JAXA Supercomputer System Annual Report April 2016-March 2017

Report Number: R16E0062

• Responsible Representative: Seiji Zenitani(National Astronomical Observatory of Japan, Division of Theoretical Astronomy)
• Contact Information: Seiji Zenitani(zenitani@rish.kyoto-u.ac.jp)
• Members: Seiji Zenitani, Iku Shinohara
• Subject Category: Space(Space science,Space philosophy)

Abstract

We conduct scientific researches on basic processes in space plasma environments, by using numerical simulations. We also develop an open-source simulation code.

Goal

To conduct a scientific research on basic processes in space plasmas

Objective

To understand the basic mechanisms of plasma processes, in particular, magnetic reconnection, in space plasmas

References and Links

N/A

Use of the Supercomputer

By using the 'OpenMHD' code (code: ACA02), we set up a magnetic reconnection problem and then study its nonlinear evolution. We consider a sufficiently large system size so that we can observe the transition to the turbulent regime. We will investigate the influence of key parameters that characterize the plasma compressibility.

Necessity of the Supercomputer

Space plasma processes are basically described by magnetohydrodynamics (MHD). However, the MHD equation system is highly nonlinear and complex. Supercomputers are necessary to explore space plasma processes.

Achievements of the Year

We have carried out large-scale MHD simulations, which is large enough to exhibit a lot of turbulent magnetic islands (plasmoids), as shown in Figure 1. We have confirmed that magnetic reconnection evolves faster in the compressible regime. Currently we are figuring out how the speed-up mechanism works.

Fig.1:MHD simulation of large-scale magnetic reconnection(In-plane magnetic field lines and the electric current density)

Publications

Presentations

1) Zenitani, S., High-speed fluid dynamics in magnetic reconnection in a low-β plasma, 'Challenges of AGN jets,' National Astronomical Observatory of Japan, Mitaka, 2017/1/19

Computational Information

• Parallelization Methods: Process Parallelization,Hybrid Parallelization
• Process Parallelization Methods: MPI
• Thread Parallelization Methods: OpenMP
• Number of Processes: 1200
• Number of Threads per Process: 1
• Number of Nodes Used: 38
• Elapsed Time per Case (Hours): 6
• Number of Cases: 10

Resources Used

Total Amount of Virtual Cost(Yen): 24,605

Breakdown List by Resources

Note: Virtual Cost=amount of cost, using the unit price list of JAXA Facility Utilization program(2016)

JAXA Supercomputer System Annual Report April 2016-March 2017