Masonry wall panels with GFRP and steel-cord strengthening subjected to cyclic shear: An experimental study

Construction and Building Materials 56 (2014) 63–73

Contents lists available at ScienceDirect

Construction and Building Materials

journal homepage: www.elsevier .com/locate /conbui ldmat

Masonry wall panels with GFRP and steel-cord strengthening subjectedto cyclic shear: An experimental study

http://dx.doi.org/10.1016/j.conbuildmat.2014.01.0560950-0618/� 2014 Elsevier Ltd. All rights reserved.

⇑ Corresponding author. Tel.: +44 01912437649; fax: +44 01912273598.E-mail address: [email protected] (M. Corradi).

Antonio Borri a, Giulio Castori a, Marco Corradi b,⇑, Romina Sisti a

a Department of Engineering, University of Perugia, Via Duranti 93, 06125 Perugia, Italyb Department of Mechanical & Construction Engineering, Northumbria University, Wynne-Jones Building, NE1 8ST Newcastle Upon Tyne, UK

h i g h l i g h t s

�We carried out shear tests on 23 masonry panels, before and after reinforcement.� Panels were reinforced with high strength steel cords and a GFRP mesh.� Non-reinforced panels underwent shear failure involving only mortar joints.� Reinforced walls presented enhanced behavior and increased mechanical parameters.� The use of scaled specimens may be considered representative of the tested masonry.

a r t i c l e i n f o

Article history:Received 8 October 2013Received in revised form 16 January 2014Accepted 16 January 2014

Keywords:MasonryShear strengthReinforcementSteel cordsGFRM

a b s t r a c t

This paper provides the results of a series of shear tests carried out on wall panels reinforced with twotechniques by means of jacketing with GFRP (Glass Fiber Reinforced Polymers) mesh inserted inside aninorganic matrix and a reinforced repointing of mortar joints using high strength stainless steel cords.The tests were done on panels assembled in the laboratory and were carried out using a widely-knowntest method (diagonal compression test). Masonry materials include stone and brick in which a variety ofunit sizes and forms are produced. Mortar was a hydraulic lime/sand with a plasticiser added to improveworkability. The failure loads, in-plane and out-of-plane deflections and failure modes were recorded.Based on the results of the experimental program, it appears that the in-plane shear strength of the rein-forced masonry wall systems increased significantly compared with that of control unreinforced masonrypanels.

� 2014 Elsevier Ltd. All rights reserved.

1. Introduction

The behavior of a Unreinforced Masonry (URM) building inresponse to static and dynamic forces caused by earthquakes isprimarily linked to the type of masonry, the poor quality of whichis a factor of serious vulnerability that often makes ineffective thework aimed at limiting local collapse mechanisms. In these cases,the achieving of an adequate degree of safety is connected to theuse of reinforcements that are able to ensure an improvement ofthe mechanical properties of the walls.

The vulnerability of unreinforced masonry constructions due tomasonry’s almost total lack of tensile strength brought to light theurgent need to improve and develop better methods of retrofittingfor existing seismically inadequate buildings. Several conventionaltechniques are available to improve the seismic performance of

existing URM walls. Surface treatments, grout injections [1–3],external reinforcement [4–7] are examples of such conventionaltechniques.

Some strengthening techniques, such as injections of cement- orlime-based grout and ferrocement, which was widely used in thereconstruction work following the earthquakes in Friuli (Italy) andSlovenia (1976), Thessaloniki (Greece) (1978), Irpinia (Italy)(1980), Erzurum (Turkey) (1983) and Umbria-Marche (Italy)(1997), have shown problems over time in terms of effectivenessand durability [8,9]. The most recent research studies have thus beendirected towards different techniques, such as those which providefor bonding with composites reinforced with FRP (Fiber ReinforcedPolymers) materials [10–15]. However, the increase in the in-planelateral resistance of URM walls upgraded with FRP was determinedto be less significant than the increase in the out-of-plane resistance[16–19].

Despite the many advantages associated with the use ofFRPs, the relevant strengthening techniques are not entirely

Fig. 1. Wall strengthened with ‘‘Reticolatus’’.

NutLocking washer

Stainlesssteel cord

New mortarStainlessbar

Fig. 2. Detail of a through connectors.

Stainlessbar

Mortar orresin

NutLocking washer

Stainlesssteel cord

New mortar

Fig. 3. Detail of a not-through connector.

Table 1Stainless steel cord mechanical properties.

