Allow me to introduce some data to the discussion.

About 5 years ago,  I needed to know what the temperature rise would be
in a high performance GA engine compartment.

I took my turbonormalized IO550 Bonanza.  August evening.  Wind was <
2knots.

OAT was 105F at 5pm.

I took off.  Flew a big pattern with the gear down, cowl flaps closed.
Mixture set to about 75F ROP.

The oil temp went up to near redline.  CHTS up to about 415F.

Landed.  Taxied about 150 yards.  Shut down in a small closed T hangar.

I had previously installed some thermocouples on various components.

Mags.  Fuel pump.  Etc.


I then monitored the temperature.  

At 32 minutes after shutdown,  the magneto case reached a maximum
temperature of about 220F.  The engine driven fuel pump was a little
hotter, but I forget the number.

The core of the engine pump would have been hotter.

Now... what happens?

Prime the engine.  Crank.  Get the F/A right - - and it will fire - -
but it WILL NOT RUN.   It will do exactly what you see all the time on
the ramp.

It will start, run for 5 or 10 seconds and die.  

WHY ?  

Because the engine driven pump is flashing fuel to vapor and vapor
locking AT THE PUMP.

If you do this a half-dozen times - - you will eventually get the pump
cooled off enough to allow it to pump liquid - - assuming the battery is
not dead by the time you do that. (It takes several hundred amps to
crank the engine.)

OTOH... if you turn on the 5 amp fuel pump for 90 seconds,  it will
circulate enough fuel through the core of the pump to cool off the core
(you can feel it cool off if you stand next to the open cowl and touch
the pump before and after the 90 second purge with your hand ...  but
you can burn your hand, so be careful) to a low enough level so that it
CAN pump liquid.


Regards,  George