You are on page 1of 8

Construction and Building Materials 16 (2002) 409–416

Shear behavior of masonry panels strengthened by FRP laminates


M.R. Valluzzi*, D. Tinazzi, C. Modena
Department of Constructions and Transportation Engineering, University of Padova, Via Marzolo, 9-35131 Padova, Italy

Received 6 July 2001; accepted 31 May 2002

Abstract

The present experimental study, performed on brick masonry panels strengthened by Fiber Reinforced Polymer (FRP) laminates,
was aimed to investigate the efficiency of an alternative shear reinforcement technique. A series of nine unreinforced masonry
(URM) panels and 24 strengthened panels have been subjected to diagonal compression tests. Different reinforcement
configurations were evaluated. Experimental results pointed out that FRP reinforcement applied only at one side of the panels did
not significantly modify the shear collapse mechanisms (diagonal splitting) of the URM; while double-side configurations
provided a less brittle failure and a noticeable ultimate capacity increase. Performances of the different reinforcement configurations
are compared in terms of strength and mechanism of failure; finally, experimental results are also used to calibrate existing
analytical formulations for ultimate shear strength prediction.
䊚 2002 Elsevier Science Ltd. All rights reserved.

Keywords: Brick masonry; Shear behavior; FRP laminates

1. Introduction behavior of the strengthening material itself; therefore,


redistribution-derived theories are not applicable. Con-
The Italian traditional masonry works, which represent sequently, investigations on alternative mechanisms pro-
a large part of its historical heritage, are particularly viding sufficient signals of incipient collapse are
susceptible to in-plane shear actions. required.
As is known, in URM brittle, shear rupture occurs A certain number of FRP masonry strengthening
either as a diagonal splitting or as step-pattern sliding applications have already been performed, involving
along the mortar joints, depending on the characteristics either FRP bars and laminates, but few analytical or
of the constituent materials (mortar and bricks). There- experimental research works have investigated the
fore, in order to predict properly the masonry shear effectiveness and reliability of that new technology w2–
capacity, it is necessary to first identify the most antic- 4 x.
ipated failure mechanism, based on the knowledge of In the present experimental work, which was per-
the involved materials. formed on coupon-size masonry panels, the diagonal
Presently, small diameter steel bars are successfully compressive test has been chosen to simply simulate the
used as reinforcement in masonry retrofitting, and sev- in-plane shear phenomenon.
eral field applications on heavily loaded historical struc-
tures have been presented w1x. 2. Materials characterization
Nowadays, FRPs represent a new opportunity to
restoring ambit, with considerable development in URM As mentioned, masonry mechanical properties depend
strengthening. A key problem is represented by FRP’s on the characteristics of the constituent elements (bricks
up-to-failure linear elastic behavior, which prevents the and mortar), as well as on the workmanship and the
ductility of the system being based on the plastic interface interaction within the assemblage. Thus, an
extensive program of tests, ranging from the units to the
*Corresponding author. Tel.: q39-049-827-5576; fax: q39-049- assemblages scale, was performed.
827-5604. The main mechanical properties of the bricks were
E-mail address: valluzzi@caronte.dic.unipd.it (M.R. Valluzzi). determined by unidirectional compressive and tensile

0950-0618/02/$ - see front matter 䊚 2002 Elsevier Science Ltd. All rights reserved.
PII: S 0 9 5 0 - 0 6 1 8 Ž 0 2 . 0 0 0 4 3 - 0
410 M.R. Valluzzi et al. / Construction and Building Materials 16 (2002) 409–416

Table 1
Physical and mechanical characteristics of the FRP laminates

Type of fiber High tensile Glass Polyvinyl-


strength type E alcohol
carbon (GFRP) (PVAFRP)
(CFRP)
Density (kgym3) 1820 2600 1300
Equivalent thickness (mm) 0.165 0.115 0.07
Characteristic tensile strength (MPa) 3430 1700 1400
Tensile modulus of elasticity (GPa) 230 65 29
Ultimate strain (%) 1.5 2.8 6

