Experimental investigation on out-of-plane behavior of masonry panels strengthened with CFRP sheets

Composites Part B 150 (2018) 14–26

Contents lists available at ScienceDirect

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

Experimental investigation on out-of-plane behavior of masonry panels strengthened with CFRP sheets

T

Mario Fagone∗, Giovanna Ranocchiai Department of Civil and Environmental Engineering, University of Florence, via di S. Marta 3, 50139 Florence, Italy

A R T I C LE I N FO

A B S T R A C T

Keywords: Masonry CFRP Mechanical anchor Out-of-plane Structural reinforcement

The use of composite materials, in particular Carbon Fibre Composite Materials (CFRP), as reinforcement of both concrete and masonry structures is more and more widespread in the structural rehabilitation and retrofitting of existing buildings, thanks to their excellent mechanical performance combined with lightness and simplicity of application. Since the bond capacity of CFRP-to-masonry bonded joints, with respect to in-plane loads, is generally lower than the composite tensile strength, several methods have been proposed in the literature to increase their structural performance. Among these, CFRP spike anchors showed to be able to effectively increase strength and dissipative capability of CFRP reinforcement sheets. Nevertheless, their use in technical practice is discouraged by the lack of specific rules that adequately support designers. The development of predictive formulas, appearing necessary to bridge this gap, requires an extensive experimental database that highlights the peculiar characteristics of these reinforcements. As a contribution in this field, this paper presents an experimental program concerning the analysis of the mechanical behavior of this type of reinforcements applied to masonry structural elements loaded by out-of-plane actions. The experimental results showed that the effectiveness of such anchors strongly depends on the shape ratio of the specimens.

1. Introduction As it is well known, a significant part of the existing buildings belonging to the World cultural and historical heritage (particularly European and Italian) has masonry structure. The peculiar characteristics of such structures (e.g. the presence of thrusting elements, quality of the constituent materials and of continuity between different structural elements, brickwork, etc.) and in particular the very low tensile strength of masonry, contribute to their seismic vulnerability. Different failure mechanisms can be activated by seismic actions: for example, both in-plane and out-of-plane mechanisms can be induced in masonry walls. As it is well known, the in-plane behavior of masonry walls (mainly defined by initial stiffness, strength and post crack characteristics) mostly contribute in the global structural response of masonry buildings. Nevertheless, out-of-plane failure mechanisms represent a serious life-safety hazard for this type of buildings, since they involve an almost instantaneous loss of the wall load bearing capacity for vertical loads. Several strengthening and consolidation techniques have been proposed in the scientific and technical literature to mitigate the seismic vulnerability of masonry structures. Among these, bonding Carbon Fiber Reinforced Polymer (CFRP) sheet demonstrated to

∗

effectively improve the mechanical behavior of masonry walls, subjected to both in-plane and out-of-plane actions [1–12], as well as of curved masonry structural elements [13–23]. Thanks to the wide experience gained in research activities concerning masonry structural elements reinforced by CFRP sheets subjected to in-plane loads [24–30], the main features of the mechanical behavior of such reinforcements are known and quite shared among specialists: tensile (in plane) force on the reinforcement sheet is transferred to the substrate mainly via shear stresses mostly concentrated in a limited portion of the bonded surface, which length is called “effective bond length”. Increasing the bond length more than the effective length do not produce a significantly increase in the (in plane) load bearing capacity of the reinforcement. Failure of such reinforcements, subjected to in-plane actions, generally occurs in the substrate, a few millimeters below the bonding surface. For this reason, the load bearing capacity strongly depends on the mechanical properties of the substrate [31], as well as other parameters like the brickwork and the sheet geometric characteristics [27]. Since the shear capacity of CFRP reinforcement sheets is generally lower than the composite tensile strength, several methods have been proposed in the literature to increase it [32,33]. Among these, CFRP spike anchors [31,34–42] proved to effectively increase strength and

Corresponding author. E-mail addresses: mario.fagone@unifi.it (M. Fagone), giovanna.ranocchiai@unifi.it (G. Ranocchiai).

https://doi.org/10.1016/j.compositesb.2018.05.031 Received 20 March 2018; Received in revised form 5 May 2018; Accepted 24 May 2018 Available online 25 May 2018 1359-8368/ © 2018 Published by Elsevier Ltd.

Composites Part B 150 (2018) 14–26

M. Fagone, G. Ranocchiai

ductility of CFRP reinforcement sheets: this “fiber-anchors”, made rolling a carbon fiber fabric sheet, are inserted into a hole drilled in the reinforced structural element and glued upward, in a “fan” shape, to the reinforcement strip, fixing the reinforcement to the substrate. Such anchors can effectively improve the CFRP-to-masonry joint and can be efficiently used in structural rehabilitation or retrofitting interventions needed, for example, after seismic events, or due to other factors such as environmental degradation, changes in usage resulting in heavier loading conditions. It is of course noteworthy that spike anchors bring some drawbacks since, for example, they require onerous preliminary preparation (surface preparation, drilling of masonry, inclusion of CFRP anchors, etc.) and longer installation time with respect to plane reinforcements. Moreover, such reinforcements are clearly irreversible so that their use in buildings belonging to the historical and cultural heritage must be carefully evaluated. Nevertheless, the pool of existing buildings in which anchored reinforcements can be effectively used is still very large. However, the use of mechanical anchors in technical practice is discouraged by the lack of specific rules adequately supporting design choices and sizing. In this regard, existing guidelines [27,43] state that the practical use of mechanical anchors must be substantiated by representative experimental testing. Of course, predictive formulas to be inserted in appropriate regulations, needed to bridge this gap, can be defined only after collection of an extensive experimental database that highlights the peculiar characteristics of these reinforcements also with reference to the substrate material. Most of the literature devoted to anchored CFRP-to-masonry reinforcements refer to in-plane actions. Therefore, given the lack of information in this respect, the experimental research described in this paper is devoted to the analysis of the effectiveness of spike anchors in CFRP reinforcements bonded to masonry structural elements loaded by out-of-plane actions. The experimental results showed that the effectiveness of such anchors strongly depends on the shape ratio of the specimens. The paper layout is the following: the experimental program is globally described in the next paragraph; the mechanical properties of the used materials are reported in section 3 and the specimens preparation is described in section 4; the test setup is described in section 5; the test results are reported an described respectively in section 6 and 7; final remarks conclude the paper.

