TY - GEN
T1 - Significance of three-dimensional unsteady flows inside the cavity on separated- flow control around an NACA0015 using a synthetic jet
AU - Abe, Yoshiaki
AU - Okada, Koichi
AU - Sato, Makoto
AU - Nonomura, Taku
AU - Fujii, Kozo
PY - 2013
Y1 - 2013
N2 - The simulation of a separation control using a synthetic jet around an NACA0015 airfoil at Reynolds number 63,000 is conducted by a large-eddy simulation (LES) with a compact difference scheme. The synthetic jet is installed at a leading edge and actuated with nondimensional frequencies F+ = 1:0 and 6:0, which is numerically modeled by a threedimensional deforming cavity: "Cavity model" and a two-dimensional boundary condition on the airfoil: " Bc model". The aerodynamic coefficients of the controlled flows are similarly recovered from those of the separated flow using both of the Cavity and Bc model. However, the time-averaged values and flow fields are significantly different in two models, and the use of Bc model on the three-dimensional analysis is not proper. In the case with F+ = 6, a turbulent transition near the leading edge occurs much earlier with the Cavity model than the Bc model. This result indicates that the spanwise disturbance from the cavity to the separated shear layer should be carefully considered when three-dimensional unsteady analysis is conducted by LES.
AB - The simulation of a separation control using a synthetic jet around an NACA0015 airfoil at Reynolds number 63,000 is conducted by a large-eddy simulation (LES) with a compact difference scheme. The synthetic jet is installed at a leading edge and actuated with nondimensional frequencies F+ = 1:0 and 6:0, which is numerically modeled by a threedimensional deforming cavity: "Cavity model" and a two-dimensional boundary condition on the airfoil: " Bc model". The aerodynamic coefficients of the controlled flows are similarly recovered from those of the separated flow using both of the Cavity and Bc model. However, the time-averaged values and flow fields are significantly different in two models, and the use of Bc model on the three-dimensional analysis is not proper. In the case with F+ = 6, a turbulent transition near the leading edge occurs much earlier with the Cavity model than the Bc model. This result indicates that the spanwise disturbance from the cavity to the separated shear layer should be carefully considered when three-dimensional unsteady analysis is conducted by LES.
UR - http://www.scopus.com/inward/record.url?scp=85088182353&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85088182353&partnerID=8YFLogxK
U2 - 10.2514/6.2013-2748
DO - 10.2514/6.2013-2748
M3 - Conference contribution
AN - SCOPUS:85088182353
SN - 9781624102141
T3 - 43rd Fluid Dynamics Conference
BT - 43rd Fluid Dynamics Conference
PB - American Institute of Aeronautics and Astronautics Inc.
T2 - 43rd AIAA Fluid Dynamics Conference
Y2 - 24 June 2013 through 27 June 2013
ER -