Validation Tools and Data for PHM Technologies

Development of accurate and effective new diagnostic and prognostic capabilities for JSF PHM will require extensive data on both health and faulted systems states. While healthy signatures can be collected quickly and easily, faulted systems are only observed rarely, and duplicating faulted conditions with seeded fault tests is time-consuming and expensive. Simulation models for faulted conditions will allow sparse seeded fault data to be augmented with accurate simulated fault signatures to provide full data coverage across operating conditions, fault types, and fault severities for situations where extensive physical testing is prohibitively expensive. In addition, these models can be used to determine optimal sensor locations and diagnostic techniques. In Phase I, Sentient Corporation developed a fault signature simulation model for bearing vibration (referred to as the Fault Vibration Simulator or FVS) that was shown to accurately reproduce data from physical tests. In Phase II, Sentient will further develop the bearing model and will also create models for gears and other rotating machine components commonly used in aircraft propulsion. Tools and techniques for gathering system-specific transfer function information needed to calibrate the model will be developed to enable accurate modeling results from simple impulse hammer tests on actual hardware, or from FEA results if no hardware is available. The FVS models will be implemented in the COTS Simulink environment for maximum flexibility. The models will undergo a series of demonstration and refinement tests on real systems, beginning with simple test rigs and culminating in a full scale demonstration model of a selected JSF propulsion subsystem, such as the lift fan gearbox.