Experimental and Numerical Assessment of

Download as pdf or txt
Download as pdf or txt
You are on page 1of 13

Journal of Engineering and Applied Sciences

Research Article

Experimental and Numerical Assessment of Masonry Infill on Seismic


Performance of RC Frame Structure
Syed Azmat Ali Shah1*, Asfandyar Ahmed2, Khan Shahzada3, Syed Muhammad Ali4, Akhtar Naeem Khan5
and Akhter Gul6
1
Centre for Disaster Preparedness and Management, University of Peshawar, Khyber Pakhtunkhwa, Pakistan; 2City University
of Science and Information Technology Peshawar, Khyber Pakhtunkhwa, Pakistan; 3University of Engineering and Technology
Peshawar, Khyber Pakhtunkhwa, Pakistan; 4Earthquake Engineering Center, University of Engineering and Technology
Peshawar, Khyber Pakhtunkhwa, Pakistan; 5Civil, Agricultural and Mining Engineering, University of Engineering and
Technology Peshawar Khyber Pakhtunkhwa, Pakistan; 6Department of Civil Engineering UET Campus-III Bannu, Khyber
Pakhtunkhwa, Pakistan.

Abstract: The behavior of solid brick masonry is presented under the application of lateral loads. Two frames
made of reinforced concrete, bare frame and masonry infilled frame, were subjected to displacement controlled
lateral loading. Experimental tests and numerical modeling were performed, and the results compared show
a good relationship. The analysis results match very well with the experimental observations. For numerical
modeling of the infill wall, a modified micro-modelling approach is selected where the head and bed joints
are lumped into the bricks and an interface model was defined between these bricks representing the behavior
of the mortar joint. Appropriate material models and damage criteria are employed for each mode of failure.
A parametric study was then performed to assess the influence of different material properties on the lateral
load capacity of the RC frame infilled with masonry. It was demonstrated that for the same gravity load level
the enhancement of brick unit properties enhances the total capacity of the RC frame with infill. Similarly,
by increasing the gravity load on the structure with the same brick unit properties, the lateral load capacity
increases considerably.
Received: May 02, 2021; Accepted: June 24, 2021; Published: June 30, 2021
*Correspondence: Syed Azmat Ali Shah, Centre for Disaster Preparedness and Management, University of Peshawar, Khyber Pakhtunkhwa,
Pakistan; Email: engrazmatalishah@uop.edu.pk
Citation: Shah, SAA, A. Ahmed, K. Shahzada, S.M. Ali, A.N. Khan and A. Gul. 2021. Experimental and numerical assessment of masonry infill
on seismic performance of RC frame structure. Journal of Engineering and Applied Sciences, 40(1): 24-36.
DOI: https://dx.doi.org/10.17582/journal.jeas/40.1.24.36
Keywords: Experimental testing, Numerical modeling, Infill, Brick masonry, RC frame, Simplified micro-modeling

Introduction the use of the structure. These structural units are very
common in some countries prone to earthquakes like

