1
MORTAR-BASED SYSTEMS FOR EXTERNALLY
BONDED STRENGTHENING OF MASONRY
Gianmarco de Felice
1,*
, Stefano De Santis
1
, Leire Garmendia
2
, Bahman
Ghiassi
3
, Pello Larrinaga
2
, Paulo B. Lourenço
3
, Daniel V. Oliveira
3
, Fabrizio
Paolacci
1
,
Catherine G. Papanicolaou
4
1
Department of Engineering, Roma Tre University, Rome, Italy
2
Tecnalia Research & Innovation, Bilbao, Spain
3
Department of Civil Engineering, ISISE, University of Minho, Guimaraes, Portugal
4
Department of Civil Engineering, University of Patras, Patras, Greece
* Corresponding author:
E-mail: gianmarco.defelice@uniroma3.it
Telephone: +39.06.5733.6268
Fax: +39.06.5733.6265
ABSTRACT
Mortar-based composite materials appear particularly promising for use as
externally bonded reinforcement (EBR) systems for masonry structures.
Nevertheless, their mechanical performance, which may significantly differ from
that of Fibre Reinforced Polymers, is still far from being fully investigated.
Furthermore, standardized and reliable testing procedures have not been defined
yet. The present paper provides an insight on experimental-related issues arising
from campaigns on mortar-based EBRs carried out by laboratories in Italy,
Portugal and Spain. The performance of three reinforcement systems made out of
steel, carbon and basalt textiles embedded in inorganic matrices has been
investigated by means of uniaxial tensile coupon testing and bond tests on brick
and stone substrates. The experimental results contribute to the existing
knowledge regarding the structural behaviour of mortar-based EBRs against
tension and shear bond stress, and to the development of reliable test procedures
aiming at their homogenization/standardization.
Keywords: Masonry, Mortar-based composites, Tensile tests, Bond tests.
1. INTRODUCTION
An increasing attention has been given in the recent years to the development
of innovative technologies based on the use of composite materials for
strengthening masonry structures by applying externally bonded reinforcement
systems. Applications of Fibre Reinforced Polymers (FRP) to vaults, columns and
walls have demonstrated their effectiveness in increasing the load-carrying
capacity and in upgrading the seismic strength (Triantafillou and Fardis, 1997;
2
Valluzzi et al., 2001; Corradi et al., 2002; Ascione et al., 2005; Grande et al.,
2011; Oliveira et al., 2011; Valluzzi et al., 2014). During the past decade, in an
effort to alleviate certain drawbacks associated to the organic character of
polymer-based composites, fibre-reinforced inorganic matrix composites have
been developed. This broad category includes Steel Reinforced Grouts (SRG,
unidirectional steel cords embedded in a cement or lime grout) and Fabric-
Reinforced Cementitious Matrix (FRCM) composites (a sequence of one or more
layers of cement-based matrix reinforced with dry fibres in the form of open
single or multiple meshes, Babaeidarabad et al., 2013)
Inorganic matrices may exhibit lower bond strength with respect to FRPs, due
to the possible occurrence of failure modes within the reinforcement rather than
within the substrate. However, they are advantageous in terms of overlay-to-
substrate compatibility, transpirability, reversibility, fire resistance, cost, and
applicability (Papanicolaou et al., 2007; 2008; Cancelli et al., 2007; Carbone and
de Felice 2008; Borri et al., 2011; Garmendia et al., 2011; Malena and de Felice,
2014). Moreover, they seem to be particularly appropriate for application to
masonry structures, since the higher bond strength of polymeric matrices cannot
be fully exploited because of the low intrinsic mechanical characteristics of the
substrate (Oliveira et al., 2011; Garmendia et al., 2012; Grande et al., 2013;
Ceroni et al., 2014). Nevertheless, a deeper knowledge needs to be gained for
designing mortar-based strengthening systems that are suitable for application to
masonry substrates, as well as for identifying their mechanical properties (e.g.,
under direct tension or shear bond stress) through standardized testing
methodologies.
Systemized research on similar systems, such as Textile Reinforced Concrete
(TRC), has been recently conducted (Brameshuber, 2006). Despite the fact that
the main target of TRC was originally integration in new civil applications rather
than strengthening of existing ones, strong analogies exist between mortar-based
reinforcement systems and TRC on numerous key issues, including testing
methods (Contamine et al., 2011; Hartig et al., 2012, Hegger et al., 2006; Häußler-
Combe and Hartig, 2007; Colombo et al., 2013), durability and mechanical
behaviour. TRC matrices usually consist of high performance finely grained
cement concrete, while lime-based mortars might be preferred when strengthening
a masonry structure to fulfil moisture compatibility and reversibility requirements.
