Uncertainty Quantification of Fatigue Life Prediction in Welded Structures Using Microstructure-Based Simulations

Takayuki Shiraiwa,Fabien Briffod,Manabu Enoki

DOI: https://doi.org/10.1007/978-3-319-70365-7_38

2017-11-19

Abstract:A framework for uncertainty quantification of fatigue life prediction in welded structures has been developed based on multi-scale finite element analysis (FEA) considering microstructure. The multi-scale FEA consists of the following procedures: (i) mechanical and thermal properties are estimated by using commercially available software and database; (ii) temperature field, residual stress and distortion generated during a welding process is calculated on the global model by thermo-mechanical FEA; (iii) macroscopic stress field under cyclic loading condition is calculated with a hardening constitutive model; (iv) the mesoscopic stress field is derived by crystal plasticity finite element analysis with sub-model technique and the cycles for a fatigue crack initiation is analyzed by strain energy accumulation on the slip plane; (v) the cycles for fatigue crack propagation is analyzed by extended finite element method and the total number of cycles to the failure is obtained. The possible uncertainties in the current model include physical variability (weld shape, residual stress, loading and microstructure), data uncertainty due to measurement error and modeling uncertainty in numerical approximations and finite element discretization. A sensitivity analysis was performed by changing the input parameters in the proposed simulation. Additionally, fatigue database of weld joint were aggregated from public databases, published papers and academic resources, and compared with simulation results to quantify the contribution of each source of uncertainty. The effects of residual stress, toe radius and microstructure on the fatigue life of welded joints were evaluated by the proposed methodology. It was shown that the fatigue life increased with the toe radius, and the scattering in high cycle region appeared due to considering microstructure. The features and limitations of our methods will be discussed.