Research On Detection And Evaluation Of Abnormal Cutting Mark Based On S-shaped Specimen

Along with the rapid development of China’s aerospace,vehicle manufacturing and ship building industry,complex free surface workpieces are widely used,such as aircraft wings,impeller vanes and so on.But its complex geometric characteristics,such as variable curvature,variable twist angle,opening and closing angle,often lead to uneven surface structure after processing,or even abnormal processing marks.These make it difficult to characterize and evaluate.At present,the evaluation of Surface quality of complex free surface is still focused on the evaluation of contour,roughness.However,the abnormal machining marks presented on the surface can not be determined by reference to the standard,and are often evaluated as roughness or waviness.Therefore,it is necessary to explore a set of methods to detect complex surface contours,extract and evaluate the abnormal cutting marks.Based on the ‘S’ test specimen,a suitable measurement platform is built,and a new method of extracting and characterizing the abnormal cutting marks is discussed in this paper.As a supplement to the traditional roughness characterization.Using wavelet multi-scale analysis method to extract the characteristics of abnormal surface cutting marks.For the boundary effect produced by the Mallat algorithm of wavelet analysis,this paper finished the comparative study of boundary extension,and found a suitable extension method for the machined surface.Based on the robust norms of M-estimator,a robust iterative method of wavelet analysis is proposed to enhance the robustness of the filter.The concepts of surface damage ratio and damage area ratio are proposed on the evaluation of marks.Finally,the uncertainty of linear wavelet analysis algorithm is deduced,and the evaluation system of abnormal machined marks is improved based on Matlab GUI.In this paper,a large number of theoretical research and comparative analysis are carried out in the following aspects,and draw the relevant conclusions:(1)According to the structural characteristics of the ‘S’ specimen,the profile measurement platform is designed and built.And the abnormal marks of surface contours are extracted by wavelet analysis,and the two factors: the selection of wavelet base,the decomposition level and its correct calculation are discussed.(2)Analyze the causes of the boundary effect of wavelet analysis,study the boundary extension methods,and compare the extension methods to get the most suitable one for machining surface.Introduce M-estimator robust norms to wavelet analysis,and establish the relationship between the adjustable parameters of the weight function and the surface profile data,realize the wavelet robust filter analysis.(3)Proposed the concept of Rsd(surface damage ratio)and Rda(damage area ratio),which facilitates the intuitive quantitative evaluation of machined surface.(4)Derive the uncertainty transfer relationship of linear wavelet analysis algorithm,and develop the evaluation software system of surface anomaly marks based on Matlab GUI.