Experimental investigation into bond behavior of CFRP sheets attached to concrete using EBR and EBROG techniques

Composites: Part B 51 (2013) 130–139

Contents lists available at SciVerse ScienceDirect

Composites: Part B journal homepage: www.elsevier.com/locate/compositesb

Experimental investigation into bond behavior of CFRP sheets attached to concrete using EBR and EBROG techniques Ardalan Hosseini ⇑, Davood Mostofinejad Department of Civil Engineering, Isfahan University of Technology (IUT), Isfahan 84156-83111, Iran

a r t i c l e

i n f o

Article history: Received 4 November 2012 Received in revised form 27 February 2013 Accepted 10 March 2013 Available online 20 March 2013 Keywords: A. Carbon fiber B. Debonding B. Strength D. Mechanical testing Grooving method (GM)

a b s t r a c t Over the last decade, an extreme increase in the application of fiber reinforced polymers (FRPs) for strengthening of reinforced concrete (RC) structures has been observed. The most common technique for strengthening of RC members utilizing FRP reinforcements is externally bonded reinforcement (EBR) technique. Despite certain benefits of the technique such as simple and rapid installation, the main problem which has greatly hampered the use of EBR method is premature debonding of FRP composite from concrete substrate. Recently, grooving method (GM) has been introduced as an alternative to conventional EBR technique. Grooving with the special technique of externally bonded reinforcement on grooves (EBROG) has yielded promising results in postponing or, in some cases, completely elimination of undesirable debonding failure in flexural/shear strengthened RC beams. Consequently, the main intention of the current study is to make a comparison between FRP-to-concrete bond behavior of EBR and EBROG techniques by means of single-shear bond tests. To do so, CFRP sheets were adhered to 16 concrete prism specimens using EBR and EBROG techniques. The specimens were then subjected to single-shear bond test and the results were compared. A non-contact, full field deformation measurement technique, i.e. particle image velocimetry (PIV) was utilized to investigate the bond behavior of the strengthened specimens. Successive digital images were taken from each specimen undergoing deformation during the test process. Images were then analyzed utilizing PIV method and load–slip behavior as well as slip and strain profiles along the strengthening CFRP strips were reported. Experimental results of the current study strongly verify the capability of GM for strengthening RC members to completely eliminate the debonding failure. Ó 2013 Elsevier Ltd. All rights reserved.

1. Introduction Fiber reinforced polymers (FRPs) have emerged as a well-accepted alternative to conventional materials for repair and rehabilitation of existing RC structures. The growing interest in FRP applications for strengthening and retrofitting can be attributed to the material’s unique advantages such as lightweight, noncorrosive, and high tensile strength [1]. Externally bonding of FRP composites to the tension face of RC members is the primitive technique for flexural strengthening of RC beams and slabs called externally bonded reinforcement (EBR) technique [2,3]. Although EBR technique has been widely utilized in strengthening projects all around the world due to certain benefits such as simple and rapid installation, however, the technique suffers from premature debonding of FRP composites from concrete substrate [4–7]. Since all strengthening calculations greatly rely on FRP-to-concrete bond strength which is seriously affected by debonding phenomenon, ⇑ Corresponding author. Tel.: +98 311 3913818, mobile: +98 937 805 8734; fax: +98 311 391 2700. E-mail addresses: [email protected], [email protected] (A. Hosseini). 1359-8368/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.compositesb.2013.03.003

lots of analytical and experimental studies have been conducted to investigate debonding failure and some analytical and empirical models have been proposed to predict the bond strength of FRP-toconcrete. Holzenkampfer [8], Neubauer and Rostasy [9], Chen and Teng [10], Seracino et al. [11], and Bilotta et al. [12] are from those who proposed well-known FRP-to-concrete bond strength models. Moreover, existing FRP guidelines such as ACI 440.2R-08 [1], fib14 [13] and CNR-DT200 [14] present debonding models usually based on aforementioned models. However, multiplicity of debonding failure modes in externally strengthened RC members including cover separation, plate-end interfacial debonding, intermediate (flexural or flexural-shear) crack (IC) induced interfacial debonding and critical diagonal crack (CDC) induced interfacial debonding, makes the problem more complex [15]. Since all debonding failure mechanisms are categorized as undesirable brittle failure modes, existing strengthening guidelines limit strain value in FRP reinforcements to the strain level at which debonding may occur [1]. As a result, the whole capacity of composite reinforcements could not be utilized when using EBR technique due to debonding failure. Consequently, numerous experimental studies have been conducted to improve the performance of EBR

