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

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

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

Pacific Earthquake Engineering Research Center

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

Theory and Preliminary Evaluation

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

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

PEER 2001/03

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

Anil K. Chopra University of California Berkeley

Rakesh K. Goel California Polytechnic State University San Luis Obispo

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

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

ii .

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

iv .

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

vi .

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

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

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

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

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

sign u = .1) is divided by m to obtain ~ u + 2zw n u + w 2 u y f s u.. in addition to the form of the force-deformation relation. deformation is denoted by um .e.ug t n where bg bg b g bg (2. This becomes evident if Eq. Section 7. or absolute (without regard to algebraic sign) maximum. and R y (Chopra . bg 4 . and z is the damping ratio of the system vibrating within its linear elastic range (i.5) For a given u g t . The peak. 2001. and the ductility factor is u m= m uy (2. and (b) bilinear hysteretic forcedeformation relation For a given excitation u g t . (2.3) ωn = k m ζ = c 2mω n f fs = s fy (2. (a) Idealized one-story structure. z y . and u y .3). z . u £ u y ). m depends on three system parameters: w n .4) and w n is the natural vibration frequency.fS m fs c ut u fy 1 αk k 1 uy ug -fy k k um u 1 1 (a) (b) Fig. the deformation u t depends on three systems parameters: w n . 2.1. Tn = 2p w n is the natural vibration period.

04 40 Base Shear. The yield strength of the inelastic system.2. 2. based on Ry = 8 . this system and excitation. m h = 3. 2001) and scaled up by a factor of 2. 2.2 kips) .826 kips) for w = 169.2 5 . Pushover curve for structure shown in Fig. α = 0. with Tn = 0.26 kN (8. For . f o w = 184 .2: the dimensions and flexural rigidity of the structural elements are noted. is f y / w = ( f o / w ) ÷ 8 = 0.2 SYSTEM AND EXCITATION CONSIDERED Consider the one-story system in Fig.1. Vb (kN) 30 20 10 0 0 2 4 6 Roof Displacement.2311 . 2. One-story. and f y = 39.26 kN. Vby = 39.3.376 cm.2.32 m • Fig.4 in Chopra.9 kN (38.5 sec and z = 5% subjected to the north-south component of the El Centro (1940) ground motion (Fig. one-bay frame 50 uy = 1. 2.66 m • EIb EIc EIc • • L = 7. 6. u (cm) 8 Fig.

134 × 107 mm 4 (75. 2. However.) and 50. the energy dissipated in 6 .65 kN-m (191.g. (c) shows the joint rotation q bt g . 2.35.. It is organized in five parts: (a) shows the deformation u t .1 kip-in. 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 . the ductility factor m = 5.4edetermined from RHAis superimposed. 2. I b = 3.3 RESPONSE HISTORY ANALYSIS Figure 2. (d) shows the rotation q p bt g of the plastic hinges at the bg bg beam ends. the peak value determined from RHA.The yield moments in the beam and columns are defined as the bending moments due to the lateral force f y . Determined from pushover analysis at the exact peak deformation. wherein the hysteretic force-deformation history of Fig.4 shows the earthquake response of the system described in the preceding section determined by response history analysis (RHA). (b) shows the lateral force f s t or base shear Vb t normalized relative to the bg weight.0217 rad . A static pushover analysis of this one-story system leads to the force-displacement relationship shown in Fig. The resulting pushover curve is shown in Fig. 2.4 ). The peak values of the various response quantities are as follows: um = 7. 2. and (e) shows the force-deformation relation. Implementing this analysis with I c = 6.017 rad . pushover analysis cannot provide any cumulative measure of response. The yielding stiffness of each structural element is defined as 3% of its initial stiffness. respectively. as apparent in Fig.6 kip-in. and q pm = 0.) for the beam and columns.36 cm. 2.4f.3 in. q m = 0.18 kN-m (444.4 PUSHOVER ANALYSIS Static analysis of the one-story nonlinear system subjected to lateral force that increases in small increments is implemented until the lateral displacement reaches um = 7.3. and E = 2 × 108 kPa (29×103 ksi) gives yield moments of 21.4 ).077 × 107 mm 4 (146 in. e.36 cm . The system is excited well beyond the yield deformation.4e. the joint rotation and beam hinge rotation are identical to values q m and q pm determined from RHA. Observe that the pushover curve matches the initial loading path of the hysteretic system.

2 −0. (c) joint rotation.4 0. (e) force-deformation relation. (d) plastic hinge rotation.2 Vb / w 0. 15 u (cm) 0 • 7.04 0.4 0.5 0.2 0 (e) 0 (f) −0.04 0.4. and (f) pushover curve 7 .yielding during the ground motion.2311 y (b) −0. 2.017 • 0 (d) θp (rad) −0. Response of one-story system to El Centro ground motion: (a) deformation.0217 • θ (rad) 0 (c) −0.36 −15 0.2 −0. This represents an inherent limitation of pushover analyses.04 0. or the cumulative rotation at a plastic hinge.2311 (a) Vb / w 0 f / w = −0.04 0 5 10 15 Time (sec) 20 25 30 0. (b) base shear.5 fy / w = 0.4 −10 −5 0 u (cm) 5 10 Fig.4 −10 −5 0 u (cm) 5 10 −0.

8 .

and p eff t = Â peff.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 . c.4) 9 .m ι u g t bg bg (3.12): mι = n =1 Â sn = N n =1 Â Γ nmφ n N (3. each element of the influence vector ι is equal to unity.snug t n= n =1 The effective earthquake forces can then be expressed as af N af N af (3.3.n t = Â . and lateral stiffness matrices of the system. This force distribution can be expanded as a summation of modal inertia force distributions s n (Chopra. m.mι ug t bg bg bg (3. classical damping.1) can be interpreted as effective earthquake forces: p eff t = . 2001: Section 13. (3. The right side of Eq. and k are the mass. Elastic Multistory Buildings 3.3) where φ n is the nth natural vibration mode of the structure.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 = .1) where u is the vector of N lateral floor displacements relative to the ground.

Then the floor displacements are u n t = φ n qn t bg bg bg bg (3.sn ug t bg bg bg bg (3. it can be demonstrated that none of the modes other than the nth mode contribute to the response.11) and substituting in Eq.8) where the modal coordinate qn t is governed by 2 qn + 2z nw n qn + w n qn = .n (t ) = N n =1 ∑ −sn u g (t ) N (3. (3.peff (t ) = n =1 ∑ peff .10) bg bg bg 10 (3.n t is entirely in the nth-mode.u g t n Comparing Eqs. with no contribution from other modes.6) Next. p eff .Gn ug t (3. (3. subjected to u g t : Dn + 2z nw n Dn + w 2 Dn = .9) to the equation of motion for the nth-mode elastic SDF system.n t = .9) in which w n is the natural vibration frequency and z n is the damping ratio for the nth mode. an SDF system with vibration propertiesnatural frequency w n and damping ratio z n of the nth-mode of the MDF system. (3.9) and (3. The equations governing the response of the system are mu + cu + ku = . The solution qn t can readily be obtained by comparing Eq. we will outline that the response of the MDF system to p eff .5) The contribution of the nth mode to s and to p eff t are: sn = Gn mφ n respectively.7) By utilizing the orthogonality property of modes.8) gives the floor displacements u n t = Gnφ n Dn t bg (3.sn ug t bg (3.12) .10) gives qn t = Gn Dn t bg bg bg (3.

and An t = w 2 Dn t n bg (3. In contrast to the classical 11 bg .9) is the standard modal equation governing qn t .15) and (3.15) rt = bg n =1 Â rn t = bg n =1 Â rnst An bt g N (3. 3. ζn ¨ g(t ) u (a) Static Analysis of Structure (b) Dynamic Analysis of SDF System Fig. 2001.1). 3.12) and (3. Section 13.can be expressed by st rn t = rn An t bg bg bg bg bg (3. (3.13) represent the response of the MDF system to p eff. Conceptual explanation of modal response history analysis of elastic MDF systems Any response quantity r t story drifts.Forces sn An(t ) st rn ωn.16) This is the classical modal RHA procedure wherein Eq. the static value of r due to external forces sn . Eqs.13) define the contribution of the nth-mode to the response. (3. However. 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. (3. these standard equations have been derived in an unconventional way.13) st where rn denotes the modal static response.n t . Equations (3. st The two analyses that lead to rn and An t are shown schematically in Fig.14) is the pseudo-acceleration response of the nth-mode SDF system (Chopra. and Eqs.1.12) and (3.1. etc. internal element forces. Therefore.16) reflect combining the response contributions of all modes.

1).8. the peak nth-mode response as in Eq. which is valid for structures with well-separated natural frequencies such as multistory buildings with symmetric plan.9) . In such a response spectrum analysis (RSA). The SRSS rule.16).3). 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. (3. provides an estimate of the peak value of the total response: b g F N r 2 I 1/ 2 ro ª G Â no J H n =1 K 3.2 MODAL RESPONSE SPECTRUM ANALYSIS The peak value ro of the total response r t can be estimated directly from the response spectrum for the ground motion without carrying out the response history analysis (RHA) implied in Eqs.17) (Chopra.19) will provide the same value of rno . 2001. z n of the pseudo-acceleration response (or design) spectrum for the nth-mode SDF system. (3. and Tn = 2p w n is the natural vibration period of the nth-mode of the MDF system.derivation found in textbooks (e. Chopra.g.1. Section 13.17) where An is the ordinate A Tn ..(3. The peak modal responses are combined according to the Square-Root-of-Sum-ofSquares (SRSS) or the Complete Quadratic Combination (CQC) rules.18) To develop a pushover analysis procedure consistent with RSA.4 and 13.3 MODAL PUSHOVER ANALYSIS (3. 2001. 3. This concept provides a rational basis for the modal pushover analysis procedure developed later. Sections 12. this response value can be obtained by static analysis of the structure subjected to lateral forces distributed over the building height according to 12 . we used the modal expansion of the spatial distribution of the effective earthquake forces. we note that static analysis of the structure subjected to lateral forces f no = Γ n mφ n An (3. Alternatively.

1 System and Excitation Considered The 9-story structure. 1660 W. in both directions. The building’s lateral load-resisting system is composed of steel perimeter moment-resisting frames (MRFS) with simple framing on the farthest south E-W frame. the peak value of the roof displacement due to the nth-mode. 46th Street. which from Eq. and California Universities for Research in Earthquake Engineering (CUREE). The levels of the 9-story building are numbered with respect to the ground level (see Fig. 1301 S. 3. 3.. this building is 45. urno . The bays are 9. Richmond. CA 94804-4698. was designed by Brandow & Johnston Associates1 for the SAC2 Phase II Steel Project. can be combined according to Eq. and 37. each determined by one pushover analysis.21) 2 where Dn = An ω n . The interior bays of the structure contain simple framing with composite floors. denoted B1. region. Typical floor-to-floor heights (for analysis purposes measured from center-of-beam to center1 2 Brandow & Johnston Associates. Los Angeles. Third St. California.73 m (150 ft) by 45.20) and the structure is pushed to the roof displacement. SAC Steel Project Technical Office.73 m (150 ft) in plan.18) to obtain an estimate of the peak value ro of the total response. shown in Fig.19 m (122 ft) in elevation. CA 90017.2). Consulting Structural Engineers. Obviously Dn and An are available from the response (or design) spectrum. the Applied Technology Council (ATC).4.s* = mφ n n (3. The building has a basement level. A benchmark structure for the SAC project. The peak modal responses. this structure meets seismic code and represents typical medium-rise buildings designed for the Los Angeles. with the ninth level being the roof. with five bays each in the north-south (N-S) and east-west (E-W) directions.4 COMPARATIVE EVALUATION OF ANALYSIS PROCEDURES 3. SAC is a joint venture of three non-profit organizations: The Structural Engineers Association of California (SEAOC).2). Although not actually constructed.15 m (30 ft) on center. (3. (3. rno . The columns are 345 MPa (50 ksi) steel wideflange sections.12) is urno = Gnf rn Dn (3. 13 .2. This modal pushover analysis (MPA) for linearly elastic systems is equivalent to the well-known RSA procedure (Section 3. 3.

