Miniature Engine Fuels

FUEL TECHNOLOGY is a modest, not very highly publicised, branch of knowledge, with the result that the average Aeromodeller probably knows less about the fuels he uses than about any other aspect of his craft. This is greatly to be regretted since engine performance—and engine life—depends not only on engine design and workmanship but also on the characteristics of the fuels used in them. These notes have been prepared for the guidance of modellers who like to experiment with fuel mixtures of their own—to help them to experiment intelligently without undue waste of time and materials—and to assist them in judging the suitability of commercial brands of fuel for whatever purpose they may have in mind. No attempt has been made to write a "Formulary" or to review existing commercial fuels. What has been attempted is a concise and simplified account of the properties and functions of the major fuel ingredients, and an outline of the basic scientific principles to be followed in working out the design of a fuel for any particular purpose.

Before it is possible to proceed to the formulation of a satisfactory "Diesel" or "Glow fuel", it is necessary to be familiar with certain fundamental properties of fuel components such as Flash Point, Heat of Combustion, S.I.T., etc., and a short explanation of the more important of these terms is given below.

EXPLOSIVE LIMITS. When the vapour of an inflammable liquid is mixed with air the mixture will only burn if the concentration of vapour lies between certain limits known as the "Explosive Limits". These limits vary considerably for different liquids, as shown in TABLE I.

TABLE I—Explosive Limits.

Taking Methanol as an example it can be seen that if the concentration of methanol vapour in the air is less than 5.5% the mixture will be too weak to fire, whilst if it exceeds 21% the mixture will be too rich.

FLASH POINT. "Flash Point" is a measure of the flammability of a liquid. If a little flammable liquid is placed in the bottom of a small metal cup it will give off vapour into the airspace above it. If this concentration reaches the lower explosive limit the mixture of air and vapour will "flash" if a small flame or spark is brought above the cup. If the liquid does not vaporise readily (paraffin oil for example) it may be necessary to warm it until a certain critical temperature is reached at which enough vapour is given off to form the explosive mixture. This temperature, below which ignition will not take place, is known as the "Flash Point", and varies widely for different liquids, as shown in TABLE II.

TABLE II—Flash Point.

SPONTANEOUS IGNITION TEMPERATURE Also known as Self Ignition Temperature, Auto-Ignition Temperature, and S.I.T. for short. This is the temperature at which a mixture of inflammable vapour and air will ignite without the application of a flame or spark. S.I.T. is totally unrelated to the Flash Point, and should not be confused with it. TABLE III gives some typical values.

TABLE III—Spontaneous Ignition Temperature.

It can be seen that paraffinic hydrocarbons (paraffin, diesel oil etc., are mixtures of these), and ethers, have low S.I.T.'s whilst "aromatic" hydrocarbons from coal-tar like benzene and toluene, and the alcohols, have very high values.

HEAT OF COMBUSTION. The Heat of Combustion, also known as the "Calorific Value"—is the total amount of heat liberated when a given quantity of a substance is completely burned. It is, therefore, a direct measure of the total intrinsic energy, and hence of the available power, of a fuel.

Some approximate values are recorded in TABLE IV, from which it can be seen why, for example, an alcohol fuel requires larger carburettor jets than petrol; more fuel must be flooded into the cylinders per stroke in order to give a comparable power output. The figures also make it clear why alcohols run" cooler" than hydrocarbon fuels, and are therefore favoured for racing engines.

OCTANE VALUE. Pure Iso-Octane is a very good antiknock fuel for spark ignition engines, since it has a high S.I.T., whilst Pentane, with a very low S.I.T. is a bad fuel. Other fuels are compared as regards performance with mixtures of iso-octane and pentane and thereby given an "Octane" rating. If the fuel is as good as iso-octane its Octane Value is l00, whilst if it is only as good as a mixture of equal parts iso-octane and pentane its Octane Value is 50.

