Tips/Techniques Hand fitting for an interference fit?

[slow-poke]

Let's say we have a bearing with an inside diameter of 0.5000"

I want to machine a round to be an interference fit in that bearing, so perhaps finished diameter of 0.5005"

Let's say I machine that stub to 0.5010" or (0.5020" to be safe).

What is the best technique to home in on that final diameter?

Stone or sandpaper? and if so what grit?

 

[Mcgyver]

Sometimes you do what you have to do, but imo there is only one right way:, turn it to the size you need. There's ways to get a decent shot a subthou accuracy. 10ths indicator on the cross slide, tight set up and super sharp HSS so you can creep up on it would be top of my list. Anything else runs the risk of things not being round or not concentric (To the extent that matters will vary with the application of course.).

You could also burnish, that can sometimes take off a few tenths as you drive the peaks into the valleys.

I would say interference fits for bearings are perhaps the most challenging work we' do in the home shop. i.e. I'm not making light of the challenges and that special effort is needed to get good results.

Also, start with looking the up tolerances in the bearing catalogue. They'll give details for each type, class and size of bearing. Too loose, it won't hold, too tight the bearings will bind right off or won't last.

Lastly, you could use a removable cylindrical retaining compound and go on the looser side of the things. Like a loctite 641.

 

[fixerup]

For accurate dimensions, I take my time and slowly creep up to the final measurements.
Sometimes I have success taking a spring pass, it will take out up to a ~.0005" or less, the second spring pass ~ .00025" the last spring almost nothing. If I don't hit the final dimension after that i usually polish with a 440 grit or less. But like Mcgyver mention there is a good chance it will out of round. A bearing burnishing tool is on my list of tool to make.

 

[OliverTO]

If you want to take very small accurate cuts, use your compound. The lines on the cross slide dial are usually marked in .001 increments. This can be hard to judge if you want a tenths cut. If you turn your compound to 14.5 degrees, then each .001 on the compound dial moves the tool .00025 closer to the work. With a rigid set up and some preplanned setup, you can make very small and accurate cuts.

 

[David]

Second the use of HSS for the final cuts.
The nose radius of a carbide insert is the minimum DoC for that insert which can often be more than the desired amount to be removed.
Stefan Gotteswinter has an excellent video on turning precision diameters. I have watched this at least ten times and I still struggle to hit his level of accuracy.

 

[Susquatch]

I want to machine a round to be an interference fit in that bearing, so perhaps finished diameter of 0.5005"

First, recognize that tenths are not easy to do on a lathe. I second what others have said here. But perhaps I can add a few other techniques.

I've found that sharp carbide inserts for Aluminium work quite well at removing small increments of steel. However, you will be more successful if you let the tool teach you what it likes by measuring, adjust, cut, and then measure again, so you can calculate the result. And realize that this method is not linear.

You are better off using the half's method to zero in on what you want. The half's method is not for the faint of heart. You have to learn to trust yourself and practicing is a good way to do that.

A very sharp shearing tool can do wonders for taking very fine cuts with a great finish.

You don't have to have a proper burnishing tool to do burnishing at such small increment levels.

Like @OliverTO says, you can use the compound to take very fine cuts with a very sharp hss tool. 5.74 degrees is 10:1. (1 thou = 1 tenth). But I just use 6 cuz it's close enough. But very fine cuts require a VERY sharp tool and lots of patience.

When I do overshoot the target, I've used gap loktite like @Mcgyver suggested very successfully.

I've also successfully used 800 grit sandpaper for sheet metal with oil to polish off an extra tenth. You just have to ignore the look and feel and be careful to be uniform with the pressure and overlap or you will not end up with a uniform diameter. It's best to do it all at once - not back and forth.

A good fine grit hone can be used on inside surfaces.

 

[PeterT]

We were just chatting about this, link below. Not to make light of 'glues' but when you look at the specs for Loctite, this is really what they are engineered for. Not sure about the end application but the inner race of a roller bearing is not going to have a lot of torsional stress on the shaft relative to shear strength of glue.

P

Post in thread 'Meehanite Cast Iron Available Locally?'

Mar 20, 2025

Example indicating mic

• PeterT

 

[Susquatch]

Not to make light of 'glues' but when you look at the specs for Loctite, this is really what they are engineered for.

I used Loctite Gap Fill to tighten up a loose motor drive pulley on my mill. Prolly a worst case application. Still going strong after several years. Very happy with it.

 

[slow-poke]

I have a bottle of green Loctite 609

If I overshoot, I will give it a try.

Plan is to heat the stub in the oven, freeze the ballscrew, and then quickly:
Put the ballscrew in the lathe chuck, the stub in tailstock and then press them together.

Sounds like a recipe for a comedy video.

Fun project du jour.

 

[slow-poke]

Well, I didn't overshoot, I was aiming for -0.0005" and I was probably closer to -0.0007", hard to tell for sure because I was measuring the bore with a telescoping gauge and it's not super accurate. I really need to get a 0.2-1.2" inside micrometer (it's on my list).

If I had been right on, I think the hot stub plus frozen ballscrew should have gone together with minimal force. When I tried pressing the two together using the tailstock feed it got about 1/2 way and then the ballscrew just wanted to slip back through the chuck. So I jumped into plan "B" mode, pulled the assembly and stood it up on end on the prearranged 2x6 and gave it a wrap with a deadblow hammer, nothing, next up the brass hammer, a few wraps with the brass hammer and it was home.

I also put a few drops of green loctite just before, so hopefully the two are now one. I measured run out at < 0.002" so not visible to the eye.

There is no power transferred across the joint, it's just along for the ride on the tailstock end rotating in a floating bearing. The bearing on the other end has the two thrust surfaces. and the servo drive.

Why all the drama? The good but surplus ballscrew is long enough for full Z-axis travel, but too short to mount in the original leadscrew mount location so I needed to ass a few inches.

I still need to trim the stub but that should be easy.

I suppose I could pin the joint, but I'm not sure it's necessary, thoughts?

 

[Mcgyver]

good job, shouldn't be any need to pin it if you used loctite.

 

[thestelster]

I also put a few drops of green loctite just before,

Loctite sets up much quicker with heat. On a long section that's been heated up, you might find that the shaft moves in well half way, and then the loctite sets up and you're stuck. (Been there done that, won't do it again!!)

 

[fixerup]

A very sharp shearing tool can do wonders for taking very fine cuts with a great finish.

You don't have to have a proper burnishing tool to do burnishing at such small increment levels.

I forgot I had a shear tool. I made one years ago, I will look for it and park it with my go to lathe tools bits because it did work very well. With the shear tool and the top slide at 14.5', it will help with my sneak up technic to get the final measurement much better.