Skip to main content

Full text of "campbell_c_the_sports_car_its_design_and_performance"

See other formats



Fourth Edition 





Its design and performance 

M.Sc., C.Eng., M.I.Mech.E. 


First published 1954 

Reprinted (three times) 1955, 1956 

Second edition 1959 

Reprinted 1960,1961,1962,1965 

Third edition ( revised) 1969 

Fourth edition 1978 

©Colin Campbell 1959, 1969,1978 

Softcover reprint of the hardcover 4th edition 1978 

Photoset in English Times by 
Red Lion Setters, Holborn, London 

ISBN-13:978-1-4613-3386-9 e-ISBN-13:978-1-4613-3384-5 
DOI: 10.1007/978-1-4613-3384-5 

All rights reserved. No part of this book may be 
reprinted, or reproduced or utilized in any form 
or by any electronic, mechanical or other 
means, now known or hereafter invented, 
including photocopying and recording, or in 
any information storage and retrieval system, 
without permission in writing from the 

To my wife 

for her toleration of much fast motoring 
and more slow writing 

Preface to fourth edition 

Preface to fourth edition 

Some of the sports cars I wrote about a quarter of a century ago are still with us. 
I saw a few of them at the 1977 BARC Easter Monday meeting as they fought it 
out so valiantly over ten laps of the Thruxton circuit in the Classic Sports Car 
Race. It is a sobering thought that many of the readers of this new edition were 
not even born when I first attempted to analyse the technical make-up of such 
wonderful sports cars as the XK120, the Aston Martin DB2 and the Austin 
Healey ‘Hundred*. 

The greatest upheaval on the technical scene in recent years has been the 
challenge thrown out by government legislation around the world to increase 
safety and to reduce pollution. The former has virtually wiped out open-air 
motoring in several countries with warm climates, which is a pity. The latter has 
made it illegal in some countries for the enthusiast to tune his own car. This new 
edition therefore contains no advice on tuning, since this must now be 
considered as a professional prerogative demanding sophisticated and expensive 

Apart from a general up-dating of all the subject matter we have added four 
design studies, partly historical, partly technical, on the Jaguar , the Lotus , the 
Mercedes and the Porsche . The author is particularly indebted to these four 
companies for their assistance in providing so much useful technical informa¬ 
tion on their products. 

Grateful acknowledgement is also given to the many sports-car manufacturers 
who supplied technical data, drawings, photographs and handbooks, and to all 
the other component and specialist equipment manufacturers who gave such 
valuable assistance. 

The question of units for this new edition is a difficult one. Great Britain, in 

viii The Sports Car 

theory at least, is now using metric (SI) units. Power should be expressed in 
kilowatts, torque in newton metres and speed in metres per second. Not only is 
the British motor industry reluctant to make the change, but the majority of our 
English speaking readers are in America, Canada and Australia where metric 
units are only used by scientists. In general in this edition we have given 
quantities in both SI units and in foot-pound-second units. There are excep¬ 
tions. Motorists still prefer to think of speed in terms of miles per hour or 
kilometers per hour and engine speed will be quoted in terms of r.p.m. for many 
years to come. Eventually we shall become accustomed to saying that ‘James 
Hunt wrecked his engine by using too many radians per second.’ I do not think 
that ‘Master James’ will ever express it in such terms. 

Suffolk C.C. 

July, 1977 


Preface to Fourth Edition page vii 

Chapter 1 The Development of the Sports Car 1 

Motor sport . The sports car . The history of the sports car . The 
first sports car . The fabulous years . Historic sports cars . The 
future of the sports car. 

Chapter 2 The Engine: Combustion 18 

Cylinder head history . Combustion chamber research . 

Volumetric efficiency . Knock . Limiting compression ratio . 

Types of combustion chamber. 

Chapter 3 The Engine: Induction and Exhaust 35 

The induction system . The 4-cylinder in-line engine. The 
6-cylinder in-line engine. The V-8 engine . Ramming induction 
pipes . Ramming pipe theory . Forward-ram intakes . Cold-air 
intakes . The exhaust system . The silencer . Ramming exhaust 
pipes . Branched exhaust pipes. 

Chapter 4 The Engine: Valve Gear 55 

Push-rod valve operation . Double overhead valve operation . 

Single overhead camshaft head . The four-valve head . 

Desmodromic operation. 


x The Sports Car 

Chapter 5 The Engine: Fuel Metering 

The carburettor . The S.U. principle . The H.I.F. model. Fuel 
injection . Lucas electronic fuel injection . Bosch K-Jetronic 
mechanical system. 

Chapter 6 The Engine: Miscellaneous Components 82 

The crankcase . The crankshaft . Crankshaft bearings: bearing 
pressures: bearing materials . General lubrication . Pistons . 

Cooling: air cooling: water cooling . The ignition system: sparking 
plugs: the conventional coil ignition: new ignition developments. 

Chapter 7 Road-Holding 112 

Tyres: the grip on the road: the tyre footprint. Cornering power . 

Tyre construction . Aquaplaning . Cornering behaviour: oversteer 
and understeer: rear wheel drive: front wheel drive: braking . The 
mid-engined sports car . Factors leading to understeer. 

Chapter 8 The Suspension 140 

Springs . Pitching . Independent suspension . Shimmy and tramp . 
Representative designs: Jaguar, Porsche, Datsun, Aston Martin . 

The suspension damper: double-tube damper: single tube damper. 

Chapter 9 The Chassis, Frame and Body 169 

Materials . Torsional stiffness . The tubular frame . Unitary body- 
chassis construction . The backbone chassis . The shape of the 
body: drag coefficients, lift at high speed, the air dam, directional 
stability at high speed. 

Chapter 10 The Transmission 186 

Torque multiplication . The gear ratios . The overdrive . Synchro¬ 
mesh . The automatic transmission . The clutch . The final drive . 
Universal joints . The limited slip differential. 

Chapter 11 The Brakes 200 

The grip on the road . Braking forces . Weight transference under 
braking . Brake fade . Disc brakes: disc brakes for the high-speed 
sports car, pad materials. 

Chapter 12 Performance 211 

Standards of performance . The meaning of power . Acceleration: 
acceleration times for 0-60 m.p.h., the concept of effective mass . 
Maximum speed. 

Contents xi 

Chapter 13 The Sports Car in the Future 223 

The Engine: promising alternatives, the gas turbine, rotating 
combustion engines, the Diesel engine, the Stirling engine, the 
steam engine . Petrol engine developments: stratified charge: 
turbocharging: air fuel ratio control . The transmission . Tyres 
and suspension: no-roll suspension . Braking. The long-life car. 

Chapter 14 Design Studies 250 

The Jaguar 
The Lotus 
The Mercedes 
The Porsche 

Index 302 

The development 
of the sports car 

'We are living in a time when 
the plural is killing the singular; 
nothing has real value 
when everything comes in thousands.' 


Motor sport 

Man is an adventurous and fun-loving creature. If we can bring ourselves to 
accept that we are descended from some branch of the monkey family, it is no 
longer difficult to understand that our boldness and our love of fun are 
instinctive. Our instincts are sometimes strong enough to break through that 
polished skin of sophisticated behaviour that we call civilized living, strong 
enough to make us do the most irrational things in the name of sport. 

Nothing could be more serious than transport. Our modern suburban life 
would break down without it, yet we cannot resist any opportunity to create fun 
and sport out of this transport, whatever form it may take. Horse racing was 
inevitable, chariot racing irresistible and to those of us with muscles and wind 
enough the invention of the penny-farthing bicycle became just another thing to 
be raced. Who can tell what the next generation will do with space vehicles! 

Motor sport is our immediate concern and with this particular form of 
transport we have given wide range to our faculty for invention. Not content 
with racing it round and about and up and down, we invent all manner of games 
of skill and chance, trials, rallies, gymkhanas, auto-cross, scavenge hunts and 
treasure hunts. 

Sport and danger are often inseparable. Sometimes we seem to go out of our 
way to ensure this. The writer was a little horrified when he saw that the 
Kentucky Auto Speed Championship was to be fought out with old stock cars 
fitted with souped-up engines around a figure-of-eight course. The centre of the 
circuit was thus an uncontrolled crossroads to be negotiated at speeds approaching 
60 m.p.h. with a field that soon became spread out, not only around the course, 
but on all sides of the cross-roads. If racing improves the breed, the future 
Kentuckians will be happy in the cut and thrust of modern traffic. 

2 The Sports Car 

The sports car 

Since motor sport embraces such a wide variety of games with the automobile, 
from the pure racing form of Formula I Grand Prix racing to the more bizarre 
antics of stock car Demolition Derbies, the reader might well ask, if this is 
motor sport what then is a sports car? And the answer is not going to be easy. 

An examination of the writings of such renowned motoring journalists as 
Joseph Lowrey, Tom McCahill, Cyril Posthumus and Rodney Walkerley 
reveals only one thought in common. They all agree that a definition is difficult 
to find. As Cyril Posthumus dryly states: ‘It is easier by far to decide what does 
not constitute a sports car — a hearse, a hotel brake, a limousine, a small 
economy car, for example; any vehicle, in fact, in which carrying capacity takes 
priority over performance.’ 

Significantly, if we reverse these priorities and add one word, we find that 
Posthumus has given us a very good definition. We can now say ‘A sports car is 
any road vehicle in which performance takes priority over carrying capacity’. 
This simple definition brings us closest to what we call the Production Sports 
Car Class in modern racing. The prototype class of sports car was not seriously 
intended to be used regularly on the public roads. Such vehicles are designed to 
comply with the bare essential requirements of the class as defined by the 
current regulations of the FI A. They carry such lights as are demanded, a mirror 
and a barely audible horn. They measure up to the minimum windscreen height 
and door sizes* etc., but they never yield an extra inch to their competitors. 
Their suitability for normal road usage is always in grave doubt. Such vehicles 
usually arrive at the race meeting inside a van or on a trailer. It was inevitable 
that the relentless drive of all-out competition would lead to such a class of 
sports car. Many of us like to think that the real old-fashioned sportsman — my 
American friends call him ‘the nice guy’ — is so blessed by fortune that he can 
drive his honest roadworthy sports car to the races, put on his helmet and racing 
overalls and proceed to battle with the leaders. Was it not Leo Durocher of the 
Dodgers who first said ‘Nice guys finish last’? 

Despite these cynicisms the idea that a sports car is a dual purpose vehicle dies 
hard. Many still believe that the true sports car must be a road car and a racing 
car all in one package. Markham and Sherwin in The Book of Sports Cars 
suggested that ‘pleasure’ is the key word. ‘A sports car,’ they said, ‘is an 
automobile designed for the enthusiast to whom pleasure is its paramount 
potential; pleasure in its performance and pleasure in its design. The sports car 
is a dual purpose car, it is equally at home in city traffic and in all-out 
competition and requires no essential modification to convert from one use to 
the other. It is, in short, a car that is meant to be driven to a race, in the race, 
and back home from the race.’ This would have been a good definition thirty 
years ago. Today it is difficult to find any cars that are raced, even in the 
Production Sports Car Class, that could meet this specification to the full. 
When a production car has been prepared for racing it has usually become a 
noisy uncomfortable vehicle. The word ‘pleasure’ has been dimmed; only the 
sheer performance remains. 

The Development of the Sports Car 3 

The history of the sports car 

Nobody knows for sure who made the first petrol-driven automobile. The 
commercial success of Gottlieb Daimler and Carl Benz, separately and almost 
simultaneously in 1885 is undisputed, but there were earlier motor vehicles. 
Etienne Lenoir, for example, drove his horseless carriage for a distance of six 
miles in 1862. Unfortunately the records are not clear on one important point, 
the nature of the fuel used to propel this vehicle. There was also an Austrian 
called Siegfried Markus who drove his ‘Strassenwagen’ on the streets of Vienna 
in 1875, but again we know so little of its behaviour or reliability and nothing of 
its subsequent history. 

The early history of the sports car is just as confusing. Probably the most 
confusing element is the lack of an agreed definition of the vehicle itself. For 
want of a sanctioned alternative, the writer proposes to use his own definition — 
any road vehicle in which performance takes priority over carrying capacity — 
as his specification by which to judge the validity of any early vehicle to the 
name ‘sports car’. 

At the turn of the century a man had to be a true sportsman to want to drive 
one of these self-propelled road vehicles that the general public still regarded as 
a great joke. For those with a belief in the future of the vehicle there was 
sufficient adventure, even danger, in a short cross-country journey to satisfy the 
requirements of a sport. There was no need to race, the competition lay in the 
journey itself. Despite all this, the continual breakdowns, the unreliable tyres, 
the dust and the panic-stricken horse-drawn traffic, it was inevitable that 
motorists would begin to race. The first race of any consequence was from Paris 
to Bordeaux and back again to Paris —in 1895. It was won by a VA -litre 
Panhard at an average speed of 15 m.p.h. 

During the first ten years of motor racing great changes were made in the 
design of the cars that were raced. The flexible ash frames that had been 
satisfactory for the leisurely pottering around the town had to be replaced by 
steel frames that would stand up to the battering given by long distance racing 
on atrocious roads. Speeds rose higher and higher, far too high for the 
indifferent tyres of the period. By 1905 engines had grown to gigantic 
proportions. The racing Fiat of that year had a bore of 180 mm (7.08 inches) 
and a stroke of 150 mm (5.90 inches), giving a capacity of 16 litres (973 cu. in.) 
and a horsepower of 110. 

A racing car was becoming a specialized and expensive vehicle, as different 
from the normal town carriage as Zsa Zsa Gabor is from the girl next door. 
Certain influential car manufacturers in Britain began to wonder if a racing 
formula could be devised that would give exciting and purposeful competition 
between ordinary touring cars. It could, of course, have been the poor showing 
of British cars in the fierce competition of the Gordon Bennett races for real 
racing cars that made them think along these new lines, but in all fairness there 
were good reasons for staging public demonstrations of the durability and road¬ 
worthiness of ordinary touring cars. Very few people believed that these noisy 
stinking oddities would really drive the horse-drawn carriages from the highways. 

4 The Sports Car 

It was the Automobile Club of Great Britain and Ireland (later to become the 
Royal Automobile Club) that drew up the regulations for the first Tourist 
Trophy race which was held in the Isle of Man in September 1905. This was the 
first of a series of races that, in the opinion of several motoring historians, gave 
the impetus to European motor manufacturers to produce the sports car — a 
touring vehicle that could be raced. But if the T.T. saw the conception of sports 
car racing it was left to Le Mans to nurture it into blooming health with the 
assistance of that able French wet-nurse l’Automobile Club de l’Ouest, the club 
that organized the Le Mans race for sports cars, le Grand Prix d’Endurance des 
Quatres-Vingt Heures du Mans, run for the first time on May 26th and 27th in 
1923. We owe a great debt to this enterprising French club for the work they 
have done in building up Le Mans into the greatest sports car race of them all. 
No club has fought so hard to prevent it from becoming a race for thinly 
disguised Grand Prix cars with one-and-a-half seats. That they have been forced 
to compromise at all in their concept of a true sports car has almost always been 
at the insistence of the entrants and usually with threats of withdrawal of 
popular contestants. The battle has been fought volubly for many years. The 
organizers on their side try in all good faith to define the cars to be raced and the 
essential equipment to be carried, to regulate minimum sizes for windscreens, 
seats, doors, ground clearance, luggage space, etc. etc. All this then only serves 
as a challenge to the cunning brains of a few of the competitors who search 
diligently for loopholes in the regulations to give them an advantage over all the 
others and then complain loudly the following year when the new regulations 
are several pages longer. Gamesmanship has been defined as ‘winning without 
actually cheating’. I hope that Rodney Walkerley was going a little too far when 
he suggested it is ‘cheating without being found out’. 

The first sports car 

It was Henry Ford who told us ‘History is bunk.’ A less destructive criticism 
might have been ‘History is not yet an exact science and evidence is sometimes 
conflicting.’ History was all so straightforward when we were at school. Our 
country was always in the right; Shakespeare was Shakespeare and Queen 
Elizabeth was a virgin Queen. Now we know our country was sometimes wrong, 
Shakespeare’s plays could have been written by Sir Francis Bacon and the Earl 
of Leicester often stayed far too late after eating supper in the Queen’s 
chambers. We once thought we knew the history of the sports car and its 
origins. Today we are not so sure. 

There were sports cars around before anyone thought of the name. The name 
‘sports car’ only appeared in catalogues after the First World War. The French 
had absorbed the English word ‘sport’ into their language and as early in 1921 
Amilcar were advertising their tiny sporting vehicle under the model name of ‘le 
Petit Sport’. 

Sporting versions of the automobile appeared all over the civilized world in 
the first decade of the century. Since so many of these were made in twos and 
threes in back-street premises with no reliable documented evidence of their 

Fig. 1.2 Ferrari Dino Type 308GTB. Typical modern mid-engined sports car. 

The Development of the Sports Car 1 

dates of manufacture it is not likely that anyone will ever be able to name the 
first ‘sports car’ with any certainty. 

William Boddy, the editor of Motor Sport , refers to the Mercedes 60 as the 
Daimler Company’s first sports car. This car was first produced in 1903 and if 
we accept it as a sports car it must have been the first in the world. For the 1903 
Gordon Bennet race in Ireland, the German Daimler Company, who made the 
Mercedes, had prepared three Mercedes 90 racing cars. These were destroyed in 
a fire at the Canstatt works a few days before the event and private owners 
handed over their 60’s to be hastily prepared for the race. Jenatzy won this race 
with one of these fine cars and in the limited time available for preparation they 
must be regarded as production touring cars (or sports cars) with mudguards 
and lighting equipment removed. The Mercedes 60 had a 9.2-litre four-cylinder 
engine with push-rod-operated inlet valves. The chassis was of pressed steel with 
four semi-elliptic springs. Transmission was through a four-speed gearbox and 
final chain drive. This chain drive and the typical wooden wheels of the period 
were the only striking differences in the specification to divide it from the typical 
sports cars that followed twenty years later. 

Cyril Posthumus traces the growth of the sports car in Europe to the impetus 
given by the early long-distance trials, the forerunners of the modern rally. The 
most famous of these were the Prince Henry Trials beginning in 1908 when 
Prince Henry of Prussia offered a prize for a long-distance rally round Germany 
in which bonus marks were awarded on the performance achieved over certain 
timed sections. The 104 horsepower Benz which won the first Prince Henry Trial 
had all the outward appearance of a sports car. It was reported to have good 
handling, would accelerate from 30 to 50 m.p.h. in 10 seconds and had a top 
speed of 90 m.p.h. The third and last of the Prince Henry Trials was won by an 
Austro Daimler, driven by 34 years old Ferdinand Porsche. This car, with a 
high-sided touring body and the performance of a sports car, had been designed 
by Ferdinand Porsche whose influence on sports car design was to grow and 
endure over half a century. 

The ‘Prince Henry’ Vauxhall was one of the earliest British sports cars. Three 
of them competed in the 1910 Trial but were no match for the 5.5-litre 
overhead-valve-engined Austro Daimlers, having only 3-litre side-valve engines. 
Nevertheless the interest in sporting cars of this type had grown so much that 
both Austro Daimler and Vauxhall were offering for sale replicas of their Prince 
Henry models at the 1911 London Motor Show at Olympia. 

It was in 1910 too that Ettore Bugatti introduced his 1.3-litre overhead 
camshaft Type 13, the forerunner of many much more elegant sports cars made 
by le Patron between the two World Wars. A modified 1.45-litre version of the 
Type 13 ran second in the 1911 Grand Prix de France at Le Mans. Its 
competitors were the 7- to 10-litre monsters of the period. Very few of the 
diminutive Type 13 were made and the first Bugatti sports cars to be made in 
what can be called production quantities were the ‘Brescia’ and ‘Brescia 
modifie’, types 22 and 23 made after the First World War. 

A little study of the early American road racers shows that many of these 

8 The Sports Car 

roadsters, such as the Apperson, the Chadwick, the Colburn, the National, the 
Stutz and the Thomas measure up to our definition of a sports car. As the first 
American sports car the writer unhesitatingly awards the palm to the Apperson 
‘Jack Rabbit’. In the June 1964 issue of Motor Trend J.L. Beardsley is very 
persuasive in claiming the ‘Jack Rabbit’ as the world’s first sports car, but the 
Mercedes 60 was one year, possibly two years, before the first Apperson sports 
roadster. This appeared in 1904, the name ‘Jack Rabbit’ being adopted a year 
later. A 1907 advertisement for the ‘Jack Rabbit’ gives us a very early definition 
of the sports car when the Apperson Brothers claim to ‘cater to that limited class 
of owners who want a car that can be put to any service — racing or touring’. 
There is also a touch of modern snob appeal in their promise that ‘only 15 cars 
of the type will be built for 1907’. 

A close contender for the American title would be Stutz, since the ‘American 
Underslung’ was built in 1905. This was a low-built racy two-seater with a 
chassis that was underslung from the springs instead of being mounted high 
above them as in the prevailing fashion. It was not until 1914, that the famous 
Stutz Bearcat appeared, the most glamorous of all American roadsters — or 
should we say — sports cars? 

Ken Purdy does not hesitate to use the name ‘sports car’ when writing of this 
great era in American automobile history. On the Mercer ‘Raceabout’ he says: 

‘There are American cars that rank higher in rarity than the Mercer 
Raceabout, but no car ever built in America is more sought after or more prized. 
There are two reasons: the sports cars of the years between 1900 and World War 
I were starkly functional, unburdened by frills and the weight of useless metal, a 
characteristic much to be desired; and of all the many contemporary two- 
seaters, the lean, high-striding Mercer is indisputedly the best looking. Second, 
most antique automobiles are not at all fast and this one is. A good Mercer 
Raceabout will cruise all day at 60, show 70 or more on demand, and it has the 
steering and road-holding to go with its speed.’ 

The fabulous years 

No apology need be made for using such a threadbare adjective, since in its true 
sense we find that the passage of time is now weaving legends around the names 
Bugatti, Bentley, Mercedes-Benz, Alfa Romeo, Aston Martin; even the plebeian 
M.G. Fables are sometimes more popular than hard facts. Every year we see 
new books appearing to give us eye-witness, blower-by-blower accounts so to 
speak, of the mighty battles fought between Bentley and Mercedes-Benz, 
between Bugatti and Alfa Romeo; but we cannot let ourselves be tempted into 
such reminiscences, for our interest in this book is design and performance of 
sports cars and our interest in this chapter is the history of the forces and human 
foibles that have moulded the shape of the vehicle during the sixty years of its 

The greatest influence of all has been Le Mans. A study of the rules for the 
first race for sports cars at Le Mans shows how the pattern was set for the design 
of a sports car that was to last for the first ten years. All cars were to conform 

Fig. 1.3 Jaguar XJ-S. The latest Jaguar sports car has grown in size, has lost a little in sporting appeal and gained 
in effortless performance. 

Fig. 1.4 MGB GT. Ageing in design, but still a popular formula. (Even the use of ‘cheesecake’ by Publicity 

The Development of the Sports Car 11 

exactly to catalogue specification of the current year, with full touring coach- 
work including mudguards, running boards, headlights, side lights, tail light, 
a folding hood, a horn and a rear view mirror. All cars of greater capacity 
than 1100 c.c. had to carry a full four-seater body, those of lower capacity a 
two-seater body. Cars fitted with electric starters had to carry a starting handle 
in the tool-box. The only work done on the cars during the race had to be carried 
out at the pits and this only by the driver in charge of the car at the time. Only 
one of the two drivers travelled in the car at any one time, but ballast of lead or 
sand had to be carried in the car to represent the weight of one passenger. In 
later years limits were set on the internal dimensions of the body to ensure that 
the competing cars were sensibly-sized sports cars and not freaks. Mudguards 
had to be no less than a certain width and had to give a minimum angular wrap 
around the wheels. Even the type of headlamp bulb had to be the cadmium 
yellow specified for use on the public roads of France. To encourage reliability 
seals were placed on petrol and oil fillers, batteries and generators. The car that 
burned too much oil was disqualified for taking on oil too soon and the car that 
burned out its generator would have to retire during the night. Hard, you may 
say, but the organizers were not planning a Sunday School picnic, but a Grand 
Prix of Endurance. Their aim was to improve the new breed of sports cars, to 
force the automobile makers to develop engines and transmissions, lights and 
dynamos, wheels and tyres that would give trouble-free high-speed motoring to 
the new motor-minded generation that had grown up since the end of the war. 

With no other thought in mind than the development of sturdy reliable 
weather protection, a new rule was introduced in 1924. All open cars had to 
come into the pits at the end of the fifth lap; there they had to erect the hood, 
then cover a further 20 laps of the old 10-mile circuit. At the end of this time 
they were allowed to come in again and stow away the hoods, but not until the 
scrutineers had examined the condition of the hood fabric and the state of the 
supporting frame. During the actual race several over-excited drivers, who no 
doubt had felt it beneath their dignity to practise anything so mundane as 
putting a hood up and down, made quite a comedy turn of the business. Others 
had difficulty in completing the requisite number of laps as they struggled to 
retain tattered fabric with one hand, while steering with the other. 

Illogically the drivers were very angry with the organizers when they refused 
to drop this rule. The drivers and entrants agitated year after year until the rule 
was dropped. By this time, however, racing had improved the breed and by the 
end of the ’twenties the hoods on standard cars had become quite durable. 
Many years had to pass before really draught-proof weather protection was 
devised for the open sports car. 

Over the years the organizers of the Le Mans race have been fighting a 
rearguard action, all the time resisting the changes that have gradually been 
made to the regulations, until today we see that the Le Mans sports car is a 
specialized sports/racing car designed to meet the Le Mans regulations, but 
no longer a car for use on the public roads. So, eventually they have failed, but 
they have left behind a long history of fine sensible sports cars that won 

12 The Sports Car 

races, yet were always sports cars. This in itself has been a fine achievement. 

Le Mans has now lost its appeal. It is no coincidence that the ‘sports cars’ that 
race there now have lost all sense of identity with the sports cars we use on the 
public roads. 

Historic sports cars 

It is no mere cynicism that ‘nobody remembers who came in second’ and the 
cars we remember best are the ones that won races. Bentleys made a habit of 
winning races, even if they finally lost financially, and the writer would be no 
Englishman if he did not put the Bentley first. The Bentley is a cult in England 
as diehard and illogical as the Eton wall game or foxhunting. W.O. Bentley, 
who started it all, served his apprenticeship to locomotive engine design and 
there is little doubt that his cars bear the stamp of this well-engineered 
conservative branch of the mechanical arts. Here we find sound solid basic 
engineering with high factors of safety —no nonsense such as making one 
component serve two functions as we might see on a Lotus — every component 
designed specifically for its job, regardless of expense or weight, and presum¬ 
ably made to last for ever, or at least as long as the railway company stays in 

The modern sports/racing car such as the Lotus is designed to be as light as 
possible, short of falling to pieces under more than average stresses, and to be as 
small as possible within the minimum dimensions of the current FIA regula¬ 
tions. In this way less power is needed to achieve a set rate of acceleration. With 
the minimum possible frontal area less power is required to reach the designed 
maximum speed. With less power, the engine is lighter, the transmission is 
lighter and the frame that has to support these components can be made lighter. 
Lightness can be said to breed lightness. As opposed to the usual vicious circle 
we could call this ‘the benevolent circle’ and no designer in modern times has 
used this technique with more success than Colin Chapman. 

With Colin Chapman it was a matter of applying a very superior intelligence 
and a sound training in engineering to the problem of designing a car that would 
win races. With Ettore Bugatti it was more the artist’s flair for doing the right 
thing and with Bugatti this intuition seldom led him to commit a gross error. 
Born in Milan, the son of an engraver and architect, he studied at first to be a 
sculptor. When his brother Rembrandt showed superior talent, Ettore’s pride 
compelled him to abandon sculpture. He turned hopefully to the automobile 
and made his first car at the age of 18. At last he believed he had found an art 
form at which he could succeed. He left Milan for Alsace and in 1910 he rented a 
small property in Molsheim where he began to build what his advertisements 
called ‘Le Pur Sang des Automobiles’. As a lover of horses he appreciated the 
full meaning of the words ‘pur sang’ or ‘thoroughbred’. By the time he died at 
the age of 66 he had produced nearly 10,000 automobiles, less than one average 
day’s output by General Motors Corporation. Names like Sheraton, Stradivari, 
Wedgwood and Bugatti will survive. The names of the members of Detroit 
design committees are as ephemeral as slogans chalked on wet pavements. 

Fig. 1.5 Panther Lima. Sports car nostalgia is on the increase. When a Thirties replica is as beautifully finished as the 
Lima, a handbuilt copy of nothing in particular created by Panther Westwinds Ltd of Byfleet, Surrey, the 
appeal is instantaneous. The performance is electrifying. 

14 The Sports Car 

Only once did Bugatti design big, when he made the Type 41 or Bugatti 
Royale. This car was so big and expensive that it excited the public imagination, 
but the car was so untypical that, for once, his intuition let him down. The 
brakes were woefully inadequate. 

Many of Bugatti’s sports cars started life as Grand Prix cars. These he would 
later detune and sell in modified form for use on the public roads. The Type 23, 
Tull Brescia*, as used so successfully by Ranmond Mays in his early days, was a 
racing car. The road version, sold in relatively large numbers for several years, 
was the ‘Brescia Modifie*. The Type 23 was eventually replaced by the Type 37, 
a Bugatti that could be used by private entrants for road racing or could be 
fitted with road equipment and used as an everyday sports car. The beautiful 
chassis carried a simple slim body with a finely tapered bonnet behind a tiny 
horseshoe radiator. The lVi-litre engine had plain bearings (very unusual for 
Bugatti) and a relatively simple four-cylinder layout. Top speed was about 95 
m.p.h. in road trim, with an acceleration time from zero to 60 m.p.h. of about 

15 seconds. 

No marque is more difficult to classify than the Bugatti, since so many of his 
pure racing cars were later converted to road use by their owners. A few of these 
might be regarded as sports cars, but in general they retained the harsh inflexible 
character of the Grand Prix car and were only regarded as sports cars by their 
loving owners. It could be a different story when the metamorphosis was carried 
out by Bugatti himself, or by his son Jean. The most desirable Type 55 is a case 
in point, being a Type 54 Grand Prix chassis fitted with a slightly de-tuned Type 
51 Grand Prix engine —a double overhead camshaft, Roots supercharged 
engine, with ball- and roller-bearing crankshaft. With full road equipment top 
speed was about 110 m.p.h. 

No matter what excuse we may give to our wives the sports car must always be 
classified as ‘luxury consumer goods*. It was not surprising therefore that the 
economic blizzard that swept across the world after the Wall Street crash swept 
so many sports car manufacturers into oblivion. Oddly enough, Bugatti, whose 
cars were amongst the most expensive ever made, in terms of cost per pound, 
was not a casualty. Bentley, after the withdrawal of the financial backing of 
Woolf Barnato, the South African millionaire who had driven Bentleys to 
glorious victory at Le Mans, was forced into dismal bankruptcy. 

Many manufacturers now completely abandoned active participation in 
motor racing — confining their activities to a little financial encouragement to 
the more promising private entrants. Such privately owned cars were usually 
tuned free of charge by the works before a race and, if necessary, straightened 
out before the next. No ‘Clash of the Giants* now — with Blower Bentley 
battling 7-litre Mercedes-Benz. It was now the ‘Battle of the Midgets* between 
such surprisingly fast small-engined cars as the J4 and R Type M.G. Midgets, 
the NA and K3 Magnettes, the Riley T.T. Sprite and the Singer 9 and 1 Vi-litre. 
Handicap races were in vogue and to the imaginative spectator there was a 
Jack-the-giant-killer quality in the sight of a 2.3-litre Alfa Romeo, a magnificent 
car that could beat all-comers in such classic races as the Mille Miglia, the Targa 

The Development of the Sports Car 15 

Florio and Le Mans, failing to catch a tiny M.G. before the checkered flag. 
Some day this spectator hoped to own an M.G.; they only cost about £200. The 
chassis alone of the Alfa Romeo cost £1,700 in Great Britain. 

There was a growing demand for races for small sports cars and some of the 
international races were divided into classes. Others gave ‘credit laps’ to the 
smaller cars as a simple method of handicapping, while the organizers of Le 
Mans devised a complicated ‘index of performance’ to equalize, in theory at 
least, the chances of all sizes of car. This growing interest in small-car racing was 
a sign of the times. Millionaire sportsmen and wealthy noblemen had supported 
the game of motor racing from its earliest days, but there were not enough of 
these left to go around. It was a healthy sign for the future of the sport when the 
very cars that young men of modest income used for everyday transport could 
be modified at no great expense to be capable of winning races. Teams of such 
cars could be maintained at a cost of no more than £5,000 per annum, as against 
£50,000 per annum or more needed to maintain a team of Grand Prix cars. 

After the reconstruction period following the Second World War we entered 
an era of economic prosperity in what was now known as the Western World. 
Never before had we had such a wide choice of exciting sports cars. Notable 
classics that appeared in this post-war period were the A.C. Ace, the Alfa 
Romeo Giulietta, in all its stages of tune, the Aston Martin 2/4 and DB3S, the 
Bristol 403 and 404 and Ferraris of so many types — we dare not suggest that 
any were not classics. We remember the 212 Inter that was invincible in its class, 
the 340 America, the 375 Mille Miglia, the 500 Mondial, the 500 Testa Rosa, the 
750 Monza, all wonderful sports cars. Donald Healey made several models after 
the war, but the ‘Silverstone’ open two-seater, with its retractable windscreen 
and the spare wheel so neatly blended into the bodywork as to act as a rear 
bumper, is the most memorable. The Frazer-Nash ‘Le Mans’ was a superb 
achievement from such a small manufacturer and it is sad the marque has now 
passed into history. 

The XK120 Jaguar was a post-war sensation and deservedly so. The success¬ 
ful C Type and D Type were developed from this basic design, but the emphasis 
had by this time shifted towards the sports/racing category. The XK120, 
however, was a very comfortable road car, luxurious in fact compared with the 
sports car we had known before. Despite this it could win races with little more 
than normal routine preparation. Contemporary styling connoisseurs judged it 
to have a beauty of line that has seldom been surpassed. 

The Lancia Aurelia started life as an unpretentious Gran Turismo car, more 
remarkable for its advanced suspension system than for its performance. 
Eventually it grew into a sports/racing tiger that won the Mille Miglia and the 
Targa Florio and, in slightly tamer form, the Monte Carlo Rally. 

The true genre of the Maserati Brothers was the Grand Prix car, but in the 
course of a long career they have made many fine sports cars. The Type A6GCS, 
with a 2-litre 6-cylinder engine designed by Colombo, was one of their more 
successful post-war designs. 

The German contribution to the sports car revival was twofold: from 

16 The Sports Car 

Porsche —as old as the sports car and from Mercedes-Benz, as old as the 
industry itself. With only Volkswagen parts available in a country smashed by 
war, Ferdinand Porsche and his son Ferry gave us a fine little sports car that was 
different, even though it was underpowered. The Porsche today is still different 
and the Porsche enthusiast would say ‘Vive la difference*, for the wide range of 
Porsche types today can offer everything from exhilarating sporty transport to 
the most formidable of turbocharged racing sports cars. Today they are all as 
far removed from the production VW as the Lincoln Continental is from the 
Model T Ford. 

When Mercedes-Benz decided to make a new sports car after the war they 
waited until their manufacturing facilities were sufficiently re-established for 
them to design and manufacture as fine a vehicle as the pre-war SSK, of which 
they were justly proud. Nothing was spared in the choice of materials and in the 
manufacturing techniques used on this new 300SL. The engine was canted at 50 
degrees to the vertical to achieve a low silhouette. Three Solex carburettors were 
used at first, later to be replaced by an expensive, but remarkably efficient direct 
fuel-injection system. Suspension was by double wish-bones at the front and 
swing-axle at the rear. It is doubtful if any model of sports car has ever had such 
an overwhelming history of victories. The 300SLR sports/racing car that 
followed the original design never lost a race. 

For those of slender purse there has always been a wide selection of small 
sports cars, especially from Great Britain. There have been the TC, TD and TF 
Midgets, the MGA and MGB and, based on the Austin Healy Sprite, the new 
series of M.G. Midgets. The Triumph Company, who had made small sports 
cars before the war, entered the market again with the rugged conventional TR 
model which has been improved and developed up to the level of the current 
TR7. The latest Triumph is still reasonably priced for a 2-litre sports car and is 
still fun to drive. 

The Lotus Company grew out of Colin Chapman’s post-war ‘specials’ that 
were so successful that he had to make more to please the public. Tax could be 
saved in Great Britain by the use of ‘build-it-yourself kits’ and the early Lotuses 
were largely sold in this form. The modern Lotus is a luxurious creation, with 
electric windows, stereo radio and air conditioning; yet it remains a sports car. 
With an engine of only 2-litres it represents Chairman Chapman’s thoughts on 
the car for the Eighties, when low fuel consumption and the long-life car will 
become an ecological necessity. 

Detroit has been fascinated by the sports car for many years, but the market is 
relatively small by Detroit standards. The Corvette, for so long the only 
series-production sports car made in America, has had a strong following since 
it first appeared in 1953. In its latest Stingray form it is still a heavy car (over 
3,000 lb) but the handling is now well above average for such a large car. The 
Ford Motor Company once made a sports car, the GT40, for the express 
purpose of winning at Le Mans. It was an expensive experiment and proved they 
had the expertise to win. The publicity was probably worth the effort. The 
Mustang is their most sporty product on sale today. It is popular and well 

The Development of the Sports Car 17 

suited to its environment. If only they would reduce the size and weight and 
improve the handling it could eventually make a very good sports car. 

There have been a few good post-war American sports cars. Notable was 
Carroll Shelby’s Cobra that was competitive enough to win the Manufacturers’ 
Championship in 1965. 

Pressures to meet future safety, emission and fuel economy regulations on 
their current models seem to be giving the Detroit top management all the signs 
of anxiety neurosis. In such a climate the word sports car or ‘fun car’ is never 
mentioned in the Board Room. Some of their younger executives still have the 
audacity to come to work in foreign sports cars. Perhaps there is some hope for 
the future, even in Automobile-City. 

The future of the sports car 

We face a future when ‘conspicuous consumption’, whether it be of expendable 
chrome-plated hardware or of irreplaceable fuel, will invoke public anger. There 
will be no need for laws to ban twenty-foot long gas-guzzlers. For those with no 
sense of public responsibility the very cost of burning vast quantities of fuel will 
make them ‘think small.’ 

In such a climate thinking small could mean a 2-litre sports car, with good 
acceleration and sure-footed handling and a chassis and body so light and 
aerodynamically effective as to use no more fuel than the contemporary 500 cc 

The sports car of the future will be made to last twenty years or more, with all 
surfaces protected against corrosion and with mechanical components designed 
for easy maintenance. 

We sometimes hear that this age of the throw-away component keeps us all in 
employment. We are forced to ask: for how long? Some day the materials to 
replace those throw-away components will no longer be found, even deep down 
in the ground. 

Porsche are already at work on their Long-Life Car. Others will soon follow. 
If members of the Vintage Car Club can do it, many of them maintaining cars in 
excellent roadworthy condition that first left the factory fifty years ago, perhaps 
we could do the same. 

The engine: combustion 

'And now I see with eye serene 
The very pulse of the machine.' 

Cylinder head history 

Where we burn the fuel and how we control its burning is so important, so 
crucial, to the whole science of the internal combustion engine that it is entirely 
fitting that we devote a whole chapter to the component where this burning 
occurs. Before the First World War the cylinder head was largely regarded as a 
convenient cover for the cylinder block and the actual shape of the combustion 
space was not considered to be of any importance. The superiority of the 
overhead valve cylinder head, especially the pent-roof twin overhead camshaft 
design, was appreciated by automobile designers, but only for its superior 
breathing and the low inertia of the valve gear. The excellent anti-knock 
characteristics were yet to be discovered. As early as 1908 Clement-Bayard won 
a Grand Prix using a 13.9-litre overhead camshaft engine fitted with inclined 
valves in a pent-roof head. For production touring cars, however, the T-head 
and L-head engines were almost universally used. Very low compression ratios 
were used, but despite this, detonation often occurred. Nobody knew what 
caused this knocking sound that came from the combustion chambers and in 
general it was ignored. Accessibility was everything to the early motorist and the 
T-head engine with its uncooled valve-caps placed over each valve head, seemed 
at the time the essence of good design. 

After the war, a young experimental engineer, Harry Ricardo, was called 
upon to design an engine for the T.T. Vauxhall. This twin overhead camshaft 
3-litre engine is still regarded as a classic example of the type. The superiority of 
this valve layout was now well established and several examples appeared during 
the ’twenties in Grand Prix cars. In 1926 the Riley Company introduced a 
cheaper form of valvegear, using twin camshafts, placed high on the sides of the 
cylinder block, and short push-rods to operate the inclined overhead valves. The 

The Engine: Combustion 19 

classic hemispherical combustion chamber was used and the inertia of the 
valvegear was very little greater than with the Ricardo design. In the hands of 
enthusiasts such as the late Freddy Dixon, this design of engine was made to give 
remarkable power outputs all from an engine that in standard form was fitted to 
a car that sold for the modest price of £298. 

Despite the work of such men as Ricardo and Whatmough and Janeway to 
refine the behaviour and efficiency of the side-valve combustion chamber, 
despite the attractions of its cheap and simple valve operation, the side-valve 
engine eventually disappeared from the scene. 

The overhead valve engine 

There are three major design decisions that confront the engine designer when 
he begins his preliminary layout of a new cylinder head: 

(a) combustion chamber shape to promote efficient combustion, 

(b) valve layout to give efficient induction and exhaust, 

(c) valve operating mechanism to give reliable operation at the designed 
operating speed. 

Experience has shown that aspects (a) and (b) call for great ingenuity if one is 
to achieve efficient combustion without making some sacrifice in breathing 
efficiency. Experience has also shown that the most effective method of valve 
operation, the provision of two overhead camshafts to each bank of cylinders, is 
also the most expensive. 

In Chapter Three we discuss the all-important subjects of induction and 
exhaust. In Chapter Four valve operating mechanisms will be examined in 

Combustion chamber research 

In the combustion chamber of the modern sports-car engine one sees the 
embodiment of thousands of hours of research work, hundreds of patent 
specifications and countless academic theses and technical papers. The work on 

20 The Sports Car 

knock continues, but the excitement must surely be over. One could hardly hope 
for new discoveries in this field. If we hope to increase the output of the 
unsupercharged petrol engine we must strive for still higher volumetric efficien¬ 
cies, leading in turn to higher engine speeds. 

The work of Mercedes-Benz in Germany and the Texas Oil Company in 
America on the prevention of knock by means of directional turbulence and 
rotational swirl in engines using direct injection of petrol is not really a new 
discovery, since it is an extension of existing knowledge gained on compression 
ignition (Diesel) engines. 

Volumetric efficiency 

The volumetric efficiency of an engine is the ratio between the volume of air at 
atmospheric pressure and temperature drawn into the cylinder during the 
induction stroke to the volume swept by the piston, i.e. the piston area times the 
stroke. It is usually expressed as a percentage. The higher the actual pressure 
and the lower the temperature of the air at the point of inlet valve closing, the 
higher will be the volumetric efficiency. The power to be obtained from a given 
engine is largely a function of the mass of air it can breathe, since this within 
limits, controls the amount of fuel that can be burned. 

Surprise is often expressed at the high compression ratios that were used in 
some side-valve engines. What is so often overlooked is that the compression 
ratio is not a ratio of pressures, but is a free volume ratio — a ratio between the 
cylinder volume at bottom dead centre to that at top dead centre. The side-valve 
engine usually gave a poor volumetric efficiency, seldom better than 70 per cent. 
A good overhead valve engine will give 85 to 90 per cent, with the hemispherical 
head engine approaching 95 per cent at maximum torque r.p.m. From this it is 
apparent that at the same compression ratio, the compression pressure will be 
much lower in the side-valve engine and, as a consequence, the brake mean 
effective pressure (b.m.e.p.) will also be lower. 

The aim of the designer who wishes to achieve the highest possible volumetric 
efficiency in his engine is to get the air into the cylinder with as little drop in 
pressure as possible and in so doing to pick up as little heat as possible in the 
passage of the air from the atmosphere to the inside of the cylinder. The factors 
that influence the pressure drop and the charge temperature are discussed 

(a) Admixture of new charge with exhaust residuals . At a compression ratio 
of 6 to 1, the admixture of the new charge of petrol/air mixture with the hot 
residual gases in the combustion chamber will raise the charge temperature by 
about 50°C. At 10 to 1 compression ratio the increase will only be about 25°C. 
In both cases the cooling effect of fuel vaporization is not taken into account, 
since this is treated separately under ( d ). Other things being equal, we can 
therefore expect about 7 per cent higher volumetric efficiency from a 10 to 1 
compression ratio engine than a 6 to 1 engine. 

(b) Heat picked up from the induction manifold and hot-spot. The hot-spot 
was invented by American automobile engineers to ensure a quick warm-up of 

The Engine: Combustion 21 

the engine for the impatient American driver. It is a wasteful expedient and is to 
be avoided except on large engines of low specific power output. It was once 
thought that by collecting the fuel that could not be carried in suspension as fine 
droplets by the air stream, and by collecting these heavier drops in an exhaust 
gas-heated well in the floor of the induction manifold, the temperature of the 
main air stream would not be raised and the volumetric efficiency would not 
suffer in consequence. This is a fallacy. All the fuel in liquid form carried in the 
air stream must be regarded as a potential source of charge cooling, since the 
latent heat to vaporize these droplets comes largely from the air stream. To the 
designer of a high-output engine, any additional heat, whether applied at a 
hot-spot or in any other way, is anathema, since it represents a certain loss of 
power. In modern designs, where a heated induction tract is considered 
necessary to ensure a reasonably good distribution between cylinders before the 
engine has reached working temperature, there has been a return to the old 
water-heated manifold. Although not giving the rapid warm-up of the hot-spot 
it is effective in evaporating the larger drops that fall out of the air stream and 
those that fail to turn the bend where the carburettor branch enters the main tract 
On some modern engines a thermostatic control is used on the hot-spot. This 
arrangement, with careful design, is not as power-wasting as some of the older 
designs of simple hot-spot. In general we can state that the fewer cylinders 
sharing a carburettor the less will be the degree of induction heating required to 
give good distribution when warming up. With one choke per cylinder addi¬ 
tional heat is not necessary. 

(c) Heat picked up from the cylinder head and walls , piston and exhaust valve 
head . When the inlet valve opens directly into the cylinder, as in the overhead 
valve engine, this effect is reduced to a minimum. In some designs the inlet gases 
tend to impinge on the hot exhaust valve head more than in others. This is a 
mixed blessing, since the anti-knock properties of a cooler exhaust valve are 
obtained at the expense of a lower volumetric efficiency. In the average 
overhead valve engine heat picked up in this way will raise the charge 
temperature by about 30°C. 

( d) Latent heat of the fuel. The temperature in the induction tract and in the 
cylinder is influenced to some extent by the latent heat of the fuel. With petrol 
and a maximum power mixture strength of about 12-13 to 1 a drop in charge 
temperature of about 25 °C can be expected from this source. Methanol (methyl 
alcohol) has about four times the latent heat of vaporization of a normal petrol 
and, with a rich mixture strength of about 6 to 1, the potential lowering of the 
charge temperature is about 180°C. In this case only a fraction of the fuel 
droplets will be evaporated before the inlet valve closes and the actual drop in 
charge temperature resulting from this evaporative cooling will be about 50°C, 
giving an increase in volumetric efficiency of about 10 per cent when compared 
with a petrol/air mixture. 

( e ) Port design. One does not need a knowledge of fluid dynamics to see that 
a tortuous or restricted passage will reduce the mass of air drawn into the 
cylinder at high engine speeds. The series of right-angle bends we used to see 

22 The Sports Car 

on older designs of induction manifold were largely responsible for their poor 
volumetric efficiency. 

Modern thoughts on port design, as exemplified by the work of Harry 
Weslake at Rye, suggests that efficient breathing comes from an experimental 
approach to the detail design of every obstacle that lies in the path of the gas, all 
the way from the inlet to the air cleaner to the cylinder itself. The use of a 
wooden model of the entire system makes it easy to modify the shape of any part 
in the search for the lowest pressure loss. This is a complete change from the old 
school of thought that insisted in abrupt changes of section in the main tract 
with buffer ends to promote turbulence. This certainly helped to break up fuel 
droplets and to equalise the distribution of mixture between cylinders, but the 
penalty was a greater pressure loss and a reduction in the maximum power 

The inlet port is often given a tangential bias as shown in Figure 2.1. This 
creates a vortex flow which persists through the induction stroke and the 
compression stroke. The degree of spin influences the early stages of combus¬ 
tion. The importance of this induced swirl was realised many years ago by Harry 
Weslake and a swirl meter incorporating a tiny vaned spinner is inserted in the 
wooden cylinder head model used in the air-flow tests at the Weslake Laboratories. 

(/) Stroke/bore ratio. Even without recourse to mathematics, one can see 
that a small stroke/bore ratio is likely to lead to a higher volumetric efficiency 
than a large one. For any basic head design, whether it be side-valve, in-line 
o.h.v. or hemispherical, the size of the inlet valve that can be accommodated is 
limited by the size of the cylinder bore. For the same bore dimensions, the 
volume of gas to be aspirated with a stroke/bore ratio of 2 will be twice that to 
be aspirated when the stroke/bore ratio is only unity. In the first case the mean 
gas velocity through the inlet valve would be twice that in the second case (if it 
could achieve the same volumetric efficiency). With twice the gas velocity, 
however, since the pressure drop across the valve orifice varies as the square of 
the velocity, the pressure drop will be four times as great. The volumetric 
efficiency of the long-stroke engine will always tend to be lower than that of the 
short-stroke engine. This factor, as much as the limitations imposed by the high 
stresses in the long connecting-rods, was responsible for the low practical engine 
speeds of the long-stroke vintage sports car. 


To understand the phenomenon known as knock, or detonation, it is necessary to 
consider the manner in which the flame spreads from the plug points to the 
furthermost point of the combustion chamber. At first, after the ignition of the 
small amount of mixture close to the plug points, burning is relatively slow and 
depends in the main upon the speed at which the unburned charge is brought 
into contact with the expanding ball of flame. Without violent agitation, or 
turbulence, in the gas, combustion would be far too slow and if it were possible 
to make the gas perfectly quiescent only a small fraction of the total charge 
would be burned before the opening of the exhaust valve. Turbulence is present 

The Engine: Combustion 23 

in varying degrees in all piston engines, being promoted in the first place by the 
passage of the charge through the relatively small restriction made by the inlet 
valve opening and in the second place by the ‘squish’ effect as the piston reaches 
top dead-centre and traps gas between the piston crown and portions of the 
cylinder head. In certain designs of side-valve and in-line overhead valve engines 
the clearance between the piston crown and a carefully chosen portion of the 
cylinder head is made small to increase the degree of ‘squish’ turbulence. 

As the rapidly expanding ball of flame spreads outwards from the sparking 
plug, heat is radiated to the unburnt charge ahead of it. This accelerates the rate 
of burning since less time is wasted in raising the temperature of the unburnt 
charge to the temperature at which combustion takes place. Without this 
radiation, direct contact of the flame, i.e. mixing of the burning and the 
unburnt charge, would be the only way in which heat would be transferred. All 
the time that combustion is proceeding there is an exchange of heat energy 
throughout the whole combustion chamber. As fast as heat energy is liberated 
by the chemical reactions between the carbon and hydrogen of the fuel on the 
one hand and the oxygen of the air on the other — that complex reaction which 
we call burning — so this heat is spread to the unburnt charge and to the walls of 
the combustion chamber and the piston crown. This heat is transferred in two 
ways, by direct mixing of the flame front with the unburnt charge — this we 
called ‘convection’ in our physics class at school — and by radiation — the 
manner in which the earth picks up heat from the sun. All the time this 
extremely rapid interchange of heat is taking place the general rise in tempera¬ 
ture of the gases while the piston is in the neighbourhood of t.d.c. causes a rapid 
rise in pressure. The pressure rise is greatest where the temperature is highest, 
i.e. immediately behind the flame front, where combustion is nearing comple¬ 
tion. The temperature behind the flame front may be as high as 2000°C, while in 
the unburnt gas it will probably be about 500°C. Since, by Charles’ law, a rise in 
temperature at a constant volume (all this happening within a few degrees of 
crank-angle movement) produces a rise in pressure, there travels outwards from 
the sparking plug, in company with the flame front, a wave of high pressure. 

Towards the end of the combustion process, when nearly all the charge has 
been burnt, the pressure and temperature of the unburnt charge can sometimes 
reach critical values. This can occur despite the continuous loss of heat to the 
surrounding walls. If these critical values of temperature and pressure are 
reached, this ‘end-gas’, as it is called, explodes, or detonates, the whole volume 
of gas burning simultaneously. The pressure wave resulting from this detonation 
in striking the walls of the combustion chamber produces the characteristic 
metallic noise we call ‘knocking’ or ‘pinking’. A third factor besides tempera¬ 
ture and pressure decides whether or not the end-gas will detonate. This factor is 
time. Certain chemical changes are now known to take place in the end-gas 
before detonation takes place and time is required for these to occur. This is the 
reason why knocking occurs more readily at low engine speeds than at high. 

At the General Motors Research Laboratory at Warren, Michigan, a special 
engine with a quartz window in the piston crown has been used to observe the 

-18 -14 -10 -7 -3 1 

27 30 34 37 40 44 


-16 -12 -8 -4 TDC 4 

8 12 16 20 23 28 


Fig. 2.2 (a) Non-knocking combustion (actual photographs). Note relatively slow 

controlled combustion of last 25 per cent of charge (from 19° to 34° after t.d.c.). 
(b) Knocking combustion. Note rapid combustion of last 25 per cent of charge 
(from 20° to 23° after t.d.c.). (c) Uncontrolled secondary ignition from hot spot. 

The Engine: Combustion 25 

-18 -14 -11 -7 -3 1 

5 8 12 16 20 23 


combustion behaviour in knocking and non-knocking engines. Figure 2.2(a) 
shows a sequence of high-speed photographs of the progress of combustion 
across a typical combustion chamber when combustion takes place in a normal 
non-knocking manner. The numbers are the crank-angle degrees relative to 
t.d.c., negative values being before t.d.c. Figure 2.2(b) shows what happens 
inside the same engine when the fuel is changed to one of lower octane value, so 
low that knock occurs. From the point where the flame front has travelled about 
half-way across the chamber (16 degrees after t.d.c. in both cases) the 
non-knocking engine takes about 18 more degrees to pass right across the 
chamber. The knocking engine completes this distance in about 7 degrees. 

The burning of a liquid fuel is a very complex multi-stage process. Prepara¬ 
tory oxidation reactions must occur before what is called the ‘cool flame’ stage 
is reached. The third stage, ‘blue-flame formation’ is the first visible stage and is 
followed by ‘auto ignition’, the final phase. While all these complex, but 
extremely rapid, reactions are taking place organic peroxides are forming and 
are decomposing into formaldehydes. Knock is now believed to occur only when 
the concentrations of these compounds reach certain critical values in the 
end-gas. These undesirable reactions occur more readily with a detonation¬ 
conscious ‘straight-chain’ paraffin than with a high octane ‘branch-chain’ 
paraffin. Tetra-ethyl lead considerably reduces these undesirable reactions in 
the cool flame period, the peroxides particularly appearing to be de-activated by 
minute quantities of tetra-ethyl lead. 

Iso-octane is a well-known example of the ‘branched-chain’ paraffins that are 
used in fuels with high anti-knock rating. One particular isomer of octane, 

26 The Sports Car 

known to the chemist as 2,2,4-trimethyl pentane, has been standardized 
throughout the world as the 100-octane reference fuel. Thus a fuel which gives 
the same degree of knock in the C.F.R. (Co-ordinating Fuel Research) engine as 
an 80/20 mixture of the good-reference fuel, iso-octane, and the poor-reference 
fuel, normal heptane is called an 80-octane fuel. The octane rating scale is no 
longer adequate, since many branched-chain hydrocarbons have been developed 
with far higher anti-knock properties than iso-octane. New test methods have 
been developed for rating these fuels, which explains the appearance since the 
end of the Second World War of ‘110 Octane’ and even ‘125 Octane’ fuels. 
Benzene and toluene, and the commercial mixture of benzene and toluene 
known as Benzol, are ring compounds and show much of the stability and 
resistance to knocking of the branched-chain paraffins. Methyl and ethyl 
alcohol are almost knock-free, but in certain engines at compression ratios of 
about 15 to 1 they produce a type of ‘rough-running’ that sounds a little like the 
knock associated with Diesel engines. This is called pre-ignition and, as the 
name implies, it is ignition of the mixture occurring ahead of the spark. This 
phenomenon can also occur with petrol, but in this case the pre-ignition is 
usually produced by a glowing piece of loose carbon, an overheated sparking 
plug, or the glowing edge of a badly-seated exhaust valve. 

Limiting compression ratio 

In the majority of engines the practical limit to the torque, which is proportional 
to the b.m.e.p., is set by the highest compression ratio that the engine can use 
without knock occurring on the particular fuel used. 

Violent knock will quickly wreck an engine. When knock occurs a large part of 
the energy normally released by the burning of the end-gas is wasted and does no 
useful work. This is revealed by the drop in b.m.e.p. indicated on the 
dynamometer when serious knock occurs. The wasted energy results in a rise in 
cylinder head temperatures, piston crown temperatures and exhaust valve 
temperatures and in bad cases engine failure can occur by piston seizure or 
exhaust valve burning. In some cases the rising temperatures lead to pre-ignition 
and the resulting uncontrolled rise in pressure before t.d.c. may overstress a 
component such as a connecting-rod or the already overheated and weakened 
piston. Knock is a serious matter and should be avoided at all costs. 

Let us not, however, develop a neurosis about every little ping or tinkle we 
hear from our engine. There is a world of difference between the barely audible 
knock we hear when an engine is a little over-advanced and the kind of knock 
that punches holes in the crowns of pistons. This latter type of knock can be 
heard quite clearly about a hundred yards away. 

The following are the major factors that have been shown to have an 
influence on the knock resistance of any design of combustion chamber: 

(a) The maximum distance to be travelled by the flame front. The shorter the 
distance between the sparking plug and the end-gas, the less will be the tendency 
to knock. This means that engines with small bores can use higher compression 
ratios than those with large bores, all other things being equal. 

The Engine: Combustion 27 

(b) The relative positions of the plug(s) and exhaust valve(s). The plug (or 
plugs) should be placed in the cylinder head with a decided bias towards the hot 
exhaust valve. This is an obvious requirement, since temperature is one of the 
factors influencing knock. It is obviously desirable for the end-gas to be situated 
near the inlet valve rather than the exhaust valve. 

(c) The end-gas should be situated in the coolest part of the combustion 
chamber . This would appear at first sight to be another way of stating (6), but in 
some designs an attempt is made to cool the end-gas by trapping it between the 
relatively cool large areas of the piston crown and the part of the cylinder head 
immediately above it. In this position the end-gas is not only provided with large 
cool surfaces for the extraction of heat from the small mass of the trapped gas, 
but the position also tends to shield it from direct radiation from the flame 

(d) A certain amount of induced turbulence is required . The importance of 
turbulence and its influence on the combustion process was first stressed by 
Ricardo. Turbulence in a gas may be regarded as the extent to which individual 
eddies are rotating and intermingling. The more violent the turbulence, the more 
rapidly will the eddies rotate, intermingle, break up and reform. Turbulence is 
used to mix the sugar which is at the bottom of our teacup. Left to itself in 
perfectly still tea, the sugar would slowly diffuse through the tea by simple 
‘molecular diffusion’, i.e. the molecules of sugar would slowly diffuse away 
from the strong concentration near the bottom towards the weak concentration 
at the top. When we stir the tea we produce turbulence and the sugar is mixed by 
‘eddy diffusion’. In the Ricardo turbulent head designed for the side-valve 
engine and on almost all side-valve heads made since the early ’thirties a portion 
of the cylinder head is brought low enough to the piston at t.d.c. to form an area 
of ‘squish’. Since the piston approaches this portion of the head at a very high 
speed the effect is that of an air ejector squirting this small volume of trapped 
gas into the man body of gas in the combustion chamber. This squish effect 
results in a great increase in the turbulence in the combustion chamber, this 
occurring when it is most required, speeding up the rate of burning in the middle 
phase. The cooling effect of the adjacent surfaces on the gas remaining in the 
squish gap, after this ejection has occurred, has been mentioned under (c) 
above, but the effect of the ejected gas on the turbulence of the main gas was the 
subject of Ricardo’s original patents. 

The final phase of burning, the critical phase when detonation can occur, is 
slowed down by the lower temperature of the end-gas left in the squish gap. 
Ricardo also showed in his experiments that a practical limit existed to the 
amount of turbulence that should be induced. With excessive turbulence, the 
rate of pressure rise could be so high that rough running was given. Another 
disadvantage of excessive turbulence has since been shown to be an increase in 
the heat losses to the cooling water. The need for induced turbulence becomes 
less at high compression ratios, since the high pressures and temperatures at 
these higher ratios increase the speed of propagation of the flame front. 

28 The Sports Car 

Types of combustion chamber 

In general, combustion chamber designs can be classified in terms of the valve 
layout. For simplification and low cost, all valves in the same bank can be 
placed with the valve stems in a straight line. Valve operation can be by a single 
overhead camshaft or by pushrods that transmit the motion from a single 
camshaft near the base of the cylinder block. 

The Wedge head 

A wedge-shaped chamber as shown in Figure 2.3 has been popular for many 
years. It fulfills the four requirements for good combustion: 

(a) short flame travel to farthest point, 

(b) short distance from plug to exhaust valve, 

(c) well cooled end-gas and 

(d) a degree of squish turbulence. 

Fig. 2.3 A good example of a wedge head: the FWA 
Coventry Climax. 

The squish area can be chosen to give the required pressure rise during 
combustion. For a high compression engine this area is usually smaller than one 
with low compression. In this excellent Coventry-Climax engine, once used in 

The Engine: Combustion 29 

several racing sports cars, the valve ports are given a downward inclination of 10 
degrees. Since the valves are inclined at 20 degrees in the same direction the gas 
flow path into the cylinder (or exit the cylinder in the case of the exhaust valve) 
is only diverted 60 degrees, against the 90 degrees change of direction with 
vertical valves and horizontal ports. 

The Heron head 

The ‘bowl-in-piston’ or Heron head has achieved a reputation for efficient 
combustion. When Jaguar Cars Ltd. decided to develop a V-12 engine for their 
cars they made two prototypes, a double overhead camshaft engine (with two 
camshafts per bank) and a single cam design with a shallow bowl-in-piston 
combustion chamber. This design is very attractive to the production engineer. 
By transferring the combustion chamber to the piston crown the cylinder head 
face becomes a flat machined surface. The problems of matching combustion 
chamber volumes are considerably eased when the volume is transferred to the 
piston crown. 

Fig. 2.4 The Heron (bowl-in-piston) Head; a cross-section of 
the Jaguar V-12 single cam engine. 

