[kumaichi]

Hi,

New to the group and the plethera of information is amazing. Thanks for putting this site together.

I was wondering if there is a formula or some way to calculate the size of a hole so that I can press a bearing into it. I have a miniture flanged bearing that has an od of 22mm. How do I figure out what size hole I need to make sure the bearing stays put? I'm pressing it into a piece of 6061 aluminum.

Also, I've been doing reading and people have suggested freezing the bearing so I wouldn't need to actually press it into the piece, it would kind of slide in and then expand as it warms. Is this really a good idea?

Thanks in advance,

Craig

[trubleshtr]

Hi welcome to the stie, usual clearance is only a .001" per 1" for a general rule.

If you want another alternative you could drill a 22mm hole and then place a plate of aluminum flush on the one side of the bearing with a hole only as big as the I.D. or shaft size, the other side do the same thing but use a shim between the bearing and the plate to "squeeze" the bearing when the plate is bolted on. This will firmly hold the bearing, you can also control the amount of pre-load on the bearing by changing the size of the shim.

Normally you can "freeze" or "heat" a bearing to install, but heating has to be done at a controlled rate, and not with a direct flame, never heat a bearing higher than 125 deg. cels. pros use an device called an induction heater. but heating is used for installing onto shafts mostly.
You could throw your bearing in the fridge or freezer for a few minutes, ideally you want a difference of about 80 deg. cels. between bearing and housing....

good luck

[kumaichi]

trubleshtr, thanks for the quick response. That is exactly what I was looking for. I'm actually going to use two flanged bearing for each side of my piece, this way I think it is a cleaner look and helps to keep the torsional affects away from the piece.

Thanks again and have a happy new year,

Craig

[HuFlungDung]

Having bored hundreds of bearing seats for the repair of electric motors, I can give you a guideline in how the fits feel: if the bearing race and housing or shaft are exactly the same size, this will give a light press fit. Typically, a shaft is left .0005" oversize and the bearing can be heated (carefully) on a waffle iron (but use the flat side of the plates ), then assembly must be done quickly and accurately. We use exactly the same interference spec no matter what size bearing it is, perhaps a bit less for a tiny bearing like yours.

For housing fits, we strive for a hand made "push fit". This is in the range of .0003" to .0005" clearance. Again, if they are exactly the same size, it is considered tight, but not harmful. It's just a real b!tch to assemble or take apart when they are that tight. The motor rewinders themselves will salvage a housing (instead of bringing it in for machining) if it is non-critical and has maybe .001 to .0015 clearance. They use loctite on it (but only one bearing can be loctited, the other must float, hence the need for the hand-made push fit).

Timken tapered roller bearings are a different story: these can have very heavy interference on the races, but that is because the bearing preload adjustment is possible after assembly.

You know your bearing is much too tight a fit if it suddenly feels like its a bit gritty when you roll it over after assembly. If you're sensitive and have lots of experience, you can feel a ball bearing get stiffer to roll when you've reduced the internal clearance with too much interference.

Tight fits tend to score the housings/shafts too, when assembling or disassembling. So if you do achieve a tight fit, be sure to lubricate the surfaces before assembly. Aluminum is especially bad for scoring when tight fits are used.

[DareBee]

Hu's info is pretty much spot-on.
If you register on the NTN website you can use their engineering charts (which I printed and keep in my desk).
The only variation I have on Hu's post is that they recomend different fits on whether the shaft is rotating or the housing is rotating. The non-rotating component usually has 0-0.0002 clearance and the rotating member is usually 0.0002-0.0005 interference on a bearing that size.
If your bearing fit is too tight it will overheat and fail.
I also recomend a little Loctite on the running fit bore.
Freezing or heating a bearing that size will create soo little expansion/contraction that it is hardly worth the effort, but is not detrimental.

[HuFlungDung]

One thing we've had some trouble with over the years, is that some manufacturers tend to push the manufactured bearing size tolerance one way, and others push the tolerance the other way.

For this reason, we actually request the actual bearing that will be used in the application, before fitting. It is not uncommon to see a bearing that measures .0002 different than nominal dimension. This makes a big difference in assembling a heavy motor.

Most people don't realize that the difference between "tight" and "loose" is extremely small...in the order of .0001". Witness sliding a bearing up a tapered mandrel (with something like .005 taper per foot). You can barely measure the difference in diameter in the zone where the bearing makes the transition from tight to loose. I think we're playing with Vanderwaals forces here.

[DareBee]

heh heh

Youre too smart for your own good.