Construction and Building Materials 326 (2022) 126640

Contents lists available at ScienceDirect

Construction and Building Materials

journal homepage: www.elsevier.com/locate/conbuildmat

Experimental characterization of the textile-to-mortar bond through

distributed optical sensors
Elisa Bertolesi a, *, Mario Fagone b, Tommaso Rotunno c, Ernesto Grande d, Gabriele Milani e
a
Department of Civil and Environmental Engineering, Brunel University London, Uxbridge UB8 3PH, United Kingdom
b
Department of Civil and Environmental Engineering, University of Florence, via di S. Marta 3, Florence 50139, Italy
c
Department of Architecture, University of Florence, Piazza Brunelleschi 6, Florence 50121, Italy
d
Department of Engineering Science, University Guglielmo Marconi, Via Plinio 44, Rome 00193, Italy
e
Department of Architecture, Built Environment and Construction Engineering (ABC), Politecnico di Milano, Piazza Leonardo da Vinci 32, Milan 20133, Italy

A R T I C L E I N F O A B S T R A C T

Keywords: TRM (Textile Reinforced Mortar) and FRCM (Fiber Reinforced Cementitious Mortar) strengthening materials are
Fiber optic sensors highly heterogeneous composites involving domains that have dramatically different mechanical properties (i.e.,
Debonding mechanisms inorganic matrix and fabric textile). Maximizing their exploitation ratio involves achieving a better bond be­
TRM & FRCM composite materials
tween textile and inorganic matrix which shifts the failure mode from textile sliding to textile rupture or cohesive
Bond-slip model
failure modes. So far, the local bond behaviour between fiber bundles and inorganic matrix has been analysed
Strengthening
indirectly through the evaluation of the global performance of TRM/FRCM materials subjected to tensile tests or
single lap shear tests. In this article, the authors adopted distributed fiber optic sensors directly installed to textile
bundles to track the strain evolutions of PBO-TRM strengthening materials. The strain evolutions were used, for
the first time, (i) to understand the behaviour of TRM coupons subjected to tensile tests and (ii) to experimentally
calibrate the interfacial tangential bond-slip law in flat masonry pillars strengthened with TRMs. The results
allowed a better insight of the response of TRM materials not possible with traditional sensors and an accurate
characterization of their bond performance. The interfacial tangential stress-slip law was then adopted in
analytical models to predict the global performance of TRM materials providing satisfactory results compared
with experimental outcomes.

1. Introduction interfacial layer, whereas composites and supports generally behave as

elastic materials. When it comes to composite strengthening materials
Cohesive material laws describe the complex detachment mecha­ involving constituent materials that have dramatically different me­
nism of strengthening materials from brittle supports with simplified chanical and physical properties such as TRMs (Textile Reinforced
mathematical relationships [1–4]. From an operational point of view, Mortar) and FRCMs (Fiber Reinforced Cementitious Mortar), the
this simplification represents a convenient solution to accurately study assessment of their bond quality became an intricated problem [9–15].
detachment phenomena that otherwise would involve highly nonlinear Giving the high heterogeneity of their constituent materials (i.e., inor­
computational approaches and damage material models [5–8]. The ganic matrix and fiber textiles) in terms of elastic and inelastic proper­
reliability of the proposed bond-slip laws, which evolved over the years ties, the observed failure modes comprise a wide range of possibilities
from piecewise linear to linear and exponentially decreasing softening [16], including: (i) cohesive debonding in the masonry; (ii) debonding at
laws, is usually assessed through experimental investigations. The the matrix-to-substrate; (iii) debonding at the textile-to-matrix interface;
technical literature is abundant in terms of experimental research works (iv) sliding of the textile within the matrix; and (v) tensile rupture of the
confirming different mathematical descriptions of the bond performance textile. According to the technical literature, failure modes involving
of Fiber Reinforced Polymer (FRP) composites applied on brittle sup­ sliding of the textile and matrix-to-textile detachment are associated
ports [1–4]. Indeed, detachment of FRPs from concrete or masonry with a poor bond and thus to a lower exploitation ratio of the
supports involves, depending on the support quality, cohesive failures of strengthening materials itself. Improving the bond between the con­
the support itself. Thus, most of the observed damages is localized in an stituent materials, especially textile and mortar matrix, ensures higher

* Corresponding author.

https://doi.org/10.1016/j.conbuildmat.2022.126640
Received 5 November 2021; Received in revised form 14 January 2022; Accepted 26 January 2022
Available online 17 February 2022
0950-0618/© 2022 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
E. Bertolesi et al. Construction and Building Materials 326 (2022) 126640

Table 1 on one length of optical fiber and be embedded into different materials
Bricks mechanical properties; C.V. = coefficient of variation. (i.e., asphalt, concrete, etc). The present paper is intended to provide, for
n. specimens Mean C.V. the first time, the local interfacial tangential-bond slip law obtained
experimentally through Fibre Optic (FO) sensors embedded into TRM/
[MPa] [%]
FRCM materials. The article is organized into five sections, including a
Compressive strength 18 20.10 10.79 comprehensive description of the experimental investigation comprising
Young modulus 6 8712 6.92
Direct tensile strength 6 2.49 16.90
tensile and single laps shear tests of PBO (polyparaphenylene benzobi­
Bending tensile strength 6 3.36 33.77 soxazole) FRCM strengthening materials applied to flat masonry pillars
(Section 2) and the adopted Fiber Optic technology. Sections 3 discusses
the calibration procedure used to obtain the interfacial bond-slip law fed
with the strain evolutions tracked using fiber optic sensors. Finally,
Table 2
Three point bending tensile strength and compressive strength of mortar. Section 4 critically discusses the outcomes obtained using an analytical
model developed by the authors and herein coupled with the interfacial
n. specimens Mean C.V.
bond law obtained from the present experimental investigation. Final
[MPa] [%] remarks and future works are highlighted in Section 5.
(Bending) tensile strength 6 1.85 9.42
Compressive strength 12 5.18 8.212 2. Experimental investigation

