Effective turbine engine diagnostics

Current turbine engine testing rejection criteria do not fully check for all observed and known operational failure modes. A failure mode indicator that is not recognized is a slow gas generator at rated power, which indicates turbine deterioration, yet no minimum speed limit exists. Instead, the slow gas generator speed is considered indication of a better engine. Knowledge Engineering has been used to develop an Artificial Intelligence (AI) package based on the six (6) sigma method to change the field recorded data into diagnostic information. The information is presented in an easily understandable Referred Engine Diagnostic Data (REDD) format that allows the decision maker to see differences in the normal hidden condition of individual engines and rank them on their relative capability. A recording instrumentation system is used to do ground or ill-flight engine tests and electronically store engine performance data. The Artificial Intelligence program changes the data into diagnostic information about the status of the engine under test. Added decision parameters clearly show each engine´s abnormal operation, revealing the hidden faults. Useful knowledge of each engine´s condition is essential to direct limited resources to the least capable engines. The demonstrated cost reductions and operability improvements associated with the use of the instrument package and the AI diagnostic program have been significant