OPPORTUNISTIC AND CONDITION BASED MAINTENANCE MODELLING FOR AVAILABILITY ANALYSIS OF REPAIRABLE MECHANICAL SYSTEM

Abstract

This project deals with opportunistic and condition based maintenance modeling for availability analysis of repairable mechanical systems using Markov approach. The conventional techniques such as reliability block diagram, fault tree analysis and reliability graphs are no more applicable when repairs and other dependencies are incorporated in the model. Therefore, the Markov approach is selected since it is capable of modeling dependencies. Most of the mechanical systems deteriorate gradually before they fail catastrophically. Availability modeling with binary state doesn’t give realistic results. So, it would be more appropriate if multi state degradation is considered. Opportunistic Maintenance models are developed with corrective maintenance, combined with condition based opportunistic maintenance. Models with three types of repair such as Perfect, Imperfect and Minimal are developed with and without opportunistic Maintenance. The Markov based condition based model is also developed for availability analysis. Aspects such as multi state degradation, random failures, periodic condition monitoring and repair actions such as ‘no repair’, ‘minimal repair’, ’perfect repair’ and ‘imperfect repair’ are considered for modeling. The solutions of the models are obtained analytically by solving system of ordinary differential equations by Ranga-Kutta method using MATLAB software and validated by Monte Carlo Simulation. The proposed methodology is demonstrated for repairable mechanical systems. The benefits of opportunistic maintenance are quantified in terms of the increased system availability. In condition based maintenance model the condition monitoring interval is determined for maximizing the system availability. The proposed methodology is helpful for maintenance engineers in deciding suitable maintenance and replacement policies.