Sudden rotor unbalance / TIR for no apparent reason 9

[electricpete]

This is a 2500hp 1800rpm horizontal sleeve bearing motor, driving a single stage hot water pump vis a shim back coupling.

We had a motor that was fine on vibration in August < 0.1ips.
It ran from August until December.

In December we shut it down for 2 days maintenance (filter change, electrical testing) and restarted it (DOL, no vfd). Within 1 day after restart we noticed a very loud noise. Vibration had undergone a dramatic increase.

We did a variety of troubleshooting, ended up swapping the rotor and sending it to repair shop (the position of the motor within the plant does not support rigging out the motor as a unit). With the new rotor, vibration was back to normal.

Repair shop evaluated the removed rotor.

TESTS THAT WERE NORMAL:
Rotor core loop test thermography did not show any hotspots. Injected current thermography did not show any hotspots. Paint blasted from end ring and inspected – no anomalies. Rotor bar tap test indicated all bars very tight (which is not necessarily normal, this rotor has tighter bars than normal).

TESTS THAT WERE ABNORMAL:

Repair shop found significant unbalance on the rotor (“7 mils” unbalance on inboard end is how they described the unbalance – it corresponds to a calculated eccentricity).

They measured TIR with machine supported in lathe at the ends. Found very low TIR on outboard end, TIR at the inboard end started at 0 at the tip of the shaft and slowly increase dto 6 mils at the winding-side lab seal surface (the last machined surface available on that side of the motor, I’ve asked to check TIR further toward the core on non-machined surface). Slide 1 shows the TIRs recorded (note the graphic upon which this info is recorded is a standard form, it does not represent the scale of this rotor).

We checked the rotor cold and hot (heated in over), there was no change in unbalance condition or TIR (of course this does not recreate non-uniform heating of a motor start).

=== ==== ===

These unbalance and TIR measurements are very abnormal and suggest a smoking gun cause of the elevated vibration (which again was fine in August). The question is what caused a sudden abnormal rotor TIR.

There were no unusual plant transients recorded or unusual maintenance done on the motor following the August normal vibration.

The rotor has a fabricated spider with 6 spokes welded to the shaft along the entire length of the spider. Slides 2-4 show the rotor or similar rotors, as labeled. The bars are swaged once in the axial middle of the rotor. The bars are tight in the slot with a copper liner remaining in the slot (that was used to facilitate original bar installation). Slide 5 shows the liner.


Additional tests/inspections planned:
- Check TIR as close to the spider as we can get on the inboard end (to attempt to localize the bend)
- Inspect accessible spider welds

What do you think could be the cause of the sudden appearance of rotor bend?
What additional checks would you suggest?

My objective: I would like to understand the cause well enough that I can have confidence in selecting a rotor repair or replacement strategy that will prevent the problem from recurring. Shop had proposed simply bend the shaft extension using heat to restore the balance, then balance, and check balance / tir repeatability including after heating. That doesn’t give me a lot of confidence the problem will not recur when we reinstall this rotor, considering we don't know why it bent to begin with.


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

 

[xnuke]

I remember from my Navy days that we put large rotating machines on a jacking motor after shutdown which slowly turned them until they cooled down. This was to avoid bending the hot shaft under the weight of the load if it sat in one place while cooling. Any chance that could've happened here?

xnuke
"Live and act within the limit of your knowledge and keep expanding it to the limit of your life." Ayn Rand, Atlas Shrugged.
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[electricpete]

I remember from my Navy days…

Username checks out! (me too).

we put large rotating machines on a jacking motor after shutdown which slowly turned them until they cooled down. This was to avoid bending the hot shaft under the weight of the load if it sat in one place while cooling.

That is typically a practice for long slender turbines subject to low pressure gland steam hitting one side of the shaft while stationary. I’ve never heard of that practice for other machinery.

Any chance that could've happened here?

Fwiw, my opinion is that it's very unlikely for this beefy shaft and spider to deform in 2 days simply under its own weight. But I hadn’t even thought of it before.... now at least it’s on my radar to ponder / explore along with all the other possibilities. I will check out where the space heaters are located.


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

 

[edison123]

I would check for weld cracks in the spiders. Normally, the spider front ends (across the thickness) are not welded.

Muthu

http://www.edison.co.in

 

[electricpete]

Normally, the spider front ends (across the thickness) are not welded.

Thanks Muthu, I didn't know that.
Do you know why they are typically not welded at that location?


I will tell you guys my two theories. The first definitely involves weld cracks.

