Hi,
I am trying to implement static current, there are 2 issues now,
Version : Elmer 9.0
1. the converge is not good. so I have to relax to 1e-3.
2. The voltage drop is huge. the potential(violtage) drop to 0 in the middle of the power plane. so it is wrong, it cannot be such high drop.
Electric Conductivity is changed to 47e6 S/M^2 for PCB copper,
BC1 : Pad1 potential 5V
BC2 : Pad2 current density 7500 on 400 amp/mm^2
the mesh unv is created by SALOME-9.10.0. My guess is that prbably the unit in the mesh is wrong, from mm to m, so it has 1000 scale. I don't know hot to check the mesh unit. I attached the mesh, sif and results.
thank you if you can give me advise.
static current on thin copper sheet. Votage drop is huge and converge not good
static current on thin copper sheet. Votage drop is huge and converge not good
- Attachments
-
- try1.zip
- (26.46 KiB) Downloaded 24 times
-
- pwr_plane_made_by_freecad_unv.zip
- (722.28 KiB) Downloaded 25 times
-
- case.sif
- (2.01 KiB) Downloaded 25 times
Re: static current on thin copper sheet. Votage drop is huge and converge not good
I just did a simple experiment, it looks like the step file need to scale down to meter.
More checks are under process...
More checks are under process...
Re: static current on thin copper sheet. Votage drop is huge and converge not good
Hello,
To check the dimensions of the mesh, there are two simple methods. After running ElmerGrid to convert the mesh from unv format into Elmer mesh format, you could:
1. Open the mesh.nodes in a spreadsheet program and create charts such as x versus y coordinates, or x versus z coordinates.
2. Open the mesh files in ElmerGUI, and click on Model, Summary, and look for the bounding box dimensions.
Using the second method, here are the results:
Rich.
To check the dimensions of the mesh, there are two simple methods. After running ElmerGrid to convert the mesh from unv format into Elmer mesh format, you could:
1. Open the mesh.nodes in a spreadsheet program and create charts such as x versus y coordinates, or x versus z coordinates.
2. Open the mesh files in ElmerGUI, and click on Model, Summary, and look for the bounding box dimensions.
Using the second method, here are the results:
Looks like the unv file was converted into meters. Note that the width to thickness ratio is 0.174 m / 5.4e-5 m = 3,222.BOUNDING BOX
X-coordinate: [ -0.0676753 , 0.0878975 ]
Y-coordinate: [ -0.0871058 , 0.0869226 ]
Z-coordinate: [ 0 , 5.4e-05 ]
Rich.
Re: static current on thin copper sheet. Votage drop is huge and converge not good
Hello,
Attached is another sif file, this one uses the direct solver, umfpack.
Edit: updated the sif file, to more closely match the original case.sif file. Looks like the unv file was converted by ElmerGUI, which means that the default action will be '-autoclean'. The smaller pad is boundary number 26, and the larger pad is boundary number 25.
Rich.
Attached is another sif file, this one uses the direct solver, umfpack.
Edit: updated the sif file, to more closely match the original case.sif file. Looks like the unv file was converted by ElmerGUI, which means that the default action will be '-autoclean'. The smaller pad is boundary number 26, and the larger pad is boundary number 25.
Rich.
- Attachments
-
- case.sif
- (1.73 KiB) Downloaded 29 times