International Journal of the Physical Sciences Vol. 5(13), pp. 1981-1998, 18 October, 2010
Available online at http://www.academicjournals.org/IJPS
ISSN 1992 - 1950 ©2010 Academic Journals
Full Length Research Paper
Analytical and experimental studies on infilled RC
Mehmet Baran1* and Tugce Sevil2
1Department of Civil Engineering, Kirikkale University, 71450, Yahsihan, Kirikkale, Turkey.
2Department of Civil Engineering, Maltepe University, 34857, Maltepe, Istanbul, Turkey.
Accepted 27 September, 2010
Although hollow brick infills, widely used as partition walls, are considered as non-structural members,
experimental studies revealed that hollow brick infills have favourable effects on strength and stiffness
of structures. In this work, analytical studies were conducted to investigate the hollow brick infill
behaviour, in which infills were modeled by diagonal compression struts. Results were compared with
experimental ones obtained from tests of one-bay, one or two story reinforced concrete (RC) frames,
tested under both vertical and reversed-cyclic lateral loads simulating earthquake. Test frames have
intentionally been constructed poorly to reflect the most common deficiencies encountered in Turkey
such as strong beam-weak column connections, insufficient confinement, low-grade concrete, poor
workmanship and insufficient lap-splice length. Experimental studies shows that hollow brick infills
increased both strength and stiffness of RC frames. Analytical studies conducted, shows that hollow
brick infills could adequately be modeled by diagonal compression struts.
Key words: Reinforced concrete, strength, stiffness, hollow brick infill, diagonal compression strut and reversed-
cyclic lateral load.
Filling reinforced concrete (RC) frames with clay tile
serving as partitions are very common, especially in
Turkey. In structural design process, such infills are
considered as “nonstructural” members. Structure is
assumed to carry horizontal loads only by the frame
elements. However, it is apparent from geometrical
considerations that infills also resist loads and impede
deformations compatible with infilled frame action.
Analytical and experimental studies shows that infilled
frames have greater strength and stiffness compared to
bare frames. Due to changes in stiffness and mass,
dynamic characteristics of the building also change.
Understanding the behaviour of infilled frames and being
capable of making a satisfactory modeling of infills during
structural design process will help engineers to have
more realistic and economical solutions. Behaviour of
infilled frames under seismic loading is complicated.
*Corresponding author. E-mail: email@example.com. Tel:
++ (90) (318) 357 42 42 – 1254. Fax: ++ (90) (318) 357 24 59
This is the most probable reason for hollow brick infills
not being considered as “structural” members during the
structural design process, resulting with inaccurate
solutions. With this approach, natural period of building,
earthquake load transferred to each beam and column,
short column mechanisms that can occur and the failure
mode of building under an earthquake loading can not be
evaluated precisely. Since the behaviour is nonlinear and
closely related to the interaction conditions between
frame and infill, analytical studies should be revised and
supported by experimental data. Earthquake regulations
of many countries (Israel, Costa Rica, France, Algeria,
European Union, Colombia, Phillipines, etc.) recommend
to take the effect of infill walls into account during the
design process (Kaplan, 2008).
In the experimental part of the present study, one and
two-story RC frames were tested. In two-story frames,
lateral load was applied at a greater height resulting in
more turning effect whereas compressive and shear
stresses are more dominant in one-story frames. By this
way, hollow brick infill behaviour can be analyzed under
tensile stresses as well as compressive and shear stresses.
1982 Int. J. Phys. Sci.
Figure 1. Dimensions and reinforcement of the test frames.
In the experimental part of the study, one-third scale,
one-bay, one and two-story RC frames were used as test
units. Taking into account the fact that the building stock
in Turkey and many countries around consists mainly of
deficient RC framed buildings, test frames have
intentionally been designed and constructed with the
most common deficiencies observed in local practice,
such as strong beam-weak column connections,
insufficient confinement, low-grade concrete, insufficient
lap-splice length and poor workmanship. The frames had
their columns fixed to the rigid foundation beams.
Dimensions and reinforcement of both types of test
frames are illustrated in Figure 1.
Low strength concrete was deliberately used in the test frames to
represent the concrete commonly used in majority of existing
buildings in Turkey. Both frame bays were infilled with scaled hollow
Baran and Sevil 1983
Figure 2. Hollow brick used as infill material and infilling method.
bricks (with void ratio of 0.52) covered with a scaled layer of plaster.
Ordinary cement-lime mortar was used for the plaster, reflecting the
usual practice. Hollow brick and infilling method is shown in Figure
2. Ordinary workmanship was intentionally employed in wall
construction and plaster application. For the same reason, mild
steel plain bars were used as longitudinal steel in both test frames.
