At 11:30 PM 3/20/95 -0500, TR4guyinVA@aol.com wrote:
>I found a local club member who had an old header off an old TR-3 race car.
> It also isn't a true header in that the four pipes aren't of equal length,
Ummmm.....
>There are two branches to it. They connect at a Y pipe somewhere under the
>starter. Then each branch splits again up under the intake manifold. One
>branch goes to cyls 1 and 4. The other goes to 2 and 3. I figured these
>cylinders fire exactly out of phase (didn't check, just guessed) and so
>probably don't compete much for exhaust space in the manifold/header.
Ah. What you mean is that the header you have isn't a true 4-into-1
header. It's a true header, it's just a 4-into-2-into-1 header. They are
different creatures, tuned for different purposes. In general, the 4:1
header works higher up in the RPM band, while the 4:2:1 works lower
down. The reason has to do with the velocity of the gas coming through
the pipe and the speed of sound.
When an exhaust pulse travels down a tube, it has a certain pressure
due to the length and diameter of the tube. When the pulse moves
out of that tube -- either into a larger tube, such as a header collector,
or into the atmosphere, such as at the end of the tailpipe -- there is a
wave of low-pressure gas generated that reflects up the exhaust
system at the speed of sound. (For the truly attentive, that's the speed
of sound at the pressure and temperature of the exhaust gas, which
may not be the speed of sound at sea level, except in certain parts of
southern California in August.) And for the EEs out there, this is
the fluid-dynamics equivalent of time-domain reflectometry (TDR).
It just works at mach 1 instead of c (both modulo the medium's
resistance characteristics).
The effect that this negative wave has on an engine is to reduce the
amount of work required to press the exhaust gases out of the
cylinder on the exhaust stroke. This is one way that headers
improve an engine's efficiency and power. If the negative wave
hits the exhaust valve just as it opens, the higher-pressure gases
inside the cylinder will be drawn out into the low-pressure area
created by this negative wave. (The other way is that by improving
the general flow characteristics, there is usually less pressure overall
in the system, making the system more efficient even when the
negative pulse is not in effect.)
The purpose of the 4:2:1 header is to put several openings in the
exhaust tube, placed at strategically measured intervals. For this
discussion, I'll call the tube from the head to the first weld the
primary, and the tube from the first weld to the next the secondary.
After the secondary is the collector, and after that the exhaust pipe.
A perfectly tuned 4:2:1 header is designed so that the lengths of the
primary tubes causes this low-pressure wave to reflect back up the
exhaust system in a fairly short time. There will in fact be a particular
RPM at which the negative pulse hits the exhaust valve just as it
opens. (No, I don't know the equation -- anyone want to help?)
Obviously, the shorter the length of the primaries, the lower in
the RPM band this negative pulse will be generated.
[MAJOR INSIGHT: This is another reason why Vizard's tip about
stepping the exhaust port-to-header dimension is so effective -- it
causes a small negative pulse at the port interface. WFC!]
In a 4:2:1 header, then, the place where the secondaries meet provides
another negative pulse, this one higher up the RPM band because it's
farther from the motor and it takes longer for the negative pulse to
hit the exhaust valve. This provides a secondary boost of power at
a slightly higher point in the RPM band. The net result is that a
4:2:1 header system typically widens the torque curve, giving the
impression of smoother power at more RPM ranges. It's a great
street header, and a great one for any performance that puts a
premium on flexibility and throttle response.
The downside to this design is that there's a certain amount of
confusion of exhaust pulses due to the fact that you've got primary
and secondary collectors. If the negative pulse hits the exhaust
valve when it's closed, it bounces back down the pipe; when the
waves interfere, you get a certain amount of diffusion that
doesn't really do anybody any good. And with waves coming
at two frequencies, you'll get a slightly messy pulse pattern inside
the tubes. Depending on how much care your car's initial
designers took with their header calculations, a 4:2:1 header may
provide better performance than your car's stock manifold and
downpipe. For the TR4, the stock setup is pretty good; the worst
thing appears to be the inequal lengths of the 1 & 2 and the 3 & 4
primaries, before they connect to the downpipe. But that appears
to be minor.
What about a 4:1 header with equal length primaries and one big
collector? This lets you optimize for one particular RPM range, and
usually one that's higher up the powerband because the tubes are
longer. You don't get the double-dip effect of the 4:2:1, but you
don't get the messy pulses, either. Race cars typically use 4:1
headers because they are meant to run best at a particular speed;
gearing, tires, and the driver are all optimized to get the car to that
engine speed and keep it there for as much of the race as possible.
In short, a well-designed 4:2:1 can put a modest negative pulse at
the exhaust valve over a fairly wide range of revolutions per
minute; the result is a modest to significant improvement in
the engine's throttle response because the negative pulses make
the engine do less work to expel the burned gases. A 4:1 header
can put a very strong negative pulse at the exhaust valve at one
specific window in the powerband. It's like the difference between
one of those reflective tanning mirrors you put under your face
to look like you just got back from Tahiti, and a parabolic mirror
that can light things on fire.
One thing I don't quite understand is the exact mechanism by
which the longer-lower/shorter-higher pipe length to RPM range
works. It's always seemed to me that when there's MORE time
between exhaust valve events, you'd want the wave to travel
SLOWER than when there is LESS time between events. There's
something else going on I'm not sure of. But this is the general
consensus of the tuning industry as I understand it; it works that
way, I'm just not sure how. Maybe I oughta ask Doc Tak, the only
person I know who plays with race cars, tunes pipe organs, and has a
Ph. D. in fluid dynamics...
--Scott "Be thankful I didn't start talking about sausages" Fisher
|