Note: This discussion is about an older version of the COMSOL Multiphysics® software. The information provided may be out of date.

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.

Stationary study failed to converge

Please login with a confirmed email address before reporting spam

Hello, I am trying to do a stationary of a MEMS device, but Comsol keeps failing to converge.

Some background on the model:
The model is a MEMS consisting of a plate attached to 4 thin cantilever beam springs, which are in turn attached to a fixed substrate. The plate is also attached to a set of combs whose teeth act as capacitors with a set of corresponding combs attached to the fixed substrate (see attached image. Everything not in purple is attached to the plate and should be free to move). A voltage is applied to the fixed combson one side of the device, while the plate itself is grounded. This causes the plate to be displaced until the capacitive force from the comb capacitors is balanced by the restorative force of the springs.

Comsol Physics setup:
I am using the Electromechanics physics node in Comsol 4.2a to do a stationary study on how much the plate is displaced. I placed all domains in Linear Elastic material model, fixed the boundaries shown in purple in the image, applied a voltage to the three fixed combs in the right of the image, and grounded the center plate. I then created a stationary study and computed. Unfortunately, the solver fails to converge. Is there something wrong with the way I have set everything up? Does anyone have suggestions about what may be going wrong?

Josh


3 Replies Last Post 18 mai 2012, 16:31 UTC−4
Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 18 mai 2012, 10:30 UTC−4
Hi

are you sure you have enough "air" all around to allow the EC part of the physics to calculate the fields correctly (zthat is not obvious to tell from your image). Then be sure you have a few elements for the comb and the air gaps, as else the fields are not well resolved.

Finally check that you have the BC settings such that the forces couples, use plot while solving and probe plots to monitor forces stress and/or displacement of critical points

And sometimes it's worth starting with only 1 physics at the time, resolving forst structural (and stressunder body or BC loads, then solve ACDC alone and get forces, then mix both


--
Good luck
Ivar
Hi are you sure you have enough "air" all around to allow the EC part of the physics to calculate the fields correctly (zthat is not obvious to tell from your image). Then be sure you have a few elements for the comb and the air gaps, as else the fields are not well resolved. Finally check that you have the BC settings such that the forces couples, use plot while solving and probe plots to monitor forces stress and/or displacement of critical points And sometimes it's worth starting with only 1 physics at the time, resolving forst structural (and stressunder body or BC loads, then solve ACDC alone and get forces, then mix both -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 18 mai 2012, 15:18 UTC−4
Ivar, Thanks for your reply and advice. I added 3 air domains in the space between the three comb electrode pairs (I didn't know this step was necessary- I assumed Comsol would break up a larger air domain as much as was needed when it created the mesh). However, when I tried to run the study, I received an input/output error from the stationary solver. I've run into this error before, and have never been able to figure out what causes it. Do you know what its source might be?
Ivar, Thanks for your reply and advice. I added 3 air domains in the space between the three comb electrode pairs (I didn't know this step was necessary- I assumed Comsol would break up a larger air domain as much as was needed when it created the mesh). However, when I tried to run the study, I received an input/output error from the stationary solver. I've run into this error before, and have never been able to figure out what causes it. Do you know what its source might be?

Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 18 mai 2012, 16:31 UTC−4
Hi

its not evident to sy much just like that, from the image one believes that there are some missing air domains between the electrodes, but it could be that you had already removed them for the image ?

Often its worth to start with a much simpler model (3 teeeth comb), build it up test the simulations and convergence, validate it and check the BC settings etc, then only add the full blown model

--
Good luck
Ivar
Hi its not evident to sy much just like that, from the image one believes that there are some missing air domains between the electrodes, but it could be that you had already removed them for the image ? Often its worth to start with a much simpler model (3 teeeth comb), build it up test the simulations and convergence, validate it and check the BC settings etc, then only add the full blown model -- Good luck Ivar

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.