Soil Dynamics and Earthquake Engineering 161 (2022) 107404

Contents lists available at ScienceDirect

Soil Dynamics and Earthquake Engineering

journal homepage: www.elsevier.com/locate/soildyn

Shake-table study on the effect of masonry infill on the seismic response of

reinforced concrete frames
Goran Baloevic *, Jure Radnic, Nikola Grgic, Ivan Grubisic
Faculty of Civil Engineering, Architecture and Geodesy, University of Split, Matice Hrvatske 15, Split, 21000, Croatia

A R T I C L E I N F O A B S T R A C T

Keywords: This paper presents results of an experimental shake-table study that examines the effects of position and stiffness
Masonry-infilled frame of masonry infill in two-story single-bay concrete frame on its behavior under an earthquake. Six small-scale
Infill distribution effect frame types were tested: bare frame (without infill) and five frames with different infill configuration. Herein,
Shake-table study
an identical reinforced concrete structure (columns, beams) was adopted for all frames. Besides self-weight load,
frames were additionally loaded with the dead load at the floor levels. Afterwards, all frames were exposed to
horizontal acceleration of an artificial accelerogram, with successive increase of peak ground acceleration (PGA)
up to 0.8g. Characteristic displacements and accelerations of the frame, concrete and steel strains in charac­
teristic points, as well as crack states in frame and infill were measured. The effect of the infill type on the
obtained results was discussed. Finally, main conclusions of this research are given.

1. Introduction partially infilled frames with flexible floors and especially with flexible
ground story, stress-strain states in beams and columns of frames under
Concrete frames are very rational structural elements. They occupy an earthquake can be unfavorable compared to bare frames (Fotos et al.
minor volume and enable simple changes in size of the interior building [5], Dolsek and Fajfar [6], Quayyum et al. [7]). Thereby, it is necessary
spaces. In fact, such structures are very favorable from both engineering to include the infill in the structural model in seismic design of such
and architectural point of view. It is the reason for their often applica­ structures, which increases structural resistance and generates greater
tion in the practice. seismic (inertial) forces in the structure at the same time.
Frames are often infilled due to architectural reasons (Mehrabi et al. Infill inside a frame is usually slightly deformable and very sensitive
[1], Caliò and Pantò [2]). The infill can be made from different mate­ to its displacements (Fiore et al. [8]). Weaker frames with such an infill
rials, with different stiffness in relation to the stiffness of a frame. Ma­ can experience severe damages/cracks under stronger earthquakes.
sonry infill made from clay or lightweight concrete units is probably Namely, besides the analysis of global bearing capacity and safety of a
most widely represented. masonry-infilled frame under earthquake, a separate analysis of possible
Masonry-infilled frame is a complex structure composed of different damages in the infill and frame should be performed.
materials and of different bearing systems. An infill is usually made after As previously stated, the infill has a significant effect on the seismic
construction of frames. It can be distributed differently inside them, response of masonry-infilled frames. So far, many numerical and
even partially due to openings. Adequate engineering calculations, and experimental researches on the behavior of infilled concrete frames
yet nonlinear numerical analyses of such structures are still much under an earthquake have been performed. Some numerical and
desired. It is especially desirable to reliably determine the actual experimental researches considering the effect of infill on the behavior
behavior and safety of such structures in seismic conditions (Asteris and bearing capacity of infilled frames are briefly described below.
et al. [3], Cavaleri and Di Trapani [4]). Hence, further researches in this Although many numerical models for analysis of masonry-infilled
area are welcomed. frames have been developed, an adequate and reliable model for simu­
An infill increases stiffness and reduces displacements of a bare lation of these structures presents challenge to most researchers. Most of
frame, and usually increases its bearing capacity. However, in case of the models were based on the nonlinear finite element method (Mallick

* Corresponding author.
E-mail addresses: goran.baloevic@gradst.hr (G. Baloevic), jure.radnic@gradst.hr (J. Radnic), nikola.grgic@gradst.hr (N. Grgic), ivan.grubisic@gradst.hr
(I. Grubisic).

https://doi.org/10.1016/j.soildyn.2022.107404
Received 13 April 2022; Received in revised form 14 June 2022; Accepted 14 June 2022
Available online 1 July 2022
0267-7261/© 2022 Elsevier Ltd. All rights reserved.
G. Baloevic et al. Soil Dynamics and Earthquake Engineering 161 (2022) 107404

