Surface radiators for water cooled engines

On Tue, 15 Jul 2003 11:54:37 -0600, "Bill Daniels"
wrote:



Another area where I would like to see some experimental data is the
"Radiator Ramjet" (just to pick a controversial term) where the radiator is
in a tube and the heated air exits the rear of the tube at a higher velocity
than the cool air entering the front of the tube, theoretically producing a
small amount of thrust that offsets the drag of the radiator.

Bill Daniels
"Jay" wrote in message
. com...
Hi Bill,

You don't need experiental data for this Bill, you just described the
P-51 Mustang cooling system.

However, even with three heat exchangers putting out heat into the
exhaust air and a 1400 horsepower engine producing the heat, the
Mustang never actually managed to get a net thrust out of the system.
In addition, the point where the cooling system was ***ALMOST***
equalling drag was a very specific speed and altitude. I forget the
exact height but it was above 20,000 feet and the speed was over 300
mph. Only under those circumstances did the power being generated and
the speed being flown produce the necessary heat to accelerate the
exhaust air flow to nearly cancel out cooling drag. By the way, most
of the cooling systems did this to some fashion, but the Mustang was
the first to actually design the cooling system to really benefit from
it. This concept was researched and written up by a British
aerodynamicist by the name of Meridith, and the produced thrust became
known as the "Meridith Effect".

North American designed the Mustang's system using the best
aerodynamicists available at the time and with virtually unlimited
resources to manufacture the kind of heat exchangers that would work
in this environment.

By the end of WWII, almost all research into liquid cooled systems
came to a halt as jet powered aircraft became the future for military
aircraft.

I'm not an aerodynamics engineer, just a home builder. But my
impression is that most relatively slow homebuilt or GA airplanes do
not produce the heat needed to really accelerate the exhaust flow to
make much out of the Meridith Effect. After all, we're always leaning
out and cruising at reduced power settings. We have big wings, for
the most part, and a lot of drag. Something really slippery like a
Long EZ or Vari EZ or Glassair or Lancair might be fast enough to
benefit, but getting the cooling system designed and fitted within the
tiny wetted area of the fuselage might be nearly impossible.

From my personal point of view, it's far more important to make sure
the cooling system does the job all day and every day and on the
ground too than to agonize over a few mph, real or imaginary.

Corky Scott

"Corky Scott" wrote in message
...
On Tue, 15 Jul 2003 11:54:37 -0600, "Bill Daniels"
wrote:



Another area where I would like to see some experimental data is the
"Radiator Ramjet" (just to pick a controversial term) where the radiator
is
in a tube and the heated air exits the rear of the tube at a higher
velocity
than the cool air entering the front of the tube, theoretically producing
a
small amount of thrust that offsets the drag of the radiator.

Bill Daniels
"Jay" wrote in message
. com...
Hi Bill,

You don't need experiental data for this Bill, you just described the
P-51 Mustang cooling system.

However, even with three heat exchangers putting out heat into the
exhaust air and a 1400 horsepower engine producing the heat, the
Mustang never actually managed to get a net thrust out of the system.
In addition, the point where the cooling system was ***ALMOST***
equalling drag was a very specific speed and altitude. I forget the
exact height but it was above 20,000 feet and the speed was over 300
mph. Only under those circumstances did the power being generated and
the speed being flown produce the necessary heat to accelerate the
exhaust air flow to nearly cancel out cooling drag. By the way, most
of the cooling systems did this to some fashion, but the Mustang was
the first to actually design the cooling system to really benefit from
it. This concept was researched and written up by a British
aerodynamicist by the name of Meridith, and the produced thrust became
known as the "Meridith Effect".

North American designed the Mustang's system using the best
aerodynamicists available at the time and with virtually unlimited
resources to manufacture the kind of heat exchangers that would work
in this environment.

