Tiny Dana air compressor repair

I've got a couple of little Dana 12 volt air compressors used to inflate tires. Photos are at

One of them has broken a plastic pulley and a few questions come to mind:

First, does anybody recognize and/or know anything about them? Probably no, but it can't hurt to ask.

Second, the broken pulley was a press fit on a 3/8" shaft with a shallow spline rolled into it. Is there any chance an aluminum pulley with dual setscrews might hold without an impossible-to- remove press fit using only the setscrews? The pulley is:

Thanks for reading, and any insights!

bob prohaska

--------------------------------

You could install and align the pulley, then remove one setscrew and dimple the shaft with a self aligning tap-size drill bit to give the setscrew more bite without raising a burr on the shaft. Then do the other one.

OW! Nope, never seen one before.

I always figured those were molded in place.

Of course there is a chance. I don't know how much of a chance, but there is one.

The pulley is:

I'd probably measure the OD of the splines, and if possible see how much runout there is. Then I'd probably do the following to give me the best chance of success.

1. Bore the new pulley to a light press fit on the spline.
2. Mill a keyway on the shaft and broach a keyway in the pulley.

3,. Drill and tap for two grub screws in the pulley. On to bear on the key, and one to bear on the shaft.

1. Drill a slight divot in the shaft for grub screw number 2 so it doesn't raise a burr.

If I can spin the shaft sans belt and pulley I might try to put a partial radius on the outsides of the splines for a better engagement as step zero. I might try it under an end mill.

I'd probably measure the OD of the splines, and if possible see how much runout there is. Then I'd probably do the following to give me the best chance of success.

1. Bore the new pulley to a light press fit on the spline.
2. Mill a keyway on the shaft and broach a keyway in the pulley.

3,. Drill and tap for two grub screws in the pulley. On to bear on the key, and one to bear on the shaft.

1. Drill a slight divot in the shaft for grub screw number 2 so it doesn't raise a burr.

If I can spin the shaft sans belt and pulley I might try to put a partial radius on the outsides of the splines for a better engagement as step zero. I might try it under an end mill. Bob La Londe

-------------------------------------

That's what I would do too, since I have a milling machine, keyway broaches, arbor press and a lathe to make custom guide bushings for the broaches.

A lower tech substitute that works fairly well is to drill and tap a small axial hole where a keyway would go, 1/2 in the shaft and 1/2 in the pulley, and tighten a short screw into it, secured with Loctite. A drill bit won't run straight in steel|aluminum but you could turn and bore a dummy steel bushing to drill and tap the shaft side of the hole straight, then replace the bushing with the aluminum pulley and do its side. I might drill the hole in both first, then tap them.

An even lower tech fix is to align and tighten the pulley, remove it and file flats where the setscrews marked the shaft. A second short setscrew will help lock the first in place.

And then there's Loctite.

At Segway a valuable prototype was mistakenly assembled with the permanent version of it. Guess whose bench that landed on to somehow repair.

I heated the screw heads with a soldering iron until the brass knurled and threaded inserts loosened, and pulled them out. More heat loosened the screws to salvage the inserts by tapping out the Loctite residue. Then I epoxied them back in using new oiled cover screws to position them. Those responsible were too embarrassed to give me credit for bailing them out.

The same trick with less heat backed out broken-off screws from a mobility scooter I bought cheap and non-functional.

Speaking of Segway... It looks like they are a Chinese owned company now according to this video from an influencer who claims they (or their promotional company) stiffed him.

Segway Reps Tried to Force me to Post Fake Review! (Segway Cube)

It was my background noise this morning while I had coffee and made breakfast.

Speaking of Segway... It looks like they are a Chinese owned company now according to this video from an influencer who claims they (or their promotional company) stiffed him.

Segway Reps Tried to Force me to Post Fake Review! (Segway Cube)

It was my background noise this morning while I had coffee and made breakfast.

Bob La Londe

------------------------------------

Little but the name remained when I left. As soon as the lean steer version went to production the engineering staff moved on, some to work on 4-wheeled electric vehicles.

Like most of the R&D projects I've been on it was fun while it lasted.

I've been wondering about Loctite and/or epoxy. The compressor gets too hot to touch in the area of the cylinder and head. I'd imagine the input shaft might remain quite a bit cooler but am not sure. If smoking heat is required to remove Loctite it'll probably work well. If boiling water softens it, likely it won't last.

I don't have the tools nor skill to make keyways and will have to content myself with boring the pulley and putting small flats/divots on the shaft to take the setscrews.

