What's new
What's new

Making very good bearing fits to 2/3 microns. Lapping ?

hanermo

Titanium
Joined
Sep 28, 2009
Location
barcelona, spain
I need to make bearing fits to very good levels.
2-3 microns. Abec9 bearings.
Timken, SKF et al say its gotta be +0.000/-0.002-4.

25.000 mm nominal shaft, mics 3-6 microns oversize.

I thought about getting some diamond grits, and making a brass lap.
To say 24.990, doable for me with a lathe.
Split lap, then squeeze the lap with opposed bolts like a 4-jaw chuck. One bolt in and one bolt out, to control it.

Use some diamond lapping compound.

Does this make sense ?
I dont have a sufficiently good internal TP grinder.

I can buy a Nakanishi, but thats 2500€.
These are prototype, about 4, in different sizes and with different bearings.

Any advice, experiences, recommendations re: lapping and or diamond compounds ?

American lapping wants about 2000$, plus taxes to me, so thats about 3000$ by the time its here.

(Internal an external, 2 sizes.)
 
that's a tenth of a thou, bit challenging. what's the part look like? through hole? big expensive casting where you've only got one shot at it?, etc.

I've done lots of lapping of bores and will I can get a very straight, round and smooth bore to tenths, but you'd have to hold you tongue just so to hit a dimension to a tenth. If you do lap, no reason it has to be diamond paste. hold the work in the lathe, make an adjustable split lap and hand hold the lap (so it floats). don't knurl the handle lol. Creep up on it through successive grits.

I'd be thinking single point bore, compound slewed at 5%, super sharp point and use the spring of the bar to creep up on it. You'll need a high end bore gauge or forget it, you need to know where you are at.

other option is find someone with Sunnen hone
 
If you have the instruments to measure what you are doing you can hit a bore dimension to tenths with a cylindrical hone if you go slow with frequent measurement cycles.
 
Hmm. Somebody check my math:

2 microns = 2 e-6 meter

2e-6 meter = 2e-3 mm

2e-3 mm = 78 microinches.

Two microns is 78 microinches.

No way yer gonna do that.
 
If you learn to read a bearing book you will find the answer.

It is under shaft and housing fits not the bearing bore tollerance
 
Hi

I did this not long ago. I bored/turned for similar fit tolerances. I was able to get within +/-0.0002" of target dimensions.

I did this by making up a table and recording the difference between dialed-in cut and actual-cut depth.
The final 4 or 5 cuts were all the same depth (about 0.003"). All the cuts before the final one were dress-rehearsals for the final cut.

By the time I got to the last and final cut, I knew how much I had to dial in to achieve the correct actual depth.
I wrote all the figures down in the table to keep track of the numbers, errors, planned cut, actual cut, difference etc.

In addition for both the bore and the shaft, I made them about 1/2" too long to create a test strip. On the last series of cuts, I would start the cut on the test strip, then stop the machine to make a measurement. This was the time to make any final adjustments before getting into the actual part. On completion, the test strips were turned down.

You have to have good instruments and technique to measure to these tolerances.
 
All good and fine to say "two microns."

What's being measured? Diameter? Cylindricity? What's the
surface finish?

What temperature are you measuring at?
 
Many good comments.

The gages to measure this are not necessarily expensive.
A 25.000 mm +0/-1 pin gage is about 20$.

At most, I might need 10 gages, on one micron increments. 24.990 to 25.000.
Maybe a 25.002 for no go.

Temperature is a total non issue- imo, since I dont *make* the bearings, I buy them.
Measuring the bearing, I get a size of needed hole. I also *know* that the ABEC-9 bearing is 25.000 and 50.000 mm.
I also trust a pin gage.
If I can mic a pin gage, I can mic the bore. If the gage mics 1 micron undersize because its cold, so will the bore.

