Engineering Structures 126 (2016) 598–617

Contents lists available at ScienceDirect

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

Experimental study on in-plane cyclic response of partially grouted

reinforced concrete masonry shear walls
Pablo Ramírez a, Cristián Sandoval a,b,⇑, José Luis Almazán a
a
Department of Structural and Geotechnical Engineering, Pontificia Universidad Católica de Chile, Casilla 306, Correo 22, Santiago, Chile
b
School of Architecture, Pontificia Universidad Católica de Chile, Casilla 306, Correo 22, Santiago, Chile

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

Article history: This article describes the experimental results of ten partially grouted reinforced concrete masonry shear
Received 8 October 2015 walls (PG-RCMSW) that were subjected to reverse lateral in-plane cyclic loads. The variables analysed in
Revised 5 June 2016 this study were: aspect ratio, shear reinforcement ratio and level of axial pre-compression. The influence
Accepted 4 August 2016
of each of these variables on different structural parameters such as degradation of stiffness, shear
strength, displacement ductility, dissipation of energy, hysteretic damping and level of drift, was evalu-
ated. In addition, the precision of certain analytical expressions reported in the literature to predict the
Keywords:
maximum shear strength of walls was examined and contrasted with the experimental results obtained.
Reinforced masonry
Concrete masonry
The results showed that the evolution of the damage was propagated in a similar way in all the walls
Cyclic loads tested until reaching the level of maximum strength. From this point, the evolution and extension of the
Partially grouted damage depended on the characteristics and loading conditions particular to each wall. Also, a strong
Shear failure interdependence of the variables studied was identified, which became evident in the evaluation of shear
Shear strength strength, dissipation of energy, hysteretic damping, and level of drift. Using a bilinear idealization, dis-
placement ductility values between 2.85 and 7.94 were found to reflect the presence of a moderate level
of ductility in the walls tested. The equivalent viscous damping ratio associated with a non-linear
response was found to range from 5% to 11%, indicating a moderate level of energy dissipation before
the peak load was reached. Finally, the comparison between the predictions of the analytical expressions
from the literature and the experimental results showed that those expressions that incorporated some
interdependence in their design variables did not possess an appropriate degree of confidence to be
applied in assessing the shear strength of PG-RCMSW, while expressions proposed by some international
codes seem to be more reliable and conservative.
Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction phragms during a seismic event, horizontal seismic actions are

Reinforced masonry is one of the most frequently used struc-
tural systems worldwide for the construction of low and
medium-height buildings in areas of moderate or high seismic
activity. This structural system is mainly comprised of reinforced
masonry shear walls, which are arranged in two major axes of a
building. Its lateral load-carrying capacity depends on the in-
plane resistances of shear walls because the in-plane stiffness of
a shear wall is far greater than its out-of-plane stiffness [6].
Because the reinforced masonry shear wall buildings are com-
monly composed by reinforced concrete slabs that act as rigid dia-
⇑ Corresponding author at: Department of Structural and Geotechnical Engineer-
ing, Pontificia Universidad Católica de Chile, Casilla 306, Correo 22, Santiago, Chile.
E-mail addresses: pnramirez@uc.cl (P. Ramírez), csandoval@ing.puc.cl
(C. Sandoval), jlalmaza@ing.puc.cl (J.L. Almazán).
mainly transferred to walls parallel to the load direction [34]. Con-
sequently, frequently observed damage after seismic events is
related to in-plane failure modes.
In Chile, reinforced masonry have been used since the mid-
seventies in the construction of social housing and residential
buildings of up to four storeys [24]. Recent post-earthquake obser-
vations have shown that the seismic response of this type of con-
structions is still deficient [3,4,30,32,37]. In fact, the earthquakes
of Tarapacá in 2005 (Mw 7.8), Maule in 2010 (Mw 8.8), and Iquique
in 2014 (Mw 8.2) caused severe structural damage in several
masonry buildings and even collapse in some cases. During these
events, major problems were observed in those buildings that were
constructed with partially grouted reinforced concrete masonry
shear walls (PG-RCMSW). The failure mechanism observed in the
majority of the buildings affected by these seismic events was by
shear failure with a pattern of diagonal cracking. As is well known,

http://dx.doi.org/10.1016/j.engstruct.2016.08.010
0141-0296/Ó 2016 Elsevier Ltd. All rights reserved.
 P. Ramírez et al. / Engineering Structures 126 (2016) 598–617 599

Nomenclature

Awh horizontal gross cross-sectional area of test wall (mm2 ) f yh yield strength shear reinforcement (MPa)
A wv vertical gross cross-sectional area of test wall (mm2 ) f uh ultimate strength shear reinforcement (MPa)
Aw horizontal cross-section area of the wall (mm2 ) f yv yield strength vertical reinforcement (MPa)
Anv net shear area of the wall (mm2 ) f uv ultimate strength vertical reinforcement (MPa)
Av area of vertical reinforcement (mm2 ) ft tensile strength of masonry (MPa)
Ah area of horizontal reinforcement (mm2 ) FR resistant factor
Arh area of single horizontal reinforcing steel bar (mm2 ) Gm shear’s modulus of masonry, based on net area (MPa)
b shear stress distribution factor G
m shear’s modulus of masonry, based on gross area (MPa)
C rh horizontal reinforcement capacity reduction factor In Moment of inertia of the net section of the uncracked
d wall length (mm) wall (mm4 )


d effective depth of the wall (mm) L length of wall (mm)
dr v diameter of one vertical reinforcement bar due to dowel M maximum moment at the section under consideration
action (mm) (N-mm)
dE elastic idealized displacement (mm) g efficiency factor of reinforcement horizontal
dSL displacement to elastic limit state (mm) n number of vertical reinforcement bars
dMR displacement to maximum resistance limit state (mm) qh horizontal reinforcement ratio
Em Young’s modulus of masonry, based on net area (MPa) qv vertical reinforcement ratio
E
m Young’s modulus of masonry, based on gross area (MPa) qv e flexural reinforcement ratio due to the cross area of
Esh Young’s modulus of shear reinforcement (MPa) edge tension bar (%)
Esv Young’s modulus of vertical reinforcement (MPa) Sv vertical separation of horizontal reinforcement (mm)
ET accumulated dissipate energy up to maximum resis- P axial load (N)
tance (kN-mm) R2 correlation factor
EH energy dissipated for a load cycle (kN-mm) sm masonry shear strength, calculated on net area (MPa)
hw wall height (mm) s
m masonry shear strength, calculated on gross area (MPa)
hef wall effective height (mm) rn axial pre-compression stress, calculated on gross area
K s;i secant stiffness of an i cycle (kN/mm) (MPa)
Ko initial stiffness to an imposed lateral displacement of ro axial pre-compression stress, calculated on net area
0.20 mm (kN/mm) (MPa)
K E;exp experimental elastic stiffness (kN/mm) ra average compression stress due to vertical load (MPa)
K E;theorical theoretical elastic stiffness (kN/mm) mn tangential stress shear calculated on net area (MPa)
KR post-cracking stiffness (kN/mm) m
m diagonal compression resistance calculated on gross
kp coefficient of the effect of flexural reinforcement area (MPa);
ku reduction factor V shear force (N)
a parameter of the stiffness degradation Vm nominal shear strength provided by masonry (N)
b parameter of the stiffness degradation Vs nominal shear strength provided by shear reinforce-
at cross area of edge tension bar (mm2 ) ment (N)
d factor concerning the type of grouting Vn nominal shear strength (kN)
dmax;i maximum displacement in the load cycle (mm) V exp experimental shear strength (kN)
c factor concerning loading method VE lateral force to idealized elastic limit state (kN)
cg grouted shear wall factor V SL lateral force to elastic limit state (kN)
0 V MR Lateral force to maximum resistance limit state (kN)
f bm cement mortar flexural strength (MPa)
0
f cu concrete block compression strength (MPa) DV max difference of peak lateral loads of an i cycle (kN)
0
f cm cement mortar compression strength (MPa) Ddmax difference of displacement corresponding to peak lateral
0 loads of an i cycle (mm)
f cr grout cylinder compression strength (MPa)
fm
0
compressive strength of masonry prism, calculated on D drift s
net area (MPa) lMR displacement t to maximum resistance limit state;
0
neq equivalent hysteretic damping (%)
fm compressive strength of masonry prism, calculated on
gross area (MPa) s spacing of shear reinforcement (mm)
t thickness of wall

shear failure in a wall is a mechanism of a brittle nature that is
characterised by a low capacity for dissipation of energy and rapid
degradation of stiffness and strength after the maximum lateral
capacity has been reached.
Given the importance of seismic action, the behaviour of
masonry buildings comprised of PG-RCMSW is receiving increasing
attention worldwide. Numerous experimental investigations car-
ried out in Chile in recent decades [21,19,20,26,35] as well as in
other countries [8,12,22,23,28,39,9,10] have demonstrated that
the properties of the constituent materials, the wall aspect ratio,
the level of axial load, and the ratio and distribution of vertical
and horizontal reinforcements are the principal design parameters
that control the response and seismic performance of PG-RCMSW.
From the above-cited experimental results it can be seen that
an increase in axial load causes a rise in the shear strength of the
walls [22] and additional frictional strength along the diagonal
cracks, which favours an increase in hysteretic energy dissipation
[23]. However, an increase in axial load also gives rise to the walls
developing lower ductility and exhibiting more brittle behaviour
than walls without axial load [26]. With regard to the influence
of the vertical reinforcement ratio, it can be seen that if this ratio
increases, shear strength also increases [8,35,39], and the walls
show a greater number of diagonal cracks but where the cracking
is less wide. Also, in agreement with Tomazevic [36], the dowel
action that develops as a result of vertical reinforcement also
contributes to an increase in shear strength. However, a greater
600 P. Ramírez et al. / Engineering Structures 126 (2016) 598–617

