There has been a lot of discussion lately on the board about subsynchronous vibration on vertical shaft machines. I have run into something lately that is new to me on a vertical pump. I have many times seen whirl on these machines in the area of 35-45% of RPM. The other day I captured data on an inverter driven vertical pump. We set the inverter so it would run at full speed (1190 RPM) and captured our data. This is a 100 HP motor and was driving a submerged pump via a 30 foot long shaft. The installation is only about 1 year old. The amplitudes were relatively low (all <0.1 ips-peak) but all measurements had 2 large peaks. One of the peaks was at 0.78XRPM and the other at 2XRPM. On the upper bearing of the motor the 2X peak was higher. On the lower motor bearing the 0.78X peak was higher. 1XRPM could not even be identified in many of the spectra captured. On some of the spectra I could see what appears to be the first harmonic of the 0.78X vibration peak.

I was capturing the data only to demonstrate the process to potential customer and not due to any request for analysis, so I may not have the opportunity to do any further testing. If I could I would do a test capturing data on the entire operating envelope and see how things change, but I may not have that opportunity.

Have any of you seen similar things? I know I don't have enough information to really diagnose the source, (pump curves, pressures, flows, etc.) but have you seen whirl at nearly 80% of RPM and have you seen harmonics of whirl?

Skip Hartman

Skip,

I have never seen what you describe.

Did you confirm the speed with a strobe or something? The reason that I ask is that I recently did some work in some boiler id fans that were run by inverters and when the operators said they were at 60 hz they were vibrating at 1075 rpm instead of 1190.

Went to the control room and checked the readout there and it shows 1203 rpm. Checked in the MCC and found the exterior control panel showing 54.8 hz.

Good Luck,

Danny

Danny,

No, I didn't strobe the pump to verify RPM and I thought that incorrect speed could be the issue. The other strong peak in the spectrum was what looked like 2XRPM, however, and if the speed was not 1190 then I have another dilema about where the peak comes from that looks like 2XRPM! You are right. I should have verified the speed with a strobe, and I will try to do so on my next visit.

Thanks for the advice. I wonder how common it is to find incorrect speed readouts on VFDs.

Skip

Skip,
I recently came across some odd data from an inverter driven crane hoist.

In my case the amplitude was high at 1.3 ips. However the frequency didn't coincide with anything.

I did find that while strobing the shaft the coupling bolts would never freeze. This indicated to me that while the shaft was turning it was also oscillating at some other rate. Researching one of the controls system message boards it turns out that it could be caused by the SCR.s, firing rate misadjustment, or phase unbalance. If I remember correctly it may also be above or below running speed. I never did get to the bottom of it as job assignments changed.
I'll look for my old data.

Regards
Spencer

I have also seen high subsynchronous vibration of a vertical pump (1785RPM) with long hollow shaft driving the impeller.

In my case it had occurred at 0.40xRPM. 1xRPM was practically nothing! If I remember correctly, PeakVue showed abnormally high impacting on the accessible pump points. The shaft was supported in radial direction by several sleeve bearings lubricated by the process fluid.

The problem was in the worn out sleeve bearing.
I think this vibration was due to whirling rather then resonance. I guess, whirling takes effect when clearance exceeds certain value.