Assessing the reliability of components with micro- and nano-structures when they are part a multi-scale system

Abstract

Products with macro- and meso-scales components have grown into multi-scale systems where some of their components have shrunk to a smaller scale such as micro- and nano-structures. In practice, engineers may be unable to accurately assess their designs to improve the reliability of components in multi-scale systems, because component parts with micro- and nano-structures cannot be realized directly in system design phase. Thus, all kinds of design processes such as engineering verifications for physics or functions become increasingly complex in a muti-scale system. For this shortcoming, we present a parametric Bayesian method that enables engineers to assess indirectly the reliability and the quality of a component with either a micro-structure or a nano-structure using experimental data on failure time of the multi-scale system. Our proposed Bayesian approach is flexible in allowing a general model for distributions of failure times of a multi-scale system and its components. Our method is applied to a simulated data set.