TY - GEN
T1 - GPU-Accelerated Implicit Large Eddy Simulation of a NACA 0018 Airfoil with Active Flow Control
AU - Laufer, Michael
AU - Greenblatt, David
AU - Frankel, Steven Howard
N1 - Publisher Copyright:
© 2022, American Institute of Aeronautics and Astronautics Inc, AIAA. All rights reserved.
PY - 2022
Y1 - 2022
N2 - In this work, high-order Computational Fluid Dynamics (CFD) simulations have been performed for a NACA 0018 airfoil with an active flow control slot at Reynolds numbers between = 1.25 · 105 and = 2.5 · 105 using the GPU-accelerated PyFR solver. Pressure distributions and resulting aerodynamic loads show better accuracy compared to previous numerical simulation attempts, and demonstrate the first numerical result to resolve the nonlinear lift curves seen in the experiments. In particular, the importance of accurately resolving the Laminar Separation Bubble, taking into account the 3D control slot, as well as the wind tunnel geometry was found to have a large impact on results. The results in this work indicate that a high-order turbulence-resolving solver with the ability to handle complex geometries (unstructured grids) is required in order to accurately predict loads for low Reynolds number cases such as this. Lastly, an overview and performance-to-cost analysis of CFD using GPUs is shown, with conclusions demonstrating the utility of the use of GPUs in the next generation turbulence resolving solvers.
AB - In this work, high-order Computational Fluid Dynamics (CFD) simulations have been performed for a NACA 0018 airfoil with an active flow control slot at Reynolds numbers between = 1.25 · 105 and = 2.5 · 105 using the GPU-accelerated PyFR solver. Pressure distributions and resulting aerodynamic loads show better accuracy compared to previous numerical simulation attempts, and demonstrate the first numerical result to resolve the nonlinear lift curves seen in the experiments. In particular, the importance of accurately resolving the Laminar Separation Bubble, taking into account the 3D control slot, as well as the wind tunnel geometry was found to have a large impact on results. The results in this work indicate that a high-order turbulence-resolving solver with the ability to handle complex geometries (unstructured grids) is required in order to accurately predict loads for low Reynolds number cases such as this. Lastly, an overview and performance-to-cost analysis of CFD using GPUs is shown, with conclusions demonstrating the utility of the use of GPUs in the next generation turbulence resolving solvers.
UR - http://www.scopus.com/inward/record.url?scp=85122947127&partnerID=8YFLogxK
U2 - 10.2514/6.2022-0471
DO - 10.2514/6.2022-0471
M3 - ???researchoutput.researchoutputtypes.contributiontobookanthology.conference???
AN - SCOPUS:85122947127
SN - 9781624106316
T3 - AIAA Science and Technology Forum and Exposition, AIAA SciTech Forum 2022
BT - AIAA SciTech Forum 2022
T2 - AIAA Science and Technology Forum and Exposition, AIAA SciTech Forum 2022
Y2 - 3 January 2022 through 7 January 2022
ER -