Reliability and Optimal Replacement via Coefficient of Variation

The paper presents an approach to obtaining reliability and optimal replacement policies at different operating conditions, based on limited data at one operating condition. Guidelines for the choice of a distribution based on the behavior of sample functions of deterioration processes such as wear and fatigue are provided. The importance of the coefficient of variation is delineated to show how it provides a dimensionless format of the distributions and optimal replacement strategies. Simple graphical solutions of the otherwise complicated optimal replacement equations are illustrated by numerical examples based on experimental data in diverse forms from the areas of cutting tools and fatigue failures.