当前位置: X-MOL 学术Math. Models Comput. Simul. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Comparison of Domain Partitioning Algorithms in the Problem of Direct Flow Simulation in Rock Samples at the Pore Scale
Mathematical Models and Computer Simulations Pub Date : 2021-03-15 , DOI: 10.1134/s2070048220060113
E. N. Golovchenko , M. V. Iakobovski , V. A. Balashov , E. B. Savenkov

Abstract

The combination of computed tomography and computational experiment is an important and promising tool in the study of the properties of various materials. State-of-the-art tomographic methods allow us to obtain a three-dimensional image of materials with a high resolution, which leads to a high dimension of discrete settings (106–109 numerical cells). They cannot be analyzed without the application of parallel computing methods. In turn, the efficiency of parallel simulations with a large number of processors depends largely on the balanced distribution of the grid across the processors. In this study, simulation of a single-phase fluid flow in the pore space of a sandstone sample with a voxel representation is used to compare the partitions obtained by various methods using the ParMETIS, Zoltan, and GridSpiderPar parallel partitioning tools. The average time spent on the interprocess exchange during one time step of the considered parallel simulation is compared when the grid is distributed over the cores in accordance with various partitions. The results demonstrate the advantages of some algorithms and reveal the crucial criteria for the problem. The DiMP-Hydro software is used as the numerical simulator.



中文翻译:

孔隙尺度岩石样品直流模拟问题中域划分算法的比较

摘要

计算机断层扫描与计算实验相结合是研究各种材料特性的重要且有前途的工具。最先进的层析成像方法使我们能够获得具有高分辨率的材料的三维图像,这导致了离散设置的高尺寸(10 6 –10 9数字单元格)。不应用并行计算方法就无法对其进行分析。反过来,使用大量处理器进行并行仿真的效率在很大程度上取决于整个处理器之间网格的均衡分布。在这项研究中,使用体素表示对砂岩样品孔隙空间中的单相流体流动进行仿真,以比较使用ParMETIS,Zoltan和GridSpiderPar并行分区工具通过各种方法获得的分区。当网格按照各种分区分布在核心上时,将比较在考虑的并行模拟的一个时间步中花在进程间交换上的平均时间。结果证明了某些算法的优势,并揭示了该问题的关键标准。

更新日期:2021-03-15
down
wechat
bug