Engineering Structures 155 (2018) 345–357

Contents lists available at ScienceDirect

Engineering Structures
journal homepage: www.elsevier.com/locate/engstruct

Seismic fragility of open ground storey RC frames with wall openings for T
vulnerability assessment
Trishna Choudhury, Hemant B. Kaushik⁎
Dept. of Civil Engineering, Indian Institute of Technology Guwahati, Guwahati 781039, India

A R T I C L E I N F O A B S T R A C T

Keywords: Reinforced concrete (RC) buildings with open ground storey (OGS) are characterized by accumulation of large
Masonry-infill reinforced concrete frame lateral displacement at the ground storey. This is due to drastic reduction in relative stiffness of the ground
Open ground storey frame storey compared to the upper stories, a phenomenon known as soft storey effect. Openings present in the ma-
Seismic performance assessment sonry infill walls reduce both the lateral strength and stiffness of the fully infill RC frames. It is a general
Fragility analysis
perception about OGS buildings that openings present in the infill walls reduce the stiffness of upper storeys, and
thus, offsets the soft storey effect. In the present study, this issue is investigated by carrying out a simplified
performance assessment of low- to mid-rise masonry infill RC frames with different infill configurations followed
by fragility analyses. It is observed from fragility analyses that there is practically no influence of openings in
masonry infill walls of OGS frames on lateral load behavior of such frames. OGS frames with any bay and storey
configuration, even with large openings in infill walls, remain highly vulnerable to earthquakes. A new re-
presentation of fragility, Fragility Flow Plot, is proposed, independent of discrete damage states, where results of
the fragility analyses can be shown for different parameters, such as, natural period of vibration, number of bays
and storeys, and openings. The present study contributes towards quantification of seismic fragility and vul-
nerability of OGS building frames and establishes an initiative for seismic fragility based design of OGS frames
using the Fragility Flow Plots.

1. Introduction
Masonry infill reinforced concrete (RC) buildings are commonly
constructed in many countries. Although infills contribute large lateral
strength and stiffness to the building, their influence on lateral load
behavior depends greatly on their distribution in the building. One such
example, where the influence of infill distribution in the frame is
dominant, is an open ground storey (OGS) frame in which the masonry
infill walls are present in all storeys except the ground storey. The
ground storey is left open for various functional purposes, such as,
parking or for shops and services. Providing infills only in the upper
storeys of a building renders the ground storey relatively flexible and
weaker compared to the upper storeys leading to high drift and strength
demands on the ground storey columns. Generally, OGS columns lack
adequate ductility capacity, stiffness, and strength needed to resist the
high demand of storey shear [1]. This leads to an undesirable column-
sway failure mechanism in OGS buildings subjected to earthquake ex-
citations in which plastic hinges are mainly formed in the columns of
the open ground storey. In contrast, the infills restrain most of the
lateral deformation of the upper storey, and thus, little or no damage is
incurred in the upper storeys. Such peculiar behavior of masonry infill
RC frames, in which most of the lateral deformation is concentrated in
the open ground storey and the upper storeys remain vertical and
mostly undamaged, was observed in several past earthquakes (Fig. 1) as
well as past analytical studies [2–7]. Fig. 1 shows two such OGS
buildings collapsed during 2004 Sumatra and 2011 Sikkim earthquake;
both buildings had big openings in infill walls of upper storeys but still
both collapsed. Thickness of infill walls was varying from 115 mm to

Abbreviations: βdsi, normalised standard deviation of the natural logarithm of displacement; βeff, effective viscous damping; β0, hysteretic damping; μd, ductility demand; κ, damping
modification factor as in ATC 40; 3B-4S, three bay-four storey; ADRS, acceleration-displacement response spectrum; C, complete damage state; CSM, capacity spectrum method; db,
diameter of longitudinal steel bar in m; ds, damage state; du, ultimate displacement; dy, yield displacement; E, extreme damage state; ED, energy dissipated by damping; ES0, maximum
strain energy; EDP, engineering demand parameter; FFP, fragility flow plot; FI, fully infill; fy, yield strength of longitudinal steel in MPa; IM, intensity measure; L, half-length of member in
meter; lp, average plastic hinge length; M, moderate damage state; MPA, modal pushover analysis; NB, number of bays; NS, number of storeys; NRHA, nonlinear response history analysis;
NSP, nonlinear static procedures; OGS, open ground storey; Op, central opening in infill walls; PGA, peak ground acceleration; PO, pushover; PP, performance point; S, slight damage
state; Sa, spectral acceleration; Sd , spectral displacement threshold; Sd , spectral displacement demand; Sdy , yield spectral displacement demand; Sdu , ultimate spectral displacement
demand
⁎
Corresponding author.
E-mail addresses: c.trishna@iitg.ac.in (T. Choudhury), hemantbk@iitg.ac.in (H.B. Kaushik).

https://doi.org/10.1016/j.engstruct.2017.11.023
Received 4 June 2017; Received in revised form 12 August 2017; Accepted 9 November 2017
Available online 22 November 2017
0141-0296/ © 2017 Elsevier Ltd. All rights reserved.
T. Choudhury, H.B. Kaushik Engineering Structures 155 (2018) 345–357

Fig. 1. Elevation, column layout and section of collapsed OGS buildings during past earthquakes: (a) a three storey building in Port Blair, India after 26 December 2004 Sumatra
earthquake, and (b) a four storey building at Singtam market in Sikkim after 18 September 2011 Sikkim earthquake (Photos by Hemant B. Kaushik).

