Machined Parts: No Clearance

[David Winer]

I’m making a cart for my canoe club. It will haul boats back and forth from the canoe shed to the river. Mostly it’s of wood, but the ¾ inch axle and ¾ inch bearings are of steel. Apparently the two are made so precisely that there is no clearance for the axle to slide through the bearings.

I have tried smoothing and lubricating both articles but to no avail—the mating is still too tight to allow slipping the axle through to the desired position in the bearing. I’ve include photos to illustrate that a ¾ inch bit won’t go through the ¾ inch bearing hole.

Suggestions from Creekers should help. Many of you are skilled in metal-crafting as well as woodworking.

[Gerry Grzadzinski]

Sand the axle down with some 400-600 grit sandpaper. It shouldn't take much at all to get it to go in.

[David Eisenhauer]

Or use Clover compound and/or emery cloth.

[Charles Taylor]

Heat the bearing (and the casting into which it fits) directly atop a 100 watt incandescent bulb for a few minutes. Also, if it will fit, put the axle in the freezer for a while beforehand. Don't forget gloves to handle the bearing.

[David Winer]

Sand the axle down with some 400-600 grit sandpaper. It shouldn't take much at all to get it to go in.
As I said, I have tried this (using 600 grit wetordry paper). I'm uncomfortable with smoothing the axle because it would need to be to kept round over its whole length.

I'm hoping for a solution based on someone's experience rather than "have you tried?" For example someone might say to "warm the bearings and cool the axle, or take the bearings to a machine shop to enlarge the holes--that's the way I do it."

[David Winer]

Heat the bearing (and the casting into which it fits) directly atop a 100 watt incandescent bulb for a few minutes. Also, if it will fit, put the axle in the freezer for a while beforehand. Don't forget gloves to handle the bearing.
Thanks, Charles. Have you done this before? In this case I would want to try it first, then take the bearings off for re-assembly later in their final position on the axle. I'd hate to have the parts seized together so I can't remove them for later assembly.

[Lee Schierer]

Thanks, Charles. Have you done this before? In this case I would want to try it first, then take the bearings off for re-assembly later in their final position on the axle. I'd hate to have the parts seized together so I can't remove them for later assembly.

If you heat and cool the parts to get them together, they won't come apart willingly without a good press.

[Russ Ellis]

I would not heat the bearings without knowing the rated operating temperature.

Dry ice to shrink the diameter of the shaft is an option. Just don't drop it.

A good machine shop can assemble these with a hydraulic press.

Bearings are typically intended to have an interference fit with the shaft. This makes sure the bearing race is spinning and not the shaft spinning inside the bearing race. The interference fit would be in the range of +0 / -0.002 inches.

[Gerry Grzadzinski]

As I said, I have tried this (using 600 grit wetordry paper). I'm uncomfortable with smoothing the axle because it would need to be to kept round over its whole length.

Do you need to slide the entire axle through it, or just the ends?
Do you have a caliper or micrometer that you can measure the axle with? I would expect that the bearing is pretty close to .750", but the shaft is probably a bit oversize. Even .001-.002" oversize and it won't go in, but should be easily sanded enough to fit. And yes, I've done this before.
For what you're using it for, it shouldn't matter how precise your sanding job is, as long as the axle slides in.

[Charles Taylor]

Thanks, Charles. Have you done this before? In this case I would want to try it first, then take the bearings off for re-assembly later in their final position on the axle. I'd hate to have the parts seized together so I can't remove them for later assembly.

I have, when reassembling a drill press. Probably will be doing it again soon with my table saw, which I have taken apart to replace the old bearings. I didn't realize you were planning a "trial run", and I agree with Lee's warning that to get them apart again, you will likely need a press or puller.

[Matt Day]

I've used the freezer/oven method and had success. I think i did so when I replaced the my TS arbor bearings.

Regarding the statement about keeping the shaft round, I don't think you're taking off nearly enough material for that to be a concern.

