TRIPODS+X:RES:Collaborative Research: Improving Templated Microstructures via Topological Data Analysis

TRIPODS X:RES:协作研究：通过拓扑数据分析改进模板化微观结构

• 批准号: 1839267
• 负责人: Dunbar Birnie
• 金额: $ 30万
• 依托单位: Rutgers University New Brunswick
• 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: 2018-10-01 至 2021-09-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1839267&HistoricalAwards=false
• 关键词: TRIPODS+X; RES; Collaborative; Research; Improving

The importance of microstructures in Material Science is well recognized. Their local and global geometry influence the functional behaviors of the materials being designed in a major way. Therefore, methodologies for their controlled manufacturing have always been a focus of intense research. Now, with continuing advancements in characterization of materials at higher resolution and faster time scales there is intensified need for data driven digital simulation and analysis of structure. This project focuses on leveraging the new area of topological data analysis in advancing the design of templated microstructure designs through a collaboration between material and data scientists at Rutgers University and data scientists at the TRIPODS center at Ohio State University. Templating is the ideal topical area for this collaboration because it so definitively directs shape development during processing and can benefit greatly from deeper topological and statistical analytics. The researchers will develop a topology-related synergy between Materials Science, Computer Science, and Statistics that will enable improved processing of materials using templating. The geometrical and topological advances developed in this program are expected to also be extensible to other areas of materials processing, each of which has unique shape novelty, alignment effects, or texture development. The project's work could also benefit a range of similar application fields such as medical image analysis, computational neuroanatomy, geographic information systems, and engineering designs. Indeed, collaborations to apply geometric/topological methods to some of these other application fields are already underway at the TRIPODS center at OSU and could benefit from close collaboration with this Materials-focused program as it develops.The proposed research involves concepts from mathematical areas of algebraic topology and geometry, applied statistics, and computational areas of algorithms and graph theory. These will be applied to materials microstructures created by templating to help understand topological interconnections, shapes, and dynamics, which would be of benefit to functional improvements in device operation. Research in topological data analysis has brought forth the need to investigate topological concepts in the presence of finite data, approximations, and noise, constraints that are always encountered in real materials characterization. Geometric and topological computation with intentionally structured materials will yield big and diverse data that can influence and improve future material and device fabrication efforts. These new data methods will be of interest to the topological and statistical communities as well as open up new avenues for predicting and intentionally creating structures with enhanced functionality.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.

微观结构在材料科学中的重要性是众所周知的。它们的局部和全局几何形状在很大程度上影响所设计材料的功能行为。 因此，其受控制造方法一直是研究的热点。现在，随着在更高分辨率和更快时间尺度下材料表征的不断进步，对数据驱动的数字模拟和结构分析的需求不断增加。该项目的重点是通过罗格斯大学的材料和数据科学家与俄亥俄州立大学 TRIPODS 中心的数据科学家之间的合作，利用拓扑数据分析的新领域来推进模板化微结构设计的设计。模板化是此次合作的理想主题领域，因为它在加工过程中明确指导形状开发，并且可以从更深入的拓扑和统计分析中受益匪浅。 研究人员将在材料科学、计算机科学和统计学之间开发与拓扑相关的协同作用，从而能够使用模板改进材料的处理。该项目中开发的几何和拓扑进步预计也可扩展到材料加工的其他领域，每个领域都具有独特的形状新颖性、对齐效果或纹理开发。 该项目的工作还可以使一系列类似的应用领域受益，例如医学图像分析、计算神经解剖学、地理信息系统和工程设计。事实上，将几何/拓扑方法应用于其他一些应用领域的合作已经在俄勒冈州立大学的 TRIPODS 中心进行，并且可以从与这个以材料为中心的项目的密切合作中受益，因为它的发展。拟议的研究涉及数学领域的概念代数拓扑和几何、应用统计学以及算法和图论的计算领域。这些将应用于通过模板创建的材料微观结构，以帮助理解拓扑互连、形状和动力学，这将有利于器件操作的功能改进。 拓扑数据分析研究提出了在有限数据、近似值和噪声以及实际材料表征中经常遇到的约束条件下研究拓扑概念的需要。使用有意结构化的材料进行几何和拓扑计算将产生大量且多样化的数据，这些数据可以影响和改进未来的材料和设备制造工作。这些新的数据方法将引起拓扑和统计界的兴趣，并为预测和有意创建具有增强功能的结构开辟新途径。该奖项反映了 NSF 的法定使命，并通过使用基金会的知识进行评估，被认为值得支持。优点和更广泛的影响审查标准。

项目成果

Topological filtering for 3D microstructure segmentation

用于 3D 微观结构分割的拓扑过滤

• DOI: 10.1016/j.commatsci.2021.110920
• 发表时间: 2022-02
• 期刊: Computational Materials Science
• 影响因子: 3.3
• 作者: Patel, Anand V.;Hou, Tao;Rodriguez, Juan D.;Dey, Tamal K.;Birnie, Dunbar P.
• 通讯作者: Birnie, Dunbar P.

Ten simple rules to cultivate transdisciplinary collaboration in data science

培养数据科学跨学科合作的十个简单规则

• DOI: 10.1371/journal.pcbi.1008879
• 发表时间: 2021-05
• 期刊: PLOS Computational Biology
• 作者: Sahneh F;Balk MA;Kisley M;Chan CK;Fox M;Nord B;Lyons E;Swetnam T;Huppenkothen D;Sutherland W;Walls RL;Quinn DP;Tarin T;LeBauer D;Ribes D;Birnie DP 3rd;Lushbough C;Carr E;Nearing G;Fischer J;Tyle K;Carrasco L;Lang M;Rose PW;Rushforth RR;Roy S;Matheson T;Lee T;Brown CT;Teal TK;Papeș M;Kobourov S;Merchant N
• 通讯作者: Merchant N