Hello everyone! 

For the last month or so I've been trying to model the behavior of a loose tube fiber optic construction under a -40-degree temperature cycle. To minimize the stress placed on the fibers during construction, they are pulled in loosely with aramid yarn and therefore already have some "waviness" to them. During temperature cycling, basically the outer jacket shrinks since it is a type of plastic, but the fibers do not because they are made from glass. The shrinkage forces the fibers further into the jacket where they interact with one another and with the inside of the jacket. 

I have tried several different approaches to simulating this. I tried for some time to keep the fibers "free moving" while completely enclosing them inside the jacket cylinder to restrict rigid body movement, but this often proved too much for the static linear simulation to handle. So in the name of simplification, I am just fixing the fiber end faces to the jacket end faces so they are forced to shrink with the jacket during the temperature load. I know this puts artificial strain on the ends, but my plan is to just omit the ends from the final figure and just look at the middle section where the fibers are interacting with one another and with the jacket. 

I keep hitting the same wall though. I can't seem to consistently get the contact interaction between the outside of the fibers and the inside of the jacket to work. When the jacket shrinks, the fibers simply clip through to the outside of the jacket. To exaggerate the interaction I have the load set to -150C instead of -40C. 

This is how I have my contact interaction set up. I have the gap range set really low because I was experimenting but have tried many values between 0.3mm and 0.00005. The Gap between the fibers and the jacket on the end is around 0.254mm so I know my gap range needs to be smaller than that. 

Then I have a self-contact defined on the fibers the same way. When I look at the contact pressure plot, there are some sections that show stress but only a couple, and the fibers clip through one another anyway. There is no overlap or interference between any of the bodies before simulation. I'm really at a loss here! Does anyone have any tips or suggestions? Or notice if I'm doing something obviously wrong? 

Thanks!

--Devin