New Dimensions in Oil Debris Analysis - the Automated, Real Time, On Line Analysis of Debris Particle Shape

Abstract : Analysis of debris particles in machinery lubricating oil has long been used to detect the onset of wear and failure of oil wetted components in machinery. This paper explores the potential for direct linkage between on line particle shape detection and the automated early detection of machinery failures. The development and application of an optimum set of feature vectors based upon extensive particle shape data bases is designed to allow automation of the shape detection process, separation of fault and non-fault related debris, tracking of specific fault types, the elimination of false calls from non-failure related debris, and the ability to provide direct fault specific corroboration of machinery faults to vibration based analysis systems, such as HUMS. The paper contrasts this capability with current on line and off line particle detection and classification practice and capability. Finally the paper briefly describes LaserNet Fines on Line, a new technology embodying this capability designed to detect and classify metallic and non metallic particles sized from 5 microns to greater than 1000 microns.