EARTHQUAKE ENGINEERING AND STRUCTURAL DYNAMICS

Earthquake Engng Struct. Dyn. 2003; 32:1237–1258 (DOI: 10.1002/eqe.271)

Inelastic displacement ratios for evaluation of existing

structures

Jorge Ruiz-Garc
a and Eduardo Miranda∗; †

Department of Civil and Environmental Engineering; Stanford University; Stanford; CA 94305-4020; U.S.A.

SUMMARY
Results of a detailed statistical study of constant relative strength inelastic displacement ratios to
estimate maximum lateral inelastic displacement demands on existing structures from maximum lateral
elastic displacement demands are presented. These ratios were computed for single-degree-of-freedom
systems with dierent levels of lateral strength normalized to the strength required to remain elastic
when subjected to a relatively large ensemble of recorded earthquake ground motions. Three groups
of soil conditions with shear wave velocities higher than 180 m=s are considered. The inuence of
period of vibration, level of lateral yielding strength, site conditions, earthquake magnitude, distance
to the source, and strain-hardening ratio are evaluated and discussed. Mean inelastic displacement ra-
tios and those associated with various percentiles are presented. A special emphasis is given to the
dispersion of these ratios. It is concluded that distance to the source has a negligible inuence on
constant relative strength inelastic displacement ratios. However, for periods smaller than 1 s earth-
quake magnitude and soil conditions have a moderate inuence on these ratios. Strain hardening de-
creases maximum inelastic displacement at a fairly constant rate depending on the level of relative
strength for periods of vibration longer than about 1:0 s while it decreases maximum inelastic dis-
placement non-linearly as the period of vibration shortens and as the relative-strength ratio increases
for periods of vibration shorter than 1:0 s. Finally, results from non-linear regression analyses are pre-
sented that provide a simplied expression to be used to approximate mean inelastic displacement
ratios during the evaluation of existing structures built on rm sites. Copyright ? 2003 John Wiley &
Sons, Ltd.

KEY WORDS: inelastic displacement ratios; evaluation of existing structures; displacement-based

design; soils conditions; displacement demands: single-degree-of-freedom systems

∗ Correspondence to: Eduardo Miranda, Department of Civil and Environmental Engineering, Terman Engineering
Center, Stanford University, Stanford, CA 94305-4020, U.S.A.
† E-mail: [email protected]

Contract=grant sponsor: Consejo Nacional de Ciencia y Tecnolog
a

Contract=grant sponsor: Pacic Earthquake Engineering Research Center, Earthquake Engineering Research Centers
Program of the National Science Foundation; contract=grant number: EEC-9701568

Received 16 July 2001

Revised 2 July 2002 and 9 October 2002
Copyright ? 2003 John Wiley & Sons, Ltd. Accepted 9 October 2002
 10969845, 2003, 8, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/eqe.271 by Universidad Tecnica Federico, Wiley Online Library on [15/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
1238 J. RUIZ-GARC
IA AND E. MIRANDA

1. INTRODUCTION

Recently introduced displacement-based seismic design criteria use displacements rather than
forces as basic demand parameters for the design, evaluation and rehabilitation of structures.
However, implementation of displacement-based seismic design criteria into structural en-
gineering practice requires simplied analysis procedures to estimate inelastic displacement
demands on structures for ground motions in which the structure is expected to behave non-
linearly. This is particularly true when the design is based on design spectra rather than
on acceleration time histories. Recent recommendations for the evaluation and rehabilitation
of existing structures have introduced simplied analysis methods in which single-degree-
of-freedom (SDOF) systems are used to estimate global inelastic displacement demands on
structures [1–3]. Furthermore, in these new resource documents global inelastic displacement
demands of structures are computed taking into account the relationship between the maximum
inelastic displacement demands of non-linear SDOF systems and the maximum elastic dis-
placement demands of linear elastic SDOF systems. Thus, recently there has been a renewed
interest on approximate methods to estimate maximum displacement demands of inelastic
SDOF systems.
The rst study that investigated the relationship between the maximum deformations of
inelastic and of elastic systems was conducted by Veletsos and Newmark [4; 5] who studied
SDOF systems with an elasto-plastic hysteretic behaviour subjected to simple pulses and to
three recorded earthquake ground motions. They observed that in the low frequency region the
maximum deformation of the inelastic and elastic systems was approximately the same. This
observation gave rise to the well-known equal displacement rule. The study also concluded
that, in the high frequency region, the inelastic displacements are signicantly higher than
their elastic counterparts. In a later study [6], the equal displacement rule was also recom-
mended for the medium frequency region. They also noted that the width of each frequency
region is generally dierent for dierent ground motions. Shimazaki and Sozen [7] noticed
that, in the short period region the ratio of maximum inelastic displacement to maximum
elastic displacement depended critically on the lateral strength of the structure relative to the
elastic strength demand and that the estimate of the inelastic displacement demand was be-
yond a simple procedure. These conclusions were also more recently conrmed by Qi and
Moehle [8].
Miranda [9; 10] studied ratios of maximum inelastic displacement to maximum elastic
displacement of SDOF systems undergoing specic levels of displacement ductility when
subjected to 124 earthquake ground motions recorded on dierent site conditions. Mean
constant-ductility ratios of maximum inelastic to maximum elastic response for three types
of soil conditions were computed as part of the investigation. The study gave a special in-
sight to this ratio in the short period range and to the limiting periods of the spectral regions
where the equal displacement rule is applicable. The results of Miranda were conrmed by
Krawinkler and his co-workers at Stanford University [11; 12] using smaller sets of ground
motions.
In order to evaluate the non-linear static procedure of the FEMA guidelines for the seismic
rehabilitation of buildings [2] Whittaker et al. [13] also conducted a study of the ratio of inelas-
tic to elastic displacements. Results of the ratio of mean inelastic displacements to mean elastic
displacements were presented corresponding to 20 horizontal components of 10 ground mo-
tions recorded on either sti soil or soft rock sites. The study concluded that for periods smaller

Copyright ? 2003 John Wiley & Sons, Ltd. Earthquake Engng Struct. Dyn. 2003; 32:1237–1258
 10969845, 2003, 8, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/eqe.271 by Universidad Tecnica Federico, Wiley Online Library on [15/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
INELASTIC DISPLACEMENT RATIOS FOR EVALUATION OF EXISTING STRUCTURES 1239

