ALGORITHMS: Parallel Level Set Method on Adaptive Cartesian Grids for Multi-Scale Interfacial Flow Computation

算法：用于多尺度界面流计算的自适应笛卡尔网格并行水平集方法

• 批准号: 0305504
• 负责人: Z.J. Wang
• 金额: $ 5.15万
• 依托单位: Michigan State University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2003
• 资助国家: 美国
• 起止时间: 2003-07-01 至 2005-06-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0305504&HistoricalAwards=false
• 关键词: ALGORITHMS; Parallel; Level; Set; Method

Numerous physical phenomena in nature involve the motion of interfaces separating two or more regions. The level-set method was a PDE-based approach designed to capture the interface motion through the evolution of a signed distance function. This exploratory project aims to further extend the level-set method to unstructured adaptive Cartesian grids. We plan to address the following research task: Development of a level-set solver supporting the adaptive Cartesian grids, and various data structures including both tree and list-based data structures. This solver will be coupled to an adaptive Cartesian grid adaptor for interface based grid adaptations to improve the interface resolution and decrease the numerical errors near the interface.A set of software tools will be developed and made available that can be used to both train students with basic interface dynamics, and to conduct research in challenging engineering and scientific disciplines.

自然界中的许多物理现象都涉及分隔两个或多个区域的界面的运动。水平集方法是一种基于偏微分方程的方法，旨在通过带符号距离函数的演化来捕获界面运动。该探索性项目旨在进一步将水平集方法扩展到非结构化自适应笛卡尔网格。我们计划解决以下研究任务：开发支持自适应笛卡尔网格以及各种数据结构（包括树和基于列表的数据结构）的水平集求解器。该求解器将耦合到自适应笛卡尔网格适配器，用于基于界面的网格自适应，以提高界面分辨率并减少界面附近的数值误差。将开发并提供一组软件工具，可用于培训学生基本界面动力学，并在具有挑战性的工程和科学学科中进行研究。