I have a 300 HP, 1200 RPM motor driving a Nash vacuum pump. I have rotor bar frequency(0.5 G) no sidebands, with 2nd, 3rd & 4th harmonic. I also have Slot frequency (3.3 G) with 1 X RPM sidebands.
The amplitudes fluctuate with the load.
I don't see any bearing frequencies.
Both frequencies have grown since installation in 2001. Rotor bar from .03 G and slot frequency from 2.4 G.
All amplitudes are higher in vertical than horizontal. Any thoughts on how long I should leave this run?
Thanks Dennis

_2_PPV_PMP.doc (40 Kb, 61 downloads)

That is a pretty unique case as far as I know. I don't know what is supposed to be the defect that causes stator slot pass frequency vibration. Does anyone know? I think there were some discussions maybe about loose stator coils on the old board but I don't immediately see why loose coils would show up at that frequency.

Out of curiosity - Did you see any pole pass patterns around running speed? Any 2*LF?

What is the order number of that stator slot peak? Have you confirmed the number of stator slots independently? It occurs to me (although unlikely) you could have some impacting at 1x with your harmonics peaking in the range of resonant frequency of the component being impacted. (Bearings has resonant frequency in that range although there is no logical reason for impacting at 1x on bearing). TWF might be interesting to get another view of the date and confirm or disprove impacting at 1x.

Yes I have confirmed the number of rotor bars & statot slots with the manufacturer. There are 60 rotor bars & 72 stator slots. The stator slot peak is at 72.004 orders.
2 X LF is .007IPS, no pole pass patterns.
Attached is the TW from the same time. I take them on all bearings. I'm no expert on them, still learning could use any tips you can give me.
Thanks
Dennis

_2_PPV_TW_PMP.doc (60 Kb, 55 downloads)

I see you use E-Monitor like me.

To zoom-in - drag a box with your mouse (left click on one corner, hold down the left-click while you move to other corner, let go of left-click and your're zoomed).

To zoom-out - double click in the lower left corner of the plot (just to the left of 0 sec).

To set sideband cursor. Use "s" key on keyboard to toggle on and off sideband cursor mode. When in sideband cursor mode, click on one peak and then ctl-click on another peak and it should give you a series of markers. Try to line up the markers on the peak to determine spacing between peaks in seconds. The inverse of spacing between peaks is in some ways a frequency (for instance fundamental frequenc of impacting).

I apologize in advance if I'm telling you what you already know (you asked for tips).

With that out of the way, you need to zoom in to get a better view at the spacing between peaks and determine if there is any periodic pattern you can recognize.

When I look at the TWF again I think you have already given enough info to disprove my idea about impacting at 1x. 1x = 1200rpm = 20hz. The period is 1/20hz = 0.05 seconds. The peaks appear much closer than 0.05 seconds.

I'm not sure what you'll see when you zoom in. It may be a complex pattern that doesn't tell you anything. Except that if it is high frequency and generally a smooth curve (not impacting), is probably electromagnetic related. (I have taken TWF's of rbpf pattern and it is sometimes complex but very smooth - I'm not saying this is part of a RBPF pattern, only RBPF is smooth and other smooth complex waveforms may be generated by electromagnetic force as well.) None of the stuff in this paragraph is a hard and fast rule, just my general impression.

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

When I zoom in the frequencies that are repetive are 7144 CPM & 76800 CPM. My resolution is +/- 75 CPM. 7144 could be 2X LF.
76800 = 64 orders which doesn't match 60 rotor bars.
Thanks for your help.
Dennis

_2_PPV_TW_PMP.doc (40 Kb, 20 downloads)

For a wandering theoretical discussion (at least on my part) with no conlcusion see here:
http://www.vtab.se/PHP-NBoard/html/images/materiali/Forum2/HTML/001229.html
(don't forget to click on the page 2 link for more of the same):

Some more comments
http://www.vtab.se/PHP-NBoard/html/images/materiali/Forum2/HTML/000463.html

On the TWF - if you zoom in and it looks choppy, don't forget to consider the effect of your sampling rate... too low can make a smooth waveform look choppy but you can usually tell that problem if you zoom way in to where tryou can see the invidual samples.

I am really guessing you have a smooth TWF.... Because there is nothing at this 2*stator slot pass frequency. So what you have probably looks roughly sinusoidal at SSPS with a little variation up and down in the the peak magnitudes (the 1x sidebands).

I realize that a few of the trainers out there assign significance to SSPF patterns. I have not seen any explanation or case where it means anything.

