High Lift Airfoils At High Angles Of Attack

Luiz Pancini · April 6, 2015, 09:22

Hello,

I've been trying to simulate high lift airfoils generated by an airfoil optimizer program, ranging from negative angles of attack (AoA) to stall, at low Reynolds number (250 thousands) using Fluent.

In order to validate my results, I have simulated a well known high lift airfoil, the Selig 1223.

Though my cl, cd and cm results are accurate enough at low and medium angles of attack (i.g. from -2º to 12º), I was unable to get reasonable values at +12º. Fluent underpredicts both the angle of stall and maximum cl, while overpredicting cd.

I know that drag overpredictions are common, besides, I am using a turbulent model, while the wind tunnel tests contain at least some laminar flow. But, I still would expect a better value for the cl, even at these high angles of attack.

Now, about the settings:

I have tried the pressure-based Spalart-Allmaras (SA), k-e, k-w, and even SST. My best results came from SA, while in many cases SST didn't even converge. I tried both steady and transient, transient came up a bit better in some cases, Second Order Upwind, Third-Order Muscl and Quick formulations, all give about the same results.

I used a very fine 2D C-grid, with 1.97 million cells, y+ lower than 1 everywhere in the airfoil. Maximum aspect ratio of 916 and minimum orthogonal quality of 0.67. I did it the in ICEM, the way Simon shows in this youtube tutorial: https://www.youtube.com/watch?v=tYrbScUH9RE. Here goes some pictures of it:

http://s18.postimg.org/oasua5uzd/icem_mesh.png
http://s3.postimg.org/rum73srs3/leading_edge.png

For the SA model, the drag sensitivity to the domain extent showed that at 60 chords the values deviated up to 10% at medium AoA, while lift was up to 5% discrepancy, even for a 15 chord. For a 15 chord domain, lift results were off by a maximum of 3%, until 12º AoA.

If anyone knows how to improve the solution at high AoA (including stall), which is my main objective, I would really appreciate your help.

Maze · April 8, 2015, 21:57

The problems solved？

Luiz Pancini · April 9, 2015, 09:01

Not yet. Trying some 3D now.

April 6, 2015, 09:22	High Lift Airfoils At High Angles Of Attack	#1
Luiz Pancini New Member Luiz Pancini Join Date: Jan 2015 Posts: 3 Rep Power: 11	Hello, I've been trying to simulate high lift airfoils generated by an airfoil optimizer program, ranging from negative angles of attack (AoA) to stall, at low Reynolds number (250 thousands) using Fluent. In order to validate my results, I have simulated a well known high lift airfoil, the Selig 1223. Though my cl, cd and cm results are accurate enough at low and medium angles of attack (i.g. from -2º to 12º), I was unable to get reasonable values at +12º. Fluent underpredicts both the angle of stall and maximum cl, while overpredicting cd. I know that drag overpredictions are common, besides, I am using a turbulent model, while the wind tunnel tests contain at least some laminar flow. But, I still would expect a better value for the cl, even at these high angles of attack. Now, about the settings: I have tried the pressure-based Spalart-Allmaras (SA), k-e, k-w, and even SST. My best results came from SA, while in many cases SST didn't even converge. I tried both steady and transient, transient came up a bit better in some cases, Second Order Upwind, Third-Order Muscl and Quick formulations, all give about the same results. I used a very fine 2D C-grid, with 1.97 million cells, y+ lower than 1 everywhere in the airfoil. Maximum aspect ratio of 916 and minimum orthogonal quality of 0.67. I did it the in ICEM, the way Simon shows in this youtube tutorial: https://www.youtube.com/watch?v=tYrbScUH9RE. Here goes some pictures of it: http://s18.postimg.org/oasua5uzd/icem_mesh.png http://s3.postimg.org/rum73srs3/leading_edge.png For the SA model, the drag sensitivity to the domain extent showed that at 60 chords the values deviated up to 10% at medium AoA, while lift was up to 5% discrepancy, even for a 15 chord. For a 15 chord domain, lift results were off by a maximum of 3%, until 12º AoA. If anyone knows how to improve the solution at high AoA (including stall), which is my main objective, I would really appreciate your help. Maze likes this.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Airfoils and lift compared to air density and TE	Jens	Main CFD Forum	0	May 4, 2006 04:31
High Lift Devices	Anshul	Main CFD Forum	5	June 21, 2005 08:10
High Angle of Attack	w0er	Main CFD Forum	3	March 8, 2005 12:05
High lift analysis using overset	kei-tee	Main CFD Forum	0	April 10, 2003 10:53
grids around airfoils at high aoa	uygun	Main CFD Forum	9	October 11, 2002 09:30

April 8, 2015, 21:57		#2
Maze New Member Join Date: Jan 2015 Posts: 1 Rep Power: 0	The problems solved？

April 9, 2015, 09:01		#3
Luiz Pancini New Member Luiz Pancini Join Date: Jan 2015 Posts: 3 Rep Power: 11	Not yet. Trying some 3D now.