Menu

Discussion Forum

New Discussion
Filtrer

Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.

Create Composite Object Error

Posted 10 nov. 2010, 18:38 UTC−5 5 Replies

Michael Purdy
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Hello,

I am very new to COMSOL and have had some troubles creating an object to simulate.
I am trying to create an electrode on top of a glass wafer. The electrode is made of two layers, one 5nm thick and one 150nm thick. I have made the wafer 1mm thick, and placed the 5nm followed by the 150nm sections of the electrode on top of the wafer. I have placed all of these inside a testable area of size 1cm by 1cm.

When I try to 'Create Composite Object' I set the formula as:
BLK1-EXT1-EXT2-EXT3-EXT4-EXT5
where EXT1 is the wafer and EXT2 through EXT5 are sections of the electrodes (there are two parts to each layer of the electrode).
However, I keep getting an error that says "internal error in geometry decomposition."

Does anyone know what this error means and how I can fix it? Like I said, I am very new to the COMSOL software, so I apologize for the likely rather simple question.
Thanks very much.

Michael
5 Replies Last Post 18 nov. 2010, 11:26 UTC−5
Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 11 nov. 2010, 08:18 UTC−5
Hi

I suppose its related to the height/size ratios 5nm versus 1mm is 1:200'000, I never use ratios (for the mesher) largeer than 10'000 or if possible less.

The true question is what do you want to resolve within your 5nm layer? , cannot you replace its presence by a "thin layer" with a bulk equation form ?

Then have you calculated what it represents to have a 1nm (to have let's say 5 meshes across your 5 nm thick layer) sized mesh elements to fill several mm^3, probably you need a nice cluster just to fill one matrix, not to talk about solving

Finite elements are limited by the "finite" size of RAM on ourPC/WS's ;)

--
Good luck
Ivar
Hi I suppose its related to the height/size ratios 5nm versus 1mm is 1:200'000, I never use ratios (for the mesher) largeer than 10'000 or if possible less. The true question is what do you want to resolve within your 5nm layer? , cannot you replace its presence by a "thin layer" with a bulk equation form ? Then have you calculated what it represents to have a 1nm (to have let's say 5 meshes across your 5 nm thick layer) sized mesh elements to fill several mm^3, probably you need a nice cluster just to fill one matrix, not to talk about solving Finite elements are limited by the "finite" size of RAM on ourPC/WS's ;) -- Good luck Ivar
Michael Purdy
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 12 nov. 2010, 12:12 UTC−5
That's a good point. Thanks very much.

Michael
That's a good point. Thanks very much. Michael
Michael Purdy
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 17 nov. 2010, 15:03 UTC−5
I have been able to create a composite object by making the box that encapsulates everything smaller. However, I have now had a lot of trouble meshing and getting a solution. I cannot mesh the entire object using Free Mesh even with an extremely coarse setting without running out of memory. If I try to Sweep Mesh the small sections of the model and Free Mesh the rest I often get a 'Singular Matrix' error (which I do not understand). I have been able to mesh the entire object under a few circumstances, but when I then try to solve it I get a 'Empty cavity generated' error.
I assume that this is because the resolution of the mesh is not small enough in the very thin regions. I can perhaps try approximating the thin layer with a 2D sheet but obviously it would be ideal if I could simulate exactly what is happening in the electrodes.

Thanks for any suggestions you might have.

Michael
I have been able to create a composite object by making the box that encapsulates everything smaller. However, I have now had a lot of trouble meshing and getting a solution. I cannot mesh the entire object using Free Mesh even with an extremely coarse setting without running out of memory. If I try to Sweep Mesh the small sections of the model and Free Mesh the rest I often get a 'Singular Matrix' error (which I do not understand). I have been able to mesh the entire object under a few circumstances, but when I then try to solve it I get a 'Empty cavity generated' error. I assume that this is because the resolution of the mesh is not small enough in the very thin regions. I can perhaps try approximating the thin layer with a 2D sheet but obviously it would be ideal if I could simulate exactly what is happening in the electrodes. Thanks for any suggestions you might have. Michael
Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 18 nov. 2010, 05:29 UTC−5
Hi

yes meshing is an art ;) but are you really sure that the you will have a temperature gradient in the 5nm thickness direction, hence obliging you to use a 3D meshing and not surface physics ?

--
Good luck
Ivar
Hi yes meshing is an art ;) but are you really sure that the you will have a temperature gradient in the 5nm thickness direction, hence obliging you to use a 3D meshing and not surface physics ? -- Good luck Ivar
Michael Purdy
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 18 nov. 2010, 11:26 UTC−5
That's a good point. I will try to simulate it as a surface rather than a 3D object.

Thanks again.

Michael
That's a good point. I will try to simulate it as a surface rather than a 3D object. Thanks again. Michael

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.

Suggested Content