Simplified Formulation For Modeling The Nonlinear Behavior of Confined Masonry Walls in Seismic Analysis

International Journal of Architectural Heritage
Conservation, Analysis, and Restoration
ISSN: 1558-3058 (Print) 1558-3066 (Online) Journal homepage: https://www.tandfonline.com/loi/uarc20
Simplified Formulation for Modeling the Nonlinear

Behavior of Confined Masonry Walls in Seismic
Analysis
Fariman Ranjbaran , Mahmood Hosseini & Masoud Soltani
To cite this article: Fariman Ranjbaran , Mahmood Hosseini & Masoud Soltani (2012) Simplified
Formulation for Modeling the Nonlinear Behavior of Confined Masonry Walls in Seismic Analysis,
International Journal of Architectural Heritage, 6:3, 259-289, DOI: 10.1080/15583058.2010.528826
To link to this article: https://doi.org/10.1080/15583058.2010.528826
Accepted author version posted online: 11

Oct 2011.
Published online: 20 Dec 2011.
Submit your article to this journal
Article views: 270
View related articles
Citing articles: 4 View citing articles
Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=uarc20
International Journal of Architectural Heritage, 6: 259–289, 2012
Copyright © Taylor & Francis Group, LLC
ISSN: 1558-3058 print / 1558-3066 online
DOI: 10.1080/15583058.2010.528826
SIMPLIFIED FORMULATION FOR MODELING THE

NONLINEAR BEHAVIOR OF CONFINED MASONRY
WALLS IN SEISMIC ANALYSIS
Fariman Ranjbaran,1 Mahmood Hosseini,2 and

Masoud Soltani3
1
Department of Civil Engineering, Science and Research Branch, Islamic Azad
University, Tehran, Iran
2
Structural Engineering Research Center, International Institute of Earthquake
Engineering and Seismology (IIEES), Tehran, Iran
3
Department of Civil Engineering School of Engineering, Tarbiat Modarres
University, Tehran, Iran
Past earthquakes have shown the vulnerability of masonry of buildings in several cases;
however, experimental results represent ductile behavior of structural elements of such
buildings, specifically confined masonry walls. Design of buildings of this type is usually
performed in a prescriptive manner without numerical modeling and performed out of the
framework of an analysis-designing process. The purpose of this research is to identify the
major factors affecting the behavior of the confined walls against lateral and vertical forces
and to then present a simplified behavioral model precise, as much as possible, and simple
enough to be used by professional engineers. The proposed model can show the wall behav-
ior before and after cracking. A series of nonlinear static analyses in parametric form has
been performed using a wide range of effective factors on confined masonry walls with or
without opening. Then, based on the numerical results, some simple formulas have been
proposed to express the relationships between the lateral strength of the confined wall and
the wall specifications, including the initial stiffness, the secondary stiffness after cracking,
the ultimate strength, and ductility, to be used in engineering programs such as SAP, which
are widely implemented in engineering firms, by practicing engineers.
KEY WORDS: confined masonry wall, pushover, DIANA, analytical model, simplified
model
1. INTRODUCTION
Confined masonry structures have been used in construction for more than 100 years.
This type of structure comprises masonry walls and confining elements (horizontal and
vertical ties), located at four sides of the wall panel. The confining elements are usually
made of reinforced concrete (R/C), steel profile, or timbers. Therefore the features of the
confined walls are a combination of those of unreinforced masonry walls and reinforced
Received 25 September 2010; accepted 29 September 2010.

Address correspondence to Fariman Ranjbaran, Ph.D. Student, Department of Civil Engineering, Science
and Research Branch, Islamic Azad University, Tehran, Iran. E-mail: ranjbaran_far@iiau.ac.ir
259
260 F. RANJBARAN ET AL.
concrete or other used frames. It should be noted that the performance of confining ele-
ments is not similar to the performance of common R/C beams and columns, but is
basically for creating the following features in the confined wall (Brzev 2007):
r Increasing the stability and integrity of the wall against seismic in-plane and out-of-
plane forces;
r Increasing the strength of the wall; and
r Decreasing the brittleness of the wall,
which together lead to better seismic performance of the wall.
The major difference between these walls and those with the R/C frame and infill
is that the infill walls do not act as the load-bearing elements, whereas the confined
masonry wall (CMW) is basically a load-bearing element. However, many of buildings
constructed with CMWs have shown unsatisfactory response in recent earthquakes (Brzev,
2007; Moroni et al. 2004), which indicates shortcomings in their seismic design. It is there-
fore necessary to evaluate their seismic behavior more thoroughly; however, for seismic
analysis of this type of building with high precision, either advanced finite element anal-
yses or distinct element analyses are required, which are very time consuming. Therefore
a more simplified method that is less time consuming and has acceptable precision is very
useful. This method will be acknowledged, particularly by practicing engineers for seismic
assessment of existing buildings for proposing their retrofit design. Stiffness, ductility,
and ultimate strength are three main design parameters of the wall. These parameters
depend on:
r the strengths of bricks and mortar,
r the mortar thickness,
r the dimensions of the wall and their ratios,
r the number and dimension of openings, if any, and their size ratio to the wall dimensions
in horizontal and vertical directions,
r the vertical loads on the wall,
r the amount of transverse reinforcement in the tie-end columns, as well as on
r the presence or absence of additional horizontal steel reinforcement inside the mortar
joints.
Results of experimental studies show that the elastic limit strength and the ultimate
strength of the CMWs increase with the increase of vertical loading, regardless of the
type of brick unit (Ruiz-Garcia and Negrete 2009). Experimental studies also indicate
that the elastic limit strength and the ultimate strength decrease with increase in height
to length ratio. Ductility of CMWs is defined as the ratio of displacement at the ultimate
strength to the displacement at elastic limit strength. In other words, CMW ductility is
the amount of slippage of the upper part of wall decreases with respect to its lower part
along the cracked area without any strength. Experimental observations in this regard
have shown that ductility is a function of vertical loading, height to length ratio, and
the tension and compression strengths of the wall and, finally, its boundary conditions
(Lang 2002). With decrease in the normal stresses, resulting from vertical loads, cracks
usually develop through mortar in horizontal splits rather than the body of bricks, and this
development results in considerable slippage of the wall upper portion along the crack
without any major decrease in the wall strength. This slippage occurs while the increase
in normal stresses leads to increase in frictional resistance in horizontal section of the
wall, and the tendency of crack development through brick bodies, which results in brittle
INTERNATIONAL JOURNAL OF ARCHITECTURAL HERITAGE 6(3): 259–289

