Prop shaft

Help Support Homebuilt Aircraft & Kit Plane Forum:

wsimpso1

Super Moderator
Staff member
Log Member
I was thinking that if the concentric shafts were both rather long, the surface area the fluid is in contact with would provide sufficient "grip" to keep slippage down to a reasonable level and give enough external surface area via the outer shaft to cool it enough.
I should clarify that I mean having only a millimeter or so of clearance between the inner and outer shafts.
I recommend you visit the equations for viscous forces between flowing fluids in contact with stationary surfaces. Any good introductory fluids textbook should do. Then you can postulate radii, clearances, how much area is needed to generate how much torque capacity, and how much energy is lost in a viscous drive. Then your can also go to a heat transfer text, and figure out how much area and delta T it would take to passively cool these things.

I will tell you it is not really practical for our levels of power and weight sensitivity. If you do not believe me, knock yourself out with the fluids and heat transfer books. I will help you get the math straight if you want, but it won't change the fact that a 200 hp engine will lose a bunch of energy and that you won't be able to make a practical device for a flying airplane this way.

Billski

Sockmonkey

Well-Known Member
If you do not believe me, knock yourself out with the fluids and heat transfer books.
Laziness has made me decide that you're trustworthy.
seriously though, It was just an idea I had that I wanted to float by you guys.

thjakits

Well-Known Member
May I suggest that you study helicopter tail rotor drive shafts? ... especially on lighter, piston-powered helicopters like Bell 47, Enstrom, Hughes 300 and Robinson?
....Robinson uses "Flex-Couplings", basically a springy steel-plate with 4 prongs - connected crosswise with 2-prong couplers, can take a lot of flex (...tailboom going from side to side, as happens when the tail-rotor thrust changes).
Robinson enganges the piston powered models with a "slip clutch" - 2 or 4 multi-rip (lengthwise) V-belts that get slowly tensioned around 2 pulleys/sheaves. The tensioning happens with a electric jack-screw. Proper tension is obtained with some electric end-switches and some spring-plate trickery! Usually as things get warm it will re-tension a few times after start-up...
There is also a sprag-type freewheel unit, that allows the rotor to autorotate without having to pull the engine along....
That sprag-clutch is quite small and can take quite a lot of power. HOWEVER it is not meant to operate continously - usually it would be a the longest possible autorotation and some safety margin - at a time - then there is a period for it to cool off....
The belt drive likely eliminates most (if not all) resonance issues and absorbes engine power pulses....

I could imagine a 2-belt drive with a spring loaded idler pulley providing the same service on an Mini-Imp or Teal.... (the R22 helicopter uses 2 belts, supposedly one is enough to deal with it's 124hp continuos rating, and up to 160hp emergency use...)
Supposedly 2 of the 4 belts on a Robinson R44 can take care of all of the engine power, which would be 260hp max emergency use (though max power is down-rated in the helicopters to 205hp continous, although just on paper - you have the power available if needed to save the day...)

Still, one would want to do a little testing and research into power pulse elimination and certainly into resonance frequencies (you need to be able to start up and then run between idle and operating rpms).

As it is - I would avoid cross-rip belts. Though they can dig in a lot stronger, they usually will start to deteriorate faster once things go wrong. The Robinson style V-belts hold up very well - get the proper V-sheaves and make sure alignement is correct and HOLDING!
(Robinson had issues with the original Gates belts to reach their TBO of 2200 hrs, but since switched to Mitsuboshi belts, they work flawless... - yes, MitsubOshi....)
You could make the idler hand lever engageable.... - or copy the Robinson system, but with one springloaded sheave (...manually enganged with a lever) - I don't think you would need a prolonged engage-time for a prop and the shaft (vs. a fairly heavy rotor-system). Just move the lever until you feel the belts starting to engage the shaft/prop and apply a little pressure and the everything should synchronize in less than 3 seconds - then engage fully, probably with some over-center action which then leaves the proper tension to a pre-set spring.

With a little designing you should be able to get it so, that the flexplates are all you need to compensate for airframe movements, so a telescoping shaft (like on most older car-rear-drives) is not needed....

Cheers all,

thjakits

Last edited:

thjakits

Well-Known Member
Except that the propeller is a flywheel at the other end of that shaft. If the engine's power pulses aren't taken out before the power is sent down the shaft, something's gonna flex and get hot and fail. In a car, the clutch plate has springs that do that, or in an automatic the torque converter does it. Billski can elaborate on this.

In a helicopter, the tail rotor is not expected to be a flywheel, it weighs almost nothing, and that shaft is slender and winds up some during flight. It's designed that way. The piston helicopter's engine has a flywheel and a clutch of some sort.
....well, a tail-rotor does not weigh NOTHING - they are quite hefty and at the rate they turn centrifugal forces are quite substantial , definitely in "propeller range", on top of that, if the helicopter is in utility work mode (lots of full power stationary situations) the gyroscopic precesion loads are ..."impressive"!

