Engineering Structures 140 (2017) 134–150

Contents lists available at ScienceDirect

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

Resistance of reinforced concrete frames with masonry infill walls

to in-plane gravity loading due to loss of a supporting column
A. Brodsky ⇑, D.Z. Yankelevsky
Faculty of Civil & Environmental Engineering and National Building Research Institute, Technion-Israel Institute of Technology, Haifa 32000, Israel

a r t i c l e i n f o a b s t r a c t

Article history: This paper focuses on typical infilled masonry reinforced concrete (RC) frame buildings, and their action
Received 29 May 2016 in the course of an extreme event. An extreme event (i.e. local impact, blast or earthquake) may severely
Revised 22 February 2017 damage a supporting frame column and lead to a full or partial progressive collapse of the frame (or
Accepted 23 February 2017
building). The increasing frame deflection at the point of the missing column support under the action
of gravity loading may be restrained due to the structural resistance of the masonry infill wall and its
composite action with the surrounding RC frame. In this study the composite action of masonry infilled
Keywords:
walls is examined through laboratory experiments of simplified specimens comprising of a masonry wall
Masonry
Reinforced concrete
surrounded by a RC frame. It aims at evaluating the contributions of infill masonry walls, in an attempt to
Frame examine the infill masonry wall added resistance to the bare frame and its possible contribution to pre-
Infill wall vent progressive collapse. Results of laboratory tests that have been conducted on half scale seven rein-
Progressive collapse forced concrete infilled frames without a supporting column, under monotonic vertical loading along that
Damage column axis, are presented. The results indicate that masonry infill walls considerably increase the frame
Structural testing resistance to a vertical load action, compared to the resistance of a bare frame (around 280% on average
Vertical load and up to 500%). Masonry block type and column shear connectors have a major effect on the mode of
Failure mechanisms
failure. Reinforcement details have a pronounced effect on the frame performance; the proposed
Column cllapse
improved reinforcement details increase the resistance by almost 100%.
Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction hardly been addressed and most of the attention has been given
to infilled masonry walls subjected to lateral loads, with relevance
Reinforced concrete frames with unreinforced masonry infill to earthquake loading. The few experimental studies aiming at
walls are commonly used worldwide in residential and public investigating the RC infilled frame behaviour were related to inte-
buildings. In recent years more attention is given to extreme loads rior column loss [1–3]. There are no experimental studies on the
on structures such as impact and blast, and to other resulting composite wall behaviour in the case of peripheral column loss,
modes of damage including progressive collapse. The latter may although the peripheral column is more likely to be damaged, in
result from a local severe damage of a supporting column of the case of car collision or a nearby explosion in proximity to the build-
RC frame at the ground story level. The local severe column failure ing façade.
is responsible for a support loss the effect of which may affect the
entire frame or structure or a major part of it and cause partial or 1.1. Masonry infill wall modelling in column loss analysis
full collapse. A resulting downward displacement of the damaged
column may distort a bare RC frame that is supported on that col- Usually the unreinforced masonry infill walls are considered by
umn thus producing bending that exceeds its capacity and cause the designer as non-structural and are not taken into account in
failure. In the case of an infilled frame, the infill masonry may the structural design. Several experimental and numerical analysis
interact with the RC frame, restrain its deformation, increase its studies aiming at evaluation of the RC building resistance to pro-
stiffness and capacity and help redistributing the loads to neigh- gressive collapse have been carried out and reported in the litera-
bouring spans thus inhibiting the possible collapse with an alterna- ture. These studies may be subdivided into two main groups that
tive stable load path. Despite its importance this problem has differ with regard to the infill wall model. In the first group the
infill wall is modelled with a continuum linear shell element [4].
This model cannot account for cracking and is suitable for the
⇑ Corresponding author. linear infill wall behaviour only. The second group uses strut

http://dx.doi.org/10.1016/j.engstruct.2017.02.061
0141-0296/Ó 2017 Elsevier Ltd. All rights reserved.
 A. Brodsky, D.Z. Yankelevsky / Engineering Structures 140 (2017) 134–150 135

elements to model the infill wall. Usually a single strut is used that 1.4. Masonry infill wall modelling based on lateral loading behaviour
is connected to the beam-column joints [4–6]; in several studies
multiple struts are used as well [7]. The single strut approach does These observations motivated development of analytical meth-
not allow evaluating the real bending moment and shear forces in ods to model the composite wall behaviour. A most common
the RC frame as they do not simulate the interaction between the model represents the infill wall by a compressed diagonal strut.
infill wall and the surrounding frame. The multiple struts with The properties of this equivalent strut have been widely investi-
off-diagonal struts introduce discrete contacts with the RC frame gated [19–21] and the Federal Emergency Management Agency
however these are simplified models with a pre-determined con- (FEMA) adopts this single equivalent strut approach as the recom-
tact with the frame that is not related to the variable continuous mended design model [22–24]. The strut approach has been
contact pressure distribution between the masonry infill and the extended to include several adjacent struts [18,25,26]; neverthe-
RC frame. In all these studies the infill wall properties are taken less, there are rather different recommendations for the equivalent
on the basis of earlier experimental, analytical and numerical stud- strut parameters the implementation of which yields different
ies of the frame-infill wall composite behaviour under the action of results. Comparison of these models predictions, including differ-
a lateral load rather than of a vertical load. This is mainly due the ent strut parameters, with experimental results shows a wide
lack of knowledge regarding the composite wall behaviour under scatter. This also indicates that the strut model oversimplifies the
the action of a vertical load that the present paper aims at investi- complex composite wall behaviour. Indeed, careful observations
gating. A brief overview at the case of a lateral force action is given of experimental results show that many times a combination of
in the following. several modes of damage is developed throughout the loading pro-
cess. The single equivalent strut approach cannot represent any of
the modes and of their combinations, and, in fact, may model only
1.2. Frames and infill masonry wall behaviour under lateral load the primary building’s stiffness to the action of the horizontal load-
ing, and has inherent deficiencies in attempting to represent the
The behaviour of RC infilled frame buildings under a strong entire wall behaviour until reaching failure.
earthquake is very much dependent on the composite action of Beyond the simplified analytical models there exist advanced
the frames and the infill masonry walls that provide considerably modelling attempts, mainly using Finite Element formulations,
larger shear resistance. Therefore, over the last decades extensive suggesting detailed description and modelling of the infill wall
efforts have been made to investigate its behaviour. Many of these by using micro-models [27], where modelling of every block and
studies focused on a typical single bay, single story wall in a mortar interface are required. The analytical and numerical models
building, with commonly locally used types of masonry blocks. support the experimental observations that the masonry infill
Extensive studies examined several major governing parameters walls have an undoubtedly significant effect on the structural
such as: the wall geometry, window opening in the wall, type of performance of composite walls. Elaboration and discussion of
the masonry blocks and their geometry, frame’s beam and column these models is beyond the scope of this paper.
stiffness, reinforcement details in the RC frame, construction
method of the wall, effect of vertical load, etc. [8–14]. Some of
1.5. Differences between the behaviour to lateral and vertical loading
the studies were extended to include the overall building
parameters such as the number of stories and number of bays
It should be noted that there are considerable differences
[10,15,16]. Buildings in Israel are built mainly with hollow con-
between the behaviour of the same typical wall to lateral and to
crete masonry blocks or with lightweight autoclave-cured aerated
vertical loading, among which one could mention the wall geome-
concrete (AAC) masonry blocks. They differ from other types of
try (aspect ratio), the relative position of the load line of action
bricks that are being used in many other countries and are the
with respect to the direction of the masonry infill wall mortar beds,
typical masonry infill that had been used in many of the above
the relative location of columns and beams with their different
mentioned experimental studies. Therefore, there was a need to
stiffnesses and reinforcement details, different location of wall
investigate the composite wall behaviour to lateral loads with RC
construction details (e.g. shear keys, cast stops), etc. Therefore
frames that were infilled with these Israeli typical masonry blocks
the use of masonry wall well-known models that are based on lat-
as well [17].
eral loading studies raises doubts with regard to their suitability in
the case of vertical load.
1.3. Masonry infill wall failure modes
1.6. Summary and gaps of knowledge
Review of accumulated experimental test results of typical
walls subjected to, mostly static or quasi-static, lateral loads, iden- The above review describes the lack of knowledge regarding the
tifies different modes of damage depending on combinations of the role of the masonry infill walls in the structure response to vertical
different parameters; the major identified damage and failure loading due to the loss of a supporting column. Attempt to simu-
mechanisms are [18]: corner crushing, shear sliding, diagonal com- late the evolution of a progressive collapse mechanism and its
pression, diagonal cracking and frame failure (Fig. 1). analysis are commonly based on knowledge and data borrowed

