Now, it has to be said that that the experience of many countries adopting higher rates of ethanol sooner has not apparently been as calamitous as feared. This could be because those countries don’t have as much of a history of running older vehicles as in the UK and US, but when it comes to the safety of our fuel systems and the seals used therein, it pays to be careful.

In 2008, the Minnesota Center for Automotive Research published a series of papers looking at the effect of E20 (20% ethanol in gasoline) on fuel system plastics and elastomers. The study was to compare the effects of E20 with those of E10 and non-oxygenated gasoline (that is, containing no ethanol) fuels by testing to Automotive Industry SAE and ASTMS standards.

In the case of the elastomers, samples were prepared and immersed in the fuels at 55 C for 500 hours, after which changes to their physical properties of volume, weight, tensile strength, elongation at break and hardness were measured. The results of testing acrylic rubber, polychloroprene (Neoprene), nitrile rubber (NBR), nitrile/PVC and fluorelastomers showed that although there was a degree of swelling for all of them when the samples were immersed, there were no significant changes in hardness or tensile strength.

Apparently sponsored by the US Department of Agriculture, however, which was encouraging the growing of wheat corn for the production of ethanol, this report was not considered to be totally independent, so when the Indian government was considering a move from E5 to E10 the work was repeated. Testing four types of elastomer – Neoprene, NBR/PVC, hydrogenated nitrile butadiene rubber (HBNR) and nitrile rubber – representing those commonly found in their markets, and using fuels more typical of those found in India (19% olefin, 28% aromatic), the results were interpreted slightly differently.

Since the weight and volume of the Neoprene samples were far greater it was concluded that, compared to the E5 fuel, Neoprene was totally incompatible with that of the E10. For nitrile and HBNR specimens there was also a significant increase in weight and volume between E5 and E10, which at the same time was accompanied by a loss in hardness and tensile strength. HBNR and nitrile rubbers were therefore also considered to be incompatible with E10. On the other hand, PVC/NBR showed better overall resistance from E5 to E10.

So while in the wet state all the elastomers apart from PVC/NBR exhibited some significant degree of swelling in both E5 and E10 fuels, when dried they all showed signs of some level of leaching – all of which proves that with hundreds of different types of elastomers currently used in engines, many older designs could experience sealing failure with the advent of higher levels of ethanol in fuel.

Fig. 1 - Test results after immersion in the fuel

Written by John Coxon

The fixed-jet carburettor works by pulling fuel from a reservoir in response to a pressure signal from the venturi in the engine intake air. The depression thus created causes fuel to flow from the fuel reservoir in the carburettor while at the same time encouraging a small flow of air at atmospheric pressure to mix with the fuel. The fuel is metered through the main jet while the air is controlled through what is generally called the ‘air correction jet’. Somewhere in between, the fuel and air are intermixed to form an emulsion – a fine dispersant of the air inside the fuel. This assists the atomisation of the fuel as it eventually enters the engine airstream.

At times of low fuel demand, the air drawn in is very small and the fuel flow regulated by the main fuel jet will contain little in the way of an emulsion. With the throttle only slightly open, drops of liquid fuel will land on the throttle plate and atomise thereafter into the air. However, with increasing engine speed or load, as the fuel demand increases then the restriction of the main jet causes an increased depression in the air correction circuit, and this air passing through the emulsion tube effectively leans off the mixture: the higher the engine speed or load, the greater this air correction effect. Metering the fuel and the air is one thing, but mixing it and presenting it in a form so it readily mixes with the intake air is quite another. This is the function of the emulsion tube.

The essential points to remember are that the fuel passes around the outside of the emulsion tube at its base while air, coming from the top, comes down the inside of the tube and is extracted through a series of drilled holes of varying sizes and heights to mix with the fuel on the outside. When the engine is stationary, the fuel level will be the same as that in the float chamber, and will come to a level within the emulsion tube. As soon as fuel is demanded, the fuel level will drop in the chamber, uncovering more holes that will allow more air to mix, thus leaning the mixture.

As well as altering the size and height of these holes, it is also possible to alter the diameter and thickness of the emulsion tube in its cavity within the body of the carburettor. This acts as a restriction to the flow of fuel which, when all fashioned together, can tailor the flow of fuel more or less precisely to that required by the engine throughout its operating map. Under wide-open throttle acceleration the emulsion tube plays little part since the overriding effect is that of the main and air correction jet. At part-throttle however, when the quality of the fuel atomisation arguably has to be significantly better, the emulsion tube can be considered more critical. Understanding this and being able to apply it in practice is therefore one of the dying ‘black arts’.

Setting up a fixed-jet carburettor may be a long and often confusing business, but when the vehicle starts and drives smoothly and progressively, the satisfaction is immeasurable.

Fig. 1 - Fuel enters through the main jet at the bottom and flows out of the large holes just above the base. Meanwhile, air comes in through the air correction jet at the top, coming out into the fuel through the small holes midway down

Fig. 2 - Emulsion tube from a fixed-jet Weber DCOE carburettor

Written by John Coxon

Let me say here that I consider that the use of specialist racing fuels as being no bad thing. Racing fuels tend to burn so much cleaner, and for those who complain about the increased cost surely this is insignificant compared to all the other costs of competing?

However, for those who because of choice or budget restrictions prefer to use ‘pump’ or service station forecourt fuels, the issue I am about to refer to may make them think again if they continue to use lightweight aluminium fuel tanks.

Although light, aluminium is a highly reactive metal that relies on an oxide layer for protection against its corrosion. This oxide layer occurs naturally, and the low levels of ethanol such as E10 (10% ethanol, 90% gasoline) or the 5% ethanol blend often found in Europe are usually not a problem. The problem occurs principally when the ethanol content in the fuel is increased beyond these levels. The issue, although complex, generally centres on the presence of water in the fuel.

The accepted reason for the corrosion of aluminium in ethanol-blended gasoline fuel is the phase separation theory. According to this, the hydrophilic (water-loving) property of ethanol allows water to be absorbed into the fuel, which helps to divide the single phase ethanol-gasoline mixture into two phases – a water-ethanol mixture and gasoline. Acting as a form of galvanic cell, and helped by the presence of alloying metals in the aluminium, the oxide layer is overcome and erosion/corrosion takes place. 

There are many ways of tackling the problem though. Less susceptible forms of aluminium can be used, but this doesn’t really help anyone with an existing aluminium tank, although the tank can be modified by anodising the internals. Anodising or hard anodising creates a more resilient oxide barrier that tends to be thicker than the normal aluminium oxide surface and better protects the metal underneath.

Other barrier systems can be coated on the inside of the tank, substances such as LLDPE (linear low-density polyethylene) or some kinds of epoxy-based resins similar to those used in glass-reinforced fibre applications. Tanks made from GRP (glass fibre-reinforced plastics) have also sometimes been recommended as an alternative for boats using aluminium tanks, but even these may be suspect at the higher levels of ethanol mentioned, so do your research carefully.

Of course, for anyone with the appropriate budget, some form of foam-filled internal bladder would be the best choice, and this would be my favoured approach. The aluminium tank would need to have access apertures bored into it for the bladder to be inserted, but the extra safety against serious impact damage in the event of an on-track incident must surely be worth it. Not compulsory in many forms of racing – perhaps it should be – but the requirement to replace it every five years irrespective of use might put off many budget racers.

The current discussions in the EU are only proposals, and are vigorously opposed by vehicle manufacturers, and although the fuel developers will introduce some form of anti-corrosion additives, the mandatory introduction of higher levels of ethanol could mean yet another raft of modifications to racers of vintage or classic machinery.

Fig. 1 - Aluminium fuel tank with internal bladder

Written by John Coxon

But while the science of injector fouling is now well understood, the one thing that can increasingly defeat many a design is its hot fuel handling capability. And if you consider that the role of the fuel in contemporary injector architecture is one of cooling then this would seem to be a major failing.

In most commonly used injectors these days, the fuel arrives via the pump and filter and accumulates in a fuel rail under pressure. In response to the ECU-activated signal energising the solenoid, and thus opening the injector, the fuel flows axially down the injector through a simple strainer and around the solenoid coil, thus cooling it. So long as the fuel continues to flow and is injected into the cylinder, all will be well, but when the fuel stops flowing then any residual heat soak from the engine or the injector solenoid will migrate into the fuel still sitting in the injector.

As the temperature of this fuel rises then fuel vapour will form, causing a phenomenon called vapour lock, making restarting the engine very difficult. With the desire for engine packages to become much smaller, and as engine performance levels increase, this extra heat and the proximity of components such as exhaust-driven compressors will make the issue of vapour lock even more critical. In future therefore, hot fuel handling could be a major issue in low-pressure port injection systems.

Fear not though, all is not lost. One way around the problem is to increase the pressure in the fuel at restart. This should condense the fuel in the injector back into its liquid form and enable the engine to start again as planned – and indeed, most current vehicle systems work this way. Another way though, and one that is highly attractive in liquid-fuelled LPG applications, is to use so-called ‘bottom feed’ injectors.

Here, instead of feeding the fuel into the top of the injector and clamping the injector between cylinder head and the fuel rail, as is normally the case, the fuel intake into the injector is via a series of drilled holes through its side towards the bottom. Flowing through a gallery, perhaps even integrated with the cylinder head, the fuel passes into the injector and is injected out into the cylinder without going anywhere near the controlling solenoids, and any vapour (which is less likely to be formed anyway) can be easily vented and out of harm’s way. This type of system works best when the fuel is re-circulated back to the tank for cooling, and will be essential if using LPG-type fuels that boil at temperatures only slightly greater than ambient at rail pressures of about 8-10 bar.

Bottom-feed injectors were typically used on throttle body systems many years ago, but with the increasing interest in alternative fuels could they be another case of ‘back to the future’?

Fig. 1 - A bottom-feed injector

Written by John Coxon

From a technical standpoint therefore, and since higher pressures must inevitably mean tighter running clearances for the moving parts (pumps, injectors and so on), a diesel fuel filtration system would seem to pose the greater challenge. Add to this changes in the fuel designed to safeguard exhaust after-treatment devices (where fitted) and the progressive movement towards fuels from ‘bio’ sources, and the challenge of fuel cleanliness at the engine becomes Herculean. So in one instance, while the reduction in fuel sulphur to less than 15 ppm helps the after-treatment devices, at the same time it reduces the diesel fuel’s natural lubricity and its consequential ability to tolerate contaminants. And although the addition of fuel components from bio sources can help to regain some of this lubricity, diesel’s affinity with water brings problems of a different nature.

Found in all diesel fuels of one description or another, the presence of water is almost inevitable. Present as a result of poor storage or environmental conditions when the tank is vented to the atmosphere, water can cause engine issues like corrosion or erosion as a result of fuel lubricity deterioration or fuel pump cavitation. Furthermore, water can also be implicated in the build-up of injector deposits as well as fuel filter plugging at the surface interface, where damaging bacteria can grow. In short, water in fuel is bad news but that found in diesel engine fuel – whether as free water, dissolved water or as an emulsion in the fuel – is doubly so.

The mixture of hydrocarbon chains that make up a typical diesel fuel have a natural tendency to attract water. This is because water is a polarised molecule that is readily attracted to those parts of the diesel fuel hydrocarbons that have a similar but opposite charge. Once this amount exceeds the saturation point of the fuel then any excess will fall out, forming an emulsion consisting of small droplets of water suspended within the fuel. The final stage is when these droplets coalesce and, since water (specific gravity: 1.0) is denser than diesel fuel (specific gravity: 0.86-0.90), they will fall to the base of the tank as free water.

At this point it is imperative that the water is removed, which is why most fuel filters designed for diesel engine applications include some kind of water separation system. Placed on the primary side of the fuel system – the suction side of the fuel pump – in this position the fuel pump is not only protected but the water is separated before the action of the fuel pump emulsifies the water-fuel mixture. Removed from the fuel by either stripping it out using a water repellent or hydrophobic medium, or by capturing the fine droplets and coalescing them to form much larger and hence more stable droplets, either way the free water will eventually fall to the base of the filter, from where it will be periodically removed.

To most gasoline engine users the fuel filter is a ‘fit and forget’ device introduced in the fuel line on a ‘just in case’ basis. For a diesel engine a fuel filter is an absolute must.

Fig. 1 - The action of the diesel fuel filter

Written by John Coxon

Still recognised as one of the most sophisticated yet easy to use fuelling systems of its time, and widely regarded as the most significant of its kind in Formula One history, the arrangement uses a moving shuttle to meter the fuel to individual cylinders. The shuttle operates between a fixed stop and a moveable stop to meter the flow and, using a rotor geared to half engine speed spinning around the metering sleeve, times the delivery according to angle of the transfer ports.

With a constant fuel pressure of up to 150 psi, the function of the injector is therefore solely to atomise the fuel rather than meter it. Fuel delivery at the normal wide-open throttle use of race engines is therefore proportional to engine speed, and is hence only an approximate (but perfectly satisfactory) estimation of the fuelling requirement. More than this, however, the system has a method of altering part-throttle fuelling using a 3D cam profile mechanically linked directly to the throttle.

Crude by 21st century standards, certainly, but the system was popular with engine builders simply because engine fuelling characteristics could be changed very easily with only limited test bed running. Nevertheless, since fuelling was timed to inject just at the optimum time, losses through fuel disappearing past the exhaust valve were minimised, producing surprisingly good fuel economy.

But all this required two fuel pumps, the first an electric one to prime the system and, once the engine was started, a second to keep the engine running. Mechanical in nature, this second fuel pump was a simple positive-flow gear pump with the gears a compact 6-7 mm in thickness according to the fuel flow desired. Fuel would come in at the low-pressure side and be drawn around the periphery between gear and pump casing to exit at the high-pressure side. Geared to the engine crankshaft, delivery per engine revolution is therefore more or less the same throughout the speed range.

Simple and very efficient, the main issue with positive pumps of this nature is the cleanliness of the fuel, as the presence of a single grain of dirt, casting sand or machining swarf can cause major scoring between the pump gears and housing, with disastrous effect. In such cases it is therefore imperative to fit a fine-mesh filter immediately before the pump to catch any contamination before it can do the damage.

Although an electric fuel pump is needed for starting, once running the engine would idle very easily, producing negligible bore wash inside the engine. Furthermore, the design of the fuel injection nozzles produced little or no dripping on engine shutdown, preventing fuel from dripping slowly into the engine and potentially causing an hydraulic lock – with disastrous results – the next time the engine was cranked. But perhaps more important than any of these as far as driving was concerned, throttle response was instantaneous, which explains to some extent why drivers often preferred this system over any other contemporary designs in the period. Both cooled and lubricated by the fuel passing through it, even by modern standards the system is incredibly simple, yet highly effective.

