Accession Number : ADA134470

Title :   Optimal Maintenance Policies: A Graphical Analysis.

Descriptive Note : Master's thesis,

Corporate Author : AIR FORCE INST OF TECH WRIGHT-PATTERSON AFB OH SCHOOL OF SYSTEMS AND LOGISTICS

Personal Author(s) : Doumit,Patrick F ; Pearce,Barbara A

PDF Url : ADA134470

Report Date : Sep 1983

Pagination or Media Count : 115

Abstract : Probability maintenance models can be categorized according to three types of uncertainty: (1) uncertainty regarding when the next failure will occur is present for all stochastically failing components; (2) uncertainty regarding the component's present condition, good or failed, is present for some components; and (3) uncertainty regarding the component's underlying failure distribution is present in most real-world applications. Unfortunately, application of most models requires exact knowledge of the underlying failure distribution; however, graphical techniques, such as Total Time on Test (TTT), estimate optimal maintenance intervals based on empirical data; thus, they eliminate error resulting from type three uncertainty. The authors apply a TTT model to estimate optimal motor oil replacement intervals under conditions of three types of uncertainty. Their conclusions are: (1) There is no significant difference between synthetic (Stauffer and CONOCO) and petroleum motor oil lifetimes; (2) determining optimal oil replacement intervals requires application of a model more complex than the one applied; and (3) models that optimize an objective function without constraint are often not realistic. Thus, the authors develop and propose models that address three types of uncertainty and allow constraints (cost, availability, or failure risk) to be imposed on the model objective. (Author)

Descriptors :   *Mathematical models, *Preventive maintenance, *Graphics, Optimization, Reliability, Solutions(General), Methodology, Motors, Replacement, Intervals, Oils, Petroleum products, Theses

Subject Categories : Numerical Mathematics

Distribution Statement : APPROVED FOR PUBLIC RELEASE