Typical properties of reinforcing bars used in this study and average
compressive strength values for frame concrete and plaster
determined on testing day are listed in Table 1.
Loading and supporting system
In Figure 3, general views of the test set-up for two and one-story
test frames are given, respectively. As can be seen in Figure 3,
tests were performed in front of a reaction wall. Frames were
subjected to reversed cyclic lateral loading resembling seismic
effects. The quasi-static test loading consisted of reversed cyclic
lateral loading besides constant vertical load applied on both
columns. The axial load on columns was provided by steel cables
post-tensioned by hydraulic jacks.
Reversed cyclic lateral loading was applied by using a double
acting hydraulic jack. The lateral loading system had pin
connections at both ends to eliminate any accidental eccentricity
mainly in vertical direction and tolerating a small rotation in
horizontal direction normal to testing plane. Lateral load was
applied on a spreader beam at one-third of its span to ensure that
the lateral load at second floor level always remains twice as the
lateral load at first floor level. A very rigid external steel ‘guide
frame’ attached to the universal base, was used to prevent any out-
of-plane deformations. During the tests, increasing load cycles were
applied up to the capacity of frame and beyond that, deformation
controlled loading was performed with increasing displacement
cycles. Load histories of all test frames are given in Figure 4.
Deformation measurement system
All deformations were measured by displacement transducers;
using either linear variable differential transformers (LVDTs) or
electronically recordable dial gages (DGs) as shown in Figure 3.
Sway displacements were measured both at first and second floor
levels. Infill wall shear deformations were determined on the basis
of displacement measurements along the diagonals. Displacement
measurements taken at the bottom of both columns were meant for
computation of rotations of the entire frame. They also provided
1984 Int. J. Phys. Sci.
Table 1. Material properties of test frames.
No. of floors
(1) Compressive Strength of RC Infil.
Figure 3. General view of test set-ups and deformation measurement systems.
Baran and Sevil 1985
Lateral Load (kN)
40.246.5 50.0 47.241.9
Lateral Load (kN)
Lateral Load (kN)
Lateral Load (kN)
Lateral Load (kN)
Lateral Load (kN)
Figure 4. Load histories of all test frames.
1986 Int. J. Phys. Sci.
Table 2. Performance indicators.
1st story drift ratio
at peak δ δ δ δ1/h
2nd story drift ratio
at Peak (δ δ δ δ2-δ δ δ δ1)/h
(1) The ratio of Maximum lateral load to that of the reference frame.
deformations; steel yielding in the tension side column,
concrete crushing in the compression side column etc.
Behaviour of test specimens
One being bare and one hollow brick infilled (Sevil et al.,
2010), one being RC infilled (Baran et al., 2009) and six
plastered hollow brick infilled frames (Sevil., 2010; Baran
and Tankut, 2009; Okuyucu and Tankut, 2009) were tested
under vertical and quasi-static lateral loading simulating
earthquake effect. Except from the bare and RC infilled, all
frames exhibited typical masonry infilled frame behaviour
characterized by: Rather rigid and linearly elastic behaviour
at the initial stages, relatively high capacity resulting from
infill wall contribution and rapid strength degradation and
very rapid stiffness degradation upon infill wall crushing.
This expected behaviour was concluded by a typical
failure accompanied by excessive permanent first story
sway deformations. It is important to note here that
although not plastered at both sides, test frame SP2 also
exhibited typical masonry infilled frame behaviour. It was
tested to observe effects of plaster application on infilled
frame behaviour. Test frame SP6 was tested to observe
for monitoring the critical column section
the RC infilled frame behaviour which forms an upper
bound. As expected, this specimen behaved as a
monolithic cantilever where failure took place at foundation
level with column bases in terms of yielding of the steel in
the tension side column and concrete crushing and
buckling of longitudinal steel in the compression side
Test of SP5 Specimen was terminated since the
diagonal crack just below the first story beam – left column
joint turned out to be a shear failure and the column broke
off due to low concrete strength of the frame. Story Drift
Ratio is a term which is frequently used in the earthquake
engineering as a measure of non-structural damage and to
control second order effects. First and second story drift
ratio values of the test frames at ultimate load levels are
given in Table 2 and lateral load-first story drift ratio curves
for the test frames are given in Figure 5. According to the
story drift ratio curves, hollow brick infill walls and
plastering increased lateral strength and stiffness. Bare
test frame SP1 reached 1.60% lateral drift at ultimate load.
This ratio was 1.13% for test frame SP2, with non-
plastered hollow brick infills. As expected, test frames SP3,
SP4 and SP5 with plastered hollow brick infills reached
story drift ratios of 0.43, 0.42 and 0.35% values
Test frame SP6, with RC infills, had a drift ratio of 0.79%
at its ultimate load which was lower than that of test frame
SP1 but higher than masonry infilled specimens. As
expected, major damage took place in the first story infill
wall for all two-story test frames. In addition, first story drift
ratio values were higher than that of second story at
ultimate load level for all test frames.
