基于可控拉深筋的高强度板拉深性能优化及预测

合理地调节拉深筋阻力可以有效地改善板料拉深的成形质量,为此提出可控拉深筋技术以提高高强度钢板成形性能.以JAC590Y高强度钢板为研究对象,首先通过正交试验设计和数值模拟软件Dyanform相结合,研究三种不同类型的拉深筋运动轨迹对平底盒形件成形性能的影响,以极限拉深深度评判成形性能优劣,确定了优化的拉深筋运动轨迹类型为上升—静止—下降路线,并通过极差分析得到其主要影响因子H1和H2,同时结果表明三类可控拉深筋运动轨迹均能提高高强度钢板的成形性能.然后基于优化的可控拉深筋运动轨迹类型,通过模拟试验数据建立其各个因子与极限拉深深度的GA-BP(遗传算法-反向传播)神经网络预测模型,检验表明该模型能够较好地预测因子对极限拉深深度的影响,预测值与测试值的误差在5%以内.

Formability Improvement and Prediction of High Strength Steel Based on Controllable Drawbead

The drawing quality could be improved effectively by a reasonable drawbead restraining force,so the controllable drawbead technology was put forward to improve the formability of high strength steel(HSS).The JAC590Y HSS was selected as an example and three different types of drawbead trajectories to the formability effects were investigated by orthogonal tests and numerical simulation.The optimal type trajectory,ascenthaltdescent,was determined according to the limit drawing depth.The main influence factors H1,H2 were obtained by range analysis and the numerical simulation results indicate that the three types of drawbead trajectories can all improve the formability of HSS.Then the genetic algorithm back propagation(GA BP)neural network of controllable drawbead trajectory factors limit drawing depth was constructed by test data of the optimal type trajectory.The estimated values of the limite drawing depth were close to the simulation data and the errors were within 5%.