Analysis Of Surface Topography In Ultra Precision Turning Considering The Relative Vibration Between Tool And Workpieces

Ultra-precision turning is widely used in the manufacture of optical precision components with excellent machining accuracy and production efficiency.In the actual machining process,the relative vibration between the tool and the workpiece inevitably exists,which causes the workpiece surface to have texture feature with specific shape,which seriously affects the surface quality of the workpiece and deteriorate its use efficiency.In order to realize the traceability of vibration error based on the analysis of the surface topography data and the extraction of texture features,and to further improve the surface quality of ultra-precision turning workpieces,the following work has been carried out in this article:Firstly,a simulation model of surface topography in ultra-precision turning considering the relative vibration between the tool and the workpiece in the circumferential,axial,and radial directions was first established based on the principle of tool contour reproduction in this paper.Subsequently,the validity of the simulation model was verified by experiments,and the influence of relative vibration in different directions on the surface topography of ultra-precision turning was studied.Secondly,the multi-scale analysis and texture feature extraction of surface topography in ultra-precision turning were studied.Due to the shortcomings of serious frequency aliasing and poor directional selectivity of traditional 2-dimensional discrete wavelet transform(2D-DWT),this paper applied 2-dimensional dual tree complex wavelet transform(2D-DTCWT)to the multi-scale analysis of the surface topography in ultraprecision turning,and in order to select the layers of wavelet transform properly,the analysis of peakedness and the root mean square error(RMSE)between the reconstructed surface and the original surface were conducted.The performance of 2D-DTCWT was compared with 2D-DWT from the aspects of energy leakage and texture feature,which validated the superiority of 2D-DTCWT in the multi-scale analysis of the surface topography in ultraprecision turning.Finally,in order to automatically identify surfaces containing vibration-induced texture features during multi-scale analysis of surface topography in ultra-precision turning,this paper implemented pattern recognition of vibration-induced texture features based on local binary pattern(LBP)operator and support vector machine(SVM),and the recognition accuracy reaches 97.50%.