Neural Network Model for Tool Wear Curve Reconstruction during Turning of Inconel 718

The maximum flank wear measured during turning operation on Inconel 718 aircraft engine components was treated by neural network, with the aim to predict the wear time behaviour. Tool wear prediction is carried out through a supervised neural network approach on the basis of experimental data obtained during turning processes. Turning tests were performed using different cutting parameters on the basis of a specific design of experiment (DoE), and tool flank wear was measured at regular time steps during each turning operation. Different configurations of 3-layers feed-forward back-propagation neural networks were constructed, trained and tested for tool wear prediction through mapping from input vectors to output values.