Composites Part B 127 (2017) 121e132

Contents lists available at ScienceDirect

Composites Part B
journal homepage: www.elsevier.com/locate/compositesb

Test methods for Textile Reinforced Mortar systems

Stefano De Santis a, *, Francesca Giulia Carozzi b, Gianmarco de Felice a, **, Carlo Poggi b
a
Roma Tre University, Department of Engineering, Via Vito Volterra 62, 00146 Rome, Italy
b
Politecnico di Milano, Department of Architecture, Built environment and Construction engineering, Piazza Leonardo Da Vinci, 32, Milan, Italy

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

Article history: Textile Reinforced Mortar (TRM) composites, also named Fabric Reinforced Cementitious Matrix (FRCM),
Received 15 December 2016 comprise high strength textiles embedded into inorganic matrices, and have been recently developed for
Received in revised form the repair and rehabilitation of structures. Their effectiveness and compatibility with the substrate
24 February 2017
fostered the industrial development of a wide number of strengthening systems, which have been
Accepted 9 March 2017
Available online 11 March 2017
applied to masonry and reinforced concrete structures. Nevertheless, an improved knowledge still needs
to be gained on their mechanical properties (tensile and bond behaviour) and on the effect they have on
the structural performance of reinforced members. Furthermore, except for the US, no standards are
Keywords:
Mechanical properties
available for testing, qualification and design. In this paper, the main features of the tensile and bond
Mechanical testing behaviour of TRM composites are described. Recommendations for performing direct tensile tests and
Test recommendations shear bond tests are proposed, based on the outcomes of a Round Robin Test organized by the Rilem TC
250-CSM (Composites for the Sustainable strengthening of Masonry) and Assocompositi (Italian Industry
Association for Composite Materials), carried out on 26 TRM systems (made of basalt, carbon, steel, glass,
PBO and aramid textiles, with cement, lime or geopolymer mortar matrices), and involving 19 European
research institutions and 11 industrial partners.
© 2017 Elsevier Ltd. All rights reserved.

1. Introduction matrix dates back to the late 1840s, at the origin of reinforced
concrete. Josef-Louis Lambot proposed the use of iron wires set into
Textile Reinforced Mortar (TRM) composites are an innovative a cement mortar for the construction of boats, water tanks, etc.
and particularly promising solution for the repair and strength- Nearly one century later, Pier Luigi Nervi effectively employed the
ening of structures. They comprise a high strength textile that is ferrocemento, a sort of reinforced mortar consisting of a cement
applied to the outer surface of the structural members by a mortar plaster applied over one or more layers of a thin steel net, to build
matrix. Textiles are made out of either continuous fibres of carbon thin, hard shell structures for roofs or water tanks. His patent was
[1], glass [2], basalt [3], and PBO [4], arranged in the form of open registered in 1943 during the self-sufficient period in Italy and this
meshes, or steel cords or ropes [5]. Depending on the properties of technique was extensively used for a number of well-known en-
the substrate and on the target performance levels foreseen in the gineering works. More recently, a similar technology, known as
design, cement, lime or geopolymer mortar matrices (possibly Textile Reinforced Concrete (TRC), was developed [6]. TRC systems
enriched with short fibres and/or polymeric additives) can be used. make use of alkali resistant glass, carbon or aramid fabrics, com-
Since many TRM systems comprise a cementitious mortar, the bined with high performance finely grained cement concrete and,
acronym FRCM (Fabric Reinforced Cementitious Matrix) is also thanks to their high strength in both compression and tension, are
widely used in the scientific literature. Differently, the acronym SRG precast to build thin structural elements, such as permanent
(Steel Reinforced Grout) is generally adopted for the systems with formworks, façades and pedestrian bridges.
steel textiles. In the last decade, TRM systems have been proposed and
The idea of embedding a high strength mesh into a mortar developed as externally bonded reinforcements [7]. They offer the
same advantages of Fibre Reinforced Polymers (FRPs), such as high
strength-to-weight ratio, relatively fast and easy installation, and
* Corresponding author. versatility, making them an effective, practical and cost-efficient
** Corresponding author. solution for the upgrade, seismic retrofitting, and repair of struc-
E-mail addresses: [email protected] (S. De Santis), gianmarco.
tures [8e10]. The inorganic matrix employed in TRMs in place of
[email protected] (G. de Felice).

http://dx.doi.org/10.1016/j.compositesb.2017.03.016
1359-8368/© 2017 Elsevier Ltd. All rights reserved.
122 S. De Santis et al. / Composites Part B 127 (2017) 121e132

