High Vibration of electrical motor

Can you have a sketch of the pump and motor? From the point description, I imagine it's a horizontal between bearing pump driven by a motor.

The readings are really high. Consider shutting down the machine for your safety.

A wild guess for the cause would be a resonance and I would consider doing an impact test to find out the natural frequencies.

Regards- Ali M. Al-Shurafa 

Is this a vertical pump, as it would seem so from the higher motor readings? I would perform phase readings around the motor and pump to give you a better idea of how the machine is moving and likely causes. Fill out a sheet similar to the one found here:

https://www.pioneer-engineerin...ding-vibration-phase

Check alignment of course (I'm sure you already have with the high axial readings) and the motor base for flexibility if it's a horizontal driven unit. It is strange to have such a dominant high amplitude sinusoidal waveform only on the motor and not on the pump however it is difficult to understand without knowing the machine component configuration. You should be looking at acceleration in peakvue (for bearing, gearmesh other high frequency faults) not velocity however this is a running speed issue so peakvue is not relevant to the diagnosis in this particular case.

Filling out these relative motion plots is always helpful when diagnosing running speed issues.

I agree with Ali that a machine arrangement sketch/photo with measurement points would be helpful, and any impact test could confirm natural frequency close to 1xSS.

Some machine history would be useful, such as: new/old machine?, new problem?, and change in machine operation (speed/load), and maintenance or upgrade work on machine or piping system?

PeakVue spectrum and waveform is generally presented as acceleration and not velocity. It always good reporting practice to put spectrum and waveform cursor on something important and not at some random location or leave cursor off.

Walt

Shurafa posted:

Can you have a sketch of the pump and motor? From the point description, I imagine it's a horizontal between bearing pump driven by a motor.

The readings are really high. Consider shutting down the machine for your safety.

A wild guess for the cause would be a resonance and I would consider doing an impact test to find out the natural frequencies.

Regards- Ali M. Al-Shurafa 

It is Single stage overhung pump. with flexible coupling arranged horizontally..Simply to say OH1 arrangement as per API code.

ckyjohn posted:

Is this a vertical pump, as it would seem so from the higher motor readings? I would perform phase readings around the motor and pump to give you a better idea of how the machine is moving and likely causes. Fill out a sheet similar to the one found here:

https://www.pioneer-engineerin...ding-vibration-phase

Check alignment of course (I'm sure you already have with the high axial readings) and the motor base for flexibility if it's a horizontal driven unit. It is strange to have such a dominant high amplitude sinusoidal waveform only on the motor and not on the pump however it is difficult to understand without knowing the machine component configuration. You should be looking at acceleration in peakvue (for bearing, gearmesh other high frequency faults) not velocity however this is a running speed issue so peakvue is not relevant to the diagnosis in this particular case.

Filling out these relative motion plots is always helpful when diagnosing running speed issues.

Have a look at the spectrum in acceleration no symptoms of high G's was noted also motor DE and NDE bearing was replaced recently with coupling pads. 

Attachments

Files (1)

I only made the statement that looking at peakvue should be in acceleration. In this case you are dealing with running speed frequencies of interest so peakvue is not what you would utilize in analyzing the potential issues. You were of course correct to use velocity spectra and time waveform but again phase is the missing element here. Do the relative motion plot with phase and analyze the results. If nothing pops out there I would move to a bump test for structural resonance but again I cannot assume the arrangement or if this is relevant to this case. A picture is worth a thousand words as they say.

Walt Strong posted:

I agree with Ali that a machine arrangement sketch/photo with measurement points would be helpful, and any impact test could confirm natural frequency close to 1xSS.

Some machine history would be useful, such as: new/old machine?, new problem?, and change in machine operation (speed/load), and maintenance or upgrade work on machine or piping system?

PeakVue spectrum and waveform is generally presented as acceleration and not velocity. It always good reporting practice to put spectrum and waveform cursor on something important and not at some random location or leave cursor off.

Walt

Thank you for your reply!

It is horizontal, single stage overhung pump with flexible coupling.

The machine is old and as per history available fault seems older. But the readings are now cross the limits defined and increasing. Suggest some solution to resolve issue on permanent basis.

Yep! you are right about the presentations of PeakVue and TWF.

ckyjohn posted:

I only made the statement that looking at peakvue should be in acceleration. In this case you are dealing with running speed frequencies of interest so peakvue is not what you would utilize in analyzing the potential issues. You were of course correct to use velocity spectra and time waveform but again phase is the missing element here. Do the relative motion plot with phase and analyze the results. If nothing pops out there I would move to a bump test for structural resonance but again I cannot assume the arrangement or if this is relevant to this case. A picture is worth a thousand words as they say.

Relative phase analysis was conducted but the phase readings didn't highlight any speed related fault like bent shaft, misalignment, imbalance or cocked bearing.

