EFFECT OF INFILL WALL STIFFNESS VARIATIONS ON THE

BEHAVIOR OF REINFORCED CONCRETE FRAMES UNDER

EARTHQUAKE DEMANDS
Egemen SÖNMEZ1 and Cemalettin DÖNMEZ2

ABSTRACT

Interstory drift concentration under seismic demands is a common issue in reinforced concrete frames
which may result in premature failures. In this study, an organized stiffness distribution along the
building height was studied to mitigate the drift concentrations. The targeted stiffness distribution was
obtained by placing infill walls whose stiffness and strength properties vary with the elevation of the
frame. IDARC-2D was used for structural analysis. Experiments performed by others were used to
verify the hysteretic parameters of the numerical model. Then, three five-bay, five-story, planar frames
were designed. These were a bare frame and the frames with infill walls – one with uniform stiffness
and the other with varying stiffness and strength properties through the elevation. Nonlinear dynamic
and pushover analysis methods were used to observe the performance of the designed frames. The
results showed that an organized stiffness distribution could mitigate the drift concentrations at the
lower stories and result in a more balanced interstory drift distribution along the height of the frame.

INTRODUCTION

Reinforced concrete (RC) structures with infill walls are the most common building types in
earthquake-prone regions of Turkey. The infill walls are generally neglected in structural design
process due to the complications encountered in modeling them and their interaction with the
surrounding frame. However, presence of the infill walls has been proved to affect stiffness, strength
and seismic behavior of the structures significantly. Depending on the capacity demand ratios, infill
walls may be either beneficial or detrimental under seismic demands. Infill walls typically increase the
global stiffness and strength of the structures. This situation may be advantageous for non-ductile
buildings up to a certain limit. On the other hand, brittle nature and a rich variety of failure modes of
infill walls may cause unforeseen and irreversible damages. Particularly, soft-story mechanisms may
occur due to drift concentrations at lower stories of multi-story structures. To mitigate the negative
effects of the infill walls, an organized stiffness distribution along the height of the structure may be
applied by using infill walls with different stiffness and strength properties.
The main purpose of the study is to investigate the effects of stiffness variations in infill walls to
the seismic behavior of the frames. In order to achieve the purpose, an analysis software which
supports infill walls, was selected. Firstly, the infill wall model parameters of the selected software
were calibrated and verified by simulating a series of previously performed experiments. Afterwards, a
planar, five-story, five-bay RC frame was designed according to the Turkish Earthquake and RC
Codes. The stirrup spacing of the members at the confinement regions were designed with the typical
deficiencies observed in residential buildings in Turkey. Then two types of infilled frames were
1
Research Assistant, İzmir Institute of Technology, İzmir, [email protected]
2
Assoc. Prof. Dr., İzmir Institute of Technology, İzmir, [email protected]

1
obtained by introducing infill walls into the bare frame. The infill walls of the first infilled frame had a
uniform stiffness and strength distribution along the height of the building. In the second infilled
frame, the infill walls had a decreasing stiffness and strength distribution through the elevation of the
frame. These three frames were analyzed using nonlinear dynamic and pushover methods. The results
were compared in terms of interstory drift ratios.

MODELING THE NONLINEAR BEHAVIOR OF INFILLED RC FRAMES

A numerical study was performed to estimate the hysteretic control parameters and to verify the
selected analysis software. In this study, a series of experiments conducted by Çankaya (2011) were
simulated using IDARC-2D Version 7.0.
Four single-bay, 1/5-scale, four-story, planar RC frames were subjected to displacement-based
pseudo-static cyclic loading in the experiments. The parameters of the experiments were the presence
of the infill walls which were built with hollow clay tiles and the reinforcement detailings. Frames #1
and #2 were bare frames with brittle and ductile detailing, respectively; and Frames #3 and #4 were
infilled frames with brittle and ductile detailing, respectively. The details of the test setup and
construction for the frames are presented by Çankaya (2011).
The selected software presents several hysteretic models and provides adjustable parameter sets
to capture the intended behavior. IDARC implements the Three-parameter Park Model (Park et al.,
1987) to represent the nonlinear cyclic behavior of RC members. This hysteretic model has four
parameters to control stiffness degradation, stiffness deterioration and pinching effects. On the other
hand, Smooth Hysteretic Model (Reinhorn et al., 1995) is employed to simulate the cyclic behavior of
infill walls. This model has a wider range of parameters to represent the complex cyclic behavior of
the infill walls.
For modeling the hysteretic behavior of the RC members, a parametric study was performed on
the bare frames. Firstly, the RC members were defined in IDARC in terms of moment-curvature
relationships. Then, the parameters of the Three-parameter Park Model were calibrated for simulating
the experimental results (Table 1).

