The development of a comprehensive machine condition monitoring program

Abstract

Modem machine condition monitoring programs generally feature the use of either oil and wear debris analysis or vibration analysis. This is the case as expertise in both techniques is commonly not available, and that the correlation of these techniques for a range of fault conditions has not been documented. In this study, the use of a Laser Scanning Confocal Microscope and image analysis software has enabled numerical descriptors to be applied for wear debris analysis, thus allowing the condition monitoring program to operate solely on numerical analysis. Fault conditions of bent drive shaft, constant overload and contamination were imposed on a laboratory test gearbox. Primary and the development of secondary faults were identified, which allowed the detection abilities and correlation of the analysis techniques to be compared. This study is part of a comprehensive analysis into the correlation of analysis techniques. Future research will include the development of an expert system for machine condition monitoring based on artificial intelligence.