Winning more than 200 Grand Prix victories throughout the 1960s, ’70s and ’80s, it was only the advent of turbocharging and the need for electronically controlled fuelling systems that led to the advanced equipment we see these days.

Fig. 1 - Cross-section of the Lucas mechanical fuel pump

Written by John Coxon

Although there are many types of oxygenate, only two types are commonly used  – the alcohol family of methanol, ethanol, iso-propanol or TBA (tertiary butanol); and the ether family of MTBE (methyl tertiary butyl ether) or ETBE (ethyl tertiary butyl ether).

In terms of the relative oxygen content in the fuel molecule, alcohols tend to contain the most, with methanol the greatest at around 50% by weight. The more oxygen in an alcohol when blended with gasoline, the less in theory that has to be drawn in through the engine intake valve, and I suppose therefore in terms of an outright performance increase, methanol should be top of the list. And so you might expect, but for the fact that the lower heating value of methanol is less than half that of your normal gasoline, so what you add in terms of oxygen content you take away in the form of heating value, and to burn all the intake air from your typical engine it will take around twice as much as that of your usual gasoline it replaces.

The great advantages of the alcohol fuels like methanol, however, are their high heat of evaporation (six times that of regular gasoline) and their exceptional resistance to combustion knock, which make them ideal for supercharged or turbocharged engines. On the negative side, methanol has poor solubility in gasoline, which is partly why ethanol, the higher alcohol, is more readily blended at the pumps – this and perhaps the reason that ethanol can be manufactured by fermentation of biomass, which makes it much more politically expedient in the modern environmentally (and politically) aware world. The blending of alcohol fuels, particularly methanol into gasoline, can also have a significant effect on the vapour pressure of the resultant fuel, especially at low treat rates [see Figs. 1 and 2]. This can make it a challenge to blend a fuel to within the narrow band of vapour pressures necessary to give consistent combustion.

Of the other oxygenates available, until ten to 15 years ago, MTBE was the most common blend component for gasoline. Containing a useful 18% oxygen by weight, its heating value is only slightly less than that of the gasoline with which it is mixed. The chief attractions of MBTE, however, are its high level of knock resistance (around 115-123 RON) and its benign effect on the fuel vapour pressure of the resulting gasoline blend, particularly at low treat rates. Added to some fuels by as much as 15%, political pressure in the US led to the banning of MTBE in fuels because of ground water contamination issues in bulk storage tanks, and was replaced by ethanol. Nevertheless, many race fuels still use MTBE which, when allowed, is far less corrosive to an engine’s internal parts than alcohol.

They may not be as environmentally acceptable as ethanol, but for naturally aspirated engines the lower ethers such as MBTE have much to recommend themselves.

Fig. 1 - Common oxygenates

Fig. 2 - Effect on vapour pressure of key oxygenates 

Written by John Coxon

But the supply side of modern manufacturing is not the only place where the supply needs to be matched to demand. Another place, one perhaps of more interest to RET-Monitor readers, is in modern automotive fuel systems, where just sufficient fuel as is needed has to be pumped out of the tank to the exact pressure required at just the right time – no more and certainly no less.

In the more traditional manifold or port-injected system, the fuel is pumped into a simple rail, at the end of which lies a fuel pressure regulator. Once the fuel is up to the pressure required, any excess flows through the regulator valve and returns to the fuel tank, taking with it any heat absorbed during its passage across the engine. Accurately metering the fuel these systems requires a constant fuel pressure difference across the injector, and to maintain this the fuel pressure regulator is linked to the intake manifold by a simple tapping that alters the fuel pressure setting by the amount equivalent to the depression in the manifold. A highly functional system, you may say, but in supplying more fuel at pressure than is strictly necessary and then spilling it back to the fuel tank it is surely grossly inefficient.

A much more efficient way that complies with the ‘just in time’ principle is the demand-controlled system. Dispensing with the pressure regulator as such and replacing it with a closed-loop control system monitoring fuel rail pressure ensures that just enough fuel is delivered to the engine at the right time – and, perhaps more important, at the pressure desired. In compressing only that fuel necessary to feed the engine, engine efficiency is improved, and rather than controlling the fuel pressure difference across the injector mechanically, it can be controlled much more simply and electronically via the engine’s ECU. 

Of course, we still need things like a pressure relief valve should things fail, but this can be packaged inside the fuel tank along with the pump. Arguably, however, the biggest advantages of these types of systems are their adaptability and ability to increase fuel pressures if required. When an engine is hot, for instance, and the heat soak into the fuel rail at, say, a pit stop generates a fuel vapour lock (increasing the fuel rail pressure) then these systems can be made to condense the fuel vapour bubbles, enabling the engine to inject fuel and fire again. Also, in turbocharging applications, the metering range of the fuel injectors can be widened by changing the pressure difference across the injector – increasing it for full-load, wide-open throttle running and reducing it when much lower power is required or when the engine is at idle, effectively improving the precision of the fuelling.

Fuel at the right pressure delivered just at the right time? What more can you ask for?

Fig. 1 - Fuel pressure regulator for a port-injected engine

Written by John Coxon

Fortunately, and for the sake of simplicity, the vehicle in question has only one tank, situated more or less alongside the driver. The problems of feeding from two or more fuel tanks are therefore thankfully avoided, although the problems of fuel surge within the tank when cornering are not. The tank will be of a safety bladder variety inside an aluminium skin, much resembling, as far as aesthetics are concerned, the current aluminium unit but far safer in the event of a major accident. Just because such an accident has never happened to date doesn’t mean it won’t, and since the driver values his skin (but probably not as much as the rest of the vehicle) the added cost as well as the extra weight is a penalty simply to be endured.

 Inside the tank is the first dilemma. Do we include a series of trapdoors leading to a single make-up tank where the sole lift/high-pressure pump will be situated, or do we dispense with that and fit a simple low-pressure lift pump and send the fuel to a smaller, half litre or so ‘surge’ tank? Whichever in-tank pump we select, it will be fitted with a simple coarse gauze strainer to remove the larger pieces of debris, rust or dirt.

No matter how careful you may be in transferring the fuel from bulk storage into jerry cans and then into the vehicle tank, there is always the risk of small grains of rust or dirt contaminating the system. While pumps may be tolerant to a certain extent of smaller foreign bodies (say >30 microns) the fuel injectors are not so, so at some point in the system a much finer filter will need to be installed. And here is the next dilemma: where do we install the second, much finer filter – immediately after the-low pressure lift pump and before the surge tank, or after the high-pressure (up to 6-10 bar) pump and before the injectors?

In essence, I suppose it all comes down to convenience, personal preference or simple packaging in the vehicle, but placing it after the low-pressure lift pump pumping into the surge tank protects the high-pressure pump from the finer debris (down to, say, 25 microns). In this position, however, there could be circumstances where the filter places some kind of restriction or depression on the intake to the pump, and when pumping volatile fluids this is never a good idea.

As a general rule, I have always maintained that any filter should be placed on the high-pressure side of the pump. Although the filters used have large surface areas over which the filtering takes place, placing them after the high-pressure pump minimises the chance of vapour lock or cavitation as a result of any depression (however slight) should you place it on the intake side, but also ultimately protects the injectors (and hence engine) should the fuel pump fail for whatever reason. For much the same reasoning of vapour lock, the high-pressure inline pump should always be placed as low as possible in the chassis and the surge tank mounted above it.

Although apparently simple, the design of the fuel system has to be carefully thought out.

Fig. 1 - Fuel system

Written by John Coxon

Defined as the movement from one state to another, the progression of a fixed-choke carburettor is the transition stage between engine idle and the point when the main fuel system comes into play, and in my opinion is the one area in tuning a carburettor that is the least understood. The difference between an engine that is ‘tuned’ to give to give good progression and one that is less well tuned is the difference, in vehicle driving terms, between ‘chalk and cheese’. While both can produce the same ultimate performance, the former can be a delight to drive while the latter is a complete dog.

On all fixed-choke carburettors – not just those with a single choke per cylinder – at the engine idle setting the throttle plate will be positioned such that only a small amount of intake air will bleed past into the engine. Mixed with fuel metered through the idle fuel jet downstream of this plate, when burned, the power produced will be just enough to overcome engine friction at this idle speed.

As the throttle is slowly opened, the pressure difference across the throttle plate will fall and the air speed into the engine will increase. During this period, however, the increase in airflow will not be enough to bring the main fuelling system – the auxiliary venturi, main jet, emulsion tubes and air correction jets – into play. If not corrected in some way, the lack of fuel from the now not so effective idle system will make the engine run lean, stumble and possibly even misfire. To the driver, the engine may feel unwilling or unresponsive when moving away from rest and lead him (or her) to complain about ‘flat spots’ or having to ‘rev’ the engine much more at pull-away.

To combat this, a small number of tiny progression drillings are introduced into the wall of the carburettor, slightly away from that of the idle fuel drilling, to increase the flow of fuel. Linked to the same circuit as that of the idle fuel circuit, as the throttle plate progressively opens then each drilling at a time is uncovered and more fuel drawn into the airstream. Usually consisting of two or perhaps three holes of differing sizes, by the time the last one has been uncovered the main jet circuit will be starting to function, making both the idle and progression circuits redundant, since the depression in the port compared to that in the carburettor float chamber will by now have almost disappeared.

It’s very much a trial-and-error procedure, and if you drill these holes in the wrong place then the carburettor could be rendered scrap – which in a funny way could be why nobody ever talked about it at the time.

Fig. 1 - Three progression holes just uncovered by the throttle plate. The single hole above is for the idle fuel  

Written by John Coxon

So while simple aluminium fuels tanks are still sold in their thousands, such is their inability to prevent or even control the spillage of fuel after a damaging impact that the only option is to use some form of safety fuel bladder. Mandatory in Formula One for many years, these have been adopted by virtually all forms of racing these days, such that the FIA has introduced controls and homologation to ensure standards. The standards cover the inclusion of baffles (to reduce fuel slosh), fittings and connections as well as limiting the age of the tank, but the biggest section is reserved for the design and performance of the materials and construction.

Thus for instance the bladder itself should be made from “a reinforced material in polyamide, polyester, aramid or equivalent” and be impregnated with and coated both internally and externally with some form of fuel-resistant elastomer. When woven, polyamide and polyester fibres are well known for their use in clothing. In the case of aramid fibre the most famous brand name of a material is Kevlar; as well as being known for its use in carbon fibre composites and anti-stab vests in military applications, its resistance to heat and puncture make it perhaps the ideal fuel tank material. Because of cost and manufacturing issues, however, bladder tank manufacturers tend to use their own versions of woven materials to meet the FIA specification.

The tensile strength test, for instance, goes into quite significant detail, and while specifying the size of the test specimens (25 mm in width, a minimum of 150 mm in length) and how they are to be selected and cut from the bladder material to be tested, the test procedure also asks to show that the material doesn’t have tensile properties biased in any particular direction. While tensile strength testing is rigorous, perhaps even more so is the testing of a material’s puncture strength and the resistance of the seams or material to tearing in the event of impact.

The puncture resistance test is similar to that of the tensile strength test, differing only in the use of a piercing tool that looks much like the blade of a screwdriver. An interesting choice of puncture tool tip, you might add, but as some of us may know, a weak bladder can be more than just an inconvenience.

Fig. 1 - Touring car fuel tank

Fig. 2 - Hillclimb single-seater. Holding only enough fuel to complete the course, I would hope this tank will have some kind of fuel tank bladder installed, even though it’s still packaged away behind the driver 

Written by John Coxon

That original Mercedes-Benz engine was very unusual, to say the least. An inline eight-cylinder with desmodromic valve gear, a fabricated roller bearing crankshaft but with only two valves per cylinder, the engine still delivered around 290 bhp at 8500 rpm (116 bhp/litre) on a rather exotic fuel blend. Remarkable for its impressive low-speed torque, the fuel was injected mechanically through the side of the cylinder, angled upwards at 12.5º, immediately below the intake valve. Injecting on the induction stroke at a fixed 30º after top dead centre and continuing for another 160 crank degrees until just after the intake valve closed, the mixture was fired by two rather large 14 mm spark plugs.

Sixty years later, fuel systems are wholly electronic (in so far as their timing and control) but in terms of actual combustion the principles are still very much the same – the fuel needs to mix with the air and not go onto the cylinder walls. The only real difference therefore is how we go about that, but with four valve combustion chambers, centrally mounted spark plugs, more compact combustion chambers and higher engine speeds, the challenge is all the greater. And while early DI injectors may have been described as ‘wall-guided’ – where the fuel is sprayed onto the piston crown and evaporates off it – more modern systems tend to be spray-guided, where a series of up to 12 small, almost micron-sized, holes in the nozzle of the injector ‘atomise’ the fuel into the air through a series of spray cones.

Injected at fuel rail pressures up to 230 bar to achieve the flow rate necessary against the cylinder pressure, the task is to match the flow rate of the fuel to that of the incoming air. So while the Mercedes engine will have injected a constant rate of fuel to deliver the power, modern injectors with response times of 0.4 ms or less can be pulsed to suit the instantaneous air flows in the cylinder. And while engineers in the past may have intuitively estimated (read ‘guessed at’) the flow conditions in the cylinder, these days the intelligent use of computational fluid dynamics together with combustion models can supply a closer approximation.

So while the principles are still the same, the digital revolution in both control and simulation is the enabler.

Fig. 1 - So-called ‘second generation’ piezoelectric direct fuel injector

Written by John Coxon

If an engine is designed and carefully assembled according to well-defined procedures and then generously run-in to acceptable standards, after a while the wear metals will reduce in size to those generally smaller than those the oil filter can trap. At this point, although I would never truly recommend it, the filter could be removed –I have tried it in the past, and it worked without any undue issues with the engine. So long as you change the oil frequently then the system will most likely remain mainly unaffected. Likewise with the air filter. Careful design of the intake system could produce a system where the larger dust particles can drop out, leaving only those that would pass through the air cleaner, so removing the air cleaner will have no effect.    

When it comes to the fuel filter, however, the risks these days are totally different. A simple carburettor for instance would have a small mesh screen. Preventing any debris from passing into the body of the device, any smaller detritus would pass through and be drawn through the engine with little apparent effect. Modern engines are far more sophisticated though. Using high-pressure fuel pumps with close-fitting manufacturing tolerances, the fuel is pumped up to the injectors. Here, at pressures up to 200 bar, the fuel is injected through a multitude of (often up to 12) tiny micron-sized holes, each carefully angled and positioned to deliver the correct amount of fuel into the turbulent air precisely as required. Any dirt or detritus interfering with the operation of the system, however little, could clearly spell disaster.

