What's new
What's new

WEDM suitable for splitting a couple of bearing races to solve an install problem?

fgen

Plastic
Joined
Jul 25, 2010
Location
tx, usa
Hello,
I have a control shaft with these properties:
1. It has stuff welded on both ends that preclude conventional bearing installation (bearing id 1.3125 too small to pass over stuff at ends).
2. The shaft is used by a human operator and travels slightly less that 45 degrees stop to stop as the operator rotates the shaft with a lever (somewhat like a tiller)
3. It would be a good thing if the operator had a nice real anti friction bearing in place of the harsh sticky bushing solution in place now.

I'd like to experiment with taking a full compliment bearing (MS27646-42G just fyi) and making two cuts in the races so I can split each bearing race into two semi-circles. Then I'd reassemble the bearing around the shaft and hope that the limited rotational travel and careful positioning of the cutlines would allow the bearing balls to either
- not ever hit a cut, or
- if they do hit it the kerf from the cut is so small there's no noticeable detent effect.
From what I can tell from the specs and in speaking with the bearing engineer at the bearing house there's plenty of margin in thrust and radial.


So, is this something that would best be attempted with WEDM? The only other things I can think of are to use an abrasive or slotting saw like approach and then try to grind the race halves to fit up nicely. This might require the kerf be put on one side of the ideal cut line and make matched bearing races from a pair of donor races.

It could turn out the 15 or so kerf from WEDM would not work but I'd like to just try and see.

I think the first shot would be just a direct cut and hence pretty easy setup but don't know any WEDM places I can ask.

Thanks
f
 
Hi fgen:
Would you consider making the cuts at an angle?
That way when a ball rolls over the cut it doesn't hump as badly.
You could also use 2 donor bearings as you suggest, and cut them in half accurately on the wire with the offset just right so you can assemble one circular race with no gaps.
That would be the easiest way forward, but be aware the races won't be round anymore as soon as you cut them.
They will sproing from stress release.

Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
Vancouver Wire EDM -- Wire EDM Machining
 
I think I see what you're saying but I have these things on the advantage side:

1. The sproingness while certainly there is mitigated by the bearing housing on both sides (in the axial direction)
and by the shaft on the id and the bearing housing (which I build out of aluminum) (again) on the od. It will be
constrained thus - for all I know that won't be sufficient but it should help.

2. The balls can be constrained by putting a "wiper" on the outer race on one end, and on the inner race of the other end,
for each half-bearing. Arranged properly this will force the balls away from the gap on the very first 45 degree traversal, after which they will tend to
simply migrate from one end of their respective tracks (defined by the "wiper" and the gap) to the other, when the shaft is rotated, without hitting either the wiper or the gap. This
depends on only populating the bearing with something less than full compliment.

We all now wish I were more articulate, but I think all the above is real/sound...

Regards,
f


ps - by wiper I am thinking of a pin that is on a radius and prevents the balls from going past it. Any kind of metal
tab that intrudes the ball path and heards them is appropriate. I got all gaga about a pin when I realized I could
(meaning pay my betters to) EDM drill a hole for it in the races.



No matter how you slice it...hah...the bearing is going to be obround and the balls may very well get stuck.
 
By all means try it, but I highly doubt it will work. Maybe some kind of greased split sleeve.

The bearing will also most likely spring open once cut. Let us know how it works out :)
 
Hi fgen:
Hi have a quick question for you that you may not have considered yet:
Do you think it's more work to cut the components off one end, install the bearing(s) and then weld the bits back on, or do you think it's more work to split the bearings and try to make the workaround function acceptably.
After all, they were welded on before and that was good enough to do the job, so maybe removing and then re-welding them is not going to be a big deal.

Knowing absolutely nothing about what it is and what it does, it's impossible for any of us to judge, but I do know from my own experience that we can occasionally miss the blindingly obvious.

What do you think?

Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
Vancouver Wire EDM -- Wire EDM Machining
 
Have you considered a teflon lined flanged sleeve bearing? These work for slow speed applications with moderate loads where a low friction bearing is needed.
The thin aluminum housing can be epoxied into the supporting housing bore. This would eliminate the warped geometry after the saw cut.

The bearings are inexpensive. They have the additional advantage of eliminating the risk of migrating grease which can contaminate optics or sensors.


SLEEVE BEARINGS | Frelon(R) Lined - Flanged, Inch
 
Last edited:
Hey,
Yes, good point, and one of three approaches that are competing for least grossness.
The approaches are :
- cut and reweld joining the ranks of those reviled fabricators who have welded a fender panel around the heater duct work (specific recollection of an early 70s ferrari restoration...). Probably best function and low weight solution. Just have to think of a tig rig
and a saw as normal tools for any replacement required.

- cut and make some kind of bolt on reattach without compromising the strength an reliability requirements.

- put in some teflon like stuff as a bushing. Simple as it just fixes the original system (al against steel) but slightly
complicated testing as the loads on this thing go from near zero to lots and I don't know just what lots is, which is why mil spec torque tube bearings exist - to unfriction a path that might have a high line or point load in use. Not that it is a bad idea at all.

I'm currently really doing proto on all three of these approaches.

For completeness that shaft is an aileron torque tube for an unboosted side stick controller in a single engine single seat airplane. At high speeds the roll control gets very heavy and is necessasrily a one wrist problem, hence my focus on the ball bearing vs bushing. The operator is me which is why I can't fall back on get 'er done, as I'll have only me to blame if it goes badly.

Regards,
f




Hi fgen:
Hi have a quick question for you that you may not have considered yet:
Do you think it's more work to cut the components off one end, install the bearing(s) and then weld the bits back on, or do you think it's more work to split the bearings and try to make the workaround function acceptably.
After all, they were welded on before and that was good enough to do the job, so maybe removing and then re-welding them is not going to be a big deal.

Knowing absolutely nothing about what it is and what it does, it's impossible for any of us to judge, but I do know from my own experience that we can occasionally miss the blindingly obvious.

What do you think?

Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
Vancouver Wire EDM -- Wire EDM Machining
 
Is there enough room for a bearing with a larger ID that will fit over one end of the shaft and a split sleeve between the ID of the bearing and the shaft?
 
Hey,

I agree with the welding approach and I'm actually making protos of four approaches :

1. Cut and weld
2. Cut and make a bolt or clamp or something stout to allow replacing the bearings in 300 years when they wear some
3. Fancy modern plastic bushings
4. Find (not likely) or make a split bearing.

Of theses 1 is the coolest for light weight and confidence in structure both of which are first order criteria here.

2 is awaiting steel stock so I can weld up some sample gizmos.

3 I can either use a bushing in the existing Al mounting blocks or make a new block from magic slick plastic but that means I have to
satisfy the strength questions that arise. Bushing doesn't have that problem I think.

4 has been prototyped with an R18 bearing (1 1/8 " id caged ball bearing) that I cut with an angle grinder. If the balls don't hit
the gap it works well but as you can imagine the gaps are large and ugly. During normal excursions the balls don't go far enough to be a
problem but having a detent of this type is nogo event. Hence my interest in small kerf precision cuts via say WEDM.
Progress on 4 is kind of stalled due to not being able to get the local stealth (apparently) EDM shop to return calls.
I don't have a surface grinder but if I did I'd try and match up the join points with that. Don't know if that would work but
standing at a surface plate from now until the end of time does not appeal, hence the EDM hope. The OD of the R18 runs afoul of surrounding stuff hence the fascination with the thin section torque tube bearing.


Just FYI the shaft is a 1 1/8 " dia .035 wall 4130 aileron torque tube for a single seat single engine airplane. The shaft is the main part of a sidestick control system and is hence operated by one hand/wrist through a lever (control stick) about 8 or 10 inches long.
It has to work through a range of torque of near zero to lots, where I don't know how much lots is but at higher speeds the controls get very heavy in roll and invite binding as the current bushings are not anti friction elements in addition to suffering misalignment issues due to airframe flex.