beams have been considered. These were made using 1:2 scale bricks. In the scheme reported in Fig. 1, L represents the length of the specimens and is equal to 935 mm and 928 mm respectively for single leaf and double leaf specimens (see section 4 for the specimen's characteristics). Two different brickwork (single leaf and double leaf masonry) and two strengthening systems (without and with spike anchors) were considered in the experimental campaign; three specimens for each series were tested, so that the experimental program involved twelve specimens, all of them manufactured at the “Laboratorio Ufficiale Prove Materiali e Strutture” of the University of Florence, where the tests were carried out. 3. Material properties The materials employed in the experimental program described in this paper are analogous to those used in Refs. [31,34,35,41]. The reader can refer to these papers for a comprehensive description of the tests performed to characterize the mechanical properties of the materials. Here, just the main mechanical parameters are summarized for completeness sake. Soft mud bricks, also called solid pressed bricks, have been used to manufacture the specimens. These were preferred to drawn bricks because their material structure resembles the one of traditional soft pressed bricks, that are used in most existing buildings [44]. In order to determine the compressive strength and elastic modulus of the bricks, six of them were randomly chosen from the brick supply and cut to obtain cubic (50 × 50 × 50 mm) and prismatic specimens (50 × 50 × 150 mm) as schematized in Fig. 2. Moreover, prismatic specimens (40 × 40 × 200 mm) were obtained from other six randomly chosen bricks and tested using a three-point bending test scheme. Then, from one of the two halves of the specimen resulting from the bending test, a specimen was obtained for a tensile test. The mechanical parameters so determined for the bricks are summarized in Table 1. Ready mixed mortar, made with lime and cement as binder, was employed in manufacturing the specimens: the compressive and tensile strength, obtained according to [45] are reported in Table 2. A composite material, made of a unidirectional carbon fiber fabric and epoxy resin, was used to realize both the reinforcements and the spike anchors. The reinforcement sheets were applied to the substrate using a wet lay-up process (with a single layer of carbon fiber fabric), after surface preparation and primer application, according to the producer's guideline. The main characteristics of the constituent materials, declared by the producer, are summarized in Table 3.

2. Overall description of the experimental program The experimental program described in this paper aims at the analysis of the mechanical behavior of masonry specimens reinforced with CFRP sheets, subjected to out of plane actions. Particular attention is paid to the evaluation of the increase in load bearing capacity and dissipation capability due to CFRP spike anchors applied to the reinforcements, also with respect to the shape ratio of the specimens. In the experimental campaign described in Ref. [41] it has been observed that masonry panels, reinforced with CFRP sheets (with and without spike anchors) having fiber direction parallel to the bed joints, subjected to out-of-plane loads, substantially behaved, after the occurrence of the first cracks, as separate (reinforced) masonry beams. Therefore, in the experimental program here described we considered masonry beams, instead of panels, reinforced by CFRP sheets, loaded out of plane using a four point bending test scheme, made with bricks assembled as in masonry panels reinforced with horizontal strips. In view of the results described in Ref. [41], such beams can be intended as representative of the effective part of a reinforced masonry panel subjected to the considered type of loads. A four point bending test scheme was preferred to a three point (used in Ref. [41]) because in so doing the maximum tensile normal force in the reinforcement can be easily estimated using the measurements of a strain gauge bonded to the reinforcement sheet in the constant bending moment region of the beam, having length equal to 118 mm (see Fig. 1). In order to analyze the effectiveness of the reinforcement system with respect to the shape ratio of the specimen, both half brick and one brick thickness masonry

4. Specimens As previously said, two brickworks, having different shape ratio, were considered in the experimental campaign. In particular, half brick (single leaf) and one brick (double leaf) thickness specimens (see Fig. 3) were manufactured using 1:2 scale bricks (dimensions 28 × 53 × 113 mm) obtained cutting solid pressed bricks, having initial dimensions of 65 × 120 × 250 mm, in eight equal portions. Ready mixed mortar (see Table 2) was used to make the joints: all of them had thickness equal to 5 mm, except the vertical longitudinal joints of double leaf specimens, having thickness equal to 7 mm (see Figs. 3 and 12b) in order to respect the length of the 1:2 scale bricks. Four different specimens type were considered in the experimental campaign, namely:

• 1T.0.M series, single leaf specimens strengthened with not anchored CFRP sheets; • 2T.0.M series, double leaf specimens strengthened with not anchored CFRP sheets • 1T.A.M series, single leaf specimens strengthened with anchored CFRP sheets;

15

Composites Part B 150 (2018) 14–26

M. Fagone, G. Ranocchiai

Fig. 1. Test scheme (measures in mm)

• 2T.A.M series, double leaf specimens strengthened with anchored

Table 1 Bricks mechanical properties.

CFRP sheets.

n. specimens

Note that, no tests were performed on unreinforced specimens since, because of their slenderness, they broke during the positioning. Since three specimens have been manufactured for each series, each one of them was identified by the label of the corresponding series followed by a number from 1 to 3. All the specimens were cured for at least 28 days at room temperature and moisture before strengthening, with CFRP sheets having width equal to 63 mm and length equal to the specimen's one. The composite sheet, having thickness equal to 1 mm, was realized using a single layer of carbon fiber fabric using a wet lay-up procedure, as suggested by the producer, after surface preparation and laying of the primer (maximum 24 h before the application of the CFRP reinforcement). Even if monotonic actions have been considered, all the specimens were reinforced at both the lateral faces. As described before, half of specimens were fastened using spike anchors. In this case, the necessary bricks were pre-drilled and properly placed according to the schemes in Figs. 3 and 4c-d. The geometric characteristics of spike anchors, described in Fig. 5, were defined scaling appropriately those of the series (called F1_40_V) that in the experimental program described in Ref. [31] exhibited the best performance in terms of increment of both load bearing capacity and dissipative capability. In fact, since the specimens used in the present experimental campaign are 1:2 scaled, in an attempt to maintain

Compressive strength

Young modulus Direct tensile strength Bending tensile strength

6 6 6 6 6 6

load direction (see Fig. 2)

x y z y y y

Mean

C·V.

[MPa]

[%]

22.17 19.90 18.22 8712 2.49 3.36

9.28 5.11 7.40 6.92 16.90 33.77

Table 2 Three point (4 × 4x16 cm, L = 10 cm) bending tensile strength and compressive (4 × 4x4 cm) strength of mortar. n. specimens

(Bending) tensile strength Compressive strength

6 12

Mean

C·V.

