10 15 Time (sec) 20 25 30 0. Excitation is 0. (c) joint rotation.02 −1 −0.02 /W V b3 0 • 0.5 1 Fig.02 −1 −0. (b) base shear.

18).23 cm .12). considering one. (3.9 and Tables 3. and it is apparent in the example considered that three modes provide most of the response (Fig. the errors in the MPA results are generally larger than in RHA (Fig. leads to the pushover curves shown in Figs. 3. Observe that the target roof displacement in each pushover analysis is identical to its exact value determined by RHA. 3.21). respectively.4. 3. story drifts. i.6. respectively.5. (3. and the errors in these estimates relative to the exact response from RHA considering all modes. 3. and ur 3o = 0. using the force distribution of Eq.10 and Tables 3.ur relations determined by RHA (Figs. Figure 3.5.7e and to the floor displacements. and 3.e.6 present estimates of the combined response according to Eq. Implementing pushover analysis for the second and third modes. the floor displacements are proportional to the mode shape φ 1 because the structure remains elastic. 3. and the computed response values are identical to the peak response values determined from RHA (Tables 3. and external joint rotations computed by pushover analysis are presented in Tables 3. story drifts.9). As for the first mode. pushing the structure.3 Modal Pushover Analysis Implementing MPA for the fundamental vibration mode. and 3.1.5. This pushover curve is consistent with the relationship between the base shear and roof displacement determined by RHA (Fig.4) to roof displacement ur1o = 9. the value determined by RHA (Fig..5d).6e and 3.8) leads to the pushover curve shown in Fig. respectively. The floor displacements.1. (3. (3.20) with n = 1 (Fig. these pushover curves are consistent with the Vb .3). implying that the modal truncation errors are small if at 22 . confirming that MPA gives the exact values of the individual modal responses.7d). except for the algebraic sign associated with Γ1 and minor round-off errors. which would provide the Dn value to be substituted in Eq.3). pushing the structure using the force distribution of Eq. and 3. 3.20) with n = 2 and 3 up to roof displacements ur 2 o = 2. 3. 3. 3. For a fixed number of modes included considered.422 cm . In RHA the errors arise only from truncating the responses due to higher modes.4.3. 3. or three vibration modes. two.1 through 3.5e. 3. although both analyses led to identical peak values of the individual modal responses. These values of the response quantities are identical to the peak response values determined from RHA (Tables 3. (3. and 3.3). In practical application.4..4.2. 3.6.12 cm. i. this value would be determined directly from the response (or design) spectrum. and 3. As suggested by Eq.6d and 3. and external joint rotations in Tables 3.2. 3.e.

(a) Roof Displacement 15 (b) Top Story Drift 3 Mode 1 ∆ (cm) u (cm) 9. Response histories of roof displacement and top-story drift from RHA for 0.1 0. Additional errors are introduced in pushover analysis due to the approximation inherent in modal combination rules.eps r 5 10 15 20 Time.5 Ground 0 0.422 r2 0 • 1.59 • 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 9.3 0.5 0. 3.2 0.685 • −3 3 Mode 2 ∆ (cm) u (cm) • 2.eps fig3_9b. 3.4 Story Drift Ratio (%) 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.8. Heightwise variation of floor displacements and story drifts from RHA for 0.12 • r1 −15 15 0 r1 0 0 0.1 0.25 × El Centro ground motion 23 .2 0.23 r2 −15 15 0 • 0. sec 25 30 Fig.least three modes are included.eps Ground 0 0.4 Displacement/Height (%) 0.48 −3 3 1.3 0.83 • r3 0 • 0.9. sec 25 30 r 5 10 15 20 Time.6 Fig.

125 0.8 -56.44E-03 -1.152 0.253 0.9 3 Modes -12.069 0.14 m) from MPA for 0.00E-03 3.9 -13.74E-03 -1.4 -11.227 0.097 0.9 -24.1 -11.31E-03 2.15E-03 1.266 0.125 0.245 0.4 -5.3 -41.2 -4.173 0.2 -11.4 -22.22E-04 2.336 0.08E-03 2.181 0.1 -19.4 -9.4 -7.3 -12.261 -0.00E-03 -1.0 -2.9 -12.121 0.1 -0.336 RHA (all modes) 0.370 0.313 0.03E-03 1.282 -0.5 Peak values of story drift ratios (as % of story height) from MPA for 0.4 1.156 0.03E-03 -1.060 0.9 -15.157 0.253 0.069 0.4 -9.7 -21.332 0.237 0.90E-04 -9.7 2 Modes -13.09E-03 1.76E-03 1.9 -11.4 -53.062 -0.023 -0.331 0.124 0.0 -46.0 -18.9 -8.6 -17.63E-03 2.2 -57.4 Peak values of floor displacements (as % of building height = 37.9 -13.001 0.42E-04 -1.024 0.4 -19.9 -15.270 0.90E-03 1.156 0.38E-03 3 Modes 2.267 -0.245 -0.8 -22.282 0.259 -0.260 -0.1 -11.89E-03 2.011 -0.89E-03 1.177 0.89E-03 -1.401 Error (%) 1 Mode 2 Modes 3 Modes -23.253 -0.002 0.04E-03 3.57E-03 1 Mode -23.5 -16.088 -0.226 0.267 0.296 -0.079 0.09E-04 -3.4 -14.203 0.3 -9.090 0.006 -0.011 0.72E-03 3.2 -20.8 -23.73E-03 3.0 -16.133 0.2 1.38E-03 2.3 -2.078 0.310 0.09E-03 Mode 2 1.73E-05 3.9 -16.6 -19.31E-03 Error (%) RHA (all modes) 2.80E-04 3.230 -0.010 0.8 -15.28E-03 2.285 0.179 0.00E-03 2.4 0.89E-03 2.042 0.3 -33.9 -14.9 -14.12E-03 1.055 0.203 0.012 -0.2 -12.045 -0.267 0.40E-04 5.22E-03 -2.230 0.022 0.0 -10.00E-03 1.0 -0.25 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Modal Response Mode 1 -2.106 0.048 0.7 -19.9 -14.3 -14.8 -15.96E-03 2.276 0.008 -0.197 0.286 0.6 -15.321 0.9 -13.009 0.63E-03 -2.65E-03 2.229 0.99E-03 Mode 3 -3.9 -18.2 Table 3.080 0.015 -0.9 -15.7 -50.385 0.048 0.179 0.9 -15.92E-04 -1.25 × El Centro ground motion Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Modal Response Mode 1 Mode 2 Drift Ratio (%) Combined (MPA) 1 Mode 2 Modes 3 Modes Mode 3 -0.8 -14.00E-03 2.237 -0.2 -16.43E-04 -1.235 -0.Table 3.272 0.25 × 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 0.259 0.09E-03 -1.042 0.3 -19.071 0.2 -0.15E-03 2.203 0.33E-04 5.260 0.47E-03 1.1 -18.03E-03 6.05E-03 3.173 0.012 0.274 0.24E-03 -2.94E-03 2.300 0.270 0.203 0.264 Error (%) Modal Response 1 Mode 2 Modes 3 Modes -23.7 24 .76E-03 -1.53E-04 -9.7 2.466 0.133 -0.03E-03 3.00E-03 2.177 0.058 -0.133 0.407 0.043 -0.106 0.7 -15.74E-04 -6.270 0.203 0.062 0.4 -4.117 0.060 -0.322 0.032 0.9 -13.374 0.6 -16.89E-03 -2.97E-03 1.44E-03 3.157 0.003 -0.3 1.4 -22.253 RHA (All Modes) 0.328 0.036 -0.4 -4.74E-03 1.227 0.026 0.152 0.3 Table 3.3 -13.229 0.235 0.231 -0.31E-03 Joint Rotation (rad) Combined (MPA) 1 Mode 2.6 Peak values of joint rotation (radians) from MPA for 0.78E-04 2 Modes 2.260 0.097 0.4 -2.038 -0.

6 Fig.4 Story Drift Ratio (%) 0.10.4 Displacement/Height (%) 0.2 0.5 0.1 0.3 0. 3.2 0.eps fig3_10b.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.5 Ground 0 0. shading indicates modal combination error 25 .3 0. Heightwise variation of floor displacements and story drifts from MPA for 0.25 × El Centro ground motion.1 0.

26 .

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

6) The solution of Eq.3) in Eq. Unlike Eq. However. where φ n are now the modes of the corresponding linear system.4) where the only term that differs from Eq.sn ug t bg b g bg bg (4. Equation (4.4) represents N equations in the modal coordinates qn . implying coupling of modal coordinates because of yielding of the structure. (3. give the same results for u t as obtained directly from Eq. (4. 4. these equations are coupled for inelastic systems.” implying that other “modes” will also contribute to the solution. premultiplying by φ T .2 UNCOUPLED MODAL RESPONSE HISTORY ANALYSIS bg Neglecting the coupling of the N equations in modal coordinates [Eq.5) This resisting force depends on all modal coordinates qn t .… N (4.6b) are mu + cu + f s u. (3. This approximate RHA procedure is the preliminary step in developing a modal pushover analysis procedure for inelastic systems. (4. sign qn = φ n f s u n . For linear systems.2). (4.2).8) because qr t will generally be nonzero for “modes” other than the nth “mode. and using the mass. (3. sign u = . The equations governing the response of the inelastic system to p eff. (4. Simultaneously solving these coupled equations and using Eq. (4.3). in principle. 2. Eq.9) involves T Fsn = Fsn qn . (4. (4.9) for linearly elastic systems.Substituting Eq. qr t = 0 for all 28 bg .4) is rarely solved because it offers no particular advantage over Eq. however.6) for inelastic systems will no longer be described by Eq. The spatial distribution s of the effective earthquake forces is expanded into the modal contributions sn according to Eq.and classical n damping-orthogonality property of modes gives F qn + 2z nw n qn + sn = . sign u n b g b g bg (4.4)] leads to the uncoupled modal response history analysis (UMRHA) procedure.n t given by Eq.Gn ug t Mn bg n = 1.2). (3. (4. (3.3) will.