U nreinforced masonry (URM) panels usually opt


as infills/partition walls in reinforced concrete
(RC) and steel frame structures. These walls have
the United States, Turkey, Italy, China and Mexico
(Change-Hai, 2014). The response of RC infilled
frames, observed in previous earthquakes, is diverse,
serious consequences not only for the safety of life ranging from acceptable to mediocre (Rosenblueth
but also for the ease of maintenance of buildings et al., 1986). This research work focused on the role
(Shing et al., 2009). Sometimes, URM filling walls, of the presence of infill brick masonry wall panel
when typically used as partitions, can adversely affect in reinforced concrete framing structures and their
June 2021 | Volume 40 | Issue 1 | Page 24
Journal of Engineering and Applied Sciences
quantitative analysis. The mechanical behavior of masonry is intricate and
its characterization by laboratory tests is among the
When designing a building, infills are considered non- important tools offered to researchers to comprehend
structural elements. However, if in this case there are it. Nevertheless, laboratory masonry tests require
no separation gaps between the frame and the filling, costly and time consuming resources, as test
they merge with the frame around it (Stavridis and specimens, costly and specific measuring equipment,
Shing, 2010). This interaction can be advantageous and the large scale dispersion of properties. Material
in the case where it increases the strength and and geometric properties of masonry usually require a
stiffness of the building. Both strength and ductility substantial number of samples to obtain characteristic
are factors involved in the seismic performance of a results.
structure and an increase in resistance are likely to
increase performance (Korkmaz et al., 2010). On the There are fascinating studies regarding the
contrary, a disadvantage of the framework and filling understanding of the damage pattern of masonry
interaction is the uncertainty of the response due to under axial load (Andreaus and Ceradini, 1992) and
large uncertainties in the filling properties (Fiore et numerical modeling of masonry by finite element
al., 2014). A range of damage patterns is possible for method (Page, 1978) to understand the behavior
infill frames depending on the comparative stiffness subjected to monotonic lateral loads (Andreaus et
and strength of the infill and adjacent RC frame al., 1985a) and reverse cyclic loads (Andreaus et al.,
(Mehrabi et al., 1994). The infill walls of the URM 1985b). Other studies have also been carried out on
modify the dynamic response of the frame regarding the assessment of masonry columns/piers subjected
the stiffness, natural frequency and damping to cyclic loading (Andreaus et al. ,1988a, b). The study
(Kauffman and Memari, 2014). of axially loaded masonry (Andreaus et al., 1995a,
b), monotonically loaded masonry (Andreaus et al.,
The formation of soft tales is also common for 1995c) and masonry under seismic load (Andreaus et
such units. When significant damage to an infill al., 1995a) are prominent. These approaches depend
panel occurs due to large seismic loads, the risk of on adopting the advanced modeling procedures as
falling debris may also be observed (Kyriakides well as on data acquired from simple tests carried
and Billington, 2008). URM infill walls are usually out on the materials employed for masonry units and
made of fragile materials that quickly lose capacity. mortar joints. Nevertheless, the key purpose of these
The mutual effect of fragility and high rigidity has studies is the corroboration of proposed numerical
a negative influence on the lateral load response of models.
these frames (Hahemi and Mosalam, 2006).
In recent years, a great number of structural models
The ultimate strength of a filled frame is not only have been proposed regarding the response of
the sum of the ultimate strengths of the infill panel structural concrete and masonry, the analytical
and frame (Bertero and Brokken, 1983) but also modeling is based on: plasticity (Chen, 1975; Kang
depends on the relationships between the two. These et al., 2000; Grassl et al., 2001; Thabet and Haldane,
units change their mechanisms of load resistance and 2001), viscoplasticity (Cela, 1998; Gomes and
damage patterns. For low drift levels, a filled chassis Awruch, 2001; Barpi, 2004) damage mechanics
system acts as a monolithic composite wall, but as the (Cervera et al., 1996; Hatzigeorgiou et al., 2001), or
drift increases, the filling may somewhat detach from a mixture of these models (Fardis et al., 1983; Chen
the surrounding frame (Koutromanos et al., 2011). In and Buyukozturk, 1985; Yang et al., 1985; Dube et al.,
this case, the lateral resistance is mainly provided by a 1996; Faria et al., 1998). The subsequent formulations
diagonal compression mechanism. In this mechanism, generally include numeral parameters whose
the compressive stresses generate between the infill assessment is essential to obtain a close association
panel and RC frame. The location of this mechanism between the behavior prophesied by the model and its
is at two diagonally opposite angles in the locality experimentally obtained results.
of the beam/column joints. This mechanism led
several researchers to use an equivalent diagonal bar Most of these relationships are uniaxial stress-strain
to describe the effect of filling, initially proposed by model (Pauw, 1960; Popovics, 1973) that are further
Polyakov (Polyakov, 1960). widespread with the assumption of orthotropic or
June 2021 | Volume 40 | Issue 1 | Page 25
Journal of Engineering and Applied Sciences
non-isotropic material to be used to describe the make the modeling of masonry exceptionally tough.
non-linear response under biaxial (Liu, 1971; Liu et The various numerical techniques could be adopted
al., 1972) or triaxial (Buyukozturk and Sareef, 1985; for masonry models (Lourenco, 2002; Asteris and
Balan et al., 1997, 2000) state of stress. Instead of Tzamtzis, 2003a, b):
developing models using simple stress-strain relation • Masonry as single-phase material (Macro-
coupled to triaxial data, it is likely to propose material Modelling)
models by simply investigating the results acquired • Masonry as double- phase material (Modified
from triaxial tests. Micro-Modeling or Meso-Modeling)
• Masonry as triple-phase material (Detailed
Masonry has distinctive directional properties, because Micro-Modeling).
of the presence of mortar joints which are potential
planes of weakness. Contingent to the positioning A typical large-scale RC frame was chosen from
of the joints concerning applied load, damage can present building practices and tested under lateral
occur only in these joints (sliding) or concurrently in loading conditions in the laboratory. Two of these
both the joints and units. According to FEMA 356, types were constructed, one without infill and with
there are four main damage modes for masonry walls masonry infill. The numerous parameters of the units
or masonry pillars: The diagonal tension of the box, were found in the laboratory, which is presented in
crushing of the toes, sliding of the bed and tilting. Table 1. These tests were used for the selection of
Massive masonry panels usually fail in peak crushing infill materials. Later, these tests were used to select
modes/wedge or sliding shear failure. Therefore, the the appropriate constitutive models for “Brick and
ductility of the infilled RC frames appears to be mortar” in the Numerical modeling studies. The
contingent on the failure mode of the infill masonry key influence of this work is to study the lateral
(Tasnimi and Mohebkhah, 2011). load behavior of RC frame elements with and
without masonry infill and to demonstrate the use
The large quantity of manipulating aspects, for of computer simulations to predict masonry behavior
example, the size and anisotropy/orthotropy of the (i.e. to conduct the parametric study with different
units, the width of the joints and the laying of the joints, mechanical properties).
the material properties and the quality of manufacture

Table 1: Details of instrument layout.


