T. Baudendistel, PC Krause and Associates, Inc; Steve Pekarek, Mario Rotea, Purdue University; E. A. Walters, PC Krause and Associates, Inc; Steve Peecher, Smiths Aerospace; Sean A. Field, Nathan E. Kumbar, Naval Air Systems Command
In this presentation, a cadre of recently developed hardware and software tools for the prognostics and health-monitoring of electric generators, motors, power electronic components, and electric power systems will be presented. One of the tools is a vibration sensor that is low cost, durable, and relatively straightforward to implement in a drive system or power electronic module. This sensor has been used to detect torque-ripple-induced vibration created by electric machines. It provides a convenient means to detect faults of both electrical and mechanical components of electric drive systems and also facilitates feedback-based control to mitigate the vibration source through control of the excitation to a machine. In addition to providing vibration feedback, it has also been shown to be effective as a back-up position sensor. Specifically, in applications where fault tolerance is critical, the sensor has been used to determine the position of the rotor of the machine when in-line or Hall-effect sensors fail.
A second tool is a thermal-based health monitor for electric machines that effectively predicts the thermal behavior of stator, rotor, and winding structures based upon input from a minimal number of thermocouples and stator current sensors. This observer has been tested on a 3.7 kW generator and is presently being used to evaluate the short- and long-term effects of pulsed loading of electric machines.
A third tool is a method of Distributed Heterogeneous Simulation (DHS) that provides a means to simulate the healthy and faulted behavior of large-scale systems at a speed and level of detail heretofore unachievable. Specifically, DHS enables the synchronized interconnection of any number of dynamic subsystem simulations, developed using any combination of a variety of programs/languages, and implemented on a single computer/workstation/supercomputer, a local area network (Intranet), a distributed, and any combination thereof. Theoretically, using an M-computer network, DHS can approach an 3M gain in computational speed over single computer, single numerical algorithm implementation.
It is shown that through coupling of these three tools, a comprehensive prognostic and health management system (PHM) for aircraft generators and associated electrical systems can be developed. Specifically, using DHS, component and system-level simulations of aircraft generator systems under nominal and failure modes can be performed efficiently. Using the simulation results obtained, the vibration sensor and thermal-condition monitor concepts are coupled to establish a multi-physics approach that can effectively detect component degradation and predict time-to-failure, and to develop feedback-based strategies for operation of generator electrical systems under component degradation or failure. Hence, maintaining the warfighting capabilities by extending the life of the generator electrical system.
2007 ISHM Conference, August 2007. Contact information: baudendistel@pcka.com