PARAMETERIC STUDY OF THE STRUCTURAL BEHAVIOR OF FRP

REINFORCED MASONRY
Mario R. Escalante
escalantem@frcu.utn.edu.ar
UTN - {FRCU,FRCon}
Ing. Pereyra 676, 3260, Concepcion del Uruguay, Entre Ros, Argentina.
Viviana C. Rougier
rougierv@frcu.utn.edu.ar
UTN - {FRCU,FRCon}
Ing. Pereyra 676, 3260, Concepcion del Uruguay, Entre Ros, Argentina.
Abstract. Masonry structures are designed to serve a certain lifetime. However, there are several masonry buildings that have been damaged in a shorter time because of different external
actions like earthquakes, impact loads, winds, changes of use or aggressive agents. The use
of fiber reinforced polymers (FRP) as external reinforcement of masonry walls considerably
improves structural stability with a minor impact over foundations. The in-plane mechanical
behavior of unreinforced masonry and masonry retrofitted with FRP has been studied by several authors. In most of them, parameters are considered as deterministic values, nevertheless
it must be remarked that numerical models are plagued by uncertainties. For example, mechanical properties of masonry are much more variables than in the case of reinforcing steel
and concrete.In the same manner the thickness of the mortar joint is not regular. In this paper,
an existing coupled damaged-plasticity model in which bricks and mortar are separately modeled (micro approach), is used to perform a parametric study (e.g. mechanical properties) of
structure load carrying capacity. Such micro model allows simulating the behavior of masonry
elements using the mechanical properties of constitutive materials and their layout and it also
permits to take into account the debonding of the interface between bricks and mortar. Composite materials are modeled with a generalization of mixture theory, where each component
exhibits orthotropic elasto-plastic behavior. A model of unreinforced and reinforced masonry
panel is analyzed. The specimen that is constructed with clay bricks and mortar joints is subjected to compression normal to bed joints. Numerical simulations are implemented by means
of a non linear dynamics finite element method. Some load-displacement curves are shown and
numerical and experimental results are compared.
Keywords: parametric study, frp reinforced masonry
CILAMCE 2013
Proceedings of the XXXIV Iberian Latin-American Congress on Computational Methods in Engineering
Z.J.G.N Del Prado (Editor), ABMEC, Pirenpolis, GO, Brazil, November 10-13, 2013

Parametric study of the structural behaviour of FRP reinforced masonry

INTRODUCTION

Masonry is a heterogeneous composite in which brick units are held together by mortar.
Brick units can be made from clay, compressed earth, stone or concrete. Mortar can be a mixture of cement, lime, sand and water in various proportions. Consequently, masonry properties
vary from one structure to the next depending on the type of brick units and mortar used. Other
factors contributing to the variability of masonry properties include anisotropy of units, dimension of units, mortar joint width, the geometrical arrangement of masonry and the quality of
workmanship. Nevertheless, bricks and mortar being the most visible components still determine the performance of masonry. Therefore, it should examine the properties of mortar and
brick units in order to gain an insight into the behavior of masonry.
Depending on the type of load, masonry presents a complex behavior normally accompanied by cracking and brittle and sudden failure. Mechanical properties degradation and structural safety loss make the rehabilitation or reinforcement necessary.
Research into the use of fiber reinforced polymers (FRP) in masonry has shown that this
system enhances masonry wall performance, increasing final strength and in some cases ductility and stiffness. In this way, brittle behavior and sudden failure of unreinforced masonry can
be avoided.
Although experimental results have shown the advantages of this retrofitting technique,
analytical or numerical tools for description of the mechanical behavior of unreinforced and
reinforced masonry elements under different load conditions are needed to improve this intervention technique. Numerical models used for FRP retrofitted masonry usually involve masonry
models. As models used for all composite materials, masonry models can be classified in three
types (Grande et al., 2008).
Micro-models: Mortar, bricks and their interfaces are independently simulated using different constitutive models appropriate for each material with properties obtained from experimental tests performed on individual bricks, masonry units or special brick mortar small elements
designed to test the brick mortar interface. This type of model is able to reproduce experimental
tests on small masonry elements but has high computational cost when applied to full scale
masonry walls or buildings.
Micromacro: Different models are considered for bricks, mortar and interfaces combined
with some kind of homogenization procedure to obtain an equivalent homogeneous model for
structural analysis. This type of model has the advantage of reducing the computational cost
without requiring too much experimental data.
Macro-models: A phenomenological constitutive law is used to model masonry behavior.
The stress strain relationships are obtained from experimental tests performed on small masonry
units subjected to different types of loads. While very simple to use in structural analysis, this
type of model requires many experimental tests that should be repeated for each type of masonry
although the bricks and the mortar are the same. These different approaches combined with
different ways of modeling the FRP reinforcement have also been used for different authors to
model FRP retrofitted masonry.
In order to analyze the influence of different mechanical properties of constituent materials
of masonry, a parametric study on unreinforced and carbon fiber reinforced polymers (CFRP)
reinforced masonry panels was performed. This is part of an on going research focusing on the
CILAMCE 2013
Proceedings of the XXXIV Iberian Latin-American Congress on Computational Methods in Engineering
Z.J.G.N Del Prado (Editor), ABMEC, Pirenpolis, GO, Brazil, November 10-13, 2013

