Hostname: page-component-7b7567c8c-tnkh4 Total loading time: 0 Render date: 2026-11-16T06:55:49.441Z Has data issue: false hasContentIssue false

Online classification for spalling detection and vibratorybehavior monitoring

Published online by Cambridge University Press:  20 October 2014

Sanaa Kerroumi*
Affiliation:
CReSTIC, University of Reims Champagne Ardenne, Moulin de la Housse, 51687 Reims Cedex 2, France
Xavier Chiementin
Affiliation:
GRESPI, University of Reims Champagne Ardenne, Moulin de la Housse, 51687 Reims Cedex 2, France
Lanto Rasolofondraibe
Affiliation:
CReSTIC, University of Reims Champagne Ardenne, Moulin de la Housse, 51687 Reims Cedex 2, France
*
a Corresponding author:sanaa.kerroumi@univ-reims.fr

Abstract

Core share and HTML view are not available for this content. However, as you have access to this content, a full PDF is available via the 'Save PDF' action button.

Vibration analysis is the most widely used tool in industrial application for machine’shealth condition assessment. Bearings, however, are very sensitive and require specialattention. Vibration analysis employs various signal processing methods such as spectralanalysis, time-scale, time frequency analysis, etc. these methods are used to analyzebearings’ vibratory behavior by monitoring the evolution of statisticalindicators.However, diagnosing the bearing depending on traditional features only isn’tsufficient to assure effective or reliable assessment of the component’ health condition.This paper proposes a multi-features online dynamic classification as a new method forfault detection and health condition monitoring for bearings; this technique uses multiplefeatures, including traditional features extracted from the raw signal, two specialfeatures extracted by wavelet analysis, the spectral kurtosis, coupled with a nonlinearprincipal component analysis and a dynamic classification to capitalize on the hiddeninformation in the time evolution of the features.Through this article, we introducedifferent measures and techniques used to characterize the health state of rolling, thenwe deploy a methodology using dynamic classification to detect early defect. To ensure analmost continuous surveillance, this methodology is based on a real-time analysis, anduses specific statistical indicators adapted to the experimental bench. Then, themonitoring of the degradation is achieved through the resulting class of the state ofdegradation. New parameters such as the speed of the class, the position of the class, theshape of the class will be discussed to inform the state of damage. The suggestedmethodology is validated by analyzing several fatigue tests from a fatigue bench bearingthrust ball referenced SNR51207.

Information

Type
Research Article
Copyright
© AFM, EDP Sciences 2014