30 The Sports Car 

A cross-section of the single cam V-12 Jaguar cylinder head is shown in Figure 
2.4. The prototype had a squish clearance of 1.25 mm (0.050 in). Cutouts were 
required in the outer rim of the piston crown to clear the valve heads. Improved 
combustion was given with reduced squish and the final design had a squish 
clearance of 3.8 mm (0.150 in) and no valve cutouts. Harry Mundy, the Chief 
Engineer in charge of Power Units at Jaguar Cars was much impressed by the 
potential of the Heron head, in particular when used on an over-square engine. 
The single cam engine has a cylinder bore of 90 mm and a stroke of 70 mm. 

The Hemispherical head 

Perhaps the most successful engine in all sports car racing has been the 4.5 and 
5-litre flat 12 engine in the Porsche 917 (see Figure 2.5a). In its final stage of 
development as an unsupercharged engine it was producing 125 bhp per litre and 
seemed to prove conclusively that the hemispherical head was indeed the 
ultimate power producer. The two valves per head in this Porsche 917 were 
indeed enormous. The inlet valve was 47.5 mm (1.87 in) in diameter and the 
exhaust 40.5 mm (1.6 in). Their combined diameters exceeded the bore size of 
86.8 mm (3.42 in). For an unsupercharged engine it does not, however, now 
appear to be the ultimate since the 4-valve-per-head design has now become the 
norm in racing engines. 

A good example of the modern hemispherical head is that used in the Lancia 
Beta 2000, shown in cross-section in Figure 2.5b. 

Much ingenuity is often shown when designing down to a price. Paradoxically 
the Chrysler Hemi-head V-8, introduced as a moderately-priced production 
engine as long ago as 1951 (see Figure 2.6) was so successful in American stock 
car racing that the finally developed competition engine was very expensive. 

The hemispherical head and its close approximation, the pent-roof head, 
when used with very high compression ratios does not give a compact 
combustion chamber and at ratios of 11.5 and 12 to 1 as used in Formula 1 
engines it begins to look in cross-section like a piece of orange-peel. In this 
respect only can the Heron head be said to be superior. If one discounts the large 
concentric ring of squish area, the cylindrical combustion chamber in the centre 
of the piston can be made quite compact. 

The ability to produce power depends upon the ability to breathe in a high 
mass flow of fresh charge and to discharge the exhaust gases freely. A high mass 
throughput of mixture largely depends upon the provision of large valves. This 
then is where the hemispherical head scores. The valve area, depending upon the 
included angle between the valves, can be anything from 25 to 35 per cent 
greater than that of a typical flat-head engine. The specific power (the power per 
litre) is correspondingly greater. 

Fig. 2.5 (a) The Hemispherical Head: two-part cross-section of the Porsche 917 , 

5-litre engine, (b) The Hemispherical Head in production: cross-sections 
of the Lancia. Beta 2000. 



32 The Sports Car 

Fig. 2.6 The Chrysler Hemi-Head V-8, in which 
a single camshaft in the valley between 
the banks operates inlet and exhaust 
valves on both banks through push-rods. 

The Pent-roof head 

The alternative shape to the hemispherical head resembles a typical roof of a 
house with the inlet valve (or valves) on one side and the exhaust on the other. It 
is the logical shape to choose for the power-producing 4-valve head, although a 
four valve head has been made by BMW that approximates to a hemispherical 
design by arranging the valve stems to radiate outwards from some central point 
in a three-dimensional plan. This creates something of a headache for the valve 
gear designer. The pent-roof, with inlet valves on one side and exhaust on the 
other, is well-known in racing circles and is usually associated with the name 
Keith Duckworth of Cosworth fame. 

After blowing the dust off the 1925 catalogue for W.O. Bentley’s new 3-litre 
sports car we see that particular attention is drawn to the provision of ‘four 
valves in each cylinder’. One reason given for using four instead of two valves is 
given as ‘efficiency — namely, to get the gas in and out of the cylinders as 
quickly as possible.’ Another reason, so the catalogue states, is ‘reliability’ and 
this reason applies as much to a lorry engine as to a private car. By using two 
valves instead of one the seating area is increased by 50 per cent, and in 

The Engine: Combustion 33 

consequence the cooling surface is greater, and a greater volume of water can be 
circulated through the space surrounding the seatings. Further the hammering 
effect on the seating of a single large valve with a strong spring is greatly 
diminished by using light valves with light springs.’ 

Who could have put the case for the 4-valve head better, even after more than 
sixty years of automotive progress, than the great ‘W.O.’ himself? In 1925 
W.O. Bentley was not only the engine and chassis designer, the chief salesman 
and financial director. He was also the copywriter for the catalogues. 

The 4-valve head became associated for many years with that other famous 
long-stroke engine, the 4-cylinder Offenhauser, so successful for so long at 
Indianapolis. As stroke/bore ratios gradually decreased and engine speeds went 
up from 4,000 rpm to 8,000 rpm and higher there became established a general 
belief that the 4-valve head was an unnecessary complication. The Ford Motor 
Company of America changed our thinking with a sharp jolt when they 
designed an engine for Indianapolis competition with a 4-valve head. The logical 
successor to this was the Ford-Cosworth Type DFV V-8 Formula 1 engine which 
completely overwhelmed the opposition for a decade and is still used in the 
majority of Grand Prix cars. 

The Type DFV has a shallow pent-roof head with an included angle between 
valve stems of only 32 degrees. The traditional pent-roof head, as used by 
Ferrari and Maserati in the past had an included angle of 65 to 80 degrees. The 
provision of two separate rocker boxes to each bank of cylinders gave these 
engines a distinctly top-heavy appearance. The Cos worth design with its smaller 
valve angle gave the designer the chance to enclose the two camshafts in a single 
neat cambox and cover, giving a substantial saving in weight and bulk. 

A larger included angle would, of course, have permitted the use of even 
larger valves and this alternative is always a future possibility. The Cosworth 
DFV engine was given two inlet valves of 33.5 mm (1.32 in) and two exhaust 
valves of 29.0 mm (1.14 in) for a bore of 85.7 mm (3.38 in). In its first year of 
competition the engine gave its peak power at 9,000 rpm. This has now been 
increased to 11,000 rpm with the provision of larger valves. 

As a typical sports car alternative to this very expensive machinery we should 
examine the latest 2-litre Triumph engine. Taking the advice of such old hands 
as Walter Hassan and Harry Munday, who had already opted for a 2-valve 
Heron head on the V-12 Jaguar, Spencer King decided that a 4-valve head would 
be commercially viable on an engine with only 4 cylinders. A cross-section of the 
Triumph Sprint engine is shown in Figure 2.7. As shown here the head joint face 
is horizontal. When fitted to the canted engine block however it is inclined 45 
degrees to the right, i.e. with the inlet port almost vertical. Only one camshaft is 
used with 8 cams. Each cam serves to open an inlet valve and the transversely 
located exhaust valve. The inlet valve is opened directly via a bucket tappet, the 
exhaust valve via a rocker. 

As in the Cosworth engine the included angle between the valves is low, only 
35 degrees and the camshaft and rockers can be enclosed under one cover. The 
sparking plus is at the centre of the four valves, the obvious location but one 

34 The Sports Car 

Fig. 2.7 The 4-valve Pent-roof Head of the Triumph Dolomite Sprint, (a) Because 
steel - instead of cast iron - rocker is used, end wiped by cam has to be 
lubricated. Seal over exhaust valve is to prevent oil dropping down into it 
under idling conditions, (b) Comparison of four valve head with earlier 
two valve design shows both to be identical height. 

difficult to achieve. 10 mm sparking plugs have been used in such tight locations 
in the past, but they are known to be troublesome, with a very narrow heat 
range. The Triumph solution is novel. A special 14 mm plug is used but slimmed 
down to take a 16 mm A/F spanner. This plug is screwed into a boss at the 
bottom of a tubular housing. 

The engine: induction 
and exhaust 

There the thundering strokes begin, 
There the press, and there the din.' 


The induction system 

The basic principle of induction in the four-stroke engine is simplicity itself. 
Efficient induction, however is not as easy, especially at high engine speeds. On 
a two-stroke engine, where we lack the positive displacement of the descending 
piston to induce the charge, a clear understanding of gas dynamics with a little 
dash of witchcraft is needed to induce even half the charge captured by the 

Fortunately, in the modern sports car engine, we are only concerned with the 
four-stroke cycle. We are, however, concerned with a mixture of air and fuel, 
much of this fuel being in the form of finely divided liquid droplets that are 
carried by the air stream into the cylinders. At light loads and at idle about 25 to 
30 per cent of the fuel is in liquid form as it enters the manifold. At full power, 
even though engine temperatures are higher, the flash-boiling effect of a liquid 
entering a region of low pressure (the induction manifold) is lost, since the 
throttle plate is wide open, and the amount of liquid fuel is much higher, as 
much as 60 per cent in some cases. 

The more bends there are in the induction system the more danger there is of 
the mixture segregating into richer and weaker streams, some cylinders receiving 
the former and the others the latter. In a competition engine the problem of 
segregation is solved by the provision of individual induction pipes and 
individual fuel metering jets. The Porsche 917 induction pipe shown in Figure 
2.5a is a good example. Production engines do not usually have one carburettor 
(or one induction pipe) to each cylinder and the behaviour of an induction 
system can make or mar an engine. 

A designer must strike a satisfactory balance between the conflicting require¬ 
ments of providing easy passage for the air flow to achieve a high volumetric 

36 The Sports Car 

/ / / / y / / y / y y x .y /■ y / / / y y y y / / / y / y // 

12 3 4 5 6 


Fig 3.1 (a) Production ‘log-type’ intake manifold for American Motors Rambler 

6-cylinder engine, (b) Full ram manifold for Rambler, (c) Modified ram 
manifold for Rambler, (d) Air-fuel ratio comparison of experimental 
and production manifolds. Tests conducted at 1,000 r.p.m. 

The Engine: Induction and Exhaust 37 

efficiency and providing passages that impart sufficient turbulence to keep the 
liquid droplets in suspension. Good distribution means supplying equal quanti¬ 
ties at the same mixture strength to all cylinders. It was not uncommon in the 
past to find induction systems that collected pools of liquid fuel on the floor of 
the manifold at low engine speeds. Figure 3.1 shows the poor distribution given 
on several experimental manifolds used on an American Motors Rambler 
6-cylinder engine fitted with a single carburettor. A log-type manifold, i.e. 
straight pipe with right-angle bends, gave the best distribution but the lowest 
volumetric efficiency. The more streamlined ‘ram’ manifolds gave good volu¬ 
metric efficiency, but unacceptable distribution. 

Good distribution would present no problem if we were to heat the manifold, 
especially where the carburettor branch enters the main gallery. Enough heat 
applied here would vaporize all the fuel and the distribution of a gas mixture 
presents few problems. Unfortunately we cannot do this, particularly on a 
sports car engine, where performance is so important. The maximum power 
developed by an engine depends upon the mass of air it can breathe in a given 
time. By heating the air density is reduced and the mass air flow is reduced in 
direct proportion. For power considerations, then, we have no escape from the 
problems of handling wet mixtures, unless we inject the fuel directly into the 

The 4-cylinder in-line engine 

Turning now to a typical system as used on a 4-cylinder sports car engine, the 
twin-carburettor manifold, we find that our efforts to distribute the mixture 
evenly between cylinders are hindered by the firing order, i.e. the order in which 
the induction pulsations occur. Taking the usual firing order of 1-3-4-2 we see 
that the negative pressure pulse caused by No. 2 cylinder’s induction is 
immediately followed — and indeed slightly overlapped — by that from No. 1 
cylinder (see Figure 3.2). On the other hand, No. 1 pulse is followed by a period 
of about 360° before the next No. 2 pulse occurs. From this we see that the 
beginning of induction on No. 1 cylinder will tend to rob No. 2 cylinder of 
charge at the critical stage when No. 2 inlet valve is on the point of closure. This 
effect will be greatest with siamesed ports; the shorter the distance between 
adjacent valve seats the greater the chance of mutual interference during this 

Fig. 3.2 

38 The Sports Car 

period of overlap. The system is asymmetrical and it is not to be expected that 
equal charges will be drawn into Nos. 1 and 2 cylinders over a reasonably speed 
range. The same state of affairs will of course exist in Nos. 3 and 4 tract. A good 
solution to this distribution problem was found many years ago in a simple 
bleed or balance pipe between the two parts of the induction system. In this way 
when No. 2 cylinder is being robbed of charge as its inlet valve is closing, there is 
a flow of gas along the balance pipe or duct from the inner end of the opposite 
tract. Since the induction in No. 3 cylinder does not occur immediately after No. 
2, this does not rob No. 3 cylinder of charge. The diameter and length of the 
balance pipe is usually determined on the test-bed by trial and error. This system 
gives satisfactory distribution, but by no means perfect distribution. An 
excellent 4-cylinder manifold giving good power at high speed and acceptable 
low speed distribution is shown in Figure 3.4 (modified). 

We must state, at this stage, that our analysis contains a simplifying 
assumption in that the more complex effects of pressure waves or pulses in the 
inlet and exhaust systems are neglected. The special cases of ‘tuned’ or ramming 
induction pipes and exhaust pipes will be treated fully later in the chapter. 

The 6-cylinder in-line engine 

When we turn to the 6-cylinder engine, in its more popular in-line arrangement, 
there are three cases to be considered, the twin carburettor, the triple carbu¬ 
rettor and the six-carburettor. At first sight it would appear that the triple 
carburettor system is a better one than the twin. In Figure 3.3 ( a ), however, it is 
seen that while no actual overlap occurs between the opening periods of paired 
cylinders, the induction pulses for Nos. 1 and 2 pair and Nos. 5 and 6 pair do 
not occur at regular intervals. The pulses for the centre pair, Nos. 3 and 4, do 
occur at regular time intervals. The irregular pulses of the two outer pairs would 
be of no great importance if it were not for the complex wave system that 
persists in the induction tract and which influences the pressures in the cylinders, 
in particular at the crucial moment when the inlet valve is nearly closed. With an 
irregular pairing of cylinders it is not difficult to see that the charging of one 
cylinder might, at one particular engine speed, be helped by the residual pressure 
waves in the induction pipe, while the paired cylinder might have its charging 

In some 6-cylinder engines using three carburettors slightly smaller main jets 
are used in the outer carburettors to compensate for the irregular pulses. At first 
sight this appears odd, but tests have shown that for a given mean air rate a 
steady air flow through a choke induces a lower mean fuel flow from a given jet 
than an irregular flow; the more irregular the air flow, the greater the mean fuel 
flow from the jet. Thus to obtain the same mean mixture strengths throughout, 
smaller main fuel jets are required on the outer carburettors where the air flow is 
more irregular. 

Figure 3.3 (b) shows diagrammatically the twin carburettor system. With the 
usual engine timing of 1-5-3-6-2-4 the induction pulses are symmetrically at 
intervals of 240°. The induction tract itself, however, cannot be made 

The Engine: Induction and Exhaust 39 

1 2 

3 4 5 6 

Firing Order: 753 624 

l ——a 

7 2 3 4 5 6 

Firing Order: 153624 

symmetrical. If the carburettors were placed directly opposite Nos. 2 and 5 inlet 
ports the volumetric efficiencies would be slightly higher on these two cylinders 
than on the other four. This in itself would not be a serious fault, but the more 
direct flow in to these two cylinders would cause them to receive a dispropor¬ 
tionate amount of fuel in the form of droplets, especially when the engine was 
cold. A good compromise is shown in Figure 3.3 where the carburettor branch 
enters between Nos. 2 and 3 cylinder ports on the front manifold and between 
Nos. 4 and 5 on the rear manifold. With this layout, Nos. 1 and 6 cylinders are 
the only ones in which the induction process is not preceded by a reversal of flow 
in the manifold. The flow of gas into No. 2 cylinder helps the beginning of 
induction on No. 1 cylinder. Similarly the flow of gas into No. 5 cylinder assists 
the start of the induction process in No. 6 cylinder. It is reasonable therefore to 
compensate for this advantage by making the overall path to Nos. 1 and 6 
cylinders greater than the paths to the other four cylinders. 

From the above it would appear that neither twin carburettors nor triple 
carburettors can guarantee perfect distribution to a 6-cylinder engine. The triple 
system gives the shortest and most direct path from carburettor to valve and is 
usually the best power producer. The twin carburettor system is usually the most 
economical. While the earlier Jaguar engines had twin carburettors the search 

40 The Sports Car 

Fig. 3.4 Changes to 4 cyl. Coventry Climax F.W.A. 
engine to increase power. 

for greater specific power outputs in the later versions such as the 3.8 and the 
4.2-litre E-Type engine led to the fitting of three H8 S.U. carburettors. Curved 
Y-pipes were used between each carburettor and the ports in the head, the curves 
being asymmetrical and of differing lengths for the outer pairs. 

To extract the maximum power from a given size of engine a separate fuel 
metering system must be provided for each cylinder. The fuel metering system 
may take the form of three twin-choke carburettors, six single carburettors, or a 
system of individual fuel injection in which a high-pressure pump meters fuel to 
nozzles fitted in the valve port or into the cylinder head itself. Using multiple 
carburettors is a clumsy and heavy solution now that several makes of fuel 
injection are available and have proved to be reliable. The beauty of the 
individually metered system lies in the smooth unbroken passage from intake to 
valve head. There are no branches, no subdivisions of the gas flow to upset 
distribution of the fuel, no possibility of one cylinder interfering with the 
breathing of the next. A good distribution is assured with a much lower heat 
input to the charge than is possible with any other system. 

The V-8 engine 

The V-8 engine is the popular choice in America today, despite the rising cost of 
gasoline. Three types of carburettor and manifold system are in common use on 
American V-8’s. The oldest system, as one might expect, is the single carbu¬ 
rettor with one choke tube or venturi. Later came the two-barrel, then the 

The Engine: Induction and Exhaust 41 

four-barrel carburettor. For competition use, as in stock car racing or drag 
racing, it is common to see three two-barrel carburettors feeding the eight 
cylinders through a plenum chamber. A more logical scheme is seen on a recent 
Ford competition engine in which two four-barrel carburettors are used. 

For a sports car engine we need not consider the single- or the two-barrel 
carburettor. Four-barrel types usually have two float chambers, each with its 
own float and needle valve, an idle circuit with two branches (feeding only the 
primary venturis) a main discharge circuit (feeding all venturis) four throttle- 
plates (one to each venturi) and one accelerator pump. The manifold is arranged 
with galleries at two levels as shown schematically in Figure 3.5, each level 

12 3 4 

Fig. 3.5 Layout of manifold used with single 
4-choke (four-barrel) carburettors on 
Ford V-8. 

feeding two cylinders on one side and two on the opposite side. The two venturis 
provided with both idle and main metering circuits are called ‘primaries’. The 
other two are called the ‘secondaries’. Each primary and secondary pair feeds 
the outer cylinders of one bank and the inner cylinders of the other. The 
throttle-plates on the secondaries remain closed at low air velocities. In this way 
a high gas velocity is maintained through the primary venturis even at low 
engine speeds, thus helping to break up the fuel spray effectively and to 
maintain good distribution. The opening of the secondary throttle-plates can be 
controlled by a mechanical delay mechanism or, as is becoming more common, 
can be triggered by a second pair of weighted auxiliary throttle-plates that only 
open when a critical differential pressure exists between the atmosphere and the 
manifolds. In this way the four-barrel carburettor is a fine, though complicated, 
instrument, designed to give good torque at low speeds and low pressure drop at 
high speeds. 

The dual four-barrel manifold, as used on the Ford 427 cu. in. (7 litres) single 
overhead cam engine, is shown schematically in Figure 3.6. Each four-barrel 
carburettor was designed so that the primary and secondary venturis diagonally 
opposite each other feed two cylinders, one in each bank. The firing order of 
this engine (using the Ford System of cylinder numbering) is 1-5-4-2-6-3-7-8. By 

42 The Sports Car 

12 3 4 

Fig. 3.6 Layout of manifold used with dual 4-choke 
(four barrel) carburettors on Ford V-8. 

feeding cylinders in the chosen pairs, successively firing cylinders are fed 
alternately from upper and lower galleries. The improvements in torque and 
power given by this system over the single four-barrel manifold is shown in 
Figure 3.7. To European ears the sound of a V-8 engine giving an output of 
616 b.h.p. at 7,000 r.p.m. (albeit S.A.E. horse-power) is an awesome thing. It is 
also impressive that the power does not appear to be limited by breathing, since 
the power curve is still climbing at the Ford-imposed limit of 7,000 r.p.m. 



3 40 
3 20 

1000 2000 3000 4000 5000 6000 7000 




Fig. 3.7 

Full throttle performance of 427 cu. in. single o.h.c. Ford V-8 

The Engine: Induction and Exhaust 43 

Fig. 3.8 Cross-section of Jaguar twin o.h.c. engine. 

One choke per cylinder on the V-8 

One cannot attempt to predict what the ultimate development might be in the 
breathing of the overhead valve engine when fitted with one choke per cylinder. 
Many believed that the hemispherical and pent-roof head design was reaching its 
limit of development and that the classical twin o.h.c. two-valve head as seen in 
the Jaguar, the Aston Martin and the Ferrari could only enjoy marginal 
improvements in the future. It was known that the Offenhauser engine which 
has powered so many successful Indianapolis cars for so long was limited more 
by the high inertia loadings of its big-bore long-stroke design than by the 
breathing potential of its 4-valve cylinder head. The Ford Motor Company were 
well aware of this when they decided to make a V-8 engine to challenge the 
Offenhauser. They made an engine with two overhead camshafts to each bank 
of cylinders and two inlet and two exhaust valves to each cylinder, thirty-two 
valves in all. Careful calculations had shown that a higher output would be 
given by such a head than could ever be achieved by a 2-valve head. 

Two port layouts were tried, as shown at the left of Figure 3.10. In both 
systems the exhaust ports turned outwards and can be referred to as ‘horizontal 
ports’ if we regard the cylinder axis as vertical (actually 45° to the vertical on a 

44 The Sports Car 

Fig. 3.9 Cross-section of Mercedes 350SL V-8 engine. Note the downstream fuel 
injection directed into the inlet port. 

90° V-8). In the upper case shown in Figure 3.10 the inlet port is also 
‘horizontal’; in the lower case it is ‘vertical’. The tests on the air-flow rig showed 
superior breathing with ‘vertical’ inlet port. 

Ramming induction pipes 

The use of ramming induction pipes, ‘tuned’ lengths of inlet pipe, to enhance 
the output of an engine over a limited speed range is no new discovery. It was 
used on stationary constant-speed Diesel engines before the war, a well-known 
case being the two-stroke Diesel engines designed by Kadenacy for Messrs 
Armstrong Whitworth. By ‘tuning for ram’ one means that the lengths of both 
inlet and exhaust pipes have been varied experimentally until the optimum 
lengths have been found which give maximum power. The engine has, of course, 
been fitted to a dynamometer while these experiments were carried out. In this 

The Engine: Induction and Exhaust 45 

Fig. 3.10 Air flow comparisons on Ford d.o.h.c. engine. 

way the peak power can sometimes be increased by as much as 15 per cent over 
that given by an engine with two or three carburettors and the usual standard 
exhaust manifold and silencer. About 10 per cent of this gain can come from the 
adoption of a separate carburettor per cylinder, with a tuned inlet pipe either 
beyond the carburettor or between the carburettor and the engine. 

The late Freddy Dixon always fitted a separate carburettor to each cylinder on 
his racing Rileys and, despite the great air of mystery surrounding all his work 
and the padlocks always fitted to the bonnets, there is good evidence that he 
used tuned lengths of inlet pipe to give him that little extra power to beat the 
works team. The designer of the modern unblown Grand Prix engine makes full 
use of induction ram, and exhaust ram too, to fatten the torque curve exactly 
where the extra torque is most needed. Sometimes the lengths of the intake 
trumpets on the carburettors are changed between races. For a fast circuit with 
long straights, short intake horns will be fitted to give maximum ram at peak 
r.p.m.; for a twisty circuit, where acceleration is more important than top 
speed, longer intake horns will be fitted to improve the torque curve in the 
middle of the operative speed range. The use of ramming induction pipes has 
now spread to sports cars and is seen on all serious competition machinery. 

Let us now consider what is happening to the pressure waves in the induction 
system to make them have such a profound influence on performance. We 
could, of course, blindly install our engine on a test-bed and lengthen and 
shorten the pipes until the brake readings tell us we have got the right answer. 
Indeed when a dynamometer is available there is no doubt that the correct induc¬ 
tion length can be found in this way, but the man who is not so blessed must depend 
upon theory alone, hoping that the theory is not too far from the truth. 

46 The Sports Car 

Ramming pipe theory 

Induction and exhaust ramming have been the subjects of much research spread 
over almost a third of a century. As yet we cannot say we have the complete 
answer to all its problems; its full possibilities are not yet exhausted. Many years 
ago the author worked under Dr G.F. Mucklow, a British authority on the 
subject. Later, during the last war, the author was able to apply the knowledge 
thus gained to the analysis of several experiments carried out on ramming 
induction pipes fitted to two sizes of single-cylinder engine. From these 
experiments the author evolved a simple general formula for the design of 
ramming induction pipes. The formula is fairly accurate and in the majority of 
cases has been found to give quite close agreement with test-bed measurements. 
When abnormal valve timings are used the formula is less accurate. 

It is the inertia of the atmosphere that usually prevents us from achieving a 
volumetric efficiency of 100 per cent. There are ways, however, in which we can 
make this inertia work for us. Inertia may be defined as an inherent property of 
all matter by which it tends to remain for ever at rest , unless acted upon by an 
external force . When in motion , its inertia tends to maintain it at the same 
velocity , unless acted upon by an external agency . This universal objection to 
being pushed around is shared by solids, liquids and gases (and quite a large part 
of mankind too!). 

The external force to persuade the air to enter the cylinder comes from the 
downward motion of the piston. This creates a depression in the cylinder and 
this depression is, in effect, the driving force to move the air, since the pressure 
in the induction pipe is now higher than that inside the cylinder. The higher the 
speed of the engine, the greater must be the depression in the cylinder before the 
volume of air in the induction pipe can be accelerated. As the piston passes 
midstroke and begins to decelerate the inertia of the air column makes the air 
tend to ‘pile into the cylinder* at a greater rate than the displacement rate of the 
descending piston. At b.d.c. the depression is much reduced, but air still 
continues to flow into the cylinder as the piston is rising again on the 
compression stroke. This is when the inertia of the air is ‘working for us*. 

In a typical modern engine the inlet valve will start to open about 20° before 
t.d.c. and close about 50° after b.d.c. This valve timing is a compromise since 
the inertia lag between the acceleration of the piston and that of the air column 
increases with speed. If we choose an inlet valve closing time to give maximum 
volumetric efficiency at 3,000 r.p.m., at lower speeds there will be a back flow 
of mixture from the cylinder into the induction pipe just before the inlet valve 
closes. At higher engine speeds than 3,000 r.p.m. mixture will still be flowing 
into the cylinder at the point of closure. Thus we find an early ‘inlet valve 
closing* in an engine designed to have a good torque at low speeds. Where 
torque at low speeds is sacrificed in order to boost the power curve at high 
speeds the i.v.c. is late. 

In ramming-pipe theory we are interested in the effects of inertia in the 
induction pipe itself. When the inlet valve opens just before t.d.c. and the piston 
begins to descend on the induction stroke a negative pressure pulse is created at 

The Engine: Induction and Exhaust 47 

the valve port which begins to travel outwards towards the open end of the 
induction system at the velocity of sound (about 335 metres per sec or 1100 ft 
per sec). For simplicity we are considering the induction system to consist of a 
single straight pipe of unvarying cross-section. This negative pressure pulse is 
similar in character, though not in magnitude or sign, to the positive pulse that 
would travel down the pipe if a small explosive charge were detonated behind 
the valve head. If we were to fit a pressure/time recorder at the valve, it would 
trace a curve roughly of the form shown in Figure 3.11 as ‘actual pressure 
diagram’. To the non-scientific man, whose mind usually becomes a little 

-Adiabatic expansion curve -* Theoretical pressure drop 

-Theoretical pressure drop C=0 7 

valve orifice coefficient 1 -Actual pressure diagram 

C = unity 

Fig. 3.11 

stubborn when the scientist or engineer begins to talk about waves or pulses, it is 
sometimes helpful to imagine the column of air in the pipe as a long line of 
railway trucks standing behind the shunting engine (the piston). The downward 
movement of the piston may be compared in this simile to the sudden pull of the 
shunting engine on the first truck. The initial ‘plug’ of air rushing through the 
inlet valve to try to fill the depression left by the descending piston may be 
likened to the leap forward made by the first truck. This stretches the spring 
coupling between the first and second truck and after about half a second the 
second truck leaps forward to follow the first. In a similar manner a second 
‘plug’ of air may be considered as leaping forward to fill the depression left by 
the first. The impulse travels down the whole line of trucks in just the same 
manner as the negative pulse travels down the induction pipe, only much more 
slowly. In both cases it should be noted that it is the movement of matter in one 
direction (the movement of air towards the piston and the movement of 

48 The Sports Car 

trucks towards the locomotive) that causes the disturbance or pulse to travel in 
the opposite direction. 

When the negative pulse reaches the open end of the pipe it will produce a 
rarefaction in the atmosphere near the end of the pipe. The surrounding air will 
rish in to fill this depression and by virtue of its inertia will produce a positive 
pulse which will travel back up the pipe with the speed of sound in air. This, in 
the language of physics is called a ‘reflected’ pulse, the first reflection. This 
reflected pulse, when it reaches the cylinder again, is reflected back towards the 
open end. The summation of these and subsequent pulses, recorded by suitable 
instrumentation, is seen as a complex wave form. Since the wave length is 
proportional to engine speed and the length of the induction system, it is 
possible to change the frequency and magnitude of the wave system by altering 
the length of the induction pipe. The science of ram tuning is to choose the pipe 
length so that these pressure waves assist the charging process. In simple terms a 

Fig. 3.12 Effect of intake pipe length on volumetric efficiency (D Type 
Jaguar engine). 

The Engine: Induction and Exhaust 49 

positive pressure behind the inlet valve towards the end of the induction stroke 
assists charging by ‘ramming’ additional charge into the cylinder before the 
valve closes; a negative pressure lowers volumetric efficiency by extracting 
mixture during this critical period. 

Ram charging is most effective on engines with a separate induction pipe to 
each cylinder, although experiments carried out at Southampton University on a 
Ford Cortina with a single carburettor gave a power increase of about ten per 
cent by the provision of separate long branches to each cylinder. 

The effect of induction pipe length on an engine with individual pipes and a 
separate metering system to each pipe is well illustrated by Figure 3.12. Short 
pipes sacrifice power at all speeds, but the choice of a very long pipe such as 
810 mm (32.5 in) would cause a serious loss of power at 5,500 rpm. A length of 
about 430 mm (17 in) appears to be a good compromise for the D Type Jaguar 
and this was the length usually used when racing. 

Figure 3.13 is the author’s guide to the best ram length for a particular engine 
speed. It is only a rough guide since the length of the exhaust pipe and the 
pressure waves existing there also influence the shape of the volumetric 
efficiency curve. This is the cause of the modulating wave-form clearly seen on 
the curves in Figure 3.12. 

Fig. 3.13 

50 The Sports Car 

Forward-ram intakes 

The forward-ram intake is an expedient sometimes used to utilize the forward 
motion of the car to ram air at a higher pressure than atmospheric into the 
carburettor intake. The device has appeared on a few sports cars, but in the 
majority of cases the external scoop seen on the bonnet top is simply a ‘cold-air 
intake’, since the air duct between the scoop and the carburettors is freely vented 
to the engine compartment. With a true forward-ram intake the duct is sealed 
against leakage into the engine compartment. To prevent a serious disturbance 
of the carburettor metering the carburettor float chamber must also be sealed 
and the cover balanced to the pressure in the duct. On rear-engined cars the 
forward ram intake is usually placed just ahead of the rear wheel arch. 

Theoretically one can recover all the velocity head of the airstream, but in 
practice, since the duct must turn through 90° to enter the carburettors, the 
most we can hope to recover is about 90 per cent of the total velocity head. 

This velocity head is 

. pv‘ 

288 g 

where p = the velocity head, lb per sq in., p = the air density, lb per cu. ft, 
v = the car velocity, ft per sec, and g = the acceleration due to gravity, ft per 
sec. 2 . 

If the car is travelling at 150 m.p.h. (229 ft per sec) 

0.076X220 2 


= 0.4 lb per sq in. 

This represents a gain in power of about 3 per cent. Several years ago the Car 
Division of the Bristol Aeroplane Company carried out special tests on a 
forward-ram intake and confirmed that a forward-ram intake would give an 
improvement of 4 b.h.p. at an air speed of 150 m.p.h., or an improvement of 
2.7 per cent on the normal test-bed 145 b.h.p. Only high-speed sports cars can 
gain much from such an intake. At 75 m.p.h. the ram would be reduced to a 
quarter of the above value and the gain in power would be less than one per 

Cold-air intakes 

At the beginning of the chapter we mentioned the loss of power that a heated 
induction system would entail. The same objection applies to heated air entering 
the carburettor intake. Under-bonnet temperatures as high as 70°C have been 
measured on older designs of sports car. Apart from serious power loss, 
temperatures as high as that would probably cause vapour-lock in the fuel 
system. It is now generally appreciated that the carburettors can benefit from a 
separate cool air supply and with the provision of large low pressure-drop air 
cleaners there is no need to fear any increase in bore wear from the intake of fine 

The Engine: Induction and Exhaust 51 

road grit. In many cases the placing of the air scoop on the top of the bonnet has 
resulted in an improvement in the cleanliness of the air fed to the carburettors. 
The cold-air intake can be used alone, or combined with the forward-ram 
intake. If required to function without ram effect a free flow of air is permitted 
out of the air box which feeds the air filters. 

The gain in power from fitting a cold-air intake must obviously differ from 
car to car, since under-bonnet temperatures differ widely. If we take an average 
under-bonnet temperature of 40°C, with an atmospheric temperature of 15°C, 
the gain can be calculated from Charles’ law: 

Pc _ 273 + ^ 

Ph 273 +t c 

where p c = the density of air with a cold-air intake, Ph = the density of air with 
a hot-air intake, t c = atmospheric temperature, °C, = under-bonnet 

temperature, °C. 

Pc = 



■X p h = 1.09 p h 

52 The Sports Car 

This represents a gain of 9 per cent in air density, or 1 per cent for every three 
degrees drop in aspirated temperature. The gain in power will be the same, 
provided the mixture strength is adjusted to allow for the change in density and 
also provided there is no detonation limit o.r other limit preventing achievement 
of this power. In some countries a cold-air intake is not advisable owing to the 
danger of carburettor icing. 

The exhaust system 

In 1911 the late L.H. Pomeroy, designer of the Prince Henry Vauxhall said: 
‘Getting the gas out of the cylinder is a simple pumping action, but it is getting it 
in that makes the engine either a pig or a horse.’ Today with our knowledge of 
the interaction between intake and exhaust we would go further. It is a poor 
exhaust system that only gets the gas out of the cylinder; a good exhaust system 
should help the fresh charge into the cylinder. 

It is not possible to design an exhaust system on steady fluid flow theory. If 
the gas were flowing at a steady rate the branches and the main pipe could easily 
be sized so that a chosen total pressure drop was not exceeded. The flow is not 
steady and in the branches it is not even continuous. In the individual branches 
it is flowing for about 35 per cent of the time. Peak velocities in the branches 
therefore tend to be higher than in the main pipe unless their cross-sectional area 
is made at least 75 per cent of the main pipe area. 

The silencer 

For maximum power we require a straight-through system. The use of an 

The Engine: Induction and Exhaust 53 

absorption type silencer (glass-pack of Hollywood) provides the system with the 
lowest pressure drop, but the effectiveness as a silencer is often in doubt, 
especially in the low-frequency end of the audible note range. Glass fibre is 
almost invariably used as the absorbent material in the annular chamber around 
the perforated centre-tube. This has inadequate strength at the operating 
temperatures of most sports car exhaust systems and the weakened glass fibres 
soon disintegrate and disappear out of the tail-pipe. 

In figure 3.15 we show three types of exhaust system. Either type (< a ) or type 
(6) are good smooth flowing designs that lend themselves to pulse tuning. Type 
(c) is shown more as a warning. Cast manifolds of this type used to be popular 
on mass-produced cars and only helped to cut down the power output of the 

Ramming exhaust pipes 

The same principles of ramming pipe design as were discussed earlier in the same 
section on ramming induction pipes can also be applied to the exhaust pipe. 
With a straight-through silencer only a small fraction of the amplitude of the 
original pulse is damped by the absorption material of the silencer, although the 
amplitude of the pulse does become of negligible size after three or four 
reflections. To obtain ram from an exhaust system our aim is to make the 
exhaust system extract residual gas from the cylinder head during the period of 
overlap between the opening of the inlet valve and the closing of the exhaust 
valve. Overlap may be a modest 30 to 40 degrees on a touring engine. On a 
competition engine it can be as much as 120 degrees. A depression in the exhaust 
port at this time helps to extract the hot residual gases from the combustion 
chamber thus helping to lower the charge temperature. By creating a depression 
in the cylinder it also helps to start the movement of the new charge into the 
cylinder through the already open inlet valve. With normal overlaps an increase 
in volumetric efficiency of from 5 to 7 per cent can be obtained in this way. If we 
are willing to sacrifice fuel economy and a smooth idling speed, greater gains are 
possible by using overlaps of 50, 60 and even 80 degrees. In this way the last 
traces of the residual gases can be sweapt out of the combustion chamber at the 
expense of a certain amount of the new charge going to waste down the exhaust 
pipe. Unfortunately when very large overlaps are used the smooth running of 
the engine is adversely affected at engine speeds where a positive pressure is 
produced in the exhaust port during overlap. Extreme overlaps are therefore not 
conducive to good idling. 

Figure 3.14 is the author’s attempt to provide a rough guide to the ramming- 
pipe length for an exhaust system using separate pipes to each cylinder. This 
would make a very bulky exhaust system on a sports car where every pipe would 
require a separate silencer. 

Branched exhaust pipes 

A tuned length of pipe is still possible with a multi-branch exhaust system. The 
modern rear-engined GP car is now a familiar sight with its ‘bunch of snakes’ 

54 The Sports Car 

system. Care is taken to tune the lengths of the branches as well as the two main 
exhaust pipes. Branches are grouped in fours on the V-8 engine, the two inner 
pipes from one bank of cylinders being carried across to join up with the two 
outer pipes from the opposite bank. In this way even pulsations occur at the 
point where the four branches join their main pipe. The dynamometer is the best 
instrument we can use for this determination of optimum lengths, but Figure 
3.14 will be of value to the tuner of modest means. 

The engine: valve gear 

'Conroe, let me read the oft-read tale again - 
Of pregnant parts and quick inventive brain.' 


Perhaps it is nothing more than a piece of English folklore, but it is told of the 
early days of Steam that the first valve gear was invented by a lazy boy whose 
job in life was to pull the right rope at the right time to admit steam to the valve 
chest and to pull the other rope to cut it off again when he saw the cross-head 
begin to reverse its direction. He must have been a very thoughtless child. 
Automation has put many of us in later generations among the ranks of the 
unemployed — but this is material for a different sort of book. 

Push-rod valve operation 

Push-rod valve operation is a little more sophisticated than using ropes. The 
cams operate tappets (cam-followers) and these transmit motion to tubular rods 
which in turn transmit the motion via rockers to the valve stem ends. To replace 
the rope to close the valve a coil spring or springs perform the closing operation. 
The system appeals to the cost-conscious production engineer since the camshaft 
can be placed low down on the engine to be operated from the end of the 
crankshaft by a relatively short chain drive. The disadvantage lies in the long 
distance between the camshaft and the valves it has to operate. This increases 
the inertia of the valve mechanism on high-speed engines. There is also a certain 
springiness in the whole system that leads to a distortion of the valve opening 
period and rates of opening and closing as defined by the shape of the cam. 

On a V-8 engine a single camshaft can be placed in the valley between the 
banks of cylinders and used to operate valves on both banks. This, naturally, 
brings a glow of happiness to the heart of the production engine. Even so the 
ingenuity of master tuners like Harry Weslake has demonstrated what can be 
done with this simple basic layout if you only have a ‘quick inventive brain.’ 

Fig. 4.1 Four-valve Chevrolet head designed by Harry Weslake using 
a single camshaft to operate 32 valves through push-rods. 

Figure 4.1 is a cross-section of the prototype 4-valve head designed at the 
Weslake establishment at Rye in Sussex for the 5-litre Chevrolet engine. In the 
standard engine the sequence of valve operation along the length of the 
camshaft for the 16 valves (4 inlet and 4 exhaust per bank) is exhaust, inlet, 
inlet, exhaust , etc. To operate 32 valves from a single camshaft, a new camshaft 
carrier casting was made to accommodate forked rockers in two rows. The cam 
sequence was also changed to exhaust, inlet, exhaust, inlet , etc. The porting and 
valves were developed by the well-known Weslake gas-flowing technique 
described in Chapter Two. In 1974 when this work was undertaken the current 
racing 5-litre Chevrolet with 2-valve head developed 490 bhp. The Weslake 
4-valve head turned out 600 bhp at 7500 rpm. This is an interesting case-history 
to illustrate that push-rod valve operation is not to be dismissed too lightly, even 
though no one in his right mind would attempt to design a competitive Grand 
Prix engine using push-rods. The Weslake engine delivered 120 bhp per litre, 
while the current GP engine has a specific output of 179 bhp per litre and 
operates at speeds up to 12,000 rpm. 

The double overhead camshaft valve operation 

The double overhead camshaft (DOHC) cylinder head has a long racing 
tradition. Figure 4.2 shows how the two camshafts were often driven by a very 
complex gear train in the traditional DOHC racing engine of the Fifties. The 
method was noisy but very reliable. In the cross-section of the Jaguar engine 
shown in Figure 3.8 in the previous Chapter the typical valve operation through 

r * \ || 4 W||MWn^W 

■! Jl^'w 

■'i'I^L *T-T[TTTVf7W 


Fig. 4.2 Twin overhead camshaft Aston Martin engine used in DBR1 racing 
sports car. Note complexity of gear train to take drive from 
crankshaft to the two camshafts. 

bucket tappets is shown. Not all designers have adopted the bucket tappet. 
Daimler-Benz (see Figure 3.9) prefer to use a rocker between the cam and the 
valve stem. Adjustment of valve clearances on the bucket tappet system is very 
time-consuming. Circular shims of different thicknesses are placed inside each 
tappet until the correct clearances are obtained throughout. The rocker on the 
Mercedes engine pivots at one end on a spherically-ended stud. Adjustment of 
valve clearance is a simple matter of screwing the stud in or out and re-locking in 
position with the lock-nut. 

58 The Sports Car 

The single overhead camshaft head 

When Jaguar introduced the 3.4 litre XK120 sports car in 1949 with a DOHC 
engine, many automotive production engineers asked the question, ‘How on 
earth do they do it at the price?’ Even so when those two experienced engine 
designers Harry Munday and Walter Hassan began to work on a new V-12 
Jaguar engine about twenty years later they found the prospect of providing 
four overhead camshafts and all the attendant drive gear a little daunting. They 
knew how to do it of course but there were still two problems to be solved; the 
obvious question of cost and the sheer size of the DOHC layout which had to be 
accommodated between the front wheel arches. 

As a design exercise they started by making two engines, one with DOHC and 
the other with SOHC. The first engine was such a power-producer that the 
Company was tempted to develop the engine as a dual-purpose power-unit, an 
engine good enough to return to the field of sports car racing again, yet 
satisfactory in a de-tuned form, as a very attractive production sports car 
engine. The provision of camboxes, tappet carriers and cylinder heads for a 
DOHC was no problem and followed established practice. The drive train to 
operate four camshafts at the top of the engine from a crankshaft at the bottom 
was no mean problem. For a pure competition engine they used a combination 
of two chain drives, followed by two separate gear trains as shown in Figure 4.3. 
This gave reliable service during the development period and a specific power of 

Fig. 4.3 Camshaft drive on Jaguar DOHC competition engine 
using two chain drives and two gear trains. 

The Engine: Valve Gear 59 

100 bhp per litre was soon attained. A boardroom decision however prevented 
further work on this racing engine and an attempt to reduce cost for a 
production version by the provision of multiple chain drives resulted in the 
layout shown on the right in Figure 4.5. This was not a reliable drive system and 
gave a noise level that was considered unacceptable for a luxury car. 

The SOHC engine was given a larger bore, 90 mm against 87 mm for the 
DOHC, but the valves were still smaller in the SOHC design and the air flow at 
maximum lift was reduced by about 30 per cent. The highest power output 
reached on the testbed with fuel injection to individual intakes and with open 







' 250 


— — — 87mmx70mm twin-cam competition 

cr 9-6 to f 2^0° com period 

1 875m <jia. inlet valve 

H20in din inlet port 

— —87mmx70mm twin-cam production 1 

cr 94 to 1 254° cam period 

1-875m dia inlet valve 

1 380in dra inlet port 





‘90mm x70mm single cam IB versien 
cr10 6to1 254° cam period A 




die ini 

et port 










,\ . 

\ ■ 


/ ^ 









1 _ 



i _ 




170 S 
160 ~ 
150 e 



1000 2000 

3000 4000 5000 6000 7000 8000 

Fig. 4.4 Power curves of single and twin-cam Jaguar engines. 

exhaust was still only 75 per cent of that reached by the DOHC engine (see 
Figure 4.4). The advantages of the SOHC design are summarised by Harry 
Munday as: 

(a) cheapness and lightness, 

(b) reduced engine bulk, in particular of the critical width across the 

60 The Sports Car 

exhaust manifolds, since .this limits wheel-lock angles with the wide tyres 
fitted to the modern Jaguar, 

(d) more space between the cylinder heads to accommodate a 12-cylinder 
ignition distributor and an air-conditioning compressor. 

In its final production form the SOHC engine with a compression ratio of 9.0 
to 1 gives 272 bhp (DIN) at 5850 rpm, a modest output of 51 bhp per litre, but a 
silky smooth torque flow reminiscent of a steam engine. 

The four-valve head 

An interesting design of engine based on the SOHC concept is the 16-valve, 
4-cylinder engine made by Triumph, originally for the Sprint Dolomite and 
intended for the TR7. The new 4-valve head was designed to increase the power 
of the older 2-litre Triumph SOHC engine with a 2-valve head. A cross-section 
of this head is shown at the right in Figure 2.7. The 4-valve head uses a bucket 
tappet for each inlet valve and a rocker actuated by the same cam to open the 
corresponding exhaust valve. The cam rotates in a clockwise direction. In the 
drawing the piston is near TDC with the inlet valve starting to open and the 
exhaust valve not yet returned to its seat. 

The ideal place for the sparking plug on a 4-valve head in the centre. To reach 
this difficult spot special plugs are used having 14 mm diameter screw threads 
where they enter the head, but with a hexagon size reduced to that of a 10 mm 
plug. This reduces the maximum plug diameter and of the tubular housing in 
which it fits. The plus is of the conical seating type to obviate the need 
for a gasket. The problems associated with retrieving a gasket from the 
bottom of this long narrow tube are obvious. (See Figure 4.6.) 

The Engine: Valve Gear 61 

Another successful 4-valve head is that used in the 2-litre Lotus Elite, Eclat 
Esprit sports cars. It was built originally for the Team Lotus racing cars. The 
engine is a slant-four canted at 45 degrees and using two overhead camshafts to 
operate the 16 valves. The manner in which the two camshafts are driven by a 
single timing belt (toothed belt) from a pulley on the end of the crankshaft is 
shown in Figure 4.7. The small pulley on the crankshaft nose is barely visible in 
the illustration, being shielded by the outer pulley that drives the alternator and 
water pump by means of a V-belt. The toothed pulley vertically above the end of 
the crankshaft drives a jackshaft to drive the oil pump and a horizontal 
distributor. The timing-belt is tensioned by an external jockey-pulley situated 
between the driving pulley and the jackshaft pulley. 

The downdraft carburettors shown here are not those used in the production 
engines, where horizontal twin-choke Dellorto carburettors are used to fit below 
the low bonnet-line. For safety reasons a sheet metal guard is fitted over the top 
of the belt drive on production engines. 

The timing belt gives a very neat inexpensive drive for the camshaft or 
camshafts and with the now proven reliability of the latest belts we can 
anticipate more designers turning to this system in the future. Modern timing 
belts use helically-wound polyester as tension members and polychloroprene as 
the basic compound. External abrasion is resisted by a two-ply nylon facing. As 
an example of what can be done if one is not afraid to festoon the front of the 
engine with whirling belts and pulleys we have given a front view in Figure 4.8 of 
‘the poor man’s instant Grand Prix engine’ as designed by Keith Duckworth. 

Fig. 4.7 Lotus LV/240 engine. Use of timing belts (internally toothed) to give 
inexpensive and reliable drive to DOHC engine. 

This Cosworth engine uses the 3-litre V-6 British Ford ‘stock-block’ as a basis 
for a 3,412 cc engine that develops ‘more than 400 blip’ at 8,500 rpm. 

The combustion chamber shape, valve and port configuration is based on the 
victorious Cosworth 3-litre Grand Prix engine and all is achieved using the 
bottom end of a perfectly ordinary mass produced push-rod engine. It is 
interesting to note that Mike Hall who was in charge of the development of this 
particular engine has found that timing belt drives are not only cheap and 
reliable, but they help to isolate camshaft and crankshaft vibrations. Many an 
expensive gear-train drive has foundered on that very problem. Gears provide 
great strength and reduce timing variations to a minimum, but there have been 

The Engine: Valve Gear 63 

Fig. 4.8 Cosworth-Ford V-6 with 4-valve DOHC arrangement using a multiplicity 

of timing belt drives. 

occasions when destructive interaction of torsional vibrations has been trans¬ 
mitted through the gears between the two vibrating systems, the crankshaft and 
the camshaft. In theory, a designer should be able to use the computer to 
eliminate this possibility. In practice, being only human, he is happy to use a 
drive that automatically absorbs these torsional vibrations. 

Desmodromic operation 

Older readers, and perhaps more youthful followers of the sport of motor¬ 
cycle racing, might well ask: what happened to desmodromic valve operation? 
On a conventional engine the cam only opens the valve, closure being brought 
about by the action of the spring or springs. As engine speeds rise so do the 
inertia forces and the only way to prevent the valve from bouncing off the end 
of the rocker or the top of the cam is to increase the strength of the springs. This 
increases the loads on the valve seats at lower engine speeds and is a source of 
valve-seat sinkage that calls for special valve-seat insert materials to resist this 
continual hammering action. Figure 4.9 is a plot of actual valve lifts as 
measured on a push-rod ohv installation. The upper plot shows that the cam 
profile is faithfully followed at an engine speed of 2,000 rpm. At 4,800 rpm 
(camshaft speed of 2,400 rpm); however, the valve bounces away from the 

64 The Sports Car 


Fig. 4.9 

Laboratory measurements of valve 
lifts on a push-rod engine with 
inadequate valve spring strength: 

(a) at 2000 r.p.m.; (b) at 4000 r.p.m. 


Fig. 4.10 

The desmodromic valve operation used by 
Delage in 1914 on their GP racing car. 

The Engine: Valve Gear 65 

rocker before reaching full opening, then falls back momentarily only to 
ricochet again off the nose of the cam. Contact is made again with the rocker 
end just before the point of closure. After a momentary closure the valve 
bounces again for a few degrees of crank rotation, making the actual point of 
closure an uncontrolled variable. Desmodromic valve operation is the logical 
solution, a solution that has been a long-standing challenge to our engineering 
ability. Desmodromic is a word created from two Greek roots: desmos meaning 
a band and dromas meaning running. It is therefore a valve gear confined to run 
in a band. When it first appeared on the 1914 Delage GP car the band was 
provided by a rectangular stirrup on the valve stem that embraced an eccentric 
cam (see Figure 4.10). A small ‘tolerance spring’ was used to pull the valve 
finally on to its seat, since the desmodromic cam mechanism only brought the 
valve within a few thousandths of an inch of the closed position. This problem 
of tolerance has been the downfall of so many attempts to provide a practical 
system. If, for example one were to provide a total clearance between the cam 
and the stirrup of, say, 0.10 mm (0.004 in) and this value was exceeded due to 
the sinkage and erosion on the valve and its seat, plus the differential expansion 
due to temperature changes between a cold engine and a hot engine, the valve 
could fail to reach its seat. This would naturally result in a burned-out valve in a 
very short time. 

Delage soon abandoned their ‘push-pull’ system, but there were many other 
attempts during the next forty years with no design reaching the level of 
reliability to warrant its adoption for a production engine. In 1954, however. 

Fig. 4.11 Desmodromic valve operation used by 
Mercedes-Benz on their 1954/55 
GP cars. 

66 The Sports Car 

Daimler Benz entered the Grand Prix arena with their world-beating Formula 1 
car fitted with a 2.5 litre straight-eight engine with a reliable desmodromic valve 
gear. Even so they never attempted to introduce desmodromic operation in a 
production engine and the suspicion remains that the excellent reliability was 
dependent upon meticulous attention to valve clearances by their team of 
dedicated racing mechanics. We need hardly mention that some production 
engines receive no attention until they fail to start. The prototype Daimler-Benz 
design used light springs for the final closure of the valves, but experience 
during development allowed them to dispense with these. With a fairly 
generous, but carefully maintained , clearance, the pressure inside the cylinder 
closed the valve on its seat. From a study of Figure 4.11 it will be seen that two 
cams per valve were used, one to close the valve, the other to open it. Each cam 
had its own rocker, positioned side by side on the same rocker shaft and with the 
forked ends turned inwards to bring them axially together. Extreme accuracy in 
cam machining is essential with such a system to maintain a close tolerance on 
clearances at all times if the valve is to follow the cam profile accurately. 

This was the last desmodromic system to be seen in motor racing to our 
knowledge, but the application to motor-cycle engines has proceeded apace. 
Modern racing motor-cycle engines reached permitted speeds of 20,000 rpm and 
will sometimes overspeed by as much as 3,000 rpm when going down through 
the gears when slowing down for a corner. In 1958, Norton modified their 
‘Manx’ racing engine to a very promising system but the necessary money to 
develop this desmodromic Norton was not available. In more recent times the 
Italian Ducati Company have spent many years developing a successful system. 
Unlike Daimler-Benz they were not able to dispense with ‘tolerance springs’ 
since, without these the engine gave no compression when cranked over slowly. 

The desmodromic system has so much to offer and we appear to be so close to 
a satisfactory solution. Surely someone somewhere can release a little brain¬ 
power from the pressing problems of anti-anti-missiles to give us a reliable valve 

The engine: fuel metering 

'So once it would have been - 'tis so no more: 
I have submitted to a new control.' 



The carburettor, like King Charles the Second, is ‘an unconscionable time 
a-dying.’ It has been developed over the years into an accurate and relatively 
inexpensive fuel metering device. In recent years the carburettor manufacturers 
have striven very hard to meet the increasingly stringent American emission- 
control regulations and their success has been not without some compensation, 
since they have not yet been put out of business. 

The simple carburettor uses a venturi to meter the fuel supply to match the 
mass air-flow and by a complicated system of compensating jets, emulsion 
tubes, accelerator pumps, idling circuits and other gadgetry it has for many 
years been able to match the air/fuel ratio requirements demanded by the 
project engineer over the whole operating range of the engine to an accuracy of 
plus or minus 5 per cent. Unfortunately, the project engineer who is responsible 
for getting his baby through the emission control tests now knows the acceptable 
limits on metering accuracy must be reduced to no more than plus or minus 2 Vi 
per cent, day in and day out, over wide ranges of ambient temperature and 
barometric pressure. This is the challenge that has faced the carburettor experts 
in recent years. 

The S.U. carburetter* 

The S.U. Carburetter Company have worked diligently for many years to 
eliminate or reduce many sources of mixture variation that are intrinsic in the 
basic method of metering by venturi. Since the S.U. instrument has been so 
popular in the past with sports car manufacturers it will serve to illustrate some 

♦There are two English and one American way to spell this word. The S.U. Carburetter Company 
have always used this one. 

68 The Sports Car 

of the problems encountered when the metering band has been tightened to plus 
or minus 2 Vi per cent. 

For those not familiar with the working principle of the S.U. carburetter here 
is a brief resume:- 

The modern S.U. carburetter comes in several different designs but the basic 
metering system is similar to that illustrated in Figure 5.1. At idling engine speed 


(«) ra 

Fig. 5.1 Cross-section of basic S.U. carburettor. 

the base of the piston (12) rests on the bridge (8), a small opening for the idling 
air supply being created by the provision of two small projections on the base of 
the piston. The depression downstream of the piston is in constant communica¬ 
tion with the upper part of the piston, the suction disc, through the passage (15). 
As the throttle is opened, the air flow through the gap between the piston and 
the bridge increases. This causes the depression downstream of the piston to 
increase. This increased depression causes the piston to rise, since the higher 
depression is transmitted to the top of the suction disc through passage (15). 
With an increased area at the choke (the choke being the space between the 
bottom of the piston and the bridge) the depression tends to fall again and the 
final new position of the piston is an equilibrium position where the depression 
above the suction disc exactly balances the weight of the piston, plus the 
additional load of the light spring (13). Without the damping action of the oil 
damper unit (14) any sudden change in throttle position would be followed by 

The Engine: Fuel Metering 69 

the oscillation of the piston as it overshot the equilibrium position at least once. 
As the throttle is opened and closed under the varying load demands dictated the 
the driver, the piston rises and falls to new equilibrium positions, the depression 
between the choke and the throttle always remaining constant close to the usual 
design value of eight inches of water head. As the piston rises and falls in this 
way under the influence of the changing air consumption of the engine, the 
position of the tapered needle (11) also changes, the annular space between the 
needle and the jet (10) being smaller for low air rates than for high. By choosing 
the profile of the needle to suit the mixture requirements of the engine at the 
various air rates almost any desired mixture can be metered over the whole range 
from idle up to full-throttle at the highest operating engine speed. 

The damper in the S.U. carburetter, besides its obvious action in preventing 
undue oscillations of the piston, has a secondary function in that it enriches the 
mixture during acceleration. Without an acceleration enrichment device a 
flat-spot would occur in carburation every time the throttle was suddenly 
opened. The effect of the damper on the S.U. carburetter is to delay the rise of 
the piston slightly when the throttle is snapped open. This causes the depression 
to rise above the normal steady value, and this enhanced depression acting on 
the fuel in the jet induces a greater flow of fuel. It will be seen that the delay in 
the rising of the piston will also keep the needle at a lower position in the jet, 
which in itself should tend to reduce the flow of fuel. The effect of the enhanced 
depression, however, is much greater and the combination of the two effects is 
still an increase in the strength of the mixture. 

In general, with a perfectly standard engine, only the maker’s recommended 
needle need be considered. If however the engine is modified in such a way as to 
increase the power output it becomes necessary to make dynamometer tests to 
choose a more suitable needle. It must be stated though that any form of 
modification to the fuel metering system, even a change in jet size or in needle 
profile, has become illegal in some countries where strict emission control 
regulations are enforced. 

We have described the simple S.U. carburetter as we have known it for half a 
century. We can now consider the modifications that were found necessary to 
give the more precise metering demanded by the current US Federal exhaust 
emission laws. Cars exported to North America are fitted with the HIF 
(Horizontal Integral Float Chamber) model. 

The HIF model incorporates the following eight new features: 

Spring-loaded jet needle . Readers familiar with the older S.U. carburetter 
will know how important it was to centralise the needle in the jet, or should we 
say try to centralise the needle. S.U. engineers working on the new problems set 
by the emission control requirements discovered that a needle set to be clear of 
the jet wall, but slightly off-centre (case a in Figure 5.2) would pass slightly more 
fuel under a given head than a perfectly concentric needle (case b). To remove 
this tuning anomaly they decided to spring-load the needle so that it always 
touched the side of the jet. The pressure is very light but contact with the 
jet wall is maintained over the entire range of needle movement. No jet centring 

70 The Sports Car 


(o) (b) (c) 

Fig. 5.2 Influence of needle centring on fuel flow. An off-centre 
needs (a) gives a higher flow than one perfectly centred 
(b). For consistent performance the latest carburettor 
has the needle spring-loaded to touch the jet wall (c). 

1. Piston rod 5. Needle guide 

2 . Transfer holes 6. Needle locking screw 

3. Jet needle 7. Needle biased in jet 

4. Needle spring 8. Etch mark 

Fig. 5.3 The spring-loaded jet needle. 

The Engine: Fuel Metering 71 

is required with this HIF model and one source of variation from specification is 

Concentric float chamber. With the float chamber placed in front of the 
main jet assembly (or the venturi in a fixed venturi carburettor) there is a 
tendency to enrich the mixture when accelerating and to give a leaner mixture 
when braking. Before emission control this was acceptable. In fact, in earlier 
days before we had acceleration pumps or other acceleration devices we 
considered it to be a very elegant solution to the problem of mixture enrichment 
during acceleration! 

1. Bi-metal assembly 4. Jet adjusting screw 7. Fuel inlet 

2. Concentric float 5. Si-metal pivot screw 8, Needle valve 

3. Jet head 6. Float fulcrum screw 9, Bottom cover-plate 

Fig. 5.4 An enlargement of the float-chamber in the S.U. Type HIF 

The HIF float chamber, shown in Figure 5.4, is designed to prevent such 
variations from forward, backwards or sideways accelerations of the car. The 
bowl is concentric with the jet assembly and the semi-circular float is shaped 
to surround the jet tube. The float pivot is parallel to the inlet flange. 

72 The Sports Car 

The concentric float is not a new idea and will surely become the normal pattern 
on all future designs. 

Fuel viscosity compensator. Even changes in fuel viscosity have created 
problems. As the temperature of the fuel rises, the viscosity is reduced and more 
fuel is metered through the jet at the same effective jet area. This means that the 
mixture tends to be weaker in winter than in summer. In the HIF carburettor the 
jet head is attached to a bi-metal strip (Part 1 in Figure 5.4). This strip is 
immersed in the fuel and lifts the jet to a slightly higher position as the fuel 
temperature rises. Temperature compensation makes it possible to pre-set the 
mixture setting (or during the official agent’s pre-delivery checks) and to seal 
this setting against illegal mixture adjustments. 

Part-throttle by-pass emulsion. The makers of fixed choke (venturi) carbu¬ 
rettors, such as Solex, Zenith and Weber have made great improvements in the 
degree of atomisation achieved during idle and part-throttle running. In the 
older S.U. carburetter the fuel is atomised fairly effectively at the jet. 
Unfortunately the throttle-plate is some distance downstream and tends to 
collect the spray of fine droplets on its front face and cause them to agglomerate 
into larger droplets as they are re-entrained in the airstream at the throttle-edge. 
In the new carburettor a small-bore passageway carries the mixture created at 
the top of the jet to a discharge point at the throttle-edge. A slot is provided in 
the base of the piston (see Figure 5.5) to direct the mixture into the small¬ 
bore passageway. During idle and slow-speed running the air velocity through 

2. Cold start enrichment outlet 4. Slot in piston 

Fig. 5.5 Part-throttle by-pass emulsion system in the S.U. Type 
HIF Carburettor. 

The Engine: Fuel Metering 73 

the passageway is much higher than the velocities achieved on the older 
carburettor design. Higher air velocities give more effective atomisation and this 
in turn leads to improved distribution between cylinders. 

