Home > Cars > How A Car Works

How a Motor Car works - what every young person should know

(Article by Colin Hughes)

This article is quite long: discussion with the Website Editor, Simon Coss, whether to have a text with hyperlinks to subsections giving greater detail, or to run it all as one thing, have resulted in it being all one thing. However if you start reading it and fall asleep but can remember where you got to, here is a list of section headings that you can click on to get straight to them.

If there are any of you who have found the information here useful for GCSE, whether you are teaching it or taking an exam, let us know. Tell us what we've left out, too.

Contents

Introduction  

At the 2004 Rolls-Royce Enthusiasts Club Conference, one of the delegates commented on the lack of interest in motorcars among boys. It may be that the continuous quest for cleanliness by their mothers works against a hobby that, however clean the car may appear, always results in black oily sludge covering one's hands and usually the soles of one's shoes. His comments spurred me, to get together a collection of pictures which I hope will do what old copies of children's magazines used to cover regularly - "How a motor-car works". This text is not designed to tell you everything, as my own understanding of the mechanics is pretty basic (and even may be wrong in some bits). I have where possible used Rolls-Royce publicity and technical illustrations, but these are padded out with other items from "The Wonder Book of Motors" of around 1930 and others which I will acknowledge. Click on any thumbnail picture to show a larger version.   

There are many who consider that the UK's loss of its engineering lead has come from the demise of "Meccano", which, for those who have never heard of it, was a metal construction outfit using perforated strips of steel, brackets, shafts, gears, nuts and bolts in a wide range of outfits numbered from 1 to 10. Most boys aspired to the No. 10, but you could start small and buy accessory outfits to upgrade to the next size - a birthday or Christmas present request that made it easy for your parents. The more recent popularity of "Lego" does not seem to have had a compensatory effect in our success in architecture or the building trade, though. This picture of a car chassis was originally in a Meccano brochure of 1915. I remember building something similar which included a two-speed gearbox and a clockwork motor in the engine. Original Meccano does turn up on eBay, but at prices only affordable by fathers for themselves. Most modern Meccano comes as a kit to make a specific model (as does a lot of Lego) rather than as an outfit with the potential to make a number of different designs, either as examples in a book or as you designed yourself. I feel that the current approach limits creativity - do you agree? 

Basic Components  

The frame that carried the mechanical parts of an early car as well as the body was known as the "chassis" - this is a French word and reflects France's early dominance of the motor car industry around 1900. If Britain had not had what was called the "Red Flag Act" which limited the speed of motor cars by demanding that they follow a man carrying a red flag, its motor industry might have flourished sooner. Probably the chassis would have been called the "carriage" which was the term used for the frame on which the body of a horse-drawn coach was mounted. The advantage of a chassis is that many different body styles can be fitted, but at the sacrifice of rigidity and/or greater weight. Most modern cars no longer have a chassis, because the whole of the body carries the forces generated by the car in motion; the first version made by Rolls-Royce of an integrated body structure was the Silver Shadow in 1965, but other manufacturers, such as Lancia and Citroen, had produced forerunners in the 1930s. 

The chassis shown above are (l. to r.): the first Royce 10hp car of 1904, two views of a 40/50hp "Silver Ghost" chassis and a 20/25hp chassis of around 1931. The first picture is from "A History of Rolls-Royce Motor Cars" by C. W. Morton (as are the pictures of the early Royce cars and engines below). The 40/50hp chassis and 20/25hp chassis shown above are from the car owner's handbooks.

The layout of the mechanical parts of a car had become near standard around 1901. Rolls-Royce legend has it that Royce decided to design a motor car because the Decauville that he owned was crude and he felt that he could do better. It was much more likely that he was becoming aware of intense competition for the electrical motors, dynamos and travelling cranes that his firm manufactured and was looking for a more profitable line that his factory had the capability to build without much extra investment in equipment. The Decauville was relatively advanced for its time, and many of the dimensions of the first Royce car were based upon it.  

If you click on this picture, it will show a Rolls-Royce 10hp chassis with most parts labelled that I am going to mention. The chassis had two sprung axles carrying wheels on rotating hubs, with the front axle fitted with king-pins and stub axles for steering; a track rod made the connection to make both wheels steer. Steering was controlled by the steering wheel and gears or a worm and nut drive in a steering box operated a drop arm connected to one stub axle via a drag link. (Rolls-Royce generally called the drop arm a steering lever, the drag link a side steering tube, and the track rod a cross steering tube). The rear axle was the driven one, and fitted with a differential gear to allow for the need on bends for the outer wheel to rotate faster than the inner one. On early cars braking to slow the car was only applied to the rear wheels via brake shoes inside brake drums mounted on the hubs. Early brake shoes were cast iron, but later had a high friction material riveted or bonded to the shoe. In the early Royce cars the handbrake lever operated these brakes. The footbrake pedal operated a transmission brake mounted behind the gearbox (not captioned in this picture). The engine was mounted over or just behind the front axle, behind a radiator that cooled the water that circulated around the engine to absorb excess heat. 

Behind the engine was mounted a clutch to connect and disconnect the power of the engine, which was attached in turn to a gearbox which provided different gear ratios to suit the power required to accelerate the car or to climb hills. The gear lever that moved the gears in the gearbox was known in early cars as the change speed lever because most early engines tended to be run at a fixed rotational speed until they were better able to be controlled - changing gear gave a different speed.

The gearbox output was fed through a propeller shaft with universal joints that allowed for the movement of the rear axle caused by unevenness of the road.  The propeller shaft connected to a pinion gear meshed with a larger crown wheel attached to the differential gear to match the engine's rotational speed to the rotational speed needed at the wheels of the car.  The crown wheel and pinion are also known as the final drive gears.  (Some early cars used a chain drive between the gearbox and the rear axle: two chains on sprockets independently powered each wheel; the differential was mounted behind the gearbox).  This arrangement of the components of a car is still common in larger motor cars today: smaller cars have generally moved to the arrangement of the drive from the gearbox being fed to the front wheels, which are also the steered wheels: a design first seen in the 1920s, but not successful until universal joints had been developed capable of giving a smooth drive through a tight angle.

The engine

In the early days of the motorcar, steam and electricity had been tried as well as the petrol fuelled "internal combustion" engine. Although some steam cars continued to be made until the 1920s, by far the majority of passenger motor cars used the petrol engine. Although the Diesel engine had been invented, its application to personal transport was rather later, mainly because it was then both heavy and noisy.

The picture on the left came from "The Wonder Book of Motors", one of a series published by Ward Lock & Co. from early in the 20th Century, including ones on Aircraft, Ships, Nature, Electricity and Railways. It shows the cross-section of the cylinder of an internal combustion engine, with the parts captioned. It does not show the parts that supply the petrol-air mixture, the spark at the sparking-plug, or circulate the cooling water or the oil; these will be described later.  

The essential moving parts are a piston that moves in a cylinder connected by a connecting rod to a crank on a crankshaft that converts the up and down movement of the piston into rotational motion. The crankshaft has journals that rotate in main bearings mounted in the crankcase. The end of the connecting rod rotating on the crank has a bearing known as the big end. The bearing at the other end of the rod is known as the little end and holds the gudgeon pin that is mounted in the piston. Gears attached to the crankshaft drive in this case two camshafts whose cams lift valves that control the inlet of petrol-air mixture and the exhaust of the burnt gases. The valves are helped to close by springs. 

The Otto Cycle

The fundamental way in which the internal combustion engine works is known as the Otto cycle after its inventor. Nicolaus August Otto was born on 14th June 1831 in Holzhausen (Germany). In 1862 he began his first experiments. Together with Eugen Langen he founded the first engine company - "N.A.Otto & Cie". For an animated version on the web, go to: http://www.keveney.com/otto.html or http://techni.tachemie.uni-leipzig.de/otto/otto_g0_eng.html which gives more history as well.  

The cycle is formally described as: Induction, Compression, Ignition and Exhaust, this is known as a four-stroke cycle in which the piston moves through an upward and a downward stroke twice, totalling four strokes. (There is a two-stroke cycle in which the piston moves only once through an upward and a downward stroke: the application of this has mainly been to small high rotational speed engines running near constant speed: you will find them on some motor-cycles, garden mowers and strimmers, although a number of cars have been two-stroke: early Saabs, Wartburgs and Trabants among them). A simpler description to remember for the Otto cycle is Suck, Squeeze, Bang, Blow.