than about 1s mean inelastic displacements exceed mean elastic displacements. Furthermore, it
was concluded that for systems with lateral strengths smaller than 20% of the strength required
to maintain the system elastic, mean inelastic displacements systematically exceed mean elas-
tic displacements, suggesting that the equal displacement rule may not be applicable in these
situations.
More recently, Miranda [14] presented a comprehensive statistical study of constant ductil-
ity inelastic displacement ratios for the design of structures on rm sites. This study provided
new information regarding the dispersion of this ratio and regarding the inuence of period,
level of inelastic deformation, magnitude, distance to the source and local site conditions.
Miranda concluded that for SDOF systems undergoing the same displacement ductility ratio,
inelastic displacement ratios were not aected by magnitude or by distance to the source.
Furthermore, the study concluded that for sites with average shear wave velocities higher
than 180 m=s (600 ft=s) in the upper 30 m (100 ft) of the site prole, local site condi-
tions do not aect signicantly constant-ductility inelastic displacement ratios. As part of
the study, a simplied equation to estimate ratios of maximum inelastic to maximum elastic
displacement as a function of period of vibration and of displacement ductility ratio was also
developed.
The simplied equation developed by Miranda [14] to estimate constant-ductility inelas-
tic displacement ratios is very useful in the preliminary design of new structures where
control of maximum inelastic deformations is desired for structures where an estimate of
the global displacement ductility capacity is known. However, in the evaluation of exist-
ing structures the main interest is to determine the global and local deformations that a
structure with known lateral strength may undergo when subjected to earthquakes of dif-
ferent intensities. As shown in the next section, the use of constant-ductility inelastic dis-
placement ratios underestimates the expected value of the maximum deformations in systems
with known lateral strength. Hence, inelastic displacement ratios corresponding to systems
with equal relative lateral strength (lateral yielding strength relative to the lateral strength
required to maintain the system elastic) are particularly useful when evaluating existing
structures.
While the studies of Shimazaki and Sozen [7] and Whittaker et al. [13] provided valu-
able information regarding constant relative strength inelastic displacement ratios of SDOF
systems, none of the two provided expressions of the ratio of maximum inelastic displace-
ment to elastic displacement that could be used during the evaluation of existing struc-
tures nor provided information on the dispersion of this ratio. Furthermore, none of the two
studies investigated the eect of earthquake magnitude, distance to the source or local site
conditions.
More recently Chopra and Goel [15; 16] and Fajfar [17; 18] proposed to estimate the max-
imum inelastic displacement of existing structures with known lateral strength, by multiplying
the yielding displacement by a displacement ductility ratio,
, computed from existing R
–
–T
relations, which typically provide an estimate of the mean strength reduction factor R
as
a function of the displacement ductility ratio and the period of vibration T . While this
approach is very simple and provides a way to estimate the maximum inelastic displace-
ment, Miranda [19] has shown that the ductility demand computed from R
–
–T relations
does not correspond to the mean displacement ductility demand of system with relative
strength equal to R
, and that the computed displacement ductility ratio is the rst ap-
proximation of the mean displacement ductility demand, hence this procedure introduces a

Copyright ? 2003 John Wiley & Sons, Ltd. Earthquake Engng Struct. Dyn. 2003; 32:1237–1258
 10969845, 2003, 8, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/eqe.271 by Universidad Tecnica Federico, Wiley Online Library on [15/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
1240 J. RUIZ-GARC
IA AND E. MIRANDA

systematic error that will tend to underestimate the maximum inelastic displacement. The
error will typically increase with increasing ductility, although the actual size of the er-
ror depends on the particular R
–
–T relation that is used. For the evaluation of existing
structures more accurate estimates of maximum inelastic displacements of systems where
the relative strength is known can be achieved by using directly results from statistical
studies.
The objective of this paper is to present the results of a statistical study of the ratio of
maximum inelastic displacement demand to maximum elastic displacement demand for SDOF
systems on rm sites with known relative strength. The eects of period of vibration, level
of lateral strength, earthquake magnitude, distance to the source and local site conditions are
investigated. The dispersion of constant relative strength inelastic displacement ratios is as-
sessed. This study makes use of improved information that has been made available recently
on the geological characteristics at accelerographic recording stations in California. The in-
vestigation is limited to rock and relatively rm soil sites with shear wave velocities higher
than 180 m=s in the upper 30 m of the site prole.

2. INELASTIC DISPLACEMENT RATIOS

The inelastic displacement ratio, CR , is dened as the maximum lateral inelastic displacement
demand, inelastic , divided by the maximum lateral elastic displacement demand, elastic , on
systems with the same mass and initial stiness (i.e. same period of vibration) when subjected
to the same earthquake ground motion. In both cases displacements are relative to the ground.
Mathematically this is expressed as
inelastic
CR = (1)
elastic
In Equation (1) the inelastic is computed in systems with constant yielding strength relative
to the strength required to maintain the system elastic (i.e. constant relative strength). Here
the relative lateral strength is measured by the strength ratio R, which is dened as
mSa
R= (2)
Fy
where m is the mass of the system, Sa is the acceleration spectral ordinate and Fy is the
lateral yielding strength of the system. The numerator in Equation (2) represents the lateral
strength required to maintain the system elastic, which sometimes is also referred to as the
elastic strength demand.
The nomenclature in Equation (1) is meant to be consistent with the nomenclature used
in NEHRP publications [2; 3] in which the letter C is used as a factor modifying elastic
displacements and is also consistent with the nomenclature previously used by Miranda [14]
in which the subscript in the inelastic displacement ratio represents the parameter that remains
constant. Thus, constant ductility inelastic displacement ratios are represented by C
, and con-
stant relative strength (or constant strength ratio) inelastic displacement ratios are represented
by CR . Both types of inelastic displacement ratios permit the estimation of maximum inelastic
displacement demands from maximum elastic displacement demands.

Copyright ? 2003 John Wiley & Sons, Ltd. Earthquake Engng Struct. Dyn. 2003; 32:1237–1258
 10969845, 2003, 8, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/eqe.271 by Universidad Tecnica Federico, Wiley Online Library on [15/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
INELASTIC DISPLACEMENT RATIOS FOR EVALUATION OF EXISTING STRUCTURES 1241

Figure 1. Relationship between normalized strength (1=R) and displacement ductility demand. (a) For
a T = 2:0 s system, (b) for a T = 0:2 s system.

Inelastic displacement ratios were computed for SDOF systems having a viscous damp-
ing ratio of 5%, a non-linear elasto-plastic hysteretic behaviour, and with the following
strength ratios R = 1, 1.5, 2, 3, 4, 5 and 6. For each earthquake record and each relative
strength, the inelastic displacement ratios were computed for a set of 50 periods of vibra-
tion between 0.05 and 3:0 s. Unlike the constant ductility inelastic displacement ratio C

that has to be computed through iteration on the lateral strength until the computed dis-
placement ductility demand is within a certain tolerance equal to the target ductility ra-
tio, the constant relative strength inelastic displacement ratio CR can be computed without
any iteration and thus, for a given acceleration time history, it is signicantly faster to
compute.
Figure 1 shows the relationship between the lateral strength of SDOF systems with a
period of vibration of 1:0 s, and the maximum displacement when subjected to 264 earthquake
ground motions recorded on rm sites [14]. In this gure, the lateral strength is normalized
by the lateral strength required to maintain the system elastic (i.e. Fy =(mSa ) = 1=R), and the
maximum deformation is normalized by the yield displacement. The continuous line represents
the relationship between the lateral strength and the expected value of the ductility demand
computed using the expected value of CR as follows:
E[
] = R · E[CR ] (3)
where
is the displacement ductility ratio and E[ ] denotes expectation. The dotted line
represents the relationship between the displacement ductility ratio and the expected value of
the normalized lateral strength computed for
= 1, 1.5, 2, 3, 4, 5 and 6 using the expected
value of C
as follows:


1 E[C
]
E = (4)
R

It can be seen that, in this case, the use of E[C
] using Equation (4) leads to very
similar results as when using E[CR ] in Equation (3). However, it can be noted that for

Copyright ? 2003 John Wiley & Sons, Ltd. Earthquake Engng Struct. Dyn. 2003; 32:1237–1258
 10969845, 2003, 8, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/eqe.271 by Universidad Tecnica Federico, Wiley Online Library on [15/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
1242 J. RUIZ-GARC
IA AND E. MIRANDA

a given R the ductility demand computed with constant ductility inelastic displacement ra-
tios is always smaller than the one computed with constant relative strength inelastic dis-
placement ratios. Furthermore, the dierence increases as R and
increase. The practical
implication of this observation is that the use of constant ductility inelastic displacement
ratios like those reported by Miranda [14] if used for the evaluation of existing struc-
tures, in which R is known for a given ground motion, can lead to underestimations of
the maximum inelastic displacement demands. The underestimation increases as the disper-
sion on C
and CR increases. For short period structures, where the dispersion on CR is
large, the underestimation can be much larger. An example similar to that shown in Fig-
ure 1(a) but for a period of 0:2 s is shown in Figure 1(b). In this case, for R = 3 the
underestimation in maximum displacement can be larger than 40%. Hence it is clear that
for the evaluation of existing structures there is a need for statistical studies on CR . Only
when there is no dispersion on C
and CR , as for example in the case of R = 1 or for a
single ground motion (for any level of R) do Equations (3) and (4) lead to the same dis-
placement. For a further discussion on the reason for this dierence the reader is referred to
Miranda [19].

