International Journal of Civil Engineering

https://doi.org/10.1007/s40999-018-0342-8

RESEARCH PAPER

Flexural Performance and Failure Modes of NSM CFRP-Strengthened

Concrete Beams: A Parametric Study
Rami H. Haddad1   · Oubaida A. Almomani1

Received: 11 January 2018 / Revised: 22 June 2018 / Accepted: 23 June 2018

© Iran University of Science and Technology 2018

Abstract
The flexural performance and failure modes of concrete beams, strengthened with near-surface mounted (NSM) carbon fiber-
reinforced polymer (CFRP) strips were investigated. For this, reinforced concrete beams (150 × 250 × 1400 mm) were cast,
and cured for 28 days, then strengthened with NSM CFRP strips at varying numbers (1–3), lateral spacing (50–100 mm),
and embedment lengths (150–450 mm). The impact of staggering of NSM CFRP strips upon strengthening efficiency was
tackled, as well. The flexural mechanical response, strain in strips at failure, and cracking and failure modes were evalu-
ated for all concrete beams under a four-point loading test setup. The findings indicated that inserting NSM CFRP strips at
a certain lateral distance from the main steel bars prevented end-cover peeling-off. Furthermore, staggering NSM CFRP
strips contributed to increasing the residual flexural capacity and toughness of strengthened beams by as much as 30 and
51%, respectively, although it led to a slight reduction in their ductility. In general, the present study confirmed the findings
of different literature works with regard to the effect of key repair parameters using NSM CFRP strips.

Keywords  Strengthening · NSM CFRP strips · Peeling off · Analytical prediction · Strain

1 Introduction their susceptibility to cracking under tension and in com-

pression zones, which impacts their load resistance mecha-
In field, reinforced concrete (RC) structural members are nism tangibly. Until recently, steel plates and elements or
designed to resist various types of loads unless their concrete reinforced concrete jacketing were mainly used in repair/
deteriorates due to physical, chemical, and/or electrochemi- strengthening of beams; both of which have its own disad-
cal attacks, or due to exposure to excessive loadings or an vantages. Steel plates or elements are hardly shaped to fit
accidental fire [1]. The extent of degradation in structural complex profiles of concrete sections as they are relatively
performance depends mainly upon the acting dead and live heavy, inflexible, and difficult to manipulate and bond to
loads, severity of the causative factor(s) and type of element, concrete under site conditions. Additionally, steel-strength-
foundation, columns, beam, or slab. Hence, unless proper ened structures would require periodic maintenance to
strengthening or repair measures are undertaken, structural prevent or delay corrosion, especially in aggressive envi-
collapse could occur without warning. Fortunately, statis- ronments [2, 3]. Reinforced concrete jacketing (RCJ), on
tics have shown that most deficient structural elements are the other hand, introduces apparent dimension changes and
repairable, especially with the progress made in repair using adds considerable deadweight to the repaired element [4].
advanced composite materials. As they possess better physical, chemical and mechanical
The available database shows that reinforced concrete properties than steel plates and RCJ, fiber-reinforced poly-
beams are the most repaired structural elements owing to mer (FRP) plates and sheets have been used to strengthen/
repair various concrete structural elements. The outcome of
* Rami H. Haddad approximately 20 years of research indicates many advan-
[email protected] tages of FRP materials over conventional repair techniques,
Oubaida A. Almomani including higher repair efficiency and easier application
[email protected] within a relatively short period of time [5, 6]. However, the
tendency of FRP composites, in the form of conventional
1
Civil Engineering Department, Jordan University of Science plates and sheets, to detach from concrete at relatively low
and Technology, P.O. Box 3030, Irbid 22110, Jordan

