Resumen de Bayesian Modeling of Accelerated Life Tests with Random Effects

Ramón V. León, Ravichandran Ramachandran, Avery J. Ashby, Jayanth Thyagarajan

• Bayesian analysis using Markov chain Monte Carlo methods is used to make inferences in accelerated life tests in which test units are from different batches and the group effect is random and significant. The approach accommodates multiple random effects and several accelerating factors. Pressure vessels wrapped in Kevlar 49 from a single spool with effects that are random illustrate application of the method. Implications that this analysis has on estimates of the reliability and safety of pressure vessels used in the Space Shuttle are discussed.