the epoxy resins of FRPs provides better behaviour at high tem- output data, graphs). Test report forms were prepared to this aim,
peratures, easier and faster installation on uneven or wet sub- to be filled by the participants. Within this framework, details of
strates, and no risks for the workers caused by toxic volatile geometry, particular curing conditions, specific setup and instru-
compounds. The use of lime-based mortars ensures the additional mentation slightly changed from laboratory to laboratory.
benefits (vapour permeability and physical/chemical compatibility In this work, the general features of the tensile behaviour of
with masonry substrates) required for applications to historic TRM composites and of the TRM-to-substrate bond performance
structures. Nevertheless, the bond strength of TRM composites may are presented and the influence of experimental setup detailing on
be lower than that of FRPs. In fact, the bond behaviour is itself more test results is discussed. Based on the outcomes of the Round Robin
complex, as failure may occur not only by cohesive debonding Test, recommendations for performing direct tensile and shear
within the substrate, but also by slippage of the textile within the bond tests are proposed, that specify (i) specimen geometry,
matrix, or by detachment at the textile-to-matrix or mortar-to- manufacturing and curing, (ii) experimental setups, test execution
substrate interface [11,12]. Finally, the effectiveness of TRM re- and measurement methods, and (iii) main data and results to be
inforcements strongly relies on the accuracy of the installation and provided in the test report, which may be useful useful for both
on the curing conditions of the mortars, which may be hardly product qualification/certification and engineering design
controllable in the construction site. purposes.
The last decade has met an increasing interest of the scientific
and professional communities towards the use of TRM composites 2. Tensile behaviour of TRM composites
for structural rehabilitation purposes, fostering both the research
and the industrial development of an increasing number of rein- A number of studies have been carried out in the recent years to
forcement systems, which are nowadays available in the market investigate the tensile behaviour of TRM composites and to develop
and applied in the field. Nevertheless, the knowledge gained so far suitable testing methods [2,20e22]. The research on TRM has
on their (complex) mechanical behaviour still appears incomplete. benefitted from the investigations carried out on TRC [6], for which
A better understating needs to be developed on the influence of the a recommendation on test methods has recently been issued [23].
layout of the textile, the coating/preimpregnation of the fibres, the In the US, AC434 [13] provides a guidance for performing direct
mechanical characteristics of the mortar matrix, and the properties tensile tests for the acceptance of TRM systems. In Europe, stan-
of the substrate (strength and stiffness, roughness, humidity), not dardization committees have been established to develop national
to mention the long-term durability in aggressive environment. and European guidelines.
Furthermore, with the only exception of the US documents AC434 Within the Round Robin Test, the organizers provided some
[13] and ACI 549.4R-13 [14], no standards are available that provide recommendations for direct tensile tests. The thickness of the
instructions for the mechanical testing of TRM composites and for specimens was set equal to 10 mm, but the width could range from
the design of strengthening works. Apart from the formal obstacle 40 mm to 100 mm, depending on the layout of the textile and the
met in some Countries (e.g., Italy) due to the absence of an official testing setup. Each laboratory could choose how to grip the ends of
qualification/certification, the lack of standardized experimental the specimens and how to measure the strain. The stress-strain
procedures may affect the reliability of the mechanical properties response curve had to be provided, together with the main data
(average and characteristic tensile and bond strength, stiffness (peak stress and corresponding strain, tensile modulus of elasticity,
before and after crack occurrence, expected failure modes) etc.), the crack pattern, and the failure mode. The classification of
assumed in the design, and compromises the possibility of the failure modes was also agreed a priori by the participants.
assessing the actual safety level of the reinforced structure.
With the aim of contributing to the current knowledge on the 2.1. Response of TRM composites under tensile loading
mechanical properties of TRM strengthening systems and to the
development of standardized experimental procedures, a Round Tensile tests are generally carried out on prismatic specimens
Robin Test (RRT) was organized by the Rilem Technical Committee comprising one or more layers of fabric embedded in the mortar
250-CSM (Composites for the Sustainable Strengthening of Ma- matrix (Fig. 1). Different shapes have also been experimented in
sonry) and Assocompositi (Italian Industry Association for Com- studies on Textile Reinforced Concrete, such as the bone-shaped
posite Materials). The RRT involved 19 research institutions from 6 one, which is larger in the gripping areas [24], or the dumbbell
European Countries (France, Germany, Greece, Italy, Poland, one, which is thicker in the gripping areas [25]. Waisted specimens
Portugal and Slovenia), and 11 industrial partners (ten Italians and are used to promote failure in the central region, but are more
one Greek). Twenty-five TRM composites were investigated, complex to fabricate and do not replicate the rectangular layout of
comprising basalt, carbon, glass, PBO, aramid, and steel textiles, the fabric, so existing guidelines [13,23] recommend rectangular
embedded into cement, lime and geopolymer mortar matrices. The specimens.
overall investigation included 66 test series, each of which The load is applied by a tensile testing machine under
comprised five direct tensile tests and five shear bond tests. displacement control and recorded by a load cell. The stress is then
Direct tensile tests and shear bond tests were carried out to conventionally derived by dividing the load by the cross section
derive the fundamental mechanical properties of TRM systems, area of the dry textile, to prevent results from being affected by the
such as stiffness, crack pattern, and tensile strength under tensile unavoidable variation of the thickness of the mortar. The cross
loading, and composite-to-substrate load transfer capacity and section area of the textile is calculated as its design thickness
failure mode under shear, ruling the structural performance of a multiplied by its fichticious width, this latter being the product of
broad range of applications. The experimental results of the RRT are the number of yarns/cords by their spacing. Strains are either
presented in detail in five papers organized by textile material, calculated as the displacement recorded by transducers (e.g., po-
namely: basalt [15], carbon [16], glass [17], aramid and PBO [18], tentiometers, LVDTs, etc.) divided by the gauge length of the device
and steel [19]. In order to reduce variability and allow comparisons, (Fig. 1a), or provided directly by an extensometer (Fig. 1b). In both
the organizers provided general instructions for both direct tensile cases, an average strain measure is derived along the gauge length
tests and shear bond tests regarding specimen manufacturing of the instrument, which should be large enough to include a sig-
(geometry, curing), test execution (experimental setup, displace- nificant number of cracks in order to correctly capture the response
ment rate) and experimental outcomes to be collected (main of the specimen up to failure. The use of strain gauges to record the
 S. De Santis et al. / Composites Part B 127 (2017) 121e132 123

Fig. 1. Experimental setup for direct tensile tests on TRM composites provided with GFRP wrapping (a) or tabs (b) in the gripping areas and with potentiometers and 50 mm gauge
length extensometer (a) or 200 mm gauge length extensometer (b) for displacement/strain measurement.

local strain in one cross section is inadequate in case of multi-

cracking behaviour. The application of non-conventional mea-
surement techniques, such as the Digital Image Correlation (DIC)
has been experienced [26e28], and their integration with tradi-
tional methods is currently under investigation.

2.1.1. Response stages

The tensile stress-strain behaviour is generally characterized by
three response Stages (Fig. 2). In Stage I, the mortar is uncracked
and significantly contributes to both the stiffness and the load
bearing capacity of the composite. The tensile modulus of elasticity
of the TRM (EI) can be estimated referring to an equivalent
homogenised cross section, assuming a perfect matrix-to-textile
bond. The occurrence of the first crack in the matrix identifies the
beginning of Stage II. The transition point between Stages I and II is
identified by the values of the stress sI and of the strain εI. The
larger are the tensile strength of the matrix and the mortar-to-
fabric bond/interlocking, the higher is sI. Since stresses are
conventionally referred to the cross section of the dry textile,
varying the density of the textile proportionally modifies the stress
(and, consequently, the stiffness) of Stage I. Therefore, this stage
mainly governs the response of TRM composites comprising sparse
textiles with small design thickness (e.g. PBO), and high-
performance matrices (e.g., fibre reinforced cement mortars with
polymeric additives) [16,18].
During Stage II, the crack pattern progressively develops.
Fig. 2. TRM tensile response Stages.
Generally, the formation of a crack produces a load drop such that
the response curve in this stage is uneven and the stiffness EII needs
to be derived through a linear interpolation in a suitable range of global displacement produces the widening of existing cracks. The
data. To this aim, the length of the specimen that is monitored by slope of the response curve is called EIII. Stage III ends with the peak
the displacement/strain transducer needs to be large enough to ft and the corresponding strain εt. Stage III mainly governs the
include a sufficient number of cracks. Since the crack pattern may tensile behaviour of mortar-based composites comprising a rela-
widely change depending on the properties of the TRM composite, tively large volume fraction of a stiff textile (e.g., most of Steel
EII needs to be evaluated based on the cracks that actually devel- Reinforced Grout systems). When very dense textiles are combined
oped and on the response curve obtained. If only one crack appears, to relatively weak/deformable matrices, the first two stages may
EII may be not identifiable. EII is often much lower than EI since even become hardly identifiable [20]. In Stage III, the stiffness and
cracking is associated to relatively large strain and increases with the load bearing capacity of the composite rely on those of the
limited stress growth. After the crack pattern has completely textile, and stiffness and tensile strength are therefore close to
developed, a stiffer response is observed and the increase of applied those of the fabric alone. Nevertheless, in some cases, a higher
displacement produces a nearly linear stress growth. This Young's modulus and a lower peak strain with respect to the dry
conventionally marks the beginning of Stage III. The intersection textile could be obtained, which is mainly due to two reasons. First,
between the linear regression line of Stage II and the response the mortar matrix between subsequent cracks may provide some
curve of Stage III identifies the transition point (εII,sII). contribution to the stiffness and may allow for a better stress
In Stage III, no further cracks develop and the increase of the
124 S. De Santis et al. / Composites Part B 127 (2017) 121e132