tests. The characteristic compressive strength of five Therefore, the Coulomb equation representing the dry
specimens was found to equal 8.83 MPa. Indirect tensile friction mechanism during joint sliding is:
tests (splitting test) on five specimens provided an
average value equal to 0.95 MPa; thus, the correspond- fvsfv0qms0s0.66q1.36s0 (1)
ing tensile strength can be adopted to equal 0.76 MPa. Compression tests with load cycles performed on
Mortar used for the masonry panels had the following
masonry panels, having nominal dimensions of
mix composition: 380 kgym3 of inorganic binder; 1140
51=51=12 cm, revealed a characteristic compressive
kgym3 of sand (Dmaxs4.8 mm); and waterybinder
strength of 5.56 MPa, a modulus of elasticity of 1400
ratios0.7. The flexural and compressive standard tests
MPa (referred to the 30% of the ultimate load) and a
on six specimens (40=40=160 mm3, after a period of
28 days of curing) revealed an average strength equal masonry Poisson ratio equal to nxzsy´x y ´zs0.03 and
to 1.48 (hence, the corresponding tensile is approx. 1.19 nyzsy´y y ´zs0.03 (referred to a transversal, x or y,
MPa) and 6.03 MPa, respectively. and the vertical direction z, respectively).
Interface friction characteristics along the mortar The FRP laminates involved in the experimental work
joints are referred to previous works involving the same consist in carbon, glass or polyvinyl-alcohol unidirec-
materials w3x. During those tests, 12 triplets where tested tional fibers embedded in epoxy resin, according with
with four different levels of confining stress (0.05, 0.1, the wet lay up technique. Their main mechanical char-
0.3 and 0.5 MPa). acteristics, declared by the manufacturer, are shown in
Interpolating the experimental points obtained on the Table 1.
friction vs. confining stresses, it was possible to detect, Pull-off test were performed in order to evaluate the
according with the Coulomb criterion, the parameters FRP-brick adhesion strength under actions orthogonal
describing the shear strength f v for sliding along a to the bond surface. The average tensile strength of six
mortar joint: the cohesion f v0 and the dry friction specimens is equal to 0.44 MPa; in all cases the rupture
coefficient m. was due to detachment of the brick superficial skin.

Fig. 1. Single-side strengthening patterns.


M.R. Valluzzi et al. / Construction and Building Materials 16 (2002) 409–416 411

Fig. 2. Double-side strengthening patterns.

Bond strength along the fiber direction was deter- of solid clay bricks (5.5=25=12 cm) and have 10-
mined by pulling two consecutive bricks connected by mm-thick mortar joints.
a stretched FRP strip. The measured average strength of Twenty-four of them have been strengthened by dif-
five specimens was equal to 2.50 MPa; again, detach- ferent FRP materials: nine with CFRP; 10 with GFRP
ment occurred within the brick superficial skin. and five with PVAFRP (see Fig. 1).
Due to the small size of the samples and the bricks
3. Specimen description
friableness, it was necessary to cut the loading diagonal
A series of 33 masonry panels having nominal dimen- corners, obtaining a plane transferring areas of 12=12
sions of 51.5=51=12 cm were built. They were made cm.

Table 2
Description of the specimens

Unreinforced panels PNR 1, PNR 2, PNR 3,PNR 4, PNR 5, PNR 6, PNR 7, PNR 8, PNR 9
Squared grid strengthening configuration panels
CFRP Double-side (1 layer of 1.2 cm) PR 1 Carb 2F
PR 2 Carb 2F
PR 3 Carb 2F
Single side (1 layer of 2.4 cm) PR 1 Carb 1F
PR 2 Carb 1F
GFRP Double-side (2 layers of 3 cm) PR 1 Glass 2F
PR 2 Glass 2F
PR 3 Glass 2F
Single side (4 layers of 3 cm) PR 1 Glass 1F
PR 2 Glass 1F
PR 3 Glass 1F
PVAFRP Double-side (4 layers of 5.5 cm) PR 1 PVA 2F
PR 2 PVA 2F
PR 3 PVA 2F
Single side (6 layers of 7.5 cm) PR 1 PVA 1F
PR 2 PVA 1F
Diagonal strengthening configuration panels
CFRP Double-side (1 layer of 1.7 cm) PRD 1 Carb 2F
PRD 2 Carb 2F
Single side (2 layers of 1.7 cm) PRD 1 Carb 1F
PRD 2 Carb 1F
GFRP Double-side (3 layers of 2.8 cm) PRD 1 Glass 2F
PRD 2 Glass 2F
Single side (6 layers of 2.8 cm) PRD 1 Glass 1F
PRD 2 Glass 1F
412 M.R. Valluzzi et al. / Construction and Building Materials 16 (2002) 409–416

