Hi Jarrett,

 I am not doing this because I think this is a good way
 to make a twin pusher. As my old late friend Vance RIP
 used to say I am doing it because it is an interesting project.
 In the mean times I am waiting for slide throttle parts.

 The problem with spiral bevel gear sets is the separating forces
 between the gears. This amounts to a point load of about
 a thousand pounds at 200 foot pounds input. The traditional
 way of handling this is use big heavy bearings and a heavy
 thick wall spool. I am thinking of using a large diameter, thin
 section bearing to take the thrust load and a fabricated sheet metal
 spool. The 3.5 inch carbon fiber prop shaft will be bonded
 directly to the ID of the spool.

 Consider the red device a rock solid buttress.
 What does an FEA look like in the spool.
 Material is 4130.

 Thanks

 Paul Lamar


 Paul, As I remember the detroit car companies mostly use hypoid gear sets.
Do they also have large separating forces. I remember something from an
applied mechanics course when I was at the University of Pittsburgh in the
late 1960's. The lecturer said that gears were designed with  one of two
curves. cycloidal was like  the trace of a fixed point on the side of a
rolling cylinder, and involute was like the end trace of a string unwinding
off a cylinder. It was suggested that these two curves would make a rolling
contact point at the right center distance spacing  with no change in
distance to either of the shaft centers. Having the wrong center distance
would cause the driven gear to increase and decrease speed at every tooth
contact due to the small change in the contact point to  shaft center
distance.

I used white grease to get the contact point right and flank space
for lash  on the carrier bearings. There was considerable practice with a
chrysler hemi in a sprint car. In looking at the conceptual design above,
how about a pair of timken bearings on the small end of the spool for
radial
and thrust loads and a large diameter spherical roller on the larger end
for
the radial loads there. This  would reduce the  required size of the thrust
bearings which seem more rpm limited in the larger diameter.Will a
universal
joint be necesary on the ford pinion carrier cartridge for access to the
shaft nut that retains the pinion bearings or could it be a flanged plate
with half round drive pins on bach that bolts to the yoke where the
universal joint caps would bolt up.

The flange might run smoother and allow
the drive shaft to be bonded to a flanged steel tube. One unrelated thought
is that if the engines can not be located far enough behind the firewall
for
CG while keeping the drive shafts inside the wing, the use of a full
length gearbox to gearbox drive shaft tube with hyvo chain reduction off
each engine would keep both props turning with a shut down engine. Also
both
engines could be right side up and pointing in the same direction. I think
the original sketch Jean posted showed one engine turned 90 degrees for
exhaust manifold construction prposes. A full length shaft eliminates
asymetrical thrust issues and prop synchronization. I am a real fan of
Tracy
Crook's PSRU but if an offset s needed to gear the drive shafts up to a
thicker section of the wing/and or CG isues are a deal buster this could be
a way out.

george grimes

Jean  It doesn,t seem to me that spoilers on the nose are far enough off
the aircraft center line to be very effective for yaw control. This idea
would  basically be large spoilers that turn laminar flow  into turbulent
flow and detach the boundary layer altogether. This would make a drag brake
but the disrupted flow over the wing would have serious implications on
wing lift. Just my opinion but I would encourage you to NOT DO THIS!!!.   As
for a hyvo drive the  shaft center  distances are critical and therefore
selection of the exact gear ratio is the first order of business. I have
somewhere a tsubaki catalog which indicates 100 HP for each inch of chain
width. Also recomended are the double fing fetters mounting which in these
sizes can transmit up to 1300 pound feet.The ring fetters require  a solid
shaft section for mounting but the upside is a wide tolerence on shaft size
of .005 and a smooh shaft with no keys or splines required. It installs and
dismounts with allen wrenches, and is self centering. The self centering is
very important with engine speeds at 6000RPM. The mazda flywheel extends
past the outline of the bellhousing flange on the bottom so sandoffs like
Tracy Crook uses or something integral to the chain case is necessary to get
past theflywheel diameter. The starter would mount to the chaincase as well.
the chaincase  back and cover would have  support bearings for the small
sprocket with oil seals required as well. the driveshaft end of the
chaincase would have  one bearing and oil seals on both case halves. The
bearings go outboard of the larger driven sprockets. One would be back case
mounted the other cover mounted assuming the two engines are facing the same
direction. This is to eliminate overhung loads. Also only two bearings can
be on this shaft, as it is large and stiff. Even small misalignment of
bearings from manufacturing and assembly will cause catastrophic loads on
the bearings, hot running and failure. For the bearings I would spec piloted
housed spherical bearings which include the oil seals and run within their
S10 life at these diameters and speeds.Two isolation bushings on each
chaincase would be the engine mounting points for a total of four bushings.
The two cases would need some kind of connection to each other -tubing? or
1/4" aluminum sheet metal. I may have the components drawn up somewhere on a
backup disk but my removable hardrive and softwear disappeared last week It
would be the bearings sprockets frin feters etc to scale with a part number
list for the various catalogs. This was for a dodge viper 498 HP  engine and
I don.t remember what reduction ratio I used but it was a large diameter
prop so reduction might be close enough to get right by changing the large
sprocket tooth count. Two problems. If the engine set is rubber mounted how
does engine movement effect  the spiral wound carbon drive shafts alignment
and loading. What ever you planned for the single engine per prop set might
work.The second is that this runs in oil and must be enclosed. If the
enclosure is similar in weight to what would mount  a set of toothed belt
sheaves it would still be heavier. I think Paul posted a sketch for
counterotating props with two engines and a line shaft last week. The major
differences are moving the line shaft from the top down to the spark plug
sides and both engines drive the same shaft (simpler) also no prop thrust
load is involved (simpler) since that occurs at the 90 degree gear boxes.The
toothed belts on our dragster engine pulled 500HP on a 1471 blower drive
with 5 inch wide belts so 250 Hp should be doable for continuous duty.As a
tip, if you drill 3 holes 1/16" diameter in the root of the pulley teeth the
blower whine sound goes away and the belts  don't go walk about so
regularly.


george grimes