impeller nuts coming off 3

[ardeIII]

On several centrifugal end suction pump of varying manufacturers, design conditions, etc., have recent problem of impeller nuts coming off shaft, followed by impeller, with associated damage to casing, impeller, shaft, etc.

Nuts are threaded opposite normal pump rotation direction and have retention devices (typically set screws). Check valves have been inspected and pumps are not known to run in reverse. Metallurgies are consistent ruling out loosening due to differences in thermal growth (several of the pumps are high temp-700 deg F range). Have looked for recommendations from OEMS on nut torquing values with common response being "tight is tight".

Compounding this problem, after couple of these events on one of the subject pumps, we modified the impeller nut to incorporate additional locking screws. Next time pump failed the shaft broke at the shoulder at the base of the thread relief.

Anyone seen this problem before? Any recommendations on nut torquing and locking devices?

 

[jet1749]

Sounds like vibration loosening the nut off, especially if the shaft subsequently snapped at the thread relief. Any cavitation or out of balance? try a vibe check whilst the pumps are running.

 

[Artisi]

Certainly sounds like an aplication problems, ie, operating away from BEP resulting in excessive radial load, poor inlet configuration resulting in uneven loading on the impeller, cavitation causing fretting of the impeller, entrained air causing uneven loading.

Have you checked the impeller/shaft fit- possibily excesseive clearance resulting in freting of the impeller.

 

[rmw]

Sounds to me like someone may have "bump checked" the motor rotation and loosened the nut which then came completely loose later in service.

Impellers with nuts can't be bump checked for rotation. If it's right it's good; if it's wrong it's hell.

rmw

 

[Artisi]

Impellers with locking nuts can be bumped started - that's why you have a locking nut so that the impeller won't unscrew if bumped in the wrong direction.

 

[Artisi]

The previous post refers to screw on impellers with locking nuts or set screws.
If keyed on impellers - the locking screw or nut cannot unscrew either as there is no torque applied to the locking device.

 

[rzrbk]

It sounds like you need to look somewhere besides the locking nut. Some of the items mentioned in the above posts are excellent places to start.

Your initial post implies that this problem recently appeared after a significant run with no problems, and on several different pump types and services. This implies that something has changed, be it process conditions, fluid properties, Maintenance practices, parts suppliers,or Operation practices.

 

[JJPellin]

We had a pair of API overhung pumps that had a problem with the impeller nuts coming off. The pumps in question had keyed impellers and nuts that tighten with rotation. To make it even more confusing, both pumps run all the time. They are never shut off unless they fail. This provided even less possible explanation. In the end, all we did was change the way we operated the system. They had been running at a lower flow with both pumps running in parallel. We increased the flow, closer to BEP and the problem went away. So it almost had to be hydraulic in nature. Perhaps suction recirculation cavitation can unscrew a nut. Otherwise, I am not able to explain why the problem went away.

 

[Pumpindex]

Some of these problems are the ones that leave you with your mouth open. I doubt that if will be vibration, since it would have been an original problem start with. I would recommend doing something similar to what the last poster described, and at the end, you will not have any explanation to the problem either.

http://www.pumpindex.com/

 

[ccfowler]

This is a dandy mystery. I would want to consider the possibility of some flow instabilities introduced near the pump suction. Are there any nearby valves, elbows, tees, ...?

Another thought. Are these pumps powered by adjustable speed drives (adjustable frequency drive with an induction motor)? Some of these drives can produce very nasty, fairly high frequency torsional shock loadings. Some early drives were bad enough to break motor or pump/fan shafts (usually fatigue failures, but there were some outright failures with high inertia driven loads). Has there been any indication of relative motion between the pump impellers and their shafts (wear on keys, keyways, shaft surface, or impeller bore)? This could account for the nuts being either loosened to tightened from torsional vibrations induced by the drive.

 

[Tmoose]

"Recent problem" - what changed? Does one tech repair them all?

What type of coupling now, and before?
How tight are the nuts being tightened?
The shaft and impeller and keys all must fit properly.
"Looks good" is just about meaningless for determining fits.
If the torque is indeed proven appropriate then Loctite and belleville spring washers come to mind.

 

[pumpking]

Strange one, when i was apprentice with LaBour pumps, 95% of our pumps had keyway driven Impellers (parallel bore) secured with locknut.

I am not aware of locking nuts EVER having fell off, however, we did used to put some 3mm slots in the nuts to allow us to 'peen' surrounding metal with centre punch to cause slight interference just in case - in 14 years, never heard of such a problem, so assumed it worked just fine !! Hope this helps ?

 

[djack77494]

You're real sure that the rotation is correct?

 

[Artisi]

As there is not torque being applied to the locking-nut, wrong direction can not unscrew it.

 

[desertfox]

Hi bigjohnson

If the pumps haven't been run the wrong way which is what appears from your post then is it possible that the thrust
from the impeller is undoing the nut and in fact when you
improved the locking of the nut the shaft itself failed.
Why not have the pump shaft analysed did the shaft fail in torsion or (tension along the principle planes) caused by the torsional load. Have a look at the pump shaft failure if the shaft ends are square to the diameter then the shaft failed in torsion. If however the shaft ends have lips on them running at roughly 45 degrees to the shaft end this would indicate tensile failure on the principle planes due to the torsional shear.Either way looking at the shaft failure may give some clue as to what forces are acting and causing the problem.
Additionally I don't agree with the pump supplier saying tight is tight its meaningless without knowing the forces
the nut is trying to work against.

regards

desertfox

 

[Artisi]

Is the impeller /shaft attachement a taper or parallel bore, or screwed on?

If tapered is the "fit" correct.
If parallel is there excessive clearance?
If threaded are the threads of correct form and parallel.

An impeller/shaft that has correct / bore/ taper / thread and fit plus a locking nut with nut retention that is coming "loose" would suggest a problem other than the impeller / shaft retaining system.

You need to expand your investigation into the failure and not just look at the "nut coming loose"