impeller mount to shaft with loose fit, no setscrew or taper 2

[electricpete]

This is a 1.5 horsepower 3600rpm auxiliary oil pump (auxiliary to chiller compressor). It is a gerotor positive displacement style pump mounted directly to the motor shaft. Example of similar design here:

http://www.learromec.com/Products/PR_Gero.htm

I just noticed that the impeller(cylindrical bore) is a loose fit onto the shaft. I can slide it on and off effortlessly. There is no setscrew, no taperloack, just a loose fit and a key. Apparently this is the design (not just a worn impeller).

That seems like a strange setup to me. Don't most impellers (or coupling hubs for that matter) either have an interference fit or a setscrew or taper or some other means than the key to prevent relative movement between impeller and shaft?

Could this condition contribute to failures of the shaft key and keyway?

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[MikeHalloran]

Since it's flooded with oil and not subject to reversing loads, I wouldn't expect it to have an unduly limited life.

I'd guess it's floated intentionally, to allow the rotor to center itself between the side faces of the pump, without being influenced by the axial position of the shaft, and to allow the rotor to run true in the housing even if the shaft is not exactly perpendicular to the housing faces. The normal axial clearance between rotor and housing is very, very small.

If yours is a multiple- rotor pump, as is the one illustrated in your link, then I'd assert that the floating drive is absolutely essential.



Mike Halloran
Pembroke Pines, FL, USA

 

[electricpete]

Yes it has two rotors like the one shown. The inside rotor mounted to the shaft and the outside rotor driven by the bottom spokes of the inside rotor at 6/7 the speed of the inside rotor/shaft.

It makes sense there needs to be a close clearance of some kind to prevent bypass flow. How close I'm not sure. The self-centering you mention - that would be based on oil flowing through the end clearance?

It seems to me that pump impellers in general (as well as coupling hubs) always have something in addition to a key (interference fit, setscrew or taper). If it's not needed, why is that done? What would be an example of a reversing load on a pump? Although this may be non-reversing, it certainly seems like there may be pulsations in the torque that might create some movement.

We have had numerous failures of the keys on these pumps. At one point we upgraded to harder key material. Then we started having shaft failures. Recent attention has focused on pipe strain as a possible cause but I wanted to explore whether looseness might be a contributor. It seems to me a very nonstandard setup from my limited experience.





=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[electricpete]

By the way Mike. I forgot to thank you for that info. I do appreciate it (starworthy).

I'm not trying to argue with you, but trying to explain the parts that don't make sense to me. Particularly... why does it seem like all other pumps require setscrew, tight fit or taper?

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[MikeHalloran]

Without vanishingly small clearances, a gear pump becomes inefficient, because the leak path (from one face of a gear tooth to the other face of the same tooth) is very short. Inefficient in hydraulic systems means failure to attain system pressure. In lube oil systems, pressures are lower, and clearances can be relaxed a bit, but they still have to be small.

Spur gear pumps in hydraulic systems, with which I am moderately familiar, are typically run with axial clearance as small as possible, e.g., by match- grinding the housing and the rotors so they are all exactly the same length, then assembling the end caps with a cellophane spacer/gasket. Radial tip clearances to the housing are also held very tight; in some designs, steel gears are pressed into an aluminum housing, and a big wrench is used to effect the first few rotations. (Then the resulting chips are sometimes cleaned out)

In a gerotor, there's another good reason to float the inner gear. The tip clearance depends on the geometry and accuracy of the teeth, that roll and slide on each other. If the inner gear were fixed to a shaft, extra tip clearance would be required to prevent binding as the inner gear followed the shaft's runout. As just one example of a source for that runout, applying a setscrew to a loose fitting shaft/hub joint necessarily moves the hub off-center. To keep the leakage down, the inner gear must be allowed to follow the path prescribed by the interaction of its teeth with the outer gear, so that minimal clearance can be held simultaneously at all lines of contact.

Some gerotor pumps do that with a floating or radially flexible shaft and/or splines. The gerotor in the linked illustration does it with a double key drive, having some radial clearance.

If the soft keys had been failing because of fatigue from repetitive stress due to normal pressure fluctuations, the hard keys should have solved that problem.

Shaft failures with hard keys indicate that the soft keys were there for a reason. Repeated failures of the soft keys suggest contamination of the oil with hard particles.



Mike Halloran
Pembroke Pines, FL, USA

 

[checman]

electricpete: Gear backlash can cause this type of wear on your pump. It can come from continuously stopping and starting your pump or from sudden pressure changes from valves opening and closing. But on your application you might want to check to be sure you are not operating to close to the systems relief valve setting. If you have a poppet type R/V and are operating close to its cracking point the poppet may start to chatter and that can also cause gear backlash.

Regards checman

 

[ccfowler]

I would not worry much about pulsations. Gerotor pumps normally operate very smoothly and with very low NPSHr. Although I would generally prefer a spline drive connection, the key drive should be quite reliable at this low power level.