Subject: Bell 47 Housing.
From: Rotary Engine
Date: 11/14/2006, 6:20 PM
To: AA-me


Hi everyone


I am now measuring and preparing to draw the Bell 47 housing parts in Rhino
so I can share with other individuals. I think before I put it back together
I may have some  molds of the front magnesium housing and the intermediate
aluminum ring made incorporating a few enhancements. I will also want to
draw up and spec out the other parts in case some of you may want a similar
unit in the future. These parts are pricey though. The planetary holder
appears to be forged and the flange that transfers power from the planetary
to the prop shaft was cut from a huge chunk of chrome alloy steel. At almost
three inches ID this internal spline flange is way over designed for our
use.


I also need to purchase a manual so I have an idea on the tolerances and
different nomenclature use in identifying the parts.  Next I'll need to seek
help in identifying the different alloys and heat treatments used.

One person that sells Bell parts recommended I should call the following
number.  Could one of you state side guys call for me?  I can't dial a toll
free number from Japan obviously. Here is the message I received from him.

Mr. Fir
You may contact Bell Helicopter Textron (PSE Light) at
1-800-363-8023.

Ask the individual for the phone # of publications in Texas; you may
purchase support manuals direct from publications."

Good luck with your project.

............

My preliminary research has shown that the Bell unit had two back to back
planetary gears sets.   Both had 3:1 ratios which together translates to a
9:1 total reduction.

First the specs.

*All measurements taken at the gear tooth mid line. Inches are
approximations.

Ring gear      234mm  - 9.25"
Sun  gear      120mm  - 4.72"
Planet gear     57mm  - 2.25"
Proof       2 x 57 + 120 = 234

Gears  are approx. 35mm thick.

 Planet gears have an internal 5mm boss and their roller bearings are
approx. 8mm in  diameter and 15mm long. They are individually inserted from
each side of the gear sort of like loading a revolver with 38 caliber
blanks. The outer diameter of the inner race OD is about 38mm.


Circumferences:
Ring            736 mm
Sun             377 mm
Planet          179 mm
Inner race      119 mm
Gear shaft       57 mm
Roller bearing   25 mm

The PSRU was designed to deliver low speed and transmit the high torque of a
large helicopter rotor.  It probably was over designed and could handle
1000Hp but I have not been able to confirm that claim. Here is what we do
know: For each revolution of the final sun gear, each planet gears rolls 2.1
times as much. (377/179)


On the Bell 47 unit dual planetary, the maximum rpm output would be about
300.  This means the INPUT to the final planetary sun gear was about 900
rpms. This translates to the planetary gears rolling around their bearings
at 2.1 times that speed or 1890 rpms.  Not so fast.  Now from here on I am
unclear about what happens in the actual dance within bearings captivity so
anyone please feel free to jump in and correct me.

They appear to react like little planetary sets.  Each roller bearing
contacts the outer part of the race (within the gear) and the inner 18mm
shaft at the same time. Well almost. Obviously some clearance is needed.
Reason would dictate that they are thrown outwards at speed, but then again
the whole gear is thrown outwards within the ring gear. Anyway,  as the
outer bearing race is 119mm in circumference, each individual bearing would
roll around the outer track about 4.76 times times the gear rotation if
calculated from the OD. (119/25)

If calculated from the ID (the inner gear shaft pin) about 2.25 times (119/
57mm)

 The high number (1890 X 4.76) is about  9,000 rpm.  and the low number is
about 4,250 rpm. Probably neither number is correct.


The elliptical paths of bearings is academic as we usually we pick a bearing
based on the shaft speed, and different loads applied to it and leave it up
to the bearing engineers to figure out what is actually going on inside the
races.  My curiosity has been piqued because I noticed that some of the used
bearings had large flat spots on them. Hmmm????   I was told the original
Bell 47 gear box  had trouble with the fiber bearing spacer/holders because
they would often deteriorate after 300 hours of use.  Dave Garber eliminated
them and stuck another bearing in the slot. His clearance between bearings
looked minimal if any at all and I wonder if his bearings could  spread out
enough at  higher speed.   Dave, has flown  his plane a few times to 1000
feet AGL but how reliable his bearings are is my biggest concern.   Dave is
determined to seek help in getting his plane prepped for the RENO air races
and of course my plane has to be reliable for much longer flights.


