Paul,
I've been looking through the Kinsler catalog and they have
everything
you need to put together a dead simple, reliable mechanical fuel
injection system for not much money -- maybe less than an Ellison
carb.
Constant flow fuel injection is used on aircraft piston engines and
works just like it sounds -- there is a constant spray of fuel coming
out of the injector nozzles, unlike the timed EFI and some mechanical
systems for cars.
Constant flow is not great for cars, but perfect for airplanes,
because
you are running at steady-state rpm and power most of the time.
Have a look at the attached Kinsler Fuel Flow Diagram. The system
works
like this: a mechanical fuel pump driven off the engine supplies more
fuel as rpm increases. It sends the fuel to a barrel valve that
distributes the fuel to the individual nozzles at the intake ports.
The barrel valve is kind of like a spider in aircraft FI systems,
but it
is directly linked to the throttle, so as the throttle is opened, the
barrel valve is opened as well, allowing more fuel to flow to the
ports.
Fuel metering is accomplished upstream of the barrel valve by
putting a
restrictor jet in a return line that goes back to the tank. So
the fuel
coming from the pump has two places to go, the barrel valve that
leads
to the engine nozzles -- and the restrictor jet. The more fuel that
flows through the restrictor jet, the less fuel goes to the engine.
Leaning is accomplished by varying the size of the restrictor jet. A
bigger jet means a leaner mixture because it allows more fuel to
circle
back to the fuel tank and less to the nozzles.
Kinsler has a cockpit adjustable mixture control that lets you dial
different jet sizes, as shown in the attached pictures. However, the
same thing could be accomplished with a simple ball valve, like
the Home
Depot valve that Paul is implementing. It would be controlled by a
Vernier control cable. This way no fuel has to go through the
cockpit.
The barrel valve is available with a minimum of four ports, so
you can
either put two nozzles on each P-port runner, one on each four-runner
manifold, or block off two of the ports and use one nozzle on each
P-port runner. I would use two nozzles per runner to get better
atomization. Also the fuel flow from two nozzles should be about
right
for a single rotor. Same thing with the fuel flow of the 4-port
barrel
valve.
You could use Paul's slide throttle with this system or you could use
Kinsler throttle bodies which already have the bosses for
attaching the
nozzles. If you roll your own you have to weld up some nozzle bosses
onto the runners. Btw, the Kinsler bosses can accept both the
constant-flow and EFI injectors.
This system can be used with a turbocharger or supercharger
because you
can run it with high pressure. You could get significant evaporative
cooling effect by placing the nozzles close to the engine and
spraying
the fuel right at the rotor. I would guess that you could
probably run
10 psi of boost without having to go to an intercooler.
In a normally aspirated engine you can run it at lower pressure.
It also
doesn't need heat, since icing is not possible. It provides all the
advantages of fuel injection in terms of precise metering and mixture
distribution.
This should make for a very dependable, completely mechanical system
that is not dependent on electrical, control boxes or computers.
If you
implement the flywheel magneto as well, you will have a real aircraft
engine.
You can download the Kinsler tech catalog here:
http://www.kinsler.com/catalog.htm
Regards,
Gordon.
Gordon
Very interesting. Good going.
One minor point though.
Please be advised that a ball valve is good as a shut off valve but
makes a
terrible leaning valve. For example the flow curve is like a J
curve. It's
design merit is in the low flow resistance when all the way open and the
ability to shut the flow off quickly. A good use would be the
emergency shut
off valve. For metering you need a metering valve or commonly called a
needle valve. The needle come down perpendicular to the flow and into a
tapered valve seat. Excellent fine adjustment for gas or liquid flows.
I have lots of experience using both of these valves with corrosive
gas and
liquid flows.
Doug Fir
Have you used them with a vernier leaning cable?
No problem. Just get a control cable you can twist
to adjust the needle valve. Do you have any
brands you can recommend or do we have to make one?
Paul Lamar ...No rotor no motor.
Doug,
Thanks for the scoop on the ball valve. Right after I posted this I
started thinking about the ball valve and began to suspect that it
might not be very good for leaning. A great shut off valve though.
A needle valve sounds like just the ticket -- hopefully there are
off-the-shelf items that will fit the bill. If not, you can always
use the Kinsler jet selector. Not as precise because it is stepped
rather than infitely variable, but still will work.
Regards,
Gordon.
I think the ball valve is not any more nonlinear than the round hole
in the slide throttle. One may get the impression it is hard
to adjust because the lever only moves 90 degree from full
on to full off while a threaded needle valve takes many turns
from full on to full off. The vernier push pull leaning cable to the
rescue. What the vernier does is allow super fine adjustments
by turning the knob. Much like a needle valve. I know some
of you have flown airplanes without vernier mixture knobs. Once
you have used one you will understand how well they work to
adjust the mixture to a fair the well.
IMHO the answer is lengthen the lever arm on the ball valve
until it travels 3 inch's on the end for a 90 degree
angular change. A lever about 2 inches long should do the
job.
However I am still open to using a needle valve if
somebody can come up with a practical design supplying a
3D drawing of same. A 2D sketch will suffice. If it looks
good and looks like it will work I will do a 3D.
------------------------------------------
OK I had an idea. We can make this out of aluminum tubing.
Three inch long needle. The valve seat is pressed in from
one end and the intake and outlet fittings are welded
or bonded on with JB Weld. Make two. One for each rotor.
The only draw back is the leaning knob works backwards.
push for shut off. Pull to richen and push to lean.
Paul Lamar ...No rotor no motor.
Paul,
Sorry to be late in replying. The boss had me raking leaves all day. Trust
me on the ball valve. You will be surprised on how unstable it is as a
metering valve. The reason is the opening cross sectional area is very non
linear. In use it seems like the flow goes from 10% to 80% in a few degrees
of rotation. Needle valves should be available in any valve catalog.
Yes vernier cable has the fine thread adjustment on the instrument panel
mount end I believe. C-152 had them but my C-150 has a Micky Mouse spring
clip. You pull and pull to overcome the resistance and suddenly you have
pulled too far. Makes for real attention getting when the engine starts
sputtering high over rugged northern Arizona terrain:)
OK of course pull to lean. I have an idea on how to make it. Give me about
6 hours and then I'll work on it tonight on Rhino. Lets see two needles one
inlet two outlets. The same principle will probably work for a bypass
mechanism for the constant spray injector. Now its 15:30 and still daylight
so the boss has my next assignment waiting for me. Probably splitting wood
for the ofuro, deep bath.
Doug
Paul,
Its now midnight in Japan and I got the dual needle valve and block
designed on Rhino. What format can I send it to you in. I did the
different parts in various colors to make it easier to understand.
When I render though all colors turn to gray. If I try to save a
perspective view I can't find a JPEG option.
Going to kitchen now for milk and cookies and awaiting your advise:)
Doug
Render it and save it as a jpg and send that.
Sleep during the day and work at night. That way the boss
won't bug you as much.
Paul Lamar ...No rotor no motor.
The Rotary Engine NewsLetter. Powered by Linux.
ACRE NL web site.
http://www.rotaryeng.net
Copyright 1998-2006 All world wide rights reserved.