Prop shaft

Homebuilt Aircraft & Kit Plane Forum

Help Support Homebuilt Aircraft & Kit Plane Forum:

wsimpso1

Super Moderator
Staff member
Log Member
Joined
Oct 18, 2003
Messages
7,960
Location
Saline Michigan
They work pretty well on picking up broad band noises and random impacts, but resonance? If you have resonance, it doesn't help. We still need a system that has it resonances out of range from the forcing functions.
 

wsimpso1

Super Moderator
Staff member
Log Member
Joined
Oct 18, 2003
Messages
7,960
Location
Saline Michigan
What if you had a torsionally rigid shaft? Say 12" diameter or whatever is needed.
The stiffness would have to be high enough to drive the natural frequency more than 1-1/4 octaves above the highest firing frequencies. That is what we normally do with crankshaft design.

Troubles with it are:
That stiff a shaft is heavy if it has length;
Tough to put in Hooke joints that are stiff enough;
Every part of that system has to be strong enough to carry the firing pulses.
 

BBerson

Light Plane Philosopher
HBA Supporter
Joined
Dec 16, 2007
Messages
14,968
Location
Port Townsend WA
That stiff a shaft is heavy if it has length;
Tough to put in Hooke joints that are stiff enough;
Could be a space frame shaft with rigid crank connection, no Hooke's joints.
Some racers have 12" spacers which is a sort of cantilever shaft extension. A five foot shaft/spacer would need a simple bearing at the prop. It would almost be part of the tail cone structure.
 

trimtab

Well-Known Member
Joined
Apr 30, 2014
Messages
189
Location
rocky mountains, rocky, usa
They work pretty well on picking up broad band noises and random impacts, but resonance? If you have resonance, it doesn't help. We still need a system that has it resonances out of range from the forcing functions.
The patents and literature specifically cite resonance damping. I don't know how they dissipate the heat. But the poiseuille shearing would have a similar response to drop the low pass crossover frequency as a spring or torsional modulus change.
 

wsimpso1

Super Moderator
Staff member
Log Member
Joined
Oct 18, 2003
Messages
7,960
Location
Saline Michigan
Could be a space frame shaft with rigid crank connection, no Hooke's joints.
Some racers have 12" spacers which is a sort of cantilever shaft extension. A five foot shaft/spacer would need a simple bearing at the prop. It would almost be part of the tail cone structure.
OK, let's get some detail...

If you want high torsional stiffness at min weight, you use a tube, and increase OD until it meets needs in torsional stiffness and in torsional strength. Materials like alloy steel and carbon fiber (wound at +/- 45 degees) are good choices.

The difficulty with stiff systems is that you have to drive the lowest natural frequency safely above the highest firing frequency of the engine. And with a long shaft like the OP is discussing, that really requires a LOT of torsional stiffness. Weight.

Want to make that long system also rigid to the crankshaft? Ugh. First off, the engine vibrates in a bunch of modes: Yaw rotation of horizontally opposed engines (and it takes the crank with it) at firing and twice firing frequencies. Pitch rotation of inline and V engines too. Then there is P-factor and gyroscopic moments and yaw-pitch rotation from stuff like climb and maneuvering. While the engine is trying to do all this, the prop out at the end of the shaft is providing inertia to resist these vibrations at the end of a long really stout lever arm, so you end up with really huge bending moments applied to the crank flange. Is that a problem?

First off, the crank end and prop bearing of our traditional airplane engines are designed to resist these forces with the prop bolted to the flange. Yes, some folks with fast cross country airplanes have extensions up to a foot, but they do little in the way of aerobatics. And the aerobatic folks run without prop extensions... Hmm.