CETANE VALUE. This is a method of assessing the values of diesel fuels by comparing their performance in a test engine with mixtures of different proportions of the excellent diesel fuel cetane and the very poor diesel fuel methylnaphthalene. Cetane and Cetene Values may also be calculated indirectly from the specific gravity and Aniline Point of the fuel, but this method is not applicable if "dopes" are present. A high Cetane Value means a low Octane Value, and vice versa.

IGNITION LAG. When a mixture of a diesel fuel vapour and air is raised to the Self Ignition Temperature, there may be a considerable delay before the explosion actually takes place. This time interval is known as the "Ignition Lag" and for smooth running should be small. The running characteristics of a poor fuel may be enormously improved by reducing the ignition lag by making small additions of certain "dopes". This must not be overdone since too short an ignition lag causes detonation, etc.

 

TABLE IV—Calorific Value

TYPES OF LIQUID FUELS

Liquid fuels for internal combustion engines are of two fundamentally different types, namely those to be fired by spark or hot-wire ignition and those designed to ignite under the heat of compression alone, without the application of a spark or other local hot-spot. The former fuels, of which petrol is the commonest example, should contain a low-boiling fraction (the "light ends") of low Flash Point to ensure starting from cold, but must have a high S.I.T. to prevent firing taking place under compression alone before the spark passes. The second type of fuel, for use in Diesel engines, need not possess a low Flash Point but must have a low S.I.T. It follows that a good petrol fuel will be a bad diesel fuel-and vice versa.

Miniature Diesel Fuels

The diesel fuels used in road transport vehicles are fairly high-boiling fractions from natural petroleum consisting mainly of certain types of "paraffinic" hydrocarbons. Such a "gas oil" has a Spontaneous Ignition Temperature around 250°C. and when forced into the cylinders in finely atomised form will fire satisfactorily under the high-temperature conditions prevailing in these very high-compression full-scale engines. But they will not ignite in a model "Diesel" unless it is hot, and to enable miniature compression-ignition two-stroke engines to be started it is customary to add a proportion of Ethyl Ether, which combines the phenomenally low S.I.T. of 188°C with very wide Explosive Limits. Since the miniature "Diesel" is a two-stroke engine, lubricant must also be incorporated in the fuel. Finally, to ensure smooth even running it is often advantageous to include a small proportion of a further component, the "dope". It is worth while to study in some detail the functions and properties of these four vital compounds.

(1) The Paraffinic Base-Fuel.

This is the main ingredient of the fuel. Its function is to provide most of the energy of the fuel, and it should therefore possess high Calorific Value and low S.I.T. Reference to TABLE III will show that, with the exception of certain ethers. the only readily available substances with relatively low S.I.T.'s are the paraffin hydrocarbons—which fortunately also possess very high Calorific Values. Ruling out individual pure hydrocarbons like pentane, hexane, heptane, etc., on the grounds of expense, this virtually narrows down our choice of base fuel to PARAFFIN OIL, COMMERCIAL DIESEL OIL and special HIGH CETANE GAS OIL FRACTIONS, if available. There is little to choose between paraffin and diesel oil, the latter having its higher viscosity and greater "oiliness" to recommend it. It can be seen, partly by reference to TABLE III, that the addition of petrol, benzene, toluene, naphthalene, turpentine, white spirit, or in fact any of the fantastic materials that have from time to time been recommended, must of necessity make the fuel worse, because of the high S.I.T.'s of these substances. Their use to "deaden down" the detonation of the ether is a case of two wrongs failing to make a right: a fuel that needs deadening down has got far too much ether in it.

(2) The Lubricant.

The lubricating component of the fuel may be any good quality lubricating oil, either mineral or vegetable. The only limitation imposed by vegetable oils like Castor Oil is that, alone, they will not blend with paraffin base fuels; castor oil can be used only in a fuel ready-mixed with ether, which will keep all the components in solution. There is scope for experimenting with different grades and qualities of oil.

