iMechanica - any idea about boundary conditions to be used when modelling a unit cell of a foam - Comments

Re: Abaqus, importing from Pro-E, and foam models

Biswajit Banerjee — Wed, 18 Jun 2008 23:07:24 -0400

Hi Prasanna,

I'm afraid that I don't know enough about Abaqus to say anything useful here. But there are quite a few others on this forum and on the Abaqus Yahoo group who use the software and will be able to help you. My own preference would be to just use Abaqus directly for creating the geometry because the foam geometry is relatively simple.

-- Biswajit

About boundary conditions

pthiyaga — Tue, 17 Jun 2008 12:35:26 -0400

Thanks a lot biswajit... I indeed found the info in 'wikipedia' very informative...

Now I have a trivial modelling question...Please dont mind my asking it...

I modelled one unit cell from the data I had in Pro-e and I tried to import it into abaqus.. I wanted it to be a wire because I wanted to model every strut as a beam (open-foam)....

However abaqus imports it as one single object... That is, all the wireframe edges are imported as one single object. Do you know if there is any way I can import this as a bunch of lines.

Thanks a tonne.

-Prasanna

Re: Unit cell of a foam

Biswajit Banerjee — Tue, 17 Jun 2008 00:33:22 -0400

Dear Prasanna,

I'm sure you've already looked at Gibson and Ashby for some pointers. That's of course the first place to go to for information on simple foam models.

The periodicity of tetrakaidodecahedra (more commonly called tetradecahedra - or 14 sided) is a fascinating subject. After seeing you post I looked at Wikipedia and there's a wealth of information there. I hadn't realized that there was a complete list of possible forms of 14-faced objects. The most familiar one (to me) is the the Kelvin cell or Truncated octahedron.

This element can be used to tile space leading to the bitruncated cubic honeycomb. The unit cell for that tiling is a subset of the image below

This is the repeating pattern that you're looking for. From this figure you should be able to figure out which faces and corners need to be assigned periodicity. It won't be as simple as a hex or a cube but it'll be much more fun. For a start you can just take the first 9 cells (any corner region in the figure) and apply standard periodic boundary conditions to the edges of the cube in which they are contained.

-- Biswajit

any idea about boundary conditions to be used when modelling a unit cell of a foam

pthiyaga — Mon, 16 Jun 2008 12:53:38 -0400

Dear friends / distinguished imechanicians,

I am trying to use micromechanics in foams to numerically compute the elastic constants using one unit cell. I see some literature available when simple unit cell shapes are assumed (like cubical or hexagonal). However there is nothing specific about modelling for tetrakaidecahedral foams. Assuming the right boundary conditions would be critical for computing the constants. I am not able to come up with a reapeating pattern for determining the places to apply the boundary conditions.

Have any of you done something similar to this and if yes, would you want to throw some light on the same.

Thanking you

Prasanna