Professional Documents
Culture Documents
Abstract: Lateral load behavior of reinforced concrete (RC) frames infilled with fly-ash brick masonry having various sizes of central
openings was investigated experimentally. Results showed that the infilled frames with openings exhibited significantly different lateral
load behavior than the corresponding infilled frames without openings. The drift limit states corresponding to commencement of major
events during the lateral loading (initiation of major cracks in frame members, peak load, and 80% of postpeak load) were found to improve
significantly due to the presence of openings when designed in accordance with current earthquake standards. The presence of openings in
masonry infills modifies two of the most important design parameters, lateral stiffness and strength, for which nonlinear prediction models
were developed by carrying out statistical analyses of the experimental results. Empirical models were used to estimate the stiffness and
strength of frames tested in past studies and were found to be more accurate and consistent compared with several existing predictive relations.
A normalized lateral load–drift relationship for different performance levels of such frames was also propounded based on the observed
damage events. The empirical equations can be used very conveniently to simulate the variations of stiffness and strength (global response) of
the infilled frames with openings under lateral loads. DOI: 10.1061/(ASCE)ST.1943-541X.0002562. © 2020 American Society of Civil
Engineers.
Introduction Celarec and Dolšek 2013; Cavaleri et al. 2017; Basha and Kaushik
2019a) were carried out to understand the lateral load behavior
Masonry infill walls are commonly used as partition elements in of infilled frames without openings for various types of infills, geo-
reinforced-concrete frame buildings. The presence of infills signifi- metric ratios, and support conditions. The in-plane behavior of fully
cantly alters the lateral load behavior of frames in terms of strength, infilled frames is well documented in state-of-the-art reports by
stiffness, energy dissipation, and failure mechanisms. Infill walls Moghaddam and Dowling (1987), CEB (1996), Asteris et al.
are generally considered nonstructural elements primarily due to (2011a), Asteris et al. (2013), and Di Trapani et al. (2015).
their brittle nature, their highly variable material properties, and Several experimental (Stafford-Smith 1962; Mainstone 1971;
their unpredictable placement in buildings. This issue becomes Mallick and Garg 1971; Mosalam et al. 1997; Moghaddam and
even more critical when openings are provided in infills. Signifi- Dowling 1987; Dawe and Seah 1989; Choubey and Sinha 1994;
cant uncertainty lies in the evaluation of lateral load behavior of Raj 2000; Yáñez et al. 2004; Kakaletsis and Karayannis 2008,
such frames due to variability in the material properties of masonry 2009; Sigmund and Penava 2014; Mansouri et al. 2014; Zhai et al.
and the size and location of the openings (Decanini et al. 2014; 2016; Su et al. 2017) and analytical investigations (Sachanski 1960;
Choudhury and Kaushik 2018a). A sensitivity analysis of seismic Liauw 1972; Achyutha et al. 1986; Dawe and Seah 1989; Durrani
response of infilled RC frames showed that the uncertainty in infill
and Luo 1994; Al-Chaar et al. 2003; Asteris 2003; Papia et al.
properties plays the most significant role in the variable lateral load
2003; Mondal and Jain 2008; Tasnimi and Mohebkhah 2011;
behavior of the frames (Choudhury and Kaushik 2018b). A large
Mohammadi and Nikfar 2013; Decanini et al. 2014; Asteris et al.
number of experimental (Stafford-Smith 1962; Mainstone 1971;
2016; Su et al. 2017) have been carried out to evaluate the influence
Mehrabi et al. 1996; Al-Chaar et al. 2002; Blackard et al. 2009;
of infill openings under seismic loading. An exhaustive review of
Basha and Kaushik 2019b) and numerical investigations (Polyakov
the past literature highlighting the importance of openings has been
1960; Mainstone 1971; Crisafulli et al. 2000; Chrysostomou et al.
2002; Ghosh and Amde 2002; El-Dakhakhni et al. 2003; D’Ayala consolidated by Moghaddam and Dowling (1987) and Surendran
et al. 2009; Chrysostomou and Asteris 2012; Asteris et al. 2016; and Kaushik (2012). It has been observed that most seismic stan-
dards often neglect the influence of openings in infills on the lateral
1
Postdoctoral Fellow, College of Civil Engineering, Huaqiao Univ., load behavior whereas some provide a few recommendations,
Xiamen 361021, China. Email: humayunbashasyed@gmail.com which are discussed here.
2
Formerly, Graduate Student, Dept. of Civil Engineering, Indian Insti- Eurocode 8 (CEN 2004) recommends disregarding the infill
tute of Technology Guwahati, Guwahati 781039, India. ORCID: https:// panels with more than one significant perforation or opening. To
orcid.org/0000-0003-4375-9001. Email: sachin.rit@gmail.com avoid premature damage or to prevent brittle failure of infill walls
3
Professor, Dept. of Civil Engineering, Indian Institute of Technology with openings, Eurocode 8 recommends safeguarding the edges of
Guwahati, Guwahati 781039, India (corresponding author). ORCID: openings with belts and posts. MSJC (2013) provides a basic set of
https://orcid.org/0000-0001-5896-6543. Email: hemantbk@iitg.ac.in
provisions for the design of both participating (designed to resist
Note. This manuscript was submitted on February 9, 2019; approved on
September 3, 2019; published online on January 22, 2020. Discussion per- in-plane loads) and nonparticipating (structurally isolated from the
iod open until June 22, 2020; separate discussions must be submitted for lateral load resisting system) infills. The standard recommends not
individual papers. This paper is part of the Journal of Structural Engineer- considering the beneficial influence of infills in the case of partial-
ing, © ASCE, ISSN 0733-9445. height infills and infills with openings due to their detrimental
sitivity of the model to material constitutive laws, the definition of Al-Chaar et al. (2003) carried out experimental and analytical in-
interface connections followed by their calibration, and primarily, vestigations and suggested an analogous reduction factor [Eq. (6)]
high computational demand are the major difficulties encountered for both Rk and Rs using an area opening ratio (Ar ) defined as the
during micromodeling. In the case of macromodeling, infill is mod- ratio of the area of opening Ao to the area of solid infill wall Ai . On
eled as equivalent diagonal strut, and generally two approaches are a similar note, NZSEE (2006) recommended a simplified empirical
followed to account for the influence of openings. In the first relation to estimate Rk and Rs for both door and window openings,
approach, multiple struts are used to model the infill around the using the length of the opening (lo ) and the length of the infill (li ),
openings (Thiruvengadam 1985; Hamburger and Chakradeo 1993; as shown in Eq. (7). The proposed empirical equation was based on
Decanini et al. 2014). This modeling analogy is based on the de- the experimental investigations of Dawe and Seah (1989) on steel
velopment of two sets of cracks (along the frame–infill boundary infilled frames
and near the openings) at smaller lateral drifts (Hamburger and Rk ; Rs ¼ 0.6A2r − 1.6Ar þ 1 for Ar < 0.6 ð6Þ
Chakradeo 1993). However, the analytical and experimental inves-
tigations for multiple strut placements and their locations are not lo
sufficient to establish the required guidelines (ASCE 2013). The Rk ; Rs ¼ 1 − 1.5 for lo < 0.6li ð7Þ
li
second approach is to indirectly reduce the strength and stiffness
of the infill walls in proportion to the opening size by reducing Based on a finite element study on infilled frames with central
the width of the single diagonal strut using approximate reduction window openings, Mondal and Jain (2008) proposed a linear
factors (Polyakov 1956; Sachanski 1960; Imai and Miyamoto stiffness-reduction factor (Rk ) for opening ratio Ar less than 40%
1989; Durrani and Luo 1994; Al-Chaar et al. 2003; Asteris 2003; as given in Eq. (8). Similarly, Asteris (2003) carried out a finite
Papia et al. 2003; NZSEE 2006; Mondal and Jain 2008; Asteris element study and developed plots to estimate the variation of lat-
et al. 2011b; Tasnimi and Mohebkhah 2011; Mohammadi and eral stiffness for various locations and sizes of the openings. At a
Nikfar 2013; Decanini et al. 2014; Asteris et al. 2016; Su et al. later stage, Asteris et al. (2011b) conducted a detailed parametric
2017). Some of these relevant studies are discussed here. study using the same results and proposed a stiffness-reduction fac-
Sachanski (1960) proposed a unique stiffness (Rk ) and tor (Rk ) as shown in Eq. (9)
strength (Rs ) reduction factor as given in Eq. (1) for practical
use based on the length ratio (length of the opening lo divided Rk ¼ 1 − 2.6Ar for Ar < 0.4 ð8Þ
by length of the infill li ) and height ratio (height of the opening
ho divided by height of the infill hi ) of the infill panel. Similarly, Rk ¼ 1 − 2Ar0.54 þ A1.14
r for Ar < 0.5 ð9Þ
Imai and Miyamoto (1989) evaluated the shear strength of ma-
sonry panels with openings using the reduction factor given in Tasnimi and Mohebkhah (2011) conducted an exhaustive exper-
Eq. (2) imental investigation on infill walls with various sizes of door and
window openings. Regression analysis was carried out using the
2l 3h
R k ; Rs ¼ 1 − o − o ð1Þ experimental results, and a strength-reduction factor (Rs ) was
5 li 5 hi suggested for opening ratios less than 40% as given by Eq. (10).
