Paul,
Thanks for the gear box/speed reducer pics - Interesting - A few
questions -
1 - Good lifetime on the needle bearings ??
2 - Looked like uncaged bearings ??
3 - Presume the 10 lb heavier unit used some other type of bearings
??
4 - What provisions for lubrication ??
5 - The drawings have all the appearance of a multi-thousand dollar
product - Price range ??
6 - Cooling problems - I would think that there might be some with
uncaged needles ??
7 - Doesn't look like a particularly difficult project from a
machining perspective -
8 - Presuming outer housings are aluminum castings - Patterns exist
??
Roland Friestad
You must keep the RPM on the needle bearings down. Particularly on
the thrust needle bearings. Those tiny needles are really zinging
around in there. We have a lot of experience on this. Some people
have worn them out. That is the reason for three back to back. The
RPM on any one is one third of the total.
The shaft, surprisingly, does not get any load on the bearings until
the aircraft stalls or spins. Characteristic of planetary gear sets
as all side load is canceled out. Hard to imagine but true. Those
give no problem and they are much lighter, as is the housing, if they
were ball or tapered rollers. Win win situation not to mention much
lower cost all around.
Note the extended point set screws are better and easier to machine
than splines. All back lash is eliminated.
I have an account set up with a local wholesaler of tranny parts. I
make no money on them I just charge cost plus shipping. The 3.17 gear
set is just under $200.
The box can be built for under 5 or 6 hundred in parts. It all
depends on what the NC shop charges.
Shares synthetic oil with the engine. BTW the oil stays clean in a
rotary. Uses the engine oil cooler.
Here is some more stuff on the subject from 16 years exp. :)
There will be more on the rubber coupling with the engine.
Paul Lamar
Paul and Roland,
Glad to see motion on this project. Plenty of people would be happy
to pay $2-4K for a competently done redrive in the 100-200 hp range.
The extra cost is for support -- good engineering and documentation
so that people can figure out where their problems are and how to fix
them. The only thing worse than being on a schedule to deliver things
is being on the hook for phone support when stuff breaks. The
planetary system and bearings are entirely known technology; where
there isn't a proven system, and a lot of trouble remains, is in
handling of torsional vibration.
Rotary engine power pulses will be at a higher frequency, which makes
some problems easier and takes some solutions off the table. It would
be worthwhile to make aany new design applicable to a range of
powerplants, not just rotary.
At some engine speed, the frequency of its power pulses will resonate
with the spring constant of its drive shaft, gearbox, propeller shaft
and the inertia of the rotating prop. At this frequency, each engine
pulse meets the reflection of the previous pulse delayed through the
compliance of the system, and causes a huge increase in torsion
throughout (at other frequencies the reflected pulse doesn't exactly
meet the outgoing pulse and therefore doesn't reinforce.) The engine
is pushing one way while the inertia is pushing back the other way,
doubling the torque in the shaft at the first pulse. The peak torsion
can exceed 20 times each individual torque peak because each
successive pulse adds more energy to the spring formed by the shaft.
The key to a successful redrive design will be to know exactly what's
going on and build something that will never get to a destructively
resonant condition. There are many methods of achieving that, but you
won't know you're there until you have modeled the system
mathematically and proved that your model is accurate by testing.
We need to: (1) design the compliance of the system so that the
resonance, with all values of propeller inertia, lies outside the
ground and flight operating range of the engine and (2) if the
resonant point is reachable in normal operations, include something
to break the resonance (lower the Q of the system) when excursions
exceed a safe value. In normal flight operation the resonance breaker
should not deflect at all. Ross used the springs of a clutch plate
with limited success. Paul has suggested a coupling built with pieces
of silicone rubber, it looks good but as far as I know has not been
tested. Tracy had his own design and wrote clearly about the design
process but without sharing any of the math. There are commercially
available rubber giubo disks used in many European cars. Tracy called
this the "BMW seduction" but it's also available from Centaflex and
commonly used to solve the same problem in industrial and marine
drives. Vassily at Airtrikes successfully uses Mercedes giubos with
his spur gear redrives. Rotax and Geared Drives (Bud Warren design)
use clutches -- Rotax to decouple when torque gets too high, Geared
to not engage until engine speed exceeds the resonant point. In every
case you need to establish how long something will last in worst
case conditions (operating at the critical resonance point) before
something breaks.
I don't use the terms "damper" or "harmonic" because there is no
damping going on, and this is mostly at the fundamental frequency,
not a harmonic. Damping of a resonance requires resistance, in which
the excess energy is dissipated as heat, and that's not happening
here. Any of these devices would burn up in an instant if they were
actually absorbing any significant part of the engine's power. And
"dampen" means to make wet, not attenuate, so it's not a "dampener."
An elastomeric member or spring breaks a resonance by compressing,
shifting the resonant frequency momentarily so the resonant point is
no longer where it was.
