10 1002@eqe 2288

EARTHQUAKE ENGINEERING & STRUCTURAL DYNAMICS
Earthquake Engng Struct. Dyn. 2013; 42:1583–1598

Published online 7 March 2013 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/eqe.2288
Residual drift demands in moment-resisting steel frames subjected

to narrow-band earthquake ground motions
Edén Bojórquez1,*,† and Jorge Ruiz-García2

1
Facultad de Ingeniería Civil, Universidad Autónoma de Sinaloa Calz. de las Américas y Boulevard Universitarios, Cd.
Universitaria, 80040 Culiacán, México
2
Facultad de Ingeniería Civil, Universidad Michoacana de San Nicolás de Hidalgo Edificio C, Planta Baja, Cd.
Universitaria, 58040 Morelia, México
SUMMARY
This paper presents the main results of the evaluation of residual inter-story drift demands in typical moment-
resisting steel buildings designed accordingly to the Mexican design practice when subjected to narrow-band
earthquake ground motions. Analytical 2D-framed models representative of the study-case buildings were
subjected to a set of 30 narrow-band earthquake ground motions recorded on stations placed in soft-soil sites
of Mexico City, where most significant structural damage was found in buildings as a consequence of the
1985 Michoacan earthquake, and scaled to reach several levels of intensity to perform incremental dynamic
analyses. Thus, results were statistically processed to obtain hazard curves of peak (maximum) and residual
drift demands for each frame model. It is shown that the study-case frames might exhibit maximum residual
inter-story drift demands in excess of 0.5%, which is perceptible for building’s occupants and could cause
human discomfort, for a mean annual rate of exceedance associated to peak inter-story drift demands of about
3%, which is the limiting drift to avoid collapse prescribed in the 2004 Mexico City Seismic Design Provisions.
The influence of a member’s post-yield stiffness ratio and material overstrength in the evaluation of maximum
residual inter-story drift demands is also discussed. Finally, this study introduces response transformation
factors, Tp, that allow establishing residual drift limits compatible with the same mean annual rate of exceedance
of peak inter-story drift limits for future seismic design/evaluation criteria that take into account both drift
demands for assessing a building’s seismic performance. Copyright © 2013 John Wiley & Sons, Ltd.
Received 14 February 2012; Revised 15 January 2013; Accepted 26 January 2013
KEY WORDS: residual inter-story drift; narrow-band ground motions; moment-resisting steel frames;
repairability; response transformation factors
1. INTRODUCTION
Nowadays, recently proposed performance-based seismic design and assessment procedures for new
and existing structures emphasize on the estimation of peak lateral drift demands. However, earthquake
field reconnaissance has evidenced that residual lateral displacement demands after earthquake
excitation (e.g., residual roof drift ratio or maximum residual inter-story drift ratio) also play an
important role in defining the seismic performance of a structure and it can have important
consequences. For instance, several dozen damaged RC buildings in Mexico City had to be demolished
after the 1985 Michoacan earthquake because of the technical difficulties to straighten and to repair
buildings with large permanent drifts [1]. As another example, Okada et al. [2] reported that several
low-rise RC buildings suffered light structural damage but experienced relatively large residual
deformations as a consequence of the 1995 Hyogo-Ken Nambu earthquake even though they had
sufficient deformation capacity. After examining 12 low-to-mid-rise steel office buildings (particularly
*Correspondence to: Edén Bojórquez, Facultad de Ingeniería, Universidad Autónoma de Sinaloa, Calz. de las Américas y
Boulevard Universitarios, Cd. Universitaria, 80040 Culiacán, México.
†
E-mail: eden@uas.edu.mx; eden_bmseg@hotmail.com
Copyright © 2013 John Wiley & Sons, Ltd.

