Methodology for the Real-Time Management of Aero Gas Turbine Engine Usage Data

Modern turbine engine lifing approaches compute damage accumulation in real time based on actual usage, substantially increasing the useable life over a-priori cycle or hour limits based on assumed worst-case operating conditions. However, these new algorithms are complex, consume substantial computing resources, and generate enormous amounts of data that must be archived to allow data mining and/or recalculation of damage post-mission. Sentient Corporation will develop intelligent real time data reduction algorithms that preserve the only those features of the usage data that are required for accurate damage accumulation according to the physics of failure. An off-board design time tool will automatically analyze the damage accumulation models to determine the specific data and data rates required by each. This information will be used by on-board logic to optimally sample data in real time. A modular implementation will allow the system to adapt as models are updated. In Phase I, a simplified demonstration system will be assembled in Matlab with at least one model representative of each major damage accumulation mode. The output of this system in response to recorded mission data will be compared to the system without the new data reduction algorithms to quantify the gains made and show that damage accuracy is not adversely affected.