(a) p 15

eff,1

= −s × 0.25 × El Centro

1

(b) p 5

eff,2

= −s × 0.25 × El Centro

2

ur1 (cm)

0

ur1 (cm)

9.112 •

Mode 1

Mode 1

0

−15 15 Mode 2

−5 5 Mode 2

ur2 (cm)

0

ur2 (cm)

0 • 2.226 5 10 15 20 25 30

−15 0 15

5

10

15

20

25

30

Mode 3

−5 0 5

Mode 3

ur3 (cm)

0

ur3 (cm)

5 10 15 20 Time (sec) 25 30

0

−15 0

−5 0

5

10 15 20 Time (sec)

25

30

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

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

motion

af

af

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

bg

bg

(a) p 80

eff,1

= −s × 1.5 × El Centro

1

(b) p 20

eff,2

= −s × 1.5 × El Centro

2

Mode 1

Mode 1 • 1.437

u (cm)

r1

0 • 48.24

ur1 (cm)

Mode 2

0

−80 80

−20 20 Mode 2

u (cm)

r2

0

3.37 •

ur2 (cm)

0 • 11.62 5 10 15 20 25 30

−80 0 80

5

10

15

20

25

30

Mode 3 0.4931 •

−20 0 20

Mode 3 • 0.7783

u (cm)

r3

0

ur3 (cm)

25 30

0

−80 0

5

10 15 20 Time (sec)

−20 0

5

10 15 20 Time (sec)

25

30

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

af

af

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

bg

bg

by Eq. (3.8),

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

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

b

g

b

g

(4.7)

30

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

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

bg

bg c h

(4.8)

c

h

(4.9)

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

b

g

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

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

bg

bg

bg

bg

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

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

4

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

31

where u g t = 3.02 −150 n=2 −150 50 n=2 14.4.575 5 10 15 20 Time (sec) 25 30 ur3 (cm) 0 • 5. Conceptual explanation of uncoupled modal response history analysis of inelastic MDF systems (a) Nonlinear RHA 150 n=1 150 (b) UMRHA n=1 ur1 (cm) • 75.n t = .Forces sn Unit mass An(t ) st rn ωn. and 3.112 5 10 15 20 Time (sec) 25 30 −10 0 −10 0 Fig. Fsn / Ln ¨ g(t ) u (a) Static Analysis of Structure (b) Dynamic Analysis of Inelastic SDF System Fig.s n u g t . n = 1. and (b) approximate solution by UMRHA bg af bg 32 . 4.89 50 r1 0 u (cm) 0 • 78.9 • 0 −50 10 −50 10 n=3 ur3 (cm) 0 • 5. Roof displacement due to p eff. ζn. 2.0 ¥ El Centro ground motion: (a) exact solution by NL-RHA.3. 4.51 • ur2 (cm) 0 ur2 (cm) n=3 14.

817 5 10 15 20 Time (sec) 25 30 −10 0 −10 0 Fig.6) by UMRHA is compared with the “exact” solution by nonlinear RHA. this intense excitation was chosen to ensure that the structure is excited well beyond its linear elastic limit.0 ¥ El Centro ground motion.5.5. (3. The errors are slightly larger in drift than in displacement.n t (n = 1.5)] is neglected. the errors in either response quantity are only a few percent. respectively. Such comparison for roof displacement and top-story drift is presented in Figs. and (b) approximate solution by UMRHA af bg bg 4. n = 1. These errors arise from the following assumptions and approximations: (1) the coupling among modal coordinates qn t arising from yielding of the system [recall Eqs.s n u g t . Top story drift due to p eff. (4. (4.n t = . and (3) the Fsn Ln .33 0 −20 20 −20 20 n=2 ∆r2 (cm) 0 ∆r2 (cm) 7.855 ∆r1 (cm) 0 ∆r1 (cm) n=2 • 6.2… N ) according to Eq.0 ¥ El Centro ground motion: (a) exact solution by NL-RHA.2. and 3. where u g t = 3.008 • 6. but even for this very intense excitation.956 5 10 15 20 Time (sec) 25 30 ∆r3 (cm) 0 • 5. (2) the superposition of responses to p eff.(a) Nonlinear RHA 20 n=1 20 (b) UMRHA n=1 • 5.Dn relation is 33 bg bg .744 • 0 −20 10 n=3 −20 10 n=3 ∆r3 (cm) 0 • 5. 4.4) and (4. 4. both for 3. 2.1 Underlying Assumptions and Accuracy The approximate solution of Eq.4 and 4.15) is strictly valid only for linearly elastic systems.

4. (4. A bilinear idealization of this pushover curve for the nth-“mode” is shown in Fig. (4.urn pushover curve to the Fsn Ln . is plotted against roof displacement urn . Although most commercially available software cannot implement such displacement-controlled analysis.20).Dn relation to be determined in Eq. the only force distribution that produces displacements proportional to f n is given by Eq.2 Properties of the nth-“mode” Inelastic SDF System How is the Fsn Ln . What is an appropriate invariant distribution of lateral forces to determine Fsn ? For an inelastic system no invariant distribution of forces can produce displacements proportional to f n at all displacements or force levels.20) to some bg pre-determined roof displacement. (4.approximated by a bilinear curve to facilitate solution of Eq. such nonlinear static analysis provides the so-called pushover curve.10) .8) before it can be solved? Because Eq.Dn curve. it can conduct a force-controlled nonlinear static analysis with an invariant distribution of lateral forces. the base shear is Vbny and roof displacement is urny . 4.9).8) governing Dn t is based on Eq. which is different than the Fsn Ln . (4. Therefore. (3. Therefore we impose this constraint in developing the UMRHA procedure in this section and modal pushover analysis in the next section. (3. and the base shear Vbn . 4. the relationship between lateral forces f s and Dn in Eq. However. How to convert this Vbn .6a. within the linearly elastic range of the structure.12) for floor displacements. At the yield point. this distribution seems to be a rational choiceeven for inelastic systemsto determine Fsn in Eq. when specialized for linearly elastic systems it is identical to the RHA procedure of Section 3. (3.8) in UMRHA.9) should be determined by nonlinear static analysis of the structure as the structure undergoes displacements u = Dnφ n with increasing Dn . Although several approximations are inherent in this UMRHA procedure.Dn relation? The two sets of forces and displacements are related as follows: V Fsn = bn Γn Dn = urn Γ nf rn 34 (4. When implemented by commercially available software. The structure is pushed using the force distribution of Eq. The overall errors in the UMRHA procedure are documented in the examples presented in Section 4. (4.2.1.

6b is w 2 . Knowing Fsny Ln and Dny from n Eq.2. (4.5). should be used in Eq.6. which is not a meaningful quantity. 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 (4.12) implying that the initial slope of the curve in Fig. 2001. 4.13) This value of Tn . (4.11). 4.11) * in which M n = Ln Γn is the effective modal mass (Chopra. Section 13. In contrast. 4. 4.8). 35 . where the yield values of Fsn Ln and Dn are Fsny Ln = Vbny * Mn Dny = urny Γ nf rn (4. the initial slope of the pushover curve in Fig.10 enables conversion of the pushover curve to the desired Fsn Ln − Dn relation shown in Fig. Properties of the nth-“mode” inelastic SDF system from the pushover curve Equation 4. The two are related through Fsny Ln 2 = w n Dny (4.6b.6a is 2 kn = ω n Ln .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. which may differ from the period of the corresponding linear system.

Consider a nonlinear static analysis of the structure subjected to lateral forces distributed over the building height according to s* [Eq.6a).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 .6b. Convert the idealized pushover curve to the Fsn Ln − Dn relation (Fig.20)].15) and (3.3 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. 5. 6. Repeat Steps 2 to 6 for as many modes as required for sufficient accuracy. 4. Idealize the pushover curve as a bilinear curve (Fig. Combine the “modal” responses using Eqs.2. 9. n (3. (4. Dn (t ) . Compute the natural frequencies. develop the base-shear – roof-displacement ( Vbn − urn ) pushover curve for the force distribution s* [Eq.21) where Dn . of the total response r t obtained in Step 8. (3.n t . (3. (3.11). for linearly-elastic vibration of the building. 4. details are available in Appendix A: 1.16) to determine the total response. 4.3b) with force-deformation relation of Fig. which is summarized next as a sequence of steps. Typically. 8. For the nth-mode. is now determined by 36 bg bg . the peak value of Dn t .6b) by utilizing Eq. of the nth-“mode” inelastic SDF system (Fig.13).12) and (3. and modes. 2. φn . Compute the deformation history. This value of the roof displacement is given by Eq. ωn . 4. 4. r o . with the structure is pushed to the roof displacement urno . An t .4. n 3.20)]. 4. and pseudo-acceleration history. Calculate histories of various responses by Eqs. Calculate the peak value. (3. 7. the first two or three modes will suffice.

