Accepted Manuscript

A comparison between the use of FRP, FRCM and HPM for concrete confinement

Jacopo Donnini, Simone Spagnuolo, Valeria Corinaldesi

PII: S1359-8368(18)31082-5
DOI: https://doi.org/10.1016/j.compositesb.2018.12.111
Reference: JCOMB 6440

To appear in: Composites Part B

Received Date: 10 April 2018

Revised Date: 27 November 2018
Accepted Date: 28 December 2018

Please cite this article as: Donnini J, Spagnuolo S, Corinaldesi V, A comparison between the use of
FRP, FRCM and HPM for concrete confinement, Composites Part B (2019), doi: https://doi.org/10.1016/
j.compositesb.2018.12.111.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
 ACCEPTED MANUSCRIPT

1 A COMPARISON BETWEEN THE USE OF FRP, FRCM AND HPM

2 FOR CONCRETE CONFINEMENT

PT
4 Jacopo Donnini1, Simone Spagnuolo2, Valeria Corinaldesi3

RI
1
6 SIMAU Department, Marche Polytechnic University, Ancona, Italy, j.donnini@univpm.it
2

SC
7 Civil Engineering and Computer Science Engineering Department, University of Rome Tor
8 Vergata, Rome, Italy, spagnuolo@ing.uniroma2.it
3
9 SIMAU Department, Marche Polytechnic University, Ancona, Italy, v.corinaldesi@univpm.it
10
U
AN
11 ABSTRACT

12 The use of new methods to strengthen and rehabilitate existing concrete and masonry structures
M

13 is one of the challenges that the engineering community is facing in recent years. In this field,
14 composite materials are acquiring more and more success, due to lower invasiveness and ease of
15 application if compared to more traditional systems (e.g. steel plates or reinforced concrete
D

16 jacketing).
17 This work, based on experimental investigations, aims to propose a comparison between three
TE

18 different methods as possible strengthening solutions for existing concrete elements. Twenty
19 compression tests were conducted on reduced scale concrete columns, realized by using a low
20 performance concrete, in order to reproduce the poor mechanical properties of most existing
21 structures. Two of them were left unconfined, while the other ones were reinforced by using
EP

22 Fiber Reinforced Polymer (FRP), Fiber Reinforced Cementitious Matrix (FRCM) or High
23 Performance Mortar (HPM) systems. The effectiveness of the different strengthening techniques
24 and the main differences in terms of structural response were investigated. Experimental results
C

25 were then compared with predictions deriving from guidelines and theoretical models from the
26 literature.
AC

27
28 Keywords:
29 FRCM, FRP, HPM, concrete, confinement, fiber, strengthening.
30
31 1. Introduction
32 The possibility of reinforcing concrete elements by externally applying Fiber Reinforced
33 Polymer (FRP) systems has become a well-established solution within the construction industry.
34 Experimental evidences have shown that FRP systems increase compressive strength, as well as
35 deformation capacity of concrete columns under vertical and lateral loads [1-8]. In many cases

1
 ACCEPTED MANUSCRIPT

1 the effectiveness of these systems has also been validated by natural events, such as the latest
2 earthquakes that occurred in central Italy in 2016 and 2017. The possibility of using these
3 systems to reinforce existing structures was translated into design guidelines by ACI in the USA
4 [9], by CNR in Italy [10], by CSA in Canada [11].
5 FRPs combine several advantages over more traditional reinforcement systems, such as high
6 strength-to-weight ratio, relative ease and speed of application, cost effectiveness. However, the

PT
7 presence of the organic resin (usually epoxy) results in some drawbacks and limitations, such as
8 poor fire and high temperature resistance, inapplicability on wet surfaces, irreversibility, low
9 compatibility with the substrate and low breathability.

RI
10 The attempt to overcome these limits has led to the development of new composite materials
11 consisting of fibers, in the form of fabric meshes or grids, coupled with inorganic matrices,
12 designed as Fiber Reinforced Cementitious Matrix (FRCM) systems. The inorganic matrix, that

SC
13 is not necessarily cement based, but can also consists of lime-based or geopolymer mortars,
14 ensures higher compatibility with the masonry or concrete substrates, guarantees better
15 performances when exposed to high temperature or in case of fire, it can be applied on wet

U
16 surfaces and is safer for operators [12-19].
17 FRCM are preferred to FRP systems when operating in high temperature environments or when
AN
18 the substrate of the element to be reinforced is in a high humidity environment.
19 FRCM systems proved to be very effective as external reinforcement for masonry or concrete
20 elements, even though the adherence developed at the interface between fabric and mortar is not
21 as high as in the case of organic based systems (FRPs) [20-22]. Due to mortar viscosity and
M

22 granularity, in fact, the penetration of the latter within the bundle of fibers (yarn) is not
23 guaranteed. Although this can be considered a characteristic of FRCM systems, which involve
24 the use of dry fibers, various coatings have been applied on the fibers surface or adhesion
D

25 promoters with the aim to enhance bond at the interface between fibers and matrices [23-25].
26 Although the research community has put considerable efforts on the study of the mechanical
TE

27 characterization of FRCM systems, both with dry and coated fibers, and their applicability to
28 reinforce concrete or masonry structures, there is still a lack of regulations and standards that
29 allows to design with these materials. ACI Committee 549 was the first to address this issue by
30 publishing a guide to design FRCM reinforcement systems [26] together with acceptance criteria
EP

31 and test methods by AC434.13 [27]. However, experimental studies on the behavior of concrete
32 elements reinforced with FRCM systems is still limited.
33 The use of composite systems with inorganic matrices to confine concrete columns was studied
C

34 by Triantafillou et al. [28] in 2006. In this study 14 concrete cylinders reinforced with FRP and
35 Textile Reinforced Mortar (TRM) were tested in compression. They used carbon fiber textiles in
AC