vertical reinforcement ratio would not appear to affect the dis-
placement ductility or the stiffness of the walls [28]. Regarding
the horizontal reinforcement ratio, it has been shown that an
increase in this quantity leads to a rise in shear strength [22], a
reduction in the degradation of stiffness and a substantial improve-
ment in the structural integrity of the walls [26]. Nevertheless,
from a certain horizontal reinforcement ratio, the shear strength
does not rise significantly [8] while at the same time a considerable
reduction in the displacement ductility can result [26,28]. On the
other hand, in agreement with Lüders and Hidalgo [19], depending
on the quality of the horizontal steel used, there is a certain hori-
zontal reinforcement ratio that will be exploited efficiently by a
wall in order to withstand shear stresses. Lastly, several research-
ers [22,39] have reported that the shear strength increases or
decreases as the aspect ratio of the wall is reduced or raised,
respectively. Notice that the majority of aforementioned experi-
mental research studies have employed relatively high flexural
reinforcement ratios in order to ensure the shear failure mode dur-
ing the tests.
On the basis of this background and in order to obtain greater
experimental evidence that permits to make recommendations
for future updating and/or modifications of the current Chilean
Code of reinforced masonry [17], the present research evaluates
the effect of three design variables: wall aspect ratio, shear rein-
forcement ratio and level of axial pre-compression on the in-
plane cyclic behaviour of PG-RCMSW. The influence of these design
variables was measured by quantifying certain parameters of seis-
mic interest, such as degradation of stiffness, shear strength, dis-
placement ductility, dissipation of energy, hysteretic damping
and drift. In addition, the accuracy of selected analytical expres-
sions of shear strength reported in the literature, including two
international codes, was evaluated by means of a comparison with
the experimental results.
2. Experimental programme
2.1. Test specimens
A total of ten full-scale partially grouted reinforced concrete
masonry shear walls were designed and built according to
NCh1928Of.93 [17]. The characteristics of the specimens tested
and their design details are summarised in Table 1 and Fig. 1,
respectively. Since the purpose of this study requires that speci-
mens respond and fail in shear, all walls were heavily reinforced
in the vertical direction. For this, each wall was reinforced with a
£22 mm bar in each vertical edge, except in those walls with
aspect ratio of 0.44 where £16 mm bars were used. In addition,
£10 mm steel bars were placed in the central region of walls in
order to comply with the horizontal spacing suggested by the Chi-
lean reinforced masonry standard (NCh1928Of.93 [17]). As men-
tioned, the variables considered in this study were wall aspect
ratio, horizontal reinforcement ratio and axial pre-compression.
Due to the characteristics of PG-RCMSW, only the cells contain-
ing vertical reinforcement were filled with grout. Three aspect
ratios were studied: hef =d ¼ 0:44 for short walls, hef =d ¼ 0:97 for
ordinary walls, and hef =d ¼ 1:95 for slender walls, which represent
walls typically found in reinforced concrete masonry buildings
characteristic for Chile. The shear reinforcement ratio considered
varied between 0.04% and 0.09%, while the axial load applied
was similar to the load produced by two and three storeys on a
wall of the ground floor. All the walls tested withstood axial pre-
compression of the order of 10% of the masonry compression
0
strength f m except for the walls M7 and M10, which were tested
without vertical compression load. It is noteworthy that the walls
examined in this experimental programme complement the avail-
able database by providing new experimental results of PG-
RCMSW with a horizontal reinforcement ratio above 0.06% and
with axial compression other than nil.
2.2. Construction of wall specimens
All the walls were constructed by experienced masons. The
building process was carried out in three stages: (1) construction
of the lower reinforced concrete beam, where vertical reinforce-
ment bars were anchored, (2) construction of the masonry wall
including the reinforcement embedded in the bed joints, and (3)
construction of the upper reinforced concrete beam, where also
vertical reinforcement bars were anchored. The pattern used in
all the walls was a running bond pattern with head and bed joint
thickness of 10 mm. For the joints a pre-mixed dry cement mortar
was used which is the typical kind used in Chilean masonry con-
structions. As horizontal reinforcement, a prefabricated ladder-
type was used made up of two longitudinal side rods of 4.2 mm
in diameter and 85 mm apart from each other, and with welded
cross rods every 300 mm. The ladders were embedded in horizon-
tal mortar joints and were distributed high up on the wall in agree-
ment with the horizontal reinforcement ratio defined by the
experimental programme (Table 1). The grout used to fill the cells
with vertical reinforcement was also of the pre-mixed dry type.
This grout was compacted manually and emptied out in two stages
according to the height of each wall. In the meantime, beams of
reinforced concrete were placed in the lower and higher parts of
the walls in order to anchor the vertical reinforcement and to
assure a uniform distribution of the vertical and horizontal loads
applied. All the walls were cured in a laboratory environment.

Table 1
Design details of the masonry specimens.

Wall Wall dimensions Aspect ratio Vertical reinforcement Horizontal reinforcement Axial pre-
compression
d (mm) hw (mm) hef (mm) hef/d Vertical Av (mm2) qv (%) Joint reinforcement sv (mm) Av (mm2) qv (%) r
n (MPa)
reinforcement
M1 1990 2000 1930 0.97 2£22 + 2£10 917 0.33 4-2£4.2 400 111 0.04 0.56
M2 1990 2000 1930 0.97 2£22 + 2£10 917 0.33 4-2£4.2 400 111 0.04 0.56
M3 1990 2000 1930 0.97 2£22 + 2£10 917 0.33 9-2£4.2 200 249 0.09 0.56
M4 1990 2000 1930 0.97 2£22 + 2£10 917 0.33 9-2£4.2 200 249 0.09 0.56
M5 2590 1400 1130 0.44 2£16 + 3£10 638 0.18 3-2£4.2 400 83 0.04 0.56
M6 2590 1400 1130 0.44 2£16 + 3£10 638 0.18 6-2£4.2 200 166 0.08 0.56
M7 2590 1400 1130 0.44 2£16 + 3£10 638 0.18 3-2£4.2 400 83 0.04 0.00
M8 990 2000 1930 1.95 2£22 + 1£10 839 0.60 4-2£4.2 400 111 0.04 0.56
M9 990 2000 1930 1.95 2£22 + 1£10 839 0.60 9-2£4.2 200 249 0.09 0.56
M10 990 2000 1930 1.95 2£22 + 1£10 839 0.60 4-2£4.2 400 111 0.04 0.00

r
n : The axial pre-compression was calculated considering top transference reinforced beam self-weight and imposed load by vertical actuator.
 P. Ramírez et al. / Engineering Structures 126 (2016) 598–617
Fig. 1. Design details of test walls.
Fig. 2. Construction of walls: (a) view of bed-joint reinforcement, (b) view of walls.
Fig. 2 shows the placement process of the horizontal reinforcement
and a general view of the specimens.
2.3. Material properties
The materials employed to the construction of the walls were
those commonly used in the homes and residential buildings of
reinforced concrete masonry in Chile. As masonry unit, two-cells
hollow industrial concrete blocks with nominal dimensions of
390 mm in length, 190 mm in height, and 140 mm in width were
used. The thickness of the face-shells and webs of the units varied
between 35 and 40 mm. An average compression strength of
6.36 MPa was obtained in agreement with NCh182.Of55 [13]. The
mortar used for joints had an average compression strength of
18.03 MPa and an average flexural strength of 4.7 MPa, both in
agreement with NCh158.Of67 [14]. The grout used for filling the
cavities with vertical reinforcement gave an average compression
strength of 31.7 MPa as according to NCh1037.Of77 [16].
For the horizontal reinforcement a high-strength electro-
welded ladder-type steel reinforcement of quality AT56-50H was
used with an average yield strength of 610 MPa and an average
ultimate stress of 660 MPa, both of which were obtained in agree-
601
ment with the direct tension test of NCh200.Of72 [15]. For the ver-
tical reinforcement, corrugated steel rods of quality A630-420H
were used with an average yield strength of 474 MPa and an aver-
age ultimate stress of 765 MPa, both also obtained according to
NCh200.Of72 [15].
To determine the properties of the masonry as a composite
material, ten prisms and ten wallets were built and tested for uni-
axial compression according to NCh1928.Of93 [18] and for diago-
nal compression according to NCh2123.Of93 [18], respectively.
0

The average compression strength (f m ) obtained for the prisms,
considering the gross area, was 5.54 MPa and their average modu-
lus of elasticity ðE
m Þ, calculated as the slope of the straight secant
between 5% and 50% of the maximum strength, was 6465 MPa.
In this way, the masonry used in the present experimental pro-
0

gramme developed an initial modulus of elasticity of 1167f m . In
the case of the diagonal compression test on masonry wallets an
average shear strength ðs
m Þ, considering the gross area, of
0.61 MPa was obtained as well as an average shear modulus (G
m )
calculated at 50% of maximum strength, of 1976 MPa. Table 2 sum-
marises the mechanical properties obtained for the concrete
blocks, joint mortar, grout, reinforcement, and the masonry as
composite material.
602 P. Ramírez et al. / Engineering Structures 126 (2016) 598–617

Table 2
Summary of material properties.