The left-hand diagram shows the engine at the induction stroke: the inlet valve has opened, the piston is descending and petrol-air mixture is being drawn into the cylinder. The next picture shows compression: on reaching the bottom of the stroke, the inlet valve closes and the piston moves upward, compressing the gas of the petrol-air mixture into the combustion chamber. This does two things: it makes the gas hotter (you may have noticed how the top end of a bicycle pump gets hot because the air in this area is compressed when you pump) and it makes it denser so that it is easier to set light to: the molecules of the gas are closer together. The third picture shows ignition: an electrical voltage applied to a spark gap makes a spark jump the gap and sets light to the petrol-air gas mixture in the combustion chamber. Because the gas is compressed, the flame propagates rapidly through it causing its temperature and pressure to rise. This pressure increase drives the piston downwards generating the power of the engine: the pressure and temperature drop as the piston completes its stroke and moves into the last stage: exhaust. The exhaust valve opens and the burnt gases escape to the exhaust pipe both under the excess pressure still remaining from the explosion of the mixture and driven by the piston on its upward stroke.

This example shows an engine with one cylinder only. It would be possible that the effort of compressing the mixture would stop the crankshaft rotating because the weight of the crankshaft and gears is not enough to overcome it. A flywheel, a circular weight on the end of the crankshaft, has sufficient inertia when rotating to keep the engine running even at slow speeds. Because the power of an engine depends on how much petrol and air mixture can be burnt at each cycle, it was found very soon that increased power could be obtained by making the cylinder and stroke bigger, but that smoother running as well as more power was achieved by having more cylinders and pistons arranged to fire at regular points during the rotation of the shaft; also, one large piston moving a long way up and down creates much more load on the crankshaft and bearings than several smaller ones that sweep through the same volume in one turn of the shaft.

A rough indication of the potential for power in an engine is its "swept volume": the volume that the pistons move through in one full cycle. A 3 litre engine could have two cylinders, each with a swept volume of 1 litres, or six cylinders, each of litre swept volume, and so on. In addition, the higher the speed of rotation, the more mixture can be sucked through the engine, and therefore the more power. The limit to this is usually the ability to get the mixture in through the valves and the exhaust gas out again. Another way to higher power is to increase the compression ratio - the ratio of the volume of the engine with the piston at the bottom of its stroke to the volume of the combustion chamber when the piston is at the top of its stroke: it makes the engine more efficient. The explosion travels faster at a high compression ratio, but the loads on the engine moving parts are higher and there is a risk that the mixture will ignite before the spark is fired (an effect called pinking from the sound the engine makes when this is happening R-R called it detonating). The quality of the petrol affects the rate that the flame travels through the compressed mixture: to work with high compression ratios it must have a slower rate of burning. The measure of this feature is based on the RON rating (Relative Octane Number), which compares the flame burning rate of a petrol sample with a standard chemical compound, Octane. If the sample is the same as Octane, it is assigned the RON number 100. If the flame burns faster, it will have a lower RON number. Early petrol was around 40 to 60 RON.

Engine power rating

The power of a motor car was expressed in "horse-power" (one horse-power is 746 Watts), but in the UK the horse-power (hp) used to describe the size of a car engine in the first part of the 20th Century was based on the RAC (Royal Automobile Club) hp rating, which was a complex formula based on the cross-sectional area of an engine's cylinders and was used for taxation purposes. It was not far from the actual power of an early car, but modern cars would probably produce ten times their RAC hp rating. Manufacturers found they could make the piston stroke very large, increasing the engine's power, but without changing the taxation class; this caused some mechanically undesirable designs, but as the tax was around One UK Pound per horsepower, this was some benefit if you consider that a 10hp car might pay 1000 a year in road tax in modern terms. Engine designers found that a short stroke and a wide cylinder bore reduced the inertial forces on the crankshaft and allowed more room for the valves to let in the mixture, giving greater power potential for the same swept volume. Much as with the "formula" used for Grand Prix racing cars, the design of engines could be affected by a regulation that might not result in a sensible direction for development to move. Power nowadays is expressed in terms of "brake horsepower" or in Kilowatts, based on tests of the maximum output produced as measured on a "dynamometer". A dynamometer allows the power generated to be measured at different engine speeds by varying the load, often by running an electrical generator (a dynamo) whose power is dissipated in electrical resistances capable of handling high powers. Early dynamometers used brake shoes rubbing on a drum; the force trying to rotate the shoes was measured with a spring balance and by varying the pressure on the shoes, the power of the engine could be calculated at different speeds. From this use of a brake, the term "brake horsepower" was derived. Royce, because he was an electrical engineer manufacturing motors and dynamos, tested the output of his prototype 10hp engine by running a dynamo and measuring the electrical power generated.

The Valves

The picture on the left shows the valve-gear - the mechanism for controlling inlet of mixture and exhaust of the burnt gases. Many early cars had the arrangement shown earlier to demonstrate the Otto cycle. This was known as a "T-Head" with side valves: valves mounted on either side of the cylinder and with the cylinder head containing the combustion chamber and spark plug. The arrangement permitted cross-flow where improved performance could be achieved by opening the inlet valve slightly before the exhaust valve closed, so that the exhaust gases leaving the cylinder could give a slight vacuum that helped to suck in the mixture. It also was relatively simple, because the drive to the camshafts could be directly from the crankshaft so that the valve mechanism could be operated virtually directly by the cam (tappets were operated by the cam that were adjustable to set the clearances of the valves to allow for wear and heat expansion of the engine so that the valves closed completely and did not rattle). Later designs generally had both side valves at the same side of the cylinder so that only one camshaft was needed: the "L- Head". Royce chose a different layout for the early cars: overhead inlet, side exhaust, a layout known as "F- Head". The diagram shows on the left the arrangement of the 10hp car and on the right that of the Wraith and post WWII six cylinder cars such as the Silver Wraith, Silver Cloud, Mark VI, R and S Type Bentleys. Overhead valves create a challenge because the mechanism is more complicated: it is necessary to mount part of the linkage operating the valves on the cylinder head and to interconnect the tappets riding on the camshaft to this. On the 10hp car this was done by pushrods and rockers, a method used by Rolls-Royce for all its production motor cars. In the 10hp car, the cylinder head was cast together with the cylinders and combustion chamber. Each overhead valve was mounted in a carrier screwed into the cylinder head, while another screwed valve cap gave access to allow removal of the exhaust valve. In the case of the later design, the whole cylinder head could be removed for servicing of the valves. This design also allows cooling water closer to the valves and the combustion chamber, as well as giving a flat sealing surface rather than having screw threads that frequently became jammed with carbon, or leaked. Rolls-Royce cars after the 40/50hp until the introduction of the Wraith had both inlet and exhaust valves mounted overhead on a detachable cylinder head and operated by pushrods from a single camshaft.

Some modern cars use an overhead camshaft mounted on the cylinder head to operate the valves. This has the advantage that there is little mass in the operating mechanism: there are no push rods or rockers, only the valves, tappets and springs. This allows the engine to run faster before the valve mechanism stops following the shape of the cam - "valve bounce". The problem on early cars was to drive the overhead camshaft: only when reliable chain drives were available did these become more common. Gear drives were expensive and difficult to design to allow for expansion of the engine as it warmed up. Rolls-Royce never felt that chain drives could be quiet enough for their clientele.

Crankshaft design

Royce's first car had just two cylinders cast side by side in a block, but when he went into partnership with Rolls, a range of cars with two, three, four and six cylinders was established. When designing his two-cylinder engine, Royce had a choice: if you look at the diagram of the Otto cycle, you can arrange that the ignition stroke is at the same time as the induction one on the other cylinder, which gives even timing of the explosions, but means that the pistons both move up and down together, giving a poor balance that tends to bounce the engine up and down. Or, one can arrange the cranks so that the pistons move in opposite directions, balancing both the vibration of the pistons and the rotation of the cranks, but with an uneven order of firing. Royce chose this approach (shown on the left is a picture of the crankshaft, flywheel and connecting rods) recognising that the car would not be running most of the time at full power, and the lack of vibration from the engine's moving parts was more important for the overall smoothness of the car than even firing pulses.

Far left picture: the three cylinder 15hp car had individual cylinders mounted on a crankcase, with cranks at 120 degrees apart on the shaft, giving good balance and even spaced explosions at every 240 degrees of rotation. Near left picture: the four cylinder 20hp arrangement had two 10hp blocks mounted on the crankcase, with the crankshaft arranged as two 10hp crankshafts mirrored, with all the cranks in one plane. Note that all these crankshafts were counterbalanced to minimise out of balance loads on the main bearings. Later Rolls-Royce crankshafts, from the 40/50hp onward until the early 1930s, did not have counterweights, because the diameter and width of the journals was felt sufficient not to need them and also to reduce the rotating mass of the crankshaft. When engines achieved higher rotational speeds, the load on the main bearings increased, especially where two cranks were next to each other (cylinders 2 & 3 in a four cylinder and 3 and 4 in a six) and although the bearing width and diameter was made bigger between these cranks, balance weights reduced the side loads on the bearings at high speeds.