230 mm in both the buildings. The sectional details and plan of the quantification of masonry infill RC frames is discussed.
buildings show that the columns were inadequately detailed and almost
all the ground storey columns failed. Considering this fact, past re-
2. Seismic fragility assessment
searchers (e.g., [5,8–14]) have reported the significant ‘negative’ in-
fluence of OGS on the overall seismic performance of such buildings. On
The primary motivation of seismic fragility assessment is to obtain
several forums (for example, [15]), various stakeholders have expressed
an estimate of the probability of exceedance of a given damage level in
that it is a common perception that presence of openings in infill walls
a building due to a seismic hazard to predict its vulnerability. The basic
of OGS frames reduce the seismic fragility of such frames.
steps include building’s capacity estimation for a given seismic hazard,
The seismic design codes of different countries are silent on specific
followed by the fragility estimation for a given limit state or damage
design procedure to be followed for open ground storey buildings.
state. The interest is to quantify fragility of OGS frames in comparison
Some, for example, the Indian seismic code [16], require the ground
to fully infill or bare frames under lateral loads using a consistent ap-
storey columns of such frames to be designed for higher member forces
proach. Though recent past studies have used nonlinear response his-
simply by using a multiplication factor of 2.5. However, it is reported
tory analysis followed by more rigorous incremental dynamic analysis
that use of a single multiplication factor for any type of OGS frame does
for estimation of seismic fragility of structures [21], nonlinear static
not reduce its vulnerability [5,17,18]. Previous studies, such as, those
procedures (NSPs) followed by simplistic HAZUS [20] procedure have
of Cosenza et al. [19] provide information on the influence of varying
also been effectively used in the past.
building parameters, e.g., number of bays and storeys on lateral load
Pinho et al. [22] evaluated different nonlinear static procedures
behavior of buildings.
(NSPs), such as, capacity spectrum method (CSM) of ATC 40 [23] and
There is a lack of specific literature that can assist in seismic as-
N2 method [24]. It was concluded on the basis of the comparative study
sessment of OGS buildings with openings in infill walls. Influence of
involving nonlinear response history analyses that all the NSPs are
openings in infill walls on the overall seismic vulnerability of OGS
capable of effectively predicting the lateral displacements. The effec-
frames has not been investigated yet. The primary objective of the
tiveness of nonlinear static methods in predicting lateral load response
present study is to quantify the fragility of low- to mid-rise OGS frames
has also been validated with respect to those obtained by nonlinear
and demonstrate their vulnerability in comparison with corresponding
dynamic analysis for a wide range of frames by Bosco et al. [25]. It was
FI and bare frames. An attempt was made to provide the designers with
observed that differences in seismic demand occurred in the two
an easy interface to carry out fragility analysis that can assist in fragility
methods when the frames are well excited into the inelastic range but
based design of OGS buildings. In order to achieve these objectives,
they are nearly always conservative in nonlinear static procedures.
nonlinear static analyses of OGS frames were carried out by varying the
Calvi et al. [26] discussed pros and cons of various empirical and
number of bays (1–6) and storeys (2–6), size of central opening in infill
analytical methods of vulnerability assessment, and observed that none
walls (0–90%), and seismic hazard in terms of peak ground accelera-
of the methods satisfy all the requirements necessary for an optimum
tion, PGA (0.05–1.0 g). As per HAZUS [20], low-rise and mid-rise
vulnerability assessment methodology. Kaushik and Choudhury [27]
concrete moment frames (with or without masonry infills) are those
elucidated a step-by-step procedure for estimation of seismic fragility
with 1–3 storeys and 4–7 storeys high, respectively. A new re-
and vulnerability of structures considering different performance as-
presentation of fragility termed as Fragility Flow Plot (FFP) was de-
sessment methodologies.
veloped. These plots are independent of consideration of any damage
Fragiadakis et al. [28] observed that all NSPs are good predictor of
states and can combine the results of several fragility analyses and re-
lateral displacement and storey drifts for RC buildings of up to four
sults of parameterized performance assessment for a given earthquake
storeys. Similarly, Dolšek and Fajfar [29] used simplified nonlinear
intensity. Further, possible use of these plots in seismic fragility
static procedures for seismic analysis and fragility quantification of