distribution and an increment in the peak stress. Secondly, in tests tests that should be excluded from result statistics (at least for the
on dry textiles, an unhomogeneous stress distribution amongst the peak values) are those in which the tensile failure of the textile was
yarns or amongst the wires/filaments within each yarn, or also the preceded by its slippage.
occurrence of localised damage in the gripping areas, may lead to In a dry (or coated) textile, the load is transferred from the
the premature rupture of some of them. Therefore, the apparent mortar matrix to the external wires of the yarn (or to the coating
gain in strength when passing from dry textile to TRM specimens layer) by chemical bond and from these wires to the inner core by
could be attributed to an underestimate of the actual tensile friction because the mortar does not fully impregnate the inner
strength of the textile when tested alone [16]. The mismatch be- wires. Since the matrix-to-wire stress transfer capacity is higher
tween the tensile behaviour of dry textiles and TRM composites than the wire-to-wire one, a telescopic pull-out failure may occur
depends on the fabric-to-mortar bond/interlocking, and is higher [6]. Such telescopic failure, deeply investigated on TRC, was
for dry textiles made of several filaments (e.g., those of carbon, PBO, observed also in TRM systems with basalt [15], carbon [16], glass
glass, or basalt) and lower for preimpregnated or coated ones, and [17], and PBO [18] dry fabrics (i.e., not preimpregnated). Steel tex-
for steel textiles. tiles, of course, do not exhibit such behaviour, because they do not
comprise filaments [19].

2.1.2. Failure modes and gripping methods

The failure modes observed in direct tensile tests include the 2.1.3. Design parameters
rupture of the textile, near the gripping area (Fig. 3a) or in the Tensile tests provide some of the fundamental parameters for
middle of the specimen (Fig. 3b), or its slippage within the matrix at the design of externally bonded reinforcements with TRM com-
the gripping area (Fig. 3c). The type of failure mainly depends on posites. As for the Service Limit State, the crack spacing and the
the gripping method. If the specimen is gripped in the wedges of stress of the first crack (sI) may be assumed as design parameters,
the testing machine with a sufficient compressive stress (clamping since the textile is exposed to the external environment when
grip), the slippage of the textile is prevented and a full mechanical cracks develop, affecting the durability of the reinforcement. The
characterization of the tensile response of the system may be ob- saturation crack spacing is also determined as the average distance
tained. A suitable reinforcement by means of composite tabs (e.g., between cracks in Stage III, while the maximum average crack
GFRP) or wrapping is needed to avoid mortar crushing and to width is estimated as the displacement at collapse divided by the
ensure a uniform stress distribution [12,20]. Failure by slippage of number of cracks.
the textile before its tensile strength is reached may occur if the As for the Ultimate Limit State (ULS), the stiffness in Stage III and
gripping pressure is insufficient, or absent by choice, as with the the tensile strength and strain are the fundamental design pa-
clevis grip method, in which two metal plates are glued to the ends rameters for applications in which the tensile strength of the
of the specimens, and the tensile load is transferred only by shear composite may be fully exploited (the rupture of the textile is ex-
stresses [21]. pected). These applications include the confinement of pillars or
The recommendations for tensile tests on TRC [23] state that if a columns, and the reinforcement of structural members (e.g., rein-
final crack occurs in the transition zones the stress-strain relation forced concrete beams in bending or shear and the extrados
has to be considered valid only up to the beginning of crack local- strengthening of masonry arches/vaults) with mechanical end an-
ization. The research carried out within the Round Robin Test on chors or pivots. In these cases, installation details (e.g., smoothing
TRM indicated that the peak loads associated to this failure mode of the corners, overlap length, etc.) are crucial to avoid the pre-
are comparable to those attained when the main crack develops in mature rupture of the fibres caused by shear stress concentrations
the middle of the specimen. Nevertheless, whilst for some systems or the inter-laminar debonding, thus ensuring the full exploitation
(e.g., SRG), the extensometer/displacement transducers provide of textile strength. In a broader range of applications, instead, the
reliable strain values because a sufficient number of cracks devel- load is transferred from the structural member to the textile by
oped within their base length [19], the tensile stiffness may result shear/adhesion through the matrix. In these cases, failure is ex-
overestimated if cracks appear only near the gripping areas (i.e., out pected to occur by detachment or debonding of the TRM from the
of the measurement segment of strain/displacement transducers). substrate, and the tensile strength of the textile cannot generally be
In the mechanical characterization of TRM composites, the only attained. In this case, ULS design parameters need to be derived

Fig. 3. Failure modes in direct tensile tests: tensile rupture of the textile near the gripping area (a), tensile rupture of the textile in the middle of the specimen (b), textile slippage at
the gripping area (c).
 S. De Santis et al. / Composites Part B 127 (2017) 121e132 125

through shear bond tests, to be carried out on the substrates (5) The overall length of the coupon should be at least 400 mm.
adopted for their application (concrete, masonry, tuff) [22]. AC434 It includes the gripping areas (i.e., the ends of the coupons
[13] recommends the use of a clevis type gripping in direct tensile that are clamped to transfer the tensile load, minimum
tests, which leads to the failure by textile slippage, in order to 60 mm each), the transition zones (the portions near the
reproduce the as-installed TRM behaviour without the need for gripping areas in which the stress state is influenced by the
running bond tests for acceptance purposes. Nevertheless, the re- clamping pressure, not shorter than B) and a central area in
sults provided by clevis type gripping may be sensitive to the which strains/displacements are measured (minimum
anchorage length, and the correlation with the actual stress transfer 200 mm).
mechanism experienced by TRM composites in field applications (6) The textile is placed with the direction of the fibres parallel to
appears not straightforward. the longitudinal specimen axis and symmetrically in both
the width and the thickness of the coupon.
(7) During manufacturing, particular attention should be paid to
the regularity of the specimen as well as to the planarity and
2.2. Experimental method for direct tensile tests parallel arrangement of fibre bundles/cords. To this aim, a
slight pre-tensioning (10 N/mm2 or less) may be applied to
Based on the knowledge developed by recent research studies the textile during mortar curing. The specimen shall be
and within the RRT, general recommendations for direct tensile measured in at least three different sections with in-
tests on TRM strengthening systems are proposed herein. Direct struments having accuracy within 1% of the sample di-
tensile tests provide the response of a TRM composite, subjected to mensions. Measurements shall be reported and should not
a tensile load parallel to its longitudinal axis in the quasi-static differ from the nominal ones by more than ±10% for the
regime. Test outcomes include the stress-strain curve, the numer- width and ±5% for the thickness.
ical data (stress, strain, modules of elasticity in the response stages) (8) Curing should last at least 28 days. In the curing phase,
and the crack pattern, which are necessary for the design of an adequate relative humidity conditions should be ensured to
externally bonded TRM reinforcement. Tests are meant to be car- prevent the development of differential shrinkage, which
ried out on a TRM system comprised of a textile (with a given could cause cracking and/or bending distortion that may
layout/grid spacing, surface mass density, and installed in a given affect test results. Before testing, the specimens should be
number of plies) and a mortar matrix supplied by the same stored in standard laboratory conditions (20e25
C and
producer. 50e60% R.H.) for at least 7 days.
A. Specimen geometry and manufacturing
B. Test execution
(1) Tensile tests are carried out on prismatic specimens with
rectangular cross section, comprising one or more layers of (1) Specimen dimensions are carefully checked before testing.
fabric embedded in the mortar matrix (Fig. 4). Visible pre-damage (e.g., cracking, local detachment, de-
(2) The specimens are casted in wooden, metal or aluminium flections) has to be documented.
moulds and can be either manufactured individually or cut (2) Tests are carried out with a tensile test machine, under
from larger panels. displacement control. The control parameter is the
(3) The thickness of the specimen should be 10 mm or equal to displacement measured by the integrated LVDT or by another
the thickness required in the installation manual and should transducer applied to the specimen, provided that its base
be kept constant along the entire length of the specimen. length is at least twice longer than the expected saturation
Thickness larger than 30 mm should be avoided. crack spacing (average distance between cracks in Stage III).
(4) The width of the specimen (B) should be at least 40 mm and The displacement is increased monotonically up to failure.
not larger than 100 mm. It should be an integer multiple of The initial rate should be not higher than 0.3 mm/min, to
the grid spacing of the textile and should include at least two record accurately the first phase of the test. After the
yarns/cords.