2 kips-sec2/ft).7 kips-sec2/ft). The building is modeled in DRAIN-2DX (Allahabadi and Powell.0 kips-sec2/ft). the vibration periods are 2. 14 .65×105 kg (66.sec2/ft). including the steel framing. The seismic mass of the structure is due to various components of the structure. mechanical/electrical. are located on the first.49 m (18 ft). for the first three modes are shown in Fig. ceiling/flooring. 3. The floor-to-floor height of the basement level is 3. fifth.4. we select a weak ground motion: the northsouth component of the El Centro (1940) ground motion scaled down by a factor of 4. s* (Eq. 0. The column lines employ two-tier construction. The strength.3. dimension.65 m (12 ft) and for the first floor is 5.01×106 kg (69. Each frame resists one half of the seismic mass associated with the entire structure. and for the ninth level is 1.49 sec. Column splices.27. This model is based on centerline dimensions of the bare frame in which beams and columns extend from centerline to centerline.07×106 kg (73.96 m (13 ft). i. for the second through eighth levels is 9. The seismic mass of the ground level is 9.of-beam) are 3. respectively. and seventh levels at 1. floor slabs. 3. partitions. The column bases are modeled as pinned and secured to the ground (at the B-1 level). 1988) using the M1 model developed by Krawinkler and Gupta (1998). These force n distributions will be used in the pushover analysis to be presented later. The first three vibration modes and periods of the building for linearly elastic vibration are shown in Fig.00×106 kg (616 kips. for the first level is 1.83 m (6 ft) above the center-line of the beam to column joint. 3. The floor system is composed of 248 MPa (36 ksi) steel wide-flange beams in acting composite action with the floor slab.. third. and 0.85. To ensure that this structure remains elastic. such as those described in Gupta and Krawinkler (1999) can be used. The force distributions. The simple model adopted here is sufficient for the objectives of this study. monolithic column pieces are connected every two levels beginning with the first level. The seismic mass of the above ground levels of the entire structure is 9.e.20). The 9-story N-S MRF is depicted in Fig.2. and shear distortion of panel zones are neglected but large deformation (P∆) effects are included. which are seismic (tension) splices to carry bending and uplift forces. Concrete foundation walls and surrounding soil are assumed to restrain the structure at the ground level from horizontal displacement.89×105 kg (67. 3. roofing and a penthouse located on the roof.0 kips-sec2/ft). if desired more complex models.

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

27 sec Ground −1. determined by RHA [Eqs. 2.38 0.7.05 s * 1 s * 2 s * 3 Fig. and 3 n 3. and 3 . (3.04 1.85 sec 2 Floor 5th 4th 3rd 2nd 1st T1 = 2.93 −1.487 −1. 3.44 1.796 0.12 0.5 Fig. Each figure is organized in four parts: (a) shows the roof displacement urn t .4. 3.12) and (3.1 −2.9th 8th 7th 6th T = 0.728 2.0272 −2.39 3.8 −2.13 −1.51 0.5.13)].03 −1.31 −0. (b) shows the base shear Vbn t normalized relative to the weight W of the building.5 0 0.6. and 3. is shown in Figs.75 1.37 2.94 2.72 −2. respectively.4.2 Response History Analysis The structural response due to individual vibration modes.5 −1 −0. 3.67 −1. 3.49 sec 3 T = 0. First three natural-vibration periods and modes of the 9story building 3. n = 1.5 Mode Shape Component 1 1.1 3. (c) shows the joint rotation q n t of an external joint at 16 bg bg bg .05 2. n = 1.33 2.61 2.3.05 1. Force distributions s* = mφn . 2.

Combining the modal response histories for all modes gives the total response [Eqs. The same method was used to determine the peak values of many response quantities. It is apparent that the first mode alone is inadequate.8. 3. which are listed in Tables 3. two.3. As expected. (3. the results for the roof displacement and top-story drift are shown in Fig. The peak values of the various response quantities are noted in these figures.1.1. The peak values of displacements of all floors.3. and the percentage errors due to truncation of higher modal contributions.12 cm.2. 3.9. respectively. errors are smallest in floor displacements. The errors due to truncating the contributions of vibration modes beyond the third mode are negligible. especially in estimating the story drifts. and rotations of external joints with moment connections are presented in Tables 3. with the first three modes provide essentially the exact response. including one. The peak values of floor displacements and story drifts determined by RHA. Also included are the combined response due to one. errors generally decrease as response contributions of more modes are included. the peak roof displacement due to each of three modes is ur1o = 9.422 cm. and ur 3o = 0. larger in story drifts. The linear relationship between the base shear and roof displacement for each mode implies that the structure did not yield.2. consistent with the increasing significance of the higher mode response among these three sets of response quantities. drifts in all stories. three. This is illustrated in Fig. respectively.8. 3. two. 3.15) and (3.16)]. and 3. or all modes. ur 2o = 2.the roof level. For a fixed number of modes included. but three modes—perhaps even two modes—are sufficient. in particular. 17 . and (d) shows the Vbn . 3. and 3. the exact response considering all modes. where the second and third modal responses are a larger percentage of the top story drift compared to roof displacement.urn relation. and three vibration modes. and even larger in joint rotations. are presented in Fig.23 cm.

173 0.088 -0.264 Error (%) 1 Mode 2 Modes 3 Modes -23.080 0.205 0.9 8.4 -7.097 0.266 0.011 0.413 RHA (all modes) 0.09E-03 Error (%) 1 Mode -23.3 19.2 -2.31E-03 Mode 2 -1.66E-05 -3.265 RHA (all modes) 0.229 0.253 0.157 0.259 0.3 Peak values of joint rotations (radians) from RHA for 0.008 -0.8 1.407 0.378 0.85E-03 3.152 0.199 0.045 0.7 7.50E-03 4.8 -15.11E-03 1.01E-04 -2.156 0.1 Building height is measured from the ground floor to the 9th floor.33E-03 2.6 -0.3 -3.78E-04 -3.173 0.8 -56.071 0.227 0.9 -22.130 0.42E-04 1.94E-03 2.00E-03 2.22E-03 2.7 3.25 × 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 0.042 0.65E-03 RHA (all modes) 2.124 0.63E-03 2.9 1.002 -0.01E-04 3.364 0.069 0.8 1.090 0.152 0.5 18.99E-03 Mode 3 3.6 -1.043 0.26E-04 -5.7 2 Modes -3.475 0.097 0.179 0.009 -0.321 0.177 0.0 -2.012 -0.6 11.9 -16.6 -1.1 3.203 0.370 0.2 9.74E-03 1.4 -1.8 -10.9 3.0 -0.062 0.91E-04 1.65E-03 2.032 -0.060 -0.9 -24.202 0.0 -2.124 0.2 -1.7 2.060 0.9 2.125 0.38E-03 2.310 0.350 0.253 0.25 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 3 Modal Response Mode 1 2.226 0.125 0.311 0.015 -0.0 7.307 0.8 -1.74E-04 6.1 4.054 0.0 -10.4 -10.336 0.069 0.266 0.266 0.038 0.2 0.2 -20.202 0.300 0.0 -46.117 0.156 0.2 1.99E-03 2.121 0.7 4.3 -8.6 1.261 -0.012 0.1 -0.060 -0.31E-03 Joint Rotation (rad) Combined (RHA) 1 Mode 2.09E-03 1.097 0.231 -0.09E-03 1.282 0.197 0.282 0.45E-03 3.042 0.453 0.6 4.237 0. 18 .28E-04 1.90E-03 3.311 0.00E-03 1.058 -0.003 0.03E-03 1.006 0.406 0.258 0.1 Peak values of floor displacements (as % of building height = 37.2 -4.03E-03 3 Modes 2.1 -11.133 -0.263 0.3 Table 3.76E-03 1.47E-03 1.00E-03 1.72E-03 3.235 0.1 -19.9 -23.8 -5.74E-03 1.399 0.235 0.03E-03 -6.400 0.159 0.035 0.5 0.13E-03 2.265 0.3 -0.14E-03 2.9 -15.9 9.4 -0.227 0.177 0.275 0.401 Error (%) 1 Mode 2 Modes 3 Modes -23.4 -22.15E-03 4.3 -0.245 0.0 3 Modes -5.6 0.2 -57.25 × 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 0.260 0.011 0.024 -0.225 0.44E-03 1.202 0.5 -2.245 0.6 9.1 Table 3.466 0.237 0.3 -33.14 m3) from RHA for 0.73E-03 3.295 -0.4 -6.4 0.50E-03 2.010 -0.229 0.022 0.089 0.4 -41.4 -1.1 -14.4 -3.6 0.5 -1.2 Peak values of story drift ratios (as % of story height) from RHA for 0.89E-03 1.181 0.226 0.192 0.062 -0.63E-03 2.24E-03 2.44E-03 3.026 0.333 0.Table 3.325 0.56E-03 2.29E-03 2.4 -53.055 0.001 -0.259 0.88E-03 2.183 0.303 0.03E-03 1.89E-03 1.09E-03 2.023 0.13E-04 9.1 -2.74E-04 9.7 -50.38E-04 2 Modes 2.7 -19.1 -2.76E-03 1.1 -2.260 0.64E-03 3.11E-04 -5.88E-03 2.08E-03 2.

Response due to first mode: (a) roof displacement.15 u (cm) 9.0435 9.5 15 −0.5 0 u (cm) r1 7.12 −0.0435 • 0 (b) V b1 −0.1 0.0435 Vb1 / W 0. 3. (c) joint rotation. (b) base shear.5.12 0 (d) 0 (e) 15 Fig. (d) force-deformation history.5 0 u (cm) r1 7.00174 • 0 (c) r1 −5 0 5 10 15 Time (sec) 20 25 30 0.25 × El Centro ground motion 19 .1 −15 −7.5 9.1 −15 −7.1 5 x 10 −3 θ (rad) 0. Excitation is 0.1 0.12 • 0 (a) r1 −15 0.1 /W 0. and (e) pushover curve.

1 −5 −2.23 0.0252 −2.05 5 x 10 −3 b2 θ (rad) 0. (b) base shear.5 0 u (cm) r2 2. (d) force-deformation history.00244 • 0 (c) r2 −5 0 5 10 15 Time (sec) 20 25 30 0.23 −5 (a) 0.5 0 u (cm) r2 2.0252 (d) 0 0 (e) −0.23 −2.5 5 Fig.05 /W 0.25 × El Centro ground motion 20 .1 −5 −2.0252 • 0 (b) V −0.6. and (e) pushover curve.1 Vb2 / W 0. (c) joint rotation. 3.1 0.5 u (cm) r2 0 • 2. Response due to second mode: (a) roof displacement.5 5 −0. Excitation is 0.

25 × El Centro ground motion 21 .000913 • 0 (c) r3 −2 0 5 10 15 Time (sec) 20 25 30 0.5 0 u (cm) r2 0.422 −1 (a) 0.7.1 u (cm) r3 0 • 0. 3.5 1 Fig.0108 −0. Excitation is 0.422 /W V b3 0 (d) 0 −0.02 0.5 1 −0.02 −1 −0.02 −1 −0.0108 x 10 −3 (b) −0.02 /W V b3 0 • 0. Response due to third mode: (a) roof displacement. (d) force-deformation history. (b) base shear. and (e) pushover curve.0108 −0.02 2 θ (rad) 0. (c) joint rotation.02 −0.422 (e) −0.5 0 u (cm) r3 0.

and the computed response values are identical to the peak response values determined from RHA (Tables 3. and 3. and 3.5e. pushing the structure. For a fixed number of modes included considered.12). implying that the modal truncation errors are small if at 22 .4. In practical application. (3.3).e.1.23 cm .5. 3. (3. and it is apparent in the example considered that three modes provide most of the response (Fig. 3. two. story drifts. (3.2.3 Modal Pushover Analysis Implementing MPA for the fundamental vibration mode.6. Figure 3.6d and 3.20) with n = 1 (Fig.ur relations determined by RHA (Figs. although both analyses led to identical peak values of the individual modal responses. leads to the pushover curves shown in Figs. 3. and 3. 3. This pushover curve is consistent with the relationship between the base shear and roof displacement determined by RHA (Fig. (3.5d).7d). and 3.1 through 3. or three vibration modes.5. which would provide the Dn value to be substituted in Eq. and ur 3o = 0. confirming that MPA gives the exact values of the individual modal responses.12 cm. pushing the structure using the force distribution of Eq. These values of the response quantities are identical to the peak response values determined from RHA (Tables 3. except for the algebraic sign associated with Γ1 and minor round-off errors. this value would be determined directly from the response (or design) spectrum. respectively. i. and external joint rotations computed by pushover analysis are presented in Tables 3. 3. As for the first mode. and the errors in these estimates relative to the exact response from RHA considering all modes.. respectively. the value determined by RHA (Fig. and external joint rotations in Tables 3.422 cm . considering one. 3. As suggested by Eq.6.1. the floor displacements are proportional to the mode shape φ 1 because the structure remains elastic. story drifts.4.21). and 3.3).e.7e and to the floor displacements.3.4) to roof displacement ur1o = 9. The floor displacements. 3. these pushover curves are consistent with the Vb . In RHA the errors arise only from truncating the responses due to higher modes.2. (3. using the force distribution of Eq. 3. Implementing pushover analysis for the second and third modes.10 and Tables 3.6e and 3. Observe that the target roof displacement in each pushover analysis is identical to its exact value determined by RHA.4. 3. i.18). 3. 3.3). respectively. the errors in the MPA results are generally larger than in RHA (Fig.8) leads to the pushover curve shown in Fig. 3.6 present estimates of the combined response according to Eq.9 and Tables 3.5.4. 3.9)..20) with n = 2 and 3 up to roof displacements ur 2 o = 2.

