点击次数：

论文类型：期刊论文

发表时间：2015-04-01

发表刊物：INTERNATIONAL JOURNAL OF AUTOMOTIVE TECHNOLOGY

收录刊物：SCIE、EI、Scopus

卷号：16

期号：2

页面范围：329-337

ISSN号：1229-9138

关键字：Hot forming; B-pillar; Geometry feature; Process parameter; Data mining

摘要：In the process of hot forming, many design variables have effects on the final results in different and complex ways, such as geometry feature and forming process parameters. It is difficult to understand the relationship between design variables and results, which is very important to guide the design. In this paper, Data Mining (DM) was introduced to explore the influence of the parts geometric feature and the hot forming parameters on hot forming results of an automobile B-pillar model, and the optimum parameter ranges were determined. Firstly, a series of variable parameters were chosen and 100 groups of experimental data were generated with super Latin method, then the FEM ananlysis results were calculated respectively. Secondly, analysis and evaluation of simulation results were carried out by making full use of the Decision Tree (DT) algorithm. Finally, a series of B-pillar hot forming rules were refined, such as the initial temperature of the sheet metal should be controlled between 720 A degrees C to 800 A degrees C. The fillet radius is recommended to be bigger than 10mm and the height gradient should be controlled under 67mm, etc. A real B-pillar model was designed to testify the rules and the result shows that the rules are correct and effective.