That is why the best fuel systems have at least two fuel filters. The first of these is more of a strainer. On the suction side of the in-tank mounted pump, this will remove much of the heavy sediment, scale or dirt which (despite every effort by fuel manufacturers) always finds its way into the tank. This filter is more to protect the pump, which will tolerate a lot of the debris but still has to be saved from the worst of it.

After the fuel but before the fuel rail or high-pressure pump, a more refined filter should be positioned. Designed to trap more of the finest dirt, the essential media doing the trapping can be synthetic fibre, cellulose or sometimes even a simple wire mesh or metal strainer, depending on the fuel used, the degree of filtration required or the time needed between cleaning/replacement. Motorsport filters for instance could use, say, a 55 micron mesh screen to get the flow rate necessary, but this would come at the loss of filtration efficiency and potential damage to the fuel system. Provided the flow rate necessary could be maintained, a better alternative could be an 8 micron pleated paper filter. With up to 55 sq in of filtration media for more modern vehicles, this should be a better solution.

Fuels for motorsport, by virtue of the fact that they are often stored in small containers, are more susceptible to the ingress of dirt or rust from the cans. Forecourt fuel on the other hand is likely to be much cleaner.

Surely therefore, in motorsport the last thing you want to do is to dispense with the fuel filter.

Fig. 1 - Inline fuel filter showing a rather old nylon mesh within a plastic frame. Later versions of this filter have an optional metal mesh or paper filter

Written by John Coxon

For port-injected engines the role of the fuel pump is more complex. Designed to supply full engine fuel flow rates at anywhere between 60 and 200 litres per hour at a constant pressure above that of the manifold intake, the pump also needs to be operated independently of the engine for events like cold starting or when the engine has stopped – as in the case of a shunt, for instance. Typical rail pressures will be of the order of 3-6 bar (45-90 psi), and one of two types of pump were traditionally used when they were more commonly mounted inside the fuel tank and integral with fuel filters.

The first was a positive displacement pump, but instead of a reciprocating diaphragm, roller cells or simple internal gear pumps would be used. These could deliver the high pressures necessary and be relatively insensitive to fluctuations in the voltage supply. Unfortunately, these pumps also produced high-pressure pulses inside the fuel rail that could influence the injector metering, so more modern electric pumps are more likely to be of the ‘flow’ type. Flow-type pumps use a complex, impeller-based system that slowly but incrementally increases the fuel pressure as the fuel progresses steadily through it. Since this pressure builds up steadily within the unit, the flow is practically pulsation-free, giving a huge advantage over other pumps.

However, when it comes to direct-injection systems, where the fuel is injected directly into the combustion chamber, the pumps tend to be very different. While an electric pump may still be used to scavenge the vehicle fuel tank and introduce the fuel to the low-pressure rail, a further high-pressure system is necessary to get it to the delivery pressure necessary to inject it into the combustion chamber. In such cases, and to increase the pressure from 3-5 bar up to 150 -200 bar, we have to revert to a mechanical system operated from the engine camshaft.

During the evolution of gasoline direct-injection systems, development has followed similar trends in the port-injected world such that systems can be either continuous delivery or demand controlled. The former are more suitable for racing applications since the quantity to be delivered is proportional to the rotational speed, and with as many as three barrels per pump positioned at 120º to each other around the camshaft drive, there is minimal fluctuation in outlet pressure. Compared to demand-controlled systems using a single barrel per cam, there is also little need for low-pressure pulsation attenuation systems. With the maximum fuel delivery configured to that of the maximum demand of the engine, any excess fuel not required is returned to the suction side of the pump, with the high-pressure rail remaining at a constant and hopefully pulse-free pressure.

The pressures may have increased but the development from carburettors to direct-injection fuel systems still need to address many of the same issues.

Fig. 1 - A selection of low-pressure fuel pumps

Written by John Coxon

But getting to that situation where all the fuel can be scavenged is an issue that has tasked engineers for many years, because as soon as the end of the fuel pick-up pipe is uncovered, the efficiency of the fuel pump falls and it will far sooner try to pump air than fuel.

Traditionally there are a number of ways of attacking the issue. The first might be to introduce a kind of foam plastic into the fuel tank. Using foam with an open-cell structure will reduce the mobility of the fuel around the tank, but under longer lateral g-loads the fuel will still move across the tank, and the foam – which is designed principally to prevent slosh and reduce the rate of spillage in the event of a tank rupture – will be largely ineffective in battling fuel surge.

Another, far more effective, way is to use a system of trap doors around a much smaller ‘fuel surge tank’ within the main tank. This allows fuel to move freely towards the pump pick-up but closes a trap door when the fuel tries to drain away. Having these trap doors arranged such that enough fuel is always trapped is highly effective and highly popular in many track applications.

Yet another way is a sort of swinging pick-up. Attached to the fuel pump at one end and having a fine mesh filter at the other, moveable end, the loose end is allowed to move freely in the tank. If the fuel pump is positioned centrally in the fuel tank, the arc of the loose end will follow the movement of the fuel side to side during surge conditions and – so the theory contends – fuel pressure will be maintained throughout. While surprisingly effective, systems such as these do take a little bit of fine tuning to get them working to their best. 

However, the most sophisticated way of avoiding fuel starvation is to have multiple fuel pumps positioned around the tanks and in those areas where the fuel eventually resides under the high-g conditions. Switched in and out according to pre-programmed software, fuel tanks can now be made in all manner of shapes and sizes to suit the space available on board the vehicle.

It may not be the easiest of tasks or the most glamorous, but scavenging the last drop of fuel out of the tank is well worth the effort.

Fig. 1 - Typical fuel cell incorporating a surge tank

Written by John Coxon

But the theory was simple. Take the fuel from the tank, pump it at high pressure (3 bar - we’re talking the 1980s here) to a common fuel rail and from there the injectors would direct it into the runners of the intake manifold. What could go wrong?

Well, plenty! The first thing was to ensure that for accurate injection volumes the fuel pressure had to be a constant value over the intake manifold pressure. The pressure relief valve recirculating much of the fuel back to the tank therefore had to be adjustable, with a pipe feeding the manifold air pressure signal back to the ‘other’ side of a spring-loaded bellows system. This was a mechanical solution to what was, in later years, achieved by a simple pressure sensor with a 0-5 V signal and, rather than altering the fuel pressure in the rail, the injector opening times were adjusted accordingly.

Designing the system to cope with both minimum and maximum fuel demand, the remaining issue was the fluctuation of the fuel pressure down the length of the rail. Repeatedly opening and closing the injectors along the fuel rail produces pressure fluctuations within it. Exacerbated by the heavily pulsated incoming flow created by some of the fuel pumps, these pressure pulsations could sometimes cause resonant effects which would alter the fuelling along the rail. Mysterious enough, but when trying to develop an intake manifold to give the best air distribution across the cylinders, this additional effect only served to confuse matters further.

In more refined applications, structure-borne noise was the issue which, happily, identified the source of our problem, the whole issue eventually solved using a fuel pressure damper. In a way, this was similar to the fuel pressure regulator having a spring-loaded diaphragm separating a fuel chamber and air chamber. At working pressure the fuel pressure regulator transfers all the excess fuel to the return line. Meanwhile the fuel pressure damper is so constructed as to maintain the fuel line pressure in response to a fuel pressure ’spike’, effectively altering the volume of the rail but only momentarily. Once this pressure ’spike’ has passed and has been attenuated in the process, the spring-loaded diaphragm moves back to its original position. This means the fuel rail volume is variable and not only accommodates the fuel when the pressure peaks occur but also releases fuel back into the system when the pressure drops.

Although generally not so much a problem now with more refined port-injected fuel systems, the high-pressure pulsations within direct injection designs are reintroducing this challenge to a new generation of engineers.

Fig. 1 - Aftermarket PFI fuel rail

Written by John Coxon

Gasoline, whether it is used for the road or track, consists of a range of different volatile hydrocarbons, each with its own boiling point, beginning at around 25-35 C (initial boiling point, or IBP) and finishing in the region of 180-215 C (the FBP, or final boiling point). The 'lighter' or low-boiling-point constituents of the fuel enable starting from cold and help initiating of the flame front, while as we move to the 'heavier' or higher boiling point ends, these tend to be more dense and deliver the power.

A fuel may therefore be blended to deliver more power, but without any additional conditioning equipment such as fuel heaters, ambient conditions of temperature (and pressure in some cases) have to be taken into account. However, few of us have the luxury of being able to design our own fuel, so what we tend to use is a compromise to give us the ability to start our engines and run them to deliver best performance at acceptable usage levels. It is nevertheless the problem of starting the engine that produces many of the issues we see when ambient conditions move to the extremes or when the fuel 'delivery' system is not up to the task. It is under these conditions that the phenomenon of vapour lock or boiling of these 'lighter' ends occurs.

A good fuel system, whether it is for indirect, port injection or direct injection into the cylinder, will at first scavenge the fuel out of the tank (or tanks), filter it and then present it to the high-pressure pumps, which will pump it towards the fuel injectors. As we pump it through the system, the pressure in the line obviously increases, but so invariably does the temperature. In addition, the heat generated in the injectors - (electric current)2 x its resistance, remember? - is considerable and needs the fuel to cool them and take it away out of harm's way.

So, while only part of the fuel is consumed through the engine, most of what is pumped needs to be directed elsewhere. For this the fuel will pass through the pressure regulator, which is designed to maintain pressure to the injectors, then drop it to a lower pressure, at which time these lighter ends may boil and create fuel vapour in the return line. It is in this return line that the best designs will fit a fuel cooler and have the fuel returned to some kind of collector or swirl pot to eliminate any air or vapour before being fed back into the high-pressure pumps.

Clearly, to avoid this phenomenon of vapour lock, many other precautions to eliminate heat into the fuel lines may have to be taken, but the best one in my opinion is to fit a fuel cooler.

Fig. 1 - A good fuel system

Written by John Coxon

Take the fixed-jet carburettor for instance, a device that young boys and old men used to drool over until they tried to re-jet - or, as we used to say, 'tune' it. Its idle jets, its progression, main jets and accelerator pump offered a myriad of ways of getting it wrong. Take the idle system, for one. With an idle jet, idle mixture adjustment screw and the engine idle speed controlled by the position of the butterfly (yet another screw) eventually it would be possible to get the engine running sweetly at the correct mixture and speed. However, for highly tuned engines the quality of the engine idle and its general roughness sometimes made life a little more difficult.

The main fuelling circuit was equally tricky. For this we had the auxiliary venturi, which provided the signal for the main jet, the air correction jet and the emulsion tube. Air passing through the auxiliary venturi created a negative pressure, which drew the fuel up from the float chamber through the main jet and then down again to be 'sprayed' - or more likely 'spluttered' - into the venturi. At the same time, air was drawn in through the air correction jet, mixed with the fuel via the emulsion tube and passed down with the fuel in the form of a fuel-air emulsion. This air assisted the fuel's atomisation as it entered the engine intake air and, together with the downstream throttle plate, broke up the fuel droplets to give a more homogeneous mixture.

In a simple fixed carburettor the main fuel jet will, if set to the correct setting at low speed, generate a much richer mixture as the engine speed increases. The air corrector jet on the other hand, has a much greater effect at higher engine speeds, and is incorporated to offset this enrichening tendency. The target for adjusting the main jet and the air correction jet is to produce an air-fuel ratio that is as close to constant as possible throughout the engine running range at both wide-open and part throttle.

To many, however, the emulsion tube is a bit of a mystery and from my observations at the time, few really knew much about them, often adjusting only the main and air correction jets. Although a form of mixing chamber, the emulsion also alters the mixture as the amount of fuel is restricted by the main jet lowering the fuel level in it and uncovering more of the air bleed holes. The emulsion tube can therefore increase the mixture enrichment at low speed or weaken it at low speed. It can also enrichen it for slight accelerations by altering the external diameter of the tube. It all depends on how it is designed.

With the complexities of the idle circuit along with those of the main system, it is no wonder that carburettor tuning was better left to the experts.

Fig. 1 - The fixed-jet carburettor

Written by John Coxon

Unlike port-injection systems and the much-lamented (by some) carburettor, direct injection gives complete independence of control of the metering process for both air and fuel. In port-injected systems, part of the injected fuel would land on the walls of the intake manifold or port, and this 'tau' effect could sometimes produce a lean or rich 'spike' during transient engine conditions. By injecting the fuel directly into the combustion chamber, similar to diesel engine practice, not only is this transient condition improved but the displacement effect of the fuel in the intake charge and the improved cooling effect of all the fuel evaporating inside the cylinder (as opposed to in the intake manifold and port) gives better volumetric efficiency.

However, it is not quite that simple. In port-injected systems introducing fuel and air together, controlling the turbulence thereafter and directing it towards the spark plug is well understood and relatively easy to achieve. In direct systems though, not only do we need to understand the flow of the air inside the chamber but also the movement of the fuel during the injection process in order for it to mix with the air and, when directed to the spark plug, burn efficiently.

One way of getting the correct fuel-air mixture to the spark plug is to inject it against the piston crown. Suitably shaped, and hot from the previous cycle of combustion, this so-called 'wall guided' technique evaporates the fuel injected onto the crown, and the tumble of the air in the cylinder takes the charge upwards to the spark plug. Mounted low down and often underneath the intake port, the injector fires the fuel obliquely across the cylinder towards the piston crown as it rises. Shaped piston crowns and hence heavy pistons, together with the limited time in which the piston is in the correct position, make this less desirable for ultimate performance, but access issues in locating the injector around the combustion make this a slightly easier prospect.

When designing a new cylinder head and combustion chamber, the preferred injection method is that of the spray-guided system. This method uses a significantly different approach. Mounted centrally high up in the chamber and next to the spark plug, in spray-guided systems the distance between the injector and the spark plug is very small. Since the fuelling and point of ignition are close together, the charge motion before ignition is of less importance, and instead the ability of the injector to produce well-atomised sprays of a wider angle than those strictly necessary for wall-guided methods is perhaps more important. Furthermore, using lighter pistons and opening up the injection timing window (in not having to wait for the piston) is also an advantage.

Although the straight eight-cylinder Mercedes-Benz M196 of 1955 may have used direct injection, it was not until 40 years later that the technology in gasoline engines is coming back to the fore. With DI allowed in Formula One as of 2014, let's see if racing can take over the mantle and move the technology on.

Fig. 1 - Wall-guided and spray-guided injection

Written by John Coxon

But while the older classic weekend racer may tolerate such fuel contaminants, modern fuel injection systems relying on finer tolerances between mating parts most certainly don't. At this point it is important to appreciate that the fuel supply industry goes to extreme lengths to ensure that the fuel delivered to your tank is as contaminant-free as is reasonably possible.