Yeah, North American did well with the Mustang given that it was just old
"slide rule" engineers one generation ahead of me working on it. I imagine
the kids these days using Computational Fluid Dynamics programs and modern
materials could improve on the Meridith Effect - maybe a lot.

I don't want to get too far form the original posters idea on skin
radiators. That idea is worth some experiments too.

Bill Daniels

The Mustang's cooling system (an external combustion ramjet) is probably
about as good as it is going to get utilizing radiators for heat exchangers,
because the thrust produced by a ramjet is very dependent on internal
efficiencies (drag). A radiator is a very high drag ramjet heat source
(combustor) because of its large surface area, and relatively poor
aerodynamics. The net thrust of a ramjet type cooling system could be
increased if a more efficient (lower drag) method is found to transfer the
heat to the internal airflow.

I agree with Corky's statement that effective cooling is more important (the
Mustang's was inadequate for prolonged ground operation) than a few miles
per hour in cruise for slower aircraft. However, for aircraft cruising
above 150-175 MPH, I believe cooling drag is certainly high enough to be of
interest to any designer.

RJ

"Corky Scott" wrote in message
...
On Tue, 15 Jul 2003 11:54:37 -0600, "Bill Daniels"
wrote:



Another area where I would like to see some experimental data is the
"Radiator Ramjet" (just to pick a controversial term) where the radiator
is
in a tube and the heated air exits the rear of the tube at a higher
velocity
than the cool air entering the front of the tube, theoretically producing
a
small amount of thrust that offsets the drag of the radiator.

Bill Daniels
"Jay" wrote in message
. com...
Hi Bill,

You don't need experiental data for this Bill, you just described the
P-51 Mustang cooling system.

However, even with three heat exchangers putting out heat into the
exhaust air and a 1400 horsepower engine producing the heat, the
Mustang never actually managed to get a net thrust out of the system.
In addition, the point where the cooling system was ***ALMOST***
equalling drag was a very specific speed and altitude. I forget the
exact height but it was above 20,000 feet and the speed was over 300
mph. Only under those circumstances did the power being generated and
the speed being flown produce the necessary heat to accelerate the
exhaust air flow to nearly cancel out cooling drag. By the way, most
of the cooling systems did this to some fashion, but the Mustang was
the first to actually design the cooling system to really benefit from
it. This concept was researched and written up by a British
aerodynamicist by the name of Meridith, and the produced thrust became
known as the "Meridith Effect".

North American designed the Mustang's system using the best
aerodynamicists available at the time and with virtually unlimited
resources to manufacture the kind of heat exchangers that would work
in this environment.

By the end of WWII, almost all research into liquid cooled systems
came to a halt as jet powered aircraft became the future for military
aircraft.

I'm not an aerodynamics engineer, just a home builder. But my
impression is that most relatively slow homebuilt or GA airplanes do
not produce the heat needed to really accelerate the exhaust flow to
make much out of the Meridith Effect. After all, we're always leaning
out and cruising at reduced power settings. We have big wings, for
the most part, and a lot of drag. Something really slippery like a
Long EZ or Vari EZ or Glassair or Lancair might be fast enough to
benefit, but getting the cooling system designed and fitted within the
tiny wetted area of the fuselage might be nearly impossible.

From my personal point of view, it's far more important to make sure
the cooling system does the job all day and every day and on the
ground too than to agonize over a few mph, real or imaginary.

Corky Scott

On 17 Jul 2003 09:34:54 -0700, (Jay) wrote:

On the Meredith Effect. Intuitively I don't see how its supposed to
work. You take air in the front of the scoop, you heat it up, and it
expands (ideal gas law). How does the air know thats only supposed to
go out the rear?

Jay, let's review the path the air takes through a radiator type
aviation cooling system and what happens to it.

Before I begin, you should understand that the air from the cooling
system MUST BE EXHAUSTED INTO A LOW PRESSURE AREA. If you do not
understand this first point, you will have huge problems with any
system you attempt to develop.