346
T. Choudhury, H.B. Kaushik
infilled reinforced concrete frames. Further, it has been reported that all
NSPs, including MPA, give comparable estimate of building response
where a soft or open first storey exists [30]. Therefore, for performance
assessment of the building frames, NSPs including capacity spectrum
method [23], which is also recommended in ASCE [31], is used in the
present study. Shinozuka et al. [32] derived fragility curves after car-
rying out structural assessment using capacity spectrum method and
found it to be quite adequate in comparison to nonlinear time history
procedures considering the raft of assumptions underlying it.
The present study considers the seismic hazard in the form of re-
sponse spectrum specified in the Indian Seismic code [16] for assessing
the seismic performance. The response spectrum is effectively reduced
for hysteretic effects through standard procedures mentioned in ATC 40
[23]. To consider various levels of seismic intensity, peak ground ac-
celeration (PGA) is varied from 0.05 g to 1 g. In the present study,
seismic fragility analyses are carried out by defining damage states for
each frame and finally interpreting the probability of exceedance of
each damage state based on HAZUS [20] methodology, which has been
efficiently used in several past studies for fragility estimation (e.g.,
[33,34]). Prerequisites for fragility estimation include nonlinear static
or dynamic analyses of the frames followed by structural performance
assessment. SAP 2000 [35] is used in the present study for nonlinear
static analyses, and performance assessment is carried out as per the
procedure laid down in ATC 40 [23], which is also recommended in
ASCE [31].
2.1. Structural modeling
Three different building frame typologies are considered in the
present study: bare frames, OGS frames, and FI frames with different
number of bays (NB) and storeys (NS) with varying sizes of central
opening in infill walls (Op), intended to cover low- to mid-rise build-
ings. Typically, all the frames have a bay width of 3 m and a ground
storey height of 4.4 m; all upper storeys are 3.2 m high. For bare frames,
full weight of the infill walls in all storeys is considered in order to
reflect a commonly adopted analysis and design procedure in which the
strength and stiffness of masonry infill walls are not considered. For
OGS and FI frames, strength and stiffness as well as the weight of ma-
sonry infills are considered. Typical structural plan, elevation, and
member cross sectional details of a 3 bay – 4 storey (3B – 4S) frame are
shown in Fig. 2.
The compressive strength of concrete cubes of 150 mm size is con-
sidered as 25 MPa (Elastic Modulus: 25000 MPa) and the reinforcing
bars have expected yield stress of 450 MPa (Elastic Modulus: 200 GPa).
Mander’s confined concrete model [36] is used to characterize the
stress-strain curve of concrete. Idealized stress-strain model proposed
by Kaushik et al. [37] is used to model the material nonlinearity in
masonry considering masonry prism strength of 4.1 MPa and elastic
modulus of 2255 MPa. Fig. 3 shows location of nonlinear regions
Fig. 2. (a) Building floor plan, (b) Structural elevation @ A-A (considered frame), and detailed sectional properties of (c) beams and (d) columns for the frames.
347
Engineering Structures 155 (2018) 345–357
considered in the frames along with the plastic hinge characteristic of a
moment hinge in RC members and axial hinge in masonry infill wall
(modeled as equivalent diagonal strut). The infill walls are 220 mm
thick with strut width equal to one-fourth of the diagonal length of the
wall [38]. The reduced width of the strut elements due to presence of
opening in infill walls is obtained using Eq. (1) [39], where r0 is the
ratio of opening area to the area of fully infill (FI) wall panel. Eq. (1)
was developed based on experimental results of several frames with
different opening sizes in masonry infill walls.
Reduced strut width = (1−r00.675) × (original strut width) (1)
RC members are modeled as 2-noded line elements with rigid beam-
column joints. Similarly, masonry infill walls are modeled as 2-noded
line elements (equivalent diagonal struts) with moment releases at both
ends. Weight of masonry infill walls is considered as uniformly dis-
tributed load on the beams (weight density of masonry = 18 kN/m3).
Columns are assumed to be fixed at their bases without accounting for
soil structure interaction. Material nonlinearity in RC members is in-
corporated in the elements using lumped plasticity approach. Plastic
hinges are assumed to form at a distance equal to one-half of the
average plastic hinge length, lp from the member ends, calculated (in
meter) using: lp = 0.08L + 0.022dbfy [38]. Here, L is half-length of
member in meter (assuming point of contra-flexure at member mid-
point), db is the diameter of longitudinal steel bar in m, and fy is the
yield strength of longitudinal steel in MPa. In RC members, flexural
plastic hinge properties (Fig. 3b) are assigned based on Table 6–7 and
6–8 of FEMA 356 [40]. Plastic hinges in masonry infill walls are as-
sumed to form at the center of the strut and the plastic hinge length is
generally considered as one-half to three-fourths of the length of the
strut. These assumptions are followed on the basis of results obtained in
past analytical and experimental studies [5,8,41]. Also, it is observed in
the present study that variation in location of plastic hinge and its
length in diagonal struts does not result in any significant variation in
the nonlinear response of the frames. Hence, same analytical model has
been consistently used to demonstrate the influence of openings in
masonry infill walls on seismic fragility quantification.
2.2. Seismic performance assessment
Seismic performance of a structure can be assessed with different
level of computational efficiency depending upon available data and
resources. The most common practice is to use either nonlinear static
analysis procedures or nonlinear response history analysis (NRHA)
procedures. Several studies in the past have used NSPs for structural
assessment of masonry infill RC frame buildings (e.g., [5,12,14,25,33]).
In the present study, an initial structural assessment of limited frames of
all three types was carried out using methods given in ATC 40 [23],
FEMA 440 [42], along with N2 method [43], and those suggested by
Chopra and Goel [44], and found that the performance points for the
T. Choudhury, H.B. Kaushik
Fig. 3. (a) RC members with moment hinges and equivalent strut model with axial hinges, (b) Typical moment rotation curve for RC members, (c) Axial stress – strain model for masonry.
Fig. 4. Comparison of capacity curves obtained from NL static and dynamic analyses of (a) bare, (b) OGS, and (c) FI frame.
considered type of frames come out to be quite comparable by all the
methods. Further, NRHA of each of the three types of frames (bare, OGS
and FI) was carried out for sixteen samples of ground motion ElCentro
scaled for different PGA values. The samples of ground motion differ in
their frequency content, modulation and energy content. The pushover
curves obtained for the three frames using nonlinear static analyses are
compared with that of the nonlinear dynamic analysis or response
history output of the three frames plotted in terms of maximum roof
displacement and corresponding base shear (Fig. 4). It is observed that
the pushover curves of all the three frames match quite well with the
response history output; in fact, the pushover curves represent more or
less the envelop response for all the three frames. Therefore, NSPs are
used in the present study for performance assessment of such buildings
primarily because of the simplicity of these methods and requirement of
less input data and less computational time as compared to NRHA.
CSM is an acceptable method recommended in several documents
[23,31] that gives detailed analysis procedure for frames with masonry
infill walls. In CSM [23], the hysteretic behavior of the structural
system under cyclic load, including stiffness and strength degradation,
is considered by equivalent viscous damping, which is defined in terms
of energy dissipated by the structure in a single cycle of motion, i.e., the
area enclosed by a single hysteresis loop. In addition, a factor (κ) is used
to simulate the imperfections in the hysteresis loops. The capacity curve
and elastic response spectrum (hazard curve), both in Acceleration-
Displacement Response Spectra (ADRS) format, are considered initially
in the CSM. The capacity curve is bi-linearized based on equal dis-
placement rule, and the elastic response spectrum is damped to an ef-
