Morgans wrote:
snip.

At every power pulse, the crankshaft winds up a little bit, and then unwinds
while waiting for the next power pulse. This can be a very large problem,
and can destroy a drivetrain. Aircraft engines get around it by making a
very stout driveshaft, but even then, some engine prop combinations are
plackered not to run at certain RPM's in continuous operations.

Do some research. There have been tons written on the subject, and even
though it does not make sense, driving a prop is way harder than a car
wheel, and avoiding the resonate problems.

Actually, aircraft engines don't deal with it very well at all. Large
radials deal with it by having a high reciprocating mass, and the 65
Continental class of engine deals with it by low impulse and again a
high reciprocating mass, but geared light aircraft engines have been
largely troublesome. The Continental Tiara was a disaster and so was
the GO-300. The planetary gear case Lycomings were a little more
successful but they were also heavy.

A good "car" engine that is suitable for ski boat use is suited to
aircraft use if 1) any resonant peaks in its internal configuration are
figured out first, (the boat will do that!) 2) a proper drive is
selected and 3) propeller loads are transferred to the airframe from
the drive and not the engine itself. Dave Blanton had no torsional
resonance problems, although he was probably a little lucky, and
careful study of Kiekhaefer's marine I/O and the Soloy Allison fixed
wing conversion (which use Allison helo turboshafts, different from
their purpose built fixed wing cousins) will be helpful.

A good autoderivative engin package solves a lot of problems aviation
users have lived with for so long they don't consider them problems
anymore, such as a prop stoppage destroying the entire lower end of the
engine, and not being able to run the engine without a prop or test
club.