Thanks very much for the suggestions - the only things that ever actually work are the simple ones and by then, and only then, they are
glaringly obvious.

Regards,
f

Hi fgen:
Hi have a quick question for you that you may not have considered yet:
Do you think it's more work to cut the components off one end, install the bearing(s) and then weld the bits back on, or do you think it's more work to split the bearings and try to make the workaround function acceptably.
After all, they were welded on before and that was good enough to do the job, so maybe removing and then re-welding them is not going to be a big deal.

Knowing absolutely nothing about what it is and what it does, it's impossible for any of us to judge, but I do know from my own experience that we can occasionally miss the blindingly obvious.

What do you think?

Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
Vancouver Wire EDM -- Wire EDM Machining
 
Hey,

I've tried twice to post this but I must be doing something wrong.



This in response to Marcus and cut/reweld bearing.


Probably will do this if the split bearing thing doesn't work out.

I'm currently in prototype for four approaches.

1. Cut install reweld. This is the lightest and arguably most secure approach.
There are issues as to making sure the reweld doesn't move things about
too much, either by misregistering or warp. Both of those things
are quite tractable. There's also the stigma of
having welded an otherwise replacable bearing into the mix.


2. Cut install and make some kind of fastenable gizmo to allow the bearings to
be serviced ( replaced ). I have come up with two means of doing this
but they are either complicated or possibly not strong enough. They also
weigh more than option one. It's very unlikely the bearings will ever
need to be replaced anyway.

3. Teflon or similar bushing. Uncertain if the friction reduction under load
is sufficient - very much subjective, but I am goint to try it.

4. Find or construct a split bearing. Fun if it works but really a bother.
I've made one proto with an R18 caged bearing but it is really clumsy
and is too large OD to fit where it must. I just cut it with an
abrasive wheel on an angle grinder. The kerfs cause a lot of detent
but the thing works well if you don't go too far and stick a ball in the
gap. Even if you to it is easy to undetent it from that position with
a little extra force by hand, but
even that is not acceptable in this application.



FYI the shaft is a 1 1/8 in diameber .035 in wall 4130 aileron torque tube
forming a major part of a sidestick control for a single engine single
seat airplane. It is rotated by a lever (the control stick) that is
8 or 10 inches long and of course that is done by one hand/wrist. It has to
operate from about zero torque to lots more, where I don't know how
much lots is. At higher speeds the roll controls get very heavy hence the
focus on rolling element bearing - the line contact (in principle) of a
bushing makes them prone to being creapy under load. The bearing of
(my) choice is a MS27646-42G torque tube bearing which fits
(od 1.75, id 1.3125) and is for
just this kind of application, except that it is not in any way restricted
to oscillating mode, until I cut it of course.

I very much appreciate the suggestions by the way. I'm searching for
glaringly obvious.

Also thanks for the pointers to the plastic bushings.

Also someone mentioned large ID and bushings to fit bearing. That would be
great except for the tight clearance required around the control shaft.



Regards,
f
 
FYI the shaft is a 1 1/8 in diameber .035 in wall 4130 aileron torque tube
forming a major part of a sidestick control for a single engine single
seat airplane.
f

LOL are you the pilot?

I mean, what could possibly go wrong?:eek:

If you were to split both inner and outer races, and reassemble the splits 90deg apart it would probably work without jamming.

If I was **** enough to try that in a single engine plane, I would set up a rig that actuated the control on the ground using a DC motor. Actuate at the approximate speed you would while flying, have a breakable fuse between the motor and torque tube, so if it jammed it would break the fuse. Actuate the torque tube with bearing maybe 5000 times, if it didn't jam and break the fuse, then it might be reasonable to think it would be ok to try in the air.:eek:
 
Oh my. Should have said airplane in the first post.
Race car or road race or motocross bike the same.
Feedback or smooth here matters so much on control bearings. A normal bearing split with EDM or other cut method will not give this.
Bob
 








 
Back
Top