Reliability analysis for systems subject to degradation and shocks

Engineering systems usually deteriorate due to some underlying degradation processes and possibly random shocks. Accurate modeling of the effect of these processes on the system leads to better system reliability estimation. This dissertation investigates the impact of degradation processes and random shocks on systems and studies the optimal maintenance policies for such systems. We develop new reliability models which offer more realistic system reliability estimation. Model 1 considers the effect of system’s age on shock damage magnitudes. This is due to the fact that systems become more vulnerable to shock damage magnitudes as it ages. Model 2 extends Model 1 by considering the correlation between degradation processes and accumulated shock damages. Both models assume that the shock damage magnitudes are s-independent and time-variant. For each model, the system reliability expression is developed and the associated parameter estimation method is presented. Numerical analyses and Monte Carlo simulation are conducted on different parameters of reliability expression to validate the analytical reliability expressions. Furthermore, we develop a generalized threshold-type condition-based maintenance (CBM) policy for a system subject to multiple competing risks including degradation process and sudden failure where the maintenance is considered to be imperfect. The objective of this study is to obtain the optimal preventive maintenance threshold maximizing system’s average availability. The model can also accommodate the correlation among multiple failure modes. The special case of such a system subject to two independent competing risks, degradation and sudden failure is studied where the degradation process is described according to Model 2. Finally, numerical optimization analyses and sensitivity analyses on optimum policy are conducted and presented accordingly.