Early fault detection in automotive ball bearings using the minimum variance cepstrum

Cited 44 time in webofscience Cited 0 time in scopus
  • Hit : 612
  • Download : 0
Ball bearings in automotive wheels play an important role in a vehicle. They enable an automobile to run and simultaneously support the vehicle. Once faults are generated, even if they are small, they often grow fast even under normal driving condition and cause vibration and noise. Therefore, it is critical to detect faults as early as possible to prevent bearings from generating harsh noise and vibration. How early faults can be detected is associated with how well a detecting method finds the information of early faults from measured signal. Incipient faults are so small that the fault signal is inherently buried by noise. Minimum variance cepstrum (MVC) has been introduced for the observation of periodic impulse signal under noisy environments. We are particularly focusing on the definition of MVC that goes back to the original definition by Bogert et al. in comparison with the recently prevalent definition of cepstral analysis. In this work, the MVC is, therefore, obtained by liftering a logarithmic power spectrum, and the lifter bank is designed by the minimum variance algorithm. Furthermore, it is also shown how efficient the method is for detecting periodic fault signal made by early faults by using automotive ball bearings, with which an Automobile is equipped under running conditions. We were able to detect incipient faults in 4 out of 12 normal bearings which passed acceptance test as well as in bearings that were recalled due to noise and vibration. In addition, we compared the results of the proposed method with results obtained using other older well-established early fault detection methods that were chosen from 4 groups of methods which were classified by the domain of observation. The results demonstrated that MVC determined bearing fault periods more clearly than other methods under the given condition. (C) 2013 Elsevier Ltd. All rights reserved.
Publisher
ACADEMIC PRESS LTD- ELSEVIER SCIENCE LTD
Issue Date
2013-07
Language
English
Article Type
Article
Keywords

ROLLING ELEMENT BEARINGS; WAVELET TRANSFORM; DIAGNOSIS; MACHINE; SIGNALS; GEAR; ENHANCEMENT; ENVELOPE; NOISE; MODEL

Citation

MECHANICAL SYSTEMS AND SIGNAL PROCESSING, v.38, no.2, pp.534 - 548

ISSN
0888-3270
DOI
10.1016/j.ymssp.2013.02.017
URI
http://hdl.handle.net/10203/175027
Appears in Collection
ME-Journal Papers(저널논문)
Files in This Item
There are no files associated with this item.
This item is cited by other documents in WoS
⊙ Detail Information in WoSⓡ Click to see webofscience_button
⊙ Cited 44 items in WoS Click to see citing articles in records_button

qr_code

  • mendeley

    citeulike


rss_1.0 rss_2.0 atom_1.0