A. Hosseini, D. Mostofinejad / Composites: Part B 51 (2013) 130–139

technique and mitigate the risk of premature debonding failure by utilizing different anchoring devices including the use of transverse sheets or straps [16–18], using mechanical anchors [19], wrapping the end of sheets in rods embedded in grooves formed into the concrete [20–22], anchoring the FRP sheets using FRP anchors or FRP fans [21–25], or anchoring by means of bi-directional fabric wrap [26]. It is important to recognize, however, all anchoring systems are highly problematic due to the brittle and anisotropic nature of composite materials; as a result, any proposed method of anchorage should be heavily scrutinized before field implementation [1]. Moreover, although many of aforementioned anchoring methods have shown acceptable results, but unfortunately existing information has not been provided in sufficient detail that would allow development of design procedures [25]. Grooving method (GM) consisted of cutting grooves into tension face of concrete beams, filling them with proper epoxy resin and bonding FRP sheets on the member’s surface over the filled grooves, has been recently introduced by Mostofinejad and Mahmoudabadi [27]. They investigated the effect of transverse, diagonal and longitudinal grooves in GM on ultimate load capacity and failure modes of concrete beam specimens strengthened in flexure by CFRP. Cutting 3 longitudinal grooves having 3  10 mm (width  depth) dimensions was reported to completely eliminate debonding of CFRP sheets from concrete beam specimens and increase the ultimate loads up to 80% compared with conventional EBR strengthened specimens [27]. Grooving method in its original form was later referred as externally bonded reinforcement on grooves (EBROG) technique; and its performance for flexural strengthening of RC beams using multilayer CFRP sheets was more investigated by Mostofinejad et al. [28] and promising results were reported in completely elimination of debonding failure and, subsequently, reaching full capacity of composite materials. Recently, Mostofinejad and Tabatabaei Kashani [29] investigated the performance of EBROG technique for shear strengthening of RC beams. Based on their experimental study, using vertical grooves as a substitute to conventional EBR technique prior to shear strengthening of beams by CFRP sheets changed the failure mode from shear to flexural and increased ultimate load capacity of the beams up to 15 percent compared to EBR strengthened specimens; while no debonding was reported in CFRP shear reinforcements [29]. More recently, another technique for GM was introduced and named as externally bonded reinforcement in grooves (EBRIG), in which the FRP is formed in contact with the surfaces of the grooves using wet layup procedure [30]. Despite certain benefits of the GM for flexural/shear strengthening of RC members by FRP sheets such as obviate the need for costly and time-consuming traditional surface preparation and, reduce the cost of strengthening projects due to postponing or, in some cases, completely elimination of debonding; no research has been conducted up to now to investigate bond behavior of FRP sheets attached to concrete using GM in the form of EBROG technique. Consequently, the main intention of the current study is to experimentally investigate the FRP-to-concrete bond behavior in GM by means of single-shear bond tests.

2. Experimental procedure 2.1. Specimens’ detail and material characteristics In order to carry out the single-shear tests, 16 concrete prisms with dimensions of 150  150  350 mm were cast using steel molds. To obtain a compressive strength of about 30 MPa, 425 kg/ m3 of normal Portland cement, 892 kg/m3 sand, 736 kg/m3 coarse aggregate and, 221 kg/m3 water were used. Moreover, to measure the compressive strength of the concrete, three 150  300 mm

131

cylindrical specimens were also cast from each mixer. The molds were removed after 24 h and all the specimens were cured in a water bath placed in the laboratory for 28 days at standard conditions [31]. Afterwards, concrete specimens were removed from water bath and were kept in laboratorial condition until testing. The FRP composite materials used in this study were fabricated via wet layup process and were made up of unidirectional SikaWrap 230C carbon fibers. Furthermore, based on the manufacturer suggestion, a two-component epoxy adhesive, Sikadur 330, was used for bonding FRP composite to the concrete substrate as well as for matrix phase of the FRP composite. Mechanical properties of carbon fibers and epoxy resin utilized in this research are provided in Table 1 according to the manufacturer’s catalogue. It should be noted that the presented value for ultimate capacity of fibers, 4300 MPa, cannot be achieved when using wet layup process and design rupture stress for the composite suggested by manufacturer is 2670 MPa. The ancillary tests which were conducted by the authors on the basis of the ASTM D3039-00 standard [32] indicated an average ultimate stress of 2795 MPa for the FRP composite fabricated through wet layup procedure. 2.2. Testing layout and strengthening methods In order to evaluate the bond strength of CFRP sheets attached to concrete using grooving method, eight out of 16 prism specimens were strengthened using GM in the form of EBROG technique and the other eight were strengthened using conventional EBR technique. The factors considered in the present test program included the FRP-to concrete bond length, Lf, and strengthening method while all other experimental parameters remained constant. The notation of tested specimens is T–m–n, where T refers to strengthening technique, i.e. EBR or GM techniques; m identifies FRP bond length in mm, Lf; and n distinguishes the ordinal number of each test (1 or 2). All strengthening CFRP sheets had a constant width of 48 mm and were cut into proper sizes before being used for strengthening. To strengthen the specimens using conventional EBR technique, first the weak layer of the concrete surface was removed using a grinding machine. Then the concrete surface was properly cleaned with air jet to remove dust. Finally, carbon fiber sheets were cut into selected sizes and were adhered to concrete surface using epoxy resin in wet layup procedure. However, traditional surface preparation was completely eliminated when using EBROG technique. Two 5  10 mm (width  depth) longitudinal grooves having an internal side-to-side distance of 20 mm were cut into each specimen face. The grooves length was equal to the bond length of each specimen, i.e. 75, 100, 125 or 150 mm. The grooves were then cleaned with air jet to remove dust and were fully filled with epoxy Sikadur 330. Immediately, carbon fiber sheets were bonded to the concrete surface over the grooves (Fig. 1a–c). In order to provide full field deformation measurements using image analysis technique, successive digital images should be taken from the field undergoing deformation. It is necessary for the images to have a texture to create features upon which image processing can properly operate. Since neither the CFRP sheets nor concrete surface show suitable texture, natural colored sand between sieve Nos. 50 and 100, obtained from mixing same fraction of five different colors was embedded to all specimens’ face just after end of strengthening process and before epoxy hardening (Fig. 1d). Afterwards, all the specimens were cured for at least 7 days in laboratorial condition before testing. 2.3. Test setup All strengthened specimens were subjected to single-shear test by means of a 300 kN displacement control hydraulic jack, spe-

