Application of quantitative accelerated life models on load sharing redundancy

In most cases, when analyzing redundancy, independence is assumed across the components within the system. In other words, it is assumed that the failure of a component does not affect the failure distributions of the remaining components. However, if a system consists of multiple components sharing a load then the assumption of independence no longer holds true. If one component fails then the component(s) that are still operating have to assume the tailed unit's load. Therefore, the reliabilities of the surviving unit(s) change. Calculating the system reliability is no longer an easy proposition. This paper explores the concept of component dependence while using accelerated life test data analysis models to determine system reliability. The cumulative damage model is used to analyze life data obtained from load sharing components, and solve for the parameters of the life-stress model. The life-stress model then used as an input to the system's reliability equation, therefore providing a comprehensive model for performing inferences on the life of load sharing components based on their operating stress. Utilizing these models provides better understanding of the behavior of the system in the field, and allows for more intelligent predictions.