• Publisher-DOI

Fatigue life evaluation of welded joints under multiaxial loading usually refers to stresses normal to the weld and shear stresses. Stresses parallel to the weld are not considered in most experiments or the well-known Gough-Pollard criterion. Hence, the Gough-Pollard criterion has recently been extended to include all stress components at the weld surface. In this paper, both the original and, for the first time, the extended Gough-Pollard criterion are applied to different welded specimens under multiaxial loading that includes stresses parallel to the weld. As shown, the original criterion is insufficient to evaluate such stress states. This is because the calculated fatigue life becomes less conservative as the stresses parallel to the weld become more significant. The extended criterion, on the other hand, shows greatly improved accuracy while significantly reducing the likelihood of non-conservative results. In conclusion, the extended Gough-Pollard criterion can describe fatigue life under multiaxial loading better than the original version and provides reliable and conservative results for welded joints. The main findings are valid for the nominal, the hot spot, and the notch stress concept.