132

A. Hosseini, D. Mostofinejad / Composites: Part B 51 (2013) 130–139

Table 1 Properties of FRP materials.

Fibers Adhesive

Type

Thickness (mm)

Tensile strength (MPa)

Elastic modulus (GPa)

Elongation at break (%)

SikaWrap 230C Sikadur 330

0.131 0.5–0.9

4300 30

238 4.5

1.8 1.5

Fig. 1. Strengthening via grooving method (EBROG technique): (a) cutting longitudinal grooves; (b) filling the grooves using epoxy; (c) adhering fibers on the surface using epoxy; and (d) embedding natural colored sand before epoxy hardening.

cially designed for single-shear bond test in structural laboratory of Isfahan University of Technology (IUT). One CCD-camera (charge couple device), i.e. Nikon D80 with resolution of 10.0 megapixels (3872  2592 pixels) having a Nikkor 18135 mm lens was placed perpendicular to the specimens’ face at a distance equal to 1.0 m and digital images were taken from each specimen undergoing deformation using a remote control at regular intervals. The specimens were illuminated using two white light projectors to eliminate any probable parasitic lights.

A digital data logger was used to monitor the load cell and image numbers simultaneously. Specimen dimensions, loading arrangement, and testing machine are presented in Fig. 2. As it is illustrated in Fig. 2a, in order to eliminate stress concentration at the loaded edge of the concrete prism, a 35 mm unbonded zone was introduced along the interface between the CFRP sheet and the concrete surface. Similar setups for eliminating stress concentration in single-shear bond tests have been reported by other researches [33,34].

A. Hosseini, D. Mostofinejad / Composites: Part B 51 (2013) 130–139

P

P CFRP

Support Support Unbonded zone (35 mm)

Lf 350 mm

Concrete prism

bf =48 mm

Support

bc =150 mm

hc =150 mm

(a)

(b)

133

(c) Normalize by cross-correlation of search patch with mask function (the most likely displacement is the peak of the function). (d) Find sub-pixel location of correlation peak using bicubic interpolation. (e) Repeat the procedure for other test patches of first image. (f) Repeat for all images in series and calculate displacement vectors. (g) Calculate strain tensor using obtained displacement field. As detailed discussion about PIV technique is beyond the scope of the current paper, the readers are referred to White and Take [39], White et al. [36], and Slominski et al. [37] for further details of the PIV technique including detailed procedure, source of errors and precision validation experiments. Further consideration required for digital image analysis technique is given by Carloni and Subramaniam [40]. In the current study, successive digital images were taken from each specimen undergoing deformation during test process. Images were then analyzed using GeoPIV8 software, developed at Cambridge University [39]. All PIV analyses due to slip measurements were undertaken using 256  256 pixels patches, while PIV analyses due to strain field measurements were undertaken using 128  128 pixels patches. A search area of 5  5 pixels for each pair of successive images was considered in all PIV analyses which provided sufficient area to give good tracking of the patches. According to White et al. [36] validation experiments, the precision of the PIV is a strong function of patch size; the larger the patch size, the smaller the scatter. Utilizing 128  128 pixels patches conservatively results an accuracy of about 0.005 pixels which surpasses 1:800,000 of FOV (field of view) by considering a 10.0 megapixel (3872  2592 pixels) camera. Consequently, the accuracy of displacement measurements of the current study surpasses 0.5 lm. 4. Results and discussions

Fig. 2. (a) Specimen dimensions and loading arrangement; and (b) test setup.

