The History of the Internal Combusion Engine

Send This Page To A Friend


Austria, France and Germany have each claimed to be the originators of the motor vehicle, but there is considerable doubt as to the validity of some of these claims. The probable sequence of events is as follows. At the Great Exhibition of 1851, in London, Etienne Lenoir produced a specification for an engine to run on petroleum vapour and air. This was patented in i860 and an actual engine was exhibited in Paris in 1862. It even had electric ignition, although it did not have a compression stroke, which is the prime essential for efficiency and economy.

The Lenoir design was much like the steam engines of the day, with a double-acting piston and slide valves. As the piston moved from one end of the cylinder, it first sucked in the mixture. Then, about half way along its stroke, the slide valve closed the inlet port and the mixture was fired, driving the piston the remainder of the way. On the return stroke, the slide valve opened the exhaust port and the cylinder was scavenged, while the induction and firing strokes took place on the other side of the double-acting piston.

Lenoir designed engine of 1860
Lenoir designed engine of 1860, bearing a strong resemblance to a steam engine. It had an electrical ignition but no compression stroke.

1885 Benz Single Cylinder Engine
The first engine to be fitted to a practical car occured in 1885 on the Benz. It was a single cylinder water-cooled unit which produced 0.9 hp @ 500 rpm.

Gottlieb Daimler's V-twin engine of 1887
Gottlieb Daimler's V-twin engine of 1887. The engine featured a hot-tube ignition system. A thin walled tube projected into the combusion chamber, and the outer end was heated by burners. Unfortunately it proved highly difficult to time engines with such a system.

Benz Velo engine of 1898
Benz Velo engine of 1898, which had a massive single cylinder of 110 mm bore x 120 mm stroke, giving a swept volume of 1140cc. The engine featured a coil-ignition system.

Leon Bollee tricar which featured a rear-mounted, single cylinder air-cooled engine
Leon Bollee tricar which featured a rear-mounted, single cylinder air-cooled engine.

Daimler 6hp Twin Engine
Another Daimler engine, this time a 6 hp twin, as fitted to a Panhard Levassor chassis.

24hp enigne as fitted to Count Zborowski's Cannstatt Daimler
24hp enigne as fitted to Count Zborowski's Cannstatt Daimler. Note that the engine's cylinders are still separate castings.

Berleit twin cylinder engine of 1897
Berleit twin cylinder engine of 1897.

Four cylinder engine used in the Berliet Targa-Bologna of 1906
Four cylinder engine used in the Berliet Targa-Bologna of 1906. The cylinders were cast in two pairs.

Ford Model-T engine cast en bloc
Ford Model-T engine cast en bloc. The unit was 2.9 litres.

1971 Wankel Rotary Engine
A cutaway of the 1971 Wankel rotary engine, as used by Mazda - however it was NSU who pioneered the design in their Prinz Spider of the mid 1960s. Unfortunately NSU had problems with the sealing on the three corners of the lobed piston, and some manufacturers lost interest. In 1974 General Motors also lost interest - but not because of seal issues, but rather that they could never achieve desired fuel consupmtion figures.

Next, in about 1875, came what the Austrians claim as the first car, but which has also been described as a powered handcart. This had a four-stroke engine, and it was made by Siegfried Markus, born in Vienna. However, the inventor of the four-stroke cycle— induction, compression, ignition and exhaust, all in two revolutions of the crankshaft—is claimed to have been Count Nicholas Otto. His patent is dated 1876, but, no doubt, some time elapsed between the conception of the idea and the filing of the patents.

Others had realised the importance of compression of the mixture of gas in an engine, but had turned to the more obvious device of using a separate cylinder for this purpose. Among these was Dugald Clerk who, in about 1880, applied the principle to a two-stroke engine; indeed, the two-stroke cycle with compression has sometimes been termed the Clerk cycle, even though he originally used a separate cylinder for compression.

