基于CAE与RBF神经网络的固定体塑件注塑工艺优化
|
摘要
针对保险杠固定体塑件注塑成型困难的问题,运用CAE仿真分析手段,首先对其采用3个点浇口进行浇注的方案模拟,得出了影响塑件注塑成型困难的主要问题:塑件中间部位设置的加强筋位置处,料流流动发生多次改变和多重叠加,导致注塑时熔体在此区域产生紊流和回旋滞留。将浇注系统优化为采用5个点浇口浇注方案,能有效消除中间部位所产生的紊流和回旋滞留问题。结合CAE仿真手段和RBF神经网络的预测功能,对5点浇注方案进行注塑成型工艺参数的优化。获得塑件注塑较合理的工艺参数组合为:料温(Tθ)=229℃,模温(Ts)=51℃,注塑压力(pI)=43 MPa,注塑时间(ti)=6. 64 s,第一段保压压力(ph1)=62 MPa,第一段保压时间(th1)=9 s,第二段保压压力(ph2)=38 MPa,第二段保压时间(th2)=5. 5 s,第三段保压压力(ph3)=32 MPa,第三段保压时间(th3)=4. 5 s,冷却水进口温度(Tw)=27℃,冷却液流速(Vw)=3. 2 L/min,冷却时间(tc)=18 s。经实际注塑试验,塑件的注塑效果良好,有效地解决了实际生产问题。
Aimed at the problem of difficult injection molding of bumper fixed body plastic part,the mean of CAE simulation analysis was used. First of all,the gate system was simulated by using three point gate. The main problems that affected the difficulty of plastic injection molding were found out as follows:the material flow at the position of reinforcing bars in the middle of the plastic parts had been changed many times and superimposed many times,which led to turbulence and cyclotron retention of the melt in this area during injection. The gate system was optimized to effectively eliminate the turbulence and cyclotron retention in the middle part after adopting the five point gate pouring scheme. Furthermore,the CAE simulation method and the prediction function of RBF neural network were combined. The five-point casting scheme was optimized for the parameter combination of injection molding process. The optimized process parameter combination of injection molding was obtained as follows:Tθ= 229 ℃,Ts= 51 ℃,pI= 43 MPa,ti= 6. 64 s,ph1= 62 MPa,th1= 9 s,ph2= 38 MPa,th2= 5. 5 s,ph3= 32 MPa,th3= 4. 5 s,Tw= 27 ℃,Vw= 3. 2 L/min,tc= 18 s. Through the actual injection test,the injection effect of plastic parts was good,and the actual production problems were solved well.
Aimed at the problem of difficult injection molding of bumper fixed body plastic part,the mean of CAE simulation analysis was used. First of all,the gate system was simulated by using three point gate. The main problems that affected the difficulty of plastic injection molding were found out as follows:the material flow at the position of reinforcing bars in the middle of the plastic parts had been changed many times and superimposed many times,which led to turbulence and cyclotron retention of the melt in this area during injection. The gate system was optimized to effectively eliminate the turbulence and cyclotron retention in the middle part after adopting the five point gate pouring scheme. Furthermore,the CAE simulation method and the prediction function of RBF neural network were combined. The five-point casting scheme was optimized for the parameter combination of injection molding process. The optimized process parameter combination of injection molding was obtained as follows:Tθ= 229 ℃,Ts= 51 ℃,pI= 43 MPa,ti= 6. 64 s,ph1= 62 MPa,th1= 9 s,ph2= 38 MPa,th2= 5. 5 s,ph3= 32 MPa,th3= 4. 5 s,Tw= 27 ℃,Vw= 3. 2 L/min,tc= 18 s. Through the actual injection test,the injection effect of plastic parts was good,and the actual production problems were solved well.
引文
[1]余冬蓉,倪亚坤.注塑制品常见的质量缺陷分析[J].上海塑料,2001(2):30-31.
[2]石波,刘斌.注塑成型过程中制品质量的控制研究现状分析[J].塑料工业,2012,40(10):9-13.
[3]王树勋.基于CAE分析的注塑模具浇口位置研究[J].制造业自动化,2011,33(5):35-36.
[4]王昌,于同敏.注塑模具浇口类型选择专家系统知识处理新方法[J].塑料科技,2004(5):12-16.
[5]付伟,范士娟,张海.注塑模具设计中浇口位置和结构形式的选用[J].工程塑料应用,2007,35(10):60-63.
[6]陈曦.注塑成型模具浇口位置优化算法研究进展[J].塑料科技,2017,45(2):98-101.
[7]李香林,党新安,何国昆,等.工艺参数对某复杂零件注塑件翘曲影响分析[J].塑料工业,2008,36(7):38-41.