NONLINEAR BEHAVIOR OF CONFINED MASONRY WALLS 261
fracture of the wall (Bruneau 1994). Furthermore, the results of experimental studies
indicate that with increase in the height-to-length ratio, due to the flexural mode, ductility
of the wall increases; conversely with decrease in the height to length ratio, the shear
fracture dominates the wall behavior, which in turn leads to reduction of the wall ductility
(Lang 2002).
Nonhomogeneity and anisotropy of the wall materials has made its modeling very
difficult, particularly when postcracking analysis is concerned, and none of the proposed
models is quite satisfactory in all aspects. Lang (2002) presented some relationships con-
cerning the capacity curve of unreinforced masonry walls. This capacity curve is in the
form of an elastic-perfectly plastic curve including primary effective stiffness and shear
capacity. Riahi et al. (2009) proposed a backbone model for simulating the seismic behav-
ior of typical confined masonry walls without opening for which the response is governed
by shear deformation. Their model was developed on the basis of limited data available
from tests and derived through linear regression analysis. Therefore, some limitation exists
in their model such as geometry of the wall, fracture mechanism, axial stress and the effect
of opening. Flores and Alcocer (1996) studied the nonlinear dynamic response of confined
masonry wall structures without opening. They presented a tri-linear force deformation
envelope curve, which is obtained from material properties and wall geometry. Teran-
Gilmore et al. (2009) presented displacement-based assessment using coefficient method
in FEMA for roof displacement demand and capacity curve according to the Flores and
Alcocer (1996). The comparison of analytical and experimental model shows that a prac-
tical method for assessment of confined masonry walls, especially after cracking, is the
push over analysis with the suitable capacity curve representing behavior of the wall and
replacing each wall with an equivalent column that flexural and shear properties of the wall
is concentrated on its centerline. Moroni et al. (1994) introduced an analytical model for
representing the seismic behavior of CMWs without opening, which is independent from
wall dimensions. Their proposed analytical model is prescriptive, and how to take into
account the effective parameters is unclear. Tomazevic and Klemenc (1998) presented an
analytical model for the CMWs, taking into account the interaction between masonry walls
without opening and the existing ties, and also the additional surcharge caused by the con-
fining effect of ties. They developed the existing relationships of the unconfined masonry
walls to CMWs. Their model was remarkably compatible with the experimental results.
In an analytical model, some of coefficients are defined in customized and practical form,
based on the damage index and primary effective stiffness of the CMWs. The main point
about the latter models is that it does not consider the existence of openings, if any, and the
effect of all main parameters, and it does not capture the post-cracking behavior precisely.
In another regard, the effect of confining elements and their interaction with masonry wall
is not considered appropriately in these models.
Lourenco et al. (1998) presented a constitutive behavior model for the unreinforced
masonry walls. In this method, masonry wall is modeled as a continuous homogenous
and anisotropic medium, with Rankine-type’s criterion is used as the yielding criteria in
tension and Hill-type’s criterion in compression. This proposed model is acceptable in
terms of precision; however, its usage is very time consuming and costly.
In past investigations, two points are considerable in the presented analytical mod-
els. First, the presented models are complicated and necessitate more time and expense,
and second, the simplified models do not fully take into account the effective parameters
in relation to the behavior of the element. Therefore, in this research, these two points
are considered together. The intent is to present an analytical model in case of confined

masonry walls with enough precision while not a complicated model. On this basis, by
performing a series of thorough analyses in this study on several samples of confined walls
with or without an opening or openings and with various surcharges and different geomet-
rical features and mechanical characteristics, a simplified analysis model is proposed in
which the interaction of wall and its confining ties and the postcracking behavior of the
wall are considered. In the detailed analyses, performed by DIANA (version 9.3) computer
program, Lourenco model (Lourenco et al. 1998) has been used for the nonlinear behavior
of wall, which takes into account the different strength in directions parallel and perpendic-
ular to bricks interfaces with mortar. More than 200 detailed models have been considered
for analyses, in which cracking in concrete of the ties, the interaction between ties and
wall, the interaction between walls and its foundation, and reinforcing steel bars in ties and
orthotropic behavior in masonry wall have been considered. Three different values ranging
from low to high have been considered for the wall length and for the amount of verti-
cal load on the wall, two values for the wall thickness, the dimensions of the opening (if
any), and the strength of masonry wall to see how these parameters affect the wall seismic
behavior. Based on the force displacement curves obtained by detailed analyses, the sim-
ple relationships between the seismic input parameters and the wall specifications have
been developed. These relationships include the initial stiffness, the secondary stiffness
after cracking, the ultimate strength, the ductility before reaching the maximum strength,
and finally the residual strength after ultimate strength. These formulas are very suitable to
be used in engineering programs that are widely used in engineering firms, such as SAP.
By using these formulas, it is possible to model the CMWs by conventional structural ele-
ments modeling. For this purpose, a beam or other similar element with the same geometric
and mechanical features of the walls substitutes for each of the CMWs and its nonlinear
behavior is expressed by the simplified formulas developed in this research.
2. STATING THE PROBLEM

It can be realized in available codes or guidelines for seismic design and evaluation
of the confined masonry buildings that their design are generally very conservative and is
mostly based on the response or reaction in the elastic zone. While the behavior of con-
fined masonry walls can be ductile, this fact may be included in some performance levels
and consequent nonlinear analyses (Ruiz-Garcia and Negrete 2009; Applied Technology
Council [ATC] 1996). One of the approaches of such analysis is the static nonlinear anal-
ysis method or push over analysis (POA). The purpose of this research is to propose
a simplified analytical model for determining the seismic capacity of confined masonry
walls and postcrack behavior and fracture mechanism. In this model, the determination of
five indices, including the initial stiffness, elastic limit, ultimate capacity, ductility value
related to final capacity, and finally the residual resistance are assumed as functions of the
effective factors. These factors include wall thickness, tension and compression strengths
of masonry unit, surcharge, dimensional ratio of wall and dimensions of opening, if any.
Using DIANA (version 9.3) software (DIANA, 2005), which has the capability of numer-
ical modeling such structural elements considering their behavior up to the facture stage,
the effort was to conduct a series of POA in parametric form on confined masonry walls of
clay bricks with or without opening. By obtaining the capacity curves based on the effective
parameters, which is in fact the backbone curve, the idealized capacity curve can be fitted
to the obtained curves (ATC 1996). In reality the earthquake action has a cyclic nature;
however, the cyclic behavior of masonry structure is very complex and not very easily

quantifiable. Some researchers have proposed a reduction factor for shear and ductility
capacity that is applied to the monotonic values for capturing cyclic behavior, allowing con-
sideration of the increased deterioration resulted from cyclic action in unconfined masonry
walls (Lang 2002). This simple approach is used for reinforced concrete structures as well.
This subject seems to need more elaboration and is beyond the scope of this discussion.
Two values of 22 and 35 cm were considered for the wall thickness, which are
common thickness values in masonry walls. Tensional strength of masonry unit (perpen-
dicular of bed joint) was considered to be tensional strength of mortar, varying from 0.03 to
0.25 Mpa, and compression strength was chosen proportional to the tension strength, vary-
ing from 5.47 to 6.23 Mpa. The wall height was assumed to be 3 m and its length 2, 3, and
4 m. The amount of surcharge was considered to be zero, 11.6, 23.2, and 34.8 N/mm. The
ratio of opening length to the wall length was assumed to be for the 2 m long wall 0.125,
0.25, and 0.375 and for the 3 m and 4 m long walls 0.25, 0.5, and 0.75, and the ratio of
opening height to the wall height was considered to be 0.17, 0.33, and 0.5. Ties were con-
sidered based on the recommendations of National Iranian Code of Practice for Seismic
Design of Buildings (Standard No. 2800), and their geometric and mechanical properties
were assumed to be the same for all cases. Horizontal and vertical ties were modeled in
the form of reinforced concrete beams with dimensions of 20×20 cm for vertical ties, and
20×20 and 20×35 cm for horizontal ties, corresponding to 22 and 35 cm walls respec-
tively. Reinforcement inside ties were assumed to comprise steel bars of 10 mm diameter
with the yielding strength of 300 Mpa. Compression strength of concrete was also assumed
to be 15 Mpa.
Using these quantities and taking into account the wall weight in loading process, and
assuming the boundary conditions of wall as a cantilever, the POA in displacement control
state were performed with a target displacement of 30 mm (0.01 h) (Moroni et al. 1994),
and the capacity curves for all of the numerical models were developed. Eventually, a
cluster of curves was created for each of these parameters. Using these curves, some simple
formulas were developed for predicting the five indices based on the effective parameters,
which are similar to FFMA 273 formulas, but have more precision.
3. MODELING
For modeling the masonry walls in DIANA (version 9.3) software, the continuum
finite element method (macromodel method) was used, which has relatively high precision
and is also more appropriate for studying the general behavior of walls, and enormous anal-
ysis cases (Lourenco et al. 1998). In this method, masonry wall is simulated in the form of
a continuous homogenized environment and Rankin-type’s criterion and Hill-type’s crite-
rion are used for expressing the inelastic behavior in tension and compression, respectively.
These models also take into account the orthotropic behavior. The value of fracture energy
in compression and tension are considered as common values in clay brick materials, which
are in the vertical direction, respectively, 0.018 N/mm in tension and 15 N/mm in com-
pression. Furthermore, the combined crack-shear-crush interaction model is believed to
be suitable for the simulating fracture caused by tension, fractional slide caused by shear
and crush resulted from compression (Lourenco 1996). Regarding the existence of friction
between the wall and its foundation after cracking, this model is also considered to be
suitable for simulating wall-foundation interaction (Mohebkhah et al. 2008).
For wall-ties-interaction the discrete crack model is used (Hashemi and Mosalam
2007). Normal and shear stresses are functions of total relative displacement (i.e., width