One of the first two-stroke engines to have a transfer port, so that the underside of the piston could be used to compress the charge in the crankcase, was that patented in 1889 by Day and Son, of Bath, England. This was a gas engine in which the inlet port to the crankcase was opened and closed by the bottom edge of the piston which, therefore, acted as a slide valve. The inlet opened near top dead centre (TDC) and the transfer port was opened as the top edge of the piston passed it, on approaching bottom dead centre (bdc), the exhaust, of course, having been opened similarly but rather earlier. Although the two-stroke engine has always had a minority following, the exhaust-gas pollution regulations of the 1960s and 1970s ultimately killed it for road-going, four-wheeled vehicles, due to its inefficiency.

As regards the application of the internal combustion engine to the car, there can be little doubt that the true pioneers were Gottlieb Daimler and Carl Benz. Daimler had worked for Otto, and left, in 1882, to set up his own workshop near Stuttgart, where he was joined by Wilhelm Maybach. By 1884, the first Daimler-Maybach engine had been built, and was later installed in a carriage, the shafts of which - for the horse - had been removed.

The Daimler engine ran up to 900 rpm, compared with the 200 rpm of the Otto engines of that time. First, a single-cylinder version was produced followed by a V-twin, with a carburettor designed by Maybach who, in 1892, introduced a new unit with a float and jet. The Benz four-stroke engine was installed at the rear of a tricycle in 1885. Among its most interesting features were mushroom valves and watel" cooling. The water was kept in a reservoir which had to be topped up so that it did not boil dry. Radiators did not come until later. Benz, with the Velo, as this three-wheeler was called, was the first to produce motor vehicles in significant quantities.

In France, Panhard and Levassor produced their first car in 1890, with an engine built under the Daimler patents. This was centrally mounted, but another model was built a year later with the engine at the front, where it was less exposed to the clouds of dust and showers of mud thrown up from the roads in those days. Emile Levassor's contribution was to establish the front engine and rear-wheel-drive layout, replacing the belt or chain transmission by a clutch and gearbox. He also introduced the gilled-tube radiator. However, it was the British Daimler Company, founded in 1895, which introduced the first honeycomb radiator with integral tank in 1899.

In the meantime, the first successful American car, the Duryea, ran in 1893, while Henry Ford's Quadri-cycle was built in 1896. The latter had a three-horsepower engine with two horizontal cylinders, of 65 mm bore by 152 mm stroke, an open crankcase and a centrally disposed flywheel. Its maximum speed was 600 rpm. By 1901, the Spyker car had been introduced in Holland, and Castro had begun production in Barcelona. The latter firm was taken over by the brilliant Swiss engineer, Marc Birkigt, in 1904, becoming Hispano-Suiza. Birkigt's engines were, for many years, considered to be the best in the world.

As has often been the case, restrictive legislation impeded progress in the UK, and forced the British to take a back seat while other countries forged ahead. Despite the infamous red flag law, in force until 1896, the Lanchester brothers made outstanding contributions to progress. Dr Frederick Lanchester patented a high-pressure lubrication system and a wick-type carburettor, and both these inventions were incorporated in the Lanchester engine of 1896. This engine also had a magneto instead of the, then common, heated-platinum-tube ignition system.

Later came a Lanchester engine with two horizontally-opposed cylinders, each piston of which had two connecting rods, coupled to a different crankshaft. The crankshafts were arranged one above and one below the common axis of the two cylinders, and rotated in opposite directions. This meant that not only were both the primary and secondary vibratory forces balanced, but so also, to a major extent, was the torque reaction. Such an engine, when most others shook their vehicles and occupants to pieces, stood out as little short of miraculous.


Following the early ventures and subsequent spread of motor-car technology throughout the world, a multitude of detailed improvements ensued, and this process is, of course, still continuing. First, the prime requirement was a suitable fuel. Prior to World War I, any volatile fractions distilled from crude petroleum were used. The word 'petrol' was originally a trade name for a product offered by a firm called Carless, Capell and Leonard, of Hackney Wick, London. So long as the wick-type carburettors—no connection—were used, volatility was the most important requirement.

