Chapter 7 - Statistical Approaches in Machinery Problem Solving

Abstract

Machinery maintenance activities, together with failure analysis and troubleshooting, can be regarded as failure fighting processes. Failure fighting calls for appropriate strategies in order to be successful. However, before applying such strategies, information about failure modes and failure frequencies needs to be gathered. Suitable statistical approaches will accompany this task and are therefore part of the failure fighting arsenal. Statistical approaches as they relate to failure analysis and troubleshooting are discussed. How machinery component failure modes appear and behave as a function of time is discussed and the concept of machinery component life considered. Weibull analysis, one of the most powerful statistical failure analysis methods, and its use as part of a planned equipment replacement program is described. The WeiBayes analysis, another probabilistic model for failure analysis, is discussed. The application of the Weibull and WeiBayes methods for the reliability analysis of components is illustrated using a product reliability analysis for an advanced bearing protector seal. Different machinery maintenance strategies (preventive, predictive, breakdown, and bad actor) are discussed. The use of statistical tools to help define failure experience trends in a given machinery population is outlined.