In the present study, a systematic methodology has been presented to determine optimal injection molding conditions for minimizing warpage and shrinkage in a thin wall relay part using modified particle swarm algorithm (MPSO). Polybutylene terephthalate (PBT) and polyethylene terephthalate (PET) have been injected in thin wall relay component under different processing parameters: melt temperature, packing pressure and packing time. Further, Taguchi’s L₉ (3²) orthogonal array has been used for conducting simulation analysis to consider the interaction effects of the above parameters. A predictive mathematical model for shrinkage and warpage has been developed in terms of the above process parameters using regression model. ANOVA analysis has been performed to establish statistical significance among the injection molding parameters and the developed model. The developed model has been further optimized using a newly developed modified particle swarm optimization (MPSO) algorithm and the process parameters values have been obtained for minimized shrinkage and warpage. Furthermore, the predicted values of the shrinkage and warpage using MPSO algorithm have been reduced by approximately 30% as compared to the initial simulation values making more adequate parts.

中文翻译：

---

使用改进的粒子群算法对PBT / PET零件的注射成型过程进行建模和优化

在本研究中，已提出了一种系统方法论，以使用改进的粒子群算法（MPSO）确定最佳的注塑条件，以使薄壁继电器零件中的翘曲和收缩最小化。聚对苯二甲酸丁二酯（PBT）和聚对苯二甲酸乙二酯（PET）已在不同的加工参数（熔体温度，填充压力和填充时间）下注入薄壁继电器组件中。田口的L ₉（3 ²）正交阵列已用于进行仿真分析，以考虑上述参数的相互作用影响。已经使用回归模型根据上述过程参数开发了用于收缩和翘曲的预测数学模型。进行了方差分析，以建立注塑参数和开发模型之间的统计显着性。使用新开发的改进的粒子群优化（MPSO）算法对开发的模型进行了进一步优化，并获得了工艺参数值，以最大程度地减少收缩和翘曲。此外，与使用更合适的零件的初始模拟值相比，使用MPSO算法的收缩和翘曲的预测值已减少了大约30％。