3. Image analysis using particle image velocimetry (PIV) An image-based deformation measurement technique, i.e. particle image velocimetry (PIV) was used for full field deformation measurements. PIV is originally a velocity-measuring technique developed in the field of experimental fluid mechanics [35]. The technique was originally implemented using double-flash photography of a seeded flow and the resulting photographs were divided into a grid of test subsets called ‘‘patches’’. For PIV analysis, displacement vector of each patch during the interval between the flashes is found by locating the peak of an autocorrelation function of each patch. The peak in the autocorrelation function indicates that the two images of each seeding particle overlying each other, so the correlation offset is equal to the displacement vector [36]. A modified approach was used to implement PIV in geotechnical testing by White et al. [36]. Based on validation experiments of White et al. [36], Slominski et al. [37], and Hajialilue-Bonab et al. [38], the modified PIV technique offers an order-of-magnitude increase in accuracy, precision, and measurement array size compared with previous image based methods of displacement measurement. PIV analysis procedure can be summarized as follows [36,37]: (a) Subdivide the first image into subsets (patches) and select a test patch. (b) Evaluate cross-correlation of selected test and search patch using fast Fourier transform (FFT).

Specimen characteristics as well as experimental test results including compressive strength, CFRP bond length, ultimate load, and failure modes of each specimen are presented in Table 2. Moreover, since the relative displacement (slip) between CFRP reinforcement and concrete substrate is a key factor affecting FRP-to-concrete bond behavior, ultimate slip at loaded end of CFRP reinforcements for all tested specimens are also presented in Table 2, and will be discussed later in the article. For comparison purposes, debonding load corresponding to EBR specimens calculated from Chen and Teng’s bond strength model, PChen & Teng, are also provided in Table 2 from following formulations [10]:

PChen

& Teng

¼ abw bf Le

sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 2  bf =bc bw ¼ 1 þ bf =bc sffiffiffiffiffiffiffiffi Ef tf Le ¼ pffiffiffiffi fc

pffiffiffiffi fc

ð1Þ

ð2Þ

ð3Þ

where bf, tf and Ef are the width, thickness, and elasticity modulus of the FRP, respectively; bc is width of the concrete block specimen, and fc is mean cylindrical compressive strength of concrete. Furthermore, a = 0.427 to obtain a mean prediction and Le = Chen and Teng’s effective bond length expressed by Eq. (3). Comparing experimental debonding loads in Table 2, Ptest, with PChen & Teng for EBR strengthened specimens, clearly verifies the

134

A. Hosseini, D. Mostofinejad / Composites: Part B 51 (2013) 130–139

Table 2 Specification of specimens and test results. Test specimen

fc (MPa)

Lf (mm)

PChen & Teng [10] (kN)

Ptest (kN)

Ptest,avg (kN)

EBR-75-1 EBR-75-2 GM-75-1 GM-75-2 EBR-100-1 EBR-100-2 GM-100-1 GM-100-2 EBR-125-1 EBR-125-2 GM-125-1 GM-125-2 EBR-150-1 EBR-150-2 GM-150-1 GM-150-2

36.5 36.5 36.5 36.5 36.5 36.5 36.5 36.5 39.1 39.1 39.1 39.1 39.1 39.1 39.1 39.1

75 75 75 75 100 100 100 100 125 125 125 125 150 150 150 150

10.03 10.03 N.A. N.A. 10.03 10.03 N.A. N.A. 10.21 10.21 N.A. N.A. 10.21 10.21 N.A. N.A.

9.80 9.50 14.42 15.03 9.95 9.89 14.95 14.52 10.07 9.35 14.60 16.10 9.12 9.42 14.81 15.34

9.65 14.73 9.92 14.74 9.71 15.35 9.27 15.08

capability of Chen and Teng’s bond strength model for evaluating FRP-to-concrete bond strength in EBR technique. An average discrepancy of 0.48 kN which corresponds to 4.7% error, is observed between Ptest and PChen & Teng in EBR strengthened specimens. Table 2 shows that when grooving method was used, the average ultimate loads increased up to 62.7%. On the other hand, all EBR specimens failed due to debonding while all GM strengthened specimens failed by FRP rupture (Fig. 3), which strongly shows the excellent performance of GM. It was observed that EBR strengthened specimens failed due to concrete failure adjacent to the adhesive-concrete interface while a thin layer of concrete was attached to the debonded FRP strip. This type of failure is assumed not to be a strict ‘‘debonding’’ as failure occurs in concrete [15]. However, the term ‘‘debonding’’ has been widely used by different researchers in this case [10]. 4.1. Load–slip behavior Load–slip diagrams for all specimens in series 2 were drawn using PIV results and are presented in Fig. 4. Image analyses were undertaken using 256  256 pixels patches generated at beginning of bond length, x = 0 (loaded end of CFRP strips). In order to obtain

Increase in ultimate load over EBR specimens (%)

Ultimate slip (mm)

Failure mode

–

0.30 0.35 0.27 0.38 0.57 0.63 0.46 0.37 0.72 0.67 0.34 0.47 0.76 0.79 0.35 0.22

Debonding Debonding Rupture Rupture Debonding Debonding Rupture Rupture Debonding Debonding Rupture Rupture Debonding Debonding Rupture Rupture

