> Tracy has argued that all your cooling air comes in
> at 200mph, then has to
> slow down to almost 0mph to go through the radiator,
> then accelerate back
> to 200mph again. All that work being done on the air
> costs you -- it's
> drag. So he says that you want to put the minimum
> amount of air through the
> radiator.

Any air you take on board; slow down, speed up, etc is going to
represent some drag.  But then, it's not like we have a choice.  It
takes a certain amount of air flow to remove a given amount of heat with
a certain increase in it's temperature.  It is readily calculable.  The
key is to handle the needed air in a way that reduces it's energy as
little as possible.
 
> I don't know whether its significant or not, but a
> couple of secondary
> effects occur to me...
>
> 1. If the air at the radiator is compressed, it will
> be able to carry more
> heat away per cubic metre. So you want to maximise
> the air pressure at the
> radiator. But compressing air heats it.

You have a given amount of energy (dynamic pressure) to work with
defined by the airspeed of your plane (half the density times the
velocity squared).  You can expand that air (slow it down) and convert
the dynamic pressure to static pressure.  The amount of pressure that
you can get is limited by the dynamic energy you started with, and in
relative terms, it is very little.  For our purposes in cooling, the air
con be treated as incompressible  because the density increase you can
achieve (and the resultant heating) is negligible.
 
> 2. Heating the air as it comes through the radiator
> will cause it to
> expand, and therefore produce thrust, which might be
> useful in accelerating
> the air back up to the 200mph exit speed. I vaguely
> recall that the P51's
> cooling system produced some thrust.

The heat added from the radiator does add energy and expands the air,
and helps to accelerate it back toward the free stream velocity, but
generally not enough to overcome the amount the energy you lost due to
friction, turbulence, etc. (pressure drop).  If you read the technical
papers on the P-51 ducting and cooling system you find that they did a
very nice job; but the notion that it produced thrust is a popular myth.

Al Gietzen