S. No Instrument (Capacity) Location Remarks
1. Load cell (500 kN, 50 ton) Horizontally To measure Horizontal load
2. String pot (500 mm, 19.68 in) Beam center To measure Horizontal displacement
main gauge
3. LVDT (50 mm, 1.97 in) R-D To measure displacement at right, diagonal of the infilled wall
4. LVDT (50 mm, 1.97 in) L-D To measure displacement at left, diagonal of the infilled wall
5. LVDT (50 mm, 1.97 in) CRRB To measure displacement at bottom, right side of the right column
6. LVDT (50 mm, 1.97 in) CLLB To measure displacement at bottom, left side of the left column
7. LVDT (50 mm, 1.97 in) CRLB To measure displacement at bottom, right side of the left column
8. LVDT (50 mm, 1.97 in) CLRB To measure displacement at bottom, left side of the right column
9. Dial gauge (20 mm, 0.79 in) CRRT To measure displacement at top, right side of the right column
10. Dial gauge (20 mm, 0.79 in) CLLT To measure displacement at top, left side of the left column
11. Dial gauge (20 mm, 0.79 in) CRLT To measure displacement at top, right side of the left column
12. Dial gauge (20 mm, 0.79 in) CLRT To measure displacement at top, left side of the right column
13. Dial gauge (10 mm, 0.39 in) BRT To measure displacement at top, right side of beam
14. Dial gauge (10 mm, 0.39 in) BLT To measure displacement at top, left side of beam
15. Dial gauge (10 mm, 0.39 in) BRB To measure displacement at bottom, right side of beam
16. Dial gauge (10 mm, 0.39 in) BLB To measure displacement at bottom, left side of beam
17. LVDT (50 mm, 1.97 in) FLC To measure base slip
18. Load Cell (250 kN, 25 ton) Vertical To measure Vertical load

June 2021 | Volume 40 | Issue 1 | Page 26


Journal of Engineering and Applied Sciences
Materials and Methods

Description of RC frame with and without infill wall


Two single bay, single-story and full-scaled
reinforced concrete (RC) frames were constructed
in the experimental phase of this research. In which
one was bare frame and the other was infilled with
brick masonry as shown the Figure 1. The frame
was designed for Seismic Zone 3 according to the
Building Code of Pakistan Seismic Provision (BCP
SP-2007). The structural detailing of the frame is
shown in Figure 2. The frame was designed as a bare
frame neglecting infilled wall using finite element
software, as it is conventional design practice in
Pakistan. The frames were constructed according to
local constructional practices in Pakistan.
Figure 3: Schematic diagram of the test setup.

Figure 1: Schematic diagram of Bare frame and Infilled


Figure 4: Schematic diagram for the instrumentation
frame (dimensions in inches).
plan of Frames.

The test measurement setup includes two load cells


of capacity 250 kN (25 tons) and 500 kN (50tons) to
monitor vertical and horizontal induce loads, thirteen
LVDTs, one string pot and four dial gauges as shown
in Figure 4, details of instrumentations are given in
Table 1.

The lateral loading protocol in-plane quasi-static


loading was selected from FEMA 461 as shown in
Figure 5, due to its gradually changing deformation
amplitude.

Figure 2: Structural detailing of Frame and Pad 125

100
dimensions in mm (inch).
75

50
Test setup and instrumentation
Displacement (mm)

25

To depict the slab load on frame, an arrangement was 0

made to induce constant axial load on frame through -25

an actuator mounted over frame beam. Point load -50

from the vertical actuator was uniformly distributed -75

-100
on the frame through a girder mounted on top frame -125
beam. A second actuator was used to induce lateral Loading Cycle

displacement as shown in Figure 3. Figure 5: Loading protocol.


June 2021 | Volume 40 | Issue 1 | Page 27
Journal of Engineering and Applied Sciences
Results and Discussion determine the behavior, and consequently discussed
more widely here, whereas the finite element
Figures 6 and 7 exhibit hysteresis curves of Bare method and the nonlinear solution are stated briefly.
frame and Infilled frame respectively. Both the Concerning research specialists in the area of applied
hysteretic loops were symmetrical in the forward mechanics and materials like Framcos, Fib, Rilem, it is
and reverse direction of loading. With the increase known that the significant features to be incorporated
of displacement, there was progressive increase in the in the concrete/brick model are tension stiffness
lateral load. The response of load deformation was and compression confinement (Negulescu, 2010).
approximately linear up to 10 mm lateral displacement Numerous material models including these features
in either direction. After the initiation of cracking, the are applied in the commercially available software
stiffness of both the frames decreased. At high load Atena, which is a set of finite elements intended for
cycles, pinching in the hysteretic loops was observed. the behavior analysis of structures.
The lateral load-deformation (F-D) envelopes are also
shown in Figure 8 for both the bare frame and infilled
frame. Figure 8 shows an obvious difference between
the initial stiffness and lateral load capacity of both
frames. The RC frame without an infill wall has a
peak strength of 51.87 kN at a lateral deformation of
36.21mm. Comparing the maximum capacity of the
bare frame with infilled frame, the maximum capacity
of the infilled frame was 132.5 kN at a displacement
of 36.72 mm. This assessment indicates that the RC
frame with an infill wall has 2.5 times more stiffness
than the RC frame without infill under to similar
loading protocol. The maximum capacity of the Figure 7: Hysteresis curve for Frame with Infill wall.
infilled frame increases 175% as compared to an RC
frame without an infill wall. Also, the lateral stiffness 160
up to ultimate load for RC infilled frame increases 140
Lateral Load (kN)

171% in comparison to bare RC frame. 120


100
80
60
40
20 RC Frame with Infill
0 RC Bare Frame
0 20 40 60 80 100 120
Displacement (mm)
Figure 8: Envelop curves for bare frame and infill frame
in the positive direction.