My personal opinion (take it with a grain of salt) - I think it is electromagnetic in nature and a function of the design of the motor and not to be concenred about. (If it has been there since day one that would tend to reinforce this theory). Being electromagnetic in nature it may be load sensitive - possibly explaining some of your changes. Also like RBPF pattern, the magnitude may be sensitive to very small changes in airgap possibly related to misalignment or foot conditions or bearing conditions. In absense of other symptoms I would not be concerned unless you see a really dramatic change in this frequency (even then I still wouldn't know what it means).

Electrical Apparatus Magazine June 2005 has a long article about stator and rotor slot combinations. Mostly there are a lot of rules developed from experience without explanation.

One of their rules for avoiding torque oscillations is that the number of rotor bars should not be divisible by the number of poles. In your case the rule has been violated ie 60 is divisible by 6.

They also mention the rules are not exact and they have a list of rotor and stator slot combinations which have been found acceptable (by speed of motor) based on experience. 72/60 combination is not present for 6-pole motor.

It may not mean anything. It would be interesting to take a reading in the tangential direction (like at the top of a vertical motor take against a vertical surface take a horizontal reading) and see if anything unusual shows up there which would suggest torque oscillations (probably a longshot). I'm not sure there's anything you can do if you find it. But it might change the way you look at the problem and increase the concern. If do you find anything unusual there let us know.

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

Does this motor have any sisters with same slot combinations and do they show the same pattern?

I have seen (heard) a case where the stator slot/ rotor bar number combination was bad. I have not seen vibration data from it but the problem was a extremely hi noise level since it came down to a design problem plaguing all individuals of that design this unit was placed in a sound damping cage.....
I think the problem was that the rotor could be matched during rotation with no overlapping from rotor to stator and it basically lost all rotating force a number of times per rev and only continued rotating by inertia a little bit. Would not be so fun to start it if it stopped in a spot like that? Can that be possible to make a thing that do that? I am not sure, but I know the guy that sold a refurbed that had that noise problem and the manufacturer quickly changed the design. Olov

Manufacture: toshiba......... not uncommon to see these characteristics IMHO. Without bearing frequencies present, I wouldn't worry a whole lot.

Any changes in environment? Any new installations nearby?

Please provide more info; thanks.

On second look: I don't think you can say you don't have bearing defects present. More detailed spectra are needed.

It would be interesting to get a better resolution to know if that big peak is exactly stator slot pass frequency or exactly rbpf + 2*2*LF. And if it is rbpf+2*2*LF, are those "running speed sidebands" spaced exactly running speed from it or maybe just harmonics of running speed that happen to peak in the same frequency range due to some resonance.

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

DK,
I don't know if this will be helpful in your situation but here is an article on current signature analysis that is interesting. I am not sure but it may have something to do with what OLI referred to. OLI it is the third case study.
http://domino.automation.rockwell.com/applications/gs/r...f/$file/Alford98.pdf

Thanks,
Ronnie

Ronnie - that is a good article on current signature analysis. In the case you mentioned, a 4-spoke rotor spider caused pole pass frequency oscillations in vibration and torque for an 8-pole motor due to the change in flux distribution caused by magnetic permeability of the spokes. It seems like he's saying this can occur if the number of poles is a multiple of the number of spokes. I had never heard of that one before.

To Oli,
The combination of rotor bars and stator slots can cause cusps in the torque speed curve which cause it to dip very low at a certain speed (possibly zero speed in some cases). It that cusp dips below load torque or below zero for an unloaded motor, the motor can fail to accelerate. The reason for this behavior has to do with space harmonic fluxes. The useful flux is sinusoidal in space as we move around the rotor. But to achieve that sinusoidal flux we have to uses slotted stators and rotors. The slots produce steps in the mmf. So if you draw the mmf from the stator around the gap at an instance in time it looks like a stepped sinusoid rather than a smooth sinudoid. That step wave has a lot of higher order space harmonics. Likewise rotor has it's own set of space harmonics in it's mmf. And the rotor and stator permeability also have theri own set of harmonics. When these harmonics combine, the result can be vibration, or backward torque that cancels the main torque or oscillating torque at certain speeds. These effects used to be problems many years ago when they weren't understood. By and large they have been understood for a long time (although not by me), and motor manufacturers use time-proven combinations that generally have no major problems. One place where you run into trouble is when a user requests a rewind shop rewinds a motor for a different speed. The slot combination may have been fine for the old speed but can cause problems for the new speed.

Pete,
Actually it was 8 spokes in an 8 pole motor and it seems it only occurs when the number of spokes equal the number of poles. It is a little confusing when they use an 8 pole motor and show a diagram of 4 spokes. But they do go on to state that it was confirmed to be 8 spokes in the rotor.

Ronnie

Thanks Ronnie. That drawing did confuse me.