fective damping present in the structure by using standard procedures
mentioned in ATC 40. Bi-linearization is needed to estimate the effec-
tive damping and appropriate reduction of spectral demand. The ef-
fective viscous damping (βeff) is determined as βeff = κβ0 + 5, where β0
represents the hysteretic damping calculated as β0 = (1/4π)(ED/ES0),
where, ED and ES0 are energy dissipated by damping and maximum
strain energy, respectively. Finally, the intersection of the capacity
curve with the reduced (or damped) response spectrum, known as de-
mand curve, gives the performance point (PP) of the structure for the
348
Engineering Structures 155 (2018) 345–357
considered elastic demand (hazard). The PP represents the spectral
displacement demand and spectral acceleration on the structure for a
given hazard. To represent a worst case scenario, type “C” building
category of ATC 40 is considered in the present study for which the
shape of hysteresis curve is pinched (damping modification factor, κ =
0.33) and maximum allowable effective damping (βeff) is limited to
20%. The PPs representing spectral displacement (Sd) demand on the
structure are obtained for each frame for different PGA.
With the objective of studying the influence of Op on lateral load
response of frames, initially a 3B – 4S frame is considered, and Op is
varied from 0% to 90% of the infill area for both OGS and FI frames. As
discussed later, based on these results only two different values of Op
(0% and 50%) are considered in other frames (in which NB and NS are
varied). Pushover analyses are carried out for different frames con-
sidering the fundamental mode of vibration as the lateral deformation
profile.
2.3. Influence of central opening size in infill
The PO curves obtained for 3B-4S OGS frames and FI frames with
different Op are shown in Fig. 5(a) and (b), respectively, and are
compared with the PO curve obtained for 3B-4S bare frame. It is ob-
served from the comparative PO curves that the peak lateral load car-
rying capacity of FI frames, even with Op up to about 50%, is much
higher (nearly 4 to 5 times) than that of the OGS frames as well as bare
frames. In the case of FI frame, the presence of infills in the ground
storey imparts tremendous stiffness and strength to the frame as seen
from Fig. 5(b). Infills being brittle, fail after a certain level of de-
formation in the frame. Thus, the PO curve for any infill frame initially
shows very high stiffness and strength that suddenly drops after
reaching the ultimate capacity of the infills. This behavior of the FI
frame is evaluated in the present study by increasing the opening sizes
in the infill panels of all the storeys. The sudden drop in strength and
stiffness is observed in the FI frame until an opening of 50% is provided
in the infill. When the opening size is further increased to about 60%,
the stiffness and strength of FI frame under the action of lateral loads
T. Choudhury, H.B. Kaushik
Fig. 5. PO curves for (a) OGS and (b) FI frames with different central opening sizes.
reduce gradually rather than exhibiting a sudden drop. This is because
of a balance achieved between the infill wall resistance and the RC
frame resistance when 60% or more Op are provided in FI frame. Such a
balancing behavior leads to sharing of lateral loads between the infill
walls and the RC frame from the beginning. This in turn prevents the
sudden drop in the lateral stiffness and strength as observed before.
Thus globally, a relatively more ductile behavior is achieved for the
considered frame as the opening size in infill is increased to about 60%.
Therefore, it can be inferred that the performance of the FI frame with
60% Op is even better than the FI frame without Op as well as the bare
frame. Though, such a behavior is observed in all the frames analyzed
in the present study, the observed optimal value of Op must be used
with caution as it may change due to presence of non-structural
members and other components, which are not considered in the ana-
lysis. Similarly, the results are subject to several modeling and other
limitations as stated later in the section on assumptions and limitations
of the study.
It is also observed that with an increase in Op, the lateral strength
and stiffness of FI frames decrease. This effect of Op is, however, found
to be negligible in the case of OGS frames as compared to the FI frames.
Thus, it is implied that the upper storey infills do not contribute sig-
nificantly to the lateral stiffness and strength of the OGS frame. Similar
observation is made for frames with other configuration of NB and NS.
Since the study focuses primarily on OGS frames and the effect of
openings is negligible on the behavior of OGS frames, further analyses
are carried out only for 0% Op, i.e., no opening in infill, and 50% Op.
Fig. 6 shows response surface plot of PPs for 3B-4S OGS and FI
frames with varying Op and PGA. With an increase in Op in OGS frames,
spectral displacement demand (Sd) decreases marginally for a given
PGA, whereas, it increases for FI frames. It is, however, noteworthy that
the Sd on OGS frames (with any Op) is always more than that for FI
frames with same Op for a given PGA. A comparison of ductility de-
mand (µD) for OGS and FI frames for three different Op (Table 1) also
Fig. 6. Response surface plots of performance points in terms of Sd for (a) OGS and (b) FI
frames with different Op and PGA.
349
Engineering Structures 155 (2018) 345–357
verifies this observation. It is interesting to note that most of the FI
frames (with 50–60% Op) are not expected to undergo nonlinear be-
havior for PGA of 0.36 g (maximum design basis earthquake as per the
Indian seismic code [16]) as the µD is more or less equal to one. On the
other hand, large ductility demand is imposed on OGS frames, most of
which is concentrated in ground storey columns. Similar trend is ob-
served in Sd and ductility demand (µD) for other PGA levels also.
2.4. Influence of number of storeys (Ns)
NS is varied from two (2S) to six (6S) for both OGS and FI frames
considering 0% and 50% Op. The PO curves and capacity spectra are as
shown in Figs. 7 and 8, respectively, with varying NS for 3B OGS and FI
frames with 0% Op. PO curves for OGS frames indicate that there is a
negligible decrease in initial stiffness but significant increase in lateral
strength of the 3B frames with an increase in NS. Capacity spectra show
the dominating influence of mass, since spectral acceleration, Sa de-
creases with an increase in NS. Thus, the effect of change in mass is
more significant than the change in stiffness due to varying NS, keeping
NB constant. In case of FI frames, with an increase in NS, both strength
and stiffness decrease as shown in Fig. 7(b). Again, the decrease in
initial stiffness is significant in case of FI frames as compared to the
corresponding OGS frames, implying that the ground storey infill walls
have a significant influence on the lateral load performance of the FI
frames only. Similar trend in the PO curves and capacity spectra are
observed for both OGS and FI frames with any other Op and NB.
2.5. Influence of number of bays (NB)
The influence of variation in number of bays (NB) on the lateral load
response of RC frames considering 0% and 50% Op, is studied by
varying NB from one (1B) to six (6B) for both OGS and FI frames. The
variations in PO curves and capacity spectra due to change in NB for a
4S frame are shown in Figs. 9 and 10, respectively. Fig. 9 indicates that
with increase in NB for a given Ns (here 4S), both strength and stiffness
increase for the OGS as well as FI frames. Capacity spectra (Fig. 10(a))
show the dominating influence of mass in the case of OGS frames, since
Sa decreases with increasing NB. The change in Sa goes on diminishing
with increasing NB. In case of FI frames, Fig. 10(b) shows that with an
increase in NB, Sa increases moderately for 1B and 2B frames after
which the change is negligible. Clearly, the influence of mass is less on
the lateral load behavior of both OGS and FI frames when NB is varied
in comparison to those cases where NS is varied keeping the other
parameter constant. Similar variations are observed in other frames
considered in the study as well with any Op and NS.
T. Choudhury, H.B. Kaushik
Table 1
Ductility demand (µD) on OGS and FI frames for a PGA of 0.36 g.
Frames 0% Op
Sdy (m) Sdu (m) µD
FI 0.013 0.012 0.96
OGS 0.021 0.059 2.81
2.6. Seismic damage and fragility assessment
Seismic damage assessment is the numerical quantification of
probable damage to a building due to a given hazard. Damage states are
specified in order to physically categorize the degree of damage to
members of a building. Several damage indices suggested in the past