However, fuel transferred from drums to your racer can easily become contaminated, so a critical part of any modern fuel system (and probably advisable on your classic racer as well) is therefore the fuel filter.

Located on the pressure side of the fuel pump, most modern filters including those in race vehicles are installed within the fuel tank whenever possible. Positioned coaxially and around the fuel pump, to save weight and produce a type of canister, the alternative is an inline filter on its own outside the tank and somewhere much closer to the engine fuel injector system. Either way, the design of the filter is such as to present as large a surface area as possible to the flow of the fuel and thereby minimise any pressure loss. Furthermore, having such a large surface area will also minimise the pressure loss over the life of the unit, enabling filters to be 'fit and forget' over the lifetime of the vehicle.

Made from resin-impregnated papers bonded to synthetic fibre layers, the level of filtration depends on the porosity of the paper and the distribution of the pores within it, but for port-injected systems a mean pore size of 10 microns is normally recommended. For directly injected engines, 5 microns is the rule.

When used in gasoline engines, filters can be either spiral-wound around the central support tube or radially from the central support to the outer diameter and then back to the central tube again, forming a system of pleats running radially around the central tube. In the former, the fuel runs from end to end through the filter in a longitudinal direction in line with the intake and outlet flows; in the latter, when the filter core forms more of a star shape, the fuel flows from the outside inwards, depositing its dirt on the outer side of the paper. When fitting inline units, therefore, the direction of flow is critical, and it is important not to get inlet and outlet ports the wrong way round.

The fuel filter may not be the most important component in the fuel system but if you want to protect the more expensive bits then make sure you install one, and make sure it is installed correctly.

Fig. 1 - Typical inline filter used for port-injected fuel systems

Written by John Coxon

Complicated by the phasing out of four-star 97 RON fuel and replacing it with a 95 RON - so-called 'premium' unleaded - the fuel companies produced a 98 RON lead replacement fuel (LRP) that replaced the octane and trusted enthusiasts to fit hardened valve seats to contain the wear. And when sales of the more expensive LRP fuel decayed a few years later, the compromise fuel was quietly dropped from the forecourt. By then specialist fuels were on the market in small amounts for owners who absolutely needed them, and those remaining used approved lead replacement additives.

These days, however, the issue is one of ethanol, and while the lead issue was all about the IQ levels of children living close to urban motorways, now the issue is one of sustainability and environmental considerations - and a UK government initiative called the Renewable Transport Fuel Obligation. The current standard for UK road fuel, BS EN 228, allows for up to a maximum of 5% of ethanol by volume. In practice, according to the individual refinery and the issues surrounding the transportation of the hygroscopic (water-absorbing) ethanol, not all fuels will contain this level, and fuel sold will contain either no ethanol or about 4-5 % depending on where you live. At this level there is no requirement to identify the ethanol content, and in most cases the general level of the fuel's performance is unchanged in any case.

However, with the advent of the RTFO by some now unspecified time in the future (at one time rumoured to be 2013 but now refuted by the Department for Transport and Rural Affairs) this maximum content is to increase to 10% by volume. And if 5% didn't exhibit any truly discernable difference then 10% most probably will. For modern vehicles with fuel injection and adaptive control designed for modern fuels there are no issues, but for those with carburettors (remember them?) many pre-2000 vehicles and even early direct-injection designs, the problems could be significant.

These issues fall into three main categories - combustion, corrosion and compatibility.

Compared to the value of 14.7:1 for non-ethanol, 'normal' fuels, E10 (as it is now called) has a stoichiometric value of 14.1:1. If the specific energies are broadly the same while the performance given by each will be similar, in those vehicles using carburettors, engines will run about 4-5% lean. For fuel economy this might be acceptable, but if mixture strengths are not adjusted to compensate, valve seat burning could result. The greater volatility of the increased amount of ethanol could also introduce fuel vapour lock in the under-bonnet fuel lines, so these will also have to be routed away from sources of heat.

It has long been appreciated that ethanol-gasoline blends encourage corrosion when exposed to aluminium. The precise mechanisms involved, however, are complicated. Ethanol would seem to be a better conductor of electricity than hydrocarbon fuels, and the galvanic effect of dissimilar metals in the presence of absorbed water may have much to do with it. For racers the answer must always be to empty the fuel tank and run the engine on neat gasoline at the end of the day, but where this is impractical, galvanic action as a result of the proximity of dissimilar metals (such as aluminium-steel, aluminium-brass or zinc-brass) has to be avoided.

Elastomer compatibility also has to be taken very seriously when using the different fuel blends. Seal swell characteristics or the progressive hardening of elastomers can lead to system leaks, and it appears that, as in the case of corrosion, ethanol-gasoline blends are even more aggressive than either pure gasoline or pure ethanol alone. In general, however, the later types of fluorinated seal compounds are more resistant to attack than non-fluorinated types, and the greater the degree of fluorine then the higher the resistance.

In modern fuel systems these compatibility issues will all have been sorted, but being 'green' and protecting your fuel system for those with historic or classic vehicles may not be that simple.

Fig. 1 - Comparison of fuel characteristics

Written by John Coxon

For engine lubricants the traditional choice is mineral oil or synthetic - PAOs (polyalphaolefins) and esters. In the case of transmissions, however, another family of synthetic products, polyglycols, may be commonly found. A member of the same family as used in some ethylene glycol, anti-freeze formulations, polyglycols are ideal for applications with high levels of the shearing action and boundary lubrication - or, if you like, when the relative motion between adjacent components, for instance in touching gear teeth, is small.

In these circumstances the coefficient of friction is very small, much smaller than other oils and possibly even half that of some synthetic products. Polyglycols also have high VIs (viscosity indices): that is, their viscosity changes less with temperature compared to other synthetic oils. The VI of a pure PAO oil might be 170-180 or so. For a polygylcol, the VI could be nearer 230, which is a major advantage at the extreme pressures reached in gear teeth.

While blends of other synthetics or mineral oils can achieve such high VIs using viscosity modifiers (VMs), the large molecules of these VMs can be readily chopped up under the shearing action of the gear teeth and quickly lose their effectiveness. In such cases therefore, the high-temperature viscosity of the oils can quickly fall away. The use of polyglycol technology can avoid such difficulties.

Unfortunately, the factor that makes polyglycols so attractive - that is, their affinity for the metal they are designed to protect - is the same reason for their undesirability. The highly polar oxygen atom, repeated regularly along the length of the chain, is readily attracted to positively charged metal surfaces. However, the hydrogen ions of water vapour are easily attracted to the same atoms, so oils based on polyglycols have a high affinity for water. And water, I don't need to remind you, is the enemy of anything ferrous. Polyglycols should never be used where water may be present and is the reason they can't be used in the engines. Since transmissions are generally sealed from any water vapour in the atmosphere, rust is rarely a problem inside most gearboxes.

Until recently, the other problem with polyglycols has been their incompatibility with other, more traditional hydrocarbon oils. So while many different types of synthetic products may be used in a Formula One gearbox, you can be almost certain that polyglycols will feature to a greater or lesser extent.

Fig. 1 - Lubrication zones in gear teeth

Written by John Coxon

This is not to say, of course, that the diesel cycle cannot be made to produce stonking amounts of power. Simply up the boost pressure (for it must have a turbo or supercharger), tip in large amounts of fuel and watch the torque increase up to the limit where a plume of black smoke billows from the exhaust! The powers produced can be prodigious but a diesel engine is essentially a lean-burn machine that is better working in defence of our dwindling oil reserves hauling all manner of goods through the extreme cold of an Alaskan winter as well as the heat of a Nevadan summer. It is little wonder therefore that diesel fuel specifications are dominated by low-temperature characteristics such as 'cloud point', 'pour point' and something that formulators refer to as the 'cold filter plugging point' when wax solidifies out of the solution. These surely have little relevance to the demands of motorsport.

However, combustion in a diesel engine is altogether different from that of our traditional racing machine, the spark ignition engine, and this is reflected in the fuel used and how it is characterised. In our spark ignition unit the fuel-air mixture enters the chamber, and the stoichiometric homogenous charge ignited by the spark plug burns with a flame front moving across the chamber towards the cylinder walls. The two important parameters for a spark ignition fuel are therefore flame speed and air:fuel ratio.

In a diesel engine combustion chamber, things are very different and much more complicated. Rather than having a flame front travelling across the chamber, the burning process is one of diffusion. Once the intake air has been compressed and heated to above the fuel's ignition point, and the fuel injected, fuel droplets penetrate the chamber and atomise along the way. At this point, and after a short delay, the fuel 'reacts' with the oxygen in the air, and spontaneous combustion takes place at the boundary where the fuel and air meet. Combustion takes place throughout the chamber and, once started, fuel will continue to be injected and the air motion within the cylinder maintained to ensure complete combustion.

But if characteristics such as 'knock' resistance, flame speed and heat of vaporisation are important for gasoline fuels then the equivalents for diesel fuel are ignition quality, density, volatility and viscosity.

The delay between the start of injection and combustion must be kept short, and is minimised by the use of high-cetane fuels (see RET-Monitor, Feb 2011). In diesel-engined road vehicles this cetane number is about 45-55, but the fuel specification used at Le Mans calls for that to be set to a maximum of 62. Not only does this reduce the ignition delay but the cycle-to-cycle variability is also enhanced.

An important consideration in any diesel fuel is its density, since the greater this is the greater the heating value, and heating value equals performance. But the greater the density then the greater the viscosity, and as fuel needs to be metered accurately - both in the volume supplied and its timing - this has to be kept within close bounds. The volatility of the fuel is also important. It's not necessarily directly associated with outright performance but ensures that the engine starts and that, when fully warm, the exhaust 'smoke' limit is not increased, limiting the potential performance of the engine. Whichever way you look at it, however, and just like gasoline, blending diesel fuel is full of compromises. But with Peugeot pulling out of Le Mans for this year, leaving only Audi and the huge advantage still gifted to them by the engine regulations, is there really much of a case for continuing with diesel race fuel?

Fig. 1 - Typical diesel combustion heat release diagram

Written by John Coxon

The modern internal combustion engine is a complex mechanical device, and none more than that used in motorcycle racing. The valvetrain, piston ring and bearings all have their own individual lubrication requirements which, along with others (for example, the ability to keep the engine clean), makes the task of producing an oil with the optimum 'balance' of properties for any particular engine quite tricky.

The oil intended for the MotoGP Honda, for example, will be different from that of the Ducati. This is down to the differences in relative motion of the parts inside the engine and the type of lubrication experienced between them, as well the different temperatures involved. However, motorcycles have other requirements not normally found in other vehicles. You see, they tend to use the same oil in the gearbox as that in the engine, and if the design also includes what is best described as a 'wet' clutch, the oil will have to cope with that as well.

Gearbox oils, while fundamentally similar to engine oils in many ways, need one characteristic not normally found in a combustion engine - the ability to cope with the high contact loads between meshing teeth. In a traditional spur gear system, as the teeth move in and out of engagement the relative motion of the surfaces produces a mixture of high-friction boundary lubrication, mixed and full-film lubrication depending on the speed of the shafts. Narrow gears and high contact tooth pressures, depending on the design of the gear sets, may therefore need certain extreme-pressure or EP additives to prevent the adjacent metal surfaces from coming into contact, which would otherwise cause fatigue damage and subsequent pitting of the teeth.

The most effective anti-wear and extreme-pressure additives still tend to be ZDDPs (zinc dialkyldithiophosphates) although other borate-based ones may also be commonly used. Extreme pressure additives are said to be 'surface active', and attach themselves by polar attraction to the surface of the metal in a similar way to the detergents (and other additives), which are also polar. Since competition oils are frequently changed, they don't generally require the same level of cleanliness as for, say, road bikes and therefore the removal of the detergent can sometimes act as a bonus, enhancing the anti-wear properties in the oil. But depending on the basestock and the additives used, the oil may need to be 're-balanced' to avoid one additive dominating.

The other complication in a motorcycle can be that of the wet clutch. More common in production motorcycles - where, for reasons of reliability, heat generated in the clutch pack needs to be cooled by the oil supply - yet another demand is placed on the lubricant. Passenger car engine oils (PCMOs) these days tend to incorporate certain friction modifiers to reduce boundary-layer friction at the top ring reversal point and at various points on the camshaft. Used within the confines of a clutch system, these additives may interfere with the operation of the clutch, which will struggle to maintain the level of grip required between the plates and rings, leading to clutch slippage. Likewise, with too high a level of friction, the operation of the clutch could be too viscous, making the bike very difficult to ride in confined places by the inexperienced rider. Removing the friction modifier will have implications to the engine lubrication, and so once more the oil may have to be 're-balanced' to suit.

But how many racers will blame the balance of the oil if they don't win?

Fig. 1 - A dry clutch cooled by air rather than oil

Written by John Coxon

For track-based vehicles where ease of cold starting and fuel economy is not that critical, the volatility of the fuel is of secondary significance. Race fuels consequently tend to have much flatter volatility curves than road transport or forecourt fuels, generally reflecting the increased content of individual hydrocarbon compounds within them. So while it may be difficult to start the engine from cold, at the other end of the temperature range the high level of mid-range volatility creates difficulties with hot restarts as the fuel vapour collects in the fuel lines. A minor inconvenience if nothing else, but by far the most critical performance attribute is the fuel's resistance to self-ignition or detonation - more commonly known as 'knock.'

Described in some textbooks as 'the noise transmitted through the engine structure caused by the spontaneous combustion of the fuel/air end-gas in advance of the combustion flame front', the fuel characteristic that best resists this is referred to as its octane. A rather arbitrary measure on a scale of 0 to 100, where 0 is the performance using n-heptane and 100 is that for iso-octane, the octane requirement of an engine varies depending on things like speed, charge temperature, combustion chamber geometry, compression ratio and engine operating conditions - not forgetting, of course, the time available at the appropriate conditions for the events to occur.

Governed by these engine factors the presence or absence of knock will eventually be down to the qualities of the fuel. So while they are made up of several hundred individual hydrocarbon species, each one with its own volatility or boiling point, each component also has its own ability to resist combustion knock. But while an individual compound on its own may exhibit one octane number, when blended into the fuel it will act as if it had a totally different octane value. This value is called the blending number, and often serves only to confuse.