Table 1. Adapted hysteresis control parameters

Parameter Frame #1 Frame #2 Frame #3 Frame #4
α
200.00 200.00 200.00 200.00
(Stiffness Degradation)
β1
0.01 0.01 0.01 0.01
(Ductility-based Strength Degradation)
β2
0.01 0.01 0.01 0.01
(Energy-based Strength Degradation)
γ
0.20 0.50 0.20 0.15
(Slip or Crack-closing)

The infill model of IDARC utilizes equivalent strut approach to model the infill walls. This
approach replaces an infill panel by two diagonal, compression-only struts. IDARC uses the prism
strength of the infill panel for calculating the strength envelope of an equivalent strut. Since the results
of the prism tests showed a wide distribution, a value in vicinity of the results that yielded the best fit
is selected for data input. Since no other experimental data is available, other strength parameters of
the infill walls were calculated using the formulae proposed by IDARC. The parameters of the Smooth
Hysteretic Model were set to default values except the ductility capacity of the infill wall which is set
to 15.0 and 13.0 to fit the experimental results for Frame #3 and Frame #4, respectively.
The analyses were performed using the displacement-controlled quasi-static loading histories
which were measured at every story level during the tests. The maximum interstory drift ratios applied
in the first story are summarized in Table 2.

2
 E. Sönmez and C. Dönmez 3

Table 2. Applied maximum interstory drift ratios (%) at the first story (each cycle applied twice)
Frame No. Cycle 1 Cycle 2 Cycle 3 Cycle 4 Cycle 5 Cycle 6
#1 0.7 1.1 1.5 1.9 - -
#2 0.7 1.1 1.5 2.1 2.5 3.4
#3 0.1 0.3 0.7 1.0 1.4 1.8
#4 0.1 0.3 0.6 0.9 1.5 2.0

Computed shear force versus interstory drift ratio relations for the first stories of all frames are
presented in a comparative fashion in Figure 1.

Figure 1. First story hysteresis curves of the experimental and analytical systems
For the bare frames (Frame #1 and Frame #2), the overall behavior matches closely with the
experimental results. In addition it can be seen that the pinching behavior observed in the experiment
may be modeled by selecting the proper parameter for the slip. For Frame #3, it can be observed from
the figure that the selected control parameter for the slip behavior simulated the pinching accurately.
Moreover, the infill model used by IDARC represents the contribution of the infill wall to the stiffness
and strength of the system, and the strength degradation at increasing drift levels satisfactorily.
However, in the experiment, Frame #4 reported to exhibit a different failure mechanism than Frame
#3. Shear failure occurred at the upper ends of the first story columns and the upper stories started to
move as a rigid body. The software failed to model this failure mechanism, since the shear modulus of
the columns are accepted infinitely high by default.

EFFECT OF INFILL WALLS ON THE DRIFT DISTRIBUTION OF INFILLED RC

FRAMES

Previous studies have shown that under seismic demands, drift concentration in lower stories of
infilled frames may cause major damages. In this study, it is anticipated that a well-organized stiffness
distribution along the height of the building may help mitigating the concentration phenomena. The
stiffness variations in frames are aimed to be controlled by the infill walls, since they have a major
effect on overall stiffness of structures. It is possible to alter the stiffness and strength properties of the
infill walls by applying available retrofitting or weakening techniques, as well as by changing the
thickness or the masonry material of the infill panels. Therefore, it is assumed that necessary
techniques are used to obtain the targeted stiffness levels of the infill walls.
At the beginning, in order to observe the effects of stiffness variations on the drift distribution of
multi-degree-of-freedom (MDOF) systems, four simple linear systems were constructed. The MDOF
systems had various stiffness distributions to represent the effect of infill walls with various stiffness
properties at the floor levels. Figure 2 presents the selected mass and stiffness properties, and first
mode vibration periods of the selected MDOF systems.

