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

[prasad]

Guys

I am running into difficulty while trying to fit ball bearings. The bearing I have is about 5.5" in diameter.
I want the bearing to stay in its place once mounted on the shaft or bore but I always end up with it sliding away. As I get close to the desired diameter I reduce the cut to bare minimum steps. The shaft is still not accepting the bearing. So I take one more tiny cut using the same depth setting on my cross slide dial with a hope that it will give me the required diameter. Then I find the shaft is too loose for the bearing. I have the same exact experience in boring too. I dont know how else to do it. One moment it appears too tight and next time it is too loose. When am I going wrong? What is the right method to get it right? I do not want to use loctite to get over the loose sliding fit.

Thanks
Prasad

 

[JimDawson]

As I get close to the desired diameter I reduce the cut to bare minimum steps.

There is your main problem. Take the last few cuts at about 0.010. Take a cut, measure, take another cut at the same amount, measure again. How much was actually removed? That is amount that is also going to be removed on the finish cut if you do everything the same each time. Consistency is the most important here.

With a little practice, this will allow you to hit +/- 0.0002 every time. Don't try to sneak up on the dimension. If needed, take the last 0.0005 with emery cloth or a file. (use good safety practices when using emery or a file on a lathe)
.
.

 

[Ulma Doctor]

i don't want to over-simplify...
if the bearing's id is slightly smaller (.0005") than the shaft, you will have a light press fit and the bearing will not move without encouragement.

something else to consider,
make the shaft to .001" oversize dimension, then freeze the shaft in your freezer.
warm the bearing (unless it has neoprene seals) to a maximum temp of 250*f (for most bearings)
install warm bearing on frozen shaft and allow to equalize
baddabing- a press fit, sans the pressing

 

[Firestopper]

You need to practice your technique and confirm that your dials reflect what you removed. Like Jim stated, the last few cuts should be varified so you know exactly where your at. Other influences can result in a poor fit, such as work piece too hot at time of measuring. If too hot to touch with bare hand, then your measurement will change when piece cools to touch.
Ulma Doc also bring up a great tip for assembly of interference fitment.
Machining has and continues to teach me patience. Adding a DRO to my machine was a game changer.

 

[zmotorsports]

There is your main problem. Take the last few cuts at about 0.010. Take a cut, measure, take another cut at the same amount, measure again. How much was actually removed? That is amount that is also going to be removed on the finish cut if you do everything the same each time. Consistency is the most important here.

With a little practice, this will allow you to hit +/- 0.0002 every time. Don't try to sneak up on the dimension. If needed, take the last 0.0005 with emery cloth or a file. (use good safety practices when using emery or a file on a lathe)
.
.

^^This. It takes practice and experience but get used to your machine so you know EXACTLY how much it will actually remove when you get down to the finish cuts. If I get to the point where I didn't take quite enough and am within a thou or so, I will use a file followed by a piece of emery cloth to "sneak" up on the final dimension. I will usually take my roughing cuts and leave one or two final cuts of about .010" to do just as Jim recommended.

Also depending on how tight of an interference fit I want, knowing what the actual dimension you want is important as well. I generally don't like creeping up on the dimension merely by fit and feel. I will accurately measure the part and then depending on the actual fitment I am after I will calculate and work to an actual dimension. For light interference fits I shoot for .0003" per inch of diameter. If I want a good pressd fit I use .0013" per inch.

Maybe others will have difference numbers but those are what I use, can't remember where I came up with them from but they have served me well over the years.

Mike.

 

[Kernbigo]

Mike is write on the amount of press, .005 is to much press

 

[prasad]

Thank you guys,

Many points learnt by reading your replies. I list them below.
1. Determine the exact shaft diameter you want to get based on "light interference fit"
2. Take very small cuts
3. Measure after each cut
4. When it is close to desired diameter go more slow in taking cuts.
5. Measure but remember the temperature can cause issues. Wait for cool down if necessary.
6. Never try to insert the bearing by hand - it may not go straight and give you an erroneous feeling that there is more machining needed. In reality the shaft may be ready to take the bearing. Depend on the measurement than you insertion by hand.
7. When it read for final mounting freeze the shaft and heat the bearing. (I am not sure if opposite of this will work for inserting bearing into the bore. I remember reading somewhere that bore shrinks when heated. Would it still shrink if the bore is about 5.5"?)