corner crushin g shear sliding diagonal compression diagonal cracking frame failure

Fig. 1. Typical failure modes of masonry infilled frames subjected to lateral loading.
136 A. Brodsky, D.Z. Yankelevsky / Engineering Structures 140 (2017) 134–150

from the case of the wall response to the action of a lateral load, masonry infilled frames. This is not only the most simple scheme
which may be fundamentally different. Therefore there is a need for an experimental study but it also represents a unit cell of the
of reliable conclusive experimental data, to understand the entire multi bay multi storey building wall. In that scheme the
masonry infill wall behaviour under the action of a vertical load, RC frame-masonry wall interaction is well represented, although
and its absence prevents the development of reliable analytical the restraining effects of adjacent walls from top, bottom and sides
and numerical models to simulate the complex nonlinear beha- of this cell are excluded. Similar considerations are valid for both
viour of the composite walls in the case of a supporting column lateral loading and vertical loading. It is worth noting that in a
failure. This lack of knowledge on the one hand and the major RC skeleton framed building masonry infill walls may (or may
importance of this issue to the safety and stability of buildings not) exist in almost every interior unit cell. In the case of a vertical
on the other hand present an urgent need to investigate this prob- load, the unit cell better represents the case of the peripheral
lem, and motivated this study. column (e.g. column A2 in Fig. 2) which is likely to be severely
The present investigation aims at studying the behaviour of a damaged in case of a nearby explosion in proximity to the building
composite RC frame with a typical masonry infill wall structure, façade. Nevertheless, investigation of the interaction between the
in the case of loss of a supporting column. A quasi static simplified masonry infill wall and the confining RC frame will enhance our
approach suggests investigation of a typical wall, the columns of understanding also in the general case of a supporting column fail-
which carry gravitational loading, where one of the column sup- ure effect on the composite wall.
ports is missing. At this stage, the dynamic aspects of the column The masonry blocks that were used are either hollow concrete
failure and the resulting frame response are not considered, and masonry blocks or lightweight autoclaved aerated concrete (AAC)
attention is given to the quasi-static loading of the new structural blocks. These blocks are very common in the Israeli building
system that is formed after loss of the supporting column. construction.
Although the composite wall behaviour under these conditions is The reinforced concrete frames are designed to represent typi-
a key element in the assessment of the building capacity to survive cal relatively old existing residential buildings of 3–5 stories
a loss of a supporting column and prevent a progressive collapse height, that have been built some 50–60 years ago, mostly before
event, there are no experimental studies that may illuminate the a seismic code was valid. These buildings are characterized by
composite wall behaviour in the case of peripheral column loss 20 cm sided columns as a minimum dimension. In a plane view
and direct to possible modelling of this problem. The results of of a typical frame it shows as a frame with narrow columns of
the first stage of this study are presented in the following. 20 cm wide or with wider columns (40 cm) that provide a larger
stiffness in the wall plane. In the scaled down models the column
smallest dimension, as well as the wall thickness, is 10 cm. The
2. Experimental program
blocks and mortar bed dimensions are also reduced by the scale
of 1:2 accordingly.
2.1. Preliminary considerations

It was decided that the specimen’s geometry and types of 2.2. Major design parameters and specimens characteristics
masonry blocks should be similar to those that had been used in
an earlier series of tests in which the specimens were subjected Seven specimens were prepared for testing with various combi-
to horizontal cyclic loading [17]. This consideration may enable nations of design parameters, as listed in Table 1. Two alternatives
future comparisons between similar walls that are subjected to dif- for each design parameter were considered and they are denoted
ferent load actions (lateral vs. vertical). Therefore scaled down by 0 and 1.
models (1:2) of single bay single storey infilled frames were used.
This scheme is common in numerous experimental studies on 2.2.1. Beam/column related stiffness
In stiff columns the cross-section is 200/100 mm (200 mm
width), simulating a full size 200/400 mm column. In the flexible
column the cross-section is 100/200 mm (100 mm width), simulat-
ing the same cross section area in a 90° rotated orientation. The
beam cross-section (100/200 mm) remained constant in all specimens.
The details of these parameters are given below.

2.2.2. Reinforcement details in the frame

The first five specimens that were tested indicated weaknesses
of the RC frame; therefore, prior to casting the last two specimens,
improved reinforcement detailing was implemented in them with
the following characteristics (Fig. 4): hoops density was increased
near the corners, in accordance with the common modern practice
of seismic design, in order to improve the RC frame shear capacity;
Fig. 2. (a) Floor plane; (b) section with removed peripheral column.
longer lap splices were provided in order to prevent formation of

Table 1
Design parameters.

Parameter Levels of design parameters

0 1
Beam/column related stiffness Relatively less stiff column (‘‘flexible column”) Relatively stiffer column (‘‘stiff column”)
Frame’s reinforcement details Regular reinforcement details Improved reinforcement details
Masonry blocks type Hollow concrete blocks Autoclaved aerated concrete (AAC) blocks
Infill construction method Non-integral infill Integral infill
 A. Brodsky, D.Z. Yankelevsky / Engineering Structures 140 (2017) 134–150 137

Fig. 3. Details of specimens with flexible columns and regular reinforcement.

Fig. 4. Details of specimens with stiff columns and improved reinforcement.

weak cross sections at the reinforcement overlapping zone; and 2.2.3. Masonry blocks type
the reinforcement anchor length was increased to prevent early Autoclave-cured aerated concrete blocks of dimensions
pull out damage. These changes added 22% in weight to the frame 250  150  100 mm and hollow concrete blocks of dimensions
reinforcement. 260  100  100 mm with approximately 40% void area were used
138 A. Brodsky, D.Z. Yankelevsky / Engineering Structures 140 (2017) 134–150

Fig. 5. Hollow concrete blocks: (a) standard 100 mm blocks; (b) blocks that are used in this research.