Filament diameter (mm) 0.33Number of filaments per cord 49Net cross section area (mm2) 4.19Failure tensile load (kN) 6.11Young’s modulus (GPa) 81.5Failure stress (MPa) 1458Cord diameter (mm) 3

64 A. Borri et al. / Construction and Building Materials 56 (2014) 63–73

problem-free. Some drawbacks are attributed to the organic resinsused to bind or impregnate the fibers: difficulty in removal of rein-forcement, poor behavior of epoxy resins at temperatures abovethe glass transition temperature, high cost of epoxies, potentialhazards for the manual worker [19]. The use of epoxy resinsalso prevents water–vapor permeability and its fire resistanceis very low. In many cases heritage conservation authorities(Soprintendenze in Italy, English Heritage in England, CRMH inFrance, etc.) do not permit an extensive use of epoxy adhesiveson historical listed buildings or monuments.

One possible solution to the above problems would be thereplacement of epoxy resins with inorganic ones, e.g. lime orcement-based mortars, leading to the replacement of FRP withFiber Reinforced Mortars (FRM) [6,20–21]. This was made with aGFRP grid embedded into an inorganic lime-based mortar. Thecompatibility of inorganic matrices with historic masonry is extre-mely high: inorganic matrices are usually similar in compositionand mechanical properties to historic lime mortars.

However the need to maintain the exposed look of the wall fac-ings, frequently the case in work done on historic-monumentalbuildings, greatly limits the choice of reinforcement methods.

In earlier studies Borri et al. [22] proposed a system called‘‘Reticolatus’’, which consists of the inserting into the mortar jointsof a continuous mesh of thin stainless high strength steel cords, theflexibility of which allows reinforced repointing also for irregularstone-masonry. Along with investigating more in depth strength-ening with the ‘‘Reticolatus’’ technique on both faces, the tests pre-sented here made it possible to evaluate the effectiveness of ahybrid type of strengthening obtained with joint repointing onone face of the masonry and applying GFRM jacketing on the other.This technique is proposed as a solution in cases where one of thetwo wall surfaces must be preserved but a significant increase instrength is still required.

2. Reinforcement details

2.1. Strengthening with ‘‘Reticolatus’’

The strengthening technique known as ‘‘Reticolatus’’ consists ofthe inserting in the mortar joints, stripped to a depth of 50–60 mm,of a continuous mesh made from stainless high strength steelcords, the nodes of which, generally one every two, are connectedto the other face of wall by means of transverse stainless steel bars,in a number of 5 per m2 according to the scheme in Fig. 1. Thecords are arranged in vertical and horizontal directions, formingapproximately square meshes, the dimensions of which, normally300–500 mm wide, depend on the size of the stone in the masonry,and as a rule must not be greater than the thickness of the wall.

The cords are connected to the transverse bars by means ofeyelets in which the cords can slide: thus it is possible to apply amoderate tension, so as to make the mesh immediately functional.When the size of the stone walling material is such as to prevent theuse of bars passing through the entire thickness of the masonry, theconnection can be made with bars about 2/3 as long as the thick-ness of the wall, anchored with the injecting of mortar or epoxy re-sin (Figs. 2 and 3). The final application of mortar, which completelycovers both the cords and the heads of the transverse bars, makes itpossible to preserve the fair-faced aspect of the masonry.

The stainless cords are made by twisting 49 0.03 mm-diameterindividual filaments together made of AISI 316 steel. The averagecord diameter is 3 mm and the specifications of the single cordare shown in Table 1. The mechanical properties of the metal cordswere verified by tensile tests carried out on 10 samples. The resultssubstantially confirmed the values given by the manufacturer onthe technical sheet, with small variations of the failure load(6.11 kN) and of Young modulus (81.5 GPa).

A. Borri et al. / Construction and Building Materials 56 (2014) 63–73 65

The lime mortar used for joint repointing is composed by a bin-der of natural hydraulic lime. The compressive strength of suchmortar, measured on 100 mm-diameter cylinders (height200 mm) after 28 days of curing, was found equal to 8.90 MPa.

Fig. 5. Detail of the GFRP mesh.

Table 2Mechanical characteristics of glass FRP mesh.

Horizontal direction (weft) Tensile strength (MPa) 530Sample size 10Cross section (mm2) 7.29Elongation at failure (%) 1.73Young modulus (GPa) 36.1

Vertical direction (warp) Tensile strength (MPa) 680Sample size 10Cross section (mm2) 9.41Elongation at failure (%) 1.93Young modulus (GPa) 39.8

Weight density [kg/m2] 0.5

2.2. Hybrid strengthening with ‘‘Reticolatus’’ and GFRM jacketing

This strengthening technique combines the ‘‘Reticolatus’’system on one wall face with the application of a coating about20–30 mm thick reinforced with a GFRM grid on the other face(Fig. 4). A single layer of glass textile was used for panelreinforcement.