Figure 2. Selected simple multi-degree-of-freedom systems and the first mode periods
The stiffness of a story with typical infill panels is represented with a stiffness of k. The
increasing stiffness values correspond to stories with either thicker or reinforced infill walls. The
decreasing values correspond to stories with weaker infill walls whose stiffness values were accepted
to be reduced. Modal analyses were performed to determine the natural periods and the modal shapes
of the systems. Considering that the deflected shapes of the selected systems are dominated by the first
modes, the interstory drift ratios were followed by the first modes of the systems.
The interstory drift distributions of the linear MDOF systems are presented in Figure 3. The
ratios were normalized with respect to the maximum drift which was observed in the first story of
MDOF-3. This value was normalized such that its value is unity. As it can be observed from the
figure, MDOF-1 has an increasing drift concentration in the lower stories caused by the regular
stiffness distribution in elevation. In MDOF-3, the soft first floor caused a significant concentration on
the drift of the first story. On the contrary, a strong first floor in MDOF-2 limited the drift of the first
story; however the drifts of the upper floors increased compared to those of MDOF-3. The stiffness
distribution of MDOF-4 caused a more balanced drift profile and the drift is well distributed among
the individual floors. Thus, in a system with higher stiffness value at the lower stories which is
uniformly decreasing along the building height, drift concentrations in the lower floors could be
mitigated.

Figure 3. Interstory drift ratios of the MDOF systems

4
 E. Sönmez and C. Dönmez 5

The target structures of this study are the typical residential buildings in Turkey. Considering
that generally the first mode governs the dynamic response of such frames, modifying the first mode
shape may help controlling the interstory drifts. For this purpose, infill walls were utilized to change
the first mode shape by altering the stiffness distribution along the height of the structure.
Three frames were analyzed in this stage of the study. Initially, a planar, five-bay, five-story RC
bare frame (BF) was designed and numerically modeled in IDARC. Then, two types of infilled frames,
namely IF-1 and IF-2, were obtained by placing infill walls into two bays of BF. The infilled frames
have similar infill wall amounts to the average infill ratios in the literature (Bal et al., 2008; İnel, 2009)
with two of the five bays having infill panels. The modeled frames were assumed to be located in a
building that is symmetrical in both directions. The plan of the building and the overall geometry and
dimensions of the modeled frames are presented in Figure 4. Proportioning of the selected frames is
based on the following material properties. The compressive strength of the concrete was assumed to
be 20 MPa. The modulus of elasticity of the concrete was calculated as 28 500 MPa using the formula
proposed by the Turkish RC Code (TS500, 2000). The yield stress and the modulus of elasticity of the
steel were taken as 420 MPa and 200 000 MPa, respectively. A uniform dead load of 0.45 t/m2 and
live load of 0.20 t/m2 were assumed on the slabs of the building. Gravity loads from the slabs were
calculated by triangular distribution to the beams and axial loads to the columns.

Figure 4. Designed frame, member types and dimension details of the RC members
The reinforcement details of the members were initially determined according to the Turkish
Earthquake Code (TEC2007) and the Turkish RC Code (TS500, 2000). Afterwards, the calculated
stirrup spacings for the confinement regions were doubled in all the members in order to represent the
common deficiencies of the residential buildings in Turkey. Modified stirrup spacings were confirmed
to be in accordance with the typical observed values in the residential buildings (İnel, 2009). The
reinforcement details of the beams and the columns are presented in Table 3.

