Advice wanted on motor shaft coupler

[calzap]

I have a 3 HP centrifugal water pump  with a failed motor.  Here's the link to the motor post mortem: https://www.eevblog.com/forum/projects/pump-motor-postmortem/new/#new .  Problem is a long extension of the motor shaft is also the pump drive shaft (aka close coupled).  Pump is OK, but I would like to replace the original motor (WEG brand) with a better brand (Marathon).  And in the future, I want to be able to service or replace the motor without having to disassemble the pump.

I'll cut the shaft near the old motor and connect the new motor to the pump shaft with a coupler.  I'll also build a sturdy steel frame to hold the pump and new motor.  And yes, there will be a safety shield over the coupler and shafts.  Shaft on pump is 3/4 inch diameter (19 mm).  Pump motor will be 3 HP, 240 VAC, 60 Hz, 1 Ph, and 3450 or 3600 RPM.  According to my sources (spicerparts.com and engineeringtoolbox.com), this represents a torque of 55 in.lb. (6.2 Nm).

What kind of coupler should I use to connect the motor shaft and pump shaft?  Will definitely want one with some flexibility.  A double U-joint coupler would allow the greatest flexibility but is the longest and most expensive.  This leaves a couple of types of flexible-shaft couplers as options.  One type is an aluminum cylinder with spiral slits in it that clamps to the shaft.  The other consists of two opposing nylon hubs with facing fingers and a polyurethane pad (aka spider) between.  It attaches to the shaft with set screws. See pictures.  I was surprised by the high torques the latter type can handle ... much higher than the spiral slit type.    What would you advise?

Mike in California

[andy3055]

Tkae the second one. Make sure it is rated for the HP and RPM. Not sure if they do rate like that though.

[duak]

The upper coupler, the one with slits, is better for servo systems where backlash has to be minimized.  If it's made from aluminum it'll have a fatigue life.

The lower coupler can handle larger axial and radial misalignment.  Of the two, I'd go for the lower one.

About the motor failing - Is it possible to upsize the motor slightly to buy some margin?  My observation of motors these days is that the margins we used to take for granted have been engineered out.

[calzap]

Couplers are rated by max torque, max rpm, environmental temperature limit, and tolerances for parallel, axial and radial misalignment.

    torque x rpm = power x k.  Value of k depends on units chosen for torque and power.

I am hesitant to use a higher HP motor because it might result in harmful stress on the pump.  The pump worked fine with the old motor.  It was some cost-cutting in the design and construction of the old motor (like the nylon fan blade) that caused it to die, not a lack of power.  I'm more concerned with long-term reliability and ease of servicing and replacement than more power.  How nice it would have been if we could have removed the old motor without having to disconnect the pump from the plumbing and then disassemble the pump.  Marathon motors on other pumps of ours have given long service with no problems; so that's what I'll probably buy. 

Mike in California

[langwadt]

I've tried the spiral ones on a CNC, if something goes wrong they just "unwind" the spider ones are much better

[jdragoset]

Go to local belt/bearing/power coupling supply (Motion Industries/Grainger) and ask for Love Joy.
https://www.lovejoy-inc.com/products/jaw-type-couplings/

[Gregg]

For more than 1 horsepower I have had good results with Lovejoy S-flex couplings as they have more surface bearing on the elastic component.
https://www.lovejoy-inc.com/products/s-flex-couplings/
For any motor driven flex coupling, the better the alignment, the longer it will live.
edit: McMaster also sells couplings

[radar_macgyver]

You can also use an Oldham coupler, which can take up significant axial shaft misalignment. When using all-steel versions, they can have high torque ratings. Or a v-belt with appropriate sheaves.

[tautech]

Mike, the pumps design will dictate what sort of coupler you can use.
As I see it in your failed moter thread the pumps top bearing is in common with the motors lower bearing and below that is a shaft seal, right ?
Further, the pumps bottom bearing is just a bush running wet in the water, right ?

Therefore IMO you need a rigid self centering coupling and one that can transfer any pump shaft axial stresses to the bottom motor bearing and also precisely control pump shaft end float.
IME they are a split clamping type sometimes with a drive pin and/or keyway.
Some examples of what I mean:





In multistage centificals I've worked on or replaced the wet end there's a forked bush that you apply below the coupling to set the pump frame to coupling measurement correctly for correct impeller clearances, hence the wide access to the shaft coupler behind the cover plates.

In doing this mod with a non-standard motor to pump coupling and motor mount you will need to check for the impeller end float clearances in order to replicate them for proper operation and good pump life.

Certainly not impossible with the right gear and access to a machine shop to true the motor mount and make the split coupling.
Good luck. 

[PaulAm]

If you're going to cut the motor shaft, you'll probably want to cut a keyway on that to fit your coupling and provide some kind of bearing if the pump depended on the motor bearing.  Without a keyway the coupling will spin on the shaft the first or second time it starts up.

Getting a little involved, maybe easier to scrounge a pump meant to be driven by a separate motor?  Check scrapyards and industrial surplus; it shouldn't be too hard to find one.

[calzap]

Thanks for the link to Lovejoy couplers.   Yeah, I've been looking at couplers supplied by McMaster.  I do a lot of business with them.  The Oldham and Lovejoy couplers are variations on the plastic "spider" type of coupler:  two metal or plastic pieces facing each other with fingers or teeth and a softer plastic cushion between them.

