INELASTIC DISPLACEMENT RATIOS FOR STRUCTURES ON FIRM SITES

By Eduardo Miranda,1 Member, ASCE

ABSTRACT: The results of a comprehensive statistical study of inelastic displacement ratios that permit the
estimation of maximum lateral inelastic displacement demands on a structure from maximum lateral elastic
displacement demands are presented. These ratios were computed for single-degree-of-freedom systems under-
going different levels of inelastic deformation when subjected to a relatively large number of recorded earthquake
ground motions. The study is based on 264 acceleration time histories recorded on firm sites during various
earthquakes in California. Three types of soil conditions with shear-wave velocities higher than 180 m/s are
considered. The influences of period of vibration, level of ductility demand, site conditions, earthquake mag-
nitude, and epicentral distance are carefully evaluated and discussed. Inelastic displacement ratios associated
with mean values are presented. Special emphasis is given to the disperson of the results. It is concluded that
Downloaded from ascelibrary.org by East Carolina University on 09/11/13. Copyright ASCE. For personal use only; all rights reserved.

for sites with average shear-wave velocities higher than 180 m/s the influence of soil conditions is relatively
small and can be neglected for design purposes. Finally, results from nonlinear regression analyses are presented
that provide a simplified expression to be used in the design to approximate mean inelastic displacements ratios
for structures on firm sites.

INTRODUCTION range of hysteretic-response types, oscillator-spring strengths,

and ground motion characteristics. Based on this information,
In general, nonlinear time history analyses of structures may they developed an equation based on the ratio of the charac-
produce a good estimation of global and local deformation teristic period of the structure to the characteristic period of
demands for a given acceleration time history. However, at the the ground motion and on the ratio of the base shear strength
present time these analyses are still considered unpractical for to the base shear for linear response to evaluate cases where
everyday design situations. Furthermore, the results of such the ratio of maximum inelastic displacement to maximum elas-
analyses tend to be sensitive to the specific characteristics tic displacement was <1. Similar to the conclusion reached by
of the ground motion and in some cases to the assumptions Veletsos, Shimazaki and Sozen (1984) noted that in the short-
regarding the behavior of individual structural components. period spectral region the ratio of maximum inelastic displace-
Thus, simple, yet reliable methods for estimating lateral in- ment to maximum elastic displacement depended critically on
elastic displacements demands on structures are needed for the the lateral strength of the structure relative to the elastic
design of new structures or during the seismic evaluation and strength demand and that the estimation of the inelastic dis-
rehabilitation of existing structures. placement demand was beyond a simple procedure. These con-
A particularly appealing simplified approach is to try to es- clusions were also more recently confirmed by Qi and Moehle
timate the maximum inelastic response, and, in particular, the (1991) while developing a displacement design approach for
maximum lateral inelastic displacement demand, using the re- reinforced concrete structures subjected to earthquakes.
sults from a linear elastic analysis. Efforts in this direction Concerned with the possible limitations of the equal dis-
began many years ago. One of the first studies was done by placement rule in the short-period region and when applied to
Veletsos and Newmark (1960) and Veletsos et al. (1965), who structures on soft soils sites, Miranda (1991, 1993a,b) studied
pointed out that ‘‘it is instructive to relate the maximum de-
the ratio of maximum inelastic displacement to maximum elas-
formation of the elastoplastic system to that of an elastic sys-
tic displacement of SDOF systems with the same initial period
tem having the same stiffness as the initial stiffness of the
and same damping ratio as the inelastic systems when sub-
inelastic system.’’ Using single-degree-of-freedom (SDOF)
jected to 124 earthquake ground motions recorded on different
systems subjected to simple pulses and to three earthquake
types of soil conditions. Mean constant-ductility ratios of max-
ground motions, they noticed that in the low frequency range
imum inelastic to maximum elastic response for three types
(frequencies smaller than 0.38 Hz) the maximum deformation
of soil conditions were computed as part of this investiga-
of the inelastic and the associated elastic systems may be con-
tion. This study gave a special insight to this ratio in the
sidered the same. This observation gave rise to the so-called
short-period range and to the limiting periods of the spectral
equal displacement rule, which is the basis for estimating max-
regions where the equal displacement rule is applicable. The
imum deformations in certain spectral regions in most building
codes. This study also concluded that, in the high frequency results of Miranda (1991, 1993a,b) were confirmed by Kra-
and moderately high frequency regions, the inelastic displace- winkler and his coworkers at Stanford University (Rahnama
ments are significantly higher than their elastic counterparts. and Krawinkler 1993; Severinatna and Krawinkler 1997), who
While searching for an energy criterion to rationalize the used a smaller set of ground motions and also studied the
use of reduced forces in design, Shimazaki and Sozen (1984) influence of stiffness degradation, pinching, and P-⌬ effects
noticed that, in the spectral region of nearly constant energy on this ratio. More recently, Gupta and Kunnath (1998) studied
response, the maximum nonlinear drift of reinforced concrete the effect of stiffness degradation, strength degradation, and
structures could be determined as a function of the linear drift. pinching on inelastic displacement demands using a three-
Additionally, they noticed that this observation held for a wide parameter model and 15 ground motions.
On the other hand, various studies have shown that the
1
Asst. Prof., Dept. of Civ. and Envir. Engrg., Stanford Univ., Stanford, global inelastic response of many multi-degree-of-freedom
CA 94305-4020. structures can be estimated, with appropriate modification fac-
Note. Associate Editor: Brad Cross. Discussion open until March 1, tors, from the response of SDOF systems (Saiidi and Sozen
2001. To extend the closing date one month, a written request must be 1981; Fajfar and Fischinger 1988; Miranda 1991, 1999; Qi
filed with the ASCE Manager of Journals. The manuscript for this paper
and Moehle 1991; Severinatna and Krawinkler 1997). Based
was submitted for review and possible publication on September 20,
1999. This paper is part of the Journal of Structural Engineering, Vol. on these results, several design recommendations now explic-
126, No. 10, October, 2000. 䉷ASCE, ISSN 0733-9445/00/0010-1150– itly take into account the ratio of the maximum inelastic dis-
1159/$8.00 ⫹ $.50 per page. Paper No. 21903. placement to the minimum elastic displacement as an analysis
1150 / JOURNAL OF STRUCTURAL ENGINEERING / OCTOBER 2000