and slide of crack). Tension and shear strengths and stiffness after crack between tie and
wall are neglected and the behavior of element is considered in brittle type without soft-
ening. For avoiding penetration of wall and tie elements into each other, a big number is
considered for the axial stiffness in compression. For the numerical stability of interaction
element, sliding is likely, with initial shear stiffness. Beam elements with moment-resisting
connections have been used in order to model the confining elements in the software. For
modeling the concrete material of ties, the total strain-rotating crack is used (Hashemi and
Mosalam 2007). In this model, stresses are defined based on the strain-stress relation, and
the average principal strain rate is calculated and transferred to the global system of coordi-
nates of the element (Mostafaei et al. 2008). The stress-strain relation defined in modeling
for the tensile stress is the Maekawa model (Maekawa and Okamura 1990), and for the
compression stress is the perfect elastoplastic considering 28-day compression strength of
concrete. Finally, for modeling the reinforcement of ties the longitudinal bars are assumed
to have full bond with concrete around them and follow von Mises criteria with perfect
elastoplastic flow criteria (Hashemi and Mosalam, 2007).
4. VERIFICATION OF NUMERICAL MODELING

For verifying the numerical modeling, those models developed based on the behav-
iors explained in the previous section are compared with some experimental models. These
models include a double-story concrete frame without infill walls for verifying the ties
modeling (Vecchio and Basil Emara 1992), one masonry wall without confinement for
verifying masonry wall modeling (Tomazevic and Klemenc 1998), and also two confined
masonry walls for verifying the modeling of the compound system of masonry wall and ties
(Tomazevic and Klemenc 1998; Marinilli and Castilla 2004). Lateral loading was mono-
lithically applied to the numerical models, and at first the vertical loading and after that
the lateral loading was applied separately. Figures 1 to 4 shows a comparison between
the experimental results and those obtained by the developed numerical models in this
study.
In cases of the two-story concrete frame without infill walls and the masonry wall
without confinement, good agreement is shown between the numerical and experimental
results. For the cases of confined masonry walls, the numerical results underestimate the
experimental results by a ratio of approximately 1.4. This difference is believed to be due
to the interaction between ties and wall, which cannot be modeled quite realistically by
the numerical model. On this basis, in this study a coefficient of 1.4 is considered as the
calibration coefficient. This calibration is applied to cyclic behavior of experimental model,
considering the previous discussion of cyclic loading. The good agreement between the
experimental and numerical fracture mechanisms can be also seen in Figures 2 to 4, and
the same trend is observed in the capacity curves, especially after calibration (Figure 3 and
Figure 4e). However, in addition to beam elements, truss elements have also been used for
modeling the confining elements; in order to consider the low reliability of the moment
resisting connection in ties, the behavior of numerical model in this case has not been
verified by experimental model (Figure 5). In fact, using truss or beam element depends on
the rigidity of connections between horizontal and vertical ties, which is dependent of the
construction procedure and details. As expected, by increasing the connections’ rigidity,
behavior of the confinement tends to that of flexural frames; by decreasing the rigidity, the
behavior of CMWs tends to that of a truss. In the latter case, the maximum strength of a
CMW does not exceed the strength of an unconfined wall and only its ductility increases

a)
b)
Figure 1. Comparison between experimental model of Tomazevic and numerical model related to the masonry
wall with confinement (color figure available online).
to some extent. In contrast, in first case, both strength and ductility increase (more alike
to the results of aforementioned experimental study). It is believed that in real conditions
the behavior varies between the two limit states, depending on construction style, as noted
previously. However, it should be pointed out that the poor detailing of connections leads
to vulnerability of building against earthquake (Brzev 2007).
5. FRACTURE MECHANISM
In past earthquakes and also in experimental studies, three types of fracture in the
confined masonry structures have been observed, including shear fracture, flexural fracture,
and slide fracture (Ruiz-Garcia and Negrete 2009). The first fracture type (shear fracture)
occurs through bricks and mortar in the form of diagonal crack. This type of fracture occurs
when principal stresses reach diagonal tension strength of the masonry wall. The second
type of fracture (flexural fracture) takes place in slender confined masonry walls when lon-
gitudinal reinforcing bars of vertical ties yields in one side and brick materials are crushed
in the toe of wall in the other side, or when the flexural deflection dominates the wall behav-
ior. Finally, the third type (slide fracture) is cracking in a horizontal section of the mortar
due to exceeding the shear stresses from the shear strength, resulting from the cohesion of

a) b)
c) d)
Figure 2. Comparison between experimental and numerical models related to the masonry wall without
confinement (color figure available online).
mortar and brick. It must be noted that among these three fracture types, the shear fracture
has been the most common type (Moroni et al. 2004).
In this section, the fracture mechanism obtained by numerical models is presented
for two walls without opening, one having the masonry tensile strength of 0.25 MPa, and
the other 0.03 Mpa, and also for one wall with an opening and the masonry tensile strength
of 0.25MPa. The parameters for which the effects have been considered in these analyses
include the distance of tie columns, the opening aspect ratios, the amount vertical surcharge
on the wall, and the wall thickness.
5.1. Walls Without Opening

The curves showing the sample wall capacity and points related to various damage
stages, as well as the corresponding crack development in walls and ties are represented

a) b)
c) d)
e)
Figure 3. Comparison between experimental and numerical models related to the masonry wall with confinement
(color figure available online).
in Figure 6. The wall is 3.0 m long, 3.0 m high, and its thickness is 220 mm. As shown
in Figure 6a, the point A at displacement value of 1.25 mm on the curve corresponds
with the first observable stiffness degradation and the start of transverse cracking at the
upper end of vertical tie. By increasing the top displacement of the wall and reaching
4.25 mm, cracks become observable, usually in its middle part. After this stage, the wall
top displacement increases to 4.76 mm, at which the maximum strength of the wall is

a) b)
c) d)
e)
Figure 4. Comparison between experimental model of Marinilli and numerical model related to the masonry
wall with confinement (color figure available online).
Figure 5. Effect of connection rigidity in the behavior of a confined masonry wall (CMW) (color figure available
online).

a)
b)
c) d)
Figure 6. Failure mechanism in the numerical model of masonry wall without opening (color figure available
online).
achieved. In the region between points A and B in the Figure 6, the formation of cracks in
the wall is in the stable state, and with increasing displacement, the wall strength increases
as well. However, from this stage, shown by point B in Figure 6, the formation of cracks
increases, the crack width starts increasing, and a sudden decrease in the strength value
occurs with unstable behavior to reach point C, showing the beginning of the final stage
of CMW behavior, which corresponds to its ductile behavior with some residual strength
(73.52 kN), resulting from the friction resistance, which is itself because of the integrity of
the wall, provided by the confinement, and the ties’ inherent strength.

a) b)
c)
Figure 7. Comparison of experimental and numerical models of failure mechanism (color figure available
online).
The fracture mechanism, obtained by numerical analyses, is in good agreement

with the results of Ruiz-Garcia and Negrete (2009) experimental studies, representing
the process of strength deterioration and crack development under cyclic loading. In their
experimental models, the first transverse cracks appear at the top end of the vertical ties.
Then, the diagonal cracks in the masonry wall, caused by the diagonal tension, take place,
which continue into the ends of the vertical ties at either side of the wall. Ultimately, with
the complete formation of diagonal cracks in the masonry wall, the CMW strength reaches
its ultimate value, as shown in Figure 7.
5.2. Walls With Opening

