[runchman]

I got a pair of 7201's from bearingsdirect.com for mounting the ballscrews on my clausing 8520. My clausing has bearing housings that are pretty much tailor-made for this size bearing, so I thought I was all set.

However, when I mount the bearings back-to-back or front-to-front, once the nut is tightened down so the inner races are clamped together, the bearings are still 'loose' in that the outer cages move in other than just a rotating plane (hope that makes sense).

So these bearings will obviously need a spacer to be preloaded at all. Is this expected or are these bearings just crap? I recently bought 'angular contact bearings for universal mounting', skf brand, for my mill spindle and they operated as I expected, no spacer required..

wondering if I should fabricate a outer-race spacer, guessing on nut torque to properly preload, or buy some real-live-name-brand bearings.

Thanks for any info. NC-Cams feel free to yell at me

- John

[HuFlungDung]

I don't think that the bearings are necessarily crap, you just didn't get the correct spec There are applications where the bearings would be placed at opposite ends of the shaft, and there might be some gears or something on the center section of the shaft. In such a case, then a flush ground pair of bearings is overkill, because you'll be making the preload in some other manner than locking the bearings face to face.

Do you have a return priviledge on these bearings? I suppose not if they were some high precision bearing, because you have opened the packages.

[NC Cams]

Welcome to the joys of buying mail order bearings.

Unless the bearings are/were SPECIFICALLY ground/fit for UNIVERSAL USE with a built in preload, that does not mean that the bearings will be preloaded when fit DB, DF or DT. A/C's, as made do NOT have preload - that comes ONLY via a special rework. You may have 7201's which are NOT the same as 7201DU (universal mount, no preload essentially no end play) or 7201DUL (light preload) or DUM (medium P/L) or DUH's (heavy P/L).

Some bearings are designed to be used with offset ground (different thickness) spacers - essentially, the preload is calc'd/built into the spacer offset. However to do this, someone had to do a load/raceway offset study of the particular brand and P/N of bearing to find out how much offset will generate whatever preload is desired. Definitely NOT a DIY task.

In an economy of words, at this point, it is not readily discernacle how to figure out what's going on.

Solution: take the bearings, pack them up and ship them to KAF Mfg in Stamford Ct. Tell them what you want and they'll, disassemble them, grind them to the required preload, re assemble them, mark them for proper insatallation and you'll be on your way.

I have no involvement with KAF - I use them and pay whatever price they charge. They have done bearing rework of this type for most if not all of the machine tool folks in the industry - they are good at what they do.

[runchman]

Thanks much for the info, NC. I'm wondering though if the cost of having them ground is worth it for the quality of the bearing - they are labeled 'ampec', a google search of which turns up diddly squat.

Live and learn I guess. I hate to throw more money at the project, but at the same time I hate to have a sub-standard final result for the sake of 4 bearings.

Guess I'll look into the KAF cost as well as the cost of real-live preloaded bearings.

Thanks again,

John

[NC Cams]

Buying A/C's is one thing, buying PRELOADED A/C's another thing entirely.

Reason: due to the tolerances, the ONLY real/accurate way to figure the amount to offset grind is to LITERALY load and then measure the bearings.

I know of and saw only 1 production scheme for doing that and it was at the NSK factory - they could literally preload, measure and grind the raceways with the bearings FULLY assembled and ROTATING. Most other's have to load, measure, disassemble, grind, clean and then reassemble.

The point is, for preloaded bearings, SOMEONE someplace has to do the measuring and custom grinding. For this, you or someone PAYS.

THe options: buy preloaded DU"s in the preload desired and pay dearly for them or buy generic A/C's and then pay to have them measured and P/L'd yourself.

If you want to dead load them (hang weights on them and measure the raceway offsets in microns (0.001mm") you can make unequal thickness shims and affect the preload without grinding. The PITA part is ACCURATELY loading and measuring the offset. This is QUITE a technique sensitive process so do NOT think it is/will be easy.

The cheapest/easiest fix: call and send them to KAF. I know how to do it and WON"T. LIfe has gotten TOO short....

[NC Cams]

"ampec" could have been the original customer or process that the parts were made for - you probably scored some surplus or overstocked pieces. These are what shows up at bearing clearing houses as "distressed stock" or "surplus stock". If you know what to buy, what to look for, you can get good deals. YOu can also get screwed if you DON'T know what or how to buy.....

