All the power is in the head. Part 3 - Burn rate

GC porting, development, valve and seat work, combustion chambers, cams, head construction, etc
Post Reply
Guy Croft
Site Admin
Posts: 5003
Joined: Sun Jun 18, 2006 9:31 am
Location: Lincoln, UK

All the power is in the head. Part 3 - Burn rate

Post by Guy Croft » Sat Jul 01, 2006 1:38 pm

In Part 2, I indicated that we need the following attributes in our competition engine:

- fast burn rate
- higher compression ratio (CR)
- high volumetric efficiency (Vr)

Burn rate

Here we explore the way the expansion of the flame front ( outer periphery of burning fuel) across the top of the piston in the cylinder with the piston around top dead centre - affects our power output.

The first thing to say is we're not talking F1 here, quality of the gasoline itself is a primary factor, and we, as clubman engine owners have little control over that, except to say that we can use additives (like toluene based octane booster) to inhibit adverse effects like detonation.
Although the quality of additives to encourage clean burn (ie: minimum of fouling) may vary between manufacturers, there should be no difference in calorific value ( the amount of energy in it) between maker's gasoline and another or one octane grade and another, octane rating has no influence on torque or power.

There is a very finite period available during which combustion is of any use to us. Complete expansion from 'ignite' to 'extinguish' of the flame has to be achieved at the highest possible rate to build up cylinder pressure because, whilst the combustion tends to take place at more or less constant speed on a given engine, the physical time envelope available for it to occur decreases as the engine speeds up. However there is a balance - we must maintain acceptable levels of engine noise and mechanical loading on the engine. Ignition timing and sufficient spark energy to set off the process is critical.

If combustion is slow and has to be started too early it will 'fight' the piston coming up to top dead centre (TDC) on the compression stroke and if it is slow to burn or spark occurs late, useful pressure at the right time in the power stroke will simply not be available to us.

Expansion of the flame front is - as the name suggests - a controlled release of pressure resulting from a progressive ignition gasoline-air mix in the combustion chamber. To achieve fast burn rate we need a number of things, some we can alter/achieve, others come with the engine:

1. Combustion region shape
We ideally need a combustion chamber that allows even and progressive expansion of the combustion in all directions from the point of propagation - ie: the spark plug - and complete usage (burn) of all our gasoline/air (fuel) mix.

This logically means having the spark plug in the centre of the bore, with a symmetric regime facing the advancing flame front in all directions. In other words a minimum of intrusion things sticking out in the way of the flamefront - intruder dome pistons, valve reliefs, sharp edges and faces, the edges of valve inserts etc. With some engines, notably 8v units there is a trade-off - that can only be derived from extensive dyno testing - between achieving a high compression ratio with a big piston dome and slowing down the combustion/wasting fuel.

The layout has a major influence on one main thing we must avoid - a build up of ‹Å“end gas¢ž¢ in some far corner of the chamber that becomes super-compressed by the advancing flame front instead of burning under its influence. This can lead to ‹Å“knock¢ž¢ ¢‚¬Å“ (detonation) where the end gas suddenly ignites with extreme ferocity, releasing a huge amount of pressure and heat in a totally separate event to the main ignition. This dreaded phenomenon can be highly destructive.

Knock can punch through a piston in one engine revolution, and even light knock can cause material damage to the head and piston. Engines operating with large piston domes and poor mixture quality regimes are particularly prone to it, and the design of a domed piston should always take into account the danger of creating a pocket for end-gas to accumulate. Overheating intake air and low quality gasoline are a couple of other causes.

The ideal piston/combustion chamber region would be a smooth, compact hemispherical combustion chamber with the plug not only on bore axis but placed, in the vertical plane, in the precise centre of the combustion region - and a flat top piston. The current generation of 4 valve combustion chambers take us part of the way there of course, if you want high power the 4 valve per cylinder is a good start for this reason alone.
But this is not the end of the story, because burn rate is affected by other factors too.

2. Mixture quality

The mixture in the cylinder will ideally be a ratio by mass of air and gasoline (this is the air/fuel or A/F ratio as it is more commonly known, though any thermodynamicist worth his name will insist on the air being considered as part of the fuel!).
A chemically correct ratio of gasoline/air is known as stochiometric and is 14.7 parts air to gasoline, and at this A/F ratio, theoretically, all the fuel should burn completely (though for other reasons that do not concern us here) it does not. Stochiometric is also referred to by the Lambda air-fuel ratio designation as Lambda = 1. A Lambda gauge used for monitoring fuelling uses this scale.

