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

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

Pacific Earthquake Engineering Research Center

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

Theory and Preliminary Evaluation

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

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

PEER 2001/03

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

Anil K. Chopra University of California Berkeley

Rakesh K. Goel California Polytechnic State University San Luis Obispo

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

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

ii .

The peak deformation of this SDF system is used to determine the roof displacement. 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. such as hinge plastic rotations.mι ug t . Comparing the earthquake-induced demands for the selected 9-story building determined by pushover analysis using three force distributions in FEMA-273.25 to 3. Instead. which retains the conceptual simplicity and computational attractiveness of current procedures with invariant force distribution. 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. where m is the n mass matrix and φ n its nth-mode. in the modal expansion of the effective earthquake forces. 2. The standard response spectrum analysis (RSA) for elastic buildings is reformulated as a Modal Pushover Analysis (MPA). is determined by combining the rno n = 1. and the structure is pushed to the roof displacement determined from the peak deformation Dn of the nth-mode elastic SDF system. ro . and identifies locations of most plastic hinges.0. at which the seismic response. is determined by pushover analysis. Second. the total demand. it is demonstrated that the FEMA force distributions greatly underestimate the story drift demands. However. a pushover analysis determines the peak response rno of the inelastic MDF system to individual modal terms. rno .urn curve is developed from a pushover analysis for force distribution s* . 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 .… according to an appropriate modal combination rule. Thus the MPA procedure is accurate enough for practical application in building evaluation and design. iii . p eff t = . p eff .ABSTRACT The principal objective of this investigation is to develop a pushover analysis procedure based on structural dynamics theory. MPA. 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. 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. The results presented for El Centro ground motion scaled by factors varying from 0. all pushover analysis procedures considered do not seem to compute to acceptable accuracy local response quantities.sn ug t . plastic hinge rotations are less accurate. but provides superior accuracy in estimating seismic demands on buildings. The MPA procedure is extended to estimate the seismic demands for inelastic systems: First. The base shear-roof displacement Vbn .n t = . This pushover curve is idealized as bilinear n and converted to the force-deformation relation for the nth-“mode” inelastic SDF system. and the MPA procedure is more accurate than all the FEMA force distributions methods in estimating seismic demands. Combining these peak modal responses by modal combination rule leads to the MPA procedure. and nonlinear RHA.

iv .

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

vi .

................................... Comparison of Modal and FEMA Pushover Analyses...............34 4..............................................3 3.....13 3.........................55 vii ..28 4...................................................................................................... Equation of Motion ............................................2 2...........4........................3 Modal Pushover Analysis with Gravity Loads ...............2 Response History Analysis ........................................................................................................................3 Modal Pushover Analysis ..........................................36 Modal Pushover Analysis ...................................................................................................3 Summary...................................... iii Acknowledgment ........1 3.....................................55 5.........27 4...........................3 2........33 4..........................................2 Response History Analysis ...........................................4...........................4...47 4..........1 4...............................38 4.37 Comparative Evaluation of Analysis Procedures ............................................................... vii 1...............................................................................................................................................................6 Pushover Analysis .................................................5 Response History Analysis ............................................................4 3......................27 Uncoupled Modal Response History Analysis ..............................................4 Modal Response History Analysis........................................13 3.................................................................................................................2.............................................22 4..........................................................1 System and Excitation Considered ......................................................3...................1 Uncoupled Modal Response History Analysis ............... Inelastic Multistory Buildings ...........................................12 Modal Pushover Analysis .............38 4.......................................2 Properties of nth-“mode” Inelastic SDF System ......................................................................................................2 3..1 One-Story Systems .......3 System and Excitation Considered ......................................1 5....................4...1 Summary..2 FEMA-273 Pushover Analysis ........................................................12 Comparative Evaluation of Analysis Procedures .........................................................36 4.... 2........................................4....................2 Modal Pushover Analysis .................................4 5..............................................................................4.v Table of Contents........... Introduction.................CONTENTS Abstract .....................................................................................................................................2.....................................................17 3......................................3 2.................................................................................................55 Comparative Evaluation.2...........................9 Modal Response Spectrum Analysis ........1 Underlying Assumptions and Accuracy .3 4......................................................................41 4...........................................................................................1 2.............6 Elastic Multistory Buildings ...................................................................9 3..

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

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

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

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

(a) Idealized one-story structure.e. sign u = . m depends on three system parameters: w n . 2001.1) is divided by m to obtain ~ u + 2zw n u + w 2 u y f s u. This becomes evident if Eq. 2.fS m fs c ut u fy 1 αk k 1 uy ug -fy k k um u 1 1 (a) (b) Fig. and u y .1. and R y (Chopra .ug t n where bg bg b g bg (2. z y .. u £ u y ). (2.3). z .4) and w n is the natural vibration frequency. bg 4 . in addition to the form of the force-deformation relation. or absolute (without regard to algebraic sign) maximum. and (b) bilinear hysteretic forcedeformation relation For a given excitation u g t . the deformation u t depends on three systems parameters: w n .5) For a given u g t .3) ωn = k m ζ = c 2mω n f fs = s fy (2. and z is the damping ratio of the system vibrating within its linear elastic range (i. and the ductility factor is u m= m uy (2. The peak. Section 7. deformation is denoted by um . Tn = 2p w n is the natural vibration period.

Vb (kN) 30 20 10 0 0 2 4 6 Roof Displacement. 6.2.1. based on Ry = 8 . with Tn = 0. 2001) and scaled up by a factor of 2. is f y / w = ( f o / w ) ÷ 8 = 0.5 sec and z = 5% subjected to the north-south component of the El Centro (1940) ground motion (Fig.32 m • Fig.2 kips) . 2.2 SYSTEM AND EXCITATION CONSIDERED Consider the one-story system in Fig. 2. The yield strength of the inelastic system.2: the dimensions and flexural rigidity of the structural elements are noted.3. One-story.376 cm.2 5 .9 kN (38. 2. this system and excitation.2. α = 0.66 m • EIb EIc EIc • • L = 7.826 kips) for w = 169.2311 . For .04 40 Base Shear. u (cm) 8 Fig. f o w = 184 . and f y = 39. Vby = 39. Pushover curve for structure shown in Fig. 2.26 kN (8.26 kN. m h = 3.4 in Chopra. one-bay frame 50 uy = 1.

1 kip-in.g.4edetermined from RHAis superimposed. (d) shows the rotation q p bt g of the plastic hinges at the bg bg beam ends. 2. The resulting pushover curve is shown in Fig.3 in. respectively. 2. as apparent in Fig. This pushover curve turns out to be bilinear because the beam and the columns are designed to yield simultaneously when f s reaches f y . and E = 2 × 108 kPa (29×103 ksi) gives yield moments of 21.The yield moments in the beam and columns are defined as the bending moments due to the lateral force f y .134 × 107 mm 4 (75.4 shows the earthquake response of the system described in the preceding section determined by response history analysis (RHA).0217 rad .) for the beam and columns. A static pushover analysis of this one-story system leads to the force-displacement relationship shown in Fig.. and (e) shows the force-deformation relation. 2. I b = 3.) and 50. (c) shows the joint rotation q bt g . Observe that the pushover curve matches the initial loading path of the hysteretic system. the energy dissipated in 6 .6 kip-in. The peak values of the various response quantities are as follows: um = 7. pushover analysis cannot provide any cumulative measure of response. and q pm = 0. e. The system is excited well beyond the yield deformation. the peak value determined from RHA.35.077 × 107 mm 4 (146 in. 2.4f.3.65 kN-m (191. the joint rotation and beam hinge rotation are identical to values q m and q pm determined from RHA.4 ). wherein the hysteretic force-deformation history of Fig.36 cm .4 PUSHOVER ANALYSIS Static analysis of the one-story nonlinear system subjected to lateral force that increases in small increments is implemented until the lateral displacement reaches um = 7. q m = 0. The yielding stiffness of each structural element is defined as 3% of its initial stiffness. Implementing this analysis with I c = 6. Determined from pushover analysis at the exact peak deformation.4e.18 kN-m (444. (b) shows the lateral force f s t or base shear Vb t normalized relative to the bg weight.017 rad . However.3 RESPONSE HISTORY ANALYSIS Figure 2. the ductility factor m = 5. It is organized in five parts: (a) shows the deformation u t .4 ). 2. 2.36 cm.

(b) base shear. 2.017 • 0 (d) θp (rad) −0.2 0 (e) 0 (f) −0.4 −10 −5 0 u (cm) 5 10 Fig.yielding during the ground motion.0217 • θ (rad) 0 (c) −0.4 0. Response of one-story system to El Centro ground motion: (a) deformation.2311 y (b) −0.2 Vb / w 0.4 −10 −5 0 u (cm) 5 10 −0.2311 (a) Vb / w 0 f / w = −0.4 0.04 0. (e) force-deformation relation. and (f) pushover curve 7 . (c) joint rotation.2 −0. This represents an inherent limitation of pushover analyses.5 0. 15 u (cm) 0 • 7.04 0 5 10 15 Time (sec) 20 25 30 0.04 0.2 −0. or the cumulative rotation at a plastic hinge.36 −15 0. (d) plastic hinge rotation.4.04 0.5 fy / w = 0.

8 .

mι ug t bg bg bg (3.3. and k are the mass.snug t n= n =1 The effective earthquake forces can then be expressed as af N af N af (3. (3. c. each element of the influence vector ι is equal to unity. Elastic Multistory Buildings 3. m. This force distribution can be expanded as a summation of modal inertia force distributions s n (Chopra.n t = Â .1) where u is the vector of N lateral floor displacements relative to the ground. and p eff t = Â peff. 2001: Section 13.m ι u g t bg bg (3.4) 9 .3) where φ n is the nth natural vibration mode of the structure.12): mι = n =1 Â sn = N n =1 Â Γ nmφ n N (3. 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 = . and lateral stiffness matrices of the system.2) The spatial distribution of these “forces” over the height of the building is defined by the vector s = m ι and their time variation by u g t .1) can be interpreted as effective earthquake forces: p eff t = . The right side of Eq.

7) By utilizing the orthogonality property of modes.5) The contribution of the nth mode to s and to p eff t are: sn = Gn mφ n respectively. an SDF system with vibration propertiesnatural frequency w n and damping ratio z n of the nth-mode of the MDF system.12) . Then the floor displacements are u n t = φ n qn t bg bg bg bg (3.11) and substituting in Eq.9) in which w n is the natural vibration frequency and z n is the damping ratio for the nth mode.9) to the equation of motion for the nth-mode elastic SDF system.8) where the modal coordinate qn t is governed by 2 qn + 2z nw n qn + w n qn = . The equations governing the response of the system are mu + cu + ku = .8) gives the floor displacements u n t = Gnφ n Dn t bg (3. with no contribution from other modes.sn ug t bg bg bg bg (3.9) and (3.sn ug t bg (3. p eff . we will outline that the response of the MDF system to p eff .6) Next. it can be demonstrated that none of the modes other than the nth mode contribute to the response.Gn ug t (3. subjected to u g t : Dn + 2z nw n Dn + w 2 Dn = .u g t n Comparing Eqs.10) gives qn t = Gn Dn t bg bg bg (3.n t = . The solution qn t can readily be obtained by comparing Eq. (3.10) bg bg bg 10 (3.n (t ) = N n =1 ∑ −sn u g (t ) N (3. (3. (3.peff (t ) = n =1 ∑ peff .n t is entirely in the nth-mode.

13) represent the response of the MDF system to p eff.15) rt = bg n =1 Â rn t = bg n =1 Â rnst An bt g N (3. 3. Section 13. internal element forces. Therefore. and Eqs. 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 (3.1).13) define the contribution of the nth-mode to the response. Equations (3. etc.12) and (3.16) This is the classical modal RHA procedure wherein Eq.n t . ζn ¨ g(t ) u (a) Static Analysis of Structure (b) Dynamic Analysis of SDF System Fig. st The two analyses that lead to rn and An t are shown schematically in Fig.Forces sn An(t ) st rn ωn.12) and (3.13) st where rn denotes the modal static response. 2001. (3. these standard equations have been derived in an unconventional way. However. 3.16) reflect combining the response contributions of all modes.9) is the standard modal equation governing qn t . In contrast to the classical 11 bg .15) and (3.1. Conceptual explanation of modal response history analysis of elastic MDF systems Any response quantity r t story drifts.can be expressed by st rn t = rn An t bg bg bg bg bg (3. the static value of r due to external forces sn .14) is the pseudo-acceleration response of the nth-mode SDF system (Chopra.1. Eqs. (3. (3. and An t = w 2 Dn t n bg (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. Chopra. This concept provides a rational basis for the modal pushover analysis procedure developed later. the peak value rno of the nth-mode contribution rn t to response r t is determined from st rno = rn An bg bg bg (3. The SRSS rule..1.18) To develop a pushover analysis procedure consistent with RSA. we used the modal expansion of the spatial distribution of the effective earthquake forces. the peak nth-mode response as in Eq.g. we note that static analysis of the structure subjected to lateral forces f no = Γ n mφ n An (3.9) . provides an estimate of the peak value of the total response: b g F N r 2 I 1/ 2 ro ª G Â no J H n =1 K 3. 2001. z n of the pseudo-acceleration response (or design) spectrum for the nth-mode SDF system. (3.derivation found in textbooks (e. and Tn = 2p w n is the natural vibration period of the nth-mode of the MDF system.16). which is valid for structures with well-separated natural frequencies such as multistory buildings with symmetric plan.3). this response value can be obtained by static analysis of the structure subjected to lateral forces distributed over the building height according to 12 .17) where An is the ordinate A Tn . Alternatively. (3.(3.8.1). Sections 12.3 MODAL PUSHOVER ANALYSIS (3. 3. Section 13.17) (Chopra. The peak modal responses are combined according to the Square-Root-of-Sum-ofSquares (SRSS) or the Complete Quadratic Combination (CQC) rules. In such a response spectrum analysis (RSA).19) will provide the same value of rno .4 and 13. 2001.