Cold start enrichment. The traditional method of providing a rich mixture on 
the S.U. carburetter was to provide a lever to pull down the jet-head, thus 
increasing the area between the needle and the jet. The latest device is a rotary 
valve which admits augmented fuel supply to a drilled passage that emerges 
behind the jet bridge (2 in Figure 5.5). The action of the rotary valve is 
progressive and controlled manually. 

Overrun valve. During overrun, i.e. deceleration with a closed throttle, 
manifold depression can rise to as much as 25 inches Hg. The level depends 
upon the size of the engine, the weight of the car and the design of carburettor. 
Good combustion is difficult to maintain with such a high manifold vacuum. 
Under such conditions the hydrocarbon and carbon monoxide levels rise to 
unacceptable levels. The simple solution is to bleed mixture into the manifold 
through a spring-loaded valve located in the throttle-plate to limit the manifold 
vacuum to the chosen value. This value will be slightly higher than the normal 
idle vacuum. 

Sealed mixture adjustment. Mixture adjustment in the HIF model is achieved 
by moving the jet tube up or down relative to the needle as in all previous 
models. A screwdriver adjustment is provided on the new carburetter (4 in 
Figure 5.4). This acts on a right-angled adjusting lever, attached to the 
body-casting by a spring-loaded screw (5 in Figure 5.4). The bi-metal strip 
responsible for temperature compensation (1 in Figure 5.4) is riveted to the 
horizontal arm of this lever. It is therefore capable of moving the jet head up 
or down independently. When screw 4 is turned inwards the jet tube is lowered 
and the mixture is enriched. To prevent unauthorised changes in the pre-set 
mixture strength in countries where the emission laws make it illegal to do so the 
mixture adjusting screw is in a recessed boss which can be sealed with a metal 

Crankcase emission control. One of the early requirements for cars supplied 
to the smog-bedevilled State of California was a system to prevent the escape of 
blow-by gases from the crankcase. Since the gases that escape past the piston 
rings are very rich in unburned hydrocarbons they contribute considerably to 
the sum of all other hydrocarbon pollutants in the urban atmosphere. In the 
S.U. installation these fumes are drawn from the crankcase into the constant 
depression zone between the throttle-plate and the carburetter piston. An oil 
separator is provided to prevent oil droplets from entering the induction system. 
Small variations in the amount of blow-by gas passing the piston rings does not 
have an appreciable influence on the strength of the mixture fed to the engine 
since these gases are largely unburned mixture. Since the whole system is sealed, 
the mixture that leaks past the rings was metered by the carburetter in the first 
place. The only contaminant is the small percentage of partially burned mixture 
present in these gases. 

The work carried out by S.U. to produce the HIF design is typical of the 

74 The Sports Car 

ingenuity demonstrated by other carburettor designers throughout the industry. 
Unfortunately the emission levels demanded in the future will be even lower 
than those of today. I say, unfortunately, simply because I hate to see so much 
effort unrewarded. If the final result is a cleaner environment, the result will 
obviously be fortunate. 

Several sports car manufacturers have abandoned the use of carburettors and, 
despite the increased expense, they now look to fuel injection to give them the 
ultimate answer. 


As every motor racing enthusiast knows the battle between the carburettor and 
fuel injection was decidedly won by the latter about ten years ago. Even with a 
separate carburettor (or individual venturi) to each cylinder there is always a 
small drop in volumetric efficiency when a venturi is placed between the cylinder 
and the air intake. Moreover the carburettor has only a small pressure available 
(a fraction of one atmosphere) to break up the fuel into a finely divided mist. 
With fuel injection the pressure can be several atmospheres depending upon the 
design of the pump. 

For a sports car, especially a moderately priced production sports car, the 
need to extract the ultimate power from a given size of engine is less vital. There 
is really no need to provide a separate choke (venturi) to every cylinder. It is 
much less expensive to give the customer a 4-cylinder 3-litre engine with two 
twin-choke carburettors than a 2.2 litre engine with 8 cylinders and four 
twin-choke carburettors. The road performance of the two cars would hardly 
differ, even though the aficionado will always prefer the latter. 

Why then need we bother with the complexity and high cost of fuel injection? 
To find the answer we must travel, in mind at least, to that sprawling urban 
community called Los Angeles, the super city that gave us Sam Goldwin and 
Shirley Temple, stole the Brooklyn Dodgers for a handful of silver and finally 
paid the price of sin under the affliction of photo-chemical smog. They were 
indeed the first to experience this phenomenon since their peculiar geographical 
location and climate prevented the dispersal of exhaust fumes and encouraged 
the photo-chemical reactions. In the last twenty years we have seen the gradual 
spread of smog to other American cities, to Tokio and to other warm countries 
where the automobile density is high. 

The automobile industry began to work, half-heartedly at first, and more in 
earnest when prodded by the California State Legislature, on the problems 
associated with the reduction of hydrocarbon and carbon monoxide levels in the 
exhaust gases. Later, it was realised that certain nitrogen oxides, that are 
produced when nitrogen and oxygen are subjected to the high temperatures 
existing during combustion, actually encourage the smog reactions. Nitrogen 
oxides were then added to the list of undesirables. 

To improve combustion they experimented with many types of combustion 
chamber, notably the stratified charge-engine, but as work continued through 
the sixties there grew an increasing awareness that a reliable system of fuel 
injection was a necessary corollary to successful emission control. 

The Engine: Fuel Metering 75 

In 1957 the Bendix Corporation of America published an SAE paper on their 
Electrojector fuel injection system and this was the basis of the electronic 
metering system developed by Robert Bosch of Germany under the trade name 
‘Jetronic’ and first used on a production car, the VW 1600, in 1967 to meet the 
1968 U.S. Federal Law for emission control. Variations on the Bendix/Bosch 
system have since been produced under licence by AE-Brico and the Lucas 
Electrical Company in Great Britain. Figure 5.6 shows the basic metering system 
applied to a 4-cylinder engine. 

filter pump fuef injectors 

Fig. 5.6 Application of electronics to control fuel 

Port injection is used since the injectors are not then subjected to the high 
temperatures and carbonising problems associated with direct injection into the 
combustion chamber. Metering of the fuel quantity is controlled on a pulse-time 
basis. When more fuel is demanded the pulse is of longer duration. To give the 
reader an idea of the maximum time available for this pulse we can state that the 
time taken for all four strokes of one cylinder at 6,000 r.p.m. is only 20 

76 The Sports Car 

milliseconds. Experience has shown that the pulse need not be timed relative to 
any point in the induction stroke. In fact for convenience it is normal to operate 
injectors in two groups. In the case of a 6-cylinder engine all 3 injectors in the 
same group will operate simultaneously. In a V-12, all cylinders on one bank of 
six will receive a pulse into the valve port at the same time. Some inlet valves will 
be open at the time, others closed. In the latter case the fuel will pass into the 
cylinder several milliseconds later when the valve opens. No detectable differ¬ 
ences in mixture strength or combustion behaviour have been observed between 
cylinders with inlet valves open during injection and those with the valves 

The pulse-time system is well suited to control by mini-computer. The 

A fuel tank 
B fuel pump 
C fuel filter 

D fuel pressure regulator 
E cold start fuel injector 
F fuel injector 
G intake manifold 
H extra air valve 
I cylinder head 
J piston 

K battery 
L main relay 
M fuel pump relay 
N ignitiondistributor 
O throttle position switch 
P pressure sensor 
Q air temperature sensor 

R coolant temperature 

S thermo-time switch 
T electronic control unit 

Fig. 5.7 Lucas Electronic fuel injection 

The Engine: Fuel Metering 77 

important parameters of engine speed, manifold vacuum, induction air tem¬ 
perature and engine water temperature can be measured. This information can 
then be fed into the computer designed specifically for this purpose. After 
processing all the information this mini-computer then transmits the correct 
pulse-times to the solenoids that operate the injectors. Other data processed by 
the computer are signals for mixture enrichment when starting from cold and 
during warm-up, for full-throttle operation, for acceleration and to cut off fuel 
during overrun. 

Lucas electronic fuel injection 

The Lucas system now fitted to the Jaguar XJS is shown diagrammatically in 
Figure 5.7. Metering accuracy depends primarily on two elements, time and 
pressure , this being the pressure maintained in the supply lines to the injectors. 
The fine control of the pressure from the fuel pump (B in Figure 5.7) by means 
of the regulator D to the correct pressure of 28 lb.f./in 2 (2 kg.f./cm 2 ) is as vital 
as the correct functioning of the computer circuits. Excess fuel from the 
regulator is returned to the fuel tank. The fuel line is manifolded to each injector 
with an additional line feeding the cold-start injector on the inlet manifold. 

Fig. 5.8 Inside the Lucas fuel injection control unit. 

The heart of the control system is the computer. Figure 5.8 shows the 
computer opened up for inspection. The two major control parameters are the 
pressure in the induction manifold and the engine speed. For every new model 
of engine the mixture requirements must be determined by dynamometer tests 
and these result in a family of curves of the type shown in Figure 5.9. From this 

78 The Sports Car 

set of curves the analysing circuits of the control unit are tailored to trigger the 
output circuits to generate the pulse-times requested by the combined signals of 
manifold pressure (plotted here in mm Hg) and engine r.p.m. An inductive 
transducer operated by evacuated aneroids is used to produce the pressure signal 
and reed switches in the ignition distributor indicate engine speed and camshaft 
timing. The current pulses sent out from the control unit energise the solenoids 
in the appropriate group of injectors. When a solenoid is energised the magnetic 
field attracts the plunger and lifts the needle valve from its seat. For the duration 
of the pulse the pressurised fuel is injected into the inlet port as a finely atomised 

0 1000 2000 3000 4000 5000 6000 

Engine speed n (rev/min)- *■ 

Fig. 5.9 Engine fuel demands are established by 

dynamometer tests to produce a family of 
curves relating the precise injection period for 
every engine speed and manifold pressure. 

The basic family of curves in Figure 5.9 only apply to a specific operating 
temperature. Corrective signals are given to the computer circuits for variations 
in air intake temperature (Q in Figure 5.7) and for variations in coolant 
temperature (R in Figure 5.7). Mixture enrichment for cold starting is given by 
energising the cold-start injector, E. The thermo-time switch, S, cuts off this 

The Engine: Fuel Metering 79 

Fig. 5.10 Schematic diagram of the solenoid-operated injection valve. 

injector when the engine coolant temperature reaches a certain value. The 
correct mixture for smooth idling is obtained by the provision of an extra air 
valve, H, which permits a controlled amount of air to by-pass the throttle. 
Cylinder head coolant temperature controls the exact quantity of extra air. The 
addition of extra air with a cold engine raises the manifold pressure. This in turn 
acts on the induction pressure sensor, P, which automatically increases the 
duration of the injection pulse and enriches the mixture. Warm up of the engine 
reduces the quantity of extra air, reduces the degree of enrichment and thus 
maintains the correct mixture for good combustion. 

During overrun the fuel supply is cut off, re-commencing when the accele¬ 
rator is depressed or when the engine speed has fallen to a certain value. 

A throttle-position switch is required to indicate to the control unit that the 
throttle is almost closed (idle or overrun). This switch is designed to transmit 
three distinctive signals. Besides the closed position contacts there are other 
contacts to indicate the final stages of throttle opening. At a chosen point near 
full-throttle contacts are closed that signal a demand for mixture enrichment, a 
necessary precaution to prevent damage to exhaust valves and pistons by 
burning a lean mixture at full-throttle. At all other throttle openings between the 
idle position and this point the mixture is lean, economical and controlled to 
give low emissions of hydrocarbons, carbon monoxide and nitrogen oxides. The 
third signal given by the throttle switch device indicates acceleration by a wiping 
action over a set of comb-like strip contacts. These contacts generate additional 
injection pulses to provide acceleration enrichment. 

This then is the highly sophisticated Bosch-Lucas fuel injection system. Even 
this description is not complete since the needs of other devices to improve 
combustion may be controlled from the fuel injection computer. When exhaust 
gas recirculation is used to reduce nitrogen oxides signals from appropriate 
contacts are needed to cut off the exhaust gas recirculation system when it is not 

The Bosch K-Jetronic mechanical system 

The Robert Bosch Company have been making fuel injection equipment longer 
than we care to remember. The jerk type pump used to operate the fuel injection 

80 The Sports Car 

1 Air cleaner 5 Fuel flow control valve 

2 Metering plate 6 Fuel pressure control valve 

3 Throttle 7 Injector 

4 Induction pipe to cylinder 

Fig. 5.11 Diagram of Bosch ‘K-Jetronic’ fuel injection with mechanical 
control, as used by Porsche. 

on the famous Porsche 917 racing engine was a development of the Bosch Diesel 
injection system. The latest K-Jetronic system however bears little resemblance 
to the traditional mechanical system where individual reciprocating plunger 
pumps supply fuel at a timed injection point to separate injectors. In this new 
system the injectors supply a continuous spray of fuel into the inlet ports, the 
injection rate being based on a measured mass air-flow. This is the system that 
Porsche adopted in 1974 when they needed a new metering system to meet the 
American exhaust emission laws. Porsche are unsurpassed as mechanical 
engineers and the author sympathises with their apparent distrust of the 250 to 
300 components in the electronic control unit. Mechanical engineers have a 
natural predilection for mechanical gadgets, since they can understand how they 
work! Even so the K-Jetronic system is not exactly simple, but the Porsche 
organisation have a very well equipped research and development centre at 
Weissach near Stuttgart-Zuffenhausen and the development of a reliable 
application to all the new production Porsches, even to the Turbo model, will 
have been no great challenge to such dedicated professionals. 

Figure 5.11 is a simplified line diagram of the system. Air enters a plenum 
chamber (1) after filtration. The pressure of the air as it passes over and round 
the metering plate (2) is used to measure the mass air-flow into the engine. The 
metering plate is attached to a counterbalanced lever which moves the control 
valve (5) to regulate the fuel flow to the injectors (7). This is the basic system. As 
in the electronic system it requires auxiliary devices to cope with cold starting, 
warm-up, idle and overrun. 

The Engine: Fuel Metering 81 

All future sports cars will carry suitable fuel metering equipment and 
improved combustion techniques. It will be fascinating to see what will 
eventually emerge as the most acceptable solution from the plethora of ideas 
under investigation. Perhaps it will be an improved Stirling engine, or a more 
effective stratified charge-engine. Perhaps it will be an electric car and this 
chapter will then become redundant in our next edition! 

The engine: miscellaneous 

'Every instrument, tool, vessel, if it does 
that for which it is made, is well.' 



In the beginning the crankcase was simply a convenient enclosure to prevent the 
splash fed oil from escaping from the confines of the engine. Cylinder pressures 
were low and any crankcase with thick enough walls to produce a good casting 
was stiff enough for the purpose. Today with the increased stresses and higher 
rotational speeds it is rigidity we require before all else in a crankcase. A 
crankcase is never stressed near the limit in normal operation, since the 
deflections produced by such stresses would seriously overload the main 
bearings by the malalignment that would be produced in the crankshaft 

The crankcase in this respect is analogous to the chassis frame, for in both 
Rigidity is next to Godliness. Weaving and lozenging of a chassis frame can 
upset the most carefully planned suspension layouts and, with even more 
devastating results, can a well-conceived set of main bearings be reduced to 
molten metal by a crankcase lacking in stiffness. The crankcase in the modern 
engine is almost invariably combined with the cylinder block and the whole 
casting is a carrier for both reciprocating and rotating components. It is 
designed to withstand the explosive loads from the combustion processes, these 
acting not only outwards, but upwards on the cylinder head holding-studs and 
downwards on the main bearings, tending to pull apart the top and bottom parts 
of the crankcase. Other loads are imposed by the centrifugal forces acting on the 
main bearings and, of a lesser magnitude, the torque reaction which the 
crankcase transmits through the mountings to the frame. Very few of the early 
designs of petrol engine had crankcases that were rigid enough and the life of the 
main bearings occasionally suffered as a consequence. A common fault in 
4-cylinder engines with three main bearings was a movement of the centre main 

The Engine: Miscellaneous Components 83 

bearing under load. Eventually, wear in this centre bearing caused loss of oil 
pressure, followed by failure. The cause in part was a lack of stiffness in the 
crankshaft, but an equal contribution to the failure came from crankcase 
flexibility. Without rigidity between the front and rear main bearing housings 
the benefits of the centre main bearings are lost. The principle remains when 
more cylinders and more bearings are involved. The crankshaft is designed to 
resist both bending and twisting. With an absolutely rigid crankcase, bending of 
the crankshaft would only occur between adjacent bearings. Absolute rigidity is 
impossible, but the greater the deflection the greater is the span over which 
bending of the crankshaft takes place. 

In some modern engines the necessary rigidity is obtained by making each 
section of the crankcase between main bearings, in effect, a separate compart¬ 
ment with dividing walls between the bearing housings. The loads imposed on 
the main bearings are often transferred to the heavier portions of the crankcase 
through suitably placed webs. 

Engine designers have not always thought it necessary to place a main bearing 
between each cylinder. Torsional vibration can be reduced by the use of the 

Fig. 6.1 The 6-cylinder Jaguar crankcase. 

84 The Sports Car 

shorter crankshaft of the four-bearing 6-cylinder engine. The makers of the 
pre-war Riley 9 claimed that its two-bearing 4-cylinder engine had a much 
stronger crankshaft with this arrangement since the crankshaft was so much 
shorter and stiffer than with the three-bearing arrangement. With 100 horse¬ 
power per litre produced today from 4-cylinder sports car engines it would be 
difficult to design a two-bearing crankshaft of robust enough proportions to 
withstand the bending moments over such a wide span. 

A typical modern crankcase for a 6-cylinder engine is shown in Figure 6.1. 
With the provision of transverse walls at each main bearing housing, four sides 
of a box are established between bearings to create a very rigid structure. 

A system known as ‘cross-bolting’ is used on some American competition V-8 
engines. As engine speeds increased above 6,000 r.p.m. it was found that the 
horizontal vectors of the inertia loads could no longer be adequately contained 
by the bearing caps. A cross-bolted system is shown in Figure 6.2 this being a 

Fig. 6.2 Cross-bolting of main bearing caps to increase rigidity 
of crankcase on Hemi-Head Chrysler V-8. 

scrap section of the Chrysler Hemi-Head crankcase. Two horizontal bolts are 
fitted into tapped holes in the sides of each bearing cap and are tightened to a 
pre-determined torque. Horizontal inertia loads are thus distributed between the 
bearing cap and the crankcase walls. Cross-bolting is usually confined to the 
more highly loaded intermediate bearings. 


The design of the crankshaft for a modern high performance petrol engine is a 
challenge to the designer, especially if the number of cylinders is six or more. 
The success or failure of the whole engine may rest on the ability of the designer 
to achieve a satisfactory compromise between the conflicting requirements 
placed on this component. To illustrate this point let us consider the problem of 
the number of main bearings to be used. At first sight one is tempted to reject 
out of hand the four-bearing design for a 6-cylinder engine. The seven-bearing 

The Engine: Miscellaneous Components 85 

design will result in a longer engine, but this we might accept if satisfied that 
seven bearings give the better design. A longer crankshaft unfortunately has a 
lower natural frequency of torsional vibration and, if the crankshaft is too long, 
this natural frequency can become low enough to enter the range of frequencies 
of the explosive and inertia force cyclical variations. If resonance occurs between 
this natural frequency and the applied frequency of the combustion and inertia 
loads the torsional stresses in the crankshaft can build up beyond the safe limit. 
A good vibration damper would mitigate this, but the possibility of fatigue 
failure would remain. If we reduce the projected area of the main bearings in 
order to shorten the crankshaft, the bearing load factor (pressure times surface 
velocity) can become too high. We are therefore forced to use a much larger 
journal diameter on a seven-bearing crankshaft than on a four. This increases 
the general stiffness and raises the natural frequency of torsional vibration to a 
safe level. 

The short-stroke crankshaft is inherently stiffer than the long-stroke design. 
Figure 6.3 serves to illustrate why this is so. With a long-stroke design there is 

Fig. 6.3 Overlap of crankpin and main bearing journals 
on a short-stroke engine increases stiffness of 

very little overlap of crankpin and main bearing journal, when viewed from the 
end. The amount of overlap is dependent upon journal diameters as well as 
stroke dimensions. In older designs with small diameter journals there was 
usually no overlap at all. Flexing of the web under bending moments is 
obviously much greater when the overlap is small. The V-12 Jaguar engine has a 
relatively short stroke (70 mm against 106 mm in the 6-cylinder engine). This 
gives the new crankshaft even greater stiffness than that in the older engine. 

86 The Sports Car 


Bearing pressures 

When speaking of the pressures in a set of bearings with a forced lubrication 
system we must distinguish between the pressure supplied by the oil pump to 
force oil through the bearings and the self-generated pressures created by the 
rotation of the journals in their bearings. The pump pressure controls the flow 
through the bearings, a higher pressure giving a greater flow in gallons per hour. 
Bearing temperatures can be controlled in this way. The hydrodynamic pressure 
created by crankshaft rotation is the pressure that keeps the metal surfaces 
apart, even under the high peak loads imposed by the combined combustion and 
inertia forces. 

Two Englishmen gave us our first insight into the phenomenon of hydro- 
dynamic bearing pressure. Beachamp Tower in 1883 discovered the existence of 

Fig. 6.4 

Rotation of crankshaft journal inside 
its bearing creates a wedge action in 
the oil film to build up a hydrodynamic 
pressure. The peak value of this 
pressure is far greater than that 
created by the oil pump. The clearance 
is, of course, exaggerated in this 

a fluid film pressure in a bearing and three years later Osborne Reynolds made a 
brilliant mathematical analysis of the forces that produce it. In simple terms we 
can say that the rotating action of the journal makes it climb slightly (within the 
confines of the existing bearing clearance) up the side of the bearing (see Figure 
6.4). The oil film is pulled round in the direction of rotation by the viscous drag 
at the surface of the journal and a pressure build-up occurs in the wedge-like 
part of the oil film where the minimum clearance exists. The pressure gradient 
created in this way depends upon the speed of rotation and the viscosity of the 
oil. It is at its highest value just before the point of minimum clearance and 
drops away rapidly after this, even becoming slightly negative at one point. A 
typical pressure profile is shown in Figure 6.5. The negative pressure area is an 
obvious point on the bearing for the introduction of the pressure oil feed. Since 
maximum pressures can be as high as 5,000 lb per sq in. and pump delivery 
pressures are in the range 80-120 lb per sq in. we could hardly hope to maintain 
an oil flow through the bearing with an oil feed at the bottom. 

The simple case of Figure 6.5 assumes a steady vertical load. In practice the 
load is fluctuating and changing in direction with the angular changes of the 
connecting-rod. As the load direction changes, so does the maximum pressure 
point move and with an adequate bearing area, correct clearances and the right 

The Engine: Miscellaneous Components 87 

Fig. 6.5 

Typical pressure profile near 
loaded area produced by 
wedging action. Note slight 
negative region. 

lubricant, the wedge action of the rotating film prevents metal-to-metal contact 
at all times, with the single exception of the initial slow-speed rotation when the 
engine first turns under the action of the starter. At this time we depend upon 
‘boundary lubrication’ to save the bearing from damage. This type of lubrica¬ 
tion will be discussed later. 

Bearing failure will occur when: 

(a) bearing clearance becomes so great that too much pressure is lost by 
leakage at the sides of the bearing. 

(b) the oil viscosity becomes so low, from overheating, contamination or 
other means, that the bearing pressure falls below the critical value 
required to offset the load 

(c) the oil flow through the bearing becomes so much reduced by blocked oil 
feed lines, or similar obstruction, that the surface temperature becomes 
too high for the particular bearing alloy. In such cases failure is by 
melting of the bearing metal, which in turn increases the clearance and 
leads to case (a). 

It is well known that the load that can safely be carried by a bearing becomes 
less as the speed increases. The heat generated by a bearing is a function of both 
pressure and speed. The running temperature is controlled by a balance between 
the rate at which heat is generated and the rate at which it is removed by a 
combination of air cooling and oil cooling. The more heat we can remove by the 
flow of lubricant, the higher will be the load the bearing can stand without 
failure. This is the reason for the use of oil-coolers on competition engines, since 
a cool oil will remove more heat than a hot oil. Moreover a cool oil will have a 
higher viscosity, leading to higher hydrodynamic pressures. 

The expression one sometimes finds in engineering textbooks: 

pressure x rubbing velocity = a constant 

cannot be of universal application, since the degree of oil cooling influences the 
value of the constant. An important implication of this relationship is that an 
increase in bearing diameter will not increase the load-carrying capacity of a 
bearing. The pressure is reduced, since the projected area of the bearing is 

88 The Sports Car 

increased, but the rubbing velocity is increased in the same ratio and the load 
factor remains unchanged. We therefore see that the critical dimension is the 
width of the bearing. Since the available space between cylinders is limited, there 
is a limit on all engines to the width of the main bearings. Any increase in main 
bearing width entails a decrease either in big-end bearing width or in crank web 
thickness. Thus, unless we are prepared to increase the spacing of the cylinders 
and to increase the overall length of the engine, there is a fairly close limit set to 
the width of the bearings that can be used in an in-line engine. On a V-6, V-8 or 
V-12 the available space is much less, but the main bearing loads are substan¬ 
tially reduced since the thrusts imparted by paired cylinders partly cancel one 
another out. 

When we speak of the hydrodynamic pressure exerted by the oil film of a 
bearing we must remember that this pressure is not constant across the width of 
the bearing. It is at its highest at the centre and falls to a negligible pressure at 
the outside edge. This is illustrated in Figure 6.6 for the case of a typical bearing 

Fig. 6.6 Improved pressure distribution of oil film 
when oil groove is eliminated. 

with an oil groove around the centre and a plain bearing with no oil groove. The 
traditional oil groove, originally intended to distribute oil around the bearing, is 
now condemned by most authorities since it breaks the pressure profile into two 
peaks as shown in the diagram. When the Chevrolet Company were developing 
their 265 cu in engine they also had doubts about the efficacy of the oil groove, 
so they took the trouble to measure oil film pressures in all five main bearings at 
different loads and speeds. 

The reduction in mean pressures in the grooveless bearing was approximately 
equal to the increase in bearing area that this modification provided. Endurance 
testing showed that the grooveless bearings were superior in service and, at the 
same operating clearances, would carry almost twice the load without failure. 
As a result of these experiments Chevrolet adopted a design in which the oil 

The Engine: Miscellaneous Components 89 

groove was omitted from the lower half. It was retained in the more lightly 
loaded upper half. 

Bearing metals 

The properties we value in a bearing metal are as follows:- 

(a) High mechanical strength to resist the maximum loads imposed in service. 

(b) High melting point to resist damage by high temperature lubricants and to 
resist bearing failure. 

(c) High resistance to corrosion from degraded acidic lubricants. 

(d) Good ‘embeddability’ to absorb particles of dirt smaller than the bearing 
clearance (larger particles should have been trapped by the oil filter). 

(e) Good comformability to yield slightly when the mating shaft is misaligned. 

(f) Sufficient hardness to resist abrasive wear and erosion. 

(g) Good compatibility with the shaft material to prevent seizure when oil has 
drained away after prolonged standing and before rotational speeds are 
high enough to provide hydrodynamic pressure. 

This is a formidable list and it is difficult to find a bearing material that scores 
highly on all of them. The type of bearing alloy that has been demonstrated over 
the years to possess the apparently conflicting requirements of ‘conformability’ 
and ‘hardness’ consists of a relatively soft matrix in which are embedded 
crystals of a harder metal. During the running-in period the soft matrix of a 
good bearing alloy flows slightly under pressure leaving the harder crystals to 
act as load bearing members. The depression left in the matrix acts as an oil 
reservoir to help establish an oil film over the whole surface during the critical 
period when starting the engine. 

White metal bearings 

The original white metal, introduced about 1860 by Isaac Babbitt, is thought to 
have contained about 85 per cent tin, 10 per cent antimony and 5 per cent 
copper. How right he was in choosing this alloy has been proved over and over 
again and the composition of the white metals used for the first fifty years, of 
this century differed only slightly from his original specification. 

The older brasses of the vintage era have given place to the modern 
steel-backed bearing in which a high degree of strength and stiffness is obtained 
from a thin steel shell covered with a film of bearing metal. The shell is usually 
about 0.9 mm (0.036 in) thick with a bearing metal lining of about 0.4 mm 
(0.015 in) thick. The bearings are stamped from large bi-metal strips, pressed to 
shape and indented at one corner to provide a location against rotation. They 
are cheap and reliable. Quality control was never very high with the old vintage 
hand-poured, cast, white-metal bearings. 

Metallurgists have developed several improved bearing alloys to withstand the 
increased bearings loads that have resulted from the higher compression ratios 
and higher operating speeds used on modern engines. 

90 The Sports Car 

Copper-lead and lead-bronze bearings 

The introduction of the high-speed Diesel engine for heavy transport in the early 
thirties, where compression ratios as high as 18 to 1 were used, brought in its 
wake a crop of bearing troubles, for the load carrying capacity of the 
white-metal bearings was sometimes exceeded. The answer was found in 
mixtures of approximately two-thirds copper and one-third lead and similar 
mixtures of lead copper and tin, usually called ‘lead-bronzes.’ These copper- 
lead and lead-bronze bearing metals are not alloys but mechanical mixtures of 
the powdered metals and great care is required in the casting operation to 
prevent segregation of the components. 

The one advantage over white metal is the higher limiting load factor. Two 
disadvantages are that they tend to wear the crankshaft more readily and they 
require higher running clearances. The former can be countered successfully by 
the provision of hardened journals and crankpins; the latter calls for a higher oil 
pressure and a larger capacity oil pump to compensate for the increased oil flow 
from the sides of the bearings. A minor disadvantage of these bearings are their 
poor resistance to corrosion by the acidic oxidation products in the lubricating 
oil. Frequent oil changes are beneficial in reducing this corrosion, but an answer 
was found in a protective layer of tin of indium. This coating is only about 
0.025 mm (0.001 in) thick and without regular attention to the air filter and 
regular changes of the oil filter the accumulation of road dust and combustion 
contaminants in the oil will gradually wear away this protective skin. 

Aluminium-tin alloy bearings 

Aluminium alloy crankshaft bearings are not new. High Duty Alloys of Slough 
developed good bearings of this type to carry higher loads than white metal 
about forty years ago. The new aluminium-tin alloys are far superior in 
load-carrying capacity than these earlier alloys, this being almost as great as the 
lead-bronzes. The tin content can be as little as 6 per cent or as high as 20 per 
cent. The attraction of these new alloys lies in their superior resistance to 

Aluminium-silicon alloy bearings 

A new bearing alloy has been developed to cope with the very high loads 
imposed on connecting rod bearings in turbocharged Diesel engines. This new 
aluminium alloy contains 11 per cent silicon and 1 per cent copper. The alloy has 
excellent strength and its chances of survival under severe conditions are 
exceptionally high. It requires no protective overlay against corrosion and has 
been shown to have a high resistance to seizure when hydrodynamic lubrication 
is almost non-existant. For all very high performance engines, in particular for 
turbocharged competition engines, this new bearing alloy holds great promise. 

The Engine: Miscellaneous Components 91 


The majority of motorists have at least been educated to change oil filter 
elements at the recommended mileages. It is not always realized that contamina¬ 
tion of the lubricating oil can come as much from the air entering the cylinders 
as from the burning of oil and fuel in the combustion chambers. Neglect of air 
filtration admits abrasive dust particles which are deposited by the highly 
turbulent conditions in the cylinders on to the cylinder walls. The oil film on the 
walls is being continually renewed from below and being scraped off again into 
the sump by the action of the rings. We must therefore regard the air filter as 
our primary oil filter — the first line of defence against contamination. 

The importance of adequate flow as well as adequate pressure is now 
appreciated by engine designers. The oil pumped through the bearings serves to 
cool them, the higher the flow rate the lower the operating temperature. With 
true film lubrication, as distinct from boundary lubrication, the surfaces of the 
bearing and journal are held apart by a film of oil. Since the two metal surfaces 
have a high relative speed it is obvious that adjacent layers of this oil film are 
suffering a continuous shearing action. This shearing action, usually called 
viscous drag, is manifested in the form of heat. Given good filtration, there is 
nothing to be lost and everything to be gained if the output of the pump is made 
as large as possible. 

The lubrication circuit for the V-12 Jaguar engine is shown diagrammatically 
in Figure 6.7. The oil pump is exceptionally large, but at low speeds the entire 

Fig. 6.7 Layout of Jaguar V-12 oil cooler circuit. 

92 The Sports Car 

output passes through the filter to the main oil gallery. As the speed increases 
the relief valve opens to by-pass surplus oil through the cooler. This cooled oil is 
returned to the pump inlet. At about 2500 r.p.m., as much oil is by-passed 
through the cooler as the amount delivered to the main gallery and at higher 
engine speeds the flow through the cooler is appreciably greater. The oil cooler 
is an aluminium oil-to-water heat exchanger bolted to the front underside of the 
sump (oilpan). From the base of the radiator cooled water passes through the 
water passages in the cooler before entering the water pump. At speeds of 130 
m.p.h. a temperature reduction of 23°C (41 °F) has been recorded with the 
engine fitted to the XJ-S sports car. Without the oil cooler the temperature of 
the oil delivered to the bearings would be about 140°C (284°F) which is a little 
higher than that recommended as an upper limit by the bearing manufacturers. 
Bearing failure can in general be anticipated when the temperature of the oil 
film at the bearing surface reaches about 170°C (338°F) and the temperature 
measured at the bearing is usually about 30°C (86°F) higher than that measured 
in the main gallery. 

The complexity of the lubrication system in a modern engine is shown in 
Figure 6.8. From the main gallery oil is fed at full pressure via the six 

Oil spill lubricates 
tappets and valve gear 

Oil delivery to 

Drilling to 
from front 


Crescent type 
oil pump 

Main sump 
suction pi\' 

Oil pick-uj 


Oil flow 

to sump 

Main oil 

Cross drillings to main 
bearings t crank pins 

Surge baffle plates 
Sump strainer 
Oil way to gudgeon pins 
Oil cooler 
^Oil cooler water inlet 

relief valve 

from pump 

Fig. 6.8 Diagram of oil system in Jaguar V-12 engine. 

The Engine: Miscellaneous Components 93 

cross-drillings to the main bearings. From the main bearings the oil enters the 
drillings in the crankshaft to provide a copious flow to each connecting rod 
bearing. The connecting rods are drilled to convey oil to each gudgeon pin (wrist 
pin) and a metered supply of oil is piped to the camshaft bearings. In general, it 
can be said that all rotating, sliding or rubbing parts in an engine must be 
lubricated if they are to survive. Small lightly loaded components are occasion¬ 
ally made from graphite-impregnated porous metals or from low-friction 
plastics, such as PTFE, but a continuous supply of lubricant is required when 
loads are high. Pistons have always been lubricated by splash from the oil that 
squirts from the sides of the connecting rod bearings and the sides of the main 
bearings. This does appear to be a hit-and-miss method since an increase in 
bearing side clearances can overtax the ability of the oil control rings on the 
pistons to prevent excessive oil consumption. Despite this lack of engineering 
finesse the designers of pistons and rings have achieved miracles of oil control in 
modern engines. 

Engine oils 

It is self-evident that a relatively high viscosity is one of the essential properties 
of an automobile lubricating oil. Not all viscous liquids are lubricants, however, 
since they do not all possess the property of ‘oiliness*. The peculiar property of 
oiliness is associated with the way the molecules of a lubricant in contact with a 
metal surface form an intimate bond with the metal molecules. This ‘boundary 
layer* is believed to be no more than one molecule thick. It is this surface effect 
that is so vital when an engine is being started, since the hydrodynamic pressure 
to hold the surfaces apart is only produced when the journals are rotating at a 
fair speed. Boundary lubrication can also postpone bearing failure for a short 
time after the flow of lubricant has been cut off. It can only be a postponement 
at best, since the heat cannot be easily dissipated when the flow of lubricant has 
stopped. This property of boundary lubrication is exhibited by all oils, but is 
much stronger in some than in others. Castor oil and rape oil are two 
outstanding examples. The strength of the molecular bond established by these 
two oils is so great that it resists repeated washings in petrol and other solvents 
and can only be completely removed by scraping or grinding the metal surfaces. 
It is this property that makes rape oil such a valuable constituent of certain 
proprietory upper-cylinder lubricants. In certain circumstances, notably when 
starting from cold, the piston rings are starved of oil and only boundary 
lubrication prevents partial seizure of the rings and the cylinder walls. A 
lubricant such as rape oil or colloidal graphite serves to strengthen the boundary 
lubrication. Special lubricants containing molybdenum disulphide are now on 
the market. Rather extravagant claims are made in some of the advertisements, 
but tests have shown that this substance does in fact help to reduce scuffing of 
tappets and cams and such components as may be imperfectly lubricated. Some 
proprietary engine oils contain very small amounts of metallic compounds that 
are claimed to reduce wear by improving boundary lubrication. Castrol, for 
example, add an oil-soluble long-chain compound with the popularised name of 

94 The Sports Car 

‘liquid tungsten’. Such additives do not work miracles but there is good evidence 
that they do reduce cylinder wear during the crucial period after a cold-weather 
start when oil is washed from the cylinder walls by liquid petrol. 

Another property demanded from an engine lubricant is a high ‘viscosity 
index’. The older types of lubricating oil can be blended to have a high viscosity 
at low temperatures, but as the oil temperature rises to its working temperature, 
the viscosity usually falls to a very low value. This was obviously a bad feature 
and the oil technicians have now developed oils with much higher viscosity 
indices. This simply means that the viscosity does not fall so rapidly with rise of 
temperature. The scale of viscosity index is purely arbitrary; asphaltic Gulf oil 
being taken as zero and the excellent Pennsylvanian oil as 100. The so-called 
multi-grade oils are simply oils with a high v.i. The S.A.E. specifications for 
automobile engine oils only call for an oil of medium v.i. and by blending 
special oils or by using special additives called v.i. improvers a normal oil can be 
improved in v.i. until it has the viscosity of an S.A.E. 10W Grade oil at 0°F, 
that of a 20W Grade oil at 100°F, that of a 30 Grade oil at 210°F and that of a 
40 Grade oil at 300°F. Such an oil possesses the low oil drag of a winter grade oil 
to assist winter starting, at the same time it maintains a high enough viscosity to 
prevent loss of oil pressure under arduous hard-driven conditions. 

Resistance to contamination 

In general two kinds of contamination can occur in an engine oil that arise from 
chemical breakdown of the oil itself. Slow oxidation of the oil and the 
emulsification of atmospheric condensate can form thick acid sludges. The 
older motorist will remember how the walls of the crankcase, even the insides of 
the pistons, became plastered with the stuff. The second kind of contaminant 
comes from the burning of oil in the combustion zone, on the piston rings, 
lands, etc., to form carbon. There is also a third contaminant which is formed 
by chemical breakdown, but is different in composition and appearance. This is 
the brown lacquer which is formed on the skirt of the piston. Sludge formation 
is now combated and almost completely prevented by the addition of certain 
chemicals which have been proved by full-scale engine tests to discourage 
oxidation of oil. Improved ventilation of the crankcase has also reduced the 
tendency for water vapour to condense on the crankcase walls. The chief danger 
of sludge formation, apart from the gradual reduction of the general oiliness of 
the lubricant, is the possibility of oilways becoming blocked or seriously 
restricted. The sludge is also difficult to disperse and in sufficient quantities can 
choke the surface of an oil filter. 

The carbon deposits in the combustion chamber and on the piston crown are 
partly formed from lubricant which splashes past the top ring of the piston or is 
drawn past the inlet valve guide during the induction process. A small amount 
of this carbon comes from the incomplete combustion of petrol. The deposits 
also contain small amounts of other elements, such as silica and calcium coming 
from road dust which has passed through the air filter. The more oil passing the 
top ring, the softer will the deposit be and, what is more important, the faster 

The Engine: Miscellaneous Components 95 

will the deposit grow. Carbon deposit is known to encourage knock. Calculations 
show that the effect cannot be fully explained in terms of the increase in 
combustion temperatures caused by the heat insulating properties of the layer. 
There is strong evidence that the carbon layer takes part in the combustion 
processes in a catalytic manner. One theory, advanced by a Canadian research 
worker, is that the carbon particles picked up by the turbulent flaming gases act 
as nuclei for the detonation of the end-gas. One of the petroleum companies has 
followed up this work by developing an additive which it is claimed ‘fire-proofs’ 
the carbon. Another harmful effect of carbon is the manner in which it can 
accumulate on the compression rings, in the grooves and on the lands, especially 
at the top ring, until at least one of the rings becomes stuck in its groove and 
inoperative. The majority of the oils now sold are strongly detergent, i.e. they 
wash the carbon from the pistons and rings as fast as it is formed, carrying it 
round and round the oil system in the form of extremely fine dispersed particles, 
so fine that they can do no harm to the bearings. 


The automobile piston, by a slow process of development, seems to have settled 
for two general types, the solid skirt and the split skirt, the solid skirt for the 
more highly rated engine where strength at operating temperature is at a 
premium, the split skirt for the less highly tuned engine in which such 
refinements as freedom from piston-slap are more important. 

The top ring has always been the Achilles’ heel of the automobile piston. The 
temperature at the centre of the piston crown under full load is about 250°C 
(480°F). The temperature of the top ring is not much less than this, even though 
about 85 per cent of all the heat passing into the crown of the piston is 
conducted to the cylinder walls through this single ring. Anything we can do to 
reduce the heat load of the top ring is good. Most piston designers provide a 
generous path from the crown to the lower rings in an attempt to divert heat 
from the top ring. One design of Diesel engine piston that the author calls to 
mind has a cast-iron insert in the form of a band around the top part of the 
piston. In this band is machined the top ring groove. The primary purpose of 
this insert is to reduce wear of the top and second-ring lands. The author 
believes an excellent design of piston for a sports car engine could be based on 
this principle. The poor conductivity of the cast iron and the temperature drop 
at the interface between the two metals would reduce the amount of heat 
flowing to the top ring, leaving a greater amount to pass out through the other 

It is customary to provide some form of heat barrier between the hot ring belt 
and the skirt. This permits a reduction in working clearance of the skirt with less 
likelihood of piston slap when cold. The heat barrier usually takes the form of slots 
or rows of holes between the two parts of the piston (see designs (a) and ( b ) in Figure 
6.9). Sometimes the heat barrier is as complete as the slots shown in design (c). 

96 The Sports Car 

Secre t 5-5 Seciiw 3-5 5*ciiaw 3 r $ 

0 b f 

Fig. 6.9 Three forms of support for the gudgeon pin bosses. Note 
thermal barrier slots on Type C. 

Oil control 

We once thought that the manner in which a set of rings controls the passage of 
oil to the combustion chamber was by the cumulative action of successive piston 
rings scraping a little more oil off the cylinder walls until only a minute trace 
remained to lubricate the top ring. This small quantity escaping past the top 
ring was assumed to be lost and to represent the actual quantity burned in the 
combustion chamber. An experimental study carried out by Dr. P. de K. Dykes 
of Cambridge University completely upset this simple picture of ring behaviour 
and presented us with an entirely new concept. Dr Dykes showed that oil does in 
fact pass the rings in both directions and in large quantities. With a normal 
engine in good condition almost all the oil passing the rings on the downstroke is 
returned to the sump on the upstroke. Even if oil is added to the cylinder walls 
from above the rings, a good ring set will pass almost all the oil downwards to 
the sump, leaving only a negligible portion behind to be burned. Dr Dykes 
introduced two new words into the automobile engineer’s vocabulary. A 
‘down-passing’ ring or ring set is defined as ‘one which when supplied with an 
ample amount of oil from both above and below, passes oil more rapidly 
downwards than upwards’. Conversely an ‘up-passing’ ring or ring set passes oil 
more rapidly towards the combustion chamber. The best ring sets are down¬ 
passing as a whole. If the lower rings are down-passing and the upper are not the 
oil consumption can still be high if the amount of oil passing the lower rings on 
the down-stroke is large. On the other hand, if the lower rings are up-passing 
and the upper rings are down-passing, a high oil consumption is inevitable since 
the up-passing lower rings cause a pressure to be built up between the rings 
which forces oil past the upper rings. When the whole ring set is down-passing 
the rate of oil consumption can, in some cases, be absolutely zero. This perfect 

The Engine: Miscellaneous Components 97 

condition will occur when all oil drops thrown into the combustion chamber can 
drain back to the cylinder walls. In general a few drops of oil get held back and 
are burned on the roof of the chamber or on the piston crown. Others are swept 
out with the exhaust gases. 

Piston ring flutter 

At high engine speeds a phenomenon sometimes occurs which is known as 
‘piston ring flutter*. It usually affects the top ring and in exceptional cases the 
second ring too. As long as a compression ring remains pressed against the lower 
face of its groove during the compression stroke, the pressure behind the ring is 
sensibly that in the combustion chamber. Near t.d.c., however, as the piston is 
decelerating, the inertia of the ring can make it move upwards against the upper 
face of the groove. When this happens, the ring acts as a valve, sealing the ring 
groove space against the cylinder pressure. The gas in the space behind the ring 
leaks away to the crankcase (or to the lower rings) and the ring collapses inwards 
under the action of cylinder pressure. Since this collapse occurs once every two 
revolutions of the engine with a recovery to normal diameter occurring when the 
cylinder pressure drops, the ring in effect ‘flutters’ at high speed and if the 
condition is allowed to persist, ring breakage is inevitable. 

Ring inertia can be decreased by reducing the ring width. Using the 
recommended limits supplied by Messrs Hepworth and Grandage, the British 
piston manufactureres, we have calculated that a typical engine with a stroke of 
80 mm would be limited to a maximum engine speed of 6,500 with a top ring 
width of 1.6 mm (1/16 in). 

If then we provide this engine with valves and porting to run up to 7,000 or 
8,000 r.p.m. we would have to use extremely narrow and fragile top rings. A 
more satisfactory engineering solution would be the use of a Dykes pattern of 
ring, an L-shaped ring in cross-section that is specially designed to maintain 
communication between the combustion chamber and the back of the ring at all 

Fig. 6.10 

Cross-section of Dykes ring for high-speed 

engine speeds. Figure 6.10 shows a cross-section through one type of Dykes 
ring. The side clearance at A is greater than at C. When the ring moves upwards 
under the action of the inertia force the gap C closes, but the gap A does not and 
the space B behind the upper part of the ring remains in communication with the 
combustion chamber. This ring then is free to move upwards under the action 

98 The Sports Car 

of maximum inertia forces, but the cylinder pressure is never cut off from the 
back face of the ring. The upper portion is usually made about twice as wide as 
the lower. 

Chromium-plated top rings are in common use today. The bonded layer of 
chromium is only about 0.1 mm (0.004 in) thick, but its high resistance to 
corrosion and wear has been found to be of great value in this fiery zone where 
even boundary lubrication is difficult to maintain. On modern top rings both 
top and bottom edges are profiled as shown in Figure 6.11. This helps the ring to 

Theoretical top ring profile 

Minimum angles: 

a — 8milli-radians 
(approx 30') 

b — 4miJli-radians 
(approx 15') 

Average worn top ring profile 

Angles on used rings; 

Dimension c -0 008/0 015mm 

approx angles 25'/50 / 

Dimension d “0 0005/0004mm 

approx angles 10/30' 

Fig. 6.11 Profile used for top ring to improve lubrication. 

ride like a surf-board over the extremely thin oil film that exists in this zone 
without destroying it. When wear does eventually occur, the ring form is still 


The amount of heat to be dissipated from a typical sports car engine cooling 
system is roughly equal to the heat equivalent of the power developed by the 
engine. As thermal efficiencies increase with the use of higher compression 
ratios the proportion of the total heat energy of the fuel passing out into the 

The Engine: Miscellaneous Components 99 

exhaust and cooling systems is correspondingly reduced. If we increase the 
power output by supercharging, however, we also increase the amount of waste 
heat to be dissipated. An increase of 20 per cent in b.h.p. by supercharging 
increases the heat given out to the water jackets by about 25 per cent, since a 
portion of the extra power developed in the combustion chamber is lost in 
driving the supercharger. 


If we are fortunate enough to be able to design an engine without prejudice or 
the dictates of a firm’s established precedent we cannot avoid making a 
fundamental decision at the very beginning. Are we to cool the engine by air or 
water? The majority of experienced designers made up their minds on this issue 
a long time ago and the majority chose water. Yet Ferdinand Porsche with a 
wide experience of both types chose air-cooling when he began work on his 
ubiquitous People’s Car. Today VW are making cars with water-cooled engines 
and Porsche recently introduced a new model with water-cooling, yet continue 
to air-cool their most expensive products. Obviously the decision is not an easy 

In one sense we can say that all engines are cooled by air. What we call a 
water-cooled engine uses water as an intermediate fluid to pick up heat from the 
engine and transfer it to the air stream at the radiator. Air is a much less 
effective cooling medium than water. The specific heat of air is only one-quarter 
that of water and if we compare the two fluids on a volume basis it requires 
about 4,000 times the quantity of air to transfer an equal quantity of heat. Air 
being a less effective heat transfer medium it is necessary to increase the cooling 
surface on the outside of the cylinder by the provision of a multiplicity of deep 
cooling fins. There is a limit to the number of fins we can provide, since a 
spacing that is closer than about one-tenth of an inch brings the laminar layers 
on the fin surfaces so close that the efficiency begins to fall rapidly. There is also 
a limit to the depth of fin that can be effectively utilized. Thus we find that with 
the best system of cooling fins that modern manufacturing techniques can 
provide a greater temperature differential is still required to transmit the waste 
heat from the combustion chamber to the outside cooling medium when air is 
that medium. Air-cooled engines must always therefore run with higher cylinder 
head temperatures than water-cooled engines at the same specific power output. 
In the limit then, we cannot anticipate as high a specific power output from the 
air-cooled engine. 

There is little doubt that the well-designed air-cooled engine can always be 
lighter than the water-cooled engine. The air-cooled engine requires a ducted fan 
and a system of ducts and baffles to guide the air around the heads and cylinder 
barrels, but these can be made of light-gauge metal. The finning on the cylinder 
heads and barrels adds substantially to the weight of the air-cooled engine, 
especially if the barrels are made of steel, but the weight of the radiator still tips 
the scales against the water-cooled engine. 

We can briefly summarize the comparison in the table below: 

100 The Sports Car 

Advantages of air-cooling 

(1) Lightness. 

(2) Compact dimensions. 

(3) No water leaks, external or internal. 

(4) No danger of frost damage. 

(5) Quick warm-up. 

Disadvantages of air-cooling 

(1) More expensive construction. 

(2) Higher operating temperatures. 

(3) Slightly lower specific power output. 

(4) More noise from combustion, pistons and fan. 

Water cooling 

The radiator 

The resistance to the flow of heat from the water inside the radiator to the air 
outside can be considered largely in terms of two additive resistances, the 
water-film resistance and the air-film resistance. The relative sizes of these vary, 
but in general the air-film resistance is from four to six times that of the 
water-film resistance. The modern cellular radiator matrix is designed with this 
limitation in mind and the area of metal exposed to the air stream is much 
greater than that exposed to the water. This gives some measure of compensa¬ 
tion for the disparity in film resistances. The rate of the dissipation to the air 
stream varies 

directly as 

(a) the effective area of the cooling surface, 

(b) the mean temperature difference (strictly, the logarithmic mean) between 
the metal surface and the air stream, 

(c) some power (approximately the 0.6 power) of the velocity, 

and inversely as 

(d) some power (approximately the 0.4 power) of the width of the air 
passages through the matrix. 

For a radiator of small frontal area, therefore, the air passages through the 
matrix must be as small as possible. One limit to this is set by the danger of 
gradual blockage by insects and other debris. Another limit is set by the 
permissible air pressure drop across the block. If the resistance to flow through 
the matrix becomes too great the velocity through it will be seriously reduced. In 
practice the water walls of the matrix are seldom spaced any closer than 0.4 in. 
and the extended surface fins on the air side are spaced no closer than 0.2 in. 
The air velocity through the matrix depends, not only upon the speed of the car 
through the air but upon the shape and position of the entry duct or radiator 
cowl. This velocity is usually about 70 per cent of the external relative velocity. 
Thus if the road velocity of the car is 130 ft per sec and there is a following wind 
of 30 ft per sec the velocity through the radiator will only be 70 ft per sec. 

The Engine: Miscellaneous Components 101 

At 100 m.p.h. the cooling-water horsepower (heat equivalent) to be dissipated 
will be more than four times that at 50 m.p.h. The cooling air velocity is only 
twice that at 50 m.p.h. (assuming zero wind velocity) and the increase in heat 
transfer coefficient, being approximately proportional to V °' 6 (where V is the 
air velocity through the matrix) is only about 50 per cent. This explains why the 
radiator temperature rises when we drive faster. This however is not the most 
exacting duty imposed on a radiator, since full power can be used in an 
intermediate gear when climbing a steep gradient. On a long alpine pass with 
hairpin bends, full power may be used for 80 per cent of the time at low cooling 
air velocities. The short distance between bends prevents the car attaining its 
ultimate speed for the particular gradient and with full power in use almost 
continuously at average speeds of only about 50 ft per sec a radiator designed 
for high-speed work only would prove inadequate. 

The power absorbed by the fan at high engine speeds can be as much as 6 or 7 
per cent of the gross power. This is a sheer waste of power and money since at 
high speeds there is no need to boost the air flow through the radiator. Many fan 
control systems have been tried and a popular design is driven by an electric 
motor which is only in action when the coolant temperature rises to a 
pre-determined level. The Datsun sports car is fitted with a viscous fan coupling 
which always slips at high speed, but has a variable point of engagement of the 
coupling depending upon the temperature of a thermostat built into the centre 
of the fan coupling (see Figure 6.12). Thus in hot weather the fan, which is 
driven from the front end of the water pump, operates at water pump speed up 
to about 2,500 r.p.m. At higher engine speeds the viscous coupling slips to hold 
the fan speed to this figure. In cold weather the coupling begins to slip at a much 
lower speed, down to a lower figure of about 1,600 r.p.m. The coiled bi-metal 
thermostat seen on the front face of the fan boss in Figure 6.12 is subjected to 
the temperature of the air leaving the rear of the radiator. As the coolant 
temperature rises and the air leaving the back of the radiator rises the bi-metal 
strip expands and opens the valve which admits silicon oil to the chamber that 
operates the clutch and puts the fan into operation. The Datsun system 
therefore not only saves power but it ensures rapid warm-up of the engine. 

For the designer who is seeking to save weight and drag, two well tried 
methods are available, pressurizing and glycol cooling. Pressurizing the system 
to 0.82 bar (12 lbf/sq in) raises the boiling point to 117.5°C (243°F). This in 
turn raises the temperature difference between the water and the cooling air by 
17.5°C (31.5°F). At the bottom of the radiator, this may even double the 
temperature difference. A saving of as much as 35 per cent in matrix weight has 
been achieved in this way. Ethylene Glycol, with a b.p. of 180°C (356°F) points 
the way to even greater reductions. To the author’s knowledge, no designer of a 
sports/racing car has been bold enough yet to use 100 per cent ethylene glycol as 
a coolant. Apart from obvious disadvantages such as expense and the corrosive 
properties of the material (despite all efforts to find the perfect inhibitor) one 
disadvantage of using a higher exit radiator temperature would be the general 
increase in temperatures throughout the system. Water-jacket temperatures 

102 The Sports Car 

Fig. 6.12 Temperature controlled fan coupling used on Datsun 260Z sports car. 

The Engine: Miscellaneous Components 103 

would also be much higher and the temperatures of the cylinder walls and rings 
would be higher. This is no light matter and on a high-output engine could mean 
the difference between trouble-free piston operation and piston failure. 

The water pump 

Without a pump the water velocity through the radiator is slow, depending as it 
does upon convection currents rising from the hotter parts of the cylinder head. 
The greatest objection to thermo-siphon cooling, however, is the erratic manner 
in which circulation is maintained. In certain hot-spots, such as exist around the 
exhaust valves, circulation is intermittent; vapour bubbles form at these points, 
momentarily blanketing the surface against effective heat transfer. When the 
bubbles break away from the surface and rise, relatively cool water is drawn in 
from the cylinder jackets to replace them. Thick cylinder head walls in these 
zones will assist in absorbing temperature fluctuations to a large extent, but 
there still remains the loss of cooling during the time the surface is blanketed by 
the steam. If the surface is blanketed with steam during 25 per cent of the time 
the effective cooling surface will be reduced by about 20 per cent. 

When a pump is used, not only is good circulation promoted in the cylinder 
and head jackets but the increased velocity through the radiator matrix, to a 
smaller degree, increases the rate of heat transfer to the air stream. In some 
engines the pump does not circulate water through the cylinder jackets, only 
through the cylinder-head passages. In this way rapid warm-up is assisted. There 
are, of course, interconnecting passages between the cylinder and head jackets. 
The cooled water from the bottom tank of the radiator is led directly to the 
exhaust side of the cylinder head in such installations. Sometimes an internal 
tube is used, from which the water emerges in the form of high-velocity jets 
directed at the hottest zones behind the exhaust valve seats. 

A cross-flow radiator is sometimes used in modern low-profile sports cars. 
Such a radiator can be made wide and shallow, with an inlet and outlet header 
tank on each side of the matrix. The top of the radiator unit is usually below the 
top of the cylinder-head level, but a make-up header tank is usually situated at a 
higher level on the engine bulkhead. 

The design of an effective pump for a high-speed engine is not easy. The 
resistance of the cooling system to the flow of water remains fairly constant; the 
pressure of the pump to overcome this resistance varies as the speed of rotation. 
If the pump is designed to run at half-engine speed, circulation at 30 m.p.h. in 
top gear would probably be inadequate. If, however, the pump is designed to 
operate at engine speed, a phenomenon known as cavitation is liable to occur 
near maximum engine speed. Cavitation is caused by the creation of a high 
enough depression at the inlet to the impeller to flash off steam, thus causing 
suction at the impeller to break down completely. When this occurs, pumping 
ceases momentarily with, in some cases, spectacular and almost explosive 
boiling occurring at the cylinder head. 

104 The Sports Car 


Magneto or coil 

The early sports car engine almost invariably had magneto ignition, especially 
on the more highly developed engines giving peak power at speeds of 5,000 
r.p.m. or higher. Such speeds were very near the limits of coil ignition, which, in 
its stage of development at the time, could not be depended upon to maintain a 
high enough voltage at these higher engine speeds. The early rotating armature 
magnetos were limited in speed by the centrifugal stresses in the windings. 
Nevertheless they could give reliable service at engine speeds of 6,000 r.p.m. in a 
4-cylinder engine. Since those days there have been improvements in both 
magneto and coil ignition. 

The magneto is a more efficient spark producer at high speed than at low. The 
coil, on the other hand, requires a certain time interval for the current to build 
up to its full value. This time is of the order of 0.01 sec. This current build-up 
can only occur during the time the contact breaker points are closed, and at 
speeds of about 6,000 r.p.m. on a 6-cylinder engine the sparking voltage on the 
old design of coil using the pre-war design of distributor would fall to about 
one-third of the low speed value. The builders of the more expensive sports cars 
in the ’thirties solved the problem of good ignition at both ends of the speed 
range by using two plugs per cylinder, one plug sparked by coil ignition, the 
other by magneto. 

Coil ignition is still the choice today, but the spark is now triggered 
electronically on the latest designs. The mechanically operated contact breaker 
will soon be a ‘vintage’ memory. The Lucas electronic ignition is described later 
in the chapter. 

Sparking plugs 

Essentially a sparking plug is nothing more than a spark gap, but a spark gap 
that is asked to function under very arduous and dirty conditions. That it 
occasionally fails to operate is not really very surprising. For a plug to operate 
satisfactorily the nose of the insulator surrounding the central electrode must be 
at a temperature within the range 350-700°C (700-1300°F). If the temperature is 
too high, pre-ignition can occur. If the temperature is too low, the insulator will 
not keep clear of carbon. Since carbon is a good conductor of electricity it is an 
obvious necessity to keep the insulator nose clear of carbon deposits by 
maintaining it at a high enough temperature. A practical indication of this is 
given by the appearance of the nose when the plug is removed immediately after 
running at full power. A perfectly clean off-white insulator nose indicates too 
high an operating temperature. A light brown colour indicates a correct 
temperature and a plug blackened by ‘lamp-black’ type of carbon is running too 
cool. These indications are only relative and are influenced by other factors, 
such as mixture strength and the knock-rating of the fuel. 

Plugs are made in a wide range of heat values (see Figure 6.13). Cool-running 
engines require ‘hot’ plugs, or ‘soft’ plugs as they are sometimes called. High 

The Engine: Miscellaneous Components 105 

Fig. 6.13 A hot running plug usually has a greater distance from 
the tip to the plug body. A very cool plug has a larger 
diameter central electrode. 

compression, high output engines developing high temperatures and pressures in 
the combustion chamber require relatively ‘cool* or ‘hard* plugs for satisfactory 
running. Some of the heat picked up by the insulator and central electrode from 
the combustion gases is given to the relatively cool charge during the induction 
and compression strokes. The major path for this heat, however, is outwards 
through the body of the insulator to the copper seating in the plug body. For a 
hot plug the distance from the nose of the insulator to this seating is made large. 
For a cool plug the distance is much shorter. Although the maker’s recom¬ 
mended plug or one of equivalent heat value is usually the one to be used, some 
designs of engines are more sensitive in this respect than others and one 
sometimes finds that plug types have to be varied to suit the driver rather than 
the engine, a cool plug being required for the pole-position Grand Prix driver 
and a hotter plug for the tyro in the back row. 

To produce a spark across the sparking-plug points at the correct time in the 
cycle calls for a voltage surge up to anything from 5,000 to 15,000 volts, 
depending upon the size of the plug gap and the prevailing pressures inside the 
cylinder. This is well within the capabilities of the modern conventional coil 
ignition system under almost all conditions, except those of high-speed opera¬ 
tion. It is this limitation that the spark begins to fail when we require a sparking 
rate much higher than 400 per second that has led us at last into an era of 
experimentation after a period of design stagnation lasting for more than a 
quarter-century. Today we see transistorized ignition systems well established 

106 The Sports Car 

and capacity discharge systems in an advanced stage of development. One of 
these will probably be the choice for the next quarter-century. 

Wankel engines have presented a challenge to the plug manufacturers since 
the plug does not benefit from a cooling draught of mixture during an induction 
stroke or a compression stroke. The plug is in the firing line, so to speak, all the 
time since the same plug or pair of plugs are required to provide ignition for a 
succession of power strokes. Surface gap plugs have been developed for this 
arduous duty. Discharge occurs across the whole concentric area around the tip 
of the central electrode. These plugs have been found to function most 
effectively when used with the latest electronic ignitions where the voltage 
rise-time is very fast. 

The conventional coil ignition system 

The conventional circuit is shown in Figure 6.14. The cycle of operations begins 
when cam rotation causes the contacts to close and a current to flow through the 
primary windings, thus producing a magnetic field around the soft-iron core at 




I M 

• • 




“ Contact “ “ 


Spark plug 


Fig. 6.14 

the centre of the coil. Breaking of the circuit as the contacts are opened causes a 
rapid collapse of the magnetic field which generates a transient voltage in the 
primary windings which can surge to a value of 200 to 300 volts. The primary 
windings consist of about 200 turns of covered wire. The purpose of the 
condenser is the storage of electrical energy when the points open. A clean break 
in the current would not be possible without this storage capacity and the points 

The Engine: Miscellaneous Components 107 

would suffer from the severe arcing. As soon as the induced current stops 
flowing, the condenser discharges back into the primary windings, this back 
flow helping to remove the remnants of the magnetic field around the core. 

