'^.
V fif \ x ^r
M ^ x
THE LIBRARY
OF
THE UNIVERSITY
OF CALIFORNIA
DAVIS
->^' xX \
,; ^ V
^
5^ AV
IT ^
.r '
A
":.' y
/y re
_ .
^T*
r,"
r
. / . *~ \^-
y .
' / r ' -,
MODERN ENGINES
AND
POWER GENERATORS
MODERN ENGINES
AND
POWER GENERATORS
A PRACTICAL WORK ON PRIME MOVERS
AND ^HE ^RANSMISSION OF POWER
STEAM, ELECTRIC, WATER, AND HOT AIR
BY
RANKIN KENNEDY, C.E.
AUTHOR OF
; ELECTRICAL INSTALLATIONS " " ELECTRICAL DISTRIBUTION BY ALTERNATING CURRENTS AND
TRANSFORMERS" "PHOTOGRAPHIC AND OPTICAL ELECTRIC LAMPS" AND NUMEROUS
SCIENTIFIC ARTICLES AND PAPERS ON MECHANICAL AND ELECTRICAL ENGINEERING
WITH MANY HUNDRED ILLUSTRATIONS
VOL I.
LONDON :
THE CAXTON PUBLISHING COMPANY
UNIVERSITY ex JALIFORNIA
DAVIS
PREFACE TO VOLUME I
THE term Modern Engines applies not only to engines of modern design, or invented
in recent years, but also to all engines found to serve useful and efficient purposes at
this date. Many old devices and designs have survived, and are worthy of a place when
presented in their more modern developments.
The most highly developed machinery cannot in every case be employed by the
engineer. In many instances rough and ready methods are necessary, and efficiency
is sacrificed for the sake of convenience and simplicity. He has often to make use of
what materials can be most readily found in his neighbourhood, and while a triple-
expansion steam engine and a multitubular boiler would be highly appreciated and very
economical as a prime mover, such an outfit would be in many cases impossible where
less complicated and more primitive machinery would be quite successful. References
are therefore made to older types still in practical use.
Each Volume of this work will contain some special feature. This First Volume
discusses the fluid pressure machines operated by fluid impulse, by air, water, and
steam flowing under a difference of pressure. These include that most important prime
mover, the steam turbine, which therefore forms the special feature of Volume I. Being
a new subject, and of great interests involved, the prior patents covering the funda-
mental inventions have been briefly noticed, as well as all the more recent inventions
worthy of notice on this subject.
The application of the steam turbine to marine propulsion will be further discussed
later on in this work under " Marine Engines."
The windmill is worthy of attention in connection with the possibility of trans-
mitting and storing the energy generated. Combined with storage, the best site for
the mill may be chosen, and the energy delivered where required at any reasonable
distance.
The water turbines selected for description are the latest of their types, with special
reference to governing. Injectors and centrifugal pumps are shortly discussed. These,
although not prime movers, are important impulse machines closely connected with
prime movers operating by momentum and fluid velocities.
With a description of the best rotary piston engines which have hitherto been made,
this Volume concludes the survey of the more elementary prime movers and their
accessories.
The next Volume will contain a full treatment of internal combustion engines gas
and oil, also of hot air and furnace gas engines ; motor car engines and locomotive
engines will follow.
RANKIN KENNEDY.
CHAPTER I
PAGE
INTRODUCTORY ........ i
CHAPTER II
CLASSIFICATION OF PRIME MOVERS. ...... 8
WINDMILLS .......... 9
HYDRAULIC MACHINERY ........ 28
WATER TURBINES ......... 33
PULSOMETERS HYDRAULIC RAMS . . , . . . 8 1
STEAM JETS INJECTORS ........ 86
WATER JET PROPELLER ........ 104
CENTRIFUGAL PUMPS . . . . . . . .no
CHAPTER III
STEAM TURBINES . . . . . . . . .123
GAS TURBINES ......... 190
DESIGNING OF STEAM TURBINES . . . . . . .193
CHAPTER IV
ROTARY PISTON ENGINES . .... 200
LIST OF PLATES
TURBINE AIR PROPELLER ..... Frontispiece
PLATE I. GUNTHER'S TWIN VORTEX TURBINES . . . Facing page 32
,, II. HYDRO-ELECTRICAL PLANT, WITH SENSITIVE GOVERNOR . ,, 66
,, III. JET IMPULSE TURBINE ......,, 78
,, IV. CENTRIFUGAL PUMP AND TURBINE COMBINED . . ,, 96
,, V. STEAM TURBINE CENTRIFUGAL PUMPS IN SERIES . . ,, 128
,, VI. SMALL DE LAVAL TURBINE MOTOR . . . . ,, 144
,, VII. PARSONS' TURBO ALTERNATOR . . . . ,, 176
MODERN ENGINES AND POWER
GENERATORS
CHAPTER I
INTRODUCTORY
THE heat engine has been the theme of many great treatises. Like most other
scientific subjects, it is easier to theorise and pursue mathematical investigation on
assumed conditions than it is to invent, improve, design, and construct working engines
or other practical appliances. While theories and mathematical conclusions are valuable
guides, the results of actual machinery in practical tests are of immensely greater value,
so that, while giving due weight to the theoretical side of the question, we shall devote
this work principally to inventions, improvements, designs, and constructions in prime
movers.
We know only two types of prime movers the heat engine and the electric
engine. Windmills and water wheels derive their energy primarily from the heat of
the sun. Steam, oil, and gas engines derive their energy from fuel heat ; they are all
heat engines.
The electric engine has yet to be made practicable. The only direct source of
electrical energy is the Voltaic battery, and that supplies only a small quantity at
immense expense. All the electric motors so much in use are driven by heat engines ;
they are at present only transmitters of the power or energy of heat engines. By their
aid, however, we are enabled to accomplish work with the heat engines which cannot
be done by any other means. The simplest and earliest heat engines were driven by
wind and water, and these have again been made more useful by the electric motor,
transmitting their power to more convenient places where it can be better used.
Within recent times considerable advances have been made in prime movers ; the
dreams of early inventors have been realised in the steam turbine, the rotary steam
engine, the gas engine, the oil engine, and motor car engines. Power from heat engines
transmitted electrically to electric motors has also revolutionised the problems of rapid
transit on street railways, underground railways, and is at present forcing the attention
of main line railways.
The education of modern engineers also attracts much attention. By this is not
merely meant the preparation of youths who intend to adopt engineering as a profession
VOL. i. i
Modern Engines
or trade, but the education of the more important class the owners, controllers, and
employers in engineering works. The staff of a concern may be ever so well educated
and skilled without much progress being made if the controlling head is not keenly alive
to improvements and new inventions. There is need for more alacrity in adopting new
and improved methods a'nd machinery, which may be due to want of knowledge regarding
them. There is no standing still in scientific manufactures at this date.
No employers or controllers can calculate, like their forefathers, upon making certain
articles without improvements for a long lifetime ; for good or ill those days are gone.
To illustrate the meaning of these remarks two instances in the author's intimate
knowledge may be quoted first, the now highly successful electric tramways in our
cities have been, 90 per cent, of them, equipped by American engineers with
American machinery. Yet there is nothing in the equipment which can be said to have
originated in America. The overhead trolley, the electromotor, the controller, the
dynamo, and all the special electrical apparatus were well known to every electrician
about twenty years ago in Great Britain, and it is quite safe to say that a score of skilled
electrical engineers in Britain could have been found in the year 1888, any one of whom
could have designed and laid out an electric tramway dynamos, motors, trolleys, and
all. There was no lack of education and skill among the youths nor among experts,
but no employers or controllers with manufacturing facilities and capital could be found
to take up what has been a very lucrative business to the Americans. The Americans
have followed up their success by establishing large factories in England, so as to retain
their hold upon the tramway and railway electric business.
Take also letterpress printing machinery : the largest newspapers, and some smaller
ones, are all printed on American presses. A new printing establishment putting in new
machinery nowadays must have the quickest and best working presses, machines with
continuous revolving cylinders and two revolution or perfector machines. Three large
press manufacturers in America have been shipping these improved presses in quantity
to our printing establishments, simply because the home manufacturers go on making
presses without improvements, and not at all for any lack of skill and knowledge in the
younger generation of engineers.
Whatever may be the cause of it, there can be no doubt that the class to whom
we must look for the encouragement of the advances in practical engineering require
educating, in some way or another, to open their eyes to the value of promptly seizing
upon every opportunity for securing improvements. It is very well for a country to
have the honour and glory as the birthplace of improvements, but it is better, very much,
if its captains of industry take advantage of them in time to advance the trade and
industry of the country before outsiders step in.
It may be said that capitalists and captains of industry can avail themselves of
expert advice ; that, however, while the best course to take in ordinary matters, is
not successful in questions regarding the adoption of new methods or manufactures.
The expert desires to be, above all things, a "safe man," and naturally is conservative
and timid when faced with responsibilities involved in new departures.
If every speculator consulted the family lawyer on the question of his investments
there would be no enterprise in this world, no risks, no new departures. It is the
robust, acute man, with a clear mind and depending upon his own education and
knowledge, who sees and grasps the possibilities of a step in advance ; and we want
some powerful stimulant to arouse a class of capitalists and controllers in this country
to bring out the best that is in the people for their own good, instead of supinely
allowing foreigners to run away with all the inventions and improvements.
Fiscal arrangements are powerless in these matters : a bolstered-up, pampered
industry makes no progress ; it does not require to progress ; nothing but energetic
educated enterprising business can affect the situation. Writings have often proved
of value in directing- movements, especially in scientific work ; to some extent this
Historical 3
work will present a re'sume' of up-to-date improvements, some already developed, others
awaiting development. All this is somewhat of a digression ; the excuse for it is that
mechanical and electrical sciences are in their practical application retarded or accelerated
largely by the enterprise of the capitalists. And it is hoped the treatment of the subject
in these volumes may help to enlighten many who would bring up their knowledge to
present date.
With the advent of the modern steam turbine the steam engine has arrived at its
meridian ; and the new Hult rotary engine may also give another finishing touch to the
great edifice of improvements on steam engines. These two departures are worthy of
the closest attention. But other agents besides steam are calling for recognition.
The internal combustion engine in its many forms has reached a stage where it will
also offer itself as a powerful competitor with steam.
Then the electrical transmission of power has brought the great water powers of
the world down from their inaccessible fastnesses in the mountains to the workshops
on the plains. It is no wild dream, but quite within the range of practical engineering,
to say that we could transmit the energy of Norwegian waterfalls by submarine cables
to Edinburgh to light the city and run its tramways. The utilisation of our own
Highland lochs and waterfalls will come before that scheme.
A brief sketch of the history of the subject may precede the description of prime
movers. It is proposed to describe the prime movers in their order first, and then to
treat details afterwards. That the ancients became acquainted with means for moving
huge masses to great heights and over long distances we have ample proof in the
pyramids and ancient ruins. All primitive machinery was at first moved by animal
and human power, using the machinery of levers, wedges, inclined planes, rollers,
ropes, and other early inventions. With these engines man early discovered that if a
weight of material could be moved at all, however small a distance in a day, he could
move it any distance given plenty of time ; and that, no doubt, is the explanation of
the vast works carried out by such simple engines. Time was of no value ; a few
generations of men spent on a great work was taken then as a matter of course.
