Event Details
On-line Current Signature Analysis Utility for Fault Diagnosis of Three-Phase Machines
Presenter: Yaju Jadeja
Supervisor: Dr. Aaron Gulliver and Dr. Subhasis Nandi
Date: Wed, April 24, 2013
Time: 14:00:00 - 15:00:00
Place: EOW 430
ABSTRACT
Abstract:
Condition monitoring of rotating electrical machines is essential for extending their life and performance while providing advance warnings of failures. Early detection of faults allows correction of incipient failures before they occur, thus avoiding unscheduled downtimes due to unexpected failures. Existing theories for detecting common induction machine faults such as broken rotor bars/end-rings, bearing faults, air-gap eccentricities and winding shorts rely heavily on spectral analysis of machine quantities such as torque, vibration and stator currents. A utility equipped with waveform and spectral analysis techniques such as FFT, Park's vector and complex double sided FFT can reveal many machine irregularities that can lead to detection of various faults. A current signature analysis based on-line, real-time fault detection utility is developed to locate various characteristic fault frequencies and signature patterns in 3-phase machines. The proposed interactive utility has been implemented in MATLAB and experimentally tested on various three-phase synchronous and induction machines. The performance and reliability of the system is verified using machines with broken rotor bar/end-ring and different severity levels of static, dynamic and mixed eccentricity faults.