| 摘要: | 本文研究，著重於平板 L型彎曲製程中折疊區有孔之實驗已完成。應用反應曲面法(RSM)，提出有系統且完整的實驗設計與分析方法。實驗設計採用中央合成設計法(CCD)，設計 4個獨立變數包含工件厚度、模具彎角半徑、孔徑和孔數作為設計參數。同時已完成探討，在 L型彎曲製程中對於回彈角及圓孔變形率的研究。頂面切線方向沿上表面軸向變形，底面軸方向沿下表面切線方向變形，其變數名稱分別為 RD_A、RD_T、Rd_a、Rd_t，結果顯示，回彈角會隨著孔之尺寸、模具彎角半徑、孔數的減小及工件厚度的增大而減小。其變形率，RD_A、RD_T 、 Rd_a、 Rd_t 會隨著模具彎角半徑的減小或增加工件厚度將會有較高之變形率。此外，上述之討論也會隨孔之尺寸和孔數的增加使所有的變化減少。本研究使用RSM的方法得到明確的二階數學公式，在有關回彈角與RD_A、RD_T、Rd_a及Rd_t實驗值與預測值影響變化之比較，其最大誤差分別為4.24％、4.30％、4.52％、4.7％和 4.21％。
In this study, a series of experimental investigation focusing on the folded zone of the plates with holes in the L-bending process have been performed. On the basis of the response surface method (RSM), a set of complete, systematic experimental design and analysis method have been proposed. The experimental design adopts the central composite design (CCD). Also, four independent variables, including thickness of work pieces, die corner radius, hole diameter and the number of holes, are regarded as the design parameters in the experimental design. The value of springback angle and the shape deformation of holes in the L-bending process also have been explored in the meanwhile. The results reveal that the value of springback angle increases with increasing the size of hole and die corner radius, and decreases with increasing the thickness of work pieces. The deformation rate along the axial direction on the top surface, the tangent direction on the top surface, the axial direction on the bottom surface and the tangent direction on the bottom surface are designated as RD_A , RD_T, Rd_a, Rd_t, respectively. According to the experimental results, to reduce the die radius or increase the thickness of work pieces will cause a high deformation rate. In addition, the variation of RD_A , RD_T, Rd_a and Rd_t will decrease with the increase of the hole diameter and the number of holes. Using the RSM method with a quadratic explicit formulas, the maximum deviations of the predicted values of relevant influencing parameters in terms of the springback angle, the variation of RD_A , RD_T, Rd_a and Rd_t, which are compared with the actual experimental data, are 4.24%, 4.30%, 4.52%, 4.79% and 4.21%, respectively. |
