Improving and optimizing of imperfect preventive maintenance policy for CNC cutting machine

In thesis study, a sequential preventive imperfect maintenance model that minimizes the long-term operational cost per unit time is improved for random degrading machining center in defence and aerospace industry. Parameters used for the proposed Markovian maintenance algorithm are gathered from real life 2013- 2016 data sets which are provided from Turkish Aerospace Industries, Inc.(TAI). In contrast the traditional maintenance plans in literature and the current application status under process, a reliability-centred preventive maintenance policy is developed by monitoring continuously and collecting instantaneous information about the system's current condition. Whenever system's reliability reaches the threshold R, imperfect repair is performed to restore the system and enhance system condition in aim to maximize the average availability, reduce the number of possible breakdowns during the system lifetime and accomplish the goal of quality assurance. Besides, the combination system's failure rate increase factor and age reduction factor is taken in consideration as variant according to system usage and age in order to build hybrid hazard model. In improved approach, as planned maintenance cost, minimal repair cost and system operating cost changes according to time and maintenance cycle, breakdown cost are assumed as constant. In this preventive maintenance algorithm for complex equipment, the optimal system's reliability and number of maintenance cycle number, the interval time are reached under the criteria of the long term cost minimization of the system. Additionally, sub-systems are analyzed in scope of Bayesian theorem and it is determined which sub-systems are repaired at each maintenance cycle.