=== theory 1 ====
The leading theory is residual stresses built into the shaft when welding the spider during initial rotor fabrication. Or equivalently, shaft / spider was assembled and had a bow, then the journals and other areas were machined to make the rotor look straight, and balancing carried out – it still results in residual stress on the shaft. Then what happened recently - maybe some welds cracked and allowed that shaft to respond to that residual stress by bending

=== theory 2 ====
My second theory is much more of a longshot, but I’d like your feedback if you ever heard of anything like it or you think it’s even credible. The second theory is that one or more bars stuck in the slot during / after uneven differential expansion from DOL start, putting the bar in tension or compression, resulting in the core/spider being the opposite (compression or tension) in that one area. Uneven compression or tension of the core / spider around the cirfumference, creates a moment arm (with the distance component being the radius of the spider) which tends to bend the shaft. Remember copper has twice the thermal expansion coefficient of steel AND the copper gets much hotter during start. That’s a lot of relative motion to accommodate. Ideally the bars are anchored by the swage in the center and expand within the slot smoothly outward, and later contract smoothly inward without sticking in the slot. It’s actually easier to imagine them sticking on laminated edges of the slot (like most rotors) than the copper lining in this particular rotor. But who knows, as I mentioned before these bars are very tight in the slot, and maybe there is some corrosion or welding of the interface.

I have two other fuzzy facts 2A and 2B which I haven’t mentioned that also tend to steer me towards this oddball theory 2:

2A – Back in 2003, we had a sister motor with thermally sensitive vibration that suddenly increased dramatically. We called the OEM to assist. They observed teardown, rebuild, uncoupled run said the rotor needed balaning, attempted balancing but they couldn’t (it seemed not repeatable). We sent the whole motor (rotor, stator, frame) back to the OEM and they attempted to troubleshoot and repair. I don’t have a written report (I had only sketchy verbal details, that have faded a little over time), but I know they tried balancing and machining the rotor and the vibration stayed high. They ended up telling us that the rotor was not repairable because it was not swageable, and they had to sell us a new rotor of a new design. The replacement rotor solved the vibration. The prominent mention of need to swage steers me toward thinking about thermal expansion (although it could’ve also been simply loose bars on that original rotor). Also the original design didn't have that copper liner, I wonder why they added that in the new design. Unfortunately I haven't established a good communication link with the OEM this time yet.

2B - When we started up the motor after 2 days maintenance in December 2018, I mentioned we heard an abnormally high noise within one day. What I didn’t mention is that a very experienced electrician (who previously worked many years in a motor shop and still teaches me lots of things about motors) was present AT THE TIME they restarted the motor. He may have stuck around for 5 minutes or so before he decided the motor had no abnormal noises. If the motor had sounded as crazy loud when we started it as it did a day later, I'm sure he of all people would have noticed it. So that makes the story even weirder: what changed between 5 minutes after start and 24 hours later when the noise was heard (with the motor running the whole time). My answer to support theory 2 - maybe the sticking occurred during the bar contraction phase of the transient when the initially hot bars and cool core were coming more into thermal equilibrium.


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

 

[electricpete]

I would check for weld cracks in the spiders. Normally, the spider front ends (across the thickness) are not welded

Are you saying:

A - the flat side of the block facing center toward the shaft is not welded since it is not accessible once the block is in position. So only the accessible corners at interface of bar and shaft are welded - along two long long sides and two short sides per block.

or

B - the accessible corners at interface of bar and shaft are welded only along the two long sides (the length of the rotor), but not at the two short sides each end.


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

 

[edison123]

pete - B.

Muthu

http://www.edison.co.in

 

[electricpete]

thanks. do you know the reason why the short sides usually aren't welded?

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

 

[edison123]

Torsional stress failures in that plane. Thanks to this OEM fumble, we are doing reshafting of a few 1250 KW motors.

Muthu

http://www.edison.co.in

 

[desertfox]

Hi electricpete

Relating to your theory 1, I doubt that the spider fabrication would have just been machined after welding, I feel sure they would have stress relieved the fabrication and then put a finishing cut onto the machined areas to minimise creating residual stress.
I do quite a lot of heavy fabrication design and we always stress relieve the fabrications before
Putting a finishing cut on rough machined areas. I found this site which gives stuff that can cause motor failure, for instance if there is corrosion on the shaft this could create an imbalance over a period of time.

http://www.uesystems.com/news/the-50-failure-modes-of-electric-motors

“Do not worry about your problems with mathematics, I assure you mine are far greater.” Albert Einstein

 

[electricpete]

thanks desertfox.

the rotor is accessible for inspection. it had expected / normal amount of dirt and corrosion in my view, nothing excessive.

more importantly, the TIR is much more than normal and so I link it to the unbalance (my assumption is the bow/TIR caused the unbalance). Neither corrosion not anything on the link would cause a bow. Any further thoughts?