3. EARTHQUAKE GROUND MOTIONS USED IN THE STUDY

A total of 216 earthquake acceleration time histories recorded in the state of California in
12 dierent earthquakes with magnitude ranging from 5.8 to 7.7 were used in this study.
A particularly large number of earthquake ground motions was selected in order to assess
the dispersion of the inelastic displacement ratios and in order to be able to obtain inelastic
displacement ratios corresponding to dierent percentiles. All the ground motions selected
have the following characteristics: (i) recorded on accelerographic stations where enough
information exists on the geological and geotechnical conditions at the site that enables the
classication of the recording site in accordance to recent code recommendations [2; 3; 20];
(ii) recorded on rock or rm sites with average shear wave velocities higher than 180 m=s
(600 ft=s) in the upper 30 m (100 ft) of the site prole; (iii) recorded on free eld stations
or in the rst oor of low-rise buildings with negligible soil–structure interaction eects; (iv)
recorded in earthquakes with surface wave magnitudes (Ms ) larger than 5.7; and (v) records
in which at least one of the two horizontal components had a peak ground acceleration larger
than 40 cm=s 2 .
The earthquake ground motions were divided into three groups according to the local site
conditions at the recording station. The rst group consisted of 72 ground motions recorded
on stations located on rock with average shear wave velocities between 760 m=s (2500 ft=s)
and 1525 m=s (5000 ft=s). The second group consisted of 72 records obtained on stations on
very dense soil or soft rock with average shear wave velocities between 360 m=s (1200 ft=s)
and 760 m=s while the third group consisted of 72 ground motions recorded on stations
on sti soil with average shear wave velocities between 180 m=s (600 ft=s) and 360 m=s.
Recording stations in the rst group correspond to site class B according to recent design
provisions [2; 3; 20] while recording stations in the second and third groups correspond to
site classes C and D, respectively. For a complete list of all ground motions including peak
ground accelerations, earthquake magnitude, site class at the recording station, and distance
to the horizontal projection of the fault rupture see Tables I–III.

Copyright ? 2003 John Wiley & Sons, Ltd. Earthquake Engng Struct. Dyn. 2003; 32:1237–1258
 Copyright ? 2003 John Wiley & Sons, Ltd.
Table I. Earthquake ground motions recorded on NEHRP site class B used in this study.
Earthquake name Magnitude Station name Station Distance Comp. 1 PGA PGV Comp. 2 PGA PGV
(Ms) number (km) (deg) (cm=s 2 ) (cm=s) (deg) (cm=s 2 ) (cm=s)

INELASTIC DISPLACEMENT RATIOS FOR EVALUATION OF EXISTING STRUCTURES

1971 San Fernando 6.5 Lake Hughes, Array Station 4 126 19.6 111 168.2 5.7 201 143.5 7.1
1971 San Fernando 6.5 Lake Hughes, Array Station 9 127 23.0 21 119.3 4.5 291 109.4 3.9
1979 Imperial Valley 6.8 Superstition Mountain 286 26.0 135 189.2 6.3 45 108.0 5.1
1984 Morgan Hill 6.1 Gilroy 1, Gavillan Coll. 47 379 16.0 230 57.5 2.9 320 93.4 2.9
1986 Palm Springs 6.0 Silent Valley, Poppet Flat 12 206 23.7 0 102.6 3.9 90 107.4 4.0
1986 Palm Springs 6.0 Winchester, Hidden Valley Farms 13 200 49.8 90 54.6 1.3 270 56.5 1.3
1986 Palm Springs 6.0 Winchester, Bergman Ranch 13 199 55.3 0 62.2 1.9 90 85.7 1.8
1986 Palm Springs 6.0 Murrieta Hot Springs, Colling Ranch 13 198 61.0 0 45.9 1.8 90 49.4 1.3
1986 Palm Springs 6.0 Whitewater Trout Farm 5072 7.3 180 482.7 34.7 270 600.4 31.5
1986 Palm Springs 6.0 Anza Red Mountain 5224 45.6 270 102.0 5.2 360 126.5 3.4
1987 Whittier 6.1 Mt Wilson, CIT Seismic Station 24 399 22.1 0 121.3 3.3 90 171.3 4.6
1987 Whittier 6.1 Los Angeles, Gritth Park Observatory 141 22.3 0 133.8 3.6 360 121.4 4.1
1989 Loma Prieta 7.1 Gilroy 1, Gavillan Coll. 47 379 10.5 90 433.6 33.9 360 426.6 31.6
1989 Loma Prieta 7.1 Hollister, SAGO south cinega road surface 47 189 32.4 261 70.7 10.3 351 65.3 9.3
1989 Loma Prieta 7.1 Monterey, City Hall 47 377 42.7 90 61.1 5.8 360 68.5 3.5
1989 Loma Prieta 7.1 South San Francisco, Sierra Point 58 539 67.6 115 57.2 7.1 205 102.7 8.8
1989 Loma Prieta 7.1 San Francisco, Dimond Heights 58 130 75.9 90 110.8 14.3 360 96.4 10.5
1989 Loma Prieta 7.1 Piedmont, Piedmont Jr. High Grounds 58 338 77.2 45 81.2 8.2 315 69.6 9.1
Earthquake Engng Struct. Dyn. 2003; 32:1237–1258

1989 Loma Prieta 7.1 San Francisco, Rincon Hill 58 151 78.5 90 88.5 10.4 0 78.6 6.7
1989 Loma Prieta 7.1 San Francisco, Pacic Heights 58 131 80.5 270 60.2 12.8 360 46.3 9.2
1989 Loma Prieta 7.1 San Francisco, Cli House 58 132 87.4 0 73.1 11.2 90 105.7 21.0
1989 Loma Prieta 7.1 San Francisco, Telegraph Hill 58 133 88.0 90 51.2 5.5 0 90.5 9.5
1989 Loma Prieta 7.1 Point Bonita 58 043 88.1 297 71.4 12.9 207 69.9 11.4
1991 Sierra Madre 5.8 Mt Wilson, CIT Seismic Station 24 399 5.3 0 270.7 13.0 90 196.2 7.5
1992 Landers 7.5 Twentynine Palms Park Maintenance Bldg 22 161 41.9 0 78.7 3.7 90 59.1 4.9
1992 Landers 7.5 Silent Valley, Poppet Flat 12 206 51.3 90 39.4 5.1 0 48.9 3.8
1992 Landers 7.5 Amboy 21 081 68.3 0 88.3 11.0 90 143.2 20.0
1994 Northridge 6.8 Malibu Canyon, Grith Observatory 5080 20.2 360 176.4 12.3 270 270.0 17.7
1994 Northridge 6.8 Lake Hughes, Array Station 9 24 272 28.4 90 221.2 10.1 360 154.5 8.4
1994 Northridge 6.8 Los Angeles, Temple & Hope 24 611 32.2 180 189.1 20.0 90 123.7 13.9
1994 Northridge 6.8 Lake Hughes, Array Station 4 24 469 34.0 0 56.4 5.1 90 82.4 5.1
1994 Northridge 6.8 Mt Wilson, CIT Seismic Station 24 399 36.9 90 130.7 5.8 360 228.5 7.4
1994 Northridge 6.8 Los Angeles, City Terrace 24 592 37.1 90 258.0 12.8 0 310.1 14.1
1994 Northridge 6.8 Antelope Buttes 24 310 48.6 90 99.7 4.3 0 44.9 3.6
1994 Northridge 6.8 San Pedro, Palos Verdes 14 159 58.5 90 93.1 6.6 0 98.9 5.6

1243
1994 Northridge 6.8 Leona Valley #3 24 307 37.8 0 82.4 8.5 90 104.0 8.1

10969845, 2003, 8, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/eqe.271 by Universidad Tecnica Federico, Wiley Online Library on [15/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
 Copyright ? 2003 John Wiley & Sons, Ltd.