The secondary windings which are wound round the primary windings are 
made of extremely fine wire, about 20,000 turns in all. The two concentric coils 
thus constitute a transformer with a step-up ratio of 100 to 1. The more rapid 
the collapse of the magnetic field when the points open, the higher will be the 
voltage build-up in the secondary windings. The actual peak voltage reached is 
always that required to jump the plug gap. The larger the gap, the greater the 
voltage needed to jump it. Discharge of this secondary charge, lasting about a 
thousandth of a second, triggers an oscillating current in both primary and 
secondary circuits which persists until all the energy is dissipated. The contacts 
close again and the cycle is repeated. The purpose of the distributor is to channel 
the secondary current to each sparking plug in turn, depending upon the desired 
firing order. For a 6-cylinder engine running at 5,000 r.p.m. the time available 
for the whole sequence of events is about five thousandths of a second. The 
whole ignition story is told in Figure 6.15 by a typical trace from an ignition 

| [ L 1 1 __1_1_J- \ --J-1--1-1 

60 50 40 30 20 10 O 50 

Degrees of Distributor Rotation 
C 6 cylinder engine ) 

Fig. 6.15 

oscilloscope. In this particular case the secondary voltage has risen to a voltage 
of about 9,000 before the plug gap has ionized sufficiently for the spark to 
jump. As soon as the gap has become conductive and for the rest of the 
‘inductive’ part of the spark the voltage falls to a lower value to complete the 
discharge with a falling alternating current. The dwell section of Figure 6.15 is 
the period when the points are closed and the magnetic field is building up in the 

108 The Sports Car 

coil. If this period is inadequate, as it might well be at high engine speeds, the 
available voltage will fall in the manner shown in Figure 6.16. Fortunately the 
required secondary voltage also falls, since the falling volumetric efficiency at 
higher speeds results in lower cylinder pressures, which in turn require a lower 
voltage for a given spark gap to be bridged. The point of intersection of the 
‘available secondary voltage’ and the ‘required sparking voltage’ lines is the 
limiting speed for reliable ignition. Thus in the case of the high-compression 
engine shown in Figure 6.16 occasional misfiring will begin to occur at about 
7,000 r.p.m. 

0 1000 2000 3000 4000 5000 6000 7000 8000 

Engine r.p.m.-6 cylinder engine 

Fig. 6.16 Typical curves of secondary voltage and required sparking voltage 
with coil ignition. 

The conventional coil ignition system has been working very close to its 
maximum capacity for many years and, with the trend to higher engine speeds, 
frequent maintenance has been required to keep it functioning in a reliable 

New ignition developments 

Transistorized ignition systems have been available for several years and, when 
properly designed and with adequate cooling provided for the transistor, have 
been found to give reliable service with a marked increase in contact-breaker 
life. The majority of the American automobile manufacturers now offer a 
transistorized ignition system as an original equipment option. In these applica¬ 
tions, the transistor simply serves as a relay switch to reduce the current carried 
by the contacts (see lower circuit of Figure 6.14). A very small current, almost 
one-quarter of an amp, is all that passes across the contacts in this system, and 
this current by the action of the crystals in the transistor, is stepped up to a 
current of about 7 amps to pass through the primary coil windings. No 
condenser is required and the life of the contacts is extended to at least 20,000 

The Engine: Miscellaneous Components 109 

Transistors can be ruined by too high operating temperatures. They are 
usually mounted on a deeply finned block of aluminium, called a heat sink , but 
the location of this heat sink where too much heat is picked up from the 
surroundings can lead to transistor failure. Silicon transistors are most resistant 
to high temperature, but are more expensive than the popular germanium units. 

The simple transistor-switching system can only be regarded as an interim 
stage in the development of the new ignition systems. For the system of the 
future we must look to the new ‘breakerless’ techniques developed over the last 
ten to fifteen years. The Lucas ‘OPUS’ system has been used on racing engines 
for more than ten years and in its latest production version is available for 4, 6, 
8, and 12 cylinder engines. 

The Mark 2 OPUS (Oscillating Pick-up System) is a contactless system and 
has a spark capability of 800 sparks per second, twice that of the best 
mechanical contact-breaker systems. Since the system will operate for very long 
periods without maintenance it has been of great benefit to British firms such as 
Jaguar and Triumph in meeting the stringent American anti-smog laws. The 
system is shown schematically in Figure 6.17. It comprises four main compo¬ 
nents, an amplifier unit, a ballast resistor unit, an ignition coil and a distributor. 

The amplifier is a continuously operating fixed frequency (600kHz) oscillator, 
transformer-coupled to an amplifier stage and an output stage which is a power 

Fig. 6.17 Schematic arrangement of Lucas ‘OPUS’ Mark 2 
ignition system. 

Fig. 6.18 The Lucas 35 DE8 Electronic Distributor, one of a new range of 4, 6, 8 and 
12 cylinder distributors designed to comply with increasingly stringent 
regulations governing emissions and service intervals. 

The Engine: Miscellaneous Components 111 

transistor performing the function of the conventional contact-breaker. The 
amplifier unit is housed in a finned aluminium sink unit, since the power 
transistor must not be allowed to overheat. 

The distributor (see Figure 6.18) resembles the conventional design externally, 
since the distributor cap performs the usually duty of distributing the high 
voltage current to the plug leads. Inside the cap one also finds the customary 
rotor arm. The familiar contact-breaker is replaced by the timing rotor which is 
a moulded fibre-glass-filled nylon disc with ferrite rods embedded in the 
periphery. The number of rods is equal to the number of cylinders in the engine. 
The pick-up module is an E-shaped transformer core. The outer limbs of the E 
carry input windings which are fed from the oscillator in the amplifier unit. The 
centre limb carries the output windings to drive the amplifier stage. By design 
the resultant magnetic flux in the centre limb is negligible and the output signal 
is consequently also negligible when one of the ferrite rods is not in line with the 
pick-up module. When a ferrite rod passes across the face of the E-core the 
magnetic circuits become unbalanced giving an increased signal voltage in the 
output coil. When fed to the amplifier this signal causes the output power 
transistor to be switched off. This breaks the primary circuit in the ignition coil 
(as in the case of a conventional contact-breaker). The voltage rise-time is very 
high and a high voltage is produced in the secondary windings to create the 
required spark at the plug. 

A centrifugal advance mechanism is incorporated in the distributor and a 
normal vacuum unit adjusts the ignition timing under varying load conditions as 
in the conventional system. 

7 / 


'O what a flowery track lies spread before me, henceforth! 
What dust clouds shall spring up behind me 
as I speed on my reckless way!' 



Tyres were a source of great trouble to the early motorist, particularly when he 
raced. Figure 7.1 shows a competitor in the 1906 French Grand Prix struggling 
with a tyre-change at the side of the road while Vincenzo Lancia hurtles past in 
his 16-litre FIAT. The spare covers strapped to the back of Lancia’s car 
suggests that he also anticipated tyre trouble. 

To Dunlop and his generation his tyre was simply a means of absorbing the 
shocks that the rough roads of the period imparted to the bicycles and horseless 
carriages that travelled them. The pneumatic tyre’s potential for clinging to the 
road surface to provide the incredible cornering power that the modern sports 
car driver takes for granted had to wait for two important things. First we had 
to have good hard-surfaced roads to give the tyre something to grip and 
secondly we had to wait for the tyre manufacturers to develop the teeth to give 
that grip, the modern tyre tread with its multiplicity of grooves and slits so 
designed to wipe away any film of water and to maintain a steady bite on the 
road surface. 

The grip on the road 

The only roads that existed for the early motorist were of crushed stone and 
these soon disintegrated until the surface became an unpredictable loose mixture 
of broken stones, smaller debris and dust. Even a modern tyre tread fails to get a 
grip on such a surface and the early motorist was forever haunted by what the 
motoring journals of the period called ‘the dreaded side-slip’. 

The grip of the tyres on the road is just as important during braking, when 
accelerating and when cornering. One might expect some relationship to exist 
between the limiting thrusts that a tyre can exert on the road surface, forwards, 

Fig. 7.1 16-litre F.I.A.T. driven by Vincenzo Lancia in 1906 French Grand Prix. Note the need for frequent tyre 
changes. (Photography from the Cyril Posthumus collection). 

114 The Sports Car 

backwards and sideways. An investigation by the Road Research Laboratory of 
D.S.I.R. showed that the same forces can be exerted by the contact patch of a 
tyre under identical surface conditions in any direction. Thus, if we are 
travelling on a good non-skid surface with a vertical load on the tyre of 500 lb 
and the limiting cornering side thrust per tyre is 400 lb, i.e. a centrifugal 
acceleration of 0.8 G, the limiting retardation and the limiting acceleration in a 
straight line will also be 0.8 G, or a thrust at the road surface of 400 lb. On the 
other hand if the surface is wet and the limiting side force is only 200 lb, the 
limiting straight-line acceleration and deceleration forces will also be about 200 

The laws of friction are no longer regarded as a simple matter of applying a 
proportionality constant (called a coefficient of friction) to every pair of 
rubbing surfaces. We now know that the conditions at the surface are often 
more important than the nature of the materials themselves. 

Five main types of friction are usually distinguished: 

(a) Smooth dry friction. This is self-explanatory. It is seldom encountered on 
the public highway. Polished granite paving stones, still used on a few roads in 
the North of England, almost approach this condition. 

(b) Rough dry friction. As in the first case the surfaces are in direct contact 
and are perfectly dry, but, owing to the uneven nature of one or both of the 
surfaces, projections on one cut through or lacerate portions of the other. This 
is encountered often when skidding on dry roads, particularly with an asphalt or 
tarred road with a rough surface. Under these conditions decelerations of 0.8 to 
0.8 G are possible. 

(c) Loose dry friction. When the two dry surfaces are held apart by a layer of 
loose particles such as sand or gravel, a sliding action takes place which is 
sometimes called a ball-bearing action. The more rounded the particles, the 
more they roll, the more angular, the more sliding and the less rolling takes 
place. The overall effect is a reduction of the coefficient of friction when 
compared with smooth dry surfaces in direct contact. This condition is 
frequently encountered on the road, especially near the road verge where loose 
gravel accumulates. 

(d) Lubricated friction. This occurs when a liquid film completely separates 
the two surfaces, as in a well-lubricated bearing. This seldom occurs between the 
tyres and the road surface, except in the case of an icy road when a film of water 
can be maintained between the road surface and the contact patch of the tyre, in 
the same way as the film of water is formed by the pressure of the blade of a 
skate. The effect is very similar in both cases. Unfortunately the sliding motorist 
cannot turn his skates sideways in the manner of an expert skater! Occasionally 
wet granite paving stones give true lubricated friction, but in general a film of 
grit is present to modify the action. 

(e) Partially lubricated friction. This occurs when separation by a liquid is 
not complete and the surfaces are in actual contact in places or, at the most, are 
only separated by a ‘liquid’ film of one or two molecules in thickness. This 
happens on wet roads and locked wheel decelerations of from 0.6 to 0.4 are 

70 series 

Road-Holding 115 

Fig. 7.2 Tyre footprints. 

116 The Sports Car 

possible in these conditions. On muddy roads and particularly after a short 
shower of rain when the dry dust has been converted to a film of mud of such 
high viscosity that the squeegee action of the tyre treads fails to expel it 
completely from the surfaces in contact, conditions can again approach the fully 
lubricated state and locked wheel decelerations can be limited to values as low as 
0.05 G. When acceleration or deceleration produces sufficient heat to melt the 
tar surface a condition of partial lubrication exists. When black streaks are 
produced in the road surface the limiting acceleration is usually only about 
0.5 G. 

The tyre footprints 

Typical static tyre footprints are shown in Figure 7.2. These are static footprints 
or contact patches for tyres of various tread designs. For maximum traction and 
cornering on both dry and wet roads it is necessary to provide, not only 
circumferential grooves but a carefully evolved system of slots and ‘sipes’ that 
clings to the surface to give us the ‘Graunchy Grippers’ of the advertisements. 
An adequate drainage pattern to remove water from the contact patch as fast as 
it enters the leading edge is also necessary. 

The dynamic behaviour of the footprint 

Figure 7.3 shows the manner in which a tyre obtains a side thrust from the road 
surface to counter the centrifugal force when turning a corner or to prevent the 
car running into the ditch on a steeply cambered road. In Figure 7.3 the tyre is 
viewed from below as if running on a sheet of glass. This diagram shows that the 
plane of the wheel makes an angle, <p, with the path along which it is travelling. 
This is called the slip angle and the greater the side force required from the tyre, 

Fig. 7.3 Distortion of tyre centre-line at contact 
patch produces cornering force and self¬ 
aligning torque. (Distortion exaggerated). 

Road-Holding 117 

the greater will be the slip angle up to the point at which it loses its grip on the 
road surface and skids. This is a similar condition to the stalling of an aeroplane 
wing when the critical angle of incidence is exceeded. The cornering force does 
not act at the centre of the contact patch, but slightly to the rear. This, plus the 
designed castor displacement of the steering layout, results in a turning moment, 
Fxd , called the self-aligning torque, which must be countered by the pull 
provided by the driver on the steering wheel. It is this torque, plus or minus any 
torque set up by the camber of the wheels, that goes to make the ‘feel’ of the 
steering in the driver’s hands. This feel can sometimes be misleading, since it 
bears no relationship to the cornering power of the tyres. 

Cornering power 

The manner in which the tyre surface creates this cornering power is as follows: 

Each portion of the tread as it approaches the contact patch must be pushed 
outwards and distorted from its natural shape to conform to the particular slip 
angle at which the tyre is operating. The rest of the tyre casing and tread must 
exert the force needed to distort the contact patch and it is this continuous 
process of distortion as each portion of the tread approaches the contact patch 
that contributes to the cornering power exerted on the vehicle. As each section 
of tread leaves the contact zone it recovers its original shape, but no work is 
done here, except a certain loss of energy, called hysteresis loss, which results in 
a heating up of the tyre. 

The cornering power of a tyre is defined as the cornering force divided by the 
slip angle required to provide this force. The usual units are lb per degree. This 
cornering power under any given load can be measured on a special laboratory 
machine. Some companies, such as Dunlop, use a large rotating drum. Others 
prefer to use what is called a flat-bed rolling tyre tester. With these machines slip 
angles can be measured accurately under any combination of vertical loads and 
side loads to produce sets of curves of the type shown in Figure 7.4. With these 
machines it is possible to explore the influence of many factors that enter into 
the design of a tyre, such factors as tyre section width, tyre pressure and the 
influence of the angle at which the textile cords are set across the tyre centre line. 

The following is a brief summary of these influences. 

Inflation pressure . The introduction of ultra low profile tyres has accen¬ 
tuated the importance of tyre pressures. The aspect ratio (height to width 
ratio) of a tyre cross-section is variously expressed by tyre manufacturers 
as ‘70 Series’ or ‘70 Profile’ or by the aspect ratio incorporated after a ‘stroke’ 
in the tyre size, i.e., 185/70 VR 15; this indicating a tyre 185 mm wide with 
a 70 profile on a 15 inch diameter wheel and a VR speed rating — usually 
up to 150 m.p.h. 

When tyre profiles were all at 100 percent the contact patch was very narrow 
and more heavily loaded. The actual tread profile (not profile ratio) presented to 
the road surface was very rounded and an increase in cornering power could 
usually be obtained by running at 10 or 15 per cent above the maker’s 
recommended pressure. All this has changed now and with 70 or 60, and 

118 The Sports Car 

especially with the 50 Series tyre it is essential to maintain the correct tyre 
pressure, cornering power being reduced appreciably by a variation, either 
upwards or downwards, of more than 2 lb per sq in. The tread profile is very flat 
on these tyres and overpressure makes this slightly convex thus reducing the area 
of the contact patch. Conversely, a hollow centre part can be given to the 
contact patch with underinflation. 

Grand Prix drivers using tyre profiles as low as 35 and even 30, are not only 
much concerned during the warming-up laps with the warming-up of their sticky 
tread compounds, but are also very finicky about tyre pressures, knowing that a 
slightly concave footprint is just as dangerous as a convex one. Most modern 
high speed road tyres are confined to the 70 and 60 Series although a few tyre 
manufacturers have ventured lower. Until tyre manufacturers can provide a tyre 
less sensitive to small changes in pressure this factor alone will provide a 
practical limit of about 50. 

Camber. The influence of wheel camber on cornering power has become 
all-important in recent years since the introduction of ultra-low profile tyres. 
Racing car designers have even been turning in despair to thoughts of beam 
axles, a simple solution to the problem of keeping the wheels upright in a corner. 
With a profile of 50 or under there is a serious loss of cornering power when the 
camber angle exceeds about 3 degrees in either direction. De Dion suspension, as 
used by Aston Martin, is one solution to this problem since this is a sophisti¬ 
cated form of beam axle. Where independent suspension is used at both ends of 

Road-Holding 119 

the car a suspension geometry is usually chosen that maintains the outer wheels 
within about 2 degrees of vertical when cornering. Since the inner wheels are less 
lightly loaded when cornering near the limit, relatively larger changes in camber 
angle can be permitted without seriously reducing the overall cornering power. 

Rim width. Wide tyres are less effective on narrow rims and a change to a 
lower profile tyre usually demands a change to a wider wheel. 

Traction, braking and acceleration. Despite all the development work carried 
out on tread patterns one surprising fact always seems to emerge when tyres are 
tested. A tyre footprint will exert almost the same limiting force in any 

Fig. 7.5 The Circle of Forces: how available cornering force 
is reduced during traction and braking. 

120 The Sports Car 

direction . Thus if we apply traction or braking effort while the footprint is also 
subjected to a cornering force the limiting cornering force will be reduced. The 
‘Circle of Forces’ shown in Figure 7.5 is a simple way to demonstrate this 
concept. Laboratory experiments carried out by the Ford Motor Company of 
America in the Sixties on a 9.00 x 14 tyre as used on a full-size sedan showed 
that traction or braking would reduce the limiting cornering force by as much as 
30 per cent. Expressed another way one could say that with more moderate 
cornering the slip angle would have to be greater to exert the same cornering 
force when the car was braking or accelerating at the same time. 

Load . The variation of cornering force with changes in the load carried by 
the wheel is best illustrated by Figure 7.4, which is a plot of side-thrust against 
load over a range of slip angles for a 6.00 x 16 tyre inflated to 28 lb per sq in. 
pressure. It should be noted that cornering force, not cornering power, has been 
plotted since the curves of cornering power against load would lie too much on 
top of each other for clarity. The influence of tyre load is seen to be small at 
small slip angles, but as the slip angle increases it becomes greater. When 
cornering near the limit, slip angles of 10° and in favourable circumstances even 
14° are possible. In such conditions the transfer of load from the inner to the 
outer wheels by the rolling couple acting on the sprung mass begins to exercise 
an important influence on the cornering behaviour of the car. No transfer when 
cornering, a load of 800 lb on each wheel and a slip angle of 5° will give a side 
thrust of 440 lb per wheel for the tyre considered in Figure 7.4. But, in general, 
some transfer does occur and with a transfer of 400 lb from the inner to the 
outer wheels it is seen that a slip angle of 6 | 0 is now required to provide an 
average side thrust of 440 lb per wheel, i.e. 340 lb on the inner wheels and 540 lb 
on the outer wheels. From this we see that, all other things being equal, a car 
with a small roll angle on corners can corner slightly faster with safety than one 
with a large roll angle. The effect is most pronounced when the mean tyre load 
coincides with the peak of the side thrust/tyre load curve. Without load transfer 
a load of 1,200 lb and a side thrust of 600 lb would give a slip angle of 7° (see 
Figure 7.4). A load transfer of 50 per cent is quite moderate, but such a transfer 
would increase the slip angles to something greater than 12° which in many cases 
would result in a bad skid. The importance of choosing the tyre sizes to operate 
on the rising part of the thrust/load curve is thus emphasized. 

Tyre construction 

Twenty years ago we could divide tyre construction into two very distinctive 
types, cross-ply and radial. Since then the American tyre manufacturers, with 
large capital investments in plants to make cross-ply tyres, have shown ingenuity 
to combine the two types of tyre into bias-belted tyres and at the same time to 
make these with very little modification to their existing plants. On this side of 
the Atlantic there have also been many variations evolved on the original 
Michilin X steel-braced radial tyre. 

The older design of tyre, the cross-ply, has the rubber carcase of the tyre 
reinforced by layers of bias-cut fabric which pass across the tyre from side to 

Road-Holding 121 

side being anchored to the bead wires on both sides. A conventional angle would 
be from 20 to 30 degrees to the circumferential centre-line. Cross-ply construc¬ 
tion is shown in Figure 7.6. In this illustration the aspect ratio is 100 per cent. 

Fig. 7.6 Internal construction of cross-ply tyre. 

Cross-ply construction is used, however, in racing tyres with profiles as extreme 
as 30 per cent. On such tyres the cord angle is decreased to about 20 degrees to 
resist centrifugal growth at high speed. This gives a very harsh uncomfortable 

Radial tyres, by definition, have the textile cords passing from bead to bead 
radially, i.e. at an angle of 90 degrees to the circumferential centre-line. This 
makes the side walls very flexible and with no attempt to stiffen the tread zone 
the natural cross-section would be circular and the unsupported tread would 
twist and squirm in all directions. In the original Michelin X tyre three-plies of 
steel mesh were placed below the tread to stiffen it and the effect was quite 
remarkable. Not only did it stabilise the tread in a way never achieved before on 
any cross-ply tyre but it reduced tread movement in the footprint area so much 
that tyre wear was reduced dramatically. Slip angles were lower for a given 
cornering force than on a cross-ply tyre and the limiting cornering forces were 
higher. The tyre crown was so resistant to sideways movement under high 
cornering forces that the gradual reduction in contact patch area that occurs on a 
cross-ply tyre as it reaches the limit was absent. This tended to give a rather 
sudden breakaway. Even so the original X tyres were, in their day, unsurpassed 
in the wet. Disadvantages were increased noise, a harsher ride and heavier 

Today after nearly thirty years of development the steel-braced tyre is a much 
more refined product and is in competition with a whole range of variants using 

122 The Sports Car 

combinations of many textile materials such as nylon and polyester to brace the 
tread zone. Typical of this modern generation of radials is the Dunlop SP Sport 
shown in cross-section in Figure 7.6. The same internal construction is used in 
the 60 Series SP Sport Super shown in Figure 7.8, but the tread bracing layers in 
this case are of steel. The SP Sport Super has a large block tread pattern with 
wide channels for water drainage. In the 70 Series it is fitted to the Jaguar XJ-S, 
and the Porsche 911. In the 60 Series it is used on the Lotus Elite. 

Bias-belted tyres have been developed by the American tyre manufacturers 

Tread Pattern 

Tread Bracing Layers 

Wall Rubber 



Fig. 7.7 Internal construction of Dunlop SP Sport sti 

Road-Holding 123 

Fig. 7.8 Dunlop SP Sport Super Tyre (60 Series) as used on Lotus Elite. 

and are cross-ply tyres in general construction, but with the addition of a 
stiffening belt of steel mesh, nylon or polyester cord. As one would expect their 
general behaviour and performance falls somewhere between that of the old 
cross-ply and the best radials. 

Tyre compounds 

Natural rubber is rather soft, but has good flexibility even at low tempera¬ 
tures. Hardness can be increased by the degree of vulcanisation and by 
compounding with synthetic rubbers. It has poor resistance to oxidation and 
attack by atmospheric ozone. 

Chloroprene copolymer (neoprene) is not as ‘bouncy’ as natural rubber, but is 
far superior in resistance to high temperature and ozone. It becomes very stiff at 
low temperatures. 

124 The Sports Car 

Styrene butadiene copolymer (SBR), when compounded with reinforcing 
fillers, has physical and chemical properties very similar to natural rubber. It 
possesses high resistance to abrasion and has contributed largely to the much 
improved grip of modern tyres in the wet. Compounds using large additions of 
SBR are sometimes called ‘high mu’ compounds. Mu (p) being the Greek letter 
used by tyre technologists for the coefficient of friction between the rubber and 
the road surface. Unfortunately, high mu compounds also exhibit high values of 
the property called ‘hysteresis’. Hysteresis in this context means the amount of 
energy dissipated internally when the rubber is compressed and then released. 
This energy is converted into heat and high mu compounds, although suitable 
for wet-weather tyres, can become dangerously overheated when used on dry 

The choice of tyres at the start of a race is a headache for team managers, but 
we, happily, are not concerned with such matters in this book. For a sports car 
to be used every day of the week we need an all-weather tyre and the rubber is 
compounded to give a satisfactory compromise between wet and dry require¬ 
ments and always with the need for excellent abrasion resistance in mind. 


Aquaplaning (hydroplaning in America) is in effect high speed skating of the 
tyre on a film of water when travelling on wet roads. It was observed many years 
ago that aircraft landing on wet runways would occasionally suffer the loss of 
all braking effort. Investigation showed that the tyres could ride up on a wedge 
film of water in a manner that is closely related to the hydrodynamic wedge 
action of a lubricated bearing. The frictional coefficient becomes negligible and 
t»ll braking and steering effort is lost. The same phenomenon can occur at speeds 
of 80 m.p.h. and over with standard automobile tyres. With a smooth tyre 
aquaplaning can occur at speeds as low as 60 m.p.h. — a reduction in the grip 
on the road can even occur at speeds as low as 35 m.p.h. Figure 7.9 shows 
photographs taken through a special glass section in a runway during experi¬ 
ments at the NASA Langley Field Center using 6.50 x 13 passenger car tyres. 
With a smooth tyre the contact patch completely disappears at 80 m.p.h. With a 
ribbed tyre the contact patch is reduced in area at 80 m.p.h. but is still making 
an effective enough contact for some braking effort to be applied to the road 
surface. At some unspecified higher speed even the ribbed tyre will lose all 
contact with the road surface and will slide as freely as on an icy surface. 

The Dunlop Rubber Company developed the C41 tread to provide rapid 
effective drainage channels for the water film. This tyre has been shown to be 
about 15 per cent more effective on fully wetted surfaces than other tread 
designs. New patterns have since been developed which eject the water sideways 
along ducts moulded below the tread. The Dunlop SP Sport radial (79 Series) 
shown fitted to an E Type Jaguar in Figure 7.10 uses these water-clearing 
‘aqua-jets’ to pump water away from the contact patch and is also given a 
central drainage groove, a technique learned by Dunlop to combat aquaplaning 
on their Formula 1 tyres. 

126 The Sports Car 

Fig. 7.10 Dunlop SP Sport radial (70 Series) fitted to E-Type Jaguar. 

Oversteer and understeer 

The stability of the motion of a car when travelling in a straight line on the 
highway at almost any speed is entirely a question of oversteer and understeer. 
Let us consider the case of a car travelling in a straight line on an uncambered 
road when the car is suddenly subjected to a small side thrust, such as might be 
caused by a gust of wind or a pot-hole. Under the action of the side thrust all 
four wheels will run at slip angles, 0! at the front and 0 2 at the rear. If 0, is less 
than 0 2 the car will steer towards the side force. The car is now following a 
curved path and the centrifugal force thus produced will augment the initial side 
force and 0i and 0 2 will both increase to new slip angles 0/ and 0 2 '. The 
increase in 0 2 will be greater than the increase in <p lt thus increasing the 
centrifugal force still more. It is easy to see that the car is in an unstable 
condition and the situation will call for continual steering correction from the 
driver. This condition is known as oversteer. 

The reverse effect is produced when 0j is greater than 0 2 , since the car turns 
slightly away from the disturbing side force, producing a centrifugal force in the 
opposite direction that tends to cancel the initial disturbing force. This is a 
stable condition and means that many of the small disturbing forces are 
automatically damped out by the steering behaviour of the car itself without any 

Road-Holding 127 

action being required from the driver. This condition is called understeer . While 
opinions may differ as to the most desirable behaviour of a car when cornering, 
there is no doubt that every car should be set up to be in an understeering 
condition when running in a straight line. The amount of understeer need only 
be slight. The manner in which the designer arranges that the slip angle is always 
greater at the front is a complex subject. Some of the factors that influence 
under steer and over steer will be considered later in the chapter. 

Steering layout 

The original Ackermann steering layout as used on the earliest horseless 
carriages was modified by Jeantend in 1878. On the older beam front axle the 
modified Ackermann geometry was given by inclining the links which connect 
the wheel steering pivots and the track-rod ends so that, if extended, they would 
intersect on the car centre-line about two-thirds of the wheel-base from the 
front. Only when making very slow turns, i.e. about 5 m.p.h., would a car turn 
about the Ackermann centre. Not only would the rear wheels be turning about a 
smaller radius than the front, but by turning about the Ackermann centre, the 
slip angles would be zero on all four wheels. It has already been demonstrated 
that a tyre must exert a side thrust when a car is cornering and this side thrust 
can only be exerted by the tyre running at a slip angle. It is reasonable to 
suppose that at moderate and high speeds the actual turning centre will be at 
some point further forward than the Ackermann centre. By turning about a 
centre such as B in Figure 7.11 the required slip angles are obtained. 

It is now realized that the modified Ackermann layout is unsound for 
high-speed motoring. The outer wheel, the one taking most of the load, is given 
a smaller slip angle than the lightly loaded inner wheel. A negative Ackermann 
layout has been adopted by several modern designers since this allows the inner 
and outer wheels to operate much closer to their natural slip angles when 

Fig. 7.11 Steering layout. 

128 The Sports Car 

The car shown in Figure 7.11 is turning on a radius of 60 ft. (The drawing is 
not to scale.) Let us assume that the gross weight is 2,000 lb and that the slip 
angles, both front and rear, are 6 V 2 degrees. If we simplify the problem by 
assuming tha the car does not roll, the load on each tyre will be identical and 
equal to 500 lb. 

The centrifugal force resisted by each wheel, 

Wv 2 500v 2 
"gT “ 32-2 X 60 

0-26v 2 

The side thrust per wheel with a vertical load of 500 lb and a slip angle of 6 V 2 
degrees we find from Figure 7.4 to be 400 lb for a tyre size of 6.00 x 16. This 
side thrust must of course equal the centrifugal force, i.e. 

F= 400 = 0.26v 2 
v = 56 ft per sec = 38 m.p.h. 

If the car were travelling at a higher velocity, but still turning about a 60 ft 
radius the centrifugal loads would be higher, increasing as the square of the 
velocity, the required slip angles would be greater and the front wheels would be 
turned further into the corner so that the Ackermann centre would now be at 
some point A'. In practice the position B, the actual turning centre, will depend 
on the relative sizes of the slip angles on the front and rear wheels and the 
amount of tractive effort applied to the driving wheels. For an oversteering car, 
B will be further in since the slip angle on the rear wheels will be greater than on 
the front. For an understeering car, B will be further out. 

Triangle of Forets Triangle of Forces 

(not accelerating} <accelerating) 

Fig. 7.12 Balance of cornering forces: rear wheel drive. 

Road-Holding 129 

Rear wheel drive car 

Let us first consider the simple case of a typical front-engined sports car with 
rear wheel drive and a fairly high power to weight ratio. 

When an understeering car enters a corner at speed the driver finds that the 
front of the car must be pointed further into the corner than appears necessary 
from the actual curvature of the corner. This, of course, is caused by the need to 
generate cornering forces, i.e. the establishment of slip angles. Photographs of 
racing and sports cars which appear to be about to run into the inside verge on a 
corner are frequently shown in the motoring press. This effect is most noticeable 
on the ‘drifting’ racing car but it takes place, although to a less noticeable 
degree, on any car. 

The four-wheel drift is a familiar sight to the race-going crowds, but the 
manner in which it works is not generally understood. Let us consider that the 
car in Figure 7.12 is a rear-wheel drive racing or sports car with a high power to 
weight ratio and that it is taking a corner at such a speed that the slip angles are 
high, let us say about 9° or 10°. The car is arranged by design to understeer on 
corners. With a slip angle of 10° on the front tyres, that on the rear tyres is only 
8°. The effects of roll are neglected for simplicity. Taking first the simple case 
when the torque applied to the rear wheels is just sufficient to overcome the 
normal resistances of the road and the aerodynamic drag, i.e. the driver is going 
through the corner with his foot on the accelerator in the identical position it 
would occupy when travelling at the same velocity on the straight. On a long 
curve this would result in a steady drop in speed through the curve, since more 
power is required to maintain a given speed on a bend than on the straight. The 
centrifugal force, acting along the line BO , is equal to \xW , where W is the gross 
weight of the car and \x is the value of the centrifugal force expressed in 
‘gravities’. If W equals 2,000 lb and \x equals 0.8 the centrifugal force will be 
1,600 lb, or 400 lb per wheel. 

To hold the car on its circle around point B the tyre cornering forces acting 
towards this point will be F x and F 2 as shown in the triangle of forces at the 
bottom of the figure. 

Now let us suppose that the driver, wishing to maintain a constant speed 
throughout the bend, steps on the accelerator and applies an additional thrust at 
the rear tyre contact surface of 100 lb per wheel. The triangle of forces to the 
extreme right of the rear wheels indicates this force as C and the resultant thrust 
from the rear wheels is seen to be F s . The immediate effect of this is to move the 
instantaneous turning centre from B to B '. The angle between F\ and F3, the 
resultant wheel thrusts when accelerating, is now greater than the angle between 
F x and F 2 . This means that to counteract the same centrifugal acceleration, the 
slip angles must be greater when accelerating round a bend, since F\ and F 3 
must be greater than F x and F 2 for the same value of juW. 

One important effect of this additional thrust from the rear wheels is its effect 
on the ‘attitude’ or angle of yaw of the car to its path. The instantaneous 
direction of the car is at right angles to B 'O, but the centre-line of the car is at an 
angle to this direction line. The yaw angle will vary in size with the extent of the 

130 The Sports Car 

acceleration. All the way round the bend the car is held at a yaw angle and if at 
any time the car gets dangerously near the verge the driver can either steer out or 
press harder on the accelerator. If he chooses to do the latter, the turning centre 
will move further forward from B' to B " (see Figure 7.12), the angles between 
F'i and F 3 will increase still further and, to counteract the same centrifugal 
force, the slip angles at the back and front will have to increase. If the driver 
overdoes the acceleration, the front tyres will skid first, since the slip angles are 
greater at the front. This skidding of the front wheels automatically reduces the 
centrifugal force by increasing the radius of the turning circle. Provided there is 
room on the outside of the bend the loss of the true drift position in this way is 
not dangerous and the experienced driver soon recovers the correct yaw angle. 
Some drivers prefer to hold a constant throttle position on a bend and to make 
steering corrections to hold the drift. Techniques in starting the drift differ from 
driver to driver. A slow-motion film of Stirling Moss in action was once made 
by the Shell Film Unit. He started the drift with a sharp jab at the brake pedal 
and a flick of the steering wheel to catch the correct yaw angle. The car was held 
at the correct yaw angle by amazingly quick small steering corrections. 

One of the characteristics of cornering near the limits of adhesion is the lack 
of feel at the steering wheel. Self-aligning torque usually increases up to a slip 
angle of 5°. At greater angles the torque falls off until at about 10° it becomes 
zero. In conditions where greater angles are possible the self-aligning torque can 
become negative. In other words, instead of the driver having to hold the 
steering wheel to prevent it flying back to the straight-forward position he finds 
it necessary to apply force in the opposite direction. To the inexperienced driver 
this can be disconcerting since the feel of the steering is contrary to his previous 

Fig. 7.13 Balance of forces: front wheel drive. 

Road-Holding 131 

Front-wheel drive 

The forces acting on a front-wheel-drive car when turning a corner are shown in 
Figure 7.13. When accelerating, the car travels at a yaw angle as in the case of the 
rear-wheel-drive car, but the line of action of the wheel thrusts, F ’ 2 and F 3 , can 
come more into line with the centrifugal force n'W' and can therefore be slightly 
smaller than in the non-accelerating case. This is the explanation of the 
oft-repeated saying that a front-wheel-drive car is ‘more stable’ when accelerating 
round a corner. It would be better to say that for a given limiting cornering thrust 
the F.W.D. car can corner faster when accelerating than when decelerating. 
Moreover, the F. W.D. car is inherently faster round a corner than the R. W.D. car 
when the accelerating technique is used correctly. From Figure 7.13 it will be seen 
that there is an optimum value for the acceleration. If C is too great, F 2 and F 3 will 
come out of line again. Here lies the danger in a front wheel drive racing car. When 
the optimum value is achieved the driver has reached a point of no return. Lifting 
the throttle, touching the brakes, even a slight variation in the line through the bend 
can result in a disastrous slide. Modern front wheel drive cars are very popular in 
Europe but the torque they can transmit to the front wheels when cornering is 
hardly enough to maintain a constant speed. Careful tuning of the suspension 
geometry and the strength of the anti-roll bars can give very forgiving handling on 
these underpowered family saloons. It is significant that no production sports cars 
are made today with front wheel drive. 


The effect of braking on the position of the turning centre is exactly the opposite to 
the effect of acceleration. The deceleration thrusts at the road surface are of 
opposite sign, as shown in Figure 7.14 and the turning centre moves backwards to 
some point such as D. The immediate effect on the car is that it drifts towards the 

Fig. 7.14 Cornering forces during braking. 

132 The Sports Car 

inner verge, since the yaw angle has reversed and if the driver does not correct this by 
immediately steering outwards the car will run into the verge. This is occasionally 
seen in racing when a driver is forced to brake to avoid a collision on the apex of a 
corner. Sometimes, however, savage braking on the soft-sprung modern sports car 
throws a greater load on the front wheels. Too great a cornering thrust is demanded 
from the front tyres and the car shoots off tangentially from its former turning 

The mid-engined sports car 

Motoring writers sometimes state that the modern Grand Prix car is set up to 
oversteer. This is misleading, or at least an oversimplification. Formula 1 cars and 
turbocharged racing sports cars such as the Porsche 917 are capable in dry 
conditions of transmitting extremely high forces through their ultra-wide rear 
tyres. With a power to weight ratio of about 700b.h.p. per ton the simple concept of 
‘the Circle of Forces’ becomes of the utmost importance if we are to understand the 
behaviour of such a vehicle in a corner. If the driver puts his foot down harder in the 
middle of a corner the slip angles at the rear increase dramatically. If he eases the 
pressure they are decreased. The degree of oversteer is therefore under the control 
of the driver and in this sense the car may be said to be steered by the right foot. This 
bears little resemblance to the oversteer demonstrated by men like Gonzales on the 
1 Vi litre supercharged BRM when he completed many corners with the steering on 
full opposite lock. This was very spectacular but only served to demonstrate a 
design deficiency. 

The tread compounds and rounded tyre profiles of the racing tyres used in the 
early fifties were vastly inferior in performance by modern standards. Such 
notoriously oversteering monsters as the pre-war P-Wagen and the ill-fated 
original BRM could have been persuaded to handle in a most gentlemanly manner 
if modern tyre technology had been available at the time. 

The Ferrari Dino, Lotus Esprit, Porsche Carrera and other mid-engined sports 
cars exploit the same cornering technique as the modern Grand Prix car, even to the 
use of wider tyres at the rear than the front. To a more modest extent, since the 
power to weight ratio is lower, the latest mid-engined sports cars can be controlled 
in a fast corner by the use of the accelerator. With the main masses of engine, 
transmission and passengers so disposed as to give a low polar moment of inertia 
the mid-enginea car has a very quick steering response. Over-enthusiastic use of 
the accelerator when cornering is soon corrected by a quick flick of the very 
responsive steering. 

Factors leading to understeer 

The following factors all tend to produce understeer when cornering. Used singly 
or in combination almost any car can be made to understeer to the required degree: 

(1) Re-distribution of the sprung weight so that a greater proportion is 
carried by the front wheels. 

(2) Fitting larger tyres at the rear than at the front. This is much in evidence 
in Grand Prix racing. 

Road-Holding 133 

(3) Arranging for the front wheels to have a greater outward camber than the 
rear wheels when the car rolls on a corner. This happens automatically on the 
most common suspension combination in use today, i.e. with wishbone I.F.S. 
and a beam or De Dion axle at the rear. When independent suspension of the 
wishbone type is used at both front and rear it is sometimes necessary to arrange 
the rear links in such a way that the camber when cornering is less at the rear 
than at the front. This can be achieved by making the lower links on the rear 
suspension disproportionately longer than the top links. This is sometimes 
referred to as providing a ‘swing-axle effect’ since a true swing axle gives a 
negative (inward) camber on a corner. The resulting camber, when longer 
bottom links are used at the rear, is still positive, but not as great as at the front 
for a given roll angle. This reduces the cornering powering of the front tyres 
relative to the rear and gives the desired understeer. 

(4) Transferring more of the roll couple from the rear to the front. The 
rolling action of the sprung mass (body, frame, etc.) on a corner is resisted by 
the springs. An understeering tendency is given by making the front springs 
provide the major resistance to rolling. The anti-roll bar which is fitted to the 
front of many modern cars produces this effect. 

Gyroscopic effects 

A factor which is always present when cornering and which is often ignored is 
the gyroscopic reaction of the rotating mass of the engine components on the 
road wheels. This effect is well known to aircraft pilots and is unconsciously 
corrected by the experienced pilot. With the normal direction of engine rotation 
a left-hand turn produces a tendency for the nose to dip, i.e. a transfer of weight 
from the rear wheels to the front, a right-hand turn throws more weight on the 
rear wheels. On the level this effect is only noticeable in cars with very high 
power to weight ratios and at high engine speeds. 

This gyroscopic torque should not be confused with the torque reaction which 
occurs during acceleration on cars with non-independent suspension at the rear. 
This causes a transfer of weight from the right-hand near wheel to the left-hand 
and is responsible for the smoking of the rear right-hand tyre when a 
big-engined car is accelerating from a standstill. 

Roll centres 

The relative positions of roll centres of the two ends of the car have a profound 
effect on the cornering behaviour and can explain the skittish manner in which 
some cars lift a wheel when cornering fast. 

The roll centre may be defined as the point in the transverse plane of the front 
or rear suspension about which the sprung mass of the car rotates under the 
action of a side load, such as the centrifugal force when cornering. The sprung 
mass is of course all that part of the car mounted on the springs, i.e. body, frame, 
passengers, etc. In the case of a De Dion or independent rear suspension it includes 
the rear drive box. The position of the roll centre on several suspension systems 
is shown in Figures 7.15 to 7.22 and is indicated by Mo in each case. 

134 The Sports Car 

M 0 

Fig. 7.15 

Figure 7.15 is the popular unequal length wishbone system which is used at 
the front on the majority of cars and is also used at the rear on a few sports cars. 
To find the roll centre a line is drawn from the instantaneous link centre O, 
through the road wheel contact point to intersect the vertical centre line of the 
car. On a few older cars parallel wishbones are used in as Figure 7.16. Since in 

Fig. 7.16 

this case the instantaneous link centre is at infinity M 0 must be at the road 
surface. The Morgan system (also used by Lancia on the early Aurelias) is 
shown diagrammatically in Figure 7.17. Here again the roll centre is at road 
level. The trailing link system, used by Porsche for so many years at both front 
and rear is simply another means of providing a parallel motion (parallel when 
viewed from the rear) to wheels on opposite sides. The roll centre for this system 

Fig. 7.17 

too is at road level. A system that has become popular is the MacPherson or 
Chapman-strut system, as used on many former Lotus designs (see Figure 7.18). 
When used to provide independent rear suspension the wheel is located laterally 
by the fixed length drive shaft. Fore and aft location is provided by a two-piece 
radius arm. The action of the sliding strut, which is a combined damper and 
spring unit, and the joint action of the radius arm plus the lateral restraint of the 
flexibly-jointed drive shaft is very similar to an unequal length wishbone system, 
but with the upper wishbone inclined downwards, thus bringing the instantaneous 
link centre close to the plane of the wheel. Small changes in the forward location 
of the radius arm will change the roll centre height. 

Road-Holding 135 

Fig. 7.18 Chapman strut on early Lotus sports car. 

The roll centre on the beam axle, supported on semi-elliptic laminated 
springs, as used at both front and rear on pre-war sports cars (see Figure 7.19), 
is situated at a point between the axle centre-line and the spring shackles. To 
reduce the roll-centre height slightly the springs can be given a reverse camber, 

thus lowering the position of the shackles. The D-Dion axle shown in Figure 
7.20, is basically the same arrangement; the only difference being that the axle is 
not located by the springs. M 0 in this case lies at the pivot point of the axle as 

Fig. 7.20 

Fig. 7.21 

136 The Sports Car 

defined by the axle-locating links, shown here as a block sliding in vertical 
guides. Aston Martin use a Watt’s parallel linkage. 

Two types of swing axle have been used. The older basic design was used on 
the early Mercedes-Benz 300SL and is shown diagrammatically in Figure 7.21. 
The roll centre height is very high, higher even than that of the rigid beam axle. 
The 300SL Mercedes-Benz, fitted with unequal-length wishbones at the front 
(see Figure 7.15) and with normal swing-axle at the rear, thus had a very low roll 
centre at the front and a very high one at the rear. Despite the use of an anti-roll 
bar at the front the two systems were unbalanced and showed a marked 
roll-oversteer effect. The Daimler-Benz Company were not satisfied with the 
behaviour of the 300SL and evolved a low-pivot swing-axle system as shown in 
Figure 7.22. The rigid swinging arms were pivoted at a point below the 
rear-drive casing, thus giving a low roll centre to the sprung mass. To increase 
the roll resistance at the rear a ‘compensating spring’ is provided. 

Fig. 7.22 

Roll resistance 

When the sprung mass of a car rolls under the action of a side force, a 
restraining couple tending to pull the whole sprung mass back towards its 
normal upright position is exerted by the suspension system. This is illustrated 
for the cases of a parallel wishbone system and a De Dion system in Figure 7.23. 
Let us take the case of a car for which the sprung mass, W s , is acted upon by a 
centrifugal force of n gravities. The side force will be and the rolling 
moment in the first case will be pW s hi and in the second, nW s h 2 . This rolling 
moment is obviously greater in the first case. For a medium-sized sports car, hi 
might be about 15 in. and h 2 could be as little as 3 in. From this we see that case 
(a) in Figure 7.23 is at a disadvantage and will give large roll angles at speeds on 
corners. This lack of roll stiffness is usually compensated by the provision of an 
anti-roll bar at the front. 


An intriguing phenomenon is the trick exhibited by some cars of lifting a wheel 
when cornering fast. Sometimes it is a front wheel, sometimes a rear. The effect 
is a little disturbing to the onlooker in all cases. 

When a mixed suspension system is used (with a high roll centre at one end 
and a low one at the other) the roll angle for a given rolling moment is roughly a 

Road-Holding 137 

Fig. 7.23 (a) Rolling moment with I.F.S. (b) Rolling moment with De Dion axle. 

mean of the angles that would be given by (a) a high roll centre at both ends and 
( b ) a low roll centre at both ends. A typical arrangement might be parallel 
wishbones at the front and a De Dion axle at the rear. This would give zero roll 
contre height at the front and about 12 in. at the rear. The approximate height 
of the roll axis at the centre will be 6 in. and the side load will act at a height of 
about 15 in. (the C.G. of the sprung mass). The arm of the rolling moment at the 
centre of the car will therefore be 15 minus 6 in. = 9 in. This gives a rolling moment 
approximately three times that given when a beam axle is used at both ends. 

This moment of magnitude 1^x9 is resisted by both front and rear springs 
and the resulting roll angle will be much greater than with the rigid-axle sports 
car. The spring deflections will also be greater and if these exceed a certain limit 
the load exerted by the inner spring at either front or rear will exceed the weight 
of the unsprung parts and this particular wheel will lift. It is customary to make 
the rear springs slightly stiffer than the front. In such cases it is the rear wheels 
that exert the greater portion of the restraining force and it is therefore a rear 
wheel that lifts first. 

This can be prevented in three ways. The front spring rate can be increased; 
the rear spring rate can be decreased; an anti-roll bar or roll-stabilizer can be 
fitted at the front. The latter method has the advantage that the front spring 
stiffness to many types of wheel disturbance is not increased, the bar only being 
fully in torsion when one front wheel is in the maximum ‘bump’ position and 
the other in the maximum ‘re-bound’ position. This is, of course, the position 
approached when cornering near the limit. The roll-stabilizer is a simple torsion 
bar, cranked at both ends and lying transversely across the frame. The outer 
cranked ends are attached to the wishbones at suitable points and as the outer 
wishbones rise and the inner ones fall under the rolling action induced by 
cornering, the ends of the bar are twisted in opposite directions. 

Occasionally, but only when an unusual combination of springing systems is 
used, too much of the overturning moment is taken by the front suspension and 
it is one of the front wheels that lifts. 

138 The Sports Car 

It is not suggested that the action of lifting a wheel is in itself dangerous, since 
most of the load is carried by the outside wheels when an independently sprung car 
is cornering near the limit. Anything we can do to raise the roll centres at both ends 
will make the inner wheels do a little more work in a corner and the slip angles on the 
outer wheels will be appreciably reduced. The phenomenon of wheel-lifting is a 
modern one. The older sports cars with beam axle front and rear and a high roll 
centre at both ends always skidded before the roll angle was at all critical. 

The centre of gravity of the sprung mass 

The importance of a low centre of gravity for the sprung mass of a sports car did 
not escape the attention of the early sports car designers, but there were those 
that hinted darkly at the dangers of too low a C.G. With the right CG. height 
there would be a warning before the tyres lost their grip of the road surface 
(always the rear end in those days); with the C.G. too low the breakaway would 
be sudden and with no warning the driver might be too late to take steering 
correction. We now realize it was the absence of roll on corners when the C.G. 
of the sprung mass coincided with the roll centre that made such designs 
dangerous. Without roll the driver had no real indication of the centrifugal 
force he was asking the tyres to resist. The degree of indication was even less if 
the bucket seat had a good grip on his back. In these old cars the roll centres at 
both front and rear were high, being fairly close to wheel-centre height. Today 
with independent suspension at both front and rear, roll centres have been 
drastically lowered. As a consequence the designer is no longer afraid of a low 
centre of gravity. On the contrary he is for ever seeking ways to lower it. 

The rear-engined sports car 

The popularity of the rear-engined sports/racing car in recent years has led some 
people to make the claim that the rear-engined car can negotiate a given corner 
faster than a front-engined car. The author can find no convincing evidence to 
support this claim and, as demonstrated earlier in this chapter, the only layout 
with a fundamental advantage when accelerating in a bend is the front-wheel- 
drive car. 

The popularity of the rear-engined car for circuit racing is well founded since 
there are several important advantages to be gained. These are: 

(1) There is no propeller shaft in the way and the driver can be seated as low 
as possible, thus reducing the overall height and the frontal area of the car. 
(Only in a competition type of car with a ‘one and a half seater’ body does the 
propeller shaft tunnel intrude. On a sensible width of body there is ample space 
on either side of the tunnel for the seats.) 

(2) There is a natural weight bias on the driven wheels since both engine and 
transmission is at this end. This reduces wheel spin. (This applies equally well to 
the front-wheel-drive car.) 

(3) Weight transfer when accelerating increases the load on the rear wheels 
and improves traction. (In the f.w.d. car there is weight transfer away from the 
driven wheels.) 

Road-Holding 139 

(4) Weight transfer to the front during braking tends to equalize the wheel 
loads and a brake distribution approaching 50/50 is possible. (On the f.w.d. 
front-engined car, the static weight bias to the front is accentuated and the front 
brakes are called upon to contribute most of the braking effort. This arrange¬ 
ment is more likely to lead to overheated front brakes.) 

(5) When the front wheels are steered and the rear wheels are driven the 
engineering design problems associated with steering and driving the same 
wheels do not arise. It must be admitted that the latest designs of constant 
velocity universal joint have worked quite successfully on many small and 
medium powered saloon cars. Time alone will tell if equal success has been 
achieved when torques of 400 to 500 lb ft have to be transmitted through the 
front wheels. 

(6) A good aerodynamic nose shape is easy to achieve on the body of a 
rear-engined car. 

It is sometimes thought that a rear-engined car, with its inevitable weight bias 
to the rear, must of necessity be an oversteering car and perhaps even an unsafe 
car. This is not necessarily so. With the right combination of all the factors that 
contribute to understeer there is seldom any difficulty in taking the correct 
action to prevent oversteer. We have already mentioned the four major factors 
that influence under steer. The General Motors Research Laboratories have 
demonstrated that there are 27 interacting design parameters involved and now 
use this information in a computer programme to help them design the desired 
handling characteristics into their latest cars while still on the drawing board. 

It is argued by some rear-engined car experts (Jim Hall who designed the 
Chaparrals is one of these) that a degree of oversteer when a car enters a corner 
is desirable. This is debatable, but the desirability of a gradual reduction in this 
oversteer as the centrifugal force in the corner increases seems to be generally 
accepted. Extreme cases of oversteer are now ancient history. To complete the 
final stages of a corner with full opposite lock is exhilarating to the driver and 
thrilling to the more naive spectators, but it is never the fastest way through a 

The suspension 

'Ye who borne about in chariots and sedans 
Know no fatigue but that of idleness.' 



It is always good to start at the beginning. Why then do we have springs? We 
could say we have them to isolate the occupants of the car from the up and down 
movements of the wheels as they run over an uneven surface. But is the word 
‘isolation’ correct? If the road wheels pass over a 2-in. step in the road surface 
what happens to the passengers — do they rise 1 in., 2 in. or 4 in.? From 
experience we can say that they rise at least 2 in., often more. This can hardly be 
called ‘isolation’. 

Taking this as a concrete example, let us consider what happens when a car 
travelling at speed encounters a step in the road surface produced by running on 
to a newly tarred surface. The size of the step will be taken as 2 in. In Figure 8.1 
the simplified car is seen to be suspended on trailing links and coil springs. The 
immediate effect of the front wheels mounting the ridge is a compression of the 

Path of point A under 
action of partially 
damped spring 

Fig. 8.1 Spring behaviour when negotiating step in road surface. 

The Suspension 141 

front springs by exactly 2 in. This is followed immediately by the recovery of the 
spring from this sudden compression. In the case of a partially damped system 
this spring recovery will lift the front of the car by an amount greater than 2 in. 
since the spring will overshoot its normal ride position. It is therefore seen that 
by fitting springs to the car we have magnified the road disturbance. Let us see, 
now, what happens if we dispense with springs altogether. The change of level 
of the front end of the car as the front wheels encountered the step would in this 
case be instantaneous. If we neglect the spring of the chassis and the upholstery 
and the natural spring of the driver’s anatomy, the change in the driver’s 
position is also instantaneous. This means an upward acceleration of infinity ! 
This is of course impossible, but the practical implication, when allowance is 
made for the spring in the cushion, etc., is that the driver would be subjected to 
a most violent upward acceleration of several gravities which would jar his spine 
and project him out of his seat. Let us calculate now what the acceleration 
would be with a typical suspension system. A loaded spring always oscillates at a 
certain ‘natural’ frequency, depending, amongst other things, upon the size of 
the load. When all damping is neglected (including any damping inherent in the 
spring itself, such as the interleaf friction in a leaf spring) it oscillates with a 
simple harmonic motion. The time of a complete oscillation, in seconds, is 

T=2ns/(d/a) (1) 

where d = the initial displacement of the spring under its static load W. 

This displacement is measured in feet 

a = the maximum acceleration, ft/sec 2 . This will occur at the upper and lower 
extremities of the spring travel. 

The frequency of a typical sprung mass can be taken as 90 oscillations per minute, 
i.e. T=0.67 

Maximum acceleration = (2ir/T) 2 X displacement 
= (2 tt/7 t ) 2 X(2/12) 

= 14.7 ft/sec 2 
= 0.45 gravities. 

It should be pointed out that T is dependent upon the load W acting on the 
spring. With the spring oscillating freely under the weight of the wheel and 
attendant unsprung parts, T will be much lower (i.e. the frequency will be 
higher) than when oscillating under the much greater weight of the sprung 
portion of the car. 

The above acceleration of 0.45 G is the maximum to which the driver and 
passengers will be subjected and will occur at the top of ‘the bounce’, at a time 
interval of 0.33 seconds after the front wheels strike the step. This maximum 
acceleration and time interval will be the same whatever the speed of the car, if 
the springs are undamped. Shock absorbers modify the spring behaviour in such 
a way that the upward acceleration is increased the higher the speed. This will be 
discussed later. 

142 The Sports Car 

When the rear wheels reach the step the behaviour of the rear sprung mass will 
be very much the same as described above. If the spring frequency is lower, i.e. 
T r is greater than 7/, the acceleration will be reduced in the ratio (Tf/T r ) 2 . We 
therefore see how desirable it is to have a low natural frequency for the springs, 
whether they be at the front or rear. Unfortunately a low frequency spring is a 
‘soft’ spring. By ‘soft’ spring we mean one with a low rate. The spring rate being 
the load in lb required to produce a deflection of 1 in. This means that the initial 
deflection of the spring under its static load will be greater with a soft spring. 
Table 8.1 shows how the spring frequency varies with initial spring deflection. 

Table 8.1 Number of oscillations per minute for various initial spring deflections 
and the maximum accelerations given by a two-inch step in road surface 

Initial spring 
deflection in. 

No. of oscillations 
per minute 


Time of oscillation 

Max. acceleration 
given by two-inch 
step , gravities 

































An initial deflection greater than 6 in. is seldom possible in practice. With a 
6-in. initial deflection the designed range between spring stops would be about 8 
in. and this would be difficult to arrange in a low-built sports car. This, 
therefore, represents the absolute limit. The upward acceleration given by a 
2-in. step would, with this spring frequency of 77, be as low as a third of a 
gravity, an upward acceleration so low that the occupants would experience no 
discomfort, especially when the spring impulse is conveyed to the person 
through a well-sprung seat. A lower frequency than 77 would be undesirable in a 
sports car, however, since the roll on corners would be excessive and the general 
behaviour much too spongy. In practice, with independent front suspension at 
the front and semi-elliptic springs at the rear, a frequency of about 80 is used at 
the front and 90-100 at the rear. 

Now let us consider what takes place when a car passes over a ripple or similar 
depression in the road surface, as shown in Figure 8.2. For simplicity the 
following assumptions are made: 

(a) the ripple affects both front wheels simultaneously and to the same extent; 

(b) the ripple is sinusoidal in form; 

(c) The car mass is divided into two parts, Wf and W r , which are capable of 
up and down motion independently. 

The Suspension 143 

Fig. 8.2 Spring behaviour over ripple in road surface. 

From the car designer’s viewpoint we are interested in two things: 

(1) How big must the depression be before a wheel loses contact with the road 
surface? (This is a most important factor, affecting as it does the 
road-holding of the car on rough roads.) 

(2) What accelerations are produced on the sprung masses, Wf and W r l 
Let us take the natural front spring frequency, under the action of the sprung 

mass at the front, Wf, to be l/Tf, i.e. the oscillation period = 7/. 

Under the action of the unsprung mass, i.e. the front wheel assembly, the 
natural period, tf, therefore becomes Tj\/(wf/Wf ), where wf is the mass of the 
front wheel assembly. 

If the horizontal span of the ripple is s ft and the car velocity is v ft/sec, the 
time t, of the half oscillation induced by the ripple is s/v secs. 

Contact of the tyre with the profile of the ripple is therefore maintained so 
long as s/v is greater than 

II 1(^l\ 

2 V (w// 

i.e. in the limiting case, 

The minimum ripple length is therefore, 

If we take typical values of 

Tf = 0.67 sec. 

Wf= 100 lb 
Wf = 500 lb 

v = 44 ft/sec. (30 m.p.h.) 

144 The Sports Car 

44X0.67 / /100 \ 

2 V \500 / 

= 6.6 ft. 

At 60 m.p.h. the limiting length for contact will be 13.2 ft. (These calculations 
neglect such effects as the expansion of the tyre contact patch as the load is 
removed from the tyre and the movement of the sprung mass.) 

The greatest depth of ripple that can be negotiated without loss of tyre contact 
is largely governed by the amount that the spring is already compressed under 
the weight of the car, i.e. the initial static spring deflection. If the initial 
deflection is 4 in. the load on the spring will be zero at the bottom of a 4-in 
depression of the ‘correct’ profile. The inertia of the unsprung mass will tend to 
carry the wheel to a slightly greater depth. For example, with no spring 
damping, if the wheel assembly weighs one-quarter of the sprung mass, the 
wheel will continue downwards for one-quarter of the initial deflection before 
starting to rebound. We can therefore say, as a rough guide, that the 
approximate maximum depth or height of ripple for limiting road contact is 
given by 

where d = the initial static spring deflection. 

Using equations (2) and (3) Table 8.2 shows the effect of changes of Tf and 
(Wf / Wf) on the ‘ripple rideability’ of a car travelling at 60 m.p.h. A value of 0.4 
for Tf represents the case of the hard sprung vintage-type sports car. A value of 
0.8 is approximately that of the Jaguar XJ-S front suspension. A value of 
one-quarter for w// Wf is typical for the front axle of a vintage car and a value 
of one-eighth is perhaps too high for the best type of modern front suspension 
but could be achieved by a modern independent rear suspension system. 

We must not lose sight of the fact that several simplifying assumptions have 
been made. One factor, for instance, which we have neglected is the movement 
of the sprung mass itself. This will lag behind the wheel movement and at a 

Table 8.2 Limiting ripple dimensions for tyre adhesion at 60 m.p.h. 

Tf Wf Limiting ripple dimensions at 60 m.p.h. 

Wf - 

seconds Length, s, ft Depth, x, in. 

















The Suspension 145 

speed of 60 m.p.h. the sprung mass will have hardly started to move downwards 
before the wheel has traversed the ripple. We therefore would not expect this 
particular assumption to invalidate the above figures. The natural frequency of 
the ‘spring’ in the tyre is much higher than that of the car springs; the static 
deflection is also relatively small and the overall effect of the tyres on the 
accuracy of the above calculations cannot be great. 

Table 8.2 shows that when we are designing our suspension system with a view 
to maintaining tyre contact with the road at all times, the most important factor 
is spring frequency, l/T, which should be as low as possible. But what controls 
the practical lower limit for spring frequency? 

The limit to spring frequency is obviously set by the mass of the wheel and the 
suspension unit, the unsprung mass. A low frequency means a low spring rate. 
If we were to use a low spring rate with a relatively high unsprung mass, the 
vertical wheel movements under the action of normal road undulations would 
be excessive. We therefore see that Tf is really controlled by wf/Wf and that so 
far as this particular property of ‘ripple rideability’ is concerned this mass ratio 
must be as low as possible, i.e. for a given weight of car the unsprung weight 
should be as small as possible. This explains the steady decrease in wheel 
dimensions during the last thirty years. A small sports car in 1935, such as the 
M.G. Magnette, had 4.75 x 19 in. tyres. Its modern counterpart, the M.G. 
Midget has 145x13 tyres, giving a much lighter wheel. Unfortunately the 
advent of the low profile tyre fitted to a very wide wheel on the higher-powered 
sports cars has increased the weight of rubber in the tyre. To offset this it is now 
normal to use wheels of light alloy. 

