Anisotropic materials

General discussion about Elmer
Post Reply
AlexR
Posts: 12
Joined: 06 Apr 2014, 16:14
Antispam: Yes

Anisotropic materials

Post by AlexR »

Hello,

Does Elmer allows to define anisotropic materials ?

I didn't saw any option to define several Youngs modulus and Poisson ratio (which is important in anisotropic materials definition).

If somebody has any information, please let me know ;)
annier
Posts: 1168
Joined: 27 Aug 2013, 13:51
Antispam: Yes

Re: Anisotropic materials

Post by annier »

Hi AlexR,
Yes Elmer allows you to define anisotropic materials.
It is specified in the linear elasticity solver of Elmer Models Manual as following:
ElmerModelsManual wrote:For anisotropic materials, the stress-strain relations may be given in somewhat different form:
τV = EεV ,
...
The matrix E is given as input for the anisotropic material model of Elmer.
The keywords related to anisotropy are:

Code: Select all

Rotate Elasticity Tensor = Logical True
and
Material Coordinates Unit Vector i (3) , i =1,2,3
These are defined in the material section of the solver input file (SIF).
The Youngs Modulus/Modulii are defined as scalar for isotropic materials whereas matrix for anisotropic material.
Yours
Anil Kunwar
Anil Kunwar
Faculty of Mechanical Engineering, Silesian University of Technology, Gliwice
AlexR
Posts: 12
Joined: 06 Apr 2014, 16:14
Antispam: Yes

Re: Anisotropic materials

Post by AlexR »

Thank you a lot for this answer. I'll see it and make a feedback :)
AlexR
Posts: 12
Joined: 06 Apr 2014, 16:14
Antispam: Yes

Re: Anisotropic materials

Post by AlexR »

Hello,

I inserted the following code in the material section of the .sif file.

Code: Select all

Material 1
     Name = "Mat1"
     Density = 7500
     Poisson Ratio = 0.30
     Youngs Modulus = 200e9
     Rotate Elasticity Tensor = Logical True
     Material Coordinates Unit Vector 1(3) Real [0.7170 0.7170 0]
     Material Coordinates Unit Vector 2(3) Real [-0.7170 0.7170 0]
     Material Coordinates Unit Vector 3(3) Real [0 0 1]
End
You can have good results, but the problem is the following : after using the solver, these code lines are erased.

I think it comes from the fact that I use Elmer GUI to load the project and then launch the solver.

- Do you think it is a non-GUI problem ? (it seems like the Tutorial 13 in the Elmer Tutorials Manual)
- In this way, how to work using Elmer GUI ?
mzenker
Posts: 1999
Joined: 07 Dec 2009, 11:49
Location: Germany

Re: Anisotropic materials

Post by mzenker »

Hi,

ElmerGUI overwrites the sif file each time you save your project with the values you have entered. All changes you make to the sif file are lost. But you can add a new material with Model->Material->Add, and use the free text field for properties you don't find a field for.

HTH,

Matthias
annier
Posts: 1168
Joined: 27 Aug 2013, 13:51
Antispam: Yes

Re: Anisotropic materials

Post by annier »

Hi AlexR,
Yes, if you use Elmer nonGUI, it is quite easy to modify the things in SIF.
Since ElmerGUI has some limitations, it is sometimes like
"All ElmerGUI tutorials can be run with ElmernonGUI method, but all ElmernonGUI tutorials cannot be run using the ElmerGUI so far now".
You can give ElmernonGUI a try.
Yours
Anil Kunwar
Anil Kunwar
Faculty of Mechanical Engineering, Silesian University of Technology, Gliwice
AlexR
Posts: 12
Joined: 06 Apr 2014, 16:14
Antispam: Yes

Re: Anisotropic materials

Post by AlexR »

Thank you mzenker and annier for your reply.

I saw in the material section (Model ->Material->Add...-> Linear Elasticity) that you can define the Youngs Modulus matrix and/or rotate the elasticity tensor.

The Youngs Modulus definition seems correct using Elmer GUI, but I saw a mistake in the elasticity tensor rotation.

For example you want to make a 45° rotation across the z axis (3-axis in Elmer). The standard code would be

Code: Select all

Material 1
     Name = "Mat1"
     Density = 7500
     Poisson Ratio = 0.30
     Youngs Modulus = 200e9
     Rotate Elasticity Tensor = Logical True
     Material Coordinates Unit Vector 1(3) Real [0.7170 0.7170 0]
     Material Coordinates Unit Vector 2(3) Real [-0.7170 0.7170 0]
     Material Coordinates Unit Vector 3(3) Real [0 0 1]
End
But Elmer generates the .sif file with this code

Code: Select all

Material 1
     Name = "Mat1"
     Density = 7500
     Poisson Ratio = 0.30
     Youngs Modulus = 200e9
     Material Coordinates Unit Vector 1(3) Real [0.7170 0.7170 0]
     Rotate Elasticity Tensor = Logical True
     Material Coordinates Unit Vector 2(3) Real [-0.7170 0.7170 0]
     Material Coordinates Unit Vector 3(3) Real [0 0 1]
End
It involves that results are false.I don't understand why the sif generation does this error...
Do you have an idea about this ?

Yours.
Alex
mzenker
Posts: 1999
Joined: 07 Dec 2009, 11:49
Location: Germany

Re: Anisotropic materials

Post by mzenker »

Hmmm, are we sure that order matters in this case?
If you open the sif file in the GUI (Sif->Edit), correct the order in the material declaration, save the sif and run the solver, does that change the results?

Matthias
AlexR
Posts: 12
Joined: 06 Apr 2014, 16:14
Antispam: Yes

Re: Anisotropic materials

Post by AlexR »

Hi !
I come back to this subject because the matrix definition is the following.
You have to insert these lines in the Material definition into the .sif file :

Code: Select all

Youngs Modulus (3,3) =
    10 10 0
    10 10 0
    0 0 10
(make the same thing using a 6x6 matrix if you have a 3D model).

But it stays another question :
What are data in the matrix ? Does the matrix defines the rigidity matrix or anything else ?
raback
Site Admin
Posts: 4832
Joined: 22 Aug 2009, 11:57
Antispam: Yes
Location: Espoo, Finland
Contact:

Re: Anisotropic materials

Post by raback »

Hi

Just a note on the syntax. If you want to rotate the material then this could work:

Code: Select all

     Rotate Elasticity Tensor = Logical True
     Material Coordinates Unit Vector 1(3) = Real 0.7170 0.7170 0
     Material Coordinates Unit Vector 2(3) = Real -0.7170 0.7170 0
     Material Coordinates Unit Vector 3(3) = Real 0 0 1
The order of keywords has no difference except if the same keyword appears several times.

-Peter
Post Reply