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.

results from frequency response is not matching the eigen v

Please login with a confirmed email address before reporting spam

Hello everyone,

I am modeling a structural mechanics with ale mesh frame and electrostatic in this frame.. in an axi-symmetric 2d model (MEMS module)..

The problem is ... When I solve for the eigenvalues I got a resonance at 1MHz and then at 4.1MHz and the third at 9MHz

but when I solve for the frequency response analysis I found different result


thanks you so much for your support


yours
Mark George

6 Replies Last Post 12.02.2012, 06:32 GMT-5

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 11.02.2012, 06:31 GMT-5
kindly find design attached
kindly find design attached


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 11.02.2012, 06:32 GMT-5
Hi

if you are in v4.2 or 4.2a you should be aware of the importance od the order in which you apply the multiple physics, either you use the coupled (pre-cooked) EMI physics, or you should use the ALE + Solid+ EC/ES physcs, but then be sure you START with the ALE before the solid, as both use the spatial frame, and if you arrange them the other way you will have some surprises with the result. COMSOL has told me they will issue a warning, or automatically arrange the order in the next release, but currently with 4.2 one must take care manually

--
Good luck
Ivar
Hi if you are in v4.2 or 4.2a you should be aware of the importance od the order in which you apply the multiple physics, either you use the coupled (pre-cooked) EMI physics, or you should use the ALE + Solid+ EC/ES physcs, but then be sure you START with the ALE before the solid, as both use the spatial frame, and if you arrange them the other way you will have some surprises with the result. COMSOL has told me they will issue a warning, or automatically arrange the order in the next release, but currently with 4.2 one must take care manually -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 11.02.2012, 06:39 GMT-5
Thanks for yours reply

but I use COMSOL 3.4 not the new version


Mark George
Thanks for yours reply but I use COMSOL 3.4 not the new version Mark George

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 11.02.2012, 07:29 GMT-5
The same problem occurs .. the output frequency response couldn't be right at all


:S
The same problem occurs .. the output frequency response couldn't be right at all :S

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 12.02.2012, 03:34 GMT-5
Thank you Ivar for your answer

But Is it important in all the analysis ( parametric and eigenvalues ) or in the frequency response analysis only ?

As when I run the parametric analysis and the eigenvalues I got results near the predicted results

but when I run the frequency response analysis I got a very far results from the predicted


So where is the problem ?????


yours
Mark George
Thank you Ivar for your answer But Is it important in all the analysis ( parametric and eigenvalues ) or in the frequency response analysis only ? As when I run the parametric analysis and the eigenvalues I got results near the predicted results but when I run the frequency response analysis I got a very far results from the predicted So where is the problem ????? yours Mark George

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 12.02.2012, 06:32 GMT-5
Hi

not sure why, as I do not have any older version (only the latest 4.2a) running I cannot test anything out, from my remindings, frequency scans and eigenmodes agreed,

I usually run an eigenfrequency first to define the modes, and then use those to define the stepping (manually with many points around the eigenfrequencies). If you leave the system free it will most proably skip all narrow peaks.

Adding some damping often helps the convergence too

--
Good luck
Ivar
Hi not sure why, as I do not have any older version (only the latest 4.2a) running I cannot test anything out, from my remindings, frequency scans and eigenmodes agreed, I usually run an eigenfrequency first to define the modes, and then use those to define the stepping (manually with many points around the eigenfrequencies). If you leave the system free it will most proably skip all narrow peaks. Adding some damping often helps the convergence too -- 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.