One-story test frames SP7, SP8 and SP9 had drift ratios
of 0.36, 0.65 and 0.53%, respectively. The value for test
frame SP7 was less when compared to the other two,
since this frame had continuous column longitudinal bars
together with higher axial column loads. Drift ratios were
0.42 and 0.43% for the test frames SP4 and SP3, which
were the equivalent pairs for test frames SP7 and SP8,
having drift ratios of 0.36 and 0.65% respectively.
Equivalent pairs had all the variables same except the
number of stories that test frames had. As expected, first
story drift ratio value for test frame SP3 was greater than
that of test frame SP4 and value for test frame SP8 was
greater than that of test frame SP7 since higher axial load
made the infills and frames much stiffer.
According to the Turkish Seismic Code (2006),
maximum story drift index is limited to 0.0035 in the elastic
analysis of the structure whereas, it is specified as 0.010
for inelastic analysis. On the other hand, according to
clause 1630.10 of UBC (Uniform Building Code, 1997), the
maximum story drift index is limited to 0.025 for the
structures with a fundamental period less than 0.7 s and
0.020 for the structures with a fundamental period greater
than 0.7 s. Hollow brick infilling reduces the amount of
deformations as compared to bare test frame SP1.
Baran and Sevil 1987
Lateral Load (kN)
Lateral Load (kN)
Lateral Load (kN)
Lateral Load (kN)
First Story Drift
Lateral Load (kN)
First Story Drift
Lateral Load (kN)
Figure 5. Lateral load-first story drift curves.
1988 Int. J. Phys. Sci.
DISCUSSION OF TEST RESULTS
Strength and stiffness
Test frame performances are evaluated in terms of load-
top displacement, energy dissipation and initial stiffness
values as summarized in Table 2. When the results in
Table 2 are examined, it can clearly be observed that
there was no significant difference between the lateral
load capacities of test frames SP4 and SP5 although one
of them had continuous longitudinal bars through the
height of the specimen whereas the other had lap splices
at both floor levels with a length of 20φ (160 mm). This
situation was owing to the level of the axial load applied
on to the columns during the experiments of these two
specimens. Total axial load on both columns was
approximately 117.7 kN during the experiments of both
specimens. This load level corresponded to 20% of the
column’s axial load capacity, which can be considered as
high. With the application of relatively higher axial load
level on both columns of test frame SP5, the lap splice
effect could not be observed at a lateral load level of
approximately 75 kN which was the lateral load capacity
of both test frames. However, when total axial load on
both columns is 10% of the column’s axial load capacity,
lateral load capacity of the frame decreased to 65 kN
level, as in the case of test frame SP3. It should be noted
that lateral load capacity of test frame SP3 is less than
that of test frame SP5 although there were no lap-splices
in test frame SP3 but had lower axial load on its columns.
This situation shows the importance of the column’s axial
load level on the strength of the RC test frame. In
addition, the decrease is more pronounced in the case of
one story test frames.
Test frame SP2 had non-plastered hollow brick infills in
contrast to test frames SP3, SP4 and SP5. As expected,
this specimen had lower lateral strength (about 50 kN) as
compared to SP3, SP4 and SP5. The ratio of strength
increase as compared test frame SP1 was almost 3.5
times of test frame SP2, where this value was
approximately 5.0 times for test frames SP3, SP4 and
SP5. This shows the importance of the effect of hollow
brick infilling on the RC frame behaviour, although it was
non-plastered. In addition, plastering the hollow bricks
obviously enhances the strength increase that is supplied
only by hollow brick infilling. However, non-plastered
hollow brick infilled frame SP2 showed more ductile
behaviour than the plastered brick infilled frame. This
can be attributed to the fact of higher stiffness of plaster
than masonry. Test frames SP4 and SP3 had maximum
lateral loads of 78.8 and 66.6 kN, respectively. These
values were 86.6 and 62.3 kN for the respective
equivalent pairs of one-story test frames.
Strength and stiffness characteristics together with the
general behaviour of specimens were evaluated by the
help of response-envelope curves shown in Figure 6,
which were constructed by connecting the peak points of
the hysteretic load-displacement curves of the test
frames for each forward and backward cycle. For two-
story specimens, second story level displacements were
used. However, in order to be able to make a
comparison; first story level displacements were used in
the comparison of all test frames. These curves indicate
that hollow brick infills significantly increase strength and
stiffness and improve ductility
photographs of all specimens after failure are given in
The initial stiffness of a specimen was calculated by
using the slope of the linear part of the first forward load
excursion (Baran, 2005). It was used as a relative
indicator in improvement of the rigidities of test frames.
