Bob,
I formed the belly pans in an aluminum die. When I machine a flywheel, the brass or aluminum stock is mounted to a shortened crankshaft that is the full diameter out to the taper. This piece is mounted in a precision four jaw chuck. The cut down crankshaft is carefully zeroed to + or - .0000" runout. It will be used to set the precise angle needed on the lathes compound when cutting the taper in the back end of the brass or aluminum flywheel stock & it will be used to machine the front side of the flywheel piece. Notice the oversize stainless steel piece the is shrunk into the flywheel stock as a first step. It will be finished machined to a size that causes the collet body to be located in two axes precisely each & every time it is mounted. The collet body is held to the flywheel face with the two 6-32 screws shown. "IT SHOULD NEVER BE MOUNTED BY USING A MACHINED SCREW THREAD"! The collet body has a hardened & ground seat for its collet to sit in & a 32 pitch thread for very aggressive tightening. The wrench size on collet body & the clamp nut is 5/8".
I machine the entire back side, including the tapered seat, "V" groove, "cut" knurl if wanted, final diameter & the 8 mm through hole. All of these machining operations will be dead concentric to the tapered seat that was cut in the beginning operation. Turn the piece around, use the zeroed crankshaft (.0000" runout) & mount very "tightly" the piece to be machined on the front side with a 8 mm screw. Now, the oversize shrunk in stainless steel piece will be machined to size as well as the front side of the flywheel. The front side surface will be square & concentric to the locating hub. I use .5000" + .0002" or .0003" interference fit for the 1/8" wide locating hub that locates inside collet body.
The collet body is machined & the collet seat in the collet body is hardened & ground. Again, I start with the back side of this piece & use the two 6-32 holes in the piece to mount it to a zeroed (.0000") holding fixture. The clamping collet, which is split from end to end, is also hardened & ground in a holding fixture. Splitting the collet from end to end cause it to close over its entire length, giving a very high clamping force! There is a hardened steel thrust washer used between the clamping nut, which is not hardened, & the hardened split collets back end. All of the pieces shown are interchangeable between nitro engine flywheels & gas engine flywheels with the universal mounting system used. I can use the same type collet body with this locating system to mount a ball & pin type universal on any flywheel. Cables that are used in these split collets are solid over the distance they sit in the split collets with their ends silver soldered solid. The collets can also be used with solid shafts. The .0002" runout I referred to is measured at the output shaft.
I have used the same collet assemblies without any failures for more than 10 years. Ask yourself how much runout can a 2.000" OD, 8 ounce flywheel withstand when operating at 20,000 to 30,000? Better yet, how much runout can the cable withstand before it begins to adversely affect the performance of the boat?
Jim Allen