Skip to main content Skip to footer

Flexural Strengthening Of Deteriorated Brick Masonry With Carbon-Fiber-Reinforced Polymer (CFRP) Bars

May 15, 2007

S Y M P O S I U M O N B U I L D I N G E N V E L O P E T E C H N O L O G Y • NO V E M B E R 2 0 0 7 S C H N E R C H • 5 1
This paper focuses on the use of nearsurfacemounted
(NSM) carbon fiberreinforced
polymer (CFRP) bars for rehabilitation and/or strengthening of existing masonry façades.
This type of “structural repointing” is suitable for strengthening historical unreinforced
masonry façades, in addition to contemporary masonry façades with significant corrosion of
the embedded steel reinforcement. Research in this area was first developed in Italy, where
the large number of historical unreinforced masonry buildings, together with a modern
understanding of earthquake forces, has generated considerable interest to reinforce these
structures using noncorroding
materials. The paper discusses guidelines and installation
techniques for the flexural strengthening of masonry façades based on the best practice
reported in the literature and practical experience in applying the technique. Structural
strengthening using NSM CFRP bars is shown to be a viable and costeffective
method to
reinforce walls for outofplane
loads such as wind and earthquake forces.
This topic, involving innovative technologies and practices for the façade structural system,
will be presented at an intermediate level and will be of interest to structural engineers
and architects dealing with concrete and brick masonry structures.
DAVID SCHNERCH has obtained considerable experience on the rehabilitation and
strengthening of structures using fiber reinforced polymer (FRP) materials through his master’s
in civil engineering from the University of Manitoba and his PhD in civil engineering
from North Carolina State University. Since joining the Cambridge office of WJE, he has
designed the FRP rehabilitation system for a 14story
masonry façade. David has published
numerous articles on the bond and flexural design using FRP materials and developed
design guidelines for the strengthening of steel structures with advanced, highmodulus,
carbon FRP materials.
5 2 • S C H N E R C H S Y M P O S I U M O N B U I L D I N G E N V E L O P E T E C H N O L O G Y • NO V E M B E R 2 0 0 7
Many existing historic unreinforced
masonry (URM) buildings exist in parts of
the world that are seismically vulnerable.
polymer (FRP) materials
are gaining acceptance as strengtheners of
these types of structures, due to FRPs’
durability, low weight, and high strength.
Typically, FRP materials are inert when
subjected to deicing salts or many other
corrosive agents. Their high strengthtoweight
ratio not only reduces the dead load
of the structure, but also has constructability
advantages that can offset the relatively
high material costs associated with these
materials. Optical fibers may be embedded
in the FRP material to monitor strains in the
masonry and have been used to monitor
and perform a load test on a historic brick
masonry structure in Italy (Bastianini et al.,
2005). The nonductile
and isotropic nature
of these materials are aspects that must be
recognized when comparing FRP materials
to traditional materials.
All FRP materials are composite, consisting
of at least two different materials,;
one typically comes in the form of a strong
fiber embedded in a resin matrix. The fibers
provide strength and stiffness, while the
resin serves to protect the fibers and to
transfer stress from fiber to fiber, and ultimately,
to the structural element. The most
commonly used fibers for structural applications
are Eglass
and carbon, derived
from either a polyacrylonitrile(
PAN) based
or coaltarpitchbased
precursor. Some
typical mechanical properties of these fiber
types are listed in Table 1. When pultruted
into bars, strips, or thin plates, the
mechanical properties of the finished glass
FRP (GFRP) or carbon FRP (CFRP) material
depend on the ratio of the fibertoresin
cross section (or fiber volume fraction) and,
to a lesser extent, the material properties of
the resin itself. In general, Eglass
and PAN
carbon fibers have similar strength properties,
but carbon has a higher stiffness in
addition to better creep performance.
Early applications of FRP strengthening
to masonry walls were by a wet layup
For this application method, a dry
fiber fabric that may be unidirectional or bidirectional
is used in conjunction with a saturating
resin. The resin is applied to the
substrate, and the reinforcing fibers are laid
into the resin. An additional layer of resin is
applied to the surface. Additional layers of
reinforcing fibers may be built up in this
way, or the fiber orientation may be varied to
achieve the required amount of strengthening
in a given direction. Curing of the resin
may be at ambient or elevated temperatures.