1244
Table II. Earthquake ground motions recorded on NEHRP site class C used in this study.
Earthquake name Magnitude Station name Station Distance Comp. 1 PGA PGV Comp. 2 PGA PGV
(Ms) number (km) (deg) (cm=s 2 ) (cm=s) (deg) (cm=s 2 ) (cm=s)

1952 Kern County 7.7 Santa Barbara, Courthouse 283 85.0 42 87.8 12.1 132 128.6 15.5
1952 Kern County 7.7 Pasadena, CIT Athenaeum 475 109.0 180 46.5 5.6 270 52.1 9.2
1971 San Fernando 6.5 Lake Hughes, Array Station 12 128 17.0 21 346.2 17.0 291 277.9 12.7
1971 San Fernando 6.5 Glendale, 633 E, Broadway 122 18.0 110 265.7 10.3 200 209.1 13.4
1971 San Fernando 6.5 Lake Hughes #1, Fire Station #78 125 25.0 21 145.5 18.0 111 108.9 11.7
1971 San Fernando 6.5 Castaic Old Ridge Route 110 26.0 21 309.4 15.6 291 265.4 25.9
1971 San Fernando 6.5 Pearblossom Pump Plant 585 36.0 0 91.5 4.7 270 120.5 5.6
1979 Imperial Valley 6.8 El Centro, Parachute Test Facillity 5051 14.0 225 106.9 17.8 315 200.2 16.1
1984 Morgan Hill 6.1 Gilroy #6, San Ysidro Microwave Site 57 383 11.5 0 214.8 11.4 90 280.4 36.7

J. RUIZ-GARCIA
1984 Morgan Hill 6.1 Gilroy Gavillan College Phys Sch. Bldg 47 006 16.0 337 85.9 2.9 67 95.0 3.6
1987 Whittier 6.1 Garvey Reservoir Abutment Bldg 709 3.4 60 367.1 15.8 330 468.2 19.0
1987 Whittier 6.1 Alhambra, 900 S. Fremont 24 461 6.7 180 286.2 22.0 270 374.3 16.3
1987 Whittier 6.1 San Marino, SW Academy 24 401 7.7 270 136.5 4.8 360 183.8 12.9
1987 Whittier 6.1 Inglewood, Union Oil Yard 14 196 22.5 0 246.1 18.1 90 219.3 8.9
1987 Whittier 6.1 Long Beach, Recreation Park 14 241 29.6 90 57.2 3.5 180 53.8 5.5

AND E. MIRANDA
1987 Whittier 6.1 Sylmar, Olive View Medical Center 24 514 45.9 0 55.8 4.0 90 50.3 3.5
1987 Whittier 6.1 Riverside, Airport 13 123 57.8 180 38.4 1.4 270 56.8 1.4
1987 Whittier 6.1 Lancaster, Medical Oce Bldg FF 24 526 72.2 10 59.4 2.9 100 59.6 3.0
Earthquake Engng Struct. Dyn. 2003; 32:1237–1258

1987 Whittier 6.1 Castaic, Old Ridge Route 24 278 77.3 0 67.2 4.4 90 65.4 4.5
1989 Loma Prieta 7.1 Corralitos, Eureka Canyon Road 57 007 0.0 90 469.4 47.5 360 617.7 55.2
1989 Loma Prieta 7.1 Gilroy, Gavillan College Phys Sch Bldg 47 006 10.9 67 349.1 28.9 337 310.0 23.0
1989 Loma Prieta 7.1 Saratoga, Aloha Ave. 58 065 12.4 90 316.2 43.5 0 494.5 41.3
1989 Loma Prieta 7.1 Santa Cruz, UCSC 58 135 12.5 90 401.5 21.2 360 433.1 21.2
1989 Loma Prieta 7.1 Gilroy 6, San Ysidro Microwave site 57 383 19.9 90 166.9 14.2 0 112.2 12.8
1989 Loma Prieta 7.1 Coyote Lake Dam, downstream 57 504 21.7 285 174.7 22.6 195 154.7 13.0
1989 Loma Prieta 7.1 Woodside, Fire Station 58 127 38.7 90 79.7 14.7 0 79.5 15.5
1989 Loma Prieta 7.1 Fremont, Mission San Jose 57 064 42.6 90 100.5 8.8 0 117.7 11.5
1989 Loma Prieta 7.1 Hayward, CSUH Stadium 58 219 56.7 90 82.6 6.4 0 72.5 5.6
1989 Loma Prieta 7.1 Berkeley, Lawrence Berkeley Lab. 58 471 83.9 90 114.4 20.9 0 47.7 9.2
1992 Landers 7.5 Joshua Tree, Fire Station 22 170 7.1 0 268.3 27.5 90 278.4 43.2
1994 Northridge 6.8 Castaic Old Ridge Route 24 278 24.6 360 504.2 52.2 90 557.1 52.1
1994 Northridge 6.8 San Marino, SW Academy 24 401 35.5 360 148.2 5.4 90 122.5 7.3
1994 Northridge 6.8 Alhambra, 900 S. Fremont 24 461 37.2 360 78.3 12.7 90 99.1 9.7
1994 Northridge 6.8 Lake Hughes #1, Fire Station #78 24 271 37.7 0 84.9 9.2 90 75.2 9.4
1994 Northridge 6.8 Littlerock, Brainard Canyon 23 595 47.9 90 70.6 5.0 180 59.0 5.3
1994 Northridge 6.8 Rancho Palos Verdes, Hawthorne Blvd. 14 404 53.8 0 71.1 5.0 90 52.7 3.3

10969845, 2003, 8, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/eqe.271 by Universidad Tecnica Federico, Wiley Online Library on [15/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
 Copyright ? 2003 John Wiley & Sons, Ltd.
Table III. Earthquake ground motions recorded on NEHRP site class D used in this study.
Earthquake name Magnitude Station name Station Distance Comp. 1 PGA PGV Comp. 2 PGA PGV
(Ms) number (km) (deg) (cm=s 2 ) (cm=s) (deg) (cm=s 2 ) (cm=s)

