PDA

View Full Version : Press fit bearing to shaft or housing?



Jmay
11-14-2016, 04:52 PM
In a situation with a stationary housing and a rotating shaft with deep groove ball bearings wich one gets the tight fit the shaft or the housing?
I have heard it said the one that is spinning gets the press fit and the stationary gets the clearance fit.

BCRider
11-14-2016, 05:42 PM
It'll be interesting to see the other answers.

On motorcycle wheels it's the other way around. Tight outer fits with a slip through axle. But then of course it would be pretty tough to do it the other way around in that case. And while a slip fit during installation of the wheel itself the spacers and axle are then tightened which clamps the inner races tightly against turning.

cameron
11-14-2016, 07:12 PM
Generally, the load will rotate in relation to one of the members. That's the one that gets the tight fit.

Think of a rotating wheel hub, for instance, with a rather loose outer bearing ring. As the hub rotates, the upward force of the axle will press the race continually against the bearing seat, so that the outer ring rolls against the bearing seat. As the outer ring perimeter is less than the perimeter of the bearing seat, the outer ring is rotating backwards in relation to the hub, which of course means that it is rotating forward at a slower speed than the hub, so, in effect, slipping in the hub.

The forces involved are much higher than might be intuitively expected, and this slippage will occur even with force fits which are inadequate for the situation.

As for the non-rotating axle, the upwards force is not rotating in relation to it, and the drag of the bearings is the only agent tending to rotate the inner race on the shaft. Unless the bearings seize, the effect of friction drag is miniscule compared to the forces that can cause slippage of the outer ring in the hub.

Jmay
11-14-2016, 10:08 PM
Generally, the load will rotate in relation to one of the members. That's the one that gets the tight fit.

Think of a rotating wheel hub, for instance, with a rather loose outer bearing ring. As the hub rotates, the upward force of the axle will press the race continually against the bearing seat, so that the outer ring rolls against the bearing seat. As the outer ring perimeter is less than the perimeter of the bearing seat, the outer ring is rotating backwards in relation to the hub, which of course means that it is rotating forward at a slower speed than the hub, so, in effect, slipping in the hub.

The forces involved are much higher than might be intuitively expected, and this slippage will occur even with force fits which are inadequate for the situation.

As for the non-rotating axle, the upwards force is not rotating in relation to it, and the drag of the bearings is the only agent tending to rotate the inner race on the shaft. Unless the bearings seize, the effect of friction drag is miniscule compared to the forces that can cause slippage of the outer ring in the hub.

So it's sounds like you agree. If the shaft is rotating it gets the press fit. Do understand you correctly?

Mcgyver
11-14-2016, 10:32 PM
JMay, you've got it right. so.....what exactly are you designing :)

6PTsocket
11-15-2016, 06:49 AM
I have a crummy Craftsman belt sander that was making all kinds of noise. I finally traced it to the belt bearings that were slipping in the casting. Some Loctite and the noise was gone. The inner race is a slip fit.

Sent from my SM-G900V using Tapatalk

Black Forest
11-15-2016, 07:07 AM
Generally, the load will rotate in relation to one of the members. That's the one that gets the tight fit.

Think of a rotating wheel hub, for instance, with a rather loose outer bearing ring. As the hub rotates, the upward force of the axle will press the race continually against the bearing seat, so that the outer ring rolls against the bearing seat. As the outer ring perimeter is less than the perimeter of the bearing seat, the outer ring is rotating backwards in relation to the hub, which of course means that it is rotating forward at a slower speed than the hub, so, in effect, slipping in the hub.

The forces involved are much higher than might be intuitively expected, and this slippage will occur even with force fits which are inadequate for the situation.

As for the non-rotating axle, the upwards force is not rotating in relation to it, and the drag of the bearings is the only agent tending to rotate the inner race on the shaft. Unless the bearings seize, the effect of friction drag is miniscule compared to the forces that can cause slippage of the outer ring in the hub.