The Six-cylinder Crankshaft  

The six-cylinder engine raised some special design problems. At the time that Rolls met Royce, it was fashionable to have six cylinders because the engines were smoother than fours. Manufacturers like Napier with its publicist S. F. Edge were pushing the benefits of six cylinders in "the Battle of the Cylinders". The pictures on the left show the further developments in 1905 and 1906 in the six-cylinder Rolls-Royce engines. The two drawings at the far and middle left, as well as the photograph, are from C. W. Morton's book, the near left one is a Max Millar one published in "Autocar". The third from the left is Albert Wheatman turning the first of the modified 30hp crankshafts, with examples of the 2 and 3 cylinder crankshafts in the foreground. The main point to note is that the original 30hp six was achieved by bolting three 10hp cylinder pairs on a common crankcase. You can also see that there is now only one camshaft operating both the overhead inlet and side exhaust valves, which were still in the "F-head" arrangement. While the 20hp four-cylinder engine was made up by bolting two 10hp pairs onto a crankcase and was a successful design, the 30 hp was not as successful because the position of the cranks along the crankshaft and the sequence of ignition firing resulted in a torsional (rotational) vibration that both caused the timing gears to rattle and in extreme conditions caused crankshaft breakages. We will say more about this later. The problems of the 30hp were overcome in the 40/50hp "Silver Ghost", an engine that with development continued for nearly 20 years. You can see that it has the cylinders cast in threes and is in effect two three-cylinder engines mirrored. In fact the arrangement of the cranks of the 40/50hp was introduced in later 30hp cars - the 30hp started with a crankshaft with pairs of cranks at 180 degrees arranged at 120 degrees along the shaft, which contributed to the vibration problem. The 40/50hp engine moved to an "L-head" arrangement of side valves, this had the inlet and exhaust side by side, operated by a single camshaft. Another change was to clamp the main bearings against the top half of the crankcase: in the earlier pictures of the crankshafts of the 10, 15 and 20hp cars, you can see that the bearings are fitted in the lower half of the crankcase. Because the engine mountings onto the chassis were on the upper half of the crankcase, driving forces on the crankshaft were taken through the lower crankcase and its attachment bolts to the upper half. By mounting the bearings in the upper half, the loads are taken directly through to the engine mountings. The lower half of the crankcase became simply the oil sump and took no engine loads.

Crankshaft vibration

I mentioned the issue of crankshaft vibration in the section about the first six cylinder R-R cars. When the 30hp six was launched, one car on test broke its crankshaft just behind its front flywheel and two others did so in customers' hands. The front flywheel had been fitted to reduce a rattle in the timing gears which drove the camshafts. The main issue was that the crankshaft lacked torsional stiffness, meaning that it would wind up and unwind as it rotated. The firing order of the engine and the weight of the crankshaft with its balance weights aggravated this torsional vibration. At that time also, the pistons were cast iron, with heavy connecting rods. Royce reduced the effect by removing the balance weights, which lightened the crankshaft and raised its natural vibration frequency, and by reducing the weight of the front flywheel. Without a redesign with bigger journals, he could not make the crankshaft stiffer, but learnt from this for later crankshaft designs.

During this time of experiment on the front flywheels, a distance piece was made up in wood for one flywheel experiment, which reduced the timing gear rattle. When replaced with a production version in metal, the rattle returned. Royce's examination of the wooden distance piece showed some charring on its surface indicating relative movement had been happening. He recognised in this evidence the principle of the "slipper-flywheel" vibration damper, although he did not design one until later. A similar effect was also discovered with an engine that had noticeably less rattle because the flywheel was not tight on the tapered nose of the crankshaft and evidence of movement could be seen. In later years there was a dispute because Dr Frederick Lanchester had patented a similar vibration damper, which was settled when Royce's record was disclosed of prior discovery of the principle, although he had not patented it at the time. The picture shows the cross-section of the slipper-flywheel, usually called the "slipper drive", fitted on R-R six cylinder engines through the pre WWII period. A split flywheel was clamped by springs through two friction linings against an inner disc mounted on the end of the crankshaft. Torsional vibrations of the crankshaft are damped by the inertia of the flywheel which rubs against the linings.

Later engines that ran at higher rotational speeds had the balance weights re-introduced to reduce the load on the bearings, but the diameter and length of the journal and big end bearings were increased to stiffen the crankshaft and raise the natural torsional vibration frequency.

Engine ancillaries

The complete 10hp engine is shown here. We have been discussing the mechanical design of the engine: how to control getting the mixture in and out, and the moving parts that transmit the power of the explosions to drive the car. A number of other parts are needed to make the engine work reliably. In this photograph, it was fitted with a Longuemare carburettor to generate the petrol-air mixture. This was the same make as the one fitted to Royce's own Decauville motor car on which the essential dimensions of his own design were based. Later engines had a carburettor to Royce's design. You can also see the governor linkage which controlled the speed of the engine by adjusting the throttle in the carburettor.  We will say more later about carburettors and governors. The water pump was driven from the end of one of the camshafts (at this stage the engine design had separate camshafts to operate the valves) and circulated water through a jacket around the cylinders and the combustion chambers out through a pipe at the top of the cylinder head to the radiator. We will say more later about cooling systems. At the end of the other camshaft was mounted a distributor which consists of a rotating carbon brush moving inside an insulated cover in which two metal contacts were placed. These contacts were connected to the low voltage side of the trembler coil which fed the sparking plugs that ignited the mixture. A mechanism was also provided to rotate the insulated cover to change the timing of the spark. We will say more later about ignition systems.

Lubrication

Early motor car engines used two sorts of lubrication (apart from the normal use of an oil can to oil the pivots of external linkages and rotating parts such as the fan): splash and drip feed. Splash lubrication involved having sufficient oil in the sump - the bottom of the crankcase - for the cranks and the lower bearings of the connecting rods - the big-ends - to dip into it and throw it around to lubricate the bearings of the shafts at the top of the connecting rods on which the pistons pivoted - the little-ends and gudgeon-pins - as well as the walls of the cylinders on which the pistons run, the timing gears and the camshaft bearings, cams and tappets.  

Drip feed involved a small pump, usually mounted on the driver's side of the dashboard of the car, feeding several controllable feeds with sight glasses showing the rate of the oil drops passing through. Each feed went to points on the engine such as the front and rear crankshaft bearings, as well as to the gearbox and clutch withdrawal mechanism. The system was a total loss one. You can see the unit in the picture on the left of the dashboard of a four-cylinder "Light Twenty", just above the clock. It contained about half a gallon (2 litres) of oil and had a sight glass for the level as well as the row of drip-feeds mounted on top of it. 

This picture is a diagram of the 10hp gears and drives, showing the belt drive to the drip-feed lubrication, fed from the back of one camshaft. It also shows the throttle governor on the front of the crankshaft that operated the levers in the photograph of the 10hp engine. The 15, 20 and 30hp cars had the governor mechanism driven from the camshaft as part of a geared shaft that fed back to the dashboard to operate the distributor (in the picture of the Light Twenty this is the box on the dashboard to the left of the drip-feed oil supply tank. To its left is the housing for the trembler coil).

Apart from the mess caused by a total loss oil feed system, the principle of splash lubrication of bearings depended on oil being thrown onto the surfaces of the cylinders or being taken into bearings by surface tension effects: the lubrication was purely dependent on the oil film covering the surface, only a few molecules thick. Under high pressures, such as at the ignition stroke, the load on the big-end bearing on the crank pin could drive the oil out of the contact area, causing metal to metal rubbing. Some car makers used ball or roller bearings instead of plain bearings, because the rolling action of the balls or rollers along their tracks avoided rubbing and allowed splashed oil to reach the tracks easily. These were, however, noisier, complex (you needed to have a crankshaft able to be taken to pieces to install ball-races) and therefore more expensive.  

