Prediction of cutting tool wear during a turning process using artificial intelligence techniques

In the manufacturing industry, cutting tool failure is a serious event which causes damage to the cutting tool and reduces the quality of the product, which increases the cost of production. A reliable, intelligent, tool wear monitoring system is required in the metal cutting manufacturing process to mitigate these negative effects. This study presents a model-based approach for tool wear monitoring based on an adaptive neuro-fuzzy inference system (ANFIS) for a cold-finished steel bar 1215 turning process. A three-input cutting force (Fx, Fy and Fz) and single-output (tool flank wear) model was designed and implemented using the ANFIS approach. The forces were measured using a piezoelectric dynamometer and data acquisition system. Flank wear was also monitored using a tool maker’s microscope. The model prediction results show that it is accurate enough to perform online monitoring of the turning process and can detect wear while operating.