Optimization Of Hot Extrusion Die For Aluminum Alloy Cartridge

Aiming at the difficulty of processing 7050 aluminum alloy cartridge at room temperature and the difficulty of forming thin-walled cartridge parts with variable wall thickness at one time,the forming process of 7050 aluminum alloy cartridge and the problem of optimization of the die structure was deeply studied.Firstly,according to the high strength properties of 7050 aluminum alloy and the simplicity of the process,the hot backward extrusion method was adopted.Then,the feasibility of preparing aluminum alloy cartridge by hot backward extrusion was analyzed and verified by DEFORM simulation software.It was found that the traditional backward extrusion method could not produce thin-walled cartridge parts with variable wall thickness.Then a new combined extrusion method was proposed based on the traditional backward extrusion method: back extrusion-stretching method,and the optimal height of inner cavity of die cavity in back extrusion-stretching method was predicted by BP neural network.The stress of die core in back extrusion-stretching method was analyzed by DEFORM software,and the feasibility of preparing aluminum alloy cartridge by back extrusion-stretching method was determined.Under the condition of the optimal height of inner cavity of die cavity,the back extrusion-stretching method of the cartridge was simulated,and the temperature,equivalent strain,equivalent stress,velocity distribution cloud and damage distribution cloud of cartridge after extrusion were analyzed.Then the experimental study on the back extrusion-stretching process of the 7050 aluminum alloy cartridge was carried out,and the cartridge with qualified size and good surface quality was successfully prepared.Finally,the uniformity of load distribution on the surface of the die core profile was taken as the objective function,and the coordinates of nine control points on the surface of the die core profile were taken as the design variables.The optimization of the die core shape was completed by combining numerical simulation,BP neural network and genetic algorithm.And the optimization results of standard genetic algorithm and multi-group parallel genetic algorithm were analyzed.This paper has completed the preliminary design of the scheme,the verification and optimization of the scheme,which effectively reduces the cost of die testing and improves the design efficiency,which effectively reduces the cost of die testing and improves the design efficiency.