Community Research and Development Information Service - CORDIS

Abstract

The subject of this research project was the development of an expert system for diagnosis of mechanical wear of a machinery park in the steel industry. The initial system was mainly based on oil and vibration analysis. Extensive research work was done in the field of vibration analysis especially for diagnosis of antifriction bearings. The envelope analysis turned out to be the most reliable method to detect bearing defects. Unfortunately an automatic diagnosis using the envelope analysis is not working just by checking amplitudes of characteristic frequencies. It is necessary to recognize patterns in the envelope spectrum, which are characteristic for bearing defects. This recognition must work in a fault tolerant way. In the second stage of the project a monitoring system was developed which is able to diagnose antifriction bearings with an expert system. It works with an artificial neural network. With its help it is possible to automatically recognize defects of an antifriction bearing even at an early stage of damage.

Additional information

Authors: PICCIRELLI A, ARBED Recherches, Esch-sur-Alzette (LU)
Bibliographic Reference: EUR 16793 DE (1998) 44pp., FS, ECU 8.50
Availability: Available from the (2)
ISBN: ISBN 92-828-1692-3
Record Number: 199810354 / Last updated on: 1998-04-08
Category: PUBLICATION
Original language: de
Available languages: de