We have considered the effects of a ripple on wheel adhesion; it now remains 
to consider the effects of such a ripple on the sprung mass. What is the 
magnitude of the downward acceleration on the car itself and on what does it 
depend? This will first be considered for the general case where the spring 
frequency, 1/7/, does not coincide with the natural frequency of the undula¬ 
tions. In this case, from Newton’s First Law, we see that the force P imparted by 
the wheel to the base of the spring will be transmitted through the spring to the 
sprung mass of the car itself. The acceleration of the wheel will be much higher 
than that of the sprung mass in the inverse ratio of the masses. 

Force = Mass X Acceleration 

Wf Wf 

where af - the acceleration of the unsprung mass 
Af = the acceleration of the sprung mass 
W f 

The vertical movements of the sprung mass will also be reduced in this same 
ratio, Wf/Wf. 

146 The Sports Car 

For maximum passenger comfort and, what is of equal importance in the case 
of a sports car, minimum stress to sprung parts of chassis and body, it is 
necessary to keep Wf/ Wf as low as possible. 

Although called here ‘the general case’, this really only applies to a limited 
range of speeds over any given series of undulations in the road surface. In 
practice the vertical movements of the sprung mass can vary from as little as 
zero to as high as x, the depth of the depression itself. The first case, where the 
body movement is zero is well demonstrated by the corrugated dust roads so 
familiar to the motorists of Africa and Australia. If the frequency of the road 
undulations coincides with the natural frequency of the springing system when 
excited by the unsprung mass, the wheels and the lower ends of the springs will 
oscillate without transmitting any force to the upper ends of the springs and the 
car will travel along the corrugated road without a single tremor reaching the 
passengers. This ideal case is never quite achieved. Even where American sedans 
abound, spring frequencies of different models are not exactly alike and the 
corrugations are the average bounce period of the average car travelling at the 
average cruising speed. A vintage-type car would have to travel at about twice 
the cruising speed of the modern cars to ‘get into the groove’ and it would 
probably wreck itself before it got there! 

If the length of the ripple from peak to bottom of depression is 10 feet a car 
with an average spring frequency of 80 per minute and a w// Wf ratio of Vs 
would travel most comfortable at a velocity of 59 ft per sec. or 40 m.p.h. (This 
comes directly from equation 2.) The initial corrugations in the dust roads were 
of course started by the natural spring oscillations of the first cars to go over 
them. Subsequent drivers chose the most comfortable cruising speeds to suit 
their particular spring frequencies and this served to dig the corrugations deeper 
and deeper. 

The other extreme, that which gives the greatest movement to the body, 
occurs when the frequency of the road undulations coincides with the natural 
frequency of the springs when excited by the sprung mass. This frequency is of 
course much lower than the natural frequency of wheel bounce, as discussed in 
the previous paragraph. 

In this case if 

s = k(vTf) 

s = 30 ft and Tf = 0.75 

v = 80 ft per sec. 

=* 55 m.p.h. 

The natural reaction of a driver when he encounters a series of undulations of 
this character at the wrong speed is to take his foot off the accelerator. In some 
cases a more effective remedy would be to open the throttle more. 


One other common suspension phenomenon remains, that of pitching. Rolling 
may be regarded as a suspension phenomenon, but this has already been 
discussed in the previous chapter under the general question of ‘Road-holding’. 

The Suspension 147 

In Figure 8.2 we treated our basic car as if it were two entirely separate parts, 
the front and the rear. This helped us to consider the problems of ‘bounce’ 
entirely divorced from that of ‘pitch’. We cannot ignore pitching, however, 
since it happens to all cars to a greater or lesser extent. Pitching is best illustrated 
by jumping up and down on the bumpers of a modern car. The body will 
oscillate about the centre of gravity of the sprung mass. The time for a complete 
oscillation depends upon what is called ‘the radius of gyration’ (see Figure 8.3). 
This is in effect the radius at which the sprung mass could be considered to be 
concentrated in a single arc and still oscillate at the same frequency as the actual 

The bouncing period from equation (1) is 


The pitching period is given by 

'- 2 ’V(i X 

From this we see that if the C.G. is symmetrically placed, i.e. /, = l 2 and K, 
the radius of gyration = h, 

T= t 

In other words, the frequencies of pitching and bouncing are equal. 

The customary layout of chassis components on a modern sports car gives an 
approximate value of fP/Qi x l 2 ) = 0.9. For the sports car of the thirties it was 
often as low as 0.7. 

Pitching is, of course, induced by a depression or bump. As the front wheels 
drop into a pothole, the dip at the front end causes a rise at the rear and if this 
should coincide with the passage of the rear wheels over a bump in the road 
surface, the impulses, being additive, will produce bad pitching. It is apparent 
from this that a wavy road surface with a distance from bump to depression 

148 The Sports Car 

exactly equal to the wheel base will produce resonance pitching. If at the same 
time the car is driven over this road at such a speed as to produce resonance 
bouncing, interference between the two frequencies of bounce and pitch, 1 IT 
and 1 It, will produce a slow ‘beat’. This objectional type of beat is occasionally 
encountered on the older arterial roads, where subsidence at the joints of the 
concrete sections has led to a wave formation. If T = t, the beat period becomes 
infinite, i.e. there is no beat. 

On some of the older cars, with the engine set well back behind the front 
wheels and no overhang at the rear x l 2 ) was sometimes as low as 0.7 and 
‘heterodyning’ of these two simple harmonic motions would lead to the most 
disconcerting leaps in the air on occasions. 

Summing up our discoveries so far, we would say that for good springing our 
fundamental needs are: 

(a) a spring frequency as low as possible; 

(b) unsprung masses, at both front and rear, as low as possible; 

(c) disposition of the front and rear masses such that the period in pitch is as 
near as possible to that in bounce. 

Independent suspension 

All this has led us logically to the main reason for the introduction of 
independent springing about forty years ago. There were certain difficulties 
about its adoption for the rear wheels but almost all cars had changed to I.F.S. 
by the time the post-war models appeared. Independent springing, at front or 
rear, gives the lightest suspension system at present known. Moreover, the 
disappearance of the front axle has allowed the designer to move the engine 
forward, with obvious advantages in the case of the family saloon where 

Table 8.3 Unsprung weights for front suspensions of 900 Kg (2000 lb) sports car 

Suspension arrangement Effective unsprung Wf 

weight per wheel 
Kg (lb) 

Wishbone or trailing link, with torsion bar or coil spring 36 (80) 6.2 

Wishbone with transverse laminated spring 41 (90) 5.6 

Front axle, with half-elliptic springs 54(120) 4.2 

passenger space is at a premium. A secondary advantage from this change which 
is of great importance to the sports car designer is the increase in the radius of 
gyration of the sprung mass, thus bringing the pitching period closer to the 
bouncing period. 

Table 8.3 gives representative figures for the unsprung weights on a typical 
sports car of 2,000 lb kerb weight when different types of front suspension are 
used. Table 8.4 gives similar values for different designs of rear suspension. The 
De Dion axle is also included. It is not of course independent, since the 

The Suspension 149 

Table 8.4 Unsprung weight for rear suspensions of 900 kg (2000 lb) sports car 

Suspension arrangement 

Effective unsprung 
weight per wheel 
kg (lb) 



I.R.S. with torsion bar or coil spring; inboard brakes 




I.R.S. with torsion bar or coil spring; outbroad brakes 41 



Swing axle with coil springs 




De Dion axle with torsion bar or coil spring; 
inboard brakes 




De Dion axle with torsion bar or coil spring; 
outboard brakes 




Live axle with trailing links and coil spring or to 
or torsion bar 




Live axle, half elliptic springs 




wheels are interconnected, but it is an attractive design, being both light and 

When we look at these two tables it is apparent that independent suspension 
offers the greatest saving in weight at the rear; a reduction of about one-third at 
the front and about one-half at the rear. Why then, we might ask, is I.R.S. not 
used as much as I.F.S.? Briefly, we can say that the complications which arise 
when we have to design an independent system for wheels that must be driven 
through two constant-velocity universal joints lead inevitably to more expense. 
Car manufacturers who wish to stay in business do not sanction expensive 
changes lightly. While I.F.S. offered so many advantages besides improved 
comfort, solving the old problems of tramp and shimmy, there are really only 
two major advantages to be gained by a change to I.R.S., a reduction in 
unsprung weight and an absence of transverse weight transfer during accelera¬ 
tion. This second advantage shows up most clearly on a sports car with a big 

Fig. 8.4 Transfer of weight from right wheel to left wheel on a 
live rear axle during acceleration (1962 Corvette). 

150 The Sports Car 

engine. Figure 8.4 shows how the static loads of 750 lb on each rear wheel of the 
non-independently sprung rear axle of the 1962 Corvette change to 917 lb on the 
left rear wheel and 583 lb on the right rear when the rear axle twists under the 
torque reaction as maximum engine torque is applied in bottom gear. Even on a 
good surface this results in bad wheel spin and loss of acceleration for the first 
ten yards or so. With a chassis-mounted rear-drive box on an I.R.S. installation 
the torque reaction is taken by the much greater sprung mass and results in a 
negligible weight transfer. 

Shimmy and tramp 

With the rigid front axle it was necessary to limit wheel movement with stiff 
springs and firm dampers or the tilting of the axle and both wheels as one wheel 
passed over a bump could introduce powerful gyroscopic forces that could cause 
the wheels to flap from side to side. The word ‘shimmy* came from a popular 
dance of the ’twenties. The word ‘tramp* was used to describe a rocking motion 
of the front axle, which in its most violent form would cause the front wheels to 
lift off the ground and ‘patter* or ‘tramp* on the road surface. There are several 
ways in which tramp or patter can be started on a beam front axle, but the 
primary cause of all the steering bothers was the high inertia of the whole front 
axle assembly with its heavy axle, brakes and wheels. The secondary cause was 
the interaction between wheel movements from side to side (see Figure 8.5). 

Enough has been said to show why the designer was glad to be well rid of the 
front axle. The live rear axle, however, seems more or less innocuous. The 
steering effect of any bouncing or waggling it may do is not very noticeable and 
any sideways displacements may be discernible to rear seat passengers but these 
are a rarity in the case of the sports car. The live rear axle is not quite as 
innocuous as it seems. The unsprung weight is twice the weight it need be (see 
Table 8.4). This we have shown is undesirable if we wish to maintain contact of 
the rear wheels on the road surface on a bad road. The sports car with a live rear 
axle will therefore corner and brake just as well as the equivalent I.R.S. sports car 
so long as the road surface is as good as a billiard table. As soon as bad ripples 
appear in the road the independently sprung wheel comes into its own. Up to a 
point, if the ripples are small and within the road-holding capacity of the w./W. 
ratio of the beam-axle car, there will still be little difference in the cornering 

The Suspension 151 

power, but with bad ripples the rigid axle car will corner at the limit in a 
succession of sideways hops, while the I.R.S. car performs a smooth well- 
controlled drift through the corner. Experience has confirmed that the De Dion 
axle, which is very little heavier than true I.R.S., gives a rear-end ride 
approaching that of the latter. 

In recent years there have been many successful designs of I.R.S. on 
production sports cars of moderate price, the Corvette Stingray, the Datsun 
280Z, the Lotus Elite and Esprit and at a lower price level still, the TVR and the 
whole range of Triumph models from TR2 to TR6. The latest TR7, however has 
a live rear axle. 

There are problems associated with the design of a good system of indepen¬ 
dent suspension as an examination of a few modern examples will show. 



Double wishbone front suspension has been used on Jaguar sports cars for 
nearly thirty years. The E Type front suspension is shown in Figure 8.6. It was 
developed from the older XK design and used a fore and aft mounted torsion 
bar, built into the lower wishbone pivot, as the spring. Anti-dive geometry was 
given to the wishbone layout when the V-12 engine was installed in the E Type 

Fig. 8.6 Front suspension on the E Type Jaguar. With the later V-12 engined model 
the front suspension geometry was revised to give an anti-dive effect. 

152 The Sports Car 

Anti-dive geometry is common practice on the softly suspended American 
cars. The method usually used is to cant the upper wishbone pivot axis upwards 
at the front and the lower pivot downwards. When braking, the kingposts react 
to the braking forces by attempting to rotate. This reaction is resisted by the 
wishbones and with the correct anti-dive angles an upward force is applied to the 
front of the chassis that is exactly equal to the inertia force that would other¬ 
wise make the front end of the car dip or ‘dive’. Unfortunately, rotation of 
the kingposts also produces undesirable changes in castor angle. Jaguar 

Fig. 8.7 Rear suspension on E Type Jaguar. 

The Suspension 153 

compromise by providing sufficient anti-dive to offset 50 per cent of the dive. This 
is given with an angle oil Vi degrees between upper and lower wishbone axes. The 
front suspension on the latest Jaguar, the SJ-S is similar in layout to that used 
on the V-12 engined E Type. Anti-dive geometry is again provided, but in this 
model the torsion bar is replaced by the more conventional coil spring. 

The rear suspension on the XJ-S is almost identical to the well-tried IRS 
design used on the E Type (see Figure 8.7). Twin coil springs are placed in front 
and behind the lower suspension arms (only one can be seen in this cross- 
section). The universally-jointed drive shaft acts as the upper transverse 
suspension arm. Radius arms (not visible in the drawing) are attached to the hub 
carrier and serve to resist braking and accelerating torques. 

Fig. 8.8 Camber angle change under roll with swing-axle, 
trailing link and semi-trailing link suspensions. 

Fig. 8.9 Swing-axle suspension is good - up to a point. ‘Jacking action’ caused by lifitng the throttle in the middle of 
a tight corner. {Motor photograph). 

The Suspension 155 


The older Porsche trailing link suspension was very effective with narrow tyres 
with rounded sections, but a new system had to be devised as tyre sections 
became wider and flatter. Both front and rear suspensions on the modern 
Porsche are of the type described as ‘semi-trailing link’. Many car designers are 
turning to this system in their new models. It is in effect a cross between trailing 
link, the old Porsche system and swing-axle, the system that Ferdinand Porsche 
Senior tried to tame before the war. Figure 8.8 serves to illustrate the problems 
associated with the two systems. 

With a trailing link the wheel is constrained to move parallel to the body side. 
For single wheel movements that produce negligible body roll the full area of the 
contact patch is maintained. If the body rolls, however, the wheels change their 
camber angles to the same degree as the body roll. This of course is unacceptable 
with modern low profile tyres. Roll is inevitable since the roll centre is at ground 
level with a trailing link suspension. The roll-centre is much higher with 
swing-axle suspension and the tendency to roll is therefore much lower. Even so, 
when cornering near the limit the dreaded ‘jacking action’ can occur as 
demonstrated in Figure 8.9. When this occurs the body is jacked upwards under 
the action of centrifugal force and pivots about the only available flexible 
member in the drive system, the inboard universal joint on the heavily loaded 
side of the car. This can cause a dangerous loss of wheel adhesion since the 
camber change is so drastic. Camber change is also quite large for single-wheel 

Fig. 8.10 Front suspension on Porsche Type 911. 

156 The Sports Car 

bumps. The weakness of the swing-axle lies in the inherently short swing-axle 
length. Since the rear drive unit must occupy some of the available space the 
swing-axle length cannot be much more than 40 per cent of the total track width 
on a rear-drive design. Early Allards used swing-axle suspension at the front, 
but even with 50 per cent of the track available the cars still suffered from 
excessive camber change. 

The semi-trailing link suspension extends the effective swing-axle length and 
thus reduces camber changes. Wheel movement is constrained to pivot about an 
inclined axis and the effective swing-axle length depends upon the angle chosen 
by the designer. When used on the driven end of the car it is necessary to provide 
a coupling with sliding splines on the drive shaft to accommodate the changes in 
drive shaft length introduced by the new pivots. This system can be given 
sufficient swing-axle length to remove all danger of jacking. The semi-trailing 
angle is chosen to strike the best compromise between the fully trailing link and 
the swing-axle. Since the trailing link gives negative wheel camber (the top of the 
wheel leaning outwards) on the outer wheel in a bend and the swing-axle gives 
positive wheel camber (except under jacking conditions) an angle close to 45 
degrees will make the outer wheel stay very close to upright under most 
conditions of roll. This angle can be varied at the will of the designer to suit his 
particular requirements. 

Typical examples of Porsche suspensions are given in Figures 8.10 and 8.11. 
This design is common, apart from small detail changes, to the Type 911, the 
Carrera and the Turbo. Damper struts and semi-trailing links are used at both 

Fig. 8.11 Rear suspension on Porsche Type 911. 

The Suspension 157 

front and rear. The front suspension comprises a lower wishbone made from 
rectangular section tubes welded to a circular section tube housing the torsion 
bar. The damper strut is the only other suspension link. It must be one of the 
lightest suspension systems in use today. The rear suspension is similar in 
layout. In both cases the upper mounting of the MacPherson strut carries a 
rubber bush to permit a small degree of angular change as the wheel rises and 


The Datsun 260Z and 280Z uses what may be called ‘pure MacPherson’ 
suspension at front and rear. Figure 8.12 gives a cross-section of the front 

[ Strut mounting insulator 8 Wheel bearing 

9 Suspension ball joint 

10 Transverse link 

11 Compression rod 

12 Stabilizer bar 

13 F-Tont suspension member 

2 Thrust bearing 

3 Bound bumper rubber 

4 Dust cover 

5 Coil spring 

6 Strut assembly 

7 Front wheel hub 

Fig. 8.12 Front suspension on Datsun 260Z. 

160 The Sports Car 

suspension with a plan view (upper left) to clarify the location of the 
compression rod and the stabiliser bar (parts 11 and 12 in the cross-section). The 
base of the MacPherson strut is located against transverse forces by part 10, the 
transverse link. Vertical wheel movements are controlled by the coil spring, part 
5, and the damper unit incorporated in the strut. Fore and aft location is 
provided by the compression rod. The rear suspension uses a MacPherson stut 
inclined inwards at the upper end. The hub carrier is located by a strong 
transverse link. The drive shaft is universally-jointed at both ends and incor¬ 
porates a sliding spline coupling, since the shaft length varies as the wheel rises 
and falls. The Datsun coupling uses hardened steel balls sliding in splines to give 
a low friction value, even under heavy torque. The coupling is filled with 
lubricant and sealed with a rubber bellow. 

Aston Martin 

Aston Martin use a conventional double wishbone and coil spring system at the 
front. It is in the rear suspension that we find our technical interest since the 
Aston Martin Company have used De Dion rear suspension for many years and 
have been happy to specify a similar system for their new ‘space age’ Lagonda 
which is not yet in production. Count De Dion first used the system at the end of 
the last century, but it was probably Harry Miller who first realised its excellent 
road-holding qualities when he adopted it for his front wheel drive racing cars in 
1924. The De Dion axle uses a light rigid tube to connect the wheel hubs 
together. Separate drive shafts, each provided with two universal joints, take the 
drive from the final drive unit which is mounted on the chassis and is therefore 
not part of the unsprung weight. Since the distance between wheel centres 
remains constant as the wheels move up and down sliding splines are provided 
inside the inner universal joints. 

The De Dion tube is located fore and aft by two parallel links on each side and 
transverse restraint is given by means of a Watt’s linkage. This comprises two 
transverse links attached to the chassis at the outer ends and to a short rocking 
lever at their inner ends. The centre of the short lever is attached to the centre of 
the De Dion tube. This form of parallel motion was the brain-child of the great 
James Watt about two hundred years ago. 

The De Dion system was used successfully in the past on racing cars. The 
control of camber change which is now so important with low profile tyres is 
admirably achieved when cornering hard with this system. Under single wheel 
bump there is some camber change however and this change, while acceptable 
on a sports car, would hardly give acceptable handling on a Grand Prix car with 
very low profile tyres. 


We have called it a shock absorber for so long that the habit is hard to break, 
but we have already demonstrated at the beginning of the chapter that it is the 

The Suspension 161 

springs that are the shock absorbers. The devices used to damp out excessive 
oscillations in the springs are better described as dampers. 

The laminated spring, even when anti-friction liners are used between the 
leaves, is never quite free from a small amount of internal damping, but the coil 
spring and the torsion bar possess no internal damping and if used on a car with 
no dampers at all would result in a car that was completely unsafe on the road 
unless driven at very low speeds, since it would bounce up and down without 
ceasing and would lean alarmingly on corners. 

Hydraulic dampers were originally made to operate in one direction only, i.e. 
when the wheel lifted in passing over a bump. Today they work on rebound too. 
In America the word ‘jounce’ is used instead of ‘bump’. The modern damper is 
usually set to give a greater degree of damping on rebound than on bump. 
Severe damping of the initial bump only results in greater vertical accelerations 
being applied to the sprung mass and it was shown earlier in the chapter that the 
reduction in these accelerations was the primary purpose of the suspension 
system. Our aim then is to damp out as quickly as possible each vertical impulse 
given by the road surface, without increasing unduly the maximum accelerations 
applied to the sprung mass. Perfect damping (neglecting the acceleration 
limitation) would mean that no bump could be followed by a rebound. With 
this there would be no possibility of residual oscillations from one impulse 
overlapping those of a subsequent impulse. Such overlapping impulses, if in 
phase, can lead to extremely high amplitudes with the springs crashing hard on 
their stops. The worst case with an undamped suspension is when the road 
impulses are of a regular character and their frequency coincides with the 
natural frequency of the springs. 

Fig. 8.15 Damping characteristics at different 
frequencies of road ripple. 

162 The Sports Car 

Figure 8.15 shows the effect of various degrees of damping on the vertical 
accelerations experienced by the sprung mass. 

fr = the frequency of the road disturbances, 

fn = the natural frequency of the suspension. 

With completely undamped springs, when f r = f n the acceleration imparted 
to the sprung mass becomes infinite (Curve 1). Fortunately there is always some 
friction present to prevent this case arising. With a small degree of damping 
present, the chassis acceleration is still uncomfortably high when the frequencies 
resonate, but is low at f r /fn ratios of 1.5 or more (Curve 2). Curve 3 represents a 
well damped system. At resonance the vertical accelerations are low enough for 
comfort. At higher ratios of f r /f n , accelerations are higher than with the 
undamped system, but are still below the discomfort level. Curve 4 represents 
the case of an overdamped system. Below and at resonance the accelerations are 
very low but at high ratios of such as are encountered at high speeds, 
spring movement is so restricted by damping that the accelerations imparted to 
the chassis become impossibly high. Spring damping, we see then, is all a matter 
for compromise and the best final setting for any car in any given circumstances 
can only be decided by road-testing. 

Early experiments with hydraulic dampers using simple jets to control the 
resistance to flow were found to give resistance/flow characteristics as shown in 
Figure 8.16. If one large jet was used the flow was turbulent over most of the 
operating range. If several small jets were used the flow through the jets became 
streamlined and the resistance to flow varied directly as the piston velocity. This 

Fig. 8.16 Flow resistance characteristics of jets 
and ride-control valve. 

The Suspension 163 

approached the ideal resistance/flow curve for ‘initial bump’ damping, giving a 
resistance increasing linearly with car velocity for a given size of bump. For 
rebound and subsequent damping strokes a constant resistance is required at all 
car velocities for a given size of bump, since the time of recovery after the initial 
bump and for all subsequent half-oscillations is controlled by the spring 
periodicity, which is constant. For a small initial bump the time of recovery is 
exactly the same as for a large initial bump. The fluid velocity through the jets 
during rebound is therefore directly proportional to the size of the bump and, to 
a large extent, to the amount of energy to be absorbed. If the resistance to oil 
flow varies directly as the vertical velocity we therefore have ideal rebound 
damping. Unfortunately a very large number of small holes would be required 
to achieve laminar flow and the danger of choking with dirt would be great. 

The modern ‘ride-control’ valve is a spring-loaded valve which opens to a 
greater or less extent with increase or decrease of fluid velocity. It gives what is, 
in effect, a variable size jet and by variation in spring strength and port shape a 
wide range of resistance/flow curves can be obtained. The principle is illustrated 




Fig. 8.17 The principle of the ‘ride-control’ or ‘linear’ valve. 

in Figure 8.17. Typical flow characteristics given by this type of valve are given 
in Figure 8.18. The other function of the ride-control valve is to provide a 
measure of compensation for variations in viscosity. An oil with a high viscosity 
index is always used in hydraulic dampers, but the variation in viscosity over the 
working temperature range can still be 2 to 1 and a valve that gives a larger 
effective area as the viscosity increases can help to reduce the ‘hardening’ of the 
dampers in cold weather and the ‘softening up’ as the fluid temperature rises on 
a journey. 

Very high frequency movements (circa 300 reversals per second) of small 
magnitude, such as those induced by travelling over a well-maintained British 
paved road at, say, 40 m.p.h., are largely absorbed by the tyres. A Continental 
pav6 in poor condition will produce larger amplitudes, still of a frequency of 100 
reversals per second or more, and the inertia of the spring-loaded ride-control or 
‘pressure’ valves is too great to allow fluid movement to keep pace with the 
extremely rapid realtively small movements of the suspension. To cope with 


164 The Sports Car 


Fig. 8.18 Typical flow characteristics given by bleed flow and linear 
valve flow. 

Fig. 8.19 ’The De Carbon (Girling Monitube) 

principle and the Woodhead Monotube 

The Suspension 165 

these high frequency low amplitude movements a bleed is provided which 
by-passes the ride-control valve. 

The double-tube damper 

Modern sports cars are fitted with dampers of the direct acting telescopic type. 
These are now made in two patterns. The older, double-tube design, uses two 
chambers in the form of concentric tubes. The inner tube is the working cylinder 
containing the piston and the space between this tube and the larger diameter 
tube is the oil reservoir. The need for this reservoir or recuperation cylinder is 
apparent when we see that the effective area of the piston is greater on the 
down-stroke by an amount equal to the cross-sectional area of the piston rod 
(see Figure 8.20). Liquids are of course almost incompressible. Since the 













Fig. 8.20 Damper details (a) twin-tube, (b) single-tube, (c) double¬ 
acting Belville valve used as ride-control valve. 

working cylinder is full of oil the excess oil pumped through the bleed holes ride 
control valves on the down-stroke must go somewhere. It passes into the 
annular chamber. On the return stroke the flow is reversed. This flow of oil is a 
considerable amount and calls for a large recuperator valve with unrestricted 
passages if aeration is not to take place when riding over rough roads. 

166 The Sports Car 

Aeration can still be a problem in all types of hydraulic damper. Since room 
for expansion is necessary in the recuperation chamber the oil surface is always 
in contact with the air. Over rough roads the oil surges badly, producing 
aeration throughout the bulk of the oil. In the telescopic damper the high flow 
backwards and forwards through the recuperator valve aggravates this condi¬ 
tion. Very few older designs of damper are entirely free from aeration on long 
journeys over rough roads. Careful design can prevent serious loss of damper 
efficiency from this cause. 

A typical modern direct-acting damper of very robust construction is the 
Dutch made Koni, shown in Figure 8.21. The piston rod has two upper drillings 













Fig. 8.21 

Cutaway view of Koni damper, 
showing arrangement of valves 
and passageways. 

that connect radially with a concentric drilling that permits the passage of oil 
into the centre of the rod. The central passage is closed at the base by a 
lock-bolt. The final connection between the upper and lower portions of the 
cylinder is made by two calibrated radial holes on opposite sides of the rod, the 
left one being slightly higher than the right. This slight difference in height is 
part of the Koni system of adjustment, since by turning the adjusting nut so that 
it screws further on to the externally threaded end of the piston rod, first the 
right-hand hole can be gradually covered, then the left-hand. In this way the 
damping action in bump or ‘jounce* is increased. 

The Suspension 167 

When the damper is compressed under bump action the fluid below the piston 
becomes compressed and begins to flow through the calibrated holes. At the 
same time, if the rate at which the damper has been compressed is rapid enough, 
the return valve will also open against the action of the dished control spring. 
Thus for low-speed bumps only the calibrated holes are open for the flow of oil. 
For more rapid movements under bump action the additional passage through 
the return valve is open. Since the fluid displacement below the piston is greater 
than above, an excess of fluid must be passed into the reservoir through the base 

On the rebound stroke the fluid in the upper part of the cylinder is 
compressed. With relatively slow movement all the fluid can be passed through 
the two calibrated radial holes above the adjusting nut. If the rebound speed is 
greater, pressure will build up to a high enough value for the by-pass valve to 
open. This valve is controlled by a conical coil spring seated around the 
adjusting nut. 

To increase the damper settings in both bump and rebound the damper should 
be completely collapsed until the slot in the base of the adjusting nut engages a 
slot in the base valve body. The adjusting nut can now be screwed further into 
the base of the piston. The first half-turn will begin to close off the lower of the 
two radial calibrated holes. Two-and-a-half more turns will completely block 
the first hole and partially block the second. It is recommended that changes of 
no more than one complete turn be made between road tests. 

It should be noted that inward rotation of the adjusting nut also compresses 
the conical coil spring which controls the by-pass valve and thus also increases 
the resistance to fluid flow on rebound. 

The single-tube damper 

The principle used in the Girling monitube damper is shown in Figures 8.19 and 
8.20. This principle was first used on the De Carbon dampers many years ago. 
With modern shaft-sealing techniques very successful applications of this old 
principle have been evolved. The recuperation cylinder is replaced by an 
extension to the working cylinder containing gas under high pressure. This 
chamber is separated from the working cylinder by a free piston made of 
glass-reinforced nylon. Movement of this low-inertia free piston accommodates 
the oil flow caused by the differences in swept volume on the two sides of the 
working piston. Valving for both bump and rebound flow is incorporated in the 

The development engineers at Woodhead-Munroe showed much ingenuity 
when they designed the new generation damper for their company. As an 
example of what is sometimes called ‘lateral thinking’ they solved the problem 
of aeration by designing constant aeration into the system. Their single-tube 
damper is charged with a finely emulsified mixture of oil and gas under 
pressure. This Woodhead Monotube damper shown in Figure 8.22 absorbs 
variations in displacement by compressing or expanding the gas bubbles. Since 
the valving is designed to suit a gas/liquid mixture there is little change in 

168 The Sports Car 

damper effectiveness as the fluid heats up after prolonged use on rough roads. 
The problem of aeration no longer exists since the fluid is always aerated! ‘If 
you can’t lick ’em, join ’em,’ is a dubious policy for a politician but it 
sometimes works for engineers. 

The chassis, frame and body 

'The Admiralty opposes the building of ironclads 
because iron is heavier than wood and will sink.' 

(First Lord of the Admiralty, 1850) 


When Lotus Cars were asked to provide a submarine version of the Elite for the 
latest James Bond film the problem of buoyancy was no doubt a new 
experience, but the normal considerations in the mind of a car designer when 
choosing his materials of construction are strength, stiffness and resistance to 
wear and corrosion. Not all designers are in agreement on the most suitable 
materials. The vintage design in which a steel ladder-type frame supports an ash 
body panelled in aluminium now only survives in a few vintage replicas and in 
that appealing hand-made product of the Morgan Motor Company at Malvern. 
Even they now make certain concessions in the design of the frame to increase 
the torsional stiffness. 

Fig. 9.1 Torsional deflection of frame (much exaggerated). 

170 The Sports Car 

Torsional stiffness 

When one wheel of a car travels over a bump in the road surface the spring 
attached to this wheel is compressed and the spring load is transmitted to the 
frame and the frame is deflected (see Figure 9.1). If the spring rate is clb/in. and 
the deflection produced on the spring by the bump is d inches, the load 
transmitted to the frame is cd lb. If the torsional stiffness of the frame is C 
lb/in., the torsional deflection of the chassis, D, will be cd/C in. Let us consider 
a case of a spring rate of 100 lb/in. and a road wheel deflection of 4 in. If we 
wish to limit the frame deflection to 1/32 in. the torsional stiffness of the frame 
will need to be (100x4) 1/32 = 12,800 lb/in. or 230 Kg/mm. 

To the older motorist with experience of the wood-framed coach-built body, a 
high degree of torsional stiffness is very desirable. The creaking of timber and 
squeaking of joints was all too much in evidence on many of the older cars and 
if one carelessly parked the car with one wheel on the grass verge and the other 
three on the road, it was not unusual to find one of the doors jammed by the 
twisting of the frame. Yet the chassis frame usually had a battleship solidity in 
appearance and weighed at least twice as much as modern designs. We saw in 
Chapter Eight that the most important factor contributing to the smoothness of 
the ride is the weight of the unsprung parts. The change to independent 
suspension at the front and the decrease in wheel sizes all round has reduced the 
unsprung weight by about one-half and roughly doubled the ability of the 
wheels to cling to the ripples in the road surface. In the same manner the vertical 
impulse given to the frame when a car travels over a bump is directly 
proportional to the unsprung weight. This explains, in part, why the older 
non-independently sprung chassis, despite its apparent robustness, suffered 
more under shocks of a rough road than the modern chassis. Nevertheless, this 
is not the whole explanation. The older chassis designer used a lot of metal in his 
frame, but he put it in the wrong places. To understand what is meant by this let 
us consider the stiffness of the simple sheet of metal in Figure 9.2. The torsional 
stiffness of this thin plate is very low, as shown at (a); the resistance to a force 
such as X in (b) is obviously high. If the reader doubts this let him take a piece of 
18 gauge mild steel (about 1/20 in. thick) about 2 in. wide and 6 in. long, clamp 

Fig. 9.2 (a), (b), (c) and (d) Torsional stiffness. 

The Frame and Body 171 

it in a vice and try to push it in the manner of Figure 9.2 (b). The first case (a) we 
shall call ‘flat-plate stiffness’, the second (b) ‘lozenge stiffness’. 

Since welding had not yet become a production technique the vintage car 
designer used riveting and his basic structural material was the rolled steel 
section. The channel section was found to be the most convenient and the 
cheapest (Bugatti, who loved to be different, once used a sandwich section with 
steel plate on the outside and wood in the middle, but the method was 
expensive). The channel section has no lozenge stiffness (see Figure 9.2(c)). Each 
of the three sides twist in the manner of Figure 9.2 (a). However, as soon as we 
close in the fourth side we find that ‘flat-plate’ twisting is now almost 
completely prevented and the manner in which the box section twists is now by 
lozenging. The stiffness from lozenging is so much greater than that from flat 
plate twisting that the latter may be neglected. If we compare two similar 
sections using the same weight of steel, the first of channel section, the second of 
box section, the torsional stiffness of the second will be about 500 times that of 
the first. The torsional stiffness of a similar weight of circular section tubing 
would be of the same order. 

One might ask at this stage: if the designer wanted increased torsional 
stiffness, why did it take him thirty years to realize that the channel section was 
so weak in torsion? The answer is that many designers believed that a certain 
amount of flexibility was required to improve road-holding. 

The flexible chassis contributed in no small way to the effective spring rate to 
single wheel movement, without affecting appreciably the spring rate when both 
front or both rear wheels passed over a bump. In this way wheel adhesion on a 
bumpy corner was much improved since the double bump, affecting both front 
wheels or both rear wheels simultaneously, is not as common as the single 
bump. What then is the price paid for this additional ‘frame springing’? If we 
consider the case of a typical ‘flexible’ frame, we might expect, when a front 
wheel passes over a 4-in. bump at the same time as the opposite rear wheel 
passes over a similar bump, that the frame deflection at each end might be as 
much as l A in. In the case of a beam front axle, this twisting of the frame has no 
effect at all on the camber angle of the wheels, which is controlled entirely by the 
movement of the axle. This movement, however, is severe enough in itself. If the 
bump is 4 in. and the track 48 in., the tangent of camber angle change is 
4/48 = 0.083 and the camber angle change is 4° 42' . This means that with a 
level road camber angle of 2° 30' the camber angles when passing over the bump 
become 7° 12' for one front wheel and —2° 12' for the other. The gyroscopic 
torque from such a camber change will be high and will demand instant steering 
correction from the driver if the car is not to change its course. 

With independent front suspension, however, the camber angle is controlled 
largely by the geometry of the suspension linkages, i.e. if double wishbones are 
used, and the frame does not twist, the designer can choose the ratio of top to 
bottom linkage lengths to give him the camber angles he wishes at all positions 
of bump and rebound. If the frame twists the camber angles change and this is 
one reason why the designer prefers, nay, insists upon, the stiffest possible 

172 The Sports Car 

frame. Having achieved an almost perfect control over such important factors 
as camber and castor angle, he cannot have his plans upset by a frame that twists 
and weaves as the car travels along a rough road. 

The tubular frame 

The use of steel tubing welded into a strong light frame is an attractive solution, 
the expense varying with the complexity of the design. Examples vary, from the 
straight-forward space-frame construction that Lotus used with such success in 
their early sports and racing cars to the spaghetti-like ‘bird-cage’ frame of the 
Maserati Type 61. 

Figure 9.3 shows the Lotus Nineteen frame. Behind the radiator there are two 
tubular bulkheads arranged to be in line with the front and rear arms of the 
front suspension wishbones. Amidships is a scuttle bulkhead, formed from 
sheet steel with suitable lightening holes. It will be noted that diagonal bracing is 
used throughout the frame, with the exception of the two-sheet steel bulkheads, 
which have sufficient depth of section to resist lozenging. The engine bulkhead, 
immediately behind the seats, is a rectangular cross-braced tubular structure. 
The parallel radius arms, used to provide fore-and-aft location for the rear 
wheel hubs, are attached to brackets on the sides of this bulkhead. The upper 
ends of the coil spring/damper units for the rear suspension system are attached 
to brackets at the top corners of the rear bulkhead. The complete frame, with all 
brackets, weighed 32 kg (70 lb). 

Unitary body-chassis construction 

Logically it would seem that the final answer must be a successful merging of the 

The Frame and Body 173 

chassis and body into one unit. This unit would provide the necessary outer 
protective envelope that is normally provided by the body and at the same time 
would be a structural framework to house the engine, transmission, fuel tank, 
seats, etc. Such an integrated unit would be designed to withstand the maximum 
loads imposed by the suspension system and would distribute them to the 
structure in such a way that no undue vibration or noise is transmitted to the 

This type of integrated body/chassis is in common use throughout the world. 
The very complexity of the structure, made up of innumerable welded box 
sections, ribbed panels, arches, door openings, etc., makes it impossible for a 
perfectly engineered structure ever to be achieved. Inevitably, however carefully 
the separate sections that make the whole are designed, some of them will be 
more heavily stressed. Despite this criticism, the inherent superiority of the 
unitary construction over the use of a separate body and chassis is still 
overwhelming. In one aspect only does it still create a problem. With a separate 
body we are free to mount it on special shock-absorbing mountings that help to 
isolate the occupants from road shocks. With a unitary construction this is not 
possible. One solution is to use special rubber mountings where the suspension 
links are attached to the frame. 

A typical unitary design is the Datsun 260Z shown in Figure 9.4. There must 
be at least a hundred different pressings in this structure, yet the pressing of the 
basic components employs very little labour. The major work content has been 
in the welding together of all these components and many large-scale manu¬ 
facturers have now started to automate this operation. This is all in direct 
contrast to the hand-built vintage car involving many hours of panel beating and 
very little welding. Needless to say the modern car is held together by spot welds 
or series of stitch welds. Continuous welding would be far too expensive and 
would create alignment problems from distortion. 

The backbone chassis 

In the Elan the Lotus Company reverted to an older conception, using a welded 
sheet steel backbone to carry a moulded fibre-glass body that is partially 
stressed. The idea is reminiscent of the R Type M.G. chassis which was an 
advanced design for 1935, in which a backbone chassis was supported on four 
independently sprung wheels. The Elan had an immensely stiff backbone (see 
Figure 9.5), the torsional stiffness between the wheel planes being about 700 kg 
m per degree (5000 lb-ft per degree). The centre section, about 60 cm (24 in) 
long, was a welded mild steel box, 15 cm (6 in) wide and 27 cm (11 in) deep. 
Forward of this section the frame forked to pass on each side of the gearbox and 
engine. At the rear, similar forks terminated in rectangular vertical pillars that 
provided upper mountings for the Chapman-struts. 

Since they evolved the Elan, the Lotus engineers have concentrated on 
building increased strength into the fibre-glass body shell and the result has been 
a body/chassis structure that can meet all the US Federal and European safety 
requirements. The Lotus body is now injection moulded, using fibre-glass 

Fig. 9.4 Typical unitary body-chassis construction. The Datsun 260Z sports car. Brc 
several box sections. Note how the simple lesson of Figure 9.2 is applied efl 

The Frame and Body 175 

moulds to take the place of the steel moulds one would use when injection 
moulding small plastic components. Not only is the surface finish as perfect as 
one would wish but the control of section thickness is very precise. Lotus have 
also used a process of paint application in which the mould is sprayed with paint 
before it is filled with fibre-glass mats and injected with resin. The penetration 
of the paint gives a well-bonded effective paint film of about 1 mm (0.040 in). 

On the Elite the bifurcated section at the rear of the chassis carries a 
cross-member which supports the final drive unit, rear suspension mountings 
and fuel tank. The torsional stiffness of this steel backbone alone is 250 kgm per 
degree (1800 lb-ft per degree). The latest design of Lotus fibre-glass body now 
contributes much more to the overall torsional resistance than was the case with 
the Elan. A steel roll-over bar passes across the roof behind the front doors and 
the door locks engage striker-plates fixed to this member. The door locks are 
built into the steel girder that is used to stiffen the door to meet the U.S. Federal 
side impact regulations. An interesting feature of the door lock design to 
prevent ‘bursting’ of the lock under side impact is that the striker-plate is 
designed to bend in such a way that the door lock and striker-plate remain in 
reasonable alignment and do not separate even when the centre of the door has 
been pushed inwards as much as 12 inches. The Federal requirements are that 
the centre of the door should not be pushed inwards more than 12 inches when 
rammed by a pole under a force of 3500 lbf (1600 kg). The Lotus doors meet this 
requirement with a good margin in hand. To meet the 30 m.p.h. frontal crash 
test requirements it was necessary to design the body structure from the bumper 
backwards as a deformable structure. Exceptional strength in the structure, 
would, of course, reduce damage to the car but would subject the occupants to 
dangerous decelerations. In the Federal test front-seated dummies wearing seat 
belts carry G-meters to register the maximum decelerations reached during the 
30 m.p.h. crash test. 

176 The Sports Car 

On all Lotus models a deformable box-section of moulded fibre-glass encloses 
each pop-up headlamp and transmits the crash load from the bumper back¬ 
wards to the double-skinned front wheel arches. Behind these further hollow 
beams carry the crash loads back to the double-skinned door sills and into the 
main structure. The bumpers are constructed of a fibre-glass outer shell filled 
with polyurethane foam. This is necessary to comply with the 5 m.p.h. crash test 
which all bumpers are required to absorb without damage. 


The Elite meets all Federal and European safety requirements with generous margins. 

1 5 mph Federal bumpers (for cars sold in the States only). Car withstands 5 mph barrier 
collision and 5 mph pendulum striker impact without sustaining damage. No contact with 
pendulum except on impact surface - therefore no damage to bodywork. 

2 Roof crush resistance. Maximum allowable deflection of 5 in when subjected to static load 
of IV 2 times vehicle weight. Elite deflected only 3 in. 

3 Side impact resistance. Pole-shaped former pushed into door with body clamped rigidly in 

Requirements Elite Performance 

Average force to deflect door by 6in 2250 lb 3400 lb 

Average force to deflect door by 12 in 3500 lb 4700 lb 

Peak force 4774 lb 8200 lb 

Steel door beam acts in tension for large deflections giving high strength. Door lock hoop is 
fixed to plate attached to roll-over bar. Plate bends (see insert) during deflection of door 
maintaining two halves of lock in alignment and thus integrity of system. 

4 Steering wheel displacement in 30 mph barrier collision. Maximum permissible: 5in. 

For Elite: Ydn. 

5 30 mph crash test; Exceptionally low decelerations and forces on onstrumented dummies 
wearing seatbelts. 

Fig. 9.6 How the Lotus Elite meets U.S. Federal and European safety requirements. 

Figure 9.6 gives a summary of the Federal and European safety requirements. 
The gradual collapse of the fibre-glass box sections in the Lotus body in the 30 
m.p.h. crash test is also illustrated. 

The Frame and Body 111 


There is always a conflict of requirements when designing an automobile body. 
The family motorist is probably the most difficult to please, since he demands a 
large seating capacity and luggage accommodation all within modest overall 
dimensions and at the same time expects the stylist to provide something with 
extravagant sweeping lines. The sports car owner is more interested in perform¬ 
ance and will sacrifice some body space in the interests of aerodynamic 
efficiency. The typical pre-war motor car wasted a lot of power in pushing a 
flat-fronted angular body through the air at a typical top speed of 60 to 70 
m.p.h. The Wa -litre M.G. saloon, post-war design, was typical of these earlier 
cars, with a rather square body, a flat windscreen, square upright radiator and 
separate headlamps mounted between the radiator and the wings. To propel 
such a car at 60 m.p.h. required about 32 b.h.p. at the road wheels. The Porsche 
911 with a body developed by extensive wind tunnel testing requires less than 20 
b.h.p. to propel it at the same speed. To reach a speed of 100 m.p.h. the Wa -litre 
M.G. saloon would have required an engine producing about 150 b.h.p. The 
Porsche requires only half this power to cruise at 100 m.p.h. 

Drag coefficients 

The resistance to forward motion of any body in its passage through the air can 
be considered to consist of two types of drag. Form drag , expressed simply, is 
the work done by the body in pushing the air out of the way. The less 
streamlined the body, the greater the form drag for a given frontal area. A flat 
plate represents the ultimate in poor streamlining since its passage through the 
air at the speeds under consideration produces eddies in the air over a much 
wider area than that of the actual plate. The perfect streamline form only 
disturbs the air in its own vicinity. Friction drag is the resistance to motion set 
up by the layers of air adjacent to the surface of the body. The replacement of 
an older square-type body by a modern streamlined design, while decreasing the 
form drag, will sometimes increase the friction drag, since the total surface of 
the streamlined design tends to be greater. 

The drag coefficient is a useful value by which the aerodynamic efficiencies of 
body shapes can be compared. The appropriate formula in self-consistent units 

F=C d A V 2 

where F is the drag (force units) 

C d is the drag coefficient 
A is the maximum cross-sectional area 
V is the velocity. 

Expressed in the popular motoring units of V in mph, A in square feet and force 
in lbf, the formula becomes: 

C d AV 2 

178 The Sports Car 

The power to overcome this drag is given by: 

C d A.v 3 

Drag horsepower =- 


where C d = drag coefficient (dimensionless) 

A = frontal area in sq ft. 
v - airspeed, m.p.h. 

For an automobile we have another resistance to add, the road resistance. 
This will vary with tyre design, with road surface, with tyre pressure and with 
load. As an indication of the size of this resistance Figure 9.7 is given. Road 



H.P. 20 

0 10 20 30 40 SO 60 70 80 90100110120/30740150160170180790200 

Fig. 9.7 Horsepower used in overcoming road resistance. 

resistance here is given in terms of horsepower per 1,000 lb of car weight. The 
power consumption increases approximately as the square of the speed. Since 
aerodynamic drag horsepower increases as the cube of the speed, the road 
resistance becomes a small percentage of the total power loss at speeds above 
100 m.p.h. 

In Chapter Twelve we will examine the performance of several modern sports 
cars and make estimates of their drag coefficients from their maximum speeds. 
Values of C d are here found to vary between a lower limit of 0.31 for the Lotus 
Esprit to 0.74 for the Lotus Seven. 

An open cockpit and an erect windscreen completely spoil the airflow pattern 
over a sports car and a higher maximum speed is always given with a hard top 
fitted and the side windows closed. The provision of a metal cover over the 
passenger’s seat on a Triumph TR2, a single aero-screen for the driver and spats 
over the rear wheel openings, reduced the drag coefficient by 10 per cent and the 
frontal area by about 20 per cent, a total saving in drag horsepower of about 28 
per cent. 

Wind-tunnel tests on scale models of cars have shown that the behaviour of 
the air flowing round a car body adjacent to a flat surface, i.e. the road surface, 
is vastly different from the flow pattern around the same body in free air. There 
is a build-up of pressure underneath the car which can lead to steering lightness 

The Frame and Body 179 

at speed. On a high-speed car it is sometimes necessary to modify the upper 
surfaces at the front end, even moving away from the true streamline shape, in 
order to produce a downward thrust on the nose of the car. The laminar, or 
streamline, nature of the flow does not persist for long on the upper and side 
surfaces of the body. The flow underneath, even when an undershield is fitted, is 
turbulent over almost the whole length of the car. Early streamlined cars had 
very long tails in an attempt to conform to the basic streamline form for the 
maximum speed. The early break-up of the laminar flow, however, produced by 
the presence of the road surface and the eddies induced by such unavoidable 
excrescences as the windscreen, mirror, exhaust system, wheel arches, etc., 
makes it impossible to derive any benefit from the use of a long tail. The flow 
over the tail is fully turbulent, even with a good design of body, and the use of a 
long tail only leads to excessive friction drag. The modern high speed tail is 
therefore no tail at all, usually having the appearance of being chopped of with a 
knife. Dr. W.I.E. Kamm of Germany was the first to apply this knowledge to 
the design of streamlined automobiles. These were not a commercial success, 
since the motorist has a prejudice against anything that he considers to be 
‘unstreamlined’. In competition there was not the same prejudice, particularly 
when it was realized that the Kamm tail also lightened the vehicle (see Figure 

entrance height need only 








Fig. 9.8 (a), (b) The design of radiator ducts. 

180 The Sports Car 

The internal air resistance of a car is almost as important as the external. Air 
must pass into the car to cool the radiator (or the cylinders on an air-cooled 
engine), to cool the brakes and to ventilate the cockpit. Air must also be taken to 
the carburettors. All this air must leave the body again (even the carburetted air 
must leave in the form of exhaust gas) and the positions chosen for these exits, if 
badly chosen, can seriously upset the airflow round the body. If the air is 
exhausted into a pressure area the flow through the duct can be reduced or even 
stopped. Internal ducting for the radiator and brakes must be designed to reduce 
turbulence to a minimum and to have as few bends as possible. The power to 
move this air through the ducts must be provided by ihe engine. Figure 9.8 (a) 
shows how the inlet to a radiator cooling duct for a sports/racing car need only 
have about one-sixth of the actual radiator area to give adequate cooling. Figure 
9.8 ( b) shows how a blanking plate across the inlet can cause excessive 
turbulence and power loss. 


Lift at high speed 

The side profile of a typical car is unsymmetrical, being relatively flat at the base 
and curved on the top, a little like the cross-section of a thick aeroplane wing. It 
is not surprising then that lift can be produced at high speed. On a typical sports 
car a lift of about 150 lb is produced at 100 m.p.h., most of this being applied to 
the rear end. At higher speeds the lift increases and can seriously affect the 
stability and controllability of the car. In the colourful words of Hap Sharp, of 
the Chaparral racing team, ‘You’re blasting down the straight ... and then you 

The Frame and Body 181 

begin to hear this kind of high-pitched whine, a kind of whirring noise, and 
everything begins to go soft on you. And you better believe you are trying to 
fly.’ The Chaparral was fitted with an uptilted flap at the rear end of the body, a 
‘spoiler’ in aeronautical terminology, to create a downward thrust on the rear 
wheels to counteract the induced lift and the handling of the car at high speed 
became much less hazardous. 

In the 1967 Chaparral the spoiler concept was developed into an adjustable 
aerofoil mounted on struts at the rear of the car. By moving a lever the driver 
could dip the leading edge of the aerofoil to produce the desired downthrust at 
high speed, to reduce wheel spin during acceleration or increase cornering power 
in fast bends. 

The followers of Motor Sport will know what a plethora of flimsy, even 
dangerous, aerofoils appeared on Grand Prix cars in the following years. The 
FIA safety committee have now got the situation under control and the spin-off 
from this intensive experimentation over the last ten years has been the 
application of this aerodynamic knowledge to the design of high-speed sports 
cars. Figure 9.9 shows a typical pressure distribution from front to rear on a 
typical sports car body. Areas of positive pressure exist, as one might expect, at 
the front of the car and in front of the windscreen. Overall, however, the mean 
pressure from front to rear is negative. As the speed increases so does the 
magnitude of this lift. The centre of pressure, which is to pressure what a centre 
of gravity is to mass , i.e. an effective point at which the summation of all the 
individual pressure zones may be considered to act, has a tendency to move 
forwards as the speed increases. Hence the need for an aerodynamic device to 
create a downthrust at the rear of the car if the car is to handle in a stable 
manner at speed. A typical aerofoil or rear spoiler as fitted to the Porsche Turbo 
is seen in Figure 9.10. 

Fig. 9.10 The Porsche Turbo body, showing the aerofoil to prevent rear-end lightness 
at speed. 

182 The Sports Car 

The air dam 

Even when a designer provides a smooth underfloor to his car, with an exhaust 
system effectively buried in a tunnel and very little else exposed to the airstream 
but the suspension components and the brake discs, the turbulent condition of 
the air that passes underneath the car contributes considerably to the overall 
drag loss. We can, however, by confining our design to sensible limits on ground 
clearance, attempt to deflect the air stream approaching the front of the car so 
that a proportion no longer passes under the car but is directed outwards and 
upwards to flow along the sides. With the help of the wind tunnel the designer 
can angle this front spoiler or air dam to create a measure of downthrust in the 
manner of the rear spoiler. 

Many modern high-speed cars are fitted with air dams and tests have shown in 
a few cases that they help to reduce fuel consumption. 

Fig. 9.11 Kamm tail profile, showing separation of boundary 
layer and break-up into vortices. 

Directional stability at high speed 

A body should be designed so that the air forces acting all round it will tend to 
heat it into the direction the nose is pointed in the manner of a flighted dart. The 
fins fitted to the tails of land-speed record cars act like the feathers on a dart, 
providing a powerful side-thrust to the rear end to bring it back into line if the 
car should veer from its true course. Designers of modern sports cars try to 


Centre Neutral Resultant 

C,G. of wind steer wind force 

Fig. 9.12 Stability under cross-wind. 

The Frame and Body 183 

introducing the additional drag that stabilizing fins always inflict and try to 
achieve the same effect from the shape of the rear section of the body itself 
(Figure 9.11). How this instability arises can be understood by referring to 
Figure 9.12, which shows the relative positions of the centre of gravity and the 
centre of pressure in a car which has aerodynamic stability. The centre of gravity 
is the point at which we could balance the mass on the point of a pin. In the 
same way, the centre of pressure is the point at which the forces produced by the 
air pressure, both positive and negative, acting all round the body, may be 
considered to be concentrated. In the case shown, the air pressures tend to hold 
the car in stable equilibrium. If the car should be deflected from its straight path 
by a bump in the road or a sudden gust of wind against the side, the pressure on 
the side of the car towards which the rear end turns will increase and the turning 
moment about the centre of gravity will bring the car on a straight path again. 
The forces from the tyres tending to produce under or oversteer (see Chapter 
Seven) are superimposed on the above forces. 

As the car travels at higher and higher speeds the positive air pressures at the 
r ront increase at a greater rate than the negative ones at the rear and the centre 
jf pressure moves forward. With the usual modern body shape, at speeds above 
120 m.p.h. the centre of pressure moves in front of the centre of gravity and a 
state of unstable equilibrium is reached in which any swing from the true course 
brings into action a turning moment about the centre of gravity which tends to 
increase the swing still further. This calls for extremely quick and frequent 
steering corrections from the driver. A palliative used in certain cases is to 
increase the slip angles of the front tyres relative to the rear, thus bringing into 
play greater understeering forces from the tyres. If overdone, this can lead to 
excessive understeer at low speeds. The remedy adopted by many aerodynamicists 

Fig. 9.13 One-eighth scale model used in wind tunnel tests of Jaguar XJ-S. 

Fig. 9.14 Stabilising ‘fins’ at rear of Jaguar XJ-S body. 

is to incorporate additional vertical surfaces at the rear end of the car. Rear and 
mid-engined cars present no problem here since the body sides tend to be high at 
the rear to accommodate the bulk of the engine and its auxiliaries. 

Fig. 9.15 Wind tunnel testing of XJ-S at MIRA laboratories. Note use of wool-tufts 
and white smoke to indicate flow lines and degree of turbulence 

The Frame and Body 185 

Wind tunnel tests on the model of the prototype Jaguar XJ-S made by the late 
Malcolm Sayer led him to incorporate curving side fins at the rear quarters. 
These can be seen on the one-eighth scale model (Figure 9.13) and more 
distinctly in the three-quarter rear view of the production XJ-S (Figure 9.14). 
The rear of the car is also seen to have a cut-off tail, a much less drastic 
treatment than the full ‘Kamm tail’ shown earlier in Figure 9.11. 

Many techniques have been used to study the flow lines around a car body, 
but the old wool-tuft method is still popular. Figure 9.15 shows the method in 
use on the prototype XJ-S. Each tuft is glued to the body at one end. The extent 
of eddy formation (turbulence) in any zone can be recorded by still photography 
and, more effectively, by the movie camera, since the speed of the tuft agitation 
can be seen in this way. As shown in this photograph white smoke, introduced at 
the tunnel inlet, can also help the aerodynamicist to study the general flow 
pattern. The air dam on the XJ-S can be clearly seen below the front bumper. 

The transmission 

'Consider that men will do the same things nevertheless, 
even though thou shouldst burst.' 


Torque multiplication 

Unlike the steam engine, the petrol engine is characterized by a lack of flexibility 
in its torque range. At tick-over speeds of 500-700 r.p.m., or thereabouts, on the 
modern car engine all the power produced in the combustion chamber is 
absorbed in overcoming the internal resistance of the engine. With the throttle 
almost closed, at tick-over, however, satisfactory carburation is achieved. If the 
clutch is engaged and the throttle opened wider the experienced driver knows 
that the engine speed must rise to 1000-1500 r.p.m. before the clutch is fully 
engaged or the torque transmitted will be insufficient and the engine will stall. 
The lack of torque at low engine speeds and large throttle openings is entirely a 
question of carburettor limitations. Large throttle openings at low speeds cause 
a breakdown in carburation since the air velocity in the induction tract becomes 
too low to carry the fuel droplets in suspension. With the relatively large choke 
sizes used on sports cars this effect is aggravated. 

The limiting torque range on a typical 125 m.p.h. sports car might be 1500 to 
6000 r.p.m. If the driver were clever enough to start in top gear, then he would 
be required to slip the clutch until the car speed was more than 30 m.p.h. before 
he could risk full engagement. Full engagement below this speed would mean a 
stalled engine. Apart from the difficulty of such sensitive clutch operation, few 
clutches would withstand such a high degree of slip for long. This is one reason 
for having a gearbox. The other is the need for ‘torque multiplication’ between 
the engine and the road wheels. Acceleration takes place when the torque 
transmitted to the driving wheels exceeds the torque necessary to overcome the 
tractive resistance, the combined rolling and air resistances; the greater the 
excess torque, the greater the acceleration. If we wish, then to achieve maximum 
acceleration over the whole speed range of the car, the overall gear ratio between 

The Transmission 187 

the engine and the road wheels must be varied in such a way that the engine is 
always running at the speed which gives the maximum torque at the road 
wheels.* This is the ideal and can only be given by an infinitely variable speed 
gear. The automatic transmission is a combination of a limited range torque 
converter and a self-change gearbox. Unfortunately the torque converter is not 
as efficient as a gear train, despite improvements in recent years. Even so there is 
a gradual drift in the minds of many designers towards the acceptance of 
automatic transmissions for sports cars. 

The manual gearbox is a compromise (see Figure 10.1), giving, as it does, 3, 4 
or 5 definite changes in ratio at intervals in the speed range of the car. In this 
way, when accelerating through the gears the engine speed varies in turn in each 
gear between approximately maximum torque engine speed and maximum 
power engine speed, thus giving no great drop in torque at either limit. The 
engine designer may have chosen his valve timing, carburettor choke sizes, etc., 
with the sole object of producing a flat torque curve over the operating range. In 
this case the design of a suitable gearbox is easy. Unfortunately, if the designer’s 
aim has been to win as much power as possible from his engine, then the torque 
at low speed is poor and careful choice of gear ratios is necessary if acceleration 
is not to suffer. This is precisely the position with some sports cars today and 
has led to the adoption of a five-speed gearbox in some cases. 

*At first sight, one would expect this to be the speed giving maximum engine torque. The torque 
curve falls off slowly, however, and by using a bigger overall speed reduction ratio, more torque is 
given at the road wheels at higher engine speeds, and in fact, for a given road speed, the greatest 
torque is given to the road wheels at maximum power engine speed. 

188 The Sports Car 

The gear ratios 

It requires no great gift for mathematics to see that to achieve the maximum rate 
of acceleration through the whole speed range the step-up ratio between 
adjacent gears must form a geometric progression, i.e. 

1st ratio _ 2nd ratio _ 3rd ratio 
2nd ratio 3rd ratio 4th ratio 

If the step-up ratio between the gears is, say, 1.5 and the final ratio is 4.0, the 
gear ratios would be as follows: 

13.5 : 9.0 : 6.0 : 4.0 

With this set of ratios and a perfect driver, i.e. one who changes gear at the 
right engine speed each time, the same speed range of the torque curve will be 
used in each gear as he accelerates from a standstill up to speed. 

The typical European car driver spends much of his time changing backwards 
and forwards between top and third gear. A change down to second is made less 
frequently. Often, in the typical European terrain, a gradient is encountered 
which calls for a step-up ratio, not as much as 1.5, but something less drastic, 
say, 1.3. This type of box with a small step-up ratio between top and third and a 
large step-up ratio between third and second and an equally large step-up ratio 
between second and first was popular for many years and is still used on some 
popular cars. A more logical set of gear ratios is given when the step-up ratio is 
gradually reduced when travelling through the gears from low to high. A typical 
set of gears would be: 

Gearratio 13.5 : 7.8 : 5.2 : 4.0 

Step-up ratio 1.7 1.5 1.3 

Compare this with the set of gears above, which cover the same range but with 
geometric progression . 

The overdrive 

Many British sports cars are fitted with the Laycock de Normanville overdrive 
unit. Not only does this lead to an improvement in fuel economy at high cruising 
speeds but it gives all the advantages of a 5- or even a 6-speed gearbox with the 
additional advantage of fingertip gear selection in the higher ratios. Some 
designers prefer to use the overdrive on top gear only, giving in effect a 5-speed 
gearbox. Others prefer to make it opeative on both top and third, so that 
overdrive third gear is an intermediate step between direct third and direct top. 

Essentially, the Laycock de Normanville overdrive is an epicyclic gear train. 
This gear train being brought into operation by means of a hydraulically 
actuated cone clutch. This clutch locks the sun wheel to the outer casing and 
releases the internally toothed annulus. When overdrive is engaged the input 

The Transmission 189 

shaft rotates the carrier holding the three planet wheels around the stationary 
sun wheel to drive the annulus, which is connected to the output shaft, at a 
higher speed (from 22 to 32 per cent higher — depending upon the model). 

Hydraulic pressure is supplied from a hydraulic accumulator and a plunger- 
type pump, cam-operated from the input shaft. Selection of overdrive is 
controlled by a small switch, conveniently placed near the driver’s hand, a relay 
and a solenoid-operated hydraulic valve. An over-riding gearbox switch is used 
to make the overdrive inoperative in the appropriate lower gears and in reverse 

In direct drive, the sliding cone clutch releases the sun wheel from the friction 
lining on the outside of the clutch and locks it by its inner friction lining to the 
annulus. This renders the epicyclic train inoperative. Engine power is trans¬ 
mitted directly through a uni-directional clutch unit which couples the input and 
output shafts. This clutch resembles the old Rover free-wheel device in that a set 
of hardened steel rollers slides up ramps on the inner clutch member to engage 
claws on the outer member. On the over-run the rollers slide inwards down the 
ramps releasing the claws and allowing the outer member to free-wheel. This 
free-wheeling clutch is required so that the annulus can disengage from the input 
shaft and rotate at a higher speed when overdrive is in operation. 

