See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/304526935

Parametric Study on Flexural Response of FRP Reinforced ECC Beams

Article · January 2009

CITATION READS
1 241

1 author:

Shamsher Bahadur Singh

Birla Institute of Technology and Science Pilani
245 PUBLICATIONS 1,163 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Response and micromechanics Based Design of Engineered Cementitious Composite Structures View project

Failure and Stability of Laminated Composite Plates with Cutouts View project

All content following this page was uploaded by Shamsher Bahadur Singh on 12 July 2016.

The user has requested enhancement of the downloaded file.

 Experimental and Parametric Investigation of Response of
NSM CFRP-Strengthened RC Beams
S. B. Singh, M.ASCE 1; A. L. Reddy 2; and C. P. Khatri 3
Downloaded from ascelibrary.org by Birla Institute of Technology - Pilani on 06/18/16. Copyright ASCE. For personal use only; all rights reserved.

Abstract: The present investigation deals with carbon-fiber-reinforced polymer (CFRP) rebar manufacturing, experimental flexural and
shear response of near-surface-mounted (NSM) CFRP-strengthened reinforced concrete (RC) beams, and detailed parametric study of
NSM-strengthened RC beams using Abaqus. The CFRP bars were manufactured locally using the twisting technique. Parametric study
was performed by considering various key parameters which affect the strengthening of the RC beams such as a CFRP bar’s orientation,
geometry, intergroove separation, type of groove, cross-section area of bars, and material strength. The parametric study is based on the
nonlinear finite-element modeling using Abaqus for examining the flexural and shear behavior. All beams were tested under a two-point
loading system. The numerical results obtained using Abaqus are validated with the experimental results. The present study results show a
significant increase in the strength of RC beams strengthened with NSM CFRP bars. Furthermore, based on this study, guidelines for realistic
structural applications and optimization of the use of NSM CFRP bars are proposed. DOI: 10.1061/(ASCE)CC.1943-5614.0000411. © 2013
American Society of Civil Engineers.
Author keywords: CFRP bars; Flexural strength; Nonlinear finite-element method; Near-surface mounting; Shear strength; Structural
strengthening.

Introduction years; therefore De Lorenzis et al. (2002) have experimentally and

numerically determined the bond slip relations for NSM FRP rods
Most of the existing structures were constructed well before the embedded in concrete.
1970s, and hence they are based on gravity load design (GLD); One of the issues with strengthening techniques is achieving
consequently, they lack many important factors such as seismic appropriate ductility. In contrast to other methods, NSM provides
loading and ductile detailing. Furthermore, over time they have lost more ductility due to its transverse strengthening and concrete
their load-carrying capacity. El-Attar et al. (1997) have experimen- confinement effects. El-Hacha and Rizkalla (2004) and Rasheed
tally depicted the behavior and reasons behind inefficiency of et al. (2010) have demonstrated that the NSM technique provides
GLD-based RC buildings subjected to earthquake. Since rehabili- a significant increase in ductility compared to other externally
tation is faster, cheaper, and also the original design is retained, bonded composite systems. When the RC beams are subjected
enhancing the performance of these structures is a better alternative to high strains, failures between the FRP-concrete interfaces
over reconstructing them. Various techniques are being used are observed, which depends on the bar diameter and its orientation
for upgrading the structures, but recently, near-surface-mounting and other parameters (Singh et al. 2012). Many analytical models
(NSM) of FRP bars has emerged as a very effective solution. are also proposed in these regards. Rasheed et al. (2004) have
NSM is not only easy for installation, application, and handling proposed an analytical model using an experimentally obtained
but also resolves the ductility issues. Ceroni et al. (2012) have ex- bilinear moment-curvature relationship for FRP-strengthened RC
perimentally shown the effectiveness of the near-surface-mounted sections in predicting the load–deflection response of concrete
technique over other externally bonded systems (EB). Other exter- beams reinforced with FRP bars. Weidong et al. (2010) and
nally bonded FRP systems are even prone to external damage from Ferracuti et al. (2006) have proposed numerical models to analyze
environmental factors such as fire, moisture, external impact, etc. the behavior of FRP bars and FRP plate with concrete, respectively,
De Lorenzis and Teng (2007) have extensively investigated the for development of bond stress versus slip relations. NSM
performance of NSM and other externally bonded reinforcement technique is both practically efficient and economical for strength-
systems. They have depicted NSM’s importance both in shear as ening the negative moment region of beams and slabs. Hence, it is
well as flexure strengthening. Because of immense applications necessary to standardize its application and usage for strengthening
of this technology, huge research has been conducted over the concrete structures.
Although NSM has become one of the most discussed topics of
1
Professor, Civil Engineering Dept., Birla Institute of Technology and research recently, and there are many works depicting its impor-
Science, Pilani Rajasthan 333031, India (corresponding author). E-mail: tance, very few papers actually suggest the effective guidelines
[email protected] for applying them in real structures. American Concrete Institute
2
Graduate Student, Civil Engineering Dept., Birla Institute of Technol- (ACI) Committee 440 [ACI 440.2R (ACI 2008b)] has given some
ogy and Science, Pilani Rajasthan 333031, India. design guidelines for external strengthening of structures using
3
Graduate Student, Civil Engineering Dept., Birla Institute of Technol-
FRP, but there are some more areas which need addition. The fac-
ogy and Science, Pilani Rajasthan 333031, India.
Note. This manuscript was submitted on December 26, 2012; approved tors which govern the performance of NSM are taken into consid-
on June 28, 2013; published online on July 2, 2013. Discussion period open eration in the current study. It deals with standardizing the CFRP
until March 2, 2014; separate discussions must be submitted for individual bar application for various strengthening purpose. The RC beams
papers. This paper is part of the Journal of Composites for Construction, were strengthened with NSM CFRP bars, and their flexural and
© ASCE, ISSN 1090-0268/04013021(11)/$25.00. shear responses were evaluated experimentally and numerically.

© ASCE 04013021-1 J. Compos. Constr.

