Knurled Bearing Fit: How to Engineer It

EPAIII · Aug 28, 2016

I didn't want to hijack the other thread but a number of people have suggested there and in other situations here and on other boards that a bearing can be made to fit in a hole that has been bored a bit too large by knurling the inside surface. I can see how this would work, the knurling tool will raise the metal on the peaks by pushing it out of the valleys. But, then the bearing will be sitting on just those peaks.

To me, this seems like a completely jury rigged solution. First, there seems to be no way of controlling the actual area over which the radial load on the bearing will be resting. That raised metal at the peaks will be poorly controlled in the knurling process. Just how wide will they be? How can you know? Will they extend completely across a peak or will there just be two hair lines of raised metal on the edges of it? And then just how much will the bearing actually compressed or displace or shear off? What will the resulting area that actually supports the bearing be?

And, with use and some vibration or radial loads, won't those narrow ridges of metal tend to be pressed back down into the original surface? What prevents this?

Are there any engineering guidelines or calculations for this situation or is it just total seat of the pants engineering? Is there any way of prediction the performance of such an installation over a period of time? Or just try it and see if any trouble develops?

I guess I have always had some doubts about this procedure.

Danny VanVoorn · Aug 28, 2016

Proper bearing mounting is as important as the bearing is in my understanding. The raised knurling process is as you describe and unpredictable, a machinist could bore to the proper size after knurling but as said the displaced metal may move back some so it seems like a crap shoot? It may be OK for certain classes of use but I don't think it is a cure-all for every situation. They used to knurl valve guides and pistons on engines but those engines never went the miles the new ones will, never performed as well either.
Dan

L Vanice · Aug 28, 2016

Loctite is an alternate solution, which can be used alone or in addition to knurling. The Loctite sales engineers can give free advice. Some newly manufactured assemblies are made with Loctite instead of super-close machined and pressed fits.

Note that "knurling" is an ancient fix for loose valve guides in gas engines. I think I saw a picture of the tool years ago and it looked more like a thread-forming tap than a wheel-type knurling tool. A bearing bore could be treated to a threading operation with a dull tool (or a sharp tool run backward) to get the same effect. It would not take much to tighten up a bore that is a tenth oversize. A really loose bore needs a sleeve and another try at getting the size right.

Fortunately, I have never had to try to fix such a situation, so I have no practical experience, just theory.

Larry

partsproduction · Aug 28, 2016

I did it a few times early on, but it seemed to me that it would wear loose very quickly. The addition of one of the permanent loctites, the green bearing and seal retainer I was thinking, well, now you have the B&S retainer captured by the ridges of the knurling. But then my own thinking went to sleeving. Sleeving is not a real good solution either unless there is plenty of excess metal around the hole to allow a heavy sleeve.

All the energy wasted and the worry that it won't last is a good reminder to approach the bore size with serious care.

But then, guys, we know that if the bearing is in a fixed element it should be perfectly sized so that it is just loose enough that the outer race can ever so slowly turn in use, so that the races are constantly presenting new surfaces to the wear of the rolling elements. That fit is so critical for size, concentricity and cylindricity that it's often done by grinding or burnishing, not usually by single point boring.

But the OP didn't say the bore was for a rolling contact bearing, a plain bearing commonly called a bushing may have been meant. In that case my own practice is for a heavier press fit and reaming the bushing to size after pressing, in hopes it will never spin.

Because it may spin anyway I've always liked to cut a shallow groove around the outside where the lubrication enters and several holes to the interior orbital grooves from that, that way if it spins it doesn't cut off the lube hole. It depends a lot on the use, fast heavy use? Slow low stress use?

ratbldr427 · Aug 28, 2016

You are using to much common sense and you know the result of that? You are exactly right!
I have knurled a lot of bores and shafts for various reasons and it all depends on some variables as to how successful it is.Some components are over built,some are marginal to begin with.The wider lightly loaded bearings work ok the narrower bearings don't last.Knurled shafts don't last as long knurled bores do.
Most people who do knurl jobs have no way of knowing or any way of comparing the knurled surface with the oe correct bore once the job is sent on its way.
Since most of the work I do will come back to the shop if it is wear prone I can sometimes compare them.
It is not simply a matter of not knowing how to correctly knurl a job it the nature of what a knurl does.
Not related to bearings but when I did a lot of automotive valve work one of the first steps is to repair the valve guides.Common practice was to knurl and ream worn guides which doesn't last.However lightly knurling a smooth new guide will tend to improve life(using the correct type of knurl).Worn guides wear in an hour glass shape and knurling leaves peaks with wide valleys to support the side loads on the valve.a knurled new guide has wide flats for bearing with narrow grooves which hold oil.
Very rare but I have come across some knurled new parts.
Using tolerance rings is an alternative to knurling in some cases and seems to work pretty well.I have seen more factory tolerance rings than knurled oe jobs.
Will I still knurl? I don't like to but I will,just depends!

John Welden · Aug 28, 2016

It's fine for some parts. It's been done nine zillion times. World is still goin.

Bearings in a jet engine and highly engineered/critical shit need to be right. No one is knurling that kind of stuff. It's not a problem.

People have also center punched for bearings too, including me. No one died.

partsproduction · Aug 29, 2016

One time a friend of mine was boring a steel wheel for a steam engine he was building, it was to be a press fit. Well, he went oversize about .001" I'm thinking looking back, and he showed me how he could use an acetylene torch to shrink the bore. He got the inner surface red hot and quenched it with coolant, this he did in the lathe. When he was done he had exactly the right amount of press.
I've tried that same thing many times, but far more often than working it didn't because I didn't have a hot enough torch tip to get the inside surface red hot.