Property Notation Average CV Test standard

(Mpa) (%)
0
Concrete block compression strength f cu 6.36 16 NCh182.Of 55
0
Cement mortar compression strength f cm 18.03 2 NCh158.Of 67
0
Cement mortar flexural strength f bm 4.73 6 NCh158.Of 67
0
Grout cylinder compression strength f cr 31.72 2 NCh103.Of 77
Yield strength shear reinforcement f yh 610 3 NCh200.Of 72
Ultimate strength shear reinforcement f uh 660 3 NCh200.Of 72
Young’s modulus of shear reinforcement Esh 189,598 8 NCh200.Of 72
Yield strength vertical reinforcement f yv 474 3 NCh200.Of 72
Ultimate strength vertical reinforcement f uv 765 3 NCh200.Of 72
Young’s modulus of vertical reinforcement Esv 209,145 9 NCh200.Of 72
Masonry compression strength 0
5.54 15 NCh2123.Of 97 Mod.2003
fm
0
fm 8.65
Young’s modulus of masonry E
m 6465 17 NCh2123.Of 97 Mod.2003
Em 10,115
Masonry shear strength s
m 0.61 11 NCh1928.Of 93 Mod.2009
sm 0.95
Shear’s modulus of masonry G
m 1976 24 NCh1928.Of 93 Mod.2009
Gm 3083
0

The properties f m , Em , sm and Gm are calculated on net area while the properties f 0m , E
m , s
m and G
m are calculated on gross area.

Fig. 3. Test set-up.

2.4. Test set-up and instrumentation
Fig. 3 shows the characteristics of the test setup. In order to
induce cyclic shear loads similar to those that an earthquake pro-
duces in the plane of the walls, a metal loading frame made up
of a horizontal actuator of a maximum capacity of 500 kN and a
vertical actuator of a maximum capacity of 700 kN was used. All
the bases of the walls were anchored to the strong floor of the lab-
oratory and the rotation of their upper beams was not restricted
simulating a condition of cantilever walls.
The horizontal actuator was connected to the metal loading
frame, which was the responsible for transmitting the lateral dis-
placements to the wall. Fig. 3 also shows the distribution of the
internal forces induced in each one of specimens tested and the
way in which the aspect ratio was defined by the relation between
the effective height (hef ) and the total length of the wall (d).
The instrumentation of the walls consisted of two load cells,
sixteen linear variable differential transformer (LVDTs) transducers
and two strain gauges as shown in Fig. 4. The load cells [0] and [1],
placed between the actuators and the metal frame, measured the
 P. Ramírez et al. / Engineering Structures 126 (2016) 598–617 603

Fig. 4. Instrumentation.

magnitude of the lateral force and vertical load, respectively. The
LVDTs [2] and [3] measured the lateral displacements at one third
of the height and at the point of inflection of the walls, respec-
tively. The LVDT [4] measured the lateral slipping of the bottom
concrete beam and the LVDT [5] registered the relative horizontal
slipping between this beam and the wall. The LVDTs [6] and [7]
measured the rotation of the bottom concrete beam. The deforma-
tions due to flexion and shear were measured by the LVDTs [8] to
[13] and [14] to [15], respectively, and the displacements outside
the plane of the wall were monitored by the LVDTs [16] and
[17]. In addition, the strain gauges [SG-1] [SG-2], which were
installed during the construction of the walls in the lower part of
the vertical bars of both lateral ends, were used to measure the
strains during the cyclic response of the vertical reinforcement.
2.5. Testing procedure
All the walls were subjected to reverse in-plane cyclic shear
loading controlled by displacement. The velocity applied to all
the walls was 7 mm/min. The displacements were increasing until
the failure of the walls was achieved. As collapse of the wall was
considered the state in which the lateral strength had dropped
by at least 20% of the maximum registered lateral strength during
the test. In order to capture any evidence of degradation in stiffness
and strength, two cycles were applied for the same amplitude of
lateral deformation. Fig. 5 shows the protocol of displacements
imposed in the tests and their graphic representation. It should
be noted that this protocol was the same used in previous Chilean
investigations [2]. For those walls with axial pre-compression, the

40
Number Number
Displacement (mm) Displacement (mm)
of Cycle of Cycle 30
1-2 0.20 27-28 10.50
3-4 0.40 29-30 11.60
Lateral Displacement [mm]

20
5-6 1.00 31-32 12.50
7-8 1.60 33-34 13.50 10
9-10 2.00 35-36 15.00
11-12 2.60 37-38 18.50 0
13-14 4.00 39-40 19.50
-10
15-16 5.00 41-42 22.40
17-18 5.80 43-44 25.40
-20
19-20 7.00 45-46 28.30
21-22 8.00 47-48 31.20 -30
23-24 8.40 49-50 35.10
25-26 9.50 51-52 39.00 -40
1 8 15 22 29 36 43 50
Cycle Number

Fig. 5. Imposed displacement history.

604 P. Ramírez et al. / Engineering Structures 126 (2016) 598–617
vertical load was applied gradually until the desired value before
the lateral displacements was started. The action of the vertical
actuator was transmitted to the metal frame through a system
made up of a rigid steel girder and metal rollers resting on a rein-
forced concrete beam that served as a load transfer system. This
beam distributed the vertical load to the specimens and the rollers
permitted the free lateral movement of the metal frame with
respect to the position of the vertical actuator.
3. Experimental results
3.1. Force-displacement diagrams
The hysteresis diagrams and envelope curves of lateral force
versus displacement for each wall tested are presented in Fig. 6.
With the exception of wall M2, all the walls showed a symmetric
hysteretic behaviour up to the level of maximum strength. The
asymmetric hysteretic response in the wall M2 was basically due
to the imposition of asymmetric cycles of displacements caused
by an accidental error in the execution of the test. In spite of this,
the results obtained from this specimen are considered to be reli-
able for inclusion in subsequent evaluations. All the walls showed
a linear-elastic behaviour up to a determined level of lateral dis-
placement, where the first diagonal cracks appeared. This initial
linear-elastic behaviour was characterised by narrow cycles of hys-
teresis, which exhibited a low dissipation capacity of energy. With
the continuous imposition of displacements, the area of the
hysteresis cycles began to increase significantly leading to the
development of high levels of dissipation of energy, inelastic defor-
mation and progression in the damage. In the walls M2, M4, M7,
and M10 the phenomenon of ‘pinching’ was observed, which man-
(a) (b)
(e) (f)
(h) (i)
Fig. 6. Lateral force-displacement diagrams.
ifested in walls with and without axial load. This finding is contrary
to that reported by Minaie et al. [23] who observed this phe-
nomenon only in walls without axial load.
In the case of ordinary walls (hef =d ¼ 0:97Þ such as M1 and M2,
with minimum horizontal reinforcement ratio (qh ¼ 0:04%) and
0
with axial pre-compression rn =f m  0:10, the deterioration of
strength was more marked in one direction of load than in the
other. The reason for this was due to the fact that in one of the
directions, the evolution in size and thickness of the predominant
diagonal crack was faster and more sudden than in the other one.
In the direction where the bigger drop in strength of the wall was
produced, the predominant diagonal crack went through the wall
in its totality, completely penetrated the compression toe, caused
face-shell spalling, and simultaneously fractured several bed-
joint reinforcements. This would explain why the lateral strength
in the other direction of load did not decrease so abruptly
(Fig. 6a and b).
Contrary to the aforementioned case, the walls M3 and M4 with
a greater horizontal reinforcement ratio ðqh ¼ 0:09%Þ and with
0
axial load (rn =f m  0:10) exhibited a different behaviour. The dete-
rioration of strength of wall M3 occurred gradually, with penetra-
tion of cracks into the toes without face-shell spalling and with a
progressive fracturing of the bed-joint reinforcement. In contrast,
the post-peak response of the wall M4 was brittle and sudden, with
simultaneous failure of several bed-joint reinforcements and a
rapid propagation and penetration of diagonal cracking into the
toes and upper ends of the walls. Given this ambiguous situation,
it was not possible to define the post-peak behaviour for this type
of walls (Fig. 6c and d).
In the case of the short walls such as M5 and M7 with aspect
ratio hef =d ¼ 0:44, the deterioration of strength was gradual in
(c) (d)
(g)
(j)
 P. Ramírez et al. / Engineering Structures 126 (2016) 598–617 605

