CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   2-D GAMBIT mesh (http://www.cfd-online.com/Forums/fluent/33979-2-d-gambit-mesh.html)

yaron June 16, 2004 01:22

2-D GAMBIT mesh
 
shalom,

I'm trying to mesh a very simple model in gambit which I know his analitic solution : fully develop laminar flow between infinite parallel plates. when I solved it (2-D fluent) I got a different solution for every meshing. when I use very fine mesh I move away from the analitic solution (instead of congregate to it) what can be the problem ?

Thanks a lot for your help

yaron


ap June 17, 2004 03:45

Re: 2-D GAMBIT mesh
 
What kind of grid are you using? Usuall a structured uniform grid is better for such a simple case.

Which discretization scheme for convective terms in the transport equations did you choose? Try to increase its order to obtain better results.

Did you try to use FLUENT adaption function?

Hi :)

ap

yaron June 18, 2004 02:06

Re: 2-D GAMBIT mesh
 
ap shalom, I used uniform grid with different intervals without adapt. when I meshed with very fine intervals, the solution moved away from the analitic one - I tring to find an explanation for this result.

when I tried to use coarse mesh & fluent adapt in the test area - the solution was satisfactory.

I didn't understand your explanation on "convective terms in the transport equations" - what do you mean by that and how can I do it?

THANKS a lot for your help

yaron

ap June 18, 2004 03:46

Re: 2-D GAMBIT mesh
 
FLUENT offers different kinds of discretization schemes for the convective terms of the transport equations you're solving in your model,

You can find them in the Solve->Control->Solution menu of the FLUENT graphical interface.

A better accuracy can be obtained using a high order discretization scheme such Second Order Upwind or QUICK.

Also the pressure-velocity coupling algorithm is important. You can change it in the same window where you choose the discretization scheme.

Hi :)

ap


All times are GMT -4. The time now is 16:50.