Predicting Mechanical Behaviour of Auxetic Lattice Structures using Finite Element Analysis and Machine Learning

Download
2024-12
Arslanca, Yaman
Mechanical properties of auxetic lattice structures have been extensively researched in both academia and industry. Recent advancements in artificial intelligence, particularly in machine learning, have also been attracting significant attention. This work aims to utilize machine learning to investigate, analyze and predict the mechanical behaviour of auxetic double arrow-head lattice structures. A total of 1401 double arrow-head lattice structures were generated using finite element analysis in an automated manner. The analysis results, along with the input features, were used to train three different machine learning models: Neural Network, Random Forest, and Extreme Gradient Boosting. Prediction results from this training for eight output variables are presented, and optimization studies using the Pareto set and a genetic algorithm are conducted to identify the optimal design parameters for the structure.
Citation Formats
Y. Arslanca, “Predicting Mechanical Behaviour of Auxetic Lattice Structures using Finite Element Analysis and Machine Learning,” M.S. - Master of Science, Middle East Technical University, 2024.