and Severn [9], Dhanasekar and Page [10], Saneinejad and Hobbs [11], a bare frame, an infilled frame with flexible ground floor, an infilled
Mehrabi and Shing [12]). Recent numerical models have been devel­ frame with flexible upper floor, an infilled frame with larger opening on
oped that comprise interface elements to simulate frame-infill interac­ both floors, an infilled frame with smaller opening on both floors, and a
tion (Mohammad et al. [13], Koutromanos et al. [14], Baloevic et al. fully infilled frame without openings. Characteristic displacements and
[15]), method of contact points developed by Asteris [16], combined accelerations of the frame, concrete and steel strains in characteristic
smeared and discrete crack model (Stavridis and Shing [17]), points, as well as crack states in frame and infill were recorded during
displacement-based fibre elements (Mucedero et al. [18]), a r-adaptive the tests. The discussion was focused on the effect of the adopted infill
finite element approach (Fabbrocino et al. [19]), tensegrity model types on the test results, followed by main conclusions of the conducted
(Fraternali et al. [20]) and discrete element method (Mohebkhak et al. research. The objective of the research was to contribute to further
[21], Pantò et al. [22]). knowledge on the effect of infill on the actual seismic behavior of
A small number of experimental shake-table tests have been per­ masonry-infilled frames, as well as to generate an experimental database
formed to study the behavior of masonry-infilled frames under earth­ that could be used in valorization of numerical models for seismic
quake loading. Fardis et al. [23] analyzed the bidirectional response of a analysis of the considered structures.
two-story RC frame structure with two adjacent sides infilled through
shaking table tests. Hashemi and Mosalam [24] studied the effect of 2. Experimental test setup
unreinforced masonry wall on the structural behavior of infilled rein­
forced concrete frame based on shake-table tests on a 3/4 scale model. Six types of 1/4 scaled, two-story, single-bay masonry-infilled con­
Koutromanos et al. [25] performed a study that investigated the effec­ crete frames were tested (see Fig. 1): a bare frame (BF), a frame with
tiveness of retrofitting unreinforced masonry infill walls with composite flexible ground floor and with infill on the upper floor (IF1), a frame
materials to enhance the seismic performance of infilled nonductile RC with flexible upper floor and infill on the ground floor (IF2), partially
frames. The shake-table tests were conducted on a 2/3 scale, three-story, perforated masonry-infilled frames with larger opening (IF3) and with
two-bay, masonry infilled RC frame that had one bottom-story wall smaller opening (IF4) on both floors, and a fully masonry-infilled frame
retrofitted with engineered cementitious composites. Magliulo et al. (IF5). Herein, all concrete frames were identical in geometry and ma­
[26] and Petrone et al. [27] performed an experimental testing on a steel terial properties, while position and stiffness of the masonry infill was
test frame with internal partitions to investigate the seismic behavior of varied.
nonstructural components. Palios et al. [28] tested a single bay,
single-story steel frame infilled with off-the-shelve solid clay bricks 2.1. Geometry and materials
reinforced with steel straps along the bed joints, subjected to in-plane
quasi-static cyclic loading and to out-of-plane shake table testing. Basic geometric data of the specimens is shown in Fig. 2. Frames
Pavese et al. [29] investigated seismic response and peculiar damage were 1500 mm high, 1660 mm wide and 80 mm thick. The specimens
pattern by means of shake table tests performed on a 1/2 scaled were fabricated with self-compacting concrete having maximal grain
three-story infilled prototype, with particular attention given to the role size of 8 mm. Mix composition of the concrete involved 320 kg of cement
of beam-column joints and frame-panel interaction. Zhang et al. [30] CEM I 42.5 R, 150 kg of stone powder, 1240 kg of fine aggregate and
considered six 1/3 scale, single-story single-bay RC frames with different 415 kg of coarse aggregate. Water to cement ratio was 0.6, with addition
infill configurations and flexible connection details through shake-table of a superplasticizer for water reduction and increased workability.
tests. Yuen et al. [31] performed shake-table tests on five 1/3 scale Tested compressive strength, tensile strength and modulus of elasticity
infilled reinforced concrete frame specimens with different combina­ of the concrete are given in Table 1. Stress-strain relationship of the
tions of slit panels, isolation gaps between the infills and columns, and concrete in compression is given in Fig. 3a. Longitudinal reinforcement
anchorage of the infills. Singhal and Rai [32] investigated the effect of placed in the beams and columns was Ø8mm bars of B500B class, while
openings on load-carrying capacity of masonry-infilled frames with and transversal reinforcement in the beams and columns was constructed
without openings, subjected to a sequence of slow cyclic in-plane drifts from a steel wire with diameter of 2 mm. Stress-strain relationship of the
and shake table–generated out-of-plane ground motions. Xue and Xu steel bars is given in Fig. 3b.
[33] tested two 1/2 scaled model specimens of a traditional Brick elements were 80 × 160 × 80 mm in size, made from cellular
column-and-tie timber structures on shaking table, one without infilled autoclaved concrete with density of 450 kg/m3. Masonry work was
wallboard and the other with infilled wallboard. Gont et al. [34] per­ performed by using a thin-layer prefabricated mortar for cellular con­
formed shaking table tests on two three-story, two-bay, 1/4-scale rein­ crete brickwork. Material properties of the cellular concrete and mortar
forced concrete frame structures, with variants of an infilled RC frame are given in Table 1. Columns of the frames were constrained at the base
without infill walls in the first story (pilotis frame) and a bare RC frame. through the metal plate welded on the longitudinal columns’ rein­
The shake-table tests of the masonry-infill systems are recent and forcement and on the upper surface of the shake-table.
quite rare. For most experiments, a relatively smaller number of speci­ Besides self-weight, all frames were loaded with additional vertical
mens was considered. Herein, tests on specimens in real and reduced permanent load of 3.0 kN/m at each floor level (to represent a floor
scale were conducted. Investigations on structures in real scale are most load). Herein, concrete blocks were used to ensure additional load (see
valid, but also the most expensive and sometimes impossible, mostly due Fig. 4).
to the limited dimensions and performance of available shake-tables.
The preparation of specimens in the reduced scale enables an 2.2. Basic free oscillation periods of the tested frames
increased number of experimental testing of these systems on the shake-
table. With this approach, it is possible to investigate the effect of in­ Before applying the adopted dynamic excitations on the considered
dividual parameters on their behavior under earthquake. frames, their basic periods of free oscillations were experimentally
In respect to distribution of infill walls along floor plan and height of evaluated (see Table 2). It can be noticed that the periods are low due to
the structure, as well as to its stiffness (size and position of openings in the adopted small-scale model. The greatest difference between them is
the infill, material properties, etc.), actual behavior and bearing capacity small (just 20%) due to low stiffness of the brick elements used, as well
of infilled concrete frames under an earthquake are still insufficiently as due to relatively greater dimensions of the adopted beams and
investigated and therefore subjected to further research. columns.
This paper presents the results of a shake-table study on the behavior
of ¼ scale, two-story, single-bay masonry-infilled reinforced concrete
frames under an earthquake. Six types of the infilled frames were tested:

