Process monitoring in milling by pattern recognition

Abstract The application of the pattern recognition technique for process monitoring in end milling is discussed in this paper. Cutting forces, torque, and spindle vibrations are monitored during machining, and are used to generate several signal features which are shown to be rather sensitive to the process conditions under consideration. Five machining conditions (classes) are considered in this study, namely; tool life end, chatter, stable cutting, air and transient cutting. A linear discriminant function-based technique is used for the identification of process conditions. The performance of the classifier is evaluated by numerous cutting tests. The results indicate correct rates of recognition for each class ranging from 85% to 100%.