10P-1000HP - Shaft Hunting 1

[macmckim]

Good Day

Have issue with a 1000HP, 10P (713 RPM) 4kv / 60HZ, TEAAC compressor motor - VFD powered - Variable torque application.

Motor is direct coupled (flexible coupling) to VRU ring compressor and is VFD power from 660 - 740 RPM (fully loaded)
The motor shaft starts hunting - movement is over 6.5mm. Hopefully you will be able to see attachment (short video)
The compressor shaft does not have any movement, only the motor shaft/coupling is hunting.
Shaft hunting starts at 660 RPM - does not happen below this speed.

When motor/coupling are disengaged from compressor (no load) - motor is tested at operating speeds, the motor shaft does not move
Vibration levels, bearing RTD temps, full load current are all in acceptable range.
There are 3 compressor motors - two operate correctly and third is hunting.

Any idea's

thanks
Mac

 

[waross]

Check the travel limits in the coupling. The motor shaft may be "bouncing" against the travel limit of the coupling.
The coupling may be slightly out of position on the shaft.

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[itsmoked]

So that's JUST the motor shaft doing that into the coupling and the compressor shaft is not doing it?

Keith Cress
kcress -

http://www.flaminsystems.com

 

[EdStainless]

So if you only slow to 665 rpm it doesn't happen?
Odd motor harmonic?

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[macmckim]

itsmoked: Yes only the motor shaft / coupling is moving - the compressor shaft/coupling is not moving.

EdStainless: Yes, below the 660 RPM and slower, the shaft/coupling is not moving back and forth.

 

[dArsonval]

The black line marked on the shaft... do we know if that line represents the rotor's
magnetic center to the stator?

Or asked another way... do you know where the shaft belongs when the motor is
running decoupled and free to find magnetic center?

John

 

[macmckim]

John;

The motor shaft does have a magnetic center groove in shaft. I cannot say what the black line in the video is.

Yes the magnetic center is known - and when uncoupled the motor shaft is in correct magnetic center position during no load operation.

Mac

 

[waross]

When coupled, there should be equal travel in the flexible coupling in either direction from the magnetic center.

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[electricpete]

< Yes the magnetic center is known - and when uncoupled the motor shaft is in correct magnetic center position during no load operation.


Yes, motors generally seek magnetic center when uncoupled. The question is: are things set up so that it runs near that same place when coupled? The centering force is highest when motor is near magnetic center and it grows lower as you move further away. So your best bet to avoid oscillations (but still not foolproof) is to couple it near magnetic center as John alluded. And also it should not be too close to either mechanical limit of travel as Bill mentioned.

Perhaps you have small movement within the clearances of your driven equipment (maybe 0.010" clearance) which excites large movement of the spring-mass system involving the coupling (axial spring) and the motor rotor mass at its resonant frequency (which looks to be a few hz). We have something like that which occurs repeatedly at certain operating conditions of a double suction pump at our plant, but luckily we do not operate long in those conditions or else we would be concerned about coupling fatigue.

What type of coupling? It looks like you have Bently Nevada prox probes, do they show anything unusual? Did this machine always act this way?



=====================================
(2B)+(2B)' ?

 

[macmckim]

Hello Pete

Per client the coupling is a "flexible coupling".
Yes the motor has BN prox. probes. Probes do not show any abnormal vibration, ever shaft movement.

Client's comments:
Motor #1 has total running hours around 15000, other two motor #2 & #3 have running hours around 19000.
Below are few test we have carried out to see cause of hunting on motor#1
Uncoupled run with speed range from 90% to 100%, no shaft hunting observed.
VFD flux current monitoring during uncoupled run to see any significant spike at flux current at the VFD, flux current observed was constant through out the no-load run up.
Coupling inspection: no any sign of coupling damage
Addition to above test, during normal operation we have not seen any high vibration or bearing temp rise issues.
No any sign of damage to compressor components.


My only idea at this time, its a bit of a long shot, somehow the rotor has moved on the shaft 5-7 mm.

Mac

 

[waross]

Check that the magnetic center mark is still at the magnetic center.


--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[electricpete]

There are no doubt a lot of causes of hunting. Verifying that the motor is coupled up to run at the same place that is magnetic center (as determined during uncoupled run) is the first obvious step.

I'll just talk some more about the particular case I mentioned above (since that's the only one I have experience with). It is a double suction pump with relatively balanced axial thrust (near zero). That makes a bad situation where under certain flow conditions, minor changes in impeller position can cause reversing of the thrust so the impeller jumps to the opposite side of the clearance, which in turn again changes the thrust and the impeller jumps back and forth within its thrust bearing clearance. Of course pump clearance is only 0.010" so you don't see it visually (you might see it on axial thrust prox indications or pump axial casing vib readings). But it creates a broadband excitation which exites the natural resonant frequency of the mass spring system consisting of coupling spring / motor rotor mass... which is very obvious visually (in our case the motor shaft shuttles axially about a quarter of an inch, I'm not sure whether that motion is being limited by endfloat which is 0.5" total). Attached is a more detailed discussion of that mechanism for pumps. I doubt the exact same explanation applies to compressors but maybe you can find out if the compressor operates near zero axial thrust, or take some axial vib measurements on the pump compressor.

=====================================
(2B)+(2B)' ?

 

[waross]

It could also be an irregularity in the ring casting or the impeller.

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[TugboatEng]

I too have seen double suction pumps oscillate back and forth. These were fire pumps and runtime was minimal so it never had a chance to escalate in to trouble.

I do see in gearboxes that may experience freewheeling in certain conditions, spring packs are used in the thrust bearing to favor load on one side.

 

[edison123]

Check the motor bearings are not askew horizontally.