comparison with the ordinary walls (aspect ratio hef =d ¼ 0:97).
Both types of walls characterised by having the minimum horizon-
tal reinforcement ratio (qh ¼ 0:04%) and pre-compression level
rn =f 0m of 10% and 0% respectively, showed successive fracturing
of the bed-joint reinforcements and penetration of diagonal crack-
ing into the compression toes and upper ends without a face-shell
spalling. In addition, the wall that was not subjected to axial load
(M7) showed a better distribution of fracturing and less severe
cracking than the wall that was subjected to axial load (M5). Sim-
ilarly, the wall M6 with a greater horizontal reinforcement ratio
0 The main damage caused by this mode of failure were sloped and
(qh ¼ 0:08%) and axial load rn =f m  0:10 had a gradual deteriora-
tion of strength, with penetration of cracks into the toes without
face-shell spalling and a better distribution of cracks and less sev-
ere cracking. This reflects the beneficial effect in controlling dam-
age when a greater horizontal reinforcement ratio is placed in
the walls. (Fig. 6e–g).
With regard to the slender walls (hef =d ¼ 1:95Þ, it was possible
to observe that the propagation of damage and the deterioration of
strength in one of the walls (M8) with minimum horizontal rein-
0
forcement ratio (qh ¼ 0:04%) and with axial load (rn =f m  0:10),
was rapid and sudden for each of the directions of load. The cause
of this was the sudden progression in size and thickness of severe
diagonal cracks in each direction of load that went through the
wall almost completely, totally penetrated the compression toes,
caused face-shell spalling and simultaneously fractured several
bed-joint reinforcements (Fig. 6h). In contrast, the deterioration
of the strength and the behaviour of the other slender wall
(M10) with the same characteristics but without axial load
0 In those areas where the vertical reinforcement intercepted, it
(rn =f m ¼ 0) was gradual. This wall showed consecutive fractures
of some bed-joint reinforcements and diagonal cracks in both
directions of load that only partially penetrated the toes of the
wall. Although the failure of this latter wall can be considered as
brittle, it was not explosive but occurred in a gradual and con-
trolled way (Fig. 6j).
The behaviour and subsequent failure of wall M9, which had a
greater horizontal reinforcement ratio (qh ¼ 0:09%) and axial load
0 viour, Fig. 8 shows, as an example, the evolution and propagation
(rn =f m  0:10), exhibited a gradual degradation of strength. The
diagonal cracking partially penetrated the toes and upper ends of
the wall. Lateral face-shell spalling was more evident in the area
between the first row and up to half the height of the wall. Unlike
the other slender walls, this wall did not show breaking of the
ladder-type reinforcements during the test. As above, a positive
effect on damage control by placing above-the-minimum horizon-
tal reinforcement ratios could be seen (Fig. 6i).
3.2. Failure modes and damage progression
As anticipated, all the walls tested failed due to diagonal shear.
stepped cracking, penetration of cracks into the upper ends and
toes of the walls, fracture of the bed-joint reinforcements, face-
shell spalling, cracking in the grout, and lateral deformation of
the vertical bars. Fig. 7 shows a summary of the main damage
observed during the tests.
The behaviour and failure of the walls were controlled by diag-
onal cracking in both directions, which was propagated in a slanted
way going through the units and mortar, or in a stepped way
through the bed and head joints or also, as a combination of both
types of propagation. The diagonal cracking was caused by the
principal stresses, which exceeded the tensile strength of the
masonry, and the stepped cracks was caused by high shear stresses
that exceeded the bond strength between the units and the mortar.
In the case of the bed-joint reinforcements that failed because of
direct tension, it could be observed that the breaking of the ladder
reinforcements occurred mainly in the areas where major diagonal
cracking had become evident and not throughout the whole wall.
could also be observed that a seriously large and thick diagonal
crack, the vertical bars showed lateral deformation as a result of
the relative movement experienced by the parts where the diago-
nal crack divided the wall (dowel effect).
With regard to the propagation of damage, all the walls suffered
the same sequence and evolution of damage until reaching the
maximum strength. To illustrate and describe this general beha-
of the damage that occurred in wall M4. As can be seen, the dam-
age began with the formation of slight, slanted and stepped cracks

Fig. 7. Typical damage observed during the tests: (a) sloped and stepped cracking, (b) penetration of cracking into upper ends and toes, (c) fracture of bed-joint
reinforcements, (d) face-shell spalling, (e) cracking of grout column, and (f) lateral deformation of vertical bars.
606 P. Ramírez et al. / Engineering Structures 126 (2016) 598–617
Fig. 8. Damage progression in wall M4: (a) slight slanted and stepped cracking, (b) increase and distribution of cracking, (c) initiation of cracking in grouted cells, (d)
penetration and extension of cracking in toes and upper ends, and (e) collapse.
in those areas where there were no vertical bars. These cracks
developed more generally in the area between the foot of the wall
and half its height (Fig. 8a). As the lateral deformation increased,
new cracks appeared and were distributed progressively across a
diagonal strip. The density of the cracks and the space between
them were dependent on the horizontal reinforcement ratio, i.e.
those walls with greater horizontal reinforcement ratio exhibited
a greater density and a better distribution of cracking (Fig. 8b).
By imposing further lateral displacement, the length and width
of the cracks increased gradually forming continuous diagonal
cracks, which spread and partially penetrated the masonry units
filled with grout. In this series of cracks, one diagonal crack of con-
siderable thickness and size was formed, which finally predomi-
nated and defined the final cracking pattern of the wall.
In several walls, it was observed that this severe diagonal crack
was more predominant in one direction of load than in the other
(Fig. 8c). Subsequently, the diagonal cracks gradually spread to
the toes and upper ends of the walls until the maximum strength
was reached. In this situation, the diagonal cracks did not pene-
trate the toes and upper ends of the wall completely and no
face-shell spalling became evident (Fig. 8d). Fig. 9 shows the crack-
ing pattern observed for each of the walls under the condition of
maximum lateral load. After of this state of damage, the evolution,
propagation and extension of the damage depended on the partic-
ular characteristics of each wall. Finally, it is important to note
that, according to measurements of the strain gauges, the yielding
point of the vertical reinforcing bars was not reached in any of the
walls.
4. Analysis of structural parameters
This section focuses on the influence of each design variable
under study (i.e. aspect ratio, shear reinforcement ratio and level
of axial pre-compression) on different structural parameters such
as degradation of stiffness, shear strength, displacement ductility,
dissipation energy, hysteretic damping and level of drift. In gen-
eral, the analyses are performed in groups to show the effect of a
variable in particular. Concerning the presence of the vertical rein-
forcement, the results indicate that their contribution depends on
the failure mode developed in the walls. In walls failing in shear, a
slight beneficial effect on the shear strength has been recognized
[35,8]. However, the benefit is not straightforward due to the pre-
dominance of diagonal cracking for medium to high levels of verti-
cal pre-compression [11]. In addition, a greater vertical
reinforcement ratio does not seem to affect the displacement duc-
tility or the stiffness of the walls [28]. Therefore, when a wall fails
in shear, the main contribution of the vertical reinforcement is
related to a better resistance to sliding and a more effective control
of diagonal cracking that in turn improves its post-cracking
performance.
4.1. Degradation of stiffness
The degradation of stiffness that a wall undergoes when it is
subjected to in-plane lateral cyclic loads depends on the level of
damage that it is going to experience during its cyclic response.
With the aim of evaluating the degradation of stiffness of the
PG-RCMSW up to a level of deformation that corresponds to the
maximum strength, the relation K s;i =K o in function of drift (D) is
represented graphically for a set of walls with different aspect
ratios, horizontal reinforcement ratios and levels of axial pre-
compression as is shown in Fig. 10. The term K s;i corresponds to
the secant stiffness for each one of the cycles of displacement
imposed on the wall, K o represents the initial stiffness of the spec-
imen in the first cycle of load corresponding to a lateral displace-
ment of 0.20 mm, and D represents the drift or relation between
 P. Ramírez et al. / Engineering Structures 126 (2016) 598–617 607

Fig. 9. Cracking patterns of PG-RCMSW in state of maximum load.

608 P. Ramírez et al. / Engineering Structures 126 (2016) 598–617

100 in degradation of stiffness than the slender walls during their

MBH-1(+)
MBH-1(-) whole response cycle. In this analysis, an influencing factor could
MBH-2(+)
MBH-2(-)
also be the high level of vertical reinforcement ratio present in
80 MBH-3(+) the slender specimens. In any case, the stiffness degradation rate
MBH-3(-)
MBH-4(+) seems not to be affected by the horizontal spacing between vertical
MBH-4(-)
MBH-5(+)
bars Dhanasekar and Haider [5] nor by different vertical reinforce-
Ks,i/Ko [%]

60 MBH-5(-) ment ratios [35] for walls with a similar aspect ratio. With regard
MBH-6(+)
MBH-6(-) to the effect of the horizontal reinforcement ratio (Fig. 11d–f) and
MBH-7(+)
MBH-7(-)
the level of axial pre-compression (Fig. 11g and h), no marked
40 MBH-8(+) influence was observed in the degradation of stiffness.
MBH-8(-)
MBH-9(+)
MBH-9(-)
MBH-10(+) 4.2. Shear strength
20 MBH-10(-)

The shear strength of the walls was expressed as the ratio of the
tangential shear stress (mn ) and the square root of the compressive
0 0
0 0.25 0.5 0.75 1 1.25 1.5 strength of the masonry prism (f m ). It should be noted that this
Δ [%] notation, also used by Matsumura [22] and Voon and Ingham
[39], allows that the variations introduced by the net area ðAnv Þ
0
Fig. 10. Stiffness degradation of the walls tested. and by the quality of the masonry ðf m Þ to be excluded from the test
qffiffiffiffiffiffi
0
results. Consequently, the term mn = f m was correlated with the
lateral displacement of the walls according to aspect ratio, horizon-
the maximum lateral deformation reached per cycle at the inflec-
tal reinforcement ratio and level of axial pre-compression [31].
tion point of the wall and its effective height (hef ). The secant stiff-
Fig. 12a indicates that as the aspect ratio increases, the shear
ness (K s;i ) was calculated as,
strength decreases. This finding has also been reported by Mat-
DV max sumura [22] and Voon and Ingham [39]. Although, in this case,
K s;i ¼ ð1Þ the tendency is even more noticeable for short walls
Ddmax
(hef =d ¼ 0:44) due to the presence of a larger number of vertical
where DV max is the difference between the maximum loads in a hys- bars compared to other aspect ratios studied. However, atypical
teresis cycle and Ddmax is the difference between the deformations cases were observed in two groups of walls. In the first group, a
associated with the maximum loads for the same cycle of short wall with three vertical reinforcing bars in its central region
hysteresis. had a lower strength than an ordinary wall with two vertical bars
As can be seen in Fig. 10, the secant stiffness of all the walls (Fig. 12b). In the second group, a short wall with three vertical steel
declined rapidly during the first levels of deformation and there- bars in its central region had a lower strength than a slender wall
after continued falling more gradually until the end of the tests. with a single vertical bar (Fig. 12c).
The shape of the function of the stiffness degradation of each one On the other hand, the shear strength increased in accordance
of the walls was appropriately adjusted to a power function con- with an increase in the horizontal reinforcement ratio, but which
firming what has been reported by Tomazevic [36], whose power was a more marked rise in the case of the ordinary and slender
function used to relate the degradation of stiffness to the drift walls than in that of the short walls (Fig. 12d–f). Similarly, the
resulted in Eq. (2), effect of axial pre-compression favoured an increase in shear