The GFRP mesh used in this research program was manufac-tured using AR-glass (Alkali-Resistant) fibers and an epoxy resin.Specimens extracted from the composite mesh has been measuredwith a tensile modulus ranging from 36.1 to 39.8 GPa. The meshhas a cross section area of 7.29 mm2 and 9.41 mm2 respectivelyin the horizontal (weft) and vertical (warp) direction and has anopening of 99 mm in both directions (Fig. 5). The main mechanicalcharacteristics of GFRP material measured via tensile tests, areshown in Table 2.

The choice of the mortar to be used for jacketing is a difficulttask; in fact, to have good durability, this mortar must be compat-ible with the existing masonry from the physical, chemical andmechanical points of view. The mortar should be strong enoughbut not too stiff, and have good bond with the stones and withthe existing mortar. For these reasons we used different kinds ofmortars for panel construction, joint repointing and GFRMjacketing.

Brick- and stone-panels were reinforced using a weak hydrauliclime. The average 60-day compressive strength of mortar obtainedfrom 6 compression tests on cylindrical sample (100 mm diameter,200 mm height) was 4.71 MPa (st. dev. 0.31 MPa) (Table 4). Pebblestone-panels were reinforced using a weaker hydraulic mortarwith a 60-day compressive strength of 4.24 MPa (st. dev.0.24 MPa). ASTM C780 was used as a reference for compressiontests [23].

The two reinforced faces are connected to each other by meansof transverse threaded stainless steel bars having a diameter of8 mm. Metal cords are passed through a ring at the end of the con-nector, so that by tightening a nut at the opposite end, it is possibleto lightly pretension the stainless steel mesh (Fig. 6).

When applied on regular brick walls, ‘‘Reticolatus’’ is made byplacing cords in the horizontal mortar joints (typically every

Fig. 4. Hybrid strengthening with ‘‘Reticolatus’’ and GFRP jack

three). The horizontal cords can be connected to each other bypairs of vertical cords spaced about 800 mm apart from each other.Transverse connectors having specifications similar to those usedin applications on irregular stone masonry are placed where thevertical and horizontal reinforcements intersect.

3. Methodology

3.1. Test specimens

Twenty-three diagonal compression tests were carried out inthe laboratory on samples with nominal dimensions of1200 � 1200 mm, which differed from each other in wall type

eting for stone-masonry panels (a) and brick-panels (b).

Stainlesssteel bar

Localreinforcement

WasherNut

GFRP mesh

Jacketing

Stainlesssteel cord

New mortar

Stainlesssteel bar

Fig. 6. Detail of a through connector used for hybrid reinforcement.

66 A. Borri et al. / Construction and Building Materials 56 (2014) 63–73

and reinforcement configuration (Table 3). More specifically, testswere done on:

8 panels of double-leaf rough hewn stone masonry, with athickness of 400 mm;7 panels of masonry in solid brick with a thickness of 250 mm(header bond);8 panels of uncut rounded stones (pebbles), with a thickness of400 mm.

In order to verify the repeatability of the results, two sampleswere provided for each wall type/reinforcement configurationcombination.

To evaluate the possible differences in behavior connected withthe number of transverse connectors on the nodes of the‘‘Reticolatus’’ mesh, for both stone and pebble masonry, two pairsof samples were prepared: the first was characterized by connec-tions passing through the entire thickness of the wall together withconnections not passing through on the face reinforced with‘‘Reticolatus,’’ the second only with connections passing through.The two reinforcement configurations are shown in Fig. 7.

To quantify the contribution made by the reinforcement in thevertical joints of the brick masonry samples, one pair of sampleswas strengthened placing the stainless steel cords only in the

Table 3Mortar types used for panel construction and reinforcement.

Panel No. Mortar used for panelconstruction

Mortar usedfor jacketing

Mortar used forjoint repointing

MP-1 AA – –MP-2 AA – –MP-1-I-N AA NHL6 M8MP-2-I-N AA NHL6 M8MP-1-I-P AA NHL6 M8MP-2-I-P AA NHL6 M8MP-1-R-N AA – M8MP-2-R-N AA – M8

MD-1 AA – –MD-2 AA – –MD-1-I-V AA NHL6 M8MD-2-I-V AA NHL6 M8MD-1-I-H AA NHL6 M8MD-2-I-H AA NHL6 M8MD-1r-I-V AA NHL6 M8

MC-1 AB – –MC-2 AB – –MC-1-I-P AB NHL5 M8MC-2-I-P AB NHL5 M8MC-1-I-N AB NHL5 M8MC-2-I-N AB NHL5 M8MC-1-R-P AB – M8MC-2-R-P AB – M8

horizontal joints, and the other inserting the cords also in the ver-tical joints, as shown in Fig. 8.