Wind power was no doubt first observed on boats upon the waters. The wind
would move a boat without a sail, and this would naturally suggest a sail to increase
the effect, and from a sail on a boat to a sail on a windmill of the old type is an easy
transition. Water power was very early used to assist man in his efforts to move
things. The Chinese records prove that water power wheels were invented by them
long before the Christian era.
In ancient times intellectual contemplation and philosophical speculations were
considered dignified and worthy of the most active and learned minds, while practical
inventions and arts were considered as unworthy of notice in the ancient biographies
and histories, and during the dark Middle Ages practical engineering and science were
strangled by superstition and false religion. Galileo had to deny the motion of the
earth round the sun to save his life. Owing to these prejudicial effects we cannot
trace the earliest beginnings of prime movers ; even the steam engine may have been
invented long before any record we can trace.
The supposed higher dignity of mental contemplation and philosophical speculation
still exists to some extent in ancient universities, but from the time of the sixteenth
century the progress of mechanical invention, scientific discovery, and application to the
work of the world has been very rapid and wonderful.
Heat engines are first found mentioned in Hero's Pneumatics, written 130 B.C., in which
he describes a steam turbine of quite a practicable design, and also a steam pump for
water raising. Another work, published in 1601, on pneumatics, by Battista della
Porta, describes a steam water pump much the same as Hero's, but working with
a vacuum to raise the water into a receiver, from which it was expelled by steam.
In 1629 Branca described and operated a steam turbine in which a steam jet drove
Modern Engines
a wheel with vanes. He managed to explode his steam boiler, and was shut up as
a prisoner on the plea that he must be mad. In England the Marquis of Worcester
set up and worked a steam pump at Vauxhall in 1656. In 1697 Savery improved
Worcester's pump, and it was introduced and much used for mine drainage. This
seventeenth century, therefore, proved one in which the steam engine inventors were
allowed to make known and use their engines ; but so far these engines were steam
turbines or water pumps, in which the steam pressure acted directly on the water and
forced it up.
The next step in the evolution of the heat engine was one which was bound to come
as better mechanical construction advanced. As soon as mechanics became capable of
forming cylinders, rods, plates, and other pieces of metal with some accuracy of shape
and form, it became possible to advance the heat engine. In 1690 Denis Papin had
succeeded in constructing a cylinder and piston, the prototype of all our powerful steam
engines up to within ten years ago.
He made a cylinder and fitted a piston into it, and in the lower part of the cylinder
he placed some water ; on placing the cylinder on a fire the steam produced forced up
the piston, on swinging the cylinder from off the fire the steam condensed and the piston
was forced down again. Here we have the elements of the reciprocating steam engine.
And Papin, true to the inventor's instincts, enthusiastically describes his dreams of
steam pumps, and, by rack and pinion, turning wheels and driving paddles on ships, all
of which were quite feasible in his mind, but far beyond the skill of the mechanics of his
day to put into practice.
It is to this day even a remarkable fact that inventors are sometimes ahead of the
practical mechanic. Many an inventor conceives a valuable improvement, works it out
in his own mind, and finds that the resources of the constructors and mechanics are not
sufficient to realise his schemes.
All the early steam turbine inventors were baffled in their attempts, simply for want
of tools and machinery sufficiently good to make their turbines. The recent success of
the steam turbine has been more due to the refinement of modern machine tools and
accuracy of metal-working machinery than to any new principles or inventions.
Nowadays the inventor has many advantages denied to the old pioneers, machine
tools are now made to produce almost anything in metal with the utmost precision,
however complicated. A piston and cylinder are now fitted to the y^Vry f an inch easily,
while James Watt was delighted when he succeeded in getting a fit so close that he
could not slip a half-crown between the piston and cylinder.
Papin invented also a safety valve for his boiler, and so escaped the misfortunes of
poor Branca.
Then followed Savery, Newcomen, and Cawley, who combined the separate boiler,
cylinder, and piston with condensing water injected into the cylinder. Potter added the
self-acting valves ; Leupold and Smeaton added improvements, so that at the end of
the seventeenth century the steam engine had become a well-known useful machine.
Although frightfully inefficient, it had plenty of power, and did good work.
James Watt took up the question in 1759, and ten years after filed the most
important patent specification the world ever saw, describing his improvements in steam
engines. And after all, it was to Scotland and a Scotchman we are indebted for the
final solution of the steam reciprocating engine problems ; aided, however, by Matthew
Boulton's financial assistance and personal energy, without which it must be admitted
that James Watt's great genius would have been lost. Dr. Roebuck, of Carron Iron
Works, had helped Watt at the outset, and enabled him to make a start. Boulton was
a discoverer, and his greatest discoveries were James Watt, and later on Murdoch,
the inventor of coal gas lighting. Matthew Boulton and James Watt carried out their
great engineering work in Birmingham, and they had a long and bitter fight in the law
courts to maintain their rights to Watt's inventions.
Historical
The chief claims of Watt's patent of 1769 are worth recording 1 here. In Watt's own
words
" First, That vessel in which the powers of steam are to be employed to work the
engines, which is called the cylinder in common fire engines, and which I call the steam
vessel, must, during- the whole time the engine is at work, be kept as hot as the steam
that enters it ; first, by enclosing it in a case of wood, or any other materials that
transmit heat slowly ; secondly, by surrounding it with steam or other heated bodies ;
and thirdly, by suffering neither water nor any other substance colder than the steam to
enter or touch it during that time.
"Secondly, In engines that are to be worked wholly or partially by condensation
of steam, the steam is to be condensed in vessels distinct from the steam vessels or
cylinders, although occasionally communicating with them ; these vessels I call con-
densers ; and, whilst the engines are working, these condensers ought at least to be
kept as cold as the air in the neighbourhood of the engines, by application of water, or
other cold bodies.
"Thirdly, Whatever air or other elastic vapour is not condensed by the cold of the
condenser, and may impede the working of the engine, is to be drawn out of the steam
vessels or condensers by means of pumps wrought by the engines themselves, or
otherwise.
" Fourthly, I intend, in many cases, to employ the expansive force of steam to press
on the pistons, or whatever may be used instead of them, in the same manner in which
the pressure of the atmosphere is now employed in common fire engines. In cases
where cold water cannot be had in plenty, the engines may be wrought by this force of
steam only, by discharging the steam into the air after it has done its office.
" Lastly, Instead of using water to render the pistons and other parts of the engines
air and steam tight, I employ oils, wax, resinous bodies, fat of animals, quicksilver, and
other metals in their fluid state."
Watt afterwards proceeded to detail inventions, the pendulum engine governor,
the engine power indicator, the parallel motion, the expansive working by cut-off at a
fraction of the piston stroke, the double acting engine, the butterfly valve, and other
inventions, brought the engine to great perfection. It may seem strange that such a
simple thing as a crank for converting reciprocating to rotary motion was unknown
in Watt's time, yet it seems it was not known then, for it was invented by Watt, but
pirated and patented by another.
When one considers that only a little more than one hundred years ago a crank and
connecting rod and fly-wheel were new things, the extraordinary progress made between
1785 and 1885 can be realised.
It is of little interest to follow the history of the steam engine after Watt's time ; all
subsequent improvements have been mere details, and more due to taking advantage
of improved materials and machine tools for better construction, using higher pressure,
greater expansion, and multiple cylinders.
The reciprocating engine has reached its zenith in our day, and will likely hold the
first place in many uses for many years to come, but the march of improvement never
ceases. The steam turbine, so long impossible for want of means to properly construct it,
has at last been evolved, and is now beginning to compete successfully with the Watt
engine, which it is destined to supersede altogether in the steam-engine world. The
rotary steam engine has been the dream of every engineer ; even Watt spent much time
upon it, but it has hitherto baffled all inventors. Recently, however, a new rotary engine
has been constructed which in an exceedingly beautiful and simple manner overcomes the
chief difficulty, that is, the internal friction of the revolving piston, so that hopes are
raised also in this direction.
Steam, however, has its limitations, now well defined by scientists, and the point
seems almost to be reached beyond which nature puts a stop to further improvement.
Modern Engines
This is not to be wondered at, for during- one hundred and fifty years the steam engine
and its details have been the subject of tens of thousands of improvements and patents,
and the research of physicists of the highest ability, mathematicians, mechanicians, and
experts on steam.
Another type of heat engine was early proposed, and has been the subject of much
study, namely, the "internal combustion engine." This title includes all that large class
of engines in which the working fluid is hot air, heated in the cylinder by combustion
with gas, or gasified oils, or fuel dust.
The fuel being burned in the cylinder direct, the efficiency should be higher than that
of the steam engine, and it is so ; but the high efficiency which might be expected has
not yet been reached, a good deal of room for improvement still exists in these hot air
internal combustion engines. High piston speed should be aimed at, and proper balanc-
ing and cushioning- of the reciprocating parts should receive more attention.
The simple external combustion hot air engine has a small field of usefulness, but is
of no account for large powers.
The designs of engines, steam, gas, oil, and hot air, are very numerous ; adopted to
many various purposes, we shall only study the modern types, leaving out the old beam,
oscillating, grasshopper, side-lever, steeple, and other obsolete types, most of these
designs being made to suit the tools of the builder. Thus if an engine maker had no
large planing machine for iron slides, he designed the engines to be made without planed
slides. Turning and boring tools preceded shaping and planing tools, hence old
engines were designed with an eye to do all the machined work for them in the lathe,
and all the motion gear in the forge or smithy.
The other source of energy, electricity, has become of great importance as a
secondary power. Electricity cannot yet be liberated from fuel, like heat, by simple
combustion. The Voltaic battery is the only available source of electricity direct from
fuel, and as it uses expensive fuel with much waste it is not of commercial importance for
power purposes. But by converting the power of a heat engine into electrical energy
we obtain the most economical and ready means for transmitting energy for power and
light. For this purpose it is only necessary to attach a dynamo-electric machine to be
driven by the heat engine.
As a secondary power, electricity has a great field of usefulness. It also converts
the heat engine power into a force which can be applied to chemical and metallurgical
purposes in producing aluminium and refining copper. The production of caustic soda,
potass, and carbide of lime, nitric acid, carborundum, potass chloride, and many other
processes are capable of being carried out electrically.
As a secondary engine, the compressed air engine is also of much use, especially for
drilling, riveting, and caulking in boiler and shipbuilding work. Also in rock-drilling
in mines we shall have occasion to consider compressed air engines very fully.
In the case of the internal combustion engines, we have in the oil engine a complete
power generator, in itself requiring no separate gas or other generator ; hence its
extremely suitable application to motor cars.
But gas engines require gas generators. These are of much interest to engineers, as
offering a means of producing gas for power purposes from cheap fuel by a simple and
cheap process.