Table 3
Experimental tests results

Panel type Cracking Crack Failure % Failure mode


Load pattern load
(kN) (kN)
Unreinforced masonry panels
PNR 1 – – 75.5 Splitting
PNR 2 – – 90.0 Splitting
PNR 3 – – 91.2 Splitting
PNR 4 – – 127.9 Splitting
PNR 5 – – 77.1 Splitting
PNR 6 – – 93.2 Splitting
PNR 7 – – 93 Splitting
PNR 8 – – 121.6 Splitting
PNR 9 – – 137.3 Splitting
Average 100.7 Ref.
Single-side reinforced panels
PR 1 Carb 1F – Sharp diag. 91.6 Splitting
PR 2 Carb 1F – Sharp diag. 90.0 Splitting
Average 90.8 y9.8
PR 1 Glass 1F – Sharp diag. 94.6 Splitting
PR 2 Glass 1F 86.0 Sharp diag. 99.1 Splitting
PR 3 Glass 1F – Sharp diag. 94.0 Splitting
Average 95.9 y4.8
PR 1 PVA 1F – Sharp diag. 102.0 Splitting
PR 2 PVA 1F – Sharp diag. 100.0- –
Average 101 q0.3
PRD 1 Carb 1F – Sharp diag. 112.8 Splitting
PRD 2 Carb 1F – Sharp diag. 119.8 Splitting
Average 116.3 q15.5
PRD 1 Glass 1F 113.8 Sharp diag. 115.3 Splitting
PRD 2 Glass 1F 108.1 sharp diag. 108.3 Splitting
Average 111.8 q11.0
Double-side reinforced panels
PR 1 Carb 2F – Spread diag. 100.5 DLM at lower corner
PR 2 Carb 2F 104.6 Spread diag. 113.3 DLM sup.qRTT centr.
PR 3 Carb 2F – Spread diag. 98.5 DLM lower, sides A and B
Average 104.1 q3.4
PR 1 Glass 2F – Spread diag. 106.7 DLM up. and low. corner
PR 2 Glass 2F 120.5 Spread diag. 135.4 RTT side A
PR 3 Glass 2F – Spread diag. 103.0 DLM low. side A and B
Average 115.0 q14.2
PR 1 PVA 2F – Spread diag. 148.5 DLM at lower corner
PR 2 PVA 2F 130.5 Spread diag. 150.4 DLM at lower corner
PR 3 PVA 2F – Spread diag. 146.0 DLM low. side A and B
Average 148.3 q47.3
PRD 1 Carb 2F – Spread diag. 143.1 DLM up. and low. side A
PRD 2 Carb 2F 125.3 Spread diag. 149.4 DLM inf.qRTT centr.
Average 146.3 q45.3
PRD 1 Glass 2F – Spread diag. 170.3 DLM low. side A
PRD 2 Glass 2F 133.2 Spread diag. 180.0 DLM inf. sides A and B q
RTT centr. side A
Average 175.2 q74.0

In order to study the influence of the eccentricity of Moreover, two configurations of the reinforcing sys-
the strengthening, the strips were applied on both sides tem were investigated: strips as grid arrangement or
or only at one side of the panels; in the latter case, the application of diagonal strips orthogonal to the loaded
FRP thickness was doubled to maintain the FRP amount diagonal. The same reinforcement amount was applied
constant. for the two different configurations, for each kind of
M.R. Valluzzi et al. / Construction and Building Materials 16 (2002) 409–416 413

4.2. Single-side strengthening

Splitting failure with a clear diagonal crack pattern


was also obtained in all single-side reinforced panels,
whereas ultimate load was in many cases lower than the
reference. The samples exhibited a clear bending defor-
mation during the loading phases along the unreinforced
diagonal; as a consequence, the main damage was
concentrated on the unreinforced side (see Fig. 3). That
bending phenomenon was caused by a noticeable differ-
ence of stiffness on the opposite sides as a result of the
asymmetrical reinforcement. Among the one-side rein-
forced specimens, diagonal strengthening configuration
always revealed a higher effectiveness than the squared
grid set-up.