The curves showing the sample wall capacity and points related to various damage
stages as well as the corresponding crack development in walls and ties are represented
in Figure 8. The wall is 2.0 m long, 3.0 m high, and its thickness is 220 mm. As shown
in Figure 8a, the point A at displacement value of 0.9 mm on the curve, corresponds to

a)
b) c)
d) e) f)
Figure 8. Failure mechanism in the numerical model of masonry wall with opening (color figure available
online).
the start of cracking in the upper and lower corners of the opening (Figure 8b). Following
this stage, cracks are observed in ties as well. By increasing the wall top displacement,
cracks spread from the opening corners toward the wall side with some inclination, which
is shown in the Figure 8a by point B, corresponding to displacement value of 2.05 mm.
Appearing of these cracks is in fact the start of formation of plastic hinges in the wall
piers which behave like a column in case of large openings in the wall. After this stage,
the wall top displacement increases with a relatively constant strength till reaching a value
of 3.5 mm, shown by point C in the figure. This stage continues with increase of cracks’
width, and development of cracks toward the wall corners. Then, wall shows a softening
behavior in which piers are deteriorated and diagonal cracks are completely observed in

them, as shown in Figure 8f. This stage is corresponding to point D in Figure 8a. Finally,
because of the existence of ties, a fully ductile behavior along with a residual strength of
30.34 KN is observed in the confined wall.
To verify the numerical modeling the obtained results are compared with those of the
experimental study (Lourenco 1996), shown in Figure 9a. The beginning of failure mech-
anism in numerical model is in good agreement with the experimental results, although
the experimental model is just an unconfined wall with opening, however, in experimen-
tal model the final stage of failure accompanies with forming four rigid blocks, resulting
from bending of failure mechanism in piers, specifically crushing in toes and horizontal
tensile crack in heels (Figure 9b). In numerical study, due to the existence of confining
elements the mentioned final failure mechanism does not happen and instead, diagonal
cracks appear in piers. This behavior is seen in Tehuacan earthquake of 1999, as shown in
Figure 9c (Brzev 2007).
a) b)
c)
Figure 9. Failure mechanism in a wall with an opening (color figure available online).

5.3. Wall Tensile Strength

An important point with regard to the fracture mechanism of wall is the effect of
tensional strength of masonry, as discussed hereinafter. In Figure 10a and 10b, the load-
displacement or the capacity curve of wall with a lower tensional strength (ft = 0.03 Mpa),
and the beginning of cracking at top of the vertical tie are shown. Also in Figure 10c
and 10d opening and development of cracks in the wall and the distribution of principal
compression stress are shown.
a)
b)
c) d)
Figure 10. The behavior of a confined masonry wall (CMW) with low tensile strength of masonry (color figure
available online).

Figure 11. Effect of tensile strength in behavior of confined masonry walls (CMWs) (color figure available
online).
In Figure 10a point A in the capacity curve corresponds with start of cracking in the
tie (shown in Figure 10b), point B corresponds with the start of diagonal cracking in the
wall, point C corresponds with widening of cracks, as shown in Figure 10c, and point D
corresponds with the ultimate strength of wall (fcy = 5.47 Mpa) as shown in Figure 10d.
As shown in Figure 10a, the capacity curve of the confined wall with very low tensional
strength is very similar to that of a concrete frames without infill panels (see Figure 1b).
For higher values of tensional strength, as shown in Figure 11a, the maximum strength
of the CMW becomes more than the residual strength and initial ductility increases, but
the consequent failure of the CMW is sudden and the stability in the softening behavior
decreases. It can be said that, after the start of crack widening in the wall, the strength of
the CMW drops suddenly and reaches the residual strength of the infilled frame (combined
action of the frame and the cracked wall). In case of low tensile strength the wall behavior
becomes similar to an elastic-perfectly plastic one. The trend of this change of behavior
is shown well in experimental models (Flores and Alcocer 1996) (see Figure 11b). In the
case of low tensile strength, the value obtained by Equations (3) or (4) for the maximum
strength becomes less than obtained by Equations (6) and (7), and the only two indices of
initial stiffness and residual strength are sufficient for modeling the CMW.
5.4. Effect of Surcharge

With increasing the amount of vertical surcharge on the CMW, the elastic limit, the
maximum and residual strengths increase, but the initial ductility decreases. This behav-
ior has been observed in experimental studies too, as mentioned previously. The trend of
capacity curve with increasing surcharge is shown in Figure 12.
5.5. Effect of Wall Width

It is observed that with increasing the width of the wall the initial stiffness, the elastic
limit, the maximum and the residual strengths increase, but the initial ductility decreases.
The capacity curve for walls with 22 cm and 35 cm thickness are shown in Figure 13a and
Figure 13b. These observations are valid for both groups of CMWs, however, the fracture
mechanism is quite different in these two groups. In 22 cm thickness walls, the fracture
mechanism is shear mode (Figure 13c and 13d) and diagonal cracks appear in the wall,
while in 35 cm thickness walls, the mechanism depends on the wall length, so that it is

Figure 12. Effect of surcharge in behavior of confined masonry walls (CMWs).
flexural for walls of 2 and 3 m long by yielding of longitudinal reinforcement in one side
(Figure 13e and 13f) and is of shear type in walls of 4 meter length. As a general rule, it
can be said that by decreasing the ratio of length to height of the wall, the initial ductility
increases due to the flexural deflection or yielding of tensile vertical reinforcements in one
side of the CMW.
5.6. Effect of Opening Dimensions

By increasing the dimensions (length and height) of openings, due to increase in the
flexural behavior of piers, the strength and the initial stiffness decrease, while the initial
ductility increases (Figure 14).
5.7. Effect of Wall Thickness

By increasing the wall thickness, the elastic limit, the maximum and residual
strengths and also the ductility of the wall increase. This behavior is shown in Figure 15.
6. SIMPLIFIED ANALYTICAL MODELS

After verification of numerical modeling the simplified analytical models can be
developed by considering some sets of values for the five main factors, affecting the behav-
ior of the confined masonry walls. The mathematical forms of the simplified analytical
models are selected based on the basic relationships in mechanics of structures. The values
of the used factors or coefficients appearing in the considered mathematical formulas are
obtained by regression analysis through the least squares method. In all formulas, dimen-
sions are in term of Newton and millimeter. A sample of the idealized curved fitted to one
set of plotted numerical results is shown in Figure 16. Furthermore, the five indices are
shown in this figure as well.
It is seen that the fitted curve is consisted of four specific linear segments, of
which the initial stiffness, the yielding displacement and secondary stiffness, the maxi-
mum strength and initial displacement ductility, the strength drop after reaching the its
maximum value, and finally, the maximum ductility (if achievable) can be obtained. These
specifications are discussed in more detail in the following discussion.

a) b)
c) d)
e) f)
Figure 13. Effect of distance between the vertical ties in behavior of confined masonry walls (CMWs) (color
figure available online).
6.1. Initial Stiffness

Based on the prevalent relations for flexural and shear stiffness of a deep beam the
stiffness coefficient values of confined walls, can be expressed by Equations (1) and (2) for
walls without opening and walls with opening, respectively:
1
K= (EQ1)
h3w b × hw
+
a × 3 × E × Iw G × Aw

a)
b)
Figure 14. Effect of an opening dimensions in behavior of confined masonry walls (CMWs).
Figure 15. Effect of wall thickness in behavior of confined masonry walls (CMWs).
1
K= lo h o (EQ2)
( lw × h w ) h3w
a× b × c×3×E×IW
+ d×hw
G×AW
In Equations (1) and (2) parameters E and G are modulus of elasticity and shear
modulus of the wall materials, respectively, Iw and Aw are moment of inertia and cross-
sectional area of the horizontal section of the masonry wall (in walls with opening these

Figure 16. A sample of the lateral force-displacement curves of confined walls obtained by numerical analyses
and the fitted curve for developing the simplified analytical formulas (color figure available online).
two parameters are calculated as same as walls without opening), hw is the wall height, lw is
the wall length, ho and lo are the opening height and length, and a, b, c and d are parameters
whose values depend on the wall specifications, respectively, which are given in Table 1.
In other words, the parameters a and b in Equation (1) and c and d in Equation (2) define
the effect of tie columns in the form of effective moment of inertia and cross-sectional area.
In Table 1. the values of correlation coefficient, R, in regression analysis, are also shown,
indicating the high precision of the proposed formulas. To verify the analytical formulas,
proposed for calculation of the initial stiffness of the CMWs without opening, the results
obtained by these formulas have been compared with those obtained by Flores and Alcocer
(1996) and Tomazevic and Klemenc (1998) formula in Table 2, the bases of these formulas
and the proposed formula are similar and only their coefficients are different. In general,
a good agreement exists between the results. It should be mentioned that all the required
data with regard to the masonry material characteristics and reinforcing details cannot be
found in the available literature for experimental studies, however, since the other formula
has been verified by experimental models, and therefore, it can be used for verification of
other formulas, including the above formulas.
6.2. Maximum Strength

For the maximum or ultimate strength, Qu , based on the curves fitted to the numerical
results, Equations (3) and (4), for wall without opening and walls with opening, can be
introduced, respectively, based on the idea from Equation (7–5) in FEMA 273.

