RBE2 and RBE3 vs natural frequency 2

[FSB1]

I have a model connected to masses with RBE3
after a normal modes analysis I observe the 1st frequency

after converting these RBE3 to RBE2 the 1st frequency is lower than before.
I expected the 1st frequency to be higher instead, since RBE2 are infinitely rigid?

 

[janFEA]

Hi,
this is interesting observing. I would say the same at first sight - if RBE3 is correctly defined, then it does not add any additional stiffness to the structure.
And this is many times the advantage compared to RBE2.
I'm thinking what case could be responding to your problem...The "magic" is in distribution of mass and stiffness.
I haven't test it but it could be a good sample for validation:
let's have a cantilever beam. The end (let say 1/4 of length) is connected to the RBE3 which holds some mass element. 1th eigen mode is bending - bending of all length.
Then we have RBE2 instead of RBE3, it means 1/4 of length is rigid. 1th eigen mode is bending as well, but only 3/4 of length. Without any additional mass I would expect this frequency is higher (shorter length, stiffer), decreasing can be caused by mass of rigid part (compared to the rest).
It would be good to make some sensitive analysis and vary with the additional mass, because if this is true, there can be some boundary when 1th frequency starts to increase.

But as I said, I haven't test this, just idea.

Regards, Jan


With best regards,
Dr. Jan Vojna
Lead Engineer Development

Siemens, s.r.o.