An artificial neural network approach for analysis and minimization of HAZ in CO2 laser cutting of stainless steel

This paper present an approach for modeling and analysis of the effects of the laser cutting parameters on the width of HAZ obtained in CO2 laser cutting of stainless steel by using artificial neural network (ANN). ANN model was developed in terms of the specific laser energy (laser power to cutting speed ratio), assist gas pressure and focus position. Using the experimental data the ANN was trained with gradient descent with momentum algorithm and the average absolute percentage errors on training and testing were 3.68 % and 3.52%, respectively. In addition to modeling and analysis, through ANN simulation optimal cutting conditions with minimal width of HAZ were identified.