2
G. Baloevic et al. Soil Dynamics and Earthquake Engineering 161 (2022) 107404

Fig. 1. Schematic illustration of the tested masonry-infilled concrete frames.

Fig. 2. Basic geometry data of the tested frames.

material properties and under corresponding dynamic excitation.

Table 1
Starting point in selecting the relevant seismic excitation is that
Mechanical properties of the used materials.
tested small-scale infilled frames exposed to this excitation experience
Variable Unit Material higher nonlinearities same as real frames under stronger earthquakes,
concrete steel steel brick mortar without actual need to link the selected excitation to a real earthquake.
bar wire In accordance with the previous statement, it is believed that the relative
modulus of GPa 36.1 205.0 120.0 1.02 – relationship between obtained results on small-scale models under the
elasticity adopted excitation and relative relationship between obtained results on
compressive MPa 49.9 653.0 – 2.18 10.0 prototype structure under corresponding excitation would be close to
strength
each other. It is also believed that the size of the adopted models would
tensile strength MPa 6.9 653.0 240.0 0.2 2.5
not have a great effect on the conclusions of the performed study.
It is well known that for a real structure, as well as for its small-scale
2.3. Applied seismic excitation model, none single expected accelerogram of an earthquake can be
relevant for the most unfavorable state of all structural elements and
Selecting an appropriate seismic excitation in the experimental their cross-sections. In that sense, design of real structures requires an
analysis of small-scale models is very important. It usually has a sig­ application of several relevant earthquakes expected for current location
nificant effect on the obtained conclusions, especially when they are and structural type. Character of the study is that it does not aspire to
brought firmly in terms of bearing capacity and safety of a full-scale such an approach, but to give global results regarding the effect of the
structure. relative relation of the considered frame stiffnesses, i.e. the considered
In the concerned study, it is thought that the adopted small-scale cases of infilling the concrete frames.
model can precisely simulate the behavior of some specific full-scale Based on the previous stated, and in particular to reduce the scope of
structure, since relative effects of different parameters have been the study costs, only one seismic excitation was adopted which is
investigated on the small-scale model (various cases for stiffness of a considered to be relevant in terms of the presented conclusions.
masonry-infilled frame). These effects are very close to the relative ef­ The considered frames were exposed to horizontal component of an
fects which would be obtained on the full-scale structure with equal acceleration time history, created from elastic design spectra according

3
G. Baloevic et al. Soil Dynamics and Earthquake Engineering 161 (2022) 107404

Fig. 3. Stress-strain relationship for the concrete and reinforcement.

Fig. 4. The specimens IF1 (left) and IF4 (right) prepared for shake-table testing.

Table 2
Basic period (T1) of free oscilaltion for the tested specimens before execution of
dynamic tests.
Variable Specimen

BF IF1 IF2 IF3 IF4 IF5

Period T1 [s] 0.119 0.120 0.074 0.106 0.105 0.103

to Eurocode 8 [35], for earthquake type 1 and soil type A. The adopted
elastic design spectra is shown in Fig. 5a and horizontal component of
the generated accelerogram with PGA = 0.1 g in Fig. 5b.
Spectral values of the adopted accelerogram in Fig. 5b are shown in
Fig. 6. By comparing the diagram of spectral values in Fig. 6 with basic
periods of free oscillation for the considered specimens in Table 2, it can
be expected that the adopted accelerogram is relevant for the analysis of
the relative effect of the infill stiffness on the behavior of the infilled
frames under an earthquake. The adopted accelerogram in Fig. 5b was
sequentially applied with successive increase of PGA = n × 0.1g; n = 1,
…,8.

