Extending a motor shaft

[Bob Engelhardt]

I have this DC motor that was originally on a kid's "motorcycle", driving a chain, & with a short shaft:


(the bearing takes up a lot of what's shown)

The shaft is 1/2" (13mm) diam & I'm thinking of extending by making it 5/8" (15mm). I'd turn the shaft down to 3/8" (10mm) and have a 3/8 bore in the extension. I'm thinking that the 1/8" (3mm) wall in the extension would be better than the 1/16" (1.5mm) that I'd get with a 1/2 shaft & 5/8 extension.

It's the fastening them together that I'm puzzling about. I'm expecting to get a lot of torque & I'd like to silver solder them, but my silver solder is 45% & takes a lot of heat to use. There is a phenolic (?) spacer close by & I wouldn't want to over heat it.

Plug weld ?

Drill & pin it?

Loctite ? If Loctite, then the 1/8 is not so important.

Thanks, Bob

[Arcane]

I'd try a short piece of 5/8" shaft with a 1/2" hole in it to slip over the original shaft with Loctite and using a boring bar with a small curve on the cutting tip to eliminate a stress riser at the bottom of the hole wouldn't hurt. Drill a small hole through the center axially so it doesn't "air lock" on you when you slide it on.

If that didn't work, then I would go to Plan B

[BCRider]

And how would you ever take it apart to do any servicing inside? Or if needed to replace the bearing? For that matter the heat on that bearing from the silver soldering would likely fry the bearing. Never mind the phenolic spacer that would certainly fry from the sort of heat needed for anything other than soft soldering.

If you don't mind the motor having a life span limited by that one bearing due to lack of access I think you're looking at a pretty firm press fit or a light press fit with roughened surfaces and one of the Loctite or other strong assembly adhesives. The idea of the roughened surfaces to give room for little channels of the adhesive and the light press across the crests of the roughness as a fit for alignment.

[old mart]

I would go for Arcanes method, then the stiffness of the shaft would not be reduced.
Rub some abrasive paper axially along the part of the shaft to be loctited to strengthen the bond.

[Bob Engelhardt]

And how would you ever take it apart to do any servicing inside? Or if needed to replace the bearing? ...

A new bearing will be fitted to the 5/8 shaft (& the housing enlarged). Didn't mention that - my neglect.

[Paul Alciatore]

I am not a big fan of reducing the diameter of a shaft to extend it. The strength of a round rod is increased or decreased by at least the second power of the diameter so reducing the diameter to 50% will reduce the strength to less than 25%. And that is probably unacceptable in a motor where the original designer was under pressure to minimize every part to save costs. On top of that, just adding a radius to the stress riser that a shoulder will produce does not eliminate that increase in stress: it only reduces it somewhat.

As for welding an extension on, beyond the danger of melting internal components of the motor, heaven only knows what that would do to the strength of the shaft. Are you going to anneal it afterwards? How? I would not be surprised if a welded shaft were to simply break in two after some thousands of revolutions with the cyclic loading as each rotation occurs.

In this case, a larger diameter is apparently OK so there should be no reason for decreasing the diameter of the motor shaft. I would leave that dimension completely alone. Instead, I would make a sleeve that fits over that diameter and fasten it to the shaft with a SHCS which is in a threaded hole down the middle of the shaft. Instead of the higher percentage of loss in the strength of the shaft that reducing the diameter would bring, removing material from it by drilling a hole in the center will have a smaller decrease in that strength as compared to the loss of cross sectional area. And even that would be, at least partially, compensated by the strength of the screw itself. Here is a drawing of what I mean:



The shaft diameter is not reduced at any point. Any stress risers at the end of the threaded hole or at the roots of the threads are combated by the wrap around nature of the sleeve. The sleeve itself is in compression from the SHCS and this increases it's strength.

As shown in the drawing, the sleeve should extend beyond the end of the screw hole: I would recommend as far as possible. I would suggest a 1/4" screw for the 1/2" shaft and fine threads instead of coarse. As an option, you can choose a left or right hand screw based on the direction of rotation. The end of the shaft should be faced and the bottom of the well in the extension should also be either faced or slightly concave in the middle: this will ensure contact at or near the OD of the shaft to prevent any flexing there.

An appropriate LocTite adhesive (thread locker) can be used to ensure that the joint does not slip. I would recommend a diameter of at least 3/4" for the extension, but you may be able to get away with less.