Table 3. Reinforcement details of the designed frames

Location of Beam Type Type of
All Columns
Reinforcement B1 B2 B3 B4 Reinforcement
Bottom 3Ø16 3Ø14 3Ø12 2Ø14
Left Top 4Ø16 + 2Ø12 5Ø14 + 2Ø12 2Ø16 + 2Ø12 3Ø12 + 2Ø12
Stirrup Ø8 / 192 mm Ø8 / 192 mm Ø8 / 192 mm Ø8 / 192 mm Longitudinal 12Ø18
Bottom 3Ø12 3Ø12 3Ø12 3Ø12
Midspan Top 2Ø12 2Ø12 2Ø12 2Ø12
Stirrup Ø8 / 235 mm Ø8 / 235 mm Ø8 / 235 mm Ø8 / 235 mm
Bottom 3Ø16 3Ø14 2Ø14 2Ø14
Stirrup Ø10 / 200 mm
Right Top 4Ø16 + 2Ø12 5Ø14 + 2Ø12 2Ø14 + 2Ø12 3Ø12 + 2Ø12
Stirrup Ø8 / 192 mm Ø8 / 192 mm Ø8 / 192 mm Ø8 / 192 mm

The first infilled frame, IF-1, was aimed to represent the common infill wall practice observed
in the residential buildings in Turkey. Hollow clay tiles with dimensions of 191913.5 cm were
assumed to be used in the infill panels. Total infill wall thickness was assumed 23 cm with 2 cm of
plaster on each side. Weight of an infill panel was calculated to be 0.4 t/m2 based on the information
provided by manufacturers. The prism strength of an infill panel was assumed to be 3 MPa depending
on Paulay and Priestley (1992). The initial stiffness of an individual infill panel was calculated by
IDARC as 80 kN/mm and the lateral yield force as 420 kN.
In the second infilled frame, IF-2, it was aimed to have a pre-organized stiffness distribution
throughout the height of the building by modifying the infill walls used in IF-1. Linear modal analyses
of simple MDOF systems showed that a descending stiffness distribution proportional to the building
height is advantageous in terms of linear interstory drift profile. Depending on this observation, the
infill panels in IF-2 were aimed to have the stiffness properties in a descending form, where the stiffer
infill walls are located in the lower stories and the flexible infill walls in the upper stories. The
stiffness values of the infill panels in IF-2 were determined to be 2.0, 1.6, 1.2, 0.8 and 0.4 times of
those of the infill walls used in IF-1. By this distribution, the lower three stories were accepted to be
reinforced using different methods to obtain the mentioned stiffness values. On the contrary, the upper
two stories were assumed to be weakened using various techniques mentioned in the literature
(Mohammadi et al., 2011; Preti et al., 2012). Since the modification in stiffness affect the strength
properties of the infill walls as well, the yield strength of the reinforced infill panels were assumed to
be increased by 40%, 30% and 20% in the first, second and third floors, respectively. In the fourth and
fifth stories, the yield strength values were assumed to be decreased by 20% and 40%, respectively.
These factors were selected arbitrarily, since no specific modifying method was selected. It is also
considered that IF-1 and IF-2 should have the similar initial period to have comparable seismic
demands for the systems. To obtain the similar periods, the initial stiffness values of the infill walls in
IF-1 were taken 10% higher than the original value. The initial stiffness and the yield strength values
of the infill walls are presented in Table 4. The first natural mode periods of BF, IF-1 and IF-2 were
calculated as 0.74, 0.50 and 0.51 seconds, respectively, by linear modal analyses.

Table 4. Initial stiffness and yield strength values of the infill walls
IF-1 IF-2
Story Initial Stiffness Yield Force Initial Stiffness Yield Force
(kN/mm) (kN) (kN/mm) (kN)
5 88 420 32 252
4 88 420 65 336
3 88 420 97 504
2 88 420 129 546
1 88 420 161 588

Nonlinear dynamic analyses were performed to calculate the seismic response of each frame.
Initially, ten strong ground motion acceleration records were adopted from the study of Lepage (1997).
Table 6 presents the information about the selected ground motions. The records encompass a wide
range of soil conditions and frequency contents. The selected ground motions were normalized with
respect to El Centro 1940 NS scaled to 0.40 g peak ground acceleration using the method proposed by
Lepage (1997) to yield the similar displacement demands. The peak ground accelerations used to
normalize each record are presented in Table 5.

