During a recent opportunity overhaul of a 575 V, 700 hp / 300 rpm vertical pump motor, IGBT VFD operated, it was discovered that the rotor iron laminations package near the non-drive end had rotated on the shaft slightly. It is observed that this rotor does not feature through-bolts to clamp the laminations together. A spider would be the closest description of the stampings ( see 3 photos attached ).

The motor has been in operation for 10 years. Our overhaul shop informs that a slight shifting of the iron laminations package is quite common in large non-clamped design rotors, and there is nothing to worry about. My question, should I trust their re-assurance ?

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

The photos ?

I am trying in vain for the past 1/2 hour or so to get them uploaded. Something is wrong here. Only one pic is showing. Any ideas ?

Here we go >>>

and the third photo >>>

Just thinking through, what problems could arise from torsional shifting of lam plates relative to each other:

Doesn't seem likely to cause unbalance, especially if you haven't seen strange vibration.

The fact that lam's are loose means vibration may be wearing down the interlaminar insulation resistance. Was a core test done for watts per pound and hot spots? That would help you to judge whether it has caused hot spots. At any rate, the rotor watts per pound clearly is not as important on the rotor as on the stator as has been discussed recently. But localized damage causing hot spots during core test might perhaps cause a bow during startup.

It seems like the degree of shifting of the laminations relative to each other would be limited by the rotor bars. If the end pack were completely loose from the shaft and the spider, then the only thing keeping that end pack tied to the shaft/spider is those rotor bars. That could put some unusual shear stresses on the rotor bars and possibly the endring joints. Odds are if things are loose now, the looseness can only get worse over time unless something is done.

Those are the only concerns I can think of. But that's just some thoughts on a problem I don't have any experience with... might be wrong (maybe others have some ideas).

What are the repair options?

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

This lamination stack must have been at one time pretty straight. Somebody must have inserted the bars into the slots. And generally it is much easier when the stack is straight than when it is crooked.
Sometimes there is a key on one of the arms, to keep the lamination in position. Obviously it is not your case. Your lamination sits on the spider and only the friction prevents it from spinning on the spider.
If the main part of the rotor lamination is really tight on the spider, and only those 2 sections have moved, it is an indication that the torque moved the lamination in the direction of the rotation.
Over 10 years the torque acting on the lamination has rotated the lamination on the spider. Since the bars are inserted with certain necessary gap, this gap has been eliminated on one side, and only the bars now prevent the lamination from further movement in those 2 sections.
You do not have to worry about the lamination “shearing” the bars.
What I would look for is a sign of a recent movement of the lamination versus spider. If there is an evidence of such a movement, the interference between the lamination and the spider is not sufficient. It would be a serious problem, but I do not think it is the case.
The movement of the lamination was so miniscule, that we can eliminate another worry: A damaged insulation between the rotor laminations. As I have shown in the thread: “Torque producing force-follow up”, the slip frequency in the rotor is so low that the eddy currents are not an issue and the rotor can be made as well of solid iron. If you want to learn more about the motors with solid iron rotors see: http://www.myoops.org/twocw/mit/NR/rdonlyres/Electrical...A2C6E/0/chapter8.pdf
or:
http://eprints.usq.edu.au/archive/00000551/02/Ahfock_T....Hewitt_A._(2005).pdf
Note: The testing of the rotor lamination for looses in Watts is another motor myth. Nobody can do it, but lots of them claim that they do do it. We have heard about it in recent postings. What a joke!
Since the losses due to slip frequency are so miniscule, the only worry is the rotor surface losses due to slotting. And again, the movement could not have caused any problem. The worry is not a worry at all.
The general conclusion is: If there is no evidence of the recent movement between the rotor laminations and the spider arms, there is no problem at all, and you can have a good, worry free weekend.
jank

That all makes good sense. If you told me it's ok I would take your word for it.

Rotor watts losses, whether they mean much or not, I said was not as important as hot spots, and then only during starting. It is a little bit different than a solid rotor which would heat uniformly during start. But starting time is brief and the losses after that are negligible so your comments make sense.

