Non-intrusive detection and characterisation of bearing faults from the temporal features of vibration

Abstract

ABSTRACT The behaviour of vibrations is widely used for non-intrusive inspection and health monitoring of bearings. However, automated methods, intended for predicting the bearing status, greatly depend on the features extracted from the vibration. Generally, time domain features are computationally simpler than frequency and time–frequency domain features. In this paper, the ability of time domain features to characterise the type of bearing fault is analysed. Types of bearing faults considered are healthy, inner race failure (IRF), roller element defect (RED) and outer race failure (ORF).The features being analysed are standard error (SE), absolute deviation of SE from the reciprocal of number of samples (β), entropy (E), variance (µ), standard deviation (SD), peak amplitude (PA), RMS, crest factor (CRF), impulse factor (IF), shape factor (SHF), energy and clearance factor (CLF). Among these, SE, ‘β’, IF and SHF characterise the status of bearing and the type of faults better than other features. The SE, ‘β’, IF and SHF corresponding to the vibrations acquired from normal and faulty bearings differ with a ‘P’ value of, 7.13866 × 10−23,7.06651 × 10−23, ≈ 0 and ≈ 0, respectively. These features can be used to distinguish defective bearings with 100% sensitivity and 94.73% specificity.