Non Crimp Fabric

[louisepb]

Hi Jost,

I'm not entirely sure what the problem is here. Does the original .inp file created by TexGen run ok?

In what way are you editing in Abaqus? If you alter the elements in the mesh in any way then the orientation file won't match any longer.

You say that you created an assembly. The implication of the error message is that it only applies to a model - could this be the problem?

Rather more questions than answers but I hope they help you sort out what the problem is!

Louise

[David]

Hi Jost,

I've not come across the problem you mention, probably because I've been exploring another, more promising avenue here. I read the documentation (and a few papers by the developers) in more detail and found that the researchers at Nottingham used Python scripts to load models into Abaqus/CAE. This is in contrast to Importing models using File > Import... which, as we've all found, causes .ori and .eld headaches.

As per Louise's advice in another thread I'm compiling TexGen from source so it links properly to my version of Abaqus 6.10-1 (i.e. linking TexGen with Python 2.6 as Abaqus 6.10 uses, not Python 2.5 as the TexGen bundle uses).
Once that is complete I think I should be able to correctly take TexGen Python scripts and load them properly into Abaqus/CAE using File > Run Script, rather than messing about with manual editing of .inp.
I should, if all goes well, finish this compiling today (first time I've ever complied a program; thankfully there is excellent documentation here) and if I get my TexGen models loaded in CAE correctly (via Python scripts) I'll let you know.

David

[dyeyanie]

Hi all,

I got the same error messages (as below) too when I tried to change the boundary condition and load cases. I was trying to impose a fixed boundary condition at one end and impose a load to the other end.

***ERROR: in keyword *DISTRIBUTIONTABLE, file "JobAI40.inp", line 442430: The
keyword is misplaced. It can be suboption for the following
keyword(s)/level(s): model

and this distribution table are referring to the orientation section of the geometry.

Is there any solution due to this?

Yanie

[louisepb]

Hi,

Not really sure what the problem is here, especially without seeing the file causing the error.

Louise

[Jensen917]

Hi all,

I have the same problem as David mentioned above. When I try to write a new input file I get an error loading the .ori file.
David seems to have found a way to solve the problem by writing a python script, but he does not seem to be active int his forum anymore.

Does anybody else have found a way to write a new input file wuthout getting errors? Or has anybody heard from David? Maybe his way of writing a script works well. Unfortunately he doesn't answer my pm.

I hope to get some help here. TexGen is s great programme and it's a pitty that Abaqus fails to deal with it in a proper way.

-Jens

[louisepb]

Hi Jens,

Are you running your Abaqus job from the command line? You need to do this in order to be able to load the .ori file.

More details of the exact problem you're having would help in trying to sort out a solution.

Louise

[Jensen917]

Hi Louise,

I have imported a voxel file of a plain weave in ABAQUS/CAE with File->Import->Model... and then set some boundary conditions to make a simple tensile test. After that I have used the Job Manager to write a new .inp file. That only works when I suppress the orientations.
From the new input file I copy the added fragments and add them to the original input file.
I try to run that extended input file from the Command Line but I get an error.

Error: Abaqus Input File Processor exited with an error.

At least it does load the .ori file now. The mistake I made before was that I changed the name of the file but not in the .inp file.
But I have no idea what goes wrong now.

[louisepb]

Hi Jens,

Abaqus should have created a .log and .dat file which will hopefully tell you where the errors were.

Louise

[Jensen917]

Hi Louise,

I found the error in the .dat file. It was a syntax mistake. Thank you very much for that hint. I am new to Abaqus and I really appreciate your help! You are doing a great job!

But back to the orientation problem: The orientations can't be used, because ABAQUS is not able to load the information from the discrete field TEXGENORIENTATIONVECTORS into the part orientation.
My idea was to write a python script that gets the orientations from the .ori file and writes a new discrete field. This new discrete field includes all the information about orientations. Further the script could assign this new discrete field to the orientations instead of the TEXGENORIENTATIONVECTOR.

Has anybody tried that before and has expierence with it? I have not been working with python yet, so it might take a while until I can share my results with you.

-Jens

[DieSure]

Hi Jen,

All you need to do is redefine the discrete field in CAE. Just go the 'Fields' under the model tree and select discrete fields. you should be able to find the field named TEXGENORIENTATIONVECTOR. Replace the current data in the Field Data window with the data in the .ori file (just right click and read from the .ori file). The first line in the .ori file is for the default values.
Hope this helps.

Shuo

[gaoxp]

Hello,
Can you please send to me a noncrimpfabric file? Email: gaoxp@imut.edu.cn
I would appreciate. Thank you.

Hi,

Welcom to the forum.

I have sent a tg3 file for a non crimp fabric by email.

Louise

[louisepb]

Hi,

I have emailed the tg3 file for the non-crimp fabric to you.

Best regards,
Louise

[paulk]

Hello Louise,

I have problems creating a non-crimped fabric, too. I use LS Dyna for analyzing friction, bending and compacting behavior on those fabrics. I tried to use the script oriented on the one which Samnite created, but it appears that there are some contact/ volume problems within the fabric as many elements collide with each other. I don´t know how to solve the problem, because I don´t have any experience in wirtting a script. Can you please send me non-crimp fabric file? I would appreciate it.

Thank You in advance.

Best regards

Paul

[louisepb]

Hi Paul,

There is a script for a non-crimped fabric in one of the Appendices in Martin Sherburn's thesis here: http://eprints.nottingham.ac.uk/10303/1 ... -final.pdf. The tg3 file is attached.

Best regards,
Louise

[paulk]

Thank You very much, Louise. You helped me a lot.