The experimental investigation concerned a FRCM coupon subjected

exploitation ratios of the strengthening intervention. This is usually
to tensile test according to [38] and a masonry pillar strengthened with
associated with failure modes (i) and (v). Results from [17] clearly
FRCM materials subjected to a single lap shear test according to [16].
indicate that the bond capacity is not directly related to the textile
The masonry pillar was manufactured using bricks and mortar having
tensile strength or stiffness but depends on the interaction between
the same characteristics of the ones utilized by the authors in previous
textile and matrix as well as to that between composite system and
experimental campaigns. Therefore, only the main mechanical proper­
substrate. Thus, understanding the bond quality between textile and
ties of such materials are reported in the following since the mechanical
mortar matrix is of paramount importance to: (i) improve exploitation
characterization of mortar and bricks has already been fully described in
ratios of TRM and FRCM strengthening materials, (ii) calibrate mathe­
[39–40]. For further details, the reader can refer to the cited papers. The
matical interfacial bond-slip laws [18–24] and (iii) provide reliable
FRCM strengthening system considered in this paper is composed of an
guidelines for their design [16,25–27]. From an experimental point of
inorganic matrix and a bi-directional PBO fiber mesh (70 g/m2 in warp
view, the study of the local bond between textiles and inorganic matrices
and 18 g/m2 in weft). The main mechanical characteristics of the ma­
in TRM or FRCM materials was performed conducting tensile tests on
terials composing the FRCM system, declared by the supplier in the
coupons [28–29] and single/double lap shear tests [30–31]. Unfortu­
technical data sheet, are reported in the following.
nately, the corpus of the experimental investigations done so far, failed
in providing accurate data on the matrix-fibre interaction, thus sug­
2.1. Materials properties
gesting further investigation is still needed. The available experimental
data refer to TRM/FRCM macroscopic behaviour (i.e., failure modes,
2.1.1. Bricks and mortar
force-slip curves, maximum elongation and peak loads) leaving several
Solid pressed bricks, also called soft mud firebricks, and ready mixed
open issues still unresolved. Tensile tests are used to provide the
mortar made with lime and cement as binder were used in
constitutive relationship of TRM/FRCM composites, which is generally
manufacturing the masonry pillar. Those materials were considered
characterized by three response stages, whereas single/double lap shear
because their characteristics are similar to those of the materials
tests are preferred to estimate the bond capacity transferred from the
commonly present in most of the existing masonry constructions in Italy.
structural element to the externally bonded strengthening. In [17], de
The mechanical characterization of the bricks was carried out by sub­
Felice et al. clearly identify that the weakest element controlling the
jecting 18 cubic specimens (50×50×50 mm3) and six prismatic speci­
bond capacity change from system to system, depending on multiple
mens (50×50×150 mm3) to compression tests, 6 prismatic specimens
aspects, which cannot be directly inferred from tensile tests. Several
(40×40×200 mm3) to three-point bending tests and, finally, 6 prismatic
attempts were done to investigate the bond quality between textiles and
specimens (40×40×90 mm3) to direct tensile tests. The test results are
inorganic matrices through pull-out tests of single bundles from matrix
summarized in Table 1.
cubic specimens [32–33]. Although those investigations provided
For the characterization of the mechanical parameters of the mortar,
valuable insights of the local bond, the scale change between single
six 40×40×160 mm3 specimens were subjected to three-point bending
bundle bond and textile bond add a further level of complexity to the
tests. Then, compression tests were performed on the twelve halves of
problem and the actual assessment of what happens locally during
the prisms. The tests were performed according to [41]. The average
detachment phenomena is still one of the most important and, so far
strength values are shown in Table 2 as well as the coefficient of
unresolved, issues for TRM/FRCM materials. This is mostly due to the
variation.
fact that traditional sensors are only able to provide global measure­
ments which are suitable for FRPs reinforced structures but not for TRM/
2.1.2. Fiber reinforced cementitious matrix (FRCM)
FRCM strengthening materials. Recently, many civil engineering fields
Bi-directional PBO mesh and stabilized inorganic matrix (a cement-
are seeing the introduction of new sensing technologies such as wave­
based mortar with short fibers) are the components of the strength­
length- or frequency-based fiber optic sensors [34–37]. Fiber optic
ening system used in the present experimental investigation. The me­
sensors are based on interrogation of strain-induced wavelength shifts.
chanical properties declared by the supplier in the technical data sheet
Introduction of Bragg gratings in optical fibers usually is performed
are summarized in Table 3.
exposing optical fiber core to intense UV laser through a mask to achieve
specific wavelength reflectivity. Conversely to traditional sensors (i.e.,
2.2. Distributed fiber optic sensors
LVDTs DIC and strain gauges), such sensors provide several advantages,
namely: (i) high sensitivity, (ii) ease of installation, (iii) capability of
This study adopted distributed fiber optic (FO) strain sensors to
static and dynamic measurements. In addition, they can be serially
measure the local interfacial bond law in FRCM/TRM strengthening
multiplexed by writing several Bragg gratings of different wavelengths
materials. The FO technology employed in the present study is based on

2
E. Bertolesi et al. Construction and Building Materials 326 (2022) 126640

Table 3
Mechanical properties (declared by the producer) of the reinforcing system components.
Equivalent thickness of Tensile Young’s Ultimate Toughness Compressive Compressive Young’s Bending
the mesh (warp modulus (Ef) of the elongation of the strength at 28 days modulus at 28 days strength at 28
direction) fibre/mesh fibre. days

[mm] [GPa] [%] [GPa] [MPa] [MPa] [MPa]

PBO fibre – 270 2.5 5.80 – – –

Bi-directional 0.045 241 – – – – –
PBO mesh
Inorganic – – – – ≥ 20 ≥ 7500 ≥ 3.5
matrix

coefficient (k). Actual (λ) and reference (λ0 ) wavelengths are automat­
ically calculated by the FO dynamic interrogator. The distributed FO
sensor comprises 6 measuring points with 8 mm grating length and 26
mm geometrical spacing reaching a total sensing length equal to 130
mm (spacing between first and last mid-section of the FOs). Two
different acquisition frequencies were used, namely 100 and 50 Hz for
the tensile (the FO sensor is schematically depicted in Fig. 3a) and the
single lap shear (the FO sensor is schematically depicted in Fig. 7a) tests,
respectively. No temperature compensation was considered in the pre­
sent study. Fig. 1

2.3. Tensile test

Fig. 1. Distributed fiber optic sensor adopted in the present investigation.
Tensile test on the coupon specimen was carried out according to
[38]. The specimen, composed of two layers of mortar and a single layer
classical Rayleigh scattering. The irregularities allowing the elastic
of PBO fiber mesh, had total thickness of 8 mm and length 500 mm; a
scattering of the transmitted light inside the glass core, are generated
textile fabric portion composed of seven yarns in the warp direction was
during the FO manufacturing, and they are directly linked to the
considered, so that the coupon had width equal to 63 mm (see Fig. 2a).
wavelengths of the light wave.
Before manufacturing the specimen, a FO sensor was installed to the
The FO sensor comprises a coated fibre having 125 μm cladding
central part of the fabric in the warp direction (see Fig. 2b). According to
diameter, a GFRP outer diameter of 0.5 mm and an additional High-
the indications provided by the supplier, the specimen was manufac­
Density Polyethylene (HDPE) outer jacket (diameter 0,2 mm). The
tured through a wet-lay-up procedure (see Fig. 2c): a first layer of matrix
coating layer was designed to protect the inner fiber during the instal­
was laid in a wood framework having thickness of 4 mm; then, a single
lation and testing operations. This study used a dynamic data acquisition
layer of PBO mesh was positioned on it taking care that this adheres to
system to perform the measurement. The frequency shift is associated
the matrix by exerting a slight pressure; a second wood framework (4
with strain changes, as described in Eq. (1):
mm thick) was positioned on the first one so that the second layer of
1 λ matrix could be applied. The specimen was cured at room conditions for
ε = ln (1)
k λ0 at least 28 days before testing.
According to [38] the upper and lower parts of the specimen, having
At a constant temperature, the spectral shift can be converted into
a length of 125 mm (Fig. 2a), were clamped using stiff steel plates with a
strain along the FO distributed sensor with a calibrated strain sensitivity
rubber sheet in between (Fig. 3). Such plates were tightened using a

Fig. 2. Coupon specimen geometry (-a), FO sensor glued to the PBO mesh (-b) and coupon manufacturing (-c).

