Académique Documents
Professionnel Documents
Culture Documents
Keywords: anchorage, fiber reinforced polymers, lime stone, masonry, out-of-plane, strengthening.
1 INTRODUCTION
There is a huge inventory of URM walls exposed to structural weakness or overloading . Most of
the traditional strengthening methods result in an added mass to the structure, extracting larger forces
during an earthquake. FRP systems can improve strength, stiffness and ductility. A considerable
progress was achieved in understanding the basic mechanics and establishing design guidelines on
the use of FRP composites in the retrofit of URM (Unreinforced Masonry) walls . In this paper we are
satisfying some of the research needs related to the reinforcement of masonry structures with FRP
systems as pointed out in the ACI 440 M (2005)1 and as investigated by Abdul Rahman, B. (2006)2
,Abdul Rahman, B. and El Shafey, N.(2005)3 which are as follows , Interaction of axial and flexural
capacity of masonry walls retrofitted with FRP systems ; determination of failure modes of masonry
walls retrofit with externally bonded FRP systems , detailing performance of FRP anchors when
anchored into masonry , and the effect of anchorages on masonry elements .
Several masonry strengthening work using FRP to resist out-of-plane loads were proposed and
investigated by, Ehsani, et. al (1997)4 , Triantafillou (1998)5, Dimas, et. al (2000)6, Tumialan, et. al
(2003)7, Hamoush, et. al (2003)8 , Tan and Patoary (2004)9 , Ghobarah and Galal (2004)10.
Triantafillou (1998)5 has performed one of a few researches that considered the interaction of axial
load and flexure as a parameter in the experimental investigations , he also proposed design
equations for FRP strengthened walls for both combined out-of-plane bending and axial load , using
the concept of balanced strain reinforcement ratio. Natural stone masonry retrofit with FRP was
investigated by some researchers such as Aiello, et. al (2003)11 who studied the bond performance
between natural masonry and FRP sheets. GFRP anchor spikes proved to have good performance in
preventing peeling off and enhancing the flexural strength of reinforced concrete slabs Piyong, et. al
(2003)12, and in clay masonry Tan and Patoary (2004)9.
2 EXPERIMENTAL STUDY
1
FRPRCS-8 University of Patras, Patras, Greece, July 16-18, 2007
according to ASTM designation composed of lime, sand and water with an average compressive
strength of 0.15 Mpa , and masonry prism compressive strength was 3.7 MPa .
A total of identical nine one-third scale stone masonry wall specimens were constructed and
reinforced with a single vertical GFRP ply on the tension side, in the reinforced concrete laboratory of
Cairo university . The specimens nominal dimensions were (200x450x1450) mm. They included two
as-built (control) specimens to represent two different axial load levels. The walls were tested
statically under combined uniformly distributed normal force and two points out-of-plane bending. The
walls were divided into three groups , all of them had approximately the same axial capacity, all the
walls had h/t ratio equal to 7.1. The numbers 1,2,3 in the wall identification denote the group number.
The letters A&G denote as-built or strengthened walls .While; the letters I, II, III denote the specimen
number within each group .
The axial load was calculated based on a percentage of the average cross sectional capacity of
the wall (N) where N was approximately 30 kN . The walls were tested one week after application of
GFRP and 28 days after construction .Figure(1) shows the dimensions , the building pattern and
reinforcement configuration of the test walls. The details of the tested walls are summarized in
Table(1). It should be noted that the percent of the reinforcement ratio is the ratio between the gross
cross sectional area of the GFRP reinforcement to the cross sectional area of the masonry section.
GFRP Anchorage
Axial load Strip Width
Group Specimen Reinforcement at each side
level (N) (mm)
Ratio *(%) of GFRP sheet
W1GI 0.06 50 Non
Group 1 W1GII 0.1 0.12 100 Non
W1GIII 0.12 100 One anchor spike
W2A Control specimen
Group 2 W2GI 0.1 0.06 50 Two anchor spikes
W2GII 0.12 100 Two anchor spikes
W3A Control specimen
Group 3 W3GI 0.2 0.06 50 Two anchor spikes
W3GII 0.12 100 Two anchor spikes
2
FRPRCS-8 University of Patras, Patras, Greece, July 16-18, 2007
construction , the prisms were laid with lime mortar using the same design mix in the walls . The
prisms were tested in compression; according to ASTM designation E447. The Pattern used for the
prism assembly was chosen to provide representative value of the compressive strength as built in the
structure , Drysdale et al.(1994)13 the test setup for the compressive strength of masonry units and for
masonry prisms is shown in Figure 2(a) and (b) respectively. Table(2) shows the test results of the
average compressive strengths of the materials.