[MPa]

[%]

1.85 5.18

9.42 8.212

constant the fiber density, the amount of fiber in the spike anchors and, consequently, the area of the cross section of the dowel, was scaled 1:2; so, a 2 scale factor was used to define the diameter (10 mm) and the

Fig. 2. Cubic and prismatic specimens obtained from one brick for compression and Young modulus tests. 16

Composites Part B 150 (2018) 14–26

M. Fagone, G. Ranocchiai

(Fig. 6a). Within 24 h, the (hardened) anchors were inserted into the holes, pre-filled with epoxy resin, and a first hand of epoxy resin was spread on the faces of the specimen to be reinforced (Fig. 6b). The carbon fiber fabric was laid on the epoxy resin allowing the passage of the protruding anchor through the fibers of unidirectional fabric (Fig. 6c). At this point, epoxy resin was spread on the composite sheet and the carbon fiber of the spike anchor, that was free from resin, was splayed over the sheet as to form a fan of 40 mm radius, 75° wide. In order to avoid the detachment of the fan anchors from the CFRP sheet during the hardening phase, these were cramped as showed in Fig. 6d. A similar procedure was followed for not anchored reinforcements, except, of course, for what concerns the insertion of the anchors.

Table 3 Mechanical properties (declared by the producer) of the reinforcing system components.

Unidirectional carbon fiber fabric Adhesive Primer

Nominal thickness

Tensile elastic modulus

Bending elastic modulus

Ultimate tensile strain

Characteristic tensile strength

[mm]

[MPa]

[MPa]

[%]

[MPa]

0.165

230000

–

1.3

2500

– –

– 1200

3300 –

– –

– > 20

5. Test setup and procedure

length (35 mm) of the dowel. The fan angle (75°) and radius (40 mm) were set equal to the corresponding of F1_40_V series, in order to avoid the sliding of the reinforcement strip from the fan anchor (called “Debonding Splay Anchor” failure mechanisms in Ref. [31]). Since single leaf specimens had thickness equal to 55 mm, it was impossible to drill two holes, 35 mm depth, at the same position of the opposite lateral faces of the specimens. So that, in this case it was chosen to drill a passing hole and to use a double fan anchor, fastening both the CFRP reinforcements, rather than shifting the anchors (see Fig. 5b) at the two opposite faces of the specimens. The round sharp external edge of the pre-drilled holes was smoothed on a 6 mm radius to prevent stress concentration as recommended by Ref. [43]. Impregnated anchors, already used in Refs. [31,34,35], were preferred to dry anchors because they showed, at least for applications to concrete substrates [46], best mechanical performance. These were obtained from a 100 mm width sheet (width refers here to the direction orthogonal to the fiber direction) according to the following procedure: a length of 133 mm or 75 mm (in the fiber direction) was cut out from the carbon fiber fabric to obtain spike anchors to be applied respectively to single leaf and double leaf specimens. Such sheets were partially impregnated with epoxy resin and rolled in the width (100 mm) direction as to form a dowel that was inserted in a pre-formed hole (having diameter 8.5 mm, so that it is 1.5 mm lower than the hole's diameter) in a polystyrene mold in order to harden while confined. The central portion, 40 mm length, and the end portion, 30 mm length, were impregnated for anchors applied respectively to double leaf and single leaf specimens. The manufacturing procedure for the anchored specimens was the following: after curing the masonry specimens (minimum 28 days), the lateral faces were sanded and cleaned; then the primer was spread

The test fixture and the test instrumentation are schematized in Figs. 1 and 7. The specimens, subjected to a four-point bending test scheme, were positioned vertically so that the self-weight did not affect the test results. In order to align the end hinges to the specimen's axes, two steel plates were glued to the end surfaces. After positioning, it was verified that at the hinges there was an appropriate play, in order to avoid an axial load being applied to the specimen at the beginning of the test. No additional vertical load but the self-weight was so applied to the specimens. Note that, since the specimens were constrained by two end-hinges, tensile forces could develop for high values of deflection. The horizontal load (perpendicular to the specimens' axis, see Fig. 7) was applied to the lateral surface of the specimens by means of two steel cylinders soldered to a steel plate, connected to a screw jack. Since the loading cylinders were directly in touch with the CFRP sheet, it was verified at the end of each test that they did not damage the reinforcement. The length of the middle constant bending moment region was 118 mm while the total length was 935 mm and 928 mm respectively for single leaf and double leaf specimens. The load, applied by a screw jack, was measured using a 50 kN load cell as schematized in Fig. 7. Two transducers (namely “lvdt 1” and “lvdt 2” in Fig. 7) measured the displacement of the loading plate and two other transducers measured the horizontal displacement of the steel plates glued to the end surfaces of the specimens, in order to determine the neat deflection of the beam. Two strain gauges, SG01 and SG02 in Fig. 7, were placed at the centerline of the CFRP sheets in order to estimate the maximum normal force in the reinforcements. Of course, since SG02 is subjected to compressive normal force, the deformations it measured can be considered reliable until any

Fig. 3. Specimen’s brickworks (measures in mm): (a) half brick and (b) one brick thickness specimens. Dashed circles indicate the position of pre-drilled holes in specimens to be reinforced by anchored CFRP sheets. 17

Composites Part B 150 (2018) 14–26

M. Fagone, G. Ranocchiai

Fig. 4. CFRP reinforcements considered in the experimental program (measures in mm): (a) not anchored CFRP reinforcement of single leaf specimens – 1T.0.M series; (b) not anchored CFRP reinforcement of double leaf specimens – 2T.0.M series; (c) anchored CFRP reinforcement of single leaf specimens – 1T.A.M series; (d) anchored CFRP reinforcement of double leaf specimens – 2T.A.M series.

Fig. 5. Spike anchor geometry (measures in mm): (a) plane view; (b) section referring to 1T.A.M series; (c) section referring to the end anchor of 2T.A.M series; (d) section referring to the intermediate anchor of 2T.A.M series. 18

Composites Part B 150 (2018) 14–26

M. Fagone, G. Ranocchiai

Fig. 6. Specimens production procedure.

Fig. 7. Test setup and instrumentation

6. Experimental results

compression instability phenomena caused the (local) detachment of the reinforcement from the substrate. Other strain gauges, according to the schemes in Fig. 7, were glued to the external surface of the “tensileside” of the CFRP reinforcement in order to estimate the strain profile (and so the tensile stress distribution) on the CFRP reinforcement subjected to tensile normal force. The load was applied increasing monotonically the displacement, at a constant rate of 0.015 mm/s, up to the specimens failure.

All the load-deflection diagrams obtained from the tests, reported in Fig. 8, exhibited a similar path: they present an initial (almost) linear branch up to the opening of the first cracks in the masonry; then, the diagrams present a second branch that, even if less regular then the first one, can be still approximated as linear. Then, a third, more scattered, branch follows, up to the specimen's failure. Therefore, in order to facilitate the comparison of the global behavior of the specimen's series considered in the experimental campaign, the load-displacement diagrams have been interpolated by a trilinear diagram (even if, as 19

Composites Part B 150 (2018) 14–26

M. Fagone, G. Ranocchiai

Fig. 8. Load-displacement diagrams and (poly-)linear interpolations: (a) sample 1T.0.M; (b) sample 1T.A.M; (c) sample 2T.0.M; (d) sample 2T.A.M; (e) all samples; (f) interpolations.