Third St. (3. The building has a basement level. 3.15 m (30 ft) on center. CA 94804-4698.18) to obtain an estimate of the peak value ro of the total response. and 37. denoted B1.2.4. region. urno . 46th Street. The columns are 345 MPa (50 ksi) steel wideflange sections.4 COMPARATIVE EVALUATION OF ANALYSIS PROCEDURES 3. in both directions. 1301 S. with five bays each in the north-south (N-S) and east-west (E-W) directions. Consulting Structural Engineers. 3. this structure meets seismic code and represents typical medium-rise buildings designed for the Los Angeles.2). 1660 W.s* = mφ n n (3. The levels of the 9-story building are numbered with respect to the ground level (see Fig.19 m (122 ft) in elevation.1 System and Excitation Considered The 9-story structure. shown in Fig. The interior bays of the structure contain simple framing with composite floors. This modal pushover analysis (MPA) for linearly elastic systems is equivalent to the well-known RSA procedure (Section 3. which from Eq.73 m (150 ft) by 45. this building is 45. The bays are 9.. Richmond.21) 2 where Dn = An ω n . rno . was designed by Brandow & Johnston Associates1 for the SAC2 Phase II Steel Project. Typical floor-to-floor heights (for analysis purposes measured from center-of-beam to center1 2 Brandow & Johnston Associates. SAC Steel Project Technical Office. and California Universities for Research in Earthquake Engineering (CUREE). the peak value of the roof displacement due to the nth-mode. Obviously Dn and An are available from the response (or design) spectrum. 13 . The peak modal responses. Although not actually constructed.12) is urno = Gnf rn Dn (3. (3. California.20) and the structure is pushed to the roof displacement. 3.2). each determined by one pushover analysis. the Applied Technology Council (ATC). SAC is a joint venture of three non-profit organizations: The Structural Engineers Association of California (SEAOC). Los Angeles. A benchmark structure for the SAC project. 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. CA 90017.73 m (150 ft) in plan. with the ninth level being the roof. can be combined according to Eq.

07×106 kg (73.01×106 kg (69. roofing and a penthouse located on the roof. i. The strength. The first three vibration modes and periods of the building for linearly elastic vibration are shown in Fig. for the second through eighth levels is 9. we select a weak ground motion: the northsouth component of the El Centro (1940) ground motion scaled down by a factor of 4. The seismic mass of the ground level is 9. such as those described in Gupta and Krawinkler (1999) can be used. if desired more complex models.65 m (12 ft) and for the first floor is 5. third.96 m (13 ft).of-beam) are 3.49 m (18 ft). 3.4. 3. ceiling/flooring. The building is modeled in DRAIN-2DX (Allahabadi and Powell. The simple model adopted here is sufficient for the objectives of this study.2 kips-sec2/ft). fifth. mechanical/electrical.20). 14 . for the first level is 1. including the steel framing. The seismic mass of the above ground levels of the entire structure is 9. respectively. s* (Eq.. The column bases are modeled as pinned and secured to the ground (at the B-1 level). The floor-to-floor height of the basement level is 3.0 kips-sec2/ft). The 9-story N-S MRF is depicted in Fig. The seismic mass of the structure is due to various components of the structure. for the first three modes are shown in Fig.3.65×105 kg (66. and 0. The floor system is composed of 248 MPa (36 ksi) steel wide-flange beams in acting composite action with the floor slab. 3. partitions.89×105 kg (67. Each frame resists one half of the seismic mass associated with the entire structure.00×106 kg (616 kips. are located on the first. To ensure that this structure remains elastic.27.49 sec. floor slabs. This model is based on centerline dimensions of the bare frame in which beams and columns extend from centerline to centerline. and shear distortion of panel zones are neglected but large deformation (P∆) effects are included. 0. The column lines employ two-tier construction.7 kips-sec2/ft). These force n distributions will be used in the pushover analysis to be presented later.e. the vibration periods are 2. Concrete foundation walls and surrounding soil are assumed to restrain the structure at the ground level from horizontal displacement. dimension. monolithic column pieces are connected every two levels beginning with the first level. 3. The force distributions.83 m (6 ft) above the center-line of the beam to column joint.85. and seventh levels at 1.sec2/ft). Column splices.2. 1988) using the M1 model developed by Krawinkler and Gupta (1998). which are seismic (tension) splices to carry bending and uplift forces. and for the ninth level is 1.0 kips-sec2/ft).

. 2000] 15 . 3.2. Nine-story building [adapted from Ohtori et al.Fig.

728 2.4.5 Mode Shape Component 1 1.03 −1. 3. and 3 n 3.12 0.37 2. (b) shows the base shear Vbn t normalized relative to the weight W of the building. 3.6. and 3. (c) shows the joint rotation q n t of an external joint at 16 bg bg bg .13)].05 s * 1 s * 2 s * 3 Fig. First three natural-vibration periods and modes of the 9story building 3.33 2. respectively.9th 8th 7th 6th T = 0.31 −0.13 −1.93 −1.51 0. 3.75 1.39 3.5 −1 −0. n = 1.05 1.3.12) and (3.49 sec 3 T = 0.0272 −2. and 3 .85 sec 2 Floor 5th 4th 3rd 2nd 1st T1 = 2.796 0.7. 2.1 −2. Each figure is organized in four parts: (a) shows the roof displacement urn t .2 Response History Analysis The structural response due to individual vibration modes.38 0.04 1.72 −2. is shown in Figs.67 −1. Force distributions s* = mφn .27 sec Ground −1.44 1.1 3.8 −2.5. n = 1.5 Fig.5 0 0. (3.61 2.94 2.487 −1. determined by RHA [Eqs.05 2.4. 2. 3.

urn relation.the roof level. two. For a fixed number of modes included. or all modes.15) and (3. larger in story drifts. The errors due to truncating the contributions of vibration modes beyond the third mode are negligible.422 cm. 3. The peak values of displacements of all floors. and 3.2. 3. 3.3.1. (3.1. consistent with the increasing significance of the higher mode response among these three sets of response quantities. The same method was used to determine the peak values of many response quantities. the results for the roof displacement and top-story drift are shown in Fig. with the first three modes provide essentially the exact response. 17 . respectively.2. Combining the modal response histories for all modes gives the total response [Eqs. This is illustrated in Fig. errors are smallest in floor displacements. and three vibration modes.8. The peak values of floor displacements and story drifts determined by RHA. The peak values of the various response quantities are noted in these figures. which are listed in Tables 3. are presented in Fig.9. drifts in all stories. respectively. and ur 3o = 0. including one. but three modes—perhaps even two modes—are sufficient. two. in particular. the peak roof displacement due to each of three modes is ur1o = 9. errors generally decrease as response contributions of more modes are included.16)]. 3. The linear relationship between the base shear and roof displacement for each mode implies that the structure did not yield.3. and even larger in joint rotations. and 3. 3.12 cm. and rotations of external joints with moment connections are presented in Tables 3. where the second and third modal responses are a larger percentage of the top story drift compared to roof displacement. the exact response considering all modes.8. Also included are the combined response due to one. and (d) shows the Vbn . As expected. It is apparent that the first mode alone is inadequate. especially in estimating the story drifts. and the percentage errors due to truncation of higher modal contributions. ur 2o = 2.23 cm. three.

2 0.202 0.192 0.03E-03 -6.159 0.15E-03 4.9 8.56E-03 2.50E-03 4.032 -0.080 0.4 -0.258 0.99E-03 2.275 0.350 0.097 0.364 0.060 -0.231 -0.035 0.295 -0.09E-03 1.22E-03 2.Table 3.9 -16.13E-03 2.237 0.66E-05 -3.3 -0.25 × El Centro ground motion Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Modal Response Mode 1 Mode 2 Drift Ratio (%) Combined (RHA) 1 Mode 2 Modes 3 Modes Mode 3 0.006 0.9 3.321 0.008 -0.3 -0.023 0.1 Building height is measured from the ground floor to the 9th floor.225 0.311 0.2 -4.2 -57.89E-03 1.300 0.24E-03 2.2 Peak values of story drift ratios (as % of story height) from RHA for 0.010 -0.400 0.1 3.264 Error (%) 1 Mode 2 Modes 3 Modes -23.069 0.38E-04 2 Modes 2.5 18.0 -2.307 0.058 -0.5 -2.183 0.090 0.202 0.5 0.054 0.266 0.237 0.156 0.370 0.7 3.4 -10.466 0.024 -0.060 -0.229 0.011 0.3 Table 3.399 0.0 -46.01E-04 -2.74E-03 1.097 0.226 0.2 9.042 0.8 1.6 1.045 0.0 -10.089 0.259 0.3 Peak values of joint rotations (radians) from RHA for 0.1 -2.01E-04 3.475 0.1 -19.9 -22.65E-03 RHA (all modes) 2.124 0.015 -0.009 -0.63E-03 2.179 0.03E-03 3 Modes 2.1 -2.226 0.8 -5.282 0.64E-03 3.282 0.263 0.6 -1.042 0.85E-03 3.6 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 0.11E-04 -5.227 0.4 -41.199 0.44E-03 3.42E-04 1.6 -0.336 0.74E-03 1.74E-04 6.245 0.245 0.202 0.097 0.0 -2.1 -11.50E-03 2.88E-03 2.13E-04 9.333 0.1 -0.8 -15.125 0.310 0.012 0.94E-03 2.38E-03 2.133 -0.90E-03 3.266 0.043 0.03E-03 1.44E-03 1.3 -8.260 0.14E-03 2.7 -19.001 -0.65E-03 2.205 0.121 0.1 4.26E-04 -5.25 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 3 Modal Response Mode 1 2.1 -2.8 -1.260 0.401 Error (%) 1 Mode 2 Modes 3 Modes -23.45E-03 3.261 -0.6 -1.2 -1.7 2.413 RHA (all modes) 0.062 0.266 0.9 9.4 -6.265 0.00E-03 1.00E-03 2.003 0.125 0.152 0.1 Peak values of floor displacements (as % of building height = 37.28E-04 1.124 0.9 -23.229 0.7 4.0 -0.9 2.47E-03 1.7 -50.3 19.00E-03 1.235 0.31E-03 Joint Rotation (rad) Combined (RHA) 1 Mode 2.088 -0.74E-04 9.130 0.152 0.071 0.406 0.4 -1.6 9.378 0.181 0.89E-03 1.91E-04 1.055 0.1 -14.31E-03 Mode 2 -1.311 0.88E-03 2.6 0.4 -53.09E-03 2.8 -56.177 0.325 0.09E-03 1.08E-03 2.14 m3) from RHA for 0.235 0.2 -20.76E-03 1.173 0.253 0.5 -1.78E-04 -3.11E-03 1.72E-03 3.227 0.09E-03 Error (%) 1 Mode -23.038 0.73E-03 3.2 -2.4 -7.4 -22.8 1.157 0.03E-03 1.3 -33.9 -24.069 0.9 -15.002 -0.265 RHA (all modes) 0.4 0.062 -0.9 1.7 2 Modes -3.026 0.63E-03 2.8 -10.011 0.060 0.4 -1.2 1.022 0.253 0.407 0.453 0.76E-03 1.29E-03 2.203 0.259 0.117 0.99E-03 Mode 3 3.1 Table 3.7 7.4 -3.177 0.0 7. 18 .197 0.6 4.173 0.303 0.0 3 Modes -5.6 11.012 -0.156 0.3 -3.33E-03 2.

00174 • 0 (c) r1 −5 0 5 10 15 Time (sec) 20 25 30 0.1 /W 0.5 9.5 0 u (cm) r1 7.1 0.1 0. (c) joint rotation.5.1 5 x 10 −3 θ (rad) 0.5 0 u (cm) r1 7.0435 9.1 −15 −7.12 0 (d) 0 (e) 15 Fig.15 u (cm) 9.0435 • 0 (b) V b1 −0. and (e) pushover curve. Response due to first mode: (a) roof displacement. (d) force-deformation history.1 −15 −7. Excitation is 0. (b) base shear.12 • 0 (a) r1 −15 0. 3.5 15 −0.0435 Vb1 / W 0.25 × El Centro ground motion 19 .12 −0.

