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.

Modelling a changing capacitance in two Steps

Please login with a confirmed email address before reporting spam

Hello @ all,

I try moddling a Capacitor, where one plate is moved by an acceleration. This should course a deformation and this should copurse a change in Capacitance, whitch I want to calculate.

I tried the following:

STEP 1: Calculate the displacement of the plate via the Solid-Structuremechanics-modul. This works fine.

STEP 2: Calculate the Capacitance via the electrostatics-modul. I am using a terminalwith 1 V at the one plate, a ground at the other plate. This solution alone works also fine. But I am not able to couple these two.

I thought: Just get the resulting deformation from step 1 as Coinstaind for STEP 2. I tried this by setting: 'Initial values solved for' in Study 2 as 'Solution, Solver 1'. And 'Values of Variables not solved for' as 'Solution, Solver 1', too. I thought that now the displacement should be assumed for STEP 2, but it isn't.

Where is my false? Is there a tutorial for coupling two physics elsewhere?

Kind reguard,
H.Schweiger

5 Replies Last Post 2012年1月31日 GMT-5 09:06
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 2012年1月30日 GMT-5 15:10
Hi

you need to get the two coupled,

One way: calculate the deformations with solid of the plate. Then remesh the deformed structure by adding some surrounding "air" and use that for the 2nd ES step to calculate the field and therefrom the capacitance. This assumes that the ES measurement with some electromagnetic field, hence EM forces too, are negligible w.r.t. the acceleration forces.

By the way you say acceleration induced deformation: do you simulate that via a body load F[N/m^3] = rho*acc ? or ?

The other way is to use a coupled physics this implies a ALE step to deform the air region such that you get a capacitance change. Check the model library I believ there is a few examples therein

--
Good luck
Ivar
Hi you need to get the two coupled, One way: calculate the deformations with solid of the plate. Then remesh the deformed structure by adding some surrounding "air" and use that for the 2nd ES step to calculate the field and therefrom the capacitance. This assumes that the ES measurement with some electromagnetic field, hence EM forces too, are negligible w.r.t. the acceleration forces. By the way you say acceleration induced deformation: do you simulate that via a body load F[N/m^3] = rho*acc ? or ? The other way is to use a coupled physics this implies a ALE step to deform the air region such that you get a capacitance change. Check the model library I believ there is a few examples therein -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2012年1月31日 GMT-5 03:33
Thanks.

Yes. I am modelling the acceleration via bodyload.

I tried it using ALE. I modelled a weak air-domain (very small E-modul). I have computed the olid part, after that I have remeshed the deformed mesh and made an electrostatic analyses witch it.

This works fine.

But later I have acceleration with sine-form and an frequency of 10 ... 100 Hz. Is there any way to modell this?
Thanks. Yes. I am modelling the acceleration via bodyload. I tried it using ALE. I modelled a weak air-domain (very small E-modul). I have computed the olid part, after that I have remeshed the deformed mesh and made an electrostatic analyses witch it. This works fine. But later I have acceleration with sine-form and an frequency of 10 ... 100 Hz. Is there any way to modell this?

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 2012年1月31日 GMT-5 04:00
Hi

careful with the small modulus approach all in solid, things have changed in v4.2a,

I would suggest to use an ALE driven by the Solid.

But important put the ALE physics node first BEFORE the SOLID otherwise you will get strange results.

It is the last node that is the driver for the deformation (Solid is a kind of ALE already with the spatial frame approach). Then be sure your SOLID is limited to the sold deformation domains, and that your ALE is limited (not stricly required but it makes the model simpler) to the surrounding Air. At the last step add a ES for the eelctric field (not sure it should be the last will have to try myself, but order ALE + SOLID last is important, I have just had a mail exchange with support on exactly that point ;).

But you should also have the EMI physics that couples all three for you, no ?

--
Good luck
Ivar
Hi careful with the small modulus approach all in solid, things have changed in v4.2a, I would suggest to use an ALE driven by the Solid. But important put the ALE physics node first BEFORE the SOLID otherwise you will get strange results. It is the last node that is the driver for the deformation (Solid is a kind of ALE already with the spatial frame approach). Then be sure your SOLID is limited to the sold deformation domains, and that your ALE is limited (not stricly required but it makes the model simpler) to the surrounding Air. At the last step add a ES for the eelctric field (not sure it should be the last will have to try myself, but order ALE + SOLID last is important, I have just had a mail exchange with support on exactly that point ;). But you should also have the EMI physics that couples all three for you, no ? -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2012年1月31日 GMT-5 04:15
Thanks.

I think I have understood.

I do have a Licence for AC/DC Module, RF-Module and Solid Mechanics here.

Is EMi in there? What is it?

Edit: I junst found out that it has to be the MEMS-Module? And then Electromechanics in there? Is that right? because we don't have a Licence for this Module

Edit2: Just togy we got a new Licence for the MST-Module. Perhaps this is a solution?

Edit 3: Now, I saw that the MST modul contains the emi physics. It runs very easysly and fine.

Now the only question is how to apply the 10 Hz acceleration?
Thanks. I think I have understood. I do have a Licence for AC/DC Module, RF-Module and Solid Mechanics here. Is EMi in there? What is it? Edit: I junst found out that it has to be the MEMS-Module? And then Electromechanics in there? Is that right? because we don't have a Licence for this Module Edit2: Just togy we got a new Licence for the MST-Module. Perhaps this is a solution? Edit 3: Now, I saw that the MST modul contains the emi physics. It runs very easysly and fine. Now the only question is how to apply the 10 Hz acceleration?

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 2012年1月31日 GMT-5 09:06
Hi

you apply/add a "structural domain body load" sub node, but 10Hz mean harmonic/frequnecy domain solvers,

But I have just combined a ALE + Struct + ES and that seems to work fine too (in that order)

I believe EMI was intended for PZT as it defines a series of local coordinates for PZT orientation, but it should also work for capacitive items

Apart that if I calculate the capacitance from 2*energy/V^2 its constant while not in my ALE+Solid+ES case

--
Good luck
Ivar
Hi you apply/add a "structural domain body load" sub node, but 10Hz mean harmonic/frequnecy domain solvers, But I have just combined a ALE + Struct + ES and that seems to work fine too (in that order) I believe EMI was intended for PZT as it defines a series of local coordinates for PZT orientation, but it should also work for capacitive items Apart that if I calculate the capacitance from 2*energy/V^2 its constant while not in my ALE+Solid+ES case -- 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.