air-assisted injectors, sort of

[Chris Conlon]

Greg wrote:

> The neatest thing about it is that I see no reason not to stab the
> Orbital air assisted injectors (the actual fuel injectors, which,
> in turn, plug into them,  are standard stuff) into the existing
> bungs on an existing TPI engine and go from there.
> ...
> Orbital claims maximum droplet size of 8 microns coming out of
> the injector. Obviously, if these units can direct inject, they
> are quick enough to match injection with inhale in a TPI system
> without resorting to staged injectors.

Presumably this means that these injectors have a minimum usable
"on" time a good bit lower than the usual 1-2 msec or so? (Or does
the engine have a fairly low ratio of power at WOT to power at idle?)

There are still several relationships and effects that I'm curious
about:

Does a decreased droplet size lower the "effective" octane of a fuel
(by increasing flame speed, leading to an increased chance of
detonation)? Or is it all vapor by TDC anyway, regardless of droplet
size?

To what extent is the fuel/air really homogeneous by TDC? On one hand
you have "stratified charge" engines, and OTOH your residual volume
is fairly small in an average-CR engine.

Ignoring the issue of detonation, how ignitable are lean mixtures? Or
rather, how lean can you make the mixture and still be able to ignite
it? What if you allow for a seriously powerful ignition, and/or bump
the CR up a lot? It seems a bit weird to me that lean mixtures are
more prone to detonation (which indicates to me a mixture near
optimum, or near one of perhaps 2 optimums) and also considered to
be hard to ignite - or is that only at a *much* leaner A/F ratio?

(I realize this is heading towards a diesel or GDI engine, I just
wonder if any of those efficiencies can be brought to a TPI setup.)

Also I'd really love to find some reference that maps flame speed
and the detonation line, flame temperature, charge density and A/F
ratio for gasoline over a wide range. (Not just near stoich.)  In
the past I was involved in making explosives, and all this kind of
data was easy to find, but this was mostly for solid and some liquid
explosives. I did get to make some air-fuel explosives, but not many,
and never really had nice complete data as for more conventional
systems.

My experience there left me with a few good rules of thumb; I'm sure
some of them apply here but doubt that they all do:

Fine grain size (in a mixture) increases flame speed. Grain shape
(in a composite) will have a profound effect on flame speed, and
often on an explosive's tendency to detonate. (We probably only
get one grain shape, spherical, which is regressive-burning, i.e.
the burn rate slows as it goes.)

Almost anything that burns really well can be made to detonate if
you have enough of it, under enough pressure, hot enough, and with
a strong enough initiation. Gas in an engine is obviously not too
very far from this limit under normal conditions.

There's a grey area (of pressure, density, temperature, flame speed)
within which an explosive might burn or it might detonate. Some
explosives easily go from burning to detonation, some do not. Pressure
helps make this transition. (Gas engines, again, are often in this
grey area.) Sometimes this grey area may be a very wide area indeed.
A very strong initiation (shock wave) can decisively ensure that the
explosive will detonate, not burn.

In a mixture, mixtures close to stoich are usually *in the neighborhood
of* the most-easily-detonated mixture.


Ok if you've suffered my post this far I'll give up my silly/brilliant
idea for people to poke holes in or run with:

Take the engine of your choice
Add the air-assisted injectors, or outright LPG injection. You're
  gonna need all the octane you can get
Raise the CR way way up
Have your chosen head-work Guru redo the head, bearing in mind
  the high CR.
Run the hottest ignition you can find or build (probably build)
Ceramic coat the piston/bore/head
Run the engine way lean
Run water/methanol injection to control detonation / EGT / NOx formation

I like the idea of squeezing a lot out of a given displacement, and out
of a given amount of fuel... what I don't know is if this is at all
workable. Maybe I'll have to wait for GDI after all.

   Comments appreciated,
   Chris C.