| The manufacturing industry is main body of national economy and has important strategic significance to the development of the country.With the progress and development of the manufacturing industry China’s science,technology,and comprehensive national strength are constantly improving.As an important part of the manufacturing industry,thin-walled parts are widely used in various fields due to their lower height and high performance.However,due to its large size and poor stiffness,the thin-walled deformation is almost inevitable during machining,resulting in the decrease of workpiece accuracy,which cause serious influence on precision and surface quality of workpiece.Therefore,In order to reduce the deformation of milling thin-walled parts,the finite element simulation is used to analyze the milling deformation.And then,optimizing the clamping layout and compensating machining error.It plays an important role in improving the machining accuracy,surface quality and machining efficiency of thin-walled parts.Based on the finite element simulation technology,this dissertation analyzes and predicts the deformation caused by clamping layout and cutting force in the machining process of thin-walled parts.The method of combining BP neural network and genetic algorithm is adopted to establish the optimization model of the clamping of thin-walled parts and realize the optimization of the clamping layout of thin-walled parts.The influence of cutting force on thin-walled parts in the milling is analyzed,and the cutter deformation is predicted.The deformation is compensated actively to reduce the machining error of thin-walled parts.The research contents are as follows :In this dissertation,aiming at the thin-walled parts for torpedo.Using ABAQUS simulation software,the finite element simulation model of workpiece deformation is established for the clamping process of thin-walled parts.Based on this model,the deformation caused by clamping layout in the milling process is analyzed.Based on the simulation model,the neural network is combined with the finite element,and the data provided by the finite element simulation are used to train the neural network.The prediction model between clamping layout parameters and machining deformation is constructed by trained BP neural network.Taking the minimization of machining deformation as the objective function,the optimization of clamping parameters is completed by using the global optimization ability of genetic algorithm,and the optimal clamping layout is obtainedThe simulation analysis of the side wall milling deformation of thin-walled parts was carried out,and the milling force model of the end mill was established.The end mill was divided into several cutting units along the axial direction and the cutting force of each cutting unit was established.Then,the cutting force model of the end mill in the cutting process was established by summation.The tool yield deformation caused by cutting force in the machining process is predicted,and the machining deformation value of the discrete point is close to the real machining deformation shape.According to the obtained deformation contour,the cutting parameters are corrected,and the error compensation of milling is realized,which provides a basis for improving the machining accuracy of thin-walled parts. |