Although floats and jets were the order of the day, by about 1908, the octane ratings of fuels, commercially available until about 1921, ranged only between forty and fifty. In consequence, compression ratios had to be limited to a maximum of about 4.5 :1, so fuel consumption rates were relatively high; the octane rating of the general run of fuels, nowadays, is about 96, and the top-grade ones about 101. In consequence, compression ratios are now mainly of the order of 8:1. A good illustration of the advantages of increasing the octane rating of fuels is the virtual doubling of power output of aero engines consequent upon an increase from 77 to 100 octane shortly before World War 2. The steady progress in fuels—and, indeed, in many other aspects of engine design - has been largely due to the truly enormous amount of research and development done by the oil companies.


Understandably, engineers were originally so engrossed in the mechanical aspects of design, that the fundamentals of combustion were largely ignored. Engine knock was taken for granted as inevitable, and compression ratios were set as high as possible within the limitation imposed by this phenomenon. Despite the fact that Hopkinson, in the early 1900s, had suggested that knock and pre-ignition were entirely separate phenomena, it was mostly believed, as late as 1910, that knock was due to premature ignition.

The great impetus to development given by World War I led to intensive study of combustion in many countries. In the UK, an outstanding figure in this work was Harry R. Ricardo. By 1920, the principles of combustion-chamber design were well known, and in August 1924, the Ethyl Gasoline Corporation was founded in America for the marketing of tetra-ethyl lead (TEL) as an anti-knock additive. This corporation was, in fact, a partnership between General Motors and Standard Oil, and the tetra-ethyl lead was manufactured by du Pont right up to 1948. In the meantime, in 1926, Graham Edgar, of Ethyl, invented the octane rating principle.

The discovery of TEL is an interesting story. Thomas Midgeley Jnr, of General Motors, had a theory that detonation was caused by the sudden evaporation, and therefore over-rapid burning, of fuel from droplets in the cylinder. He concluded that this was due to the absorption of heat by the droplets and, therefore, if they could be coloured—to absorb radiant heat more rapidly from the walls of the chamber - they might evaporate before the spark occurred, and the flame would spread at a steadier rate from the centre of spark ignition. Actually, this theory was untenable, but it did set him on the right path.

Midgeley scoured all sources available to him for soluble dyes to colour the fuel. He found that iodine suppressed knocking, and then went on to investigate aniline. His first ethyl compound was ethyl iodide, but eventually, in 1922, he got round to tetra-ethyl lead. Subsequent developments included the introduction of bromide and chlorine compounds to assist in preventing the build-up of lead deposits in the cylinders. Tetra-methyl lead was discovered later still. It had a boiling point not much more than half that of TEL, and so tended to distribute more uniformly between the cylinders of an engine.

Combustion-Chamber Design

Developments in combustion-chamber design have gone hand-in-hand with advances in fuel technology. There have been three essentials, all related to knock rating: compactness, turbulence of the mixture and good cooling. Limits are set, however, by the need for manufacturing economy. Even though production quantities were nothing like so large in the early days, side-valve engines were the norm—on grounds of low cost—until well into the 1920s, and the mark IV Hillman Minx saloon with a side-valve engine was not superseded by an overhead-valve version until 1955.

Up to about 1920, the two valves were generally together on one side of the cylinder. The cylinders and their heads were integral, and the valves removed through detachable screwed plugs above them. Then the T-head came into vogue, with the inlet valves on one side of the cylinder and the exhaust on the other. This had the advantage that there was more room for larger valves and the gas flow was unidirectional, so breathing was improved. It also had the merit of accessibility. However, the combustion chamber was less compact than that of the side-by-side valve engine, and there was less turbulence, so compression ratios had to be restricted to prevent detonation.

By 1920, the overhead-valve engine, with push-rod actuation, was making inroads into the market. In the next few years, there came the famous Bentley 3-litre and Sunbeam engines with overhead-camshaft actuation of the valves. For compactness, the trend in combustion-chamber shape went to hemispherical, penthouse, wedge or bathtub shapes. In the meantime, the side-valve design made a come-back by virtue of redesign of its combustion chamber to reduce the length of flame travel and to induce extra turbulence.

