|
Almost all systems of interest in reliability applications are designed to be repaired, rather than discarded, after their first failure. Nevertheless, most reliability texts overemphasize nonrepairable items (henceforth, "parts"); if repairable systems (henceforth, "systems") are addressed, they usually are assumed to be same-as-new after repair. Such renewal by repair is neither plausible, nor mathematically tractable, nor even desirable since reliability growth is sought. Moreover, even with the utmost care in distinguishing between parts and systems, failure-data-sets for parts and systems look similar, their mathematical models look similar and even fundamentally different analysis results often look similar. Most reliability texts are impeccably rigorous when addressing parts but, unfortunately, many become extremely sloppy when treating systems - which require much more rigor! All these interacting factors have caused widespread misconceptions about even basic systems' reliability concepts. For example, what could be simpler than the idea that a system's reliability is improving if it fails less often with increasing operating time? In general, there is no connection between this concept and decreasing "failure rate" since the blatant misnomer "failure rate" almost always is defined as a property of a part's distribution of time-to-failure. Moreover, even under the definition for parts, increasing "failure rate" does not imply a monotonically increasing average number of part failures per unit time. This course presents basic concepts and models for parts and systems and stresses their up to infinite differences, rather than their superficially striking but relatively unimportant similarities.
|
Cart
