Bayesian analysis for estimating statistical parameter distributions of elasto-viscoplastic material models

High temperature design methods rely on constitutive models for inelastic deformation and failure typically calibrated against the mean of experimental data without considering associated scatter. Variability may arise from acquisition process, heat-to-heat material property variations, or both need to be accurately captured predict parameter bounds leading efficient component design. Applying Bayesian Markov Chain Monte Carlo (MCMC) method produce statistical capturing underlying uncertainty in is an area ongoing research interest. This work varies aspects MCMC explores their effect posterior distributions a uniaxial elasto-viscoplastic damage model using synthetically generated reference data. From our analysis with we determine that informed prior distribution including different types test conditions results more accurate distributions. The distributions, however, do not improve when increasing number similar Additionally, changing amount scatter affects quality especially less sensitive parameters. Moreover, perform sensitivity study parameters likelihood function analysis. help reliability reduce dimensionality problem by fixing insensitive comprehensive described this demonstrates how efficiently apply methodology capture uncertainties high models. Quantifying these will engineering practices lead safer, effective designs.