36 2 or 3 layers with a cement-based mortar of compressive strength in the range from 15.24 to
37 21.18 MPa. The aim of the research was to determine the influence of the matrix and layer’s
38 number on the axial strength. Experimental results showed a great enhancement in strength and
39 ductility, with a hardening behavior of stress-strain curves.
40 Bournas et al. [19] carried out some tests on reinforced concrete columns strengthened both with
41 TRM and FRP jackets. Results showed that TRM jackets are slightly less effective in terms of
42 increasing strength and deformation capacity by approximately 10%. An experimental study on
43 FRCM strengthening systems was carried out by Ombres [29], to analyze the performances of
44 plain concrete elements wrapped with PBO fiber meshes embedded into an inorganic matrix.
45 The fibers reinforcement ratio, fibers orientation and compressive concrete strength were the

2
 ACCEPTED MANUSCRIPT

1 parameters investigated. The PBO-FRCM system showed to be effective in confining concrete

2 by significantly increase both the peak strength and axial strain. The observed failure mode was
3 due to a loss of compatibility and separation at the fabric to matrix interface.
4 De Caso et al. [16] carried out an experimental research on concrete cylinders confined with a
5 fiber reinforced composite system made of glass fiber sheets and hydraulic cement-based matrix.
6 Despite the poor fiber impregnation, the selected system allowed to obtain a substantial increase

PT
7 in axial strength and deformability with respect to unconfined cylinders.
8 Trapko [30] in 2012 focused on the influence of the temperature on concrete columns confined
9 with PBO fabrics coupled with an inorganic matrix. The experimental activity proved that the

RI
10 initial heat treatment did not change significantly the final experimental results.
11 Another possibility to reinforce existing concrete elements is the use of High Performance
12 Mortar (HPM) jackets. HPM jackets are used when the objective is to significantly increase the

SC
13 carrying capacity of a structural element or when the element is structurally deficient due to
14 damage, which may be caused by seismic events, from design or construction faults, or due to
15 material degradation.

U
16 If compared to normal reinforced concrete jackets or steel jacketing, the use of HPM results in
17 many advantages: reduction in jacket size, ease of execution, reduction or elimination of need for
AN
18 additional steel reinforcement and cost effectiveness [31-34].
19 Meda et al. [32] showed that by applying a high-performance mortar jacket to existing concrete
20 columns with corroded rebars, it is possible to increase the bearing capacity of the columns,
21 reaching a maximum strength greater than the one of the undamaged elements.
M

22 The performance of HPM, as external reinforcement, strongly depends on the mortar mechanical
23 properties and on the strength developed at the interface between the mortar and the old
24 concrete. HPM can achieve compressive strength greater than 100 MPa and flexural strength
D

25 higher than 30 MPa. Mechanical characterization of HPM and the improvement of the
26 performances of cementitious matrices through the addition of steel, glass or carbon fibers have
TE

27 been studied by Corinaldesi and Donnini [35,36]. Moreover, the use of Calcium Oxide showed
28 to further improve the mechanical properties of fiber reinforced HPM [37].
29
30 In this study, the use of FRP, FRCM and HPM jackets to reinforce scaled concrete columns has
EP

31 been investigated. In particular, the possibility of combining different fibers and matrices, the
32 effect of using epoxy resin or inorganic mortars, coupled with the same fabric reinforcement on
33 the compressive behavior of concrete columns are analyzed and discussed. In addition, a
C

34 comparison of experimental results with analytical predictions deriving from guidelines or

35 theoretical models has been presented and discussed. Finally, a preliminary study on the use of
AC

36 different HPMs as external jackets and their effect on the compression behavior of the confined
37 columns has been carried out.
38
39 2. Materials and methods
40 The experimental investigation included 20 cylindrical concrete cylinders having a diameter of
41 140 mm and height of 460 mm. Two specimens were left unreinforced, while the others were
42 strengthened by externally applying three different composite materials (FRP, FRCM and HPM).
43 The application of the same fabric reinforcement, made of carbon or PBO fibers, coupled with
44 organic or inorganic matrices, was investigated and results were compared with the aim to

3
 ACCEPTED MANUSCRIPT

1 quantify their effectiveness. The use of HPM applied as external jacketing with a thickness of 30
2 mm was also investigated. A test matrix summarizing all types of applied reinforcements is
3 reported in Table 1.
4
5 Table 1 - Test Matrix

PT
Element Reinforcement Label
Ref_1
Unreinforced

RI
Ref_2
Epoxy resin + E_CS_1
Carbon sheet E_CS_2

SC
Epoxy resin + E_CF_1
Carbon fabric E_CF_2
Epoxy resin + E_PBO_1
PBO fabric E_PBO_2

U
Concrete
Mortar 15 + M15_CF_1
cylinder
Carbon fabric M15_CF_2
AN
(Height: 460 Mortar 45 + PBO M45_PBO_1
mm, fabric
Diameter: 140
M45_PBO_2
mm) Mortar 45 + M45_CF_1
M

Carbon fabric M45_CF_2

HPM + Fly Ash + HPM _FA_G_1
Glass fibers HPM _FA_G_2
D

HPM + Fly Ash + HPM _FA_S_1

Steel fibers HPM _FA_S_2
TE

HPM + Fly Ash + HPM _SF_S_E_1

Glass fibers +
CaO HPM _SF_S_E_2
6
EP

7
8 2.1 Material properties
9 A concrete with poor mechanical properties (Table 2) was used to cast the cylinders, in order to
C

10 reproduce the real conditions that can be found in most existing concrete structures in need for
11 reinforcement.
AC

12 Two FRCM mortars with different strength classes were used, denoted as M15 and M45, while a
13 two-component epoxy resin was used for FRP systems. Mechanical properties of the inorganic
14 matrices were determined on specimens 40x40x160 mm, after 28 days of curing at RH=50 ± 5%
15 and T=20 ± 2°C, according to UNI EN 1015-11:2007 [38]. Results are reported in Table 2.
16 The use of different fabric reinforcements was investigated. In the case of FRCM, carbon or
17 PBO fabrics were used. The PBO fabric is formed by yarns with 10 mm and 20 mm spacing in
18 the two orthogonal directions, nominal equivalent thickness of 0.046 mm in the longitudinal
19 direction and 0.0224 mm in the transversal direction. The carbon fabric has nominal thickness of
20 0.048 mm in the two orthogonal directions and spacing between yarns of 20 mm. Mechanical
21 properties of the fabrics are reported in Table 3. The same fabrics were impregnated and bonded

