Vibration Signal Analysis of Induction Motor Bearing Faults: Some Aspects

Vibration monitoring and analysis techniques are among the most commonly used methods in identifying defects in induction motors. Motor defects like bent shafts and bearing defects are analyzed, focusing on twice-line-frequency (100 Hz) components for the vibration data of an induction motor belonging to a petrochemical industry. The motor defect in this case was a bent shaft. A marginal correlation between the vibration data and the 100 Hz component could be seen. A similar study is attempted using another data set collected from web resources. The tracking twice-line-frequency data reveals progressive deterioration of the motor condition with time; this is in spite of the motor exhibiting vibrations within the acceptable limits as per ISO 10816-3. As the vibration signals are non-stationary, the second data set is analyzed using discrete wavelet transform (DWT). The sub-band D4 of DWT showed a definite correlation with the ball-bearing faults.