First of all, you have treated wires like beams and that is not good at all.
Everything must be transformed into volume.
Then, to make it simple and the quickest, then is to put a plate of a few tenths (in a much softer material (aluminum or plastic) than your wires) which will be attached (without penetration) to the tangency points of the outer wires. After that, all you have to do is apply pressure to the mini-plate so that all your deformations are observable.
Can you make a partial image (in the pattern so that I can see how the wires are hung on what appear to be tubes.
Also note that you don't have to test the entire model
Excuse me: but if for this type of simulation I say that we must ban the beams it is not without reason.
Without going into technical details (cf. the difference between a second-order tetrahedral element and a two-knot beam element, etc...) ;-)
In the meantime, I put forward a single argument, which is that of the mesh.
Given the size of the wires, the mesh will be sporty. ;-) Especially since we don't have a detail of the image allowing us to see how the threads are braided (or not braided if they are strictly straight).
Kind regards
Can you go to the end of your simulation in the meantime it will surely be rich in lessons. ;-)
However, I understand that you can't give us the file for privacy reasons, but there is always more than one solution.
Can you give us a picture of the fence with the connections in a corner (a very partial image therefore). The two questions I asked myself were: is the thread braided (warp thread t weft thread as for textiles and some wire mesh) and how is the connection made on the external mechanically welded (attached, welded, others).
With this minimum of information, I must be able to reconstruct a digital twin and look at how to do this simulation.