Quote:
Originally Posted by Dimman
Exhaust/header/turbo manifold questions?
There've been a few in another thread.
Sometimes those questions are hard, so if anyone wants to throw in their 2 cents or explain the following, have at it:
- Equal length vs unequal length headers
- 4-2-1 vs 4-1 headers
- venturi merge collectors
- inertial scavenging
- pulse scavenging
- pipe diameter (primary and secondary)
- pipe length (primary and secondary)
- stepped pipe sizes
- how cam timing plays a role
- other stuff I've forgotten/don't know about...
|
This is a series of questions with a single, long answer so I'll give a shot at doing the whole set at once. Bear in mind that I'm not an engine guy, just science-y and generally knowledgeable, so this isn't likely to be 100% complete or accurate.
Any old set of tubes will get the exhaust from your engine to the back of your car, but a properly designed exhaust system does it in a very specific way. The name of the game is scavenging; promoting the flow of gasses into and out of the cylinder, it's the foremost goal of any performance oriented exhaust system for a non-turbo car. There are two types of scavenging, inertial and wave.
Inertial Scavenging uses the kinetic energy of the escaping exhaust gasses to create a partial vacuum in the cylinder just before the intake valve opens. When the exhaust valve first opens, the high pressure in the cylinder and the low pressure in the primary cause the exhaust molecules to accelerate out into the exhaust system. When the pressure in the cylinder equalize with and then fall below the pressure in the exhaust primary (after TDC) the exhaust gasses begin to decelerate. But their accumulated kinetic energy means that they continue down the exhaust pipe increasing the gas pressure ahead of them and reducing the pressure behind them. This way, when the intake valve opens the exhaust gasses help pull fresh air/fuel mixture into the cylinder. This is what "overlap" is all about, and also why overlap sucks at low rpm: there isn't enough velocity in the exhaust gasses for there to be any inertial scavenging, and you just pull exhaust back into the cylinder instead, ruining VE. A similar effect occurs between cylinders as well. When the exhaust gasses flow past a merge collector they leave a region of low pressure behind them. When this is timed right, that low pressure will help pull exhaust gasses from the next set of exhaust valves to open.
Wave scavenging uses the resonant frequency of the gases in the primary and secondary exhaust tube to help pull fresh air/fuel mixture into the cylinder. As soon as the exhaust valves open, a pressure wave (sound) begins traveling down the primary ahead of the exhaust gasses. Where ever this wave encounters a change in cross-sectional area, a portion of it's energy will be reflected back up the exhaust tube in a reversion wave. The thing to understand about pressure waves is that they create they create a region of high pressure ahead of them, and a region of low pressure behind them. The trick is to have the reversion wave bounce off the exhaust valve just as the intake valve opens, further lowering the pressure in the cylinder. For this to happen, the time between the opening of the exhaust and intake valves needs to match up with the length of the primary/secondary and the speed of the wave. This means that wave scavenging only occurs at certain rpms, and will work against you at other rpms when the pressure wave arrives at exactly the wrong time. When the pressure wave bounces back from a merge collector, it travels up multiple primaries. In this way cylinders can help scavenge each other if the timing is right.
With
unequal length headers, half of the cylinders will have exhaust primaries and/or secondaries of a different length than the other half of the cylinders. The reuslt is that each set of cylinders will have different wave scavenging characteristics, reaching peak volumetric efficiency at a different rpm. This effect can be used to create a broader powerband at the expense of peak power. The problem being that whenever one set of cylinders is producing peak power, the other set is not. With
equal length headers every cylinder will create peak power at the same time, maximizing peak power at the expense of powerband width. It should be noted that "equal length" is usually defined as ±1". An obvious side effect of unequal length headers is the sound. Pressure pulses from the cylinders will leave the engine evenly spaced, but half take longer to reach the merge collector. The result is that the exhaust pulses leaving the muffler are not even, producing a lumpy, irregular exhaust note. This is the source of the "Boxer Rumble", not the engine configuration.