J. Compos. Constr., 2014, 18(1): 04013021

 The beams are strengthened with 2,000 MPa bars throughout the is cured, it is cut at the ends. The typical length of manufactured
numerical study, unless explicitly specified. Different schemes CFRP bar is 2 m, while the nominal diameter of CFRP bars is 7, 8,
for strengthening in flexure and shear were considered. Various and 10 mm consisting of 75, 85, and 100 numbers of carbon fibers,
parameters such as bar diameter (BD), intergroove spacing (GS), respectively. The ASTM D3916 (ASTM 2008) method for tensile
groove width (GW), bar strength (BS), groove distance from edge testing of FRP bars is implemented, and tensile strengths of various
(GD-E), groove orientation (GO) in shear with grooves being CFRP bars manufactured at STC are presented in Table 1.
vertical (SV), grooves inclined at an angle (SI-GO), and mixed
pattern (SM-GO), affecting the performance of strengthened beams
are considered for the present study. What should be the shape of Experimental Study of Beams
the bars used for strengthening? It is one of the most important Three different categories of beams were cast based on their
concerns for the designers. Hence, in the current study, some reinforcement detailing, and they are classified as (1) underrein-
crucial conclusions are drawn based on the numerical results
Downloaded from ascelibrary.org by Birla Institute of Technology - Pilani on 06/18/16. Copyright ASCE. For personal use only; all rights reserved.

forced beams: three numbers, (2) overreinforced beams: three

obtained for this scheme [F–bar cross section (F-BCS)]. numbers, (3) shear beams: 10 numbers. Underreinforced and
Nonlinear finite-element modeling was performed using overreinforced beams were designed to fail in flexure, while the
Abaqus (Dassault Systèmes Simulia Corporation 2010), and the shear beams were designed to fail in shear as per IS-456 (Bureau
obtained results are validated with the experimental ones. Multiple of Indian Standards 2000). All sixteen test beams were of 1,300 ×
validation models are produced to reach proper accuracy in the 130 × 130 mm dimensions. One or two unstrengthened beams in
properties. NSM is a novel technology, and much more research each category acted as control beams. In nomenclature of beams, E
is yet to be done to validate the existing guidelines and to propose refers to the specimens tested experimentally. Details and nomen-
additional ones. Therefore, current parametric results will be of the clature regarding the beam specimens are presented in Table 2.
most practical use in the design of NSM FRP-strengthened concrete Underreinforced beams are reinforced with two numbers of
structures. As this technique has several advantages over other 8-mm steel bars, while overreinforced and shear beams are
conventional strengthening schemes, the results obtained through reinforced with three numbers of 12-mm steel bars as internal
the current study will definitely assist in setting up the guidelines reinforcement. Complete reinforcement details of beam specimens
so that the overall usage of this technique is standardized. are provided in Table 3 and Fig. 1. An effective cover of 30 mm to
the center of the steel reinforcement is provided from the bottom for
all beams. The compressive and tensile strength of concrete are 30
Experimental Investigation and 3.83 MPa, respectively. The yield and ultimate tensile strength
of steel are 415 and 425 MPa, respectively.
The entire experimental investigation of this study was carried out
in four subsequent steps: (1) CFRP bars were manufactured locally
at the Structural Testing Centre (STC) using twisting technique; Installation of NSM CFRP Bars on RC Beams
(2) sixteen half-scale reinforced concrete beams were cast;
To install the NSM FRP rebars, a groove is cut into the concrete
(3) beams were strengthened with NSM CFRP bars by inserting
surface in the desired direction, and then the groove is filled half-
the bars in beams in various patterns; and (4) tests were performed
way with epoxy; thereafter the FRP rod is placed in the groove, and
on control and NSM CFRP-strengthened RC beams. Details of
light pressure applied. This forces the resin to flow around the rod
bar and beam test specimens are presented in Tables 1 and 2,
and fill completely between the rod and the sides of the groove.
respectively.
After leveling the groove by filling it with a large enough amount
of epoxy, it is left for the epoxy to harden. Generally, the size of the
Manufacturing and Testing of CFRP Bars groove is governed by the diameter of the FRP bar. ACI 440.2R
(ACI 2008b) recommends the groove width and depth to be at least
In the current study, mechanically developed specialized twisting 1.5 times the bar diameter. This study incorporates the same groov-
equipment is used in the manufacture of bars by hand lay-up ing patterns suggested by ACI 2R-08 (2008b).
process. In this technique, hooks are connected to the handle at For predicting the flexural strength, out of six beams designed to
the ends of the equipment, and CFRP fiber rovings are tensioned fail in flexure, two beams acted as control beams. Further, one
between the hooks. Adhesive resin is prepared by mixing epoxy beam from each category (i.e., underreinforced and overreinforced)
and hardener TriEthyleneTetraAmine (TETA) with a dosage of is strengthened with an 8-mm-diameter FRP bar, while another
hardener 4% by weight of epoxy. These hooks are movable in beam from each category is strengthened with a 10-mm-diameter
the sense that after tensioning the carbon fibers between the two bar. Hence, two different sizes of grooving are done in beams de-
ends, these hooks can be held apart by rotating and outward twist- signed for flexure failure. The sizes of groove were 12 × 12 mm for
ing. This way the CFRP fibers also get entangled with each other 8-mm FRP bars and 15 × 15 mm for 10-mm FRP bars. The spacing
and then get stretched outward causing excess resin to drip out. between the two grooves is 50 mm for both types as indicated
This also helps in maintaining uniform bar diameter without in Fig. 1.
sagging. After the proper number of rotations, the hooks and Similarly, two beams out of the ten beams designed to fail in
handles are fixed, and the entire set is left to cure. Once the bar shear acted as control beams. Further, three different strengthening
patterns and two different groove sizes (Table 2 and Fig. 2) are
Table 1. Tensile Strength of CFRP Bars made within eight remaining shear beams for two different bar
diameters (8 and 10 mm). The first pattern grooves (SI) are made
Serial Effective Tensile Young’s
at an inclination of 45° having a spacing of 100 mm between them.
number Specimen diameter (mm) strength (MPa) modulus (GPa)
The second pattern grooves (SV) are made vertically with a spacing
1 Specimen 1 7 918 45 of 50 mm on one side while it is 100 mm on the other side, so that
2 Specimen 2 7 1,016 44 the failure can be observed on the latter side of the beam; while the
3 Specimen 3 8 1,036 43 third pattern consists of both vertical and inclined patterns,
4 Specimen 4 8 1,025 42
i.e., mixed pattern (SM).