Couple the rotor to the compressor in the motor axial magnetic center (by moving the stator/motor). Most likely the rotor hunting goes away.

Muthu

http://www.edison.co.in

 

[Tmoose]

Sometimes (often?) flexible couplings have "limited end float" features when used with a sleeve bearing motor. I believed that is to use the driven equipment's thrust bearing to prevent the rotor being able to rub the end stops.

I think the only time a coupling needs large axial travel is when both driver and driven have their own thrust bearings, and the axial thermal growth would start a fight.

 

[Gr8blu]

macmckim: You never did mention whether the travel is all one way, or whether it is some in each direction (toward/away from compressor). From what I can see in the video, it looks like the shaft (and thus motor rotor) is trying to move AWAY from the compressor.
1) This is highly indicative of a misalignment during installation: as in the fitters "pulled" the rotor toward the compressor to finish up the coupling bolt tightening. Even if the intent was to have the rotor operate on magnetic center, it wouldn't be in that case. The magnetic centering force is trying to pull the rotor back into alignment with the stator, and the spring component of the flexible coupling is pulling it back toward the compressor. The problem is not all that uncommon, either, since the sleeve-type bearings usually have far more range of movement compared to a typical rolling element bearing arrangement.
2) Did you ever check whether the coupling halves are correctly aligned? A bit of torsion resulting from misalignment (usually face-to-face, rather than center-to-center) there could well introduce a "hunting" response in the motor rotor as the coupling tries to "flex" to compensate.
3) You do mention having proximity probes (Bently Nevada). Are they X-Y oriented, or axial, or multi-planar? What I'm getting at is that a "standard" X-Y probe arrangement cannot tell you anything about axial movement unless the shaft moves far enough to take itself completely out of the range of the probe sensing. In most cases, this is not going to ever happen without some other catastrophic event.
3) If you can measure 6.5 mm (roughly 0.250 inch) axial travel, what you really should have installed is a "limited endfloat" coupling, not a flexible one. This is because there is relatively little resistance to axial movement in a sleeve bearing operating at speed. As a rough approximation, the axial centering force for a machine of this rating (1000 HP at 713 rpm) is going to be on the order of a few hundred pounds - which means you could just lean on the end of the shaft when it's running and easily move it. It doesn't take much from the drive train to reach this kind of force - and if you run back into the shaft/bearing shoulder, you could just bounce off and start a continuous "hunting" cycle as the rotor skids from one end of the allowable travel to the other. When installed correctly, the "limited endfloat" coupling design prevents the shaft from moving far enough to start this "bump and run" condition.

Converting energy to motion for more than half a century

 

[macmckim]

Muthu, Tmoose & Gr9blu

thankyou for comments and questions.

The motor is driving a VRU (Vapor Recovery Unit) Ring Compressor in a refinery. The refinery staff / team has large number of electrical, mechanical and reliability engineer.
They know how to set up motor's alignment / magnetic center & coupling. There are 3 units and two have been operating without issue, third is hunting.

The refinery team (millwright, mechanical and reliability) confirmed that they have motor alignment and coupling installation done correctly. I have not been to site, as units are few thousand miles away.
The BN prox. probes are X-Y axes for vibration - not shaft hunting movement. Per client, there are no vibration issues.

The motor mfg is blaming it on the compressor (client has confirmed that compressor shaft/compressor coupling is not moving)
We estimated this oscillation was occurred because of some exciting frequency from compressor

The refinery team disagrees. So my saga continues.

 

[Gr8blu]

mamckim: Thanks for the quick response. Was just looking at the video again. Is there ANY chance that there may be a blockage or other eccentricity in the internal air flow path within the induction machine itself? For example, having one of the two fans within the motor installed backward? The change in internal air pressure could theoretically be enough to apply sufficient force to start the rotor moving axially - and as it moves, it develops a "spring effect" as the rotor acts like a piston in a cylinder alternately increasing and decreasing the pressure at each end. I admit - I have not seen it on something moving this slowly .. but have seen it on higher speed designs (3000-9000 rpm range).
I also reread your initial post: as this is a totally enclosed (air-air) design, it is possible the internal (rotor mounted) fans are fighting the air movement from the air-air assembly and setting up a similar "pump action" cycle from the ventilation scheme. This could be exacerbated if the airflow within the motor frame was unidirectional (one end to the other) and not double-end vent (in at both ends and out at mid-frame).
As to the motor manufacturer stance: if the machine does not move axially when uncoupled, it does NOT mean the manufacturer is completely off the hook. It just means that at the unloaded condition, there isn't enough force generated to make the shaft move. If the rotor or stator has skewed slots, it might be enough under higher current conditions to cause the hunting effect. Other "magnetic" dissymmetry within the machine can also be causes for relative rotor movement.

Converting energy to motion for more than half a century

 

[waross]

I have been there so many times.

The refinery team (millwright, mechanical and reliability) confirmed that they have motor alignment and coupling installation done correctly

To me this means two things:
The refinery team believe that the coupling is aligned properly.
The refinery team will resist any effort to verify the alignment.
If they are missing a crucial step in the alignment, they may have to live with this problem until the end of time.

Anecdote alert.
I was working on a control panel while a small crew was changing out a 5 HP, belt drive motor.
They seemed to know what they were doing, but when they started the motor the belt wobbled.
They shut down and found that the motor was out of alignment.
They re-aligned the motor and tried again,
The belt wobbled and the motor went out of alignment.
At one point they fabricated a tapered base plate to try to correct the alignment. No Joy.
After two and a half days, they discovered that the pulley was not square on the motor shaft.
This was a good crew who knew what they were doing but had missed one little step.

--------------------
Ohm's law
Not just a good idea;
It's the LAW!