The steam boiler is a necessary part of the steam plant, and has received almost as
much attention as the engine ; in its many forms adapted to many different purposes
we have a large subject to study of very great interest. The fuels available for
producing heat are not many, coal, wood, oil, natural gas, waste products such as
straw, cane, sawdust, and dust destructors. To coal the British Isles are indebted for
their huge manufacturing industry ; it has been cheap and plentiful, giving about
14,000 B. th. units per Ib. burned. When coal becomes scarce, or is at last all used
up in these islands, a vast change in the industrial world must have occurred. Coal is
Introductory
being- used up at an enormous rate, which goes on increasing- yearly. The only hope
is that engineers will soon discover an engine in which coal can be more economically
used ; the best heat engines are at present very wasteful of fuel, the steam engine
wasting about 80 per cent, of its fuel at this present day. What a field for invention
and improvement this fact discloses !
In many manufacturing processes fuel is also recklessly wasted to a large extent.
In the iron-blast furnaces coal consumption has been gradually reduced, and the
utilisation of the furnace gases for power purposes promises a further saving. Coke is
used for many purposes, and, as at present made, all the gaseous and liquid products
are wasted in its manufacture ; the blazing coke heaps and coke ovens are a terrible
waste of fuel. The prodigal abuse of Nature's greatest store of energy on earth is
pitiful. Coal can never be restored in the ages of man, and a coalless country can never
be anything but a thinly populated agricultural and pastoral land. The saving of coal
is one of the engineer's chiefest problems. Mineral oil has been found plentiful in some
localities in Russia and America, and can be made from Scottish shale. As a fuel, it is
far inferior to coal, but it is convenient for transhipment. Whether the oil supplies will
last for long or short times we cannot know. Texas seems to have a large natural
supply of crude petroleum, which, being of little use for anything else, is sold cheap
for power purposes. In a suitable engine it gives i horse-power-hour for one
farthing.
In colonial countries boilers are in many cases fired by straw and wood, the furnaces
being constructed specially for the purpose, as the fuel is very bulky compared with
its output of heat. At timber works, saw mills, etc. the waste wood and sawdust is
used for fuel to work the engines.
The refuse destructor in large towns and cities produces heat sufficient for raising
a large quantity of steam, which in some cases is used for electric lighting purposes.
In tropical countries direct sun heat has been used for raising steam for small
engines, the boiler being placed in the focus of a large mirror.
All these fuels and their uses are well worthy of the engineer's attention ; and the
accumulated knowledge of their values, and the best uses to make of them, is to be
gathered in this work in later chapters.
Condensers, economisers, pumps, valves, and all the many accessories of modern
prime movers we will discuss in their proper places.
CHAPTER II
PRIME MOVERS, CLASSIFICATION OF
IN treating 1 this extensive subject there are two methods of procedure. First, to begin
with the sources of energy fuel, heat, electricity, and the elements of mechanism,
thermodynamics, electrodynamics, fluid pressures ; and thence to engine designs and
constructions. This course, however suitable as an educational curriculum, is not of so
much practical value as the second method ; in this we follow more closely the natural
course of development of the prime mover. Engines were successfully constructed and
used without a knowledge of the exact sciences connected with their theory of action,
this knowledge forerunning the theories. As an analogy, we may refer to the science
of physiography. A surveyor could possibly by scientific measurements and observa-
tions and exact calculations delineate on a map the approximate course of a river
from his knowledge of the watershed ; while an ordinary traveller, beginning at the
mouth of the river, could by simple observations obtain a far more accurate and exact
knowledge by following the course from mouth to source. In the same way we shall
begin where the pioneers in engine making started, namely, where they commenced
to make engines without inquiring closely into the more abstruse problems. These
deeper scientific questions follow naturally as the subject develops. The fluid pressure
engines are primarily engines the working substance in which are fluids.
CLASSES OF PRIME MOVERS
Fluids are of two kinds liquids and gases.
i. Natural Fluid Pressure Engines
Natural fluid pressures are used in windmills air pressure ; and in water engines,
wheels, and turbines liquid pressure.
2. Steam Engines
Artificial fluid pressure engines are those in which the pressure is produced by
heat from fuel combustion, and comprise steam and hot air engines. Steam engines
are of three classes :
(1) Steam turbines.
(2) Reciprocating piston engines.
(3) Rotary piston engines.
8
Classification
3. Hot Air Engines
Hot air engines are of two kinds :
(1) External combustion engines, in which the fuel is burned outside the
engine.
(2) Internal combustion engines, in which the fuel is burned inside the engine ;
to this class belongs the gas and oil engines.
4. Electric Engines
In this class the energy is applied by electric pressure generated from fuel by
chemical combination, which may be properly described as combustion without heat,
electric pressure being the result, instead of high temperature.
5. Secondary Engines
These are engines operated by fluid pressure or electric pressure, generated by
some prime mover of the types belonging to Classes i and 2, such as compressed
air, turbines driven by water which has previously been pumped up to a height by
a windmill, water turbine, steam or gas engine, and electromotors driven from dynamos.
WINDMILLS
The windmill and water wheel naturally preceded the heat engine in practice. We
therefore begin with these engines, and will follow in the order here given, with the
consideration of the construction of the others.
Naturally, wind power would be the earliest, and windmills the first prime movers.
The force of the wind is self-evident, and the first attempts to utilise it were no doubt
in the direction of propelling boats by sails. At the present day sails and masts are
still used on large vessels for propulsion. The sailing vessel can be moved in opposite
directions by the same wind, and can sail on a course at an angle in part against the
wind. This feat would at first be found by experiment ; but it can be demonstrated
by the principle of the resolution of forces as given in text-books on Mechanics.
The old windmills so conspicuous on the landscape were mechanically simply four
masts, with one sail on each, set on a rotating shaft. In a lo-mile breeze a 15-foot
wheel gave about i horse-power, and the power given was roughly about proportional
to the square of the diameter from tip to tip of the sails, and the best results obtained
when the speed in feet of the sail tips was about 2.5 times the speed of the wind.
The old windmill did good work in its day, and its more scientific successor is
capable of still better performances. Far from regarding the windmill as a sailing
body like a boat, the modern engineer now considers it in its true light as a fluid
pressure turbine. Like water power, the best sites for the wheel are generally far
away from the points where the power is required, so that some system for transmitting
the power must be adopted to obtain the best results. And again, as with all natural
forces beyond the control of man, some system of storage must be employed if the
force is to be ready for use daily at command.
In electricity we have both requirements provided a means of storage and a
means for transmission so that we can store up the power when the wind blows ;
and in designing the pjant, place the wheel on the best available site, even miles from
where the power is wanted, and then transmit the power by electric current from the
wheel to the work to be done. This electrical transmission puts both wind and water
power on quite a different footing from that on which they stood twenty years ago.
At that time steam had carried all before it, but since then it has been found that
10
Modern Engines
steam does not accomplish everything" desired. It has its limitations, which are now
clearly defined ; and more attention is paid to other working- fluids for prime movers.
Now, when we can choose the highest and most suitable site for a windmill, and
transmit the power any reasonable distance for use, it becomes a more generally
applicable prime mover.
Windmills are, however, very limited in power. A pleasant breeze flows about
10 miles per hour ; and as that can be counted upon for about 8 hours a day on an
average, it is taken as a minimum standard. The power, however, increases nearly as
the square of the speed of the wind, so that the power varies largely.
In practice it is found that well-made modern mills and water pumps will do the
work as under.
TABLE I.
Diameter of Sail
in Feet.
Water raised in Gallons
100 Feet high per
Hour with lo-mile Wind.
Horse-power in
Work done.
Wind at 20 Miles
per Hour
Horse-power.
7
80
*Vth
Aths
10
200
j> 5 th
$ths
12
35
1th
^ths
16
700
Jrd
ifrd
20
1 200
fibs
3
The above figures are for mills driving water pumps direct without gearing.
Geared mills work the pumps slower, and are generally used where power is taken
direct from the mill.
TABLE II.
Diameter of Sail
in Feet.
Water raised TOO
Feet per Hour
jo-mile Wind.
Horse-power
approximate with
2O-mile Wind.
25
30
35
40
Gallons.
2000
3000
4500
6060
4
5i
74
12
(Actual Brake
j Horse-power.
The power should be as the square of the diameter of the wheel and as the square
of the velocity of the wind. The horse-power in water raised in feet per hour equals
gallons x i o x f eetjieigjit^ multiplying> ga ii O ns by 10 to get Ibs. Thus for a 2 5 -foot mill
60x33,000
2000 x io x IPO _ i horse-power ; and as the power is as the square of the velocity of
60 x 33,000
wind, we would eet for a 2O-mile wind =4 horse-power at that velocity.
io' 2
USEFUL FORMULA
The force of wind increases as the square of its velocity.
a = area exposed at right angles to the wind in square feet ; F = force of the wind
in Ibs.; H = horse-power ; and v = velocity of the plane a in direction of the wind,
+ when it moves opposite, and when it moves with the wind.
F = 0.0022880 V 2 , when v = o
F = o.oo2288*(V v y
H
.
240384.6
Wind Power 1 1
Example. A rail train running E.N.E. 25 miles per hour exposes a surface of
1000 square feet to a pleasant brisk gale N.E. by N. Required the resistance to the
train in the direction, it moves and the horse-power lost.
E.N.E. -N.E. by N. = 3 points = 33 45'; V = 14 feet per second, a brisk gale;
v= 25 x 1.467 = 36.6 feet per second; and F = o.oo2288 sin. 33 45' x 1000 (i4 + cos.
33 45' x 36.6) 2 = 305. i Ibs.
TT ^O^.I X 76.6
H = ^-^ a = 20 horses.
550
The motions and effects of gases by the force of gravity are analogous to those of
liquids.
The altitude or head of the atmosphere at uniform density will be the altitude of a
column of water 33.95 feet divided by the specific gravity of the air, 0.0012046, or
33 ' 95 =28, 183 feet.
0.0012046
The velocity due to this head will be V = 8.O2 v/28, 183 = 1346.4 feet per second, the
velocity with which the air will pass into a vacuum.
VELOCITY OF WIND
When air passes into an air of less density, the velocity of its passage is measured
by the difference of their density.
H and h = density of the air in inches of mercury; /= temperature at the time of
passage ; and V = velocity of the wind in feet per second.
/j_j[ _ fa
V= 1346.4* / - (i +0.00208^).
As the wind averages between 10 and 20 miles per hour, we might count on an
average between i and 4 horse-power as the output of a mill of 25 feet diameter on a
jo-foot tower placed in a favourable situation. These observations refer to the ordinary
modern mill as shown in Fig. i ungeared, and in Fig. 2 geared, in which the sails are
moving in a plane at right angles to the wind, and are set at an angle such that the
tips of the sails move about 2.5 times the velocity of the wind. Thus the 25-foot wheel
would in a 2O-mile wind make 24 revolutions per minute, not an excessive speed.
This type of wheel has the advantage over others of presenting every blade constantly
to the cylinder of wind which it intercepts, and runs at a fairly high speed. The whole
of the blade is not equally effective, however, as its effect decreases towards the centre,
where the wind has less leverage, and in high gales it is apt to be damaged, so that
attempts have been made to design other forms, which we will also describe.
