Email from James Zhang

[Martin]

I downloaded TexGen at the link you gave me and have been playing with TexGen today, creating geometric model and exporting it into ABAQUS. I am so excited about this software and the work you have done. I have a question for you, I looked at your presentation, some models were created by TexGen, but I can not reproduce them, such as the 3D woven fabric and knit interlace. I was wondering if the version I have downloaded is the same as you used in the presentation?

To answer your question, the version you have is indeed the same one used to generate the models in the presentation you saw. However, currently the graphical user interface is limited to creating 2d woven fabrics. This is simply because I haven't had time to program the graphical user interfaces for creating other types of fabrics yet. This does not mean you can't do it. It just means that you have to use the python scripting interface to do it. And indeed that is how I have created the 3D models.

There is some documentation on writing python scripts to create models here:

http://texgen.sourceforge.net/index.php/Scripting_Guide

In the future I will create user interfaces for other types of fabrics, however if you are serious about creating fabric geometric models I suggest you use the python scripting because you have more control over the models created.

On a side note, my colleague Jon Crookston has written a python script to export TexGen model to ABAQUS. You may find this works better than using the IGES and STEP export. You may be able to persuade him to upload his code

[Robitaille]

Jon, Martin,

We're about to resume Simon's work based an actual reinforcements - textile prepreg panels were made & are measured for geometry and thermal. I am planning to forge ahead with v2 for the time being - previous issues more related to our models than software i.m.o.

What kind of arguments would be needed for this capability to be shared towards possible use of v3? Jon, do you get any control over mesh density as allowed by Gambit? Arguably this will not be as important now, but still nice to have.

Cheers

[Martin]

Thanks for making use of the forum Francois! (I'm not sure if you intended this as a reply or a new topic, it doesn't seem related to the original post?).

We're about to resume Simon's work based an actual reinforcements - textile prepreg panels were made & are measured for geometry and thermal. I am planning to forge ahead with v2 for the time being - previous issues more related to our models than software i.m.o.

Good stuff! Can I ask why you prefer to stick with v2? If its because you don't want to get into the python scripting I understand that. But I've been working on something recently on v3 that you may find interesting, its a graphical user interface for creating textile geometry interactively. I think it should be much easier to use than the previous interface available in TexGen v2. I plan to release a new version sometime soon, its pretty much functional already, I am just adding a few finishing touches and doing some testing.

What kind of arguments would be needed for this capability to be shared towards possible use of v3? Jon, do you get any control over mesh density as allowed by Gambit? Arguably this will not be as important now, but still nice to have.

I'm not quite sure what you are talking about here, what capability?

As for mesh density, yes you can vary it. Basically Jon's script does the same as what I had implemented for gambit through the use of journal files. With a few enhancements such as applying material properties and such. I suspect it is just a question of changing a single parameter to affect mesh density.

[joncrookston]

What kind of arguments would be needed for this capability to be shared towards possible use of v3? Jon, do you get any control over mesh density as allowed by Gambit? Arguably this will not be as important now, but still nice to have.

I'm not quite sure what you are talking about here, what capability?

As for mesh density, yes you can vary it. Basically Jon's script does the same as what I had implemented for gambit through the use of journal files. With a few enhancements such as applying material properties and such. I suspect it is just a question of changing a single parameter to affect mesh density.

Yes - that's right. Mesh density is specified as an argument within the script. If you can use this kind of approach, Francois, I think you'll see a significant productivity increase compared with the old v2/Gambit method.

[Robitaille]

Hello,

Thanks. The link was that there was mention of ABAQUS in the first post.

On the contrary, I'm very keen to get into v3 and now I have time to do it. I guess reason for sticking to v2 is that the process is all mapped - geo to mesh to export - and we've done it before. Simon does calculations in ABAQUS but we've not been through v3 to ABAQUS yet. The change will happen, maybe not for immediate work that I'm trying to complete - but I have many other ideas!

As far as arguments, what I meant is favours really. But it will be more sensible to discover v3 properly before considering an actual switch.

Thanks!