Typically, they took generic A/C's and then turned them into preloaded parts via custom duplex grinding. The parts you scored may also NOT have been ABEC 7 certified parts. This is a common misconception, namely, that any/all A/C's are "machine tool grade" - not hardly. Ditto that with preloading - it won't be preloaded until or unless it is specifically labeled as DUL, DUM or DUH or some other such designator.

It might have a CA or CG. If CA, that usually means that there is an axial clearnce when DU mounted. CG means in some speak that it has a certain micron, not force, preload. Again, it all depends on the bearing maker.

Regardless of the issue, you now know more about A/C's than the average bear. I bet you also won't buy "deal bearings" again and expect them to "bolt in" - not anymore anyway.....

[runchman]

That old axiom 'you get what you pay for' pretty much always rings true in the end, doesn't it !

There is definitely axial play, as in easily felt by hand. This will cause you to shudder, but I'm going to try making a slight spacer for between the outer races, tighten the nut by just winging it, and see how it performs.

For the light duty hobbiest that I am, that may turn out fine. If it is less than desirable once I get up and running, I'll get some proper bearings then.

If nothing else, learning about bearings is at least interesting. I didn't know angular contacts existed before starting this cnc conversion. Always good to be learning.

- John

[NC Cams]

Using dial indicators and proper support, you SHOULD be able to figure out how much offset is in the bearing ring side when there is ZERO clearance in the bearing and the thing is shoved in its thrust absorbtion direction.

Once you see how much offset occurs with NO clearance/load, you can dead weight it with some REASONABLE preload (say 25-50 lbs). Measrue offset again.

Once you know the offset, you should be able to figure out how many turns or fractions of turns of the nut will duplicate the same load/offset. YOu can then either torque/turn or make up an appropriate shim to affect the same load/offset.

Before you do that, give a KAF a call. It may not be worth the hassle.

Keep in mind that you can put a small TON of force onto a ball bearing and not see a corresponding change in the turning torque. DON'T even think of using rotating torque as a method of quantifying preload - this is NOT a taper but a rolling ball bearing.

You can load the hell out of them to the point where you damage the bearing and NOT see a prioportional change in rotating torque.

Be careful - even though it is a hobby deal, there is no need to trash bearings if you don't have to.

[M100]

nccams is my understanding of this process correct?

Assume the bearings are to be fitted in a back to back arrangement



Fixture:

There is a test fixture that can support alternately either the inner or outer race of the bearing with the bearing race positioned on a vertical axis

The level of the bearing races can be measured with a dti to an appropriate accuracy maybe 1/10 thou? (a dti with 0.00005" resolution/0.00012" accuracy is available)

A test preload can be applied to the bearing inner race whilst maintaining access for measurement of race offset


Measurement Procedure:

The bearing is vertically supported (inverted) on the front face of the inner race - the internal clearance is thus zero

Measure the difference in height between the rear face inner race and the rear face outer race

Support the bearing (non inverted) vertically on the rear face outer race - the internal clearance is again zero

Measure the difference in height between the front face inner race and front face outer race

The two dimensions obtained are as shown in the first picture below


Now apply a specified test preload to the inner race

Measure the difference in height between the front face inner race and front face outer race

Manufacture shims that will ensure identical dimensions between the outer race and inner race on both the front and back face "i.e flush" when the bearing is preloaded to the same level as the test preload

This is as shown in the second picture below


When the bearing is assembled, with appropriate shims and preloaded on the inner ring such that the inner race and the outer race plus shims are at identical heights the assembly is effectively equivalent as shown in the third picture below.

[NC Cams]

You have the process right - up until the reassemble with shim part.

By dead loading the rings with the desired preload, you generate a ring offset. If you figure out how much ring side/facehas to be removed to make the rings flush under load, when unloaded, the raceways will be unloaded offset.

When/if you reclamp the rings back flush with each other, you've put back the preload you dead weighted them at.

The trick is, "how do I remove the metal from the rings?". Simple. Heat the outer ring and the ring will fall off the inner - no muss, fuss or force. Of course, record what needs to be recorded, grind precisely, reassemble and you should have proper preload.

If the bearings are NOT flush ground, you can simply make up unequal spacers and put whatever offset is needed to get whatever preload into unequal spacer thicknesses. Noth ways work - the trick is to dead force load the rings to figure out how much offset you need for a given/desired preload.