J. Struct. Eng. 2000.126:1150-1159.

 tool to estimate the inelastic response of structures to earth- persion of the inelastic displacement ratios. All of the ground
quake ground motions [Applied Technology Council (ATC) motions selected have the following characteristics: (1) Re-
1996; FEMA 1997)]. The proposed ratios Rd in ATC-32 and corded on accelerographic stations where detailed information
C1 in FEMA-273 have been primarily based on the results exists on the geological and geotechnical conditions at the site
from the study by Miranda (1991), whereas the modification that enables the classification of the recording site in accor-
factors proposed in the FEMA-273 guidelines (1997a) are pri- dance to recent code recommendations (ATC 1996; FEMA
marily based on the results by Krawinkler and his coworkers 1997a,b; International Conference of Building Officials
(Rahnama and Krawinkler 1993; Severinatna and Krawinkler (ICBO) 1997); (2) recorded on rock or firm sites with average
1997). shear-wave velocities higher than 180 m/s (600 ft/s) in the
The objective of this paper is to present the results of a upper 30 m (100 ft) of the site profile; (3) recorded on free
comprehensive statistical study of the ratio of maximum in- field stations or in the first floor of low-rise buildings with
elastic displacement demand to maximum elastic displacement negligible soil-structure interaction effects; (4) recorded in
demand for firm sites. This study can be considered as a fol-
earthquakes with surface-wave magnitudes Ms larger than 5.7;
low-up of the study of Miranda (1991). However, this study
and (5) records in which at least one of the two horizontal
Downloaded from ascelibrary.org by East Carolina University on 09/11/13. Copyright ASCE. For personal use only; all rights reserved.

tries to more specifically quantify the influence on soil con-

components had a peak ground acceleration larger than 40
ditions, earthquake magnitude, and epicentral distance as well
as to present quantitative results of the dispersion on these cm/s2.
ratios. This study makes use of improved information that has Recent seismic design provisions in the United States in-
been recently made available on the geological characteristics clude a new site classification. In particular, for firm sites with
at accelerographic recording stations in California. The inves- average shear-wave velocities higher than 180 m/s in the upper
tigation is limited to rock and relatively firm soil sites with 30 m of the site profile there are four different site classes.
shear-wave velocities higher than 180 m/s. In particular, this For these four site classes the linear elastic design spectra in-
study evaluates whether the maximum inelastic displacement cluded in these design recommendations may be significantly
to maximum elastic displacement ratios need modifications for different from each other. Thus, it is particularly important to
various firm site classes for which recent seismic design pro- know whether inelastic displacement ratios to be used in de-
visions recommend different elastic design spectra. sign for estimating maximum inelastic displacements from
maximum elastic displacements for these site classes also need
INELASTIC DISPLACEMENT RATIOS to be different from each other. For this purpose, the earth-
quake ground motions were divided into three groups accord-
The inelastic displacement ratio C␮ is defined as the maxi-
ing to the soil conditions at the recording stations. The first
mum lateral inelastic displacement demand ⌬inelastic divided by
group consisted of ground motions recorded on stations lo-
the maximum lateral elastic displacement demand ⌬elastic on a
system with the same mass and initial stiffness (i.e., same pe- cated on rock with average shear-wave velocities higher than
riod of vibration) when subjected to the same earthquake 760 m/s (2,500 ft/s). The second group consisted of records
ground motion. Mathematically this is expressed as obtained on stations on very dense soil or soft rock with av-
erage shear-wave velocities between 360 m/s (1,200 ft/s) and
⌬inelastic 760 m/s, and the third group consisted of ground motions re-
C␮ = (1)
⌬elastic corded on stations on stiff soil with average shear-wave ve-
locities between 180 m/s (600 ft/s) and 360 m/s. Recording
Thus, if information on this ratio is available, an estimation stations on the first group correspond to site classes A and B
of the maximum inelastic displacement can be obtained from
according to recent design provisions (ATC 1996; FEMA
the maximum elastic displacement. In the study presented
1997a,b; ICBO 1997), and recording stations on the second
here, inelastic displacement ratios were computed for SDOF
systems having a viscous damping ratio of 5% and a nonlinear and third group correspond to site classes C and D, respec-
elastoplastic hysteretic behavior. tively. Table 1 lists all of the records considered in the study.
For earthquake design purposes it is useful to compute in- For each record earthquake magnitude, NEHRP site class,
elastic displacement ratios corresponding to different values of shortest distance to the horizontal projection of the rupture,
displacement ductility ratios ␮ in accordance to the inelastic epicentral distance, and peak ground acceleration are given. It
deformation capacity of a structure. In this study, maximum can be seen that the smallest earthquake magnitude considered
inelastic displacement demands ⌬inelastic corresponding to spe- here corresponds to the 1991 Sierra Madre earthquake (Ms =
cific values of ␮ were computed by iteration on the lateral 5.8) and the largest magnitude corresponds to the 1957 Kern
strength of the system until the displacement ductility demand County earthquake (Ms = 7.7).
was, within a tolerance, equal to the specified ductility ratio. Fig. 1 shows the distribution of magnitudes and distances
The tolerance was chosen such that ⌬inelastic was considered to the horizontal projection of the rupture of the earthquake
satisfactory if the computed ductility demand was within 1% ground motions selected in the study. It can be seen that with
of the specified displacement ductility. the exception of one recording station, all other stations were
Inelastic displacement ratios were computed for SDOF sys- closer than 120 km (75 mi) from the rupture. Similarly, most
tems subjected to six different levels of inelastic deformation of the records selected are from earthquakes with magnitudes
corresponding to the following ductility ratios: 1.5, 2, 3, 4, 5, between 6 and 7.5. Fig. 2 shows peak ground accelerations
and 6. For each earthquake record and each target displace- and corresponding distances to the rupture for the 264 records
ment ductility ratio, the inelastic displacement ratios were considered in this study. It can be seen that the set of ground
computed for a set of 50 periods of vibration between 0.05 motions selected, with the exception of a few records, follows
and 3.0 s. the overall trend that one would expect from known attenua-
tion relationships; high levels of maximum ground accelera-
EARTHQUAKE GROUND MOTIONS CONSIDERED tion correspond to recording sites with small distances to the
A total of 264 earthquake acceleration time histories re- rupture, and, as this distance increases, the maximum ground
corded in the state of California for 12 different earthquakes acceleration decreases. Distances to the horizontal projection
were used in this study. A particularly large number of earth- of the rupture range from 1 to 156 km, and peak ground ac-
quake ground motions was selected in order to assess the dis- celerations range from 27 to 770 cm/s2.
JOURNAL OF STRUCTURAL ENGINEERING / OCTOBER 2000 / 1151

J. Struct. Eng. 2000.126:1150-1159.

 TABLE 1. Recorded Earthquake Ground Motions Used in This Study

Magnitude Station NEHRP Dist. 1 Dist. 2 Comp. 1 PGA Comp. 2 PGA

Date Earthquake name (Ms ) Station name number site class (km) (km) (deg) (cm/s2) (deg) (cm/s2)
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)
02/09/71 San Fernando 6.5 Los Angeles, Gritfith Park 141 AB 18.0 33.6 180 183.7 270 173.7
Observatory
02/09/71 San Fernando 6.5 Lake Hughes, Array Station 4 126 AB 19.6 26.6 111 168.2 201 ⫺143.5
02/09/71 San Fernando 6.5 Pasadena, CIT Cal Tech 266 AB 22.0 25.5 180 92.9 270 198.5
‘ Seismo Lab
02/09/71 San Fernando 6.5 Lake Hughes, Array Station 9 127 AB 23.0 26.6 21 119.3 291 ⫺109.4
10/15/79 Imperial Valley 6.8 Superstition Mountain 286 AB 26.0 60.6 135 ⫺189.2 45 108.0
04/24/84 Morgan Hill 6.1 Gilroy 1, Gavillan College 47379 AB 16.0 38.6 230 ⫺57.5 320 93.4
07/08/86 Palm Springs 6.0 Silent Valley, Poppet Flat 12206 AB 23.7 27.9 0 102.6 90 107.4
07/08/86 Palm Springs 6.0 Winchester, Hidden Valley 13200 AB 49.8 54.5 90 54.6 270 56.5
Farms
Downloaded from ascelibrary.org by East Carolina University on 09/11/13. Copyright ASCE. For personal use only; all rights reserved.

