Structures 28 (2020) 2488–2506

Contents lists available at ScienceDirect

Structures
journal homepage: [Link]/locate/structures

Repairing and strengthening of reinforced concrete cantilever slabs using

Glass Fiber–Reinforced Polymer (GFRP) wraps
Nehal M. Ayash ([Link])*, Ahmed M. Abd-Elrahman, Abd-Elaziz Soliman
Civil Engineering Department, Faculty of Engineering at Mataria, Helwan University, Cairo, Egypt

A R T I C L E I N F O A B S T R A C T

Keywords: This study represents the behavior of reinforced concrete cantilever slabs repaired and strengthened using GFRP
Cantilever slabs experimentally; then a 3D finite element model using “ANSYS” was constructed to test the validity of analytical
Reinforced concrete models in capturing the experimental cases. Additionally, the experimental results are compared with theoretical
Repairing
calculations per the ACI and ECP codes to test the validity of codes equations. Finally, more numerically models
Strengthening
Reloading
are applied to review the effect of various parameters on the behavior of cantilever slabs strengthened with
GFRP wraps GFRP. Also, the theoretical calculations consistent with the ACI and ECP codes are compared with numerical
GF anchors results obtained using the ANSYS finite element program for parametric study.
Debonding The using GFRP as a top mesh on cantilever slabs increased the strength and therefore the deflection of the
Crushing cantilever was increased due to the more additional load that resisted by Fibers. The strengthened slabs were
provided more ductility, more strength capacity and higher deflection than the repaired slabs. The slabs
strengthened/repaired with GFRP Wraps additionally to glass anchors get more strength capacity than slabs
strengthened/repaired with GFRP Wraps only. For (ACI) and (ECP) codes the identical in predicting the share of
flexure strength increasing but need adding amplification factor about 1.25. (ACI) and (ECP) codes do not
capture the effect the FRP anchors in flexure strength increasing, but it can be reached that these needs adding
amplification factor about 1.12 above the case of strengthening without anchors. An analytical model using
ANSYS software program can present a good simulation for the failure loading, deflections and debonding of
fiber for the cases of strengthening and repairing. This encourages continuing the parametric study. From the
parametric study, it can reach that the greater the cantilever length, the more the deflection but the lower the
load carrying capacity although of the more the use of fiber glass in strengthening. The greater the thickness of
the strengthening cantilever slab, the lesser the deflection value but the more the load carrying capacity.

1. Introduction be a significant issue with steel plate bonding. The utilization of fiber
reinforced polymer composite materials is a pretty option because it
Reinforced concrete cantilever beams and slabs are common features offers many advantages over the standard steel plate bonding technique,
in buildings. Worldwide, the utilization of R.C. canopies and balconies including a high strength/weight ratio, high corrosion resistance and
are widespread. As cantilever slabs are statically determinate structures, suppleness with the architectural uses. Accordingly, several researchers
they sudden failed once their moment capacity is exceeded at support. recommended the utilization of assorted end anchorage systems to avoid
The failures of cantilever slab structures were due to mistakes in placing such a sudden premature debonding failure and to make sure that
the steel reinforcement bars, severe environment and errors within the strengthened members will develop their full flexural capacity.
design. There are two common ways to retrofit deficient cantilever slabs; The previous studies were started by Teng et al. [1] which study
the primary way is to extend the slab thickness with additional top steel experimentally the strengthening of RC cantilever slabs by bonding
reinforcement which can conflict with the architectural uses. The second GFRP strips/sheets on the top surface. The effectiveness of various
way is to bond steel plates to the top surface of slab; this way has been anchorage systems for GFRP was focused. Seven tests on RC cantilever
used widely to strengthen R.C structures. Since the top surface of slabs with or without strengthening using GFRP strips had been con­
cantilever slabs are mostly exposed to weather attacks, corrosion could ducted. They reached that the GFRP strips anchored into the walls

* Corresponding author.
E-mail addresses: Nehal82ayash@[Link] (N.M. Ayash), ahmousa2@[Link] (A.M. Abd-Elrahman), ceabdsoliman@[Link] (A.-E. Soliman).

[Link]
Received 20 July 2020; Received in revised form 20 October 2020; Accepted 21 October 2020
Available online 5 November 2020
2352-0124/© 2020 Institution of Structural Engineers. Published by Elsevier Ltd. All rights reserved.
N.M. Ayash et al. Structures 28 (2020) 2488–2506