Whatever wind-up you get on the TR-shaft is not a problem - even with fast power changes and heavy pedal work - actual force changes are rather slow (as slow/fast as a human can act on the controls...).

But of course you are right - one needs to be aware and careful with resonance frequencies!
Again - a proper belt reduction drive would probably take care of most of the problems....

thjakits

Dan Thomas

Well-Known Member
....well, a tail-rotor does not weigh NOTHING - they are quite hefty and at the rate they turn centrifugal forces are quite substantial , definitely in "propeller range", on top of that, if the helicopter is in utility work mode (lots of full power stationary situations) the gyroscopic precesion loads are ..."impressive"!

Whatever wind-up you get on the TR-shaft is not a problem - even with fast power changes and heavy pedal work - actual force changes are rather slow (as slow/fast as a human can act on the controls...).

But of course you are right - one needs to be aware and careful with resonance frequencies!
Again - a proper belt reduction drive would probably take care of most of the problems....

thjakits
I used to have a pair of tail rotor blades. They were hollow, made of aluminum sheet, and weighed maybe a pound or pound and a half apiece. Bell 206, I think.

Pops

Well-Known Member
HBA Supporter
Log Member
I recommend you visit the equations for viscous forces between flowing fluids in contact with stationary surfaces. Any good introductory fluids textbook should do. Then you can postulate radii, clearances, how much area is needed to generate how much torque capacity, and how much energy is lost in a viscous drive. Then your can also go to a heat transfer text, and figure out how much area and delta T it would take to passively cool these things.

I will tell you it is not really practical for our levels of power and weight sensitivity. If you do not believe me, knock yourself out with the fluids and heat transfer books. I will help you get the math straight if you want, but it won't change the fact that a 200 hp engine will lose a bunch of energy and that you won't be able to make a practical device for a flying airplane this way.

Billski

Dan

BBerson

Light Plane Philosopher
HBA Supporter
The helicopter v- belt reduction is about 9 to one ratio from flywheel to rotor mass.
A direct drive prop shaft is one to one ratio, so more likely to resonate, I think.

Turd Ferguson

Well-Known Member
HBA Supporter
The one on my lathe is incredible. Never fails.

Pops

Well-Known Member
HBA Supporter
Log Member
One of my grandson's has his name on a patent that has something to do with magnetic coupling. Over my pay scale.

thjakits

Well-Known Member
23 years in automatic transmission design engineering, 18 of it in torque converters. I know these machines.

... They are still a two mass system with inertia at each end and a spring in between, and you will get as many pulses per prop rev as you have blades, so you still have to make sure your operating range is safely separated from resonance modes...

... Seriously, it is way better to have a selected amount of springiness in a designed system so resonance occurs away from the engine operating vibrations and other inputs of the system... Yeah, it means doing some serious engineering, but there are no coolers, no oil loops, no pumps, no seals, no turbomachinery, and lot less weight....

Billski
....and we are right back with what Robinson did with their TR-driveshafts!

Robinson is also know to find the lightest solution to any problem - .....

thjakits

thjakits

Well-Known Member
I used to have a pair of tail rotor blades. They were hollow, made of aluminum sheet, and weighed maybe a pound or pound and a half apiece. Bell 206, I think.
...right, probably a little more than a pound.
Then there is the hub, spider, pitch-links .... TR ARE light!! But with the above mentioned forces at work, attaching gear has to be designed rather tough - at the end - a R44 TR has a tip speed of 614 fps while doing 2428 rpm....
I'd guess the TR weighs in about the same as the 206 blades (Robinson makes a SUBSTANTIAL TR...)
....for those of you who want to throw numbers right away!!

thjakits

Last edited:

TFF

Well-Known Member
I would never have called a Robinson tail rotor substantial. Call it adequate. A helicopter blade system is at constant speed or as close as it can. It’s not changing speed with throttle More or less torque is applied with load and load does change but most is only seen as blade pitch change not rpm change. Flex plates are only strong enough. I’m use to the multiple layer type not the one piece Robby. It doesn’t take too much to make them shred.

Bigshu

Well-Known Member
It would need a flywheel, much like the original big flywheel on the Wright Flyer with the prop shafts. Prop shaft failures were a major problem for the Wrights from the start. Some failed in seconds while others were 2 hours.
Never saw the engine setup on the Big dipper, or the Cloudster 2 for that matter. There's no flywheel in the Mini-IMP, unless the flexidyne takes the place of one. Don't think there was one on Teal, but I hope to get a close look at that this year.

