Chiang Mai Journal of Science

     Fatigue is the most important failure mode of welded structures in engineering applications. However, unlike steel and aluminum alloy welds, the fatigue design guideline of titanium-welded structures are still limited. In this study, Ti-6Al-4V titanium alloy welds with butt joint were produced by electron beam welding (EBW). We examined the suitability of related design codes and established an optimal fatigue-life analysis method for titanium structures. Fatigue tests with constant amplitude loading and variable amplitude loading were performed. Two mean stress correction methods usually adopted in high cycle fatigue analysis, Goodman and Gerber, were evaluated. The results showed that the fracture site of the EBW fatigue sample was located in the base metal. The EBW butt joints of the Ti-6Al-4V alloy apparently illustrated a higher fatigue strength compared to the corresponding fatigue design curve in the AWS D1.9 standard. Under constant amplitude loading with tensile mean stress, the experimental fatigue life fell between the Gerber and Goodman curves. The Miner rule combined with the Goodman mean stress correction approach was recommended to obtain a better fatigue life prediction in the case of variable amplitude loading with tensile or slightly compressive mean stress. All predicted fatigue lives fell within a three-fold change boundary for both SAE transmission and bracket histories.