through horizontal slots using fiber anchors is an easy and effective is slabs strengthened using two strips of two layers of GFRP covering
method. Lam et al. [2] examined the strengthening of R.C. cantilever 60% and 90% from the cantilever length. The third group is slabs
slabs with bonded GFRP strips both experimentally and theoretically. strengthened with two steel plates either glued using epoxy resin or
The GFRP strips were anchored to the supporting wall using horizontal fixed using expansion bolts. The loading was conducted in two phases.
slots to limit deboning. It had been concluded that deboning between the Firstly, the samples were subjected to 40% of the ultimate load of the
slab and GFRP strips was less likely to occur with sufficient anchorage of control slab. Secondly, the specimens reloaded after being strengthened
the FRP strips to the cantilever slab by fiber anchors. The slabs might fail up to failure. They concluded that the jacket strengthening technique
by either FRP rupture or concrete crushing. The preloading effect was has significant influence on all measured responses, regardless of the
not significant if the FRP rupture is occurred. This effect becomes more installation the shear connector bars. De-bonding occurred of GFRP
significant and detrimental if the slab failed by concrete crushing. Teng sheets, however, the strengthening with GFRP sheets using 60% of
et al. [3] presented an experimental study on R.C. cantilever slabs cantilever length showed superior stiffness and ductility compared to
bonded with GFRP strips. Ten tests in three series were conducted on strengthening with 90% of cantilever length. The strengthening by steel
such slabs with different amounts of internal steel reinforcement and plate with epoxy resin caused an early de-bonding.
external FRP reinforcement. The fiber anchors were installed near the Although of the benefit of the numerical method that supplemented
fixed end of the slab. They concluded that Debonding results in more experimental testing and can assist in exploring effects of different pa­
ductile behavior of the slab. The effect of internal steel reinforcement rameters, which the experimental program left out, because of the
ratio on debonding behavior within the slabs is not strong. Instead, the limitations in time, apparatus or cost, there are few finite element
important factors appear to be the FRP strip width and thickness. studies on this field and most are limited only on validation of F.E. model
Moussa [4] studied experimentally and theoretically the behavior of such as; Fahmy A. Fathelbab et al. [10] studied analytically the
strengthened and repaired R.C. cantilever slabs and evaluated the effi­ strengthening of a RC bridge slabs using externally bonded FRP sheets. A
ciency of various strengthening and repairing methods like using carbon finite element program ANSYS was analyzed strengthened slab models
fiber sheets, adding new tension steel bars. It was concluded that the using several arrangements of FRP sheets. The evaluation of the effect of
CRFP sheets presented a decent solution for the strength achievement changing FRP stiffness and FRP schemes in strengthening RC slabs are
without any extra dead loads regarding the simplicity of application. performed. It is observed that attaching FRP sheets to the RC slab in­
Using of second layer of CRFP sheets can cause to significant increase in creases its capacity and enhances the ductility. Mehdi M. Lima et al. [11]
the flexure capacity and stiffness of slab. The effect of preloading on the constructed a finite element analysis that validated with the results of
behavior of steel repaired slabs is more significant way than that of CFRP the experiments of two reinforced concrete beams, two columns, two
slabs. Tamer El Maaddawy et al. [5] examined the using of slabs and six walls. The ultimate failure load of these members is ob­
mechanically-anchored un-bonded fiber reinforced polymer (MA-UFRP) tained as they have different CFRP orientations to the loading direction.
for improving the flexural strength of the RC slabs. Six RC slabs were The analysis result showed god agreement with the experimental data
tested under four-point bending to failure. One slab was used as a con­ regarding the ultimate failure load of RC samples, except for the RC wall.
trol; two slabs were strengthened with externally-bonded FRP (EB-FRP) The FRP orientation has significant effect on the ultimate failure load in
system, one slab with end-anchorage and one slab without end- all strengthened RC members. Tejendra Tank et al. [12] analyzed ret­
anchorage. The remained three slabs were strengthened with MA- rofitted slabs using in ANSYS software. The (Slab-A) was non-
UFRP system having different anchors’ locations. It showed that MA- strengthened slab; the (Slab-B) was strengthened with GFRP sheet and
UFRP system enhanced the flexural strength with up to 43%. The the (Slab-C) was strengthened with GFRP sheet at the entire soffit of the
strength of the slabs strengthened with MA-UFRP system was on 18% slab. Analytical results showed good agreement with experimental re­
and 10% lower than that of the slab strengthened with EB-FRP system sults for all cases. The deflection values by ANSYS, are lower by nearly
with end anchorage and without end-anchorage, respectively. The by 10–15% than those obtained from experiments and this is due to the
deflection of the slabs strengthened with MA-UFRP system was 56% and shrinkage and creep parameters in experimental are not in ANSYS. Dina
5% higher than that of the slab strengthened with EBFRP without end- A. Emarah [13] investigated the behavior of RC slabs strengthened with
anchorage and with EB-FRP with end-anchorage respectively; while CFRP sheets using nonlinear finite-element analysis by ANSYS. A com­
15% lower than that of the control. Scott T. Smith et al. [6] tested six parison between the results of ANSYS, experimental works, and equa­
specimens of one-way simply supported RC slabs that strengthened with tions of Egyptian Standing Code (ESC) was made for strengthened slabs
FRP composites on tension face and anchored with various arrange­ with different sheet areas. It was investigated the effects changing the
ments of FRP anchors. The enhancement in capacity and deflection for main reinforcement, concrete compressive strength, slab thickness, and
slabs strengthened with FRP plates and anchors was 30% and 110%, FRP sheet thickness on strengthened slab behavior. The results of ANSYS
respectively, over the control slab. Closer spaced anchors reduced the and the equation of ESC are in good agreement. The strengthening of
rate of debonding crack propagation and also achieved higher de­ slabs with CFRP sheets is improved the flexural strength capacity. The
flections. Sergio F. Bren et al. [7] performed the experimental tests and distribution of CFRP sheets enhanced the performance of slabs. The
finite element simulations to understand the behavior between FRP increase in concrete compressive strength, slab thickness, and sheet
sheets and FRP anchors. The results showed that Longitudinal spacing of thickness leads to increase in load capacity of strengthened slabs. The
anchors should be limited to a distance located within the stress transfer increase in the main reinforcement ratio increased the failure load up to
zone of the FRP sheet. The anchor diameter should be sized according to a certain limit. Despite of the intensity of the studies on the behavior of
the force demand in the sheet. Tarek Ali et al. [8] presented the effect of simple slabs using FRP; there are a few studies that focused on cantilever
using different methods to enhance the ultimate capacity of flexural slabs using FRP, also, there are a lack in studying the main parameters
behavior in RC slabs. Four RC specimens were casted. Specimens were specially focused on the slab properties not the fibers properties which
strengthened with different methods such as usage of GFRP sheets, affecting on the behavior of cantilever slabs. In addition, there are vague
carbon fibers laminate strips and near surface mounted steel rebars. All in the codal design equations in the case of FRP repairing and in case of
specimens were subjected to two-point loads. The behavior of anchors existence in repairing and strengthening.
strengthened specimens was compared with the control specimen. They The objective of the current work is presenting more specialization in
reached that Strengthening RC slabs using GFRP Sheets, CFRP Laminate the behavior of reinforced concrete cantilever slabs repaired or
Strips or using NSM steel rebars is an effective technique. Aymen H. strengthened using GFRP wraps with/without GF anchors experimen­
Khalil et al. [9] tested six strengthened R.C. cantilever slabs specimens tally. The numerical analysis by using finite element program ANSYS
divided into three groups. The first group is slabs strengthened using R. software are conducted to calibrate with these experimental works for
C. jacket above with and without shear connector bars; the second group presenting more studies on the various parameters that affected on the

2489
N.M. Ayash et al. Structures 28 (2020) 2488–2506

Fig. 1. Test Setup and Details.

Fig. 2. Steel Reinforcement Details of All Tested Slabs.

behavior of R.C. cantilever slabs strengthened with GFRP. The design bottom and adjacent spans have 5Ø12 bottom steel. The transverse re­
provisions for using FRP to reinforced concrete in different international inforcements are 5Ø8/m’ top/bottom. All slabs are designed based on
codes are revised with these experimental and numerical works. Egyptian code ECP 203–2018 [15].
Fig. 1 showed the Test setup and details. All the slab specimens were
2. Experimental works subjected to two- points loading; one is at the end of the cantilever and
the second at the mid span between two supports. The test was carried
2.1. Test program out at an approximate rate of loading of 15–25 kN per step by the hy­
draulic jack of 250 kN maximum capacity. The load was increased until
The experimental test program included five one-way simple slabs failure occurred, either by failure of FRP or by crushing of the concrete.
with cantilevers. The test consisted 5 slabs which (S01) represented the Two groups of strain gauges were used to monitor the strain component
control slab without repairing or strengthening, (S03 and S04) repre­ at the specified locations. The first group was the steel reinforcement
sented the repaired slabs which previously loaded (up to 0.7Pu) of gauges that glued on the reinforcing bars at cantilever support and at the
cracking load (Pcr) and (S06 and S07) represented the strengthened slabs mid-span of the adjacent span. The second group was the concrete strain
before loading. The slabs (S03) and (S06) were repaired and strength­ gauges, they assigned on the top face of cantilever slab and on the
ened with Glass Fiber Wraps as externally bonded on the negative bottom face of the mid-span of the adjacent span. Two Linear Variable
moment zone, respectively. The slabs (S04) and (S07) were repaired and Differential Transformers (LVDTs) have been used to determine the
strengthened with both Glass Fiber Wraps and anchors, respectively. All deflections at cantilever end and at the mid-span of the adjacent span.
slabs had total length of 3000 mm divided into a centerline span of 1950 The layout of the test slab showing the geometric and reinforcement
mm length and a cantilever that had 950 mm effective length. All slabs details, the statically system, loading and bending moments were shown
had a cross section of (600 × 120 mm). All cantilevers had 5 Ø 10 top/ in Fig. 2 and Fig. 3.

2490
N.M. Ayash et al. Structures 28 (2020) 2488–2506

Fig. 3. Concrete Details of All Tested Slabs Showing the Statically System, Loading and Bending Moment.

Fig. 4. (a) GFRP Wrap and (b) GF-Anchors Application on Top Slab Surface.

