Fuel, Combustion and Engine Performance, an Overview

This rather rambling monograph (it wasn’t intended to be so…) was provoked by a question on Startline about whether one would see any benefit in running a Monoposto engine on higher octane fuel. Would a change from 97 to 99 RON be noticed? I thought I ought to know the answer to this, but soon realised that it may not be as straightforward as I first thought, and the following is the way in which I reminded myself of the basic principles. As combustion tends to be rather misunderstood, I thought I would take the risk of boring my reader with going through some of the basics, and thus improve his life. Use of the singular is intentional.

As background, in the dim and distant past, I used to know quite a lot about fuel technology and its effect on engine combustion/performance as I was employed by the Ford Motor Company as a development engineer on petrol engine performance and economy at their Dunton R & D Centre and did a lot of work with BP which was their partner in this type of thing, at their Sunbury Research Centre. Later I was with Lotus Cars and actively involved with engine performance and design work. The basic facts of how engines react to different fuels have not really altered, so I’m pretty sure I’m talking sense here, but fuel composition has changed enormously, and I know I’m out of date, but again, some things don’t alter. Let’s start with some basic definitions and explanations.

1. Octane Numbers.

There is always too much emphasis on research octane numbers (RON) in this country, as that is what is quoted on the pumps. There is another, more significant anti-knock rating and that is the motor octane number, MON. Both are established on very crude, single cylinder, research engines with variable compression ratios. As everyone knows (?), the anti-knock quality of a fuel is rated on a scale where n-heptane (which has very poor anti-knock qualities) is 0 and iso-octane (which is good) is 100. The fuel being tested is compared with a mixture of these two hydrocarbons. If it is deemed to be 98 it has the same anti-knock performance as a fuel consisting of 2% n-heptane and 98% iso-octane. Over 100 (there are fuels with better anti-knock characteristics than iso-octane) the method is different, and the octane number is replaced with the “Performance Number”. The research engine is now supercharged and the number is derived from the increase in performance when the ignition is knock limited in comparison with running on iso-octane. For example if the engine gives 8% more power with knock limited spark advance then the RON of this fuel is 108. The mixture strength is kept the same to avoid another variable.

The difference between RON and MON is that the MON is determined at conditions slightly more relevant to real world engines, at a higher engine speed and at elevated intake temperatures. MON is always lower than RON and the difference between the two is called the Sensitivity of the fuel. Different fuel compositions can have the same RONs but very different MONs. In America the number quoted on pumps is the average between the RON and MON which is termed the Anti-Knock Index. The important thing to remember is that MON should be the rating that concerns us but finding it out is surprisingly difficult for the end user.

2. Fuel Composition.

The problems faced by oil companies is in producing acceptable motor fuels from deteriorating crude oil stocks at an economical price, bearing in mind the severe lack of refinery capacity (world wide), the processes such as cracking and reforming that are available at these refineries and the many other products that have to be derived from the same input stock. As the demand for the “other products” vary from time to time (such as ever increasing demands for diesel fuel), there is a constant juggling act. On top of his there is legislation that started off with gradual reductions in lead content and now other requirements about more environmentally friendly fuels. As much of this legislation comes from Government, not all is well founded even though their intentions may be right, but could also be seen as cynical vote catching public relation exercises.

Motor gasoline is composed of four types of hydrocarbons, straight and branched chain paraffins, straight chain partially double bonded olefins, ring chains napthenes and six carbon molecules called aromatics. The best known ring structure in terms of schoolboy chemistry is the benzene ring, but health problems have been identified and their use is probably now more restricted than it used to be. When I was at Fords it was the olefin content that was considered of major importance and was causing detonation related problems that lead to piston failures, disgruntled customers and expensive warranty claims. Olefins have good RONs but high sensitivity leading to low MONs, a bit like the bio-ethanol which is currently in vogue.

These days there is a requirement for a certain percentage of bio-content, but my understanding is that this is satisfied largely by adding it to diesel fuel rather than gasoline, as it is an overall requirement. Gasolines that do contain bio-ethanol tend to be retailed by Supermarkets, and the problem is their high sensitivity and the short life before they evaporate away. I’m told that the oil companies only expect a “life” of 4 weeks, 2 weeks from refinery to retailer and two weeks once it has been sold! Typical substances are MBTE (methyl butyl tertiary ether) and ETBE (ethyl butyl tertiary ether). Both are becoming discredited as they can find their way into drinking water and are carcinogenic. The oil companies like them because they use up surplus butane.

Perhaps this is enough basic chemistry. The message that comes across is that fuel quality is generally declining.