To get maximum power we have run a competition engine on the rich side of stochiometric - anything as gasoline rich as 12.6/1 at full throttle. Lean gives economy, but we have to rich for power. 12.6/1 A/F ratio is 0.86 Lambda (12.6 14.7). This richer setting can only be determined by testing and calibration and the burn rate will suffer it doesn't suit the engine. Over-rich can cause aggravated ring and bore wear, but highly turbocharged units are often run richer than equivalent normally aspirated units to cool the cylinder.

When the charge is fully homogenised (mixed) it burns faster and more completely than if the gasoline droplets are too large to mix with air in the time available. With fuel injection we can assure that not only is the charge quality excellent, but we can inject the gasoline at precisely the right time to optimise the firing cycle and get rid of the problem of impaired fuel distribution from one cylinder to another ¢‚¬Å“ an especially bad problem on engines equipped with a single carburettor. We also the avoid the problem of gasoline collecting on the lower sections of the inlet manifold under varying conditions of pressure and temperature.

The quality of the mixture once in the cylinder can be severely affected by contamination of the combustion region by 'other things'. Contaminants displace fuel and reduce the burn rate. Exhaust gas is an example - residual gas left over from the previous cycle is a bad thing, and so effective in reducing burn rate and power that it is used in emission control with EGR systems (exhaust gas recirculation). With EGR exhaust gas is bled into the chamber at part load where the combustion is relatively poor and giving high emission of unburned hydrocarbon (HC). The loss of power from deliberate contamination makes the driver to press the throttle harder to restore power, thus taking the engine into a more efficient regime.

Many things can contribute to a high level of residuals, here are a few basic ones for a start:
1. The exhaust valve cycle may not be optimised.
2. There may be a pumping loss on the exhaust stroke
3. There may be back-pressure in the exhaust system

The issue of exhaust valve cycle is very much tied in with Vr; but in the context of purging the cylinder we want the exhaust valve to open at such a point in the power stroke that we extract as much energy from the power stroke as we can without dropping the cylinder pressure too early before bottom-centre (BDC), but at the same time capturing maximum benefit from a strong pressure wave down the ex port and header at the right time. Spark-ignition engines, especially normally aspirated (not turbocharged or supercharged) rely on pressure waves in inlet and exhaust, without them the engine will never develop its full racing potential.

Overlapping the inlet and exhaust valves around top centre (TDC). We hold the exhaust valve open well past TDC so that incoming charge from the inlet port - with the inlet valve opening up - can sweep across the top of the piston and upper cylinder to 'purge' it free of residual. This is one of the most significant aspects of race engine tuning, and computer prediction methods with race engines demonstrates that overlap is far more efficient than previously considered. Of course, overlap is tied in with inlet charge momentum and cylinder filling (sometimes called 'column inertia), but I will consider that under Vr.

Of course, holding both valves open tends to make the engine response rather poor at lower engine speeds. But, provided that the inlet tract (from rampipe to valve) and exhaust tract (from valve to header junctions) are designed to enable inlet and exhaust pressure waves, both negative and positive, arrive at the right place at the right time, overlapping can be use to huge advantage in developing a flat power curve at high rpm with almost no 'bottom-end penalty'.

A pumping loss occurs when the flow characteristic of the exhaust port is simply too low to evacuate the cylinder of gas under transfer of pressure (high) in the cylinder to (low) in the exhaust header on its own. The whole system is mass and pressure dependent.
Once the cylinder pressure drops to atmospheric, the piston naturally pumps out remaining gas, but the more restrictive the ex port cross section the harder this job is. The whole system is mass and pressure dependent and very much governed by the flow potential of the ex port and valve and the exhaust valve timing and duration. When the cylinder won't evacuate effectively on its own - the piston has to do the work, and it saps engine power.

Exhaust back pressure has nothing to do with exhaust pressure waves - these travel at high but sub-sonic velocity thru the exhaust gas in both directions in the upstream sections of pipework. A quick mention on waves for clarity at this point might be helpful:

A pressure wave will have a negative or positive value. It will of course have a certain magnitude. When a pressure wave hits a closed pipe end or similar solid face it is reflected as the same positive value. When a negative wave meets an increase in volume of critical size it will be reflected as a positive wave. A rampipe is good example of negative wave acting in our favour - at a certain valve lift if cylinder depression is high enough (a function of valve timing and piston velocity) a negative wave coming out of the inlet tract will be reflected back into the cylinder from the open end and help build up charge momentum as it goes.