The bump test was not perform. Are you recommend to do the bump test?

Configuration can be imagine by considering the attached picture. it is the same configuration.

Attachments

Photos (1)

Please submit the phase readings as there is little to no possibility that they do not highlight a fault as there definitely is one. If the entire assembly is on the base and the pump vibration levels are as low as I saw from your report, I highly doubt it is structural resonance unless you have concrete or grout poured under the pump but not the motor or some other anomaly. Check the base under the motor for cracks or loose motor bolts to the base. The higher than normal vertical readings are generally a base looseness issue however the high axial is normally attributed to alignment although you are not seeing it on the pump which is strange. Remove the coupling and run the motor by itself and see if the vibration levels remain high and if so replace the motor and inspect the base at that time.

ckyjohn posted:

Please submit the phase readings as there is little to no possibility that they do not highlight a fault as there definitely is one. If the entire assembly is on the base and the pump vibration levels are as low as I saw from your report, I highly doubt it is structural resonance unless you have concrete or grout poured under the pump but not the motor or some other anomaly. Check the base under the motor for cracks or loose motor bolts to the base. The higher than normal vertical readings are generally a base looseness issue however the high axial is normally attributed to alignment although you are not seeing it on the pump which is strange. Remove the coupling and run the motor by itself and see if the vibration levels remain high and if so replace the motor and inspect the base at that time.

Thanks : i will share phase readings to you by Monday & i will definately plan to conduct solo test of motor.

Any further suggestion or recommendations from your side is highly appreciated.

"The bump test was not perform. Are you recommend to do the bump test?"

Yes, of course it is easy to do. Refer to marked image for measurement locations.

Also verify correct shaft end gap as part of a precision shaft alignment.

Conduct sounding test of baseplate to detect voids in grout (if present at all), and install epoxy grout as necessary. 

Search: machine base grout sounding test

https://www.bing.com/search?q=...4702A4E81E3A829E39F3

Pump Baseplates and Installation can impact Pump Reliability

From <https://www.waterworld.com/art...ump-reliability.html>

Walt

Attachments

Photos (1)

Did you correct for the direction of the accelerometer in an axial direction on the diagram? In other words are these corrected values (+-180 to one reading) or as read? Coherence values can be kept out of the diagram (but pay attention while collecting and of course this depends upon averaging as well) and you should instead include vibration amplitude typically in displacement or velocity so there is recognition of where the issue is and where to focus on machine component issues. I am not a big fan of two roaming sensor cross channel phase. Personally I just go into the balance program and use absolute phase with phase tape and a tacho then record the data on a bubble plot with one roaming sensor. I find with this approach it is much easier to analyze. If using cross channel phase I personally would use a single fixed sensor and one roaming sensor and pick a reference direction and correct your axial phase reading (and marked as "corrected" or "as taken"). Ensure the hardware is setup to look at the phase of your running speed frequency (F3 Set Frequency) when you open the analyze program / manual analyze / set analyze mode / cross channel phase on your CSI box.

ckyjohn posted:

Did you correct for the direction of the accelerometer in an axial direction on the diagram? In other words are these corrected values (+-180 to one reading) or as read? Coherence values can be kept out of the diagram (but pay attention while collecting and of course this depends upon averaging as well) and you should instead include vibration amplitude typically in displacement or velocity so there is recognition of where the issue is and where to focus on machine component issues. I am not a big fan of two roaming sensor cross channel phase. Personally I just go into the balance program and use absolute phase with phase tape and a tacho then record the data on a bubble plot with one roaming sensor. I find with this approach it is much easier to analyze. If using cross channel phase I personally would use a single fixed sensor and one roaming sensor and pick a reference direction and correct your axial phase reading (and marked as "corrected" or "as taken"). Ensure the hardware is setup to look at the phase of your running speed frequency (F3 Set Frequency) when you open the analyze program / manual analyze / set analyze mode / cross channel phase on your CSI box.

Yes!! the accelerometer direction is corrected.

No resonance was observed during bump test (reported by my reliability team)

I'll update all the members regarding the root cause, if we find it.

we will plane for the solo test of motor in next week may be.

 

I can't interpret the "phase data" with the way it is presented/displayed. Apparently the bottom numbers are for Coherence, so you must have used two vibration sensors (accelerometers) and a 2-channel analyzer. If this was the method, then please indicate the Reference Location-Direction and the amplitude ratio should be the bottom number. The frequency, such as 1xSS, should also be stated. It is also good to provide the absolute vibration level (average) or spectrum for the Reference measurement. This data is usually called an Operating Deflection Shape (ODS) Test.

While running motor uncoupled may be helpful, it is not the most important test for the situation you have. I still recommend following my suggestions for impact test, sounding test on baseplate, and verifying shaft end gap as part of a precision shaft alignment.

Add Reply

×
×
×
×