<

The solution to this was pressure lubrication, first used by Royce in the 40/50hp engine, in which an oil pump takes oil from the sump at the bottom of the crankcase and feeds it through pipes to the bearings of the crankshaft. Other oil supplies were taken to the timing gears, camshaft bearings and, in the case of overhead valves, to the shaft on which the valve rockers were mounted. These pictures from the car handbook show the arrangement on a 1930 20/25hp engine. The middle picture shows the oil pump in pieces: two gears in mesh (R) act as the pump; a spring-loaded pressure relief valve (O) bleeds excess oil to a pipe that feeds the timing gears and overhead valves (green in the right-hand diagram). The main oil flow (blue in the right-hand diagram) goes to an oil gallery (called by Rolls-Royce the "oil distributing main") that feeds to the crankshaft main bearings. Circumferential grooves in the bearings take oil into the hollow crankshaft (red in the diagram) holes in the cranks feed the connecting rod big-end bearings. Holes in the big end bearings let oil run up pipes on or in the connecting rods to feed the little-ends and spray oil under the piston both to lubricate the cylinders and to cool the piston. The oil circulation in a car engine contributes to the cooling of the engine. As oil gets warmer, it gets thinner, which helps it circulate and gives less friction, but also means that the lubricating film can break down more easily under load. Air passing over the surface of the crankcase and the sump cools the oil: the engines of R-R cars were fitted with undersheets which not only kept the engine clean, but controlled the air flow around the bottom of the engine. Removal of the undersheets can make an engine run hotter.

The Cooling System

Most cars use water as a cooling medium, although air cooling has been used for cars, notably the Volkswagen "Beetle" of the 1950s and 60s, and is still used on many motorcycles. The water cooling system has a water-pump, generally made of a rotating impeller driven by the engine inside a circular housing. Water is fed to the centre of the impeller, which pushes it out through a pipe at the outside of the pump housing. It circulates water through the water jacket of the engine and a radiator that cools the water, assisted by a fan. Although some heat is lost by radiation, most is transferred by conduction to the air passing over the radiator matrix. In the picture on the extreme left, you can just see a regular pattern of a gilled tube matrix. The metal of the radiator is usually a good heat conductor such as copper or brass. In this radiator, vertical tubes carry water down from the top (the header tank); they are passed through flat sheets of copper, to which they are soldered. The air passing over the sheets and the tubes cools the water. Other designs of matrix included the honeycomb made up of a stack of circular tubes slightly wider at each end, soldered together at the ends so that water flows between the tubes and air flows through the tubes. The centre picture shows a honeycomb matrix on a 1910 40/50 hp.

As the radiator was designed to give enough cooling on the hottest day, if there was no other means of controlling the air or water flow, the engine could run at a lower temperature than was most efficient. In some cases a thermostat was used to control the water flow: a flexible capsule contained a wax that changed volume on melting; this would open a valve once the temperature melted the wax. In the picture on the right, a 20/25 hp car, the radiator is fitted with air control shutters that on earlier cars were controlled by the driver, who watched the temperature gauge, or in later cars by a thermostat that operated levers controlling the shutter opening. Although late versions of the 40/50hp car used a thermostat to control the water flow in the cooling system, Rolls-Royce went to radiator shutters in 1925 and only went back from thermostatically controlled air flow by shutters to thermostat control of the water flow in the late 1930s. My belief is that this was because antifreeze mixtures to be added to the cooling water were not very reliable in the early 1920s: if the car was not warmed up before moving off on a cold day, by the time the thermostat opened to let warm water from the engine into the radiator, the water in the radiator could have frozen.

The Carburettor

Before fuel injection - where the fuel is sprayed under pressure into the air stream in the area just before the inlet valves (the inlet manifold which supplies air to all the cylinders) - the petrol-air mixture was controlled by a device known as the carburettor. Petrol from the supply tank was fed by pressure, gravity or suction to a float-chamber, rather similar in action to the ball-valve in a water cistern, which maintained the petrol at a constant level. The carburettor is an air pipe feeding the inlet manifold, with the flow of gas controlled by a throttle, usually a circular disc mounted on a shaft through the pipe, but a piston valve on early cars. Within the pipe is a section that is restricted in area called a "venturi". This makes use of the Bernouilli effect - the same phenomenon that gives lift to an aeroplane's wing. Within the restricted zone, the air flows faster, moving its molecules farther apart. A gas with its molecules farther apart is one at lower pressure, therefore in this area the air is at lower pressure. A hole in the wall of the tube at this point will therefore experience suction; if a pipe is inserted at this point and attached to the petrol supply from the float chamber, petrol will be sprayed into the air stream. By making the end of the pipe a controlled size hole (the jet), the amount of petrol sprayed into the air in the tube can be the right amount to cause the mixture to burn when a spark is applied. Most carburettors have a needle valve inserted into the jet that can be moved in and out to adjust the amount of petrol flowing. If not enough petrol is supplied, the mixture is weak and if too much, it is rich or strong.  Unfortunately, as the flow of air through the tube increases, the amount of petrol sucked into it increases faster than the increased air flow rate, making the mixture rich. In the case of Royce's carburettor, he built in an air valve, which was a piston in the side of the tube, arranged so that the pressure in the tube was applied to the top of the piston and normal atmospheric pressure was applied to the lower side. The lower pressure in the tube resulting from higher air flow made the piston rise and open an additional supply of air to compensate for the additional amount of petrol being sucked in. The diagram shows the air valve in operation. Later carburettors, such as the "S.U." used an air valve to control a tapered needle valve in the jet, but in this case it was designed to lift the needle to enrich the mixture at low air flow rates and lower it at higher flow.

The pictures on the left from the R-R car Service Instructions show a later version of the Royce carburettor, used on the 20 and 20/25hp cars between 1923 and 1934. The left hand picture shows the "butterfly valve" throttle at top left, the float chamber with needle valve to control petrol level, and the starting carburettor. When a car is cold, the rate of evaporation of the petrol is slower: even if the correct ratio of petrol and air is mixed, a lot of the petrol does not become vaporised, but remains as liquid coating the sides of the inlet manifold. To achieve a richer mixture for starting, early cars used a choke which restricted the air flow ahead of the jet so that the suction from the engine pulled in more petrol. The starting carburettor was also designed to give a richer mixture, and operates through a small tube that feeds the mixture into the manifold after the throttle. The engine is started by opening the tube, with the main throttle shut. Once the engine is running, the throttle is opened to bring the main carburettor into operation, usually with the mixture set rich until the engine warms up. The diagram of the early carburettor in the previous section shows that the area around the throttle had a water jacket, to keep the mixture warm: evaporation of petrol causes cooling and in certain conditions the carburettor can partially block with ice from water vapour in the air. Later cars used engine heat by a water jacket around the manifold fed from the cooling system to aid the evaporation of the petrol, or by having a "hot spot" where the inlet manifold was in contact with the hot exhaust manifold (which collects the exhaust gas from the exhaust valves and feeds it to the exhaust pipe through a silencer). 

The right hand picture shows that the carburettor had two jets: one for slow running that is permanently in operation to maintain the engine's idling speed, and a main jet. Each jet has its own venturi, but the main jet is the only one provided with the air valve to give extra air as the flow increases when the throttle is opened.  The air valve is arranged so that the piston that controls the extra air rises slowly to its balance point when the throttle opens. This makes the mixture fed to the cylinders initially richer when the throttle is opened quickly to accelerate the car.

The Ignition System

The petrol-air mixture in the cylinder of a car engine is ignited by an electrical spark across a gap between two conductors. Very early cars had other means of ignition: hot tube was operated by a unit like a blow-lamp heating the end of a tube that went into the cylinder; ignition depended on the mixture reaching the red hot tube at the top of the compression stroke, but this only worked reliably if the inlet port and the tube were well separated: some engines had a moving cover for a chamber holding the end of the tube, which opened at the top of the compression stroke. This was a method originally developed for stationary gas engines. Low-tension ignition had a pair of contacts inside the cylinder carrying current and moved apart by a linkage to create a spark as they separated. Keeping the linkage gas tight was a challenge.

All modern ignition systems use high tension ignition in which a high voltage pulse is fed to a sparking plug. A sparking plug is screwed into a hole in each combustion chamber. It consists of a central conductor surrounded by a ceramic heat-resistant insulator mounted on a metal body carrying a metal finger that overlaps the end of the central conductor and forms the spark gap. The gap is adjusted by bending the finger. The other end of the conductor carries a terminal to which the electricity is supplied. Various different methods have been used to generate the electricity to produce the spark at the right point of the engine's cycle. Early Rolls-Royce cars were fitted with a trembler coil. Later ones had coil ignition or a magneto. Some were fitted with both coil and magneto - "dual ignition" with two sparking plugs in each cylinder. Some had a magneto for stand-by use. All of these devices use the principle of magnetic induction discovered by Michael Faraday to generate the pulses of high voltage electricity, but use different methods to change a magnetic field's strength to induce an electric current.  