(Fig. 5). The blocks dimensions comply with the adopted half scale Table 3
modelling. Reinforcing steel properties.

Bar size Mean yield Ultimate Mean elongation

2.2.4. Infill construction method [mm] strength [MPa] strength [MPa] at rupture [%]

Two different construction methods were examined: 3 220 390 23.0

6 602 695 2.75
8 613 816 9.03
a. Integral infill wall: infill wall with unreinforced shear con-
nectors. In that method, the frame casting followed the
masonry wall construction thus forming unreinforced shear
keys between the columns and the masonry wall.
Standard [29] that was mandatory in the times when the existing
b. Non-integral infill wall: frame casting preceded the masonry
buildings that are simulated in this study had been constructed.
wall construction.
The actual average uniaxial compressive strength of these cubes
was in the range of 26.6–28.3 MPa (listed in Table 4), which is in
The seven specimens with the corresponding design parameters
agreement with the requirement of relevant Israeli standards for
are listed in Table 2. Specimens are identified by a four-digit code
20 MPa characteristic strength (24–28 MPa, depending on quality
denoting the levels of the four design parameters in the order (left
control conditions).
to right): Beam/column related stiffness; frame’s reinforcement
details; masonry blocks type; infill construction method.
2.4.2. Reinforcement
2.3. Specimens details The diameter of the reinforcing bars was defined on the basis of
relevant scaling laws, as half of the diameter size of the corre-
Details of selected panel configurations are presented in sponding reinforcing bars, usually utilized in full scale construc-
Figs. 3 and 4. Details for other configurations can be derived by tions of the tested type, in order to adjust them to the half-scale
inference. Reinforcement information is given in the figures and specimens. In addition to the 8 mm hot-rolled deformed bars also
the rebars are plotted in scale. The reinforcement details emulate, 6 mm cold drawn deformed bars were used, in spite of their lim-
at half-scale, typical details employed in old residential buildings ited elongation at rupture, since hot-rolled deformed bars of this
incorporating RC frameworks. The characteristic strength of con- diameter are not available in Israel. The reinforcing bars were
crete of at least 20 MPa [28] representing concrete requirements tested according to the relevant Israeli standards [30–32]. In addi-
at that time and represent the usually implemented concrete in tion to the longitudinal reinforcement bars, 3 mm diameter plain
such frameworks at the time. bars were used for the beam and column stirrups. It should be
noted that 3 mm diameter bars are not covered by the aforemen-
2.4. Material properties tioned standards, since they are not used in ordinary reinforced
concrete structures. In spite of this, the diameter of the stirrups
2.4.1. Concrete had to be adjusted for the half-scale specimens, according to rele-
Concrete compressive strength was determined from tests of vant scaling laws. Values of mean tensile strength and elongation
sets of three 100 mm cubes, as stated in the version of the Israeli are detailed in Table 3.

Table 2
Specimens configurations.

No. Code Beam/column related stiffness Frame’s reinforcement details Masonry blocks type Infill construction method
1 0011 Flexible column Regular Hollow concrete Integral
2 1010 Stiff column Regular Hollow concrete Non-integral
3 1011 Stiff column Regular Hollow concrete Integral
4 1000 Stiff column Regular AAC Non-integral
5 1001 Stiff column Regular AAC Integral
6 1100 Stiff column Improved AAC Non-integral
7 0110 Flexible column Improved Hollow concrete Non-integral
 A. Brodsky, D.Z. Yankelevsky / Engineering Structures 140 (2017) 134–150 139

Table 4
Test results.
 kN   kN 
No. Code Concrete compressive strength [MPa] Ki mm
Dcr [mm] Kcr mm
Pu [kN] DPu [mm] Df [mm] Pre [kN] Dre [mm]

1 0011 27.4 27.1 1.4 2.9 71.1 12.9 12.9 23.5 45.0
2 1010 28.3 34.5 0.6 4.8 77.1 19.7 35.7 39.8 81.0
3 1011 28.3 15.4 3.3 2.1 72.1 12.7 22.1 33.1 60.0
4 1000 27.0 6.8 4.5 0.8 44.5 37.7 57.3 26.0 128.0
5 1001 26.8 31.0 0.9 2.2 71.0 22.0 22.0 40.4 69.0
6 1100 28.2 27.4 0.67 5.6, 1.3a 88.51 43.16 61.62 62.48 105. 0
7 0110 26.6 139.9 0.3 5.9 122.0 24.7 42.2 59.7 59.0
a
Specimen 6 has two noticeable different stiffness K cr .

2.4.3. Masonry blocks type 2.5. Testing procedure

Standard 10 cm thickness hollow concrete blocks were used
(Fig. 5), rather than cutting the typical 20 cm blocks. Since the 2.5.1. Test setup
AAC blocks are solid blocks, there was no difficulty to produce The experimental setup and instrumentation used for monitor-
the small scale masonry blocks by simple mechanical sawing. ing during a typical test are shown in Fig. 6. To allow in-plane
The mean compressive strength of three individual AAC blocks of deformation only, the possible out of plane displacement was
dimensions 250  150  100 mm compressed across their restrained at the upper beam level. The left hand side column base
250  100 mm faces according to the relevant Israeli standard is hinge supported in both (Horizontal and vertical) directions;
[33] is 3.1 MPa. The mean compressive strength of three hollow these supports represent the lower floor column and the horizontal
concrete blocks with approximately 40% void area of dimensions elements at the next span. Thus, this column is referred to as the
260  100  100 mm compressed across their 260  100 mm faces ‘‘supporting column”. These supports do not allow the supporting
is 10.9 MPa. column to rotate. Thus the frame action is taken into considera-
tion; a negative moment develops at the beam-supporting column
2.4.4. Mortar connection taking into account the beam continuity. The top of the
The tested compressive strength of 6 individual 40 mm cubes of column is supported in the horizontal direction, and a vertical con-
cement mortar used for construction of the masonry infill walls is stant load (130 kN) is applied to represent the upper floor column
in the range of 13.4–15.0 MPa. The flexural strength of 3 individual axial force. This force is applied by a hydraulic actuator against two
40  40  160 mm blocks of cement mortar used for construction steel bars positioned along the sides of the column. A strain gauge
of the masonry infill walls is in the range of 3.5–3.8 MPa. was attached to each bar to monitor the constant load. There is no
vertical support under the right hand side column in order to rep-
2.4.5. Initial shear strength of a composite masonry wall resent the missing supporting column underneath. Monotonic ver-
Specimen of three blocks were used to obtain the initial shear tical load was applied by a controlled loading hydraulic actuator at
strength of a composite masonry wall according to the EN 1052- the top of this column (the ‘‘loaded column”). This actuator was
3 standard [34]. Shear failure was observed at the mortar interface. connected to an identical actuator and an aligned load cell that
The initial shear strength of a concrete blocks composite wall is in were placed inside a stiff steel frame that was place aside of the
the range of 0.85–0.96 MPa and 0.07–0.13 MPa for AAC blocks test setup (Fig. 6). Hydraulic pressure was increased gradually.
composite wall. These values indicate a low level of adherence Displacement transducers (LVDTs) measuring the displacements
between the mortar and the AAC blocks. at the loaded column base and at the supports. The change in

Fig. 6. Test setup and instrumentation.