literature suggest damage state thresholds in terms of different para-
meters (e.g., strength, ductility, energy dissipation). A particular da-
mage state and the corresponding damage accumulated in the structure
are correlated through these damage indices, and for the same frame,
there may exist several non-identical damage indices. Damage states
were also expressed in the past in terms of capacity curve parameters,
such as, yield displacement (dy) and ultimate displacement (du) [45,46].
Other parameters, such as, maximum base shear capacity [47] or inter-
storey drift ratio [20] have also been used to quantify the damage.
Fig. 11 shows an idealized capacity curve with four damage states,
thresholds of which are demarcated by vertical lines based on yield and
ultimate displacement.
Based on [46], four damage states considered in the present study
are: Slight (S), Moderate (M), Extreme (E), and Complete (C). For de-
fining the damage state thresholds, yield displacement (dy) is defined as
the initiation of first plastic hinge in any ground storey columns. For
bare and OGS frames, failure point or ultimate displacement (du) is
defined as a point on the pushover curve when at least 50% of the
ground storey columns exceed hinge rotation corresponding to life
safety limit state in FEMA 356 [40], i.e., the plastic rotation capacity is
less than the corresponding demand calculated from the inelastic ana-
lysis. The life safety limit state corresponds to a displacement or da-
mage level in which the columns reach the deformation equal to three-
fourth of the deformation specified for the collapse prevention level,
i.e., before shear strength of columns is reached. Physically, the failure
point or the ultimate displacement (du) of the frame under considera-
tion corresponds to a global lateral displacement of the frame at which
at least half of the ground storey columns reach deformation levels of
life safety limit state. For fully infill (FI) frames, failure of the ground
storey infill walls is an additional criterion considered to obtain the
ultimate displacement capacity of the frame. At this level of lateral
global displacement, at least half of the ground storey columns undergo
Fig. 7. Influence of variation in NS on PO curve for (a) OGS frames and (b) FI frames for a 3 bay (3B) frame with 0% Op.
Engineering Structures 155 (2018) 345–357
50% Op 60% Op
Sdy (m) Sdu (m) µD Sdy (m) Sdu (m) µD
0.018 0.018 1.00 0.019 0.021 1.08
0.021 0.051 2.43 0.021 0.05 2.38
severe damage with some of them might even reach collapse prevention
state (in case of bare and OGS), and in addition, all the masonry infill
walls of the ground storey fail (in case of FI frames). The criterion of
50% or more “failures” of columns in a storey is an assumption to strike
a balance between the usual over conservative approach, according to
which failure coincides with failure of the first lateral load resisting
member, and the reality witnessed after earthquakes that buildings do
not collapse even after a number of lateral load resisting members fail
[33].
Given the damage state thresholds and PP of the frames for different
PGA, seismic fragility can be obtained using Eq. (2) [20].
1 ⎛ Sd ⎞ ⎤
P[ds|Sd] = Φ ⎡
⎢ β ln ⎜ Sd,dsi ⎟ ⎥
(2)
⎣ dsi ⎝ ⎠⎦
Here, P[ds|Sd] represents the conditional probability of being in, or
exceeding, a particular damage state (ds), given the spectral displace-
ment demand (Sd) of the performance point, Ф is the standard normal
cumulative distribution function, and βdsi is the normalized standard
deviation of the natural logarithm of the displacement threshold (Sd,dsi )
indicating uncertainties in capacity curves, damage levels, modeling
errors, and seismic hazard.
Estimation of uncertainty in the fragility analysis with some degree
of confidence is extremely difficult due to unavailability of highly
variable material properties of masonry, large variation in construction
methods, and scarcity of ground motion records, especially in devel-
oping countries like India. Therefore, a preliminary analysis was first
carried out for the considered type of frame, material, and seismic
hazard for obtaining a range of βdsi values required to be considered in
the fragility analysis. It was observed in the preliminary analysis that
the variance in probability of exceedance of a particular damage state
in the considered frames diminishes with increasing values of βdsi,
especially for βdsi values above 0.7 for higher displacement thresholds
[27]. On the other hand, the variance in probability of exceedance of a
particular damage state does not change much with changing βdsi for
lower displacement thresholds, and again it starts diminishing after βdsi
reaches a value of about 0.7. While FEMA P-58 [48] suggests a default
value of 0.6 for βdsi, HAZUS [20] suggests values of 0.75, 0.70, and 0.65
350
T. Choudhury, H.B. Kaushik
Fig. 8. Influence of variation in NS on capacity spectra for (a) OGS frames and (b) FI frames for a 3 bay (3B) frame with 0% Op.
for buildings designed to old, moderate, and modern codes, respec-
tively. These uncertainty values are obtained as a combination of un-
certainty arising from variation in capacity, seismic demand, and da-
mage state thresholds, thus taking care of all the parameters affecting
fragility estimates. The suggested values of βdsi for mid-rise RC bare
frames designed for moderate code level is 0.7 for all damage states
except for the complete damage state for which the value is 0.89 [20].
However, βdsi for masonry infill RC frames designed for moderate code
is not recommended by HAZUS [20]. Also, if different variability as-
sociated with each damage state is considered in the analysis, un-
realistic fragility curves (for instance, intersecting) result in cases where
median values are closely spaced [49]. Therefore, results of the fragility
analysis for all the frames considered are shown in the manuscript for a
βdsi value of 0.7 as an approximate value of beta corresponding to each
damage state. However, the seismic fragility can be estimated for other
values of uncertainties as discussed in the paper. Moreover, it is im-
portant to note that the observations and conclusions made in the
present study will remain valid for any value of uncertainty.
3. Seismic fragility curves
Seismic fragility curves with respect to spectral displacement as an
intensity measure (IM) are developed for the 3B-4S frames (bare, OGS,
and FI) with Op varying from 0% to 90%. For reasons discussed later in
this section, seismic fragility for frames with other combinations of NB
and NS are developed only for 0% and 50% Op.
Fig. 9. Influence of variation in NB on PO curves for (a) OGS and (b) FI frames for a four storey (4S) frame with 0% Op.
351
Engineering Structures 155 (2018) 345–357
3.1. Influence of central opening size in infill
Fig. 12 shows the fragility curves obtained for 3B-4S OGS and FI
frames with different Op with Sd as the IM. These fragility curves are
also compared with those obtained for 3B-4S bare frame. In both the
OGS and FI frames, though the seismic fragility decreases with in-
creasing Op, the decrease is not considerable in case of OGS frames. The
FI frames are very stiff and hence undergo much less lateral displace-
ment compared to a relatively flexible bare and OGS frame. As seen
from the comparison of pushover curves, the lateral load behavior of FI
frames also changes drastically with increase in opening sizes in infills
(Fig. 5b). There occurs a significant reduction in both the lateral
strength and stiffness of the FI frames with increasing opening sizes in
infill walls. Since there is a large change in the flexibility of the frames
with changing opening sizes in infills, there occurs a large change in the
displacement capacity (ultimate displacement before failure) of the
frames. The displacement capacity of the FI frames with larger opening
size is higher compared to one with a lesser opening size. Again, a
greater displacement capacity implies lower fragility. In contrast, in
OGS frames, the change in opening size in infills merely affects the
overall performance of the frame. There is an insignificant reduction in
the stiffness as well as strength of OGS frames with increase in opening
size in infills of upper storeys (Fig. 5a). This implies that the displace-
ment capacity, and hence the seismic fragility of the OGS frames, do not
alter much due to opening size.
Further, although the lateral load carrying capacity of the bare and
OGS frames is nearly the same (Fig. 5), OGS frames, even with 90% Op,
are more fragile compared to the bare frames. On the other hand,
T. Choudhury, H.B. Kaushik Engineering Structures 155 (2018) 345–357

