Real Time Prediction of Flank Wear by Neuro Fuzzy Technique in Turning

Machine tool automation requires reliable online tool wear monitoring techniques for automated manufacturing. Automatic detection of the state of tool wear in metal cutting operations is an important industrial problem. In this paper, an ultrasonic system is presented to monitor the tool wear in turning. The ultrasonic waves reflected from the wear region were analysed and correlated with flank wear. The reflected ultrasonic signal is analysed in both time and frequency domain. Artificial intelligence techniques such as artificial neural networks, fuzzy logic and the neuro-fuzzy technique have proved their potential in monitoring the manufacturing processes. Here, Adaptive Neuro Fuzzy Inference System (ANFIS) is used to identify the tool wear. The experimental validation runs show the system can predict the tool wear with average error of 2.5%. The decision making algorithm (DMA) is presented to determine the status of wear.