Bearing precision levels

[AndySomogyi]

It would appear that the OEM Timken bearings on my Clausing 5428 are “standard precision”. Spoke with Clausing on the phone, and the bearing numbers they gave me are identical to many automotive bearings.

Using the original bearings, with some rear journal damage, I measured TIR inside the spindle bore, in the taper, outside the spindle, on the chuck mounting flange, and by the rear bearing, on the spindle shaft itself. I measure a TIR of around 1-2 tenths.

I’m not very well versed in the different bearing classes and tolerances.

On a medium duty, and medium precision lathe like the Clausing 5400, is there any advantage to going with the “precision” class bearings. The cost difference is massive, I got the standard precision ones for about $130, open box on eBay. But the precision class bearings, even open box would bump the cost up to around $1000 for the races and cones.

A related question, is it possible to measure bearing TIR with V-blocks and surface plate? I’d like to orient or ‘clock’ the two cone bearings so the out of round is in phase, i.e. to minimize spindle wobble. It’s much easier to compensate for a uniform runout parallel to the rotation axis than a wobble.

I did get lucky and found one precision class bearing race as an open box, I’ll fit that on the chuck side of the spindle.

 

[tertiaryjim]

By no means can I claim to have great knowledge of bearings but do have some basics.
A precision bearing ( class 3, 5, or 7, will always be a matched cone and race sold as a set.
They may have marks on the races which must be aligned during installation
Most standard quality lathes have class 3 bearings as I have seen.

3 tenths run out isn't bad unless you plan to be doing some really close work. If it repeats in the same position each
rotation it's in the spindle and not the bearings.
Try going to the web site of the brand bearings you want and looking at their info.
Get the info and makings of the class 3 bearings so you can recognize them when doing a search.

 

[Richard King 2]

Timken tapered cone bearings uses a different numbering system thn angular contact ball bearings. The higher the number on a Timken the worse it is. So a 3 is automotive. a 00 is precision. The reason most they use a precision on the back bearing is to eliminate wobble as if the back one runs out it multiplies the error or wobble. I get a kick out of new or rookie "rebuilders" always trying to re-invent the wheel. Don't you think a machine builder would use cheaper bearings on there machines to hold down on the sales price? I have pulled apart a few machines spindle heads and the bearings were cheap lower class bearings after some rookie tried to save money and now wonders why he gets terrible chatter in their cuts. I suppose if your working in a blacksmith shop and never want to hold tight tolerances you can buy cheap. But you get what you pay for. Here is a cute skit on bearing installs. 2 neighbors who have the same job at different factories....https://www.youtube.com/watch?v=1Hhdqh9NCps

 

[AndySomogyi]

I'm certainly no rebuilder, and no intention of ever being one, just trying make do with minimal resources, and get a functional lathe so I make parts for my other broken tools, so I can finally get to making project car parts, and try to get my performance automotive engineering business going. Rebuilding this lathe is a massive distraction for me, and I'd really like to get back to automotive engineering and fab. But, I can't afford a new lathe, so have to make do with this old and busted one.

I recently had a email conversation with an engineer at Clausing, who was nice enough to get me the original Timken bearing part numbers and precision classes. He informed me that the machine was originally built with "standard" class 2 precision bearings, and that all the original unmarked bearings that they installed were standard precision level 2. I have no idea if the bearings were ever replaced on this lathe, but all of them were unmarked Timkens, which according to Clausing indicates they they are standard precision.

According to Timken, Imperial Bearing Precision Classes are: Standard: 4, 2, and Precision: 3, 0, 00, 000.
Automotive being standard precision level 2.

The original Clausing and Timken part numbers are that I got direct from Clausing are:

Spindle bearings
Clausing number: 045-001, 045-002,
Original Timken cones: 369A and 368A
Original Timken outer races: 362B

Appearance wise, the rear race did show some pitting and damage, and there was a higher indicator runout when with the tenths indicator placed near the rear bearing. The front race looked pretty good, however there is some bearing rattle noise coming from the front. The indicator showed around 1-2 tenths runout directly on the spindle at the front.

