Subject: Dave Garber's bell 47 PSRU.
From: Rotary Engine
Date: 2/5/2010, 9:06 PM
To: AAA Put this in the To box


Paul,

On your website in the PSRU section you have a picture of Dave Garber's
engine mount
plate and the sun gear drive plate attached to a object that is obviously
bolted to the
counter weight.

The object in question is a plate with some sort of cup in the center. I
forgot to ask
Dave what that was.  I would guess it was for a bearing. Do you know?

He didn't use any type of torque absorber on the his turbocharged 12A. Did
you notice any
undue vibrations when you ran it at Sun and Fun?

Doug in Japan

We never let it idle below 2000 RPM.

I think you are talking about these pictures.
The big flaw in Dave's design was not enough spread in the
prop shaft bearings. Those big props would have one heck
of a gyro bending load on the prop shaft. All he had in
there were two closely spaced giant ball bearings.

From the looks of it all the bending load was transfered to
the rear main bearing of the 12A. I would make a 3D of the
whole thing as you have the parts so you have the dimensions.

Paul Lamar


Paul;
   20years ago I was building a PSRU out of the top 1/2 of a Bell 47
gear box. The project got lost during a messey divorce.
Does Dave Gaeber live in Florida?. I used to know someone in Florida that
was building a twin engine push pull air raceing rotary. He had a landing
gear collaps during taxie tests and I lost contact with him. Would you be
so
kind to email me his email address?
          Thanks in advance,
                         Russ Mac Farlane

Yes that was one and the same. Tragically Dave died in a crash last
year. The airplane is still around someplace.

Paul Lamar

--

Paul,

Yes it appears rather short coupled. There are several features in the
Bell 47 'as is' that I don't like. One is using a spine between the
planetary mount plate and the outer bearing. The slop, though minute,
pretty much guarentees the need for an inner bearing like Dave used.

For the purposes of proper alignment only, a transmissions and Mistral's
PSRU, both rely on a simple nose on the end of the drive shaft after the
spline to insert into the e-shaft.
Like Mistral I have an intermediate bell housing with a central bearing
there to take any load that it might see from the prop gyroscopic behaviour.

I doubt I'll be doing acrobatics in the SeaWind but it will be experiencing
turbulence.
Distance from the PRSU front to back of the prop flange is about an inch.
From the front to the center of the MT prop arc is about six inches


Doug in Japan

Interesting item, looks robust.

A few questions:

!: What is the reduction ratio?

2:I don't see any torsional resonance damping. Any ideas ?

3: Are any of these items still around?

Thanks.

James Freeman


James,

The unit reduction is 3:1.    That implies with a CS prop, engine would be at 8200 rpm at flat take-off pitch.

I designed a combination six inch pulley and torque dampener that would bolt to the counter weight.  I never made it yet and will probably go with something larger in diameter closer to what Tracy did but with harder and more rubber inserts.  This issue is pending.

Someone on this site mentioned finding a few units on the east coast of USA  a few months back.

If you ask  the Bell 47 user groups they usually quote a figure of about $30k for the whole unit which is actually two planetary units back to back.  Please understand the original reduction unit was  9:1.    (3 times 3). I mention this only to give you an idea that the outer planetary bearings were design to experience a maximum output shaft rpm of 300. This was done by a smaller gear set feeding the larger gear set I am using.  Dave Garber tossed the parts of the smaller gear set but used a sawed-off outer planetary plate as the input into his gear set.
You can see it attached to his counter weight on the photos Paul posted.

All in all the Bell 47 unit is robust and the case could be duplicated if someone is interested. No one seemed to have any interest so I haven't  gone to the trouble to have it 3-D scanned yet.  The expensive parts would be the forged ring gear which has a lip built into its outer edge.   This allows the attach bolts to react the engine torque directly.  It also will be instrumental in containing the centrifigal forces of the planetary gears. In Tracy's design the torque is transmitted to the aluminum housing and from there to the attach bolts.  Fine as his has a thick robust  case.

The Bell's planetary gears 'as is' will not accept  high speed roller bearings.  The gears are hard enough for the max line speeds (but their internal diameter is too large and there is a ridge inside).  For testing purposes only I will have the ridge grounded out and press in a sleeve to reduce the diameter down to accept a caged roller bearing.

