At 01:18 AM 3/21/01 +0000, you wrote:
> <SNIP>
> What I would like to know about is exhaust backpressure and just general
>exhaust design as it relates to horsepower and torque.
> <SNIP>
>I have several of the SA Design books, and in the "How to Build Horsepower"
>series by David Vizard, there is a section on exhaust design. Vizard
>specifically mentions how people say that an exhaust should have some
>backpressure and he goes on to explain in no uncertain terms that this is
>pure
>BS.
> <SNIP>
>Unfortunately, in the entire exhaust section, I don't
>see any references to torque, just HP. Perhaps free flow does increase HP
>but
>at the expense of low end torque?
> <SNIP>
> I guess the least backpressure would be had by open headers. The
> musclecar
>and hot rod crowd will always "uncork the headers" to get the best times
>at the
>strip. If this didn't work, I suspect they'd stop doing
>it. :-) However, as
>I mentioned before, we have personal experiences from folks on the list that
>show open headers on our trucks really drains the power. Is this merely
>because the system is no longer tuned as a whole, or is it really decreasing
>the power potential?
>
> Related to this would be the whole "exhaust tuning" area. Carefully
> measured
>full length headers, carefully placed H & X pipes, tri-y headers, etc.
>Obviously there are things you can do to an exhaust system beyond the basics
>such as single or duals and pipe size. These items obviously also have an
>effect on performance. Unfortunately, except for the few formulas available
>from the usual sources, this whole area appears to be a black art.
> -Jon-
HAZARDOUS TO YOUR HEALTH ... DO NOT ATTEMPT TO READ AT ONE SITTING!!!
Feel free to send your opinions. There are some assertions that I think can be
debated.
The Physics
It is valuable to keep in mind what is really going on in exhaust systems
in a modern
high-speed engine. Don't think of a continuous stream of exhausts gases
flowing
smoothly out the pipes; think of a complex fluttering, pulsing, resonating
system in
which the overall movement is hugely outbound, but where exhaust gases
start and
stop rapidly, bumping into each over, with pressure waves and zones
continuously
reflecting off each other and portions of the exhaust system.
If thought of in this way, it is easy to imagine why, by some miracle, at
certain speeds
for an unexpectedly good air flow when everything resonates together just
right whereas
at other speeds, the exhaust system is, to some degree, fighting
itself. Predicting the
complex, interacting behavior of an exhaust system is too complicated for
easy modeling.
Backpressure affects an engine's efficiency in two ways:
Higher backpressure provides greater resistance to the upward movement
of a position on its exhaust stroke. This, in turn, results in more
exhaust gases
being left in the space remaining in the combustion chamber when a
piston is at
TDC, leaving less room for the fresh intake charge entering subsequently on
the intake stroke (meaning there will be less oxygen and fuel to be burned
which translates to less power). Higher backpressure can also lead to
greater
combustion chamber heat (possibly increasing detonation, or glowing hot
spots
and pre-ignition).
Lower backpressure always improves horsepower and fuel economy.
Momentary or transient lower pressure can be optimized with the right
exhaust
system. The difference between negligible pressure at the exhaust ports
and the
5 or 10 or even 15psi of backpressure that is possible with some factory
or turbo
exhaust systems can be a lot of power. Typically, a 40% increase in
power can
double backpressure. Any time you measure a pressure over 4 or 5 psi in
backpressure, you should be thinking about modifications to the exhaust to
improve airflow for performance.
Low backpressure is essentially always good (even true on turbo
engines). So
why might an exhaust systems be better than open ports vented directly
to the
atmosphere? Because, in part, the blast of exhaust gases moving through
the exhausts system away from the exhaust ports at speeds of up to several
hundred miles per hour have momentum or inertia. Inertia is a law of
physics which
says that stopped objects tend to stay stopped, and moving objects tend
to keep
moving. Once a pulse of exhaust gases has been expelled from the
exhaust port
and is moving away from the exhaust valve at high speed, it will tend to
keep moving,
even when the valve closes behind it. Thus a transient low pressure
area is created
behind the gases in the exhaust port and exhaust header or manifold,
creating lower back
pressure for better scavenging the next time the exhaust valve
opens. Naturally, the
speed of the engine and its displacement and number of cylinders, and
the size and
length of the exhaust headers influences the speed and inertia of
exhaust pulses
(and the timing, as well), allowing the possibly of tuning the exhaust
system for very
high efficiencies at certain speeds. Changing the diameter and length of
exhaust
header pipes affects the rpm range at which tuned exhausts help to
scavenge the
clearance volume. Long narrow header pipes help low rpm operation by
producing
shock waves at the exhaust port. The tradeoff is that such pipes hurt
higher rpm
operation. In any case, if all the header tubes are all equal length
and diameter all
will have the optimal rpm range.