0 sffiffiffiffiffiffiffiffiffi1 Mohammadi and Nikfar (2013) proposed empirical relations based
l lo h o A on a comprehensive statistical analysis of finite element results for
Rk ; Rs ¼ min@1 − o ; 1 − ð2Þ estimating stiffness and strength-reduction factors for steel and
li li h i
RC-confining frames for different opening sizes (doors and win-
dows) as given in Eqs. (11)–(13)
Durrani and Luo (1994) developed a predictive model based on
the finite element analysis to estimate the strength and stiffness- Rs ¼ 1.49A2r − 2.238Ar þ 1 for Ar < 0.4 ð10Þ
reduction factors for infilled frames with central openings as given
in Eq. (3). The factor Ad is given by Eq. (4). Here the variables d Rk ¼ 1.1859A2r − 1.6781Ar þ 1 for Ar < 0.4 RC and steel frame
and do are the diagonal length of the infill wall and the diagonal ð11Þ
length of the opening, respectively; θ and θo are the tangents of the
height-to-length ratio of infill and opening, respectively, and Ai is Rs ¼ −1.085Ar þ 1 for Ar ≤ 0.4 RC frame ð12Þ
the area of solid infill wall
Rs ¼ −2.122Ar þ 1 for Ar ≤ 0.25 steel frame ð13Þ
Ad
Rk ; R s ¼ 1 − ð3Þ
li × h i
Decanini et al. (2014) studied the effect of door and window
2
openings on the lateral strength and stiffness of infilled frames
ðd sin 2θ − do sinðθ þ θo ÞÞ
Ad ¼ Ai − ð4Þ by considering approximately 150 numerical and analytical studies.
2 sinð2θÞ A unique expression for reduction of strength and stiffness was
frame. Asteris et al. (2016) proposed an analytical fifth order different regions. Therefore, the primary motivation of the present
polynomial expression [Eq. (17)] for the determination of the re- study was to develop an empirical model that could predict the stiff-
duction factor for stiffness, considering the opening percentage and ness and strength of a variety of frames with reasonable accuracy,
vertical-load distribution based on the past works of Asteris (2003) even for higher opening ratios instead of achieving high accuracy
and Cavaleri et al. (2005) for some frames while sacrificing the quality of prediction for other
frames.
Rk ¼ 1 − 2Ar ð16Þ
175
175 6ϕ@90 175 6ϕ@120
A B
15 15
2-10ϕ 2-10ϕ C C
Section AA 110
Section BB
175 175 6ϕ@120
6ϕ@ 90 6ϕ@ 110 110 1500
15
11 2-8ϕ 115 2-8ϕ
D D
Downloaded from ascelibrary.org by Indian Institute of Technology Guwahati on 01/28/20. Copyright ASCE. For personal use only; all rights reserved.
10ϕ@ 4 corners
15 15
Section of Lintel
12ϕ@ 4 corners 12ϕ@ 4 corners
Section CC 675 675 8-16ϕ
Section DD
8ϕ@
Key: All dimensions are in mm
120 400
3-8ϕ = 3 bars of 8 mm diameter
6ϕ@90 = 6 mm diameter bars at 90 mm 40
center to center spacing 3200 350
prisms and the shear strength of 1,200 × 1,200 × 110-mm masonry capacities, cyclic hardening or softening effects, and deterioration
wallettes obtained from diagonal compression tests was found to behavior. Considering the slow varying nature of the amplitude
be about 3.9 and 0.14 MPa, respectively. The compressive strength modulation of the ground motion, it is advisable to load the spec-
(3.9 MPa) and elastic modulus of the fly ash brick masonry prism imens by applying multiple cycles of each drift level to observe the
(2,667 MPa) was found to be significantly lower than the com- strength and stiffness degradation of the specimens in each drift
monly used burnt clay brick masonry (Kaushik et al. 2007). level. In the present study, three cycles of each drift level were ap-
Detailed material characterization of fly ash brick masonry under plied on the specimens as shown in the loading protocol in Fig. 3.
various loading conditions can be obtained from Basha and The considered loading protocol permits explicit control of force
Kaushik (2015). The experimental loading setup and arrangement and deformation histories and allows visual observation and mark-
of instrumentation of the test specimens is shown in Fig. 2. All ing of damage. The drift level was initially incremented in equal
specimens were subjected to a combination of incremental horizon- steps of 0.15% until 1.54% drift level because of higher stiffness
tal loads applied cyclically and constant vertical loads applied by of the specimens in initial stages. Later, the drift increment was
placing reinforced concrete plates over the slab. doubled (0.30%) in the inelastic excursion stages to capture the
In the present study, displacement-controlled cyclic loading ultimate limit state quickly as most of the damage was concentrated
was applied on specimens in a slow, controlled, and predetermined in the original locations, until the capacity of the specimen was
manner by disregarding the dynamic effects. The cyclic loads reduced to at least 75% or the failure of the frame became immi-
were applied using a 250 kN capacity servo-controlled hydraulic nent. The loading protocol must be chosen carefully, as the number
actuator of 125-mm stroke length. Cyclic tests provide basic in- of inelastic excursions increases with a decrease in time, and at the
formation on strength and stiffness characteristics, deformation same time, the increase in rate is very high for short time-period
Actuator
175
Strong Wall
1
1260 2
1500
675 750 3
180
4 70
400
Strong Floor
3200 350
Front View Side View
1.3
in the columns were observed at a drift level of 0.76% near beam–
-1.3 column joints, followed by flexural cracks at 0.92% drift level. At
-3.9
4.61% drift level, out-of-plane collapse of the infill panel was ob-
served. Finally, at 5.8% drift level, the test was terminated due to
-6.5 extensive crushing of core concrete in columns and failure of lon-
0 1000 2000 3000 4000 5000
Time (sec) gitudinal reinforcement [Fig. 5(a)]. In the case of IF-Ar -3, cracks
originated near the loading corners in both push and pull directions
Fig. 3. Loading history for slow-cyclic test. as diagonal stepped cracks passing through a mortar bed and head
Downloaded from ascelibrary.org by Indian Institute of Technology Guwahati on 01/28/20. Copyright ASCE. For personal use only; all rights reserved.
joints of infill panel. The cracks often passed through brick units
using vertical splitting, as was observed in the case of IF-Ar -0.
Most of the cracks in the masonry infill below the opening were
frame systems (Su et al. 2017). The specimens were instrumented vertical in nature, passing through the mortar head joints and brick
with load cell, linearly varying displacement transducers (LVDT), units. At 0.62% drift level, brick units below the lintel beam near
and strain gauges. The lateral deflections along the height of the the left and right edge of the opening in both the masonry piers
frame were monitored using LVDTs. The strains in the longitudinal
disintegrated and stopped contributing to the lateral load resistance.
bars were recorded using strain gauges bonded to the rebar in the
At the same drift level, diagonal tension cracks were observed near
critical regions of columns and beams (Fig. 2).
the top of the columns along with flexural cracks along the length
of the columns. With further increases in drift levels, bed-joint slid-
Experimental Observations ing cracks, vertical cracks in masonry, and diagonal and flexural
Experimental behavior of infilled frames with openings can be seg- cracks in columns were observed (up to 2.46% drift level). In
regated into three categories based on the initiation of cracks and the subsequent drift level (3.07%), the entire masonry below the
their subsequent propagation. The first category consists of the in- lintel beam disintegrated. The lintel beam dislodged at a drift level
filled frame with 3% opening (IF-Ar -3), whose behavior was found of 4.61%, and the infill above the lintel beam collapsed in the sub-
to be quite similar to that of the fully infilled frame (IF-Ar -0). In the sequent drift levels, after which the test was terminated at a 5.84%
second category, infilled frames with 10% opening size (IF-Ar -10) drift level [Fig. 5(b)].
and 20% opening size (IF-Ar -20) are placed where the masonry In cases of the second category of frames (IF-Ar -10 and
piers adjacent to the opening were of intermediate size. The third IF-Ar -20), initially the cracks formed in infill as interface cracks
category belongs to infilled frames with opening sizes of 30% near both the loading corners and spread toward the bottom of the
(IF-Ar -30), 40% (IF-Ar -40), and 50% (IF-Ar -50), in which the lintel, along with a vertical crack below the opening. In subsequent
width of the piers was narrower compared with the first two cat- drift levels, diagonal stepped cracks were observed below the lintel
egories. The undamaged state of the specimens before the com- beam, emanating from the corners of the openings toward the
mencement of the test is shown in Fig. 4. opposite beam–column joint. In the initial drift levels, most of
In the case of IF-Ar -0 (fully infilled frame), the cracks in infill the cracks were concentrated in masonry piers leading to disinte-
walls were formed at the frame–wall interface at 0.15% lateral drift gration of the brick units near the edges of the openings at 0.92%
level, followed by formation of stepped cracking along the diagonal and 1.23% drift levels in IF-Ar -10 and IF-Ar -20, respectively.