With regard to the quantity of oil to incorporate in the fuel, this again is a matter for experiment. Many miniature engine fuels are grossly over-lubricated, with the result that they are unnecessarily messy in use, and also require more ether than they otherwise would. In designing a diesel fuel it should be borne in mind that the oil has one function only—to provide adequate lubrication—and that it should not be expected to burn, to moderate the explosive tendencies of excess ether, or to do anything else. A two-stroke motorcycle engine runs on the road for long periods at a time under much greater (and varying) load than any model engine, and with considerably greater bearing and piston speeds, yet seldom does the percentage of lubricant in the fuel exceed 7.5%. It is desirable in formulating a model diesel fuel to increase this proportion for the following reasons:-

1. A new engine may have tight spots and require excessive lubrication till it is run-in.
2. In a very old, or badly made engine, the piston may be a poor fit in the bore, so that a fairly thick viscous fuel is needed in order to seal the compression, and
3. The manufacturer must allow a reasonable safety factor.

Point 2 normally affects only the ease of starting: once the engine has been started it will usually continue to run perfectly satisfactorily even on a very thin fuel. With old engines starting can usually be facilitated by injecting a drop or two of lubricating oil through the ports.

For a normal fuel for use in a run-in engine in good condition, oil percentages in the region 30% to 50% are unnecessarily high. If the aeromodeller experiments with proportions of oil in the range 12%-20% for racing blends and 20%-30% for general-purpose and running-in fuels, he will not go far wrong. Diesel oil based fuels tend to require rather less than those blended with paraffin.

(3) Ether.

Apart from its low S.I.T., which enables it to start easily, and its wide Explosive Limits which ensure that throttle (needle valve) settings are not critical, ether is a bad diesel fuel. It has a considerably lower Calorific Value than the paraffinic base fuel (kerosine) and it detonates or "knocks" badly. Excess of ether means correspondingly less base-fuel in the formulation, and hence a fuel of lower calorific value than need be, whilst its detonating propensities when present in excess cause diesel knock and impose undue strains on the con-rod. Ether should, therefore, be added to a diesel fuel for one purpose only, namely to make the engine start. Just enough for this purpose should be added—and no more. 30%-35% is excessive, and modellers are recommended to experiment in the range 20%-30%. It cannot be overstressed that the function of the ether is solely to bring about easy starting; it should not be expected to usurp the function of the base-fuel.

There seems to be some confusion regarding the grades of ether suitable for use in fuels. Ether is manufactured from ordinary ethyl alcohol, two molecules of which join together, with the elimination of water, thus:-

The process is usually carried out by heating the alcohol with concentrated sulphuric acid, which absorbs the water formed—which is why the product is sometimes called "sulphuric ether". The ether which distils over is washed free from acid, purified, dried and re-distilled. It therefore contains no acid whether it is sold as "Anaesthetic Ether", "Ether '720", "Ether B.S.S. 759", "Sulphuric Ether" or "Ether Meth.". All these materials are, effectively, the same thing; and if properly manufactured are all harmless to model engines. The '720 refers to the specific gravity of the product and shows the substantial absence of water; B.S.S. 579 refers to the appropriate British Standards Specification laying down the standard of purity; "Ether Meth." indicates that the ether was not manufactured from pure ethyl alcohol but from methylated spirits, which contain a few percent of methanol—this will give traces of methylethyl and di-methyl ethers in the product, which are not harmful. Anaesthetic ether is made from pure alcohol and usually contains a proportion of deliberately added alcohol, and sometimes other additives, to prevent peroxide formation on storage. It is more expensive than other grades and, if anything, is slightly less suitable for fuel work.

The di-ether, Methylal, with the chemical formula CH₃O-CH₂-O-CH₃, may be used partly or wholly to replace ethyl ether in certain specialised fuel formulations. The higher ethers Amyl Ether and Butyl Ether are too high boiling to be valuable alone, but may be used mixed with ethyl ether. Isopropyl Ether, unlike the straight-chain ethers above, has a very high S.l.T. and is not suitable for use in diesel fuels. It is a possible ingredient of glo-fuels.