On top of this, simply changing the position of one of the methyl groups in a hydrocarbon molecule - keeping the same number of hydrogen and carbon atoms and the same number and type of covalent bonds - can also have a significant effect. Decreasing the length of a simple hydrocarbon chain and incorporating a methyl group as a side chain has the effect of improving the octane. For instance, normal, straight-chained octane (n-octane) has a boiling point of 126 C but blends into the fuel as if its octane was -17. On the other hand, the branched chain iso-octane, with the same chemical formula and a slightly lower boiling point of 99 C, has an octane of 100. In general, reducing the number of carbon atoms in the main chain reduces the tendency to knock, but at the same time reduces the component boiling point, and so increasing its volatility.

The task of the fuel blender is therefore to take all these properties (and others) to create the correct 'balance' for your engine. And if he (or she) can do that successfully then why not give them the task of sorting out the global economy? Surely they can't do any worse than our politicians.

Fig. 1 - Comparison of actual octane and blending octane for some of the more common fuel hydrocarbons

Written by John Coxon

Despite all this, however, in essence any lubricated system is principally a trade-off between overall friction or power loss and that of protection against wear. An oil that gives good protection may consume large amounts of power, while that intended to provide only the minimal amount (of protection) is highly likely to induce a much higher degree of wear. Thus while we may be looking at low-viscosity oils for sprint racing, for endurance racing the grades used will almost inevitably be much more viscous; for short races we may be looking at SAE 20 grades or thinner while for much longer, perhaps 24-hour endurance racing then SAE 40, 50 or even 60 grades may be the norm.

But selecting the correct viscosity for the start of the race is one thing, ensuring that it is still suitable at the end is quite another. I'll explain.

Consisting of a blend of differing types of base stocks, viscosity modifiers and additives, the viscosity of any oil will change with temperature, the duration at this temperature as well as other mechanical factors such as the amount of shearing action it has experienced and the rate of shear. Some of these are permanent while others are only temporary. In a bearing, for instance, oil 'sees' a very high rate of shear when the viscosity modifying co-polymers, which are present to control the high-temperature kinematic viscosity, lose their effectiveness. Once this shearing action is removed then the oil may be restored to its static, non-shearing value.

More permanent losses to viscosity can also occur as a result of this shearing action, while conversely the action of high temperature and oxidation or nitration may increase it. I'm sure we've all seen that big black 'gloopy' mess if used oils have been left in old cars for many years. The amount to which this 'ageing' happens depends on the quality of the initial base stock, but the natural impact on any crankcase oil will be to increase in viscosity, the higher the temperature, the longer the time.

The greatest impact on the viscosity of any oil, however, is likely to be the level of fuel dilution contained within it. In a normal family passenger car the sump may contain around 1-2% fuel dilution as a result of cold-start and other fuel enrichment processes associated with full-throttle running. In a racecar - particularly one at, say, Le Mans where the throttle may be held wide open for 70-80% of the lap - the fuel dilution factor will be much greater.

Depending on the wide-open throttle fuel enrichment strategy prevailing, reports of 10% or more are therefore highly likely. And even if much of the fuel will boil off at sump oil temperatures, the remaining residue will still cause the oil to fall in viscosity quite markedly, to the point where say a 60 grade becomes a 50, or a 40 grade ceases to lubricate much at all! The combination of shear, whether permanent or not, degradation and fuel dilution can therefore give a viscosity very different from the oil you originally put in.

After 24 hours of high revs and extreme temperatures it might be reasonable to expect that the crankcase oil may be - shall we say - a little past its best. However, with a typical LM P1 engine consuming something like 8 litres or more during the race, the oil in the sump will have been totally replaced by the end with regular topping up.

Looked upon as bad by some, topping up with oil - a mere inconvenience with your roadcar - is much more beneficial than you might have thought with an endurance racer.

Fig. 1 - Typical U-tube oil viscometer for measuring kinematic viscosity

Written by John Coxon

Consisting primarily of methane and smaller quantities of the higher saturates (ethane, butane and so on) together with 'inerts' (mainly nitrogen but also carbon dioxide), natural gas is found in many parts of the world deep underground. In its raw state the gas may also contain water, hydrogen sulphide (H2S) and some of the higher hydrocarbons, but these have to be removed before transportation to prevent condensation at the higher pressures or to avoid corrosion in the pipeline.

When using the fuel in a gas cooker the major concern is that of the gases' Wobbe Index. Defined as the volumetric heating value of the gas divided by the square of its density, natural gas is sold to the domestic user on this basis. Used in an engine, however, this is not quite so relevant. For an engine our main concern is octane requirement or, more specifically, the amount of compression the air-gas mixture can withstand before spontaneous ignition.

While the octane of the primary constituent, methane, is extremely high (around 130 RON, 122 MON) those of the possible contaminants (ethane, propane and butane) can be considerably less. The Motor Octane number for ethane is 101, while those for propane and butane are 96 and 89 respectively. Consequently, for use in an engine, the purer the methane the better.

Even in its purest form, however, natural gas comprising almost 100% methane still has other issues, two of which are its laminar flame speed and the comparatively poor lean flammability. The fact that methane can be readily seen bubbling up through rotting vegetation in ponds and reed beds suggests it is a fairly stable molecule, not readily reacting chemically with other components in the natural world. By the same token it is difficult to ignite, and even when finally burning in the combustion chamber of an engine its flame speed can at times be appreciably slower than gasoline, particularly when running rich. So while its anti-knock potential is great, the potential to extract that performance is compromised by its burning characteristics.

Hydrogen, on the other hand, has characteristics that are almost complementary in that it has a much wider flammability range and is easily ignited. When added to methane in small amounts (up to as much as 5-7%) therefore, it greatly assists the ignition characteristics. Furthermore, with a flame speed of up to eight times higher than methane, once ignited - and particularly in the initial 0-10% burn period - the fuel-air mixture will burn quicker and more completely.

Referred to as HCNG or in some circumstances as Hythane, the use of such mixtures could eventually bring us one step closer to that other Utopia - that of the hydrogen economy.

Fig. 1 - Laminar flame velocity of methane compared to a typical gasoline mixture

Written by John Coxon

The real concern for the formulators is the increasing exploitation of the two-car draft when two cars, literally running nose to tail, bumper to bumper, can gain so much extra speed and effectively run away from the pursuing pack. The car in front runs in clean air, and engine coolant/oil temperatures of about 220-230 F (105-110 C) will be typical. The car behind, however, almost totally excluded from any form of cooling air, will see temperatures much nearer 300 F (150 C). Running ten to 12 laps at a time in this way, on and off throughout a 500-mile race, puts serious demands on the engine oil.

Drafting of course is nothing new; we are all familiar with multi-car crocodiles snaking their way around the superspeedways. Positioned only a few feet apart as engine temperatures rose, drivers would jink to one side to pick up cool air before ducking back under again and regaining their tow. But as the advantages of tucking right underneath the rear spoiler of the car in front have become more readily appreciated - despite the potential damage to the engine - engineers have been forced to look at ways of making the engines last. For the lube oil supplier this means making sure the oil is at its optimum viscosity for the entire race and with the minimum of degradation, despite these temperature changes.

As a rough and general rule, the higher the oil viscosity in an engine, the greater the protection it offers, but this comes at the expense of friction. Too low a viscosity and the oil film breaks down, and wear will result; too high and you lose power. In formulating engine oils, therefore, the idea is to run at the lowest possible viscosity compatible with an acceptable amount of wear.

Running at 300 F for long periods rather than 230 F, oil blenders need to re-blend their normal PAO/ester blends for the superspeedways to take this into account. One way of doing this would be to re-blend the oil by increasing the amount of viscosity modifier (VM) to limit this fall-off in viscosity with temperature, a property known as Viscosity Index or VI. Added to the blend in small amounts, VMs (sometimes referred to as polymers, co-polymers or Viscosity Index Improvers) cause the oil to become more viscous than it would otherwise have been as the temperature increases.

These long-chain molecules can, however, break up under the action of the high shear rates, when loss of viscosity results over time. Oils running at optimum viscosity at the beginning of the race would therefore be seriously 'thinner' at the end of it. So the challenge for the oil blender is to ensure that his blend has the minimum change in viscosity for only the slightest amount of co-polymer, relying more on the properties of the base oil.

Used many years ago to classify motor oils but no longer widely appreciated, you could say that in NASCAR at least, high-VI oils are once more back on track.

Fig. 1 - Variation of viscosity with temperature, and the effect of viscosity index of two 0W-20 oils, each with a kV100 of 8.7 cSt

Written by John Coxon

The fact that water and ethanol are fully miscible in each other with no limits is by now I hope evident to all, and naturally therefore, when exposed to the humidity of the atmosphere, such water will inevitably migrate into the alcohol without limit. When high levels of ethanol are used, for example E100 (100% ethanol) or E85 (85% ethanol), the presence of comparatively small amounts of water can be tolerated - indeed, industrial-quality ethanol can have as much as 4.9% water in it simply as a result of the manufacturing process and the energy required to refine the product to a anhydrous higher level.

But when only small amounts of the stuff are added to the likes of E10 or E15 - as in the case of many modern-day gasoline blends or when the octane-enhancing benefits of MTBE (methyl tertiary butyl ether) have been replaced by less polluting ethanol - the presence of water can be a problem. Let me explain.

For all practical purposes, while water is fully miscible in ethanol, it is totally immiscible in gasoline. That's not quite true since most gasoline will have a very small amount of water in it - maybe 0.01% or 0.02% by volume - but, by and large, water will not dissolve in the fuel. In a fuel tank full of gasoline, once the water has exceeded this level the surplus water will fall to the bottom of the tank and hopefully out of harm's way. In the presence of an oxygenate such as MTBE, the maximum amount (of water) absorbed can increase slightly to about 0.06-0.07% at ambient temperature, and any excess will fall to the bottom of the tank again.

However, if this oxygenate is replaced by a less polluting ethanol, the mixture will quite readily absorb as much as 0.4-0.5% of water, and since ethanol is highly hygroscopic, it will do so. As a guide, a drum of fuel left unsealed in an atmosphere of 70% relative humidity will absorb this amount in three months. Once the water exceeds this level, 'phase separation' occurs and the water-ethanol mixture, not just the water, will sink to the bottom of the tank, effectively denuding the gasoline of its octane. The more water that is absorbed, the less octane in the remaining fuel. I'll leave you to imagine the rest.

So while a smidgeon of water may enhance your drinking pleasure, in your fuel tank it can create so many more issues with your engine.

Fig. 1 - Variation of the maximum water absorbed in fuels with temperature

Fig. 2 - Water - good or bad? Model of the hydrogen bond: the 'spare' electrons in the outer shell of the water oxygen atom (in pink) are attracted to the electron-deficient hydrogen molecules of the alcohol (in white)

Written by John Coxon

However, irrespective of the duty cycle of either engine, in any reciprocating engine there are three main critical regions - the bearings, piston rings and valvetrain - with different relative velocity, temperature and pressure within them. The engine lubricant (base stock, viscosity modifiers and additives) will therefore be the result of the compromises in blending the oil for these three zones.

The most critical property in any lubricant is its viscosity, and this is highly dependent on the temperature, the rate of shear and the pressure within it. Variations in viscosity with temperature are described by the Vogel equation, a principle that we all use when draining engine oil, say, as the oil drains more freely at higher temperature. The shear rate - the rate at which the successive layers of oil slide across each other in the lubricant film - has been described by MM Cross in his paper "Rheology of Non-Newtonian Fluids", while the variation in viscosity with pressure can be approximated using the Barus equation.

Combined together, and with knowledge of the various coefficients of each equation, a reasonable assessment of the viscosity within a given lubricant can be calculated in each of the systems above. Using this, a minimum oil film thickness in each of the zones can be modelled and the frictional forces therein derived. By blending a number of oil components together in the virtual world, a numerical model of an oil can be developed at a basic level before the actual blending and physical bench testing in the real one.

While this might sound straightforward - and, dare I say it, even pedestrian - nothing could be further from the truth. To generate more accurate data, the precise geometry of the components under study needs to be known. For instance, in the case of the piston ring pack, we need to know the gas pressures on either side of the piston rings, the temperature of both ring and liner at the point of contact and any possible effects as a result of any restrictions to the flow of the lubricant by the oil control ring.

In the bearings this will mean the temperature around the bearing and the elasticity of the components under the loads generated. As both temperature and pressure change as we travel around the bearing, so too will the oil viscosity. Indeed, while exact figures may be difficult to produce, the real benefit of such analysis is to investigate trends that will point towards improved design.

However, from an engine performance perspective, the real interest may lie in the total friction within these units. In a typical touring car at high speed using techniques described above, the total engine friction was estimated at 7.5 kW using an SAE 15W/40 oil. Of this, 55% came from the bearings, only 39% from the pistons and 6% from the valvetrain. Reducing the lubricant viscosity lowered the friction in the bearings and piston ring pack but brought about an increase in the friction in the valve train.

Using similar techniques at the maximum speed of a Formula One engine, the optimised figure was a total of 65 kW, of which 46% came from the piston assembly and 38% from the bearings. The remaining 16% (or 10 kW) was attributed to the valvetrain.

The approach to minimising engine friction in Formula One, or indeed any engine, would therefore appear to be elementary. For high levels of boundary lubrication (when relative sliding velocities are low), high viscosities should be used together with friction modifiers. Where an engine has less boundary lubrication (and therefore higher relative velocities), lower viscosities may be more acceptable. All fine in theory but in any complex lubrication system to establish the minimum viscosity acceptable requires regular sampling of the lube and subsequent screening for wear metals.

Fig. 1 - Variation of oil viscosity with shear rate and temperature

Written by John Coxon

There was a time in the early 1990s of course, when, because of the increasing demands of gasoline emissions legislation, the only form of high-performance motoring for the future envisaged by some was likely to be diesel-powered. But this had more to do with the state of the gasoline car passenger vehicle emissions legislation relative to that of diesel than any desirable property of the fuel. Indeed, I think most will agree that the current domination of 24-hour endurance racing is due solely to a quirk of engine regulations favouring compression ignition than any inherent benefit of the fuel. But while the fuel comes from a similar source and is slightly higher up the hydrocarbon fuel volatility curve, in terms of combustion the properties used it could not be so very different.

In gasoline, the quality of the fuel is denoted by its Octane Number. A measure of its anti-knock property, once injected into the airstream and the mixture compressed, the higher the Octane Number the greater the resistance to spontaneous burning, and the charge can be ignited at the precise time required by means of the spark plug. In a diesel engine the situation is totally different. Here, instead of being injected into the port, the fuel is injected at high pressure into the combustion chamber and, rather than delaying the onset of combustion, since there is no spark plug present, the intention is to encourage burning of the fuel rather than delay it.

