Abstract:The long-term performance of engineering structures in a corrosive environment will be significantly affected by the coupled action of corrosion and fatigue. In this article, a probabilistic corrosion fatigue analytical model is proposed by taking into account the effects of corrosion-induced reduction of the cross-sectional area and deterioration of the fatigue strength of structural components. The proposed model is exemplified to evaluate the probabilistic corrosion fatigue life of a typical welded joint in the suspension Tsing Ma Bridge instrumented with a long-term structural health monitoring system. A genetic algorithm–based mixture parameter estimation method is developed to facilitate the multimodal modeling of stress spectrum derived from the long-term monitoring data of dynamic strain. The achieved results demonstrate that with the increase in the service life, the reliability index of the investigated typical welded joint is dramatically reduced under the combined effect of corrosion and fatigue.

Probabilistic corrosion fatigue life assessment of a suspension bridge instrumented with long-term structural health monitoring system