Another measure that can be taken would be to round the corners of the tap drill used to make the shaft hole. This would further reduce any stress riser at the bottom of that hole. It also increases drill bit life as the outer corners of a bit usually wear the most. I have done this by hand on several of my drill bits to extend their life between regular resharpening.

[darryl]

Make the extension like Paul shows, but then slit across the bored end several times to turn that into fingers. When that is pushed onto the motor shaft, the fingers can be compressed in some way to lock the extension to the motor shaft. A two part, screw together ring would do it, or perhaps a compression ring or a compression nut would be used.

[dian]

make a light press fit addapter, pin it and thats it.

[J Tiers]

Paul A's solution would be even better if the joint is tapered and not parallel. Tapered joints can be made very tight but still removable. A parallel joint cannot be made both tight and easily removable, there must always be some clearance or it becomes a press fit.

[Duffy]

Why not a suitably sized taper pin. Of course this presupposes that you have a matching taper pin reamer

[Bob Engelhardt]

I am not a big fan of reducing the diameter of a shaft to extend it. ...

Yeah, I'm thinking that 3/8" is really pretty small.

As for welding an extension on, beyond the danger of melting internal components of the motor, heaven only knows what that would do to the strength of the shaft. ...

The welding alternative was a way to secure a sleeved extension to the shaft. It was just using a couple of plug welds. The advantage would be much less heat than silver soldering.

... Here is a drawing of what I mean:

Nice drawing.

...An appropriate LocTite adhesive (thread locker) can be used to ensure that the joint does not slip. I would recommend a diameter of at least 3/4" for the extension, but you may be able to get away with less. ..

I have some LocTite 609, 660, and 680 retaining compounds. The 609 is for close fitting assembly (from press fit up to .006 clearance) & has a shear strength of 2300 psi. Which means for my 1/2" diam, 1" long joint, the shear torque limit would be 150 ft-lbs (203 N-m)! I do need to test that it hasn't timed out (exceeded its shelf life).

[Arcane]

.......... I do need to test that it hasn't timed out (exceeded its shelf life).

From what I have read Loctite has an extremely long shelf life regardless of what the company says. Don't forget, they want you to buy new stuff regularly. I have some that I got back must be close to 35 years now and it still works!

[ed_h]

I had to do something similar to a garage door opener motor. I turned about 5/8" of the the 7/16" motor shaft down to 1/4", and threaded it. Then drilled the end of a short piece of 1/2" rod and tapped it 1/4-20. Screwed the extension on, then turned it down to 7/16 in situ. Voila--a same size shaft extension that runs true.

Ed

[Paul Alciatore]

I have mused about tapered fit extensions in the past for small diameter shafts. The taper does seem to be a stronger way to do it and, as you say, it can be easily removed. But then, perhaps too easily. You would not want the constant vibration and cyclic loading from the motor to work it loose. Taper plus a screw perhaps? That is used for tooling.

The 1/2" size here does border on small, but is probably large enough to be OK with the method I showed. A taper fit would require both a taper reamer and a matching taper on the shaft.

When you said "press fit" a wild thought came to me. Yes, make it per my sketch with a press fit. But add a larger size, internal thread in the extension. Normally that thread would not be in use and the retaining bolt would pass through them without engaging. But if disassembly is needed, a bolt with that thread and a smaller diameter for it's first inch or so could be used to disassemble it. This disassembly tool would reach past the threads in the shaft to the bottom of that threaded hole and the threads on the section in the extension would allow it to push the extension off. It would be a somewhat long bolt/tool.

The biggest problem I see with this is the disassembly tool would almost certainly be lost long before disassembly was needed and the poor fool looking at how to do it would be hard pressed to guess how to make a replacement. I did say it was a wild idea, didn't I?

Paul A's solution would be even better if the joint is tapered and not parallel. Tapered joints can be made very tight but still removable. A parallel joint cannot be made both tight and easily removable, there must always be some clearance or it becomes a press fit.

[Bob Engelhardt]

From what I have read Loctite has an extremely long shelf life regardless of what the company says. Don't forget, they want you to buy new stuff regularly. I have some that I got back must be close to 35 years now and it still works!

I had never considered its shelf life until I had a red and a blue thread lockers go bad. They were probably 5 years old and 95% unused. Really annoying to throw out so much unused product. I replaced them with much smaller sizes.