1584 E. BOJÓRQUEZ AND J. RUIZ-GARCÍA
10 with structural system based on steel moment-resisting frames) structurally damaged and leaned after
the same earthquake, Iwata et al. [3] highlighted that the cost of repairing leaned steel buildings linearly
increased as the maximum and roof residual drift increased. On the basis of their study, the authors
suggested that steel buildings should be limited to maximum and roof residual drift limits of about
1.4% and 0.9%, respectively, to satisfy a repairability limit state that meet both technical
and economical constraints. More recently, a field investigation in Japan highlighted that a residual
inter-story drift of about 0.5% is perceptible for building occupants and a residual inter-story drift of
about 1.0% could cause human discomfort [4]. These field observations also imply that the total
expected economic losses computed from peak drift demands and peak floor acceleration demands
could be smaller than those computed taking into account permanent drift demands due to necessity of
demolishing structures having excessive residual deformations due to, again, the technical difficulties to
move back the structure to its original position, although they did not experience severe structural
damage [5]. Therefore, several researchers have highlighted that the estimation of residual drift
demands should also play an important role during the design of new buildings [e.g., 6, 7, 13] and the
evaluation of the seismic structural performance of existing buildings [e.g., 8–12].
Motivated by earthquake field reconnaissance observations, researchers have performed analytical
investigations aimed at gaining further understanding on the parameters that influence the amplitude
and height-wise distribution of residual drift demands in existing multi-story buildings [e.g., 6, 8–12].
They have reported that the residual drift demand amplitude and distribution over the height depends
on the component hysteretic behavior [6, 8, 9], building frame mechanism [6, 8, 9], structural
overstrength [8, 9], and ground motion intensity [6, 8–11].
However, it should be noted that previous studies have employed earthquake ground motions
recorded in accelerographic stations placed on rock or firm sites, which have very different
frequency content and duration characteristics than those recorded in stations placed on soft-soil
sites (e.g., the lake-bed zone of Mexico City or the Marina district in San Francisco Bay). These
types of records are characterized by relatively long predominant period of the ground motion and
low-frequency content (i.e., narrow-band earthquake ground motions). Therefore, the response of
structures under narrow-band earthquake ground motions strongly depends on the relationship
between the first-mode period of vibration of the structure and the predominant period of the ground
motion [e.g., 8, 14]. In fact, previous studies have shown that peak inelastic displacement demands
become large when the period of vibration of the structures shortens with respect to the predominant
period of the ground motion [e.g., 8]. Furthermore, a recent study [13] showed that ductile steel
moment-resisting frames designed under seismic loading following the ASCE 7-05 standard [15],
which is the up-to-date standard for seismic design in the USA, would experience excessive residual
inter-story drifts when subjected to earthquake ground motions representative of firm and rock sites
scaled to reach the maximum credible excitation level. Therefore, this study motivates the need to
evaluate if buildings designed on soft-soil sites with modern seismic provisions would experience
excessive residual inter-story drift levels. For this task, this type of study requires a probabilistic
approach such as those proposed in [8, 10, 11].
The objective of this paper is to evaluate the levels of residual inter-story drift demands during the
seismic performance-based assessment of moment-resisting steel buildings designed with a modern
seismic code when subjected to narrow-band earthquake ground motions. For this purpose, four
moment-resisting steel buildings with different heights designed accordingly to the 2004 Mexico
City Seismic Design Provisions (MCSDP) [16] were evaluated under a set of 30 earthquake ground
motions recorded on stations placed in soft-soil sites of Mexico City during five historical
earthquakes, and scaled to reach different levels of ground motion intensity. Although this study
provides new information on the importance of residual drift demands triggered by soft-soil records,
findings are limited to the study-case steel buildings.
2. BUILDING FRAME MODELS EMPLOYED IN THIS STUDY
Typical regular steel buildings for office occupancy and having different number of stories (N = 4, 6, 8, and 10)
were considered in this investigation. Figure 1 shows the plan view of the buildings. All frames were
Copyright © 2013 John Wiley & Sons, Ltd. Earthquake Engng Struct. Dyn. 2013; 42:1583–1598
DOI: 10.1002/eqe
RESIDUAL DRIFT DEMANDS IN MOMENT-RESISTING STEEL FRAMES 1585
Analyzed frames
D
5.00 m
5.00 m EBF EBF

B
5.00 m
A
8.00 m 8.00 m 8.00 m
1 2 3 4
Figure 1. Plan view of moment-resisting steel frames.
designed by an experienced structural engineering office following the MCSDP [16]. Therefore, moment-
resisting frames were provided in both the longitudinal and transverse direction, while additional eccentric
brace frames (EBF) were incorporated in the transverse direction for drift control because the weak axis of
the columns is oriented in this direction. However, it should be pointed out that the goal of this and other
investigations [e.g., 17, 18] was to access the seismic response of the moment-resisting steel frames in the
longitudinal direction. The buildings were assumed to be located in the soft zone of Mexico City, which is
the zone with the highest seismic hazard in the former ‘lake-bed’ zone of the city. Elastic acceleration
design spectrum ordinates were reduced by a response modification factor equal to 3 in both directions,
which takes into account the ability of the structure to undergo inelastic deformations, without
consideration of structural overstrength, but with limited ductility. Thus, an equivalent static linear
analysis, which is commonly used in the Mexican design practice, assuming a triangle inverted
distribution of code-specified base shear was employed for sizing the frame members. To provide a
context with US standards, the frames in the longitudinal direction are similar to the ordinary steel
moment frames specified in Table 12.2-1 of the ASCE 7-05 standard [15].
The buildings were analyzed with the nonlinear dynamic analysis computer program RUAUMOKO
[19]. Only half of each building was modeled because of symmetry in the building’s plan. Exterior and
interior frames were modeled as two-dimensional centerline models, assuming fixed columns, which
imply that base flexibility and soil–structure interaction was neglected. Both frames were attached
through rigid frame elements to experience the same lateral deformation at each floor assuming a
rigid diaphragm. It should be mentioned that in Mexican design practice, all frames in the same
direction are designed as moment-resisting frames, which carry the gravity loads proportional to
their tributary area (i.e., interior columns and beams are not considered only as load-carrying gravity
elements). For dynamic analyses, second-order effects were explicitly considered (i.e., large
displacement analysis). Mass-and-stiffness proportional Rayleigh damping was considered for the
analysis, where 3% of critical damping was assigned to the first two modes of vibration of the frames.
Beams and columns were modeled as frame elements, which concentrate their inelastic response in
plastic hinges located at their ends (i.e., lumped plasticity nonlinear frame elements). Plastic hinge
length was assumed as 80% of the section depth. Nondegrading bilinear moment–curvature
relationship with strain-hardening ratio equal to 3% that considers axial load–flexural bending
interaction was considered to model the hysteretic behavior of the steel columns. The beam’s
behavior was modeled through a bilinear moment–curvature relationship with strain-hardening ratios
equal to 1% and 3% (baseline value). However, slab contribution to the beam’s bending capacity
was neglected in this study. Baseline results are presented on the basis of flexural moment capacity
of beams and columns determined from a nominal steel yield stress of 2530 kg/cm2, consistent with
the design process of the frames, but expected steel yield stress of 3450 kg/cm2 was also considered
to discuss the influence of material overstrength on residual drift demands. In addition, the flexibility
of the panel zone was not explicitly modeled, but rigid offsets in both beams and columns
were taken into account in further analyses, which are discussed in a separate section. Finally, it
should be mentioned the numerical modeling of the study-case frames for computing their
DOI: 10.1002/eqe
nonlinear response followed well-known modeling strategies (e.g., lumped-plasticity elements, mass-
and-stiffness proportional Rayleigh damping, and rigid diaphragm assumption). However, to
validate the response of the numerical models against experimental evidence, a series of small-scale
or large-scale steel specimens representative of the building’s Mexican design practice should have
been subjected to quasi-static protocols representative of the soft-soil earthquake ground motions
employed in this study, or even shaking-table testing using narrow-band earthquake ground motions.
Unfortunately, this task has not been carried out in Mexico, and there is a lack of information about
the experimental performance of steel structures under this type of records.
Before conducting nonlinear dynamic time-history analyses, modal and nonlinear static
(pushover) analyses of each frame were performed to obtain relevant dynamic and mechanical
characteristics such as the fundamental period of vibration, T1; the building’s yield coefficient
(i.e., base shear normalized with respect to the building’s weight), Cy; and roof displacement at
yielding, Dy, as it is reported in Table I. It should be mentioned that pushover analysis was
conducted with the software RUAUMOKO [19] by using a slow ramp loading function
assuming a triangular-inverted loading distribution.
After this task, dynamic time-history analyses were carried out for each frame subjected to a set of
narrow-band earthquake ground motions described as follows.
3. SELECTION OF NARROW-BAND EARTHQUAKE GROUND MOTIONS
The study-case frame models were subjected to a set of 30 narrow-band and long duration ground
motions recorded at soft-soil sites of Mexico City from five historical earthquakes. The seismic
events had seismic magnitudes ranging between 6.9 and 8.1, with epicenters located at distances
about 300 km, or more, from Mexico City. All records were gathered from accelerographic stations
located at sites having soil periods equal to, or approximately, 2 s. It should be mentioned that sites
having soil periods of 2 s are fairly common within the Lake Zone and that the higher levels of
shaking in terms of peak ground acceleration have been consistently observed at these sites.
Furthermore, Meli and Avila [14] found that most of the structural damage in Mexico City due to
the September 19, 1985 earthquake were recorded in buildings with structural periods from 0.5 to
1.5 s, which are smaller than those of the soil (about 2 s), because these types of buildings were
particularly subjected to significant number of cycles that could have diminished their lateral
strength and stiffness capacity. Moreover, ground motions recorded on soft soils lead to large energy
demands on structures as compared with those recorded on firm soils [20, 21]. These important
issues motivated the use of the study-case frames, which fall into the range of fundamental periods
from 0.5 to 1.5 s.
Some important characteristics of the selected records are summarized in Table II. In this table,
whereas PGA and PGV stand for the peak ground acceleration and velocity, respectively, tD is the
strong-motion duration estimated according to the Trifunac and Brady [22] criterion, which is
defined as the time interval delimited by the instants of time at which the 5% and 95% of the Arias
Intensity occurs. Note that the average duration of the records equals 74.4 s.
Table I. Structural properties of baseline steel frames (with material post-yield stiffness ratio equal to 3% and
nominal properties) under consideration.
Frame Number of stories T1 (s) Cy Dy (m)
F4 4 0.90 0.45 0.136
F6 6 1.07 0.42 0.174
F8 8 1.20 0.38 0.192
F10 10 1.37 0.36 0.226
DOI: 10.1002/eqe
Table II. Selected narrow-band ground motions.

