Probing the Evolution of Granular Microstructures during Dynamic Annealing via Integrated Three-Dimensional Experiments and Simulations

许多金属材料是多晶的，其中固体由许多具有不同取向的微小晶体（“晶粒”）组成。在某些情况下，当材料以单晶形式制成时，就可以获得更好的性能，例如喷气发动机涡轮叶片、形状记忆合金和太阳能电池。然而，生产单晶既昂贵又耗时。最近研究人员发现，通过反复加热和冷却金属合金的循环热处理引起的异常晶粒生长可以产生单晶。在异常晶粒生长过程中，少数晶粒优先吞噬邻近晶粒生长。这个项目的目标是发现这个过程发生的原因和方式。该项目整合了实验表征和晶粒生长过程模拟的新兴研究，利用实时观察多晶微观结构演变的能力，并利用高性能计算来模拟微观结构演变。发展对异常晶粒生长的基本理解可能会导致单晶材料制造的范式转变，从而促进制造业和技术的全球竞争力以及国家繁荣。该项目还促进了训练有素的未来劳动力的发展；两名研究生在实验、建模、模拟和数据分析方面接受了最先进技术的培训。拓展活动是通过“女性在数学、工程和科学方面更出色”项目和华盛顿小学科学奥林匹克竞赛进行的，其中包括一个虚拟现实演示，让学生们在热处理过程中经历不断变化的微观结构。循环热处理是通过热循环反复形成和溶解析出相的过程，它有望通过异常晶粒生长对单晶和其他大晶粒材料进行固态加工。然而，由于对这一进程的基本机制了解不足，其全部潜力尚未实现。本项目的目的是促进非等温退火过程中异常晶粒生长的科学研究。研究了以下几个基本问题：动态热环境下晶粒异常生长的机制是什么？哪些晶粒最有可能变得异常，它们的显微结构特征是什么？异常晶粒如何生长成新的微观结构邻域？为了回答这些问题，结构表征、相场和相场晶体建模以及图论方法的新兴研究得到了协同整合。基于同步加速器的高分辨率x射线衍射显微镜用于实时可视化和量化晶粒和亚晶粒网络的演变，并基于测量的空间、时间和方向分辨数据集作为初始条件，实现微观结构演变模拟。然后将所得的高维数据提炼成一个网络模型，该模型简洁地描述了晶界运动的颗粒和局部驱动力，从而大大降低了预测微观组织演变的计算成本。对热循环过程中晶粒生长异常的科学理解最终为单晶制造工艺设计提供了依据。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。

Non-technical SummaryMany metallic materials are polycrystalline, in which a solid consists of many tiny crystals (“grains”) with different orientations. In some cases, superior properties can be achieved when the material is made in a single-crystal form, as in the case for jet-engine turbine blades, shape-memory alloys, and solar cells. However, single crystals are expensive and time-consuming to produce. Recently researchers have found that single crystals can be produced through abnormal grain growth induced by cyclic heat treatment, where the metal alloy is heated and cooled repeatedly. During abnormal grain growth, a few grains preferentially grow by engulfing the neighboring grains. The goal of this project is to discover why and how this process occurs. The project integrates emergent research in experimental characterization and simulations of the grain growth process, capitalizing on the ability to watch the evolution of the polycrystalline microstructure in real-time and leveraging high-performance computing to simulate microstructural evolution. Developing the fundamental understanding of abnormal grain growth could lead to a paradigm shift in the manufacture of single-crystalline materials, and therefore it promotes global competitiveness in manufacturing and technology and national prosperity. The project also promotes the development of a highly trained future workforce; two graduate students are trained in state-of-the-art techniques in experiments, modeling, simulations, and data analysis. Outreach activities are carried out through the Females Excelling More in the Math, Engineering, and the Sciences program and Washtenaw Elementary Science Olympiad and include a virtual reality demonstration that allows students to walk through an evolving microstructure during heat treatment. Technical SummaryA cyclic heat treatment, in which precipitates are repeatedly formed and dissolved by thermal cycling, holds promise for solid-state processing of single crystals and otherwise large-grained materials via abnormal grain growth. However, its full potential has not been realized due to the poor understanding of the mechanisms underlying the process. The objective of this project is to advance the science governing the growth of the abnormal grains during non-isothermal annealing. The following fundamental questions are addressed: What is the mechanism by which abnormal grain growth is initiated under dynamic thermal environment? Which grains are most likely to become abnormal and what are their microstructural signatures? How does the abnormal grain grow into new microstructural neighborhoods? To answer these questions, emergent research in structural characterization, phase-field and phase-field-crystal modeling, and graph theory methods are synergistically integrated. High-resolution synchrotron-based X-ray diffraction microscopy is utilized to visualize and quantify the evolution of the grain and subgrain network in real-time and also enable microstructural evolution simulations based on the measured space-, time-, and orientation-resolved datasets as initial conditions. The resulting high-dimensional data is then distilled into a network model that succinctly describes the granular and the local driving forces for grain boundary motion, which significantly reduces the computational cost of predicting the microstructural evolution. Scientific understanding of abnormal grain growth upon thermal cycling ultimately informs the process design of single-crystal fabrication.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.