Multi-point tool condition monitoring system: A comparative study

In the metal removal process, the condition of the tool plays a vital role to achieve maximum productivity. Hence, monitoring the tool condition becomes inevitable. The multipoint cutting tool used in the face milling process is taken up for the study. Cutting inserts made up of carbide with different conditions such as fault-free tool (G), flank wear (FW), wear on rake face (C) and tool with broken tip (B) are considered. During machining of mild steel, vibration signals are acquired for different conditions of the tool using a tri-axial accelerometer, and statistical features are extracted. Then, the significant features are selected using the decision tree algorithm. Support Vector Machine(SVM) algorithm is applied to classify the conditions of the tool. The results are compared with the performance of the K-Star algorithm. The classification accuracy obtained is encouraging hence, the study is recommended for real-time application.