The Evaluation And Calculation Methods On Abrasive Grains Of Single-layer CBN Thin-grinding-wheel Based On Image Processing

The process of grinding is replacing the turning and the boring in modern manufacturing, which application is more and more widely spreading, and the utilization rate of super hard abrasive material tools is getting higher and higher. Especially, the single-layer CBN grind ing-wheel is popularized in application, and its production scale is increasing year by year, but the production process of the grinding wheel is still not perfect. There are lots of disadvantages such as the abrasive grains falling off, the failure of grinding caused by the lack of space among the grains and the absence of testing measurement for the single-layer CBN grinding wheel. An important reason is that the distribution of CBN abrasive grains on the grinding wheel isn’t fully taken into consideration during the manufacturing process.In this thesis, the research aims are committed to the distribution of CBN abrasive grains. While to the CBN grains’micron-level size, the conventional means could not meet the testing requirements. So, an evaluation method on abrasive grains of single-layer CBN thin-grind ing-wheel must be presented based on the image processing. Firstly, the image formation of the grinding wheel in the cutting area must be solved. For realizing the terminal goals, an SCM controlling automatic imaging instrument was designed, which was driven by an X-Y workbench and the image formation of the CBN grains is acquired with a given industry digital microscope.And the control programs were written in assembly language that achieved the semi-automatic control of the imaging process.The noise reduction, the segmentation and the edge detection process were achieved for the acquired image through the imaging toolbox in MATLAB. The noise reduction processes include the image smoothing by threshold method, using median filter to eliminate the impulse noise and the elimination of white Gaussian noise using wavelet. The image segmentation was completed to obtain the clear distinction between abrasive and background by the HSI threshold method. The edge detection using LOG operator had highlighted the profile of abrasives.The studies of abrasive grains were carried out in the lateral direction and the longitudinal direction. In the lateral direction, a new evaluation method had been proposed on the abrasives distribution. It combined the variance of abrasive in loose parts and the variance of area ratio in other parts. In the longitudinal direction, the3-D reconstruction of abrasive grain was completed by stereo vision formation which need to image twice at the same abrasive grain, and then the surface landforms of the grinding wheel could be reflected perfectly with the3-D reconstruction.