WIDTH (B)
40-100mm
TABS/REINFORCEMENT TABS/REINFORCEMENT

LATERAL PRESSURE LATERAL PRESSURE

EXTENSOMETER/DISPLACEMENT TRANSDUCER

THICKNESS
~10MM

TRANS. MEASUREMENT BASE OF THE TRANS.

GRIPPING AREA EXTENSOMETER/DISPLACEMENT TRANSDUCER GRIPPING AREA
ZONE ZONE
>60mm >B >200mm >B >60mm

Fig. 4. Specimen geometry for direct tensile tests.

126 S. De Santis et al. / Composites Part B 127 (2017) 121e132

multiple cracking phase (Stage II), it can be increased grad- (vi) Number of cracks and saturation crack spacing.
ually up to a maximum of 0.5 mm/min.
(3) The specimen is gripped in the clamping wedges of the Results should be provided for all the specimen. Mean, standard
testing machine. The normal pressure applied in the gripping deviation and coefficient of variation of each parameter should also
area by the wedges is such to prevent the textile from slip- be calculated for the set of specimens. Peak values of specimens
ping out of the mortar matrix. The specimen should be that exhibited failure mode C should be considered unreliable and
reinforced in the gripping areas to ensure a homogeneous should be excluded from statistics.
stress distribution and to prevent excessive crushing damage
of the mortar. To this aim, composite (e.g., C/GFRP) wrapping 3. TRM-to-substrate shear bond behaviour
and/or composite or metallic (preferably aluminum or lead)
tabs can be used. Tabs should be soft enough to avoid local 3.1. Bond behaviour of TRM reinforcements
stress concentrations.
(4) The applied load is recorded by a load cell with suitable In a wide number of applications, the effectiveness of the
resolution and accuracy (to be certified). The stress is derived externally bonded reinforcement relies on the composite-to-
as the load divided by the cross section area of the dry textile. substrate bond strength. The only exceptions are represented by
This area is calculated as the design thickness of the textile the full wrapping (feasible on pillars or columns) and by the
multiplied by its width, this latter being the product of the strengthening layouts that include transversal strips to delay the
number of yarns/cords by their spacing. detachment (as it may be done in reinforced concrete beams
(5) Strains are either measured directly by an extensometer or strengthened in bending and in shear), or mechanical end anchors
calculated as the displacement recorded by transducers (e.g., to prevent it (as it may be done for the extrados strengthening of
potentiometers, LVDTs, etc.) divided by the gauge length of masonry arches). In these cases, the tensile rupture of the textile is
the device. The transducers must have suitable resolution expected to occurs.
and accuracy (to be certified). The gauge length of the in- Differently from Fibre Reinforced Polymers, in which the
strument should be large enough to include a significant detachment occurs by cohesive debonding within the substrate,
number of cracks, and a minimum of 200 mm is recom- TRMs may exhibit different failure modes, involving the textile-to-
mended. Measurements from devices with shorter gauge matrix or the mortar-to-substrate interface, or also the slippage of
length may be considered reliable only in the first (un- the textile out of the mortar matrix [11,12,29]. Numerous studies
cracked) Stage. Global displacement measures (from end have been devoted the TRM-to-substrate bond behaviour, with the
plate to end plate) can be used provided that slipping is aim of investigating the effect of the layout of the textile, the
prevented between tabs and clamping wedges, specimen coating/preimpregnation of the fibres, the mechanical character-
and tabs, and textile and matrix. istics of the mortar matrix, and the properties of the substrate
(6) The sampling frequency should be at least 5 Hz. (strength and stiffness, roughness, humidity) on bond strength and
(7) At least five samples should be tested. failure mode, develop design formulations for the estimate of the
bond strength and identify suitable testing methods.
C. Test report In the RRT, shear bond tests were carried out on masonry prisms
The test report shall include at least the following information: made out of five 55 mm thick clay bricks, having 14.8 N/mm2
compressive strength, 2.5 N/mm2 tensile strength and 5.76 kN/
(1) General information mm2 Young's modulus (supplied by SanMarco Terreal, Italy, under
(i) Origin and indication of the test specimens (constitu- the name SanMarco Rosso Vivo A6R55W). Depending on the ge-
ents, supplier). ometry of the TRM strip to be applied to the substrate and on
(ii) Specimen geometry (overall dimensions, number of laboratory facilities, some institutions decided to build the prisms
textile layers). with whole bricks, while others cut the bricks in smaller portions.
(iii) Curing and storage conditions (temperature in
C and The four bed joints were made of lime mortar with compressive
relative humidity in %). strength of about 5 N/mm2 (which differed from participant to
(iv) Age of the specimens at testing. participant). The bonded length was 260 mm for all tests and the
(v) Anchorage length and details of the gripping method. first bonded section was 30 mm far from the edge of the prism. The
(vi) Experimental laboratory (institution, location) and width of the bonded area was equal to that of the coupons used in
characteristics of the instrumentation: testing machine direct tensile tests. Also in this case, the laboratories chose the
(type and load capacity), load cell (accuracy and reso- setup (18/19 preferred a Single Lap setup, while one used a Double
lution), displacement transducers/extensometer (gauge Lap setup), the way to grip the end of the unbonded textile, and the
length, accuracy and resolution). Date of the test. way to measure the slip (relative displacement between substrate
(vii) Control parameter and displacement rate. and reinforcement at the loaded end of the bonded area). The
(2) Test results stress-slip response curve had to be provided, together with the
(i) Stress-strain response curve. Stresses must refer to the peak stress, the corresponding slip, and the failure mode (whose
cross section of dry textile. classification was agreed amongst participants, as in tensile tests).
(ii) Tensile stiffness in the three response Stages (EI, EII and
EIII). 3.1.1. Experimental setups
(iii) Stress and strain values of the transition points between Shear bond tests were carried out on specimens manufactured
Stages I and II and between Stages II and III (εI,sI and by applying a reinforcement strip to a substrate (brick, stone unit,
εII,sII, Fig. 2). concrete block, etc.), following either a single-lap scheme or a
(iv) Peak stress and strain (ft,εt). double-lap scheme. In the former, the reinforcement is bonded to
(v) Failure mode, according to the following classification: one side of the substrate, which is kept fixed (e.g., by a steel frame).
A: main crack near the gripping area; B: main crack in A portion of textile is left unbonded, clamped, and pulled by means
the middle of the specimen; C: textile slippage in the of either a testing machine or a movable actuator [19] (Fig. 6a). The
gripping area, without tensile rupture (Fig. 5). single-lap setup is easier to implement than the double-lap one,
 S. De Santis et al. / Composites Part B 127 (2017) 121e132 127