4
 ACCEPTED MANUSCRIPT

1 to the concrete surface with an epoxy resin (FRP) or with a cement-based matrix (FRCM). In the
2 case of FRP, a unidirectional carbon sheet (300 g/m2) coupled with epoxy was also applied.
3
4 Table 2 – Mechanical properties of concrete and matrices
Material Compres Flexural Elastic

PT
sive strength modulus
strength
(MPa) (MPa) (GPa)
Concrete 14 1.3 11.4

RI
Mortar M15 17 3.6 12.5
Mortar M45 50 6.2 34.5
HPM _FA_G 91 13.1 45.3
HPM _FA_S 128 34.1 49.7

SC
HPM _SF_S_E 144 27.1 50.1
Epoxy resin 59 48.0 1.9
5
6 Table 3 – Mechanical and geometric properties of the fabrics

U
Tensile Elastic Ultimate Nominal Fabric
AN
Material Orientation/type strength modulus strain thickness weight
(kN/m) (GPa) (%) (mm) (g/m2)
Carbon fabric (CF) Bi-directional 240 235 1.5 0.048 172
M

Carbon sheet (CS) Unidirectional 800 240 1.8 0.164 300

PBO fabric Unidirectional 264 270 2.5 0.046 146
7
8 Three HPM with different compositions and mechanical properties were prepared, using a
D

9 commercial Portland-limestone blended cement type CEM I 52.5 R, according to the European
10 Standards EN-197/1 [39] and, as aggregate, a quartz sand with particle size up to 1.0 mm. A
TE

11 water-reducing admixture constituted of a carboxylic acrylic ester polymer was used to maintain
12 a low water to cement ratio. A low-calcium fly ash (ASTM C 618 Class F) produced by a
13 thermal generating station was used. The Blaine fineness of fly ash is 480 m2/kg and its relative
EP

14 density (specific gravity) is 2.25. Finally, short fibers made of brass-coated steel or glass, with
15 nominal length of 13 and 12 mm respectively, were used as randomly dispersed reinforcement
16 for the three HPM mixtures. The mixture proportions are reported in Table 4.
17
C

18 Table 4 – High Performance Mortars mixtures (kg/m3)

AC

CEM I Brassed Glass

Mortar Water Sand Fly ash Superpl. CaO SRA
52.5R steel fibers fiber
HPM _FA_G 960 240 960 250 96 - 66 - -
HPM _FA_S 960 240 960 250 96 192 - - -
HPM _SF_S_E 960 240 960 250 96 192 - 35 4

19
20 Due to brittle behavior of the matrix and possible slippage of the fabric within the mortar, it is
21 important not to consider the mechanical properties of the fabric, but rather those of the entire

5
 ACCEPTED MANUSCRIPT

1 composite material. Therefore, in this study, the FRCM mechanical properties have been
2 evaluated following the AC434.13 Annex A instructions.
3 Tensile tests on FRCM specimens were carried out in order to determine the mechanical
4 properties of the composite. Specimens preparation and test setup have been described in
5 previous works by the authors [27]. Ultimate tensile strength σu, elastic modulus (or stiffness
6 modulus) in the cracked phase E2 and ultimate strain εu have been reported in Table 5 and used

PT
7 to predict the confinement effectiveness of the different FRCM systems.
8

RI
U SC
AN
9
M

10 Figure 1 – Carbon fabric, PBO fabric and carbon sheet

11
D
TE
EP

12
13 Figure 2 – Glass and brassed steel fibers
C

14
15 Table 5 – FRCM tensile properties [According to AC434.13-Annex A]
AC

Ultimate Elastic Ultimate

tensile modulus strain
FRCM
strength
system
σu E2 εu
(MPa) (GPa) (%)
M15_CF 846 79 0.93
M45_CF 975 80 1.03
M45_PBO 1542 132 1.80
16
17 2.2 Specimens preparation and test setup

6
 ACCEPTED MANUSCRIPT

1 All specimens were manufactured by the same concrete batch, cast into a cylindrical plastic
2 formwork, demoulded after 7 days and cured at laboratory conditions (70% RH, 20 °C) for 28
3 days. The external surface of the specimens was smoothed with a sander and cleaned before the
4 application of the reinforcement.
5 FRP reinforcements were applied by qualified personnel from the manufacturer. First, the
6 concrete surface was treated with an epoxy filler to eliminate defects and a primer was applied in

PT
7 order to improve the adhesion of the next resin layer. Then, carbon or PBO fabrics were applied
8 by using a two-component epoxy resin. Fabric overlap was equal to 100 mm in each confined
9 specimen.

RI
10 FRCM systems were applied as indicated by the manufacturer, manually spreading a first layer
11 of mortar with a thickness of 4 mm and applying the PBO or carbon fabric, slightly pressing it
12 into the mortar. Fabric overlap was equal to 150 mm in each confined specimen. The second

SC
13 mortar layer was applied with the same thickness.
14 HPM were cast within a cylindrical plastic formwork, with an internal diameter equal to 200
15 mm, to ensure a reinforcement thickness of 30 mm. The high workability and self-compacting

U
16 properties of the HPM mixtures allowed to easily cast them inside the formworks, without the
17 need for vibration.
AN
18
M
D
TE
EP

19
20 Figure 3 - Compression test setup
C

21
22 After strengthening, all specimens were cured at laboratory conditions (70% RH, 20 °C) for 28
AC

23 days. Top and bottom surfaces of all columns were smoothed to assure parallel surfaces and
24 uniform load distribution. All specimens were tested under uniaxial compression, using a 3000
25 kN compression machine, through monotonically applied loading at a rate of 0.3 mm/min.
26 Applied loads were measured from a load cell while axial displacements were measured by using
27 external linear variable differential transducers (LVDT) mounted on two opposite sides of the
28 specimen (Figure 3). The average deformations determined through the relative displacements
29 measured by two LVDTs have been reported in Table 6. The peak strength of the unconfined
30 specimen fc0 was defined as the maximum load reached during the test, divided by the net area of
31 the concrete cylinder. In the case of reinforced specimens, the ultimate strength fcc was calculated

7
 ACCEPTED MANUSCRIPT

1 considering the maximum load reached in the post-cracking phase, divided by the same net area
2 of the cylinder.
3
4 3. Experimental results
5 Results of uniaxial compression tests for unreinforced (Ref) and reinforced concrete specimens

PT
6 are summarized in Table 6 and Table 7, while stress-strain curves are reported in Figure 4. The
7 maximum load (Fmax), peak strength (fcc) and strain (εcc) of reinforced specimens have been
8 reported. The average peak strength of the unconfined specimen fc0 was equal to 11.40 MPa

RI
9 while the ultimate strain εc0 was 0.359 %. In Table 6, the values of the ratios between the peak
10 strength of the confined specimens fcc and that of the reference specimen fc0, and between the
11 ultimate axial strain of the confined (εcc) and unconfined (εc0) specimens are also reported.