An outstanding feature of this clever design is that engine braking is given at 
all times. In direct drive the sun wheel and annulus are locked together by the 
cone clutch and engine braking is obtained through the locked train. In 
overdrive, the sun wheel is locked to the outer casing and rotation of the annulus 
in the reverse direction is resisted by rotation of the planet carrier against the 


The baulk-ring (called blocker synchro in America) type of synchromesh has 
now replaced the older constant-load type of synchromesh on all quality cars. It 
was invented by Professor Ferdinand Porsche and first used on his 1947 
Cisitalia racing car. In this system it is physically impossible to engage the gear 
dogs until synchronization is achieved. There have been many versions of 
baulk-ring synchromesh since the Porsche original and they all work on the 
same basic principle whereby a slotted ring is caused to rotate several degrees 
under the action of the torque created at the mating surfaces of the cones. This 
small amount of rotation of the baulk ring moves the slot (or slots) in the ring so 
that they block the further movement of the sliding clutch member. When the 
speeds become synchronized there is no longer a torque applied to the baulk ring 
and it returns to its original position. The slots in the ring now line up with the 
pins or keys in the sliding clutch member and the dog clutch is now free to move 
into full engagement with the external driving dogs at the side of the gear wheel. 
It is obvious from this that the dogs cannot clash since it is impossible for them 
to be engaged until the cones are perfectly synchronized. 

Fig. 10.2 The Jaguar gearbox with details of baulk-ring synchromesh. 

The clutch 

The single-plate diaphragm clutch is used on many British sports cars (see Figure 
10.3). The Datsun 260Z clutch is very similar in design, but in this case a 
ball-bearing design of clutch release is used instead of the older carbon-faced 
ring. In the diaphragm clutch the ring of coil springs that were formerly used to 
hold the pressure plate in contact with the driven plate are replaced with a single 
dished spring consisting of an outer rim and a number of integral fingers 
radiating inwards from this rim. The clutch release bearing moves the inner ends 
of these fingers to release the load on the pressure plate and free the driven plate 
from the flywheel face. With large engines, requiring the 10 or 11 in. diameter 
clutch, the centrifugal loading on this design of clutch becomes so great that it is 
not advisable to use it for engine speeds above 6,500 r.p.m. For higher engine 
speeds on competition engines a more reliable dlutch would be the multi-plate 
clutch that Borg and Beck developed specially for this arduous duty. 

The automatic transmission 

The automatic transmission is the norm in America; it has been developed to a 

The Transmission 191 

Fig. 10.3 Borg and Beck diaphragm clutch. 

high degree of reliability and is only about 8 per cent less efficient than the 
manual gearbox in terms of overall fuel consumption. The ZF Company of 
Germany have developed a new type of automatic gearbox that is more 
attractive to the sporting motorist and this is the basis of the ‘Sportomatic’ 
gearbox that is currently available on Porsche cars. 

The ZF philosophy is based on a realisation that a simple plate clutch is one of 
the most abused and overworked components fitted to the modern motor car 
when one considers the number of engagements and disengagements made in a 
short suburban journey during the rush hour. It could be argued that this is not 
the milieu intended for the sports car, but the typical owner does, unwillingly, 
spend many hours of his motoring life under these tedious conditions. Friends in 
the motor trade assure me that the average life of a clutch today in city traffic is 
less than 40,000, yet the engine is still in excellent condition after twice this 

The WSK transmission, first shown by ZF at the Frankfurt Show, comprises 
a three-element torque converter with automatic lock-up and a free-wheel to 
give engine braking effect. The torque converter is connected by a single-plate 
clutch to a standard ZF synchromesh gearbox with 3, 4 or even 5 ratios. On the 
Porsche 3 ratios are provided. There is a weight penalty of about 12 per cent 
when compared with a simple manual gearbox and single-plate clutch, but the 
additional expense is recovered by the inherent long-life of the transmission. 

On the Porsche an induction vacuum-operated control system has been 
devised to operate the clutch, which is never used to take up full torque, but is 
only used to lock-up the drive to eliminate converter losses when the car is on the 
move. The wear and tear on the clutch surfaces is therefore minimal. With the 
converter in lock-up normal gearchanges can be carried out by the driver 
without the use of a clutch pedal. It is a simple two-pedal system as in the 
popular American automatic transmission, but there is no possibility of an 

Fig, 10.4 ZF automatic transmission 

The Transmission 193 

unexpected change into a higher gear if the accelerator is released momentarily 
and the driver is always in full control of the gearbox. In heavy traffic the driver 
can let the torque converter take over, thus reducing the number of operations 
of the lock-up clutch to about one-tenth the number that would occur with a 
normal clutch and manual gearbox. 

The cross-section of the transmission shown in Figure 10.4 shows the main 
elements of the WSK system. This is not the Porsche automatic transmission but 
it serves to illustrate the principle. 




6 . 
15 . 
20 . 
21 . 

23 . 



Pinion shaft drive flange 
Pinion shaft outer bearing 
Pinion shaft 
Front pinion bearing 
Rear pinion bearing 
Differential cage outer bearing 
Drive shaft flange 
Differential cage 
Cover plate 
Chassis mounting plate 
Differential gear shaft 
Differential gear, free 
Differential gear* fixed 
Crown wheel 

Fig. 10.5 Rear drive unit on Datsun 260Z. 

3052/56 MrertgThen^d 
clinch 240 rr 

Fig. 10.6 Porsche transaxle as used on Turbo model. The notes on the drawing refer to modifications made to the 
1976 model. 

The Transmission 195 

The final drive 

The bevel drive, and its popular variant, the off-set bevel or hypoid drive, have 
been with us so long that we take their quiet reliability for granted giving no 
thought to the long history of development behind them. 

Originally introduced with great zest by the gear specialists to please the body 
stylists (whose desire to provide a low floor-line for the rear-seated duchess was 
very commendable), the hypoid gear is now almost universal and is in use on 
many sports cars that do not have rear seats. 

The name ‘hypoid’ is an abbreviation for ‘hyperboloid of revolution’, since 
this is the form of the envelope traced out by the pitch surfaces of this particular 
gear tooth profile. The pitch surfaces of a spiral bevel and pinion, however, lie 
on cones with a common apex. A discussion of hypoid tooth forms is beyond 
the scope fo a book of this kind. It is enough to state that the hypoid gains 
fractionally from the greater number of teeth in mesh and the load-carrying 
capacity is a little greater. Nevertheless there is a serious disadvantage. The 
spiral bevel gives a pure rolling contact. With the hypoid gears a certain amount 
of sliding occurs between the mating teeth. Much research has been carried out 
by the oil companies and special hypoid oils have been developed which prevent 
scuffing of the tooth surfaces under this sliding action. 

A good example of the rear drive unit on a front-engined sports car with IRS 
is shown in Figure 10.5. This is a broken section on the Datsun 260Z, the 
cross-section being on two levels since, with a hypoid gear set, the pinion is 
situated at a lower level in this plan view than the centre-line of the crown wheel. 
The pinion shaft is carried in one ball bearing and two tapered roller bearings 
which are preloaded during assembly. This is common modern practice. 

Cars with rear or mid-engine location usually have a combined gearbox and 
final drive assembly called a transaxle . An excellent example is the latest 
Porsche Type 930 transaxle as used on the Turbo model. In this case, it will be 
seen that the spiral bevel gear set is not a hypoid, i.e. the pinion shaft centre-line 
intercepts that of the crown wheel. Porsche obviously prefer to use large enough 
gears rather than accept the undesirable scuffing action of the hypoid gears. The 
single-plate clutch, seen on the right of Figure 10.6, transmits engine torque via 
the drive shaft across the top of the differential cage to the gearbox at the rear. 
The four forward gears are contained in the main casting, which is of silicon 
aluminium alloy. The reverse gears are behind the rear bearings in a separate 
compartment. The gear selector shaft can be seen at the bottom of the housing. 

Porsche have a policy of incorporating improvements and modifications on a 
continuous basis, not at the behest of ‘Motor Show’-conscious sales managers. 
Figure 10.6 is a drawing issued by the Porsche Service Department to indicate 
new features on the 1975/76 Turbo model. 

Universal joints 

If it is true that ‘necessity is the mother of invention’, it is also true that the 
period of gestation is unpredictable. The need for a good reliable constant- 
velocity universal joint existed for two decades or more before Rzeppa found 

196 The Sports Car 

the answer. The lack of a good constant-velocity universal joint led to the failure 
or ‘limited success’ of several front wheel drive cars in the past. Today we have 
the Rzeppa and the Weiss joint, both using balls located in matching half¬ 
grooves in the two halves of the joint and both providing a constant velocity 

Let us explain what we mean by constant velocity. The majority of the older 
universal joints operated in the manner of a simple Hooke’s joint, two of which 
are shown in Figure 10.7. If the two shafts are at an angle and the driving shaft 

Fig. 10.7 Double Hooke’s joint. 

is rotating at a constant angular velocity, the angular velocity of the driven shaft 
goes through a cyclical variation every 90 degrees, being slower than the driving 
shaft at first, then faster, the mean velocity over each quarter revolution being, 
of course, equal. For an angular displacement of 8 degrees between driven and 
driving shafts, the maximum velocity variation is only 2 per cent, but if the angle 
between the shafts is trebled to 24 degrees the velocity variation is increased 
nine-fold to 18 per cent, which is a serious variation. 

The use of two Hooke’s joints, 90 degrees out-of-phase, as shown in Figure 
10.7, is one way to transmit a constant velocity, but the centre shaft still goes 
through a cyclical velocity variation and this in itself can transmit unpleasant 
vibrations to the whole assembly and to the car itself if shaft angles of 20 degrees 
are exceeded. For a front wheel drive car, angles of 40 degrees or more are 

The Rzeppa joint and the British equivalent, the Birfield-Rzeppa which is 
used on the Leyland front wheel drive cars, are shown in cross-section in 

Fig. 10.8 (a) Rzeppa constant-velocity joint, (b) Birfield-Rzeppa constant- 

velocity joint. 

The Transmission 197 

Figure 10.8. The drive is transmitted through the six hardened steel balls which 
roll in grooved raceways in the two halves of the joint. Constant velocity is 
achieved by the geometry of the ball grooves which maintains the six balls and 
their locating cage in a half-angle position at all times. That is to say that the six 
balls and their cage always lie in the same place, this being the plane which 
bisects the angle between the driving and the driven shaft — whatever that angle 
may be. This is illustrated in Figure 10.9. The driven shaft rotates at exactly the 

Fig. 10.9 Action of the Rzeppa joint. 

same angular velocity as the driving shaft at all times. The Rzeppa uses a 
circular contour for the ball grooves. The Birfield-Rzeppa uses an elliptical 
contour giving a two-point contact with each ball instead of a line. Both designs 
are quiet and cool running and will give long life if not loaded beyond their 
designed capacity. 

A good example of modern front wheel drive transmission using constant 
velocity universal joints is the Lancia Beta shown in Figure 10.10. 

The limited slip differential 

If one wheel loses traction on a normal differential no torque can be applied to 
the opposite wheel. This is an obvious disadvantage when driving in snow and 
ice and can also be found to reduce acceleration on good road surfaces, 
particularly when the driven wheels are not independently sprung and load 
transference can occur between the driven wheels, as discussed in Chapter Eight 
and illustrated in Figure 8.4. An early solution to the problem of excessive 
one-wheel spin was to provide a ready means for locking the differential. This 
was a drastic solution since we still would like the benefits of the differential 
when turning corners. The solution that is used today is a device which allows 
the differential to operate, but limits the relative speeds at which the two driven 
wheels can rotate. In other words, slip is limited. A good modern example is the 
Thornton Power-Lok , made by the Dana Corporation in America and Salisbury 
Transmission in England and fitted to the XJ-S Jaguar. In this design multi-disc 

The Transmission 199 

clutches are fitted on the outside of each differential side gear. Each clutch unit 
comprises four discs loaded by a Belleville (cone spring) washer. When one 
wheel starts to spin, axial thrust produces a cam action between the wedge- 
shaped ends of the pinion shafts and corresponding wedge-sided holes cut in the 
differential carrier. The differential carrier is split and is permitted outward 
axial movement. This outward movement occurs under the cam action from the 
pinion shaft ends and puts pressure on the clutches, automatically restricting 
rotation of the differential carrier. Anyone who has driven in the wet in one of 
the older 3.8-litre Jaguars with non-independent suspension at the rear and with 
no limited slip differential will appreciate how much this device can contribute 
to road safety. 

The brakes 

'Let it make no difference to thee 
whether thou art cold or warm, 
if thou art doing thy duty.' 


The grip on the road 

Whatever power we may apply to the brake shoes, whatever we may do to try to 
keep the drums cool, our final limitation on the rate of retardation is the grip of 
the tyres on the road surface. Locked wheels are of little use on a slippery 
surface as many of us have learned from experience of winter motoring. 

In Table 11.1 are given average stopping distances from 30 m.p.h. on several 
typical surfaces. The values were obtained with smooth tyres; improved figures 
could, of course, be obtained with a good tread. 

Table 11.1 Average stopping distance from 30 m.p.h. with 
smooth tyres 


Clean dry surface (without ruts) 


Clean dry surface with tar melting 


Dry surface with loose sand 


Wet ‘non-skid surface 


Average wet surface 


Wet surface with loose sand 


Icy surface 


Wet surface with poor skid resistance 


Experiments by the Road Research Laboratory have shown that locked-wheel 
stopping distances of the order of 30 to 40 ft from 30 m.p.h. can be obtained 
under the best conditions on the following surfaces: 

(a) concrete; 

The Brakes 201 

(b) macadam; 

(c) hot-rolled asphalt with precoated chippings; 

(d) hot-rolled asphalt without precoated chippings, but with stone in the mix; 

(e) dense tar surfacing; 

(f) fine-texture asphalt. 

Stone setts, wood blocks, cast-iron paving, rubber blocks and rock and mastic 
asphalts without precoated chippings all give stopping distances of 50 ft or more 
in the most favourable conditions. These experiments, however, showed that a 
wide range of stopping distances are given by all surfaces. Concrete, for 
example, over 70 tests showed a range of 34 to 130 ft. The best figures were 
given by dense tar surfacing and rock asphalt with precoated chippings for 
which the ranges were 32 to 64 and 44 to 60 ft respectively. Rock asphalt gave in 
one test a stopping distance from 30 m.p.h. of 495 ft — a truly magnificent skid! 
Since the above surfaces were chosen at random throughout the British Isles, 
one shudders to think of the consequences of a pedestrian crossing placed on 
one of the worst of these. 

One aspect of road surfacing which must not be neglected is the abnormal 
wear that can be produced on the tyres of a high-speed car when abnormally 
large chippings are used to provide a non-skid surface. Since a surface with the 
same non-skid properties can be made with a relatively smooth finish which is 
less injurious to the tyres the continued use of such abrasive surfaces only 
increases the cost of motoring. 

Braking forces 

The above is a broad picture of the limitations set by the frictional forces 
between the tyres and the road surface. However great the braking forces we are 
able to apply to the brake drums, the upper limit on the rate of retardation is 
always set by the grip of the tyres on the road. There is no need to stress how 
much depends upon the driver’s ability to recognize road surfaces and to drive 
accordingly. The designer nevertheless must provide adequate braking for the 
best conditions where decelerations of 1.0 to 1.3G are possible. Since a certain 
falling off in braking efficiency is inevitable between overhauls, it is advisable to 
increase this slightly and provide braking forces to give a maximum retardation 
of 1.5G. 

Weight transfer under braking 

Under hard braking the load on the front wheels is increased considerably, that 
on the rear wheels correspondingly reduced. We usually call this ‘weight 
transfer’. Pedants could object to this, since thepull of gravity is not changed. It 
is really a transfer of force, not weight, but our interest is in the loads on the 
tyres, since this helps the designer to determine relative areas of the brake pads 
or brake linings at front and rear. 

In Figure 11.1 it is seen that under a braking effort of 1G, the moment Wx is 
resisted by the couple Hy, where is the load transferred from rear to front. 

202 The Sports Car 

Wf=% + W b 

Fig. 11.1 Weight transfer under braking. 

This therefore calls for a braking distribution of approximately 70/30. If we 
were to design on this figure what would happen if, on a wet road surface we 
found the limiting deceleration to be only 0.2 G? The braking transfer couple 
would become 


- : — = 75 lb. 


1075 lb. 

925 lb. 

With a brake actuation giving a 70/30 distribution, the retarding force at the 
road surface at the rear would be 

0.2X2000X0.3 = 120 lb 

and at the front, 

0.2X2000X0.7 = 280 lb. 

With the actual loads at the front and rear with this low rate of deceleration, 
the forces at the rear tyre will tend to give a deceleration of 


W b = 
• Wf = 

K = 

The Brakes 203 

and at the front, 



= 0.13 




= 0.26 G 

Since the limiting deceleration on the wet road surface is 0.2 G, the front 
wheels will slide and the rear wheels will still be turning. 

We are now faced with a problem. Is it better to choose a 70/30 distribution 
which will lock the front wheels on wet roads or a 55/45 distribution which will 
lock the rear wheels when braking hard on dry roads? 

Experiments all over the world, beginning as far back as Darwin and Burton’s 
work on ‘Side slip on motor cars’, reported in Engineering , September 1904, 
have confirmed the following: 

Locking the rear wheels only, produced a condition which resulted in partial 
or complete loss of control of the vehicle. This state was very unstable and the 
car could easily go into a spin. 

Locking the front wheels only, produced a stable condition in which the 
vehicle travelled in a straight line. In this condition there was almost complete 
loss of steering control. 

Locking all four wheels produced a condition which was fairly stable, 
resulting in the car travelling in a straight line. If the wheels on one side 
encountered a higher coefficient of friction than those on the other side, the car 
would turn towards the side with the higher drag. (A good example of this 
would be braking hard on a road with a loose surface near the verge. The loose 
‘ball-bearing’ friction on the near-side would offer the least resistance to a 
locked wheel and the car would gradually turn towards the crown of the road, 
the back end sliding more and more into the ditch. If on the other hand the road 
were icy, with a lower coefficient of friction near the centre of the road than on 
the gravelled surface near the verge, the car would tend to dive nose first into the 

From all this we see that if we must lock any pair of wheels, it is safer to lock 
the front ones, since this ensures a stable straight line ditch-free halt. This has 
led to the adoption of a 60/40 distribution in general and a 65/35 distribution 
for racing cars and sports cars. There is no doubt that an easy adjustment of the 
braking distribution operated from the driver’s seat would be invaluable to the 
keen motorist and especially to the rally competitor. 

Brake fade 

This phenomenon has led to a revolution in brake design in which the disc brake 
has gradually replaced the drum brake on the front wheels of nearly every car 
sold in Europe with any pretensions to high performance. In a few more years 
the changeover will be complete at the rear end too. 

Brake fade is caused in two ways. The first cause is the drop in the coefficient 
of friction that can occur rather suddenly when the temperature of the brake 

204 The Sports Car 

lining rises above some critical temperature, usually about 400°C (750°F). 
Attempts to combat fade on drum brakes by using a hard lining were not too 
successful since the second cause of fade is the loss of ‘bedding’ of the linings 
when the drums expand away from the linings. Since the curvature of the 
expanded drum no longer conforms to the curvature of the linings there is a 
serious loss of effective brake area when rigid unyielding linings are used — and 
linings that retain a high coefficient of friction at high temperatures are 
inevitably hard linings. Coefficients measured on hard and soft Mintex linings 
are shown in Figure 11.2 

Friction surface temperature°C 

Fig. 11.2 Variation of coefficient of friction with 
rising temperatures on several Mintex 
lining materials. 

When we brake a medium-sized sports car from 100 to 30 m.p.h. using the 
brakes to the full, we generate enough heat in the brakes to bring a 2-quart size 
kettle of water to the boil and all in about five seconds. 

There is little hope that we could ever dissipate it as fast as it is generated. The 
temperature of the linings, the shoes and the drums rise during braking and fall 
again during the intervals between brake applications. On a twisty road taken at 
speed, the brake applications follow in quick succession and the heat from one 
application has not been fully dissipated before the next one occurs. In these 
circumstances the temperature of the linings can rise towards the dangerous 
critical point at which fade begins. With certain poor designs with full disc 
wheels and enclosed wings with spats, the temperature can rise to 500°C 

The Brakes 205 

(930°F). The effect of this on the coefficient of friction varies from lining to 
lining, but in most cases it results in a serious falling off in the braking effort. 

To expose the wheel again to the full blast of the air stream as on the older 
sports cars would be a simple solution but the attendant drag from the wheel 
and suspension system becomes formidable at speeds above 120 m.p.h. and this 
solution, except in certain special cases, cannot be entertained. 


Disc brakes proved to be the answer to the problem of brake fade. The idea was 
not new. Dr F. Lanchester took out a British patent on a type of disc brake in 
1902 and it was used for the first time on the 1906 25 h.p. Lanchester car. The 
Lanchester brake used calipers to grip both sides of a thin disc. Operation was 
mechanical and was not outstandingly efficient, even by 1906 standards. 

During the war the Aviation Division of the Dunlop Rubber Company 
produced a successful aircraft disc brake. In 1949 they began to apply this 
experience to the automobile field. In this first experimental automobile design 
a 3 A- in. thick, 12-in. diameter cast-iron disc, mounted on the wheel hub and 
rotating with the wheel, was straddled by a heavy U-section caliper member 
containing three pairs of circular friction pads. Braking was achieved by the 
three pairs of pads gripping the sides of the disc when hydraulic pressure was 
applied to three pairs of pistons behind the pads. Instead of the normal 
hydraulic system a hydraulic servo system was used in which pedal pressure 
controlled the output of a hydraulic pump driven from the gearbox mainshaft. 
Racing experience with these brakes in 1952 led to their appearance in modified 
form on the 1953 Jaguars at Le Mans. The effect was electrifying, especially at 
the end of the 160 m.p.h. Mulsanne Straight, where the Jaguar drivers could 
delay the application of their brakes to a point 300 yards nearer the hairpin 
than their Continental rivals. There is little doubt that these new brakes won 
the race that day. This form of brake was later used on the ‘D’-Type Jaguar, 
three pairs of pads being used on each front wheel and two pairs on each 
rear wheel. 

Experience of disc brakes suggested at first that they were the complete and 
final answer to the problem of brake fade. Improvements in tread compounds 
and the arrival of low profile tyres have made it possible to use even higher rates 
of deceleration and it has now become necessary on racing cars to use ventilated 
disc brakes. These are hollow discs with the passage in the central section 
designed like a shrouded impellor or fan to produce a radial air flow. Ventilated 
disc brakes are now being used on some of the faster production sports cars, 
such as the Ferrari and Porsche. 

The freedom from fade we associate with the disc brake comes in the main 
from its direct exposure to a blast of cooling air. The use of a cooling fan on a 
drum brake will draw cooling air over the inner surface of the drum, but the 
direct air-blast cooling of the disc brake is far superior. Even when the 
temperature of the disc and the pads rises under the action of fierce and frequent 

206 The Sports Car 

brake applications the effects of differential expansion are negligible since the 
mating area of the two surfaces suffers no distortion. In the case of the drum 
brake, differential expansion can change the relative radii of curvature and 
seriously reduce the effective mating area. Another problem associated with 
drum brakes is the accumulation of dust in the drums — dust produced by wear 
of the lining material. Erratic braking can sometimes be caused when the leading 
edges of the linings fail to wipe the drum surfaces clean. Such debris cannot 
collect on the exposed surface of a disc brake and the pads are presented at all 
times with a clean braking surface. Wetting of the disc with rain gives no 
appreciable loss of braking effort since the pressures applied to the relatively 
small pads are far greater than those applied between the linings and drums on 
conventional brakes and the film of water is completely removed by the wiping 
action of the pads. 

Fig. 11.3 Girling disc brake caliper - exploded view. 

It is important to stress that the disc brake does not provide any greater rate 
of retardation than the drum brake. This limiting deceleration is set by the grip 
of the tyres on the road and it is not difficult to design drum brakes to provide 
this value for one brake application. The unique feature of the disc brake is its 
ability to provide this braking effort time and time again without any decrease in 

An exploded view of a typical Girling disc brake caliper is shown in Figure 
11.3. The action of the disc brake is simple. When the brake pedal is depressed, 
hydraulic fluid under pressure from the master cylinder forces the pistons 
towards the disc until the friction pads grip the disc from their respective sides, 

The Brakes 207 

under full hydraulic line pressure. Many of us have experienced how effective 
the method can be on the ordinary bicycle. The heat generated in the disc is 
continuously removed by the airstream, since about 80 per cent of the disc area 
is subjected to a blast of cooling air. This is a much more effective cooling 
system than that available to the drum brake where the heat must first pass 
through the metal drum itself before it reaches the air stream. The limiting rate 
of heat input to a drum brake is about 2.5 b.h.p. per sq in of lining area. For a 
disc brake this limit is about 8 b.h.p. per sq in of pad area. 

Girling Ltd (who absorbed Dunlop’s disc brake division some years ago) 
developed a caliper brake that incorporates hydraulic operation from the foot 
brake and mechanical operation from the hand brake. This design is shown in 
Figure 11.4. It consists of a caliper body that straddles the disc, two opposed 



















Fig. 11.4 Girling rear brakes with cable handbrake operation. 

friction pad levers and an operating lever. The pad levers are mounted on 
fulcrum pins attached to the caliper and linked by an adjustable cross bolt. One 
end of the cross bolt is pinned to the outer pad lever and the opposite end is 
secured to the nut of the self-adjusting mechanism housed in the operating lever. 
Connected to the other end of the operating lever are the handbrake cable and 
the operating cylinder assembly. The friction pad assemblies are housed within a 
steel frame bolted to the caliper. Any tendency for the pads to tilt is resisted by 
the right-angled backing plates moulded to the friction material. 

After each brake application the retraction plates withdraw the pads off the 
disc to maintain a constant small clearance between the pads and the disc. The 

208 The Sports Car 

operating lever incorporates a ratchet mechanism that automatically adjusts the 
pad clearance as the friction material wears. 

Disc brakes for the high-speed sports cars 

The disc brakes on the ‘D’-type Jaguars used a hydraulic servo pump driven by 
the gearbox to step up the pressures in the system to a much higher level than in 
a normal hydraulic system. Disc brakes have been successfully applied to small 
light European sports cars without recourse to servo mechanisms or special 
pumps. The larger European sports or G.T. cars such as the Aston Martin, the 
Jaguar and the Ferrari, now use a vacuum servo mechanism to boost operating 
pressures on 4-wheel disc brake systems. When Zora Arkus-Duntov decided the 
time had come to fit disc brakes to the Corvette he decided to design a system 
that would stop this 1600 kg (3500 lb) car effectively without any servo 
assistance of any kind. The brakes were developed by the Delco Moraine 
Division of General Motors and are notable for the use of a ventilated disc. In 
this way the effective cooling area of the disc is doubled. The better cooling 
afforded by the vented disc is shown by the curves of Figure 11.5. The left-hand 

Fig. 11.5 Improved cooling of pad material with vented disc. 

curves are the lining temperatures during 25 repeat stops of 0.62 G deceleration 
at one mile intervals. The right-hand curves show how the lining temperatures 
rise to new higher levels when the deceleration rate is stepped up to 0.77 G. The 
final temperature of the vented disc was 130°F (72°C) lower than the solid disc. 

The development work that led up to the ventilated disc brakes on the 
fabulous turbocharged Type 917/10 Can-Am Championship-winning Porsche is 
a fine example of recent practice in racing sports car design. With such 
experience behind them no one who buys a production Porsche in the future 
need fear that the performance of the car will be allowed to outstrip its ability to 
stop! The earlier unsupercharged Type 917s were unofficially timed on the 

The Brakes 209 

Fig. 11.6 Porsche-designed front brake of Type 917/10 Can-Am car. 

Mulsanne Straight at Le Mans in 1971 at 240 m.p.h. The development engineers 
were well aware by 1969 that the solid disc brake had reached the limits of its 
heat dissipating capacity. In the Daytona 24-Hour Race in 1970 they used a 
ventilated disc fabricated from an aluminium alloy spider that formed the fan 
blades with copper-chromium alloy discs rivetted on each side. They still 
experienced brake fade. 

The solution, which has proved adequate for the 850-950 b.h.p. turbocharged 
racing cars is shown in Figure 11.6. For lightness the ventilated disc is attached 
to the steel wheel hub via an aluminium alloy conical member (distinguished by 
the larger lightening holes in the photograph). The disc itself is in cast-iron with 
curved vanes inside the two flanges. Drilling of the disc flanges (the ‘Gruyere’ 
design in the factory jargon) was originally undertaken to reduce unsprung 
weight. It was found that the modification improved pad cooling and main¬ 
tained a high coefficient of friction of almost constant value over a very wide 
temperature range. Water and mud could be cleaned from the surface much 
more effectively and the wear pattern on the pads was much more regular. 

210 The Sports Car 

Ventilated disc brakes now feature as standard equipment on the Carrera and 
the Turbo model Porsches. 

Pad materials 

Makers of brake-lining materials are now able to produce a formulation for the 
brake pads that will maintain a fairly constant coefficient of friction of about 
0.4 up to a temperature of at least 500°C (930°F) and over a wide range of pedal 
pressures. The new materials have much longer life than the earlier pads and do 
not give any undue wear or scoring of the discs. Brake squeal is usually caused 
by vibrations set up in the calipers. It can be reduced by a change in pad 
material, but a permanent cure can only be made by a redesign of the caliper. 

To achieve these results conventional brake-lining materials are still used, 
these being asbestos fibre, mineral or metallic oxide fillers and an organic 
bonding agent. Sintered metal pads of identical material to the drum-type 
linings that have worked so well under racing conditions on drum brakes did not 
perform well on disc brakes. Durability was poor and the high thermal 
conductivity caused them to conduct too much heat to the brake fluid, thus 
introducing a danger of vapour lock from vaporized brake fluid, which of 
course can lead to a complete loss of brake effort. Much research is still to be 
done in this new field. 


'Often think of the rapidity with which 
things pass and disappear.' 


Standards of performance 

Standards are changing all the time, in the way we live, in our morals, in our 
automobiles and in the performance we expect from them. Many a seemingly 
modest young lady will appear in public today in clothes that would have led to 
her arrest for indecent exposure fifty years ago. And so it is with the speed and 
acceleration of our cars. When the author, as a young schoolboy, first crashed 
through the magical one-mile-a-minute barrier as he crouched down in the 
passenger’s seat of his father’s Studebaker tourer, he experienced not only a 
great sense of achievement but a certain feeling of guilt. Was it right to travel at 
this speed on the public highway, even without another car in sight? Today this 
same middle-aged schoolboy has to move over in to the slow lane if he wishes to 
dawdle at 60 m.p.h. 

It is intriguing to compare the fabulous performances of the giants of the past 
with those of today. Fifty years ago the 41/2 litre Bentley, representing the best in 
British sports car design would accelerate from zero to 60 m.p.h. in about 15 
seconds. Today, a perfectly ordinary family car of only one-third the engine 
capacity of the Bentley will reach 60 m.p.h. from zero in about the same time. 
Larger engined sports cars, such as the Aston Martin, the Maserati Bora and the 
Ferrari Dino will reach 60 m.p.h. in about 6 seconds. The mind boggles when we 
turn to the acceleration potential of modern racing machinery. The following 
figures were calculated by the factory computer for the Porsche turbocharged 
Type 917/10 Group 7 racing car with the turbocharger controls set to produce 
maximum power (about 950 b.h.p. DIN): 

0 to 60 m.p.h. ( 96 k.p.h.).2.1 seconds 

0 to 100 m.p.h. (161 k.p.h.).3.9 seconds 

0 to 200 m.p.h. (322 k.p.h.).13.4 seconds. 

212 The Sports Car 

These figures were supplied by Dr Fuhrmann, Porsche’s Managing and 
Technical Director and we have no reason not to believe them. No driver would 
indulge in such fireworks in a race since the tyre wear would be unacceptable. It 
is interesting to note however that this fantastic Porsche would reach 200 m.p.h. 
before the old Bentley had reached 60 m.p.h.! 

The meaning of power 

Before considering the performance of modern sports cars we must refer briefly 
to that pampered and rather sluggish horse that contributes to the measurement 
of the ‘SAE gross horsepower.’ Corrupted by a horse-power race in the sixties 
the American motor manufacturers were content to quote horsepowers in the 
Press using an SAE approved method in which power was measured on the 
dynamometer with the dynamo and cooling fan disconnected, the exhaust front 
pipe exhausting into an underground ducting maintained at a pressure slightly 
below atmospheric pressure and the ignition timing set to give maximum power, 
with no regard to the actual ignition advance curve dictated normally by the 
distributor. The final cheat was the application of a correction in which the 
engine was motored over with water and oil at normal operating temperatures to 
measure the power absorbed by internal friction and pumping losses. This 
friction horsepower was then added to the previously measured value to give the 
completely fictitious figure called the ‘gross SAE horsepower.’ 

This sort of chicanery allowed claims of 400 ‘horsepower’ for engines that 
gave little more than 300 b.h.p. at the flywheel when fitted into the car. Several 
European manufacturers used the same methods for many years. 

There is an air of realism abroad in Detroit now and the net SAE horsepowers 
they quote today are very close to the values measured on British, French, 
German and Italian test-beds. The Deutsche Industrie Norm (DIN) horsepower 
has always been a fair one and the author has adopted this one in this chapter 
when the value is available. The International Standards Organisation (ISO) 
have set up another slightly different procedure and one can only hope that this 
will eventually be adopted by all countries. 


What determines a given rate of acceleration? Is it power to weight ratio — is it 
torque to weight ratio — or is it a mixture of the two? It will be useful, before we 
look further into this question, to clarify our ideas on horsepower and torque. 
Horsepower is a rate of doing work, i.e. the time element is involved. It is 
measured in units of work per unit time. Torque, on the other hand, is a twisting 
moment, simply a force times a distance. If a rear wheel is 12 inches radius and 
is exerting a force of 400 lb at the road surface to propel the vehicle forwards, 
the torque transmitted is 4,800 lb in. or 400 lb ft. If the wheel is turning at a 
speed of 500r.p.m. the number of radians turned through in a minute while the 
torque is acting will be 27rx500. The rate of doing work is, therefore, 
2n x 500 x 400 ft lb per minute. The horsepower per wheel is therefore 

Performance 213 



= 27TX400X500 
= 38.1 

If the same torque of 400 lb ft were transmitted to the road surface by the 2-ft 
diameter wheel at twice the speed, i.e. 1,000 r.p.m., the power transmitted 
would be double, or 76.2 b.h.p. 

Let us now consider what happens to the power of the engine on its way from 
the flywheel to the driving wheels. In top gear one would not expect to lose 
much of the power in the gearbox, but the churning of the oil can still absorb a 
little power, as can be verified by feeling the rise in temperature of the 
gearbox-casting after a long run. The transmission efficiency of the hypoid gear 
in the rear axle will be 94-95 per cent and in all about 7 per cent of the power 
leaving the flywheel will disappear in the form of heat and noise before it 
reaches the road wheels. In the intermediate gears about 88-90 per cent of the 
engine power will perform useful work. 


Fig. 12.1 Comparative power and torque curves of Indianapolis engines. 

The Ford engines were # used later in the Lotus 30 competition 
sports car and the G.T. Fords. 

Figure 12.1 gives two examples of the power and torque curves for the highly 
tuned V-8 engines developed specially by the Ford Motor Company for the 1963 
and 1964 Indianapolis Lotus-Fords and used later in the Lotus 30 and the Ford 
G.T. sports cars. It is of interest to compare the performance of these short 
stroke, medium bore V-8s, one with push-rod operated valves, the other with a 

214 The Sports Car 

twin o.h.c. head to each bank and 4 valves to each cylinder, and the 
near-culmination of the Meyer Drake Company’s development of the big-bore, 
long-stroke 4-cylinder Offenhauser engine. The bore and stroke of the Ford V-8 
is 95.6 mm x 73 mm, of the Offenhauser 109 mm x 111 mm. The push-rod 
Ford gave slightly less power, the d.o.h.c. Ford slightly more power than the 
1963 Offenhauser, but the big 4-cylinder Offenhauser develops peak power at a 
much lower speed and the torque is much superior. Such an engine will give 
greater acceleration, for acceleration is dependent on torque, as we shall 
demonstrate later in this chapter. 

It was shown in our companion volume Design of Racing Sports Cars that a 
good approximation to the mathematical relationship between bhp and the 
cylinder dimensions is 

Max. b.h.p. a d 1-65 j 0-5 

where d is the cylinder bore 
s is the stroke. 

When the stoke/bore ratio is constant, 

b.h.p. <*d 2 ' 15 

or approximately, 

b.h.p. « piston area. 

The torque, however, with a given fuel and cylinder head design and 
compression ratio is directly proportional to the swept volume. 

With a constant stroke/bore ratio 

torque °c d 3 . 

For our purpose it is more convenient to measure engine size in terms of the 
swept volume, V. 

In terms of this then, 

b.h.p. oc y\ 
torque V. 

Based on this relationship we would expect a series of DOHC engines as used 
in modern sports cars to give the values for maximum power and torque given in 
Table 12.1. 

The 6-litre engine is seen to have three times the torque of the 2-litre, but 
only slightly more than twice the power. Every engine tuner looks for higher 
‘revs’ if he wants more power. In this manner some engine tuners obtain quite 
impressive increases in power from otherwise standard engines. The torque, 

Performance 215 

Table 12.1 Approximate variation of power and torque in high-performance sports 
car engines 



cu in 

Net power 


Net torque 
lb ft 
































however, does not benefit from increased r.p.m. as such, but only from the 
improved breathing that has led to the higher engine speeds. Putting this 
another way, an increase in volumetric efficiency from the use of larger inlet 
valves, may increase the maximum torque by, say, 10 per cent. The power curve 
will now peak at 15-20 per cent higher r.p.m. and the overall effect will be an 
increase in maximum power of as much as 30 per cent. 

When engines are designed for different basic duties it is possible for two 
engines to have the same power output but very different torques. The pre-war 
American engine was designed for the lazy driver who wanted to get into top 
gear as soon as possible and to stay there, only using the lower gears on steep 
hills and when brought to a standstill at traffic lights. A single small choke 
carburettor was always used and the inlet valves were correspondingly small. 
Good torque at low speeds was obtained at the expense of maximum power. The 
power curve peaked at about 3,500 r.p.m. A typical 4-litre engine of the period 
would develop 100 b.h.p. with a maximum torque of 190 lb ft. A modern 
1 Vi -litresports engine could deliver 100 b.h.p., but the torque would probably not 
exceed 90 lb ft. If we took two similar cars with similar bodies and installed the 
two engines, the 4-litre American engine and the Wi -litre sports car engine, in 
such a way that the kerb weights were identical, they would give identical top 
speeds, provided that the right rear axle ratios were chosen. The American 
engined car however would out-accelerate the sports car engined car, and would 
cut the acceleration times of the latter car by two. 

Acceleration times for 0-60 m.p.h. 

It would be of value, especially to the ‘special’ designer, if we could estimate the 
probable acceleration to be expected from a given specification of car. For 
general road conditions the most important acceleration range is 0-60 m.p.h. 
Let us then see if we can devise an empirical formula giving a fair approximation 
to the time taken to accelerate through the gears to 60 m.p.h. on any given car. 

The basic acceleration formula, from simple mechanics, is in self-consistent 

P = Mf 

where P = the force producing acceleration, 

216 The Sports Car 

M = the mass to be accelerated, 

/ = the acceleration. 

The force to produce acceleration, in our case, is directly proportional to the 
excess torque; the excess torque being the torque remaining when we deduct 
from the total torque the torque absorbed in overcoming road resistance and air 
resistance. Over the speed range of 0-60 m.p.h. only a small percentage of the 
total torque is absorbed in overcoming the resistance to motion. At 60 m.p.h. on 
a typical sports car rolling resistance will be about 30-40 lb; the air resistance 
about 70-80 lb — a total of about 110 lb. If we take a typical engine torque of 
120 lb ft and a second gear ratio, overall, of 8 to 1, a transmission efficiency of 
90 per cent and a tyre diameter of 26 inches, the propelling thrust at the rear 
tyres is 

-= 800 lb. 


The excess thrust is therefore 800—110 = 690 lb or 86 per cent of the total. 

At 30 m.p.h. the total resistance will be about 30 lb. The excess thrust is 
therefore 800—30 = 770 lb or 96 per cent of the total. 

With a bigger engined car the excess torque will be a higher percentage of the 
total. For a car like the Aston Martin it will rise to about 95 per cent of the total 
over this speed range and for a smaller engined car such as the M.G.B. it will fall 
to about 80 per cent. In general, however, it seems reasonable to take P into the 
formula as being proportional to T, the total engine torque. 

M M 

where T is the maximum engine torque, lb. ft. 
Our formula then becomes: 

Time in seconds, 0-60 m.p.h. <*M/T 
or, if we take the car test weight in lb = W 

to-6o = KW/T (1) 

However, when we try to apply this formula, it is soon apparent that 
something has been neglected. Let us test it by choosing a value of K = 0.4 to 
give agreement with the road-test value for the modest powered Alfa Sud. As we 
use this value to estimate the performance of higher powered sports cars the 
error increases with each increase in power to weight ratio. The table opposite 
illustrates this. 

Errors of this magnitude cannot be dismissed as the effects of wheel-spin. It is 
in fact caused by making no allowance for the losses from internal accelerations, 
the torque absorbed in accelerating the rotating and reciprocating components 
in the engine, the rotating parts of the clutch, gearbox and final drive. Even the 

Performance 217 

Make and Type 

Time to accelerate 0 to 60 mph 

by road-test by formula 




Alfa Sud L 




Lotus 7 








Maserati Bora 




Porsche 917 (1971 Le Mans 




wheels and tyres must be accelerated in a rotary sense as well as a linear sense. In 
the case of a high-powered sports/racing car such as the Porsche Type 917 
approximately 50 per cent of the total torque that would be available in steady 
motion is lost in rotational accelerations, leaving only about 50 per cent for 
linear acceleration of the car’s mass. 

The concept of effective mass 

Our simple Newtonian formula of P = Mf must be replaced by a more 
sophisticated concept if we are to make an accurate estimate of the acceleration 
potential of any new design of sports car at the drawing board stage, especially 
if the projected specification is for a high-performance vehicle. 

The author is indebted to J.L. Koffman (Automobile Engineer, December, 
1955) for the concept of effective mass. In the simple Newtonian acceleration 
formula Mis replaced by M 1 , where 

1 +OC+0) 

oc is a correction factor for the polar moment of inertia of the engine. (This 
correction factor is dependent upon the overall gear ratio, being higher in 1st 
gear than in top gear.) 

j3 is a correction factor for the polar moment of inertia of the road wheels. 
The polar moment of inertia of the gearbox, propeller shaft and final 
drive gears are all very small and can be neglected. 

To apply the Koffman formula to any vehicle we need the following 

(1) the net torque curve for the engine, 

(2) the gear ratios, 

(3) the rolling radius of the driving wheels (this increases with speed), 

(4) the rolling resistance of the car (the road resistance plus the air resistance), 

(5) the test weight of the car, 

(6) an approximate value for the polar moment of inertia of the engine, 
flywheel and clutch. 

The author was able to collect the above data for a Jaguar XK150S and to 
calculate the acceleration times through the gears from 10 m.p.h. to 120 m.p.h. 

218 The Sports Car 

The calculated values give remarkably close agreement with the road test values 
taken from The Autocar test of September 18, 1959. The influence of the polar 
moment of inertia of the rotating masses on the performance of this car can be 
judged from the following figures. The true mass of the XK150S, the test 
weight, was 3,590 lb. The effective mass in each gear was calculated to be: 


M* in first gear 

= 5,060 

M* in second gear 

= 4,130 

M in third gear 

= 3,880 

M in fourth gear 

= 3,840 

M in overdrive 

= 3,770 

The effective mass in first gear is 141 per cent of the true mass. Thus in first 
gear about one-third of the torque developed at the crankshaft is absorbed in 
accelerating the engine, the clutch and all the rotating masses, even the road 
wheels. The actual polar moment of inertia of the road wheels is usually very 
high, especially on a big-engined competition vehicle with large tyres. The 
rotational speed of the wheels, however, is only a small fraction of that of the 
engine. In first gear on the XK150S the overall ratio is 12.2 to 1. In this gear the 
torque absorbed in rotating the wheels is only about 3-4 per cent of the total 
engine torque. In overdrive a higher percentage of the ‘lost’ torque is used in 
rotating the wheels. Of the total lost torque of 5 per cent, V/i per cent is 
absorbed in rotating the engine and IVi per cent in rotating the wheels. 

The Jaguar XK150S was no mean performer for its day, and could accelerate 
from 0 to 100 m.p.h. in 22.4 seconds. Several modern sports cars can halve this 
acceleration time. When this is achieved by using a highly tuned 3-litre engine 
instead of a more modestly tuned 5-litre engine less power is wasted in 
accelerating the rotating and reciprocating components. In this sense the smaller 
engine can be said to be more efficient. 

It can be shown that the effective mass is highest when 

(a) the ratio engine r.p.m./m.p.h. is high 

(b) the ratio engine polar moment of inertia/total mass is high. 

It can also be shown that every car has a limiting effective bottom-gear ratio. 
At this limiting ratio any change to a lower ratio (actually to a higher ratio when 
expressed as a numerical value) would increase the effective mass more than the 
gain in torque given by the ratio change. The overall effect would thus be a 
reduction in the rate of acceleration. For the SK150S the limiting effective gear 
ratio would be 17 to 1. The actual bottom-gear ratio of 12.2 to 1 is well inside 
this limiting value. For many modern racing/sports cars the limiting ratio will 
approach 10 to 1 and the designer of such projectiles should beware that he does 
not exceed this limiting effective ratio. Even though the bottom gear is a ratio 
only used once, from the starting line, the use of a ratio that is higher 
(numerically) than the limiting value will only result in a drop in acceleration. 

While the Koffman method is the most accurate means available to 
the designer or project engineer for calculating the acceleration of a 

Performance 219 

modified design over any desired speed range, the method is very laborious. 
Moreover many of us do not have access to the necessary data. A much simpler, 
but admittedly less accurate, method has been devised by the author. 

Table 12.2 is based on the following formula: 

t 0 -6o = (2W/T) 0 ' 6 

where W = the test weight in lb. 

T = the max. engine torque, net in lb ft. 

A comparison of the formula values and the Road Test values shows good 
agreement in the majority of cases. 


There was a time when one could take out a high performance sports car on the 
public roads and find out its top speed in reasonably safety — but not so today. 
Today with maximum speeds often approaching 150 m.p.h. and in a few models 
even exceeding 200 m.p.h. to use the public roads for this purpose is asking for a 
brief notoriety in the daily papers, perhaps even a longer spell in one of Her 
Majesty’s prisons! The testing staffs of the motoring journals are now finding it 
so difficult to make accurate measurements of this important yardstick of 
performance that they sometimes substitute an estimated value in their reports. 

In Chapter Nine in the section on Drag coefficients we introduced the general 
formula for drag horsepower: 

HP drag 

C d Av 3 

where Q = drag coefficient 
A = frontal area, sq ft 
v = speed, m.p.h. 

The horsepower absorbed in aerodynamic drag is about 70 per cent of the net 
b.h.p. at the flywheel. Thus if we know the net maximum power of an engine, 
the frontal area of the car and can make an inttelligent guess at the drag 
coefficient, this formula will give us a good approximation to the maximum 
speed. Since v varies as the cube root of these quantities no great accuracy is 
required in our estimate of Q or A or even the net horsepower. For most 
modern shapes A can be estimated fairly accurately at: 

0.9 x width X height 

For older sports cars and particularly for cars with separate wings like the 
Lotus Seven and the Morgan the constant should be reduced to 0.8. To help the 
reader in his choice of a value for C d when estimating the probable performance 

Table 12.2 Acceleration time 0—60 mph (0—96 kph) 














'O i/) 00 ON 00 On vd t^’ VO t^’ 00 On C"* 


^ © oo oo vo uo t> o co r»* oo vo vo 

in vo oo vo vo d t"** oo* oo* vo* r- \6 r-*’ oo’ iri os’ r- 




















oo\ | o^a\r)ONO\'tfoo\^xhn^h 
Nt n -H CN fS rs 






> CO 
CO ® 

« X 

a > 

■s U © 

ed -m 00 

s % " 

n > 3 
O > CO 

+H 1) +J 

V3 ^ cd 

< U Q 













s <? s 

O H-, <u 

cn x « 

’S b . 

cd cb 
£ 3 2 

*h hn c 






bo *3 

•pn C 3 

W co .ti 

c c 

•5 « 
si > 
bO S _ 
g 00 M 

2 a 

o ._ 

Ph Os KSl 


E 3 

3 3^ 

- O O 
J J J 

iJ 1 : ^ « 




bO o 

O O 3 

S Ph Pd 

t"* .J 

Dd S 
H 2 

a ^ 


H H 

*1968A/oro/* road test 
f Estimated at 85% of SAE torque 

^Estimated by Motor staff (figures not available from manufacturer) 

Table 12.3 Maximum speeds and drag coefficients 

•« §> 
*3 ^ 

S r? 

o o o o o o o o o o o o o o o o o 



c >3 

5 a 
S s 






(N(N’- h (N(N’- h (N-h 














r- (N 
(N (N 

a\(Na\^ H mO(N(Nmm^H'ooo'0 
oo^HOO'O v o<NmO(NC'iomr'-o 
-h(N (N -h (N -h -h -h -h (N 


i/N O ON i/N i/N ON 

rf rf T$- IT) OO 

m m *-h (N (N *-H 



m i—i co »—i »—i »—i »—i cn *—< *- 1 





ma\a\m*-Hcna\a\ v O‘nTtL0*-H*-H00<NL0 



o do o 

uo r- 

o oo oo r- o\ oo on 

OC VO 00 Q\ t 00 NO 

























O N 
U O 

+j oo 
^D (N 

o c 

b S3 



* £ 

H ^ 

o w 

oo «J S 
O I—» D 
m x pq 
*£ *h cd 

s 3 g 

b oo 5 



cd tJ- 

•a « 


w 5 

w C/3 


2 C/3 






c g 

*3 « 

JG > _ 

00 a) c/3 

g CO ffl 

r M M L 
C 3 3 <£ 

C -M -M CQ 
Cd O O ^ 
hJ hJ hJ a 

> 2 
H * 






ON ^ 

g Jc c S 
ff 8 ^ H 

I £ (2 •§ 

H -= 

o, § 



H K 

*1968 Motor road test 
f Estimated at 85% of SAE horsepower 

=f Estimated by Motor staff (figures not available from manufacturer) 

222 The Sports Car 

of a new design of car Table 12.3 has been compiled from published road test 
data. From this we see that a typical modern sports car has a drag coefficient of 
about 0.4. The best examples, invariably the result of painstaking testing in the 
wind-tunnel, give values approaching 0.3. A typical design of pre-war sports 
car, with flared wings, a slab fuel tank at the rear, a flat screen with only about 
20 degrees rake, but with hood erect and side screens in place would have a value 
for C d of 0.7 to 0.75. The Lotus Seven is in many respects a replica of this older 
design, but with modern materials and design techniques the weight is almost 
halved. Where the object is sheer acceleration good handling and a modest top 
speed the Lotus Seven formula still makes sense. If only we could persuade 
Colin Chapman there is still a market for an updated Lotus Seven perhaps he 
would design one, still at a kerb weight of about 1700 lb, but this time with a 
drag coefficient of 0.4 to 0.5. 

The sports car in the future 

'We are all working together to one end, 
some with knowledge and design, and others 
without knowing what they do.' 



However much we may hate them we cannot escape the consequences of exhaust 
emission regulations on the future of the sports car engine. American car 
manufacturers have staggered from one crisis to another as the Federal emission 
standards have been tightened in successive stages to levels that have resulted in 
the addition of catalytic converters containing expensive and not very durable 
materials, air injection pumps, exhaust gas recirculation devices and very 
complex fuel injection systems. To add to the driver’s misery some of these 
changes have resulted in poor drivability in traffic and an increase in fuel 

Most governments in Europe have adopted the ECE specification known as 
ECE15. These standards are not as severe as the U.S. Federal or the Californian 
standards, since smog is not yet such a problem in our cooler climate. Even so 
several European sports car manufacturers have found the North American 
market quite profitable in the past and the larger manufacturers, such as 
Leyland and Mercedes-Benz and some enterprising small firms such as Lotus 
and Porsche are loth to abandon the market. If the Americans can do it — so 
can we! 

It is difficult to make a direct comparison between the American standards 
and those specified in ECE 15 since the driving cycles used in the acceptance tests 
of acceleration, cruise, deceleration, braking and idling are, following the inane 
decisions of most large committees, completely different. Sweden and Australia, 
however, have adopted the 1973 Federal test method and Table 13.1 serves to 
show the wide gulf that exists today between the Federal and Californian 
standards on the one hand and the rest of the world on the other. There is some 
consolation in the fact that the small-engined car is now favoured by a test 

224 The Sports Car 

method that measures emissions in grams per mile. The seven-litre gasoline 
guzzler carrying a single commuter into the city must eventually disappear from 
the scene. 

Table 13.1 

Exhaust emission standards 



NO x 


(Hydrocarbons) (Nitrogen oxides) 









Present U.S. Federal 




Present Californian 




Proposed 1978 U.S. Federal 





We know that many of our brightest automobile engineers are not at all happy 
when they consider the problems of coaxing conventional petrol engines 
through the 1978 Federal test schedule. They are less happy when they think of 
the work involved maintaining these standards after the car has left the factory. 
Some even predict that the engine that has served us so well since the be ginnin g 
of the century will be replaced by an entirely different engine within the next ten 
years. The automobile industry has not ignored these predictions. Large scale 
programmes have been undertaken to investigate all promising alternatives. At 
the same time they have not neglected to try out any feasible schemes to improve 
combustion and reduce the exhaust emissions from the conventional engine. 
The capital investment in the machinery to make the conventional internal 
combustion engine throughout the whole world is an astronomical figure. This 
will not be cast aside lightly. It will be of some value at this stage if we make an 
appraisal of all the promising alternatives. 

The gas turbine 

In 1951 Air Commodore F.R. Banks, in his James Clayton Lecture to the 
Institution of Mechanical Engineers, gave the following summary of the 
position of the automobile gas turbine at that time: 

In the first place, the most suitable size of gas turbine to give reasonable 
efficiency is better fitted to the needs of the larger vehicle rather than the 
automobile — since it is more easy to build an efficient gas turbine of 250 shaft 
horsepower than one of 50 or 100 shaft horsepower. Scaling the engine down to 
these comparatively low powers demands lengthy and expensive development, 
to obtain the required efficiency of components such as the compressor, the 
combustion chamber, and the turbine. 

The Sports Car in the Future 225 

Smoothness of operation and lack of vibration, inherent in the gas turbine, 
are now so good in the piston engine that a change to the former on these 
grounds alone could hardly be justified. 

Working principle 

For traction applications the gas turbine is fundamentally different from the 
aero gas turbine. Figure 13.1 serves to illustrate the ‘two-turbine’ principle as 


used in car and truck applications. The air/fuel mixture is burned in a 
combustion chamber and the resulting expanding gases are used to drive the first 
turbine, the compressor turbine. The compressor turbine provides the power to 
drive the compressor which compresses the air before it enters the combustion 
chamber. This compression of the intake air is as necessary in the gas turbine as 
in the piston engine; the greater the compression ratio, the more efficient the 
conversion of heat energy into useful work. The hot gases leaving the first 
turbine are expanded further through a second turbine, this being the power 
turbine which is connected through suitable gearing to the driven wheels of the 
vehicle. To improve the efficiency of the power unit the hot gases leaving the 
power turbine are passed through a heat exchanger or ‘regenerator’ as it is 
usually called by turbine engineers. This extracts heat from the gases before they 
are exhausted to the atmosphere and raises the temperature of the air entering 
the combustion chamber. The reader will have realised a fundamental difference 
between the gas turbine and the piston engine. In the former the processes of 
compression, combustion, expansion and exhaust are continuous , in the latter 
they are cyclical. 

When we analyse the evidence reported during the 26 years since Air 
Commodore Banks made his authoritative statement we find that, despite 
considerable progress in this field, the goal of a commercially viable gas turbine 

226 The Sports Car 

to replace the automobile piston engine still eludes the thousands of engineers 
involved. General Motors, Ford and Chrysler in America and Leyland in Great 
Britain have spent millions of dollars on the development of the gas turbine. 
Only in the field of trucks and buses, where the size-factor, stressed by Banks, is 
in their favour, have they made any real progress. One of the current projects 
sponsored by the American Environmental Protection Agency is a seventh 
generation Chrysler automotive gas turbine. If all goes well full-scale produc¬ 
tion of this engine is planned for 1983. 

The gas turbine gives very low hydrocarbon and carbon monoxide emissions 
and the problem of high concentrations of nitrogen oxides (NOx) that caused so 
much concern about five years ago has been solved by vaporising and 
pre-mixing the air/fuel mixture fed to the combustion chamber. Great advances 
have been made in the development of ceramic materials for the turbine blades 
thus permitting maximum operating temperatures to be raised from about 
1000°C (1830°F) to about 1350°C (2450°F). This makes the gas turbine much 
more efficient and economical. 

Fig. 13.2 GM Detroit turbine with power-transfer system giving economy and 
engine braking. 

There remains one very challenging obstacle to the future of the gas turbine, 
the problem of cost. The mass production of investment castings for the 
aluminium alloy impellors, the development of inexpensive techniques to mould 
the turbine blades in sintered ceramix-metal mixtures, the development of cheap 
reliable regenerator cores; these and many other problems still occupy many 
fertile brains. It is very discouraging, but the small or medium-size gas turbine 

The Sports Car in the Future 227 

for automotive use seems very little nearer to-day than when we reviewed the 
position in 1969. 

The application of the gas turbine to heavy road transport vehicles has made 
rapid strides in the last decade. A schematic layout of the 400 shaft horsepower 
turbine made by the Detroit Diesel Division of General Motors is shown in 
Figure 13.2. These engines have operated successfully in Greyhound long¬ 
distance buses. A power-transfer system is used to connect the power turbine 
shaft and the compressor (gasifier) turbine shaft under the automatic control of 
a ‘torque sensor*. This system, not only gives the very desirable feature of 
overrun braking, but improves acceleration and economy. 

Rotation combustion engines 

It is 22 years since Felix Wankel invented his remarkable rotary engine. Today 
there are firms all over the world working under licence on the development of 
this engine. From Figure 13.3 we see that the ‘rotating piston* is a three-sided 
rotor, eccentrically mounted on the output shaft and geared to it by an internal 
gear and pinion with a ratio of 3:1. Thus the output shaft makes three 
revolutions for every revolution made by the rotor. Both rotate in the same 
direction. Chambers are formed between the specially shaped outer casing (the 
geometric shape is called an ‘epitrochoid*) and the rotor, these chambers 
varying in volume as the rotor turns, thus giving alternate compressions and 
expansions as in the reciprocating engine. Intake and exhaust ports are provided 
in the periphery of the casing or in the end walls. Figure 13.3 will help the reader 
to understand how the 4-stroke cycle is obtained. Three expansion (working) 
strokes occur for every revolution of the rotor. Since the output shaft makes 
three revolutions in this time period only one working stroke occurs per 
revolution of the output shaft, which is the same number that occurs with a 
2-cylinder 4-stroke reciprocating engine. The decision of the FI A racing 
committee to rate the Wankel engine at twice the swept volume of one of its 
chambers is therefore logical. 

It will be seen that ignition is provided by the same sparking plug for 
successive firings. Conditions are obviously arduous for the single plug, since it 
receives no cooling from the induced charge as in a conventional engine. To 
prevent leakage across the rotor tips it is necessary to limit the size of the hole 
between the plug socket and the combustion chamber. 

Gas sealing at the three tips of the rotor and at the end faces was a difficult 
problem in the early Wankel engines. Some designs suffered from excessive 
wear rates, others would flutter at a critical speed and this fluttering would soon 
gall the surface of the chamber and destroy the gas seal. Effective seals have 
now been developed by several of the companies working in this field, the 
sealing usually being least effective at low speeds. The intersection of the apex 
seals and the end-face seals on the NSU-Wankel engine is shown in Figure 13.4. 
The apex seal blade is free to slide in a radial direction, the inner portion being 
guided by a hardened steel bolt. The end-face seals extend from apex to apex 
and provide an interlock for the bolt. All parts are preloaded towards the faces 

228 The Sports Car 







Fig. 13.3 Four-cycle operation of Wankel rotating combustion engine. An 
eccentrically mounted three-lobe rotor, turning within an 
epitrochoidal combustion casing, drives a centrally mounted 
output shaft through an internal gear and pinion. The enclosed 
volumes at A, B and C are successively expanded and compressed 
in the manner of the four-cycle piston engine. The single 
sparking plug carries a very high heat-load since it fires three 
times per revolution of the rotor, or once per revolution of the 
output shaft. 

The Sports Car in the Future 229 

Components of the apex 
sealing system on the 
NSU-Wankel rotor. 

to be sealed by corrugated springs. These springs do not provide sufficient 
pressure for sealing, this being provided by gas pressure acting behind the 
blades, as in the piston rings in a conventional engine. 

Cooling of the outer casing is by water jackets, but the rotor is oil-cooled, 
being fed under pressure through the hollow shaft and by various passages in the 
hollow centre of the rotor. Circulation is maintained by means of a stationary 
extraction scoop. 

Toyo Kogyo, the Japanese Company who make the Mazda car, have worked 
very hard on the many problems associated with the rotating combustion engine 
and have now produced a smog control system for their latest engine that 
operates on a lean mixture. This confounds many American experts who have 
been insisting on the need to operate on the rich side of a chemically-correct 
mixture in order to reduce the nitrogen oxides formed during combustion and to 
use a catalytic muffler or thermal reactor to burn away the excess hydrocarbons 
and carbon monoxide. The Mazda thermal reactor is maintained at a high 
temperature by means of an exhaust gas heated jacket and an efficient insulating 
cover. The former criticisms that the Wankel engine is uneconomical and 
unreliable could therefore be challenged by the New Mazda, but the weight of 
invested capital sunk into the manufacture of piston engines is still the greatest 
threat to this ingenious engine. Even so, if Toyo Kogyo can give a convincing 
demonstration that their engine is superior to the current piston engines, the 
changeover will eventually occur. 

The diesel engine 

The engine associated with the name Dr Diesel operates on the principle of 
‘compression ignition’. By this we mean that the compression ratio is so high 
that the temperature of the air towards the end of the compression stroke is high 
enough to ignite a spray of fuel injected a few degrees before TDC. No sparking 

[- >3.5 Cutaway view of the Mazda twin-rotor engine. Note the use of twin sparking plugs to each combusion chamber. 

