Fatigue Life Uncertainty Evaluation of Steel Box Girder Considering Time-Varying Model of Vehicle Load and Bridge

The failure risk of steel box girders gradually increases during service due to the accumulation of fatigue load effects and the deterioration of fatigue resistance. Reliability evaluation is an effective method for predicting the capacity of a bridge to withstand hazardous events during service. However, due to the difficulty of updating bridge models based on monitoring data, traditional fatigue load effect evaluation methods rely on fatigue loads and bridge models from the design stage, resulting in a large deviation in the fatigue reliability uncertainty quantification. To address this issue, a fatigue life uncertainty evaluation method for steel box girders considering time-varying vehicle loads and a bridge model is proposed. First, a framework for determining fatigue load effects is established, including the development of a refined stochastic traffic load model that considers spatiotemporal distribution, as well as the updating and calibration of the multiscale model based on stochastic vehicle-induced responses. Second, time-dependent reliability theory is employed to assess fatigue life uncertainty through the developed fatigue load effect model, integrating the effects of fatigue resistance degradation and nonlinear fatigue damage accumulation. Finally, the effectiveness of the proposed method is demonstrated through its application to a long-span steel box girder suspension bridge. The results demonstrate that the proposed stochastic traffic load modeling method can predict fatigue damage accurately and that fatigue performance can be dynamically evaluated through real-time calibration of the multiscale model.