ABAQUS dry fabric simulation

[eswe004]

Hi,

I am hoping to simulate the compaction of a textile between two compression plates.
Initially I have been doing this by exporting an Abaqus dry fibre file with compression plates and setting up / running the simulation in Abaqus.

I see that this can all be done from within TexGen (with the capabilities of showing the deformed geometry?) however I haven't managed to get this to work. I have used the sample script on http://texgen.sourceforge.net/index.php ... oldid=3104 and it all ran fine, but didnt produce any deformation. The following warning was displayed:
<Warning: No nodal displacements found while parsing temp.dat.>

I think this is because no compression plates were actually generated before the simulation was executed. Would you be able to provide me with information on how to define this? (or do you happen to have a script which already does this that I can use?).

Thanks,
Elinor

[louisepb]

Hi Elinor,

If you use the script exactly as it is on the web page it will perform the simulation with a tension in the x direction.

If you want to produce a compaction with compression plates you will need to change the line "tension.AddScale(1.1, 0, 0)" to give a deformation in the z, eg tension.AddScale(0, 0, 0.9)

You will also need to add a line to add the compression plates: deformer.SetIncludePlates( True )

Hopefully that should work for you,

Best regards,
Louise

[eswe004]

Hi Louise,

Thanks for the assistance.
I changed the linear Transformation scale to what you suggested, and added the compression plates. For some reason though, when I execute the simulation through TexGen (using the standard material definition), I get the following "RuntimeError: ABAQUS has failed with error code: 1".

In the .sta file, it says that the error is due to excessive distortion of elements, and I can see this when looking at the simulation results. However, if I simply execute the .inp file that was generated (through the Abaqus GUI) it produces a realistic deformation.

Any suggestions as to what went wrong?

Cheers,
Elinor

[louisepb]

Hi Elinor,

I think that the problem may be caused by an error in the script which I didn't register, which I then continued in my last answer. For axes with no tension/compression the scale should be 1.0 (ie stays the same length). I tried running a version with scale(1.1, 0, 0) which gave very strange results. So, you need a scale of something like (1.0, 1.0, 0.9) - or whatever you want your z deformation to be. I did a run with this input and still got the warning but there seemed to be deformations when I loaded the tg3 file produced.

Let me know how you get on. This process was developed by my predecessor so it's not something which I've used (and checked is still working!).

Louise

[eswe004]

Thanks, it now works.

However I can't seem to see the deformed geometry in TexGen. I use the following to save the deformed geometry:
SaveToXML('pweave-tensioned.tg3')
But this seems to be the original undeformed geometry (just with a different domain defined).
How did you save the deformed tg3 file?

I know that it is possible to get the deformed geometry from the Abaqus results (linking up the nodal displacements with the original node definition, but I would rather stay away from that if there is a simpler solution).

Elinor

[louisepb]

Hi Elinor,

I've taken another look at the tg3 file and realise that what I'd taken to be a deformation was simply the rendering trimming the textile to the domain. The calls to Abaqus and the reconstruction of the deformed TexGen model are in the Abaqus.py script included in the TexGen installation. I will have to look into why this isn't finding the nodal displacements. I suspect that, as this was written some time ago, the abaqus output file format has changed so that the script no longer reads the file correctly. I'm currently trying to get a new release of TexGen out so I won't be able to investigate straight away but I'll try and get to it as soon as I can.

Louise

[maryam_sodagar]

Dear Louise

Hi
Me again!
My model looks better now. I would appreciate you if you or anybody else can answer these critical questions for me:
1.What does x y z scale exactly mean? I know for example when we want to achieve a unixial load in x direction, we should put it (1.1, 1,1) which means no deformation in y and z direction. what is the difference when we put it (1.2, 1, 1)? not only + 0.1 but also +0.2...
2.what are these displacement/rotation boundary conditions which are exporting to abaqus? If I suppress them what will happen?

[louisepb]

Hi Maryam,
The scale will specify the amount of deformation that you require.
Which are the displacement/rotation boundary conditions you are refering to?

Best regards,
Louise

[maryam_sodagar]

Thank you so much. I am near my defense time but still I have some problems.

(The scale will specify the amount of deformation that you require) - I understand what you mean. When we apply the amount of deformation so it is our input right? and what would be the output? It should be the Stress/strain/modulus...?
(Which are the displacement/rotation boundary conditions you are refering to? ) - I mean the displacement/rotation BC which you can see in the image. They have been automatically exported to abaqus. for uniaxial loading, if we suppress them or not what will happen?

11.jpg (86.55 KiB) Viewed 14673 times

[louisepb]

Hi Maryam,

The nodes closest to the centre of the yarns are constrained to prevent rigid body motion. The deformations are also applied at these points.

Best regards,
Louise

[maryam_sodagar]

Hi Louise

I have a basic question again!
X, Y and Z scales are the same as Displacement/Rotation BC in abaqus? are they the same as Tensile Strain(%) ?

[louisepb]

Hi Maryam,
The scale specifies the deformation (ie strain) that you are applying to the model.

Best regards,
Louise