 strength, where a greater influence was seen in the short walls
K s;i
¼ aðDÞb ð2Þ than in the slender ones (Fig. 12g and h). Table 5 summarises for
K0
each one of the walls tested the maximum strength values in terms
qffiffiffiffiffiffi
where a and b are the parameters of the stiffness degradation. The 0
to the ratio mn = f m , which varies between 0.25 and 0.46.
values of these parameters together with the correlation factor R2
are given in Table 3 for all the walls tested.
The resulting stiffness degradation-drift relationships for the 4.3. Ductility
ten tested walls, grouped according to the design parameters con-
sidered in this study, are presented in Fig. 11. As can be observed, Ductility is defined as the parameter that measures the ability
the degradation of stiffness is mainly affected by the aspect ratio of of an element to resist repeated reversed cycles of inelastic defor-
walls. From the results shown in Table 4 and from Fig. 11a–c one mation without significant degradation of strength [27]. A com-
can see that the ordinary and short walls experienced a greater fall monly used approach to estimate the ductility is to idealize the
experimental load-displacement envelope through a bilinear enve-
lope. Slightly different approaches can be found in the literature for
Table 3
evaluating the parameters defining this idealized envelope. See
Stiffness degradation parameters and correlation factor R2 of the walls tested.
Shedid et al. [33] for a wider discussion. In Fig. 13, a possible def-
Wall a b R2 inition of the parameters of the bilinear envelope is given. The
M1 0.12 0.45 0.82 approach adopted here is based on methods suggested by Lüders
M2 0.03 0.72 0.88 and Hidalgo [20] and El-Dakhakhni et al. [7]. In this way, the dis-
M3 0.11 0.54 0.97
placement ductility ðlMR Þ was established as the relationship
M4 0.05 0.61 0.96
M5 0.06 0.61 0.94 between the last displacement (dMR ) associated with the point
M6 0.06 0.61 0.95 where maximum strength was reached and the idealized elastic
M7 0.04 0.72 0.97 displacement (dE ) as is shown in Eq. (3):
M8 0.15 0.49 0.97
M9 0.14 0.50 0.93 dMR
M10 0.17 0.46 0.94 lMR ¼ ð3Þ
dE
 P. Ramírez et al. / Engineering Structures 126 (2016) 598–617 609

(a) (b) (c)

(d) (e) (f)

(g) (h)

Fig. 11. Stiffness degradation according to different design parameters: (a–c) walls with different aspect ratios; (d–f) walls with different horizontal reinforcement ratios;
(g and h) walls with different axial pre-compression levels.

Table 4 way that it went through the point of maximum strength

Effective stiffness results (K s;i =K o ) for three aspect ratios. (V MR ; dMR ) and made the amount of energy under the bilinear
Walls Drift (%) envelope the same as the energy under the experimental envelope.
0.00 0.05 0.10 0.20 0.30 0.40 0.50
In agreement with this criterion, the point (V E ; dE ) was located
where the two segments intersected and where it was believed
Short walls ðhef =dÞ ¼ 0:44 1.00 0.44 0.27 0.16 0.11 0.08 0.07
that the elastic behaviour ended. The decision to define the end
Ordinary walls ðhef =dÞ ¼ 0:97 1.00 0.47 0.31 0.21 0.16 0.12 0.10
Slender walls ðhef =dÞ ¼ 1:95 1.00 0.70 0.53 0.38 0.29 0.24 0.20
of the curve as the maximum strength is because the majority of
the tested walls showed a very quick strength degradation after
reaching this level of lateral load.
On the other hand, the theoretical elastic stiffness (K E;theorical ) for
each one of the specimens tested was calculated based on the
In order to construct the bilinear envelope shown in Fig. 13, a bending theory of beams. In this calculation the behaviour of can-
straight-line segment was drawn initially from the origin of the tilever walls, the average mechanical elastic properties of prisms
curve up to a point where it was estimated that the elastic beha- and wallettes measured in the laboratory, and the contribution of
viour would end (V SL , dSL ). This point corresponded to the level of the components of deformation owing to shear und flexion were
displacement where the first diagonal cracks on the surface of considered. The formula used is shown in Eq. (4),
the wall were visually detected and also where a remarkable
change in the slope of the tangent of the envelope curve with 1
K E;theorical ¼ ð4Þ
h3ef 1:2h
respect to the slope of the tangent measured at a level of deforma- þ Gm Anefv
3Em In
tion of 0.2 mm was observed. Using this criterion, the experimental
secant stiffness (K E Þ was calculated as the relation between V SL and where Em and Gm are the average moduli of elasticity and shear of
dSL . It should be noted that El-Dakhakhni et al. [7] have recently the masonry, both refer to the net area, In is the moment of inertia
used a similar approach. As second segment of the bilinear ideal- of the net section of the uncracked wall, hef is the effective height of
ization, a straight line was chosen with a slope (K R ) in such a the wall, and Anv is the net area of the wall.
610 P. Ramírez et al. / Engineering Structures 126 (2016) 598–617

(a) (b) (c)

(d) (e) (f)

(g) (h)
Fig. 12. Shear strengths: (a–c) walls with different aspect ratios; (d–f) walls with different horizontal reinforcement ratios; (g and h) walls with different axial pre-
compression levels.

Table 5
Maximum shear strength and total energy dissipation of the walls.

Wall Variables

hef =d qh rn =f 0m pmffiffiffiffi
n
fm0
ET (kN-mm)

M1 0.97 0.04 0.10 0.33 8320

M2 0.97 0.04 0.10 0.32 10,209
M3 0.97 0.09 0.10 0.45 20,366
M4 0.97 0.09 0.10 0.46 27,594
M5 0.44 0.04 0.10 0.36 9536
M6 0.44 0.08 0.10 0.40 8750
M7 0.44 0.04 0.00 0.25 4575
M8 1.95 0.04 0.10 0.29 6480
M9 1.95 0.09 0.10 0.37 31,752
M10 1.95 0.04 0.00 0.28 10,071

Table 6 summarises the results obtained from the bilinear ide-

alization of the walls tested in agreement with the procedure
described above. From these results, it can be seen that the theo-
Fig. 13. Bilinear idealization diagram.
retical stiffness in all cases exceeded the experimental stiffness,
indicating that using the theoretical equation based on elastic
properties and exclusively for isotropic and homogenous materials
overestimates the value of elastic stiffness. In relation to the ductil- ular characteristics of each wall (aspect ratio, steel reinforcement
ity, it could be observed that it varied over a wide range of values ratio, etc.) and on the loading conditions, some walls have a more
between 2.85 and 7.94, which means that depending on the partic- ductile mode of failure due to diagonal shear stress than others.
 P. Ramírez et al. / Engineering Structures 126 (2016) 598–617 611

Table 6
Summary of the results of bilinear idealization.

Wall Direction Load V SL (kN) dSL (mm) V E (kN) dE (mm) V MR (kN) dMR (mm) K E;exp (kN/mm) K E;av er:exp (kN/mm) K E;theorical (kN/mm) lMR
M1 + 128.29 1.56 182.96 2.23 221.08 9.16 82.21 66.64 205.36 4.11
– 126.26 2.47 164.94 3.23 211.67 9.22 51.06 2.85
M2 + 68.31 0.58 99.10 0.95 196.34 6.72 118.20 108.77 205.36 7.07
– 102.14 1.03 145.58 1.55 220.85 9.01 99.34 5.81
M3 + 126.51 1.00 189.33 1.60 283.58 10.20 126.68 96.85 205.36 6.38
– 160.09 2.39 228.06 3.19 303.14 11.34 67.02 3.55
M4 + 136.10 1.94 294.84 3.08 308.32 14.26 70.02 98.08 205.36 4.63
– 165.32 1.31 240.63 1.82 291.55 11.22 126.14 6.16
M5 + 160.75 0.52 238.62 0.84 316.35 3.11 310.77 288.84 686.93 3.70
– 204.43 0.77 249.46 0.97 290.14 6.67 266.91 6.88
M6 + 239.88 0.71 308.29 0.88 330.38 2.72 335.79 309.52 686.93 3.09
– 218.39 0.77 292.56 1.07 342.41 5.34 283.24 4.99
M7 + 133.68 0.63 153.82 0.72 199.86 5.64 213.29 221.60 686.93 7.83
– 159.41 0.69 183.75 0.80 221.78 4.24 229.90 5.30
M8 + 53.15 1.98 72.94 2.59 99.21 9.28 26.78 32.67 51.49 3.58
– 42.54 1.10 55.01 1.63 100.31 10.53 38.55 6.46
M9 + 50.38 1.81 75.30 3.36 116.49 26.71 27.79 23.79 51.49 7.95
– 64.79 3.28 85.33 4.42 141.68 23.02 19.78 5.21
M10 + 47.56 1.88 69.37 2.77 99.89 14.18 25.25 24.34 51.49 5.12
– 43.35 1.85 72.56 3.07 98.43 19.44 23.43 6.33

Note: Direction Load as in Table 5.

