3D cube with 1D-Line in the middle

General Discussion
1

Hello!

I want to create a 3D cube, with labeled line* (or border) in the middle, where the tetraeders or tetrahedralization adapts or uses the vertexes of the 1D-line. Just a 3D cube with a labeled line in it as one mesh and labeled 1D entities.

Is this possible with FreeFEM and its plugins or do I need external Mesh-Generator-Software?

*or helix

2

I think it is not possible with FreeFem. You could use gmsh. But then the label attached to the 1d curve will be impossible to use within FreeFem.
Instead of the label attached to the 1d curve, within FreeFem you could eventually deal with the list of vertex indices on this curve, and build a 1d meshL, independently of the 3D mesh.

3

Thank you for your answer, but I don’t understand.

After loading (gmshload) the gmsh-Mesh, I use the line- oder curve vertices from it, to build a meshL?

Greetings

4

There is no built-in command to build this meshL from the 3d mesh. What you can do is copy by hand the 1d data from the file .mesh and put them to another file, so that you can read it with FreeFem to build the meshL, by adapting to x,y,z components the example of Input Border Data from an External File in FreeFEM++ - #3 by fb77
which has only x,y components.
At the end instead of mesh Th=buildmesh(a(nborder));
you can do meshL ThL=buildmeshL(a(nborder));
Anyway I don’t see what you want to do afterwards, since in FreeFem line integrals cannot be used in variational formulations for 3d finite elements.

5

I rather meant from kind of vertices-coordinates-array, or similar. But doesn’t matter.

I learned on COMSOL and want to switch to FreeFEM++ now.
There I could have a line or edge helix, which I could use as current source for a 3D magnetic vector-potential field. But as I learned today (if correct), the geometrical approach with an integrated mesh isn’t neccesary.
I just don’t want to use an volumetric current source (or helix or more complex one), because of performance reasons.

Greetings