Massively Parallel Algorithms for Modeling the Structure of Liquids and Liquid-Solid Interfaces

用于模拟液体结构和液固界面的大规模并行算法

• 批准号: 9217287
• 负责人: James Chelikowsky
• 金额: $ 49.56万
• 依托单位: University of Minnesota-Twin Cities
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 1992
• 资助国家: 美国
• 起止时间: 1992-09-15 至 1997-02-28
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=9217287&HistoricalAwards=false
• 关键词: Massively; Parallel; Algorithms; Modeling; Structure

This grant provides support for a collaborative research project between materials science and computer science to develop and implement massively parallel algorithms for modeling the structure of liquids and liquid-solid interfaces. The grant is funded equally by the Divisions of Materials Research and Advanced Scientific Computing through the Computational Approaches to Real Materials program which itself is part of the High Performance Computing and Communications Initiative. The development of novel, massively parallel computers has opened new avenues for the application of electronic structure codes to complex materials systems. New codes will be implemented which will take full advantage of parallel machines to examine the structure and electronic properties of liquids and liquid-solid interfaces. Currently, these systems, in particular the liquid- solid interface, are not accessible to ab initio methods. The research will employ molecular dynamics simulations using quantum forces determined from ab initio pseudopotentials. Our first effort will focus on liquids involving tetrahedral semiconductors. We will calculate the electronic, structural and dynamical properties of the elemental liquids such as silicon and simple compounds such as gallium arsenide. With respect to the liquid- solid interface, we will examine the structural and dynamical properties of the interface during the melting process. %%% This research will develop and implement computer programs for the new generation of parallel computers in order to study the structure and electronic properties of liquid semiconductors and their interface with solid semiconductors. This work has important ramifications in the processing of semiconducting materials, in addition to its fundamental physical interest. The work will also develop new computer algorithms which will have importance in many fields of science. The research is funded jointly by the Divisions of Materials Research and Advanced Scientific Computing through the High Performance Computing and Communications Initiative.

该赠款为材料科学和计算机科学之间的合作研究项目提供支持，以开发和实施用于建模液体和液固界面结构的大规模并行算法。 该赠款由材料研究和高级科学计算部门通过真实材料计算方法计划平等资助，该计划本身是高性能计算和通信计划的一部分。 新型大规模并行计算机的发展为电子结构代码在复杂材料系统中的应用开辟了新途径。 将实施新的代码，充分利用并行机来检查液体和液固界面的结构和电子特性。 目前，这些系统，特别是液-固界面，无法通过从头算方法实现。 该研究将采用分子动力学模拟，使用从头算赝势确定的量子力。 我们的第一个努力将集中于涉及四面体半导体的液体。 我们将计算硅等元素液体和砷化镓等简单化合物的电子、结构和动力学特性。 对于液-固界面，我们将研究熔化过程中界面的结构和动力学特性。 %%% 这项研究将为新一代并行计算机开发和实施计算机程序，以研究液体半导体的结构和电子特性及其与固体半导体的界面。 除了其基本的物理意义外，这项工作在半导体材料的加工中也具有重要的影响。 这项工作还将开发新的计算机算法，这在许多科学领域都具有重要意义。 该研究由材料研究和先进科学计算部门通过高性能计算和通信计划联合资助。