As it can be seen in Table 2, hollow brick infills increased
initial stiffness of specimens significantly.The increase
was nearly 20 times for two story hollow brick infilled test
frames and approximately 30 times for one story test
frames. The variation in the initial stiffness values for the
two groups can be owing to the number of stories. It
should be noted that quality of the workmanship in the
construction of the hollow brick infill wall and plastering of
the specimen played an important role in the
displacement history in early cycles.
Energy dissipation capacities
Energy dissipation capacity is an important indicator of
the structure’s ability to withstand severe ground motions.
Energy dissipation capacity (Baran, 2005) is an important
indicator of the improved behaviour. For specimens, the
amount of dissipated energy was determined by
calculating and adding the areas under the lateral load -
hydraulic jack level displacement curves for each cycle.
For one-story test frames, hydraulic jack is at the level of
first story beam which means that lateral load-top
displacement graphs were used for energy dissipation
calculations. It is important to note here that the energy
dissipation characteristics of the test frames strongly
depends on the loading history. The loading histories of
the test frames were intended to be the same, but when
the response of a test frame became non-linear,
backward and forward half cycle loadings were controlled
by top story level displacements. The same top story
level displacements were intended to be reached for the
forward and backward half cycles. Total amount of
dissipated energy of each specimen is tabulated in Table
2. As it can be seen in this table,that hollow brick infilling
improve the energy dissipation characteristics of the test
Displacement ductility is defined by the ratio of the
ultimate displacement to yield displacement. The
of frames. The
Baran and Sevil 1989
Figure 6. Envelope load-displacement curves (for two-story and one-story test frames).
ultimate displacement is defined as the top story level
displacement at which the lateral load dropped to 85% of
the maximum applied load at post peak region. The yield
displacement was described with a secant drawn starting
from the origin and passing from the point on which
lateral load is 70% of the maximum applied load. This
secant line was extended up to the horizontal line drawn
from the maximum load and corresponding displacement
was accepted as yield displacement (Sezen and Moehle,
2004; Sevil, 2010). The calculated ductility values are
listed in Table 2. As it can be observed, one-story test
frames SP7 and SP8 showed more ductile behaviour
than equivalent two-story test frames SP4 and SP3,
respectively. In addition, two-story test frame SP4 and
one-story test frame SP7, which had higher column axial
loads (nearly 25% of column axial load capacity) showed
more ductile behaviour than two-story test frame SP3 and
one-story test frame SP8, respectively which had lower
column axial loads (nearly 10% of column axial load
capacity). This situation can be owing to the more
dominant compressive and shear stresses in one-story
frames and more efficient behaviour of the infill, which
can be positively influenced by the confining effect of
compressive forces. Although test frame SP5 had two-
stories, it showed more ductile behaviour than one-story
test frame SP9 which had lower column axial load. It
should be noted here that, test frame SP2, which had
non-plastered hollow bricks, showed more ductile
behaviour than bare test frame SP1 and RC infilled test
1990 Int. J. Phys. Sci.
Figure 7. Test frames after failure.
Infill wall modeling
Beginning with the first study conducted by Polyakov
(1957), analytical and experimental studies on infills have
been conducted for nearly fifty years. During his studies,
Polyakov observed diagonal cracks in the center region
of the infill, seperation over a finite length of the beam
and the column between the frame member and the infill
at the unloaded corners and full contact between them
adjacent to two opposite loaded corners. In the 1960’s,
Baran and Sevil 1991
Figure 8. Equivalent diagonal compression strut replacing infill and orthotropic
model for infill.
Smith (1962, 1966, 1967, 1968) and Smith and Carter
(1969) modeled the infill walls as equivalent diagonal
compression struts. In the 1970’s (Mainstone and Weeks,
1970; Mainstone, 1974; Klingner and Bertero, 1978) in
the 1990’s (Paulay and Priestley, 1992; Angel et al.,
1994; Saneinejad and Hobbs, 1995) and in the early
2000’s (Al-Chaar, 2002; El-Dakhakhni et al., 2003)
conducted analytical and experimental studies on the infill
walls and contribute to better understanding of the infilled
frame behaviour.Results obtained by Smith and Carter
(1969) showed similarity to experimental results obtained
by Mainstone (1974) and Al-Chaar (2002).
