TY - GEN
T1 - Hybrid turbulence model simulations of internal and external flows
AU - Gross, A.
AU - Fasel, H. F.
N1 - Publisher Copyright:
© 2015 American Institute of Aeronautics and Astronautics Inc. All rights reserved.
PY - 2014
Y1 - 2014
N2 - The large computational expense of direct numerical simulations provides motivation for the continuous development of new and improved turbulence models. In our research the standard k-ω and k-ε two-equation turbulence models were modified by limiting the turbulence length scale. The resulting hybrid models and a renormalization group hybrid model were tested for a plane channel flow at Reτ= 395 and for the Stanford asymmetric diffuser. All three models were then employed for investigating the separated flow on the suction side of a wing section with a modified NACA643-618 airfoil for Re=300,000 and α=15deg. As a reference, in addition to the hybrid simulations, a direct numerical simulation was carried out. Although plausible results were obtained for the plane channel flow, the diffuser flow simulations were inconclusive. For the wing section, very good agreement with the DNS data was obtained with the RG model. For the two-equation hybrid models, the turbulent suction side boundary layer separated too early compared to the DNS.
AB - The large computational expense of direct numerical simulations provides motivation for the continuous development of new and improved turbulence models. In our research the standard k-ω and k-ε two-equation turbulence models were modified by limiting the turbulence length scale. The resulting hybrid models and a renormalization group hybrid model were tested for a plane channel flow at Reτ= 395 and for the Stanford asymmetric diffuser. All three models were then employed for investigating the separated flow on the suction side of a wing section with a modified NACA643-618 airfoil for Re=300,000 and α=15deg. As a reference, in addition to the hybrid simulations, a direct numerical simulation was carried out. Although plausible results were obtained for the plane channel flow, the diffuser flow simulations were inconclusive. For the wing section, very good agreement with the DNS data was obtained with the RG model. For the two-equation hybrid models, the turbulent suction side boundary layer separated too early compared to the DNS.
UR - http://www.scopus.com/inward/record.url?scp=84938257644&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84938257644&partnerID=8YFLogxK
M3 - Conference contribution
AN - SCOPUS:84938257644
T3 - 52nd Aerospace Sciences Meeting
BT - 52nd Aerospace Sciences Meeting
PB - American Institute of Aeronautics and Astronautics Inc.
T2 - 52nd Aerospace Sciences Meeting 2014
Y2 - 13 January 2014 through 17 January 2014
ER -