13
Vol.:(0123456789)
International Journal of Civil Engineering

loads undermined the benefit of its use as a repair/strength- significant improvements to the performance of repaired
ening technique. structural elements. However, the main disadvantage of
To overcome the shortcomings associated with the use using NSM CFRP strips for repairing flexural members is
of conventional FRP plates in repair, a new form of FRP represented in their tendency to peel off at their termination
composite is currently emerging: near-surface mounted zones, followed by accelerated crack propagation towards
(NSM) carbon FRP (CFRP) strips. These are inserted in the compression region. Hence, there is a need to reach a
the man-created groove that is filled with a strong adhe- better understanding as for the effect of strips’ configura-
sive. Compared to externally bonded FRP plates or sheets, tion and their staggering upon repair significance. Further-
NSM CFRP strips are protected against accidental mechani- more, the present findings would provide additional data
cal damage, sun, and moisture, and hence, it would be very for future development and validation of reliable empirical
attractive to promote negative moment capacity in areas models capable of predicting ultimate strain in NSM CFRP
exposed to mechanical damage or environmental attacks. strips at failure.
Furthermore, NSM CFRP strips are less prone to de-bonding
from concrete substrate, and their application introduces no
esthetic changes to the repaired structural elements [7–12].
Consequently, various studies were recently undertaken to 3 Methodology of Study
investigate the mechanical performance and failure modes
of beams repaired for shear and flexure using NSM CFRP To fulfill the objectives of the proposed study, 19 reinforced
strips. The major findings indicated: (a) higher contribu- concrete beams (150 × 250 × 1400  mm) were designed
tion of NSM CFRP strips to improving structural perfor- according to ACI committee 318 to show ductile flexural
mance as compared to externally attached strips of similar failure [13]. For this, all beams were reinforced at similar
sectional area and strength properties; (b) groove’s thickness longitudinal steel reinforcement ratio of 1% with the shear
and spacing having a significant impact on the efficiency reinforcement determined based upon a constant shear/flex-
of repair, and (c) premature concrete cover delamination ural theoretical ultimate load capacity ratio of about 1.5.
being a major disadvantage. Most studies reported soffit Beams were cast in wooden molds, and cured for 28 days
cover delamination as the dominant failure mode in shear- under moist conditions before being divided into three main
strengthened beams, especially when relatively close strip sets: the first set consisted of two control beams, whereas
spacing is used [7–12]. the second consisted of 16 beams, strengthened with NSM
The present paper aims to help the reader gain a better CFRP strips at different embedment lengths and strips’ num-
understanding of the effect of the key parameters that sig- ber, spacing, and positioning relative to main steel; embed-
nificantly influence the repair efficiency when NSM CFRP ment length being defined as extension distance of strips
strips are used. Two of the parameters were rarely studied beyond maximum moment zone. The third set pertained to a
(if none); specifically, lateral position of strips relative to single beam, strengthened using three staggered NSM CFRP
that of reinforcing steel bars, and longitudinal staggering of strips, two of which were terminated at a shorter embedment
NSM CFRP strips. The mechanical response of strengthened length relative to the third one.
beams was evaluated in terms of load–deflection response The present beams were strengthened with different
and its characteristics as well as strain induced in NSM numbers, spacing values, and embedment lengths for NSM
CFRP strips at failure. Furthermore, the present findings CFRP strips to experimentally determine critical repair
with regard to the impact of key parameters upon flexural configuration(s) which would contribute to delaying or pre-
repair efficiency when using NSM CFRP strips in terms of venting undesirable flexural failure by concrete cover peel-
strain induced (in strips) were validated against three of the ing-off. The entire testing program is summarized in Table 1
most recent relevant research works. and illustrated schematically in Fig. 1. The different beams
were given letter–number designations, reflecting number of
NSM CFRP strips, and their extension and spacing.
2 Problem Statement and Objectives Finally, all beams were tested under the effect of four-
point loading with load, mid-span deflection, and strain
Premature detachment of conventional carbon FRP (CFRP) in NSM CFRP at failure acquired using a data acquisition
plates in strengthened beams before the compression con- system. The data were processed and analyzed to establish
crete develops its ultimate strain capacity had limited the the impact of various parameters on flexural performance,
benefit of using such materials in repair works. As an alter- ultimate strain in strips, and mode of failure. The impact of
native, NSM CFRP strips were manufactured and com- various key parameters upon ultimate strain in NSM CFRP
mercialized because their positioning in grooved concrete strips was substantiated through comparisons with the most
significantly augments their bond with concrete, imparting recent literature works.

13
International Journal of Civil Engineering

Table 1  Designation of test Set # NS SCONF SN EL (mm) SP (mm) Beam designation

specimens and parameters
1 2 None – – – C
2 2 NSM 1 250 – C_SN1 _EL(25)
2 NSM 2 150 7.5 C_SN2 _EL(15)_SP7.5
2 NSM 2 250 7.5 C_SN2 _EL(25)_SP7.5
2 NSM 2 450 7.5 C_SN2 _EL(45)_SP7.5
2 NSM 2 450 5 C_SN2 _EL(45)_SP5
2 NSM 2 450 10 C_SN2 _EL(45)_SP10
2 NSM 3 150 5 C_SN3 _EL(15)_SP5
2 NSM 3 250 5 C_SN3 _EL(25)_SP5
1 NSM 3 2 (150)–450 5 CSN3 _EL(2 × 15 − 1 × 45)_SP5

SCONF strengthening configuration, SN strips number, NS number of samples, EL embedment length

beyond maximum moment zone, SP spacing between strips, C control (intact), NSM near-surface mounted

4 Experimental Program of steel plates of 10 mm thickness were attached to the high
shear zones of the beams to ensure flexural failure, as shown
Material properties, mix design, concrete mixing and cast- in Fig. 2b. The steel plates covered the shear zones to a dis-
ing, detailing of test beams and reinforcement, procedure tance of 30 mm from the bottom surface of the beams where
of repair applications and testing setup and procedure have the NSM CFRP strips were inserted in concrete grooves.
been discussed in detail in this section. With such a configuration, the attached steel plates would
have no direct effect upon possible end-concrete cover
4.1 Concrete Ingredients and Proportions separation.

Type I ordinary Portland cement, tap water, coarse lime-

4.3 CFRP Composites
stone aggregate with 19 mm maximum aggregate size, and
a mixture of fine limestone aggregate and silica sand at 70
NSM CFRP strips and their prescribed adhesive from SIKA
and 30%, respectively, were used in casting the concrete
were used for strengthening different specimens. Their prop-
mixture used in this work. A commercial superplasticizer
erties, as provided by the manufacturer (SIKA), are shown
was added to the mixture to achieve the target workabil-
in Tables 4 and 5, respectively.
ity. The coarse aggregate gradation met the requirements
of the ASTM method C33 [14]. The physical properties of
the different aggregate particles were measured according to 4.4 Concrete Mixing, Casting, and Curing
ASTM specifications with the results listed in Table 2 [14].
The proportions of concrete ingredients were determined A tilting drum mixer of 0.15 m3 was used for mixing the
using the ACI-211.1 mix-design method to achieve a cyl- concrete ingredients. At first, a little amount of water was
inder compressive strength of 35 MPa at 28 days of moist added in the mixer to wet the surface, then coarse aggregates
curing and a slump of about 100 mm [15]. The mixture’s were added and wetted with some water while the mixer
proportions of cement, coarse aggregate, fine aggregate, was rolling. Afterwards, cement, fine aggregate, silica and
silica sand, water, and superplasticizer were 430, 870, 506, water were added gradually and alternatively, before add-
217, 236, and 3.23 kg/m3, respectively. ing a superplasticizer to the mixture at 0.75% by cement
weight to achieve the required workability. Finally, the
4.2 Test Specimen ingredients were blended for about 1 min in the mixer to
obtain a cohesive mixture. Concrete was placed in the molds
Nineteen beams (150 × 250 × 1400 mm) were reinforced in three layers, and each was compacted using a vibrator,
with 2φ12 mm steel bars in compression side (top), and before the third layer was finished smooth by a trowel. The
2φ14 mm in tension side (bottom), with steel stirrups of φ8 specimens were demolded after 24 h, and then covered with
mm spaced @ 75 mm (center-to-center); the steel detail- wet burlap for another 27 days. Standard cylinder speci-
ing is shown in the schematics of Fig. 2a. The mechanical mens (100 × 200 mm) were also cast, cured in wet burlap
characteristics for the steel bars were determined at room and tested for compressive strength after 28 days according
temperature from their stress–strain diagrams and are listed to ASTM C 39 [14].
in Table 3. Additional shear reinforcement using two pairs