4. which is summarized next as a sequence of steps. plastic hinge rotations. (4. of the nth-“mode” inelastic SDF system (Fig. 5. story drifts. At this roof displacement.solving Eq.2.. we will refer to rno as the peak “modal” response even in the case of inelastic systems. 6. The peak “modal” responses rno . each determined by one pushover analysis. 2001.6b by solving Eq. (4. the lateral force distribution used possesses two properties: (1) it appears to be the most rational choice among all invariant distribution of forces. bg bg bg bg Thus. (3. and (2) it provides the exact modal response for elastic systems.3. However. e. it can be determined from the inelastic response (or design) spectrum (Chopra. As shown in Sections 3.n t . The response value rno is an estimate of the peak value of the response of the inelastic system to p eff. Calculate the peak roof displacement urno associated with the nth-“mode” inelastic SDF system from Eq. governed by Eq.6 and 7. Dn . alternatively.18). Steps 1 to 4 of the MPA are same as those for UMRHA. (3.2 and 3.6). 4. This application of modal combination rules to inelastic systems obviously lacks a theoretical basis. as mentioned earlier. This pushover analysis. details are available in Appendix B. are combined using an appropriate modal combination rule. 4. it seems reasonable because it provides results for elastic buildings that are identical to the well-known RSA procedure described in Section 3.1 Summary The peak inelastic response of a building to earthquake excitation can be estimated by the MPA procedure just developed.8). etc.8). joint rotations. or from the inelastic response (or design) spectrum.g. 37 .2. the pushover analysis provides an estimate of the peak value rno of any response rn t : floor displacements. seems reasonable. It provides results for elastic buildings that are identical to the well-known RSA procedure (Section 3. (4.4.3b) with force-deformation relation of Fig. to obtain an estimate of the peak value ro of the total response.3. for elastic systems.3) because. Compute the peak deformation. although somewhat intuitive for inelastic buildings.21). Sections 7.12). as described in Section 4. Eq. rno also represents the exact peak value of the nth-mode contribution rn t to response r t .

the “exact” results. and the “exact” response from nonlinear RHA for the roof displacement and top-story drift. Figure 4. This deficiency also implies that. respectively. 4. 4.7 cm. and three “modes”.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. and three modes are compared with the “exact” values in Fig. the first two or three “modes” will suffice. At urno . the UMRHA procedure lacks a rigorous theory. The peak values of floor displacements and story drifts including one.2). Determine the total response by combining the peak “modal” responses using the SRSS combination rule of Eq.18).2.1 and 4. ur 2o = 11. The peak values of response are as noted. with. and compared with the results of a rigorous nonlinear RHA using the DRAIN-2DX computer program. This is to be expected. determined by the UMRHA procedure.0 to 3. 8. the peak roof displacement due to each of the three “modes” is ur1o = 48. To ensure that this structure responds well into the inelastic range the El Centro ground motion is scaled up by a factor varying from 1.1). also included are the combined responses due to one. subtract the yield value of hinge rotation to determine the hinge plastic rotation. and the errors in the approximate results are shown in Fig.9.53 cm.7 shows the individual “modal” responses. Typically.3 cm. two. in particular.4.1 Uncoupled Modal Response History Analysis The structural response to 1.4. (3.7. three “modes” included. although the trends are not systematic as when the system remained elastic (Section 3. the combined response due to three “modes”. Repeat Steps 3 to 8 for as many “modes” as required for sufficient accuracy. 9. Observe that errors tend to decrease as response contributions of more “modes” are included. extract from the pushover database values of other desired responses. the 38 .8. two. is presented next. The peak values of displacements of all floors and drifts in all stories are presented in Tables 4. 4.0. 4. and ur 3o = 2. in contrast to modal analysis (Section 3.5 ¥ the El Centro ground motion including the response contributions associated with three “modal” inelastic SDF systems. and the percentage errors in the approximate results. From the total rotation of a plastic hinge. say. rno .

0. sec 25 30 r −12 0 Fig. 1. 0.5. just as for elastic systems.7 • −80 80 ∆r1 (cm) u (cm) "Mode" 1 −12 r2 0 ∆r2 (cm) 12 0 u (cm) −80 80 −12 r3 0 • 2.24 • NL−RHA r −12 12 ∆r (cm) u (cm) 0 5 10 15 20 Time. recall that the computed errors have been presented earlier for ground motion multipliers 0.1 and 4. For this purpose the UMRHA and exact analyses were repeated for ground motions of varying intensity. in particular. (a) Roof Displacement (b) Top Story Drift 80 12 0 • 3.62 5.2) if the system yields significantly versus if the system remains within the elastic range (Tables 3. 0.25.2).1 and 3.5.5 × El Centro ground motion: individual “modal” responses and combined response from UMRHA.44 • "Mode" 2 r1 0 • 48. 2.0. and total response from NL-RHA 39 . the degree to which the system deforms beyond its elastic limit.1 and 4.6 5 10 15 20 Time.25 (Tables 3.53 ∆r3 (cm) 12 0 • 2. Next.2) and 1. However. we investigate how the errors in the UMRHA vary with the deformation demands imposed by the ground motion.5 (Tables 4.0.2).85. the errors in responses computed by UMRHA including three “modes” relative to the “exact” response were determined.88 u (cm) "Mode" 3 −80 80 0 −80 80 0 −80 0 • 44.1 and 3. for a fixed number of “modes” included. the errors in story drifts are larger compared to floor displacements. defined as the El Centro ground motion multiplied by 0.75.7.38 6. For each excitation.1 −12 12 ∆r (cm) u (cm) UMRHA 3 "Modes" 0 • 7. and 3. 1. 0. sec 25 30 • 48.response is much less accurate (Tables 4.3 11. 4. Response histories of roof displacement and top-story drift due to 1.

For this purpose. Heightwise variation of error in floor displacements and story drifts estimated by UMRHA including one. in each story drift (Fig.10 summarizes the error in UMRHA as a function of ground motion intensity. 4.5 1 1. To interpret these results.11.5 Fig. the pushover curves using force distributions s* [Eq.5 1 1. Heightwise variation of floor displacements and story drift ratios from UMRHA and NL-RHA for 1. 4. Shown is the error in each floor displacement (Fig. 4.5 Displacement/Height (%) 2 Ground 0 0. indicated by a ground motion multiplier.5 ¥ El Centro ground motion Figure 4.9.20)] for the first three n modes of the system are shown in Fig. or three “modes” for 1. it will be useful to know the deformation of the system relative to its yield deformation. (3. Two versions of the pushover curve are 40 . 4.10a).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.(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. two.8. with the peak displacement of each “modal” SDF system noted for each ground motion multiplier.5 Story Drift Ratio (%) 2 2.10b). and the error envelope for each case. 4.

Additional errors are introduced in UMRHA of systems responding beyond the linearly elastic limit for at least two reasons. as mentioned in Section 4. The structure is pushed using the force distribution of Eq. UMRHA lacks a rigorous theory and is based on several approximations.10) for each “modal” inelastic SDF system (Figs 4. Figure 4. The system remains elastic up to ground motion multiplier 0. The location of plastic hinges and their rotations. 3. even though the system remains essentially elastic.2.3 cm. (2) increase rapidly as the ground motion multiplier increases to 1.included: the actual curve and its idealized bilinear version.01 (Fig. Each of these three pushover analyses provides the pushover curve (Fig. The idealized curve for the first “mode” deviates most from the actual curve near the peak displacement corresponding to ground motion multiplier 1.7 cm.0. 4. determined from “exact” analyses.4). 2.11). were noted but are not shown here.20) with n = 1. 4. 11. First.75. 4. 41 .5).2.6 and 4. the pushover curve for each “mode” is idealized by a bilinear curve in solving Eq.75. For more intense excitations. and 2. the values determined by RHA of the nth-mode inelastic SDF system (Fig. Second. and 3 (Fig. (4. and (4) are larger in story drift compared to floor displacements. and. the peak values of displacements at all floors (Table 4. and plastic hinge rotations at the external beam end at each floor level (Table 4. the first reason mentioned above seems to be the primary source for the errors.4.7). drifts in all stories (Table 4.0.4.10 permits the following observations regarding the accuracy of the UMRHA procedure: the errors (1) are small (less than 5%) for ground motion multipliers up to 0.11a). the ductility factor for the first mode system is only 1. Perhaps this explains why the errors are large at this excitation intensity.4) to roof displacements urno = 48.2 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. as mentioned in Section 3. (3) maintain roughly similar values for more intense ground motions. respectively. the errors in truncating the higher mode contributions are negligible.53 cm. (3.1).3). 4.5 ¥ the El Centro ground motion.

4.863 0.2 22.260 -15.8 1.9 12.6 4.877 0.513 -0.811 1.490 1.220 -0.055 -0.376 -1.096 0.088 10.126 0.820 -19.554 1.490 -1.938 -1.5 9.283 1.472 1.372 -1.8 1.079 0.201 -0.009 -0.4 -1.366 0.248 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.707 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 0.214 0.3 6.751 1.1 1.806 -0.201 -1.072 -1.315 -0.7 31.5 2 GM Multiplier 2.241 1.914 -0.071 -0.900 -10.372 1.473 -22.1 0.130 0.982 9.2 Peak values of story drift ratios (as % of story height) from UMRHA for 1.942 -0.3 25.003 0.373 -0.033 0.6 2.226 -0.293 1.104 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” -0.722 0.513 0.407 -10.049 -0.065 0.526 -0.072 1.133 1.220 0.135 9.9 31.938 1.3 8.4 4.9 16.4 1.138 1.370 -0.0 1.5 28.478 0.5 1 1.9 12.676 0.495 1.121 -0.10.256 -1.235 -0.983 1.727 1.772 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 1.187 -0.6 1.317 0.8 1.1 8. Errors in UMRHA as a function of ground motion intensity: (a) floor displacements.844 -25.426 -1.5 18.9 31.136 1.819 2.379 1.914 2.668 -23.169 0.5 -3.298 0.154 0.044 1.7 Table 4.120 1.5 3 Fig.298 -0.256 1.(a) Floor Displacements 60 50 40 Error Envelope Error for Floor No.291 0.366 -0.616 -0.410 -1.0 11.852 1.5 28.201 1.2 1.8 1.8 0.1 Peak values of floor displacements (as % of building height = 37.820 -0.1 1.070 1.241 -1.350 -0.8 14.945 -37.071 0.1 3.0 11.698 1.333 0.338 1.971 1.5 10.371 0.5 9.0 2.057 -0.2 12.806 0.5 3 0 0 0.763 -15.3 0.2 1.018 0.575 -41.3 1.9 5.5 1 1.663 -0.5 2 GM Multiplier 2.256 1.3 42 .376 1.5 × El Centro ground motion Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Modal Response “Mode” 1 “Mode” 2 Drift Ratio (%) Combined (UMRHA) 1 “Mode” 2 “Modes” “Mode” 3 -1.0 7.484 0.202 11.2 6.0 9.371 -0.8 1. and (b) story drifts Table 4. Noted 60 50 40 (b) Story Drifts Error Envelope Error for Story No.2 4.430 1.216 1.7 14.540 0.410 1.200 8.0 -9.003 -31.010 0.663 0.214 -0.4 -7.942 1.14 m) from UMRHA for 1.338 -1.863 1.856 2.068 0.