20

Composites Part B 150 (2018) 14–26

M. Fagone, G. Ranocchiai

Table 4 Characteristic load values F1, F2 and Fmax, average slope of the first (K1) and second (K2) branches of the load-displacement diagrams and failure model occurred during the tests. “MasF” = masonry failure; “CF” = cohesive failure; “PF” = prismatic failure. Increase in load bearing capacity (Fmax) due to spike anchors: “1T” series +0.6%, “2T” series +22.3%. Specimen

1T.0.M.1 1T.0.M.2 1T.0.M.3 1T.0.M 1T.A.M.1 1T.A.M.2 1T.A.M.3 1T.A.M 2T.0.M.1 2T.0.M.2 2T.0.M.3 2T.0.M 2T.A.M.1 2T.A.M.2 2T.A.M.3 2T.A.M

Failure mode

Mas-F Mas-F Mas-F mean c.v. [%] Mas-F Mas-F Mas-F mean c.v. [%] CF CF CF mean c.v. [%] CF-PF CF-PF CF-PF mean c.v. [%]

F1

F2

Fmax

K1

K2

[N]

[N]

[N]

[N/mm]

[N/mm]

1013 1125 898 1012 11.24 1126 1182 909 1073 13.44 2896 2463 3394 2918 15.96 3597 4024 3108 3576 12.82

3326 3135 2223 2895 20.36 2816 3051 2454 2774 10.85 4270 3467 5868 4535 26.95 5581 4293 5904 5259 16.21

3510 3389 2521 3140 17.18 3180 3407 2891 3159 8.19 5388 5052 6162 5534 10.29 7271 5244 7787 6768 19.87

549 459 470 493 9.99 485 462 335 428 18.84 1616 1955 2433 2001 20.51 2334 1460 2320 2038 24.58

202 213 140 185 21.21 184 190 160 178 8.89 623 911 840 791 18.99 759 742 860 787 8.09

Table 6 Load and stiffness ratios. Series

1T.0.M 1T.A.M 2T.0.M 2T.A.M

1T.0.M 1T.A.M 2T.0.M 2T.A.M

εF1

Series

1T.0.M 1T.A.M 2T.0.M 2T.A.M

c.v.[%]

Mean [μ]

c.v.[%]

667 665 662 694

2.89 6.47 14.50 10.19

6315 6297 3659 5114

19.15 12.42 23.26 39.54

F3/Fmax [%]

K2/K1 [%]

mean

c.v.

mean

c.v.

mean

c.v.

mean

c.v.

32.55 33.85 52.53 55.37

10.50 6.23 6.35 34.52

91.81 87.66 81.04 78.14

3.63 2.82 16.53 4.16

90.47 95.34 88.47 92.84

11.34 3.57 3.44 6.71

37.69 42.28 39.88 40.13

22.10 11.78 15.44 23.73

δ3/δFmax [%]

δ3/δ2 [%]

mean

c.v.

mean

c.v.

151.78 108.32 109.61 129.26

29.12 12.16 10.81 39.20

194.91 199.64 240.42 292.11

44.48 20.21 47.53 14.33

stroke of the interpolating load-displacement graph, cracks grow both in the masonry substrate and behind the CFRP reinforcement (a few millimeters below the CFRP sheet-substrate interface). Then, after the load value F2, the diagrams show a third, very scattered, branch having a very low average slope: at this stage, the crack pattern grows up to the failure of the specimens. Note that the load value F2 is not very different from the maximum load Fmax. The ratio F2/Fmax is about 90% for single leaf series and 80% for double leaf specimens (see Table 6). It is noteworthy that the load ratio F1/Fmax, F2/Fmax and F3/Fmax is almost not affected by spike anchors for both single leaf and double leaf specimens, as well as the stiffness ratio K2/K1 and the deflection ratio δ3/δ2 (Table 7). The ratio between δ3/δFmax does not show a clear trend since the maximum load occurs within the third branch of the diagram having, as previously said, a very irregular path. It is apparent, from the average values reported in Table 4, that spike anchors did not affect the characteristic load and stiffness values of single leaf specimens. For double leaf specimens, they produced, instead, an increase of F1 and Fmax of about 22% and of F2 of about 16%, while, again, did not affect the average stiffness values. For single leaf specimens, cracks in the substrate were mainly localized in the central portion, close to the loading plate (see Fig. 9a–b); It is apparent, from the crack pattern, that the stress distribution in the specimen was such that a compressed internal arch (within the thickness) occurred, for which the reinforcement acted as a chain. Failure of one leaf specimens always occurred because of masonry failure (“MasF” in Table 4) as showed in Fig. 9b. Contrariwise, even if double leaf specimens exhibited, at failure, a diffuse crack pattern in the substrate, failure always occurred because of the detachment of the reinforcement from the substrate, a few millimeters below the reinforcement-substrate interface (“CF” – cohesive failure – in Table 4, see Fig. 9c). This was associated to the pull out of a prismatic shaped portion of brick for anchored specimens (“PF” – prismatic failure – in Table 4, see Fig. 9d). Of course, in this case spike anchors modified the failure mode of the specimens preventing the “simple” detachment of the reinforcement from the substrate, so that the load bearing capacity was increased of about 22%. In order to analyze the behavior of the reinforcement sheets during the tests, the strain profiles, obtained using the average values of strain measured by the strain gauges (Fig. 7) during the test, are reported in Fig. 10. In the diagrams, the half-length of the vertical segments corresponds to the standard deviation of the recorded values, while “dist” represents the distance of the strain gauge from the upper end of the specimen (see Fig. 7). Such strain profiles have been determined with reference to four load level (L1 = F1/2; L2 = F1; L3 = F1 + (0.75 Fmax –

εFmax

Mean [μ]

F2/Fmax [%]

Table 7 Displacement ratio.