4.3. Double-side strengthening

Fig. 3. Single-side reinforced panel failure mode: diagonal splitting In all these cases, the failure mechanism consisted in
with a single large crack on the unreinforced side. Notice the bending sudden loss of collaboration between reinforcement and
along the free diagonal. substrate, due to either de-lamination (peeling) of the
superficial part of masonry or rupture of the FRP strips
reinforcing material, so that the ‘stiffness by mechanical (see Fig. 4).
ratio’ rE can be maintained for the two mentioned Grid-reinforced specimens determined spread-cracks
arrangements. patterns, whereas a clear splitting crack appeared in all
The design reinforcement criterion of the FRP amount the diagonally reinforced panels. The ultimate strength
was based on the expectation of an increase of the 50% increase was noticeable in almost all cases; while only
of the URM ultimate shear strength by applying the the CFRP-reinforced panel was seriously affected by
principal tensile stress limitation method on the homog- de-lamination.
enized section (following the Frocht theory, as in Yokel The URM typical sudden failure was noticeably cor-
and Fattal w5x). rected by the FRP strengthening, especially by the grid
As a consequence, due to the different mechanical configuration, where crack wide spreading provided
characteristics of the fibers, each test condition is char- sufficient signals of incipient crisis well before collapse.
acterized by a different width of the strips and different Deformations increased visibly up to failure, and the
number of layers to be glued (see Figs. 1 and 2). The global behavior resulted was less brittle, as shown in
panel typologies are shown in Table 2. Figs. 5 and 6 (up to 80% of the ultimate load, when
sensors have been removed).
4. Experimental results

The samples were subjected to diagonal compression 5. Analysis of the results


test, and both vertical and horizontal deformations were
measured by displacement transducers. Quantitative An FEM simulation was implemented by means of a
results are summarized in Table 3, where abbreviations commercially available code to validate the assumptions
are as follows: DLM means de-lamination of FRP on stress distributions corresponding to the different
stripes, RTT means FRP tensile rupture; the location of strengthening configurations of the panels. Results are
the possible concentration of the rupture is also indicat-
ed. In the following, a description of the different panel
typology behaviors is given:

4.1. Plane panels

As expected from the material characterization and


the triplet test results, all the unreinforced specimens
presented brittle failure due to splitting along the loaded
diagonal.
The average failure load, used as reference value for Fig. 4. Double-side reinforced panel failure mode: either peeling in
the comparison with the strengthened specimens results, the anchoring zones or rupture of the fibers are reached depending on
is equal to 100.7 kN. both the reinforcement stiffness and bonding area.
414 M.R. Valluzzi et al. / Construction and Building Materials 16 (2002) 409–416

Fig. 5. Stress–strain diagram of panels reinforced with carbon-FRP


strips in a diagonal configuration, either on one or both sides. Values
refers to the average measurement on the opposite faces.

shown in Fig. 7 in terms of principal tensile stress


(splitting stress). According to the Frocht theory, in
URM panels the highest splitting stresses are concen-
trated in the core, whilst partial stress relocation is given
by the diagonal reinforcement. Finally, the squared grid
reinforcement configuration presents a more uniform
distribution of stress on the panel, while stress peaks
are shifted toward anchoring zones.
Analytical models available in literature to evaluate
the shear strength VRd of URM shear walls are here
reported. All of them are based on the linear effects
superposition, which derive from the implicit assumption
of plastic stress redistribution. Even though the latter
assumption is not properly introduced when dealing with
FRP, at present no more appropriate approaches are
available. Formulations (2) and (4) (ENV, Eurocode 6
w6x and Tomazevic et al. w7x, respectively) are proposed
for masonry reinforced with steel bars; whereas formu-
lation (5) w4x is introduced specifically for FRP. Essen- Fig. 7. FEM simulation: highest splitting stress in the core of the URM
tially, they are the sum of the original masonry shear panel; partial stress relocation with diagonal reinforcement; and shift-
ing of stress peaks to anchoring areas with squared grid reinforcement
distribution.

capacity and the reinforcement effect, which is reduced


for FRPs to take into account different issues descending
from their non-ductile behavior. Particularly, in the
mentioned formulas are included the characteristic
strength of the masonry f vk and the geometrical and the
mechanical characteristics of the masonry wall and of
the reinforcement (t is the wall thickness, l its length,
ds0.8l is the effective depth; rfrp is the FRP ratio
computed on the wall section; Ar is the area of the
reinforcement; f tk is the characteristic tensile stress of
the FRP; ´frp,u is the FRP tensile ultimate strain). In
particular, the formulation (5) considers a factor of
Fig. 6. Stress–strain diagram of panels reinforced with glass-FRP efficiency r, which depends on the failure mode (FRP
strips in a diagonal configuration, either on one or both sides. Values rupture or de-bonding). The expression of r, given by
refers to the average measurement on the opposite faces. the Eq. (6), was found by Triantafillou w4x for concrete
M.R. Valluzzi et al. / Construction and Building Materials 16 (2002) 409–416 415

Fig. 8. Comparison between experimental and analytically predicted shear strength for URM and FRP-strengthened panels.

