Research On Wear Monitoring Of Engineering Ceramic Precision Grinding Wheel Based On Acoustic Emission Technology

As a typical hard and brittle material,advanced ceramics are machined mainly by precision grinding and lapping of ultra-hard fine abrasives.The wear of the grinding wheel is one of the main factors affecting the surface quality and accuracy of the workpiece.There will be loss of abrasive particles,broken abrasive particles,bond fracture and grinding wheel blockage,which will reduce the grinding ability of the grinding wheel and affect the surface quality of the ceramic workpiece.At present,the judgment of the wear degree of the grinding wheel is usually based on manual experience,and it is difficult to make a timely and accurate judgment.Therefore,in order to better ensure the grinding quality,improve the grinding efficiency and prolong the service life of the grinding wheel,it is necessary to formulate a reasonable method to conduct in-depth research on the monitoring of the grinding wheel wear state.In this paper,based on the acoustic emission technology,the monitoring of the grinding wheel wear state in the grinding process of alumina ceramics is carried out.The processing methods of the acoustic emission signal are compared and analyzed,and the influence of different process parameters on the acoustic emission signal is explored.The preferred combination has been discussed and analyzed in depth.The specific research work is as follows:(1)This paper firstly analyzes and compares the commonly used modal decomposition methods for acoustic emission signals,selects VMD algorithm to extract features of grinding acoustic emission signals;and proposes a signal reconstruction method with adaptive parameters VMD based on whale optimization algorithm.Acoustic emission signal has been tested to verify its feasibility.(2)By comparing the characteristics of acoustic emission signals under different process parameters,the frequency distribution of the acoustic emission events in alumina grinding was discussed.The whole-cycle acoustic emission signal acquisition experiment of grinding wheel wear,set the node to collect the acoustic emission signal and grind with constant parameters until the grinding wheel is completely passivated.Classification,providing theoretical basis and data support for the characterization of acoustic emission of grinding wheel wear.(3)The VMD algorithm based on adaptive parameters decomposes and reconstructs the acoustic emission signal,analyzes the corresponding relationship between each frequency band of the signal and the grinding event of the grinding wheel,and deeply discusses the change law of the time domain and frequency domain characteristics of the signal accompanying the performance degradation process of the grinding wheel.A random forest model was built,and the signal frequency band features were sorted and screened with the average Gini index drop as an indicator,and the optimal feature combination was selected to identify the grinding wheel wear state.Through the test,the comprehensive recognition accuracy of the model to the grinding wheel wear state was 85%.