1 −5 −2.05 /W 0.5 0 u (cm) r2 2. Response due to second mode: (a) roof displacement.1 Vb2 / W 0.5 5 −0. and (e) pushover curve.05 5 x 10 −3 b2 θ (rad) 0. Excitation is 0.0252 (d) 0 0 (e) −0. 3.00244 • 0 (c) r2 −5 0 5 10 15 Time (sec) 20 25 30 0. (c) joint rotation.23 −2.25 × El Centro ground motion 20 .5 0 u (cm) r2 2.1 0.5 5 Fig.0252 −2.1 −5 −2.0252 • 0 (b) V −0.23 −5 (a) 0.6.23 0.5 u (cm) r2 0 • 2. (b) base shear. (d) force-deformation history.

02 2 θ (rad) 0. and (e) pushover curve.25 × El Centro ground motion 21 . Excitation is 0.02 −0.5 1 −0.1 u (cm) r3 0 • 0. (c) joint rotation.422 /W V b3 0 (d) 0 −0.02 /W V b3 0 • 0.5 0 u (cm) r2 0.7. Response due to third mode: (a) roof displacement.02 0.0108 −0.02 −1 −0.5 0 u (cm) r3 0.0108 −0.02 −1 −0.422 −1 (a) 0.422 (e) −0.0108 x 10 −3 (b) −0.5 1 Fig. 3.000913 • 0 (c) r3 −2 0 5 10 15 Time (sec) 20 25 30 0. (b) base shear. (d) force-deformation history.

and external joint rotations computed by pushover analysis are presented in Tables 3. (3.5d). or three vibration modes. and the errors in these estimates relative to the exact response from RHA considering all modes.20) with n = 2 and 3 up to roof displacements ur 2 o = 2. 3.9 and Tables 3.6. Observe that the target roof displacement in each pushover analysis is identical to its exact value determined by RHA.7e and to the floor displacements.6. 3. (3.5e. 3. and 3.3). considering one. 3. This pushover curve is consistent with the relationship between the base shear and roof displacement determined by RHA (Fig. 3. implying that the modal truncation errors are small if at 22 .4. As suggested by Eq. In RHA the errors arise only from truncating the responses due to higher modes. 3. In practical application.7d). respectively.. two. and external joint rotations in Tables 3. 3. pushing the structure.5.4. confirming that MPA gives the exact values of the individual modal responses.2.8) leads to the pushover curve shown in Fig.12 cm. 3. 3.12).5.2. i. and 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. using the force distribution of Eq.23 cm .3). story drifts. Figure 3. which would provide the Dn value to be substituted in Eq. 3.6e and 3.4. pushing the structure using the force distribution of Eq..1 through 3.3). (3. the floor displacements are proportional to the mode shape φ 1 because the structure remains elastic. (3. respectively. i.1.4) to roof displacement ur1o = 9.6 present estimates of the combined response according to Eq. the value determined by RHA (Fig.e.e. 3. For a fixed number of modes included considered. and 3. Implementing pushover analysis for the second and third modes. this value would be determined directly from the response (or design) spectrum.ur relations determined by RHA (Figs. 3.18).10 and Tables 3.21). except for the algebraic sign associated with Γ1 and minor round-off errors. respectively. leads to the pushover curves shown in Figs. The floor displacements. story drifts. 3.3 Modal Pushover Analysis Implementing MPA for the fundamental vibration mode. and ur 3o = 0. and 3. the errors in the MPA results are generally larger than in RHA (Fig.5.3. and 3.422 cm .9). and the computed response values are identical to the peak response values determined from RHA (Tables 3.6d and 3. and it is apparent in the example considered that three modes provide most of the response (Fig.4. these pushover curves are consistent with the Vb . As for the first mode.1.20) with n = 1 (Fig.

59 • Total (All Modes) 0 −15 0 ∆ (cm) u (cm) 0 −3 0 fig3_8.9.8. Heightwise variation of floor displacements and story drifts from RHA for 0.6 Fig.1 0.5 Ground 0 0.23 r2 −15 15 0 • 0.12 • r1 −15 15 0 r1 0 0 0.eps Ground 0 0.03 −3 3 Mode 3 ∆ (cm) u (cm) r3 −15 15 9.685 • −3 3 Mode 2 ∆ (cm) u (cm) • 2.25 × El Centro ground motion 23 .5 0. 3. 3. sec 25 30 r 5 10 15 20 Time.least three modes are included. sec 25 30 Fig. Response histories of roof displacement and top-story drift from RHA for 0.eps fig3_9b.422 r2 0 • 1. Additional errors are introduced in pushover analysis due to the approximation inherent in modal combination rules.3 0.3 0.4 Displacement/Height (%) 0.48 −3 3 1.2 0.25 × El Centro ground motion: first three modal responses and total (all modes) response (a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th (b) Story Drift Ratios Floor 4th 3rd 2nd 1st Floor 5th RHA All Modes 1 Mode 2 Modes 3 Modes 5th 4th 3rd 2nd 1st RHA All Modes 1 Mode 2 Modes 3 Modes fig3_9a. (a) Roof Displacement 15 (b) Top Story Drift 3 Mode 1 ∆ (cm) u (cm) 9.4 Story Drift Ratio (%) 0.1 0.eps r 5 10 15 20 Time.83 • r3 0 • 0.2 0.

011 -0.313 0.157 0.157 0.038 -0.270 0.266 0.276 0.25 × El Centro ground motion Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Modal Response Mode 1 Mode 2 Drift Ratio (%) Combined (MPA) 1 Mode 2 Modes 3 Modes Mode 3 -0.0 -10.7 2.106 0.4 -4.78E-04 2 Modes 2.00E-03 -1.3 1.090 0.010 0.9 -15.003 -0.96E-03 2.6 -16.310 0.385 0.156 0.012 -0.1 -18.7 -15.133 -0.6 -15.260 0.261 -0.336 RHA (all modes) 0.0 -18.133 0.203 0.9 -13.282 -0.03E-03 6.322 0.9 -15.125 0.80E-04 3.57E-03 1 Mode -23.0 -16.8 -14.253 0.2 -12.230 -0.89E-03 -2.9 -18.069 0.203 0.9 -13.088 -0.2 -11.179 0.53E-04 -9.060 0.253 -0.94E-03 2.3 -9.023 -0.106 0.15E-03 2.286 0.2 -16.76E-03 1.048 0.43E-04 -1.285 0.267 0.8 -15.3 -13.90E-03 1.28E-03 2.0 -2.8 -56.33E-04 5.22E-03 -2.203 0.9 3 Modes -12.328 0.4 -22.133 0.03E-03 3.401 Error (%) 1 Mode 2 Modes 3 Modes -23.370 0.38E-03 2.235 -0.47E-03 1.270 0.226 0.331 0.6 -19.4 Peak values of floor displacements (as % of building height = 37.032 0.00E-03 1.181 0.3 Table 3.00E-03 2.63E-03 2.24E-03 -2.152 0.05E-03 3.9 -11.374 0.260 -0.267 -0.04E-03 3.74E-04 -6.74E-03 1.08E-03 2.7 24 .03E-03 1.8 -22.259 -0.001 0.0 -46.006 -0.7 2 Modes -13.40E-04 5.31E-03 Joint Rotation (rad) Combined (MPA) 1 Mode 2.9 -15.09E-03 -1.42E-04 -1.9 -12.253 0.230 0.2 Table 3.25 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Modal Response Mode 1 -2.9 -14.2 -57.7 -21.65E-03 2.080 0.270 0.09E-03 1.235 0.89E-03 1.3 -12.296 -0.14 m) from MPA for 0.4 -14.25 × 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 0.4 -9.055 0.2 -0.062 -0.97E-03 1.272 0.009 0.90E-04 -9.097 0.125 0.4 -4.078 0.22E-04 2.09E-04 -3.31E-03 Error (%) RHA (all modes) 2.03E-03 -1.156 0.4 0.264 Error (%) Modal Response 1 Mode 2 Modes 3 Modes -23.1 -11.173 0.73E-05 3.2 -20.3 -19.4 -9.3 -33.1 -11.1 -19.332 0.72E-03 3.117 0.12E-03 1.237 0.245 0.282 0.31E-03 2.229 0.9 -24.227 0.92E-04 -1.Table 3.011 0.9 -16.9 -14.09E-03 Mode 2 1.002 0.229 0.4 -53.5 Peak values of story drift ratios (as % of story height) from MPA for 0.152 0.008 -0.274 0.022 0.3 -41.060 -0.4 -22.4 -2.89E-03 2.6 Peak values of joint rotation (radians) from MPA for 0.042 0.7 -19.44E-03 3.2 1.0 -0.069 0.4 -11.267 0.259 0.245 -0.466 0.8 -23.177 0.9 -13.024 0.121 0.177 0.179 0.042 0.045 -0.097 0.73E-03 3.071 0.38E-03 3 Modes 2.203 0.321 0.4 -7.043 -0.026 0.012 0.300 0.4 -5.74E-03 -1.1 -0.079 0.2 -4.231 -0.5 -16.173 0.00E-03 2.015 -0.237 -0.9 -14.336 0.89E-03 2.3 -2.407 0.6 -17.9 -8.197 0.227 0.253 RHA (All Modes) 0.260 0.3 -14.89E-03 -1.00E-03 3.124 0.00E-03 2.036 -0.7 -50.76E-03 -1.8 -15.15E-03 1.048 0.4 -19.99E-03 Mode 3 -3.062 0.63E-03 -2.44E-03 -1.9 -13.9 -15.203 0.4 1.058 -0.

Heightwise variation of floor displacements and story drifts from MPA for 0.25 × El Centro ground motion.4 Displacement/Height (%) 0.10.1 0.eps fig3_10b.2 0.1 0.6 Fig.2 0.4 Story Drift Ratio (%) 0.(a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th (b) Story Drift Ratios Floor 4th 3rd 2nd 1st Floor 5th RHA All Modes MPA 1 Mode 2 Modes 3 Modes 5th 4th 3rd 2nd 1st RHA All Modes MPA 1 Mode 2 Modes 3 Modes fig3_10a.3 0.3 0.5 0. 3.eps Ground 0 0. shading indicates modal combination error 25 .5 Ground 0 0.

26 .

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

4) where the only term that differs from Eq. Equation (4. The equations governing the response of the inelastic system to p eff. (4. (4.Substituting Eq. sign u n b g b g bg (4. For linear systems.2 UNCOUPLED MODAL RESPONSE HISTORY ANALYSIS bg Neglecting the coupling of the N equations in modal coordinates [Eq.Gn ug t Mn bg n = 1. (4.” implying that other “modes” will also contribute to the solution. however.and classical n damping-orthogonality property of modes gives F qn + 2z nw n qn + sn = . 4. (3.3). where φ n are now the modes of the corresponding linear system. premultiplying by φ T .4) represents N equations in the modal coordinates qn .6) for inelastic systems will no longer be described by Eq.6) The solution of Eq. Eq.6b) are mu + cu + f s u. Unlike Eq. qr t = 0 for all 28 bg .9) involves T Fsn = Fsn qn . in principle. these equations are coupled for inelastic systems.2). (3. (4. (3.3) will. and using the mass. (4.4)] leads to the uncoupled modal response history analysis (UMRHA) procedure. (4. This approximate RHA procedure is the preliminary step in developing a modal pushover analysis procedure for inelastic systems. (3. Simultaneously solving these coupled equations and using Eq.… N (4. sign qn = φ n f s u n .5) This resisting force depends on all modal coordinates qn t . (4. (4. sign u = .2).8) because qr t will generally be nonzero for “modes” other than the nth “mode.9) for linearly elastic systems. give the same results for u t as obtained directly from Eq. However.n t given by Eq.4) is rarely solved because it offers no particular advantage over Eq.3) in Eq.2). (3. The spatial distribution s of the effective earthquake forces is expanded into the modal contributions sn according to Eq. implying coupling of modal coordinates because of yielding of the structure.sn ug t bg b g bg bg (4. 2.

(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

Roof displacement due to p eff. where u g t = 3.0 ¥ El Centro ground motion: (a) exact solution by NL-RHA.Forces sn Unit mass An(t ) st rn ωn. and (b) approximate solution by UMRHA bg af bg 32 . ζn.51 • ur2 (cm) 0 ur2 (cm) n=3 14. 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) • 75.112 5 10 15 20 Time (sec) 25 30 −10 0 −10 0 Fig.s n u g t .02 −150 n=2 −150 50 n=2 14.9 • 0 −50 10 −50 10 n=3 ur3 (cm) 0 • 5. 4.3.89 50 r1 0 u (cm) 0 • 78.575 5 10 15 20 Time (sec) 25 30 ur3 (cm) 0 • 5.4.n t = . n = 1. 2. 4. Fsn / Ln ¨ g(t ) u (a) Static Analysis of Structure (b) Dynamic Analysis of Inelastic SDF System Fig. and 3.