Each test is identified with a code of four indices, the first ofwhich indicates the masonry type (MP = stone-masonry,MD = brick-masonry, MC = uncut rounded stones (pebbles)), thesecond the progressive number identifying the panel, the thirdthe type of shear strengthening done (I = Hybrid, R = Reticolatus)and lastly the fourth the strengthening layout (P = with throughconnectors, N = with both through and un-through connectors,H = Reticolatus made of only horizontal cords, V = Reticolatusmade of only horizontal or vertical cords).

The panel brick panel No. 1 was subjected to mechanical testbefore (MD-1) and after repair (MD-1r-I-V) in order to determinethe shear strength.

In-plane resistance of un-reinforced masonry walls is based onmortar strength and masonry unit proportions. The properties de-pend on the characteristics of constituents (bricks/stones and mor-tar) and on their interaction and the mechanical properties ofconstituents vary in function of their nature (solid bricks, stones,composition of mortar, etc.). In this work in order to simulate a his-toric lime-based mortar stone and brick panels were built with ahydraulic lime-based mortar (AA) characterized with low mechan-ical properties (Table 4). For pebble stone panels a weaker lime-based mortar (AB) has been used.

3.2. Test set up

The diagonal compression test was designed in order to evalu-ate the shear strength, the shear elastic modulus and the ductilityof the masonry [24,25]. The diagonal test was carried out on panels1200 x 1200 mm.

The test mechanism is composed of a set of metallic elementsfixed at the two corners of a diagonal of the panel. A jack, placedat one corner, is interposed between two metallic elements whichpermit it, on the one hand, to act directly on a corner of the panel,while at the same time being rigidly connected to an analogousmetal element located at the opposite corner. A closed system is

Fig. 7. Configuration of specimens and strengthening (stone-masonry panels).

Fig. 8. Configuration of specimens and strengthening (brick-masonry panels).

Table 4Mortar mechanical properties.

Type Ratio by volume ofhydraulic lime (sand,water)

Youngmodulus(GPa)

Compressionstrength (MPa)

AA Lime-based

2.3, 0.42 – 1.87

AB Lime-based

3, 0.45 – 0.96

M8 Lime-based

Ready-to-use – 8.90

NHL5 Lime-based

2, 0.42 14.53 4.24

NHL6 Lime-based

Ready-to-use 15.09 4.71

GFRPJacketing

ReticolatusLoad

Transducers

Transducers

Fig. 9. Load and displacement transducer positions.

Table 5Test results.

Panel No. Reinforcementtype

LoadPmax

(kN)

Tensilestrength ft

(MPa)

Shear strengths0 (MPa)

s0,R/s0,UR

MP-1 URM 161.6 0.176 0.117 –MP-2 URM 107.0 0.121 0.081 –MP-1-I-N Hybrid 246.2 0.262 0.175 1.76MP-2-I-N Hybrid 273.7 0.298 0.198 2.01MP-1-I-P Hybrid 233.3 0.254 0.169 1.71MP-2-I-P Hybrid 284.3 0.296 0.197 1.99MP-1-R-N Reticolatus 162.4 0.175 0.117 1.18MP-2-R-N Reticolatus 158.5 0.171 0.114 1.15

MD-1 URM 120.6 0.209 0.139 –MD-2 URM 108.2 0.187 0.124 –MD-1-I-V Hybrid 135.3 0.222 0.148 1.12MD-2-I-V Hybrid 197.6 0.332 0.222 1.68MD-1-I-H Hybrid 192.0 0.318 0.212 1.61MD-2-I-H Hybrid 162.4 0.266 0.178 1.35MD-1r-I-V Hybrid 127.1 0.216 0.144 1.04

MC-1 URM 46.7 0.052 0.034 –MC-2 URM 49.8 0.055 0.037 –MC-1-I-P Hybrid 132.5 0.131 0.088 2.47MC-2-I-P Hybrid 109.1 0.110 0.074 2.07MC-1-I-N Hybrid 108.5 0.105 0.070 1.97MC-2-I-N Hybrid 107.5 0.104 0.069 1.96MC-1-R-P Reticolatus 63.8 0.070 0.047 1.32MC-2-R-P Reticolatus 69.7 0.077 0.051 1.44

A. Borri et al. / Construction and Building Materials 56 (2014) 63–73 67

obtained in which the jack stretches the panel along one of twodiagonals. The panels rest on the lab ground and it has been as-sumed that the boundary conditions do not affect the shearstrength increment of reinforced panels compared to unreinforcedones.

Both diagonals of the panel were instrumented on both sideswith four (potentiometers) transducers. Panel out-of-planedisplacements due to dissymmetric hybrid reinforcements weremeasured and plotted in real-time under measured load with threehorizontal potentiometers: two in the wall panel center and one onthe top unloaded edge (Fig. 9). The total number of the channels of

acquisition was nine (seven transducers, jack’s force, time) and thefrequency of acquisition was equal to 10 Hz. The test consisted inequal couples of cycles of loading and un-loading, with increasesof 20–25 kN, up to the point of failure.

