Kiia Grönlund, Antti Ahola, Shahriar Afkhami, Tuomas Skriko
• Experimental fatigue tests on welded joints under variable amplitude loading. • Successful employment of a 3D-scan-based finite element model within the 4R method. • Established mesh convergence to validate scan-based finite element model accuracy. • The 4R method was able to consider the effect of varying loading conditions. • Evaluation of local approaches for fatigue assessment under variable amplitude loads. The aim of this study was to validate a local stress-based fatigue assessment approach, the 4R method, for assessing the fatigue strength of common welded ship structural details subjected to variable amplitude (VA) loads. The objective was to provide increased accuracy in fatigue strength estimations through the consideration of local elastic-plastic material behaviour, possible residual stress relaxation, and sequential effects of loading conditions. The VA load effects on fatigue strength of welded joints made of high-strength steel (690QT) were investigated by means of analytical calculations and experimental testing. Random VA load spectra were created for the fatigue tests based on different mean stress levels according to a two-parameter Weibull distribution. Additionally, a 3D scan-based solid finite element model of the longitudinal double-sided gusset joint was employed within the notch-based local fatigue assessments. The use of the scanned geometry reduced the scatter of the fatigue test results among specimens in the local approaches highlighting the importance of accurate consideration of real weld geometry in the determination of fatigue notch factors. Furthermore, the 4R method provided additional accuracy by considering the loading sequence and mean stress via mean-stress correction.