Table 5. Ground motion records used in dynamic analyses

Record Characteristic Original Peak Normalized Peak
Record Duration Period, Tg (sec) Ground Ground
(sec) Acceleration (g) Acceleration (g)
Hachinohe 1968 EW 40 1.14 0.19 0.19
Santa Barbara 1952 S48E 40 1.03 0.13 0.21
Sendai 1978 NS 40 0.95 0.26 0.23
Seattle 1949 S02W 50 0.89 0.07 0.25
Taft 1952 N21E 30 0.72 0.16 0.31
Kobe 1995 NS 30 0.70 0.83 0.31
El Centro 1940 NS 30 0.55 0.35 0.40
Llolleo 1985 N10E 50 0.50 0.71 0.44
Tarzana 1994 NS 30 0.44 0.99 0.50
Castaic 1971 N21E 30 0.35 0.32 0.63

6
 E. Sönmez and C. Dönmez 7

The presented acceleration records were applied to BF and IF-1 to determine the most critical
records in terms of interstory drift distributions. Examination of the results showed that the presence of
the infill walls limited the roof displacement up to a certain point in most earthquakes. However, in
Sendai, Seattle, Kobe and El Centro, the roof displacements of IF-1 were almost equal or higher than
that of BF. In these earthquakes, the drift concentrations at the lower stories of IF-1 are also quite
significant. Individual infill wall hystereses of the frames were then examined and it was seen that the
infill walls at the lower stories had passed to the post-peak stage of their response in the mentioned
ground motions. Beyond the post-peak stage, dramatic decrease in the stiffness and strength of the
infill walls caused drift concentrations at the lower stories, thus formation of soft-story mechanisms.
Based on these observations, dynamic analyses of IF-2 were carried out with these four ground motion
records, namely Sendai, Seattle, Kobe and El Centro, which were determined to be the most critical
motions.
The interstory drift ratios of BF, IF-1 and IF-2 at maximum roof displacements are compared
for each of the selected critical ground motions in Figure 5. In Sendai and Seattle earthquakes, the
interstory drift demands dropped and a well distributed demand obtained. In Kobe and El Centro,
dramatic drops in the lower story interstory drift ratios took place. On the other hand, the infill walls in
those stories still passed to the post-peak stage.

Figure 5. Interstory drift ratios of BF, IF-1 and IF-2 for the selected critical ground motions
The pushover analyses were performed to determine the performance of the three frames
according to TEC2007 using SAP2000, since the infill model in IDARC does not support pushover
analysis. The frames were assumed to be located in the first seismic zone (peak ground acceleration
0.4g) and on a type Z3 (medium dense soil) of local site. The lateral load pattern was calculated
according to the equivalent earthquake load procedure and assigned to the frames at the floor levels.
Using the procedures described in TEC2007, seismic displacement demand was determined for each
frame, and the frames were pushed to the demanded displacements to determine the performance
levels.
In SAP2000, linear behavior of the RC members is determined through the cross sectional and
the material properties. The nonlinear behavior of the RC members was defined by assigning plastic
hinges at both ends of the members. The hinge properties were determined by simplified moment-
curvature with tri-linear relation as in IDARC. Using the strain limits defined by TEC2007 for
concrete and reinforcement, curvature values were calculated for damaged-limit states for each
member. The calculated values defined in SAP2000 for each hinge; therefore it was possible to
determine the damage region of each RC member and the performance of the entire structure,
consequently.
The infill walls were modeled as compression-only equivalent struts in SAP2000. To determine
the nonlinear behavior of the infill walls, displacement-controlled incremental quasi-static analyses
were performed in IDARC for each infill wall. Then, the results were simplified as tri-linear curves
and defined as plastic springs at the mid-point of each equivalent strut.
The pushover curves of the three frames are presented in Figure 6. As expected, the presence of
the infill walls substantially increased the strength and stiffness of the frame. The maximum base shear
versus the frame weight ratio is 0.17 for BF. This ratio increased to 0.27 and to 0.29 for IF-1 and IF-2,
respectively. The maximum shear force was reached in relatively low displacements for the infilled
frames. The initial stiffness of IF-1 and IF-2 is 2.80 and 3.07 times that of the BF, respectively.