” respectively.11c). especially in estimating the story drifts (Fig. 4. and the errors in these estimates relative to the exact response from nonlinear RHA. 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.13 and Table 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%. Furthermore.” two “modes.13 and Tables 4. Obviously. however. the third “mode” does not contribute because this SDF system remains elastic (Fig.3 through 4. 4.3 and 4. overestimated by up to 18% in the middle stories. Observe that the primary contributor to plastic hinge rotations in the lower stories is the first “mode” and the second “mode” in the upper stories. this error is not especially significant because the hinge plastic rotation is very small. 4. significant improvement is achieved by including response contributions due to the second “mode”.” and three “modes. but the third “mode” contributions do not seem especially important (Fig. although the error is recorded as 100% if MPA estimates zero rotation when the nonlinear RHA computes a non-zero value.4 with Tables 4. the errors in the modal pushover results are. 4. One “mode” pushover analysis was unable to identify the plastic hinges in the upper stories where higher mode contributions to response are known to be more significant.Figures 4. the results were not always accurate. in general. Pushover analysis seems to be inherently limited in computing accurately hinge plastic rotations.1 and 4.3 and 4. even if three “modes” are included (Fig. but was more successful when the excitation was more intense (results are not included). MPA failed to identify the plastic hinges at the column bases in Fig. considering one.5). As shown in Figs. Fortuitously.13 and Tables 4. 4. The drifts are underestimated by up to 13% in the lower stories. two.4). 4.4. The second “mode” was necessary to identify hinges in the upper stories. and three “modes. For example.” and nonlinear RHA. for two or three modes included.9 and Tables 4. 4.5 present estimates of the combined response according to Eq. the hinges identified in beams at the 6th floor were at variance with the “exact” results. were determined by four analyses: MPA considering one “mode.4). and are within a few percent of the exact values for the upper stories.3 and 4.12 and 4. significantly smaller than in UMRHA (compare Fig.12 and Tables 4. 4. The first “mode” alone is inadequate.14.12 and Tables 4.2).14. However.3 and 4.18). (3. 43 . The locations of plastic hinges shown in Fig. The errors are especially large in the hinge plastic rotations estimated by the MPA procedure.13 with Fig.

5 0. 4. and 3.75 Actual Idealized 20 40 60 Roof Displacement (cm) (b) "Mode" 2 Pushover Curve 80 12000 10000 u = 9.25 5 10 15 20 Roof Displacement (cm) 25 (c) "Mode" 3 Pushover Curve 12000 10000 u = 4. V = 5210 kN. “Modal” pushover curves with peak roof displacements identified for 0.0.5 2 3 4000 2000 0 0 Actual Idealized 0.5 0.5 2 3 Actual Idealized 2000 0 0 0.5 1 0.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36. 0.0 × El Centro ground motion 44 . α = 0. α = 0.5 0. 1.5.75.19 y by 3 2 1.85 0.2 cm.11.14 y by Base Shear (kN) 8000 6000 4000 1.0.75 0.25 0. V = 4952 kN. 2. V = 7616 kN.25.75 0. α = 0.9 cm. 0.13 y by Base Shear (kN) 8000 6000 1 1. 1.85 Base Shear (kN) 8000 6000 4000 2000 0 0 0.6 cm.25 1 2 4 6 Roof Displacement (cm) 8 10 Fig.5.85 0.

now standard in engineering practice.5 × El Centro ground motion. the MPA procedure is equivalent to the RSA procedure. the errors in MPA were fortuitously smaller than in UMRHA (compare Figs. indicated by a ground motion multiplier. identified in Section 3.5 1 1. shading indicates errors in MPA including three “modes” Figure 4.5 Displacement/Height (%) 2 Ground 0 0. In this case. an observation with practical implications.4 also apply to MPA. as discussed in Sections 3.15a). implying that the modal combination errors contained in these procedures are acceptable. each story drift (Fig.15).(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 0.15b).10 and 4.15) for ground multipliers larger than 1. However. whereas MPA contains errors inherent in modal combination rules.5 Story Drift Ratio (%) 2 2. and the error envelope for each case. While various sources of errors in UMRHA. 4. 4. implying excitations weak enough to limit the response in the elastic range of the structure.5 1 1.15 summarizes the error in MPA considering three “modes” as a function of ground motion intensity. 4. As mentioned in Section 3.5 Fig. The errors are only weakly dependent on ground motion intensity (Fig.3 for elastic systems (or weak ground motions).4.4. Shown is the error in each floor displacement (Fig. the errors in MPA were larger for ground motion multipliers less than 0. implying excitations intense enough to cause significant yielding of the structure.12 Heightwise variation of floor displacements and story drift ratios from MPA and NL-RHA for 1.75. 4.0. UMRHA is essentially exact. 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. 4.2 and 3. 45 .3.

Errors in floor displacements. and three “modes” for 1. 4.5 ¥ El Centro ground motion 46 . two. and hinge plastic rotations estimated by MPA including one.13.(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. story drifts.

00E+00 0.640 -1.399 0.8 -29.9 5.3 -3.0 3 “Modes” -32.3 1.1 62.298 0.Table 4.304 1.36E-03 6.2 -100.7 1.372 0.0 -50.222 0.015 0.02E-03 3.8 1.071 0.1 62.942 6.2 6.5 10.879 1.5 × El Centro ground motion Hinge Plastic Rotation (rad) “Modal” Response Combined (MPA) “Mode” 1 7.99E-03 6.018 -0.0 -100.00E+00 0.37E-04 Error (%) “Modal” Response 1 “Mode” -32.8 -6.683 1.945 -49.244 0.02E-03 0.049 -0.266 -0.498 1.667 1.426 15.018 -0.00E+00 0.72E-03 7.6 -7.478 0.00E+00 0.36E-03 6.2 -100.9 -100.5 7.220 1.820 -7.0 -100.36E-03 6.6 1.00E+00 3.02E-03 0.1 1.728 1.336 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.7 -12.5 2.190 -0.911 0.9 1.209 1.1 -8.1 13.4 1.018 0.414 1.00E+00 3 “Modes” 7.00E+00 0.202 8.00E+00 0.55E-03 3.781 0.00E+00 2 “Modes” 7.611 0.581 0.667 -1.116 1.053 1.00E+00 “Mode” 3 0.5 × 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 0.6 -44.76E-03 4.154 0.057 0.1 46.00E+00 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th “Mode” 2 0.331 1.0 -5.8 -6.55E-03 3.298 -0.9 -100.220 1.0 -100.135 -7.00E+00 0.2 -3.8 -4.351 -0.666 Table 4.233 1.8 -29.6 13.980 0.00E+00 0.640 1.5 1.50E-10 3.00E+00 0.88E-03 1.0 2 “Modes” -32.687 0.168 -0.37E-03 1.737 0.55E-03 3.066 -0.6 7.18E-03 7.980 -0.053 -1.125 -1.652 1.503 -1.311 0.0 15.72E-03 7.233 1.5 Peak values of hinge plastic rotations (radians) from MPA for 1.6 13.895 1.068 0.1 46.473 -15.72E-03 7.5 2.222 0.752 1.407 -27.516 0.371 -0.00E+00 1 “Mode” 7.05E-03 2.5 2.514 -1.614 0.5 Table 4.7 1.00E+00 0.8 7.304 -1.3 -3.315 0.1 18.2 1.8 -29.19E-10 3.5 -6.015 0.2 -100.012 1.76E-03 4.8 “Mode” 3 -1.105 0.804 1.527 -0.76E-03 4.200 8.00E+00 0.101 -0.5 × El Centro ground motion Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th “Modal” Response “Mode” 1 “Mode” 2 Drift Ratio (%) Combined (MPA) 1 “Mode” 2 “Modes” 3 “Modes” Error (%) NL 1 2 3 RHA “Mode” “Modes” “Modes” 1.88E-03 0.435 0.0 -50.4 Peak values of story drift ratios (as % of story height) from MPA for 1.088 12.3 Peak values of floor displacements (as % of building height = 37.26E-04 9.055 0.033 -0.429 0.37E-03 1.518 1.6 -9.176 0.72E-03 7.130 0.2 0.003 -16.745 1.8 0.007 1.756 0.895 1.007 1.00E+00 0.60E-04 7.5 10.900 -0.00E+00 0.8 1.581 0.37E-03 1.071 -0.118 0.594 -1.250 0.733 1.9 2.9 0.305 -0.668 -13.1 46.844 -7.009 0.02E-03 3.694 1.3 11.3 13.503 1.756 0.8 -12.399 0.116 1.9 -4.156 -0.4 -8.705 1.1 62.00E+00 0.116 1.00E+00 0.737 1.338 1.9 1.00E+00 0.738 1.00E+00 0.60E-03 2.00E+00 0.982 13.9 7.0 1.14 m) from MPA for 1.76E-03 4.2 11.6 -8.37E-03 1.8 17.53E-03 7.10E-02 9.36E-03 6.910 1.260 -14.6 -44.641 1.781 0.0 47 .763 -14.259 1.88E-03 0.2 -4.197 -0.705 -1.066 -0.2 0.575 -53.517 1.22E-10 NL RHA 1.00E+00 0.

Figures 4. MPA including two “modes” underestimates the displacements of lower floors by up to 6% and overestimates the upper floor displacements by up to 22%. The response contributions of the second and third modes are the same as before (Tables 4.18) considering one.7 cm and ur 3o = 2.7).11a).18 and Tables 4. but it is inadequate in estimating the story drifts (Fig. resulting in the pushover curve shown in Fig.18 and Tables 4. 4.6).4) to roof displacement ur1o = 52.5). Static analysis of the structure for gravity loads provides the initial state—forces and deformations—of the structure. Also presented at these roof displacements are the displacements at all floors (Table 4.4.5 × El Centro ground motion are presented below.” these effects will modify the combined modal response as well as the results of nonlinear RHA. The story drifts are underestimated by up to 7% in the lower stories.20) with n = 1 to a target roof displacement.8). 4.” but not the contributions of higher “modes. drifts in all stories (Table 4.6 and 4.17 and Tables 4.6 through 4. 4. and overestimated by up to 29% in the middle stories and by up to 13% in the upper stories. The errors are especially large in the plastic hinge rotations estimated by the MPA procedure even if three “modes” are included (Fig. the structure is pushed using the force distribution of Eq. 4. Results of such analyses for the selected building subject to 1. as are the roof displacements ur 2o = 11.16a—which is slightly different than the one excluding gravity loads (Fig. and three “modes. They are now included in the pushover analysis of the structure for the first “mode” only.7).0 cm. 4. and the structure is pushed using the force distribution of Eq.6 and 4. The first “mode” alone provides adequate estimates of floor displacements.16 are unchanged.7). (3. (3. As shown in Fig.8 present estimates of the combined response according to Eq.” and the errors in these estimates relative to the exact response from nonlinear RHA. (3. The pushover curves for the second and third modes included in Fig.4.53 cm. two. the results presented so far did not include gravity load effects.17 and 4. 4.3 Modal Pushover Analysis with Gravity Loads To evaluate the accuracy of the dynamic response of the system. Obviously gravity load effects will influence the seismic demands due to the first “mode.20) with n = 1 (Fig.” however. 3. Starting with its initial state under gravity loads. and plastic hinge rotations at selected external beam end at each floor level (Table 4.18 and 48 . 4.4.6 and 4. the third “mode” contributions do not seem especially important (Fig.17 and Tables 4.3 . Significant improvement is achieved by including response contributions due to the second “mode.7.