© ASCE 04013021-2 J. Compos. Constr.

J. Compos. Constr., 2014, 18(1): 04013021

 Table 2. Nomenclature of the Beam Specimens Tested Experimentally
Specimen number Beam ID Details of the strengthening patternsa
1 E-CBURF1 Underreinforced flexure control beam
2 E-URF2 Underreinforced flexure strengthened with NSM-FRP of 8-mm diameter
3 E-URF3 Underreinforced flexure strengthened with NSM-FRP of 10-mm diameter
4 E-CBORF1 Overreinforced flexure control beam
5 E-ORF2 Overreinforced flexure strengthened with NSM-FRP of 8-mm diameter
6 E-ORF3 Overreinforced flexure strengthened with NSM-FRP of 10-mm diameter
7 E-CBS1, E-CBS2 Control beam for shear-1 and control beam for shear-2
8 E-SV3 SB strengthened in vertical position of with 8-mm bars
9 E-SV4 SB strengthened with vertical position of NSM-FRP bars (8-mm)
10 E-SV5, E-SV6 SB strengthened with vertical position of NSM-FRP bars (10-mm)
Downloaded from ascelibrary.org by Birla Institute of Technology - Pilani on 06/18/16. Copyright ASCE. For personal use only; all rights reserved.

11 E-SI7 SB strengthened with bars (8-mm) 45° inclined to longitudinal axis of beam
12 E-SI8 SB strengthened with NSM-FRP bars (10-mm) inclined at 45°
13 E-SM9 SB strengthened with bars (8-mm) inclined at both 45° and 90°
14 E-SM10 SB strengthened with bars (10-mm) inclined at both 45° and 90°
a
SB = shear beam.

Table 3. Reinforcement Details of Beam Specimens

Type Number of beams Compression steel Tension steel Shear steel
Underreinforced 3 Two numbers of 8-mm diameter Two numbers of 8-mm diameter 6-mm diameter at 50-mm c=c
Overreinforced 3 Two numbers of 8-mm diameter Three numbers of 12-mm diameter 6-mm diameter at 50-mm c=c
Shear beams 10 Two numbers of 8-mm diameter Three numbers of 12-mm diameter 6-mm diameter at 100-mm c=c

Fig. 1. Reinforcement details of beams

Fig. 2. Various types of strengthening patterns in shear

© ASCE 04013021-3 J. Compos. Constr.

J. Compos. Constr., 2014, 18(1): 04013021

 Table 4. Load Spans in Various Specimens Table 5. Load Carrying Capacity and Failure Modes of NSM CFRP-
Strengthened RC Beams
Load span
(mm) Specimens Diameter of Peak Load
Beam FRP (mm) (kN)a Failure mode
150 Flexure beams (underreinforced and overreinforced beams)
410 Shear control beams, shear beams with NSM FRP bars E-CBURF1 Not applicable 30 Flexural tension
inclined at 45° and with mixed pattern (vertical and E-URF2 8 51 FRP rupture
inclined at 45°) E-URF3 10 54 FRP debonding
500 Shear beams with vertical pattern (NSM FRP bars at 90°) E-CBORF1 — 50 Concrete crushing
E-ORF2 8 71 Concrete crushing
E-ORF3 10 68 FRP debonding
E-CBS1 — 70 Shear failure
Instrumentation and Testing E-CBS2 — 76 Shear failure
Downloaded from ascelibrary.org by Birla Institute of Technology - Pilani on 06/18/16. Copyright ASCE. For personal use only; all rights reserved.

E-SV3 8 89 Shear failure

An LVDT (stroke 100 mm, sensitivity 0.1 mm) was used to E-SV4 8 94 Shear failure
measure the vertical deflection kept at the soffit of the beam at E-SV5, E-SV6 10 95, 90 Concrete crushing
the midspan. The applied load was measured by a load cell kept E-SI7 8 88 Concrete crushing
under one of the supports. A hand-operated hydraulic jack was used E-SI8 10 75 FRP debonding
to apply the load. A sixteen-channel data acquisition system was E-SM9 8 109 Concrete crushing
used to collect the data. All beams were subjected to monotonic E-SM10 10 86 Concrete crushing
loading till failure. Two-point loading was applied for all the a
Ten kN ¼ One ton.
beams. Three different load spans (Table 4) were used for the test-
ing of beams (Fig. 3) to ensure desired flexural or shear failures in a
particular beam. These different load spans were provided with a
sufficient number of stirrups to avoid premature shear failure in the
midspan zone.

