|
In a message dated 12/12/2007 9:05:59 A.M. Central Standard Time,
fredmoreno@optusnet.com.au writes:
Referring to Scott’s
set up using a pitot tube in the cowl inlet to pressurize the fuel injector
air inlets, Walter wrote:
“Very nice set-up. Did you
use a turbo-type injector to hook to the ram air? [Yes]”
This topic raises
some questions in my mind that just won’t go away. Pesky
things.
The fuel injectors
have holes in them so that air can flow from a high pressure region into the
injector to the lower pressure region in the intake port. Along the way
the flow of air helps to atomize the
fuel assisting combustion.
There are the bits
of “data” (I use the term loosely) rumbling around in my mind that suggest the
injector air flow is not important at wide open throttle and thus irrelevant
to high power cruise performance.
Fred, this is the only time it actually matters to me.
At anything other than WOT, the pressure in the manifold is, by definition and
observation, lower than that at WOT. The amount of flight (or taxi)
time spent at low power is inconsequential. At any time the throttle (air
valve controller) reduces the MAP below the pressure of the air used to feed the
injector bleed hole (whether ram or upper cowl cooling air), everything is fine
as the assist air is being sucked in.
This issue is important in a turbocharged engine because
a great deal of engine operational time is spent with the MAP at a higher
pressure than normally available air to be supplied to each injector -
especially at higher altitudes. Thus, "deck pressurized" air is tapped and
fed to the injectors to at least keep the bleed hole air at the same pressure as
the induction system.
But, that is not the only issue. I will again get to
that later.
The question is:
True or False?
Data bit #1: I
recall the following when talking to an auto engine “expert.” Modern
electronic fuel injection systems control fuel flow by varying pulse length,
but the fuel pressure remains constant. A jet of fuel is pointed at the
back of the intake valve. The intake valve which runs moderately hot
helps to vaporize the
fuel. Further vaporization occurs as the intake air is drawn by the
intake valve into the cylinder. The space between valve and valve seat
creates an annular venturi of sorts, and as we know, venturis are highly
effective at atomizing liquid fuel. So in auto engines higher fuel
pressure squirting on the intake valve plus a bit of heat plus the high shear
stresses that occur when the flow passes the valve and seat all combine to
provide good atomization over a wide range of power conditions.
It is time to talk to more modern engine experts (I am not one, but I
stayed at a Holiday Inn Express once). There are probably no single outlet
injectors in any of the efficient water cooled automobile engines extant
today. Electronic injectors do not need air to atomize the fuel.
Even my 1800cc Honda motorcycle uses 50 psi fuel delivery to electronic
injectors with 12 holes in each to better atomize the fuel in the
induction system (also utilizes 2 intake valves, 2 spark plugs and is
water cooled). Auto/motorcycle engines don't fly (much).
Data bit #2: I spent some years working in gas
turbine combustion to reduce NOx emissions. This led me to Arthur
Lefebvre’s book Gas Turbine Combustion which reports on the huge amounts
of work done on fuel atomization in jet engines. The problem is that
between cold day sea level take off power and flight idle at the start of
descent at 36,000 feet, there is a huge difference in fuel flow. This
creates major problems for atomization over the entire range of operation, and
has resulted in a lot of fancy fuel injector designs.
Yes, designs for kerosene powered
turbines.
Our simple constant
flow fuel injection systems are more like turbine injectors than modern
electronic auto injectors.
Data point #3:
Aircraft fuel injectors tend to squirt at the wall of the intake port (also
warm) and not at the back of the intake valve. So cold, low power
(idle) atomization maybe particularly poor. And low power which calls
for low air flow in the intake port also leads to poor atomization. Help
needed.
Again, so little tach time is spent at very low power, it is
not an issue. But the sucking (vacuum) produced by reducing the
throttle and the draw of the cylinder supplies adequate air for
atomization.
So methinks the
following:
In our constant flow
fuel injectors, at idle and low power settings, the fuel flow is so low that
the injector by itself does not squirt, but dribbles fuel out in a low
pressure stream. Poor atomization = lousy combustion. So to
improve atomization, we add air assist atomization. Fortunately, low
power also means low manifold pressure, and so ambient pressure (or upper deck
pressure) will force air through the fuel injector where it can atomize the
fuel dribble and – magic! – good (or at least improved) combustion. One
can expect at least 5 or 10 inches of mercury delta P between intake port and
ambient, good enough to spray gasoline.
Yes, fine.
At high power and in
cruise (in particular, wide open throttle at altitude) the manifold pressure
virtually equals the ambient pressure. No delta P means no atomization
air through the injector. But not to worry – the cylinder head is hot so
the fuel is hitting a hot surface inside the intake port, and moreover, the
MASS FLOW RATE of air through the intake port is high so that the fuel is well
sheared and mixed when air and fuel pass through the venturi between intake
valve and valve seat.
Not with a good ram air system.
Conclusion (Sorry
Scott, if I am right) – The extra effort to gain a few inches of WATER
pressure using the pitot ram tube set up to pressurize the injector air is not
worthwhile at higher power settings. At lower power settings, the delta
P IS large enough to provide atomization.
So for higher power
settings or wide open throttle, there is little to be gained by fancy air
distribution systems for air injection flow. More importantly for
Scott, the speed contribution may be zip.
Maybe. The air available to each injector is EQUAL in
pressure. There is an observed effect. After installation a
different cylinder is the first to reach peak when leaned. Other users
have reported a tighter CHT temperature spread. I have seen a tighter fuel
flow spread at peak EGTs. I do not have sophisticated equipment to
measure the difference in a meaningful way. I don't care. My
indications are that I have a better balanced system than I did before and
without running with different sized injectors. That is, I may have
accomplished some fine tuning at little cost. I hope that no one
else makes use of this technique - there are some untested areas, such as
what happens by blocking all of the injector air.
There are two things
that can modify this argument.
The first is that
the pressure in the intake port is not constant at the “manifold pressure” but
surging up and down as the cylinder goes gulp-gulp-gulp, and so there may well
be some delta P during the intake stroke when the cylinder is sucking on the
intake port. If this pressure fluctuation is not too fast for the flow
in the injector, it may assist in atomization. But the effect is due to
periodic low pressure in the intake port due to induction, not a tiny bit of
additional pressure from the pitot ram air set up.
The second thing
that can modify the argument is DATA! One sound data point is worth 1000
speculations. Has a good before and after test under carefully
controlled conditions shown performance improvements in terms of power,
mixture distribution, or beneficial effect on the harmonic convergence?
If so, Publish or Perish!
Yes, I am relying on my observations and the implementations
that all turbocharged engines use - I. E. the deck pressure (MAP) shall not
exceed the air pressure available at each injector.
What say you,
Walter? Teach us some more about fuel atomization at higher power settings
and wide open throttle. What does the data show? If no data, you are
invited to speculate as well. One Walter speculation = 1000 Fred
speculations.
Always Curious
Fred
>>>>>>>>>>>>>>>>
Scott
Krueger
MotorMouth, Charlatan, Bon Vivant, Escapee, Etcetera...
images[1].jpg
|