(4) Dopes.

There are a number of well recognised "dopes" which may be added to diesel fuels, best known of which are

• Ethyl and Amyl Nitrites
• Ethyl and Amyl Nitrates
• β-Chloro-ethyl Nitrate
• Paraldehyde
• Various organic peroxides like Tertiary Butyl Hydro-Peroxide, Di-Tertiary Butyl Peroxide, etc.

The choice of dope is usually determined by price and availability.

(the modern usage of the word "dope" rather distorts the meaning here. What the author is referring to are fuel additives which are intended to improve its performance this is rather the opposite meaning to today where "dope" is a personal additive which decreases performance.)

The function of the dope is to reduce "Ignition Lag" and thereby give smooth powerful running. Very little dope is needed for this purpose, the precise amount depending on the particular fuel formulation, and is a matter for experiment in each case. Seldom is more than 3% required, and modellers would be well advised to start with about 1% of dope and gradually increase, by not more than 0.5% at a time up to a maximum of about 2.5%, until smooth even running is obtained—and then to STOP. This is a case of "a little of what you fancy does you good"—but a little bit more can play hell. Dopes should be used solely for the purpose described above and should under no circumstances be used in excess to assist starting. They do, indeed, lower S.l.T. somewhat, but their effect in this direction is most marked with the first few per cent. and then falls off very rapidly It should be remembered that nitrate dopes are, in effect, high explosives and that when they burn they generate nitrous fumes. An over-doped fuel requires the compression setting of the engine to be drastically reduced as the engine warms up it sets up unnecessary strains in the engine, and it is corrosive.

A proprietary brand of fuel will be a carefully balanced blend of ingredients with the correct amount of dope; no attempt should be made to "improve" it by further dope additions.

Following the basic principles discussed above, and bearing in mind that each component of the mixture has its own specialised part to play in the performance of the final fuel, it is now possible to set about designing a good diesel fuel for a particular engine or for a specific purpose. A good running-in fuel for new engines and for general purpose flying would look something like this :

whilst a Racing or Competition fuel might well be :

If the fuel is of the ready-mixed variety all the ingredients are mixed together, and the lubricant may be castor oil. But if the fuel is to have its ether added immediately before use, only the first three components are mixed in each case; in which event mineral lubricant must be employed.

Starting with either of the above basic formulations as a guide, the ideal fuel for a particular purpose and individual engine can readily be worked out on the test bench by modifying the components of the appropriate formula a very few per cent at a time until optimum performance is obtained. It should be borne in mind that the perfect fuel for one engine may not be ideal for another with totally different design characteristics and the really scientific flying enthusiast will study the individual fuel requirements of all the more important engines in his "armoury". It should also, of course, be appreciated that different fuels may require different starting and running settings—and the careful experimenter has to develop a considerable amount of patience.

RUNNING-IN ENGINE TEMPERATURE.

It follows from the increased proportion of base-fuel and the reduced proportion of ether that a "racing" fuel will run hotter than a running-in or general-purpose fuel, because of its higher Calorific Value. This relatively high-temperature running has been known to worry some modellers, who sometimes attribute it to frictional heat arising from under-lubrication. Any well-formulated racing fuel is, by its very nature, bound to run hot—and it is advantageous that it should do. The efficiency of operation of the internal combustion engine increases, within reasonable limits, with increase in temperature of running, hence the modern practice of cooling full-scale aero engines with ethylene glycol (b.p. 198°C), instead of with water (b.p. 100°C).