2 0.422 r2 0 • 1.5 Ground 0 0.685 • −3 3 Mode 2 ∆ (cm) u (cm) • 2.eps fig3_9b.8.1 0. Heightwise variation of floor displacements and story drifts from RHA for 0.least three modes are included.3 0.25 × El Centro ground motion: first three modal responses and total (all modes) response (a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th (b) Story Drift Ratios Floor 4th 3rd 2nd 1st Floor 5th RHA All Modes 1 Mode 2 Modes 3 Modes 5th 4th 3rd 2nd 1st RHA All Modes 1 Mode 2 Modes 3 Modes fig3_9a.83 • r3 0 • 0.4 Displacement/Height (%) 0.12 • r1 −15 15 0 r1 0 0 0.4 Story Drift Ratio (%) 0. Additional errors are introduced in pushover analysis due to the approximation inherent in modal combination rules. 3.eps r 5 10 15 20 Time.6 Fig. 3.3 0.5 0.eps Ground 0 0.59 • Total (All Modes) 0 −15 0 ∆ (cm) u (cm) 0 −3 0 fig3_8.2 0.1 0. Response histories of roof displacement and top-story drift from RHA for 0. sec 25 30 Fig.23 r2 −15 15 0 • 0. (a) Roof Displacement 15 (b) Top Story Drift 3 Mode 1 ∆ (cm) u (cm) 9.48 −3 3 1.25 × El Centro ground motion 23 . sec 25 30 r 5 10 15 20 Time.03 −3 3 Mode 3 ∆ (cm) u (cm) r3 −15 15 9.9.

2 1.90E-04 -9.8 -23.63E-03 -2.3 -33.106 0.1 -0.0 -2.261 -0.44E-03 -1.270 0.09E-03 Mode 2 1.6 -17.4 -4.062 -0.9 -8.401 Error (%) 1 Mode 2 Modes 3 Modes -23.25 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Modal Response Mode 1 -2.024 0.328 0.3 Table 3.227 0.Table 3.14 m) from MPA for 0.321 0.53E-04 -9.0 -10.259 0.296 -0.313 0.2 -57.203 0.259 -0.010 0.4 -2.22E-04 2.28E-03 2.04E-03 3.9 -18.92E-04 -1.7 -21.023 -0.012 0.045 -0.09E-03 1.2 -11.31E-03 Error (%) RHA (all modes) 2.009 0.89E-03 2.179 0.177 0.0 -18.3 -13.008 -0.088 -0.89E-03 -1.33E-04 5.42E-04 -1.235 0.6 -15.8 -15.4 -19.31E-03 2.011 0.038 -0.96E-03 2.3 -41.43E-04 -1.106 0.9 -15.097 0.9 -14.9 -15.73E-03 3.264 Error (%) Modal Response 1 Mode 2 Modes 3 Modes -23.5 Peak values of story drift ratios (as % of story height) from MPA for 0.253 -0.74E-03 -1.157 0.24E-03 -2.267 0.1 -19.002 0.9 -12.9 -13.2 -20.4 Peak values of floor displacements (as % of building height = 37.40E-04 5.177 0.90E-03 1.4 -22.0 -46.226 0.231 -0.282 -0.44E-03 3.012 -0.203 0.9 -13.76E-03 -1.043 -0.8 -22.310 0.133 0.032 0.89E-03 2.285 0.38E-03 3 Modes 2.322 0.156 0.011 -0.266 0.00E-03 2.282 0.6 -16.042 0.015 -0.8 -15.3 -2.229 0.042 0.9 -14.73E-05 3.12E-03 1.048 0.03E-03 6.260 0.048 0.253 0.203 0.65E-03 2.15E-03 1.78E-04 2 Modes 2.260 -0.237 0.260 0.370 0.9 -15.89E-03 1.203 0.173 0.15E-03 2.125 0.1 -18.179 0.63E-03 2.4 -11.97E-03 1.197 0.4 -5.097 0.94E-03 2.003 -0.235 -0.31E-03 Joint Rotation (rad) Combined (MPA) 1 Mode 2.336 RHA (all modes) 0.001 0.4 -7.8 -14.4 -4.276 0.47E-03 1.227 0.2 -4.9 -14.062 0.267 0.4 -53.7 -19.055 0.74E-03 1.7 2.9 -16.08E-03 2.267 -0.74E-04 -6.99E-03 Mode 3 -3.006 -0.272 0.466 0.181 0.1 -11.03E-03 -1.9 -13.0 -16.3 -9.89E-03 -2.80E-04 3.078 0.4 -9.060 -0.4 -22.25 × 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 -0.9 -24.4 -9.9 -11.080 0.245 0.00E-03 2.7 24 .336 0.25 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Displacement /Height (%) Modal Response Combined (MPA) Mode 1 Mode 2 Mode 3 1 Mode 2 Modes 3 Modes 0.374 0.3 -12.57E-03 1 Mode -23.72E-03 3.133 0.8 -56.09E-04 -3.245 -0.7 2 Modes -13.4 1.060 0.286 0.069 0.05E-03 3.38E-03 2.2 -12.6 Peak values of joint rotation (radians) from MPA for 0.7 -50.00E-03 3.3 -19.270 0.5 -16.300 0.331 0.2 Table 3.058 -0.124 0.6 -19.133 -0.022 0.03E-03 1.121 0.117 0.270 0.036 -0.157 0.3 -14.069 0.00E-03 2.3 1.071 0.22E-03 -2.026 0.156 0.09E-03 -1.76E-03 1.152 0.385 0.173 0.00E-03 -1.7 -15.090 0.9 3 Modes -12.1 -11.079 0.9 -15.229 0.2 -16.2 -0.9 -13.152 0.00E-03 1.253 0.03E-03 3.230 -0.332 0.274 0.230 0.203 0.125 0.4 0.253 RHA (All Modes) 0.0 -0.237 -0.407 0.4 -14.

1 0. shading indicates modal combination error 25 .4 Story Drift Ratio (%) 0.5 0. Heightwise variation of floor displacements and story drifts from MPA for 0.2 0.25 × El Centro ground motion.(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. 3.4 Displacement/Height (%) 0.6 Fig.3 0.1 0.eps fig3_10b.eps Ground 0 0.10.3 0.5 Ground 0 0.2 0.

26 .

Both systems have the same mass and damping.mι ug t b (4.4 4. Thus.1) becomes mu + cu + f s u. but depend on the history of the displacements: f s = f s u. the initial loading curve is idealized as bilinear. Therefore. sign u b g g bg (4. Eq.3) .1 Inelastic Multistory Buildings RESPONSE HISTORY ANALYSIS For each structural element of a building. (4. and the unloading and reloading curves differ from the initial loading branch. Each structural element of this elastic system is defined to have the same stiffness as the initial stiffness of the structural element of the inelastic system.1) With this generalization for inelastic systems. sign u = . it is useful for later reference to transform Eq.2) The standard approach is to solve directly these coupled equations.2) to the modal coordinates of the corresponding linear system. 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). Expanding the displacements of the inelastic system in terms of the natural vibration modes of the corresponding linear system we get ut = bg n =1 Â φ nqn bt g 27 N (4. (3. Although classical modal analysis (Section 3.1) is not valid for inelastic systems. the relations between lateral forces f s at the N floor levels and the lateral displacements u are not single valued. leading to the “exact” nonlinear response history analysis (RHA).

(4. premultiplying by φ T .… N (4. qr t = 0 for all 28 bg . (3.2 UNCOUPLED MODAL RESPONSE HISTORY ANALYSIS bg Neglecting the coupling of the N equations in modal coordinates [Eq.and classical n damping-orthogonality property of modes gives F qn + 2z nw n qn + sn = .9) for linearly elastic systems. (4. This approximate RHA procedure is the preliminary step in developing a modal pushover analysis procedure for inelastic systems.4) where the only term that differs from Eq. however.” implying that other “modes” will also contribute to the solution.3) will.4) represents N equations in the modal coordinates qn .5) This resisting force depends on all modal coordinates qn t . (4. these equations are coupled for inelastic systems. (4.2). (3.sn ug t bg b g bg bg (4. implying coupling of modal coordinates because of yielding of the structure.4)] leads to the uncoupled modal response history analysis (UMRHA) procedure. sign qn = φ n f s u n . For linear systems. sign u = . The spatial distribution s of the effective earthquake forces is expanded into the modal contributions sn according to Eq. The equations governing the response of the inelastic system to p eff.2).Substituting Eq. (4. 2. Eq. (3.6) The solution of Eq. (4. sign u n b g b g bg (4.6b) are mu + cu + f s u. 4.Gn ug t Mn bg n = 1. give the same results for u t as obtained directly from Eq. (3.3). Equation (4. (3. where φ n are now the modes of the corresponding linear system. and using the mass.4) is rarely solved because it offers no particular advantage over Eq. However.8) because qr t will generally be nonzero for “modes” other than the nth “mode.9) involves T Fsn = Fsn qn . Simultaneously solving these coupled equations and using Eq.6) for inelastic systems will no longer be described by Eq. (4. Unlike Eq.n t given by Eq.3) in Eq. in principle.2). (4.

(a) p 15

eff,1

= −s × 0.25 × El Centro

1

(b) p 5

eff,2

= −s × 0.25 × El Centro

2

ur1 (cm)

0

ur1 (cm)

9.112 •

Mode 1

Mode 1

0

−15 15 Mode 2

−5 5 Mode 2

ur2 (cm)

0

ur2 (cm)

0 • 2.226 5 10 15 20 25 30

−15 0 15

5

10

15

20

25

30

Mode 3

−5 0 5

Mode 3

ur3 (cm)

0

ur3 (cm)

5 10 15 20 Time (sec) 25 30

0

−15 0

−5 0

5

10 15 20 Time (sec)

25

30

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

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

motion

af

af

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

bg

bg

(a) p 80

eff,1

= −s × 1.5 × El Centro

1

(b) p 20

eff,2

= −s × 1.5 × El Centro

2

Mode 1

Mode 1 • 1.437

u (cm)

r1

0 • 48.24

ur1 (cm)

Mode 2

0

−80 80

−20 20 Mode 2

u (cm)

r2

0

3.37 •

ur2 (cm)

0 • 11.62 5 10 15 20 25 30

−80 0 80

5

10

15

20

25

30

Mode 3 0.4931 •

−20 0 20

Mode 3 • 0.7783

u (cm)

r3

0

ur3 (cm)

25 30

0

−80 0

5

10 15 20 Time (sec)

−20 0

5

10 15 20 Time (sec)

25

30

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

af

af

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

bg

bg

by Eq. (3.8),

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

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

b

g

b

g

(4.7)

30

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

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

bg

bg c h

(4.8)

c

h

(4.9)

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

b

g

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

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

bg

bg

bg

bg

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

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

4

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

31

ζn.51 • ur2 (cm) 0 ur2 (cm) n=3 14. Roof displacement due to p eff. n = 1.9 • 0 −50 10 −50 10 n=3 ur3 (cm) 0 • 5. 2.Forces sn Unit mass An(t ) st rn ωn. 4. and 3.112 5 10 15 20 Time (sec) 25 30 −10 0 −10 0 Fig. Fsn / Ln ¨ g(t ) u (a) Static Analysis of Structure (b) Dynamic Analysis of Inelastic SDF System Fig.02 −150 n=2 −150 50 n=2 14.3.n t = . and (b) approximate solution by UMRHA bg af bg 32 . 4.s n u g t . Conceptual explanation of uncoupled modal response history analysis of inelastic MDF systems (a) Nonlinear RHA 150 n=1 150 (b) UMRHA n=1 ur1 (cm) • 75.89 50 r1 0 u (cm) 0 • 78.575 5 10 15 20 Time (sec) 25 30 ur3 (cm) 0 • 5.4. where u g t = 3.0 ¥ El Centro ground motion: (a) exact solution by NL-RHA.

n = 1. and (b) approximate solution by UMRHA af bg bg 4. where u g t = 3.5. this intense excitation was chosen to ensure that the structure is excited well beyond its linear elastic limit. (2) the superposition of responses to p eff. the errors in either response quantity are only a few percent.817 5 10 15 20 Time (sec) 25 30 −10 0 −10 0 Fig.744 • 0 −20 10 n=3 −20 10 n=3 ∆r3 (cm) 0 • 5.956 5 10 15 20 Time (sec) 25 30 ∆r3 (cm) 0 • 5. 4. and 3.4 and 4. both for 3.1 Underlying Assumptions and Accuracy The approximate solution of Eq. Such comparison for roof displacement and top-story drift is presented in Figs.0 ¥ El Centro ground motion: (a) exact solution by NL-RHA. and (3) the Fsn Ln .0 ¥ El Centro ground motion. 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.5. 4.855 ∆r1 (cm) 0 ∆r1 (cm) n=2 • 6.2… N ) according to Eq.(a) Nonlinear RHA 20 n=1 20 (b) UMRHA n=1 • 5.33 0 −20 20 −20 20 n=2 ∆r2 (cm) 0 ∆r2 (cm) 7. (4.6) by UMRHA is compared with the “exact” solution by nonlinear RHA.n t (n = 1. respectively.008 • 6.s n u g t .4) and (4. but even for this very intense excitation. The errors are slightly larger in drift than in displacement.5)] is neglected. (3.15) is strictly valid only for linearly elastic systems.n t = . 2.2.Dn relation is 33 bg bg . (4. Top story drift due to p eff.

