Menu

Discussion Forum

New Discussion
Filtern

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 The information provided may be out of date.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.

building piezo ring projector (electroacoustic transducer)

Posted 02.11.2012, 16:13 GMT-4 Low-Frequency Electromagnetics, MEMS & Nanotechnology, Piezoelectric Devices, Acoustics & Vibrations, Structural Mechanics Version 4.2a, Version 4.3, Version 4.3a 7 Replies

Erhan Ozer
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Hello,

I plan to build piezo ring projector, but I don't decide to build it as 2d axis symmetry or entire 3d model. My electrodes would be located circumferentially. 2d model would be problem because of the electrodes locations.

Thank for your answer.

Erhan
7 Replies Last Post 02.05.2013, 11:06 GMT-4
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 03.11.2012, 07:31 GMT-4
Hi

I do agree that if you have electrodes non symmetrically aroud the circumference, you are obliged to go to 3D which is far heavier to solve than a nice 2D-axi ;)

--
Good luck
Ivar
Hi I do agree that if you have electrodes non symmetrically aroud the circumference, you are obliged to go to 3D which is far heavier to solve than a nice 2D-axi ;) -- Good luck Ivar
Erhan Ozer
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 20.11.2012, 05:26 GMT-5
Thank you Ivar for your response,

How can I create a circumferential polarization for my model which provide to reduce computation load. I sent you a picture of polarization, maybe you have any idea. It would be possible or not, I can't decide on this model?


thanks.

Thank you Ivar for your response, How can I create a circumferential polarization for my model which provide to reduce computation load. I sent you a picture of polarization, maybe you have any idea. It would be possible or not, I can't decide on this model? thanks.

Attachments:

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 20.11.2012, 07:50 GMT-5
Hi

I would say asplit up your PZT into pie slices, or rectangles + pie slices, then define a series of rotated coordinates (or you could try a cylindrical, but that I have never used not sure it works for anisotropic material definitions, should be carefulla verified) and define each slice accordingly. It's quite some knitting, perhaps easier as external functions or via matlab or Java

--
Good luck
Ivar
Hi I would say asplit up your PZT into pie slices, or rectangles + pie slices, then define a series of rotated coordinates (or you could try a cylindrical, but that I have never used not sure it works for anisotropic material definitions, should be carefulla verified) and define each slice accordingly. It's quite some knitting, perhaps easier as external functions or via matlab or Java -- Good luck Ivar
Erhan Ozer
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 20.11.2012, 20:07 GMT-5
Thank you Ivar,

I have same Idea in my mind,but I thought never about external function.

Firstly, you mean with external function, generate a function with live link to MATLAB for changing coordinate system?

Secondly, There would be 72 pieces in Total, so it would be make difficult to solve model. Maybe It can't be possible.

Is there any example related with circumferentially polarized ring?

Erhan
Thank you Ivar, I have same Idea in my mind,but I thought never about external function. Firstly, you mean with external function, generate a function with live link to MATLAB for changing coordinate system? Secondly, There would be 72 pieces in Total, so it would be make difficult to solve model. Maybe It can't be possible. Is there any example related with circumferentially polarized ring? Erhan
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 21.11.2012, 01:19 GMT-5
Hi
I do not know of any circular polarized items, neither as HW nor as model, but I do certainly not know all out there. Probably one can use a cylindrical coordinate, but one must just validate it carefully, something slightly tricky for such tensors, and the particular notation ;)

--
Good luck
Ivar
Hi I do not know of any circular polarized items, neither as HW nor as model, but I do certainly not know all out there. Probably one can use a cylindrical coordinate, but one must just validate it carefully, something slightly tricky for such tensors, and the particular notation ;) -- Good luck Ivar
Erhan Ozer
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 21.11.2012, 01:21 GMT-5
Ok,

thank you.
Ok, thank you.
RVN
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 02.05.2013, 11:06 GMT-4

Is there any example related with circumferentially polarized ring?

Erhan



Hi!

You need to redefine the basis vector system since the Comsol default is a cartesian with the polarization in x1 (if I remember correctly). It sounds like you need one that has the polarization in radial direction (cylindrical).

Right click Definitions -> Coordinate systems-> Base vector system, and start from there. Define your x1, x2,x3 as follows:

e_{rho}= e_x* cos(phi)+e_y*sin(phi)
e_{phi} = -e_x*sin(phi +e_y*cos(phi)
e_z = e_z

where e_x, e_y and e_z are the unit vectors and phi can be expressed as phi= atan(y/x).

Now just take care to place the e_z as best suits your model (where you want your polirazation to point).

A similar model you can take a look at, is the Tutorial model "piezoacoustic_spherical" where the same tactic is employed with a spherical shell piezo so that the polarization is always in radial direction.

Hope this helps
--
/RVN
[QUOTE] Is there any example related with circumferentially polarized ring? Erhan [/QUOTE] Hi! You need to redefine the basis vector system since the Comsol default is a cartesian with the polarization in x1 (if I remember correctly). It sounds like you need one that has the polarization in radial direction (cylindrical). Right click Definitions -> Coordinate systems-> Base vector system, and start from there. Define your x1, x2,x3 as follows: e_{rho}= e_x* cos(phi)+e_y*sin(phi) e_{phi} = -e_x*sin(phi +e_y*cos(phi) e_z = e_z where e_x, e_y and e_z are the unit vectors and phi can be expressed as phi= atan(y/x). Now just take care to place the e_z as best suits your model (where you want your polirazation to point). A similar model you can take a look at, is the Tutorial model "piezoacoustic_spherical" where the same tactic is employed with a spherical shell piezo so that the polarization is always in radial direction. Hope this helps -- /RVN

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