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

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

Pacific Earthquake Engineering Research Center

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

Theory and Preliminary Evaluation

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

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

PEER 2001/03

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

Anil K. Chopra University of California Berkeley

Rakesh K. Goel California Polytechnic State University San Luis Obispo

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

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

ii .

However. in the modal expansion of the effective earthquake forces. 2. is determined by combining the rno n = 1.0. p eff . This pushover curve is idealized as bilinear n and converted to the force-deformation relation for the nth-“mode” inelastic SDF system. Thus the MPA procedure is accurate enough for practical application in building evaluation and design. The base shear-roof displacement Vbn .urn curve is developed from a pushover analysis for force distribution s* .sn ug t . The results presented for El Centro ground motion scaled by factors varying from 0. and identifies locations of most plastic hinges. the total demand. 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.ABSTRACT The principal objective of this investigation is to develop a pushover analysis procedure based on structural dynamics theory. all pushover analysis procedures considered do not seem to compute to acceptable accuracy local response quantities. ro . MPA. such as hinge plastic rotations. iii .25 to 3. where m is the n mass matrix and φ n its nth-mode.mι ug t .… according to an appropriate modal combination rule. p eff t = . and the structure is pushed to the roof displacement determined from the peak deformation Dn of the nth-mode elastic SDF system. The peak deformation of this SDF system is used to determine the roof displacement. The standard response spectrum analysis (RSA) for elastic buildings is reformulated as a Modal Pushover Analysis (MPA). show that MPA estimates the response of buildings responding well into the inelastic range to a similar degree of accuracy when compared to standard RSA for estimating peak response of elastic systems. but provides superior accuracy in estimating seismic demands on buildings. and the MPA procedure is more accurate than all the FEMA force distributions methods in estimating seismic demands. Combining these peak modal responses by modal combination rule leads to the MPA procedure. Instead. plastic hinge rotations are less accurate.n t = . it is demonstrated that the FEMA force distributions greatly underestimate the story drift demands. The MPA procedure is extended to estimate the seismic demands for inelastic systems: First. at which the seismic response. Second. is determined by pushover analysis. which retains the conceptual simplicity and computational attractiveness of current procedures with invariant force distribution. 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. 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. a pushover analysis determines the peak response rno of the inelastic MDF system to individual modal terms. The peak response of the elastic structure due to its nth vibration mode can be exactly determined by pushover analysis of the structure subjected to lateral forces distributed over the height of the building according to s* = mφ n . Comparing the earthquake-induced demands for the selected 9-story building determined by pushover analysis using three force distributions in FEMA-273. rno . and nonlinear RHA.

iv .

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

vi .

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

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

. 1997]. Bracci et al. 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. However. Gupta and Krawinkler. 2000] have led to good estimates of seismic demands. 2000]. they are conceptually complicated and computationally demanding for routine application in structural engineering practice. Obviously. 2000. Krawinkler and Seneviratna. 1999. 1999]. Kim and D’Amore. several researchers have proposed adaptive force distributions that attempt to follow more closely the time-variant distributions of inertia forces [Fajfar and Fischinger. Gupta and Krawinkler. 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. While these adaptive force distributions may provide better estimates of seismic demands [Gupta and Kunnath. 1988. 1998. Skokan and Hart.and medium-rise structures in which inelastic action is distributed throughout the height of the structure [Krawinkler and Seneviratna. Gupta and Kunnath. Lawson et al. 1981. To overcome these limitations. 1998. While nonlinear response history analysis (RHA) is the most rigorous procedure to compute seismic demands. requires explicit consideration of inelastic behavior of the structure. current structural engineering practice uses the nonlinear static procedure (NSP) or pushover analysis in FEMA-273 [Building Seismic Safety Council. such satisfactory predictions of seismic demands are mostly restricted to low. but investigations [Saiidi and Sozen.. None of the invariant force distributions can account for the contributions of higher modes to response. 1994. 1997. 1999. or for a redistribution of inertia forces because of structural yielding and the associated changes in the vibration properties of the structure. such as life safety and collapse prevention. Fajfar and Fischinger. 2000]. 1988. 1991. Attempts 1 . 1999.1 Introduction Estimating seismic demands at low performance levels. Maison and Bonowitz. Miranda. after the structure yields both assumptions are approximate.

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

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

and (b) bilinear hysteretic forcedeformation relation For a given excitation u g t . Section 7.3). (a) Idealized one-story structure. sign u = .fS m fs c ut u fy 1 αk k 1 uy ug -fy k k um u 1 1 (a) (b) Fig.. z y . m depends on three system parameters: w n . This becomes evident if Eq. 2001.e. and z is the damping ratio of the system vibrating within its linear elastic range (i.1.ug t n where bg bg b g bg (2.3) ωn = k m ζ = c 2mω n f fs = s fy (2. u £ u y ).4) and w n is the natural vibration frequency.5) For a given u g t . Tn = 2p w n is the natural vibration period. the deformation u t depends on three systems parameters: w n . z .1) is divided by m to obtain ~ u + 2zw n u + w 2 u y f s u. 2. in addition to the form of the force-deformation relation. deformation is denoted by um . bg 4 . and the ductility factor is u m= m uy (2. (2. and R y (Chopra . or absolute (without regard to algebraic sign) maximum. and u y . The peak.

4 in Chopra.9 kN (38. one-bay frame 50 uy = 1. m h = 3.2 SYSTEM AND EXCITATION CONSIDERED Consider the one-story system in Fig. Pushover curve for structure shown in Fig.2 kips) . with Tn = 0.26 kN.04 40 Base Shear. α = 0. 6.32 m • Fig. f o w = 184 .26 kN (8. The yield strength of the inelastic system. is f y / w = ( f o / w ) ÷ 8 = 0. u (cm) 8 Fig.66 m • EIb EIc EIc • • L = 7. 2. 2. One-story. 2001) and scaled up by a factor of 2. and f y = 39.826 kips) for w = 169.376 cm. this system and excitation.2311 . Vby = 39.2 5 .2: the dimensions and flexural rigidity of the structural elements are noted.1. 2. based on Ry = 8 .2. For . Vb (kN) 30 20 10 0 0 2 4 6 Roof Displacement.2.3.5 sec and z = 5% subjected to the north-south component of the El Centro (1940) ground motion (Fig. 2.

pushover analysis cannot provide any cumulative measure of response.017 rad . and q pm = 0. 2.g.3 RESPONSE HISTORY ANALYSIS Figure 2.4e. the peak value determined from RHA.4 shows the earthquake response of the system described in the preceding section determined by response history analysis (RHA). The resulting pushover curve is shown in Fig. 2. 2.4f.35. as apparent in Fig. the joint rotation and beam hinge rotation are identical to values q m and q pm determined from RHA.) and 50.0217 rad . Determined from pushover analysis at the exact peak deformation.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. 2.36 cm .) for the beam and columns.077 × 107 mm 4 (146 in. (b) shows the lateral force f s t or base shear Vb t normalized relative to the bg weight. wherein the hysteretic force-deformation history of Fig. and E = 2 × 108 kPa (29×103 ksi) gives yield moments of 21. The peak values of the various response quantities are as follows: um = 7. e. (d) shows the rotation q p bt g of the plastic hinges at the bg bg beam ends.18 kN-m (444.134 × 107 mm 4 (75.4 ). Observe that the pushover curve matches the initial loading path of the hysteretic system.36 cm. the ductility factor m = 5.3. It is organized in five parts: (a) shows the deformation u t .1 kip-in. I b = 3. The system is excited well beyond the yield deformation. and (e) shows the force-deformation relation. 2. the energy dissipated in 6 .65 kN-m (191. 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 .3 in.6 kip-in. 2. The yielding stiffness of each structural element is defined as 3% of its initial stiffness. A static pushover analysis of this one-story system leads to the force-displacement relationship shown in Fig.The yield moments in the beam and columns are defined as the bending moments due to the lateral force f y . (c) shows the joint rotation q bt g .4 ).. However.4edetermined from RHAis superimposed. Implementing this analysis with I c = 6. q m = 0. respectively.

15 u (cm) 0 • 7.4 −10 −5 0 u (cm) 5 10 Fig. and (f) pushover curve 7 .yielding during the ground motion.017 • 0 (d) θp (rad) −0.04 0.2311 (a) Vb / w 0 f / w = −0.04 0 5 10 15 Time (sec) 20 25 30 0.2311 y (b) −0. (c) joint rotation.4 −10 −5 0 u (cm) 5 10 −0.5 fy / w = 0.2 −0.4. 2.2 0 (e) 0 (f) −0.04 0. (e) force-deformation relation. or the cumulative rotation at a plastic hinge. This represents an inherent limitation of pushover analyses.36 −15 0.2 Vb / w 0. (b) base shear.0217 • θ (rad) 0 (c) −0. (d) plastic hinge rotation.4 0.5 0. Response of one-story system to El Centro ground motion: (a) deformation.2 −0.04 0.4 0.

8 .

4) 9 . classical damping. c. and lateral stiffness matrices of the system. Elastic Multistory Buildings 3.1 MODAL RESPONSE HISTORY ANALYSIS The differential equations governing the response of a multistory building to horizontal earthquake ground motion u g t are as follows: mu + cu + ku = . (3.n t = Â . The right side of Eq. This force distribution can be expanded as a summation of modal inertia force distributions s n (Chopra. m.m ι u g t bg bg (3.1) where u is the vector of N lateral floor displacements relative to the ground.2) The spatial distribution of these “forces” over the height of the building is defined by the vector s = m ι and their time variation by u g t .3. 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) where φ n is the nth natural vibration mode of the structure. 2001: Section 13.1) can be interpreted as effective earthquake forces: p eff t = . and k are the mass.mι ug t bg bg bg (3.12): mι = n =1 Â sn = N n =1 Â Γ nmφ n N (3. and p eff t = Â peff.

n t is entirely in the nth-mode.sn ug t bg bg bg bg (3.9) and (3.7) By utilizing the orthogonality property of modes. it can be demonstrated that none of the modes other than the nth mode contribute to the response.u g t n Comparing Eqs. Then the floor displacements are u n t = φ n qn t bg bg bg bg (3.9) to the equation of motion for the nth-mode elastic SDF system. The solution qn t can readily be obtained by comparing Eq. subjected to u g t : Dn + 2z nw n Dn + w 2 Dn = .9) in which w n is the natural vibration frequency and z n is the damping ratio for the nth mode.10) gives qn t = Gn Dn t bg bg bg (3.8) gives the floor displacements u n t = Gnφ n Dn t bg (3.11) and substituting in Eq. an SDF system with vibration propertiesnatural frequency w n and damping ratio z n of the nth-mode of the MDF system.sn ug t bg (3.12) . (3. we will outline that the response of the MDF system to p eff . The equations governing the response of the system are mu + cu + ku = . (3. (3.5) The contribution of the nth mode to s and to p eff t are: sn = Gn mφ n respectively.n t = .n (t ) = N n =1 ∑ −sn u g (t ) N (3.peff (t ) = n =1 ∑ peff .Gn ug t (3.8) where the modal coordinate qn t is governed by 2 qn + 2z nw n qn + w n qn = . with no contribution from other modes.6) Next.10) bg bg bg 10 (3. p eff .

2001.9) is the standard modal equation governing qn t . Equations (3. However. (3. Conceptual explanation of modal response history analysis of elastic MDF systems Any response quantity r t story drifts.13) st where rn denotes the modal static response.12) and (3.16) reflect combining the response contributions of all modes. internal element forces.14) is the pseudo-acceleration response of the nth-mode SDF system (Chopra.1). Section 13. (3. and An t = w 2 Dn t n bg (3. Therefore.15) rt = bg n =1 Â rn t = bg n =1 Â rnst An bt g N (3. these standard equations have been derived in an unconventional way. (3.1. Eqs. st The two analyses that lead to rn and An t are shown schematically in Fig. 3.n t .13) represent the response of the MDF system to p eff. etc. 3.can be expressed by st rn t = rn An t bg bg bg bg bg (3.16) This is the classical modal RHA procedure wherein Eq. In contrast to the classical 11 bg .12) and (3. the static value of r due to external forces sn .15) and (3.13) define the contribution of the nth-mode to the response.1. ζn ¨ g(t ) u (a) Static Analysis of Structure (b) Dynamic Analysis of SDF System Fig. the response of the system to the total excitation p eff t is ut = bg bg bg n =1 N Â un bt g = Â Gnf n Dn bt g n =1 N N (3. and Eqs.Forces sn An(t ) st rn ωn.

g.17) where An is the ordinate A Tn . the peak nth-mode response as in Eq. 2001. (3.derivation found in textbooks (e.17) (Chopra.16). Section 13.1). Alternatively. and Tn = 2p w n is the natural vibration period of the nth-mode of the MDF system. This concept provides a rational basis for the modal pushover analysis procedure developed later. (3.19) will provide the same value of rno .4 and 13. the peak value rno of the nth-mode contribution rn t to response r t is determined from st rno = rn An bg bg bg (3.18) To develop a pushover analysis procedure consistent with RSA. The peak modal responses are combined according to the Square-Root-of-Sum-ofSquares (SRSS) or the Complete Quadratic Combination (CQC) rules. Sections 12. z n of the pseudo-acceleration response (or design) spectrum for the nth-mode SDF system. which is valid for structures with well-separated natural frequencies such as multistory buildings with symmetric plan.3).9) . we used the modal expansion of the spatial distribution of the effective earthquake forces. 2001. we note that static analysis of the structure subjected to lateral forces f no = Γ n mφ n An (3.1. Chopra. 3. In such a response spectrum analysis (RSA).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.8. provides an estimate of the peak value of the total response: b g F N r 2 I 1/ 2 ro ª G Â no J H n =1 K 3..(3.3 MODAL PUSHOVER ANALYSIS (3. this response value can be obtained by static analysis of the structure subjected to lateral forces distributed over the building height according to 12 . The SRSS rule.