Records Date Magnitude Station PGA (cm/s2) PGV (cm/s) tD (s)
1 19/09/1985 8.1 SCT 178.0 59.5 34.8
2 21/09/1985 7.6 Tlahuac deportivo 48.7 14.6 39.9
3 25/04/1989 6.9 Alameda 45.0 15.6 37.8
4 25/04/1989 6.9 Garibaldi 68.0 21.5 65.5
5 25/04/1989 6.9 SCT 44.9 12.8 65.8
6 25/04/1989 6.9 Sector Popular 45.1 15.3 79.4
7 25/04/1989 6.9 Tlatelolco TL08 52.9 17.3 56.6
8 25/04/1989 6.9 Tlatelolco TL55 49.5 17.3 50.0
9 14/09/1995 7.3 Alameda 39.3 12.2 53.7
10 14/09/1995 7.3 Garibaldi 39.1 10.6 86.8
11 14/09/1995 7.3 Liconsa 30.1 9.62 60.0
12 14/09/1995 7.3 Plutarco Elías Calles 33.5 9.37 77.8
13 14/09/1995 7.3 Sector Popular 34.3 12.5 101.2
14 14/09/1995 7.3 Tlatelolco TL08 27.5 7.8 85.9
15 14/09/1995 7.3 Tlatelolco TL55 27.2 7.4 68.3
16 09/10/1995 7.5 Cibeles 14.4 4.6 85.5
17 09/10/1995 7.5 CU Juárez 15.8 5.1 97.6
18 09/10/1995 7.5 Centro urbano Presidente Juárez 15.7 4.8 82.6
19 09/10/1995 7.5 Córdoba 24.9 8.6 105.1
20 09/10/1995 7.5 Liverpool 17.6 6.3 104.5
22 09/10/1995 7.5 Sector Popular 13.7 5.3 98.4
23 09/10/1995 7.5 Valle Gómez 17.9 7.18 62.3
24 11/01/1997 6.9 CU Juárez 16.2 5.9 61.1
25 11/01/1997 6.9 Centro urbano Presidente Juárez 16.3 5.5 85.7
26 11/01/1997 6.9 García Campillo 18.7 6.9 57.0
28 11/01/1997 6.9 Est. # 10 Roma A 21.0 7.76 74.1
29 11/01/1997 6.9 Est. # 11 Roma B 20.4 7.1 81.6
30 11/01/1997 6.9 Tlatelolco TL08 16.0 7.2 57.5
4. SEISMIC PERFORMANCE OF STUDY-CASE STEEL FRAMES
4.1. Framework
The seismic performance of the selected steel frames is estimated in terms of peak and residual drift
demands. At a first phase, incremental dynamic analysis [23] is used to assess the seismic
performance of the steel frames under narrow-band motions at different intensity levels. Next, the
well-known seismic performance-based assessment procedure suggested by the Pacific Earthquake
Engineering Center [24] in the USA was employed in this study, which indicates that the mean
annual frequency of exceedance of an engineering demand parameter (EDP) of interest exceeding a
certain level edp can be computed as follows:
Z
lðEDP > edpÞ ffi P½EDP > edpjIM ¼ imjdlIM ðimÞj (1)
IM
where IM denotes the ground motion intensity measure (e.g., peak ground acceleration, spectral
acceleration at the first-mode period of vibration, and inelastic displacement demand at the
building’s fundamental period of vibration) and P[EDP > edp | IM = im] is the conditional
probability that a EDP exceeds a certain level of edp given that the IM is evaluated at the ground
motion intensity measure level im. In addition, dlIM(im) refers to the differential of the ground
motion hazard curve for the IM. In this context, whereas the first term in the right-hand side of
Equation (1) can be obtained from probabilistic estimates of the EDP of interest (i.e., via
incremental dynamic analyses), the second term in Equation (1) is represented by the seismic hazard
curve, which can be computed from conventional probabilistic seismic hazard analysis, evaluated at
DOI: 10.1002/eqe
the ground motion intensity level im. Note the importance of the ground motion intensity measure for
assessment of seismic performance, which is the joint between earthquake engineering and
seismology. Particularly, three features are desirables in a selected ground motion intensity measure:
sufficiency, efficiency, and scaling robustness [e.g., 25]. Sufficiency means that given an IM the
structural response is not sensitive to other ground motion parameters (e.g., magnitude and
distance). Efficiency is defined as good explanatory power of the IM with respect to some EDP; this
may help in reducing the records-to-record variability to estimate the structural response in such way
that it is one of the most important features of an IM. Finally, robustness means that the amplitude
(linear) scaling of records does not induce bias in the estimation of the seismic demand. In the
present study, the spectral acceleration at the first mode of vibration Sa(T1) was selected as IM. It is
important to point out that the records used herein allow the use of a scaling criteria based on Sa(T1)
because of the following reasons: (1) sufficiency of Sa(T1) with respect to magnitude and distance;
(2) similar spectral shape of the records because the ground motion records selected have similar
values of Np [26, 27] (defined later), which is observed in Figure 2, where the response spectra of
the records scaled for similar values of Sa(T1) for a period of T = 0.90 s (the fundamental period of
frame F4) are shown; and (3) the property known as scaling robustness is satisfied, and this is valid
although significant bias usually occurs when increasing nonlinear structural behavior. Parameter Np
is defined as follows:
Saavg ðT1 ; . . . ; ; TN Þ
Np ¼ (2)
SaðT1 Þ
where Saavg(T1, . . .,TN) represents the geometrical mean between the periods T1 and TN. Bojórquez and
Iervolino [26] demonstrated that for scale factors in a range of 1 to 100, no significant bias occurs for
important levels of nonlinear behavior (ductility demands up to six) if the records are selected with
similar values of Np. Thus, when Sa(T1) is selected as IM, Equation (1) can be expressed as follows:
Z