3. Combustion Processes.

How do engines react to varying fuel quality? Here we are concerned with combustion defects, rather than other aspects such as hot fuel handling and cold starting issues. There are two chief culprits here, often confused and generally misunderstood, which are pre-ignition and detonation, the latter also called knocking, pinking or pinging (or klopfen in German, only remembered as Ford of Europe tended to be a bilingual operation). Years ago one of my sons described the sound as grasshoppers in the engine. Detonation and pre-ignition are quite different phenomena. Detonation is rapid and uncontrolled combustion after normal combustion has been initiated by the spark. It is akin to an explosion rather than the desirable slow rate of controlled burning and progressive advancement of the flame front that is the hallmark of normal combustion. The still unburned end gas is heated by radiation from the burning part of the charge and is also being compressed further by this combustion. It gets to the stage where it spontaneously ignites in an uncontrolled way. Turning off the ignition will stop the engine. Normal combustion is comparatively slow which is why engines need more spark advance. A rule of thumb to identify a good combustion chamber design is that much less spark advance is required than a poor one would want. 20 to 30 degrees BTDC at high speeds is good, 40 or more, decidedly poor. And the time for combustion to take place (the time for the slowly moving flame front to propagate) is much the same at low and high ends of the engine speed range, which is why we need variable spark timing.

Pre-ignition is, as its name implies, combustion taking place before the spark, and is due to hot spots such as an overheated exhaust valve, spark plug electrodes, sharp edges in the combustion chamber or glowing carbon deposits. Turning off the ignition makes no difference in a single cylinder engine, but in multi cylinder engine generally only one cylinder is affected and its contribution is not enough to keep the whole engine running.

What types of failure are expected and what are the signs to look for? Detonation is not, in itself, hugely damaging. In its most benign form it causes a shot blasted appearance of the piston, especially around the periphery of the crown, which is hardly detrimental at all, but it can get worse as the uncontrolled explosion causes a shock wave across the combustion chamber and which accounts for the familiar knocking clattery sound. This shock wave breaks up the boundary layer which insulates the piston from combustion heat and the piston overheats, expands and partially seizes in the bore. The other pistons keep the engine running, and the piston start to soften and melt. Rings stick and combustion gasses get down the side and generally the whole thing is ruined.

Pre-ignition also results in overheating but in a different way. The peak firing pressures become very high and are concentrated on the PV (pressure against volume) diagram before TDC is reached, in effect trying to push the piston back down again whilst on its way up. The other cylinders keep it going and the high pressures lead to high temperatures and the piston can melt. The melting is typically away from the periphery and the hole is neat and looks molten.

Detonation can lead to pre-ignition. This is the real danger of detonation. Destruction takes place within a few working cycles, i.e. fractions of a second. There is a school of thought that says audible detonation need not be worried about. The damaging sort at higher revs is drowned out behind the general hubbub of the engine. Funnily enough detonation is best detected, audibly that is, (meaning without the benefit of instrumentation), by placing ones fingers in ones ears, thus blocking out spurious noise.

Detonation is made worse by advanced ignition, low octane numbers, especially the MON (implying high sensitivity, highly olefinic fuels), high engine temperatures and poor piston and combustion chamber design. The converse applies.

Pre-ignition is made worse with fuels that have little olefin content but are highly aromatic, high temperatures, excess carbon deposits and is generally only damaging at high engine speeds. Road cars can exhibit run on, which is really pre-ignition, but I haven’t come across this with race engines even with their high idle speeds which are often contributory.

4. Effects of engine and fuel parameters on performance.

What we all want is the maximum performance we can extract from the fuel together with a greater or lesser safety margin between survival and self destruction. Big budget teams may choose to sail closer to the wind and rebuild engines more often. They certainly do not want failures in the middle of races.

The energy in the fuel is a given amount that we, as end users, have no control over apart from the dubious practice of adding additives. The amount of energy is not related to the octane number of the fuel, and all normal gasolines have almost identical calorific values.

Alcohols have lower calorific values than gasoline and the reason that more power can be realised from an engine running on methanol is mainly to do with the charge cooling effect. Its high latent heat of evaporation acts like an intercooler on a turbocharged engine and increases the volumetric efficiency of the engine because the cold charge is denser and can contain more oxygen. Of course fuel consumption is increased. I think there were also safety issues, both for and against. Methanol is not prone to exploding like gasoline but it burns with an invisible flame.

What is important is the shape of the so called full load ignition loop. This is a plot of power, torque or BMEP, (all are related for a given engine speed and give the same shape of curve with only the numbers altering) taken at constant engine speeds with varying ignition advance. A series of these, taken at all the engine speeds that we are likely to operate in, maps the ignition advance. For a given speed, as the ignition is advanced, power increases, reaches a maximum and then tails off. You would naturally want to set the ignition to optimum, but, for a given fuel, the onset of detonation could be before this happy stage is reached, forcing a deliberate retard and a consequent loss of power. If a better fuel would allow the optimum advance to be reached, then a power increase will be seen, even though the fuel has the same calorific value.