Experimental Results

Response of Beams Strengthened in Flexure

As shown in Table 5, the NSM FRP-strengthened underreinforced
beams E-URF2 and E-URF3 have about 69 and 80% higher
load-carrying capacity than that of the corresponding control beam
(E-CBURF1), respectively. It is also shown that underreinforced
beams strengthened with 8-mm-diameter NSM bars failed by
rupture of FRP bars, while those strengthened by 10-mm-diameter
Fig. 4. Flexural response of NSM-strengthened overreinforced RC
bars failed by debonding of NSM FRP bars.
beams
In the case of overreinforced beams (Fig. 4), the beam E-ORF2
exhibits 42% higher load capacity than that of the corresponding
control beam (E-CBORF1). However, the beam E-ORF3 strength-
ened with 10-mm-diameter NSM bars failed due to premature de- The load versus deflection response of the shear beams strength-
bonding of the NSM bars (Fig. 5b) and demonstrates 36% higher ened with 8-mm-diameter NSM CFRP bars is shown in Fig. 6.
load capacity than that of the control beam. Modes of failure of all Similarly, the beams (E-SV5 and E-SV6) strengthened using
the schemes are mentioned are mentioned in Table 5. 10-mm-diameter NSM CFRP bars in the vertical direction have
average load capacity of about 27% higher than that of the control
Response of Shear Beams shear beams. Similarly, the beams E-SI7 and E-SI8 strengthened
The peak loads and failure modes of beams designed to fail in using 8- and 10-mm NSM FRP bars with inclined strengthening
shear are presented in Table 5. It is observed that the average pattern are observed to have increased load-carrying capacities
load-carrying capacity of the beams E-SV3 and E-SV4 is 26% by about 20 and 3%, respectively. The ultimate failure modes of
higher than that of the control beams (E-CBS1 and E-CBS2). beams E-SI7 and E-SI8 are concrete crushing and FRP debonding,

Fig. 3. Load setup of beam for flexure and shear failure test

© ASCE 04013021-4 J. Compos. Constr.

J. Compos. Constr., 2014, 18(1): 04013021

Downloaded from ascelibrary.org by Birla Institute of Technology - Pilani on 06/18/16. Copyright ASCE. For personal use only; all rights reserved.

Fig. 7. Typical deflected shape and failure pattern of RC beams

strengthened with NSM FRP vertical bars in shear

the effective strengthening of FRP bars orientated in both vertical

and inclined directions providing better shear strength and prevent-
ing early development of tensile cracks at the bottom of the
strengthened beams. The failure of the beams E-SM9 and
E-SM10 is governed by the flexural compression due to crushing
of concrete at the top of the beam.

Parametric Study

In this section, the effect of various parameters such as BD, BS,

Fig. 5. (a) FRP rupture in E-URF2; (b) premature FRP debonding in BCS, groove inclination in shear (GO-SI), GW, GS, GD-E, and
E-ORF3 type of beam underreinforced (UR) / overreinforced (OR) are
evaluated using general purpose code Abaqus (Dassault Systèmes
Simulia Corporation 2010). The results obtained by Abaqus are
validated with the experimental studies performed. Table 6 depicts
the nomenclature and overview of all the parameters considered as
part of the study.

Numerical Modeling of Strengthened RC Beams Using

Abaqus
The behavior of composite beam with varying bar diameters, bar
shape and orientation, intergroove separation, and groove widths
cannot be fully covered experimentally as this would be both
time-consuming and expensive. Hence, besides the experimental
investigations, a validated numerical modeling is essential for
further study on behavior of flexure- and shear-strengthened
beams with different parameters. Therefore, numerical modeling
Fig. 6. Response of RC beams strengthened with 8-mm NSM FRP using Abaqus (Dassault Systèmes Simulia Corporation 2010) is
bars in shear incorporated for detailed parametric study.
The concrete-damaged plasticity model of Abaqus (Dassault
Systèmes Simulia Corporation 2010) is used to accurately account
respectively. The mode of failure in scheme E-SV with 8-mm bars for nonlinear behavior under compression and cracking under
is shear failure, while with 10-mm bars the mode of failure is con- tension. These two kinds of failure (Table 5) are observed through
crete crushing (Fig. 7). It is also observed that the premature FRP the experimental investigations. Hence, for the present study, a
debonding failure of the beam E-SI8 leads to the reduced concrete-damaged plasticity model is used for the modeling of
load-carrying capacity as compared to beam E-SI7, which failed concrete.
in flexural compression. From Table 5, it is also observed that Cohesive elements are used to model the interface between
the beams E-SM9 and E-SM10 strengthened using both vertical concrete and adhesive (epoxy). A linear elastic traction-separation
and inclined strengthening patterns (mixed pattern) have 49 and law prior to damage and a linear damage evolution based on energy
19% higher load capacity than the average load capacity of dissipation is assumed for defining the interface behavior as
the shear control beams, respectively. Thus it is shown that the mentioned in section 29.5.1 of the Abaqus analysis user’s manual
mixed-shear-strengthening pattern is more effective than the (Dassault Systèmes Simulia Corporation 2010). When a cohesive
vertical-shear-strengthening pattern. This can be explained with element is completely damaged, then that element is deleted.

© ASCE 04013021-5 J. Compos. Constr.

J. Compos. Constr., 2014, 18(1): 04013021

 Table 6. Nomenclature of the Beam Specimens Modeled Using ABAQUS
Beam Details of the strengthening patterns
NF-CBOR1 Numerical model of overreinforced flexure control beam
F-BD-6 Beam strengthened in flexure with 6-mm bars of 2,000-MPa tensile strength
F-BD-8 Beam strengthened in flexure with 8-mm bars of 2,000-MPa tensile strength
F-BD-10 Beam strengthened in flexure with 10 mm bars of 2000 MPa tensile strength
F-BD-12 Beam strengthened in flexure with 12 mm bars of 2000 MPa tensile strength
F-BS-1500 Beam strengthened in flexure with 1500 MPa tensile strength
F-BS-2500 Beam strengthened in flexure with 2500 MPa tensile strength
F-BS-2800 Beam strengthened in flexure with 2800 MPa tensile strength
F-GS 1.5d Beam strengthened in flexure with three bars of 2000 MPa tensile strength and clear distance between grooves at bottom is 1.5d
F-GS 2.5d Beam strengthened in flexure with three bars of 2000 MPa tensile strength and clear distance between grooves at bottom is 2.5d
Downloaded from ascelibrary.org by Birla Institute of Technology - Pilani on 06/18/16. Copyright ASCE. For personal use only; all rights reserved.

