X-Mozilla-Keys:
Fred wrote:
Looks like a good design to me. Compare to late model King Airs with the
"pitot" cowl which were designed and introduced to reduce drag and produce
more ram pressure at the turbine inlet to make the King Air faster. It
worked. Looks very similar. Below the spinner is also a good inlet location
since it is a high pressure region during climb at higher angles of attack.
I would say Steve has good model for others to copy.
As long as the flow is slowed adequately, the turn into a highly tipped
radiator is not that bad. It is commonly done in Formula 1 cars and Indy
cars. The loss of the turn is proportional to the local velocity squared.
So if the local velocity in front of the radiator is low (well "diffused" or
well slowed and uniform prior to radiator) the turn into the radiator comes
with a very low pressure loss penalty. Rule of thumb: for uniform flow
straight into a radiator, the inlet loss (due to inlet coefficient effects)
is generally less than 10% of the total radiator pressure drop which is
dominated by friction losses in the core. So if you get good pressure
recovery in the diffuser, and then lose a bit on the turn into the radiator,
you are still net ahead.
Better yet is to take advantage of the tipped radiator to make it a bit
bigger (larger frontal area) and a bit thinner since the pressure loss in the
radiator is proportional to thickness (if the area is held constant) and
proportional to one over the frontal area squared (if the thickness is held
constant). So bigger and thinner rapidly drops pressure drop. The problem
is packaging and diffusing the flow well prior to radiator inlet. If you are
hard core, you might consider two cores bent into a shallow V to fit into the
bottom curvature of the cowl better.
Same thing holds once the flow exits the radiator. It is going slow because
the radiator exit area is big so if the flow turns and twists (as it will)
the associated losses are not a major penalty. So as long as the flow areas
inside the cowl are large, the flow can find its way to the exit which can be
located anywhere it is convenient and appropriate. This is where you locate
your cowl flap to control the flow rate and maximize exit velocity to
recapture lost momentum. A cowl flap on the top of the cowl well forward of
the base of the windshield (a local high pressure region) would be ideal
because in climb it is a lower pressure area than below the cowl which would
improve the pressure drop across the radiator and improve cooling (but with
increased drag). But one needs to think about hot air blowing over the
windshield and possibility of oil and such on the windshield. Might be good
for deicing in winter, but uncomfortable in summer. Because of this cowl
flaps are normally out the side (see W.W.II radial engine fighters) or out
the bottom as in "modern" general aviation aircraft.
Incidentally, flowing the air from the radiator over the engine is good
practice to carry away exhaust manifold heat and keep surfaces comparatively
cool that would otherwise get sunburned in the glow of the red hot exhaust
pipes.
Steve has a good model to copy, particularly if the diffuser is smooth and
recovers pressure effectively. The way to check is to take another airspeed
indicator along, and put a line up to the face of the radiator and use this
pressure to drive the air speed indicator. Compare to the actual IAS. If
the airspeed shown in the cowl is 90% of the IAS, then the pressure recovery
is 81% of the ram pressure (point 9 squared), a very good value. Generally I
figure diffusers seldom do better than about 75%. It would be interesting to
try this on Steve's plane and see how well his inlet is working. Further
optimization might make the inlet smaller and pointier to reduce airframe
drag, but one runs the risk of compromising cooling at low speed and
increasing cooling drag. One has to work more and more to get less and less.
Fred Moreno
The Aircraft Rotary Engine Newsletter. Powered by Linux.
http://home.earthlink.net/~rotaryeng/ http://www.linux.org