As for the reinforcement textiles, beyond those typically used in TRC (glass,
carbon or aramid fibre bundles), steel cords (Borri et al., 2011), basalt (Balsamo et
al., 2011), and natural fibres (Pacheco-Torgal and Jalali, 2011) may be potentially
selected for the strengthening of masonry, provided that fabric layouts are
designed to ensure adequate interlocking within a weaker matrix.
In the perspective of using mortar-based composites as strengthening system,
more research is needed to explore the bond performance, for which only few
contributions have been provided to date,(see for instance: Ortlepp et al., 2006;
Faella et al., 2008; Carbone and de Felice, 2009; D’Ambrisi et al., 2013, D’Antino
et al., , 2014; Carozzi et al., 2014).
The present work describes the results of an experimental campaign devoted to
the investigation of the mechanical performance of reinforcement systems
comprising fibrous textiles embedded in inorganic matrices. The research is
currently on-going within the RILEM TC CSM (Composites for sustainable
strengthening of masonry). Three research laboratories, affiliated with the
University Roma Tre (Rome, Italy, UNIRM3), the University of Minho
3
(Guimarães, Portugal, UMINHO), and Tecnalia Research & Innovation (Bilbao,
Spain, TECNALIA) were involved. The experimental programme comprised
Steel Reinforced Grouts (SRG), Carbon Textile Reinforced Mortars (CTRM) and
Basalt Textile Reinforced Mortars (BTRM). Both cement-based and lime-based
mortars have been used as matrices. The three composite systems were
characterised through direct unidirectional tensile tests. Then, the composite-to-
substrate bond performance was investigated using different test setups (single or
double lap scheme) and considering various anchorage lengths, substrates (brick
and stone), and surface preparation techniques.
2. MATERIALS
The properties of the materials used to manufacture the specimens are listed in
Tables 1 and 2. The former includes the type, the compressive strength (f
b
, f
m
) and
the Young’s modulus (E
b
, E
m
) of the substrates and mortar matrices. The tensile
strength of the mortar matrices (f
mt
), derived through three point bending tests, is
also reported. EN 1926 (2006), EN 772-1 (2002), and EN 1015-11 (2007)
standards were followed for the tests on natural stones, bricks, and mortars,
respectively.
Table 1 Mechanical properties of substrates and matrices
Institution
Substrate
Matrix
Name
Acronym
Country
Type
f
b
N/mm
2
E
b
N/mm
2
Type
f
m
N/mm
2
E
m
N/mm
2
f
mt
N/mm
2
Roma Tre
University
UNIRM3
Italy
Brick
55.2
16000
Fibre-reinforced
cement-based
mortar
38.0
15000
7.5
University
of Minho
UMINHO
Portugal
Brick
14.2
9500
Lime-based
mortar
13.0
14000
3.2
Tecnalia
R&I
TECNALIA
Spain
Stone
21.0
5900
Cement-based
polymer-
modified mortar
22.6
15700
2.5
Table 2 contains the type and the properties of the filament/wire (tensile
strength, f
fil
, Young’s modulus, E
fil
, and ultimate strain,
u,fil
), the properties of the
textile (tensile resistance, f
t
, Young’s modulus, E
f
, ultimate strain,
u,
and weight,
W) and its equivalent thickness (t).
Table 2 Textile tensile properties
Type
Institution
Filament
(1)
Textile
f
fil
N/mm
2
E
fil
N/mm
2
u,fil
%
f
t
(2)
N/mm
2
E
f
(2)
N/mm
2
u
(2)
%
W
(1)
g/m
2
t
(1)
mm
Steel
UNIRM3
2474
207000
2.30
3186
192857
1.61
2110
0.256
UMINHO
3200
206000
–
3070
190000
1.62
1800
0.227
TECNALIA
3200
206000
–
3165
170000
2.20
600
0.075
Carbon
UNIRM3
4800
240000
1.80
1914
189361
1.18
168
0.047
Basalt
TECNALIA
2100
89000
3.10
1160
67000
1.91
235
0.035
(1)
Data provided by the manufacturers
(2)
Data determined experimentally
A pozzolan-cement mortar used by UNIRM3, specifically designed for use
with fibre meshes; the 28-days’ compressive strength and the Young’s modulus of
4
the mortar were derived from five 50 mm cubic specimens. Solid clay bricks were
used for bond testing, their properties deriving from compression tests on five 50
mm cubic specimens. Both steel and carbon meshes were used; the former was a
commercial tape-like product consisting of high carbon steel cords
unidirectionally oriented, having 12 cords/inch (2.11 mm spacing between cords,
designated by the producer as ‘medium density’). The carbon fibre mesh was a
balanced textile comprising 4 mm wide carbon fibre rovings arranged in two
orthogonal directions at a net spacing of 6 mm.