13
International Journal of Civil Engineering

Fig. 1  Strengthening configurations using NSM CFRP strips as illustrated by bottom and cross-sectional views of the concrete beams

4.5 Bonding NSM CFRP Strips

Table 2  Physical properties of aggregate used in the present work
Properties Aggregate In order to install the NSM CFRP, grooves at a depth of
CL FL SS 70% FL + 30 SS 22.5 mm and width of 8 mm were created within the beams’
concrete cover at different spacing values using an electri-
BSG (D) 2.41 2.60 2.59 2.62 cal drill, as shown in Fig. 3. The dimensions for the grooves
FM NA 3.25 1.6 2.8 were chosen upon the recommendation of Seo et al. [10]
Absorption (%) 2.1 1.7 0.5 NA and taking into account the geometric dimensions of the
UW (kg/m3) 1401 1612 NA NA NSM CFRP strips used. The grooves were cleaned by air
CL coarse limestone, FL fine limestone, SS silica sand, BSG (D) bulk vacuuming and direct brushing, before being wetted with
specific gravity, FM fineness modulus, UW unit weight a volatile compound to reduce surface humidity. Later, the

13
International Journal of Civil Engineering

Fig. 2  Schematics of present
control and strengthened beams,
designed to failure in flexure

bonding epoxy was squeezed into the grooves until filled the epoxy, squeezed out, and used to cover the strips by a
completely to expel air bubbles before the NSM CFRP special spatula.
strips were pressed into the full depth of the grooves with

13
International Journal of Civil Engineering

Table 3  Physical and mechanical properties of the reinforcing steel at 4.6 Test Setup

ambient temperature
Bar diameter YS (MPa) US (MPa) EAF (%) All beams were simply supported while being subjected to
(mm) two-point load around the center, as shown in Fig. 2. One
8 390 570 13
linear variable displacement transducer (LVDT) was posi-
12 441 641 8
tioned at the mid-span to measure the vertical deflection,
14 608 732 7.5
whereas another LVDT was used to measure the strain in
CFRP strips. The latter was mounted on two small exten-
YS yield stress, US ultimate stress, EAF elongation at failure sions of NSM CFRP strips, attached to the main NSM CFRP
strips at two points spaced at 160 mm around the center of
the beams (Fig. 2b illustrates LVDTs distribution). The load
Table 4  Physical and mechanical properties of SIKA NSM CFRP was applied using a hydraulic testing machine of 250-kN
strips capacity under a loading rate of 0.02 kN/s. Measurements of
Fiber type NSM CFRP load versus vertical deflection, and NSM CFRP elongation
were acquired electronically by a data acquisition system.
Fiber orientation 0° (unidirectional)
Fiber density 1.6 g/cm3
Strip width 15 mm
5 Mechanical Response of Beams
Strip thickness 2.5 mm
Strengthened with NSM CFRP Strips
Cross-sectional area 37.5 ­mm2
Mean tensile strength 3100 N/mm2
The mechanical behavior of NSM CFRP-strengthened beams
Tensile E-modulus 165,000 N/mm2 (nominal)
was investigated considering the effect of key parameters,
Strain at break > 1.7% (nominal)
namely NSM CFRP strips’ embedment length, number,
spacing, positioning relative to main steel and staggering.

Table 5  Properties of SIKA Color Cement grey

epoxy used for bonding CFRP
strips Mixed density at 25 °C 1.8 g/cm3 (approx)
Sag flow Non-sag on vertical surface
Flashpoint NA
Shrinkage 0.04%
Compressive strength-ASTM C579 85 N/mm2 at 3 days
Flexural strength according to ASTM 25 N/mm2 at 7 days
Tensile strength according to ISO 527 18 N/mm2 at 7 days
Shear strength according to ASTM 21 N/mm2 at 7 days
Bond to concrete > 4 N/mm2 at 1 day (concrete fracture)
E-modulus-ASTM and ISO 527 10,000 N/mm2 (compression and tension)

Fig. 3  Creating single and multiple grooves on the tension side of the beam: a sawing procedure; b a schematic showing the dimensions of the
groove