M. Escalante, V. Rougier

study of the stochastic model. The following parameters were considered: ultimate compression
strength (uc ) and Youngs modulus (E) of the clay bricks and mortar, and Youngs modulus of
CFRP.
To carry out that study an existing coupled damaged-plasticity model (Luccioni and Rougier,
2005) was used. Finally, some load-displacement curves are shown and numerical and experimental results are compared.

2.1

UNIAXIAL BEHAVIOR OF UNREINFORCED AND RETROFITTED

MASONRY UPON LOADING NORMAL TO BED JOINT
Behavior of unreinforced masonry under compression normal to the
bed joints

Masonry is very weak in tension because it is composed of two different materials distributed at regular intervals and the bond between them is weak. However, masonry transmits
compressive loads very effectively. As failure occurs by splitting due to transverse tension in
the brick caused by the differential lateral expansion of the stiffer brick and more flexible mortar (Page, 1978; Gabor et al., 2006), compression capacity is governed by the tensile properties
of the units. The bed joint has a tendency to expand laterally more than the bricks because of
lesser stiffness. However, mortar is confined laterally at the brickmortar interface by the bricks
because of the bond between them; therefore, shear stresses at the brick-mortar interface result in an internal state of stress which consists of triaxial compression in mortar and bilateral
tension coupled with axial compression in bricks (see Fig.1a). In case of low stiffness bricks,
they tend to expand laterally faster than high stiffness mortar. But the mortar prevents this
lateral expansion and it creates a triaxial compression stress state in bricks with the resulting
tensile stresses in the mortar (Fig.1b). As soon as the principal tensile stress reaches the tensile strength of the mortar, cracks appear and the failure occurs (Rougier, 2007; Prakash and
Alagusundaramoorthy, 2008)
Generally, failure modes under uniaxial compression also depend upon the type of mortar
used. Strong mortars are usually associated with brittle and explosive failure; weak mortars
are associated with ductile and slower rate of crack propagation (Rougier, 2007; Lourenco,
1996). Examples of load-displacement curves obtained in 500x250x600 [mm3 ] prisms of solid
soft mud brick are shown in Figure 2. As mortar strength increases, compression strength of
masonry increases. However, Rougier (2007) found that such increase is not lineal in 490
280 130 [mm3 ] prisms of solid clay units, when the mortar strengths were of 1.87 MPa and
4.93 MPa, an average of uniaxial compression stresses of 3.09 MPa and 3.84 MPa was obtained,
respectively. So 2.6 times stronger mortar corresponds to an increase of masonry strength of
approximately of 25%.

