True multiphysics

Numerical methods and mathematical models of Elmer
Post Reply
matt
Posts: 2
Joined: 17 Jun 2021, 16:00
Antispam: Yes

True multiphysics

Post by matt »

Hi,

I am new to Elmer, and been going through some of the tutorials which are all very helpful. However I want to model heat transfer at cryogenic temperatures, and I'm not sure if I don't understand something fundamentally, or that I'm running into some limitation.

To boil down & simplify the problem: I want get the temperature of an object that has current running through it, while it's being cooled at some interfaces of that object. E.g.a rod that's cooled on both ends, while current runs through. I would expect something like combining heat equation and the static current solver to the same body, and somewhere to couple the equations. Ideally, I would also like to include radiation heating/cooling into account which adds another equation.

Can this be done using Elmer, or are the solvers always independent?

Thanks!

Kind regards,
Matt
raback
Site Admin
Posts: 4173
Joined: 22 Aug 2009, 11:57
Antispam: Yes
Location: Espoo, Finland
Contact:

Re: True multiphysics

Post by raback »

Hi Matt,

This is spot on the type of application Elmer likes.

The tutorials are just tutorials for beginners. If you have 10 equations then you can easily create 10 basic single-equation tutorials. But for two or more equations the type of possible problems easily explodes and it is not easy to enumerate all the types of problems possible.

Here coupling would be through Joule heating and temperature dependent parameters. The coupling can be weak. If there are some convergence problems then some relaxation helps usually.

There is also a strongly coupled thermoelectric model but that is mainly for cases cases where these two fields are inherently coupled. Not for this type of case.

If you have problems just create a minimal case with HeatSolver & StatCurrentSolver and share the case for comments. The coupling is just minor icing on the cake.

-Peter
kevinarden
Posts: 1210
Joined: 25 Jan 2019, 01:28
Antispam: Yes

Re: True multiphysics

Post by kevinarden »

This is a case that was combining the heat equation with the static current, including joule heating. Good discussion, but not sure this case was completely solved, but provides the overall approach.

viewtopic.php?f=3&t=7062&hilit=8mm
matt
Posts: 2
Joined: 17 Jun 2021, 16:00
Antispam: Yes

Re: True multiphysics

Post by matt »

Thanks for the responses all :). I've gone through the initial learning curve, I just misunderstood that you can just 'check/uncheck' various solvers for each body. Anyway I've gotten much further:
2021-06-23 14_35_40.png
(41.05 KiB) Not downloaded yet
I've also attached my SIF file and I think this makes sense. However I'm not sure if I addressed the boundary conditions correctly. What I want is to define a fixed current, and compute the voltage drop. Currently I'ved added a 0V BC on one face, and a currrent density at another. Is this the way to go?

Secondly, I would like to get things like temperature profiles and wattages (energy dissipation) as well, but these are not in the default Elmer VTK viewer. Is there a way to view these? I'm sorry if these are very obvious questions.

Thanks!
matt
Attachments
case.sif
(8.7 KiB) Downloaded 16 times
Post Reply