Abstract This study proposed fatigue failure criteria and investigated its effectiveness to both single-sided and double-sided gusset joints through three different fatigue assessment methods: the first one is the nominal stress method; the second method is the hot spot stress method, which requires stress calculation using FEA; and the third method is the master S-N curve using the equilibrium-based structural stress method adopted by ASME BPVC VIII-2 and API 579-1/ASME FFS-1, which requires structural stress calculation using FEA. The failure criteria for welded in-plane gusset joints were investigated based on loading and geometry conditions. When the proposed failure criteria were investigated using the equilibrium-based structural stress method, all fatigue data sets were consolidated within the master S-N curve, and the design S-N curve was appropriately conservative. However, nominal and/or hot spot stress-based approaches showed inconsistent trends compared to the fatigue data sets. Some applications can use rigorous failure criteria, such as finite crack sizes, rather than complete failures. The modified equilibrium-based structural stress procedure was proposed for partial failure cases to consider the rigorous failure criteria. Once the modified procedure was applied, the partial failure data sets were mingled with the complete failure data sets. Therefore, the design S-N curve shows excellent conservatism regardless of complete or partial failures. This study shows that when the failure criteria for welded in-plane gussets were applied, the master S-N curve approach using the equilibrium-based structural stress method exhibited excellent data collapse and proper conservatism regardless of the geometry conditions and failure criteria.
Hong et al. (Sun,) studied this question.