Back to Search Start Over

ACCURACY, EFFECTIVENESS AND IMPROVEMENT OF VIBRATION-BASED MAINTENANCE IN PAPER MILLS: CASE STUDIES

Authors :
Basim Al-Najjar
Source :
Journal of Sound and Vibration. 229:389-410
Publication Year :
2000
Publisher :
Elsevier BV, 2000.

Abstract

Many current vibration-based maintenance (VBM) policies for rolling element bearings do not use as much as possible of their useful lives. Evidence and indications to prolong the bearings' mean effective lives by using more accurate diagnosis and prognosis are confirmed when faulty bearing installation, faulty machinery design, harsh environmental condition and when a bearing is replaced as soon as its vibration level exceeds the normal. Analysis of data from roller bearings at two paper mills suggests that longer bearing lives can be safely achieved by increasing the accuracy of the vibration data. This paper relates bearing failure modes to the observed vibration spectra and their development patterns over the bearings' lives. A systematic approach, which describes the objectives and performance of studies in two Swedish paper mills, is presented. Explanations of the mechanisms behind some frequent modes of early failure and ways to avoid them are suggested. It is shown theoretically, and partly confirmed by the analysis of (unfortunately incomplete) data from two paper mills over many years, that accurate prediction of remaining bearing life requires: (a) enough vibration measurements, (b) numerate records of operating conditions, (c) better discrimination between frequencies in the spectrum and (d) correlation of (b) and (c). This is because life prediction depends on precise knowledge of primary, harmonic and side-band frequency amplitudes and their development over time. Further, the available data, which are collected from relevant plant activities, can be utilized to perform cyclic improvements in diagnosis, prognosis, experience and economy.

Details

ISSN :
0022460X
Volume :
229
Database :
OpenAIRE
Journal :
Journal of Sound and Vibration
Accession number :
edsair.doi...........89d35c98195ffa3373283f70dcccecc0