F-GW 1.3d Over-reinforced (OR) flexure strengthened with NSM-FRP (2000 MPa) of groove width 1.3d
F-GW 1.5d OR flexure strengthened with NSM-FRP of groove width 1.5d
F-GW 1.7d OR flexure strengthened with NSM-FRP of groove width 1.7d
F-GW 2d OR flexure strengthened with NSM-FRP of groove width 2d
F-GD-E1d OR flexure strengthened with NSM-FRP of groove distance from edge 1d
F-GD-E2d OR flexure strengthened with NSM-FRP of groove distance from edge 2d
F-GD-E3d OR flexure strengthened with NSM-FRP of groove distance from edge 3d
F-OR-8 OR flexure strengthened with 1025 MPa NSM-FRP 8 mm bar
F-BCS − 10 × 3 Beam strengthened with NSM-FRP bars of cross section of 10 mm × 3 mm
F-BCS-10 × 4 Beam strengthened with NSM-FRP bars of cross section of 10 mm × 4 mm
F-BCS − 7 × 7 Beam strengthened with NSM-FRP bars of cross section of 7 mm × 7 mm
F-BCS − 10 × 5 Beam strengthened with NSM-FRP bars of cross section of 10 mm × 5 mm
NS-CBOR1 Numerically modeled over-reinforced control beam for shear
SV-GW 1.3d Shear beam strengthened in vertical position of NSM-FRP of groove width 1.3d
SV-GW 1.5d Shear beam strengthened in vertical position of NSM-FRP of groove width 1.5d
SV-GW 2d Shear beam strengthened in vertical position of NSM-FRP of groove width 2d
SV-GW 2.5d Shear beam strengthened in vertical position of NSM-FRP of groove width 2.5d
SV-GS 1s Shear beam strengthened in vertical position of NSM-FRP of groove spacing 1s (s is spacing between two stirrups.)
SV-GS 1.5s Shear beam strengthened in vertical position of NSM-FRP of groove spacing 1.5s
SV-GS 2s Shear beam strengthened in vertical position of NSM-FRP of groove spacing 2s
SI-GO 30 Shear beam strengthened in inclined position of groove orientation 30°
SI-GO 45 Shear beam strengthened in inclined position of groove orientation 45°
SI-GO 52.4 Shear beam strengthened with inclined position of groove orientation 52.4°
N-SV Shear beam strengthened with vertical NSM-FRP bars
N-SM-GO 52.4 Shear beam strengthened in mixed pattern of groove orientation 52.4°
Note: Strength of the NSM-FRP bars used for numerical study is 2000 MPa, unless specified.

The results obtained from the numerical study are compared The ultimate concrete compressive and tensile strengths for a
with those obtained from the experimental results of the control concrete beam model are calculated by Eqs. (1) and (2), respec-
and strengthened beams. tively (ACI 318) (ACI 2008a):

2
Material Properties Ec
In any numerical investigation, it is of utmost importance to provide fc0 ¼ ð1Þ
4,733
the material properties as realistically as possible. Hence, during
validation, the maximum error is limited to less than 10%. The
properties of concrete, FRP, steel reinforcement, and epoxy used pffiffiffiffiffi
in the numerical modeling are presented in Table 7. FRP shows fr ¼ 0.62 f c0 ð2Þ
a linear behavior up to the ultimate load and then fails suddenly
due to brittle failure. Hence, linear properties for FRP bars are where fc0 , Ec and f r are in MPa; f c0 = ultimate compressive strength
incorporated in the numerical models. of concrete; Ec = compressive elastic modulus; and f r = ultimate
tensile strength of concrete in bending. Poisson’s ratio for concrete
is assumed to be 0.2.
Table 7. Material Properties of Concrete, CFRP, and Epoxy Used for Here, it is assumed that the uniaxial stress-strain relations can be
Numerical Modeling
converted into stress-equivalent plastic strain curves which are
Serial automatically obtained from the provided inelastic strain data.
number Material Parameter Properties (MPa) Uniaxial compressive stress-strain values for concrete are evaluated
1 Concrete Compressive strength 30 from Eqs. (3) and (4) (Desayi and Krishnan 1964; Gere and
Tensile strength 3.83 Timoshenko 1997).
Young’s modulus 25,980 Under uniaxial compression, the response is linear until the
2 CFRP bar Ultimate tensile strength 2,000 initial yield point, typically 30% of ultimate compressive stress.
Young’s modulus 150,000 In the plastic regime, response is typically characterized by strain
3 Epoxy Tensile strength 82 hardening followed by strain softening beyond the ultimate stress.
Young’s modulus 3,400 This representation, although somewhat simplified, captures the
4 Steel Tensile strength 425
main features of the response of concrete.

© ASCE 04013021-6 J. Compos. Constr.

J. Compos. Constr., 2014, 18(1): 04013021

 Table 8. Concrete Yield Stress and Corresponding Inelastic Strain Values where f c = the compressive strength of concrete and Gf = the frac-
Yield stress (MPa) Inelastic strain ture energy of concrete measured in N=m. In the present study, for
M30 concrete, fracture energy of 90.1 N=m is used.
7.02 0
The linear elastic behavior prior to initiation of damage is often
15.29 0.00034
21.90 0.00069
described in terms of a penalty stiffness that degrades under tensile
26.33 0.00104 and shear loading but is unaffected by pure compression. Delami-
29.18 0.00147 nation at concrete and epoxy interface is defined in terms of traction
30.00 0.001931 and separation. The current traction-separation model assumes
27.43 0.00213 three components of separation: one normal to the interface and
24.50 0.00257 two parallel to it; and the corresponding stress components are
21.91 0.003 assumed to be active at a material point. Uncoupled cohesive
behavior is considered by defining stiffness components in normal,
Downloaded from ascelibrary.org by Birla Institute of Technology - Pilani on 06/18/16. Copyright ASCE. For personal use only; all rights reserved.

shear, and tangential directions. The off-diagonal terms in the

 
ϵEc elasticity matrix are set to zero as the normal, shear, and tangential
f¼   ð3Þ components are uncoupled. For a cohesive layer, stiffness (K c ) of
1 þ ϵϵ0 2
the interface (relating the normal traction to the displacement) can
be computed using Eq. (6):