A trembler coil is like an induction coil that I remember from school Physics lessons producing long sparks and coloured light in gas discharge tubes. For more information try this link: http://www.sparkmuseum.com/INDUCT.HTM It acts as a transformer with a primary coil of a few thick windings wrapped around a soft iron core. This coil is in circuit with a battery and a sprung vibrating arm which makes and breaks the circuit at a pair of contacts. The vibrating arm works on the same principle as a traditional electric bell: it carries an iron disc that is attracted by magnetism towards the end of the iron core against the spring loading of the arm when current flows through the primary coil. The movement opens the contacts at the end of the arm, which breaks the circuit, stopping the magnetism. The spring returns the arm, the contacts close again and current flows again through the coil so that the cycle repeats. There is an adjuster on the arm to control its movement to make it cycle rapidly - it makes a buzzing sound. Around the primary coil is another coil of fine wire with many more turns (the secondary coil), making a step-up voltage transformer. The build-up and collapse of the magnetic field in the core generates a sequence of high voltage pulses in the secondary coil, which create sparks at the gap of a sparking plug connected across the ends of the coil. To distribute the voltage pulses to the different plugs in a car engine, and at the right time in the cycle, there is a rotating insulated wheel, the rotor arm, carrying a conducting segment connected to the high voltage supply by a contact, usually a carbon brush in a unit known as a distributor, shown in the right hand picture. The conducting segment moves past a set of carbon brushes, one for each cylinder, to supply the voltage to the appropriate sparking plug at the beginning of its firing stroke. When stationary, following cranking the engine round to suck in fuel/air mixture, if the trembler coil is switched on, a shower of sparks will occur in the cylinder that is on the firing stroke, which generally will start the engine (this was known as "starting on the switch"). The trembler coil is effective at low speeds, but at high speeds there is a risk that the frequency of voltage pulse production will not reliably coincide with the frequency that the engine needs sparks - the car will tend to prefer to run at multiples of the speed of the trembler coil vibration. In the left hand picture, the extreme left hand box on the dashboard contains the trembler coil and the box next to it with the circular window contains the distributor. In the picture of a 1912 40/50 engine, the trembler coil is on the engine side of the dashboard (the wooden box with a metal top) and the distributor is on the right of the picture in the position on the engine in which it remained for all 6 cylinder Rolls-Royce cars until the Second World War.

The magneto was introduced because it was more efficient at high engine speeds. In early Rolls-Royce 40/50hp cars dual ignition was provided by a trembler coil to one set of sparking plugs and a magneto to a second set. A magneto works by having the same arrangement of a primary coil of thick windings surrounded by a secondary coil of fine windings on an iron core, but placed to rotate between the poles of a strong magnet. The movement of the windings through the magnetic field generates a current in the primary coil, which is highest twice in each rotation, with the current flowing in alternate directions. In circuit with the primary coil is a contact breaker mounted on the coil shaft, which has an arm riding on a lobed cam that causes contacts to open at the point where the coil is moving through the strongest magnetic field and the current highest. The current flowing in the primary coil induced by the magnetic field collapses, generating a voltage pulse in the secondary coil to produce a spark at the sparking plug. The contact breaker has two cam lobes so that it generates a spark at both of the strongest magnetic field points of its rotation. For a 6 cylinder car, it needs to produce a spark three times for each crankshaft revolution (with the Otto cycle, each cylinder fires once in two turns of the crankshaft), so the magneto is driven at one and a half times the crankshaft speed. A distributor is driven at half the coil shaft speed through gears to take the voltage pulses to the individual plugs. (The illustration is of a 4 cylinder car magneto which has two lobes on the contact breaker cam and four plug connections on the distributor and is driven a half crankshaft speed - it comes from an Austin "Twelve" handbook). The main feature of the magneto is that it does not require a battery, which was a good situation when batteries were unreliable or where the no means of charging the battery was provided on the car. Also, because increasing the rate of rotation of a magneto increases the current flowing in the primary coil, the spark gets better the faster it rotates.

Finally, the conventional ignition coil and distributor can be seen as a combination of the two devices previously described. The engine drives a contact breaker that breaks the primary coil circuit at the firing stroke for each cylinder, producing a voltage in the secondary coil that is passed through a distributor to the correct sparking plug. Across the points of the contact breaker in a coil system is a condenser (capacitor) designed to produce an oscillating current in the primary coil during the collapse of the current, giving a very rapid multiple spark. The contact breaker is mounted beneath the distributor and driven at one-half of engine rotation speed. For a 6 cylinder car there are six lobes on the cam to give six sparks in each rotation of the cam and the distributor rotor: three sparks for each turn of the crankshaft. Until WWII the Rolls-Royce ignition coil had a ballast resistance in series with the primary coil to limit the current flowing through it. There were two reasons for this: one was that the resistance of the thick coil windings was lower than needed and lower than achievable even if they were made of thinner wire of higher resistance (later coils wound the primary on the outside of the coil, rather than around the core; the extra diameter meant that more wire was used, increasing its resistance); the second was that the wire of the ballast resistor was chosen to be of a material that had low resistance when cold and higher resistance when hot. This gave a higher current through the primary coil for a few seconds, just at the time when the car was being started and needed a strong spark because the starter motor current would lower the battery voltage. Modern cars still use coil ignition although the timing of the spark and the breaking of the current is now done using a solid state electronic circuit.

Engine Controls

A modern car has only two controls for the engine: the accelerator and the clutch (for a manual gearbox car). The accelerator is usually connected to the throttle and controls the flow of petrol-air mixture to the engine. The clutch connects the engine to the gearbox and the drive to the rear wheels: it is used to take the load off the gear-wheels when changing gear and to engage the engine gradually when moving off from a standstill. I will describe the construction of the clutch under the section describing the transmission. Early cars had other controls: a hand throttle, ignition timing and petrol-air mixture controls. There was also a control mounted on the dashboard for turning on the starting carburettor.

The hand throttle was used to set the engine idling speed when the car was stationary: in the early R-R cars it operated through a governor. This was a device originally used in stationary steam engines as a speed control: rotating weights moved out under centrifugal force as the speed increased and moved a collar on the rotating shaft that gradually closed a valve controlling the steam supply. It worked against a spring. It compensated for variation in steam pressure and the power demand to the engine. The governor weights are just to the right of the safety valve in the picture of part of a traction engine. The speed of the engine could be adjusted by varying the tension on the spring.

In early cars, because the carburettor did not control the mixture well (if you remember, the mixture on a simple single jet carburettor would vary from lean at slow air flow to rich at fast), the governor would constantly vary the throttle opening to adjust for misfiring of the engine, especially if it was running lean at idling speeds. The speed of the engine was set on the hand throttle control, which was connected to the throttle through a spring acting against the force exerted by the governor weights trying to close the throttle. The accelerator pedal over-rode the governor, opening the throttle directly. The governor could also be used as a primitive cruise control, but did not have an automatic cut off when the brakes were applied. In the picture of the early distributor for a 40/50hp fitted with trembler coil, the governor weights are mounted on the shaft beneath the distributor and move the control lever by a collar on the shaft. When the distributor was redesigned for coil ignition, with the contact breaker mounted on the shaft below the rotor arm, the collar driven by the governor weights was also used to rotate the upper part of the shaft so that as the speed increased, the spark generated by the contact breaking was earlier. The time the explosion of the mixture in the cylinder takes to develop full pressure on the piston is the same at any engine speed, therefore the faster the engine speed, the earlier the spark has to be for the explosion to reach the piston at the right time to get the maximum energy from the hot gases. Instead of firing just after the piston is at its highest point in the compression stroke, at high speeds the spark starts the explosion before the piston reaches its highest point.

The ignition control, marked Early and Late on R-R cars, rotates the body of the distributor and the contact breaker to vary the timing of the spark. Generally the engine would run with the lever nearly all the way towards the Early setting, but, for starting the car using the starting handle, it was set fully Late to avoid the engine trying to go backwards when it fired: otherwise there was a high risk of breaking the fingers or wrist of the person turning the handle (never have your thumb on the opposite side of the handle from your fingers when starting: it is very easy to break your thumb if the engine backfires). Once started, the lever would be brought up towards Early. The trick of starting on the switch was still possible: the ignition control rotated the distributor through a wide range such that it was likely to open the contact breaker and generate a spark. In the cars fitted with a governor, when the ignition was switched off and the engine came to a stop, the governor would open the throttle to compensate for the drop in speed such that the cylinders would have a good charge of mixture ready for starting on the switch. An old Rolls-Royce car in good condition can often start on the switch after standing for an hour.

The mixture control operated on the main and slow running jets through levers that moved needles up and down in the jets (marked in the illustration). Generally the control was set to Strong when starting the car on the starting carburettor; the hand throttle would be moved up to bring the main carburettor into operation and the starting carburettor turned off. As the engine warmed up, the control would be moved to the centre of its range. The car's handbook suggested moving the control slightly towards Weak for touring, but with the ignition set fully advanced because a weak mixture burns more slowly. For hill-climbing and when slow running, it suggests setting slightly towards Strong.

The Transmission

This describes the connection between the engine and the wheels. It is made up of the clutch, gearbox, propeller shaft, crown wheel and pinion, differential, half-shafts and hubs.

The clutch is operated by a foot pedal and is used to take up the drive from the engine smoothly when moving off from a standstill and to take the load off the gears in the gearbox when changing gear. The pedal operates a lever that moves a spring-loaded plate away from the back of the engine flywheel. Although the design varied through the early period (the 40/50hp had a "cone clutch" in which the friction material was mounted on a conical surfaced drum), Rolls-Royce clutches usually had two sets of friction linings: one attached to the flywheel and the other set on a ring moved by the clutch pedal, with an intermediate slotted steel plate between. The drawback of this design was that heat generated when the clutch slipped would be absorbed in the intermediate plate and have nowhere to go except through the lining material, which was a poor conductor. Often the plate would distort. Later clutches worked with the friction material attached to the plate and engaging with the flywheel surface: the metal of the flywheel would dissipate the heat from any slipping. One of the problems with early cars was that oil seals to stop oil getting out around shafts coming out of the oil filled crankcase and gaskets to seal mating joints on castings were not very efficient. Rolls-Royce cars had an "oil thrower" that was a threaded section between the flywheel and back crankshaft journal designed to screw the oil back in as the crank rotated: there was no physical seal. It was therefore not unknown for clutches to get oily and slip.

The gearbox contains a number of combinations of gearwheels to give different rotational speeds between the input shaft and the output shaft. Generally it was arranged that the input shaft rotated more rapidly than the output shaft except in the direct drive ("top gear") where the input and output shafts were coupled together. The picture shows the three-speed gearbox of an Austin "Seven" car, which was rather simpler than the R-R gearbox, but explains the principles. In this, the output shaft is splined and carries two sliding gears (splines are generally square edged straight slots machined along a shaft so that the slots are the same width as their separation and allow something with a hole similarly machined on its inside to slide freely on the shaft and to be driven by rotation of the shaft). The gear "A" constantly drives gear "F" on the layshaft "N" which carries the wheels for 1st "H" and 2nd "G" gears and the pair of wheels "I" and "J" for Reverse. The lowest gear, 1st, is engaged by moving gear "E" into mesh with "H", 2nd by moving "D" into mesh with "G". Top gear couples the input and output shafts "P" and "M" by moving gear "D" which has a dog clutch "C" on its face that engages with another dog clutch "B" on the face of gear "A" . Reverse is engaged by moving "E" into mesh with "I". "Selectors" control the movement of the gears. They are levers that run in the slotted rings on the side of gears "D" and "E" and operated by the gear lever moving in a "gate".

The picture shows a gear lever (and hand brake lever). When not moved into any of the slots, the gearbox is in "Neutral", with no drive being transmitted. The lever is moved into the marked gates for each of the four speeds.  If Reverse is required, the button on the top of the lever must be pressed to release a catch that allows the lever to be moved into the gate "R". In early cars the gear lever was known as the "change speed lever" because, with a governed engine running at fixed revolutions, changing gear changed the car's speed.

The type of gearbox illustrated is what was known as a "crash" gearbox, because there was no means of synchronising the relative speeds of the gear wheels when changing gear. The technique of "double declutching" was something that many drivers never mastered and because a feature of R-R cars was the flexibility of the engine, some would start in whichever gear they thought the car would be able to use for the whole of a journey: if the route was level, they would start in top gear. If they came to a hill, they would stop the car and start off in the appropriate lower gear. Later gearboxes used gears in constant mesh carrying dog clutches on their faces, engaged by the selector moving a dog clutch on the splines of the output shaft into contact with those on the gear wheel. This allowed helical fine teeth to be used on the gears to reduce noise and allowed the wearing areas to be on the dog clutch faces, which could also have coarser teeth. Later, in the early 1930s "synchromesh" gearboxes were introduced: these had cone clutches that would meet before the dog teeth engaged and synchronise the gear wheel and output shaft rotation speeds.

With a crash gearbox, it was usually not necessary to double declutch when changing to a higher gear if a pause was made to let the engine slow down to match the rotation speed of the next gear, but the technique was essential for changing to a lower gear. I quote a 1930 handbook for the 20/25hp: -

"When changing up, it must be borne in mind that in order to bring the gears into silent engagement, a perceptible pause must be made with the gear lever in the neutral position and the clutch withdrawn. This will give the clutch shaft time to slow down until the gears to be engaged are rotating at a relative speed approximately equal to that which will obtain when they are in mesh. The lever can then be moved into the required position without effort or noise, the clutch re-engaged and the accelerator pedal depressed.

When changing down the converse is the case, i.e., the speed of the clutch shaft requires to be increased before engaging a lower gear.  This can be done by double-clutching, which consists in quickly letting in the clutch and speeding up the engine while the gear lever is in the neutral position.  The clutch must then be withdrawn again and the gear lever moved to the next lower gear position.  It is better to speed up the clutch shaft in this manner rather too much than too little, as the period which must necessarily elapse before the gear is engaged will result in a slight decrease in the clutch shaft speed, and the driver is able to feel the way into the gear and make a good change.  On the other hand, if the engine is not speeded up sufficiently, either the gear will be missed or a noisy change effected."

With some cars, drivers find it may be easier not to declutch completely, controlling the engine firstly to take the load off the gears to be able to shift into neutral, then raising the engine speed to that for the next gear down and controlling the engine speed with the accelerator while feeling the engagement of the gear.

The propeller shaft. This takes the drive from the gearbox to the rear axle. Because the rear axle is sprung, the propeller shaft needs to move through an angle while maintaining free rotation. This is achieved by fitting universal joints at each end of the propeller shaft (later versions of the 40/50hp had a torque tube, which was a tube rigidly attached to the rear axle and pivoted on a spherical bearing mounted just behind the gearbox, with a single universal joint inside the sphere - the rear axle shown below right is an example of an axle with torque tube). The earlier picture of the Austin "Seven" gearbox shows a rubberised fabric disc universal joint, known as a "Hardy disc". Rolls-Royce used this type of universal joint for the drives for the water pump and dynamo on engines in the 1920s and 30s. The picture here shows the front universal joint of a 1930 R-R 20/25hp car. The joint is made of two "Y" shaped pieces with round stubs at the tip of each branch of the "Y" that face each other 90 degrees apart and engage with bearings inside a ring. The ring contains an oil bath to lubricate the joint. Later universal joints used very thin roller bearings - "needle rollers" to transmit the load. In the picture there is a sliding joint behind the universal joint to allow for the variation in distance of the rear axle from the gearbox in the course of its movement on the springs.

The rear axle. This divides the power from the engine between the two rear wheels via drive shafts called half-shafts. Generally the rotation speed of the engine is around four times faster than the rotation rate of the wheels, so a reduction gear known as the crown wheel and pinion is used, shown on the left hand picture of a demonstration R-R rear axle, to convert the rotation of the propeller shaft, which runs in line with the motion of the car, to the rotation of the axle at right angles to its motion. Attached to the surface of the crown wheel is a set of gears that allow for the different rates of rotation of the rear wheels when the car goes round a corner; these are known as the differential gears. Some early cars had a gear on the inner end of each half-shaft that engaged with one of two sets of planetary gears in a housing fixed on the crown wheel. The planetary gears would be stationary unless the rate of rotation of one axle was faster than the other, when one set would revolve in the opposite direction to the other. The two rings of planetary gears and their mounting were heavy, so later differential gears used four bevel gears in mesh, as shown in the right hand pictures of a demonstration unit, with the output to the half-shafts from two opposite gears either on splines on the half-shaft or formed on the inner end of each half-shaft. A cage that held the bevel gears was fixed to the crown wheel: in the pictures, the ends of the blue shaft carrying the central black rod would be held by the cage. All four black rods would rotate together and the gears would be stationary within the rotating cage when the car was moving straight forward. In the lower picture, the effect of making a very sharp left hand turn is shown: the left hand half-shaft has not rotated, but the cage has been driven to rotate through a quarter turn and the bevel gears have driven the right hand half-shaft to turn through half a turn.