(a) (b) (c)

(d) (e) (f)

(g) (h)

Fig. 14. Idealized ductility curves: (a–c) walls with different aspect ratios; (d–f) walls with different horizontal reinforcement ratios; (g and h) walls with different axial pre-
compression levels.
612 P. Ramírez et al. / Engineering Structures 126 (2016) 598–617
In order to evaluate the effects of the three experimental design
variables on the ductility, the bilinear envelopes were superim-
posed on a series of walls with different aspect ratios, horizontal
reinforcement ratios and axial pre-compression as is illustrated
in Fig. 14. From Fig. 14a–c, it can be seen that the aspect ratio
did not have an obvious effect on the ductility to several groups
of walls with similar horizontal reinforcement ratios and axial
load. Although one would have expected that with a greater aspect
ratio the walls would have shown a greater level of ductility, as has
been reported by Vaughan [38], such behaviour did not manifest
itself clearly in the present results. This finding can be observed
in Fig. 14a and c, where slender walls with hef =d ¼ 1:95 show a
lower values of lMR in comparison to short walls with
hef =d ¼ 0:44. In addition, though in some walls an increase in the
level of ductility as a result of an increase in the horizontal rein-
forcement ratio could be observed (Fig. 14e and f), neither was a
clear behaviour for this effect displayed (Fig. 14d).
In the same way, although the influence of axial pre-
compression led to lower levels of ductility for one direction of
load (East), in the other direction this same behaviour could not
be observed (Fig. 14g and h). It should be stressed therefore, that
from the point of view of ductility the behaviour was generally
asymmetric as can be seen by the imbalance in the majority of
(a) (b)
(d) (e)
(g)
Fig. 15. Energy dissipation: (a–c) walls with different aspect ratios; (d–f) walls with different horizontal reinforcement ratios; (g and h) walls with different axial pre-
compression levels.
the ductility values obtained for both directions of load. On the
other hand, and as was presented at the beginning of this section,
the horizontal spacing between vertical reinforcing bars would
have a limited but beneficial effect on the deformability of the
specimens when a shear failure mode occurs [11]. In any case, a
greater amount of experimental data is needed to support any con-
clusion in this regard.
4.4. Energy dissipation
In this study, the total capacity of energy dissipation (ET ) was
measured up to the level of maximum strength. This is because
the majority of the specimens exhibited unstable cycles that led
to sudden and brittle failure after reaching maximum strength.
Owing to the fact that for each level of displacement imposed
two load cycles were applied, the energy dissipated for each level
of displacement (Es;i ) was calculated as the summation of the areas
enclosed by the two respective hysteretic cycles. Table 5 sum-
marises the capacity of energy dissipation developed by each one
of the walls. Notice that, according to Sierra [35], the horizontal
spacing of vertical steel bars has no significant effect on energy
dissipation.
(c)
(f)
(h)
 P. Ramírez et al. / Engineering Structures 126 (2016) 598–617
In order to evaluate the effect of the three experimental design
variables on the dissipation of energy, the relationship Es;i versus
drift (D) is represented graphically for a group of walls with differ-
ent aspect ratios, horizontal reinforcement ratios and axial pre-
compression, as is shown in Fig. 15. From these graphs, it can be
seen that the amount of energy dissipation was influenced by the
aspect ratio (Fig. 15a–c). In this case, the slender walls required a
greater level of deformation and a greater number of hysteretic
cycles to dissipate the same amount of energy than the short and
the ordinary walls. This would explain why the short and the ordi-
nary walls concentrated a greater dissipation of energy during
minimal demands of seismic displacements and therefore suffered
greater structural damage than the slender walls.
On the other hand, it can be seen that the increase in the hori-
zontal reinforcement ratio increased the capacity of energy dissi-
pation in the ordinary and slender walls, while for the short
walls such increase did not affect the energy dissipation capacity
(Fig. 15d–f). In addition, it could be observed that the energy dissi-
pation capacity in the ordinary and slender walls increase from an
approximate drift value of 0.45%, that is, when considerable diago-
nal cracking had already manifested itself in the specimens. This
would mean that the contribution of the horizontal steel in
strength, ductility and energy dissipation was developed from
some state of advanced damage in the wall.
(a) (b)
(d) (e)
(g)
Fig. 16. Equivalent viscous damping: (a–c) walls with different aspect ratios; (d–f) walls with different horizontal reinforcement ratios; (g) and (h) walls with different axial
pre-compression levels.
613
Regarding the level of axial pre-compression, the results show
that the action of the axial load increased the capacity of energy
dissipation in the short and slender walls. This increase is more
noticeable in short walls in comparison with slender walls if a
same level of drift is considered (Fig. 15g and h). As an atypical
case, it was found that wall M5, despite improving the capacity
of energy dissipation by applying an axial load, did not reduce its
level of deformation for one of the directions of load (Fig. 15g).
4.5. Equivalent viscous damping
One of the simplest ways of representing the real damping of
structures is through equivalent viscous damping ratio. In this
study, the technique of setting the energy dissipated from each
one of the hysteretic cycles equal to the energy dissipated from a
system with linear viscous damping is used to determine the
equivalent viscous damping ratio neq;i , i.e.:,
Es;i
neq;i ¼ ð5Þ
2  p  K s;i  d2max;i
where Es;i represents the energy dissipated for a load cycle, K s;i is the
secant stiffness of the load cycle, and dmax;i is the maximum dis-
placement in the load cycle.
(c)
(f)
(h)
614 P. Ramírez et al. / Engineering Structures 126 (2016) 598–617

In order to evaluate the effect of the three experimental design
variables on the equivalent damping ratio, the relationship neq;i is
represented graphically according to the level of drift (D) for a
group of walls with different aspect ratios, horizontal reinforce-
ment ratios and levels of axial pre-compression as is shown in
Fig. 16. It can be seen that the equivalent damping ratio at the
beginning of the cyclic response exhibited high values, which ran-
ged from 9% to 23%. Subsequently, once the walls made incursions
into the non-linear range, the damping values tended to stabilise
within a range that varied between 5% and 11%. This stabilisation
occurred for different levels of drift in a non-linear regimen.
From the graphs in Fig. 16a–c, it is evident that damping tended
to be greater for the slender walls than for short and ordinary
walls. This trend could be observed when the walls developed

(a) (b) (c)

(d) (e) (f)

(g) (h)

Fig. 17. Drift: (a–c) walls with different aspect ratios; (d–f) walls with different horizontal reinforcement ratios; (g and h) walls with different axial pre-compression levels.

Table 7
Summary of selected shear strength expressions.

Reference Shear equation Eq.

" !qffiffiffiffiffiffiffiffi #
Matsumura [22] qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 6
0
0


V n1 ¼ ku kp 0:76


ðhe f =d Þþ0:7
þ 0:012 f m þ 0:2ro þ 0:18cd qh f yh f m  ð0:875td Þ


kp ¼ 1:16q0:3v e ; qv e ¼ at =t:d 7
qffiffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffi
Tomazevic [36] ft
2 0
V n2 ¼ Aw b rf a þ 1 þ C rh ð0:90d Asrh f yh Þ þ 0:806n drv f cr f yv 8
t

V n3 ¼ V m þ V s 9


NTCM [29] V m ¼ F R ð0:5m m Awh þ 0:3PÞ 6 1:5F R
m Awh ;
m Ah
V s ¼ F R g f yh Awh
st
10
V n4 ¼ cg ðV m þ V s Þ 11
h  i qffiffiffiffiffiffiffiffi  
MSJC [25] 0

12
V m ¼ 0:083 4:0  1:75 Vd M

Anv f m þ 0:25P; V s ¼ 0:5 Asrh f yh d

qffiffiffiffiffiffiffiffi
qffiffiffiffiffiffiffiffi

0

0
13
where M=Vd 6 0:25 V n 6 cg 0:498Anv f m ; where M=Vd P 1:00 V n 6 cg 0:332Anv f m
 P. Ramírez et al. / Engineering Structures 126 (2016) 598–617 615

levels of drift between 0.1% and 0.4%. With regard to the influence
of the horizontal reinforcement ratio, it could be seen that an
increase in this value led to an increase in damping in all the walls.
However, this increase occurred when the short and the ordinary
walls had levels of drift of around of 0.1%, while that in the case
of the slender walls it occurred when the levels of drift were
greater than 0.4% as is shown in Fig. 16d–f. This behaviour was
due to the fact that the capacity of the horizontal reinforcement
for this type of walls developed partially or completely from a con-
dition of significant diagonal cracking, which happened earlier in
the short and the ordinary walls than in the slender ones. As con-
sequence, an increase in the dissipation of energy and also in an
increase in equivalent viscous damping were observed. On the
other hand, an increase in axial pre-compression led to an increase
in equivalent damping ratio, the effect of which was more pro-
nounced in the slender walls than in the short walls (Fig. 16g
and h). In the case of one slender wall, the increase in the damping
by axial load developed from the beginning of the application of
the lateral load, while in one short wall the rise in damping devel-
oped from a higher level of drift. Again, according to Sierra [35], the
equivalent viscous damping is shown independent of the number
of vertical bars present in a wall.
4.6. Drift
Drift is one of the most important parameters in the seismic
design of structures. In this study, and as has already been men-
tioned, all the PG-RCMSW were designed based on Code
NCh1928.Ofc.93 and all of them exhibited a diagonal shear failure
mode. The drift (D) was obtained for each one of the displacements
imposed as the relation between the horizontal displacement mea-
sured at the height of the inflection point of the wall and its effec-
tive height (hef ). The horizontal displacement included the
deformations owing to shear and to flexion but not the displace-
ments owing to slipping and rocking, which were disregarded for
this analysis. It is also important to mention that, because the spac-
ing between vertical reinforcing bars has a limited influence on
drift level [35,28], this parameter has not been considered in the
analysis.
Fig. 17 summarises the envelope curves that relate the lateral
load with the level of drift for a group of walls according to aspect
ratio, horizontal reinforcement ratios and axial pre-compression.
In these graphs, the drift was considered up to the level of maxi-
mum strength. As can be seen in Fig. 17b and c, the drift increased
as the aspect ratio increased. However, this effect was not very

