Can't find equilibrium on simscape multibody/ can't put the steady state by default

Hello,

I am simulating the following system: It is a double pendulum haning on 4 wires. The 4 wires are considered as damp/springs (prismatic joint).

I also attached the simulink model.

Each wire is connected to the mass by a gimbal joint having its stifness and damping.

My first question is:

• How should I find the steady state position ? I would like to make my system be at the equilibrium at t=0s. But because of the complexity of the geometry, It is difficult to find this position by calculation. Is that possible to let the model run, let it reach the equilibrium position and then use these coordinates to set the new initial position?

My second question is:

• My system can't find a stable steady state. When I start the simulation, it finds a steady state, and then, it starts to vibrate with an increasing amplitude and then it collapses. So, to avoid this, I can put very high damping and it reaches the equilibrium, but when I zoom a lot on a wire for exemple, it jumps from a position to an other so the system oscilate a bit. I think I am over constrained but I can't find the reason.

Thank you for your help and your time,

Vermeer

[Vermeer Bonhomme]

[J Chen]

Can you attached the Simulink file?

[Vermeer Bonhomme]

I can but it works with parameters and solidworks that are defined in the same folder. So I will share all the folder. Like it is too heqvy I have splited it in 3 zip files that you have to "merge" in one folder.

Thank you very much for your help

Tell me If you can't load it. If you run the code with these parameters, it should diverge.

[Vermeer Bonhomme]

You might have to change the path of the STEP file in the Support subsystem .. (./Support.STEP is the right one)

Best

Answer 1

[Steve Miller]

There are a few adjustments you should make to your model.

1. Set the Consistency Tolerance in the Solver Configuration block to 1e-9 (the default). In your version, it is set to 1e-5. This is 10000 times larger than the default.
2. Set the solver to daessc

With these changes, it reaches steady state in about 50 seconds - see screenshot below. Attached is the model with the changes above, plus with relative paths for the CAD geometry.

Another change I would recommend is changing some of the Gimbal Joints to Universal Joints. This would prevent the wires rotating about their primary axis.

You can save the final state of a simulation and use it as a starting point. Look at the Solver Configuration panel, Data Import/Export pane.

--Steve

[Vermeer Bonhomme]

Hello,

Thank you for your so detailed answer and help. Thats awesome. Indeed it works way better right now.

By the way, what is going on at ~ 80s for the velocity? Is that a jump ?

Again thank you so much.

Vermeer