Shaft won't fit bearing

Home Model Engine Machinist Forum

Help Support Home Model Engine Machinist Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

swarf

Well-Known Member
HMEM Supporting Member
Joined
Oct 12, 2020
Messages
80
Reaction score
34
Location
Welland Ont
Why after machining a shaft to fit a bearing lol, it wont fit.
In fact I'm finding in some cases I have to go undersize three or four thou.
Used a deburring tool on the bearing and a file on the corner edge of the shaft and no luck.
What am I doing wrong?
 
Why after machining a shaft to fit a bearing lol, it wont fit.
In fact I'm finding in some cases I have to go undersize three or four thou.
Used a deburring tool on the bearing and a file on the corner edge of the shaft and no luck.
What am I doing wrong?
Doing ball bearings you working most time to ±0.000,1"
A trick can use is 1/2" wide ball bearing is under size the first 3/16" by 0.001" make easy to start.
Press fit on 1 diameter is 0.000,3"

Dave
 
Are you saying you have to remove 3 or 4 thou to get a bearing on a shaft that it previously fit on? Sounds like a lot. Do you mean tenths?
 
My local automotive repair shop has been complaining that replacement bearings won't fit the shaft the old one had been removed from. Sometimes it takes a couple tries to get a bearing that fits.
 
Are you saying you have to remove 3 or 4 thou to get a bearing on a shaft that it previously fit on? Sounds like a lot. Do you mean tenths?
No, I machined a shaft to fit a bearing. Making it .0015 undersize to fit inside a plain bearing.
It wouldnt fit. I tried breaking the edges of the shaft with a file still no good. So then machined another couple thou and went from being to tight to to loose.
 
Not an interference fit.
Dave
Myself I done hundreds of thousands of ball bearing.
The fit is very important for ball bearing. Basic rule of thumb is the shaft 5/8" needs a 0.000,1 fit.
The outside for a 1½" need a press fit 0.000,3". There are charts for press fit for ball bearings witch I used. Then tolerance is less than ±0.000,1".
My first said same thing. It not hard to do.
I have seen machinist use sand paper to get that close
Witch is a good idea for some new at doing bearing bores and shafts.
Some will use a lot of Loctite for there errors.

Dave
 
Myself I done hundreds of thousands of ball bearing.
The fit is very important for ball bearing. Basic rule of thumb is the shaft 5/8" needs a 0.000,1 fit.
The outside for a 1½" need a press fit 0.000,3". There are charts for press fit for ball bearings witch I used. Then tolerance is less than ±0.000,1".
My first said same thing. It not hard to do.
I have seen machinist use sand paper to get that close
Witch is a good idea for some new at doing bearing bores and shafts.
Some will use a lot of Loctite for there errors.

Dave
I often machine to exact size, pat myself on the sholder, take a break, get surprised that the part is undersized (yet again undersized and surprised).

a) use sandpaper b) get impatient c) skim of the smallest step you can cut d) end up under/oversized and use Loctite :cool:

At 0.003-0.004" undersize I would expect it to be a case for the Loctite approach. ( aka rattlefit)

@swarf: Any chance that you created a taper by accident? I had some taper caused by the live center for small mild steel shafts, making it difficult to press the bearing on.

Greetings Timo
 
I often machine to exact size, pat myself on the sholder, take a break, get surprised that the part is undersized (yet again undersized and surprised).

a) use sandpaper b) get impatient c) skim of the smallest step you can cut d) end up under/oversized and use Loctite :cool:

At 0.003-0.004" undersize I would expect it to be a case for the Loctite approach. ( aka rattlefit)

@swarf: Any chance that you created a taper by accident? I had some taper caused by the live center for small mild steel shafts, making it difficult to press the bearing on.

Greetings Timo

It takes time to get a good bearing fit.
I first time I was doing bearing fit
For errors I use Loctite for ball bearing fit and made knurling tool bores.
It looking at mic and each mark is divided in 10 space and realizing that how close I had to machine. It took a little time to learn. Now on any lathe I turn the bore to ±0.000,1".

You can turb your compound to 11° for finer feed.

Dave
 
I often machine to exact size, pat myself on the sholder, take a break, get surprised that the part is undersized (yet again undersized and surprised).

a) use sandpaper b) get impatient c) skim of the smallest step you can cut d) end up under/oversized and use Loctite :cool:

At 0.003-0.004" undersize I would expect it to be a case for the Loctite approach. ( aka rattlefit)
Ah, the famous "tighten the screw till it strips then back it off ¼ turn" approach ☺️😉
 
Hi Guys,

Surface roughness can make a terrific difference to a bearing fit, on both inside and outside diameters !