5 1 1. two.5 2 GM Multiplier 2.5 1 1.5 3 Fig.15. Most pushover analysis procedures do not seem to compute to acceptable accuracy plastic hinge rotations. (a) MPA. 4. 4.5 2 GM Multiplier 2. 1−"Mode" • • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • •• • •• • • • • • •• •• • •• •• •• •• (b) MPA. Errors in MPA as a function of ground motion intensity: (a) floor displacements.8). Locations of plastic hinges determined by MPA considering one. and three “modes” and by NL-RHA for 1.5 × El Centro ground motion (a) Floor Displacements 60 50 40 Error Envelope Error for Floor No.Table 4. Noted 60 50 40 (b) Story Drifts Error Envelope Error for Story No. 2−"Modes" •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• • •• (c) 3−"Modes" • • • • • • • •• •• • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• • •• • • • • • • • • • (d) Nonlinear RHA •• •• • •• •• •• •• • •• •• •• •• •• •• •• • •• •• • •• •• •• •• • • • • • • •• •• •• Fig.14. and (b) story drifts 49 .5 3 0. 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.

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

0.85 0.5 2 3 Actual Idealized 2000 0 0 0. α = 0.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36.9 cm. V = 5210 kN.5 0.85 Base Shear (kN) 8000 6000 4000 2000 0 0 0. 0.75 0.75 Actual Idealized 20 40 60 Roof Displacement (cm) 80 (b) "Mode" 2 Pushover Curve 12000 10000 u = 9. α = 0. “Modal” pushover curves with gravity loads included.14 y by Base Shear (kN) 8000 6000 4000 1.6 cm. V = 4952 kN.25 0.50. 0.5 0. α = 0. 0. 2.5 2 3 4000 2000 0 0 Actual Idealized 0.25.0 × El Centro ground motion 51 .3 cm. 4.75 0.5 0. 1.25 5 10 15 20 Roof Displacement (cm) 25 (c) "Mode" 3 Pushover Curve 12000 10000 u = 4.19 y by 3 2 1.25 1 2 4 6 Roof Displacement (cm) 8 10 Fig.85 0.85.13 y by Base Shear (kN) 8000 6000 1 1.16.0. noted are peak values of roof displacement for 0.5 1 0. and 3.75. V = 7433 kN.

4.5 Displacement/Height (%) 2 Ground 0 0.5 1 1. and 52 .5 Fig. 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. Errors in floor displacements. Heightwise variation of floor displacements and story drift ratios from MPA and NL-RHA for 1.17.18. and hinge plastic rotations estimated by MPA including one. 4.5 1 1.5 Story Drift Ratio (%) 2 2. 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. gravity loads included.5 × El Centro ground motion. story drifts.

5 × El Centro ground motion.5 2 GM Multiplier 2. 4. two.20. Locations of plastic hinges determined by MPA considering one. 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 (a) Floor Displacements 60 50 40 Error Envelope Error for Floor No. 4. and three “modes” and by NL-RHA for 1.5 3 Fig.5 3 0. 1−"Mode" • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • •• •• •• • • • • • •• •• •• •• •• •• •• (b) MPA.5 1 1. Noted 60 50 40 (a) Story Drifts Error Envelope Error for Story No. 3−"Modes" (d) Nonlinear RHA • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• •• •• • • • • • • • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • • • • • •• •• •• Fig. Errors in MPA as a function of ground motion intensity: (a) floor displacements.(a) MPA.5 2 GM Multiplier 2. gravity loads included 53 . 2−"Modes" •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• •• •• (c) MPA.5 1 1. and (b) story drifts.19.

197 -0.667 0.2 21.854 0.330 1.4 1.5 1.19E-03 1.3 -100.214 0.8 28.4 20.372 0.9 -3.064 -10.3 1.266 -0.2 9.998 1.0 16.75E-03 0.78E-03 1.35E-03 8.637 0.933 1.35E-03 8.9 -6.0 2 “Modes” -32.19E-04 5.11E-03 9.00E+00 0.8 9.009 0.049 -0.19E-03 1.399 -0.066 -0.88E-03 1.6 1.2 4.836 -0.00E+00 “Mode” 2 0.850 -1.311 0.530 1.728 1.35E-03 8.637 0.107 1.237 0.594 -1.00E+00 0.983 1.527 -0.6 2.3 -22.102 1.319 1.353 -23.7 -2.88E-03 0.101 -0.744 1.754 0.4 -4.466 0.00E+00 Combined (MPA) 1 “Mode” 8.35E-03 8.927 1.270 -12.114 -1.0 -100.9 31.8 0.2 1.315 0.310 1.1 1.5 × El Centro ground motion.877 -46.5 × El Centro ground motion.888 1.057 0.7 -2.831 0.4 -6.26E-03 3. 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” 1.0 37.00E-03 5.00E+00 0.921 1.237 0.490 -11.478 0.13E-03 5.996 0.8 Table 4.436 1.156 -0.3 9.263 0.665 0.55E-03 3.55E-03 3.908 1.305 -0.831 0.0 37.00E+00 0.2 12.00E+00 0.2 16.603 -1.371 -0.114 1.8 0.5 54 .4 -4.00E+00 3 “Modes” 8.00E+00 Error (%) “Modal” Response 1 “Mode” -32.23E-02 1.00E+00 “Mode” 3 0.037 0.996 -0.0 -30.2 1.461 0.213 1.00E-03 5.176 0.04E-02 8.6 0.2 9.515 -50.190 -0.8 Peak values of hinge plastic rotations (radians) from MPA for 1.00E+00 3.5 -5.783 1.998 21.998 0.7 Peak values of story drift ratios (as % of story height) from MPA for 1.330 1.6 19.8 -32.5 3 “Modes” -32.5 “Mode” 3 -1.015 0.813 1.213 1.399 0.507 1.8 -32.23E-03 3.00E-10 NL RHA 1.3 13.Table 4.6 2.105 0.7 19.263 1.109 0.821 1.00E+00 0.983 1.5 0.130 0.858 2.257 0.00E+00 0.068 0.11E-03 9.434 0.037 -0.5 21.0 37.23E-03 3.00E+00 0.00E+00 0.00E+00 0.04E-10 3.213 1.687 0.3 -100.434 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.55E-03 3.2 0.7 16.530 1.5 × El Centro ground motion.17E-03 9.1 13.6 Peak values of floor displacements (as % of building height = 37.938 1.854 0.2 4.530 1.207 18.168 -0.00E+00 0.066 -0.00E+00 0.908 -1.434 0.2 4.5 1.35E-10 3.125 -1.00E+00 0.860 1.821 -1.5 10.2 -2.4 26.454 1.00E-03 5.822 1.913 7.3 -22.00E+00 0.19E-03 1.516 0.319 1.199 16.7 1.2 -0.953 15.00E-03 5.429 -1.018 -0.754 1.850 1.2 2.351 -0.603 1.88E-03 0.7 4.19E-03 1.2 -3.830 -12.1 4.00E+00 0.098 20. gravity loads included Hinge Plastic Rotation (rad) Floor “Modal” Response “Mode” 1 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 8.00E+00 0.00E+00 0.0 -30.414 28.23E-03 0.102 1.2 9.071 0.11E-03 9.00E+00 0.154 0.00E+00 2 “Mode” 8.11E-03 9.686 -7. 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” 0.0 -100.9 -4.514 -1.128 -1.4 1.3 -22.14 m) from MPA for 1.071 -0.429 1.5 21.8 1.673 Table 4.055 0.00E+00 0.3 -100.033 -0.1 21.23E-03 0.

joint rotations. j 2. 2… N ). 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. k = 2 for T1 ≥ 2. Specified in FEMA-273 are three distributions for lateral forces: 1.2 COMPARATIVE EVALUATION Compared in this section are the earthquake-induced demands for the selected building determined by five analyses: pushover analysis using the three force distributions in FEMA-273. and the exponent k = 1 for fundament period T1 £ 0. plastic hinge rotations.5. a plot of base shear versus roof displacement. computed at the target displacement represent the earthquakeinduced demands on the structure. 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. MPA considering three “modes. assumed to be linearly elastic. “Uniform” distribution: s* = m j (where the floor number j = 1. Comparison of Modal and FEMA Pushover Analyses 5.5 sec . gravity load effects were included in all 55 .5 sec . 1997] The pushover curve. story drifts.” and nonlinear RHA. 5. and 3. SRSS distribution: s* is defined by the lateral forces back-calculated from the story shears determined by response spectrum analysis of the structure. etc. The floor displacements. and varies linearly in between..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.