4. Analysis of panels

The hypothesis made is that the system can be studied as anelastic problem of an in-plane loaded slab with two forces P actingon two opposite edges of a diagonal, the tensile strength of the ma-sonry was determined:

ft ¼ aPmax

Anð1Þ

where An is the area of the horizontal section of the masonry panelconsidered net of the increase in thickness due to the application ofthe jacketing. For all the panels tested parameter a was assumed tobe equal to 0.5, in accordance with the indications of the RILEMstandards [25]. The choice not to differentiate the value of a accord-ing to the wall type, unlike that proposed in Brignola et al. [26],arises from the need to determine the tensile strength of reinforcedmasonry to which it is not possible to extend the considerationspresented in that article.

The shear strength value s0 was evaluated according to the for-mula proposed by Turnšek and Cacovic [27]:

s0 ¼ft

1:5ð2Þ

Table 5 summarizes the results obtained from the tests givingthe maximum load (Pmax), the tensile strength (ft), the shearstrength (s0) and the ratio between the shear strength of thereinforced sample and the mean of the unreinforced samples(s0,R/s0,UR).

Using lc,A to indicate the initial length of the measurement baseof the transducer on the diagonal on the A panel face and Dlc,A toindicate the relative shortening in length (Dlt,A is the relativeelongation) the average compressive strain ec and tensile strain et

and shear strain c are respectively defined as:

Table 6Out-of-plane deflections at maximum load Pmax.

Pmax (kN) Deflection at Pmax (mm)

MP-1-I-N 246.2 7.00MP-2-I-N 273.7 7.20MP-1-I-P 233.3 8.90MP-2-I-P 284.3 5.00

MD-1-I-V 135.3 1.00MD-2-I-V 197.6 0.50MD-1-I-H 192.0 0.80MD-2-I-H 162.4 0.20MD-1r-I-V 127.1 19.40

MC-1-I-P 132.5 9.40MC-2-I-P 109.1 7.10MC-1-I-N 108.5 10.90MC-2-I-N 107.5 9.10

Fig. 10. Unreinforced brick panel: crack pattern.

Fig. 11. Unreinforced pebble s

68 A. Borri et al. / Construction and Building Materials 56 (2014) 63–73

ec ¼12

Dlc;Alc;Aþ Dlc;B

lc;B

� �ð3Þ

et ¼12

Dlt;A

lt;Aþ Dlt;B

lt;B

� �ð4Þ

c ¼ jecj þ jetj ð5Þ

For the panels with different reinforcement on the two faces,Table 6 gives the values of the out-of-plane displacement of theupper vertex of the diagonal in tensile of the face reinforced withGFRM jacketing when reaching the maximum load.

For the stone and pebble masonry panels with hybrid strength-ening, different displacement readings were found with the trans-ducers applied on the diagonals of the two faces. However, it wasdecided to consider a mean value and then to process the acquireddata according to the formulations (1)–(5) by reason of the factthat the out-of-plane displacement of the panel was small up toreaching the maximum load. For the brick masonry panels againwith hybrid strengthening, the eccentric response was negligible,as the in-plane stiffness of the masonry and that of the jacketingapplied are comparable.

The unreinforced panels of each wall type had diagonal crackingthat involved only the mortar joints (Figs. 10 and 11). Upon re-moval of the load at the end of the test, the pebble panels werecompletely disjointed, providing evidence of the weakness typicalof this wall type. The tests on panels reinforced with ‘‘Reticolatus’’on both faces showed that, by holding together the various partsmaking up the sample, this strengthening technique is able toovercome this problem (Fig. 12). The panels with hybrid strength-ening also showed diagonal cracking which, however, is morewidespread. Figs. 13–15 gives the envelope curves of the diagonalload vs. shear strain diagrams for the three wall types tested.

Strengthening with ‘‘Reticolatus’’ on both faces brings about anincrease in strength of 17% for stone masonry and 40% for pebblemasonry. The increases for hybrid strengthening instead were87%, 44% and 112% for stone, brick and pebble masonry, respec-tively. Thus it is clear that the increase in strength following theapplication of strengthening techniques depends on the mechani-cal characteristics of the masonry being reinforced, becoming morenoticeable with the decrease in the strength before reinforcement,as in the case of pebble masonry.

In order to quantify the improvement in the post-failure behav-ior of reinforced panels, parameter K was used to evaluate the load

tone panel: crack pattern.

A. Borri et al. / Construction and Building Materials 56 (2014) 63–73 69

value that the panel is able to support upon reaching a givendeformation:

K ¼ P10cy

Pmaxð6Þ

where P10cy is the residual load corresponding to a shear strainequal to 10 times the deformation cy determined by the intersectionbetween the horizontal line to shear load vs. strain curve at thepoint of maximum load and the secant line at 70% of the maximumload (Fig. 16).

