Before we start a definition from that nice Mr. Wiki Pedia so that there is no confusion: “A ball bearing is a type of rolling-element bearing that uses balls to maintain the separation between the bearing races.”
Extending the above: “A taper-roller bearing is a type of rolling-element bearing that uses cylinders to maintain the separation between the two bearing races. The working faces of the races are at an angle to the rotational axis of the bearing”
I have seen in a few topics discussions relating to the mounting of bearings in pairs, and specifically to the need to have one or more of the races floating. By floating I mean that a race is free enough to slide axially under reasonable force without there being significant radial clearance.
The rule is (for the vast majority of applications) that where two rolling- element bearings are mounted on the same shaft one of the four races must float. The floating race may be an inner race or an outer race.
Why? Well, ball-bearings are designed to take radial load, but only little or moderate axial load. Thus, if all four races were locked it would be next to impossible to assemble the bearings onto the shaft so that the axial load was not excessive (because the bearings are designed with minute clearances, and to maintain those clearances the bearings wound have to be mounted with astounding precision. Also, even if that precision could be readily achieved simple thermal expansion could generate those adverse axial loads on the bearings.
By making, say, one inner race float the ball bearings take radial load in the intended manner, but there can’t be the excessive axial load generated within the assembly that will grossly shorten bearing life and cause problems such as “lumpy” running due to brinelling.
With taper-roller bearings the reason for having a floating race is different. Taper-roller bearings are designed to take axial loads as well as radial loads. The nature of the bearing is that they are axially pre-loaded to a degree that is dependent on the magnitude and direction of the loading that they will experience in service. A typical installation instruction would be “torque nut to X Nm and back off ½ a flat”. Applying the specified torque to a nut on a known thread would apply a known pre-load. If all four races were all locked by interference fits for example, it would not be practically possible to determine the necessary torque because the force needed to push the race along the shaft would not be known. Further, if the nut was over-tightened, and if the frictional force was greater than the pre-load, there would be no way to correct the pre-load.
This brings me to the Repton RT1 Radii Turning Tool (Manufactured by Serepton Ltd.).
This device looks a bit Heath-Robinson in my view, but works well once the bearing problem has been ironed out. When I got mine, the handle worked fairly freely, but the cutter moved around a jerkily leaving visible discontinuities in the work piece. I know I am not the only person to experience this problem with the Repton.
I e-mailed Serepton Ltd. and to Serepton’s credit a replacement pair of bearings and a new shaft turned up in the post very quickly.
Whilst waiting for the bearings I stripped the Repton down, and the two causes of the problem were immediately apparent.
Firstly: The bearing pair consisted of one ball-bearing and one taper-bearing. A non-sense arrangement as the taper-bearing needs to be axially loaded to give the tool the stiffness that it needs, but the ball race that it was paired with is not intended to take such axial loads.
Secondly: All four races (two inner, two outer) were locked. Both outer races were locked by being press fits in the body of the tool and by the axial loading on the bearings. Both inner races were locked by being a tight press fit on the shaft.
When the bearings were fitted the axial load applied caused the brinelling of the ball race, which was why the handle moved in jerks. Backing off the nut didn’t help because that did not reduce the pre-load as both inners were locked by the tight press fit (not that would have solved the problem as the ball bearing was knackered by the brinelling which is irreversible).
For me the solution was simple, I replaced the knackered ball bearing with the taper-roller bearing provided by Serepton. If you have a dodgy Repton, you will have to source a taper-roller bearing of the right size. Two taper-roller bearings suitably preloaded yield a high stiffness support capable of taking loads axially and radially – just what is needed in this application.
The fix runs like this:
Knock or press the shaft through from the ball bearing side (Protect the thread in the shaft from damage by fully screwing a SHCS into it). Take great care how you support the body of the tool as the side of the tool at the bearing is thin section aluminium. I used several ground steel blocks, one of which supported the thin section on each side.
You now have a taper-roller bearing inner with a shaft. Protecting the bearing from any debris, use fine emery to polish down the shaft over half its length until the inner of the replacement taper bearing can be pushed along the shaft using firm finger pressure. (I had a spare shaft, so put that in the lathe to work on and pushed the old shaft out of the original taper-bearing using a suitable spacer and being very careful not to damage the cage). I found that once the black finish on the shaft had just about disappeared due to the polishing, then the fit of the bearing on the shaft was just right (Goldilocks agreed with me on this).
Now knock or press out the knackered ball bearing from the taper bearing side. Again, be very careful to support the tool body along the side where the wall section is thin. Throw bearing into bin.
Fit outer race of new taper roller into the tool body using a nut, bolt and appropriate spacers and washers. You all know how to do this sort of thing. Before you pull the outer race home, be very careful to CHECK, double CHECK and then triple CHECK that the outer race is facing the right way. Then get a friend to CHECK it again. Get it wrong here and you are stuffed.
What you are aiming at is shown below - two opposed taper roller bearings installed.
Grease up the new inner and re-grease the old inner. Apply a thin smear of grease to the shaft. Re-assemble.
Adjusting the pre-load is a matter of feel. Adjust until there is no axial play and then tweak up a smidgeon. You know it’s right when there is no play and the handle moves without too much effort. You’ll know it when you feel it.
Test that all is well by putting a radius or two on a bit of scrap.
Job done, and the finished tool works very nicely. Have a beer.