140 A. Brodsky, D.Z. Yankelevsky / Engineering Structures 140 (2017) 134–150

length of the masonry infill wall diagonals were measured by Method c is less sensitive to cracking resulted displacement due
means of thin cables attached to LVDTs. This transducers setup to tension of the top beam and of the masonry and therefore yields
aims at providing the measurements that are needed to calculate a relatively low displacement estimate compared to the first two
the distortion and the net vertical deformations of the masonry methods, which yield rather similar results. The vertical displace-
wall panel. ment is therefore defined as the average of the first two methods.

2.5.2. Testing procedure

A monotonic vertical load was applied by the hydraulic actuator 3.3. Presentation of test results
at the top of the loaded column. Controlled hydraulic pressure was
provided by the pump. Throughout the experiment the LVDT’s and 3.3.1. General
load cell’s readings were recorded and the cracks were numbered In the following sub-sections, major results of the different tests
during their appearance and development and were marked on are presented. The results include the load-displacement curve in
the specimens’ whitewashed face. Usually, the experiment was general and the capacity in particular as well as description of
terminated at the maximal actuator opening (around 130 mm). the damage and its development. A general view of the damaged
walls at their final state is shown in Fig. 8. Different damage pat-
3. Test results terns are identified in the masonry walls and the surrounding RC
frame that will be discussed in more detail in the following.
3.1. General All load-displacement curves are characterized by a nonlinear
ascending branch with decreasing stiffness until the ultimate resis-
Test results are described by the force-displacement relation- tance is reached (Fig. 9); thereafter the displacement increases at a
ship i.e. the variation of the vertical applied load with the vertical relatively constant load until reaching a certain displacement, after
displacement of the loaded (unsupported) column and via inspec- which resistance drops with increasing displacements. The resis-
tion of the cracking and damage evolution of the tested wall. In the tance decreases to its residual value and the displacement then
following, this information will be presented and discussed. As an keeps growing at with that constant magnitude residual resistance.
introductory stage, attention is given below (Section 3.2) to the These displacements may be large and the test was terminated
evaluation of the vertical displacement that is a major parameter. when the actuator reached its maximum stroke.
The measured vertical displacement may be affected by rigid body The key characterizing parameters constructing the force-
displacement and/or rotation and these influences should be elim- displacement curve are: the variable stiffness of the ascending
inated to yield the net vertical displacement as a result of the com- branch, the ultimate resistance and the residual resistance and
posite wall deformation. the corresponding displacements. The values of these parameters,
Then, the load-displacement and damage descriptions of the for the tested specimens, are summarized in Table 4 and provide
different walls will be presented and evaluated (Section 3.3). an overall view of the effect of the combinations of these parame-
ters on the infill frame resistance.
3.2. Net vertical displacement calculation A glance at Table 4 shows that all the parameters vary strongly -
with the specimen properties, however further examination of the
The net vertical displacement can be calculated by three differ- results indicates some trends related to the combined effect of the
ent methods (Fig. 7): above variables. The ultimate resistance (Pu) of the tested speci-
mens varies within a wide range (44.5–122 kN). However, within
a. Based on the vertical LVDT measurement from which a rigid the given range, the resistance of most specimens varied between
body displacement component resulting from the rigid body 70 kN and 90 kN. Improved reinforcement details contribute con-
rotation is subtracted (Fig. 7a); siderably to the infill frame resistance. The displacement corre-
b. Based on the elongation measurement of the tension diago- sponding to the ultimate resistance (DPu) in specimens with
nal (Fig. 7b); hollow concrete blocks is significantly smaller than in specimens
c. From the shortening measurement of the compression diag- with AAC blocks. The residual resistance (Pre) varies in the range
onal (Fig. 7c). 23.5–62.5 kN, where specimens with improved reinforcement

(a) (b) (c)

LV11
///H///

LV
10 LV7
LV6
LV2

Fig. 7. Assessment of the net vertical deflection: (a) vertical LVDT; (b) diagonal elongation; (c) diagonal shortening.
 A. Brodsky, D.Z. Yankelevsky / Engineering Structures 140 (2017) 134–150 141

Fig. 8. Final state of damage of the different specimens.

- Initial stiffness

Pu - Displacement at first stiffness change

- Stiffness after first stiffness change
- Ultimate resistance
Kcr
- Displacement at ultimate resistance
Pre Pre - Displacement at failure
Ki
- Residual resistance
- Displacement at residual resistance

cr Pu f re

Fig. 9. A typical force-displacement curve.

details (specimens 6 and 7) shows almost identical values (about This same bare frame has not been tested in the present series
60 kN) and are the highest values among all the specimens. The of tests to the action of a vertical load. Therefore 2D Analysis is car-
displacement at residual resistance (Dre) is considerably larger in ried out on the laterally loaded bare frame, to validate the compu-
non-integral frames. The effect of each design parameter will be tational method through comparison with this test and the same
examined in detail in Section 4.2. bare frame that is being used in the present study, will be evalu-
ated to the action of a vertical load. The analysis is carried out with
3.3.2. The bare frame ATENA 2D [35], a non-linear Finite Element (FE) software. Four-
In order to evaluate the contribution of the infill wall, the beha- node Sbeta elements were used to simulate the concrete. The
viour of the concrete bare frame subjected to monotonic vertical concrete material properties are listed in Table 5. The concrete
loading should be examined first. In an earlier series of tests, in non-linear behaviour model is shown in Fig. 10 and represents
which similar infill frames were examined to the action of a lateral the major features of concrete in compression and tension
load, a bare frame was tested as well [17] under quasi-static cyclic (nonlinear stress-strain, and unloading–reloading), it accounts for
reversal loading. compressive strength and shear stiffness reduction after cracking,
142 A. Brodsky, D.Z. Yankelevsky / Engineering Structures 140 (2017) 134–150

