JAXA Supercomputer System Annual Report April 2020-March 2021

Report Number: R20EACA37

Subject Category: JSS Inter-University Research

PDF available here

• Responsible Representative: Soshi Kawai, Professor, Department of Aerospace Engineering, Tohoku University
• Contact Information: Hiroyuki Asada(h.asada@tohoku.ac.jp)
• Members: Soshi Kawai, Ryo Kamogawa, Hiroyuki Asada, Aya Aihara

Abstract

To accurately simulate turbulent flows with separation and reattachment using wall-modeled large-eddy simulation (LES), non-equilibrium effects, such as the effects of the pressure gradient, should be incorporated in the wall model. However, the existing non-equilibrium wall-models require solving the partial differential equations (PDEs), which increase the difficulty of implementation, especially for complex geometries. Therefore, we construct and validate a non-equilibrium wall-model based on the ordinary differential equation (ODE). First, the non-equilibrium effects of turbulent boundary layers (i.e., the pressure-gradient and convective terms) and the eddy viscosity are consistently modeled using the wall-resolved LES database of the flat-plate separating/reattaching turbulent boundary layers. Then, we perform wall-modeled LES of the flat-plate turbulent boundary layer with separation/reattachment and a near-stall airfoil flow using the proposed ODE-based non-equilibrium wall model. In both the validation cases, the wall-modeled LES using the proposed wall model show a good agreement with the reference wall-resolved LES results.

Reference URL

Please refer to http://www.klab.mech.tohoku.ac.jp/ .

Reasons and benefits of using JAXA Supercomputer System

In this study, a newly-proposed wall model is validated. Since the wall-modeled LES is an unsteady 3D calculation, parallel computations using a supercomputer are required. Also, to propose and validate a new wall model, several computational cases with different settings need to be conducted. Therefore, our research requires massively parallel computations using JAXA’s supercomputer.

Achievements of the Year

(a)An ODE-based non-equilibrium wall model is constructed based on the wall-resolved LES database of the flat-plate separating/reattaching turbulent boundary layers. In the proposed wall model, the non-equilibrium terms, pressure-gradient and convective terms, are modeled using the flow quantities at the matching location of the wall model. Also, the eddy viscosity is modeled consistently with the pressure-gradient and convective terms. The implementation of the proposed model is easier than the existing PDE-based non-equilibrium wall model. Then, we conduct the wall-modeled LES using the proposed ODE-based wall model (ODE-NEQWM) for the flat-plate separating turbulent boundary layers and compare the results with the wall-resolved LES database as well as the wall-modeled LES using the existing PDE-based non-equilibrium wall model (PDE-NEQWM). The mean streamwise velocity and Reynolds shear stress profiles obtained by the proposed ODE-NEQWM show a good agreement with those of LES and PDE-NEQWM. The results indicate that the proposed ODE-based non-equilibrium wall model has comparable accuracy with the existing PDE-based wall model.
(b)As a more practical validation case, a high Reynolds number (the Reynolds number based on the chord length is 2.1*10^6) flow around the A-airfoil at a near-stall condition (the angle of attack is 13.3 degrees) is simulated using the proposed wall-modeled LES (ODE-NEQWM). The mean pressure coefficient and the skin-friction coefficient obtained by ODE-NEQWM show a good agreement with the reference data of the wall-resolved LES database and the wall-modeled LES using the existing PDE-based wall model. The results obtained by the proposed wall-modeled LES show a robustness to predict a more practical airfoil flow near a stall condition.

Fig.1: Isosurfaces of the Q-criterion colored by the streamwise velocity in the flat-plate separating/reattaching turbulent boundary layer

Fig.2: Profiles of (a) the mean streamwise velocity and (b) Reynolds shear stress in separating/reattaching turbulent boundary layer (circles, wall-resolved LES; red solid lines, the proposed ODE-based non-equilibrium wall-modeled LES; black dashed lines, the existing PDE-based non-equilibrium wall-modeled LES)

Fig.3: Isosurfaces of the Q-criterion colored by the streamwise velocity in the flow around A-airfoil

Fig.4: Distributions of mean skin friction and pressure coefficients along the A-airfoil (circles, wall-resolved LES; red solid lines, the proposed ODE-based non-equilibrium wall-modeled LES; black dashed lines, the existing PDE-based non-equilibrium wall-modeled LES)

Publications

– Oral Presentations

(1) R. Kamogawa, Y. Tamaki, and S. Kawai, “Modeling of the non-equilibrium effects of separated turbulent boundary layer for ODE-based wall-modeled LES”, 34th CFD symposium, Online, 2020.

Usage of JSS

Computational Information

• Process Parallelization Methods: MPI
• Thread Parallelization Methods: OpenMP
• Number of Processes: 416
• Elapsed Time per Case: 200 Hour(s)

Resources Used(JSS2)

Fraction of Usage in Total Resources*1(%): 0.15

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.

Resources Used(JSS3)

Fraction of Usage in Total Resources*1(%): 0.01

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.

JAXA Supercomputer System Annual Report April 2020-March 2021