Exhaust System Thoughts & Theory

[adamjp]

Guys,

First some definitions:
NA Naturally Aspirated
FI Forced Induction

An engine is an air pump. Air goes in, mixes with fuel, provides the oxygen to sustain an explosion, and then leaves. The key thing is that air and exhaust are gases. The measure of air in against capacity is referred to as volumetric efficiency (VE). It is possible to increase the size of the intake charge beyond that of the empty cylinder - a very good thing. Engines with this capability have a VE exceeding 1, meaning they get the advantages of forced induction without the disadvantages. An S2000 engine has a VE of 1.12 (112%). B6/BP in a NA MX5 probably works at a VE of around 0.8 (80%). Most performance modifications to an engine are about getting the biggest packet of air into and out of the engine as quickly as possible in a given operating range (Maximising VE).

For any given diameter of pipe the gas will flow at a given speed. If the pressure is constant, yet the diameter is reduced, the speed will increase. Increase the diameter, and the speed will reduce. Unlike liquids, gas WILL locally compact (increase pressure), causing pulsation in the system. It is this pulsation that we try to use to ‘tune’ the engine by using the high and low pressure components of the pulse. We want a high pressure pulse to hit the inlet valve ‘just’ as it opens. We want a low pressure pulse to be present at the exhaust valve ‘just’ as it opens. The aim is to get more air into an engine, and then out.

If you have extractors (tuned length headers) then you are attempting to 'pull' the exhausted gasses out of the combustion chamber with pulse tuning (harmonics). These match with the tuned length intake manifold to set up a resonance within the combustion chamber that has the air charges moving in well ordered packets (at least that is the theory) to maximise VE.

People often equate low backpressure in exhaust systems (due to large pipes) with loss of power. What they are actually observing is a loss of gas velocity at low RPM due to large pipes. Without the velocity, you get the pressure pulses stalling and diluting with a consequent loss of beneficial harmonics, and subsequent reduction in VE. It starts to come back at higher RPM as the engine passes more gas. The optimum outcome of all of this is that the gas comes out of the exhaust port travelling really fast, moves into the exhaust system and tries to keep the speed up as it cools, contracts and meets other exhaust gases from other pistons. Your engine can only move so much gas, and the trick is to tune the exhaust system to keep gas velocity UP in the operating band you desire. You can go too big, and some will tell you ‘you must have some backpressure’.

The concept of backpressure on an NA car was originally to keep some of the exhaust gases lingering around the exhaust port to stop the incoming charge sweeping straight through the combustion chamber. This will happen with cams that have overlap. The lack of backpressure would substantially increase the fuel costs (as the fuel is not being burned in the combustion chamber, but in the exhaust system and is therefore wasted) and will raise the temperature of the exhaust system (leading to early failure). Afterburning is not generally applicable to piston engines.

[adamjp]

Performance NA camshafts usually increase overlap (inlet and exhaust open at the same time) through longer valve opening times (duration) to produce increased filling and consequent scavenging of exhaust gasses. An S2000 achieves that 112% VE through V-Tec. I won't go into valve angles, but this also has an effect. In the old and bold days of pushrod V8s and holley carbies, lumpy idle (mucho overlap cam) and single plane intake manifolds developed good power high in the range, but were only saved down low by the sheer displacement of the engine generating torque. Keeping backpressure on common cars (not the performance GT types) allowed low range operation without using prodigious amounts of fuel. It impacted high range operation too by reducing flow when it really was needed (but the ol’ 351ci station wagon still went well enough).

Keep one thing in mind: Backpressure is the enemy of VE by reducing the amount of waste air that can be removed from an engine.

Application to Forced Induction.

The common thing that you will hear from the turbo guys is

'I put this sik dump pipe on, and now the turbo spools up quicker.'

They are kinda right, but first a little more theory.

The theory of equilibrium when applied to gas under pressure says that gas will move from a high pressure state to a low pressure state until the two pressure areas reach equilibrium.

