CNC銑削智能化顫振控制系統之研究

Abstract

銑削製程由於能製作複雜形狀而被廣泛用於機械製造，然而不適當的切削條件之選用會容易產生加工時的顫振。再生式顫振屬於加工顫振中最常見的一種，再生式顫振發生時會造成刀具的損傷與降低工件的表面光度並且產生大量的噪音。為了避免發生再生式顫振，應在加工前選擇適當的切削條件，維持較佳的加工效率又不會導致再生式顫振，即使如此，亦可能有突發的再生式顫振，當再生式顫振出現時，就已造成刀具磨損及工件表面的損壞，因此應在顫振尚未造成影響前預測出顫振即將發生，並且及時抑制顫振的出現，保護刀具及工件的表面。 針對加工前選擇的切削條件，本文提出一結合切削經濟性與工具機穩定性的決定方法。計算出使加工的時間及成本最佳的轉速，配合切削穩定性圖及工具機主軸馬達的功率，得到可使工具機發揮最大效能的切削條件。目前在實務上因動力計成本昂貴，無法直接求得切削穩定性圖中必須的切削力係數，本文亦提出一簡便的策略，建立出切削負載率與平均切削力的線性模型，藉由此模型，可直接由CNC工具機控制器中量得並紀錄的切削負載率求得對應的平均切削力，以得到當時加工的切削力係數，讓切削穩定性圖及求得的切削條件更為準確。在顫振預測與顫振抑制方面，優化改良過去實驗室所提出的再生式顫振預測法則與顫振抑制策略，使其適用在更多的加工狀況，並將其整合實作出程式模組，可直接與CNC工具機的控制器連線以實際應用在加工上。實驗驗證發現，以本文提出的方法決定的切削條件，相較經驗及查表的條件加工效率最高可提升23%，且刀具壽命最高亦可多達3.6倍的壽命。而本文提出的切削力係數估算策略，經過不同材料的測試，估算結果與實際的切削力係數誤差在5.5% 以內，在容許的範圍內。顫振預測與抑制方面，在時域訊號還未有明顯變化前，程式模組即可預測出顫振將要出現，並自動調變至穩定主軸轉速，在顫振發生前即時抑制顫振的出現，達到預測顫振與抑制顫振的效果。本文從尚未加工的過程到切削過程為止，皆避免了再生式顫振的出現，成功地開發出一套實用且具商業價值的CNC銑床智能化顫振控制系統。

Milling process can produce complicated workpieces, so it was used widely in manufacturing. However, the inappropriate cutting conditions will be easy to induce regenerative chatter, which is the most common of chatter in milling, and it will damage tools and the surface of workpieces, and cause a lot of noise. To prevent regenerative chatter from affecting manufacturing quality, the select of appropriate cutting condition is necessary. Besides, the understand of machine tool dynamics will also increase production efficiency. On the other hand, in cutting processes, there may be unexpectedly regenerative chatter, so the cutting process should be monitored on-line. Nevertheless, when the regenerative chatter happens, the tool and surface of workpieces were injured. Therefore, the chatter is supposed to be predicted and be suppress to occur before it damage tool and workpieces. This work proposes a decision of appropriate cutting condition before machining, combining the economics of cutting and the stability of machine tool. It offers the cutting speeds for minimum cost and for minimum production time, and considers the stability lobe diagrams and the power of spindle to get appropriate cutting conditions, making the most of efficiency of machine tool. Moreover, in practice, the dynamometers are expensive , so it can’t be used to get the cutting coefficient , essential to build the stability lobe diagram . This work also proposes a convenient strategy, and sets up a linear model of the cutting load and average cutting forces . By the model, the cutting load from CNC controller can be transfered to the average cutting force to get the real cutting coefficient , it makes the stability lobe diagrams and cutting conditions more accurate . For the prediction of chatter and the suppression of chatter , this work improves the strategies , presented by the lab , makes them more useful to machining , and implements a modular program , able to connect the CNC controller . The experiment result shows that, the material removal rate of cutting condition decided by this work increases by 35% and the tool life goes up to 2 times, compared to the experienced cutting conditions . Addtionally, the differences between the cutting coefficients by the strategy and the real cutting coefficients are less than 5.5%. The modular program can preditct the chatter occur before the chatter amplitude is still small, and quickly suppress the chatter.