2.4. Measured quantities

Mechanical quantities showed in Fig. 7 were measured for each

applied dynamic excitation. Namely, the measurements included hori­
zontal displacements (u1, u2) and horizontal accelerations (a1, a2) at
each floor level (Fig. 7a), strains in the longitudinal steel reinforcement
of the columns (ϵS) at points A, B, C (see Fig. 7b), strains in concrete at Fig. 5. The adopted acceleration time history.
exterior side of the columns (ϵC) at points D, E, F, G (see Fig. 7c), as well
as crack states in the concrete frames and masonry.

4
G. Baloevic et al. Soil Dynamics and Earthquake Engineering 161 (2022) 107404

Fig. 6. Spectral values of the adopted accelerogram.

Fig. 7. Scheme of measuring points with measured quantities.

2.5. Equipment and instrumentation inductive displacement transducers HBM type WA/100-L. The acceler­
ations were measured by a glued KISTLER type 8044 accelerometer with
A shake-table in Laboratory for seismic testing at Faculty of civil measuring range of ± 5g and with automatic compensation of the
engineering, architecture and geodesy, University of Split was used for gravitational acceleration. The steel strains were measured by glued
the test. The shake-table has a layout size of 4 × 4 m, with uniaxial strain gauges HBM type K-LY11-5/120, while the concrete strains were
horizontal component of motion. A high-speed data acquisition system measured by glued strain gauges HBM type K-LY11-60/120. The sam­
HBM Quantum X with 16 channels was used to acquire digital data pling rate of 200 Hz was used on the data acquisition system during the
during shake-table tests. The displacements were measured by means of shake-table tests. Structural deformations and crack patterns in the

Fig. 8. Horizontal displacement of the top of the upper floor (u2).

5
G. Baloevic et al. Soil Dynamics and Earthquake Engineering 161 (2022) 107404

masonry and concrete were monitored and recorded by high quality significant difference. Accelerations are also increased with increase of
video camera. PGA, but not proportionally. The greatest acceleration a2 for PGA = 0.8
g was for the specimen BF in amount of approx. 19 ms− 2 (about 2.37
3. Test results and discussion times greater than PGA).
Vertical reinforcement strain at the bottom of the column at ground
Only most important measured results are graphically illustrated and floor at point B is shown in Fig. 11. It can be noticed that greatest strains
briefly discussed. The diagrams in Figs. 8–14 show results in time occurred for the specimen IF1, and afterwards for the specimen BF.
domain for all tested frames under PGA = 0.4g and 0.8g. The results can Strains in the other frames are significantly lower. At the end of the
eventually be used for validation of numerical models for dynamic excitation with PGA = 0.8 g, irreversible strains remained in the spec­
analysis of infilled frames. In valorization of these results, it should be imens IF1 and BF.
noted that several specimens experienced irreversible displacements Fig. 12 shows vertical reinforcement strains at the bottom of the
and strains under previously applied excitations. Evolution of natural column on the upper floor, at point A. The specimen BF had greatest
frequencies before and after shaking table excitations are given in strains, while IF1 and IF2 had slightly lower strains. Stronger infilled
Fig. 15. Some diagrams with characteristic displacement-strain re­ frames IF3, IF4 and IF5 had significantly lower strains, where IF3 had
lationships are given in Figs. 16–18. Maximal values for some measured greatest among them. Under excitation with PGA = 0.8 g, irreversible
quantities are given in Figs. 19–24, while final crack states in the tested strains remained in the specimens BF, IF1 and IF2, as well as smaller
specimens are shown in Fig. 25. irreversible strains for IF3.
Concrete strains at the bottom of the column at the ground floor at
point F (see Fig. 13) are dominantly compressive, and greatest for the
3.1. Time history results
specimens BF and IF1. All compressive strains in the concrete are elastic.
They are small for the other frames. Small irreversible tensile strains
Horizontal displacement at the top of the upper floor (u2) is pre­
occurred for the specimens IF1 and IF3.
sented in Fig. 8. On the first sight it can be noticed that displacements are
Concrete strains at the bottom of the column at upper floor at point G
increased with increase of PGA, but not proportional. It can also be
are presented in Fig. 14. All of them are relatively small. Maximal
noticed that displacement for the weak frames BF, IF1, IF2 are signifi­
compressive and especially tensile strains were for the specimen IF2,
cantly greater than of the strong frames IF3, IF4, IF5. The greatest dis­
which is the result of its great stiffness at the ground story and small
placements were for the specimen BF, as well as for the specimen IF1
stiffness at the upper story.
which were slightly lower. The specimen IF5 had smallest displace­
ments. In accordance with the excitation in Fig. 5b, the greatest dis­
placements occurred approximately from 5th to 25th sec of its duration. 3.2. Natural frequencies and hysteretic response
Horizontal displacements at the top of the ground floor (u1), pre­
sented in Fig. 9, are similar to those in Fig. 8. The difference is that u1 is At the end of last applied excitation with PGA = 0.8 g (final test),
greatest for the specimen IF1 and slightly lower for the specimen BF. basic periods of free oscillations for all considered frames were experi­
Displacement u1 for the specimen IF2 is significantly lower and almost mentally evaluated, which is presented in Fig. 15 together with their
equal to that of the specimen IF5, which has the smallest displacement. initial periods of elastic structure. It is obvious that the final period was
Fig. 10 shows horizontal relative acceleration at the top of the upper significantly increased for the specimen BF. It is a consequence of sig­
floor (a2). Maximal accelerations for all tested frames show no nificant damages that occurred in the frame, which resulted in reduction

Fig. 9. Horizontal displacement of the top of the bottom floor (u1).