07/08/86 Palm Springs 6.0 Winchester, Bergman Ranch 13199 AB 55.3 60.1 0 62.2 90 85.7
07/08/86 Palm Springs 6.0 Murrieta Springs, Collings 13198 AB 61.0 65.8 0 45.9 90 49.4
Ranch
10/01/87 Whittier 6.1 Mt. Wilson, CIT Seismic Sta- 24399 AB 22.1 18.2 0 ⫺121.3 90 ⫺171.3
tion
10/01/87 Whittier 6.1 Los Angeles, Gritfith Park 141 AB 22.3 21.5 0 ⫺133.8 360 ⫺121.4
Observatory
10/04/87 Whittier aftershock 6.1 Mt. Wilson, CIT Seismic Sta- 24399 AB 17.9 17.2 90 136.0 0 142.5
tion
10/17/89 Loma Prieta 7.1 Gilroy 1, Gavillan College 47379 AB 10.5 28.4 90 433.6 360 426.6
10/17/89 Loma Prieta 7.1 Hollister, SAGO Vault Tunnel 1032 AB 29.9 49.0 360 ⫺60.1 270 ⫺36.3
10/17/89 Loma Prieta 7.1 Hollister, SAGO South 47189 AB 32.4 47.4 261 70.7 351 65.3
Cinega Road surface
10/17/89 Loma Prieta 7.1 Monterey, City Hall 47377 AB 42.7 49.0 90 61.1 360 68.5
10/17/89 Loma Prieta 7.1 South San Francisco, Sierra 58539 AB 67.6 84.0 115 57.2 205 102.7
Point
10/17/89 Loma Prieta 7.1 Bear Valley, Station 7, Pin- 1476 AB 69.0 87.5 310 ⫺40.6 220 45.2
nacles National Mon.
10/17/89 Loma Prieta 7.1 San Francisco, Dimond 58130 AB 75.9 92.0 90 110.8 360 96.4
Heighs
10/17/89 Loma Prieta 7.1 Piedmont, Piedmont Jr. High 58338 AB 77.2 92.6 45 81.2 315 69.6
Grounds
10/17/89 Loma Prieta 7.1 San Francisco, Rincon Hill 58151 AB 78.5 94.5 90 88.5 0 78.6
10/17/89 Loma Prieta 7.1 San Francisco, Pacific 58131 AB 80.5 96.6 270 60.2 360 46.3
Heights
10/17/89 Loma Prieta 7.1 San Francisco, Cliff House 58132 AB 87.4 99.4 0 ⫺73.1 90 ⫺105.7
10/17/89 Loma Prieta 7.1 San Francisco, Telegraph Hill 58133 AB 88.0 96.9 90 51.2 0 90.5
10/17/89 Loma Prieta 7.1 Point Bonita 58043 AB 88.1 103.6 297 71.4 207 69.9
06/28/91 Sierra Madre 5.8 Pasadena, CIT Cal Tech 266 AB 19.0 19.7 360 273.8 270 ⫺172.6
Seismo Lab
06/28/91 Sierra Madre 5.8 Mt. Wilson, CIT Seismic Sta- 24399 AB 5.3 6.4 0 270.7 90 196.2
tion
06/28/92 Landers 7.5 Twentinine Palms Part Main- 22161 AB 41.9 44.2 0 ⫺78.7 90 ⫺59.1
tenance Building
06/28/92 Landers 7.5 Silent Valley, Poppet Flat 12206 AB 51.3 39.9 90 39.4 0 48.9
01/17/94 Northridge 6.8 Malibu Canyon, Griffith Ob- 5080 AB 20.2 19.5 360 ⫺176.4 270 270.0
servatory
01/17/94 Northridge 6.8 Los Angeles, Griffith Park 141 AB 24.5 26.0 360 162.9 270 282.1
Observatory
01/17/94 Northridge 6.8 Lake Hughes, Array Station 9 24272 AB 28.4 44.7 90 221.2 360 154.5
01/17/94 Northridge 6.8 Los Angeles, Temple & Hope 24611 AB 32.2 31.9 180 189.1 90 123.7
01/17/94 Northridge 6.8 Lake Hughes, Array Station 4 24469 AB 34.0 49.4 0 56.4 90 82.4
01/17/94 Northridge 6.8 Mt. Wilson, CIT Seismic Sta- 24399 AB 36.9 44.6 90 ⫺130.7 360 228.5
tion
01/17/94 Northridge 6.8 Los Angeles, City Terrace 24592 AB 37.1 38.3 90 258.0 0 310.1
01/17/94 Northridge 6.8 Antelope Buttes 24310 AB 48.6 63.3 90 99.7 0 44.9
01/17/94 Northridge 6.8 San Pedro, Palos Verdes 14159 AB 58.5 58.3 90 93.1 0 98.9
07/21/52 Kern County 7.7 Santa Barbara, Courthouse 283 C 85.0 89.8 42 ⫺87.8 132 128.6
07/21/52 Kern County 7.7 Pasadena, CIT Athenaeum 475 C 109.0 126.2 180 ⫺46.5 270 ⫺52.1
02/09/71 San Fernando 6.5 Lake Hughes, Array Station 128 C 17.0 23.1 21 ⫺346.2 291 277.9
12
02/09/71 San Fernando 6.5 Glendale, 633 E. Broadway 122 C 18.0 32.1 110 265.7 200 ⫺209.1
02/09/71 San Fernando 6.5 Lake Hughes #1, Fire Station 125 C 25.0 30.2 21 ⫺145.5 111 108.9
#78
02/09/71 San Fernando 6.5 Castaic Old Ridge Route 110 C 26.0 27.6 21 ⫺309.4 291 ⫺265.4
02/09/71 San Fernando 6.5 Pearbbssom Pump Plant 585 C 36.0 45.2 0 ⫺91.5 270 120.5
10/15/79 Imperial Valley 6.8 El Centro, Parachute Test Fa- 5051 C 14.0 50.2 225 106.9 315 200.2
cility
04/24/84 Morgan Hill 6.1 Gilroy #6, San Ysidro Micro- 57383 C 11.5 35.9 0 ⫺214.8 90 ⫺280.4
wave Site
04/24/84 Morgan Hill 6.1 Gilroy Gavillan College 47006 C 16.0 38.6 337 85.9 67 95.0
Physical Science Building