Strengthened slab is implemented by gluing the fiber wraps onto the

Table 1
top surface along the slab axis before loading; while repairing slab is
Concrete Mix Proportions.
loaded firstly by 70% of the ultimate loads and then cracks were
Coarse Aggregate (Kg) Fine Aggregate (Kg) Water (Liter/Kg) Cement (Kg) repaired and then glued the fiber wraps. The glass fiber anchors are used
1400 700 175 400 with holes depth is 6 cm and holes area are 1.5 times the area of anchor.
Fig. 4 showed the GFRP Wrap and GF-anchors application on top slab
surface.

Table 2
Physical of reinforcing steel bars. 2.2. Material properties
Material Nominal Yield Strength Yield Strain Ultimate
Diameter (mm) (N/mm2) (mm/mm) Strength (N/ 2.2.1. Concrete
mm2)
The concrete was designed to have a 28-day cube compressive
Steel 10 360 0.002 520 strength of 24 MPa, with 10 mm maximum aggregate size and 0.44
12 360 0.002 520 water content ratios. The coarse aggregate used was a crushed calcar­
8 240 0.002 350
eous aggregate. Table 1 presented the mix proportions to produce 1 m3

2491
N.M. Ayash et al. Structures 28 (2020) 2488–2506

Table 3 3. Analytical works

Technical data for GFRP Wrap (SIKAWRAP-430 G).
Property Value The analytical works were constructed using finite element model
3 (FEM) by ANSYS 15 software [17]. These analytical works were per­
Dry Fiber Density (ɣf) 2.56 g/cm
Dry fiber Tensile Strength (ʃf) 2500 N/mm2 formed to check their validation with the results obtained experimental
Fiber Modulus of Elasticity in Tension (Ec 72000 N/mm2 (EFRP) works. Then, further analytical works were constructed for assisting in
Dry fiber Elongation at break 2.08% exploring effects of various parameters.
Laminate Nominal Thickness 0.168 mm
Laminate Nominal Cross Section 168 mm2 per m width
Laminate Tensile Strength 1500 N/mm2 3.1. Element types

The element types for finite element model are shown in Table 6. The
SOLID 65 element was used to model the concrete. This element is
Table 4
Technical data for EPOXY Resin (SIKADUR- 330). capable of plastic deformation, cracking in three orthogonal directions,
and crushing. A LINK180 element was used to model steel reinforce­
Property Value
ment. This element is a uniaxial tension-compression element. The
Density 1.30 ± 0.1 kg/L SOLID45 is used for the three-dimensional modeling of solid structures.
Modulus of Elasticity in Flexure 3800 N/mm2
This element has plasticity, creep, swelling, stress stiffening, large
Modulus of Elasticity in Tension 4545 N/mm2
Tensile Strength 30 N/mm2 deflection, and large strain capabilities.
Elongation at Break 0.66%
3.2. Real constants
of concrete mix.
The real constants for model are shown in Table 6. Real constant Set
1 is used for the solid 65 element. Real constant Sets 2, 3, and 4 are
2.2.2. Reinforcing steel
defined for the link 180 element.
High tensile steel bars with diameters 10 mm and 12 mm were used
for reinforcing all slabs in longitudinal reinforcement and mild steel bar
with diameter 8 mm was used for transverse reinforcement. Table 2 3.3. Material properties
showed the mechanical properties of reinforcing steel.
For Solid65 element, the linear isotropic and multi-linear isotropic
2.2.3. Glass fiber wrap and epoxy resin material properties are defined to model concrete. The linear isotropic
The FRP Wraps were provided by SIKA-EGYPT for construction properties are defined by the modulus of elasticity Ex was based on
√̅̅̅̅
chemicals. The commercial name is (SikaWrap-430G); it is woven uni­ equation Ex = 4400 f ’c , which fc’ is specified compressive strength of
directional E-Glass fiber fabric. The epoxy used with the fiber called concrete and the passion’s ratio. The multi-linear curve is used to help
(Sikadur-330) and it is a two-components, thixotropic epoxy based with convergence of the nonlinear solution algorithm. Fig. 5 shows the
impregnating resin and adhesive. The main characteristics of fiber and stress strain curve used in this study. The compressive uniaxial stress–­
resin are provided by manufacturer are given in Table 3 that showed the strain relationship for the concrete model was divided into two parts.
technical data for FRP Wrap and Table 4 that showed the technical data The first part shows the linear elastic zone up to point 1 with stress equal
for Epoxy. 0.3 fc’ . The second part from point 2 to 6 shows the elastic–plastic zone at
the maximum compressive stress of concrete material, which obtained
2.3. Details of parameter’s groups Ec ε
from stress f = 1+( ε )2 where the strain at ultimate compressive strength
εo

The tested specimens were divided into 3 groups. Table 5 showed the
details of all specimens’ groups. Group No.1: Consists of slabs S01, S03 Table 6
Real Constants for different materials.
and S06. Slab S01 used to represent the reference slab. Slab S03 used to
represent the repaired slab with GFRP-Wrap only as it was previously Real constant Element type Constant Value
loaded with the cracking load (Pcr) up to the service load (0.70 Py). Slab 1 Solid 65 Material number 0
S06 represent the strengthened slab with GFRP-Wrap only without any Volume ratio 0
previously loading. Group No.2: Consists of slabs S01, S03, and S04. Slab 2 Link 180 / ᴓ 12 Cross section area (mm2) 113.04
3 Link 180 / ᴓ 10 Cross section area (mm2) 78.5
S04 used to represent the repaired slab with GFRP-Wrap and GF- 4 Link 180 / ᴓ 8 Cross section area (mm2) 50.25
Anchors. Group No.3: Consists of slabs S01, S06, and S07. Slab S07
used to represent the strengthened slab with GFRP-Wrap and GF-
Anchors.

Table 5
Details of parameter’s groups.
Group Slab No. Type of repair or strengthening Parameter effect Initial Load Total Depth (mm) Cantilever steel Using Anchors

(1) S01 — Reference 0.0 120 5T10 No

S03 Repaired Wrap Initial Load 0.7 Pu 120 5T10
S06 Strengthened Wrap Initial Load 0.0 120 5T10
(2) S01 — Reference 0.0 120 5T10 No
S03 Repaired Wrap Initial Load 0.7 Pu 120 5T10
S04 Repaired Wrap and Anchors Initial Load 0.7 Pu 120 5T10 Yes
(3) S01 — Reference 0.0 120 5T10 No
S06 Strengthened Wrap Initial Load 0.0 120 5T10
S07 Strengthened Wrap and Anchor Initial Load 0.0 120 5T10 Yes

2492
N.M. Ayash et al. Structures 28 (2020) 2488–2506

Fig. 5. Uniaxial Stress-Strain Curve for concrete.

