Home
LINKS
Email:
moderator@CH601.org
Cooling your Engine
First, air will not blow through a radiator...it must be sucked through. That
means that the exhaust side of the radiator must dump into a low pressure
area...a low pressure area created with louvers, lips on the cowl, slats, flaps
or by exhaust system education. A radiator sticking out into the air stream is
an air-brake, not a cooling system. Now before someone says that so-in-so did it
that way and it worked, let me say he was lucky that some oddity of his
installation caused it to work. Such installations are not reproducible except
on some really slow flight type low hp type aircraft.
The most efficient cooling system from a simplicity and dependability point of
view has the radiator in the engine cowling with a dedicated air supply.
Starting with an opening at the lower front of the cowl of about 55 square
inches (works for 200 HP + or - 25%). Air enters into a duct work extending from
the lower nose bowl opening back to the face of the radiator. The cross section
of the duct should expand from the intake size back to the face of the radiator
with walls of the duct diverging at about 7 degrees. This is optimum but in
practical application most ducts are not long enough to diverge at that angle
and get large enough to cover the correct size radiator in the distance from
nose bowl to firewall. In the real world walls diverging at 15 degrees works
essentially as well. The object here is not to "blow" air through but to give
the air room to slow down and build pressure. A cooling system through which air
slows, expands, builds pressure, gets sucked through the radiator into a plenum
then sucked out of the cowling works with virtually no drag penalty (certainly
no worse and a similar hp air-cooled engine with a potential for much less
drag). The setup must have a pressure differential from the front side of the
radiator to the back side of the radiator. The catch phrase is that the duct
should be "divergent-convergent." Divergent to allow slowing down and pressure
to build (trading velocity for pressure). Expansion through the radiator then
convergent to allow for some acceleration of the air velocity during compression
to more closely match the outside-the-cowl velocity into which the, now hot, air
will dump.
In this HP range an adequate radiator should be about 19 X 28 inches, two rows
thick and made out of aluminum (one from a Ford Taurus, Mustang, T-Bird, etc.
SHO or equivalent from Summit Racing, Summit's own brand is relatively
inexpensive. Salvage yards are good...don't spend more than $200 even for a new
rad) cross flow type.
The radiator can be laid horizontally or up to an approximately 45 degree angle
below the engine close to the firewall. The Zodiac cowling has more than enough
room for this style setup with flat 4s (Soobs). The intake duct can be
created with baffling on the upper side and the cowling on the bottom side. The
radiator must be absolutely sealed to the termination end of that duct. Any gap,
as little as 1/4" around the perimeter will allow most of the air to bypass the
radiator rather than flow through it (sounds impossible? build it any other way
and you'll have cooling problems). The engine exhaust system should be routed or
shielded so that it dumps no heat into this intake duct. If the exhaust pipe
must exit such that it goes through the duct, create an oversize sheet metal
through-pipe passage through the duct through which the exhaust pipe extends.
Air from the upper engine compartment will flow out around the exhaust pipe
cooling it and preventing too much heat exchange into the dedicated radiator
airflow.
The cooling air can flow up or down through the radiator because, when the
system ducts are built this way the prop wash is adequate to supply cooling air
for the longest ground holds. Since the air is slowed before entering the
radiator there is little necessity for streamlining the duct at the radiator.
The fact that the air appears to have to make a right angle turn through the
radiator in a horizontal layout does not cause problems...we are at this point
working with pressure differentials not streamline high velocity flow.
Streamlining does come into play in the exit plenum/duct after the radiator. The
"back" wall of the duct, the area into which the airflow exiting the back of the
radiator will impinge should curve, gently or relatively abruptly but
continuously towards the exit from the cowling. Sidewalls should also smoothly
converge to the exit, meaning no square steps or direction changers, radiuses as
gradual as possible but fitting the space available. You will hear rules of
thumb that the exit openings, because the air is heated and thus presents more
volume than the original intake air, should be 2.5 times the area of the intake
opening. With good extraction of this heated air, allowed by the fact that it is
exhausted into a low pressure area, the exit cross section can be as small as
the intake in some cases. In others larger, but I have never seen this exit area
have to exceed about 1.5 to twice the area of intake. Remember this air is
compressing and accelerating producing a high flow rate.
