The surfaces produced by milling and turning are ordinary in tribology research. With the unceasing improvement of the quality of surface manufacture, the method of evaluation and measure with objective and scientific to the surface roughness of milling and turning sets a higher request.It is well known that fractal is an upsurge branch in math fields in recent years. It has been an important interactive research branch in the field of mechanical surface to be united with fractal theory. This paper has researched the nature of mechanical surface produced by milling and turning based on fractal theory.According to the machined characterization of milling and turning parts, analyzes that 1) because of the shafts surface produced by milling and turning being 3D, this paper simplifies the 3D evaluation to 2D. 2) the primitive image of shafts is sensitive to illumination, and affected by reflected light and nick, at the same time including speckle, and is improved in the spatial and frequently area.3) the nick influence is removed by Stat. during data processing.Through the above analysis, in order to complete non-contact measurement, this paper built proper measure system with the concrete problem. The system is composed of the image capture system, processing system, and output system. The hardware elements includes an examination shelf with xyz direction as the platform, supported by the digital camera with 8 million pixels as the core part, the specialized LED lamp-house, and milling specimen with Ra0.8μm -0.1μm,turning specimen with Ra6.3μm -0.8μm.The main work is to measure three milling and turning shafts with unknown roughness. It takes Matlab as the software, which has enough merits in image processing and data analysis. The roughness of shafts is measured qualitatively and quantitatively. The system is simple and applicable, and is able to complete the roughness measure of milling and turning shafts under existing condition.Based on the fractal theory, this paper analyzed the characterization of fractal, distinguished anisotropy and isotropy, built the mathematical model and obtained the fractal parameters. The traditional roughness parameter are mostly based on statistics, they are scale-dependent. In order to overcome the deficiency, the fractal parameters are used to characterize the surface appearance effectively, which is complex, irregular and rough. Through the data processing, drew out the next conclusion.The fractal dimension is scale-independent, reflecting the complexity and irregularity of the fractal surface; the scale constant is the absolute measure parameter, evaluating the regularity in unit scale.The fractal dimension is same in every direction except parallel to the lay. The fractal dimension along the lay is less than that across the lay. It confirmed T.R.Thomas'theory and S.Davies'production.When Ra is same, the fractal dimension of milling is higher than that of turning. At the same time, the scale constant of milling is also higher than that of turning.Application of curve fitting method based on least square method and construction of 3-degree algebraic polynomial, we can get the relation of Ra and fractal parameters and compute the approximate value of roughness. This paper is composed of six chapters. The first chapter introduced the significance of this work, and summarized several existing roughness measuring technique and principle, finally elaborated the main content of this paper; The second chapter introduced the concept and nature of fractal, carried on a more detailed analysis on the relation between the feature of surface appearance and fractal,and specified the equipment of measure system including its hardware and software; The third chapter firstly narrated the fractal characterization of the anisotropy of rough surfaces, secondly built a mathematical model; The fourth chapter solved the uneven caused by illumination, which had big influence on image capture, finally improved the primitive image; The fifth chapter removed the nick influence by Stat., drew out the fractal parameters and analyzed the regularity ,then fitted the curve based on least square method and computed the roughness of machined shafts ; The sixth chapter drew out the conclusion, and pointed out the deficiency of this paper. |