"Blow, or Hardly Blown," etc.

[Fisher, Scott]

This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.
--
[ Picked text/plain from multipart/alternative ]
Jim Accordino writes:

> "bowls" directly above the valves.  On this list
> there's been discussion about NOT port matching on the
> exhaust side.  Something about pulse reversion.

Exactly.  I've done this (and the equivalent on the intake side, too, but in
the other direction of course) on naturally-aspirated engines, and it seems
to work.

The principle is to introduce a "step" between the port and the manifold.
When gas flows out of the smaller port into the larger manifold, it does so
easily, about as easily as if there were no step at all.

Reverse flow, however, is impeded, because the outermost portion of the
exhaust gas charge hits the "step" and curves inwards in turbulence, choking
off the flow of any high-pressure gases that are heading back towards the
head.

And they do.  When a "slug" of exhaust gas reaches an opening in the pipe --
at the header collector, for example, or at a resonator/muffler, or when it
finally exits at the tailpipe -- this causes a low-pressure pulse to travel
back UP the exhaust pipe, towards the engine, at the speed of sound.
High-performance exhaust systems are designed to take advantage of this,
because if the low-pressure pulse hits the exhaust valve at just the right
time, one or more good things can happen.  The easiest to visualize is that
if the cam has a fair amount of overlap -- meaning that there's a period
during compression at which the intake and exhaust valves are both open at
the same time, even only fractionally -- this negative pulse (a spike of low
pressure) can hit the exhaust valve and actually "suck" a little more intake
charge into the cylinder because there's a vacuum (relatively speaking) in
the exhaust port, which causes a vacuum in the cylinder head, which fills up
from the intake port.

One effect of the anti-reversionary step at the exhaust (and, on carbureted
cars, intake) manifolds is to let you run cams with more overlap and retain
low- and mid-range drivability.  Here's why: there's a high-pressure zone
just in front of the negative pulse that travels backwards up the exhaust.
Without the anti-reversion step, this high pressure results in contamination
of the combustion chamber with exhaust gases, because there's more pressure
in the exhaust port than in the cylinder head, so gas flows backwards into
the CC, and long valve overlap makes this worse.  Not good.  The
anti-reversionary step is one of those neat little tricks of fluid dynamics
that make science fun: when the high-pressure wave hits the step, it causes
turbulence and impedes the actual FLOW of gases backwards.  But when the
low-pressure (vacuum) wave hits the step, there's no real impedance (if I
may use a term more often used with electronics) because it's a vacuum
that's moving.

I will admit that I have no real understanding of how the spinning vanes of
a turbocharger affect the pulse reversion phenomenon, and I leave that
discussion to the turbo experts on the list, of which I am not one.  On a
naturally-aspirated car, however, the net effect of building steps into the
port/manifold interface is to reduce the "camminess" of an
aggressively-cammed engine, while retaining most of its high-end benefits.
Neat stuff, and easy to do with a Dremel in your garage -- just do it BEFORE
you send the head out to be pressure-washed and have new valves fitted.

Note that the same basic laws of physics apply to intake as well, but with
some differences in fuel-injected cars.  Aggressively-cammed cars with
carburetors can have an additional problem with lumpy idle because the
overlap can actually cause reversion up the intake, disturbing the "signal"
(the vacuum that the carb uses to determine how much fuel gets sucked into
the airstream) and causing the mixture to go off.  Since there's no such
"signal" in injected cars, that problem at least does not exist.  I suspect
it wouldn't HURT anything to grind a 1/16" step into the head/manifold
interface -- as long as you make the port larger than the manifold -- but I
also suspect the gains wouldn't be as noticeable as they are in, say, a
British sports car with siamesed intake ports and dual SU carburettors.  I
haven't tried this on any of my Audis (yet).

One other interesting point: when I was practicing this, it was legal in
SCCA Solo II, at least in Street Prepared, as none of the work needed to go
more than 1" from the manifold face.  You may be able to pick up some
additional benefit at the same time if you pay attention to the casting
inside the exhaust manifold; on the old British sports car I was doing this
to, there was a lot of casting flash in the #1 and #4 ports, and removing
that probably made as much difference as the stepped exhaust port.  I
haven't yet yanked the manifold on my Audi so I don't know how clean it is;
for the three or four of us on this list who also have Alfa Romeos (hi Tess!
You feeling better yet?), the insides of all the exhaust manifolds I've
looked at are as clean as you'd expect from a car with so many beautiful
castings, so there are no easy gains likely to be had there.

The best explanation of this whole subject that I've read is in "Tuning
BMC's A Series Engine" by David Vizard.  While it doesn't discuss our cars
at all, the book does an excellent job of explaining the principles behind
things like cam timing, the tradeoffs of velocity versus volume in port
design, and all the underlying science that makes hot-rodding work (or not).

--Scott Fisher
  Tualatin, Oregon

--