1152 / JOURNAL OF STRUCTURAL ENGINEERING / OCTOBER 2000

J. Struct. Eng. 2000.126:1150-1159.

 TABLE 1. (Continued )
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)
10/01/87 Whittier 6.1 Garvey Reservoir Abutment 709 C 3.4 11.3 60 ⫺367.1 330 ⫺468.2
Building
10/01/87 Whittier 6.1 Alhambra, 900 S. Fremont 24461 C 6.7 12.3 180 286.2 270 374.3
10/01/87 Whittier 6.1 San Marino, SW Academy 24401 C 7.7 12.9 270 136.5 360 183.8
10/01/87 Whittier 6.1 Inglewood, Union Oil Yard 14196 C 22.5 25.0 0 246.1 90 219.3
10/01/87 Whittier 6.1 Long Beach, Recreation Park 14241 C 29.6 31.9 90 ⫺57.2 180 53.8
10/01/87 Whittier 6.1 Sylmar, Olive View Medical 24514 C 45.9 44.8 0 ⫺55.8 90 50.3
Center
10/01/87 Whittier 6.1 Riverside, Airport 13123 C 57.8 59.6 180 38.4 270 ⫺56.8
10/01/87 Whittier 6.1 Lancaster, Medical Office 24526 C 72.2 70.1 10 ⫺59.4 100 59.6
Building FF
10/01/87 Whittier 6.1 Castaic, Old Ridge Route 24278 C 77.3 75.9 0 ⫺67.2 90 ⫺65.4
10/17/89 Loma Prieta 7.1 Corralitos, Eureka Canyon 57007 C 0.0 6.9 90 469.4 360 617.7
Downloaded from ascelibrary.org by East Carolina University on 09/11/13. Copyright ASCE. For personal use only; all rights reserved.

Road
10/17/89 Loma Prieta 7.1 Gilroy, Gavillan College 47006 C 10.9 28.7 67 349.1 337 310.0
Physical Science Building
10/17/89 Loma Prieta 7.1 Saratoga, Aloha Avenue 58065 C 12.4 27.4 90 316.2 0 494.5
10/17/89 Loma Prieta 7.1 Santa Cruz, UCSC 58135 C 12.5 16.6 90 401.5 360 433.1
10/17/89 Loma Prieta 7.1 Gilroy 6, San Ysidro Micro- 57383 C 19.9 35.2 90 166.9 0 112.2
wave Site
10/17/89 Loma Prieta 7.1 Coyote Lake Dam, down- 57504 C 21.7 30.9 285 174.7 195 154.7
stream
10/17/89 Loma Prieta 7.1 Woodside, Fire Station 58127 C 38.7 54.8 90 79.7 0 79.5
10/17/89 Loma Prieta 7.1 Fremont, Mission San Jose 57064 C 42.6 54.8 90 100.5 0 117.7
10/17/89 Loma Prieta 7.1 Hayward, CSUH Stadium 58219 C 56.7 71.0 90 82.6 0 72.5
10/17/89 Loma Prieta 7.1 Yerba Buena Island 58163 C 80.2 95.4 90 65.8 0 28.1
10/17/89 Loma Prieta 7.1 Berkeley, Lawrence Berkeley 58471 C 83.9 98.6 90 114.4 0 47.7
Laboratory
06/28/92 Landers 7.5 Joshua Tree, Fire Station 22170 C 7.1 11.3 0 268.3 90 278.4
01/17/94 Northridge 6.1 Castaic Old Ridge Route 24278 C 24.6 40.1 360 504.2 90 557.1
01/17/94 Northridge 6.1 San Marino, SW Academy 24401 C 35.5 39.3 360 148.2 90 122.5
01/17/94 Northridge 6.1 Alhambra, 900 S. Fremont 24461 C 37.2 39.0 360 78.3 90 99.1
01/17/94 Northridge 6.1 Lake Hughes #1, Fire Station 24271 C 37.7 53.4 0 ⫺84.9 90 75.2
#78
01/17/94 Northridge 6.1 Wrightwood, Jackson Flat 23590 C 42.5 76.4 90 ⫺55.4 180 ⫺36.2
01/17/94 Northridge 6.1 Littlerock, Brainard Canyon 23595 C 47.9 60.1 90 70.6 180 59.0
01/17/94 Northridge 6.1 Rancho Palos Verdes, Haw- 14404 C 53.8 53.2 0 ⫺71.1 90 ⫺52.7
thorne Blvd.
07/21/52 Kern County 7.7 Los Angeles, Hollywood Sto. 135 D 107.0 118.8 90 41.2 180 ⫺58.1
PE Lot
04/09/68 Borrego Mountain 6.7 El Centro, Imperial Valley Ir- 117 D 45.0 29.5 180 ⫺127.8 270 ⫺56.3
rigation District
02/09/71 San Fernando 6.5 Los Angeles, Hollywood 135 D 23.0 24.0 90 ⫺207.0 180 167.3
Storage Building
02/09/71 San Fernando 6.5 Vernon, Cmd Terminal 288 D 33.5 49.2 187 80.5 277 104.6
02/09/71 San Fernando 6.5 Santa Ana, Enginering Build- 281 D 71.5 88.2 176 26.8 266 28.2
ing
10/15/79 Imperial Valley 6.8 El Centro #7, Imperial Valley 5028 D 0.6 29.4 230 453.7 140 326.8
College
10/15/79 Imperial Valley 6.8 El Centro #6, 551 Huston 5158 D 1.3 29.8 140 ⫺368.7 230 ⫺428.1
Road
10/15/79 Imperial Valley 6.8 Bonds Corner 5054 D 2.6 9.0 140 ⫺575.7 230 770.4
10/15/79 Imperial Valley 6.8 El Centro #8, Cruickshark 5159 D 3.8 29.6 140 598.3 230 457.4
Road
10/15/79 Imperial Valley 6.8 El Centro #5, James Road 952 D 4.0 30.5 140 ⫺517.2 230 367.2
10/15/79 Imperial Valley 6.8 El Centro, Dogwood Road 5165 D 5.0 29.5 90 ⫺284.9 180 371.9
10/15/79 Imperial Valley 6.8 El Centro #4, Anderson Road 955 D 6.8 29.7 140 483.6 230 ⫺349.7
10/15/79 Imperial Valley 6.8 Hotville, Post Office 5055 D 7.5 22.7 225 ⫺246.2 315 213.1
10/15/79 Imperial Valley 6.8 Brawley, Municipal Airport 5060 D 8.5 46.3 225 162.2 315 ⫺217.5
10/15/79 Imperial Valley 6.8 El Centro #10, Community 412 D 8.5 29.8 320 ⫺221.7 50 ⫺168.2
Hospital
10/15/79 Imperial Valley 6.8 Calexico, Fire Station 5053 D 10.6 17.4 225 269.6 315 ⫺196.9
10/15/79 Imperial Valley 6.8 El Centro #11, McCabe 5058 D 12.6 30.1 140 355.4 230 ⫺374.5
Union School
10/15/79 Imperial Valley 6.8 El Centro #3, Pine Union 5057 D 12.7 31.7 140 ⫺261.7 230 218.1
School
10/15/79 Imperial Valley 6.8 El Centro #12, 907 Brock- 931 D 18.0 32.1 140 138.7 230 113.4
man Road
10/15/79 Imperial Valley 6.8 El Centro #13, Strobel Resi- 5059 D 22.0 35.8 140 114.6 230 136.2
dence
10/15/79 Imperial Valley 6.8 El Centro #1, Borchard 5056 D 22.0 38.5 230 ⫺136.2 140 ⫺139.4
Ranch
10/15/79 Imperial Valley 6.8 Plaster City, Storehouse 5052 D 32.0 54.4 135 55.5 45 ⫺41.9
10/15/79 Imperial Valley 6.8 Coachella, Canal #4 5066 D 49.0 86.8 45 ⫺113.6 135 ⫺125.7
04/24/84 Morgan Hill 6.1 Gilroy #2, Hwy 101/Bolsa 47380 D 1.0 38.0 0 153.7 90 ⫺210.0
Road Model

