Cellular solids and micro-lattices are a class of lightweight architected materials that have been established for their unique mechanical, thermal, and acoustic properties. It has been shown that by tuning material architecture, a combination of topology and solid(s) distribution, one can design new material systems, also known as metamaterials, with superior performance compared to conventional monolithic solids. Despite the continuously growing complexity of synthesized microstructures, mainly enabled by developments in additive manufacturing, correlating their morphological characteristics to the resulting material properties has not advanced equally. This work aims to develop a systematic data-driven framework that is capable of identifying all key microstructural characteristics and evaluating their effect on a target material property. The framework relies on integrating virtual structure generation and quantification algorithms with interpretable surrogate models. The effectiveness of the proposed approach is demonstrated by analyzing the effective stiffness of a broad class of two-dimensional (2D) cellular metamaterials with varying topological disorder. The results reveal the complex manner in which well-known stiffness contributors, including nodal connectivity, cooperate with often-overlooked microstructural features such as strut orientation, to determine macroscopic material behavior. We further re-examine Maxwell's criteria regarding the rigidity of frame structures, as they pertain to the effective stiffness of cellular solids and showcase microstructures that violate them. This framework can be used for structure-property correlation in different classes of metamaterials as well as the discovery of novel architectures with tailored combinations of material properties.
Hyperuniform monocrystalline structures by spinodal solid-state dewetting
在大的长度尺度上具有密度波动异常抑制的材料是被称为无序超均匀的新兴系统。潜在的隐藏顺序使它们在几个应用中具有吸引力，例如光管理和拓扑保护的电子状态。这些应用需要可扩展的制造，这是很难实现与可用的自上而下的方法。从理论上讲，它是已知的，spinodal分解可以导致无序超均匀结构。因此，自发形成稳定的图案可能是自下而上制造这些材料的可行途径。在这里，我们表明，单晶超导体为基础的结构，特别是Si $_{1-x}$ Ge $_{x}$层沉积在绝缘体上的硅衬底，可以进行旋节固态去湿功能与有效的超均匀字符相关的障碍。可以获得针对特定形态和超均匀特性的纳米到微米尺寸的结构，证明了该方法的通用性，并为无序超均匀超材料的技术应用铺平了道路。相场模拟解释了潜在的非线性动力学和物理起源的新兴模式。
Multi-fidelity Bayesian Data-Driven Design of Energy Absorbing Spinodoid Cellular Structures
Department of Electronic Components, Technology and Materials, Delft University of Technology（电子组件、技术和材料系，代尔夫特理工大学）；School of Engineering and Materials Science, Queen Mary University of London（工程与材料科学学院，伦敦女王学院）
