CFD Online Discussion Forums

CFD Online Discussion Forums (
-   enGrid (
-   -   Volume mesh algorithm (

Tensian September 17, 2013 13:10

Volume mesh algorithm
Hi everyone,

I need to know which algorithms does Engrid use for surface meshing and for volume meshing. Anybody knows?

Thanks in advance.

ogloth September 18, 2013 02:30

Hi James,

yes I do.

First a question: Why are you interested in the internals of enGrid?

Now an attempt to answer your question:

Surface meshing is done by an iterative Delaunay mesher, i.e. points are inserted or removed based on the local mesh resolution, then the mesh is smoothed and edges are swapped according to a local evaluation of the Delaunay criterion.

Prismatic boundary layers are extruded based on local propagation vectors from the surface. The computation of these propagation vectors is not trivial and is done up-front (based on different criteria like small gaps, sharp corners, etc.).

Tetra meshing is done by the NETGEN library and I think the best would be to look on their website for more details (

We are, however, thinking of using a similar approach to the surface mesher for tetra meshing. As it stands we have access to an old F77 code which takes an existing tetra mesh and optimises it by inserting and deleting nodes, smoothing it and applying local reconfigurations. It all depends a bit on the time -- or money;-) -- we find to dedicate to the inclusion of this tool into enGrid.

I hope this helps a little bit!


Tensian September 18, 2013 09:57

Thanks for your fast reply Oliver,

I am searching the most suitable meshing software for my PhD. I read in the terminal somethig about Delaunay but also I can see some items related to Octree. So I am wondering what was the algorithm used just when you click at "Create/Improve Volume mesh".

In NETGEN website I saw that Advancing Front and Delaunay are used, but I am not sure the one that is used in NETGEN configuration for Engrid.

I worked with ICEM and Delaunay algorithm is fast and used to give good results for my CFD model.

Thanks again and if itīs possible please let me know the algorithm choice you have done for Engrid volume meshing.

All times are GMT -4. The time now is 15:30.