1w b fa c
Qu = a × 1.4 × ft × Aw × × 1+ (EQ3)
hw ft
1 h
1w b fa c 0 + 0
Qu = a × 1.4 × ft × Aw × × 1+ × d 1w hw (EQ4)
hw ft
In Equations (3) and (4) the basic parameters are the same as those introduced in
Equations (1) and (2), and ft and fa are the tensional strength of the wall material and the

Table 1. Numerical values of parameters used in Equations (1) to (8)
Equation number and the wall thickness (in cm)

Parameter (1) 22, 35 (2) 22, 35 (3) 22 (3) 35 (4) 22 (4) 35 (5) 22 (5) 35 (5∗ ) 22 (5∗ ) 35 (6) 22 (6) 35 (7) 22 (7) 35 (8) 22 (8) 35
a 1.98 0.95 636 143 505 410 0.76 0.74 0.97 0.95 52339 60567 11.11 11.3 29.46 32.6
279
b 1.77 237 0.43 0.55 0.45 0.49 — — — — 0.98 0.95 0.064 0.067 −5.55 −5.64
c — 2.79 0.55 0.83 0.79 0.88 — — — — 18.64 53.58 1.23 1.28 −4.31 −7.83
d — 1.65 — — 0.33 0.28 — — — — — — 0.91 0.91 4.88 8.45
R2 0.97 0.91 0.94 0.98 0.86 0.92 0.85 0.93 0.85 0.88 0.98 0.98 0.95 0.89 0.91 0.90
∗ Related to wall with opening.
Table 2. Comparison results between Flores and also Tomazevic model and proposed model for initial stiffness
K K(F)/ K(T)/
Ft An L/h Fa/Ft (Tomazevich) K(Flores) K(Model) K(M) K(M)
t = 22 cm
0.25 440000.00 0.67 0.42 39415.95 43700.29 56023.17 0.78 0.70
0.25 660000.00 1.00 0.42 97268.39 115973.85 116959.87 0.99 0.83
0.25 880000.00 1.33 0.42 167517.78 211601.40 180730.25 1.17 0.93
0.20 440000.00 0.67 0.53 38035.56 42169.86 54061.18 0.78 0.70
0.20 660000.00 1.00 0.26 93861.94 111912.31 112863.80 0.99 0.83
0.20 880000.00 1.33 0.00 161651.11 204190.88 174400.87 1.17 0.93
t = 35 cm
0.25 700000.00 0.67 0.27 62707.19 69523.19 89127.78 0.78 0.70
0.25 1050000.00 1.00 0.27 154745.16 184503.85 186072.52 0.99 0.83
0.25 1400000.00 1.33 0.13 266505.56 336638.60 287525.41 1.17 0.93
0.20 1400000.00 1.33 0.33 257172.22 324849.12 277455.93 1.17 0.93
0.15 1050000.00 1.00 0.22 144570.25 172371.69 173838.85 0.99 0.83
compressional stress created in the wall because of the surcharge. To verify the analytical
formulas proposed for calculation of the ultimate strength of the CMWs without open-
ing, the results obtained by these formulas have been compared with those obtained by
Tomazevic and Klemenc (1998) and also Riahi et al. (2009) formula in Table 3. For walls
with thickness of 22 cm the proposed formula is much closer to Tomazevic formula and for
wall with thickness of 35 cm is much closer to Riahi formula. In general a good agreement
exists between the results.
6.3. Yielding or Elastic Limit Strength

Numerical results show that the wall without opening reaches its yielding strength
with start of cracking in ties, and this yielding strength is well below the maximum strength
of the wall. However, the wall with opening reaches its yielding strength with formation
of plastic hinges at both ends of piers, and its yielding strength is approximately the same
as its maximum or ultimate strength, since the existence of opening gives a frame-like
behavior to the wall. The amount of this strength, which can be called the elastic limit of the
wall, is obtained based on the curve fitted to the force- displacement curve from numerical
analyses. Using regression analysis, the following simple formula can be suggested for the
yielding strength of the wall.
Qp = a × Qu (EQ5)
The value of parameter a is given in Table 1, and as it can be seen, it is very close
to unity for walls with opening. Also, for walls without opening, the values of this param-
eter is very close to the values of 0.7 to 0.8 given by Tomazevic and Klemenc (1998)
and 0.8 given by the Flores and Alcocer (1996) analytical formula. Furthermore, for wall
with opening the results obtained by the proposed formula have been compared with that
obtained by Riahi formula in Table 4. It should be mentioned that the proposed formula by
Riahi is based on a linear relationship that was fitted to the limited experimental data in the
form of ratio of the two values of Qp in walls with and without opening as function of the

Table 3. Comparison between results of Tomazevic and the Riahi model and proposed model for maximum
strength
Qu Qu Qu Qu(R)/ Qu(T)/
Ft An L/h Fa/Ft (Riahi) (Tomazevic) (Model) Qu(M) Qu(M)
t = 22 cm
0.25 440000.00 0.67 0.00 85.36 127.66 110.11 0.78 0.86
0.25 440000.00 0.67 0.21 93.75 134.54 122.34 0.77 0.91
0.25 440000.00 0.67 0.42 102.13 140.91 133.63 0.76 0.95
0.25 440000.00 0.67 0.63 110.52 146.86 144.19 0.77 0.98
0.25 660000.00 1.00 0.00 128.04 162.52 156.06 0.82 0.96
0.25 660000.00 1.00 0.21 140.62 173.22 173.38 0.81 1.00
0.25 660000.00 1.00 0.42 153.20 183.08 189.38 0.81 1.03
0.25 660000.00 1.00 0.63 165.78 192.26 204.35 0.81 1.06
0.25 880000.00 1.33 0.00 170.72 198.25 199.86 0.85 1.01
0.25 880000.00 1.33 0.21 187.49 212.71 222.04 0.84 1.04
0.25 880000.00 1.33 0.42 204.26 226.00 242.54 0.84 1.07
0.20 440000.00 0.67 0.00 75.66 105.99 100.04 0.76 0.94
0.20 440000.00 0.67 0.53 92.59 119.01 126.28 0.73 1.06
0.20 660000.00 1.00 0.26 125.95 144.48 161.25 0.78 1.12
0.20 880000.00 1.33 0.00 151.32 162.47 181.58 0.83 1.12
0.20 880000.00 1.33 0.53 185.18 189.70 229.20 0.81 1.21
t = 35 cm
0.25 700000.00 0.67 0.00 135.80 191.66 122.85 1.27 0.64
0.25 700000.00 0.67 0.13 144.06 198.65 136.22 1.21 0.69
0.25 700000.00 0.67 0.27 152.95 205.28 149.33 1.16 0.73
0.25 700000.00 0.67 0.40 161.21 211.62 162.21 1.12 0.77
0.25 1050000.00 1.00 0.00 203.70 247.12 191.90 1.22 0.78
0.25 1050000.00 1.00 0.13 216.09 258.01 212.79 1.16 0.82
0.25 1050000.00 1.00 0.27 229.43 268.30 233.27 1.12 0.87
0.25 1050000.00 1.00 0.40 241.82 278.10 253.38 1.08 0.91
0.25 1400000.00 1.33 0.00 271.61 303.96 263.34 1.18 0.87
0.25 1400000.00 1.33 0.13 288.12 318.67 292.00 1.13 0.92
0.25 1400000.00 1.33 0.27 305.91 332.57 320.10 1.08 0.96
0.25 1400000.00 1.33 0.40 322.43 345.79 347.70 1.05 1.01
0.15 700000.00 0.67 0.00 104.93 122.73 92.76 1.13 0.76
0.15 700000.00 0.67 0.44 121.70 135.94 125.68 0.97 0.92
0.15 1050000.00 1.00 0.22 169.98 158.25 171.01 0.99 1.08
0.20 1400000.00 1.33 0.00 240.74 262.31 232.92 1.07 0.89
0.20 1400000.00 1.33 0.33 274.28 290.19 295.39 0.96 1.02
ratio β of the opening area to the wall area. It seems that this method may be accompanied
with error because the value of β does not show the dimension of piers that leads to differ-
ent behavior for CMW with opening (the same β but different values of width and height
in piers).
6.4. Residual Strength