(a) (b)
Fig. 2 Test Walls
3
FRPRCS-8 University of Patras, Patras, Greece, July 16-18, 2007
A test setup was designed to subject the masonry wall specimens to an increasing horizontal load
causing out-of-plane bending combined with a constant uniformly distributed axial compressive normal
force maintained at a specified axial load level while applying the horizontal load. The out-of-plane
load was created by means of a hydraulic jack activated by a manual pump. The force created by this
jack reacted against a vertical stiff steel girder .This force was transmitted to the exterior face of the
masonry wall using two steel rods attached to two horizontal steel beams at a quarter height from the
top and the bottom of the wall. On the reaction side, two 45 cm length hinges were used to prevent top
and bottom bed joint sliding "Walking", they were placed 50 mm from the top and bottom of the wall
edges. A uniformly distributed axial compressive normal force was generated by 500-ton testing
machine at a specified axial load level; the wall was placed between an R.C. footing at the wall base
and the rigid steel plate of the testing machine at the top of the wall. This prevented the wall edges
from outward movement. A scheme of the test setup and the actual test in the lab is shown in
Figure(4) . A constant vertical prestress was maintained during the tests by adjusting the axial load
manually.
The first group consisted of W1GI, W1GII, and W1GIII . It was subjected to an axial load
equivalent to 0.1 axial load level W1GI and W1GII were strengthened using 50 and 100 mm GFRP
plies respectively without anchorage. A single anchor was used in W2GIII in the second masonry unit.
W2A & W3A were tested under the condition of “as-built” the axial load was increased from 0.1 to 0.2
axial load level, W2GI and W3GI were strengthened using 50mm single GFRP layer. To study the
effect of increasing the axial load level in case of low reinforcement ratio. W2GII and W3GII were
strengthened with a single 100 mm GFRP layer. To study the effect of increasing the axial load when
using high reinforcement ratio . For all tested walls mid span deflections , strains at the tension side
and the applied loads have been measured.
N N
Channels
LVDTs
Load
Cell
GFRP Sheet
Strain Gauge
Hydraulic
Jack
Masonry wall
4
FRPRCS-8 University of Patras, Patras, Greece, July 16-18, 2007
70
60
Out-of-plane load (kN)
W1GIII
50 Sliding shear
Debonding
40 W1GII
Debonding
30
20 W1GI
W2A W1GI W1GII
10 W1GIII
W2A
0
0 10 20 30 40
Mid span deflection mm
Fig 5 Load Deflection Curves of the First Group with the Control Specimen
Mortar
Mortar Joint
Joint Crack
Crack
Location
of a single
GFRP Sliding
anchor shear
Hinge spike failure
Hinge
5
FRPRCS-8 University of Patras, Patras, Greece, July 16-18, 2007
70
60
Tensile rupture W2GI W2GII
Out-of-plane load (kN)
in GFRP
50
Sliding
40 shear failure
30
W2A
20
W2A
10 W2GI
W2GII
0
0 10 20 30 40
Mid span deflection mm
Fig 7 Load Deflection Curves and Failure Modes Second Group
Mortar
Mortar joint
Cracks
joint
Cracks Shear
GFRP Failure
Rupture
6
FRPRCS-8 University of Patras, Patras, Greece, July 16-18, 2007
W3GII
50
W3GI Sliding
40 shear
failure
30
W3GII
20 W3A W3A
10 W3GI
W3GII
0
0 10 20 30 40 5
Mid span deflection mm
Fig 9 Load Deflection Curves and Failure Modes Third Group
GFRP
anchor Stepped
spike shear
Crack
Mid height
mortar joint Rupture of
crack GFRP sheet
The flexural capacity of the strengthened walls was calculated using the design equations by
Triantafillou (1998)5 for the case of out-of-plane bending with axial force .The additional moment due
to the P-delta effect was included in calculations based on γ m and γ frp =1 , a MathCAD worksheet
was constructed to perform the calculations .