3
E. Bertolesi et al. Construction and Building Materials 326 (2022) 126640

Fig. 3. Tensile test setup and instrumentation: schematization (-a) and positioning of the specimen in the loading machine (-b).

Fig. 4. Tensile stress – strain diagram obtained with the LVDTs and the fiber
optic sensors.

torque wrench, so that a compression corresponding to 65% of the

minimum strength of the mortar declared by the supplier (M20) was
imposed to the coupon at the load introduction areas. Each of the
clamping systems was connected to the test machine using a double joint
as to make spherical hinges. The test was carried out in deformation-
controlled mode, imposing a strain rate of 2 × 10− 3 /min. During the
test, two Linear Variable Displacement Transducers (LVDTs) measured
the elongation of the central portion of the specimen (base length 200
mm) and the FO sensor measured the axial deformation at six points of
the central yarn of the PBO mesh (“S1′′ -”S6′′ in Fig. 3a).
The stress-strain diagram obtained from the test is reported in Fig. 4;
the strain values reported in abscissa refer to the average of the LVDTs Fig. 5. Geometry of reinforced masonry pillars – SLS test. Dimensions in mm.
and fiber optic (FO) sensors measurements, while the stress values in
ordinate refer to the nominal cross section of the textile mesh. The stress-
following, where the corresponding theoretical model is presented and
strain diagrams show a typical trilinear shape, as expected.
validated). Indeed, while the diagram of LVDT globally accounts for the
Although the trend of the stress-strain diagrams of the LVDT and FO
formation of cracks (the corresponding strains were deduced dividing
sensors is similar since it underlines the three common phases charac­
the displacements of LVDT by their length, 200 mm, which covers a
terizing the tensile behavior of FRCM: pre-crack, cracking developing
portion of the specimen where the majority of cracks occurred), the
and post-crack, the diagrams of FO sensors are neither coincident among
measurements provided by FO sensors are instead locally influenced by
them nor with the diagram of LVDT. This feature is strictly related to the
the crack formation. In particular, since the formation of each crack
behavior of the specimen observed during the test which was charac­
leads to a local redistribution of stresses between matrix and grid, the
terized by a progressive formation in the matrix of six main cracks
measurements provided by the FO sensors (particularly the ones close to
located at different sections (a more detail explanation is reported in the

4
E. Bertolesi et al. Construction and Building Materials 326 (2022) 126640

Fig. 6. Strengthened masonry pillar manufacturing; (-a) preparation of the surface to be reinforced and formwork positioning; (-b) application of the PBO mesh
equipped with the FO sensor to the first layer of mortar; (-c) application of the second layer of mortar.

Fig. 7. SLS test – setup and instrumentation: schematization (-a) and positioning of the specimen in the loading machine (-b).

the cracks) are able to capture this phenomenon. following a wet-lay-up procedure, analogous to the one previously
described for the coupon manufacturing (see Fig. 6). Also, in this case,
an FO sensor was installed to the central yarn of the PBO mesh before
2.4. Single lap shear test manufacturing the strengthening.
The test was carried out according to [16]. The specimen was con­
The geometry of the strengthened masonry specimen subjected to a strained between two steel plates, positioned at the top and bottom,
single lap shear test is schematized in Fig. 5. The masonry specimen was connected by four threaded bars; a slight precompression was applied to
manufactured with five bricks (250 × 120 × 65mm3 ) and four mortar the specimen in order to avoid its rotation during the test (Fig. 7). The
joints having thickness of 10 mm. One face of the so obtained masonry load was applied clamping the free edge of the PBO mesh, properly
pillar was strengthened with a FRCM strengthening having dimensions equipped with CFRP tabs (see Fig. 7). In order to avoid tensile failure of
63 × 315 × 8mm3 ; the thickness and the width of the strengthening are the free mesh outside the FRCM strengthening, this was also reinforced
the same as considered for the coupon. with a CFRP composite (see Fig. 7b). During the test, two couples of
The geometric characteristics of the strengthening and the distances LVDTs measured the slip (relative displacement between the mesh and
from the edges of the reinforced surface agree with the indications the masonry substrate) at the upper and lower ends of the re­
provided by [16]. The masonry specimen was manufactured and cured inforcements. Moreover, the FO sensor measured the strain of the mesh
at room conditions for at least 28 days. Then, the surface to be reinforced at the positions “S1”-“S6” indicated in Fig. 7a. The test was carried out
was sanded and cleaned and the FRCM reinforcement was realized under displacement control at a machine stroke rate of 0.2mm/min.

5
E. Bertolesi et al. Construction and Building Materials 326 (2022) 126640

Fig. 8. SLS test: load-slip diagram (-a) and failure mode (-b).

Fig. 9. Strain evolution tracked with the fiber optic sensor (colormap defined on the average displacement from vertical LVDTs – mm).

