SEARCH

SEARCH BY CITATION

Abstract

An earlier nonparametric statistical study of GE F414 engine removals from operational F/A-18 aircraft in US Navy service provided insights into the lifetime patterns of engine removals for various causes. Inspection of the estimated hazard function for engine removals for foreign object damage (FOD) suggested that a parametric analysis using Erlang distributions might be fruitful, bolstered by a hypothesized relevance to the maintenance procedures governing engine removals for this cause, and their outcomes. The objective was both a better model to forecast engine removals and to provide insight into the number of FOD incidents it took to drive an engine removal. Gamma and Erlang distributions did better fit the removals data and provide a tool for predicting engine removals, aircraft availability impact, and the resultant maintenance workload. A parametric model using a cascade of Erlang functions was developed to simulate the combined FOD/line maintenance process, which provides insight into the outcomes expected under reasonable simplifying assumptions. This model predicts that the key research issue, the probability that a typical FOD event prompts a removal, cannot be estimated from engine removals data alone. Field data must be collected to gain understanding of the underlying frequency of FOD and the utility of the present inspection criteria.