I need to get within 2 to 3 microns of internal / external size of bearing. This 25.000 dimension is diameter of the shaft-
.. or outer bearing size 50.000 mm.
As long as the TIR is also less than 2 microns, I am fine.
I feel confident the TIR of a bore I make is less than 2 microns.

I dont consider my surface finishes good enough to bore this on the lathe, at the moment.

I dont have a way to get these honed.
These are cartridge spindles, that I make.
Would not be practical or economical to have these done elsewhere.

I need the right size of shaft and bore, to make sure the bearings are straight and work as intended.
They must not be expanded on the shaft (shaft too big) or squeezed in the boore.

These are standard fits for Abec 9 bearings.
7205 P2 25.000 / -3 50.000 /-3
7210 P2 (50.000 mm /-3 shaft and 90.000 /-3 housing).

Can anyone recommend a toolbit that might be good enough for boring the last 10 microns ?
I dont mind buying inserts and or holders specifically for this.

I have 1 micron resolution on the x axis on the lathe, so resolution on increasing the bore is not a problem.

My goal is to develop the process, so I can do this repeatably without too much skill or feel involved.
I hope I can do this without a custom diamond tool.
 
Many good comments.



Can anyone recommend a toolbit that might be good enough for boring the last 10 microns ?
I dont mind buying inserts and or holders specifically for this.

I use Tungaloy cermets. I found they produce a really good finish when boring.
Minimum DoC is around 0.002".
 
so these are spindles you are making. i wonder how you are going to align the two bores.
 
Hi

I did this not long ago. I bored/turned for similar fit tolerances. I was able to get within +/-0.0002" of target dimensions.

I did this by making up a table and recording the difference between dialed-in cut and actual-cut depth.
The final 4 or 5 cuts were all the same depth (about 0.003"). All the cuts before the final one were dress-rehearsals for the final cut.

By the time I got to the last and final cut, I knew how much I had to dial in to achieve the correct actual depth.
I wrote all the figures down in the table to keep track of the numbers, errors, planned cut, actual cut, difference etc.

In addition for both the bore and the shaft, I made them about 1/2" too long to create a test strip. On the last series of cuts, I would start the cut on the test strip, then stop the machine to make a measurement. This was the time to make any final adjustments before getting into the actual part. On completion, the test strips were turned down.

You have to have good instruments and technique to measure to these tolerances.

Great! you are within 3X the tolerance. Not much help
 
I use Tungaloy cermets. I found they produce a really good finish when boring.
Minimum DoC is around 0.002".
Your depth of cut, unless my math is really bad, is almost 30X the tolerance. Are you going to be able to hold .000078" with that?
 
Your depth of cut, unless my math is really bad, is almost 30X the tolerance. Are you going to be able to hold .000078" with that?

No but I did get within +/- 0.0002" of target.

The DOC has to be enough for the tool to cut and not rub. I was able to consistently hold a tolerance 20x to 30x the DOC.
 
Are you going to be able to hold .000078" with that

Remind us again where he said that? I'm sure I read he was doing plus / minus 2 tenths.
I was able to get within +/-0.0002" of target dimensions.
Your depth of cut, unless my math is really bad, is almost 30X the tolerance. Are you going to be able to hold .000078" with that?
Just for the record if you were sneaking up on a 2 tenths tolerance, jig boring. How much stock are you leaving on the bore for a final cut? Perhaps your getting Rozens ramblings confused with this guy?
 
IMHO i think you need to be considering surface ruffness in this, Just because the points of the id thread the lathe has cut measure that size does not make it a ideal bearing hole. Equally the lathe spindle accuracy is starting to come into play at these levels. IMHO your firmly in lapping - precision grinding territory and well outside measuring it with a std mic for shaft diameters!

Alternatively drop the bearings down a grade or 2 and match them to the levels you can hold the housing bore too.
 
If it's a prototype, then make it a neat fit when fitting by hand and Loctite the bearing in place: instant bore fit to +0.000 mm. any preloading mechanism can be on the inner race and shaft.
 








 
Back
Top