注塑件体积收缩率变化的数值模拟优化与预报  

Optimization and Forecast of Numerical Simulation of Volumetric Shrinkage Variation for Injection Molding Products

作　　者：许建文[1] 刘斌[1] 

机构地区：[1]华侨大学机电及自动化学院,福建泉州362021

出　　处：《华侨大学学报（自然科学版）》2010年第3期241-245,共5页Journal of Huaqiao University(Natural Science)

基　　金：福建省自然科学基金资助项目(E0810040);福建省青年创新基金资助项目(2004J033)

摘　　要：运用Moldflow分析软件,结合田口方法的实验设计,对注塑成型过程进行数值模拟计算,得到各个工艺参数对体积收缩率变化的影响次序及最优化的工艺参数组合.利用BP(Back Propagation)人工神经网络对注塑件的体积收缩率的变化进行预测,以最优化的工艺参数组合为基准,通过微调各个工艺参数来安排正交实验,并将结果作为神经网络的样本数据.经过训练后的神经网络能够准确地预测体积收缩率的变化,从而达到以较少的试验实现注塑成型工艺的优化与控制.The influence order of each process parameter on volumetric shrinkage variation of injection molding products and optimum process parameters can be obtained by numerical simulation and calculation of injection molding process with combination of experimental design of Taguchi method and Moldflow software.The volumetric shrinkage variation of injection molding products is predicted by back propagation neural network,in which the arrangement of orthogonal trials by adjusting each process parameter is made on the basis of optimum process parameters and the experimental results are used as the sample data of neural network.The trained neutral network can accurately predict the volumetric shrinkage variation so that the optimization and control of injection molding process could be achieved using fewer experiments.

关 键 词：注塑件 体积收缩率 MOLDFLOW 田口方法 BP神经网络 

分 类 号：TQ320.66[化学工程—合成树脂塑料工业] TP183[自动化与计算机技术—控制理论与控制工程]