View Full Version : How to deal with loose bearing fit?

john hobdeclipe
12-14-2012, 07:40 PM
These three pictures show the drive pulley assembly for a woodworking shaper. Two #6210 ball bearings are pressed onto the aluminum pulley, then pressed into the cast iron housing. They are restrained at the bottom of the housing by a shoulder and at the top by a snap ring (not shown). In operation, the pulley turns at about 7000 rpm


Problem is, the bearings are loose in the housing bores. Indeed, I can get a .002 feeler between the outside of the bearings and the housing bores.

So, what are my options to tighten everything back up to a tight enough fit that the bearings (one new, one old) won't spin in the housing?

ps. It occurred to me just now as I was writing this that the bores in the housing may be worn out of round and off center, toward the motor side because of the belt tension. I'll have to check that tomorrow.

Deus Machina
12-14-2012, 07:42 PM
If it's out of round, bore it and press in a sleeve.
If it's not out of round, that should be the proper clearance for a drop or two of Loctite bearing retainer compound.

12-14-2012, 08:19 PM
If it's only a .002 gap, wouldn't Loctite fill the void?

12-14-2012, 08:24 PM
The "right" way is to bore, press in a sleeve and bore again. However... if the existing bore isn't too bad you can lightly knurl the bore and press the bearing in. Even "pricking" the bore with a fine punch in many locations can suffice. All depends on how critical the application is.

Use loctite with care - you might want to get the bearings out one day and that will require heat.

12-14-2012, 08:29 PM
Housing needs to be bored and bushed and if the shaft is worn...turn down approx 0.050" or so under/
machine up a new sleeve with a 0.002 to 0.003" tight bore...heat up ...shrink fit on.
Leave the od +0.100" . wait to cool...machine to new bearing bore fit...see below chart for proper
sizes for your bearing


12-14-2012, 08:54 PM
Ball bearing shim rings,made to fit the bearing.In use the housing is bored to the bearing OD plus 2x's the shim thickness,the shim is inserted into the housing bore and then the bearing pressed in.


They come in all common ball bearing sizes,better description at the bottom of this page-

12-14-2012, 09:09 PM
I like those rings...now if it wasn't so hard to get McMaster to ship to Canada!!

12-14-2012, 09:09 PM
I occasionally rebuild antique auto generators that have loose front bearing plates. These are a stiff slip-fit (not pressed) so rather than risk having a sleeve work out, I made a small ID knurler I run on the bearing seat with the lathe. It doesn't take a lot of knurling to tighten up a bearing. The displaced metal can rise .001-.006" easily depending on fine to course knurling wheels. The bearing is pressed on the armature shaft so I've also had to knurl a few shafts then skim it on the lathe to get it back to spec.

12-15-2012, 01:04 AM
Loctite make a product specifically for this type of problem ,its usually aluminium housings that this happens with . I have done a couple of altenator front housings for my vehicles and usually get around twevle months of operation , change the bearings and replace brushes regs and back into service.

The bearings just press out of the loctite , cleanup and replace.

J Tiers
12-15-2012, 09:34 AM
Since that is cast iron, knurling etc won't work quite as well as with steel.

No doubt the bush idea is the best. Bit it will involve fixturing the housing while leaving both bores clear, and line boring the two together to fit the bearings, then putting in the snap ring grooves. It appears you have holes and a surface to use as attachment and reference, so it's possible to do without too much trouble.

Or, you could loctite them as several have suggested... The bearing seat locktite is not a "stud retainer" type, so as mentioned you shouldn't be needing a torch if you ever need to remove the bearings....

Normally only the ID or OD (ID in general) is a press fit, not both, so maybe the existing housing wore simply from the natural "creep" of the outer race which was not a heavy press fit. it does not appear that differential expansion will account for a lot of relative movement in that part, aluminum or no.

12-15-2012, 11:53 AM
Since that is cast iron, knurling etc won't work quite as well as with steel.

The OP said it was the aluminum pulley to bearing fit that was loose.

