Source for precision spindle for testing ABEC 7/9 bearings

[rhb]

I am trying to find a source for a 25 mm x 300 mm spindle suitable for testing ABEC 7 & 9 class bearings prior to installation.

One of the two precision ground test bar companies I contacted said they could not provide such a bar accurate to 0.0002" TIR and straightness. I've not yet heard from the other.

I've been using a Morse taper test bar from India and clocking about a tenth TIR and vertical error between centers on a small lathe. But with the bar in cantilever I measure 1' 2" of arc misalignment between the bar and the axis of rotation which I attribute to the spindle bearings. However, I don't have a means of verifying the bar other than with the bar in the spindle and supported by the tailstock. The test between centers suggests that the bar is very good. Testing the bearings seems the natural next step.

I bought a pair of "ABEC 7" bearings for $60+ each, but when they arrived I discovered they were from China. Before I install them, I'd like to make sure they are better than 1' of arc.

Centerless ground drill rod is specified at 0.0005" TIR. I'm looking for something better.

Reg

 

[Bakafish]

There's a lot going on here, the first point I'd make is leaving out the exact bearings (identified by the standardized numbering system) you are wanting seems like a serious omission. Saying the new ones were 'made in China' tells us nothing about who manufactured them or if they are any good, what actual research did you do before buying them? You have also said nothing about the lathe itself, the precision you seem to be chasing is only going to matter on a machine that has the rigidity to take advantage of it. You mention a small lathe (I'm assuming this is the device in question) and you are talking about micron precision, I think you're going to need to explain why you think this pursuit makes any sense at all.

And assuming (for the sake of argument) that the morse taper test bar is infinitely precise, the fact that you are seeing good numbers between centers, but not when seated would lead me to check the matching taper of the lathe spindle before jumping to the conclusion it is a bearing issue. Without knowing what machine you have, the probability is - as with most machines - the existing bearings in it are likely the most accurate thing about it.

 

[rhb]

None of the above is remotely relevant to my question.

Where can I get a single bar 25 mm x 300 mm which is ground between centers to +0.0 -5 um tolerance? Sandpaper to size is not acceptable.

 

[Richard King]

If I were you I wouldn't waste my time and money on those bearings.

mounting precision angular contact bearings skf - Search Videos

www.bing.com

 

[Cole2534]

None of the above is remotely relevant to my question.

Where can I get a single bar 25 mm x 300 mm which is ground between centers to +0.0 -5 um tolerance? Sandpaper to size is not acceptable.

Have you tried any tool holder makers? I know Maritool offers a 40mmx250mm, that is +/- 3 microns, perhaps they'd do a custom?

 

[Cole2534]

If I were you I wouldn't waste my time and money on those bearings.

mounting precision angular contact bearings skf - Search Videos

www.bing.com

No joke, the gaging will eclipse their cost ten fold.

 

[Richard King]

No joke, the gaging will eclipse their cost ten fold.

If the steel is a low grade or the heat treating is lousy, mounting cheap and it sounds unknown to the OP the bearings were made in China? Why go to all the trouble finding the TIR if the failure rate is going to be high. Buy a good reputable bearing and be done with it.

 

[jim rozen]

"Where can I get a single bar 25 mm x 300 mm which is ground between centers to +0.0 -5 um tolerance?"

Once you press these onto the mythical bar, and then press them back off, chances are they're not ABEC9 anymore. Also you'll need to have more than just the bar, you'll need
to constrain the outer races as well to develop the correct preload. So a precison bore with clamping setup is needed. Why not just install them?

 

[rhb]

I'd forgotten the enthusiasm here for seizing on irrelevant details to offer unneeded "advice". I should have stuck to:

"Where can I get a single bar 25 mm x 300 mm ground between centers to +0.0 -5 um tolerance?"

My big mistake was saying what it was for. So I got 6 replies telling me things I was already acutely conscious of. And nothing useful such as "XYZ does custom grinding work to those tolerances."

 

[Richard King]

With that attitude ------ Goodbye

 

[Cole2534]

I'd forgotten the enthusiasm here for seizing on irrelevant details to offer unneeded "advice". I should have stuck to:

"Where can I get a single bar 25 mm x 300 mm ground between centers to +0.0 -5 um tolerance?"

My big mistake was saying what it was for. So I got 6 replies telling me things I was already acutely conscious of. And nothing useful such as "XYZ does custom grinding work to those tolerances."

I gave you a relevant suggestion. What else do you want, a part number?

 

[rhb]

Cole2534:

*Every* tool holder maker *might* do a custom. I was hoping for the name of a company that was known to do customs. But yours was the only post that at least addressed the sourcing question.

Richard King: Thank you!

 

[Robert R]

The Danly catalog lists ground between centers case hardened guide posts in both metric and inch sizes. The centers
of the posts are intact after finish grinding for the designs without center mounting bolts. The plus or minus tolerance on the diameter may not be useful for your application. I have used the inch size posts with centers as alignment tools. The posts are available with and without a clamping collar.

Danly IEM

https://file.daytonlamina.com/pdf/DS562_Pins_and_Bushings-Metric.pdf

 

[rhb]

Robert R: Thank you! I'll look into the details to see what the tolerances are, but those are generally very tight tolerance items. Don't know why I didn't think of them.

