|
|
The analysis of ram air benefits needs to consider the entire system
from where the air enters the cowl up to the inlet to the carburetor or
fuel injection butterfly valve. Many induction systems are simply awful
creating major pressure losses while others provide superior recovery of
ram pressure and delivery to the engine. Replacing the former with the
latter will provide substantial benefits in the form of one to two
inches Hg improvements in manifold pressure.
Let's take the induction process one step at a time. First step is to
slow the oncoming air converting kinetic energy of velocity into the
potential energy of increased static pressure. If you mess up this
step, you can lose all ram air benefits before entering the induction
system. The slowing process occurs ideally in two stages. The first is
in front of the cowl as the air slow to go around the blunt leading
portions of the air frame. These lead to areas of local high pressure,
and this is where you want to take the induction air in from the main
stream. Under the spinner is usually a pretty good place. If there is
no readily accessible high pressure area (usually a place of negative or
concave curvature where the flow streamlines curve AWAY from the
centerline of the fuselage), then a scoop may be required. But keep in
mind that while a scoop may create a nice source of ram air, improperly
placed it can also create a lot of drag.
Once the air is slowed externally, it is brought through an opening in
the cowl and further slowed, virtually to zero velocity (tens of feet
per second, or less) before passing through an air filter. This second
stage deceleration is through a duct of increasing cross sectional area
called a diffuser, and diffusers are tricky to design well. Suffice it
to say that a smooth internal surface with no steps or abrupt changes in
cross sectional area or direction is required. The NACA scoop under the
spinner of the Lancair IV is a good example of a combined scoop and
diffuser and is the result of much testing. The key is to have
sufficient area to permit the air flow required by the engine to enter
the cowl opening at a low speed, say less than 20-25% of the free stream
velocity, perhaps a bit more with the NACA scoop. (Purists and
performance nuts should consult the technical papers on NACA scoops for
ideal velocity ratios.)
Cessna blew it here with the early pressurized 210. They made the NACA
inlet scoop too small, appropriate for low altitude, but inappropriate
for high altitude thin air. At high altitude, the volume of air
necessary is about 2.5 to 3 times the volume at sea level. The small
scoop forced the turbo to suck through the scoop like a kid sucking on a
soda straw in a milk shake. This resulted in a pressure at the inlet to
the turbocharger compressor well below ambient. (A small air cleaner
can do the same thing). The compressor then compresses the flow to the
required pressure for the engine, and the pressure ratio necessary is
quite high because of the relative vacuum at the inlet to the
compressor. High pressure ratio yields high temperature ratio, and the
induction air temperature to the engine skyrockets. Result: high
altitude detonation, and a lot of engines broke up as a result until
enlarged inlet scoops were retrofitted to the fleet. Intercoolers were
also added in many cases.
If you locate the induction air inlet properly and design the diffuser
properly, then you can probably get 75% of the total ram pressure, the
balance being lost to friction during the deceleration process. If you
really work at it you might get 80%+, and this is what the turboprop
guys strive for. But it takes a lot of work.
Now that the air has been slowed and pressurized, it is time to go
through the air filter, and then the duct work into the engine. If the
air filter is too small (and I believe most aircraft air filters are too
small, particularly for high altitude turbocharged engines), you lose
pressure due to friction. Pressure drop in air filters is due to
laminar flow in the porous element, and depends on the volume of the air
flow, independent of the air density. High performance racing filters
are a big help having been compromised in the direction of lower
pressure drop, not lower cost as with mass market auto air cleaners.
The other thing is to keep the filter dry if possible. Imagine the air
filter on a C-172, caked with dust, on the outside of the cowl, facing
forward when flying in cloud or rain. The dust turns to mud, clogs the
filter, and with more water the mud is drawn through the filter and
hopefully runs out the bottom of the ducting somewhere. Better to let a
little of the incoming air escape out the bottom of the duct or cowl in
front of the filter carrying rain water with it, and let the flow to the
air filter make some type of gentle curve sideways or upward to separate
the water. Again, the Lancair IV filter can do this with the ring air
filter if there is a place for the water to exit out the bottom of the
duct (normally not included).
Once through the filter, it becomes a question of good duct work design
to get to the engine and past carburetor heat boxes or alternate air
inlets that permit alternate air entry if the filter ices up. I am of
the opinion that this duct work should be sized for air velocities of
less than about 100 feet per second and preferably less than 70 feet per
second. (Keep in mind on aspirated engines that the inlet flow pulsates
a lot as each piston inhales a gulp of air. The pressure drop under
these conditions is higher than would occur with an average, steady
flow.) The ram pressure of a sea level density stream of air at 100
feet per second is about 0.1 inches of mercury, and it is easy to lose
two or three times this amount in a sharp ninety degree bend. Smooth
bends with generous radius and rounded corners are best, but difficult
to make in sheet metal. Bell mouth inlets to carburetors or fuel
injection air inlets also reduce inlet losses. (But think about fuel
fires on carbureted engines when building your induction ducting.)
Most aircraft installations are square, sharp cornered, and too small in
my opinion, and these cause losses in performance. You may argue that
if you plane has a turbo, no problem, just push the knob in more to make
up for the induction losses, but recall that the turbo must compress the
flow more and this raises induction temperature. The problem worsens in
the summer time when air density is lower and ambient temperatures are
higher. So both aspirated and super charged aircraft can benefit
substantially from good induction system design that strives for minimum
pressure drop.
To recap:
1) Put the induction air inlet location at a high pressure point on the
cowl. Under the spinner on in another location of concave curvature is
best. Where the flow is parallel to the fuselage centerline and a NACA
duct is used is second best. (See Turbo-210 or Turbo-182 NACA inlet
scoops for examples.) Regions where the flow curvature is convex (flow
curves toward the fuselage centerline) are worst because these are
regions of lower pressure and higher local velocity, and it is harder to
recover the ram pressure from a high velocity region. But you may have
no choice, and have to do the best you can.
2) Make the inlet generously sized so that the flow that enters the
throat of the inlet is no more than 20-25% of the free stream velocity.
This will reduce subsequent friction losses in the diffuser.
3) Be careful in diffuser design and construction to complete the flow
deceleration smoothly without major bends or discontinuities in cross
sectional area.
4) Have a big enough air filter, and try to let the water separate and
flow out of the diffuser duct before passing through the air filter.
5) After passing through the air filter, keep duct size generous with
smooth changes in cross sectional area and radiused bends without sharp
corners. This is hard to do in sheet metal, but very beneficial if you
are fuel injected and can use carefully fabricated curving, smooth
composite ducts in areas where there is no chance of a fuel fire.
Hope this helps. Ram pressure is free horsepower. Scoop it up.
Fred Moreno
|
|