Table 5 Characteristic axial strain values at first crack (εF1) and maximum load (εFmax). Series

F1/Fmax [%]

underlined above, the third branch is not very regular). In these diagrams (see Fig. 8-f), P1 represents the intersection point between the first and the second branches and P2 is the intersection point between the second and the third ones. P3, that is the final point of the third linear interpolation branch, has abscissa equal to the deflection recorded during the test at failure. Note that, according to their definition, P1, P2 and P3 do not necessarily belong to the equilibrium path; in the following, F1, F2 and F3 represent the ordinate and δ1, δ2 and δ3 represent the abscissas of P1, P2 and P3. The failure mode, the maximum load recorded during the tests (Fmax), as well as the load values F1 and F2 are reported in Table 4, together with the slope K1 and K2 of the first and second linear branch of the interpolating diagrams. In all the tests, the first crack opened in the masonry substrate at the end of the first branch (at P1): at this point, one or two “bending-type” fractures, involving both bricks and perpend masonry joints, occurred close to the end of the constant (central) bending moment region of the specimens. Even if, of course, the first crack load F1 is very different for single leaf and double leaf specimens because of the different inertia moments, first crack occurred, for all the specimen's series, at very close values of axial strain (εF1 in Table 5), ranging from 662μ and 694μ. These values were recorded using the strain gauge (SG01 in Fig. 7) glued at the center of the reinforcement sheet. Note that these values are about twice the tensile crack deformation of the bricks (286μ — 386μ), that can be evaluated as the ratio between the (direct — bending) tensile strength and the elastic modulus of bricks, using the values reported in Table 1. After P1, that is within the test stage corresponding to the second 21

Composites Part B 150 (2018) 14–26

M. Fagone, G. Ranocchiai

Fig. 9. Failure modes: (a) sample 1T.0.M (“MasF”); (b) sample 1T.A.M (“MasF”); (c) sample 2T.0.M (“CF”); (d) sample 2T.A.M (“CF-PF”).

specimens, 9.4 kN for 2T.0.M series and 13.1 kN for 2T.A.M series (see Table 8). In order to estimate the difference between these values and the capacity of the reinforcement bonded to the masonry substrate (not only carbon fiber fabric, as previously done), these can be compared with the load bearing capacity of such reinforcements subjected to single lap shear tests [31,34,35]: of course, such comparison can provide only a rough estimation of the ratio between the maximum axial load in the reinforcement and its capacity since the mechanical behavior of the considered reinforcements are expected to be different if they are applied to specimens subjected to bending or single lap shear conditions, because of the differences in the stress distribution and in the boundary conditions. Seeing that in Ref. [47] it has been proved that the ratio between the (double lap) shear capacity and the width of not anchored CFRP reinforcements is almost constant for width ranging from 60 mm to 100 mm, the shear capacity of the CFRP reinforcements considered in the present experimental campaign can be evaluated, as a first approximation, properly scaling the load bearing capacity of the reinforcements considered in Ref. [35], where the specimens were manufactured using the same materials and the same procedure described in this paper. It was obtained in Ref. [35] that the average shear capacity is 13.4 kN and 19.6 kN, respectively for not anchored and anchored CFRP reinforcements, having width equal to 100 mm and length higher than the effective bonding length. Scaling the previous values using a factor of 63/100, it is estimated a shear capacity of 8.4 kN for not anchored reinforcements and of 12.4 kN for anchored reinforcements (see Table 8). The ratio between the average maximum tensile force evaluated via strain gauge measure and the shear capacity estimated according to the width ratio is 1.93 for 1T.0.M series and 1.31 for 1T.A.M series, so that it appears that the bearing capacity of such reinforcements bonded to masonry specimens subjected to bending actions is higher than the one corresponding to simple shear. This can be due to confinement effects induced by bending, that produced high curvature of the masonry beams: the maximum deflection of single leaf specimens (see Fig. 8) is, in fact, about 20÷35 mm, so that 38÷66% of the specimen's thickness and 2÷4% of the specimen's length. The high load bearing capacity of the reinforcement bonded to single leaf specimens produce a failure mode related to masonry crushing (“MasF” in Table 4). For double leaf specimens (which failure mode is related to the strengthening system) the evaluated maximum tensile force in the reinforcement is close to (even if a few higher than) the estimated shear capacity: in view of the higher shape ratio (thickness/length) and lower maximum deflection values of these specimens with respect to single leaf series, we can suppose that, in this case, there

F1)/2 and L4 = 0.75 Fmax) as indicated in the legend of Fig. 10. Even if, according to their definition, such load levels depend on the average values of F1 and Fmax of each series, they almost coincide for 1T.0.M and 1T.A.M that exhibited almost the same characteristic loads values. It is apparent that, except for L3 and L4 in 1T.0.M series and L4 in 2T.A.M series, all the strain profiles are almost linear so that they agree with the bending moment diagrams in the specimens. This indicates that, despite the cracks, occurred close to the CFRP reinforcement, this almost remained attached to the substrate also for not very low load levels. On the contrary, the shape of the strain profiles corresponding to L3 and L4 in 1T.0.M series and L4 in 2T.A.M series indicate that the reinforcement detached from the substrate in its central portion. Note that, comparing the strain profiles in Fig. 10, it is apparent that spike anchors did not affect the reinforcement behavior for small load levels; in fact, the strain profile corresponding to L4 = 4.2 kN in Fig. 10c almost coincide with the one corresponding to L3 = 4.3 kN in Fig. 10d. 7. Discussion The failure mode of the specimens tested in the present experimental campaign did not involve tensile failure of the reinforcement (see Table 4); therefore, it is obvious that the maximum tensile force in the reinforcement is lower than its tensile strength. The ratio between these values (“maximum tensile load in the reinforcement”/“tensile strength”) can be estimated as the ratio between the tensile strain εFmax corresponding to the maximum load, reported in Table 5 (measured by SG01, see Fig. 7), and the ultimate tensile strength of the carbon fiber fabric declared by the producer (1.3%, see Table 3). It is obtained that such “exploitation ratio” is about 48% for 1T.0.M and 1T.A.M series, 28% for 2T.0.M series and 40% for 2T.A.M series, so that the load level in the reinforcement is by far lower than its strength. As described in section 3.3, the composite reinforcement sheet, having total thickness equal to 1 mm, was made with a single layer of carbon fiber fabric. The homogenized elastic modulus of the composite (in the fiber direction) can be easily evaluated using the rule of mixtures as Ef = 40706 MPa (see mechanical parameters summarized in Table 3). By means of this value, the average tensile force acting in the central portion of the reinforcement (corresponding to the constant bending moment region of the specimens) when the maximum load Fmax is applied to the specimens, can be estimated as εFmax × Ef × Af, being Af = 63 mm2 the area of the composite cross section. In so doing it is found that the average value of the maximum tensile force in the reinforcement, at Fmax, is equal to about 16.2 kN for single leaf 22

Composites Part B 150 (2018) 14–26

M. Fagone, G. Ranocchiai

Fig. 10. Strain profiles: (a) sample 1T.0.M; (b) sample 1T.A.M; (c) sample 2T.0.M; (d) sample 2T.A.M. “dist” has origin at the upper end of the specimens; dots represent the mean of the values recorded by the strain gauges, while dashed lines simply link these values. Vertical segments half length correspond to the standard deviation.