2.2

Behavior of reinforced masonry under compression normal to the bed

joints

Recent researches carried out over masonry specimens subjected to axial compression normal to the bed joints showed different results, depending on material properties of masonry
CILAMCE 2013
Proceedings of the XXXIV Iberian Latin-American Congress on Computational Methods in Engineering
Z.J.G.N Del Prado (Editor), ABMEC, Pirenpolis, GO, Brazil, November 10-13, 2013

Parametric study of the structural behaviour of FRP reinforced masonry

Figure 1: Uniaxial behaviour of masonry prism upon loading normal to bed joints: a) State of stress corresponding to high stiffness bricks. b) State of stress corresponding to low stiffness bricks

Figure 2: Uniaxial behavior of masonry upon loading normal to the bed joint: stress-displacement diagrams
for 500x250x600 [mm3 ] prisms of solid soft mud brick obtained by Binda et al.1988 (Lourenco, 1996).

components, retrofitting schemes, amount and layout of the reinforcement. Prakash and Alagusundaramoorthy (2008) studied the effectiveness of GFRP composite retrofitting on the behaviour of masonry wallettes constructed with high stiffness cement mortar and low stiffness
brick, using two layers of GFRP composites. They found enhancing in load resistance up to
20 %, increasing in stiffness of more than 100% and reducing up to 40% in average ultimate
strain due to GFRP reinforcing. However, Luccioni and Rougier (2011) tested small solid clay
masonry panels retrofitted with one layer of CFRP sheets and strips and they found that FRP
reinforcement did not increase strength and stiffness substantially, but it improved ductility and
modified the failure mode (see Fig.3).
CILAMCE 2013
Proceedings of the XXXIV Iberian Latin-American Congress on Computational Methods in Engineering
Z.J.G.N Del Prado (Editor), ABMEC, Pirenpolis, GO, Brazil, November 10-13, 2013

M. Escalante, V. Rougier

Figure 3: Stress-strain curves for CFRP retrofitted specimens under compression. Comparison with unreinforced specimen (Luccioni and Rougier 2011).

TEST DESCRIPTION

Several tests were carried out in order to evaluate the uniaxial compression behavior of
unreinforced and CFRP retrofitted masonry normal to the bed joints. Two group of panels
were built and tested according to INPRES-CIRSOC 103 recommendations and trying to reflect material and workmanship qualities similar to those used in actual masonry construction
in Argentina: Group I (580x 610x130 [mm3 ] panels) and Group II (560 x 550 x 125 [mm3 ]
panels).The specimens dimensions differ the dimensions of the solid clay bricks used were not
the same. The dimensions of the bricks for Groups I and II are shown in Table 1. All panels
had a 15 mm thick mortar joint (V. and B., 2012). Two types of retrofitting schemes were used:
total reinforcement and reinforcement with 75 mm width strips (Fig.4). The strip width was
chosen based on numerical results (Rougier, 2007). In all cases, both the front and back sides
of the panels were reinforced with one CFRP layer and fibers were laid out horizontally, that
is, parallel to the bed joints and normal to the load direction. In this work, only specimens of
Group II and reinforcement with 75 mm width strips were analyzed.
The strength and elastic properties of bricks, mortar and brick masonry panels were evaluated through testing (Rougier, 2007) and they are also included in Table 1.
Unidirectional carbon fiber fabric with high content of carbon and high modulus and strength,
saturated in situ with an epoxy system was used for the reinforcement and retrofitting. The
properties of the lamina were obtained from the manufacturer (SIKA) specifications and were
numerically validated (Rougier 2007) using a constitutive model for unidirectional fiber reinforced lamina. They are presented in Table 2.
The test setup for compression normal to the bed joint is shown in Fig.5. To ensure uniform
pressure and to reduce the lateral displacement, 20 mm thick steel plates with a rubber layer of
CILAMCE 2013
Proceedings of the XXXIV Iberian Latin-American Congress on Computational Methods in Engineering
Z.J.G.N Del Prado (Editor), ABMEC, Pirenpolis, GO, Brazil, November 10-13, 2013

Parametric study of the structural behaviour of FRP reinforced masonry

Figure 4: CFRP retrofitted and repaired specimens tested under compression. CFRP layout. (a) Completely
reinforced. (b) Band reinforced .
Table 1: Mortar and bricks mechanical properties obtained from tests

Mortar
Properties

Grupo I

Bricks

Grupo II

Grupo I

Grupo II

(a)

(b)

Elasticity modulus E (MPa)

3380

4312

1528

1662

1400

Comp. ult. strength, uc (MPa)