11 0. The first four floor displacements are within 5% of the “exact” value.042 0. Also included in these presentations are the MPA results considering three “modes” and the “exact” demands from nonlinear RHA.1 through 5. the floor displacement demands in Fig. The pushover curves are given in Fig.4.1.177 0.0381 0.0 cm. The MPA procedure is more accurate than all the FEMA force distributions. 0. and the displacements of upper floors are overestimated by 9-22%.126 0.00719 0. pushover analyses are implemented for a target roof displacement of 52. and the locations of all plastic hinges in Fig. Force distributions in FEMA-273: (a) “uniform”.4.0702 0. 5. Using each of these force distributions.3.281 0.0446 0. both presented in Section 4. the story drift demands in Fig. with the errors being larger for the SRSS distribution.11 0.11 0.4a. 5.3.062 0.3a and 5. 5.1. 5.11 0.3a and Table 5.1 demonstrate that the displacement demands are underestimated by the ELF and SRSS force distributions by 12 to 30 % at the lower six floors of the building. wherein the first two are obvious and the third is determined from response spectrum analysis of the building (Appendix C).0913 0. Figures 5.0466 0. The “uniform” distribution overestimates all floor displacements by 17-28%.165 0. 5. The three FEMA force distributions are presented in Fig. 5.4 and Tables 5. The errors in the FEMA and MPA estimates of seismic demands relative to the “exact” demands are presented in Fig.5 times the El Centro ground motion.analyses.0981 0.367 (a) Uniform (b) ELF (c) SRSS Fig.112 0.119 0. and Table 5.2.2.11 0.0896 0. 5.1.0654 0. (b) ELF.3.3b and Table 5. and (c) SRSS 56 .11 0. the value determined from RHA of the first-mode inelastic SDF system for 1.0197 0.21 0. plastic hinge rotation demands in Table 5.11 0.

(a) Uniform Distribution 12000 10000 u = 33 cm. V = 7456 kN. α = 0. 5.2.3 cm. V = 6897 kN.14 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized 10 20 30 40 Roof Displacement (cm) (b) ELF Distribution 50 60 12000 10000 u = 38. α = 0. (b) ELF. Pushover curves using three force distributions in FEMA-273: (a) “uniform”.8 cm. and (c) SRSS. α = 0. gravity loads are included 57 .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 = 8530 kN.18 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized 10 20 30 40 Roof Displacement (cm) (c) SRSS Distribution 50 60 12000 10000 u = 39.

5 1 1.5 1 1. Modal pushover analysis procedure gives estimates better than all the FEMA force distributions. reaching 64%. Heightwise variation of floor displacements and story drift ratios estimated using FEMA-273 force distributions.3. 5.3 demonstrate that the hinge plastic rotations estimated by all three FEMA force distributions contain unacceptably large errors. and overestimated by no more than 32%. For the uniform distribution. errors are largest in the upper stories.) The pushover analysis procedures considered seem incapable of computing accurately local response quantities. with story drifts under estimated by. reaching 35%. 7%. MPA including three “modes.4c and Table 5.3b and 5. For the ELF distribution. reaching 31%. at most. but it is still inaccurate. (a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th (b) Story Drift Ratios Floor 4th 3rd 2nd 1st NL−RHA MPA. such as hinge plastic rotations. (The 100% error in a 6th floor hinge is ignored as it simply represents that the MPA estimated zero rotation. and Table 5.” and NLRHA. the errors are largest in the lower stories. 3 "Modes" FEMA Uniform ELF SRSS 0. 3 "Modes" FEMA Uniform ELF SRSS 0. gravity loads included 58 . Figure 5. In contrast. For the SRSS distribution.5 Story Drift Ratio (%) 2 2. whereas nonlinear RHA computed an insignificantly small value.4b.5 Ground 0 Ground 0 Fig. with errors reaching 37% in this example. larger errors are noted in the upper and lower stories.Figures 5. This seems to be an inherent limitation of pushover analysis.2 demonstrate that the story drift demands are greatly underestimated by all the FEMA force distributions.5 Displacement/Height (%) 2 Floor 5th 5th 4th 3rd 2nd 1st NL−RHA MPA. the MPA procedure is more accurate than all the FEMA force distributions.

the SRSS distribution fails to identify yielding of beams in the middle floors. Based on the results presented here. it appears that structural performance evaluation should be based on story drifts that are known to be closely related to damage and can be estimated to a higher degree of accuracy by pushover analyses. the “uniform” distribution fails to identify yielding of the beams above the fourth floor. story drifts.” nonlinear RHA (“exact”). gravity loads included The structural engineering profession is now comparing these hinge plastic rotations against rotation limits established in FEMA-273 to judge structural component performance. 5.(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. and the three FEMA analyses. While pushover estimates for floor displacements are more accurate. and hinge plastic rotations estimated using FEMA-273 force distributions and MPA (including three “modes”).4. The locations of the plastic hinges are not identified correctly by the FEMA force distributions. The locations of plastic hinges shown in Fig. Errors in floor displacements. 5.5 were determined by five analyses: MPA considering the three “modes. they are not good indicators of damage. and the ELF distribution fails 59 .

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

8 -32.3 -11.436 Error (%) NL Uniform RHA 0.6 22.16E-03 0.5 61 .399 1.09E-03 4.372 1.51E-03 4.4 -50.3 23.667 0.9 SRSS -22.45E-03 3.736 0.007 1.84 1.490 0.1 0.673 Error (%) NL RHA 1.128 1.9 16.207 1.9 15.6 -41.34E-03 2.314 1.0 -100.323 1.953 0.26E-03 3.00E+00 0.1 10.098 1.78E-03 0.5 -3.94E-03 2.547 -27.5 -27.061 1.13E-03 5.6 -6.5 10.809 0.2 16.623 1.0 -63.53E-02 1.4 16.52E-04 1.083 1.8 2.0 37.927 1.4 -23.3 31.4 MPA -2.011 1. gravity loads included Displacement /Height (%) FEMA ELF SRSS MPA 1.0 -30.335 2.11E-03 9.724 0.7 7.294 1.00E+00 Error (%) FEMA Uniform ELF SRSS MPA 24.462 1.064 1.5 -32.399 1.2 -4.19E-04 5.50E-03 0.55E-03 3.03E-03 5.672 1.04E-02 8.992 1.318 2.466 0.089 1.8 -63.19E-03 1.7 MPA -2.209 0.8 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 0.178 1.562 1.168 1.3 -4.597 0.7 19.5 -33.2 -100.015 0.4 -55.9 11.8 7.351 0.783 1.2 16.6 17.1 -3.5 -33.23E-03 3.836 0.75E-03 0.7 Table 5.998 1.9 -70.00E-10 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 1.2 19.530 1.984 1.708 0.975 1.7 -29.154 1.2 4.9 -77.93E-03 1.839 0.6 -17.0 Table 5.888 0.306 1.5 Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 2.2 13.195 0.Table 5.2 6.0 -100.14 m) from FEMA force distributions and MPA.88E-03 1.00E+00 2.6 14.3 29.566 1.262 1.938 1.877 1.199 27.4 -12.1 163.4 16.00E+00 0.3 -14.5 22.7 -15.0 -59.7 26.00E+00 0.399 -27.998 1.234 1.7 -28.344 0.2 10.10E-02 7.9 16.00E+00 0.109 1.4 0.270 0. Peak values of floor displacements (as % of building height = 37.5 -29.8 44.7 -60.0 -100.8 -2.367 1.8 -100.417 1.62E-03 0. gravity loads included Displacement /Height (%) FEMA ELF SRSS MPA 0.487 0.8 17.353 1.00E+00 0.310 1.2 5.23E-02 1.4 21.2 -12.00E+00 NL RHA 1.6 4.00E-03 5.686 0.5 -29.35E-03 8.7 21.8 16.1 -26.3 Peak values of hinge plastic rotations (radians) from FEMA force distributions and MPA Hinge Plastic Rotation (rad) MPA FEMA ELF SRSS 4.6 -11.78E-03 1.9 0.1.17E-03 9.5 26.515 FEMA Uniform ELF 27.875 0.6 -73.00E+00 4.9 -20.2 9.341 1.560 1.854 0.3 -22.263 0.414 1.611 0.221 1.6 -4.9 28.00E+00 8.0 -71.00E+00 0.860 1.65E-03 7.58E-04 6.524 0.530 1.789 0.3 -100.6 -25.355 0.830 2.730 1.59E-03 5.0 -57.35E-10 3.6 21.2 9.5 15.7 FEMA ELF SRSS -22.214 1.0 -100.067 0.330 1.858 1.8 -100.8 -35.913 0.2 Peak values of story drift ratios (as % of story height) from FEMA force distributions and MPA.

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

5 3 Fig.(a) FEMA: Uniform 60 50 40 60 50 40 3 (b) FEMA: ELF Error Envelope Error for Floor No.5 1 1.5 2 GM Multiplier 2.5 3 0 0 0. gravity loads included 63 .6.5 2 GM Multiplier 2.5 2 GM Multiplier 2.5 1 1.5 2 GM Multiplier 2.5 3 (c) FEMA: SRSS 60 50 40 60 50 40 2 3 1 5 4 (d) MPA (3 "Modes") Error (%) 30 20 10 0 0 8 Error (%) 30 20 6 8 1 5 2 4 3 9 7 6 7 9 10 0 0 0.5 3 0.5 1 1. 5. Errors in floor displacements from three force distributions in FEMA-273 and from MPA including three “modes”. Noted 2 Error (%) Error (%) 1 2 30 20 10 0 0 1 4 6 7 30 20 10 8 3 5 8 4 9 7 9 6 5 0.5 1 1.

5 1 1.(a) FEMA: Uniform 100 60 50 9 8 (b) FEMA: ELF Error Envelope Error for Story No.5 3 Fig.5 2 GM Multiplier 2. 5.5 2 GM Multiplier 2.5 1 1.7.5 2 GM Multiplier 2. Noted 80 40 9 1 Error (%) 7 Error (%) 60 30 20 10 2 5 6 4 8 40 1 6 20 2 5 3 4 3 7 0 0 0.5 3 0 0 0.5 2 GM Multiplier 2.5 3 0 0 0.5 3 (c) FEMA: SRSS 60 50 40 60 50 40 3 2 1 (d) MPA (3 "Modes") Error (%) 30 20 Error (%) 30 5 4 8 9 20 9 6 7 7 10 0 0 4 10 2 3 1 8 5 6 0.5 1 1.5 1 1. gravity loads included 64 . Error in story drifts from three force distributions in FEMA-273 and from MPA including three “modes”.

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

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

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

they are not good indicators of damage. 68 . Instead. 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. with the excitation scaled to cover a wide range of ground motion intensities and building response. such as life safety and collapse prevention. This new method for estimating seismic demands at low performance levels.11. should obviously be evaluated for a wide range of buildings and ground motion ensembles. This report has focused on development of the MPA procedure and its initial evaluation in estimating the seismic demands on a building imposed by a selected ground motion. While pushover estimates for floor displacements are even more accurate. 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.

