Unable to converge for higher angle of attack for airfoil

Shiberu · December 29, 2015, 11:11

Hello!

Recently I've embarked on a project to study the aerodynamics behavior for the NACA0009 airfoil. I used ansysltd youtube tutorial on ICEM to create a mesh for the related airfoil geometry, and Ansys Fluent to solve it under a k-epsilon model. The solution was able to converge easily at 0 degrees angle of attack, and results were quite satisfactory.

However, when i rotated the geometry by 20 degrees, recreated the meshing and tried solving, it couldn't converge even after a long time. I've been digging around and tried following tweaks to coax it to convergence, yet no avail. Do you guys have any suggestions? I would be more than happy to post details such as my mesh, geometry etc. as I am unable to do it now (posting from my laptop).

Thanks guys!

DrDevice · January 15, 2016, 04:59

In my experience, what tends to happen is as the flow separates more and more off of the trailing edge of the wing (Which is after about 15º for a naca0009) larger recirculating vorticies tend to appear. These tend away from a steady state solution.

There are a couple of things that you can do to help fix it. Firstly, you can try using a different turbulence model. I would recommend Spalart Allmaras for basic analysis, or Reynolds Stress Transport for a more in depth analysis. The k-epsilon model only works well for fully turbulent flows, which you aren't experiencing.

The second thing to do is to increase the mesh size downstream of the wing to try to remove transience by numerical diffusion.

And the last thing is to make sure your mesh sizing on the wing surface is appropriate, and that you fluid domains are large enough.

December 29, 2015, 11:11	Unable to converge for higher angle of attack for airfoil	#1
Shiberu New Member Lester Join Date: Dec 2015 Posts: 1 Rep Power: 0	Hello! Recently I've embarked on a project to study the aerodynamics behavior for the NACA0009 airfoil. I used ansysltd youtube tutorial on ICEM to create a mesh for the related airfoil geometry, and Ansys Fluent to solve it under a k-epsilon model. The solution was able to converge easily at 0 degrees angle of attack, and results were quite satisfactory. However, when i rotated the geometry by 20 degrees, recreated the meshing and tried solving, it couldn't converge even after a long time. I've been digging around and tried following tweaks to coax it to convergence, yet no avail. Do you guys have any suggestions? I would be more than happy to post details such as my mesh, geometry etc. as I am unable to do it now (posting from my laptop). Thanks guys!

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
[OpenFOAM] Plot Angle of Attack Next to Transient Pitching Airfoil	dancfd	ParaView	6	October 24, 2013 01:37
[Netgen] Import netgen mesh to OpenFOAM	hsieh	OpenFOAM Meshing & Mesh Conversion	32	September 13, 2011 06:50
introducing angle of attack on ICEMCFD HEXA	icem beginner	CFX	2	December 24, 2008 12:00
angle of attack with icem hexa	icem beginner	Main CFD Forum	0	December 5, 2008 17:54
angle of attack	kiran	FLUENT	0	September 10, 2004 09:18

January 15, 2016, 04:59		#2
DrDevice New Member Join Date: Feb 2015 Posts: 5 Rep Power: 11	In my experience, what tends to happen is as the flow separates more and more off of the trailing edge of the wing (Which is after about 15º for a naca0009) larger recirculating vorticies tend to appear. These tend away from a steady state solution. There are a couple of things that you can do to help fix it. Firstly, you can try using a different turbulence model. I would recommend Spalart Allmaras for basic analysis, or Reynolds Stress Transport for a more in depth analysis. The k-epsilon model only works well for fully turbulent flows, which you aren't experiencing. The second thing to do is to increase the mesh size downstream of the wing to try to remove transience by numerical diffusion. And the last thing is to make sure your mesh sizing on the wing surface is appropriate, and that you fluid domains are large enough.