52.6 – 48.6 – 58.1 – 62.7

accurate measurements and eliminate effect of probable concrete specimens’ rigid body motion and elastic deformation of supports, displacements of both upper corners of concrete specimens were evaluated using PIV results by means of generating 256  256 pixels patches and the averaged values were subtracted from CFRP displacements obtained from PIV analyses. Fig. 4 obviously illustrates that using FRP bond length beyond effective bond length, Le, does not increase the ultimate debonding load, since all EBR strengthened specimens failed around a load of 9.5 kN. It should be noted that effective bond lengths were calculated from Eq. (3) as 71.8 and 70.6 mm for the specimens with concrete compressive strength of 36.5 and 39.1 MPa, respectively. It may be concluded from Fig. 4 that although the experimental results strongly verify the concept of effective bond length, however, longer FRP bond length increases the ultimate slip of the bonded joint and leads to a more ductile failure. As it is plotted in Fig. 4, two major zones can be considered in load–slip diagrams. The first zone as a linear part from beginning of loading process up to a load of approximately 8 kN for specimens with EBR method and 10 kN for specimens with GM. The second zone for EBR specimens is where the load is approximately constant up to debonding. This zone for the specimens strengthened by GM linearly

Fig. 3. Specimens failure: (a) debonding in specimen EBR-150-2; and (b) FRP rupture in specimen GM-150-1.

135

A. Hosseini, D. Mostofinejad / Composites: Part B 51 (2013) 130–139

18 EBR-75-2 GM-75-2 EBR-100-2 GM-100-2 EBR-125-2 GM-125-2 EBR-150-2 GM-150-2

16 14

Load (kN)

12 10 8 6 4 2 0 0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

Slip at loaded end (mm) Fig. 4. Load–slip curves for series 2 specimens.

concrete substrate when using GM, the slip values sharply decrease toward the free end of the strip. Strain profiles along bonded length of CFRP reinforcements are presented in Figs. 7 and 8 for specimens EBR-150-1 and GM-150-1, respectively. The strain values are derived from slip profiles obtained from PIV analyses for the whole length of CFRP strips; except for the start point of bond length (x = 0) which were deduced directly from applied load values. It is due to local stress concentration at this point as previously mentioned by Yao et al. [15]. As it is illustrated in Fig. 7, upon reaching Pmax, only a limited length of the CFRP strip (about 70 mm) in EBR-150-1 specimen resisted the applied load which strongly verifies the concept of effective bond length. However, after reaching Pmax, debonding rapidly propagated toward the free end of CFRP strip causing strain redistribution along the strip (Fig. 7). It may be noted here that similar strain profiles for EBR specimens were presented by Yao et al. [15], obtained from conventional strain gauges’ reading. This similarity despite minor differences between material characteristics, obviously verifies the capability of the PIV technique for deformation

continues upward with a lower stiffness compared to that of the first zone, up to FRP rupture around a load of 15 kN. 4.2. Slip and strain profiles along CFRP reinforcements Slip profiles along the bond length corresponding to different load levels for specimen EBR-150-1 are plotted in Fig. 5 using PIV analyses. Similarly, the slip profiles are presented for specimen GM-150-1 in Fig. 6. Both EBR-150-1 and GM-150-1 specimens reached a maximum slip of 0.4 mm at their maximum load level, Pmax, which were equal to 9.12 kN and 14.81 kN, respectively. After Pmax, however, the CFRP strip which was externally bonded to EBR150-1 specimen experienced higher slip values along the whole bond length under approximately constant load (Fig. 5), and finally debonded from the concrete substrate. In GM-150-1 specimen, however, only the first 90 mm of the CFRP strip experienced deformations and the slip profiles gradually increased up to FRP rupture at maximum load level, Pmax = 14.81 kN (Fig. 6). It can be concluded from Fig. 6 that due to high strength bond between CFRP strip and

0.80

0.99 P 0.70

0.95 P 0.97 P

0.60

Pmax 0.95 P

Slip (mm)

0.50

0.87 P 0.75 P

0.40

0.56 P 0.30 0.20 0.10 0.00 0

10

20

30

40

50

60

70

80

90

100

110

120

130

Distance from loaded end, x (mm) Fig. 5. Slip profiles corresponding to different load levels for specimen EBR-150-1.

140

150

136

A. Hosseini, D. Mostofinejad / Composites: Part B 51 (2013) 130–139

0.40

Pmax 0.35

0.92 P 0.83 P

0.30

0.71 P 0.59 P

Slip (mm)

0.25

0.49 P 0.20

0.36 P 0.23 P

0.15 0.10 0.05 0.00 0

10

20

30

40

50

60

70

80

90

100

110

120

130

140

150

Distance from loaded end, x (mm) Fig. 6. Slip profiles corresponding to different load levels for specimen GM-150-1.

0.9 0.99 P 0.95 P 0.97 P Pmax 0.95 P 0.87 P 0.81 P 0.75 P 0.68 P 0.56 P

0.8 0.7

Strain (%)

0.6 0.5 0.4 0.3 0.2 0.1 0.0 0

10

20

30

40

50

60

70

80

90

100

110

120

130

140

150

Distance from loaded end, x (mm) Fig. 7. Strain profiles corresponding to different load levels for specimen EBR-150–1.