Often, the wet layup
method requires one
or more primer resins to be applied to the
surface of the masonry to penetrate into
pores and voids and also to level the surface
prior to application of the fibers. An outer
coating is also necessary
to cover the finished surface.
Another application technique is by
directly bonding flat FRP plates or strips to
the masonry. These strips are produced by
impregnating the fibers with a resin and
pultruding the uncured fiberresin
through a die into a continuous, uniform,
cross section shape such as a bar or
plate. This process is highly controlled,
allowing a uniform orientation of the fibers,
a higher fiber volume fraction, and a higher
glass transition temperature due to a heated
cure. A roughened surface or peelply
often used to enhance the bond characteristics
of the material and reduce the
amount of surface preparation required in
the field.
Fiber Tensile modulus
(106 psi)
Tensile strength
Strain to failure
10.5 500.0 4.8
Carbon (PAN) 34 530.0 1.4
Carbon (PITCH) 55 275.0 0.5
Table 1 – Properties of commonly used types of fiber (adapted from ACI 440, 2002).
S Y M P O S I U M O N B U I L D I N G E N V E L O P E T E C H N O L O G Y • NO V E M B E R 2 0 0 7
A more recent technique is to bond the
FRP to the structure by means of a nearsurfacemounted
(NSM) technique. This
method typically uses smalldiameter
or rectangular crosssection
bars that may
be embedded within a mortar bed joint or
slot. Whereas the first two techniques alter
the appearance of the wall visually, depending
on the amount of strengthening
required, NSM FRP bars may be placed
within all or some of the bed joints. Slots
may also be cut in the vertical direction. For
concrete structures, the NSM technique has
been found to result in improved bond
between the concrete and FRP adherends
and better protection against fire, vandalism,
and impact.
Especially for application by means of
wet layup
of FRP sheets, the effect of moisture
transport through the wall must also
be considered (Feickert et al., 2003). The
potential exists, given the hydrothermal
environment of the wall for vapor liquefaction
at the interface between the masonry
and the epoxy resin interface. A finite element
study has shown that gaps between
the FRP sheets are necessary. However,
even if the surface of the wall is 60 percent
covered by FRP material and an extreme
combination of temperature and humidity
is assumed, the resulting bond stresses are
less than the tensile strength of CMU by a
factor of three, indicating that adequate
performance can be achieved, even for
these methods.
Under seismic loading, URM masonry
may fail in plane or out of plane. For
masonry structures subjected to wind
loads, typically exterior wall sections will be
loaded out of plane. This paper focuses on
the outofplane
behavior of URM walls
before and after rehabilitation with FRP
materials. Even reinforced masonry buildings
may be suitable for rehabilitation with
FRP if the existing reinforcement has deteriorated
due to corrosion of the cross section
or if errors were made in the original
construction, such as leaving portions of
the cores ungrouted. Deterioration or a deficiency
in the construction of a reinforced
masonry wall may leave the wall vulnerable
to outofplane
S C H N E R C H • 5 3
Previous work has shown that FRP
materials are suitable to enhance the resistance
to both inplane
and outofplane
loads. Strengthening with FRP materials is
highly effective for walls that can be treated
as simply supported. For a wall with a low
slenderness ratio, built between rigid supports,
the FRP is less effective because
arching action of the wall dominates since
crushing of the masonry units at the
boundary regions controls the ultimate
behavior (Tumialan et al., 2002).
Several researchers have investigated
strengthening of masonry walls with FRP
sheets. Ehsani and Saadatmanesh (1996)
found that the flexural capacity of smallscale
URM walls could be greatly increased
with the addition of GFRP sheets. Tested
walls typically failed by crushing of the
masonry. For walls strengthened on both
sides with GFRP sheets and tested in
reversed cyclic loading investigated by
Kuzik et al. (2003), it was found that the
walls retained their integrity through multiple
loading cycles. VelazquezDimas
et al.
