Fatigue tests and numerical analyses were carried out in order to investigate the effect of hammer peening on fatigue life of U-rib structure. The fatigue test results showed that the fatigue life of peened specimens are longer than the non-peened ones. In addition, the numerical results showed that compressive residual stresses were induced at weld toe and root by hammer peening, moreover that the accumulated plastic strains during cyclic loading can explain the experimental results in terms of fatigue life extension.

The existing methods of assessing the fatigue life of welded joints fail to consider local strain ranges and mean stress at the weld toe. The present work proposes a novel approach to assessing the fatigue life of welded joints by conducting measurements with digital image correlation (DIC) and X-ray diffraction (XRD) in combination. Local strain ranges at the weld toe of gusset welded joints were measured by DIC. Hammer peening was conducted on the welded joints to introduce different initial stresses. The influence of mean stress was investigated by considering initial residual stress measured by XRD and a perfect plastic material model. The fatigue experiment was carried out on specimens with and without hammer peening. The results showed that hammer peening could offset adverse welding deformation effectively, and introduce significant residual compressive stress. The fatigue failure life increased by more than 15 times due to hammer peening. The fatigue initiation life assessed by the proposed method was close to that based on nominal stress, indicating that the proposed method is reliable for predicting the fatigue initiation life of welded joints.