Research On 3D Reconstruction Of Porous Aerostatic Bearing Material And Analysis Internal Flow Field

Porous gas hydrostatic bearings are commonly used as key components of ultraprecision machine tools,and are used as bearings for shafts in ultra-precision machine tools due to their advantages of cleanness,low friction,low loss,and high speed.In the past studies,the porous material was generally idealized.However,due to the complicated internal microstructure of the porous material,simply using the fitting model for equivalent characterization does not correctly reflect the properties of the porous material and the internal flow field.Therefore,the porous material is taken as the research object,and the porous material is reconstructed by DCGAN(Deep Convolution AntiGeneration Neural Network)method.Then the material performance parameters are calculated and analyzed according to the obtained reconstruction model,and the reconstructed model is obtained.Morphology and permeability parameters.Finally,the three-dimensional model is used to analyze the three-dimensional flow law of the porous flow field,and the static characteristics of the porous gas static thrust bearing are further obtained.The main research contents of the thesis are as follows:(1)A three-dimensional reconstruction method of porous materials based on deep learning was studied.First,a SEM high-precision electron microscopy system was used to obtain a continuous tomographic image of the porous material,establishing a data set needed to combat the generation training.Secondly,based on the constructed DCGAN anti-generation neural network,the data set is trained to obtain the same intermediate layer image as the actual pore feature.Again,the intermediate layer image is spliced and reconstructed using the slice combination method to obtain a reconstruction model of the porous material.Finally,analyze the influence of training parameters on the training process and training results,and find the best training parameters.(2)The material properties of the porous reconstruction model at the microscale were analyzed.Firstly,based on the watershed algorithm,the pores and the physical boundary are distinguished,and the morphology parameters of the reconstructed model are obtained by statistical methods.Secondly,the permeability of the reconstructed model is obtained by using the numerical analysis method based on process image and the simulation simulation method based on reconstruction model.Finally,the experimental platform for permeability measurement was built,and the permeability obtained by the above two methods was compared and verified to obtain a more accurate measurement method of permeability.(3)The flow law of the flow field in the porous material is studied,and the influence of the three-dimensional internal flow model on the static performance of the bearing is calculated.Firstly,based on the reconstruction model,the flow field simulation analysis is carried out to obtain the flow velocity and pressure distribution of the three-dimensional flow field in the porous material.Secondly,the static characteristics of the bearing under the one-dimensional flow model and the three-dimensional flow model are calculated separately,and compared with the experimental measurements,the accuracy of the threedimensional flow model is verified.Finally,the OBZO software is used to realize the hole plugging model of porous materials,and the variation of the internal flow field before and after plugging and the influence on the static characteristics of the bearing are summarized and analyzed.