Research On Milling Deformation Control Of Thin-walled Parts Based On Digital Twin

Thin-walled parts are widely used in aerospace,automobile,shipbuilding and other fields due to their compact structure and light overall quality.Due to the high removal rate of thin-walled parts,the stiffness of processed parts is reduced,resulting in the reduction of machining accuracy and the difficulty of ensuring machining quality.The machining accuracy is reduced and the machining quality is difficult to guarantee.With the integration and application of new generation technologies such as Internet of things,artificial intelligence and digital twin with manufacturing industry.Therefore,a twin platform for rapid simulation,accurate and efficient prediction and control of machining deformation of thin-walled parts is constructed to improve the machining accuracy of thin-walled parts.The specific research content is as follows:（1）The five functions of virtual simulation,communication,thin-walled parts machining deformation prediction and control,service and application are analyzed.Use UG to draw the machine tool model,render the model through 3Ds Max,and then import it into Unity 3D to establish the virtual machine tool processing scene.This system is modularized,and the twin system of thin-walled parts processing process is built.（2）The finite element model of titanium alloy milling was established in ABAQUS by using finite element simulation technology.The parameters of materials,failure criteria and tool-workpiece friction model were set,and the milling force and deformation data were obtained by simulating the milling process.The simulation results are consistent with the experimental results,so the simulation deformation data can be used as the data support source for the prediction model of machining deformation of titanium alloy thin-walled parts.（3）For thin-walled parts processing,the processing data are firstly classified and a framework for monitoring data acquisition is established.The Convolution Neural Network（CNN）and Gated Recurrent Unit（GRU）are used to train and fit the machining deformation model of thin-walled parts,and the machining deformation prediction framework of twin data is constructed.By comparing the predicted results with the actual deformation requirements,the machining deformation of the thin-walled parts was optimized by adjusting the spindle speed and the feed ratio.（4）According to the function of twin system for prediction and control of thinwalled parts machining deformation,a server and a client are created.The communication between physical machine tool and virtual machine tool is completed by OPC and API8070 protocol,and the real-time data transmission requirements are met.The feasibility and accuracy of the twin system are verified by the comparison between the system’s prediction of thin-wall parts machining deformation and the simulation and experimental data.According to the prediction results of machining deformation of thin-walled parts,the spindle speed and feed rate in the machining process of thin-walled parts are adjusted to reduce the deformation in the cutting in and cutting out stages of thin-walled parts.