Table 5 plastic hinge at top right hand side column-beam joint, similarly
Concrete material properties in the FE model. to the experiment. Following this validation stage, the same bare
Ec [MPa] Elastic modulus 28,160 frame was analysed to the action of a vertical load (Fig. 11).
l [–] Poisson ratio 0.200 The analysis shows that after ascending at a considerably smal-
ft [MPa] Tensile strength 2.063 ler stiffness, the specimen reaches an ultimate load of 24.1 kN at a
fc [MPa] Cube compressive strength 25.20
Gf [MN/m] Specific fracture energy 5.16E 05
considerably larger (about twice as large) displacement and exhi-
bits a pronounced ductility and energy absorbing capacity. A sim-
ilar analysis (not shown in the figure) on the stiffer columns case
yields an even higher capacity (30.6 kN). Comparisons of the bare
frame resistance performance with the overall composite wall
resistance demonstrate the contribution of the infill masonry wall
ft'ef to the resistance and stiffness of the composite wall action.
Ec
3.3.3. Results of the tested walls
d c g t 0 3.3.3.1. Specimen 1 (0011). At about 38 kN a horizontal crack (crack
din
loa 1 – Fig. 12) is developed in the masonry between the upper block
un
course and the one below; it significantly reduces the wall stiffness
(Fig. 13). The infilled frame rotates around the supported column
fc'ef and the lower beam connection (point O, Fig. 14a). The horizontal
forces in the upper beam (HB, Fig. 14a) and the infill wall (hw)
oppose the moment created by the vertical load. Infill vertical crack
Fig. 10. Uniaxial stress-strain law for concrete.
(crack 2) that is passing through the upper beam determines the
specimen capacity (71.1 kN). As a consequence the vertical resis-
hardening and softening in compression and biaxial strength fail- tance drops significantly while increasing the vertical displace-
ure criterion. A detailed description of the constitutive model can ment. Crack 2 detaches the connection keys from the masonry
be found in the ATENA Program Documentation [35]. A multi- wall thus redistributing the load carrying system; from that stage
linear model was used to represent the steel reinforcement and on, a truss like system is formed in which the vertical load is
behaviour according to experimental test results that are shown carried by a horizontal upper beam in tension and diagonal com-
in Section 2.4. A Newton-Raphson method was used with tangent pression (Fig. 14b). Diagonal shear crack (crack 3 – Fig. 12) devel-
stiffness updated after each iteration. ops at the left hand side of the lower beam due to the increasing
Fig. 11 shows the experimental response curves’ envelopes of local compression of the diagonal strut. The wall resistance
lateral cyclic loading until its maximal value in both positive and decreases with increasing vertical displacement and stabilized at
negative loading directions (±x). To facilitate the comparison, the around 25 kN while the vertical displacement continues to grow
absolute values of experimental negative direction loading results gradually. The test was terminated after 90 mm displacement
were taken. In addition this figure includes the ATENA monotonic due to a technical problem.
loading analysis of the same frame that is subjected to lateral load- View of the specimen with the detailed cracking pattern at the
ing. A very good match is obtained between the predicted nonlin- end of the test is shown above (Fig. 8a).
ear load-displacement behaviour of the flexible column’s bare
frame with the lateral load action. The initial stiffness, the maxi- 3.3.3.2. Specimen 2 (1010). At about 27 kN vertical separation
mum resistance and the corresponding displacement are well pre- begins to develop between the infill wall and the supported
dicted. The maximum resistance was obtained after formation of column (crack 1 – Fig. 15) resulting in stiffness degradation

Fig. 11. Comparison of load-displacement curves of bare frames with flexible columns.
 A. Brodsky, D.Z. Yankelevsky / Engineering Structures 140 (2017) 134–150 143

2
1 2

450
3

2 4

Fig. 12. Damage pattern specimen 0011.

Fig. 15. Damage pattern specimen 1010.

3, 4 (78kN)

2 (30kN)

1 (27kN)

Fig. 13. Load-displacement curve for specimen 0011. Fig. 16. Load-displacement curve for specimen 1010.

(Fig. 16). Tensile cracks (crack 2) are developed along the upper 3.3.3.3. Specimen 3 (1011). At a low load level of 16 kN a horizontal
beam at 30 kN. The masonry at the upper right hand side of the separation occurs between the upper block course and the one
wall starts crushing under compression (crack 3 – Fig. 15) at a load below course (crack 1 – Fig. 17), accompanied by tensile cracks
level of 78 kN. The crushing causes splitting of the hollow concrete that appear at the top of the upper beam near the supported
blocks and exposure of the beam’s and column’s reinforcement due column. As the vertical load further increases the upper beam
to the concrete splitting. Afterwards, tensile cracks (crack 4 – tensile cracks spread especially near the zone of the upper rein-
Fig. 15) are developed on the outer side of the loaded column until forcement lap slices at 1/3 and 2/3 beam length. A vertical infill
failure of this column longitudinal reinforcement occurs and crack (crack 2 – Fig. 17) appears when the vertical load reaches
causes a significant reduction in the specimen resistance. 52 kN and significantly reduces the stiffness (Fig. 18). The load
A view of the specimen after the test is shown above (Fig. 8b). keeps growing at a considerably reduced stiffness accompanied

Fig. 14. Schematic load transfer before (a) and after (b) the vertical crackopening.
144 A. Brodsky, D.Z. Yankelevsky / Engineering Structures 140 (2017) 134–150

350
500
2

1 1
4 1

2 4 2
1
6

600
5

3
3

Fig. 17. Damage pattern specimen 1011.

Fig. 19. Damage pattern specimen 1000.

4 (44.5kN)
3 (40kN)

2 (43kN)
5 (43kN)
1 (32kN)
6 (36kN)

Fig. 18. Load-displacement curve for specimen 1011.

Fig. 20. Load-displacement curve for specimen 1000.

by the vertical crack opening until reaching a critical stage at
which the hoop in the upper beam fails next to crack 2 at
72.1 kN. As a consequence, the vertical resistance drops signifi-
cantly as the vertical displacement increases. Diagonal shear cracks
develop at the left hand side of the lower beam (crack 3 – Fig. 17)
due to the increasing local compression of the diagonal strut indi-
cating this beam receives the vertical load through the infill wall
close to the supported column. Similarly to specimen 1, the upper
beam is subjected to tension, where the horizontal load at the
cracked section is carried just by the longitudinal reinforcement
crossing the crack, and the infill wall carry diagonal compression. 3
At a displacement of almost 60 mm, the descending branch turns
into a stabilized residual resistance of about 33 kN for further
1
increasing vertical displacement. The test was terminated after
75 mm displacement due to a technical problem. 2
A view of the specimen after the test is shown above (Fig. 8c).

3.3.3.4. Specimen 4 (1000). An extensive amount of diagonal and

horizontal cracks is developed at the infill wall above 30 kN (cracks
1, 2, 4, 5 – Fig. 19) and causes stiffness degradation (Fig. 20). At a
load level of 40 kN a shear crack (crack 3 – Fig. 19) is developed
at the lower beam near the supported column. In addition tensile Fig. 21. Damage pattern specimen 1001.
cracks are developed on the outer side of the loaded column, at
about its mid height, starting from 37 kN. Infill cracks 4 and 5
(Fig. 19) dictate the maximum specimen resistance at 44.5 kN.
The loaded column’s curvature could be observed during the crack 6 – Fig. 19. The experiment was terminated when the
loading progress. The descending branch of the vertical load actuator maximum stroke was reached and the resistance dropped
continues while the vertical displacement grows until failure of to 27 kN.
the loaded column longitudinal reinforcement is occurred in A view of the specimen after the test is shown above (Fig. 8d).
 A. Brodsky, D.Z. Yankelevsky / Engineering Structures 140 (2017) 134–150 145

2 (71kN)
370
3 (55kN)
0 1 2

7 5 6
800
1 (35kN)

Fig. 22. Load-displacement curve for specimen 1001.

4

1 3 Fig. 25. Damage pattern specimen 0110.

3-4 (122kN)
4 5-7 (106kN)

4 5
0’, 1’ (86kN)
2
1 (27kN)
600

Fig. 26. Load-displacement curve for specimen 0110.

Fig. 23. Damage pattern specimen 1100.