SC
12 Results of tests are presented and discussed in terms of stress-strain response, ductility and
13 failure modes.
14 Table 6 – Results of compression tests on concrete specimens

U
Specimen Fmax fc0 fcc εc0 εcc fcc/fc0 εcc/εc0
(kN) (MPa) (MPa) (%) (%)
AN
Ref_1 170 11.05 - 0.346 - - -
Ref_2 181 11.76 - 0.372 - - -
M

E_CS_1 497 11.40 32.31 0.359 1.778 2.92 5.13

E_CS_2 477 11.40 31.01 0.359 1.742 2.72 4.85
D

E_PBO_1 330 11.40 21.45 0.359 1.271 1.94 3.67

E_PBO_2 321 11.40 20.87 0.359 1.259 1.83 3.51
TE

E_CF_1 305 11.40 19.83 0.359 1.297 1.79 3.74

E_CF_2 297 11.40 19.31 0.359 1.107 1.69 3.08

EP

M15_CF_1 205 11.40 13.32 0.359 0.566 1.20 1.63

M15_CF_2 215 11.40 13.98 0.359 0.550 1.23 1.53
M45_PBO_1 279 11.40 18.14 0.359 1.117 1.64 3.23
C

M45_PBO_2 265 11.40 17.27 0.359 1.062 1.51 2.96

AC

M45_CF_1 213 11.40 13.85 0.359 0.488 1.25 1.41

M45_CF_2 207 11.40 13.46 0.359 0.481 1.18 1.34
15
16 Test results showed that both FRP and FRCM were able to enhance the performance of
17 unconfined cylinders, with a significant gain in strength and ductility.
18 In the case of FRP reinforcements, the stress strain curve is characterized by three different
19 stages. Initially, most of the load is carried out by the concrete and the reinforcement is not
20 activated; the stress strain curves are almost linear (Figure 4) and the slope is similar to that of
21 unconfined specimens. Once the concrete core starts to get damaged, the external reinforcement

8
 ACCEPTED MANUSCRIPT

1 is activated. The slope of the stress-strain curves decreases, due to the lower elastic modulus of
2 cracked concrete, and becomes linear up to failure. The carbon sheet (E_CS, 300 g/m2) provided
3 the best performing solution, by increasing the compressive strength of about 2.7 times.
4 Compressive strength of cylinders reinforced with PBO (E_PBO, 146 g/m2) and carbon (E_CF,
5 172 g/m2) fabrics increased by 1.8 and 1.7 times respectively. Specimens confined with FRP
6 systems failed abruptly when the fibers reached the ultimate strength (Figure 5).

PT
7 Cylinders reinforced with FRCM systems showed a different behavior under compression. In
8 this case the stress strain curve can be divided into 4 stages. After the failure of the inner
9 concrete core, the specimen maintains high stiffness, due to the presence of the FRCM inorganic

RI
10 matrix. At the end of this stage, the inorganic matrix starts to crack and the fabric reinforcement
11 is activated. The third phase is characterized by the slippage of the fabric within the inorganic
12 matrix. This phenomenon has been clearly highlighted in other studies concerning the

SC
13 mechanical characterization of FRCM systems [40-42]. The slope of the curve in this stage
14 depends on the friction developed at the interface between fabric and inorganic matrix. The
15 failure of concrete cylinders confined with FRCM systems was more ductile than that of FRP

U
16 confined specimens, due to the formation of multiple cracks in the matrix and to the slow
17 slippage of the fabric within the mortar (Figure 5). FRCM systems increased the compressive
AN
18 strength of the concrete cylinders of 1.2 up to 1.6 times the initial strength. Nevertheless, best
19 performances were obtained by using a unidirectional PBO fabric coupled with an inorganic
20 matrix specifically designed for concrete substrates. This system was able to significantly
21 increase the ductility of the concrete element. Ultimate strain was 3 times higher than that of the
M

22 unreinforced specimen. Failure mode was always by fibers slippage and breakage, after the
23 formation of several longitudinal cracks on the mortar surface.
24 FRCM confinement systems with dry carbon fabrics were less effective than those with PBO,
D

25 due to lower mechanical properties of carbon fibers and lower adhesion between fibers and
26 matrix. The confinement ratio (fcc/fc0) was about 1.2. No substantial differences in the peak load
TE

27 were noticed by changing the mortar strength class (passing from M15 to M45). This
28 phenomenon was also observed by other authors [43] and could be expected by looking at the
29 tensile properties of the two FRCM systems (Table 5).
30
EP

31
32
C
AC

9
 ACCEPTED MANUSCRIPT

PT
RI
U SC
1
2 Figure 4 – Stress-strain curves of uniaxial compression tests on concrete cylinders
AN
3
4
M
D
TE
C EP
AC

10
 ACCEPTED MANUSCRIPT

PT
RI
U SC
1
AN
2 Figure 5 – Concrete cylinders with different reinforcement systems at failure
3
4 3.1 FRP vs FRCM
M

5 Although FRCM systems offer considerable advantages if compared to FRP, as previously

6 reported, it has been recognized that FRCM are slightly less effective than FRP, due to the
7 difficulty of the inorganic matrix to penetrate within the filaments of the reinforcement, and
D

8 therefore to the lower adhesion developed at the interface between matrix and fibers [19,25].
9 The different behavior of concrete cylinders confined with FRP or FRCM systems can be easily
TE