JOURNAL OF STRUCTURAL ENGINEERING / OCTOBER 2000 / 1153

J. Struct. Eng. 2000.126:1150-1159.

 TABLE 1. (Continued )
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)
04/24/84 Morgan Hill 6.1 Gilroy #7, Mantnilli Ranch, 57425 D 13.7 37.6 0 183.0 90 111.5
Jamison Road
04/24/84 Morgan Hill 6.1 Gilroy #3 Sewage Treatment 47381 D 14.4 37.6 0 177.0 90 189.8
Plant
10/01/87 Whittier 6.1 Whittier Narrows Dam (up- 289 D 5.1 11.1 33 294.0 62 225.4
stream)
10/01/87 Whittier 6.1 Bell Los Angeles Bulk Mail 5129 D 10.6 12.3 10 ⫺322.1 280 436.9
Center
10/01/87 Whittier 6.1 Vernon, Cmd Terminal 288 D 11.1 13.1 7 ⫺267.3 277 ⫺239.9
10/01/87 Whittier 6.1 Downey, County Maintenance 14368 D 16.2 17.0 180 193.2 270 150.7
Building
10/01/87 Whittier 6.1 Los Angeles, Hollywood 24303 D 23.8 25.0 0 201.3 90 103.7
Storage Building
Downloaded from ascelibrary.org by East Carolina University on 09/11/13. Copyright ASCE. For personal use only; all rights reserved.

10/01/87 Whittier 6.1 Century City, Los Angeles 24390 D 29.6 32.0 0 57.6 90 ⫺67.2
Country Club South
10/01/87 Whittier 6.1 Pomona 4th, and Locust FF 23525 D 29.9 30.4 12 68.4 102 49.0
10/01/87 Whittier 6.1 Long Beach, Harbor Admin- 14395 D 32.8 35.9 0 ⫺48.2 90 ⫺68.9
istraton Building
10/01/87 Whittier 6.1 Rancho Cucamonga, Law and 23497 D 45.5 47.0 90 55.5 360 45.3
Justice Center
10/01/87 Whittier 6.1 Arleta, Nordhoff Avenue Fire 24087 D 45.7 39.0 180 87.1 270 84.2
Station
10/01/87 Whittier 6.1 Rosamond, Goode Ranch 24274 D 89.0 87.0 0 ⫺73.8 90 50.4
10/17/89 Loma Prieta 7.1 Gilroy 2, Highway 101 Bolsa 47380 D 12.1 29.5 90 316.3 0 394.2
Road Motel
10/17/89 Loma Prieta 7.1 Gilroy 3, Sewage Treatment 47381 D 14.0 31.1 90 362.0 0 531.7
Plant
10/17/89 Loma Prieta 7.1 Gilroy 7, Mantelli Ranch 57425 D 24.3 39.6 90 314.3 0 205.6
Jamison Road
10/17/89 Loma Prieta 7.1 Agnews, Agnews State Hos- 57066 D 27.0 40.0 90 157.6 0 163.1
pital
10/17/89 Loma Prieta 7.1 Salinas, John Work Street 47179 D 29.3 45.9 250 110.2 160 88.2
10/17/89 Loma Prieta 7.1 Hayward, John Muir School 58393 D 58.9 71.0 90 136.0 0 166.5
10/17/89 Loma Prieta 7.1 Oakland, two story building 58224 D 76.3 91.9 290 238.3 200 187.3
10/17/89 Loma Prieta 7.1 Richmond, City Hall Parking 58505 D 92.7 107.7 280 103.6 190 122.7
lot
06/28/92 Landers 7.5 Yermo, Fire Station 22074 D 26.3 85.8 270 ⫺240.0 360 ⫺148.6
06/28/92 Landers 7.5 Palm Springs, Airport 12025 D 28.2 41.8 0 ⫺74.2 90 ⫺87.2
06/28/92 Landers 7.5 Fort Irwin 24577 D 65.5 120.9 0 ⫺111.4 90 119.8
06/28/92 Landers 7.5 Hemet, Stetson Avenue Fire 12331 D 66.8 72.6 0 79.8 90 95.2
Station
06/28/92 Landers 7.5 Baker, Fire Station 32075 D 88.3 123.9 50 105.6 140 ⫺103.6
06/28/92 Landers 7.5 Boron, Fire Station 33083 D 92.4 142.5 0 ⫺116.6 90 88.4
06/28/92 Landers 7.5 Pomona, 4th and Locust FF 23525 D 117.6 122.0 0 65.5 90 43.2
06/28/92 Landers 7.5 Downey, County Maintenance 14368 D 156.0 162.6 0 50.4 90 38.7
Building
01/17/94 Northridge 6.8 Arleta, Nordhoff Avenue Fire 24087 D 4.0 9.9 360 302.0 90 337.3
Station
01/17/94 Northridge 6.8 Los Angeles, Hollywod Stor- 24303 D 24.8 22.9 360 381.4 90 227.0
age Building
Note: Dist. 1 is the closest distance to the horizontal projection of the rupture; Dist. 2 is the epicentral distance; PGA = peak ground acceleration.

