A systematic approach to assess the impact of engine measurement uncertainty on the analysis uncertainty of maintenance induced performance recovery is presented. For that purpose a methodology is developed in a first step to estimate the random measurement uncertainty of production instrumentation based on engine monitoring data. The methodology is specifically designed for the requirements of a maintenance provider without Original Equipment Manufacturer (OEM) information on the measurement chain's components and their elemental uncertainties. The filtering steps, needed to process the raw measurement data until a conservative estimation of the random measurement uncertainty is obtained, are described in this paper. The estimated random measurement uncertainty is used to assess the analysis uncertainty of engine performance parameters and their recovery during an engine shop visit using a Monte Carlo Simulation (MCS). The uncertainties of different analysis methods are compared with respect to their sensitivity to the estimated measurement uncertainty and discussed with respect to their applicability to the problem of analysing performance recovery. These reflections include the prospect of differently workscoped engines. Finally, the basic ideas of the student-t problem are used to establish a confidence interval for the analysed performance recovery as a function of available in-service data. A limit value is established for the number of averaged engine snapshots resulting in an improved performance recovery analysis.
展开▼