In this present form the blades are flat metal sheets, geared so that they can
be feathered in order to present them at different angles to the wind, in some cases
to turn them edge on in high gales so as to reduce the pressure. This geared mill
is shown in Fig. 2. The little windmill at the back winds the large wheel round,
so that it always automatically presents the same face to the wind. The blades also
automatically adjust themselves in heavy winds so as to prevent excessive speeds.
These automatic gears are necessary on all large mill wheels of this type. The small
wheel is known as the fantail. In some old mills it was little else than a tail, a flat
board on a tail rod on the axis of the wheel.
Referring to Fig. i, which illustrates the smaller mills ungeared for pumping water,
made by Messrs. Robert Warner & Co., the mills are provided with fixed galvanised
vanes, fixed to circular rings and arranged to turn partly out of the wind when the
velocity ot the latter is higher than is required for the proper working of the mill.
Fantails are provided to each mill for maintaining the sail to the wind. Suitable
gearing is arranged for conveniently putting the mill into or out of work from the
12
Modern Engines
ground level, so that the attendant need not go to the top except for the purposes of
lubricating, which is very seldom required. The towers are made square (excepting for
the larger sizes, which would be hexagon), having four corner posts of angle section,
with suitable strong strengthening rings and bracings. These are made mainly of
steel, wrought iron being used for bolts and other smaller parts. The towers are
FIG. i Simple Ungeared Windmill.
FIG. 2 Geared Windmill.
arranged for bolting down to four blocks, which may be of concrete, brickwork, or other
suitable materials that may be available.
The sail and pump shafts are of best steel, the bearings, including the turntable
and millhead, being very carefully designed so as to give the smallest amount of friction.
Special attention is given to the parts requiring lubrication, so as to need the least
possible attention.
Windmills
The geared mills are constructed of larger sizes, from 20 to 40 feet in diameter. The
Figure 2 is from a photograph of a 4o-foot mill supplied to the Government for Egypt.
These mills are substantially built with very large wearing surfaces, which are
well provided for means of lubrication, thus rendering them very durable. They are
automatic in their action during heavy winds, so that excessive speeds which might
cause damage are avoided. The winding gear consists of a geared back fantail which
is far the most efficient and satisfactory type to adopt with geared windmills. They
can also be quickly started and stopped by hand from ground level. The power is
transmitted from the sail shaft to the ground or any other level of the tower by a pair of
bevel wheels and upright shaft. The power can be taken from the latter or from a
horizontal as may be most convenient. The towers are made mainly of steel, wrought
iron being used for bolts and sqme of, the smaller parts. The smaller towers are
square and the larger ones hexagon, all hav-
ing suitable strengthening rings and brac-
ings, and are arranged for being bolted
down to suitable foundations, such as con-
crete, brickwork, etc.
The first cost of windmills is not high.
A 2O-foot mill ungeared with a 3o-foot iron tower costs
about 140, giving from about i to 3 horse-power. A
4O-foot mill giving from 6 to 12 horse-power, complete
with a 4O-foot tower, costs about ^470. When we
consider the simplicity of the machine, its small cost for
upkeep, and no cost whatever for fuel, these prices are
not excessive.
These prices, however, do not include the cost of the
power transmitting or utilising gear ; say, for instance,
the pump, if it is for water raising, is not included. We
will return to this point after we have discussed the other
two forms of wind wheels.
A wind wheel which rotates with its axis vertical
and the wheel horizontal was designed by Robinson to
register the average speed of the wind over a given
period. This form of wheel depends for its action
upon the difference of resistance offered to the wind
by the concave and convex sides of a hollow hemi-
sphere. The convex sides offer much less resistance,
hence the wheel revolves in the direction pointed by the
convex side. It is thus evident that the whole pressure of the wind is not available,
as the back pressure on the convex side must be deducted from the forward pres-
sure on the concave side ; and the pressures act only for about one quarter of a
revolution.
On the other hand, the speed can never exceed that of the wind, and the arms
may be made long and strong. A considerable leverage can be given to the acting cup.
The Robinson anemometer is shown in Fig. 3, as made by Casella.
On this same principle the wind wheel of Professor Blyth is designed. It is
shown in Fig. 4, the hemispherical cups being replaced by wooden buckets of
large size ; the construction explains itself pretty clearly. The difficulty with these
wheels is their great size required for comparatively small powers. A better design
of the same type has been made in the form of a water wheel, a portion of which
is shown in Fig. 5, giving the delineation of the vanes; and in Fig. 6, where it is
seen mounted on a vertical axis on a tower. Here again the torque is due to the
difference of resistance between the concave and convex sides of the vanes ; only, the
FIG. 3. Robinson Anemometer.
Modern Engines
one partially screens the other, so that the difference is greater than it is with cups
Neither of these types are as efficient as these shown in Fig's, i and 2.
FIG. 4. Blylh's Windmill.
tf
FIG. 5. Diagram of Wind Wheel Vanes.
FIG. 6. Horizontal Wind Wheels,
Transmission of Wind Power 15
POWER TRANSMISSION GEAR FOR WIND WHEELS
The gear from the wheel shaft itself consists of a vertical shaft and bevel wheels,
a pump and crank to work the plunger where simple water raising is required. As
a rule, however, the water lies low, in a well or pond or river, while the best place for
the wheel is high up. And as we cannot bring the water up, we must take the wheel
down to the water, and thus it becomes less effective. This difficulty might be over-
come by transmitting the power by some means from the high wind wheel down to
the low water pump. Ropes and pulleys and compressed air have been used for this
purpose, but they are expensive and inefficient, and the source of much trouble ; of the
two, compressed air is the best. In this system the wind wheel pumps air under
pressure into a receiver, generally a cylindrical steel boiler. From this an underground
iron pipe connects down to the pump placed at the water source, the compressed air
here, by means of an air motor and water pipe combined (to be described under pumps),
pumps the water up to a reservoir to the desired level, from whence it is supplied by
gravitation to the users. For pumping purposes this is a satisfactory system, for
both wind wheel and water pump can be placed just where they are most advantageous,
and even more than one pump may be used from same wind wheel if several sources of
water can be tapped.
For power purposes the same arrangement can also be used with advantage, and
by providing large storage vessels for the air under pressure a fairly constant supply
of air can be relied upon for daily work. This system for transmitting and utilising wind
power possesses much to recommend it in Egypt, Westralia, and other places where fuel
and water are scarce. The other system is the electric transmission system, in which the
wind wheel drives a dynamo-electric generator, from which wires are run to the motor
or motors where the power is required.
Like compressed air, it requires for its reliable working a storage reservoir, in
the shape of an electric storage battery. And as the speed of the wheel varies with
the power and velocity of the wind, a special battery must be used capable of charging
between very great differences of electric pressures, so as to make use of all winds
blowing between 10 and 30 miles an hour.
The problem to solve in the storage of the energy of the wind wheel is much the
same for compressed air and electricity. In the case of compressed air the pump
must work between great differences of speed ; and as the torque required to move
its pistons, rods, etc. must be sufficient to work it at a lo-mile breeze speed, it is
evident this torque being constant at all speeds the speed of the wheel will go up
nearly as the square of the wind velocity, and the increased work of the pump will
be due to and proportional to the increased speed and not to increased torque ; the
variation in speed would be very large. A variable stroke compression pump would
greatly reduce this large variation, but that is a difficulty in itself. The next best
solution of the problem is to use a compound air pump with three or four cylinders
and pistons, in which one, two, three, or four pumps may be worked from the wind
wheel, one on the lightest breeze, 10 miles, two on the 15-mile speed, three on the
20-mile, and four on the 25-mile speed, each pump being also larger in capacity than
the preceding one. By this means the speed may be very much moderated, for the
torque required increases as pumps are put on, so that the work done is proportional
to the wind pressure or torque.
The electrical dynamo generator for storing electricity into the battery acts to
some extent (on a wind wheel) similarly to the compound pump. If the speed of
the dynamo at the lowest speed of the wind wheel is such as to generate an electric
pressure slightly greater than the battery pressure, then as the speed increases the
torque required also increases, for the pressure of the generator and the current
also increases into the battery. Hence the speed of the wind wheel is opposed by
1 6 Modern Engines
this increasing: drag put on by the generator, and the variation in speed much
moderated. For the purpose the dynamo should be compound wound and with
field magnets far below saturation-point at the small load. The battery must be
calculated to take the charging current at the maximum speed without excessive
heating.
The electrical points will be elucidated under their own headings.
If compressed air is employed for transmitting the power, then some further con-
siderations must be taken into account. When air or any gas is compressed by
force, sensible heat appears; the gas is raised in temperature during the compression.
The question as to whence cometh the heat is an important one, which is discussed
under the science of heat. In the meantime, suffice it to say that if the air is carried
away in pipes or stored in a reservoir this heat is dissipated and lost by conduction
and radiation, and the condensed gas in losing the heat falls in pressure to some
extent. And further, we shall find that if the gas is afterwards expanded doing
work in an engine, we must make good this loss of heat if we are to get efficiency
out of the use of the compressed air ; for the gas on expanding must get back the
heat produced when it was compressed, and if this amount of heat is not supplied to
it while expanding from some extraneous source, it takes the heat from itself and the
containing pipes and cylinders and falls in temperature, so much that in air motors
of any great power the passages, valves, and cylinders
become filled with ice and snow, produced from the
moisture in the air freezing from the extraction of the
heat by the expanding air.
If compressed air is therefore allowed to cool down
before use in an engine, then to get good efficiency we
must provide some source of heat to "reheat" it before
it enters the engine.
In fact, this absorption of heat by previously com-
pressed gas which has been cooled under compression is
IG ' 7 '~ the fundamental principle upon which some ice making
and refrigeration by machinery is based. Air is compressed in these machines, then cooled
as much as possible, then expanded, when it thereupon falls below zero temperature.
This freezing air is then led through vessels, in which it abstracts the heat and produces
ice.
In the air transmission of power not only is the ice and snow a nuisance in the
engine, but it reduces the efficiency, for the mean pressure is less as the temperature
falls. Reheating is therefore an important engineering question in compressed air
power, which cannot be fully gone into in this chapter, but we may refer to a simple
means for supplying the necessary heat suitable for simple transmission under the
considerations of wind power.
Into the supply pipe next to the engine a steel box is inserted large enough to hold
an oil lamp with several large wicks. The box has a movable cover, easily secured,
air tight, and capable of withstanding the maximum air pressure ; under these conditions
all the air entering the engine must pass over the lamp wicks. To start the engine the
lamp is ignited, the cover shut down, and the air turned on ; the engine starts, and
as the air passes over the lamp it maintains the combustion, and at the same time
heats the air entering the engine. This is an example of an internal combustion
heater in which the products of combustion also pass into the engine. There is no
chimney loss, hence it is very efficient.
The efficiency of the air motor is much improved, and no troubles from ice and snow
occur. Diagram Fig. 7 shows the apparatus.