The overall errors in the UMRHA procedure are documented in the examples presented in Section 4. (4.10) . 4. 4. (4.2 Properties of the nth-“mode” Inelastic SDF System How is the Fsn Ln .12) for floor displacements.Dn relation? The two sets of forces and displacements are related as follows: V Fsn = bn Γn Dn = urn Γ nf rn 34 (4.Dn relation to be determined in Eq.8) in UMRHA. The structure is pushed using the force distribution of Eq.20). the only force distribution that produces displacements proportional to f n is given by Eq. However. the base shear is Vbny and roof displacement is urny . Therefore. (4. within the linearly elastic range of the structure.urn pushover curve to the Fsn Ln . How to convert this Vbn .6a. it can conduct a force-controlled nonlinear static analysis with an invariant distribution of lateral forces.approximated by a bilinear curve to facilitate solution of Eq. is plotted against roof displacement urn .8) governing Dn t is based on Eq. A bilinear idealization of this pushover curve for the nth-“mode” is shown in Fig.2. (3.4.9) should be determined by nonlinear static analysis of the structure as the structure undergoes displacements u = Dnφ n with increasing Dn .Dn curve. Although most commercially available software cannot implement such displacement-controlled analysis. When implemented by commercially available software. which is different than the Fsn Ln . What is an appropriate invariant distribution of lateral forces to determine Fsn ? For an inelastic system no invariant distribution of forces can produce displacements proportional to f n at all displacements or force levels.8) before it can be solved? Because Eq. (3. (4. Although several approximations are inherent in this UMRHA procedure. (3. when specialized for linearly elastic systems it is identical to the RHA procedure of Section 3. and the base shear Vbn . Therefore we impose this constraint in developing the UMRHA procedure in this section and modal pushover analysis in the next section. the relationship between lateral forces f s and Dn in Eq. such nonlinear static analysis provides the so-called pushover curve.20) to some bg pre-determined roof displacement. At the yield point.9).1. this distribution seems to be a rational choiceeven for inelastic systemsto determine Fsn in Eq. (4.

should be used in Eq. 4. 4.6a is 2 kn = ω n Ln .5).6. Properties of the nth-“mode” inelastic SDF system from the pushover curve Equation 4.6b.6b is w 2 . 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. which may differ from the period of the corresponding linear system. 35 .10 enables conversion of the pushover curve to the desired Fsn Ln − Dn relation shown in Fig.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.8). 4.12) implying that the initial slope of the curve in Fig. (4. Section 13. the initial slope of the pushover curve in Fig. 4.11) * in which M n = Ln Γn is the effective modal mass (Chopra.2.13) This value of Tn . The two are related through Fsny Ln 2 = w n Dny (4. which is not a meaningful quantity.11). (4. Knowing Fsny Ln and Dny from n Eq. where the yield values of Fsn Ln and Dn are Fsny Ln = Vbny * Mn Dny = urny Γ nf rn (4. 2001. In contrast.

4. ωn . details are available in Appendix A: 1. (3. 4. of the nth-“mode” inelastic SDF system (Fig.11). φn .20)].16) to determine the total response.15) and (3. Typically. Dn (t ) . Compute the deformation history. Compute the natural frequencies. is now determined by 36 bg bg .6b) by utilizing Eq. Consider a nonlinear static analysis of the structure subjected to lateral forces distributed over the building height according to s* [Eq. n 3. 4. develop the base-shear – roof-displacement ( Vbn − urn ) pushover curve for the force distribution s* [Eq. 9. the peak value of Dn t .12) and (3. which is summarized next as a sequence of steps. r o .21) where Dn .13). 4. 6. An t . and pseudo-acceleration history.3b) with force-deformation relation of Fig.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 . 7. (4.6a). For the nth-mode.4. and modes. for linearly-elastic vibration of the building. (3. n (3. (3.3 Summary The inelastic response of an N-story building with plan symmetric about two orthogonal axes to earthquake ground motion along an axis of symmetry can be estimated as a function of time by the UMRHA procedure just developed. 5. Repeat Steps 2 to 6 for as many modes as required for sufficient accuracy. This value of the roof displacement is given by Eq. 4. Convert the idealized pushover curve to the Fsn Ln − Dn relation (Fig. Calculate the peak value.6b. Calculate histories of various responses by Eqs. 2. (3. 8.20)]. Idealize the pushover curve as a bilinear curve (Fig. with the structure is pushed to the roof displacement urno .n t . 4. of the total response r t obtained in Step 8.2. Combine the “modal” responses using Eqs. the first two or three modes will suffice.

As shown in Sections 3. as mentioned earlier. It provides results for elastic buildings that are identical to the well-known RSA procedure (Section 3. (4. the lateral force distribution used possesses two properties: (1) it appears to be the most rational choice among all invariant distribution of forces. to obtain an estimate of the peak value ro of the total response. The response value rno is an estimate of the peak value of the response of the inelastic system to p eff. plastic hinge rotations. At this roof displacement. we will refer to rno as the peak “modal” response even in the case of inelastic systems. 4. 5.n t .3b) with force-deformation relation of Fig. Eq. or from the inelastic response (or design) spectrum. 4.6). are combined using an appropriate modal combination rule. the pushover analysis provides an estimate of the peak value rno of any response rn t : floor displacements.8). 37 .2. The peak “modal” responses rno . although somewhat intuitive for inelastic buildings.3) because.2 and 3. for elastic systems. each determined by one pushover analysis.solving Eq. etc. governed by Eq.2.1 Summary The peak inelastic response of a building to earthquake excitation can be estimated by the MPA procedure just developed.3. e. Calculate the peak roof displacement urno associated with the nth-“mode” inelastic SDF system from Eq. (4. it can be determined from the inelastic response (or design) spectrum (Chopra. it seems reasonable because it provides results for elastic buildings that are identical to the well-known RSA procedure described in Section 3. as described in Section 4. Sections 7. (3. alternatively.21). Steps 1 to 4 of the MPA are same as those for UMRHA. (3.8). Dn . However.18). story drifts. This application of modal combination rules to inelastic systems obviously lacks a theoretical basis. Compute the peak deformation. This pushover analysis.12). which is summarized next as a sequence of steps. seems reasonable. of the nth-“mode” inelastic SDF system (Fig.g. rno also represents the exact peak value of the nth-mode contribution rn t to response r t . (4. 2001.3.6 and 7. 4. and (2) it provides the exact modal response for elastic systems.. bg bg bg bg Thus. details are available in Appendix B.6b by solving Eq. joint rotations.4. 6.

To ensure that this structure responds well into the inelastic range the El Centro ground motion is scaled up by a factor varying from 1.53 cm.7. the UMRHA procedure lacks a rigorous theory. also included are the combined responses due to one.1).1 and 4.7 cm. 4. is presented next. 4.4.0. the combined response due to three “modes”.8. Determine the total response by combining the peak “modal” responses using the SRSS combination rule of Eq. determined by the UMRHA procedure. 4. 4. the peak roof displacement due to each of the three “modes” is ur1o = 48. the “exact” results. and compared with the results of a rigorous nonlinear RHA using the DRAIN-2DX computer program. the 38 . in contrast to modal analysis (Section 3. ur 2o = 11. subtract the yield value of hinge rotation to determine the hinge plastic rotation. Typically.0 to 3. This is to be expected. with.1 Uncoupled Modal Response History Analysis The structural response to 1.7 shows the individual “modal” responses.4.3 cm. and three modes are compared with the “exact” values in Fig. and the errors in the approximate results are shown in Fig. The peak values of response are as noted. and the “exact” response from nonlinear RHA for the roof displacement and top-story drift. From the total rotation of a plastic hinge. extract from the pushover database values of other desired responses.9. and ur 3o = 2.2. Figure 4. Observe that errors tend to decrease as response contributions of more “modes” are included.2). and the percentage errors in the approximate results. (3.18). 9. respectively.5 ¥ the El Centro ground motion including the response contributions associated with three “modal” inelastic SDF systems.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. two. the first two or three “modes” will suffice. This deficiency also implies that. The peak values of floor displacements and story drifts including one. although the trends are not systematic as when the system remained elastic (Section 3. two. 8. and three “modes”. in particular. say. rno . At urno . Repeat Steps 3 to 8 for as many “modes” as required for sufficient accuracy. three “modes” included. The peak values of displacements of all floors and drifts in all stories are presented in Tables 4.

25 (Tables 3.3 11.5 × El Centro ground motion: individual “modal” responses and combined response from UMRHA.1 and 4. in particular. the errors in story drifts are larger compared to floor displacements.53 ∆r3 (cm) 12 0 • 2. Response histories of roof displacement and top-story drift due to 1. For each excitation.24 • NL−RHA r −12 12 ∆r (cm) u (cm) 0 5 10 15 20 Time. just as for elastic systems.5. However.62 5.2).2). we investigate how the errors in the UMRHA vary with the deformation demands imposed by the ground motion.0. the errors in responses computed by UMRHA including three “modes” relative to the “exact” response were determined. For this purpose the UMRHA and exact analyses were repeated for ground motions of varying intensity.5 (Tables 4.44 • "Mode" 2 r1 0 • 48.response is much less accurate (Tables 4. and 3.1 and 3.75.2) if the system yields significantly versus if the system remains within the elastic range (Tables 3. for a fixed number of “modes” included.0. defined as the El Centro ground motion multiplied by 0. 1.38 6. recall that the computed errors have been presented earlier for ground motion multipliers 0.1 −12 12 ∆r (cm) u (cm) UMRHA 3 "Modes" 0 • 7.1 and 3. 4.2) and 1.25. and total response from NL-RHA 39 . sec 25 30 r −12 0 Fig.5. the degree to which the system deforms beyond its elastic limit. 0. (a) Roof Displacement (b) Top Story Drift 80 12 0 • 3. 0.7. 0. 1. sec 25 30 • 48.85.0.7 • −80 80 ∆r1 (cm) u (cm) "Mode" 1 −12 r2 0 ∆r2 (cm) 12 0 u (cm) −80 80 −12 r3 0 • 2. 2.6 5 10 15 20 Time.1 and 4. Next.88 u (cm) "Mode" 3 −80 80 0 −80 80 0 −80 0 • 44.

For this purpose. Two versions of the pushover curve are 40 . two.20)] for the first three n modes of the system are shown in Fig. Shown is the error in each floor displacement (Fig. 4. the pushover curves using force distributions s* [Eq. Heightwise variation of error in floor displacements and story drifts estimated by UMRHA including one. or three “modes” for 1.5 1 1. in each story drift (Fig. 4. (3.10 summarizes the error in UMRHA as a function of ground motion intensity. To interpret these results.10b).11.10a). and the error envelope for each case. 4. indicated by a ground motion multiplier. 4.5 Fig.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. 4.9.5 Story Drift Ratio (%) 2 2. Heightwise variation of floor displacements and story drift ratios from UMRHA and NL-RHA for 1.(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 0.8.5 Displacement/Height (%) 2 Ground 0 0. it will be useful to know the deformation of the system relative to its yield deformation.5 1 1. with the peak displacement of each “modal” SDF system noted for each ground motion multiplier.

even though the system remains essentially elastic.0. as mentioned in Section 4. 4.6 and 4.11). Second. (3) maintain roughly similar values for more intense ground motions.75. 4.11a).4).10) for each “modal” inelastic SDF system (Figs 4. (3. The structure is pushed using the force distribution of Eq.2. 11. 41 .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. drifts in all stories (Table 4. and 3 (Fig.included: the actual curve and its idealized bilinear version.01 (Fig.2. UMRHA lacks a rigorous theory and is based on several approximations. the pushover curve for each “mode” is idealized by a bilinear curve in solving Eq.75. 3. Each of these three pushover analyses provides the pushover curve (Fig. and.5 ¥ the El Centro ground motion. were noted but are not shown here. the values determined by RHA of the nth-mode inelastic SDF system (Fig.1). the peak values of displacements at all floors (Table 4. For more intense excitations. The idealized curve for the first “mode” deviates most from the actual curve near the peak displacement corresponding to ground motion multiplier 1.20) with n = 1.4) to roof displacements urno = 48. 4. 2.0. The system remains elastic up to ground motion multiplier 0.53 cm. determined from “exact” analyses.3). and plastic hinge rotations at the external beam end at each floor level (Table 4. and (4) are larger in story drift compared to floor displacements. and 2. First. the ductility factor for the first mode system is only 1. as mentioned in Section 3. the errors in truncating the higher mode contributions are negligible.4.5). (4.4.10 permits the following observations regarding the accuracy of the UMRHA procedure: the errors (1) are small (less than 5%) for ground motion multipliers up to 0. The location of plastic hinges and their rotations. Additional errors are introduced in UMRHA of systems responding beyond the linearly elastic limit for at least two reasons. 4.7 cm. (2) increase rapidly as the ground motion multiplier increases to 1. Perhaps this explains why the errors are large at this excitation intensity. the first reason mentioned above seems to be the primary source for the errors.3 cm. Figure 4. respectively.7).