Bigshu

Well-Known Member
....Robinson uses "Flex-Couplings", basically a springy steel-plate with 4 prongs - connected crosswise with 2-prong couplers, can take a lot of flex (...tailboom going from side to side, as happens when the tail-rotor thrust changes).
Robinson enganges the piston powered models with a "slip clutch" - 2 or 4 multi-rip (lengthwise) V-belts that get slowly tensioned around 2 pulleys/sheaves. The tensioning happens with a electric jack-screw. Proper tension is obtained with some electric end-switches and some spring-plate trickery! Usually as things get warm it will re-tension a few times after start-up...
There is also a sprag-type freewheel unit, that allows the rotor to autorotate without having to pull the engine along....
That sprag-clutch is quite small and can take quite a lot of power. HOWEVER it is not meant to operate continously - usually it would be a the longest possible autorotation and some safety margin - at a time - then there is a period for it to cool off....
The belt drive likely eliminates most (if not all) resonance issues and absorbes engine power pulses....

I could imagine a 2-belt drive with a spring loaded idler pulley providing the same service on an Mini-Imp or Teal.... (the R22 helicopter uses 2 belts, supposedly one is enough to deal with it's 124hp continuos rating, and up to 160hp emergency use...)
Supposedly 2 of the 4 belts on a Robinson R44 can take care of all of the engine power, which would be 260hp max emergency use (though max power is down-rated in the helicopters to 205hp continous, although just on paper - you have the power available if needed to save the day...)

Still, one would want to do a little testing and research into power pulse elimination and certainly into resonance frequencies (you need to be able to start up and then run between idle and operating rpms).

As it is - I would avoid cross-rip belts. Though they can dig in a lot stronger, they usually will start to deteriorate faster once things go wrong. The Robinson style V-belts hold up very well - get the proper V-sheaves and make sure alignement is correct and HOLDING!
(Robinson had issues with the original Gates belts to reach their TBO of 2200 hrs, but since switched to Mitsuboshi belts, they work flawless... - yes, MitsubOshi....)
You could make the idler hand lever engageable.... - or copy the Robinson system, but with one springloaded sheave (...manually enganged with a lever) - I don't think you would need a prolonged engage-time for a prop and the shaft (vs. a fairly heavy rotor-system). Just move the lever until you feel the belts starting to engage the shaft/prop and apply a little pressure and the everything should synchronize in less than 3 seconds - then engage fully, probably with some over-center action which then leaves the proper tension to a pre-set spring.

With a little designing you should be able to get it so, that the flexplates are all you need to compensate for airframe movements, so a telescoping shaft (like on most older car-rear-drives) is not needed....

Cheers all,

thjakits
I think this approach has some merit. I'm concerned the availability of the flexidyne units might be a serious problem.

BBerson

Light Plane Philosopher
HBA Supporter
The flexidyne must have some flywheel effect but not as much as a prop according to this:

Active Member
Look at the c 5-6 vette drive system.
Look at 63 Pontiac lemans.
I had a 63' Lemans on entitle with that "torque tube". It had 4 wheel independent suspension as well. I wish they'd continued refining that chasis for another 10 years it would have been a keeper. Different times - I think the 64' was changed considerably and rebadged as "GTO". Also fun but..

Back to torque tubes - I thought the carbon fiber tubes with proper layups were considerably stiffer and so could be lighter, or longer, for a given power transfer..

thjakits

Well-Known Member
I would never have called a Robinson tail rotor substantial. Call it adequate. A helicopter blade system is at constant speed or as close as it can. It’s not changing speed with throttle More or less torque is applied with load and load does change but most is only seen as blade pitch change not rpm change. Flex plates are only strong enough. I’m use to the multiple layer type not the one piece Robby. It doesn’t take too much to make them shred.
Well, you have to compare according to the actual machine - the R44 IS smaller than a 206!
However - it is well known and proven that the R44 (...and now possibly the R66) have about the strongest TR in the industry! A Robinson TR - essentially - doesn't stall (LTE) - and it still works at 70% (most other helicopter's TR are long gone "ineffective" by then...
The 22 is less efficient, but just as effective as the 44 TR....
[Don't get me wrong - though I have way too many hours in the 44, I wouldn't want to go back on a 44 from a 350B3....]

thjakits

TFF

Well-Known Member
I don’t have much experience with 206s, but I pretty much touch an Enstrom every day of the week. I know I can do left pedal turns 50 rotor RPMs low which is similar to the factory testing pedal rigging.

Derswede

Well-Known Member
Personally I think the latest model
Turbo Encabulator
would be the best application, no friction, and less than a billion \$!

(OK, a day late....sorry!)

Derswede