members (´frp,e is the effective FRP strain). On the basis as it can noticeably affect the FRP strengthening
of the present experimental database it would be possible effectiveness.
to provide a better calibration of the r factor for
masonry; in fact, by measuring the effective strain for 6. Conclusions
each different FRP type it is possible to newly determine
all coefficients of Eq. (6) by polynomial interpolation. The contribution of FRP strips on the shear behavior
Note that f vk0 wEqs. (2) and (3)x is obtained by sliding of clay brick wallets has been investigated; far from
tests on triplets, whereas f 9vk0 wEq. (4)x is obtained by being exhaustive, the results of the present study indi-
diagonal compressive tests. cated that asymmetrical applications (single-side rein-
VRdsfvktdq0.9drfrpftkt (2) forcement) on masonry panels offer a limited
effectiveness.
where: The diagonal configuration is more efficient in terms
fvksfvk0qms0; (3) of shear capacity than the grid set up; however, the
latter offers a better stress redistribution that causes a
B E
VRdsD0.9tlf9vk0y1.5qy1qs0yf9vk0Gq0.4Arftk
C F
(4) crack spreading and a less brittle failure.
In most cases, less stiff FRP material appeared to be
VRdsfvktdq0.9drfrpEfrpr´frp,ut more effective both in terms of ultimate strength and
(rsefficiency factor) (5) stiffness (not reported for brevity) increase of the panels.
That was due to the particular design criterion used
r´frp,us´frp,es0.0119y0.0205ŽrfrpEfrp.
(weaker material has a larger adhesion area), and also
q0.0104ŽrfrpEfrp.2 (6) to the fact that stiffer material is more vulnerable to de-
The comparison among the experimental and the bonding, especially when the number of plies increase.
predicted values of the shear strength is reported in Fig. Presently, our effort is oriented to formulate a relation
8. The efficiency factor given by Eq. (6) appears to be involving stiffness, thickness, width and number of
excessively conservative as it provides very low shear layers of FRP to quantify the susceptibility of the
strengths, so a better calibration of the formula is reinforcement to de-bond.
necessary. Anyway, despite the same mechanical param- Low increments in the shear strength, in the particular
eter rE was maintained for diagonal and grid set-up, as experience here described, are also attributable to the
mentioned before, the related shear strength differed of peeling occurring in the portions next to the applied
more than 40% for the two configurations. Therefore, compressive loads (where high stress causes premature
for a better calibration of the r factor, also the geomet- cracks) and to the peculiar lower tensile strength of the
rical reinforcement arrangement should be considered, bricks, which causes splitting failure through them as
416 M.R. Valluzzi et al. / Construction and Building Materials 16 (2002) 409–416

main failure mechanism. Nevertheless, bricks used rep- References


resent the most common typology in Italy.
Alternative anchoring methods appear to be a key w1x Binda L, Modena C, Valluzzi MR. Mechanical effects of bed
issue to evaluate during further experimentations in joints steel reinforcement in historic masonry structures. Struc-
order to prevent loss of effectiveness due to de-bond. tural faultsqrepair. Engineering Techics Press, Edinburgh, UK:
The type of test conducted and the specimen dimen- 1999.
sions appear an easy and efficient system to check w2x Tinazzi D, Arduini M, Modena C, Nanni A. FRP-structural
repointing of masonry assemblages. Ottawa, Canada: ACMBS
elementary strengthening configurations.
III, 2000.
Finally, the application of Triantafillou’s formula wEq. w3x Valluzzi MR, Modena C, Marchetti M. Shear strengthening of
(5)x requires further calibrations, especially for failure masonry panels using FRP. Proceedings of the Twelfth IB2MaC
mechanisms affecting the efficiency parameter and the Conference. Madrid, Spain: 2000.
configuration of reinforcement. Anyway, until new w4x Triantafillou TC. Strengthening of masonry structures using
approaches to predict the shear capacity are available, a epoxy-bonded FRP laminates. J Compos Constr ASCE
better definition of the efficiency factor r appears to be 1998;2(2):96 –104.
w5x Yokel FY, Fattal GS. Failure hypothesis for masonry shear
the right direction to follow.
walls. J Struct Div 1976;102(ST3):515 –32.
w6x ENV 1996-1-1. Eurocode 6: design of masonry structures.
Acknowledgments 1996.
w7x Tomazevic M, Lutman M, Pertrovic L. In plane behavior of
All the materials used in the research, including the reinforced masonry walls subjected to cyclic lateral loads.
fibers and the adhesion system, were supplied by Mod- Report to the Ministry of Science and Technology of Republic
ern Advanced Concrete (MAC S.p.A.) of Treviso, Italy. of Slovenia, parts 1 and 2. Liubljana, Slovenia, 1993.

You might also like