A Probabilistic Analysis of the Benefits of In-Service Fatigue Damage Monitoring for Turbine Engine Prognosis

Prognosis and structural health management based on data acquisition and data fusion from on-line sensors, combined with analytical methods for data interpretation and decision- making offer the potential for significant enhancements in assuring the reliability and readiness of high-value assets. The specific objective of the current study was to assess the potential benefits of continually monitoring fatigue damage with in-service sensors. Specifically, a probabilistic, damage-based life prediction analysis was employed to assess the trade-off between sensor sensitivity and interrogation frequency on the probability of failure. Analyses were performed using an enhanced version of the DARWIN ® probabilistic life prediction code that was modified to employ actual (F16/F100) usage data from flight data recorders. The methodology is demonstrated for the specific case of a titanium alloy compressor disc—a fracture critical component in turbine engines. The results demonstrate the reliability benefits of in-service monitoring of fatigue cracks in turbine engine discs, even if on-board sensors are 6 to 10 times less sensitive than those currently used for depot inspection. This benefit is due to the continual feedback provided by in-service sensing that enables continual updating of the decision making process. Target sensitivities for the development of on-board sensors are also suggested.