Vibration-Based Power Spectral Density Analysis for the Detection of Multiple Faults in Rolling Element Bearings

摘要

The traditional industry is shifting towards fourth industrial revolution called “Industry 4.0” that includes automatic fault detection of machinery. The detection of mechanical faults mainly depends upon accurate condition monitoring of the rotating machinery that requires effective predictive maintenance techniques. The modern industry focusing on predictive maintenance techniques so that these faults can be detected automatically. This article proposed vibration-based power spectral density analysis to detect multiple faults in rolling element bearings. The power spectral density plots were computed at 16.7, 25, 40, and 50 Hz operating speed with a bearing loader and without installing a bearing loader on the vibration data taken from SpectraQuest's Machinery Fault Simulator? through the accelerometers mounted on inboard and outboard bearing housing. The four various types of faults: outer race, inner race, ball, and combined fault have been analyzed separately in this study. The power spectral density plots of damaged rolling element bearings showed high amplitude vibrations with harmonic behavior. This research article proposed that vibration-based power spectral density is an excellent measurement to detect faults in rolling element bearings that endows technical support to Industry 4.0.