Discussion Forum

Your model looks good to me, you just need to display an Impedance plot in the results.

-Create a new 1D Global Plot for your frequency sweep (solution 2) (you could do a line or point graph too, same results)
-plot freq vs. 1/abs(pzd.Y11). (usually abs(pzd.Z11) works, there's something funny there since that has no units)

Hello Marc,
Thanks for your reply. I already did it. But it does not seems to me correct. Because there are nothing change in the impedance(1D plot ). If you have time, would you please look at it again. Please see the attached the model.

Can anybody help me regarding impedance analysis of piezo(ID plot)?


Regards
Bhuyan

Bhuyan,

I agree there's something wrong- your piezo is behaving as a capacitor because there's no displacement in the entire model when the frequency domain study is run. Maybe 2D for piezo doesn't make sense because there is no depth? If you apply a prescribed displacement to the top of the Ti model you get displacement but the piezo impedance is still behaving like it's stationary.

I've built 2D axisymmetric and 3D models the same way you did and they turn out fine. Sorry I can't help further, try Comsol support they're great.

Marc

Hello Marc,
Thanks for your encouraging reply. I am also agree with your comments for 2D Piezo.
I have also built for 2D axial symmetric and 3D and they are working. But did you find the impedance of piezo(1D plot) for 3D or 2D axial? If so, would you please tell me the procedure?

Thanks
D Bhuyan

After thinking about this, Comsol is perfectly capable of solving the 2D problem.

The trick to remember is the material properties for the default piezoelectric materials in the comsol library are set up so that the "3" dimension (normally Z) is in the direction of poling. For these 2D models, the default setting is to have the "3" direction be out-of-plane...not too helpful. Let's change this...

1) Define a Base Vector system for the material with "2" as the out-of-plane index.
2) Base vectors x1 = x, x3 = y. (this is a right-handed orthonormal coord system, but check "Assume orthonormal" just for good measure.
3) Open your Piezoelectric Material Model. Change the Coordinate System Selection to your new Base Vector System.
4) solve away!

I can't seem to attach files right now, but it's not a plain capacitor anymore...

As for your question about plotting Impedance (vs. frequency, I'd assume), same procedure as my first reply post: create a global plot of 1/abs(pzd.Y11) vs freq . I like to define a probe of this plot before I solve so that it updates while solving. Note that if you don't use abs( ) and don't have damping in your model, the result is imaginary and your plot shows nothing!

Regards,

Marc

Dear Marc,

Thanks a lot........I think you are very good in COMSOL. But I am a new user in comsol multiphysics and trying to improve myself. I already did the simulation model for impedance anlysis according your instructions. It's working fine. But how can i do it for time dependent anylysis?

For your consideration, I am also trying to see the simulation results for time dependent analysis (for example range 0 to 5ms, step size 1us). When I compute the time dependent study, it is working but the pressure or displacement in the Ti does not changing with respect time(for 1D plot). I think there are some errors inside the model. But I can't notice the error. If you have time can you please look at my model. I have attached the model.


Regards
D Bhuyan

Dear Marc,

Thanks a lot........I already did the simulation model for impedance analysis. It's working fine. But how can I do it for time dependent analysis?
I am more interested to see the time dependance behavior of this piezo. Can you please give me some idea?
I am trying to see the simulation results for time dependent analysis (for example range 0 to 5ms, step size 1us). When I compute the time dependent study, it is working but the pressure in the Ti does not changing with respect time(for 1D plot). The simulation result is not expected for me.

Regards
D Bhuyan