lðEDP > edpÞ ffi P EDP > edpjSa ðT1 Þ ¼ sa dlSa ðT1 Þ ðsa Þ (3)
Sa ðT1 Þ
where dlSa ðT1 Þ ðsa Þ ¼ lSa ðT1 Þ ðsa Þ lSa ðT1 Þ ðsa þ dsa Þ is the hazard curve differential expressed in terms
of Sa(T1). Equation (3) was used to evaluate the structural reliability of the study-case frames in
terms of two EDPs: peak and residual inter-story drift demands. For evaluating the first term in the
integrand for residual drift demands, parametric lognormal cumulative probability distribution was
Figure 2. Elastic response spectra for the records scaled at the same spectral ordinate Sa(T1) = 100 cm/s2 for T
equals 0.9 s and 3% of critical damping.
DOI: 10.1002/eqe
proposed in [8, 10] for characterizing the empirical cumulative probability distribution. This proposal
was verified in this study by using the well-known Kolmogorov–Smirnov goodness-of-fit test being
positive at all intensity levels. Likewise, a parametric lognormal cumulative probability distribution
was employed for peak inter-story drift demands [24]. Therefore, the term P(EDP > edp|Sa(T1) = sa)
is analytically evaluated as follows:

lnedp m^ lnEDPjSa ðT1 Þ¼sa
PðEDP > edpjSa ðT1 Þ ¼ sa Þ ¼ 1 Φ (4)
^ lnEDPjSa ðT1 Þ¼sa
s
where m^ lnEDPjSa ðT1 Þ¼sa and s

