Predicting the fatigue survival probability of High-Strength steel Wires: A Weibull model incorporating corrosion and stress ratios

Fatigue behavior of high-strength steel wires in bridge stay cables is critical to bridge reliability, as these wires experience varying stress ratios, stress ranges, and corrosion levels that significantly impact their fatigue life. Existing models typically predict fatigue life under different stress ranges and corrosion conditions but often neglect the critical influence of stress ratio, leading to potential inaccuracies. This paper introduces an innovative R-C-S-N Weibull model that integrates stress ratio, stress range, and corrosion effects to improve fatigue life predictions. Model parameters are determined through maximum likelihood estimation (MLE) and validated using both experimental data and literature sources. The model achieves a Mean Absolute Error (MAE) of 0.10, indicating minimal deviation from observed values, while a Skewness of 0.13 reflects a balanced yet slightly conservative distribution. Additionally, the 0.4-0.6 Probability Interval Coverage of 0.6 shows that 60% of predictions fall within this central range, underscoring model stability. Results indicate that corrosion has a more pronounced impact on fatigue life at lower stress ranges, with stress ratio playing a secondary but significant role. In the model, parameter D quantifies the effect of corrosion, and the shape parameter beta governs variability. Sensitivity analysis highlights stress range and corrosion as the most influential factors in determining fatigue life. The model's reliable predictive capabilities make it valuable for assessing the remaining life of corroded steel wires in bridge structures, particularly when combined with non-destructive testing methods, providing a robust and partially conservative tool for structural health monitoring.