INELASTIC DISPLACEMENT RATIOS FOR EVALUATION OF EXISTING STRUCTURES

1952 Kern County 7.7 Los Angeles, Hollywood Storage PE Lot 135 107.0 90 41.2 6.0 180 58.1 5.3
1968 Borrego Mtn 6.7 El Centro, Imperial Valley Irrigation District 117 45.0 180 127.8 26.3 270 56.3 13.2
1971 San Fernando 6.5 Los Angeles, Hollywood Storage Bldg 135 23.0 90 207.0 18.9 180 167.3 14.9
1971 San Fernando 6.5 Vernon, Cmd Terminal 288 33.5 187 80.5 6.4 277 104.6 9.8
1971 San Fernando 6.5 Santa Ana, Engineering Bldg 281 71.5 176 26.8 2.9 266 28.2 3.1
1979 Imperial Valley 6.8 Calexico, Fire Station 5053 10.6 225 269.6 21.2 315 196.9 16.0
1979 Imperial Valley 6.8 El Centro #11, McCabe Union School 5058 12.6 140 355.4 24.7 230 374.5 29.8
1979 Imperial Valley 6.8 El Centro #3, Pine Union School 5057 12.7 140 261.7 46.8 230 218.1 39.9
1979 Imperial Valley 6.8 El Centro #12, 907 Brockman Road 931 18.0 140 138.7 11.1 230 113.4 10.7
1979 Imperial Valley 6.8 Plaster City, Storehouse 5052 32.0 135 55.5 4.2 45 41.9 3.8
1979 Imperial Valley 6.8 Coachella, Canal #4 5066 49.0 45 113.6 12.5 135 125.7 15.6
1984 Morgan Hill 6.1 Gilroy #2, Hwy 101/Bolsa Road Motel 47 380 1.0 0 153.7 5.1 90 210.0 12.6
1984 Morgan Hill 6.1 Gilroy #7, Mantnilli Ranch, Jamison Rd 57 425 13.7 0 183.0 7.4 90 111.5 6.0
1984 Morgan Hill 6.1 Gilroy #3, Sewage Treatment Plant 47 381 14.4 0 177.0 11.2 90 189.8 12.7
1987 Whittier 6.1 Bell Los Angeles Bulk Mail Center 5129 10.6 10 322.1 31.1 280 436.9 39.7
1987 Whittier 6.1 Vernon, Cmd Terminal 288 11.1 7 267.3 25.4 277 239.9 22.9
1987 Whittier 6.1 Downey, County Maintenance Bldg 14 368 16.2 180 193.2 28.8 270 150.7 13.4
1987 Whittier 6.1 Los Angeles, Hollywood Storage Bldg 24 303 23.8 0 201.3 9.0 90 103.7 6.9
Earthquake Engng Struct. Dyn. 2003; 32:1237–1258

1987 Whittier 6.1 Century City, LA Country Club South 24 390 29.6 0 57.6 3.7 90 67.2 4.2
1987 Whittier 6.1 Pomona, 4th and Locust FF 23 525 29.9 12 68.4 2.4 102 49.0 2.3
1987 Whittier 6.1 Long Beach, Harbor Administration Bldg 14 395 32.8 0 48.2 5.5 90 68.9 3.6
1987 Whittier 6.1 Rancho Cucamonga, Law and Justice Center 23 497 45.5 90 55.5 1.4 360 45.3 1.4
1987 Whittier 6.1 Arleta, Nordho Avenue Fire Station 24 087 45.7 180 87.1 4.8 270 84.2 5.7
1987 Whittier 6.1 Rosamond, Goode Ranch 24 274 89.0 0 73.8 3.5 90 50.4 3.1
1989 Loma Prieta 7.1 Gilroy 2, Hwy 101 Bolsa Road Motel 47 380 12.1 90 316.3 39.1 0 394.2 32.9
1989 Loma Prieta 7.1 Gilroy 3, Sewage Treatment Plant 47 381 14.0 90 362.0 44.7 0 531.7 35.7
1989 Loma Prieta 7.1 Agnews, Agnews State Hospital 57 066 27.0 90 157.6 17.6 0 163.1 26.0
1989 Loma Prieta 7.1 Hayward, John Muir School 58 393 58.9 90 136.0 11.5 0 166.5 13.7
1989 Loma Prieta 7.1 Oakland, 2 story 58 224 76.3 290 238.3 36.1 200 187.3 19.9
1989 Loma Prieta 7.1 Richmond, City Hall parking lot 58 505 92.7 280 103.6 14.2 190 122.7 17.3
1992 Landers 7.5 Yermo, Fire Station 22 074 26.3 270 240.0 51.5 360 148.6 29.7
1992 Landers 7.5 Palm Springs, Airport 12 025 28.2 0 74.2 10.9 90 87.2 13.8
1992 Landers 7.5 Fort Irwin 24 577 65.5 0 111.4 9.7 90 119.8 16.4
1992 Landers 7.5 Baker, Fire Station 32 075 88.3 50 105.6 9.4 140 103.6 11.0
1992 Landers 7.5 Pomona, 4th and Locust FF 23 525 117.6 0 65.5 12.3 90 43.2 8.5

1245
1994 Northridge 6.8 Los Angeles, Hollywood Storage Bldg 24 303 24.8 360 381.4 22.3 90 227.0 18.1

10969845, 2003, 8, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/eqe.271 by Universidad Tecnica Federico, Wiley Online Library on [15/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
 10969845, 2003, 8, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/eqe.271 by Universidad Tecnica Federico, Wiley Online Library on [15/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
1246 J. RUIZ-GARC
IA AND E. MIRANDA

4. STATISTICAL RESULTS

4.1. Mean ratios for di
erent local site conditions

A total of 64,800 inelastic displacement ratios were computed as part of this study (corre-
sponding to 216 ground motions, 50 periods of vibration and 6 levels of relative strength).
Mean inelastic displacement ratios were then computed by averaging results for each period,
each strength ratio and each group of local site conditions at the recording station. For the
site classes considered here current seismic design provisions in the US [3] specify linear
elastic design spectra that are signicantly dierent from each other. Thus, it is particularly
important to know if inelastic displacement ratios to be used for estimating maximum inelastic
displacements from maximum elastic displacements while evaluating existing structures are
aected by local site conditions, and if so, to quantify these dierences.
Figure 2 shows mean inelastic displacement ratios corresponding to the three groups of site
conditions considered here. It can be seen that, in general, constant relative strength exhibits
the same trend regardless of the local site conditions. These ratios are characterized by being
larger than 1 in the short period spectral region (maximum inelastic displacements larger than
maximum elastic displacements) and relatively close to 1 (maximum inelastic displacements
on average approximately equal to maximum elastic displacement) for long periods. For
periods smaller than 1:0 s inelastic displacement ratios are strongly dependent on the period of
vibration and on the lateral strength ratio. In general, in this spectral region maximum inelastic
displacements become much larger than maximum elastic displacements as the lateral strength
ratio increases (i.e. as the lateral strength decreases with respect to the lateral strength required
to maintain the system elastic) and as the period of vibration decreases. Furthermore, constant
relative strength inelastic displacement ratios tend towards ∞ as the period of vibration tends
to zero, which means that existing structures with very short periods may undergo very large
inelastic displacement demands relative to their elastic counterparts unless they have lateral
strengths that allow them to remain elastic or nearly elastic. It is important to notice that the
limiting period dividing spectral regions where the equal displacement rule is applicable from
those where this rule is not applicable and is unconservative (and produces an underestimation
of the maximum lateral displacement demand) depends primarily on the lateral strength ratio,
although it is also inuenced by local site conditions. In general, this limiting period increases
as lateral strength ratio increases and as the average shear wave velocity in the upper 30 m of
the site prole decreases (i.e. as site conditions become softer). For example, for structures
on a site class C the equal displacement rule is applicable on average for periods longer than
about 0:4 s for a lateral strength ratio of 1.5 but the rule is only approximately correct for
periods longer than about 1:0 s for lateral strength ratios of 6. Similarly, for a lateral strength
ratio of 2, the equal displacement rule is approximately correct for periods longer than about
0.45, 0.65 and 0:8 s for structures on site classes B, C and D, respectively. It can be seen
that in the short period spectral region CR increases as the average shear wave velocity in the
upper 30 m of the site prole decreases.