RESULTS OF STATISTICAL STUDY

Mean Ratios for All Site Classes
A total of 78,600 inelastic displacement ratios was com-
puted as part of this investigation (corresponding to 264
ground motions, 50 periods of vibration, and 6 levels of in-
elastic deformation). Mean inelastic displacement ratios were
then computed for each period and each displacement ductility
ratio. Fig. 3 shows mean constant-ductility inelastic displace-
ment ratios corresponding to all ground motions, regardless of
their site class. It can be seen that mean inelastic displacement
ratios are characterized by being larger than 1 in the short-
period spectral region (i.e., maximum inelastic displacements
larger than maximum elastic displacements) and approxi-
mately equal to 1 (i.e., maximum inelastic displacements on
average approximately equal to maximum elastic displace-
ment) for periods longer than about 1.0 s. For periods smaller
than 1.0 s, inelastic displacement ratios are strongly dependent
FIG. 1. Magnitude versus Distance to Horizontal Projection of on the period of vibration and on the level of inelastic defor-
Rupture for Earthquake Ground Motions Considered mation. In general, in this spectral region maximum inelastic
1154 / JOURNAL OF STRUCTURAL ENGINEERING / OCTOBER 2000

J. Struct. Eng. 2000.126:1150-1159.

 to the mean. Fig. 4 shows COVs of inelastic displacement
ratios corresponding to ground motions from all site classes
considered herein. It can be seen that, as expected, dispersion
increases, as the level of inelastic deformation increases. Fur-
thermore, with the exception of very short periods (smaller
than 0.15 s), for a given level of ductility demand the COV
does not exhibit strong differences with changes in the period
of vibration. Mean COVs corresponding to periods of vibra-
tion between 0.15 and 3.0 s range from 0.15 for a displacement
ductility ratio of 1.5 to 0.38 for a displacement ductility ra-
tio of 6.
Because mean inelastic displacement ratios are approxi-
mately constant and equal to 1 for periods of vibration larger
than 1.0 s, and because COVs have only small variations with
Downloaded from ascelibrary.org by East Carolina University on 09/11/13. Copyright ASCE. For personal use only; all rights reserved.

changes in period of vibration in this spectral region, then

inelastic displacement ratios associated to mean plus one stan-
dard deviation will also be approximately period independent
for periods longer than 1.0 s. This can be confirmed in Fig. 5,
FIG. 2. Peak Ground Acceleration (PGA) versus Distance to where values of inelastic displacement ratios corresponding to
Horizontal Projection of Rupture of Earthquake Ground Motions mean plus one standard deviation are shown. It can be seen
Considered that in the spectral region where mean inelastic displacement
ratios are constant and approximately equal to 1 (i.e., periods
longer than 1.0 s) the equal displacement rule is no longer
applicable for C␮ associated to mean plus one standard devi-
ation. Furthermore, using a constant C␮ equal to 1.5 in this
spectral region as representative of ratios associated to mean

FIG. 3. Mean Inelastic Displacement Ratios from 264 Ground

Motions from Site Classes A, B, C, and D

displacements become much larger than maximum elastic dis-

placements, as the ductility demand increases and as the period
decreases. Furthermore, inelastic displacement ratios tend to- FIG. 4. COVs of Inelastic Displacement Ratios from 264
Ground Motions from Site Classes A, B, C, and D
ward ␮ as the period of vibration tends to zero. It is important
to notice that the limiting period that divides the spectral
regions where the equal displacement rule is not applicable
and is unconservative (produces an underestimation of the dis-
placement demand) from the region where this approximation
is applicable depends on the level of ductility. In general this
limiting period increases, as the displacement ductility demand
increases. For example, for a ductility ratio of 1.5, the equal
displacement rule is applicable for periods longer than about
0.2 s, and for a ductility ratio of 3, the rule is applicable for
periods longer than about 0.8 s. Similarly, for a displacement
ductility ratio of 4, the equal displacement rule is only appli-
cable for periods longer than about 1.0 s.

Dispersion
Although mean inelastic displacement ratios are very im-
portant, as they represent what can be expected on average, it
is equally important to know the scatter that exists in the re-
sults about the mean, and, in particular, it is important to quan-
tify the level of dispersion. A common and effective way to FIG. 5. Mean Plus One Standard Deviation of Inelastic Dis-
quantify the dispersion is through the coefficient of variation placement Ratios from All Ground Motions from Site Classes A,
(COV), which is defined as the ratio of the standard deviation B, C, and D

JOURNAL OF STRUCTURAL ENGINEERING / OCTOBER 2000 / 1155

J. Struct. Eng. 2000.126:1150-1159.

 corresponding to the three groups of site conditions considered
herein. It can be seen that the results for the three groups are
very similar suggesting that, for sites with average shear-wave
velocities higher than 180 m/s in the upper 30 m of the site
profile, inelastic displacement ratios are not significantly af-
fected by local site conditions. Mean inelastic displacement
ratios computed from ground motions recorded at stations
classified as site class D are slightly smoother than those in
the other two groups because it includes a larger number of
ground motions.
To assess the errors that would be introduced by neglecting
the effects of site conditions on inelastic displacement ratios
for firms sites (site classes A, B, C, and D), ratios of mean C␮
on each group to mean C␮ computed from all 264 ground
Downloaded from ascelibrary.org by East Carolina University on 09/11/13. Copyright ASCE. For personal use only; all rights reserved.

motions were computed. Figs. 8(a), 8(b), and 8(c) show mean
inelastic displacement ratios for site classes A and B, site class
C, and site class D normalized by mean inelastic displacement
of the four site classes, respectively. It can be seen that for
site classes A, B, and C normalized mean inelastic displace-
ment ratios are, in general, smaller than 1. Thus, for these site

FIG. 6. Inelastic Displacement Ratios Corresponding to Dif-

ferent Percentiles

plus one standard deviation, as recently recommended (FEMA

1997a; Whittaker et al. 1998) is too conservative for displace-
ment ductilities equal or smaller than 6. Rather inelastic dis-
placement ratios associated to mean plus one standard devia-
tion in this spectral region should be specified as a function
of the level of inelastic deformation. For periods longer than
1.0 s the mean plus one standard deviation value of C␮ varies
from approximately 1.07 for a displacement ductility ratio of 1.5
to approximately 1.47 for a displacement ductility ratio of 6.
Another way to evaluate the scatter about mean results is to
compute inelastic displacement ratios corresponding to different
percentiles. Inelastic displacement ratios for a displacement duc-
tility ratio of 3 corresponding to percentiles of 10, 30, 50, 70,
and 90% are shown in Fig. 6(a). It can be seen that although
median inelastic displacement ratios (p = 50%) are approxi-
mately equal to 1 for periods longer than about 0.8 s, in 80%
of the cases inelastic displacement ratios will be between the
curves associated with 10 and 90%, which in this spectral region
implies that in 80% of the cases C␮ varies between approxi-
mately 0.65 and 1.35. Similarly, inelastic displacement ratios in
this spectral region in 40% of the cases (between p = 30 and
70%) are larger than 0.8 and smaller than 1.15. Inelastic dis-
placement ratios corresponding to the same percentiles but for
a displacement ductility ratio of 5 are shown in Fig. 6(b). In this
case it can be seen that for periods of vibration larger than about
1.0 s there is an 80% probability that maximum inelastic dis-
placement demand will be approximately between 0.6 and 1.5
times the maximum elastic displacement demand.