After reaching its ultimate strength, the wall shows a reduced or residual strength,
which is almost constant with increase in the wall top displacement, until reaching the ulti-
mate displacement. The existence of the residual strength and the capacity of considerable
displacement are mainly because of the existence of ties and their protecting effect on the

Table 4. Comparison results between Riahi model and proposed model for yielding strength in walls with
opening
Qp Qp(Riahi) Qp(R)/
Ft An L/h Fa/Ft (Model) β β = 0 Qp(M)
t = 22 cm
0.25 440000.00 0.67 0.00 55.88 0.08 80.17 1.43
0.25 440000.00 0.67 0.21 64.97 0.08 87.24 1.34
0.25 440000.00 0.67 0.42 73.72 0.08 94.31 1.28
0.25 440000.00 0.67 0.63 82.21 0.08 101.38 1.23
0.25 660000.00 1.00 0.00 60.87 0.17 93.59 1.54
0.25 660000.00 1.00 0.21 70.77 0.17 101.84 1.44
0.25 660000.00 1.00 0.42 80.30 0.17 110.10 1.37
0.25 660000.00 1.00 0.63 89.55 0.17 118.35 1.32
0.25 880000.00 1.33 0.00 59.70 0.25 89.23 1.49
0.25 880000.00 1.33 0.21 69.41 0.25 97.10 1.40
0.25 880000.00 1.33 0.42 78.76 0.25 104.97 1.33
0.20 440000.00 0.67 0.00 50.55 0.08 64.13 1.27
0.20 440000.00 0.67 0.53 70.73 0.08 78.41 1.11
0.20 660000.00 1.00 0.26 66.09 0.17 83.04 1.26
0.20 880000.00 1.33 0.00 54.00 0.25 71.38 1.32
0.20 880000.00 1.33 0.53 75.56 0.25 87.27 1.15
t = 35 cm
0.25 700000.00 0.67 0.00 79.41 0.08 127.54 1.61
0.25 700000.00 0.67 0.13 88.42 0.08 134.50 1.52
0.25 700000.00 0.67 0.27 97.99 0.08 142.00 1.45
0.25 700000.00 0.67 0.40 106.77 0.08 148.97 1.40
0.25 1050000.00 1.00 0.00 85.73 0.17 148.89 1.74
0.25 1050000.00 1.00 0.13 95.47 0.17 157.02 1.64
0.25 1050000.00 1.00 0.27 105.80 0.17 165.77 1.57
0.25 1050000.00 1.00 0.40 115.28 0.17 173.90 1.51
0.25 1400000.00 1.33 0.00 82.57 0.25 141.95 1.72
0.25 1400000.00 1.33 0.13 91.95 0.25 149.70 1.63
0.25 1400000.00 1.33 0.27 101.90 0.25 158.05 1.55
0.25 1400000.00 1.33 0.40 111.03 0.25 165.80 1.49
0.15 700000.00 0.67 0.00 61.82 0.08 76.52 1.24
0.15 700000.00 0.67 0.44 85.21 0.08 90.67 1.06
0.15 1050000.00 1.00 0.22 79.51 0.17 97.59 1.23
0.20 1400000.00 1.33 0.00 74.02 0.25 113.56 1.53
0.20 1400000.00 1.33 0.33 95.14 0.25 129.30 1.36
wall integrity, and to some extent due to the friction existing in the crack area of wall.
Based on regression analyses Equations (6) and (7) have been obtained for walls without
opening and walls with opening respectively, which give the residual strength as a function
of length to height ratio and amount of surcharge of the masonry unit. According to the
results of section 5.3, the ratio of residual to maximum strength depends on the wall ten-
sile strength. It is shown that with increasing the wall tensile strength the mentioned ratio
increases and it seems that it should not be limited to an absolute value.
b
lw
Qr = a × × cfa (EQ6)
hw


b

1w 10 h
+ hw0
Qr = exp a × ×c ×d
fa lw
(EQ7)
hw
6.5. Displacement Ductility

Referring to Figure 16, the displacement ductility can be defined as ratio of the dis-
placement value at the maximum or ultimate strength to displacement value at the elastic
limit strength. Lack of existing models for the deformation capacity in various states, non-
linearity of the response and contribution of both characteristic of panel and confinement
elements to this parameter made model specification difficult. Considering that ductility
values are dependent on the amounts of vertical loading or surcharge on the wall, length
to height ratio, and tension and compression strengths of masonry materials. For walls
without opening, based on regression analyses, Equation (8) can be introduced.

lw fa
D = [a × ft ] + b × + c× +d ≥1 (EQ8)
hw ft
The proposed formula for calculating displacement based on the maximum strength
is compared with Flores and also Tomazevic formulas in Table 5. Flores and Alcocer
(1996) proposed an absolute value of 0.003 for drift independent of wall characteristics.
It may result in error in comparison with the exact value. Also in Tomazevic formula, as
shown in Table 5, the trends of the two formulas in both 22 cm and 35 cm thickness of
wall are similar and especially a good agreement exists between the results of walls with
22 cm thickness. Some deviation is observed in walls with 35 cm thickness. In Tomazevic
model for walls with 35 cm thickness despite the decrease of surcharge, displacement
has a decreasing trend in comparison with walls of 22 cm thickness, whereas in the pro-
posed model this trend is vice versa. It was shown in section 5.7 that with increasing the
wall thickness the displacement corresponding to the maximum strength increases as well.
Table 5. Comparison results between Flores and also Tomazevic model and proposed model for displacement
corresponding to maximum strength
(Model) (Flores) (Toma.) (F)/ (T)/

Ft An L/h Fa/Ft [mm] [mm] [mm] (M) (M)
t = 22 cm
0.25 440000.00 0.67 0.42 12.21 9.00 8.24 0.74 0.67
0.25 660000.00 1.00 0.42 6.01 9.00 4.34 1.50 0.72
0.25 880000.00 1.33 0.42 3.10 9.00 3.11 2.91 1.00
0.20 440000.00 0.67 0.53 8.50 9.00 7.21 1.06 0.85
0.20 660000.00 1.00 0.26 4.45 9.00 3.55 2.02 0.80
0.20 880000.00 1.33 0.00 2.67 9.00 2.32 3.37 0.87
t = 35 cm
0.25 700000.00 0.67 0.27 13.30 9.00 7.55 0.68 0.57
0.25 1050000.00 1.00 0.27 8.21 9.00 4.00 1.10 0.49
0.25 1400000.00 1.33 0.13 6.06 9.00 2.76 1.49 0.45
0.20 1400000.00 1.33 0.33 3.83 9.00 2.60 2.35 0.68
0.15 1050000.00 1.00 0.22 4.35 9.00 2.52 2.07 0.58