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.817 5 10 15 20 Time (sec) 25 30 −10 0 −10 0 Fig. and (3) the Fsn Ln .2.5. 4.008 • 6. (4.744 • 0 −20 10 n=3 −20 10 n=3 ∆r3 (cm) 0 • 5. 2. but even for this very intense excitation.s n u g t .0 ¥ El Centro ground motion.(a) Nonlinear RHA 20 n=1 20 (b) UMRHA n=1 • 5. 4.855 ∆r1 (cm) 0 ∆r1 (cm) n=2 • 6.Dn relation is 33 bg bg . and (b) approximate solution by UMRHA af bg bg 4.33 0 −20 20 −20 20 n=2 ∆r2 (cm) 0 ∆r2 (cm) 7. and 3. (2) the superposition of responses to p eff.15) is strictly valid only for linearly elastic systems.2… N ) according to Eq. n = 1. The errors are slightly larger in drift than in displacement.4) and (4. (4.n t (n = 1.5)] is neglected.n t = .6) by UMRHA is compared with the “exact” solution by nonlinear RHA. Such comparison for roof displacement and top-story drift is presented in Figs. where u g t = 3.0 ¥ El Centro ground motion: (a) exact solution by NL-RHA.5. (3. Top story drift due to p eff. respectively. this intense excitation was chosen to ensure that the structure is excited well beyond its linear elastic limit.1 Underlying Assumptions and Accuracy The approximate solution of Eq.4 and 4.956 5 10 15 20 Time (sec) 25 30 ∆r3 (cm) 0 • 5. the errors in either response quantity are only a few percent. both for 3.

(3. A bilinear idealization of this pushover curve for the nth-“mode” is shown in Fig. (3. 4. How to convert this Vbn . (4.2. However.20).1. (3.Dn relation? The two sets of forces and displacements are related as follows: V Fsn = bn Γn Dn = urn Γ nf rn 34 (4. (4.12) for floor displacements. the only force distribution that produces displacements proportional to f n is given by Eq. Therefore. is plotted against roof displacement urn .8) governing Dn t is based on Eq. (4.6a.2 Properties of the nth-“mode” Inelastic SDF System How is the Fsn Ln . (4.8) before it can be solved? Because Eq. Although most commercially available software cannot implement such displacement-controlled analysis. the base shear is Vbny and roof displacement is urny .8) in UMRHA.9). the relationship between lateral forces f s and Dn in Eq. When implemented by commercially available software. Therefore we impose this constraint in developing the UMRHA procedure in this section and modal pushover analysis in the next section.Dn curve.9) should be determined by nonlinear static analysis of the structure as the structure undergoes displacements u = Dnφ n with increasing Dn . within the linearly elastic range of the structure.urn pushover curve to the Fsn Ln . which is different than the Fsn Ln . such nonlinear static analysis provides the so-called pushover curve. The overall errors in the UMRHA procedure are documented in the examples presented in Section 4.4.20) to some bg pre-determined roof displacement. (4.approximated by a bilinear curve to facilitate solution of Eq. this distribution seems to be a rational choiceeven for inelastic systemsto determine Fsn in Eq. Although several approximations are inherent in this UMRHA procedure. when specialized for linearly elastic systems it is identical to the RHA procedure of Section 3.Dn relation to be determined in Eq. 4. At the yield point. and the base shear Vbn .10) . The structure is pushed using the force distribution of Eq. it can conduct a force-controlled nonlinear static analysis with an invariant distribution of lateral forces. 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.

Knowing Fsny Ln and Dny from n Eq. which may differ from the period of the corresponding linear system. In contrast. Properties of the nth-“mode” inelastic SDF system from the pushover curve Equation 4.13) This value of Tn .8). Section 13. 4. 2001. (4. the initial slope of the pushover curve in Fig.11).10 enables conversion of the pushover curve to the desired Fsn Ln − Dn relation shown in Fig. The two are related through Fsny Ln 2 = w n Dny (4. where the yield values of Fsn Ln and Dn are Fsny Ln = Vbny * Mn Dny = urny Γ nf rn (4. 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. should be used in Eq.6b is w 2 .11) * in which M n = Ln Γn is the effective modal mass (Chopra.6. (4. which is not a meaningful quantity.2.12) implying that the initial slope of the curve in Fig. 4.6b. 4. 4.V bn (a) Idealized Pushover Curve F sn /L n (b) F sn / L − D Relationship n n Idealized Vbny Actual 1 αnkn V bny /M * n 1 α ω2 n n kn 1 1 ωn 2 u u rny rn Dny = ur n y / Γn φ r n D n Fig.5).6a is 2 kn = ω n Ln . 35 .

16) to determine the total response. (3. 4. 4. of the total response r t obtained in Step 8. r o .11). 7. Idealize the pushover curve as a bilinear curve (Fig. 4. 6. Calculate histories of various responses by Eqs. Compute the natural frequencies.13). 4. An t .21) where Dn . 2. φn .6b. (4. Dn (t ) . n (3.3b) with force-deformation relation of Fig. For the nth-mode. which is summarized next as a sequence of steps. ωn . (3. the first two or three modes will suffice. (3. (3. Convert the idealized pushover curve to the Fsn Ln − Dn relation (Fig. 5. 8.2. 4. details are available in Appendix A: 1. 4. 9. is now determined by 36 bg bg . This value of the roof displacement is given by Eq.12) and (3. and modes. n 3.15) and (3.20)].4.20)]. Consider a nonlinear static analysis of the structure subjected to lateral forces distributed over the building height according to s* [Eq. for linearly-elastic vibration of the building. Repeat Steps 2 to 6 for as many modes as required for sufficient accuracy.6b) by utilizing Eq.6a).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. Typically.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 . with the structure is pushed to the roof displacement urno . the peak value of Dn t . Calculate the peak value.n t . develop the base-shear – roof-displacement ( Vbn − urn ) pushover curve for the force distribution s* [Eq. Compute the deformation history. of the nth-“mode” inelastic SDF system (Fig. and pseudo-acceleration history. Combine the “modal” responses using Eqs.

8).21). as described in Section 4. The peak “modal” responses rno .2.g. and (2) it provides the exact modal response for elastic systems.1 Summary The peak inelastic response of a building to earthquake excitation can be estimated by the MPA procedure just developed. The response value rno is an estimate of the peak value of the response of the inelastic system to p eff. governed by Eq.18). are combined using an appropriate modal combination rule. bg bg bg bg Thus.12). joint rotations. (3. Dn . alternatively. story drifts. (3. 4.solving Eq. 4. Compute the peak deformation. as mentioned earlier. etc.3) because. 4. 37 . 6. It provides results for elastic buildings that are identical to the well-known RSA procedure (Section 3. However. details are available in Appendix B. rno also represents the exact peak value of the nth-mode contribution rn t to response r t . Eq. although somewhat intuitive for inelastic buildings.2 and 3. At this roof displacement. it seems reasonable because it provides results for elastic buildings that are identical to the well-known RSA procedure described in Section 3. the lateral force distribution used possesses two properties: (1) it appears to be the most rational choice among all invariant distribution of forces. e. seems reasonable. 5. This application of modal combination rules to inelastic systems obviously lacks a theoretical basis. for elastic systems. plastic hinge rotations. (4. or from the inelastic response (or design) spectrum.n t . Sections 7.6b by solving Eq. (4.4.6 and 7.3.. the pushover analysis provides an estimate of the peak value rno of any response rn t : floor displacements. to obtain an estimate of the peak value ro of the total response.3b) with force-deformation relation of Fig. This pushover analysis.6). we will refer to rno as the peak “modal” response even in the case of inelastic systems. which is summarized next as a sequence of steps.3. of the nth-“mode” inelastic SDF system (Fig. As shown in Sections 3. it can be determined from the inelastic response (or design) spectrum (Chopra. Calculate the peak roof displacement urno associated with the nth-“mode” inelastic SDF system from Eq. (4.2. each determined by one pushover analysis. 2001.8). Steps 1 to 4 of the MPA are same as those for UMRHA.

and compared with the results of a rigorous nonlinear RHA using the DRAIN-2DX computer program. respectively. and the errors in the approximate results are shown in Fig.7 shows the individual “modal” responses. 4. 8. in contrast to modal analysis (Section 3.0 to 3. ur 2o = 11. This is to be expected.4. extract from the pushover database values of other desired responses.1). although the trends are not systematic as when the system remained elastic (Section 3. is presented next.53 cm.1 Uncoupled Modal Response History Analysis The structural response to 1. determined by the UMRHA procedure.0. 9.7. This deficiency also implies that. three “modes” included. The peak values of displacements of all floors and drifts in all stories are presented in Tables 4. Repeat Steps 3 to 8 for as many “modes” as required for sufficient accuracy.2). subtract the yield value of hinge rotation to determine the hinge plastic rotation. also included are the combined responses due to one. and the percentage errors in the approximate results. Typically. the 38 . (3. Figure 4. the “exact” results. 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. the combined response due to three “modes”.7 cm. Determine the total response by combining the peak “modal” responses using the SRSS combination rule of Eq.2. the UMRHA procedure lacks a rigorous theory.4. say.1 and 4. with. 4. 4. and the “exact” response from nonlinear RHA for the roof displacement and top-story drift. The peak values of response are as noted.18). Observe that errors tend to decrease as response contributions of more “modes” are included. and ur 3o = 2.5 ¥ the El Centro ground motion including the response contributions associated with three “modal” inelastic SDF systems. At urno . rno . From the total rotation of a plastic hinge. two. in particular. the first two or three “modes” will suffice. The peak values of floor displacements and story drifts including one. 4. two.8. and three “modes”.3 cm. and three modes are compared with the “exact” values in Fig.4 COMPARATIVE EVALUATION OF ANALYSIS PROCEDURES The response of the 9-story building described earlier is determined by the two approximate methods: UMRHA and MPA.9.

0.25 (Tables 3.5 (Tables 4.44 • "Mode" 2 r1 0 • 48.62 5. Next.0.6 5 10 15 20 Time.7 • −80 80 ∆r1 (cm) u (cm) "Mode" 1 −12 r2 0 ∆r2 (cm) 12 0 u (cm) −80 80 −12 r3 0 • 2. However. 0. for a fixed number of “modes” included.response is much less accurate (Tables 4. For each excitation.75.2). the errors in responses computed by UMRHA including three “modes” relative to the “exact” response were determined. and total response from NL-RHA 39 .1 and 4.0.5. in particular. the errors in story drifts are larger compared to floor displacements. 1. 0.1 and 4.88 u (cm) "Mode" 3 −80 80 0 −80 80 0 −80 0 • 44. defined as the El Centro ground motion multiplied by 0.85. sec 25 30 • 48.5 × El Centro ground motion: individual “modal” responses and combined response from UMRHA. For this purpose the UMRHA and exact analyses were repeated for ground motions of varying intensity.7. Response histories of roof displacement and top-story drift due to 1.3 11. just as for elastic systems. 1.1 and 3.2).25. the degree to which the system deforms beyond its elastic limit. 4. we investigate how the errors in the UMRHA vary with the deformation demands imposed by the ground motion.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 0.0. sec 25 30 r −12 0 Fig.5.1 and 3. (a) Roof Displacement (b) Top Story Drift 80 12 0 • 3.1 −12 12 ∆r (cm) u (cm) UMRHA 3 "Modes" 0 • 7.53 ∆r3 (cm) 12 0 • 2.2) if the system yields significantly versus if the system remains within the elastic range (Tables 3. and 3.2) and 1. 2.38 6.

5 1 1. the pushover curves using force distributions s* [Eq. Two versions of the pushover curve are 40 . with the peak displacement of each “modal” SDF system noted for each ground motion multiplier.11.20)] for the first three n modes of the system are shown in Fig. indicated by a ground motion multiplier.5 ¥ El Centro ground motion Figure 4.5 × El Centro ground motion (a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th (b) Story Drift Ratios Floor 4th 3rd 2nd 1st UMRHA 1 "Mode" 2 "Modes" 3 "Modes" Floor 5th 5th 4th 3rd 2nd 1st UMRHA 1 "Mode" 2 "Modes" 3 "Modes" Ground −60 −40 −20 0 Error (%) 20 40 60 Ground −60 −40 −20 0 Error (%) 20 40 60 Fig. 4. Shown is the error in each floor displacement (Fig. 4. 4.8.5 Fig. Heightwise variation of error in floor displacements and story drifts estimated by UMRHA including one. it will be useful to know the deformation of the system relative to its yield deformation.5 1 1. 4.10b). in each story drift (Fig.10 summarizes the error in UMRHA as a function of ground motion intensity.5 Story Drift Ratio (%) 2 2. 4. or three “modes” for 1. and the error envelope for each case. (3.9.(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 0.10a). For this purpose. two. Heightwise variation of floor displacements and story drift ratios from UMRHA and NL-RHA for 1. To interpret these results.5 Displacement/Height (%) 2 Ground 0 0.

