The Experiment Analysis of Bonding Performance between Near Surface Mounted CFRP Strip and Concrete at Different Curing Temperatures

To investigate the bonding performance of near-surface mounted (NSM) carbon fiber-reinforced polymer (CFRP) technique under different curing temperatures, a total of five test specimens strengthened with NSM CFRP strip were tested by direct pull-out methods. The strain of CFRP strip, interfacial bonding shear stress, ultimate load and failure mode were analyzed through the test. The experiment results showed that the ultimate load and bonding shear stress of the specimens increased gradually when curing temperature was lower than the glass transition temperature (Tg) of the bond material. Afterward, they gradually decreased with the increase in temperature. The bonding shear stress was gradually transmitted from the loading end to the free end. The failure modes were changed from CFRP strip–adhesive interface bonding failure to CFRP strips fracture failure with the temperature rising. In this paper, the bonding performance experiment results of NSM CFRP strip could provide a reference for the rapid strengthening of monorail traffic track beams.


Introduction
Nowadays, monorail traffic has got widespread attention due to its low noise and adaptability to complex geological environment. However, when the track beam of monorail traffic suffers damages, it should be strengthened rapidly so as not to affect the normal operation of the monorail vehicle. In view of the special function and construction of the track beam, it is necessary to strengthen it with nearsurface mounted (NSM) technique, which will not change the size of the original structure. The preferred reinforcement material of NSM technique is carbon fiber-reinforced polymer (CFRP) because of its lightweight, high strength, corrosion resistance and convenient construction. In order to achieve the purpose of rapid strengthening, the most effective method is to increase the cure temperature of the test specimen so as to improve the solidification rate of the adhesive material. Therefore, the bonding performance of NSM CFRP technique is more worthy of attention when the curing temperature of adhesive materials is elevated.
So far, a large number of research results show that the reinforced concrete (RC) beam strengthened with NSM technique exhibits good flexural performance at ambient condition [1,2]. In addition, many researchers made indepth studies of bonding performance between NSM CFRP and concrete interface at ambient condition. The main parameters affecting the bonding performance are the groove size, concrete strength, bonded length, distance from the concrete edge, adhesive type, CFRP strip size and so on [3][4][5][6][7]. The bonding performance between NSM CFRP and concrete has been studied fully at ambient conditions. However, there are few research results on the effect of temperature on the bonding performance between NSM CFRP and concrete. For example, Hu [8,9] investigated the bonding performance between NSM CFRP and concrete at temperature range of 4-180°C. The results show that there was a glass transition temperature of bonding material. Glass transition temperature is a watershed that affects the performance of bonding materials. Lee [10] studied the bonding properties of three kinds of filling materials (epoxy mortar EM, anti-shrink mortar NM and epoxy putty EP) subjected to 100 times temperature cycle (-15 to 55°C). Mohammed [11] investigated the bond performance of NSM CFRP using the high-strength selfcompacting cement base with graphene oxide (IHSSC-CA) as a filling material that was subjected to 400, 600, 700 and 800°C. Yu B investigated the strength and stiffness properties of NSM CFRP through the single shear pull-out test when the specimens were subjected to 100, 200, 300 and 400°C. The results showed that the bond strength was, respectively, reduced by 80% and 100% compared to that of the specimens in room temperature when the specimens were heated to 200°C and 400°C [12]. At present, the research results of the effect of curing temperature on the bond performance of NSM CFRP were not sufficient. In this paper, the direct pull-out tests of NSM CFRP applied for the concrete block were carried out by considering the effect of curing temperature. The research contents include pull-out load, bonding shear stress, local bond-slip and failure mode. The research results would lay a foundation for using NSM CFRP in rapid reinforcement of monorail track beam.

