Re: 3D-1D coupling

Mon, 15 Mar 2010 03:40:25 +0000

I forgot to mention :

I want to solve this sequence, both domains interact :
- solve 3D domain
- solve 1D domain with boundary values of solved 3D domain
- solve 3d domain with new boundary values from 1D domain
- and so on

3D-1D coupling

Mon, 15 Mar 2010 03:36:03 +0000

Hi everyone,

I would like to couple a bunch 1D domains at each and every point of a 3D domain boundary (2D surface). See the attached .jpg to understand.

Here is what I tried until now for a simple conduction problem :

1. I work in 3D.
2. I draw 2 adjacent 3D cubes.
3. I set the boundary conditions, etc.
4. For my 1D domain (1D in theory, drawn in 3D), I do this:
In Physics\Equation System\Subdomain Settings\keep only the terms for 1 coordinate (the one normal to the boundary) in the Weak equations and the Variables equations.

It solves quickly, no error messages, but I have to validate my solution. It looks a little weird to me since 1D nodes should be aligned, and here I still have tetrahedrons with unaligned nodes...

I can't find nothing related to my problem in the integration-extrusion-projection coupling variables theory.

François

COMSOL Forums: 3D-1D coupling

Re: 3D-1D coupling

3D-1D coupling