Font Size: a A A

Laser Sharpening Evaluation Of Diamond Grinding Wheels Based On 3D Shape Recognition

Posted on:2023-05-30Degree:MasterType:Thesis
Country:ChinaCandidate:M Y GaoFull Text:PDF
GTID:2531307097976939Subject:Mechanical engineering
Abstract/Summary:PDF Full Text Request
Diamond grinding wheels have the advantages of large grinding ratio,long life cycle,and can maintain good size and shape,which are good tools for processing high hard and brittle materials.The resharpening quality of diamond grinding wheels affects the grinding performance,so it is necessary to evaluate the quality of diamond grinding wheels after laser resharpening.However,there is no unified standard for grinding wheel resharpening quality evaluation in the industry,and most of the evaluation methods are contingent and subjective,and there are metamorphic and recondensed layers on the surface of the grinding wheel after laser resharpening,which makes the topography of the grinding wheel complex,and it is difficult to accurately obtain the three-dimensional information of the grinding wheel surface based on the traditional inspection methods,and the efficiency is low.In this paper,we propose a three-dimensional grinding wheel surface topography inspection algorithm,and investigate the evaluation indexes of laser sharpening quality of bronze diamond grinding wheels,and propose three evaluation indexes of abrasive grain height and depth distribution index,abrasive grain density,and bond surface roughness,and conduct sharpening tests and grinding tests to verify the three indexes.Firstly,the 3D inspection algorithm is studied,and the 3D inspection algorithm of the grinding wheel surface matching the 2D image and 3D point cloud is proposed.The 2D surface image of the grinding wheel and its corresponding 3D point cloud are acquired simultaneously,and the U-Net semantic segmentation neural network is improved to segment the abrasive and bond pixels,and the abrasive pixels are filtered to remove the edge abrasives and adherent abrasives.The 3D point cloud is corrected to eliminate the curvature effect and match the abrasive pixels.Finally,the connected domain is processed to extract two-dimensional information such as the area of the abrasive grains,the number of abrasive grains,and three-dimensional information such as the height of the protruding abrasive grains.Next,the laser sharpening quality evaluation index is studied.The buried depth of abrasive grains is obtained from the two-dimensional area of abrasive grains,and the evaluation index of the high and deep distribution of abrasive grains is proposed.By counting the number of abrasive grains,the evaluation index of abrasive grain density is proposed.By extracting the bond surface height information,the evaluation index of bond surface roughness is proposed.Finally,laser sharpening tests were carried out to obtain grinding wheels with different sharpening qualities,and grinding tests were conducted to verify the validity of the sharpening evaluation indexes by the magnitude of grinding force when grinding the workpiece and the surface roughness of the workpiece after grinding.
Keywords/Search Tags:sharpening quality evaluation, diamond grinding wheel, convolutional neural network, 3D inspection, image segmentation, laser sharpening, grinding performance
PDF Full Text Request
Related items