CFD Online Discussion Forums

CFD Online Discussion Forums (
-   Open Source Meshers: Gmsh, Netgen, CGNS, ... (
-   -   Preferences on OpenSource Meshers (

kmooney November 26, 2012 13:08

Preferences on OpenSource Meshers
Howdy Foamers,

The Gambit license is expiring for my research group in a short while. Since we have been using it exclusively for several years now, we are out of the loop in terms of foam-compatible open source or free meshing software.

What is your favorite free meshing software? If you have an extra second a few thoughts on why would be a great help!

Thanks everyone!


Lieven November 26, 2012 19:35

Howdy Kyle,

The past few months I've tested quite some mesh tools and there is quite some free stuff to find. However, I was quite disappointed after testing several of them (engrid, gmsh, snappyhexmesh, ...). Non of the free tools available provided me the flexibility that I enjoyed when using Gambit.
To my experience, only ICEM-CFD and CUBIT manage to compete with Gambit (there actually even more powerful) but these tools are not free.

So it all depends on what type of mesh you want to be able to generate. If it is a tetrahedral mesh you want, you will probably find something that fits your needs (e.g. engrid). But if you want to generate high-quality hexahedral meshes, don't get your hopes up too high of finding something for free.



elvis November 27, 2012 10:27

Dear Lieven,

yesterday Oliver Gloth showed the coming Engrid 1.4 version, meshing structures ready to make Conjugate Heat Tranfer simulations easier is the main improvement.

Oliver mentiond that Engrid does not use TETs!!! They are bad for FVM simulation.
Engrid converts TET into Polyhedrals (but it does not use PolyDualMesh).
He said that even if checkmesh complains, that the meshes are most of the time good for fast simulation runs.

Another opensource hero could be HELYX OS at least for those working with SHM and want a nice GUI.

kmooney November 27, 2012 11:52

Thank you! I appreciate all of the great advice.

Lieven November 27, 2012 13:08

Dear Elvis,

I fully agree that tets are to be avoided when performing FVM simulations but I don't agree with the statement that Engrid is not capable of producing meshes constructed from TETs (cfr.Engrid does not use TETs!!!). At least the 1.2 version I have installed here on my pc is definitely capable of doing this.

So in my previous post I only want to suggest that if you want a TETs mesh, you can use Engrid fro this. I didn't want to imply that this is the only functionality of engrid. You are indeed correct that hex meshes can also be produced with this tool.

ogloth November 29, 2012 05:16


just to clarify what enGrid does ;-)

enGrid is basically an unstructured mesher which creates a mixed tet/prism grid with boundary layer resolution. Hexes can only be created by extruding boundary faces. This will change soon in order to support hex-core grids -- the functionality is almost there but it still needs to be connected to the rest. I might release this as 1.4.1 if I find the time in the next month or two.

If you use OpenFOAM, tet grids lead to very dissipative solutions. For this reason enGrid has the option to export as polyhedral cells. As opposed to polyDualMesh, enGrid only converts tetrahedral regions and not the prismatic boundary layer or extruded regions. The resulting grids are, according to my experience, a lot better than tets and in many cases even comparable to pure hex grids. For other codes (e.g. SU2 from Stanford University) tetra grids might be okay, because these codes internally convert into dual cells.

If you want to use enGrid in a commercial environment, you might want to consider our support offer -- see


All times are GMT -4. The time now is 09:20.