Current Solver Test

Elmer cases by the users for the users
Post Reply
madstamm
Posts: 37
Joined: 07 Apr 2010, 09:56

Current Solver Test

Post by madstamm »

Hi All,

This is a complete file package with which you can test the CurrentSolve solver.
Current through a AWG20 wire is solved for, when applying 0.1V. The Whitney EM solver is also used to calc the resulting magnetic field.

To make sure my models were being correctly calculated, I modelled the following:
- 5cm AWG20 copper wire with expected resistance of 1.72mOhm
- 0.1V across the wire, should result in 5.8W power loss
- Aircube with 10cm sides at zero potential
- the two objects are meshed in conformal fashion using salome, allowing correct electromagnetic modeling. Note that this is key to making this work.

Result: It works nicely! Part of the solver log:
-------------------------------------------------------------------------------------------
StatCurrentSolve: Total Heating Power : 5.8175872743848025
StatCurrentSolve: Effective Resistance : 1.71892565222538573E-003
-------------------------------------------------------------------------------------------

The Salome python script that generates the model and mesh can be found here:
currentSolveCheck_AWG20_b.zip
(1.21 KiB) Downloaded 836 times
The sif file is here:
case_sif.zip
(1.25 KiB) Downloaded 930 times
And finally, the project folder is below. You can extract the folder and open it as a project in elmer - but note that you have to generate the sif file, then edit it, load case2.sif, save it as case.sif and only then start the solver. I dont have the project set up for ElmerGUI, and the GUI overwrites the case by default.
currentSolveAWG20.zip
(1.58 MiB) Downloaded 937 times
VTK showing the electrical potential on the wire
Image

VTK showing the potential and the magnetic flux
Image
orcaja
Posts: 4
Joined: 01 Dec 2011, 23:57
Antispam: Yes

Re: Current Solver Test

Post by orcaja »

Thank you for sharing, I have a question about this case. In the .sif file there are two lines in the boundary conditions, they are:

P {e}= real 0
P = Real 0

and I'm tryin to understand what they mean, I also saw in the solve log, the following:

Model Input: Unlisted keyword: [p {e}] in section: [boundary condition 1]
Model Input: Unlisted keyword: [p] in section: [boundary condition 1]

can anyone explaint to me what do they mean?
thank you in advance.
orcaja
raback
Site Admin
Posts: 4812
Joined: 22 Aug 2009, 11:57
Antispam: Yes
Location: Espoo, Finland
Contact:

Re: Current Solver Test

Post by raback »

Hi

This is a special solver which includes nodal degrees of freedom used for the scalar potential and edge degrees of freedom used for the vector potential. The latter choice ensures that the vector potential by construction is divergence free i.e. satiesfies the Coulomb gauge.

In Elmer the Dirichlet conditions are high-level matrix operations. If the variable name is "varname" then the normal way to set Dirchlet conditions would be "varname = 0". Now this only sets the nodal dofs. In this case we also want to set the edge dofs and for the reason a convention "varname {e} = 0" was introduced.

So in this particular case "p=0" set the scalar potential to zero, and "p {e}=0" sets the normal component of the vector potential to zero.

-Peter
Post Reply