Geometry and boundary conditions


Figure 6: Hysteresis curves for Bare Frame. Modeling of RC frame
To model the RC frame for both with and without
Numerical modeling infill walls, a material model named “3D Non-linear
The response of the structural components can be cementitious-2” was used for concrete which is a
obtained by a non-linear finite element (FE) analysis. Fracture-plastic model that joins the constitutive
The overall procedure for non-linear analysis has behaviors for tension (fracture) and compression
essentially three (03) fundamental parts: The finite (plastic) (Pukl et al., 2001). The tensile behavior
element approach, the material model and the non- depends on the conventional smeared crack function
linear solutions procedures, which essentially form and cracks band behavior. It uses Rankine criteria of
a balanced estimate. However, the material models exponential softening and can be employed as a fixed
June 2021 | Volume 40 | Issue 1 | Page 28
Journal of Engineering and Applied Sciences
or rotated crack approach. The plasticity model is the micro-modeling approach was adopted where the
Menétrey-Willam failure approach. The model adopts mortar joints are lumped into the brick units and the
a return mapping procedure for the combination of brick units were enlarged to keep the same structural
fracture and plasticity equations. The properties used geometry. This approach allows replicating the
for the model are given in Table 2. The compressive local failure mechanism of the tested structure with
strength reduction factor is used as 0.8 which means reasonable accuracy and less time requirements as
that after tensile cracking the concrete compressive compared to the detailed micro modeling approach.
strength should not drop below 80% of the ultimate It is also worth mentioning that the macro-modeling
strength. Similarly, the fixed crack model coefficient approach has not opted here since it is effective in
is set to be 0.8 which means that after the initiation large structures and saves a lot of computational time
of cracks the crack will be allowed to rotate up to 20% but the local damage mechanism cannot be predicted
reduction of the strength and after reaching the 80% by that approach, also it idealizes the masonry as a
of compressive strength the crack direction will get homogeneous material (Zucchini et al., 2009) which
fixed. The reinforcement in the frame was modelled is not right because masonry is a heterogeneous
as discrete rebars with the same area of the bar material with the distinct unit and mortar properties,
and using the same no of bars as per experimental thus make a big limitation on adopting the macro-
details. The material model used for rebars was “cyclic modeling approach. A detailed description of all the
reinforcement”. A bilinear model with hardening was three modeling methodologies has been presented in
used with a yield strength of 350 MPa. section 1 of this study. As already mentioned, the brick
units were modeled connected by interface elements
Table 2: Properties of concrete for RC frame. (representing the mortar joints and brick-mortar
Material Property Value Units interface) therefore the material models were defined
for both the elements and the properties obtained in
Elastic modulus Ec 24.8 GPa
the test program were adopted to verify the numerical
Poisson’s ratio ʋ 0.2 - models.
Compressive strength fc 20.6 MPa
Tensile strength ft 1.23 MPa
Fracture energy Gf 57.93 N/m
Tension Stiffening coefficient Cts 0.5 -
Plastic strain at strength fc ɛcp 1.02x10 -3
-
-4
Critical compressive disp. wd 5x10 m
Compressive strength reduction rc,lim 0.8 -
Fixed Crack model coefficient 0.8 -

The bare frame model is shown in Figure 9. Where


only the columns and beam are modelled along
with the concrete pad. Since the beam was cast
monolithically with the columns, also the columns
were cast monolithically with the concrete pad, also
the column reinforcement was continuous in the
beam and bottom pad to build a mechanical bond,
therefore a perfect connection was used between all
these elements to avoid any relative displacement of
the elements. To account for the reinforcement bond
with surrounding concrete, the ‘‘perfect connection’’
option in Atena was opted, which applies full bond.

Modeling of the infill panel


To reasonably capture the damage pattern of the
experimentally tested infilled frame, a simplified Figure 9: RC frame model.
June 2021 | Volume 40 | Issue 1 | Page 29
Journal of Engineering and Applied Sciences
Material models joints with reasonable accuracy. This contact model
Modeling the brick unit allows not only the sliding between the contact
The brick unit is modeled as a discrete element with surfaces but also allows the contact to be opened by
distinct properties defined using 3D Nonlinear adjusting the contact opening stiffness (Figures 12,
Cementitious-2 material (Pukl et al., 2001). The main 13).
material relations adopted for this model are presented
in Figure 10. The behavior of the compressive uniaxial
component has been described by a linear total stress
curve (Figure 10) that starts at the beginning.

Figure 10: Material Model for Brick Element (Pukl et Figure 11: Geometry details of Isoparametric brick
al., 2001). element (Pukl et al., 2001).

The behavior of uniaxial tensile material was assumed


linear elastic until failure with a constant elastic
modulus equal to the slope of the first segment of
the compression curve. Based on the laboratory tests
carried out on the brick prisms, various properties
were determined which were employed in the model
description. The main characteristics of this model,
namely elastic modulus, compressive strength, tensile
strength, etc., were taken directly from the results
obtained by laboratory tests of masonry prisms.
The primary parameters used were determined
experimentally and are summarized in Table 2. Figure 12: Failure surface for Interface element (Pukl et
al., 2001).
An iso-parametric brick element was used to model
the masonry unit. This solid element has six degrees
of freedom (DOF) at each node: translations and
rotations in the directions x, y and z. The shape of
the rectangular element corresponds to the regular
pattern of the masonry. The element has plasticity,
tensile stress, large deviation, and large deformation
proficiencies, see Figure 11. Figure 13: Typical interface model behavior in (a) shear
and (b) tension (Pukl et al., 2001).
Modeling the mortar joints
Since the properties of mortar generally dominate the The joints represent the connections between the
response of masonry (Zucchini and Lourenco, 2009), microelements. The Mohr-Coulomb connection has
thus appropriate modeling of these joints is critical. been employed to signify the joints. The primary
“No separation” contact model was adopted in the parameters used for the interface element were
analysis given its ability to simulate the behavior of determined experimentally and are given in Table 3.
June 2021 | Volume 40 | Issue 1 | Page 30
Journal of Engineering and Applied Sciences
Table 3: Summary of material properties.
Topology Contact - Joint Brick - Element
Material 3D Interface 3D Non-Linear Cementitious2
Parameters Knn Ktt Ft c Φ E GF v ft fc
MN/m3 MN/m3 MPa MPa GPa N/m MPa MPa
Values 85000 56500 0.12 0.08 0.65 14.34 36.49 0.2 0.46 8.75