10 observed in the stress-strain curves of Figure 6. The first part of the stress-strain curves is almost
11 the same for all specimens.
12 In FRP systems, the slippage of the fabric (both carbon and PBO) is prevented by the resin and
EP

13 once the reinforcement has been activated, the curve is almost linear up to failure.
14 Compression behavior of cylinders reinforced with FRCM systems is different: the slope of the
15 curve changes when the perfect adhesion between fabric and matrix is no longer guaranteed. The
16 inorganic mortar starts to crack and the the fabric slippage takes place. In this phase the FRCM
C

17 effectiveness depends only on the friction between fabric and mortar. The molecular structure of
18 PBO is capable of establishing chemical bonds with the inorganic mortar, which helps to
AC

19 improve the overall FRCM performance. The smooth surface of carbon fibers, the absence of
20 strong chemical bonds at the interface fibers-mortar and the telescopic behavior of dry carbon
21 yarns [20,22,23] do not allow the composite material to fully develop its effectiveness. This
22 phenomenon can be observed by looking at the different slope of the stress-strain curves once the
23 reinforcement has been activated. In the case of PBO-FRCM reinforcement, a strain-hardening
24 behavior is observed during the fabric slippage, while, looking at the Carbon-FRCM systems, the
25 stress is almost constant and a plateau is exhibited before failure.
26 The reduction of the confinement ratio (fcc/fc0) passing from organic to inorganic matrices was
27 equal to 17% when using PBO fabrics and 30% in the case of dry carbon fabrics.

11
 ACCEPTED MANUSCRIPT

1 Regarding the effect of the reinforcement on the strain ductility, the higher ductility ratio was
2 attained by the E_PBO confined specimens, with an increase of the ultimate strain of about 3.5
3 times while the M45_CF increased the ultimate strain of 1.38 times.
4

PT
RI
U SC
AN
5
M

6 Figure 6 – Stress-strain behavior of concrete columns reinforced with FRP or FRCM systems
7 subjected to uniaxial compression
D

8
9 3.2 HPM
TE

10 Experimental results of compression tests on concrete cylinders confined with HPM are reported
11 in Table 7. The experiments highlighted a significant strength improvement for cylinders
12 confined with HPM. All mortars proved to be very effective in increasing the ultimate
EP

13 compressive strength and axial stiffness of the unstrengthened specimen. However, cylinders
14 reinforced with HPM showed a different behavior with respect to FRP and FRCM confinement
15 systems. The compression load is carried out only by the external HPM, which is more resistant
16 and stiffer than the internal concrete. The stress strain curve is almost linear up to reach the
C

17 failure of the HPM (Figure 7).

18 Failure in HPM strengthened cylinders under uniaxial load was due to the compressive failure of
AC

19 the HPM layer (Figure 8). As the load increases, vertical cracks appear on the external
20 reinforcement layer. However, crack initiation did not cause a load reduction, nor a loss of
21 stiffness, due to the bridging effect of short glass and steel fibers. Once the maximum load has
22 been reached, the cracks continue to increase, and a softening branch can be observed in the
23 stress-strain curves (Figure 7).
24 Cylinders confined with HPM_FA_G failed due to the complete breakage of the external mortar
25 layer, which detached from the internal concrete element. Specimens reinforced with HPM and
26 steel fibers (HPM_SF_S_E, HPM_FA_S) showed a higher peak load and a different failure

12
 ACCEPTED MANUSCRIPT

1 mode. In this case the HPM cracked but no debonding between HPM jacketing layer and inner
2 concrete was detected.
3
4 Table 7 – Results of compression tests on concrete columns confined with HPM
Specimen Fmax fcc εcc fcc/fc0 εcc/εc0

PT
(kN) (MPa) (%)
HPM _FA_G_1 1200 38.23 0.451 3.45 1.30
HPM _FA_G_2 1144 36.43 0.417 3.20 1.16

RI
HPM_FA_S_1 1640 52.23 0.777 4.71 2.25
HPM _FA_S_2 1601 50.99 0.730 4.47 2.03

SC
HPM _SF_S_E_1 1430 45.54 0.584 4.11 1.69
HPM _SF_S_E_2 1356 43.18 0.504 3.79 1.40

U
5 AN
M
D
TE
EP

6
C

7 Figure 7 – Stress-strain behavior of concrete columns reinforced with HPM subject to

8 compression
AC

9
10 The experimental stress-strain curves of compression tests on cylinders strengthened with HPM
11 and steel fibers (HPM_FA_S, HPM_SF_S_E), showed a remarkable softening branch and a
12 significant deformation capability after the formation of multiple cracks, up to complete failure
13 of the specimen. Cylinders confined with HPM with short glass fibers (HPM_FA_G) showed a
14 more brittle failure. This phenomenon can be attributed to the higher effectiveness in the
15 bridging action provided by steel fibers, when the HPM is subject to combined compression-
16 tensile biaxial stresses. The use of an expansive agent within the HPM worsened the adhesion
17 with the existing substrate, leading to a detachment at the interface between HPM and concrete
18 and so the breakage occurred for loads lower than those expected (Figure 9, HPM_SF_S_E).

13
 ACCEPTED MANUSCRIPT

1 On the contrary, better adhesion and no debonding were observed in cylinders confined with
2 HPM without the addition of expansive agents (Figure 9, HPM_FA_S).
3

PT
RI
U SC
AN
4
5 Figure 8– Failure modes of concrete columns reinforced with HPM
M

6
D
TE
EP

7
C

8 Figure 9 – HPM-concrete interface after failure

9
AC

10
11 4. Prediction of the reinforcement effectiveness
12 4.1 Confinement with FRP and FRCM
13 Different models have been formulated so far in the scientific literature to predict the
14 confinement effects provided by FRP and FRCM systems. Some of these models have been
15 calibrated based on experiments carried out with FRP reinforcements, such as the ones by
16 Spoelstra and Monti [44], Toutanji [45], Shehata [1]. In most cases, due to lack of analytical
17 models calibrated on FRCM systems, the same formulas used for FRPs have been used with
18 FRCM reinforcement. However, the different nature of the matrix (organic or inorganic)

