Dry Fiber Orientation Definition

[eswe004]

Hi,

I am conducting unit cell compaction simulations in Abaqus using Abaqus scripting functions.
In the script, when I suppress the orientation definitions and model the material as an isotropic , I get a realistic deformation (as shown below).
However, when I leave the orientation definitions as they were and specify the material's engineering constants as E1=E2=E3=E(isotropic), etc. (essentially still isotropic material), then the resulting deformation clearly has problems (as shown). Would you know why this has happened? Could this be due to a problem with the .ori file?

Thanks in advance for your assistance,
Elinor

[louisepb]

Hi Elinor,

I haven't seen anything like this before. We haven't had any problems with material orientations - particularly for a simple weave like this. Have you set up your shear moduli and Poisson's ratios correctly as well as E1, E2 etc? Also it looks like the boundary conditions might not be set up correctly - they don't appear to have been set up to be periodic.

I'm afraid that's all I can come up with at the moment.
Best regards,
Louise

[eswe004]

Hi,

I have checked the material properties and they seem to be realistic.
Perhaps it could be an issue with the periodic boundary conditions (or lack of). When I import the .inp model that was produced in TexGen, I can't seem to find any reference to periodic boundary conditions that were imposed. I can see that corresponding node sets were defined (Bound0A / Bound0B etc.) but can't see if these were used anywhere. Is that the case? Or is there a problem with the way that I have exported the TexGen model?

Thanks in advance,
Elinor

[louisepb]

Hi Elinor,

In the dry fibre export there is just a simple set of equations which specify that deformations on opposite boundaries match. Also the central node in each yarn is constrained to prevent rigid body motion. You will probably need to specify boundary conditions yourself to make this work for your particular simulation.

Also just to check: you are running from the command line, not importing the model into Abaqus CAE? If you just load into CAE all the orientations etc will be lost.

Best regards,
Louise