# 2D mesh generation for CFD analysis of a wing profile

Hello,

I´m interested in creating a mesh similar to the one from this tutorial, but I´m having trouble since the tutorial starts with the simulation setup.
https://www.simscale.com/projects/eduardocclc/tutorial_compressible_flow_over_airfoil/

First, I would like to know how can I create that “C” type inlet, or if it´s ok to use a square bounding box as usual.

Second I would like to confirm if I should use 1m for the size of the bounding box in the z coordenate and then just apply the 2D Empty boundary condition to the sides of the domain.

Hi @eduardocclc,

you might have a look at the CFD master class where the airfoil is covered: CFD Master Class - NACA0012 Airfoil

There you can see how the mesh can be generated. Boundary conditions will also be covered there.

Some small hints concerning the setup:

• In general the C-grid is the best one for this case because the mesh quality will be the best at the trailing edge

• Since the C-grid is symmetric, it is optimal for various angles of attack.

• When using non-reflective boundary conditions L_Z & L_Y can be 20-25 times the length of the profile. L_B has to be a little bit higher 25-30 times the length of the profile.

• You need approx. 30-50 points between the profile and L_Z

• On the profile itself you need approx. 100 points to see pressure and velocity gradients properly

Feel free to add more important points to this project that might be interesting. (@pfernandez, @dylan, @1318980)

Best,

Jousef

1 Like

I don’t think there’s a way to create 2D mesh with the hex-dominant parametric algorithm on SimScale yet, because the underlying SnappyHexMesh tool generates 3D cells only. There is a way to get a 2D slice from a 3D mesh, but you will need to work with native OpenFOAM. https://openfoamwiki.net/index.php/SnappyHexMesh

Because the mesh tool doesn’t create 2D cells, you can’t use empty as BC on the sides.

You don’t need a C-grid in this case, because of the unstructured nature of the cells. It won’t matter even if you want to vary AOA.

i have a one problem in relaxation factor it creates the instability in the solution. can you look what is the problem

https://www.simscale.com/workbench/?pid=1963253659363330458#tab_1-0