Wherever wind power is seriously considered it is worth while weighing the advan-
tages of the two systems of storage, compressed air and electric. In out of the way
Windmills
places compressed air seems the simpler system to maintain, as it is wholly mechanical
in operation and of common mechanism in construction ; while the electric system is
the more flexible in application, overhead wires on poles carrying the power wherever
required.
CONSTRUCTION OF WIND WHEELS
The old wheels had only four or six sails carried on poles like masts. The
sails were canvas spread on a brander of wood, with cords and pulleys whereby
the sails could be "reefed" to reduce the pressure in a heavy gale. Wood was
the material used throughout. They are to be seen yet in the eastern counties of
England, and several good examples are still working in the Orkney Islands grind-
ing grain.
The modern wheels herein illustrated are built of mild steel, strong, compact, and of
light weight, and the principles of design are the same as those of a water turbine, for
they are fluid pressure turbines of a simple design.
Properly designed, the blades of the wheel should be made of
sheet steel, and curved so as to gradually change the direction of
the air stream striking them. For common practical work, and
owing to the facility with which they can be feathered, flat blades
are used, as shown in Figs, i and 2. Curved blades cannot be
turned edge on to the wind, so that in a heavy gale they would
be damaged ; but in any wind wheel of importance arrangements
can be made whereby the whole wheel can be turned edge on to
the blast.
The blades when curved should be shaped like the blades of
a parallel flow water wheel, as shown in Fig. 5. The pump to be
employed for compression of air in cases where that system of
power storage is adopted must be simple and effective. The whole
question of compressed air power and apparatus will be fully
treated later ; meanwhile, it will suffice to show a very simple
form of water pump suitable for the windmill (Fig. 8). The pump
and the driving shaft are fixed conveniently to a cast-iron hollow
column, which forms a receiver of small capacity. It is a simple
trunk piston pump with large valves. For large storage of air, wrought-iron drums
are of greatest capacity, with a minimum weight ; those should be made of a size
easily transported, and a number connected together by a pipe to obtain the desired
capacity.
For electric storage the dynamo chosen should be a series wound machine, feeding
into a battery of large capacity, and the terminal electric power delivery at the battery
should be at least 2 volts less than the dynamo terminal power delivery when the mill is
running on the lowest speed it may be designed for, and an automatic cut-out must be in
circuit in order to open the circuit the moment the battery and dynamo power deliveries
are equal. By this arrangement, as the speed increases with the speed of the wind,
the torque will increase also.
In charging, the power delivery of each cell may be 2.5 volts, and may be 2.7 volts ;
in fact, the charging volts may range from 2.2 to 2.8 volts. At 2.2 volts the current
entering the cells will be very small, while at 2.8 volts it may be very large, the
difference of 0.6 volt giving a large current through the cell's very small resistance.
The cell may have a resistance of o.oi ohm, so that the current would be j^ = 6
amperes.
Any series wound dynamo will not do for this work. Common series dynamos are,
VOL. i. 2
FIG. 8. Windmill
Pump.
1 8 Modern Engines
as a rule, made with large armatures and small fields, so that they are nearly constant
current generators. A dynamo for a windmill must have, on the contrary, a large field
and small armature, so that the field strength goes on increasing proportionally with
increased speed. Such machines are not to be found in stock, but must be specially
designed for the wind wheel to be used with it.
Attempts have been made to combine two machines to be coupled in series parallel
for different speeds on a windmill to be carried by ships, but the simple series dynamo
specially designed is much better and far simpler.
The electric system of storage is the best for wind power, and is well worthy of the
consideration of engineers ; while the compressed air system offers advantages in colonial
and outlying work where the electric battery and motors might not get the required
skilled attention.
In the spring of this year the Royal Agricultural Society of Great Britain took
up the question of wind engines. The following particulars and illustrations are from
their report thereon :
Having regard to the increasing demand for an inexpensive pumping plant, and
the very material alteration in the design of the modern wind engine, the society decided
to institute trials of wind pumping engines in competition for prizes, so that trustworthy
data might be provided for the guidance of intending users. No independent competitive
trials of these engines have previously been carried out in this country.
As the use to which these engines are principally applied is that of pumping, and
as with an engine running for several hours throughout the week a very considerable
amount of water may be pumped with comparatively small power, the limit of 4 horse-
power was decided upon as a maximum. It will be seen from the results of the trials
that this was a very ample allowance, as none of the competing mills attained such
power under the stipulated wind velocity, which was fixed at ten miles per hour. This
velocity is considerably in excess of the mean velocity at inland stations throughout the
year in this country. On the other hand, there are periods at which the engines should
be able to work efficiently at very much higher velocities. It was obviously undesirable
to fix upon a high wind velocity, as it is all important that, though the engine should
control itself and work efficiently in a high wind, it should also run in light winds, which
are generally prevalent during the driest periods of the year when there is the greatest
necessity for pumping.
With a view to ensure that the trials should be held under the most variable
conditions as to wind velocity, they were fixed to take place during March and April.
The records of the wind velocities during the period of the trials proved the wisdom of
this choice of season, as during the greater part of the time there was plenty of wind,
with just a few days of light wind, which afforded the opportunity that was desired of
comparison between the several engines under those conditions.
To gauge the power developed by each engine, it was stipulated that the amount
of water pumped against a head of 200 feet should be measured, and that such should be
taken as the standard upon which to compare the work done by each.
It is not to be assumed that 200 feet head is in any way the average head at which
such engines would ordinarily work. In most cases they would raise the water to a very
much smaller height. On the other hand, they might be required to raise water to a
greater height. This is a point which merely affects the size of the pump ; and the
particular head of 200 feet was adopted in order to keep the size of the pump and
its connections within moderate dimensions ; 1 1 gallons per minute representing a
horse-power.
Twenty - two competitors entered, and the tests came off as arranged on the
society's permanent show ground at Baling. Sixteen of the competitors for various
reasons dropped out of the contest before the final, and six entered for the final tests,
with the result shown in Table III.
Windmills
TABLE III. PERFORMANCE OF WIND ENGINES IN PUMPING WATER.
I.
II.
No. 3.
No. 7.
No. 8.
No. 14.
No. 16.
No. 17.
Goold,
Shapley, &
Muir Co.
Messrs.
Thomas
& Son.
Mr. John
W. Titt.
Messrs.
R. Warner
& Co.
Mr. John
W. Titt.
Messrs.
Henry
Sykes Ltd.
Price ....
70
77
61, 75. 6d.
^79
N?3 N?7 N8.
SMtraWR MESS" ITXV*
INUinC"! THOMASSSON TUT
= ===
N?I4. NI8. WIT.
MESS 1 ?* MTJ.W. MESS"H!
WAHNEBiC- TITT. SYKEStT"
no/fflo
84000
^ 1
= ;f= 1 =
--S &
g====
I 1 '==j
S s
ir^TOC
iZ^COO
tootooc
940C3
8 j|
: S:
== : 1= i : =
1 ^ ?
Ig I i
^: =ifF=|f =
l\ =l=ll =
1=1 = I =
toooo
ow
now
w
1- m =
1= :& =1 =
Si !' i 1
III
1 1 i 1
IN i i
! ! aM
IN i M|M
CZD
PARTICULARS OF PUMPS.
No. 3 = 4 inches by 22 inches. Double-acting.
,, 7 = 4 8 ,, Single-acting.
>, 8 = 3 i 64
14 = 3i 5 -. Double-acting.
,, 16 = 4-^ ,, 8 ., Single-acting.
,, 17 = 2^ ,, 8 ,, Double-acting.
NOTE. Nos. 8 and 16 are " Bucket and Plunger "
Pumps.
FULL BORE CAPACITY PER STROKE.
No. 3
7
16
17
3636
.183
.348
.4601
.284
gallons.
RATIO BETWEEN WHEEL AND PUMP.
No. 3 = 2^ to
16
17
FIG. 9. Diagram Illustrating the Working of six selected Wind Engines during Final
Trials. 150 working hours.
subsequently adjusted. Credit for this facility of adjustment was given in estimating the
design. It would not be fully represented in the recorded performance.
Others of the exhibits had arrangements in the head gear for adjusting the length
of stroke, as will be seen from the descriptions given above.
Tests on Windmills 2,1
One requirement was regarded as inexorable. The engine and pump were
expected to stand the strain of the two months' run with only such attention as might
be supposed to be available on the spot in ordinary conditions of working. Any break-
down that required the maker to be called in to get the machine into proper working
order again was regarded as terminating the competition so far as that machine was
concerned, on the ground of inadequate stability and durability.
The final stage of the competition dealt with the selection of two mills for the
First and Second Prizes respectively from among the selected six. The diagram
(Fig. 9) represents the results obtained by the second month's run of the six
engines. It shows the number of revolutions made by each of the engines, the
number of strokes of the pumps, and the amount of water pumped. The last item is
indicated in the diagram by the length of the full black column the third column
in each space. Above the top of the black column is the outline of an addition to
its height, intended to show the loss of water due to the slip or defective efficiency of
the pump.
In order to give further indication of the process leading up to the decision, it is
desirable to refer to the several points to which specific attention was paid, and with
regard to these to record the following notes :
i. The design of each tower was examined. Its practical stability was sub-
jected to severe experimental test during the course of the trials, under winds
which reached 45 miles per hour in one instance, a very high velocity for an inland
exposure.
Price has been taken into account in relation to the amount of water which can
be delivered for a given outlay. The price is to a considerable extent determined by the
size of the engine.
In the efficiency of delivery No. 3 stands conspicuously at the head in this respect.
No. 17 comes next, and loses not a little by the inefficiency of the pump ; then follow
No. 8, No. 7, No. 14, and No. 16, in the order named. The inefficiency of the pump
of No. 14 placed it at a considerable disadvantage in this respect.
With regard to the delivery of water, it should be remarked that, as may be
inferred from its design, No. 3 delivers water equally for practically every part of the
stroke. There is hardly any idle or slack time in the revolution of the wheel, and
consequently the engine works with a very long pump stroke, and does a great deal of
pumping for very few revolutions.
The pumps of the six engines were double acting pumps, or bucket and plunger
pumps, with the exception of that belonging to No. 7, which was single acting.
The conditions of the competition were favourable for the use of double acting
pumps.
In forming a judgment upon the cost of maintenance in respect of repairs, regard
has been had not only to the excellence of design and workmanship, but to the probable
wear and tear under what may be called "agricultural" conditions. In this respect the
number of revolutions of the wheel and the number of strokes of the pump for
a given quantity of water delivered are very important items, as in ordinary wear it
is upon these two that the life of the engine or pump depends. In both these par-
ticulars No. 3 is conspicuously successful. Then follow No. 16, No. 7, No. 14, No.
17, and No. 8. No. 7 is distinguished by the uniform thoroughness of execution of the
design.
The ease of erection, which is an important item from the agricultural point of
view, as it involves serious considerations of expense, has been estimated partly from an
inspection of the designs and partly from experience on the ground. On the whole,
No. 7 is regarded as entitled to special mention in this respect, and No. 17 presented
obvious elements of disadvantage from this point of view.
Considering Table III. and Fig. 9, we get the performance of the mills. No. 3 is
22
Modern Engines
clearly ahead of all the others in nearly every point, and easily takes the first prize.
