JAXA Supercomputer System Annual Report February 2025-January 2026

Report Number: R25ETET01

Subject Category: Skills Acquisition System

PDF (to be added)

• Responsible Representative: Yoshikazu Makino, Aviation Technology Directorate, Re-BooT project Team
• Contact Information: Hiroaki Ishikawa(ishikawa.hiroaki2@jaxa.jp)
• Members: Hiroaki Ishikawa, Ryo Kanazawa, Sowa Moriyama, Ren Nimura, Kaito Shimizu, Kento Sakuma, Takao Tsuchiya, Kazuma Yamanaka

Abstract

The system integration design technologies for achieving low sonic-boom, low aerodynamic drag, low landing and take-off noise, and light weight simultaneously are the key technologies for future supersonic airplanes. JAXA is promoting the R&D for these technologies based on our experiences of demonstrating the advanced low-drag and low-boom design concepts.

Reference URL

Please refer to http://www.aero.jaxa.jp/eng/research/frontier/sst/ .

Reasons and benefits of using JAXA Supercomputer System

To achieve low sonic-boom, low aerodynamic drag, low landing and take-off noise, and light weight simultaneously, the multi-objective optimization tools are utilized in the design study. The super computer is necessary to obtain the efficient evaluation of multiple evaluation indicators.

Achievements of the Year

To enhance the airframe/propulsion integration design of supersonic passenger aircraft, this study investigates duct shape design methodologies for supersonic inlets as well as approaches to mitigate the aerodynamic effects induced by the airframe.

As a potential approach for realizing highly integrated supersonic inlets, it is conceivable to employ a subsonic diffuser with a larger inlet aspect ratio (AR) than those used in conventional designs. To clarify the key design considerations for subsonic diffusers with large inlet AR, computational fluid dynamics (CFD) analyses using FaSTAR were conducted on the JSS3 supercomputing system for a simplified diffuser configuration. A comparison of exit Mach number distributions for diffusers with different inlet ARs (Fig. 1) shows that, when the inlet AR becomes extremely large (e.g., AR = 40), the diffuser's fundamental function-flow diffusion-is not effectively achieved. This indicates the existence of an upper limit to the inlet AR that can be practically applied. Furthermore, total pressure losses generated inside the diffuser were evaluated from the CFD data and compared with estimates obtained using a simplified prediction method developed for diffusers with conventional AR. The results reveal that the rate of increase in total pressure loss with respect to inlet AR is substantially higher than expected (Fig. 2). This behavior is presumed to arise from non-uniform velocity distributions near the diffuser inlet, which violate the common assumption of constant friction coefficients in both the circumferential and streamwise directions.

Additionally, when a supersonic inlet is installed directly on an aircraft wing or fuselage surface to increase the level of airframe integration, the thick boundary layer that develops on these surfaces enters the inlet. Flow separation inside the inlet caused by this boundary‑layer ingestion can significantly degrade propulsion performance. Wind tunnel experiments have demonstrated that the total pressure recovery of the inlet can be improved by placing a pair of vortex generators (VGs) upstream of the inlet compression surface (ramp). To reveal this mechanism, CFD analyses of a simplified model were performed using FaSTAR on JSS3. The results show that the cross flow induced by the VGs thins the boundary layer reaching the central region of the ramp and enhances the spanwise pressure gradient formed on the ramp surface (Fig. 3). This, in turn, facilitates outward turning and discharge of the boundary layer, thereby reducing the amount of boundary‑layer flow entering the inlet entrance plane (Fig. 4).

Fig.1: Subsonic diffuser model and exit Mach number distribution

Fig.2: Variation of total pressure loss with inlet aspect ratio and Mach number distribution near the diffuser inlet

Fig.3: Mach‑number distribution and surface‑pressure distribution around the ramp model

Fig.4: Amount of boundary‑layer ingestion at the cross‑section corresponding to the inlet entrance

Publications

N/A

Usage of JSS

Computational Information

• Process Parallelization Methods: MPI
• Thread Parallelization Methods: Automatic Parallelization
• Number of Processes: 288 - 1536
• Elapsed Time per Case: 6 Hour(s)

JSS3 Resources Used

Fraction of Usage in Total Resources*1(%): 0.17

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