(2000) also tested walls under cyclic, outofplane
loading. Some of the walls were
strengthened on both sides with GFRP
strips. It was found that although the URM
and FRP represent brittle materials that are
nearly elastic to failure, the combination of
the two materials resulted in a system capable
of dissipating energy. Aiello and Sciolti
(2006) indicate that simplified theoretical
models utilized to predict the strain distribution
within the composite sheets at the
service conditions give satisfactory results,
assuming mechanical and geometrical
properties are accurately defined.
Galati et al. (2006) prepared wall panels
composed of clay bricks or CMU. These
specimens were strengthened with either
round GFRP bars or rectangular CFRP bars
using epoxyor
grout. Failure modes included
debonding of the FRP reinforcement, crushing
of the masonry in compression, rupture
of the FRP in tension, or shear failure of the
masonry at the supports. The flexural
strengthening has been proven to increase
the flexural capacity of the walls from two to
14 times, depending on the type and
amount of reinforcement used.
Apart from flexural loading, masonry
walls may also be strengthened for inplane
shear loading using FRP materials. A comprehensive
program to investigate the reinforcement
of masonry walls for inplane
loading was conducted by Marshall et al.
(1999). Strengthening with FRP materials
was shown to increase the inplane
of masonry walls and enable greater story
drift before failure occurs. Lower modulus
glass fiber was preferred, due to its greater
ductility. NSM techniques may also be
applied to increase the shear capacity of
URM walls. For walls with GFRP bars
inserted into every bed joint, the shear
increase was 80 percent (Li et al., 2005).
When designing the FRP strengthening,
the designer must take into account the
adhesion between the two adherends.
Limited studies are available pertaining to
the bond of FRP materials to masonry materials,
although considerable work in this
area has been conducted on concrete structures.
For concrete structures, Hassan and
Rizkalla (2004) found that the bond characteristics
of NSM bars or strips in concrete
beams are influenced by the groove dimensions,
groove spacing, and the limited adhesive
cover. The groove dimension also influenced
the failure of the NSM FRP bars and
strips. Other work in concrete performed by
De Lorenzis and Nanni (2002) indicated
that bars used for NSM strengthening
should have some degree of surface deformations
or roughness to prevent interfacial
failure between the FRP and the adhesive.
Between the adhesive and the concrete, the
amount of roughness generated by saw cutting
of the groove was sufficient to prevent
interfacial failure at this interface.
For masonry walls, Roku et al. (1999)
noted that absorption of epoxy is limited in
extruded brick units as compared to molded
brick units, leading to a possible reduction
of the bond strength at the resintomasonry
interface. For masonry walls strengthened
by the NSM technique, the cutting of
the grooves prepares a surface in which the
epoxy may flow into the pores of the brick
more easily. As in concrete structures, the
size of the groove cut into the masonry also
influences the bond behavior. Galati et al.
(2006) observed that round GFRP bars
exhibited a better performance when the
size of the groove was 2.25 times the diameter
for a latexmodified
cementitious paste
or 1.5 times the diameter for epoxy paste.
It is possible to bond FRP to stone, clay,
or concrete masonry. Aiello and Sciolti
(2006) conducted a study of the bond
between FRP sheets and limestone ashlars.
This type of stone is reported as the most
common utilized for the construction of
masonry structures in Italy. Tests were conducted
using either a type of doublelap
shear test or a commercially available machine.
Depending on the type of stone, the
strain distribution, and the bond stressslip
law, the ultimate load was found to vary.
Parameters studied included bond lengths,
anchoring of the FRP sheets with transverse
layers, and changes in the specimen configuration.
The type of substrate was found to
be more significant than increasing the development
length from 70 to 150 mm. The
mechanical strength of the stone had an
influence on the bond strength–both the
ultimate and the service conditions. Interface
deformability was determined to be very
low, indicating the effectiveness of the system.
Failure of URM walls occurs when the
masonry reaches its tensile strength limit.
The tensile strength of masonry is relatively
low and is associated with a brittle failure
mode. A masonry wall strengthened with
FRP materials will not necessarily fail once
the masonry reaches its tensile strength
limit, since FRP reinforcement is much
stronger in tension than the masonry. In
addition to allowing greater tensile strain
prior to failure, cracking occurs with smaller,
flexural cracks formed
before failure.