Pump is horizontal, not vertical.  It's made by Gorman-Rupp and sold under their AMT badge or rebadged as Dayton and sold by Grainger/Zoro. It's also sold as Teel brand.  As sold, the pump frame supports the close-coupled motor.  See the picture.

I'm definitely sticking with this pump model.  After 10 years and more than 6000 hrs run time, the pump that had the motor die had been faultless and performed exactly as expected.  The backup pump (same model) has been faultless after a year of use.  If I change the pump model,  I'll have to redo the plumbing ...  ugh, no, not happening!   I can get a new backup motor/pump combo (same model) on sale now for about US$1000.  In terms of money and time that would probably be more economical than putting a motor with a coupler on the pump that has the dead motor.  The pump that had the motor die could be kept for parts.  So, it's a learning project that might save time down the road because if the motor should need servicing or replacement, the pump won't have to be disconnected from the plumbing and disassembled.

Mike in California

[tautech]

Thanks for the link to Lovejoy couplers.   Yeah, I've been looking at couplers supplied by McMaster.  I do a lot of business with them.  The Oldham and Lovejoy couplers are variations on the plastic "spider" type of coupler:  two metal or plastic pieces facing each other with fingers or teeth and a softer plastic cushion between them.

Pump is horizontal, not vertical. It's made by Gorman-Rupp and sold under their AMT badge or rebadged as Dayton and sold by Grainger/Zoro. It's also sold as Teel brand.  As sold, the pump frame supports the close-coupled motor.  See the picture.

I'm definitely sticking with this pump model.  After 10 years and more than 6000 hrs run time, the pump that had the motor die had been faultless and performed exactly as expected.  The backup pump (same model) has been faultless after a year of use.  If I change the pump model,  I'll have to redo the plumbing ...  ugh, no, not happening!   I can get a new backup motor/pump combo (same model) on sale now for about US$1000.  In terms of money and time that would probably be more economical than putting a motor with a coupler on the pump that has the dead motor.  The pump that had the motor die could be kept for parts.  So, it's a learning project that might save time down the road because if the motor should need servicing or replacement, the pump won't have to be disconnected from the plumbing and disassembled.

Mike in California

That's not in discussion, only the principles the coupler MUST adhere to.
Please re-read my post again without the bias of thinking I recommending a change to vertical.
There are basic engineering requirements to swap to a coupled motor so your project is not to be a waste of time and effort.

For the plumbing to be a hassle to replace/service/repair the pump is the fault of the original installer where Mac unions of some type should have been used to quick and simple swaps of pump units.

[calzap]

Tautech,

I know you weren't saying I should switch to vertical.  My response about not switching pump models was in response to PaulAm.   The plumbing does have unions, but a change in geometry or inlet/outlet diameters, would necessitate some significant plumbing changes.  But it's not happening anyway.

I believe I understood your post and appreciate the advice.  The picture below shows the shaft, seal and impeller arrangement.  What is called "adapter" in the diagram is actually part of the pump case, and the motor bolts to it.   It's called adapter because the shafts  are slightly different for the 1 Ph vs 3 Ph versions of the motor.  A more expanded view of the adapter (labeled #2) is in the second picture.

Mike in California

[calzap]

Now I can see that a flexible coupling could allow a disastrous amount of wiggle because the far end of the shaft is not supported and the impellers might contact the case.   And a rigid coupling would require some tricky alignment.  Sounds like it's time to give up my idea of a separate motor and coupler.   I'll either buy a replacement close-coupled motor  or a motor/pump.  Thanks for all the responses.  I learned a lot.

Mike in California

[tautech]

Tautech,

I know you weren't saying I should switch to vertical.  My response about not switching pump models was in response to PaulAm.   The plumbing does have unions, but a change in geometry or inlet/outlet diameters, would necessitate some significant plumbing changes.  But it's not happening anyway.

I believe I understood your post and appreciate the advice.  The picture below shows the shaft, seal and impeller arrangement.  What is called "adapter" in the diagram is actually part of the pump case, and the motor bolts to it.   It's called adapter because the shafts  are slightly different for the 1 Ph vs 3 Ph versions of the motor.  A more expanded view of the adapter (labeled #2) is in the second picture.

Mike in California

Sure, the adapter in #2 is just the motor mount and likely a standard SAE mount (or somesuch) and completely rigid just as any shaft coupling must also be to transfer all pump axial and thrust forces to the motor bearings. That the shaft rotates @ 3000+rpm requires the shaft coupling to be precisely engineered/machined so to not introduce additional vibration or misalignment.

If it were me....and it's not, I'd be replacing like with like to get things up and running and spending some good time and effort to devise a robust solution for both the motor adaption and shaft coupling.
I don't see it as shit easy however it's very possible.
A shaft coupling remedy should be devised first then extension to the motor mount to allow room for the coupling WRT overall length and possible differences in the motor mount/coupling measurements. You might get lucky and only need spacers and longer bolts however care need be taken without a motor mount alignment spigot to keep it all perfectly aligned.

Good luck. 

[tautech]

Now I can see that a flexible coupling could allow a disastrous amount of wiggle because the far end of the shaft is not supported and the impellers might contact the case.   And a rigid coupling would require some tricky alignment.  Sounds like it's time to give up my idea of a separate motor and coupler.   I'll either buy a replacement close-coupled motor  or a motor/pump.  Thanks for all the responses.  I learned a lot.

Mike in California

Yep, without support it would fail, and bad !