本研究透過射出成型模具內感測系統，結合實驗設計法之品質最佳化理論，建立製程參數的品質特性函數，找尋最佳製程參數以達到智慧製造的目標。首先透過安裝模具內感測系統，線上擷取熔膠之溫度與壓力數據，監控射出成型製程中試片之成型狀態。將溫度、壓力數據轉換成比容，根據比容變異程度作為判定品質特性之依據。接續以田口方法與二階反應曲面法逐步解析製程參數對於品質特性之影響，其中針對二階反應曲面法之中央合成設計與Box-Behnken設計進行比較，結果顯示此兩種反應曲面法能夠建立射出成型製程參數與體積收縮率計算值之迴歸模型。本研究發現保壓階段對於體積收縮之影響最為顯著，以及由於Box-Behnken設計之設計點分佈狀況，對於本案例之優化能力比中央合成設計弱。實驗最終以中央合成設計將體積收縮率計算值降低至1.22%，其最佳製程參數之總保壓時間比Box-Behnken設計節省0.4秒，提升生產效率與試片品質。;This research expects to establish a regression model between the injection molding process parameters and quality characteristics through the design of experiment. First of all, an in-mold sensing system was equipped to obtain the temperature and pressure of the melt. The melt state of the molded part was being monitored during the injection molding process. The analysis of quality characteristics was referring to the variation of specific volume which was calculated by the temperature and pressure data. Then the influence of control factors on quality characteristics was analyzed by Taguchi method and second order response surface methodology. The results showed that the packing stage had the most significant effect on the volumetric shrinkage. The statistical predictability of the Box-Behnken design (BBD) is less accurate than that of central composite design (CCD) due to the distribution of design points in the BBD. The experimental results showed the CCD reduced the calculated volumetric shrinkage to 1.22%. The optimum total packing time of CCD was reduced by 0.4 seconds compared with that of BBD.
