Natural frequency of turbine rotor

[suviuuno]

I get peculiar results from turbine rotor. Adding constraints to bearing locations lowers natural frequencies. Higher frequencies with no constraints comparing to same mode shapes as in constrained model. No damping included at bearing locations. Beam and shell elements are used. Adding any constraints should add onto natural frequencies, right?

Thanks a bunch

 

[Tmoose]

Are any of the bearing locations hollow shafts modeled with shell elements, and constrained at a single node on the OD?
I was getting some unrealistic results from deflections at the bearing locations that looked kind of like this C>. Kind of like a tent pole poking a very localized high spot.

With my shaft modeling I started using 2 pairs of spring elements attached tangentially at the neutral axis OD. 4 springs = one bearing. This is probably somewhat stiffer than a real bearing, but I got pretty good results with it.