Probabilistic Life Prediction for High Temperature Low Cycle Fatigue Using Energy-Based Damage Parameter and Accounting for Model Uncertainty

Probabilistic methods have been widely used to account for uncertainty from various sources to predict fatigue life for components or materials. The Bayesian approach can potentially give more accurate estimates by combining test data with technical knowledge available from theoretical analyses and/or previous experimental results. The aim of the present paper is to develop a probabilistic methodology for high temperature low cycle fatigue life prediction using an energy-based damage parameter and to demonstrate the use of an efficient probabilistic method. Accordingly, a Black-box approach is used to quantify model uncertainty for three damage parameters (the generalized damage parameter, SWT and plastic strain energy density (PSED)) using measured differences between experimental data and model predictions. The proposed model was verified using experimental data for nickel-base Superalloy GH4133 under different temperatures from literature. The results show that the uncertainty bounds using the generalized damage parameter for life prediction are tighter than that of SWT and PSED methods, which leads to better decision making based on the same available knowledge.