基于正交试验法和神经网络的上饰板注塑工艺优化  

Optimization Analysis of Injection Molding Process of Decorative Plate Based on Orthogonal Test Method and Neural Network

作　　者：盘承军[1] 李炜[2] 杨达飞[1] PAN Chengjun;LI Wei;YANG Dafei(Department of Information Engineering, Liuzhou Vocational ＆ Technical College, Liuzhou, Guangxi 545006, China;School of Computer and Communication Engineering, Guangxi University of Science and Technology, Liuzhou,Guangxi 545006, China)

机构地区：[1]柳州职业技术学院电子信息工程系,广西柳州545006 [2]广西科技大学计算机与通信工程学院,广西柳州545006

出　　处：《塑料》2018年第3期92-97,共6页Plastics

基　　金：广西教育厅科研课题(2017KY1054)

摘　　要：针对产品结构复杂而导致注塑后熔接线较多、缩痕明显及翘曲变形大的注塑质量问题,通过CAE辅助分析,对浇口位置浇注系统进行了优化设计,获得了良好的注塑充填效果。运用正交试验方法优化了成型工艺参数:T1=60℃,T2=270℃,V=40 cm3/s,pi=80 MPa,ti=7 s,ts=16 s(模具温度T1、熔计温度T2、注射速率V、保压压力pi、保压时间ti、冷却时间tc)。进一步运用SOM神经网络的预测功能,对翘曲变形进行了参数控制优化,获得的最佳工艺参数为T1=58℃,T2=262℃,V=35 cm3/s,pi=75 MPa,ti=6 s,tc=18 s。所优化的工艺参数经实际注塑检验,翘曲量可以控制在0.89 mm以下,能够较好地满足生产要求。Aimed at the quality of the injection molding problems such as more weld lines, obvious shrinkage marks and large warpage after injection molding appeared because of the complexity of the product structure, Through the CAE aided analysis, the gate position and gating system were optimized,good injection filling effect had been obtained. The parameters of forming process were optimized by orthogonal test,the optimum process parameters T1 = 60 ℃ , T2 = 270 ℃ , V = 40 cm3/s ,pi = 80 MPa, ti =7 s, t c= 16 s were obtained. Further more, the SOM neural network was used to optimize the parameters of warping deformation,the optimum process parameters T1=58 ℃, T2 =262℃, V=35 cm3/s, pi =75 MPa, ti =6 s,t c=18 s were obtained. The optimized technological parameters were ehecked by the actual injection molding, the warpage was controlled below 0.89 mm and the production requirements could be well met.

关 键 词：SOM神经网络 正交试验 注塑成型 CAE分析 优化 

分 类 号：TQ320.66[化学工程—合成树脂塑料工业]