14
 ACCEPTED MANUSCRIPT

1 produces different micro-mechanical effects, different behavior at the interface between matrix
2 and fabric and different failure modes. In FRP composites, the resin does not crack and the
3 slippage of the fibers within the matrix is avoided. In FRCM systems the cement-based matrix
4 cracks at low load levels and failure is often governed by fabric slippage within the matrix. For
5 this reason, the models proposed to predict the response of concrete elements confined with FRP
6 systems are not very accurate and reliable in the case of FRCM reinforcements. Some

PT
7 researchers proposed new analytical models to consider the different behavior of inorganic based
8 systems, such as the models proposed by Triantafillou [28], Cascardi [46] and Ombres [47]. The
9 American guideline ACI549.4 R13 also provides analytical formulas to predict the compressive

RI
10 behavior of concrete columns confined with FRCM systems. In this study, the models listed in
11 Table 8 were applied to predict the peak strength fcc and the ultimate axial strain εcc of the
12 confined columns. It is worth mentioning that mechanical properties of the FRCM systems used

SC
13 in this study have been determined according to AC434.13 [27].
14
15 Table 8 – Analytical expressions for the maximum strength of confined concrete fcc

U
16 and maximum axial deformation εcc AN
Analytical expressions of fcc and εcc Model
.
= . + ∙
,
∙ = 1 + 5 ∙ −1 ∙ [Spoelstra&Monti 1999]

. '
= + . ∙
,
∙ = 1 + (1.9 + 310.57 %) ∙ −1 (∙
M

'
[Toutanji 1999]
.0
'-,. '
=) + ∙ , * = 1 + 632 ∙ ∙ ( ∙
' /
[Shehata 2002]
D

. 2 . 34 -.33
= + . 1 ∙ ∙ = 1+ ∙ ∙
, 6,788
5
[Triantafillou 2006]
TE

.0
'-,.
= + :. 2 ∙ . 2
= 1+ (
9 ,
'
[Cascardi 2017]

/
= + . : ∙ ∙ = 0.0035 + 0.015 ∙ <
, 6,788
[CNR-DT200 R1/2013]
EP

'-,. .30
= 9 + . ∙ κ= , = 1.5 + 12>?
%
'
[ACI 549.4 R13]
C

17
18 The value f1,eff was determined:
AC

19
'-,. = >. @ / %
20
21 with @ = 4B /C.
22
23 In the analysis, the value ke=0.5 was used, as indicated in [10]. Predictions of each considered
24 model and comparisons with experimental results in terms of peak strength (fccth/fccexp) and axial
25 strain (εcc,th/εcc,exp) are reported in Table 9 and 10.
26
27

15
 ACCEPTED MANUSCRIPT

1 Table 9 – Confined concrete strength (fcc): comparison of experimental results (fccexp) with
2 theoretical models and design guidelines (fccth)
Specimen fcc/fc0 fccth/fccexp
Exp Spoelstra Toutanji Shehata Triantafillou Cascardi DT200 ACI549
& Monti (1999) (2002) (2006) (2017) (2013) (2013)

PT
(1999)
E_CS 2.82 1.09 1.50 0.99 0.95 - 1.22 -
E_PBO 1.89 1.20 1.49 1.03 0.91 - 1.38 -

RI
E_CF 1.74 0.85 1.04 0.76 0.68 - 1.03 -
M15_CF 1.22 0.69 1.04 0.91 0.86 1.16 1.11 0.95

SC
M45_PBO 1.56 0.84 1.06 0.82 0.74 1.13 1.09 0.92
M45_CF 1.22 0.72 1.06 0.91 0.87 1.18 1.13 0.96

U
3
4
AN
5 Table 10 – Confined concrete strain (εcc): comparison of experimental results (εcc,exp) with
6 theoretical models and design guidelines (εcc,th)
Specimen εcc/εc0 εcc,th/εcc,exp
M

Exp Spoelstra Toutanji Shehata Triantafillou Cascardi DT200 ACI549

& Monti (1999) (2002) (2006) (2017) (2013) (2013)
(1999)
D

E_CS 4.99 2.02 2.93 4.46 0.31 - 1.00 -

E_PBO 3.59 1.37 2.37 4.80 0.39 - 1.07 -
TE

E_CF 3.41 1.50 1.88 2.97 0.35 - 0.82 -

M15_CF 1.58 1.48 1.45 5.67 0.82 0.90 1.37 1.19
EP

M45_PBO 3.10 1.24 1.70 3.40 0.39 1.03 0.83 0.64

M45_CF 1.38 1.29 1.30 5.09 0.72 0.79 1.23 1.05
7
C

8 The analysis of results highlighted that, with reference to the peak strength, the models proposed
9 by Spoelstra, Shehata and Triantafillou are very close to experimental results of columns
AC

10 confined with FRP made of carbon sheet (E_CS), while DT200 overestimate the peak strength of
11 about 22%. The same models underestimate the peak strength of columns confined with FRCM
12 systems. The model proposed by Cascardi slightly overestimates the ultimate peak strength (13-
13 18%) while the American guideline AC434 slightly underestimates it (4-8%).
14 DT200 is able to predict with reasonable accuracy the ultimate strength and ultimate strain of
15 confined specimens, providing that the mechanical properties of FRCM are calculated according
16 to AC434.
17
18

16
 ACCEPTED MANUSCRIPT

1 4.2 Confinement with HPM

2 For the preliminary evaluation of the compressive strength of columns confined by HPM, the
3 contribution of the inner concrete column is neglected, thus considering the reinforcement
4 section as a hollow section. The net reinforcement area considered in the calculation is equal to
5 16014 mm2 (HPM with a thickness of 30 mm).