be used to this purpose, since the compressed reinforcement sheet, where SG02 is glued to, in correspondence of Fmax was locally detached from the substrate because of instability phenomena. Assuming that the (rectangular) cross section of the specimens remains plane during bending, the distribution of the axial deformation on the thickness can be assumed to be linear as represented in Fig. 11. Of course, the closer the load is to failure, the rougher is the approximation of this assumption. According to the linear simple bending theory, the maximum (tensile) deformation, corresponding to the CFRP reinforcement sheet, can be evaluated as εf = χ (h-a), while the minimum (compressive) deformation in the masonry can be evaluated as εm = χ a, being a the distance of the neutral axis from the upper (compressed) face and χ the curvature. Within the linear elastic response of the materials, the maximum and minimum normal stresses can be evaluated respectively as

Table 8 Comparison between the maximum tensile force in the reinforcement sheet (estimated via εFmax) and the Single Lap Shear capacity, estimated scaling the values obtained in Ref. [35] with respect to the reinforcement width. Series

EfAf εFmax (1) [N]

Estimated SLS capacity (2) [N]

Evaluated (1)/estimated (2)

1T.0.M 1T.A.M 2T.0.M 2T.A.M

16195 16149 9384 13116

8374 12361 8374 12361

1.93 1.31 1.12 1.06

was no confinement effect increasing the reinforcement capacity. Moreover, in order to have a first approximation of the maximum compressive stress in the masonry at Fmax, we can use the standard bending beams theory applied to composite cross section, as it is usual, for example, in the analysis of reinforced concrete beams. The deformation values measured by the strain gauge SG02 (see Fig. 7) cannot

σf = Ef εf = Ef χ (h − a) σm = Em εm = Em χa being. 23

(1)

Composites Part B 150 (2018) 14–26

M. Fagone, G. Ranocchiai

Fig. 11. Internal strain and stress distribution according to the hypothesis of rigid plane cross section and (unilateral) linear elastic materials.

The elastic modulus of the masonry Em can be estimated via homogenization with reference to the periodic cell of the brickwork (see Fig. 12), subjected to uniform compressive load. Being the elastic modulus of the bricks and of the mortar respectively Eb = 8712 MPa (see Table 1) and Emor = 2172 MPa (as obtained by ad hoc experimental tests), using a standard procedure it is obtained Em = 7126 MPa and Em = 6682 MPa respectively as the homogenized elastic modulus of single leaf and double leaf specimens. The maximum bending moment M can be evaluated as (see Fig. 1)

Ef: elastic modulus of the CFRP composite Em: elastic modulus of the masonry Following a standard procedure, the distance a of the neutral axis from the upper face of the specimen can be evaluated from the force equilibrium equation as

σf Af − σmb

a a2 = 0 → E f Af (h− a) − Em b = 0 2 2

(2)

while the curvature χ can be evaluated from the equilibrium of moments equation as

(

a M=σf Af h− 3

→ χ=

)=

a σmb 2

M

( )

E f Af (h−a) h−

a 3

=

(

a h− 3

M=

)

( ) a 3

(4)

The maximum compressive stress in the masonry, as well as the tensile strain and normal force in the CFRP reinforcement obtained using the procedure previously described are summarized in Table 9. The estimated values of maximum compressive stress in the masonry for single leaf specimens are close to the compressive strength of bricks (see Table 1); this agrees with the failure mode of such specimens (see Table 4), occurred always because of “compressive-like” masonry failure. Conversely, the estimated maximum compressive stress in masonry for double leaf specimens (failed because of the CFRP reinforcement) is far from the compressive strength of the bricks. Regarding the CFRP composite, the previous procedure leads to underestimate, with respect to the recorded values reported in Table 5, the tensile strain at

2M Em ba2 h−

Fmax L 118mm ⎛ + 51mm − ⎞ 2 ⎝2 2 ⎠

(3)

where: Af = 63 mm2 area of the CFRP reinforcement cross section b = 93 mm width of the masonry beam cross section h = 53 mm height of single leaf beam cross section 113 mm height of double leaf beam cross section Ef = 40706 MPa homogenized elastic modulus of the composite Em homogenized elastic modulus of the masonry

Fig. 12. Periodic cell of (a) single leaf and (b) double leaf specimens. 24

Composites Part B 150 (2018) 14–26

M. Fagone, G. Ranocchiai

reinforcement remained substantially bonded to the substrate at least up to this load level; - the ratio between the estimated value of the maximum tensile load in the reinforcement and its nominal tensile strength is about 48% for 1T.0.M and 1T.A.M series, 28% for 2T.0.M series and 40% for 2T.A.M series; so that the maximum load level in the reinforcement is by far lower than its strength; - the estimated values of maximum compressive stress in the masonry for single leaf specimens are close to the compressive strength of bricks, according to the failure mode of such specimens; conversely, the estimated maximum compressive stress in masonry for double leaf specimens is far from the compressive strength of the bricks

Table 9 Maximum stress in masonry (σm), maximum deformation (εf) and normal force (Nf = εf × Ef × Af) in the CFRP reinforcement estimated sing the model schematized in Fig. 11. Series

1T.0.M 1T.A.M 2T.0.M 2T.A.M

σm

εf

Nf

[MPa]

[μ]

[N]

19.53 19.65 9.77 11.94

5933 5969 4726 5780

15214 15306 12120 14823

the maximum load for single leaf specimens, and to overestimate this value for double leaf specimens. This can be due to the difference between the (roughly) supposed internal force distribution, defined according to the hypothesis of rigid-plane cross sections and (unilateral) linear elastic behavior of the material, and the real one. In any case, the estimated values agree with the global behavior and the failure modes experienced by the specimens. Note that, since the values of maximum stress and strain predicted by the model schematized in Fig. 11 are lower than the elastic limit of the materials, the maximum (experimental) bending moment is, of course, lower than the elastic limit bending moment that could be estimated assuming a unilateral linear elastic behavior of the materials. Therefore, if more refined constitutive models were used, taking into account the post-peak branch of the stress-strain curve of the masonry, they would have estimate moment values higher than the experimental one. In fact, the failure mechanisms of the specimens, especially double leaf ones, involve other phenomena with respect to the simple crisis due to compression of the cross section.

