Magnetic field produced by linear current in COMSOL

[Caleydos]

TL;DR Summary

Issue with COMSOL 5.5 when trying to calculate the magnetic field produced by linear current in 3D

Hello,

I have used an edge current of 10 A through a 0,45 cm (lenght) wire inside an air sphere. The thing is that, according with Ampere law, the magnetic field (B) produced at a 1 mm of distance from the wire shall be 0,002 T, and I am obtaining much higher values in this simulation (around 0,3 T at his point).

I have done it in 2D and result is OK, according with theoretical values, but I do not know why in 3D it changes. Any idea please?

Many thanks in advance

 

[DrClaude]

Can you post a picture of the geometry in 2D and in 3D?

 

[Caleydos]

Sure! I attach them here. Thanks a lot!

First one is the 3D model (lateral view), showing the value of the magnetic flux density in COMSOL 5.5 (this value do not matches with theoretical)

The second one is 2D in COMSOL 3.5, this value really matches with theoretical

 

[DrClaude]

There is something strange going on, for instance:

Check the boundary conditions.

 

[Caleydos]

Thanks a lot for your reply.

This is one of the things that I supossed in the beggining, but I made the boundary sphere wider and the result is the same (see attached). It seems the magnetic flux shown in the boundary layer is debt to the sphere is very small. Any other idea?

 

[Caleydos]

I have tried modifying the length of the wire, magnetic flux should remain the same, but it isn´t. It seems the longer wire produces more flux at 1mm (it makes no sense).

On the other hand, I do not understan why this shape in the magnetic flux, it should be purely circular.