Architected cellular materials are a class of artificial materials with cellular architecture-dependent properties. Typically, designing cellular architectures paves the way to generate architected cellular materials with specific properties. However, most previous studies have primarily focused on a forward design strategy, wherein a geometry is generated using computer-aided design modeling, and its properties are investigated experimentally or via simulations. In this study, we developed an inverse design framework for a disordered architected cellular material (Voronoi lattices) using deep learning. This inverse design framework is a three-dimensional conditional generative adversarial network (3D-CGAN) trained based on supervised learning using a dataset consisting of voxelized Voronoi lattices and their corresponding relative densities and Young’s moduli. A well-trained 3D-CGAN adopts variational sampling to generate multiple distinct Voronoi lattices with the target relative density and Young’s modulus. Consequently, the mechanical properties of the 3D-CGAN generated Voronoi lattices are validated through uniaxial compression tests and finite element simulations. The inverse design framework demonstrates potential for use in bone implants, where scaffold implants can be automatically generated with the target relative density and Young’s modulus. Graphical Abstract

Triply periodic minimal surfaces (TPMS) have drawn widespread attention out of their biomimetic features. This work compares skeletal-TPMS lattices with of strut-TPMS lattices in terms of their geometric and mechanical properties. Skeletal-TPMS lattices consist of continually smooth surfaces, while their counterparts, strut-TPMS lattices, are composed of cylindrical beams. These lattices derived from four kinds of TPMS, namely Gyroid (G), Schwarz Diamond (D), Schwarz Primitive (P), and iWp (W), having incremental nodal connectivity of 3, 4, 6, and 8, respectively. The scaling laws of Young’s modulus and yield strength are determined as power functions of volume fraction according to finite volume method (FVM) simulations. Their uniaxial compression behaviors are characterized by FVM simulations and are experimentally validated using additive manufacturing and uniaxial compression tests, showing that skeletal-TPMS lattices exhibit more homogeneous stress distribution. This work would not only broaden the material space for scaffolds, but shed light on the design criteria for tissue engineering.

Silica aerogels are highly porous 3D nanostructures and have exhibited excellent physio-chemical properties. Although silica aerogels have broad potential in many fields, the poor mechanical properties greatly limit further applications. In this study, we have applied the finite volume method (FVM) method to calculate the mechanical properties of silica aerogels with different geometric properties such as particle size, pore size, ligament diameter, etc. The FVM simulation results show that a power law correlation existing between relative density and mechanical properties (elastic modulus and yield stress) of silica aerogels, which are consistent with experimental and literature studies. In addition, depending on the relative densities, different strategies are proposed in order to synthesize silica aerogels with better mechanical performance by adjusting the distribution of pore size and ligament diameter of aerogels. Finally, the results suggest that it is possible to synthesize silica aerogels with ultra-low density as well as high strength and stiffness as long as the textural features are well controlled. It is believed that the FVM simulation methodology could be a valuable tool to study mechanical performance of silica aerogel based materials in the future.