weigthed link, unreal simulate results

Paul,

I think the only thing I can add is ... If we want to apply a gravity load then we cannot have inertia relief. So one should remove the load acting on the part such that the inertia relief acts as both gravity and the small accelerations required for balance. The answers are different. Using inertia relief restricts the problem to small deflections and prevents the use of contact (in this case, contact doesn't matter).

I would want to measure the forces at the suspension points to see what errors there were.

why such a fine mesh in your model?

Thanks

Just some general sympathetic thoughts/comments for Antonio otherwise ...

Rotations about the single suspension point are required to achieve balance. These are too big leading to failure insufficiently constrained.

Using light ground springs is a standard technique and one expects it to work. But unless the balance is pretty much perfect initially then the natural subtended movement to an equilibrium position is still too big. Stiffening the springs reduces the movement but then the answers will be wrong, the springs should carry zero load ideally.

We could use large deflection but then we can't use beams, ground springs, weighted links, inertia relief

Modelling cables is problematic as using beams or springs, one cable will carry an unreasonable share of the load whereas in reality the cables would probably pass over a pulley and (ignoring friction) have equal tensions and exchange length. I haven't got a technique for allowing cables to naturally exchange lengths over a pulley(s) and the geometry of the angle of wrap and the point where the cable begins to tough the pulley is not simple. Resorting to contact so that the contact point sorts itself out gets us a little way along another inevitably blind alley because we cannot define a sensible cable material property.

Cables tend to be stiff in tension but not bending/torsion/compression. What material property is that? I tried and failed defining Hyper-elastic material that would have negligible stiffness with negative strain. A transversely isotropic material strong one way but not transversely might be useful but non-isotropic materials cannot be used in large defection studies. Besides, the cable will have the same strength in compression.

We can have 'tension only' springs by using a force deflection curve rather than constant stiffness, ok for fixed length cables (no pulley). A study will be automatically large deflection if a non-linear spring is introduced. Some stiffness must be provided against 'free' rotations else the model will fail and because we cannot have ground springs, simple springs and additional constraints can be used. Exchanging cable length over the pulley is still a problem.