Shape Gradients in Metamaterial Unit Cells: A Comparison of FEA and IGA
Please login to view abstract download link
The emerging desire for the application of metamaterials results, amongst others, in a deeper analysis of their unusual mechanical properties, which make them attractive for special application areas. Metamaterial behavior is commonly investigated using representative unit-cell models, where the geometry governs the effective response of lattice-like structures. In this context, one subgroup of metamaterials is represented by auxetic structures, which are known for negative Poisson’s ratios, high energy dissipation and absorption properties, as well as high resistance against penetration. These exemplarily mentioned properties make auxetic structures interesting for aerospace engineering, cf. e.g. [1]. Analyzing such metamaterials with finite element analysis (FEA) lead to high numerical and timeconsuming effort. Therefore an isogeometric analysis (IGA) approach can provide advantages considering precise results, sensitivity analysis besides possible optimization operations while requiring less resources. This study focuses on the methodology and modeling of unit cells rather than on the simulation of a specific metamaterial class. Within this study the modeling and analysis of auxetic unit cells are differentiated between FEA and IGA approaches. Especially the computations of shape gradients of physical quantities are juxtaposed. The comparison is conducted on representative cell geometries and addresses criteria such as numerical efficiency, accuracy and robustness of the gradient information, and implementability of the respective workflows. The work is based on variational design sensitivity analysis (VDSA) approach, cf. e.g. [2,3], focusing on geometry design and taking into account linear elasticity, geometrically nonlinear elasticity, and elastoplasticity. The obtained gradients are intended to be suitable for gradient-based shape optimization, and the presented sensitivity framework is applicable independently of the discretization method, allowing it to be combined consistently with either FEA or IGA. REFERENCES [1] Grünfelder, N., Kälber, L., Waschinsky, N., Ricken, T., Shape Optimization of Auxetic Unit Cells under Dynamic Loading in Macroscopic Components, AeroBest 2025 III ECCOMAS Thematic Conference on Multidisciplinary Design Optimization of Aerospace Systems Proceedings, 1, 37-51, 2025. [2] Manque Roa, N., Liedmann, J., Barthold, F.-J., Valdebenito, M., Faes, M., Interval Isogeometric Analysis for coping with geometric uncertainty, Comp
