Vibe "analysis" by natural frequencies only?

[dag113]

I have a situation where a subcontractor is designing a structure to be attached to our system (a moving vehicle). Per the requirements from our customer, their structure needs to withstand a certain vibration profile (broadband with narrow band spikes). However, we are unable to share the details of the profile with the subcontractor for a variety of reasons. The sub, therefore, wants to provide us with the natural frequencies of their structure. Is there a way to take their natural frequencies and determine if the structure meets the given vibe profile? (My initial thought is "no," but I'm assuming I'm missing something.)

I greatly appreciate any help any of you vibe experts can provide.

 

[drawoh]

dag113,

What are you trying to work out, the worst case deflection?

--
JHG

 

[dag113]

"What are you trying to work out, the worst case deflection?"

Normally, I would think I'm working out stresses in the structure caused by the vibrations. If this was my own structure, I'd be running a modal and then a vibration analysis based on the inputs defined in the vibe profile, review the stresses, and compare them to the strength of the material. The customer's spec simply says that the various structures need "to meet their performance requirements when subjected to the given vibe profile"; the only "performance requirements" for the structure while subjected to the vibe profile is the stress safety factor requirement.

Does that help? ('Cause I'm still thoroughly confused.)

 

[WARose]

Me personally, I’ve never done a vibration analysis without at least an estimate of what the unbalanced force is. A modal analysis will certainly let you know if you are anywhere near a natural frequency of the system (ergo minimizing the dynamic magnification factor [DMF])…….but without an actual number causing displacement: how do you know what the steady-state amplitude is?

So I guess my point is: getting the frequencies of their system may not be enough: the system as a whole (along with the unbalanced forces) must be part of the analysis.

 

[dag113]

WARose,

Thank you very much for your response. That's just what I was thinking. Their natural frequency results will just tell me if the applied vibration profile will have a magnification issue like you said. The stresses at other frequencies (steady-state amplitude?) could still be unacceptably high, though, right?

 

[WARose]

The stresses at other frequencies (steady-state amplitude?) could still be unacceptably high, though, right?

That’s certainly a possibility. But what worries me about any sort of comparison like you are talking is: it may not take into account a localized affect (that’s why I talked about modeling the whole system). Let me give you an example of what I am talking about: recently I modeled a foundation system that had a compressor sitting on it. The entire foundation had a frequency that wasn’t anywhere near the forcing frequency (for any mode). Sounds like it would be ok right? Well, a portion of my foundation was shaking like crazy in the computer model. The reason for that was a portion of the mat that spanned between the piles had a different natural frequency that got pretty close to the forcing frequency.

So although we are talking situations here: I think/hope you catch my drift. Your model needs to be something that can capture every possible motion.

 

[dag113]

Thank you for the example. That's a very good point and gives me more backup for saying that this isn't a good idea. I really appreciate your help, WARose.

Apart from actually running the vibe profile against the model (which they can't do because they can't have the profile and I can't do (easily) because I don't have the model (and theirs isn't compatible with our FE system, etc.)), I think I'm a bit stuck.

 

[IRstuff]

"withstand" how? Fatigue?

Note also, that no mounting structure is ever infinitely rigid, despite our claiming it. So when this "structure" is attached to the vehicle, how is it going to change its, or the vehicle's, natural frequencies?

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

 

[dag113]

I understand your point about the finitely-rigid mounting structure. Considering the mass and stiffness of the particular vehicle involved, though, I'm not really concerned about significant changes to its natural frequencies due to what we're mounting to it.

And the customer spec is vague regarding "withstanding." It just says that components shall meet their performance requirements when subjected to the vibration profiles given. The only performance requirements I have for the structure is that it must meet a certain safety factor. No "life" requirements, etc.