Ed, The big sun gear is 2.25 inches in diameter.
so tooth torque  load is 200 foot pounds or 2400 inch pounds.
so the tooth shear load is 2400 inch pounds divided by 1.125 or
2,133 pounds. The same tooth load is applied to the ring gear teeth.

The 2.85:1 ring gear is 5 inches in diameter with a circumference of
15.7 inches. It is 2 inches wide so the area is 31.4 square inches.
Therefor the shear load is 2,133 pounds divided by 3.4 or 68 pounds
per square inch shear. Well within the capability of red RTV. In
other words we can just glue the ring gear to the housing with red
silicone rubber. Make a real thin aluminum steel RTV sandwich one
square inch in area and see how much load it will take in shear.

Your front planet shafts are the same size as mine. .440 or so.
The torque load on the bolts is 2.85 times 200 foot pounds or 570
foot pounds. That is 6840 inch pounds. Bolt circle is 3.3 inches.
Radius is 1.65 inches so total shear on the six bolts is 4,145
pounds. There are six bolts so each bolt must take 690 pounds shear.
If we EDM them out to 5/16 and use 5/16 bolts.  The bolt diameter
.3125. Pi X r squared for the area. Area is.0767 each bolt.
690/.0767 then bolt shear stress is then 8,996 [psi] in each bolt.

Piece of cake. Even cheap bolts will take 100,000 psi.

As a belt and suspenders we can add six 3/8th dowel pins between the
bolts. The 5/16th bolts will then just hold the works together.


Check my arithmetic.

Paul Lamar


Only 690lbs shear applied to each bolt? An 8 penny nail can handle that
load. So yes, a 5/16 or 8mm rod bolt ought to be plenty.

The best setup is the alum prop shaft with the planet cage machined on
one end. Use longer shafts, No bolts required.

Ed Carver

The cost of the shafts is the problem. They need to be hardened and
ground. Also NC machining the planet carrier is tricky and expensive.
The 5/16 bolts will work fine for the 2.85:1 and 3/8th for the 3.17:1.
They have an over design factor of around 10.


Paul Lamar

Paul,  Just  running the math through again. 8996 pounds /.07666 square
inches appears to be 117,344 psi which would be a 17% overload not a
safety factor of 10.


From the posted pictures of the 6 pinion
planetary there appears to be 6  roughly trapizoidal shaped block
spacers located   between the planet gears. Possibly you could drill
here through the entire planetary  for 7/16 bolts and at a slightly
smaller bolt circle than the planet pins so as to work with the 7075
shaft you  have. Then ARP studs could be screwed into the end of the
propshaft to bolt on the planetary.


I apparently don't understand how the
EDM hollowed pins would work. If there is enough clearance to install
bolts then the thin shell of the pin that remains is pretty much
unsupported. It sounds like the plan is to have the 690 pound shear load
retrained by a shell .040? thick. The rotary engine may not have power
reversals on the output shaft but there are peaks and valleys of the
power pulses. It would seem that a safety factor of 1.25 over the peak
of the pulse would be necessary from a  safety standpoint.
George grimes

Thanks for checking my numbers George. I frequently make mistakes in arithmetic.

The 8996 is the shear stress in psi. Not the load. The load is only 690 pounds per bolt. Therefor the shear stress is 690 pounds of shear/.07666 bolt area = 8996 psi. Well under 100,000 psi by a factor of ten.

I did not make that clear by leaving out the units. I corrected the units above.

You bring up a good point about the thin shell of the EDM'ed shaft.
To calculated the stress we would need an FEA of one cylinder inside another with a few thou between them. My gut tells me it would be OK.
The load is only 690 pounds. This factor we hope to prove in ground testing. However the shafts are extremely hard.

We could also build a complete test rig and static test it with 570 foot pounds torque or more by a factor of three or four. Of course We would need a lever arm about ten feet long and I happen to have just the thing. A three inch channel we use to get the nut off the flywheel :)

Paul Lamar