Elmer Discussion Forum

HI

I have been unsuccessfully trying to input orthotropic 2D thermal conductivities into the heat equation model.

I thought this was supported in Elmer based on the discussion of Model 1 - Heat Equation where it says that "For solids, conduction may be anisotropic and the conductivity a tensor"

I followed the Tutorial 10 approach for an anisotropic Youngs modulus on page 50 to no avail. A result from one of my many attempts is given below:

Excerpt from the SIF file:

Material 1
Name = "solid"

Density = 1550
Heat Capacity = 1450
Heat Conductivity
size 2 2
real 0.1 0.6
0.6 0.1
end
End


Elmer tells me:

Error:: Model Input:
Error:: Model Input: Unknown specifier: [end]
Error:: Model Input: In Section: [material 1]
Error:: Model Input: For propert name: [0.6 0.1]

Is it possible to have solve a solids heat conduction problem with anisotropic thermal conductivity using Elmer?
If so, can you please explain how to do it?

thanks

gsw

I would also be very interested in the answer.

Hi

I think that the expression for heat conductivity is missing a line feed between the two rows. In fact you can have all the entries on the same row, the size decleration alone decleares the size, not the layout of the numbers. I would try with the following:
or as the variabletype is known a priori simply
-Peter

Hi,
According to the above view, I think

(a) If I have a 2D model, kx = 1.0 & ky = 0.0, so
Heat conductivity(2,2) = Real 1.0 0.0 0.0 0.0

(b) Same as (a), kx = 0.0 & ky = 1.0, so
Heat conductivity(2,2) = Real 0.0 0.0 0.0 1.0

(c) If I have a 3D model, kx = 0, ky = 0 & kz=1.0, so
Heat conductitvity(3,3) = Real 0.0 0.0 0.0 0.0 0.0 1.0

Is that correct?
Please help me to confirm.

BR,
Taco.

Yep, except in 3D you need 9 components, not six. -Peter

Hi, Peter,

Thanks a lot.

Taco

Hi,
Seeing the much relevance of anisotropic heat conduction in solids, I am trying to refresh this discussion started 7 years ago.
2D Model
kx = 0, ky = 1.0, so 3D Model
kx =1. 0, ky = 5.0 & kz=1.50, so Yours Sincerely,
Anil Kunwar

Hi All,
Is it possible to vary the anisotropic material property with coordinates in Elmer?
Relevant threads:
1. viewtopic.php?f=3&t=3139&start=0&hilit=tensor+variable
2. viewtopic.php?f=7&t=2267&start=0&hilit=user+subroutine

Variation of a tensor with Coordinate
1.viewtopic.php?f=3&t=3802&start=0&hilit=tensor+matc
2. viewtopic.php?f=3&t=110 (length limit of matc)


Yours Sincerely,
Anil Kunwar