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 The information provided may be out of date.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.

Simple contact mode using electrostatic cantilever model

Please login with a confirmed email address before reporting spam

Hi everyone,

I was going through the electrostatic cantilever model (3D) and was trying to simulate a contact mode. What I did was choosing the boundary of the cantilever and the opposite ground surface as a contact pair, after that I just increase the parameter V_in hoping that the cantilever will come into contact with the ground surface. But the generates an error 6170, while the simulation is running.

The reason why I want to simulate this is I wanted to find out the mode shape if an excessive voltage is applied onto the cantilever, in both DC/AC, steady state and transient. The actual simulation will be imposed on some 1D nano material like nanowire and nano wall.

Hope you guys can help me with this. Any suggestion is really appreciated. Thanks

Rgds,
Jeff

3 Replies Last Post Mar 12, 2011, 5:12 a.m. EST
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 Mar 11, 2011, 1:03 a.m. EST
Hi

one of the difficulties with an electrostatic pulling of a cantilever is that the force - voltage relation is non linear as the electric filed depends on the distance and this distance is reducing, hence increases the electric field again and increases the forces, which further reduces the gap => you end up with a pull in effect (positive feedback). This can also be seen a s a bifurcation or a loop-back of the domain of stable solutions. Such jumps are very difficult for a solver to pass through, should it go left or right ?

That is why driving a contact case with a linearly driven force/pressure is far better (for the solver purpose)

--
Good luck
Ivar
Hi one of the difficulties with an electrostatic pulling of a cantilever is that the force - voltage relation is non linear as the electric filed depends on the distance and this distance is reducing, hence increases the electric field again and increases the forces, which further reduces the gap => you end up with a pull in effect (positive feedback). This can also be seen a s a bifurcation or a loop-back of the domain of stable solutions. Such jumps are very difficult for a solver to pass through, should it go left or right ? That is why driving a contact case with a linearly driven force/pressure is far better (for the solver purpose) -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Mar 12, 2011, 2:23 a.m. EST
Hi Ivar,

Thanks for the reply. I have search the forum and found similar cases due to non-linear solution between the electrostatic force and the closing gap and hence given up the electrostatic model for the mode shape investigation.

As you suggested, now I am using a linear system, with pure structural cantilever with distributed load and I was able to get a contact mode with a relatively higher load so that the cantilever collapses onto the opposite contact.

My question is:

1. When I tried to simulate this in transient/parametric and it gives me an singular matrix error. Is it possible to to parametric / transient analysis in COMSOL with contact pair?

2. ALE is always necessary to generate post solution that shows the deformed boundary displacement?

3. In electrostatic case, is there anyway to simulate electrostatic pull-out voltage, with initial contacted surface?

Thanks

Rgds,
Jeff
Hi Ivar, Thanks for the reply. I have search the forum and found similar cases due to non-linear solution between the electrostatic force and the closing gap and hence given up the electrostatic model for the mode shape investigation. As you suggested, now I am using a linear system, with pure structural cantilever with distributed load and I was able to get a contact mode with a relatively higher load so that the cantilever collapses onto the opposite contact. My question is: 1. When I tried to simulate this in transient/parametric and it gives me an singular matrix error. Is it possible to to parametric / transient analysis in COMSOL with contact pair? 2. ALE is always necessary to generate post solution that shows the deformed boundary displacement? 3. In electrostatic case, is there anyway to simulate electrostatic pull-out voltage, with initial contacted surface? Thanks Rgds, Jeff

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 Mar 12, 2011, 5:12 a.m. EST
Hi

a singular matrix normally means tat there are not enough BC defined to get an unique solution (or some are lineraly dependent)

contacts work from what I remember for most solvers, certainly transient.

One advice, use linear elements for the contact region at least

ALE is required for large dispalcements, but do not forget that if you simulate true contact the mesh size becomes =0 between the parts, this is a topolog change and I doubt that the solvers would appreciate.
Use an offset to allow a thin region of the ALE mesh to exist between the parts

Normally you should be able to reverse the operation, but you need a way to start in a close to collapsed situation (i.e. pressure load it and save the reuslts and restart from there with a voltage above the pull-in effect, then release gently the voltage with a paramterical sweep. In theory I belive this should work, but I haventtried it out, not sure how much effort it is. Might have already been discussed in one of the COMSOL conference papers as, for me, this is a natural way to try

--
Good luck
Ivar
Hi a singular matrix normally means tat there are not enough BC defined to get an unique solution (or some are lineraly dependent) contacts work from what I remember for most solvers, certainly transient. One advice, use linear elements for the contact region at least ALE is required for large dispalcements, but do not forget that if you simulate true contact the mesh size becomes =0 between the parts, this is a topolog change and I doubt that the solvers would appreciate. Use an offset to allow a thin region of the ALE mesh to exist between the parts Normally you should be able to reverse the operation, but you need a way to start in a close to collapsed situation (i.e. pressure load it and save the reuslts and restart from there with a voltage above the pull-in effect, then release gently the voltage with a paramterical sweep. In theory I belive this should work, but I haventtried it out, not sure how much effort it is. Might have already been discussed in one of the COMSOL conference papers as, for me, this is a natural way to try -- 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.