Experimental Design
(1) The test plan To investigate the bonding performance of NSM CFRP technique applied to concrete block under different curing temperatures, a total of five direct pull-out specimens were carried out considering the effect of curing temperature. Curing temperature of the specimen was ranged from 35 to 65°C, and curing time was constant at 12 h based on consideration of ambient temperature and adhesive allowable temperature. The dimension of the concrete block was 150 mm (width) 9 150 mm (height) 9 300 mm (length). The dimension of the CFRP strip was 20 mm (width) 9 1.4 mm (thickness) 9 400 mm (length), and the bonded length of CFRP was 200 mm. The dimension of the groove was 15 mm (width) 9 30 mm (depth). The specific form of specimen is shown in Fig. 1. CFRP and epoxy were provided by Carbon Composites (Tianjin) Co., Ltd. The specific test plan is given in Table 1.
(2) Material properties The concrete grade of the specimens was C50, and concrete cube compressive strength was 50.12 MPa on 28 days.
The concrete elastic modulus was 3.45 9 10 4 MPa, and temperature expansion coefficient of the concrete was 10.2 9 10 -6°C-1 . The ultimate tensile strength of CFRP was 2000 MPa, and its elastic modulus was 140 GPa. Temperature expansion coefficient of CFRP was 0.3 9 10 -6°C-1 . The density of epoxy resin was 1.6 g/cm 3 , and its tensile strength was 40 MPa. The specific material properties are given in Table 2.
(3) Preparation process of specimen Preparation process of specimen includes four steps. First, the groove was injected with epoxy resin close to one-third of its depth. Second, the CFRP strip with strain gauge was embedded into the groove along the centerline of the groove (Fig. 2). Third, the remaining space continued to fill with epoxy. Fourth, the CFRP strip was lightly pressed and the overflow epoxy was removed with a spatula. Finally, the specimen was cured for a predetermined time at a set curing temperature, and the curing temperature was adjusted by an infrared heating lamp (Fig. 3).

Measuring Point Arrangement and Test Loading
In order to measure the strain value of CFRP during loading, five strain gauges were arranged at an unequal spacing on the CFRP strip. In order to avoid direct contact between the strain gauge and the epoxy resin, the paraffin was pasted on the strain gauge. Two linear variable differential transformers (LVDT) with a measuring range of 50 mm were arranged on the loading end and free end of the tested CFRP, respectively. Strain and displacement data were collected from the XL2118C force and strain integrated parameter tester. The specific layout position is shown in Fig. 4. After the specimen was cured for a predetermined time, it was conducted on the pull-out test immediately using hydraulic universal testing machine with a capacity of 1000 kN (shown in Fig. 5). To prevent eccentric loading, a horizontal detector was used to monitor the loading device. The graded loading system was strictly enforced during the tests in order to observe the test phenomenon. The test loading process was described as follows.
First, in order to check the stability of the test device, it is necessary to preload on CFRP with 1 kN. During preloading, the loading device and test instrument need to be carefully checked. Second, the test specimens are loaded to half of the theoretical ultimate load at every grade 1 kN and then are loaded continuously at every grade 0.5 kN until the CFRP strip was pulled out. After each grade load is loaded, the load should be stabilized for 3 min. When the test data become stable, the displacement and strain data were collected, and the experiment phenomenon is observed and recorded. Data collection machine is shown in Fig. 6. 2 Data Processing and Results Analysis 2.1 Data Processing Figure 4 shows that the free end is defined as the origin of the coordinate X. The corresponding strain is nðx i Þ at the location of x i . According to the relation between the interface bonding shear stress and strain in the literature    [13], the bonding shear stress sðxÞ of CFRP strip can be expressed: where b h is the thickness of CFRP strip; E cf is the elastic modulus of CFRP strip; x i is the location of strain gauge. The slip uðxÞ of CFRP strip can be attained: where u 0 is the displacement of CFRP strip at the free end.

