2.1 Analysis of failure mode and failure mechanism
CFRP-concrete interface failure usually occurs in the interaction zone between concrete and binder resin near the concrete side[10]. The interaction zone between the concrete base layer and the binder resin can be subdivided into three layers of the permeation layer (I), the weak layer (II) and the hard layer (III), as shown in Fig. 2. The area from the binder resin to the depth of 1–2 mm of the concrete is a permeation layer of the binder resin, and the force of this region is mainly the adhesion force and the mechanical bite force. Deep into the concrete 3-5mm is a weak layer, mainly composed of cement slurry and fine aggregate. Further deep is the hard zone of concrete with high strength. With the prolonged erosion time of the sulfate dry-wet cycle and the change of the CFRP bond length, the failure mode of the test piece also changed.
Since the shear strength of the binder resin is much larger than the shear strength of the concrete, the damage of the uneroded specimen occurs in the II region. CFRP has a large amount of concrete debris and particles pulled down. The concrete surface is uneven and produces a triangular shear zone at the loading end. The shorter the bond length, the larger the range of the triangular shear zone. The shear failure surface starts from 2 to 5 cm from the loading end and is at an angle of about 45 degrees to the CFRP axis. The triangular concrete block that is pulled out is better bonded to CFRP (Fig. 3(a), (c)). It can be seen from the test results that the concrete with a small bond length is more severely damaged at the free end, as shown in Fig. 3(a), which is caused by the concrete reinforcement at the free end.
At the beginning of the sulfate dry-wet cycle, the hydration product of the cement reacts with the sulfate to form ettringite and gypsum with swelling[22–23], and these swell products and salt crystals continuously accumulate in the pores of the concrete. However, at this time, the expansion force of the salt crystals is less than the tensile strength of the concrete, which will not cause cracking of the concrete, and the bonding interface will not be damaged. The failure morphology is basically the same as that of the specimen not subjected to sulfate attack; As the erosion time increases, most of the specimen damage still occurs in Zone II, but the probability of a triangular shear zone appearing at the loading end is reduced, and the concrete debris and particles under CFRP become thinner. As the sulfate attack time increases further, the triangular shear zone no longer appears at the loading end, and the concrete layer adhered to the CFRP becomes thinner and thinner. When the erosion time reached 150 times, only small concrete particles were left on the CFRP, but the loose concrete particles that were pulled down increased, and the free end of the test piece with a short bond length did not appear to be pulled down, as shown in Fig. 3(b) and 3(d).
2.2 Ultimate bearing capacity
Figure 4 shows the ultimate bearing capacity of CFRP-concrete interface as a function of CFRP bond length. It can be seen from the figure that the ultimate bearing capacity increases with the increase of CFRP bonding length, but the ultimate bearing capacity does not increase linearly with the increase of bonding length. When the bonding length is short, the ultimate bearing capacity increases with the increase of the bonding length; when the bonding length exceeds a certain value, the ultimate bearing capacity increases with the increase of the bonding length. The reason is that when the bonding length is less than the effective bonding length, the increase of the bonding length directly increases the force transmission area of the bonding interface, and the ultimate bearing capacity increases greatly. When the bonding length is greater than the effective bonding length, the increase of the bonding length does not increase the effective force transmission area of the interface, and the improvement of the bearing capacity is mainly provided by the mechanical biting force and frictional force of the interface after peeling.
Figure 5 shows the ultimate bearing capacity of double-shear specimens as a function of erosion time for different CFRP bond lengths. It can be seen from the figure that the test pieces with the bonding length of 60 mm and 80 mm have a significant decrease with the increase of the sustained immersion time of the sulfate, and the decrease is more than that of the test piece with the bonding length exceeding 120 mm. The bearing capacity of the test pieces with the bonding length of 120mm, 150mm and 180mm is basically the same as the erosion time. The reason is that with the extension of erosion time, the effective interfacial bond length increases constantly, while the bond length of specimens with the lengths of 60 and 80 has been smaller than the effective bond length after sulfate erosion, leading to an increase in the reduction of ultimate bearing capacity.
2.3 Load-slip curve
Figure 6 is a corresponding load-slip (P-s) curve for the number of times of dry and wet sulphate cycles of 0, 30, 60, 90, 120, and 150 cycles. It can be seen from Fig. 6(b) that at the initial stage of loading, the curve is approximately linear, and at this time, CFRP and the concrete are subjected to force, and the interface is in an elastic state. As the load increases, cracks appear in the concrete at the interface, and the curve gradually becomes nonlinear. When a certain load is reached, the bonding surface begins to peel off, and the curve shows a significant inflection point. The load at this time is the peeling load. Since then, the amount of slip has increased rapidly and the load has remained essentially unchanged. With the increase of the number of dry and wet cycles of sulfate, the peeling load, ultimate load, ultimate end slip and interface stiffness (slope of the linear part) of the test piece showed a decreasing trend. The ultimate load carrying capacity of the test piece decreased by about 35% after 150 cycles of sulfate dry and wet cycles. It is indicated that the dry and wet cycle of sulfate causes damage to the bonding performance of the interface of the test piece.
It can be seen from Fig. 6(a) that when the CFRP bonding length is short, the peeling load, ultimate load, ultimate end slip and interface stiffness of the specimen are also reduced with the increase of the number of dry and wet cycles of sulfate. trend. However, the P-s curve does not appear as the horizontal section in Fig. 6(b), only the ascending section, which is due to the short bonding length of CFRP, and the interface is quickly peeled off after the specimen reaches the ultimate load.
2.4 Interfacial strain distribution
The CFRP strain distribution under different loads can be obtained by the strain gauge attached to the CFRP surface. The strain \(\varepsilon\) in the loading process varies with the distance x from the loading end as shown in Fig. 7.
For specimens with a relatively long bond length (Fig. 7(a) and Fig. 7(b)), at the initial stage of loading, the strain mainly exists near the loading end, and the strain is zero far from the loading end. As the load increases, the strain value increases and gradually transfers to the free end. After reaching the shed load, load CFRP end start paring, reach maximum strain loading end, with the increase of load in fluctuation near the loading end strain in scope, the curve started to level, strip surface, which is moving towards the free end until the interface stripping damage, but the specimen until destruction, a distance near the free end strain value is small.
For specimens with a short bond length (Fig. 7(c) and Fig. 7(d)), the development of CFRP strain does not occur in the horizontal section, and the interface peels off quickly when the load reaches the peeling load. At the beginning of the load, the free end is producing strain, the strain values at that point and slightly above the middle period of strain values, this is due to the smaller bond length of CFRP, the load is small interface will be the overall force, the strengthening effect of the free end interface makes it load more hours free end strain value is slightly larger than middle area, but eventually destroyed free end also far less than the maximum strain value.
Figure 8 shows the distribution of CFRP surface strain when the time of sulfate dry-wet cycle is different. It can be seen from the figure that the maximum strain of the CFRP surface gradually decreases with the increase of erosion time, and the slope of the slope of the curve also decreases with the increase of erosion time. It indicates that the dry and wet cycle of sulfate degrades the bond performance of CFRP-concrete interface. At the same time, as the erosion time increases, the bond length required for large strain at the free end increases. When it is not eroded, only the test pieces with a bond length of 60 mm and 80 mm have a large strain at the free end. After 150 days of sulfuric acid dry-wet cycle, the test piece with a bond length of 120 mm began to show relatively large strain near the free end.