[8]茆茂,朱先琦.基于正交试验法的镜框注塑工艺参数优化[J].南昌工程学院学报,2015,34(3):51-54.
[9]刘兆栋,苏文斌,姚海剑.基于数值模拟的注塑模具优化及注塑成型工艺参数快速设置[J].塑料科技,2009,37(1):68-71.
[10]孙宝寿,邓益民,顾伯勤,等.基于DOE的注塑成型工艺参数优化[J].机械制造,2008,46(8):55-57.
[11]孔轶艳,黄力,陆大同.基于CAE和RBF神经网络的注塑工艺优化分析[J].塑料,2017,46(3):121-125.
[12]项小慧,项慧发,朱羽辰,等.凸起偏心类杯状件注塑成型CAE分析及工艺优化[J].塑料,2015,44(1):113-118.
[13]陈哲.基于RBF神经网络的注塑成型工艺多目标优化及评价研究[D].宁波:宁波大学,2011.
[14]赵宏伟,李圣普.基于粒子群算法和RBF神经网络的云计算资源调度方法研究[J].计算机科学,2016,43(3):113-117.
[15]赵新宇,费良军. LM算法的神经网络在灌区地下水位预测中的应用研究[J].沈阳农业大学学报,2006,37(2):213-216.
[16]肖忠良,李智勇.一种改进的RBF神经网络多目标优化算法[J].科学技术与工程,2009,9(21):6379-6384.
[17]陆广华,王匀,张乐莹,等.基于BP神经网络的注塑成型收缩率预测[J].塑料,2017,46(2):86-88.
[18]陈叶娣.基于正交实验和CAE技术的梳子注射成型工艺参数优化设计[J].模具制造,2013,13(5):40-43.
[19]王锦红,何冰强.基于Moldflow的棘架套件浇注系统方案CAE分析[J].塑料,2016,45(1):79-82.
[20]刘荣亮,胡泽豪.注塑件成型工艺参数优化研究[J].机械研究与应用,2014,27(3):53-55.
[2]石波,刘斌.注塑成型过程中制品质量的控制研究现状分析[J].塑料工业,2012,40(10):9-13.
[3]王树勋.基于CAE分析的注塑模具浇口位置研究[J].制造业自动化,2011,33(5):35-36.
[4]王昌,于同敏.注塑模具浇口类型选择专家系统知识处理新方法[J].塑料科技,2004(5):12-16.
[5]付伟,范士娟,张海.注塑模具设计中浇口位置和结构形式的选用[J].工程塑料应用,2007,35(10):60-63.
[6]陈曦.注塑成型模具浇口位置优化算法研究进展[J].塑料科技,2017,45(2):98-101.
[7]李香林,党新安,何国昆,等.工艺参数对某复杂零件注塑件翘曲影响分析[J].塑料工业,2008,36(7):38-41.
[8]茆茂,朱先琦.基于正交试验法的镜框注塑工艺参数优化[J].南昌工程学院学报,2015,34(3):51-54.
[9]刘兆栋,苏文斌,姚海剑.基于数值模拟的注塑模具优化及注塑成型工艺参数快速设置[J].塑料科技,2009,37(1):68-71.
[10]孙宝寿,邓益民,顾伯勤,等.基于DOE的注塑成型工艺参数优化[J].机械制造,2008,46(8):55-57.
[11]孔轶艳,黄力,陆大同.基于CAE和RBF神经网络的注塑工艺优化分析[J].塑料,2017,46(3):121-125.
[12]项小慧,项慧发,朱羽辰,等.凸起偏心类杯状件注塑成型CAE分析及工艺优化[J].塑料,2015,44(1):113-118.
[13]陈哲.基于RBF神经网络的注塑成型工艺多目标优化及评价研究[D].宁波:宁波大学,2011.
[14]赵宏伟,李圣普.基于粒子群算法和RBF神经网络的云计算资源调度方法研究[J].计算机科学,2016,43(3):113-117.
[15]赵新宇,费良军. LM算法的神经网络在灌区地下水位预测中的应用研究[J].沈阳农业大学学报,2006,37(2):213-216.
[16]肖忠良,李智勇.一种改进的RBF神经网络多目标优化算法[J].科学技术与工程,2009,9(21):6379-6384.
[17]陆广华,王匀,张乐莹,等.基于BP神经网络的注塑成型收缩率预测[J].塑料,2017,46(2):86-88.
[18]陈叶娣.基于正交实验和CAE技术的梳子注射成型工艺参数优化设计[J].模具制造,2013,13(5):40-43.
[19]王锦红,何冰强.基于Moldflow的棘架套件浇注系统方案CAE分析[J].塑料,2016,45(1):79-82.
[20]刘荣亮,胡泽豪.注塑件成型工艺参数优化研究[J].机械研究与应用,2014,27(3):53-55.