The points worthy of special commendation are
1. Its general excellence of design, especially as regards the engine and pump.
2. Its efficiency as determined by the amount of water pumped.
3. Its successful governing.
4. The arrangement for the automatic application of the brake.
5. Its economy in upkeep, due to the slow motion of its moving parts, and its
good workmanship.
6. Its reasonable price.
For the Second Prize a similar prominence in so many points is not to be expected,
but when the qualifications under the several points enumerated are all taken into
account, No. 7 stands distinctly ahead
of the remaining four. The points for
special notice with regard to this mill
are
1. The generally satisfactory nature
of the design of the engine and of the
tower, and the thoroughness of the work-
manship in its execution.
2. The provision of an efficient brake.
3. The comparative economy of up-
keep arising from the soundness of the
design, and the comparatively slow motion
of the pump.
4. The ease of erection.
5. Its reasonable price.
There was considerable diversity in
the details of the" sails, their shape, angles,
and areas, and the trials were not con-
clusive as to the best form ; for some were
evidently good mills, but handicapped
by the conditions. However, the form
of the sails, their angles, and dimen-
sions of the prize winner are such as the
experience in turbines generally would
lead one to adopt. We will now briefly
describe the final competitors, beginning
with
FIG. 10. No. 3 Windmill.
No. 3. Messrs. Goold, Shapley, &* Muir Co. Ltd., Brantford, Ontario, Canada.
This Canadian engine, illustrated in Figs. 10 to 12, has a wheel 16 feet in diameter,
with 18 blades grouped in six sections, having an area of 131.32 square feet, with an
available clearance area of 67.93 square feet between the blades and 30.88 square feet
at centre of wheel.
The blades are fixed after the American fashion of threading them on the outer
ring and fixing them thereto by means of a stamped steel bracket riveted to the sail and
ring, the inner end of the blade being riveted to the inner ring. This method probably
gives a maximum of strength for a minimum of weight, but it is open to the objection
that should the middle blade of any one section of the wheel fail, in order to dis-
mantle and replace it it is necessary first to cut off the adjoining blade. The
wheel is mounted on a horizontal shaft running in two roller bearings of 6| inches
in length. The boss of the wheel is made very deep in order to allow of the effectual
Details of Windmills
bracing- of the arms, but in order to avoid consequent increased overhanging the
boss is recessed at the back and the roller bearing off the main shaft is carried well
forward into it.
The gear of this engine differs from all others. The object to be attained was to
get as long a stroke of pump as possible, a pinion on the main shaft gears into a mangle
rack of such a length as to give one stroke of the pump shaft to two and a half of the
wheel.
This g-ear is illustrated in Fig.
n, B, C, and D. It will be seen that
there are two vertical parallel racks,
with a connecting- semicircular rack
top and bottom, in which the pinion
alternately engages ; the rack is alter-
nately thrown over at the end of each
stroke by means of cams, thus re-
versing the direction of the travel of
the pump rod.
The rack is guided between four
steel rollers, which at either end of
the stroke engage the cams ; and to
ensure the even working of the pinion
in the rack a steel guide plate work-
ing against a flanged roller is pro-
vided.
The swing of this mangle action
is only if inch, consequently there is
very little angular motion of the pump
connecting rod. The pump rod, which
is of 3 inches square white maple, is
connected to the rack casting by
means of a piece of if-inch wrought-
iron pipe guided through the tower
head.
The r pump is a double acting-
syphon pump, 4 inches diameter by
22-inch stroke, the working barrel
being of gun-metal and the valves of
vulcanised indiarubber.
The main feature of this engine
is the method adopted of governing,
which differs entirely from that of any
of the other engines. This is shown
in Fig. n, E. On the frame of the
wind vane a lever is pivoted, one end
being controlled by a spiral spring anchored to the lower frame of the vane. To
the other end of the lever is attached a chain H, connected with the pull-in wire,
the tightening of which is done in opposition to the force of the spring and pulls the
wheel into the wind. On freeing the pull-in wire, or in the event of the breaking of
same, the spring pulls the wheel over parallel with the vane ; consequently the edge
of the wheel only is facing the wind ; at the same time the brake is put on and the
mill stops.
The tower consists of four angle steel posts, with five intermediate angle iron frames
and diagonal tie rods. At the bottom of each post are anchor plates, which are
FlG. ii. Details of No. 3. The Prize Winner.
Modern Engines
bolted to the timbers let into the ground about 5 feet deep. The general design and
workmanship of this engine leaves little to be desired.
This engine was started on Monday, March 2, and ran most satisfactorily all
through the trials. At the end of the first month it was one of the selected six, and
ultimately gained the First Prize.
No. 7. Messrs. Thomas & Son, 64 Broad Street, Worcester.
This engine is illustrated in Fig. 14. The wheel is 16 feet in diameter with
24 blades, having 141.55 square feet area, with an available clearance space be-
tween same of 55.5 square feet, and 23.04 square feet area at centre of wheel. The
blades in this wheel are separately riveted to brackets fastened to the rings of the
t<- 91-10" *4
Utfj
_:^>j_ _L _^i _' _"^s4_ "J. _^^ v JL SA 1*
i / * r ' '
j- 7-/3fe JT - - -/-/.% -j 7-7** 1
FIG. 12. No 3. Details of Sails.
wheel, so that any one blade may be easily removed and another fixed in its place. The
outer end of the blades are very materially strengthened by having two corrugations,
which are seen in the illustration. The wheel is further stiffened by six stay rods from
the outer ring of the wheel to a casting on the projection of the wheel spindle.
Immediately behind the wheel is an automatic band brake, which is actuated by a
projection on the tail vane when it is blown over by a gust of wind, or when the wheel
is pulled out of the wind by the windlass.
The wheel spindle is carried in two horizontal roller bearings of ample propor-
tions, and the thrust of the wheel is taken by ball bearings at the end of the spindle.
The tail vane itself is mounted on ball bearings. On the wheel spindle a pinion is
keyed, which gears into a wheel on the pump crankshaft, speeded so as to give one
Details of Windmills
stroke of the pump for two and a half revolutions of the wheel. The teeth of these
wheels are all tooled, being cut out of the solid. The cast-iron revolving head is
carried on ball bearings at the cap of the tower, and a cylindrical sleeve descends
through the tower head and is guided by four rollers. This gives a very efficient support
to the head. The pump rod is of i-inch wrought-iron tube, the weight of which is
counterbalanced by a vibrating lever and weight.
The tower consists of four angle posts inclined approximately i to 5, with angle iron
stiffening frames every 5 feet, and f inch diagonal bracing rods. Placing the stiffening
frames at the intervals named materially facilitates the work of erection, as the tower
forms its own scaffold as the work proceeds.
The pump is of the single acting type, fitted with gun-metal bucket and suction
valves.
The design and workmanship both in the engine and tower were most carefully
thought out and executed, and throughout the
trial the engine worked very satisfactorily, and was
ultimately awarded the Second Prize.
No. 8. Mr. John Wallis Titt, Warminster.
The wheel of this engine is 16 feet in diameter,
sail area 113.89 square feet, clearance area between
blades 74.59 square feet, and 38 square feet area at
the centre of the wheel.
The stroke of the pump can be varied. The
wheel is mounted on a crankshaft with roller
bearings. The head is built of steel angles and
plates, and revolves on roller bearings.
Another point in this mill is the provision of
a ladder revolving with the head gear, so that a
man can without danger lubricate or attend to other
small matters in connection with the gear while the
engine is running.
The wheel is controlled by a tail vane and
weighted lever. This lever is raised by the starting
wire, which passes up through the centre of the mill
and over two small guide pulleys. During the trial
the wire failed, and the absence of any brake on
this gear was very apparent. The general arrange-
ment of the head is shown in Fig. 13.
The pump is of the bucket and plunger type, 3^ inches diameter and 6^ inches
stroke, and there being no intermediate gear the efficiency of the pump was good.
At times, however, the speed of the pump is high, and unless the plunger is properly
lubricated it would tend to wear more than a slower running pump.
FIG. 13. No. 8. Mr. John W. Titt's Mill.
No. 17. Messrs. Henry Sykes Ltd. , 66 Bankside, London, S.E.
This engine presents several novel features which had evidently been carefully
thought out.
Unlike other engines having fixed blades in their wheels, in which the tendency is
rather to increase the size of the blades and to diminish their number, the opposite
course is adopted of diminishing the width and increasing the number of the sails, on the
ground that most work is done by the wind on the leading edge of the blade ; and
consequently it is well to have as many leading edges as practicable. The periphery of
26
Modern Engines
the wheel has a hoop round it, to which the end of the blades is fixed, which gives it a
very strong- and, at the same time, neat appearance. Another novel feature of the
engine is the jointed tail vane, described later.
The wheel is 16 feet diameter, with 42 blades 4 feet 6 inches long, 9 inches
FIG. 14. No. 7. Messrs. Thomas & Son's Mill. Second Prize.
wide at the periphery, and 6 inches at the inner end. The sail area is 142.82 square
feet, with 54 square feet available clearance between them, and 38.48 square feet
at the centre. The curvature of these blades is shown in Fig. 16. The wheel is
mounted on a plain single-throw crankshaft, carried in an inclined position in two
gun-metal bearings in the revolving head, the thrust of the wheel being taken by a
Details of Windmills
2.7
collar on the shaft, bearing against the side of the main gun-metal bearing. The head
revolves on roller bearings.
The tail vane is mounted on two rods attached to the revolving head, which always
tends to make the wheel face the wind, the
inclined hinged axis of vane, a fixed portion
of tail vane, and a side vane, which tends
to make the wheel face away from the wind
and only present its edge to it.
Diagram Fig. 15 shows the position
of the wheel and vanes when the engine
is running in a moderately light wind, the
wheel facing the wind.
We need not describe more of them
from the report, but may notice the head
gearing of the "Samson" Mill by Messrs.
J. S. Millar & Sons, Annan.
This "Samson" mill is very similar to
the " Ideal" mill ; the wheels are the same,
but the gearing is somewhat different, in that
two small pinions on the wheel shaft gear
into two wheels with a pin between, forming
the crank for the pump rod, taking the place
of the annular gear in the " Ideal " mill. The
governing arrangements are the same in
each. The " Samson " tower is of the same
design, but is a 4-post tower instead of FlG> I5 ._ De tails of No. 17.
tripod. The gearing is illustrated in Fig. 17. Henry Sykes' Mill.
i
FIG. 16. No. 17. Henry Sykes' Mill. Details of Sail.
Modern Engines
An exceedingly simple little windmill
(Fig. 18) may be constructed of two half
discs of sheet metal, zinc for preference,
or aluminium. The sails are at right
angles to each other, and the shaft
passes between at an equal angle to
FIG. 17. Windmill Head Gear.
FIG. 18. Simple Windmill.
each sail. It works best with the shaft
parallel with the wind. The half discs
are tied to the shaft and to each
other by steel wires, as shown. It
has been called by a fanciful name, the
Pantanemone.
HYDRAULIC MACHINES
The turbine wheel has completely superseded the water wheel. For all falls of water
the turbine can be designed to give better results.
