airfoil at low Reynlods

fulkrum · April 17, 2009, 08:00

Hi!
I have to draw a polar, or at least a cl - alpha plot (in order to evaluate the Clmax) of an airfoil (Eppler 420 or similar) at Reynolds of about (1.5-3)*10^5.

Xfoil,the Eppler's program and similar (base on potential + boundary layer) are not much reliable due to the laminar flow separation bubble.

A CFD, with 2D and incompressible flow, would do a better job with reasonable calculation resources.
However, with CFD, will be the results reliable? At what extent?
Which is the best turbulence modeling to adopt?
Are there some free codes well suited for this task?
Thank you

A_Lyaskin · April 17, 2009, 12:33

To get reliable CFD data you'll need a turbulence model capable for simulating free transition.

The other problem is that such kind of flow is very sensitive to external conditions - look for experimental data and you'll see that there's a huge data scattering between different wind tunnels. So the questions is - what will be the criteria for reliability of the results?

vishyaroon · April 17, 2009, 17:06

I had a comment for the question you have posted. You are carrying out a 2D calculation for the airfoil. I have come across studies that show that even for 2D studies a 2D/3D calculation (convert 2D case to 3D with periodic BCs in the spanwise direction) can predict certain characteristics better. For example if you are planning to analyze the shear layer, it is better if you do a 2D/3D calculation as compared to 2D, as 2D supresses the vortex interaction in the third dimension.

harishg · April 18, 2009, 02:26

If you want to study the LSB, I would suggest you to use the Spalart-Allmaras model. The model has been specifically developed for external flows with separation. If you need unsteady flow field data, you can use the DES version of the model. The model has been implemented in OpenFOAM.

Regarding reliability of results you would find abundant data on measurements for airfoils and you can validate with any one of the different shapes. NACA 0012/0015 have lot of data available for low Re.

fulkrum · April 18, 2009, 07:39

1)
It is true that even wind tunnel experimentation show an important scattering, since the problem (laminar separation - transition - boundary layer reattachment) is very sensitive to external conditions.
In fact, I should first answer myself what would I consider reliable.
I would say to evaluate the clmax the airfoil is capable of, and at the same time have an estimation of the cd at clmax.
However the great sensitivity to external conditions may avoid any certainty a priori.
I don't know if the Eppler e420 FX doesn't present a strong sensitivity around 1.5-3 Reynlods.
However, in order to obtain more reliable results, even if not always the best results, I would consider to add something "disturbing" the flux in order to achieve a turbulent flow early (the model I am designing has to just to take off with the maximum possible weight, and after a very very brief flight it should land, thus the design point is near CL = CLmax).
In such a situation, simulation with turbulent flow would give "good" Clmax estimation (and at least an acceptable one for the drag).

2)
It's true that turbulence has a tree-dimensional nature, thus a bidimensional simulation would not comply with reality. Thank you, I forgot this :-)

3)
Thank you very much for the technicl advices!

April 17, 2009, 08:00	airfoil at low Reynlods	#1
fulkrum New Member Rauno Join Date: Apr 2009 Location: San Diego (USA); Pisa (Italy); Munich(Germany) Posts: 10 Rep Power: 17	Hi! I have to draw a polar, or at least a cl - alpha plot (in order to evaluate the Clmax) of an airfoil (Eppler 420 or similar) at Reynolds of about (1.5-3)*10^5. Xfoil,the Eppler's program and similar (base on potential + boundary layer) are not much reliable due to the laminar flow separation bubble. A CFD, with 2D and incompressible flow, would do a better job with reasonable calculation resources. However, with CFD, will be the results reliable? At what extent? Which is the best turbulence modeling to adopt? Are there some free codes well suited for this task? Thank you

April 17, 2009, 17:06	2D vs 3D	#3
vishyaroon Senior Member Aroon Join Date: Apr 2009 Location: Racine WI Posts: 148 Rep Power: 17	I had a comment for the question you have posted. You are carrying out a 2D calculation for the airfoil. I have come across studies that show that even for 2D studies a 2D/3D calculation (convert 2D case to 3D with periodic BCs in the spanwise direction) can predict certain characteristics better. For example if you are planning to analyze the shear layer, it is better if you do a 2D/3D calculation as compared to 2D, as 2D supresses the vortex interaction in the third dimension.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Low Speed Airfoil	Mancusi	FLUENT	7	April 3, 2014 07:11
2D Low Speed Airfoil Problem when altering Inlet	mike wilson	CFX	12	August 3, 2010 12:06
Low Reynolds number airfoil.	Pablo Cornejo	FLUENT	14	October 19, 2005 10:41
About low Reynolds number airfoil experiment data.	zqnwpu	Main CFD Forum	5	December 25, 2004 04:52
Multicomponent fluid	Andrea	CFX	2	October 11, 2004 06:12

April 17, 2009, 12:33		#2
A_Lyaskin Member Anton Lyaskin Join Date: Mar 2009 Location: Samara, Russia Posts: 66 Rep Power: 17	To get reliable CFD data you'll need a turbulence model capable for simulating free transition. The other problem is that such kind of flow is very sensitive to external conditions - look for experimental data and you'll see that there's a huge data scattering between different wind tunnels. So the questions is - what will be the criteria for reliability of the results?

April 18, 2009, 02:26		#4
harishg Senior Member N/A Join Date: Mar 2009 Posts: 189 Rep Power: 17	If you want to study the LSB, I would suggest you to use the Spalart-Allmaras model. The model has been specifically developed for external flows with separation. If you need unsteady flow field data, you can use the DES version of the model. The model has been implemented in OpenFOAM. Regarding reliability of results you would find abundant data on measurements for airfoils and you can validate with any one of the different shapes. NACA 0012/0015 have lot of data available for low Re.

April 18, 2009, 07:39		#5
fulkrum New Member Rauno Join Date: Apr 2009 Location: San Diego (USA); Pisa (Italy); Munich(Germany) Posts: 10 Rep Power: 17	1) It is true that even wind tunnel experimentation show an important scattering, since the problem (laminar separation - transition - boundary layer reattachment) is very sensitive to external conditions. In fact, I should first answer myself what would I consider reliable. I would say to evaluate the clmax the airfoil is capable of, and at the same time have an estimation of the cd at clmax. However the great sensitivity to external conditions may avoid any certainty a priori. I don't know if the Eppler e420 FX doesn't present a strong sensitivity around 1.5-3 Reynlods. However, in order to obtain more reliable results, even if not always the best results, I would consider to add something "disturbing" the flux in order to achieve a turbulent flow early (the model I am designing has to just to take off with the maximum possible weight, and after a very very brief flight it should land, thus the design point is near CL = CLmax). In such a situation, simulation with turbulent flow would give "good" Clmax estimation (and at least an acceptable one for the drag). 2) It's true that turbulence has a tree-dimensional nature, thus a bidimensional simulation would not comply with reality. Thank you, I forgot this :-) 3) Thank you very much for the technicl advices!