The “load”-“average top slip” equilibrium path obtained from the 3. Calibration procedure for the interfacial bond-slip law
test is reported in Fig. 8a, while the failure mode observed at the end of
the test is shown in Fig. 8b. As clearly visible from Fig. 8a, the specimen The single lap shear test results are considered in this section to
showed an initial elastic phase with gradually decreasing stiffness calibrate the interfacial bond-slip law obtained using FO sensors applied
approaching the peak load. The decreasing stiffness was caused by the on the textile grid. The total bond length considered in this study was
progressive cracking of the mortar matrix. No slippage was observed at 315 mm. This value is higher than the effective bonding length esti­
this point. After reaching the peak load, the specimen started experi­ mated in [30] in 260 mm with reference to the same reinforcing system
encing debonding phenomena involving the detachment of the textile considered in the paper but bonded to concrete substrate. The portion of
from the mortar matrix. the reinforcement monitored during the test using FOs sensors had al
length of 130 mm (see Fig. 7 and Fig. 9). Such length, although less than
the effective length, was sufficient to catch all the branches of the

6
E. Bertolesi et al. Construction and Building Materials 326 (2022) 126640

Where, ER, AR and bR are the Elastic modulus, cross section and width of
PBO bundles, respectively. In this study, the following values were
considered to obtain the tangential stress distribution: ER=240 GPa,
AR=0.405 mm2 and bR=4 mm. It is worth mentioning that the Elastic
modulus was assumed equal to the value provided by the manufacturer,
the bundle cross section was estimated from the equivalent thickness
provided by the manufacturer and the bundle width was measured
experimentally.
In order to define the slip distribution in the FRCM sample, the
following assumptions were made: (i) perfect adhesion between PBO
textile and mortar matrix at the beginning of the test at the strength­
ening free end; (ii) the debonding process is considered to take place at
the interface between textile grid and mortar matrix only; (iii) linear
variation of strains in the textile grid between two subsequent fiber optic
sensors. Then, integration of strain profile, starting from the free end of
the strengthening material and assuming the slip at the free end to be
Fig. 10. Interfacial tangential bond-slip law calibrated using fiber zero, leads to the following expression for the slip at a general abscissa
optic sensors. ×, with xi ⩽ x ⩽ xi+1:
(εi + εi− 1 )*(xi − xi− 1 )
interface law. Fig. 9 depicts the strain evolutions obtained with the fiber si = si− 1 + + εi− 1 *(xi − xi− 1 ) (3)
2
optic sensors (Sn, n=1,2,3,4,5,6) plotted against their position along the
Finally, the average slip is calculated as:
FRCM sample. As clearly visible, the strain evolutions indicate the rising
of strain levels close to the loaded end and the progressive strain shifts in (si + sj )
the more distant parts as the anchorage length activates and the si− j = (4)
2
debonding process takes place. It is worth noting, FO sensors S6 and S5
In the following section, the tangential bond-slip law obtained
recorded lower maximum strain which may be explained by edge effects
experimentally (Fig. 10) using the procedure described in this section, is
or telescopic failures well documented in inorganic matrix-based
used to calibrate the analytical interfacial bond-slip law to be used in
strengthening materials [10,27–28]. This effect is currently under
analytical models developed by the authors.
investigation with suitable sophisticated numerical and analytical
models. The calibration procedure herein discussed resemble the one
4. Analysis of the shear and tensile experimental behaviour by
provided in [42] used for FRP strengthening materials neglecting the
means theoretical models
contribution coming from the support deformability. Indeed, although
in current literature are available studies which specifically investigate
The experimental outcomes and derived data are here analyzed
the interaction between the grid and matrix by also taking into account
throughout theoretical models finalized to simulate the behavior of
the different observed behaviour at the lower and upper grid-matrix
specimens subjected to shear-lap tests, where the behavior at the
interface [43], in the present study the calibration of the parameters
reinforcement-matrix interface is the governing phenomenon, and, in
characterizing the grid-matrix interface law on the basis of the experi­
addition, the tensile behavior of TRM/FRCM coupon, where the damage
mental measures is carried out by assuming the same constitutive
of matrix generally governs the response.
behaviour at the both lower and upper interface and, moreover,
considering an elastic behaviour for the textile grid and assuming that
the deformation takes place at the interface between the textile and the 4.1. Specimens subjected to shear-lap tests
lower mortar layer, average value of shear stress between two subse­
quent fiber optic sensors can be written as follows: The numerical model proposed in the present Section is aimed at
reproducing the experimental results obtained previously for specimen
ER *AR *(εi+1 − εi )
τ1− 2 = (2) subjected to shear-lap test. The preliminary target is to fit the global
bR *(xi+1 − xi )
load-displacement response of the specimen experimentally tested. To
this scope, all the non-linearities are lumped in a fictious interface

Fig. 11. Flat strengthening pack. Geometry of the strengthening system (-a) and governing equilibrium equation at the interface, with the definition of slip (-b).

7
E. Bertolesi et al. Construction and Building Materials 326 (2022) 126640

d2 s(x)
ER tR = τ(s) (6)
dx2
Eq. (6) is coupled to initial conditions imposed on the free edge. In
this way, the following Cauchy problem is obtained, which can be
numerically solved by means of a standard ODE45 Runge-Kutta
algorithm:

d2 s(x) τ(s)
=
dx2 ER tR
s(0) = s0 (7)
⃒
⃒
ds(x) ⃒ σ R (0)
= =0
dx ⃒x=0 ER

It is interesting to point out that the imposition of a given slip s0 at

the free edge allows to transform a Boundary Value Problem BVP into a
Cauchy problem (which can be tackled with explicit algorithms and
therefore does not exhibit numerical convergence problems). From a
Fig. 12. Comparison among experimental interface τ(s) − s relationships and mechanical point of view, the reproduction of a possible snap-back
three numerical models (linear-exponential, exponential and multi-linear) during the debonding of the interface at late stages of the loading pro­
adopted to fit experimental data. cess is very straightforward. In the present study, three different τ(s) − s
relationships are assumed for the interface: (i) linear-exponential [2],
located between the support and the strengthening material. Even (ii) exponential [344] and (iii) multi-linear.
though the authors are aware that such an assumption represents a The linear-exponential function is constituted by two phases; the first
rough simplification of the actual local behavior of a TRM/FRCM rein­ phase is linear and characterized by a stiffness equal to k. Phase 1 holds
forcing system, because simplistically derived from approaches already before the peak tangential stress fb is reached. The second phase is
in use for FRP, it can be considered reliable and globally predictive when exponential with softening, ruled by the fracture energy GII. The
the mode of failure is known, for instance from experimental evidence as analytical expressions of the two phases are given below:
in the case here analyzed. Having at disposal an experimental interface
τ(s) = ks(x) (8)
relationship to which to refer, a fair reproduction of the global load-
displacement response in a single lap shear test is possible. Consid­ τ(s) = fb ∙e
− (s− s* )∙fb
GII −
s∙fb
= τ0 ∙e , with τ0 = fb ∙e (9)
GII
s* fb
GII

