BK12 Bearing is tight..how to adjust..

[newpatch]

Hi All,

I was wrestling with this bearing over the weekend, and after installing it on my ballscrew rod I installed the nut but the bearing was tight. I took it apart, relubed it and played around with the oil seals. I could get the oil seals to seat better resulting in less drag, but in the end it was no where near as loose as the supporting bearing on the opposite end which consists of just 1 bearing pressed into the block.



The fixed bearing I had problems with has 2 bearings pressed together and 2 outer oil seals. There's also a little metal sleeve at each end that goes into the oil seal so when you tighten everything up using the nut it sandwiches everything together.

I have a feel that those 2 bearings have a slight amount of space between them that is causing the drag when I tighten everything...I was thinking on getting better bearings...the stock ones seems cheap...

Any idea on how to install these so that there's no drag?

newpatch

[louieatienza]

My opinion is it's the seals. There are a couple ways the bearings are preloaded, and both require a tiny space between the bearings, with a spacer (or either the inner or outer race is slightly larger.) In one way, the bearings are face to face with the outer races pressed together, the other way the bearings are back to back with the inner races pressed together.

[newpatch]

Thanks Louie,

I'm going to try to remove the seals but leave the little spacers in to see how it works.....

What I don't understand is why you need 2 bearings and 2 seals on this side of the ballscrew. The other side has 1 bearings and no seals and works so smoothly without any drag....

Why couldn't you have just 2 bearings sandwiched together? Or 1 long bearing like a thrust bearings???

I hate it when people make simple things complicated....

newpatch

[louieatienza]

Simple, the fixed end is designed for fixing the screw axially, hence the two angular contact bearings. The free end is supposed to allow for thermal expansion of the screw, hence it doesn't need to hold the screw axially.

In fact, theoretically, the bearing on the free end slould be able to slide in the block, but for some reason they're tight. I know on all my higher-end blocks this is true. So I wonder how the free end on the Chai block allows for axial play when it seens the bearing is pressed in?

Anyway, a radial bearing (which is on the free end) is not designed to take axial loads. Inexpensive or low load rating blocks have two radial bearings with a slight preload.

On my higher end blocks, there is some drag from the seal, but they spin much freer than the Chai blocks. They also cost 4-10 times more, retail.

Thinking this through, you might be able to turn down the collar or spacer slightly so that it doesn't rub against the seal as tightly. The other side shouldn't be a problen since you don't need the spacer there.

[newpatch]

Thanks again Louie,

I'll fiddle around with it some more.....

newpatch

[wildwestpat]

Hi Newpatch

The fixed end bearing to the generic BK design exists in several different qualities. The simple mounting by sandwiching the bearings together as in the sectional views should have a pair of angular contact bearings and the pair should be matched. This means they can be simply clamped up tight on both the shaft and housing. Often in the lower priced iterations of the BK matched pairs are not used on cost grounds. This can cause either excessive pre load or leaving some end float. Cheaper still are those BK housings supplied with normal deep row ball bearings. Look at any bearing supplier for 7203 bearings as individual units pairs and matched pairs prices are around $7each - $20 pair unmatched - $200 matched. It all depends on the life expectancy and the main factor is speed of rotation and frequency of reversal as well as the axial loading.

Yes there are other bearing arrangements that can give as satisfactory life. Also the humble deep grove bearing has the ability to cope with both radial and axial forces but life expectancy calculations should be done.

The use of oil seals places appreciable drag on the shaft. It is possible the BF end uses metal shields not seals and these are almost no extra load on the shaft. If you look at the data sheets on ball screws you may find that the far end can be left unsupported.

Your comment about the simplification making things more complicated is in part historical and in-part the demands for simple bolt together manufacture. Then the cost cutting starts and bits get down graded leading to either excessive pre-load and reduced life or axial slop when carried to extreme a fault that is obviously wrong on initial assembly. For maximum shaft support the AC bearings should be reversed to that shown in the section but this makes for even less tolerance to not having a matched pair of bearings.

As is too often the case it is difficult to know what one is buying.

Hope this helps - regards - Pat

PS If the quality of the bearings is suspect then you would need to proceed with care as the outer race ways are held together by the four bolts in the flange and the inner race ways by the nut and associated lock nut on the shaft. Suggest tighten both flange and nuts down gently and observe what happens to the resistance as you rotate the shaft with your fingers. The Pre load should only just be detectable.

[louieatienza]

Your comment about the simplification making things more complicated is in part historical and in-part the demands for simple bolt together manufacture. Then the cost cutting starts and bits get down graded leading to either excessive pre-load and reduced life or axial slop when carried to extreme a fault that is obviously wrong on initial assembly. For maximum shaft support the AC bearings should be reversed to that shown in the section but this makes for even less tolerance to not having a matched pair of bearings.

Pat was wondering... It was my assumption that having the outer races "facing" each other as in the drawing provided the most ridgidity, since the races are farther apart; as opposed to the the other way, where those same races would be back-to-back? For ecample, the drawing shows the pressure angle of the bb/outer race, that "X" would decrease if the bearings were reversed?

[Walky]

Just a quick note regarding problematic blocks:

If there is axial play with the whole bearings moving back and forth on the block:

Place a shim on the OD of the bearings, at one of the external sides (not between them).


If there is axial play because of angular contact bearing's mismatch (ODs don't move, IDs do):

Place a shim on the ID of the bearings, between both bearings.


DO NOT use an OD shim between both bearings, as the balls can get slightly out of place. Each AC bearings can only take force in one direction, as excessive force in the wrong direction can force the balls off their races (that's why the blocks use two of them). forcing both ID's closer together when having an OD spacer between the bearings can cause the balls to get slightly off track, damaging the races or (maybe, can't say for sure) get them stuck.

I learned about how AC bearings work the hard way, hammering one of them on the ID to get it out of a block, only to get a bunch of balls rolling on the floor. They won't fall off by themselves, but it takes far less force to take them apart than, say, a skate bearing.

[newpatch]

Walky,

Funny you should say this:

"I learned about how AC bearings work the hard way, hammering one of them on the ID to get it out of a block, only to get a bunch of balls rolling on the floor."

I had the same experience this past weekend.....

Live an learn. Fortunately I was able to put everything back together...

newpatch

[wildwestpat]

Hi louieatienza

The rigidity of the bearing pair is a function of the angle the resultant force takes through the ball. With the outer race way larger land facing each other the resultant angle is projecting outwards.

SKF - Image

Typical angles are 30 to 40 degrees and the greater the distance between the line of action and the shaft crossing points the more rigid the resultant bearing support of the shaft. But the more accurate the alignment of shaft with other components is also required.

This link shows it more clearly than I can describe it. The rest of the SKF site is worthy of a look as much of the information is applicable to bearings in general.

Regards - Pat

[newpatch]

So I think I found the problem...

If I slide the bearing on the ballscrew without tightening the nut it will spin freely, granted it's not as free and the bearing on the other side, but if I tighten the nut, it starts to get drag. So I think I need some kind of thin washer between the bearings so when I tighten the nut the bearings don't move....or, I can check to see if the bearings are seated properly. Which I did on one bearing block and helped a little.

Also, the drag I'm talking about is not a lot. If I spin the bearing holding the ballscrew rod it will spin 2-3 times around before it stops....I think it should be looser but this my first experience so I don't have anything to judge against.

newpatch

[Walky]

You will get drag when tightening the nut because of the bearings' preload and the nut/screw rubbing against the seals. Mine don't spin freely when tightened either, I don't think it will be a problem if the drag is moderate (it could actually be helpful on the Z axis to prevent backdriving).