It is clearly not the wish of the reputable fuel manufacturer to ruin his customers' engines, and his branded fuels will have undergone extensive tests on a range of engines before being launched on the market. There should, therefore, be no cause for uneasiness in using well known proprietary brand fuel. But if the modeller is still anxious, it is suggested that he feel, not the cylinder head where combustion of powerful fuel is taking place, but the crankshaft main-bearing. If this remains moderately cool he need have no fear of a seizure.

WARNING. In fairness to the manufacturer, as well as in his own interests, the modeller should. of course, be careful only to use a fuel for the purpose for which it is intended. A "Competition" or "Racing" mixture is, as its name implies, intended for high-speed work, and the manufacturer assumes his customer will not be expecting to develop maximum power and revs with a new engine straight out of its box. A "Standard" or "Running-In" fuel should always be used with new engines, which should first be run on the bench for some time with an oversize propeller. After the engine has loosened up it should be run for another half-hour or more with a standard prop., still on the same type of fuel. Only after proper running in, and after a fair amount of work, should peak output with racing fuels be attempted.

Spark Ignition Fuels

The usual fuel for an ordinary spark plug engine is "Petrol". A motor spirit which is a simple "cut" from the distillation of natural petroleum—what is known in U.S.A. as a "straight-run gasoline"—consists mainly of paraffinic and naphthenic hydrocarbons boiling over the range 40°-190°C. Because of its high paraffinic content it has a fairly low S.l.T. and tends to "knock" or "pink" badly in a modern high compression automobile engine. Its low Octane Value is raised by either of two methods. The first is to incorporate a small amount of a dope having precisely the opposite effect of a diesel dope, in order to suppress pre-ignition, i.e. to raise the S.l.T. Lead Tetra-Ethyl is pre-eminent for this purpose, although when used alone it has the disadvantage of giving hard deposits of lead oxide inside the engine, Modern "Ethyl Fluid" contains ethylene dibromide to minimise this trouble. The second method is to enrich the straight-run fuel by additions of benzene (benzole), toluene, other hydrocarbons of high Octane Value (high S.l.T.), or alcohols. The high octane hydrocarbons may be obtained from coal-tar distillation or from the gasoline itself by various high-pressure high-temperature "cracking" processes known as "aromatisation", "preforming", "alkylation", etc.

Alcohol blends containing methyl and ethyl alcohols may also be used satisfactorily in spark ignition engines. They perform best in engines with high compression ratios and are therefore most suited to motor-cycles and racing cars, where their high S.I.T.'s ensure immunity from "knocking". Such blends are, of course, eminently suited to miniature spark ignition engines, castor oil lubricant being incorporated for two-stroke engines. The relatively low calorific values of alcohol blends, and their higher price, makes their use far ordinary purposes uneconomic if hydrocarbon fuels are available. But the increased volume c.f. fuel that has to be flooded into the cylinders in order to obtain comparable power output tends to keep the engine moderately cool at high speeds, an important consideration with racing engines. The calorific value of methanol blends may be increased by replacement of part of the methanol by Methylal, the di-ether already referred to above. which is not prohibitive in cost for specialised fuels. Methylal can be used alone as a motor fuel.

The higher the compression ratio of an engine the higher must be the Octane Value of its fuel. But the spark-ignition engine possesses a certain measure of tolerance for poor fuels resulting from the ability to vary the ignition timing retarding for starting and with fuels of low rating, and advancing for high speeds and with high octane fuels. This flexibility is lacking with glo-plug motors.

Glo-Plug Motor Fuels

The glo-plug engine is without ignition control, and fuel formulation might therefore be expected to be more critical than for spark ignition engines. For maximum racing performance this is undoubtedly true, yet it is surprising on how many weird and wonderful concoctions the average glo-motor will run passably well. A good general purpose fuel on which any glo-plug engine will run is a simple mixture of

but performance may not be outstanding. The castor oil proportion may with advantage be increased for some engines for the preliminary running-in; it should seldom be reduced below 20% even with well seasoned engines. Methanol does not have the natural inherent oiliness of diesel oil, and glofuels must have a higher oil content than diesel fuels. In order to develop the high revs of which it is capable the glo-engine must be fairly "sloppy", and to ensure adequate compression for starting a fairly oily viscous fuel is needed. Castor oil, and not a blended lubricant like Castrol "R", is to be preferred since it does not contain additives insoluble in methanol and therefore yields a clear fuel without sediment.

