标题：Optimal Design of Second-Step Welding Chamber for a Condenser Tube Extrusion Die Based on the Response Surface Method and the Genetic Algorithm
作者：Xuemei Sun;Guoqun Zhao;Cunsheng Zhang
作者机构：[Sun, X] Key Laboratory for Liquid-Solid Structural Evolution, Processing of Materials of Ministry of Education, Shandong University, Jinan, Shandong 更多
通讯作者地址：[Zhao, GQ]Shandong Univ, Minist Educ, Key Lab Liquid Solid Struct Evolut & Proc Mat, Jinan 250061, Shandong, Peoples R China.
来源：Materials and Manufacturing Processes
关键词：Aluminum alloy extrusion process;Box-Behnken experimental design;Genetic algorithm;Response surface method;Second-step welding chamber.
摘要：Second-step welding chamber of porthole die plays an important role in controlling metal flow and improving the quality of the extrudate. In this article, using a condenser tube aluminum alloy extrusion profile as the example, the shape and the height of second-step welding chamber of the extrusion die are selected as the design variables. Standard Deviation of the Velocity field in bearing exit (SDV) is used as the objective function. By combining Box Behnkcn experimental Design (BBD) with Response Surface Method (RSM). a prediction model for SDV is established. The model is optimized by means of genetic algorithm (GA), and the optimal extrusion die for the condenser tube profile is obtained. In comparison with the initial die design, a more uniform velocity distribution in the cross-section of the profile in the bearing exit is achieved by using the optimal die design. The SDV is reduced from 4.37 nim/s of the initial die design to 0.29 mm/s of the optimal die design. The distortion of the cxlrudate is controlled effectively. Thus the extrudate quality is improved significantly, and its shape and dimension are satisfactory. Finally, the metal flow pattern and die strength for the optimal scheme are analyzed. The optimization strategy proposed in this article has practical meaning in improving the die design, shortening the production development cycle, and reducing the production cost.