I did get lucky and find one 362B outer race in the higher class 3 precision. Is it better to install this higher spec one in the chuck or tail end of the spindle? I drew a little diagram here with the precision bearing at the front vs. the rear. The geometry here would indicate that with the precision bearing set at the front minimizes the error near the chuck. I'm not sure if there are other factors to consider however.

 

[AndySomogyi]

The standard precision race showed up today. I came up with a way of estimating runout. I’m not sure how accurate / precise this method is, but it does seem reasonably repeatable.

What I did was lightly clamp a v block on the surface plate, and rotate the bearing snug against the block and plate while reading with a tenths indicator. The surf plate ensures the indicator always reads in the same depth, and ensures perpendicular.

The standard precision race here showed a total deviation of under a tenth. Measured out of round on all the internal and external surfaces.

I can’t guarantee the validity of this method, but hey, when you don’t have fancy measurement instruments, you make do. I also don’t know how appropriate it is clamping onto a surf plate, so just be careful.

 

[mikey]

Andy, not sure what you're trying to accomplish here. A typical bearing race will be round to much less than you can measure with a dial test indicator and even if you could measure it, it will not give you any useful information about how much run out that race will have when installed in a machine spindle. You sort of have to install it, set preload and measure run out once the spindle is in place. Just an opinion.

 

[AndySomogyi]

Andy, not sure what you're trying to accomplish here. A typical bearing race will be round to much less than you can measure with a dial test indicator and even if you could measure it, it will not give you any useful information about how much run out that race will have when installed in a machine spindle. You sort of have to install it, set preload and measure run out once the spindle is in place. Just an opinion.

I was reading through the Timken manual, and looked up the max allowed runout measurements for class 2 standard precision races, and was a bit concerned as the max seemed pretty high. So I wanted to see if these were near the max spec.

I measured the assembled spindle runout with the existing damaged standard precision bearings, and that showed about 2-3 tenths. I just wanted to make sure that this new race would not add any extra error.

This is literally my first experience with setting up precision spindles, so just trying to be as careful as I can with the limited tools I’ve got and trying to make sure I don’t screw anything up.


Sent from my iPhone using Tapatalk Pro

 

[mikey]

I totally get what you mean, Andy. I installed P5 bearings in my mill spindle and I contacted the manufacturer to make sure I had the installation procedure down before I moved on it. Bearings at this level are not cheap so I do understand your concern.

I did want to mention that races are matched to their bearings so you have to keep them paired. Cleanliness is critical, too, so clean everything really well during installation.

Good luck!

 

[AndySomogyi]

Pulled the old bearings out, and I’m not sure how to identify them. I sent pics to Clausing, and they said all the OEM bearings were unmarked, so these are not original. The Clausing engineer confirmed that these are NOT original, but he was unable to identify them.

looking at the markings, I’m guessing these might be class-0 precision bearings, as they look like a matched set. But I’m not familiar with how to read these codes.

Since I’m guessing these are precision class 0 bearings, and I’ve measured the front spindle runout at around a tenth, and the front looks very good, I think I’ll leave the front bearing alone.

The rear one here does have some damage, mostly from some idiot beating on it with a chisel. So I think I will have to replace the rear one.

with these markings, does this look like a marched precision class 0 to you guys?

 

[mikey]

I looked up a Timken 362B and it looks like a typical automotive type tapered roller bearing that costs less than $20.00 each. I don't think they fall into an accuracy class, Andy.

One idea might be to contact Timken and see if they do have accuracy class rated bearings in this size. Since you're already in there, makes sense to just put good bearings in there and I would change the pair if it was me. I have ABEC 3 bearings in my Sherline machines and P5/ABEC 5 angular contact bearings in my mill; my larger lathe has P6 bearings. Run out on all these machines is from 0.0000 to 0.0001" TIR.

Precision bearings make a difference and I would consider them.