Discussion Forum

Good day,

I am trying to simulate an electrical synchronous generator with the rotating machines physics. Everything works fine so far.

The only problem I occur is when trying to use an interpolated nonlinear mu(rmm.normB) curve.
Before the computation process reaches 1%, it is being cancelled and the dialog window says:

"Failed to find consistent initial values.
Last time step is not converged."

Interesting is, that if I don't divide the curve by mu0 (to get mur),
[ int3(rmm.normB) instead of int3(rmm.normB)/(4*pi*1*e-7) ]
the computation successes, but the result looks wrong of course, since the relative permeability is just too low in that case.

When I choose mur as constant, the computation works fine and the results seem to be right, since they equal a simulation I made with an other FEM software.

I already tried to use a very fine mesh and tiny timesteps (1us) . Usually I'm using the time-dependent direct solver but it didn't lead to success yet.

Is there anybody who can help me with this?

Thanks in advance!