This picture of a non R-R rear axle shows that the axle casing carries the rear springs and the rear braking mechanism. The part of the axle casing on either side of the differential housing is called the axle tube. This particular design of axle has the propeller shaft mounted inside a tube: the "torque tube", which is attached to the rear of the gearbox via a spherical joint surrounding the front universal joint. This allows the axle to move up and down with the springs, but prevents it twisting the springs with the torque of acceleration or braking. Later 40/50hp and "New Phantom" cars had this arrangement, but it increased the unsprung weight of the axle and a normal open propeller shaft was used in later cars, with the twisting controlled by the rigidity of the springs, or, later by a steadying rod known as a "Panhard rod", named for its inventor. The Royce 10hp car was fitted with one because its springs did not resist twisting of the rear axle. 

I mentioned "unsprung weight": the axles and wheels of a car are not sprung (except slightly by the flexibility of the tyres). The energy passed to the axles on hitting a bump is transmitted to the chassis via the springs and is felt by the passengers to an amount dependent on the softness of the springs and the firmness of the damping by the shock absorbers, but the lighter the wheels and axles are the lower is the energy transferred to the rest of the car and needing to be absorbed by the springs and shock absorbers. Many improvements in the ride of motor cars have resulted from reducing the weight of the wheels and axles while maintaining their strength and rigidity.

The wheel hubs. These are mounted on the ends of the axles. They rotate and are retained on ball-races mounted on the outside of the axle tubes of the rear axle and on the stub axles of the front axle which turn on a king-pin when the car is steered. Some R-R hubs are shown in cross-section in the picture. The rear hubs are driven by the half-shafts which are splined on their outer ends and carry driving dogs which engage with splines on the inside of the hubs. The arrangement means that no side loads are passed to the half-shafts - the axle is described as fully-floating. The hubs carry splines on the outside which engage with others on the inside of the road wheels. A wheel nut screws onto threads at the outer end of each hub and clamps the wheel against a coned surface at its inner end. The brake drum is mounted on the inner end of the hub. Later R-R cars and most modern cars use nuts screwed onto bolts mounted on a flange around the hub to hold the wheels on, or hexagon-headed bolts that screw into threaded holes in the hub.. Where necessary to stop them unscrewing in use, left-hand threads have been used on one side of the car.  

The Braking System

Early cars generally had brakes on the rear wheels only: traffic was light, so the need to stop quickly was small. It was also an engineering challenge to provide a mechanism to operate brakes on wheels that were steered. It was recognised early that two independent braking systems were desirable: one, hand operated, that could be locked on for parking use, the other, foot operated so that the driver's hands were free for steering. In the 10hp Royce car, the hand brake lever operated brakes that were internal expanding shoes in drums on the rear hubs. The force applied to the brake on each side of the chassis was fed via a brake compensator, which was a miniature differential gear designed to make sure that the same load was applied to each brake, even if the play in the connection was different from one side to the other. The foot brake worked a transmission brake mounted behind the gearbox that applied braking effort through the propeller shaft and rear axle gears and half-shafts. Many early cars had transmission brakes, but their disadvantages were that they applied high loads on the rear axle gears and generated vibration because early universal joints were not that smooth in transmitting load. On a wet surface, if the brake locked, the differential could result in the rear wheels skidding in opposite directions, contributing to what was known as "the dreaded side-slip." In development of the 40/50hp car, the transmission brake was replaced by another set of brake shoes in wider rear brake drums, a similar layout to the picture above in the section about the rear axle.

When travelling forwards, the force of braking tended to push the front of the car downward, transferring the load onto the front wheels away from the back. With rear wheel brakes, this meant that rear wheels' grip on the road was reduced, reducing the maximum braking possible before the wheels skidded. It was recognised that adding front wheel brakes would greatly improve the ability for a car to stop. As for many features in a car, the engineering challenge was to find a way of controlling the braking effort on axles that were steered.

The operating mechanism for a 1930 R-R car is shown in the pictures at the left and included front wheel brakes, introduced around 1924 by Rolls-Royce, rather later than many other makers. The rear brakes were operated by cables to levers mounted on the brake back plates that carried the mechanism operating the brake shoes inside the drums. A double cam was mounted on the end of the operating shaft that pushed the two brake shoes apart, bringing them into contact with the drum. The shoes were pivoted at their opposite ends and fitted with return springs. Cables to the front brakes operated levers (B in the right hand picture) on shafts running to the outer ends of the front axle beam. Levers at the ends of the shafts operated rods running to other levers at the top of the brake back plates that worked the brake shoe actuators. The geometry of the levers was arranged so that the clearance remained nearly constant for all steering positions. As for the rear brakes, compensators were fitted to the front brakes to allow for wear.

The Rolls-Royce four-wheel braking arrangement is shown in the picture on the left (the front of the car is to the left). It shows that the handbrake was still a separate system working on the rear wheels. Additionally, the system was fitted with a mechanical servo to reduce the braking forces needed. In essence this was a rotating disc driven by a worm drive from the gearbox output shaft against which was pressed another disc carrying a friction brake lining. Levers attached to the second disc would be rotated by the frictional forces to operate the brakes. The footbrake pedal both operated the pressing of the second disc and levers for direct braking force to the rear brakes. The servo levers added to the force applied to the rear brakes and operated the front brakes. At the time the system was designed, it was felt that the ability to steer in a skid was a good thing, so no more than one-third of the braking effort could be applied to the front brakes. Also, if the rear wheels locked, the servo disc would stop turning and prevent more braking at the front, however hard the brake pedal was pressed. Later R-R cars increased the amount of braking at the front. Post WWII cars used hydraulic operation of the front brakes (hydraulic brakes overcame all the issues of geometry of front wheel braking systems by feeding the operating pressure through flexible tubes to pistons working the brake actuators) and went up to around 50 per cent of the braking at the front when the system was superseded in 1965 with the introduction of the Silver Shadow with modern vacuum servo hydraulic brakes using disc brakes instead of drum brakes. A vacuum servo has a flexible disc mounted on a shaft connected to the brake pedal inside a chamber connected to a vacuum supply (often the suction from the engine induction system is used) controlled by the movement of the pedal shaft. As the pedal is pressed, the suction is fed to one side of the disc which then pulls on the shaft to increase the pressure on the brakes. The shaft operates a piston inside the brake master cylinder which pushes the hydraulic fluid through the brake pipes.

Until the introduction of anti-lock braking systems, most other car makers fitted a brake limiter to the rear brakes: their philosophy was that it was better to have maximum braking with the front wheels locked and the back wheels rotating so that you had the accident head-on rather than sideways.

Disc brakes, fitted at the front on most modern cars and all round on many, use a U shaped calliper carrying on one side a hydraulic piston bearing against a brake pad with another pad on the other side, both clamping a circular disc on the wheel hub. The calliper is mounted on the axle and has sufficient free movement in and out to allow the pressure of the piston to clamp both pads by reaction.

Steering

At an early stage, cars used a steering wheel to control the direction of travel. Some very early cars used tiller steering with an arm held by the driver like the tiller of a boat, which was directly coupled to the wheels: this was satisfactory while cars remained very light in construction, but when cars got heavier a steering wheel operating through a gear system in a steering box allowed greater turning force to be applied. Although the amount of movement was small, the forces needed to turn the wheels when the car was stationary meant that gears with a large area of contact were needed. One arrangement used on most R-R cars through the 1920s and 30s was worm and nut: the shaft attached to the steering wheel carried a worm drive at the end, with a long nut on it that carried two stubs engaging with a lever at the top of the steering drop arm. The drop arm had a pivot shaft that led through the side of the chassis to an arm that moved the side steering tube forward and backward to operate the steering arm that turns the wheels. You can see this in the right hand picture of the front brakes above. The wheels are connected by a steering cross-tube pivoted on arms projecting back from the bottom of the king-pins to couple them together.

Later R-R cars used a cam and roller, shown in cross-section in the picture. The cam is more in the form of an hour-glass worm with spiral teeth on a curved surface. The roller has a profile that engages with the teeth of the cam over a broad surface of it, and rather than sliding on the cam in the way the nut slides on the worm, it rolls as the cam is turned. The roller is on a pivot mounted eccentrically on the shaft holding the steering drop arm so that it turns the shaft as it moves along the cam. The rolling movement avoided "sticky steering" suffered by the worm and nut design. Sticky steering was caused by the oil being squeezed out of the surfaces under pressure and causing seizure: most often happening when a steady force was applied to the steering wheel as in going round a roundabout. The car would continue to turn until a hard push was given to bring the wheels back to straight ahead.

Most modern cars use rack and pinion steering. This replaces the centre section of the steering cross-tube by a straight gear rack driven by a pinion gear turned by the steering wheel.

