time domain analysis converge issue

[aaronlsd11]

Hi,

I'm using the windturbine example to do a time domain analysis. Instead of using point load, I added a uniformly distributed loads on the whole beam with a magnitude of 100kN/m(extremely large), with a time vector of [0:0.1:10]. However, the result did not converge at 8th step. I have tried to adjust the force_scale factor, and it does help for another few steps, but still failed to converge at less than 15 steps.

Any ideas of why this could be happening would be really helpful.

[Cardoza2]

Sorry for the late response, but I've found that often the time domain analysis won't converge if you have unrealistic stiffness and/or mass matrices. In fact, it's much less forgiving if your mass matrices are wildly off. If I remember correctly, the wind turbine blade example is just a cobbled together example to demonstrate some capability. So it might not generalize well.

I mean you might get somewhere with the force scaling factor, but I don't think it would be a robust way to solve the problem.

Also, have you tried decreasing your time step? With such a large loading, you might need a smaller time step to resolve the temporal portion of the problem.