Investigations on tool wear behavior in turning AISI 304 stainless steel: An empirical and neural network modeling approach

Abstract

Machining with a cutting edge with extensive damage or a fractured cutting edge significantly influences the machining performance. Therefore, investigations on tool wear behavior, their forms, and wear mechanisms will be very helpful in the current environment of sustainable manufacturing. On the other hand, the machining economy is negatively impacted by replacing the tool well before its useful life. This proactive maintenance planning reduces the risk of sudden tool failure and potential workpiece damage. Accordingly, the current work creates empirical and ANN models to predict flank wear growth for turning AISI 304 stainless steel using a MTCVD-TiCN/Al₂O₃ coated carbide tool. The experiments were designed to cover a broad range of operating conditions to ensure the model's accuracy and applicability in practical machining scenarios. An ANN was modeled using a feedforward backpropagation machine learning technique. In this study, a higher prediction accuracy of 0.9975 was achieved with ANN model as compared to the empirical model. The most common wear mechanism observed is metal adhesion, followed by fracture due to the pulling away of adhered material. The developed models have been found to be valuable for optimizing cutting parameters and enhancing tool life in machining.