Reliability modeling of mixtures of one-shot units under thermal cyclic stresses

Modeling components' thermal fatigue life due to cyclic temperature fluctuation based on Coffin-Mason principle has been extensively investigated. However, sparse research assesses the thermal fatigue life by providing the reliability metrics of components/systems under thermal fatigue. The Birnbaum-Saunders (BS) distribution is developed to model the unit's fatigue failure induced by mechanical stresses and provides the unit's reliability metrics. In this paper, we investigate a generalized Birnbaum-Saunders (GBS) distribution and its performance in predicting fatigue failure caused by thermal cyclic stresses. We then apply the GBS distribution to model the reliability metrics of a system with mixtures of nonhomogeneous one-shot units subject to thermal fatigue. An extensive simulation model is developed to validate the system reliability metrics accuracy. Numerical examples are presented to illustrate the use of the models.