Probabilistic Models for Creep-Fatigue in a Steel Alloy
Publication or External Link
In high temperature components subjected to long term cyclic operation, simultaneous creep and fatigue damage occur. A new methodology for creep-fatigue life assessment has been adopted without the need to separate creep and fatigue damage or expended life. Probabilistic models, described by hold times in tension and total strain range at temperature, have been derived based on the creep rupture behavior of a steel alloy. These models have been validated with the observed creep-fatigue life of the material with a scatter band close to a factor of 2. Uncertainties of the creep-fatigue model parameters have been estimated with WinBUGS which is an open source Bayesian analysis software tool that uses Markov Chain Monte Carlo method to fit statistical models. Secondly, creep deformation in stress relaxation data has been analyzed. Well performing creep equations have been validated with the observed data. The creep model with the highest goodness of fit among the validated models has been used to estimate probability of exceedance at 0.6% strain level for the steel alloy.