Menu

Discussion Forum

New Discussion
Filter

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 The information provided may be out of date.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.

Saw sensor with boundary load

Posted Aug 15, 2011, 8:31 a.m. EDT MEMS & Nanotechnology, Acoustics & Vibrations, MEMS & Piezoelectric Devices, Modeling Tools & Definitions, Parameters, Variables, & Functions Version 4.2 6 Replies

Peter Kropelnicki
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

hello together,

Im facing a problem in comsol 4.1. I'm putting a boundary load onto the backside of my membran (piezo 2um, SiO2 1um, silicon 100um). From both side the mambran is fixed. the Saw fingers have a 10 um distance. I found the resonance frequency of the SAW (only horizontal SAW wave are active according to Scheerschmidt comsol paper). Unfortunately the boundary load is not stored and works only until the stationary state in the solver. Even with a prediscriped displacement it doesnt work, the displacement is reset to zero during after solving the eigenfrequency solution. So Im not seeing the membran-bow but only the SAW displacement on the surface. Can someone help me with this problem? Is this problem solvable with comsol 4.2?
6 Replies Last Post Aug 18, 2011, 5:02 a.m. EDT
Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Aug 17, 2011, 2:24 a.m. EDT
Hi

eigenfrequency analysis is a linear analysis, that is your model is linearised around the initial condition state and external loads (i.e forces, pressure) are not considered but deformation and stress are considered, so you must ensure that your load is applied by a stationary analysis and then that your eigenfrequency analysis is using this first stationary RESULTS as initial conditions.

try out the buckling analysis example of COMSOL Model library and see how COMSOL set's up the defualt solver to a stationary solver node + store solution node + eigenfrequency solver node (its a special wigenfrequency set-up for buckling, not exactly your case but the methodology remains valid)

--
Good luck
Ivar
Hi eigenfrequency analysis is a linear analysis, that is your model is linearised around the initial condition state and external loads (i.e forces, pressure) are not considered but deformation and stress are considered, so you must ensure that your load is applied by a stationary analysis and then that your eigenfrequency analysis is using this first stationary RESULTS as initial conditions. try out the buckling analysis example of COMSOL Model library and see how COMSOL set's up the defualt solver to a stationary solver node + store solution node + eigenfrequency solver node (its a special wigenfrequency set-up for buckling, not exactly your case but the methodology remains valid) -- Good luck Ivar
Peter Kropelnicki
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Aug 17, 2011, 4:41 a.m. EDT
Dear Ivar,

I cant find the buckling analysis. In which module is it? What is the exact name?

Best regards,

Peter
Dear Ivar, I cant find the buckling analysis. In which module is it? What is the exact name? Best regards, Peter
Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Aug 17, 2011, 6:17 a.m. EDT
Hi

I find them here, but indeed you need the structural module, forgot about that

--
Good luck
Ivar
Hi I find them here, but indeed you need the structural module, forgot about that -- Good luck Ivar

Attachments:

Peter Kropelnicki
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Aug 17, 2011, 9:45 p.m. EDT
Hey Ivar,

got it. Hm, in the eigenfrequency it is clear to me not to see the bending, but what about the simple frequency analysis? There is not linearization, so the statistic preload and the mesh bending should be visible, isn't it?
Hey Ivar, got it. Hm, in the eigenfrequency it is clear to me not to see the bending, but what about the simple frequency analysis? There is not linearization, so the statistic preload and the mesh bending should be visible, isn't it?
Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Aug 18, 2011, 1:42 a.m. EDT
Hi

for harmonic solvers the values you use for the external forces are assumed to be the amplitude of a sinus type excitation, hence you will be fooled there too, in the sense you will not get a static load deformation on top.

Good question now: How to apply a deformation and do a harmonic analysis on top of the static case?

perhaps to perform a stationary case, save the initial conditions, get the static stress as preload and then do a harmonic run, but there will be noting to avoid that the parts straightens up, to release the stress if there is no counter force ...

Remains the time transient solver, here you can apply a static + superposed time variable load but it requires many iterations to get to a steady state

--
Good luck
Ivar
Hi for harmonic solvers the values you use for the external forces are assumed to be the amplitude of a sinus type excitation, hence you will be fooled there too, in the sense you will not get a static load deformation on top. Good question now: How to apply a deformation and do a harmonic analysis on top of the static case? perhaps to perform a stationary case, save the initial conditions, get the static stress as preload and then do a harmonic run, but there will be noting to avoid that the parts straightens up, to release the stress if there is no counter force ... Remains the time transient solver, here you can apply a static + superposed time variable load but it requires many iterations to get to a steady state -- Good luck Ivar
Peter Kropelnicki
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Aug 18, 2011, 5:02 a.m. EDT
Hello Iyar,

I found another problem in my simulation. I have simulated the acoustic wave based on the 2D axisymmetric simulation. Due to this I allways get a serious problem here, even Im not in the resonance frequency. Please find attached my result. There is no nice SAW in, because of the symmeric condition. Need your advice here,

Best regards,

Peter
Hello Iyar, I found another problem in my simulation. I have simulated the acoustic wave based on the 2D axisymmetric simulation. Due to this I allways get a serious problem here, even Im not in the resonance frequency. Please find attached my result. There is no nice SAW in, because of the symmeric condition. Need your advice here, Best regards, Peter

Attachments:

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.

Suggested Content