Effect of Soil Conditions

Most sites of new and existing structures will typically be
classified as firm sites (site classes A, B, C, and D). Thus, it
is important to quantify the differences of inelastic displace-
ment ratios computed from ground motions recorded in these FIG. 7. Mean Inelastic Displacement Ratios for Different Site
site classes. Fig. 7 shows mean inelastic displacement ratios Conditions

1156 / JOURNAL OF STRUCTURAL ENGINEERING / OCTOBER 2000

J. Struct. Eng. 2000.126:1150-1159.

 earthquake magnitude affects inelastic displacement ratios. To
study the effect of earthquake magnitude, mean inelastic dis-
placement ratios were computed from ground motions in three
groups. The first group comprises 68 ground motions recorded
on earthquakes with surface-wave magnitudes between 5.7 and
6.2. The second group includes 92 ground motions recorded
on earthquakes with magnitudes between 6.3 and 6.9, and the
third group contains 92 ground motions recorded on earth-
quakes with magnitudes between 7.0 and 7.8. A comparison
of mean inelastic displacement ratios in these three groups of
magnitudes for displacement ductility ratios equal to 2 and 4
are shown in Fig. 9. It can be seen that despite relatively large
changes in earthquake magnitude, changes in inelastic dis-
placement ratios are very small.
Downloaded from ascelibrary.org by East Carolina University on 09/11/13. Copyright ASCE. For personal use only; all rights reserved.

Effect of Distance to Rupture

To study the effects of distance to the horizontal projection
of the rupture on inelastic displacement ratios, mean inelastic
displacement ratios were computed from earthquake ground
motions in three groups with a different range of distances to
the rupture. The first group comprises 78 ground motions re-
corded at distances between 1 and 20 km from the horizontal
projection of the rupture. The second group includes 92
ground motions recorded at distances between 20 and 45 km
from the horizontal projection of the rupture, and the third
group contains 82 ground motions recorded at distances be-
tween 45 and 160 km. A comparison of mean inelastic dis-
placement ratios in these three groups of distances to the rup-
ture for displacement ductility ratios equal to 2 and 4 are
shown in Fig. 10. It can be seen that, for this range of dis-
tances, changes in mean inelastic displacement ratios are rel-

FIG. 8. Mean Inelastic Displacement Ratios on Each Group

Normalized by Mean Ratios from All Ground Motions: (a) Site
Classes A and B; (b) Site Class C; (c) Site Class D

classes neglecting the effect of local site conditions on C␮

would result in a small overestimation of inelastic displace-
ment demands. For site class D normalized mean inelastic dis-
placement ratios are larger than 1, which means that for this
site class neglecting the effect of local site conditions on C␮
would produce a small underestimation of inelastic displace-
ment demands. It can be seen that the difference in the elastic
displacement ratio produced by local site conditions increases
with increasing displacement ductility ratio. However, with the
exception of periods around 1.0 s, the errors produced by ne-
glecting the effect of local site conditions on C␮ are typically
smaller than 5% for a displacement ductility ratio of 2 and
smaller than 10% for a displacement ductility ratio of 6. For
practical design situations, these differences are relatively
small and can be neglected.
Effect of Earthquake Magnitude
Elastic spectral ordinates are dependent on the magnitude FIG. 9. Effect of Earthquake Magnitude on Mean Inelastic Dis-
of the earthquake. Thus, it is important to know to what extent placement Ratios

JOURNAL OF STRUCTURAL ENGINEERING / OCTOBER 2000 / 1157

J. Struct. Eng. 2000.126:1150-1159.

 tios computed with (2). In this figure it can be seen that this
equation yields correct limits to C␮ , namely, C␮ = ␮ as T →
0 and C␮ = 1 as T → ⬁. Additionally, (2) correctly addresses
the fact that the period limiting spectral regions, where the
equal displacement rule is applicable from those where it is
not, changes with the level of inelastic deformation. Further-
more, the proposed equation does not depend on any charac-
teristic site period, and thus it can be used for all sites with
average shear-wave velocities larger than 180 m/s without the
need of estimating a characteristic or corner period for the site.
Fig. 12 presents a comparison of mean inelastic displace-
ment ratios for periods <1.0 s with those computed with (2).
It can be seen that, despite its simplicity, this equation provides
very good results and is able to capture the effect of ␮ and T
Downloaded from ascelibrary.org by East Carolina University on 09/11/13. Copyright ASCE. For personal use only; all rights reserved.

FIG. 11. Inelastic Displacement Ratios Computed with Eq. (2)

FIG. 10. Effect of Nearest Distance to Horizontal Projection of
Rupture on Mean Inelastic Displacement Ratios

atively small. However, in a recent study using only near-field

records, Báez and Miranda (2000) concluded that inelastic dis-
placement ratios for periods between 0.1 an 1.3 s for near-
field records with forward directivity effects (i.e., those re-
corded in the horizontal component oriented perpendicular to
the fault strike and where rupture moves towards the site) can
be larger than those recorded farther away from the rupture or
those not affected by forward directivity.

NONLINEAR REGRESSION ANALYSES

For displacement-based design and, in general, in earth-
quake resistant design, it is desirable to have a simplified ex-
pression to estimate mean inelastic displacement ratios that
would then allow the estimation of maximum inelastic dis-
placement demands from maximum elastic displacement de-
mands. Using the Levenberg-Marquardt method, nonlinear re-
gression analyses were conducted to derive a simplified
expression for estimating mean inelastic displacement ratios.
This method combines the steepest-descent method and a Tay-
lor-series-based method to obtain a fast and reliable technique
for nonlinear optimization (Bevington and Robinson 1992).
Regression analyses were done using mean inelastic displace-
ment ratios corresponding to all earthquake ground motions.
The resulting equation is given by

冋 冉 冊 册
⫺1
1
C␮ = 1⫹ ⫺1 exp(⫺12T␮⫺0.8) (2)
␮
where ␮ = displacement ductility ratio; and T= period of vi-
bration. Eq. (2) represents a surface in the C␮-␮-T space and FIG. 12. Comparison of Inelastic Displacement Ratios in
provides estimates of mean inelastic displacement ratios as a Short-Period Spectral Region: (a) Mean Values of C␮ from Site
function of ␮ and T. Fig. 11 shows inelastic displacement ra- Classes A, B, C, and D; (b) C␮ Computed with Eq. (2)

1158 / JOURNAL OF STRUCTURAL ENGINEERING / OCTOBER 2000

J. Struct. Eng. 2000.126:1150-1159.

 on inelastic displacement ratios. The correlation coefficient be- it properly takes into account the fact that the period at which
tween the mean values and those computed with (2) is 0.979, mean inelastic displacement ratios become equal to 1 depends
and the standard error is 0.048. on the level of inelastic deformation.

CONCLUSIONS ACKNOWLEDGMENTS
The primary purpose of this investigation was to assess in- Special thanks are given to José Ignacio Báez and Francisco Garcı́a-
elastic displacement ratios that permit the estimation of max- Álvarez, former students at the National Autonomous University of Mex-
imum inelastic displacements from maximum elastic displace- ico and La Salle University, respectively, who conducted some of the
ments for structures on firm sites. A statistical study has been computer runs of which the results are presented in this paper. The writer
also gratefully acknowledges the suggestions of two anonymous review-
presented of inelastic displacement ratios computed for SDOF
ers of the manuscript.
systems undergoing different levels of inelastic deformation
when subjected to a large number of earthquake ground mo-
APPENDIX. REFERENCES
tions. The following conclusions can be drawn from the results
of this study. Applied Technology Council (ATC). (1996). ‘‘Improved seismic design
Downloaded from ascelibrary.org by East Carolina University on 09/11/13. Copyright ASCE. For personal use only; all rights reserved.

