Collaborative Research: Knowledge and Data-driven Design of Mechanical Metamaterials

协作研究：机械超材料的知识和数据驱动设计

• 批准号: 1825521
• 负责人: Daniel Selva Valero
• 金额: $ 28.05万
• 依托单位: Texas A&M Engineering Experiment Station
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2018
• 资助国家: 美国
• 起止时间: 2018-09-01 至 2023-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1825521&HistoricalAwards=false
• 关键词: Collaborative; Research; Knowledge; Data; driven

Traditionally, materials have been designed through choices of molecular level composition and structure. With the advent of increasingly sophisticated material-forming techniques like additive manufacturing, structures repeated at microscale are also now being used to realize effective overall properties; these materials are termed "metamaterials". The most obvious gain for metamaterials versus traditional fully dense materials is in weight and material consumption, but there are also responses that are simply not achievable with fully dense polymers. In particular, 3D printing of polymers has reached a critical threshold of quality, speed, and size at which it can be used for production rather than just prototyping. The geometry within the repeated structural cell of a metamaterial critically influences the overall properties. Determining the optimal geometry requires a design framework distinct from that used for dense materials. This work will explore innovative ways of combining expert knowledge (i.e., physical laws, models, heuristics) and databases of actual and simulated material behaviors, using advanced machine learning and search algorithms to foster the discovery of metamaterials with desired properties. Progress in the project will promote the new field of data-driven design as well as advance the national health, prosperity, and welfare by facilitating the design of advanced materials with hitherto unknown, yet desirable combination of properties. Beyond this technological impact, this grant will serve to prepare the next generation of students for a new era of design for intelligent materials and structures. Doctoral, undergraduate, high school, and middle school students will be reached through in-lab research experiences and design outreach activities.The central objective of this work is to create a design method for 3D printable elastomeric metamaterials that leverages both available engineering knowledge and data. The design space of interest will include two distinct geometry classes -- lattice materials and minimum energy surfaces. The methodology in this project will leverage physics-based models, existing knowledge, and data to minimize the resources needed to reach an acceptable design. The intermediate research objectives are to: (1) formulate and validate a comprehensive set of low computational cost mechanics models for lattice and minimum surface energy style metamaterials, together with a set of heuristics for designing such materials; (2) develop data-driven surrogate models and identify sources of and quantify uncertainty in predicted mechanical properties of 3D printed mechanical metamaterials; (3) develop knowledge representations and data fusion strategies to incorporate expert knowledge including physical laws, heuristics, and beliefs into the design of 3D printed metamaterials. In contrast to the current state-of-the-art for metamaterial design, the design framework that is produced by this grant will be well oriented to accommodate large deformation. This will facilitate design of printed metamaterials for properties such as toughness and failure strain.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.

传统上，材料是通过选择分子水平的组成和结构来设计的。随着增材制造等日益复杂的材料成型技术的出现，在微观尺度上重复的结构现在也被用于实现有效的整体性能；这些材料被称为“超材料”。与传统的全致密材料相比，超材料最明显的优点是重量和材料消耗，但也有全致密聚合物无法实现的响应。特别是，聚合物的3D打印已经达到了质量、速度和尺寸的关键门槛，可以用于生产，而不仅仅是原型制作。超材料重复结构单元内的几何形状对其整体性能有重要影响。确定最佳几何形状需要一个不同于用于致密材料的设计框架。这项工作将探索结合专家知识（即物理定律，模型，启发式）和实际和模拟材料行为数据库的创新方法，使用先进的机器学习和搜索算法来促进发现具有所需特性的超材料。该项目的进展将促进数据驱动设计的新领域，并通过促进设计迄今为止未知的、但理想的性能组合的先进材料，促进国家的健康、繁荣和福利。除了这种技术影响之外，这项资助还将帮助下一代学生为智能材料和结构设计的新时代做好准备。博士，本科生，高中和初中生将通过实验室研究经验和设计推广活动。这项工作的中心目标是创建一种3D打印弹性体超材料的设计方法，利用现有的工程知识和数据。感兴趣的设计空间将包括两种不同的几何类型——晶格材料和最小能量表面。这个项目中的方法将利用基于物理的模型、现有的知识和数据来最小化达到可接受的设计所需的资源。中间的研究目标是：(1)制定和验证一套全面的低计算成本的晶格和最小表面能型超材料力学模型，以及一套设计此类材料的启发式方法；(2)开发数据驱动的替代模型，识别和量化3D打印机械超材料预测力学性能的不确定性来源；(3)开发知识表示和数据融合策略，将物理定律、启发式和信念等专家知识纳入3D打印超材料的设计中。与目前最先进的超材料设计相比，这项拨款产生的设计框架将很好地适应大变形。这将有助于设计具有韧性和失效应变等特性的印刷超材料。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。

项目成果

Evaluating Designer Learning and Performance in Interactive Deep Generative Design

评估设计师在交互式深度生成设计中的学习和表现

• DOI: 10.1115/1.4056374
• 发表时间: 2023
• 期刊: Journal of Mechanical Design
• 影响因子: 3.3
• 作者: Chaudhari, Ashish M.;Selva, Daniel
• 通讯作者: Selva, Daniel

Leveraging Design Heuristics for Multi-Objective Metamaterial Design Optimization

利用设计启发法进行多目标超材料设计优化

• 发表时间: 2021
• 期刊: IDETC/CIE2021
• 作者: Kumar, Roshan S;Srivasta, Srikar;Silberstein, Meredith N;Selva, Daniel
• 通讯作者: Selva, Daniel

Examining the impact of asymmetry in lattice-based mechanical metamaterials

• DOI: 10.1016/j.mechmat.2022.104386
• 发表时间: 2021-10
• 期刊: Mechanics of Materials
• 影响因子: 3.9
• 作者: Srikar Srivatsa;Roshan Suresh Kumar;Daniel Selva;M. Silberstein

Identifying and Leveraging Promising Design Heuristics for Multi-Objective Combinatorial Design Optimization

识别和利用有前途的设计启发法进行多目标组合设计优化

• DOI: 10.1115/1.4063238
• 发表时间: 2023
• 期刊: Journal of Mechanical Design
• 影响因子: 3.3
• 作者: Suresh Kumar, Roshan;Srivatsa, Srikar;Baker, Emilie;Silberstein, Meredith;Selva, Daniel
• 通讯作者: Selva, Daniel