My background is in EE and acoustics. I design and make acoustic
transducers for a living, so resonance, damping and compliance are a
little bit familiar. I have studied redrive design for some time (I
am still looking for an engine/redrive that will work for a low speed
propeller design project) and have met a few of the problems, but
know only that I am certainly not an expert in any of this -- but I
can now tell when "expert" opinions aren't supported by physics.
David Josephson
Let me know what you come up with. A lot of people have been fooling
with this for a long time. Gear and spline back lash probably enters
in to it and that changes over time. So too rotor bearing clearance.
Two nine pound rotors have a significant effect on the polar moment
of inertia. Prop weight also varies. Beware of the rotary's high RPM
when looking at open rubber couplings. I think that had something to
do with the RotoMax crash.
Then there is the question of shock load on the gears. Seems like a
bit of rubber is called for there. Some Brit engineer said "An ounce
of rubber is worth a pounds worth of engineering :) Silicone rubber
is tough and high temperature. Ideal for this use.
It seems no matter what we do the resonance happens at 1800 engine
RPM.
No problem idling the rotary engine above 1800 RPM as the prop is
only going 600 RPM roughly. I was rather surprised to find the same
resonance RPM on the RotoMax engine with an entirely different gear
box, engine design and coupling.
When I shut off the O-470 in my Cessna 182 the whole airplane
shutters as it passes through resonance RPM.
Paul Lamar
I'd make provisions for a hydraulically operated prop and governor
pad if I was to do a new PSRU product.....make it an option-----even
Geared Drives is going to make a 0 offset geared unit now, with their
clutch and their hydraulic prop and gov........... Marc W.
I knew you were going to say that Marc :)
In order to use a hyd prop the whole design must be changed to
pressure fed plain bearings. The vast majority don't need a 50 pound
hyd prop. The power to weight of the rotary airplane is so good they
leap off short runways :) Cost is also one of the negative issues.
Paul Lamar
There are also a few electrically actuated CS props available. MT
makes certified versions with composite blades.
http://www.mt-propeller.com/en/entw/pro_elec.htm
There is no need to re-invent the wheel to add a hyd gov unit. The MT
composite props are so light that the biggest complaint is the aft
moving CG. I fly a Decathlon with an MT composite prop, and the
biggest downside I can see is the aft CG, and the low inertia.
Removing the 60+ lb Hartzell and replacing it with the 40 lb MT made
a noticeable difference especially when flying solo from the front
seat. A friend with an RV4 with an IO 360 recently replaced his
aluminum prop with a fixed MT. Overall he is happy, but his CG moved
aft enough that a 200 lb passenger makes his plane out of CG. Also,
when you pull the prop back it stops almost instantly and will even
stop windmilling at best glide speed, requiring a dive to get it
spinning again. The plus side is instant acceleration. The low
inertia allows the engine to go from idle to WOT almost as fast as
you can push in the throttle.
Another plus is that if you damage a blade, they will send you a
single blade so you dont have to replace the whole prop. In the event
of a prop strike, the blade pretty much shear off. You still have to
tear down the engine, but most times engine damage is minimal.
Overall, I like these props.
Kevin Alderman
Hi, I am new to the group, and thanks for my being allowed to
participate. Is the gearbox design that is being talked about in this
e-mail feed this one, by Paul Lamar,
http://www.rotaryeng.net/High-tech-complete-PSRU.jpg.
Thanks, Steve Carlisle
In theory that is lighter than the conventional and traditional prop
flange and shaft. The drawback is it takes a high HP NC lathe and the
prop shaft is made from very expensive and very costly 4340 solid
billet bar. LOTS of waste chips :)
Paul Lamar
Hi Paul, How is the drive taken through the planet set in the "high
tech gearbox", I don't see any drive flange from it and the ring
appears to be floating. Chris.
The ring gear is locked to the housing for a 3.17 GB. The back of the
output shaft is drilled and taped for a whole bunch of machine screws
holding the planet carrier. Similar to this power sport design.
Paul Lamar
If needles or balls are not caged, they will bunch together and seize
the shaft. I was working at Rocketdyne when they were making the Beta
Gimbal Actuators that are used to position the Solar Array Wings toward
the sun on the International Space Station. On the design, one of the
bearing assemblies had balls that were individually caged; there was not
one whole cage that kept all the balls equally spaced, but rather a
separate cage for each ball. It didn't work. They spent untold hours
trying to figure out why the actuator was locking up. The reason was
that some of the balls were bunching together. You must have a
monolithic cage that keeps all the balls, or needles, equally spaced.
Gary Schwarz
Yes of course. I just got lazy and did not draw the cage in.
Paul Lamar
The Rotary Engine News Letter. Powered by Linux. ACRE NL web site.
http://www.rotaryeng.net You Tube
http://tinyurl.com/beqqxas
Copyright 1998-2014 All world wide rights reserved.