The same interface element was used for connecting Comparison of experimental vs numerical results
brick units with the adjacent RC frame. The backbone curve was drawn from the Hysteresis
curves (Figures 6 and 7) and compared to the results
Meshing, loading and solution parameters obtained from numerical modeling values. Figure 15
A 3D micro modeling of masonry with geometric shows the comparative results of the bare frame for
characteristics equal to those confirmed in the both the experimental testing as well as numerical
laboratory (Figure 14) was adopted using FE software modelling. The numerical results obtained show good
(Atena 3D) that had employed the material model relation with the experimental results.
adopted. The iterative method of Newton-Raphson
was employed to solve the numerical equations. 60
The loading was simulated by imposing a constant 50
Lateral Load (kN)
vertical load and a monotonic lateral displacement.
The assessment of monotonic and cyclic testing 40
consequences proposes that the loading protocol is not 30
significant if one is concerned about the load capacity of
the structure. Since the main purpose of the numerical 20
modeling was to perform the parametric analysis of 10 Experimental Results
the experimentally tested frames to understand the Numerical Modeling Results
0
behavior regarding its lateral load response. Thus, the
0 20 40 60 80 100 120
assumption was considered valid and only monotonic
Displacement (mm)
loading was applied while calibrating the model and
also in the parametric analysis. The displacements Figure 15: Results comparison of bare frame.
were imposed in small steps (displacement of 1 mm
per step), with linear variation and uniform growth. Figure 16 shows the comparative results obtained
for RC frame with infill wall from experimental
testing and numerical modelling. The initial stiffness
of numerical curve matches with experimental one
up to a load value of 40 kN however, after that the
numerical model shoes a relatively less stiff behavior.
Similarly, the maximum lateral load for numerical
model occurs at a displacement of 18 mm, while that
of experimental one occurs at 37 mm. The lateral load
resistance of numerical model starts decreasing after
20 mm but on the other hand, it increases up to 38
mm for the experimental model and starts reducing
after that. Also, the post peak (softening) path of
both the curves differ up to some extent, with the fact
that numerical model stops at a displacement of 88
mm and lateral load of 84 kN while the experimental
model reaches 100 mm displacement with lateral load
of 110 kN. It is seen that the proposed model can
detect the key aspects of the tested behavior of the
walls. The damage mechanism and load-deformation
Figure 14: Simplified micro modeling approach. diagrams are partially well replicated.
June 2021 | Volume 40 | Issue 1 | Page 31
Journal of Engineering and Applied Sciences
160 infilled frame. After the agreement of the results of
140 experimental and numerical models, a parametric
study was performed to check the effect of different
Lateral Load (kN)

120
100 parameters by keeping the boundary conditions and
80 frame dimensions similar to the originally tested
60 frame.
40
20 Experimental Results Parametric study
Numerical Modeling Results Numerical model of the parametric study
0
0 20 40 60 80 100 120 In parametric study, the properties used for steel and
Displacement (mm) concrete as well as the geometry of the assembly were
Figure 16: Results comparison of infilled frame. kept the same, since the main objective was to evaluate
the influence of masonry infill wall on the response of
Figure 17 shows the comparison of the damage RC frame structure. Therefore, the material properties
pattern of the infilled frame by experimentally testing of brick units were changed to see their effect on the
and numerical modelling. The damage pattern follows behavior of the structure.
the actual damage pattern of the system very much.
However, the crack formation in the panel indicates 400

the same damage pattern and predicts the same stress 350
Lateral Load (kN)
concentration point as obtained from lab tests as well 300

as the micro-modeling approach. 250

200

150

100

50
0 kN 78 kN 156 kN 234 kN 312 kN
0
4 6 8 10 12 14 16
Brick Compressive Strength (MPa)

Figure 18: Effect of gravity load and brick compressive


strength on the lateral load response of the RC infilled
frame.