6.73

7.72

4.00

11.82

8.28

Charac. Flex. Rup.Modulus (MPa)

2.83

2.83

2.65

Poisson ratio,

0.21

0.21

0.21

0.16

0.15

Length

280

260

Width

130

125

Height

20

20

50

55

Dimensions (mm):

approximately 8 mm thickness were placed on the top and bottom of the specimens. Vertical
and horizontal relative displacement between two fixed points were measured on both sides of
the panels and then extrapolated to the total width and height of the panel. In most cases, the
measurement equipment had to be removed before failure for safety reasons.
All the tests were performed with Instron 8204, 600 kN testing machine under increasing
load with displacement control (Rougier, 2007).

NUMERICAL STUDY

In this work the masonry and CFRP reinforced masonry elements under compression normal to bed joints are simulated with a micro-model in which bricks and mortar are separately
CILAMCE 2013
Proceedings of the XXXIV Iberian Latin-American Congress on Computational Methods in Engineering
Z.J.G.N Del Prado (Editor), ABMEC, Pirenpolis, GO, Brazil, November 10-13, 2013

M. Escalante, V. Rougier

Table 2: Composite mechanical properties

Volume fraction of fibers, kf

0.3

Longitudinal elasticity modulus, El (MPa)

72500

Transversal elasticity modulus, Et (MPa)

6200

Longitudinal-transversal Poissons ratio , lt

0.08

Transversal-longitudinal Poissons ratio, tl

0.017

Transversal-transversal Poissons ratio, tt

0.20

Longitudinal tensile strength, lt (MPa)

960

Transverse tensile strength, ut (MPa)

51

Figure 5: Figure 5. Experimental setup for uniaxial compression test on masonry panels.

modeled (Luccioni and Rougier, 2011). The model was implemented in a non-linear plane finite
element program.
The application of this model to masonry requires the proper definition of the finite element model and the definition of several functions (yield function, potential function, damage
function, hardening functions, etc.) and parameters for the constitutive models. Functions normally used for concrete were used for bricks and mortar. Some of the mechanical properties of
bricks and mortar, like elasticity modulus, Poison ratio, compression strength, tension strength,
fracture energy and compression strain hardening were obtained from tests performed on bricks
and mortar specimens (Rougier, 2007). The rest of the parameters were indirectly obtained
through numerical simulation of small unreinforced masonry specimens and comparison with
experimental results (Luccioni and Rougier, 2011).
The interface between bricks and mortar allowing possible debonding is also approximately
considered without explicitly modeling the interface but properly modifying the mortar constitutive equation(Luccioni and Rougier, 2011; Luccioni et al., 2005). The mortar was supposed
to be anisotropic with reduced shear strength in the interface plane. The value of the shear
strength was adjusted to reproduce the experimental results of small masonry specimens tested
in (Luccioni and Rougier, 2010). FRP reinforcement is simulated without explicitly modelCILAMCE 2013
Proceedings of the XXXIV Iberian Latin-American Congress on Computational Methods in Engineering
Z.J.G.N Del Prado (Editor), ABMEC, Pirenpolis, GO, Brazil, November 10-13, 2013

Parametric study of the structural behaviour of FRP reinforced masonry

ing reinforcement elements but with a generalization of the classic mixture theory (Oller et al.,
2006). In this way, FRP reinforced bricks or FRP reinforced mortar are considered as composite materials made of bricks or mortar and FRP composite respectively. The mixture theory
is applied to these composites and the modified mixture theory is used for the FRP composite
itself (Luccioni, 2006; Toledo et al., 2008).

4.1

Constitutive model for solid clay bricks, mortar

Orthotropic plastic-damage models (Betten, 1988; Luccioni et al., 1995; Luccioni and
Rougier, 2005; Luccioni et al., 1996) are used for bricks and mortar in the micro-model. Even
though different materials with different mechanical properties are dealt with, solid clay bricks,
mortar and masonry are frictional materials, that is, their behavior is influenced by hydrostatic
pressure. Bricks are normally made of isotropic materials and mortar is also approximately
isotropic. However, mortar and bricks are modelled as orthotropic materials in order to account
for the weakness of the interfaces and the possibility of relative displacements without explicitly
modeling the interfaces (Luccioni et al., 2005).

