Herb Sanders wrote:

Here is a real world example of scoop design. We bought a Long EZ that

had the oil

cooler mounted behind the main spar in a box fed by a scoop constructed

by cutting

the cowl back, pulling it down and glassing the sides, and leaving a 2" x

4" scoop at

the cowl surface feeding into the box.  It did not cool well. It would go

to red line

in flight at wide open throttle and have to be step climbed - still ran

at near

redline, even at reduced throttle.

We took a scoop available from Velocity that hangs down from the surface

about 1",

has a nice teardrop shape, and about 3" dia round opening (7 sq in

compared to 8 sq

in on the scoop we were replacing). It came with sharp edges, but I took

a 5/16 dia

nylon tube, split it length ways, filled it with 5 min epoxy, and forced

it on to the

lip. After cure, we filled in behind the tubing with micro inside and

outside, and

faired it into the sides. Now, in 100 deg Memphis weather and at gross

wt, we can

climb WOT to 10,500' and not see oil temp over 195 deg F.

>From what I have read, a perfect NACA flush inlet is about 85% efficient

compared to

a free air ram inlet at pressure recovery. On my Velocity project, I have

built a

plenum over the top of engine fed by a 9" x 3" free air ram scoop 1"

above the fus

top. My plan is to feed combustion air, oil cooler air, and engine

cooling air all

from the plenum chamber. Any redneck who likes the look of aCamaro with a

hood scoop

would love it. Herb Sanders

This correlates with my opinion that the boundary layer in this area will be
about
3/4" thick or maybe more.  The pressure is realtively low here, as I dump my
oil cooler exit air out into the boundary here, which makes it a lot lower
than the
pressurized side of my cowl.   If you pick up this low pressure, low energy air
(like the original scoop in Herb's bird was doing) you won't have very much
massflow
and also very little dynamic pressure, since the velocity is low in the
boudnary.
You need to go out about 3/4 ore more to pick up clean air which is flowing at
the freestream velocity.  Slowing this air down smoothly in a good diffuser
gets you a nice flow of lower velocity, high pressure air - perfect for a
radiator
or oil cooler.

The fact that the scoop extends beyond the boundary layer will usually keep the
boundary (bad) air out of the scoop.  If you want to make sure the boundary air
stays out you can use a splitter, as recommended by Jeff.  Basically make up a
thin plate parallel to the airflow (parallel to the surface, too) which is
spaced off
of the surface more than the expected boundary layer thickness and extends
upstream somewhat ahead of the inlet to keep the boundary air from getting into
the inlet.

Hope this helps.  Modestly rounded lips (like Herb is describing) help the inlet
work better if slightly misaligned to the local flow direction, which is
difficult to
be absolutely sure of.  Extremely thick rounded lips will start providing more
overall frontal area for the scoop than the inlet area, which leads to external
diffusion, probably not a good thing.

For example, if the duct inner dimensions (air flow duct) are 3" x 4", and the
edges are very thin, the total frontal area might be only 3.1" x 4.1".  This is
6% "excess frontal area".  The risk is that if the airflow is misaligned to
the duct
inlet (or vice versa depending on your viewpoint) you can get separation at
the inlet
and a loss of flow.  With 3/8" diam. rounded lips the duct frontal area has
to go up
to 3.75"  x 4.75" if the inlet stays the same.  This duct will work about
the same as
the thin edged inlet if the flows are well aligned, but it will work much
better if the
flow is misaligned.  The price is the additional 48% "excess" frontal area
of the duct.

If the lip shape is more like a straight hole on the inside with a nice
rounded outer contour and a small radius inner nose (see attached image)
the duct will have substantial external diffusion.  The problem with a
protruding duct using external diffusion is that it greatly increases the
"excess" frontal area, so internal diffusion with a thinwalled duct will
provide lower frontal area, and lower drag.  However, where the frontal area
is  preexisting (wing or fuselage LE, for example) an external diffusion
inlet is a good choice.  If the duct wall thickness for the example
goes up to 1", the new frontal area is based on 4" x 5", which is 66%
more frontal area than the original duct, and a bunch more drag.

So, thin duct walls produce minimum "excess" frontal area, but you
make the duct alignment more critical as the lip radius decreases.
Also, don't ingest the boundary layer if you plan on significant
recovery of dynamic pressure.

Bill

 
The Aircraft Rotary Engine Newsletter.              Powered by Linux.

http://home.earthlink.net/~rotaryeng/            http://www.linux.org