Cutting Tool Condition Monitoring Based on a Machine Learning Approach by Using Vibration Signals

In metal cutting operations, tool wear could result in higher energy utilization, excessive tool vibration, and deteriorated surface finish quality. Therefore, tool wear monitoring is critical to enhance manufacturing productivity and quality. This study proposes a new approach for monitoring and classification of tool wear with 192 data sets of turning experimental tests at different cooling approaches. The Daubechies wavelet transform is used to state the level of flank wear at different cooling approaches. The acquired vibration signals were processed in both the time and frequency domain with various features extracted. An artificial neural network (ANN) model was applied to detect the flank wear at dry, flood, and minimum quantity lubrication(MQL) cooling states. The results confirmed the ability of the monitoring system to detect the flank wear of the cutting insert by utilizing the tool vibration signals.