Bigger dump pipe = lower pressure area immediately after the turbine. Lower pressure area increases the velocity of the gasses through the turbine, thereby increasing possibility for the rate of change or acceleration of the gas flow. This gives the turbo the ability to accelerate faster, increasing pressure until the blow off valve relieves pressure. There is a limit to this from a tuning point of view, and a practical size consideration as well. The rally guys also utilise anti-lag, to keep the turbo spooling when they are off throttle. Anti-lag is hell on turbines however (this IS afterburning for piston engines).

A major thing that the duff-duff boys forget with their ever increasing exhaust sizes, their gas velocity is dropping. A turbo does not work on the pressure differential between the combustion chamber and the exhaust system. It works because of the mass of the gas flowing through the turbine as it moves from the combustion chamber (high pressure) to the exhaust (low pressure). It is possible to get some pulse tuning happening here too. Turbo extractors are a good thing, they are the bunch of twisted pipes you sometimes see that connect the exhaust port with the turbine.

Note that the concept of scavenging has only small application AFT of the turbine. The turbine provides so much backpressure on the engine that harmonics induced by the 4 stroke process are largely cancelled out. If the gas then stalls in the exhaust system due to inadequate velocity (caused by a 4in pipe on a 2000cc engine) then you are going to lose responsiveness and power across most of the range. Where the duff-duff crowd go wrong is the ASSumption that large size is necessary for the whole system. Good exhausts have a larger dump pipe straight off the turbo, but by the time the CAT comes into play, the exhaust is reducing to a smaller diameter. A CAT is also a resonator, and it slows the gasses, and by absorbing heat from them, causes them to contract.

For a race car which operates at high RPM most of the time you need a bigger exhaust to keep the backpressure down due to the increased volume of air passing through. Of course, you have fitted cams, injectors, timing systems and inlet modifications to optimise this operating range, and VE at that range.

Street cars generally don't have to operate at constant high-rpm. This is one reason the car companies are starting to build turbochargers with variable flow characteristics (by vanes directing the airflow onto the turbine). The idea is to increase responsiveness when flow volumes are low, but not inhibit them when they are high.

[adamjp]

Factories build exhaust systems which must conform to the laws of the land, laws of physical fitment, and the laws of cost effective manufacture. Can they be improved on? Hell yes! Do you need 4” fully sik all the way back? Hell no! It is possible to have an exhaust system that is quiet, high flow and power increasing. No accounting for style and decibel induced power however.

An ideal exhaust system would be small at low RPM, but grow to larger diameters as the RPM increases. If you watch the exhaust of a jet engine, you will see it has a variable nozzle to adjust thrust across the operating range (they have variable intakes too). As backpressure is the enemy of VE, this would maximise VE at any point in the RPM range. Trouble is there is not any commercially available technology that does this for piston engines. Due to the wide operating range of a car engine, you will have backpressure somewhere in the higher range in order to retain gas flow in the lower range. Like any other part of the engine, you tune the exhaust to suit the operating band desired, and since we cannot change our exhaust sizes as required, you have to choose one size.

So the accepted wisdom of 2\" for the 1.6 is largely (and empirically) correct. The key is to get the spent gasses out of the way quickly. 2.25\" is probably good for the 1.8, but may actually be a little large. For turbo engines, you will need bigger, but 3” is probably too big, I would suggest 2.5” for most MX5 turbo engines would be a good place to be.

As an aside, the sound from an exhaust will have a lower note (do reh mi) with a larger exhaust as the gasses are leaving the tip at a lower velocity and pulse pressure. Higher speed pulses come from smaller pipes and these consequently produce higher pressure variations in the system and sharper sound. Louder comes from having less resonating chambers (which slow the gasses down and capture really intrusive audible harmonics). A catalytic converter, muffler and a resonator are resonating chambers. The pitch or 'note' of an exhaust is governed by the stroke, the extractor design and the orientation of the various resonating chambers in the exhaust system.