3.3.3.6. Specimen 6 (1100). At about 20 kN tensile cracks are
developed at the top of the upper beam near the supported
column and at the upper beam reinforcement overlapping zone
4 (88.5kN) 5 (88kN) (num. 1 – Fig. 23). At 54 kN infill diagonal cracks (crack 2 –
Fig. 23) appear and cause stiffness decrease (Fig. 24). At a load level
3 (67kN) of 67 kN shear crack (crack 3 – Fig. 23) develop at the upper beam
near the supported column. The resistance keeps growing to a level
of 88.5 kN where maximum resistance is reached while crushing of
2 (54kN) the AAC blocks located near the loaded column and the upper
beam corner is observed, as well as blocks crushing close to the
loaded column at the column mid height level (crack 4 – Fig. 23).
1 (20kN) A rather constant resistance at the ultimate strength characterizes
this wall resistance for the following 40 mm. Then tensile cracks
are developed on the outer side of the supported column at a level
of about 600 mm. Failure of the longitudinal reinforcement
(crack 5 – Fig. 23) in the loaded column is observed near the
Fig. 24. Load-displacement curve for specimen 1100.
crushed zone of the masonry. This causes a sharp drop in the wall
resistance from 88 kN to about 70 kN.
A view of the specimen after the test is shown above (Fig. 8f).
3.3.3.5. Specimen 5 (1001). Diagonal and horizontal cracks (crack 1
– Fig. 21) are developed in the infill wall at about 35 kN and cause 3.3.3.7. Specimen 7 (0110). First tensile cracks appear at 10 kN at
stiffness degradations (Fig. 22) similarly to specimen 4. The maxi- the top of the upper beam near the supported column (crack 0 –
mal vertical load (71 kN) is reached at a vertical displacement of Fig. 25). A horizontal separation crack develops between the infill
20 mm as a consequence of the development of another infill diag- wall and the upper beam at 27 kN (crack 1). At a load level of
onal crack (crack 2 – Fig. 21). At a displacement of 60 mm a major 86 kN crack 0 and the separation crack 1 considerably expand
damage occurred when the longitudinal reinforcement of the (Fig. 26). The vertical limit resistance (122 kN, see Fig. 26) is deter-
loaded column failed (crack 3 – Fig. 21) after which the specimen mined by a diagonal crack in the upper beam close to the loaded
resistance abruptly decreased. Then a constant resistance of about column (crack 2).
40 kN accompanied the increasing displacement. This diagonal crack cross meets the ends of the additional lower
A view of the specimen after the test is shown above (Fig. 8e). reinforcement (Fig. 27b). This crack continues through the infill
146 A. Brodsky, D.Z. Yankelevsky / Engineering Structures 140 (2017) 134–150

masonry wall (crack 3) and splits it into two parts (Fig. 27c). A the load, the infill masonry vertical crack and the beam shear crack
diagonal crack (crack 4) develops at the lower beam close to the expand until failure occurs in the upper beam hoops. At this stage
supported column, thus determining the location of the compres- the wall resistance drops considerably.
sive forces line of action. Afterwards the infill wall applies perpen-
dicular forces to the loaded column axis. It causes tensile cracks to 4.1.3. Mode 3
develop on the column outer side (Fig. 27d). After the development This mode of damage is observed in specimen 7, and is typical
of the infill diagonal crack, the diagonal compression at the upper to strong infill blocks (hollow concrete) with weak columns. It is
right hand side of the wall is transmitted through a rather narrow characterized by infill corner crushing and shear failure near the
zone thus causing crushing of the hollow concrete blocks in this top of the loaded column.
area (crack 5 – Fig. 25). As a result shear failure is developed close
to the upper part of the loaded column (crack 6 – Fig. 27e) and 4.2. Comparative analysis
additional infill crack (num. 7 – Fig. 25). This damage causes a
sharp decrease of the resistance. This section starts with the infill masonry wall influence on the
A view of the specimen after the test is shown above (Fig. 8g). frame vertical resistance. Then, the effect of a specified parameter
is analysed while the other three design parameters are kept con-
4. Analysis and interpretation of the results stant. The effects on the capacity and the failure mode are being
assessed. Table 2 summarizes the major parameters of the differ-
4.1. Failure mechanisms ent specimens and may assist in the following analysis.

The above results indicate three major failure modes (Fig. 28). 4.2.1. Infill wall effect
The presence of the infill wall effect on the failure mode and the
4.1.1. Mode 1 vertical resistance. The most significant contribution of the infill
An extensive amount of diagonal and horizontal cracks are wall was observed in specimen 7. Fig. 29 shows the contribution
developed at the infill wall during the loading process. In addition of the infill wall to the vertical resistance. The RC frame with
tensile cracks are developed on the outer side of the loaded column flexible columns and hollow concrete, non-integral infill wall resis-
until failure of the column longitudinal reinforcement occurs. This tance was five times higher than the resistance of same bare frame
mode of damage is observed in specimens 2, 4, 5 and 6, and is typ- calculated in Section 3.3.2.
ical to weak infill blocks (the AAC).
4.2.2. Beam/column related stiffness
4.1.2. Mode 2
This mode of damage is observed in specimens 1 and 3, and is Specimens 1 (0011) and 3 (1011)
typical to strong concrete masonry blocks applied in an integral Constant parameters: regular reinforcement, integral infill
construction method. The usage of strong hollow concrete blocks wall and hollow concrete blocks. Capacity: There is much similar-
reduces the diagonal cracking in the wall. A single vertical crack ity between these two specimens in terms of their ultimate
in the infill is first observed. It is immediately followed by a shear resistance. The failure has not occurred at the columns, thus the
crack in the same location at the upper RC beam. With increasing column differences are not affecting the failure. The columns dif-

Fig. 27. Damage patterns specimen 7: (a) deformed free body diagram, (b) upper beam’s shear crack near the loaded column, (c) Infill diagonal crack, (d) tensile cracks on the
loaded column outer side, and (e) loaded column shear failure.
 A. Brodsky, D.Z. Yankelevsky / Engineering Structures 140 (2017) 134–150 147

Fig. 28. Main failure mechanisms.

Fig. 29. Load-displacement curve for specimen 7 and bare frame.