Then mounting the prop end of the shaft to the airplane... Having worked a system like this, I say that you have to carry reactions in vertical and lateral axes, but you either have to let the shaft slide axially (and let the engine mounts carry thrust) or you have to carry prop thrust and figure out how to let the engine float longitudinally. Why? Well first there is prop thrust and P-factor both pushing on the thing in rigid body manner, and the fact that the shaft modestly changes length under loads. Then the fuselage and the shaft are unlikely to have the same thermal expansion behaviour. Steel moves less than aluminum moves less than carbon composites. Now maybe you can match the shaft and aft fuselage, but parts bolted to the crankshaft and in the hot internal air off the engine will usually be warmer than the fuselage shell. No, I suspect you will have to find a way to let the shaft slide fore-aft through the prop bearing out at the tail. And this bearing will cause us to have big pitch and yaw moments at the crank due to nominal vibration of the engine in those axes...

One other point is that long shafts have another resonant behavior called critical rpm. Any shaft will have a vibe mode where the middle of the shaft vibes out of axis. When its rotation speed lines up with this natural frequency, the shaft whirls around like a jump rope. Does this again at twice that rotation speed too. This can drive more shaft stiffness (more diameter is common) or it drives support bearings along its length. big delivery trucks have three and four part driveshafts for exactly this reason. That 20' drive shaft as a one piece shaft would be prohibitively large in diameter. So they break it up into shorter pieces where each one has a critical speed above the operating range.

I shall emphasize again, that in stiff systems, with first resonance well above the forcing functions from engine operation, the entire thing then rotates and accelerates while vibrating together. Crank, shaft, prop. While this is just like with no extension torsionally, even that has tails. The Army-Navy engine specs in the mid- to late-1930's drove the development of various torsional pendulums - to keep props from flinging their tips. On our engines, not so much.

The usual solutions arrived at, first in steam turbines, then in fixed wing airplanes, helos, gas turbines, you name it, is usually to either drive lowest forcing function frequencies high or to use soft systems, and commonly both. Soft systems. This isolates firing vibe at the engine and you do not need to carry big torques to other parts of the machine. Look at the long slender shafts on helo tail rotors and the prop drives on the post-WWII bombers.

If you want low torsional stiffness, you use a solid shaft and increase its diameter until it meets your torsional strength needs, then see if it is torsionally soft enough. It gets softer the longer you make it and the lower the G of the material.

When you make the shaft torsionally soft, if can be the soft element that drives the first natural frequency of the system below min excitation frequencies. And even if it does not get the natural frequency low enough, it gets it part way there, so you can get the rest of the way there with other soft elements in the system.

No, you do not just have one or two things to make "safe", there are a bunch, and soft systems are common for a bunch of reasons...

Billski
 

trimtab

Well-Known Member
Joined
Apr 30, 2014
Messages
189
Location
rocky mountains, rocky, usa
An important and common means of reducing resonance gain and increasing resonance frequencies is to employ a redrive to match inertia. This will be a problem for the application here but it's an important concept to know about. I've used the concept quite a bit to kill large moving load resonances in large mass control systems without resorting to feed forward schemes.

 

wsimpso1

Super Moderator
Staff member
Log Member
Joined
Oct 18, 2003
Messages
7,960
Location
Saline Michigan
The patents and literature specifically cite resonance damping. I don't know how they dissipate the heat. But the poiseuille shearing would have a similar response to drop the low pass crossover frequency as a spring or torsional modulus change.
Marketing BS. If they solved real problems, the automakers would use them. I know, I have been there. We turn over heaven and earth to fix NVH and durability issues. The money and effort spent on the topic is amazing.

First, the lowest crankshaft resonant mode in torsion must be at least 2 octaves above maximum firing frequency... This is a common criteria in IC engine design, and is a primary driver of crankshaft journal diameters and cylinder layout. If the crank will wind up and unload at a frequency within range of the firing frequency and of the pistons accelerating back and forth in the bores (twice firing frequency), the crank is in resonance and will not last at all. This is what racing teams find out when they run the rpm range higher on an engine by things like tuning valve timing and breathing. They get into crankshaft resonance (or its close cousin, crankcase resonance) and break stuff. So that sets highest rpm they will tolerate with a case and crank design. And these viscous absorbers hung on the crank DO NOTHING for that mode.

