Circuit + CoilSolver - Transformer Test

Numerical methods and mathematical models of Elmer
Post Reply
gforti
Posts: 71
Joined: 11 Jan 2022, 20:39
Antispam: Yes

Circuit + CoilSolver - Transformer Test

Post by gforti »

Hello,

I've been studying Elmer FEM through the webinars and tutorials. To try to understand it better, I am trying a simple simulation of a transformer.
The transformer has a very simple geometry and I am using GMSH to create the geometry + mesh.
The driving fields are established through "CircuitsAndDynamics" and 3D coil is calculated using "CoilSolver".
I am comparing the results with some simple analytical calculations.

My issue here is the difference between the results from 2D and Analytical simulations (which agree) with the 3D one (specially at higher frequencies).

All the files are attached.
GMSH files: trafo_geo_v4.py + trafo3d_geo.py. The command for ElmerGrid is in cmd_elmerGrid.txt
Analytical calculation: Trafo_Ind.py (calculated results for 6 Hz, 60 Hz and 600 Hz are commented here)
Elmer FEM files are the remainder ones.

I am running Elmer in an Ubuntu 20.04 based distro (Elementary OS 6.1)
Version: 9.0 (Rev: 1c73bb548, Compiled: 2022-11-23)
Attachments
Elmer_Trafo_Test.zip
(9.19 KiB) Downloaded 62 times
mesh_3d_2.zip
(812.08 KiB) Downloaded 46 times
mesh_2d.zip
(98.67 KiB) Downloaded 42 times
vencels
Posts: 63
Joined: 20 Sep 2016, 17:05
Antispam: Yes
Location: Latvia
Contact:

Re: Circuit + CoilSolver - Transformer Test

Post by vencels »

Hi,

This is not a commercial post of mine, but I would like to suggest you to try TRAFOLO https://trafolo.eu/download-page/
It generates Elmer's sif files and runs it so you can get some ideas out of it.

Free for non-commercial use. We would be very happy if you could do comparisons with other solutions.

If any issues, you can find my contact details on the web.

Juris
gforti
Posts: 71
Joined: 11 Jan 2022, 20:39
Antispam: Yes

Re: Circuit + CoilSolver - Transformer Test

Post by gforti »

Hello all,

I was doing some studies, and comparing my simulation against Ansys Maxwell (a college provided the results).
It seems that the simulation was ok. Difference was with inappropriate mesh density for 3D case (used Salome for this). Anyways,
I am going to attach all simulation files here again. Also, the results are discussed in the pdf file.

Share.7z has:
analyticalCounter - python scripts showing the results for self and mutual induction calculations
Elmer_2d - files for 2D simulation
Elmer_3d - files for 3D simulation (except mesh)
femmCounter - femm file for 2D comparison

mesh3d.7z and mesh3d_elements files must be copied to a folder named 'mesh_3d' for Elmer 3D simulations (place this folder in the Elmer_3d folder)
Attachments
mesh3d_elements.7z
(768.94 KiB) Downloaded 43 times
mesh3d.7z
(380.46 KiB) Downloaded 44 times
Share.7z
(626.2 KiB) Downloaded 45 times
gforti
Posts: 71
Joined: 11 Jan 2022, 20:39
Antispam: Yes

Re: Circuit + CoilSolver - Transformer Test

Post by gforti »

report.
Attachments
results_discussion.pdf
(860.75 KiB) Downloaded 94 times
gforti
Posts: 71
Joined: 11 Jan 2022, 20:39
Antispam: Yes

Re: Circuit + CoilSolver - Transformer Test

Post by gforti »

Continuing, I faced some problems using CoilSolver and Circuits & Dynamics with Open Coil. Deriving from the case above, I tried first a simple circuit. I got that if I did not switched the "Electrode Boundaries" order, simulation failed. Error message: WARNING:: CoilSolver: No negative current sources on coil end! ERROR:: CoilSolver: Crappy potentials, cannot continue!
Had to:

Code: Select all

Electrode Boundaries(2) = Integer 3 2

for it to work.
But, with two circuits, it did not work at all. Can anybody help?

Files are attached.
case_3dOpen_ld_h_t2.sif and transient_3dOpen_1_h.definition for one circuit
case_3dOpen_ld_h_t3.sif and transient_3dOpen_3_h.definition for two circuits
mesh_3d_open - mesh
Attachments
Elmer_t3.zip
(740.84 KiB) Downloaded 44 times
Post Reply