Turbulence was obtained partly by suitable shaping of the induction port and passages in the head. The aim was to direct the flow at a speed and in a manner such that eddies were shed from both the rim of the valve seat and the head of the valve as the gas entered the cylinder. Additionally, part of the roof over one side of the combustion chamber was brought down almost to the level of the piston crown at top dead centre. This formed what has been termed a 'squish shelf' because the gas trapped between the two surfaces is squished (squirted) out into the main part of the combustion chamber. In side-valve engines, the resulting jet of gas is directed towards the valves and sparking plugs. The placing of the squish shelf on the side, remote from these components, helped to ensure that the main combustion space was compact.

As far as cooling was concerned, many of the early engines were faulty. For example, the 1896 Leon Bollee 2 hp, single-cylinder, air-cooled engine of 650 cc swept volume had cooling tins only on the cylinder barrel. In other respects, this engine was very advanced. It had a Maybach type of spray carburettor and a detachable cylinder head. However, it had hot-tube ignition and an automatic inlet valve. The exhaust valve was mechanically actuated, and its operation was regulated by a centrifugal governor to restrict the speed of the engine to about 700 rpm. In 1897, C. McRobie Turrell produced a modified version of this design for the Bollee-Coventry Motette. Turrell incorporated fins in the cylinder head to prevent overheating in the vicinitv of the exhaust valve.

Vibration and Noise

Vibration was the bugbear of the early engines. It shook motors and vehicles to pieces, and subjected the occupants to extremes of noise and discomfort. A major factor was the early engineers' fear that increasing the number of cylinders would increase the risk of unreliability. However, with the demise of the automatic inlet valve—opened by suction in the cylinders— and the introduction of magneto ignition, these fears were allayed. By 1897, the cylinders were already being arranged one behind the other, instead of in the V-layout previously favoured. The first four-cylinder engine was built in 1898 by Daimler in England. Ader of France built the first V8 engine in 1903, and the first in-line-six engine for a car was designed by S. F. Edge working with Montague Napier.

In those days, the competition from the relatively quiet and smooth-running steam engine did a lot for the new machines. As early as 1897, Lanchester had produced the fully balanced engine previously described. Later, Lanchester's contra-rotating weight balancer, for counteracting the secondary vibrations of a four-cylinder engine, was introduced. It was driven by a spiral gear from the crankshaft, but at twice engine speed, and was positioned beneath the centre main bearing. Several manufacturers have used it over the years, but it is expensive and can be noisy. In the meantime, Gobron-Brillie in France and Arrol-Johnston in Scotland developed engines with combustion occurring between two pistons moving in opposite directions in each cylinder, but connected to a single crankshaft. The French invention had long connecting rods as in the more modern Doxford marine engine, while rocking beams were used in the Scottish engine, as in the Commer two-stroke diesel engine introduced in 1954. It was an Arrol-Johnston car that won the first Tourist-Trophv race in the Isle of Man.

The fundamental answer to the balancing problem lay in the adoption of the in-line-six-cylinder layout with a 120-degree crankshaft, which became almost universal in the USA before the advent of the V-engines. However, the less expensive four-cylinder unit has remained popular in Europe. The latter owes its success in no small measure to flexible mounting and, in particular, the Chrysler Floating Power system introduced in the early 1920s. Other detail advances in reducing noise have included improvements in camshaft design, the total enclosure of valve mechanisms, accurate machining, and balancing of rotating and reciprocating parts, and the adoption of light alloy instead of cast iron or steel pistons. Valve gear open to the dusty atmosphere, and therefore subject to rapid wear, was not uncommon even in the period 1910-1920.

Leading in engine balancing was De Dion Bouton. For example, the company's Populaire model of 1903 had an 846 cc single-cylinder engine developing 8 hp at i500rpm, and an experimental version of this design had run up to 3000 rpm for short periods. The first manufacturer to use a crankshaft-balancing machine in quantity production was General Motors, in 1924, for its Cadillac engine.