Table 7 Table 9
Material Models Solid 65 for Concrete. Material Models Solid 45 for Steel Plates.
Linear isotropic Linear isotropic

Modulus of elasticity Ex 21556 N/mm2 Modulus of elasticity Ex 200000 N/mm2

Passion ratio PR xy 0.2 Passion ratio PRxy 0.3
Multi-linear isotropic
Strain Stress N/mm2
0.000334 7.20
0.000902 16.71 Table 10
0.001091 18.96 Material Models Solid 45 for Glass Fiber Wrap.
0.001318 21.04
0.001591 22.70 Linear Orthotropic
0.001909 23.72
Modulus of elasticity Ex 72000 N/mm2
0.002227 24.00
Modulus of elasticity Ey 4545 N/mm2
Concrete
Modulus of elasticity Ez 4545 N/mm2
Open shear transfer 0.3
Passion ratio PRxy 0.3
Open shear transfer 0.8
Passion ratio PRyz 0.3
Uniaxial cracking stress 2.93 N/mm2
Passion ratio PRxz 0.3
Uniaxial crushing stress 24.00 N/mm2
Shear modulus Gyz 27,692
Shear modulus Gxy 1748
Shear modulus Gxz 1748
Multi-linear
Table 8
strain Stress N/mm2
Material Models link 180 for steel reinforcement. 0.0208 1500
Linear isotropic

Modulus of elasticity Ex 200000 N/mm2

length and a cantilever that have 950 mm effective length, and a cross
Passion ratio PRxy 0.3
section of (600 × 120 mm). The loading plates and steel supports are
Bi-linear isotropic 100 × 600 × 30 mm. Fig. 6 shows the Volumes Created in ANSYS for
Yield stress fy 360 N/mm2
slab before and after adding GFRP wraps.
Tangent modulus E’ 360 N/mm2

3.5. Modeling and Meshing

2fc’
εo = Ec . The point 7 is defined at fc’ and εo .
To obtain good results from the Solid65 element, the use of a rect­
The Link180 element is being used for all steel reinforcement in the
angular mesh is recommended. The volume mapped sweep command
slabs and it is assumed to be bilinear isotropic. The material properties
was used to mesh the steel plate, concrete and GFRP. Meshing of rein­
for Solid 45 which are being used for steel plates for supports and
forcement is divided with size such that their nodes matched with vol­
loading plate is assumed to be linear isotropic. The material properties
umes mesh nodes as shown in Fig. 7. The selection of the mesh density in
for Solid 45 which are being used for steel plates for GFRP wraps is
finite element modeling is important. A convergence of results is ob­
assumed to be linear orthotropic. The Material Models for solid 65, link
tained when an adequate number of elements is used in a model and this
180 and solid 45 are shown from Tables 7–11.
can achieve when an increase in the mesh density has a negligible effect
on the results. So, it was reached that the use of element mesh size equal
3.4. Modeling
25 mm is suitable that it does not consume the time in the analyses and
giving accurate results.
The Slab, plates, and supports were modeled as volumes. The model
is 3000 mm total length divided into a centerline span of 1950 mm

2493
N.M. Ayash et al. Structures 28 (2020) 2488–2506

Table 11 restart an analysis after the initial run or load step has been completed.
Material Models Solid 65 for Resin. The time at the end of the load step refers to the ending load per load
Linear isotropic step. The time step size is set to indicate load increments used for this
analysis. Typical commands utilized in a nonlinear static analysis are
Modulus of elasticity Ex 4545 N/mm2
Passion ratio PR xy 0.2 shown in Table 12. All values for the nonlinear algorithm are set to
Concrete defaults. The nonlinear equations are adopted using Full Newton-
Open shear transfer 0.2 Raphson method with a sufficiently large number of solution sub-steps
Open shear transfer 0.8 during the loading process to capture all different stages of the
Uniaxial cracking stress 2.93 N/mm2
Uniaxial crushing stress − 1
behavior, including cracking, yielding, and failure. The automatic time
stepping, which regulates the sub-step size according to the convergence
Multi-linear isotropic
of the solution, is as program chosen to help reduce computational time.
Strain Stress N/mm2
0.0066 30 A convergence tolerance of 5% is assumed based on the displacement
degree of freedom for concrete problems.
For strengthened specimens, the analysis is started from 0 to the
3.6. Boundary conditions and loading failure at the time at end of load step equal 50000 N. On the other hand,
the analysis of repaired specimens is applied in two steps; firstly,
To ensure that the model acts the same way as the experimental slab, without adding GFRP elements and loaded from load 0 to 0.7* Pu and
boundary conditions need to be applied where the supports and loading obtained the deflection at the end of this step. The second step is by
exist in Fig. 8. The support was modeled so that a roller and a hinged adding GFRP elements and applied the deflection obtained previously
were created. All nodes under the plate were given constraint in the UY, and again loaded from 0 to the failure.
and UZ directions for roller support and all nodes under the plate were
given constraint in the UX, UY, and UZ directions for hinged support. 4. Results and discussions
The force, P, applied at each node on the steel loading plates.
4.1. Experimental results

3.7. Analysis type

The results contained mode of failure, cracking load, yielding load,
failure load, and load–deflection curves. Fig. 9 showed the different
The static analysis type is utilized. The restart command is utilized to

Fig. 6. Volumes Created in ANSYS for slab before and after adding GFRP wraps.

Fig. 7. Mesh of the Concrete, Steel Plate, and Steel reinforcement.

2494
N.M. Ayash et al. Structures 28 (2020) 2488–2506

Fig. 8. Boundary Conditions for supports and loading plates.

the concrete and steel strain in strengthened slabs are more than
Table 12 repaired slabs. Existence of anchors caused excessive increase in con­
Commands Used to Control Nonlinear Analysis. crete and steel strain than without anchors.
Analysis option Small displacement static As shown in Fig. 13, the ductility indexes for strengthened slabs are
Calculated prestress effects No more than repaired slabs either in cases without or with anchors when
Time at end of load step 50,000 compared with control slab (S01). The effect of GFRP increase the
Automatic time stepping Prog chosen ductility of slab when used as strengthening method (S06) by about
Time step size 500 188%, while when using as repairing method (S03) the ductility is
Minimum time step 500
Maximum time step 25,000
increased by 115%. The effects of anchors in strengthened slabs (S07)
Frequency Write every sub step increase the slab ductility by 129% and it is opposite to the case of using
anchors in repaired slab (S04) which reduced the ductility. Hence,
strengthening the cantilever slabs (with/ without anchors) is more
failure mode for reference, repaired and strengthened slabs. Figs. 10–12 effective than repairing (with/ without anchors).
showed the load – deflection, load – concrete strain and load – steel
strain curves for all tested cantilever slabs. The ductility index is the 4.1.2. Group (2) and (3): Effect of anchors in repairing and strengthening
ratio of deflection at the ultimate load of cantilever slab to the deflection methods
at yielding point of steel. Fig. 13 showed the ductility index of cantilever For slab S04 when compared with S03, the failure load capacity
slabs. Table 13 showed a summary of experimental test results. increased by 112% but the ductility is reduced by 83%. The concrete
strain is reduced by 65% but the steel strain increased by 118%.
4.1.1. Group (1) comparing between repairing and strengthening When comparing slab S07 with S06, the failure load capacity
As shown in Fig. 9, a typical flexural failure mode was occurred for increased by 111%. The ductility is increased by 147%. The concrete
control slab S01. Slab S03 and S06 were expressed de-bonding of the strain is increased by 456% and the steel strain increased by 122%.
glass fiber wrap then a sudden flexural failure. But for slabs S04 and S07 Hence, the effect of anchors “either in repairing or strengthening
a more cutting in fiber near the anchors was occurred. It was noted that method” has the same effect on load capacity of slab and steel strain but
using anchors increase the stress concentration around them. the uses of anchors in strengthening method caused increases in the
From Fig. 9, the failure load had a significantly effect by using the ductility for cantilever and the concrete strain more than when use these
Glass fiber wrap in repairing and strengthening. The failure load ca­ anchors in repaired method.
pacity; compared to control slab (S01); for repaired slab S03 and
strengthened slab S06 (both without anchors) were increased by 143% 4.2. Validation works
and 154%, respectively, with more deflection by 124% and 186%,
respectively as the fiber resists more loading. In addition, the failure This part divided two sections; the first is comparing the experi­
load capacity for repaired slab S04 and strengthened slab S07 (with mental results with the Design Code provisions such as ACI 440.2R-17
anchors) were increased by 161% and 172%, respectively, with more [14] and ECP 208-05 [16]. Secondly, construct the finite element
deflection by 175% and 196%, respectively as the fiber and anchors model using ANSYS software program and check its validity with the
resists more loading. results obtained from experimental tests.
In Figs. 11 and 12, concrete strain and steel strain at cantilever were
increased for all strengthened and repaired slabs (with/without an­
chors) when these were compared with controlled slabs. It is noted that

2495
N.M. Ayash et al. Structures 28 (2020) 2488–2506

Fig. 9. Failure modes for all tested slabs.