ering a TRM/FRCM or an FRP strip externally bonded to a flat surface, as In Eq. (9) s* represents the slip value at the end of the linear elastic
in Fig. 11, it is assumed that the debonding occurs at the interface be­ phase and all the other symbols have been already introduced.
tween the cementitious matrix and the reinforcing pack, without any The second interface model is characterized by a single exponential
damage inside either the support or the strengthening material [1–3]. phase that describes both the pre-peak and the softening behavior. The
Under such hypothesis, the obtained results turn out to be independent corresponding τ(s) − s relationship adopted is the following:
from the substrate properties. In addition, the strengthening material ( )
(width equal to BR) is assumed elastic with an equivalent stiffness equal λ s2
( 2
)
to ER tR (Fig. 11-b), where ER is the elastic modulus and tR the equivalent s 2 1− s20 − 2λ s2
τ(s) = (τ* − τr ) e + τ r 1 − e s0 (10)
thickness. Furthermore, the inelastic interface between reinforcement s0
and substrate is supposed characterized by a Mode II fracture behavior
In the previous equation s0 is a slip parameter, τ* and τr two stresses
ruled by a tangential stress τ(x)-slip s(x) relationship tuned on experi­
with a clear physical meaning and λ a numerical coefficient ruling the
mental data available.
shape of the softening phase. It is interesting to point out that Eq. (10)
As far as the numerical parameters adopted for the reinforcement in
has the property that lim τI = τr , so that τr exhibits the physical
the simulations are concerned, the following numerical values are s→+∞

adopted, in agreement with experimental measurements: ER=240 GPa, meaning of residual tangential stress, a feature that allows the utilization
which corresponds to PBO dry elastic modulus; tR=0.405/4 mm (cross of Eq. (10) for curved substrates, but that appears also suitable to study
section of a single bundle divided by its width, assuming the cross sec­ TRM/FRCM when the reinforcement grid slips inside the mortar.
tion rectangular for the sake of simplicity), BR=4×7 mm, considering 7 The third interface model adopted for the simulations is multi-linear.
bundles with width equal to 4 mm. The adoption for ER of the PBO dry The corresponding τ(s) − s relationship is derived from the knowledge of
elastic modulus is justified by the mode of failure experimentally the tangential stresses at n different slips. Assuming that at two
observed, which relies into a slippage of the grid inside the cementitious consecutive slips si and si+1 the tangential stresses are respectively equal
matrix. to τi and τi+1 , the i-th branch of the interface (s) − s law is expressed
With reference to Fig. 11b, by imposing the equilibrium along the analytically as follows:
longitudinal direction on a portion of the reinforcing pack, the following
s(x) − si
equation is obtained: τ(s) = τi + (τi+1 − τi ) (11)
si+1 − si
dσ R
dxtR BR − τ(s, x)BR dx = 0 (5) According to the experimental data collected in the previous Sec­
dx
tions, the three numerical τ(s) − s laws depicted in Fig. 12 are considered
Assuming the reinforcement elastic, then σ R = ER εR (x) = ER ds(x) dx
, and their performance in reproducing the experimental global load-
where εR (x) is the axial strain of the reinforcement and s(x) is the slip of displacement curve carefully evaluated.
the reinforcement over the support. The parameters assumed for the first and second interface models are
Substituting such relation in the previous equation, the field equa­ the following: for the first model fb =2.1 MPa, s* =0.075 mm and
tion for the debonding of the reinforcement pack is obtained: GII=0.3 N/mm; for the second models0 =0.035 mm, τ* =1.9 MPa, τr =0.3
MPa andλ=0.42. The solution of the Cauchy problem of Eq. (7) allows to
obtain the global force–displacement curves depicted in Fig. 13 and

8
E. Bertolesi et al. Construction and Building Materials 326 (2022) 126640

Assuming a linear behaviour along the zone dx of both matrix and

reinforcement:
duR
σ R = ER
dx

duc
σ c = Ec (13)
dx

Table 4
Parameters introduced into the theoretical model for the tensile behavior of
TRM/FRCM-coupon.
Parameter symbol [unit] value

reinforcement Young’s modulus ER [MPa] 240000

matrix Young’s modulus Ec [MPa] 7500
reinforcement equivalent thickness tR [mm] 0.045
matrix one layer thickness tc [mm] 4
reinforcement width bR [mm] 63
matrix width bc [mm] 63
Fig. 13. Comparison between experimental global load-displacement curve
specimen length L [mm] 250
and numerical predictions provided by linear-exponential, exponential and
matrix tensile strength fct [MPa] 2.0
multi-linear models. bond strength fb [MPa] 1.9
slip at the end of the ascending branch s0 [mm] 0.05
compared with the experimental data. As it is possible to notice slip at the end of the descending branch sf [mm] 0.17

comparing the numerical results with the experimental load-

displacement curve, the fitting is quite satisfactory in all the cases,
considering also the rough simplifications adopted in the models and the
straightforwardness of the procedure proposed. It is worth observing
that the multi-linear numerical model seems the most adequate to fit
global experimental data. This is not surprising, because the interface
τ(s) − s relationship adopted is capable, among the others, to closely
replicate the experimental local behavior of the interface, see Fig. 12.

4.2. Specimen subjected to tensile test

A theoretical model similar to the one previously proposed for the

case of shear lap tests is here employed to analyse the results deduced
from the tensile test. The model is directly based on previous studies
carried out by the Authors and already presented in [18,45,46].
Considering the static conditions of an infinitesimal longitudinal
portion of reinforcement and a layer of matrix (Fig. 14), the translational
equilibrium equations along the longitudinal direction of the rein­
forcement allow to write the following system of differential equations:
{
dσR BR tR − 2τ(s, x)BR dx = 0
(12)
dσc Bc tc − τ(s, x)BR dx = 0
Fig. 15. Accounted crack pattern deduced from both experimental outcomes
where, additionally to symbols used in the previous model, Bc and tc are
(left picture) and theoretical model (right picture).
respectively the width and thickness of the matrix layer.

Fig. 14. Schemes accounted for deriving the equation at the basis of the theoretical model.

9
E. Bertolesi et al. Construction and Building Materials 326 (2022) 126640

Fig. 16. Results deduced from the theoretical model at the following steps: attainment of matrix tensile strength after the formation of the third crack (-a); formation
of the sixth crack (-b).