TENSILE FAILURE

A
TEXTILE MATRIX GRIPPING AREA

TENSILE FAILURE

B
TEXTILE MATRIX GRIPPING AREA

SLIPPAGE

C
TEXTILE MATRIX GRIPPING AREA

Fig. 5. Classification of failure modes in direct tensile tests.

and facilitates the identification of the load. Particular attention is that their mismatch tends to increase during test progression, and
however needed on the alignment of the specimen, to ensure that a cannot be neglected when failure does not occur simultaneously on
pure shear stress is applied to the reinforcement and prevent the the two sides [30]. A rotating cylinder can be used to minimize the
result from being affected by parasitic normal stresses on the difference of the load applied to the two sides of the specimen [12].
substrate-to-matrix interface [5]. The inherent eccentricity be- Otherwise, resistive strain gauges can be glued to the unbonded
tween the tensile load applied to the textile and the restraining textile to determine the actual load rate that is applied to each side
force exerted by the substrate may lead to a mixed failure mode [11]. The correct alignment of the specimen in the setup is neces-
(mode I, by normal stress, and mode II, by shear stress), but it af- sary to ensure the homogeneous load distribution among the fibre
fects the bond strength only for short bonded lengths [4]. bundles of the unbonded textile. Due to the stochastic nature of the
In the double-lap scheme, a U-shaped textile strip is bonded to local bond properties, an increase of the number of the bonded
the two sides of a substrate block/prism, which is kept fixed. A areas (from the one of single-lap setup to the four ones of the
saddle or a cylinder is used to pull the textile [12] (Fig. 6b). As an double-lap double-prism one) is associated to a slight decrease of
alternative, two substrate blocks/prisms can be joined by means of the measured bond strength and to a relatively lower scatter of
two textile strips (one per side) and moved apart by means of steel results [30].
plates connected by bars clamped in the wedges of the testing As in direct tensile tests, shear bond tests are carried out under
machine [30] (Fig. 6c). With this setup, it is more difficult to mea- displacement control. In addition to the force, the relative
sure the slip at the loaded end of each bonded area. The symmetry displacement between reinforcement and substrate at the loaded
of the geometry of double-lap setups does not guarantee that the end of the bonded area (slip) is also recorded by means of
load distribution on the two sides is equal. Actually, it was found displacement transducers. In principle, the slip could also be

Fig. 6. Experimental setups for composite-to-substrate shear bond tests: single-lap (a), double-lap single-prism (b), and double-lap double-prism (c).
128 S. De Santis et al. / Composites Part B 127 (2017) 121e132

calculated by integrating the strain measured along the bonded generally give rise to a stress-slip curve with a nearly flat branch,
area (e.g., by resistive strain gauges). Apart from the higher cost followed by a brittle detachment (Fig. 8a).
associated to the use of strain gauges and the effort needed for their TRM systems with relatively weak textile-to-matrix bond, such
application (which has to be carried out on the textile, before its as those with carbon [16], PBO [4,18], or steel ropes [19] very often
installation on the substrate), this method may underestimate the exhibit slippage failure. In this case, the stress-slip response shows
slip. Strain gauges provide very local measurements (thus possibly a post-peak soft load decrease due to the progressive loss of friction
failing in capturing strain concentrations) and/or may partially of the textile sliding within the mortar (Fig. 8b). Finally, TRM sys-
detach from the textile [19]. tems comprising sparse meshes of relatively weak textiles (e.g.,
basalt [15] or glass [17]) may fail by textile rupture. Proper
3.1.2. Failure modes dimensioning of the reinforcement should preclude this type of
Depending on the properties of the substrate, the shear strength failure. The stress-slip response exhibits either instantaneous load
of the matrix, the tensile strength of the textile, and the textile-to- drops at the textile rupture of the bundles, if it occurs out of the
matrix bond/interlocking, different failure modes may occur, such bonded area, or a sudden reduction and a further slight load
as: (A) debonding with cohesive failure in the substrate (Fig. 7a), (B) decrease if a telescopic failure occurs within the matrix, then fol-
debonding at the reinforcement-to-substrate interface (Fig. 7b) or lowed by textile pull-out (Fig. 8c).
(C) at the textile-to-matrix interface (Fig. 7c), (D) sliding of the
textile within the mortar matrix (within the reinforcement thick- 3.2. Shear bond test method
ness) (Fig. 7d), and, finally, (E) tensile rupture of the textile (Fig. 7e).
Cohesive debonding within the substrate is the same failure mode Recommendations for shear bond tests on TRM strengthening
that typically takes place with FRPs. It may occur when a strong systems are proposed herein. Shear bond tests provide the stress-
matrix is bonded to a relatively weak substrate. The use of a stiff slip response of a TRM composite bonded to a substrate block/
matrix may also cause local stress concentrations and promote this prism, subjected to a shear load at the reinforcement-to-substrate
type of detachment that, once activated, tends to propagate sud- interface applied in the quasi-static regime, and the correspond-
denly to the whole bonded area [19]. Detachment at the matrix-to- ing failure mode, which are necessary for the design of an exter-
substrate interface may take place on relatively smooth surfaces. It nally bonded TRM reinforcement. As for direct tensile tests, bond
has been observed on clay brick substrates [5]. The unevenness of tests are meant to be carried out on a TRM system comprised of a
masonry substrates generally inhibits this failure type, but it may textile (with a given layout/grid spacing, surface mass density, and
still be caused by inaccurate installation (Fig. 7b). Both proper installed in a given number of plies) and a mortar matrix supplied
substrate preparation (removal of dust, wetting, improvement of by the same producer.
surface roughness, consolidation, if allowed when dealing with A. Specimen geometry and manufacturing
cultural heritage) and adequate curing conditions (sufficient hu-
midity) should therefore be ensured [11]. Detachment at the (1) Shear bond tests should be carried out on the standard
textile-to-matrix interface may occur with dense fabrics, inhibiting substrates (concrete, brick masonry, tuff masonry) that are
a proper protrusion of the matrix through the voids [5]. Meshes proposed by the supplier as perspective field applications of
with sufficient grid spacing and matrices with small grain size the TRM system.
range should be used to avoid, or delay, this failure, which was (2) The TRM strengthening system is installed with the direction
however observed in combination with the tensile rupture of the of the fibres parallel to the longitudinal specimen axis and
cords in SRG composites (Fig. 7c). Failure modes A, B, and C symmetric to the specimen width (Fig. 9).