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. The structure is pushed using the force distribution of Eq.4) to roof displacements urno = 48.6 and 4. the ductility factor for the first mode system is only 1.1).0. and 2. and. UMRHA lacks a rigorous theory and is based on several approximations. 4.53 cm. 4.2. 4. as mentioned in Section 4.75.included: the actual curve and its idealized bilinear version.75. 11. (3) maintain roughly similar values for more intense ground motions. The system remains elastic up to ground motion multiplier 0. as mentioned in Section 3.3).01 (Fig. For more intense excitations. were noted but are not shown here. The idealized curve for the first “mode” deviates most from the actual curve near the peak displacement corresponding to ground motion multiplier 1. Figure 4.0.11a).11).4. Each of these three pushover analyses provides the pushover curve (Fig. the first reason mentioned above seems to be the primary source for the errors.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. (3. Additional errors are introduced in UMRHA of systems responding beyond the linearly elastic limit for at least two reasons. 4. and plastic hinge rotations at the external beam end at each floor level (Table 4. First. 3. and (4) are larger in story drift compared to floor displacements. The location of plastic hinges and their rotations. 2.5 ¥ the El Centro ground motion. the pushover curve for each “mode” is idealized by a bilinear curve in solving Eq. drifts in all stories (Table 4. the peak values of displacements at all floors (Table 4. Perhaps this explains why the errors are large at this excitation intensity.4).10) for each “modal” inelastic SDF system (Figs 4. even though the system remains essentially elastic.2.7 cm.3 cm. Second. (4. respectively. the values determined by RHA of the nth-mode inelastic SDF system (Fig.5).7).20) with n = 1. (2) increase rapidly as the ground motion multiplier increases to 1. and 3 (Fig. determined from “exact” analyses. the errors in truncating the higher mode contributions are negligible.4. 41 .

8 1.201 -1.126 0.350 -0.942 -0.2 22.9 16.241 1.372 1.3 8.554 1.852 1.8 14.136 1.2 6.256 1.248 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.293 1.370 -0. 4.070 1.096 0.214 -0.942 1.371 -0.763 -15.373 -0.820 -0.4 4.698 1.472 1.9 31.372 -1. and (b) story drifts Table 4.7 31. Errors in UMRHA as a function of ground motion intensity: (a) floor displacements.811 1.5 3 0 0 0.540 0.426 -1.8 1.5 18.200 8.5 10.410 -1.220 -0.2 1.317 0.226 -0.072 1.2 12.806 0.0 1.490 1.4 -7.1 1.722 0.772 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 1.1 3.0 11.072 -1.260 -15.727 1.5 9.009 -0.410 1.366 0.018 0.9 12.071 -0.575 -41.3 1.10.473 -22.201 -0.938 1.478 0.366 -0.068 0.983 1.003 -31.241 -1.7 14.513 -0.256 1.4 1.071 0.938 -1.1 8.5 2 GM Multiplier 2.914 2.315 -0.104 0.049 -0.663 -0.9 12.088 10.5 -3.490 -1.5 1 1.138 1.663 0.971 1.8 1.187 -0.513 0.1 1.1 Peak values of floor displacements (as % of building height = 37.0 9.2 1.5 28.877 0.495 1.5 2 GM Multiplier 2.5 1 1.676 0.010 0.202 11.7 Table 4.5 9.5 × 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” -0.3 6.044 1.235 -0.291 0.806 -0.844 -25.003 0.0 11.14 m) from UMRHA for 1.135 9.707 1.668 -23.256 -1.298 0.856 2.376 -1.120 1.407 -10.169 0.3 0.3 42 .8 1.133 1.5 28.033 0.338 1.216 1.6 2.376 1.130 0.220 0.065 0.333 0.214 0.379 1.8 0.1 0.484 0.079 0.8 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 0.9 5.0 -9.914 -0.057 -0.945 -37.9 31.863 0.616 -0.430 1.0 7.982 9.5 3 Fig.6 1.121 -0.298 -0.154 0.5 × El Centro ground motion Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Modal Response “Mode” 1 “Mode” 2 Drift Ratio (%) Combined (UMRHA) 1 “Mode” 2 “Modes” “Mode” 3 -1.900 -10.055 -0.2 4.526 -0.6 4.863 1.(a) Floor Displacements 60 50 40 Error Envelope Error for Floor No.4 -1.820 -19.2 Peak values of story drift ratios (as % of story height) from UMRHA for 1.0 2.371 0.201 1.819 2.751 1.3 25.283 1.338 -1. Noted 60 50 40 (b) Story Drifts Error Envelope Error for Story No.

the errors in the modal pushover results are.12 and Tables 4. overestimated by up to 18% in the middle stories.13 and Tables 4.14. 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. The errors are especially large in the hinge plastic rotations estimated by the MPA procedure. MPA failed to identify the plastic hinges at the column bases in Fig. considering one. the results were not always accurate. The drifts are underestimated by up to 13% in the lower stories. As shown in Figs. the third “mode” does not contribute because this SDF system remains elastic (Fig. 4. Furthermore. 4. 4. this error is not especially significant because the hinge plastic rotation is very small. were determined by four analyses: MPA considering one “mode.11c).2).12 and Tables 4.Figures 4. 4. and are within a few percent of the exact values for the upper stories. However.14. 4.9 and Tables 4.5).3 and 4. however.” two “modes. for two or three modes included.3 and 4.5 present estimates of the combined response according to Eq. significantly smaller than in UMRHA (compare Fig.3 and 4.4).12 and 4. although the error is recorded as 100% if MPA estimates zero rotation when the nonlinear RHA computes a non-zero value.” and nonlinear RHA.” respectively. especially in estimating the story drifts (Fig. 4. 4.4 with Tables 4. Obviously. 4.18). but was more successful when the excitation was more intense (results are not included).4). For example. The locations of plastic hinges shown in Fig.1 and 4. The second “mode” was necessary to identify hinges in the upper stories. MPA including three “modes” underestimates the displacements of the lower floors by up to 8% and overestimates the upper floor displacements by up to 14%. Fortuitously.” and three “modes.13 and Table 4. but the third “mode” contributions do not seem especially important (Fig. The first “mode” alone is inadequate.3 through 4. significant improvement is achieved by including response contributions due to the second “mode”. and three “modes. 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. (3. two.4. even if three “modes” are included (Fig. the hinges identified in beams at the 6th floor were at variance with the “exact” results. 43 . and the errors in these estimates relative to the exact response from nonlinear RHA.13 and Tables 4. One “mode” pushover analysis was unable to identify the plastic hinges in the upper stories where higher mode contributions to response are known to be more significant. Pushover analysis seems to be inherently limited in computing accurately hinge plastic rotations.13 with Fig. in general.3 and 4. 4.

5 0.13 y by Base Shear (kN) 8000 6000 1 1.0. “Modal” pushover curves with peak roof displacements identified for 0.85 Base Shear (kN) 8000 6000 4000 2000 0 0 0.25 0.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36.0.25 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. 0.5 1 0. 0. V = 5210 kN. V = 7616 kN. 1. 4.75.5 2 3 Actual Idealized 2000 0 0 0.2 cm.85 0.9 cm.5 0.25.6 cm.5 2 3 4000 2000 0 0 Actual Idealized 0. V = 4952 kN.0 × El Centro ground motion 44 .5.75 0.5. 2.85 0. α = 0. 1. α = 0.5 0.75 0.14 y by Base Shear (kN) 8000 6000 4000 1. and 3.75 Actual Idealized 20 40 60 Roof Displacement (cm) (b) "Mode" 2 Pushover Curve 80 12000 10000 u = 9. α = 0.11.

the errors in MPA were larger for ground motion multipliers less than 0. an observation with practical implications. As mentioned in Section 3.5 Displacement/Height (%) 2 Ground 0 0.4 also apply to MPA. Shown is the error in each floor displacement (Fig. The errors are only weakly dependent on ground motion intensity (Fig.15) for ground multipliers larger than 1. implying excitations intense enough to cause significant yielding of the structure.3. In this case. implying that the modal combination errors contained in these procedures are acceptable.15 summarizes the error in MPA considering three “modes” as a function of ground motion intensity.5 Story Drift Ratio (%) 2 2. now standard in engineering practice. 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.(a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th NL−RHA MPA 1 "Mode" 2 "Modes" 3 "Modes" (b) Story Drift Ratios Floor Floor 5th 4th 3rd 2nd 1st 5th 4th 3rd 2nd 1st NL−RHA MPA 1 "Mode" 2 "Modes" 3 "Modes" Ground 0 0.4. and the error envelope for each case. shading indicates errors in MPA including three “modes” Figure 4. implying excitations weak enough to limit the response in the elastic range of the structure.5 × El Centro ground motion. identified in Section 3.75.4.15b). the errors in MPA were fortuitously smaller than in UMRHA (compare Figs.0. whereas MPA contains errors inherent in modal combination rules. While various sources of errors in UMRHA. each story drift (Fig. the MPA procedure is equivalent to the RSA procedure. as discussed in Sections 3. 4. 4. 4.12 Heightwise variation of floor displacements and story drift ratios from MPA and NL-RHA for 1.15).10 and 4.5 Fig.15a). However. indicated by a ground motion multiplier.5 1 1. 45 .5 1 1. 4.2 and 3. UMRHA is essentially exact.3 for elastic systems (or weak ground motions). 4.

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. and hinge plastic rotations estimated by MPA including one.5 ¥ El Centro ground motion 46 . two. and three “modes” for 1. Errors in floor displacements. 4.

00E+00 0.6 -44.00E+00 3.2 -100.10E-02 9.667 1.0 3 “Modes” -32.8 -4.687 0.220 1.745 1.8 0.156 -0.015 0.60E-04 7.8 -29.116 1.399 0.068 0.2 -100.118 0.667 -1.02E-03 0.0 15.260 -14.3 11.00E+00 0.125 -1.88E-03 0.088 12.00E+00 0.007 1.00E+00 0.516 0.00E+00 0.8 7.2 1.007 1.820 -7.00E+00 0.6 -44.5 -6.6 13.407 -27.1 62.666 Table 4.8 1.05E-03 2.209 1.9 -4.581 0.980 0.00E+00 0.781 0.1 62.053 1.14 m) from MPA for 1.00E+00 0.003 -16.5 Table 4.5 × 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” 1.9 5.6 -9.197 -0.2 0.2 -4.76E-03 4.00E+00 0.900 -0.372 0.6 -8.5 2.804 1.116 1.76E-03 4.02E-03 0.694 1.503 1.00E+00 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th “Mode” 2 0.4 -8.298 -0.351 -0.1 -8.0 -50.00E+00 0.116 1.88E-03 1.55E-03 3.331 1.22E-10 NL RHA 1.304 1.371 -0.895 1.982 13.266 -0.36E-03 6.88E-03 0.130 0.683 1.0 47 .057 0.498 1.0 -100.1 13.942 6.6 -7.5 2.135 -7.1 1.37E-03 1.00E+00 0.575 -53.009 0.2 -3.6 1.76E-03 4.00E+00 0.259 1.844 -7.980 -0.02E-03 3.72E-03 7.018 -0.72E-03 7.4 Peak values of story drift ratios (as % of story height) from MPA for 1.071 -0.5 10.8 -6.55E-03 3.752 1.2 6.033 -0.0 1.8 1.9 0.101 -0.37E-03 1.652 1.00E+00 0.222 0.705 1.055 0.945 -49.233 1.233 1.0 2 “Modes” -32.738 1.8 -29.00E+00 0.733 1.3 Peak values of floor displacements (as % of building height = 37.5 7.3 1.76E-03 4.5 × El Centro ground motion Hinge Plastic Rotation (rad) “Modal” Response Combined (MPA) “Mode” 1 7.37E-04 Error (%) “Modal” Response 1 “Mode” -32.763 -14.00E+00 3 “Modes” 7.517 1.737 0.9 7.55E-03 3.37E-03 1.00E+00 2 “Modes” 7.705 -1.Table 4.910 1.8 -29.176 0.015 0.8 17.00E+00 “Mode” 3 0.2 0.72E-03 7.220 1.298 0.399 0.36E-03 6.99E-03 6.518 1.305 -0.668 -13.8 “Mode” 3 -1.6 13.514 -1.7 1.8 -6.1 46.1 62.02E-03 3.9 -100.26E-04 9.190 -0.00E+00 0.5 Peak values of hinge plastic rotations (radians) from MPA for 1.168 -0.00E+00 0.244 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 0.154 0.53E-03 7.728 1.066 -0.336 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.311 0.36E-03 6.018 -0.473 -15.50E-10 3.6 7.018 0.200 8.00E+00 0.426 15.5 1.315 0.304 -1.0 -5.5 10.503 -1.18E-03 7.756 0.19E-10 3.614 0.7 -12.72E-03 7.066 -0.781 0.338 1.9 2.641 1.071 0.00E+00 0.737 1.5 2.9 -100.581 0.222 0.2 11.4 1.594 -1.611 0.049 -0.756 0.3 13.053 -1.1 46.0 -100.414 1.640 1.1 46.9 1.1 18.8 -12.00E+00 1 “Mode” 7.202 8.478 0.0 -50.37E-03 1.012 1.429 0.7 1.105 0.435 0.640 -1.00E+00 0.527 -0.36E-03 6.3 -3.60E-03 2.895 1.0 -100.250 0.2 -100.879 1.9 1.3 -3.911 0.