4.2. Mean and median ratios for all site classes

Figure 3 shows mean constant-ductility inelastic displacement ratios corresponding to all 216
ground motions, regardless of the site conditions at the recording station. Ratios shown in

Copyright ? 2003 John Wiley & Sons, Ltd. Earthquake Engng Struct. Dyn. 2003; 32:1237–1258
 10969845, 2003, 8, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/eqe.271 by Universidad Tecnica Federico, Wiley Online Library on [15/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
INELASTIC DISPLACEMENT RATIOS FOR EVALUATION OF EXISTING STRUCTURES 1247

Figure 2. Mean inelastic displacement ratios for NEHRP site classes B, C and D.

this gure represent what, on average, can be expected for existing structures built on rm
sites. Again, it can be seen that in the short period region the ratio of inelastic to elastic
displacement demand is strongly dependent on the relative strength of the system. Furthermore,
in this period region the equal displacement rule can result in signicant underestimations of
the maximum inelastic displacement demand.

Copyright ? 2003 John Wiley & Sons, Ltd. Earthquake Engng Struct. Dyn. 2003; 32:1237–1258
 10969845, 2003, 8, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/eqe.271 by Universidad Tecnica Federico, Wiley Online Library on [15/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
1248 J. RUIZ-GARC
IA AND E. MIRANDA

Figure 3. Mean inelastic displacement ratios for all 216 ground motions
recorded in NEHRP site classes B, C and D.

Figure 4. Coecients of variation of inelastic displacement ratios for all 216 ground motions
recorded in NEHRP site classes B, C and D.

4.3. Dispersion

While mean inelastic displacement ratios are very important, as they represent what can be
expected on average, it is equally important to quantify the level of dispersion in CR . A
common and eective way to quantify the dispersion is through the coecient of variation
(COV), which is dened as the ratio of the standard deviation to the mean. Figure 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 inelas-
tic deformation increases. Dispersion is particularly high for periods of vibration (T ¡0:4 s)
regardless of the lateral strength ratio. Relatively high dispersion also occurs for high values
of R and periods up to 1:5 s. Furthermore, with the exception of very short periods (smaller
than 0:15 s), for a given level of ductility demand the COV decreases with increasing periods.

Copyright ? 2003 John Wiley & Sons, Ltd. Earthquake Engng Struct. Dyn. 2003; 32:1237–1258
 10969845, 2003, 8, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/eqe.271 by Universidad Tecnica Federico, Wiley Online Library on [15/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
INELASTIC DISPLACEMENT RATIOS FOR EVALUATION OF EXISTING STRUCTURES 1249

Figure 5. Inelastic displacement ratios corresponding to dierent percentiles for (a) R = 3 and (b) R = 5.

This trend is more noticeable for systems with smaller relative strengths (i.e. with higher val-
ues of R). In general, dispersion in CR is larger than the dispersion reported by Miranda [14]
for C
, particularly for periods smaller than 1:0 s.
Another way to consider the dispersion on CR is to compute inelastic displacement ratios
corresponding to dierent percentiles. Inelastic displacement ratios for a lateral strength ratio
of 3 corresponding to percentiles of 10%, 30%, 50%, 70% and 90% are shown in Figure 5(a).
It can be seen that although median inelastic displacement ratios (p = 50%) are approximately
equal to one for periods longer than about 1:0 s, there is an 80% probability that inelastic
displacement ratios will be between the curves associated to percentiles 10 and 90%, which
in this spectral region implies that in 80% of the cases CR varies approximately between 0.65
and 1.5. 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 displacement
ratios corresponding to the same percentiles but for a lateral strength ratio of ve are shown in
Figure 5(b). In this case it can be seen that for periods of vibration larger than about 1:0s there
is an 80% probability that maximum inelastic displacement demand will be approximately
between 0.6 and 1.85 times the maximum elastic displacement demand.

Copyright ? 2003 John Wiley & Sons, Ltd. Earthquake Engng Struct. Dyn. 2003; 32:1237–1258
 10969845, 2003, 8, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/eqe.271 by Universidad Tecnica Federico, Wiley Online Library on [15/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
1250 J. RUIZ-GARC
IA AND E. MIRANDA

Figure 6. Mean ratios of maximum deformation of bilinear to elastoplastic systems for all 216 ground
motions recorded in NEHRP site classes B, C and D.

4.4. E
ect of post-yield sti
ness

In order to study further the eect of positive post-yield stiness on the maximum inelastic
displacement demands, maximum deformations of bilinear systems with post-yield stiness
ratios (post-yield stiness normalized by initial stiness) of  = 3, 5 and 10% were computed
when subjected to all 216 ground motions. Then, ratios of the maximum inelastic deformation
of bilinear systems to maximum deformation of the elastoplastic system was obtained for each
record, each period of vibration, and each lateral strength ratio. Figure 6 shows mean ratios of
maximum inelastic displacement demand of bilinear systems having post-yield stiness of 3%
of the initial stiness to the maximum inelastic displacement demand of elastoplastic systems.
It can be seen that the maximum inelastic deformation of the bilinear systems becomes smaller
with respect to the one of the elastoplastic system as the strength ratio increases. For periods
of vibrations larger than about 1:0s this ratio of maximum deformation remains approximately
constant. However, for periods smaller than about 0:5 s the maximum deformation of systems
with positive post-yield stiness can be signicantly smaller than that of elastoplastic systems.
Figure 7 shows the coecient of variation of inelastic displacement ratio of bilinear systems.
Comparing Figures 4 and 7 it can be seen that the dispersion of inelastic displacement ratios
of systems with positive post-yield stiness is essentially the same as that with elastoplastic
systems except for periods smaller than about 0:25 s where small reductions in dispersion are
observed.

4.5. E
ect of soil conditions

Most structures are built on sites that are classied as rm sites (site classes B, C and D).
Thus, it is important to quantify the dierences on inelastic displacement ratios computed
from ground motions recorded in these site classes. In order to assess the eect of local site
conditions, that is to evaluate the dierences in CR for dierent site conditions within rms
sites (site classes B, C and D), ratios of mean CR on each group to mean CR computed from

Copyright ? 2003 John Wiley & Sons, Ltd. Earthquake Engng Struct. Dyn. 2003; 32:1237–1258
 10969845, 2003, 8, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/eqe.271 by Universidad Tecnica Federico, Wiley Online Library on [15/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
INELASTIC DISPLACEMENT RATIOS FOR EVALUATION OF EXISTING STRUCTURES 1251

Figure 7. Coecients of variation of maximum inelastic displacement ratio of bilinear systems with
 = 3% computed with all 216 ground motions recorded in NEHRP site classes B, C and D.

all 216 ground motions were computed. Figures 8(a)–(c) show mean inelastic displacement
ratios for site class B, site class C and site class D normalized by mean inelastic displacement
of the four site classes, respectively. It can be seen if one neglects the eects of local site
conditions for structures on rock (site class B) and uses mean values from all 216 ground
motions one would overestimate maximum inelastic displacement demands. For periods be-
tween 0.2 and 1:2 s the overestimation is less than 15% while for periods longer than 1:2 s
the overestimation is less than 5%. For structures on site class C the use of mean inelastic
displacement ratios from all site classes considered here would produce small underestima-
tions of maximum displacements for T ¡0:2 s, small overestimations for 0:2 s¡T ¡0:9 s and
practically no errors on average for T ¿1:0 s. Meanwhile for structures on sites classied as
D ignoring the eects of site conditions in the estimation of CR could result in small under-
estimations of maximum displacement for structures with periods smaller than 1:4 s. It can
be seen that the dierence in inelastic displacement ratio produced by local site conditions
increases with increasing lateral strength ratio. Thus, for lateral strength ratios smaller than
3 the errors produced by neglecting the eect of local site conditions on CR are typically
smaller than 10%, while for strength ratios of 4 and 5 site conditions are typically smaller
than 20%.
Dierences in mean inelastic displacement ratios for these site classes can be slightly re-
duced if periods of vibration are normalized by a characteristic period for each site class as
suggested by Chopra and Chintanapakdee [21]. Figure 9 shows a comparison of mean inelastic
displacement ratios for R = 4 and 6 for sites classes B, C and D when periods of vibration
are normalized by characteristic site periods of 0.75, 0.85 and 1:05 s, respectively. In this case
mean inelastic displacement ratios are closer to each other.