Guidelines for the strengthening of
masonry structures have been developed
based on limit state principles (Galati et al.,
2005). This flexural design process investigates
several different failure modes and
limit states that are consistent with the
building code requirements for masonry
structures and are consistent with code
requirements established by ACI and TMS.
Governing failure modes for the design may
be the crushing of the masonry prior to rupture
of the FRP material or the rupture of
the FRP material. The debonding failure
mode, due to its inherent nonductile
nature, must also be avoided due to the
brittle nature of this failure mode. Typically,
strengthening limits are established so that
the structure must support its nominal service
load in the event that the FRP becomes
As with steelreinforced
masonry reinforced with NSM FRP follows
similar design assumptions. Plane sections
have been shown to remain plane and the
tensile strength of the masonry is neglected
in the design. As has been shown in the
research, the bond between the FRP and
the masonry is assumed to be perfect, with
no slip, until debonding failure. Finally, the
wall can be assumed to behave under simplysupported
5 4 • S C H N E R C H S Y M P O S I U M O N B U I L D I N G E N V E L O P E T E C H N O L O G Y • NO V E M B E R 2 0 0 7
The design procedure begins with the
properties of the FRP material. The limit on
the FRP strength is based on the guaranteed
tensile strength of the FRP provided by
the manufacturer. This strength is reduced
by a factor based on the environmental
exposure, which in turn is a function of the
type of fiber. Another reduction is used to
limit the potential for debonding at the ultimate
strength. For walls subjected to sustained
loads, consideration of creep rupture
should be taken into account to ensure the
tensile stress in the FRP does not exceed
values given by ACI 440 (2002).
As masonry materials exhibit nonlinear
behavior in compression, the stress distribution
for the portion of the masonry in compression
may be represented by an appropriate
representation. The stress distribution
used should be appropriate given
the level of strain in the masonry. For example,
if the FRP reaches its rupture strain
prior to the masonry reaching its crushing
strain, then the stress in the masonry will be
less than what is predicted by the rectangular
stress block at the masonry ultimate
strain. The maximum strain in the FRP is
assumed to be 0.0035 for clay masonry and
0.0025 for concrete masonry.
Apart from a debonding failure mode,
failure by masonry crushing may occur for
walls with a high FRP reinforcement ratio.
For walls with a low reinforcement ratio,
failure will be by FRP rupture. Masonry
crushing is a more desirable failure mode,
since it provides greater ductility. However,
the guidelines accept both failure modes,
provided that the appropriate strength and
serviceability criteria are satisfied. As with
any strengthening design, it is important to
consider if alternate failure modes (such as
shear failure) become prevalent. The guidelines
only consider the design of walls with
strengthening applied to a single side, since
most of the experimental work to this point
has considered this configuration.
Calculation of the flexural capacity of
the wall is based on conditions of force equilibrium
and strain compatibility. The failure
mode of the wall is determined by comparing
the balanced FRP reinforcement ratio to
the selected FRP strengthening ratio. For a
masonry crushing failure, when the FRP
reinforcement ratio is greater than the balanced
ratio, the stress distribution in the
masonry can be approximated by a rectangular
stress block. When the FRP reinforcement
ratio is less than the balanced ratio,
an iterative approach is necessary to determine
the equivalent stress block. For bearing
walls, the axial load is taken by summing
the ratio of the axial load to the axial
capacity and the flexural load to the flexural
capacity. The sum of this equation
should be less than unity.
When a wall is built in between supports
that restrain outward movement,
membrane compressive forces act in the
plane of the wall. After cracking, this results
in arching action that increases the capacity
of the wall relative to the same configuration
with simple boundary conditions. In
this case, it is recommended that one consider
the arching mechanism to avoid overestimating
the strengthening, since crushing
of the masonry controls the failure.
Furthermore, it is stated that the capacity
of an unstrengthened URM wall with end
restraints is far superior to an identical simply
supported wall with FRP strengthening.
Adhesives are used to bond the FRP
strengthening material to the masonry substrate.
Due to the highly anisotropic nature
of composite materials, mechanical fasteners
cannot be used. However, just as with a
bolted connection, a bonded joint must be
designed through proper selection of the
adhesive, preparation of the surfaces to be
adhered, and detailing of the joint geometry.