0

2f c Ec
ε0 ¼ ð4Þ Kc ¼ ð6Þ
Ec Tc
where f = stress (in MPa) at any strain any strain level, ε; ε = strain where T c = initial thickness of the interface and Ec = adhesive
at stress f; and ε0 = strain at the ultimate compressive strength, fc0 . material stiffness.
The final yield stress versus inelastic strain values used for
modeling the concrete are presented in Table 8. The yield stress
and corresponding strain values are obtained from analytical Geometric Modeling
equations proposed in the Abaqus analysis user’s manual (Dassault In this section, modeling details of the specimens (control and
Systèmes Simulia Corporation 2010). strengthened specimens) using Abaqus (Dassault Systèmes Simulia
Under uniaxial tension, the stress-strain response follows a lin- Corporation 2010) are discussed. Geometries of the models corre-
ear elastic relationship until failure stress is achieved. Beyond this sponding to the underreinforced and overreinforced specimens are
stress, the formation of microcracks is represented macroscopically dealt with separately. Here, both steel and FRP reinforcements are
with a softening stress-strain response, which induces strain modeled as discrete truss elements inside concrete. The stirrups
localization in the concrete structure. The Modulus of rupture is are modeled using 2D wire elements. The reinforced concrete
computed to be 3.8 MPa. beams with grooves are modeled using 3D solid elements. Global
mesh size of 25 mm is taken.
Other Parameters The major embedded elements are the internal stirrups and two
The flow potential, yield surface, and viscosity parameters for the 8-mm bars at the top and two 12-mm bars at the bottom. The bond
concrete-damaged plasticity model are defined through concrete between FRP and epoxy is assumed to be perfect, while the bond
damage plasticity input. Herein, the parameters in Table 9 are used between epoxy and concrete is modeled using cohesive elements
for all numerical models. The flow potential eccentricity value considering response in terms of traction and separation.
ε ¼ 0.1, meaning that the material has almost the same dilatation
angle over a wide range of configuring pressure stress values. The
ratio f bo =fco is set to the default value of 1.16, while the value of Validation of Numerical Models
K c is taken as 1.0, so that the yield surface has a perfect cone shape
in three-dimensional spaces. The viscosity parameter, μ, is used The Abaqus (Dassault Systèmes Simulia Corporation 2010) flexure
for the viscoplastic regularization of the concrete constitutive control beam model (NF-CBOR1) is validated with the correspond-
equations. The default value of 0.0 is taken, assuming a rate- ing experimental result (E-CBOR1). The strength of FRP rebar ob-
independent analysis. tained from experiments is used for validating the strengthened

Modeling Concrete-Epoxy Interface

Cohesive behavior is enforced only for portions of surfaces that are
in contact at the start of the analysis. The concrete-epoxy interface
behavior is initially linear elastic followed by damage initiation and
evolution based on energy dissipated due to failure. The fracture
energy of concrete is calculated using the Remmel (1994) method
as given in Eq. (5)


f
Gf ¼ 65 ln 1 þ c ð5Þ
10

Table 9. Plasticity Parameters Used for Concrete Damage Plasticity Model

Dilation angle Eccentricity, ε f bo =f co KC Viscosity parameter, μ Fig. 8. Numerical validation of overreinforced control beam for flexure
36° 0.1 1.16 1 0 with experimental results

© ASCE 04013021-7 J. Compos. Constr.

J. Compos. Constr., 2014, 18(1): 04013021

Downloaded from ascelibrary.org by Birla Institute of Technology - Pilani on 06/18/16. Copyright ASCE. For personal use only; all rights reserved.

Fig. 9. Numerical validation for shear control beam Fig. 10. Numerical validation of overreinforced strengthened flexure
beam

Table 10. Load-Carrying Capacity of NSM CFRP-Strengthened RC

Beams Modeled Using Abaqus
Basic description
Beam (bar diameter) Peak load (kN)a
NF-CBOR1 8 56(50)b
F-BD-6 6 80
F-BD-8 8 87
F-BD-10 10 93
F-BD-12 12 97
F-BS-1500 8 78
F-BS-2500 8 85
F-BS-2800 8 87
F-GS1.5d 8 80
F-GS 2.5d 8 81
F-GW 1.3d 8 103 Fig. 11. Numerical validation for shear-strengthened beam
F-GW 1.5d 8 102
F-GW 1.7d 8 102
F-OR-8 8 74 The numerical model NS-CBOR1 is observed to tend toward the
F-GW 2d 8 85 experimental result (E-CBS1) with a deviation of 9% as shown in
F-GD-E 1d 8 75 Fig. 9 and Table 10. This substantiates the evidence indicating the
F-GD-E 2d 8 96 effectiveness of the material properties of the concrete used in
F-GD-E 3d 8 90 the study.
NS-CBOR1 8 63 Further, the Abaqus (Dassault Systèmes Simulia Corporation
N-BCS-10 × 3 10 × 3 77 2010) flexure-strengthened beam (F-OR-8) is validated (keeping
N-BCS-10 × 4 10 × 4 81 nature the same) with the corresponding experimental result
N-BCS-7 × 7 7×7 88
(E-ORF2). Numerically, the Abaqus overreinforced model
N-BCS-10 × 5 10 × 5 90
SV-GW 1.3d 8 100
(overreinforced beam) for flexure is in close proximity to the
SV-GW 1.5d 8 100 experimental result with a deviation of 4% (Fig. 10). Similar val-
SV-GW 2d 8 100 idation is conducted for an Abaqus shear-strengthened beam
SV-GW 2.5d 8 93 (N-SM-GO 52.4) with the corresponding experimental beam
SV-GS 1s 8 99 (E-SM9). Both the load versus deflection curve and maximum
SV-GS 1.5s 8 81 load-carrying capacity of the shear-strengthened beam are match-
SV-GS 2s 8 80 ing with the experimental result with a deviation of 2% (Fig. 11).
N-SI 30 8 84 Hence, the properties of epoxy and FRP model (giving interfa-
N-SI 45 8 102 cial behavior and crack modeling) are very well validated using
N-SV 8 80
both the strengthened beams (flexure and shear). For the parametric
N-SI 52.4 8 92
N-SM-GO 52.4 8 106
study, properties of commercially available FRP bars have
been used as mentioned in Table 7. Load-carrying capacity and
a
Ten kN ¼ One ton. strength enhancement provided by strengthened beams modeling
b
Experimental results for corresponding beam as presented in Table 5. with Abaqus (Dassault Systèmes Simulia Corporation 2010) are
presented in Table 10.
beams. The nature of the load versus deflection curve is observed to
be similar. Numerically, the validated model (control beam) for Results and Discussion
flexure is in close proximity to the experimental result with a
deviation of 5% (Fig. 8). Therefore, this model is considered as
the validated one, and the same material properties are used for Response of Strengthened-Flexure Beams
parametric study. Furthermore, the nature of the curve is observed As shown in Table 10 and Fig. 12, the tensile strength of the NSM
to be similar to the experimental result. Similarly, numerically mod- CFRP bars affects the performance of beams. The beam F-BS-2800
eled shear control beam is also validated with experimental results. strengthened with CFRP bars of strength of 2,800 MPa exhibits

