A simplified probabilistic fatigue-life model for FRP reinforcing bars using a normal distribution approach

Fibre-reinforced polymer (FRP) reinforcing bars are increasingly used in structural engineering applications due to their high tensile strength, corrosion resistance, and favourable durability characteristics. Despite these advantages, the fatigue behaviour of FRP bars remains a critical design concern, as fatigue test results exhibit significant scatter that must be addressed through rational and reliable modelling approaches. Probabilistic fatigue-life assessment of FRP materials is commonly conducted using the wear-out model originally proposed by Sendeckyj, in which fatigue failure is described as a progressive degradation of residual strength. This framework is typically combined with a Weibull distribution to account for statistical variability in material properties. While such Weibull-based formulations have demonstrated good predictive capability, their practical implementation often involves iterative procedures and calibration of distribution parameters. In the present study, the original Sendeckyj wear-out model serves as the base framework, and a simplified probabilistic formulation is proposed to describe the variability of the effective static strength using a normal distribution. Unlike conventional approaches that directly fit fatigue life data, the proposed method uses a normal distribution on transformed strength parameters derived from the wear-out model, while preserving the deterministic fatigue degradation mechanism originally introduced by Sendeckyj. The proposed approach is verified using fatigue datasets for glass, carbon, and basalt FRP reinforcing bars collected from the literature. The results demonstrate that the normal distribution–based formulation provides reliability estimates comparable to those obtained using Weibull-based Sendeckyj models, while significantly reducing the computational effort required to determine target reliability levels. The proposed method, therefore, offers a practical and computationally efficient alternative for probabilistic fatigue assessment of FRP reinforcing bars in structural engineering applications.