The second way that exhaust systems can make power is by utilizing the
transient low pressure zones of reflected shock waves moving backward
toward the
exhaust ports at the speed of sound as each pulse of exhaust exits the
system.
Since the velocity of exhaust gases changes with rpm, at certain speeds the
reflected low-pressure shock waves reach the exhaust ports exactly at
the right time
to help scavenge gases from the combustion chamber. If inertia and
pressure wave
effects at the exhaust ports coincide, this "tuning" effect is even more
pronounced. If
the exhaust tuning rpm range coincides with similar intake system tuning
and the
cam timing and lift is similarly optimized for this same rpm range, a
"sweet spot" of
tremendous power can be built-in to an engine's horsepower and torque
curves.
On all-out automotive race engines and high performance street motorcycles,
volumetric efficiencies (the amount of air actually drawn into a
cylinder compared to
cylinder filled with atmospheric pressure) of over 100percent can be
produced, for
tremendous power from very small engines. Of course, when you optimize
on rpm
range, you are not optimizing other rpm ranges which is why highly tuned
breathing
systems are good only on vehicles with very high power-to-weight
ratios. A superbike
has plenty of power for its weight at all rpms. Most street cars, on
the other hand,
with highly tuned breathing systems would have much higher peak
horsepower and
be overall slower due to the fact that you have to get through 3000rpm
to get to 8000,
and a broad power band and high average power is often preferable to a
high, "peaky"
one.
Effects of Exhaust Changes on Engine Management Systems
A free-flowing exhaust system with proper tuning is an unqualified
benefit except
when installation throws off the intake airflow mixture by confusing the
computer
engine management system or the physical metering systems of a carburetor
system. This is why the myth has arisen than an engine needs a little
backpressure
to run right. Backpressure is not per se good, but a change in behavior
of exhaust
gases resulting from a modified exhaust system does not stop at the exhaust
valves. If it did, there could not be power benefits from the
modification. Changes in
exhaust backpressure and resonant tuning have effects which move backward
through the combustion chamber and into the intake system, affecting the
intake
system resonant tuning and inertial ram-effects. Manifold pressure may
be impacted,
reversion behavior in carburetors may be impacted, and in Mass Airflow
Sensors in a
computerized EFI engine may be fooled by reversion changes. Reversion occurs
when the fluttering pulsing intake gases temporarily move the backwards
through the
intake system perhaps picking up fuel twice in a carb or impacting the
hot wire or film
which measure air mass on an injected engine, adversely impacting
injector open
time and air-fuel mixture in some vehicles. This would tend to be
corrected by the
oxygen sensor at idle or light cruise; it would not be corrected at wide open
throttle which is just when you'd expect the biggest benefits from a
free-flowing
exhaust system
Header Systems
All street vehicles require some sort of plumbing to gather the exhaust
gases from
the various exhaust ports and funnel it into mufflers and catalytic
converters. The
factory answer is almost always a cast iron exhaust manifold which is
fastened to
the exhaust ports via studs to collect the exhaust. In turn, it is bolted
to a single
output "head pipe" no more than a few inches from the cylinder head. The
manufacturer priorities include safety, reliability, emissions, costs,
and power-
probably in that order. Usually some attention is given to a streamline
flow, but this
maybe greatly compromised by space and cost consideration to the point where
exhaust impulses essentially enter a "log" manifold chamber which becomes
extremely turbulent and full of conflicting shock wave as the exhaust
impulses enter
the manifold dozens or hundreds of times per second.
