Abstract
Two novel damage accumulation rules are proposed for fatigue life prediction of welded details in long-span aluminum bridges, addressing the limitations of existing models, including Miner's rule. These models are developed based on continuum damage mechanics and the sequential law concept combined with the strain energy density, to improve the accuracy under random variable amplitude (VA) loading. They utilize common S-N curves and material parameters. The models are verified through fatigue tests on welded joints from a long-span aluminum bridge under random VA loading, including friction stir welded (FSW) joints, for which no standardized S-N curves are available. Fatigue lives of specimens under random VA loading are predicted using established cumulative damage models and compared with the proposed models. The results demonstrate the improved accuracy of the proposed models under random VA loadings, offering a more reliable approach for fatigue life assessment of welded joints in aluminum bridges.