In describing the property of the fuel, instead of Octane Number, compression-ignition engine engineers refer much more to the Cetane Number of the fuel, which is a measure not of anti-knock but of ignition delay in the engine. But in some sort of convoluted and strange way, the two - Octane and Cetane - are actually linked. For although a fuel designed for a diesel engine would not be suitable for use in a gasoline unit, a fuel that has a high Octane Number will most likely have a low Cetane Number, and vice versa.

Surprising or not to the layman, the proof for such a statement can be found in the properties of the individual components or classes of components that make up each of the fuels. Aromatic compounds, for instance those that include the benzene ring, have a very poor ignition quality in diesel engines, while their performance improvement characteristics are renowned in gasoline units. Alcohols - methanol and ethanol, well known for their use in race engines - also offer poor ignition performance in diesel engines. For use in compression-ignition engines, alcohols have to be turned into fatty acid methyl esters (from methanol) or fatty acid ethyl esters (from ethanol), a process to be found generally in the manufacture of biodiesel.

But contrasting fuel properties or not, with only Audi and Peugeot so far having developed truly high-performance compression-ignition diesel units, it is difficult to see what the future is for diesel fuel in sportscar racing. Given the sheer expense of developing suitable engines, and the advantage over gasoline steadily eliminated by changes to the rulebook, it is difficult to see any other manufacturer entering the fray. And as we all know, once successful, manufacturers always eventually pull out, leaving the independents to fight among themselves.

I have no crystal ball but, like the turbo years in Formula One, maybe the era of the whispering giants in sportscar racing and the use of anti-, anti-knock fuel is slowly coming to an end.

Fig. 1 - Relationship between Cetane and Octane

Written by John Coxon

Although experiments to use such fuels in reciprocating race engines have never met with much success, there is a 'rocket fuel' that has found favour in Top Fuel drag racing - nitromethane. With the chemical formula CH3NO2 it has some interesting properties.

The composition of hydrocarbon fuels means they burn only within a narrow range of air:fuel ratios (the ratios of air coming through the engine intake to the fuel consumed by mass for complete combustion). The oxygen required for combustion comes solely through the engine intake system, and therefore needs highly efficient engines with excellent volumetric efficiencies to achieve maximum power output. Nitromethane, however, already contains 52.5% of the oxygen needed for complete combustion, so the volume of oxygen in the air coming through the intake is arguably not quite so critical. Indeed, the more fuel that is supplied, the more oxygen that is imbibed and the more power is potentially available.

To put it into numbers, the stoichiometric air:fuel ratio is about 1.67:1; this compares with that of traditional hydrocarbon fuels that are somewhere nearer 14.7 :1. The downside with nitromethane though is that for every kilogramme burnt, only 11.3 MJ of heat is released; for most hydrocarbon fuels it is nearer 44 MJ. Nevertheless, if we can tip in up to nine times more fuel we can still generate more than twice the energy while burning the same amount of air. And with a heat of vaporisation of about twice that of the hydrocarbon fuel to create even denser air charges, it all begins to sound very encouraging. If we also take into account the faster flame speed of nitromethane, in theory requiring less ignition advance than other oxygenated fuels such as methanol, then nitromethane would seem to be the ideal fuel for motorsports.

It gets better. The real beauty of nitromethane is its ability to burn and therefore release its heat in the complete absence of air, so the emphasis on engine volumetric efficiency is arguably non-existent. However, when it burns, it forms a mixture of carbon dioxide, carbon monoxide, hydrogen, nitrogen and water - instead of just carbon dioxide and water - which can combine to form a mixture of nitrous/nitric acids, so the exhaust fumes are not only very unpleasant but poisonous too. So while the breathing of the engine may not be impaired, that of the driver most likely will be!

Also, because the fuel is so explosive - for want of a better word - it can only be used for short periods of time. In practical terms, high combustion temperatures of up to 2400 C limit its usefulness when lower compression ratios, rich mixtures and retarded ignition are necessary to counteract combustion knock.

Generally blended with methanol up to a maximum of 90%, nitromethane is therefore quite worthy of its name - 'rocket fuel'.

Fig. 1 -The nitromethane molecule

Written by John Coxon

But in the world of fuels, simply removing the 'if' or 'si' in French, produces a totally different connotation. RVP, you must understand, stands for Reid Vapour Pressure. A measure of the pressure of the fuel vapour at 100 F (37.8 C), this is just one of the parameters used to measure the volatility of a fuel. The RVP is therefore the vapour pressure of the fuel obtained when the air:liquid ratio is 4:1 at the given temperature, and measuring it cannot be simpler.

The procedure involves filling a metal chamber with a chilled sample and connecting it to another, this time an air chamber of precisely four times the volume. The equipment is totally immersed in a water bath at 100 F and the whole lot agitated until a constant pressure reading is obtained. Unlike some of the other measurements of volatility - such as distillation curve or vapour-liquid ratio - the RVP is simple, quick and therefore cheap. But after going through all this rigmarole what does it actually tell us?

In the world of road transport engines the RVP is intended to give an indication of cold-weather driveability. Too low a vapour pressure and the engine may be difficult to start, even with pre-heating as in a race unit. To high a vapour pressure and the engine could be difficult to start when hot and suffer from a phenomena known as 'vapour lock'.

A common issue with competition engines, vapour lock is more likely to affect carburettor systems under hot restart conditions. Here vapour formation in the fuel pump normally created as a result of 'heat soak' can interfere with the pumping, making the fuel-air mixture too lean and difficult to re-ignite. In fuel-injected engines the issue is essentially the same but the higher fuel rail pressures involved tend to delay the onset of the problem to the point where vapour lock is only rarely experienced.

In Europe, service station forecourt fuels are separated into eight volatility classes, and the volatility available from summer to winter across the continent will depend on the prevailing weather conditions expected in that part of the globe at that time of the year. Buying a pump fuel in one place and then moving on to race somewhere warmer or colder can therefore introduce fuel system issues.

The RVP is much more controlled for race fuels, and will remain the same summer or winter - but only provided the containers are tightly sealed! However, if travelling to warmer climes it might be as well to inform your fuel supplier so that he can best advise not only on your choice of fuel but how best to transport it.

The RVP is only one measure of fuel volatility though, and applies to the 'light' ends of the fuel 'boiling' off at the lower temperatures only. In terms of overall fuel volatility there is a bit more to it than that.

Fig. 1 - Fuel barrel

Written by John Coxon

In the real world of engine lubrication and the engines we use in our everyday life, these requirements are strangely very similar, but for one significant difference - that is, exhaust emissions or, to be more precise, exhaust system protection.

In the US and many other parts of the world, but not Europe, the automotive passenger car industry sees the introduction in 2011 of the ILSAC (International Lubrication Standardization and Approval Committee) GF-5 engine lubrication specification. Replacing the GF-4 specification, which came into effect around 2004 MY (Model Year), this latest in oil performance targets is designed to push the boundaries of passenger car factory-fill oils to new limits.

Designed to assist future CAFÉ (Corporate Average Fuel Economy) standards for passenger cars with higher oil quality, GF-5 formulations are also intended to be better at coping with the effects of the ethanol now seen increasingly in forecourt fuels. With increased levels of Group III base oils commonly found in many European ACEA specifications, more low-friction, 0W-20 formulations are therefore likely to be seen in service stations under the guise of API (American Petroleum Institute) SN service-fill specification from early next year.

Apart from introducing new tests such as those for seal compatibility, improved E85 rust protection, better sludge protection and better piston cleanliness, the main aim of GF-5 has been towards fuel economy (as mentioned) and emission system (particularly catalyst) protection. For many years now, the presence of phosphorus in engine oil has been blamed for a gradual fall-off in catalyst performance throughout the life of a vehicle. Phosphorus, as many of you will know, is a major constituent of ZDDP, a highly effective anti-wear agent necessary for cam protection under boundary-layer and mixed-lubrication regimes.

For GF-4, the level of phosphorus in the oil was reduced to 0.08 % by weight. For GF-5 this limit is retained, but the amount of volatile phosphorus - that which is most likely to be 'boiled off' and find its way into the catalyst - is now limited. So while the amount of phosphorus is still reduced, at least there is some form of control over the amount of it retained in the oil, hopefully to lubricate the cam and piston ring later in its life.

While this might give a glimmer of hope for better cam wear protection on passenger cars, it is of little interest to the average competition engine builder. With no requirement for catalyst protection technology, a market now exists for low-viscosity 0W or 5-20W oils where the engine bearing size and clearances are acceptable but with much higher levels of anti-wear, ZDDP-type protection.

So while GF-5 factory-fill/API SN service-fill oils may be ideal for the new generation of fuel-efficient passenger cars, the time has surely come when oils for everyday transport and those for competition have diverged.

Fig. 1 - Comparison of ILSAC GF-4 and GF-5 oil performance

Written by John Coxon

Nevertheless, there are times when race organisers wish to restrict the amount of fuel carried on board and therefore stipulate a maximum tank capacity to which all competitors must comply. In such cases, in order to finish the race and assuming refuelling is not allowed (or even desirable), it will be necessary to eke out the fuel supply in some way or another.

The most obvious way is to trim back on the fuel map and instead of, say, running at 12:1 air-fuel ratio, nudge it back a little to 12.5:1 for example. Little power will be lost and, with luck, you may make it to the end of the race. At the same time, however, you may finish last!

But there is more than one way to obtain fuel economy. Take the blend of fuel used, for instance. A typical gasoline fuel used for motorsports consists of up to 200 or more individual hydrocarbon species, each of which has its own heat of combustion, octane rating, heat of vaporisation and specific gravity. Furthermore, each of these compounds - whether it is a member of the paraffin, olefin, cycloparaffin or aromatic group - will have a temperature at which it boils such that the whole mixture we call gasoline will have a volatility (or boiling point) curve similar to the one shown in the figure here.

For most fuels this is in the form of a gentle S-shaped curve, which rises progressively from the left-hand side to a maximum at the right-hand side. Once fully blended, this volatility curve is a design attribute of the fuel, with the smaller and lighter hydrocarbons boiling off towards the lower portion of the curve and the heavier, longer-chain molecules towards the top end. At the lower end of this curve, if the volatility is too high, the engine may be difficult to start when cold; if it's too low then the vapour produced in the fuel line can make restarting when hot almost impossible. At the higher end of the curve, the idea of fuel economy introduces itself.

At this higher end, and with everything else being equal, moving to hydrocarbon components with a greater density will pack more heat energy into a given volume. In practice, the heavier hydrocarbons from the aromatics group with the generalised formula CnH2n-6 are used and, so long as the overall heat of energy and octane rating aren't unduly affected, will enable the engine to run slightly leaner.

Of course, it is important to ensure that the overall fuel density doesn't exceed that specified or that the overall volatility characteristic expressed in the RVP (Reid Vapour Pressure) becomes too low - or indeed that one of any number of constraints put upon the fuel is not transgressed. With this level of complexity it is easy to see why gasoline fuel blends can become so complicated and so different between suppliers.

The real skill in fuel blending therefore is to balance the volatility curve to suit the characteristics of the engine and yet still keep within the fuel regulations.

And in the hunt for fuel economy I would start with the fuel.

Fig. 1 - A typical gasoline volatility curve

Written by John Coxon

To claim it is a revolution is no hyperbole. For while in both sets of regulations - the 2009 version and its updated 2010 successor - the purpose of the Article is "to ensure that the fuel used in Formula One is petrol as this term is generally understood", the thinking behind the changes goes much further than anyone can possibly imagine even now, some four months into the start of the season.

Since the regulations have changed quite significantly, one might be forgiven for thinking that the fuel raced at Silverstone this year will be radically different from that raced at the same venue last year. To a certain extent this might be true. Since in-race refuelling has now been banned, cars will of necessity have to come to the grid with much higher fuel loads, so the density of the fuel will be more critical.

In addition, since the fuel is now in the car, it can't be kept cool until the pit stop, and its bulk temperature in the vehicle tank will inevitably be higher, particularly later in the race. A higher temperature means more chance of it vapourising in the fuel system unless this is altered at the blending stage.

To counteract these effects the make-up of the fuel, compared to 2009, could be described as 'significantly different'. But a fuel that is 'significantly different' by any stretch of the imagination is hardly one that is revolutionary. After all, it's still gasoline isn't it?

Well, let me explain further. Having said that the fuel at Silverstone this year could be little changed from 2009, the real revolution is in the thinking behind the hydrocarbon components allowed and the changes that could bring about to all of us in terms of sustainable fuel technology at the gas station forecourt.

From 2004 and up to 2009, bio components were introduced to the fuel in response to the ecological conscience of the participants. The FIA wanted to prove to the world with the minimum of effort that the formula was pushing forward the boundaries of fuel technology in terms of lessening the use of fossil fuels.

Introducing a minimum content of 5.75% (by mass) in the form of oxygenates "derived from biological sources" went some way towards defusing any undesirable criticism and shadowed political events in the EU, which at the time were pushing towards even higher levels for forecourt fuels. More important, it bought the FIA and the fuel suppliers time to review their overall objectives regarding the type of fuels used and dovetail them to the possible future demands on the hydrocarbon fuels business. It is no secret that the fuel suppliers look to Formula One as a means of showcasing fuel technology, developing new tools and methods in what is after all a critically important area.

So while other classes of racing moved towards blends of inefficient ethanol, Grand Prix racing saw a future using advanced fuels not too dissimilar to the ones we use today but made up from components that were produced from non-fossil fuel sources. Thus, as rule 19.4.5 in the 2009 regulation simply states that "a minimum of 5.75% (m/m) of the fuel must comprise of oxygenates derived from biological sources", for the 2010 version this has been opened up to include "hydrocarbons and oxygenates". With the relaxation in the distillation characteristics, namely E70° C, E100°C and E150°C, and the abolition of the 'PONA' table, the future could see some very interesting new hydrocarbons derived from biological sources - which of course could eventually benefit us all.

It may be revolutionary but nobody yet knows how much.

Fig. 1 - The gas chromatogram is designed to 'fingerprint' the fuel but can't distinguish the source of those components.

Written by John Coxon

But LPG is nothing new in tin-top racing. Twelve years ago, Vauxhall introduced a prototype LPG-fuelled car to its racing series for the V6 Vectra SRi saloon. With rumoured alterations to the cams and exhaust compared with those required by the championship, and a prototype fuel injection system, the car flattered to deceive. Ineligible for championship points, the car undoubtedly had a competitive edge, but the question was always one of the modifications to the engine rather than the characteristics of the fuel. Although it was an interesting technical exercise, the project was more an example of marketing than pushing forward the boundaries of engineering.

In 2004 the fuel surfaced again, this time in the hands of the amiable but under-funded BTCC team of John George. And while the Vectra six years earlier used a prototype liquid fuel injection system, the BTTC-spec Honda Civic used a gaseous system.