Results Analysis of Bonding Performance Between NSM CFRP and Concrete
The free end of T35 specimen began to slip at the pulling load of 15 kN, and then, the CFRP strip occurred torn sound when load increased to 24 kN. Finally, the interface between adhesive and CFRP strip was destroyed due to bonding failure at 25.2 kN (shown in Fig. 7a). Because no cracks were observed on the outer surface of the specimen, the failure was speculated occurring at the interface of the CFRP-adhesive. The failure mode of T40 specimen was the interface destruction of CFRP and adhesive accompanied by adhesive cracking (shown in Fig. 7b), and the final failure load was 29.5 kN. The failure modes of T50, T60 and T65 specimens were mainly broken on CFRP strip accompanied by crack of the adhesive layer. The ultimate load was 37.1 kN, 36.3 kN and 35.2 kN, respectively.  About the experiment phenomenon of T50 specimen, a crisp sound was heard when the specimen was loaded to 18.2 kN, and the sound continued to appear at 23.9 kN. It was found that there were a crack at the loading end adhesive and precipitation of a small amount of white powder. It continued to be loaded to 35 kN accompanied by the sound of the CFRP strips torn. At this time, the load increased slowly, the loading end slip increased obviously and the free end began to slip. The precipitation of white powder increased at the interface between the adhesive and CFRP. Finally, the CFRP strip was broken and a number of fine cracks appeared at 37.1 kN.
As it is given in Table 3, first, with the increase in curing temperature, the failure modes of the specimen were changed from bond failure to the fracture of the CFRP strip. Second, when curing temperature was less than 50°C, the ultimate load increased gradually, which was increased by 47.2% from 35 to 50°C. When curing temperature exceeded 50°C, the ultimate load decreased gradually, which was reduced by 5% from 50 to 65°C. The ultimate load tended to first increase, and then decrease with the increase in curing temperature. The ultimate load reached the maximum value 37.1 kN when curing temperature was 50°C. Figure 8 shows the strain distribution of CFRP strip (specimen T40). As it is shown in the figure, when the load is lower than 24 kN, the effective transfer length of strain is 110 mm. When the specimen is loaded to 29.5 kN, the free end of CFRP begins to occur strain, which indicates the effective length of strain extends to the entire bonding area of CFRP strip. With the increase in external load, the strain of CFRP gradually develops from the loading end to the free end. Figure 9 represents the strain distribution of CFRP strip (specimen T50). Compared with the strain distribution of the test specimen T40, the strain distribution of the test specimen T50 is relatively full with the increase in load. Figure 10 shows bond-slip curve of NSM CFRP strip at the different positions (specimen T60). As it is shown in the figure, the bonding shear stress reaches a peak value 8.45 MPa at the location of x = 170 mm, and the slip value is 220.23 lm. With the increase in external load, the peak shear stress is transmitted to the free end. At the location of x = 145 mm, the peak bonding shear stress reaches 8.02 MPa, and the slip value is 809.7 lm. At the location of x = 110 mm and 65 mm, the bonding shear stress does not reach the peak value. The results show that the peak bonding shear stress is also transmitted to the free end, but its value is reduced gradually. Figure 11 shows the relationship between the external load and the bonding shear stress (specimen T65). At the location of x = 170 mm, the bonding shear stress reaches the maximum value 8.02 MPa when the specimen is loaded to 18 kN. At the location of x = 145 mm, the peak bonding shear stress reaches 7.14 MPa when the load reaches to  27 kN. As the load continues to increase, the bonding shear stress of the location of x = 65 mm and x = 110 mm also increases, but it does not reach the peak value when the specimen is damaged. From the analysis results of Figs. 8, 9 and 10, the interfacial bonding shear stress between CFRP and concrete increases gradually with the increase in the external load acting on the specimens. The peak shear stress is gradually transmitted from the loading end to the free end. Figure 12 shows the bond-slip curve of the loading end at different curing temperatures. With the increase in interface slip, the bonding shear stress first increases and then decreases. The peak bonding shear stress is related to the curing temperature. When curing temperature is 50°C,    the peak bonding shear stress reaches maximum value 9.17 MPa. In addition, with the increase in curing temperature, the change trend of bonding stiffness is basically consistent with peak shear stress. According to the above analysis, curing temperature has a great influence on the bond performance of NSM CFRP strip. When its curing temperature is lower than glass transition temperature, the bond performance increases gradually with the curing temperature elevated. When it exceeds its glass transition temperature, the bond performance will gradually weaken with the increase in curing temperature.

Conclusions
In this paper, the influence of curing temperature is considered in the bond performance tests of NSM CFRP technique carried out by single shear pull-out methods. Through the pull-out experiments and analysis of test data, following conclusions are summarized.
(1) With the increase in curing temperature, the strain of CFRP strip shows a relatively full trend with load changes. When curing temperature was less than 50°C, the ultimate load increased gradually, which was increased by 47.2% from 35 to 50°C. When curing temperature exceeded 50°C, the ultimate load decreased gradually, which was reduced by 5% from 50 to 65°C. With the increase in curing temperature, the ultimate load tends to first increase and then decrease. The failure modes were changed from CFRP strip-adhesive interface slip failure to broken CFRP strip. The bonding shear stress between CFRP and concrete gradually increases with external load, which was transmitted from the loading end to the free end. (2) With the cure temperature increaing, the peak bonding shear stress first increases and then decreases. When curing temperature is 50°C, the peak bonding shear stress reaches maximum value 9.17 MPa. In addition, the change trend of bonding stiffness is basically consistent with peak shear stress with the increase in curing temperature. Therefore, curing temperature of bonding materials has a great influence on the bond performance of NSM CFRP strip. (3) The overall research results mainly represent that the bonding performance of NSM CFRP specimens changes with the curing temperature, which can provide reference for the rapid reinforcement of monorail traffic track beams to select suitable curing temperature.