Edit: I don't yet know the tolerances on the stock parts, but am sending an inquiry. In any case, they do custom work and have a catalog sheet for that:

https://www.danly.com/PDFs/SpecialComponentsFlyer.pdf

Given my requirements, I'll probably need a custom bar. Finding a manufacturer of parts with similar process requirements who advertises that they do custom work is a huge help and greatly appreciated.

 

[dana gear]

The Danly catalog lists ground between centers case hardened guide posts in both metric and inch sizes. The centers
of the posts are intact after finish grinding for the designs without center mounting bolts. The plus or minus tolerance on the diameter may not be useful for your application. I have used the inch size posts with centers as alignment tools. The posts are available with and without a clamping collar.

Danly IEM

https://file.daytonlamina.com/pdf/DS562_Pins_and_Bushings-Metric.pdf

We have used there plain bearings and caged ball/shafts in a number of our die sets, always excellent and precise to listed spec.

 

[rhb]

Robert R and dana gear:

Gentlemen,

I'd like to express my great appreciation for your replies. I'm interested in alternative suppliers if anyone wishes to suggest one based on personal experience, but Danly IEM seems to be exactly the sort of company I was looking for in my original post. And it seems unlikely anyone can top their bona fides.

As I'm almost certainly going to need a custom bar in any case, I'm going to design a bar from scratch for the task. It appears that if one half of the bar has a taper from say -.0005" to +.0005" over 3" , then the reverse taper followed by a 6" cylindrical portion which is 0.001" undersize I should be able to easily install and remove bearings for inspection.

I'd like to emphasize this is only a preliminary estimate. I need to analyze errors, installation forces, etc in detail before finalizing the design.

Some may say I'm being overly paranoid. My bearings were purchased from a reputable manufacturer who it turns out also distributes bearings from other manufacturers. But the whole counterfeit bearing situation is very worrisome. Anyone in the supply chain can slip a few counterfeits in the box and then sell the real ones on the side. At the price of close tolerance bearings, one fake in an order for 20 is hugely profitable. Woe unto the recipient of the fake in a shipment to a licensed distributor.

Counterfeit parts are probably a much bigger problem in the electronics industry. There are companies whose sole business is using per part x-ray analysis to ensure that parts for critical systems perform as specified. They do very detailed analyses of known genuine parts and then compare all incoming parts to that data before selling them to defense contractors.

Despite the job being a completely off the wall request, it's from from a long time customer, so we really want to deliver despite the demanding spec and the need to fundamentally rebuild a machine for the job.

You've been a huge and greatly appreciated help.

Thanks for taking your time to answer my question,
Reg

 

[triumph406]

Despite the job being a completely off the wall request, it's from from a long time customer, so we really want to deliver despite the demanding spec and the need to fundamentally rebuild a machine for the job.

Reg

I think if I was your customer, and read this thread, I'd wonder if I picked the right vendor for the job. I wish your customer luck, Me thinks he's going to need it.

 

[rhb]

triumph406:

Thank you for your very generous and helpful comment. Particularly your focus on my original question.

Academic research scientists are a different crowd. Having equal academic credentials is a strong plus. I can talk to them in terms of their needs in their domain and then explain the impact of their budget on the ability to meet the stated physical requirements. Moreover, I can suggest alternate designs. That's a pretty compelling alternative to dealing with a machinist who doesn't understand what the goal of the research instrument is and blindly quotes whatever was requested which is often naively over specified. While lots of universities have excellent machine shops, not all do.

 

[triumph406]

If I had to try and verify a thrust bearing I would:
1) turn up a piece of barstock in a lathe that was a slip fit into the inner diameter, tap the center of the turned part.
2) push the bearing on and use a washer etc to clamp the inner race to the turned part.
3) make a disk that can be used to push against the outer race, maybe .2"-.25" smaller then the OD od the outer race.
4) push against the outer race using a live center in the tailstock. But not directly. Maybe use a 12" long rod between the center of the disk and tailstock.so an off center tailstock axis doesn't influenec the outer race.
5) preload the outer race sufficiently to provide some preload.
6) Spin the outer race by hand, with a .0001" indicator on each face of the bearing, and the outer diameter.
7) with 30 minutes invested in making parts, and a few $'s in material you could quickly determine everything you need to know about the bearings.

This could be done on a 50 year old clapped out SB lathe if desired. By not turning the spindle while verifying the bearing your not having to account for spindle bearing anomalies

That's my 2c

 

[Cole2534]

triumph406:

Thank you for your very generous and helpful comment. Particularly your focus on my original question.

Academic research scientists are a different crowd. Having equal academic credentials is a strong plus. I can talk to them in terms of their needs in their domain and then explain the impact of their budget on the ability to meet the stated physical requirements. Moreover, I can suggest alternate designs. That's a pretty compelling alternative to dealing with a machinist who doesn't understand what the goal of the research instrument is and blindly quotes whatever was requested which is often naively over specified. While lots of universities have excellent machine shops, not all do.

That's the kind of shop I was raised in! The boss was excellent about walking back part/machine specs to deliver required results without unnecessary costs/labor. Stuff like explaining how although pin spacing needed to be +/-0.002" that doesn't mean the whole drawing needs the same precision.