4.6. E
ect of earthquake magnitude

Elastic spectral ordinates are dependent on the magnitude of the earthquake. Thus, it is impor-
tant to know to what extent earthquake magnitude aects inelastic displacement ratios. In order

Copyright ? 2003 John Wiley & Sons, Ltd. Earthquake Engng Struct. Dyn. 2003; 32:1237–1258
 10969845, 2003, 8, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/eqe.271 by Universidad Tecnica Federico, Wiley Online Library on [15/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
1252 J. RUIZ-GARC
IA AND E. MIRANDA

Figure 8. Mean inelastic displacement ratios on each group normalized by mean ratios from all ground
motions: (a) site class B; (b) site class C; (c) site class D.

to study the eect of earthquake magnitude, mean inelastic displacement ratios were computed
for ground motions recorded on class D sites and then grouped according to the magnitude
of the earthquake in which they were recorded. A comparison of mean inelastic displacement
ratios computed ground motions in three ranges of magnitudes for lateral strength ratios equal

Copyright ? 2003 John Wiley & Sons, Ltd. Earthquake Engng Struct. Dyn. 2003; 32:1237–1258
 10969845, 2003, 8, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/eqe.271 by Universidad Tecnica Federico, Wiley Online Library on [15/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
INELASTIC DISPLACEMENT RATIOS FOR EVALUATION OF EXISTING STRUCTURES 1253

Figure 9. Mean inelastic displacement ratios normalized by a characteristic period for each
site class for (a) R = 4 and (b) R = 6.

to 2 and 4 is shown in Figure 10. It can be seen that for a lateral strength ratio equal to 2,
earthquake magnitude has practically no eect on CR . However, for weaker structures relative
to the intensity of the ground motions (i.e. for higher values of R) magnitude can inuence CR
for systems with short periods of vibration. It can be seen that for R = 4 the inelastic displace-
ment ratios of ground motions recorded in earthquakes with surface-wave magnitudes between
5.7 and 6.2 tend to be smaller than those of ground motions recorded in earthquakes with
higher magnitudes. In particular, for R = 4 and T = 0:5s, mean values of CR of ground motions
recorded in earthquakes with magnitudes higher than 6.3 are approximately two times higher
those from ground motions recorded in earthquake with magnitudes between 5.7 and 6.2.

4.7. E
ect of distance to the rupture

In order to study the eects of distance to the source on inelastic displacement ratios mean
inelastic displacement ratios in site class D were computed from earthquake ground motions
in three groups having a dierent range of distances to the horizontal projection of the rupture
(the so-called Joyner and Boore distance). A comparison of mean inelastic displacement ratios
for dierent distances to the rupture for lateral strength ratios equal to 2 and 4 is shown

Copyright ? 2003 John Wiley & Sons, Ltd. Earthquake Engng Struct. Dyn. 2003; 32:1237–1258
 10969845, 2003, 8, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/eqe.271 by Universidad Tecnica Federico, Wiley Online Library on [15/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
1254 J. RUIZ-GARC
IA AND E. MIRANDA

Figure 10. Eect of earthquake magnitude on mean inelastic displacement

ratios for (a) R = 2 and (b) R = 4.

in Figure 11. It can be seen that, for this range of distances, changes in mean inelastic
displacement ratios are relatively small. However, it should be noted that the ensemble of
records considered here does not include near-fault records. In a recent study using only near-
eld records, B
aez and Miranda [22] concluded that constant ductility inelastic displacement
ratios for periods between 0.1 and 1:3 s for near-eld records with forward directivity eects
(i.e. those recorded 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 aected by forward directivity.

5. NON-LINEAR REGRESSION ANALYSES

For displacement-based design and, in general, in earthquake-resistant design it is desirable

to have a simplied expression of inelastic displacement ratios to estimate maximum inelastic
displacement demands from maximum elastic displacement demands for structures where the

Copyright ? 2003 John Wiley & Sons, Ltd. Earthquake Engng Struct. Dyn. 2003; 32:1237–1258
 10969845, 2003, 8, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/eqe.271 by Universidad Tecnica Federico, Wiley Online Library on [15/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
INELASTIC DISPLACEMENT RATIOS FOR EVALUATION OF EXISTING STRUCTURES 1255

Figure 11. Eect of nearest distance to the horizontal projection of rupture on mean inelastic displace-
ment ratios for (a) R = 2 and (b) R = 4.

lateral strength is known. In this investigation non-linear regression analyses were done in
order to minimize the mean relative error given by
1  n  50 6 C 
R elastic; i; j − inelastic;i; j; k
E = (5)
300n i=1 j=1 l=1 inelastic; i; j; k

where elastic; i; j is the elastic displacement for the jth period when subjected to the ith
record, elastic; i; j; k is the inelastic displacement for the ith record, jth period and kth strength,
CR elastic; i; j is the approximate inelastic displacement, and n is the number of records. The
proposed equation is given by


1 1
CR = 1 + − (R − 1) (6)
a(T=Ts )b c
where R is the lateral strength ratio, T is the period of vibration of the system. Ts is the
characteristic period at the site and a b, and c are constants that also depend on the site

Copyright ? 2003 John Wiley & Sons, Ltd. Earthquake Engng Struct. Dyn. 2003; 32:1237–1258
 10969845, 2003, 8, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/eqe.271 by Universidad Tecnica Federico, Wiley Online Library on [15/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
1256 J. RUIZ-GARC
IA AND E. MIRANDA

Table IV. Site-dependent coecients to be used in Equation (5).

Site class a b c Ts (s)
B 42 1.60 45 0.75
C 48 1.80 50 0.85
D 57 1.85 60 1.05

conditions. The Levenberg–Marquardt method [23] was used to compute the parameters a,
b and c that minimize Equation (5). The resulting values of these parameters are given in
Table IV. Equation (6) corresponds to a surface in the CR –R–T space that provides estimates
of mean inelastic displacement ratios as a function of R and T . This equation can be further
simplied, with only slightly larger mean relative error, if coecients a = 50, b = 1:8 and
c = 55 are kept constant for all three site classes, which means that only Ts is changed from
one class to another.

6. CONCLUSIONS

The purpose of this study was to assess inelastic displacement ratios that permit the estima-
tion of maximum inelastic displacements from maximum elastic displacements for existing
structures built on rm sites whose lateral strength is known. A statistical study has been
presented of inelastic displacement ratios computed for SDOF systems with an elasto-plastic
hysteretic behaviour with dierent levels of lateral strengths relative to the strength required to
maintain the system elastic when subjected to a large number of earthquake ground motions.
The following conclusions are drawn from the results of this investigation.
In the short-period spectral region, maximum inelastic displacement demands of systems
with constant relative strength are on average much larger than maximum elastic demands. In
this spectral region the ratio of maximum inelastic to maximum elastic displacement demand is
strongly dependent on the period of vibration and on the lateral strength ratio. Constant relative
strength inelastic displacement ratios increase as the lateral strength of the system decreases
with respect to the lateral strength required to maintain the system elastic. For periods longer
than 1:2 s and lateral strength ratios smaller than 6 maximum inelastic displacement demands
are approximately equal to maximum elastic demands.
Coecients of variation of inelastic displacement ratios increase with increasing lateral
strength ratios. Dispersion is relatively large for lateral strength ratios higher than 4 and
periods of vibration smaller than 1:5 s. With the exception of periods shorter than 0:15 s,
coecients of variation decrease with increasing period of vibration.
It was found that the eects of local site conditions on constant relative strength inelastic
displacement ratios are slightly larger than those of constant ductility inelastic displacement
ratios, however, in general the eect is still relatively small, particularly for periods longer
than 1:2 s. Neglecting the eect of site conditions for structures with periods smaller than
1:5 s built on rm sites will typically result in errors less than 20% in the estimation of mean
inelastic displacement ratios, whereas for periods longer than 1:5 s the errors are smaller than
10%. Dierences are even smaller if the lateral strength ratio is equal to or smaller than 3.