Fig. 10. Influence of variation in NB on capacity spectra for (a) OGS and (b) FI frames for a four storey (4S) frame with 0% Op.

however, the change is marginal with respect to NB. Almost all the
configurations of frames with different combinations of NB and NS have
high probability of undergoing slight and moderate damage, but the
probability of undergoing complete damage is more for frames with
higher NB and NS. Similar trends are observed for OGS frames with
other Op and PGA levels, for example, as shown in Fig. 14(b) and (c) for
OGS frame with 0% opening and PGA of 0.24 g, and for OGS frame with
50% opening and PGA of 0.36 g, respectively.

4. Seismic fragility flow plot

Damage is a progressive and irreversible process, and discrete

quantification of damage (such as those given by [46]) based on its
Fig. 11. Discretized damage states defined on an idealized capacity curve. intensity with the help of strength or displacement parameters brings
uncertainty into the picture. In order to represent the damage and as-
sociated fragility in a continuous manner, the concept of Fragility Flow
seismic fragility of FI frames with any Op is higher than that of the bare
Plot (FFP) is introduced (Fig. 15), where the damage is represented in a
frame for slight and moderate damage states. However, for complete
continuous form without any discretization of damage states. FFP is a
damage state, fragility of FI frames with about 60% Op nearly coincides
bivariate plot of seismic fragility (shown as radial lines) with respect to
with that of the bare frame.The faster rate of reduction in the fragility of
displacement threshold (Sd ) and displacement demand (Sd). FFPs relate
FI frames with increasing Op relative to OGS frames is because of faster
several parameters (natural period of vibration, number of storeys,
rate of increase in displacement threshold (Sd ), which in turn depends
number of bays, displacement demand, displacement capacity of the
on dy and du of FI frames. With an increase in Op, both dy and du in-
building, probability of exceeding a building function limit state) re-
crease in OGS and FI frames (Fig. 13), however, the rate of increase is
quired for rapid quantification of fragility of structures for a given level
much higher in FI frames. When the Op is increased from 0% to 90%,
of uncertainty and seismic hazard. FFPs provide more information than
the increase in dy is 93% and 23%, and increase in du is 42% and 9% in
that provided by the parameterized fragility discussed in the past lit-
case of FI and OGS frames, respectively. The dy and du values in FI frame
erature (e.g., [50–52]). In addition to providing a direct relation be-
exceeds the OGS frame values after Op of about 50% and 20%, re-
tween seismic capacity and seismic displacement demand for a given
spectively. In fact, the du value in the FI frame with 60% Op is similar to
uncertainty and probability of exceeding a building function limit state,
that in the bare frame value due to which their fragility also coincides.
FFPs combine results of parameterized performance assessment for a
Beyond 60% Op, the du value in FI frame exceeds the bare frame value,
given earthquake intensity.
and hence, the fragility of FI frame is lower than that of the bare frame.
Therefore, FFP is useful in multiple ways: for instance, if the dis-
From the fragility analyses of OGS and FI frames, it is observed that
placement threshold is known for a given seismic demand on any
for any combination of NB and NS, the FI frames have a faster rate of
structure, the user can graphically quantify probability of exceedance of
reduction in seismic fragility for higher damage states compared to the
the displacement threshold value. Thus, FFPs give a continuous flow of
corresponding OGS frames. In addition, as already discussed, because of
probability of exceedance for any damage threshold and for a given
higher displacement threshold values for FI frames, the fragility of
spectral displacement demand. Moreover, for a given seismic fragility
complete damage for FI frames with different NB and NS reduces even
of some damage level, if the seismic displacement demand is known (Y-
below the fragility of bare frame when 50–60% Op is provided in FI
axis), then the threshold or limit of displacement capacity that should
frames. This is not so in the case of OGS frames, and thus, the fragility
be available in the structure can be determined from the X-axis. The
of OGS frames always remains higher than that of the bare frame with
structure can then be designed for that minimum displacement capacity
any Op and for any NB and NS with respect to Sd as intensity measure.
to avoid damage due to a seismic event that imposes a demand, Sd on
the structure. In this way, one can possibly include seismic fragility in
3.2. Influence of number of bays and storeys building design procedures.
A few past studies (e.g., [50–52]) suggested probabilistic seismic
Fig. 14a shows contour plot for seismic fragility obtained for OGS demand models in the form: ln(EDP) = a + b × ln(IM), where EDP is
frames with 0% Op with respect to NB and NS for a PGA of 0.36 g. any engineering demand parameter, IM is the intensity measure para-
Seismic fragility is found to increase with increasing NB and NS, meter, and a and b are constants to be determined from regression

352
T. Choudhury, H.B. Kaushik Engineering Structures 155 (2018) 345–357

Fig. 12. Comparison of fragility curves for 3B-4S OGS and FI frames with varying Op.

analysis. This model was then related to damage parameters in order to range of parameters. Moreover, results of such linear fits cannot be
develop damage fragility curves. Unlike FFPs, such models cannot extrapolated beyond the range of data available, and paucity of ex-
combine all the above mentioned parameters, and therefore, cannot perimental or analytical data increases the uncertainty associated with
give an overall picture of the seismic fragility of buildings considering a the damage models. Though use of the FFPs is also subject to some