So let's presume we have situated the exhaust opening such that it
empties into the afore mentioned low pressure zone somewhere on the
fuselage, and we're moving through the air at cruising speed.

The air is directed into the intake opening because we designed it
such that it scoops in air and is pointed at the airflow. It flows
into the intake duct and the duct expands at anywhere from 7 to 15
degrees or so until it gets to the radiator, which is necessarily
larger than the intake opening. While the duct is expanding to the
size of the radiator, the air has lost some velocity but it's gained
pressure. We have selected a fin density compatible with the speed we
have slowed the air down to, so as to allow the air time to absorb
heat from the fins.

It flows through the radiator and removes heat from the fins. The air
is now heated much more than it was when it entered the duct. We've
sealed the space surrounding the radiator because air is lazy, it
doesn't want to go where you want it to go, it will seek easier
passages if you let it. If it can bypass the radiator, it will, thus
robbing you of it's ability to remove heat from the radiator.

Now the air is behind the radiator and it's greatly heated and
expanded. Why doesn't it want to blow forward? Because it's lazy,
remember? It wants to take the easiest path and that path is toward
the low pressure area we first stipulated was necessary (high pressure
at the inlet, low pressure at the outlet). But the duct is now
narrowing. This compresses the air and accelerates it. It cannot do
anything but head to the rear at ever increasing velocity, blowing out
the exit outlet parallel to the slipstream.

It's the exit parallel to the slipstream in it's accelerated state
that recovers the drag imposed while it entered the inlet and pushed
through the radiator face.

In effect, it's a low grade pulsejet engine with the radiator and the
rearward velocity of the air through the system acting as the vanes
that normally slam shut against the combustion of a real pulsejet,
forcing the products of the combustion out the tailpipe.

Does this clear things up?

Corky Scott

PS, the system described above is optimal. The reality is that the
confines of the engine compartment, protrusions and the usual
narrowness of the fuselage contribute to making almost all cooling
systems less than optimal. That's why I emphasized that in the real
world, the most important thing is that the system cool properly
first, and worry about the drag it produces later.

Excellent answer A, in cruise flight, there is little difference between a
propeller's slipstream and freestream velocity.

Today's props are very efficient (near 90% max) and create little
turbulence, with the exception of the tip's vortex . I have opened the
window on a c-150 and put my arm out to feel the propwash, and other than a
slight pulsing the only substantial disturbance is where my hand penetrates
the prop's tip vortex ring. Try it some time. This is not to say the flow
across a fuselage behind a prop is laminar, I'm nearly certain it isn't as
laminar boundary layers needs very little disturbance to become turbulent,
but total chaos does not exist within the slipstream behind a prop.

The stagnation point can also be defined as the point on the leading edge of
a surface/component that the velocity normal (perpendicular) to the skin is
0.

The Meredith effect, and ramjets in general, create stream flow by
maintaining a negative pressure gradient (high to low pressure) streamwise
within the duct/engine. This gradient is created by area differentials,
fixed or variable, between the inlet and exhaust. The inlet area is smaller
than the exhaust area, and thrust is a product of the duct's massflow times
the inlet's and exhaust's differential velocities. In the case of a cooling
system, if the internal aero losses are to high, due to friction across
radiators, etc., then the net thrust may be negative.

RJ



wrote in message ...
You'd be surprised how *little* a prop accellerates the air. The thrust
is a
product of how much air is accellerated and the change in velocity. A
prop
going 150 mph is working on a lot of cubic feet of air per second - a cyl
of air
the diameter of the prop and the distance the airplane goes in a second.
Changing the velocity of that amount of air just 10 mph gives oodles of
thrust.
The back prop on a 337 is seeing air that's moving backward at about 15
mph.

The slower the airplane is going, the more it has to accellerate the air
to get
the same thrust, because it's accellerating less air.