PT
6
7 ' ,DE = )FGHI ∙ ' J,GHI */FDKD
8

RI
9 Theoretical and experimental compressive strength have been compared and results are reported
10 in Table 11.
11

SC
12 Table 11 – Comparison of experimental results with theoretical predictions on concrete columns
13 confined with HPM
Specimen fcc,exp fcc,th fcc,th /fcc,exp

U
HPM _FA_G 37.33 46.40 1.24
HPM_FA_S 51.61 65.25 1.26
AN
HPM _SF_S_E 44.36 73.44 1.65
14
15 The theoretical prediction overestimates the effective resistance of about 25%. This fact is
M

16 probably due to a slight eccentricity of the HPM reinforcement with respect to the vertical axis
17 of the specimen, produced during the casting of the self-compacting HPM. Specimens reinforced
18 with HPM_SF_S_E reached compressive strength lower than the one expected, due to premature
D

19 failure at the interface between reinforcement and concrete column (Figure 9). The presence of
20 the expansive agent within the mixture did not improve the adhesion at the interface HPM-
TE

21 concrete. Also, in this case a certain eccentricity can be noticed by the formation of cracks on a
22 limited portion of the specimen.
23
EP

24
25 5 Conclusions
26
27 The confinement effect of FRP, FRCM and HPM was evaluated based on uniaxial compression
C

28 tests of concrete cylinders. All the composite systems showed to be effective in increasing
29 compressive strength and ductility of the concrete elements, even if the mechanical behavior,
AC

30 final performances and failure modes were different. Experimental results allowed to draw the
31 following conclusions:
32
33 • FRCM systems proved to be less effective than FRP in concrete confinement. The peak
34 strength of cylinders confined with Carbon-FRCM systems was about 30% lower than
35 that of the same fabric coupled with an epoxy resin and 17% lower when using PBO
36 fabrics.
37 • Concrete cylinders confined with FRP failed due to fibers’ breakage, while the failure
38 observed in FRCM confined specimens was due to the slippage of the fabric within the

17
 ACCEPTED MANUSCRIPT

1 inorganic matrix. This phenomenon depends on the different ability of the organic (FRP)
2 and inorganic (FRCM) matrices to impregnate the carbon or PBO fabrics.
3 • HPM jacketing layer of 30 mm thickness showed to significantly increase the
4 compressive strength, stiffness and ductility of the concrete cylinders. The use of short
5 steel fibers enhanced the post peak response in compression loading.
•

PT
6 DT200 and ACI 549 furnished reliable predictions of the response of FRP and FRCM
7 confined concrete elements. In the case of FRCM confined elements, it is fundamental to
8 consider not only the properties of the fabric, but especially those of the whole composite
9 material (consisting of a mortar coupled with a fabric), as suggested by the American

RI
10 guideline AC434.13.
11
12

SC
13 REFERENCES
14
15 [1] Shehata IAEM, Carneiro LAV, Shehata LCD. Strength of short concrete columns

U
16 confined with CFRP sheets. Materials and Structures. 2002; 35: 50-58.
17 [2] Berthet JF, Ferrier E, Hamelin P. Compressive behavior of concrete externally confined
AN
18 by composite jackets. Part A: experimental study. Construction and Building Materials. 2005;
19 19: 223-232.
20 [3] Nanni A, Bradford NM. FRP jacketed concrete under uniaxial compression. Construction
21 and Building Materials. 1995; 9(2): 115-124.
M

22 [4] Ilki A, Peker O, Karamuk E, Demir C, Kumbasar N. FRP retrofit of low and medium
23 strength circular and rectangular reinforced concrete columns. Journal of Materials in Civil
24 Engineering. 2008; 20(2): 169-188.
D

25 [5] Realfonzo R, Napoli A. Concrete confined by FRP systems: confinement efficiency and
26 design strength models. Composites Part B: Engineering. 2011; 42: 736-755.
TE

27 [6] Ozbakkaloglu T, Lim JC, Vincent T. FRP-confined concrete in circular sections: Review
28 and assessment of stress-strain models. Engineering Structures. 2013; 49: 1068-1088.
29 [7] Faella C, Martinelli E, Paciello S, Camorani G, Aiello MA, Micelli F, Nigro E. Masonry
EP

30 columns confined by composite materials: Experimental investigation. Composites Part B:

31 Engineering. 2011; 42: 692-704.
32 [8] Faella C, Martinelli E, Paciello S, Camorani G, Aiello MA, Micelli F, Nigro E. Masonry
33 columns confined by composite materials: Design formulae. Composites Part B: Engineering.
C

34 2011; 42: 705-716.

35 [9] American Concrete Institute (ACI) Committee 440. Guide for the design and construction
AC

36 of externally bonded FRP systems for strengthening concrete structures – ACI440.2R-08.

37 Farmington Hills (MI): ACI; 2008.
38 [10] CNR-DT 200 R1/2013. National Research Council. Guide for the design and construction
39 of externally bonded FRP systems for strengthening existing structures - CNR DT200, Rome,
40 Italy; 2013.
41 [11] Canadian Standards Association (CSA). Canadian highway bridge design code –
42 CAN/CSA-S6-06. Mississauga, Ontario, Canada: CSA; 2006.
43 [12] Trapko T. Confined concrete elements with PBO-FRCM composites. Construction and
44 Building Materials. 2014; 73: 332-338.

18
 ACCEPTED MANUSCRIPT

1 [13] Colajanni P. De Domenico F. Recupero A. Spinella N. Concrete columns confined with

2 fibre reinforced cementitious mortars: Experimental and modelling. Construction and Building
3 Materials. 2014; 52: 375-384.
4 [14] Trapko T. Stress-strain model for FRCM confined concrete elements. Composites: Part
5 B. 2013; 45: 1351-1359.
6 [15] Corinaldesi V. Donnini J. Mazzoni G. Experimental study of adhesion between FRCM

PT
7 and masonry support. Key Engineering Materials. 2015; 624: 189-196.
8 [16] De Caso y Basalo FJ, Matta F, Nanni A. Fiber reinforced cement-based composite system
9 for concrete confinement. Construction and Building Materials. 2012; 32: 55-65.

RI
10 [17] Papanicolaou CG, Triantafillou TC, Karlos K, Papathanasiou M. Textile-reinforced
11 mortar (TRM) versus FRP as strengthening material of URM walls: In-plane cyclic loading.
12 Materials and Structures. 2007; 40(10): 1081-1097.

SC
13 [18] Donnini J, De Caso y Basalo F, Corinaldesi V, Lancioni G, Nanni A. Fabric-reinforced
14 cementitious matrix behavior at high-temperature: Experimental and numerical results.
15 Composites Part B: Engineering. 2017; 108: 108-121.