In Switzerland and in the United States of America water powers to great extent
are used. In 1880 there were 55,000 water wheels working in North America, with
a total aggregate power of 2,000,000 horse-power. Since then this has been largely
increased by the introduction of electrical transmission, so that the horse-power is now
considerably over 2,000,000.
At Geneva water power to the extent of over 7000 horse-power is derived from the
Rhone, and 12,000 horse-power more from the same river about 5 miles farther down
the stream.
In the British Isles there are no great falls of water, but in the more hilly and
mountainous districts water powers could be developed of quite considerable amounts
in Wales, Ireland, and North Scotland. Coal, however, is cheap and plentiful, so that
the available water powers are neglected.
Water power has been estimated to cost in America about ^4, los. per horse-power
per annum, and steam power from coal at 8s. a ton costs about the same. This, however,
is for low falls ; for high falls the cost is much less, and in some falls over 100 feet it has
been as low as 2 per horse-power per annum.
The first cost of water power varies very much with the district and conditions, and
cannot be laid down on any fixed basis. The large works at Geneva cost ^30 per
horse-power, the smaller works 60 per horse-power.
With water power the bulk of the cost is due to rental of land and water rights,
Hydraulic Machinery
interest on capital and depreciation of plant, only a small fraction for working" expenses ;
with steam power about half the cost is due to wages and fuel, and the other half to
permanent charges.
A steam plant therefore costs less when idle, whereas in a water power plant costs
go on all the time full. If the water cannot be used night and day the whole cost falls
upon the hours during which it is used.
Storage of water is of great importance in all water powers, and in some cases it is
the simplest and best means for storing power to pump water up to a high reservoir ;
and in some cases hydraulic accumulator storage can be utilised where a high level
reservoir is not available. Lord Armstrong's high pressure hydraulic distribution of
power system is a good example of this method.
The hydraulic accumulator is simply a tall cylinder with a heavily loaded ram,
as shown in Fig. 19. The water is pumped in continuously in small quantity, and
maintained under the pressure of the load on the ram, the ram rising and falling as
FIG. 19. Hydraulic
Accumulator.
FIG. 20. Hydraulic Crane.
FlG. 21. Hy-
draulic Tele-
scopic Lift.
the pressure water is used up or stopped. If a is the area of the plunger or ram, and
P
P the load upon it, then - =p, the pressure per square inch on the water delivered in
Ibs. If L is the length of the cylinder full of water, then aL is the maximum storage,
and the energy stored in foot-lbs. is equal to the quantity of water stored multiplied
by the pressure per square foot.
Thus one of the accumulators in the London Hydraulic Power Company's station,
working at a pressure of 100,000 Ibs. per square foot with a 2O-inch ram, a = 2. 2 feet,
and a stroke of 23 feet, the energy stored is only
2 2 X TOO OOO X
r
= 2.5 horse-power for
-^
33,000 x 60
an hour.
But it could deliver this energy very quickly. In one minute it would work at the
rate of 150 horse- power.
It is a costly storage apparatus, but for the purposes to which it is applied its great
rate of discharge for a short period makes it worth the money. This form of hydraulic
storage is used largely for lifts and cranes, and an excellent but small system exists in
30 Modern Engines
Glasgow and London very successfully. At Birkenhead Docks and Woolwich Arsenal
hydraulic cranes are so worked. A simple hydraulic accumulator will work a number
of cranes and hoists.
In hydraulic cranes the ram carries a sheave of pulleys with chains for multiplying 1
the motion, as shown in Fig. 20. The chain is fixed at F, passes over one pulley at D
on the ram, then round fixed pulley M, and again over D, and from thence over guide
pulleys C, A, B to weight W. A chain H driven by a separate ram is used for slewing
the jib.
The hydraulic lift as now most commonly made has a telescopic ram, one being
fitted inside of the other as shown in Fig. 21, so that when closed down only a shallow
well beneath is necessary to accommodate a high lift. Thus with one cylinder and
three concentric rams, each lifting 10 feet, a height of 40 feet may be obtained. In
Richmond's patent differential telescopic high lift hydraulic lift the water under each
piston is forced into the next cylinder above, so that the rams all travel up simultaneously
at same speed, all reaching the top at same time. This is by far the best and safest lift,
and is the only one to be recommended for heavy passenger work.
WATER PRESSURE ENGINES
These are of two classes, reciprocating pumping and blowing engines and engines
with rotatory shafts and cranks. In the design of these engines it must be borne in
mind that water is a heavy inexpansible fluid, which has considerable inertia, reacting
on anything which it meets to divert its course of motion. Many water pressure engines
have been designed without regard to this elementary fact, and with certainty of failure.
The simplest form is that designed to work a bellows or pump by a to-and-fro
movement only. The more complex forms are those producing rotary motion. A
moving column of water has velocity and pressure, and head or elevation, i.e. the differ-
ence in level between the place where it starts from to the place it falls to.
The total energy is the sum of its momentum, its pressure head, and elevation head.
If V is the velocity in feet per second, p = pressure in Ibs. per square foot, A = the
elevation, and g= the weight per cubic foot, the total energy per Ib. of water is equal to
\7-2 j.
L^+P- + h in feet.
2 - g
All hydraulic engines are arranged to work with one or other of these quantities. For
example, in an old overshot or breast wheel the water drops into the buckets at the high
level with as little force as possible, only falling a few inches into the buckets ; it then
merely acts by its dead weight on the loaded side of the wheel ; the energy is simply
proportional to h x w, that is, height of fall multiplied by the weight falling. A cubic
foot weighs 62.5 Ibs. roughly, and, falling in a bucket a height of 10 feet, would give
62.5 x 10 = 625 foot-lbs.
The pressure head is used in all piston engines. Here the water flows from the high
level at a slow speed through a pipe, the pressure on the piston is that due to the height
of the column of water ; thus a lo-foot head would give a pressure of 625 Ibs. per
square foot on a piston at the bottom ; thus we can utilise /, the pressure.
In the third method we utilise V, the velocity. In this case the water is allowed to
flow with the full velocity due to the pressure, and to strike against a free moving blade
or vane, which arrests its motion and turns the jet aside. The energy at any velocity V
. mV 2
is -_ . If we take a lo-foot fall, again, we find the V for this fall to be equal to 8 ^h
_
_ V 2 2v2 2
^8 ^10 = 25. 2 = velocity, and = energy per Ib. Hence -^- -= 10 approximately, the
energy per Ib. of water, so that the cubic foot would have 10x62.5 = 625 foot-lbs. of
energy. It will be seen, then, that we get the same amount of energy out of every cubic
Hydraulic Engines
foot of water at the same fall whether it acts by A, its dead weight on a wheel, or by
pressure p on a piston, or by velocity V on a vane or movable blade.
In some water power engines we have p and V acting together, as in Professor
James Thomson's vortex turbine.
In the water pressure engines now under consideration p is the acting force, V being
kept below 2 or 3 feet per second by using large supply pipes and ports.
Hydraulic pressure engines are only to be used for very high pressures and for slow,
FIG. 22. Hydraulic Engine.
steady motion. The volume of water used per stroke by pressure engines is constant,
however much energy is exerted. Water can be saved only by reducing the number or
length of strokes on small load. The efficiency at full load is about 80 per cent., at
half-load under 35 per cent. To get over this difficulty many devices have been proposed,
such as the use of two or more cylinders in cranes or lifts, so that for light loads one
cylinder is used, and for full load all the cylinders are used.
For pumping water the load is constant, hence a simple single cylinder is used, and
also for bellows blowing, such as shown in Fig. 22.
FIG. 23. Hydraulic Engine.
FIG. 24. Hydraulic Engine.
The first successful engine of this type was probably that of the design by Mr. Joy,
who applied the auxiliary valve. There being no rotary motion, the valves must be
operated in the first instance by some tappet motion, like that shown in Fig. 23, in
which the piston B makes the little valve rods projecting through the cylinder cover
move when it comes to either end of its stroke. Now, if these rods are made to actuate
a slide valve like D direct, there being no momentum and no expansion, as soon as the
valve opens very slightly the piston stops and the valve ceases to open ; hence only a
small amount of water flows and the speed is restricted.
Modern Engines
The idea of the auxiliary valve is to provide a small easily worked piston to which
the tappets admit water under pressure. This auxiliary piston, having only to move the
D valve, moves easily and quickly, and being 1 balanced by a piston or other means moves
with little friction, and opens the ports full.
Another form with a rotary auxiliary valve is shown in Fig. 24, the valve being
operated by a tumbler motion which falls over and opens the valve wide.
In organ blowing engines the auxiliary valve may with advantage be operated by
air pressure from the bellows and started by a hand valve.
The rotary shaft engines are usually oscillatory, thus obtaining a very simple valve
mechanism. One which has been much used is that shown in Fig. 25, the Schmidt
engine. The oscillation of the cylinder opens and shuts the ports ; the cylinder is carried
on trunions E, concentric with the
curved faces of the valves. These
trunions are carried on stiff levers
pivoted to the bed, and by means of
screws can be depressed so as to
keep the valve faces close up as they
wear away. An air vessel is also
provided on the supply to give some
elasticity and prevent shocks.
In Hastie's engine we have an
example of an effort to construct a
3 - cylinder engine with a variable
piston stroke, so that the water used
will be more in proportion to the
work done, and it met with
.some degree of success. The
crank pin is moved on a
slide against the tension of
springs, so that the greater
the resistance offered by the
load the longer becomes
the centre of the pin from
the crank shaft centre, and
hence the greater the lever-
age of the piston and the
longer the stroke.
In the same direction of
improvement a later type of
engine has been developed
by Mr. Arthur Rigg with
four cylinders. In Hastie's
engine the cylinders were
oscillatory, but otherwise stationary while the crank revolved. In Rigg's engine the
crank pin does not revolve, but is capable of a lateral movement horizontally, while
the cylinders revolve round the pin and also oscillate. The plungers are pivoted to
a revolving crank disc, as shown in Fig. 26. If the central hollow pin, which
also acts as valve, is central, then the engine stands still, and if moved to one side
excentric ; the length of the stroke is twice the eccentricity. The central pin is carried
on differential hydraulic plungers, by means of which it can be moved horizontally to
right or left. There are two ports, an exhaust and pressure port, on the central boss,
which is faced to press against the fixed ports, and each cylinder receives water during
one-half of a revolution and exhausts during the other half.
FIG. 25. Schmidt's Hydraulic Oscillating Engine.
o
Water Turbines
33
If the pin is moved to one side the disc revolves in one direction, and if moved to
the other side it revolves in the opposite direction, and the stroke and consumption of
water is adjusted to the work to be done.
In Hastie's engine the adjustment of stroke is automatically proportionate to the
load ; the Rigg engine requires to be adjusted by hand by valves to the load.
The water enters by pipe on the left hand, and by the hollow plunger reaches the
plung-er ports. The water pressure in the open-ended hollow ram tends to move the ram
to the right, while the right-hand larger plunger can be moved to the left by admitting
water pressure ; the water is admitted to or released from the larger plunger by two
valves, so that it can be adjusted at any point, and by shutting both valves the plunger
is locked in position.
