Novel rolling bearing diagnosis technology using spectral kurtosis and the wavelet higher-order spectra

L. Gelman, T. H. Patel

Research output: Contribution to journalArticle

2 Citations (Scopus)


A novel technology based on a combination of spectral kurtosis and the wavelet higher-order spectra for the detection of local fatigue defects in rolling bearings by vibration measurement is studied. The bearing test-rig has a coupled motor driving a shaft supported on three identical rolling bearings. The drive provides a 20-60 Hz supply frequency. Tests were conducted at a full-speed and full-load condition. The full-speed and full-load condition corresponds to a 60 Hz supply frequency and 196 N resultant radial load. The test bearing was placed on the non-drive end. A systematic methodology for the selection of parameters of the wavelet higher order spectra is developed and successfully validated experimentally. Experimental validation of the technology is performed using test-rig data related to undamaged bearings and bearings at an early stage of local fatigue damage. The high effectiveness of early bearing diagnostics by the proposed technology has been experimentally demonstrated using the Fisher criterion and impact detection rate. The mean Fisher criterion and the impact detection rate for all impacts is 5.8% and 98%, respectively.
Original languageEnglish
Pages (from-to)452-456
Number of pages5
JournalInsight: Non-Destructive Testing and Condition Monitoring
Issue number8
Publication statusPublished - Aug 2015
Externally publishedYes


Cite this