4.2

Anisotropy treatment

The anisotropic model is based on the assumption that two spaces can be defined (Betten,
1988; Luccioni et al., 1995): (a) a real anisotropic space and (b) a fictitious isotropic space.
The problem is solved in the fictitious isotropic space allowing the use of elastoplastic models
originally developed for isotropic materials (Luccioni and Rougier, 2013).

4.3

Isotropic model: general form

The plastic process is described by a generalization of classical plasticity theory that takes
into account many aspects of geomaterials behavior (Luccioni et al., 1996). The elastic threshold is described by a yield function,
F (ij , k ) = f p (ij ) K(ij , p ) 0

(1)

where f p (ij is the equivalent stress defined in the tension space and that can take up the form of
any of the yielding functions of classic plasticity ( Tresca, Von Mises, Mohr Coulomb, Drucker
Prager, etc). If this model is used for mortars or bricks a suitable approach for frictional materials as Mohr Coulomb or Drucker Prager must be adopted. K(ij , p ) is the yielding threshold
and p is the plastic hardening variable.
The following rules are used for the evolution of plastic strains:
pij =

G(mn , p )
ij

(2)

where is the plastic consistency factor and G is the plastic potential function.
The plastic hardening variable p is obtained normalizing energy dissipated by the plastic
process to unity and varies from 0, for the virgin material, to 1 when the maximum energy is
plastically dissipated (Luccioni and Rougier, 2005; Luccioni et al., 1996; Rougier and Luccioni,
2007).
CILAMCE 2013
Proceedings of the XXXIV Iberian Latin-American Congress on Computational Methods in Engineering
Z.J.G.N Del Prado (Editor), ABMEC, Pirenpolis, GO, Brazil, November 10-13, 2013

M. Escalante, V. Rougier

The damage threshold is described by a damage function in the following way:

F d = f d (ij ) K d (ij , d ) 0

(3)

where f d (ij ) is the equivalent tension, K d (ij , d is the equivalent damage threshold and is
the hardening variable (Luccioni et al. 1996). The equivalent tension f d (ij may be evaluated
using known yielding functions (Tresca, Von Mises, MohrCoulomb or DruckerPrager) or any
function specially developed for damage (Luccioni et al., 1996; Luccioni and Rougier, 2005).
The scalar damage variable d varies from 0 to dc . that is 0 d dc where 0 dc 1 is
the level of damage correspondent to the material failure.
The evolution of permanent strains and damage is obtained from the simultaneous solution
of the following equations called the consistency conditions of the problem (Luccioni et al.,
1996),

F p = 0
(4)
F d = 0
Eq. (13) are two linear equations in and d

4.4

Composite materials modeling

The reinforcement material made up of polymeric matrix and carbon fibers is itself a composite material formed by a matrix with embedded long fibers. To simplify the numerical simulation and to reduce calculus volume, it was modeled with an equivalent homogeneous model.
An orthotropic elasto-plastic model with the composite properties was used for that purpose.
As the properties provided by the manufacture were not enough to model this material, a generalization of mixture theory (Luccioni, 2006; Toledo et al., 2008) was used to obtain all the
reminding mechanical properties. In this calculus, the properties of the composite were obtained from the properties of the fibers and the epoxy matrix and the fibers volume ratio. In this
way, the lamina properties already given by the manufacture were also verified (Rougier, 2007).

4.5

CFRP reinforced masonry

The CFRP laminas applied on both faces of the masonry panels were not modeled independently but together with brick or mortar. This consideration gives place to a composite material
consisting of brick or mortar and two sheets of composite. In all cases the finite element mesh
should be carefully defined so that the elements match the reinforcement zones.
Mixtures theory can be used to model the in plane behavior of this type of composite where
the strains are the same for both component materials (Rougier, 2007).

