| Historically, radial
engine fighter aircraft are considered to have greater drag
than for liquid cooled fighters, which have a much reduced
frontal area. There are quite a few reasons for believing
this consideration to be facile and false. Not the least of
these is that the fastest piston engine aircraft in the
world is Rare Bear. Also, the finest fighter aircraft of WW
2 was not the Spitfire or Mustang, but the FW 190 D9.
This amazing aircraft, while using a liquid cooled in-line
Jumo engine, used a radial-configuration cowling for the radiators!
Here is food for thought indeed! The underlying problems
in extracting engine power are piston speed, engine cooling
and cooling drag. We have little control over piston speed,
but engine cooling and cooling drag are often very poorly
implemented, due to a comprehensive ignorance of how to
optimise them. These two factors are linked, but first let's
look at the classical mistakes made by modellers (NB:
not by aircraft engineers).
The typical modeller thinks that the faster the air flows
through the cowl, the better the cooling. And if the engine
is too wide, he shaves off the cooling fins on the engine
sides. The air blasts straight in on the front of the
cooling fins, races past the sides and on passing the rear,
misses the fins at the rear altogether.
Lesson 1: The greater the area of cooling
fins available, the greater the amount of power that can
be extracted from the engine. More fins, more nitro, more
power.
Lesson 2: The faster the air rushes thru
the cowl, the more uneven the cooling, the more nefficient
the heat transfer, and the higher the cooling drag.
Lesson 3: Uneven cooling causes power loss
thru cylinder distortion. The object is to cool the cylinder
uniformly, to keep it round and maintain the piston/cylider
seal: leakage past the piston loses power and upsets
tuning. Return now to Rare Bear (full-size) and
the Sea Fury. If you care to look inside the cowl, you notice
some remarkable things.
1. There is a large space in front of the engine! This
is the plenum chamber.
2. There is a very narrow annular air inlet, yet sufficient
air gets in to cool 3000 HP!
3. The INSIDE of the cowl is streamlined!
4. There is a fairing over the crankshaft not unlike a
mirror image of the spinner!
Lesson 4: Air is most effective at cooling
a cylinder when it passes over the cylinder at LOW speed
and high pressure
Lesson 5: The idea behind the plenum chamber
structure it to slow the air by expansion after it enters
the cowl. When the air expands, its pressure increases, just
a fact of nature. Remember, we live in a sea of compressed
air, at roughly 14.7 psi. By being clever with our aerodynamics,
we can raise or lower this pressure. We use wings to lower
the pressure, and plenum chambers to increase it.
Lesson 6: When the 250 MPH air over our
model enters the flow annulus, it doesn't want to expand nicely
into the plenum just because we want it to. The shape of the
INNER cowl has to avoid turbulating the airflow and/or permitting
stalled flow to exist. The INNER shape of the cowl is probably
more important than the OUTER shape. It certainlyis from the
point of view of cooling and cooling drag. The latter can
be 40% of the total airframe drag!!
Lesson 7: The lip of the cowl inside the
cowling has two components: the spinner side and the external
cowl side. By placing a mirror-imaged spinner over the crankshaft,
the air avoids stalling on that side and expands down toward
the crankshaft smoothly. On the externalcowl side, the lip
must be rounded and flow smoothly back toward the cylinder
head area,again to avoid stalling and allow smooth expansion.
The smoother the expansion, the greater the pressure increase
in the cooling air. That spells more cooling and less cooling
drag.
Notice that nothing has been said about getting the air
out of the cowling! While that is important, thehard
work in cowling design is getting the air IN. Getting it out
again is not nearly so bad. There needs to be a big enough
hole for the air to get out, and it needs to be at a point
wherethe pressure over the fuselage is low. That point
is where the speed of the external flow is large. The
scale location is fine, but the size of the hole can affect
cooling drag.
OK. So far, we have really been talking about full-size radials,
where there are cylinders everywhere you look. But in
a model, such as the Herbrandson 289 powered Rare Bear, there
are only 2, and they are horizontally opposed. Also, there
are carbi are intakes stuck in there as well. So how to handle
this situation? Well here is a fun idea. Have a
look inside a full size light aircraft with a horizontally
opposed engine. Chances are you won't see a cooling fin
anywhere! Just some baffles and a big hole.
Lesson 8: The space above a horizontally
opposed engine is the plenum chamber. Those holes in the front
of the cowl don't go to the cylinders, they go to the plenum
chamber. The first thing the hotshot guys do is make those
holes as small as they can to reduce excess flow and hence
cooling drag.
Lesson 9: The air from the plenum chamber
is then directed DOWNWARDS through the cooling fins, via a
cunningly arranged set of baffles. These baffles have the
work cut out for them, as we really want that high pressure
air to pass over all of the fin surface, which is probably
not possible, but it is what we aim for. Back
to the model. We have a problem: we need air for the carbies
AND air for cooling. That is NOT the same air. we need
different air! So here is the trick. We want two plenum chambers:
one for the carbies, and one for cooling. This is why
I like radial cowls. Make the upper volume of the cowl
for the carbies, and the lower cowl volume for the cooling
air. This way the flow into each can be optimised, assuming
we are really smart to begin with! (warning: the carbies
may not be set up for high pressure air from a plenum).
Lesson 10: When we look into the annular
cooling ring, we should not see the cooling fins at all. All
we should see are upper and lower intakes into the two plenums.
Lesson 11: The two plenums must be completely
separate.
Lesson 12: The lower plenum must be baffled,
so that UPDRAUGHT flow takes place past the front and rear
fin surfaces. The front updraught air must then be directed
back over the upper fin surfaces.
Lesson 13: There must be no leakage from
either plenum, except for air going over the fins and into
the carbie intakes.
Lesson 14: the baffles may be flexible in
cases where flow velocity is not a problem. ie upper plenum.
Ok, that is basically it. Now to comment on your existing
model cooling systems: here is what I see.
1. The design is set up for high speed airflow thru the
cowl, leading to poor cooling and high cooling drag
2. While the engine may not show signs of overheating,
the cylinders will be out-of round while hot, power
will be lost and nitro content limited
3. Some of the fins have been machined off. Start again
with new cylinders, you need all the fins you can get.
4. A real attempt has been made to provide cooling flow,
which is good, but it is also misconceived.
I regret that to achieve low cooling drag and efficient cooling,
most existing designs have to be scrapped and a great deal
of work done to implement the above principles and lessons.
I am sure a learning curve is involved. If the existing
system does work, then don't bother. With a new design, you
can easily run into a fresh set of problems which will need
to be worked thru. But you should end up with more power and
much less cooling drag.
Finally, radial cowls require a propeller design matched
to the cowling shape. Stock propellers can lose a lot of thrust
on radial cowls. Read my website on the AT6 prop.
|