Before the motorcar, it was realised that the steered wheels must be arranged to turn by different amounts for a vehicle to travel smoothly round corners. In horse-drawn coaches, the shafts pulled by the horses were attached to a pivoted front axle with the wheels mounted on each end. When turning a corner, the circle traced by the inner wheels is smaller in diameter than that followed by the outer wheels, therefore the inner steered wheel needs to turn more than the outer one to be at right angles to the centre of the circle being turned to avoid the wheels scrubbing sideways. It was invented by a German called Langensperger in 1816, but was patented in London in 1817, by a man named Ackermann whose name has become associated with it. Langensperger worked out that by making the ends of the arms holding the steering cross-tube narrower than the pivot points of the wheels, the angle moved by the inner wheel when turning was increased. He developed a simple calculation of this angle dependent on the wheelbase (distance between front and back wheels) and the track (separation of the wheels). At higher speeds, the simple Ackermann calculation does not solve all the problems. Other changes to the angles of rotation of the wheels have been made to improve a car's stability: castor angle relates to arranging that the centre that the wheels pivot about is ahead of the point of contact with the road. This tends to help the wheels return to the straight ahead position, making the steering self-centring. This can also be helped by arranging that the wheels are not quite parallel in the straight ahead position, but "toed-in" so that they are closer together at the front edges.

The Suspension

The wheels and axles of a car are mounted on the chassis flexibly so that bumps in the road are not felt strongly by the passengers. This is normally done using springs. On early cars these were leaf springs similar to those used on horse-drawn vehicles. These are shown in a number of pictures in the earlier sections. A pile of curved steel strips of reducing length is clamped at its centre and mounted by U-bolts to the axle. The top leaf of the spring has eyes formed at the end into which drilled bushes are fitted through which shackle pins pass. The shackle pins are attached to the chassis, usually rigidly at one end of the spring leaf and via a swinging arm at the other. This allows for the change in length of the spring leaf as its curvature varies under different loads. The location of the axle in the correct position is by the section of the top leaf from the rigid end shackle to the centre of the spring.

A spring when deflected and released tends to oscillate. The action of the spring leaves rubbing against each other tends to damp down the oscillation, so early cars depended on this to stop them bouncing about. However, as cars increased in weight and because lubrication of the spring leaves stops them wearing, additional damping was necessary. This was achieved with shock absorbers. Early cars used friction shock absorbers, which carried a friction material like brake lining between spring loaded metal discs attached one to the chassis and the other to the axle. Greater damping was achieved by interleaving several discs with friction material. Later cars used hydraulic shock absorbers in which a lever connected to the axle moved a piston that pumped oil through a small hole to give viscous damping. The picture shows a Rolls-Royce shock absorber. The blue parts are the lever that moves the piston, the piston and the square shaft that is linked to the arm connected to the axle. The unit at the right hand end varies the pressure on a valve that controls the damping force by varying the size of the restrictor controlling the oil flow. This unit was fed by oil pressure from a pump on the gearbox to vary the overall damping amount: this was the "ride control". Many modern cars use telescopic hydraulic shock absorbers in which a shaft attached to the axle operates the piston directly.

Early cars used beam axles with leaf springs mounted close to the ends of the axle next to the brake back plates. Even so, when one wheel hits a bump, the axle tends to turn and compress the spring at the opposite end of the axle by a small amount. This is not a major issue for the rear axle, but for the front axle, the springs and the chassis on which the shackles are fixed have to be further away from the axle ends in order to allow the wheels to turn without rubbing against them. This means that when one wheel hits a bump, the axle and the wheels tilt more. A rotating wheel at speed is like the flywheel in a gyroscope, which, if tilted in one direction will try to turn at right angles, an effect called "bump steering". One cure for this is to have independent front suspension: instead of a beam axle, the hub is supported on parallel arms that are mounted on the chassis and can move up and down while keeping the hub upright. Various systems were used in developing this type of suspension: some cars had two transverse leaf springs, one above the other and clamped to the chassis at their centres, with the hubs supported between their ends so that each side of the spring worked independently. Most modern cars use coil springs, often surrounding a telescopic shock absorber: essential because coil springs have very little internal damping. Coil springs also can be easily deflected sideways, so, unlike leaf springs, other devices are needed to make sure that the axle is well located, moving only in directions the car designer wants.

Axle Location

As cars got heavier and acquired things like front wheel brakes, the loads on the axles and suspension could cause them to move in ways that made the car difficult to control and more uncomfortable in ride. Car designers have found that ride and handling can be improved by several methods: improve the axle location; improve the stiffness of the chassis; and reduce the "unsprung weight" - this is those parts of the car that get the direct shocks from the road: the wheels, hubs, brakes, axles and part of the springs.

The addition of front brakes increased the unsprung weight and put the axle under new stresses, notably twisting of the axle by the braking reaction between the brake shoes and drum and thrust on the springs from the deceleration of the car when braking. The first tended to bend the top leaf of the spring into an "S" shape and could affect the geometry of the wheels by affecting the angle of the king-pins. Royce set the leaf spring rigid pivot at the rear of the main spring leaf, so that the braking load was taken by the rear half of the spring in compression and transmitted to the chassis. Some makers had the rigid pivot at the front so that the load put the spring in tension, but meant that front end of the chassis which was the thinnest and least rigid part was under greater load. To avoid the twisting of the axle, Royce made the link between the shock absorber arm and the axle into an "A" shape with the bottom of the "A" attached to the axle, preventing the axle twisting by pulling horizontally on the shock absorber arm. A similar effect of twisting arose at the rear axle, both under braking forces and from the torque of the car accelerating. In the early Royce cars and some later ones, a radius rod was fitted running from the rear axle casing to a chassis cross-member near the gearbox to prevent the axle twisting.

All of the modifications to deal with the braking forces tended to increase unsprung weight and at the same time tyres were increased in cross-section: so-called "balloon tyres" were run at lower pressures and although they made the ride softer, tended to make the handling less precise. To compensate, the springs tended to be made stiffer, hardening the ride once more, and the chassis was reinforced to improve stiffness. Only in the second half of the 1930s, when independent front suspension with coil springs was introduced and gave a reduction of unsprung weight, and because deeper and therefore more rigid chassis side members were needed to support the suspension arms, was it possible to soften the ride again to the situation that existed just before front brakes were introduced.

With the introduction of independent suspension, it was also found necessary to control the way that the car rolled in corners. A characteristic of early cars was that as it rolled on cornering, the weight was transferred to the outer wheels and turned the car more sharply into the corner, an effect called "roll oversteer". If one entered a corner too fast, this would make it essential to reduce the degree of turn of the steering wheel to avoid the car going into a spin. An "anti-roll" bar would control this: a steel rod with ends bent forward and attached to the axle spring mountings. In roll, when one wheel would compress the spring and the other expand it, the rod would resist being twisted. If both wheels moved in the same direction, as when hitting a bump across the road, the rod would simply turn and offer no resistance.

Throughout the development of Rolls-Royce cars, the chassis was gradually strengthened because it was recognised that rigidity improved the control of the car, as well as reducing loads to the body. Until WWII all Rolls-Royce cars were fitted with coach-built bodies, usually with metal panels on a wooden frame, although some coachbuilders introduced metal framing in the late 1930s. Rolls-Royce did not fit its own bodies until after WWII with the introduction of the Bentley Mark VI and Rolls-Royce "Silver Dawn" using bodies manufactured by The Pressed Steel Co. Ltd. Although some of the coachbuilt body structures improved a car's rigidity (later 1930s cars with front mudguards and running boards as single units bolted also to the rear mudguards were more rigid than cars with these parts separate), the main stride forward was the introduction of the unitary construction "Silver Shadow" in which the body frame contributed to the strength of the whole structure. This and the need to crash test each different motorcar design resulted in the death of the traditional coach building industry.  

To finish this account, the picture at far left is of the chassis of the Silver Cloud, which in longer form was the basis of the later Phantom V and VI, the last coach-built R-R cars. You can see the cruciform member that braced the chassis to prevent distortion under load. Early chassis were riveted and with "U" section frames, but this was welded and of box section to increase its strength. The right hand picture is of the Silver Shadow, introduced in 1965 with unitary or monocoque construction of the body, and shows the mechanical components that were attached to it laid out in front. This is another of those wonderful sectioned illustrations of cars that motoring magazines have excelled at - here a Vic Berris picture originally published in "Autocar". You can work out for yourself which part is the same as those shown for earlier cars.

Colin W. Hughes & Rolls-Royce Enthusiasts' Club 2006