Is that the language of Bassoom?��

cameron
11-15-2016, 08:40 AM
So it's sounds like you agree. If the shaft is rotating it gets the press fit. Do understand you correctly?

Yes, I agree.... Except for when the load is rotating with the shaft.

Too dang early in the morning to come up with a plausible example!

garyhlucas
11-15-2016, 06:59 PM
Simple thing here. If the shaft is the moving part and is not a press fit then the shaft will tend to walk around inside the bearing because of the very slight difference in the diameters and may get progressively worse. If the loose fit is on the stationary part it won't move, in most cases. I am sure someone can suggest a case where it will!

cameron
11-15-2016, 07:51 PM
Simple thing here. If the shaft is the moving part and is not a press fit then the shaft will tend to walk around inside the bearing because of the very slight difference in the diameters and may get progressively worse. If the loose fit is on the stationary part it won't move, in most cases. I am sure someone can suggest a case where it will!

Yes.

Suppose you wanted a shaker for a gravel screen, or something similar, and you made this by mounting an unbalanced mass on a rotating shaft.

The force to be dealt with is the centrifugal force of the unbalanced mass. This force is rotating, but so is the shaft, and so the force is not rotating in relation to the shaft. It is rotating in relation to the housing and the outer races of the bearings.

In this case, the bearings must be a press fit in the stationary part, the housing, or the outer rings will walk around the housing.

There is not much, other than bearing drag, tending to turn the inner bearing races on the shaft. However, since the whole shebang is being shaken rather vigorously, you might want something a bit more secure than, for example, the fit often used for the inner races of the front wheel bearings on a rear drive car, or the rear wheel bearings on a front drive car.

Jmay
11-19-2016, 11:01 AM
Another question! This is the scenario for a pair opposing tapered rollers with rotating shaft with load on shaft and the outer race press fit in stationary housing. Can the bearings be slip fit on the shaft with shaft collar or stop on one end and nut or something on opposing side for preloading both bearings?

Sorry for all the dumb questions I am just trying to get something's straight in my mind about bearings, spindles, and such. My simple mind is currently fixated on the subject. I have read through some SFK and Timken literature but its not real clear cut. I am seeking real world experience not paper and ink experience lol!

becksmachine
11-19-2016, 01:50 PM
Another question! This is the scenario for a pair opposing tapered rollers with rotating shaft with load on shaft and the outer race press fit in stationary housing. Can the bearings be slip fit on the shaft with shaft collar or stop on one end and nut or something on opposing side for preloading both bearings?

Sorry for all the dumb questions I am just trying to get something's straight in my mind about bearings, spindles, and such. My simple mind is currently fixated on the subject. I have read through some SFK and Timken literature but its not real clear cut. I am seeking real world experience not paper and ink experience lol!

Ok, if the shaft rotates with the load, which is the opposite of the automobile front wheel scenario, it would be best to have the cones up against shoulders and tight on the shaft and preload is adjusted by moving the cups in their bores.

This does make the design somewhat more complicated as providing the means to adjust cups usually requires some sort of ring that is retained by a bolt pattern. Preload is then adjusted by shims between the ring and the housing.

Dave

Jmay
11-19-2016, 04:33 PM
Ok, if the shaft rotates with the load, which is the opposite of the automobile front wheel scenario, it would be best to have the cones up against shoulders and tight on the shaft and preload is adjusted by moving the cups in their bores.

This does make the design somewhat more complicated as providing the means to adjust cups usually requires some sort of ring that is retained by a bolt pattern. Preload is then adjusted by shims between the ring and the housing.

Dave

What I have imagined is the races pressed in the housing with lip in housing for a stop wide ends facing out . Then place shaft collar on shaft at pre determined location tightened and tacked on front side of collar then the opposite end if shaft will be threaded at predetermined location to set preload with a nut for both bearings. One end of shaft is input side the other is output. 1" shaft max 300 rpm less than .75 hp so not really extreme condition of high speed or high load.

cameron
11-19-2016, 05:28 PM
What I have imagined is the races pressed in the housing with lip in housing for a stop wide ends facing out . Then place shaft collar on shaft at pre determined location tightened and tacked on front side of collar then the opposite end if shaft will be threaded at predetermined location to set preload with a nut for both bearings. One end of shaft is input side the other is output. 1" shaft max 300 rpm less than .75 hp so not really extreme condition of high speed or high load.