ference in stiffness also has a minor effect due to the protruding

shear keys into the masonry wall, thus adding pronounced stiffness Fig. 30. Stiffness ratio comparison specimens 0011 and 1011.
to the flexible column and preventing its global deformation and
restraining its bending. It is expected that difference between stif-
fer and stronger columns and flexible and weaker columns would Capacity: improved reinforcement details have a pronounced
be pronounced in walls without shear keys, where wall curvature effect on the frame resistance (by almost 100%, see Fig. 31).
is not restrained and column damage may develop, depending on Failure mode: failure of both specimens occurs due to failure of
its stiffness and strength parameters. Therefore, there is small dif- the longitudinal column’s reinforcement. Improved reinforcement
ference between the maximum vertical loads (Fig. 30). details reduce the deformation and significantly increases the ulti-
Failure mode: These specimens have a similar failure pattern mate vertical load. The stiffness degradation occurs due to infill
characterized by a wall vertical crack that continues to the upper wall cracks development. These cracks develop with increasing
beam. Hoops failure in the cracked section governing the maxi- deformation. Therefore the stiffness degradation in these two spec-
mum vertical load. The difference in the descending branch beha- imens occurs almost at the same deformations (7, 19, 60 and
viour between the two specimens obtained due the number of 96 mm) at different vertical load as may be observed by the piece-
hoops in the upper beam cracked section: two hoops in specimen wise linear envelope curves plotted in Fig. 31. Increasing the rein-
3 and only single hoop in specimen 1. In the flexible columns spec- forcement anchoring length improved the lap splice detailing
imen (specimen 1) the wall vertical crack occurred within half considerably by preventing tensile crack development at the upper
block distance from the shear connectors ends, whereas in the stiff beam and loaded column corner (specimens 2, 4 and 5). Instead,
columns specimen the crack path is close to the shear connectors diagonal shear crack appear in the upper beam at the cross section
ends. This determines the crack location in the upper beam and where the additional lower reinforcement ends (Fig. 27b).
its relationship with the crossing rebars and the overlapping
details. These may affect the tensile stresses in the rebars, their
4.2.4. Masonry blocks type
possible slip and govern the rate of the beam crack opening. The
latter may be responsible for force-displacement behaviour
beyond the peak load. Specimens 2 (1010) and 4 (1000)
Constant parameters: Non integral infill wall, stiff columns and
regular reinforcement. Capacity: the capacity of the hollow con-
4.2.3. Frame’s reinforcement details crete block specimen is 73% larger than the capacity of the AAC
blocks infill wall (see Fig. 32).Fig. 33
Specimens 4 (1000) and 6 (1100) Failure mode: Diagonal cracks widened considerably in the
Constant parameters: Non integral infill wall with AAC blocks AAC blocks infill wall developed already at a low level of the verti-
and stiff columns. cal load. However, almost no cracks develop in the hollow concrete
148 A. Brodsky, D.Z. Yankelevsky / Engineering Structures 140 (2017) 134–150

Envelope Capacity: Although there are significant differences in the beha-

Improved reinforcement viour of both specimens, the maximum vertical load is quite simi-
Specimen 6
Regular reinforcement lar (see Fig. 32).
Specimen 4 Failure mode: Many diagonal and horizontal cracks are devel-
oped at the infill wall during the loading process of the AAC blocks
specimen (specimen 5). In addition tensile cracks are developed on
the outer side of the loaded column until failure of the column’s
longitudinal reinforcement is reached. This specimen is character-
ized by ductile behaviour when every infill wall crack causes
changes in the stiffness. The hollow concrete specimen fails due
to a single vertical crack in the masonry wall that passing to the
upper beam near the end of the shear connectors. In contrast, this
specimen is characterized by brittle behaviour when the vertical
resistance drops significantly after the hoop in the upper beam
Fig. 31. Frame’s reinforcement details comparison specimens 1000 and 1100. fails, next to the infill vertical crack. Thus, the type of masonry
block has a major effect on the mode of failure. The effect on the
blocks infill wall (specimen 2). An exception is specimen 7 where capacity depends on the interaction with other parameters. For
diagonal cracks have also evolved due to the high load magnitude specimens with shear connectors, the effect appears to be small,
that exceeded 120 kN; this is due to improved structural details of due to the anchoring effect of the shear keys in the masonry blocks,
the RC frame. In relatively high loads (77 kN) corner crash is devel- that restrain the column bending.
oped in the hollow concrete blocks infill wall. In both specimens
extensive tensile cracks develop in the loaded column during the
loading until failure of the column’s longitudinal reinforcement 4.2.5. Infill construction method
occurs. Therefore in this case the column properties have a great
influence on the capacity. Specimens 2 (1010) and 3 (1011)
Constant parameters: Hollow concrete blocks, stiff columns
Specimens 3 (1011) and 5 (1001) and regular reinforcement. Capacity: These two specimens show
similar resistance.
Constant parameters: Integral infill wall, stiff column and reg- Failure mode: The failure location is different; in the integral
ular reinforcement. specimen (with shear connectors) the failure occurs at the upper
beam’s reinforcement after formation of a vertical crack in the

Fig. 32. Masonry blocks type comparison specimens 1000 vs. 1010 and 1011 vs. 1001.

Fig. 33. Infill construction method comparison specimens 1011 vs. 1010 and1000 vs. 1001.
 A. Brodsky, D.Z. Yankelevsky / Engineering Structures 140 (2017) 134–150 149

wall. Vertical separation develops between the infill wall and the is particularly important to investigate the behaviour of this sys-
supported column in the non-integral specimen. This specimen tem. Such an investigation will enhance our understanding of the
fails at the column longitudinal reinforcement. complex interaction between the masonry infill wall and the sur-
rounding RC frame and contribute to improved design criteria
Specimens 4 (1000) and 5 (1001) leading to enhanced performance and robustness of such buildings
in the case of an extreme event.
Constant parameters: AAC blocks, stiff column and regular
reinforcement. Acknowledgement
Capacity: The shear keys anchorages in the masonry improve
the column performance and limit its column curvature. Thereby This paper is dedicated to the memory of our late colleague and
the column’s shear connectors (integral specimen) increase the friend, Stephan Schwarz, who had participated in the first part of
frame resistance by 57%. this project with much dedication and talent.
Failure mode: These two specimens have a similar failure The help of our engineering and technical staff is highly
mechanism, extensive infill diagonal and loaded column tensile appreciated.
cracking until failure of the longitudinal column reinforcement
occurs. References

