Tool Condition Monitoring (TCM) systems, which aim to identify tool wear automatically to avoid damage to the machined part, are often not suitable for industrial applications due to: (i) sensing methods which can be expensive and invasive to install and (ii) testing regimes that only evaluate a narrow operating window under controlled conditions. To combat these issues, the research outlined in this paper explores the feasibility of TCM using electrical power consumption of an industrial Computer Numeric Control (CNC) cutting machine in combination with a mechanistic model for end milling operations. End milling of aluminium 6082 was performed over a range of cutting parameters (i.e. spindle speed, feed rate, tool diameter) and repeated with increasing levels of tool wear to ascertain the suitability of this approach. Reasonable correlation between the predicted and observed tool wear was found using the mechanistic model (R² = 0.801), however variance between the cutting parameters highlights the limitations of the predictions. To combat this variance, an averaging window is taken over the course of a cutting program to reduce the overall error. A concept TCM system is then proposed, utilising cyber-physical models of the milling process to determine the width and depth of cuts for complex geometries which change in real time, with the challenges of implementing such a system discussed.

旨在自动识别工具磨损以避免损坏机加工零件的工具状态监测（TCM）系统通常由于以下原因而不适用于工业应用：（i）可能昂贵且具有侵入性的传感方法；以及（ii）仅在受控条件下评估狭窄操作窗口的测试方案。为了解决这些问题，本文概述的研究探索了使用工业计算机数字控制（CNC）切割机的电能消耗结合端铣削操作的机械模型进行TCM的可行性。在一定范围的切削参数（即主轴转速，进给速度，刀具直径）上进行铝6082的立铣，并随着刀具磨损水平的提高而重复进行，以确定这种方法的适用性。²  = 0.801），但是切削参数之间的差异突出了预测的局限性。为了克服这种差异，在切削程序的过程中采用了一个平均窗口以减少整体误差。然后，提出了一种概念中医系统，利用铣削过程的网络物理模型确定实时变化的复杂几何形状的切口的宽度和深度，并讨论了实现这种系统的挑战。