CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   Main CFD Forum (http://www.cfd-online.com/Forums/main/)
-   -   2D-simulation of flow (http://www.cfd-online.com/Forums/main/12437-2d-simulation-flow.html)

 Jesper SÃ¸rensen October 20, 2006 02:07

2D-simulation of flow

Hey

Im a an konstruction ingineer student from Denmark, and I have a qustion regarding:

"How to make a 2D-simulation of windflow around i.e an airfoile profile??"

I have no problem setting up 3D simulations in a windtunnel, but do not know how to do the simulation i 2D - Please help!

\\Jesper

 ganesh October 20, 2006 11:41

Re: 2D-simulation of flow

Dear Jesper,

If you want experimental data for a 2D numerical simulation via experiments for the flow past an airfoil, you need to either

a) Use a wing with larger span (ie a long wing), so that the AR is high enough to lead to a lower induced drag and therfore lesser three dimensional effects

b) Use end plates for a finite wing model, to eliminate any possibility of circulation at wing tips and essentially diminishing three dimesionality effects.

On a wind tunnel, clearly option (b) is most feasible.

Hope this helps

Regards,

Ganesh

 Jesper SÃ¸rensen October 20, 2006 11:54

Re: 2D-simulation of flow

Thanks for that input Ganesh -

but was not in perticular the awnser I were looking for. The model has to have a specific size! The problem is how to "draw", so to speak, the 2D model in the CFX program-place it in a windtunnel, and then generate a 2d mesh for the problem in hand!

\\Jesper

 ganesh October 21, 2006 00:03

Re: 2D-simulation of flow

Dear Jesper,

If you are looking at a numerical simulation, there are two possibilities, once again.

1. Get the coordinates of the airfoil, and enerate the airfoil using a suitable spline fit. Then generate a frafield, say 10-20 chords away and mesh the closed domain thereby produced. This mesh can then be used to simulate the flow past the airfoil.

2. Generate the three dimensional wing and a suitable farfield, these now being 3d rather than 2d. To simulate the 2D problem, we need to eliminate the three dimesionality effects, which can be achieved by using periodic BCs at the ends (it is as if you kept endplates when doing experiments).

I believe you are looking out for Option 2

Hope this helps

Regards,

Ganesh