Investigations into Some Parameters of Vibration Signal of Faulty Bearings with Wavelet Transform

The vibration signals generated by a rotor bearing system are greatly influenced by the occurrence of a fault on the bearings. Monitoring of parameters like peak, overall rms, kurtosis, crest factor, and power from the time domain signal reveals the status of health of bearings supporting the rotor and serves as the easiest approach. However, the majority of these parameters give an estimation of the overall health of the system and not just the bearings. This work addresses the use of Discrete Wavelet Transform (DWT) to filter the bearing related signal and then monitor the health of bearing. The work also discusses on the selection of appropriate sampling frequency for the collection of time domain signal and demonstrates the dependence of parameters on the same. The results indicate that the use of DWT for a signal with higher sampling frequency has increased the accuracy of prediction of estimation of the defect.