Fig. 4. Specimens before commencement of the tests: (a) IF-Ar -0; (b) IF-Ar -3; (c) IF-Ar -10; (d) IF-Ar -20; (e) IF-Ar -30; (f) IF-Ar -40; (g) IF-Ar -50;
and (h) BF.
Fig. 5. Failure mechanisms observed in various specimens: (a) IF-Ar -0; (b) IF-Ar -3; (c) IF-Ar -10; (d) IF-Ar -20; (e) IF-Ar -30; (f) IF-Ar -40; (g) IF-
Ar -50; and (h) BF.
Flexural and diagonal tension cracks in columns were initiated at observed flexural cracks in the columns near the top beam–column
0.46% drift level in the case of IF-Ar -20, whereas similar cracks joints at a drift level of 0.62%. In the subsequent drift levels
were initiated at a higher drift level of 0.92% in the case of (1%–2%), flexural cracks developed along the length of both
IF-Ar -10. With further increases in drift levels, crushing of mortar columns. At 3% drift level, concrete cover spalled near the top
bed joints above and below the opening was observed. At 1.84% column–beam junctions. Finally, at 6.15% drift level, the experi-
drift level, masonry brick units near the openings started falling in mental investigation was aborted when the postpeak capacity of
the out-of-plane direction after complete disintegration from the the BF was reduced to 75% of its maximum capacity [Fig. 5(h)]
infill wall. At 4% drift level, the crushing of concrete was observed by the formation of flexural hinges at both ends of the columns.
in columns of IF-Ar -10, near the beam–column joint. The lintel
beam supporting the spandrel masonry dislodged at drift levels of
5.84% and 3.69% in the case of IF-Ar -10 and IF-Ar -20, respec- Influence of Openings on Lateral Load Response
tively, and the tests were terminated in the subsequent drift levels Experimental observations revealed the differences in crack forma-
(6.15% and 4.30%) when complete collapse of the frames was im- tion profiles and failure modes of different frames. The influence of
minent [Figs. 5(c and d)]. opening size on the lateral load response of infilled frames in terms
In the cases of IF-Ar -30, IF-Ar -40, and IF-Ar -50, the width of of hysteretic response, lateral strength, lateral stiffness, and energy
the masonry pier was much shorter, so the diagonal stepped cracks dissipation is assessed in this section.
were not observed in the masonry piers in the initial drift levels. Fig. 6 shows the hysteretic response of all the frames. It can be
The cracks in the infill initiated as interface cracks from the loading observed that the hysteretic loops are more symmetric in nature in
corners spreading toward the bottom of the lintel, along with a ver- both push and pull directions for specimens with larger openings,
tical crack in the sill masonry. Subsequently, diagonal stepped reflecting that the influence of infill on the overall response was less
cracks were observed in masonry piers in the case of IF-Ar -30, in these specimens. In the initial drift levels, the lateral load resis-
whereas no such cracks developed in the other two frames. Flexural tance of IF-Ar -0 and IF-Ar -3 was higher compared with that of the
cracks developed in the columns at drift levels of 0.46%, 0.62%, other infilled frames, due to the major contribution of infill in the
and 0.46%, respectively, in the cases of the three specimens of initial stages. As soon as the infill panel cracked, gaps developed
IF-Ar -30, IF-Ar -40, and IF-Ar -50. Diagonal tension cracks were along the infill–frame interface, and the loops became irregularly
observed in the columns near the beam–column joints at a drift spaced. Nonuniformity (kinks) in the lateral load response, ob-
level of 0.62% (IF-Ar -30) and 0.92% (IF-Ar -40 and IF-Ar -50). served especially in the cases of IF-Ar -0 and IF-Ar -3, was due to
Disintegration of bricks was observed below the lintel beam near the complete contribution of the infill in the initial drift levels by
the opening at 0.77% drift level in the case of IF-Ar -30, and out-of- forming diagonal cracks followed by readjustment of the interlock-
plane falling of the bricks in both masonry piers was observed at ing brick units in the infill wall panels. In the cases of other spec-
2.15% drift level. The contribution of infill was substantially less in imens (IF-Ar -10, IF-Ar -20, IF-Ar -30, IF-Ar -40, and IF-Ar -50), the
the case of IF-Ar -40 and IF-Ar -50, as the width of the masonry piers contribution of the infill was found to be gradual by observing com-
was smaller and very few diagonal stepped cracks were formed in bined diagonal cracks and bed-joint cracks. Pinching was found to
both piers even at higher drift levels. The tests were terminated at be significantly higher in the cases of IF-Ar -0 and IF-Ar -3 in the
6.46%, 5.84%, and 5.23% drift levels in the cases of IF-Ar -30 postpeak regime when compared with other specimens with open-
[Fig. 5(e)], IF-Ar -40 [Fig. 5(f)], and IF-Ar -50 [Fig. 5(g)], respec- ings, which may be due to the extensive cracking observed in the
tively, when crushing of the core concrete was observed near the infill. The variation in lateral load response parameters (lateral
ends of both columns. The reference specimen of bare frame strength, stiffness, and energy dissipation) of different frames with
-70
-140
-7 -5 -3 -1 1 3 5 7 -7 -5 -3 -1 1 3 5 7
(a) Drift (%) (b) Drift (%)
Downloaded from ascelibrary.org by Indian Institute of Technology Guwahati on 01/28/20. Copyright ASCE. For personal use only; all rights reserved.
140
IF-Ar-10 IF-Ar-20
Lateral Load (kN)
70
-70
-140
-7 -5 -3 -1 1 3 5 7 -7 -5 -3 -1 1 3 5 7
(c) Drift (%) (d) Drift (%)
140
IF-Ar-30 IF-Ar-40
Lateral Load (kN)
70
-70
-140
-7 -5 -3 -1 1 3 5 7 -7 -5 -3 -1 1 3 5 7
(e) Drift (%) (f) Drift (%)
140
IF-Ar-50 BF
Lateral Load (kN)
70
-70
-140
-7 -5 -3 -1 1 3 5 7 -7 -5 -3 -1 1 3 5 7
(g) Drift (%) (h) Drift (%)
Fig. 6. Hysteretic response of the specimens: (a) IF-Ar -0; (b) IF-Ar -3; (c) IF-Ar -10; (d) IF-Ar -20; (e) IF-Ar -30; (f) IF-Ar -40; (g) IF-Ar -50; and (h) BF.
respect to the bare frame is given in Table 2 and shown in Fig. 7. (0.45%). The frames with 30%, 40%, and 50% openings reached
Further, Table 2 also compares the lateral load behavior of all spec- their respective capacities at drift levels ranging from 1.85% to
imens in terms of lateral drifts, recorded at various salient points on 2.46%. The lateral load capacity reduced to 80% of its peak in the
the envelop response of lateral load–drift curves [Fig. 7(a)]. range of 2.46%–4.92% drift level, considering both push and pull
The lateral load carrying capacity of infilled frames with open- directions. IF-Ar -20 observed degradation at the lowest drift level
ings ranged from 2.1 to 1.25 times that of the bare frame specimen, (−2.46% in the push direction), primarily due to disintegration of
whereas the lateral load resistance of the fully infilled frame the masonry piers and their out-of-plane failure.
(IF-Ar -0) was about 2.5 times that of the bare frame. The lateral Kakaletsis and Karayannis (2007, 2009) reported that the lateral
load capacity of infilled frames decreased significantly with in- load resistance of infilled frames with concentric openings
creasing opening sizes from 3% to 20%, whereas the capacity re- (10%, 22%, and 38% opening sizes) made using low-strength ma-
mained almost constant (about 1.25 times higher than that of BF) sonry was in the range of 1.47–1.5 times that of the bare frame.
for higher opening ratios (30%, 40%, and 50%). The lateral load However, in cases of specimens with eccentric openings moving
capacity of infilled frames with opening sizes from 3% to 20% was from the leeward column to the center, the average lateral strength
attained at a lateral drift of 0.62%–1.08%, which was slightly was found to vary from 1.38 to 1.64 times that of the bare frame.
higher than that observed in the case of the fully infilled frame The lateral stiffness of the infilled frames was found to be about
Note: K and K b represent the initial stiffness of infilled and bare frames, respectively; ED and EDb represent the cumulative energy dissipation of infilled and
bare frames, respectively; F is the maximum lateral load; DRi , DRf , DRs , DRmax , DR80 , and DRu represent the drift at initiation of crack in infill (sliding/
diagonal), initiation of flexural cracks in columns, initiation of shear cracks in columns, peak load, 80% of peak load in postpeak region, and ultimate lateral
deformation, respectively.