13
International Journal of Civil Engineering

5.1 Effect of Embedment Length NSM CFRP at embedment lengths of 150, 250, and 450 mm
were (120, 107 and 136%), (95, 109 and 113%) and (85, 113
Figure  4 shows the load–deflection curves for different and 111%), respectively. As can be deduced, the beams with
beams, strengthened by two NSM CFRP strips at embed- the longest embedment length achieved the highest flexural
ment lengths of 150, 250 and 450  mm, respectively, as load capacity and stiffness, yet attained reduced rotational
compared to that of control (C) ones. The curves showed ductility. The results from direct measurements of strains
linear behavior at the initial stages of loading before became in NSM CFRP strips at failure (reported in Table 6) are in
nonlinear behavior thereafter. The beams, strengthened with agreement with earlier conclusions related to the positive
NSM CFRP at embedment lengths of 150, 250 and 450 mm, impact of longer embedment length upon repair efficiency.
achieved a load capacity of 115, 133 and 154% of that for The positive impact of extending further NSM strips towards
control beams, respectively, reflecting the benefit of extend- the beams’ supports upon flexural load capacity and stiff-
ing further NSM CFRP strips beyond the high moment zone. ness of strengthened beams may be referred to two major
The curves’ characteristics are shown in Table 6, with their factors: (a) shifting of termination points of the strips in
residual properties computed as the percentage of mechani- concrete beams towards zones where fewer load-generated
cal properties for strengthened beams to that of control cracks exist, and (b) reducing average bond stress between
beams. the NSM CFRP strips and concrete, and consequently, the
Results of Table 6 show that the residuals for rotational average horizontal shearing stresses in the vicinity of the
ductility, stiffness, and toughness, pertaining to beams with strips’ ends. Accordingly, the initiation and progression of
concrete cover separation required relatively higher trans-
verse loads. The reduced ductility and toughness for the
300 repair cases with higher embedment lengths of NSM CFRP
strips reflected the increase in overall flexural stiffness of
250
strengthened beams.
200
5.2 Effect of Number of CFRP Strips
Load (KN)

150
C_SN2_EL(45)_SP_7.5
100 Figure  5 shows load-deflection curves for the beams
C_SN2_EL(25)_SP_7.5
50 C_SN2_EL(15)_SP_7.5 strengthened with different numbers of NSM CFRP strips at
C a constant embedment length of 250 mm as compared to that
0 of control (C) ones. The curves showed linear behavior at the
0 5 10 15 20 25 30
Deflec on (mm)
initial stages of loading before turned nonlinear afterwards.
Beams with one, two, and three NSM CFRP strips achieved
load capacity values of 137, 133 and 146%, respectively, of
Fig. 4  Load versus deflection for beams strengthened using two NSM
CFRP strips at a spacing value of 75 mm and embedment lengths of that for control beams, reflecting the benefit of increasing the
150, 250, and 450 mm total cross-sectional area of NSM CFRP strips. Furthermore,

Table 6  Mechanical characteristics of control and strengthened beams

Beam designation ULC (kN) AULC (kN) RD S (MN/m) T (J) 𝜀f (µ) 𝜀f

𝜀fu

C 165 (100%)a NA 2 (100%) 21.9 (100%) 1891 (100%) NA NA

C_SN1 _EL(25) 226 (137%) 210 1.75 (88%) 23.2 (105%) 2003 (106%) 7600 42%
C_SN2 _EL(15)_SP7.5 190 (115%) 214 2.4 (120%) 23.5 (107%) 2567 (136%) 4300 24%
C_SN2 _EL(25)_SP7.5 219 (133%) 241 1.9 (95%) 23.8 (109%) 2140 (113%) 6700 38%
C_SN2 _EL(45)_SP7.5 254 (154%) 282 1.7 (85%) 24.8 (113%) 2089 (111%) 11,400 64%
C_SN2 _EL(45)_SP5 265 (161%) 281 1.9 (95%) 24.2 (111%) 2674 (141%) 11,000 61%
C_SN2 _EL(45)_SP10 270 (164%) 280 1.8 (90%) 24.6 (112%) 2137 (113%) 12,000 67%
C_SN3 _EL(15)_SP5 188 (114%) 214 2.5 (125%) 23 (105%) 2760 (146%) 3400 20%
C_SN3 _EL(25)_SP5 240 (146%) 254 1.5 (75%) 26.7 (112%) 1550 (82%) 5200 30%
CSN3 _EL(2 × 15 − 1 × 45)_SP5 236 (143%) 266 1.9 (95%) 24.2 (111%) 1793 (95%) 6000 34%

ULC ultimate load capacity, RD rotational ductility, S stiffness, T toughness, 𝜀f strain in NSM CFRP, 𝜀fu ultimate strain of NSM CFRP, AULC
analytical ultimate load capacity, NA not applicable
a
 Residual properties

13
International Journal of Civil Engineering

300 300

250 250
200 200
Load (kN)

Load (kN)
150 150
C_SN3_EL(25)
100 C_SN2_EL(45)_SP_7.5
C_SN2_EL(25) 100
50 C_SN1_EL(25) C_SN2_EL(45)_SP_10
C 50 C_SN2_EL(45)__SP_5
0 C
0 5 10 15 20 0
Deflec on (mm) 0 5 10 15 20
Deflec on (mm)
Fig. 5  Load versus deflection for beams strengthened using different
number of NSM CFRP strips at an embedment length of 250 mm Fig. 6  Load versus deflection for beams strengthened using two
NSM CFRP strips at spacing values of (50, 75, and 100 mm) and an
embedment length of 450 mm

