Estimation of Expected Lifetime and Reliability During Burn in and Field Operation Using Markov Chain Monte Carlo Simulations

Estimation of the mean lifetime and reliability of sophisticated systems is a challenging problem in many engineering applications. This is especially important when the study is concerned with both burn in and field operation periods since the hazard rate is no longer constant and the underlying processes are non-homogenous. In such cases, theoretical development of the solution is very tedious and obtaining results for the expected lifetime and reliability of complex systems is almost impossible. Monte Carlo simulations provide a viable alternative for estimation of the expected lifetime and reliability in such situations. Predictive calculations for the mean time to failures may be carried out using MIL-HDBK-217F for expected operating conditions of the system. However, these estimates are only valid for the field operating conditions assuming that the parts lifetime obeys an exponential probability distribution.