AA8030铝合金挤压工艺参数的神经网络分析  

Neural Network Analysis of Extrusion Process Parameters for AA8030 Aluminum Alloy

作　　者：刘振海[1] 

机构地区：[1]盐城工学院工程学院,江苏盐城224051

出　　处：《铸造技术》2015年第12期2978-2980,共3页Foundry Technology

摘　　要：应用神经网络模型分析AA8030铝合金的力学性能对等通道转角挤压变形工艺参数的响应。进行了SEM分析和金相组织观察,对模型进行了实验验证。计算结果表明,利用神经网络模型预测的铝合金拉伸力学性能和试验值之间的误差没有超过1.5%,平均误差0.8%。说明该模型对拉伸力学性能的预测能力及预测精度良好。利用该模型,得到合适的等通道转角挤压工艺条件为:挤压方式为前一道次完成后旋转180°再进入下一道次;挤压道次为4;挤压速度为5 mm/s。A neural network model was used to analyze the the response of the mechanical properties of AA8030 aluminum alloy in equal channel angular extrusion process parameters. Then the SEM analysis and metallographic observation were carried out to verify the model. The calculation results show that the error of predicting the tensile properties of the aluminum alloy by using the neural network model is not more than 1.5%, and the average error is only0.8%. It is showed that the model has good prediction ability and accuracy of the mechanical properties. At last, the most appropriate equal channel angular pressing process parameters were obtained by using the method of equal channel angular pressing process, as follows： the extrusion method is whirling 180° to the next pass after completing one pass, the extrusion pass is 4, and the extrusion speed is 5 mm/s.

关 键 词：等通道转角挤压 强变形 神经网络 

分 类 号：TG376[金属学及工艺—金属压力加工]