6
G. Baloevic et al. Soil Dynamics and Earthquake Engineering 161 (2022) 107404

Fig. 10. Horizontal relative acceleration of the top of the upper floor (a2).

Fig. 11. Vertical reinforcement strain at the bottom of the column at the ground story (point B).

of its stiffness. Since the basic period of a structure is approximately the specimens IF1 and IF2 was 66.6% and 64.7% greater than their
proportional to the square root of its stiffness, it follows that the final initial stiffness, respectively. Stronger frames IF3, IF4 and IF5 showed
stiffness for the BF amounts approximately 31.2% of its initial (elastic) very small change in their initial stiffness at the end of test.
stiffness. Final period for the specimen IF1 was 22.5% greater than its The final period of the frames with stronger infill, i.e. of the fully
initial period, and 24.3% greater for the specimen IF2. Final stiffness for infilled frame (IF5) and of the frames with perforated infill (IF3, IF4),

7
G. Baloevic et al. Soil Dynamics and Earthquake Engineering 161 (2022) 107404

Fig. 12. Reinforcement strain at the bottom of the column at the upper floor (point A).

Fig. 13. Concrete strain at the bottom of the column on the bottom story (point F).

remained practically the same at the end of dynamic tests. This is a displacements and reinforcement strains are greatest for the specimens
consequence of quite small reduction of initial stiffness of these frames, BF and IF1, with expressed nonlinear relationship.
which is in accordance with the observed smaller damages (cracks) in Relationship between displacement u2 and reinforcement strain εS at
them. point A for PGA = 0.8 g is presented in Fig. 17. The relationship is
Relationship between displacement u2 and reinforcement strain εS at similar to that in Fig. 16, where the hysteresis behavior is most expressed
point B for PGA = 0.8 g is presented in Fig. 16. It is obvious that for the specimen IF1.

8
G. Baloevic et al. Soil Dynamics and Earthquake Engineering 161 (2022) 107404

Fig. 14. Concrete strain at the bottom of the column on the upper story (point G).

This also shows unfavorable state of stresses in the columns of the

ground story of the frame IF1 than of the frame BF. Peak upper story
drift ratio for the specimen IF2 under PGA = 0.8 g was approx. 1/280.
Displacements and inter-story drifts for the stronger frames IF3, IF4 and
IF5 are smaller.
Peak tensile strains in vertical reinforcement at the bottom of col­
umns at points C, A during action of successive excitations are given in
Figs. 21 and 22. In general, reinforcement strains are greater by
increasing the PGA, however their increase is nonlinear. Peak strains at
ground floor at point C for PGA = 0.8 g were approx. 3.0‰ and 2.4‰ for
the specimens IF1 and BF, respectively. Since maximal elastic steel strain
is approximately 2.4‰ (see Fig. 3b), it can be assumed that tensile
reinforcement in the column of the specimen IF1 experienced nonlinear
strains. Peak reinforcement strains in the other tested specimens at point
Fig. 15. Graphical presentation of basic periods for the specimens before and C were elastic and significantly lower. Peak strains at upper floor at
after the tests. point A for PGA = 0.8 g were measured as approx. 1.4‰ and 1.2‰ for
the specimens BF and IF2, respectively. It can be noticed that steel
Relationship between displacement u2 and concrete strain εC at point strains were greater for IF2 than for BF for excitations with PGA <0.6 g.
D for PGA = 0.8 g is presented in Fig. 18. Peak strain for IF1 was approx. 1‰. Stronger frames IF3, IF4 and IF5 had
lower values of peak steel strains.
Peak compressive concrete strains at the bottom of columns at points
3.3. Peak values and crack patterns
D and E are given in Figs. 23 and 24, respectively. The specimen IF1
under PGA = 0.8 g had the greatest concrete strain at point D of approx.
Peak horizontal displacements u1, u2 of the frame during action of
1.33‰. The specimen BF had slightly lower strain of about 1.13‰.
successive excitations are given in Figs. 19 and 20. In general, the dis­
According to Fig. 3a, the strains reached non-linear behavior of the
placements u1 and u2 are disproportionally greater by increasing the
concrete. Peak concrete strains in the columns for other tested frames
PGA. Peak displacement at the top of the upper floor (u2) for PGA = 0.8 g
were relatively small. At point E, the specimen IF2 under PGA = 0.8 g
was approx. 16.6 mm for the specimen BF, while peak displacement at
had the largest peak concrete strain of approx. 1.16‰. For the same
the top of the ground floor (u1) was approx. 11.5 mm for the specimen
PGA, the specimen BF had concrete strain at point E of approx. 0.38‰, i.
IF1. Displacements for the remaining tested frames were significantly
e. about 3 times lower. This confirms the cognition that the presence of
lower due to their higher stiffness. For the specimen BF under PGA = 0.8
the infill at upper story can result in unfavorable significant stresses in
g, peak drift ratio of its upper story was approx. 1/90, while peak drift
its columns compared to the bare frame.
ratio of the bottom story was approx. 1/77. Greater drifts, and especially
Schematic illustration with photographs of the crack states at the end
bottom story drift, are the result of the greater applied base accelerations
of all tests is given in Fig. 25. The bare frame (BF) and the frame with
and appearance of significant nonlinearities (damages) in the column
flexible ground story (IF1) experienced cracks at some joints between
joints. For the specimen IF1 under PGA = 0.8 g, peak bottom story drift
lower beam and columns. The frames with openings in the infill (IF3,
ratio was 1/66, which is approx. 17% greater than for the specimen BF.