Typical floor-to-floor heights (for analysis purposes measured from center-of-beam to center1 2 Brandow & Johnston Associates. the peak value of the roof displacement due to the nth-mode.2). 13 . The bays are 9. The levels of the 9-story building are numbered with respect to the ground level (see Fig. The columns are 345 MPa (50 ksi) steel wideflange sections. with the ninth level being the roof. (3. SAC Steel Project Technical Office. 3. 46th Street.20) and the structure is pushed to the roof displacement. The building has a basement level.15 m (30 ft) on center. Obviously Dn and An are available from the response (or design) spectrum. 3. 3.12) is urno = Gnf rn Dn (3.18) to obtain an estimate of the peak value ro of the total response. in both directions. SAC is a joint venture of three non-profit organizations: The Structural Engineers Association of California (SEAOC). Third St. urno . 1301 S. shown in Fig. CA 94804-4698. Richmond.21) 2 where Dn = An ω n .19 m (122 ft) in elevation. CA 90017. the Applied Technology Council (ATC).. California. Although not actually constructed. region. Los Angeles.2). The interior bays of the structure contain simple framing with composite floors. 1660 W. with five bays each in the north-south (N-S) and east-west (E-W) directions.4 COMPARATIVE EVALUATION OF ANALYSIS PROCEDURES 3. was designed by Brandow & Johnston Associates1 for the SAC2 Phase II Steel Project. and 37. (3. The building’s lateral load-resisting system is composed of steel perimeter moment-resisting frames (MRFS) with simple framing on the farthest south E-W frame. The peak modal responses. rno .73 m (150 ft) in plan. each determined by one pushover analysis. this building is 45. can be combined according to Eq.s* = mφ n n (3. Consulting Structural Engineers. and California Universities for Research in Earthquake Engineering (CUREE).1 System and Excitation Considered The 9-story structure. which from Eq. This modal pushover analysis (MPA) for linearly elastic systems is equivalent to the well-known RSA procedure (Section 3. A benchmark structure for the SAC project.73 m (150 ft) by 45. this structure meets seismic code and represents typical medium-rise buildings designed for the Los Angeles.4. denoted B1.2.

Concrete foundation walls and surrounding soil are assumed to restrain the structure at the ground level from horizontal displacement.of-beam) are 3. 3. The seismic mass of the above ground levels of the entire structure is 9.2 kips-sec2/ft). and for the ninth level is 1. The column bases are modeled as pinned and secured to the ground (at the B-1 level). for the second through eighth levels is 9.3. floor slabs.0 kips-sec2/ft). mechanical/electrical. The first three vibration modes and periods of the building for linearly elastic vibration are shown in Fig.65 m (12 ft) and for the first floor is 5. This model is based on centerline dimensions of the bare frame in which beams and columns extend from centerline to centerline. the vibration periods are 2. To ensure that this structure remains elastic. 3. The force distributions.49 m (18 ft). 14 .2. These force n distributions will be used in the pushover analysis to be presented later.85. partitions. Column splices. if desired more complex models. The column lines employ two-tier construction. fifth.0 kips-sec2/ft).65×105 kg (66. The floor-to-floor height of the basement level is 3. s* (Eq. for the first level is 1. third. respectively.01×106 kg (69.83 m (6 ft) above the center-line of the beam to column joint.7 kips-sec2/ft). dimension.89×105 kg (67. 0.27.sec2/ft). i.4. and 0. The 9-story N-S MRF is depicted in Fig. The simple model adopted here is sufficient for the objectives of this study. 3. we select a weak ground motion: the northsouth component of the El Centro (1940) ground motion scaled down by a factor of 4. roofing and a penthouse located on the roof. ceiling/flooring. The seismic mass of the structure is due to various components of the structure.00×106 kg (616 kips.e. The seismic mass of the ground level is 9. 3.07×106 kg (73.49 sec. The building is modeled in DRAIN-2DX (Allahabadi and Powell. which are seismic (tension) splices to carry bending and uplift forces. including the steel framing. The floor system is composed of 248 MPa (36 ksi) steel wide-flange beams in acting composite action with the floor slab.. The strength. for the first three modes are shown in Fig.96 m (13 ft). and shear distortion of panel zones are neglected but large deformation (P∆) effects are included. are located on the first. Each frame resists one half of the seismic mass associated with the entire structure. 1988) using the M1 model developed by Krawinkler and Gupta (1998). and seventh levels at 1. such as those described in Gupta and Krawinkler (1999) can be used. monolithic column pieces are connected every two levels beginning with the first level.20).

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

2 Response History Analysis The structural response due to individual vibration modes.94 2. is shown in Figs. Each figure is organized in four parts: (a) shows the roof displacement urn t .728 2.33 2.9th 8th 7th 6th T = 0.85 sec 2 Floor 5th 4th 3rd 2nd 1st T1 = 2.1 3. 3.44 1.49 sec 3 T = 0.13)].05 2.27 sec Ground −1.487 −1.05 s * 1 s * 2 s * 3 Fig. respectively.4. and 3 .31 −0.51 0.04 1.5 −1 −0. and 3 n 3.3. 3.61 2. (c) shows the joint rotation q n t of an external joint at 16 bg bg bg . First three natural-vibration periods and modes of the 9story building 3. (b) shows the base shear Vbn t normalized relative to the weight W of the building.75 1.67 −1.13 −1.12) and (3.5 0 0.4.1 −2.5 Mode Shape Component 1 1.72 −2. 3.37 2. determined by RHA [Eqs.0272 −2. 2. Force distributions s* = mφn . n = 1.03 −1. and 3.8 −2.38 0.05 1. n = 1.12 0.5. 3.7. 2.796 0.5 Fig. (3.39 3.6.93 −1.

The peak values of floor displacements and story drifts determined by RHA. respectively. and ur 3o = 0. larger in story drifts. the results for the roof displacement and top-story drift are shown in Fig. where the second and third modal responses are a larger percentage of the top story drift compared to roof displacement.2. and even larger in joint rotations.16)]. the peak roof displacement due to each of three modes is ur1o = 9. errors generally decrease as response contributions of more modes are included. ur 2o = 2. consistent with the increasing significance of the higher mode response among these three sets of response quantities. with the first three modes provide essentially the exact response. which are listed in Tables 3.12 cm. It is apparent that the first mode alone is inadequate. but three modes—perhaps even two modes—are sufficient. The errors due to truncating the contributions of vibration modes beyond the third mode are negligible.9. This is illustrated in Fig.3.422 cm. the exact response considering all modes. The peak values of displacements of all floors. and the percentage errors due to truncation of higher modal contributions.2.15) and (3. and rotations of external joints with moment connections are presented in Tables 3. drifts in all stories. The same method was used to determine the peak values of many response quantities. including one. For a fixed number of modes included. or all modes. Combining the modal response histories for all modes gives the total response [Eqs. 3. and (d) shows the Vbn . two. in particular.the roof level. two. and three vibration modes. respectively. As expected. three. The linear relationship between the base shear and roof displacement for each mode implies that the structure did not yield.3. are presented in Fig. 3. Also included are the combined response due to one. 3. 17 . and 3. especially in estimating the story drifts. 3.1. and 3.8.urn relation.1. 3. errors are smallest in floor displacements. (3.23 cm. The peak values of the various response quantities are noted in these figures.8.

85E-03 3.4 -1.235 0.1 3.282 0.74E-03 1.226 0.090 0.89E-03 1.265 RHA (all modes) 0.09E-03 Error (%) 1 Mode -23.042 0.1 -2.2 -1.475 0.56E-03 2.024 -0.03E-03 1.235 0.311 0.9 -24.307 0.3 -0.265 0.011 0.3 Peak values of joint rotations (radians) from RHA for 0.8 -56.0 -0.156 0.6 -1.022 0.097 0.012 0.9 -22.227 0. 18 .177 0.1 Table 3.413 RHA (all modes) 0.370 0.2 -57.003 0.7 -50.0 -10.026 0.133 -0.13E-04 9.29E-03 2.237 0.0 -2.9 8.466 0.058 -0.09E-03 1.74E-03 1.2 0.062 0.63E-03 2.50E-03 2.9 9.060 -0.258 0.14E-03 2.0 -2.38E-04 2 Modes 2.333 0.002 -0.65E-03 2.261 -0.99E-03 Mode 3 3.264 Error (%) 1 Mode 2 Modes 3 Modes -23.6 -1.350 0.09E-03 2.00E-03 1.173 0.032 -0.73E-03 3.202 0.7 4.001 -0.069 0.15E-03 4.8 1.010 -0.125 0.1 -11.45E-03 3.08E-03 2.78E-04 -3.229 0.156 0.28E-04 1.6 0.14 m3) from RHA for 0.225 0.229 0.47E-03 1.054 0.25 × El Centro ground motion Displacement /Height (%) Modal Response Combined (RHA) Mode 1 Mode 2 Mode 3 1 Mode 2 Modes 3 Modes Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 0.11E-03 1.0 3 Modes -5.4 -6.260 0.1 -19.90E-03 3.03E-03 -6.325 0.023 0.9 -16.1 -14.1 Building height is measured from the ground floor to the 9th floor.260 0.4 -10.8 -1.64E-03 3.38E-03 2.364 0.5 -1.069 0.4 -22.336 0.6 4.44E-03 1.74E-04 6.1 4.4 0.66E-05 -3.00E-03 2.042 0.65E-03 RHA (all modes) 2.130 0.303 0.26E-04 -5.0 -46.1 -0.42E-04 1.400 0.183 0.3 19.5 18.7 3.453 0.055 0.035 0.6 11.94E-03 2.3 -0.4 -1.03E-03 3 Modes 2.2 -20.121 0.5 -2.63E-03 2.8 -15.25 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 3 Modal Response Mode 1 2.259 0.3 -8.253 0.9 2.2 -4.311 0.009 -0.399 0.6 0.231 -0.199 0.159 0.192 0.03E-03 1.7 2.2 9.6 -0.097 0.2 Peak values of story drift ratios (as % of story height) from RHA for 0.01E-04 3.202 0.1 -2.008 -0.6 1.76E-03 1.205 0.300 0.157 0.259 0.50E-03 4.9 -15.7 2 Modes -3.7 -19.089 0.6 9.179 0.09E-03 1.011 0.321 0.9 -23.253 0.88E-03 2.9 1.245 0.038 0.080 0.227 0.152 0.3 -33.4 -3.202 0.197 0.152 0.071 0.282 0.24E-03 2.72E-03 3.00E-03 1.3 -3.295 -0.125 0.9 3.124 0.203 0.2 1.8 1.8 -10.173 0.01E-04 -2.8 -5.31E-03 Joint Rotation (rad) Combined (RHA) 1 Mode 2.2 -2.088 -0.266 0.062 -0.060 -0.11E-04 -5.045 0.177 0.266 0.88E-03 2.275 0.060 0.99E-03 2.181 0.378 0.263 0.015 -0.76E-03 1.124 0.1 Peak values of floor displacements (as % of building height = 37.266 0.006 0.407 0.401 Error (%) 1 Mode 2 Modes 3 Modes -23.31E-03 Mode 2 -1.7 7.0 7.406 0.043 0.91E-04 1.13E-03 2.226 0.33E-03 2.5 0.74E-04 9.22E-03 2.4 -0.4 -41.310 0.237 0.44E-03 3.1 -2.4 -7.097 0.Table 3.012 -0.117 0.25 × El Centro ground motion Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Modal Response Mode 1 Mode 2 Drift Ratio (%) Combined (RHA) 1 Mode 2 Modes 3 Modes Mode 3 0.245 0.3 Table 3.4 -53.89E-03 1.

12 0 (d) 0 (e) 15 Fig.1 /W 0. (d) force-deformation history.1 −15 −7.0435 Vb1 / W 0. Response due to first mode: (a) roof displacement.1 0.5.5 9. (c) joint rotation.1 −15 −7. (b) base shear.12 −0.5 15 −0. Excitation is 0.5 0 u (cm) r1 7.1 0.5 0 u (cm) r1 7.1 5 x 10 −3 θ (rad) 0.00174 • 0 (c) r1 −5 0 5 10 15 Time (sec) 20 25 30 0.0435 9. 3.0435 • 0 (b) V b1 −0. and (e) pushover curve.25 × El Centro ground motion 19 .15 u (cm) 9.12 • 0 (a) r1 −15 0.

0252 • 0 (b) V −0.5 u (cm) r2 0 • 2.5 5 −0. and (e) pushover curve.5 5 Fig.25 × El Centro ground motion 20 .23 0.00244 • 0 (c) r2 −5 0 5 10 15 Time (sec) 20 25 30 0. Excitation is 0.23 −5 (a) 0.1 −5 −2.0252 −2. (d) force-deformation history.1 0.6.23 −2. (b) base shear. 3.05 /W 0.5 0 u (cm) r2 2.1 −5 −2.0252 (d) 0 0 (e) −0.5 0 u (cm) r2 2.1 Vb2 / W 0. (c) joint rotation.05 5 x 10 −3 b2 θ (rad) 0. Response due to second mode: (a) roof displacement.