Parameters under consideration


To study the effect of brick properties on the response
of infilled RC frame by varying the brick strength of
4, 6, 8, 10 and 12 MPa. All the five types of frames
were subjected to five different types of gravity load
Figure 17: Comparison of damage pattern of infill wall. levels i.e. 0 kN, 78 kN, 156 kN, 234 kN and 312 kN.
Thus, making a total of twenty-five (25) studies to be
The main purpose of this study was to determine performed and the lateral load capacity (i.e. maximum
the effect of infilled brick wall on the lateral load lateral load of the frame) was calculated. Results are
performance of the RC structure. As presented presented in Figure 18, where it is seen that increasing
previously in detail, the inclusion of infilled wall not the brick strength from 4 MPa to 12 MPa, increases
only stiffens the RC frame (about 171%) but it also the maximum capacity of the infilled frame from 115
increases its lateral load performance considerably (i.e., kN to 205 kN, in the absence of any gravity load.
175%). The models were then generated in FEA based Similarly, for the gravity load of 312 kN, the effect
software with the material properties and boundary of brick compressive strength is more obvious on the
conditions similar to the experimentally tested frames. response of the structure, thus increasing it from 160
The numerical model seemed to replicated the actual kN for 4 MPa brick strength to 360 kN for 12 MPa
tested frames with marginal but acceptable accuracy of brick compressive strength. Thus, it can be stated
of predicting the maximum lateral load as well as that for the same masonry properties, the maximum
the damage pattern of both bare frame as well as capacity of the infilled frame increases by increasing
June 2021 | Volume 40 | Issue 1 | Page 32
Journal of Engineering and Applied Sciences
the gravity load on the structure. Similarly, under the Novelty Statement
same gravity load level, the capacity of the infilled
frame enhances by enhancing the properties of the The key influence of this work is to study the lateral
masonry (in this case only brick unit properties are load behavior of RC frame elements with and with-
considered and mortar properties are kept constant). out masonry infill and to demonstrate the use of com-
puter simulations to predict masonry behavior (i.e., to
Conclusions and Recommendations conduct the parametric study with different mechan-
ical properties).
This paper presents both the experimental and
numerical study of the influence of Masonry infill on Author’s Contribution
the behavior of RC frames. In experimental work, the
quantitative comparison was made between both types Syed Azmat Ali Shah: Experimental testing and
of frames (with and without infill panel) based on initial analysis.
stiffness, ultimate strength and ultimate displacement. Asfandyar Ahmed: Numerical analysis
The same structures were then modelled numerically Khan Shahzada: Contributed to the design and
in an FEA software Atena, using a simplified micro implementation of the research.
modeling approach. The model predicts the non-linear Syed Muhammad Ali: Conceived and planned the
response of infilled frame by capturing all potential experiments
modes of failure. Appropriate material models and Akhtar Naeem Khan Main: Conceptual ideas and
biaxial damage criteria were chosen, and masonry proof outline
is supposed to be an anisotropic material. The shear Akhter Gul: Involved in planning and supervised the
and bending distortions, which play a significant role fabrication of the specimen.
in the overall behavior of the walls, have been taken
into account. The presented computing method can Conflict of interest
be used to predict the response of infilled frames with The authors have declared no conflict of interest.
diverse material and geometric properties. The peak
lateral load determined by numerical model gives References
90% and 98% of the maximum lateral load obtained
from experimental testing for bare frame and infilled Andreaus, U. and A. di Paolo. 1988a. A 3d
frame, respectively. finite element model for the analysis of
masonry structures. In: Proceedings of the
The precision of the method accepted is obtained by 8th international brick and block masonry
comparison of experimental and numerical results. conference, Dublin, 3P: 1405–1416.
It is presented that the precision of the prophesied Andreaus, U. and A. di Paolo. 1988b. 3d analysis of
results is satisfactory. After calibrating the numerical masonry columns with grouted reinforcement
model, a parametric analysis was led in which the bars. In: Proceedings of the 8th international
brick unit properties were changed, subjected to brick and block masonry conference, Dublin; 3:
different gravity load levels. It was demonstrated 1507-1518.
that for the same gravity load level the enhancement Andreaus, U. and G. Ceradini. 1992. Failure modes
of brick unit properties enhances the total capacity of solid brick masonry under in-plane loading.
of the RC frame with infill. Similarly, by increasing Masonry Int. 6(1): 4–8.
the gravity load on the structure with the same brick Andreaus, U. and L. Ippoliti. 1995a. A two story
unit properties, the lateral load capacity increases masonry wall under cyclic loading: A comparison
considerably. It should be noted that current study is between experimental and numerical results. In:
based on single bare frame and single infilled frame Proceedings of the 3rd international symposium
tests and hence the scope is very limited and cannot on computer methods in structural masonry,
be taken as a representative for all kind of infilled Lisbon, 1: 68–77.
structures. Andreaus, U. and L. Ippoliti. 1995b. Masonry
panel under in-plane loading: A comparison
between experimental and numerical results. In:
Proceedings of the 3rd international symposium
June 2021 | Volume 40 | Issue 1 | Page 33
Journal of Engineering and Applied Sciences
on computational methods and experimental Building Code of Pakistan, Seismic Provision,
measurements, Capri, Italy; 1: 603–610. SP-2007. Ministry of Housing and works,
Andreaus, U. and L. Ippoliti. 1995c. A two story Government of Islamic republic of Pakistan,
masonry wall under monotonic loading: 2007.
A comparison between experimental and Buyukozturk, O. and S.S. Sareef. 1985. Constitutive
numerical results. In: Proceedings of the 4th modeling of concrete in finite element analysis.
international conference on structural repair Comput. Struct., 21: 581–610. https://doi.
and maintenance of historical buildings, Creta, org/10.1016/0045-7949(85)90135-X
1: 319–326. Cela, J.J.L., 1998. Analysis of reinforced concrete
Andreaus, U., 1996. Failure criteria for masonry structures subjected to dynamic loads with a
panels under in-plane loading. J. Struct. Eng., visco plastic Drucker, Prager model. Appl. Math.
122(1): 37–46. https://doi.org/10.1061/ Model, 22: 495–515. https://doi.org/10.1016/
(ASCE)0733-9445(1996)122:1(37) S0307-904X(98)10050-1
Andreaus, U., G. Ceradini, M. Cerone and P. Cervenka, V., J. Cervenka and R.P. Atena. 2002.
D’Asdia. 1985a. Masonry columns under A tool for engineering analysis of fracture in
horizontal loads: A comparison between finite concrete, Sadhana 27(4): 485–492. https://doi.
element modeling and experimental results. org/10.1007/BF02706996
In: Proceedings of the 7th international brick Cervera, M., J. Oliver, O. Manzoli. 1996. A rate-
and block masonry conference, Melbourne, pp. dependent isotropic damage model for the seismic
469–478. analysis of concrete dams. Earthq. Eng. Struct.
Andreaus, U., M. Cerone, P. D’Asdia and F. Dyn., 25: 987–1010. https://doi.org/10.1002/
Lannozzi. 1985b. A finite element model for (SICI)1096-9845(199609)25:9<987::AID-
the analysis of masonry structures under cyclic EQE599>3.0.CO;2-X
actions. In: Proceedings of the 7th international Chang-Hai, Z., 2014. Experimental investigation of
brick and block masonry conference, Melbourne, in-plane seismic behavior of full-scale masonry
1: 479–488. infilled RC frames. Int. Works. Seismic Perf.
Asteris, P.G. and A.D. Tzamtzis. 2003a. Nonlinear Non-Struct. Elem. (SPONSE), August 29-31,
seismic response analysis of realistic gravity 2014.
dam-reservoir systems. Int. J. Nonlin. Sci. Num. Chen, A.C.T. and W.F. Chen. 1975. Constitutive
Simul., 4(4): 329–338. https://doi.org/10.1515/ relations for concrete. J. Eng. Mech. Div.
IJNSNS.2003.4.4.329 ASCE, 101: 465–481. https://doi.org/10.1061/
Asteris, P.G. and A.D. Tzamtzis. 2003b. On the JMCEA3.0002034
use of a regular yield surface for the analysis of Chen, E.S. and O. Buyukozturk. 1985. Constitutive
unreinforced masonry walls. Electron. J. Struct. model for concrete in cyclic compression.
Eng., 3: 23–42. J. Eng. Mech. Div. ASCE, 111: 797–815.
Balan, T.A., E. Spacone and M. Kown. 2000. A 3D htt ps://doi.org/10.1061/(ASCE)0733-
hypoplastic model for cyclic analysis of concrete 9399(1985)111:6(797)
structures. Eng. Struct., 23: 333–342. https:// Dube, J-F., G. Pijaudier-Cabot and C. La Borderie.
doi.org/10.1016/S0141-0296(00)00048-1 1996. Rate dependent damage model for
Balan, T.A., F.C. Filippou and E.P. Popov. 1997. concrete in dynamics. J. Eng. Mech. Div.
Constitutive model for 3D cyclic analysis of ASCE, 122: 359–380. https://doi.org/10.1061/
concrete structures. J. Eng. Mech. Div. ASCE, (ASCE)0733-9399(1996)122:10(939)
123: 143–153. https://doi.org/10.1061/ Fardis, M.N., B. Alibe and J.L. Tassoulas. 1983.
(ASCE)0733-9399(1997)123:2(143) Monotonic and cyclic constitutive law for
Barpi, F., 2004. Impact behaviour of concrete: A concrete. J. Eng. Mech. Div. ASCE, 109: 516–
computational approach. Eng. Fract. Mech., 536. https://doi.org/10.1061/(ASCE)0733-
71(2): 197–213. 9399(1983)109:2(516)
Bertero, V. and S. Brokken. 1983. Infills in seismic Faria, R., J. Olivera and M. Cevera. 1998. A strain-
resistant building. ASCE J. Struct. Eng., based plastic viscous-damage model for massive
109(6): 1337-1361. https://doi.org/10.1061/ concrete structures. Int. J. Solids Struct., 35:
(ASCE)0733-9445(1983)109:6(1337) 1533–1558. https://doi.org/10.1016/S0020-
June 2021 | Volume 40 | Issue 1 | Page 34
Journal of Engineering and Applied Sciences
7683(97)00119-4 10: 2305–2316. https://doi.org/10.5194/
Federal Emergency Management Agency (FEMA). nhess-10-2305-2010
2000. Prestandard and commentary for the Koutromanos, Stavridis A., Shing B.P., and Willam,
seismic rehabilitation of buildings. Report no. K. 2011. Numerical modeling of masonry-
FEMA 356, FEMA, Washington, DC; 2000. infilled RC frames subjected to seismic loads.
Federal Emergency Management Agency. 2007. Comp. Struct., 89(2011): 1026–1037. https://
FEMA 461. Interim protocols for determining doi.org/10.1016/j.compstruc.2011.01.006
