5 Order on 6 joint flexible coupling?

[Sofistioelevib]

Good morning Guys,
yesterday on my vibration inspection i found a problem on a flexible coupling as in a picture.
The configuration is motor and pump simply without gearboxes or belt, the speed is 1480 RPM
Simple to see 6 join on a coupling but i found a strong spectra with only 5 times per revolution (8 mm/sec) especially on axial plane.
To ensure to 5 times exactly i performed a 200 Hz fmax with 6400 spectral lines
I thought to find 6X instead of 5X, have you any idea why?
No information on time domain
The spare unit is running and we are going to change coupling soon

Regards

 

[Strong]

Just a little bit of wear!! When components get that bad, the vibrations don't necessary follow linear dynamics. There can also be a resonant structure (lateral or torsional) that amplifies the 5xSS and not the 6xSS. I have never seen an elastomer-type coupling with metal hub wear that extensive. You could investigate the root cause and make modifications to minimize future failure.

Walt

 

[electricpete]

You want to understand why you are seeing 5x. I suggest put your spectrum on log scale and look for one of two patterns:
[ol 1]
[li]PATTERN A - Harmonics of 1x which gradually rise up to 5x and then decrease from there. It suggests you have something like impacting at 1x, exciting a resonant response at 5x as Walt suggested.[/li]
[li]PATTERN B – 5x stands far above the rest, 10x, 15x 20x etc also stand out from the nearby peaks. This suggests a forcing function at 5x. It could be something like blade pass frequency (do you know how many blades on the pump?). Perhaps the coupling malfunction somehow resulted in change of the axial or radial clearances around the impeller.[/li]
[/ol]
You might also get clues on which scenario by looking at the TWF.


=====================================
(2B)+(2B)' ?

 

[Sofistioelevib]

Thank you guys for your point of view.

Effectively the frequency seems excited from 5th haromonic of 1X even if really higher then the rest of Harmonics but looking in in logarithmic scale i found all harmonics with 5th higer.
Last things:

suppose to know the resonant frequency and it is 5,1XM.... could be reasonable to think that 5XM will excite 5,1XM and i will see on a spectra 5,1XM or do you think i will see ONLY 5XM?

To clarify...when i'm really near to resonant frequency i think the resonance will be excited so in this case my aspectation will be 5,1XM high instead exactly 5XM unless the forcing frequency at 5XM is really strongher than resonant frequency.

Hoping to be clear what do you think about my last thought?

Regards

 

[GregLocock]

Perhaps it would help if you actually uploaded an image of the spectrum or the data as a time history.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[Sofistioelevib]

UnfortunatelyGreg i can't.
I have instrument with no software (analysys made on the display) and display is not so easy to photograph.
Furthermore i've no information about time

 

[electricpete]

Effectively the frequency seems excited from 5th haromonic of 1X even if really higher then the rest of Harmonics but looking in in logarithmic scale i found all harmonics with 5th higer.

So you have many harmonics of 1x which happen to peak at 5x. (I assume when you look out around 10x, 15x, those harmonics don’t stand up above the others). Then you likely have what I called pattern A: the fundamental forcing frequency is 1x and you are exciting a resonant response near 5x.

Last things:

suppose to know the resonant frequency and it is 5,1XM.... could be reasonable to think that 5XM will excite 5,1XM and i will see on a spectra 5,1XM or do you think i will see ONLY 5XM?

My assumptions about your notation:
[ol 1]
[li]I think 5,1 is what we in US would call 5.1[/li]
[li]I think 5,1XM means 5.1 times running speed.[/li]
[/ol]
Then I think you will see ONLY 5XM. This is under an assumption that your forcing function is periodic with fundamental frequency 1XM, in which case the response will likely be periodic at 1XM and will will have only exact harmonics of 1XM. It is a common scenario for rotating equipment (*).

To clarify...when i'm really near to resonant frequency i think the resonance will be excited so in this case my expectation will be 5,1XM high instead exactly 5XM...

I disagree with what you stated in bold. Under the assumptions I made above (that 1XM is forcing frequency and response is periodic with fundamental frequency of 1XM), then the response contains ONLY exact harmonics of 1XM. The magnitude of those response peaks may be influenced by resonances, but their frequencies will not.

...unless the forcing frequency at 5XM is really strongher than resonant frequency.

I don't know what that means. How do we compare characteristics of the excitation to be stronger or weaker than characteristics of the excited system? It seems to me like a non-sensical comparison.

* There may be circumstances in rotating equipment when a 5.1XM resonance would be excited. Some examples would include broadband excitation such as flow noise or rubbing. These tend to excite any resonance within their frequency band. There are also a variety of discrete non-syncronouse excitation frequencies that may be generated such as bearing defect.


=====================================
(2B)+(2B)' ?