Part of Kohler's engine test lab includes spin-testing flywheels. Each flywheel is properly installed to a clean, dry, motor-driven-and-monitored shaft, then hand-torqued just like the manual instructs. The flywheel is mounted into what resembles a large safe with very thick walls and a big door. RPM is then run up to the point of burst. They need to survive at least 2X the normal rated speed of 3,600 RPM.
After bursting, everything that still stays on the shaft is out of balance and the lab floor will shake as the speed comes back down. When the door is opened, there will be a pile of broken parts with magnet bits stuck all over the inside of the chamber. All flywheels pass the 2X3600 mark -- 7,200 RPM -- and many of them hit 10,000 RPM.
Anytime a new flywheel is developed, whether for a new engine or as a new flywheel design for a current engine, it gets qualified/tested. And anytime a modification is made to an existing flywheel, it gets qualified/tested.
Many Kohler engines are set to run 3750 no-load speed.
Keep in mind that lubricating the tapers or using an impact wrench to torque a flywheel are always bad ideas. Broken flywheels result from lubricating the tapered shaft or over-torquing the flywheel bolt/nut. But properly hand-torquing a stock Kohler flywheel on a clean, dry taper will safely allow in excess of 5,000-6,000 RPM.
As noted above, this is Kohler's flywheel-test information. I'm still trying to get information on Tecumseh, Briggs, Kawasaki and Honda flywheels.
Now, having said that, and having watched Knoot's video, I need to say that the Kohler we ran to 5,700 RPM did not have any counterbalance mechanism inside, just a well-balanced crank, rods, pistons and cam. The Tecumseh V-twins being raced don't have counterbalancing mechanisms, either, but I don't know what RPMs they're producing. What I saw on Knoot's Briggs video looked like the oscillating counterbalance weight self-destructing, and the oil that came out of the crankcase looked quite dirty and black, which would tell me that the internals had a lot of wear from the dirty oil. This may have helped contribute to the self-destruct. What was the RPM at the point of disintegration?
And Paul keeps mentioning "insurance companies and a 3,650 RPM maximum." Which insurance companies? Would a stock Kohler running it's governed 3,750 RPM be uninsurable? Maybe the insurance companies are using outdated information from when the flywheels included the cast-in cooling fins. The new unfinned flywheels with separate plastic fans can safely spin much faster, as Kohler's tests -- and my own experience -- prove.
Respectfully,
Rodney