Vibration-based fault diagnostic platform for rotary machines

This paper provides a vibration-based diagnostic platform to systematically monitor and diagnose of rotary machine faults. Commonly rotary machine faults described in this paper are misalignment fault, bearing cage defect, ball bearing defect, bearing outer race fault and inner race fault. The use of structural resonance frequency, ISO 10816 for vibration level assessment, spectrum assessment for misalignment and bearing faults have been detailed. Beside fault diagnosis, repair action has been included to recommend different maintenance plans according to the faulty conditions. These methods form the basis of a knowledge-based system for diagnosis. The results have been successfully tested on a Hitachi Seiki high speed milling machine. The developed diagnosis platform minimized the need for human intervention in rotary machine performance monitoring and degradation detection.