Bearing Fits?

[EOsteam]

What is a good dimension for the press fit of the I.D. of a bearing on a shaft? Assume a 1" I.D. on the bearing to make it simple. What dimension should one turn the shaft? .001" over? .0005" over? Another dimension?

Thanks,
Harper

[10KPete]

It depends upon materials, length of brng, thickness of brng. All affect the tension developed in the ring/bearing when pressed on. And that determines the force required.

But, life is easy these days. Almost all the bearing manufacturers, and many distributors, have all the technical information you would ever need. All on line!

Don't guess. Ruins bearings...

Pete

[EOsteam]

The tables appear to be written by mechanical engineers specifically for other engineers. I was hoping there was a general guideline to use for a pressed fit. If it helps my bearing is a 6208 2RS deep groove ball bearing. The ID is 40 mm. The tables list a tolerance from -5 microns to +13 microns. Why the negative tolerance?

[warmstrong1955]

Welcome to the world of bearing engineers! Quite the bunch. I've worked with several of 'em over the years.
Per my SKF data, in microns, for a 40mm ID deep groove:
-5 to 11 is a J6 spec, which is for light loads. In some applications, a slip, or loose fit is adequate, and makes assembly easier.
Heavy loads, K5, I show 2 to 13. Can't get away with a slip fit in heavy loads in most applications, as the bearing may spin on the shaft.

You will also find different manufactures have a touch different tolerances.
NTN spec for J6: -5 to 23, and K5 is 2 to 25. Go figure....

A good bearing guy, who speaks English, is difficult to find.
There is a lot involved, speeds, loads, impact.....but for most stuff, shootin' from the hip usually works, if you fall within the tolerances. When in doubt, talk to your bearing house, and give him some operating parameters.


Bill

[Atkinson_Railroad]

Agree with Bill. Welcome to the world of Bearing Fits!

Bearing fit charts have grown exceedingly complicated and difficult to navigate on-line.

The older SKF Bearing Installation Guides and the even older New Departure bearing guide books were/are much easier to read and decipher.

It helps to locate a chart published for a specific industry and start from there. Even then, you’ll find asterisks related to whether a shaft is hollow or solid. From there, it goes into an art form of opinion based on actual application and environment.

As an example, here are Radial Ball Bearing Fit tolerances adapted from ABMA Standard 7, Tables 1, 2, 3, and 4 for a 6208 bearing used in an electric motor. (The fit pertains to: Shaft rotates- Outer ring stationary)

40 mm bearing bore I.D. would mate with a shaft having an O.D. tolerance of 1.5753 to 1.5749 inches.
80 mm bearing O.D. would mate with a housing having an I.D. tolerance of 3.1496 min to 3.1503 max inches.

Exceeding the tolerances above would result in a fit that’s either too loose, or too tight for the application.

Cylindrical and taper roller bearing fits having the same 40 mm / 80 mm dimensions would have a whole different set of tolerances for use in a gearbox for instance. (And sleeve bearings are a very different subject in of themselves.)

A vibrator motor used to shake out castings, or shake poured concrete would be an example of how a standard bearing chart is not useable. With a vibrator motor application, the housing fit is a press fit, and the shaft fit is free to move. Hence, a very different set of tolerances.

The chart attached can get you started, but again… it pertains to a specific industry.

John

[warmstrong1955]

Good example John,

I have had shafts and housings built to repair shaker screens and their gearboxes, and ore bin vibrators, and always have use the tightest end of the tolerance, ID & OD. If you don't, it gets ugly quickly!

Bill

[mcostello]

As a general rule I have always used .001 per inch of dimension, be it shaft or bore. If it gets much tighter it could get to be too much. This is general and I have not worked on shaker screens or ultra abusive machinery.

[Rich_Carlstedt]

There are 25 Microns to one thousandth of an inch.
That means every 2 1/2 Microns is one Tenth ( 2 1/2 Microns =.0001)
So the spec calls for your shaft to be .0002" to .0005"oversize.
This size not only locks the inner race to the shaft, but adds the correct pre-load radially to the bearing.
Anything over will led to shorter life most likely. Anything less will allow the bearing to spin, unless captured axially ( clamped)

Rich

[GlennW]

There are 25 Microns to one thousandth of an inch.
That means every 2 1/2 Microns is one Tenth ( 2 1/2 Microns =.0001)

Rounding off the least significant digit to keep it simple and relative to this discussion.

[EOsteam]

After messing with the bearing calculator on the SKF page for a 6208 2RS bearing, there are some questions on how the calculations are figured. I know that for Imperial fits a general rule to use is .00X" per inch of diameter. How are the calculations figured for metric bearings. The bearing fit calculator gives you the tolerance based on a specific bearing and a particular class of fits (Example: j5). It will spit out your maximum and minimum allowed dimensions for a particular class of fit. Is the tolerance figured .0X mm/cm of width? If one knew the numbers used to generate the calculations you could extrapolate out to any diameter and have your tolerance figures without having to look up the calculators online. Does anyone know how the metric figures are derived?

Harper

[warmstrong1955]

I don't know of a generic formula to calculate bearing fits. Besides types & styles, there are so many variables in application, and design, that it doesn't make that a good solution. RPM, load, vibration, impact, heat.....and the relative thickness of inner & outer race to the shaft & housing varies. Sometimes the same basic bearing is uses, with a larger or smaller bore, or OD. Things like that skew any formula. There are also many bearings which are and were specifically designed for one particular application. Many, are only manufactured by one company. Some, have also been introduced as a redesign, to correct an engineering problem, in a situation where the component could not be modified, or was only capable of slight modification. I could go on....but you get the idea.....
If you look at the SKF site, you can see all the different formulas for calculating many of these variables. It will make your head explode!

http://www.skf.com/us/products/bearings ... index.html

So....I refer to the published recommendations for the particular bearing(s) I'm going to use, and the fit classes they suggest, and the duty it will see, and I go for it. If I'm still not sure, I contact a bearing guy. Their heads have already exploded, but they are good at figuring out all things bearing! Most of my experience with them has been roaming around the world fixing problems and failures.
I have even done so online with SKF a few times to ask questions. Not sure if they still offer that service, that was a few years ago, and they have had a couple of website changes since.

Maybe this will help: http://www.ntnglobal.com/en/products/ca ... 2E_a07.pdf

Bill

[EOsteam]

Thanks to you all for your help. I have been reviewing the Manufacturer’s data and spec tables and have come to the conclusion that, yes, they will make your head explode.

At this point, the most relevant question would be the following: The bearing is a 6208 2RS and the application is main axles for a 1.5” scale 4-8-4 Northern. What class of fit would be the recommendation to follow? There is j5 all the way through n6. Once that is decided it’s an easy matter to use one of the online plugins to get the machining specs. On one of the pages they were recommending n6 for full-size railroad axles but thats probably overkill for the smaller scales. I look forward to hearing what you all have to say. Thanks again!

Harper