计算机科学 ›› 2019, Vol. 46 ›› Issue (8): 89-94.doi: 10.11896/j.issn.1002-137X.2019.08.014

• 2018 全国高性能计算学术年会 • 上一篇    下一篇

基于超算暨HPIC-LBM的大时空尺度三维湍流磁重联

颜辉1, 朱伯靖2, 万文1, 钟英1, DavidAYune3   

  1. (中山大学国家超级计算广州中心 广州510006)1
    (中国科学院云南天文台暨中国科学院天文大科学中心 昆明650216)2
    (美国哥伦比亚大学应用物理和数学系 纽约10027)3
  • 收稿日期:2018-07-23 出版日期:2019-08-15 发布日期:2019-08-15
  • 通讯作者: 朱伯靖(1974-),男,博士,副教授,主要研究方向为大时空尺度湍流磁重联耗散区磁能释放-转移-耗散、等离子体加热及粒子加速,E-mail:bjzhu@ynao.ac.cn
  • 作者简介:颜辉(1987-),男,博士,工程师,CCF会员,主要研究方向为高性能计算,E-mail:hui.yan@nscc-gz.cn;万文(1987-),男,硕士,工程师,主要研究方向为超级计算机架构和异构系统;钟英(1989-),女,硕士,工程师,主要研究方向为高性能计算和应用软件优化;DavidAYune(1944-),博士,教授,主要研究方向为地球动力学
  • 基金资助:
    中国科学院前沿局重点项目(QYZDJ-SSW-SLH012),国家自然科学基金广东联合基金(U1501501),广东省引进创新创业团队项目(2016ZT06D211),中山大学超算应用培育专项

HPIC-LBM Method Based Simulation of Large Temporal-Spatial Scale 3D Turbulent Magnetic Reconnection on Supercomputer

YAN Hui1, ZHU Bo-jing2, WAN Wen1, ZHONG Yin1, David A YUNE3   

  1. (National Supercomputer Center in Guangzhou,Sun Yat-Sen University,Guangzhou 510006,China)1
    (Yunnan Observatories,Center for Astronomical Mega-Science,Chinese Academy of Sciences,Kunming 650216,China)2
    (Applied Physics and Applied Mathematics Department,Columbia University,New York 10027,USA)3
  • Received:2018-07-23 Online:2019-08-15 Published:2019-08-15

摘要: 大时空尺度湍流磁重联(LTSTMR)是空间物理、太阳物理、宇宙学领域中一种普遍存在的爆发现象。磁能转移-释放-耗散、等离子体加热及高能粒子加速是研究该类现象的核心内容,而研究湍流在LTSTMR中如何起作用是解决以上问题的关键所在。2D/2.5D磁重联模型因其在物理图像简化方面忽略了自然界爆发现象的3D属性和本质。文中基于新型HPIC-LBM粒子算法,在天河2号平台上开展了LTSTMR十万核心级别的数值实验工作。首次获得了太阳大气活动磁重联精细演化(0~500 km) 耗散区内oblique不稳定性证据。证明了耗散区内微观集体集合效应宏观表现的3种具体形式:湍流磁场self-generating-organization、湍动等离子体self-feeding-sustaining及磁场与等离子体间的相互作用。该研究为在超算平台上应用三维磁重联模型开展大时空尺度湍流磁重联提供了一种新途径。

关键词: HPIC-LBM, 磁能转移-释放-耗散, 大时空尺度, 高能粒子加速, 高性能计算, 三维湍流磁重联

Abstract: Large temporal-spatial scale turbulent magnetic reconnection (LTSTMR) is a general explosive astrophysical phenomena in space physics,solar physics and cosmology.To investigate this phenome,the mechanism of magnetic energy transfer-release-dissipation,plasma heating and acceleration of high energy particle should be learned,where the key is the role of turbulence.Previous 2D/2.5D simulation can only provide simplified physical pictures,which ignores the 3D nature and properties.By HPIC-LBM simulation on Tianhe-2 from National Supercomputer Center in Guang Zhou (NSCCGZ) with up to 100 000 CPU cores,the existence of oblique instability was firstly proved from fine structure simulation (0~500 km) of solar atmosphere activities.It also presented three kind of macroscopic representation inside the dissipation area:self-generating-organization,self-feeding-sustaining and interaction between magnetic field and plasma.This research provides a new tool for investigating 3D LTSTMR phenomenon on supercomputer