Another, and more secure way for a rotating shaft, is to have the outside face of one inner race against a solid stop (collar or shoulder), a spacer between the inner races, and a nut clamping them tight. Bearing clearance or preload is adjusted with shims between the spacer and one of the inner races.

tyrone shewlaces
11-19-2016, 06:00 PM
Press fit both !!

Spin Doctor
11-19-2016, 06:04 PM
Press fit both !!

Depending on the class of the bearing used, yah they can be both be a press fit. Another consideration is just how the outer race is clamped in the bore

oldtiffie
11-19-2016, 09:35 PM
Most/many of the better bearing company catalogues have adequate guidance on these very subjects as regarded where to where "tight" and "not so tight" pressed diameters are. They also give the tolerance/limits as well as the class or type of finish.

I prefer the "know so" answers which are "objective" as apposed to "rules of thumb" and "off the top of your (my??) head" as that is "subjective".

In other words I prefer to "know so" answers instead of the "think so" answers.

But to each his own as he sees the needs to suit his purpose.

dian
11-20-2016, 07:37 AM
Press fit both !!

glue them both.

AntonLargiader
11-20-2016, 07:41 AM
In the BMW motorcycle transmissions that I repair, the bearings (tapered roller or ball) are press-fit to the shaft and the outer races are an interference fit in the case. They are shimmed either at the outer race in the case, or at the inner race on the shaft.

I have disassembled two Japanese motorcycle transmissions and found them to have a slip-fit for the outer race in the case, but I think the inner races were press-fit. Not 100% certain.

Jmay
11-20-2016, 08:20 AM
glue them both.

I have never used bearing retaining adhesive how well does it due? Is it affected by the oil and grease that it's in and degrade over time or is it good for the life of the bearing?

JoeLee
11-20-2016, 08:32 AM
Another, and more secure way for a rotating shaft, is to have the outside face of one inner race against a solid stop (collar or shoulder), a spacer between the inner races, and a nut clamping them tight. Bearing clearance or preload is adjusted with shims between the spacer and one of the inner races.I've seen this set up with a lot of duplex bearings. Shims or spacers in between the two bearings and a spanner nut on the out side squeezing them together. There are also special bearings designed to run in a duplex configuration where the outer race is extended slightly further than the inner so when they are loaded together no side load is applied to the inner races.

JL.................

JoeLee
11-20-2016, 08:41 AM
I've seen it both ways. It all depends on the design and how it will be assembled. Electric motors for example.... the bearings are usually a press fit to the shaft and a slip fit to light hand push fit into the housing. I've taken two of the same motors apart before and found that one has a loose fit in the housing and the other quite a bit tighter. I'm guessing tolerances come into play on that as they are not held to tenths in the bearing housings or end bells.
I you really want to control this everything needs to be ground.

JL.................

Spin Doctor
11-20-2016, 11:03 AM
As I stated in my previous post. A lot matters on the class of the fit of the bearings being used. Take a spindle the is using ABEC Class 7/ISO Class P4 Angular contact bearings. On the working end of the spindle the Inner Race will have a very slight press fit usually in the somewhere in the j or js fit schedule. The bore is going to be a J or JS fit. Both the bore and shaft are going to be in the somewhere around a 3 to 5 on the fit chart. On the back or drive end the shaft will be in the same range of fits while the bore is going to be an H Class fit. Most likely 3 to 5 on the fit chart. This is a 0/+ fit that allows axial expansion of the shaft due to heat. Heat is another factor that plays into fits. If the bearings are running in an elevated temperature situation the bore may be opened up and the bearing class may be of a lower grade to allow for radial expansion.