Comparing the theoretical results and experimental observations it can be concluded that the given
equations were effective in predicting the ultimate capacity for the case of tensile flexural failure,
7
FRPRCS-8 University of Patras, Patras, Greece, July 16-18, 2007
where the tensile strength of GFRP is governing the wall behaviour. In the case of unanchored
specimens when debonding is controlling the wall behaviour (un-anchored specimens) the equations
where unconservative . A comparison between the experimental and predicted wall ultimate loads is
given in Table(3).
Table 3 Comparison between Experimental and Predicted Ultimate Loads
Predicted Experimental
Ultimate Ultimate out- Experimental Failure
Group Specimen
out-of-plane of-plane load Predicted Mode
load (kN) (kN)
W1GI 63.5 35.0 0.55 D
Group1 W1GII 70 42.5 0.61 D
W1GIII 70 47.0 0.67 SS
W2A Control 30.5 Control O2C
Group2 W2GI 63.5 63.3 1 R
W2GII 70 60.0 0.86 SS
W3A Control 49.4 Control O1C
Group3 W3GI 54 58.7 1.09 R
W3GII 71 49.2 0.69 SS
R is flexural rupture failure of GFRP, SS is sliding shear mode of failure, D is debonding failure,
O2C is overturning on two cracks, O1C is overturning on a single crack, the percentage of the strain
levels in GFRP are measured when debonding initiated ( for the specimens failed by debonding) , or
the strain attained at failure for different modes of failure.
Compared to as-built walls the strengthened walls exhibited significantly improved performance .
The composite overlays adhered very well to the stone substrate and apparently helped to prevent a
total masonry collapse. The GFRP anchor spikes gave a good performance in anchoring the sheet to
the wall till the end of the experiments, transferring the cause of failure from premature peeling off to
other modes. A fact , which must be clearly indicated, is that there are some limited variations in the
properties of the test specimens which were fabricated using natural dimension stone because of the
nature of the material .
Increasing the GFRP reinforcement ratio led to another premature failure caused by sliding shear
at the second mortar joint. That was true for specimens W2GII and W3GII which were reinforced using
100 mm strip .The same observation was stated in the ACI 440 M (2005)1 in case of strengthening
flexural and bearing walls for out-of-plane loads . Indicating the fact of observing shear mode of failure
in case of providing a large amount of FRP for flexural reinforcement .Figure(11) shows the
enhancement in the walls’ behaviour due to anchorage for the higher and lower reinforcement ratios at
0.2 axial load.
70
70
Tensile rupture
60
60 W2GI GFRP W2GII
Out-of-plane load (kN)
debonding Sliding
50 W1GII
50 shear failure
40 40
Debonding
30 30
W1GI
W2A
20 W2A W1GI 20
W2A W2A
10 W2GI 10 W2GII
W1GII
0 0
0 10 20 30 40 0 10 20 30 40
Mid span deflection mm Mid span deflection mm
8
FRPRCS-8 University of Patras, Patras, Greece, July 16-18, 2007
Table(4) shows the increase in the ultimate load carrying capacities of the strengthened walls
calculated as a ratio of the capacity of the control specimen for each group and the mode of failure .
Table 4 Increase in the Ultimate Load Carrying Capacity Referred to the Control Specimen
For the strengthened specimens variable modes of failure were noted depending on the
dominant parameters. It was noted that the increase in the ultimate load carrying capacity was
maximum when the failure of the wall was controlled by tensile rupture for lower axial load level . All of
the specimens failed by undesirable brittle modes of failure which didn’t give enough signs of warning
before failure this was by either premature debonding or tensile rupture in GFRP or by sliding or
stepped shear failure, to obtain more desirable mode of failure compressive crushing failure should be
insured this was investigated by Abdul Rahman, B.(2006)2 .