Exhaust "headers", on the other hand, maintain streamline individual
runners from
each port for considerable distance until they are brought together in a
large
collector in such a way that the velocity vector of the gas pulses are
lined up and do
not conflict. The individual header pipes are welded to a flange that
bolts on the
head(s) in the same way as cast-iron manifolds (although different
fasteners, such as
bolts, with under-sized heads, are often required).
As mentioned earlier equal-length header pipes will optimize engine
airflow at
particular speeds. Equal-length headers tend to be more expensive, and
are generally
found as original equipment on expensive, high performance vehicles.
Equal-length
headers usually have more bends and take up more space. Available from the
aftermarket, they are more sometime complex to install, particularly on
"V"-type
engines sometimes requiring removal of the engine or at least motor mount
disconnection and engine jacking.
For reliability reasons, the car companies always use cast-iron
manifolds; in a few
cases manufactures have constructed large cast "headers" or "semi-headers" to
improve performance although these are rarely equal-length. "Shorty"
headers are
occasionally found as original equipment (for example on some Mustang
V-8s) and
are commonly available from the aftermarket. They are virtually never
equal-length.
Headers must be selected to work with the other design considerations
that affect
engine breathing (intake and exhaust runner sizes, cam selection, compression
ratio, etc. The benefits of headers depend completely on how good or bad
the stock
manifold was. A set of tube headers installed on a stock engine with no other
modifications will probably give you a little horsepower; a set of equal
length headers
installed on an engine with better heads and compression ratio, a hotter
cam, a
bigger throttle body, and a large mandrel bent exhaust systems with high flow
honeycomb cat and muffler will make a very large difference and will
allow the rest of
the performance modifications to really work.
Headers are usually constructed as 4-into-1units for 4-cylinder engines (and
two-times 4 into-1 for V8), or as two 3-into-one for V6 and straight six
engines. Ideally,
the header pipes feeding into each collector on a dual header system
would originate
with the correct cylinders such that the exhaust impulses into each
collector were
evenly spaced-at least as evenly as the engine itself. Evenly spaced may
not be
practical with the V-configuration engines although so-called
"180-degree" headers
have been constructed for V-engines in which selected ports header pipes
crossover
to join a collector on the opposing of the engine. Ford, using a
different approach,
once built a V-8 Indy car engine with the exhaust valves on the inboard
side of each
head with header pipes exiting the heads toward what is usually the
lifter-valley at
which point 180-degree header fabrication was easy. Most V-6 and V-8
headers are
" bank separated" units which are slightly less efficient than 180-degree
headers
("240-degree" on V-6).
Research shows most street engine are not sensitive to the shape of the
exhaust
gas passages after leaving the head ports. Nor are they sensitive to
header pipe
length within a fairly broad range (varying length between as little as
22" and as much
as 48" produces similar power curve) on high performance street machine
meaning
equal-length headers are not a great advantage unless everything else is
working
together as on a full-race engine. In fact "Tri-Y" headers take
advantage on this street
vehicle length and sensitivity by joining parts of headers pipes with a
Y-joint, then
joining those pipes together in a Y. Such 4-into-2-into-1 design are
cheap and work
fine on lower speed engines with milder cams (duration below
260-270). Although on
longer cams, they can make considerably less power. When you consider
the terrible
exhaust manifolds "headers" delivered as original equipment and many
vehicle Tri-Y
headers can be a good step up from a restrictive cast-iron manifold.
Street engines should be designed to optimize low and mid-range torque in
order to
be fun to drive, and headers should be selected with this in mind-meaning
long-length
headers with relatively small diameter tubing, designed to maintain high
gas velocity
through the engines of heavier vehicles at low and medium rpm.
Unfortunately, such
headers will offer restriction to high-end torque and power. Large
diameter header
tubes offer unrestricted high-end velocities, but low-end performance is
compromised.
Anti Reversion (A.R.) headers are designed to have the best of both
worlds. Cyclone
and Blackjack hold a trademark for "Anti-Reversion" equipment, which uses
fairly
large diameter primary header tubing for excellent top end power, but
also contains
smaller diameter direction-sensitive anti-reversion cones inside the main
tubing at the
exhaust ports, designed to act a little like a one-way valve to prevent
exhaust gases
from reentering the exhaust ports at low engine speeds. Such A.R.
headers are less
sensitive to diameter considerations, although diameter and length still
matter. It is
extremely important to install a cross-over pipe between the two halves
of any dual
exhaust system; although it is not clear why, without a crossover, the
low-end power
and benefits will be lost. A.R. headers may require a switch to smaller
jets (a size or
two) on carburetted vehicles due to improved booster venturi "signal"
strength.