Fig. 7. Failure modes in shear bond tests: debonding with cohesive failure in the substrate (a), debonding at the reinforcement-to-substrate interface (b) or at the textile-to-matrix
interface (c), sliding of the textile within the reinforcement thickness (d), and tensile rupture of the textile (e).
 S. De Santis et al. / Composites Part B 127 (2017) 121e132 129

(a) (b) (c)

Stress in the textile

Stress in the textile

Stress in the textile

Failure mode: A, B, C Failure mode: D Failure mode: E

Slip Slip Slip

Fig. 8. Stress-slip response curve associated to different failure modes in shear bond tests.

(3) The instructions provided by the supplier in the product particular attention should be paid in the regularity of the
technical data sheet must be followed. In particular, if spe- reinforcement as well as in the planarity and parallel
cific surface preparation (roughness improvement by bush arrangement of fibre bundles/cords. To this aim, a slight pre-
hammer, consolidation, application of primer to promote tensioning (10 N/mm2 or less) may be applied to the textile
adhesion, etc.) are required, this has to be followed in the during mortar curing. The reinforcement shall be measured
installation and stated in the test report. in at least three different sections with instruments having
(4) The specimen is manufactured by applying the TRM accuracy within 1% of the mortar thickness. Measurements
strengthening system to one (in case of Single Lap testing shall be reported and should not differ from the nominal
setup) or two (in case of Double Lap testing setup) sides of ones by more than ±10% for the width and ±5% for the
the substrate. The portion of the substrate where the system thickness.
is applied is named bonded area. (6) The length of the bonded area (L) should be at least 250 mm.
(5) The thickness of the strengthening system applied to the This latter is not intended as an effective transfer length.
substrate should be the same as those used for direct tensile Longer bonded areas could lead to higher peak loads if failure
tests. In order to reduce the effects of possible nonhomoge- occurs by textile slippage within the mortar, due to the
neous stress distribution widthwise, it is recommended that friction contribution activated at the textile-to-matrix
the width of the bonded area (B) is the same of the coupons interface.
tested under tensile loading. During manufacturing, (7) In order to reduce edge effects, the distance between the
lateral edge of the bonded area and the side of the substrate
should be at least 20 mm, and that between the first bonded
section at the loaded end (i.e., on the end where load is
BONDED
applied) should be at least 30 mm.
WIDTH (B)
(8) Curing should last at least 28 days. In the curing phase,
THICKNESS >20mm 40-100mm >20mm adequate relative humidity conditions should be ensured to
~10mm prevent the development of differential shrinkage, which
would cause cracking and/or premature detachment. Before
testing, the specimen should be stored in standard laboratory
conditions (20e25
C and 50e60% R.H.) for at least 7 days.

B. Push-pull Single Lap and Double Lap testing setups

BONDED AREA

BONDED LENGTH
SUBSTRATE
SUBSTRATE

L>260mm

(1) Either a Single Lap (SL) or a Double Lap (DL) setup can be
BONDED used (Fig. 10).
AREA (2) In the case of SL, the TRM system is bonded to one side of the
substrate and a portion of textile is left unbonded on one side
of the bonded area (Fig. 10a). The specimen is placed in a
steel frame (made out of plates to form an angle of 90
),
which is stiff enough to avoid rotations and distortions. The
LOADED
frame is fixed to one set of wedges of the testing machine on
>30mm

END
one side, and kept fixed, while the unbonded textile is
clamped in the wedges on the other side, and pulled. The
UNBONDED TEXTILE

UNBONDED TEXTILE

loaded end of the composite strip can be glued between two

plates (aluminum, steel, plumb, or GFRP) in order to guar-
antee a homogeneous distribution of the stresses and avoid
slippage phenomena at the clamps. A careful positioning of
the specimen is necessary to ensure the alignment of the
upper plate of the supporting steel device and the textile
strip which is pulled. By doing so, a pure shear load is applied
to the reinforcement preventing the result from being
affected by parasitic normal stresses on the substrate-to-
Fig. 9. Specimen geometry for bond tests (detail of the bonded area).
130 S. De Santis et al. / Composites Part B 127 (2017) 121e132

STEEL BAR
WELDED TO
STEEL PLATE
(a) (b) (c)

SUBSTRATE

SUBSTRATE
SUBSTRATE
BONDED BONDED BONDED BONDED BONDED
AREA AREA AREA AREA AREA

REACTING UNBONDED UNBONDED

STEEL TEXTILE TEXTILE
UNBONDED
FRAME
TEXTILE
REACTING
UNBONDED UNBONDED

SUBSTRATE
STEEL
TEXTILE TEXTILE BONDED
GRIPPING FRAME BONDED
AREA AREA
AREA

CYLINDRICAL
JOINT

STEEL BAR
WELDED TO
STEEL PLATE

Fig. 10. Schemes of experimental setups for bond tests: single lap (a), double-lap single-prism (b) and double-lap double-prism (c).

matrix interface. Nevertheless, in order to reduce possible (4) Both in the SL and in the DL setups, the unbonded textile can
normal stresses at the reinforcement-to-substrate interface, be impregnated (e.g., with epoxy resin) to improve the dis-
which may be caused by slight misalignments, the length of tribution of the load width wise.
the unbonded textile should be at least 400 mm, including
the gripping area. C. Test execution
(3) In the case of DL, a continuous U-shaped textile strip is
bonded to the two sides of a substrate block/prism (Fig. 10b). (1) Specimen dimensions are carefully checked before testing.
The substrate is placed in a steel frame (composed of two Visible pre-damage (e.g., cracking, local detachment, de-
transversal beams (or a beam and a plate) connected by two flections) has to be documented.
or four threaded bars, which is fixed to the testing machine (2) Tests are carried out with a tensile test machine, under
by means of a steel bar gripped in the clamping wedges. The displacement control. The control parameter should be the
textile is wrapped around a cylinder, whose diameter is equal displacement measured by the integrated LVDT or by another
to the distance between the two sides of the specimen to transducer applied to the specimen to record the slip. The
ensure that a pure shear stress is applied to the displacement is increased monotonically up to failure at a
reinforcement-to-substrate interface. To this aim, the align- rate not higher than 0.2 mm/min. Note that this rate is lower
ment between the specimen and the cylinder should also be than that recommended for tensile tests, since smaller
ensured and spherical joints could be used to this purpose. displacement are generally recorded in bond tests.
The cylinder is connected to the wedges of the testing ma- (3) The applied load is recorded by a load cell with suitable
chine and pulled. The possibility of the cylinder to rotate resolution and accuracy (to be certified). The stress is derived
ensures that the load is equally distributed between the two as the load divided by the cross section area of the dry textile.
sides of the specimen. Otherwise, two Teflon sheets should In case of DL, the load applied to each side of the specimen
be used to minimize friction. The overall length of the can be derived as the resultant load divided by two only if a
unbonded textile should be at least 800 mm. A DL setup with cylinder allowed to rotate or two Teflon sheets are used. As
four bonded areas can also be used, that is made of two an alternative, a reliable method (e.g., by gluing resistive
substrate blocks/prisms, connected by two TRM strips, one strain gauges on the unbonded textile) should be used to
per side (Fig. 10c). The two substrate blocks are moved apart establish the percentage of total load that is actually applied
thus inducing a shear load at the four reinforcement-to- to each side of the specimen. In this case, the unbonded
substrate interfaces. Due to the higher complexity and to textile can be locally impregnated with epoxy resin to ensure
the difficulty in measuring the slip at the loaded end of each the proper adhesion of the strain gauge.
of the four bonded areas, the use of this setup is not rec- (4) The slip has to be determined. It is defined as the relative
ommended, unless it is made necessary by specific needs, displacement between reinforcement and substrate at the
related to the available facilities in the laboratory or to the loaded end of the bonded area. The slip can be either
geometry of the specimens (e.g., provided with curved sur- measured directly by means of displacement transducers
faces [31]). with suitable resolution and accuracy (to be certified) or
 S. De Santis et al. / Composites Part B 127 (2017) 121e132 131