On reflection the maximum torque doesn't seem huge. At 100 PSI there's roughly 100 pounds force on the piston with a half-inch lever arm. With a 3/16 inch shaft radius it's only 270 inch pounds. Have I got that right? The new pulley seems to be coming direct from China and should arrive sometime late in the month. When I get a good look at it there will probably be more questions.

Thanks very much for writing!

bob prohaska

I don't have the tools nor skill to make keyways and will have to content myself with boring the pulley and putting small flats/divots on the shaft to take the setscrews.

---------------------- I don't know who has machine tools here. You can't do much serious metalworking without them, or the skills of an 18th century clockmaker.

The semi-magical toymaker Drosselmeyer in The Nutcracker was similar to real master craftsmen of the time (1816), as was the creator of the automated female android Coppelia by the same author. Androids look human, robots may not, though writers play loose with the terms.

If the setscrews have cup points that leave a ring indentation you can use it as a guide to file the flats. File deep enough to see the entire ring and try to make it disappear simultaneously all around. It may help to blacken the developing flat with a marker. Match-drilling a divot in the steel shaft without the guidance of a milling machine risks the setscrew threads in the aluminum pulley.

Intuition suggests there should be a reasonably direct relation between press fit force and torque holding ability, at least when all dimensions are close to 1. Is there a rule of thumb or table? A few hundred pounds of press fit wouldn't be very hard to apply.

Here's a photo (sorry for the blur) of the compressor shaft:

Here's a view of the back of the crank: the crank in a press (vise) or against an anvil might make it fairly easy to assemble by force.

The tip shaft diameter is .350 inch, the splines appear to be knurled in before hardening (the shaft feels file- hard) with a diameter of .355 inch. The basic diameter of the crankshaft is .375 inch. Looks like I'll have to do some boring. The pulley bore is .316 or .317, nominally 8 mm.

The pulley showed up direct from China, postage paid, in about two weeks. Really rather amazing for $8.65 .

Thanks for reading,

bob prohaska

Intuition suggests there should be a reasonably direct relation between press fit force and torque holding ability, at least when all dimensions are close to 1. Is there a rule of thumb or table? A few hundred pounds of press fit wouldn't be very hard to apply.

Here's a photo (sorry for the blur) of the compressor shaft:

Here's a view of the back of the crank: the crank in a press (vise) or against an anvil might make it fairly easy to assemble by force.

The tip shaft diameter is .350 inch, the splines appear to be knurled in before hardening (the shaft feels file- hard) with a diameter of .355 inch. The basic diameter of the crankshaft is .375 inch. Looks like I'll have to do some boring. The pulley bore is .316 or .317, nominally 8 mm.

The pulley showed up direct from China, postage paid, in about two weeks. Really rather amazing for $8.65 .

Thanks for reading,

bob prohaska

---------------------------------------- I struggled unsuccessfully with such low tech repair methods until I bought a lathe and milling machine that have probably paid for themselves with the money saved by repairs and custom tools, and again by advancing my project responsibilities at work. For example I bought a "worn out" snowblower for $100 and brought it back to like-new function if not appearance by correcting its wear. Bushing the worn hole that I just mentioned to Bob was a part of it. It's old enough to be all metal that can be repaired better (thicker) than new.

At work I demonstrated that I could handle the mechanical design and machining of a project, not just the electronic work that my job title suggested. That's what got me into Segway, in fact they gave me the work that required old-fashioned hand craftsmanship as well as machining.

bob prohaska snipped-for-privacy@www.zefox.net> wrote: ... A few hundred pounds of press fit wouldn't be very hard to apply.

Here's a view of the back of the crank:

the crank in a press (vise) or against an anvil might make it fairly easy to assemble by force.

---------------------- The exposed end is the offset crank throw. Push too hard and you might bend it. Pushing against the counterweight may be safer, if you can.

understood. Ideally the pressure would be applied in balance on both. That would require some sort of stepped (or undercut) mandrel.

After thinking it over, I believe the force required to create a press fit between a smooth shaft and a smooth hole would be roughly equal to the torque applied at the hole radius needed to make it slip.

On paper, if the torque generated by the piston/rod/crank was

270 inch pounds a press fit nominally tighter than 270 linear pounds would be expected to hold. I'm assuming shearing friction is equal in both axial and azimuthal direction. That obviously won't be true in my case, thanks to the axial splines knurled into the shaft.

If somebody sees an error in this reasoning please point it out.

It's not clear I want to use a press fit (disassembly will be extremely difficult if it's ever needed) but it looks like an option.