Solid clay bricks with dimensions of 200 mm × 100 mm × 50 mm were used at
UMINHO for bond tests. The compressive strength of the bricks was
characterized through compressive tests on six 40 mm cubic specimens, in the
flatwise direction. A commercial ‘medium density’ steel mesh (12 cords/inch)
was used as the reinforcement inserted in a pozzolan lime-based mortar with a
compressive strength much lower than that used at UNIRM3, which contained
cement, and a comparable Young’s modulus. The mortar was characterized by
performing compressive tests on five cylindrical specimens with 50 mm diameter
and 100 mm height at 28 days.
TECNALIA laboratory used basalt fibre and steel wire meshes. Basalt textile
was a balanced bi-directional grid comprising bitumen-coated fibre bundles.
Basalt textile grid spacing was 20 mm × 20 mm, while steel wire fabric had a
density of 4 cords/inch (6.35 mm spacing between cords, designated by the
producer as ‘low density’). A cementitious mortar was used containing less than
4% of organic resins. After a 28-day curing period, five 40 mm ×40 mm ×160 mm
prisms were tested to determine the compressive strength and the Young’s
modulus of the mortar. Stone units were used for bond testing, whose compressive
strength and Young’s modulus were derived by means of five compression tests
on 50 mm cubic specimens.
3. TENSILE TESTS
Despite the ultimate load may be sometimes difficult to be exploited, the
tensile behaviour of EBRs may be important in some structural applications, such
as shear reinforcement of masonry panels, extrados strengthening of arches and
vaults, and confinement of columns. For this reason, direct tensile tests are
required by standard codes (CNR, 2012; ICC, 2013) for the mechanical
characterization of mortar-based reinforcement systems and are expected to
become a fundamental step of product qualification process. Besides the
maximum attainable stress, tensile tests also provide the Young’s modulus of the
composite, which is a key property for the reinforcement design, and the matrix-
to-textile bond properties, which significantly affect cracking, thus influencing
adhesion to the substrate and durability.
An overview of the tensile tests is shown in Table 3, in which the size of the
specimens and the number of textile layers embedded in the mortar are also listed.
UNIRM3 performed tests on CTRM and SRG specimens having an overall cross
section of 40 mm × 7 mm and a length of 800 mm and 700 mm, respectively.
Tests were carried out after 28 days of curing by means of a universal MTS
testing machine, equipped with a 500 kN hydraulic actuator, under displacement
control at 0.005 mm/s rate (machine compliance < 0.05%). The applied load was
5
measured by a load cell integrated in the testing machine, while the strains were
recorded through resistive strain gauges (having length of 10 mm and 0.07%
precision), positioned vertically, parallel to each other, and applied directly on the
fibres. For SRG specimens, in order avoid the detachment of the strain gauges
from the steel cords, a small portion of the textile was impregnated with
polymeric resin. This was not necessary for CTRM specimens, as the strain gauge
and the carbon yarn had approximately the same width (4 mm). In multi-layer
CTRM specimens, strain gauges have been applied to the intermediate layer in
order to record a local measure of the strain in the reinforcement. It should
however be considered that the strain distribution among layers may display slight
differences.
The SRG specimens tested at UMINHO had a 3 mm mortar cover on each side
of the steel cords resulting in a total thickness of 6 mm. Monotonic tests were
performed after 28 days of curing using a universal testing machine with a
maximum load capacity of 200 kN, under displacement control at 0.033 mm/s rate
(machine compliance < 0.2%). The applied load was recorded by a load cell
integrated in the testing machine, while deformation was monitored by a clip
gauge placed mid-height on the specimens. Finally, TECNALIA manufactured
specimens with a 100 mm × 10 mm cross section area and 600 mm length. The
ends of each specimen were reinforced with two additional layers of textile
(extending 200 mm from end to mid-height), in order to promote the failure of the
specimen in its middle third portion. Tests were performed on a 100 kN hydraulic
testing machine, under displacement control at 0.008 mm/s rate (machine
compliance < 0.3%) and load measurement precision better than 0.3%. The
deformation was measured by two LVDTs. According to current strengthening
design practice, multi-layer reinforcement systems were tested for carbon and
basalt textiles, as shown in Table 3.
Table 3 Overview of the direct tensile tests on strengthening systems
Institution
Basalt
Carbon
Steel
Specimen
dimension
[mm
3
]
2 layers
3 layers
3 layers
4
cords/inch
12
cords/inch
UNIRM3
4
40×7×800
4
40×7×700
UMINHO
4
50×6×450
TECNALIA
7
7
7
100×10×600
Based on the specific properties of the specimens tested and on available
laboratory facilities, different solutions were developed to ensure adequate
clamping of the specimen, which is necessary to guarantee a uniform load transfer
and avoid stress concentration in the gripping area (Fig. 1). UNIRM3 and
UMINHO chose to leave each end of the specimen free of mortar and to apply
aluminium tabs by means of a structural glue. Then, the tabs were clamped within
the wedges of the testing machine. TECNALIA gripped the ends of the TRM
specimens by means of a mechanical device made out of two steel plates having
rough surfaces (knurl).
