Coupling Magnetodynamics to Fluid dynamics

Numerical methods and mathematical models of Elmer
metermera
Posts: 11
Joined: 14 Feb 2023, 17:08
Antispam: Yes

Re: Coupling Magnetodynamics to Fluid dynamics

Post by metermera »

including the MMG lines belowin my .sif file (taken from the test case Peter suggested) seems to work without any issues (or changes to my Elmer install on Windows). Seems a bit too good to be true. I'm still re-running and post-processing so it's early days but the results seem encouraging.


MMG Remesh = Logical True
MMG hmin = Real 0.05
MMG hmax = Real 0.15
MMG hsiz = Real 0.1

mmg No surf = Logical True
mmg no move = Logical True[/b]
raback
Site Admin
Posts: 4812
Joined: 22 Aug 2009, 11:57
Antispam: Yes
Location: Espoo, Finland
Contact:

Re: Coupling Magnetodynamics to Fluid dynamics

Post by raback »

Great to hear! This is really fresh stuff! MMG has been used by the Elmer/Ice community for quite some time for their specific challenges. So this is just minor extension of that work. Need to add some additional features still... -Peter
raback
Site Admin
Posts: 4812
Joined: 22 Aug 2009, 11:57
Antispam: Yes
Location: Espoo, Finland
Contact:

Re: Coupling Magnetodynamics to Fluid dynamics

Post by raback »

Hi

There are now two more variations of the test case. One where the mesh density is solved from an diffusion-reaction equation and another one in 3D.

-Peter
metermera
Posts: 11
Joined: 14 Feb 2023, 17:08
Antispam: Yes

Re: Coupling Magnetodynamics to Fluid dynamics

Post by metermera »

Thanks Peter. Elmer is magic! I will check these out.

I have a slight problem where the heat solver is giving me a static temperature across the conductor, even though joule heating is on (and a significant amount of heating is being solved).

This is a transient mode with a conductor moving through a magnetic field (generated by a coil) and the resulting JxB forces are solved and joule heating is solved but the temperature scalar field across the conductor isn't right e.g. it is constant and then cooling.

I just can't see what is wrong. Sorry to bother...again!
Attachments
hgff_0325.sif
(9.65 KiB) Downloaded 42 times
raback
Site Admin
Posts: 4812
Joined: 22 Aug 2009, 11:57
Antispam: Yes
Location: Espoo, Finland
Contact:

Re: Coupling Magnetodynamics to Fluid dynamics

Post by raback »

Hi,

I was aware of some functionality missing needed for transient simulation. Just added it to "devel" branch. The the field values and most importantly the previous values need to be interpolated to the new mesh. The current values mainly just help in convergence of linear system but if the problem is transient you really need the previous values too.

Didn't really test it too much. But it should be a simple addition since the routine doing the interpolation established since long.

Now there is a caveat. The interpolation only works for nodal entities. In 3D the magnetic fields are solved using Hcurl conforming elements. For this the interpolation does not work. It would be moderate effort to make it work for regions where the elements stay the same. Luckily for magnetostatics you do not need the time derivatives.

-Peter
metermera
Posts: 11
Joined: 14 Feb 2023, 17:08
Antispam: Yes

Re: Coupling Magnetodynamics to Fluid dynamics

Post by metermera »

Thanks Peter. I look forward to when it is available.
Post Reply