2 12.6 1.10. and (b) story drifts Table 4.104 0.220 0.5 10.096 0.707 1.226 -0.241 -1.4 1.945 -37.0 1.366 -0.317 0.914 2.333 0.126 0.3 25.376 -1.079 0.5 9.763 -15.371 0.1 3.5 18.5 9. Errors in UMRHA as a function of ground motion intensity: (a) floor displacements.772 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 1.379 1.0 2.256 1.6 2.811 1.982 9.376 1.668 -23.9 31.200 8.478 0.350 -0.5 -3.1 Peak values of floor displacements (as % of building height = 37.472 1.138 1.863 0.072 -1.3 0.2 1.5 2 GM Multiplier 2.135 9.856 2.133 1.5 1 1.3 1.248 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.938 -1.806 -0.5 28.495 1.9 5.484 0.9 12.3 8.065 0.7 14.3 42 .1 1.260 -15.0 11.914 -0.136 1.942 1.241 1.8 1.0 9.1 0.942 -0.072 1.722 0. 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 0.033 0.121 -0.366 0.554 1.6 4.4 4.616 -0.4 -1.018 0.256 -1.430 1.214 0.983 1.540 0.7 Table 4.938 1.169 0.202 11.003 -31.8 1.370 -0.071 -0.7 31.863 1.049 -0.490 -1.526 -0.751 1.2 4.490 1.5 3 Fig.315 -0.120 1.1 8.14 m) from UMRHA for 1.8 1.9 12.338 1.214 -0.154 0.291 0.372 -1.338 -1.852 1.201 -0.820 -0.5 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Displacement /Height (%) “Modal” Response Combined (UMRHA) “Mode” 1 “Mode” 2 “Mode” 3 1 “Mode” 2 “Modes” -0.3 6.8 1.283 1.055 -0. 4.130 0.070 1.256 1.8 0.2 1.068 0.216 1.9 16.8 1.009 -0.220 -0.877 0.4 -7.010 0.371 -0.2 22.044 1.1 1.407 -10.819 2.473 -22.575 -41.0 7.8 14.410 -1.298 0.698 1.844 -25.298 -0.0 -9.5 1 1.410 1.513 -0.057 -0.0 11.727 1.(a) Floor Displacements 60 50 40 Error Envelope Error for Floor No.071 0.235 -0.2 Peak values of story drift ratios (as % of story height) from UMRHA for 1. Noted 60 50 40 (b) Story Drifts Error Envelope Error for Story No.5 28.971 1.5 × El Centro ground motion Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Modal Response “Mode” 1 “Mode” 2 Drift Ratio (%) Combined (UMRHA) 1 “Mode” 2 “Modes” “Mode” 3 -1.201 1.663 0.373 -0.820 -19.372 1.003 0.187 -0.293 1.676 0.9 31.2 6.426 -1.088 10.5 2 GM Multiplier 2.513 0.806 0.201 -1.663 -0.5 3 0 0 0.900 -10.

4. As shown in Figs.3 through 4. The errors are especially large in the hinge plastic rotations estimated by the MPA procedure. Fortuitously. but the third “mode” contributions do not seem especially important (Fig. and three “modes. although the error is recorded as 100% if MPA estimates zero rotation when the nonlinear RHA computes a non-zero value. 4. even if three “modes” are included (Fig.12 and 4. the hinges identified in beams at the 6th floor were at variance with the “exact” results. were determined by four analyses: MPA considering one “mode. 4. For example. (3.14.3 and 4. 4.3 and 4.18). Furthermore. the third “mode” does not contribute because this SDF system remains elastic (Fig.5). MPA failed to identify the plastic hinges at the column bases in Fig. 4. this error is not especially significant because the hinge plastic rotation is very small. considering one.4). 4. two. The locations of plastic hinges shown in Fig. However.Figures 4.” and three “modes. 43 . significantly smaller than in UMRHA (compare Fig.3 and 4.11c).4).4 with Tables 4. 4. and are within a few percent of the exact values for the upper stories. Pushover analysis seems to be inherently limited in computing accurately hinge plastic rotations. and the errors in these estimates relative to the exact response from nonlinear RHA.13 with Fig.12 and Tables 4. The drifts are underestimated by up to 13% in the lower stories. The first “mode” alone is inadequate.12 and Tables 4. MPA including three “modes” underestimates the displacements of the lower floors by up to 8% and overestimates the upper floor displacements by up to 14%. in general. The second “mode” was necessary to identify hinges in the upper stories. for two or three modes included. Obviously.13 and Table 4.9 and Tables 4. however. but was more successful when the excitation was more intense (results are not included). 4.14. the results were not always accurate.5 present estimates of the combined response according to Eq.” respectively.13 and Tables 4. the errors in the modal pushover results are. 4. One “mode” pushover analysis was unable to identify the plastic hinges in the upper stories where higher mode contributions to response are known to be more significant. 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. overestimated by up to 18% in the middle stories.13 and Tables 4. significant improvement is achieved by including response contributions due to the second “mode”.2).” and nonlinear RHA.1 and 4.” two “modes. especially in estimating the story drifts (Fig.3 and 4. 4. Observe that the primary contributor to plastic hinge rotations in the lower stories is the first “mode” and the second “mode” in the upper stories.

V = 5210 kN.5.85 0.25 0.5 1 0.5 2 3 Actual Idealized 2000 0 0 0.75.0 × El Centro ground motion 44 . and 3.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36.6 cm.2 cm.75 Actual Idealized 20 40 60 Roof Displacement (cm) (b) "Mode" 2 Pushover Curve 80 12000 10000 u = 9. “Modal” pushover curves with peak roof displacements identified for 0.75 0.13 y by Base Shear (kN) 8000 6000 1 1.85 Base Shear (kN) 8000 6000 4000 2000 0 0 0.11.5 0. 1. 0.5 0. 0.19 y by 3 2 1. 2. V = 4952 kN. α = 0. 1.5.25.0. α = 0.25 1 2 4 6 Roof Displacement (cm) 8 10 Fig.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 = 4.75 0.9 cm.0. V = 7616 kN. α = 0.5 0.85 0.5 2 3 4000 2000 0 0 Actual Idealized 0. 4.

5 × El Centro ground motion. the MPA procedure is equivalent to the RSA procedure. UMRHA is essentially exact. shading indicates errors in MPA including three “modes” Figure 4.75. now standard in engineering practice. However. 4.4 also apply to MPA. an observation with practical implications. 4. implying that the modal combination errors contained in these procedures are acceptable. The errors are only weakly dependent on ground motion intensity (Fig. implying excitations weak enough to limit the response in the elastic range of the structure. the errors in MPA were fortuitously smaller than in UMRHA (compare Figs.5 1 1. 4. While various sources of errors in UMRHA. indicated by a ground motion multiplier.15 summarizes the error in MPA considering three “modes” as a function of ground motion intensity.5 Fig.0.3 for elastic systems (or weak ground motions).(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. The fact that MPA is able to estimate the response of buildings responding well into the inelastic range to a similar degree of accuracy indicates that this procedure is accurate enough for practical application in building retrofit and design.15). 4. implying excitations intense enough to cause significant yielding of the structure. the errors in MPA were larger for ground motion multipliers less than 0. and the error envelope for each case.10 and 4.4.5 1 1.2 and 3.5 Story Drift Ratio (%) 2 2. In this case. identified in Section 3. As mentioned in Section 3.3.15) for ground multipliers larger than 1.4. whereas MPA contains errors inherent in modal combination rules.15a).15b). Shown is the error in each floor displacement (Fig. as discussed in Sections 3. each story drift (Fig. 4. 45 .12 Heightwise variation of floor displacements and story drift ratios from MPA and NL-RHA for 1.5 Displacement/Height (%) 2 Ground 0 0.

4. story drifts. Errors in floor displacements.5 ¥ El Centro ground motion 46 . and three “modes” for 1. and hinge plastic rotations estimated by MPA including one.(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.13. two.

3 1.1 62.066 -0.980 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 0.763 -14.5 2.37E-03 1.900 -0.5 10.72E-03 7.116 1.72E-03 7.9 0.015 0.705 1.53E-03 7.1 1.76E-03 4.130 0.640 -1.197 -0.50E-10 3.266 -0.5 2.8 -4.176 0.209 1.118 0.00E+00 0.233 1.72E-03 7.705 -1.220 1.135 -7.9 -4.015 0.88E-03 1.305 -0.351 -0.5 1.2 11.498 1.9 -100.6 -44.311 0.7 1.503 -1.02E-03 3.26E-04 9.018 0.066 -0.233 1.071 -0.5 Table 4.304 1.012 1.071 0.00E+00 0.8 -29.00E+00 0.0 -100.049 -0.435 0.36E-03 6.00E+00 0.00E+00 0.2 -4.4 Peak values of story drift ratios (as % of story height) from MPA for 1.00E+00 0.2 -100.37E-03 1.007 1.5 10.478 0.00E+00 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th “Mode” 2 0.00E+00 3 “Modes” 7.407 -27.7 1.00E+00 0.009 0.190 -0.Table 4.738 1.00E+00 0.222 0.1 62.429 0.518 1.6 -9.003 -16.8 17.125 -1.8 -29.668 -13.9 -100.414 1.5 2.527 -0.00E+00 0.55E-03 3.053 -1.76E-03 4.60E-03 2.88E-03 0.72E-03 7.00E+00 “Mode” 3 0.1 -8.10E-02 9.910 1.088 12.6 1.728 1.315 0.895 1.154 0.14 m) from MPA for 1.820 -7.37E-03 1.338 1.200 8.00E+00 0.694 1.00E+00 1 “Mode” 7.756 0.068 0.057 0.982 13.260 -14.4 -8.594 -1.752 1.053 1.250 0.781 0.055 0.687 0.1 13.667 1.0 15.88E-03 0.5 7.9 7.2 -3.76E-03 4.37E-03 1.00E+00 0.304 -1.8 “Mode” 3 -1.1 62.00E+00 0.02E-03 3.6 7.756 0.9 5.0 -5.614 0.945 -49.844 -7.980 -0.581 0.5 × El Centro ground motion Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th “Modal” Response “Mode” 1 “Mode” 2 Drift Ratio (%) Combined (MPA) 1 “Mode” 2 “Modes” 3 “Modes” Error (%) NL 1 2 3 RHA “Mode” “Modes” “Modes” 1.00E+00 0.4 1.156 -0.220 1.804 1.00E+00 0.2 0.033 -0.399 0.371 -0.911 0.5 Peak values of hinge plastic rotations (radians) from MPA for 1.259 1.2 0.399 0.516 0.00E+00 0.8 1.0 -100.8 1.298 0.737 1.0 -100.36E-03 6.2 1.55E-03 3.0 1.8 -6.8 -12.36E-03 6.652 1.683 1.99E-03 6.8 -6.6 -44.76E-03 4.895 1.733 1.640 1.298 -0.879 1.0 -50.244 0.02E-03 0.202 8.00E+00 0.737 0.331 1.22E-10 NL RHA 1.0 2 “Modes” -32.2 6.942 6.3 -3.3 Peak values of floor displacements (as % of building height = 37.473 -15.1 18.101 -0.5 -6.37E-04 Error (%) “Modal” Response 1 “Mode” -32.00E+00 0.55E-03 3.6 -7.018 -0.116 1.7 -12.60E-04 7.0 47 .575 -53.9 1.8 -29.00E+00 0.1 46.36E-03 6.0 3 “Modes” -32.0 -50.018 -0.503 1.3 13.02E-03 0.667 -1.641 1.581 0.05E-03 2.666 Table 4.007 1.5 × El Centro ground motion Hinge Plastic Rotation (rad) “Modal” Response Combined (MPA) “Mode” 1 7.9 1.2 -100.116 1.105 0.6 13.514 -1.426 15.00E+00 2 “Modes” 7.19E-10 3.00E+00 0.3 -3.6 13.1 46.168 -0.372 0.18E-03 7.781 0.6 -8.00E+00 3.517 1.611 0.9 2.2 -100.1 46.8 7.3 11.336 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.222 0.745 1.8 0.