4 Matsumura(1988) Vexp / Vn(+) 4 Tomasevic(1999) Vexp / Vn(+)

Min =0.51 Min =0.56
V / V (-) V / V (-)
Mean =0.69 exp n Mean =0.83 exp n
3.5 3.5
Max=0.77 Max=1.08
σstd =0.07 σ std =0.18
3 Cov=0.10 3 Cov=0.21

2.5 2.5
Vexp /Vn

Vexp /Vn

2 2

1.5 1.5

1 1

0.5 0.5

0 0
MBH1 MBH2 MBH3 MBH4 MBH5 MBH6 MBH7 MBH8 MBH9 MBH10 MBH1 MBH2 MBH3 MBH4 MBH5 MBH6 MBH7 MBH8 MBH9 MBH10
(a) (b)
Wall Wall

4 NTCM(2004) Vexp / Vn(+) 4 MSJC(2013) Vexp / Vn(+)

Min =1.00 Min =0.95
V / V (-) V / V (-)
Mean =1.21 exp n Mean =1.71 exp n
3.5 3.5
Max=1.43 Max=3.31
σstd =0.12 σstd =0.67
3 Cov=0.10 3 Cov=0.40

2.5 2.5
Vexp /Vn

Vexp /Vn

2 2

1.5 1.5

1 1

0.5 0.5

0 0
MBH1 MBH2 MBH3 MBH4 MBH5 MBH6 MBH7 MBH8 MBH9 MBH10 MBH1 MBH2 MBH3 MBH4 MBH5 MBH6 MBH7 MBH8 MBH9 MBH10
(c) (d)
Wall Wall