I'm not worried about the bars shearing. I said there were unusual shear stresses. I was thinking of the bar as a beam, shear force is integrated to give bending moment, which is integrated to give slope, which is integrated to give displacement. My concern would be that this allows the bars to bend and therefore might possibly stress the end ring joint. You can see the bend of the bar slot in the 2nd photo of first post and it might try to bend that end ring joint. Of course, there should be some testing done while the rotor is out to check for cracked joints and if it shows no problems we feel better. (Which test works best... that's another thread.)

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

Ungereng:

The rotor may not be clamped (which it actually looks like it is), but there should still be keys located in the shaft in order to keep the rotor core laminations from spinning on the shaft. In this type of design, there are usually at least two.

The movement of laminations of the rotor in this size vertical machine is actually unusual. It usually results from the load torque/inertia effects between the components of the driver (motor) and driven (pump). I will return to this in just a second, in the meantime, let’s take a look at the issue that you have right now.

First, the movement of the laminations: When the rotor is re-installed, or while it was operating, the additional skew may be causing some additional noise. Almost a transformer noise or whine, on top of the noise that you would hear from the VFD.

Second, the good news is that if you were to look at the air spacing between each section, you will notice that there is additional metal between the laminations of one section and the next (splines). These are not rotor bars, but are spacers that connect one section of laminations to the next in order to keep the rotor together as one large mass. You will want to inspect and see if any of them are damaged or broken as that may affect the mechanical strength of the rotor.

Finally, I would not worry too much about the inter-laminar insulation on the rotor. It is not used as a glue but may have actually allowed some slippage if the laminations were relatively loose on the shaft or the keys, if they are there, do not extend all the way through from the front to end of the core. The keys may be hidden behind the clamps at either end.

So, main issues, some additional noise and losses due to the interaction of the fields between the stator and components of the rotor. I would also suspect that, if this were under full load torque at startup, that you should watch for rotor bar fractures down the road.

However, the only times I have really seen this type of thing occur is on applications where the motor is 'across the line.' The reason has to do with the inertia of the load, the twisting of the shaft (torque) and the torque produced by the rotor system. With the VFD, that should not be occurring, unless you have a pulsating load or the ramp-up time is too fast, or that it starts while the column is falling and the rotor is running backwards when it is started.

Basically, what is happening in the system is that the load has both an inertia and torque based upon the impellor. When the motor starts, this torsional load is transferred through the shaft and is seen primarily on the drive end of the motor shaft. The back end of the motor shaft has a relatively minute inertia and only the bearing system as a load. This generates a twisting in the shaft before the rotor, a different one under the rotor core (and spider, in this case) and a different one at the opposite drive end.

In a vertical pump, the pump starts and the rotor and assembly is ‘pulled down’ (hence the reason for the thrust bearing at the top) and the rotor shaft twists (rotor shaft material is quite elastic). On the load side of the rotor, the twisting is significant. Under the rotor core, it is stiffened and the twisting is not as much and at the top (ODE) the twisting is either near zero, or depends on how much load is being handled by the thrust bearing (friction load).

When the shaft twists, if the laminations (with the exception of the metal clamps at either end, the laminations also make up the spider so that each lamination is touching the shaft) are not held in place with a key or weld, they will skew a little. Depending on the manufacture and stiffness of the system, this may occur very quickly, or may take a long time.

As far as the reliability of the system:

1. Are you seeing a torsional pulsation in the pump?
2. Are the speeds changed often and does the motor come up to speed or change speeds quickly?
3. If the drive is a PWM drive, are you using the ‘s-curve’ feature – normally not done in pumps, but if you are maintaining a column…?
4. Is there any evidence of fractures or bending stresses in the splines or rotor bars?
5. And, this may also be evidence of issues with the pump operation at variable speeds.

Sincerely,
Howard

quote:

motordoc
Finally, I would not worry too much about the inter-laminar insulation on the rotor. It is not used as a glue...

I searched for the logical context for someone to make the negated statement "It is not used as a glue...". The only logical conclusion is that it was directed toward me, since I stated the link between core looseness and degradation of interlaminar insulation. (I stated on 24 August 2007 03:17 PM: "The fact that lam's are loose means vibration may be wearing down the interlaminar insulation resistance")

But I never said anything about lack of glue action of lamination insulation causing core looseness
(lamination insulation degradation => core looseness).

