We have 5 identical pumps running side by side all at the same speed, and I know what you're going to say... "Their vibration frequencies are modulating each other!"

Unfortunately I don't think that's case (but I could still be wrong). We've only noticed on one of the motors that there is this relatively long beat frequency. See attached. Bearing faults aren't lining up, rotor bar and stator slot frequencies aren't lining up either.

The motor is a 400HP, GE L5011VP, 460Volts, 1180 RPM.

It is driven by a VFD. Turning speed on the attached TWF and spectrum is 1080 RPM.

Could this be an electrically induced vibration?

We've already rebuilt two of the five motors and found that the ODE bearing had fluting from arcing. We spec'd the rebuilt motors to have ground rings and insulated bearings. The remaining three motors (one of which is the one with the beat freq.) do not have ground rings.

Thanks for any help!

Attachments

Photos (2)
Original Post

More information is needed!

What is the period between beats (seconds)?

Is 1080 rpm the exact shaft speed by measurement? Also need the other pumps exact speed to evaluate beating between machines (why guess?).

The spectrum shows modulation sidebands at 18 Hz (1xSS) and 120 Hz (2x line freq.)

A high resolution low frequency spectrum of motor current and/or vibrations can confirm a motor rotor electrical fault.

Walt

Hi Walt!

Period of beat is 3.473 S (0.29Hz). Looking closer at a velocity spectrum I can see that the actual speed is 17.75Hz (1065 RPM). See attached for some more spectra.

Attachments

Photos (2)
ThomasAH posted:

I know what you're going to say... "Their vibration frequencies are modulating each other!"

 

Exactly. With additional comment that the mentioned pump seems to be more resonant prone than the others (attachment, base plate, concrete, srews ....). Can you measure the pump while the others turned off?

Last edited by Becar

If you measure the speed of one or more other pumps and find the speed difference of 0.29 Hz or 17.4 rpm with the pump in question, then there is a beating effect between the vibrations of the two machines. You may not be able to turn off the machines, but you could bump the speed up or down with the VFD and see what the change in beat frequency is. Typically beating vibrations are more noticeable when the vibrations of each machine have similar amplitude. A resonant structure can also amplify vibrations from both machines. A high resolution low frequency spectrum should be able to split the two shaft speed frequencies apart and correlate with each machine speed. The beat frequency may also occur at the pump pulsation frequency where the speed difference is divided by the number of vanes or blades.

You could also calculate the electrical rotor bar fault frequency based on slip rpm times the number of poles. Where the slip rpm is equal to the VFD frequency minus the actual shaft speed.
Walt

Add Reply

Post
×
×
×
×
Link copied to your clipboard.
×