Thanks
Prasad

 

[John Hasler]

Thank you guys,

Many points learnt by reading your replies. I list them below.
1. Determine the exact shaft diameter you want to get based on "light interference fit"
2. Take very small cuts
3. Measure after each cut
4. When it is close to desired diameter go more slow in taking cuts.
5. Measure but remember the temperature can cause issues. Wait for cool down if necessary.
6. Never try to insert the bearing by hand - it may not go straight and give you an erroneous feeling that there is more machining needed. In reality the shaft may be ready to take the bearing. Depend on the measurement than you insertion by hand.
7. When it read for final mounting freeze the shaft and heat the bearing. (I am not sure if opposite of this will work for inserting bearing into the bore. I remember reading somewhere that bore shrinks when heated. Would it still shrink if the bore is about 5.5"?)

Thanks
Prasad

A bore expands when heated.

 

[RJSakowski]

Guys

I am running into difficulty while trying to fit ball bearings. The bearing I have is about 5.5" in diameter.
I want the bearing to stay in its place once mounted on the shaft or bore but I always end up with it sliding away. As I get close to the desired diameter I reduce the cut to bare minimum steps. The shaft is still not accepting the bearing. So I take one more tiny cut using the same depth setting on my cross slide dial with a hope that it will give me the required diameter. Then I find the shaft is too loose for the bearing. I have the same exact experience in boring too. I dont know how else to do it. One moment it appears too tight and next time it is too loose. When am I going wrong? What is the right method to get it right? I do not want to use loctite to get over the loose sliding fit.

Thanks
Prasad

I don't believe that anyone mentioned making a spring cut yet. Any time you turn a piece you will experience deflection of the tool and/or workpiece. The amount of deflection depends upon the depth of cut, the relation of the turned length of the workpiece to its diameter, additional support (tailstock center, follower, etc.), workpiece material, and tool geometry and sharpness.

When you get close to your final dimension, make a second (spring) cut at the same setting. Since the tool loading is lighter, the amount of deflection or spring is less and you will remove a small amount of additional material. This will often bring you in to your final dimension. If not, then make a slight adjustment and repeat.

Taking a .01" depth of cut and getting close to the point where you only have a thousandth or so to go and making an adjustment without a spring cut often will give you an undersized piece. Every situation is different. After many iterations, you can develop a sense of the expected behavior. My routine for critical diameters is to stop short by a a suitable amount and make a trial finishing pass. That will tell me how the turning is behaving for a given situation. It does take a little practice to dial it in.

Bob

 

[hanermo2]

Lots of good advice.
There are 2 schools of thought .. and both are correct.

Spring cuts, ie advance nothing, will usually take off minuscule amounts of material in the 0.01 mm - 0.0005" range.
Will work well, and can get you there when you should be at the correct size, and it more/less measures correctly, but "it just wont fit".

The other option.
Take off about 0.1 (-0.06mm) / 0.004-0.005" as the minimum amount.
Calculate your near- to- last cuts, so that the last 2-3 are about this amount.

Both work, both are correct.

What happens is that everything has slop, ie minuscule clearance, unless its preloaded.
EVERYTHING bends.
EVERYTHING yields.
You add up all 3 and thats your wobble/slop/backlash. It contributes to all three.

On small lathes, the slop is usually 0.01 mm (-0.03) mm.
BUT,they easily bend about 0.05 mm (or more) with the small loads of finishing cuts. About as much as one-finger pressure .. circa 10 kg.
(A 1000 kg bridgeport will bend near/about 0.04 mm with one finger).

When doing spring cuts, the parts settle into the middle range, and typically keep taking off a little, for 2-3 passes.
When doing calculated cuts, the error is already added into the results. This error is very consistent (to less than 0.01 mm variance).

On really good industrial lathes, or cnc lathes, where there is no slop (much less than 0.01 mm), and very little bend, other techniques can be used.
There, the dials will actually tell you where the controlled point is, and you can just advance to that point, and you are done.
They are extremely consistent.