where σc and Ec are respectively the normal stress and the Young’s where u is the displacement imposed during the test at the loaded end
modulus of matrix. and L is the bond length of the specimen.
Introducing the slip s between the reinforcement and the matrix: On the other hand, the formation of cracks along the matrix requires
to introduce additional conditions along the length, because the field
s = uR − uc (14)
equilibrium equation on the matrix holds on single portions of the
where uR and uc are respectively the axial displacement of reinforcement coupon separately, with an increase of the number of equations gov­
and matrix, it follows: erning the problem. They depend on the number of cracks and their
position, and generally consist of imposing the same value of displace­
⎧
⎪
⎪ d2 s ment and normal stress of the reinforcement and a null value of normal
⎪ − K1 τ(s, x) = 0
⎨ dx2 stress of matrix, at the two sections across the same crack [18,45,46].
(15)
⎪
⎪
( 2
d uR d2 s
) Additionally, further conditions could be necessary depending on the
⎪
⎩
dx2
−
dx 2
+ K2 τ(s, x) = 0 behaviour and the assumed constitutive shear stress-slip law of the
reinforcement-matrix interface, as extensively discussed in [18,45,46]
where K1 and K2 are two constants respectively equal to: where the reader is referred for further details. Indeed, since both the
cracking of matrix and the de-bonding of grid from matrix are the two
K1 =
2 possible phenomena specifically accounted for in the theoretical model,
ER tR the set of equations governing the problem (number of equations and
parameters characterizing the equations themselves) is strictly depen­
bR
K2 = (16) dent on their occurrence during the increasing of the applied displace­
bc Ec tc
ment. In particular, while the cracking of matrix mainly leads to an
The above system of equations can be solved as a BVP (Boundary increase of the number of equations depending on the number of cracks,
Value Problem) adding boundary conditions, which depend on both the the occurrence of de-bonding introduces modifications of the set of
setup configuration (conditions imposed at the ends of reinforcement) equations governing the problem depending on the type of constitutive
and on the pattern charactering the formation of cracks along the matrix law assumed for the interface [18,45,46]. For instance, considering a
[18,45,46]. multilinear law for the mesh-matrix interface, the attainment of the
As far as the first conditions are concerned, taking into account the bond strength in some sections of specimen leads to the need to intro­
configuration of the setup here employed for tensile tests, they are the duce additional equations necessary to describe the behaviour of the
following: portions of specimen where the interface is in the pre-peak phase and the
ones where the interface is in the post-peak phase [18,45,46].
x = 0 : uR = 0; uc = 0
The solution of the above system allows to simulate the tensile
x = L : uR = u; uc = u (17) behaviour of TRM/FRCM coupons by analysing their behaviour at the
local level.

10
E. Bertolesi et al. Construction and Building Materials 326 (2022) 126640

Fig. 17. Average shear stresses between two contiguous sensors derived at different load levels registered by the load cell during the tests at different level of the
peak load Fmax: 5%Fmax (-a); 10%Fmax (-b); 15%Fmax (-c); 20%Fmax (-d); 30%Fmax (-e); 40%Fmax (-f); 50% Fmax (-g); 60%Fmax (-h); 70%Fmax (-i); 80%Fmax (-l).

11
E. Bertolesi et al. Construction and Building Materials 326 (2022) 126640

mentioning:
- the first two cracks are located at the loaded ends of the specimen
whilst the third one at x=160 mm: the position of such cracks has been
directly deduced from experimental observations carried out during the
tests and considering in the theoretical model the attainment of the
tensile strength of the matrix in such positions (Fig. 16a) – the position of
the subsequent three cracks is directly deduced from the theoretical
model at the sections where the tensile strength of the matrix is pro­
gressively attained (Fig. 16b).
The obtained crack pattern well approximates the experimentally
observed one (see Fig. 15). Therefore, this crack pattern is here selected
in order to perform a comparison at the local level between the theo­
retical and experimental results, the latter deduced from the measure­
Fig. 18. Max and Min values of average shear stresses between two contiguous ments provide by optical sensors.
sensors evaluated at the different load levels. In particular, considering the experimental data in terms of strains
directly obtained from optic sensors, an approximate estimation of the
average value of shear stress τij between two contiguous sensors i and j at
one of the two interface layers can be obtained similarly to shear-lap test
(see Eq. (2)), for the case of tensile test, where an equal value of shear
stresses at the two interfaces is considered:
ER tR εj − εi
τij = (18)
2 xj − xi

The obtained values are presented in Fig. 17 at different levels of

applied load (from 5% of the peak load to 80% of the peak load) by
reporting with circular symbols the values of ij at the sections halfway
between two contiguous sensors. Significant values of shear stresses can
be observed (with consequent slip between reinforcement and matrix)
for load levels greater than 10%. The subsequent load steps are char­
acterized by values of average shear stresses with opposite signs: this
effect is certainly due to the formation of cracks in the matrix. In
Fig. 19. Accounted load steps for deriving average shear stresses from strain particular, for both steps at 15%Fmax and 20%Fmax shear stresses with
measurements. the highest value are located next to the section where the third crack
occurred (x=160 mm).
As far as the tensile test experimentally analysed in this paper is A further important observation from the average shear stresses
concerned, the above model is applied by considering the parameters obtained from strain measurements can be deduced by analysing their
reported in Table 4 (meaning of s0 and sf is depicted in Fig. 12) and peak values (both negative and positive) at different load steps, graph­
assuming a crack pattern composed of six macro-cracks obtained by ically presented in Fig. 18.
considering both the experimental outcomes and results of the theo­ Indeed, it is possible to observe that, until a load level equal to about
retical model itself (see Fig. 15). In addition, one of the shear stress-slip 20%Fmax, an increase of the peak values of shear stresses is observed
laws previously deduced from the shear-lap test is also adopted. (the maximum occurs at a load level equal to 18%Fmax); differently, the
Regarding the crack pattern, the following aspects are worth subsequent load steps are characterized by peak values of shear stresses

Fig. 20. Comparison between average shear stresses deduced from strain measurements (circular symbols) and shear stresses obtained from the theoretical model at
the formation of six cracks (dotted lines). In the plot vertical dotted lines schematize the cracks position.

12
E. Bertolesi et al. Construction and Building Materials 326 (2022) 126640

not varying in a significant manner. Such outcome is a direct effect of the of the matrix damage, and the bond behaviour at the level of the matrix-
crack pattern: most macro-cracks form indeed in the range 15–20% strengthening interface.
Fmax, whilst the subsequent steps are mainly influenced by the opening
of the cracks. This phenomenon can be also deduced from the Force- CRediT authorship contribution statement
Strain curve (see Fig. 19), where the zone characterized by the lower
slope lies in this load range. Elisa Bertolesi: Methodology, Conceptualization, Formal Analysis,
In conclusion, in order to compare the results deduced from the Data curation, Investigation, Writing – original draft, Writing – review &
theoretical model and those obtained from the strain measurements, the editing, Visualization. Mario Fagone: Methodology, Conceptualization,
average shear stress values deduced from Eq. (1) at the load step 18% Formal Analysis, Data curation, Investigation, Writing – original draft,
Fmax (i.e. when all macro-cracks form in the experimentation carried Writing – review & editing, Visualization. Tommaso Rotunno: Meth­
out) are compared with the corresponding results deduced from the odology, Conceptualization, Formal Analysis, Data curation, Investiga­
theoretical model when the formation of all six cracks occurs. The tion, Writing – original draft, Writing – review & editing, Visualization.
comparison presented in Fig. 20 shows a good prediction of the theo­ Ernesto Grande: Methodology, Conceptualization, Formal Analysis,
retical model particularly near the zone where the third crack arises (i.e., Data curation, Investigation, Writing – original draft, Writing – review &
the one imposed in the theoretical model on the base of the experimental editing, Visualization. Gabriele Milani: Methodology, Conceptualiza­
observation). tion, Formal Analysis, Data curation, Investigation, Writing – original
Finally, it is also important to observe that the values of shear stresses draft, Writing – review & editing, Visualization.
obtained from both experimental data and theoretical model do not
reach the bond strength: this confirms in the performed tensile tests the Declaration of Competing Interest
predominance of the damage of matrix instead of the de-bonding at the
reinforcement-matrix interface. The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
5. Conclusions the work reported in this paper.