© ASCE 04013021-8 J. Compos. Constr.

J. Compos. Constr., 2014, 18(1): 04013021

Downloaded from ascelibrary.org by Birla Institute of Technology - Pilani on 06/18/16. Copyright ASCE. For personal use only; all rights reserved.

Fig. 14. Effect of bar diameter on response of beams strengthened in

flexure
Fig. 12. Effect of strength of NSM CFRP bar on response of beams
strengthened in flexure

Table 11. Load-Carrying Capacity of Strengthened RC Beams Modeled

Using Abaqus
Percent strength enhancement
Model with respect to control beam
F-BD-12 73
F-BS-2800 55
F-GW 1.3d 84
F-GD-E 2d 71
SV-GW 1.3d 59
SV-GS 1s 56
N-SM-GO 52.4 68 Fig. 15. Performance of the beams strengthened with bars of different
cross sections

appropriate distance between the two. The model F-GD-E-2d

exhibits 29% more strength than F-GD-E-1d. Also a large slip
is observed between bar and concrete in the case of F-GD-E 1d.
Therefore, the specimens with the very least distance between
groove and edge (Fig. 13 and Table 10) would be unable to achieve
the ultimate strength, as peeling-off or slip can occur. It is observed
that F-GD-E-2d achieved 70% more ultimate strength than the con-
trol beam. Furthermore, it is observed that the bar diameter also
plays a significant role for the strengthening purpose (Fig. 14).
The bars with 6-mm diameter (F-BD-6) have displayed 18% lesser
enhancement than that strengthened with 12-mm CFRP bars
(F-BD-12). From Fig. 15, it is clear that the beam strengthened with
two NSM CFRP rectangular bars with dimensions 10 × 5 mm
Fig. 13. Effect of distance of grooves from edge in beams strengthened
(N-BCS-10 × 5) depicts 19% more strength than the beam
in flexure
strengthened with 10 × 3 mm (N-BCS-10 × 3) bars. Models
N-BCS-10 × 5 and N-BCS-7 × 7, which have rectangular cross
section, depicted slightly more strength than the beam strengthened
55% enhancement in the beam’s strength (Table 11). Also, for
with circular bars (F-BD-8) which have similar area, but the
groove-width patterns under the F-GW scheme, it is observed that
difference obtained is negligible.
varying the groove width does not result in any significant variation
in the strength of the beam, but a minimum groove width is neces-
sary for epoxy to hold the bar effectively in concrete. Specimen Shear Response of Strengthened Beams
F-GW 1.3d shows an enhancement of 84% over the control beam Under the scheme SV-GW, it is observed that the ultimate strength
(Table 11). Further, when the effect of spacing between grooves is of specimens with groove widths lesser than 2 d (2 times the diam-
considered under scheme F-GS where three bars are used for eter of the bar) remained almost in the same range. However, when
strengthening, it is observed that for groove distance of 1.5 and the groove width is 2.5 d, strength of the beam declines. This can be
2.5 diameter, the performance remained more or less the same. explained by the fact that with the increase in groove width there is
Hence, spacing does not affect the performance significantly, but a a decrease in the effective area of concrete along the cross section,
minimum distance of 1.5 diameter is suggested. which leads to declined performance compared to SV GW 1.3d, as
In scheme F-GD-E, which represents the effect of the distance can be seen from Fig. 16. The specimen SV-GW-1.3d shows 59%
between groove and edge, it is concluded that there should be an higher shear strength than that of the control beam.

© ASCE 04013021-9 J. Compos. Constr.

J. Compos. Constr., 2014, 18(1): 04013021

 is the most efficient one after the beam strengthened in mixed
pattern (N-SM-GO 52.4), then comes N-SV.

Conclusion
The study presents an experimental investigation into the perfor-
mance enhancement of flexure- and shear-strengthened beams over
the control beams. Furthermore, a detailed parametric study is also
conducted and presented so as to examine the effect of various
parameters, through numerical modeling using Abaqus (Dassault
Systèmes Simulia Corporation 2010). These numerical models
Downloaded from ascelibrary.org by Birla Institute of Technology - Pilani on 06/18/16. Copyright ASCE. For personal use only; all rights reserved.