9
G. Baloevic et al. Soil Dynamics and Earthquake Engineering 161 (2022) 107404

Fig. 16. Relationship between displacement u2 and reinforcement strain εS at point B for PGA = 0.8 g.

Fig. 17. Relationship between displacement u2 and reinforcement strain εS at point A for PGA = 0.8 g.

IF4) experienced cracks at the interface between the infill and the shake- strains and natural oscillation frequencies of the tested masonry-infilled
table. The frame with flexible upper story (IF2) and the fully infilled frames, most important remarks are summarized in the conducted
frame (IF5) had no visible cracks. The crack patterns are in agreement experimental study. For the case of a frame with flexible ground story
with the determined natural oscillation frequencies in Fig. 15. The bare (IF1) under PGA = 0.8 g, in relation to the next unfavorable case of bare
frame (BF) and the frame with flexible ground story (IF1) experienced frame (BF), it was confirmed that:
the greatest reduction in the stiffness, i.e. the greatest reduction in
natural frequencies. These specimens had greatest visible cracks that - A bottom story drift was 17% increased.
occurred at the joint between column and beam, which affected their - Tensile strains in longitudinal reinforcement at the bottom of column
structural behavior under dynamic action. at ground story were 34% greater.
- Vertical compressive concrete strains at the bottom of column were
19% greater.
3.4. Discussion
Unfortunately, sometimes in the infilled frames there are present
An acceptably designed infill at all stories is often present in many and/or flexible floors above the ground floor, which is also unfavorable
buildings with frame load-bearing structure. Unfortunately, in some solution. For the case of a frame with flexible upper floor (IF2) under
buildings there is a flexible ground story due to various and sometimes PGA = 0.8 g, in relation to the next unfavorable case of bare frame (BF),
unnecessary reasons, which is very unfavorable in seismic areas. This the following was also experimentally confirmed:
results in generating of high seismic (inertial) forces above the ground
story, where resistance and bearing capacity of the building is drastically - Upper story drift was 3 times lower.
reduced.
Based on the measured values of displacements, accelerations,

10
G. Baloevic et al. Soil Dynamics and Earthquake Engineering 161 (2022) 107404

Fig. 18. Relationship between displacement u2 and concrete strain εC at point D for PGA = 0.8 g.

Fig. 19. Peak displacements at the top of upper floor (u2).

Fig. 20. Peak displacements at the top of bottom floor (u1).

- Vertical compressive concrete strains at the bottom of column at damage level in columns and beams of an infilled frame.
upper floor were 3 times greater, even though the drift was 3 times The results of the conducted experimental study can be correlated
lower. with the results of the previously conducted numerical study on the
behavior of masonry-infilled reinforced concrete frames [36,37].
In the conducted tests, after shake-table excitation with PGA = 0.8 g,
the final stiffness of the frames BF, IF1 and IF2 was 31.2%, 66.6 and 4. Conclusions
64.7% of their initial stiffness, respectively. For the cases with an infill at
all stories IF3, IF4 and IF5, the final stiffness was 94.6%, 98.1 and 98.1% On the basis of the conducted experimental shake-table study on the
of their initial stiffness, respectively. The infill significantly reduces the behavior of single-bay two-story reinforced concrete frames under the

11
G. Baloevic et al. Soil Dynamics and Earthquake Engineering 161 (2022) 107404

Fig. 21. Peak tensile reinforcement strains at the bottom of column at point C.

Fig. 22. Peak tensile reinforcement strains at the bottom of column at point A.