140 27 IF-Ar-0
IF-Ar-0
IF-Ar-3 IF-Ar-3
IF-Ar-30 IF-Ar-30
IF-Ar-40 9 IF-Ar-40
IF-Ar-50 IF-Ar-50
BF BF
0 0
-9
-70
-18
-140 -27
-7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7
(a) Drift (%) (b) Drift (%)
IF-Ar-3 IF-Ar-3
IF-Ar-10 3500 IF-Ar-10
Energy Dissipation (kNmm)
IF-Ar-20 IF-Ar-20
12000 IF-Ar-30 3000 IF-Ar-30
IF-Ar-40 IF-Ar-40
IF-Ar-50 IF-Ar-50
BF
2500 BF
8000 2000
1500
4000 1000
500
0 0
0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7
(c) Drift (%) (d) Drift (%)
20
IF-Ar-0
Equivalent Visocus Damping (%)
IF-Ar-3
IF-Ar-10
IF-Ar-20
15 IF-Ar-30
IF-Ar-40
IF-Ar-50
BF
10
0
0 1 2 3 4 5 6 7
(e) Drift (%)
Fig. 7. Lateral load behavior of infilled frames with openings: (a) envelop of lateral load response; (b) variation in lateral stiffness; (c) cumulative
energy dissipation; (d) reduced energy dissipation per cycle; and (e) equivalent viscous damping at different drift levels.
of infilled frames with openings at different drift levels. The deg- drift levels, it can be inferred that the presence of openings in infill
radation in lateral stiffness was found to be in the range of 30%– panels significantly influences energy dissipation both in pre- and
41% from the first drift level (0.15%) to the subsequent drift level postpeak regimes up to disintegration of masonry infill walls.
(0.31%). The lateral stiffness decreased to 80% of its maximum From the analysis of the strain-gauge data, it was observed that
at a drift level ranging from 1.23% to 3%. The degradation pattern the major damage events in the columns (formation of flexural and
of lateral stiffness did not follow any specific trend, unlike lateral shear cracks) initiated prior to the yielding of the rebar. In the
strength variation, and the lateral stiffness of infilled frames cases of IF-Ar -0, IF-Ar -3, and IF-Ar -10, the rebar yielded at a drift
with different opening sizes reached 80% of their maximum at level of 1.23%, whereas in the case of IF-Ar -20, IF-Ar -30,
almost the same drift level (1.23%–1.38%), except in the case IF-Ar -40, and IF-Ar -50 specimens, yielding of the rebar was ob-
of IF-Ar -50, in which it was attained at a significantly higher drift served at a slightly higher drift level of 1.84%, 1.84%, 2.15%, and
level of 3%. 2.15%, respectively. In comparison, the major damage in the col-
Cumulative energy dissipation was calculated as the summa- umns initiated at a drift level of 0.46%–0.92%. Additionally, it
tion of area under the hysteretic loops for all drift levels. The cu- was affirmed from the strain responses that the lateral load carry-
mulative energy dissipation obtained for all the specimens is ing capacity of most of the specimens (IF-Ar -0, IF-Ar -3, IF-Ar -10,
normalized with respect to that of the bare frame and is given in IF-Ar -20, and IF-Ar -30) was attained before yielding of the rebar
Table 2. The variation in the energy dissipation capacity, reduced- in the columns (Table 2). In the cases of IF-Ar -40 and IF-Ar -50,
energy dissipation per cycle, and the variation of equivalent specimens with larger openings, both the yielding of rebar and
viscous damping ratio for all the specimens at various drift levels lateral load capacity were attained at nearly the same drift level
are also shown in Figs. 7(c–e). The equivalent viscous damping (1.84%–2.15%), as the influence of the infill was reduced with
(β e ) of specimens was calculated following the recommen- an increase in the opening ratio. In comparison, the bare-frame
dations of ATC 40 (ATC 1996) and FEMA 368 (2001) given in specimen observed yielding of reinforcement at a drift level of
Eqs. (22) and (23) 2.33% prior to the attainment of the lateral load capacity (drift
level of 3.08%).
Pþ − P−
K ef ¼ ð22Þ
Δþ − Δ−
Evaluation of Stiffness and Strength-Reduction Factors
2 Ecy It was clearly observed from this experimental study that the lateral
βe ¼ ð23Þ
π K eq ðΔþ − Δ− Þ2 stiffness and strength of infilled frames decrease with an increase in
opening size. As observed in past studies, the reduction in the lat-
where K ef = effective stiffness; Δþ and Δ− = peak lateral dis- eral stiffness and strength of infilled frames due to the presence of
placements corresponding to peak lateral forces Pþ and P− in pull openings can be taken into account by using appropriate stiffness
and push directions, respectively; and Ecy represents the energy and strength-reduction factors [ratio of stiffness (K o ) and strength
dissipated per displacement cycle. The energy-dissipation capacity (Fo ) of the infilled frame with opening to the corresponding stiff-
of the infilled frames with 3%, 10%, 20%, and 30% opening ratios ness (K i ) and strength (Fi ) of the fully infilled frame]. In simplified
was found to be higher when compared with a bare-frame speci- macromodeling techniques, these reduction factors are used to ap-
men. The energy dissipation was found to be highest in the case of propriately modify the properties of the equivalent diagonal struts
IF-Ar -10 (2.44 times BF), primarily due to the contribution of in- in linear as well as nonlinear analysis (Choudhury and Kaushik
fill even at higher drift levels [Fig. 7(a)], lesser amounts of pinch- 2018a; Basha and Kaushik 2019a). Therefore, in the present study,
ing, and fewer numbers of kinks observed under lateral loading. the stiffness and strength-reduction factors were obtained using un-
The equivalent viscous damping ratio [Fig. 7(e)] was found to constrained regression analysis of the experimentally obtained ef-
vary irregularly in the initial drift levels due to the influence of fective secant stiffness for the first drift level and the peak strength,
the masonry infill, and with an increase in drift levels, the ratio respectively. The obtained stiffness (Rk ) and strength (Rs ) reduction
was more or less linearly amplified as the contribution of the factors are given by Eqs. (24) and (25), respectively, and the fitted
frame members was initiated. As discussed in the previous study prediction models are shown in Fig. 8. The coefficient of determi-
by Basha and Kaushik (2016), energy dissipation can be quanti- nation (R2 ) and the standard error of estimate (σ) between the
fied based on lateral strength, displacement sustained, and the experimentally obtained reduction factors and those obtained using
effect of pinching. Additionally, a nonquantifiable parameter, the proposed model are found to be approximately 0.94 and 0.06,
i.e., the damage observed under lateral loading, may be considered respectively, for the stiffness-reduction factor, and approximately
to evaluate the variation in energy dissipation. The case of IF- 0.97 and 0.04, respectively, in case of the strength-reduction factor.
Ar -40 exhibited a lower energy dissipation capacity (0.79 times) Fig. 8 shows the developed prediction models estimate the lateral
compared with that of the BF, primarily because the infill walls stiffness and strength with high accuracy. Standard deviation and
stopped contributing to the lateral load resistance after a 3.67% coefficient of variation in the prediction of stiffness-reduction fac-
drift level until the termination of the test (5.85% drift level). tors were found to be approximately 0.06 and 0.07, respectively.
Experimental Rk duction factors (stiffness and strength) with the corresponding error
0.20 Proposed Rk 0.20 in prediction. In addition to the prediction model developed in the
Experimental Rs present study, the empirical relations proposed by Sachanski (1960),
Proposed Rs
Imai and Miyamoto (1989), and ASCE (2013) are found to estimate
0.00 0.00 the stiffness-reduction factor with high accuracy, which is demon-
0.00 0.10 0.20 0.30 0.40 0.50 strated in Fig. 9(a) where the mean of the reduction-factor ratio is
Area Opening Ratio (Ar) nearly equal to 1.0. Apart from this, models by Su et al. (2017) for
hollow bricks and by Durrani and Luo (1994) also predicted the
Fig. 8. Stiffness and strength reduction models for infilled frames with Rk values satisfactorily. The models by Al-Chaar et al. (2003),
central openings. Mohammadi and Nikfar (2013), and Su et al. (2017) for solid bricks
slightly overestimated the actual stiffness values. Conversely, Papia
et al. (2003), NZSEE (2006), Mondal and Jain (2008), Asteris et al.