353
T. Choudhury, H.B. Kaushik
Fig. 13. Change in (a) yield and (b) ultimate displacement with Op for OGS and FI frames.
limitations and assumptions discussed later, FFPs can be useful for
rapid and large-scale fragility estimation of buildings.
Fig. 15 shows the proposed FFP for OGS frames with 0% Op in-
cluding a scatter plot of displacement demand and capacity for different
damage states (S, M, E, C) for a PGA of 0.36 g. Here, the radial lines give
the conditional probability of exceedance for a given threshold. The
plot covers buildings from low height (2S) to mid-height (6S) having a
natural period in the range of 0.36 s to 0.81 s, thus, including a wide
range of buildings constructed over hard soil type as per the Indian
seismic code [16]. Similar curves can be plotted for different values of
Op, βdsi, and PGA. The dotted horizontal lines delineate the natural time
period obtained from modal analysis of the frames analyzed for dif-
ferent combination of NB and NS. The color bar represents the value of
conditional probability of exceedance of a given damage threshold
based on Eq. (2). The blue, green, yellow, and red color markers in the
scatter represent PPs obtained for the four damage states – S, M, E, and
C, respectively. Each set of scatter plots (for individual damage states)
consists of five points showing variation in fragility for frames with
different NB (2B to 6B) keeping NS constant. Further, such set of scatter
plots are shown for five different NS (2S to 6S).
A wide range in displacement threshold values is observed in be-
tween extreme and complete damage states because of consideration of
discrete damage states from literature [46]. For instance, the observed
dy and du values from the analysis of 3B-4S OGS frame with 0% Op are
21 mm and 79 mm, respectively, using which the displacement
threshold values for S, M, E, and C come out to be 15 mm, 21 mm,
36 mm, and 79 mm, respectively (Fig. 11). These threshold limits are
marked as circles (for 3B-4S) in different color fills for the four damage
states in the FFP developed for the frame (Fig. 15). The relative dif-
ference between the displacement threshold values between S-M, M-E,
Fig. 14. Contour plot for seismic fragility showing influence of change in NB and NS for OGS frames with (a) 0% Op and for a PGA of 0.36 g (b) 0% Op for PGA of 0.24 g, and (c) 50% Op
for PGA of 0.36 g.
354
Engineering Structures 155 (2018) 345–357
and E-C damage states are 6 mm, 14 mm, and 43 mm, respectively.
Clearly, a wide range of displacement thresholds between E and C has
no description of damage, and hence seismic damage as well as fragility
cannot be determined for any displacement threshold value between E
and C using conventional fragility curves. FFP provides an advantage of
determining the seismic demand as well as fragility for any intermittent
displacement threshold values. For the 3B-4S OGS frame, the spectral
displacement demand (58 mm) and the seismic fragility in terms of
probability of exceedance for different damage state thresholds (nearly
100% for S, 92% for M, 75% for E, and 34% for C) can be picked out
easily using the FFP. It is also interesting to note a slight variation in
fragility with changing NB, keeping NS constant (as shown by each
group of scatter plot). This variation is more pronounced for higher NS
(>4S). The variation in fragility is, however, quite noticeable when the
NS is varied keeping NB constant. Such comparison of bivariate data
with other parameters, such as, natural period, NB, and NS, is possible in
FFP. Fig. 15 can also be used as a design aid for integrating fragility
with the design process. It is possible to estimate the range of Sd and Sd
from FFP for given values of NB, NS, and fragility level. Inversely, for an
assumed displacement threshold (e.g., design lateral drift in seismic
codes) required in a frame to be designed, the number of maximum
bays and storeys can be estimated from FFP for considered seismic
fragility.
Fig. 16(a) shows another FFP for the four damage states in 3B-4S
OGS frame with Op varying from 0% to 90% for a PGA of 0.36 g and
with βdsi of 0.7. The horizontal discontinuous lines represent the outer
bounds of the scatter, i.e., for 0% and 90% Op for different damage
states. As already discussed in conventional fragility curves, seismic
fragility decreases uniformly, though marginally, from 0% to 90% Op
for each damage state. A large gap is again observed between the E and
T. Choudhury, H.B. Kaushik
C damage states covering wide displacement threshold values of about
0.045 m to 0.08 m. Seismic fragility for such missing displacement
threshold range can be estimated easily using the proposed FFP. While,
the conventional fragility curves (Fig. 12) give information on variation
in Sd only, a clear variation in fragility is observed in FFP both with
respect to Sd and Sd with the change in Op. For example, for a 3B-4S
OGS frame the variation in fragility for a uniform increase in Op from
0% to 90% for the four damage states can be easily obtained using FFP
as: 97% to 91% for S, 92% to 79% for M, 76% to 57% for E, and 34% to
19% for C.
Fig. 16(b) and (c) shows the conventional fragility curves for the
Extreme damage state in OGS frames with different Op considering PGA
and Sd as the IM, respectively. Say, a region or zone has a design PGA of
0.36 g, then from Fig. 16(b) the corresponding seismic fragility for
frames with 0%, 50%, and 90% Op are obtained approximately as 0.75,
0.65, and 0.55, respectively. This, however, does not give information
on Sd, for which Fig. 16(c) is additionally required to be used con-
sidering the seismic fragility values already obtained. For a given
structure and hazard level, complete information on seismic fragility
and Sd for different Op, NB, and NS can be easily obtained for any
Fig. 16. Fragility flow plot and conventional fragility curves for 3B-4S OGS frames with varying Op for PGA of 0.36 g and βdsi of 0.7: (a) fragility flow plot, (b) fragility curves with PGA as
IM, and (c) fragility curves with Sd as IM.
355
Engineering Structures 155 (2018) 345–357
Fig. 15. Fragility Flow Plot for OGS frame with
0% Op for a range of 2B-2S to 6B-6S frames for
PGA of 0.36 g and βdsi = 0.7.
displacement threshold using FFP.
5. Assumptions and limitations of the study
Results of the study are applicable subject to a few assumptions and
limitations. The frames are designed and detailed using the relevant
Indian codes [16,53] whose requirements are quite similar to some
other commonly used codes of practice [54,55] as clear from the de-
signed RC sections (Fig. 2). Columns are assumed to be fixed at their
bases without accounting for soil structure interaction. Only weight of
the infill walls in all storeys is considered in the bare frames, while for
OGS and FI frames, weight of masonry infills as well as their strength
and stiffness are considered. This is done to reflect a commonly adopted
analysis and design procedure in some countries including India. Plastic
hinges are assumed to form near beam-column joints in RC members
and at the center of the equivalent single diagonal strut used for
modeling masonry infill. RC members are expected to exhibit ductile
response, therefore, it is assumed that shear failure will not take place.
The diagonal struts are assumed to fail in axial compression only. For
nonlinear static analysis, all the frames are assumed to predominantly
T. Choudhury, H.B. Kaushik
deform according to their fundamental mode of vibration, thus the
results are not applicable to buildings with dominating torsional mode
shapes. The influence of cracking in concrete are assumed to be taken
care of by the Mander’s confined concrete model. The frames are con-
sidered to be of structural behavioral type “C” as per ATC 40; this type
represents poor hysteretic behavior with a substantial reduction of loop
area (severely pinched).
The study is applicable to low- and mid-rise RC frames with max-
imum six bays and six storeys with and without masonry infill walls.
Consideration of any other type of structural configuration and irre-
gularity will be a subject matter of further research as it will require
additional details and analyses. The performance assessment of the
frames is limited to the seismic hazard specified in the Indian seismic
code [16]. To consider various epistemic and aleatoric uncertainties in
fragility analyses, average uncertainty factors that account for all the
uncertainties in material properties, modeling, seismic hazard, damage
state thresholds, and capacity of the structure are considered for all the
damage states. Shear-flexure interaction is not considered in the study.
Limitations on use of nonlinear static procedures are explained in the
section on seismic performance assessment. Failure of the frames is
considered when at least half of the ground storey columns exceed
hinge rotation corresponding to life safety limit state along with failure
of ground storey infill walls. Simplistic material behavior defined by
FEMA 356 is considered for defining the plastic hinge behavior of
ductile RC members. Masonry material model available in literature is
adopted for defining plastic hinge behavior of diagonal struts.
6. Summary and conclusion
Ground storey in large number of multistory buildings is left open
(i.e., without infill walls), even in seismically active regions, in order to
accommodate shopping or parking facilities or for other functional
purposes. Such buildings have performed poorly during several past
earthquakes. In such a scenario, it is extremely important that seismic
vulnerability of such buildings be quantified using simple scientific
methods that can remove the conceived misperception about safety of
such buildings when opening are present in infill walls. This issue is
addressed in the present study by carrying out structural assessment
and fragility analyses of OGS frames with varying number of bays,
number of storeys, size of the central opening in the infill walls, and
seismic hazard. A simplified method of quantifying the seismic fragility
of open ground storey buildings is suggested. Following major conclu-
sions can be drawn from the study:
• Lateral load behavior of OGS buildings remains unaffected by the
amount of openings in infill walls primarily because infill walls are
not present in the ground storey.
•
Seismic fragility analysis for four damage states (slight, moderate,
extreme, and complete) proved that presence of openings or in-
creasing the opening size in infill walls of OGS frame does not re-
duce its seismic fragility.
• A lower seismic fragility may be achieved for higher damage states
in case of fully infill frames in comparison with bare frames if
opening area in infill walls exceeds about 60% of the wall area. This
is because performance, and hence fragility, depends not only on the
initial stiffness but also on several other parameters, such as, post
yield stiffness, the ultimate or collapse point, mode shape, and
natural period of vibration.
• Ductility demand imposed on columns of OGS frames for any level
of seismic hazard is always higher than that imposed on columns of
FI frames as well as bare frames with any opening size in the infill