LPG is one of those 'funny' fuels. It's generally considered a by-product of the oil and gas industry, as an impurity in natural gas or a product of petroleum refining, and depending on where you live in the world, its composition may vary. A mixture of mainly propane and butane, and in some cases propene and butanes as well, it is usually stored as a liquid but at ambient temperatures quickly reverts to a gas.

This characteristic can be very useful in an engine, and can be used to cool the incoming charge when injected into the port as a liquid. If injected as a gas, however, this advantage is lost and the displacement of air by the expanded fuel coming out of the evaporator leads to a further power loss. So the John George Civic was at a power disadvantage right from the off, and it was no wonder that the team ran in gasoline trim for most of the season.

In the UK, LPG (or Autogas as it is sometimes known) is about 99% commercially pure propane, and while its re-introduction to BTCC is clearly for marketing reasons, a number of technical challenges did face the development team. Funnily enough, one of these was getting the liquid fuel to vaporise fully inside the cylinder.

With a density of 0.54 kg/litre, liquid propane needs about 25% greater flow rate to supply the same amount of calorific value of fuel to the engine. Injecting this greater amount - and despite the slightly lower latent heat effect of the fuel of about 360 kJ/kg as opposed to 400-plus with gasoline - the fuel was accumulating on the cylinder walls and washing away the lubricant. Injecting the fuel well upstream of the traditional intake port position, as well as one or two other 'tweaks', eventually solved the problem but not before a couple of engines had been lost on the dyno.

With LPG currently selling in the UK at about £0.57 ($0.87) per litre at the pump and gasoline priced at an average of £1.20 (about $1.83), the interest in LPG by the typical motorist is bound to rise, despite the differences in calorific content and hence consumption in miles per gallon. But while marketing can have a strong influence on the vehicles and products used in the sport, we must never allow the organisers to forget that it is the engineering that makes it all work.

Fig. 1 - Gaseous injection manifold of the Honda Civic

Written by John Coxon

Pump gasoline you see, is a composite product made up from hundreds of different hydrocarbon compounds all carefully blended to a given performance specification. This means that it will have a minimum resistance to engine 'knocking', the octane quality (RON and MON) as well as maximum levels of hydrocarbons like olefins (alkenes) and aromatics. The octane quality is present to ensure that road going engines don't destroy themselves through premature ignition while the limits on the types of hydrocarbons has been regulated as a result of on-road vehicle emissions legislation. Introduced to underpin a minimum engine performance such a specification is not truly appropriate to a race engine. In Europe EN228 calls for a minimum RON of 95 while UK specification BS7800 raises this to 97 RON for Super unleaded. In addition to this, road fuels can vary from batch to batch, week-to-week and though suppliers will try to ensure a level of consistency, the amount of each hydrocarbon component will nevertheless still vary. During the year, summer or winter pump fuels will also be changed to suit the ambient conditions. In the winter the volatility will have to be increased because of the low air temperatures and this will require a greater proportion of lower boiling point constituents each with a differing octane value.

Race fuels on the other hand are altogether different. To start off with, race fuels will be specified to a maximum octane value. Regulated by the sporting authority in the UK these are 100 RON and 89 MON but in other countries may be higher. The fuel sold will therefore be much closer to the maximum limit than any minimum specified. Furthermore, every batch of fuel summer-to-winter, batch-to-batch or week-to-week will be precisely the same giving the same ignition properties irrespective of where and when purchased. Engine ignition mapping can therefore run much closer to the limits of combustion or if you prefer a higher level of safety on the base level tune. Because they tend not to incorporate the higher end components, racing fuels will also tend to burn faster and cleaner producing fewer in the way of carbon deposits keeping the engine cleaner and delivering more power for longer. If this wasn't enough, these faster burning characteristics mean less ignition advance and the probability that more of the fuel will burn inside the combustion chamber giving more power than simply releasing its heat into the exhaust port. With more power and a cooler running engine, tell me, what more could you ask for?

Even considering any marginal extra expenditure, you simply know it makes sense.

Fig. 1 - Volatility curve of a typical racing fuel compared to the limits for a pump fuel. Note how the curve is quite flat in the middle part indicating perhaps that the fuel has a higher proportion of fewer hydrocarbon components than would be present in a pump fuel.

Written by John Coxon

However it is as well to remember that the conditions in the combustion chamber of the CFR engine, that rather old and some might say antiquated unit used to evaluate both RON and MON, are totally different to that in a racing engine. This CFR engine, designed in the1930s, runs at 600 rpm for RON testing and a mind numbing 900 rpm for the MON test procedure. Any readers thinking that a fuel in an engine running at 10,000 rpm plus will behave in a similar way is perhaps rather optimistic. As it is a mixture of 84% toluene supplemented with 14% n-Heptane while producing the test results comparable to a more normal 102 RON fuel, is so much more resistant to detonation in the combustion chamber of a high speed racing engine. Anyone involved in the blending of fuels for racing will surely be aware of this. For the rest of us, the sight of a Formula One car powering out of the old chicane at Silverstone with a rooster tail of a plume of black smoke emanating from the exhaust pipe will be a sight we will never forget! 1500 bhp out of a 1½ litre engine running at 5 bar boost was really something! However non PC it may be, that's my kind of alternative fuel!

Fig. 1 - Fuel comparison.

Written by John Coxon

Mention the word 'biobutanol,' however and you will be surely met by any number of blank looks and quizzical expressions. But before long there is every chance that biobutanol may be on the lips of every tree-hugging politician in the land if recent developments in fuels technology come to full fruition.

To the chemists amongst you butanol, or to use its chemical formula, C4H9OH is the fourth in the series of simple alcohols: methanol, ethanol, propanol and now butanol. Currently derived mainly from the fermentation of sugars in organic feedstocks such as sugar cane/beet, corn or wheat, the future suggests that with only small changes, the same second and third generation manufacturing plants can be used as those for bioethanol.

With four carbon atoms in the molecule, biobutanol has a much higher energy content per litre than ethanol. To the farmer this means a more valuable crop and hopefully higher prices. To the fuel chemist, an energy density of 29.2MJ/L compared to that of 19.6 MJ/L for ethanol and 32 MJ/kg for gasoline is a much more practical proposition for blending into current day fuels. So as bioethanol fuels require bigger fuel injectors and larger fuel tanks, biobutanol blends, in theory at least should be able to use the same gasoline systems currently used. Being less corrosive than bioethanol such systems wouldn't necessarily have to be so highly specified against corrosion, while at the same time it is also less aggressive towards elastomer seals.

But as they say, you can't have everything and while the Octane numbers of the substance are broadly comparable with that of a high performance gasoline, it is the relatively poor (compared to ethanol) heat of vaporisation, which begins to let the side down. At around 0.43 MJ/kg compared with 0.36 MJ/kg for gasoline and a whopping 0.92 MJ/kg for ethanol, the opportunity to keep charge temperatures to an absolute minimum, especially in turbocharged engines, is lost.

The big advantage of biobutanol in a strange way however is its low vapour pressure. Let me explain. Gasoline fuels are blended from a wide range of hydrocarbons to produce a fuel that gives easy starting in cold temperatures but good fuel economy at higher ambient temperatures. The fuel therefore has a range of individual fuel boiling points, which make up a typical fuel volatility curve. When blended into gasoline on its own, biobutanol could seriously distort this curve producing poor cold start performance. However, our old friend bioethanol has quite the opposite effect when only small amounts of the product (less than 10%) can have a major influence on the light end of this curve. By combining the effect of the bioethanol with that of biobutanol this distortion can be minimised and therefore biobutanol is said to have co-synergy with gasoline fuels containing bioethanol.

At the moment the use of biobutanol as a blended component into gasoline is still in its infancy. While it is still only early days there is every chance however that the technology may see the continued use of gasoline type fuels well into the foreseeable future. And while we won't necessarily get all the benefits accrued from ethanol based fuels, as far as most users will be concerned, nothing will have changed.

Is that not good news for all this New Year?

Fig. 1 - Primary butanol models.

Written by John Coxon

Designed to cling to the metal parts when the oil film disappears, the EP additive is an essential element to all gearbox oils. Until recently these additives have been based on Zinc DialkylDithioPhosphate or ZDDP and is the substance that gives off that somewhat sweet pungent smell which hangs around any gearbox or differential. However, in gearboxes, as in engine crankcase oils, the use of these is now not always desirable and so the race is on to find other less active sulphur-phosphate compounds. One solution has been the addition of solid lubricants – graphite and molybdenum disulphide, to the mix. Here these dry powder layers when added to the oil slide over each other reducing friction and preventing wear, but issues with each - graphite settling out in the oil filter and the intense colour of ‘black moly’ staining everything it comes into contact with, make them not particularly attractive for widespread use. Another approach and one developed only recently, has been the introduction of inorganic fullerenes, or IFs for those in the know.

Discovered in 1985, fullerenes are a family of carbon allotropes – molecules composed entirely of carbon in the form of hollow spheres, ellipsoids, tubes or planes. When spherical, they can often be referred to as ‘Buckyballs’ after Richard Buckminster Fuller whose shell-shape or lattice-structure they resemble. Indeed the shape of the F.A. football is based around a C60 fullerene comprising of a mixture of hexagonal rings interspersed with pentangles. The beauty of the fullerene is that apart from being very stable, having no or very few ‘lose ends’ and with the ability to ‘nest’ rather like onion rings, they work on a nano level as opposed to the considerably smaller molecular level of other solid lubricants. If you like, fullerenes act more like ball bearings, the covalent bonds linking the inorganic components together while the weak Van-der-Waal bonds allow each layer to slide over the next. Chemically unreactive and therefore highly resistant to oxidation, fullerenes also do not degrade much over time. With characteristics such as this, and impressing all those teams having used gearbox oils containing it, expect to see more of this nanotechnology in the future.

And what can be more important than life and death, but the durability of your gearbox?

Written by John Coxon.

Specific power, specific weight, manoeuvrability, cost of manufacture, ease of maintenance, durability, NOx emissions and even fuel consumption in the case of smaller engines, can be far superior to its less controversial 4-stroke brother, but why then don’t we see more of these engines in competition?

With one strike for power and one to wear it out as opposed to one of power and three to wear it out for the 4-stroke, used in most forms of karting and in some smaller engined motorcycle racing, the basic two-stroke engine can appear visibly dirty to the casual observer. High amounts of toxic unburned fuel and an occasional slight tinge to the exhaust suggest a high level of particulate matter when running on traditional gasoline fuels. But if that fuel were changed to a more environmentally benign ethanol or blend of ethanol, then there is increasing evidence to suggest that not only will these objections be overcome but that when fully optimised, two stroke engines can produce even more performance.

Essentially, it is all about charge density. In a 4-stroke naturally aspirated engine, any benefit given up by the fuel is mainly due to its increased octane number or resistance to detonation. This allows an increase in compression ratio and more efficient combustion. In a two-stroke engine however, not only can we use the increased octane available to create more efficient combustion, but the process whereby the intake charge is compressed in the hot crankcase can be used to extract the maximum benefit out of the latent heat of evaporation which at 904 kJ/kg is something like 2½ - 3 times that of gasoline. Not only that, but a greater cooling effect in the crankcase produces an increased charge density, which in turn expels a greater pulse of gas into the exhaust pipe. This greater pulse is consequently reflected back and used to pull in even more intake charge into the cylinders. And since the performance of a two-stroke is all about the exhaust system and these pulse-tuning effects, the effect on the performance of the engine using ethanol is significant, to put it mildly.

In blind back-to-back testing with gasoline on a 125 cc kart, after doing a few modifications to the engine and changing to a blend of ethanol, the driver reported a more responsive engine with much better pulling power from a lower engine speed. At the end of the session asked if he wanted to revert to the earlier specification his comments were un-publishable but needless to say the offer was refused. And while running rich, the amounts of carbon monoxide produced will most probably be of the same magnitude to that of gasoline, there is some evidence to suggest that the un-burned hydrocarbons typical in a basic 2-stroke, may be less toxic.

It’s not all good news, however. The corrosive effect of alcohol fuels on brass or aluminium in the carburettor or engine crankcase has to be carefully managed. As well as running on regular gasoline for a period at the end of the day and regular changes in crankcase lube, consideration has to be made on the type of lube oil added to the fuel. While many of the more usual two-stroke oils may not be suitable, those containing high levels of esters are claimed to be much more miscible.

With thermal efficiencies approaching 55% in some types of sophisticated 2-stroke designs, rather than trying to ban them in motorsport, shouldn’t the authorities be trying to encourage the use of more sophisticated versions? A change to a more environmentally benign fuel such as ethanol, could therefore just be the starting point.

Written by John Coxon.

After coal, it will surprise many to hear that natural gas is the most abundant fuel around with almost twice as much proven reserves to that of crude oil. With a Lower Calorific Value (LCV) of around 47 MJ/kg (compared with gasoline at 43 MJ/kg) and an octane (RON) rating much nearer 130, it has always surprised me that the fuel hasn’t been taken seriously as a possible high performance fuel. The news that motor sports diesel pioneers VW/Audi, entered two of its CNG-powered Sciroccos in this years Nurburgring 24 hr with one car coming home 17th overall and first in the AT (alternative powertrain) class, is finally making people sit up and take notice. While the actual fuel used was HCNG, a mixture of hydrogen and compressed natural gas, the project nevertheless and at long last demonstrated the potential of this otherwise underrated utility.

If we assume that natural gas consists of predominately methane with small amounts of the higher ‘saturated’ alkanes (ethane, propane, butane, etc.) together with an even smaller amount of inerts (principally nitrogen) then the number of hydrogen atoms to every carbon will approximate to four. Comparing this to an average carbon to hydrogen ratio of gasoline of C1H1.78 and a few further calculations will prove that natural gas (or for the purposes of this calculation - 100% methane) will produce about 19-20% less carbon dioxide per kilogram of fuel compared with most gasoline fuels. At a time when most other methods to reduce gasoline consumption (and hence carbon) struggle to achieve only small savings, you will begin to see the validity of the cause.

