Fabric Compression

[ajroman]

Hello,

A little over 2 months ago I posted a thread asking for help with a model that I was working on that dealt with fabric compression (Original Thread:

http://eng-tips.com/viewthread.cfm?qid=255686&page=1).

Since then I have slowly inched closer at a solution but have yet to fully compress the model. I have tried several combinations of the suggestions from previous posts, but have yet to gain the desired final result.

The most up-to-date model (please see attached link below) has ran to roughly 17% of its final compression distance before aborting due to too many increments.
Could you please review the model and post any suggestions or feedback that would be helpful in completing the job? Please also let me know if you need any additional information or background on the model, I would be glad to answer.

Thank you for your time and help and I look forward to hopefully finally resolving this problem.

-ajr

Previous Related Thread Posted:

http://eng-tips.com/viewthread.cfm?qid=256519&page=1

 

[rstupplebeen]

I am looking at your model and many of my suggestions remain unchanged.

In addition every fiber has a space between other fibers. That means that every fiber can have a RBM. You need stabilization on.

You probably need to run this in explicit due to the very difficult contact.

Are you sure that this initial geometry is representative of the actual fabric? I have never seen a weave that does not touch other fibers before compaction. Even if this were possible the mechanical properties in all but the 1 direction would basically be matrix properties because all the other fibers would be compressed and not be able to develop a tensile load.

I hope this helps.

Rob Stupplebeen

 

[ajroman]

Thank you for the help once again.

Just to clear some things up, the initial model's geometry is not an actual representation of the weave. To achieve an actual representation, we need to eliminate all the excess spacing in the model as well as further compressing it to its actual weave thickness, all while maintain periodic boundary conditions so that the final compressed model would mimic a unit cell version of a dry fabric weave.

The compression step seems to be where our troubles lies. Many improvements have been made since the original model posted, such as each tow interactions with one another as well as each plate. Each interaction was paired manually to all surrounding surface that each tow may encounter contact with during compression. This approach was done because by automatically finding contact pairs would not detect all possible fiber interactions.

I have tried using automatic stabilization in the past without success. I have recently, however, tried switching it back on with hopes to allow the model to run to completion. What I have found is that the model is compressing to the point where all excess space looks to be eliminated. Unfortunately this is where the Job continues to run (increase in increments) but seems to come to a stand still for Total Time/Freq progression. It took about 370 increments to reach 29% completion (0.295291 Total time/Freq). From increments 371 to 944 (where I terminated the job) no advancement was made in total time/freq column (0.295291). Any words of advice to improve the progression of the job?

I am also concerned with the model deforming. I restricted each fiber about its perimeter at its center only. I was hoping to see once all the excess space was eliminated, the fibers to begin to compress and deform into a more elliptical cross section. I have yet to see this occur.

Thanks again for your assistance and please let me know if you need any additional information.


Do you have any suggestions as to what setting to switch on? (Specify damping factor, dissipated energy fraction, or from previous general step) Any suggestion as to what value to use? I also tried adding a contact control setting that allows for stabilization of a factor of 1. Is this needed? Would this effect the stabilization settings in step 1?

 

[rstupplebeen]

Model 2 fibers forming an "X" being compressed by the plates. If you can get that to solve those settings will be a good starting point for the full laminate.

To pretension your fibers you could apply a tensile displacement where you have symmetry planes.

You might also want to check out Explicit for difficult contact like this.

I hope this helps.

Rob Stupplebeen