Velocity Stacks

[Scott Fisher <sefisher@domain.elided>]

In AD V7 #867, Kim Boydell asks, "Now, is it possible to tune velocity
stacks for different engine speeds?"

Yes, that's the key to velocity stacks.  Briefly: shorter stacks work
better at high RPM, while longer stacks work better at low RPM.  The
simple explanation for this lies with the inertia that the incoming
charge of air has; a longer column of air (moving down a longer stack)
has more mass -- therefore higher inertia -- than a shorter column of
air (moving down a shorter stack).

What happens next is sometimes called resonant supercharging.  Let's
take the long stack/low RPM operation example.  As the intake stroke
nears its end, piston speed drops and the intake valve begins to close
- -- two factors that reduce the available air/fuel mix that can be burned
two strokes later in the cycle.  

But if you increase the inertia of the incoming column of air, the
reduction in piston speed has less effect on the intake charge.  The
incoming air can still be moving at relatively high speed even after the
piston slows, effectively shoving itself through the closing intake
valve and getting in more air/fuel mix than would be possible
otherwise.  Taken to extremes, this can actually result in moving more
than (say) 1 liter of air into a 1-liter engine, at least at the optimum
RPM; said another way, it is possible for the volumetric efficiency of
an engine to exceed 100%, even without mechanical supercharging.  It
just isn't *easy*...

Kim also describes his [aside: I'm making the somewhat sexist assumption
that because Kim had a motorcycle, this particular Kim is a he, in spite
of the fact that my Delicious and Adorable Wife, also named Kim, is
decidedly NOT a he] old Moto Guzzi as having no filters at all, but
simply "a couple of velocity stacks with some screening on the ends
suitable for stopping small birds and big rocks."

These were quite the rage in the Fifties, Sixties and Seventies.  The
bad news is they do almost no good and quite a bit of bad -- that is,
they don't work either as filters or as low-restriction devices.  The
problem is that the mesh occludes a significant portion of the intake
area, reducing the amount of air that can flow past it.  (K&N filters,
to pick one example, also use mesh to hold the fiber elements in place,
but the deep pleating around the perimeter means that the available
surface area is so much higher that it evens out.  Or, if you lose 25%
of your surface area to the mesh, then simply make the surface area 200%
bigger and you've still got 50% more flow area available.)  And consider
that fine dust such as can be found at the apex of any corner on a road
circuit (kicked up when the guy in front of you drops a wheel in the
dirt) is, when mixed with hot motor oil, a very effective grinding
paste.  

Kim's final question is, "Or is it just the shape of the stack that
facilitates good breathing, i.e. the venturi effect?  Every velocity
stack I have seen is tulip or bell shaped."

That's also an important part of the equation.  A bare carburetor throat
- -- that is, one that forces the air to make a 90-degree bend -- actually
works like a restriction.  Air that enters a bare carb throat is pinched
just as though there was a smaller choke installed just inside the carb
throat.  By radiusing the entry to the carb (or injector tube), it's
possible to avoid this flow restriction and increase the potential
airflow by a measurable amount.  

Final comments on velocity stacks: a number of top-dollar racing efforts
have been using velocity stacks that adjust their length, in real time,
to adapt to the engine's current RPM and intake requirements.  Mazda's
prototype racers (such as those which won Le Mans several years back)
were the first cars I was aware of to use this technology; I remember
seeing video of the driver blipping the throttle, and the intake trumpet
would move back and forth as the RPM changed.  Apparently modern
Mercedes FIA-GT cars (or whatever they're called this month -- the ones
that keep crashing so spectacularly at the Sarthe) are using a similar
technology.  This is also in use in high-priced production cars; the
highest-performance (and priced) models from both Porsche and BMW vary
the length of the intake tract as a way of optimizing performance at
multiple RPM levels.  

And as a piece of -- I almost want to call it hubris -- consider the
McLaren Can-Am cars of the late 1960s.  I'd wondered about the uneven
ram tubes on the Hilborn injectors that topped the aluminum big-block
Chevrolet engines on those cars for years -- all other cars on track had
identical, tidy-looking matched ram tubes, while the McLarens looked
like a field of silver mushrooms, sprouting at all different heights and
angles from the intake valley of the V8.  Turns out that this was an
effort to improve drivability at all RPM levels -- rather than optimize
for one particular RPM, they made each piston work best at a different
engine speed.  So while the cars could have been tuned for a higher
output at one specific RPM, at the detriment of all other RPM, instead
they deliberately gave the cars lower peak output but improved power
throughout the rev range.  They were absolutely dominant, till Ferdinand
Piech got the Porsche 917 working right and dominated Can-Am competition
so thoroughly that the series died out.

For the best reference book on the subject of velocity stacks (and
virtually everything else related to the science of high-performance
engine design), check out David Vizard's "Tuning the A Series Engine"
some time.  Yes, it's about non-Alfa engines, but it has the best
description I've ever read of such issues as velocity stacks, cam
timing, exhaust system configuration and other high-performance issues. 
Every statement is backed up with graphs, measurements on a flow bench,
chassis dynamometer readings, and other hard data; for example, Vizard
tested several shapes of velocity stack and there's a picture of the
results, with the improvement (or reduction) in flow listed beside each
illustration. 

Regardless of whether you ever own a car with the BMC A Series engine,
this book is still one of the most valuable resources a high-performance
enthusiast could own.

And on the subject of Alfas and velocity stacks, I've been giving a lot
of thought to an item that AR Ricambi sells (or at least that's in my
very old Ricambi catalog) -- a set of velocity stacks designed for
Spica-equipped cars.  Now that I have one, I'm thinking about it -- but
this message is long enough, I'll write up my thoughts on the proper way
to get cold air AND have resonant supercharging in a separate message.

- --Scott

------------------------------