000913 • 0 (c) r3 −2 0 5 10 15 Time (sec) 20 25 30 0.02 2 θ (rad) 0. Excitation is 0. Response due to third mode: (a) roof displacement. (c) joint rotation.5 1 Fig.422 (e) −0.0108 −0.0108 −0.422 /W V b3 0 (d) 0 −0. (d) force-deformation history.02 0.0108 x 10 −3 (b) −0.02 −1 −0. (b) base shear.5 0 u (cm) r2 0.02 −1 −0.5 1 −0.5 0 u (cm) r3 0.1 u (cm) r3 0 • 0.422 −1 (a) 0.02 −0.25 × El Centro ground motion 21 . and (e) pushover curve.02 /W V b3 0 • 0. 3.7.

5.2. 3.e. leads to the pushover curves shown in Figs. 3.4) to roof displacement ur1o = 9. (3.6 present estimates of the combined response according to Eq.7d). Figure 3. and 3. 3. considering one. 3. confirming that MPA gives the exact values of the individual modal responses. story drifts. 3. these pushover curves are consistent with the Vb .1 through 3. and the errors in these estimates relative to the exact response from RHA considering all modes. the value determined by RHA (Fig.8) leads to the pushover curve shown in Fig.e. and ur 3o = 0. (3. 3. although both analyses led to identical peak values of the individual modal responses.9).4.6. 3.5.. i.3.12 cm. except for the algebraic sign associated with Γ1 and minor round-off errors. the floor displacements are proportional to the mode shape φ 1 because the structure remains elastic. 3. and external joint rotations in Tables 3.10 and Tables 3.4. For a fixed number of modes included considered..5d).7e and to the floor displacements. This pushover curve is consistent with the relationship between the base shear and roof displacement determined by RHA (Fig. two. (3. story drifts. Observe that the target roof displacement in each pushover analysis is identical to its exact value determined by RHA.ur relations determined by RHA (Figs. respectively.21). i. implying that the modal truncation errors are small if at 22 .1.6d and 3. and 3.4. respectively. The floor displacements.20) with n = 1 (Fig. the errors in the MPA results are generally larger than in RHA (Fig.1.20) with n = 2 and 3 up to roof displacements ur 2 o = 2.4. These values of the response quantities are identical to the peak response values determined from RHA (Tables 3. (3.23 cm .3). pushing the structure using the force distribution of Eq. and external joint rotations computed by pushover analysis are presented in Tables 3. this value would be determined directly from the response (or design) spectrum. and 3. and 3. which would provide the Dn value to be substituted in Eq. As for the first mode.12).3 Modal Pushover Analysis Implementing MPA for the fundamental vibration mode.3). and it is apparent in the example considered that three modes provide most of the response (Fig.18). pushing the structure. As suggested by Eq.3). or three vibration modes.9 and Tables 3.5. In RHA the errors arise only from truncating the responses due to higher modes.2. 3. 3. (3. using the force distribution of Eq. and 3. 3.6. 3.422 cm . and the computed response values are identical to the peak response values determined from RHA (Tables 3.5e.6e and 3. respectively. 3. Implementing pushover analysis for the second and third modes. In practical application.

8. Heightwise variation of floor displacements and story drifts from RHA for 0.83 • r3 0 • 0.eps r 5 10 15 20 Time. 3. sec 25 30 Fig.23 r2 −15 15 0 • 0.59 • Total (All Modes) 0 −15 0 ∆ (cm) u (cm) 0 −3 0 fig3_8.1 0.03 −3 3 Mode 3 ∆ (cm) u (cm) r3 −15 15 9.422 r2 0 • 1. Additional errors are introduced in pushover analysis due to the approximation inherent in modal combination rules. Response histories of roof displacement and top-story drift from RHA for 0. 3.2 0.1 0.least three modes are included.9. sec 25 30 r 5 10 15 20 Time.eps fig3_9b.48 −3 3 1.2 0.6 Fig.4 Story Drift Ratio (%) 0. (a) Roof Displacement 15 (b) Top Story Drift 3 Mode 1 ∆ (cm) u (cm) 9.685 • −3 3 Mode 2 ∆ (cm) u (cm) • 2.3 0.25 × El Centro ground motion 23 .5 0.4 Displacement/Height (%) 0.12 • r1 −15 15 0 r1 0 0 0.5 Ground 0 0.eps Ground 0 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.3 0.

106 0.173 0.253 -0.260 -0.062 0.94E-03 2.261 -0.25 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Modal Response Mode 1 -2.266 0.0 -18.03E-03 -1.38E-03 3 Modes 2.285 0.260 0.7 -15.9 -15.157 0.9 -11.15E-03 2.270 0.09E-03 -1.009 0.322 0.89E-03 2.177 0.89E-03 2.44E-03 -1.31E-03 Joint Rotation (rad) Combined (MPA) 1 Mode 2.3 1.235 0.097 0.22E-04 2.3 -41.1 -0.0 -2.069 0.407 0.57E-03 1 Mode -23.203 0.8 -14.124 0.4 -4.9 -15.036 -0.89E-03 -1.9 -13.31E-03 Error (%) RHA (all modes) 2.230 0.44E-03 3.92E-04 -1.157 0.264 Error (%) Modal Response 1 Mode 2 Modes 3 Modes -23.058 -0.9 -12.156 0.99E-03 Mode 3 -3.022 0.300 0.133 0.7 -50.76E-03 1.00E-03 1.09E-03 Mode 2 1.9 -16.1 -11.76E-03 -1.270 0.267 0.7 2.042 0.63E-03 2.282 -0.78E-04 2 Modes 2.00E-03 2.72E-03 3.245 -0.4 -22.63E-03 -2.9 -13.31E-03 2.0 -10.173 0.9 -13.226 0.6 Peak values of joint rotation (radians) from MPA for 0.9 -14.89E-03 -2.197 0.3 -13.3 Table 3.12E-03 1.276 0.038 -0.090 0.0 -0.7 -21.4 -5.03E-03 1.385 0.8 -23.133 0.00E-03 -1.7 2 Modes -13.74E-03 -1.25 × El Centro ground motion Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Displacement /Height (%) Modal Response Combined (MPA) Mode 1 Mode 2 Mode 3 1 Mode 2 Modes 3 Modes 0.65E-03 2.270 0.05E-03 3.235 -0.310 0.062 -0.25 × El Centro ground motion Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Modal Response Mode 1 Mode 2 Drift Ratio (%) Combined (MPA) 1 Mode 2 Modes 3 Modes Mode 3 -0.43E-04 -1.6 -16.1 -11.4 -4.03E-03 3.73E-05 3.011 0.048 0.04E-03 3.96E-03 2.125 0.9 -15.9 -18.259 -0.9 3 Modes -12.2 -11.9 -15.117 0.9 -14.374 0.097 0.229 0.1 -19.011 -0.2 -12.09E-03 1.3 -33.4 -2.260 0.0 -16.331 0.08E-03 2.6 -19.14 m) from MPA for 0.4 -9.274 0.4 -19.9 -8.055 0.90E-04 -9.080 0.203 0.001 0.42E-04 -1.203 0.253 RHA (All Modes) 0.045 -0.4 -9.33E-04 5.133 -0.4 -22.2 -0.7 24 .15E-03 1.048 0.010 0.8 -15.227 0.008 -0.229 0.286 0.89E-03 1.9 -13.328 0.401 Error (%) 1 Mode 2 Modes 3 Modes -23.088 -0.026 0.74E-04 -6.00E-03 3.106 0.00E-03 2.28E-03 2.73E-03 3.023 -0.24E-03 -2.03E-03 6.3 -9.370 0.336 RHA (all modes) 0.321 0.3 -14.09E-04 -3.9 -24.267 0.47E-03 1.156 0.38E-03 2.237 0.060 -0.22E-03 -2.237 -0.6 -17.2 -4.00E-03 2.152 0.0 -46.296 -0.2 -16.012 -0.97E-03 1.282 0.4 Peak values of floor displacements (as % of building height = 37.7 -19.177 0.042 0.4 0.012 0.Table 3.8 -15.043 -0.1 -18.152 0.231 -0.230 -0.4 -14.4 -7.267 -0.015 -0.002 0.078 0.313 0.069 0.272 0.4 -11.8 -22.006 -0.5 -16.90E-03 1.466 0.40E-04 5.259 0.179 0.2 -57.245 0.079 0.203 0.024 0.032 0.3 -12.336 0.6 -15.253 0.060 0.2 1.071 0.181 0.3 -2.4 -53.3 -19.80E-04 3.2 Table 3.53E-04 -9.253 0.5 Peak values of story drift ratios (as % of story height) from MPA for 0.121 0.74E-03 1.125 0.227 0.4 1.332 0.8 -56.203 0.2 -20.179 0.9 -14.003 -0.

2 0.4 Story Drift Ratio (%) 0.eps Ground 0 0. 3.25 × El Centro ground motion.1 0.3 0.10.6 Fig.5 Ground 0 0. Heightwise variation of floor displacements and story drifts from MPA for 0. shading indicates modal combination error 25 .2 0.5 0.1 0.(a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th (b) Story Drift Ratios Floor 4th 3rd 2nd 1st Floor 5th RHA All Modes MPA 1 Mode 2 Modes 3 Modes 5th 4th 3rd 2nd 1st RHA All Modes MPA 1 Mode 2 Modes 3 Modes fig3_10a.3 0.4 Displacement/Height (%) 0.eps fig3_10b.

26 .

(4.1) is not valid for inelastic systems. but depend on the history of the displacements: f s = f s u. 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. Eq. and the unloading and reloading curves differ from the initial loading branch.1 Inelastic Multistory Buildings RESPONSE HISTORY ANALYSIS For each structural element of a building.1) becomes mu + cu + f s u. (3.4 4.2) to the modal coordinates of the corresponding linear system. the initial loading curve is idealized as bilinear. Therefore. the relations between lateral forces f s at the N floor levels and the lateral displacements u are not single valued.mι ug t b (4.2) The standard approach is to solve directly these coupled equations.3) . leading to the “exact” nonlinear response history analysis (RHA). Thus.1) With this generalization for inelastic systems. Although classical modal analysis (Section 3. Both systems have the same mass and damping. sign u b g g bg (4. 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. it is useful for later reference to transform Eq. the natural vibration periods and modes of the corresponding linear system are the same as the vibration properties of the inelastic system undergoing small oscillations (within the linear range). sign u = .

2 UNCOUPLED MODAL RESPONSE HISTORY ANALYSIS bg Neglecting the coupling of the N equations in modal coordinates [Eq.3). Unlike Eq. This approximate RHA procedure is the preliminary step in developing a modal pushover analysis procedure for inelastic systems. (4.9) involves T Fsn = Fsn qn . (4. (4. (3.3) in Eq. (4.4) represents N equations in the modal coordinates qn . sign u n b g b g bg (4. however. sign u = . (4. 2.and classical n damping-orthogonality property of modes gives F qn + 2z nw n qn + sn = . The spatial distribution s of the effective earthquake forces is expanded into the modal contributions sn according to Eq.” implying that other “modes” will also contribute to the solution. 4.5) This resisting force depends on all modal coordinates qn t . qr t = 0 for all 28 bg .4) where the only term that differs from Eq.9) for linearly elastic systems.Substituting Eq.n t given by Eq. where φ n are now the modes of the corresponding linear system.… N (4. For linear systems.sn ug t bg b g bg bg (4. (3. give the same results for u t as obtained directly from Eq.4) is rarely solved because it offers no particular advantage over Eq. (3. (4. (4. The equations governing the response of the inelastic system to p eff. Equation (4.2).2).8) because qr t will generally be nonzero for “modes” other than the nth “mode. However. Simultaneously solving these coupled equations and using Eq.6b) are mu + cu + f s u. premultiplying by φ T . implying coupling of modal coordinates because of yielding of the structure. (3.3) will. (3.4)] leads to the uncoupled modal response history analysis (UMRHA) procedure.Gn ug t Mn bg n = 1. Eq.6) for inelastic systems will no longer be described by Eq. these equations are coupled for inelastic systems. (4.6) The solution of Eq. in principle.2). sign qn = φ n f s u n . and using the mass.

(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

s n u g t .51 • ur2 (cm) 0 ur2 (cm) n=3 14.575 5 10 15 20 Time (sec) 25 30 ur3 (cm) 0 • 5. 4. ζn.112 5 10 15 20 Time (sec) 25 30 −10 0 −10 0 Fig. 2. 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) • 75.9 • 0 −50 10 −50 10 n=3 ur3 (cm) 0 • 5.n t = .Forces sn Unit mass An(t ) st rn ωn.3. Fsn / Ln ¨ g(t ) u (a) Static Analysis of Structure (b) Dynamic Analysis of Inelastic SDF System Fig.0 ¥ El Centro ground motion: (a) exact solution by NL-RHA. Roof displacement due to p eff.4. and (b) approximate solution by UMRHA bg af bg 32 . and 3.89 50 r1 0 u (cm) 0 • 78.02 −150 n=2 −150 50 n=2 14. where u g t = 3. 4. n = 1.