Figure 6. Pushover curves of BF, IF-1 and IF-2

The plastic states of the members were evaluated at the end of the analyses and the performance
level of each frame was determined. The bare frame was determined to be in “Collapse Prevention”
level and it did not satisfy the life safety requirements while IF-1 was in “Life Safety” zone and IF-2
was in “Immediate Occupation” zone.
Interstory drift ratios at the code displacement demands for each frame are presented in Figure
7. The maximum interstory drift ratio was observed in the second story of BF as 1.78%. The interstory
drifts of IF-1 was lower that those of BF; however the drift concentration at the lower stories
dominated the behavior. In IF-2, the interstory drifts at the lower floors decreased and those at the
upper floors increased when compared to IF-1. This leads the frame to have a more balanced interstory
drift distribution while preventing the drifts to concentrate in the lower stories.

8
 E. Sönmez and C. Dönmez 9

Figure 7. Interstory drift ratios of BF, IF-1 and IF-2 for pushover analysis

DISCUSSION OF THE RESULTS

As mentioned above, the study consists of two phases. In the first phase, it was intended to calibrate
and verify the selected analysis algorithm, IDARC, by simulating previously performed experiments.
The experiments included two planar bare and two infilled RC frames. In these simulations, IDARC
satisfactorily represented the behavior of the tested frames and the contribution of the infill walls to
their response. When the hystereses for individual loading groups were examined separately, it was
seen that the simulation results showed a good approximation with the experimental results. In Frame
#4, algorithm failed to model the behavior accurately. This is due to the shear failure of the first story
columns during the experiments. IDARC assumes infinite shear modulus in its models. Since, shear
failures were not considered in the planned study, the subsequent analyses were carried out using
IDARC.
In the second phase of the study, it was aimed to investigate the effects of infill wall stiffness
variations on the seismic behavior of the designed frame. To achieve this purpose, seismic responses
of the three designed frames were compared in terms of interstory drift ratios. In the regularly infilled
frame, IF-1, the infill walls limited the interstory drift ratios and consequently, the peak displacements
up to a certain point. For the ground motions in which the infill walls at the lower stories had passed to
the post-peak stage of their response, the stiffness and strength of the lower stories decreased rapidly
and soft-story mechanisms were formed.
It is observed that, the presence of the infill walls can be advantageous, since they shorten the
period of the structure; hence, the seismic displacement demand decreases. However, the uniform
stiffness distribution along the height of the structure could cause drift concentrations in the lower
stories which may force the infill walls to exceed their lateral strength capacity. To mitigate the drift
concentrations at the lower stories, infill walls with varying stiffness and strength properties were used
to obtain an organized stiffness distribution through the height. Observations in linear systems
indicated that uniformly decreasing stiffness - with larger values in the lower stories - provides a better
drift distribution along the height of the structures. Analyses of the nonlinear systems confirm this
observation, the interstory drift distributions of IF-2 were more balanced than those of BF and IF-1.
The organized stiffness distribution also limited the maximum interstory drift ratios when compared to
BF and IF-1, except El Centro, in which the ratios are lower than that of IF-1 but higher than that of
BF.
Pushover analysis was also performed, since it is a common application used to determine the
seismic performance of the buildings in practice. The results indicated that the organized stiffness
distribution improved the performance of the frame when compared to those of BF and IF-1 by
decreasing the drift concentration at the lower stories. Particularly, this limits the damage in the lower
story columns. The interstory drift distribution of IF-2 at the demanded displacement was observed to
have a balanced profile when compared to the other frames.
CONCLUSIONS