A very large number of substances have been suggested from time to time as useful additives to simple Castor Oil/Methanol blends in order to give increased performance. This list includes Amyl Acetate, Ethyl and Amyl Nitrates, Acetone, various cellulose solvents, Nitrobenzene, and many more. Extensive experiments in the Author's laboratory with these, and a host of other materials have led to the conclusion that whilst one or two may have a slight effect in glo-plug engines of early type, most of them are valueless in a modern glo-engine. In work with, for example, the latest type "Yulon", replacement of part of the methanol in a methanol/castor oil blend by

Ethyl Nitrate
Amyl Nitrate and Nitrite
Amyl, Butyl, Ethyl and Isopropol Ethers
Ethyl and Amyl Acetates
Paraldehyde
Acetaldehyde
Nitrobenzine

and many other solvents was found to have little or no useful effect, even when added in quite substantial quantities. It is true that in some instances the engine developed a very satisfying statico note suggestive of increased revs, a very portent exhaust flavour, or both, but in no case was any significant speed improvement recorded by the instruments.

An approach to the problem of improving simple methanol blends can be made by replacing part of the methanol by a fuel of higher calorific value such as benzene, toluene, acetone, ethyl alcohol or methylal. In some cases these materials effect a slight improvement, but usually more in the direction of improved fuel consumption than in increased speed. In any case there is a limit to the proportion of such substances that can be added since without exception, they have narrower Explosive Limits than methanol; after quite a small percentage has been added the throttle setting may become too critical for reasonably easy control. Furthermore, excess of some of these compounds of higher calorific value can cause an engine to run very hot indeed and to eject showers of red sparks so that risk of seizure becomes very real. Acetone was invariably found to give erratic running, which is surprising.

METHANOL. Some straight methanol castor oil blends have been found to run more smoothly than others. Modellers would be well advised to purchase only the purest methanol. Methyl and Ethyl alcohols come on the market in various "Proof" strengths, i.e. containing varying proportions of water and for best results only 74° over-proof methanol should be used (this contains over 99% of methanol).

METHANOL/CASTOR OIL RATIO. Unlike diesel fuels, the speed is not greatly influenced by variations in the base fuel/oil ratio. If a particular engine is adequately lubricated by, say, 20% of oil and 80% of methanol, there is no significant loss of speed when the ratio is altered to 30 : 70. On the other hand, if the former mixture is somewhat under-lubricating the engine, there may be a substantial increase in r.p.m. when the oil ratio is raised.

NITRO PARAFFINS. Whilst most of the substances so far discussed are without any profound effect on the speed of a glo-engine, this is certainly not true of the nitroparaffins. Replacement of part of the methanol in a Methanol/castor oil blend by Nitromethane, Nitroethane or Nitropropane may increase engine speed by between 1,000 and 2,000 r.p.m. In this respect the nitroparaffins appear to be unique—and are indispensable for really high speed work. Unfortunately, they have not hitherto been readily available in this country, they have been fantastically expensive, and except in carefully balanced fuel formulation they involve high fuel consumption. However, the outlook is improving; nitromethane and at least two proprietary nitroparaffin fuels are now on the British market, and nitropropane is on its way.

Just why nitroparaffins are so effective is not clear. They have very low energy contents, as reference to their Calorific Values in TABLE IV will show. Nitromethane, for example has only half the Calorific Value of Methanol and a nitromethane fuel might more logically, in fact, be described as a "cool" fuel than a "hot" fuel. Their effectiveness would seem to lie not in their intrinsic energy contents (which are very low) but rather in the extreme rapidity with which this energy can apparently be liberated. In their effectiveness, Nitromethane and the Nitropropanes are closely similar on the test-bench, nitropropane possibly giving a slightly more stable mixture under flight conditions. They would appear to be interchangeable in glo-fuel formulations, the choice depending mainly on price and availability. Nitroparaffin blends require a slightly wider throttle setting than non-nitrated blends, and are hence a little less economical in use.

