Drag Coefficient undepredicted by CFD?

AamirSultan · February 5, 2019, 07:05

I am simulating a 3D airfoil .

The values of Lift coefficients are within the bounds of 10% error compared to
the experimental values .

However , the values of drag coefficient I am getting have an error of above 20% compared to experimental values .
I am using SA turbulence model .

If the above discrepancy is due to the wake region/flow separation and wing tip vertices . How can I accurately capture these phenomena ?

arungovindneelan · February 5, 2019, 08:55

For airfoil major contributor of drag is skin friction drag if you are dealing with sub-sonic flows. To get accurate result make sure you are using NS-equation. Quality of the grid should be good to get an accurate result and you should resolve the boundary layer especially laminar sub-layer to get accurate drag value. For grid size next to wall estimation, you can consider the following link.
http://www.pointwise.com/yplus/

Then gradually stretch to the boundary of the domain. I personally prefer SST-k-W turbulence model for drag calculations. I believe solution procedures and turbulence models have a little effect if your grid is fine and has good orthogonality, less skewness and uniformly stretched.

My conclusion is: do LES if you have computational power, else do RANS with SST-k-W turbulence model and make sure you are resolving laminar sublayer.

I hope you can get an accurate result if you are able to make good mesh no matter whether you are doing RANS or LES.

AamirSultan · February 5, 2019, 10:39

Quote:

Originally Posted by arungovindneelan

For airfoil major contributor of drag is skin friction drag if you are dealing with sub-sonic flows. To get accurate result make sure you are using NS-equation. Quality of the grid should be good to get an accurate result and you should resolve the boundary layer especially laminar sub-layer to get accurate drag value. For grid size next to wall estimation, you can consider the following link.
http://www.pointwise.com/yplus/

Then gradually stretch to the boundary of the domain. I personally prefer SST-k-W turbulence model for drag calculations. I believe solution procedures and turbulence models have a little effect if your grid is fine and has good orthogonality, less skewness and uniformly stretched.

My conclusion is: do LES if you have computational power, else do RANS with SST-k-W turbulence model and make sure you are resolving laminar sublayer.

I hope you can get an accurate result if you are able to make good mesh no matter whether you are doing RANS or LES.

Thank you so much.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Low Y+ vs High Y+ CFD simulation for determining DRAG	surajp92	Main CFD Forum	2	September 19, 2017 03:15
wrong SU2 calculation for lift and drag coefficient for NAC4421	mechy	SU2	7	January 9, 2017 05:18
CFD Design...The CFD Future	John C. Chien	Main CFD Forum	20	November 19, 2015 23:40
Automotive test case	vinz	OpenFOAM Running, Solving & CFD	98	October 27, 2008 08:43
drag and lift coefficient	Noé	Siemens	5	July 13, 2004 10:21

February 5, 2019, 08:55		#2
arungovindneelan Member AGN Join Date: Dec 2011 Posts: 70 Rep Power: 14	For airfoil major contributor of drag is skin friction drag if you are dealing with sub-sonic flows. To get accurate result make sure you are using NS-equation. Quality of the grid should be good to get an accurate result and you should resolve the boundary layer especially laminar sub-layer to get accurate drag value. For grid size next to wall estimation, you can consider the following link. http://www.pointwise.com/yplus/ Then gradually stretch to the boundary of the domain. I personally prefer SST-k-W turbulence model for drag calculations. I believe solution procedures and turbulence models have a little effect if your grid is fine and has good orthogonality, less skewness and uniformly stretched. My conclusion is: do LES if you have computational power, else do RANS with SST-k-W turbulence model and make sure you are resolving laminar sublayer. I hope you can get an accurate result if you are able to make good mesh no matter whether you are doing RANS or LES.