Dn relation is 33 bg bg . and (3) the Fsn Ln .33 0 −20 20 −20 20 n=2 ∆r2 (cm) 0 ∆r2 (cm) 7.n t (n = 1.4) and (4. and 3.0 ¥ El Centro ground motion: (a) exact solution by NL-RHA. 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. (4.1 Underlying Assumptions and Accuracy The approximate solution of Eq. 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. this intense excitation was chosen to ensure that the structure is excited well beyond its linear elastic limit.4 and 4. 2. respectively. 4.744 • 0 −20 10 n=3 −20 10 n=3 ∆r3 (cm) 0 • 5. (3. The errors are slightly larger in drift than in displacement.5.0 ¥ El Centro ground motion.2.5.n t = .5)] is neglected. where u g t = 3.s n u g t .6) by UMRHA is compared with the “exact” solution by nonlinear RHA. but even for this very intense excitation. Top story drift due to p eff.15) is strictly valid only for linearly elastic systems. (2) the superposition of responses to p eff.2… N ) according to Eq. both for 3. and (b) approximate solution by UMRHA af bg bg 4. n = 1.956 5 10 15 20 Time (sec) 25 30 ∆r3 (cm) 0 • 5.(a) Nonlinear RHA 20 n=1 20 (b) UMRHA n=1 • 5. 4. Such comparison for roof displacement and top-story drift is presented in Figs.008 • 6. (4.855 ∆r1 (cm) 0 ∆r1 (cm) n=2 • 6.

4.Dn relation? The two sets of forces and displacements are related as follows: V Fsn = bn Γn Dn = urn Γ nf rn 34 (4.approximated by a bilinear curve to facilitate solution of Eq. is plotted against roof displacement urn .9). The structure is pushed using the force distribution of Eq.8) governing Dn t is based on Eq. Therefore. Although most commercially available software cannot implement such displacement-controlled analysis.10) . the only force distribution that produces displacements proportional to f n is given by Eq.2 Properties of the nth-“mode” Inelastic SDF System How is the Fsn Ln . it can conduct a force-controlled nonlinear static analysis with an invariant distribution of lateral forces.9) should be determined by nonlinear static analysis of the structure as the structure undergoes displacements u = Dnφ n with increasing Dn .8) in UMRHA. At the yield point. (4. such nonlinear static analysis provides the so-called pushover curve. the relationship between lateral forces f s and Dn in Eq. (3. this distribution seems to be a rational choiceeven for inelastic systemsto determine Fsn in Eq. (4.20) to some bg pre-determined roof displacement.Dn curve. (4. Although several approximations are inherent in this UMRHA procedure. (4.Dn relation to be determined in Eq. How to convert this Vbn .8) before it can be solved? Because Eq.1.12) for floor displacements. The overall errors in the UMRHA procedure are documented in the examples presented in Section 4. (3.20). However. (4. within the linearly elastic range of the structure.2. A bilinear idealization of this pushover curve for the nth-“mode” is shown in Fig.6a. and the base shear Vbn . When implemented by commercially available software. 4.urn pushover curve to the Fsn Ln . 4. What is an appropriate invariant distribution of lateral forces to determine Fsn ? For an inelastic system no invariant distribution of forces can produce displacements proportional to f n at all displacements or force levels. the base shear is Vbny and roof displacement is urny . (3. when specialized for linearly elastic systems it is identical to the RHA procedure of Section 3. Therefore we impose this constraint in developing the UMRHA procedure in this section and modal pushover analysis in the next section. which is different than the Fsn Ln .

where the yield values of Fsn Ln and Dn are Fsny Ln = Vbny * Mn Dny = urny Γ nf rn (4. the elastic vibration period Tn of the nth-mode SDF system is computed from F Ln Dny I 1/ 2 Tn = 2p G H Fsny JK (4. 4. 4. (4.6. 4.11). should be used in Eq.11) * in which M n = Ln Γn is the effective modal mass (Chopra. the initial slope of the pushover curve in Fig. 2001.2.12) implying that the initial slope of the curve in Fig. The two are related through Fsny Ln 2 = w n Dny (4.5).8).10 enables conversion of the pushover curve to the desired Fsn Ln − Dn relation shown in Fig. 35 . (4. Knowing Fsny Ln and Dny from n Eq. which is not a meaningful quantity. Section 13.6b is w 2 .13) This value of Tn . 4. which may differ from the period of the corresponding linear system.6a is 2 kn = ω n Ln . Properties of the nth-“mode” inelastic SDF system from the pushover curve Equation 4.V bn (a) Idealized Pushover Curve F sn /L n (b) F sn / L − D Relationship n n Idealized Vbny Actual 1 αnkn V bny /M * n 1 α ω2 n n kn 1 1 ωn 2 u u rny rn Dny = ur n y / Γn φ r n D n Fig. In contrast.6b.

7. for linearly-elastic vibration of the building. Combine the “modal” responses using Eqs. of the total response r t obtained in Step 8. Calculate the peak value. details are available in Appendix A: 1. (3. is now determined by 36 bg bg .16) to determine the total response. and modes. ωn . 4. develop the base-shear – roof-displacement ( Vbn − urn ) pushover curve for the force distribution s* [Eq. n (3. Calculate histories of various responses by Eqs. 5.12) and (3. r o .4. with the structure is pushed to the roof displacement urno . Consider a nonlinear static analysis of the structure subjected to lateral forces distributed over the building height according to s* [Eq. the peak value of Dn t .21) where Dn . Repeat Steps 2 to 6 for as many modes as required for sufficient accuracy.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.3b) with force-deformation relation of Fig. 8. n 3. Typically.3 MODAL PUSHOVER ANALYSIS bg bg A pushover analysis procedure is presented next to estimate the peak response rno of the inelastic MDF system to effective earthquake forces p eff .6a). 4. which is summarized next as a sequence of steps. of the nth-“mode” inelastic SDF system (Fig.6b) by utilizing Eq.13). φn . 6. This value of the roof displacement is given by Eq.15) and (3. For the nth-mode. Compute the deformation history. (4. 4. 2.11). Idealize the pushover curve as a bilinear curve (Fig. (3. (3. 4. the first two or three modes will suffice.20)].2. (3. Compute the natural frequencies. and pseudo-acceleration history. 4.20)]. Dn (t ) . Convert the idealized pushover curve to the Fsn Ln − Dn relation (Fig. 9.6b.n t . An t . 4.

(4. of the nth-“mode” inelastic SDF system (Fig. as described in Section 4.g. The peak “modal” responses rno .12). as mentioned earlier.4. or from the inelastic response (or design) spectrum. 2001. (4. This application of modal combination rules to inelastic systems obviously lacks a theoretical basis. 6. bg bg bg bg Thus. it can be determined from the inelastic response (or design) spectrum (Chopra. Dn . are combined using an appropriate modal combination rule. although somewhat intuitive for inelastic buildings. seems reasonable.3b) with force-deformation relation of Fig. and (2) it provides the exact modal response for elastic systems.2.3. The response value rno is an estimate of the peak value of the response of the inelastic system to p eff. plastic hinge rotations. joint rotations. to obtain an estimate of the peak value ro of the total response. each determined by one pushover analysis.solving Eq.18).3.6). 4. story drifts. e.2.6 and 7. As shown in Sections 3. etc. It provides results for elastic buildings that are identical to the well-known RSA procedure (Section 3.3) because. the pushover analysis provides an estimate of the peak value rno of any response rn t : floor displacements.1 Summary The peak inelastic response of a building to earthquake excitation can be estimated by the MPA procedure just developed. it seems reasonable because it provides results for elastic buildings that are identical to the well-known RSA procedure described in Section 3. Sections 7. (3. However. details are available in Appendix B. 4. (3. (4. Eq.8).. 37 . 5. rno also represents the exact peak value of the nth-mode contribution rn t to response r t .n t . for elastic systems. Steps 1 to 4 of the MPA are same as those for UMRHA. alternatively.8). Compute the peak deformation.21). At this roof displacement.2 and 3. This pushover analysis. 4. which is summarized next as a sequence of steps. the lateral force distribution used possesses two properties: (1) it appears to be the most rational choice among all invariant distribution of forces. Calculate the peak roof displacement urno associated with the nth-“mode” inelastic SDF system from Eq. governed by Eq. we will refer to rno as the peak “modal” response even in the case of inelastic systems.6b by solving Eq.

say. From the total rotation of a plastic hinge.7 cm. also included are the combined responses due to one.18). the combined response due to three “modes”. respectively. the “exact” results. in contrast to modal analysis (Section 3.53 cm. determined by the UMRHA procedure. with. Figure 4. This is to be expected. rno . and three modes are compared with the “exact” values in Fig.2). Determine the total response by combining the peak “modal” responses using the SRSS combination rule of Eq. extract from the pushover database values of other desired responses. The peak values of displacements of all floors and drifts in all stories are presented in Tables 4.9. 9. This deficiency also implies that.1 and 4. subtract the yield value of hinge rotation to determine the hinge plastic rotation. Typically. (3.8. and compared with the results of a rigorous nonlinear RHA using the DRAIN-2DX computer program.7. The peak values of response are as noted. and three “modes”. the first two or three “modes” will suffice.0. 8.0 to 3. 4. the UMRHA procedure lacks a rigorous theory. 4. in particular. and ur 3o = 2. and the errors in the approximate results are shown in Fig. Repeat Steps 3 to 8 for as many “modes” as required for sufficient accuracy. The peak values of floor displacements and story drifts including one. the 38 .4. ur 2o = 11. Observe that errors tend to decrease as response contributions of more “modes” are included. two.2. the peak roof displacement due to each of the three “modes” is ur1o = 48.3 cm. and the percentage errors in the approximate results.5 ¥ the El Centro ground motion including the response contributions associated with three “modal” inelastic SDF systems. 4. two. and the “exact” response from nonlinear RHA for the roof displacement and top-story drift.1). 4.1 Uncoupled Modal Response History Analysis The structural response to 1.4 COMPARATIVE EVALUATION OF ANALYSIS PROCEDURES The response of the 9-story building described earlier is determined by the two approximate methods: UMRHA and MPA.7 shows the individual “modal” responses.4. 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. is presented next. although the trends are not systematic as when the system remained elastic (Section 3. At urno . three “modes” included.

24 • NL−RHA r −12 12 ∆r (cm) u (cm) 0 5 10 15 20 Time. sec 25 30 r −12 0 Fig. For each excitation.1 and 4.0.75. (a) Roof Displacement (b) Top Story Drift 80 12 0 • 3.85.7.response is much less accurate (Tables 4. recall that the computed errors have been presented earlier for ground motion multipliers 0. for a fixed number of “modes” included.25 (Tables 3.2).2) if the system yields significantly versus if the system remains within the elastic range (Tables 3. in particular. 1. the errors in story drifts are larger compared to floor displacements. sec 25 30 • 48. For this purpose the UMRHA and exact analyses were repeated for ground motions of varying intensity. However. the degree to which the system deforms beyond its elastic limit.88 u (cm) "Mode" 3 −80 80 0 −80 80 0 −80 0 • 44.5. 2. and total response from NL-RHA 39 . Response histories of roof displacement and top-story drift due to 1. the errors in responses computed by UMRHA including three “modes” relative to the “exact” response were determined.5.5 × El Centro ground motion: individual “modal” responses and combined response from UMRHA.53 ∆r3 (cm) 12 0 • 2. Next.1 −12 12 ∆r (cm) u (cm) UMRHA 3 "Modes" 0 • 7. we investigate how the errors in the UMRHA vary with the deformation demands imposed by the ground motion. and 3. just as for elastic systems.3 11.2) and 1.44 • "Mode" 2 r1 0 • 48.1 and 3. 4. 0.38 6.7 • −80 80 ∆r1 (cm) u (cm) "Mode" 1 −12 r2 0 ∆r2 (cm) 12 0 u (cm) −80 80 −12 r3 0 • 2.0.1 and 4.6 5 10 15 20 Time. defined as the El Centro ground motion multiplied by 0.25. 1.0.5 (Tables 4. 0.1 and 3.62 5.2). 0.

