Research On Design And Performance Of Surface Grinding Tool Based On Random-grid

Ultra-precision machining technology has become one of the essential indicators to evaluate the level of a country’s advanced manufacturing technology,and is widely used in national defense,military and civilian industry.In the ultra-precision machining processes,abrasive machining has unparalleled advantages.Compared with free-abrasive machining,fixed-abrasive machining has the advantages of controllable abrasive distribution and high material removal efficiency.It has gradually become the major development direction in ultraprecision grinding processing technology.The preparation process of traditional fixed abrasive tools has high energy consumption and low flexibility,and it is difficult to balance the composite requirements of material removal rate and surface quality in the abrasive machining process.Therefore,this thesis proposes a random-grid structure abrasive tool based on rapid prototyping technology,and conducts in-depth research and discussion on the performance of this tool through the analysis of abrasive particle motion trajectory,grinding processing experiment and machine learning.The main tasks completed in this paper are as follows:Firstly,the random-grid structure grinding plate is modeled and prepared.Using threedimensional software Pro/E,Rhino and 3D Max,several design methods and modeling ideas of surface grinding plate with spatial structure were proposed.The optimal ratio of resin bond and abrasive was determined to be 11.7 by the light curing rapid prototyping experiment method.Finally,the random grid spatial structure grinding plate was prepared by using resin and alumina mixture.Secondly,according to the fixed abrasive grinding process,the mathematical model of fixed abrasive plane grinding is established,and the trajectory equation of abrasive particles consolidated on the grinding plate is obtained.The trajectory simulation of grinding plates with different surface morphology and different spatial structure is carried out by using MATLAB.The area ratio of grinding trajectory between different grinding plates is compared,and the abrasive layer by layer of grinding plates with different spatial structure is analyzed.The results show that the grinding path of the grinding plate with random grid space structure is more uniform,and the coverage area of abrasive particles is 99.89%,which can achieve better high efficiency and low damage grinding.Then,in order to further evaluate the grinding performance of the grinding plate with random grid spatial structure,and quantify the influence of grinding parameters and the surface morphology of the grinding plate on the grinding performance.The orthogonal experiment of fixed abrasive grinding was carried out with 4S double-sided grinder.The surface roughness and material removal rate of the workpiece were measured,and the main influencing factors of grinding experiment were analyzed.The results show that in the parameters studied,the concentration of grinding fluid has the greatest influence on the surface roughness,followed by the morphology of the consolidated abrasive on the grinding plate,the motor speed has the greatest influence on the material removal rate,and the concentration of grinding fluid has the second influence on the material removal rate.Finally,BP neural network algorithm is used to establish the prediction model between grinding parameters and grinding results.Python programming is used to train the network with experimental data,so as to predict new experimental samples.Finally,the super parameters of the neural network are optimized to further enhance the prediction ability of the network.