[ymc11]

Hello everyone,
I did some simulations of NACA0012 airfoil using SU2. I used the grid in testcases/rans/naca0012/mesh_NACA0012_turb_897*257.su2 and SST turbulent model. The angle of attack is 15deg. However, the cl value(1.499) matchs well with the experiment(1.5), but the cd(0.028845) is much larger than the experiment value(0.018).
Could anyone please help me? Thanks a lot!

[hlk]

Originally Posted by ymc11

Hello everyone,
I did some simulations of NACA0012 airfoil using SU2. I used the grid in testcases/rans/naca0012/mesh_NACA0012_turb_897*257.su2 and SST turbulent model. The angle of attack is 15deg. However, the cl value(1.499) matchs well with the experiment(1.5), but the cd(0.028845) is much larger than the experiment value(0.018).
Could anyone please help me? Thanks a lot!

Did it match at lower angles of attack?
At a 15 degree angle of attack there could easily be separation and unsteady effects in the experimental results. Also, note that turbulence effects are generally difficult to match between experiment and CFD, and will likely require tuning of the turbulence model parameters as well as making sure that the mesh is appropriately refined (look up "y+" on cfd-online if you are not familiar with that).

[DLuo]

Hi Yang,
CLMax for this airfoil in 2D is around 13 degrees. CFD in it's current state, especially with RANS solutions is not well suited for predicting performance in the stall regime. Like Heather mentioned, at that angle there's a reasonable amount of separation and unsteadiness therefore a RANS solution by definition (Reynold averaged NS) wouldn't be the proper method of prediction.
For more reading on the subject I suggest reviewing the results from AIAA High Lift Prediction workshops