When we pump the Bell unit up to the speeds Dave and I  will be using,
things become interesting.  With shaft inputs of 8200 rpms, and using only
the larger of the planetary clusters, the prop rpm will max at 2733 rpms for
take off. The input to the sun gear is still 8200 rpm  however and as such
and the planetary gears will be  spinning on their bearings 2.1 times as
fast or  17,200 rpm.  Ouch!

 I can just imagine what those large roller bearings logs are doing at those
speeds. The phrases 'oil heater' and 'spin weld' come to mind.

Obviously I need another bearing and possibly another design profile for the
planet gears.

Notice I haven't discussed gear line speeds,  planetary bearing loads due to
centrifugal forces, or oiling and windage  yet. That is for another day. I
have noticed though that the Bell unit as well as Mistral's PSRU support the
prop shaft and planetary cluster via an internal sleeve or bearing that is
integral to the e-shaft.  That is wise IMHO. Those planetary gears  have
enough work to do without handling prop flywheel permutations as well.

 About now is the time I am starting to feel all warm and fuzzy seeing the
fine work Mistral has done:)  But hey I am not bailing just thinking.


DougFir

I think it is 2.4:1. As I recall we measured it.

 You have to put .jpg extensions on those file names
or browsers will not display them. In fact they are
not even jpg's. What gives? What kind of camera did you use?
Where did you get those files? What kind of computer are
you using? Perhaps you need an American computer instead
of a Japanese computer :)

OK I have some pictures of the Bell box so here they are.

 Can you add a hyd constant speed prop feature?

  Paul Lamar ...No rotor no motor.


Doug,
Great work !- if we can just take a step back and reassess the facts, so
everyone can follow your assessment.

By the Photo's, he sun gear has 46 teeth and the ring gear has 92 teeth,
therefore the calculations for the ratio is Ring Gear ( RG) divided by(/)
Sun Gear ( SG) = ratio, therefore RG92/ SG46 = 2 so we have a ratio of 2 :
1.
As the calculations for turning in the same direction require the ring gear
to be held, the calculations are RG/SG +1= Ratio ( 92/ 46+1) = 3 : 1 Ratio,
as you have stated. Obviously the ring gear shown in your Photo's are
designed to be held so that's an added bonus and confirms the ratio of 3 :1
is the ratio initially it was designed for.

Next is the planet gears, for which the calculations are (RG - SG) / 2,
92 - 46 = 46/ 2 = 23. I couldn't count the  planet gear teeth, but assume
that number is correct.

So we have even number toothed sun gear and ring gear with an odd number
planet gear which makes this a perfect set-up for tooth wear patterns. Also
straight cut teeth don't have the thrust problems associated with the Ford
unit, although may be noisier.

One thing I would check with the manufacturer is the RPM limitations,
especially in regard to the bearings. If information is unavailable from the
manufacturer, I would check with a bearing supplier to see what RPM the
needle bearings can handle. I would also ask for the HP limits and life
expectancy of the gears- even if you have to have an independent engineering
assessment made.

If you want to go further you could have metallurgy assessments made on the
gears, they may vary from gear to gear.

One interesting point to make is the gear tooth (size) selection may be for
a lower RPM - which may limit the RPM your seeking. This is something people
generally don't think of. I don't think it's a problem in this case, by just
viewing the Photo's; however it would be a good thing to know for sure. The
general rule of thumb is that the bigger the tooth in relation to the size
of the gear, the better it handles the RPM - sounds counter intuitive to me,
but them's the rules.

Hope that helps you Doug and also those trying to follow your development.
George (down under)

Doug sent only one picture that I could read George.
The rest of the photos I took in Florida at Dave Garbers house.
I also scanned and uploaded the calculation chart.

Paul Lamar

The Rotary Engine NewsLetter. Powered by Linux.
ACRE NL web site. http://www.rotaryeng.net
Copyright 1998-2006 All world wide rights reserved.