Acknowledgement The Authors gratefully acknowledge the financial support provided by the Italian Department of Civil Protection and ReLUIS (Rete dei Laboratori Universitari di Ingegneria Sismica), 2014–2016 Grant Innovative Materials. Moreover, BASF and TERREAL-SANMARCO are gratefully acknowledged for providing material. References [1] Hosseini A, Mostofinejad D, Emami M. Influence of bonding technique on bond behavior of CFRP-to-clay brick masonry joints: experimental study using particle image velocimetry (PIV). Int J Adhesion Adhes 2015;59:27–39. http://dx.doi.org/ 10.1016/j.ijadhadh.2015.01.015. [2] Carloni C, Focacci F. FRP-masonry interfacial debonding: an energy balance approach to determine the influence of the mortar joints. Eur J Mech Solid 2016;55:122–33https://doi.org/10.1016/j.euromechsol.2015.08.003. [3] Sistani Nezhad R, Kabir MZ. Experimental investigation on out-of-plane behavior of GFRP retrofitted masonry panels. Construct Build Mater 2017;131:630–40. http:// dx.doi.org/10.1016/j.conbuildmat.2016.11.118. [4] Sorrentino L, D'Ayala D, de Felice G, Griffith MC, Lagomarsino S, Magenes G. Review of out-of-plane seismic assessment techniques applied to existing masonry buildings. Int J Architect Herit 2016;0. http://dx.doi.org/10.1080/15583058.2016. 1237586. 15583058.2016.1237586. [5] Bruggi M, Milani G. Optimal FRP reinforcement of masonry walls out-of-plane loaded: a combined homogenization–topology optimization approach complying with masonry strength domain. Comput Struct 2015;153:49–74https://doi.org/10. 1016/j.compstruc.2015.02.004. [6] Anania L, D'Agata G, Giaquinta C, Badalà A. Out of plane behavior of calcareous masonry panels strengthened by CFRP. APCBEE Procedia 2014;9:401–6. http://dx. doi.org/10.1016/j.apcbee.2014.01.070. [7] De Felice G, Al Shatva O, Mauro A, Sorrentino L. On the seismic behavior of out-ofplane loaded masonry walls Wroclaw: Dolnoslaskie Wydawnictwo Edukacyjne-Dwe 2012. [8] Hamed E, Rabinovitch O. Failure characteristics of FRP-strengthened masonry walls under out-of-plane loads. Eng Struct 2010;32:2134–45https://doi.org/10.1016/j. engstruct.2010.03.016. [9] Lunn D, Rizkalla S. Design of FRP-strengthened infill-masonry walls subjected to out-of-plane loading. J Compos Construct 2013;18:A4013002. http://dx.doi.org/ 10.1061/(ASCE)CC.1943-5614.0000412. [10] Sinicropi D, Perria E, Galassi S, Paradiso M, Borri A. Artificial ageing of mortar prisms reinforced through steel, glass and organic fibres vol. 624. Trans Tech Publications; 2015. http://dx.doi.org/10.4028/www.scientific.net/KEM.624.542. [11] Paradiso M, Galassi S, Borri A, Sinicropi D. Reticolatus: an innovative reinforcement for irregular masonry. A numeric model. Struct. Archit. Concepts, appl. Challenges proc. 2nd int. Conf. Struct. Archit. ICSA 2013; 2013. p. 841–8. [12] D'Ambrisi A, Feo L, Focacci F. Experimental and analytical investigation on bond between Carbon-FRCM materials and masonry. Compos Part B-Engineering 2013;46:15–20. http://dx.doi.org/10.1016/J.Compositesb.2012.10.018. [13] Anania L, Badalà A, D'Agata G. The post strengthening of the masonry vaults by the Ω-Wrap technique based on the use of C-FRP. Construct Build Mater 2013;47:1053–68https://doi.org/10.1016/j.conbuildmat.2013.05.012. [14] Borri A, Castori G, Corradi M. Intrados strengthening of brick masonry arches with composite materials. Compos Part B-Engineering 2011;42:1164–72. http://dx.doi. org/10.1016/j.compositesb.2011.03.005. [15] Foraboschi P. Strengthening of masonry arches with fiber-reinforced polymer strips. J Compos Construct 2004;8:191–202. http://dx.doi.org/10.1061/(Asce)10900268(2004)8:3(191). [16] Oliveira DV, Basilio I, Lourenco PB. Experimental behavior of FRP strengthened masonry arches. J Compos Construct 2010;14:312–22. http://dx.doi.org/10.1061/ (asce)cc.1943-5614.0000086. [17] Valluzzi MR, Valdemarca M, Modena C. Behavior of brick masonry vaults strengthened by FRP laminates. J Compos Construct 2001;5:163–9. http://dx.doi. org/10.1061/(Asce)1090-0268(2001)5:3(163).

8. Conclusions An experimental program aimed to the analysis of the mechanical behavior of masonry specimens reinforced with CFRP sheets, subjected to out of plane actions has been described in this paper. In particular, the influence of spike anchors on the load bearing capacity and dissipation capability of the reinforcements has been analyzed, also with respect to the shape ratio of the specimens. The experimental results showed that: - for all the specimen's series considered in the experimental program, regardless to the shape ratio, first cracks occurred at very close values of axial strain, ranging from 662μ and 694μ; these values are about twice the tensile crack deformation of the bricks (286μ − 386μ), so that it appears that first cracks are related not only to the tensile strength, but also to other phenomena, such as confining effects occurring in the considered reinforced specimens subjected to bending; - failure mode of single leaf and double leaf specimens were very different from each other. Single leaf specimens exhibited a cracks pattern mostly concentrated in their central portion and failed because of masonry compressive failure. Double leaf specimens exhibited, at failure, a diffuse crack pattern in the substrate, but failure always occurred because of the detachment of the reinforcement from the substrate; this was associated to the pull out of a prismatic shaped portion of brick for anchored reinforcements. Therefore, spike anchors modified the failure mode of double leaf specimens; - according to the previous point, spike anchors did not affect the characteristic load and stiffness values of single leaf specimens; for double leaf specimens, anchors produced an increase of F1 and Fmax of about 22% and of F2 of about 16%; - for almost all the specimens, the shape of the strain profiles corresponding to load levels up to 75% of the maximum load almost agree with the bending moment diagram. This suggests that the 25