Fig. 23. Peak compressive concrete strains at the bottom of column at point D.

action of a single horizontal accelerogram, with variants of bare frame • When designing an infilled frame in practice, it is necessary to
and with five different infill configurations, the following can be reliably include an adequate stiffness and resistance of the infill in the design
concluded. model of a frame.
• It is very important that an infill is evenly distributed in a frame
• Infilled frames generally have a completely different behavior through all its stories, i.e. that its stiffness and resistance at each story
(displacement, resistance, safety) under an earthquake in relation to is correctly designed. It is desirable to increase the stiffness of the
bare frames. An infill increases the structural mass and the stiffness infill from the top to the ground story, which is the same principle for
of the bare frame, which generates greater inertial (seismic) forces in frames and other load-bearing structures. Namely, it is desirable that
the structure. An infilled frame can have greater but also smaller the level of structural stiffness and resistance corresponds to the level
bearing capacity and safety than a bare frame, which depends on of internal forces in it, including the effect of an earthquake.
design of infill distribution in a frame, together with the effect of its • Irregular distribution of masonry infill across the elevation in a frame
stiffness and resistance. can be very unfavorable in seismic area. The conducted shake-table
study showed that, compared to the bare frame, the frame with
flexible ground story experienced a significant increase in bottom

12
G. Baloevic et al. Soil Dynamics and Earthquake Engineering 161 (2022) 107404

Fig. 24. Peak compressive concrete strains at the bottom of column at point E.

Fig. 25. Final crack patterns in the tested frames.

story drift, longitudinal reinforcement strains and vertical concrete Acknowledgement

strains. Further, the frame with the flexible upper floor had greater
vertical concrete strains at the upper floor compared to the bare This research is partially supported through project KK.01.1.
frame. February 1, 0027, a project co-financed by the Croatian Government and
• The conducted experiments have also confirmed that in cases where the European Union through the European Regional Development Fund
an infill is present at all stories, even with weaker infill due to - the Competitiveness and Cohesion Operational Programme.
opening, such frames have significantly low displacements, concrete
and reinforcement strains, i.e. they have higher load-bearing ca­ Appendix A. Supplementary data
pacity than the case of a bare frame.
• An infill not only increases the initial stiffness of a bare frame, but Supplementary data to this article can be found online at https://doi.
also ensures that the overall stiffness of a structure is less reduced org/10.1016/j.soildyn.2022.107404.
when it is exposed to an action of a stronger earthquake.
References
Declaration of competing interest
[1] Mehrabi AB, Shing PB, Schuller MP, Noland JL. Experimental evaluation of
masonry-infilled RC frames. J Struct Eng 1996;122(3):228–37.
The authors declare that they have no known competing financial
[2] Caliò I, Pantò B. A macro-element modelling approach of Infilled Frame Structures.
interests or personal relationships that could have appeared to influence Comput Struct 2014;143:91–107.
the work reported in this paper. [3] Asteris PG, Antoniou ST, Sophianopoulos DS, Chrysostomou CZ. Mathematical
macromodeling of infilled frames: state of the art. J Struct Eng 2011;137(12):
1508–17.
[4] Cavaleri L, Di Trapani F. Cyclic response of masonry infilled RC frames:
experimental results and simplified modeling. Soil Dynam Earthq Eng 2014;65:
224–42.