[David Kumm]

Heating bearings is a common way to mount them. Get a heat stick and mark the inner and heat until the mark liquifies. Once on you won't likely get it apart without wrecking the bearing but they don't look like expensive ones. If heat or cool doesn't do it, you could take the shaft to a machinist. Shouldn't cost much to take a .001 off the shaft. Dave

[Keith Weber]

A good machine shop can assemble these with a hydraulic press.

Bearings are typically intended to have an interference fit with the shaft. This makes sure the bearing race is spinning and not the shaft spinning inside the bearing race. The interference fit would be in the range of +0 / -0.002 inches.

Russ beat me to the punch. Bearings usually are press fit on to the shaft. Find someone with a hydraulic press, or hefty arbor press. You'll need to know where you want to stop the bearings when being pressed onto the shaft, so you can assemble everything later. Another option might be to source a bearing with a slightly larger inside diameter. When I say slightly larger, I mean a few thousandths, not the next fractional size up. This is a canoe cart and not a planer head, so a press fit bearing is probably overkill anyway.

[ian maybury]

Hi guys. Pardon a different view. Stock pillow blocks and flanged bearing units like David's are normally sized so that standard bright or drawn shafting is a close sliding/clearance fit in the bore of the bearing. The bearing inner race is locked on to the shaft so it doesn't turn/spin using the provided set screws. It's basically intended as a cheap and cheerful (= no machining, sealed, self aligning, easily assembled) means of mounting a not too heavily stressed shaft in a bearing. They can over time with a bit of corrosion become locked on and need cleaning and/or a pullers to remove. (it won't anyway survive very long if dunked in water)

The table near the top of the first page of this chart lists typical shaft and bearing inner race bore tolerances in inches: http://www.emersonindustrial.com/en-US/documentcenter/PowerTransmissionSolutions/Installation%20and%20Maintenance/Form_9579E.pdf There's a similar table in metric units on page 13 in this FAG bearing write up: http://www.schaeffler.com/remotemedien/media/_shared_media/08_media_library/01_publications/schaeffler_2/tpi/downloads_8/tpi_219_de_en.pdf

Drawn/bright shafting is typically produced a hair under nominal size for this reason (perhaps 1/2 thou at this size), and bearing inner races a shade over. They have long since been made to a stock tolerance class determined by standards - so fit problems are unusual. At 3/4in dia the total clearance is likely to be something less than 1 1/2 thou - which should mean the shaft slides freely in the race, but doesn't rattle about.

All it takes it a small ding or a bit of rust on the shaft to mess up the fit, or maybe a burr on the inside of the bearing race. It'll jam tight on a few specks in the wrong place. Maybe as the guys say a touch with emery cloth or whatever would sort it out? Go easy on the bearing getting the shaft out or back in - support the inner race using a square ended tube/sleeve something. Unsupported hammering will destroy the ball race.

If cleaning up doesn't sort it and the shaft is truly oversize for the bearing, then something is likely wrong and it's a case of changing for replacement parts. If you have a micrometer you could measure the shaft and decide whether the problem is the shaft or the bearing...

[Ellery Becnel]

Ian is spot on for this application. Not all bearings are made to have an interference fit with the shaft. The issue is most likely with the shaft. The only way of knowing is to measure both.

Ellery Becnel

[Rod Sheridan]

David, did you buy a precision axle or merely a piece of cold rolled round steel?

Cold Rolled Round often isn't accurate enough to be used in a bearing...............Rod.

[David Winer]

David, did you buy a precision axle or merely a piece of cold rolled round steel?

Cold Rolled Round often isn't accurate enough to be used in a bearing...............Rod.

FIRST, I want to thank everyone who responded with advice about a matter that I got myself into from ignorance. I didn't know enough even to state that the axle and pillow block type bearings were not designed to be mated. I purchased them from different sources on the internet. The shaft is stainless steel. My plan for the cart has been to make the axle go from wheel to wheel instead of using short pieces for an axle on each wheel. And this scheme means that the bearings would need to be put on after the cart's wooden frame is finished with full length axle in place. (I see that it all might not work out the intended way.)