Fig. 10. Load – Cantilever deflection for tested slabs.

2496
N.M. Ayash et al. Structures 28 (2020) 2488–2506

Fig. 11. Load – concrete strain at top of cantilever for tested slabs.

Fig. 12. Load – steel strain at top of cantilever for tested slabs.

4.2.1. Validation with design code provisions

[Link]. ACI 440.2R-17 (USA). The effective stress and strain in FRP at
failure are calculated from:
ffe = Ef εfe (1)
√̅̅̅̅̅̅̅̅̅̅
h− c f ’c
εfe = εcu ( ) − εbi ≤ εfd (= 0.41 ≤ 0.9εfu ) (2)
c nEf tf

The stress and strain in steel reinforcement are calculated from:

fs = Es εs ≤ fy (3)

( ) d− c
εs = εfe + εbi ( (4)
h− c
Fig. 13. Ductility index of cantilever slabs. The strain in concrete is calculated from:
( ) c
εc = εfe + εbi ( (5)
h− c

2497
N.M. Ayash et al. Structures 28 (2020) 2488–2506

Table 13
Summary of experimental Test Results.
Specimen No. Specimen type Cracking Yield Failure Mode of failure

Pcr(kN) Δcr(mm) Py(kN) Δy(mm) Pmax (kN) Δmax (mm)

S01 Control ts = 12 cm 17 3.5 39.7 11.12 57.7 23.89 FF

S03 Repair GFRP 19 6.17 47.7 11.9 82.5 29.54 DFF
S04 Repair GFRP + Anchors 20 8.68 66.5 20.18 92.8 41.69 DFF
S06 Strengthen GFRP 27 7.2 51.58 14.79 88.9 44.47 DFF
S07 Strengthen GFRP + Anchors 33 7.5 44.35 11.54 99 46.84 DFF

Pcr and Δcr = cracking load and deformation, Py and Δy = yeild load and deformation, Pmax and Δmax = failure load and deformation, FF = Flexural Failure and DFF =
Debonding of Fiber and flexural failure.

The depth of the equivalent rectangular stress block of the com­ ( a) ( a)

pressed concrete is calculated as: Mu = As fs d − + Af ffe h − (14)
2 2

c=
As fs + Af ffe
(6) Table 14 showed the difference in load capacity between manual
α1 fc’ β1 b solution using ACI 440.2R-17 and ECP 208–2005 and experimental test
The ultimate flexural moment is calculated as: results. It is obvious that the same amplification effect in flexure due to
FRP strengthening obtained from (ACI) and (ECP). The strengthening
( β c) ( β c)
Mn = As fs d − 1 + ΨAf ffe h − 1 (7) amplification using manual solutions using (ACI) and (ECP) are less than
2 2 that obtained from experimental test result by about 25–30%. The effect
of adding anchors is not captured in manual solution using either ACI
where, εbi is the strain level in concrete substrate at the time of FRP
440.2R-17 or ECP 208–2005. Manual solution needed correlations in
installation., εcu is the maximum usable strain of concrete = 0.003, εc is
their equations to consider the effect of anchors in increasing the flexure
the strain level in concrete, εfe is the effective strain level in FRP at
strength. From experimental test, it can be suggested that the amplifi­
failure, εfd is the debonding strain limit of FRP, εfu is the design rupture
cation factor to consider the effect of anchors in flexure strengthening is
strain of FRP, Ec , Es , andEf is modulus elasticity of the concrete, steel and about 1.12. While; in order to get more specifically the amplification
FRP, respectively, h is the depth of section, d is the effective depth of factor of the influence of using the anchors; further experimental tests
section, ffe is effective stress in FRP at failure, fs are the stresses in steel are needed.
respectively, fc’ is specified compressive strength of concrete, f y is In addition, the values of load capacity using manual solution using
specified yield strength of steel reinforcement, Ψis FRP strength reduc­ (ACI) and (ECP) are underestimated less than that obtained from
tion factor = 0.85 for flexure, α1 is multiplier on fc’ to determine intensity experimental test result by about 32%. ECP 208 get lesser load capacity
of an equivalent rectangular stress distribution for concrete, β1 is ratio of than ACI 440.2R by about 4%; i.e. there is no main difference between
depth of equivalent rectangular stress block to depth of the natural axis, both codes. In both codes, there are reservations in values obtained from
As is the steel reinforcement area, Af is the area of FRP sheet = ntf wf ; manual solution equations in obtaining the flexure strength using FRP.
where n, tf , andwf are the number of FRP sheet, thickness and width of From experimental test, it can be suggested that the correction factor to
FRP sheet. calculate the flexure moment in case of strengthening by FRP is about
1.3 and this value needs more experimental tests to be approved.
[Link]. ECP 208–05 (EGYPT). The effective stress and strain in FRP at
failure are calculated from 4.2.2. Validation with analytical model
In this section a comparison between experimental and numerical
ffe = Ef εfe /γf (8) analysis for current experimental works will be illustrated in Table 15.
Fig. 14 showed the load-deflection relationship for slab S01, S03 and
h− c
εfe = εcu ( ) − εbi ≤ km εfu (9) S06 in both analytical end experimental tests, respectively.
c
As shown the maximum difference in load capacity between the
Where bond –depend coefficient analytical and the experimental is about 8% at yielding and 7% at
⎧ ( ) failure, while the maximum difference in cantilever deformation is
1 nEf tf
⎪
⎪
⎪
⎨ 60εfu 1 − ≤ 0.9nEf tf ≤ 180000 about 3% at failure. Therefore, the numerical model can be used to
360000
km = ( ) validate the experimental work and can be used for further parametric
⎪
⎪
⎪
⎩
1 90000
≤ 0.9nEf tf > 180000 studies.
60εfu nEf tf

The stress and strain in steel reinforcement are calculated from

4.3. Parametric studies
fy
f s = Es ε s ≤ (11)
γs Strengthening of cantilever slabs was depended on several variables
such as: concrete strength, reinforcement ratio, slab thickness and
(
εs = εfe + εbi (
) d− c
(12) cantilever length. This study was focused on slab thickness and canti­
h− c lever length. Group (4) “S01, S03, S08 and S09”, Group (5) “S10, S11,
The depth of the equivalent rectangular stress block of the com­ S12 and S13” and Group (6) “S14, S15, S16 and S17” are focused on the
pressed concrete is calculated as: effect of the cantilever length with maintain slab thickness constant as
120, 140 and 160 mm, respectively. While, Group (7) “S01, S03, S10,
As fs + Af ffe S11, S14 and S15” and Group (8) “S08, S09, S12, S13, S16 and S17” are
a= (13)
focused on the effect of the cantilever slab thickness with maintain
(0.67fcu b)
γc
cantilever slab length constant as 1000and 1500 mm, respectively. As
The ultimate flexural moment is calculated as:
shown in Tables 16 and 17, the results of the studied cases which divided
into 5 groups to focus on the certain variables.

