Collaborative Research: CDI-Type II: Cyber-Enabled Design of Functional Nanomaterials

当纳米颗粒分散在基质中时，会产生1028299 KumarNovel杂化材料。纳米粒子在基质中的排列方式对所得到的杂化材料的性能和性能具有至关重要的影响。目前在这一领域的大多数工作都处于“发现”模式：从给定的纳米颗粒构建块(NBB)开始，然后使用理论/模拟来模拟工艺步骤，以创建NBB组装件并计算其性质。虽然这种发现模式对于事后理解实验结果是有用的，但真正感兴趣的是先验地设计NBB，它可以组装成具有预定性质的功能结构。该结构可以在“平衡”状态下发展，或者可以冻结在可以通过使用适当的处理来调谐的结构中，例如通过应用诸如在挤出机中的强流场。目前，这种“反向”或设计目标只能通过反复试验的方式来实现。智力价值这个项目有两个主要主题：第一，将开发一个网络基础设施，以实现先验的NBB设计目标。这种情况下的一个关键困难是，这样的设计过程需要材料科学和科学计算/软件工程方面的专业知识，例如，对NBB进行适当的建模、集成各种软件以及将关键组件并行化以在诸如图形处理单元(GPU)的专用计算机体系结构上运行。这种基础广泛的能力是目前大多数从业者无法企及的。为了绕过这一困难，这项工作的第二个相关目标是开发基于模型集成计算(MIC)的工具，以便材料的计算设计可以从基于模拟的实验的设计和实施中分离出来。该项目是一项高风险、高回报的努力，因为它试图抽象出目前开发此类基于模拟的实验所需的深厚人类专业知识。更广泛的影响研究工作将丰富和补充创新的教育和推广活动。特别是，将举办一年一度的计算“网络营”，所有学生(本科生和毕业生)、博士后和教职员工以及外部合作者都将参加。与HBCU机构的合作将允许招募少数族裔REU学生，他们可以在网络夏令营度过一段时间，最终目标是招募和留住学生。

1028299KumarNovel hybrid materials result when nanoparticles are dispersed in a matrix. How the nanoparticles are arranged in the matrix critically affects the properties and the performance of the resulting hybrid materials. Most current work in this field is in a "discovery" mode: one starts with a given nanoparticle building block (NBB), and then uses theory/simulation to model the processing steps to create the NBB assemblies and compute their properties. While this discovery mode is useful for understanding experimental results a-posteriori, the real interest is in a-priori designing NBB that can assemble into functional structures with pre-determined properties. The structures may develop under "equilibrium" or may be frozen in structures that can be tuned through the use of appropriate processing, e.g., through the application of strong flow fields such as in an extruder. Currently, this "reverse" or design goal is only achieved in a trial-and-error manner. Intellectual MeritThis project has two overarching themes: First, a cyber-infrastructure will be developed to realize the goal of a-priori NBB design. A crucial difficulty in this context is that such a design process requires expertise in both materials science and scientific computing/software engineering, e.g., to properly model the NBBs, integrate various pieces of software, and parallelize critical components to run on specialized computer architectures, such as graphical processing units (GPUs). Such broad-based capabilities are out of reach of most current practitioners. To circumvent this difficulty, the second interrelated goal of this work is to develop model-integrated computing (MIC) based tools, so that the computational design of materials can be decoupled from the design and implementation of simulation-based experiments. The project is a high-risk, high-reward endeavor because it attempts to abstract the deep human expertise currently required for the development of such simulation-based experiments. Broader ImpactThe research effort will enrich and complement innovative educational and outreach activities. In particular, an annual Computational "cyber camp" will be conducted, in which all students (undergraduates and graduates), post-docs and faculty funded by this grant, and external collaborators will participate. Partnerships with HBCU institutions will allow for recruiting minority REU students, who can spend time at the cyber camp, with the ultimate goal of recruitment and retention.