5). the results presented so far did not include gravity load effects. (3.4. and three “modes.20) with n = 1 (Fig. Starting with its initial state under gravity loads. 4.4.” and the errors in these estimates relative to the exact response from nonlinear RHA. (3.18) considering one.4) to roof displacement ur1o = 52.18 and Tables 4.17 and Tables 4. As shown in Fig.8 present estimates of the combined response according to Eq.20) with n = 1 to a target roof displacement. (3. Results of such analyses for the selected building subject to 1. 4.” but not the contributions of higher “modes.7). as are the roof displacements ur 2o = 11. 3.17 and 4. Figures 4. 4. Static analysis of the structure for gravity loads provides the initial state—forces and deformations—of the structure.” these effects will modify the combined modal response as well as the results of nonlinear RHA.7).16a—which is slightly different than the one excluding gravity loads (Fig. The pushover curves for the second and third modes included in Fig.3 .4. the third “mode” contributions do not seem especially important (Fig. 4.18 and Tables 4.6 and 4. and overestimated by up to 29% in the middle stories and by up to 13% in the upper stories.16 are unchanged.17 and Tables 4. Significant improvement is achieved by including response contributions due to the second “mode. resulting in the pushover curve shown in Fig.5 × El Centro ground motion are presented below. MPA including two “modes” underestimates the displacements of lower floors by up to 6% and overestimates the upper floor displacements by up to 22%.6 and 4. 4. 4. The response contributions of the second and third modes are the same as before (Tables 4.7).8). The errors are especially large in the plastic hinge rotations estimated by the MPA procedure even if three “modes” are included (Fig. and the structure is pushed using the force distribution of Eq. They are now included in the pushover analysis of the structure for the first “mode” only.7.6 and 4. but it is inadequate in estimating the story drifts (Fig.0 cm. Obviously gravity load effects will influence the seismic demands due to the first “mode. the structure is pushed using the force distribution of Eq. Also presented at these roof displacements are the displacements at all floors (Table 4.11a). 4. drifts in all stories (Table 4.7 cm and ur 3o = 2. and plastic hinge rotations at selected external beam end at each floor level (Table 4.18 and 48 .6 through 4.” however.6).3 Modal Pushover Analysis with Gravity Loads To evaluate the accuracy of the dynamic response of the system. The story drifts are underestimated by up to 7% in the lower stories. The first “mode” alone provides adequate estimates of floor displacements. two.53 cm.

Most pushover analysis procedures do not seem to compute to acceptable accuracy plastic hinge rotations.8). Locations of plastic hinges determined by MPA considering one.5 3 0. 2−"Modes" •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• • •• (c) 3−"Modes" • • • • • • • •• •• • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• • •• • • • • • • • • • (d) Nonlinear RHA •• •• • •• •• •• •• • •• •• •• •• •• •• •• • •• •• • •• •• •• •• • • • • • • •• •• •• Fig. Noted Error (%) 30 20 10 0 0 2 8 3 4 1 9 Error (%) 30 20 5 4 6 5 7 2 9 10 0 0 1 8 7 3 6 0.5 3 Fig.5 × El Centro ground motion (a) Floor Displacements 60 50 40 Error Envelope Error for Floor No.5 1 1. and (b) story drifts 49 .5 1 1.5 2 GM Multiplier 2. 1−"Mode" • • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • •• • •• • • • • • •• •• • •• •• •• •• (b) MPA. 4.15. 4. and three “modes” and by NL-RHA for 1. two. Errors in MPA as a function of ground motion intensity: (a) floor displacements. Noted 60 50 40 (b) Story Drifts Error Envelope Error for Story No.14. (a) MPA.Table 4.5 2 GM Multiplier 2.

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. Shown is the error in each floor displacement (Fig. each story drift (Fig. MPA provides response values accurate enough for practical application in building retrofit or design.20a). 4. excluding gravity load effects. 4. this procedure is able to predict plastic hinge locations essentially consistent with nonlinear RHA. With two modes included in MPA.” two “modes”.20 summarizes the error in MPA considering three “modes” as a function of ground motion intensity.20b). 4.15. 4.The locations of plastic hinges shown in Fig.19 were determined by four analyses: MPA considering one “mode. and the error envelope for each case. 50 . and the errors are only weakly dependent on ground motion intensity. Figure 4. indicated by a ground motion multiplier. The second “mode” is necessary to identify hinges in the upper stories. and nonlinear RHA. These errors are only slightly larger than those in Fig.

75 0.50. 1. 2. “Modal” pushover curves with gravity loads included. 4.16.13 y by Base Shear (kN) 8000 6000 1 1.25 1 2 4 6 Roof Displacement (cm) 8 10 Fig.0 × El Centro ground motion 51 .25 0.5 1 0. and 3.75 0.5 0. 0.6 cm.3 cm.75.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36. α = 0.85 0.5 2 3 4000 2000 0 0 Actual Idealized 0.5 0.25 5 10 15 20 Roof Displacement (cm) 25 (c) "Mode" 3 Pushover Curve 12000 10000 u = 4.0.85.19 y by 3 2 1.9 cm.85 0. noted are peak values of roof displacement for 0. V = 5210 kN.14 y by Base Shear (kN) 8000 6000 4000 1. 0.0.5 2 3 Actual Idealized 2000 0 0 0. α = 0. 0. α = 0.85 Base Shear (kN) 8000 6000 4000 2000 0 0 0.75 Actual Idealized 20 40 60 Roof Displacement (cm) 80 (b) "Mode" 2 Pushover Curve 12000 10000 u = 9.25. V = 7433 kN. V = 4952 kN.5 0.

Errors in floor displacements.5 Displacement/Height (%) 2 Ground 0 0. story drifts.(a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th NL−RHA MPA 1 "Mode" 2 "Modes" 3 "Modes" (b) Story Drifts Floor Floor 5th 4th 3rd 2nd 1st 5th 4th 3rd 2nd 1st NL−RHA MPA 1 "Mode" 2 "Modes" 3 "Modes" Ground 0 0. 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.18.5 Fig. Heightwise variation of floor displacements and story drift ratios from MPA and NL-RHA for 1. and hinge plastic rotations estimated by MPA including one.5 Story Drift Ratio (%) 2 2.5 1 1. 4. two. and 52 .5 × El Centro ground motion.5 1 1.17. 4. gravity loads included.

19.(a) MPA.20.5 3 0.5 2 GM Multiplier 2. gravity loads included 53 .5 1 1.5 3 Fig. gravity loads included (a) Floor Displacements 60 50 40 Error Envelope Error for Floor No.5 × El Centro ground motion. 1−"Mode" • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • •• •• •• • • • • • •• •• •• •• •• •• •• (b) MPA. 3−"Modes" (d) Nonlinear RHA • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• •• •• • • • • • • • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • • • • • •• •• •• Fig. 4.5 2 GM Multiplier 2. 2−"Modes" •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• •• •• (c) MPA. Noted 60 50 40 (a) Story Drifts Error Envelope Error for Story No.5 1 1. two. 4. and three “modes” and by NL-RHA for 1. and (b) story drifts. Locations of plastic hinges determined by MPA considering one. Noted Error (%) 30 20 6 8 1 5 2 4 3 9 7 Error (%) 30 4 5 20 6 9 7 1 10 0 0 10 0 0 8 2 3 0. Errors in MPA as a function of ground motion intensity: (a) floor displacements.

3 -22.888 1.836 -0.860 1.9 -4.00E+00 3 “Modes” 8.4 -6.831 0.2 1.728 1.102 1.130 0.213 1.594 -1.821 -1.00E+00 0.8 0.4 1.8 -32.00E+00 0.1 4.2 21.783 1.00E+00 0.8 Table 4.996 0.603 1.015 0.310 1.754 1.19E-04 5.5 × El Centro ground motion.213 1.3 -22.17E-03 9.372 0.514 -1.5 0.953 15.414 28.9 31.429 -1.8 9.23E-03 3.311 0.6 2.Table 4.5 10.0 -100.850 1.14 m) from MPA for 1.19E-03 1.2 4.4 26.353 -23.78E-03 1.1 1.921 1.3 -100.371 -0. gravity loads included Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th “Modal” Response “Mode” 1 “Mode” 2 Drift Ratio (%) Combined (MPA) 1 “Mode” 2 “Modes” 3 “Modes” Error (%) NL 1 2 3 RHA “Mode” “Modes” “Modes” 1.102 1.3 -100.168 -0.7 -2.270 -12.2 2.037 0.237 0.066 -0.7 4.23E-03 3.26E-03 3.068 0.7 19.04E-02 8.00E+00 0.101 -0.00E+00 0.55E-03 3.351 -0.7 -2.237 0.6 0.530 1.850 -1.5 1.927 1.00E+00 0.466 0.00E+00 2 “Mode” 8.00E+00 0.854 0.330 1.156 -0.35E-03 8.00E+00 0.5 54 .908 -1.0 37.55E-03 3.2 -3.6 1.2 0.190 -0.1 21.266 -0.00E+00 0.23E-03 0.11E-03 9.00E+00 3.00E-10 NL RHA 1.821 1.00E+00 0.35E-03 8.071 -0.098 20.998 0.207 18.88E-03 0.436 1.665 0.176 0.3 -22.744 1.813 1.830 -12.319 1.530 1.687 0.88E-03 0.8 0.009 0.2 16.8 Peak values of hinge plastic rotations (radians) from MPA for 1.478 0.055 0.2 9.00E-03 5.2 1.754 0.637 0.8 1.686 -7.3 13.7 1.19E-03 1.128 -1.305 -0.263 1.2 9.319 1.399 0.00E+00 0.114 -1.5 21.983 1.3 -100.064 -10.434 0.00E+00 Combined (MPA) 1 “Mode” 8.35E-10 3.908 1.998 1.23E-02 1.0 -30.88E-03 1.673 Table 4.0 37.0 -30.9 -3.2 9.8 -32.4 20.330 1.00E+00 0.603 -1.515 -50.429 1.5 -5.1 13.527 -0.0 -100.6 19.2 12.00E+00 0.913 7.3 1.983 1.854 0.199 16.9 -6.7 Peak values of story drift ratios (as % of story height) from MPA for 1.11E-03 9.5 21.507 1.3 9.667 0. gravity loads included Hinge Plastic Rotation (rad) Floor “Modal” Response “Mode” 1 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 8.530 1.114 1.00E-03 5.6 2.516 0.637 0.13E-03 5.125 -1.2 -0.831 0.8 28.214 0.5 3 “Modes” -32.00E+00 Error (%) “Modal” Response 1 “Mode” -32.033 -0.490 -11. 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” 0.6 Peak values of floor displacements (as % of building height = 37.263 0.858 2.877 -46.55E-03 3.00E+00 0.00E+00 “Mode” 3 0.933 1.4 -4.315 0.2 -2.213 1.434 0.454 1.11E-03 9.04E-10 3.996 -0.00E+00 0.0 37.0 16.19E-03 1.75E-03 0.5 “Mode” 3 -1.0 2 “Modes” -32.11E-03 9.109 0.461 0.105 0.4 1.00E+00 “Mode” 2 0.7 16.2 4.066 -0.938 1.19E-03 1.257 0.5 × El Centro ground motion.197 -0.071 0.35E-03 8.00E+00 0.00E+00 0.2 4.399 -0.00E+00 0.154 0.00E+00 0.434 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.057 0.23E-03 0.00E-03 5.4 -4.5 × El Centro ground motion.822 1.00E-03 5.107 1.5 1.049 -0.018 -0.998 21.037 -0.35E-03 8.

Comparison of Modal and FEMA Pushover Analyses 5.. MPA considering three “modes. and 3. a plot of base shear versus roof displacement. SRSS distribution: s* is defined by the lateral forces back-calculated from the story shears determined by response spectrum analysis of the structure. “Uniform” distribution: s* = m j (where the floor number j = 1. assumed to be linearly elastic. Equivalent lateral force (ELF) distribution: s* = m j h k where h j is the height of the j j jth floor above the base. 5. plastic hinge rotations. k = 2 for T1 ≥ 2. j 2. 1997] The pushover curve. and varies linearly in between. The floor displacements. computed at the target displacement represent the earthquakeinduced demands on the structure. and the exponent k = 1 for fundament period T1 £ 0. etc.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. 2… N ). is determined by nonlinear static analysis of the structure subjected to lateral forces with invariant distribution over height but gradually increasing values until a target value of roof displacement is reached. gravity load effects were included in all 55 .5 sec . story drifts. Specified in FEMA-273 are three distributions for lateral forces: 1.1 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.5.” and nonlinear RHA. joint rotations.5 sec .

