Reliability analysis for multi-component systems with interdependent competing failure processes

Abstract

In this paper, two reliability models for multi-component systems suffering from dependent competing failure processes is studied. All components in the systems are subject to two types of failures, i.e., soft failure owing to internal degradation and hard failure due to external shocks. A soft failure happens when the total degradation amount of a component exceeds its soft failure threshold, and a hard failure of a component occurs when a single shock load or the accumulated shock load exceeds its hard failure threshold. A component fails when a soft failure or a hard failure happens, whichever occurs first. The external shock arriving according to a non-homogeneous Poisson process has impact on all components of the systems simultaneously. The soft failure and hard failure processes are dependent because the external shock can not only break down components stochastically, but also accelerate the degradation process of the components. By using the multivariate analysis technique, the analytic forms of the reliability functions for systems under cumulative shock model and extreme shock model are derived. The random point method in Monte Carlo simulation and vector program in Matlab are applied to calculate multiple integral in the expressions of the reliability functions. Finally, a case study on the Micro-Electro Mechanical Systems with multiple different components is provided to illustrate the proposed model.