Simulation-based fatigue assessment using the 4R method in different load conditions

Abstract

This study examines the use of simulation-based model for the fatigue assessment that combines computational weld mechanics (CWM) and mechanical analysis to obtain local fatigue-effective stresses. The applied CWM analysis includes sequentially coupled heat source modeling and mechanical analysis to predict local material properties, residual stresses, and welding deformations. Subsequentially, fatigue assessments are conducted based on the material cyclic simulations considering different load conditions, welding-induced residual stresses and estimated material condition changes at the heat-affected zone (HAZ). The current work evaluates the fatigue strength of a longitudinal gusset joint subjected to different load conditions: constant amplitude, variable amplitude, and quasi-statically overloaded constant amplitude loads. The computational fatigue assessments are performed using a multiparametric fatigue assessment approach, namely the 4R method, and computational results are verified with the experimental fatigue tests. The simulated welding-induced residual stresses and material properties at the HAZ showed reasonable agreement with the experimental results. In addition, the S–N master curve fitted to the experimental fatigue data and simulation stress results shows a relatively small scatter range index, equal to T_σ,sim = 1.25 considering different load conditions. The model sensitivity study to the estimated material properties showed that the cyclic strength coefficient is the most influential factor in fatigue life estimation.