Topology optimization and anisotropic design of 3D conformal surface-derived lattice structures

In this paper, we propose an advanced structural design method inspired by bionics, utilizing surface-derived lattices and incorporating multi-scale, gradient, and conformal design principles. To enable multi-scale structural optimization, we first address two key challenges. First, we develop an integrated implicit modeling approach for Conformal Lattice Structures (CLS), enhancing efficiency through the application of topological automatic equivalence transformation technology. Second, we adopt a Transfer Learning (TL) model to efficiently predict the performance of arbitrary lattice types, enabling online adaptation and eliminating the need for pre-training on specific configurations. Considering the multi-scale nature of CLS, the topology optimization framework is reconstructed to generate gradient lattices, including conducting Finite Element Analysis (FEA) and re-deriving sensitivity formulas for equivalent structures. Furthermore, the inherent anisotropic properties of surface-derived lattices expand the design dimensions of CLS. By employing Fourier transform for numerical projection of the macro-mesh, we derive a macro-scale design that aligns optimally with the stress transfer path. Through extensive numerical simulations and mechanical experiments, we demonstrate that the proposed optimization method significantly enhances the stiffness of lattice structures, providing interfaces for additive manufacturing. Additionally, the CLS designed with macro-scale anisotropy outperforms traditional arrangements, showcasing superior performance.