Developing Criteria for Sample Sizes in Jet Engine Analytical Component Inspections and the Associated Confidence Levels

摘要

This thesis provides managers of USAF jet engine programs with a model to help determine an appropriate sample size of engines/components to be inspected in a Lead-the-Force/Analytical Component Inspection (LTF/ACI) program. The major purpose of such a program is to identify problems and failure trends in engines/components before the problems are experienced by the majority of the fleet, based on the concept that a sample of engines with accelerated operating hours can represent the future status of the entire fleet. Initial engine/component inspection intervals for the fleet are set low and extended as LTF/ACI engines/components pass inspection criteria. Specific objectives are: (1) to determine what sample size of components is required to reach some specific level of confidence that the inspection intervals for the fleet can be increased, (i.e., the fleet can continue flying past that initial interval safely); and (2) to determine the risk or decrease in confidence associated with a less-than-optimum sample size. Small sample binomial statistics were used for the analysis due to the small number of engines/components usually inspected in an LTF/ACI program and the pass/fail nature of the inspection plan. The study found that the increase (decrease) in confidence attained by varying the sample size of engines/components slightly is significant enough to warrant careful consideration by managers attempting to balance cost, logistical, and engineering constraints. Data tables and graphs present the required sample sizes to ensure varying confidence level for varying levels of an acceptable number of components/engines that pass inspection within specified error limits. (edc)