One of the limiting factors inherent in an engine has always been piston speed, which is also affected by stroke to bore ratio. Power output is solely a function of brake mean effective pressure (bmep) and piston speed. In 1910, piston speeds of about 1500 ft per min were normal. By 1920, they had increased, by virtue of the use of aluminium pistons, to 2500 ft per min and they have risen to little more than 2700 ft per min since. In 1910, mean effective pressures of 85-90 lb per sq in were considered very good and compression ratios were about 4:1.

Ten years later, compression ratios had increased to about 5:1 on 45 octane fuel, thanks almost solely to good combustion-chamber design—improvements in fuel were to come later. In consequence, mean effective pressures were up to 110-120 lb per sq in, with peak pressures of about 500 lb per sq in. Nowadays, the corresponding figures are of the order of 140-150 lb per sq in, with compression ratios of 8 :1. Peak cylinder pressures are about 800-900 lb per sq in.

The early engines mostly had aluminium crankcases with cast-iron blocks and heads. Stroke/bore ratios of as high as 2j : 1 were common. The result of such a long stroke was a deep structure having good beam stiffness. By 1910, however, ratios had, in general, settled down to between 2:1 and 1:1. Notable exceptions were the Lanchester four and six-cylinder engines with 3 in stroke by 4 m bore. For high performance cars, on the other hand, long strokes remained. Notable examples were the Hispano Suiza with 80 mm bore by 180 mm stroke, the Vauxhall 30/98, with 95mmx 140 mm and the Bentleys.

The Ford Model T, in 1908, set the fashion with its integral cast-iron cylinder block and crankcase. The inherent stiffness of this material, and the design, enabled stroke/bore ratios in general, to come down to less than unity. In the UK, however, the long stroke persisted until 1947, when the old RAC formula for basing taxation on the bore dimension was abandoned. Then, shorter strokes made the British engines more compact as regards overall height, and to some extent facilitated handling and machining in production. Engine speeds increased, also, to obtain the same piston speeds and therefore equivalent power.

Aluminium was impracticable for really large quantity production until a reliable die-casting process was developed. High hopes were expressed for the Reynolds A390 alloy and the General Motors Accurad process, both announced in 1969. The process has been viewed with some scepticism in certain quarters, but other methods have since been perfected for use with the same alloy. The development that really heralded mass production of the car engine was the shell-type bearing, w hich was pioneered by Allison for the Liberty aircraft engines in the early 1920s. Previously, each bearing had to be hand-scraped to match the journal.

L-head, I-head, and F-head Engines

The terms used for the head of an engine came down through the years as engineers' slang for conveniently designating cylinder-block construction. Supposedly, the mechanical elements involved were so named because they resembled these letters of the alphabet. In the L-head engine, which was a type widely used, both the intake and the exhaust valves were in the cylinder block, at the sides of the cylinders. This type of engine was generally considered to be the most economical to build and the easiest to keep quiet, which accounts in large part for its popularity.

The I-head engine was an overhead type. Both the intake and the exhaust valves were directly over the pistons. This arrangement allowed for constricted valves and more compact combustion chambers. As for the F-head engine, each intake valve was placed in the cylinder head over its piston, but the exhaust valves were located in the cylinder block beside the pistons. Unrestricted by space limitations, the intake valves could be made unusually large: the result was greater volumetric efficiency. This type of engine produced more power per cubic inch of displacement and was more economical to operate.

In the 1960s it was only Willys that was building an F-head engine, but in England there were three - Rolls-Royce, Bentley, and Rover. Which design was best is a matter of opinion - but of the three designs, the F-head engine had less sensitivity to combustion-chamber deposits and fuel-octane ratings than any other type of engine with the same compression ratio.

Also see: The History of the Diesel Engine | The History of the Automobile

Latest Classic Car Classifieds

You may also like...