seismic performance characteristics of structural Kyriakides, M.A. and S.L. Billington. 2008.
and non-structural components through Seismic retrofit of masonry-infilled non-
laboratory testing. Redwood City, CA. ductile reinforced concrete frames using
Fiore, A., F. Porco, G. Uva and M. Sangirardi. 2014. sprayable ductile fiber-reinforced cementitious
The influence of uncertainties of infill panels composites. Proc. 14th World Conf. Earthq.
relative to the seismic response of RC existing Eng., October 12-17, 2008, Beijing, China.
buildings. Structures Under Shock and Impact Liu, T.C.Y., 1971. Stress–strain response and
XIII. https://doi.org/10.2495/SUSI140411 fracture of concrete in biaxial compression.
Gomes, H.M. and A.M. Awruch. 2001. Some Research report no. 339. Dept. of Structural
aspects on three-dimensional numerical Engg. Cornell University,
modelling of reinforced concrete structures Liu, T.C.Y., A.H. Nilson and F.O. Slate. 1972.
using the finite element method. Adv. Eng. Biaxial stress–strain relations for concrete.
Softw., 32: 257–277. https://doi.org/10.1016/ ACI Struct. J., 98: 1025–1034. https://doi.
S0965-9978(00)00093-4 org/10.1061/JSDEAG.0003222
Grassl, P., K. Lundgren and K. Gylltoft. 2001. Lourenco, P., 2002. Computations on historic
Concrete in compression: A plasticity theory masonry structures. Prog. Struct. Eng. Mater.,
with a novel hardening law. Int. J. Solids Struct., 4(3): 301–319. https://doi.org/10.1002/pse.120
39: 5205–5223. https://doi.org/10.1016/ Mehrabi, A.B., P.B. Shing, M.B. Schuller and J.L.
S0020-7683(02)00408-0 Noland. 1994. Performance of masonry-infilled
Hashemi, A. and K.M. Mosalam. 2006. Shake- R/C frames under in-plane lateral loads. Report
table experiment on reinforced concrete CU/SR-94/6, Department of Civil.
structure containing masonry in fill wall. J. Eq. Negulescu, C., 2010. Evaluarea vulnerabilităţii
Eng. Struct. Dyn., 9(1): 73–83. fondului construit înainte de 1977 în Bucuresti,
Hatzigeorgiou, G., D. Beskos, D. Theodorakopoulos în funcţie de condiţiile locale de teren. Ph. D.
and M. Sfakianakis. 2001. A simple concrete Diss., Technical Univ. of Constr., Bucharest,
damage model for dynamic FEM applications. 2010.
Int. J. Comput. Eng. Sci., 2: 267–286. https:// Page, A., 1978. Finite element model for masonry.
doi.org/10.1142/S1465876301000325 J. Struct. Div., 104(8): 1267–1285. https://doi.
Kang, H.D., K. William, B. Shing and E. Spacone. org/10.1061/JSDEAG.0004969
2000. Failure analysis of RC columns using a Pauw, A., 1960. Static modulus of elasticity of
triaxial concrete model. Comput. Struct.,77: concrete as affected by density. ACI J., 57: 679–
423–440. https://doi.org/10.1016/S0045- 688. https://doi.org/10.14359/8040
7949(00)00006-7 Polyakov, S.V., 1960. On the interaction between
Kauffman, A. and A.M. Memari. 2014. Performance masonry infill walls and enclosing frame when
evaluation of different masonry infill walls with loaded in the plane of the wall. Earthq. Eng.,
structural fuse elements based on in-plane cyclic EERI, San Francisco, pp. 36-42.
load testing. Buildings, 4: 605-634. https://doi. Popovics, S., 1973. A numerical approach to
org/10.3390/buildings4040605 the complete stress-strain curve of concrete.
Korkmaz, S.Z., Kamanli M., Korkmaz H.H., Cement Conc. Res., 3: 816–825. https://doi.
Donduren M.S., and Cogurcu M.T. 2010. org/10.1016/0008-8846(73)90096-3
Experimental study on the behaviour of Pukl, R., J. Cervenka and V. Cervenka. 2001.
nonductile infilled RC frames strengthened Simulating a response of connections. Proc.
with external mesh reinforcement and plaster RILEM Symp. On Connections between Steel
composite. Nat. Hazard. Earth Syst. Sci., and Concrete, Stuttgart, Germany.
June 2021 | Volume 40 | Issue 1 | Page 35
Journal of Engineering and Applied Sciences
Rosenblueth, E. and R. Meli. 1986. The 1985 and analytical approaches. Eng. Struct.,
Mexico earthquake: Causes and effects in 33(3): 968–980. https://doi.org/10.1016/j.
Mexico City. Conc. Int., pp. 23-34. engstruct.2010.12.018
Shing B.P., Stavridis A., Koutromanos I., Willam Thabet, A, and D. Haldane. 2001. Three-
K., Blackard B., Kyriakides A.M., Billington dimensional numerical simulation of the
L.S., and Arnold S. 2009. Seismic performance behaviour of standard concrete test specimens
of non-ductile RC frames with brick infill. when subjected to impact loading. Comp.
Proceedings of ATC and SEI 2009 conference Struct., 79: 21–31. https://doi.org/10.1016/
on improving the seismic performance of S0045-7949(00)00109-7
existing buildings and other structures. https:// Yang, B-L., Y.F. Dafalias and L.R. Herrmann.
doi.org/10.1061/41084(364)102 1985. A bounding surface plasticity model for
Stavridis, A. and P.B. Shing. 2010. Finite element concrete. J. Eng. Mech. Div. ASCE, 111: 359–
modeling of nonlinear behavior of masonry- 380. https://doi.org/10.1061/(ASCE)0733-
infilled RC frames. J. Struct. Eng., 136(3): 285– 9399(1985)111:3(359)
296. https://doi.org/10.1061/(ASCE)ST.1943- Zucchini, A. and P.B. Lourenco. 2009. Validation
541X.116 of a micro-mechanical homogenization model:
Tasnimi, A.A. and A. Mohebkhah. 2011. Application to shear walls. Int. J. Solids Struct.,
Investigation on the behavior of brick-infilled 46(3–4): 871–886. https://doi.org/10.1016/j.
steel frames with openings, experimental ijsolstr.2008.09.034

June 2021 | Volume 40 | Issue 1 | Page 36

You might also like