论文部分内容阅读
现有企业生产大口径三通管要通过多道次翻边工艺,本文根据冲压变形规律设计的预制孔管件可进行一次翻边生产等径三通管,同时应用Deform软件对翻边工艺进行数值模拟。然后,利用神经网络结合遗传算法对大口径管件预制孔孔形及凸模运动速度进行优化,可得到成形质量较佳的三通管。最后,通过对大口径三通管多级翻边扩孔工艺研究得出结论:以冲压变形规律设计的预制孔管件一次成形特定高度的三通管是可行的,但需对工艺参数进行优化。
The existing enterprises to produce large-diameter tee pipe through multi-pass cuffing process, this article according to the rules of stamping deformation prefabricated pipe fittings can be a flange production of tee, while the application of Deform software on the flanging process values simulation. Then, by using neural network and genetic algorithm to optimize the hole shape of the prefab hole and the movement speed of the punch, the three-way pipe with better forming quality can be obtained. Finally, through the research on the large-diameter three-way pipe multi-step flanging and reaming process, it is concluded that the prefabricated pipe fittings designed by the stamping deformation method are feasible to form the three-way pipe with a certain height at one time, but the process parameters need to be optimized.