Corky Scott
The concept as outlined by Meredith, was that the air to cool the
engine should be directed to the radiator via a duct that expands at
the face of the radiator (which slows the velocity down and increases
pressure), then reduced behind it (which re accelerates the air). The
idea was to slow the air down such that it passed slowly enough
through the radiator to actually do some work (remove heat from the
fins), then be accelerated again to exit parallel to the slipstream.

I believe that both Hoerner and Kuchemann & Weber both published some
research on that very subject. Essentially, one wants to have trumpet bells
facing the front and back of the radiator. The trumpet bells serve convert
the dynamic pressure to static pressure (and back again). The downside for
us homebuilders (ok, "you homebuilders" :-) is that the length of the
trumpet needs to be a substantial length. IIRC, it is something like 2x the
maximum dimension of the radiator. So for a 20" x 20" x 1" radiator, one
would need something like a 40" long trumpet under the cowl. I don't think
I'm alone here in saying that ain't gonna happen.

Then along came two engineers, Kays & London, who researched ducting for
radiators ["Compact Heat Exchangers", Kays & London]. Their research
indicates that a simple wedge duct on the front face feeding the radiator
was only slightly less effective at recovering the dynamic pressure than the
Hoerner trumpet. And for our purposes, a wedge duct is completely doable.

What Kays & London also discovered was that clear space behind the radiator,
and (relatively) gentle bends in the ducting all around, were critical to
good airflow. IIRC, the clear space behind the radiator needed to be a few
multiples of the thickness of the radiator. So a 2" radiator might need 6"
to 10" of unobstructed clear space behind it, and even then it couldn't end
with a perpendicular flat plane like the firewall.

Some builders/pundits have proposed dumping the radiator air directly into
the slipstream by mounting the radiator so that the back face is pointed
straight out the side or bottom of the cowl (with a hole in the cowl,
obviously). The thought process goes: the radiator wants an unobstructed
exit, the exit air is slow, the slipstream is slow, why not just dump it
overboard immediately? The jury is still out on this line of thought.

Here's an intriguing addendum that never got utilized by North
American in the Mustang, or by any other fighters: Meredith wrote that
the jet effect would be greatly enhanced if the exhaust system could
be piped to discharge within the exhaust ductwork that carried the
heated air from the radiator. This is because any additional heat
would expand the air, increasing the velocity of the discharge and
therefore the thrust attained.

Atwood described this is being a quasi jet engine. However, the
problems of routing the exhaust back to the exit duct were considered
insurmountable and doing so was never seriously considered.

That was then, this is now. Routing the exhaust tubes into the
radiators exhaust duct is exactly what I am doing with my V6
installation in the Christavia Mk4.

Others have researched and even implemented exhaust augmentors in homebuilt
aircraft with some success. They aren't a free lunch, however. For
starters, exhaust systems are notorious for developing leaks at the most
inopportune times & places. The longer and more complex the plumbing, the
greater the opportunity for Murphy to make his presence known. Exhaust
systems get HOT. Do not underestimate the potential for the exhaust
augmentor to radiate significantly inconvenient quantites of heat into the
backside of the radiator.

With the radiator just in front of and at the bottom of the firewall,
there is room to bring the exhaust system in behind the radiator and
have it discharge facing out the exit duct.

The radiator wants an unobstructed path for the air leaving. Facing the
firewall is non-optimal. Adding an exhaust pipe makes it worse. Also, be
certain the exhaust pipe cannot radiate heat into the radiator.

The idea is to have the pipes terminate inside the duct so
that the exhaust pulses not only heat the air, they accelerate the air.

Great in theory. Good luck on your execution.

Do I expect this to waft me through the skys at 200 mph while burning
4 gallons per hour? No. You can't make a silk purse out of a sow's
ear. The Christavia will not cruise beyond 130 mph (if that) at any
engine setting below full throttle because it's a four seat fabric
covered STOL type. In addition, the less the power available (heat)
and the slower the speed, the less the effect.

What it will (should) do is make sure that the radiator/cooling system
functions properly by pulling the air through the system at all times.
The neat thing here is even during climb, when traditionally the
airspeed, and therefore air through the cooling system, is low and the
heat produced in the engine high, the air flow through the system
increases automatically because of the increased exhaust flow. More
power, more airflow, less power, less airflow.

Great in theory. I cannot help but wonder, however, if you could get more
"bang" for your effort if we instead concentrated on ensuring the propering
ducting of the radiator(s). Never forget that the oncoming airstream has
only about 0.1 to 0.2 psi available to shove the air through the radiator.
That isn't much, and every little obstruction hurts.

Also, try not to make the classic mistake of using a "NACA duct" to feed
your radiator. In almost every case it doesn't work. Even the writers of
the NACA papers describing the ducts said not to use them to feed radiators
(the ducts were designed to feed jet engine intakes).

At this point I'm just about finished with both headers. I have to
weld up the rightside flange that bolts the exhaust tubes to the
collectors so that I can unbolt the exhaust system and remove the
engine, then the exhaust system is finished and I can tighten
everything up, adjust the carburator float level and see how/if it
runs.

I have to make adjustments to the PSRU during the initial run to make
sure the belt is properly tensioned, then the prop goes on and I begin
a long series of engine runs during which I'll be keeping a record of
engine coolant temps, oil pressure, oil temp and coolant pressure.
The record will be available for the FAA once the airplane is ready
for it's inspection... sometime down the road.

Once you're to the point of taxiing (or even flight testing), one can
fashion a simple multi-point water manometer system using clear drip
irrigation tubing & fittings, cut up kitchen sponges (for diffusers), and
colored water. It might be a useful instrument in determining where the
airflow is and how to make the engine cool better.

Good luck!

Russell Kent