Progress of Measurement of Mechanical Wear With Wear Debris

The wear of mechanical parts and elements is the main reason of the failures of machines, and causes enormous economic loss as results of failures of mechanical wear. Thus, the operation conditions of equipment are monitored with respect to wear process and wear state. The wear mechanisms and the wear stages can be studied with the detection techniques in terms of wear debris, and the maintenance based on machine condition and the prediction of the remain service life of mechanical elements and parts can be achieved. The principles of wear debris detection, statistical analysis, morphological analysis, image segmentation and recognition and intelligent examine based on artificial intelligence were reviewed. The ferrography methods have direct reading ferrography, analytical ferrography, rotary ferrography and on-line ferrography. The principles of debris detection are classified as ferrography, Hall-effect, laser net fins, microfluidic detection, radial inductive sensor, electrostatic sensor, ultrasonic echo detection and detection method with parallel resonance structure. The state and the art of measurements using wear debris of above different principles were comparatively reviewed. On the other hand, the challenges and existed problems, which need studied further, of wear debris for detection of mechanical wear are proposed and presented. The automatic analysis of debris size, edge profile and surface texture based on machine vision would obtain its feature parameters. The statistical computation with group theory can avoid the dependence on experience of researchers, and achieve more efficient and accuracy analysis of debris features. The characterization of debris morphology with fractal theory can obtain their scale-invariant parameters, and the inverse fractal analysis from small scale to larger scale can provide new information for the debris morphology. The morphological feature extraction based on three dimensional view images for wear debris analysis can present full description of the wear debris. The introduction of the rough set theory to the fractal analysis debris morphology can improve the efficiency and accuracy of the segmentation of wear debris. The birth rate of new wear debris is an effective parameter to assess the wear stage and its tendency, with which remain service lives of mechanical elements can be predicted. The integration of different principal methods can improve the accuracy of wear monitoring, and provide more reliable prediction results for the wear failure of machines. Moreover, an exited problem and useful topic is the integration of the intelligent analysis methods presented by different researchers, and develop one software for the monitoring of wear failure of machines, which is a trend of monitoring method of mechanical wear. The final goals are the automatic failure diagnosis of mechanical wear based on the information of wear debris, and the prediction of remain service life of mechanical elements as results of mechanical wear.