High-speed coupling? 1

[McLeod]

I'm searching for a small (~.187 in OD) flexible coupling for a low-torque (~2 oz-in), high-speed (80,000 rpm) application in which the angle between the rotating shafts ranges from 0 deg. to 20 deg.

Of the flexible coupling options that I've explored so far (U-joint, ball, bellows, spring, jaw, Oldham, Schmidt, flex-shaft, etc.), there are many options of workable size which can handle the torque, but none that can come close to meeting the speed requirement even when run in-line, much less at 20 deg. angulation.

Does anyone know of a component which meets the specs?

 

[CouplingKing]

There is no standard product available to do this duty.

Rexnord offer a metal disc type coupling from their BSD series that can cope with 55000rpm, but it is limited to about 1 degree angular misalignment. They are quite specialised but they may be able to help.

The only couplings used for misalignments above 10 degrees are universal joints, and these normally can't handle the speed.

Goodluck

CK

 

[McLeod]

CouplingKing - thanks for your feedback!

One thing I should add is that the duty cycle is intermittent - one minute on, one minute off, over approx. 60 min. for one "use". It has occurred to me that the vendors' published ratings may carry an assumption of continuous duty over longer periods. I've asked one vendor (whose products are rated to 20,000 rpm) to clarify the test criteria, but have not heard back yet.

Is there an applicable standard for rating the top speed of couplings?

 

[CouplingKing]

Catalogue ratings are based on an acceptable service life. Coupings are generally expected to outlast gearbox bearings, with the replacement of occasional wearing parts (e.g. rubber spiders in jaw couplings).

The short service life will give you considerably more options for both speed and misalignment. If a metal disc coupling is misaligned beyond it's normal operating range, the discs take on an S shape when viewed from above. The high stress points are where the S shapes start and finish, this is also where the discs will flex during the rotation. Hence fatigue failure is the expected failure mode. With the short life time, you may get away with it, the manufacturer will have to advise you.

Mayr make a disc coupling of a similar type, where the discs are Star shaped, allowing significantly more misalignment capability. I think they call it a DS-W. DS is the standard high performance metal disc coupling, I think the W variant offers bigger misalignment, but I am not sure about the speed. Their website is

http://www.mayr.de

and follow the links to the English language section.

Hope it goes well.

CK

 

[amorrison]

some maybe options

1. electric generator->electric motor power transfer
Coupling via wires (and suitable electronics?)
2. gear down the input rpm - go through a "regular" coupling and then gear up again
3. do gaseous couplings exist? This is how low rpm power is extracted from very high rpm jet turbines
4. Maybe a gaseous viscosity friction coupling - two impellers "connected" by gas friction using a bellows container enclosing the system.
5.magnetic coupling system?

 

[EdDanzer]

Amorrison4,
I like your magnetic coupling idea, but mass and size could be a problem.
A hex socket and hex ball ( on the end of a hex wrench) might work. Picture a socket head cap screw and a ball hex allen wrench, and a dry lub.

 

[Warpspeed]

Just a stab in the dark, but how about a short lenth of flex cable ?

 

[fredt]

Because the dimensions are so small and the torque very low it is likely that something quite simple will work - the 20 deg angle is difficult however and the geometry of the axes intersection point will also be significant. A magnetic coupling is probably ruled out because the halves have to be in close proximity.

I would try a helical spring set up to tighten with rotation. It may be a better option than the flex cable suggested which is also worth considering. Although the speed is high the very small diameter means that centrifugal forces will not be great. You will need a bit of luck to avoid instability problems. On that scale it should not be an expensive exercise to test various options on a simple rig. I much doubt that you will find a stock design to do the job.