The analysis results firstly indicated that within the scope of the presented study the selected software,
IDARC-2D, and its infill wall model which is based on the equivalent strut approach can be used to
simulate the behavior of the RC frames with infill walls under cyclic loading satisfactorily. Though,
the program did not have the capability to simulate the shear failure in the columns that occurred due
to forces developed by the infill walls. The calibration of the hysteretic control parameters using the
experimental data was satisfactory.
The existence of the infill walls increased the stiffness and strength of the frame substantially.
The base shear capacities of the infilled frames were increased by approximately 100%. Moreover, the
initial stiffness values of the infilled frames were as high as three times of the bare frame. Since the
maximum displacement induced by strong ground motion is sensitive to stiffness (Shimazaki and
Sozen, 1988; Lepage, 1997); with decreasing periods, the displacement demand for the infilled frames
decreased. However, the drift concentration in the lower stories of IF-1 caused a soft-story mechanism
in some of the selected ground motions and the upper stories moved together almost as a rigid body.
In some of the selected ground motions, the interstory drift ratios of the lower stories of IF-1
exceeded those of BF. The organized stiffness distribution managed to limit these drifts in IF-2. The
drifts observed in the lower stories of IF-2 were smaller than those of BF, except El Centro ground
motion. The maximum interstory drift levels of IF-2 in El Centro are about the bare frame demands.
Still, it can be concluded that the selected distribution was proved to be advantageous in controlling
drift distribution of the frames.
In the pushover analyses, the existence of an organized stiffness distribution revealed a better
interstory drift distribution along the elevation of the frame. The evaluation of the member damage
states revealed that the selected stiffness distribution improved the seismic performance of the frame
when it was compared to those of BF and IF-1.
The results of the study showed that neglecting the infill walls in structural design can be
extremely misleading while analyzing the seismic response. Expected behavior in structural designs
which ignore the infill walls may not be the actual behavior and unforeseen damages may occur in the
buildings during earthquakes. The results also indicated that, existing infill wall retrofitting and
weakening techniques can be used to improve the seismic performance of the frame structures if they
are executed appropriately throughout the structure. Buildings with stiffer lower stories and a uniform
stiffness change throughout the height were determined to exhibit better seismic performance. This
targeted stiffness distribution along the elevation can be obtained by using infill walls which have
varying stiffness properties.

REFERENCES
Bal IE, Crowley H, Pinho R, Gülay F (2008) “Detailed assessment of structural characteristics of Turkish RC
building stock for loss assessment models”, Soil Dynamics and Earthquake Engineering, 28(10-11), 914-
932.
Çankaya MA (2011) “Dynamic behavior of reinforced concrete frames with infill walls” (Master’s Thesis), Civil
Engineering Department, İzmir Institute of Technology, İzmir, Turkey.
Inel MM, Ozmen HB, Senel SM, Kayhan AH (2009) “Mevcut Betonarme Binaların Yapısal Özelliklerinin
Belirlenmesi”, Sakarya Uluslararası Deprem Sempozyumu: “İlk on yılında Marmara Depreminin
Öğrettikleri”, 1–2 Ekim 2009, Sakarya, Türkiye, 295-304 (In Turkish).
Lepage A (1997) “A method for drift-control in earthquake-resistant design of RC building structures” (Ph. D.
thesis), University of Illinois, Urbana.
Mohammadi M, Akrami V and Mohammadi-Ghazi R (2011) “Methods to improve infilled frame ductility”,
Journal of Structural Engineering, 137(6), 646-653.
Park YJ, Reinhorn AM, Kunnath SK (1987) “IDARC: Inelastic Damage Analysis of Reinforced Concrete Frame
- Shear-Wall Structures”, Technical Report NCEER-87-0008, State University of New York at Buffalo.
Paulay T and Priestley MJN (1992) “Seismic Design of Reinforced Concrete and Masonry Buildings”, John
Wiley & Sons, New York.
Preti MM, Bettini NN, Plizzari GG (2012) “Infill walls with sliding joints to limit infill-frame seismic
interaction: Large-scale experimental test”, Journal of Earthquake Engineering, 16(1), 125-141.

10
 E. Sönmez and C. Dönmez 11

Reinhorn AM, Madan AA, Valles RE, Reichman Y, Mander JB (1995) “Modeling of masonry infill panels for
structural analysis”, Technical Report NCEER-95-0018, National Center for Earthquake Engineering
Research, State University of New York at Buffalo.
Shimazaki K, Sozen M (1988). “Strong ground motion drift and base shear coefficient for RC structures”, In:
Proc. 9th world conference on earthquake eng. vol. 5, 165-70.
TEC2007. Turkish Earthquake Code (2007) “Regulations on Structures Constructed in Disaster Regions”,
Ministry of Public Works and Settlement, Ankara (In Turkish).
TS500 (2000) “Requirements for Design and Construction of Reinforced Concrete Structures”, Turkish
Standards Institution, Ankara.