Dear all
I'm a student in Japan and simulate electric field using Gmsh and Elmer.
I have one trouble on Elmer.
When I use Elmer to calculate electric field after meshing geometry using Gmsh, Elmer sometimes fail to calculate (no error) the field and the given electric potential is 0 (an attached picture show a map of the electric field). I can not find clear reason why Elmer gives these results.
Are there any suggestion to solve this problem?
Best regard.
Electric field calculation using Gmsh and Elmer.
Electric field calculation using Gmsh and Elmer.
- Attachments
-
- スクリーンショット 2019-04-24 23.44.04.png (226.15 KiB) Viewed 4138 times
Re: Electric field calculation using Gmsh and Elmer.
Hi,
you should give more information if you want someone to help you. What is your geometry (geo file if you use gmsh would be fine)? You should also post the sif file as it contains the settings for ElmerSolver.
Matthias
you should give more information if you want someone to help you. What is your geometry (geo file if you use gmsh would be fine)? You should also post the sif file as it contains the settings for ElmerSolver.
Matthias
Re: Electric field calculation using Gmsh and Elmer.
Thank you for the comment.
I was creating GEM.
I was trying to improve "Solver 1" in .sif file
Best regards.
I was creating GEM.
I was trying to improve "Solver 1" in .sif file
Best regards.
- Attachments
-
- gemcell.sif
- (2.89 KiB) Downloaded 251 times
-
- gemcell.geo
- (15.24 KiB) Downloaded 242 times
Re: Electric field calculation using Gmsh and Elmer.
Hi,
I cannot see an obvious reason why your calculation should not produce a valid result.
The densities of your materials are approximative, but density does not affect the electric field.
Since at least one of your materials is a conductor, you could try to use the static current solver instead of the electrostatic solver. Then you would have to set the conductivities of the materials, and calculate the electric field yourself from the simulated potential (there are some posts in this forum in tha subject).
In general, to gain confidence in the results, you could start with a very simple geometry (e.g. a cube) and verify that the result is correct for that case.
HTH,
Matthias
I cannot see an obvious reason why your calculation should not produce a valid result.
The densities of your materials are approximative, but density does not affect the electric field.
Since at least one of your materials is a conductor, you could try to use the static current solver instead of the electrostatic solver. Then you would have to set the conductivities of the materials, and calculate the electric field yourself from the simulated potential (there are some posts in this forum in tha subject).
In general, to gain confidence in the results, you could start with a very simple geometry (e.g. a cube) and verify that the result is correct for that case.
HTH,
Matthias
Re: Electric field calculation using Gmsh and Elmer.
Dear Matthias
Thank you for the comment.
Actually, I found an output message, which is as follows.
>Elmergrid loading data:
>-----------------------
>Format chosen using the first line: $MeshFormat
>Loading mesh in Gmsh format 4.0 from file gemcell.msh
>Reading 50 entities in 0D
>Reading 96 entities in 1D
>Reading 50 entities in 2D
>Reading 4 entities in 3D
>The program encountered a major error...
>No elements to load in Gmsh file!
According to this massage. it seems Gmsh has some troubles. When I tuned distances of the geometry a little bit, the field was calculated w/o any failure. I guess that an available size is restricted, probably...
Best regards.
Thank you for the comment.
Actually, I found an output message, which is as follows.
>Elmergrid loading data:
>-----------------------
>Format chosen using the first line: $MeshFormat
>Loading mesh in Gmsh format 4.0 from file gemcell.msh
>Reading 50 entities in 0D
>Reading 96 entities in 1D
>Reading 50 entities in 2D
>Reading 4 entities in 3D
>The program encountered a major error...
>No elements to load in Gmsh file!
According to this massage. it seems Gmsh has some troubles. When I tuned distances of the geometry a little bit, the field was calculated w/o any failure. I guess that an available size is restricted, probably...
Best regards.
Re: Electric field calculation using Gmsh and Elmer.
Hi,
try to export the mesh in gmsh format 2 and import that to Elmer. gmsh format 4 import to Elmer has not been possible for some time - apparently this has now changed, but I suspect that the implementation is rather fresh and possibly might still have some bugs...
Matthias
try to export the mesh in gmsh format 2 and import that to Elmer. gmsh format 4 import to Elmer has not been possible for some time - apparently this has now changed, but I suspect that the implementation is rather fresh and possibly might still have some bugs...
Matthias
Re: Electric field calculation using Gmsh and Elmer.
Thank you for the comment.
In fact, I used following commands.
gmsh ***.geo -3 -order 2 -optimize_netgen -refine
ElmerGrid 14 2 ***.msh -autoclean
ElmerSolver ***.sif
Best regards.
In fact, I used following commands.
gmsh ***.geo -3 -order 2 -optimize_netgen -refine
ElmerGrid 14 2 ***.msh -autoclean
ElmerSolver ***.sif
Best regards.
Re: Electric field calculation using Gmsh and Elmer.
Hi,
to get msh format 2, add
-format msh2
to your gmsh command (see viewtopic.php?f=3&t=6654&start=10).
HTH,
Matthias
to get msh format 2, add
-format msh2
to your gmsh command (see viewtopic.php?f=3&t=6654&start=10).
HTH,
Matthias
Re: Electric field calculation using Gmsh and Elmer.
Thank you very much!
I did not know about msh2
After using msh2, the number of times Gmsh/Elmer fail to load the mesh-calculation was reduced. Actually, it is not completely solved, but, it is the very good improvement
Best regards.
I did not know about msh2
After using msh2, the number of times Gmsh/Elmer fail to load the mesh-calculation was reduced. Actually, it is not completely solved, but, it is the very good improvement
Best regards.
Re: Electric field calculation using Gmsh and Elmer.
Hi,
what kind of problem do you have with msh2 format?
Matthias
what kind of problem do you have with msh2 format?
Matthias