Combining Two Meshes
 

Hello,

I'm new to posting on these forums. I can't seem to find the answer to my question online anywhere, so I thought I would try here.

Question: What is the procedure for combining two different meshes together.

For example, if I have a square 3D domain that is full of tetrahedral meshes and inside of the domain there is a spherical surface ( representing a particle). The mesh is generated between the spherical surface and the bounding edges of the cubical domain.

Now if I want to simulate what flow would result if one particle was placed in front of another, I could stitch two domains together at a common edge and use that combined mesh for my simulation.

I can't seem to find any discussion about 1.) How to go about doing this and 2.) Any open source software that can do this.

Am I thinking about this wrong?

Thanks,
Chris

Generally, it's not so easy to merge two meshes - not many programs give this option, though in principle it's not so difficult. Basically, the nodes at the boundaries of the two meshes should match exactly, so that the corresponding nodes can be merged. I don't know what kind of software you use for simulation, but FreeFEM++ (http://www.freefem.org/ff++/) can combine meshes - I suppose it automatically aligns nearby nodes. The command is as simple as

mesh somemesh = mesh1 + mesh2;

You can export the combined mesh in the MEDIT .mesh format.

Hi Christopher,

I don't know about any open source software but I'm sure that ICEM is able to merge two different meshes. It's pretty much easy, you just need to import the first mesh domain and then import the second one. It will ask you if you want to replace or merge and you easily choose merge.

It's also possible to read multiple mesh files in Fluent. for more info follow this link https://www.sharcnet.ca/Software/Flu...ug/node171.htm

Combining meshes
 

As mentioned before there are several mesh generation softwares that can combine meshes even with boundaries that are not coinciding.
But of course the solver needs to be able to digest it.

In NUMECA sofware it is called a full non-matching connection. Non-matching connections still require coinciding boundary edges (even if the point distribution can be different) but in FULL non-matching the addition "FULL" refers to the fact that even the boundary edges are not coinciding.

Best regards,
Colinda
NUMECA Academic group.