This paper describes a new concept for the implementation of the direct simulation Monte Carlo (DSMC) method. It uses a localized data structure based on a computational cell to achieve high performance, especially on workstation processors, which can also be used in parallel. Since the data structure makes it possible to freely assign any cell to any processor, a domain decomposition can be found with equal calculation load on each processor while maintaining minimal communication among the nodes. Further, the new implementation strictly separates physical modeling, geometrical issues, and organizational tasks to achieve high maintainability and to simplify future enhancements. Three example flow configurations are calculated with the new implementation to demonstrate its generality and performance. They include a flow through a diverging channel using an adapted unstructured triangulated grid, a flow around a planetary probe, and an internal flow in a contactor used in plasma physics. The results are validated either by comparison with results obtained from other simulations or by comparison with experimental data. High performance on an IBM SP2 system is achieved if problem size and number of parallel processors are adapted accordingly. On 400 nodes, DSMC calculations with more than 100 million particles are possible.

Stefan Dietrich

Iain D. Boyd

Journal of Computational Physics

JOURNAL OF COMPUTATIONAL PHYSICS 彻底处理了物理问题的计算方面，介绍了物理学各个领域中出现的数学方程的数值求解技术。该杂志力求强调跨越学科界限的方法。《计算物理学杂志》还出版 4 页或更少的简短注释（包括图表、表格和参考文献，但不包括标题页）。对已在本期刊上发表的文章发表评论的致编辑的信件也将被考虑。笔记和信件都不应该有摘要。

The Journal of Computational Physics focuses on the computational aspects of physical problems, presenting numerical solution techniques for mathematical equations arising in various fields of physics. The journal emphasizes interdisciplinary approaches. Additionally, the Journal of Computational Physics accepts brief notes of up to 4 pages (including figures, tables, and references, but excluding the title page). Letters to the editor commenting on articles published in this journal will also be considered. Notes and letters should not include abstracts.

《Journal of Computational Physics》专注于物理问题的计算方法，介绍了在物理学各个领域中出现的数学方程的数值求解技术。该杂志强调跨学科的方法。此外，《Journal of Computational Physics》还接受最多4页的简短注释（包括图表、表格和参考文献，但不包括标题页）。对已在本期刊上发表的文章的评论信件也将被考虑。注释和信件不需要摘要。

Scalar and Parallel Optimized Implementation of the Direct Simulation Monte Carlo Method

This paper describes a new concept for the implementation of the direct simulation Monte Carlo (DSMC) method that uses a localized data structure based on a computational cell to achieve high performance, especially on workstation processors, which can also be used in parallel.

Category	Quartile
物理与天体物理	2
物理与天体物理, 物理数学物理	1
物理与天体物理, 计算机跨学科应用	2

Scalar and Parallel Optimized Implementation of the Direct Simulation Monte Carlo Method

ivySCI AI Smartly Parses PDF, Answers Researchers' Questions, and Helps You Understand Papers in Seconds

Journal Info

Category	Quartile
PHYSICS, MATHEMATICAL	1