Magnetism

General discussion about Elmer
Post Reply
goosst
Posts: 2
Joined: 19 Jun 2010, 12:52

Magnetism

Post by goosst »

Hi,

I'm used of using FEMM http://www.femm.info/ for doing magnetic 2D and axisymmetric simulations. But now I'm looking for a suitable program that has the ability of doing 3D simulations. So I am very very new to Elmer :-) .

I have the impression that magnetism is not something standard available in Elmer.
So I'm wondering if somebody already simulated some magnetic things in elmer.
For example: a reluctance motor, permanent magnets etc.

Thanks,
madstamm
Posts: 37
Joined: 07 Apr 2010, 09:56

Re: Magnetism

Post by madstamm »

Hello Goost

Magentism does indeed seem to be a little-used package.
The forums, in this regard at least, are quiet.

I'm working on modeling a reluctance motor in 3D, and am convinced that it will work. There are problems, however, it is certainly not like using a commercial package with regards to electromagnetism.

1
magnetodynamics solver is available only with the latest release of elmer
- it does seem to reliably converge at this time, but there are a few non-GUI steps involved right now
- it is not officially integrated into the package (GUI, keywords, etc). I made some efforts in this direction, so let me know if you are interested.....
- probably only useful right now if you are at least mildly linux-friendly, or willing to get so

2
in general: making FEM-useful 3D models without commercial CAD software is quite tricky.
getting into the freeware tools takes time.

Bottom line (my opinion, of course)
if you have lots of time to allocate (student), it's worth the effort - you will learn a lot, probably get your work done, and advance the tool (elmer) in the process. If you are worried about sales next quarter, you're better off dropping USD10k+ on a commercial package.

regards
Mike
goosst
Posts: 2
Joined: 19 Jun 2010, 12:52

Re: Magnetism

Post by goosst »

Hello Mike,

Of course I'm interested in what you are doing in elmer for magnetism.
But maybe I have to come back in a year or something to try something myself to see if it gets a little bit more useable by then :).

Thanks,
Stijn
ghjmaui
Posts: 1
Joined: 29 Nov 2012, 05:33
Antispam: Yes

Re: Magnetism

Post by ghjmaui »

Hello Mike, et.al-
I am also new to Elmer and am in a similar situation. FEMM is great (with EXCELLENT examples), but limited in that it is only 2D and sort of hard for some configurations. Are there any examples out there in the last couple years since this last post, that could help someone in getting up to speed in using Elmer for magnetic simulation? The manual is really quiet, and while ElmerFront comes up with magnetics enabled, I can't really figure out how to set anything up, and there is almost no reference in the ElmerGUI to using it at all. The writeups on the magnetism packages are impressive, but no examples. Can anyone help me out? A simple example or two (a permanent magnet with a coil in the same model) would help a lot of people out, I think. Maybe something that could be added to the Tutorials?...

A humble newbie-

-Glen
raback
Site Admin
Posts: 4812
Joined: 22 Aug 2009, 11:57
Antispam: Yes
Location: Espoo, Finland
Contact:

Re: Magnetism

Post by raback »

Hi

There are consistency tests in the distribution. Those one starting with mgdyn* are related to the modern AV solver. You can also find pretty impressive stuff on the forum with complete examples.

There will soon be an update on the documentation. Writing tutorials is is a form of altruism and unfortunately the 1st in the agenda. Still I hope there will soon be more tutorials on the subject.

-Peter
nuriam123
Posts: 1
Joined: 13 Jul 2021, 11:29
Antispam: Yes

Re: Magnetism

Post by nuriam123 »

Very interesting, but what about magnets like the ones from this website.
raback
Site Admin
Posts: 4812
Joined: 22 Aug 2009, 11:57
Antispam: Yes
Location: Espoo, Finland
Contact:

Re: Magnetism

Post by raback »

Hi,

I'm not expert in magnetism but what Elmer has is a) given magnetization b) possibility to define B/H curve and c) Zirka hysteresis model where the up and down slope for B/H curve are defined separately. The model c) is only available for 2D. Maybe you can from these deduce whether the magnets you refer to can be modeled.

-Peter
Post Reply