(1999). A. 6th U. Stanford University.P.M. G. Pushover analysis procedure in earthquake engineering. Proc. S. S. (2000). 29:1287-1305. J. Krawinkler.D. ASCE 123(1):3-10. Report No.H. Prentice Hall: New Jersey. University of California. and Seneviratna. (1997). Struct. P. Spectra. Kunnath. Fajfar.K. Kim.K. Earthq.. Earthquake Engineering Research Center. (2000). ASCE. I: Equivalent nonlinear system. M. Earthq. Engrg. Dyn. H. Seismic demands for performance evaluation of steel moment resisting frame structures (SAC Task 5. FEMA-273.3). A. and Powell. A. Story drift demands for steel moment frame structures in different seismic regions. Nat. DRAIN-2DX user guide. A. Pros and cons of a pushover analysis of seismic performance evaluation. Washington. Estimation of seismic drift demands for frame structures. (1998)..K. and Kunnath. and Gupta. 15:417-434. Method of reliability-based seismic design. J. and D’Amore. 9th World Conf.C.K. Struc.4. Engng..W. Engrg. Gupta. NEHRP Guidelines for the Seismic Rehabilitation of Buildings.. 5:111-116. D.. Gupta. and Reinhorn. J. 16(2):367-392 Han. 132.. Blume Earthquake Engineering Center. Y. Struct. Engrg. (1998). Calif. Earthq. 20(4-6):452-464. (1988). H. Berkeley. Struct. R. N2—a method for nonlinear seismic analysis of regular structures. and Wen. Japan. (1999).. Spectra. Tokyo-Kyoto. 69 . Gupta.K. UCB/EERC-88/06. S. Dynamics of Structures: Theory and Applications to Earthquake Engineering. and Fischinger. E. Calif. Chopra. Stanford. S. and Krawinkler. B. Building Seismic Safety Council (1997). Engrg. A. John A. on Earthq.. J. Seattle.S. (1988). Krawinkler. Engrg. H. (1997). and Krawinkler. (2001).M. Conf. 123:256-265. Bracci. Report No.. Earthq. H. Proc. G.. Adaptive spectra-based pushover procedure for seismic evaluation of structures. Seismic performance and retrofit evaluation for reinforced concrete structures. Federal Emergency Management Agency.7 References Allahabadi. Washington.

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

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

0. Abn .9. i 3. Calculate the yield displacement.9. A.4 to 3.9. 3.7. i i 3.2. i +1 i i 3. Calculate the post-yielding strain-hardening ratio. If desired.4. Replace i+1 with i and repeat Steps 3. Calculate area under the bilinear curve OAB.6 × Vbny . at base shear i equal to 0. i i i 3. Calculate the error = 100 × Abn − Apn ( ) ( ) ( ) ) Apn .1. Let the point with base shear = Vbny and roof displacement = urny be denoted as A. 3.6 × Vbny ( ) i urn . i i i αn = Vbno Vbny − 1 urno urny − 1 i 3. O.1. 3.5. This value. urn. If the error exceeds some pre-specified tolerance. and B with straight-line segments to obtain the idealized bilinear curve.6. obtained by judgment. by connecting a straight line between origin. Calculate the slope.4. i 3.2.6 . Vbny . i 3.6 × Vbny . 72 . urny = Vbny kn . determine the roof displacement. other appropriate methods can be ( used. corresponding to the estimated yield i i i base shear.0.6 . iterations are necessary.4. will be refined by an iterative procedure that seeks to equate areas under the actual and the idealized pushover curves.3.i 3. Draw the curve OAB by connecting the three points O.8. Calculate initial slope of the idealized bilinear curve. Vbny . From the pushover data. and a point on the actual pushover curve with base shear equal to i 0. Estimate the yield base shear.8. This step gives the secant stiffness at a base shear equal to 60% of the yield base shear. kn . Calculate Vbny = Vbny × Apn Abn . kn = 0.

10b and 4. Combine the “modal” responses using Eqs. (3. Calculate histories of various responses using Eqs. A. 8. Develop the Fsn Ln − Dn relation (Fig.10a and 4. 4. Repeat Steps 3 to 6 for as many modes as required for sufficient accuracy. Scale the horizontal axis by Γnφrn to obtain Dno = urno Γ nφrn and Dny = urny Γ n φrn (Eqs. In general first two or three modes will suffice. 4.2.6 ur n y ur n o ur n Fig. Compute deformation history. (3.3. 5.1. An (t ) .15) and (3.4.1 Idealization of nth-“mode” pushover curve 73 . 6. 4.6 × Vbny A • Idealized αnkn B • 1 Actual k 1 • O n figA_1. 7. A. Calculate peak values. Compute the Ln and Γ n from Eq. (3. Vbn Idealization of Pushover Curve Vbno Vbny 0.11b).13). Dn (t ) . A. * 4.3b) with unit-mass and force-deformation relation of Fig. Scale the vertical axis with * Mn to obtain * Fsno Ln = Vbno M n and * Fsny Ln = Vbny M n (Eqs.2. and pseudo-acceleration history. 4. r o .11a).eps ur n. of the combined responses obtained in Step 8. of the nth“mode” inelastic SDF system (Fig. 4.4) and effective modal mass from M n = Ln Γn .16).2). 9.0.12) and (3.

2 Properties of the nth-“mode” inelastic SDF system A. First three mode shapes and frequencies of the selected building were computed and are shown in Fig. Ap1 = 360777 kN-cm. to the north-south component of the El Centro (1940) ground motion scaled up by a factor of 1. respectively.eps Dny = ur n y / Γnφr n Dno = ur n o / Γnφr n Dn Fig. Target displacements used to generate these pushover curves are 63. The pushover curves for the first three modes.).2 EXAMPLE The UMRHA procedure is implemented to calculate the response of the 9-story building described in Section 3. Area under the actual pushover curve.1. 3.2. 3.5 cm and Vb1o = 8729. A.2. 3. B. 3. generated using DRAIN-2DX. computed for the first three modes are shown in Fig. and 12. for the first. 2.4 cm (10 in.7 cm (5 in.Fsn / Ln Fsn / Ln − Dn Relationship Vbno / M* n Vbny / Mn * 1 α ω2 n n ω2 1 n figA_2. The anchor point. The base-shear – roof-displacement (Vbn .3.4.urn ) pushover curve for the force distribution s* : n 2.).1. 25. Idealized bilinear curves for each of the three modes are included in Fig.3.1.5 cm (25 in. The force distributions. and third mode. s* . is defined at the target roof displacement. second. The following steps illustrate the procedure to develop the idealize curve for the first “mode”. ur1o = 63. 74 . A. n 2. 3.).3.6 kN. are shown in Fig. A. At this point. Following is step-by-step implementation of the procedure described in Section A. 1.5.4.

and α1 = 0. i 3.2.9.1. The Fs1 L1 − D1 relation for the first “mode” is developed as follows. i i i 3. 3.86 cm at 0.1.4.3. The initial slope of the idealized bilinear curve. α1 = Vb1o Vb1 y − 1 ur1o ur1 y − 1 = ( ) ( ) (8729. Ab1 = 365100 kN-cm.9 kN. This value exceeds the prespecified tolerance of 0.3666.4 × (360777 365100 ) = 7911.4 ) − 1 (63.6 8006. ur1. 3.18 kN/cm.4 kN. Therefore.8 kN. The post-yielding strain-hardening ratio.9. i +1 3.6 × Vb1 y ( ) i ur1. The first estimate of the yield base shear Vb1 y = 8006.6 kN.01%.1. The procedure converged after nineteen cycles to give ur1 y = 36. A.198%.0. L1 = 2736789 kg.86 = 210. and M1 = 2736789 × 1.2.23 cm.194. ur1 y = Vb1 y k1 = 8006. i 3. i i 3. Determined from the pushover database.6.6 = 22. k1 .i 3.4.6 × Vb1 y = 4803.8 22.1.2.09 cm. The point A on the i i bilinear curve is defined by ur1 y = 38.13).9. * 4. i i i 3. The results for other modes are summarized in Table A. The next estimate of the yield shear is Vbny = 8006.8. The yield displacement. is calculated as follows.09 cm and Vb1 y = 8006.135. (4.5 38. 3. iterations are necessary. and B with straight-line segments gives the idealized bilinear curve. 75 . Error = 100 × (365100 − 360777 ) 360777 = 1. 4. The results of the iterative procedure are summarized in Table A. Also included are the modal damping ratios and the periods calculated from Eq.7.5.0.4 kN.4 210. The curve OAB obtained by connecting the three points O. Γ1 = 1.3666 = 3740189 kg. i i 3.18 = 38.4. Area under the bilinear curve OAB. Vb1 y = 7615.6 = 4803. k1 = 0.09 ) − 1 = 0.

The peak values are also plotted in Fig.4. A.4. 4.7. Histories of roof displacement and top story drifts for the first “mode” are computed and presented in Fig.2. The peak values are computed and are summarized in Tables 4. Deformation and pseudo-acceleration histories of the inelastic SDF systems for the first “mode” with unit mass and force-deformation relation developed in Step 4 are plotted in Fig. 4.62 (cm/sec2).51 cm. Scaling the vertical axis by Fs1 y L1 = 203. 6.7. 4. * M1 gives Fs1o L1 = 233. 7.7.2.40 (cm/sec2) and 76 . The combined modal responses are presented in Fig. 9.46 cm and D1y = 26. The results were generated for first three “modes’ and are included in Fig. 5. 8. 4. Scaling the horizontal axis by Γ1φr1 gives D1o = 46.8.3.1 and 4. 4.

(a) "Mode" 1 Pushover Curve 12000 10000 u = 36.3.eps 0 0 2 4 6 8 10 Roof Displacement (cm) 12 Fig. V = 7616 kN. V = 5210 kN. A. α = 0. α = 0.9 cm.14 y by 6000 Base Shear (kN) 4000 2000 Actual Idealized figA_3c. α = 0.6 cm.2 cm.19 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized figA_3a.eps 0 0 5 10 15 20 Roof Displacement (cm) (c) "Mode" 3 Pushover Curve 25 8000 u = 4.13 6000 Base Shear (kN) 4000 2000 Actual Idealized figA_3b. V y by = 4952 kN.eps 10 20 30 40 50 Roof Displacement (cm) (b) "Mode" 2 Pushover Curve 60 8000 u = 9. “Modal” pushover curves for the example building 77 .

