Determination of the parameters of two-beam laser splitting of silicate glasses using regression and neural network models

Abstract

The current work takes the results of the numerical experiment implemented in the Ansys finite element analysis program to create the neural network and regression models of two-beam laser splitting of silicate glasses. The regression models of two-beam laser glass cutting have been obtained in the DesignXplorer module of Ansys Workbench using a face-centered version of the central composite design. The processing speed, the parameters of laser beams, the glass plate thickness, and the distance between the laser radiation and the refrigerant affected zones were used as variable factors. The maximum temperatures and thermoelastic tensile stresses in the laser processing area were used as responses. The artificial neural networks have been constructed and trained using the TensorFlow package. The results of determining the maximum temperatures and thermoelastic stresses in the laser treatment area using the neural network and regression models have been compared.