However, life is never that simple or easy and while most gasoline engines can be run on natural gas with a few modifications, there is a fundamental objection to using the fuel in an internal combustion engine. Gasoline engines are generally optimised to run on a particular grade of fuel. Be that ULG95 for road transport or 102RON for competition, the compression ratio and combustion system will be fully optimised to ensure efficient and repeatable combustion for that fuel chosen. While the quality of these liquid fuels will be blended to consistent levels of octane, for natural gas coming from a gas utility company through the mains gas supply, the requirements are different. Customers for this type of fuel are more interested in their energy or heating value than in its detonation characteristics and in this case the control factor is that of the Wobbe Index and not octane number. Defined as the volumetric heating value of the gas divided by the square root of its specific gravity, this index is proportional to the heating value of the quantity of gas that will flow through an orifice in response to a given pressure drop. Thus to ensure a constant calorific value in response to fluctuations in composition, propane or other alkanes may have to be added to maintain the heating value of the mixture. Since the Motor octane of methane is very high (around 122) but that of propane is only 96( butane 89 and ethane 101), with the variability of the gas, maintaining the Wobbe number at a given value will invariably lead to wide changes in octane value, which would be unacceptable from an engine calibration perspective. This shouldn’t stop us from using the fuel but simply highlights some of its limitations.

Low carbon fuels? Been using them for years.

Written by John Coxon.

All gasoline fuels contain additives. Whether this is to reduce oxidation in the form of inhibitors or metal deactivators to minimise corrosion, these are added in small amounts to ensure that the fuel reaches the user in the best possible condition and causes as little damage to the supply network as is physically possible. Other performance-boosting additives may also be introduced and might include anti-ORI (octane requirement increase) or anti-pre-ignition. Spark-aiding additives, to assist lean burn combustion, might also be present as well as several other types designed to ‘help’ combustion along the way.

During the course of history the type and amount of these additives might be subject to change, partly as a result of changes in technology or for environmental considerations or both. One of these, Tetra Ethyl Lead or TEL was the subject of much debate twenty years ago.

Primarily an anti-knock additive, TEL was the most effective way of increasing the octane level of any gasoline. Decomposing in the combustion chamber at temperature, TEL forms a catalytically active cloud of metal oxide particles, which interfere with the rapid chain combustion that constitute detonation or engine ‘knock.’ Progressively limited by statute and then finally phased out altogether, the octane provided by the additive was steadily replaced first by improved refining introducing more aromatics / olefins to the mix but later, saturated hydrocarbons and various oxygenates.

One of the side effects that made TEL so attractive was its tendency to leave lead deposits in critical areas of the engine. In particular the lead used to accumulate on both intake and exhaust valve seats and act as a lubricant as the valve rotated on its seat. As a direct consequence of this, engine manufacturers would often choose to run fairly ‘soft’, unhardened valve seats which, when lead was removed from fuels, created prodigious valve seat wear issues. Easily cured at the manufacturing stage with hardened valve seats, in some engines, where the valves were too close together, a demand for aftermarket additives providing this missing lubrication was swiftly generated. The main constituents of these replacement additives are well known. Consisting of compounds containing sodium, potassium, phosphorus or manganese, these are now allowed by the motor sporting authorities, which is where much of the controversy begins.

In order for these to mix evenly in the fuel and ensure more accurate dosage, these compounds will need to be pre-mixed in a hydrocarbon ‘carrier’. This has to be clean burning and influence the fuel in the least way possible. In most cases toluene is the ideal choice but with blending octane numbers well in excess of the base fuel, inevitably some level of octane enhancement of the fuel is likely. At the dose rate required this is likely to be around 2 octane numbers. Furthermore while clearly phosphorus compounds cannot be used in conjunction with catalyst-equipped vehicles there is increasing evidence to suggest that all metallic fuel additives can be harmful to many exhaust gas sensors and control equipment. Thus while another additive, Methyl cyclopentadienyl Manganese Tricarbonyl or MMT may give excellent protection at the stated dose for older engines, its continued use in newer vehicles on a regular basis for its Octane increase alone, apart from being unnecessary, is likely to lead to long term damage of any engine closed loop control system. Unless engines are modified to take advantage of the extra octane available - for instance by increasing the compression ratio or advancing the ignition (where desirable), there would seem to be little point in using such additives.

The messages coming out from the UK Motor Sports Association in particular, is also confusing. While sodium, phosphorus, potassium and manganese are allowed for the purposes of lubrication, for some reason, manganese alone is limited to 0.005 gm/litre, a concentration which is only a fraction of that required for adequate valve seat protection. Maybe the MSA are heeding warnings from the vehicle manufacturers, but allowing its existence at such a low level even to historic vehicles, is surely confusing.

Whatever the additive at whatever the dosage, fuel additives will always be a contentious issue.

Written by John Coxon.

Methanol, a bio-fuel I here you say? Well, strangely enough in 2009, yes! I know that methanol always used to be manufactured from the fossil fuel, natural gas and when converted into syngas (or synthesis gas – a mixture of carbon monoxide and hydrogen) it was only a short hop over a catalyst to get methanol. Today however, as a result of the increasing demand for bio-diesel, a by-product of making this product can be used for methanol manufacture. As any eco-warrior will know, bio-diesel results from the transesterification of vegetable or animal fats but a significant by-product of this process is glycerine. Used in small amounts in soap and the food industry (and given the additive number E422 – remember last month?), the industry, if you will pardon the pun, is now literally awash with the stuff. Fortunately however, there is another use for it – to make methanol or, since it comes from natural sources, bio-methanol.

To avoid wasteful disposal issues the glycerine is shipped from the bio-diesel manufacturers into the refinery, used as a feedstock for the syngas and the resulting methanol shipped out in the same tankers that brought it in. Thus wherever you see bio-diesel there’s a good chance that methanol or should we now say ‘bio-methanol’, won’t be all that far away. And as we already know, methanol in its non-bio form and as a fuel for racing has been around for very many years.

The simplest of the alcohols, methanol has many advantages over all other hydrocarbon fuels. With a higher specific energy than gasoline, it’s comparatively low heating value is more than offset by its very low stoichiometric air:fuel ratio (6.45:1). Furthermore, and while gasoline gives its maximum power at 20% enrichment, methanol will take up to 40% more to reach its peak! Better anti-knock characteristics enabling higher compression ratios (up to 16:1) and with charge temperatures kept low as a result of the very high heat of vaporisation, the only real downside to using the fuel, assuming you can cope with the extra tank capacity needed, is its highly corrosive nature. With this easily controlled by the inclusion of corrosion inhibitors, the future for methanol, bio-methanol that is, must be looking bright.

In an ironic twist of fate I always thought it unfortunate that after 40 years of using methanol, albeit derived from natural gas at the time, in 2007 Indycars transferred their allegiance to the apparently more environmentally friendly bio-ethanol. A marketing ploy or possibly even a political move designed to support the mid-west farmers will always be debated, but in a funny sort of way with its issues surrounding the diversion of grain crops towards fuel production and the increased price of food as a result, the environmental benefit or as I prefer to put it, the ‘balance of green’ has arguably swung back towards methanol again.

Politics and motorsport are frequently found together, but getting it right is often simply a matter of waiting.

Written by John Coxon.

The tale is a sad one but one completely avoidable once you understand the situation. In recent years the oil designed for passenger cars has changed in formulation. In response to the need to reduce precious metal loadings in modern vehicle catalysts and at the same time increase their durability, certain key components have had to be reduced or replaced in the latest generation of PCEOs (Passenger Car Engine Oils). Critically for gasoline engines, these key compounds are those containing sulphur and phosphorus which rather regretfully are the same components found in one of the most effective anti-wear oil additives commonly in use, that of ZDDP, or zinc dialkyldithiophosphate.

In any crankcase engine oil the formulation has to contain certain anti wear additives to protect the surface of the engine components when the relative motion of the mating parts falls below a certain threshold level. At this point, should no other means of protecting the surface be available, there is a great risk that the oil film is insufficiently thick to prevent surface to surface contact when hydrodynamic lubrication progressively falls back into the boundary type. Metal to metal contact will take place and result in friction, high temperatures and ultimately the failure of the parts. Of all the components in an engine, the cam - flat tappet interface, whether in the form of the direct acting mechanical bucket in overhead cam engines or to a slightly lesser extent mushroom tappets in pushrod engines, is that most at risk. In these circumstances the oil entrainment velocity, the relative velocity of the cam-tappet contact point on the tappet with respect to the velocity of that point on the cam, drops to zero and then reverses at slow speed. If this relative velocity is insufficient to generate a ‘wedge’ of oil between the two, then these surfaces will rely on the presence of the highly polar ZDDP molecules to keep the surfaces apart and protect them.

Up until recently typical zinc levels of 1200 ppm were sufficient to protect the flat tappet and once the temperature of the oil was above 55-60 deg C, these components would become highly surface active, attracting themselves to the nearest metal surface. Thus while the viscosity of the oil decreased these additives would protect all metal surfaces and enhance the property of the lubricant as the temperature increased. Come the arrival of ‘Emission System Protection’ oils in Europe (ACEA C1- C3) and GF-4 fuel economy oils in the US, the level of zinc has been regulated to much lower levels of 500-900 ppm depending upon the viscosity grade. And while other forms of anti-wear have been developed they are not necessarily as effective nor perhaps rather more importantly as cheap as their earlier counterparts. Designed for and proved on the more modern, low friction valve trains of the latest generation of passenger cars, these low zinc, phosphorus and sulphur oils should not be used on older flat tappet engines.

While in the UK the ‘Protect and Survive’ booklet of the 1970s was somewhat fatalistic in the event of nuclear war, the current threat to our older flat tappet valve train is easily parried by using the appropriate engine crankcase oil for the task. In general most oils of the right grade for your engine and designed specifically for competition will satisfy this requirement.

Written by John Coxon

Today in the motoring and motor sports world we have our own E numbers: E10, E20, E100 and of course, E85 where the number following the ‘E’ stands for the percentage of ethanol (normally bio-ethanol) contained in a gasoline blend. And in many ways the confusion surrounding them is much the same. To start off with, E85 is now referred to as Ed85. There is no change in composition, it is just that the authorities would prefer to emphasize that the fuel is not intended to be for human consumption, the ‘d’ standing for ‘denaturised.’ Fortunately gasoline conforming to the appropriate ‘local’ standards, as is stated in the various fuel regulations, is generally considered a ‘denaturant’ in its own right and so an 85 percent blend of ethanol in gasoline, with or without the denaturant, amounts to much the same thing.

Now I don’t know what your gasoline is like or whether it conforms to ‘local’ standards or not, but I would imagine even the thought of sitting in your favourite chair with your favourite cigar and a glass of RON102 doesn’t truly appeal. But believe it or not, if your gasoline doesn’t conform to these ‘local’ EU standards, then under these same standards you are ‘heavily recommended’ (read ‘legally obliged’) to add either methyltertiobutylether (MTBE) and/or ethyltertiobutylether (ETBE) and/or isobutanol/tertiary-butyl alcohol (TBA) to satisfy this element of the rules. These compounds are normally considered to be oxygenates in any other context to do with fuels, but any or all of them can be used either singly or together to create your blend, but apparently not isobutanol on its own. Isobutanol on its own, with a much higher boiling point than ethanol and of limited solubility in water, can be quite easily separated out and so it is ‘advised’ to use this in combination with one or more of the others.

In the United States where MTBE has been phased out (because of groundwater contamination concerns) these denaturant regulations refer mainly to straight refinery products: LSR gasoline – a low octane poor quality product distilled directly from the crude, Raffinate – the waste liquid from a refining process or ‘natural gasoline’- generally referred to as a refinery ‘blend stock’ product. However, whichever way you look at it, and no matter which side of the ‘pond’ you sit, the authorities seem to have beaten us to it and the only way we can have that celebratory drink is by standing at the bar and paying our liquor tax in the normal way.

Ethanol for your car or ethanol for your jar – the two in many ways are so very similar and yet so very different in many other ways. But I still prefer mine out of a glass.

Written by John Coxon.

However, this was carried out with very little fanfare, and few even knew about it.Oxygenates, no matter what their source, are good for anti-knock behaviour and hence power, and have been widely used in pump fuels ever since the earlier anti-knock additive, tetraethyl lead (TEL) was phased out in the 1990s. Initially introduced in the form of MTBE (methyl tertiary butyl ether) – CH3.O.C4H9 – in response to US and now EU legislation, they are now being replaced by their bio-derived counterparts.The high octane, bio-oxygenates of most interest are ethanol and ETBE (ethyl tertiary butyl ether). Whereas MBTE was produced from fossil fuel, ETBE is manufactured from ethanol, which itself comes from bio sources and therefore ETBE can claim to have some ‘green’ credentials. However, it is the blending of ethanol into gasoline that raises a number of issues.The first of these is that the addition of only a very small amount of ethanol (vapour pressure – 16 kPa) to a gasoline mix (maximum vapour pressure - 60 kPa) actually increases the vapour pressure of the overall fuel. Explained by the formation of hydrogen bonds, when using small quantities of ethanol, this is contrary to what one might intuitively think. Thus for our hydrocarbon fuel of say, 60 kPa vapour pressure, the introduction of 2% ethanol will increase the vapour pressure to 67.3 kPa. At 5% the vapour pressure will increase still further to a maximum of just over 68 kPa. After this, the vapour pressure will begin to fall again as this hydrogen bond effect is overtaken by the reduced vapour pressure of the additive such that at 5.75%, the maximum stipulated by the FIA, the mixture will fall to 68 kPa. Thus a fuel, which was initially within the limits in the rules (45.0 – 60.0 kPa), could now be illegal once ethanol is added.However, unlike EN228, the legal requirement for forecourt fuels, Article 19 of the Formula One regulations does take this into account allowing a maximum of 680 hPa vapour pressure when a minimum of 2% or more of bio-ethanol is used.Another issue in the use of ethanol is that it distorts the volatility curve. In particular, it pushes up the E70 value (the percentage of the fuel evaporated at 70 degrees C) and means that careful re-blending with a low volatility blend stock is necessary. Other than pushing the fuel outside of the designated bands, this is not a major issue within Formula One. But for commercial fuels, problems with cold starting and drivability could be encountered unless this is corrected at the blending stage.

Another major issue with ethanol, perhaps not widely publicised, is its affinity for water. Most could not have failed to notice that water mixes ever so well with your favourite alcoholic tipple, well, the same applies to ethanol when mixed into gasoline. Water in very small amounts will almost inevitably be found in any liquid fuels made on an industrial scale. In the case of gasoline, depending on its temperature and composition, this water will dissolve only slightly (less than 0.01%) and the rest will accumulate at the bottom of the tank. When mixed with ethanol, however, the water is completely miscible and may cause the ethanol-gasoline blend to separate out with an alcohol-enriched water layer at the bottom of the tank effectively denuding the gasoline–ethanol mix above it. Again not a problem with Formula One fuels, when contamination can be eliminated with careful handling and storage, but for commercial forecourt fuels the risk is ever present.Thus while the teams struggle to develop their so-called ‘green’ KERS systems, the fuel suppliers just quietly get on with the job.