This is 200hp 4-pole motor running at 1780rpm.

There was some slight abnormal noise reported on the motor last month. This month we shut down for work on the pump (no work on motor). Realigned motor and restarted and got the attached pattern which we had not seen before.

Magnitudes are low. The center frequencies of the clusters in orders are: 1.000,2.000,3.000,4.043(2LF),5.04,6.04,7.04,8.08(4LF),9.04,10.04

All of these clusters (except 1.000) have multiple pole pass sidebands (80cpm) around the center frequency.

Two possibilities occur:
1 – Electrical vibration. We have 2*LF and 4*LF and what might be 1x sidebands around 2*LF?
2 – Bearing problem. I haven’t been able to identify any bearing patterns here

Will post some more data and get current signature analysis when we get the time.

Any thoughts from the data so far?

quote:

Originally posted by electricpete:

The center frequencies of the clusters in orders are: 1.000,2.000,3.000,4.043(2LF),5.04,6.04,7.04,8.08(4LF),9.04,10.04

All of these clusters (except 1.000) have multiple pole pass sidebands (80cpm) around the center frequency.

Pete,

The fact that all clusters except 1x have PPF won't overturn the diagnosis that this might be an electrically induced vibration. ( BTW, if you switch to a Log scale you may see PPF SBs around 1x as well ). Is it possible that while working on the pump they have shifted the CL (preloaded) of the motor rotor?

An important thing is this pattern:
1.000,2.000,3.000,4.043(2LF),5.04,6.04,7.04,8.08(4LF),9.04,10.04

You have centered the cursor on the highest amplitude within a cluster, thus making 5.04, 6.04, 7.04 look strange. Instead, I suggest positioning the cursor on 4.00, 5.00, 6.00, 7.00, 8.00, 9.00, 10.00. The sidebands will be higher in this case then respective 1x multiple, so what ??? (Of course, 4.043 and 8.08 are LF related).

Yes, the pole pass sidebands are 0.04 orders, so what I called the center frequency could be the upper sideband. If it is, it is still a strange pattern... haven't seen that before.

Yes, there are teensy weensy pole pass sidebands visible around 1x on a log scale.

Isn't this pattern typical (except for higher SBs) for rotor bar problems or dynamic eccentricity? If so, taking el. current signature will be warranted now in order to confirm presence of PP around 60HZ harmonics.

Those two are probably the most likely scenario's in my book but still need the current data to evaluate further. I'll post the results when I get them.

E.P. & Dave,

By chance do you know the number or rotor bars and or the number of slots in the stator?

It sounds like you are heading in the right direction suspecting either rotor bar, eccenticity or perhaps stator mechanical/electrical. These would be easily picked up with MCSA.

At 1780 RPM, your RS would be 29.66666 hz. and one would find the rotor bar peak at 58.666666 hz.

The center frequency for eccentricity would be RS * bars, then plus/minus 60 and plus/minus 180 for your 4 spectrum peaks.

CF for stator mechanical would be RS * slots. plus/minus 60 (and no multiples), then if you see RS around your plus and minus 60, would be an electrical stator problem indicator.

And please post the results and mention the equipment used, if possible.

Thanks

For current signatures we use our Entek datapac along with a clamp-on probe. Long ago we tried to look at the high frequency stuff and there was a dynamic range problem. All the activity down there is in the last 1 or 2 bits so you can't see anything. So all we look at is pole pass sidebands around line frequency.

I need to get my butt in gear and take another look at that.

Has anyone used an Entek datapack for current analysis to get higher frequencies around RBPF? What kind of clamp-on and what kind of collection setup?

This message has been edited. Last edited by: electricpete,

Pete
The few I have seen like that have been rotor
problems.

I have a case study I did a few years back hid some where in this PC.
I will send it to you when I find it.

Here was current signature analysis we took in October 06. It's the same pattern we had previously in July 06.

Pole pass sidebands are very low (~65db). I did notice there are a lot of sidebands... does that tell me anything?

Back in October we did do some checks on the motor including endplay. There was around 40 mils endplay easily obtainable by hand, with additional 10-15 mils (compression of washer) obtainable using prybar. We contacted the OEM who said this was exactly normal/expected for this particuar motor (although that seems to me like a weird setup).

We have some uncoupled run data. I'm looking through that and will post soon.

Any ideas on this motor?

Pete -

Horizontal config???

Sleeve bearings??? Have you looked at the possibility that during the realign the motor was pulled out of its 'electrical/magnetic center'?

Were soft foot checks (parallel and angular)performed during the alignment? Base/Foundation flatness? Any reason that any of this might have changed during the alignment? Old shims reused / discarded?

This message has been edited. Last edited by: Jamie Dugan,

It's a horizontal 200 hp 1800 rpm ball-bearing open-drip-proof motor.

We did do soft foot checks.

I don't think the condition appeared during an alignment. But we did check alignment and uncouple for troubleshooting of the vibration.