5 1 1. with the peak displacement of each “modal” SDF system noted for each ground motion multiplier. the pushover curves using force distributions s* [Eq. Two versions of the pushover curve are 40 . Shown is the error in each floor displacement (Fig.11. Heightwise variation of error in floor displacements and story drifts estimated by UMRHA including one.9. 4.10a).5 ¥ El Centro ground motion Figure 4.5 Displacement/Height (%) 2 Ground 0 0.10 summarizes the error in UMRHA as a function of ground motion intensity.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. or three “modes” for 1. two. 4. Heightwise variation of floor displacements and story drift ratios from UMRHA and NL-RHA for 1.10b). (3. it will be useful to know the deformation of the system relative to its yield deformation. For this purpose. 4. indicated by a ground motion multiplier.5 Fig. in each story drift (Fig. 4.5 1 1.20)] for the first three n modes of the system are shown in Fig. To interpret these results.5 Story Drift Ratio (%) 2 2.(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. and the error envelope for each case.

and 3 (Fig. The idealized curve for the first “mode” deviates most from the actual curve near the peak displacement corresponding to ground motion multiplier 1. and. For more intense excitations.10) for each “modal” inelastic SDF system (Figs 4. 4. respectively. (2) increase rapidly as the ground motion multiplier increases to 1. Additional errors are introduced in UMRHA of systems responding beyond the linearly elastic limit for at least two reasons. (3) maintain roughly similar values for more intense ground motions.4.4).7). Perhaps this explains why the errors are large at this excitation intensity. the pushover curve for each “mode” is idealized by a bilinear curve in solving Eq. 2. The structure is pushed using the force distribution of Eq. as mentioned in Section 3. First. the errors in truncating the higher mode contributions are negligible.11). the first reason mentioned above seems to be the primary source for the errors. and (4) are larger in story drift compared to floor displacements.5).7 cm.5 ¥ the El Centro ground motion.2. The system remains elastic up to ground motion multiplier 0.1).53 cm. (4. 41 .0.2 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 1.4) to roof displacements urno = 48. (3. and plastic hinge rotations at the external beam end at each floor level (Table 4. the peak values of displacements at all floors (Table 4. were noted but are not shown here.75.01 (Fig.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. 11.11a). and 2. UMRHA lacks a rigorous theory and is based on several approximations. Each of these three pushover analyses provides the pushover curve (Fig. drifts in all stories (Table 4. 4.3 cm. 4. the values determined by RHA of the nth-mode inelastic SDF system (Fig. Second. 3. as mentioned in Section 4. determined from “exact” analyses. even though the system remains essentially elastic. the ductility factor for the first mode system is only 1.6 and 4.0. 4.3). Figure 4.20) with n = 1.included: the actual curve and its idealized bilinear version.4. The location of plastic hinges and their rotations.75.2.

484 0.5 3 0 0 0.5 10.0 11.4 1.473 -22.942 -0.235 -0.938 -1.044 1.201 1.256 1.1 Peak values of floor displacements (as % of building height = 37.982 9.010 0.8 0.8 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 -1.2 12.526 -0.9 16.241 1.4 4.410 -1.1 1.136 1.202 11.14 m) from UMRHA for 1.2 1.065 0.820 -19.049 -0.472 1.540 0.133 1.3 42 .138 1.055 -0.376 -1.9 31.5 28.201 -0.9 5.727 1.820 -0.772 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 1.663 0.971 1.200 8.2 Peak values of story drift ratios (as % of story height) from UMRHA for 1.003 -31.220 -0.914 2.379 1.298 0.104 0.5 9.1 3.201 -1.407 -10.5 3 Fig.852 1.806 0.938 1.8 1.914 -0.371 -0.942 1.4 -7. 4.373 -0.6 1.3 8.707 1.371 0.554 1.8 1.763 -15.033 0.410 1.372 1.844 -25.248 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.003 0.819 2.983 1.018 0.856 2.806 -0.333 0.317 0.5 1 1.863 0.241 -1.8 1.256 -1.490 1.169 0.2 22.126 0.2 4.256 1.068 0.945 -37.668 -23.298 -0.478 0.214 -0.072 -1.8 14.338 -1.366 0.5 28.5 × 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 1.6 2.009 -0.7 14.079 0.315 -0.722 0.2 6. and (b) story drifts Table 4.4 -1.7 31.3 25.260 -15.096 0.751 1.698 1.130 0.877 0.135 9.088 10.663 -0.575 -41.070 1.120 1.154 0.513 0.1 8.293 1.0 2.1 0.376 1.3 6.5 2 GM Multiplier 2. 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 0.676 0.2 1.071 -0.5 18.5 1 1.338 1.5 9.426 -1.350 -0.057 -0.9 12.372 -1.370 -0.(a) Floor Displacements 60 50 40 Error Envelope Error for Floor No. Errors in UMRHA as a function of ground motion intensity: (a) floor displacements.072 1.220 0.226 -0.187 -0.0 1.616 -0.811 1.7 Table 4.0 -9.3 0.430 1.0 11.5 2 GM Multiplier 2.513 -0.8 1.283 1.0 7.291 0. Noted 60 50 40 (b) Story Drifts Error Envelope Error for Story No.9 12.10.900 -10.071 0.863 1.5 -3.216 1.495 1.121 -0.9 31.0 9.366 -0.490 -1.6 4.1 1.214 0.

considering one. The second “mode” was necessary to identify hinges in the upper stories. significantly smaller than in UMRHA (compare Fig. two. 4.13 with Fig. As shown in Figs.11c).3 and 4.3 through 4. and the errors in these estimates relative to the exact response from nonlinear RHA. the results were not always accurate.12 and Tables 4. overestimated by up to 18% in the middle stories. and three “modes. For example.4). The locations of plastic hinges shown in Fig. 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%.5 present estimates of the combined response according to Eq. Pushover analysis seems to be inherently limited in computing accurately hinge plastic rotations. 4.4 with Tables 4. One “mode” pushover analysis was unable to identify the plastic hinges in the upper stories where higher mode contributions to response are known to be more significant. The first “mode” alone is inadequate.18). especially in estimating the story drifts (Fig.3 and 4. The drifts are underestimated by up to 13% in the lower stories.3 and 4. 4. (3.” two “modes.Figures 4. in general. 43 .2). Furthermore. Fortuitously.9 and Tables 4.4. the errors in the modal pushover results are. were determined by four analyses: MPA considering one “mode. 4.12 and Tables 4. although the error is recorded as 100% if MPA estimates zero rotation when the nonlinear RHA computes a non-zero value. the third “mode” does not contribute because this SDF system remains elastic (Fig.13 and Tables 4. the hinges identified in beams at the 6th floor were at variance with the “exact” results. MPA failed to identify the plastic hinges at the column bases in Fig.13 and Tables 4. but the third “mode” contributions do not seem especially important (Fig.3 and 4.” and nonlinear RHA. 4. Obviously.4). 4. 4. significant improvement is achieved by including response contributions due to the second “mode”. The errors are especially large in the hinge plastic rotations estimated by the MPA procedure.” respectively. for two or three modes included.12 and 4.” and three “modes. even if three “modes” are included (Fig. and are within a few percent of the exact values for the upper stories. However.14. the additional errors due to the approximation inherent in modal combination rules tend to cancel out the errors due to the various approximation contained in the UMRHA. however.14. 4. this error is not especially significant because the hinge plastic rotation is very small.5). but was more successful when the excitation was more intense (results are not included). 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.1 and 4.13 and Table 4.

2.9 cm.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36. α = 0. V = 4952 kN. 0. 1.0 × El Centro ground motion 44 .5.5. α = 0.25 1 2 4 6 Roof Displacement (cm) 8 10 Fig. 1.14 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 = 4.13 y by Base Shear (kN) 8000 6000 1 1.0.19 y by 3 2 1.25.5 2 3 4000 2000 0 0 Actual Idealized 0.5 0.5 0. “Modal” pushover curves with peak roof displacements identified for 0.5 2 3 Actual Idealized 2000 0 0 0.0.75. 0.11.5 0.85 0. V = 7616 kN.2 cm.75 0. 4.6 cm. V = 5210 kN.25 0.75 0.5 1 0.75 Actual Idealized 20 40 60 Roof Displacement (cm) (b) "Mode" 2 Pushover Curve 80 12000 10000 u = 9. α = 0.85 0.85 Base Shear (kN) 8000 6000 4000 2000 0 0 0. and 3.

However. an observation with practical implications. as discussed in Sections 3.3 for elastic systems (or weak ground motions).5 Story Drift Ratio (%) 2 2.12 Heightwise variation of floor displacements and story drift ratios from MPA and NL-RHA for 1.2 and 3. now standard in engineering practice.15 summarizes the error in MPA considering three “modes” as a function of ground motion intensity. While various sources of errors in UMRHA. indicated by a ground motion multiplier. each story drift (Fig. UMRHA is essentially exact. shading indicates errors in MPA including three “modes” Figure 4.15).75.5 1 1.5 1 1.15a). implying that the modal combination errors contained in these procedures are acceptable.5 Fig.4.(a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th NL−RHA MPA 1 "Mode" 2 "Modes" 3 "Modes" (b) Story Drift Ratios Floor Floor 5th 4th 3rd 2nd 1st 5th 4th 3rd 2nd 1st NL−RHA MPA 1 "Mode" 2 "Modes" 3 "Modes" Ground 0 0. 4. the errors in MPA were fortuitously smaller than in UMRHA (compare Figs.0. identified in Section 3.3. In this case. implying excitations intense enough to cause significant yielding of the structure. 4. 45 . Shown is the error in each floor displacement (Fig. implying excitations weak enough to limit the response in the elastic range of the structure.5 × El Centro ground motion.4. 4. 4. The fact that MPA is able to estimate the response of buildings responding well into the inelastic range to a similar degree of accuracy indicates that this procedure is accurate enough for practical application in building retrofit and design.4 also apply to MPA.15b). whereas MPA contains errors inherent in modal combination rules. 4. the errors in MPA were larger for ground motion multipliers less than 0. As mentioned in Section 3.15) for ground multipliers larger than 1. The errors are only weakly dependent on ground motion intensity (Fig. the MPA procedure is equivalent to the RSA procedure.5 Displacement/Height (%) 2 Ground 0 0. and the error envelope for each case.10 and 4.

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

498 1.3 -3.0 -100.266 -0.304 1.8 -6.00E+00 0.00E+00 0.018 -0.304 -1.053 -1.00E+00 0.6 -9.00E+00 0.250 0.611 0.88E-03 1.00E+00 0.752 1.666 Table 4.311 0.003 -16.980 -0.6 -44.00E+00 0.9 2.705 -1.116 1.2 0.02E-03 0.2 -100.00E+00 0.640 -1.4 1.351 -0.72E-03 7.00E+00 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th “Mode” 2 0.5 7.9 7.6 7.156 -0.7 -12.435 0.683 1.60E-04 7.3 Peak values of floor displacements (as % of building height = 37.197 -0.00E+00 1 “Mode” 7.336 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.517 1.5 2.5 × El Centro ground motion Displacement /Height (%) “Modal” Response Combined (MPA) “Mode” 1 “Mode” 2 “Mode” 3 1 “Mode” 2 “Modes” Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 0.8 -29.015 0.407 -27.00E+00 3.00E+00 0.9 1.5 × El Centro ground motion Hinge Plastic Rotation (rad) “Modal” Response Combined (MPA) “Mode” 1 7.733 1.8 -4.50E-10 3.9 5.745 1.9 -100.6 1.72E-03 7.756 0.503 -1.55E-03 3.911 0.00E+00 0.1 1.018 -0.331 1.233 1.1 62.5 10.705 1.5 -6.8 1.900 -0.518 1.5 Peak values of hinge plastic rotations (radians) from MPA for 1.00E+00 0.220 1.015 0.895 1.00E+00 3 “Modes” 7.2 -4.3 -3.5 1.1 46.9 0.879 1.19E-10 3.222 0.168 -0.910 1.8 “Mode” 3 -1.1 62.3 11.116 1.298 0.259 1.009 0.781 0.5 Table 4.9 -100.478 0.233 1.0 -100.895 1.0 -50.503 1.209 1.1 46.9 -4.1 18.8 -6.00E+00 2 “Modes” 7.945 -49.652 1.116 1.Table 4.066 -0.426 15.76E-03 4.55E-03 3.527 -0.781 0.60E-03 2.6 -8.2 -100.5 2.36E-03 6.033 -0.372 0.00E+00 0.36E-03 6.667 1.130 0.298 -0.3 13.5 2.8 -12.55E-03 3.2 0.222 0.0 1.756 0.641 1.26E-04 9.0 3 “Modes” -32.6 13.071 0.737 0.00E+00 0.4 -8.00E+00 “Mode” 3 0.055 0.36E-03 6.068 0.02E-03 3.00E+00 0.982 13.516 0.687 0.05E-03 2.76E-03 4.37E-03 1.00E+00 0.414 1.190 -0.315 0.6 13.018 0.88E-03 0.668 -13.1 13.72E-03 7.37E-04 Error (%) “Modal” Response 1 “Mode” -32.399 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” 1.057 0.2 1.00E+00 0.37E-03 1.581 0.844 -7.202 8.260 -14.473 -15.00E+00 0.6 -44.125 -1.18E-03 7.118 0.37E-03 1.244 0.575 -53.804 1.0 47 .7 1.371 -0.007 1.429 0.9 1.2 11.8 7.053 1.8 17.37E-03 1.8 1.7 1.4 Peak values of story drift ratios (as % of story height) from MPA for 1.00E+00 0.2 6.6 -7.76E-03 4.763 -14.007 1.101 -0.2 -3.1 46.088 12.154 0.0 2 “Modes” -32.338 1.135 -7.305 -0.0 -100.980 0.640 1.694 1.2 -100.105 0.1 -8.066 -0.76E-03 4.02E-03 0.00E+00 0.012 1.99E-03 6.72E-03 7.0 -5.667 -1.399 0.14 m) from MPA for 1.594 -1.220 1.049 -0.88E-03 0.942 6.0 15.8 -29.200 8.00E+00 0.728 1.10E-02 9.36E-03 6.02E-03 3.738 1.820 -7.176 0.5 10.1 62.8 -29.0 -50.53E-03 7.737 1.22E-10 NL RHA 1.581 0.8 0.514 -1.614 0.3 1.071 -0.

