modeling of a piezoelectric composite plate 1

[alinalin]

Use ANSYS to analyze a piezoelectric composite plate:

Is there any way of inputing the material matrices of piezoelectric composite before transformation and then add an angle? I searched for the elements of
the structural layered composite, and they don't have the DOF of voltage-that's why I can not use this kind of elements. So, the only way is to input the material matrices (elastic, piezoelectric and dielectric matricex) after transformation?

Because I need to keep the DOFs of three displacements and voltage, I can only choose SOLID5 as the element type. It causes a problem that I can not mesh a thin plate with small size elements-the thickness of the plate can not be
divied up to the requirement of accuracy. How to solve this problem?

Appreciate your reply!

 

[ansysfreak]

Hi

It looks like you want to do an Coupled-Field analyse.
So you need to take an look at that.

Gtrz
Garry

 

[alinalin]

Thank you! I've read through the part of coupled-field analysis, but I can not solve my problem.

 

[ansysfreak]

Hi

You want to model piezoelectric composite, and look at the displacemants what are caused by Volt?

You can make this work by using an struc-elek coupled-field analyse with an shell element.

Grtz
Garry

 

[alinalin]

Thanks! I have to use 3-D elements because I need to calculate three displacements and voltage. Then, I got too many DOFs for a thin plate and the thickness can not be discritized by enough nodes.

 

[GeorgeEllwood]

I'm not quite sure if I understand the question but you can use SOLID98 elements and change one of the keyopts to give you ux,uy,uz and volts DOFs only.

 

[ansysfreak]

Hi

The solid98 is the high order elemnt of solid5. And it will not help him. The best thing for him to do is to use a shell element( because of the thin elements).

Grtz
Garry

 

[alinalin]

Thanks for the reply! I need to have the DOFs: Ux, Uy, Uz, Volt. And the version of ANSYS I used can not handle more than 256,000 nodes. For the thin piezoelectric composite plate which is made of several layers of piezoelectricity with different layer-up angles, the simplest element I can use is SOLID5. I read through the elements for modeling layered composites(SHELL99, 91, 181 etc.). They don't have have DOF of voltage.

 

[ansysfreak]

Hi

There are two ways of doing it:
-using solid element 5/98, with this element you can do an direct coupled analysis, but when your volume is very thin you maybe get bad results. Also it is easier to model.
-using shell elements, with this element you can model thin elements. Disadvantage you need to do and couple-field analysis. Its also difficult to model the layers. Advantage is the amongt of elements which can be very low. So you decide what you want to do!

Grtz
Garry

 

[alexos1975]

Hello!

I got an ANSYS error message, which consequence is still unclear to me:

Specified constraint not allowd on slave DOF in coupled set. Coupled set No.1 First node=2 Slave node= xxxx DOF=volt
Move constraints to first node or remove coupling and constrain all.

Does anyone can give me a hint, where to find some more information concerning this error message? Or can anyone explain me the backgrounds of this error?

Thanks

Alex

 

[alexos1975]

Sorry....that was the wrong placement of my message..

alex

 

[alinalin]

ansysfreak, thanks for the post!
I used the direct method to model the thin structure and can not get good results. I don't quite understand what you said about the second method (sequential method). I only apply electric potential to the plate. How can I model with shell elements without the DOF of volt? To do a piezoelectric analysis, the manual have specialized six kinds of elements:
plane13/223, solid5/98, solid226/227.