The present work provides valuable experimental results to calibrate

References
the interfacial bond-slip law in TRM/FRCM strengthening materials and
a validation of those results developed using closed form solutions [1] J. Vaculik, P. Visintin, N.G. Burton, M.C. Griffith, R. Seracino, State-of-the-art
already published by the authors of this work. The novelty is represented review and future research directions for FRP-to-masonry bond research: Test
methods and techniques for extraction of bond-slip behaviour, Construction and
by the adoption of distributed fiber optic sensors comprising 6 sensing
Building Materials 183 (2018) 325–345.
points installed on PBO textiles embedded into TRM/FRCM strength­ [2] G. Milani, E. Grande, E. Bertolesi, T. Rotunno, M. Fagone, Debonding mechanism
ening materials. The FOs sensors were used to track the local strain of FRP strengthened flat surfaces: Analytical approach and closed form solution,
evolution in TRM/FRCM materials subjected to: (i) tensile tests and (ii) Construction and Building Materials 302 (2021), 124144.
[3] G. Milani, M. Fagone, T. Rotunno, E. Grande, E. Bertolesi, Development of an
single laps shear tests. From the experimental and analytical in­ interface numerical model for C-FRPs applied on flat and curved masonry pillars,
vestigations, the following conclusions can be drawn: Composite Structures 241 (2020), 112074.
[4] H. Yuan, X. Lu, D. Hui, L. Feo, Studies on FRP-concrete interface w hardening and
softening bond-slip law, Composite Structures 94 (12) (2012) 3781–3792.
• Compared with traditional sensors (i.e., LVDTs and Strain Gauges), [5] U. De Maio, F. Greco, L. Leonetti, P.N. Blasi, A. Pranno, An investigation about
FOs provided valuable data on the local behaviour of TRM/FRCM debonding mechanisms in FRP-strengthened RC structural elements by using a
coupons subjected to tensile and single laps shear tests. cohesive/volumetric modeling technique, Theoretical and Applied Fracture
Mechanics 117 (2022), 103199.
• The obtained data allow for studying the tensile behaviour by ana­ [6] S. Marfia, E. Sacco, J. Toti, A coupled interface-body nonlocal damage model for
lysing at the local level the influence of crack pattern on the tensile the analysis of FRP strengthening detachment from cohesive material, Frattura ed
response of coupons. the results obtained from the theoretical model Integrità Strutturale 18 (2011) 23–33.
[7] R. Fedele, G. Milani, ‘A numerical insight into the response of masonry reinforced
provide both the assessment of the laws obtained from experimental by FRP strips, The case of perfect adhesion’, Composite Structures 92 (10) (2010)
tests and, at the same time, the reliability of data furnished by the 2345–2357.
monitoring system itself. [8] C.A. Coronado, M.M. Lopez, Numerical modeling of concrete–FRP debonding using
a crack band approach, J Compos Constr ASCE 14 (2010) 11–21.
• The local FO strain evolutions allowed to monitor the behaviour of
[9] F.G. Carozzi, P. Colombi, G. Fava, C. Poggi, A cohesive interface crack model for
TRM/FRCM strengthening materials tested in single lap shear tests the matrix–textile debonding in FRCM composites, Composite Structures 143
and thus, the definition of an interfacial bond-slip law. (2016) 230–241.
• The experimentally obtained bond-slip law was successfully used to [10] L.A.S. Kouris, T.C. Triantafillou, State-of-the-art on strengthening of masonry
structures with textile reinforced mortar (TRM), Construction and Building
calibrate different analytical models developed by the authors. The Materials 188 (2018) 1221–1233.
analytical force-slip curve obtained using the calibrated bond-slip [11] V. Alecci, M. De Stefano, R. Luciano, L. Rovero, G. Stipo, Experimental
law showed a very good agreement with the global experimental investigation on bond behavior of cement-matrix–based composites for
strengthening of masonry structures, Journal of Composites for Construction 20 (1)
force-slip curve. (2016) 04015041.
[12] G. De Felice, S. De Santis, L. Garmendia, B. Ghiassi, P. Larrinaga, P.o.P.B. Lourenc,
Finally, it is important to underline that the paper shows how the use D.V. Oliveira, F. Paolacci, C.G. Papanicolaou, Mortar-based systems for externally
bonded strengthening of masonry, Materials and structures 47 (12) (2014)
of optical fiber sensors coupled with TRM materials can shed light on 2021–2037.
aspects that have so far been neglected in previous studies. The cali­ [13] P.D. Askouni, C.G. Papanicolaou, Experimental investigation of bond between
bration of the interface law on the basis of the results obtained and its glass textile reinforced mortar overlays and masonry: the effect of bond length,
Materials and Structures 50 (2) (2017) 164.
subsequent use in the analytical model demonstrates that such sensors [14] F.G. Carozzi, A. Bellini, T. D’Antino, G. de Felice, F. Focacci, Ł. Hojdys, L. Laghi,
can be effectively applied in these materials and, the results obtained E. Lanoye, F. Micelli, M. Panizza, et al., Experimental investigation of tensile and
analytically (but based on experimental data not previously available) bond properties of carbon-FRCM composites for strengthening masonry elements,
Composites Part B: Engineering 128 (2017) 100–119.
are consistent with traditionally obtained measurements and available
[15] S. Barducci, V. Alecci, M. De Stefano, G. Misseri, L. Rovero, G. Stipo, Experimental
in the technical literature. and analytical investigations on bond behavior of Basalt-FRCM systems, Journal of
In conclusion, the monitoring system based on the use of optic sensor Composites for Construction 24 (1) (2020) 04019055.
is a valuable strategy for experimentally characterizing the behaviour of [16] G. De Felice, M.A. Aiello, C. Caggegi, F. Ceroni, S. De Santis, E. Garbin, N. Gattesco,
Ł. Hojdys, P. Krajewski, A. Kwiecien, et al., Recommendation of RILEM ́ technical
TRM and FRCM strengthening systems at local level, both regarding the committee 250-CSM: Test method for textile reinforced mortar to substrate bond
tensile behaviour of coupons, important for understanding the influence characterization, Materials and Structures 51 (4) (2018) 95.