However, in Tomazevic formula some simplifying assumptions have been used causing
some errors.
For walls with opening, since the wall piers have a great role in the wall ductility,
the ductility is a function of dimensions ratio of piers and also the amount of surcharge.
In Equations (9) and (10) the ductility values are given respectively for walls of 22 cm
thickness and walls of 35 cm thickness.
⎧ lp
⎪ 0.6 × [5.88 − (24.7 × fa )] >1
⎨ hp
lp
D = 5.88 − (24.7 × fa ) 0.75 ≤ ≤ 1; ≥1 R2 = 0.94 (EQ9)
⎪
⎩
hp
lp
1.3 × [5.88 − (24.7 × fa )] hp
< 0.75
⎧ lp
⎪ 0.68 × [5.68− (31.32 × fa )] >1
⎨ hp
lp
D = 5.68 − (31.32 × fa ) 0.75 ≤ ≤ 1; ≥1 R2 = 0.97 (EQ10)
⎪
⎩
hp
lp
1.8 × [5.68 − (31.32 × fa )] hp
< 0.75
In Equations (9) and (10) lp and hp , are the length and the height of piers (the part
of wall locating at either side of an opening). Difference between the formulas related to
walls with opening of 22 cm thickness and wall of 35 cm thickness is basically due to the
difference between the behaviors of these walls, which is itself due to remarkably different
values of compression stresses in walls with opening, having different thicknesses as well
as the dimensions of opening. In other words, it seems that the existence of opening in
the wall leads to the creation of some local submechanisms resulting in a complicated
behavior of the CMW. It may be seen in Equation (8) that, with an increase in the amount
of surcharge and also the amount of length-to-height ratio, the wall ductility decreases,
in case of walls without opening. The same is true for walls with opening, and as the
length-to-height ratio of piers and also the amount of surcharge increase, the wall ductility
decreases, and vice versa.
7. SUGGESTION OF A MACROMODEL
By using the proposed analytical formulas it is possible to simulate the CMW
buildings in a three-dimensional configuration by introducing some macromodels in any
conventional engineering software for push over analysis. For this purpose, each CMW is
substituted by an element of linear configuration, having the geometrical properties of the
corresponding wall unit, and a plastic shear hinge at the middle of the substitute element
(Shiga et al. 1980), whose boundary conditions are defined as a hinge at the bottom, and
a moment bearing roller at the top of the element (Figure 17). The plastic behavior of the
substitute hinge is given by the proposed formulas. As a sample of substituting the CMWs
with ordinary elements, the capacity curves of a CMW of 35 cm thickness, 4.0 m length,
and 3.0 m height without opening, having tensional strength of masonry unit of 0.25 MPa
and surcharge of 0.12 MPa, obtained by the mentioned method and the proposed formulas,
is presented in Figure 18, which shows a good agreement between the two curves.
8. VERIFICATION OF MACROMODELING
In order to verify the proposed formulas for making macromodels of CMW buildings
a full scale two-story building, studied before by Alcocer et al. (1996) under cyclic loading

Figure 17. A confined masonry wall (CMW). model for static nonlinear analysis.
Figure 18. Comparison of capacity curve obtained by the proposed macromodel and that of analytical model
(color figure available online).
is modeled by the proposed formulas and the results are compared. The geometric features
and test set-up of the building are shown in Figure19a. In fact, the model consists of eight
elements in direction of loading and four elements perpendicular to them symmetrically,
which represents 12 walls without openings that are connected together by the slab.
A displacement control POA was performed by a target displacement of 2.5 cm
(0.01h) and a lateral load pattern proportional with the story masses multiplied by their
height above the foundation level (Salonikios et al. 2003). The failure mechanisms of the
two models are very similar as shown in Figure 19b. In experimental model, the first story
walls have suffered heavy damage; in the second story, the walls of 210 cm length have
experienced minor damage; finally, the walls of 160 cm length have exhibit elastic behav-
ior. For comparison, damage in the macro-model is measured by lateral displacement of
plastic shear hinge, In the first story, the lateral displacement of CMW is more than that
obtained based on the maximum strength given by the formula (in the residual strength
region) and in the second story that of the wall of 210 cm it is more than the lateral dis-
placement obtained based on the elastic limit, and less than the maximum strength (before
reaching the strength peak). In the wall of 160 cm, the lateral displacement of CMW is less
than that obtained based on the elastic limit (in the elastic region, Figure 20). Also the soft
story mechanism is shown in this figure similar to the experimental results and the actual
reaction of CMW subjected to the earthquake in August 2007 in Pisco, Peru (Brzev 2007).
The capacity curves of the considered building, obtained by experiment and the
macromodel are shown in Figure 21, which shows the similar trend of the two curves.

a)
b)
Figure 19. Experimental study of a two-story masonry building.
Figure 20. Damage status and distribution of plastic hinge in structure (color figure available online).

Figure 21. Comparison of capacity curve obtained by experimental study and that of numerical model (color
figure available online).
A value of 1.14 can be obtained from this figure for the ratio of the maximum capacity of
the numerical model to that of the experimental model and proposed macromodel underes-
timates the initial stiffness in the elastic range. It seems that the actual boundary condition
in experimental model and disability numerical model to capture of its real condition and
the lack of information about properties of masonry wall such as modules of elasticity,
shear modules, and some uncertainty such as the effect of workmanship that affect on the
response of masonry structures, cause the poor prediction in initial stiffness region; how-
ever, a good agreement exists between the results of the experimental and numerical model
for the maximum strength and displacement corresponding to maximum strength and also
the post cracking behavior. The proposed analytical model has an acceptable precision
level in addition to simplicity and saving the time and the costs of analyses. The important
feature of this model is capability of capturing the post-cracking and softening behavior of
the walls in buildings.
9. LIMITATIONS OF THE ANALYTICAL MODEL

Limitations of the analytical model follow.
r The general behavior of CMW models is based on their shear performance, therefore,
any violation from this behavior could disturb the accuracy of the model.
r The applicability of the derived equations is limited by the range of variables included
in section 2.
r According to this research, if existing longitudinal reinforcement in tie columns is
less than 48 with mentioned dimensions for tie columns in section 2, the failure
mechanism changes to flexural failure.
r In this research the units and mortar of masonry wall are clay bricks and sand-cement,
therefore the assumption of modeling independent of workmanship is based on these
kinds of materials.
r In walls, especially those with opening, the increasing of opening length and height and
surcharge out of the range of variables in section 2 may result in premature masonry
crushing in spandrel beam.

10. CONCLUSIONS
In this research, by use of DIANA Software (9.3), the focus was first to perform
a set of nonlinear static analyses on confined masonry walls with and without opening.
These analyses were performed in the form of sensitivity analysis of the numerical models
versus the known effective factors on the behavior of confined masonry walls. Numerical
modeling has been verified by experimental and other verified numerical models. Then,
based on the results of numerical analyses, some simplified relationships for lateral force-
displacement behavior were developed. In the proposed simplified formulas, the effects of
geometric factors of walls as well as the mechanical specifications of the wall materials,
and also the amount of the existing surcharge are considered. The relatively high values of
coefficient of correlation for the parameters used in the proposed simplified formulas show
their relatively high precision. To use the simplified formulas in engineering practice for
evaluation of buildings with confined walls, some simple elements, such as plastic shear
hinge, can be used in the conventional software such as SAP or ETABS. The specifications
of these plastic shear hinges can be easily determined based on adjusting their values with
the corresponding values in the proposed formulas.
REFERENCES
Alcocer, S. M., J., Ruiz, A. Pineda, and Z. Zepeda. 1996. Retrofitting of confined masonry walls
with welded wire mesh. In Proceedings of the Eleventh World Conference on Earthquake
Engineering. Mexico, 1996, Elsevier Science, Ltd. paper 1471.
Applied Technology Council (ATC). 1996. Seismic evaluation and retrofit of concrete buildings:
ATC–40, vols. 1–2. ATC.
Bruneau M. 1994. Seismic evaluation of unreinforced masonry buildings: A state-of-the-art report.
Canadian Journal Civil Engineering 21(3):512–539.
Brzev, S. 2007. Earthquake-resistant confined masonry construction. Kanpur, India: National
Information Center of Earthquake Engineering (NICEE).
DIANA. 2005. DIANA finite element analysis: User’s manual—Element library. Delft, The
Netherlands: TNO Building and Construction Research.
Flores, L. E., and S. M. Alcocer. 1996. Calculated response of confined masonry structures.
In Proceedings of the Eleventh World Conference on Earthquake Engineering. Mexico, 1996,
Elsevier Science, Ltd. paper 1830.
Hashemi, A., and K. Mosalam. 2007. Seismic evaluation of reinforced concrete buildings including
effects of masonry infill walls. PEER Technical Report, 255 pp.
Lang, K. 2002, February. Seismic vulnerability of existing buildings. Zurich, Switzerland, : Institute
of Structural Engineering/Swiss Federal Institute of Technology.
Lourenco, P., G. Rots, and J. Blaauvendraad. 1998. Continuum model for masonry: parameter
estimation and validation. Journal of Structural Engineering 124(6): 642–652.
Lourenco, P. 1996. Computational strategies for masonry structures. Delft, The Netherlands: Delft
University.
Maekawa, K., and H. Okamura. 1990. Nonlinear analysis and constructive models of reinforced
concrete. Tokyo, Japan: University of Tokyo.
Marinilli, A., and E. Castilla. 2004. Experimental evaluation of confined masonry walls with sev-
eral confining columns. In Proceedings of the Thirteenth World Conference on Earthquake
Engineering. Vancouver, Canada, August 1–6, 2004, paper 2129.
Mohebkhah A., A. A. Tasnimi, H. A. Moghadam, et al. 2008. Nonlinear analysis of masonry–
infilled steel frames with openings using discrete element method. Journal of Constructional Steel
Research 64(12): 1463–1472.