You all have come up with several approaches that I may use, and may indeed use some of the ideas in combination. One thing I have gleaned is that abrading the shaft to fit for this application does not involve precise fitting in either dimension or roundness. So "sanding" could end up a dirty way to get 'er done if more sophisticated ways are not practical. Taking the problem to a machinist (one of the suggestions) sounds like the most straight forward.

Further ideas are welcome. This forum is a gold mine.

[Ronald Blue]

Ian and Ellery are correct. The issue is with your axle. As Ian said look closely for a nick or ding or even possibly a swell on the end. If you can measure it accurately check it's diameter. You can easily remove excess stock with emery cloth in the 150-200 range. If you work your way around evenly you won't get it out of round enough to bother anything. After all it's just an axle on a hand cart. Another thing to watch is if the clearance is tight if you get even slightly cocked it will lock on the shaft. Also if it starts on when you remove it if there is a ding or swell at fault it will be shiny at that contact point. You should only need to go in a few inches on each end unless this is different than I envision. I am not so sure the forstner bit is a good way to gauge the bearing size.

[Greg Urwiller]

Another thing to consider: I assume your wheels will spin freely on the axle? Otherwise your cart won't turn well if the wheels are locked onto the shaft. When turning your cart each wheel is going to need to rotate at a different speed just like your vehicle axle/differential. If so then you don't need the axle bearings. The axle can be locked solid to the cart by other means rather than a bearing. Or am I looking at this wrong?

[David Winer]

"You can easily remove excess stock with emery cloth in the 150-200 range. If you work your way around evenly you won't get it out of round enough to bother anything. After all it's just an axle on a hand cart. Another thing to watch is if the clearance is tight if you get even slightly cocked it will lock on the shaft. Also if it starts on when you remove it if there is a ding or swell at fault it will be shiny at that contact point. You should only need to go in a few inches on each end unless this is different than I envision."
________________________________________

Now that I have a better understanding of the precision needed in this case I am leaning toward a trial with your specific grit (150-200). This should bring down the axle diameter enough to slide readily. Mating of axle to the inside diameter of the bearing does not need to be exact (you all assure me). Once the pillow blocks are attached to the wood frame the inner-bearings will be tightened to the axle and the axle will then turn within the pillow blocks on their races.

[David Winer]

Another thing to consider: I assume your wheels will spin freely on the axle? Otherwise your cart won't turn well if the wheels are locked onto the shaft. When turning your cart each wheel is going to need to rotate at a different speed just like your vehicle axle/differential. If so then you don't need the axle bearings. The axle can be locked solid to the cart by other means rather than a bearing. Or am I looking at this wrong?
You are correct in pointing out that the axle bearings are not actually needed. But as you say, locking the axle to the cart "by other means" raises other design problems. Use of pillow block bearings seemed straight-forward until I ran into the problem of machined surfaces tolerances.

[David Winer]

David, did you buy a precision axle or merely a piece of cold rolled round steel?

Cold Rolled Round often isn't accurate enough to be used in a bearing...............Rod.

Rod's comment seems to address the problem on the nose.

My attempts to reduce axle diameter by abrasives were hopeless--minutes of work with coarse grits have not allowed an end of the axle to fit into the bearing, both "nominal" 3/4 inch.

The wheels arrived by UPS. Lo and behold, they have built-in bearings and fit my stainless steel rod very nicely. So, in good ol' American boy tradition I will drop back three paces and punt. I am changing the cart design to eliminate pillow block bearings, and am now going to mount the wheels on the axle simply with stainless washers and cotter pins. This scheme will also allow the wheels to turn independently to avoid skidding during tight turns, a possible problem mentioned in one comment.

When the project is finished I will send a picture of the cart to Sawmill Creek. Everyone's comments have been most helpful.