resulting in the pushover curve shown in Fig.3 Modal Pushover Analysis with Gravity Loads To evaluate the accuracy of the dynamic response of the system.16a—which is slightly different than the one excluding gravity loads (Fig.4. and overestimated by up to 29% in the middle stories and by up to 13% in the upper stories. 4. but it is inadequate in estimating the story drifts (Fig. Results of such analyses for the selected building subject to 1.7.0 cm. (3. They are now included in the pushover analysis of the structure for the first “mode” only.18 and Tables 4. and plastic hinge rotations at selected external beam end at each floor level (Table 4. and the structure is pushed using the force distribution of Eq. two.4.6 and 4.” however. 4. drifts in all stories (Table 4. Obviously gravity load effects will influence the seismic demands due to the first “mode.7). 4. 4. (3. as are the roof displacements ur 2o = 11. the structure is pushed using the force distribution of Eq. 4.53 cm. and three “modes.6 and 4. Also presented at these roof displacements are the displacements at all floors (Table 4. 4. The response contributions of the second and third modes are the same as before (Tables 4.6).8).17 and Tables 4.” these effects will modify the combined modal response as well as the results of nonlinear RHA. Significant improvement is achieved by including response contributions due to the second “mode. the third “mode” contributions do not seem especially important (Fig.11a). The errors are especially large in the plastic hinge rotations estimated by the MPA procedure even if three “modes” are included (Fig. the results presented so far did not include gravity load effects. Figures 4.18 and Tables 4.5 × El Centro ground motion are presented below.4) to roof displacement ur1o = 52.7).6 through 4.5).17 and Tables 4. The story drifts are underestimated by up to 7% in the lower stories. 3.7). 4. Static analysis of the structure for gravity loads provides the initial state—forces and deformations—of the structure.18 and 48 . (3.8 present estimates of the combined response according to Eq.18) considering one. As shown in Fig.20) with n = 1 to a target roof displacement. The first “mode” alone provides adequate estimates of floor displacements.16 are unchanged.” but not the contributions of higher “modes.6 and 4.3 .20) with n = 1 (Fig.4.7 cm and ur 3o = 2.17 and 4. The pushover curves for the second and third modes included in Fig.” and the errors in these estimates relative to the exact response from nonlinear RHA. MPA including two “modes” underestimates the displacements of lower floors by up to 6% and overestimates the upper floor displacements by up to 22%. Starting with its initial state under gravity loads.

Locations of plastic hinges determined by MPA considering one. and (b) story drifts 49 . Most pushover analysis procedures do not seem to compute to acceptable accuracy plastic hinge rotations.15. 4.Table 4. two.5 × El Centro ground motion (a) Floor Displacements 60 50 40 Error Envelope Error for Floor No. Noted 60 50 40 (b) Story Drifts Error Envelope Error for Story No. 1−"Mode" • • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • •• • •• • • • • • •• •• • •• •• •• •• (b) MPA.5 2 GM Multiplier 2.8). 2−"Modes" •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• • •• (c) 3−"Modes" • • • • • • • •• •• • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• • •• • • • • • • • • • (d) Nonlinear RHA •• •• • •• •• •• •• • •• •• •• •• •• •• •• • •• •• • •• •• •• •• • • • • • • •• •• •• Fig. and three “modes” and by NL-RHA for 1.5 1 1.5 3 0. (a) MPA.5 2 GM Multiplier 2.5 3 Fig. Noted Error (%) 30 20 10 0 0 2 8 3 4 1 9 Error (%) 30 20 5 4 6 5 7 2 9 10 0 0 1 8 7 3 6 0.5 1 1. 4. Errors in MPA as a function of ground motion intensity: (a) floor displacements.14.

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

5 2 3 4000 2000 0 0 Actual Idealized 0. α = 0.9 cm. V = 4952 kN. 0.85 0.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36. 1.3 cm. 0. α = 0.75.0.25 0. 0.5 0.13 y by Base Shear (kN) 8000 6000 1 1. α = 0. 2.5 0.5 0.25. 4.85 Base Shear (kN) 8000 6000 4000 2000 0 0 0.25 1 2 4 6 Roof Displacement (cm) 8 10 Fig.75 Actual Idealized 20 40 60 Roof Displacement (cm) 80 (b) "Mode" 2 Pushover Curve 12000 10000 u = 9. V = 7433 kN. noted are peak values of roof displacement for 0.75 0. V = 5210 kN. and 3.85 0.5 2 3 Actual Idealized 2000 0 0 0.50.0.6 cm. “Modal” pushover curves with gravity loads included.19 y by 3 2 1.75 0.16.5 1 0.0 × El Centro ground motion 51 .14 y by Base Shear (kN) 8000 6000 4000 1.25 5 10 15 20 Roof Displacement (cm) 25 (c) "Mode" 3 Pushover Curve 12000 10000 u = 4.85.

5 Fig. 4.5 Displacement/Height (%) 2 Ground 0 0. gravity loads included. story drifts. 4. Heightwise variation of floor displacements and story drift ratios from MPA and NL-RHA for 1.5 × El Centro ground motion.5 Story Drift Ratio (%) 2 2. and 52 . Errors in floor displacements.(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 0.18. two. and hinge plastic rotations estimated by MPA including one.5 1 1. shading indicates errors in MPA including three “modes” (a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th (b) Story Drift Ratios Floor 4th 3rd 2nd 1st MPA 1 "Mode" 2 "Modes" 3 "Modes" Floor 5th 5th 4th 3rd 2nd 1st MPA 1 "Mode" 2 "Modes" 3 "Modes" Ground −60 −40 −20 0 Error (%) 9th 8th 7th 6th 20 40 60 Ground −60 −40 −20 0 Error (%) 20 40 60 (c) Hinge Plastic Rotations Floor 5th 4th 3rd 2nd 1st MPA 1 "Mode" 2 "Modes" 3 "Modes" Ground −120 −80 −40 0 Error (%) 40 80 120 Fig.5 1 1.17.

20.5 × El Centro ground motion. 1−"Mode" • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • •• •• •• • • • • • •• •• •• •• •• •• •• (b) MPA. and (b) story drifts.5 1 1.5 3 Fig. Locations of plastic hinges determined by MPA considering one. 3−"Modes" (d) Nonlinear RHA • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• •• •• • • • • • • • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • • • • • •• •• •• Fig. and three “modes” and by NL-RHA for 1.5 2 GM Multiplier 2. gravity loads included (a) Floor Displacements 60 50 40 Error Envelope Error for Floor No.19. Errors in MPA as a function of ground motion intensity: (a) floor displacements. gravity loads included 53 . 2−"Modes" •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• •• •• (c) MPA. 4. two. Noted Error (%) 30 20 6 8 1 5 2 4 3 9 7 Error (%) 30 4 5 20 6 9 7 1 10 0 0 10 0 0 8 2 3 0. Noted 60 50 40 (a) Story Drifts Error Envelope Error for Story No.5 1 1. 4.5 2 GM Multiplier 2.(a) MPA.5 3 0.

00E+00 2 “Mode” 8.2 -3.098 20.102 1.830 -12.2 4.5 54 .3 -100.2 2.00E+00 0.5 × El Centro ground motion.2 4.637 0.00E+00 0.00E+00 “Mode” 2 0.996 -0.14 m) from MPA for 1.603 -1.0 2 “Modes” -32.8 0.744 1.037 0.23E-02 1.156 -0.00E+00 Combined (MPA) 1 “Mode” 8.821 -1.04E-10 3.00E+00 0.8 Table 4.330 1.933 1.00E+00 “Mode” 3 0.7 -2.2 -0.11E-03 9.015 0.953 15.7 19.033 -0.0 -30.257 0.998 0.687 0.068 0.101 -0.7 1.7 16.3 -22.049 -0.913 7.9 -6.35E-03 8.11E-03 9.305 -0.813 1.19E-03 1.018 -0.2 0.9 31.927 1.35E-03 8.434 0.0 -30.5 3 “Modes” -32.00E+00 0.5 10.00E+00 0.266 -0.2 -2.00E+00 0.754 1.263 1.3 1.102 1.009 0.00E+00 0.728 1.514 -1.399 -0.23E-03 0.2 1.434 0.854 0.454 1.071 -0.330 1.2 12.125 -1.9 -4.315 0.921 1.3 -22.00E-03 5.1 1.637 0.429 -1.594 -1.35E-03 8.4 -4.351 -0.5 21.8 -32.00E+00 3 “Modes” 8.19E-03 1.8 -32.107 1.207 18.371 -0.1 4.414 28.00E+00 0.8 0.6 Peak values of floor displacements (as % of building height = 37.372 0.938 1.6 1.5 × El Centro ground motion.507 1.5 1.530 1.00E+00 0.850 -1.8 Peak values of hinge plastic rotations (radians) from MPA for 1.4 -4.434 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.673 Table 4.2 9.00E+00 0.263 0.75E-03 0.2 9.00E+00 3.00E+00 0.8 28.00E+00 Error (%) “Modal” Response 1 “Mode” -32.353 -23.2 4.5 -5.213 1.11E-03 9.00E+00 0.00E+00 0.466 0.2 21. 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” 1.00E-03 5.00E-10 NL RHA 1.527 -0.822 1.530 1.667 0.35E-10 3.23E-03 0.515 -50.850 1.877 -46.983 1.00E-03 5.399 0.5 “Mode” 3 -1.4 -6.888 1.037 -0.310 1.983 1.5 0.665 0.55E-03 3.00E+00 0.00E+00 0.9 -3.168 -0.3 -100.6 19.2 1.13E-03 5.Table 4.114 -1.26E-03 3.7 4.436 1.066 -0.213 1.5 × El Centro ground motion.0 -100.109 0.4 1.071 0.128 -1.686 -7.270 -12.214 0.8 1.429 1.478 0.7 Peak values of story drift ratios (as % of story height) from MPA for 1.821 1.831 0.00E+00 0.00E+00 0.197 -0.783 1.057 0.3 13.88E-03 1.176 0.23E-03 3.998 1.04E-02 8.0 37.836 -0.3 -22.6 2.854 0.4 26.4 1.2 16.530 1.00E+00 0.319 1.0 16.55E-03 3.055 0.1 13.5 21.908 1.7 -2.114 1.1 21.23E-03 3. gravity loads included Hinge Plastic Rotation (rad) Floor “Modal” Response “Mode” 1 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 8.858 2.19E-03 1.78E-03 1.105 0.213 1.831 0.0 -100.066 -0.490 -11.4 20.5 1.00E+00 0.3 9.0 37.190 -0.3 -100.35E-03 8.860 1.11E-03 9.754 0.199 16.516 0.00E-03 5.237 0.461 0.88E-03 0.908 -1.6 2.88E-03 0.998 21.8 9.603 1.19E-04 5.311 0.2 9.0 37.319 1.130 0.6 0.154 0.064 -10.996 0. gravity loads included Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Displacement /Height (%) “Modal” Response Combined (MPA) “Mode” 1 “Mode” 2 “Mode” 3 1 “Mode” 2 “Modes” 0.17E-03 9.55E-03 3.19E-03 1.237 0.

2… N ). MPA considering three “modes.5 sec . gravity load effects were included in all 55 . and the exponent k = 1 for fundament period T1 £ 0. SRSS distribution: s* is defined by the lateral forces back-calculated from the story shears determined by response spectrum analysis of the structure. joint rotations. Comparison of Modal and FEMA Pushover Analyses 5. a plot of base shear versus roof displacement. story drifts..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-273.1 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. and varies linearly in between. plastic hinge rotations. k = 2 for T1 ≥ 2.5. and 3. etc. assumed to be linearly elastic. 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. “Uniform” distribution: s* = m j (where the floor number j = 1. 5. j 2. The floor displacements.” and nonlinear RHA.5 sec . 1997] The pushover curve. Specified in FEMA-273 are three distributions for lateral forces: 1. 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.

11 0. 5. Using each of these force distributions. The pushover curves are given in Fig. Force distributions in FEMA-273: (a) “uniform”.3.367 (a) Uniform (b) ELF (c) SRSS Fig. both presented in Section 4. The MPA procedure is more accurate than all the FEMA force distributions.21 0. Figures 5.112 0. 5.3.analyses.0381 0.11 0.5 times the El Centro ground motion. The “uniform” distribution overestimates all floor displacements by 17-28%.11 0.1 through 5.2.00719 0.165 0.0913 0.0446 0. 5. 0. Also included in these presentations are the MPA results considering three “modes” and the “exact” demands from nonlinear RHA.0981 0.4 and Tables 5. the value determined from RHA of the first-mode inelastic SDF system for 1.3a and Table 5. and the displacements of upper floors are overestimated by 9-22%.1.119 0.177 0.062 0. The first four floor displacements are within 5% of the “exact” value. and (c) SRSS 56 .0702 0.3. (b) ELF. 5.0 cm. and Table 5.281 0. 5.0466 0. 5.11 0. the floor displacement demands in Fig. wherein the first two are obvious and the third is determined from response spectrum analysis of the building (Appendix C).3b and Table 5.1.1.126 0.3a and 5.11 0.4a.0197 0. The errors in the FEMA and MPA estimates of seismic demands relative to the “exact” demands are presented in Fig. the story drift demands in Fig. and the locations of all plastic hinges in Fig.11 0. pushover analyses are implemented for a target roof displacement of 52.1 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. plastic hinge rotation demands in Table 5.4.0654 0. The three FEMA force distributions are presented in Fig. 5.042 0.2.4.11 0.0896 0. with the errors being larger for the SRSS distribution.

