How To Fit A Bearing On A Shaft Or A Bore?

[Downunder Bob]

Maybe you're misunderstanding my question. I was talking about the bearing in use heating up and expanding not heating it in oil to install it.

I am wondering why in both cases you wouldn't have a tight fit on the shaft so as the bearing heated up from use it would expand and tighten up inside the housing. Seems no different to me whichever race is moving or stationary.

Just wondering the theory behind it. Thank you.

The theory is based on the idea that with a rotating shaft then the inner race becomes part of the shaft. In the case of a wheel then the outer race is part of the wheel. It's common engineering practice.

 

[fitterman1]

@ Hellkell - The bearing in use will heat up and expand, and so will every other part in contact with it. This maintains equilibrium in the assembly during operation. That's the whole point of bearing fits.
Also I haven't seen this mentioned either - that if you leave your measuring tools on a shelf with sun streaming through a window onto them or some other heat source nearby, your tools will measure your work undersize because they are in an expanded state. The error may be minor but for bearings its crucial in a high speed environment. Keeping your instruments at an ambient temperature and recognising the fact your workpiece is different and will require stabilizing is a sign of experience and the tighter the tolerance the more vigilant you have to be.
No good in a machineshop where you turn 10 shafts for one that is correct.
Even I failed to notice the age of this thread, hopefully there will be lessons learnt.

 

[gheumann]

Prologue - the most recent post brought this thread to my attention. I didn't see the original post was many months ago. But I'll leave my answer here in case it helps someone down the road.

You certainly can't do an accurate press-fit sizing by feel. The bearing won't go on at all until the shaft is already too small. The heating and cooling method is good to know but I wonder if you're even attempting to get to a known diameter by measuring as opposed to "by feel". You can't use a digital calipers that's reads to .0005 and is only accurate to .001" for this. You need to use a micrometer. And on that topic while I own the classic $150+ Mitutoyo digital calipers I just bought an "iGaging" 0-1" digital micrometer for $40 on Amazon to see how good it is. Know what? It is worth the $$. It takes a little more care to measure accurately but I measured several different gage blocks many times and feel its real world accuracy is good to about +/-.00015", good enough for this job and good enough for most of the jobs I do.

When I'm sizing a bore or a shaft for a press fit I always try to stage my passes so that the last 3 passes are .005" each. Heavier cuts will bend the tool and work away and the diameter from one end to the other will vary. Never count on being able to take .001 or less. The spring cut is about the only way I know to do that - so if you CAN stop and measure your final diameter BEFORE moving the tool, do. If it is the right size, retract the tool. It is slightly over, start back up and take the spring cut.

But back to my original intent - if you want to size stuff for press fits you need accuracy to about .001" - and so you need measuring tools that measure to .0001" or you're going to be fighting measuring error every time.

 

[jamby]

Just a random thought but if you are using high speed steel or carbide to cut with they take very different spring cuts. Carbide cuts with a lot more pressure and will take a deeper spring that high speed. I've had problems with carbide spring cuts tearing the surface and giving more that expected removal. Also carbide will some times leave a harder surface. I prefer to use high speed for close fits.

Jim

 

[fitterman1]

Jim, I like to use CCGT tips designed for Aluminium, regardless of material (in unhardened state). They have nice sharp positive rake and give an excellent finish even on stainless. Very good for finishing cuts down to 001"

 

[Tozguy]

I remember reading somewhere that bore shrinks when heated. Would it still shrink if the bore is about 5.5"?)
Prasad

A bore expands when heated.

I believe the confusion comes from a specific situation like when the bore is in a large casting or hub. If you heat the bore with a torch from the inside the metal immediately surrounding the bore will expand. But because the metal farther out is still cooler, then the expanding metal has no place to go but inwards, making the bore temporarily smaller. But if the whole part is heated evenly or from the outside, the bore will expand.

 

[FanMan]

The theory is based on the idea that with a rotating shaft then the inner race becomes part of the shaft. In the case of a wheel then the outer race is part of the wheel. It's common engineering practice.

Correct. To visualize it better, picture if the bearing is a press fit on the shaft and very loose in the hub. As the hub rotates, the bearing would roll around inside the hub. Now picture it the other way, tight in the hub and loose on the shaft. The bearing won't move relative to the hub, and since the load direction isn't changing relative to the shaft, it won't move there either.

For a rotating shaft in a fixed hub, opposite fits, same principle.

Another aspect of it is that bearings are designed with a certain amount of internal clearance, which closes up slightly (and predictably) when the bearing is pressed onto the shaft or into the hub. If you had a press fit on both sides, it would close up too much and possibly bind. If a press fit in both is necessary, bearings with increased internal clearance are available by special order.

I'm too much of a machining novice to advise on how to achieve the required fit, but as an engineer who designs machines and specifies bearings I will say that the bearing manufacturers publish tables of recommended fits and tolerances for their bearings, and those fits will give good results in almost all situations.

 

[Downunder Bob]

Correct. To visualize it better, picture if the bearing is a press fit on the shaft and very loose in the hub. As the hub rotates, the bearing would roll around inside the hub. Now picture it the other way, tight in the hub and loose on the shaft. The bearing won't move relative to the hub, and since the load direction isn't changing relative to the shaft, it won't move there either.

For a rotating shaft in a fixed hub, opposite fits, same principle.

Another aspect of it is that bearings are designed with a certain amount of internal clearance, which closes up slightly (and predictably) when the bearing is pressed onto the shaft or into the hub. If you had a press fit on both sides, it would close up too much and possibly bind. If a press fit in both is necessary, bearings with increased internal clearance are available by special order.

I'm too much of a machining novice to advise on how to achieve the required fit, but as an engineer who designs machines and specifies bearings I will say that the bearing manufacturers publish tables of recommended fits and tolerances for their bearings, and those fits will give good results in almost all situations.

Excellent advice from FanMan. For more advanced information on bearing fits, interference and clearances email a major bearing manufacturer, I'd be very surprised if they didn't send you some literature on the subject. Also google the question "Ball race interference and clearances.