GenerateFromCoarseHexMesh

[rwendner]

The command GenerateFromCoarseHexMesh does not produce a valid LDPM mesh.

In the simplest case of generating a single Hex element the resulting Tet mesh is severely distorted.

The attached gen file contains two versions (one manually defining the single hex element and the other one using the Generate command).

[zhouxinwei]

This method generation scheme was implemented many years ago, early on in the LDPM mesh generation schemes. As explained in the manual, this method is not reliable and "tetgen meshing may not be sucessful". We stopped using it in favor of GenerateFromFineTetMesh. The latter method makes it possible to control the density of the external triangular mesh that will be used in the LDPM mesh.

[rwendner]

I am aware of GenerateFromFineTetMesh. Thanks for pointing this out.

However, there are also disadvantages to using a FineTetMesh. That is why we also looked Into the method based on a CoarseHexMesh where we do not have to prescribe surface nodes.

Interestingly, I get the posted mesh using the latest library and MarsO100 on my Macbook together with tetgen 1.4.3.
A student gets a perfectly fine mesh running tetgen 1.4.0 and the compiled version from May on the cluster.

[vorelj]

Hi, I accounted for the same problem. I know that you suggest to use fine mesh, but either cancel this method or provide some check, which controls the mesh shape. I think that some small angle in the triangular mesh is good parameter which could be checked.

[vorelj]

Hi, the problem with the mesh generated using "GenerateFromCoarseMesh" is somehow tetgen version dependent. The weird behavior is accounted for tetgen1.4.3 but it doesn't happen with tetgen1.4.0.

[zhouxinwei]

This method does rely on tetgen.

This method is not reliable and marked as discontinued in version 2016.8.10. Please use other methods, such genterateFromFineTetMesh.

For historical reasons, the "Make ParticleList" in hex solid list will be maintained. Its warning message is removed in version 2016.8.10.