A B C
SUBSTRATE SUBSTRATE SUBSTRATE

MATRIX TEXTILE MATRIX TEXTILE

MATRIX TEXTILE

D E1 E2
SUBSTRATE SUBSTRATE SUBSTRATE

MATRIX TEXTILE MATRIX TEXTILE MATRIX TEXTILE

Fig. 11. Classification of failure modes in shear bond tests.

derived after the test by post-processing test data. In this Results should be provided for each specimen. Mean, standard
case, all the displacement components contributing to the deviation and coefficient of variation of each parameter should also
overall displacement recorded by the transducer integrated be calculated for the set of specimens.
in the testing machine (e.g., elastic elongating of the
unbonded textile, deflections of the steel frame) must be 4. Conclusions
carefully accounted for.
(5) The sampling frequency should be at least 5 Hz. The outcomes of a wide Round Robin Test led to the develop-
(6) At least five samples should be tested. ment of recommendations for performing direct tensile tests and
shear bond tests on TRM composites, to derive the fundamental
D. Test report mechanical parameters for the qualification and design of exter-
The test report shall include at least the following information: nally bonded reinforcements.
Direct tensile tests provide the stress-strain response curve,
(1) General information which is characterized by three response stages (uncracked, crack
(i) Origin and indication of the test specimens (constitu- development, and cracked), the stress and strain values of the
ents and supplier of TRM system, type, geometry and transition points between one stage and the following one, and
mechanical properties of the substrate). those of the peak, the stiffness in the three stages, and the crack
(ii) Specimen geometry (type of setup SL/DL, dimensions of pattern. In the testing procedure, the specimen has to be gripped on
the bonded area and distance from the edges of the the mortar matrix with sufficient pressure to avoid slippage and
substrate, number of textile layers, length of the after suitable reinforcement to avoid damage in the matrix. The
unbonded textile). If DL setup with four bonded areas is gauge length of the device for displacement/strain measurement
used, the reasons that led to this choice have to be should be large enough to include a sufficient number of cracks to
specified. prevent the results from being affected by the crack distribution
(iii) Curing and storage conditions (temperature in
C and which may vary randomly from test to test.
relative humidity in %). TRM-to-substrate shear bond tests provide the stress-slip
(iv) Age at testing of the specimens. response and the corresponding failure mode. Correct alignment
(v) Details of the gripping method for the unbonded textile and setup implementation ensure that the peak load is recorded
(in case of SL). reliably. Detachment may occur by cohesive debonding within the
(vi) Experimental laboratory (institution, location) and substrate, by detachment at the textile-to-matrix or mortar-to-
characteristics of the instrumentation: testing machine substrate interface, by slippage of the textile within the matrix,
(type and load capacity), load cell (accuracy and reso- or, finally, by tensile rupture of the unbonded textile. In addition to
lution), displacement transducers/extensometer (gauge the TRM-to-substrate load transfer capacity, information for the
length, accuracy and resolution). Date of the test. optimization of the strengthening systems and for the design of the
(vii) Control parameter and displacement rate. reinforcement can be also determined from the outcomes of bond
(2) Test results tests.
(i) Stress-slip response curve (stresses being referred to the The results of both direct tensile tests and shear bond tests may
cross section of dry textile) be affected by the manufacturing and curing conditions. This
(ii) Peak stress and slip (fb,s(fb)). highlights the importance of a careful installation for ensuring the
(iii) Exploitation ratio of tensile strength, in %, defined as effectiveness of the strengthening work. The development of offi-
h ¼ fb/ft, ft being the tensile strength of the TRM system cial standards for mechanical testing, installation and design would
derived from direct tensile tests. improve the reliability of strengthening works with TRM compos-
(iv) Failure mode, according to the following classification: ites and the possibility of controlling their effectiveness, thus pro-
A: debonding with cohesive failure of the substrate; B: moting their safe use in the rehabilitation and upgrade of existing
debonding at the matrix-to-substrate; C: debonding at structures as well as in the safeguarding of the built heritage.
the textile-to-matrix interface; D: textile slippage within
the matrix, with or without cracking of the outer mortar Acknowledgements
layer; E1/E2: tensile rupture of the textile out of the
bonded area or within the matrix, respectively (Fig. 11). The authors gratefully acknowledge the financial support
132 S. De Santis et al. / Composites Part B 127 (2017) 121e132