1.1 Pmax 1.0

0.92 P 0.83 P

0.9

0.71 P

0.8

0.59 P

Strain (%)

0.7

0.49 P 0.36 P

0.6

0.23 P 0.5 0.4 0.3 0.2 0.1 0.0 0

10

20

30

40

50

60

70

80

90

100

110

120

130

140

Distance from loaded end, x (mm) Fig. 8. Strain profiles corresponding to different load levels for specimen GM-150-1.

150

137

A. Hosseini, D. Mostofinejad / Composites: Part B 51 (2013) 130–139

Longitudinal strain (%)

0

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

30

Distance from loaded end, x (mm)

60 90 120 150

(a) 0.56 Pmax

(b) 0.81 Pmax

(c) Pmax

(d) 0.97 Pmax

(e) 0.95 Pmax

(f) 0.99 Pmax

0 30 60 90 120 150

Fig. 9. Strain fields corresponding to different load levels for specimen EBR-150-1 (dash lines indicate 48 mm width of CFRP strip, Pmax = 9.12 kN).

measurements. On the other hand, in GM-150-1 specimen, strain profiles gradually increased up to FRP rupture and only the first 90 mm of the CFRP strip experienced strain values up to ultimate

load levels (Fig. 8). This is due to the fact that GM creates a stronger bond which transfers the shear stresses into the concrete substrate more efficiently compared with EBR technique.

Longitudinal strain (%)

0

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

30

Distance from loaded end, x (mm)

60 90

120 150

(a) 0.29 Pmax

(b) 0.54 Pmax

(c) 0.71 Pmax

(e) 0.92 Pmax

(f) Pmax

0 30

60

90

120 150

(d) 0.83 Pmax

Fig. 10. Strain fields corresponding to different load levels for specimen GM-150-1 (dash lines indicate 48 mm width of CFRP strip, Pmax = 14.81 kN).

138

A. Hosseini, D. Mostofinejad / Composites: Part B 51 (2013) 130–139

0.7 0.95 P 0.6 Pmax

Strain (%)

0.5

0.95 P

0.4

0.87 P

0.3

0.81 P

0.2

0.75 P 0.68 P

0.1

0.56 P 0.0 -20

-15

-10

-5

0

5

10

15

20

Distance from centerline of CFRP strip (mm) Fig. 11. Strain profiles along width of CFRP strip corresponding to different load levels for specimen EBR-150-1.

4.3. Strain fields As it was mentioned earlier, deformation fields in tested specimens could be monitored during test process by means of PIV technique. So in order to investigate more detailed behavior of CFRP reinforcements bonded to concrete using EBR and EBROG techniques, digital images which were taken during the tests were analyzed and strain fields were obtained for EBR-150-1 and GM-150-1 specimens. An area of 130  150 mm (width  length) was considered for each specimen and patches of 128  128 pixels, spaced at 32 pixels center-to-center were generated. Strain fields corresponding to different load levels for EBR-150-1 and GM-150-1 specimens are presented in Figs. 9 and 10, respectively. Fig. 9 obviously illustrates the propagation of debonding along the CFRP strip after Pmax, up to failure for the specimen strengthened by EBR technique. On the other hand, Fig. 10 shows the capability of GM for creating stronger FRP-to-concrete bond since larger strain values in a limited bond length could be reached. Careful inspection of Fig. 10 reveals that strain distribution along the width of CFRP strip in specimen GM-150-1 is not uniform. In order to investigate the effect of strengthening method on strain distribution along the transverse direction of CFRP sheets, strain profiles along the width of CFRP strips corresponding to different load levels are presented in Figs. 11 and 12 for specimens EBR-150-1 and GM-150-1, respectively. The strain values are derived from slip profiles at a distance equal to 15 mm from loaded

end of CFRP strip (x = 15 mm), and the PIV analyses due to slip measurements were undertaken using 128  128 pixels patches. As it is illustrated in Fig. 11, the strain distribution in specimen EBR-150-1 is more and less uniform along the width of CFRP strip, and the strain profiles gradually increase up to debonding of CFRP sheet from the concrete substrate. Note that strain profile along the width of the CFRP sheet is not expected to be symmetrically distributed due to intrinsic local differences of bond thickness and workmanship in wet layup systems. The strain distribution in specimen GM-150-1, however, is not uniform along the width of CFRP strip. As it is observed in Fig. 12, the CFRP strip edges and the area between the grooves reach higher strain values. This irregularity can be attributed to high stiffness of epoxy-filled grooves which makes the bond of FRP to concrete undeformable beneath the grooves. It may be noted that the intention of the current study is to generally investigate the efficiency of GM (EBROG technique) compared with conventional EBR method, and extra studies are definitely needed to investigate the effect of the grooves characteristics on FRP-to-concrete bond behavior in GM.

