Topology optimization using artificial intelligence

Abstract

An analysis of topology optimization employing deep learning, namely Generative Adversarial Networks (GANs), and topology optimization utilizing the Solid Isotropic Material with Penalization (SIMP) method is presented in this research. We describe the theoretical foundations of GANs and the SIMP technique. A cantilever beam with predetermined boundary conditions was the topic of a static study to show the practical efficacy of these methods. The structural performance parameters, such as maximal directional displacement, maximal Von Mises stress, and deformation energy. The findings show that deep learning-based topology optimization, as demonstrated by TopologyGAN, provides considerable benefits in terms of improved design correctness and computing performance.