^ lnEDPjSa ðT1 Þ¼sa are the geometric mean and standard deviation of the natural
logarithm of the EDP, respectively, and Φ() is the standard normal cumulative distribution function.
4.2. Evolution of peak and residual drift demands at different intensity levels
Figure 3 shows the structural response in terms of median peak (maximum) inter-story drift ratio,
IDRmax, and median maximum residual inter-story drift ratio, RIDRmax, for all the frames analyzed
at different levels of ground motion intensity. It can be seen that the values of IDRmax and RIDRmax
increase as the ground motion intensity tends to increase. To provide a context of the results, 2004
MCSDP suggests that IDRmax should be limited to 3% to avoid collapse [16], as shown in the
figure. Therefore, it can also be observed that the limiting IDRmax is reached when Sa(T1) is in a range
from 0.8 to 1.1 g, for all the study-case frames. In addition, Figure 4 shows the ratio of median residual
inter-story drift demands divided by median peak inter-story drift demands for all frames and intensity
levels under consideration. It can be observed that this ratio could be in a range from 15% to 18% for
the F4 and F6 models, respectively, when the frames also reach peak inter-story drift ratio demands in
the order of 3%, which indicates that residual inter-story drift demands could be as high as 0.45–0.58%.
Figure 3. Median values of IDRmax and RIDRmax using nonlinear incremental dynamic analysis for frame:
(a) F4, (b) F6, (c) F8, and (d) F10.
DOI: 10.1002/eqe
Figure 4. Ratio of median values of RIDRmax and IDRmax at different intensity levels in terms of Sa(T1) for
all frames under consideration.
A closer look at the ratio of residual to peak inter-story drift demands for each record scaled to three
different intensity levels corresponding to frame F4 is shown in Figure 5. It can be seen that this ratio
has significant record-to-record variability (e.g., this ratio reach values from almost zero to
approximately 0.5 at the same intensity level), which is not surprising because previous studies have
shown that residual drift demands computed from earthquake ground motions recorded in firm
sites involve large levels of record-to-record variability [e.g., 8–10]. To explain such a large record-
to-record variability, the evolution of IDRmax and RIDRmax as well as the height-wise distribution
of peak and residual inter-story drift for frame F4 at increasing intensity levels of records 1 (SCT station)
and 3 (Alameda Station), which have very different ratios for the same intensity level, is shown in
Figures 6 and 7, respectively. It can be observed that three patterns of the ratios can be identified:
(1) both peak and residual drift responses increase monotonically; (2) peak drift increases, but residual
drift response exhibits a lower response at some intensity level than that of the previous intensity; and
(3) peak drift increases and residual drift increases again after a lower responses for smaller intensities.
This drift behavior might suggest the following explanation: although record 3 triggers smaller peak
drift demand than that of record 1 up to, for example, a ground motion intensity of 0.8 g, residual drift
demand due to record 3 also increased monotonically (e.g., because of residual drift concentration in
the first story) while peak residual drift due to record 1 is decreasing from previous intensities.
Furthermore, it is interesting to note that record 1 leads to smaller residual/peak ratios for intensities of
Figure 5. Ratio of RIDRmax and IDRmax for three different intensity levels in terms of Sa(T1), frame F4 and
for all the records under consideration.
DOI: 10.1002/eqe
Figure 6. Evolution of IDRmax and RIDRmax at different intensity levels for frame F4 under records 1 and 3
considered in this investigation.
Figure 7. Height-wise distribution of peak and residual inter-story drift demands in the F4 frame when
subjected to different intensity levels of two earthquake ground motions: (a) record 1 and (b) record 3.
1.0 and 1.2 g (Figure 5) because residual drifts grows at a low rate after a descending rate for intensities
between 0.7 and 0.9 g, whereas peak drift grows at a faster rate. It should be noted that Vamvatsikos and
Cornell [23] also observed this twisting pattern for peak drift demands. They suggest that as the
earthquake ground motion intensity increases, earlier yielding in a story could occur at strong intensity
levels than yielding under lower intensities, which would modify subsequent nonlinear response
(e.g., constraining peak drift demands). Therefore, previous findings support that the record-
to-record variability in the estimation of residual drift demands should be explicitly incorporated
during performance-based assessment of existing buildings.
DOI: 10.1002/eqe
5. STRUCTURAL RELIABILITY ANALYSIS
5.1. Baseline results

The numerical assessment of the mean annual rate of exceeding IDRmax and RIDRmax for the frame
models having post-yield stiffness ratio equal to 3% and nominal material properties, which serves
as a baseline for discussing the influence of post-yield stiffness ratio and material overstrength in
further sections, is addressed in this section. To this task, the spectral acceleration hazard curves
corresponding to the first-mode period of vibration of each study-case building and for the
Secretaría de Comunicaciones y Transportes (SCT) site in Mexico City were developed following
the procedure suggested by Alamilla [28]. Therefore, numerical integration of Equation (3) yields
the seismic hazard curves in terms of the mean annual rate of exceeding IDRmax and RIDRmax
shown in Figure 8 for each of the four study-case frames. For both drift cases, the curves follows a
similar trend as the number of stories increases, suggesting that although the frames designed with
the MCSDP [16] were not originally developed for specific structural reliability values, the structural
designs obtained with this seismic code tend to have similar annual rates of exceeding IDRmax,
which is one of the main parameters used by the MCSDP to satisfy the earthquake-resistant design
criteria of buildings. In addition, the results illustrate that also the RIDRmax hazard curves is
practically the same for all the analyzed frames, particularly for RIDRmax demands less than 1.0%,
which means that the structural reliability of the study-case frames is quite similar if RIDRmax is
used as seismic performance parameter.
A direct comparison of seismic hazard curves of peak and residual inter-story drift demands for each
frame is shown in Figure 9, which allows a better seismic performance assessment of the frames as
suggested in [8, 10]. Recall that a residual inter-story drift in a range of about 0.5–1.0% is perceptible
for building occupants and could cause human discomfort according with reference [4]. Firstly, the
IDRmax and RIDRmax seismic hazard curves for frame F4 are shown in Figure 9(a). It is observed that
for a limiting IDRmax equals to 3%, the residual inter-story drift demand is about 0.8% to have the same
mean annual rate of exceedance of both parameters. Likewise, frame F6 could experience RIDRmax
equals to 0.65% to meet the same mean annual rate of exceedance corresponding to an IDRmax
equal to 3.0% as is shown in Figure 9(b). Therefore, although the inter-story drift threshold
recommended for the MCSDP [16] aims to guarantee a good seismic performance and to avoid
structural collapse of steel buildings, it might not avoid the necessity of repairing the building after the
earthquake because the residual inter-story drift could be larger than 0.5%, if this drift limit could be
used as a threshold to take the decision of repairing or not repairing the building. In the case of
frames F8 and F10 (Figure 9(c) and 9(d), respectively), for a mean annual rate of exceeding about
0.001 (i.e., return period of 1000 years), the IDRmax is equals to 3%, and RIDRmax is around 0.5%.
a) b)
Figure 8. Seismic hazard curves for all the study-case steel frames in terms of: (a) peak residual inter-story
drift demands and (b) residual inter-story drift demands.
DOI: 10.1002/eqe
Figure 9. Peak and residual inter-story drift hazard curves for the study-case steel frames considered in this
investigation: (a) F4, (b) F6, (c) F8, and (d) F10.
5.2. Influence of post-yield stiffness ratio

