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.
Parallel Plate Capacitor
Posted Jun 13, 2012, 4:40 p.m. EDT MEMS & Nanotechnology, MEMS & Piezoelectric Devices 4 Replies
Please login with a confirmed email address before reporting spam
This is my first time using Comsol. My project is to model the capacitance of a system with the dimensions as follows:
Top Electrode: 2e-3m x 2e-3m x 100e-9m
Silicon Substrate: 2e-3m x 2e-3m x 2e-6m
Gap: 1e-6m
Bottom plate: 3.5e-3m x 3.5e-3m x 2e-6m
Bottom Electrode: 3.5e-3m x 3.5e-3m x 100e-9m.
Because Comsol cannot automatically mesh over these dimensions, I decided to reduce the system down into something Comsol could mesh first. Then I would build the 2e-3 and 3.5e-3m dimensions from this smaller one.
These dimensions are:
Top/Bottom Electrodes: 1e-4m x 1e-4m x 100e-9m
Top/Bottom Substrates: 1e-4m x 1e-4m x 2e-6m
Gap: 1e-6m
I managed to get Comsol to solve this system by following the documentation for the Capacitive 3D Comb Drive:
file:///C:/COMSOL35/doc/mems/wwhelp/wwhimpl/common/html/wwhelp.htm#href=memsactuatormodels.18.4.html&single=true
When I try to add another version of the above system by using Copy+Paste, the array tool, or the mirror tool, Comsol will not mesh this structure either (Error 4100). I tried using the "Interactive Meshing" feature and meshed each boundary, but then I get Error 7067: System Matrix is 0.
I would appreciate any help and suggestions anyone has... I've been stuck on this problem a while.
Top Electrode: 2e-3m x 2e-3m x 100e-9m
Silicon Substrate: 2e-3m x 2e-3m x 2e-6m
Gap: 1e-6m
Bottom plate: 3.5e-3m x 3.5e-3m x 2e-6m
Bottom Electrode: 3.5e-3m x 3.5e-3m x 100e-9m.
Because Comsol cannot automatically mesh over these dimensions, I decided to reduce the system down into something Comsol could mesh first. Then I would build the 2e-3 and 3.5e-3m dimensions from this smaller one.
These dimensions are:
Top/Bottom Electrodes: 1e-4m x 1e-4m x 100e-9m
Top/Bottom Substrates: 1e-4m x 1e-4m x 2e-6m
Gap: 1e-6m
I managed to get Comsol to solve this system by following the documentation for the Capacitive 3D Comb Drive:
file:///C:/COMSOL35/doc/mems/wwhelp/wwhimpl/common/html/wwhelp.htm#href=memsactuatormodels.18.4.html&single=true
When I try to add another version of the above system by using Copy+Paste, the array tool, or the mirror tool, Comsol will not mesh this structure either (Error 4100). I tried using the "Interactive Meshing" feature and meshed each boundary, but then I get Error 7067: System Matrix is 0.
I would appreciate any help and suggestions anyone has... I've been stuck on this problem a while.
4 Replies Last Post Jun 18, 2012, 1:53 p.m. EDT
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
when you have such thin layers, the best is to use "surface thin" boundaries, as if you want to know something about what happens inside the 3D thin layers you must mesh them with 3 or more elements in the thickness, and your RAm and solving time explodes
check the doc
--
Good luck
Ivar
when you have such thin layers, the best is to use "surface thin" boundaries, as if you want to know something about what happens inside the 3D thin layers you must mesh them with 3 or more elements in the thickness, and your RAm and solving time explodes
check the doc
--
Good luck
Ivar
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Are the "surface thin" boundaries part of a specific module, if so which one?
The problem with designating the mesh with 3 or more elements is that I made a composite object with the surrounding air and the two electrodes (as in the Capacitive Comb model documentation). Can I still manually mesh this object?
Also, which "doc" are you talking about?
Thanks for the quick response.
The problem with designating the mesh with 3 or more elements is that I made a composite object with the surrounding air and the two electrodes (as in the Capacitive Comb model documentation). Can I still manually mesh this object?
Also, which "doc" are you talking about?
Thanks for the quick response.
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
you should always be capable of meshing a model, and it should solve with only 1 element in the thickness, the thing is how representative is it ?, are variations of your dependent variables, in the thickness of the part, important at all ?
Sometimes, to make the messhing easier it is worth subdividing the geometry into several domains with "internal" boundaries, that are there "just to help" the mesher. If your model can be made from only recangles or blocks, its often worth to make many of them and its not always worth to make unions and differences, just leave the last "geometry - Union" mode athe the Finish line (geomanalyze in 3.5) do the domain sorting
--
Good luck
Ivar
you should always be capable of meshing a model, and it should solve with only 1 element in the thickness, the thing is how representative is it ?, are variations of your dependent variables, in the thickness of the part, important at all ?
Sometimes, to make the messhing easier it is worth subdividing the geometry into several domains with "internal" boundaries, that are there "just to help" the mesher. If your model can be made from only recangles or blocks, its often worth to make many of them and its not always worth to make unions and differences, just leave the last "geometry - Union" mode athe the Finish line (geomanalyze in 3.5) do the domain sorting
--
Good luck
Ivar
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
I believe any variation in the thickness are important. I have tried subdividing the geometry up, but I did not use the Union at all. Should I use this after I define all boundary conditions?
When defining boundary conditions I was wondering if you could tell me the difference between using Distributed Capacitance vs Zero Charge/Symmetry? I am currently defining the Silicon substrates in between the two gold electrodes (the dielectric) with the Zero Charge.
Thanks,
--Jordan
When defining boundary conditions I was wondering if you could tell me the difference between using Distributed Capacitance vs Zero Charge/Symmetry? I am currently defining the Silicon substrates in between the two gold electrodes (the dielectric) with the Zero Charge.
Thanks,
--Jordan