Angular contact bearing question

[NC Cams]

Be carefulf of vernaculars.

The bearing pictured in post one is a classic "double row conrad bearing assembly". They are built with one piece inner and outer rings. To assembe the bearing (as all such bearings are assembled) they use the "conrad process" wherein they literally plastically deform the outer rings, drop in the balls, then undeform the ring and move the balls into even spaces and then pop in a cage.

Not until or unless they find a precision way to measure the raceways, calc a way to establish EXACTLY the ball size that will generate the fixed preload for the counteracting opposing forces in the raceways from the preload, could they come up with a "preloaded" bearing.

Now, it IS possible to put in oversized balls, Iit is possible to load them with C2 clearance which has a pretty "tight" feeling but this is technically NOT a preload.

Now, if the bearnigs ARE in fact two single rows of bearings mounted as a pair, these are technically NOT "double rows". Rather, they are duplexed - yes there are two, yes they are mounted in matched pairs, NO they are not double row bearings at least as the industry defines the.

Guys have "preloaded" or more appropriately TRIED to preload double rows by running real heavy press fits. This does make them feel tighter due to the reduction of RADIAL clearance. But it does not take care of removing the AXIAL clearance which is what the member is trying to remove.

I've seen many a "duplex" bearing incorectly identified as "double rows" for the reasons already cited - sadly, this is also done by people in the industry who should know better. When/ if you REALLY want to absorb high amounts of axial thrust with NO deflection potential, the best/most appropriate way is with a high contact angle, heavily preloaded, DUPLEXED pair of A/C bearings.

Duplexed bearings have two rows of bearings but they are not NOT NOT the same as a double row conrad bearing.

[Robin Hewitt]

Originally Posted by NC Cams they literally plastically deform the outer rings, drop in the balls, then undeform the ring and move the balls into even spaces and then pop in a cage.

Amazing, so if I removed the cages and pushed all the balls to one side I still couldn't fish one out?

Conrad must have been one heck of an engineer

[NC Cams]

Unless there is a "filling slot" or the shoulders of the rings are quite low (as in/on the NON thrust absorbing side of the A/C's) yes, you could NOT remove the cage(s) and simply remove the balls without first deforming the outer ring. Try it and see for yourself.

A/C's are not assembled that way.

They put the cages on the balls and then set the inner ring atop the outer and then heat the outer ring with induction or convection heat (hot plate). When the outer ring swells up adequately, the inner ring and balls and cage "drop" into place.

If you should, perhaps shove the inner ring apart by pushing in the wrong direction, you have a 99.99999% chance of hurting the balls and/or raceway by shoving the balls over the edge of the raceway. This typically irrepairably cuts the balls and/or deforms/brinnels the raceway.

And, yes, Conrad was a pretty clever engineer as every ball bearing made uses the "conrad method" (ring deformation) to construct bearings.

When you know how conrad style bearings are made (double row bearings like that pictured in post 1 are definitely conrads), it becomes intuitively obvious that axially preloading them is essentialy all but impossible to do properly or consistently.

For the application in question, buy a pair of properly sized A/C's that are already preloaded (the heavier the better), bolt them in so that the offsets are proper, tighten them down until the offset rings touch and go to town machining.

You can buy some unpreloaded bearings and shim them (PITA to do if you don't know what to do or how) or you can buy some A/C's and have them preloaded by a rework house. One such house is KAF MFG in Stamford CT.

They do exceptionally good work, I have NO affiliation to/with them and they are NOT necessarily cheap. The work is STELLAR - I have my bearing preload work done there and have NEVER been disappointed.

Send them your bearings, tell them how much preload you want, whether you want DF or DT or DB mount and pay the bill when it comes. They should be ready to go out of the box - simply wash carefully, do NOT air spin them, grease them and install them properly.

[aubrey]

Hi all
Why are "ball" bearings used in situations where the bearing has to take a not inconsiderable loading sideways?
Surely taper roller bearings are far more suited to the job.
OK, the design of the assembly would be a bit more complicated having to have an add a method of "adjusting" the play and seperate seals but apart from that it should be a far better solution.
Furthermore, roller bearings are able to take far greater loads in the direction we are putting them in a leadscrew situation and would be infinitely adjustable in the way that they were designed to handle.
I'm not sure that ball bearings were actually designed to be jacked together to reduce slop. Doing so is a short term thing and causes accelerated wear leading to more "slop".
Also, due to the larger contact area, rollers would last forever if no dirt got in.

Just a thought
Aubrey

[Rube]

Originally Posted by aubrey ...taper roller bearings are far more suited to the job...

Thats the stance I originally took myself. I made a very meager attempt at finding some small tapered roller bearings, but couldn't find anything small enough. Again, just a very cursory look though. I then decided that for this particular application that was overkill anyway when a set of $10/pr 40* A/C ball bearings were available. I also think that part of the reason would be that there is more power needed to turn a tapered roller?

NC...what is "matched" when two single row A/C bearings are installed in a duplex arrangement? For my Y axis in my X3 conversion I was going to use two of the 7200B's back to back (or is face to face better for any reason?) in a fixed /floating arrangement (the far side of the screw left hangin'). Is there a real need to "match" them in this application? Why would they not need to be matched if they were installed on opposite ends of the shaft (like I plan on doing with my X and Z)?

[NC Cams]

The reason you "match" the contact angle is so that you have the same thrust and/or radial capacity regardless of the direction of load/thrust. Check out the Fx and Fy loading factors for the bearings and you'll probably see a/the difference.

I guess you could run mismatched pairs - this is sometimes done in triplex or quads to gain radial or axial thrust capacity as needed. However, I can't say that I've seen any cases where an A and C contact angle were matche as a duplex set.

DB gives better resistance to overturning moments applied to/thru the bearing.

DF gives better allowance for misalignment that may exist to/thru the bearing.

Single fixed versus non-floating ends depends on the length of the screw and/or whether or not you have a lot of thermal growth potential to deal with. A long slender screw will stretch easier than a short stubby one so you''d tend to use one method over the other depending on the needs. It all gets taken into the mix when the machine/application is being engineered.

Tapered rollers do not have as low a rootating friction in thrust as ball bearings for a very simple reason - ball bearings absorb thrust via pure rolling whereas tapered rollers ultimately absorb thrust by the ends of the rollers rubbing/sliding against the flanges.

Simply because the the tapers have so much metal carrying load over long/wide rollers, they have a HUGE capacity for load. Just because a tapered roller has high radial capacity does not make it unilaterally better/best for any specific application, especially axial thrust in ball screws.

[Rube]

Thanx, I appreciate your input...I think I'm on my way

[aubrey]

Originally Posted by NC Cams Tapered rollers do not have as low a rootating friction in thrust as ball bearings for a very simple reason - ball bearings absorb thrust via pure rolling whereas tapered rollers ultimately absorb thrust by the ends of the rollers rubbing/sliding against the flanges. Simply because the the tapers have so much metal carrying load over long/wide rollers, they have a HUGE capacity for load. Just because a tapered roller has high radial capacity does not make it unilaterally better/best for any specific application, especially axial thrust in ball screws.

Thanks for that. Makes good sense for the home hobby mill seeing as they are not pushed as hard as the commercial ones.
May be an alternative if you are for ever having to adjust for excessive backlash AND you can locate suitably sized units.

Go Well
Aubrey