PARAMETRIC STUDY

A parametric study is presented in this section. This is based on an existing finite element
modeling (Luccioni and Rougier, 2005) and it was performed on unreinforced panels and panels
reinforced with CFRP band subjected to compression normal to the bed joints. In order to
CILAMCE 2013
Proceedings of the XXXIV Iberian Latin-American Congress on Computational Methods in Engineering
Z.J.G.N Del Prado (Editor), ABMEC, Pirenpolis, GO, Brazil, November 10-13, 2013

Parametric study of the structural behaviour of FRP reinforced masonry

analyze the effects on the load carrying capacity of unreinforced and retrofitted masonry the
influence of different values of some mechanical properties of bricks, mortar and CFRP were
studied.
First, the effect of slight variation of Youngs modulus and compression strength of the units
and mortar on the ultimate load of unreinforced masonry was analyzed. Then, Youngs modulus
of CFRP was varied, while the mechanical properties of bricks and mortar remained constant.
The aim was to understand the global behavior of CFRP retrofitted masonry.

5.1

Influence of the Youngs modulus and compressive strength of the clay

bricks and mortar on load carrying capacity of unreinforced masonry

The results of a numerical study carried out in order to know the influence of the Youngs
modulus and compression strength of the units and mortar in masonry final strength are shown
in this section. Compression normal to bed joints tests were simulated with a non linear plane
finite element program in which the models described in sections 5.1 and 5.2 were implemented.
The results of these simulations are hereafter presented. We recall that only specimens of Group
II (560 x 550 x 125 [mm]3 ) were analyzed.
All the specimens were modeled with three node triangular plane stress finite elements.
Masonry elements were meshed distinguishing bricks and mortar elements. The mechanical
properties of mortar and bricks were experimentally obtained and are summarized in Table 4
(Rougier, 2007; Luccioni et al., 1996).
The finite elements mesh, boundary and loading conditions for masonry panels tested under
compression normal to the bed joints are presented in Fig.6. Taking advantage of the specimen
symmetry only a quarter of them was modeled.

Figure 6: Unreinforced panel under compression normal to bed joints: (a) Load and boundary conditions,
(b) Finite element mesh.

Numerical load - axial displacement curves (P-1 ) of unreinforced masonry panels and their
comparison with experimental results are presented in Figures 7 and 8. Displacements represented correspond to total vertical relative displacements of the panels sides. Variations of a
CILAMCE 2013
Proceedings of the XXXIV Iberian Latin-American Congress on Computational Methods in Engineering
Z.J.G.N Del Prado (Editor), ABMEC, Pirenpolis, GO, Brazil, November 10-13, 2013

M. Escalante, V. Rougier

Table 3: Composite mechanical properties.Specimen:560 x 550 x 125 [mm3]

Properties

Mortar

Brick

Elasticity Modulus, E (MPa)

1528

1400

Poissons ratio,

0.21

0.15

Tension ultimate strength, ut (MPa)

0.54

0.414

Compression ultimate strength, uc (MPa)

8.28

Uniaxial compression elastic threshold, f c (MPa)

3.5

Initial compression/tension strength ratio, R0p

10

20

Plastic damage variable for the peak stress, pcomp

0.20

Fracture energy, Gfp (MPa m)

1.01-5

3.0E-5

Crushing energy, Gc p (MPa m)

1.01-3

2.0E-3

Yield criterion

Mohr Coulomb

Drucker Prager

Plastic flow

Mohr Coulomb

Drucker Prager

Damage Criteria

Drucker-Prager

Drucker-Prager

Friction angle for damage function ()

Uniaxial compression damage threshold, cd (MPa)

3.7

7.5

Damage fracture energy, Gd (MPa m)