A similar (but less effective) benefit to A.R. headers can be achieved by
sizing the
header flange and pipes to be larger than the exhaust port. The mismatch
will have a
minor detrimental effect on outward flow, but a much greater restriction
on reversion.
The overall effect is improved exhaust flow. The mismatch should only
occur near the
floor of the port, never the roof, because the flow velocity is greater
near the roof, and
a mismatch there could restrict flow more seriously in the outbound
direction. To
implement the mismatch correctly, it might be necessary to weld and
re-drill the
header mounting flanges, or install adapter plates with the correct
mismatch built-in.
If you cannot find headers for you engine/vehicle- particular engine-swap
situations or
where the engine layout is heavily customized, there are shops around the
country
with fabricators who can build custom headers. Custom headers are built
by bolting
header flanges to the block, and then piecing together the lengths of the
mandrel-bent
tubing, and the flanges can be custom fabricated by having a machine shop
build the
flange(s) from flat plate steel, usually by using a exhaust manifold
gasket as a
template. Custom-head fabricators use a band saw or tubing cutter and
grinder to cut
sections of strait and short or long radius tubing, which are then
tack-welded in place.
When the first tube has been constructed as far as the collector
location, a generic
fitting is tack welded in place. Now the fabricator constructs other
tubes, one by one,
from the exhaust ports to the collector. A fabricator with good
technique, good
spatial-visualization skills, and artistry can build a beautiful header
which fits perfectly
and has equal length runners.
A strategy between store-bought headers and custom-built headers is buying a
semi-custom kit of various "good-guess" header tubing pieces and bends of
a given
diameter, intended for a particular engine. A decent welder can build
good headers
this way faster than starting from scratch.
A custom turbo manifold header is built the same way as an ordinary
header, only the
individual header pipes are pieced together to a turbo flange instead of
a standard
header exhaust flange. The turbo itself is usually bolted to the turbo
flange and the
whole thing held in place with a jig or brace. Once the header is
constructed,
depending on design, the header itself may be sufficient to support the
turbo without
bracing. Turbo headers are sometimes constructed with tight radius weld
pipe bends
inserted of ordinary mild or stainless tubing, because the pipe bends are
much
thicker and more durable.
Headers are often built from mild steel tubing, and then may be finished
with a
protective coating from outfits like HPC or Jet-Hot. Even high-temp
paint will not last
on headers, and bare headers will rust. The best headers are built from
stainless
steel. Stainless a steel alloy blended with large amounts of hard chrome-
is much
more expensive than mild steel and the metal is very hard and relatively
difficult to
work with. However, it is very durable. Chrome-plated headers will not
rust on the
outside, but they turn yellow and blue spots at hot spots and will not
retain the
beautiful silver finish in these areas.
The major disadvantage of headers compared to cast-iron manifolds is their
susceptibly to cracking and breaking and warping such that they leak
exhaust or
require frequent new header gaskets. The extreme heat of exhaust headers
components has the unfortunate side affect that lock washers tend to lose
their
temper, and Loctite doesn't work, so you may find yourself frequently
tightening
header bolts. After a few tightening, depending on the level of
vibration and so forth.
the bolts may then stay tight, some of this can be the fault of the
header gasket
shrinking or eroding. If the header flange is machined absolutely flat,
dong away with
gaskets entirely and coating the header flanges with muffler seal paste
can be an
excellent solution to prevent leaks.
There are a number of aftermarket thermal wraps for containing the heat
in headers
and exhaust systems components. Keeping in heat in the exhaust gases rather
than having it dissipated through exhaust system into the atmosphere can be
beneficial where exhaust gas velocity is very important to effective
exhaust port
scavenging. Warming exhaust components is even more important where radiated
heat might raise under-hood temperature of intake air. Every 10-degree
increase in
intake temperature reduces power by one percent. Thermo-Tec and Swain
technology-and even J.C. Whitney sell an effective canvas like wrapping
material for
as little as $35 for a 2-inch by 50 foot roll (50 feet required for 4
cylinder headers, 100
feet for a 6-cylinder or small V-8). The cons on thermal wraps have been
addressed
many times by the DMLers.
