Date of Award

Fall 2013

Document Type

Thesis

Degree Name

Master of Science (MS)

Department

Electrical and Computer Engineering

First Advisor

Demerdash, Nabeel A. O.

Second Advisor

Yaz, Edwin E.

Third Advisor

Ionel, Dan

Abstract

This thesis contains a comprehensive analysis of experimental data collected from a case-study interior permanent-magnet electric machine experiencing winding short-circuit faults of varying severity. The experimental data collected from the faulted machine is from both motoring operation energized by a PWM sensorless ac drive as well as generating operation driven by the test bed dynamometer with a resistive winding load. For both modes of operation a complete set of three-phase voltage and current signals was recorded and analyzed. The three fault diagnosis techniques applied and compared regarding their ability to diagnose and prognosticate a winding fault in an interior permanent-magnet machine are motor current spectrum analysis (MCSA), negative sequence components analysis utilizing a symmetrical components transformation, and the space-vector pendulous oscillation method. The applicability of these diagnosis techniques to this case-study experimental interior permanent- magnet machine demonstrate that motor current spectrum analysis is inconclusive in diagnosing a winding fault when the machine is operated as a motor, but useful for diagnosing a winding fault when the machine is operated as a generator. Negative sequence components analysis successfully diagnoses the winding fault in both the motor operation and the generator operation cases. Finally, the space-vector pendulous oscillation method results are inconclusive for both the motor operation and the generator operation test cases.

COinS