Large Eddy Simulations of an Airfoil in Turbulent

Large Eddy Simulations of an Airfoil in Turbulent Inflow Lasse Gilling, lg@civil. aau. dk Department of Civil Engineering, Aalborg University, Denmark Extended visit at Chalmers, Sweden 1

Numerical Setup • • • The flow past a NACA 0015 airfoil is studied Turbulence intensity (TI) and angle of attack (AOA) is varied The Ellip. Sys 3 D flow solver is used Wall boundary layer is modeled by detached eddy simulation The k-ω SST turbulence model is used Inlet BC Symmetry BC Flow dir. c AOA 3 c Outlet BC ~4 c Symmetry BC 2

Results: Lift and Drag 1. 5 2 D RANS LES, TI=0. 0 % LES, TI=0. 5 % LES, TI=2. 0 % Measurements 0. 3 0. 25 1 C L C D 0. 2 0. 15 0. 1 2 D RANS LES, TI=0. 0 % LES, TI=0. 5 % LES, TI=2. 0 % Measurements 0 0 2 4 6 8 10 12 Angle of attack [deg] 14 16 18 0. 05 20 0 0 2 4 6 8 10 12 Angle of attack [deg] 14 16 18 20 • Flow sensitive to turbulence • LES with no inflow turbulence predicts stall too late • LES with 0. 5 % TI gives good agreement before stall • LES with 2. 0 % TI gives poor results for low AOA but better after stall • 2 D RANS is good for low AOA, but fails to predict stall 3

Results: Surface Pressure • • Good agreement in general Flow very sensitive at 16° AOA 7 LES, TI=0. 0 % LES, TI=0. 5 % LES, TI=2. 0 % Measurements 6 5 AOA=14 deg -c p 4 3 2 1 0 -1 7 LES, TI=0. 0 % LES, TI=0. 5 % LES, TI=2. 0 % Measurements 6 0. 2 0. 3 0. 4 0. 5 x/c 0. 6 0. 7 0. 8 0. 9 1 LES, TI=0. 0 % LES, TI=0. 5 % LES, TI=2. 0 % Measurements 6 5 AOA=16 deg 3 3 2 2 1 1 0 0. 1 0. 2 0. 3 0. 4 0. 5 x/c AOA=18 deg 4 -c p 0. 1 7 5 -1 0 0. 6 0. 7 0. 8 0. 9 1 -1 0 0. 1 0. 2 0. 3 0. 4 0. 5 x/c 0. 6 0. 7 0. 8 0. 9 1 4

Flow Visualization 1. 5 TI=0. 1 % 1. 4 g d a c C L 1. 3 h 1. 2 i b f 1. 1 e 1 20 40 60 80 100 120 140 160 180 t. U/c • AOA=16° and TI=0. 1 % • Low separation gives high lift and vice versa Large variations in spanwise direction and time • 5

Future Plans • Implement a momentum disc approach of imposing the inflow turbulence • Save mesh points upstream of the airfoil • Reduce computational costs • Investigate influence of turbulence length scale 6