(3.5). Static analysis of the structure for gravity loads provides the initial state—forces and deformations—of the structure.4.7).17 and 4. the third “mode” contributions do not seem especially important (Fig. 4. Results of such analyses for the selected building subject to 1.7).17 and Tables 4. The errors are especially large in the plastic hinge rotations estimated by the MPA procedure even if three “modes” are included (Fig. and plastic hinge rotations at selected external beam end at each floor level (Table 4. 4. The pushover curves for the second and third modes included in Fig. the results presented so far did not include gravity load effects.18 and Tables 4.11a).8 present estimates of the combined response according to Eq. resulting in the pushover curve shown in Fig.0 cm.3 Modal Pushover Analysis with Gravity Loads To evaluate the accuracy of the dynamic response of the system. two. drifts in all stories (Table 4.6).18) considering one. They are now included in the pushover analysis of the structure for the first “mode” only.6 and 4. 4.6 through 4.18 and 48 .6 and 4. MPA including two “modes” underestimates the displacements of lower floors by up to 6% and overestimates the upper floor displacements by up to 22%. Figures 4.18 and Tables 4.20) with n = 1 (Fig.4. (3.8).7 cm and ur 3o = 2.6 and 4.4. 4. and three “modes.20) with n = 1 to a target roof displacement. Significant improvement is achieved by including response contributions due to the second “mode. The first “mode” alone provides adequate estimates of floor displacements. and overestimated by up to 29% in the middle stories and by up to 13% in the upper stories. 4.7. (3. 4.16 are unchanged. 4.7).” but not the contributions of higher “modes.53 cm.5 × El Centro ground motion are presented below.” these effects will modify the combined modal response as well as the results of nonlinear RHA.3 . The story drifts are underestimated by up to 7% in the lower stories. The response contributions of the second and third modes are the same as before (Tables 4.16a—which is slightly different than the one excluding gravity loads (Fig. the structure is pushed using the force distribution of Eq. Obviously gravity load effects will influence the seismic demands due to the first “mode.17 and Tables 4. Starting with its initial state under gravity loads. Also presented at these roof displacements are the displacements at all floors (Table 4.” however. As shown in Fig. but it is inadequate in estimating the story drifts (Fig.” and the errors in these estimates relative to the exact response from nonlinear RHA.4) to roof displacement ur1o = 52. as are the roof displacements ur 2o = 11. 3. and the structure is pushed using the force distribution of Eq.

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. Most pushover analysis procedures do not seem to compute to acceptable accuracy plastic hinge rotations. 4.5 3 Fig.Table 4. and (b) story drifts 49 . two.14. 1−"Mode" • • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • •• • •• • • • • • •• •• • •• •• •• •• (b) MPA.5 1 1.5 1 1. (a) MPA. Errors in MPA as a function of ground motion intensity: (a) floor displacements.5 2 GM Multiplier 2.5 3 0. Noted 60 50 40 (b) Story Drifts Error Envelope Error for Story No.15. Locations of plastic hinges determined by MPA considering one. and three “modes” and by NL-RHA for 1.8).5 × El Centro ground motion (a) Floor Displacements 60 50 40 Error Envelope Error for Floor No.5 2 GM Multiplier 2. 4.

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

0.75. V = 5210 kN.85 0.75 0. V = 4952 kN.85 0.0.14 y by Base Shear (kN) 8000 6000 4000 1.75 Actual Idealized 20 40 60 Roof Displacement (cm) 80 (b) "Mode" 2 Pushover Curve 12000 10000 u = 9.5 0.5 2 3 4000 2000 0 0 Actual Idealized 0.5 2 3 Actual Idealized 2000 0 0 0. 1.25 0. 4.50.25.19 y by 3 2 1. α = 0.25 5 10 15 20 Roof Displacement (cm) 25 (c) "Mode" 3 Pushover Curve 12000 10000 u = 4. and 3. 0.5 0.85 Base Shear (kN) 8000 6000 4000 2000 0 0 0.85. α = 0. noted are peak values of roof displacement for 0.13 y by Base Shear (kN) 8000 6000 1 1.16. 2.5 0.5 1 0.25 1 2 4 6 Roof Displacement (cm) 8 10 Fig.6 cm. “Modal” pushover curves with gravity loads included.3 cm. 0.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36.9 cm. α = 0.75 0. V = 7433 kN.0.0 × El Centro ground motion 51 .

17.(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. and hinge plastic rotations estimated by MPA including one. and 52 . Errors in floor displacements.5 1 1.5 Story Drift Ratio (%) 2 2. gravity loads included.5 Fig. 4. two.5 Displacement/Height (%) 2 Ground 0 0. Heightwise variation of floor displacements and story drift ratios from MPA and NL-RHA for 1.5 1 1.5 × El Centro ground motion. shading indicates errors in MPA including three “modes” (a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th (b) Story Drift Ratios Floor 4th 3rd 2nd 1st MPA 1 "Mode" 2 "Modes" 3 "Modes" Floor 5th 5th 4th 3rd 2nd 1st MPA 1 "Mode" 2 "Modes" 3 "Modes" Ground −60 −40 −20 0 Error (%) 9th 8th 7th 6th 20 40 60 Ground −60 −40 −20 0 Error (%) 20 40 60 (c) Hinge Plastic Rotations Floor 5th 4th 3rd 2nd 1st MPA 1 "Mode" 2 "Modes" 3 "Modes" Ground −120 −80 −40 0 Error (%) 40 80 120 Fig. 4. story drifts.18.

5 1 1.20. gravity loads included 53 . 4. and three “modes” and by NL-RHA for 1. 3−"Modes" (d) Nonlinear RHA • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• •• •• • • • • • • • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • •• •• • •• •• •• •• • • • • • • •• •• •• Fig.5 2 GM Multiplier 2. Locations of plastic hinges determined by MPA considering one.5 2 GM Multiplier 2.5 × El Centro ground motion.5 3 Fig.(a) MPA.19. 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. two. gravity loads included (a) Floor Displacements 60 50 40 Error Envelope Error for Floor No. and (b) story drifts.5 3 0. Errors in MPA as a function of ground motion intensity: (a) floor displacements. 2−"Modes" •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • • • •• •• •• (c) MPA. Noted 60 50 40 (a) Story Drifts Error Envelope Error for Story No. 1−"Mode" • • • • • • • •• •• •• •• •• •• •• •• •• •• •• •• •• •• •• • • •• •• •• • • • • • •• •• •• •• •• •• •• (b) MPA. 4.5 1 1.

105 0.5 -5.830 -12.429 1.7 -2.434 0.00E+00 Error (%) “Modal” Response 1 “Mode” -32.4 26.197 -0.00E+00 2 “Mode” 8.5 3 “Modes” -32.2 12.821 1.9 -4.13E-03 5.5 × El Centro ground motion.35E-03 8.0 16.263 1.813 1.6 2.8 0.2 1.263 0.7 -2.00E+00 0.35E-03 8.156 -0.319 1.8 28.908 -1.35E-03 8.00E+00 0.00E+00 0.0 37.214 0.4 -4. 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” 0.109 0.514 -1.1 13.130 0.5 21.75E-03 0.3 -100.00E+00 0.330 1.315 0.1 21.2 1.4 1.11E-03 9.0 -30.4 1.00E+00 0.478 0.19E-03 1.8 -32.00E-10 NL RHA 1.2 -0.5 “Mode” 3 -1.594 -1.00E-03 5.311 0.850 -1.461 0.04E-02 8.2 -3.207 18.04E-10 3.00E+00 Combined (MPA) 1 “Mode” 8.033 -0.266 -0.5 0.213 1.4 -6.0 2 “Modes” -32.783 1.237 0.237 0.305 -0.11E-03 9.850 1.5 × El Centro ground motion.7 1.00E+00 0.5 × El Centro ground motion.530 1.3 13.4 20.00E+00 “Mode” 3 0.154 0.78E-03 1.821 -1.23E-03 0.213 1. gravity loads included Hinge Plastic Rotation (rad) Floor “Modal” Response “Mode” 1 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 8.399 -0.2 4.2 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” 1.637 0.998 0.114 -1.3 1.88E-03 0.55E-03 3.858 2.0 37.2 9.998 1.2 -2.0 37.998 21.399 0.429 -1.23E-03 3.983 1.933 1.00E+00 0.101 -0.2 9.2 4.3 9.8 1.14 m) from MPA for 1.8 -32.2 4.687 0.098 20.5 54 .434 Error (%) NL 3 1 2 3 RHA “Mode” “Modes” “Modes” “Modes” 0.6 0.068 0.88E-03 1.2 16.88E-03 0.831 0.530 1.927 1.454 1.00E+00 “Mode” 2 0.490 -11.8 Table 4.6 19.11E-03 9.414 28.9 31.913 7.00E+00 0.6 Peak values of floor displacements (as % of building height = 37.351 -0.071 -0.836 -0.00E+00 0.310 1.983 1.114 1.2 9.0 -30.066 -0.00E-03 5.0 -100.102 1.00E+00 0.603 1.3 -22.270 -12.831 0.5 21.037 -0.371 -0.3 -22.00E+00 0.9 -6.686 -7.6 1.637 0.00E-03 5.466 0.066 -0.744 1.434 0.822 1.667 0.665 0.5 1.1 1.372 0.19E-03 1.1 4.055 0.7 16.35E-10 3.17E-03 9.00E+00 0.071 0.5 10.009 0.8 0.00E+00 3 “Modes” 8.754 1.4 -4.6 2.888 1.7 Peak values of story drift ratios (as % of story height) from MPA for 1.19E-03 1.128 -1.00E+00 0.049 -0.176 0.26E-03 3.00E+00 3.55E-03 3.507 1.854 0.55E-03 3.938 1.330 1.516 0.19E-04 5.9 -3.037 0.527 -0.190 -0.921 1.673 Table 4.23E-03 0.064 -10.3 -100.35E-03 8.23E-03 3.7 19.102 1.854 0.00E+00 0.00E+00 0.3 -100.213 1.00E+00 0.23E-02 1.00E+00 0.530 1.603 -1.3 -22.8 Peak values of hinge plastic rotations (radians) from MPA for 1.2 2.436 1.125 -1.199 16.107 1.00E+00 0.018 -0.5 1.7 4.319 1.015 0.11E-03 9.168 -0.00E-03 5.754 0.057 0.877 -46.353 -23.257 0.996 0.953 15.Table 4.2 21.0 -100.515 -50.996 -0.19E-03 1.908 1.728 1.8 9.860 1.00E+00 0.

Specified in FEMA-273 are three distributions for lateral forces: 1. joint rotations.2 COMPARATIVE EVALUATION Compared in this section are the earthquake-induced demands for the selected building determined by five analyses: pushover analysis using the three force distributions in FEMA-273.5 sec . gravity load effects were included in all 55 . j 2. computed at the target displacement represent the earthquakeinduced demands on the structure. and 3.5 sec .1 FEMA-273 PUSHOVER ANALYSIS In this investigation we focus on one step in the nonlinear static procedure in the FEMA-273 document [Building Seismic Safety Council. k = 2 for T1 ≥ 2. 2… N ). etc. 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. a plot of base shear versus roof displacement.. and varies linearly in between.5. Comparison of Modal and FEMA Pushover Analyses 5. SRSS distribution: s* is defined by the lateral forces back-calculated from the story shears determined by response spectrum analysis of the structure. story drifts. and the exponent k = 1 for fundament period T1 £ 0. plastic hinge rotations. 1997] The pushover curve. 5. The floor displacements. MPA considering three “modes.” and nonlinear RHA. 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. “Uniform” distribution: s* = m j (where the floor number j = 1.

3. (b) ELF.0197 0.4.5 times the El Centro ground motion.11 0. The “uniform” distribution overestimates all floor displacements by 17-28%.367 (a) Uniform (b) ELF (c) SRSS Fig.3b and Table 5. The first four floor displacements are within 5% of the “exact” value. The three FEMA force distributions are presented in Fig. 0. The errors in the FEMA and MPA estimates of seismic demands relative to the “exact” demands are presented in Fig.3a and 5.11 0. the value determined from RHA of the first-mode inelastic SDF system for 1.3.062 0. 5. Also included in these presentations are the MPA results considering three “modes” and the “exact” demands from nonlinear RHA.2. 5.0 cm. and the displacements of upper floors are overestimated by 9-22%. and the locations of all plastic hinges in Fig.126 0.0381 0. both presented in Section 4.11 0.analyses.3a and Table 5.0896 0.4.281 0.11 0. with the errors being larger for the SRSS distribution. 5. The pushover curves are given in Fig.0981 0.1.4 and Tables 5. the floor displacement demands in Fig.0446 0.0913 0. pushover analyses are implemented for a target roof displacement of 52.0702 0. and Table 5.2.3.1. 5.1 demonstrate that the displacement demands are underestimated by the ELF and SRSS force distributions by 12 to 30 % at the lower six floors of the building.00719 0.11 0.11 0. 5.0654 0. Using each of these force distributions.1 through 5.165 0. Force distributions in FEMA-273: (a) “uniform”.1. wherein the first two are obvious and the third is determined from response spectrum analysis of the building (Appendix C). 5.11 0. plastic hinge rotation demands in Table 5. the story drift demands in Fig.21 0.112 0.119 0.177 0. Figures 5. The MPA procedure is more accurate than all the FEMA force distributions.042 0. and (c) SRSS 56 .0466 0.4a. 5.

