Research On Precision Simulation Method Of Barrel Finishing Based On Digital Twin Surface Model

Improving the surface quality of parts is one of the important ways to improve the performance of parts and prolong the life of parts.As one of the processes to improve the surface quality,barrel finishing technology can effectively improve the surface quality of parts through the micro-grinding effect of granular media on the surface of parts.In order to improve the surface quality of parts,various factors including the processing equipment and parameters,the shape and size of the granular media need to be considered,and it is difficult to provide enough test pieces to find the optimal processing parameters by traditional trialand-error experiments.The use of discrete element method to simulate the process of barrel finishing is mainly used to analyze the mechanism of barrel finishing and optimize the process parameters,which can effectively reduce the cost of trial-and-error.At present,the simulation process treats the part surface as an ideal surface,but the actual surface is not smooth,and there are morphological errors generated by the machining process.In order to improve the accuracy of the simulation,this paper constructs a non-idealized surface model of the part used in the barrel finishing process,and studies the refined simulation of the barrel finishing process based on this model,and finally explores the splicing method for a wide range of surface profiles.The main research contents are as follows:(1)The modeling method of part surface model in discrete element refinement simulation of barrel finishing is studied,and a digital twin surface model containing surface microform information is proposed,and the modeling process is clarified.In the modeling process,the wavelet transform is used to separate the different morphological components of the complex surface morphology of the part.And the reference surface and roughness surface are extracted from the surface morphology.Poisson reconstruction method is applied to complete the 3D reconstruction of the surface morphological data.Finally,the construction of the digital twin surface model is completed by selecting the example object,and the reconstruction accuracy is characterized by the error size of its three-dimensional roughness relative to the actual surface,and the model is constructed and evaluated at four different test positions of the specimen,in which the maximum error does not exceed 5%,so the correctness of the digital twin surface model is verified.(2)In order to verify the superiority of the digital twin surface model containing microscopic morphology data in the discrete element simulation of barrel finishing,and to contrast with the ideal surface model,and the ideal surface model and digital twin surface model with the same number of meshes were imported into the EDEM simulation software.Simulates the machining process in the BJL-LL05 centrifugal barrel finishing machine,finally,the actual wear of the specimen was verified experimentally.In the similarity calculation of the wear images,it was found that the similarity between the wear images of the digital twin surface model with different morphological features and the actual wear images of the ideal surface model increased by 0.6049 and 0.333,respectively,thus confirming the superiority of the digital twin surface model in the discrete element simulation of barrel finishing.(3)In response to the problem that the 3D morphology measuring instrument has high accuracy but small measuring range,which leads to the small size of non-ideal surface model of the part for fine simulation,the image stitching technology is applied to the stitching of 3D morphology of the part,and the stitching method of the large range of 3D morphology is studied,so that the digital twin surface model of larger size is obtained,and the value of the digital twin surface model in the discrete element simulation of barrel finishing is improved.The value of the digital twin surface model in the discrete element simulation of rolling mill is improved.The surface morphology obtained from the stitching is finally constructed as a digital twin surface model and applied to the discrete element refinement simulation of barrel finishing.