A Helicopter Machine MTTR Simulation Study

With the continuous improvement of its technical performance, the structure of the helicopter equipment is more and more complex, which requires the increase of its maintainability. For a long time, MTTR (Mean Time to Repair) is a common index to measure the maintainability of equipments. The way of the verification of MTTR in engineering mainly is according to the three verification schemes given by GJB2072-1994"Maintenance test and evaluation" as1-A,1-B,9of appendix A, and each of the three verification schemes require the amount of samples of LRU (Line Replaceable Module) should be more than30. As complex equipment, a helicopter may have about300LRUs. The total amount of maintenance work samples needed for the overall unit MTTR verification of helicopter will be very large, so it’s difficult to implement in the practical engineering application. Moreover, the acquisition of samples will extend the verification period, thereby affecting the final shaping of the equipment.To solve these problems, this dissertation designs and implements a simulation method of the overall unit verification of helicopter to reduce the required sample size of each LRU. For the actual situation of verification work, as failure rate of each LRU known or unknown, it collects the natural fault repair work samples and virtual fault repair work samples in a time period, and these samples are fitted for distribution in the simulation platform to get the distributions and parameters of corrective maintenance time. Then, it takes samples according to the proportion of different natural fault repair work samples of LRU on the distributions, and calculates in accordance with the definition of MTTR value. Finally, it obtains the mean of these simulation results to reduce the random errors, achieving the overall unit MTTR verification of helicopter.Under the premise of ensuring the accuracy of the verification, this dissertation reduces the sample size of each LRU by fitting the distribution of a small amount of repair work samples, achieving the goal of reducing the total samples. Finally, it verifies the validity of the model by simulating with the actual data, and provides a new method to solve the problem.