Failure Distribution of Repairable Units Approximation through Markov Chains

Failure distributions of components and assemblies are commonly approximated by analytical functions, e.g. by exponential or Weibull distributions. The advantages of this approach are its simplicity and the staightforwardness of the calculations. Another possibility is to describe failure distributions by means of Markov or semiMarkov chains, however several constraints are thereby introduced. The present paper deals with the use of Markov chains as an approximation of failure distributions of repairable units and assemblies due to ageing, enlighting the main constraints, advantages and disadvantages as well as proposing some guidelines for the use of such approaches.