the results, as listed in Table 6, showed that the residuals for

(rotational ductility, stiffness, and toughness) were (88, 106 NSM CFRP strips at a spacing value of 75 mm, showed clear
and 106%), (95, 109 and 113%), and (75, 122 and 82%), degradation with main characteristics reduced to (154, 85,
respectively. Thus, the duplicate beams, strengthened with 113 and 111%), respectively. Here, the strips were exactly
three NSM CFRP strips, achieved the highest flexural load aligned at the same lateral position as the main reinforcing
capacity and stiffness, yet the lowest rotational ductility and steel. Therefore, the relatively thin concrete layer, located
toughness. It can be stated that the higher improvement in between the strips and reinforcing steel, demonstrated low
ultimate load capacity upon the use of more NSM CFRP resistance to horizontal shearing stress. This resulted in a
strips is justified by the increase in the cross-sectional area premature separation of concrete cover at relatively low
of the strips, extended to a satisfactory embedment length loads [10].
of 250 mm. On the contrary, the increase in horizontal shear
stress interaction upon the implementation of more NSM 5.4 Effect of Staggering of Embedment Lengths
CFRP strips had caused premature concrete peeling-off at
relatively lower deflections, undermining both ductility and Figure  7 shows load–deflection curves for the concrete
toughness [6, 10, 11]. beams, strengthened using three NSM CFRP strips without
and with embedment length staggering, as compared to that
5.3 Effect of Lateral Location of NSM CFRP Strips of control (C) ones. The curves for load versus deflection
followed similar trend behavior as that observed for other
Figure  6 shows load–deflection curves for the beams, strengthened beams. As can be deduced from Table 6, the
strengthened with two NSM CFRP strips at an embedment residuals’ characteristics in terms of (load capacity, stiffness,
length of 450 mm and different spacing values of 50, 75,
100 mm, were compared to that of control (C) ones. The
curves showed linear behavior at the initial stages of loading 250
before turning nonlinear thereafter. According to Table 6,
200
the residuals for (load capacity, rotational ductility, stiffness,
toughness) for concrete beams with two NSM CFRP strips 150
Load (kN)

extended beyond the high moment zone to 450 mm, and

spaced laterally at 50 and 100 mm, were (161, 95, 111, and 100
141%) and (164, 90, 112 and 113%), respectively. These C_SN3_EL(2x15 -1x45)_SP_5
C_SN3_EL(15)_SP_5
mechanical responses reflected relatively low sensitivity 50
C
towards strips’ spacing, contrary to the case with residual
0
ultimate strain, which experienced increasing trend as strips’
0 5 10 15 20
spacing was increased: the residual ultimate strain was Deflecon (mm)
increased from 61 to 67% when spacing was increased from
50 to 100 mm with a noticeable change in the failure mode
Fig. 7  Load versus deflection for beams strengthened using three
of strengthened beams, as discussed later on. On the other NSM CFRP strips at a spacing value of (50  mm) without and with
hand, the mechanical behavior for beams, strengthened with embedment lengths staggering

13
International Journal of Civil Engineering

rotational ductility, and toughness) were (114, 111, 95, and 6 Mode of Cracking and Failure Under
95%) and (143, 105, 125, and 146%) for the cases without Loading
and with strips’ staggering, respectively. This suggests that
staggering of NSM CFRP strips in concrete beams imparted 6.1 Effect of Embedment Length of CFRP NSM Strips
additional improvement in the residuals for flexural capacity,
rotational ductility, and toughness of strengthened beams The failure modes for beams, strengthened with two
by about 29, 30, and 51%, respectively, yet caused a slight NSM CFRP strips at embedment lengths of 150, 250 and
reduction in stiffness by 6%. The strain in NSM CFRP strips 450 mm, are shown in photos in Fig. 8. With load applica-
of the case with staggering increased by 70% as compared tion, flexural cracks initiated at the middle zone at loads
to the case without staggering, supporting the significant of 50, 52 and 59 kN, respectively, before spreading along
improvement reported in flexural capacity. The noticeable the entire span. With further load application, end-cover
benefit from staggering can be attributed to the reduced separation was initiated by the formation of vertical crack
interaction of horizontal shearing stresses between NSM across the concrete cover at the termination of the NSM
CFRP strips due to the subsequent termination of these strips CFRP strips at loads of 75, 100 and 175 kN, respectively.
beyond the high moment zone [6, 10, 11]. This helped in This was followed by the propagation of a major horizon-
improving the ultimate load capacity, ductility and tough- tal crack at the level of the tension reinforcement towards
ness at failure due to delayed concrete cover detachment. the high-tension zone, until the crushing of concrete com-
The above results encourage further investigation to substan- pression zone. Hence, debonding-type failure (end-cover
tiate the contribution of staggering to promoting structural separation) was recognized.
performance.

Fig. 8  Mode of failure for beams reinforced with different embedment lengths of NSM CFRP strips (side view: left; bottom view: right)

13
International Journal of Civil Engineering

6.2 Effect of Number of CFRP Strips 6.4 Effect of Staggered NSM CFRP Strips

The failure modes for concrete beams, strengthened with The failure modes for beams, strengthened with three NSM
one and three NSM CFRP strips at an embedment length of CFRP strips at an embedment length of 150 mm, were simi-
250 mm, are shown in photos in Fig. 9. With load applica- lar to those, strengthened with three CFRP strips at embed-
tion, flexural cracks initiated at the middle zone at loads ment lengths of 450 mm for the middle strip and 150 mm for
of 52 and 50 kN, respectively, before spreading along the the edge strips. Flexural cracks initiated at the middle zone
beams’ spans. Later, end-cover separation was initiated by at loads of 50 and 52 kN before spreading along the beams’
the formation of a vertical crack across the concrete cover spans, respectively. End-cover separation was initiated by
at the termination of the NSM CFRP strips at loads of 130 the formation of vertical crack across the concrete cover
and 95 kN, respectively. This was followed by propagation at the termination zones of the 150-mm-NSM CFRP strips
of a major horizontal crack at the level of the tension rein- when loads of 65 and 60 kN were reached, respectively. This
forcement towards the high-tension zone, until crushing of was followed by propagation of a major horizontal crack
concrete with end-cover separation clearly indentified. at the level of the tension reinforcement towards the high-
tension zone, until crushing of concrete in the compression
zone.
6.3 Beams Strengthened with NSM CFRP Strips
at Different Lateral Locations
7 NSM CFRP Strips’ Key Repair Parameters:
The failure modes for concrete beams, strengthened with two An Overview
NSM CFRP strips at same embedment length of 450 mm,
yet having different spacing values of 50 and 100 mm, are The present findings with regard to the impact of repair
shown in photos in Fig. 10. With load application, flexural key parameters were validated through comparisons with
cracks were initiated at the middle zone at loads of 45 and three literature works similar to the present one. Further-
42 kN, respectively, before spreading along the beams spans. more, other key parameters, not considered in the present
Prior to failure, cracks extended deeper into the compression work, were highlighted and a short critique review presented
zone, and flexure-type failure was recognized. This was not regarding the efficiency of published empirical models for
the case when the two strips were placed at 75 mm along the prediction of repair efficiency using NSM CFRP strips.
the same longitudinal alignment as main steel reinforcement, The findings of three different literature works are sum-
where failure occurred by end-cover separation, as shown marized in Table 7 along with those of the present study [8,
in Fig. 9. 12, 16]. The data reflected the impact of most key parameters

Fig. 9  Mode of failure for beams reinforced with different numbers of NSM CFRP strips (side view: left; bottom view: right)

13
International Journal of Civil Engineering

Fig. 10  Mode of failure for beams reinforced with two CFRP strips at embedment length of 45 cm and different lateral spacing (side view: left;
bottom view: right)

considered in literature. These included embedment length, CFRP strips upon spacing increase between strips). Gener-
number of NSM CFRP strips, spacing between strips, and ally, as spacing and embedment lengths were increased, ulti-
stiffness of NSM CFRP strips. Other key parameters, not mate strains in NSM CFRP strips tended to increase because
reflected in the presented data, included man-made groove of the reduction in the effect of strip–strip interaction [6, 10,
size, spacing between strips and reinforcement, and rein- 11]. This means that interactions between load-induced hori-
forcement magnitude and distribution [17]. zontal shearing stresses at concrete cover level become inef-
The present findings indicate that higher embedment fective, especially near the termination points of NSM CFRP
lengths help in delaying NSM CFRP strips’ end separation. strips. Consequently, peeling-off concrete cover would be
Therefore, using the highest possible extension of strips, delayed or prevented, and hence, the ultimate load capacity
from points of maximum moment to supports, would have and stiffness as well as strain induced in NSM CFRP strips
additional advantage for repair represented with increased would be increased.
load capacity, stiffness and toughness. These improvements Using NSM CFRP strips of similar embedment lengths
cannot be related solely to the increased bond between the but at a higher number in the range of 1–3 per beam had
strips and concrete as a result of using higher embedment a beneficial impact on improving flexural performance in
lengths, but rather to the shift in the termination point of terms of induced strain in the strips at failure, as can be
strips towards supports, thus delaying the transverse crack deduced from the present findings as well as those by
triggered due to sudden moment reduction [7]. The spacing Haddad and Almomani [16] and Hong et al. [12]. Of course,
between constant numbers of NSM CFRP strips across the further investigation is needed to establish the impact of
width of the strengthened beams had a tangible impact upon using strips at numbers greater than three per beam on
flexural performance, hence inducing strain in strips at fail- the structural characteristics of NSM CFRP-strengthened
ure (Table 7 shows 17% increase in strain at failure of NSM beams.

13
International Journal of Civil Engineering

Table 7  Present and literature Beam designation Ref. fc′ (MPa) n Geometric properties (mm) ϵf ϵf
ultimate strains strips for (µs) ϵfu
(%)
varying material and geometric GW GD tf Wf S Lf
properties of NSM CFRP-
strengthened beams C_SN2_EL (15) P 35 2 8 25 2.5 15 75 150 4300 25
C_SN2_EL (25) 35 2 8 25 2.5 15 75 250 6700 39
C_SN2_EL (45) 35 2 8 25 2.5 15 75 450 11,400 67
C_SN1_EL (25) 35 1 8 25 2.5 15 75 250 7600 45
C_SN3_EL (25) 35 3 8 25 2.5 15 50 250 5200 31
C_SN3_EL (15) 35 3 8 25 2.5 15 50 150 3400 20
C_SN3_EL (2 × 15 − 1 × 45) 35 3 8 25 2.5 15 50 250 6000 35
C_SN2_EL (45) 35 2 8 25 2.5 15 50 450 11,000 65
C_SN2_EL (45) 35 2 8 25 2.5 15 100 450 12,000 71
HD_SN2_EL (15) Ref. [16] 10 2 8 25 2.5 15 75 150 2200 13
HD_SN2_EL (25) 10 2 8 25 2.5 15 75 250 3400 20
HD_SN2_EL (45) 10 2 8 25 2.5 15 75 450 4200 25
HD_SN3_EL (15) 10 3 8 25 2.5 15 50 150 2000 12
ALII-2 Ref. [12] 27 2 4 15 2.8 10 78 1500 10,410 61
AMI-2 34 2 4 15 2.8 10 79 1500 11,285 66
AMII-2 34 2 4 15 2.8 10 78 1500 10,034 59
AMIII-1 34 1 4 15 2.8 10 125 1500 10,953 64
AMIII-2 34 2 4 15 2.8 10 76 1500 11,359 67
AMIII-3 34 3 4 15 2.8 10 51 1500 12,013 71
PMIV-1 34 1 4 15 2.8 10 200 1500 9873 58
PMIV-2 34 2 4 15 2.8 10 151 1500 11,702 69
PMIV-3 34 3 4 15 2.8 10 101 1500 11,456 67
B-N-1-2 Ref. [8] 35 1 15 25 2.4 25 75 1950 9700 57
B-N-2-2 35 2 15 25 1.2 25 50 1950 12,800 75

P present results, b beam’s width, n number of strips, GW groove width, GD groove depth,
tf thickness of NSM CFRP strips , Wf width of NSM CFRP strips, S spacing between strips, Lf embedment
length from point of high moment region to support, 𝜀f strain in NSM CFRP strips at failure