CONCLUSIONS
The following conclusions are deducted from the experimental results
RECOMMENDATIONS
It is suggested to do further experimental investigations to analyse the various parameters , and to
develop reliable relationships for the cases considered. It is also recommended to investigate the
reinforcement ratio at which compressive crushing occurs, which is the most desirable mode of failure.
9
FRPRCS-8 University of Patras, Patras, Greece, July 16-18, 2007
REFERENCES
[1] ACI 440 M, Guide for the Design and Construction of Externally Bonded FRP Systems for
Strengthening Unreinforced Masonry Structures , Draft of the State- of-the-Art Report on
externally bonded Fiber Reinforced Plastic (FRP) Reinforcement for Masonry Structures,2005.
[2] Abdul Rahman, B., “Out-Of-Plane GFRP Strengthening of Natural Masonry Walls Enhanced
Using Glass Fiber Composite Anchors” , Master Degree Thesis Department of Structural
Engineering, Cairo University, Egypt January 2006.
[3] Abdul Rahman, B. and El Shafey, N. “Out-Of-Plane GFRP Strengthening Of Natural Masonry
Walls Enhanced Using Glass Fiber Reinforced Polymer Anchors”, Fourth Middle East
Symposium Of Structural Composites For Infrastructure Applications, May 20-23, Alexandria,
Egypt.
[4] Ehsani, M., Saadatmanesh, H., Al-Saidy, A., “Shear Behaviour of URM Retrofitted with FRP
Overlays” , Journal of Composites for Construction ASCE, 1(1), 1997, pp 17-25.
Web site: http://www.u.arizona.edu/~ehsani/Publications.htm.
[5] Triantafillou,T. “Strengthening of Masonry Structures Using Epoxy-Bonded FRP Laminates”,
Journal of Composites for Construction, ASCE, V. 2, No. 2,1998, pp. 96-104.
[6] Dimas, V., Ehsani, M. and Saadatmanesh, H., "Out-of-Plane Behavior of Brick Masonry Walls
Strengthened with Fiber Composites ", ACI Structural Journal Vol. 97 No. 3 May June, 2000
pp. 377-387.
[7] Tumialan, J., Galati, N., Nanni, A., “Fiber-Reinforced Polymer Strengthening of Unreinforced
Masonry Walls Subject to Out-of-Plane Loads”, ACI Structural Journal, V. 100, Issue 3, May-
June, 2003, pp 321-329
[8] Hamoush, S., Mlakar, P., Bastos, A., Costa, A., “The Shear Strength of Fiber Reinforced
Masonry Walls”, Proceedings of the International Conference Structural Faults and Repair 1-3
July,2003, London UK.
[9] Tan, K., Patoary, M., “Strengthening of Masonry Walls against Out-of-Plane Loads Using
Fiber-Reinforced Polymer Reinforcement”, Journal of Composites for Construction, ASCE
January/February,2004, pp 1-9.
[10] Ghobarah, A., Galal, K., ”Out-of-Plane Strengthening of Unreinforced Masonry Walls with
Openings”, Journal of Composites for Construction ASCE, July/August, 2004,pp 298-305.
[11] Aiello, M., Sciolti, M., Pecce, M., “Experimental Investigation on Bond between FRP Sheets
and Natural Masonry Blocks”, Proceedings of the International Conference Structural Faults
and Repair, 1-3 July,2003, London UK.
[12] Piyong, Y., Silva, P., Nanni, A., “Flexural Strengthening of Concrete Slabs by a Three-stage
Prestressing FRP System Enhanced with the Presence of GFRP Anchor Spikes”, Proceedings
of CCC 2003 - Composites in Construction International Conference, 16-19 September,2003,
University of Calabria Rende (CS), Italy.
Web site: http://campus.umr.edu/rb2c/research/2003/03-46.pdf
[13] Drysdale, R., Hamid, A., Baker, L., “Masonry Structures Behaviour and Design” , Textbook and
design reference”, ISBN 0-13-562026-0, Sponsored by Brick Institute of America National
Concrete Masonry Association published by Prentice-Hall,1994.
10