In the short-period spectral region, maximum inelastic dis- criteria for California bridges: Provisional recommendations.’’ Rep. No.
placements demands are, on average, larger than maximum ATC-32, Redwood City, Calif.
Báez, J. I., and Miranda, E. (2000). ‘‘Amplification factors to estimate
elastic demands. In this spectral region the ratio of maximum inelastic displacement demands for the design of structures in the near
inelastic to maximum elastic displacement demand depends on field.’’ Proc., 12th World Conf. on Earthquake Engrg.
the period of vibration and on the level of inelastic deforma- Bevington, P. R., and Robinson, D. K. (1992). Data reduction and error
tion. In the medium- and long-period spectral regions, the analysis for the physical sciences, McGraw-Hill, New York.
maximum inelastic displacement demands are, on average, Fajfar, P., and Fischinger, M. (1988). ‘‘N2—a method for nonlinear seis-
equal to maximum elastic demands for the range of displace- mic analysis of regular structures.’’ Proc., 9th World Conf. on Earth-
quake Engrg., Vol. 5, 111–116.
ment ductility ratios studied here. Periods at which mean in- Federal Emergency Management Agency (FEMA). (1997a). ‘‘NEHRP
elastic displacement ratios become equal to 1 depend on the guidelines for the seismic rehabilitation of buildings.’’ Rep. FEMA 273
level of inelastic deformation. These limiting periods increase (Guidelines) and Rep. 274 (Commentary), Washington, D.C.
with increasing displacement ductility ratios and range from Federal Emergency Management Agency (FEMA). (1997b). ‘‘NEHRP
0.2 s for a ductility ratio of 1.5 to 1.2 s for a ductility ra- recommended provisions for seismic regulations for new buildings and
tio of 6. other structures.’’ Rep. FEMA 302 (Provisions) and Rep. 303 (Com-
mentary), Washington, D.C.
The COVs of inelastic displacement ratios increase with in- Gupta, B., and Kunnath, S. K. (1998). ‘‘Effect of hysteretic model pa-
creasing level of inelastic deformation. With the exception of rameters on inelastic seismic demands.’’ Proc., 6th Natl. Conf. on
periods shorter than 0.15 s, COVs are approximately constant Earthquake Engrg., Earthquake Engineering Research Institute, Oak-
with changes in the period of vibration. For periods longer land, Calif.
than about 1.2 s, the mean plus one standard deviation value International Conference of Building Officials (ICBO). (1997). Uniform
of C␮ varies from approximately 1.07 for a displacement duc- building code, Whittier, Calif.
Miranda, E. (1991). ‘‘Seismic evaluation and upgrading of existing struc-
tility ratio of 1.5 to approximately 1.47 for a displacement tures.’’ PhD thesis, University of California, Berkeley, Berkeley, Calif.
ductility ratio of 6. Therefore, current recommendations of us- Miranda, E. (1993a). ‘‘Evaluation of site-dependent inelastic seismic de-
ing 150% of maximum elastic displacements as representative sign spectra.’’ J. Struct. Engrg., ASCE, 119(5), 1319–1338.
of mean plus one standard deviation maximum inelastic dis- Miranda, E. (1993b). ‘‘Evaluation of seismic design criteria for highway
placements associated in this spectral region are too conser- bridges.’’ Earthquake Spectra, 9(2), 233–250.
vative for structures undergoing displacement ductility de- Miranda, E. (1999). ‘‘Approximate seismic lateral deformation demands
in multistory buildings.’’ J. Struct. Engrg., ASCE, 125(4), 417–425.
mands smaller than 6. Rather, inelastic displacement ratios Qi, X., and Moehle, J. P. (1991). ‘‘Displacement design approach for
associated with mean plus one standard deviation should be reinforced concrete structures subjected to earthquakes.’’ Rep. No.
specified as a function of the level of inelastic deformation. EERC 91/02, Earthquake Engrg. Res. Ctr., University of California,
For sites with average shear-wave velocities higher than 180 Berkeley, Richmond, Calif.
m/s in the upper 30 m of the site profile, inelastic displacement Rahnama, M., and Krawinkler, H. (1993). ‘‘Effects of soils and hysteresis
ratios are not significantly affected by local site conditions. models on seismic design spectra.’’ Rep. No. 107, John A. Blume Earth-
quake Engrg. Ctr., Stanford University, Stanford, Calif.
Neglecting the effect of site conditions for firm sites will typ- Saiidi, M., and Sozen, M. A. (1981). ‘‘Simple nonlinear seismic analysis
ically result in errors <10% in the estimation of mean inelastic of R/C structures.’’ J. Struct. Div., ASCE, 107(5), 937–953.
displacement ratios. Similarly, inelastic displacement ratios are Seneviratna, G. D. P. K., and Krawinkler, H. (1997). ‘‘Evaluation of in-
not affected by changes in earthquake magnitude. With the elastic MDOF effects for seismic design.’’ Rep. No. 120, John A. Blume
exception of very near-field sites that may be influenced by Earthquake Engrg. Ctr., Stanford University, Stanford, Calif.
forward directivity effects, inelastic displacement ratios are not Shimazaki, K., and Sozen, M. A. (1984). ‘‘Seismic drift of reinforced
concrete structures.’’ Tech. Res. Rep. of Hazama-Gumi Ltd., Tokyo,
significantly affected by changes in the epicentral distance or 145–166.
the closest distance to the horizontal projection of the rupture, Veletsos, A. S., and Newmark, N. M. (1960). ‘‘Effect of inelastic behavior
thus making this parameter a particularly useful tool in earth- on the response of simple systems to earthquake motions.’’ Proc., 2nd
quake-resistant design. World Conf. on Earthquake Engrg., Vol. 2, 895–912.
Eq. (2), derived from nonlinear regression analyses, pro- Veletsos, A. S., Newmark, N. M., and Chepalati, C. V. (1965). ‘‘Defor-
duces very good approximations of mean inelastic displace- mation spectra for elastic and elastoplastic systems subjected to ground
shock and earthquake motion.’’ Proc., 3rd World Conf. on Earthquake
ment ratios of sites with average shear-wave velocities higher Engrg., Vol. II, 663–682.
than 180 m/s. This equation does not require the estimation Whittaker, A., Constantinou, M., and Tsopelas, P. (1998). ‘‘Displacement
of a characteristic or corner period for the site and is simple estimates for performance-based seismic design.’’ J. Struct. Engrg.,
enough to be used in practical design situations. Furthermore, ASCE, 124(8), 905–912.

JOURNAL OF STRUCTURAL ENGINEERING / OCTOBER 2000 / 1159

J. Struct. Eng. 2000.126:1150-1159.