PREDICTING FAILURE MODES TO IMPROVE RELIABILITY

This paper describes the evolutionary process leading to the conclusions that for equipment experiencing a constant failure rate: the ability to provide failure-free production for a given time, t, results in greater customer satisfaction than providing a system with a high mean-time-between failure that a powerful tool in providing failure-free production for a given time, t, is the capability to predict failure modes In fact, major efforts to increase the mean-time-between failure while equipment is in a constant failure rate mode can result in increased customer frustration. The main task in improving customer satisfaction is identifying the real customer requirements and then providing for those requirements. After many discussions and joint work sessions with customers, a few customers have identified their requirements in terms of uninterrupted operation for a given period of time. After the given period of time preventive maintenance is performed and the process is repeated. The methods required to meet this requirement include: operator training, raw material control, properly trained maintenance engineers, and a method to predict future failure modes. The road that led to the need to predict future failure modes and how this contrasts with an effort to increase the mean-time-between failure is the topic of the paper. Additional benefits to be gained from the Data Acquisition System used to predict failures is provided in the Appendix.