A quantitative weld sizing criterion for fatigue design of load-carrying fillet-welded connections

Abstract Fillet welded connections are most commonly used for construction of engineering structures that are often subjected to various forms of cyclic loading in service. There exist two potential fatigue failure modes, i.e., weld root cracking through weld throat and weld toe cracking into plate thickness. The former should be prevented at design stage through a proper weld sizing. However, due to difficulties in determining stress concentration at weld throat, existing fatigue design rules, particularly on weld sizing, are largely empirical and often result in excessive conservatisms, leading to oversized welds. As structural light weighting becomes increasingly important, there is an increasing demand for more quantitative fatigue-based fillet weld sizing criterion. By taking advantage of a comprehensive set of fatigue test results on load-carrying fillet-welded cruciform joints and associated analytical developments reported by Xing et al. (2016), an analytical weld throat stress model taking into account of weld penetration and joint misalignments is presented in this paper for determining critical weld size beyond which weld throat failure become unlikely. Then, it is shown that a weld sizing criterion can be analytically developed and expressed as a function of weld penetration and joint misalignments, which are shown to agree well with the large amount of test data reported by Xing et al. (2016), including a systematic validation by means of a logistic regression method on the same test data. As a result, the present developments provide both analytical and experimental basis for achieving quantitative weld sizing so that weld throat fatigue failure mode can be effectively prevented with a clearly defined confidence level without the need of over-sizing.