In their studies, Smith and Carter (1969) assumed that
the frame and the infill are not bonded together. When
the load is applied, the frame and the infill seperates over
a finite length of the beam and the column and the
contact between them remains adjacent to two opposite
corners. At this stage, a line drawn from one loaded
corner to the other represents the direction of the
principal compression. Therefore, the panel transfers
compression along this line. In fact, it can be assumed
that the infill behaves as a diagonal strut and the
structure can be analyzed with equivalent struts replacing
the infills, as shown in Figure 8. A diagonal compression
strut can adequately represent load transfer mechanism
observed from the experiments and conducted finite
element analysis. Here α and β are the interaction
distribution parameters as presented in Figure 8. In the
case of infills with masonry materials, Equations 1 and 2
are proposed by FEMA (1998) for the determination of
the mechanical and the geometrical properties of the
equivalent diagonal strut;
1992 Int. J. Phys. Sci.
where ainfill is the effective width of the equivalent
diagonal strut, λ is a dimensionless parameter, hcol is the
column height between centerlines of beams, d is the
diagonal length of infill panel, Einf is Young’s modulus of
the infill, bw is the thickness of the infill, βs is the angle
whose tangent is infill height to length, E is Young’s
modulus of the column, I is the moment of inertia of the
column and hinf is the height of the infill.
The equivalent compression strut shall have the same
thickness as the infill it represents.
In the analytical studies conducted, plastered hollow
brick infill walls were modeled by equivalent diagonal
compression struts. Therefore, axial strength (fcm) and
stiffness of the struts should be computed. The axial
strength (fcm) and stiffness of the struts can be obtained
from the tests of square plastered hollow brick infill
panels under diagonal compression. However, in the
absence of the panel tests, Equations 3 and 4 proposed
by Binici and Ozcebe (2006), can be used to predict the
strength and stiffness of the plastered hollow brick infill in
Ebrick is the Young’s modulus of the infill material.
Binici and Ozcebe (2006) proposed its value to be
varying between 500 to 1500 times the compressive
strength of the infill. Hollow bricks used in the infills of the
test frames were loaded in the direction of (Duvarci,
2003) and perpendicular to the holes and the results are
given in Table 3.
Since the infill is diagonally compressed when the
infilled frame is loaded laterally, El-Dakhakhni et al.
(2003) made a justifiable assumption that the properties
in the diagonal direction are the governing material
properties. Plastered hollow brick infills are anisotropic.
At this point, another assumption is made by considering
the anisotropic infill as orthotropic. Since the infill of the
test frames behave as it is under compression, Equation
5 derived by using constitutive relations of orthotropic
plates and axes transformation matrix, can be used to
obtain the Young’s modulus of the infill in the diagonal
Einfill-0 and Einfill-90 are Young’s modulus of the infill in the
direction parallel to and normal to mortar bed joints
respectively, υ0-90 is Poisson’s ratio defined as the ratio of
the strain in the direction normal to the mortar bed joints
to the strain in the direction parallel to the mortar bed
joints. υ0-90 can be taken as 0.25 and Einfill-0 as half of Einfill-
90. G is shear modulus.
The use of Equation 5 for unreinforced concrete infill
walls reduces Young’s modulus in the inclined direction
to about 75% of that in the direction perpendicular to the
mortar bed joints. Although depending mostly upon the
hollow brick’s void ratio, an average ratio of 70% can be
taken as for the case of plastered hollow brick infills.
Initial Young’s modulus is commonly related to ultimate
compressive strength of concrete or masonry like
materials. It would be a justifiable assumption that not
only Young’s modulus, but also the ultimate strength of
the infill in the θ direction, finfill-θ, also changes. A
simplification can be made at this point for taking into
account the variation in direction by using a smaller factor
relating Einfill-θ to finfill-θ and Einfill-90 to finfill-90, since the infill
wall is anisotropic. The assumption that compressive
strength of the infill varies according to the angle of
loading was investigated by Hamid and Drysdale (1980)
and a value of finfill-θ = 0.7finfill-90 was suggested by Seah
(1998). The orthotropic model for the infill given by El-
Dakhakhni et al. (2003) is illustrated in Figure 8.
Nonlinear finite element
Saneinejad and Hobbs (1995) suggested that the secant
stiffness of the infilled frames at the peak load to be half
the initial stiffness. This suggestion might be adapted to
the calculation of the Young’s modulus at peak load, Einfill-
p = 0.5Einfill-θ.
A trilinear relation stress-strain diagram for concrete
masonry infill is suggested by El-Dakhakhni et al. (2003)
instead of the parabolic one as shown in Figure 9.
Accordingly, this approximation is simpler and more
practical for analysis. Accepting the strain ε2 equal to the
strain ε1, using an average value of Einfill-0 and Einfill-90 for
modulus of elasticity and accepting finfill-θ = 0.6finfill-90 yield
satisfactory estimations for the deformation capacity of
the equivalent compression strut. Hence, a stress-strain
diagram as shown in Figure 10 for the equivalent
compression strut modeling the plastered hollow brick
infill wall, was used in the analytical studies. Test results
showed that axial load applied on the frames increased
the push over capacity of the specimens. Therefore, axial
load should have effect on the ultimate load carrying
analysis conducted by
Baran and Sevil 1993
Figure 9. Simplified stress-strain diagram of concrete.