That same lumpiness is also seen on I4 and V8 engines with crossplane crankshafts. Because of the firing order, the exhaust pulses do not leave the engine evenly spaced, and so naturally do not leave the muffler evenly either. This also means that the scavenging effects between cylinders will occur at different rpms, creating a wider powerband.
180 degree headers correct this in V8s by connecting two cylinders from one bank with two cylinders from the other with equal length primaries/secondaries such that the exhaust pulses reaching each merge collector are evenly spaced. This means that inter-cylinder scavenging will create peak VE in each cylinder at the same rpm, increasing peak power at the expense of powerband width. The most famous example of this is the old Ford GT40 and its "bundle of snakes". The most noticeable side effect of this is that it completely eliminates the characteristic "V8 rumble" and produces a sound just like a flat crank Ferrari V8 (except deeper due to the larger displacement). A similar but lesser effect is achieved by connecting the two cylinder banks after the merge collectors, with an
X-pipe or
H-pipe etc.
4-2-1 vs 4-1 headers are usually described as, 4-1 is for peak power and 4-2-1 is for torque. This is the simplistic way of describing it. It comes down to scavenging. With a
4-1 header, the pressure wave is reflected from only one merge collector, so the wave that returns to the exhaust valve still has most of it's energy. It will only produce a wave scavenging effect at one rpm, but that effect is as strong as it can be. Further, scavenging between cylinders occurs evenly across all cylinders, producing more power at one (usually high) rpm by sacrificing powerband width. With
4-2-1 headers the pressure wave reverts twice, after the primary and after the secondary. The primary is the shorter of the two and creates wave scavenging at high rpm while the reversion from the secondary is tuned to scavenge at a low rpm. But the effectiveness of both is reduced as neither wave contains as much energy as the wave in a 4-1. Further, inter-cylinder scavenging peaks at different rpms for different cylinders. The result is a broader power band, with less potential for peak power. It should be noted however that a well designed 4-2-1 header will make more peak power than a poorly designed 4-1, and vice versa with low end torque.
A similar effect is achieved using
stepped pipe sizes, at each change in exhaust tube diameter, a small reversion wave is created. The length of each tube section can be tuned to improve VE at certain rpms.
A
venturi merge collector is shaped to prevent any increase in volume where exhaust primaries/secondaries meet in order to preserve exhaust gas velocity and by extension maximize inertial scavenging between cylinders.
A smaller
Primary diameter will improve inertial scavenging at low rpms and increase back pressure. Back pressure will improve volumetric efficiency at low rpms by counteracting charge loss due to overlap, but will create pumping losses and reduce inertial scavenging at high rpm. Larger primaries will improve peak power at the cost of low rpm torque (lower exhaust velocity). If you want to get precise about it, there's an equation based on primary length and displacement per cylinder.
Primary Length is calculated using the number of degrees between valve openings and the rpm at which you wish to make more power. There's an equation.
Secondary diameter is mathematically related to primary diameter, there's an equation. All the same science-y stuff applies.
Secondary length is calculated using the same equation as primary length, however you calculate the overall length including the primaries. The actual length of the secondaries is the overall length minus the length of the primaries.
Primary and secondary lengths are the means by which wave tuning is accomplished. They determine how long it will take the pressure waves to return to the cylinder. How long they
need to take is entirely dependent on your cam specs. The headers that work great with stock cams will not produce optimal results with a crazy 294 degree cam. Changes to the separation of valve events will change the rpms at which a set of headers is effective. This even includes adjustable cam gears.
There is a caveat though. All those fancy equations are approximations. The pressures waves traveling through the exhaust move at the speed of sound. The speed of sound depends on the density of the material it is traveling through, which in the case of a gas means its pressure and temperature. But neither the pressure nor the temperature of exhaust gasses are constant, or even necessarily similar from one vehicle to another. The equations will get you close, but they can't give you the perfect headers. In an ideal world, you'd produce several prototypes and dyno test them all back to back.