翹曲變形一直都是塑膠射出成型產品開發最大困擾之一,傳統上利用試誤法(trial-and-error)嘗試延長保壓時間,加大保壓大小或其他可能之方式處理,但常常無法有效改善品質與精度。本研究透過Moldex3D模流分析軟體進行框架型產品實驗設計法(DOE)及三維體積收縮補償法(3DVSCM)探討其對翹曲變形之優化結果,其中實驗設計法(DOE)使用L18(21×37)直交表配置,探討其對S/N比與翹曲變形之影響,最後再由變異數分析(analysis of variance)評估控制因子之影響力;三維體積收縮補償法(3DVSCM)則是先經由模擬分析原始設計產品後得知翹曲變形的區域位置,最後再透過補償修正產品模具的方式進行改善。研究結果顯示,原始設計之產品平坦度偏移量達0.481mm,利用實驗設計法優化後之最佳化設計,平坦度偏移量降為0.129mm,改善率達73% ;使用三維體積收縮補償法修正後之產品分析結果,平坦度偏移量降為0.061 mm,改善率達87%。 Warpare is one of the crucial problems in injection molding product development. Conventionally, people usual apply packing pressure, packing time, or many other operation conditions based on trial-and-error way to improve the warpage problem. In this study, Moldex3D analysis software was used to study the optimal results of warpage by frame-type product using both design of experiment (DOE) method and 3D volume shrinkage compensation method (3DVSCM). Specifically, L18 (21×37) orthogonal array of Taguchi method was introduced to deploy the factors. Furthermore, using analysis of variance investigates the significance of the factors that are displayed on the S/N ratio and warpage. Moreover, 3D volume shrinkage compensation method (3DVSCM) is applied to compensate the deviation of warpage for the original design. The results show that the flatness deviation of the original design is 0.481mm. That deviation of the DOE optimized part is 0.129 mm. The improvement based on DOE is about 73%. Meanwhile, when 3DVSCM is applied, the flatness deviation of the optimized part is only 0.061 mm. The improvement is up to 87%.