U
16 [19] Bournas DA, Lontou PV, Papanicolaou CG, Triantafillou TC. Textile-Reinforced Mortar
17 (TRM) versus FRP Confinement in Reinforced Concrete Columns. ACI Structural Journal.
AN
18 2007; 104(6): 740-748.
19 [20] Donnini J, Corinaldesi V, Nanni A. Mechanical properties of FRCM using carbon fabrics
20 with different coating treatments. Composites Part B: Engineering. 2016; 88: 220-228.
21 [21] Naaman AE. Textile reinforced cement composites: competitive status and research
M

22 directions. In: International RILEM conference on material science. 2010; 1: 3-22.

23 [22] Hegger J, Voss S. Investigations on the bearing behaviour and application potential of
24 textile reinforced concrete. Engineering Structures. 2008; 30(7): 2050-2056.
D

25 [23] Banholzer B, Brameshuber W, Jung W. Analytical simulation of pull-out tests – the

26 direct problem. Cement and Concrete Composites. 2005; 27: 93–101.
TE

27 [24] Nadiv R, Peled A, Mechtcherine V, Hempel S, Schroefl C. Micro- and nanoparticle

28 mineral coating for enhanced properties of carbon multifilament yarn cement-based composites.
29 Composites Part B: Engineering. 2017; 111: 179-189.
30 [25] Donnini J, Lancioni G, Corinaldesi V. Failure modes in FRCM systems with dry and pre-
EP

31 impregnated carbon yarns: Experiments and modeling. Composites Part B. 2018; 140: 57-67.
32 [26] ACI 549.4 R-13. American Concrete Institute. Guide to Design and Construction of
33 Externally Bonded Fabric-Reinforced Cementitious Matrix (FRCM) Systems for Repair and
C

34 Strengthening Concrete and Masonry Structures – ACI Committee 549, 2013.

35 [27] AC434-13. Acceptance criteria for masonry and concrete strengthening using fabric-
AC

36 reinforced cementitious matrix (FRCM) composite systems, ICC Evaluation Service, 2013.
37 [28] Triantafillou TC, Papanicolaou CG, Zissimopoulos P, Laourdekis T. Concrete
38 confinement with textile-reinforced mortar jacket. ACI Structural Journal. 2006; 103(1): 28-37.
39 [29] Ombres L. Concrete confinement with a cement based high strength composite material.
40 Composite Structures. 2014; 109: 294-304.
41 [30] Trapko T. The effect of high temperature on the performance of CFRP and FRCM
42 confined concrete elements. Composites Part B: Engineering. 2013; 54(1): 138-145.
43 [31] Rabehi B, Ghernouti Y, Li A and Boumchedda K. Comparative behavior under
44 compression of concrete columns repaired by fiber reinforced polymer (FRP) jacketing and

19
 ACCEPTED MANUSCRIPT

1 ultra-high-performance fiber reinforced concrete (UHPFRC). Journal of Adhesion Science and

2 Technology. 2014; 28 (22-23): 2327-2346.
3 [32] Meda A, Mostosi S, Rinaldi Z and Riva P. Corroded RC columns repair and
4 strengthening with high performance fiber reinforced concrete jacket. Materials and Structures.
5 2015; 49 (5): 1967-1978.
6 [33] Beschi C. Meda A. Riva P. Column and joint Retrofitting with High Performance Fiber

PT
7 Reinforced Concrete Jacketing. Journal of Earthquake Engineering. 2011; 15(7): 989-1014.
8 [34] Wang YC. Lee MG. Ultra-high strength steel fiber reinforced concrete for strengthening
9 RC frames. Journal of Marine Science and Technology. 2007; 15(3): 210-218.

RI
10 [35] Corinaldesi V, Nardinocchi A, Donnini J. The influence of expansive agent on the
11 performance of fibre reinforced cement-based composites. Construction and Building Materials.
12 2015; 91: 171-179.

SC
13 [36] Donnini J, Bellezze T, Corinaldesi V. Mechanical, electrical and self-sensing properties
14 of cementitious mortars containing short carbon fibers. Journal of Building Engineering. 2018;
15 20: 8-14.

U
16 [37] Corinaldesi V, Donnini J, Nardinocchi A. The influence of calcium oxide addition on
17 properties of fiber reinforced cement-based composites. Journal of Building Engineering. 2015;
AN
18 4: 14-20.
19 [38] UNI EN 1015–11, Metodi di prova per malte per opere murarie - Determinazione della
20 resistenza a flessione e a compressione della malta indurita, 2007.
21 [39] EN-197/1, Cement – Part 1: Composition, Specifications and Conformity Criteria for
M

22 Common Cements, 2000.

23 [40] Donnini J, Corinaldesi V. Mechanical characterization of different FRCM systems for
24 structural reinforcement. Construction and Building Materials. 2017; 145: 565-575.
D

25 [41] Carozzi FG, Milani G, Poggi C. Mechanical properties and numerical modeling of Fabric
26 Reinforced Cementitious Matrix (FRCM) systems for strengthening of masonry structures,
TE

27 Composite Structures. 2014; 107: 711–725.

28 [42] Arboleda D, Carozzi FG, Nanni A, Poggi C. Testing Procedures for the Uniaxial Tensile
29 Characterization of Fabric-Reinforced Cementitious Matrix Composites. Journal of Composites
30 for Construction. 2016; 20 (3).
EP

31 [43] Minafò G, La Mendola L. Experimental investigation on the effect of mortar grade on the
32 compressive behaviour of FRCM confined masonry columns. Composites Part B. 2018; 146: 1-
33 12.
C

34 [44] Spoelstra MR, Monti G. FRP-Confined concrete model. Journal of Composites for
35 Construction. 1999; 3(3): 143-150.
AC

36 [45] Toutanji HA. Stress-strain characteristics of concrete columns externally confined with
37 advanced fiber composite sheets. ACI Materials Journal. 1999; 3: 397-404.
38 [46] Cascardi A, Aiello MA, Triantafillou T. Analysis-oriented model for concrete and
39 masonry confined with fiber reinforced mortar. Materials and Structures. 2017; 50: 202.
40 [47] Ombres L, Mazzuca S. Confined concrete elements with cement-based composites:
41 confinement effectiveness and prediction models. Journal of Composites for Construction. 2017;
42 21(3).
43

20