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) • 22.5 × El Centro ground motion for the first “mode.747 5 10 15 20 Time (sec) figa_4. Histories of deformation and pseudo-acceleration due to 1.” second “mode.52 5 10 15 20 Time (sec) 25 30 0 • 1.” and third “mode” inelastic SDF systems.06 0 • 1. A. 78 .(a) Deformation 80 n=1 3 (b) Pseuso−acceleration n=1 • 0.4.m −80 0 −3 0 25 30 Fig.2769 D1 (cm) 0 • 35.

192 0.18 210.017 0.64 37.30 37.4 7672.85 0.18 210.18 210.51 2 “Mode” 2 -920860 -0.18 210.3 7658.7 7639.022 0.910 0.2671 1.180 0.4 7911.948 ζ n (%) 79 .194 (kN-cm) 365100 364060 363276 362684 362235 361892 361631 361432 361279 361162 361073 361004 360951 360911 360880 360856 360838 360824 360813 Table A.013 0.9 4570.3666 3740189 203.6 × Vb1 y i ur1.18 (cm) 38.3 7628.182 0. Results of iterative procedure to develop the idealized bilinear curve for the first “mode” inelastic SDF system Itr.18 210.56 19.6 i k1 i ur1 y i α1 i Ab1 Error (%) 1.4 4595.8525 1.44 36.082 0.18 210.35 36.18 210.063 0.6 4583.18 210.9 7615.1 4574.0 4588.194 0.5 (cm) 22.75 21.5 3109.8 4747.78 21.39 36.237 0.529 0.74 (kN/cm) 210.0.309 0.05 52.8 4647. i 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 i Vb1 y i 0.4 7647.4927 1.7 4580.693 0.59 22.170 0.46 2.23 22.151 0.86 22.190 0.85 36.0 7619.23 0.107 0.18 210.56 47.29 36.3 7786.029 0.18 210.79 21.198 0.2 4628.176 0.18 210.Table A.26 36.193 0.18 210.0 4704.2.74 21.0 4577.1.4 7714.3 7745.404 0.95 21.037 0.62 26.32 36.2406 167531.76 21.18 210.135 0.11 22.5 7624.194 0.9 4573.2 4671.191 0.18 210.193 0.24 36.09 37.1 7616.139 0.38 22.05 36.5309 488839.194 0.5 7633.136 Fsny Ln (cm/sec2) Dny (cm) Fsno Ln (cm/sec ) Dno (cm) Tn (sec) 233.50 36.9 (kN) 4803.1 1013.103 “Mode” 3 696400 0.28 36.010 (kN) 8006.8 7618.186 0.18 210.77 21.02 21. No.2 4571.12 3876.048 0.25 36.86 21.59 36.3 4603.75 21.65 1226.40 46.70 36.09 18.18 210.193 0.25 36.79 0.5 4614.18 210.1 4569.8 7622.83 21.162 0.6 7840.2 7690.18 210.184 0.9 4570.188 0.90 21. Properties of “modal” inelastic SDF systems Properties Ln (kg) Γn * M n (kg) “Mode” 1 2736789 1.81 21.

80 .

1. Dn . At urno . This is the same example as solved in Appendix A. (3. 10. subtract the yield hinge rotation to determine the plastic hinge rotation. 11.21). the first two or three “modes” will suffice. or from the inelastic response (or design) spectrum.8).5. Compute the peak deformation. Following is step-by-step implementation of the procedure described in Section B. Calculate the peak roof displacement urno associated with the nth-“mode” inelastic SDF system from Eq.3b) with unit mass and force-deformation relation of Fig. 4. (4. 4.4. Steps 1 to 4 of the MPA are the same as those for UMRHA presented in Appendix A.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.Appendix B Modal Pushover Analysis B. extract from the pushover database values of other desired responses. Repeat Steps 3 to 7 for as many “modes” as required for sufficient accuracy. rno .33 cm. 81 .18).1 to the north-south component of the El Centro (1940) ground motion scaled up by a factor of 1. Determine the total response by combining the peak “modal” responses using the SRSS combination rule of Eq.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. 6. From the total hinge rotation. Typically. B. Solving Eq.6b by solving Eq.2 EXAMPLE The MPA procedure is implemented to calculate the response of the 9-story building described in Section 3. 13. (3. of the nth-“mode” inelastic SDF system (Fig. 12. 10. (4.

(3.3 for the floor displacements normalized by the building height and in Table 4.4.11. Also included in Table 4.3 and 4. values of floor displacements and story drifts are extracted.” and the results are included in Tables 4.4 for the story drifts normalized by the story height. Steps 3 to 7 are repeated for first three “modes.1. At ur1o = 48. 9.366 × 1 × 35.33 = 48. The total response computed by combining the peak “modal” responses using the SRSS combination rule of Eq. 8. 12.28 cm.28 cm. Results of Steps 5 and 6 are summarized in Table B. The values are summarized in Table 4.18) are also included in Tables 4.5 are the plastic hinge rotations computed by subtracting the yield hinge rotation from the total hinge rotations.4. Peak roof displacement ur1o = Γ1 × φr1 × D1 = 1.3 and 4. 82 . where results for other ground motion intensities are also included.

367 1.312 1.50 35.8451 5.748 63.184 0.117 5.676 6.36 1.735 3.38 22.37 57.52 0.154 78.71 1.06 1.5 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 2.33 1.660 14.139 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.379 21.901 8.755 27.755 0.229 8.007 36.0 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 1.969 0.275 1.504 18.0 µ = Dn Dny urno = Γnφrn Dn (cm) 83 .678 0.28 46.Table B.690 10.18 27.551 2.85 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 1.332 13.03 0.13 2.07 “Mode” 2 4.27 0.450 4.395 0.436 7.268 0.05 0.03 26.252 9.513 0.766 7.52 “Mode” 3 1.59 0.25 Quantity Dn (cm) “Mode” 1 6.267 5.35 0.535 14.577 16.691 0.70 0.332 48.185 11. Calculation of roof displacements urno from peak deformation of inelastic SDF systems Ground Motion Multiplier 0.1.25 20.457 12.856 31.225 2.35 1.4222 3.467 14.75 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.73 24.200 0.023 0.37 1.126 13.913 22.5 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.82 1.0 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 3.79 0.

84

**Appendix C FEMA Force Distribution Calculations
**

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

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

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

Floor, j

mj (10-3×kg)

s* = j

∑i mi

0.112 0.110 0.110 0.110 0.110 0.110 0.110 0.110 0.119

mj

1 2 3 4 5 6 7 8 9

503.5 494.7 494.7 494.7 494.7 494.7 494.7 494.7 534.1

C.2

EQUIVALENT LATERAL FORCE (ELF) DISTRIBUTION 85

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

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

Floor j 1 2 3 4 5 6 7 8 9

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

-3 k

s* j

=

m jhk j

∑i mihik

0.007 0.020 0.038 0.062 0.091 0.126 0.165 0.210 0.281

C.3

SRSS DISTRIBUTION

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

N

∑ n (V jn )

2

to get column 8 of

8 -326.4 1759.1 -525. the lateral forces are normalized by the base shear to obtain column 10 in Table C.0 1476.6 1233.2 fj = ∑ fi i fj (2) (3) (4) 59.2 97.065 0.1 87.2 277.8 374.7 (10) 0.3.7 1578.3 -6.042 0.0 176.For convenience. Table C.2 285.047 0.7 355.0 136.5 159.6 -732.1 -438.2 Lateral Force fj (kN) (9) 203.5 -320.6 V j3 (kN) (7) 478.0 1231. FEMA 273 SRSS force distribution Lateral Forces Floor j (1) 1 2 3 4 5 6 7 8 9 f j1 (kN) f j2 (kN) f j3 (kN) V j1 (kN) (5) 1917.0 354.090 0.7 1622.7 694.6 319.367 87 .1 832.2 105.1 Story Shears V j2 (kN) (6) 1114.5 -973.7 374.4 250.5 215.6 -359.098 0.6 366.2 148.9 1683.7 -46.177 0.1 -967.045 0.2 200.3 222.0 381.0 980.7 234.6 -352.1 -646.4 1842.9 1446.7 2065.3 240.9 832.7 Vj (kN) (8) 2268.7 101.9 366.9 880.9 446.1 1857.9 -166.5 -350.9 -153.070 0.2 95.5 320.6 286.0 -5.3.3 -646.7 525.8 1381.8 430.4 400.7 95.

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

a research report modal pushover analysis

a research report modal pushover analysis

- Pushover Analysis using ETABS and SAP2000
- Pushover analysis
- Pushover Analysis Final
- EVALUATION OF PUSHOVER ANALYSIS PROCEDURES
- Pushover Ppt
- A comparison of single-run pushover analysis techniques for
- Non Linear Analysis Pushover
- Nonlinear Pushover Analysis of Reinforced Concrete Structures
- 08_EC8-ReLUIS_2
- 1-s2.0-S1877705813003858-main
- 3 General Requirements
- WCEE2012_4135 DBDSOFT
- Non-Linear Static Push Over Analysis
- C Earthquake Engineering
- Performing Pushover Analysis in Etabs
- Structural Dynamics
- A Modal Pushover Analysis Procedure to Estimate
- 2013 Pushover Analysis Guide Final RC Building With Masonry Infill Walls EditedDraft January 2013
- Pushover
- Performance Based Seismic Design
- nistgcr11-917-11
- Performance Based Design Using Nonlinear Analysis
- Pushover Analysis Using Etabs
- Action Plan for Performance Based Seismic Design
- ETABS Pushover Analysis
- SAP-Etabs-Course-360
- Comp Slope Fredlund
- Meland Afmc 16 07
- Controlling Physics-Based Characters Using Soft Contacts
- Behavior of Braced Frames

Are you sure?

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

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

Get the full title to continue

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

scribd