(2011b), Decanini et al. (2014), and Asteris et al. (2016) underesti-
A slightly higher standard deviation (0.18) and coefficient of varia- mated the stiffness values. As far as strength prediction is con-
tion (0.17) were found for the strength-reduction factor, primarily cerned, the model proposed in the present study and that by
due to the brittle nature of masonry Al-Chaar et al. (2003) leave all other models behind in accuracy,
as shown in Fig. 9(b) and Table 4. The models suggested by
Rk ¼ 1 − A0.55
r for Ar ≤ 0.5 ðR2 ¼ 0.94; σ ¼ 0.06Þ ð24Þ Tasnimi and Mohebkhah (2011), Mohammadi and Nikfar (2013),
and Su et al. (2017) for hollow bricks, and Su et al. (2017) for solid
bricks also predicted the strength reduction satisfactorily. Imai and
Rs ¼ 1 − A0.65
r for Ar ≤ 0.5 ðR2 ¼ 0.97; σ ¼ 0.04Þ ð25Þ
Miyamoto (1989), Sachanski (1960), Durrani and Luo (1994),
NZSEE (2006), and Decanini et al. (2014) underestimated strength
To validate the proposed prediction models, a comparative study values. It is clear from Fig. 9 that although there are existing em-
was carried out to estimate the lateral stiffness and strength reported pirical models that can predict the stiffness and strength of some
in the past experimental and analytical studies (Choubey and Sinha cases with very good accuracy, the equations developed in the
1994; Raj 2000; Yáñez et al. 2004; Kakaletsis and Karayannis present study are more consistent in providing reasonably good
2008, 2009; Mansouri et al. 2014; Sigmund and Penava 2014; predictions in almost all considered cases. This is quite evident
Zhai et al. 2016; Decanini et al. 2014; Akhoundi et al. 2016) using from the mean of the reduction-factor ratio, which is nearly equal
the proposed Eqs. (24) and (25). Further, the values predicted by the to 1.0 for both stiffness and strength in the case of the proposed
2.0 2.0
Mean Mean
Actual /Predicted (Rs )
Actual/Predicted (Rk )
1.5 1.5
1.0 1.0
0.5 0.5
0.0 0.0
Fig. 9. Variation in the ratio of actual to predicted reduction values for (a) stiffness; and (b) strength.
Table 3. Comparison of experimentally and analytically obtained stiffness with their predicted values
Actual
© ASCE
values Predicted stiffness-reduction factor (Rk ) using different empirical models
Past experimental Proposeda M&J AL N At M&N Su Su S D&L D Ast P I&M AS
or analytical studies Ar Rk [Eq. (24)] [Eq. (8)] [Eq. (6)] [Eq. (7)] [Eq. (9)] [Eq. (11)] [Eq. (20)] [Eq. (21)] [Eq. (1)] [Eq. (3) [Eqs. (14)] [Eq. (17)] [Eq. (5)] [Eq. (2)] [Eq. (16)]
Present study 0.00 1.00 1.00(0)b 1.00(0) 1.00(0) 1.00(0) 1.00(0) 1.00(0) 1.00(0) 0.99(2) 1.00(0) 1.00(0) 0.99(1) 1.00(0) 1.24(24) 1.00(0) 1.00(0)
Present study 0.03 0.79 0.86(8) 0.93(16) 0.96(20) 0.74(7) 0.72(10) 0.96(20) 0.95(20) 0.95(19) 0.83(4) 0.90(13) 0.79(2) 0.92(15) 0.95(19) 0.83(4) 0.94(19)
Present study 0.10 0.72 0.73(0) 0.75(3) 0.85(18) 0.53(26) 0.5(31) 0.85(18) 0.84(16) 0.86(19) 0.69(5) 0.73(0) 0.60(18) 0.69(5) 0.71(2) 0.69(5) 0.81(12)
Present study 0.20 0.59 0.59(0) 0.49(18) 0.71(20) 0.33(44) 0.33(45) 0.72(21) 0.67(13) 0.74(25) 0.56(6) 0.52(12) 0.42(29) 0.42(30) 0.49(18) 0.56(6) 0.61(2)
Present study 0.30 0.43 0.49(13) 0.23(48) 0.58(34) 0.18(58) 0.22(51) 0.61(41) 0.50(15) 0.61(41) 0.46(5) 0.37(14) 0.31(29) 0.24(46) 0.32(28) 0.46(5) 0.41(6)
Present study 0.40 0.38 0.40(5) NA 0.46(21) 0.05(87) 0.14(65) NA 0.33(15) 0.48(27) 0.37(3) 0.26(33) 0.23(40) 0.12(70) 0.17(57) 0.37(3) 0.20(48)
Present study 0.50 0.26 0.32(22) NA 0.36(35) NA 0.08(70) NA 0.16(40) 0.36(37) 0.30(13) 0.17(36) 0.18(34) 0.06(79) 0.04(85) 0.30(13) 0.01(100)
Kakaletsis and 0.10 0.70 0.72(1) 0.73(4) 0.84(20) 0.63(11) 0.49(31) 0.84(19) 0.83(17) 0.86(21) 0.66(7) 0.70(2) 0.62(12) 0.48(33) 0.54(24) 0.75(7) 0.80(13)
Karayannis (2008,
2009)
Kakaletsis and 0.16 0.71 0.64(10) 0.59(17) 0.77(8) 0.43(39) 0.39(46) 0.77(8) 0.74(4) 0.79(11) 0.60(15) 0.60(16) 0.49(31) 0.48(33) 0.54(25) 0.62(12) 0.69(3)
Karayannis (2008,
2009)
Kakaletsis and 0.21 0.80 0.58(28) 0.46(43) 0.70(14) 0.25(69) 0.32(61) 0.71(13) 0.65(19) 0.73(10) 0.56(31) 0.51(38) 0.40(51) 0.48(41) 0.54(34) 0.50(38) 0.59(27)
Karayannis (2008,
2009)
Kakaletsis and 0.10 0.95 0.72(25) 0.73(23) 0.84(12) 0.63(34) 0.49(49) 0.84(12) 0.83(13) 0.86(10) 0.66(32) 0.70(27) 0.62(35) 0.48(50) 0.54(44) 0.75(21) 0.80(17)
Karayannis (2008,
2009)
Yáñez et al. (2004) 0.38 0.52 0.42(21) 0.01(98) 0.48(9) 0.05(91) 0.15(72) 0.54(2) 0.36(32) 0.51(4) 0.39(26) 0.28(48) 0.24(55) 0.16(71) 0.22(58) 0.37(30) 0.24(54)
Yáñez et al. (2004) 0.38 0.57 0.42(28) 0.01(98) 0.48(16) 0.05(92) 0.15(74) 0.54(6) 0.36(37) 0.51(11) 0.39(32) 0.28(52) 0.24(59) 0.16(73) 0.22(61) 0.37(36) 0.24(58)
Yáñez et al. (2004) 0.15 0.95 0.65(32) 0.61(37) 0.78(19) 0.62(35) 0.40(59) 0.78(19) 0.75(22) 0.80(17) 0.54(44) 0.55(42) 0.56(42) 0.16(84) 0.22(77) 0.75(22) 0.70(27)
04020021-11
Yáñez et al. (2004) 0.35 0.47 0.44(6) 0.09(82) 0.52(9) 0.07(87) 0.17(64) 0.56(19) 0.41(13) 0.54(16) 0.42(12) 0.31(35) 0.26(46) 0.21(56) 0.29(39) 0.38(20) 0.30(37)
Yáñez et al. (2004) 0.14 0.95 0.67(30) 0.65(33) 0.80(17) 0.64(33) 0.42(56) 0.80(17) 0.77(19) 0.81(15) 0.57(41) 0.59(38) 0.58(39) 0.21(78) 0.29(70) 0.76(21) 0.73(24)
Yáñez et al. (2004) 0.35 0.35 0.44(23) 0.09(76) 0.52(44) 0.07(83) 0.17(53) 0.56(57) 0.41(14) 0.54(53) 0.42(16) 0.31(15) 0.26(29) 0.21(42) 0.29(20) 0.38(6) 0.30(17)
Yáñez et al. (2004) 0.14 0.80 0.67(18) 0.65(20) 0.80(2) 0.64(21) 0.42(48) 0.80(2) 0.77(5) 0.81(1) 0.57(30) 0.59(27) 0.58(28) 0.21(74) 0.29(65) 0.76(6) 0.73(10)
J. Struct. Eng.
Downloaded from ascelibrary.org by Indian Institute of Technology Guwahati on 01/28/20. Copyright ASCE. For personal use only; all rights reserved.
Table 3. (Continued.)
