JAXA Supercomputer System Annual Report February 2025-January 2026

Report Number: R25EG3205

Subject Category: Research and Development

PDF (to be added)

• Responsible Representative: Hideaki Nanri, Research and Development Directorate, Research Unit IV
• Contact Information: Shun Takahashi Research and Development Directorate, Research Unit IV(takahashi.shun@jaxa.jp)
• Members: Masaaki Fukui, Susumu Hasegawa, Taku Inoue, Masatoshi Kodera, Masaru Koga, Haruyuki Kato, Toshihiko Munakata, Takayuki Nagata, Yoichi Ohnishi, Tetsuji Ogawa, Takafumi Tano, Sadatake Tomioka, Kenta Tsukuda, Masahiro Takahashi, Shun Takahashi

Abstract

In this project, we conduct engine combustion analysis, aerodynamic analysis of the vehicle, thermal-structural analysis of the components, and flight analysis of the vehicle, all in connection with the technological advancement of the air-breathing propulsion system, which is one of the future transportation systems.

Reference URL

N/A

Reasons and benefits of using JAXA Supercomputer System

The project aims to advance the technology of air-breathing transport systems—one of the future modes of transportation-by conducting analyses such as engine combustion analysis, vehicle aerodynamic analysis, thermal-structural analysis of components, and vehicle flight analysis. However, many of these analyses involve extremely high computational costs. Furthermore, since the analyses will be performed using tools such as FaSTAR, developed by JAXA, the project is highly compatible with JAXA's supercomputers. Based on this background and these reasons, there is great significance in utilizing JAXA's supercomputers.

Achievements of the Year

To improve the performance and expand the operational range of scramjet engines, we are studying 3D-shaped air inlets, which are highly efficient. Using the 3D-shaped inlet design tool developed last fiscal year, we designed and evaluated a REST (Rectangular-to-Elliptic Shape Transitioning) inlet using CFD as shown in Fig. 1. In this design, the flow-path cross-section transitions continuously from rectangular to elliptical. Shown are an example of wall pressure distribution normalized by the main flow static pressure and Mach number contours within the central symmetry plane. We gained knowledge about the basic flow characteristics of the 3D-shaped inlet and the impact of angle of attack on internal flow. Based on the CFD evaluation, we then selected a candidate shape and fabricated the wind tunnel test model. We plan to conduct wind tunnel tests on the model soon.

Also, in this study, 18 pressure measurement points (Fig. 2) are installed on an experimental model to investigate optimal combinations of sensor locations and machine learning algorithms capable of accurately predicting flight conditions. First, wind tunnel experiments (Fig. 3) are conducted to obtain pressure measurement data over a wide range of Mach numbers. Based on these results, numerical simulations are performed, and their validity is confirmed through comparisons with the experimental data. Furthermore, by combining machine learning algorithms with combinatorial optimization techniques, a high-accuracy flight condition estimation model is developed.

This rearch presents the results of Mach 2 tests in which the angle of attack alpha and sideslip angle beta are varied. A comparison between CFD and experimental pressure measurements confirms the validity of the obtained data. Using the pressure values at each measurement point derived from CFD, Partial Least Squares (PLS) regression is performed, and the Variable Importance in Projection (VIP) is evaluated. The results are shown in Fig. 4.

Fig.1: The wall pressure distribution and the Mach number contours within the central plane of a REST inlet model; the main flow Mach number is 5.45 and the angle of attack is 0 degrees. Wall pressure is shown in dimensionless form based on main flow static pressure.

Fig.2: Experimental model and pressure tap numbering

Fig.3: Schlieren photograph for the Mach 2 test

Fig.4: VIP ranking for the prediction of angle of attack (left) and sideslip angle (right)

Publications

- Oral Presentations

Shun Takahashi, TatsushiIsono, Takuo Onodera, Masao Takegoshi, Shuto Yatsuyanagi, SadatakeTomioka, "Numerical simulation of instability in hydrocarbon fuel flow under supercritical conditions", The 35thInternational Symposium on Space Technology and Science (ISTS), 2025

Shun Takahashi, Takayuki Nagata, Kouichiro Tani, Masaru Koga, Tatsushi Isono, Masao Takegoshi, Sadatake Tomioka, Haruyuki Kato, Kenta Tsukuda, Daisuke Sasaki, "Hypersonic vehicle design based on mode decomposition", AIAA SciTech 2026, 2026

Usage of JSS

Computational Information

• Process Parallelization Methods: MPI
• Thread Parallelization Methods: N/A
• Number of Processes: 256 - 1800
• Elapsed Time per Case: 2 Hour(s)

JSS3 Resources Used

Fraction of Usage in Total Resources*1(%): 0.51

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 2025-January 2026