Key/Keyway for Torque Transmission

[ddelaiarro]

I have the following set-up:

A gear is mounted to a ferrofluidic feedthrough (vacuum to atmosphere seal that allows rotational motion to be transmitted via a ferrofluidic vacuum seal). A shaft (1.5"-2.0" diameter) travels through the axis of the gear and also moves 0.875" up and down.

I need to transfer the torque from the gear to the shaft. I'm thinking I can put a key on the gear and a keyway in the shaft. Does anyone have any literature on a situation like this? I'm looking for information on recommended materials, design, tolerances, etc. I want backlash to be minimal, but, at the same time, I don't want to cause too much drag on the linear portion of the motion.

Thanks in advance for any help.

Dan

 

[GregLocock]

For any reasonably sized, simple, design it won't work. the mechanical advantage of the PCD over the keyway will cause the keyway to lock.

Why not use a long (axially) gear on one shaft and a short one on the other shaft?

If you insist on using a gear (say if you are using a chain drive) then investigate ballsplines.



Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[ddelaiarro]

I can't use a long (axially) gear on the shaft that moves linearly because the shaft must 'disappear' into a bellows vacuum seal that is only slightly larger than the shaft diameter while the gear stays where it is (attached to the ferro-seal). I don't have much room below the ferro-seal to work with either.

Perhaps a very small linear rail (like those made by THK) mounted on the shaft would work. But, I'd have to 'sink' this design into the shaft and I also need to run four gun-drilled bores (for cooling gas and water) up the shaft, so I'm not sure if there will be room.

 

[GregLocock]

How much time and effort are you willing to put into this?

If you analyse the forces throughout the system you may find a feasibel solution.

Also, how is the gear located axially?

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[electricpete]

Just to clarify - You want the shaft to move axially while the gear stays in the same position axially?

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[MikeHalloran]

The ballspline that Greg suggested functionally comprises an array of ball slides, built into the shaft instead of inserted.





Mike Halloran
Pembroke Pines, FL, USA

 

[MintJulep]

And presumably this is all in the vacuum side of the seal, right?

What is the level of vacuum?

How much torque?

Does the linear motion have to occur simultaneously with rotation?

 

[ddelaiarro]

Thanks for all the responses. The gear is located on the atmospheric side of the the ferro-seal. This seal has an inner hub that is free to rotate and that is what the gear is bolted to. The shaft moves linearly and rotates so that a subassembly in vacuum can both translate and rotate, although not at the same time. Rotation needs to be transmitted both in the up and down positions. Torque ranges from 1E-3 Torr to 1E-7 Torr. I'm looking at putting about 90 ft-lbs on the shaft.

 

[MikeHalloran]

So, don't try to work within the gear.

Ex: Bolt the gear to the seal hub, but not with regular bolts; use shafting with threaded ends, or shoulder bolts, so you get an array of cylindrical features extending out into the air from the gear face. Fit a large flange to the end of the shaft and equip it with bearings to slide on the shoulder screws. You will probably need to matchmark the gear and the flange; it's an awkward assembly, and if it fits together at all, it will fit in only one orientation.

It's not the sort of thing I would put in production, but a good machinist can make _one_ that works.



Mike Halloran
Pembroke Pines, FL, USA

 

[ddelaiarro]

Mike -

Thanks for the response. Unfortunately, that idea won't work for two reasons: A) this needs to be a production design and B)I can't make the overall diameter of the shaft bigger than 2" due to some other design constraints.

All of these ideas are great though. It gives me different paths to go down. It's amazing what another set of 'eyes' can do within design.

 

[Cockroach]

Taper the shaft into the hub of the gear. This will provide a high precision alignment of gears while permitting axial movement in the transverse direction. You may need to knurl the shaft just before full contact in order to increase surface friction between mating pieces.

Kenneth J Hueston, PEng
Principal
Sturni-Hueston Engineering Inc
Edmonton, Alberta Canada

 

[ddelaiarro]

I overlook a simple solution. Why not use a square shaft with a square linear bearing press fit (possibly with set-screws to prevent rotation) into the seal. The gear is bolted onto the rotary part of the seal, thus transferring the torque from the gear to the shaft. Pretty simple, right?

 

[strokersix]

Diamond section driving members are often used for telescoping action because they provide less friction than square due to more favorable pressure angle.

 

[handleman]

Just out of curiosity, have you contacted a vendor of UHV equipment to see what they may have in stock that you could modify? (Or are you a vendor of UHV equipment developing new product? )

 

[ddelaiarro]

handleman, I work the latter. We are a company that specializes in ion beam etch equipment for the data storage and semiconductor industry. All our products involve UHV.

 

[EdDanzer]

This may provide some help.

http://www.dodge-pt.com/pdf/catalog/pt_components/2004_pdf/eng_tech.pdf

The side of the surface of the key that is sliding should be large enough that the pressure does not exceed 6,000 psi or it will not slide. The force required to slide it is affected by the normal force and the coefficient of friction.

 

[handleman]

I know linear-plus-rotary feedthroughs are relatively common, although in the catalogs I have they have much lower torque limits than what you're looking for (100 in-lb was the largest I saw). Do those mechanisms just not scale up well?