The Sports Car in the Future 231 

plug is, of course necessary. A Diesel engine runs completely unthrottled, taking 
in a full charge of air at every induction stroke. Power is regulated entirely by 
the amount of fuel injected near TDC at the beginning of the working stroke. 
With no throttle to regulate the air flow only about 5 per cent of the oxygen is 
required for combustion at idle. At full power about 90 per cent of the oxygen is 
consumed. Any attempt to increase the metered quantity of fuel beyond this 
point results in incomplete combustion, exhibited by the emission of black 
smoke from the exhaust pipe. A petrol engine consumes 100 per cent of the 
oxygen supplied (neglecting the very small amounts of fuel only partially 
oxidised, i.e. the carbon monoxide). The petrol engine can also operate at much 
higher engine speeds. The speed of combustion in the Diesel engine is not as 

In brief the Diesel engine, for the same overall bulk, produces about 70 per 
cent of the power of a petrol engine. Since the Diesel uses compression ratios as 
high as 20 to 1, heavier wall thicknesses are required in the combustion chamber 
walls and the cylinder walls, and a general increase in crankcase rigidity is 
necessary to withstand the higher loads transmitted to the crankshaft. For 
engines of the same weight we can therefore only anticipate the Diesel to 
produce about 60 per cent of the power of the petrol engine. On the credit side, 
the Diesel is very economical and a Diesel car will use only about two-thirds the 
fuel of the conventional car. This in itself is enough to explain the renewed 
interest shown by the automobile industry in the Diesel since the Middle East oil 
producers began to show their economic strength with such devastating results. 
Today 40 per cent of the cars made by Mercedes-Benz are Diesel-powered and 
many other well-known European car makers offer Diesel engines as an option 
on at least one model in their current range. 

Turbocharging will be discussed later in the chapter but this could be the 
secret potion needed to achieve a successful mating of the overweight Diesel and 
the sports car of the future. Friedrich Van Winsen, the chief development 
engineer at Mercedes-Benz, recently expressed a conviction that a turbocharged 
Diesel could be produced in the near future that would be refined enough to 
meet the exacting requirements of his company’s luxury cars and at the same 
time conform with European pollution legislation . In the author’s opinion, such 
an engine could readily be adapted to power a future Mercedes sports car. 

The Stirling engine 

One of the problems that faced the Reverend Robert Stirling, his brother James 
and his son Patrick, when they tried to raise some interest in the Reverend 
Stirling’s novel engine was to explain the working principle to businessmen with 
no knowledge of themodynamics. This was not surprising since little was known 
in 1843 about the laws that govern the conversion of heat energy into work. This 
was the year when the first Stirling engine was installed to drive the machinery at 
a Dundee foundry. Poor technology and unsuitable construction materials led to 
the comatose state of Stirling engine development for nearly a century. A few of 
the original engines survived as disused pumping engines to interest industrial 

232 The Sports Car 

archaeologists and a few were used in India to replace the punka wallah, the 
man servant who waved a fan from side to side to cool his master. 

Modern materials and an understanding of the thermodynamics of the engine 
revived interest in this engine in 1937 when the N.V. Philips Company of 
Endhoven in Holland began a research programme to study its potential. This 
has grown into a world-wide interest, with General Motors and Ford in America 
and Mann and MWM in West Germany all participating in the search for a 
commercially viable engine. 

The simple idea conceived by this Scottish Presbyterian minister in 1816 is 
illustrated in Figure 13.6. In this illustration no attempt is made to show the 



in cold spa ce 
at bottom 





Gas transferred 

Gas expands 


to hot space by 
movement of 
both pistons 

forcing bottom 
piston downwards 

Fig. 13.6 The Stirling cycle. 

mechanism used to convert the reciprocating motion of the two pistons or to 
arrange the phasing of their relative movement. Many mechanisms have been 
devised and it is difficult today to predict which will prove to be the final 
solution. In the original engine a single cylinder is effectively divided into two 
working chambers by a heat exchanger or regenerator. The upper hot chamber 
and the lower cold chamber each contain pistons which are interconnected in 
some way to produce a pumping action (in many of the original engines) or to 
produce rotary motion. By suitable phasing of the cranks the gas in the lower 
chamber is first compressed, then transferred to the hot side through the 
passages in the regenerator which is designed to have a high heat storage 
capacity. The regenerator contains heat stored from the previous cycle. The 
heated gas enters the hot side which is maintained at a high temperature by 
external combustion of a suitable fuel and the gas expands and does useful work 
on the upper piston. The piston movements are reversed and the gas is 
transferred back through the regenerator. As the lower piston returns to its 

The Sports Car in the Future 233 

original lower position the hot gases, passing through the regenerator passages 
give up valuable heat to the regenerator. Heat is supplied continuously to the 
upper chamber; cooling by water or air is supplied continuously to the cold 
chamber. It will be realised that the Stirling engine like the steam engine is an 
external combustion engine. 

The modern Philips engine uses hydrogen as the working fluid. Hydrogen has 
a high thermal capacity and because of its low viscosity the pumping losses 
through the minute regenerator passages are low. A compact ‘rhombic drive’ 
using contra-rotating cranks geared together gives the required phasing of the 
two pistons and a matrix of small diameter crimped wires acts as a regenerator. 

There are no near-explosive pressure rises inside the Stirling engine; in fact no 
explosions inside or outside the cylinders. The Stirling engine is as quiet as the 
proverbial sewing machine, uses negligible amounts of oil and has shown few 
problems of wear or erosion. The working cycle is close to the ideal Carnot 
cycle. The Stirling engine therefore possesses the potential for a much higher 
thermal efficiency than the petrol engine. The extremely clean exhaust that can 
be given by a modern continuous combustion chamber operating on an excess of 
oxygen would be an added bonus. It is the ability to meet all reasonable future 
exhaust pollution demands that spurs on Stirling engine development. On the 
debit side the Stirling engine is still very heavy and bulky. It requires a radiator 
twice the size of the current petrol engines and is not yet able to cope with 
demands for quick changes in power output. The Stirling engine has much to 
offer as a stationary continuous speed engine, but its future in the automobile is 
still a distant prospect. 

The steam engine 

The steam engine is another power unit with the inherent advantages of a clean 
exhaust. Older Americans will remember the fine steam cars made by Abner 
Doble. His flash boiler used about 180 metres (600 ft) of small-bore seamless 
steel tubing as a boiler and an electrically-driven blower to control the excellent 
combustion of the kerosene/air mixture in the combustion chamber. The Doble 
steam car could be made ready to drive away 40 seconds after starting from 
cold; slow by modern standards but acceptable in the early twenties. 

Two companies are active today in their unfailing belief in the steam engine. 
The Lear Motor Corporation, have an engine installed in a long-distance coach 
that is used for publicity purposes and the STP Corporation, have developed a 
completely sealed (no steam loss) system. The exhaust steam is condensed in a 
jet condenser and after cooling in a large radiator is returned to the boiler feed 
pump as shown in Figure 13.7. All the steam engines produced in recent years 
have been very bulky. When installed in a car they usually require the whole of 
the engine compartment and the boot (trunk) to house the engine, condenser, 
boiler and other auxiliaries. It would be pleasant to see a lightweight steam 
engine in a sports car some day. The excellent low-speed torque would be very 
welcome. Somehow it seems to be many many years away. 

234 The Sports Car 







The STP steam system 
including the jet condenser 
— what they call a 
'closed hydroloop cycle* 






Jet condenser 





Start/Alt \ Boiler 

Piston engine 
Independent suspension 

Fig. 13.7 The STP steam car project. 

The Sports Car in the Future 235 


Stratified charge 

Otto patented a system in which combustion starts in a rich mixture and 
progresses into a lean one. His aim was to reduce the shock-loading on the 
piston. This is the first known reference to what we now call ‘stratified charge’ 
where a small quantity of mixture on the rich side of chemically correct is 
situated near the sparking plug, and a very lean mixture occupies the rest of the 
combustion chamber. The concept is that the rich mixture, once ignited, will 
carry combustion like a flaming torch through the lean mixture. By designing 
the combustion chamber to effect this action it is hoped to burn a much leaner, 
overall mixture than would be possible using a homogeneous mixture. Harry 
Ricardo was one of the first to see the possibilities of the stratified charge engine 
and he experimented on such an engine as early as 1915. Since then hundreds of 
patents have been taken out on special cylinder head designs to achieve efficient 
stratification followed by good combustion. Some of these patents have resulted 
in experimental engines and a few of these engines have shown a measure of 

The weakest homogeneous mixture that a conventional spark-ignition engine 
will run on without an occasional misfire is about 16 to 17 lb of air to 1 lb of 
fuel. Overall air/fuel ratios as lean as 60 to 1 have been burned successfully in 
stratified charge engines, but such an engine can only operate at a constant 
speed and may be regarded as a very clever laboratory experiment. On a more 
realistic level several practical engines have been produced in recent years that 
will operate in the range 20 to 25 to 1. 

Stratified charge engines fall into two generic classes, single chamber designs 
and dual chamber designs. Fuel injection is a sine qua non for the success of the 
single chamber system. A very high rate of air swirl is required into which the 
fuel is injected tangentially. The mixture resulting from this spray becomes 
stratified by the centrifugal action of the intense swirl, the richer mixture being 

Fig. 13.8 Diagram of the Texaco combustion process. 

236 The Sports Car 

centrifuged to the outside. This principle was used in the Texaco TCCS engine 
and appears in a modified form in the German MAN bowl-in-piston engine 
where the fuel is sprayed on the hot surface of the bowl. Combustion proceeds 
radially inwards as the fuel evaporates from the hot surface. 

The majority of workers in this field show a preference for the dual chamber 
system. The author worked on an experimental engine of this type fifteen years 
ago. This was the ‘Spitfire’ system, patented by Claude May and developed in 
the laboratory of the Walker Manufacturing Company in Racine, Wisconsin. A 
close approach to the Spitfire system was described in a recent paper by 
Professor Lev A. Gussak on the Russian LAG process. This engine carries a 
small primary combustion chamber connected to the main chamber by a very 
small diameter throat. A rich mixture is fed to the prechamber and a very lean 
mixture to the main chamber. A secondary, small inlet valve is used in the 
prechamber. By using such a valve the admission of the carburetted rich mixture 
can be controlled to begin at the optimum time before TDC. According to 
Professor Gussak, the burning gases violently ejected through the tiny throat 
carry micro-eddies of gas pre-conditioned for combustion as well as other small 
centres of ignited mixture that spread combustion throughout the main cham¬ 
ber. The pre-conditioning of the combustible mixture before true ignition is a 
complex physical and chemical phenomenon involving ionisation of the mole¬ 
cules followed by pre-flame oxidation. It is argued by Professor Gussak that the 
ignition points and the pre-conditioned gas pockets act like an avalanche as 
combustion spreads throughout the lean mixture. 

In conventional engines a mixture leaner than about 17 to 1 is not only 
difficult to ignite but the spread of combustion is so slow that it is not completed 
before the exhaust valve opens. 

Fig. 13.9 

The Leyland stratified charge 
engine using carburetted 
mixture to both chambers. 

The Sports Car in the Future 237 

Fig. 13.10 

Honda’s CVCC engine, the first in 
regular production. 

The Leyland dual chamber engine shown in Figure 13.9, designed for use in 
the Triumph Dolomite, bears a strong resemblance to the Russian design. The 
sparking plug is on the left of the prechamber, this chamber being fed by a 
carburetted rich mixture as in the LAG engine. The Honda CVCC engine 
(Figure 13.10) is used in the production Honda Civic. In this case again separate 
carburetted mixtures are used to feed the two chambers. Mercedes-Benz, 
however, prefer to use a fuel injector in the prechamber and a carburettor to 
supply the lean main mixture. 

The Honda engine has not yet achieved any great economy but their Civic 
saloon complies with the current Californian emission standards and gives 
nitrogen oxide levels that are low enough to meet the current Japanese standards 
which are even lower than those in California. Nitrogen oxides are formed at 

Fig. 13.11 

The Mercedes-Benz pre-chamber 
design. A carburetted mixture is 
supplied to the main chamber, 
with fuel injection in the 

238 The Sports Car 

high temperatures and the major part of the mixture in this type of engine is 
burned at a lower pressure and temperature than in the conventional engine. 

It is of interest to sports car owners that Porsche have an experimental 
dual-chamber engine. This may not be the first to appear in a production sports 
car, but the author will be very surprised if we do not see a stratified charge 
engine in a sports car in the early eighties. 


It is now commonplace for medium and heavy truck Diesel engines to be 
turbocharged. As a consequence compact, low-cost reliable turbochargers are 
available as off-the-shelf components for any specialist sports car manufacturer 
who wishes to boost the power of his engine. Turbocharging for passenger cars 
was used for a period in the sixties when the firm of AiResearch, in conjunction 
with General Motors, designed turbocharger installations for the Corvair, 
Oldsmobile and various Pontiac models. Using 100 octane fuel it was possible to 
increase power output by about 100 per cent with complete reliability. 

A turbocharger uses the pressure and temperature energy in the exhaust gases 
to drive a compressor to increase the induction pressure substantially above 
atmospheric pressure. Supercharging as we knew it on pre-war racing cars used 
mechanically driven compressors and these absorbed a considerable amount of 
the gross power that came from the increase in induction pressure. Apart from a 
certain increase in exhaust back pressure the additional power produced by 
turbocharging is all in the American vernacular ‘for free*. 

A typical turbocharger, an AiResearch Type T-04B, is shown in cross-section 
in Figure 13.12. The compressor is on the left and is an investment casting in 
aluminium alloy. The air flow is in the normal centrifugal supercharger 
direction, i.e. inwards from the left and outwards from a tangential passage 
from the scroll chamber. The impellor carries 16 radial blades. The flow 
through the gas turbine is in the reverse direction, the gas passing from the 
manifold into dual scroll chambers that form vortices converging on the curved 
blades of the turbine rotor. Expansion of the exhaust gas through the turbine, 
with the exhaust manifold pressure maintained at a much higher pressure than is 
normal in unsupercharged engines, provides the power to drive the compressor. 
The unit is quite compact since rotor diameters seldom exceed 80 mm (3.2 in) 
but they rotate at speeds as high as 120,000 r.p.m., about 10 times the maximum 
speed of a Grand Prix engine. The double rotor unit is dynamically balanced at 
assembly and by careful design very little end-thrust need be taken up by the 
fully floating bushes. Oil is supplied to these bearings at engine oil pressure and 
is drained back to the sump. The bearing housing has seals at each end. Oil 
consumption with good maintenance should therefore be nil. 

One difficulty in matching the turbocharger installation to the demands of a 
petrol engine lies in the rising delivery characteristics of a centrifugal blower 
with increase in speed. A turbocharger designed to give the desired pressure in 
the middle range of the engine speed will produce a boost pressure that is much 
too high at maximum engine speed, leading to heavy detonation and almost 

The Sports Car in the Future 241 

certain engine failure. Three control methods to overcome this problem are in 
use today. 

Waste-gate. The control scheme illustrated in Figure 13.13 spills exhaust gas 
through a waste-gate, thus by-passing the turbine. The simple control device 
shown here is a diaphragm valve that opens the waste-gate as the intake 
manifold pressure rises to the control point and closes it again as the engine 
speed falls. More sophisticated control valves that would give a boost curve that 
rises or falls at a chosen rate are of course possible. 

Compressor blow-off valve . Excess air delivery can be blown off by a suitable 
valve as shown schematically in Figure 13.14. This simple system cannot be used 
of course with the carburettor fitted on the atmospheric side of the compressor, 
but is well suited to fuel-injected engines. The thermal efficiency suffers since 
work is performed in compressing air which is then discharged to atmosphere. 
Turbine outlet restriction. A fixed orifice is sometimes placed downstream of 
the turbine to reduce the available turbine energy at high speeds. This tends to 
give higher maximum exhaust back-pressures than the other two methods. 

The Turbo Porsche is a fully developed system using the waste-gate control 
system. The caption under Figure 13.15 explains the workings of the system. 
The Bosch K Jetronic fuel injection has already been described in Chapter five. 
Many firms with emission problems are now adopting fuel injection since it 
gives such a close control of air/fuel ratios. The compressor recirculation circuit 
controlled by valve 3 is the Porsche answer to the old problem that plagued 
General Motors in the sixties. The Corvair Monza Spyder exhibited an agonising 
flat-spot when full power was again required after coasting. It could take up to 
three seconds after flooring the accelerator again before the gases in the exhaust 
system built up a high enough pressure ratio across the turbine to give any useful 
power from the compressor. 

Air/fuel ratio control 

Cars like the Turbo Porsche have been developed expressly with the North 
American market in mind and with the turbocharger controls described above 
and the latest K Jetronic fuel injection it is well able to meet the current Federal 
emission standards. Many development engineers however have recently 
expressed grave doubts of their ability to meet the proposed 1978 Federal 
standards. The Swedish Saab company is one firm that has not yet given up 
hope. In fact, as we write this chapter it is reported that the Saab Turbo has 
passed these very severe 1978 emission tests and has achieved this with a turbine 
control system largely similar to the Porsche system. They have the assistance of 
two additional devices the three-way catalyst and an automatic air/fuel ratio 

The three-way catalyst was found in the late sixties by the Ford specialists 
working in this field. In theory, to oxidise CO and HC in a catalyst bed one 
needs a lean mixture. To reduce NOx the mixture should be rich. Despite these 
simple theoretical considerations it was found that certain catalysts, largely 
platinum with about 10 per cent of rhodium, could do a ‘three-way clean-up’ of 

242 The Sports Car 

The engine draws in atmospheric air through air filter (1), mixture control (2), and 
induction pipe (4) which then flows through compressor (5) of the supercharger, pressure 
line (6), throttle housing (7), air manifold (8) and enters the engine. 

The engine exhaust gases pass through exhaust manifold (10), supercharger turbine (13), 
muffler (14) and then discharged to atmosphere. The exhaust gas flow drives turbine (13) 
which again drives compressor (5), supplying compressed air to the engine. The supply 
pressure of compressor (5) is limited by bypass valve (12) in exhaust manifold (10); when 
the supply pressure of compressor (5) exceeds a predetermined value, bypass valve (12) is 
opened by the excess pressure in control pipe (15) so that the exhaust gas flow passed through 
bypass line (11) around turbine (13) directly to muffler (14). For maintaining the supercharger 
speed, e.g. under coasting conditions or to ensure quick engine response when accelerating, a 
connection pipe with blowoff valve (3) is provided between induction pipe (4) and pressure 
line (6). With the throttle in closed position, blowoff valve (3) in control line (17) is opened 
due to the differential pressure so that the inlet air passing around compressor (5) ensures the 
required supercharger speed. 

Fig. 13.15 Diagram of Porsche turbocharging control system as used on 
Turbo model. 

all the pollutants. The problem with the three-way catalyst, insoluble in the 
sixties, was the extremely narrow band of air/fuel ratios required before all 
three reactions could occur. If for example the fuel in use had a stoichometric 
(chemically correct) air/fuel ratio of 14.6 the mid-point of the control band had 
to be very slightly on the rich side of this value, i.e., 14.5. Moreover, and this 

The Sports Car in the Future 243 

was the real challenge, all three reactions would only be achieved if the control 
band was held to limits of plus or minus 2 per cent of the mid-point under all 
operating conditions. When we remember that many induction manifolds in 
pre-emission days gave variations of plus or minus 10 per cent from cylinder to 
cylinder we see why the three-way catalyst was put on the shelf. 

A few years ago the breakthrough came with the invention of Lambda-Sonde. 
Lambda is the Greek letter used by German engineers to represent air/fuel ratio 
and Sonde is the German word for sensor. The Lambda Sonde is more correctly 
described as an oxygen sensor. The German Bosch Company is one of the firms 
active in this field and their sensor has now an assured life of 15,000 miles and 
can control air/fuel ratio to the tight band demanded by the three-way catalyst. 
The complete metering system developed by Bosch uses K Jetronic injection in 
conjunction with a ‘feed-back system’ using a signal from the Lambda-Sonde, 
situated in the exhaust manifold, to indicate to the electronic control module 
when the exhaust gas oxygen content has moved away from the control value by 
a small amount. The signals from the sensor are interpreted by the control 
module which then proceeds to regulate the quantity of fuel injected. The Volvo 
application is shown in Figure 13.16. When this engine was tested in California 
the chairman of the California Air Resources Board called it ‘the most 
significant breakthrough ever achieved.’ 

One thing is certain. Engines will become much more complex. Apart from 

Fig. 13.16 Volvo’s thinking engine. Schematic view of Volvo’s Lambda-Sonde 
equipped engine where the oxygen content of the exhaust gas is 
measured and the air/fuel ratio automatically corrected to maintain 
a clean exhaust. 

244 The Sports Car 

the increased cost, which we all regret, it is now apparent that the days of the 
do-it-yourself mechanic are numbered. Those familiar with any of the earlier 
editions will now realise why the chapter on Tuning has now been omitted. 
Under American law, carburettors are now sealed against unauthorised mixture 
adjustments. The practice will probably be adopted by other countries in the 


In our companion book, Design of Racing Sports Cars , a strong case was 
advanced for the use of automatic transmission in racing sports cars. So many 
long-distance races have been lost through a clumsy gear change on a manual 
gearbox by a tired driver. The shock loads created by one careless gear change 
can strip teeth from a gearwheel or break some other component in the drive 
train. The use of sticky tread compounds and the uncanny grip of wide tyres has 
increased this danger. 

When we consider the less arduous conditions of a production sports car used 
by a reasonably experienced driver, the danger of complete transmission failure 
recedes but the provision of an automatic transmission will remove one chore 
from the driver in a traffic-packed situation and leave him more time to 
concentrate on weaving his way through the traffic stream. With a modern 
well-designed 3-element torque converter coupled to a 3-speed automatic 
gearbox a sports car loses very little in terms of acceleration, fuel economy and 
maximum speed. Current automatic gearboxes use hydraulic control systems to 
operate the friction bands that engage the planetary gear trains. With the proven 
reliability of solid-state electronic equipment it is probable that future automatic 
gearboxes will be controlled electronically. 


One cannot predict the tyre profile ratio that will eventually become the norm 
on production sports cars but a fair guess would be that the 50 Series will seldom 
be exceeded in the downward trend since lower profiles become unacceptable on 
practical grounds. This limitation is very apparent in the case of the front- 
engined car where ultra wide tyres intrude on the space occupied by the engine 
and its increasing number of auxiliary components. 

Suspension geometry is now largely dictated by the tyres since the low profile 
tyre must be maintained very close to the vertical plane if it is to perform well. 

No-roll suspensions 

It is always possible to adopt one of the no-roll suspensions developed over the 
last twenty years, since with a body that never rolls one can design a suspension 
geometry that maintains all four wheels perfectly upright in a tight turn. The 
gain in cornering power is quite substantial. When tested by Motor a Rover 3500 
saloon fitted with the AP ‘stabilised suspension’ could negotiate a test chicane at 

The Sports Car in the Future 245 

54.5 m.p.h., an increase of 6 m.p.h. over the standard car. Since centrifugal 
force varies as the square of velocity this represents a gain of about 25 per cent 
in cornering power. 

Several engineers have designed workable no-roll suspension systems. As 
early as 1961 the writer tested a Chevrolet fitted with the Kolbe Curve-Bank 
suspension which was designed to lean inwards on corners. This would be a 
mistake with modern low-profile tyres but the system could easily be set up to 
give a flat ride. About six years later came Norbert Hamy’s Trebron system in 
which a sideways displacement of the body under centrifugal force is used to 
generate forces to hold the wheels upright in a corner. There have been others 
but none we know that have reached production or have been used successfully 
in racing. The latest sophisticated design, as tested on the Rover by Motor , 
comes from that enterprising component manufacturer Automotive Products of 
Leamington Spa. It uses the self-levelling principle developed by Citroen, but 
with an important difference, the hydro-gas strut has a split-second reaction 
time. If you stand on the front bumper of a CX Series Citroen (with the engine 
running) the front end dips and then slowly rises back to the controlled height. 
The delay time is about ten seconds. If you stand on the front of a car fitted with 
the AP system no apparent movement occurs. 

The secret of the AP control system is a small pendulum mass, mounted on a 
spring and provided with a damper. The power to correct the ride height of each 
suspension leg is provided by a large hydraulic pump capable of a high flow of 
oil at a pressure of 290-360 kN/m 2 (2000-2500 lbf/in 2 ). Full pump output, only 
required when taking a corner or chicane at speed, is 11 kW (15 b.h.p.), 
although the system is not designed simply to prevent roll. It also prevents pitch, 

Fig. 13.17 One suspension leg of the Ap stabilised 
suspension, showing schematically the 
hydraulic control system. 

246 The Sports Car 

dive and squat. In other words it is in the words of the makers a ‘stabilised 

Figure 13.17 is a schematic representation of a single suspension unit. The 
inert gas contained in the ball and separated from the hydraulic fluid by a 
flexible diaphragm is the suspension spring. By adding or subtracting flud from 
the space between the diaphragm and the piston in the suspension strut the level 
of the car body at that particular corner of the car can be adjusted upwards or 
downwards. The signal to add or extract hydraulic fluid comes from the 
balanced three-way valve. Change in level is sensed by the ‘pendulous mass’, 
which is supported on a coil spring and supplied with its own hydraulic damper. 
This mass-spring-damper unit is tailored to match the particular suspension 
system precisely, in frequency and damping characteristics. The behaviour 
under single wheel bumps is as follows: As the suspension arm moves upwards 
to compress the suspension spring (the gas contained in the ball) an upward 
force is created which lifts the body upwards. The pivot of the offset pendulum 
moves upwards with the body and, with no movement of the pendulum, the 
spool of the three-way valve would move to the left to extract fluid from the 
suspension leg. As already stated the pendulum, spring and damper unit have 
been designed to behave as a perfect model of the car suspension. Consequently, 
the pendulum rises at the same speed as the body. The net effect on the valve 
spool is to create no signal, both on bump and rebound of a single wheel. A 
change due to roll, however, is not as transient as a single wheel bump or 
rebound, which is completed in about a hundredth of a second. Roll, pitch, 
brake dive and acceleration squat are all suspension movements of relatively 
long duration and the system responds to these to maintain a level ride at all 
times. In the experimental Rover system two hydraulic valves are used at the 
front with a single central valve at the rear. The two front valves control roll, 
pitch and body height. The single rear valve only controls pitch and height. 
Additional hydraulic cylinders are incorporated in the rear suspension struts. 
These are diagonally connected to the front control valves. In this way a roll 
couple can also be applied to the rear when cornering. Moreover, by varying the 
size of these additional rear units the front and rear roll couples can be matched 
to give well-balanced behaviour when cornering. 

All this, as it stands today, represents an expensive and complicated piece of 
equipment. Even with the help of a large order for a popular car it is difficult to 
see the additional retail cost being held below £100. On a specialist sports car 
costing £10,000 a specially tailored system might add £300 to the retail price — 
an extra 3 per cent — not a great price to pay for improved cornering power 
and comfort. From the designer’s viewpoint a fully stabilised suspension system 
means that he can forget all about roll-centre heights, roll steer effects, anti-roll 
bars and the influence of wheel camber change on cornering behaviour. For the 
designers of luxury cars it means that a very soft suspension can be used. 
Perhaps Rolls Royce or Mercedes-Benz will be the first customers. 

The Sports Car in the Future 247 


Despite all the work, particularly in the U.S.A., to cocoon the occupants of a 
car against the results of an accident the author still believes that more thought 
should be given to improving handling and braking under adverse conditions so 
that accidents are less likely to happen. Anti-lock braking could give a 
worthwhile contribution to this safety philosophy. 

Those drivers who feel confident they can handle a car safely on slippery 
roads should consider the results of an exhaustive survey recently carried out by 
Calspan Corporation under the auspices of General Motors. Sixty men and 
forty women, all carefully selected as representative of typical road users, with 
an average age of 38 and an average driving experience of 19 years, were 
subjected to a series of driving tests. The report stated ‘The typical driver did 
not use the full potential of the car in terms of its cornering capabilities and 
handling qualities. In most instances, the driver resorted to hard or panic 
braking in simulated emergency situations, often locking up all four wheels and 
thus losing steering control.’ 

An interesting sidelight was that the men were shown to be more aggressive in 
their driving. They drove at higher speeds, but were not shown to be any more 
capable of handling a car in an emergency than the women. When asked to rate 
their driving ability before taking the tests 54 of the 60 men said they were above 
average; only 20 of the 40 women made this claim. There is no need for 
comment on this aspect of the tests, but the case for the provision of a braking 
system that is less dependent on driver ability is well demonstrated. 

Many accidents occur through panic braking. With locked rear wheels the car 
either spins or yaws into the approaching traffic. With locked front wheels the 
car slides in a straight line, despite the efforts of the driver to steer the car. An 
experienced driver with a cool head can use the quick on-off-on-off brake 
application technique on slippery surfaces to take advantage of the relatively 
efficient braking given just before the wheels lock-up completely. A system that 
senses the onset of wheel lock-up and uses the on-off-on-off technique 
automatically would turn us all into top rally drivers — in one aspect at least. It 
has been shown that on most flooded, greasy or icy surfaces maximum braking 
force is given with a slip of 10 to 15 per cent relative to the road surface. At 100 
per cent slip, i.e. a locked wheel, the braking force is negligible. On a dry surface 
in good condition, however, the degree of slip is not important and good 
braking is given with locked wheels. 

Dunlop was the first firm to make a serious attempt to introduce anti-lock 
braking with their Maxaret system in the late fifties. Pioneer work on this 
system led to the introduction of small-scale production on the Jenson FF 
(Formula Ferguson) four wheel drive sports saloon. The Maxaret system had a 
wheel-locking sensor with an on-off frequency of about six cycles per second. 
Later systems, such as the WSP (Wheel Slide Protection) system developed by 
Girling and Lucas in Great Britain and the Bosch system in Germany have a 
re-cycle time of 10 to 12 per second. A very high speed of response is essential 

248 The Sports Car 

since a wheel may pass from a dry surface, across a patch of ice, then back to a 
dry surface again in even less time than a sixth of a second. 

Solenoid Valve 

Figure 13.18 is a diagram of the Bosch system. The pressure-relieving piston is 
normally held in the lower position during braking by the anti-lock pressure 
system. Release of pressure on the wheel cylinder is triggered by a signal from 
the wheel-lock sensor to the electronic control box. An electronic pulse from this 
opens a solenoid-operated ball valve, thus releasing the pressure above the 
piston. The piston rises, thus relieving the pressure on the wheel cylinder and, by 
means of a stalk on the base of this piston, isolating the normal brake pressure 
by closing another ball valve. In this Bosch system the pressure relieving piston 
is held in the lower position by the superior pressure of the anti-lock system. 
This hydraulic pressure is supplied from a reservoir and an electrically-driven 
pump that continues to operate from battery voltage if the engine stops. 

With the Girling-Lucas WSP system the pressure relieving piston is held in the 
minimum volume position by a strong spring. Failure of the electrically-driven 
pump in this system therefore leaves the normal braking fully operative. Most of 
the difficulties associated with anti-lock braking have now been solved. Some 
development engineers are approaching the concept of full-scale production 
with great caution, perhaps a commendable failing when lives are at stake if 
unexpected defects are revealed by the hard school of public abuse. 


Since we pay so much for a well-engineered sports car today, it is a logical step 
to design the car for long life. This is a very complex subject and is certainly 
contrary to the policy of planned obsolescence that has kept Detroit in business 

The Sports Car in the Future 249 

for so long. However, we are the customers and the idea of a car that will last 
about twenty years appeals to many of us. Today we work hard to earn money 
so that we can buy a new car every two or three years so that other men in the 
motor industry can also keep on working hard, etc, etc. Meanwhile the raw 
materials we need for new cars are becoming more and more scarce and the 
mountains of rusty old cars are such an embarrassment that some authorities are 
dumping them in the sea. A few companies are beginning to look at the concept 
of the Long-Life Car, notably Volvo and Porsche whose cars are already well 
respected for their durability. We should give them every encouragement. 

Design studies 


William Lyons began as a body builder fifty-five years ago. It was therefore 
fitting that his final act before retiring was to assist the late Malcolm Sayer to 
design the body for the XJ-S sports car. Malcolm Sayer was a brilliant 
aerodynamicist and over the years Sir William had developed an uncanny flair 
for designing bodies that appealed to the public. 

The early bodies that were made by the Swallow Side Car and Coach Building 
Company in Blackpool were rather flamboyant and many connoisseurs of the 
period suggested they were even vulgar. Perhaps the bonnet and scuttle of the 
SSI was too long for a car with such a small engine and a top speed of only 
about 65 m.p.h., but William Lyons was selling his cars to the public, not to 
connoisseurs and the public associated long bonnets with dashing exotic sports 
cars. After the long years of the Depression they yearned for exciting sports 
cars. Unfortunately, they were not able to afford them. The SSI was in truth a 
‘dream car’. 

The years 1933 to 1935 marked a turning point in the William Lyons story. 
The motor-cycle sidecar business was sold to a sub-contractor and a public 
company was formed to make nothing but cars. This new company, SS Cars 
Ltd, would no longer build bodies to fit chassis made by other companies. They 
began to design and build their own cars. Bill Heynes left the Humber Company 
in 1935 to start work on a new chassis, a conventional design, but one with an 
exceptionally stiff frame and excellent Girling brakes. Harry Weslake designed 
two new six-cylinder overhead valve engines and it was in September 1935 that 
the new SS Jaguars were announced. At a price of £285 for the 1 Vi litre model 
and £365 for the 2Vi litre they were remarkable value for money. Many 
wondered how long they would hold together and must have been rather 

Design Studies 251 

disappointed when they eventually saw them survive to become valuable classic 
cars. The Jaguars were handsome, comfortable and well constructed and the 
2Vi and 3‘/2 litre 2-seater SS 100s that followed in 1936 and 1937 were not only 
fine cars, they were real sports cars. That, after all is the subject of this book. 
The 3'/2 litre model gave a genuine 100 m.p.h., very rare in pre-war sports cars 
and would accelerate from 0 to 60 m.p.h. in 12 seconds. 

After the war the company name was changed again, this time to Jaguar Cars 
Ltd, and their first sports car, shown for the first time at the 1948 Motor Show in 
London, was the memorable XK120, a car that is now cherished as a major 
classic. Few cars have been welcomed with such enthusiasm as the XK120. In 
those early post-war years the British public was still rationed for food, 
excitement and personal transport. The news of this truly beautiful magnificent 
sports car spread so fast there was no need for high-pressure publicity. Jaguar 
had certainly built ‘a better mousetrap’ and the public beat a path to their door. 

Many of the small companies that existed by assembling fairly effective and 
rugged sports cars from bits and pieces supplied from various component 
manufacturers found they could no longer compete with this fine new sports 
car. During the fifties the XK120, the XK140 and XK150 continued to challenge 
the best in Europe. Later variants were more close to racing cars, being 
developed to meet the increasing opposition from Ferrari, Maserati and 
Mercedes-Benz on the racing circuits of the world. This was the heyday of sports 
car racing for Jaguar and the C-Type and D-Type scored an impressive list of 
successes including five outright wins at Le Mans. The production Jaguars had 
many successes. The XK120 shown in Figure 14.1 was a replacement for 
NUB 120, now in the Jaguar museum. NUB 120 driven by Ian Appleyard with his 
wife Pat (n6e Lyons) as navigator, won a Coupe des Alpes three times in 

After a serious fire at the Browns Lane Works in 1957 production plans for a 
modified version of the D-Type, to be sold on the American market as the XK-SS, 
were abandoned. The design and development departments concentrated their 
efforts on a completely new sports car, one using independent rear suspension on a 
Jaguar for the first time. This was the XK-E usually called the ‘E-Type.’ 

The E Type, introduced in 1961, maintained the tradition that had started 
with the XK120, that a production sports car should be docile in traffic, 
well-mannered at all times and comfortable on a long journey. Many American 
housewives, taught to drive at High School on an American sedan with 
automatic transmission, are not daunted by this piece of exotic machinery and 
love to use it to bring home the groceries. The engine in the E Type was still the 
long stroke 6-cylinder engine shown in Figure 3.8. This engine first appeared in 
1948 with a bore of 83 mm and a stroke of 106 mm to give a swept volume of 
3.44 litres. The bore of the E Type was increased to give a capacity of 3.8 litres 
and later, with a bore of 92.07 mm the capacity rose to 4.235 litres. Since the 
stroke remained at 106 mm the engine lost some of its extreme stroke/bore 
ratio. Even so, Ferrari, Lotus, Maserati, Mercedes and Porsche had all gone 
over-square by this time. 

Appleyard in the Alpine Rally. (Motor photograph). 

Design Studies 253 

The E Type body was as advanced in style as its engine was old-fashioned. In 
some respects the oval cross-section was reminiscent of the earlier D Type and 
the appearance of the car was welcomed by the aficionada who write in such 
magazines as ‘Car and Driver’ and ‘Road and Track’. The latter magazine 
described it as ‘The greatest crumpet collector known to man.’ This could be 
true, but we are concerned here with technical matters and their criticism of the 
archaic gearbox with its slow synchromesh and very long travel between gears 
was well merited. A car with such potential can be ruined if so much time is 
wasted changing gears. Jaguar became aware of this criticism and an excellent 
4-speed gearbox appeared in 1964 to make the car a delight to drive. 

Fig. 14.2 V-12 engine cross-section showing carburettor layout on the left and 

fuel injection on the right. 

In 1971 the E Type appeared with a new engine, a 12-cylinder of 5.342 litres 
with the cylinders arranged in a 60 degree V. The bore was 90 mm and the stroke 
70 mm, in striking contrast to the stroke/bore ratio of the older 6-cylinder 
engine. In the development programme leading up to this engine there were two 
parallel projects, the first engine having a double overhead camshaft head on 
each bank of cylinders, the second having only a single OHC head per bank. 
The first engine, as one would expect developed more power — about 30 per 
cent in competition tune and about 10 per cent more in its planned production 
form. For several reasons the DOHC design was dropped, the major reason 
being the width between the outer camboxes and the additional height. The 
width conflicted with the space needed to give sufficient steering lock and the 

254 The Sports Car 

height of this engine could not be accommodated under the falling bonnet-line 
planned for the new generation of Jaguars. 


The new series of Jaguar saloons offer a choice of three engine sizes, two 
6-cylinder of 3.4 and 4.2 litres and the new 12-cylinder of 5.3 litres. For the new 
sports car, the XJ-S, introduced in September 1975 only the V-12 is fitted. The 
finalised production engine is fitted with Lucas electronic fuel injection as 
described in Chapter five and Lucas electronic ignition as described in Chapter 
six. It will be seen from Figures 14.3 and 14.4 that the XJ-S bears little 
resemblance to the E Type. It is also substantially larger, being 11 cm (4.5 in) 
wider and 43 cm (17 in) longer. 

Fig. 14.3 Front view of Jaguar XJ-S. 

The new XJ-S has been classified by some motoring journalists as a ‘sports 
saloon’, not a sports car. This could be true. In the words of Motor it is ‘one of 
the world’s most desirable cars.’ They also say it is ‘no replacement for Jaguar’s 
classic two-seater.’ The two-seater sports car has always been a young man’s car 
and it would appear that the Leyland marketing organisation has decided the 
supply of young men who can afford a two-seater with a top speed of 150 
m.p.h., a fuel consumption of about 14 m.p.g. and an insurance premium well 
into three figures is fast running out. For these young men they offer the 
Triumph Spitfire and TR7 and the MG Midget and MGB. The market they 

Fig. 14.4 Ghosted view of XJ-S. 

CfcAisk - l& rr 

Fig. 14.5 The V-12 engine installed. A very full engine compartment. 

Fig. 14.6 Torque characteristics of 6, 8 and 12-cylinder engines. 

Design Studies 257 

envisage for the XJ-S is the more mature and certainly the more financially 
successful businessman. In this class the new Jaguar ‘sports saloon’ is highly 
competitive. The Aston Martin DBS which could be bought for under £8000 in 
1972 now costs £16,999. In these hard times even a well-paid top executive will 
be tempted when replacing his older Aston Martin by the new Jaguar at £12,500. 
Even today there are still a few customers left for Aston Martin. There is already 
a long waiting list for the new Lagonda recently announced at an estimated 
selling price of £30,000. 

The engine 

Many aspects of the engine have already been covered in the rest of the book. 
An external view of the engine is given in Figure 14.5. It lacks the finely 
sculptured beauty that we once saw under the bonnet of a Bugatti or the early 

Fig. 14.7 The XJ-S ‘hydraulic’ bumper. The working fluid is 
a silicone wax. 

258 The Sports Car 

Ferraris. The impression is of a plumber’s nightmare. Fortunately the rest of the 
XJ-S has more aesthetic appeal and the inside of the engine compartment should 
only be seen today by garage mechanics — at fairly infrequent intervals. 

The smooth torque given by the V-12 is well illustrated in Figure 14.6. Torque 
variations are not great on a V-8; a V-12 is almost as smooth as a turbine. The 
power of the new engine is 285 b.h.p. DIN. The 4.2 litre E Type engine had an 
advertised power of 265 b.h.p., but these were SAE gross figures, a gross 
exaggeration as explained in Chapter twelve. Aluminium alloy is used for the 
major structural components of the new engine, an open deck sand casting in 
LM 25 for the block/crankcase and LM 25WP for the cylinder head. The tappet 
block, oil cooler, induction system, and all external covers are in aluminium 
alloy. The sump, however, is a steel pressing, a sensible precaution against the 
more venturesome drivers who occasionally knock holes in cast sumps on rocky 

Special features of the XJ-S 

The body 

The steel monocoque body shell is designed to meet all current Federal safety 
regulations. It incorporates ‘5 m.p.h. — no damage’ bumpers which act like the 
hydraulic buffers at railway terminals. The hydraulic medium is a silicone wax 
which absorbs the kinetic energy of the impact and slowly restores the bumper 


Design Studies 259 

Fig. 14.10 The XJ-S rear seats. 

260 The Sports Car 

to its original position. Anyone with a pre-disposition to ‘parking by ear’ should 
remember that the two struts take about thirty seconds to regain their normal 

The new Jaguar body is well sound-proofed by means of damping pads and 
deep-pile felt-backed carpeting. In this respect the car is a serious challenge to 
the most expensive luxury cars. Anyone equating sports cars with noise and fury 
should look elsewhere for his next car. The XJ-S is designed to give high 
performance with the minimum of fuss. 

The instruments 

The XJ-S instruments are contained in a single nacelle as shown in Figure 14.8. 
Between the speedometer and the tachometer the four critical indicators of 
water temperature, oil pressure, fuel contents and battery conditions are 
grouped together for easy reading. Along the top of the nacelle are 18 warning 
lights to monitor 18 major mechanical and safety functions. The instrument 
nacelle is wired by printed circuits and connected into the main wiring loom by 
two multi-pin plugs. 

General specification 


Type Four-stroke - petrol engine - 

water cooled 

No. of 

cylinders 12 in 60 Vee 

Bore 90mm 3.54 ins 

Stroke 70mm 2.76 ins 

Capacity 5343 cc 326 cu ins 

Piston area 763.2 cm 2 118 sq ins 


ratio 9.0:1 


Power 285 DIN HP at 5800 rpm 
Torque 40.7 mkg 294 lbs/ft at 3500 rpm 

Cylinder block 
Type Open deck 

Material Aluminium alloy LM25 

Pistons Aluminium alloy solid skirt with 

combustion chamber in top. 
Piston rings Three - two compression and 
one multi-rail oil control. 
Crankshaft Three plane, seven main bearings. 

Tufftrided manganese molyb¬ 
denum steel. 

Cylinder heads 

Material Aluminium alloy LM25 WP 

Camshafts Two - one per bank 

Valve layout Single overhead camshafts 

operating bucket type tappets. 
Valve lift 9.525 mm 0.375 ins 

Engine continued 
Tappets - 

type Inverted bucket 



inlet 0.30-0.35 mm 0.012-0.014 ins 

exhaust 0.30-0.35 mm 0.012-0.014 ins 


inlet 17° BTDC 59° ABDC 

exhaust 59° BBDC 17° ATDC 


Type Steel pressing with internal 



Type Pressure 

Pump type Internal and external gear with 
crescent type cut-off. 



pressure 4.9 kg/sq cm 70 psi 

Filter Full flow paper element. 



Firing order 

Spark plugs 

Lucas Opus Mk II electronic, 
la, 6b, 5a, 2b, 3a, 4b, 6a, lb, 2a, 
5b, 4a, 3b. 

Lucas magnetic impulse type. 

Stroboscopic 10 BTDC at 750 rpm 

Champion N10Y 

0.625 mm 0.025 ins 


Type Lucas electronic manifold 


Enrichment Automatic cold start injector. 

manifolds Two 6-branch aluminium alloy. 

Fuel system 

Type Recirculating. 

Pump Lucas electric permanent magnet 


Fuel speci¬ 
fication 97 octane - four star. 

Electrical equipment 
Polarity Negative. 

Battery Lucas CP 13 


capacity 68 amps at 20 hour rate. 

Starter Pre-engaged 

Alternator Lucas 20 ACR 


capacity 60 amps at 3500 engine rpm. 

Horns Twin Lucas self earthing. 

Cooling system 

Type Water pressurised. Impeller 

pump belt driven off crankshaft. 
Pressure 1.056 kg/cm 2 15 psi 

Radiator Marston Superpak crossflow. 

Thermostat Two wax type opening at 82 C 

Fans 12 bladed steel fan with viscous 

coupling and thermostatically 
controlled electric 4 bladed fan. 


Layout Four downpipes merge into two 

double skinned pipes. Two main 
and two rear silencers. 

Exhaust Exhaust emission controls incor¬ 
emissions porate exhaust port air injection, 

(North exhaust gas recirculation and a 

America) catalytic reactor for each bank. 

Evaporative Engine anti-run-on valve. Vapour 

loss from the fuel tank is piped via a 

separator cannister to a charcoal 
cannister which is purged by 
manifold depression. 



Gearbox 4-speed all synchromesh. 

Clutch Single dry plate. 


diameter 267 mm 10.5 ins 


Gearbox Borg Warner Model 12 3-speed. 


Convertor 2.0:1 ratio 

Design Studies 261 

Transmission continued 




1st 3.238:1 


2nd 1.905:1 


3rd 1.389:1 
4th 1.0:1 


Rev 3.428:1 


Axle ratios 



Overall ratios Manual 


1st 9.94:1 


2nd 5.85:1 


3rd 4.26:1 

4th 3.07:1 


Rev 10.51 



Type Disc brakes all round. 

Layout Dual circuit split front to rear 

with pressure differential 
warning actuator. 

Servo In line tandem vacuum servo. 

Discs - front Type Ventilated cast iron. 

Diameter 284 mm 11.18 ins. 
Discs - rear Type Cast iron with damper 
ring in periphery. 
Diameter 263 mm 10.38 ins 

Calipers - 

front 4 piston caliper 

Calipers - 

rear 2 piston caliper 


materials Discs - Ferodo 2430 

Rubbed area 

- front 1624 cms 252 sq ins 

- rear 956 cms 2 148 sq ins 

-total 2580 cms 2 400 sq ins 


Front layout Fully independent semi-trailing 
wishbones and coil springs. Anti¬ 
dive geometry. Girling Monitube 
dampers. Anti-roll bar. 

King pin 

inclination Vn ± % 

Castor angle V/i ± Va° 


angle Positive Vi ± V* 

Alignment 1.39-3.12 mm Vi6 -0.125 ins 
Springs - 

free length 12-14 ins 
- rate 423 lb/ins 
Anti-roll bar 

diameter 22.2 mm 0.875 ins 

Rear layout Lower transve;se wishbones with 
drive shafts acting as upper links. 
Radius arms. Twin coil spring and 
damper units. Girling Monitube 
dampers. Anti-roll bar. 

262 The Sports Car 

Suspension continued 

angle % negative ± % 

Anti-roll bar 

diameter 14 mm 0.562 ins 

Body continued 

locks. Separate keys for ignition, 
doors and boot and glove locker. 
Bonnet release and locking lever 
below facia. 

Wheels and tyres 

Wheels - GKN Kent Alloy. Light alloy 

type wheels to Jaguar design. 

Size 6JK rim. 38.1 cms (15 ins) 


Tyres Dunlop SP Super steel braced 

with block tread pattern. 

Size 205/70 VR 15 



Turns lock 
to lock 
Overall ratio 

Ad west power assisted rack, and 
pinion with energy absorbing 

393.7 mm 15.5 ins. 

3 turns 

16:1 with an 8 tooth pinion. 
Saginaw rotary vane. 







Locks and 

All steel monocoque construction. 
Two door four seater with forward 
opening bonnet and large boot. 
Complete underbody protection. 
Driver’s door mirror. Front air 
dam and undershield. Radio 
aerial. Flush fitting filler cap 
cover. Recessed door handles. 
Wrap-around front and rear bum¬ 
pers consisting of a steel armature 
mounted to Menasco struts and 
with a synthetic rubber cover. 
Designed to meet 5 mph impact 

Laminated windscreen with 
toughened side and rear windows. 
Tinted glass is standard. Electri¬ 
cally heated rear window. 

Doors are fitted with high anti¬ 
burst load locks with flush fitting 
interior and exterior handles. 
Electrically operated central door 


Overall - 

4.87 m 

191.72 ins 


1.26 m 

49.65 ins 


1.79 m 

70.60 ins 


2.59 m 

102.00 ins 

Front track 

1.47 m 

58.00 ins 

Rear track 

2.49 m 

58.60 ins 



140 mm 

5.50 ins 

Interior dimensions 
Headroom - 


914 mm 

36.0 ins 



826 mm 

32.5 ins 



56.0 ins 


1346 mm 

53.0 ins 

Seat squab to 

brake pedal 

521-362 mm 

20.5-14.25 ins 

Luggage compartment 
Maximum - 


565 mm 

22.25 ins 


572 mm 

22.5 ins 


991 mm 

39.0 ins 


0.43m 3 

15 cu ft 


1687 kg 


1 (Weight includes automatic transmission, 
automatic air conditioning, energy absorbing 
bumpers, door side intrusion members, electric 
windows, central door locking, radio with 
electric aerial but less fuel). 


Design Studies 263 

Automobiles, like people, suffer the ignominy or enjoy the privileges of Class. 
Class is decided to some extent by money, but not entirely. The position in the 
Class structure depends upon public recognition, as many a Texas cattle-baron 
has learned. Rolls Royce, Cadillac, Ferrari and Mercedes-Benz have never 
known any other class but that of aristocracy, while Chevrolet, Volkswagen and 
Morris are unashamedly working class. In one generation Colin Chapman has 
pulled Lotus out of its lowly origins of the fifties when he made kit-cars to be 
assembled by young drivers with very little money but lots of enthusiasm, 
through the middle class status of the first Elite, the Elan and the Europa into 
what is now accepted by the motoring press as the top echelon of the Middle 
Classes, if not yet the Aristocracy. In January 1977 Motor compared the Esprit, 
the new Lotus mid-engined sports car costing £8000 very favourably with the 
Porsche 911 Lux at £11,500 and the Ferrari Dino 308 GT4 at £11,700. The 
front-engined Elite (the second model to carry this name) is now available with 
automatic transmission, air conditioning and electric windows. The Elite has 
four comfortable seats and is sometimes seen as a chauffeur-driven company 
car. Company cars with a touring fuel consumption of 25-30 miles per Imperial 
gallon are beginning to appeal today even to company accountants. 

A coloured spread of the Lotus Elite with thoroughbred horses in the back¬ 
ground is evocative (to those of us old enough to remember) of the Bugatti, lepur 
sang of the automobile world. One can look for points of similarity in the style and 
thoughts of these two creative engineers, but one essential difference is that ‘value 
analysis’ was not invented in Ettore Bugatti’s day and would have been scorned by 
him if it had. Colin Chapman would be the first to point out that the Bugatti com¬ 
pany did not survive. To survive as a sports car manufacturer one must strive, not 
only for near-perfection, but for near-perfection at an acceptable price. 

Lotus Cars Ltd of Hethel, near Norwich, make about 3,000 cars a year, fewer 
than General Motors make before lunch. Colin Chapman has no desire to 
emulate General Motors in output, only in their ability to stay in business. They 
admit at Hethel that their slant-4 engine is in effect one half of a V-8, but they 
will study the market very carefully before they are tempted to put a 4-litre 
engine in one of their cars. In a world starved for fuel a 2-litre engine is very 
close to the optimum for a sports car. 

Self-sufficiency is now the watchword at Hethel. For the first time since Colin 
Chapman began to build his own cars in a small lock-up garage in North 
London in 1948 he is no longer dependent on outside suppliers for engines. The 
Lotus 4-cylinder 16-valve DOHC engine has been developed into a remarkably 
smooth reliable power unit. The cylinder head and cylinder block are die-cast in 
aluminium alloy and all components are now manufactured in the new Lotus 
machine shop on tape-controlled machines. Bodies and chassis components are 
made at Hethel. Even the air conditioning units are made on site. The only 
bought-out major components are the gearboxes and final drive units, a practice 
not unknown among much larger companies. 

In Chapter seven we advanced reasons for rejecting front-wheel drive for a 

Design Studies 265 

high performance vehicle. One can design a front wheel drive car to outperform 
all others in a fast bend, but it is possible to reach a point of no return. It is 
flattering to think that Colin Chapman is motivated by similar thoughts. Only two 
layouts are considered at Lotus. For a four-seater or 2 -I- 2 (what we once called an 
‘occasional 4-seater*) Lotus prefer a front engine driving the rear wheels. For a 
more sporting 2-seater the engine is placed immediately behind the seats, a mid¬ 
engine layout with rear drive. The Elite and Eclat are variants on the first theme. 
The Esprit, designed for those with ‘spirit* is a challenge to all those expensive 
continental mid-engined sports cars from the Ferrari Dino to the Lamborghini 
Urraco and, of course, to the Porsche 911 with its overhung rear engine. 


The 500 Series begins with the basic 501, no spartan specification this, since it is 
equipped with heater and radio, electric windows, heated rear window and rear 
window wiper and washer and inertia-reel front seat belts. The 502 specification 
carries the additions of air conditioning, stereo radio tape deck, quartz-halogen 
headlights, tinted glass and a more expensive trim. The 503 specification 
extends to power steering and the 504 carries an automatic transmission. Since 
lovers of automatics are assumed to be lazy the 504 specification also includes 
electric aerial extension. 

The engine 

This is a slant-4 with a cylinder bore size of 95.2 mm (3.75 in) and a stroke of 
69.2 mm (2.72 in). The head, block, crankcase oilpan, camshaft carriers and 
cam covers are all die-cast in aluminium alloy. The engine is canted at an angle 
of 45 degrees to give the low bonnet-line necessary in the Elite to achieve the 
wedge profile. The provision of a timing-belt drive for the two overhead 
camshafts has been described in Chapter Four and an external view of the 
engine as used in the Team Lotus racing cars can be seen in Figure 4.7. An 
exploded view of the camshaft and cylinder head components is given in Figure 
14.11. The inherent stiffness that an oversquare bore/stroke design gives to the 
short block and crankcase is well illustrated in Figure 14.2. The Lotus head 
shows a strong Cosworth influence, having a narrow valve angle and siamesed 
inlet porting to feed the 8 inlet valves. The 4 siamesed ports are fed by two 
Dellorto twin-choke DHL A 45E carburettors. 

With a specific power of 80 b.h.p. per litre from a 4-cylinder engine one 
would have accepted a few years ago that a certain amount of roughness was 
inevitable. The continuous development of this Type 907 engine over a period of 
about five years has reduced mechanical noise and vibration to a low level. Oil 
consumption was a little high on early engines but this is now reduced to an 
acceptable negligible amount. Since the Elite was introduced in 1974 an intensive 
development programme has been carried out to refine the engine. Revised inlet 
ports have given improved torque at low speeds and a lighter flywheel has 
improved acceleration. The new standard of smoothness and freedom from 
vibration is given by careful matching of combustion chamber volumes and a 
higher standard of balancing of rotating and reciprocating components. 

Fig. 14.12 Cylinder block, lower bearing housing and sump on the Lotus engine. 

Note the increased stiffness given by integrating the lower main bearing 
halves into a single ladder-like component. 

V £ ■ 

Fig. 14.1 3 The Lotus Elite. 

Fig, 14,14 The Lotus Elite; a ghosted view. 

Design Studies 269 

The Elite’s wirtd-cheat lug shape 
was not evolved without a good 
deal of wind-tunnel development 
work. The graph shows the drag co¬ 
efficients of the different 
configurations tried* some of which 
are illustrated, ^ 


Quarter-scale model with no 
radiator ducting or underbody detail. 
Drag coefficient: 0.34 
Lift at 100 mph,— front 50 lb 
rear 150 lb 

Quarter-scale model with underbody 
detail, conventional bumper and 
air inlake beneath. 

Drag coefficient: 0.3& 

Lift at 100 mph— front 00 lb 
rear 30 lb 

Quarter-scale model as In 2 but 
with spoiler under air in lake. 
Drag coefficient: 0.37 
Lift at 100 mph— front 70 lb 
rear 34 ib 

Full sized “lash-up" body—no 
spoiler and many details improvised, 
Drag coefficient: 0.3B 

Fully detailed running prototype 
with all parts to production 
accuracy. No spwler. 

Drag coefficient: 0.38 
Lilt at 100 mph— front 111 lb 
rear 31 !b 

As for 5 but with metal plate 2in 
deep at 45 L to horizontal beneath 
air intake acting as spoiler 
Drag coefficient: 0.33 
Lift as loo mph— from 58 lb 
near 37 lb 

As for 6 but With properly shaped 

Drag coefficient 0 30 
Lift at 100 mph— front 36 Ib 
rear 38 lb 

Fig. 14.15 Wind tunnel testing of the Elite. Step by step improvements. 

The gearbox 

Originally introduced with a 4-speed box the Elite now uses a Lotus designed 
box with five speeds. Fourth drive is a direct drive with an effective ratio of 0.8 
to 1 on fifth. 116 m.p.h. is given at 7,000 r.p.m. on fourth gear. The Elite can 
potter down to 30 m.p.h. in top gear, but for real acceleration the driver must 

270 The Sports Car 

Fig. 14.16 The interior of the Elite 

make full use of the gearbox. Since this has such an excellent slick gear change 
this is more of a pleasure than a chore. 

For those who prefer it a ZF automatic transmission is available. 

The body 

The clean wedge profile of the Elite body can be seen in Figure 14.13. The 
ghosted drawing in Figure 14.14 also shows body construction details and other 
features such as the suspension layout. The GFRP body and its incredibly 
smooth surface finish has been fully described in Chapter nine. Figure 14.15 is a 
resume of the wind tunnel development work that led to the final drag 
coefficient of 0.30, a reduction of 15 per cent on that given by the Europa. It is 
to be noted that the total drag of the Elite body will be higher than that of the 
Europa since the frontal area of the Elite is so much greater. The Elite is actually 
half an inch wider than the Rolls Royce Silver Shadow! It is of course much 
shorter and lower. By European standards this is wide for a 2-litre car and it tends 
to take up more than its share of the road in its native Norfolk lanes. In North 
American and Australia the Elite is still a small car. The Chevrolet Caprice and the 
Australian-built Ford Fairmont are both about three inches wider. In its own coun¬ 
try the American sedan is given wide roads, wide parking slots and big garages. 
In such surrounding the Elite makes the typical ‘compact’ look very big. 

With a front track of 4 ft 10 Vi in and a rear track of 4 ft 11 in, even wider than 
the Jaguar XJ-S, it would have been difficult to pare down the overall body width 
since the doors must be deep enough in section to accommodate the box-section 
girders needed to withstand the side impact test specification of the U.S. Federal 
laws. The interior of the Elite is carefully planned, giving space for the driver 

Design Studies 271 

and front passenger of fairly ample proportions. By seating the rear passengers 
in a knees-up reclining position, that is much more comfortable than it sounds, 
it is possible to give quite acceptable rear seating for two 6 foot passengers in a 
car with a height below 4 feet. The interior of the Elite is shown in Figure 14.16. 

General specification 



4 in line 


1973 cc (120.4 cu in) 


95.2/69.2 mm 
(3.75/2.72 in) 








Valve timing 

Dohc, 4 per cylinder 

inlet opens 

25° btdc 

inlet closes 

65° abdc 

ex opens 

65° bbdc 

ex closes 

25° atdc 




Two Dellorto DHLA 45E 


5 main 

Fuel pump 

SU electrical 

Max power 

155 bhp (DIN) at 6500 rpm 

Max torque 

135 1b ft (DIN) at 5000 rpm 


Type: 5 speed manual or automatic 
Clutch: 8.5 ins (21.59 cms) diaphragm 
spring cable operated. 

Manual Internal Ratios and 
mph/1000 rpm 

5th (O/D) 

















Final drive 



Automatic Internal Ratios and 

mph/1000 rpm 














Final drive 




GRP body, steel back¬ 
bone chassis 


Underseal on chassis 
with flame retarding resin 
and intumescent paints 



Independent by wishbone, 
transverse link, coils, tele¬ 
scopic dampers, anti-roll bar 



Independent by transverse 
link semi-trailing arm, 
coils, telescopic dampers 


Rack and pinion 




0.125 ±0.06 in 


0° ± Vi° 


3° + 1° = Vi° 

King pin 


Rear toe-in 

0.25 ± 0.06 in. 



Disc front, inboard 
drum rear 







Alloy 7J x 14 


Dunlop SP Sports 

Super 205-14/60 



22 psi f/r (high speed) 


12 v 50 Ah 




Alternator, 60A 




2 x 75/70 W 

272 The Sports Car 
The suspension 

The shape of the backbone frame described in Chapter nine begins to make 
sense when the rear suspension is studied. A transverse beam at the rear of this 
frame provides a rigid structure to transmit the vertical loads from the coil 
spring mountings to the backbone. The narrowness of the backbone itself gives 
full scope to the designer to use long semi-trailing arms to locate the wheel hubs 
and to provide the desired amount of swing-axle effect. A perimeter frame as 
used by some earlier designers to support GFRP bodies would encroach on the 
space needed for long semi-trailing arms. The mid-engine position used on the 
Esprit does in fact encroach to a certain extent on this space as will be discussed 
later. Lateral location of the wheel hub on the Elite is provided by the fixed 
length drive shaft and a secondary lateral link. 

The forked front section of the backbone frame carries vertical pillars on each 
side which act as upper spring locations for the front suspensions (See Figure 
9.5). An upper wishbone link is used on the front suspension with a single lower 
transverse link and an anti-roll bar. Telescopic dampers are used front and rear. 


The Elite cannot quite match the top speed or the vivid acceleration of the more 
powerful competition. On the credit side most of the competition needs at least 
50 per cent more fuel to travel between two points. Maximum speed of the Elite 
is 20-25 m.p.h. lower than that of the 3-litre Ferrari Dino 308 GT and the 3-litre 
Maserati Merak, but a top speed of about 120 m.p.h. is high enough for most 
drivers. With a zero to 60 m.p.h. acceleration time of just under 8 seconds the 
Elite yields about 1.5 seconds to the Ferrari and the Maserati. On the credit side 
again it is difficult to find another 4-seater to match the cornering power of the 
Elite. The combination of Lotus suspension finesse and the new Dunlop 60 
profile steel mesh radial tyres seems unbeatable. When tested by Motor they 
concluded their report as follows:- 

‘So highly do we regard this superb car in its latest form that all members of 
the test team were duty bound to sample it in order to renew the standards by 
which they must now judge rivals, some of them much more expensive than the 


The Eclat is not a full 4-seater like the Elite. It is a more sporty 2 + 2 for those 
who prefer the engine at the front and will sacrifice a little space and headroom 
at the rear to reduce the frontal area slightly and reduce the weight by about 40 
kg (90 lb). The Eclat does not have an opening rear window as on the Elite. The 
roof line is lower at the rear and the rear view afforded to the driver is also 
reduced. The luggage space, oddly enough, is slightly larger. 