Key words: 3D turbulent magnetic reconnection, High performance computing, HPIC-LBM, Large temporal-spatial scale, Magnetic energy conversion-release-dissipation, Particle accelerating

中图分类号: 

  • O124
[1]PONTIN D I.Three-dimensional magneticreconnection regi- mes:A review [J].Advances in Space Research,2011,47(9):1508-1522.
[2]YAMADA M,YOO J S,ALMONTE J J,et al.Conversion of magnetic energy in magnetic reconnection layer of a laboratory plasma[J].Nature Communications,2014,5:4774.
[3]PARNELL C E,MACLEAN R C,HAYNES A L,et al.3D magnetic reconnection[J].Astrophysical Dynamics:From Ga-laxies to Starts Proceedings IAU Symposium,2010,6(S271),227-238.
[4]THURGOOD J O,PONTIN D I,MCLAUGHLIN J A.Three-dimensional Oscillatory magnetic reconnection[J].The Astrophysical Journal,2017,844(2):1-12.
[5]PARNELL C E,HAYNES A L,GALSGAARD K.Structure of magnetic separators and separator reconnection[J].Journal of Geophysical Research,2010,115:A02102.
[6]NAKAMURA T M,HASEGAWA H,DAUGHTON W,et al.Turbulent mass transfer caused by vortes induced reconnection in collisionless magnetospheric plasmas[J].Nature Communications,2017,8(1):1582.
[7]DAUGHTON W,ROYTERSHTEYN V,KARIMABADI H,et al.Role of electron physics in the development of turbulent magnetic reconnection in collisionless plasmas[J].Nature Physics,2011,7(7):539-542.
[8]MUÑOZ P A,JAIN N,KILIAN P,et al.A new hybrid code (CHIEF) implementing the inertial electron fluid equation without approximation[J].Computer Physics Communications,2018,224:245-264.
[9]FUJIMOTO K,SYDORA R.Plasmoid-induced turbulence in collisionless magnetic reconnection[J].Physical Review Letters,2012,109(26):265004.
[10]FUJIMOTO K.Studies on large-scale evolution of magnetic reconnection using full particle simulations with adaptive mesh refinement technique[D].Kyoto:Kyoto University Japan,2005.
[11]FUJIMOTO K.Possible plasma instabilities and electron hea- tings in the downstream region of the X-type neutral line[D].Kyoto:Kyoto University.Japan,2003.
[12]FUJIMOTO K.Three-dimensional outflow jets generated in collisionless magnetic reconnection[J].Geophysical Research Letters,2016,43(20):557-564.
[13]ZHU B J,LIN J.Applications and Development in Particle-in-Cell Methods for Investigating Large-Scale Turbulent Magnetic Reconnection[J].Progress in Astronomy,2016,34(4):459-476.(in Chinese) 朱伯靖,林隽.粒子云网格方法在大尺度湍流磁重联研究中的应用和进展[J].天文学进展,2016,34(4):459-476.
[14]ZHU B J.Application ofsupercomputing-based HPIC-LBM (hybrid particle in cell & lattice Boltzmann method) onmagne-tic energy dissipation mechanism in large scale turbulent magne-tic[C]∥Asis-Pacific Regional IAU Meeting.2017.
[15]CHENG H H,QIAO YC,LIU C,et al.Extended hybrid pressure and velocity boundary conditions for D3Q27 lattice Boltzmann model[J].Applied Mathematical Modelling,2012,36(5):2031-2055.
[16]ZHU B J,YAN H,YUEND A.Electron acceleration in interac- tion of magnetic islands in large temporal-spatial turbulent magnetic reconnection[J].Earth and Planetary Physics,2019,3(1):17-25.
[1] 李浩东, 胡洁, 范勤勤.
