Hello All
I would like to share a bit of work with you.
First, the visualization of magnetic flux, with the VTK postprocessor (100 Isosurfaces with low opacity and vectors):
large image can be found here: http://www.stamm.sh/forum/elmer/eMag/20 ... gflux2.jpg
Zoomed out, so the whole Geometry minus a hidden face can be seen:
This was calculated with the following software and files:
- Latest Elmer Linux release from Debian testing repository
- Latest NetGen from Debian testing repositiory (have to run this as superuser)
- geometric input file in CSG format for netgen (See NetGen chapter 2 for CSG help)
-> http://www.stamm.sh/forum/elmer/eMag/em ... gincyl.geo
- meshed output from netgen, exported in Abaqus format (have to import this in elmer with "File>Open..."
-> http://www.stamm.sh/forum/elmer/eMag/ai ... gincyl.inp
- magnetostatics.xml file (does 3D only, no harmonic solutions), to be placed in the edf folder in Elmer directoy:
-> http://www.stamm.sh/forum/elmer/eMag/magnetostatics.xml
- solver keywords file, updated with keywords for magnetostatics solver in 1D and 3D:
-> http://www.stamm.sh/forum/elmer/eMag/SOLVER.KEYWORDS
- SIF file (you can get the body force and boundary condition settings from here):
-> http://www.stamm.sh/forum/elmer/eMag/case.sif
I solved this on a Dell 690 with dual quad-core 2.66GHz Xeons in about 2.5s.
Using MPI and all cores, of course.
Note that the solver was using an ILU1 preconditioner.
Best Regards
Mike
3D Magnetostatics: Aircore Coil Magnetic Flux Simulation
Re: 3D Magnetostatics: Aircore Coil Magnetic Flux Simulation
Hello,
I am interested in using Elmer for some magnetic work. I was wondering, how did you get the absolute value of the magnetic flux in the post-processor? I tried running your example in the latest windows build (Svn Rev: 4411) and I only get the components to show up/
I am interested in using Elmer for some magnetic work. I was wondering, how did you get the absolute value of the magnetic flux in the post-processor? I tried running your example in the latest windows build (Svn Rev: 4411) and I only get the components to show up/
Re: 3D Magnetostatics: Aircore Coil Magnetic Flux Simulation
Hi Hvrontes
To get the Absolute Magnetic flux, I used the VTK postprocessor
( Run > Postprocessor (VTK) )
Then (Edit > matc...)
Then enter "MagFluxAbs= abs(vdot(Magnetic.flux.density, Magnetic.flux.density))" , without the quotes, in the top field and press enter.
I picked the name "MagfluxAbs" myself, you can use any other variable name you wish, i think. Ex "AbsB", etc...
You should now have "MagFluxAbs" in the dropdown list...
By the way, please post on the forum when you start to run into magnetic solving problems.. Im working on this myself, using the new Magnetodynamics solvers, which are available in the latest Windows version (4440) and linux sources. Might have some tips.
Id also appreciate if you post a complete example or two....
Cheers
Mike
To get the Absolute Magnetic flux, I used the VTK postprocessor
( Run > Postprocessor (VTK) )
Then (Edit > matc...)
Then enter "MagFluxAbs= abs(vdot(Magnetic.flux.density, Magnetic.flux.density))" , without the quotes, in the top field and press enter.
I picked the name "MagfluxAbs" myself, you can use any other variable name you wish, i think. Ex "AbsB", etc...
You should now have "MagFluxAbs" in the dropdown list...
By the way, please post on the forum when you start to run into magnetic solving problems.. Im working on this myself, using the new Magnetodynamics solvers, which are available in the latest Windows version (4440) and linux sources. Might have some tips.
Id also appreciate if you post a complete example or two....
Cheers
Mike
Re: 3D Magnetostatics: Aircore Coil Magnetic Flux Simulation
Thanks for sharing, I would like to know how or where do you define which is the body one and which one is body two. I'm trying to figure out.
joc
joc
-
- Posts: 41
- Joined: 25 Feb 2011, 18:43
- Antispam: Yes
- Location: Italy
Re: 3D Magnetostatics: Aircore Coil Magnetic Flux Simulation
I tried to solve with the standard formula for infinitelly long coil:
B=m0 * # turns * current / lenght
but the vaule is far from 41.9 mT.
I uset a current density of 2.031 A/m^3 that is equivalent of 1000 turns at 10 A.
I know that the magnetostatic solver does not fullfil the Coulomb Gauge.
I think that the suited solutor is magneto dynamic with Whitney element.
B=m0 * # turns * current / lenght
but the vaule is far from 41.9 mT.
I uset a current density of 2.031 A/m^3 that is equivalent of 1000 turns at 10 A.
I know that the magnetostatic solver does not fullfil the Coulomb Gauge.
I think that the suited solutor is magneto dynamic with Whitney element.
Re: 3D Magnetostatics: Aircore Coil Magnetic Flux Simulation
Need help in setting up my first magnetics 3D problem in ELMER. Have mesh from SALOME. But there is no tutorial for 3D magnetics in the manual. Only solver keywords, which do not help a bit.
Can't access any of the links in this topic (dead links? or denied access?).
Can't access any of the links in this topic (dead links? or denied access?).