Reliability assessment for flexural FRP-Strengthened reinforced concrete beams based on Importance Sampling

Due to the uncertainties raised by model error, materials (e.g., concrete, fiber reinforced polymer (FRP) and steel) and environment (e.g., loads), research aimed at reliability analysis and risk assessment of FRP-strengthened structures is necessary. In this paper, reliability analysis of flexural FRP-strengthened reinforced concrete (RC) beams with the ultimate limit state is studied. Four failure modes are considered: concrete crushing, FRP rupture, intermediate crack-induced (IC) debonding and end debonding. First, the computational models for each failure type are assessed in terms of model error based on a comprehensive database that contains 624 FRP strengthened RC beams. Then, the stochastic information for mechanical properties of the FRP and concrete materials is further identified through a series of mechanical tests on 600 concrete compression blocks and 50 FRP tensile specimens. Subsequently, the reliability index is estimated by Importance Sampling (IS) for its satisfied accuracy and efficiency. The results show that: (1) The probability of failure is different for each failure mode. It is necessary to consider all the possible failure modes to provide a comprehensive guide in the design process. (2) Based on the partial safety factor design format, the effect of the FRP partial safety factor γf on the reliability index for concrete crushing, IC debonding and end debonding is negligible. The calibration of γf is only valid for FRP rupture in order to meet the target reliability index. (3) Based on the resistance reduction factor design format, it is effective to calibrate resistance reduction factor ϕ for concrete crushing, FRP rupture, IC debonding and end debonding. Each failure mode should be designed with corresponding ϕ. (4) Additional anchors can improve the reliability index for IC debonding and helps to transfer to other failure modes for end debonding.