14 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized 10 20 30 40 Roof Displacement (cm) (b) ELF Distribution 50 60 12000 10000 u = 38. α = 0. V = 8530 kN.(a) Uniform Distribution 12000 10000 u = 33 cm.24 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized 10 20 30 40 Roof Displacement (cm) 50 60 Fig. α = 0.3 cm. gravity loads are included 57 .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 = 39.2. 5. (b) ELF.8 cm. Pushover curves using three force distributions in FEMA-273: (a) “uniform”. and (c) SRSS. V = 6897 kN. α = 0. V = 7456 kN.

with story drifts under estimated by. 3 "Modes" FEMA Uniform ELF SRSS 0. reaching 31%. the errors are largest in the lower stories.3 demonstrate that the hinge plastic rotations estimated by all three FEMA force distributions contain unacceptably large errors.5 Displacement/Height (%) 2 Floor 5th 5th 4th 3rd 2nd 1st NL−RHA MPA. For the SRSS distribution.5 1 1. For the ELF distribution.) The pushover analysis procedures considered seem incapable of computing accurately local response quantities. gravity loads included 58 . In contrast.5 Ground 0 Ground 0 Fig.” and NLRHA. Figure 5. reaching 64%.2 demonstrate that the story drift demands are greatly underestimated by all the FEMA force distributions.Figures 5. at most. errors are largest in the upper stories. and overestimated by no more than 32%.3b and 5. the MPA procedure is more accurate than all the FEMA force distributions. whereas nonlinear RHA computed an insignificantly small value. MPA including three “modes. such as hinge plastic rotations.4c and Table 5.5 Story Drift Ratio (%) 2 2. 7%.4b. This seems to be an inherent limitation of pushover analysis.3. 5. (The 100% error in a 6th floor hinge is ignored as it simply represents that the MPA estimated zero rotation. Modal pushover analysis procedure gives estimates better than all the FEMA force distributions. reaching 35%. but it is still inaccurate.5 1 1. larger errors are noted in the upper and lower stories. 3 "Modes" FEMA Uniform ELF SRSS 0. Heightwise variation of floor displacements and story drift ratios estimated using FEMA-273 force distributions. (a) Floor Displacements 9th 8th 7th 6th 9th 8th 7th 6th (b) Story Drift Ratios Floor 4th 3rd 2nd 1st NL−RHA MPA. and Table 5. with errors reaching 37% in this example. For the uniform distribution.

the SRSS distribution fails to identify yielding of beams in the middle floors.4. The locations of the plastic hinges are not identified correctly by the FEMA force distributions. and the ELF distribution fails 59 . story drifts. 5.5 were determined by five analyses: MPA considering the three “modes. and hinge plastic rotations estimated using FEMA-273 force distributions and MPA (including three “modes”). and the three FEMA analyses. they are not good indicators of damage. While pushover estimates for floor displacements are more accurate. 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.” nonlinear RHA (“exact”).(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. Errors in floor displacements. the “uniform” distribution fails to identify yielding of the beams above the fourth floor. Based on the results presented here. 5. The locations of plastic hinges shown in Fig.

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

341 1.064 1.789 0.0 -30.00E-03 5.839 0. gravity loads included Displacement /Height (%) FEMA ELF SRSS MPA 1.0 -63.154 1. gravity loads included Displacement /Height (%) FEMA ELF SRSS MPA 0.708 0.6 -6.04E-02 8.011 1.530 1.234 1.8 -100.6 21.098 1.490 0.566 1.611 0.270 0.399 -27.560 1.436 Error (%) NL Uniform RHA 0.938 1.4 MPA -2.2 6.8 -35.9 -70.335 2.6 -25.736 0.45E-03 3.195 0.23E-03 3.487 0.372 1.462 1.367 1.00E+00 0.2 5.19E-03 1.7 MPA -2.913 0.5 -29.318 2.13E-03 5.355 0.562 1.2 9.128 1.323 1.007 1.94E-03 2.0 -71.1 10.178 1.2 13.6 17.953 0.353 1.344 0.88E-03 1.7 FEMA ELF SRSS -22.8 -2.083 1.015 0.221 1.310 1.35E-10 3.9 11.4 21.2 9.11E-03 9.809 0.00E+00 4.4 0.515 FEMA Uniform ELF 27.1.78E-03 0.8 44.417 1.52E-04 1.26E-03 3.6 -4.314 1.998 1.875 0.4 -50.686 0.294 1.0 37.672 1.5 26.1 -26.0 -100.5 -32.00E+00 0.730 1.55E-03 3.2 16.351 0.109 1.262 1.209 0.3 -100.6 14.9 16.673 Error (%) NL RHA 1.992 1.00E-10 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 1.2 10.667 0.00E+00 0.50E-03 0.8 -63.17E-03 9.53E-02 1.1 0.6 -11.998 1. Peak values of floor displacements (as % of building height = 37.5 -33.8 -32.877 1.6 -73.7 -60.3 -11.2 19.399 1.7 7.1 163.724 0.783 1.7 -28.5 15.09E-03 4.3 23.4 -23.3 -22.4 -55.10E-02 7.58E-04 6.061 1.35E-03 8.263 0.3 -14.00E+00 Error (%) FEMA Uniform ELF SRSS MPA 24.2 -100.5 Story 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 2.466 0.00E+00 0.9 SRSS -22.836 0.547 -27.623 1.5 -29.214 1.59E-03 5.3 29.199 27.3 31.00E+00 0.858 1.51E-03 4.6 -41.9 -77.65E-03 7.330 1.03E-03 5.4 -12.1 -3.0 -57.4 16.168 1.5 10.9 0.5 22.860 1.34E-03 2.84 1.23E-02 1.4 16.7 26.414 1.8 16.9 16.8 17.75E-03 0.3 Peak values of hinge plastic rotations (radians) from FEMA force distributions and MPA Hinge Plastic Rotation (rad) MPA FEMA ELF SRSS 4.524 0.5 -3.0 -59.8 7.2 -4.530 1.2 -12.7 21.14 m) from FEMA force distributions and MPA.7 Table 5.888 0.93E-03 1.975 1.5 -27.5 61 .0 -100.306 1.9 28.8 Floor 1st 2nd 3rd 4th 5th 6th 7th 8th 9th Uniform 0.399 1.2 Peak values of story drift ratios (as % of story height) from FEMA force distributions and MPA.3 -4.0 Table 5.0 -100.8 -100.6 22.7 -15.2 16.8 2.Table 5.16E-03 0.0 -100.7 -29.00E+00 NL RHA 1.597 0.5 -33.9 -20.089 1.62E-03 0.00E+00 2.00E+00 8.6 -17.6 4.00E+00 0.7 19.78E-03 1.984 1.854 0.207 1.19E-04 5.9 15.067 0.830 2.927 1.2 4.

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

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

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

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

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

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

68 . While pushover estimates for floor displacements are even more accurate. they are not good indicators of damage. such as life safety and collapse prevention.11. 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. Instead. This new method for estimating seismic demands at low performance levels. 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. 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.

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

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

20): s* = mφn n n 2. Since the target roof displacement may not be known at the start of the procedure. Let the roof displacement and base shear at the anchor point be urno and Vbno . DRAIN2DX (Allahabadi and Powell. (3. e. Compute natural frequencies. B.2.1 STEP-BY-STEP PROCEDURE A detailed step-by-step implementation of the uncoupled modal response history analysis (UMRHA) procedure is presented in this section and illustrated by an example in the next section. using any numerical integration method. 3. Define the anchor point. of the bilinear curve at the target roof displacement. A.1. 1..1) using the FEMA-273 procedure (Building Seismic Safety Council. Apply force distribution of Step 2.. The structure should be pushed just beyond the target (or expected) roof displacement in the selected mode. 1988). develop the base-shear – roof-displacement (Vbn . e. 1977). Apn . 3. Idealize the pushover curve as a bilinear curve (Fig. 3.Appendix A Uncoupled Modal Response History Analysis A. φn . ω n . iterations may be necessary.urn ) pushover curve for the force distribution s* : n 2.2.g. and modes.g. 71 . This step can be conveniently implemented in any commercially available software. Calculate the area under the actual pushover curve. respectively. for linear-elastic vibration of the building. trapezoidal rule. 2. Define the force distribution s* from Eq. For the nth-“mode”.1 incrementally and record the base shears and associated roof displacements.1.

will be refined by an iterative procedure that seeks to equate areas under the actual and the idealized pushover curves.i 3.3. From the pushover data. corresponding to the estimated yield i i i base shear. O.8.2. Draw the curve OAB by connecting the three points O. Let the point with base shear = Vbny and roof displacement = urny be denoted as A. 72 .6 . other appropriate methods can be ( used.7. Calculate the post-yielding strain-hardening ratio.4 to 3. 3.4.5. and a point on the actual pushover curve with base shear equal to i 0. at base shear i equal to 0. 3. A. Calculate the error = 100 × Abn − Apn ( ) ( ) ( ) ) Apn . i 3. iterations are necessary. This step gives the secant stiffness at a base shear equal to 60% of the yield base shear. Estimate the yield base shear. Calculate initial slope of the idealized bilinear curve. by connecting a straight line between origin. If the error exceeds some pre-specified tolerance. i i i 3.9. Calculate Vbny = Vbny × Apn Abn .2. kn = 0. Calculate area under the bilinear curve OAB.6 × Vbny . obtained by judgment. If desired.0. i i i αn = Vbno Vbny − 1 urno urny − 1 i 3.6 × Vbny ( ) i urn .4.0. Replace i+1 with i and repeat Steps 3. kn .6. 3.1. and B with straight-line segments to obtain the idealized bilinear curve. urny = Vbny kn .1.6 .6 × Vbny . i +1 i i 3.8. Abn . i 3. determine the roof displacement.9. urn. Calculate the slope.4. Calculate the yield displacement. Vbny .9. i i 3. Vbny . i 3. This value.

An (t ) . 4. Calculate histories of various responses using Eqs. Vbn Idealization of Pushover Curve Vbno Vbny 0. Combine the “modal” responses using Eqs. A.6 ur n y ur n o ur n Fig. Develop the Fsn Ln − Dn relation (Fig. (3. of the combined responses obtained in Step 8. (3.3.1. and pseudo-acceleration history.13).2.0. Repeat Steps 3 to 6 for as many modes as required for sufficient accuracy. 4. In general first two or three modes will suffice.2).11a).10b and 4. 4.4) and effective modal mass from M n = Ln Γn . 4. A.10a and 4.15) and (3.12) and (3. 4.eps ur n.2.1 Idealization of nth-“mode” pushover curve 73 .3b) with unit-mass and force-deformation relation of Fig. of the nth“mode” inelastic SDF system (Fig. 9. r o . 5.11b). 8.4. 6. Calculate peak values. Scale the vertical axis with * Mn to obtain * Fsno Ln = Vbno M n and * Fsny Ln = Vbny M n (Eqs. Compute deformation history. (3.16). * 4. 7. Dn (t ) . Scale the horizontal axis by Γnφrn to obtain Dno = urno Γ nφrn and Dny = urny Γ n φrn (Eqs.6 × Vbny A • Idealized αnkn B • 1 Actual k 1 • O n figA_1. A. Compute the Ln and Γ n from Eq.

computed for the first three modes are shown in Fig. 25.3. A.6 kN. ur1o = 63. The anchor point.eps Dny = ur n y / Γnφr n Dno = ur n o / Γnφr n Dn Fig. 1. 3. 3.).2 Properties of the nth-“mode” inelastic SDF system A. 74 . Idealized bilinear curves for each of the three modes are included in Fig.1.).5. to the north-south component of the El Centro (1940) ground motion scaled up by a factor of 1. and third mode.urn ) pushover curve for the force distribution s* : n 2.5 cm (25 in. Following is step-by-step implementation of the procedure described in Section A.Fsn / Ln Fsn / Ln − Dn Relationship Vbno / M* n Vbny / Mn * 1 α ω2 n n ω2 1 n figA_2. 3. n 2.4 cm (10 in.2. Target displacements used to generate these pushover curves are 63. First three mode shapes and frequencies of the selected building were computed and are shown in Fig.).3. are shown in Fig.2 EXAMPLE The UMRHA procedure is implemented to calculate the response of the 9-story building described in Section 3. The base-shear – roof-displacement (Vbn . second. 3. Area under the actual pushover curve. The pushover curves for the first three modes. 3. A. A.3.2.4. The force distributions. for the first. Ap1 = 360777 kN-cm. respectively. B. 2. and 12. At this point. s* . is defined at the target roof displacement.4.1. generated using DRAIN-2DX. The following steps illustrate the procedure to develop the idealize curve for the first “mode”.5 cm and Vb1o = 8729.1.7 cm (5 in.