Figures 5.177 0. 5.00719 0.11 0.0 cm.0913 0. and the displacements of upper floors are overestimated by 9-22%. 5. the floor displacement demands in Fig.0654 0. and the locations of all plastic hinges in Fig.0981 0. The pushover curves are given in Fig.0466 0.3b and Table 5.0197 0. with the errors being larger for the SRSS distribution.3a and Table 5. plastic hinge rotation demands in Table 5.042 0.2.3.0446 0.11 0.1.367 (a) Uniform (b) ELF (c) SRSS Fig.2.1. and Table 5.4.1.3a and 5. 5.0381 0. the value determined from RHA of the first-mode inelastic SDF system for 1. 5.21 0.0896 0.11 0. The first four floor displacements are within 5% of the “exact” value. 5. the story drift demands in Fig. 5. The “uniform” distribution overestimates all floor displacements by 17-28%.4 and Tables 5. The errors in the FEMA and MPA estimates of seismic demands relative to the “exact” demands are presented in Fig.11 0.3.0702 0.11 0.5 times the El Centro ground motion.3. Using each of these force distributions. wherein the first two are obvious and the third is determined from response spectrum analysis of the building (Appendix C). pushover analyses are implemented for a target roof displacement of 52.126 0.1 through 5. 0.1 demonstrate that the displacement demands are underestimated by the ELF and SRSS force distributions by 12 to 30 % at the lower six floors of the building.11 0. The three FEMA force distributions are presented in Fig.119 0.281 0. (b) ELF. Also included in these presentations are the MPA results considering three “modes” and the “exact” demands from nonlinear RHA. Force distributions in FEMA-273: (a) “uniform”.112 0.11 0.4a.4.062 0. and (c) SRSS 56 .165 0. 5. both presented in Section 4. The MPA procedure is more accurate than all the FEMA force distributions.analyses.

Pushover curves using three force distributions in FEMA-273: (a) “uniform”.2. α = 0. V = 8530 kN. 5. V = 6897 kN.3 cm.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.(a) Uniform Distribution 12000 10000 u = 33 cm. gravity loads are included 57 . α = 0.24 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized 10 20 30 40 Roof Displacement (cm) 50 60 Fig. (b) ELF.14 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized 10 20 30 40 Roof Displacement (cm) (b) ELF Distribution 50 60 12000 10000 u = 38.8 cm. and (c) SRSS. α = 0. V = 7456 kN.

gravity loads included 58 . Heightwise variation of floor displacements and story drift ratios estimated using FEMA-273 force distributions. 5. the errors are largest in the lower stories.” and NLRHA. reaching 35%. larger errors are noted in the upper and lower stories. but it is still inaccurate. 7%. MPA including three “modes.) The pushover analysis procedures considered seem incapable of computing accurately local response quantities. reaching 64%.5 1 1. reaching 31%. and Table 5.3 demonstrate that the hinge plastic rotations estimated by all three FEMA force distributions contain unacceptably large errors. For the ELF distribution. whereas nonlinear RHA computed an insignificantly small value.5 Ground 0 Ground 0 Fig. such as hinge plastic rotations. Modal pushover analysis procedure gives estimates better than all the FEMA force distributions.3. and overestimated by no more than 32%.5 Story Drift Ratio (%) 2 2. For the SRSS distribution. (The 100% error in a 6th floor hinge is ignored as it simply represents that the MPA estimated zero rotation. 3 "Modes" FEMA Uniform ELF SRSS 0.3b and 5. errors are largest in the upper stories. This seems to be an inherent limitation of pushover analysis. at most.4b.Figures 5. For the uniform distribution. In contrast. 3 "Modes" FEMA Uniform ELF SRSS 0. with errors reaching 37% in this example. Figure 5. the MPA procedure is more accurate than all the FEMA force distributions.2 demonstrate that the story drift demands are greatly underestimated by all the FEMA force distributions.4c and Table 5. with story drifts under estimated by.5 Displacement/Height (%) 2 Floor 5th 5th 4th 3rd 2nd 1st NL−RHA MPA. (a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th (b) Story Drift Ratios Floor 4th 3rd 2nd 1st NL−RHA MPA.5 1 1.

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

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

524 0. Peak values of floor displacements (as % of building height = 37.128 1.7 -29.8 17.11E-03 9.7 19.Table 5.789 0.2 9.4 MPA -2.178 1.19E-03 1.730 1.199 27.2 -100.623 1.5 26.399 -27.0 -59.0 -63.098 1.562 1.3 -11. gravity loads included Displacement /Height (%) FEMA ELF SRSS MPA 1.566 1.323 1.466 0.207 1.00E+00 4.330 1.724 0.686 0.992 1.351 0.8 16.6 -73.6 -11.938 1.88E-03 1.083 1.168 1.6 -6.23E-03 3.0 -30.0 -100.089 1.7 -28.3 29.953 0.667 0.109 1.154 1.462 1.314 1.0 -100.6 -4.26E-03 3.84 1.58E-04 6.4 -12.2 Peak values of story drift ratios (as % of story height) from FEMA force distributions and MPA.3 23.9 15.061 1.6 17.2 6.3 Peak values of hinge plastic rotations (radians) from FEMA force distributions and MPA Hinge Plastic Rotation (rad) MPA FEMA ELF SRSS 4.011 1.7 21.00E+00 0.306 1.8 -100.8 44.51E-03 4.93E-03 1.9 -77.00E+00 Error (%) FEMA Uniform ELF SRSS MPA 24.310 1.2 5.064 1.53E-02 1.00E+00 8.7 MPA -2.65E-03 7.399 1.3 -4.9 16.0 37.4 -55.4 16.9 11.10E-02 7.7 26.09E-03 4.5 61 .00E-03 5.530 1.3 -100.6 22.830 2.8 2.55E-03 3.00E+00 0.355 0.927 1.3 31.294 1.6 4.78E-03 0.14 m) from FEMA force distributions and MPA.00E+00 NL RHA 1.9 28.00E-10 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 1.836 0.335 2.673 Error (%) NL RHA 1.34E-03 2.611 0.8 -63.4 -23.975 1.45E-03 3.998 1.221 1.00E+00 0.00E+00 0.353 1.913 0.2 13.94E-03 2.736 0.3 -22.1 -26.414 1.2 -12.195 0.1 0.3 -14.877 1.854 0.1 10.318 2.5 -29.17E-03 9.1.263 0.6 -25.8 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 0.809 0.52E-04 1.23E-02 1.2 10.59E-03 5.5 10.2 16.4 0.1 163.515 FEMA Uniform ELF 27.0 -71.436 Error (%) NL Uniform RHA 0.417 1.75E-03 0.8 -35.344 0.5 22.5 -33.7 FEMA ELF SRSS -22.78E-03 1.8 -32.067 0.399 1.0 -57.547 -27.875 0.5 -27.5 -32.007 1.2 16. gravity loads included Displacement /Height (%) FEMA ELF SRSS MPA 0.16E-03 0.8 -100.8 -2.9 0.262 1.888 0.4 16.35E-10 3.708 0.2 4.00E+00 0.04E-02 8.03E-03 5.50E-03 0.5 15.0 -100.2 19.234 1.6 21.19E-04 5.214 1.209 0.0 -100.6 -17.6 14.998 1.4 21.490 0.7 -15.9 -20.858 1.62E-03 0.5 -29.9 16.367 1.6 -41.487 0.4 -50.860 1.2 9.5 Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 2.9 SRSS -22.839 0.270 0.0 Table 5.530 1.672 1.341 1.00E+00 0.9 -70.7 Table 5.984 1.783 1.372 1.5 -3.015 0.597 0.8 7.2 -4.35E-03 8.560 1.13E-03 5.00E+00 2.1 -3.5 -33.7 7.7 -60.

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

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

5 2 GM Multiplier 2.5 1 1.5 2 GM Multiplier 2.5 2 GM Multiplier 2.5 3 0 0 0. Error in story drifts from three force distributions in FEMA-273 and from MPA including three “modes”.5 1 1.5 3 (c) FEMA: SRSS 60 50 40 60 50 40 3 2 1 (d) MPA (3 "Modes") Error (%) 30 20 Error (%) 30 5 4 8 9 20 9 6 7 7 10 0 0 4 10 2 3 1 8 5 6 0.7.5 1 1.5 3 0 0 0. gravity loads included 64 .5 3 Fig.5 1 1.5 2 GM Multiplier 2. Noted 80 40 9 1 Error (%) 7 Error (%) 60 30 20 10 2 5 6 4 8 40 1 6 20 2 5 3 4 3 7 0 0 0. 5.(a) FEMA: Uniform 100 60 50 9 8 (b) FEMA: ELF Error Envelope Error for Story No.

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

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

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

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

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

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

3. and modes.1) using the FEMA-273 procedure (Building Seismic Safety Council. of the bilinear curve at the target roof displacement. ω n .Appendix A Uncoupled Modal Response History Analysis A.g. respectively. Define the anchor point. iterations may be necessary. for linear-elastic vibration of the building. e.1 STEP-BY-STEP PROCEDURE A detailed step-by-step implementation of the uncoupled modal response history analysis (UMRHA) procedure is presented in this section and illustrated by an example in the next section. The structure should be pushed just beyond the target (or expected) roof displacement in the selected mode. Define the force distribution s* from Eq.20): s* = mφn n n 2. DRAIN2DX (Allahabadi and Powell. develop the base-shear – roof-displacement (Vbn . Let the roof displacement and base shear at the anchor point be urno and Vbno .1.2.1.urn ) pushover curve for the force distribution s* : n 2. Idealize the pushover curve as a bilinear curve (Fig. 3. B.2. 2. A. Apply force distribution of Step 2. Since the target roof displacement may not be known at the start of the procedure. using any numerical integration method. 71 . 1977). Compute natural frequencies.1 incrementally and record the base shears and associated roof displacements. e. This step can be conveniently implemented in any commercially available software. (3. φn .g. 3.. For the nth-“mode”.. 1. Calculate the area under the actual pushover curve. trapezoidal rule. 1988). Apn .

and a point on the actual pushover curve with base shear equal to i 0.6.4. Estimate the yield base shear.4.2.2.0. Calculate the yield displacement. Calculate the post-yielding strain-hardening ratio. A. This value. by connecting a straight line between origin.1.0. urn. Let the point with base shear = Vbny and roof displacement = urny be denoted as A. urny = Vbny kn . Abn . i +1 i i 3.6 × Vbny . 3. Vbny . corresponding to the estimated yield i i i base shear. kn . If the error exceeds some pre-specified tolerance. at base shear i equal to 0.6 .6 .i 3.9.7.4. i i i αn = Vbno Vbny − 1 urno urny − 1 i 3. If desired. O.4 to 3. Calculate initial slope of the idealized bilinear curve. i 3. Vbny . other appropriate methods can be ( used. obtained by judgment. i 3. i i i 3. Calculate area under the bilinear curve OAB. i 3. and B with straight-line segments to obtain the idealized bilinear curve.9. will be refined by an iterative procedure that seeks to equate areas under the actual and the idealized pushover curves.9. i i 3. From the pushover data. 3.8. This step gives the secant stiffness at a base shear equal to 60% of the yield base shear. kn = 0. 3. determine the roof displacement. Calculate the slope. 72 .1. Calculate Vbny = Vbny × Apn Abn . Replace i+1 with i and repeat Steps 3.6 × Vbny ( ) i urn . Draw the curve OAB by connecting the three points O. Calculate the error = 100 × Abn − Apn ( ) ( ) ( ) ) Apn .3.8.6 × Vbny .5. iterations are necessary.

of the combined responses obtained in Step 8. A.2. Vbn Idealization of Pushover Curve Vbno Vbny 0. Dn (t ) .12) and (3.3b) with unit-mass and force-deformation relation of Fig. 9.6 ur n y ur n o ur n Fig. (3. Compute deformation history. An (t ) .eps ur n. Compute the Ln and Γ n from Eq. Calculate peak values.2. 4.0.2). Calculate histories of various responses using Eqs. 4.3. Scale the horizontal axis by Γnφrn to obtain Dno = urno Γ nφrn and Dny = urny Γ n φrn (Eqs.15) and (3. 5. Combine the “modal” responses using Eqs. and pseudo-acceleration history. 4.10b and 4.16). 4. 8.11b).10a and 4. A. r o . 7.4. * 4. Repeat Steps 3 to 6 for as many modes as required for sufficient accuracy.13). In general first two or three modes will suffice. of the nth“mode” inelastic SDF system (Fig.1. (3. 4. A. 6.1 Idealization of nth-“mode” pushover curve 73 .11a). Scale the vertical axis with * Mn to obtain * Fsno Ln = Vbno M n and * Fsny Ln = Vbny M n (Eqs.4) and effective modal mass from M n = Ln Γn .6 × Vbny A • Idealized αnkn B • 1 Actual k 1 • O n figA_1. Develop the Fsn Ln − Dn relation (Fig. (3.

