Bond-slip behavior between corroded I-shaped steel and concrete in a subsea tunnel

Abstract

Owing to the high chloride ion content in the service environment, corrosion of I-shaped steel in the primary support of subsea tunnels is inevitable; this leads to the degradation of bond performance between I-shaped steel and concrete and seriously affects the durability of subsea tunnel support systems. However, the bond-slip behavior between concrete and corroded I-shaped steel may be different from that occurring in reinforcement concrete (RC) and ordinary steel reinforcement concrete (SRC) structures for two main reasons: (1) the contact area between the I-shaped steel and concrete is larger and the shape of the I-shaped steel cross-section is more complicated than that of steel bars; (2) compared with ordinary SRC structures, those in a tunnel have no stirrup without considering shear resistance. The main objective of this study was to study the effect of corrosion on the bond-slip behavior between I-shaped steel and concrete. First, accelerated corrosion tests and push-out tests were conducted on SRC specimens and the failure mode and bond-slip curves of the specimens were investigated. The experimental results indicated that the failure mode of the experimental specimens was mainly splitting failure. Meanwhile, a three-stage bond-slip constitutive model for concrete and corroded I-shaped steel was proposed, which correlated well with the experimental curve. Finally, the degradation process of interface bond-slip caused by corrosion was analyzed via the energy principle. The dissipated energy ratio increased and the elastic energy ratio gradually decreased with increasing corrosion rate, indicating that the greater corrosion rate of I-shaped steel led to more serious internal damage. Furthermore, the damage variable was defined by the dissipated energy ratio evolution characteristics. The damage evolution models of the bond strength eigenvalues and the characteristic slip were established, which was in good agreement with the experimental data.