I’m using a curved polyline as a beam and a cable attached to both ends of it with some additional supports as my setup. Can someone explain why the beam does not straighten out, no matter how weak the cable is?
I would like to use curves that are already in shape as my primary inputs for the beams, and then hook them up to multiple cables and by adjusting the pretension get them back into desired shape. I tried bending them from straight but it gets more difficult once there’s more beams involved.
Hey, you have to define that the curved beam was initially straight. If you don’t do this, the program won’t know and will assume this shape as relaxed shape. You have two options: prestress properties or reference element. I would recommend prestress properties. Just set L, AB1,AB2,AT to true.
It kind of worked, the beam bent backwards slightly after decreasing the pretension on the cable, however I didn’t manage to get it straight (which was my initial goal). Could you explain the other approach ? How do I reference the element? Do I need to run 2 non-linear analysis - first without the cable on a straight rod, and second on a curved rod with the cable with the reference of the first one?
In order to reference a geometry, you have to generate two beam elements, one straight (line with only two control points!) and one bent. Connect them to the reference element component. Supports and loads have to be defined on the deformed geometry. That’s basically it. However, there will only be a minimal difference if everything else stays the same. As the beam straightens, the cable becomes longer and stresses result. They counteract the bending stresses of the beam. The beam straightens only until the forces of cable and beam are in equilibrium. So depending on the structure that you want to simulate, the result that you get might already be correct. If the cable is just arbitrary and you only care about the force, i.e. prestress, you could also decrease the diameter or youngs modulus of the cable. If you just want the straight beam, you can just omit the cable and let it relax.
Understood. The goal was to simply simulate accurate rod behaviour from curved beam as an input. So I guess the first option with prestress properties already does that and then it comes down to cable material settings. Thanks!
