Wear Prognostic on Turbofan Engines

Jérôme Lacaille, Aurelie Gouby, and Olivier Piol
Submission Type: 
Full Paper
AttachmentSizeTimestamp
phmc_13_009.pdf1.76 MBOctober 9, 2013 - 2:02am

One of the most evident characteristic of wear for a turbofan engine is the exhaust gas temperature (EGT). It seems clear that this temperature increases when some carbon deposits on the turbine, when the compressor efficiency diminishes so the fuel flow should increase to produce the same amount of thrust, or even when some unbalance opens the spaces between the turbine and the casing. In any cases, an increase of the EGT should be analyzed because it is a wear symptom of the engine. It is mostly concluded by a water wash in the best case or a shop visit inspection and repair in the worst case. The engine manufacturer defines a schedule plan with its customer based on consumption of the EGT margin. This margin is the amount of available increase of the exhaust temperature before an inspection. Contractually, the engine is restored with a minimum EGT margin after each repair. Thus it is up to the manufacturer to understand how this margin is used to plan shop visits and to the company to estimate the current state of its engine. However the EGT measurement is subject to a lot of noise and the company regularly washes their engines to increase randomly the margin and their capabilities. In this article we present a simple, automatic and embeddable algorithmic method to transform the successive EGT measurements in a delay indicator computed after each flight giving the amount of available use time. One challenge is to take care of the random wash or repair executed by the user. Finally this indicator may be transmitted automatically with the other data broadcasted by the aircraft computer (ACMS/ACARS) and it is used by the manufacturer to prepare his shop logistic.

Publication Year: 
2013
Publication Volume: 
4
Publication Control Number: 
009
Page Count: 
8
Submission Keywords: 
Prognostic
wear
aircraft engine
Submission Topic Areas: 
Data-driven methods for fault detection, diagnosis, and prognosis
Submitted by: 
  
 
 
 

follow us

PHM Society on Facebook Follow PHM Society on Twitter PHM Society on LinkedIn PHM Society RSS News Feed