6 = 4803. i 3.6.4 210.5.01%.9.4 × (360777 365100 ) = 7911. ur1. Therefore. The first estimate of the yield base shear Vb1 y = 8006.6 × Vb1 y = 4803.2.4 kN.18 kN/cm.09 cm.09 ) − 1 = 0. This value exceeds the prespecified tolerance of 0.2. 4.09 cm and Vb1 y = 8006.9 kN.23 cm. i i i 3.3666 = 3740189 kg.4 ) − 1 (63.4. Ab1 = 365100 kN-cm. and B with straight-line segments gives the idealized bilinear curve. i i 3.6 × Vb1 y ( ) i ur1. α1 = Vb1o Vb1 y − 1 ur1o ur1 y − 1 = ( ) ( ) (8729.86 = 210.5 38.86 cm at 0. i +1 3. and M1 = 2736789 × 1. iterations are necessary. i i 3.9. Area under the bilinear curve OAB.3.194. Also included are the modal damping ratios and the periods calculated from Eq.0. 75 . ur1 y = Vb1 y k1 = 8006. The post-yielding strain-hardening ratio.1.1. i 3. The point A on the i i bilinear curve is defined by ur1 y = 38.4 kN.135. The results of the iterative procedure are summarized in Table A.6 8006. k1 = 0.6 = 22. The curve OAB obtained by connecting the three points O.4. is calculated as follows.8 kN.9.2.8 22. 3.i 3.8. The next estimate of the yield shear is Vbny = 8006.7. (4. The results for other modes are summarized in Table A. The yield displacement. and α1 = 0. The initial slope of the idealized bilinear curve.18 = 38. Determined from the pushover database. 3. Vb1 y = 7615. The procedure converged after nineteen cycles to give ur1 y = 36. L1 = 2736789 kg. Γ1 = 1.13).1. k1 . Error = 100 × (365100 − 360777 ) 360777 = 1.198%.0. 3.3666. * 4. The Fs1 L1 − D1 relation for the first “mode” is developed as follows.6 kN.1. i i i 3. A.4.

* M1 gives Fs1o L1 = 233.4. Scaling the horizontal axis by Γ1φr1 gives D1o = 46. 4. Scaling the vertical axis by Fs1 y L1 = 203. The peak values are computed and are summarized in Tables 4.51 cm. 4. A.1 and 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.62 (cm/sec2). 8. 4. Histories of roof displacement and top story drifts for the first “mode” are computed and presented in Fig. 4. 9.40 (cm/sec2) and 76 . 5.3.7.7. 7. The peak values are also plotted in Fig.2. 6.7. 4. The combined modal responses are presented in Fig.2.8.4.46 cm and D1y = 26. The results were generated for first three “modes’ and are included in Fig.

3.19 y by Base Shear (kN) 8000 6000 4000 2000 0 0 Actual Idealized figA_3a.eps 0 0 5 10 15 20 Roof Displacement (cm) (c) "Mode" 3 Pushover Curve 25 8000 u = 4.eps 0 0 2 4 6 8 10 Roof Displacement (cm) 12 Fig.14 y by 6000 Base Shear (kN) 4000 2000 Actual Idealized figA_3c. α = 0. V = 5210 kN. A.9 cm. α = 0.13 6000 Base Shear (kN) 4000 2000 Actual Idealized figA_3b. “Modal” pushover curves for the example building 77 .eps 10 20 30 40 50 Roof Displacement (cm) (b) "Mode" 2 Pushover Curve 60 8000 u = 9. V = 7616 kN. α = 0.6 cm.(a) "Mode" 1 Pushover Curve 12000 10000 u = 36.2 cm. V y by = 4952 kN.

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) • 22.5 × El Centro ground motion for the first “mode.(a) Deformation 80 n=1 3 (b) Pseuso−acceleration n=1 • 0.m −80 0 −3 0 25 30 Fig.223 −80 80 n=3 −3 3 n=3 D3 (cm) 0 A3 (g) • 10. 78 .” and third “mode” inelastic SDF systems.4.06 0 • 1.52 5 10 15 20 Time (sec) 25 30 0 • 1. A.” second “mode.2769 D1 (cm) 0 • 35.747 5 10 15 20 Time (sec) figa_4.

2406 167531.135 0.12 3876.9 4573.182 0.18 210.18 210.910 0.38 22.44 36.35 36.18 210.3 4603.5 4614.1 1013.50 36.010 (kN) 8006.948 ζ n (%) 79 .3666 3740189 203.151 0.85 0.18 210.0 4588.56 47.02 21.180 0.78 21.18 210.2 4628.7 7639.0 4704.5309 488839.193 0.4 4595.022 0.65 1226.18 210.59 22.2.75 21.5 (cm) 22.237 0.09 37.28 36.64 37.76 21.25 36.09 18.Table A.8525 1.86 22.24 36.86 21.18 (cm) 38.23 22.74 (kN/cm) 210.8 7622.46 2.194 (kN-cm) 365100 364060 363276 362684 362235 361892 361631 361432 361279 361162 361073 361004 360951 360911 360880 360856 360838 360824 360813 Table A.4 7911.6 × Vb1 y i ur1.18 210.85 36.0.107 0.3 7658.4 7672.51 2 “Mode” 2 -920860 -0.309 0.18 210.9 4570.529 0.05 52.103 “Mode” 3 696400 0.75 21.063 0.83 21.193 0.26 36. i 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 i Vb1 y i 0.29 36. Results of iterative procedure to develop the idealized bilinear curve for the first “mode” inelastic SDF system Itr.18 210.5 3109.188 0.18 210.404 0.5 7633.11 22.3 7745.017 0.18 210.18 210. Properties of “modal” inelastic SDF systems Properties Ln (kg) Γn * M n (kg) “Mode” 1 2736789 1.40 46.74 21.1 4569.0 4577.693 0.082 0.62 26.186 0.25 36.18 210.4927 1.32 36.18 210.2 4671.193 0.05 36.162 0.136 Fsny Ln (cm/sec2) Dny (cm) Fsno Ln (cm/sec ) Dno (cm) Tn (sec) 233.190 0.2 4571.194 0.39 36.4 7647.30 37.013 0.3 7628.198 0.8 4647.194 0.9 4570.192 0.194 0.18 210.18 210.3 7786.1.9 7615.176 0.1 4574.9 (kN) 4803.8 7618.2671 1.79 0.0 7619.139 0.8 4747.2 7690.59 36.029 0.170 0.56 19.4 7714.81 21.18 210.23 0.048 0.6 4583.77 21.1 7616.18 210.184 0.7 4580.191 0.5 7624.037 0.95 21. No.70 36.6 7840.79 21.6 i k1 i ur1 y i α1 i Ab1 Error (%) 1.90 21.

80 .

rno . (4.2 EXAMPLE The MPA procedure is implemented to calculate the response of the 9-story building described in Section 3.4.1. 11.1 to the north-south component of the El Centro (1940) ground motion scaled up by a factor of 1. Steps 1 to 4 of the MPA are the same as those for UMRHA presented in Appendix A.18). Calculate the peak roof displacement urno associated with the nth-“mode” inelastic SDF system from Eq.6b by solving Eq.3b) with unit mass and force-deformation relation of Fig. of the nth-“mode” inelastic SDF system (Fig. Dn . Compute the peak deformation. Typically. 6. Following is step-by-step implementation of the procedure described in Section B. (3. Repeat Steps 3 to 7 for as many “modes” as required for sufficient accuracy. the first two or three “modes” will suffice.Appendix B Modal Pushover Analysis B. 10. From the total hinge rotation. 10.8) for the peak deformation of the first-“mode” inelastic SDF system with unit mass and force-deformation relation developed in Step 4 gives D1 = 35.33 cm. 12. extract from the pushover database values of other desired responses. At urno .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.8). 4.5. B. 13.21). Solving Eq. Determine the total response by combining the peak “modal” responses using the SRSS combination rule of Eq. (4. 81 . This is the same example as solved in Appendix A. 4. subtract the yield hinge rotation to determine the plastic hinge rotation. or from the inelastic response (or design) spectrum. (3.

8.18) are also included in Tables 4.3 for the floor displacements normalized by the building height and in Table 4.3 and 4. The values are summarized in Table 4. 9. Results of Steps 5 and 6 are summarized in Table B.28 cm.33 = 48. 12.4.4 for the story drifts normalized by the story height.11.366 × 1 × 35.28 cm. Steps 3 to 7 are repeated for first three “modes.3 and 4. At ur1o = 48. Also included in Table 4.4. The total response computed by combining the peak “modal” responses using the SRSS combination rule of Eq. 82 . values of floor displacements and story drifts are extracted.1.” and the results are included in Tables 4. (3. where results for other ground motion intensities are also included. Peak roof displacement ur1o = Γ1 × φr1 × D1 = 1.5 are the plastic hinge rotations computed by subtracting the yield hinge rotation from the total hinge rotations.

35 0.50 35.27 0.59 0.225 2.18 27.139 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.5 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.678 0.535 14.52 “Mode” 3 1.504 18.25 Quantity Dn (cm) “Mode” 1 6.25 20.79 0.85 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 1.229 8.969 0.154 78.275 1.007 36.0 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 3.676 6.03 26.748 63.690 10.268 0.312 1.Table B.200 0.70 0.13 2.457 12.38 22.0 µ = Dn Dny urno = Γnφrn Dn (cm) 83 .913 22.1.577 16.267 5.75 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 0.82 1.36 1.660 14.856 31.332 48.0 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 1.332 13.551 2.35 1.735 3.367 1.184 0.450 4.185 11.467 14.755 27.71 1.126 13.37 1.5 µ = Dn Dny urno = Γnφrn Dn (cm) Dn (cm) 2.73 24.03 0.766 7.755 0.117 5.33 1.37 57.379 21.8451 5.395 0.28 46.252 9.513 0.901 8.05 0.4222 3.07 “Mode” 2 4.436 7.52 0.06 1.023 0. Calculation of roof displacements urno from peak deformation of inelastic SDF systems Ground Motion Multiplier 0.691 0.

84

**Appendix C FEMA Force Distribution Calculations
**

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

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

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

Floor, j

mj (10-3×kg)

s* = j

∑i mi

0.112 0.110 0.110 0.110 0.110 0.110 0.110 0.110 0.119

mj

1 2 3 4 5 6 7 8 9

503.5 494.7 494.7 494.7 494.7 494.7 494.7 494.7 534.1

C.2

EQUIVALENT LATERAL FORCE (ELF) DISTRIBUTION 85

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

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

Floor j 1 2 3 4 5 6 7 8 9

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

-3 k

s* j

=

m jhk j

∑i mihik

0.007 0.020 0.038 0.062 0.091 0.126 0.165 0.210 0.281

C.3

SRSS DISTRIBUTION

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

N

∑ n (V jn )

2

to get column 8 of

1 87.9 446.0 1231.0 -5.7 355.9 1446.5 215.3 222.7 694.2 105.5 320.4 400.042 0.2 97.6 366.7 234.065 0.9 1683.7 (10) 0.6 -352.9 880.7 Vj (kN) (8) 2268.6 319.0 176.7 -46.045 0.6 -359.6 -732.6 V j3 (kN) (7) 478.1 1857.2 Lateral Force fj (kN) (9) 203.0 980.0 354.7 1622.367 87 .098 0. FEMA 273 SRSS force distribution Lateral Forces Floor j (1) 1 2 3 4 5 6 7 8 9 f j1 (kN) f j2 (kN) f j3 (kN) V j1 (kN) (5) 1917.3 240.9 832.2 95.8 374.2 200.5 -350.2 277.6 1233. the lateral forces are normalized by the base shear to obtain column 10 in Table C.1 -525.1 832.0 136.2 fj = ∑ fi i fj (2) (3) (4) 59.2 148.9 366.1 -646.177 0.1 Story Shears V j2 (kN) (6) 1114.070 0.6 286.1 -438.3 -646.8 1381.7 525.For convenience.5 -973.8 430.4 1759.7 101.7 2065.3.5 -320.7 1578. Table C.7 95.7 374.2 285.0 381.9 -166.047 0.5 159.3 -6.090 0.1 -967.8 -326.3.0 1476.9 -153.4 1842.4 250.

- 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

Sign up to vote on this title

UsefulNot usefula research report modal pushover analysis

a research report modal pushover analysis

- Nonlinear Pushover Analysis of Reinforced Concrete Structuresby betacivil
- Performance Based Seismic Designby ijump12
- Action Plan for Performance Based Seismic Designby Daniel Reynolds
- 2013 Pushover Analysis Guide Final RC Building With Masonry Infill Walls EditedDraft January 2013by Juan Carlos Jiménez Pacheco

- Pushover analysis
- Pushover Analysis using ETABS and SAP2000
- EVALUATION OF PUSHOVER ANALYSIS PROCEDURES
- Pushover Analysis Final
- Pushover Ppt
- Performance Based Design Using Nonlinear Analysis
- Non Linear Analysis Pushover
- Non-Linear Static Push Over Analysis
- Nonlinear Pushover Analysis of Reinforced Concrete Structures
- Performance Based Seismic Design
- Action Plan for Performance Based Seismic Design
- 2013 Pushover Analysis Guide Final RC Building With Masonry Infill Walls EditedDraft January 2013
- Pushover Analysis Using Etabs
- A comparison of single-run pushover analysis techniques for
- SAP-Etabs-Course-360
- Performing Pushover Analysis in Etabs
- Thesis
- Etabs-step by Step
- Pushover Analysis
- ATC-40 Vol 2
- ETABS Pushover Analysis
- A Modal Pushover Analysis Procedure to Estimate
- 15WCEE 1940 pushover analysis.pdf
- Pushover
- Pushover
- Etabs Guide
- pushover
- ATC-40
- About Pushover Analysis and SAP 2000 Method
- Seismic+Design+Aids+for+Nonlinear+Pushover+Analysis+of+Reinforced+Concrete+and+Steel+Bridges
- PUSHOVER Report Chopra[1]

Are you sure?

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

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

Get the full title to continue

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

scribd