[1] Shan S, Li S, Xu S, Xie L. Experimental study on the progressive collapse

5. Conclusion performance of RC frames with infill walls. Eng Struct 2016;111:80–92. http://
dx.doi.org/10.1016/j.engstruct.2015.12.010.
[2] Stinger SM, Orton SL. Experimental evaluation of disproportionate collapse
This paper presents results of an experimental investigation of resistance in reinforced concrete frames. ACI Struct J 2013;110:521–9. http://
masonry infilled RC frame response to loss of a supporting column. dx.doi.org/10.14359/51685609.
[3] Li S, Shan S, Zhai C, Xie L. Experimental and numerical study on progressive
Seven specimens were tested with various combinations of four collapse process of RC frames with full-height infill walls. Eng Fail Anal
design parameters: Beam/column related stiffness, frame’s rein- 2016;59:57–68. http://dx.doi.org/10.1016/j.engfailanal.2015.11.020.
forcement details, masonry blocks type and infill construction [4] Sasani M. Response of a reinforced concrete infilled-frame structure to
removal of two adjacent columns. Eng Struct 2008;30:2478–91. http://dx.
method. The results enhance the understanding regarding the wall doi.org/10.1016/j.engstruct.2008.01.019.
behaviour, the damage evolution and its ultimate resistance. The [5] Akbarpoor S. Effect of brick infill panel on the seismic safety of reinforced
infill wall can play a major role in maintaining the structural sys- concrete frames under progressive collapse; 2014
[6] Tsai M, Huang T. Effect of interior brick-infill partitions on the progressive
tem’s integrity and reducing the likelihood of a progressive col-
collapse potential of a RC building: linear static analysis results. Eng Technol
lapse and therefore its contribution should be incorporated in the 2009.
structural model. It was found that masonry infill walls increase [7] Farazman S, Izzuddin BA, Cormie D. Influence of unreinforced masonry infill
the frame resistance to vertical load up to 500% with an average panels on the robustness of multistory buildings. J Perform Constr Facil
2013;27:673–82. http://dx.doi.org/10.1061/(ASCE)CF.1943-5509.0000392.
of around 280%. The loss of a supporting column causes significant [8] Fiorato AE, Sozen MA, Gamble WL. An investigation of the interaction of
increase of the frame deflection that is restrained by the shear reinforced concrete frames with masonry filler walls. Illinois Univ Urbana Dept
resistance of the masonry infill wall, thus developing interaction of Civil Engineering; 1970.
[9] Mehrabi BA, Shing PB, Schuller PM, Noland LJ. Experimental evaluation of
forces between the infill wall and the surrounding frame. These masonry-infilled RC frames. J Struct Eng 1996.
interaction forces have an important role in the overall infilled [10] Bertero V, Brokken S. Infills in seismic resistant building. J Struct Eng
frame resistance. It was observed that the frame failure dictates 1983;109:1337–61. http://dx.doi.org/10.1061/(ASCE)0733-9445(1983).
[11] Mosalam KM, White RN, Gergely P. Static response of infilled frames using
the maximal vertical resistance. Therefore frame reinforcement quasi-static experimentation. J Struct Eng 1997;123(11):1462–9. http://dx.doi.
details have a pronounced effect on the frame performance. The org/10.1061/(ASCE)0733-9445(1997).
suggested improved reinforcement details (Section 2.2.2) increase [12] Zovkic J, Sigmund V, Guljas I. Cyclic testing of a single bay reinforced concrete
frames with various types of masonry infill. Earthq Eng Struct Dyn
the resistance by almost 100%. Beam column stiffness ratio has a 2013;42:1131–49. http://dx.doi.org/10.1002/eqe.2263.
major effect particularly in bare RC frames (about 130%). However, [13] Liu Y, Manesh P. Concrete masonry infilled steel frames subjected to combined
in case of infilled frames with strong blocks and shear key, the in-plane lateral and axial loading – an experimental study. Eng Struct
2013;52:331–9. http://dx.doi.org/10.1016/j.engstruct.2013.02.038.
beam column stiffness ratio effect diminishes because the failure
[14] Jiang H, Liu X, Mao J. Full-scale experimental study on masonry infilled RC
does not include the columns cross sections, but is rather passing moment-resisting frames under cyclic loads. Eng Struct 2015;91:70–84.
through the upper beam. http://dx.doi.org/10.1016/j.engstruct.2015.02.008.
The infill wall acts as an intermediate adjusting element, which [15] Fiorato AE, Sozen MA, Gamble WL. An investigation of the interaction of
reinforced concrete frames with masonry filler walls Rep no UILU-
interacts with the inner RC boundaries; its boundary deformations ENG. Urbana-Champaign (IL, USA): Dept Civ Eng Univ Illinois; 1970. p. 70–100.
and stresses causes damage and stiffness changes in the masonry [16] Henderson RC, Fricke KE, Jones WD, Beavers JE, Bennett RM. Summary of a
infill wall and affect the loading on the surrounding frame that is large- and small-scale unreinforced masonry infill test program. J Struct Eng
2003;129(12):1667–75. http://dx.doi.org/10.1061/(ASCE)0733-9445(2003).
responsible for its deformation and damage. The resulting overall [17] Schwarz S, Hanaor A, Yankelevsky DZ. Experimental response of reinforced
damage affects the overall stiffness and resistance. This composite concrete frames with AAC masonry infill walls to in-plane cyclic loading.
interaction results in a significant redistribution of internal loads, Structures 2015;3:306–19. http://dx.doi.org/10.1016/j.istruc.2015.06.005.
[18] El-Dakhakhni WW, Elgaaly M, Hamid AA. Three-strut model for concrete
where the effect of the masonry wall geometry and mechanical masonry-infilled steel frames. J Struct Eng 2003;129(2):177–85. http://dx.doi.
properties are of great importance. The masonry block type and org/10.1061/(ASCE)0733-9445(2003).
column shear keys have a major effect on the failure mechanism [19] Kadysiewski S, Mosalam KM. Modeling of unreinforced masonry infill wall
considering in-plane and out-of-plane interaction. Seattle: Washington Univ.,
mode. The results show three main groups of failure modes Dept. of Civil and Environmental Engineering; 2009.
(Fig. 28): Mode 1 (Section 4.1.1) is typical to weak infill blocks. [20] Fiore A, Netti A, Monaco P. The influence of masonry infill on the seismic
Mode 2 (Section 4.1.2) is typical to strong concrete masonry blocks behaviour of RC frame buildings. Eng Struct 2012;44:133–45. http://dx.doi.
org/10.1016/j.engstruct.2012.05.023.
applied in an integral construction method, where the end of the
[21] Perera R. Performance evaluation of masonry-infilled RC frames under cyclic
shear keys is a weakness section. Mode 3 (Section 4.1.3) is typical loading based on damage mechanics. Eng Struct 2005;27:1278–88. 10.1016/j.
to strong infill blocks with weak columns. These observed mecha- engstruct.2005.03.012.
nism modes are clearly different from the known mechanism [22] FEMA 307. FEMA307 – evaluation of earthquake damaged concrete and
masonry wall buildings. Tech Resour 1998.
modes that had been identified in the case of lateral loading [23] FEMA 306. FEMA 306 – evaluation of earthquake damaged concrete and
(Fig. 1), as well as from the bare frame failure mode. Therefore it masonry wall buildings, council; Appl Technol (ATC-43 Project); 1998.
150 A. Brodsky, D.Z. Yankelevsky / Engineering Structures 140 (2017) 134–150

[24] FEMA 356. FEMA 356 – Prestandard and commentary for the seismic [30] IS 4466 – Part 2. IS 4466 – Part 2: Steel for reinforcement of concrete: plain
rehabilitation of buildings. Rep FEMA-356. Washington, DC: Federal bars. Israeli Standards Institute; 2013.
Emergency Management Agency; 2000. [31] IS 4466 – Part 4. IS 4466 – Part 4: Steel for the reinforcement of concrete: Cold
[25] Jeon JS, Park JH, DesRoches R. Seismic fragility of lightly reinforced concrete drawn steel meshes. Israeli Standards Institute; 1999.
frames with masonry infill. Earthq Eng Struct Dyn 2015;41:1549–68. http:// [32] IS 4466 – Part 3. IS 4466 – Part 3: Steel for reinforcement of concrete: ribbed
dx.doi.org/10.1002/eqe. bars. Israeli Standards Institute; 2013.
[26] Crisafulli FJ, Carr AJ. Proposed macro-model for the analysis of infilled frame [33] IS 268. IS 268: Autoclaved aerated-concrete blocks. Israeli Standards Institute;
structures. Bull New Zeal Soc Earthq Eng 2007;40:69–77. 2000.
[27] Asteris PG, Cotsovos DM, Chrysostomou CZ, Mohebkhah A, Al-Chaar GK. [34] BS EN 1052-3. BS EN 1052-3:2002: Methods of test for masonry – Part 3:
Mathematical micromodeling of infilled frames: State of the art. Eng Struct determination of initial shear strength. Br Stand Inst; 2002.
2013;56:1905–21. http://dx.doi.org/10.1016/j.engstruct.2013.08.010. [35] Cervenka V, Jendele L, Cervenka J. ATENA program documentation Part 1
[28] IS 118. IS 118: Concrete: specifications, performance and production. Israeli theory; 2013.
Standards Institute; 1986.
[29] IS 118. IS 118: Concrete: specifications, performance and production. Israeli
Standards Institute; 2008.