Analysis And Prediction Of Residual Stress In Milling Of Aluminum Alloy Annular Thin-Walled Parts

Thin-walled parts are a kind of parts with light weight and excellent bearing capacity,which are widely used in various fields.Different from the traditional thin wall parts with single shape and relatively fixed,the shape of modern thin wall parts tends to be diversified and complicated,among which the circular thin wall parts are widely used in military aviation and other fields.The residual stress is one of the main reasons for the failure of annular thin-walled parts.Therefore,it is very important to understand and master the source and distribution of residual stress,as well as how to control and utilize residual stress to improve product quality.Therefore,in this paper,a certain aviation ring parts as the research object,the first research and analysis of residual stress generation principle;Then,the model of annular thinwalled parts was simplified,and the milling simulation model of annular thin-walled parts was established by using ABAQUS software to study the distribution of residual stress on the machined surface,sub-surface and deep layer,in order to provide theoretical support for improving and reducing the machining quality of annular thin-walled parts and reducing workpiece deformation.Finally,combined with neural network,the prediction model of maximum residual stress value is established.The specific research contents are as follows:1.7075-T651 aluminum alloy annular thin-wall simulation model establishment.Based on the mechanism of metal cutting,the heat generation and transfer process in the cutting process are explored.The ABAQUS simulation software and the key technologies used are briefly introduced.The simulation model of annular thin-walled parts is established,which lays a foundation for the simulation in the following chapters.2.Study on the residual stress distribution of annular thin-walled parts and the influence of process parameters on them.Firstly,the formation mechanism of residual stress in the cutting process under ideal conditions is explored from the aspect of numerical theory.The variation of residual stress is explored by simulating the unloading process of temperature.Secondly,the effects of different process parameters on the residual stress distribution of machined surface,sub-surface and deep surface were studied.Finally,aiming at the maximum residual stress and the minimum deformation of the workpiece,the optimal process parameters of annular thinwalled parts were studied.3.Establishment of residual stress prediction model for annular thin-walled parts.Combined with intelligent algorithm,BP and GRNN neural network algorithm,respectively take different milling speed,the radial cutting depth and the axial cutting depth and each tooth feeding process parameters combination of simulation,establish the BP neural network and GRNN neural network forecasting model,and by comparison with the results showed that the GRNN model,the maximum residual stress prediction more accurate.