On Feb 2, 4:09 pm, Rip wrote:
According to Lycoming, an even bigger danger is that, at -10 deg. F,
engine tolerances are such that there may be zero, none, nada clearance
between important surfaces, like the crankshaft and it's bearings. The
aluminum case shrinks more than the steel crank, and one start under
those conditions will, not might but will, destroy the engine.

Rip

Here's Lycoming's take on it:
http://www.lycoming.textron.com/main...ldWeather.html

I don't see that they warn about destroying the engine.
Damaging it certainly is likely if care isn't taken. The aluminum case
shrinks at about twice the rate of the steel crank, but the crank's
bearings are steel half-circle shells that are in crush (ends butted
against each other) in the case bores. They will shrink at the same
rate as the crank and prevent the case from binding the crank too
much. Bigger danger in cold weather is running the engine too hard
before the steel cylinder barrels have warmed up; the aluminum piston
can expand enough to cause scuffing as the clearances disappear.
Engine shops see that frequently enough. The two-stroke engines in
ultralights are particulary fussy that way.

IIRC: Coefficient of linear expansion of aluminum is .1244
units per degree F; of steel it's .0655. I'd have to look it up to get
it exact.