Guys, I'm glad we got a spirited discussion going on this issue because I
really feel that if we solve this, we'll eliminate the last objection to a
rotary aircraft engine. I say that because that's the question it always
gets down to "Yeah, but aren't you totally dependant on the engine computer
and electrical supply?", and the answer is "yes". I think for about a
thousand dollars, we could add a back-up self contained "system" that would
run the engine without external electrical power, or the engine computer.
I'd like to respond to some of the items that we're brought up in the
recent email discussion:
1. The advantage of modern EFI's are well known and they evolved from carbs
to mechanical FI, and finally EFI. The reason they went to EFI from
mechanical is precise "automatic" control over all ranges of engine
operation (temps (i.e. cold start), air pressure (high altitude), power vs.
economy cruise, emissions, equal fuel distribution, ignition timing
coordination, operating smoothness. etc.). They also realized better
economy (>10%) because they were able to get better atomization due to
higher fuel pressure. I understand a/c FI systems operate under 20 psi,
whereas modern EFI is 40 psi. The new high pressure (>60psi) electronic
injectors are currently being developed.
I think most everyone would agree that this is the EFI primary system is
what we want to operate the engine, but if this fails, we want a simple
device we could add to deliver metered fuel to the engine? EFI's generally
use a air valve (throttle body) with a throttle position sensor (tps) to
tell the computer what your foot is requesting (power vs. economy). Could
we use an (aircraft) carb as an air valve and add a tps onto the throttle
shaft? I don't think this would be a big engrg job. In operation you would
need to shut off the fuel to the carb or servo (using the mixture control)
while the EFI is operating the engine.
2. Since Mazda's apparently never came with mechanical fuel pumps, somehow
you would need this to supply fuel pressure to the a/c carb (< 5psi). The
only direct driven (non belt) devce is the oil injector metering pump.
Maybe you could adapt a small (fuel) pump to that (as Richard Sohn did)?
Otherwise, you'd probably need to have a belt driven fuel pump which I know
are available.
3. The final piece is the electromechanical ignition (aka a magneto) which
would complete the isolation of the critical engine functions. You would
want to power the leading plugs with the electronic ignition and the
trailing plugs with the magneto. If the EI quit, the engine would still
operate, although with less power, on the trailing plugs. I would think
that you could adapt a small magneto (2 plug for a 13B, or 3 for a 20B) to
the existing crank angle sensor (reluctor).
The critical advantage is that if somehow your electrical system (or engine
management system) fails, your engine will continue to operate, and you're
not a glider. This is probably not a true FADEC system, but it does meet
the intent, which is to have the advantages of optimized automatic engine
management, and yet in an emergency, to have an independent and reliable
back-up engine operating system. If someone had said to me that I could
install this kind of system on my 20B at a reasonable price, I would have
done it in a heartbeat.
I'd be interested to know how many other rotary builders would be
interested in having this type of system?
p.s. I will be doing the reverse of this on the a/c 540 engine. I will
normally operate the engine using an EFI/EI engine controller, similar to
the one I'm using on the 20B, but I will have the complete back-up of the
conventional a/c carb, magneto, and fuel pump. This is not a/c heresy, in
that most experimentals convert one mag to an electronic ignition (Jeff
Rose's system) and the engine runs much smoother and more efficiently. I'm
just taking it one more step, by adding EFI. In normal mode, only one lever
(throttle) will control all of the engine functions (power/economy &
richen/leaning), except in an emergency, when the back-up manual
electromechanical system will be employed.
Chuck Harbert
The AirCraft Rotary Engine NewsLetter. Powered by Linux.
http://home.earthlink.net/~rotaryeng/ http://www.linux.org