5. Conclusions Performance of grooving method (GM) in the form of EBROG as an alternative to conventional EBR technique for bonding CFRP sheets to concrete surface was evaluated in this research. Sixteen

1.1 Pmax

1.0 0.9

0.92 P

0.8

Strain (%)

0.83 P 0.7 0.6

0.71 P

0.5

0.59 P

0.4 0.49 P

0.3

0.36 P

0.2 0.1 0.0 -20

0.23 P -15

-10

-5

0

5

10

15

20

Distance from centerline of CFRP strip (mm) Fig. 12. Strain profiles along width of CFRP strip corresponding to different load levels for specimen GM-150-1.

A. Hosseini, D. Mostofinejad / Composites: Part B 51 (2013) 130–139

concrete prisms were strengthened using EBR and EBROG techniques, and the strengthened specimens were subjected to single-shear bond test. An image-based full field deformation measurement technique, i.e. particle image velocimetry (PIV) was used to investigate FRP-to-concrete bond behavior of strengthened specimens. Based on the experimental and PIV analyses results of the current study, the following conclusions can be drawn: 1. Test results are found to be in close agreement with the predictions of Chen and Teng’s bond strength model [10] for CFRP sheets adhered to prism specimens by EBR technique, since an average discrepancy of 4.7% between experimental debonding and predicted loads by Chen and Teng’s model was observed. Moreover, debonding loads of all EBR specimens were observed to be almost the same, regardless of CFRP strips bond length. 2. An average increase of 55.5% in ultimate load capacity of CFRP sheets attached to concrete prism specimens using GM was observed compared with those attached using conventional EBR technique. Furthermore, debonding failure was completely eliminated when EBROG technique was used, and all GM specimens failed due to FRP rupture regardless of grooves length. Conventional EBR strengthened specimens, however, failed due to debonding failure. 3. Obtained strain profiles and strain fields in different load levels implied that debonding rapidly propagated along the CFRP strip after reaching ultimate load level in specimens strengthened by EBR technique. However, due to high stiffness bond of FRP to concrete when GM was used, only a limited length of the bonded CFRP strip experienced strain up to FRP rupture. 4. Cutting two 5  10  75 mm (width  depth  length) grooves into the concrete surface not only obviated the need of costly and time-consuming surface preparation, but also made the bond of FRP to concrete strong enough that the FRP composite reached its full tensile capacity. However, due to limited results of the current study, further studies are surely needed in order to investigate the effect of material characteristics on bond behavior of FRP sheets attached to concrete using GM and propose optimum and effective groove characteristics. 5. Slip and strain profiles as well as the whole strain fields in strengthened specimens can be obtained utilizing non-contact accurate measurements of particle image velocimetry (PIV) technique. Consequently the PIV technique can be widely used as an alternative to conventional measuring techniques to investigate FRP-to-concrete bond behavior in experimental tests.

References [1] ACI 440.2R-08. Guide for the design and construction of externally bonded FRP systems for strengthening concrete structures. American Concrete Institute; 2008. [2] Meier U. Bridge repair with high performance composite materials. Mater Technik 1987;4:125–8. in German. [3] Rostasy FS. Bonding of steel and GFRP plates in the area of coupling joints. Talbrucke Kattenbusch. Research report No. 3126/1429, Federal Institute for Materials Testing, Braunschweig, Germany; 1987. [in German]. [4] Ohlers DJ, Moran JP. Premature failure of externally plated reinforced concrete beams. J Struct Eng 1990;116(4):978–95. [5] Teng JG, Chen JF, Smith ST, Lam L. FRP-strengthened RC structures. John Wiley & Sons; 2002. [6] Teng JG, Smith ST, Yao J, Chen JF. Intermediate crack-induced debonding in RC beams and slabs. Constr Build Mater 2003;17(6–7):447–62. [7] Aram MR, Czaderski C, Motavalli M. Debonding failure modes of flexural FRPstrengthened RC beams. Composites Part B 2008;39(5):826–41. [8] Holzenkampfer P. Ingenieurmodelle des verbundes geklebter bewehrung fur betonbauteile, TU Braunschweig; 1994. [in German].