The adhesive must have sufficient strength
to prevent cohesive failure of the joint and
have a viscosity that will allow it to generate
adhesive forces through mechanical interlocking
to both the masonry and FRP substrates.
Even within a type of adhesive such
as epoxies, the viscosities and pot life vary
greatly. Both of these properties will affect
how the adhesive interacts with the
adherend. Proper surface preparation is
fundamental. Improper surface preparation
may lead to an adhesive type of failure
between the substrate and the adhesive. Li
et al. (2005) reports that debonding at the
interface is the predominant
failure mode for failure of the NSM
FRP bars in masonry.
Rather than attempting to predict bond
failure, Galati et al. (2006) suggested limiting
the FRP strain as a means to control
bond stresses. The effective FRP strain is
given by the product of the design rupture
strain of the FRP reinforcement and a bonddependent
coefficient. Recommendations
for these coefficients were also developed
based on the results of this and an earlier
study by Turco et al. (2003).
The designer must understand the
importance of generating adequate bond
between the FRP and the masonry and the
requirements for achieving this bond. Direct
contact between the adhesive and the
adherends is necessary to achieve adequate
bond, as is the removal of weak layers or
contamination at the surface. Understanding
at the outset all of the steps
involved in the strengthening is necessary,
since the FRP material should be installed
as soon after the surface preparation is
completed as is practical.
The FRP material must be shipped to
the site in a rigid, sealed container that prevents
impacts from damaging the FRP
material. Inspection of the FRP before its
use should reveal any nicks or cuts that
would cause the material to be rejected. If
replacement material is required, it should
be obtained before the surface preparation
of the masonry is completed. It is necessary
to store FRP materials off the ground without
exposure to extremes in temperature or
direct sunlight. The cleanliness of the material
is essential to ensure adequate bond.
Cutting of the slots in the masonry can
be accomplished with a grinder that cuts
the entire slot at once with a constant
width. Methods to ensure the straightness
and consistent depth of the groove should
be verified on a mockup
panel before
installation proceeds on the façade. The
slotting process creates a considerable
amount of dust, and dust collection systems
or protective enclosures should be
considered. Following cutting of the
grooves, the grooves must be thoroughly
cleaned out to remove any dust that may
affect the interface between the adhesive
and the masonry. Blowing with compressed
air may merely redistribute dust from one
location to another. The preferred method is
to vacuum and brush the groove to thoroughly
collect and remove the dust. As part
of the preparation before bonding, the bars
should be cut to length, generally by hand
with a finetooth
blade. Most types of bars
have a roughened or indented surface that
results in good mechanical bond to the
Most epoxies consist of two parts that
should be mixed in accordance with the recommendations
of the manufacturer. The
adhesive can then be applied directly into
the groove, as shown in Figure 1. Care
should be taken to ensure that air voids are
not trapped by the adhesive. Once the adhesive
is placed into the void, the bar is
S Y M P O S I U M O N B U I L D I N G E N V E L O P E T E C H N O L O G Y • NO V E M B E R 2 0 0 7 S C H N E R C H • 5 5
pressed into the adhesive and additional
adhesive is placed to completely embed the
bar. If it is so desired, the adhesive may be
tooled to achieve an appearance similar to
the surrounding mortar lines.
Considerable research efforts have
shown that FRP materials may be used for
strengthening URM to resist outofplane
and other types of loadings. Not only is the
strength of these walls considerably
enhanced, but strengthened walls exhibit
improved ductility due to the increased
strain capacity of the cross section. Even
under reversed cyclic loading simulating
seismic loading, FRPreinforced
walls maintain
their integrity after multiple loading
cycles. Design guidelines have been proposed,
but additional work is required to
fully comprehend the behavior of the
bond. Analytical procedures
are necessary to quantify bond
stresses and to define standard bond tests
that can yield reliable design data, given the
great variety in the types and qualities of
masonry walls.
Aiello, M.A., S.M. Sciolti, “Bond Analysis
of Masonry Structures Strengthened
with CFRP Sheets,” Construction and
Building Materials, V. 20, No. 12,
2006, pp. 90100.
American Concrete Institute (ACI).