13
G. Baloevic et al. Soil Dynamics and Earthquake Engineering 161 (2022) 107404

[5] Fotos A, Foskolos F, Tsaris AK, Repapis CC, Asteris PG. Inelastic response of [25] Koutromanos I, Kyriakides M, Stavridis A, Billington S, Shing PB. Shake-table tests
masonry infilled reinforced concrete structures. In: COMPDYN 2017 - proceedings of a 3-story masonry-infilled RC frame retrofitted with composite materials.
of the 6th international conference on computational methods in structural J Struct Eng 2013;139(8):1340–51.
dynamics and earthquake engineering, vol. 2; 2017. p. 3312–22. [26] Magliulo G, Petrone C, Capozzi V, Maddaloni G, Lopez P, Manfredi G. Seismic
[6] Dolšek M, Fajfar P. Soft storey effects in uniformly infilled reinforced concrete performance evaluation of plasterboard partitions via shake table tests. Bull Earthq
frames. J Earthq Eng 2001;5(1):1–12. Eng 2014;12(4):1657–77.
[7] Quayyum S, Alam MS, Rteil A. Seismic behavior of soft storey mid-rise steel frames [27] Petrone C, Magliulo G, Manfredi G. Shake table tests for the seismic assessment of
with randomly distributed masonry infill. Steel Compos Struct 2013;14(6):523–45. hollow brick internal partitions. Eng Struct 2014;72:203–14.
[8] Fiore A, Porco F, Raffaele D, Uva G. About the influence of the infill panels over the [28] Palios X, Fardis MN, Strepelias E, Bousias SN. Unbonded brickwork for the
collapse mechanisms actived under pushover analyses: two case studies. Soil protection of infills from seismic damage. Eng Struct 2017;131:614–24.
Dynam Earthq Eng 2012;39:11–22. [29] Pavese A, Lanese I, Nascimbene R. Seismic vulnerability assessment of an infilled
[9] Mallick DV, Severn RT. The behaviour of infilled frames under static loading. In: reinforced concrete frame structure designed for gravity loads. J Earthq Eng 2017;
Proceedings of the institution of civil engineers (london), vol. 38; 1967. p. 639–56. 21(2):267–89.
[10] Dhanasekar M, Page AW. Influence of brick masonry infill properties on the [30] Zhang H, Kuang JS, Yuen TYP. Low-seismic damage strategies for infilled RC
behaviour of infilled frames. Proc Inst Civ Eng 1986;81(pt 2):593–605. frames: shake-table tests. Earthq Eng Struct Dynam 2017;46(14):2419–38.
[11] Saneinejad A, Hobbs B. Inelastic design of infilled frames. J Struct Eng 1995;121 [31] Yuen TYP, Zhang H-H, Kuang JS, Huang Q. Shake table tests on RC frame infilled
(4):634–50. by slitted masonry panels. Bull Earthq Eng 2018;16(9):4027–52.
[12] Mehrabi AB, Shing PB. Finite element modeling of masonry-infilled RC frames. [32] Singhal V, Rai DC. Behavior of confined masonry walls with openings under in-
J Struct Eng 1997;123(5):604–13. plane and out-of-plane loads. Earthq Spectra 2018;34(2):817–41.
[13] Mohammad AF, Khalid F, Khan RA. Finite element micro-modelling of RC frames [33] Xue J, Xu D. Shake table tests on the traditional column-and-tie timber structures.
with variant configurations of infill masonry. Struct Eng Mech 2022;81(4): Eng Struct 2018;175:847–60.
395–409. [34] Gong M, Zuo Z, Wang X, Lu X, Xie L. Comparing seismic performances of pilotis
[14] Koutromanos I, Stavridis A, Shing PB, Willam K. Numerical modeling of masonry- and bare RC frame structures by shaking table tests. Eng Struct 2019;199:109442.
infilled RC frames subjected to seismic loads. Comput Struct 2011;89(11–12): [35] European Committee for Standardization (CEN). Eurocode 8—design of structures
1026–37. for earthquake resistance, part 1: general rules, seismic actions and rules for
[15] Baloevic G, Radnic J, Matesan D, Grgic N, Banovic I. Comparison of developed buildings. 2004.
numerical macro and micro masonry models for static and dynamic analysis of [36] Baloevic G, Radnic J, Harapin A. Numerical dynamic tests of masonry-infilled RC
masonry-infilled steel frames. Lat Am J Solid Struct 2016;13(12):2251–65. frames. Eng Struct 2013;50:43–55.
[16] Asteris PG. Finite element micro-modeling of infilled frames. Electron J Struct Eng [37] Radnic J, Baloevic G, Grgic N, Harapin A, Buzov A. The effect of flexibility in
2008;8:1–11. ground storey of concrete walls and infilled frames on their seismic response. Mater
[17] Stavridis A, Shing PB. Finite-element modeling of nonlinear behavior of masonry- Werkst 2014;45(4):244–57.
infilled RC frames. J Struct Eng 2010;136(3):285–96.
[18] Mucedero G, Brunesi E, Parisi F. Nonlinear material modelling for fibre-based
progressive collapse analysis of RC framed buildings. Eng Fail Anal 2020;118: Nomenclature
104901.
[19] Fabbrocino F, Farina I, Berardi VP, Ferreira AJM, Fraternali F. On the thrust
Variable: Unit (SI) Description
surface of unreinforced and FRP-/FRCM-reinforced masonry domes. Compos B Eng
u2: mm Horizontal displacement at the top of the upper floor
2015;83:297–305.
u1: mm Horizontal displacement at the top of the bottom floor
[20] Fraternali F, Carpentieri G, Modano M, Fabbrocino F, Skelton RE. A tensegrity
a2: m/s2 Horizontal acceleration at the top of the upper floor
approach to the optimal reinforcement of masonry domes and vaults through fiber-
a1: m/s2 Horizontal acceleration at the top of the bottom floor
reinforced composite materials. Compos Struct 2015;134:247–54.
εs,A: mm/m (‰) Strain in the longitudinal steel reinforcement at point A
[21] Mohebkhah A, Tasnimi AA, Moghadam HA. Nonlinear analysis of masonry-infilled
εs,B: mm/m (‰) Strain in the longitudinal steel reinforcement at point B
steel frames with openings using discrete element method. J Constr Steel Res 2008;
εs,C: mm/m (‰) Strain in the longitudinal steel reinforcement at point C
64(12):1463–72.
εc,D: mm/m (‰) Vertical concrete strain at point D
[22] Pantò B, Caliò I, Lourenço PB. A 3D discrete macro-element for modelling the out-
εc,E: mm/m (‰) Vertical concrete strain at point E
of-plane behaviour of infilled frame structures. Eng Struct 2018;175:371–85.
εc,F: mm/m (‰) Vertical concrete strain at point F
[23] Fardis MN, Bousias SN, Franchioni G, Panagiotakos TB. Seismic response and
εc,G: mm/m (‰) Vertical concrete strain at point G
design of RC structures with plan-eccentric masonry infills. Earthq Eng Struct
PGA: m/s2 Peak ground acceleration
Dynam 1999;28(2):173–91.
[24] Hashemi A, Mosalam KM. Shake-table experiment on reinforced concrete structure
containing masonry infill wall. Earthq Eng Struct Dynam 2006;35(14):1827–52.

14