Engine system prognosis

Littles Jr. Jerrol W., Buczek Matthew B.

Abstract:
The vision of the engine system prognosis program (ESP) is to provide military commanders data and quantitative performance predictions to allow them to adaptively manage, deploy, and use individual combat systems to the limit of their individual capability. Military engines are currently managed to life limits based upon fleet wide statistics and average expected useage. While this approach has for the most part provided a high degree of reliability, it also has negative aspects. For example, this approach forces the replacement of life-limited parts when a very small percentage (say 1 in 1000) would be expected to have expended their useful life. Hence, 999 "good" parts (parts with remaining life) are discarded along with the one "bad" part. The current approach utilizes a statistically conservative philosophy due to the numerous sources of variability that affect component, and system, life predictions. GE Aircraft Engines and Pratt & Whitney have teamed in a DARPA/DSO sponsored effort develop an Engine System Prognosis (ESP) tool to more accurately predict system health and near-term capability by significantly reducing, and accounting for, the variability inherent to gas turbine engine manufacturing and operation. This tool will allow ESP-based health management systems to significantly reduce the need for service-interrupting inspection or maintenance action, and will result in significant safety, readiness, and cost of ownership improvements. The ESP program utilizes a modular approach to predict asset-specific component health, and associated prediction uncertainty, for the key engine subsystems. The ESP system implements advanced physically based models to predict component capability, while tracking prediction-related variability. Structural transfer functions (STFs) are generated to quantitatively link sensor data to remaining capability for key components and subsystems. A novel system architecture is being used to fuse information from various sensors and subsystems to provide a system-wide capability assessment and predicted effect of various usage scenarios for military commanders, and forecast pending action for maintenance personnel.

回上页