JAXA Supercomputer System Annual Report February 2023-January 2024

Report Number: R23EDA201G24

Subject Category: Aeronautical Technology

PDF available here

• Responsible Representative: Kazuyuki Nakakita, Aviation Technology Directorate, Aircraft Lifecycle Innovation Hub
• Contact Information: Andrea Sansica(sansica.andrea@jaxa.jp)
• Members: David Lusher, Yuya Ohmichi, Andrea Sansica, Markus Zauner

Abstract

FaSTAR’s default routine calculations were accelerated with OpenACC. We analyze High-Lift configuration of the NASA CRM aircraft (CRM-HL) at low-speed stall conditions using GPU-based FaSTAR. For buffet analysis, we performed GPU-based calculations using the OpenSBLI code.

Reference URL

N/A

Reasons and benefits of using JAXA Supercomputer System

To develop a GPU version of CFD solvers, the GPU nodes available on JSS3 can be used to verify the code and run large scale simulations

Achievements of the Year

About FaSTAR-GPU: Acceleration and optimizion of FaSTAR on GPU using OpenACC. The results of the multi-GPU version of the code and those of the multi-CPU version of the code are in excellent agreement (figure 1). For a typical setup, the multi-GPU version of the code allows a speedup of 2.

About OpenSBLI: OpenSBLI has been developed and used for wide-span transonic buffet simulations on up to 10 billion mesh points. World first scale-resolving high-fidelity simulations of wide-span buffet were performed up to aspect ratio of AR=3 (Figure 2). These simulations were awarded the Numerical Simulation Grand Prize at ANSS conference 2023 [Grand Prize (Aerospace Numerical Simulation Technology Category) Issued by 41st Aerospace Numerical Simulation Technology Symposium]

Fig.1: Comparison between FaSTAR-GPU and FaSTAR-CPU results on the NASA CRM-HL full-aircraft configuration.

Fig.2: Transonic Buffett OpenSBLI simulation for NASA-CRM wing on JSS3 TOKI-RURI GPU nodes.

Publications

– Peer-reviewed papers

[1] DJ Lusher, A Sansica, A Hashimoto. Effect of Tripping and Domain-Width on Transonic Buffet on Periodic NASA-CRM Airfoils. arXiv preprint arXiv:2401.14793. AIAA-J (Under review, 2024).

[2] David J. Lusher, Andrea Sansica, Neil D. Sandham, Jiangping Meng, Atsushi Hashimoto. OpenSBLI v3.0: High-Fidelity Multi-Block Transonic Aerofoil CFD Simulations using Domain Specific Languages on GPUs. Computer Physics Communications (Under review, 2024).

– Non peer-reviewed papers

[1] P. Zehner, A. Sansica, A. Hashimoto. Investigation of the Mesh and Flow Influence on GPU Performance of Time Integration Methods for an Unstructured CFD Solver. ANSS Conference Tokyo, 2023.

[2] P. Zehner, A. Sansica, A. Hashimoto. Contribution of JAXA to APC-9 using GPU-accelerated FaSTAR: free-air and in-tunnel RANS calculations for CLmax prediction on the CRM high-lift configuration. ANSS Conference Tokyo, 2023.

[3] D.J Lusher, A. Sansica, M.Zauner, A. Hashimoto. High-fidelity simulations of transonic buffet on wide-span airfoils in the OpenSBLI automatic code-generation framework on GPUs. ANSS Conference Tokyo, 2023.

[4] A. Sansica, D.J Lusher, A. Hashimoto. Mach Evolution of the Cylinder Wake Flow Bifurcations. International Symposium on Shockwaves ISSW34. Daegu, Korea, 2023.”

– Invited Presentations

[1] D.J. Lusher, N.D. Sandham. OpenSBLI/OPS workshop Training course: OpenSBLI automatic code- generation framework. UK Turbulence Consortium, Oxford. United Kingdom 2023.

– Oral Presentations

[1] P. Zehner, A. Sansica, A. Hashimoto. Investigation of the Mesh and Flow Influence on GPU Performance of Time Integration Methods for an Unstructured CFD Solver. ANSS Conference Tokyo, 2023.

[2] P. Zehner, A. Sansica, A. Hashimoto. Contribution of JAXA to APC-9 using GPU-accelerated FaSTAR: free-air and in-tunnel RANS calculations for CLmax prediction on the CRM high-lift configuration. ANSS Conference Tokyo, 2023.

[3] D.J Lusher, A. Sansica, M.Zauner, A. Hashimoto. High-fidelity simulations of transonic buffet on wide-span airfoils in the OpenSBLI automatic code-generation framework on GPUs. ANSS Conference Tokyo, 2023.

[4] A. Sansica, D.J Lusher, A. Hashimoto. Mach Evolution of the Cylinder Wake Flow Bifurcations. International Symposium on Shockwaves ISSW34. Daegu, Korea, 2023.

[5] D.J Lusher, A. Sansica, A. Hashimoto. High-fidelity simulations of wide- span transonic airfoil buffet on multi-GPUs using automatic code-generation. JSFM, CFD37 Symposium Nagoya, 2023.”

Usage of JSS

Computational Information

• Process Parallelization Methods: GPU
• Thread Parallelization Methods: N/A
• Number of Processes: 128
• Elapsed Time per Case: 100 Hour(s)

JSS3 Resources Used

Fraction of Usage in Total Resources*1(%): 4.12

Details

Please refer to System Configuration of JSS3 for the system configuration and major specifications of JSS3.

*1: Fraction of Usage in Total Resources: Weighted average of three resource types (Computing, File System, and Archiver).

*2: Fraction of Usage：Percentage of usage relative to each resource used in one year.

ISV Software Licenses Used

*2: Fraction of Usage：Percentage of usage relative to each resource used in one year.

JAXA Supercomputer System Annual Report February 2023-January 2024