Thanks for reading,

bob prohaska

understood. Ideally the pressure would be applied in balance on both. That would require some sort of stepped (or undercut) mandrel.

After thinking it over, I believe the force required to create a press fit between a smooth shaft and a smooth hole would be roughly equal to the torque applied at the hole radius needed to make it slip.

On paper, if the torque generated by the piston/rod/crank was

270 inch pounds a press fit nominally tighter than 270 linear pounds would be expected to hold. I'm assuming shearing friction is equal in both axial and azimuthal direction. That obviously won't be true in my case, thanks to the axial splines knurled into the shaft.

If somebody sees an error in this reasoning please point it out.

It's not clear I want to use a press fit (disassembly will be extremely difficult if it's ever needed) but it looks like an option.

Thanks for reading,

bob prohaska

----------------------------------------

Here is an alternative. Drill and tap a bolt at the seam. Then a close slip or light press fit might just do. Might be hard without machine tools if its aluminum against steel. Of course if you had machine tools you would have made a pulley a long time ago and been done with it.

Loctite retainer or sleeve and bearing locker might be an alternative for you as well.

That calculator solves the inverse of the question I'm asking, using dimensions and material properties. I'm thinking in the reverse direction: I know only the force required to press a pulley onto a shaft and the shaft diameter. How can I estimate the torque the joint can transmit?

My simplistic estimate is that friction is friction, and friction that resists axial motion will equally resist rotation. So, for a given assembly force, that force multiplied by the joint radius will be the absolute torque limit. Far as I can see that's the whole story, apart from material properties and stiction.

Thanks for writing!

bob prohaska

That calculator solves the inverse of the question I'm asking, using dimensions and material properties. I'm thinking in the reverse direction: I know only the force required to press a pulley onto a shaft and the shaft diameter. How can I estimate the torque the joint can transmit?

My simplistic estimate is that friction is friction, and friction that resists axial motion will equally resist rotation. So, for a given assembly force, that force multiplied by the joint radius will be the absolute torque limit. Far as I can see that's the whole story, apart from material properties and stiction.

Thanks for writing!

bob prohaska

-------------------------------------- As you wrote yourself, the splines break that relationship. Personally I'd determine an interference on a separate sample that's a little too tight to press on, bore the pulley without changing the lathe boring bar setting, then heat the pulley to expand its bore and press it on, against a stop, with one rapid stroke. Do you have a gear puller to undo a failed attempt?

The sample could be bored in short steps to see how the force increases with the amount of interference. Less than 0.001" can make quite a difference. You don't have to measure the bore, just keep the last setting. This is probably hopeless on a drill press with the ~0.010" gap between letter sized drills.

No lathe? Then you are limited to buying what you can't make.

That calculator solves the inverse of the question I'm asking, using dimensions and material properties. I'm thinking in the reverse direction: I know only the force required to press a pulley onto a shaft and the shaft diameter. How can I estimate the torque the joint can transmit?

---------------------

When a calculator doesn't let me enter what I want as input I close in on it by changing what can be entered until the results are close enough.

I do have a lathe. A well-worn South Bend 10L. While trying to sneak up on the ideal pulley bore I overshot the mark (by less than the radial play in the crankshaft) and ended up using the setscrews.

The setscrews haven't slipped in the initial testing. If they do slip I'll mark the spots and try to grind seats for them in the shaft. That can be done fairly easily with a Dremel grinder. It isn't at all obvious how to get the crankshaft out of the case. Or, for that matter, how it was assembled. It wouldn't be a bad thing to clean and lubricate at least the crank, rod and wristpin.

Thanks for writing!

bob prohaska

There's very little space to bring a puller to bear. At least the setscrew approach will make further repairs a little easier.

<snip>

Doesn't matter now... but check out Bearing Separators, may be handy to know about them someday if you don't already. Short video, ~2 minutes:

Nice repair project you've done there👍

I do have a lathe. A well-worn South Bend 10L.

----------------------------------

I have a 10L too, a 1965 model from a trade school that abused it, such as using the tailstock spindle as an anvil horn. The dealer replaced it with another that's a slightly loose fit. Some day I may bore out and sleeve the tailstock casting, which might not be the original.

They are what the National Bureau of Standards used before they bought Hardinges, and an old review Google can't find anymore claims the SB was more versatile as general purpose lathe. I bought a 6" threaded and a 4" 5C mount 6-jaw chucks for it that I use quite often on jobs that don't need repeatable centering accuracy.