| 摘要: | 傳統的振動輔助磁力研磨之方式皆為平行或垂直於工件表面的單軸向往復振動機制，其缺點為易造成工件表面研磨刮痕加深、不易得到鏡面效果。 本研究則是突破以往的振動輔助機構設計限制，以二維（X-Y平面）振動輔助磁力研磨的方式進行加工，其主要目的為利用磁刷旋轉運動與工件二維振動的相互作用，形成緻密的十字交叉加工路徑，進而改善加工效率與提高工件的表面品質，經一系列的實驗探討後可以得知二維振動輔助磁力研磨加工相較無振動輔助磁力研磨更能提昇磁力研磨效率與表面品質，進而使工件達到鏡面效果。 經由田口驗證實驗可以得知，對於表面粗糙度的改善最佳參數組合為：磁極間距1mm、SiC重量1g、鋼砂重量1.5g、研磨液重量3g、振動平台轉速1000rpm、SiC號數＃8000、鋼砂號數＃120、磁極轉速500rpm(A1B1C2D3E3F3G3H2)。 實驗證明振動輔助磁力研磨加工法在最佳參數組合下，能於5分鐘內有效改善不鏽鋼表面粗糙度由Ra0.13μm降低至0.03μm，改善率達77%，研磨25分鐘後更能達Ra0.01μm，表面改善率提高至92.3%。若無振動輔助加入則研磨5分鐘後為Ra0.05μm，改善率僅61.54%；由此可證明二維振動輔助磁力研磨加工法能節省加工時間並能有效的提高研磨效率、節省成本。 經由統計分析可得一針對表面粗糙度之非線性迴歸模型，並使用各種檢定、分析與評估，足可證明此表面粗糙度非線性迴歸模型間各參數交互作用不明顯，且具備顯著之預測能力。 The normal vibration directions of vibration-assisted magnetic abrasive finishing are all parallel or perpendicular to the surface of workpiece. Its shortcomings are easily lead to more scratches on the surface, and difficult to obtain mirror effect. This study break through previous studies, design a mechanism of two-dimensional (XY plane) vibration assisted to magnetic abrasive finishing. After a series of experiments, we get a better surface quality and enhanced efficiency compared to non-vibrational assisted magnetic abrasive finishing. Taguchi experiment verifies that, the best parameters which improve surface roughness most is: working gap 1mm, SiC weight of 1g, steel weight 1.5g, motor oil weight 3g, platform rotational speed 1000rpm, SiC mesh # 8000, steel mesh # 120, magnet pole rotational speed 500rpm (A1B1C2D3E3F3G3H2). The experiments proved that vibration-assisted magnetic abrasive finishing method in the best parameter combinations, can effectively improve the stainless steel surface roughness from Ra0.13μm to 0.03μm within 5 minutes, improvement rate 77%. After finishing for 25 minutes, the surface roughness is down to Ra0.01μm, the surface improvement rate increase to 92.3%. Without the assisted vibration, after finishing 5 minutes the roughness is Ra0.05μm, the improvement rate is only 61.54%. This proves that two-dimensional vibration assisted magnetic abrasive finishing method can save processing time and can effectively improve the finishinging efficiency and then save the cost of process. Through the statistical analysis, we have a non-linear regression model, and then use a variety of test, analysis and evaluation, getting that the interactions between parameters are not obvious, and the model have significant prediction ability. |