Copyright ? 2003 John Wiley & Sons, Ltd. Earthquake Engng Struct. Dyn. 2003; 32:1237–1258
 10969845, 2003, 8, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/eqe.271 by Universidad Tecnica Federico, Wiley Online Library on [15/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
INELASTIC DISPLACEMENT RATIOS FOR EVALUATION OF EXISTING STRUCTURES 1257

Limiting periods dividing regions where the equal displacement rule is applicable from
those where this rule is not applicable depend primarily on the lateral strength ratio, although
they are also inuenced by local site conditions. In general these limiting periods increase
primarily as the lateral strength ratio increases and to a lesser degree as the average shear
wave velocity in the upper 30 m of the site prole decreases.
For periods of vibration longer than 1:0 s changes in earthquake magnitude do not aect
constant relative strength inelastic displacement ratios. Nevertheless, for systems with periods
smaller than 1:0 s some dependence on magnitude was observed. In the short period region
inelastic displacement ratios computed from ground motions recorded during earthquakes with
magnitudes higher than 6.3 were found to be larger than those computed from records obtained
in earthquakes with magnitudes between 5.7 and 6.2. Inelastic displacements are not aected
by distance for sites located more than 10 km away from the horizontal projection of the
rupture.
A simplied equation was derived using non-linear regression analyses to estimate inelastic
displacement demands of existing structures with known lateral strength. Coecients corre-
sponding to dierent local site conditions were computed. The proposed equation minimizes
mean relative errors between approximate and exact inelastic displacements.

ACKNOWLEDGEMENTS

J. Ruiz-Garc
a acknowledges nancial support from the Consejo Nacional de Ciencia y Tecnologa
(CONACYT) in Mexico to pursue his doctoral studies at Stanford University. The last stage of this re-
search was supported by a grant to E. Miranda by the Pacic Earthquake Engineering Research (PEER)
Center with support from the Earthquake Engineering Research Centers Program of the National Science
Foundation under Award Number EEC-9701568. These nancial supports are gratefully acknowledged.
Special thanks are given to Dr Sinan D. Akkar who conducted the nal regression analyses described
in Section 5 of this paper. Records used in this investigation were provided by the California Strong
Motion Instrumentation Program of the California Geological Survey and by the PEER ground motion
database. The authors are grateful to Professors A.S. Veletsos and A.K. Chopra and to one anonymous
reviewer who provided useful comments that helped to improve the nal version of this work.

REFERENCES

1. Applied Technology Council. Seismic evaluation and retrot of concrete buildings. Report ATC-40. Applied
Technology Council, Redwood City, CA, 1996.
2. Federal Emergency Management Agency (FEMA). NEHRP guidelines for the seismic rehabilitation of buildings.
Reports FEMA 273 (Guidelines) and 274 (Commentary). Washington, DC, 1997.
3. Federal Emergency Management Agency (FEMA). NEHRP recommended provisions for seismic regulations
for new buildings and other structures. Reports FEMA 302 (Provisions) and 303 (Commentary). Washington,
DC, 1997.
4. Veletsos AS, Newmark NM. Eect of inelastic behavior on the response of simple systems to earthquake
motions. Proceedings of the 2nd World Conference on Earthquake Engineering, Japan, vol. 2, 1960; 895 – 912.
5. Veletsos AS, Newmark NM, Chelapati CV. Deformation spectra for elastic and elastoplastic systems subjected to
ground shock and earthquake motions. Proceedings of the 3rd World Conference on Earthquake Engineering,
New Zealand, vol. II, 1965; 663 – 682.
6. Veletsos AS, Vann WP. Response of ground excited elastoplastic systems. Journal of the Structural Division
(ASCE) 1971; 97(4):1257–1281.
7. Shimazaki K, Sozen MA. Seismic drift of reinforced concrete structures. Technical Research Reports of
Hazama-Gumi Ltd. Tokyo, Japan, 1984; 145–166.

Copyright ? 2003 John Wiley & Sons, Ltd. Earthquake Engng Struct. Dyn. 2003; 32:1237–1258
 10969845, 2003, 8, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/eqe.271 by Universidad Tecnica Federico, Wiley Online Library on [15/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
1258 J. RUIZ-GARC
IA AND E. MIRANDA

8. Qi X, Moehle JP. Displacement design approach for reinforced concrete structures subjected to earthquakes.
Report No. EERC 91=02, Earthquake Engineering Research Center, University of California at Berkeley,
Berkeley, CA, 1991.
9. Miranda E. Seismic evaluation and upgrading of existing structures. Ph.D. Thesis, University of California,
Berkeley, CA, 1991.
10. Miranda E. Evaluation of site-dependent inelastic seismic design spectra. Journal of Structural Engineering
(ASCE) 1993; 119(5):1319–1338.
11. Rahnama M, Krawinkler H. Eects of soft soils and hysteresis model on seismic demands. Report No. 108,
The John A. Blume Earthquake Engineering Center, Stanford University, Stanford, CA, 1993.
12. Seneviratna GDPK, Krawinkler H, Evaluation of inelastic MDOF eects for seismic design. Report No. 120,
The John A. Blume Earthquake Engineering Center, Stanford University, Stanford, CA, 1997.
13. Whittaker AS, Constantinou M, Tsopelas P. Displacement estimates for performance-based seismic design.
Journal of Structural Engineering (ASCE) 1998; 124(8):905 – 912.
14. Miranda E. Inelastic displacement ratios for structures on rm sites. Journal of Structural Engineering (ASCE)
2000; 126(10):1150 – 1159.
15. Chopra AK, Goel RK. Capacity-demand-diagram methods based on inelastic design spectrum. Earthquake
Spectra 1999; 15(4):637 – 656.
16. Chopra AK, Goel RK. Evaluation of NSP to estimate seismic deformations: SDF systems. Journal of Structural
Engineering 2000; 126(4):482– 490.
17. Fajfar P. Capacity spectrum method based on inelastic demand spectra. Earthquake Engineering and Structural
Dynamics 1999; 28(9):979 – 993.
18. Fajfar P. A nonlinear analysis method for performance-based seismic design. Earthquake Spectra 2000;
16(3):573 – 592.
19. Miranda E. Estimation of inelastic deformation demands of SDOF systems. Journal of Structural Engineering
(ASCE) 2001; 127(9):1005–1012.
20. International Conference of Building Ocials. Uniform Building Code. Whittier, CA, 1997.
21. Chopra AK, Chintanapakdee C. Comparing response of SDF systems to near-fault and far-fault earthquake
motions in the context of spectral regions. Earthquake Engineering and Structural Dynamics 2001; 30:
1769 – 1789.
22. B
aez JI, Miranda E. Amplication factors to estimate inelastic displacement demands for the design of structures
in the near eld. Proceedings of the 12th World Conference on Earthquake Engineering, New Zealand, 2000.
23. Bevington PR, Robinson DK. Data Reduction and Error Analysis for the Physical Sciences. McGraw-Hill:
New York, 1992.

Copyright ? 2003 John Wiley & Sons, Ltd. Earthquake Engng Struct. Dyn. 2003; 32:1237–1258