Fig. 12. Reinforced panels: crack pattern (stone panels: (a),

As shown in Figs. 13–15, there is a considerable decrease instrength following the formation of cracks for brickwork masonryfor unreinforced panels, while it is less for the other two types,which show a good dissipation capacity. The panels strengthenedwith the same technique on both faces and those with hybridstrengthening both showed a marked improvement in post-peakbehavior, with the ultimate load value remaining almost unalteredwith the increase in tangential deformation.

In panels with hybrid strengthening on irregular masonry theaddition of connections not passing through the nodes of the

(e) and (f); brick panels: (c) and (d), pebble panel (b)).

70 A. Borri et al. / Construction and Building Materials 56 (2014) 63–73

‘‘Reticolatus’’ mesh did not cause any significant difference eitherin terms of strength increases or improvement of post-peakbehavior (Tables 5 and 7). For stone panels (MP) with not passing

Fig. 13. Envelope curves of the load–angular s

Fig. 14. Envelope curves of the load–angular

Fig. 15. Envelope curves of the load–angular s

connections an average increase of shear strength of 88.5% hasbeen measured, compared to 81% for panels with passing connec-tions. A similar observation can be made for hybrid strengthening

train response for stone masonry panels.

strain response for brick masonry panels.

train response for pebble masonry panels.

Fig. 16. Parameter definition.

A. Borri et al. / Construction and Building Materials 56 (2014) 63–73 71

tested on pebble stone panels (94.5% and 99.5% for panels withpassing and not passing connections, respectively). However, theinfluence of these additional connections should be verified inexperimental tests that call into question the flexural behavior ofthe masonry.

5. Design cosiderations

A comparison between experimental results and the predictionsobtained by existing analytical formulations was made to evaluatethe possibility of the use of such formulations in predicting theshear behavior of strengthened panels. Even though, in fact,the proposed formulations are not specifically introduced for thestrengthening techniques proposed in this paper (formulation (8)is proposed for masonry reinforced with steel bars, whereas formu-lation (7) and (9) are introduced specifically for FRM), at present nomore appropriate approaches are available. In such a context theproposed comparison could represent a first step for the develop-ment of code recommendations for the design of strengtheningof masonry panels.

Table 7Post peak behavior and residual compressive strength of masonry panels.

Panel No. cPmax (‰) c0,7 (‰)

MP-1 0.324 0.105MP-2 0.580 0.117MP-1-I-N 4.958 0.484MP-2-I-N 5.337 0.623MP-1-I-P 6.102 0.594MP-2-I-P 3.720 0.421MP-1-R-N 3.184 0.616MP-2-R-N 5.646 0.642

MD-1 0.184 0.069MD-2 0.136 0.081MD-1-I-V 2.152 0.466MD-2-I-V 0.601 0.076MD-1-I-H 1.192 0.024MD-2-I-H 0.341 0.157MD-1r-I-V 9.486 0.864

MC-1 0.270 0.072MC-2 0.664 0.197MC-1-I-P 7.093 0.484MC-2-I-P 4.817 0.930MC-1-I-N 5.920 0.547MC-2-I-N 6.193 0.392MC-1-R-P 6.599 0.395MC-2-R-P 5.295 0.531

According to this, the maximum strength obtained during thetests was compared with those given by the formulas proposedby CNR-DT 200/2004 [28]:

Vcr ¼fvktdcRdcm

þ 0:6Ar;hd0fyk

cRdð7Þ

by Tomazevic [29]:

Vcr ¼ 0:9tlf �vk0

b

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi1þ r0

f �vk0

sþ 0:4Ar;hfyk ð8Þ

and by Triantafillou [10]:

Vcr ¼ fvktdþ 0:9dqr;hErrer;ut ð9Þ

where

rer;u ¼ er;e ¼ 0:0119� 0:0205ðqr;hErÞ þ 0:0104ðqr;hErÞ2 ð10Þ

In all listed design models, the total shear load of the strength-ened specimens is assembled from two components: the shearload carried by equivalent unreinforced masonry specimen andthe shear load contribution of strengthening.

Specifically, the symbols used in the formulations are as fol-lows: Vcr is the critical lateral load; t is the panel thickness, l itslength, d = 0.8l is the effective depth; d0 is the distance betweenthe compression side of the masonry and the centroid of GFRMflexural strengthening; b is the shear stress distribution coefficient(= 1.5 for parabolic distribution); cm (= 1) and cRd (= 1.2) arecoefficients from CNR-DT 200/2004; r0 is the design compressivestress; qr,h is the horizontal reinforcement ratio computed on thepanel section; Ar,h is the area of the horizontal reinforcement, fyk

is the characteristic tensile stress of the reinforcement, Er is thereinforcement elastic modulus, er,u is the reinforcement tensileultimate strain.

