Hi,

I've coded a NS solver based on cartesian grids and I'm now extending it to structured grids. My main objective is to simulate airfoils, possibly extending to 3D. Hence, is there a clear choice to use c or o-grid for airfoils?

Please explain why.

Also I've download a grid generator for airfoil known as UGRID which can generates 2d structured o-grid for airfoils.

Is there similar type of program which can generate c-grid for airfoil as well?

Thanks alot!

Dear Quarkz,

As far as airfoil simulations as concerned, I find no distictive reason why an O-grid should be preferred over a C-grid or viceversa for all kinds of problems. However, a C-grid CAN be preferred as the grid will be aligned to the wake or slipstream at the trailing edge of the airfoil. Particularly, in context of adaptation, wherin you have a viscous flow past airfoil, a boundary layer and the wake are common features. The BL and the wake are to be adapted in the streamwise direction ( to resolve it better). The use of a C-grid will help you achieve this easily, whereas it is not the case for an O-grid. However on a relatively fine grid, results on both O- and C- grids are found satisfactory. If you are intersted in handling more complicated geometries you can try UNSTRUCTURED grids as well.

I am not sure of any program for C-grid generation available, but if you are able to use an O-grid generator with a cartesian meshing and match them, a C-grid can be generated, though some effort is needed in that direction.

Hope this helps

Regards,

Ganesh

For viscous flow a C-grid is preferred because it allows you to resolve the wake. To really resolve the wake, you would actually need a different C-grid for each angle of attack. Often enough this is not done, which somewhat weakens this argument for using a C-grid. More important than resolving the complete far wake is the nice resolution of the region right behind the trailing edge.

O-grids are good enough for inviscid flow, where they have the advantage of reduced cell count as compared to C-grids. The resolution in normal direction is high only where you really need it: right on the airfoil surface. The region behind the trailing edge is not well resolved, which makes the use of O-grids questionable in case of viscous flow. You may still get good results even for viscous flow, depending on the particular case.

Hi,

thank everyone! I'm writing an NS solver hence it seems that I will need to use C-grid.

However, I only 've got a free 2D O-grid generator, Ugrid which is simple to use and user-friendly. Does anyone know of a c-grid generator (free/commercial) for airfoil. I prefer one where i just enter the x/y coord and a reasonable mesh can be generated.

Thanks again

Hi,

just to add that my main interest is in finding lift and drag of airfoil. in that case, is c-grid still more suitable than o-grid?

thanks again

There is C-grid Generator for free.

The name of the program is "MeshPilot"

And you can download this program through the following website.

http://www.shore-cfd.com/html/shore_...eshpilot.shtml

and maybe this program support naca series airfoil.

And I have questions. Now I trying to make 2-D airfoil grid for N-S(Spalart-Allamas).

Airfoil is RAE 2822.

But it is difficult to make this grid.

Do you know how to determine

1. Boundary layer thickness 2. Grid spacing 3. Etc. ^^

I'm poor at CFD.

Is there reference or technical guide book for grid making?

Thanks, in advance~

You can see again the "Fletcher book" T2, there is a source programm (in Fortran) for mech generation (C-grid) around naca airfoil utilising the algebraica methods (multi-surfaces elements)