摘要
目的解决板料拉深过程中出现拉裂、起皱、拉深不充分等缺陷的问题。方法利用专业分析板料成形的软件Dynaform,研究分析了非轴对称件矩形盒,在几种典型的变压边力下的拉深成形性能,获得了成形效果较好的加载模式,进而利用仿真软件Dynaform获取了样本数据。结果建立了矩形盒拉深成形变压边力网络模型并对其学习训练,最后对神经网络预测结果及仿真结果所得到的变压边力加载曲线进行多项式拟合,获取了最佳压边力控制曲线。结论在板料拉深过程中,通过控制压边力的大小,能够较好地发挥材料的流动性,改善制件的最终成形效果。
In order to eliminate the defects of cracking, wrinkling and inadequate stretching in the stretching process, the tensile forming properties of a rectangular box with a non-axial symmetric rectangular box under the typical variable blank holder force were studied by using the software Dynaform of the professional analysis of sheet metal forming. A more ideal loading mode of forming effect was obtained. Then the sample data was obtained by the simulation software Dynaform. A network model of variable blank holder force in rectangular box drawing forming was established and then studied and trained. Finally, the polynomial fitting of the curve of variable blank holder force was obtained by the neural network prediction results and the simulation results, and the optimal blank holder force control curve was obtained. In the stretching of sheet metal, control on the blank-holder force can better give play to the mobility of materials and improve the final shape effect.
出处
《精密成形工程》
2016年第4期33-37,共5页
Journal of Netshape Forming Engineering
关键词
矩形盒件
变压边力
拉深成形
神经网络
预测
rectangular box parts
variable blank-holder force(VBHF)
drawing forming
neural networks
prediction