Exhaust Pipes, Cats, and Mufflers
When it comes to exhaust systems, the car manufacturer's priorities are
not the
same as enthusiasts. Cost, even to the cent, legality, and decibels of
noise are
mighty weight consideration to the car manufacturers, which is why you
don't find
really great performing systems on factory cars, and why there is often
potential for
aftermarket parts to improve performance via the exhaust system.
Enthusiasts who
are willing to tolerate additional volume, and pay a little more, can
unlock power-
sometimes a lot.
Although the right header system can improve an engine's performance, all
the rest
of the exhaust system; the pipings, the cats, the mufflers, is a
necessary evil. The
exhaust piping is used to route header effluent to the cat and muffler,
and to get the
exhaust gases to the rear or side of the car for safe discharge where
they won't
poison anyone in the car from the CO2. For highest performance, piping
should be big
and short where possible and mufflers should be minimal to meet noise
standards
(more about mufflers in a moment). Turbo-equipped cars are especially
lucky, since
the turbo itself acts as a silencer of sorts, leaving little or no sound
deadening in the
muffler). Catalysts should be large-diameter and free flowing (honeycomb
design, with
sufficient CFM airflow). It goes without saying that all V8s and V6's
should have
dual exhaust and dual cats if at all possible. (In-line sixes should too.)
High-performance exhaust systems should always be fabricated from large
diameter
mandrel-bend tubing. Various diameter mandrel-bends are obtainable from
header
manufacturers, muffler shops, turbocharging supply outfits, large-truck
parts houses
(in huge sizes), and speed shops and racing supply stores- and even shops
that build
race-car frames. There is no such thing as too large diameter, unless ground
clearance or other problems force additional bends that defeat the
purpose of the
large diameter. Two-and-a-quarter inches should be considered minimum,
and 2.5
inches is better. Each 90-degree mandrel bend has a minor detrimental
effect that
can add up. Each 90-degree crimp-bend is much worse. The narrowest crimp
becomes the bottleneck, acting as a venturi, potentially causing
turbulence and swirl
and disturbing laminar flow in the pipe.
Catalyst require large amounts of heat to work, and also remove a lot of
heat to work,
and also remove a lot of heat from exhaust gases, which is why they should be
located as close as possible to the ports, and why dual cats may not
receive enough
heat to work properly. Three-way catalytic converters are extremely
effective at
removing HC, CO, and Nox pollutants as long as the air fuel mixture is near
stoichiometric (14.7:1) (they are so effective that very few secondary
emissions
devices are required on modern computer-controlled EFI cars). The platinum,
palladium, and rhodium noble metals in the cat combine with pollutants,
change the
nature of these gases, and are then re-deposited onto the converter,
releasing
harmless non-polluting nitrogen, carbon dioxide, and water into the
atmosphere. Cats
do not even begin to work until the platinum surfaces are 400-500
degrees, and peak
efficiency occurs at 900-1600degrees.
Mufflers use reflecting, absorbing, and restrictors to deaden the sound
waves that
come from the exhaust ports with the exhaust gases. Most factory-original
mufflers
use restriction, forcing the gases through small passages, to silence
them, which
produces both quiet exhaust and high backpressure and loss of
power. Glasspack
mufflers from the aftermarket generally send exhaust gases through a
perforated pipe
inside a chamber filled with fiberglass shavings. You should use only the
type in
which the perforation is punched outward into the fiberglass in order to
avoid installing
glasspacks even more restrictive than original equipment mufflers. Glass
packs can
be loud. Flowmaster-type mufflers are fairly large, but they are quiet and
low-restriction, using both reflective and absorption to quiet exhaust, while
maintaining large-diameter internal passages for low restriction.
Every exhaust system design (even of fuel dragsters) is a compromise.
Anonymous
This archive was generated by hypermail 2b29 : Fri Jun 20 2003 - 12:00:19 EDT