摘要
针对塑件潜在的缺陷问题,结合CAE辅助分析,对塑件基于经验设定的浇注系统进行了流动仿真分析,CAE分析结果显示,塑件常发生的8种缺陷中,在材料所推荐的工艺参数下,4种潜在的质量缺陷发生的概率比较高,对注塑所需的参数进行多种参数正交寻优,优化组合后,通过进一步结合DPA-BP神经网络预测方法,寻优获得了塑件的较佳注塑工艺参数,塑件的最佳注塑成型工艺参数为:料温235℃,模温55℃,保压分两段保压(为85 MPa-6 s,45 MPa-5 s),冷却时间60 s。实际生产验证表明,通过神经网络寻优的工艺参数,能有效地保证塑件的成型性能,有效降低了模具设计周期,降低了模具生产的潜在风险。
Aimed at the production of potential defects of plastic parts,the gating system of plastic parts based on experience was carried out by combining CAE auxiliary analysis method simulation analysis,the analysis results showed that the often occur eight kinds of defects of plastic,the plastic parts under the recommended process of the material parameters,the four potential quality defects occur probability were quite high,After orthogonal optimization and combination of various parameters needed for injection molding,by further combining with the DPA-BP neural network forecasting method,the optimization for the better injection molding process parameters of plastic parts was obtained,the best injection molding process parameters of plastic parts were plastic temperature 235℃,mould temperature 55℃,subsection pressure maintaining two pieces,for 90 MPa-16 s,70 MPa-9 s,cooling time 60 s.The optimal process parameters through neural network could effectively guarantee the molding performance of plastic parts,effectively reduce the mold design cycle,and reduce the potential risk of mold production.
作者
邓其贵
姜思佳
罗永有
DENG Qigui;JIANG Sijia;LUO Yongyou(Department of Mechanical Engineering,Liuzhou Vocational Technical College,Liuzhou,Guangxi 545006,China;Liuzhou City Vocational College,Liuzhou,Guangxi 545036,China)
出处
《塑料》
CAS
CSCD
北大核心
2019年第6期100-105,共6页
Plastics