TABLE OF CONTENTS Failure data from a given mixed population may represent various failure modes. The necessity of determining the life regions where these failure modes occur is apparent when it is realized that the times to failure for each mode follow a distinct statistical distribution, thus requiring individual mathematical treatment. Another reason is that each failure mode may require a different design change to improve the component's reliability.
A combined analytical-graphical approach for separating a population into its subpopulations, according to their failure mode, is presented here as an alternative to a physics-of-failure analysis. Graphical methods are used to arrive at the unified reliability, probability density, and failure rate functions of the whole population. These functions are of great interest to reliability engineers. These methods are illustrated with an example. The Kolmogorov-Smirnov (K-S) goodness-of-fit test is applied to the results as an aid in selecting the best representation of the component's life, reliability, and failure rate characteristics.
Various distributions have been used to describe different times-to-failure characteristics and the associated failure rate. A decreasing failure rate is usually encountered during the early life period of components when the substandard components fail and are removed from the population [1, p. 4]. The failure rate continues to decrease until all such substandard components fail and are removed. To this corresponds a decreasing times-to-failure distribution. The Weibull distribution with the shape parameter, ?, less than 1 is often used to depict this life characteristic.
A second type of...