6.0E-3

5.0E-2

proportional factor k of the ultimate compressive strength and Youngs modulus of bricks are
shown in Figures 7 (a) and (b), respectively. Then, the same variations of ultimate compression strength and Youngs modulus were considered for the mortar and the load-displacement
diagrams are presented in Figures 8 (a) and (b).
Numerical values of maximum load of masonry corresponding to different values of proportional factor (k) applied to ultimate strength (uc ) and Youngs modulus (E) of bricks and
mortar are shown in Table 4.
It can be seen that maximum load of masonry under uniaxial compression depends chiefly
on the ultimate strength of the brick units (see Fig. 7 a). On the other hand, Youngs modulus of
the bricks has practically no influence on the ultimate load capacity but significantly influences
the stiffness of masonry (Fig. 7 b).
The influence of brick strength in the specimens response to compression normal to bed
joints is also clear in Table 4. The maximum difference of maximum load of masonry is about
15.0% and it corresponds to an increase of compression strength of bricks of 1.2 times.
As a counterpart, variations of ultimate compressive strength and Youngs modulus of mortar produce small variations of load carrying capacity and stiffness of masonry, as can be observed in Fig. 8 (a) and (b) respectively. The maximum difference of ultimate load of masonry
is about 4% and it corresponds to an increase of mortar strength of 1.2 times, as can be seen in
CILAMCE 2013
Proceedings of the XXXIV Iberian Latin-American Congress on Computational Methods in Engineering
Z.J.G.N Del Prado (Editor), ABMEC, Pirenpolis, GO, Brazil, November 10-13, 2013

Parametric study of the structural behaviour of FRP reinforced masonry

(a)

(b)
Figure 7: Axial load versus axial displacement curves for compression normal to the bed joints of unreinforced panels. a) Variation of the ultimate compressive strength of the bricks; b) Variation of the Youngs
modulus of the bricks

Table 4.
Table 4: Maximum load of unreinforced masonry for different values

k factor

Maximum load (kN)

Brick

Mortar

uc

Eb

uc

Em

0.8

188

222.4

212

222.7

0.9

206

222.5

217

222.8

1.0

223

222.6

224

222.6

1.1

239

222.7

228

222.6

1.2

256

222.8

233

222.6

CILAMCE 2013
Proceedings of the XXXIV Iberian Latin-American Congress on Computational Methods in Engineering
Z.J.G.N Del Prado (Editor), ABMEC, Pirenpolis, GO, Brazil, November 10-13, 2013

M. Escalante, V. Rougier

(a)

(b)
Figure 8: Axial load versus axial displacement curves for compression normal to the bed joints of unreinforced panels. a) Variation of the ultimate compressive strength of the mortar; b) Variation of the Youngs
modulus of the mortar

5.2

Influence of the Youngs modulus of CFRP on load carrying capacity

of reinforced masonry.

The panels retrofitted with bands parallel to bed joints and normal to the applied load were
numerically analyzed. The study variable was Youngs modulus of the reinforcement strips.
Finite elements mesh boundary and loading conditions are presented in Figure 9. Since the
panels are symmetric, a quarter of them is modeled too. Triangular plane stress elements with
three nodes were used for the simulations. The load versus longitudinal displacements curves
are presented in Fig. 10. No appreciable increase in ultimate load is observed while stiffness
is kept practically unmodified. However, under in plane shear loading retrofitting with CFRP
increases ductility, prevents the joints sliding and increases the ultimate strength.

CONCLUSIONS

The behavior and strength of unreinforced masonry made of solid clay bricks present a
high variability depending on the mechanical and geometrical properties of units and mortar,
the characteristics of their interfaces, the geometrical arrangement of masonry, the boundary
conditions and the quality of workmanship.
CILAMCE 2013
Proceedings of the XXXIV Iberian Latin-American Congress on Computational Methods in Engineering
Z.J.G.N Del Prado (Editor), ABMEC, Pirenpolis, GO, Brazil, November 10-13, 2013

Parametric study of the structural behaviour of FRP reinforced masonry

Figure 9: Unreinforced panel under compression normal to bed joints: (a) Load and boundary conditions,
(b) Finite element mesh.

Figure 10: Axial load versus axial displacement curves for compression normal to the bed joints of CFRP
retrofitted panels. Variation of the Youngs modulus of the CFRP laminate .

Generally, masonry capacity to support uniaxial compression loading normal to bed joints
is good presenting also a ductile failure mode. From the parametric study performed, it can
be concluded that under compression normal to the bed joints, the ultimate strength of the
brick units is the most influential factor on load carrying capacity of unreinforced masonry.
An increase of 1.2 times of brick strength allows to obtain an increase of about 15% of the
maximum load of masonry.
Due to the cost, extension and complexity that experimental programs may have, it is important to have a numerical tool with which the unreinforced and FRP retrofitted masonry behavior under different in plane stress states can be satisfactorily reproduced. Once the model has
been adjusted, a numerical study allows the analysis of different loading conditions and repair
or reinforcement assemblages, which is translated into smaller number of laboratory tests. At
present, the authors are working on implementing a stochastic model in which the parameters
herein analyzed are considered as random variables.

