X-Mozilla-Keys:
Fred wrote:
Al:
Thanks for your questions. They are good ones.
1) Drag shown in the graphs is for the flow through the radiator only, and is
proportional to the net loss of momentum from this stream of air in going
from the free stream velocity to the outlet exit velocity. I assume the
outlet is pointed directly backward. Pointing it down is a waste of time and
speed.
2) How much drag due to inlet diffuser losses? I haven't broken it out in
relative amounts, although I could go back and calculate the distribution of
losses. I assumed for all cases that the overall diffuser performance is
75%, that is, 75% of the total ram pressure is obtained at the face of the
radiator. For a REALLY good diffuser, you might get 90%, and if you do an
awful job, maybe 50%.
Keep in mind that the compression of air ("diffusion" - it comes from the
19th century) occurs in two phases: a slowing (and compressing) in front of
the inlet due to the bluntness of the front of the airplane, and further
slowing after passing through the inlet into the diffuser itself. The first
phase is frictionless and is thus preferred, if you want to ignore all other
things. Just make the airplane real blunt in front and make a large inlet.
You will get good overall pressure recovery. The second phase has friction
(which causes boundary layers and flow separation and turbulence, etc.) which
reduces the pressure presented to the front of the radiator. If you build a
pointy airplane (and I hope you do -- they are faster) then you need to take
air in through a smaller inlet (to keep it pointy) and then do more of the
slowing in the diffuser. So faster airplanes (pointy airplanes) need better
diffusers since more pressure is recovered in the diffuser instead of in
front of the airplane. So.... you are likely to get good overall pressure
recovery in a blunt airplane without working hard, but you have to work hard
to get good pressure recovery in a pointy airplane with small inlets.
Taking all this into account, and placing my finger wetted and into the air,
I arrived at the figure of 75%. (In case you did not know, without testing,
diffusers are a large part guesswork.)
3) The inlet area is done by guess, (slow the air stream to 50% of the free
stream velocity at the inlet, based largely on the work of Miley and
Lopresti) but does not enter the analysis directly. Instead I use the
overall diffuser efficiency to compute pressure available at the front of the
radiator. The inlet area by itself is not enough to compute inlet
performance. It depends on shape, location, etc. If you are looking for a
real rough rule of thumb to size inlets based on the radiator area, I
discourage you from going so. It can be misleading. If you insist, I would
say make the inlet 20-25% of the radiator frontal area. Better to compute
the volume of air needed (which I did for each of these analyses) and then
select the inlet size that produces velocity that is 50% of the free stream
at the inlet. Unfortunately I did not provide enough information for to do
so. It varies depending on the effectiveness of the radiator (how well the
radiator heats the air and thus dumps heat). I can go back and come up with
some recommendations. I can also compute some radiator face velocities if
that would be helpful.
4) If we fly at constant indicated air speed, we fly at constant available
ram pressure as you stated. However, the pressure drop across a radiator to
maintain a constant mass flow rate (pounds per second) goes (roughly) as one
over the air density. So as the air density goes down (altitude goes up),
the pressure drop across the radiator goes up, but the ram pressure is
constant (fixed IAS) so cooling flow does down. And the engine gets hotter
unless you make a change, like open the cowl flaps. That is why turbo
airplanes cool just fine at low altitudes, and not so fine at high altitudes.
In an aspirated airplane, power drops with altitude so you produce less heat
as you climb where the cooling is worse. In a turbo, power is constant, so
the heat problem gets worse with altitude. I have a chart for a Lycoming
that shows for constant power output, the pressure drop required across the
engine increases from 6 inches of water column at sea level to 17 inches at
25,000 feet. You can see that the same IAS at sea level and 25,000 feet will
not produce the same cooling flow with such a dramatic increase in pressure
drop.
Why? What is happening? As the density goes down, the velocity through the
radiator (or over the cylinders) must go up to compensate and deliver the
same pounds per second of air. However, pressure drop goes as velocity
squared and as one over the density. Since velocity goes like one over
density for a perfect gas, we end up with pressure drop going like one over
density. Density down, pressure drop up. (all to first approximations).
Go high, make power, and you will run hot unless you are careful.
Hope this helps.
Fred
The Aircraft Rotary Engine Newsletter. Powered by Linux.
http://home.earthlink.net/~rotaryeng/ http://www.linux.org