provided by the research program ReLUIS, funded by the Italian Performance assessment of basalt FRCM for retrofit applications on masonry.
Compos Part B-Eng. http://dx.doi.org/10.1016/j.compositesb.2017.05.003.
Department of Civil Protection, Executive Project 2016. S. De Santis
[16] Carozzi FG, Bellini A, D'Antino T, de Felice G, Focacci F, Hojdys L, et al.
and G. de Felice acknowledge funding by the Italian Ministry for Experimental investigation of tensile and bond properties of Carbon-FRCM
Foreign Affairs within the research project “Composites with composites for strengthening masonry elements. Compos Part B-Eng. http://
inorganic matrix for sustainable strengthening of architectural dx.doi.org/10.1016/j.compositesb.2017.06.018.
[17] Leone M, Aiello MA, Balsamo A, Carozzi FG, Ceroni F, Corradi M, et al. Glass
heritage” (Year 2016, Grant N. PGR00234). fabric reinforced cementitious matrix: Tensile properties and bond perfor-
mance on masonry substrate. Compos Part B-Eng. http://dx.doi.org/10.1016/j.
References compositesb.2017.06.028.
[18] Caggegi C, Carozzi FG, De Santis S, Fabbrocino F, Focacci F, Hojdys L, et al.
Experimental analysis on tensile and bond properties of PBO and Aramid
[1] D'Ambrisi A, Feo L, Focacci F. Experimental and analytical investigation on fabric reinforced cementitious matrix for strengthening masonry structures.
bond between Carbon-FRCM materials and masonry. Compos Part B-Eng
Compos Part B-Eng. http://dx.doi.org/10.1016/j.compositesb.2017.05.048.
2013;46:15e20. http://dx.doi.org/10.1016/j.compositesb.2012.10.018. [19]  A, et al. Round
De Santis S, Ceroni F, de Felice G, Fagone M, Ghiassi B, Kwiecien
[2] Carozzi FG, Milani G, Poggi C. Mechanical properties and numerical modeling Robin test on tensile and bond behaviour of steel reinforced grout systems.
of Fabric Reinforced Cementitious Matrix (FRCM) systems for strengthening of Compos Part B-Eng. http://dx.doi.org/10.1016/j.compositesb.2017.03.052.
masonry structures. Compos Struct 2014;107:711e25. http://dx.doi.org/ [20] De Santis S, de Felice G. Tensile behaviour of mortar-based composites for
10.1016/j.compstruct.2013.08.026.
externally bonded reinforcement systems. Compos Part B-Eng 2015;68:
[3] Balsamo A, Di Ludovico M, Prota A, Manfredi G. Masonry walls strengthened 401e13. http://dx.doi.org/10.1016/j.compositesb.2014.09.011.
with innovative composites. Am Concr Inst, ACI Spec Publ 2011;2(275):
[21] Arboleda D, Carozzi F, Nanni A, Poggi C. Testing procedures for the uniaxial
769e86. http://dx.doi.org/10.14359/51682454. tensile characterization of fabric-reinforced cementitious matrix composites.
[4] D'Antino T, Sneed LH, Carloni C, Pellegrino C. Effect of the inherent eccen- J Compos Constr 2015:04015063. http://dx.doi.org/10.1061/(ASCE)CC.1943-
tricity in single-lap direct-shear tests of PBO FRCM-concrete joints. Compos
5614.0000626.
Struct 2016;142:117e29. http://dx.doi.org/10.1016/j.compstruct.2016.01.076. [22] Ascione L, de Felice G, De Santis S. A qualification method for externally
[5] De Santis S, de Felice G. Steel reinforced grout systems for the strengthening
bonded Fibre Reinforced Cementitious Matrix (FRCM) strengthening systems.
of masonry structures. Compos Struct 2015;134:533e48. http://dx.doi.org/ Compos Part B-Eng 2015;78:497e506. http://dx.doi.org/10.1016/
10.1016/j.compstruct.2015.08.094. j.compositesb.2015.03.079.
[6] Brameshuber W. Textile reinforced concrete. State-of-the-art report of RILEM [23] RILEM Technical Committee 232-TDT (chair: wolfgang Brameshuber).
Technical Committee 201-TRC. Report 36. Bagneux: RILEM; 2006. Recommendation of RILEM TC 232-TDT: test methods and design of textile
[7] Papanicolaou CG, Trinantafillou TC, Papathanasiou M, Karlos K. Textile rein-
reinforced concrete. Mater Struct 2016;49(12):4923e7. http://dx.doi.org/
forced mortar (TRM) versus FRP as strengthening material of URM walls: out- 10.1617/s11527-016-0839-z.
of-plane cyclic loading. Mater Struct 2008;41:143e57. http://dx.doi.org/ [24] Orlowsky J, Raupach M. Durability model for AR-glass fibres in textile rein-
10.1617/s11527-007-9226-0. forced concrete. Mater Struct 2008;41(7):1225e33. http://dx.doi.org/10.1617/
[8] Valluzzi MR, Modena C, de Felice G. Current practice and open issues in s11527-010-9586-8.
strengthening historical buildings with composites. Mater Struct 2014;47(12):
[25] Hegger J, Will N, Bruckermann O, Voss S. Load-bearing behavior and simu-
1971e85. http://dx.doi.org/10.1617/s11527-014-0359-7. lation of textile reinforced concrete. Mater Struct 2006;39(8):765e76. http://
[9] De Santis S, Casadei P, De Canio G, de Felice G, Malena M, Mongelli M, et al.
dx.doi.org/10.1617/s11527-005-9039-y.
Seismic performance of masonry walls retrofitted with steel reinforced grout. [26] Arboleda D, Yuan S, Giancaspro J, Nanni A. Comparison of strain measurement
Earthq Eng Struct Dyn 2016;45(2):229e51. http://dx.doi.org/10.1002/ techniques for the characterization of brittle, cementitious matrix composites.
eqe.2625. Research and applications in structural engineering, Mechanics and Compu-
[10] De Canio G, de Felice G, De Santis S, Giocoli A, Mongelli M, Paolacci F, et al. tation. Cape Town, South Africa. 2013.
Passive 3D motion optical data in shaking table tests of a SRG-reinforced
[27] Napoli A, de Felice G, De Santis S, Realfonzo R. Bond behaviour of steel rein-
masonry wall. Earthq Struct 2016;10(1):53e71. http://dx.doi.org/10.12989/ forced polymer strengthening systems. Compos Struct 2016;152:499e515.
eas.2016.10.1.053. http://dx.doi.org/10.1016/j.compstruct.2016.05.052.
[11] de Felice G, De Santis S, Garmendia L, Ghiassi B, Larrinaga P, Lourenço PB, et al. [28]  A, Roscini F. Application of digital
Tekieli M, De Santis S, de Felice G, Kwiecien
Mortar-based systems for externally bonded strengthening of masonry. Mater image correlation to composite reinforcements testing. Compos Struct
Struct 2014;47(12):2021e37. http://dx.doi.org/10.1617/s11527-014-0360-1.
2017;160:670e88. http://dx.doi.org/10.1016/j.compstruct.2016.10.096.
[12] Carozzi FG, Poggi C. Mechanical properties and debonding strength of Fabric [29] De Santis S, de Felice G. Bond behaviour of Steel Reinforced Grout strength-
Reinforced Cementitious Matrix (FRCM) systems for masonry strengthening.
ening systems applied to the extrados of masonry vaults [Leuven, Belgium].
Compos Part B-Eng 2015;70:215e30. http://dx.doi.org/10.1016/ In: Proceedings of the 10th international conference on structural analysis of
j.compositesb.2014.10.056. historical constructions SAHC; 2016.
[13] ICC-ES. AC434. Acceptance criteria for masonry and concrete strengthening [30] Sneed LH, D'Antino T, Carloni C, Pellegrino C. A comparison of the bond
using fiber-reinforced cementitious matrix (FRCM) composite systems. behavior of PBO-FRCM composites determined by double-lap and single-lap
Whittier, CA, USA: ICC-Evaluation Service; 2013.
shear tests. Cem Concr Compos 2015;64:37e48. http://dx.doi.org/10.1016/
[14] American Concrete Institute. ACI 549.4R-13 Guide to design and construction j.cemconcomp.2015.07.007.
of externally bonded Fabric-Reinforced Cementitious Matrix (FRCM) systems [31] De Santis S. Bond behaviour of Steel Reinforced Grout for the extrados
for repair and strengthening concrete and masonry structures. Farmington strengthening of masonry vaults. Constr Build Mater 2017;150:367e82.
Hills, MI, USA. 2013.
[15] Lignola GP, Caggegi C, Ceroni F, De Santis S, Krajewski P, Lourenço PB, et al.