2498
N.M. Ayash et al. Structures 28 (2020) 2488–2506

4.3.1. Effect of cantilever length

The Group (4) contains four slabs S01, S08 (without strengthening)
and S06, S09 (with strengthening) for the cantilever length (1000 mm

strengtheningTheor. /
and 1500 mm); respectively; with constant slab thickness 120 mm. Also,
the Group (5) contains four slabs (S10, S12) and (S11, S13) but with slab

Difference %
thickness 140 mm; in addition, the Group (6) contains four slabs (S14,
S16) and (S15, S17) with constant slab thickness 160 mm.

81%

73%
Exp.
[Link]. Cantilever load vs. deflection curve. Fig. 15 showed the canti­
lever deflection for slabs in group (4), (5) and (6), respectively. As
shown the failure loads in both control and strengthened slabs are
MmaxTheor. /
Difference %

reduced with increasing the cantilever length from 1000 mm to 1500

mm for all slab thicknesses average 70% for controlled specimens 73%
79%
64%
79%
58%
Exp.

for strengthened specimens and this reduction is due to more bending

moments. In the other words, with increasing the cantilever length, the
effect of strengthening caused smaller reduction in load carrying ca­
pacity than non-strengthened slabs. From these it can be reached that
strengthening

the strengthening does not significantly improve the load carrying load
in cantilever slabs with increasing cantilever length.
Theor.

125%

125%
% of

On the other hand, the cantilever deflection in both control and

strengthened slabs are increased with increasing the cantilever length
ECP 208–2005

from 1000 mm to 1500 mm for all slab thicknesses average 175% for
Mmax (kN.m)
Theoretical

controlled specimens 200% for strengthened specimens and this

amplification is due to more bending moments. With increasing the
22.84

22.84
18.2

18.2

cantilever length, the effect of strengthening caused larger amplification

in deflection than non-strengthened slabs; this is also due to effective
effect of fiber that increasing the slab ductility. It can be observed that
the strengthening led to significant improvement in cantilever deflection
strengtheningTheor. /

with increasing cantilever length.

Difference %

[Link]. Concrete and steel strain. Table 18 and Figs. 16–18 showed the
strain of concrete for slabs in group (4), (5) and (6), respectively.
85%

76%
Exp.

As shown, the maximum strains in concrete and reinforcement are at

the top negative point above the cantilever support in all control slabs
with different thicknesses and lengths (S01, S08, S10, S12, S14 and S16)
which indicated the failure location at this position. While in strength­
MmaxTheor. /
Difference %

ened slabs (S06, S09, S11, S13, S15 and S17), the maximum strains in
concrete and reinforcement are noted either occurred at top face of
73%
62%
73%
55%
Exp.

cantilever or at bottom face of simple slab adjacent to cantilever.

In cases of strengthened slabs and with increasing the cantilever
length (S09, S13 and S17) “length = 1500 mm“ compared to (S06, S11
and S15) “length = 1000 mm”, it is observed that the maximum strains
strengthening

were transferred from the top face of cantilever to the bottom face of
adjacent simple slab, this is due to using more fiber that increase the
Theor.

130%

130%
% of

ductility of cantilever and the failure transfer to mid span of adjacent

slab to cantilever.
Comparison between experimental result and theoretical solution.
ACI 440.2R-17

The compressive strain of concrete and the reinforcement strain were

Mmax (kN.m)
Theoretical

reduced in strengthening slabs compared to control slabs. With

increasing the cantilever length in strengthened slabs the strains were
16.79
21.91
16.79
21.91

reduced and these observed in all slabs thickness.

4.3.2. Effect of cantilever slab thickness

strengthening

The Group (7) contains six slabs S01, S10, S14 (without strength­
ening) and S03, S11, S15 (with strengthening) for slab thickness (120
154%

172%

mm, 140 mm and 160 mm); respectively; with constant slab length
% of

Exp.

1000 mm. As the previous, the Group (8) contains six slabs (S08, S12,
S16) and (S09, S13, S17) but with slab length 1500 mm.
Experimental
Mmax (kN.m)

[Link]. Cantilever load vs. deflection curve. Fig. 19 showed the canti­
lever deflection for slabs in group (7) and (8), respectively. For constant
23.08
35.56
23.08
39.6

cantilever length of 1000 mm and 1500 mm with the slab thickness

increased; the load carrying capacity are increased in both controlled
and strengthened slabs average 145% for controlled specimens 136% for
Specimen
Table 14

strengthened specimens. So, the strengthening does not have significant

change in the load carrying load in cantilever slabs with increasing their
S01
S06
S01
S07
No.

2499
N.M. Ayash et al. Structures 28 (2020) 2488–2506

Table 15
Comparison between experimental and numerical analysis.
Slab Specimen type Experimental kN Analytical kN Pyexp. Pfexp. Experimental mm Analytical mm Δyexp. Δfexp.
Pyanaly Pfanaly Δyanaly Δfanaly
PY PF PY PF ΔY ΔF ΔY ΔF

S01 Control 39.7 57.7 37.8 55 1.05 1.05 11.12 23.9 10.47 23.4 1.06 1.02
S03 Repaired 47.7 82.5 44 76.5 1.08 1.07 11.9 29.5 10.4 28.9 1.14 1.02
S06 Strengthened 51.6 88.9 50 88 1.03 1.01 14.8 44.5 12.8 43.0 1.16 1.04

Fig. 14. Load-Deflection relationships for slab S01, S03 and S06 experimentally and numerically.

Table 16
Load – deflection results for slabs based on parametric study.
Group Slab No. Slab Type Constant parameter Variable parameter Failure loadkN % Cantilever deflection mm %

4 S01 control Thickness 120 mm Cantilever length 1000 55.0 69 23.4 162
S08 control 1500 38.0 37.97
S06 strengthen 1000 88.0 61 43.0 81.5
S09 strengthen 1500 54.0 35.00
5 S10 control Thickness 140 mm Cantilever length 1000 86.0 73 18.00 183
S12 control 1500 63.0 32.94
S11 strengthen 1000 144.0 83 7.94 211
S13 strengthen 1500 120.0 16.79
6 S14 control Thickness 160 mm Cantilever length 1000 113.0 67 14.9 179
S16 control 1500 75.6 26.74
S15 strengthen 1000 120.5 75 6.75 138
S17 strengthen 1500 90.0 9.34
7 S01 control Cantilever length 1000 mm thickness 120 55.0 144 23.4 79
S10 control 140 86.0 18.00
S14 control 160 113.0 14.9
S06 strengthen 120 88.0 124 43.0 52
S11 strengthen 140 144.0 7.94
S15 strengthen 160 120.5 6.75
8 S08 control Cantilever length 1500 mm thickness 120 38.0 142 37.97 84
S12 control 140 63.0 32.94
S16 control 160 75.6 26.74
S09 strengthen 120 54.0 148 35.00 52
S13 strengthen 140 120.0 16.79
S17 strengthen 160 90.0 9.34

thickness. the slab thickness. It can be noted that the effect of strengthening with
As the slab thickness increased the cantilever deflection is reduced in increasing cantilever slab thickness has more improvement in cantilever
both controlled and strengthened slabs. With increasing the slab thick­ deflection.
ness; there are more reductions in strengthened cantilever slab than non-
strengthened slabs for both cantilever length of 1000 mm (from 77% to [Link]. Concrete and steel strain. Figs. 20 and 21 showed the strain of
27%) and 1500 mm (from 84% to 53%). This reduction is more signif­ concrete for slabs in group 7 and 8, respectively. For control cantilever
icant in short cantilever length 1000 mm than 1500 mm with increasing slabs, the typical of flexural failure is at the top negative point above the