To illustrate the influence of post-yielding stiffness in the residual drift demands of moment-resisting steel
frames, the seismic performance of the study-case frame F4 was estimated by considering a post-yielding
stiffness, Kp, of 1% in the moment–curvature hysteretic behavior of their elements (both in beams and
columns) in addition to the baseline case (Kp = 3%). For instance, whereas Figure 10(a) compares the
capacity curves obtained for cases with Kp = 1% and 3%, Figure 10(b) shows a comparison of residual
inter-story drift demand hazard curves computed for both cases. Likewise, in findings from previous
studies [e.g., 6–9], residual inter-story drift demands become smaller when Kp is increased at a certain
a) b)
Figure 10. Influence of post-yielding stiffness for frame F4 with Kp equals 1% and 3% for: (a) capacity
curves and (b) residual inter-story drift demand hazard curves.
DOI: 10.1002/eqe
intensity level. From this study, it is shown that the effectiveness of Kp to constraint residual drift demands
becomes more important for earthquake ground motion intensities with low exceedance probability than
for low intensities.
Figure 11 presents a comparison of the IDRmax and RIDRmax seismic hazard curves for frames F4
and F6 having member’s post-yielding stiffness of 1%. From the figure, it can be seen that when a
limiting IDRmax equals to 3% (as shown in red solid line), RIDRmax reach values of about 1.0% to
have the same mean annual rate of exceedance of IDRmax. These levels of residual inter-story drift
would lead to a high cost of repairing the buildings as noted by Iwata et al. [3] that may lead to take
the decision of demolishing the buildings. If building owners would like to avoid human sensibility
and discomfort, IDRmax should not exceed a value of 2%, as illustrated in green dashed line, to
guarantee that RIDRmax would not exceed a threshold value of 0.5%. Other option for allowing
larger IDRmax should be to promote an increment in Kp, as illustrated in Figure 10(b).
5.3. Influence of material overstrength

Baseline results were previously presented for flexural moment capacity of beams and columns
determined from a nominal steel yield stress of 2530 kg/cm2. In this section, expected steel yield
stress of 3450 kg/cm2 was also considered to discuss the influence of material overstrength on
residual inter-story drift demands. Figure 12(a) illustrates the capacity curves obtained for frame F4
a) b)
Figure 11. Peak and residual inter-story drift hazard curves for two study-case steel frames having member’s
post-yielding stiffness of 1%: (a) F4 and (b) F6.
a) b)
Figure 12. Influence of overstrength for frame F4 with Kp equals 3% for (a) capacity curves and (b) residual
inter-story drift demand hazard curves.
DOI: 10.1002/eqe
taking into account nominal and expected material properties and Kp equal to 3%. As a consequence of
including material overstrength, the lateral strength capacity of the frame F4 is significantly increased.
Next, Figure 12(b) shows a comparison of residual inter-story drift demand hazard curves for the frame
having nominal and expected (overstrength) material properties. Likewise, in prior studies [8, 9],
RIDRmax demands become smaller when overstrength is taken into account than when nominal
properties are assumed in the analyses. However, it is also found that this difference tends to
increase as the mean annual rate of exceedance decreases. A comparison of peak and residual drift
hazard curves for frame F4 considering overstrength and post-yielding stiffness Kp equals 1% and
3% is shown in Figure 13. As expected, the mean annual rate of exceedance of both parameters is
reduced compared with the values obtained when the nominal steel yield stress was assumed in the
frame analytical models. Nevertheless, the results still suggest that for a target IDRmax equals to 3%,
RIDRmax is about 1.0% to have the same annual rate of exceedance of both parameters. These
observations confirm the importance of constraining residual drift demands to avoid the possible
demolition of the buildings even if material overstrength is considered in the seismic assessment.
6. PROBABILISTIC RESPONSE TRANSFORMATION FACTORS
New codified seismic design rules must be capable of leading to consistent reliability levels for the
most frequently used types of complex structural systems. The control of maximum and residual
displacement demands in buildings at the same time when subjected to earthquakes relies on the
development of practically applicable models for the estimation of their expected seismic exceeding
rates of thresholds associated to both parameters (peak and residual displacement). For very
important or special systems, this can be achieved by means of studies of the dynamic response of
nonlinear multi-degree-of-freedom systems. For cases of typical of engineering practice, probabilistic
response transformation factors, TP, can be estimated as the ratio between RIDRmax and IDRmax for a
similar rate to exceed these parameters as suggested in Bojórquez et al. [18, 29, 30] and defined
mathematically as follows:

RIDR max ðlRIDR max Þ
Tp ¼ (5)
IDR max ðlIDR max Þ
In the aforementioned equation, IDR max ðlIDR max Þ and

RIDR max ðlRIDR max Þ are the median peak and
residual inter-story drift associated to a specific mean annual rate of exceedance. The response
transformation factors between RIDRmax and IDRmax can be used as a tool for performance-based or
Figure 13. Peak and residual inter-story drift demands hazard curves for frame F4 taking into account
material overstrength and two levels of member’s post-yield stiffness ratio: (a) Kp = 1% and (b) Kp = 3%.
DOI: 10.1002/eqe
Figure 14. Probabilistic response transformation factors (TP) between RIDRmax and IDRmax related to
different exceedance rates and all the frames under consideration.
reliability-based seismic design procedures or for the improvement toward the next generation of
seismic design methodologies based on structural reliability that explicitly accounts for residual
displacement demands.
Figure 14 shows the numerical results of TP computed for all the frames under consideration. From the
figure, it can be observed that TP follows a decreasing trend that is period dependent, with values in a range
of 0.05 to 0.3 for the whole range of mean annual rate of exceedance, corresponding to return periods from
200 until 10 000 years, which is more than enough for civil engineering structures. In addition, it can be
observed that TP tends to increase as the structural period decrease for lower values of the mean annual
rate of exceedance, which is particularly true for the frame F4 that have the largest values of TP. It
should be noted that frame F10, with the longest structural period of the selected moment-resisting steel
frames, has larger values of TP than the other frames when the mean annual rate of exceedance
becomes large. The transformation factors here assessed could be used as a tool for establishing a new
target limit state (e.g., repairability limit state) aimed at controlling both the maximum and residual
inter-story drift demands for a specific level of mean annual rate of exceedance.
7. CONCLUSIONS
The performance-based assessment of four moment-resisting steel buildings designed accordingly to

