[Diy_efi] fuel injection for an airplane engine

Mon May 27 20:46:00 GMT 2002

The state of the art in Aviation is 70 years old.  The stuff we have
certainly works, but product liability certainly keeps R&D on the back
burner.

This has spawned a whole cottage industry around aircraft.  People like GAMI
have gone in and tested the flow characteristic of an engine, and then they
go get certified aircraft injectors off the shelf and test them as well.
They know they need to flow X amount in each cylinder, and they use an
injector that was tested to flow that amount.  They aren't using anything
that wasn't already approved, they just came up with the process and
obtained STC's for it.

The politics around aviationis strange.

We have magnetos ($450 each) that require inspection every 250 or so hours
of use.  Does that sound reliable?  Vacuum pumps ($300 that run gyros) that
require replacement every so many hours, and they usually fail without
warning.  I could go on, but we are really stuck with this antiquated
equipment.

So why do we have mixture knobs?  Good question.  When I trained, you lean
aggressively on the ground so that the plugs don't foul.  The mags make a
weak spark, and the 100 low lead fuel really builds up.  Low lead is a
misnomer.  It's got 4x the amount of lead that good old leaded car gas had.
So anyway, you lean on the ground, and if you push in the go knob, the
engine will quit because it's too lean.  Better that than to run too lean.
So everything else is full rich which helps cool the engine at high power
settings.  Once you reach cruise, you lean until the engine coughs, then
richen up two or three turns of the knob.  Real scientific.

These days the state of the art in mixture management is to watch the EGT
probes as you lean, and run anywhere from 50 to 100 degrees rich of peak.
Oh, there may be a fuel flow meter that is there to humor you as well.  It's
kind of a stigma to run lean of peak.  Mostly because the fuel distribution
is so uneven in the cylinders, so you play safe and run ROP.  If you have
the GAMI STC, then you can run LOP and people seem to be having luck with
that.

The problem with aviation is that there are so few general aviation
airplanes that fly piston engines.  Most of the data is anecdotal.  The car
guys get 1000's of a particular engine to statistically know where the weak
points are.  GM, Ford et. al run these things in dynos many, many hours on
end and know exactly where to place spark timing, fuel ratios, etc.  They've
optimised the intakes for even flow, and done lots of other things that
aviation engines don't have.

As far as computer control is concerned, one could go open loop.  a Lambda
sensor won't last very long in this fuel.  The Bendix injection system costs
around $3,000 if I recall, and runs at about 12PSI.   It spits out a
continuous stream of fuel maybe 1mm in diameter.  It relies on air swirl to
atomize the fuel.

Just rambling.

 Brian Michalk  <http://www.michalk.com>
Life is what you make of it ... never wish you had done something.
Aviator, experimental aircraft builder, motorcyclist, SCUBA diver
musician, home-brewer, entrepreneur and SINGLE!

> -----Original Message-----
> From: diy_efi-admin at diy-efi.org [mailto:diy_efi-admin at diy-efi.org]On
> Behalf Of William Shurvinton
> Sent: Monday, May 27, 2002 3:19 PM
> To: Diy_efi at diy-efi.org
> Subject: Re: [Diy_efi] fuel injection for an airplane engine
>
>
> I have been resisting on this topic, mainly due to a lack of
> knowledge about
> what is needed for a 'safe' aircraft, but here goes, feet first
>
> The issue is being able to do a thorough FMEA on the system and to gather
> requirements for what is needed.
>
> The FMEA bit usually drive redundant systems (heading into an area where I
> am getting dangerous, although my high availability analysis is limited to
> telecomms which is not considered life critical). For example, if
> you have 4
> cylinders then 2 ECUs can be used, each driving 2 cylinders, so
> if one fails
> you are still in the air. Dual pumps, dual sensors etc, are all possible
> depending on what is needed.
>
> Where is gets interesting is the requirements. Aircraft, if i have it
> correctly have about 3 operating modes
>
> 1. Idle on the ground
> 2. Take off and climb (max welly)
> 3. Cruise (max BSFC)
>
> All of which in a fairly narrow operating RPM range. A stock ECU
> is well OTT
> for this. The simpler the better. Where it gets hazy for me is the mixture
> knob which most aviation FI systems still seem to have. Is this because
> pilots don't trust ECUs or is it because more inputs are needed
> than the ECU
> is aware of to make these decisions. Possibly a bit of both.
>
> Anyway, if having a mixture control is considered a good thing it would
> point to a very simple ECU design, not dissimilar to the old Bosch
> LE-jetronic with minimum HW to handle the 3 operating modes. From what I
> know of the GM ECMs you could strip that down to a minimum map,
> but without
> exorcising all the redundant code it might be iffy. Ideally you want
> something based on one of the newer MCU with a TPU chip (or an
> external chip
> doing the same) such that once running steady state you could put a nail
> through the CPU and it would keep going until told otherwise.
>
> Interesting area.
>
> Bill
> ----- Original Message -----
> From: Bruce <nacelp at bright.net>
> To: <Diy_efi at diy-efi.org>
> Sent: Monday, May 27, 2002 7:37 PM
> Subject: Re: [Diy_efi] fuel injection for an airplane engine
>
>
> >
> > ----- Original Message -----
> > From: "Brian Michalk" <michalk at awpi.com>
> > > The SDS has the simpler of the two with batch firing, plus
> they are not
> > > afraid of aircraft.
> >
> > And, this is is from what?,
> > The litagation from the families of those that depart this
> world.   While
> > you might take responsibility, which is fine, too often the
> families take
> > the lose as more then the risk taker figures, and then sue the
> manufacturer.
> >   Designing a Aero EFI sounds interesting, using a car based system
> sounds,
> > like a problem waiting to happen in my book.   Two totally different
> design
> > specs..  You need more then a limp home mode for aero use in an
> aircraft,
> > IMO.
> >   What might be slick in a car, maybe fatal in the air.
> > Bruce
>
>
>
> _______________________________________________
> Diy_efi mailing list
> Diy_efi at diy-efi.org
> http://www.diy-efi.org/mailman/listinfo/diy_efi
>

_______________________________________________
Diy_efi mailing list
Diy_efi at diy-efi.org
http://www.diy-efi.org/mailman/listinfo/diy_efi