Actual
© ASCE
values Predicted stiffness-reduction factor (Rk ) using different empirical models
Past experimental Proposeda M&J AL N At M&N Su Su S D&L D Ast P I&M AS
or analytical studies Ar Rk [Eq. (24)] [Eq. (8)] [Eq. (6)] [Eq. (7)] [Eq. (9)] [Eq. (11)] [Eq. (20)] [Eq. (21)] [Eq. (1)] [Eq. (3) [Eqs. (14)] [Eq. (17)] [Eq. (5)] [Eq. (2)] [Eq. (16)]
Akhoundi et al. (2016) 0.25 0.44 0.54(21) 0.35(21) 0.64(44) 0.25(44) 0.26(41) 0.66(48) 0.58(30) 0.67(51) 0.50(13) 0.44(1) 0.36(20) 0.31(30) 0.39(12) 0.50(13) 0.50(13)
Akhoundi et al. (2016) 0.30 0.38 0.49(26) 0.22(44) 0.58(50) 0.18(54) 0.21(46) 0.61(58) 0.49(28) 0.61(58) 0.45(18) 0.37(5) 0.31(21) 0.23(41) 0.31(20) 0.45(18) 0.40(4)
Akhoundi et al. (2016) 0.36 0.33 0.43(32) 0.07(80) 0.51(54) 0.10(69) 0.17(51) 0.55(68) 0.40(20) 0.54(62) 0.40(22) 0.30(10) 0.26(22) 0.16(53) 0.22(33) 0.40(22) 0.28(14)
Akhoundi et al. (2016) 0.39 0.30 0.41(35) NA 0.47(57) 0.07(79) 0.14(53) 0.53(76) 0.35(14) 0.50(65) 0.38(26) 0.27(12) 0.24(22) 0.13(58) 0.18(40) 0.38(25) 0.22(26)
Akhoundi et al. (2016) 0.42 0.27 0.38(37) NA 0.44(57) 0.03(91) 0.12(57) NA 0.29(4) 0.46(65) 0.35(27) 0.23(16) 0.22(23) 0.10(64) 0.14(51) 0.35(27) 0.16(44)
Akhoundi et al. (2016) 0.45 0.25 0.36(40) NA 0.40(58) NA 0.11(60) NA 0.24(8) 0.42(64) 0.33(30) 0.21(20) 0.20(23) 0.08(69) 0.10(62) 0.33(29) 0.10(64)
Akhoundi et al. (2016) 0.49 0.23 0.33(42) NA 0.37(58) NA 0.09(64) NA 0.18(25) 0.37(61) 0.30(32) 0.18(25) 0.18(23) 0.06(74) 0.05(78) 0.30(32) 0.02(91)
Zhai et al. (2016) 0.20 0.66 0.59(10) 0.49(26) 0.71(8) 0.36(45) 0.33(51) 0.72(9) 0.67(2) 0.74(12) 0.56(16) 0.53(20) 0.43(35) 0.39(42) 0.46(31) 0.58(13) 0.61(8)
Decanini et al. (2014) 0.05 0.73 0.81(11) 0.87(19) 0.93(26) 0.78(6) 0.64(13) 0.92(26) 0.92(25) 0.92(26) 0.74(1) 0.82(12) 0.77(5) 0.65(12) 0.68(8) 0.85(16) 0.90(23)
Decanini et al. (2014) 0.07 0.69 0.78(12) 0.83(20) 0.90(30) 0.70(1) 0.59(16) 0.90(29) 0.89(29) 0.90(30) 0.72(4) 0.79(14) 0.71(2) 0.65(7) 0.68(2) 0.80(16) 0.87(26)
Decanini et al. (2014) 0.08 0.64 0.75(16) 0.79(22) 0.88(36) 0.63(2) 0.54(16) 0.87(36) 0.86(34) 0.88(37) 0.70(9) 0.76(17) 0.65(1) 0.65(0) 0.68(5) 0.75(17) 0.84(30)
Decanini et al. (2014) 0.10 0.58 0.72(24) 0.74(28) 0.85(46) 0.55(5) 0.50(15) 0.85(46) 0.84(43) 0.86(48) 0.68(17) 0.72(23) 0.60(3) 0.65(11) 0.68(16) 0.70(21) 0.80(38)
Decanini et al. (2014) 0.12 0.53 0.70(31) 0.70(31) 0.83(55) 0.48(10) 0.46(13) 0.83(55) 0.81(51) 0.84(58) 0.66(25) 0.68(28) 0.55(4) 0.65(21) 0.68(27) 0.65(23) 0.77(45)
Decanini et al. (2014) 0.13 0.49 0.67(37) 0.66(33) 0.80(63) 0.40(18) 0.43(13) 0.80(63) 0.78(58) 0.82(66) 0.64(31) 0.65(31) 0.51(3) 0.65(31) 0.68(37) 0.60(22) 0.74(50)
Decanini et al. (2014) 0.05 0.79 0.81(2) 0.87(10) 0.93(17) 0.78(2) 0.64(19) 0.92(16) 0.92(16) 0.92(16) 0.74(6) 0.82(4) 0.77(3) 0.65(19) 0.68(15) 0.85(8) 0.90(14)
Decanini et al. (2014) 0.07 0.74 0.78(5) 0.83(12) 0.90(21) 0.70(5) 0.59(21) 0.90(21) 0.89(20) 0.90(21) 0.72(3) 0.79(6) 0.71(5) 0.65(13) 0.68(9) 0.80(8) 0.87(17)
Decanini et al. (2014) 0.08 0.69 0.75(8) 0.79(14) 0.88(26) 0.63(9) 0.54(22) 0.87(26) 0.86(25) 0.88(27) 0.70(1) 0.76(9) 0.65(6) 0.65(7) 0.68(2) 0.75(9) 0.84(21)
Decanini et al. (2014) 0.10 0.64 0.72(12) 0.74(16) 0.85(32) 0.55(14) 0.50(23) 0.85(32) 0.84(30) 0.86(34) 0.68(6) 0.72(12) 0.60(7) 0.65(0) 0.68(5) 0.70(9) 0.80(25)
Decanini et al. (2014) 0.12 0.59 0.70(17) 0.70(18) 0.83(39) 0.48(19) 0.46(22) 0.83(39) 0.81(36) 0.84(42) 0.66(12) 0.68(15) 0.55(7) 0.65(9) 0.68(14) 0.65(10) 0.77(30)
Decanini et al. (2014) 0.13 0.54 0.67(24) 0.66(21) 0.80(48) 0.40(26) 0.43(21) 0.80(48) 0.78(43) 0.82(51) 0.64(19) 0.65(19) 0.51(6) 0.65(19) 0.68(25) 0.60(11) 0.74(36)
Note: NA = Not available due to upper limit of the predicted models.
a
Proposed = proposed model; M&J = Mondal and Jain (2008); AL = Al-Chaar et al. (2003); N = NZSEE (2006); At = Asteris et al. (2011b); M&N = Mohammadi and Nikfar (2013); Su = Su et al. (2017);
04020021-12
S= Sachanski (1960); D&L = Durrani and Luo (1994); D = Decanini et al. (2014); Ast = Asteris et al. (2016); P = Papia et al. (2003); I&M = Imai and Miyamoto (1989); and AS = ASCE (2013).
b
Figures in parentheses indicate the percent error between experimental (or analytical) and predicted values.
