Study On 3D Surface Topography Decomposition Technical Based On Wavelet Transform

Workpiece surface texture includes surface roughness, waviness and shape deviation, which are caused by the influence of different factors during the machining process. While the use performance and longevity of the workpieces need higher surface roughness precision, it is necessary to separate the workpiece surface texture more precisely. The traditional separation technologies of surface topography are limited by the Fourier transform, and the surface roughness gotten by this method are not precise enough. In order to get higher reliability of microscopic surface morphology, to separate and extract the surface roughness quickly and accurately, this paper studies the wavelet filtering method based on the second linear phase shift wavelet transform theory.In this essay, several common methods to separate and extract the information of surface morphology are analyzed, their advantages and disadvantages are also compared in their application fields. A new method to extract the information of surface roughness based on the SGWT(second generation wavelet method) theory is proposed by studying the research status and problems existing in the separation of the information of surface morphology. The improved algorithm is obtained by analyzing the Mallat algorithm and its lifting principle based on the second linear phase shift wavelet transform theory. The algorithm is implemented by constructing lifted wavelet filter, and the improvements of the wavelet decomposition function compared with the MATLAB toolbox wavelet function are given. The mathematical model is built according to the composite components of surface morphology and the model is solved based on the principle of wavelet construction. The dividing frequency and decomposition layers of surface morphology information can be determined by analyzing the spectrum, and the wavelet function used in this thesis can be determined according to the selection principle and characteristics of the commonly used wavelet function. Through software implementation of the algorithm and separation model of the surface morphology, the information of surface morphology located in different scales can be obtained through simulating the process of decomposing the surface morphology by using the second generation wavelet filtering, and the information of surface roughness can be gotten after reconstruction. The method used in this paper is adopted to separate the surface morphology of grinding surface roughness of standard specimen. In these experiments, the standard value based on the arithmetic average deviation of 2D contour is regarded as a benchmark, at the same time, the 3D surface arithmetic average deviation is assessed and compared with the results measured by the instruments. All the results in this paper can be used to verify the validity and accuracy of the method using wavelet to extract surface roughness.This paper decomposes the surface morphology of grinding surface roughness of standard specimen using the second generation wavelet filter and gets the arithmetic average deviation of 2D contour of the information of the surface roughness Ra. The results show that Ra locates in the allowable error range compared with the standard values and the arithmetic average deviation of 3D surface Sq deviates little from the results measured by the instruments. It can be proved that the method used in this thesis is feasible.