Spectrum Prediction Indicating Bearing State in Induction Motor by Forced Vibration Analysis and Fuzzy Logic Technique

Abstract

Bearing faults (BF) are the main cause of induction motor failures; virtually, all techniques for diagnosing these faults have been based on expensive and time-consuming procedures. Earlier fault detection may minimize financial losses brought on by unforeseen failures in the industrial sector. This article predicts a spectrum envelope with a healthy and faulty bearing with inner race (IR) and outer race (OR) fault frequencies indicating the bearing's faults. Forced vibrational analysis (FVA) is implemented to determine whether the components will lose functionality over time, predicting one or more recognizable fault frequencies in the spectrums. The spectrum envelope indicates the low magnitude for healthy bearing and the high magnitude for faulty bearing of vibration velocity (mm/s) and acceleration (G) as per ISO 10861. The fault frequency of various defects in bearing components like IR, OR fault, spinning frequency, and turning frequency of balls is represented with peak values that are noticeable in a spectrum from the test (FVA). The experimental waveforms include the vibration frequencies that the bearing component that typically vibrates can cause sudden failure. Finally, the experimental data set is used to implement the fuzzy logic technique and investigate whether or not the motor condition meets industrial needs.