Note that the characteristic shear strength of masonry (fvk) canbe expressed in the following way:

fvk ¼ fvk0 þ 0:4r0 ð11Þ

Conversely, the value (fvk0) of the characteristic initial shearstrength has been obtained by diagonal compressive tests.

cy (‰) P10cy (kN) K (–)

0.150 144.0 0.890.167 97.9 0.910.691 243.0 0.990.889 269.4 0.980.848 229.7 0.980.601 269.9 0.950.879 144.4 0.890.918 157.4 0.99

0.099 102.2 0.850.115 79.5 0.730.665 124.0 0.920.108 178.6 0.900.034 185.6 0.970.224 131.8 0.811.235 120.1 0.94

0.103 42.6 0.910.281 43.4 0.870.691 131.5 0.991.328 98.6 0.900.781 108.4 1.000.560 106.4 0.990.564 61.3 0.960.758 67.4 0.97

Fig. 17. Comparison of masonry shear strength and contribution of strengthening by different authors for the set of performed tests.

72 A. Borri et al. / Construction and Building Materials 56 (2014) 63–73

Also, the formulation (9) considers a factor of efficiency r, whichdepends on the failure mode (reinforcement rupture ordebonding). The expression of r, given by the Eq. (10), was foundby Triantafillou [10] for concrete members (er,e is the effectivereinforcement strain).

In the chart in Fig. 17 the experimental results are comparedwith calculations by different approaches. In each bar the contribu-tion of plain masonry (bottom part) and reinforcement (upperparts) to shear force of strengthened wall can be observed.

On average the CNR and Tomazevic approach gives the closestvalues to experimental values. It is possible to observe that theerror of the model varied in fact from 10% (formula proposed byCNR) to 15% (formula proposed by Tomazevic). A single exceptionwas in the case of MP-I panels where the formula proposed byTriantafillou provided the best fit (�15%), whereas CNR andTomazevic proposal grossly underestimates the strength ofreinforced panels (�35%). Furthermore, the CNR approach leads,in almost all cases, to conservative values when comparing withthe experimental results (other approaches tends to overestimatesthe strengthening contribution).

6. Concluding remarks

Testing of stone- and brick- wall panels reinforced with GFRMand/or joint repointing reinforced with high strength steel cordson the external surfaces has validated the simplicity of its applica-tion and the effectiveness of this strengthening system as itnotably improves the mechanical behavior of the structure interms of lateral load and deformation capacity.

The large number of diagonal compression tests performed inthe laboratory on full-scale wall panels of rough-hewn stone, brickand river pebble masonry have shown that the technique proposedin this paper bring about substantial increases in shear strength.However further in-site experimental activity is needed in orderto confirm positive results found in laboratory.

The flexural behavior in the panels with hybrid strengthening,although depending on the ratio between the stiffness of the jac-keting applied and the support masonry, was nonetheless foundto be limited and did not preclude the achieving of the objectivesof improving the shear strength.

Design equations proposed by Italian CNR for wall panels rein-forced by GFRM grids and Reticolatus were in acceptable agree-ment with experimental results. Therefore, until new approachesto predict the shear capacity are available, the formula proposedby CNR appears to be adequate for design purposes, showinghow the same approach used for masonry reinforced with FRPcan be satisfactory used for GFRP mesh and Reticolatus.

Acknowledgments

This project was sponsored by the Italian Ministry of Education[ReLUIS (2010-2013) Linea 2.3.1]. Special thanks are directed toprof. Natalino Gattesco, dr. Allen Dudine and dr. Andrea Cernigoi.The authors would like to acknowledge Fibre Net S.r.l for financialsupport and material supply.

References

[1] Binda L, Modena C, Baronio G, Gelmi A. Experimental qualification of injectionadmixtures used for repair and strengthening of stone masonry walls. In:Proceedings of the 10th international brick/block masonry conference,Calgary; 1994.

[2] Binda L, Modena C, Baronio G, Abbaneo S. Repair and investigation techniquesfor stone masonry walls. Constr Build Mater 1997;11(3):133–42.

[3] Van Rickstal F. Grout injection of masonry, scientific approach and modeling.Internationale Zeitschrift für Bauinstandsetzen und Baudenkmalpfkege2001;7(3/4):407–32.

[4] Modena C. Repair and upgrading techniques of unreinforced masonrystructures utilized after the Friuli and Campania–Basilicata earthquake.Earthquake Spectra 1994;10(1):171–85.

[5] ElGawady M, Lestuzzi P, Badoux M. A review of conventional seismicRetrofitting techniques for URM. In: 13th International brick and blockmasonry conference amsterdam; July 4–7, 2004.

