Projection Methods for Multiscale Problems in Plasma Physics and Applications

等离子体物理中多尺度问题的投影方法及其应用

• 批准号: 0317511
• 负责人: Alexandre Chorin
• 金额: --
• 依托单位: University of California-Berkeley
• 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: 2003-08-15 至 2007-12-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0317511&HistoricalAwards=false
• 关键词: Projection; Methods; Multiscale; Problems; Plasma

Chorin and Tokman The investigators develop numerical methods that will beeffective in long-time simulations of the multi-scale phenomenaevident in plasma dynamics, particularly in modeling magneticreconnection. They also use the methods in studies of plasmadynamics in the solar corona and in laboratory experiments. Theymake available a software implementation of the methods. Themethods combine projection techniques with schemes of exponentialpropagation type for time-stepping. The dynamical behavior seen in astrophysical and laboratoryplasmas is complicated, and to understand it requires computersimulations. But the phenomena of interest span a wide range oflength and time scales. Consequently, to compute accurately thesolutions of the equations that describe these phenomena,conventional numerical methods require small steps in time oververy long time periods. Similarly, accurate spatial resolution ofthe solutions requires fine spatial meshes. Together, thesefactors require solving very large systems of equations at eachtime step, for a great many time steps. Such calculations areexpensive. The investigators develop and analysze numericalmethods that promise to reduce the computational costs, and applythem to study plasma physics problems. They make the methodsavailable in software. This widens the potential impact of thework, because problems with multiple scales arise in a variety ofscience and engineering applications.

Chorin和Tokman 研究人员开发的数值方法，将是有效的，在等离子体动力学，特别是在建模磁重联明显的多尺度现象的长期模拟。 他们还在日冕等离子体动力学研究和实验室实验中使用这些方法。 他们提供了这些方法的软件实现。 该方法将联合收割机投影技术与指数传播型时间步进方案相结合。 在天体物理和实验室等离子体中看到的动力学行为是复杂的，要理解它需要计算机模拟。 但有趣的现象跨越了很宽的长度和时间尺度。 因此，为了精确地计算描述这些现象的方程的解，传统的数值方法需要在很长的时间段内进行小的时间步长。 同样，精确的空间分辨率的解决方案需要精细的空间网格。 这些因素加在一起，需要在每个时间步求解非常大的方程组，需要很多时间步。 这样的计算是昂贵的。 研究人员开发和分析了有望降低计算成本的数值方法，并将其应用于研究等离子体物理问题。 他们使这个方法在软件中变得可用。 这扩大了这项工作的潜在影响，因为多尺度问题出现在各种科学和工程应用中。