Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

COMSOL _ frequency response of piezoelectric device

  1. Apr 8, 2009 #1

    I'm trying to model the frequency response of a thin PZT disc using COMSOL, but I am not sure how to calculate the impedance Z vs frequency. I defined the current density at the top electrode as I=-imag(nJ_smpz3d) and then plotted V/I using the domain plot parameters in the postproc menu. This gives me a V/I vs freq plot, but V/I is not equal to Z!

    Does anybody know how to calculate Z? And the Phase?

  2. jcsd
  3. Apr 8, 2009 #2


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member


    Did you compute the solution for the entire disc or did you make use of the symmetry and only compute the solution for a segment?
    Last edited: Apr 8, 2009
  4. Apr 9, 2009 #3
    In fact I did both cases, the main difference is that when I used symmetries then defined the variable I as I=-imag(nJ_smpz3d)*N, where N is the number of segments contained in the whole disc, i.e., if a 10 deg segment then N=36. I obtained in both ases the same solution.

    However the question still remains, and is: how do we, in either case, calculate Z? Is V/I = Z using the difinition above?? Or not? How do I use the current density to calculate Z? and the Phase? Is there a way of calculating Z without defining I?
  5. Apr 9, 2009 #4


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    If you have set up your model correctly V/I should give you the impedance.

    How have you defined the potential difference?
  6. Apr 9, 2009 #5
    From the boundary settings dialogue box, I grounded the base of the disc and applied 1V to the top. The rest of boundaries are set at "zero charge/symmetry"
    I am using the "frequency response" analysis, and I am varying the frequency from 3 to 4.5 MHz in steps of 0.02MHz (I expect the disc to resonate at around 3.8MHz since it is PZT 0.6 mm thick. I defined I at the top electrode as I=-imag(nJ_smpz3d)

    Should I define I=nJ_smpz3d, rather than the imaginary part?
    Are you sure Z=V/I, i.e., V/(-imag(nJ_smpz3d))? Do you know how do I find the phase?

    Looking at the plot of the results I obtain, there is a resonance at around the right frequency (plus another one at 4.2MHz I don't know where it comes from...), but somehow does not look totally right: Z is flat and near 0 at frequencies lower that the resonance. Then I assume that my definition of I is giving me information about the resonance but Z is not well calculated...

  7. Dec 10, 2009 #6
    if your imedance is zero at resonance, then it means you have not defined damping in your model. By defining damping you lower the Q *quality factor) and hence will provide you with the right resistance (impedance).

    Good luck
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Similar Threads - COMSOL frequency response Date
Model Heat Transfer with COMSOL Aug 26, 2016
Matlab Filtering Frequencies Apr 27, 2016
Parameteric Sweep in comsol Apr 9, 2015
Help, comsol won't change my mesh size Feb 8, 2015
Can I prevent setting up standing waves in my COMSOL model? Aug 12, 2014