Artificial neural network for predicting machining performance of uncoated carbide (WC-CO) in miling machining operation

Abstract

Surface roughness (Ra) is one of the most common responses in machining and an effective parameter to represent the quality of a machined surface. This paper presents the capability of an Artificial Neural Network (ANN) technique to develop a model to predict the Ra value of milling process. The model, presented as a network structure, is developed using the MATLAB ANN toolbox. Four different network structures were developed and assessed. The result of the modeling shows that a 3-7-1 network structure is the best model for end milling a titanium alloy using an uncoated carbide (WC-Co) cutting tool. The result of the ANN model has been compared to the experimental result, and ANN gave a good agreement between predicted and experimentally measured process parameters. The ANN technique has decreased the minimum surface roughness value of the experimental sample data by about 0.0126 �¿m, or 5.33%.