the current Mexican design engineering practice and the 2004 MCSDP when subjected to narrow-
band earthquake ground motion records with different levels of intensity were presented in this
paper. For an adequate screening of seismic performance, residual and peak (maximum) inter-story
drift demand hazard curves were computed as part of this investigation. The influence of member’s
post-yield stiffness ratio and material overstrength in the seismic hazard curves was also investigated.
By comparing peak and residual drift demand hazard curves, it was shown that if the study-case
steel structures exhibit peak drift demands of about 3%, they could experience perceptible residual
drifts (i.e., in excess of 0.5%, which is the threshold residual drift that could be perceptible to
human occupants and it could lead to human discomfort identified from recent field investigations
[4]) when subjected to narrow-band earthquake ground motions of high intensity. The amplitude of
residual inter-story drift demands strongly depends on the global post-yield stiffness shown in the
capacity curves (which was controlled by the member’s strain-hardening ratio). Results indicates
that increasing member’s post-yield stiffness ratio (e.g., from 1% to 3%) decreases the residual inter-
story drift demand as the ground motion intensity increases. For the F4 and F6 study-case frames
including member’s post-yield stiffness ratio of 1%, residual inter-story drift demands in the order
of 1% could be reached for peak inter-story drift demands of 3%, which imply that the building
owners might need to decide whether the building be repaired or demolished, on the basis of the
DOI: 10.1002/eqe
repairability limits that take into account technical and economical aspects suggested in [3]. It is noted
that to limit residual drift demands to 0.5% or 1%, peak inter-story drift demands should be limited
to 2% or 3%, respectively, even for steel frames with inherent member overstrength and member’s
post-yield stiffness ratio equal to 3%.
These new findings for structures located in soft-soil sites confirm the importance to recognize the
use and control of residual inter-story drift as performance parameter for seismic design/assessment of
structures in addition to peak inter-story drift demand. For instance, if a new performance level
associated to building’s repairability (i.e., allowing local/global retrofit along with technical and
economical feasibility of straight tilted buildings after earthquake excitation) would be included in
future editions of the Mexican Seismic Design Provisions, probabilistic response transformation
factors (i.e., ratios of RIDRmax and IDRmax for a similar rate of exceeding these parameters) such as
those computed in this investigation are a very useful tool. However, although this provided new
information about residual drift demands that moment-resisting steel frames could experience under
narrow-band earthquake ground motions, it is recognized that similar studies for other type of
structural systems should be carried out to supplement the former proposal.
ACKNOWLEDGEMENTS
While the first author would like to acknowledge financial support given by the Universidad Autónoma de
Sinaloa, through the grant PROFAPI 2012/026, the second author would like to thank the Universidad
Michoacana de San Nicolás de Hidalgo in Mexico. Both authors express their gratitude to the Consejo
Nacional de Ciencia y Tecnología (CONACYT) in Mexico for funding the research reported in this paper.
Finally, the authors wish to express their gratitude to the two anonymous reviewers that helped to improve
the quality of the paper.
REFERENCES
1. Rosenblueth E, Meli R. The 1985 Mexico earthquake: causes and effects in Mexico City. Concrete International
(ACI) 1986; 8(5):23–34.
2. Okada T, Kabeyasawa T, Nakano T, Maeda M, Nakamura T. Improvement of seismic performance of reinforced
concrete school buildings in Japan—part 1: damage survey and performance evaluation after the 1995 Hyogo-Ken
Nambu earthquake. Proceedings of the 12th World Conference on Earthquake Engineering, Auckland,
New Zealand, Paper 2421, 2000.
3. Iwata Y, Sugimoto H, Kuguamura H. Reparability limit of steel structural buildings based on the actual data of the
Hyogoken-Nanbu earthquake. Proceedings of the 38th Joint Panel. Wind and Seismic effects. NIST Special Publication
1057, 2006; 23–32.
4. McCormick J, Aburano H, Ikenaga M, Nakashima M. Permissible residual deformation levels for building structures
considering both safety and human elements. Proceedings of the 14th World Conference on Earthquake Engineering,
Beijing, China, Paper No. 05-06-0071, 2008.
5. Ramirez CM, Miranda E. Significance of residual drift in building earthquake loss estimation. Earthquake Engineering
and Structural Dynamics 2012; 41(11):1477–1493.
6. Pampanin S, Christopoulos C, Priestley MJN. Performance-based seismic response of framed structures including
residual deformations. Part II: multiple-degree-of-freedom systems. Journal of Earthquake Engineering 2003; 7(1):
119–147.
7. Pettinga D, Christopoulos, C., Pampanin S., Priestley MJN. Effectiveness of simple approaches in mitigating residual
deformations in buildings. Earthquake Engineering and Structural Dynamics 2006; 36(12):1763–1783.
8. Ruiz-García J, Miranda E. Performance-based assessment of existing structures accounting for residual displacements.
Report No. 153, The John A. Blume Earthquake Engrg Ctr, Stanford Univ., Stanford, Calif., 2005 (http://blume.
stanford.edu/Blume/TechnicalReports.htm).
9. Ruiz-Garcia J, Miranda E. Evaluation of residual drift demands in regular multi-story frame buildings for performance-
based assessment. Earthquake Engineering and Structural Dynamics 2006; 35(13):1609–1629.
10. Ruiz-Garcia J, Miranda E. Probabilistic estimation of residual drift demands for seismic assessment of multi-story
framed buildings. Engineering Structures 2010; 32:11–20.
11. Uma SM, Pampanin S, Christopoulos C. Development of probabilistic framework for performance-based seismic
assessment of structures considering residual deformations. Journal of Earthquake Engineering 2010; 14(7):
1092–1111.
12. Yazgan U, Dazio A. Post-earthquake damage assessment using residual displacements. Earthquake Engineering and
Structural Dynamics 2012; 41(8):1257–1276.
13. Erochko J, Christopoulos C, Tremblay R, Choi H. Residual drift response of SMRFs and BRB frames in steel buildings
designed according to ASCE 7-05. Journal of Structural Engineering 2011; 137:589–99.
14. Meli R, Avila JA. The Mexico earthquake of September 19, 1985—analysis of building response. Earthquake
Spectra 1989; 5(1):1–18.
DOI: 10.1002/eqe
15. ASCE. Minimum design loads for buildings and other structures. ASCE/SEI 7-05 Including Supplement No. 1,
Reston, VA.
16. MCSDP. Normas Técnicas Complementarias para el Diseño por Sismo. Departamento del Distrito Federal, 2004
(In Spanish).
17. Bojórquez E, Reyes-Salazar A, Terán-Gilmore A, Ruiz SE. Energy-based damage index for steel structures. Steel and
Composite Structures 2010; 10(4):343–360.
18. Bojórquez E, Ruiz SE, Reyes-Salazar A, González TE, Bojórquez J, Terán-Gilmore A. Response transformation factors
for reliability-based earthquake resistant design. Proceedings of the 8th International Conference on Structural
Dynamics, EURODYN 2011 Leuven, Belgium, 4–6 July 2011.
19. Carr AJ. RUAUMOKO—inelastic dynamic analysis program. Department of Civil Engineering, University of
Canterbury: Christchurch, New Zealand; 2008.
20. Bojórquez E, Ruiz SE. Strength reduction factors for the valley of Mexico taking into account low cycle fatigue
effects. 13º World Conference on Earthquake Engineering, paper 516, Vancouver Canada (CD-ROM), 2004.
21. Terán-Gilmore A, Jirsa JO. Energy demands for seismic design against low cycle fatigue. Earthquake Engineering
and Structural Dynamics 2007; 36:383–404.
22. Trifunac MD, Brady AG A study of the duration of strong earthquake ground motion. Bulletin of the Seismological
Society of America 1975; 65(3):581–626.
23. Vamvatsikos D, Cornell CA. Incremental dynamic analysis. Earthquake Engineering and Structural Dynamics 2002;
26(3):701–716.
24. Deierlein GG. Overview of a comprehensive framework for performance earthquake assessment. Report PEER 2004/
05, Pacific Earthquake Engineering Center; 2004; 15–26.
25. Luco N, Cornell CA. Structure-specific scalar intensity measures for near-source and ordinary earthquake ground
motions. Earthquake Spectra 2007; 23(2):357–392.
26. Bojórquez E, Iervolino I. Spectral shape proxies and nonlinear structural response. Soil Dynamics and Earthquake
Engineering 2011; 31(7):996–1008.
27. Bojórquez E, Iervolino I, Reyes-Salazar A, Ruiz SE. Comparing vector-valued intensity measures for fragility
analysis of steel frames in the case of narrow-band ground motions. Engineering Structures 2012; 45:472–480.
28. Alamilla JL. Reliability-based seismic design criteria for framed structures. PhD thesis, Universidad Nacional
Autónoma de México, UNAM. (in Spanish, Advisor: Luis Esteva)
29. Bojórquez E, Ruiz SE, Esteva L. Funciones de transformación de respuestas máximas entre marcos estructurales y
sus correspondientes S1GL con probabilidades de falla similares. XV Congreso Nacional de Ingeniería Sísmica,
Ciudad de México 2005. (In Spanish)
30. Bojórquez E, Ruiz SE,Terán-Gilmore A. Reliability-based evaluation of steel structures using energy concepts.
Engineering Structures 2008; 30(6):1745–1759.
DOI: 10.1002/eqe