基于并行分区搜索的多模态多目标优化及其应用
Multimodal Multi-objective Optimization Based on Parallel Zoning Search and Its Application
计算机科学, 2022, 49(5): 212-220. https://doi.org/10.11896/jsjkx.210300019
[2] 陈国良, 张玉杰.
并行计算学科发展历程
Development of Parallel Computing Subject
计算机科学, 2020, 47(8): 1-4. https://doi.org/10.11896/jsjkx.200600027
[3] 汪洋, 李鹏, 季一木, 樊卫北, 张玉杰, 王汝传, 陈国良.
高性能计算与天文大数据研究综述
High Performance Computing and Astronomical Data:A Survey
计算机科学, 2020, 47(1): 1-6. https://doi.org/10.11896/jsjkx.190900042
[4] 贾迅, 钱磊, 邬贵明, 吴东, 谢向辉.
FPGA应用于高性能计算的研究现状和未来挑战
Research Advances and Future Challenges of FPGA-based High Performance Computing
计算机科学, 2019, 46(11): 11-19. https://doi.org/10.11896/jsjkx.191100500C
[5] 张云泉.
2018年中国高性能计算机发展现状分析与展望
State-of-the-art Analysis and Perspectives of 2018 China HPC Development
计算机科学, 2019, 46(1): 1-5. https://doi.org/10.11896/j.issn.1002-137X.2019.01.001
[6] 司雨濛,韦建文,Simon SEE,林新华.
星系分组算法的并行设计与优化:SGI系统与分布式集群对比
Parallel Design and Optimization of Galaxy Group Finding Algorithm on Comparation of SGI and Distributed-memory Cluster
计算机科学, 2017, 44(10): 80-84. https://doi.org/10.11896/j.issn.1002-137X.2017.10.015
[7] 徐琪,程耀东,陈刚.
高能物理环境中混合存储系统的设计与优化
Design and Optimization of Hybrid Storage System in HEP Environment
计算机科学, 2017, 44(10): 75-79. https://doi.org/10.11896/j.issn.1002-137X.2017.10.014
[8] 黄秋兰,李海波,石京燕,孙震宇,伍文静,程耀东,程振京.
基于Openstack的高能物理虚拟计算集群系统及应用
Openstack-based Virtualized Computing Cluster and Application for High Energy Physics
计算机科学, 2017, 44(10): 59-63. https://doi.org/10.11896/j.issn.1002-137X.2017.10.011
[9] 钱迎进,李永刚,汪毅,周琳琦.
Lustre文件系统元数据服务恢复机制的改进
Improvement of Recovery Mechanism for Lustre Metadata Service
计算机科学, 2015, 42(9): 177-182. https://doi.org/10.11896/j.issn.1002-137X.2015.09.034
[10] 王一超,秦强,施忠伟,林新华.
在Intel Knights Corner和NVIDIA Kepler架构上OpenACC的性能可移植性分析
Performance Portability Evaluation for OpenACC on Intel Knights Corner and NVIDIA Kepler
计算机科学, 2015, 42(1): 75-78. https://doi.org/10.11896/j.issn.1002-137X.2015.01.017
[11] 黄秋兰,李莎,程耀东,陈刚.
高能物理计算环境中KVM虚拟机的性能优化与应用
Performance Optimization and Application of KVM in HEP Computing Environment
计算机科学, 2015, 42(1): 67-70. https://doi.org/10.11896/j.issn.1002-137X.2015.01.015
[12] 程耀东,汪璐,黄秋兰,陈刚.
高能物理计算环境中存储系统的设计与优化
Design and Optimization of Storage System in HEP Computing Environment
计算机科学, 2015, 42(1): 54-58. https://doi.org/10.11896/j.issn.1002-137X.2015.01.012
[13] 李智佳,胡翔,焦莉,王伟锋.
基于随机Petri网的高性能计算系统作业调度及InfiniBand网络互连的性能分析
Performance Evaluation of Job Scheduling and InfiniBand Network Interconnection in High Performance Computing System Based on Stochastic Petri Nets
计算机科学, 2015, 42(1): 33-37. https://doi.org/10.11896/j.issn.1002-137X.2015.01.007
[14] 李静梅,王雪,吴艳霞.
一种改进的优先级列表任务调度算法
Improved Priority List Task Scheduling Algorithm
计算机科学, 2014, 41(5): 20-23. https://doi.org/10.11896/j.issn.1002-137X.2014.05.004
[15] 王文义,王春霞,王杰.
基于CMP多核集群的混合并行编程技术研究
Research on Hybrid Parallel Programming Technique Based on CMP Multi-cure Cluster
计算机科学, 2014, 41(2): 19-22.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!