Accurate life prediction and monitoring for gas turbine engines has become increasingly important in recent years as commercial aircraft fleets are being offered through guaranteed engine maintenance programs, where plan rates are based on mission profiles, operating environment, operational hours and cycles accumulated. Hence, accurate monitoring and life predictions of critical engine components is associated with a tremendous financial incentive. A state of the art gas turbine engine carries up to 5000 sensors, which can be used to evaluate the performance of the engine. This data can be used to monitor engines in real-time, as well as collecting and analyzing that data after being streamed via satellite during flight, where algorithms can evaluate and prevent technical issues before they occur. The data collected provides engine manufacturers with early warnings related to failure diagnosis, and it enables airlines to schedule engine maintenance efficiently and in a cost effective manner. Due to the nature of the engine’s operational environment, sensors cannot be placed in certain areas of interest inside a gas turbine engine. Furthermore, thermo-mechanical models are often complex and computationally expensive to run in real time. Hence, in this work we describe the development of thermo-mechanical reduced models that can act as virtual sensors, in locations where real sensors cannot survive, and hence approximate damage variables at critical locations on a component of interest, which can be used for real-time diagnostics.

This content is only available via PDF.
You do not currently have access to this content.