Second, please read the first couple posts and skim the rest in:

When the frequency that the engine is supplying pulses in coincides with rate that the system wants to oscillate. This is resonance and it will grow in amplitude without an upper bound. Well, until something else intervenes. Like hitting stops or breaking stuff.

These guys are proposing to place a steel ring in a tight enclosure with silicone goo on the other end of the crankshaft. This takes off energy on each swing, converting it to heat. The energy absorbed on each swing would have to exceed the energy being added in each cycle. Firing pulses are big. When the manifold pressure is approximately atmospheric pressure, firing pulses peak at around eight times mean torque of the engine and that is some big energy. These dampers would have to be big indeed to pluck off enough of that to keep it from amplifying at resonance.

So, even if it could pluck off this major resonance, it would have to do it by absorbing engine power. Do you want it taking engine power and dumping it as heat? I know where I want my engine power going - turning the prop! And if it is sucking off any measurable fraction of engine power, you need to cool it. They are not cooled. A few square inches of steel hung on a hot crankshaft. As they get hot, the silicone gel inside has its viscosity drop and it becomes less effective. Terrific, in steady state operation, it gets hot and decreases effectiveness. The commercial gadgets are small, not even finned, and they are proposed for hanging on engines that have proven durable of their applications already, but now can pluck off some annoying vibrations. The drag racers like them for toning down vibration at the front end of the crank driving superchargers. Their race is over in a few seconds, so they do not worry much over cooling. Many of them do not even have cooling systems for the engine...

But to take off crank vibration as they imply in their marketing materials? Hogwash... We ran them on dynos at both Ford and Chrysler. The ones out there change the vibe spectrum in front end accessories, and not necessarily for the better. The commercial ones on the front pulley do not help with interactions between engine and tranny and drivelines. What does work great on FEAD is a frequency tuned harmonic damper (been around for many decades) built into the front crank pulley, tuning spring rates from belt and tensioners, and sometimes incorporating a OWC in the hub of the alternator.

In the auto industry, we applied effective isolator design between engine and tranny and then frequency tuned absorbers on driveshafts where they help. We did not apply things that cost a lot and do not work. Like these viscous absorbers on the crankshaft pulley.

Billski
 
Last edited:

BBerson

Light Plane Philosopher
HBA Supporter
Joined
Dec 16, 2007
Messages
14,968
Location
Port Townsend WA
. And with a long shaft like the OP is discussing, that really requires a LOT of torsional stiffness. Weight.
He said five feet but might settle for 3 feet. The engine needs flywheel weight in any case.
Extend the rigid flywheel to 3 feet.
 

mm4440

Well-Known Member
Joined
Jan 14, 2012
Messages
201
Location
LA area, CA
Hi, plan on educating your self and doing a lot of ground testing first. Home built aircraft driveshafts can work but they have killed people.
 

Primaris22

Member
Joined
May 8, 2020
Messages
5
Thanks for all of the info. I didn’t reply sooner because the discussions quickly went over my head. I could in no way do the suggested math required and could only resort to trial and error which some posters suggested against. Also, although I could handle a little extra weight, the discussions led me to believe that a 5 foot shaft system would add enough weight/complexity to put the idea outside of the design mission. There is a design ( French I believe) that uses this setup with a BMW engine. It is a really nice looking aircraft and I’ll try to find pics of it and post. My design objectives required extremely light weight and unexcelled visibility. I could go with a swept forward wing but I really don’t like that idea. Thanks again.
 

Pilot-34

Well-Known Member
Joined
Apr 7, 2020
Messages
879
Sorry took a while to find it,
Twin drive shafts


“The drive is carried out by means of shafts and deflection gearboxes, which transmit the power from the engine in the fuselage to the two-blade propellers.”
 

jedi

Well-Known Member
Joined
Aug 8, 2009
Messages
2,456
Location
Sahuarita Arizona, Renton Washington, USA
Helicopter driveshafts .. here's a study on a Huey's tail rotor.