[6] Corradi M, Borri A, Vignoli A. Experimental evaluation of the in-plane shearbehaviour of masonry walls retrofitted using conventional and innovativemethods. J British Masonry Soc 2008;21(1):29–42.

[7] Ashraf M, Khan AN, Naseer A, Ali Q, Alam B. Seismic behavior of unreinforcedand confined brick masonry walls before and after ferrocement overlayretrofitting. Int J Archit Heritage 2012;6:665–88.

[8] Valluzzi MR. On the vulnerability of historical masonry structures: analysisand mitigation. Mater Struct, Springer 2007;40:723–43.

[9] Carocci CF. Small centres damaged by 2009 L’Aquila earthquake: on siteanalyses of historical masonry aggregates. Bull Earthquake Eng2012;10:45–71.

[10] Triantafillou TC. Strengthening of masonry laminates using epoxy-bonded FRPlaminates. J Compos Constr 1998;2(2):96–104.

A. Borri et al. / Construction and Building Materials 56 (2014) 63–73 73

[11] Valluzzi MR, Tinazzi D, Modena C. Shear behavior of masonry panelsstrengthened by FRP laminates. Constr Build Mater 2002;16:409–16.

[12] ElGawady M, Lestuzzi P, Badoux M. In-plane response of URM walls upgradedwith FRP. J Compos Constr 2005;9(6):524–35.

[13] Al-Salloum YA, Almusallam TH. Walls strengthened with epoxy-bonded GFRPsheets. J Compos Mater 2005;39:1719–45.

[14] Santa Maria H, Alcaino P, Luders C. Experimental response of masonry wallsexternally reinforced with carbon fiber fabrics. In: Proceedings of the 8th USnational conference on earthquake engineering; 2006. Paper no. 1402.

[15] Corradi M, Borri A, Vignoli A. Strengthening techniques tested on masonrystructures struck by the Umbria–Marche earthquake of 1997–1998. ConstrBuild Mater 2002;16:229–39.

[16] Roca P, Araiza G. Shear response of brick masonry small assemblagesstrengthened with bonded FRP laminates for in-plane reinforcement. ConstrBuild Mater 2010;24:1372–84.

[17] Borri A, Castori G, Corradi M, Speranzini E. Shear behavior of unreinforced andreinforced masonry panels subjected to in situ diagonal compression tests.Constr Build Mater 2011;25(2011):4403–14.

[18] Ehsani MR, Saadatmanesh H, Velazquez-Dimas JI. Behavior of retrofitted URMwalls under simulated earthquake loading. J Compos Constr 1999;3(3):134–42.

[19] Papanicolaou CG, Triantafillou TC, Papathanasiou M, Karlos K. Textilereinforced mortar (TRM) versus FRP as strengthening material of URM walls:out-of-plane cyclic loading. Mater Struct 2008;41:143–57.

[20] Costa AA, Arêde A, Costa A, Oliveira CS. In situ cyclic tests on existing stonemasonry walls and strengthening solutions. Earthquake Eng Struct Dyn2011;40:449–71.

[21] Prota A, Marcari G, Fabbrocino G, Manfredi G, Aldea C. Experimental in-planebehavior of tuff masonry strengthened with cementitious matrix-gridcomposites. J Compos Constr 2006;10(3):223–33.

[22] Borri A, Corradi M, Speranzini E, Giannantoni A. Reinforcement of historicmasonry with high strength steel cords. Masonry Int 2010;23(3):79–90.

[23] ASTM C780-12a. Standard test method for preconstruction and constructionevaluation of mortars for plain and reinforced unit masonry.

[24] ASTM E519-2002. Standard test method for diagonal tension (Shear) inmasonry assemblages.

[25] RILEM TC 76-LUM. Diagonal tensile strength tests of small wall specimens,1991. In: RILEM, Recommendations for the Testing and Use of ConstructionsMaterials. London: E& FN SPON; 1994, p. 488–89.

[26] Brignola S, Frumento S, Lagomarsino S, Podestà S. Identification of shearparameters of masonry panels through the in-situ diagonal compression test.Int J Archit Heritage 2009;3:52–73.

[27] Turnšek V., Cacovic F. Some experimental results on the strength of brickmasonry walls. In: Proceedings of the 2nd international brick masonryconference, Stoke-on-trent; 1970, p. 149–56.

[28] CNR. Guide for the design and construction of externally bonded FRP systemsfor strengthening of existing structures. In: Technical document no. 200/2004.Rome: Italian National Research Council (CNR); 2004.

[29] Tomazevic M, Lutman M, Petrovic L. In plane behavior of reinforced masonrywalls subjected to cyclic lateral loads. In: Report to the Ministry of Science andTechnology of Republic of Slovenia, Liubljana, Slovenia; 1993.