With regard to the amount of nitromethane or nitropropane to include in a glo-fuel formulation, it is the Author's considered opinion that the proportions sometimes advocated are excessive. A fuel with 25% to 40% of nitromethane. apart from its exorbitant cost, usually seems to kill off gloplugs with fair rapidity. Secondly, careful speed tests on a number of engines have shown that at first there is a considerable speed increase when nitromethane is added. but that the effect gets progressively less with each further addition until it becomes insignificant. The experimenter is recommended to start off with a fairly small percentage of nitroparaffin in his fuel mixture and to carry out several speed determinations on his engine. Another mixture should then be prepared with the same base-fuel/oil ratio, but with a leper cent. more nitromethane, and further speed readings taken. This process should be repeated with small nit:roparaffin increases until there is no further speed increase measurable. In this way the most effective, and at the same time most economical, fuel will be worked out with the minimum waste of expensive materials. It will often be fcund by trials of this sort that 20% of nitromethane is just as useful as 30%.

The response of an engine to changes in fuel composition depends to a very considerable extent on the design of the engine, particularly as regards timing, porting and compression ratio. One engine may be found on test to be very much faster on a nitroparaffin blend than on a straight castor oil/methanol, whilst the performance of another engine may be found to be almost identical on either fuel. The moral is, clearly, do not run on expensive nitroparaffin blends if a non-nitrated racing methanol blend will give as good results. And equally, a commercial fuel should not be condemned because it does not improve the performance of your engine; it may be giving your friend another 1,000 revs on his engine of identical make. Engine manufacturers are constantly experimenting and incorporating minor design changes so that two apparently similar engines may, in fact, differ noticeably in compression ratio, timing, or both.

Finally. there is ample scope for studying the effect of combining nitroparaffins with other additives like amyl acetate etc., which are ineffective by themselves. The guiding principle in all such work always being to make only one change at a time, and to make the changes small gradual ones.

FUEL TESTING

Smoothness of running the absence of "missing" etc. can be tested fairly well with a critical ear—although an electronic stroboscope is better if you can borrow one. Adequacy of lubrication can be checked by feeling the crankshaft bearing (not the head!), by holding a plate behind the engine when it is running and noting how much oil is ejected, by noting whether the engine slows of its own accord when hot even with correct throttle and compression settings, and by seeing whether the engine runs any better when a few per cent. more oil is added to the fuel.

But SPEED cannot be checked by ear—USE INSTRUMENTS. An electronic stroboscope, if available, is the ideal instrument since it puts no load on the engine and since it shows variations in speed from second to second as well as overall average speed. Failing this, use a good Revolution Counter and watch, or Tachometer. The vibrating reed type of Revolution Indicator, if properly calibrated and carefully used. is capable of detecting reasonable variations in r.p.m. at the slower speeds. but is not capable of showing up small speed differences. It is suitable, therefore, for the preliminary experiments with diesel fuels, but is too insensitive at the higher revs to be of much value in glo-fuel development. In all cases the engine should be reasonably flexibly mounted; a well balanced engine fitted with a properly balanced prop, if firmly clamped in a vice, seldom gives a reading at all on a reed indicator.

In conclusion. do not be satisfied with a single speed reading—take half a dozen and average them. It is surprising what a difference twentieth of a throttle turn can make to a precision engine running near its flat-out maximum speed. And check back from time to time the values of your earlier fuel mixtures—the apparent increases in speed you have been getting with the later mixtures may be due to the engine loosening up with prolonged running. Elementary, but it happens every day.