5.14 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. V = 6897 kN. α = 0. α = 0. (b) ELF.8 cm. V = 8530 kN.24 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized 10 20 30 40 Roof Displacement (cm) 50 60 Fig.2.(a) Uniform Distribution 12000 10000 u = 33 cm. Pushover curves using three force distributions in FEMA-273: (a) “uniform”.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. α = 0. gravity loads are included 57 . and (c) SRSS.3 cm. V = 7456 kN.

MPA including three “modes. with errors reaching 37% in this example. This seems to be an inherent limitation of pushover analysis. Figure 5.3. with story drifts under estimated by.2 demonstrate that the story drift demands are greatly underestimated by all the FEMA force distributions. gravity loads included 58 . the MPA procedure is more accurate than all the FEMA force distributions. In contrast.) The pushover analysis procedures considered seem incapable of computing accurately local response quantities. Heightwise variation of floor displacements and story drift ratios estimated using FEMA-273 force distributions. 5. 7%. reaching 31%. whereas nonlinear RHA computed an insignificantly small value. such as hinge plastic rotations. but it is still inaccurate. reaching 64%.5 1 1. (The 100% error in a 6th floor hinge is ignored as it simply represents that the MPA estimated zero rotation. For the ELF distribution. 3 "Modes" FEMA Uniform ELF SRSS 0.4c and Table 5.” and NLRHA.5 1 1.3 demonstrate that the hinge plastic rotations estimated by all three FEMA force distributions contain unacceptably large errors. at most.5 Ground 0 Ground 0 Fig. For the SRSS distribution. errors are largest in the upper stories. Modal pushover analysis procedure gives estimates better than all the FEMA force distributions. reaching 35%. larger errors are noted in the upper and lower stories.4b. For the uniform distribution. and Table 5. 3 "Modes" FEMA Uniform ELF SRSS 0. the errors are largest in the lower stories.5 Displacement/Height (%) 2 Floor 5th 5th 4th 3rd 2nd 1st NL−RHA MPA.Figures 5. and overestimated by no more than 32%.5 Story Drift Ratio (%) 2 2. (a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th (b) Story Drift Ratios Floor 4th 3rd 2nd 1st NL−RHA MPA.3b and 5.

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

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

6 14.9 15.007 1.830 2.913 0.6 4.154 1.8 -32.1 -26.5 61 .5 -33.9 -77.1 0.2 9.4 0.65E-03 7.5 10.0 -100.8 -35.7 MPA -2.530 1.178 1.4 16.875 0.45E-03 3.26E-03 3.330 1.09E-03 4.8 17.4 -55.2 16.11E-03 9.51E-03 4.168 1.6 21.00E+00 0.372 1.8 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 0.877 1.736 0.3 -14.0 -100.6 -41.0 -57.4 MPA -2.083 1.35E-10 3.04E-02 8.061 1.5 -32.938 1.487 0.927 1.708 0.199 27.6 -25.6 -73.998 1.414 1.2 -100.270 0.5 -3.524 0.00E+00 0.998 1.306 1.234 1.8 2.436 Error (%) NL Uniform RHA 0.9 11.310 1.562 1.4 -23.0 -30.462 1.6 -17.0 -100.93E-03 1.78E-03 0.6 -6.783 1.3 29.7 -15.3 Peak values of hinge plastic rotations (radians) from FEMA force distributions and MPA Hinge Plastic Rotation (rad) MPA FEMA ELF SRSS 4.263 0.1 10.3 -11.88E-03 1.8 -100.2 5.62E-03 0.00E-03 5.836 0.854 0.294 1.55E-03 3.2 10.128 1.5 -33.262 1.8 -2.58E-04 6.3 -100.730 1.17E-03 9.2 4.566 1.367 1.53E-02 1.7 -29.2 16.7 -60.00E+00 0.Table 5.5 -29.0 37. Peak values of floor displacements (as % of building height = 37.314 1.888 0.4 21.6 -4.52E-04 1.34E-03 2.8 7. gravity loads included Displacement /Height (%) FEMA ELF SRSS MPA 0.3 -22.318 2.341 1.089 1.8 -63.214 1.03E-03 5.353 1.7 FEMA ELF SRSS -22.789 0.221 1.686 0.2 -4.23E-03 3.2 6.809 0.344 0. gravity loads included Displacement /Height (%) FEMA ELF SRSS MPA 1.860 1.1.195 0.6 22.6 -11.560 1.7 Table 5.9 SRSS -22.547 -27.9 -20.3 -4.8 16.992 1.3 31.9 -70.7 21.515 FEMA Uniform ELF 27.667 0.2 19.6 17.5 Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 2.9 28.9 16.7 26.530 1.00E+00 Error (%) FEMA Uniform ELF SRSS MPA 24.35E-03 8.59E-03 5.4 16.9 16.19E-03 1.673 Error (%) NL RHA 1.5 26.5 15.8 44.0 -100.109 1.623 1.8 -100.098 1.399 1.490 0.597 0.00E+00 0.466 0.84 1.335 2.00E+00 0.4 -50.858 1.5 22.78E-03 1.00E+00 2.975 1.00E+00 NL RHA 1.067 0.23E-02 1.94E-03 2.10E-02 7.209 0.16E-03 0.75E-03 0.399 -27.015 0.19E-04 5.2 9.00E+00 8.351 0.7 7.011 1.839 0.0 Table 5.7 19.323 1.50E-03 0.1 163.00E+00 0.14 m) from FEMA force distributions and MPA.672 1.417 1.3 23.4 -12.9 0.00E-10 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 1.724 0.5 -29.13E-03 5.064 1.0 -63.7 -28.1 -3.984 1.2 Peak values of story drift ratios (as % of story height) from FEMA force distributions and MPA.611 0.399 1.953 0.207 1.5 -27.00E+00 4.355 0.0 -59.2 -12.2 13.0 -71.

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

5 1 1. 5.5 1 1.5 2 GM Multiplier 2.5 1 1.5 3 Fig.5 3 0 0 0.5 2 GM Multiplier 2.(a) FEMA: Uniform 60 50 40 60 50 40 3 (b) FEMA: ELF Error Envelope Error for Floor No.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. Errors in floor displacements from three force distributions in FEMA-273 and from MPA including three “modes”.5 1 1.5 2 GM Multiplier 2.6.5 2 GM Multiplier 2. gravity loads included 63 . 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.(a) FEMA: Uniform 100 60 50 9 8 (b) FEMA: ELF Error Envelope Error for Story No. Noted 80 40 9 1 Error (%) 7 Error (%) 60 30 20 10 2 5 6 4 8 40 1 6 20 2 5 3 4 3 7 0 0 0. Error in story drifts from three force distributions in FEMA-273 and from MPA including three “modes”.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. 5.5 3 0 0 0. gravity loads included 64 .5 1 1.5 1 1.5 2 GM Multiplier 2.5 2 GM Multiplier 2.7.5 1 1.5 3 0 0 0.5 2 GM Multiplier 2.5 3 Fig.5 2 GM Multiplier 2.

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

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

plastic hinge rotations. and nonlinear RHA. The MPA procedure is more accurate than all the FEMA force distributions in estimating floor displacements. 67 . However. Thus the MPA procedure is accurate enough for practical application in building evaluation and design. and hinge plastic rotations. and identifying locations of most plastic hinges.25 to 3. Pushover analyses seem to be inherently limited in computing accurately 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. story drifts. the errors in the MPA procedure are shown to be only weakly dependent on ground motion intensity. such as hinge plastic rotations. Comparing the earthquake-induced demands for the selected 9-story building determined by pushover analysis using three force distributions in FEMA-273.0. etc. 8. 7. MPA. story drifts.(iv) At the roof displacement determined from Dn . Comparing the peak inelastic response of a 9-story SAC building determined by the approximate MPA procedure—including only the first two or three rno terms—with rigorous nonlinear RHA demonstrates that the approximate procedure while providing good estimates of floor displacements and story drifts. the pushover analysis provides the peak value rno of any response quantity: floor displacements. demonstrates that the FEMA force distributions greatly underestimate the story drift demands. it fails to compute with acceptable accuracy plastic rotations of the hinges. 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. and lead to unacceptably large errors in the hinge plastic rotations.” 10. Based on results presented for El Centro ground motion scaled by factors varying from 0. joint rotations. all pushover analysis procedures do not seem to compute with acceptable accuracy local response quantities. 9.

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

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

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

ω n . 1988).g. Idealize the pushover curve as a bilinear curve (Fig. 3. DRAIN2DX (Allahabadi and Powell. e. and modes. φn . Apn .. using any numerical integration method. respectively. B.2. Calculate the area under the actual pushover curve. Since the target roof displacement may not be known at the start of the procedure. This step can be conveniently implemented in any commercially available software. Define the anchor point. The structure should be pushed just beyond the target (or expected) roof displacement in the selected mode. Let the roof displacement and base shear at the anchor point be urno and Vbno .g. iterations may be necessary. for linear-elastic vibration of the building. e. trapezoidal rule.1. Define the force distribution s* from Eq. of the bilinear curve at the target roof displacement. develop the base-shear – roof-displacement (Vbn . 71 .1) using the FEMA-273 procedure (Building Seismic Safety Council. (3.1 STEP-BY-STEP PROCEDURE A detailed step-by-step implementation of the uncoupled modal response history analysis (UMRHA) procedure is presented in this section and illustrated by an example in the next section. Apply force distribution of Step 2.. 3.2. 1977).urn ) pushover curve for the force distribution s* : n 2. 2. 3. A. Compute natural frequencies.Appendix A Uncoupled Modal Response History Analysis A. For the nth-“mode”. 1.1 incrementally and record the base shears and associated roof displacements.20): s* = mφn n n 2.1.

will be refined by an iterative procedure that seeks to equate areas under the actual and the idealized pushover curves. 3. and a point on the actual pushover curve with base shear equal to i 0. obtained by judgment. i +1 i i 3. Calculate the post-yielding strain-hardening ratio. Calculate area under the bilinear curve OAB. i 3.9.1. at base shear i equal to 0.0. This value. determine the roof displacement. kn = 0. by connecting a straight line between origin. From the pushover data. i i i αn = Vbno Vbny − 1 urno urny − 1 i 3.6 × Vbny . If the error exceeds some pre-specified tolerance.i 3. Vbny . 3. i 3. urn. Abn . i 3. Calculate Vbny = Vbny × Apn Abn .2.3.5.4. and B with straight-line segments to obtain the idealized bilinear curve.2.9. Calculate the error = 100 × Abn − Apn ( ) ( ) ( ) ) Apn . urny = Vbny kn . Draw the curve OAB by connecting the three points O. iterations are necessary. A. Let the point with base shear = Vbny and roof displacement = urny be denoted as A. Replace i+1 with i and repeat Steps 3. O. 72 . corresponding to the estimated yield i i i base shear.8.6 .6 × Vbny ( ) i urn .4. Calculate initial slope of the idealized bilinear curve. If desired. Vbny . Estimate the yield base shear.0. other appropriate methods can be ( used.9.7.6 . i i 3.6 × Vbny .4.8. Calculate the yield displacement.1.4 to 3. i i i 3.6. This step gives the secant stiffness at a base shear equal to 60% of the yield base shear. Calculate the slope. kn . 3.

15) and (3. Combine the “modal” responses using Eqs. In general first two or three modes will suffice.6 × Vbny A • Idealized αnkn B • 1 Actual k 1 • O n figA_1.10b and 4. 4. Scale the vertical axis with * Mn to obtain * Fsno Ln = Vbno M n and * Fsny Ln = Vbny M n (Eqs. Develop the Fsn Ln − Dn relation (Fig. Repeat Steps 3 to 6 for as many modes as required for sufficient accuracy. and pseudo-acceleration history.13). A.0.16). 8. 4.11a).6 ur n y ur n o ur n Fig. 9. Scale the horizontal axis by Γnφrn to obtain Dno = urno Γ nφrn and Dny = urny Γ n φrn (Eqs.12) and (3. r o . 4.1. (3. 6. of the combined responses obtained in Step 8.4. Dn (t ) . Compute deformation history. 5. Calculate histories of various responses using Eqs. An (t ) . * 4. (3. A.11b).1 Idealization of nth-“mode” pushover curve 73 . Vbn Idealization of Pushover Curve Vbno Vbny 0.2). (3.4) and effective modal mass from M n = Ln Γn .3.2.10a and 4. Compute the Ln and Γ n from Eq. A. 7.3b) with unit-mass and force-deformation relation of Fig. Calculate peak values. 4. 4.eps ur n.2. of the nth“mode” inelastic SDF system (Fig.