Here's a weird thing about the uncoupled run. During the uncoupled run we have a similar pattern - still harmonics of running speed and highest vibration at 2*LF. But we also have sidebands during uncoupled run except those were spaced at 0.16 orders during the uncoupled run (they are consistently 0.04=pole pass during normal loaded operation). I'll post that uncoupled spectrum tomorrow.

Here is the uncoupled run I mentioned. The overall pattern is the roughly the same as when running (harmonics 2LF and harmonics of running speed) except a little higher magnitude and the sideband spacing is different (0.16 orders uncoupled vs 0.04=pp when loaded). Weird about those 0.16 orders sidebands, huh?

At least the uncoupled run rules out shaft misalignment for sure.

I was wrong about some details of the history.

The 2LF/4LF have always been there, but at lower magnitudes. They had a sharp increase in Spring 06. Exactly what was done at that time I have to review some more but it did include alignment. Also if I get time I'm going to try to see if the sidebands were there before that and also compare to the sister machines.

I was wrong about the soft foot check. There was some change in vibration when adjusting feet. I'll try to find more details on that (a very important detail).

A small apology (excuse) for not having the history - This has unfolded over a long time with a lot of people involved but I didn't follow it closely. Now I'm trying to decide if this needs any further action and I'm wishing I had followed it more closely.

A lot of homework for me.

In the meantime, any thoughts in general on this?

This message has been edited. Last edited by: electricpete,

Hey EP,
Do believe this is electrically generated. What about the stator being loose in the motor frame? Take a reading at the motor frame midpoint and see if it is higher that your bearing cap readings. I am assuming you checked to see if this energy immediately disappated on pulling the power. Also, alway remember that the electrical vibration at 2x lf is almost always linear to voltage and not alleviated by a reduction in load.

A quick review of what was said above:
200 hp 1800 rpm horizontal motor. The symtpoms on that particular motor were pole-pass sidebands while running which changed to 300 cpm while uncoupled. Current signature showed no significant sidebands.

Endplay check as discussed 10 January 2007 10:39 AM "Back in October we did do some checks on the motor including endplay. There was around 40 mils endplay easily obtainable by hand, with additional 10-15 mils (compression of washer) obtainable using prybar. We contacted the OEM who said this was exactly normal/expected for this particuar motor (although that seems to me like a weird setup)"

Update:
Motor is still running October 2008 with same symptoms. I never really thought that the endplay could be related to those symptoms, especially the pole pass.

But now considering comments in another thread
http://maintenanceforums.com/eve/forums/a/tpc/f/3751089...481065893#7481065893
it seems easy to imagine the uncoupled 300 cpm sidebands were due to endplay.
My question: Is there any way the endplay could somehow be connected to the pole pass sidebands?

quote:

Originally posted by electricpete:
My question: Is there any way the endplay could somehow be connected to the pole pass sidebands?

If you meant 0.16xRPM sidebands instead of PP sidebands, then an answer to this question, I believe, has been floating around before.

The followng mechanism sounds logical to me.

If, for some reason, the shaft is moving periodically in axial direction, then, due to the "bath tub" shape of the bearing cross section, a small vertical motion component ( maybe around 3 mil ) will also appear at the same frequency. This creates an effect of "quasy" dynamic eccentricity which, in turn, leads to modulation by this frequency of the running speed harmonics due to unbalanced magnetic pull.

Most of the time a coupling will supress axial motion. I am not sure as to what is the excitation force . Even if 300 CPM is the resonance, what is exciting this resonance?

It would be nice to verify presence of axial and vertical motion of the shaft at this frequency in the uncoupled state with a proximity probe. If not available, a seismic device also could be used with proper acquisition set up since we are talking of Fmax = 500-600 CPM and fundamental signal at 200-300 CPM. An accelerometer has a big roll-off in this range.

David

This message has been edited. Last edited by: David_G,

I think your motor may have an open rotor. You said it had a strange noise to it correct? Let me ask you, the noise disappears when uncoupled right? Uncoupled you won't see the same sidebands because of the load. I've attached one case I've experienced myself.
Sidebands (slip times number of poles) of running speed harmonics and we found an open rotor.

This message has been edited. Last edited by: THMotorMan,

6701019893_Sample_case.pdf (68 Kb, 16 downloads)

Thanks for the input David and THMM

THMM - I can't read your attachment. Seems like a zip file but none of the files contained is recognizeable as a word file. We have done current signature analysis under load as posted above. Wouldn't that rule out open/cracked rotor bar?

Converted to pdf. Try again.

Also, did you check rotor pass frequency?

Check your amp draw on the motor when fully loaded. If there is any severe variation, likely an open rotor. The sample case I referred to had two legs draw equal amps, and the third leg was 5 amps higher. The amperage would fluctuate up and down between the high and lows. Also, the amp draw was above nameplate for the motor and ran consistently in the service factor of the motor.

This message has been edited. Last edited by: THMotorMan,