Collaborative Research: GOALI: Experimentally-Validated Computational Approach to Developing and Predicting Kinetics in Anisotropic Systems

非技术总结：减轻车辆的重量可以提高燃油效率，从而抵消能源开采和使用的经济和社会成本的上升。镁作为钢和铝的潜在替代品具有吸引力，因为它的密度低，刚度/重量和强度/重量比高。然而，合金的开发可能是一个缓慢而昂贵的过程。为了加快新铸造合金的设计，本研究使用了经过实验验证的计算方法，特别强调了实现动力学信息的方法。本着材料基因组计划的精神，该项目将计算和理论（伊利诺伊大学厄巴纳-香槟分校）、实验（佛罗里达大学）和工业建模（ThermoCalc有限责任公司）结合在一起，共同努力开发科学和数据，以预测加工、材料结构和最终性能之间的关系。随着研究凝固和预测原子在原子尺度上如何运动的新方法的发展，科学和工程方面的进展将适用于其他合金。该项目利用了计算能力的新进展、新的理论发展和新的高通量实验方法。该项目将资助两名博士生，教师既参与本科生研究，也参与高中科学教师的教育。技术概述：由于镁合金的低密度，为了减少排放和提高交通运输车辆的燃油经济性，人们已经做出了相当大的努力来了解和提高镁合金的结构-性能关系。这个合作项目将计算（伊利诺伊大学厄巴纳-香槟分校）和实验（佛罗里达大学）结合在一起，以一种综合的方式显著推进铸造镁合金运输的知识。在材料基因组计划的精神下，pi以一种集成和自一致的方式进行，以理解和预测凝固过程中的传输现象。该方法将计算和实验相结合，提供了验证并将基本机制与中尺度运输联系起来。通过与ThermoCalc， LLC的合作，他们将创建新的动力学数据库，以实现材料开发。该方法的目标是在预测新镁合金相变现象的框架中对Mg中的溶质和空位输运产生新的基本理解。对多尺度行为的统一理解为材料基因组框架中未来合金开发提供了科学驱动的方法。该项目的数据将通过公共数据库（NIST MatDL）和中尺度相场和连续水平模拟数据库的开发进行传播。通过考虑温度、时间和成分的影响，这项工作的预测允许加工-结构和结构-性质预测的完全合并。该项目有四个相互依存的任务来预测铸态Mg的中尺度微观结构：1)确定HCP和液态Mg在二元和高阶体系中的迁移率。密度泛函理论（DFT）将通过计算活化能和迁移率引领这一方法。2)通过实验和DFT中的分子动力学模拟，确定所有相关体系的固液表面能和摩尔体积。3)在储存库和迁移率数据库中组装摩尔体积、表面能和迁移率。DICTRA模拟与扩散倍数进行了比较，并通过实验数据验证了偏析行为和均质温度。4)通过迁移率、摩尔体积和表面能的集成来预测凝固的微观结构。模拟包括预测核数密度、大小和组成作为温度和冷却速率的函数，并通过类似加工路线获得的实验粒度分布进行验证。预期的结果是对HCP金属中原子输运的新的基本理解，以及迁移率、体积和表面能数据库，这些数据库既可以作为存储库，也可以作为增加系统复杂性（附加相、溶质等）的基础。数据将以逻辑、可管理和可访问的方式提供，以便转换为商业实践。该平台可用于预测未来其他相关相的成核行为。

Non-technical summary:Reducing the weight of vehicles can increase fuel efficiency, which counters the rising economic and social costs of energy extraction and use. Magnesium is attractive as a potential alternative to steel and aluminum because of its low density, and high stiffness/weight and strength/weight ratios. However, alloy development can be a slow and costly process. To expedite the design of new casting alloys, this research uses experimentally validated computational methods, with particular emphasis on enabling a kinetically-informed approach. In the spirit of the Materials Genome Initiative, this project brings together computation and theory (Univ. Illinois, Urbana-Champaign), experimentation (Univ. Florida), and industrial modeling (ThermoCalc, LLC) in a joint effort to develop the science and data to predict relationships between processing, material structure, and eventually properties. The science and engineering advances will be applicable to other alloys as new methods are developed to study solidification and predict how atoms move at the atomic scale. This project takes advantage of new advances in computing power, new theoretical developments, and new high-throughput experimental methods. The program will support two PhD students, and the faculty are involved both in undergraduate research and education of high-school science teachers.Technical summary:To combat emissions and increase the fuel economy in transportation vehicles, considerable effort has been made to understand and enhance the structure-property relationships in magnesium (Mg) alloys due to their low density. This collaborative project unites computation (Univ. Illinois, Urbana-Champaign) and experimentation (Univ. Florida) in an integrated fashion to significantly advance the knowledge of transport in cast Mg alloys. In the spirit of the Materials Genome Initiative, the PIs proceed in an integrated and self-consistent manner to understand and predict transport phenomena during solidification. The approach integrates computation and experiment throughout, providing validation and connecting fundamental mechanisms to mesoscale transport. In collaboration with ThermoCalc, LLC, they will create new kinetic databases to enable material development.The goal of this approach is to produce a new fundamental understanding of solute and vacancy transport in Mg in a framework to predict phase transformation phenomena for new Mg alloys. A unified understanding of multiscale behavior provides a science-driven approach to future alloy development in a materials-genome framework. The data from this program will be disseminated via public databases (NIST MatDL) and the development of databases for mesoscale phase-field and continuum-level simulations. By accounting for temperature, time and compositional effects, the predictions from this work allow for the complete merger of processing-structure and structure-property predictions.The project has four tasks that build upon each other to predict the mesoscale microstructure for cast Mg: 1) Determine mobilities in binary and higher order systems for HCP and liquid Mg. Density functional theory (DFT) will lead this approach by calculating activation energies and mobilities. 2) Determine solid-liquid surface energies and molar volumes in all relevant systems via experiments and molecular dynamics simulations in DFT. 3) Assemble molar volumes, surface energies and mobilities in repositories and mobility databases. DICTRA simulations are compared to diffusion multiples, and segregation behavior and homogenization temperatures validated with experimental data. 4) Predict solidified microstructures through the integration of mobilities, molar volumes and surface energies. Simulations include the prediction of nuclei number density, size, and composition as a function of temperature and cooling rate, and are validated with experimental grain size distributions obtained from similar processing routes.The expected outcomes are a new fundamental understanding of atomic transport in HCP metals, with mobility, volumetric and surface energy databases that will serve both as a repository and a foundation for increasing system complexity (additional phases, solutes, etc.). The data will be supplied in a logical, manageable, and accessible fashion for transition into commercial practice. This platform may allow predictions of nucleation behavior in other relevant phases in the future.