2 Properties of the nth-“mode” inelastic SDF system A. 3. 74 . respectively.4.3. Area under the actual pushover curve.4 cm (10 in.).urn ) pushover curve for the force distribution s* : n 2.5 cm (25 in. and 12.2.3.Fsn / Ln Fsn / Ln − Dn Relationship Vbno / M* n Vbny / Mn * 1 α ω2 n n ω2 1 n figA_2. generated using DRAIN-2DX.1.4. n 2. 3. Following is step-by-step implementation of the procedure described in Section A. The force distributions.). The pushover curves for the first three modes. Idealized bilinear curves for each of the three modes are included in Fig.eps Dny = ur n y / Γnφr n Dno = ur n o / Γnφr n Dn Fig. is defined at the target roof displacement. 3. 3. At this point. Target displacements used to generate these pushover curves are 63. to the north-south component of the El Centro (1940) ground motion scaled up by a factor of 1. are shown in Fig. 25. Ap1 = 360777 kN-cm. s* . A.2 EXAMPLE The UMRHA procedure is implemented to calculate the response of the 9-story building described in Section 3.1. The anchor point. computed for the first three modes are shown in Fig. and third mode. The following steps illustrate the procedure to develop the idealize curve for the first “mode”. 1.2.5 cm and Vb1o = 8729. B. 2. second.6 kN.5. ur1o = 63. 3.1. for the first. The base-shear – roof-displacement (Vbn . First three mode shapes and frequencies of the selected building were computed and are shown in Fig.7 cm (5 in. A.). A.3.

is calculated as follows. 4.9 kN. This value exceeds the prespecified tolerance of 0. i 3. The initial slope of the idealized bilinear curve.1. 3.6 kN. and M1 = 2736789 × 1.1.4. The results for other modes are summarized in Table A. The post-yielding strain-hardening ratio. * 4.1. and α1 = 0.5.09 cm and Vb1 y = 8006.23 cm. Vb1 y = 7615. The next estimate of the yield shear is Vbny = 8006.4. 3.9.86 = 210.1.4 210. i 3. i i i 3. (4.09 cm. L1 = 2736789 kg.194. The first estimate of the yield base shear Vb1 y = 8006.18 kN/cm. Therefore. k1 = 0.7. i i 3. 3. k1 .6 = 4803.4 × (360777 365100 ) = 7911.6. The procedure converged after nineteen cycles to give ur1 y = 36.3. The curve OAB obtained by connecting the three points O.4.6 × Vb1 y ( ) i ur1. The Fs1 L1 − D1 relation for the first “mode” is developed as follows.198%.0.8.6 8006.09 ) − 1 = 0.8 kN.5 38. The yield displacement.18 = 38. Error = 100 × (365100 − 360777 ) 360777 = 1.01%. The point A on the i i bilinear curve is defined by ur1 y = 38. i +1 3. and B with straight-line segments gives the idealized bilinear curve. A.3666 = 3740189 kg. ur1.13).9. Area under the bilinear curve OAB. Also included are the modal damping ratios and the periods calculated from Eq. The results of the iterative procedure are summarized in Table A.4 kN. i i i 3.0.8 22. Γ1 = 1.2.135.6 × Vb1 y = 4803. Ab1 = 365100 kN-cm.2.4 ) − 1 (63. i i 3.6 = 22. Determined from the pushover database.4 kN. α1 = Vb1o Vb1 y − 1 ur1o ur1 y − 1 = ( ) ( ) (8729.3666.2.9. iterations are necessary. ur1 y = Vb1 y k1 = 8006.i 3.86 cm at 0. 75 .

4. 7. 5. 6. 4. The peak values are also plotted in Fig.40 (cm/sec2) and 76 .3. The peak values are computed and are summarized in Tables 4.7.7. A. * M1 gives Fs1o L1 = 233. 9.7.1 and 4. Scaling the vertical axis by Fs1 y L1 = 203. 4.4.51 cm.2. 4. 4. Deformation and pseudo-acceleration histories of the inelastic SDF systems for the first “mode” with unit mass and force-deformation relation developed in Step 4 are plotted in Fig.2. 8. Histories of roof displacement and top story drifts for the first “mode” are computed and presented in Fig. The results were generated for first three “modes’ and are included in Fig.62 (cm/sec2). The combined modal responses are presented in Fig. Scaling the horizontal axis by Γ1φr1 gives D1o = 46.46 cm and D1y = 26.8. 4.

14 y by 6000 Base Shear (kN) 4000 2000 Actual Idealized figA_3c. “Modal” pushover curves for the example building 77 .9 cm.19 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized figA_3a. α = 0. α = 0. V = 5210 kN.eps 10 20 30 40 50 Roof Displacement (cm) (b) "Mode" 2 Pushover Curve 60 8000 u = 9.2 cm. α = 0.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36.eps 0 0 5 10 15 20 Roof Displacement (cm) (c) "Mode" 3 Pushover Curve 25 8000 u = 4. V = 7616 kN.6 cm. A.13 6000 Base Shear (kN) 4000 2000 Actual Idealized figA_3b.eps 0 0 2 4 6 8 10 Roof Displacement (cm) 12 Fig. V y by = 4952 kN.3.

” and third “mode” inelastic SDF systems.” second “mode.52 5 10 15 20 Time (sec) 25 30 0 • 1.2769 D1 (cm) 0 • 35.747 5 10 15 20 Time (sec) figa_4. A.06 0 • 1.m −80 0 −3 0 25 30 Fig. 78 . Histories of deformation and pseudo-acceleration due to 1.33 A1 (g) n=2 0 −80 80 −3 3 n=2 D2 (cm) 0 A1 (g) • 22.4.5 × El Centro ground motion for the first “mode.(a) Deformation 80 n=1 3 (b) Pseuso−acceleration n=1 • 0.223 −80 80 n=3 −3 3 n=3 D3 (cm) 0 A3 (g) • 10.

136 Fsny Ln (cm/sec2) Dny (cm) Fsno Ln (cm/sec ) Dno (cm) Tn (sec) 233.1 1013.79 0. Results of iterative procedure to develop the idealized bilinear curve for the first “mode” inelastic SDF system Itr.86 21.25 36.62 26.7 4580.194 0.9 7615.190 0.3 4603.0 4577.022 0.082 0.5309 488839.74 (kN/cm) 210.029 0.18 210.18 210.4 7714.3 7658.5 4614.192 0.3 7628.6 i k1 i ur1 y i α1 i Ab1 Error (%) 1.9 4570.404 0.4 7672.75 21.2. Properties of “modal” inelastic SDF systems Properties Ln (kg) Γn * M n (kg) “Mode” 1 2736789 1.18 210.180 0.18 210.037 0.194 0.29 36.18 210.59 22.18 210.32 36.18 210.237 0.8 7622.18 210.139 0.26 36.4 4595.186 0.23 22.193 0.135 0.50 36.188 0.74 21. No.09 37.1 4569.18 (cm) 38.Table A.90 21.05 36.78 21.18 210.65 1226.1.103 “Mode” 3 696400 0.8525 1.18 210.176 0.193 0.162 0.18 210.85 36.010 (kN) 8006.86 22.2406 167531.28 36.02 21.2 4628.79 21.70 36.6 4583.25 36.0 7619.013 0.75 21.5 7633.1 7616.18 210.05 52.5 (cm) 22.40 46.2671 1.184 0.309 0.81 21.51 2 “Mode” 2 -920860 -0.910 0.4927 1.0.3 7745.5 7624.83 21.063 0.2 4671.23 0.4 7911.56 19.529 0.11 22.85 0.35 36.107 0.18 210.948 ζ n (%) 79 .64 37.38 22.46 2.194 0.18 210.3666 3740189 203.048 0.0 4704.9 4573.4 7647.56 47.6 × Vb1 y i ur1.5 3109.193 0.1 4574.18 210.44 36.194 (kN-cm) 365100 364060 363276 362684 362235 361892 361631 361432 361279 361162 361073 361004 360951 360911 360880 360856 360838 360824 360813 Table A.8 4647.18 210.18 210.9 (kN) 4803.2 7690.170 0.693 0.198 0.8 4747.12 3876.151 0.6 7840.09 18.3 7786.182 0.39 36.95 21.191 0.8 7618.2 4571.017 0.77 21.59 36.9 4570.24 36.7 7639.0 4588.30 37.18 210. i 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 i Vb1 y i 0.76 21.

80 .

10. Solving Eq.4.1 to the north-south component of the El Centro (1940) ground motion scaled up by a factor of 1. (3. B. 4. This is the same example as solved in Appendix A.6b by solving Eq. 12. From the total hinge rotation.Appendix B Modal Pushover Analysis B. 13.8) 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. Compute the peak deformation. 10. 4. Typically.8).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. of the nth-“mode” inelastic SDF system (Fig.2 EXAMPLE The MPA procedure is implemented to calculate the response of the 9-story building described in Section 3. Steps 1 to 4 of the MPA are the same as those for UMRHA presented in Appendix A.1. or from the inelastic response (or design) spectrum. subtract the yield hinge rotation to determine the plastic hinge rotation. extract from the pushover database values of other desired responses.21).33 cm. Following is step-by-step implementation of the procedure described in Section B. At urno . Repeat Steps 3 to 7 for as many “modes” as required for sufficient accuracy.3b) with unit mass and force-deformation relation of Fig. (3. (4. 11. Determine the total response by combining the peak “modal” responses using the SRSS combination rule of Eq. Calculate the peak roof displacement urno associated with the nth-“mode” inelastic SDF system from Eq. Dn . 81 .5.18). rno . the first two or three “modes” will suffice. (4. 6.

18) are also included in Tables 4. (3. 8.3 and 4.” and the results are included in Tables 4. 9. 12. Also included in Table 4. Steps 3 to 7 are repeated for first three “modes. At ur1o = 48. Peak roof displacement ur1o = Γ1 × φr1 × D1 = 1.28 cm.33 = 48. values of floor displacements and story drifts are extracted.4.4. The values are summarized in Table 4.11. The total response computed by combining the peak “modal” responses using the SRSS combination rule of Eq. Results of Steps 5 and 6 are summarized in Table B. where results for other ground motion intensities are also included.28 cm.366 × 1 × 35. 82 .4 for the story drifts normalized by the story height.3 for the floor displacements normalized by the building height and in Table 4.5 are the plastic hinge rotations computed by subtracting the yield hinge rotation from the total hinge rotations.1.3 and 4.

436 7.52 “Mode” 3 1.969 0.0 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 3.735 3.691 0.35 0.71 1.450 4.856 31.33 1.267 5.0 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 1.38 22.03 0.35 1.513 0.229 8.8451 5.73 24.5 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.379 21.85 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 1.275 1. Calculation of roof displacements urno from peak deformation of inelastic SDF systems Ground Motion Multiplier 0.25 Quantity Dn (cm) “Mode” 1 6.755 27.79 0.126 13.82 1.755 0.37 1.901 8.660 14.678 0.50 35.551 2.225 2.007 36.535 14.367 1.03 26.06 1.395 0.766 7.139 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.312 1.184 0.023 0.676 6.268 0.457 12.748 63.913 22.1.75 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.252 9.18 27.200 0.25 20.577 16.185 11.05 0.332 48.504 18.07 “Mode” 2 4.28 46.36 1.117 5.154 78.4222 3.467 14.13 2.37 57.27 0.690 10.59 0.5 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 2.Table B.52 0.70 0.0 µ = Dn Dny urno = Γnφrn Dn (cm) 83 .332 13.

84

**Appendix C FEMA Force Distribution Calculations
**

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

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

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

Floor, j

mj (10-3×kg)

s* = j

∑i mi

0.112 0.110 0.110 0.110 0.110 0.110 0.110 0.110 0.119

mj

1 2 3 4 5 6 7 8 9

503.5 494.7 494.7 494.7 494.7 494.7 494.7 494.7 534.1

C.2

EQUIVALENT LATERAL FORCE (ELF) DISTRIBUTION 85

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

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

Floor j 1 2 3 4 5 6 7 8 9

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

-3 k

s* j

=

m jhk j

∑i mihik

0.007 0.020 0.038 0.062 0.091 0.126 0.165 0.210 0.281

C.3

SRSS DISTRIBUTION

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

N

∑ n (V jn )

2

to get column 8 of

1 -967.9 832.7 Vj (kN) (8) 2268.6 -359.6 286.6 1233.3 240.7 355.7 101.3.4 250.5 -350.7 (10) 0.6 V j3 (kN) (7) 478.2 148.7 694.9 -153.0 1476.5 -973.5 159.0 136.070 0.6 366.5 215.0 354.0 -5.7 1622.1 87.367 87 .8 374.042 0.9 -166.2 Lateral Force fj (kN) (9) 203.045 0.1 Story Shears V j2 (kN) (6) 1114.1 -646.090 0.2 fj = ∑ fi i fj (2) (3) (4) 59.4 400.6 -732.2 200.5 320.2 95.9 880.9 366.3 -6.065 0.7 95.0 176.2 105.1 832.7 -46.8 -326.9 1683.1 1857.3.7 1578.7 525.0 1231.For convenience.177 0.8 430. Table C.7 2065.1 -438. 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) 1917.9 446.4 1759.6 -352.7 234.0 381. the lateral forces are normalized by the base shear to obtain column 10 in Table C.4 1842.2 277.9 1446.5 -320.047 0.6 319.098 0.3 222.2 285.1 -525.8 1381.3 -646.7 374.0 980.2 97.

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

a research report modal pushover analysis

a research report modal pushover analysis

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

Are you sure?

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

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

Get the full title to continue

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

scribd