基於控制器訊號觀測刀具磨耗對切削力係數之影響

Abstract

傳統更換銑刀是根據加工方式(粗加工、精加工)、刀具廠商提供之壽命等來作為評判標準。而刀具磨耗是判斷刀具壽命的方法之一。除此之外，刀具磨耗亦會影響到切削加工的過程，例如材料移除率下降造成加工效率降低，或是加工表面粗糙度變差等。而觀察磨耗方式有圖像處理法與切削力監測法等方法，前者無法實現實時觀測，後者雖實現實時觀測，卻需要外接動力計來蒐集訊號，對於實驗進行有一定影響，因此本研究將探討機台控制器訊號是否可用於觀測刀具磨耗。 本研究將透過實驗討論考慮刀具磨耗時，機台控制器訊號推算切削力之可行性，並且透過推算切削力係數變化，觀察其與磨耗之間的關係。實驗中所蒐集之控制器訊號，包含主軸扭矩命令訊號、伺服軸扭矩命令訊號以及伺服軸位置誤差等，其中主軸扭矩命令訊號與切削實驗之切線方向受力相關，而伺服軸扭矩命令與進給方向受力相關，並且伺服軸傳動相較主軸複雜許多，因此需要找出位置誤差與切削力變化之頻率響應。透過控制器訊號估算之切削力，再透過切削週期法鑑別出切削力係數。由切削力係數的變化觀察與刀具磨耗相關性，並且與加工訊號觀察結果比較，得到邊緣力係數K_te、K_re與刀具磨耗呈高度正相關，以此建立一判斷刀具壽命之指標，應用於不同切削條件之實驗中。

Traditionally, tool replacement in milling operations is determined based on machining methods such as roughing and finishing, the tool manufacturer's recommended life, and other criteria. Tool wear is a crucial indicator for assessing tool life, and it significantly affects the machining process. For example, excessive tool wear can lead to reduced material removal rates, resulting in decreased machining efficiency, and deteriorated surface roughness. To monitor tool wear, methods such as image processing and cutting force monitoring are commonly used. However, image processing cannot achieve real-time observation, while cutting force monitoring, despite its real-time capabilities, requires the use of an external dynamometer to collect signals, which can interfere with the experiment. Thus, this study explores the feasibility of using machine controller signals to observe tool wear, providing a less intrusive and potentially more practical solution. This research involves experimental investigations into the feasibility of using machine controller signals to estimate cutting forces while considering tool wear and to observe the relationship between changes in cutting force coefficients and tool wear. The controller signals collected in the experiment include spindle torque command signals, servo axis torque command signals, and servo axis position errors. Spindle torque command signals are related to the tangential force direction in the cutting experiment, while servo axis torque commands correspond to forces in the feed direction. The servo axis transmission system is considerably more complex than the spindle, necessitating the determination of the frequency response of position error to changes in cutting force. By estimating cutting forces from the controller signals, the cutting force coefficients are identified using the cutting cycle method. The relationship between the changes in these coefficients and tool wear is then analyzed. The correlation between cutting force coefficient changes and tool wear is observed and compared with results from machining signal observations. It is found that the edge force coefficients, K_te and K_re, exhibit a high positive correlation with tool wear. This correlation helps establish an index for determining tool life, which can be applied to experiments under various cutting conditions. The findings of this study aim to provide a new approach for real-time tool wear monitoring, potentially improving the efficiency and effectiveness of machining processes.