CDS&E/Collaborative Research: Data-Driven Inverse Design of Additively Manufacturable Aperiodic Architected Cellular Materials

由于其非凡的性能，工程超材料是不同行业（如材料和能源）广泛功能产品的基础。这项计算和数据驱动科学与工程（CDS&amp；E）合作研究奖将为可制造机械超材料的发现和优化建立一种数据驱动的方法，通过利用人工智能的探索和外推能力，共同设计用于超光能设备和电子设备等产品的非周期细胞材料的几何和性能，实现先进建筑材料的全部潜力可变形软机器人，帮助美国先进制造业的复苏。将研究成果整合到教育活动中，将有助于培养数据科学、工程设计和先进制造方面的学生，扩大未被充分代表的少数民族和第一代大学生在设计和3D打印研究和教育方面的参与。本研究弥补了对三维非周期结构细胞材料（AACM）结构-性能关系的基本理解的知识差距，并实现了具有所需性能的增材可制造细胞材料的逆设计。该项目将利用数据驱动的方法，为具有特定属性的可增材制造细胞材料建立一个合理的设计范式。它将解决由非常大的几何空间、未知的属性空间理论极限、病态逆问题以及几何兼容性和可制造性约束所带来的挑战。研究工作包括：(1)通过计算发现框架扩展AACM单元力学性能空间的理论极限；(2)阐明细胞结构的几何-性质关系，推导出一种计算效率高的数据驱动逆映射，用于生成具有规定性质的多种AACM结构；(3)考虑到非周期结构模式组合设计中的兼容性和增材制造性挑战。增强对结构-制造-性能内在关系的认识将推动新型建筑材料设计与开发的基础研究。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

Due to their extraordinary properties, engineered metamaterials are the basis for a wide range of functional products across different industry sectors, such as materials and energy. This Computational and Data-Enabled Science and Engineering (CDS&E) collaborative research award will establish a data-driven approach for manufacturable mechanical metamaterials discovery and optimization to realize the full potential of advanced architected materials by harnessing the exploration and extrapolation capability of artificial intelligence for the co-design of the geometry and properties of aperiodic cellular materials used in products such as ultra-light energy devices and shape-morphing soft robotics, helping to revitalizing advanced manufacturing in the US. Integrating the research findings into educational activities will help train students in data science, engineering design, and advanced manufacturing, broadening the participation of underrepresented minorities and first-generation college students in design and 3D printing research and education.This research bridges the knowledge gap in the fundamental understanding of the structure-property relation of three-dimensional aperiodic architected cellular materials (AACM) and achieving the inverse design of additively manufacturable cellular materials with desired properties. This project will establish a rational design paradigm for additively manufacturable cellular materials with specified properties by leveraging data-driven approaches. It will address the challenges posed by a very large geometry space, unknown theoretical limits of the property space, ill-posed inverse problems, and geometric compatibility and manufacturability constraints. The research activities include: (1) extending the theoretical limits of mechanical property space of AACM units via a computational discovery framework; (2) elucidating the geometry-property relation of cellular structures to derive a computationally efficient data-driven inverse mapping for generating diverse AACM structures with prescribed properties; (3) respecting the compatibility and additive manufacturability challenges in the combinatorial design of aperiodic structural patterns. The enhanced understanding of intrinsic structure-manufacturing-property relation will advance fundamental research of novel architected materials design and development.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.