The compressive strength of concrete had a clear impact in the latter model is represented in ignoring the impact of
on the flexural performance, as supported by present findings strip–strip interaction, which would have a clear effect upon
as well as those by Haddad and Almomani [16] and Hong end-cover separation of strips, especially those inserted in
et al. [12]. The ultimate strain of NSM CFRP strips showed concrete beams at close lateral distances. Accordingly, it is
drastic reductions that exceeded 50% when compressive essential to develop new empirical models based on a vast
strength was reduced by more than 70% and a mild reduction range of literature data, while accounting for different key
of less than 8% when compressive strength was reduced by variables, including NSM CFRP strip–strip interaction [10].
20%. Of course, the impact of strength is reflected in tension Such a model would be a great asset for the future design of
cracking resistance as well as in the bond between strips and flexural repair using NSM CFRP strips.
concrete [10]. The stiffness of NSM CFRP strips, not studied
in the present work, had a clear impact on the ultimate strain
of NSM CFRP strips, resulting in repair efficiency. The work 8 Prediction of Ultimate Load Capacity
by Khalifa [8] showed that the latter was increased by about of Strengthened Beams
32% when a single strip, instead of two NSM CFRP strips,
was inserted in 15-mm-wide grooves. The experimentally obtained ultimate load capacity for NSM
In fact, most literature models that proposed to predict CFRP-strengthened beams were compared to analytical pre-
ultimate strain in NSM CFRP strips at failure had considered dictions according to ACI 440R [18]. First, strains in concrete
either very limited number of the key parameters, as is the and steel were determined using Eq. (1) based upon the geo-
case with the models by Barros and Kotynia and ACI com- metric characteristics of the beams’ cross section and strains
mittee 440.2 [18], or overlooked major key factors, as is the measured in NSM CFRP strips at failure moment, while
case with the model by Teng et al. [19]. The shortcoming assuming that the plain section before bending remains plain

13
International Journal of Civil Engineering

after bending. The compressive force in the compression side ( )

3c
was then computed according to the Hognestad model, incor- Cc 8
+ Cs� (c� )
porated into Eq. (2) [20], with εo and εu taken as 0.002 and dc =
Cc + Cs�
0.003, respectively, whereas the forces in compression and
tension steel and NSM CFRP strips were determined using
T (d) + Tf (df ) (6)
dt = s
Eq. (3). Following this, the neutral axis location, c, was deter- Ts + Tf
mined based upon the equilibrium of the different forces, pro- 2M
Pu =
vided in Eq. (4). Accordingly, the resultant compression forces x
in concrete, C, and tension, T, acting at distances dc and dt,
respectively, from the top fiber of the concrete were obtained. where Pu is the ultimate load capacity, and x is the distance
Finally, the theoretical ultimate moment and load were com- from either support to the nearest point load.
puted using Eqs. (5) and (6). Theoretical load capacities for different strengthened
{
c d−c c − c� beams along with those experimentally measured are listed in
𝜀c = 𝜀f 𝜀s = 𝜀f 𝜀�s = 𝜀 , (1) Table 6. The error of prediction ranged from 4 to 13%, with
df − c df − c df − c f
an error average of 8%. This showed that the measured load
where 𝜀c , 𝜀s , 𝜀′s and 𝜀f are strains in concrete, tension and capacity for different beams was accurate and proportional to
compression steel, and CFRP strips, respectively, in terms reinforcement quantities and configuration.
of distance between NSM CFRP strips and top concrete sur-
face, df, depth of neutral axis from the top surface of beams,
c, and distance between compression steel center and upper 9 Conclusion
concrete surface, c′.
� Based on the above discussion, the following conclusions
� � � ��
⎧ C = 2 ∗ f � ∗ 2*𝜀c − 𝜀c 2 bc can be derived:
0 < 𝜀c < 𝜀o
⎪ 3 c
� �
𝜀o 𝜀o
⎨ � � � � ,
⎪ C = 2 fc� 1 − 0.15 (𝜀c − 𝜀o
2
bc 𝜀 < 𝜀 < 𝜀 1. Strengthening reinforced concrete beams with NSM
⎩ 3 𝜀u −𝜀o o c u
CFRP strips contributed to improving their load capacity
(2) while reducing their ductility, depending on the embed-
where fc is the compressive strength of concrete whereas b
′
ment length. The residual ultimate load capacity and
is the width of the beams. rotational ductility achieved its highest and lowest at an
embedment length of 450 mm, at 154 and 85%, respec-
⎧ C� = A� E 𝜀� Es 𝜀�s ⩽ fy tively.
⎪ s s s s
2. The noticeable improvement in load capacity and stiff-
⎨ Ts = As Es 𝜀s Es 𝜀s ⩾ fy , (3)
⎪ Tf = Af Ef 𝜀f ness of concrete beams, strengthened with three instead
⎩ of one or two NSM CFRP strips, was undermined by
the reduction in rotational ductility. Strengthening with
where Cs′ and Ts are compression and tension forces in top NSM CFRP strips at one or two layers resulted in lim-
and bottom steel, respectively, and Tf is the tension force ited improvements in ultimate load capacity and stiffness
in NSM CFRP strips. A′s and As are cross-sectional areas while preserving ductility and toughness.
of top and bottom steel reinforcement, whereas Af is the 3. The findings of the present work showed that positioning
cross-sectional area of NSM CFRP strips. Es′ and Es are NSM CFRP strips at different longitudinal alignments
cross-sectional areas of top and bottom steel reinforcement, of main reinforcing steel of present beams contributed
whereas Ef is the cross-sectional area of NSM CFRP strips. to avoiding failure through end-cover peeling-off.
∑ ∑ 4. Staggering NSM CFRP strips in strengthened beams
F = (T − C) = 0, (4) imparted improvement in the residual flexural capac-
ity and residual toughness by as much as 30 and 51%,
where C and T are compression and tension forces acting
respectively, yet caused a slight reduction in stiffness by
on the beam’s cross section.
6% as compared to the case without.
M = (C or T) (dt − dc ), (5) 5. The literature data relevant to the effect of NSM CFRP
where M is the moment capacity of the beams, and dt and strips’ repair parameters confirmed the present findings
dc are computed as follows: and emphasized the need for more representative empiri-
cal models to predicted ultimate strain in strips at failure.
6. The analytical predictions for ultimate load capacity
experimentally confirmed the results with an average