Back pressure is a build up of static pressure above atmospheric in the exhaust pipework from port to first silencer (muffler) caused by the network being too small in diameter, or over-baffled in the silencer (s), and with normally aspirated engines is always a bad thing, except, to a point, from a noise reduction point of view. Turbocharged engines (but not supercharged) rely on high back pressure ¢‚¬Å“ but only between the cylinder and turbine ¢‚¬Å“ to power the turbocharger. Turbocharged engines require less silencing than normally aspirated ones because the turbine damps out most of the cylinder noise ¢‚¬Å“ and downstream of the turbine the turbocharged engine should ideally run with almost no back pressure at all.

Introducing a ‹Å“controlled¢ž¢ degree of turbulence by means of squish bands improves the homogeneity of the charge (mix of gasoline molecules with air) and aids fast and complete combustion. Squish is also beneficial in reducing the risk of knock (detonation). Squish bands on the head interact with flat crown regions of the piston to create vortices that enhance the mixing of the charge at top centre. As a means of enhancing burn speed, squish is less significant on competition engines because its effect is less pronounced at higher speeds, and squish bands are often sacrificed to an extent during modification in order to enhance valve flow.

3. Ignition timing

Crossflow effect.JPG
crossflow (scavenging) takes place as inlet charge flows across the chamber from inlet to exhaust valve. Lost charge is more than compensated for by the power released from cylinder purging.
Crossflow effect.JPG (36.54 KiB) Viewed 12944 times
AL 019 final cc with 45-40 valves.jpg
... to achieve a high CR on an 8V combustion chamber like this rally spec Fiat TC head (45/40 valves)
AL 019 final cc with 45-40 valves.jpg (44.45 KiB) Viewed 12944 times
MH piston clearance_01.JPG
We need a big intruder dome with some head configurations..
MH piston clearance_01.JPG (18.83 KiB) Viewed 12939 times
AK Cos  16v head rtb.jpg
Burn rate - small chamber means smaller intruder dome and any 16v head (this is a Sierra Cosworth) is more effective than 8v, from superior flame front development alone. Just one of many reasons for better power..
AK Cos 16v head rtb.jpg (26.64 KiB) Viewed 12934 times
Squish band on Vauxhall 1600 SOHC unit. Milling the head down on many heads can create far too much squish - and cause a power loss.
Slide23.JPG (36.29 KiB) Viewed 12931 times
Suzuki 16v ex port before.JPG
Suzuki 16v exhaust port, a bit on the small side, but still capable of flowing 67% of the inlet port - when both ports and valve-seat combos are fully modified. A head with restrictive ports due to coolant galleries.
Suzuki 16v ex port before.JPG (25.87 KiB) Viewed 12925 times
GD 16V 011.jpg
This full spec 1/4 mile drag racing 16v Integrale head generates an E/I (exhaust to inlet) flow ratio of 88%, next-to-no pumping loss in the port.
GD 16V 011.jpg (112.49 KiB) Viewed 12922 times
Polished cc.JPG
Part-finished chamber relieved and radiused around inlets to give smooth and uninterrupted surface for expansion of the flame-front and cross-flow effect valve-valve. Always insist on this op during head mods, it doesn't show much on the flowbench but it
Polished cc.JPG (32.61 KiB) Viewed 12914 times
Detonation - head.JPG
The begining of detonation - maybe over-lean mixture (bad mapping job on this Vauxhall 16v XE head). The alloy has gone 'pasty' and is pinpricked with tiny holes. Gasket blown as well. Detonation nearly always starts on the (cooler) inlet side.
Detonation - head.JPG (27.12 KiB) Viewed 12921 times
Detonation - piston scuff.JPG
Piston scuff from detonation (far more aggressive than bore-wash from over-fuelling). The piston crown has melted and probably the upper ring land too, then the metal has been dragged down the bores.
Detonation - piston scuff.JPG (29.89 KiB) Viewed 12928 times
Detonation - piston cracked.JPG
Detonation - cracked piston - one firing cycle can do this. You rarely hear chronic detonation, the first sign is usually major power loss, smoking or - seizure.
Detonation - piston cracked.JPG (37.48 KiB) Viewed 12926 times

Post Reply

Who is online

Users browsing this forum: No registered users and 1 guest