摘要
为减少CAE分析时间,提高寻优计算效率,提出基于Kriging代理模型并结合多目标粒子群算法(MOPSO算法)对塑件的注塑成型质量进行多目标优化。以塑件的翘曲变形量、缩痕指数为优化目标,以影响塑件成型质量的模具温度、熔体温度、保压时间、保压压力、注射时间、冷却时间等注塑工艺参数为试验影响因素,应用最优拉丁超立方试验设计方法结合模流分析(MFI分析)建立分析样本,基于Isight参数优化软件构建优化目标与影响因素之间的Kriging代理模型,基于MOPSO算法在代理模型内进行全局寻优,获得了一组使塑件翘曲变形量和缩痕指数最小的最优工艺参数组合并给出了优化目标的预测值。结果表明,Kriging代理模型的预测值与模拟试验结果基本吻合,优化后的翘曲变形量降低15. 3%、缩痕指数降低19. 7%,本文提出的方法能有效、快速实现注塑成型质量的多目标优化,为工程实践提供了有益的参考价值。
In order to reduce the computer aided engineering( CAE) analysis time and improve the optimization calculation efficiency,a multi-objective optimization of the injection molding quality of plastic parts based on the Kriging agent model and a multiobjective particle swarm algorithm( MOPSO algorithm) were presented. Taking the warpage deformation amount and shrinkage index of the plastic parts as optimization targets,the injection molding process parameters,such as,the mold temperature,melt temperature,holding time,holding pressure,injection time,cooling time,and so on,which could affect the molding quality of the plastic parts,were used as test influence factors. Applying the optimal Latin hypercube test design method combined with mold flow analysis( MFI analysis) to establish analysis samples,based on the Isight parameter optimization software to construct a Kriging agent model between the optimization target and the influencing factors,and based on the MOPSO algorithm to perform a global analysis within the agent model through optimization,a set of optimal process parameter combinations that minimize the warpage deformation and shrinkage index of the plastic parts were obtained and the predicted values of the optimization targets were given. The simulation test results show that the predicted value of the Kriging agent model is basically consistent with the simulation test results. The amount of warpage and deformation after optimization reduces 15. 3%and the shrinkage index reduces 19. 7%. The method proposed in this paper can effectively and quickly achieve the quality of injection molding. Multi-objective optimization has useful reference value for engineering practice.
作者
季宁
张卫星
于洋洋
侯英洪
JI Ning;ZHANG Wei-xing;YU Yang-yang;HOU Ying-hong(Department of Mechanical Engineering,Tianjin University Renai College,Tianjin 300636,China;Department of Mathematics,Tianjin University Renai College,Tianjin 300636,China;State Key Laboratory of Engine,Tianjin University,Tianjin 300072,China;Tianjin Xinyang Mould Products Co.,Ltd.,Tianjin 300350,China)
出处
《塑料工业》
CAS
CSCD
北大核心
2020年第5期67-71,共5页
China Plastics Industry
基金
天津市教委科研计划项目资助(2019KJ152)。