Figure 10. Stress-strain diagram for the compression strut modelıng plastered hollow brick infill.
capacity of the strut and should be taken into
account.The ultimate load carrying capacity and the yield
deformation of the strut were calculated by using
In Equation (6) γ is a variable due to column axial load
effect on the ultimate load carrying capacity of the strut.
When test results are analyzed, adjusting γ as in
Equation (7) seems to be a practical and safe
An upper level of 1.3 for γ is proposed since a maximum
axial load of approximately 58.8 kN corresponding to
3 . 1
30% of column’s axial load capacity was applied during
the tests. Taking the ultimate strain of the equivalent
compression strut as εu = 0.018 yields satisfactory results
in analytical studies.
In case of test frames with lapped-splices on column
steels, the yield stress could not be reached at some
regions due to insufficient lapped-splice lengths at floor
levels. At the joints, the yield stress was decreased
proportional to the splice length of the longitudinal steel
and interaction curves of these sections were calculated
by using reduced yield stresses. For these test frames, it
was intended to compute the column capacities by using
the actual lapped-splice strengths. It is known that nearly
the full yield stress of longitudinal steel can be used in the
calculations when the lapped-splice length is not less
than 40 φ. Hence, yield stress of the longitudinal steel, fy,
1994 Int. J. Phys. Sci.
Figure 11. Analytical model of the test frames.
Table 3. Results of compression tests of tiles (MPa).
Tile no. Failure load (kN) Compressive strength (Net area)
Compressive strength (Gross area)
17.18 10.05 2.77
can be decreased proportional to the square root of
lapped-splice length of the steel. Since the lapped-splice
length at the floor levels were 20 φ, the reduced yield
stress of the longitudinal steel, fy′ can be calculated using
Equation (9) given by Canbay and Frosch (2005):
Push-over curves of the test frames were drawn to be
able to compare the experimental results with those
obtained from analytical studies conducted. Push-over
analysis is a kind of nonlinear static analysis procedure
that is generally used to evaluate the performance of the
structures under lateral loads. In the push-over analysis,
a load pattern is selected first and applied to the structure
in incremental steps. The computer program accepts
axial load-moment interaction curve or just yield moment
values of the members. In the present study, interaction
curves were used for columns whereas just yield moment
values were used for the beams idealizing beam
behaviour as elasto-plastic. Lapped-splices in the column
longitudinal steels (if exist) were taken into account.
Analytical model of the test frames is given in Figure 11.
It is assumed that the equivalent compression struts were
hinge-connected to the frames at both ends.
As it can be seen in Figures 12, 13 and Table 4,
Baran and Sevil 1995
-40-30-20-100 10 20 30 40
Lateral Load (kN)
2nd Storey Level Displacement (mm)
Lateral Load (kN)
2nd Storey Level Displacement (mm)
Lateral Load (kN)
Figure 12. Comparison of analytical and experimental load-displacement curves (two-story).
push-over analysis (Baran et al., 2010) made by the
proposed analytical method, where plastered hollow brick
infills modeled by equivalent diagonal compression struts,
gave safe and sound results in estimating the ultimate
load capacities of the test frames. With the proposed
method, the deviation in the estimation of the ultimate
load carrying capacities of the test frames stated in ±
about 10% range of the experimental values. In addition,
post-peak portions (descending portions) of the push-
over curves were adequately simulated by the proposed
method. However, initial stiffness values of the infilled
test frames could not be estimated within acceptable
closeness for all test frames. This can be owing to the
quality of the workmanship in the hollow brick infill wall
and plastering of the specimen which played an important
role in the displacement history in early cycles. Since the
proposed method is for hollow brick infilled RC frames,
push-over analysis for bare test frame SP1 and RC
infilled test frame SP6 were not conducted.
The conclusions presented below are based on the
limited data obtained from tests of RC frames and
analytical studies conducted. The plastered hollow brick
infills, used as partition walls, increased both strength
and stiffness of frames. In the test of frames with
masonry infills, the increase in strength was nearly 6
times as compared to the bare frame for both frame
types. This increase in initial stiffness was nearly 20 and
30 times for two and one-story test frames, respectively.
For RC infilled frame, the increase in strength and initial
stiffness was nearly 15 and 60 times as compared to the
bare frame. This proved the effectiveness of the method
in improving the overall seismic structural performance.
Application of plaster on both sides of the hollow brick
infill increased lateral load carrying capacity of the frame.
