1. Systematic topology optimization of elastic metamaterials for broadband bandgaps and customized mechanical properties.
- Author
-
Yan, Gengwang, Li, Yingli, Yao, Song, Yin, Guohui, and Huang, Xiaodong
- Subjects
- *
UNIT cell , *OPTIMIZATION algorithms , *METAMATERIALS , *TOPOLOGY , *FINITE element method - Abstract
[Display omitted] • An effective framework is proposed for multi-objective optimization of elastic metamaterials. • The robust post-processing is introduced to improve the manufacturability of optimized topologies. • An in-depth study of the effects of key design parameters on multi-objective optimization is conducted. • The evolutionary mechanisms of multi-objective optimizationare explored in detail. • The physical properties and wave propagation are investigated numerically and experimentally. With respect to the wide variety of emerging applications, load-bearing capacity and harmful vibration attenuation of substructures are the principal concerns, although simultaneously satisfying both requirements is challenging. The multi-objective topological optimization framework for single-phase porous elastic metamaterials (EMs) with high efficiency and accuracy is established. Through the combination of gradient-free optimization algorithms and finite element methods, the non-dominated feasible solutions can be obtained by tailoring the periodic microstructure topology to achieve desired bandgaps (BGs) and in-plane mechanical properties. An important novelty of the proposed approach is the comprehensive investigation of the initial designs, mesh resolutions, and the symmetries of the unit cells, as well as the post-processing techniques. The results indicate that Pareto-front solutions derived from various initial designs both feature centrally distributed mass blocks that are connected with finer ligaments. In addition, the two-stage topology optimization method is introduced to effectively expand the feasible solution space at the cost of computational efficiency, and the mixed populations are utilized to accelerate convergence by 16% at least. Note that utilizing a 2D simplified model under the small deformation assumption (within the linear elasticity phase) for the optimization process requires about 3%-5% or even less time compared to utilizing a 3D model with the same optimization settings. The numerical results and experimental validation of wave propagation behavior are then carried out for wave modulation and utilization. The proposed method can be applied to structure optimization or inverse design, providing novel opportunities for the design of multifunctional periodic structures. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF