A Markov Chain–Based Method for Crack Propagation Prediction of Welded Joints in Steel Bridges: Experimental Verification and Engineering Application

ABSTRACT To address the stochastic characteristics of fatigue damage, this study proposed a Markov chain–based crack propagation prediction model integrating linear elastic fracture mechanics. First, the state transfer process for the rib‐to‐deck weld was analyzed. Then, the state transfer probability matrix was calculated by analyzing the relationship between the stress intensity factor range and the transfer probability in Paris equation. In the process of calculating the stress intensity factor ranges, the randomness of multiple parameters needs to be considered including initial cracks. Subsequently, the evolution of fatigue states can be derived. After that, the fatigue test of the segment model was conducted to verify the correctness of the proposed model. Finally, the model was applied to evaluate the fatigue state of the rib‐to‐deck weld of the steel bridge. The results reveal that the proposed model is consistent with the experimental results and provides a probabilistic fatigue life prediction.