2500
N.M. Ayash et al. Structures 28 (2020) 2488–2506

Table 17
Concrete and steel strain results for slabs based on parametric study.
Group Slab No. Slab Type Constant parameter Variable parameter Concrete strain Location Steel strain Location

4 S01 control Thickness 120 mm Cantilever length 1000 0.019972 Top of cantilever 0.001906 Top of cantilever
S06 strengthen 1000 0.043974 Top of cantilever 0.001928 Top of cantilever
S08 control 1500 0.024456 Top of cantilever 0.001932 Top of cantilever
S09 strengthen 1500 0.030069 Bottom of mid span 0.001291 Bottom of mid span
5 S10 control Thickness 140 mm Cantilever length 1000 0.045739 Top of cantilever 0.001728 Top of cantilever
S11 strengthen 1000 0.031138 Top of cantilever 0.001584 Bottom of mid span
S12 control 1500 0.028844 Top of cantilever 0.001925 Top of cantilever
S13 strengthen 1500 0.014827 Bottom of mid span 0.001909 Bottom of mid span
6 S14 control Thickness 160 mm Cantilever length 1000 0.078182 Top of cantilever 0.001755 Top of cantilever
S15 strengthen 1000 0.016844 Bottom of mid span 0.000979 Bottom of mid span
S16 control 1500 0.015468 Top of cantilever 0.00175 Top of cantilever
S17 strengthen 1500 0.018654 Bottom of mid span 0.00114 Top of cantilever
7 S01 control Cantilever length 1000 mm Thickness 120 0.021872 Top of cantilever 0.001706 Top of cantilever
S06 strengthen 120 0.045974 Top of cantilever 0.001725 Top of cantilever
S10 control 140 0.044739 Top of cantilever 0.001628 Top of cantilever
S11 strengthen 140 0.027138 Top of cantilever 0.001486 Bottom of mid span
S14 control 160 0.084182 Top of cantilever 0.001954 Top of cantilever
S15 strengthen 160 0.006844 Bottom of mid span 0.000779 Bottom of mid span
8 S08 control Cantilever length 1500 mm Thickness 120 0.022456 Top of cantilever 0.001822 Top of cantilever
S09 strengthen 120 0.020069 Bottom of mid span 0.000991 Bottom of mid span
S12 control 140 0.024844 Top of cantilever 0.001828 Top of cantilever
S13 strengthen 140 0.010827 Bottom of mid span 0.001807 Bottom of mid span
S16 control 160 0.015465 Top of cantilever 0.00192 Top of cantilever
S17 strengthen 160 0.013653 Bottom of mid span 0.00107 Top of cantilever

Fig. 15. Cantilever Load vs Deflection for Slabs in Group (4), (5) and (6).

cantilever support for different slab thickness and different slab length. 4.3.3. Verification of codes provision’s
In the control slab, with increasing thickness the compressive strain Table 18 presented the Comparison between analytical result using
of concrete is increased for short cantilever length and reduced for long ANSYS and theoretical solution using ACI 440.2R-17 and ECP 208-2005.
cantilever length. While the reinforcement strains are slightly changed This comparison is conducted in order to covers the applications of codal
with increasing the slab thickness. provisions in wide range of slabs with different dimensions. As shown
For strengthening slabs, with increasing slab thickness the the percentage of the failure flexure moment difference between manual
compressive strain of concrete and the reinforcement strain are reduced solution using ACI 440.2R-17 and ECP – 208-2005 Code and analytical
and transformed from top face of cantilever support to bottom face at result ranged between 20%. The effect of strengthening on the failure
mid span of slab for different cantilever length. flexure moment using manual solution by ACI 440.2R-17 and ECP – 208-
2005 Codes are under-estimated less than that obtained from analytical

2501
N.M. Ayash et al. Structures 28 (2020) 2488–2506

result using ANSYS by average 35%.

For the parametric study that focused in this part; the failure flexure
moments from the manual solution by ACI 440.2R-17 and ECP-208-

Difference of %
2005 Codes are not changed regardless the increase in cantilever

Theor. /Analy.
strengthening
length and this is opposite to the results obtained from finite element
analyses. This is due to the nonlinearity considerations that taken in

78%

88%

61%

71%

87%

75%

77%
analytical models and cannot capture in manual solutions. This
nonlinearity effect appears mainly due to change in slab cross section;
but for slabs having the same depth and variable length, the results do
Difference of %

Theor. /Analy.

not change.
For the second factor that studied which concerned on the slab
thickness; the manual solution by ACI 440.2R-17 and ECP-208-2005
Mmax

83%
65%
80%
70%
65%
40%
56%
40%
56%
49%
87%
65%
63%
Codes can predict the increases the failure flexure moment for both
control and strengthened slabs with increasing the slab thickness as
behavior not as values which are under-estimated than that obtained
strengtheningTheor.

from analytical result.

5. Conclusions
125%

125%

102%

129%

134%

133%

This study focused on the behavior of reinforced concrete cantilever

% of

slabs repaired or strengthened using GFRP experimentally and numer­

ically. The results presented in this research are limited to the di­
ECP 208-2005

mensions, loads and conditions studied. The following can be concluded:

Theoretical

moment
Failure

22.84

22.76
22.27
22.81
22.27
28.77
26.34
35.22
26.34
35.09
kN.m

18.2

18.2

1. Strengthening and repairing techniques using GFRP as a top mesh

on cantilever slabs has a significant effect on increasing the load
carrying capacity. Also it has a considerable effect on deflection,
ductility and stiffness.
2. Strengthening techniques using GFRP as a top mesh on cantilever
Difference of %

Theor. /Analy.
strengthening

slabs either in cases without or with anchors has a significant

effect more than repairing techniques using GFRP as a top mesh
on cantilever slabs either in cases without or with anchors. This is
82%

92%

80%

74%

90%

78%

82%

based on the increasing of the load carrying capacity, ductility

and stiffness.
Difference of %

3. The effects of anchors increased the stress concentration around

Theor. /Analy.

them. At failure the fibers are cutting at location of anchors while

when non using anchors the fiber is debonding only.
Mmax

76%
62%
74%
67%
59%
48%
52%
38%
52%
46%
80%
62%
60%

4. The effects of anchors are significant in slabs with GFRP

strengthened techniques more than in slabs with GFRP repaired
techniques.
5. Strengthened or repaired slabs failed firstly due to debonding in
strengtheningTheor.

fibers or cutting in fiber at anchors locations then crushing in

concrete. This failure in fiber led to more ductile behavior of the
Comparison between analytical result using ANSYS and theoretical solution.

slab. As this ductility is associated with a small load increase,

130%

130%

134%

134%

139%

139%

debonding can be considered as a desirable feature in terms of

% of

giving warning of imminent structural collapse.

6. The strengthening using FRP that amplified the flexure strength
ACI 440.2R-17

by theoretical solutions using (ACI) and (ECP) is less than that

Theoretical

obtained from experimental test result by about 25–30% i.e. there

moment
Failure

16.79
21.91
16.79
21.86
20.45
27.49
20.45
27.49
24.12
33.44
24.12
33.42
kN.m

must be applied amplification factor ranging by 1.25–1.3 in both

codes equations to predict the flexure strength. The amplification
effect in flexure strength is the approximately the same when
using (ACI) and (ECP) equations.
strengtheningAnaly.