2. for the first. A. Ap1 = 360777 kN-cm. respectively. Idealized bilinear curves for each of the three modes are included in Fig. s* .1. and third mode.6 kN. A. The base-shear – roof-displacement (Vbn . First three mode shapes and frequencies of the selected building were computed and are shown in Fig.5 cm (25 in. generated using DRAIN-2DX.3.5 cm and Vb1o = 8729.3. to the north-south component of the El Centro (1940) ground motion scaled up by a factor of 1. is defined at the target roof displacement. n 2. Following is step-by-step implementation of the procedure described in Section A. 74 . B. The force distributions.2 Properties of the nth-“mode” inelastic SDF system A.2. The pushover curves for the first three modes. A. 25.2 EXAMPLE The UMRHA procedure is implemented to calculate the response of the 9-story building described in Section 3.7 cm (5 in.1. 3.5.4.eps Dny = ur n y / Γnφr n Dno = ur n o / Γnφr n Dn Fig. 3.1. 3. 3. and 12. 1.Fsn / Ln Fsn / Ln − Dn Relationship Vbno / M* n Vbny / Mn * 1 α ω2 n n ω2 1 n figA_2. computed for the first three modes are shown in Fig. 3.urn ) pushover curve for the force distribution s* : n 2. At this point.2.). are shown in Fig.). second.4 cm (10 in. The anchor point.3. Area under the actual pushover curve.).4. ur1o = 63. Target displacements used to generate these pushover curves are 63. The following steps illustrate the procedure to develop the idealize curve for the first “mode”.

9. i i 3. Also included are the modal damping ratios and the periods calculated from Eq.09 ) − 1 = 0.1.6 × Vb1 y = 4803. 75 . i i i 3. The results for other modes are summarized in Table A.2. (4. i 3.198%. The procedure converged after nineteen cycles to give ur1 y = 36.2.6 kN. i i i 3. The Fs1 L1 − D1 relation for the first “mode” is developed as follows. * 4.3666 = 3740189 kg. The first estimate of the yield base shear Vb1 y = 8006. Area under the bilinear curve OAB.3666. is calculated as follows.0. Vb1 y = 7615. ur1.8. The yield displacement.1.4.i 3. i i 3.4 kN. Γ1 = 1. i 3. and M1 = 2736789 × 1.86 = 210.18 kN/cm. Ab1 = 365100 kN-cm. ur1 y = Vb1 y k1 = 8006. The next estimate of the yield shear is Vbny = 8006.86 cm at 0. 4.1. k1 = 0.4 210.09 cm and Vb1 y = 8006. L1 = 2736789 kg.13).3. The point A on the i i bilinear curve is defined by ur1 y = 38.8 kN.6 8006.9 kN.5 38. This value exceeds the prespecified tolerance of 0.194. α1 = Vb1o Vb1 y − 1 ur1o ur1 y − 1 = ( ) ( ) (8729.135.6 = 4803.4 × (360777 365100 ) = 7911. Therefore. and α1 = 0.7.4 kN.6 × Vb1 y ( ) i ur1.09 cm. The results of the iterative procedure are summarized in Table A. A. The post-yielding strain-hardening ratio.8 22. The curve OAB obtained by connecting the three points O. Determined from the pushover database.4. 3.0.6. i +1 3. k1 . 3.9.23 cm.4 ) − 1 (63. Error = 100 × (365100 − 360777 ) 360777 = 1. 3.18 = 38.6 = 22.01%.2. iterations are necessary.9.5.1. and B with straight-line segments gives the idealized bilinear curve. The initial slope of the idealized bilinear curve.4.

Histories of roof displacement and top story drifts for the first “mode” are computed and presented in Fig.7.51 cm.4. 9. Scaling the vertical axis by Fs1 y L1 = 203. 8. 4. 4.7.7.2.1 and 4. The combined modal responses are presented in Fig. The peak values are also plotted in Fig. 4.4. 4.8. 6.40 (cm/sec2) and 76 . 4.46 cm and D1y = 26. Deformation and pseudo-acceleration histories of the inelastic SDF systems for the first “mode” with unit mass and force-deformation relation developed in Step 4 are plotted in Fig.3. 7. The peak values are computed and are summarized in Tables 4. 5. * M1 gives Fs1o L1 = 233. Scaling the horizontal axis by Γ1φr1 gives D1o = 46.2. A. The results were generated for first three “modes’ and are included in Fig.62 (cm/sec2).

V y by = 4952 kN.6 cm. V = 5210 kN.19 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized figA_3a. α = 0.2 cm.3.eps 0 0 2 4 6 8 10 Roof Displacement (cm) 12 Fig. “Modal” pushover curves for the example building 77 . α = 0.14 y by 6000 Base Shear (kN) 4000 2000 Actual Idealized figA_3c.eps 10 20 30 40 50 Roof Displacement (cm) (b) "Mode" 2 Pushover Curve 60 8000 u = 9.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36.9 cm.eps 0 0 5 10 15 20 Roof Displacement (cm) (c) "Mode" 3 Pushover Curve 25 8000 u = 4. A.13 6000 Base Shear (kN) 4000 2000 Actual Idealized figA_3b. α = 0. V = 7616 kN.

” second “mode.06 0 • 1.m −80 0 −3 0 25 30 Fig.747 5 10 15 20 Time (sec) figa_4.(a) Deformation 80 n=1 3 (b) Pseuso−acceleration n=1 • 0.2769 D1 (cm) 0 • 35. Histories of deformation and pseudo-acceleration due to 1.223 −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. 78 . A.5 × El Centro ground motion for the first “mode.4.52 5 10 15 20 Time (sec) 25 30 0 • 1.” and third “mode” inelastic SDF systems.

3 7658.09 18.18 210.9 (kN) 4803.3 7628.2 4628.76 21.4927 1.948 ζ n (%) 79 .4 4595.693 0.18 210.176 0.29 36.35 36.90 21.5 3109.3 7786.59 36.037 0.12 3876.188 0.95 21.404 0.6 7840.198 0.75 21.18 210.26 36.0.5 7633.30 37.11 22.5 7624.18 210.86 21.85 36.9 4573.103 “Mode” 3 696400 0.5309 488839.186 0.18 210.136 Fsny Ln (cm/sec2) Dny (cm) Fsno Ln (cm/sec ) Dno (cm) Tn (sec) 233.063 0.180 0.65 1226.1 4574.4 7647.18 210.38 22.39 36.78 21.18 210.6 × Vb1 y i ur1.309 0.1.0 4704.1 1013.6 i k1 i ur1 y i α1 i Ab1 Error (%) 1.022 0.18 210.18 210.9 4570.05 52.2.029 0.7 4580.18 210.18 210.62 26.64 37.Table A.010 (kN) 8006.194 0.5 4614.0 4577.162 0.0 4588.193 0.44 36.75 21.70 36.6 4583.237 0.23 0.9 7615.193 0.56 19.082 0.2 4571.25 36.184 0.23 22.40 46.32 36.8 4647.81 21.59 22.191 0.910 0.139 0.1 4569.182 0.194 (kN-cm) 365100 364060 363276 362684 362235 361892 361631 361432 361279 361162 361073 361004 360951 360911 360880 360856 360838 360824 360813 Table A.09 37.79 0.2671 1.18 210.017 0.194 0.013 0.77 21.0 7619.4 7714.151 0. Properties of “modal” inelastic SDF systems Properties Ln (kg) Γn * M n (kg) “Mode” 1 2736789 1.193 0.25 36.8525 1. i 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 i Vb1 y i 0.24 36. Results of iterative procedure to develop the idealized bilinear curve for the first “mode” inelastic SDF system Itr.8 7618.74 (kN/cm) 210.28 36. No.2 7690.2406 167531.79 21.74 21.18 210.18 210.3 4603.7 7639.529 0.18 210.3 7745.190 0.05 36.9 4570.83 21.18 210.18 210.1 7616.192 0.50 36.194 0.107 0.86 22.02 21.46 2.5 (cm) 22.4 7672.135 0.18 210.8 7622.048 0.2 4671.170 0.3666 3740189 203.51 2 “Mode” 2 -920860 -0.56 47.85 0.18 (cm) 38.4 7911.8 4747.

80 .

Steps 1 to 4 of the MPA are the same as those for UMRHA presented in Appendix A. Repeat Steps 3 to 7 for as many “modes” as required for sufficient accuracy.8). 4. Solving Eq. 10. Compute the peak deformation.Appendix B Modal Pushover Analysis B. (4. B.5. This is the same example as solved in Appendix A. 6. 10. 4.1. 12. (4. Dn . or from the inelastic response (or design) spectrum.3b) with unit mass and force-deformation relation of Fig. (3. of the nth-“mode” inelastic SDF system (Fig.6b by solving Eq. Calculate the peak roof displacement urno associated with the nth-“mode” inelastic SDF system from Eq. 81 . At urno . (3. 13. 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.33 cm.18). the first two or three “modes” will suffice.21). rno . Following is step-by-step implementation of the procedure described in Section B.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. 11.1 to the north-south component of the El Centro (1940) ground motion scaled up by a factor of 1.4. Typically. extract from the pushover database values of other desired responses. From the total hinge rotation.2 EXAMPLE The MPA procedure is implemented to calculate the response of the 9-story building described in Section 3. subtract the yield hinge rotation to determine the plastic hinge rotation.

11. 82 .5 are the plastic hinge rotations computed by subtracting the yield hinge rotation from the total hinge rotations. 8. 12. 9.4. Steps 3 to 7 are repeated for first three “modes. Also included in Table 4.4. Results of Steps 5 and 6 are summarized in Table B. where results for other ground motion intensities are also included. The values are summarized in Table 4.28 cm.3 and 4. (3. Peak roof displacement ur1o = Γ1 × φr1 × D1 = 1. values of floor displacements and story drifts are extracted.3 and 4.18) are also included in Tables 4.1. At ur1o = 48. The total response computed by combining the peak “modal” responses using the SRSS combination rule of Eq.” and the results are included in Tables 4.33 = 48.3 for the floor displacements normalized by the building height and in Table 4.366 × 1 × 35.28 cm.4 for the story drifts normalized by the story height.

73 24.75 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.275 1.185 11.678 0.367 1.184 0.690 10.25 Quantity Dn (cm) “Mode” 1 6.139 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.36 1.0 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 3.252 9.52 “Mode” 3 1.117 5.535 14.25 20.660 14.85 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 1.457 12.5 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.59 0.28 46.577 16.71 1.229 8.07 “Mode” 2 4.05 0.007 36.82 1.913 22.551 2.52 0.126 13.Table B.50 35.901 8.200 0.06 1.023 0.735 3.268 0.755 0.225 2.154 78.35 0.332 48.0 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 1.03 0.766 7.691 0.504 18.03 26.27 0.4222 3.312 1.0 µ = Dn Dny urno = Γnφrn Dn (cm) 83 . Calculation of roof displacements urno from peak deformation of inelastic SDF systems Ground Motion Multiplier 0.755 27.436 7.5 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 2.18 27.267 5.79 0.450 4.33 1.856 31.1.395 0.35 1.513 0.13 2.8451 5.969 0.37 1.37 57.332 13.467 14.676 6.379 21.70 0.38 22.748 63.

84

**Appendix C FEMA Force Distribution Calculations
**

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

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

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

Floor, j

mj (10-3×kg)

s* = j

∑i mi

0.112 0.110 0.110 0.110 0.110 0.110 0.110 0.110 0.119

mj

1 2 3 4 5 6 7 8 9

503.5 494.7 494.7 494.7 494.7 494.7 494.7 494.7 534.1

C.2

EQUIVALENT LATERAL FORCE (ELF) DISTRIBUTION 85

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

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

Floor j 1 2 3 4 5 6 7 8 9

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

-3 k

s* j

=

m jhk j

∑i mihik

0.007 0.020 0.038 0.062 0.091 0.126 0.165 0.210 0.281

C.3

SRSS DISTRIBUTION

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

N

∑ n (V jn )

2

to get column 8 of

2 105.0 1476.8 1381.1 87.2 277.3 240.1 -438.0 136.3.5 215.8 -326.0 381.7 1578.9 -166.090 0.6 1233.8 430.2 95.7 2065.7 -46.0 176.7 374.7 234.6 319.0 -5.9 446.7 (10) 0.2 97.9 366.6 -732.2 200.5 159.7 Vj (kN) (8) 2268.6 V j3 (kN) (7) 478.4 1759.For convenience.5 -350.367 87 .045 0.8 374.1 832.177 0.1 -646.9 880. Table C.4 1842.4 400.0 980. 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) 1917.9 832.6 -359.5 -320.042 0.7 95.4 250.070 0.5 -973.3.1 -525.7 694.2 Lateral Force fj (kN) (9) 203.1 1857.5 320.9 1683.6 366.0 1231.047 0.3 222.9 1446.1 Story Shears V j2 (kN) (6) 1114.9 -153.6 -352.7 355.7 525.7 1622.3 -646.6 286.065 0.1 -967.7 101.0 354. the lateral forces are normalized by the base shear to obtain column 10 in Table C.098 0.2 285.3 -6.2 148.2 fj = ∑ fi i fj (2) (3) (4) 59.

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

Are you sure?

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

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

Get the full title to continue

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

scribd