References
Betten, J., 1988. Application of tensor functions to the formulation of yield criteria for
anisotropic materials, International Journal of Plasticity, vol. 4, pp. 2946.
CILAMCE 2013
Proceedings of the XXXIV Iberian Latin-American Congress on Computational Methods in Engineering
Z.J.G.N Del Prado (Editor), ABMEC, Pirenpolis, GO, Brazil, November 10-13, 2013

M. Escalante, V. Rougier

Gabor, A., Ferrier, E., Jacquelin, E. and Hamelin, P., 2006. Analysis and modelling of the
in-plane shear behaviour of hollow brick masonry panels, Constr Build Mat, vol. 20, pp.
308321.
Grande, E., Milani, G. and Sacco, E., 2008. Modelling and analysis of frp-strengthened masonry
panels, Engineering Structures, vol. 30, pp. 18421860.
Lourenco, P., 1996. Computacional strategies for masonry structures, PhD thesis, Delf University of Technology, Netherlands.
Luccioni, B., 2006. Constitutive model for fiber reinforced composite laminates, Journal of
Applied Mechanics, vol., n. 6, pp. 901910.
Luccioni, B., Lpez, D. and Danesi, R., 2005. Bond slip in reinforced concrete elements, Struct
Eng ASCE, vol. 131, n. 11, pp. 571574.
Luccioni, B., Oller, S. and Danesi, R., 1995. Plastic damaged model for anisotropic materials,
Applied Mechanics in the Americas, vol. I, pp. 124129.
Luccioni, B., Oller, S. and Danesi, R., 1996. Coupled plastic-damaged model, Computer Methods in Applied Mechanics and Engineering, vol. 129, pp. 8189.
Luccioni, B. and Rougier, V., 2005. A plastic damage approach for confined concrete, Computer
& Strutures, vol. 83, pp. 22382256.
Luccioni, B. and Rougier, V., 2010. Shear behaviour of brick mortar interface in cfrp retrofitted
or repaired masonry, Int J Mech Sci, vol. 52, pp. 602611.
Luccioni, B. and Rougier, V., 2011. In-plane retrofitting of masonry panels with fibre reinforced
composite materials, Construction and Building Materials, vol. 25, pp. 17721788.
Luccioni, B. and Rougier, V., 2013. Numerical analysis of fibre reinforced polymer retrofitted
masonry panels, Engineering Structures, vol. 49, pp. 360372.
Oller, S., Oate, E., Miquel, J. and Botello, S., 2006. A plastic damage constitutive model for
composite materials, Int J Solids Struct, vol. 33, n. 17, pp. 20512518.
Page, A., 1978. Finite element model for masonry, Journal of the Structural Division, ASCE,
vol. 104, n. 8, pp. 12671285.
Prakash, S. and Alagusundaramoorthy, P., 2008. Load resistance of masonry wallettes and shear
triplets retrofitted with gfrp composites, Cement & Concrete Compos, vol. 30, pp. 74561.
Rougier, V., 2007. Refuerzo de muros de mampostera con materiales compuestos, PhD thesis,
National University of Tucuman, Argentina.
Rougier, V. and Luccioni, B., 2007. Numerical assessment of retrofitting systems for reinforced
concrete elements, Engineering Structures, vol. 29, pp. 16641675.
Toledo, M., Nallim, L. and Luccioni, B., 2008. A micro-macromechanical approach for composite laminates, Mechanics of Materials vol. 40, pp. 885906.
V., R. and B., L., 2012. Brick and Mortar Research, Nova Science Publisher, New York, USA,
chapter In-Plane Behavior of CFRP Retrofitted Masonry: Experimental and Numerical Assessment, pp. 93132.
CILAMCE 2013
Proceedings of the XXXIV Iberian Latin-American Congress on Computational Methods in Engineering
Z.J.G.N Del Prado (Editor), ABMEC, Pirenpolis, GO, Brazil, November 10-13, 2013