Table 4. Comparison of experimentally and analytically obtained strength with their predicted values
Actual values Predicted strength-reduction factor (Rs ) using different empirical relations
© ASCE
a
Past experimental or Proposed AL T N M&N Su Su S D&L D I&M
analytical studies Ar Rs [Eq. (25)] [Eq. (6)] [Eq. (10)] [Eq. (7)] [Eq. (12)] [Eq. (18)] [Eq. (19)] [Eq. (1)] [Eq. (3)] [Eq. (14)] [Eq. (2)]
Present study 0.00 1.00 1.00(0)b 1.00(0) 1.00(0) 1.00(0) 1.00(0) 0.97(3) 0.98(3) 1.00(0) 1.00(0) 0.99(1) 1.00(0)
Present study 0.03 0.82 0.90(9) 0.96(16) 0.94(14) 0.74(10) 0.97(18) 0.94(14) 0.95(15) 0.83(1) 0.90(10) 0.79(5) 0.83(1)
Present study 0.10 0.70 0.78(12) 0.85(22) 0.80(14) 0.53(24) 0.90(28) 0.86(23) 0.88(26) 0.69(2) 0.73(3) 0.60(15) 0.69(2)
Present study 0.20 0.73 0.65(11) 0.71(3) 0.62(16) 0.33(55) 0.79(8) 0.75(2) 0.78(7) 0.56(24) 0.52(29) 0.42(43) 0.56(24)
Present study 0.30 0.53 0.55(3) 0.58(9) 0.47(12) 0.18(66) 0.68(27) 0.63(19) 0.69(29) 0.46(14) 0.37(30) 0.31(42) 0.46(14)
Present study 0.40 0.49 0.45(8) 0.46(7) NA 0.05(90) NA 0.52(5) 0.59(20) 0.37(25) 0.26(49) 0.23(54) 0.37(25)
Present study 0.50 0.52 0.37(30) 0.36(32) NA NA NA 0.41(23) 0.50(5) 0.30(43) 0.17(68) 0.18(67) 0.30(43)
Kakaletsis and Karayannis 0.10 0.82 0.78(6) 0.84(3) 0.79(4) 0.63(24) 0.89(9) 0.86(4) 0.88(7) 0.66(20) 0.70(15) 0.62(24) 0.68(17)
(2008, 2009)
Kakaletsis and Karayannis 0.16 0.81 0.70(14) 0.77(6) 0.69(16) 0.43(47) 0.83(2) 0.79(3) 0.82(1) 0.60(26) 0.60(27) 0.49(40) 0.61(26)
(2008, 2009)
Kakaletsis and Karayannis 0.21 0.80 0.64(20) 0.70(13) 0.60(25) 0.25(69) 0.78(3) 0.74(8) 0.78(3) 0.56(31) 0.51(37) 0.40(51) 0.50(37)
(2008, 2009)
Kakaletsis and Karayannis 0.10 0.93 0.78(18) 0.84(10) 0.79(16) 0.63(33) 0.89(5) 0.86(9) 0.88(7) 0.66(30) 0.70(26) 0.62(34) 0.68(27)
(2008, 2009)
Yáñez et al. (2004) 0.38 0.63 0.47(26) 0.48(24) 0.37(42) 0.05(92) 0.59(7) 0.54(15) 0.61(4) 0.39(39) 0.28(57) 0.24(63) 0.37(42)
Yáñez et al. (2004) 0.38 0.63 0.47(27) 0.48(25) 0.37(43) 0.05(93) 0.59(8) 0.54(16) 0.61(5) 0.39(39) 0.28(57) 0.24(63) 0.37(42)
Yáñez et al. (2004) 0.15 0.87 0.71(18) 0.78(11) 0.70(20) 0.62(28) 0.84(4) 0.80(8) 0.83(5) 0.54(38) 0.55(37) 0.56(36) 0.61(30)
Yáñez et al. (2004) 0.35 0.53 0.50(7) 0.52(4) 0.40(25) 0.07(89) 0.62(16) 0.57(7) 0.64(19) 0.42(22) 0.31(43) 0.26(53) 0.38(30)
Yáñez et al. (2004) 0.14 0.90 0.73(20) 0.80(12) 0.72(20) 0.64(30) 0.86(6) 0.82(10) 0.84(7) 0.57(37) 0.59(35) 0.58(36) 0.63(30)
Yáñez et al. (2004) 0.35 0.51 0.50(4) 0.52(0) 0.40(23) 0.07(88) 0.62(20) 0.57(11) 0.64(23) 0.42(20) 0.31(41) 0.26(51) 0.38(27)
Yáñez et al. (2004) 0.14 0.77 0.73(6) 0.80(3) 0.72(6) 0.64(18) 0.86(11) 0.82(6) 0.84(9) 0.57(26) 0.59(24) 0.58(25) 0.63(18)
Raj (2000) 0.11 0.93 0.76(19) 0.83(11) 0.77(18) 0.63(33) 0.88(6) 0.85(10) 0.87(7) 0.63(32) 0.67(28) 0.61(35) 0.67(29)
04020021-13
Raj (2000) 0.11 0.88 0.76(14) 0.83(6) 0.77(13) 0.63(29) 0.88(0) 0.85(4) 0.87(2) 0.63(28) 0.67(24) 0.61(31) 0.67(24)
Raj (2000) 0.11 0.87 0.76(13) 0.83(5) 0.77(12) 0.63(28) 0.88(1) 0.85(4) 0.87(1) 0.63(28) 0.67(24) 0.61(30) 0.67(24)
Raj (2000) 0.11 0.87 0.76(13) 0.83(5) 0.77(12) 0.63(29) 0.88(0) 0.85(4) 0.87(1) 0.63(28) 0.67(24) 0.61(31) 0.67(24)
Mansouri et al. (2014) 0.16 0.78 0.70(11) 0.76(3) 0.68(14) 0.47(40) 0.83(6) 0.79(1) 0.82(5) 0.59(25) 0.58(26) 0.49(37) 0.60(24)
J. Struct. Eng.
Proposed = proposed Model; AL = Al-Chaar et al. (2003); T = Tasnimi and Mohebkhah (2011); N = NZSEE (2006); M&N = Mohammadi and Nikfar (2013); Su = Su et al. (2017); S = Sachanski (1960); D&L =
[Eq. (2)]
0.65(10)
0.60(11)
model, even for high opening ratios. No other empirical model
0.72(3)
0.69(7)
I&M
gives such consistent results for both stiffness and strength predic-
tion, a fact that is also evident from Tables 3 and 4.
To throw more insight into these observations, the stiffness and
[Eq. (14)] strength values are predicted using all these empirical models and
0.65(6)
0.60(7)
0.55(7)
0.51(6)
directly compared with results of each of the considered experi-
D
0.72(12)
0.68(15)
0.65(19)
0.76(9)
[Eq. (1)]
0.66(12)
0.64(19)
are more accurate than those by the other models. It is clear from
0.70(1)
0.68(6)
S
Fig. 12 that the linear Eq. (8), proposed by Mondal and Jain
0.88(27)
0.86(34)
0.84(42)
0.82(51)
Su
Durrani and Luo (1994); D = Decanini et al. (2014); P = Papia et al. (2003); and I&M = Imai and Miyamoto (1989).
0.91(32)
0.90(39)
0.88(48)
0.86(58)
0.55(14)
0.48(19)
0.40(26)
Figures in parentheses indicate the % error between experimental (or analytical) and predicted values.
0.63(9)
error is reasonably low for both stiffness and strength for all the
N
nis (2008, 2009) on strong and weak masonry, it was found that the
T
[Eq. (25)]
Proposed
0.81(16)
0.78(21)
0.76(28)
0.74(35)
Durrani and Luo (1994), and Decanini et al. (2014) predicted the
reduction in strength with higher errors in the case of the aforemen-
tioned frame (Ar ≈ 20%). Yáñez et al. (2004) conducted an exper-
imental investigation on frames with door and window openings
involving significantly high-strength and stiff concrete and hollow-
0.69
0.64
0.59
0.54
Actual values
Rs
(2014)
(2014)
(2014)
(2014)
in Tables 3 and 4.
analytical studies
RC and steel frames infilled with stiff solid-brick masonry and soft
et
et
et
et
Decanini
Decanini
Decanini
Decanini
0.60
0.40
0.20
Downloaded from ascelibrary.org by Indian Institute of Technology Guwahati on 01/28/20. Copyright ASCE. For personal use only; all rights reserved.
0.80
0.60
0.40
Prediction Models
Fig. 10. Comparison of predicted lateral stiffness with individual experimental or analytical study.
estimate the reduction in stiffness and strength. Similarly, ASCE Al-Chaar et al. (2003), Mohammadi and Nikfar (2013), and Su et al.
(2013), Mohammadi and Nikfar (2013), and Al-Chaar et al. (2003) (2017) significantly overestimated stiffness reduction for area ratios
also overestimated the stiffness and strength-reduction factors >10%. Conversely, Papia et al. (2003), NZSEE (2006), Mondal
(i.e., stiffness and strength were underestimated). The other empir- and Jain (2008), Decanini et al. (2014), and Asteris et al. (2016)
ical models (Durrani and Luo 1994; Imai and Miyamoto 1989; significantly underestimated the stiffness reduction. In addition to
NZSEE 2006; Papia et al. 2003; Asteris et al. 2016) considered the the accurate estimation in the present study, Sachanski (1960), Imai
length ratio (lo =li ) as a varying parameter and predicted the stiff- and Miyamoto (1989), and Durrani and Luo (1994) predicted stiff-
ness reduction reasonably well. The empirical model of Decanini ness reduction reasonably well. It was found that almost all predic-
et al. (2014) [Eq. (14)] was also developed using length ratio as one tion models underestimated strength reduction, whereas the models
of the parameters. The empirical models proposed in the present proposed by Su et al. (2017) and Mohammadi and Nikfar (2013)
study and by Sachanski (1960) and Imai and Miyamoto (1989) es- predicted strength reduction satisfactorily up to an opening ratio of
timated the stiffness-reduction factors accurately in the case of an only 25%.
experimental investigation carried out by Mansouri et al. (2014) on Once again, it is important to highlight from Tables 3 and 4 and
low-strength and soft solid-clay bricks for various sizes of window Figs. 12 and 13 that the error in estimation of both stiffness and the
and door openings. Mohammadi and Nikfar (2013), Al-Chaar et al. strength-reduction factors increases significantly with an increase
(2003), and Su et al. (2017) predicted the strength reduction for in Ar regardless of the empirical model used. The equations devel-
hollow bricks reasonably well for the two opening ratios (16% oped in the present study estimate stiffness and strength reduction
and 27%). with good accuracy not only for the present experimental results
In the case of the analytical study on RC frames with central but also for the past experimental and analytical results. From
window openings by Akhoundi et al. (2016), it was found that the comparative study, it is clear that the past empirical equations
0.60
0.40
0.20
Downloaded from ascelibrary.org by Indian Institute of Technology Guwahati on 01/28/20. Copyright ASCE. For personal use only; all rights reserved.