My challenge is whether to machine up a combination bell housing (with central bearing) or go with the RX-4 bell housing I have.   Starting from scratch would save me an inch in over-all length engine prop length look nicer, and be repeatable.  I must admit I worry about trusting the strength of an automobile bell housing to keep the PSRU attached, even if if does have a nice oil feed bearing in the middle of it.

Conical housings and suitable attach methods to the engine are desirable.  With the risk of sounding like a broken record, Mistral did the PSRU correctly and I would like to follow their lead but create some engine mounts into it.

I tend to over engineer things on the first round of innovation.

Here are some Notes you may find useful.
Bell 47 PSRU notes â?? Doug Fir

I count 23 teeth on the planet gears. Sun gear has 46. Ring gear has 92

The sun gear rpm is the same as the engine of course. The overall ratio is 3:1 so the planet cluster (output) travels at 1/3 rd. the speed. So for each engine revolution each individual planet gear rolls only 1/3 along the circumference of the outer ring gear. [see position 360 in the website picture] This information is important when calculating the gear line velocity because that distance divided by time will result the velocity of Meters/sec.

Gear line velocity.

Outer diameter is a little less than 12 inches so circumference = pia D or 3.14 ft.

In max red line limit of 9,000 rpms in each engine rev. the planet would roll 1/3 rd of the circumference or 1.046 feet. At redline 9000 rpm that equals 9,414 ft. per min.  *Note max redline occurence assumes 10:1 compression Renesis rotors, a MT constant speed prop, and a malfunction in pilot or prop control.

*NASA considers gear line velocities over 15,000 ft/per min. to be high.

At continued cruise speeds, the gear line velocities will be around 6276 ft/min. or 40% of NASA limits. Go here for many papers from Glen Research Center

http://www.grc.nasa.gov/WWW/StructuresMaterials/TribMech/publications.html

Planet gearâ??bearing speed.

In 3 engine revolutions each planetary gear rotates 4 times around its own shaft. At half this amount  it can be seen from the pictures that the planet gear has rotated twice. and travels 1/2 the way around the perimeter of the ring gear.  [see position 540 degrees in website picture] As can be determined by manipulating the device and by a simple math calculation the planet gear moves 1/4 of the way around the ring gear  (one revolution around its planetary axis) as the sun gear moves 270 degrees.

*This is important to understand when calculating planet gear bearing speeds. If for instance engine/ prop redline are considered to be max. 9000 rpms then, the planetary gear bearings must be able to spin to 12,000 rpms. The selected roller bearing is rated at 20,000 rpms.

Bearing loading

Centrifical force throws the bearings and gear outwards. Ring gear contains the loading force to a main degree but the bearing will have some load on them.

Oiling

An oiling mist is the preferred method as flooded roller bearings may skid instead of roll.  Like the RX-4 semi auto engine, a hole, dowel and 0-ring can be placed to interconnect with both the engine oil chamber and the bell housing oil chamber.  Oil is introduced to both the front bearing and from there through the hollow sun gear shaft into nozzles spraying at or thorough the  the sun gear.





Spline size

I used the Tilton sport discs for a Volvo in my Mazda input shaft Saenz transmission. And Chevy 15/16 fine spline on the Richmond gear transmission. The rotary, even the three rotor, have dismal torque and there is little problem with shaft size. Drag racing with a heavy flywheel would need a deeper evaluation.  Lynn E. Hanover [Ed. Note the prop is a heavy flywheel. Use a 32mm spline which is the size of the Sun gear shaft. ]

The standard 22mm OD size is OK for testing.



The two rotor side port has only about 165 foot pounds of torque and the p-port is about 200 foot pounds. Given the same RPM here should be less of a problem with the rotary say compared to a four cylinder. The main reason we use a damper is the resonant frequency of a planetary with a prop is about 2000 engine RPM. Any time you separate a large rotating mass (prop) with a gear train or long shaft to the engine you are going to have torsional problems at some RPM.

Paul Lamar

Ed. Note.   A turbo charge p-port over 300 hp may see 220 foot/pounds of torque or more. The Bell 47 unit would multiple this torque to 660 lb pounds and/or more.   Something to think about when designing the PSRU mount is the breakfast cereal jingle for Rice Crispees.   "Snap, crackle, Pop"


Doug in Japan




-- 
The Rotary Engine NewsLetter. Powered by Linux.
ACRE NL web site. http://www.rotaryeng.net
Youtube key word PaulLamar2
Copyright 1998-2009 All world wide rights reserved.