[jntan]

Anyone has this problem? This always happens when I'm using auto tet meshing and use existing surface mesh as input to generate volume mesh. I'm sure this is not the computer RAM problem because the mesh quantity is only 900,000 and I have 4GB RAM and GForce 9600 512MB graphic card.

[PSYMN]

If you have a closed surface mesh, then change the Tetra method to Delaunay or Advancing front or something like that to generate a volume mesh from existing mesh.

The Octree with the option to "Use Existing Mesh" is really mean for just a few locations of existing mesh.

Behind the scenes, it generates a full mesh and then tries to make that mesh consistent with the existing mesh... It was not meant to be used this way and will be much much slower than the Delaunay or Advancing front methods.

[matteoL]

Hello,
I am having the same issue of the post: disconnecting orphan cells takes ages.
Unfortunately my geometry is a a box (already surface meshed) and an internal wall ,( i.e. a baffle in the middle) and it seems that that onlt the octree methid manage to mesh it... all other meshers (bottom up) fail because they say that there are single edges (of course there are, it is a baffle and i cannot split it (with split internal wall) before having created the volume mesh...)..

What should i do?

thanks,
matteo

[Chris_321]

I have the same problem.

Can someone explain how i can compute a delauny mesh without having a octree mesh first? If i try to do this it just creates "subsets" and delete my surface mesh.

[diamondx]

You don't need an octree mesh to generate a delaunay mesh , but you definitely need a surface mesh at least. Delaunay starts and grows from surface mesh...