Present
Experimental Study Decanini et al. (2014) Akhoundi et al. (2016)
0.00
0.0 0.1 0.2 0.3 0.4 0.5 0.0 0.1 0.2 0.3 0.4 0.5 0.0 0.1 0.2 0.3 0.4 0.5
(a) Opening Ratio (Ar) (b) Opening Ratio (Ar) (c) Opening Ratio (Ar)
1.00
Strength Reduction Factor (Rs)
0.80
0.60
0.40
Prediction Models
Fig. 11. Comparison of predicted lateral strength with individual experimental or analytical study.
100 100.0
Error in Stiffness Reduction (%)
80 80.0
60 60.0
40 40.0
20 20.0
0 0.0
0.00 0.10 0.20 0.30 0.40 0.50 0.00 0.10 0.20 0.30 0.40 0.50
(a) Area Opening Ratio (Ar) (b) Opening Area Ratio (Ar)
Fig. 12. Comparison of prediction error in relation to the opening-area ratio for (a) stiffness-reduction factor; and (b) strength-reduction factor (same
markers are used as in Figs. 10 and 11).
Fig. 13. Comparison of prediction error across all the studies for (a) stiffness-reduction factor; and (b) strength-reduction factor.
of column shear failure on pushover-based assessment of masonry in- Memphis, TN: Central US Earthquake Consortium.
filled reinforced concrete frame structures: A case study.” Soil Dyn. Imai, H., and M. Miyamoto. 1989. “Seismic behavior of reinforced ma-
Earthquake Eng. 100 (Sep): 98–112. https://doi.org/10.1016/j.soildyn sonry walls with small opening.” In Vol. 2 of Proc., 5 Jornadas Chi-
.2017.05.032. lenas de Sismología e Ingeniería Antisísmica, 965–973. Santiago,
Cavaleri, L., M. Fossetti, and M. Papia. 2005. “Infilled frames: Develop- Chile: Ministerio de Vivienda y Urbanismo.
ments in the evaluation of cyclic behaviour under lateral loads.” Struct. Kakaletsis, D. J., and C. G. Karayannis. 2007. “Experimental investigation
Eng. Mech. 21 (4): 469–494. https://doi.org/10.12989/sem.2005.21 of infilled R/C frames with eccentric openings.” Struct. Eng. Mech.
.4.469. 26 (3): 231–250. https://doi.org/10.12989/sem.2007.26.3.231.
CEB (Comite Euro-International Du Beton). 1996. RC frames under earth- Kakaletsis, D. J., and C. G. Karayannis. 2008. “Influence of masonry
quake loading: State of the Art Report. London: Thomas Telford. strength and openings on infilled RC frames under cyclic loading.”
Celarec, D., and M. Dolšek. 2013. “Practice-oriented probabilistic seismic J. Earthquake Eng. 12 (2): 197–221. https://doi.org/10.1080
performance assessment of infilled frames with consideration of shear /13632460701299138.
failure of columns.” Earthquake Eng. Struct. Dyn. 42 (9): 1339–1360. Kakaletsis, D. J., and C. G. Karayannis. 2009. “Experimental investigation
https://doi.org/10.1002/eqe.2275. of infilled reinforced concrete frames with openings.” ACI Struct.
CEN (European Committee for Standardization). 2004. Design of struc- J. 106 (2): 132–141.
tures for Earthquake Resistance—Part 1: General rules, seismic ac- Kaushik, H. B., D. C. Rai, and S. K. Jain. 2007. “Stress-strain character-
tions and rules for buildings. BS EN 1998-1. Eurocode 8, Brussels, istics of clay brick masonry under uniaxial compression.” J. Mater.
Belgium: CEN. Civ. Eng. 19 (9): 728–739. https://doi.org/10.1061/(ASCE)0899-1561
Choubey, U. B., and S. N. Sinha. 1994. “Cyclic response of infilled (2007)19:9(728).
frames.” J. Struct. Eng. 21: 203–211. Liauw, T. C. 1972. “An approximate method of analysis for infilled frames
Choudhury, T., and H. B. Kaushik. 2018a. “Seismic fragility of open with or without opening.” Build. Sci. 7 (4): 233–238. https://doi.org/10
ground storey RC frames with wall openings for vulnerability assess- .1016/0007-3628(72)90004-7.
ment.” Eng. Struct. 155 (Jan): 345–357. https://doi.org/10.1016/j Mainstone, R. J. 1971. “On the stiffness and strength of infilled frames.”
.engstruct.2017.11.023. In Proc., Transportation and Institute of Civil Engineers, 57–90.
Choudhury, T., and H. B. Kaushik. 2018b. “Seismic response sensitivity to London: Institute of Civil Engineers.
uncertain variables in RC frames with infill walls.” J. Struct. Eng. Mallick, D. V., and R. P. Garg. 1971. “Effect of openings on the lateral
144 (10): 04018184. https://doi.org/10.1061/(ASCE)ST.1943-541X stiffness of infilled frames.” ICE Proc. 49 (2): 193–209. https://doi
.0002190. .org/10.1680/iicep.1971.6263.
Chrysostomou, C. Z., and P. G. Asteris. 2012. “On the in-plane properties Mansouri, A., M. S. Marefat, and M. Khanmohammadi. 2014. “Experimental
and capacities of infilled frames.” Eng. Struct. 41 (Aug): 385–402. evaluation of seismic performance of low-shear strength masonry infills
https://doi.org/10.1016/j.engstruct.2012.03.057. with openings in reinforced concrete frames with deficient seismic
Chrysostomou, C. Z., P. Gergely, and J. F. Abel. 2002. “A six-strut details.” Struct. Des. Tall Special Build. 23 (15): 1190–1210. https://doi
model for nonlinear dynamic analysis of steel infilled frames.” .org/10.1002/tal.1115.
Int. J. Struct. Stab. Dyn. 2 (3): 335–353. https://doi.org/10.1142 Mehrabi, A. B., P. B. Shing, M. P. Schuller, and J. L. Noland. 1996.
/S0219455402000567. “Experimental evaluation of masonry-infilled RC frames.” J. Struct.
Crisafulli, F. J., A. J. Carr, and R. Park. 2000. “Analytical modelling of Eng. 122 (3): 228–237. https://doi.org/10.1061/(ASCE)0733-9445
infilled frame structures—A general review.” Bull. New Zealand Soc. (1996)122:3(228).
Earthquake Eng. 33 (1): 30–47. Moghaddam, H. A., and P. J. Dowling. 1987. The state-of-the-art in infilled
Dawe, J. L., and C. K. Seah. 1989. “Behaviour of masonry infilled steel frames. London: Imperial College of Science and Technology.
frames.” Can. J. Civ. Eng. 16 (6): 865–876. https://doi.org/10.1139 Mohammadi, M., and F. Nikfar. 2013. “Strength and stiffness of masonry-
/l89-129. infilled frames with central openings based on experimental results.”
D’Ayala, D., J. Worth, and O. Riddle. 2009. “Realistic shear capacity as- J. Struct. Eng. 139 (6): 974–984. https://doi.org/10.1061/(ASCE)ST
sessment of infill frames: Comparison of two numerical procedures.” .1943-541X.0000717.
Eng. Struct. 31 (8): 1745–1761. https://doi.org/10.1016/j.engstruct Mondal, G., and S. K. Jain. 2008. “Lateral stiffness of masonry infilled
.2009.02.044. reinforced concrete (RC) frames with central opening.” Earthquake
Decanini, L. D., L. Liberatore, and F. Mollaioli. 2014. “Strength and stiff- Spectra 24 (3): 701–723. https://doi.org/10.1193/1.2942376.
ness reduction factors for infilled frames with openings.” Earthquake Mosalam, K. M., R. N. White, and P. Gergely. 1997. “Static response of
Eng. Eng. Vib. 13 (3): 437–454. https://doi.org/10.1007/s11803-014 infilled frames using quasi-static experimentation.” J. Struct. Eng.
-0254-9. 123 (11): 1462–4169. https://doi.org/10.1061/(ASCE)0733-9445(1997)
Di Trapani, F., G. Macaluso, L. Cavaleri, and M. Papia. 2015. “Masonry 123:11(1462).
infills and RC frames interaction: Literature overview and state of the MSJC (Masonry Standards Joint Committee). 2013. Building code require-
art of macromodeling approach.” Eur. J. Environ. Civ. Eng. 19 (9): ments and specifications for masonry structures. Farmington Hills, MI:
1059–1095. https://doi.org/10.1080/19648189.2014.996671. MSJC.
Durrani, A. J., and Y. H. Luo. 1994. Seismic retrofit of flat-slab buildings NBC. 1994. Nepal national building code for mandatory rules of thumb
with masonry infills. Technical Rep. No. NCEER-94-0004. Taipei, for reinforced concrete buildings with masonry infill. NBC 201.
Taiwan: National Center for Earthquake Engineering Research. New York: NBC.