In actual practice the ignition loop is fairly flat and the ignition would be set to the MBT figure. This means “minimum advance for best torque” and often the point giving 0.1% power loss is chosen for road engines to help the safety margin. The line plotting the onset of detonation at different engine speeds is called KLSA which stands for knock limited spark advance. Knock is usually worse at the highest combustion pressures which occur at peak torque. If KLSA is always well above MBT then the spark advance may be optimised for performance, but if it is close or below the MBT plot, the spark has to be retarded and performance suffers. KLSA is usually at its lowest point where the peak combustion pressures are found, which occurs at the peak torque revolutions. At higher engine speeds the margin between KLSA and MBT is usually widening.

When I was at Ford low lead level fuels were just being introduced in Germany. Up to then lead levels had been uncontrolled and was typically 0.7 gm/litre, set by the oil companies perception of what was the most economical way of achieving the desired fuel quality. The first European legislated reduction was to 0.4 gm/litre and this seemed to have little effect on engine durability. Of course, it put up the cost of fuel as adding lead (in the form of TEL or TML, that is tetra-ethyl lead or tetra-methyl lead) was cheaper than having to produce higher quality gasoline through different refinery practices. The next stage was down to 0.14g/litre and again this put up fuel prices but had a more marked effect on engine durability and also caused increased fuel consumption as compression ratios had to be reduced. The final death knell for leaded fuels was emission requirements so stringent that catalysts had to be introduced. These could tolerate no lead at all (there is always a tiny trace and current “zero lead fuels can contain up to .005g/litre. Put rather crudely, anti-knock additives such as TEL and TML00 work by releasing a shower of solid lead particles that slows combustion. Should I go into valve seat recession or the need for lead scavengers? No.

Engine power depends very much on compression ratio. There are big improvements to be made but the law of diminishing returns applies. Simplifying the thermodynamics to the air-standard cycle shows that changing from 10:1 to 12:1 will increase efficiency by about 5%, but changing from 12:1 to 14:1 will only yield 3.4%. In real life, other factors mean the air-standard cycle is quite optimistic, but does show why F3 engines run such high, potentially dangerous compression ratios. A turbocharged diesel is not much higher than this and requires no spark to start combustion!

Running these high compression ratios means F3 engines are rather sensitive to both fuel quality and incorrect mixtures which is why special race fuels are used. The Sunoco FR F3 fuel has a 102 RON with a sensitivity of 12. The manufacturers state a “high oxygen content” which I think implies an alcohol content. If normal road fuels are used the ignition has to be so retarded that efficiency suffers severely and a 10% drop in output could be expected. It is also the reason why F3 engines are forced to run at artificially low coolant temperatures, with 50 to 60 degrees being typical. Optimum engine power is generally at about 90 to 100 deg. C coolant temperatures. Interestingly modern F1 engines run in the 120 to 130 degree range so that less drag is induced by the smaller radiators.

5. Conclusions.

Unfortunately none of this has answered the initial question which was “would changing from 97 to 99 RON fuel make a noticeable improvement to a Monoposto car”? As we are now all experts on fuel technology (!), it is easy to see that an increase in RON will only make a difference if it allows the ignition to be advanced to a more favourable part of the ignition loop. In other words was the ignition deliberately retarded to give an adequate safety margin on the 97 RON fuel? Obviously this depends on the particular engine, and having access to meaningful test data is something which few of us have the luxury of and it is generally necessary to resort to acts of faith.

In a road car if the engine has a knock detector, then more power would be obtained from 99 RON providing the spark was knock limited on 97 RON fuel, without the driver having to change anything, as the ignition is always advanced up to the knock threshold.

Without knock detection, and most (all?) Monoposto engines are like this, there will be no effect at all, unless the ignition is deliberately advanced by remapping and then only if the engine were knock limited.

The amount of extra power would probably be no more than 2% but, as they say, every little helps, but, and I hate to say this, the Kiwi was probably right. However we always run on Optimax, and I think the very slight extra cost is justified, not from the power point of view, but the extra safety margin that it gives before needing new pistons.

Martin Cliffe, B.Sc (Hons), C.Eng. M.I. Mech.E

BP Sunbury

An old octane test engine

A more modern (but judging by the PC, not that much more modern) octane test engine

Isomers of octane (C8H18)

3-D Octane

A piston (from a Jaguar engine, as it happens) suffering from detonation.(jagweb.com)

Another damaged piston, this time from preignition (aa1car.com)

Deryk Young's hillclimb Gould runs on Bio-Methanol. That would be "methanol" then. (Toby Moody)

Methanol was also used in 1950's 500's. Click pic for movie of a methanol fire in a big bottle.

Biofuels aren't welcomed by all. This was a protest against large scale biofuel plantations (Not Tesco, though. They sue.) (Indymedia.org).

Sunoco F3 fuel

Which Mono driver do you think is modelling for ad?