The increase was nearly 3.5 times as compared to the
bare frame. For both types of frame, one of the main
difference is the application level of loading. In two-story
frames, the lateral load was applied at a greater height
and therefore moment arm is greater resulting in more
overturning effect. Therefore, more tensile stress occurs
at the tension side column of two-story frames. However,
compressive and shear stresses are more dominant in
one-story frames. This is the most possible reason for
higher initial stiffness of one-story frames.
1996 Int. J. Phys. Sci.
Lateral Load (kN)
1st Storey Level Displacement (mm)
Lateral Load (kN)
-40-30 -20 -100 102030 40
1st Storey Level Displacement (mm)
Lateral Load (kN)
Figure 13. Comparison of analytical and experimental load-displacement curves (one-story).
Table 4. Comparison of experimental response curves with the analytical push-over curves.
Ultimate load (kN)
SP2 50.3 55.4
SP3 66.6 69.0
SP4 76.8 79.0
SP5 74.2 84.2
SP7 86.6 82.3
SP8 62.3 71.4
SP9 65.5 68.7
(1)Ratio of the experimental data to the analytical data.
Initial stiffness (kN/mm)
Two-story and one-story equivalent test frames showed
very similar behaviour, especially lateral load capacities
of equivalent pairs were close. Presence of inadequate
(20 bar diameter) lapped-splices on column longitudinal
steels did not seem to adversely affect the infill
effectiveness significantly, if the column axial load was
not less than 20% of its axial load capacity. Hence, bond
problems due to lapped-splices on column steels would
not be critical in the cases when the axial load level on
the columns are not very low. Independent from the
presence of lapped-splice in steel, lower axial load on
columns created a negative effect on the lateral strength.
Hence, it can be concluded that higher column axial
loads made the infills stronger which provided higher
lateral load capacity to the frame. This phenomenon was
taken into account in calculating the ultimate load
carrying capacity of a compression strut modeling the
plastered hollow brick infill.
The proposed equivalent diagonal compression strut
modeling showed good correlation with the test results. In
the structural design process, equivalent diagonal
compression struts modeling the plastered hollow brick
infills can easily be added to the existing frame model of
the buildings. By this way, considerable amount of time
and work might be saved by the use of this method which
enables the quick determination of the ultimate load
carrying capacities of the frames with plastered hollow
Baran and Sevil 1997
Lıst of symbols
: Reinforcing bar diameter
: Angle whose tangent is infill height to length
: Relative displacement between two successive floors
: A variable due to column axial load effect on the ultimate load carrying capacity of the
: A dimensionless parameter
: Poisson’s ratio defined as the ratio of the strain in the direction normal to the mortar bed
joints to the strain in the direction parallel to the mortar bed joints
: Ultimate strain of the equivalent compression strut
: Effective width of the equivalent diagonal strut
: Thickness of the infill
: Diagonal length of infill panel
: Young’s Modulus of the column
: Young’s modulus of the infill material
: Young’s Modulus of the infill
: Young’s modulus at peak load
: Young’s modulus of the infill in the direction parallel to mortar bed joints
: Young’s modulus of the infill in the direction normal to mortar bed joints
: Young’s modulus of the infill in the θ direction
: Young’s modulus of the plaster
: Ultimate load carrying capacity
: Ultimate strength of the infill material
: Axial strength
: Ultimate strength of the infill in the direction parallel to mortar bed joints
: Ultimate strength of the infill in the direction normal to mortar bed joints
: Ultimate strength of the infill in the θ direction
: Ultimate strength of the plaster
: Yield stress of the longitudinal steel
: Reduced yield stress of the longitudinal steel
: Shear modulus
: Story height
: Column height between centerlines of beams
: Height of the infill
: Moment of inertia of the column
: Column axial load level
: Thickness of the infill material
: Thickness of the plaster
Al-Chaar G (2002). Evaluating Strength and Stiffness of Unreinforced
Masonry Infill Structures. Construction Engineering Research
Angel R, Abrams DP, Shapiro D, Uzarski J, Webster M (1994).
Behaviour of Reinforced Concrete Frames with Masonry Infills.
Structural Research Series No.589, University of Illinois at Urbana-
Champaign, UILU ENG 94-2005: 183.
Baran M (2005). Precast Concrete Panel Reinforced Infill Walls for
Seismic Strengthening of Reinforced Concrete Framed Structures.
Ph D Thesis, Middle East Technical University, Ankara, Turkey.
Baran M, Okuyucu D, Tankut T (2009). Seismic Strengthening of R/C
Framed Structures by Precast Concrete Panels (Experimental
Studies) (in Turkish). Int. J. Eng. Res. Dev., 1(1): 63-68.
