JAXA Supercomputer System Annual Report April 2017-March 2018

Report Number: R17ETET05

Subject Category: Skills Acquisition System

PDF available here

• Responsible Representative: Takashi Aoyama, Aeronautical Technology Directorate, Numerical Simulation Research Unit
• Contact Information: Himeko Yamamoto himeko@toki.waseda.jp
• Members: Himeko Yamamoto

Abstract

For the development of a jet engine combustor with high environmental compatibility, we develop the combustion calculation code that can capture the pressure propagation and chemical reaction with practical calculation cost. In addition, a verification calculation of this calculation code is conducted on the scramjet test engine of DLR.

Reference URL

N/A

Reasons for using JSS2

It is necessary to use supercomputer to conduct development and verification of the combustion calculation code.

Achievements of the Year

In this research, the implementation of the combustion calculation code based on compressible flamelet model to a density-based fluid analysis solver, FaSTAR (FAST Aerodynamic Routines) was conducted. Then, an artificial neural network (ANN) model was applied to reduce the calculation cost when referring to the flamelet library. The ANN model has 2 input units (mixture fraction and scalar dissipation rate), 3 hidden units, and output units corresponding to mass fraction of each chemical species (As an example, Fig.1 shows the ANN model of OH mass fraction). Approximation of the flamelet library using this model achieved reduction of memory usage (more than 50%) during numerical simulation compared with conventional linear interpolation. For evaluation of this combustion calculation code in actual combustion field, a numerical simulation of the hydrogen combustion in a scramjet test engine of the German Aerospace Center (DLR) was conducted (Fig.2 shows time-averaged density distribution by ANN model).

Fig.1: OH mass fraction generated by ANN

Fig.2: Time-averaged density distribution [kg/m3]

Publications

■ Non peer-reviewed papers

1) Himeko Yamamoto, Rui Toyonaga, Yasuhiro Mizobuchi, Tetsuya Sato, Numerical analysis of jet lifted flames using G-equation, JAXA Special Publication, 2017.12

2) Rui Toyonaga, Himeko Yamamoto, Yasuhiro Mizobuchi, Tetsuya Sato, Numerical investigation on supersonic combustion in SCRAMJET engine by FaSTAR, JAXA Special Publication, 2017.12

■ Presentations

1) Himeko Yamamoto, Rui Toyonaga, Yasuhiro Mizobuchi, Tetsuya Sato, Numerical study on flame structure of jet lifted flames by FaSTAR, 57th Aerospace Propulsion Conference of Japan Society for Aeronautical and Space Science, 2017.3

2) Himeko Yamamoto, Rui Toyonaga, Yasuhiro Mizobuchi, Tetsuya Sato, Numerical analysis of jet lifted flames using G-equation, 49th Fluid Dynamics Conference/ 35th Aerospace Numerical Simulation Symposium, 2017.6

3) Himeko Yamamoto, Rui Toyonaga, Yusuke Komatsu, Koki Kabayama, Yasuhiro Mizobuchi, Tetsuya Sato, Improvement of compressible flamelet model using artificial neural network, The Asian Joint Conference on Propulsion and Power, 2018.3

Usage of JSS2

Computational Information

• Process Parallelization Methods: MPI
• Thread Parallelization Methods: N/A
• Number of Processes: 2 – 1024
• Elapsed Time per Case: 120.00 hours

Resources Used

Fraction of Usage in Total Resources*1(%): 0.60

Details

Please refer to System Configuration of JSS2 for the system configuration and major specifications of JSS2.

*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.

JAXA Supercomputer System Annual Report April 2017-March 2018