Software package for modeling and optimizing parameters of laser processing of brittle non-metallic materials using laser splitting methods

Abstract

In this work, with help of a software package implemented in Python, at the first stage, a finite element calculation of the temperature fields and thermoelastic stress fields formed during the processing of quartz glass plates by laser splitting was performed. At the second stage, based on the data obtained as a result of finite element modeling, the parameters of laser cutting of quartz plates were approximated using artificial neural networks. Neural networks and their training algorithms were implemented using the Keras library. At the third stage, the search for optimal laser cutting parameters was carried out using the author's genetic algorithm using the constructed neural network approximators. Finite element modeling in the developed software package was implemented using the FEniCS library and verified by comparing the obtained results with the corresponding values of maximum temperatures and thermoelastic stresses in the laser processing zone of quartz plates, determined using the APDL programming language in the Ansys software package.