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

 

[desertfox]

Hi electricpete

Well I guess I am stabbing in the dark lol, only other thing I can think of would be a sudden overload on the shaft itself which from what you are saying is unlikely.
Going back to non-uniform thermal expansion is it possible that the shaft was prevent from growing longitudinally along its axis, it would then be possible to put the rotor in compression and buckle it. Slightly related I once saw a motor operated butterfly valve jam because the valve operating shaft couldn't expand during service.

“Do not worry about your problems with mathematics, I assure you mine are far greater.” Albert Einstein

 

[crshears]

To me, TIR is a TLA for Technical Interface Requirements, something it clearly isn't in this instance...so just what is TIR in this context?

CR

"As iron sharpens iron, so one person sharpens another." [Proverbs 27:17, NIV]

 

[Hoxton]

Total Indicated Reading

This means if you are using a dial type run out indicator it is the difference between maximum and minimum indications; ie the diameter of the run out.

 

[crshears]

Gotcha,Hoxton; thanx.

CR

"As iron sharpens iron, so one person sharpens another." [Proverbs 27:17, NIV]

 

[panter]

E-pete,
maybe it is possible that the shaft has been bent due to some impact during dismantling or transport rotor . Maybe it's a little crazy considering the gauges and diameters of the shaft , but I see some signs of a potential impact at the spider .
Good luck !

 

[Compositepro]

If there are large temperature differences caused by not rotating the rotor during cooling it can lead to temporary distortion of massively stiff parts. This could lead to some movement between press-fit or clamped parts which involves a lot of friction. If motion occurs in a press fit it will not return to the same shape due to the hysteresis caused by friction. Just conjecture.

 

[rhatcher]

This size motor with sleeve bearings should have received an alignment before startup, there should be an alignment report that would have indicated a bent shaft before start-up had the motor been damaged during transport or installation. Having said that, it's hard to imagine the shaft being bent in service under load, my experience is that a high torsional load would shear the shaft. Is there any way that the shaft could have taken a large hit after alignment but before startup?

 

[Strong]

"What do you think could be the cause of the sudden appearance of rotor bend?
What additional checks would you suggest?"

What did noise sound like, how long did it last, and did sound event occur at same time of vibration event?

Inspect both bearings for Babbitt condition and loose or crack liner or housing.


Inspect shaft coupling for cracked discs and sheared bolts.

Inspect for rotor contact with stator.

Inspect for loose/missing balance weights.

Conduct Growler test for cracked rotor bar (or any rotor electrical circuit fault)

Conduct rotor natural frequency (bending modes) test to identify possible cracked shaft or change in NF caused by rotor structure change.

I did similar tests on a 4500 hp 3600 rpm steam generator feed pump motor (Allis Chalmers), but it was a long time ago. That rotor had natural frequency (balance resonance) slightly above 1/2 x shaft speed. Your rotor may have 1st critical slightly above 1xSS and migrated down to 1xSS by loose or cracked rotor structure/shaft.

Walt

 

[waross]

From the first post:

We had a motor that was fine on vibration in August < 0.1ips.
It ran from August until December.

In December we shut it down for 2 days maintenance (filter change, electrical testing) and restarted it (DOL, no vfd). Within 1 day after restart we noticed a very loud noise. Vibration had undergone a dramatic increase.

Mechanical impact damage:
I would expect that it would take an impact that is large in relation to the motor to bend a shaft.
Unless someone dropped a couple of tons onto the coupling we may rule out impact damage.
Transport damage. The motor was not transported. It sat for two days.
Heat damage:

This is a 2500hp 1800rpm horizontal sleeve bearing motor, driving a single stage hot water pump vis a shim back coupling.

How hot is the water?
Is the coupling conducive to transmitting heat to the motor shaft?
Is it possible that heat from the water kept the shaft hot while the rotor was cooling?
Is it possible that the shaft became even hotter from conducted heat when it was stationary?
Such a condition could lead to slippage between the shaft and the spiders that was not reversible when the shaft cooled.
Turning gear may not prevent such a condition.
The solution may be to remove the heat.
Possibly by draining the pump or changing the coupling to a type that does not conduct heat.




Bill
--------------------
"Why not the best?"
Jimmy Carter