A.K. Boomer
12-15-2012, 11:59 AM
Go the loctite - it falls withing the clearance recommendations and the stuff is amazing if both surfaces are clean, stud and bearing mount, or sleeve retainer...

if you have viewing access in a section then mark it and also mark the mounting surface and keep an eye on these for assurance that things are not rotating on you in the future, this will at least save you from tearing things up further, so if it's easy access what do you have to lose?
My guess is it will be a done deal and you get to keep full housing strength - pressing in a sleeve can sometimes weaken things...

J Tiers
12-15-2012, 08:01 PM
The OP said it was the aluminum pulley to bearing fit that was loose.

Izzat so?

Problem is, the bearings are loose in the housing bores. Indeed, I can get a .002 feeler between the outside of the bearings and the housing bores.

12-15-2012, 08:16 PM
LOl... Tonight with my glasses it all seem different. ;)

yes, knurling the cast iron is not the best choice.

Rich Carlstedt
12-16-2012, 12:23 AM
On ball bearings, your press fit is the part that rotates, and the other part is a snug, but sliding fit.
So the inner race needs a press and the housing does not in your case. The bore may have had a .0005" space or clearance originally.
The "wear" if you can call that is caused by bearing wear and increased drag. This puts the outer race in a precession load.
Over lubrication or contamination for example can restrain the cage from rotating . If the cage does not rotate, the outer race must move the same linear speed as the inner race, which means it turns in it's housing, but in the opposite direction. If the balls donot want to turn, the same thing happens, except the outer race turns in the same diretion as the shaft
Anytime outer races have spun, Its best to replace the bearings (all) as you may return to the same problem shortly after the rebuild.

Having a .002 clearance is no problem. I would sooner have that, then a mis-bored housing !
First, the suggestions to use locktite bearing mount are superb !
If you work with high speed precision spindles you will find that rather than encounter ball bearins with mis-alligned housing bores, They over-size the bearing mounting bores a few thou and then loctite the bearings when the spindle shaft is installed. The bearings run much cooler, as they are installed without any "housing preload". They also are quieter , again, because the balls were not compressed in any manner , but instead are in a free state.

Since this is a high speed unit, go for the Loctite bearing Mount.

If the bore is really bad, say .010 to .020" over , then there is a product ( Loctite or Permatex ?) Called "Metal Mender" that does the same thing


12-16-2012, 04:19 AM
Ive had bearings spin in their housings due to load distorting the bearing and locking it up, but I "cure" the problem by loctiting them into their housing. Post spin damage is usually in the same order of magnitude as the o/p, and this offers a reliable cure for it even with 150+ hp going through the drive

Its quite common to split a big japanese 4 motorcycle engine thats been tuned a bit and see fret marks around the final drive bearing housing where this has happened. Usually we try to build extra bearings into the area to cure it when we go to even higher hp levels. My current drag bike has 3 extra bearings supporting that shaft area over factory.

john hobdeclipe
12-16-2012, 08:39 PM
Thanks, everyone, for all the input.

I think I'll go with locktite for now. It's cheap, it's quick, it's not going to damage anything if it doesn't work. I can center the bearings in their bores using 3 small strips of .001" shim material. Since this pulley drives the work spindle through a splined insert, I don't think it will hurt if the drive pulley is off center by a thousandth or so.

Opening up the bores and pressing in a bushing probably won't work in this application...the cast iron around the bores is only about 3/8" thick. That's not going to be very strong after removing enough for a bushing. Besides, I don't have a good way to bore it accurately, and at this stage of my development as a metalworker I don't feel confident to tackle such a job.

Thanks to weirdscience for pointing out the existence of ball bearing shim rings, or "tolerance rings." I'll definitely file that away in my memory bank.

AK, I'll do as you suggested and apply a small dab of paint to the bearings and housing as alignment marks that I can check periodically

Knurling: Are there special tools made for knurling the insides of bearing bores?

Anyways, I'm going to try to get this together next week, so hopefully the machine can do something more than just take up space. Thanks again.

J Tiers
12-16-2012, 09:54 PM
Thanks to weirdscience for pointing out the existence of ball bearing shim rings, or "tolerance rings." I'll definitely file that away in my memory bank.

There are tolerance rings, and maybe shims, and they aren't the same..... The "tolerance rings" often are crimped to form a spring gripper that takes up a lot more than 2 thou. That's the type I have seen advertised for OEM use.