walls. Since columns in ground storey of OGS buildings lack ade-
quate ductility, stiffness, and strength required to resist high storey
shear, seismic fragility of OGS frames is always higher than the FI
and bare frames.
• In order to get a clear insight on seismic performance and fragility of
356
Engineering Structures 155 (2018) 345–357
OGS frames, bivariate plot representing fragility, termed as Fragility
Flow Plots (FFP), are introduced. In addition to providing a direct
relation between seismic capacity and seismic displacement demand
for a given uncertainty and probability of exceeding a building
function limit state, FFPs combine results of parameterized perfor-
mance assessment for a given earthquake intensity. Such plots can
directly relate the parameters (natural period of vibration, number
of storeys, number of bays, displacement demand, displacement
capacity of the building, probability of exceeding a building func-
tion limit state) required for building design based on fragility for a
given uncertainty and earthquake intensity.
The study contributes towards classifying the seismic fragility of
low- to mid-rise (natural period 0.36–0.81 s) OGS frames on the basis of
several structural parameters. FFP is also intended to be used as a de-
sign aid for OGS frames with different bay and storey configurations
facilitating development of fragility based seismic design. The results of
the study can be used in development of design guidelines for OGS
buildings depending upon the fragility obtained for different combi-
nation of structural parameters.
Acknowledgments
The authors acknowledge the financial assistance provided by the
Ministry of Human Resource Development (MHRD), Government of
India.
References
[1] Fardis MN, Negro P, Bousias SN, Colombo A. Seismic design of open-storey infilled
RC buildings. J Earthq Eng 1999;3(2):173–97.
[2] Jain SK, Lettis B, Murty CVR, Bardet JP. Bhuj, India Earthquake of January 26,
2001. Earthq Spectra 2002;18(Supplement A):398. July.
[3] Murty CVR, Rai DC, Jain SK, Kaushik HB, Mondal G, Dash SR. Performance of
structures in the Andaman and Nicobar Islands (India) during the December 2004
Great Sumatra earthquake and Indian Ocean tsunami. Earthquake Spectra, 22(S3),
EERI, USA, Special Issue June 2006; 2006. p. S321–54.
[4] Kaushik HB, Jain SK. Impact of great December 26, 2004 Sumatra earthquake and
tsunami on structures in Port Blair. J Perform Constr Facil, ASCE
2007;21(2):128–42.
[5] Kaushik HB, Rai DC, Jain SK. Effectiveness of some strengthening options for ma-
sonry-infilled RC frames with open first storey. J Struct Eng, ASCE
2009;135(8):925–37.
[6] Dutta SC, Mukhopadhyay PS, Saha R, Nayak S. 2011 Sikkim earthquake at Eastern
Himalayas: lessons learnt from performance of structures. Soil Dyn Earthq Eng
2015;75:121–9.
[7] Yuen YP, Kuang JS. Nonlinear seismic responses and lateral force transfer me-
chanisms of RC frames with different infill configurations. Eng Struct
2015;91:125–40.
[8] Moghaddam HA, Dowling PJ. ESEE Research Report No. 87–2, The state of the art
in infilled frames. London: Imperial College of Science and Technology. Civil Eng.
Dept; 1987.
[9] Dolšek M, Fajfar P. Soft storey effects in uniformly infilled reinforced concrete
frames. J Earthq Eng 2001;5(1):1–12.
[10] Chintanapakdee C, Chopra AK. Seismic response of vertically irregular frames: re-
sponse history and modal pushover analyses. J Struct Eng, ASCE
2004;130(8):1177–85.
[11] Özhendekci N, Özhendekci D. Rapid seismic vulnerability assessment of low-to mid-
rise reinforced concrete buildings using Bingöl's regional data. Earthq Spectra, EERI
2012;28(3):1165–87.
[12] Al-Nimry H, Resheidat M, Qeran S. Rapid assessment for seismic vulnerability of
low and medium rise infilled RC frame buildings. Earthq Eng Eng Vibr
2015;14(2):275–93.
[13] Davis PR, Padhy TK, Menon D, Prasad AM. Seismic fragility of open ground story
buildings in India. Proc, 9th U.S. National and 10th Canadian Conference on
Earthquake Eng., Toronto, Ontario, Canada. 2010.
[14] Mondal G, Tesfamariam S. Effects of vertical irregularity and thickness of un-
reinforced masonry infill on the robustness of RC framed buildings. Earthq Eng
Struct Dyn 2014;43(2):205–23.
[15] NICEE Workshop. Seismic safety agenda in north-eastern region of India,
Proceedings of the workshop organized by National Information Centre of
Earthquake Engineering (NICEE) and Indian Institute of Technology (IIT) Guwahati
on 22 March 2014 at IIT Guwahati; 2014.
[16] BIS. IS 1893 (Part 1): 2002, Criteria for Earthquake Resistant Design of Structures:
Part 1 General Provisions and Buildings. In: 5th Rev., Bureau of Indian Standards,
New Delhi, India; 2002.
[17] Bhat SA, Setia S, Sehgal VK. Seismic response of moment resisting frame with open
T. Choudhury, H.B. Kaushik
ground storey designed as per code provisions. Advances in Structural Engineering.
India: Springer; 2015. p. 869–83.
[18] Pragalath DH, Avadhoot B, Robin DP, Pradip S. Multiplication factor for open
ground storey buildings–a reliability based evaluation. Earthq Eng Eng Vibr
2016;15(2):283–95.
[19] Cosenza E, Manfredi G, Polese M, Verderame GM. A multilevel approach to the
capacity assessment of existing RC buildings. J Earthq Eng 2005;9(01):1–22.
[20] HAZUS. Multi-hazard Loss Estimation Methodology, Earthquake Model, HAZUS-
MH MR4 Technical Manual. National Institute of Building Sciences and Federal
Emergency Management Agency (NIBS and FEMA), Washington, DC; 2013.
[21] Vamvatsikos D, Cornell CA. Incremental dynamic analysis. Earthq Eng Struct Dyn
2002;31(3):491–514.
[22] Pinho R, Marques M, Monteiro R, Casarotti C, Delgado R. Evaluation of nonlinear
static procedures in the assessment of building frames. Earthq Spectra, EERI
2013;29(4):1459–76.
[23] ATC 40. Seismic Evaluation and Retrofit of Concrete Buildings. Applied Technology
Council, Volumes 1 and 2, Redwood City, CA, USA; 1996.
[24] Fajfar P. A nonlinear analysis method for performance based seismic design. Earthq
Spectra, EERI 2000;16(3):573–92.
[25] Bosco M, Ghersi A, Marino EM. On the evaluation of seismic response of structures
by nonlinear static methods. Earthq Eng Struct Dyn 2009;38(13):1465–82.
[26] Calvi GM, Pinho R, Magenes G, Bommer JJ, Restrepo-Vélez LF, Crowley H.
Development of seismic vulnerability assessment methodologies over the past 30
years. ISET J Earthq Tech 2006;43(3):75–104.
[27] Kaushik HB, Choudhury T. Vulnerability analysis of buildings for seismic risk as-
sessment: a review. The Bridge and Struct. Engineer, The J. Indian National Group
(ING) of Int. Association for Bridge and Struct Eng (IABSE), vol. 45(1); 2015. p.
63–76.
[28] Fragiadakis M, Vamvatsikos D, Aschheim M. Application of nonlinear static pro-
cedures for the seismic assessment of regular RC moment frame buildings. Earthq
Spectra, EERI 2014;30(2):767–94.
[29] Dolšek M, Fajfar P. Simplified non-linear seismic analysis of infilled reinforced
concrete frames. Earthq Eng Struct Dyn 2005;34(1):49–66.
[30] Goel RK. Evaluation of modal and FEMA pushover procedures using strong-motion
records of buildings. Earthq Spectra, EERI 2005;21(3):653–84.
[31] ASCE. Seismic evaluation and retrofit of existing buildings. ASCE/SEI 41–13,
Reston, VA; 2013.
[32] Shinozuka M, Feng MQ, Kim HK, Kim SH. Nonlinear static procedure for fragility
curve development. J Engg Mech, ASCE 2000;126(12):1287–95.
[33] Kappos AJ, Panagopoulos G, Panagiotopoulos C, Penelis G. A hybrid method for the
vulnerability assessment of R/C and URM buildings. Bull Earthq Eng
2006;4(4):391–413.
[34] Ellingwood BR, Celik OC, Kinali K. Fragility assessment of building structural sys-
tems in Mid-America. Earthq Eng Struct Dyn 2007;36(13):1935–52.
[35] CSI. Structural Analysis Program (SAP2000) – advanced, static and dynamic finite
element analysis of structures. Berkeley, USA: Computers and Structures Inc.; 2015.
[36] Mander JB, Priestley MJ, Park R. Theoretical stress-strain model for confined
357
Engineering Structures 155 (2018) 345–357
concrete. J Struct Eng, ASCE 1988;114(8):1804–26.
[37] Kaushik HB, Rai DC, Jain SK. Stress-strain characteristics of clay brick masonry
under uniaxial compression. J Mater Civil Eng, ASCE 2007;19(9):728–39.
[38] Paulay T, Priestley MJN. Seismic design of concrete and masonry structures. New
York: John Wiley and Sons; 1992.
[39] Surendran S. Lateral load behaviour of masonry infilled RC frames with central
opening M.Tech. Thesis India: Indian Institute of Technology Guwahati; 2012.
[40] FEMA 356. Prestandard and Commentary for the Seismic Rehabilitation of
Buildings. Building Seismic Safety Council (US), and Applied Technology Council,
Redwood City, California; 2000.
[41] Basha SH, Kaushik HB. Behavior and Failure Mechanisms of Masonry-Infilled RC
Frames (in Low-Rise Buildings) subject to Lateral Loading. Eng Struct
2016;111(15):233–45. Elsevier, https://doi.org/10.1016/j.engstruct.2015.12.034.
[42] FEMA 440. Improvement of nonlinear static seismic analysis procedures. Applied
Technology Council (ATC). Rep. No. FEMA-440, Washington, D.C.; 2005.
[43] Fajfar P, Gašperšič P. The N2 method for the seismic damage analysis of RC
buildings. Earthq Eng Struct Dyn 1996;25(1):31–46.
[44] Chopra AK, Goel RK. Capacity-demand-diagram methods based on inelastic design
spectrum. Earthq Spectra, EERI 1999;15(4):637–56.
[45] Lagomarsino S, Giovinazzi S. Macroseismic and mechanical models for the vul-
nerability and damage assessment of current buildings. Bull Earthq Eng
2006;4(4):415–43.
[46] Barbat AH, Pujades LG, Lantada N. Seismic damage evaluation in urban areas using
the capacity spectrum method: application to Barcelona. Soil Dyn Earthq Eng
2008;28(10):851–65.
[47] Silva V, Crowley H, Varum H, Pinho R, Sousa R. Evaluation of analytical meth-
odologies used to derive vulnerability functions. Earthq Eng Struct Dyn
2014;43(2):181–204.
[48] FEMA P-58. Next-Generation Methodology for Seismic Performance Assessment of
Buildings, prepared by the Applied Technology Council (ATC) for the Federal
Emergency Management Agency, Report No. FEMA P-58, Washington, D.C.; 2012.
[49] Gueguen P. Seismic vulnerability of structures. New Jersey: ISTE Ltd. and John
Wiley & Sons Inc; 2013.
[50] Mackie KR, Stojadinović B. Post-earthquake functionality of highway overpass
bridges. Earthq Eng Struct Dyn 2006;35(1):77–93.
[51] Ghosh J, Rokneddin K, Padgett JE, Dueñas-Osorio L. Seismic reliability assessment
of aging highway bridge networks with field instrumentation data and correlated
failures, I: methodology. Earthq Spectra, EERI 2014;30(2):795–817.
[52] Lallemant D, Kiremidjian A, Burton A. Statistical procedures for developing earth-
quake damage fragility curves. Earthq Eng Struct Dyn 2015;44(9):1373–89.
[53] BIS. IS 13920 ductile detailing of reinforced concrete structures subjected to seismic
forces-code of practice. New Delhi, India: Bureau of Indian Standards; 1993.
[54] ACI. ACI 318: Building code requirements for structural concrete, ACI Committee
318. Farmington Hills, MI: American Concrete Institute; 2007.
[55] CEN. EN 1998–1, Eurocode 8: Design of structures for Earthquake Resistance-Part 1:
General rules, seismic actions and rules for buildings. Brussels: European
Committee of Standardization; 2004.