13
International Journal of Civil Engineering

prediction error of 8%. Further modeling using nonlinear 9. Peng H, Zhang J, Cai CA, Liu Y (2014) An experimental study
finite element would allow generating more data while on reinforced concrete beams strengthened with pre-stressed near
surface mounted CFRP strips. Eng Struct 79:222–233. https:​ //doi.
considering various key parameters that could aid in the org/10.1016/j.engst​ruct.2014.08.007
development of empirical models for practical design of 10. Seo S, Feo L, Hui D (2013) Bond strength of near surface mounted
flexural repair with NSM CFRP strips. FRP plates for retrofit of concrete structures. Compos Struct
95:719–727. https​://doi.org/10.1016/j.comps​truct​.2012.08.038
11. Lee D, Cheng L (2013) Bond of NSM systems in concrete
strengthening—examining design issues of strength, groove
Acknowledgements  The authors acknowledge the technical and finan- detailing and bond-dependent coefficient. Constr Build Mater
cial support provided by the research deanship at Jordan University of 47:1512–1522. https:​ //doi.org/10.1016/j.conbui​ ldmat​ .2013.06.069
Science and Technology (project number 153/2015) and the assistance 12. Hong KN, Han JW, Seo DW, Han SH (2011) Flexural response
by the technicians at the structural and materials laboratory via the of reinforced concrete members strengthened with near-surface-
Department of Civil Engineering. mounted CFRP strips. Int Phys Sci 6(5):948–961. https​://doi.
org/10.5897/IJPS1​0.222
Funding  Dean of Scientific Research, Jordan University of Science 13. American Concrete Institute (2008) Building code requirements
and Technology, P.O. Box 3030, 22110 Irbid, Jordan. for structural concrete and commentary (ACI-318). American
Concrete Institute, Farmington Hills
14. ASTM International (2004) American Society for Testing and
Materials International (ASTM). ASTM International, West
References Conshohocken
15. ACI Manual of Concrete Practice (2008) Standard practice for
1. Sun J, Huang Q, Ren Y (2015) Performance deterioration of cor- selecting proportions for normal, heavyweight, and mass concrete
roded RC beams and reinforcing bars under repeated loading. (ACI 211.1), Part I: Materials and general properties of concrete.
Constr Build Mater 96: 404–415. https​://doi.org/10.1016/j.conbu​ ACI, Detroit
ildma​t.2015.08.066 16. Haddad R, Almomani O (2017) Recovering flexural performance
2. Demir A, Tekin M, Turali T, Bagci M (2014) Strengthening of RC of thermally damaged concrete beams using NSM CFRP strips.
beams with prefabricated RC U cross-sectional plates. Struct Eng Constr Build Mater 154:632–643. https:​ //doi.org/10.1016/j.conbu​
Mech 49(6):673–685. https:​ //doi.org/10.12989/​ sem.2014.49.6.673 ildma​t.2017.07.211
3. Jones R, Swamy RN, Charif A (1988) Plate separation and anchor- 17. Barros J, Kotynia R (2008) Possibility and challenges of NSM for
age of reinforced concrete beams strengthened by epoxy-bonded the flexural strengthening of RC structures, In: Fourth interna-
steel plates. Struct Eng 66(5):85–94 tional conference on FRP composites in civil engineering (CICE
4. Raval S, Dave U (2013) Effectiveness of various methods of 2008), Zurich, p 6
jacketing for RC beams. Proc Eng 51:230–239. https​: //doi. 18. ACI Committee 440 (2008) Guide for the design and construc-
org/10.1016/j.proen​g.2013.01.032 tion of externally bonded FRP systems for strengthening concrete
5. Haddad RH, Al-Rousan RZ, Al-Sedyiri BK (2013) Repair of structures (440.2R-08). American Concrete Institute, Farmington
shear-deficient and sulfate-damaged reinforced concrete beams Hills, p 76
using FRP composites. Eng Struct 56:228–238. https​: //doi. 19. Teng JG, Zhang SS, Chen JF (2016) Strength model for end cover
org/10.1016/j.engst​ruct.2013.05.007 separation failure in RC beams strengthened with near-surface
6. Ceroni F, Pecce M, Bilotta A, Nigro E (2012) Bond behavior of mounted (NSM) FRP strips. Eng Struct 110:222–232. https​://doi.
FRP NSM systems in concrete elements. Compos B 43:99–109. org/10.1016/j.engst​ruct.2015.11.049
https​://doi.org/10.1016/j.compo​sites​b.2011.10.017 20. Hognestad E, Hanson NW, McHenry D (1955) Concrete stress
7. Haddad RH, Almasaeid HH (2016) Recovering shear capacity of distribution in ultimate strength design. ACI J Proc 52(4):455–480
heat-damaged beams using NSM CFRP strips. Constr Build Mater
105:448–458. https​://doi.org/10.1016/j.conbu​ildma​t.2015.12.152
8. Khalifa AM (2016) Flexural performance of RC beams strength-
ened with near surface mounted CFRP strips. Alex Eng J
55(2):1497–1505. https​://doi.org/10.1016/j.aej.2016.01.033

13