The stagnation point is the point on the leading edge of the wing where
the air
is splitting - some goes over the wing, some goes under the wing. Above
the
stagnation point the air goes over, below, under. AT the stagnation point
(which is very, very small) it piles up then falls off one way or the
other.


In article ,
says...

I'm not clear on what the stagnation point is. Maybe you can expand
you explanation more?

My line of thinking was basically this:
Airplane flies in clean air, propeller bites into clean air at high
sub sonic speeds, creates column of high velocity air, but at the same
time, wastes bunches of energy also churning it up. As air moves back
past fuselage and wing roots, churning energy disipates (decaying
exponentially with distance from prop). Now I'm sure that the
airframe itself introduces some turbulence of its own, but this is
distributed all over and is less concentrated at its source(s)

I think a lot of people are thinking about airplanes minus the
propeller or thinking of pusher configurations. This may be one of
the reasons why pusher power plants seem often have over-heating
issues. Taking the 337 for example: On the surface you'd say, the
back of the airplane must be traveling at the same speed as the front
of the aiplane so what's the deal? But going to the second order
effects, the front engine lives in an air flow with speeds much higher
than IAS and the turbulence of the cooling stream for the front engine
is MUCH higher than it is the intake scoop for the rear engine because
of the aformentioned exponential decay of the turbulence in the prop
wash.

On the Meredith Effect. Intuitively I don't see how its supposed to
work. You take air in the front of the scoop, you heat it up, and it
expands (ideal gas law). How does the air know thats only supposed to
go out the rear?

Regards

"RJ Cook" wrote in message
...
Jay,

The turbulence within a boundary layer over an aerodynamic surface
INCREASES
with distance DOWNSTREAM of the stagnation point. The cowl of an
aircraft
would have less turbulence within the boundary layer than nearly
anywhere
else on the fuselage.

RJ

Thanks for contributing to the fray...

I wasn't really talking about cruise so much as I was talking about
climb-out: high engine output, low vehicle speed, high AOA. Arm test-
If you can feel pulsing 5 feet back from the prop, then you know that
there is air of varying velocities washing over your arm. And since
you also know that the distubance decays very quickly at first and
then more gradually as time (and airplane) goes by, the disturbance is
much higher a one foot behind the propeller. I'm sure its not total
chaos, and it isn't laminar either, something between the two.
Question is, does the amount of turbulence produced, boost the heat
transfer capability of that air enough for a surface radiator on the
front/under surface of the cowl to work for a 100-200HP engine?

And yes, the first step is to make sure that you've got reliable
cooling. Okay, check, you've done that. Now what can be done to
eliminate one of the largest sources of drag? If this actually
worked, it would be a huge incentive for more water cooled (AKA auto
conversion) power plants in aircraft.



"RJ Cook" wrote in message ...
Excellent answer A, in cruise flight, there is little difference between a
propeller's slipstream and freestream velocity.

Today's props are very efficient (near 90% max) and create little
turbulence, with the exception of the tip's vortex . I have opened the
window on a c-150 and put my arm out to feel the propwash, and other than a
slight pulsing the only substantial disturbance is where my hand penetrates
the prop's tip vortex ring. Try it some time. This is not to say the flow
across a fuselage behind a prop is laminar, I'm nearly certain it isn't as
laminar boundary layers needs very little disturbance to become turbulent,
but total chaos does not exist within the slipstream behind a prop.

The stagnation point can also be defined as the point on the leading edge of
a surface/component that the velocity normal (perpendicular) to the skin is
0.

The Meredith effect, and ramjets in general, create stream flow by
maintaining a negative pressure gradient (high to low pressure) streamwise
within the duct/engine. This gradient is created by area differentials,
fixed or variable, between the inlet and exhaust. The inlet area is smaller
than the exhaust area, and thrust is a product of the duct's massflow times
the inlet's and exhaust's differential velocities. In the case of a cooling
system, if the internal aero losses are to high, due to friction across
radiators, etc., then the net thrust may be negative.

RJ