Date: Sun, 17 May 1998 04:04:17 -0700
From: Robert Harris 
Subject: FIREBALL - it lives.

Fireball - the implementation thereof - or the begats and how to steal it
without casting new heads.

The key is swirl.  Heywood (lots of dry facts - which Heisler's picture snap
into place - some thing to be said for synthesis of multiple references)
states that with a flat top piston and all the weird high swirl ports that
you reach a swirl ratio ( gas velocity after swirl is induces / piston
speed ) of between 2 and 3 to one.  However, change page, go to diesel.
There the classic bowl in piston with a diameter of .5 bore and centered,
can, with minimal clearance, reach a swirl ratio of 15 ONE FIVE to 1 or at
least 5 times better than the best of the induction impacting swirl stuff.
The taller the cylinder with respect to the same size, the faster the swirl,
however the "friction" of the rotating mass also goes up. Not particularly
sensitive to minor off centering - but ratios do fall.  Anyway, the fuel is
injected into this cyclone where it very rapidly burns.

On to a practical chamber.  Since we need a sparking plug, we can't put the
bowl in the piston.  The outstanding advantage of the bowl is the
centri-whatever of the fuel concentrating it on the outside.  We'd have to
spark into the lean center and that would suck. So, to get the wild swirl,
we simply invert the bowl.  We form the bowl in the chamber - but all the
bowl in piston rules apply.  This will be done by sculpting the piston crown
rather radically.

Next, since modifying the head other than minorly is out, the only plastic
we have is the piston shape.  Heads HAVE to have a "Closed Chamber" and
roughly in line valves or we be fucked.  Cleveland 2V's - out - 4V's in.
Now I be in shit.  The later 385 heads MAY be better than the earlier heads.
We'll work with that head for this thought.  We need to angle mill the
head - mainly to reduce the angle of the exhaust valve and to reduce chamber
volume somewhat as we are going for 15+ to 1.  Also, it would be good to
recess the valve upward if possible - read again VIZARD on Chevy heads - can
help scavenging and desirably reduce the diameter to height ratio.
Everything I have read indicates that the swirl is not that much until the
clearance volume is reduced to under .100 inch and then the action starts to
take place. The higher the compression, the faster the swirl -
overcompensating for the compression increase.  We can also shape the squish
to get a "squirt" to really spin it up. Thats what the Fireball did.  The
key factor is that the bowl contain most (at least 75% - preferably a lot
more) of the compressed volume.  If its spread out, we are back to a non
swirl high compression chamber and its problems.

Some bowl in piston phenom.  Swirl is slowest at the bottom of the bowl.
This is where the gases first enter. Since we inverted the bowl, this would
be the top of the chamber.  This area is less sensitive to friction losses
than the top of the bowl.  This implies the shape is less critical here.
Offsetting the bowl slows the swirl moderately - but apparently, with a
squirt like the fireball, some of this loss can be recovered.  Then, the
shape does not HAVE to be round. Izusu uses a square bowl with some nice
attributes in improved burn and mixing.  Velocity is meaningful.  The higher
the swirl the quicker we run up into the MACH number limiting factor,
specially since the gasoline engine has a much higher piston velocity than
the diesel.

So, using diesel rules, here's where we go.  Swirl bowl - approximately .5
piston diameter or less.  Vertical to piston.  "Left" Edge lined up with
exhaust side of chamber. "Right" Edge lined up with the sparking plug. These
define the approximate size and shape of the bowl - from a cylinder to a
squared cylinder.  Taller is better.  What we are after is the maximum
horizontal velocity practical.  Since the swirl is slowest at the top, the
valve angle and pocket makes less difference - doesn't have to be vertical -
the angle may also introduce tumble which will increase the burn rate.  What
we will do is make the bowl in the crown by milling out the exhaust valve
pocket flat. This is why the angle mill - to reduce the depth of the in
crown bowl by bringing the valve closer to vertical. Later 385 heads would
probably let me use a flat top in this bowl.  This flat top is probably
relatively critical.  If there is no recess for the exhaust valve, the
"squish" squirts in along the bottom and adds to the turbulent burn.

Now, we "fill" up the non bowl with a super high dome. Ideally, it would
totally fill the rest of the chamber - leaving just the "bowl" under the
exhaust valve - going metal to metal at TDC.  The clearance underneath for
the intake valve should be shrouded in the crown so that it "jets" out near
TDC (probably on the spark plug side) and the jet accelerates the swirl.
Backside should blend into the wedge with absolute minimum clearance.

So what we have is a monster dome piston that completely fills the intake
side and milled flat for a flat top on the exhaust side.  With some minor
head and further sculpting of the dome, we can probably get most of the bowl
in piston/fireball advantages without casting our own head. Lets say that we
achieve half of the practical 15 to 1 swirl ratio - which gives us 7 to 1 -
that's still up to three to one over the best of induction swirl - and its
taking place right when we need it most and not decaying like induction
induced swirl does near TDC.  This should be a verrra fast burn chamber

With this chamber, long rod short rod is out.  Longest rod possible -
period. Reason - reverse squish.  This chamber counts heavily on the last
moment violent inrush of gases.  Going past TDC just as quickly opens the
chamber to these squish - now quenching areas.  Holding TDC longer means
more complete burn of bowl before heating end gases - further reducing
knock.  Also great argument for pin offset - longer time after ignition to
hold gases in violently swirling bowl.

Best candidate heads - early closed chamber designs that were increased in
volume by simply increasing depth of chamber (like 385) rather than opening
outward. Makes the bowl deeper without having to mill out exhaust valve
pocket.

 The Usual Sources say that flat top or reverse dome pistons make more power
because of the
"better" flame propagation.