7. The effect of adding FRP anchors is not captured in manual so­

lution using either (ACI) or (ECP). There are needed to correlate
the theoretical equations to consider the anchors effect. From
experimental results, it can be reached that the amplification
160%

142%

167%

182%

154%

179%
% of

factor to consider the anchors effect is about 1.12 multiplies to

that obtained for cases with FRP sheet only and this must verify
by more experimental tests.
Analytical

8. An analytical model using ANSYS software was presented a

moment
Failure

30.24
kN.m

reliable prediction of the failure load and deflection, failure

35.2
22.8
32.4
34.4
57.6
39.6

46.8
72.3
22

72

54

modes and can capture the de-bonding failure of the strengthened

slabs.
Table 18

Average

9. The maximum concrete / reinforcement strains are transformed

Slab

S01
S06
S08
S09
S10
S11
S12
S13
S14
S15
S16
S17
No.

from top face of cantilever slab above support location to the

2502
N.M. Ayash et al. Structures 28 (2020) 2488–2506

Fig. 16. Concrete and steel strain distributions for Slabs in Group (4).

Fig. 17. Concrete and steel strain distributions for Slabs in Group (5).

bottom face for middle span of slab adjacent to cantilever because decreased with increasing the slab cantilever length and trans­
of existence of fiber as strengthening above cantilever length. ferred from the top face of cantilever slab above support location
10. From parametric studies, using fiber glass in strengthening to the bottom face for middle span of slab adjacent to cantilever.
cantilever slabs at constant slab thickness with variable canti­ 11. Using GFRP in strengthening cantilever slabs at constant canti­
lever length, as the cantilever length increased the reduction in lever length with variable slab thickness, there is a slightly in­
load bearing capacity has slightly changes. The amplification in crease in load capacity with more reduction in cantilever
cantilever deflection is increased leading to more slab ductility deflection with increasing the slab thickness. The maximum
obtained. The maximum concrete and reinforcement strain are concrete and reinforcement strain are decreased with increasing

2503
N.M. Ayash et al. Structures 28 (2020) 2488–2506

Fig. 18. Concrete and steel strain distributions for Slabs in Group (6).

Fig. 19. Cantilever Load vs. Deflection for Slabs in Group (7) and (8).

the slab thickness and transferred from the top face of cantilever 3. Using near surface mounted FRP as method of strengthening or
slab above support location to the bottom face for middle span of repairing of cantilever slabs.
slab adjacent to cantilever. 4. More other parametric analysis is recommended such as miss-
12. Increasing in cantilever length with remaining the constant slab positioning of cantilever steel, concrete and steel material proper­
thickness showed no differences in the values of the failure ties and length of FRP wrapped length, width and configurations.
flexure moment that obtained from the manual solution by (ACI)
and (ECP) Codes. CRediT authorship contribution statement
13. The manual solution by (ACI) and (ECP) Codes can predict the
increases in the failure flexure moment for both control and Nehal M. Ayash: Methodology, Supervision, Writing - original draft.
strengthened slabs with increasing the slab thickness as behavior Ahmed M. Abd-Elrahman: . Abd-Elaziz Soliman: Methodology, Su­
not as values. pervision, Writing - review & editing.

6. Recommendations for future works Declaration of Competing Interest

1. More layers of GFRP can be used in the strengthening or repairing the The authors declare that they have no known competing financial
cantilever slabs. interests or personal relationships that could have appeared to influence
2. Comparing between different types of fibers based on the capacity the work reported in this paper.
increasing and its cost.

2504
N.M. Ayash et al. Structures 28 (2020) 2488–2506

Fig. 20. Strain distributions for Slabs in Group (7).

Fig. 21. Strain distributions for Slabs in Group (8).

2505
N.M. Ayash et al. Structures 28 (2020) 2488–2506

Acknowledgment [8] Tarek Ali and Sameh Yehia (2016), Study on Strengthening of RC Slabs with
Different Innovative Techniques, Open Journal of Civil Engineering. Vol.6 No.4,
September 2016.
The authors gratefully acknowledge the technical staff of the struc­ [9] Aymen H. Khalil, Ashraf M. Heniegal, Bassam A. Abdelsalam (2018) Experimental
tural laboratory in the Civil Engineering Department at Helwan Study on Strengthening Systems of Reinforced Concrete Cantilever Slabs.
University. Proceeding of 10th international conference on Nano-technology in construction
(NTC 2018) 13-17 April 2018, Hurgada, Egypt.
[10] Fathelbab FA, Ramadan MS, Al-Tantawy A. Strengthening of RC bridge slabs using
References CFRP sheets. Alexandria Eng J 2014;2014(53):843–54.
[11] Lima MM, Doh J-H, Hadi MNS, Miller D. The effects of CFRP orientation on the
[1] Teng JG, Lam L, Chan W, Wang J. Retrofitting of deficient RC cantilever slabs using strengthening of reinforced concrete structures. Struct Design Tall Spec Build 2016;
GFRP strips. J Compos Constr 2000;4(2):75–84. 2016(25):759–84.
[2] Lam L, Teng JG. Strength of RC cantilever slabs bonded with GFRP strips. J Compos [12] Tank T, Modhera CD. Finite element modelling of RC slab strengthened with GFRP.
Constr 2001;5(4):221–7. Mater Today: Proc 2017;4(9):9784–91.
[3] Teng J, Cao S, Lam L. Behavior of GFRP-strengthened RC cantilever slabs. Constr [13] Emarah DA. Numerical study to investigate the behavior of reinforced concrete
Build Mater 2001;15(7):339–49. slabs with CFRP sheets. Water Sci 2019;33(1):142–53.
[4] A. Moussa (2003) “Repair and strengthening of Cantilever Reinforced Concrete [14] ACI 440.2R-17. Guide for the design and construction of externally bonded FRP
Slabs” Scientific Research Bulletin, Faculty of Eng., Ain-Shams University, Cairo, systems for strengthening concrete structures. Farmington Hills, MI: American
Egypt. Dec. 2003. Concrete Institute; 2017. pp. 76.
[5] El Maaddawy T, Soudki K. Strengthening of reinforced concrete slabs with [15] ECP 203-2018 “Egyptian Code of Practice for Design and Construction of Concrete
mechanically-anchored unbonded FRP system. Constr Build Mater 2008;22(4): Structures Code”, Standing Committee to prepare the Egyptian,“ HBRC, Cairo-
444–55. Egypt, 2018.
[6] Smith ST, Hu S, Kim SJ, Seracino R. FRP-strengthened RC slabs anchored with FRP [16] ECP 208-2005 “Egyptian Code Egyptian Code for the use of fiber reinforced
anchors. Eng Struct 2011;33(4):1075–87. polymer (FRP) in the construction fields”, Standing Committee to prepare the
[7] Sergio F. Bren and Geoffrey N. McGuirk (2013) Advances on the Behavior Egyptian,“ HBRC, Cairo-Egypt, 2005.
Characterization of FRP-Anchored Carbon Fiber-Reinforced Polymer (CFRP) Sheets [17] ANSYS R15.0, “Release 14.5 Documentation for ANSYS,” in ANSYS, 15.0 ed.
Used to Strengthen Concrete Elements. International Journal of Concrete Canonsburg, PA, USA, 2015.
Structures and Materials Vol.7, No.1, pp.3–16, March 2013.

2506