Adaptive neuro-fuzzy approach to predict tool wear accurately in turning operations for maximum cutting tool utilization

In industry, the capability to predict tool wear is an essential matter as the quality and performance of a cutting tool directly affect the product quality. Developing a model to predict tool wear can help control tool wear and maximize tool utilization. Therefore, this study presents a method of predicting tool flank wear of coated carbide inserts while machining AISI 1050 low carbon steel with a turning operation. The adaptive neuro-fuzzy approach (ANFIS) was implemented in this research. Experiments were conducted based on the Design of Experiments (DOE) technique by developing experiments with four factors at four levels corresponding to the L₁₆ (4⁴) experimental array to measure tool flank wear. At the end, a verification test was conducted to illustrate the effectiveness of this approach. Using ANFIS, average prediction accuracy of 92.42% was obtained and the ANFIS tool wear model developed indicated how the interaction between factors influenced tool wear.