Fig. 16. Effect of groove width on response of beams strengthened in are first validated with experimental results. The parametric study
flexure provides guidelines on the strengthening patterns for effective
placement and usage of FRP bars. These guidelines are in good
agreement with the ACI-440 ACI (2008b) guidelines for near-
surface mounting of RC beams strengthened with FRP bars. In case
of flexure and shear strengthening, four groove-specific parame-
ters: Clear spacing between grooves (GS), groove width (GW),
grooves inclination (SI), and groove distance from edge (GD-E)
apart from the other common parameters: FRP bar diameter and
strength of FRP bars are considered. The effects of most of the
parameters which determine the effectiveness of the NSM tech-
nique have been discussed. Based on the above investigations,
the following concluding remarks are made regarding the efficient
application of the NSM technique:
1. The FRP bar diameter and strength plays an important role.
2. Orientation of FRP bars used for shear strengthening plays a
significant role; hence they should be used effectively. The
Fig. 17. Effect of groove spacing on response of beams strengthened in NSM bars installed at 45° to the axis of the beam are found
shear to be most efficient for shear-strength enhancements, while the
bars placed vertically are least effective.
3. It is necessary to maintain a minimum groove distance of at
least 1.5 times the diameter of CFRP bar from the edge.
4. A minimum distance of 1.5 times the diameter of NSM FRP
bars should be ensured between the centers of the grooves.
5. Increase in the groove width beyond 2 times the NSM FRP bar
diameter is detrimental to shear strength.
6. The closer the bars are for NSM FRP shear strengthening, the
greater is the performance of the beam in shear. However,
the spacing between NSM FRP bars should not be less than
the distance between the steel stirrups in the beams.
7. Varying the shape of NSM FRP bars being used does not
significantly influence the enhancement in the strength.
Although rectangular bars depicted slightly higher enhance-
ment over the beams strengthened with circular bars with
similar area, the difference obtained is unsubstantial.
Fig. 18. Effect of orientation of grooves on response of beams
strengthening in shear
Acknowledgments

When the spacing between the grooves is considered, it is This paper is a part of a research project funded by University Grant
observed that the beam’s strength is inversely proportional to Commission (UGC) of India, New Delhi, sanctioned to the first
the spacing between the grooves. When the NSM CFRP bars author under the scheme of Major Research Projects [F. No. 34-
are nearer, they add to the shear strength (Fig. 17). The ultimate 389/2008 (SR)] and Aditya Birla group (ABG project).
strength of the beam SV-GS 1s strengthened for shear under this
scheme is observed to be 99 kN which is about 56% higher than
that for the control beam. References
Regarding the orientation of CFRP bars for shear strengthening,
American Concrete Institute (ACI). (2008a). “Building code requirements
it is clear from Fig. 18 that specimen N-SI 45 gives maximum
for structural concrete and commentary.” ACI 318, Farmington Hills,
strength because in this case the bars are perpendicular to the MI.
cracks, while it is the least in the case of vertical bars, where American Concrete Institute (ACI). (2008b). “Guide for the design and
the bars make approximately 45° to the cracks. Hence, the construction of structural concrete reinforced with FRP bars.” ACI
shear-strengthening pattern with orientation of the bar being 45° 440.2R, Farmington Hills, MI.

© ASCE 04013021-10 J. Compos. Constr.

J. Compos. Constr., 2014, 18(1): 04013021

 ASTM. (2008) “Standard test method for tensile properties of Ferracuti, B., Savoia, M., and Mazzotti, C. (2006). “A numerical model for
pultruded glass-fibre-reinforced plastic rod.” ASTM D3916-08, FRP-concrete delamination.” Compos. Part B, 37(4–5), 356–364.
West Conshohocken, PA. Gere, J., and Timoshenko, S. (1997). Mechanics of materials, PWS
Bureau of Indian Standards. (2000). “Plain and reinforced concrete—Code Publishing, Boston.
of practice.” IS 456-2000, New Delhi, India. Rasheed, H. A., Harrison, R. R., Peterman, R. J., and Alkhrdaji, T.
Ceroni, F., Pecce, M., Bilotta, A., and Nigro, E. (2012). “Bond behavior (2010). “Ductile strengthening using externally bonded and near
of FRP NSM systems in concrete elements.” Compos. Part B, 43(2), surface mounted composite systems.” Compos. Struct, 92(10),
99–109. 2379–2390.
Dassault Systèmes Simulia Corporation. (2010). Abaqus analysis user’s Rasheed, H. A., Nayal, R., and Melhem, H. (2004). “Response prediction
manual 6.10, ABAQUS Inc., Providence, RI. of concrete beams reinforced with FRP bars.” Compos. Struct., 65(2),
De Lorenzis, L., Rizzo, A., and Tegola, A. L. (2002). “A modified pull-out
193–204.
test for bond of near-surface mounted FRP rods in concrete.” Compos.
Remmel, G. (1994). Tensile and shear behavior of high strength concrete
Part B, 33(8), 589–603.
Downloaded from ascelibrary.org by Birla Institute of Technology - Pilani on 06/18/16. Copyright ASCE. For personal use only; all rights reserved.

members (in German), Deutscher Ausschuss für Stahlbeton, Beuth

De Lorenzis, L., and Teng, J. G. (2007). “Near-surface mounted FRP
reinforcement: An emerging technique for strengthening structures.” Verlag, Berlin.
Compos. Part B, 38(2), 119–143. Singh, S. B., Madappa, V., Reddy, A., Khatri, C. K., Nagarjuna, A., and
Desayi, P., and Krishnan, S. (1964). “Equation for the stress-strain curve of Hima, K. (2012). “Performance of NSM-FRP RC beams in flexure
concrete.” J. Am. Concr. Inst., 61(3), 345–350. and shear using locally developed CFRP rebars.” Int. J. Sustainable
El-Attar, A. G., White, R. N., and Gergely, P. (1997). “Behaviour of gravity Mater. Struct. Syst., 1(1), 42–67.
load designed reinforced concrete buildings subjected to earthquakes.” Weidong, W., Juan, C., and Zhao, L. (2010). “Analysis of the bond
ACI Struct. J., 94(2), 133–145. performance of strengthened with NSM CFRP rods in concrete
El-Hacha, R., and Rizkalla, S. H. (2004). “Near-surface-mounted fiber- based on Ansys.” Int. Symp. on Multi-field Coupling Theory of Rock
reinforced polymer reinforcements for flexural strengthening of and Soil Media and its Applications, Oriental Academic Forum,
concrete structures.” ACI Struct. J., 101(5), 717–726. Australia.

© ASCE 04013021-11 J. Compos. Constr.

View publication stats J. Compos. Constr., 2014, 18(1): 04013021