There is a basic Eclat, priced £1100 below that of the cheapest Elite. This 
Type 520 Eclat is the only option with a 4-speed manual gearbox, 5Vi J x 13 
steel wheels and 70 profile Goodyear G800 Grand Prix tyres. The 521, 522 and 


Fig. 14.17 The Eclat 

Fig. 14.18 the mid-engined Esprit. 

276 The Sports Car 

523 are equipped with 7J x 14 alloy wheels and the more expensive Dunlop SP 
Sports Super 205/60 VR tyres. They also have a five-speed manual gearbox as in 
the Elite. The 522 carries the air conditioning option and the 523 has power 
steering. Automatic transmission is not available on the Eclat. A rear view of 
the Eclat is shown in Figure 14.17. 


If the Elite is the replacement for the Elan it is logical to see the Esprit as the 
replacement for the Europa. This is an oversimplification. There was little in 
common in the external appearance of these earlier models. The three new 
Lotuses, however all bear a strong family likeness. 

The Europa came in for much criticism for its poor visibility, particularly 
over the rear-quarters. Not all the criticism was fair. One editor of a New York 
car magazine made fun of the difficulties he experienced getting in and out of 
the car. In truth he was really too old and too fat for a Europa! Very few active 
young men complained on this score, but the Renault power unit only produced 
67 b.h.p. in the original form and 78 b.h.p. in the S2 version. The car was 
aerodynamically efficient and this gave it a top speed of 115 m.p.h., but the car 
sadly lacked acceleration for a car with such potential. 

The body on the new Esprit follows the general wedge outline of the Elite. In 
fact the lower half of the body is made in the same mould as the Elite. Body and 
interior were styled by the European designer Gingiaro. He gave the body an 
even lower profile than the Elite and Eclat. With a height of 3 ft 7% in, the 
Esprit is 3 3 A in lower than the ferrari Dino and the Porsche 911 at 4 ft 4 ins in 
almost towers above it. 

The kerb weight of the Esprit is 1980 lb. That of the Europa was 1465 lb. With 
an allowance of 300 lb for 2 passengers the Esprit has an effective power to 
weight ratio of 157 b.h.p./ton; the Europa had only 99. Where the Europa took 
10.7 seconds to reach 60 m.p.h. from rest, the Esprit takes 6.8 seconds and will 
reach 100 m.p.h. in slightly more than 20 seconds. To complete the comparison 
the power to weight ratio of the Ferrari Dino 308 GT4 is 182 b.h.p./ton. The 
Dino will accelerate from rest to 100 m.p.h. in about 16 seconds. The Esprit will 
give a touring fuel consumption of 25 miles per Imperial gallon, the Dino only 
18, but does this really matter when one has paid about £12,000 for the car 
delivered to the door? The Esprit is not only £3,000 less expensive than this 
lowest prices of the Ferrari range, but a comparison of replacement part costs 
must carry some weight, even to pop stars. To replace a broken windscreen on 
the Lotus costs about £80, on the Ferrari about £220, on a Lamborghini about 

An external view of the Esprit is given in Figure 14.18, a ghosted line drawing 
in Figure 14.19. The interior is shown in Figure 14.20. 

The engine 

The same engine is used in all three models. In the Esprit, however, it occupies 

Design Studies 277 

Fig. 14.20 The interior of the Esprit. 

the space allocated to the rear passengers in the Elite, driving the rear wheels 
through a transaxle, the gearbox being overhung as in modern racing cars. 

The gearbox 

Maserati, now under the financial control of the Citroen company, make the 
5-speed transaxle for this Lotus model. The gear linkage was originally designed 
for the Maserati-powered Citroen SM and is found to give the proverbial 
butter-slicing changes, no mean achievement with such a remote gearbox 
location. As in the front-engined models fifth gear is used as an overdrive, 
although in this gearbox none of the gears give a direct drive. 

The body 

The chisel-nose of the Esprit is even lower than that of the Elite and Eclat, since 
the only major bulk housed at the front is the spare wheel. This, in practice can 
be a disadvantage since the front extremities of the car cannot be seen when 
parking. A similar objection can be raised to the rear visibility. On the open 
road this is no great handicap since there is a rearview mirror and very few cars 
attempt to pass. There has been a marked improvement on the letter-box slot 
rear window with built-in side blinkers that damned the Europa in the motoring 
press. The Esprit demands a change in driving technique when entering angled 
road junctions. The drivers of vans (panel trucks) have a good survival rate in 
modern traffic conditions and they have an inferior rear quarter vision. If we 
may throw out a wild suggestion, is it not possible to use fibre-optics or some 
other periscopic system to transfer a view of what lies behind the car on a 
display panel in front of the driver? 

278 The Sports Car 
The suspension 

Front and rear suspension are identical in principle to that used so successfully 
in the Elite. One noticeable difference is the shortening of the semi-trailing links 
in the Esprit. This is necessitated by the mid-engine location. The Esprit has 
205/60 HR 14 tyres at the front and 205/70 HR 14 at the rear. The tyre pressures 
of 18 p.s.i. front and 28 p.s.i. rear reflect this imbalance, accentuated further by 
the rear-end weight bias resulting from the engine location. 

Despite these fundamental differences from the front-engined Lotuses the 
Esprit corners and handles so well that the majority of professional testers are 
tempted to ‘chicken-out’ from testing to the ultimate. The low polar moment of 
inertia makes the Esprit more responsive to rapid changes in direction than the 
Elite and the Elite is no sluggard by contemporary standards. If the driver ever 
becomes so carried away as to exceed the limits of adhesion on a tight bend the 
car spins off, which is usually much safer than ploughing straight ahead. 


When tested by Motor the Esprit carried a total weight, included 100 lb of test 
equipment, of 2630 lb. This was 370 lb lighter than the fully equipped Elite 
tested 18 months earlier by the same magazine. The Esprit is higher geared; 5.69 
m.p.h. per 1000 r.p.m. in bottom gear and 8.56 m.p.h. per 1000 r.p.m. in 
second versus values of 5.2 and 8.3 on the Elite. This higher gearing tends to 
offset the weight advantage and the acceleration figures for 0-60 m.p.h. and 
0-100 m.p.h. were only marginally reduced. The top speed was not measured 
but was adjudged to be in excess of 130 m.p.h. or about 10 m.p.h. faster than 
the Elite. Such Figures do not express the appeal of the Esprit. It is not sufficient 
to say ‘the Esprit is an enthusiasts car’ since Lotus have been making nothing 
else since they went into business. The Esprit is much closer to the modern 
racing car which is controlled in a corner as much by the accelerator as by the 
steering wheel. 

General specification 


Engine continued 


4 in line 




1973 cc (120.4 cu in) 

Fuel pump 

SU electrical 


95.2/62.9 mm 

Max power 

160 bhp (DIN) at 6200 rpm 

(3.75/2.72 in) 

Max torque 

140 lb ft (DIN) at 4900 rpm 







5 speed manual 




8.5 in dia diaphragm spring 



Internal ratios 

and mph/100 rpm 

Valve timing 



inlet opens 

27.5° btdc 



inlet closes 

52.5° abdc 



ex opens 

52.5° bbdc 



ex closes 

27.5° atdc 








2 Dellorto DH LA 45 E 

Final drive 


Design Studies 279 

Body /chassis 

Construction GRP body, steel backbone 



Undersealant on chassis 



Ind. by unequal length 


wishbones, coil springs 
and anti-roll bar 

Ind. by lower transverse 



links, fixed length drive- 
shafts, semi-trailing arms 
and coil springs 

Rack and pinion 




3-5 mm (0.12 - 0.20 in) 


O-V 2 0 


3° ±Vi° 

King pin 


Rear toe-in 

8-10mm (0.32 - 0.39 in) 



Discs all round (inboard 
at rear) 




Split, front/rear 

Rear valve 






Alloy sports pattern; 7J 
rear; 6J front 


205/70 HR 14 rear; 205/60 
HR 14 Dunlop SP Sport front 



18 psi front; 27 psi rear 


12V, 44 Ah 


Negative earth 






4x5” Sealed beam 75/60 


The history of the Daimler-Benz Company is the history of the automobile 
itself. If we dismiss the vague claims of Lenoir in 1862 and Markus in 1875 as 
the inventors of the first petrol driven automobile we are left with the 
well-authenticated claims of Gottlieb Daimler and Karl Benz between 1885 and 
1886. After working on the test bed in 1885, Daimler’s four-stroke petrol engine 
was installed in what became his first limited production road vehicles at the end 
of that year. Karl Benz was working on a two-stroke engine as early in 1880, but 
his motor vehicle patent only appeared in 1886. The merits of the four-stroke 
engine were known to Benz but Otto had already patented the system in 1877. 
Benz worked for five years on a three cylinder engine in which one cylinder 
pre-compressed the air, one pre-compressed the combustible gas and the third 
was the working cylinder. By the time Benz was able to finance the construction 
of his first motor vehicle in 1885 Otto’s patent had been challenged in the courts 
and invalidated. It was July 1886 before the first Benz three-wheeler petrol- 
driven carriage appeared on the public roads, fitted with a single cylinder 
four-stroke engine having cam-operated valves and electric ignition. On this 
evidence it would appear that Daimler had the prior claim to the first practical 
automobile. Professor Kurt Schnauffer of the Technische Hochschule at 
Munich, who has made a study of the subject, insists that ‘Carl’ Benz made the 
first successful automobile. He also spells his Christian name with a ‘C\ while 

280 The Sports Car 

the Daimler-Benz literature use a ‘K’. It is purely a question for historians. Since 
the two companies started by these great pioneers amalgamated in 1926 all 
rivalry on this issue must surely have died. 

Gottlieb Daimler must have guessed that sales would soon take off like a 
rocket but he died at the turn of the century and it was left to his old friend 
Wilhelm Maybach to take over the technical and commercial control of the 
company. Karl Benz died in 1929, three years after the amalgamation. Like 
Henry Ford he had always hated motor racing, but the success of Daimler under 
the forceful direction of Wilhelm Maybach and the excellent sales growth of this 
company that many attributed to their racing triumphs, eventually led to a 
change in the board policy at the Benz company. Karl Benz decided to retire in 
1903 at this time and the new generation of Benz cars carried much larger 
engines and competed, with only modest success, in the hectic motor racing of 
the period. Their 60 horsepower 8-litre car won the Prince Henry Trial in 1907, 
but little else during the early years. They became more and more committed to 
the sport and their racing models grew to alarming proportions. The formidable 
Blitzen Benz of 1909 had a capacity of 21.5 litres from 4 OHV cylinders and 
developed 200 b.h.p. It achieved world fame as a record breaker when Barney 
Oldfield recorded a flying start mile of 131.72 m.p.h. in 1910. 

Maybach’s success with his own design of Daimler owed a little to the shrewd 
commercial presentation of the new ‘Mercedes’ model to the public, but much 
more to the excellence of the car itself. Emil Jellinek, a Hungarian entrepreneur 
ordered a relatively large number of these new cars on the understanding that he 
would receive the sole selling rights in America, Austria, Belgium, France and 
Hungary. Moreover he demanded a change of name to that of his daughter, 
Mercedes, It was not entirely sentiment that motivated this request. He knew 
that a German-sounding name would be a handicap to sales in some European 
countries. In Germany the new car was still called a Daimler, but after two years 
they decided to register the new name as a trade mark and to abandon the brand 
name Daimler. 

In the spirit of the times Maybach designed the 1901 Mercedes as a production 
car that could be raced. He broke away from the concept of bigger and bigger 
engines epitomised by the Blitzen Benz since big inefficient engines only increase 
the weight and size of the transmission, axles, springs and chassis until the 
exercise becomes self-defeating. He also knew that the tyres of the period were 
not good enough to cope for long with such enormous weights. Barney Oldfield 
must have been a very brave man! 

In a modest way Maybach adopted the design philosophy that we now see 
demonstrated in the modern Grand Prix car. He began to discard all unneces¬ 
sary weight. For his four-in-line engine he adopted mechanically-operated inlet 
valves, at a time when suction-operated inlet valves were the norm. This made 
his engine more efficient and able to operate at higher r.p.m. Cylinders were 
cast in pairs in iron, but a light aluminium alloy crankcase was used. The 
customary heavy and clumsy serpentine tube radiator was replaced by a 
honeycomb radiator. Not only was the radiator much lighter but the cooling 

Design Studies 281 

system content was reduced drastically giving a saving of about 30 lbs in water 
alone. There was a four-speed gearbox and a final drive by chains. With a power 
output of 35 b.h.p. from a capacity of 5.9 litres the new car was not quite as 
heavy as the 4 horsepower Daimler made four years earlier. 

The 35 was followed by a 40/45, a 60 and a 90. The important Gordon- 
Bennett race in Ireland in 1903 was won by stripped and tuned 60 b.h.p. touring 
cars borrowed from customers since the 90 b.h.p. team cars had been destroyed 
in a works fire a few weeks before the race. 

When Maybach left in 1907 to start his own company, Gottlieb Daimler’s 
eldest son Paul became the technical director. One of his early innovations was 
to replace the chain drive to the rear wheels with a shaft drive to the rear axle. 
Similar drives had been tried before by Renault and others and there were still 
problems to be solved before a reliable shaft drive emerged. 

During the 1914-18 war the Daimler Company worked on the application of 
supercharging to aero-engines. Roots-type blowers were used in an attempt to 
maintain ground level engine power at higher altitudes. Paul Daimler saw a 
future for supercharging in post-war racing and sports cars and he encouraged 
his development department to continue this work in the twenties. He was, of 
course, not aware of the Chadwick supercharged racing car from Pottstown, 
Pennsylvania, that had been so successful in American hill-climbs in 1907 and 
1908. This work in the early twenties eventually resulted in the fabulous SSK 
and SSKL Mercedes sports cars that were such a thorn in the side of W.O. 

The Daimler engineers developed a system of supercharging that is usually 
described as ‘blowing through the carburettor.’ The more popular method is to 
place the carburettor on the inlet side of the blower. To function correctly the 
Mercedes system requires a sealed carburettor float chamber with the top of the 
chamber pressurised by the blower. The Daimler engineers had experienced in 
their road testing a frustrating boost-lag after the throttle was closed at the 
approach to a bend. With the large volume induction manifolds used at that 
time this is not surprising. By blowing through the carburettor a pressure 
build-up occurred upstream of the throttle plate during overrun and this gave 
a much improved throttle response when the driver floored the accelerator after 
a bend. 

A two-litre blown Mercedes won the Targa Florio and the Coppa Florio in 
1924. The more famous of the supercharged Mercedes sports cars, the SSK and 
the SSKL were made after the amalgamation of the Daimler Company with 
Benz, but the team of Daimler engineers provided the expertise that made these 
cars so memorable. 

Paul Daimler’s early attempts to produce supercharged production touring 
cars resulted at first in a reputation for unreliability. One of the troubles, 
plug-fouling at low throttle openings, only demonstrated that plug technology 
had not yet found an answer to widening the operating temperature range of 
their plugs. The way of the innovator is fraught with pitfalls like this. There were 
other troubles with the early supercharged cars and the Board of directors made 

282 The Sports Car 

an approach to a young Austrian engineer called Ferdinand Porsche who had 
already made a reputation as the designer of fine cars for Austro-Daimler. 

At the end of 1922 Paul Daimler resigned from the company his father had 
established in 1885 and in April 1923 Dr Porsche left Austro-Daimler to take up 
his new appointment as Chief Engineer at Stuttgart-Unterturkheim. What a 
bargain that was! When one thinks of the tens of thousands of man-hours 
devoted on design committees today before a single new model rolls off the 
assembly line one finds it difficult to comprehend how Ferdinand Porsche was 
able to direct design teams while they produced 65 entirely different vehicles, 
touring cars, sports cars, racing cars trucks and tractors — and this accom¬ 
plished in less than six years! No wonder that he thought it advisable to take a 
short vacation before moving to a new appointment with the Steyr Company in 
his native Austria. 

Porsche had not been too happy working for the much enlarged amalgamated 
Daimler-Benz Company which showed the usual disorientation that occurs at 
such times. He need not have worried about the future products. Fine engineers 
such as Hans Nibel carried on where he left off. They introduced independent 
front suspension, swing-axle rear suspension and a new design of straight-eight 
engine. To the typical German the Mercedes is the finest car in the world and it 
was in the late twenties that the image of Mercedes began to be firmly 
established, as makers of luxury cars and as makers of the finest in sports cars. 
There has always been this dichotomy. Our interest in this book, however, is 
only in their sports cars. 

The famous S Series sports cars were designed by Ferdinand Porsche. The 
blower design was based on the earlier work by Paul Daimler. All the S Series 
had a supercharged SOHC six-cylinder engine. A unique feature for the period 
was the casting of cylinder block and crankcase as a single unit in aluminium 
alloy. Aluminium cylinder heads with durable valve inserts were still in the 
future and Ferdinand Porsche, confined to the technology of the period, used a- 
cast-iron cylinder head. The use of these two materials with different coeffi¬ 
cients of expansion, particularly on such a long engine, gave them much trouble 
at first with leaking head gaskets. This and other problems were solved in the 
painstaking manner characteristic of the man and the engine soon achieved a 
reputation for reliability in service. The Roots blower was mounted vertically in 
front of the cylinder block and was driven through a pair of bevel gears and a 
multi-disc clutch from the front of the crankshaft. The fan was also driven 
through a clutch from the front of the camshaft. This was designed to be 
disengaged at high speed when not required, thus conserving power. It is 
intriguing to those accustomed to carrying out their own maintenance that the S 
Series engine only required one special tool for maintenance. The cylinder head, 
though, was so heavy that lifting gear was required to remove it. 

The S model was introduced in 1926, the year when the amalgamation took 
place. It had a wheelbase of 3.4 metres (11 ft 2 in) and an engine capacity of 6.8 
litres (415 cu in). A low centre of gravity (low for the period) resulted from the 
use of underslung rear springs. The power with natural aspiration was 120 

Design Studies 283 

b.h.p., increasing to 180 b.h.p. with the blower engaged. The top speed, using 
the blower, was about 100 m.p.h., a very impressive performance in 1926. In 
1927 the S Model won the Nurburgring Grand Prix of Germany, a race for 
sports cars, with two more S Models finishing in second and third place. 

Competition, chiefly from Bentley Motors, speeded the development of a 
‘super’ model, the SS and in 1928, the SSK (K for kurz , short) with a wheelbase 
shortened to 2.9 metres (9 ft 6 in) and the supercharged power increased to 250 
b.h.p. from a bored out version of the original engine. The new capacity of 7.07 
litres (430 cu in) was retained for the ultimate development the SSKL (L for 
leight , light). Hans Nibel was now the design head at Daimler-Benz and his 
lightened chassis frame had so many holes down the sides that it looked like a 
slice of gruyere cheese. In a final attempt to crush the opposition this potent 
vehicle was fitted with the ‘Elephant’ blower, an oversize blower giving a boost 
pressure 0.83 bar (12 lb/in 2 ). This blower raised the maximum power to 300 
b.h.p., but this could only be used for short periods when required to pass a 
competitor. This car won the Mille Miglia on its first outing and would have had 
many more victories in later years if the effects of the thirties depression had not 
begun to bite so savagely. In 1931 Bentley Motors went into liquidation and 
Bugatti seemed to have lost his sense of direction. Only Alfa Romeo seemed to 
be able to finance an active racing department and this was not a rewarding 
exercise when the opposition had disappeared. 

The ear-splitting scream of the SSKL with the blower in action is now only a 
memory, but the Daimler-Benz Company were more soundly based financially 
than many smaller companies. They weathered the difficult times of the early 
thirties and concentrated on the manufacture of luxury cars. Even in hard times 
there are still a few buyers left for cars in the top class. 


Before Hans Nibel’s untimely death in 1934, he designed the advanced Type 
170, a 1.7 litre Mercedes touring car with an all-independent suspension system 
and was responsible for the design of the new racing car that was to start a new 
era in Grand Prix racing. Fritz Nallinger became the new Engineering Director. 
His chief assistant was Rudolf Uhlenhaut, who was gaining valuable experience 
to be applied later when he directed the design team working on the highly 
successful 300 SL in later years. 

From the end of World War 2 it took Daimler-Benz only twelve months to get 
their bombed-out factory turning out a dressed-up version of the pre-war Type 
170. Five years later they were so soundly based that their minds turned once 
more to the concept of a sporting model. In 1951 they introduced the Type 300 
luxury car to be followed by the 300S in 1952 and the 300SL sports car in 1954. 
Daimler-Benz were once more making a real sports car. 

284 The Sports Car 

THE 300SL 

The new sports car had a great appeal to engineers. The author knew several 
who could not resist the attraction of its engineering excellence. One Texas 
engineer actually bought two, a black one and a white one. The appeal to the 
public in general was the performance and the good looks. The coupe body with 
its gull-wing (up and over) doors was enhanced by a shape that was both 
aesthetic and efficiently streamlined. The final touch of showmanship was the 
large Mercedes three-pointed star in the centre of the radiator grill. The tubular 
frame had enormous strength. The body sills so wide the driver and passenger 
had to learn how to slide their posteriors across them when entering the vehicle. 
Few people found this objectionable. In fact when the body shape was changed 
in 1956 to provide more normal side-hinged doors and narrow sills there were 
many voices raised in protest. 

The engine of the 300SL was developed from the six-cylinder 3-litre engine 
already used in the Type 300. This was a SOHC design with a robust seven 
bearing crankshaft. In the 300SL the engine was tilted at 50 degrees to the 
vertical as shown in the cross-section of Figure 14.21. This helped to lower the 

Fig. 14.21 The 300SL sports car, showing a cross-section of the 
slant-six SOHC engine. 

bonnet-line. The joint-face of the block and head was also angled in the 
opposite direction at 18 degrees to the engine centre-line. An off-set yet compact 
combustion chamber was formed between the raised piston crown and a side 
pocket formed between the angled head and the cylinder block. A new head was 
made for the sports car engine in aluminium alloy. The ports were enlarged and 
an additional boss was provided to take the direct fuel ihjection nozzle. 

As a design study it is of interest to consider why the decision was taken to 
abandon the well-tried system of carburation for direct fuel injection, a system 
remarkably close to that used in Diesel engines. The answer lies in the experience 
gained by Dr Nallinger in his pre-war work on the Mercedes Diesel. Experience 

Fig* 14.22 Ghosted drawing of the 350SU 

286 The Sports Car 

often guides us in our decisions, sometimes unconsciously. Mercedes had a long 
working partnership with the Robert Bosch Company who had designed 
jerk-type pumps for Diesel engines for thirty years and were able to develop a 
suitable pump and injectors for petrol (much more abrasive than Diesel fuel) in 
a relatively short time. Part of the pump development programme involved a 
control system that compensated for changes in ambient air temperature and 
density. The combustion chamber is a hot dirty environment for such a delicate 
device as a fuel injector, but the weight of long experience won through. It is of 
interest though that Daimler-Benz turned to the less hostile location of port 
injection on later models. With fuel injection the power output of the 300SL was 
220 b.h.p. (DIN). The suspension system was based on the firm’s racing 
experience. Front suspension was by coil springs with two wishbones per wheel. 
Rear suspension was the controversial system using coil springs and swing-axles. 
Experience showed that the combination of a low roll-centre at the front and a 
high one at the rear resulted in too much roll-steer effect. Later models changed 
to a more complex arrangement as shown in Chapter seven (Figure 7.22). This 
design gave a lower pivot point below the rear drive casing. Roll resistance was 
also increased by the addition of a horizontal ‘compensating spring’. 

Fig. 14.23 The fuel-injected SOHC V-8 engine of the 350SL. 

Design Studies 287 

The transmission had four forward speeds giving maxima, quoting from ‘The 
Autocar’ road test of 1955, of 70 m.p.h. in second, 98 in third and 135 in top. 
The time to accelerate to 100 m.p.h. was 21.0 seconds, an exceptionally good 
figure for a sports car of that period. 

The 300 SLR appeared in 1955. This was an eight-cylinder racing-sports car 
using an engine that was very close to the contemporary Mercedes Grand Prix 
car engine. With regret we must omit it from this review since it was not a car 
sold to the public. 


With the introduction of the 230SL Daimler-Benz began to draw together the 
two major interests of the company. Their interest in racing in the sixties was 
reduced to a few works-supported entries in Rallies and their new sports car, the 
230SL, was soon seen to be the forerunner of the new generation of Mercedes 
sports cars — the luxury sports car. Perhaps this trend is inevitable. The supply 
of Spartans who like to feel the wind in their hair seems to be running out. The 
230SL, introduced in 1964 was still available as a convertible, but was more 
popular with coupe bodywork. It had a six-cylinder in-line engine of 2300 cc 
capacity fitted with Bosch fuel injection. The maximum speed was 106 m.p.h., 
but the car was obviously not designed with competition in mind. The series has 
been popular and has been developed over the last thirteen years with 
progressive increases in engine size and the addition of greater luxuries. The 230 
SL was followed by the 250SL, the 280SL and the 350SL. The current model is 
available with a choice of two engines, a 3.5 litre or a 4.5 litre. The new engines 
are both V-8s and are fitted with the latest Bosch electronically controlled fuel 
injection as described in Chapter Five. 

Many years had been spent in developing a good well-tamed swing-axle rear 
suspension for the Mercedes but, as happened to so many chassis designers, a 
completely new approach was demanded with the introduction of low profile 
tyres. Swing-axle suspension gives camber changes that are too large for the new 
tyres. There was very little change in the front suspension; double wishbones 
being used here, hardly differing in principle from the front suspension on the 
1934 Type 170. The rear suspension is the semi-trailing link design which has 
become so popular in Europe today. With a touch of stubbornness Mercedes- 
Benz literature insists on calling it ‘diagonal swing-axle’. What ever name we use 
it is still a ‘rose’. The author speaks with experience since his BMW is fitted with 
a similar rear suspension. 

The SOHC cylinder head design used on the V-8 engine is the well-tried 
Mercedes method of valve operation using a rocker to transfer the cam action to 
the valve rather than a bucket tappet (see Figure 3.9). A longitudinal cross- 
section of the engine is given in Figure 14.23. 




of cylinders 8 

arrangement single, overhead camshafts 
in cu ins 213.6 

in litres 3.5 

Engine output 

(DIN) 200 net bhp/5800 rpm 


rpm 6500 

in 3.62/2.59 

mm 91.7/65.6 

ratio 9.5:1 

Fuel Premium commercial fuel 

or benzine - benzol 



(DIN ft lbs/rpm 211 /4000 

mkg/rpm 29.2/4000 
Fuel Bosch, electronically con- 

injection trolled fuel injection with 

system automatic starting and 

warming up, responding 
to absolute suction pipe 
pressure, engine speed, 
cooling water temperature 
and air temperature. 


nozzles Bosch 


sequence 1-5-4-8-6-3-7-2 

Ignition Automatically, acc. by 

timing centrifugal force and 


Crankshaft 5 multi-layer bearings 

bearings with steel-backed shells 

Connecting- Multi-layer bearings with 

rod bearing steel-backed shells 

Oil filter Lubricating-oil filter with 

paper cartridge, full-flow 
Oil cooling Oil cooler in air-stream 

Capacity of 

crankcase Imp/U.S. pts 13.2/15.9 


Engine continued 

Cooling Water circulation through 

pump drive, by-pass 
thermostat, visco fan 

Electric 12 Volts, three-phase 

system current generator 

770 Watts max 


capacity 66Ah 

Power transmission 
Clutch Single-plate dry clutch 

Transmission MB-4-speed-transmission, 
fully synchronized in all 
gears, floor-type gear shift 




transmission) Optional 

MB-Automatic transmission, 

floor-type gear shift 


3.96 3.98 


2.34 2.39 


1.43 1.46 


1 1 


3.72 5.48 



43 % 43 % 



41 % 42 % 


22 % 22 % 


13.5% 13.5% 

Top speed in 

Approx, m.p.h. 

the individual 


34 m.p.h. 27 m.p.h. 



56 m.p.h. 56 m.p.h. 


93 m.p.h. 93 m.p.h. 


130 m.p.h. 127 m.p.h. 

Rear axle 




Engine speed 

at 60 m.p.h. 

2945 r.p.m. 3085 r.p.m. 


Frame Frame-floor unit, closed 

central member, box¬ 
shaped side and cross 
linked by welded sheet 
iron floor plates, self- 
supporting body, Engine- 
transmission block resting 
on two rubber bearings, 
on front axle carrier; in 
the rear with one rubber 
bearing on the chassis. 

Design Studies 289 

Chassis continued 

Front wheel Independent suspension 
suspension through twin wish-bones 

with anti-dive control 
device, coil springs with 
progressively acting rubber 
helper springs, dual-effect 
hydraulic telescopic 
absorbers, anti-roll 
bar, front axle carrier 
suspended front side 
members through rubber 

Rear axle MB-diagonal-swing axle, 
coil springs with progres¬ 
sively acting rubber helper 
springs, dual-effect 
hydraulic telescopic shock 
absorbers, anti-roll bar 
limited slip differential 

Brake system Hydraulic dual-circuit 
brakes with vacuum 
booster; disc brakes front 
and rear. 

Steering MB-Power steering with 

automatic bleeding. 

Tyre size 205/70 VR 14 with tubes 

Dimensions inches metres 

Wheelbase 96.9 2.46 

Track front 57.2 1.45 

Track rear 56.7 1.44 


length 172.1 4.37 

Dimensions continued 

width 70.5 1.79 


(unloaded) 51.2 1.30 

Turning circle ft (approx) metres (approx) 

diameter 33.9 10.0 

Tank capacity Imp/U.S. gal 19.8/23.8 

incl. reserve Imp/U.S. gal 2.9/3.4 

Weights lb kg 

Kerb weight 

(DIN 70020) 3405 1543 


total weight 4355 1977 


Top speed Manual transm. 130 mph app. 

with MB Autom.transm. 127 mph app. 


ratio lb/HP (DIN) 14.8 


(0-62 m.p.h.) 

when shifting 

through gears 8.8 seconds approx. 


according to 

DIN 70030+) m.p. Imp/U.S. gals 22/18 
(measured at 68 m.p.h.) 

at cruising 

speed m.p. Imp/U.S. gals 25/21 

+At % of top speed (max. 68 m.p.h.) with 10% 


In 1969, Daimler-Benz revealed an experimental sports car, the Cl 11 shown in 
Figure 14.24 fitted at this time with a 3-rotor Wankel engine mounted behind 
the driver and later to be fitted with a 4-rotor version of 350 b.h.p. (DIN). There 
was much speculation at the time that a Mercedes might be seen again in sports 
car racing, but the vehicle has, so far, been nothing more than a design exercise. 
It is interesting to note that the wedge profile body is fitted with gull-wing doors 
as in the original 300SL. 

The development of the Wankel engine has been overshadowed in recent 
years by two problems, high fuel consumption and poor emission control. Not 
all workers in this field have been discouraged but Daimler-Benz have decided to 

Fig. 14.24 The Cl 11 experimental sports car, first used with 4-rotor Wankel power and now used as a mobile test bed for a 3-litre 
turbocharged Diesel. 

Design Studies 291 

discontinue development work on the Wankel engine and to put renewed effort 
into a Diesel engine programme. The company still denies that the Cl 11 has any 
future as a production car and they are currently using it as a mobile test bed for 
a turbocharged 3-litre 5-cylinder Diesel engine. The car has already broken the 
long-distance Diesel-engined car world record by completing 10,000 miles at an 
average speed of 156 m.p.h. 

Mr Van Winsen, who was responsible for the development of the new 
turbocharged diesel, believes that a Mercedes sports car powered by this unit 
will be produced by 1979 and that this new sports car will have no difficulty in 
meeting the tighter emission and fuel consumption regulations that will be in 
force in the United States at that time. 

‘Diesel engines we can make without headaches,’ says Mr Van Winsen: 
‘Petrol engines are more of a problem.’ 


This story must start at the turn of the century when Ferdinand Porsche, in his 
late twenties, began to compete in early motor races with cars of his own design. 
In 1905, after taking his doctorate in engineering, Dr Porsche became the chief 
engineer for the Austro-Daimler company. An Austro-Daimler won its class in 
the 1922 Targa Florio and Hans Stuck drove one to success in the European Hill 
Climb Championship. Between 1922 and 1928 Dr Porsche was in charge of the 
design office at the Daimler Motoren Gesellschaft and as already related in the 
Mercedes Design Study he supervised the design of 65 widely different classes of 
vehicle. It was a very tired man who eventually returned to his native Austria to 
work for the much smaller Steyr company. Before leaving the amalgamated 
Daimler-Benz company he was responsible for the design of the SSK Mercedes 
sports car. 

In 1930 Dr Porsche formed his own company in Stuttgart. His intention was 
to act in a consultative capacity and to build up a design and development 
service for the benefit of the expanding motor industry. In 1933 he was asked to 
design a new Grand Prix car for a newly formed group of car manufacturers to 
be called Auto-Union. This was the famour P-Wagen with the controversial 
combination of a heavy rear engine and swing-axle rear suspension. Despite the 
handling problems associated with this fascinating vehicle that started life with a 
295 b.h.p. engine and finished in 1937 with 520 b.h.p., Dr Porsche was still 
convinced the combination could be made to work. Since he had already 
designed about 80 vehicles by this time, very few of us would have had the 
temerity to contradict him. Consequently, his next major design assignment, 
one of the most famous in the whole history of the motor car, is now known 
affectionately as ‘the Beetle’. Inevitably he designed it >\ith a rear engine and 
swing-axle rear suspension. 

After the war with the German automobile industry in ruins Dr Ferdinand 
Porsche, Jnr., usually called ‘Dr Ferry’ to distinguish him from his father, 

Fig. 14.25 Ghosted drawing of the Type 911 

Design Studies 293 

decided the only hope of re-creating the pre-war Porsche Technical Office was 
to design a sporting vehicle based on the Volkswagen. His father, Professor 
Porsche was no more optimistic but they hoped that sufficient second-hand VW 
would be available to make a start. They must have been surprised at the rapid 
recovery made by the VW factory and the first post-war Porsche, designed by 
Dr Ferry, went into production in 1949. 

The Porsche Technical Office existed before the war purely as a design and 
development establishment. The Porsche baulk-ring synchromesh had been 
designed and covered by world patents at this time. In 1948 an agreement was 
signed with the newly established Volkswagen company in which VW had full 
use of all Porsche patents free of charge. In exchange Porsche were to receive a 
royalty on every VW built. It is this dichotomy within the Porsche company, its 
manufacturing interests on the one hand and its design and development 
function on the other, that explains the employment of more than 1,000 
technicians and engineers in the research complex at Weissach. This research 
and development complex is well outside the industrial area of Stuttgart- 
Zuffenhausen where the Porsche cars are made. 

The manufacturing side of Dr.Ing. h.c. F.Porsche Aktiengesellschaft turns out 
about 15,000 cars per annum. Porsches have never been inexpensive, but they 
would certainly be more expensive if the lavish facilities at Weissach were 
supported entirely by the sale of Porsche cars. The research centre has 13,400 sq 
metres (144,000 sq ft) of workshops, foundries, test-rooms and laboratories. 
There is a ‘Can-Am’ high-speed circuit and a Mountain circuit and all the rolling 
dynamometers, exhaust emission test equipment, arctic weather chambers, etc 
that one would expect to find at the research establishment of a large 
manufacturer. It is here that the motor racing development and testing work is 
carried out, the story of which has been so admirably told by Paul Fr£re in ‘The 
Racing Porsches’. 

The history of Porsche as a manufacturer from the post-war start in 1949 to 
the present day is one of a gradual shedding of the ‘hotted-up VW’ image and 
the establishment of a true identity. The Porsche 911, which has now been in 
production for eleven years, does not employ a single component in its entire 
body and chassis from the VW Beetle. It is therefore ironic that an entirely new 
Porsche is now being marketed in Europe that is constructed from many 
components used in the current Audi and VW cars. The original 924 project, 
using a water-cooled front engine driving the rear wheels, was undertaken on 
behalf of VW, who later decided not to proceed to production. Porsche, 
however, saw possibilities in this new car and have proceeded to establish a new 
production line to make the 924 as a medium-priced sports car using many 
mass-produced Audi and VW components. They have also developed a 
water-cooled V-8 version, the Type 928. The author is not yet familiar with these 
new cars and does not propose to discuss them in this study. 

294 The Sports Car 


There is something logical in the Porsche philosophy of avoiding the idea of 
annual model changes. Every attempt is made to identify weaknesses that are 
thrown up by customer liaison, to work on these problems and introduce 
changes and modified components as soon as they have been approved by the 
development department. Every attempt is meanwhile made by the staff at 
Weissach to improve the reliability, road-holding, handling, braking, etc based 
on new information emanating from racing experience. Henry Ford also saw the 
logic of continuity when he refused to replace the Model T. He did very little, 
though, to improve the car. Porsche avoid this mistake. The original Type 911 
in 1966 had a 2-litre engine. Since then the engine size has increased to 2.2, then 
2.4 and is now available in 2.7 or 3.9 litre form. A ghosted view of this most 
popular of Porsches is given in Figure 14.25 

The engine 

Porsche have used horizontally opposed air-cooled engines since the early 
post-war days. The first VW-based Porsche was a flat-4. Since then there have 



flywheel side 



Bearing 1: with shoulder, sollc 



pulley side 



Bearing 2i split 



flywheel side 



Bearing 3: solid 



pulley side 



Bearing 4: solid (pulley side] 

Fig. 14.26 Simplified layout of a flat-4 engine. The same layout, with 
main bearing between pairs of opposed connecting rods is 
used on 6, 8 and 12 cylinder Porsche engines. 

Design Studies 295 

been 6s, 8s and 12s. The racing engines, including the flat-12 in the fabulous 
Type 917, have all been described in Paul FrSre’s book. What is difficult to 
describe in a few words are the benefits that accrue when a company is able to 
test their products in the harsh fire of competition. Some large companies claim 
that modern laboratory techniques make racing and rallying completely out¬ 
dated as test methods. They could be right, but it is difficult to infuse a 
competitive spirit into a computer-controlled series of laboratory tests. Motor 
racing soon reveals if your product is much inferior to the opposition. 
Laboratory testing only tells you how good your own product is. 

Porsche use both methods, but their racing programme involves a consider¬ 
able expenditure. The publicity gained by these racing successes is difficult to 
evaluate in financial terms, but it is very reassuring to the owner of a Type 911 to 
know that similar engines to the one in his car have been tested in long distance 
races and with the engine tuned to produce much more power than in the 
production model. 

Figure 14.28 gives a longitudinal and a transverse cross-section of the Turbo 
engine which is similar in basic layout to the Type 911. For those overwhelmed 
by the mass of detail Figure 14.26 shows how banks of cylinders on one side are 
staggered relative to those on the opposite side. This is the earlier flat-4 engine. 
American readers familiar with the V-8 should visualise the flat-6 as a V-6 with 
the included angle between banks increased to 180 degrees. Perhaps this 
description appeals more to our Irish readers! A flat-6 fits into a rear-engined 
sports car as admirably as a flat-12 into a Formula 1 racing car. Dr Porsche 
always had a firm commitment to air-cooling. The advantages of air-cooling 

1. Low weight (even alloy radiators are heavy), 

2. Greater reliability, since that constant source of trouble, the burst or 
leaking hose, is absent. 

The disadvantages are: 

1. Slightly higher internal operating temperatures than those in a water- 
cooled engine of the same specific power output, 

2. A higher level of mechanical and combustion noise, since there are no 
water jackets to help damp out some of the noise. 

Air-cooling on the Type 911 is provided by a large ducted fan mounted above 
the engine and belt driven from a pulley on the front of the crankshaft. A SOHC 
head is used, rockers being used to transfer the cam lift to the valves which are at 
an included angle of 55 degrees. The two camshafts are chain-driven from the 
crankshaft. The cylinder barrels are in aluminium alloy with chrome-plated 
bores. The crankcase, which is made in two halves, was originally an aluminium 
alloy casting but, as a result of racing experience, it is now a much lighter 
magnesium alloy casting in the 2.7 litre engine. The Bosch K Jetronic fuel 
injection has already been described in Chapter Five. To comply with future 
legislation in several countries to reduce the lead content of fuels the Porsche 
911 is now designed to run on 2 star petrol. 

Design Studies 295 

296 The Sports Car 


The transaxle used on the type 911 (Figure 14.27) is cast in silicon aluminium 
alloy. The gearbox is overhung behind the final drive section. The gear selector 
rod can be seen at the bottom of the box. The Porsche design of automatic 
transmission is available as an optional extra. 

Fig. 14.27 Gearbox and final drive unit as used on the Type 911. 

The body 

The Porsche is a small car compared with the Jaguar, the Lotus and the 
Mercedes. Porsche sports cars have always been small high performance sports 
cars with two comfortable front seats and two tiny seats in the rear for children. 
The Porsche 911 is 14 ft 1 in long and 5 ft 2>Vi in wide. The Jaguar XJ-S is 16 ft 
long and 5 ft \QVi in wide. The Jaguar has 75 per cent more power than the 
Porsche, but it weighs 45 per cent more in road trim with two people aboard. 
These figures are not given to denigrate the Jaguar, only to show that Porsche 
have always believed that ‘small is beautiful’. Drag coefficients as low as 0.31 
have been achieved by several racing Porsches, but the addition of aerofoils and 
other spoiler devices to aid stability at speed increased this figure. In the same 
way, but less seriously, the addition of those bulky ‘5 m.p.h. collision’ bumpers 
has increased the total drag of the latest 911 slightly. The coefficient in the 
region of 0.32 is still very low for a production sports car. 

The body is well equipped and carries the usual Porsche high standard of 
finish. The galvanised under-body is guaranteed for six years. Besides the 
normal coupe body, those with a taste for fresh air can choose the Targa model 
which carries a detachable roof section that can be stored in the boot. The fixed 
roof section acts as a roll-over bar. 

The suspension 

The suspension has been described in Chapter eight. In recent years Porsche 

Design Studies 297 

have concentrated, not only on high cornering power, but on safe predictable 
handling. Under moderate cornering forces the car understeers to a small extent. 
Acceleration through a corner gives a slight roll-steer effect. With a Porsche it is 
always safest to brake before a bend, then to accelerate through it. If a need to 
lift-off then occurs the behaviour remains stable and predictable. 

At speeds above 120 m.p.h. the standard Porsche body, with no aerofoil, can 
be a little twitchy, especially in a cross-wind. The Turbo with a maximum speed 
exceeding 150 m.p.h. has a substantial aerofoil blended attractively into the 
body to create the necessary downthrust for high speed stability. 

General specification 

Engine Body/chassis 


6, horiz opposed 


All steel integral 


2687 cc (163.97 cu in) 


Floorpan, sills and wheel- 


90x70.4 mm 

arches galvanised; PVC 


(3.54x2.77 in) 


underseal; all cavities Tec- 
tyl treated. 6 year warranty 


Light Alloy 



Light Alloy 


Independent by MacPher- 


Sohc per bank 

son Struts, longitudinal 

Valve timing 

(at 1 mm valve clearance) 

torsion bars, anti-roll bar. 

inlet opens 

6° atdc 


Independent by semi- 

inlet closes 

50° abdc 

trailing arms, transverse 

ex opens 

24° bbdc 

torsion bars, and anti-roll 

ex closes 

2° btdc 






Bosch K Jetronic fuel 


Rack and pinion 





8 main 

Toe in 


Fuel pump 

Bosch electric 




Max power 

165 bhp (DIN) at 5800 rpm 


6° 5' ±0° 15' 

Max torque 

173.51b ft (DIN) at 

Rear toe in 

0° 20' 


4000 rpm 



Discs, front and rear 


5 speed manual 




Sdp diaphragm spring 


Dual, split front/rear 

Internal ratios and mph/1000 rpm 

Rear Valve 








Light Alloy, 6 in rim 




185/70 VR 15 




29 psi front; 34 psi rear 






Final drive 



66 Ah, 12V 








60/55 W Halogen 

Design Studies 299 


The name Carrera once indicated a highly competitive racing sports car that 
yielded very little of its performance and handling to creature comforts. The 
latest Carrera has a top speed in excess of 140 m.p.h. and can accelerate from 
zero to 60 m.p.h. in 6 seconds, yet the same lavish equipment available on the 
Type 911 is standard equipment on the Carrera. With thermostatic heater 
control, even an electrically-heated outside door mirror the new Carrera 
combines the luxurious interior comfort of the Type 911 with the handling of a 
racing car. 

The Carrera is equipped with wider wheels and tyres at the rear. The front 
wheels on the Carrera are 185/70 VR 15 on 6 inch wide wheels, identical in fact 
to those at both ends of the Type 911. At the rear 215/60 VR 15 tyres are fitted 

Fig. 14.29 Comparative power and torque curves for the 2.7 
litre Carrera and the 3 litre Turbo. 

300 The Sports Car 

to 7 inch wheels. Forged aluminium alloy wheels are used on the Carrera to 
resist the higher stresses liable to be applied with the higher performance. With 
the increased output of 200 b.h.p. from the 3-litre engine and an increased area 
of rubber at the rear the traction and cornering power of the Carrera is even 
more impressive than that of the 911. The rear wheel arches are widened to take 
the wider tyres. 


The engine 

The principle of turbocharging has been discussed in Chapter Thirteen. Full use 
of the experience gained in turbocharging the Type 917 racing sports cars has 
been made in applying the technique to the new Turbo model. Longitudinal and 
transverse cross-sections of the Turbo engine are given in Figure 14.28. The 
turbocharged Type 917 produced anything from 175 to 200 b.h.p. per litre, 
depending upon the boost control setting, the Production model is controlled to 
a boost pressure of 0.8 bar (11.8 lb per sq in). This modest amount of 
supercharge increases the maximum power of the 3-litre engine from 205 to 260 
b.h.p., as shown in Figure 14.29. An interesting aspect of the boost control 
system is the useful lift given to the power curve in the middle speed range. At 
4,000 r.p.m. the power is increased by 50 per cent. At 6,200 r.p.m. where the 
Carrera engine peaks the power increase is less than 25 per cent. Tuned in this 
way the mid-range acceleration is much improved yet the maximum power 
output is kept within safe limits. 

The transmission 

Apart from the use of a high strength silicon aluminium alloy in the transaxle 
casings the transmission layout is largely similar to that of the 911. The final 
drive ratio varies depending upon the choice of tyre section since the rear wheels 
can be fitted with either 50 Series or 60 Series tyres. Wider gearwheels are used 
in the Turbo gearbox to withstand the increased torque. A five-speed gearbox is 
not required on the Turbo since the low speed torque is so good. 

Suspension, wheels and tyres 

Suspension layout is largely similar to the Type 911. The rear radius arms are 
made as aluminium castings of generous section to give increased strength. The 
rear hubs are also more robust, being the pattern used on the Type 917. Bilstein 
gas-filled dampers are used and the suspension geometry is designed to suit the 
ultra-low profile tyres. As in the Carrera forged aluminium alloy wheels are 
used, the front being 7J section, the rear 8J. 205/50 VR 15 tyres are fitted at the 
front and a choice between 225/50 VR 15 or 215/60 VR 15 at the rear. 


The body changes introduced with the Turbo are pronounced flares at the wheel 
arches, a small spoiler added to the front apron and a very elegant aerofoil at the 

Design Studies 301 

rear. The widened wheel arches are neatly blended into the body contours. The 
rear ‘spoiler’ (the name used by Porsche for what is technically an aerofoil) is 
not only aesthetically pleasing but with characteristic ingenuity the Porsche 
engineers have incorporated in the upper surface additional cooling vents to 
improve engine cooling when idling, to cool the air conditioning condenser and, 
by means of a separate passage, to admit air to the engine cooling blower. 

All windows on the Turbo are tinted. The rear window has its own wiper. 
There is two stage heating of the rear window and single stage heating of the 
front screen. Interior heating is thermostatically controlled and the external 
mirror is not only heated but can be adjusted manually by the driver from an 
internal lever. Headlamp washers are also a part of comprehensive equipment. 
In keeping with the image of this new leader in the Porsche range the Turbo has 
a Blaupunkt stereo-casette player and radio with 4 speakers and an electric 
aerial. With all this lavish equipment the kerb weight of 1195 kg (2635 lb) is 
remarkably low. 


The maker’s figures for the Turbo maximum speed is in excess of 250 k.p.h. 
(155 m.p.h.). The acceleration time for 0-100 k.p.h. (0-62 m.p.h.) is given as 5.5 
seconds, which appears to put it ahead of all the European competition. 


Acceleration, 212-219 
time, 0-60 mph, 215-219, 220 
Ac kermann centre, 127,128 
Aerodynamic drag, 178 
lift, 180 

stability, 179,182-183 
Aerodynamics, 177-185 
Aerofoil, 181 

Air cooled engines, 99-100 
dam, 182,185 
fuel ratio, 67 
control, 241-243 

AiResearch turbocharger, 238, 239 
Alcohol, 21, 26 
Alfa Romeo, 8,14-15 
Aligning torque, see Self-aligning torque 
Angle of yaw, see Yaw angle 
Anti-dive, 151,152-153,246 
Anti-lock braking, 247-248 
Anti-roll bar, 133,137 
Anti-squat, 246 
Apperson ‘Jack Rabbit’, 8 
AP stabilised suspension, 245-246 
Aquaplaning, 124,125 
Aston Martin, 8 
DBR1 engine, 57 
DBS performance, 220, 221 
suspension, 159, 160 

Austro-Daimler, 7, 282, 291 
Automatic transmission, 187,190-193 
Axle, live, 118,135,149,151 

Balance pipe, induction, 38 
Baulk ring synchromesh, 189 
Beam axle, see Axle, live 
Bearing loads, 86, 87 
metals, 89-90 
pressures, 86-89 
Bentley, 8,12,14, 32-33,211 
four-valve head, 32-33 
performance, 211 
Benz, 3, 7, 279, 280 
Benzol, 26 

Blocker synchro, see Baulk ring 
Body, fibreglass, 173-175 
shape, 177-185 

Bosch anti-lock braking, 247-248 
fuel injection, 79-80 
Bounce frequency, 146,147 
resonance, 146 
Brake cooling, 205, 209 
disc, 205-210 
drum, 203,204,205 
fade, 203-204,209 
forces, 201-203, 205 
lining material, 204 
pad material, 204, 210 

Index 303 

Braking, 131-132 
anti-lock, 247-248 
weight transfer, 201, 203 
Bugatti, 7,8, 12-14, 171 
Bump, suspension, 137, 161, 163, 166, 


Camber, 118, 133, 153 
Camshaft drive, DOHC, 18, 19, 56-57, 58 
SOHC, 58-61 
Carbon deposits, 94-95 
Carburettor metering, 67 
SU, 67-74 

Carburettors, triple, 38-40 
twin, 37-40 
V-8, 40-44 
Castor angle, 152 
Cavitation, 103 
Centre of gravity, 138 
of pressure, 180, 181, 183 
Centrifugal forces, 128, 129 
CFR engine, 26 

Chapman, Colin, 12, 16, 263, 265 
strut, 134, 135 
Charles’ Law, 23, 51 
Chassis design, 169-176 
torsional stiffness, 169-172 
Chevrolet Corvette, 16, 208 
Chevrolet-Weslake four-valve head, 55-56 
Chrysler hemi-head, 30, 32 
crankcase, 84 

Citroen self-levelling suspension, 245 
Clutch, 190, 191 
Coil ignition, 104-111 
Cold air intake, 50-52 
Combustion chamber research, 19-20 
shape, 19, 26-27 
types, 28-34 

Compression ratio, limiting, 26 
Connecting rod, 93, 294 
Constant velocity universal joint, 195, 

196, 197 

Contact patch, see Tyre footprint 
Contamination, oil, 94, 95 
Cooling, air, 99-100 
water, 100-103 

Cord angle, see Tyre, cord angle 
Cornering, see Tyre, cornering 
Cosworth-Ford Formula 1 engine, 33 
V-6 engine, 61-63 

Coventry-Climax cylinder head, 28-29 
Crankcase, 82-84 

Crankshaft, 84-85 

Cylinder dimensions and power, 214-215 
and torque, 214-215 
head design, 19, 28-34 
history, 18 

Daimler, 3, 7, 279, 280, 281,282 
Daimler-Benz, 279, 280, 282, 283, 287, 

Damper, suspension, 160-168 
characteristics, 161-165 
double-tube, 165-167 
Girling, 164, 165, 167 
Koni, 166 

single-tube, 167-168 
Woodhead-Munroe, 167-168 
Datsun, 260Z body, 173, 174 
front suspension, 157-160 
rear drive unit, 193 
suspension, 158, 160 
280 Z performance, 220, 221 
De Dion axle, see Suspension 
Desmodromic valve gear, 63-66 
Detonation, see Knock 
Diesel engine, 229-231 
Differential, limited slip, 197, 198 
Disc brake, see Brake, disc 
Drag, 177-180 

coefficients, 177-178, 219, 221,222 
Drift, four wheel, 129-130 
Drive, front wheel, 131, 198, 263, 265 
rear wheel, 128, 129-130 
Duckworth, Keith, 32, 61 
Dyke’s piston ring, 97-98 

Emission control, crankcase, 73 
exhaust, 67, 74, 223-224, 226, 229, 231, 
233, 237, 241-244 
End-gas, 23, 26, 27 
Ethyleneglycol, 101 
Exhaust pipe design, 52-54 
pressure waves, 54 
residuals, 20 
silencer, 52-53 

Fan, 101, 102 
Ferrari, 6, 15,33,263 
308 GT4, 6 

performance, 220, 221 
Fibre-glass body, 173-175 
Final drive unit, 193, 194, 195, 198 
Flame front, 23, 26 

304 The Sports Car 

Flow, laminar, 179 
streamline, 179 
Ford GT40,16 
Mustang, 16-17 
V-8 racing engine, 213, 214 
Ford-Cos worth engines, see 
Cos worth-Ford 
Forward ram intake, 50 
Four-valve cylinder head, 60-63 
Four wheel drift, 129-130 
Frame, chassis, 169 
tubular, 172 

Front wheel drive, 131,198, 263, 265 
Fuel injection, 74-81 
Bosch K-jetronic 79-80 
Lucas, 75-79 

Future developments, brakes, 247-248 
engine, 223-244 
suspension, 244-246 
transmission, 246 
tyres, 244 

Gas turbines, 224-227 
Gearratios, 187,188 
Gearbox, Jaguar E Type, 190 
Porsche, 194 

Girling anti-lock brakes, 247 
brakes, 206,207 
Glass-fiber, see Fibre-glass 
Gyration, radius of, 147 
Gyroscopic effects, 133 

Healey Silverstone, 15 
Hemispherical combustion chamber, 19, 

Heron cylinder head, 29 
Honda CVCC engine, 237 
Horsepower comparisons, 213 
definition, 212 
true measurement, 212 
versus cubic capacity, 214-215 
Hotspot, 21 

Hydroplaning, 124,125 
Hypoid gears, 195 

Ignition, conventional system, 104-108 
Lucas OPUS system, 109-111 
magneto, 104 

transistorized, 106,108-109 
Independent suspension, 133,134,148- 
150, 171-172 
Induction, 35-49 
ram manifold, 36-37 
ramming pipes, 44-49 

In-line engines, 37-40 
Intake, cold air, 50-52 
forward ram, 50 
Iso-octane, 25-26 

Jacking action, 155,156 
Jaguar C Type brakes, 205 
design study, 250-262 
E Type gearbox, 190 
six-cylinder engine, 43 
suspension, 151,152 
XJ-S cylinder head, 29-30 
description, 254-260 
engine, 253-254, 256, 257-258 
lubrication circuit, 91-93 
performance, 220,221 
specification, 260-262 
suspension, 151,153 
valve operation, 58-60 
Jounce, see Bump 

Kamm tail, 179,182,185 
K-jetronic fuel injection, 79-80 
Knock, 22-26 

Lambda-Sonde, 243 
Lamborghini Espada performance, 220, 

Lancia Beta cylinder head, 30, 31 
front wheel transmission, 197, 198 
performance, 272 
Latent heat of fuel, 21 
Le Mans, 4, 8-12 
Lift at high speed, 180 
Limited slip differential, 197-198 
Live axle, see Axle, live 
Long-life car, 248-249 
Lotus, 12,16 
design study, 263-279 
Eclat acceleration, 220 
description, 272, 273, 276 
Elan body-chassis, 173 
Elite body-chassis, 173-176, 268, 270- 

description, 265-272 
engine, 264-266 
specification, 271 
suspension, 272 

Esprit, description, 274, 275, 276-278 
maximum speed, 221 
suspension, 278 
Seven performance, 220,221 
Lubricants, 93-95 

Index 305 

Lubrication pressure, 86-89 
system, 91-93 
Lucas fuel injection, 75-79 
ignition system, 109-111 

MacPherson strut, see Suspension 
Magneto, 104 
Manifold, exhaust, 52 
induction, 35-42 
Maserati, 15 

Bora performance, 220, 221 
Maximum speed, 219-222 
Mazda-Wankel engine, 229 
Mercedes, 7 

Cl 11 Prototype, 289-291 
design study, 279-291 
300SL, 16, 283,284-287 
300SLR, 16, 65-66, 287 
350SL engine, 44, 286, 287 
specification, 288-289 
valve operation, 57 
SSK, 16, 281 

Mercedes-Benz, 8,14,16 

Mercer, 8 

MG, 8,10,14,16 

MGB GT V-8 performance, 220,221 
Michelin X tyres, 121 
Mid-engined sports car, 132 
Morgan Plus Eight performance, 220,221 
suspension, 134 
Motor sport, 1-2 

No-roll suspension, 244-246 
NSU-Wankel engine, 227-229 

Octane rating, 25-26 
Offenhauser engine, 213,214 
Oil, 93-97 
Overdrive, 188-189 
Overhead valve engine, 18,19,43,44, 

Oversteer, 126,127,128,132,139,183 

Panther Lima, 13 
Pent roof cylinder head, 18, 32-34 
Performance of typical sports cars, 220, 

standards, 211-212 
Pinking, see Knock 
Piston design, 95-96 
lubrication, 96-97 
rings, 95, 96, 97-98 
Pitching, 24,146-148, 245, 246 

Pneumatic tyre history, 112 
Porsche Carrera, 299-300 
design study, 291-301 
engine, 294-295 
suspension, 155,156, 296-297 
transmission, 194,296 
Turbo, 181, 241, 242, 300-301 
Type 911,294-297 
performance, 220,221 
Type 917, 30, 31, 35, 208-209 
performance, 211,212 
Port design, 21-22 
Power, see Horsepower 
Pre-ignition, 26 
Pump, oil, 86, 91-92 
water, 92,103 

Pushrod valve operation, 55-56 

Radiator, 100-101 
Radius of gyration, 147 
Ramming pipe, exhaust, 54 
inlet, 46-49 

Rear suspension, Aston Martin, 159,160 
Datsun, 157-160 
Jaguar, 152 
Lotus, 272, 278 
Mercedes, 136,282,286, 287 
Porsche, 156-157 
wheel drive, 129-130 
Rear-engined sports car, 138-139 
Rebound, suspension, 137,161,163, 


Reliant Scimitar performance, 220, 221 
Resonance bouncing, 146 
crankshaft vibration, 85 
pitching, 146-148 
Ricardo, 18 
cylinder head, 19,27 
Road ripples, effect on suspension, 142- 
surfaces, 114-116 
Roll, 133,138,153,155,246 
angle, 133,136,137 
centre, 133,134,135,136,138,153 
couple, 133 
resistance, 136 
stabiliser, 137 
Rollover bar, 175,295 
Rotating combustion engine, 227-229, 230 
Rzeppa universal joint, 195,196,197 

Saab Turbo, 241 
Safety regulations, 176 

306 The Sports Car 

Self-aligning torque, 116,130 
Semi-trailing link, 153,155 
Shelby Cobra, 17 
Shimmy, 150 

Shock absorber, see Damper 
Side valve cylinder head, 19 
Silencer, exhaust, 152-153 
Slip angle, see Tyre, slip angle 
Sparking plug, 104-106 
Speed maximum, 209,219-221, 301 
Spoiler, 181 

Sports car, definition, 2 
history, 3-17 

Spring acceleration, 141,142,145 
coil, 153,160 
deflection, 142,144 
frequency, 141,142,143,145,146 
semi-elliptic, 135,142 
torsion bar, 148,149,153 
Sprung mass, 138,141,142,143,145,146 
Squish, 23, 28 
Steam engine, 233-234 
Steering layout, 127-128 
Stirling engine, 231-233 
Stopping distances, 200 
Stratified charge engine, 235-238 
Streamlining, 177-185 
Stroke-bore ratio, 22 
Stutz, 8 

SU carburetter, 67-74 
Supercharging, 281, 282, 283 
Suspension, Chapman strut, 134,135 
Damper, see Damper 
DeDion, 118,135,136,137,149,159, 

live (beam) axle, 135,148,149,150 
MacPherson strut, 134,135,157,160 
semi-trailing link, 153,155 
swing-axle, 136,153,155 
trailing link, 134,153,155 
wishbone, 134,137,148,149,151 
Swing-axle, see Suspension 
Synchromesh, 189 

Torque, 186,187, 212, 213, 214, 215, 216, 
definition, 212 
multiplication, 186-187 
versus cubic capacity 214-215 
Torsional stiffness, of chassis-body, 
169-172, 173, 175 
vibrations, of crankshaft, 85 
Traction, effect on cornering, 119-120 
Trailing link, see Suspension 

Tramp, 150 
Transaxle, 194 

Transistorized ignition, 106,108-109 
Transmission, automatic, 187,190-193 
final drive, 193,194,195,198 
manual, 187-190 

Triumph four-valve cylinder head, 33-34 
TR7 performance, 220, 221 
Turbo-charging, 238-241 
Turbulence, 22, 23, 27, 28 
TVR 300ML performance, 220, 221 
Tyre construction, 120-126 
compounds, 123-124 
cornering force, 117,119,120,128, 
power, 117-120 
footprint, 115-117,121 
grip on the road, 112-116,200-201 
pressure, 117-118 
profiles, 117,118 

slip angle, 116,117,118,119,120,126, 
tread pattern, 115,116,122,124 

Understeer, 126,127,128,129,183 
factors leading to, 132-133 
Unitary body-chassis, 172-173,174 
Universal joints, 195-197 
Unsprung mass, 143,144,145,146,148, 

Upper cylinder lubricants, 93 

Valve operation, DOHC, 56-57, 58 
push-rod, 55-56 
SOHC, 58-61 
Vauxhall, 7 
TT engine, 18 
Venturi, 67,71,74 
Viscosity, fuel, 72 
Volumetric efficiency, 20-22 
Volvo’s ‘thinking engine’, 243 

Wankel engine, 227-229, 230 
Water cooling, 100-103 
Wedge cylinder head, 28 
Weight transfer under braking, 201-203 
Weslake, 55-56 
Wheel lifting, 136-138 
sizes, 145 

Wishbone suspension, see Suspension 
Yaw angle, 129,130,131 
ZF automatic transmission, 191-192