The cause-effect relationship of this link goes in the opposite direction:
(core looseness => lamination insulation degradation )

See "Electrical Insulation for Rotating Machines", Section 10.3.1: "When the laminations in a stator core become loose they can move relative to one another under the influence of mechanical vibration and/or electromagnetic forces, and the insulation on them degrades due to abrasion." (This should apply similarly for the rotor core as it does for the stator core, although the significance of interlaminar insulation degradation is much different for rotor and stator as we have discussed.)

See also http://www.mpinstrumentation.ca/Portfolio_Tad.pdf page 14:
"Shorts between laminations most often occur due to:
1....
2. Insulation loss due to lamination fretting in a loose core."

Epete:

No comments were made about anyone or directed towards anyone.

I didn't even read the other posts.

There is one more possible explanation of the shift. The motor was a subject to a sudden stop in the past (pump failure or something like that). It caused some lamination to slip on the shaft. (Initially I thought there was a spider, but there is none). It could have been those two sections or the rest of the lamination depending on the direction of the rotation. This explanation looks to me at this moment fairly plausible.
I have seen a similar reaction on an 800 hp motor. In that case it was not the rotor lamination that slipped. It was the stator lamination that turned in the housing 30 degrees shutting the power off by ripping off the leads.
jank

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

Ungereng,

Does the motor have an anti-rotation device on top as is typical of many vertical pumps? An abrupt stop at top could cause the rotor to shear (torque from momentum reaction). A mechanical guy has to think of mechanical things first before considering electrial issues!!

Walt

Walt:

Actually, excellent point as I did not consider the deceleration settings of the drive. The twisting of the shaft does not require just an abrupt stop, but can be caused also by the column falling while the motor is still trying to turn if the deceleration is too slow, or the column falling quickly if the deceleration is too fast. Especially if the system was designed without any type of backflow valve because it is operated with a VFD.

Howard

I don't think a VFD has the oomph to oppose a rotating load and create rotor damage like that. The motor would regen into the DC bus, and the bus voltage would elevate until the VFD tripped out on DC Bus Overvoltage.

Wally,

Only if the VFD decelerated too fast.

I still think it is during acceleration.

Howard

Gents,
Why not consider asynchronous transient torques? When motor starts direct-on-line there can be excessive transient peak torque. Also, when motor is disconnected from the mains and immediately thereafter is re-connected to the mains then there can be large transient torque as the motor is re-connected to the mains while having residual stator voltage. These transients can result in deformation of stator or rotor laminations if there is insufficient welding quality, shrink-fit quality, etc. Of course I have read that there is a VFD and the DC bus of a VFD is not powerfull enough to create transient torques like the transients that can occur when operated DOL on a high MVA short-circuit capacity mains.
Now I wonder: did this motor have a pre-life without a VFD? The available info does not provide any time line info! Has this motor once been running DOL and did the deformations occurred long time ago?
If so, you can trust their re-assurance.
Regards,
Arie Mol
NL

Do you have a torsional analysis of the system, including a torsional transient study?

There could be dynamic excition, either steady state excitation of a resonance or transient torsional response. It's always better to be safe.

One can also get a torsional vibration measurement (test in field).

Thank you ALL for your valuable comments and observations on the subject, it is much appreciated.

Sorry, I had said that the laminations shift occured on the NDE, it actually happened on the DE. I did some detective work on the rotor. As you can see from the new photos attached, the black paint coat where the spider and end laminations meet is not broken / cracked. I cofirmed that this is the very first overhaul of this motor since it was put into service about 10 years ago, it had never been disassembled. Thus we are looking at the original paint job applied by the motor manufacturer, and we can safely deduct that the shifted laminations at the DE existed already when the motor left the factory.

The upper photo shows the NDE where the spider is still properly centered over the end stampings. Removing some rust we found the key depicted in the lower photo. The key extends all the way, visible at both ends. Of course, why the partial iron package rotated on the shaft with the key in place we will never find out.

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