Multiple-Point Adaptive Performance Simulation Tuned to Aeroengine Test-Bed Data

Abstract

Adaptive simulation technology enables the calibration of a performance simulation code to a given in-service gas turbine and provides correct prediction of its performance. This is a fundamental prerequisite for reliable gas-path diagnostics and performance health monitoring. In this paper, a new offdesign performance adaption algorithm is introduced. Cranfield University's consolidated engine performance simulation code PYTHIA is enhanced with the capability of offdesign performance adaptation to model available field data. The software minimizes, via a genetic algorithm, an objective function that measures the error between an initial engine model output and the real engine data by varying some characteristics' scaling factors. In this study, a multiple-point adaptation procedure was applied to a two-shaft aeroengine. This generated an optimized engine model that minimized its deviations from a set of test-bed data. The adapted model was then tested against different real data, resulting in an average error, over 8 measured parameters, of less than 0.35%.