Composites Part B 150 (2018) 14–26

M. Fagone, G. Ranocchiai

(Asce)Cc.1943-5614.0000276. [33] Grelle SV, Sneed LH. Review of anchorage systems for externally bonded FRP laminates. Int J Concr Struct Mater 2013;7:17–33. http://dx.doi.org/10.1007/ s40069-013-0029-0. [34] Caggegi C, Pensee V, Fagone M, Cuomo M, Chevalier L. Experimental global analysis of the efficiency of carbon fiber anchors applied over CFRP strengthened bricks. Construct Build Mater 2014;53:203–12. http://dx.doi.org/10.1016/j. conbuildmat.2013.11.086. [35] Fagone M, Ranocchiai G, Briccoli Bati S. An experimental analysis about the effects of mortar joints on the efficiency of anchored CFRP-to-masonry reinforcements. Compos Part B-Engineering 2015;76:133–48. http://dx.doi.org/10.1016/j. compositesb.2015.01.050. [36] Contrafatto L, Cosenza R. Behaviour of post-installed adhesive anchors in natural stone. Construct Build Mater 2014;68:355–69https://doi.org/10.1016/j. conbuildmat.2014.05.099. [37] Contrafatto L, Cosenza R. Prediction of the pull-out strength of chemical anchors in natural stone. Frat Ed Integrità Strutt 2014:8. http://dx.doi.org/10.3221/IGF-ESIS. 29.17. [38] Briccoli Bati S, Fagone M, Ranocchiai G. The effects of mortar joints on the efficiency of anchored CFRP sheets reinforcements of brick-masonry. Key Eng Mater 2015;624:575–83. http://dx.doi.org/10.4028/www.scientific.net/KEM.624.575. [39] Ceroni F. Bond tests to evaluate the effectiveness of anchoring devices for CFRP sheets epoxy bonded over masonry elements. Compos B Eng 2017;113:317–30. http://dx.doi.org/10.1016/j.compositesb.2017.01.042. [40] Foraboschi P. Effectiveness of novel methods to increase the FRP-masonry bond capacity. Compos B Eng 2016;107:214–32. http://dx.doi.org/10.1016/j. compositesb.2016.09.060. [41] Fagone M, Ranocchiai G. On the mechanical behaviour of anchored CFRP-to-masonry reinforcements. In: Modena C, da Porto F, Valluzzi MR, editors. 16th int. Brick block mason. Conf. Padova, Italy: Taylor & Francis Group; 2016. p. 387–93. [42] Fagone M, Ranocchiai G. An experimental analysis of the mechanical behaviour of anchored CFRP-to-masonry reinforcements loaded by out-of-plane actions. In: Di Tommaso A, Gentilini C, Castellazzi G, editors. Mech. Mason. Struct. Strengthened with compos. Mater. II - MuRiCo5 Bologna, Italia: Trans Tech Publications; 2017. p. 204–11. http://dx.doi.org/10.4028/www.scientific.net/KEM.747.204. [43] Institute AC. ACI 440.2R-08: Guide for design and construction of externally bonded FRP systems for strengthening concrete structures. 2008. ACI 440.2R. [44] Briccoli Bati S, Ranocchiai G. A critical review of experimental techniques for brick materials. Proc. 10th int. Brick block mason. Conf, vols. 1–3. 1994. p. 1247–55. Calgary. [45] UNI EN 1015–11. Methods of test for mortar for masonry Part 11: determination of flexural and compressive strength of hardened mortar. 2007. UNI EN 101. [46] Zhang HW, Smith ST, Kim SJ. Optimisation of carbon and glass FRP anchor design. Construct Build Mater 2012;32:1–12. http://dx.doi.org/10.1016/J.Conbuildmat. 2010.11.100. [47] Briccoli Bati S, Fagone M. An analysis of CFRP-brick bonded joints. XVIII Convegno Ital. di Mecc. Comput., Siracusa, Italy. 2010. p. 1–4.

[18] Basilio I, Fedele R, Lourenço PB, Milani G. Assessment of curved FRP-reinforced masonry prisms: experiments and modeling. Construct Build Mater 2014;51:492–505https://doi.org/10.1016/j.conbuildmat.2013.11.011. [19] Fagone M, Loccarini F, Ranocchiai G. Strength evaluation of jute fabric for the reinforcement of rammed earth structures. Compos B Eng 2017;113:1–13. http:// dx.doi.org/10.1016/j.compositesb.2016.12.054. [20] Briccoli Bati S, Fagone M, Rotunno T. Lower bound limit analysis of masonry arches with CFRP reinforcements: a numerical method. J Compos Construct 2013;17:366. http://dx.doi.org/10.1061/(asce)cc.1943-5614.0000350. [21] Cancelliere I, Imbimbo M, Sacco E. Experimental tests and numerical modeling of reinforced masonry arches. Eng Struct 2010;32:776–92. http://dx.doi.org/10. 1016/j.engstruct.2009.12.005. [22] Lucchesi M, Pintucchi B, Zani N. Dynamic analysis of FRP-reinforced masonry arches via a no-tension model with damage. Key Eng Mater 2015;624:619–26. http:// dx.doi.org/10.4028/www.scientific.net/KEM.624.619. [23] Fagone M, Rotunno T, Briccoli Bati S. The groin vaults of st. John hospital in jerusalem: an experimental analysis on a scale model. Int J Architect Herit 2016;10:903–18. http://dx.doi.org/10.1080/15583058.2016.1158331. [24] Valluzzi MR, Oliveira DV, Caratelli A, Castori G, Corradi M, de Felice G, et al. Round Robin Test for composite-to-brick shear bond characterization. Mater Struct 2012;45:1761–91. http://dx.doi.org/10.1617/s11527-012-9883-5. [25] Malena M, Focacci F, Carloni C, de Felice G. The effect of the shape of the cohesive material law on the stress transfer at the FRP-masonry interface. Compos B Eng 2017;110:368–80. http://dx.doi.org/10.1016/j.compositesb.2016.11.012. [26] Freddi F, Sacco E. An interphase model for the analysis of the masonry-FRP bond. Compos Struct 2016;138:322–34. http://dx.doi.org/10.1016/j.compstruct.2015. 11.041. [27] National Research Council. CNR-DT 200/R1: istruzioni per la Progettazione, l'Esecuzione ed il Controllo di Interventi di Consolidamento Statico mediante l’utilizzo di Compositi Fibrorinforzati. 2013. [28] Ceroni F, de Felice G, Grande E, Malena M, Mazzotti C, Murgo F, et al. Analytical and numerical modeling of composite-to-brick bond. Mater Struct 2014;47:1987–2003. http://dx.doi.org/10.1617/s11527-014-0382-8. [29] Ghiassi B, Oliveira DV, Lourenco PB, Marcari G. Numerical study of the role of mortar joints in the bond behavior of FRP-strengthened masonry. Compos Part BEngineering 2013;46:21–30. http://dx.doi.org/10.1016/j.compositesb.2012.10. 017. [30] Briccoli Bati S, Fagone M. Lunghezza ottimale di ancoraggio per rinforzi in FRP su elementi in laterizio. Valutazione e riduzione della vulnerabilità sismica di Edif. esistenti c.a., Roma. 2008. p. 1–8. [31] Fagone M, Ranocchiai G, Caggegi C, Briccoli Bati S, Cuomo M. The efficiency of mechanical anchors in CFRP strengthening of masonry: an experimental analysis. Compos B Eng 2014;64:1–15. http://dx.doi.org/10.1016/j.compositesb.2014.03. 018. [32] Kalfat R, Al-Mahaidi R, Smith ST. Anchorage devices used to improve the performance of reinforced concrete beams retrofitted with FRP composites: state-of-theart review. J Compos Construct 2013;17:14–33. http://dx.doi.org/10.1061/

26