Modeling Dynamic Behavior of VIG Plates Using Neural Networks

The subject of this thesis is a VIG (Vacuum Insulated Glass) plate. The aim of this study was to find a numerical model in order to create an effective and precise numerical imitation of the real panel, considering the appropriate support conditions.

Damian Kozanecki
Technical University of Lodz, Poland
Faculty of Civil Engineering, Architecture, and Environmental Engineering

Furthermore, the aim was also to analyze the influence of various geometrical and mechanical parameters on the natural frequencies values, as well as the description and selection of artificial intelligence methods and data sets, in order to determine the value of Young’s modulus of pillars located between the glass panes, using the created predictive models. The thesis is divided into four parts. The first part presents the methods of analysis used, the second part analyzes the dynamic behavior of the VIG plates, and the third part presents the applied methods of artificial intelligence. In the fourth part, the methods and results from the first three parts were used in relation to laboratory tests.