Tricks here that produce the cooling desired is that curved exhaust duct. Where
the exhaust duct forms with the sidewall of the cowling add some rear facing
louvers in the cowling (louvers create their own low pressure area). Where that
curved duct back wall leads the air to exit at the bottom of the firewall exit
low pressure is created with the offset to the cowling extending below the
firewall. A downward angled lip on that bottom trailing edge of the cowl will
add to the low pressure area creation. This sounds complicated...it is not and
is actually completely consistent with all that info on WWII cooling of water
cooled engines...including the belly scoop design of the P51 Mustang. This
system is no more difficult to create than good baffling for a Lycoming.
The other method that can be made to work is to place the radiator flat against
the bottom of the cowling and use the non-dedicated air flow that comes into the
cowling through the nose bowl holes around the prop. This has a slight
disadvantage of picking up some heat from the engine and most detrimentally from
the exhaust pipes. But it can be made to work even on the hottest days and may
be easier to create by those who are sheet-metally challenged.
Using a standard radiator (meaning something you do not have to get specially
made) this is typically accomplished by placing the radiator flat against the
bottom of the cowling. The radiator can be oriented length wise side to side or
front to back of the bottom of the cowling. The bottom of the cowling obviously
has to have an opening to allow the escape of hot air. And just as with the
dedicated air system the hot air has to exhaust into a low pressure area for air
flow to actually take place.
The radiator needs to be sealed to the bottom of the cowling...any gap at all
will allow the air to bypass the radiator core...and it will. One builder I know
went with the radiator-against-the-bottom-of-the-cowl method but could not get
the engine to cool once ambient air temps got above 80 degrees F. I finally had
a chance to see it at Sun 'N Fun. Everything was perfect EXCEPT that the rubber
sheeting that he used to fit against the cowl had gotten rolled inward towards
the radiator. When the plane was sitting still the rubber laid down tightly
against the inside of the cowl with no visible gaps. The radiator was large
enough, but temps just would not stay down. After looking for a long time I
suggested that the seal needed to roll to the outside so that under pressure it
was forced down tightly against the cowl. The builder did not believe me, or I
should say that he did not believe that so simple a solution was the answer. We
made the change right there in the display area by just poking the rubber seal
back the other way with the eraser ends of a couple of pencils. He didn't have a
chance to fly it until it was time to return home. Temps that day were 95
degrees F.
When he got home he called to say that the temps stayed lower than they did even
with ambient temps around 50 degrees with the seals turned the wrong way. (he
had originally flown the engine with no seals between the radiator and the cowl,
but temps were cold so the problems didn't show up)
The low pressure area can be created several ways. The whole bottom of the
cowling under the radiator can be louvers...tall ones placed close together can
have a square inch opening virtually equal to the dimensions of the radiator.
Another method takes a arched sheet of aluminum to create a plenum from the
front of the radiator to the rear.
Another method, my favorite, is the movable slat or louver. This is similar to
the fan louvers you'll see over a ventilation fan in a warehouse. The advantage
of these is that you can adjust them to control heat and drag.
This set up allows full power climb-out with plenty of air flow then at cruise
altitude close 'em down to just trickle enough air to cool...streamline!
The heat input from the exhaust pipes can be significant. Below is a picture of
Hirschfield's setup showing the heat shield to reduce radiation and convection
to reduce heating of the air intended to cool the engine. Note the shield
attached to the engine mount. also note that air also exits the cowl through the
hole in the firewall through which the exhaust pipe exits. This lets the air
that is heated by those pipes to exit rather than getting mixed into the
radiator air (Hirschfield also installs mufflers behind the firewall).
I hope this helps you get a head start on building your own cooling system.