Baran M, Tankut T (2009). Effect of Lapped Splice Deficiency on
Behaviour of R/C Frames (in Turkish). Int. J. Eng. Res. Development,
Baran M, Canbay E, Tankut T (2010). Seismic Strengthening with
Precast Concrete Panels – Theoretical Approach (in Turkish).
UCTEA, Turkish Chamber of Civil Engineers, Tech. J., 21(1): 4959-
Binici B, Özcebe G (2006). Seismic Evaluation of Infilled Reinforced
Concrete Frames Strengthened with FRPS. Proceedings of the 8th U.
S. National Conference on Earthquake Engineering. San Francisco,
California, USA, Paper No. 1717.
Canbay E, Frosch RJ, (2005). Bond Strength of Lapped-Spliced Bars.
ACI Structural Journal, July-August 2005, Title no: 102-S62.
Duvarci M (2003). Seismic Strengthening of Reinforced Concrete
Frames with Precast Concrete Panels. MS Thesis, Middle East
1998 Int. J. Phys. Sci.
Technical University, Ankara, Turkey.
El-Dakhakhni WW, Elgaaly M, Hamid AA (2003). Three-Strut Model for
Concrete Masonry-Infilled Steel Frames. J. Struct. Eng., pp. 177-185,
Federal Emergency Management Agency (FEMA) (1998). Evaluation of
Earthquake Damaged Concrete and Masonry Wall Buildings, FEMA
Hamid AA, Drysdale RG (1980). Concrete Masonry under Combined
Shear and Compression Along the Mortar Joints. ACI J., 77: 314-320.
Kaplan SA (2008). Dolgu Duvarlarin Betonarme Taşiyici Sistem
Performansina Etkisi. Türkiye Mühendislik Haberleri, 452-2008/6.
Klingner RE, Bertero V (1978). Earthquake Resistance of Infilled
Frames. J. Struct. Division ASCE, 104, June.
Mainstone RJ, Weeks GA (1970). The Influence of Bounding Frame on
the Racking Stiffness and Strength of Brick Walls. 2nd International
BrickMasonry Conference held at Watford, England, pp. 165-171.
Mainstone RJ (1974). Suplementary Note on the Stifness and Strengths
of Infilled Frames. Building Research Station UK, Current Paper
Okuyucu D, Tankut (2009). Effect of Panel Concrete Strength on
Seismic Performance of RC Frames Strengthened by Precast
Concrete Panels. WCCE-ECCE-TCCE Joint Conference: earthquake
& tsunamı held at Istanbul, Turkey, IMO Publication Nr.: E/09/03,
Paulay T, Priestley MJN (1992). Seismic Design of Reinforced Concrete
and Masonry Buildings, New York, John Wiley.
Polyakov SV (1957). Masonry in Framed Buildings; An Investigation into
the Strength and Stiffness of Masonry Infilling (English Translation),
Saneinejad A, Hobbs B (1995). Inelastic Design of Infilled Frames. J.
Struct. Eng., 121(4): 634-650.
Seah CK (1998). Universal Approach for the Analysis and Design of
Masonry-Infilled Frame Structures. Ph D Thesis, University of New
Brunswick, Fredericton (NB), Canada.
Sevil T (2010). Seismic Strengthening of Masonry Infilled R/C Frames
with Steel Fiber Reinforcement. Ph D Thesis, Middle East Technical
University, Ankara, Turkey.
Sevil T, Baran M, Canbay E (2010). Investigation of the Effects of
Hollow Brick Infills on the Behavior of Reinforced Concrete Framed
Structures; Experimental and Analytical Studies. International Journal
of Research and Development, Accepted for Publication.
Sezen H, Moehle JP (2004). Shear Strength Model for Lightly
Reinforced Concrete Columns. J. Struct. Eng. ASCE, 130: 1692-
Smith BS (1962). Lateral Stiffness of Infilled Frames. ASCE J. Struct.
Div., 88: 183-199.
Smith BS (1966). Behaviour of Square Infilled Frames. ASCE J. Struct.
Div., 92, ST. 1.
Smith BS (1967). Methods for Predicting the Lateral Stiffness and
Strength of Multi-Storey Infilled Frames. Build. Sci., 2: 247-257.
Smith BS (1968). Model Test Results of Vertical and Horizontal Loading
of Infilled Specimens. ACI J. August, pp. 618-624.
Smith BS, Carter CA (1969). Method of Analysis for Infilled Frames.
Proc. ICE., pp. 44: 31-48.
Turkish Seismic Code (2006). Ministry of Public Work and Settlement,
Government of Republic of Turkey, Ankara.
Uniform Building Code (1997). 5360 Workman Mill Road, Whittier,
California 90601-2298, USA.