Among a broad range of promising applications, the use of cellular solids as lightweight structural components is an important field of research that requires reliable predictions of their stiffness and strength. Predictive and general models should not depend on extensive parameter-fitting experiments and should not rely on computationally intensive numerical calculations for each new set of geometric parameters and loading conditions. An overview of models for 2D, 2.5D, and three-dimensional structures will be presented. Most 2D and 2.5D models neglect out-of-plane behaviour and the face sheets used in sandwich panels. 3D studies, mainly by finite element models (FEMs), are often limited to a narrow set of geometries and simple loading conditions. Elastic anisotropy is well covered, but calculating yield surfaces remains a challenge. Simplified models based on structural mechanics are rare and often limited in scope. They offer a flexible, computationally efficient approach for simulating truss-based materials. For more advanced designs, parameter-based FEMs must be developed for any loading condition to facilitate the generalised incorporation of 3D cellular solids in mechanical design. Artificial intelligence and machine learning are promising approaches for making optimal use of experimental and FEM results across multidimensional parameter spaces.
Alarcón et al. (Wed,) studied this question.