Research On Structure Optimization And Parameter Optimization Of Complex Aluminum Extrusion Dies

Nowadays,resources and energy issues have become increasingly prominent,aluminum profiles as a key factor in transportation,construction,shipbuilding and other areas of lightweight,get more and more people's attention.The alu minum products are often obtained by extrusion molding.With the increasing variety of alum inum extrusion products and com plex cross-sectional shape,extrusion die structure design has become increasingly dif ficult,in addition the choice o f extrusion process parameters of extrusion process also can not be ignored.The traditional production methods about test mode or maintenance were can not meet the needs of the produc tion design.With the continuous development of the finite element simulation technology matures,the extrusion die extrusion process optim ization and parameter optimization provides a new way of thinking.This paper briefly introduces th e basic principles of extruded aluminum,the main characteristics of aluminum extrusion,and the finit e element simulation software nam ed HyperXstrude.Then using t his software to sim ulate the extrusion process about t he complex hollow aluminum profiles with a cantilever.In order to obtain products of acceptable quality extruded profiles,the profiles corresponding m old structure was m odified design.We redesigned the distribution of m old vent hole,the corresponding shunt br idge and the welding cham ber,so that the uniform ity of the velocity distributi on profile extrusion get a substantial increase.According to the simulation results,taken a series of structural optimization measures like adjust the size of the vent holes,set a choke blocks where th e flow rate too fast and twice with length adjustment.So exit velocity profiles become m ore stable.Profile deformation degree small and the strength of the mold are assured.Then study the af fect about the pa rameters of profile extrusion process for metal flow rate uniformity and metal mold strength.Got the t influence of extrusion process pa rameters about m old preheating temperature,preheating tem perature of the billet,billet diameter and extrusion speed to the square deviati on of velocity of extrusion profil e cross-sectional.And the m old suffered the greatest stress with th e variation of the process parameters.Finally,make use of the artificial intelligence algorithm s to optimize the process parameters.Use MATLAB software to establish BP neural network prediction model r ealized by input parameter values within a range of forecasts for SDV values.And verify the reliability of its predictive power.Make the BP neural network combined with genetic algorithm to search the best process parameters.Through simulation and test verification HyperXtrude optim ized reliability of the results.Obtained the optimum parameter value of the profiles tha t extrusion speed is 2mm/s,mol d preheating temperature is 453?,billet preheating temperature is 479?,billet diameter is 261 mm.