| 摘要: | 本研究利用模內感測技術蒐集成型階段模內各處比容之間的差異作為反應值，並以實驗設計法之方式設定製程參數，建立參數與反應值間的關係。以感測器跟澆口之間的距離區分為近處、中間段以及遠處，蒐集三處的比容資料，計算比容值凝固時到平衡時的改變量用於預估體積收縮率；另一方面，取出三處凝固時的比容，透過公式來量化三處比容間的非均勻性，用作預估產品可能發生翹曲之指標數值。對於凝固時的比容取值方式，本研究不同於一般單看溫度或壓力來估計澆口凝固的方式，而是透過壓力曲線變化來找出三處感測點個別的凝固時間。將比容的改變量與非均勻性的組合作為反應值進行二階段實驗，先以田口法篩選重要因子，再利用反應曲面法建立數學模型。以料溫、模溫、射速、保壓壓力、保壓時間和冷卻時間作為實驗的六項因子，經過田口法實驗並分析結果後選出保壓壓力和保壓時間為最重要的兩項因子。接著將保壓改為三段，以三項保壓壓力和總保壓時間做四因子的反應曲面法實驗，得到的最佳組合預測值為0.0088 ，將最佳參數經過實驗驗證的結果為0.0086 ，預測誤差為2.3%。反應曲面法的中心點組別實驗結果為0.0125 ，優化結果與之相比改善了31.2%。顯示實驗設計法對於射出成型之非線性問題能得到有效的預測結果並改善產品品質特性。;This study used the in-mold sensor to measure the specific volume difference in molding process as the response. The process parameters in experimental design were set to establish the relationship between the parameters and the response value.
The mold part was divided into three regions such as near, middle and far positions. The specific volume data at three positions were collected, and the variation of specific volume from freezing to equilibrium is used to estimate the volumetric shrinkage rate.
Additionally, the three positions were compared with one another. By measuring the specific volume of the three positions at freezing point, we can use the formula to quantify the non-uniformity of the specific volume, represents as an index value for predicting the possible warpage of the product.
Regarding the method of measuring the specific volume at freezing point, this study is different from the general method of estimating the freezing state of the gate through temperature or pressure. Instead, it finds the individual freezing point of the three sensing locations through the variation of pressure line.
The combination of the variation and the non-uniformity of specific volume is used as the response to conduct a two-stage experiment. First, Taguchi method is used to screen the important factors. After that, Response surface method is used to establish a mathematical model. The melt temperature, mold temperature, filling velocity, packing pressure, packing time and cooling time were selected as the control factors in the experiment. After Taguchi method experiment and analysis, packing pressure and packing time were selected as the two most significant factors.
Then adjust the packing pressure to three stages, and run a four-factor Response surface method experiment with packing pressure of three stages and total packing time. The best combination predict value is 0.0088 , and the best parameter set verified through experiment is 0.0086 , the error of prediction is 2.3%.
The experimental result of the center point group through Response surface method is 0.0125 , which is an improvement of 31.2% compared with the optimized result. Result shows that the experimental design method can effectively predict the nonlinear problem of injection molding and improve the product quality characteristics eventually. |