10 1002@eqe 2288

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

10 1002@eqe 2288

Uploaded by

Copyright:

Available Formats

EARTHQUAKE ENGINEERING & STRUCTURAL DYNAMICS

Earthquake Engng Struct. Dyn. 2013; 42:1583–1598

Residual drift demands in moment-resisting steel frames subjected

Edén Bojórquez1,*,† and Jorge Ruiz-García2

Received 14 February 2012; Revised 15 January 2013; Accepted 26 January 2013

Copyright © 2013 John Wiley & Sons, Ltd.

2. BUILDING FRAME MODELS EMPLOYED IN THIS STUDY

5.00 m EBF EBF

Figure 1. Plan view of moment-resisting steel frames.

3. SELECTION OF NARROW-BAND EARTHQUAKE GROUND MOTIONS

Table II. Selected narrow-band ground motions.

4. SEISMIC PERFORMANCE OF STUDY-CASE STEEL FRAMES

where m^ lnEDPjSa ðT1 Þ¼sa and s

5. STRUCTURAL RELIABILITY ANALYSIS

5.1. Baseline results

5.2. Inﬂuence of post-yield stiffness ratio

5.3. Inﬂuence of material overstrength

6. PROBABILISTIC RESPONSE TRANSFORMATION FACTORS

In the aforementioned equation, IDR max ðlIDR max Þ and

The performance-based assessment of four moment-resisting steel buildings designed accordingly to

You might also like