Vibratory Limit Acceptance for Pump Purchase 1

[GrimesFrank]

Purchasing a 3"x1-1/2"x11" (fluid: water)/(OP: 155 gpm @ 320ft)/(40hp @ 3450 RPM) centrifugal pump/motor ass'y. The vendor has asked me to accept a vibration limit of .28 in/s RMS at running speed.

I have a hard time accepting this as our operating predictive maintenance program says to issue a monitor alert at 0.139 in/s RMS for our running pumps. Even looking at Fig 9.6.4.9 of ANSI/HI 9.6.4 has a limit of 0.176 in/s RMS.

Now I understand about the dampening effect of grouting in a pump but what quantitative (or qualitative I guess) method can I use to associate a vibe spec in a pump purchase (test on bench) to how it will vibe in the field?

Frank "Grimey" Grimes
Rule 25. of Swanson's "Unwritten Rules of Management"
Have fun at what you do. It will reflect in your work. No one likes a grump except another grump.

 

[GregLocock]

That's about 0.3 g . Sorry I haven't got any machinery books to hand, but that sounds high for balance on a rotating machine of about the size of a small car engine.

Not destructively high, but not great.

Grout won't damp out first order much, it is a powerful mechanism, used to deliberately generate vibration on many rigs.

You'd need to know a lot about the mounting stiffness before you can estimate the operating vibration from a bench test, even then the factor of two you are looking at is inside my guess as to accuracy of that prediction. The problem is that you are very interested in how much of the baseplate mass you are coupling the pump into.

I suppose you could pull one of your current pumps and bench test it to give some idea of the difference between operating and bench vibration.



Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[sms]

That is way too high for shop acceptance. Even if you were talking overall that is too high, for running speed that is way too high.

-The future's so bright I gotta wear shades!
Please see FAQ731-376 for tips on how to make the best use of the Eng-Tips Forums.

 

[electricpete]

Most folks in the US use ips on a peak/0 basis (even when referring to overalls). .28 ips RMS is abouto 0.4 ips pk/0.

API 610-2003 tables 7 and 8 has some limits much lower. For horizontal pumps they limit bearing housing vib to 0.3 mm/sec rms which I think is around 0.15 ips pk/0.

Is the vib at 1x or blade pass? Blade pass vibration of course often varies with flow and often gets larger at low flows.

It is true that vib is expected different at the shop than installed, but typically the shop test should be performed removed from resonance so no resonance amplification expected.

Most important - what standard or spec was invoked at the time of purchase? That's what counts in terms of what you can hold your vendor accountable to.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[GrimesFrank]

Thanks guys I didn't think I was insane questioning this. The vendor is thinking I am a nitpicker with this, they almost had me convinced to change my mind.

I was initiated talk of negotiating a clause to say we accept a certain limit on the bench but if we can't get under the "real" limit in-situ its cause for rejection. They were squirming, but didn't run when I broached it.

Any thoughts?

Frank "Grimey" Grimes
Rule 25. of Swanson's "Unwritten Rules of Management"
Have fun at what you do. It will reflect in your work. No one likes a grump except another grump.

 

[GregLocock]

Oo, have I misunderstood - does "The vendor has asked me to accept a vibration limit of .28 in/s RMS at running speed."

mean first order (~60Hz) only?

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[GrimesFrank]

Greg,
Yes they have asked me to accept it at ~60 Hz (3450RPM / 60s/min = 57.5 Hz)

Our original spec called for a limit of 0.001 in. (pk/pk) = 0.181 in/s (0/pk)
(I was not responsible for the original spec.)
They came back and said they can meet 0.003 in. (pk/pk) = 0.542 in/s (0/pk)

My knowledge of vibratory engineering was limited so I questioned it and did research, found out this is excessively rough for a pump my size so I rejected it.

They came back with 0.28 in/s (RMS) = 0.396 (0/pk). This came from ANSI/HI Std Fig 9.6.4.10 for what I found out was a solids handling pump. My pump handles water, which according to ANSI/HI Std Fig 9.6.4.9 should be about 0.176 in/s (RMS) = 0.249 in/s (0/pk).

Now I have RMS and peak numbers rolling around in my head and I'm now trying to determine if our original spec was out to lunch or is the manufacturer just asking for a wide margin with no reasonable justification.

Discuss.

Frank "Grimey" Grimes
Rule 25. of Swanson's "Unwritten Rules of Management"
Have fun at what you do. It will reflect in your work. No one likes a grump except another grump.

 

[sms]

I live by API 610/ ISO 13709 which are specifically for pumps in hydrocarbon service, in other words not water, but:

Overall vibration is limited to 3 mm/s rms = 0.12 in/sec rms overall

Any specific frequency is limited to 67% of that or 0.08 in/sec overall.

This is less than half of what your guys are quoting.

Also be careful about translating Pk-Pk displacement to velocity. Usually Pk-Pk displacement is specified for shaft vibration using a proximity sensor on a pump with journal bearings. Velocity on the other hand is usually bearing cap vibration and could apply to either a journal bearing or a rolling element bearing machine.

Since these are really two different measurements, it is not really valid to convert shaft displacement criteria to bearing housing velocity criteria.

-The future's so bright I gotta wear shades!
Please see FAQ731-376 for tips on how to make the best use of the Eng-Tips Forums.

 

[GrimesFrank]

Sanity check people.

I have had time to dig through the surface of this.

The pump is an ANSI pump, not API I'm afraid. According to ANSI/ASME B73.1 - 5.1.4;

5.1.4 Vibration.
The vibration level measured on the pump bearing housing at the manufacturer’s test facility at rated condition point (speed ±5%, flow ±5%) shall not exceed twice the limits shown in Fig. 9.6.4.4 of HI 9.6.4.

I looked up Fig 9.6.4 and the limit for a 40hp edition. The limit is 0.155 in/s-RMS, hence 2x 0.155 = 0.31 in/s-RMS > 0.28 in/s-RMS. I hate blindly following standards but the code don't lie, would you let this go, I guess I have no choice do I?

Au secours!

Frank "Grimey" Grimes
Rule 25. of Swanson's "Unwritten Rules of Management"
Have fun at what you do. It will reflect in your work. No one likes a grump except another grump.