139

[9] Neubaure U, Rostasy FS. Design aspects of concrete structures strengthened with externally bonded CFRP plates. In: Proceedings of seventh international conference on structural faults and repair, vol. 2. Edinburgh: ECS Publications; 1997. p. 109–18. [10] Chen JF, Teng JG. Anchorage strength models for FRP and steel plates bonded to concrete. J Struct Eng 2001;127(7):784–91. [11] Seracino R, Raizal Saifulnaz MR, Ohlers DJ. Generic debonding resistance of EB and NSM plate-to-concrete joints. J Compos Constr 2007;11(1):62–70. [12] Bilotta A, Ceroni F, Nigro E, Pecce M. Design by testing procedure of debonding load for RC elements strengthened with EBR FRP materials. In: Proceedings of tenth international symposium on fiber reinforced polymer reinforcement for reinforced concrete structures, vol. 275. American Concrete Institute; 2011. p. 1–20. [13] FIB. Externally bonded FRP reinforcement for RC structures. FIB Bulletin 14, International Federation for Structural Concrete; 2001. [14] CNR-DT 200. Guide for the design and construction of externally bonded FRP systems for strengthening; 2004. [15] Yao J, Teng JG, Chen JF. Experimental study on FRP-to-concrete bonded joints. Composites Part B 2005;36(2):99–113. [16] Brena SF, Bramblett RM, Wood SL, Kreger ME. Increasing the flexural capacity of RC beams using CFRP composites. ACI Struct J 2003;100(1):36–46. [17] Coronado CA, Lopez MM. Experimental characterization of concrete-epoxy interfaces. J Mater Civ Eng 2008;20(4):303–12. [18] Kotynia R, Baky HA, Neale KN, Ebead UA. Flexural strengthening of RC beams with externally bonded CFRP systems: test results and 3D nonlinear FE analysis. J Compos Constr 2008;12(2):190–201. [19] Elsayed WE, Ebead UA, Neale KW. Studies on mechanically fastened fiberreinforced polymer strengthening systems. ACI Struct J 2009;106(1):49–59. [20] Khalifa A, Alkhrdaji T, Nanni A, Lansburg S. Anchorage of surface mounted FRP reinforcement. Concr Int 1999;21(10):49–54. [21] Eshwar N, Nanni A, Ibell TJ. Performance of two anchor systems of externally bonded fiber-reinforced polymer laminates. ACI Mater J 2008;105(1):72–80. [22] Ceroni F, Pecce M. Evaluation of bond strength in concrete elements externally reinforced with CFRP sheets and anchoring devices. J Compos Constr 2010;14(5):521–30. [23] Orton SL, Jirsa JO, Bayrak O. Design considerations of carbon fiber anchors. J Compos Constr 2008;12(6):608–16. [24] Ozbakkaloglu T, Saatcioglu M. Tensile behavior of FRP anchors in concrete. J Compos Constr 2009;13(2):82–92. [25] Niemitz CW, James R, Brena SF. Experimental behavior of carbon fiberreinforced polymer (CFRP) sheets attached to concrete surfaces using CFRP anchors. J Compos Constr 2010;14(2):185–94. [26] Al-Mahaidi R, Kalfat R. Investigation into CFRP laminate anchorage systems utilizing bi-directional fabric wrap. Compos Struct 2011;93(4):1265–74. [27] Mostofinejad D, Mahmoudabadi E. Grooving as alternative method of surface preparation to postpone debonding of FRP laminates in concrete beams. J Compos Constr 2010;14(6):804–11. [28] Mostofinejad D, Shameli SM, Hosseini A. Experimental study on the effectiveness of EBROG method for flexural strengthening of RC beams. In: Proceedings of sixth international conference on FRP composites in civil engineering (CICE 2012), Rome, Italy; 2012. [29] Mostofinejad D, Tabatabaei Kashani A. Experimental study on effect of EBR and EBROG methods on debonding of FRP sheets used for shear strengthening of RC beams. Composites Part B 2013;45(1):1704–13. [30] Mostofinejad D, Shameli SM. Externally bonded reinforcement in grooves (EBRIGs) technique to postpone debonding of FRP sheets in strengthened concrete beams. Constr Build Mater 2013;38:751–8. [31] ASTM C192. Standard practice for making and curing concrete test specimens in the laboratory; 2002. [32] ASTM D3039/D3039M. Standard test method for tensile properties of polymer matrix composite materials; 2000. [33] Ali-Ahmad M, Subramaniam K, Ghosn M. Experimental investigation and fracture analysis of debonding between concrete and FRP sheets. J Eng Mech 2006;132(9):914–23. [34] Mazzotti C, Savoia M, Ferracuti B. An experimental study on delamination of FRP plates bonded to concrete. Constr Build Mater 2008;22(7):1409–21. [35] Adrian RJ. Particle imaging techniques for experimental fluid mechanics. Ann Rev Fluid Mech 1991;23:261–304. [36] White DJ, Take WA, Bolton MD. Soil deformation measurement using particle image velocimetry (PIV) and photogrammetry. Geotechnique 2003;53(7):619–31. [37] Slominski C, Niedostatkiewicz M, Tejchman J. Application of particle image velocimetry (PIV) for deformation measurement during granular silo flow. Powder Technol 2007;173:1–18. [38] Hajialilue-Bonab M, Azarnya-Shahgoli H, Sojoudi Y. Soil deformation pattern around laterally loaded piles. Int J Phys Modell Geotechnics 2011;11(3):116–25. [39] White DJ, Take WA. GeoPIV: particle image velocimetry (PIV) software for use in geotechnical testing. Cambridge University Engineering Department Technical, Report, D-SOILS-TR322; 2002. [40] Carloni C, Subramaniam KV. Direct determination of cohesive stress transfer during debonding of FRP from concrete. Compos Struct 2010;93(1):184–92.