Concerned that Modal Frequencies are Close Together

[cobra46]

Hi All,

Been having fun using the vibration side of SW simulation over the last couple months.

I am concerned about the modal analysis results on an aluminum billet housing assy. I ran the study out to 48 modes and they start at 538 Hz and go to 865 Hz. There are a lot of resonant frequencies over a quite narrow range. Does this seem strange to anyone else? Check out modes 30-36 below:

30 842.9
31 844.0
32 844.2
33 845.3
34 845.8
35 847.0
36 848.1

Being so close doesn't pass my sniff test but I am new to frequency analysis. What are your thoughts on this?

I have another question regarding how to setup the study to simulate my housing being mounted to a thicker more rigid plate. It seems many of the resonant shapes would be impossible if the housing was mounted to a thick plate. No penetration contact between the thick plate and my housing is not available with the frequency study. I was forced to simply fix the mounting holes.

Thanks,

Kevin

 

[FEA way]

Check what happens if you refine the mesh. Also, take a look at the mode shapes associated with those frequencies.

In modal analysis, nonlinear contact cannot be used. You could connect the parts using bonded contact.

 

[cobra46]

@FEA way,

Thanks for the reply. Changing the mesh yields the same results.

I think using a bonded contact between the housing and thicker rigid plate would not give accurate results as the entire mounting surface would be 'stuck' to the rigid plate. In reality, the plate could slide and separate from the mounting plate. This seems like a serious weakness of SW simulation.

 

[FEA way]

It’s not a weakness of SW Simulation, it’s a fundamental assumption of linear dynamic analyses - no nonlinearities, including changing contact conditions. The contact state is basically frozen for such simulations and perfect bonding is assumed.