Committee 440. Guide for the Design
and Construction of Externally
Bonded FRP Systems for Strengthening
Concrete Structures, October
Bastianini, F., M. Corradi, A. Borri, A.D.
Tommaso, “Retrofit and Monitoring
of an Historical Building Using
“Smart” CFRP with Embedded
Fibreoptic Brillouin Sensors,”
Construction and Building Materials,
V. 19, No. 7, 2005, pp. 525535.
De Lorenzis, L., and A. Nanni, “Bond
Between NearSurface
Polymer Rods and
Concrete in Structural Strengthening,”
ACI Structural Journal, V. 99
No. 2, Mar.Apr.,
pp. 123132,
Ehsani, M. R. and H. Saadatmanesh.
“Seismic Retrofit of URM Walls with
Fiber 16 Composites,” The Masonry
Society Journal, V. 14, No. 2, 1996,
pp. 6372.
Feickert, Carl, A., Mark W. Lin, Jonathan
C. Trovillion, Ayo O. Abatan, and
Justin B. Berman, “Hygrothermal
Modeling in the Application of FiberReinforced
Polymers for Structural
Upgrade of Unreinforced Masonry
Walls,” U.S. Army Corps of Engineers,
Engineer Research and Development
Center Report, TR0320,
September 2003, 68 p.
Galati, N., E. Garbin, G. Tumialan and
A. Nanni, “Design Guidelines for
Masonry Structures: OutofPlane
Loads,” ACI SP23016,
October 1,
2005, pp. 269288.
Galati, N., G. Tumialan, A. Nanni,
“Strengthening with FRP Bars of
URM Walls Subject to OutofPlane
Loads,” Construction and Building
Materials, V. 20, No. 12,
2006, pp.
Hassan, T., and S. Rizkalla, 2004,
“Bond Mechanism of NSM FRP Bars
for Flexural Strengthening of Concrete
Structures,” ACI Structural
Journal, V. 101, No. 6, Nov.Dec.,
pp. 830839.
Kuzik, M.D., A.E. Elwi, J.J.R. Cheng,
“Cyclic Flexure Tests of Masonry
Walls Reinforced with Glass Fiber
Reinforced Polymer Sheets,” Journal
of Composites for Construction, V. 7
No. 1, 2003, pp. 2030.
Li, Tong, Nestore Galati, J. Gustavo
Tumialan, and Antonio Nanni, “Analysis
of Unreinforced Masonry Concrete
Walls Strengthened with Glass
Polymer Bars,” ACI
Structural Journal, V. 102, No. 4,
2005, pp. 569577.
Marshall, O.S., S.C. Sweeney, and J.C.
Trovillion, “Seismic Rehabilitation of
Unreinforced Masonry Walls,” Fourth
International Symposium on FiberReinforced
Polymer Reinforcement for
Reinforced Concrete Structures, SP138,
American Concrete Institute,
1999, pp. 287295.
Roko, K., T.E. Boothby, and C.E. Bakis,
“Failure Modes of SheetBonded
Fiber Reinforced Polymer Applied to
Brick Masonry,” Fourth International
Symposium on FiberReinforced
(FRP) for Reinforced Concrete
Structures, Baltimore, Maryland,
Nov. 1999, pp. 305311.
Tumialan, J.G.; N. Galati, S.M.
Namboorimadathil, and A. Nanni,
“Strengthening of Masonry with FRP
Bars,” Third International Conference
on Composites in Infrastructure (ICCI
2002), San Francisco, CA, June 1012,
2002, 11 p.
Turco, V., N. Galati, J.G. Tumialan, and
A. Nanni, “Flexural Strengthening of
URM Walls with FRP Systems,”
Proceedings of the International
Symposium FRPRCS6,
July 2003.
J., M. Ehsani, and H.
Saadatmanesh, “OutofPlane
of Brick Masonry Walls
Strengthened with Fiber Composites,”
ACI Structural Journal, V. 97,
No. 3, MayJune
2000, pp. 377387.
5 6 • S C H N E R C H S Y M P O S I U M O N B U I L D I N G E N V E L O P E T E C H N O L O G Y • NO V E M B E R 2 0 0 7
Figure 1 – Application of adhesive into groove.