Moroni, M. O., M. Astroza, and C. Acevedo. 2004, Performance and seismic vulnerability of
masonry housing types used in Chile. Journal of Performance of Constructed Facilities 18(3):
173–179.
Moroni, M. O., M. Astroza, and S. Tavonatti. 1994. Nonlinear models for shear failure in confined
masonry walls. TMS Journal 12(2):72–78.
Mostafaei, H., F. J. Vecchio, and T. Kabeyasawa. 2008. Nonlinear displacement based response
prediction of reinforced concrete columns. Journal of Engineering Structures 30(9): 2436–2447.
Riahi, Z., K. Elwood, and S. M. Alcocer. 2009, Backbone model for confined masonry walls for
performance-based seismic design. Journal of Structural Engineering ASCE 135(6):644–654.
Ruiz-Garcia, J., and M. Negrete. 2009. Drift-based fragility assessment of confined masonry walls
in seismic zones. Journal of Engineering Structures, September 18:170–181.
Salonikios, T., C. Karakostas, V. Lekidis, and A. Anthoine. 2003. Comparative inelastic pushover
analysis of masonry frame. Journal of Engineering Structures 25:1515–1523.
Shiga, T., A. Shibata, J. Shibuya, and J. Takahashi. 1980. Performance of the building of faculty of
engineering, Tokyo University, during the 1978 Miagi-ken-oki earthquake. In Proceedings of the
7th WCEE. Istanbul, Turkey, vol. 7, 357–364.
Teran-Gilmore, A., O. A. Zuniga-Cuevas, and J. Ruiz-Garcia. 2009. Displacement-based assessment
of low-height confined masonry buildings. Earthquake Spectra 25(2):439–464.
Tomazevic, M., and I. Klemenc I. 1998. Seismic behavior of confined masonry walls. Earthquake
Engineering and Structural Dynamics 26: 1059–1071.
Vecchio, F., and M. Basil Emara. 1992, January–February. Shear deformations in reinforced concrete
frames. ACI Structural Journal, 89(1).

Simplified Formulation For Modeling The Nonlinear Behavior of Confined Masonry Walls in Seismic Analysis

Încărcat de

Informații document

Titlu original

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

Simplified Formulation For Modeling The Nonlinear Behavior of Confined Masonry Walls in Seismic Analysis

Încărcat de

Drepturi de autor:

Formate disponibile

International Journal of Architectural Heritage

Conservation, Analysis, and Restoration

ISSN: 1558-3058 (Print) 1558-3066 (Online) Journal homepage: https://www.tandfonline.com/loi/uarc20

Simplified Formulation for Modeling the Nonlinear

Fariman Ranjbaran , Mahmood Hosseini & Masoud Soltani

To link to this article: https://doi.org/10.1080/15583058.2010.528826

Accepted author version posted online: 11

Submit your article to this journal

Article views: 270

View related articles

Citing articles: 4 View citing articles

Full Terms & Conditions of access and use can be found at

SIMPLIFIED FORMULATION FOR MODELING THE

Fariman Ranjbaran,1 Mahmood Hosseini,2 and

Received 25 September 2010; accepted 29 September 2010.

INTERNATIONAL JOURNAL OF ARCHITECTURAL HERITAGE 6(3): 259–289

INTERNATIONAL JOURNAL OF ARCHITECTURAL HERITAGE 6(3): 259–289

2. STATING THE PROBLEM

INTERNATIONAL JOURNAL OF ARCHITECTURAL HERITAGE 6(3): 259–289

INTERNATIONAL JOURNAL OF ARCHITECTURAL HERITAGE 6(3): 259–289

4. VERIFICATION OF NUMERICAL MODELING

INTERNATIONAL JOURNAL OF ARCHITECTURAL HERITAGE 6(3): 259–289

INTERNATIONAL JOURNAL OF ARCHITECTURAL HERITAGE 6(3): 259–289

5.1. Walls Without Opening

INTERNATIONAL JOURNAL OF ARCHITECTURAL HERITAGE 6(3): 259–289

INTERNATIONAL JOURNAL OF ARCHITECTURAL HERITAGE 6(3): 259–289

INTERNATIONAL JOURNAL OF ARCHITECTURAL HERITAGE 6(3): 259–289

INTERNATIONAL JOURNAL OF ARCHITECTURAL HERITAGE 6(3): 259–289

The fracture mechanism, obtained by numerical analyses, is in good agreement

5.2. Walls With Opening

INTERNATIONAL JOURNAL OF ARCHITECTURAL HERITAGE 6(3): 259–289

INTERNATIONAL JOURNAL OF ARCHITECTURAL HERITAGE 6(3): 259–289

INTERNATIONAL JOURNAL OF ARCHITECTURAL HERITAGE 6(3): 259–289

5.3. Wall Tensile Strength

INTERNATIONAL JOURNAL OF ARCHITECTURAL HERITAGE 6(3): 259–289

5.4. Effect of Surcharge

5.5. Effect of Wall Width

INTERNATIONAL JOURNAL OF ARCHITECTURAL HERITAGE 6(3): 259–289

Figure 12. Effect of surcharge in behavior of confined masonry walls (CMWs).

5.6. Effect of Opening Dimensions

5.7. Effect of Wall Thickness

6. SIMPLIFIED ANALYTICAL MODELS

INTERNATIONAL JOURNAL OF ARCHITECTURAL HERITAGE 6(3): 259–289

6.1. Initial Stiffness

INTERNATIONAL JOURNAL OF ARCHITECTURAL HERITAGE 6(3): 259–289

INTERNATIONAL JOURNAL OF ARCHITECTURAL HERITAGE 6(3): 259–289

6.2. Maximum Strength

INTERNATIONAL JOURNAL OF ARCHITECTURAL HERITAGE 6(3): 259–289

Equation number and the wall thickness (in cm)

6.3. Yielding or Elastic Limit Strength

INTERNATIONAL JOURNAL OF ARCHITECTURAL HERITAGE 6(3): 259–289

6.4. Residual Strength

INTERNATIONAL JOURNAL OF ARCHITECTURAL HERITAGE 6(3): 259–289

INTERNATIONAL JOURNAL OF ARCHITECTURAL HERITAGE 6(3): 259–289

6.5. Displacement Ductility

(Model) (Flores) (Toma.) (F)/ (T)/

INTERNATIONAL JOURNAL OF ARCHITECTURAL HERITAGE 6(3): 259–289

INTERNATIONAL JOURNAL OF ARCHITECTURAL HERITAGE 6(3): 259–289

INTERNATIONAL JOURNAL OF ARCHITECTURAL HERITAGE 6(3): 259–289

Figure 19. Experimental study of a two-story masonry building.

INTERNATIONAL JOURNAL OF ARCHITECTURAL HERITAGE 6(3): 259–289

9. LIMITATIONS OF THE ANALYTICAL MODEL

INTERNATIONAL JOURNAL OF ARCHITECTURAL HERITAGE 6(3): 259–289