This is one reason that you can buy heaters in order to expand the inside of a bearing. Trouble with this is that it becomes extremely difficult to remove a bearing fitted in this way. Conversely freezing a bearing so that it drops into place makes it difficult to remove.

Many manufacturers just super glue bearings in place !
 
As I have not really much professional experience in working to close tolerances my strategy is to cut the last about five steps to final dimension in equal fine cuts that give good surface finish. On my Schaublin 102VM lathe that is 0,02 mm or slightly less then 0,001 ". Creeping in these small increments to the final dimension with measurement after each step gives me the possibility to adapt five times the depth of cut a little bit under about the same load conditions for workpiece and cutting tool. It gets me predictable to anywhere between zero and plus 0,01 mm above target dimension. Zero is fine and plus 0,01 mm is fine after a bit of sanding.
 
No, I machined a shaft to fit a bearing. Making it .0015 undersize to fit inside a plain bearing.
So it's a bush, not a ball bearing as everyone assumes.
1.5 thou is a LOT of clearance (I'd say sloppy fit) for any ball bearing.

Not using vernier calipers to measure by any chance?
There seems a lot we are NOT being told here.
 
So it's a bush, not a ball bearing as everyone assumes.
1.5 thou is a LOT of clearance (I'd say sloppy fit) for any ball bearing.

Not using vernier calipers to measure by any chance?
There seems a lot we are NOT being told here.
It's a plain bearing block made from bronze. Drilled and bored to size. I machined a shaft for it (surface finish a little course) to the dimensions mentioned and it wouldn't fit.
I think as mentioned in a previous reply. The finish may have something to do with it.
 
How and what are you using to accurately measure the bore? Having to go .003" -.004" under size on the shaft means something sure isn't right. On smaller holes .0005" under on the shaft should allow the parts to slide together as long as the surface finish on both parts is reasonable and the bushing is bored and not a simple drilled hole. I could see the problem easily happening using calipers for the internal measurements. Due to there mechanical design, they simply can't make highly accurate internal bore measurements. In fact I don't rely on any caliper measurement I.D or O.D. that has to be accurate to less than about .005". And the smaller the bore the less accurate those caliper measurements are. It's due to the internal jaws not having a knife edge and instead they both have narrow flats for longevity and durability. Even when measuring something like a grooves or slots without any radius, I don't use the calipers reading at all. Instead I use mine as a gauge to find the slot width and then lock the setting. Then measure the distance across those internal jaws with a micrometer if the accuracy requirements are high enough. Anything more accurate for something like a slot, I'd use a set of telescoping gauges, stack of gauge blocks or my two point internal mikes.

But those telescoping gauges are also highly dependent on them being very smooth, well made, burr free and just as important, require some well practiced technique. With everything correct, they can be accurate to within a few 10ths. It does take quite a bit of practice with them and measuring known bore sizes to double check you can actually get accurate and dependable measurements though. I also take at least 3 measurements just to be 100% sure of an accurate size measurement. But measuring with any accuracy even down to a few 10ths is at least an order of magnitude tougher than being certain of your measurement at .001" Any caliper of any type most of us can afford are at best somewhere within about .003". And I can say that with some assurance after double checking myself and my own calipers against proper gauge blocks. Under less than perfect conditions and with parts still in a mill vise or lathe chuck, I wouldn't bet bet a dime on less than .005" unless it's by pure luck.
 
When I started seriously machining in about 2009, the thing that held me back initially was impatience.
I expected to take a few cuts, spend about 2 minutes, and have a perfect piece/shaft.

I overshot or undershot every time.
And I was using vernier calipers to measure, which I discovered are good for rough sizing, but not good for final measurements.

And as others have mentioned, I discovered fine sandpaper, perhaps 600 grit or finer.
I eventually learned to get the shaft about 1 thou oversized, and then use the sandpaper to take it very slowly down to final size, with a lot of measuring as I went.

Like everyone else, I had to learn how to do it.
We all have had to learn how to do it, and we all had to start with that first shaft.
Nobody was born with a micrometer in their hand.
.
 
Last edited:

Latest posts

Back
Top