Someone mentioned about a 20% power consumption, for a tail rotor system, of the total .. this data seems to agree with that, .. up to 170 HP at times, for a Huey. So, if you can satisfy the flexing issues, (tubes are springy things, in the radial direction) .. these driveshafts, eh? And find the nearest Huey junkyard?

Attached, is one page of that PDF.

I was around this aircraft, some. Quite a bit of maintenance, the tail rotor system .. and the alignment of everything, had to absolutely be right on.
Not able to quickly find the tail rotor shaft RPM. This would be critical to know if the plan was to use a Huey tail rotor shaft for aircraft propulsion. The rotor rpm in a helicopter does not vary all over the map as it does in a propeller aircraft. That would be another issue.

Does anyone know the tail rotor shaft rpm for typical helicopters?
 

BBerson

Light Plane Philosopher
HBA Supporter
Joined
Dec 16, 2007
Messages
14,968
Location
Port Townsend WA
Long V belts can be a shaft alternative. The engine would be vertical shaft like a Briggs 810 or two- stroke mounted vertical. The belt is turned 90° around rollers to the horizontal prop shaft pulley sheave.
Two Briggs 810 could power one prop for about 60hp.
 

Vigilant1

Well-Known Member
Lifetime Supporter
Joined
Jan 24, 2011
Messages
6,237
Location
US
Long V belts can be a shaft alternative. The engine would be vertical shaft like a Briggs 810 or two- stroke mounted vertical. The belt is turned 90° around rollers to the horizontal prop shaft pulley sheave.
When configured that way, does the continual twisting/untwisting of the V-belt increase the belt temperature or decrease belt life?
Two Briggs 810 could power one prop for about 60hp.
One challenge is that, in single engine mode, a 30HP engine driving a prop designed for 60hp won't produce as much thrust as a prop designed for a 30 HP engine.

Two B&S 810s would be a very economical way to make 60hp, and the weight penalty isn't very high compared to a single 60hp 4-stroke. Running them in their "native" vertical shaft mode simplifies lubrication, etc.
 

BBerson

Light Plane Philosopher
HBA Supporter
Joined
Dec 16, 2007
Messages
14,968
Location
Port Townsend WA
One guy on the yahoo small engine forum flew his vertical shaft Briggs with the 90° belt drive on his powered parachute. I would think three feet of straight belt should twist ok.
I wouldn't call it a twin engine airplane because if one engine seized it would stop the other engine same as any single piston failure on a Lycoming.
 

Pale Bear

Well-Known Member
HBA Supporter
Joined
Feb 22, 2009
Messages
48
Location
Minnesota
Not able to quickly find the tail rotor shaft RPM. This would be critical to know if the plan was to use a Huey tail rotor shaft for aircraft propulsion. The rotor rpm in a helicopter does not vary all over the map as it does in a propeller aircraft. That would be another issue.

Does anyone know the tail rotor shaft rpm for typical helicopters?
Good question, .. I should have researched that, first. It looks to be 1650 RPM, .. on page 5, of what I attached, .. which is a bit slow, isn't it.
 

jedi

Well-Known Member
Joined
Aug 8, 2009
Messages
2,456
Location
Sahuarita Arizona, Renton Washington, USA
This would depend on the rotor diameter. But if I remeber well from a UH-60 Blackhawk OH, the main rotor was around 250 rpm while the tail was at around 1200 rpm. Neither would be suitable for a homebuilt aircraft.
I think the discussion was about the long shaft down the tail cone to the tail rotor not the tail rotor rpm.
I expect the 90 degree gear box to the tail rotor would likely have a step up or step down gear ratio. Maybe not.

DZ, do you know if the tail shaft and tail rotor are a 1:1 speed ratio? Thanks for the input.
 
Top