IH-GAN: A Conditional Generative Model for Inverse Design of Heterogeneous Cellular Structures

Cellular structures with controlled local structures can realize heterogeneous material properties and hence enable a much wider range of functions than homogeneous structures. However, the design of heterogeneous cellular structures is challenging due to the high degrees of design freedom. We propose a simple yet principled way to achieve the fast design of heterogeneous cellular structures. This method uses physics-based optimization to find the optimal material property distribution under given design requirements, and uses a conditional generative model, named Inverse Homogenization Generative Adversarial Network (IH-GAN), to find the local structures that correspond to the optimal material properties. Results show that compared to the conventional variable-density approach, our method achieves a 3.03% reduction in displacement in a compliance minimization problem, and reduces the errors by around 80% in two target deformation problems.