Menu

Discussion Forum

New Discussion
Filter

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.

Time-domain Study of a Piezoelectric Actuator

Posted May 18, 2017, 8:17 a.m. EDT 5 Replies

Suketu Naik
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Hello all, I have gone through multiple threads on time-domain studies for piezoelecteric transducers and followed all the help but have run into problems. I am trying to apply a sinusoidal voltage to a piezoelectric beam actuator with a frequency that coincides with the largest displacement that I found in eigen-mode and frequency-domain analyses. To best of my knowledge, all the boundary conditions in Solid Mechanics and potentials in the Electrostatics are correct (I created a sinusoid waveform and used its label as Sinu(t) under Electric Potential as suggested by this thread: www.comsol.com/community/forums/general/thread/128071/)
I have tried every thing but my plot (point probe on the free end or the tip of the beam) shows jagged waves and does not follow the input voltage shape (sinusoid). Am I setting that up incorrectly?
I tried restricting the steps taken by the solver but that didn't help. I have attached the excitation voltage waveform, the output, and the time domain solver setup (Comsol forum won't let me post my model). Can someone please take a look and let me know what I am doing wrong?
Thank you very much for your help.
5 Replies Last Post Nov 6, 2017, 7:12 a.m. EST
Edgar J. Kaiser Certified Consultant
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 8 years ago May 18, 2017, 10:36 a.m. EDT
Hi,

the 2 ms timestep is not sufficient for your waveform. You should use some ten steps per period. You also didn't check the checkboxes in the time stepping.
If you excite with a continuous harmonic wave and all is linear in your model you can use the frequency domain solver which is much more effective in that case.

Cheers Edgar



--
Edgar J. Kaiser
emPhys Physical Technology
www.emphys.com
Hi, the 2 ms timestep is not sufficient for your waveform. You should use some ten steps per period. You also didn't check the checkboxes in the time stepping. If you excite with a continuous harmonic wave and all is linear in your model you can use the frequency domain solver which is much more effective in that case. Cheers Edgar -- Edgar J. Kaiser emPhys Physical Technology http://www.emphys.com
Suketu Naik
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 8 years ago May 18, 2017, 11:15 a.m. EDT
Thank you Edgar, I would like to see the actuator response based on geometric nonlinearity hence I only used freq-domain to perform a check on the eigen-mode analysis.
I tried changing the time-step to 0.1 ms (75 steps per period) but I see the same output. Looking at this response, it seems like there is a numerical integration issue (I tried different solvers to no avail)? Do you suppose it will help if I somehow stored the previous solution (I don't know if there is an option to do that somewhere or if it is doing that already). Another option would be to run stationary solver, store that value as the initial value and use it to perform time-domain study.
How can I resolve this issue?
Thank you again for your help.
Thank you Edgar, I would like to see the actuator response based on geometric nonlinearity hence I only used freq-domain to perform a check on the eigen-mode analysis. I tried changing the time-step to 0.1 ms (75 steps per period) but I see the same output. Looking at this response, it seems like there is a numerical integration issue (I tried different solvers to no avail)? Do you suppose it will help if I somehow stored the previous solution (I don't know if there is an option to do that somewhere or if it is doing that already). Another option would be to run stationary solver, store that value as the initial value and use it to perform time-domain study. How can I resolve this issue? Thank you again for your help.
Suketu Naik
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 8 years ago May 18, 2017, 11:50 a.m. EDT
Ok, so I resolved this issue.
I took the following steps:
1) changed the time-step and made is very small (e.g. 100 steps per period)
2) plotted Z-displacement and not the total displacement (assuming the actuation will happen in z-dir).
3) included step1: stationary. step2: time dependent.
4) clicked on step2: entered range(0,0.05,37.5) ms. For my simulation, this creates 5 periods with small step size.
5) In step2, I made sure that the time-step is restricted. (Solver Configurations->Time-Dependent Solver1: panel on right, expand TimeStepping->Initial step=0.001ms, Max Step=0.05ms, which matched
with my step size.

I got a very nice sinusoidal displacement.
Hope this helps someone else (sorry comsol forum wouldn't let me post my files for some reason).
Thank you.
Ok, so I resolved this issue. I took the following steps: 1) changed the time-step and made is very small (e.g. 100 steps per period) 2) plotted Z-displacement and not the total displacement (assuming the actuation will happen in z-dir). 3) included step1: stationary. step2: time dependent. 4) clicked on step2: entered range(0,0.05,37.5) ms. For my simulation, this creates 5 periods with small step size. 5) In step2, I made sure that the time-step is restricted. (Solver Configurations->Time-Dependent Solver1: panel on right, expand TimeStepping->Initial step=0.001ms, Max Step=0.05ms, which matched with my step size. I got a very nice sinusoidal displacement. Hope this helps someone else (sorry comsol forum wouldn't let me post my files for some reason). Thank you.
Suketu Naik
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 8 years ago May 18, 2017, 11:54 a.m. EDT
Sorry I forgot to mention step 6.
6) Went back to Step2: time dependent panel (main panel) and set up the following:
Values of Dependent Variables->Intial values of variables solved for->Settings: User Controlled, Method: Solution, Study: <Study name>, Stationary
Values of variables not solved for->Settings: Physical Controlled.
I didn't change anything in the stationary.
Sorry I forgot to mention step 6. 6) Went back to Step2: time dependent panel (main panel) and set up the following: Values of Dependent Variables->Intial values of variables solved for->Settings: User Controlled, Method: Solution, Study: , Stationary Values of variables not solved for->Settings: Physical Controlled. I didn't change anything in the stationary.
Frank Turner
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 7 years ago Nov 6, 2017, 7:12 a.m. EST

Hi, Suketu,

I am trying to build a model of the piezoelecteric transducer, too. My research is about the acoustic emission signals generated by a sphere particle impacting onto a plate and I need a piezoelectric transducer to aquire the acoustic emission signals. I think the part of piezoelectric transducer simulation is similar with yours. May I ask how to achieve this part with your model in details? I have read your comments on your simulation but I just could not get the points. Can you tell me how to set the bound conditions about the piezoelectric? Thank you in advance.

Frank

Hi, Suketu, I am trying to build a model of the piezoelecteric transducer, too. My research is about the acoustic emission signals generated by a sphere particle impacting onto a plate and I need a piezoelectric transducer to aquire the acoustic emission signals. I think the part of piezoelectric transducer simulation is similar with yours. May I ask how to achieve this part with your model in details? I have read your comments on your simulation but I just could not get the points. Can you tell me how to set the bound conditions about the piezoelectric? Thank you in advance. Frank

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