Skip to main content

Hi Analysts,

I am running into this problem of an AC electric motor in a vertical electric motor driven centrifugal sea cooling pump configuration. The overall velocity vibration of the motor NDE is oscillating from 2.8mm/s RMS to 13.5mm/s RMS as picture below.

The overall velocity of Pump NDE is also oscillating from 0.6mm/s to 1.8mm/s.

Have you guys encountered this behavior and could you please suggest me the possible causes and share your related experiences about this case.

I've also attached some spectrum and time waveform, motor-pump arrangement drawing in the attached file. (Noted for this vertical configuration, V direction refer to the direction inline with suction/discharge piping, H direction is the direction perpendicular to V direction)

Kind regards and success.

Attachments

Images (14)
  • Overall Pump NDE (1)
  • Overall Pump NDE (2)
  • Spectrum Motor DE-A
  • Spectrum Motor De-H
  • Spectrum Motor DE-V
  • Spectrum Motor NDE-H
  • Spectrum Motor NDE-V
  • Spectrum Pump DE-A
  • Spectrum Pump DE-H
  • Spectrum Pump DE-V
  • Spectrum Pump NDE-H
  • Waveform Motor NDE-H
  • Waveform Motor NDE-V
  • zoom spectrum 1xRPM motor NDE
Files (1)
Tags: Motor, Oscillating, vibration

Replies sorted oldest to newest

@Registered Member posted:

Go to Troubleshooting Induction Motors and review in particular the content on page 10 about beating.

Hi John,

Thank you so much for your information. For slip-beat vibration, when was on site, I tried to record a spectrum with high resolution and low fmax to separate the beat frequencies (if any). As in the "zoom" spectrum below, I don't see the present of 4xPole frequency (this motor is 4 pole).

RM

Are there other machines operating at same speed nearby?

What has been checked/verified so far?

Have you inspected the coupling?

Are there bearings in the pump?

1x rpm in your fft shows "skirting" at the base of the peak, this can indicate a natural frequency is present, have you done a bump test with the unit off or with it running? Monitor live time with machine running, continuously bump the motor with a dead blow hammer, watch the fft for a "mound" of energy to show up while you bump the machine and decrease after you stop bumping machine. Your vibration data shows a very large amplitude difference from motor drive end H to V, this too indicates a natural frequency issue.

With the pump running, monitor vibration live time, loosen 1 foot bolt at a time and watch what happens to the vibration, if vibration changes, then there could be a soft foot present.

AMS software I see, can you post a "multiple point plot", tag the motor HVA on both ends and any pump points, looking to compare all points at once in all directions. See attached example

Dave

Attachments

Images (1)
  • multiple point plot
RM

Hi Dave,

Thank you so much for your help! For your question:

1. We have 2 next MSWC-pump to this pump with the 3x50% configuration. 2 MSWC pumps is running at a time.

2. We weren't able to physically check the pump at the moment

3. Pump doesn't have bearings

I will manage to check pump for soft foot and resonant in the soonest opportunity as you suggest. For the fluctuating of the overall amplitude, what is your theory? Does resonant or soft foot alone can create such phenomenon?

I've also attached the multipoint spectrum as you ask.

Again, I am very happy to hear about your thought.

Attachments

Images (1)
  • Multipoint
RM

The beat, in my opinion, is coming from: loose bearing fit to shaft/housing in motor, soft foot, worn coupling or the 2 running pumps are beating with each other. The imbalance from 1 pump eventually lines up with the imbalance from pump 2 then eventually the separate again runs smoother.

Your recent plot, thank you, shows a "skirt" just to the left of 1x rpm, this is typically a natural frequency. So when the pump starts, just before full speed, the vibration will increase, then drop slightly after up to full speed. If the pump is running, and you continuously bump the motor and watch fft, that skirt will increase and decrease with the impacting of dead blow hammer. Design engineers try to keep the operating speed 20-30% away from this natural frequency, yours appears to be less than that...

Keep in mind that each pump will have its own natural frequency in all 3 axis, one of the 3 pumps may be operating right at the natural frequency.

What details do you have on these pumps? Age, recent overhaul, your first readings, piping/piping support changes, base anchor flatness/rust/broken bolts

Dave

RM
@Registered Member posted:

Hi Analysts,

I am running into this problem of an AC electric motor in a vertical electric motor driven centrifugal sea cooling pump configuration. The overall velocity vibration of the motor NDE is oscillating from 2.8mm/s RMS to 13.5mm/s RMS as picture below.

The overall velocity of Pump NDE is also oscillating from 0.6mm/s to 1.8mm/s.

Have you guys encountered this behavior and could you please suggest me the possible causes and share your related experiences about this case.

I've also attached some spectrum and time waveform, motor-pump arrangement drawing in the attached file. (Noted for this vertical configuration, V direction refer to the direction inline with suction/discharge piping, H direction is the direction perpendicular to V direction)

Kind regards and success.

Thinh,

Hello,

I have a couple of questions. See the M2V Point Spectrum image.

Referencing to Point # M2V
The sideband frequency appears to be of a low frequency range, similar to, # of poles times slip frequency, {(1800 –1780)}x 4 = 80 cpm/60 Hz = 1.33 Hz. But this particular pattern could be because you captured this spectrum at just the “right” time to give the impression of electrical problems. Have you more data, on a different day, on this point M2V?

But on the other hand, if this is a speed difference of 2 motors, coming in phase and causing an increases in amplitude, on this particular motor, then this spectrum pattern would, be constantly changing patterns at a rate close to, or equal to the beat frequency you hear.

Does the other motor, which is running at the same time this one is running, have the same beat frequency and/or amplitude change (not necessarily the same amplitude) when this happens to this “problem” motor?
Are all these pumps connected to the same discharge piping, that might allow transfer of vibration from one pump to another?

Thanks and Have a Safe Day,

Ralph

Attachments

Images (1)
  • Sidebands on 1x rpm question 06-22-2022: Point M2V
RM

Many good suggestions have already been given by the forum members.

What I advise is to compile the suggestions into a checklist to systematically go through them.

In order to eliminate some of the mentioned possibilities, there is a need to take additional measurements with specic configurations or perform more tests. For example, to verify the sidebands and confirm the frequencies, I would take high-resolution data with a high number of averages say 50. I would also perform a shutdown test either with overall amplitude monitored or with spectrum on peak hold averaging. If possible temporarily switch off the sister machines and see the effect.

The above steps focus on the possibility of having external forces. You may be able to develop similar tests to focus on the forces coming from the internal parts.

Regards- Ali M. Al-Shurafa

RM

Add Reply

×
×
×
×
Link copied to your clipboard.
×