But really all of this is stuff to remember when sneaking up on a bore, before it goes oversize!

Spruewell · Aug 29, 2016

Knurling a bearing fit is IMO is a bandaid at best. If there is heavy loading or vibration in the assembly those points of contact will break down prematurely allowing the bearing to skate in the housing. This typically results in disaster. I have also seen cases where dirt got in between the bearing race and the housing causing localized deformation of the bearing race and premature failure of the bearing.
I have knurled bearing fits before, but only because I didn't have another choice.

Tyrone Shoelaces · Aug 29, 2016

First time I did it as an apprentice I got threatened with the sack ! I quote - " If I ever catch you doing that again you'll be down the road ".

I've seen it done since. I've seen guys squeeze brass bearings OD oval in the vertical borer chuck to make them fit into bearing housings. They were then bored out to size and round on the ID later on so they could get away with it.

Not a good solution.

" Loctite " the way to go in this circumstance. As I recall you could use " Locmetal " up to a clearance of 0.010".

Regards Tyrone.

RC99 · Aug 29, 2016

I have done a chinese knurl and devcon on a bearing housing to fix a flogged out bore. The machine was 40 years old, does little hours and not worth spending the $ on.

Then we had a windmill "reconditioned" by experts who knurled a ~1 1/2" shaft that is supposed to be a press fit in a cast iron hub. It lasted two years instead of fifty years+. Idiots.

As others have said. It depends.

adama · Aug 29, 2016

IME its all to how you do it, if your takeing up a few thou of worn hole its a great rebuild trick, But its very much a Knurl then bore, done right theres plenty of contact to support any load the bearing can see. If your trying to regain 5-20 thou or not boring after the knurl, yeah you will have problems.

But regaining a slight loss of fit on old parts, it can work great all the more so if you use it with loctite.

HuFlungDung · Aug 29, 2016

I've done this a few times, but only to take up a thousandth or two of looseness on a non-critical shaft. I use a fine tooth knurl because I want to get a lot of contact points. Then I press on an old bearing without any premachining, in order to mash the knurl back down towards the original size. Then I press the bearing off again and machine the fit, because it typically will have too much interference if I do not.

Jim Williams · Aug 29, 2016

Knurling to improve a bearing fit is a patch job to avoid scrapping a part. Depending on the lack of fit and use, it may have some lasting value. Knurling is properly used to provide an improved gripping surface.

JIm

gadget73 · Aug 29, 2016

I've done the center punch "fix" before. Last was on a set of aluminum wheels for a pallet jack. If the bearing is a little sloppy in there, quite honestly it just doesn't matter. I skimmed it just a CH oversize when I had to modify the wheel to fit properly. "New" part wasn't bored deep enough to fit my jack. It was a BS fix but for what it is, it was enough.

adama · Aug 29, 2016

I disagree with it being called a patch job, more than a few of the machines i maintain the shafts are little more than bright mild steel, a light knurl and the bearings a interferance fit. Yeah this may not be the way to patch highly mission critical high speed stuff, but theres a whole bunch of things out there that rotate with minimal loads it works just fine on!

metalmagpie · Aug 29, 2016

Last time I centerpunched a bore was on the aluminum end cap of a Baldor motor. It was loose enough to make noise in use. Centerpunches have held up fine in medium use for several years now. When they inevitably fail I'll try green Loctite.

How does one keep the shaft central to the bore when using Loctite?

metalmagpie

EPAIII · Aug 29, 2016

So it sounds like I feared, no real way to engineer this, just seat-of-the-pants stuff. And a lot of hope.

dave powelson · Aug 29, 2016

EPAIII said:
So it sounds like I feared, no real way to engineer this, just seat-of-the-pants stuff. And a lot of hope.

have you looked at tolerance rings as mentioned in post 5?

ewlsey · Aug 29, 2016

I made a knurling machine/fixture one time. The application was a smallish shaft that was zinc plated after machining. The plating thickness was all over. They had the shafts machined a little small and we made a dimpling machine to give the bearing something to press onto so the shaft would not spin in the bearing.

It worked fine, but the application was non critical and the OD of the bearing was actually pressed into a plastic housing.

On critical assemblies there are much better ways. Press in a sleeve and rebore, bore for the next size bigger bearing, put the bearing in a separate housing and pin it in location, spray weld if weldable, etc.

HuFlungDung · Aug 29, 2016

metalmagpie said:
Last time I centerpunched a bore was on the aluminum end cap of a Baldor motor. It was loose enough to make noise in use. Centerpunches have held up fine in medium use for several years now. When they inevitably fail I'll try green Loctite.

How does one keep the shaft central to the bore when using Loctite?

metalmagpie

Generally the aluminum endbells have a steel insert because aluminum itself is not durable enough as a slip fit for a bearing. Even the factory steel shell will wear out.

Punch marks in aluminum: wasted effort, bad fix.

Knurled Bearing Fit: How to Engineer It

Diamond

Titanium

Diamond

Titanium

Titanium

Diamond

Titanium

Hot Rolled

Diamond

Diamond

Diamond

Diamond

Stainless

Hot Rolled

Diamond

Titanium

Diamond

Cast Iron

Diamond

Diamond

Similar threads