Study On Prediction Of Sliding Running-in Wear Based On Surface Topography

Running-in process is an inevitable wear stage of machine system. Unworn machined surfaces reach low wear rate by dynamic wear during running-in wear process, which is important to prevent seizure, establish steady work condition and extend service life. The modification of surface topography is a significant feature of friction pairs, and surface topography after running-in wear directly influences the performance of friction pairs during steady wear stage. Moreover, in contrast with steady wear process, the wear volume of running-in process is higher, and further influences the service life of friction pair. Since the contact area of sliding friction pair is larger, surface topography in contact area plays more important role in sliding friction pair in contrast with rolling friction pair. Investigating the evolution of surface topography during running-in wear process of sliding friction pairs, studying the relevance of surface topography before and after running-in wear process, and establishing wear model predicting surface topography after running-in and running-in wear volume based on surface topography before running-in are of great importance in optimization design of surface topography of friction pairs to control running-in result, and improve performance and service life. The main work and innovation are as follow:Based on contact mechanics, an analytical running-in model of sliding friction pair was established. Based on contact mechanics and mixture lubrication feature, according to Greenwood-Williamson model and mechanical principle of surface microstructure, considering adhesive contact of friction pair, this paper proposed a new analytical model of running-in wear.Based on the analytical running-in model, running-in wear process was simulated. The relevance of surface topography before and after running-in was demonstrated by analysis of simulation result. Aiming at traditional conclusions, running-in experiment was designed to investigate the influence of work condition and surface topography on surface topography after running-in. According to the analysis result of areal surface topography evaluation, the relevance of surface topography before and after running-in was confirmed.Based on minimum redundancy-maximum relevance feature selection, a new method of extracting feature papameters of surface topography of running-in wear prediction model was proposed. Since the areal surface topography evaluation has many parameters, the model covering all parameters is too complicated to achieve expecting performance. Moreover, the evaluation based on all parameters is redundant from the view of function. Therefore, it is necessary to extract feature parameters to present surface topography and relate with the objective features of prediction model. Based on the feature analysis of surface parameters, the feature parameter sets were built by minimum redundancy-maximum relevance feature selection.A new method of modeling running-in wear feature based on least squares support vector machine was proposed. According to the extracted feature parameters of surface topography, establishing the model of predicting surface topography after running-in and running-in wear volume enable the control of running-in result. Considering the complication of running-in wear process, based on the relevance conclusion of surface topography before and after running-in, this method regarded the running-in as black box and ignored the interior wear mechanism, and studied the relationship of input (work condition and surface topography) and output (surface topography) of running-in wear process. Meanwhile, the running-in wear volume was taken into the consideration of running-in output of prediction model, the influence of surface topography on running-in wear volume was investigated. Running-in experiments confirmed the efficiency and accuracy of running-in wear prediction model.