Fig. 18. Comparison of experimental vs analytical results: (a) Matsumura [22], (b) Tomazevic [36], (c) NTCM [29] and (d) MSJC [25].
616 P. Ramírez et al. / Engineering Structures 126 (2016) 598–617
marked for walls with a minimum horizontal reinforcement ratio
and with axial load (Fig. 17a). Similarly, Fig. 17e and f shows that
the drift increased as the horizontal reinforcement ratio increased.
Nevertheless, this effect was less influential in the short walls with
different horizontal reinforcement ratios and with axial load
(Fig. 17d).
In the same way, from Fig. 17h it is evident that the drift tended
to decrease in the case of the walls with axial load. However, this
effect was less pronounced in the short walls with the minimum
horizontal reinforcement ratio (Fig. 17g). From these results, it is
clear that the variation of drift is hardly appreciable in the walls
with a low aspect ratio (hef =d ¼ 0:44) regardless of the horizontal
reinforcement ratio placed in the walls (0.04–0.09%) and the
0 2. All the walls exhibit a similar propagation of damage until the
applied axial load (0–0.10f m ).
5. Experimental versus analytical results
In this section, the results of maximum shear strength obtained
from the ten walls tested are compared with some predictions of
strength reported in the literature. The expressions selected to pre-
dict the maximum shear strength were the equations proposed by
Matsumura [22], Tomazevic [36] and expressions recommended
by the Mexican Code [29] and by the Masonry Standards Joint
Committee’s Building Code [25]. It is important to stress here that
expressions prescribed by the Code NCh1928.Of93 [17] were not
considered in this comparison owing to the fact that its formulae
are based on allowable stress. Table 7 shows the expressions
selected for this comparison and Fig. 18 shows their respective
performance.
As it can be seen this comparison was carried out by means of
the ratio between the experimental strength to the predicted
strength, denoted here as V exp =V n . To illustrate the variability of
this relationship, the standard deviations and coefficients of varia-
tion are presented in the same graphs. As can be observed from
Fig. 18c and d, the predicted expressions of shear strength recom-
mended by codes are conservative, and a greater degree of disper-
sion is seen in the values estimated by the Masonry Standards Joint
Committee’s Building Code [25] than in those calculated by the
Mexican Code [29]. Equally, one can tell that the prediction given
by the expression of the MSJC [25] to estimate the maximum
strength of the slender walls (M8, M9, M10) turned out to be
highly conservative.
On the other hand, from Fig. 18a and b, it is evident that the
expressions of shear strength recommended by Matsumura [22]
and Tomazevic [36] are, in general, not conservative. The estimates
of shear strength given by the formula of Tomazevic [36] for the
short and the ordinary walls are fairly satisfactory; however, the
prediction for the slender walls (M8, M9, M10) was overestimated.
In the same way, it could be seen that the predictions given by the
formula of Matsumura [22] were unreliable overestimated in all
cases. These findings are agreement with a recent work by Aguilar
et al. [1].
6. Conclusions
An experimental study of the cyclic behaviour of ten partially
grouted reinforced concrete masonry shear walls (PG-RCMSW)
exhibiting a diagonal shear failure mode was carried out. The walls
were designed and built according to the Chilean reinforced
masonry standard NCh1928Of.93 [17]. The effect of three experi-
mental design variables (aspect ratio, horizontal reinforcement
ratio, and axial pre-compression) on the cyclic behaviour measured
in terms of stiffness degradation, shear strength, displacement
ductility, dissipation of energy, hysteretic damping, and drift was
analysed and discussed. The main findings and conclusions pre-
sented in this paper can be summarized as follows:
1. Owing to the fact that all the specimens tested were heavily
reinforced in the vertical direction to prevent failure by flexion
or slipping, all the walls failed as a result of shear stress. This
failure mechanism exhibits diagonal cracking, breakage of hor-
izontal rods, face-shell spalling, and lateral deformation of ver-
tical rods. It is important to remark that the contribution of the
vertical reinforcement was tentatively analysed in terms of the
main findings reported by some past investigations. However, a
greater amount of experimental data is needed to support any
conclusion in this regard.
maximum strength was reached. From this stage, the evolution,
propagation and extension of the damage depended on the
characteristics and particular loading conditions of each wall.
In addition, it was observed that the strength degradation is
more severe in some walls than others. Therefore, to foresee
the seismic behaviour of this type of walls beyond their maxi-
mum strength seems to be uncertain and complex.
3. The design variable that predominantly influenced the stiffness
degradation was the aspect ratio. It could be seen that the ordi-
nary and the short walls experienced a stronger decrease in
stiffness than the slender walls during the cyclic response.
The effect of the horizontal reinforcement ratio and axial pre-
compression did not appear to have an appreciable influence
on the stiffness degradation.
4. It could be seen that as the aspect ratio increased, the shear
strength decreased. However, two cases were found to be atyp-
ical: one, where a short wall exhibited less strength than an
ordinary one and another case where a short wall was less
strong than a slender one. Since these results were not reflected
in other studies, further experimental research is needed to
clarify this behaviour. On the other hand, an increase in the hor-
izontal reinforcement ratio led to an increase in shear strength,
where a greater influence was observed in the ordinary and the
slender walls than in the short ones. Similarly, the application
of axial pre-compression favoured an increase in shear strength,
an effect that was more pronounced in the short walls than in
the slender walls.
5. With regard to ductility, there was no clear effect of the three
design variables considered; in fact, a rather asymmetric beha-
viour was seen for both loading directions. Furthermore, the
ductility for the group of walls studied was found to fluctuate
over a wide range of values, i.e. between 2.85 and 7.94. This
would mean that depending on the particular characteristics
of each wall (aspect ratio, steel reinforcement ratio, etc.) and
on the loading conditions some walls were more ductile than
others when exposed to failure by diagonal shear.
6. On the other hand, it could be established that the energy dis-
sipation capacity was clearly dependent on aspect ratio. With
the same level of lateral displacement, the short and the ordi-
nary walls dissipated a greater amount of energy than the slen-
der walls. Similarly, it was found that an increase in the
horizontal reinforcement ratio increased the energy dissipation
capacity of the ordinary and the slender walls. In the case of the
short walls this effect was not significant. Besides, it could be
observed that axial load led to a greater capacity of energy
dissipation.
7. Equivalent viscous damping ratio ranges between 5% and 11%. It
could be seen that the equivalent damping ratio depended on
the aspect ratio. The slender walls developed greater levels of
damping than the short and the ordinary walls. Furthermore,
an increase in the horizontal reinforcement ratio tended to
 P. Ramírez et al. / Engineering Structures 126 (2016) 598–617
increase the level of damping while the application of axial load
produced a greater increase in damping in the slender walls
than in the short walls.
8. Drift was found to depend on all the three design variables con-
sidered in this study. The results showed that the level of drift
increased when the aspect ratio increased or when the horizon-
tal reinforcement ratio also increased. However, drift was
reduced when the level of pre-compression rose. Nevertheless,
the walls with low aspect ratios showed very little sensitivity to
the variation of drift when the horizontal reinforcement ratio
changed from 0.04% to 0.09% or when the levels of axial load
0 [15] INN. Instituto Nacional de Normalización. Norma Chilena Oficial NCh 200-Ofc.
varied between nought and 0.10f m .
9. Finally, these results demonstrated that the expressions to pre-
dict the shear strength of the PG-RCMSW, recommended by the
[25] and by the Code [29], turned out to be conservative. How-
ever, the estimates established by the latter Code were closest
and most reliable. From these results it can be deduced that
diagonal compression stress is one of the most important
parameters that should be considered in a predictive equation.
On the other hand, these results also showed that using expres-
sions that included some degree of interrelation of variables
were non-conservative and thus require some modification in
order to be applicable in the type of walls studied.
Acknowledgments
This investigation was partly supported by the FONDEF
Research Project No. D10I1027 ‘‘Desarrollo de plataforma de ame-
naza sísmica para el norte de Chile y propuesta de actualización de
normas de diseño sismo resistente: Análisis de los efectos del ter-
remoto del Maule 2010 y caracterización de terremotos tipo en el
segmento Taltal-Arica in Chile” and by the Chilean Fondo Nacional
de Ciencia y Tecnología (Fondecyt de Iniciación) through Grant No.
11121161. The first-named author was supported by the scholar-
ship programme ‘‘Ayudante-Becario/2011-201200 of the Vicerrec-
toría de Investigación de la Pontificia Universidad Católica de
Chile and by the scholarship programme ‘‘Beca de Doctorado en
Chile para Estudiantes Extranjeros” of the Comisión Nacional de
Investigación Científica y Tecnológica – CONICYT-PCHA/
Doctorado Nacional 2013/No. 6313015.
References
[1] Aguilar V, Sandoval C, Adam JM, Garzón-Roca J, Valdebenito G. Prediction of
the shear strength of reinforced masonry walls using a large experimental
database and artificial neural networks. Struct Infrastruct Eng 2016;1–14.
http://dx.doi.org/10.1080/15732479.2016.1157824.
[2] Alcaino P, Santa-María H. Experimental response of externally retrofitted
masonry walls subjected to shear loading. J Compos Constr 2008;12
(5):489–98.
[3] Almazán JL. Comportamiento de Estructuras Antisísmicas durante el
Terremoto del Maule y su posible efecto en las Normas de Diseño Sísmico en
Chile. Paper presented at the Jornadas Sudamericanas de Ingeniería
Estructural, San Juan. Argentina; 2010.
[4] Astroza M, Moroni O, Jaramillo C. Efecto en los Edificios de Albañileria.
Mw = 8.8. In: Primera Edición editor, Terremoto en Chile, 27 de Febrero de
2010; 2012.
[5] Dhanasekar M, Haider W. Effect of spacing of reinforcement on the behaviour
of partially grouted masonry shear walls. Adv Struct Eng 2011;14(2):281–94.
[6] Drysdale RG, Hamid AA, Baker LR. Masonry structures: behavior and
design. Englewood Cliffs (NJ): Prentice Hall; 1994.
[7] El-Dakhakhni WW, Banting BR, Miller SC. Seismic performance parameter
quantification of shear-critical reinforced concrete masonry squat walls. J
Struct Eng 2013;139(6):957–73.
[8] Elmapruk JH. Shear strength of partially grouted squat masonry shear walls
Master of Science in Civil Engineering. Washington State University; 2010.
[9] Ghanem GM, Essawy AS, Hamid AA. Effect of steel distribution on the behavior
of partially grouted reinforced masonry shear walls, 6th Canadian Masonry
Symposium, 15–17 June 1992. Saskatoon, Canada: University of
Saskatchewan; 1992.
617
[10] Ghanem GM, Salama AE, Elmagd SA, Hamid AA. Effect of axial compression on
the behavior of partially grouted reinforced masonry shear walls. In: 6th North
American masonry conference, 6–9 June 1993, Philadelphia, PA; 1993.
[11] Haach V, Vasconcelos G, Lourenço P. Parametrical study of masonry walls
subjected to in-plane loading through numerical modeling. Eng Struct 2011;33
(4):1377–89.
[12] Ingham J, Davidson B, Brammer D, Voon K. Testing and codification of partially
grout-filled nominally-reinforced concrete masonry subjected to in-plane
cyclic loads. Masonry Soc J 2001;19(1):83–96.
[13] INN. Instituto Nacional de Normalización. Norma Oficial Chilena NCh 181-Ofc.
65. Bloques huecos de hormigón de cemento. Santiago, Chile; 1967.
[14] INN. Instituto Nacional de Normalización. Norma Oficial Chilena NCh 158-Ofc.
67. Cementos – Ensayo de flexión y compresión de morteros de cemento.
Santiago, Chile; 1968.
72. Productos Metálicos. Ensayo de Tracción. Santiago, Chile; 1974.
[16] INN. Instituto Nacional de Normalización. Norma Chilena Oficial NCh 1037-
Ofc. 77. Hormigón – Ensayo de compresión de probetas cúbicas y cilíndricas.
Santiago, Chile; 1977.
[17] INN. Instituto Nacional de Normalización. Norma Chilena Oficial NCh 1928-
Ofc.93. Modificada 2009. Albañilería Armada – Requisitos para el diseño y
cálculo. Santiago, Chile; 1993.
[18] INN. Instituto Nacional de Normalización. Norma Chilena Oficial NCh 2123-
Ofc.97. Modificada 2003. Albañilería Confinada, Requisitos para el diseño y
cálculo. Santiago, Chile; 1997.
[19] Lüders C, Hidalgo P. Influencia del Refuerzo Horizontal en el Comportamiento
Sísmico de muros de Albañilería Armada. Paper presented at the IV Jormadas
Chilenas de Sismología e Ingeniería Antisísmica, Viña del Mar, Chile; 1986.
[20] Lüders C, Hidalgo P. Ductilidad y Degradación de Rigidez de Muros de
Albañilería Armada. Paper presented at the XXIV Jornadas Sudamericanas de
Ingeniería Estructural. Porto Alegre, Brasil; 1987.
[21] Lüders C, Hidalgo P, Gavilán C. Comportamiento Sísmico de Muros de
Albañilería Armada. Proyecto No.70/83. Departamento de Ingenieria
Estructural. Pontificia Universidad Católica de Chile. Santiago, Chile; 1985.
[22] Matsumura A. Shear strength of reinforced masonry walls. Paper presented at
the Proc., 9th world conf. on earthquake engineering; 1988.
[23] Minaie E, Mota M, Moon F, Hamid A. In-plane behavior of partially grouted
reinforced concrete masonry shear walls. J Struct Eng 2010;136(9):1089–97.
[24] Moroni M, Astroza M, Acevedo C. Performance and seismic vulnerability of
masonry housing types used in Chile. J Perform Constr Facilities 2004;18
(3):173–9.
[25] MSJC. Masonry Standards Joint Committee. Building code requirements and
specification for masonry structures: containing TMS 402-13/ACI 530-13/ASCE
5-13, TMS 602-13/ACI 530.1-13/ASCE 6-13 and companion commentaries.
American Concrete Institute. Detroit; 2013.
[26] Muñoz WE. Estudio experimental del comportamiento de muros de albañilería
de bloques de hormigón sometidos a carga lateral alternada. (Memoria para
optar al Título de Ingeniero Civil), Universidad de Chile; 1992.
[27] NIST. Seismic design of special reinforced masonry shear walls: A guide for
practicing engineers, NIST GCR 14-917-31, prepared by the Applied Technology
Council for the National Institute of Standards and Technology. Gaithersburg
(MD): National Institute of Standards and Technology (NIST); 2014.
[28] Nolph SM, ElGawady MA. Static cyclic response of partially grouted masonry
shear walls. J Struct Eng 2011;138(7):864–79.
[29] NTCM. Normas Técnicas Complementarias para Diseño y Construcción de
Estructuras de Mampostería Gaceta Oficial del Distrito Federal. México; 2004.
[30] Nuñez M. Análisis de Daños provocados por el terremoto del 27 de Febrero del
2010 a los Edificios de Villa Cordillera Comuna de Rancagua Tesis de Ingeniería
Civil. Santiago de Chile: Universidad de Chile; 2010.
[31] Ramírez P, Sandoval C, Almazán J. Experimental assessment of shear strength
and displacement capacity of partially grouted reinforced concrete masonry
shear walls. In: 11th Canadian conference on earthquake engineering. Victoria
(BC); 2015.
[32] Santander D. Análisis de Daños de un Conjunto Habitacional provocado por el
Sismo del 13 de Junio de 2005 en Pozo Almonte, Iquique. Memoria para optar
al título de Ingeniero Civil, Universidad Catolica del Norte. Antofagasta, Chile;
2007.
[33] Shedid MT, Drysdale RG, El-Dakhakhni WW. Behavior of fully grouted
reinforced concrete masonry shear walls failing in flexure: experimental
results. J Struct Eng 2008;134(11):1754–67.
[34] Shing PB, Schuller M, Hoskered VS. In-plane resistance of reinforced masonry
shear walls. J Struct Eng 1990;116(3):619–40.
[35] Sierra GA. Estudio experimental de la influencia del refuerzo vertical en muros
de albañilería armada sometidos a carga lateral alternada. (Memoria
presentada para optar al Titulo de Ingeniero Civil mención Estructuras y
Construcción), Universidad de Chile, Santiago, Chile; 2002.
[36] Tomazevic M. Earthquake-resistant design of masonry buildings. London: Imperial
College Press; 1999.
[37] Valdebenito G, Alvarado D, Sandoval C, Aguilar V. Terremoto de Iquique
Mw = 8.2 - 01 Abril 2014: Daño observado y efectos de sitio en estructuras de
albañilería. In: XI Congreso Chileno de Sismología e Ingeniería Sísmica.
Santiago de Chile; 2015.
[38] Vaughan TP. Evaluation of masonry wall performance under cyclic loading M.
Sc. Thesis. Washington: Washington State University; 2010.
[39] Voon KC, Ingham J. Experimental in-plane shear strength investigation of
reinforced concrete masonry walls. J Struct Eng 2006;132(3):400–8.