Magnetostatic pair of coils

[federico]

Hi all,

I am working on a two coils case surrounded by a sphere of air. I would like to know if the lines in the coil without electrical current is the magnetic flux density LINKED from the alimented coil.

If yes, i would like to see it value. How can i do??

Regards
federico

[raback]

Hi,

You could try the sif from here to this case to get the inductance matrix:
viewtopic.php?p=29786#p29786

My guess is that you cannot induce a current to a closed coil since it cannot have a potential gradient. You can generate a current to a closed coil with special tricks included in CoilSolver. Then this current may be used with circuit simulator as a basis with unknown multiplier OR as "test current" to detect the coupling to other coils as is done in the case above.

-Peter

[federico]

Hi,
Thanks for your reply.

I managed to set a coilcurrent = 1 and i obtained this results.
My idea is make a division between the magnetic flux linked and the current i set, so i have calculated the mutual inductance.

The question is: how can i understand what is the correct value of the magnetic flux linked? (Here seems to be something like 5e-7).

regards
federico

[raback]

Hi

Maybe try to find a reference case where the solution is known. There seems to be semianalytic solutions like in:
https://www.researchgate.net/publicatio ... ils_in_Air
or probably other FEM codes would have could have some examples.

If you have a ready-made mesh we could easily try the method in Elmer to compute the mutual inductance.

-Peter

[federico]

Hi Peter,

Thanks for your reply a lot.
I am working in Elmer to try to get some results, it’s like a test to try the accuracy of Elmer.
Here is the L (self inductance of a coil) obtained analytically. ( you can see everything in the matlab script (as case reference) ).
Now what I want is a very close result in Elmer compared to the script in matlab.
I think the Magnetic Field Energy approach is the best (W=1/2*L*I^2) to calculate L.
Here I need the Magnetic field energy of the whole domain and a current setted.

My question are:

1) Is the line in the ‘solver log’ (using coil solver, Mgdyn and mgdynpost) -Electromagnetic field energy- the energy I need? (Here it is too high).

2) Is it correct to set the value equal to 1 Amp of the current by checking in coilsolver -coil closed- , -desired current density- =1, and -normalize coil
current- and then in Body force using "Equals coilcurrent 1" ..... ?
"Equals coilcurrent 2"
"Equals coilcurrent 3"

After all, if this approach gives the right results, i’ll be happy to add a new coil and calculate the Mutual induction.

Here is attached the matlab script of the analytical reference case, the .geo file and sif file of the case i am working in Elmer.

So working step by step.. i am a beginner in Elmer

[raback]

Hi, It seems a .step file is missing that is referred to in the .geo file. -Peter

[federico]

Oh i am sorry Peter, you're right.

Here it is.

[raback]

Hi

With two times "refine by splitting" in Gmsh + this sif file I get E=2.914821108708E-007 which should be close L/2.

-Peter

[federico]

Thank you so much Peter for the help you are giving me, this is exactly what i needed.

But, when i run the sif file you attached i have this problem.

[raback]

Your Elmer is older than mine...