方盒形件拉深成形变压边力加载规律优化  

Optimization on loading law of variable blank-holder force in deep drawing of square box

作　　者：张效林 李奇涵[2] Zhang Xiaolin;Li Qihan(Guangdong TAYO Motorcycle Technology Co.,Ltd.,Jiangmen 529000,China;College of Mechanical and Electrical Engineering,Changchun University of Technology,Changchun 130012,China)

机构地区：[1]广东大冶摩托车技术有限公司,广东江门529000 [2]长春工业大学机电工程学院,吉林长春130012

出　　处：《锻压技术》2019年第11期68-74,共7页Forging & Stamping Technology

基　　金：吉林省省级经济结构战略调整引导资金专项项目(20141131)

摘　　要：选择方盒形非轴对称件为模拟研究对象,分析了方盒形件拉深成形的工艺特点和常见的失效形式及判定标准,通过专业CAE分析软件DYNAFORM,研究分析了方盒形件在不同典型压边力加载模式下的拉深成形性能和极限拉深比(LDR),确定了V型或者类似于V型的变压边力加载状态下坯料的成形效果和LDR最优。建立了方盒形件成形过程中变压边力加载规律的径向基(RBF)神经网络智能预测模型,并完成预测模型的训练和性能检验,对比发现预测结果与模拟结果吻合较好,而且RBF神经网络预测变压边力加载时板料拉深成形质量更好,也更趋近实际生产状态。最后对神经网络预测结果进行多项式拟合优化,获得了成形效果较为理想的变压边力加载曲线。For square box parts of non-axisymmetry,the process characteristics and common failure modes as well as judgment criteria in deep drawing of square box parts were analyzed.Then,the deep drawing performance and the limit drawing ratio(LDR)of square box parts under different loading modes of typical variable blank holder force(VBHF)were studied and analyzed by the professional CAE simulation software DYNAFORM,and the best forming effect and LDR of blank under the V-type or similar V-type of VBHF loading were confirmed.Furthermore,the RBF neural network intelligent forecasting model of VBHF loading law in the deep forming process of square box parts was established,and the training and performance test of the forecast model were completed.The predicted results are in good agreement with the simulation results,and the RBF neural network predictes that the sheet drawing quality with VBHF loading is better and closer to the actual production state.Finally,the neural network prediction results were optimized by the polynomial fitting,and the VBHF loading curve with ideal forming effect was obtained.

关 键 词：方盒形件 变压边力 拉深成形 加载模式 智能预测 

分 类 号：TH16[机械工程—机械制造及自动化] TG386[金属学及工艺—金属压力加工]