13
E. Bertolesi et al. Construction and Building Materials 326 (2022) 126640

[17] G. De Felice, T. D’Antino, S. De Santis, P. Meriggi, F. Roscini, Lessons learned on [33] Y. Li, J. Bielak, J. Hegger, R. Chudoba, An incremental inverse analysis procedure
the tensile and bond behavior of fabric reinforced cementitious matrix (FRCM) for identification of bond-slip laws in composites applied to textile reinforced
composites, Frontiers in Built Environment 6 (2020) 5. concrete, Composites Part B: Engineering 137 (2018) 111–122.
[18] E. Grande, M. Imbimbo, E. Sacco, Numerical investigation on the bond behavior of [34] C. Du, S. Dutta, P. Kurup, T. Yu, X. Wang, A review of railway infrastructure
FRCM strengthening systems, Composites Part B: Engineering 145 (2018) monitoring using fiber optic sensors, Sensors and Actuators A: Physical 303 (2020),
240–251. 111728.
[19] E. Grande, B. Ghiassi, M. Imbimbo, Theoretical and FE models for the study of the [35] L. Fan, Y. Bao, W. Meng, G. Chen, In-situ monitoring of corrosion-induced
bond behavior of FRCM systems, Woodhead Publishing Series in Civil and expansion and mass loss of steel bar in steel fiber reinforced concrete using a
Structural Engineering, Numerical Modeling of Masonry and Historical Structures distributed fiber optic sensor, Composites Part B: Engineering 165 (2019) 679–689.
From Theory to Application, 2019, pp. 685–712. [36] B. Schranza, M.F. Nunesa, C. Czaderski, M. Shahverdia, Fibre optic strain
[20] E. Grande, G. Milani, Modeling of FRCM strengthening systems externally applied measurements for bond modelling of prestressed near-surface-mounted iron-based
on curved masonry substrates, Eng Struct 233 (2021), 111895. shape memory alloy bars, Construction and Building Materials 288 (2021),
[21] E. Grande, G. Milani, Procedure for the numerical characterization of the local 123102.
bond behavior of FRCM, Compos Struct 258 (2021). [37] M. Saidi, A. Gabor, Experimental analysis of the tensile behaviour of textile
[22] E. Grande, G. Milani, Numerical simulation of the tensile behavior of FRCM reinforced cementitious matrix composites using distributed fibre optic sensing
strengthening systems, Compos Part B Eng 189 (2020). (DFOS) technology, Construction and Building Materials 230 (2020), 117027.
[23] E. Grande, G. Milani, Interface modeling approach for the study of the bond [38] W. Brameshuber, et al., Recommendation of RILEM TC 232-TDT: test methods and
behavior of FRCM strengthening systems, Compos Part B Engineering 141 (2018) design of textile reinforced concrete: Uniaxial tensile test: test method to determine
221–233. the load bearing behavior of tensile specimens made of textile reinforced concrete,
[24] E. Grande, G. Milani, M. Imbimbo, Theoretical model for the study of the tensile Mater. Struct. Constr. 49 (2016) 4923–4927.
behavior of FRCM reinforcements, Constr Build Mater 236 (2020). [39] T. Rotunno, M. Fagone, E. Bertolesi, E. Grande, G. Milani, Single lap shear tests of
[25] E. Bertolesi, B. Torres, J.M. Adam, P. Calderon, J.J. Moragues, Effectiveness of masonry curved pillars externally strengthened by CFRP strips, Compos. Struct.
Textile Reinforced Mortar (TRM) materials for the repair of full-scale timbrel 200 (2018) 434–448.
masonry cross vaults, Engineering Structures 220 (2020), 110978. [40] M. Fagone, G. Ranocchiai, S. Briccoli Bati, An experimental analysis about the
[26] E. Bertolesi, M. Buitrago, E. Giordano, P.A. Calderon, J.J. Moragues, F. Clementi, J. effects of mortar joints on the efficiency of anchored CFRP-to-masonry
M. Adam, Effectiveness of textile reinforced mortar (TRM) materials in preventing reinforcements, Compos. Part B-Engineering 76 (2015) 133–148.
seismic-induced damage in a u-shaped masonry structure submitted to pseudo- [41] European Standard, EN 998-2 Specification for mortar for masonry – Part 2:
dynamic excitations, Construction and Building Materials 248 (2020), 118532. Masonry mortar, (2016).
[27] L. Ascione, G. de Felice, S. De Santis, A qualification method for externally bonded [42] B. Ferracuti, M. Savoia, C. Mazzotti, Interface law for FRP–concrete delamination,
fibre reinforced cementitious matrix (FRCM) strengthening systems, Composites Composite Structures 80 (4) (2007) 523–531.
Part B: Engineering 78 (2015) 497–506. [43] C. Carloni, T. D’Antino, L.H. Sneed, C. Pellegrino, Role of the Matrix Layers in the
[28] S. De Santis, F.G. Carozzi, G. de Felice, C. Poggi, Test methods for Textile Stress-Transfer Mechanism of FRCM Composites Bonded to a Concrete Substrate,
Reinforced Mortar systems, Composites Part B: Engineering 127 (2017) 121–132. Journal of Engineering Mechanics 141 (6) (2015).
[29] E. Bertolesi, F.G. Carozzi, G. Milani, C. Poggi, Numerical modeling of fabric [44] G. Milani, E. Grande, Simple bisection procedure in quickly convergent explicit
reinforce cementitious matrix composites (FRCM) in tension, Construction and ODE solver to numerically analyze FRCM strengthening systems, Composites Part
Building Materials 70 (2014) 531–548. B: Engineering 199 (2020) #108322, https://doi.org/10.1016/j.
[30] L. Sneed, T. D’Antino, C. Carloni, C. Pellegrino, A comparison of the bond behavior compositesb.2020.108322.
of PBO-FRCM composites determined by double-lap and single-lap shear tests, [45] Grande E., Imbimbo M., Sacco E. (2017), ‘Local bond behavior of frcm
Cement and Concrete Composites 64 (2015) 37–48. strengthening systems: Some considerations about modeling and response’, Key
[31] P. Askouni, C. Papanicolaou, Comparison of Double-Lap/Double-Prism and Single- Eng. Mater., vol. 747 KEM, 2017, p. 101–7.
Lap/Single-Prism Shear Tests for the TRM-to-Masonry Bond Assessment, [46] E. Grande, M. Imbimbo, S. Marfia, E. Sacco, Numerical simulation of the de-
International Conference on Strain-Hardening Cement-Based Composites, RILEM bonding phenomenon of FRCM strengthening systems, Frat Ed Integrita Strutt 13
Bookseries 15 (2018) 527–534. (2019) 321–333.
[32] A. Dalalbashi, B. Ghiassi, D.V. Oliveira, A. Freitas, Fiber-to-mortar bond behavior
in TRM composites: Effect of embedded length and fiber configuration, Composites
Part B: Engineering 152 (2018) 43–57.

14