TEAM 7 - Asymmetrical Conductor with a Hole

Main Forum Old-Forum - do not post here
1

Hi there,

Im currently working on solving the TEAM 7 problem in NGSolve (Compumag).

That problem describes an aluminum plate with eccentrically placed hole. This plate is set asymmetrically inside a magnetic field which is produced by a coil. That coil is not perfectly cylindrically shaped. And that’s where my problem is.

As of now I used:

CoefficientFunction((-y,x,0))

which results not the correct results for that geometry.

I also have access to a working solution in FEniCS. There the impressed current vector potential (T0) is created in a separate file saved to .xml and read again in the main code. I currently don’t know if there is an equivalent method in NGSolve.

I read about BSpline and VoxelCoefficientFunctions, but I don’t completely understand how to implement them as well as if they would be an appropriate approach to the solution.

TLDR: How can I correctly implement the impressed current vector potential for a not perfectly cylindrically shaped coil in NGSolve.

I attached the volume and my current code.

Any help is appreciated.

Best
Sebastian

Attachment: ngsolve_team7.py

Attachment: team7.vol

2

Hello,

I’m somehow in the same situation. I tried to do the following: I solve the equation for voltage potential in a conducting part which can then be derived to the current density. See the screenshot in the attachement.
The problem hereby is that I have to cut the coil to impress a potential. This is not ideal for the magnetic field distribution.

800×454 87.9 KB

Is there any other idea to impress a current in a plane to a solid coil with arbitrary shape?

Kind regards