A small engine with 2|-inch rams working at 700 revolutions per minute at 500 Ibs.
pressure runs well, and has power enough to work a heavy crane.
These and other hydraulic pressure engines have a very limited sphere of useful-
ness, although of much importance wherever they can be used, such as for intermittent
work. The high pressure of accumulators is required to give them power, and the
internal strains are very great, so that they are heavy machines and somewhat costly.
FIG. 26. RiggT s Engine.
An hydraulic system for working hydraulic cargo lifts and cranes is exceedingly
useful on steamships, as they work silently, efficiently, and with a maximum of safety,
and do not create the intolerable clatter and noise raised by the steam winch.
WATER TURBINES
This class of hydraulic machinery is the important one. They have been roughly
sorted into classes according to the direction of the water flow in relation to the turbine
wheel. As a matter of fact, it makes little difference whether the flow is in one direction
or another ; but the classification serves as a brief description of the general arrange-
ment of the machine.
We have, ist, axial flow turbines, in which the water flows parallel to the axis of
the wheel.
2nd. Inward radial flow turbines, in which the water flows in radially from the outer
to the inner periphery.
3rd. Outward radial flow turbines, in which the water flows from the inner to the
outer periphery.
4th. Mixed flow turbines, in which the water flows radially and axially.
5th. Free jet turbines or tangential flow turbines, in which a jet or jets of water
strike the buckets or vanes tangentially.
VOL. i. 3
34
Modern Engines
GUIDE BLADE
energy m
FIG. 27. Diagram of Pressure Turbine.
But the real fundamental classification of turbines divides them into two classes
only
ist. Pressure turbines.
2nd. Impulse turbines.
The first act partly by the pressure of the water and partly by its momentum, that
is, by V and p. When the water enters the turbine it has pressure, as all the energy of
the fall has not been developed into velocity. It has velocity also, which it loses in
passing- through the turbine, giving
up the corresponding-
doing so.
As there is pressure on the water
as it enters it must fill the whole
turbine case, which must be water-
tight except at the outflow ; the
water must act on the whole wheel.
In the second class the whole
head of water is developed into a jet
or jets of great velocity, due to the
fall ; such jets have no side pressure.
This may be seen by simple inspec-
tion, for a free jet moves for a dis-
tance from the orifice as a straight
rod of water, not expanding side-
ways as it would do under pressure.
Water under pressure transmits that pressure sideways, forward, and in every
direction ; but in a column of water moving at the full velocity due to the head there is
no side pressure, the whole energy is in the forward movement.
Pressure turbines work best when drowned, so that no part of the fall is lost. But
they will work equally well up to 25 feet above the tail race if the outflow pipe mouth is
carried down below the tail water surface, this part of the fall being utilised by the
suction of the water falling down the pipe. Acting like a syphon, this outflow tube is
called a draught tube. This facility is of value in many cases, for it is not always
advisable nor possible to have the turbine
at the tail water level. And when the
tail water level is variable in height this
draught tube allows of the rise and fall
of the level without affecting the turbine,
except to reduce the suction when the tail
water level rises. Whereas the impulse
turbine must be placed above the highest
level of the tail water, it will not work
drowned ; hence the whole fall cannot be
utilised at low water level in the tail race.
Some turbines have been designed to
work as impulse and pressure turbines, so
that when the tail water is low they work by impulse, but when high and the wheel
drowned, work as pressure turbines. Such turbines are called "limit turbines." They
are not successful ; it is by far better to use a pressure turbine with a draught tube long
enough to keep the wheel above high water mark.
The two forms of turbines may be illustrated in diagrams (Figs. 27 and 28). Fig. 27
is a pressure turbine inward flow, with guide blades so as to give direction to the water
and regulate the quantity ; one guide blade and three wheel passages are shown filled
with water. The water under pressure enters the wheel tangentially, and its course is
RELATIVE PATH ABSOLUTE PATH
FIG. 28. Diagram of Impulse Turbine.
Water Turbines
35
reversed in the passages through which it flows and falls in pressure, so that when it
emerges it has no pressure.
The Jonval type of turbine, upon which system so many pressure turbines are made,
is shown in Fig. 29 as a diagram. It consists of a ring of guide blades curved to guide
the water into a ring of wheel blades curved the opposite way. The water presses
FIG. 29. Jonval's Turbine.
FIG. 30. Impulse Turbine.
through the full wheel passages, strikes and presses against the wheel blades, losing
its pressure in driving the wheel.
The impulse turbine is shown in diagram (Fig. 28) with two jets full and two wheel
buckets nearly full. Here the water is spurted at full velocity against the vanes in the
wheel, which are wide enough apart to allow the water to be deviated without filling
FIG. 31. Impulse Turbine. Guide Vanes
and Wheel Blades.
FIG. 32. Fontaine Turbine.
the buckets ; and in order to allow the water to spread out as it flows through the
wheel, the vanes are splayed out towards the outfall as shown in the cross section
in Fig. 30. The guide blades and buckets are shown in Fig. 31. K is the guide blade,
G and H are vent holes in the side of the wheel buckets to allow air to escape and so
prevent back pressure. In the impulse turbine there may be as many guide blades as
buckets, and the water may flow full bore
all round, but that is not necessary ; there
may be only a few water jets on a large
wheel, in which case it is called a partial
flow turbine.
In the old Fontaine turbine, shown in FlG 33 ._Blades of Fontaine Turbine,
part section, the jets could be closed separ-
ately by sliding gates, as shown in Fig. 32. This shows in the two partially closed
jets that the water does not flow steadily, but is broken up in the wheel, and so its
power is wasted. Again, in the full flow the water chokes the wheel buckets (Fig. 33),
because they are not splayed out at the exit, and the water requires pressure to force
it out.
Modern Engines
In Fig. 34 the Girard improvement is seen, wherein the buckets are curved and
splayed so that the water at full gate never fills them, but leaves an open ventilating
space.
In these impulse turbines, partially closing
the jets is bad regulation ; for regulating purposes
the active jets should be reduced or increased in
number by stopping them off one at a time.
The simplest impulse turbine is the Pelton
wheel, driven by a free jet. This is shown in
diagram in Fig. 35, the wheel carrying on its cir-
cumference a series of double cupped buckets ; the
jet at full velocity due to
FIG. 34. Girard Turbine.
strikes the mid-rib,
splits into two, and glances off reversed in motion.
Now, in reversing it exerts a great pressure on the
cups delivering up its momentum.
In choosing a turbine we must be guided first
by the condition of the tail race. If the tail water
level is liable to considerable variations, then a
pressure turbine should be selected to work either
drowned or through a draught tube. But the
impulse turbine is much easier regulated where water supply is variable and none too
much of it.
Again, pressure turbines run up high in rotational speed
on high pressures, while impulse turbines can be made with
partial flow and with wheels large in diameter, so that with
the same vane speed the rotations are smaller.
The pressure turbine is cheaper than the impulse on
moderate falls with lots of water, but for the higher falls the
impulse turbine, like the Pelton wheel, is cheaper and better.
To sum up the features of the two classes. The impulse
turbine must always discharge above the tail water. It will
work with complete or partial flow for full load. Regulated FlG . 3S ._ PeltonWh ~ 1Bucket .
by closing one or more guide passages.
The pressure turbine discharges either below tail water or above tail water through
a draught tube. It is regulated by closing some of the guide passages or by varying
their area or throttling the water supply. At full load turbines
are all highly efficient, some as high as 85 per cent.; but at lower
speeds the efficiency falls off rapidly. The most efficient governor
at low loads is that applied to the Vortex pressure turbine (Fig.
27), where the inlet to the wheel vanes is closed or opened by
the guide blades moved by a governor.
Here the impulse turbine has the advantage, as we can stop
off the jets without causing losses. A Girard turbine of 200
horse-power on a test gave an efficiency of 0.8 full load and 0.802
on half load, the speed the same in both cases.
There is another type of turbine, not much used but of some
interest scientifically, as it illustrates the true pressure turbine.
Two nozzles mounted on a hollow shaft, and pointing tangen-
FIG. 36. Reaction Wheel, tially as shown in Fig. 36, will rotate with power when pressure
water is supplied through the shaft. To understand this turbine
we begin with the knowledge that a fluid water under pressure exerts an equal
pressure on all sides of a containing vessel. Now, if we remove a part of one
side of a containing vessel under water pressure the water will flow out of the
Water Turbines
37
opening, but the total pressure on the side of the vessel in which the opening- is
made will be so much less than the pressure on the opposite side by the area of the
hole multiplied by the pressure. Thus if the vessel had four sides of 9 inches area,
and out of one side we cut a i square inch hole and the pressure per square inch is
10 Ibs., then on the side out of which the inch hole is cut the total pressure will be
8 x 10 = 80 Ibs. on that side, while the pressure on the opposite side will be 9 x 10 = 90 Ibs.
A difference of 10 Ibs. pushing in that direction would move the vessel away opposite to
the issuing water. And to obtain efficiency the vessel should move at a speed at least
about .8 \/2 gh.
The chief point to observe in the design of this class of turbine is that the supply of
water to the vessel with the nozzle or hole is abundant and sufficient to maintain the full
pressure behind the hole. And if the water is to pass through a hollow shaft this means
a very large shaft.
But this can be overcome and a good construction made of a turbine on this
principle. At the speed required for high falls the water in the wheel acquires an
enormous centrifugal force, and the friction of the water is considerable. By making a
compound wheel this speed can be reduced to half. It is of importance in steam
turbines, this modification for reducing speed without loss of efficiency.
TURBINES, THEIR CONSTRUCTION
We shall now take up each class of turbine with a view to examine their con-
struction. In the preceding division we have only considered their general mode of
action and application. When one comes
to consider turbines practically, not only
is the direction of flow axial, or radial or
mixed flow merely a matter of convenience,
but there is little or no difference in action
between a pressure turbine and an impulse
turbine. We have followed this classifi-
cation so far because it is the usual
text-book classification. In the following
descriptions we shall see that all turbines
work by reaction, by entering a wheel in
one direction, deviating in the wheel, and
emerging as nearly as possible in the
opposite direction. It is a familiar fact
that any body, solid or fluid, moving in
any direction and meeting another body
which intercepts the moving one and re-
verses its movements, experiences a pres-
sure proportional to the square of the
speed before impact.
In this respect all turbines are alike :
the water must be deviated by the wheel.
And so we best can select a starting-point from a turbine which can be dissected into
either pressure or impulse wheels, observe the deviation of the water and the produc-
tion of the acting pressures.
All turbines can be developed from Barker's mill, the original form of which is
shown in Fig. 37, and which was merely a development of Hero's steam turbine
invented 2000 years before.
Whitelaw and Stirrat, in Scotland, made a special study of this wheel, and con-
cluded that the arms should be curved into an Archimedean spiral, which of course
FIG. 37. Barker's Mill.
Modern Engines
would finish at a tangent to the periphery, as shown in Fig. 38. The black spiral line
is the curve of the line of the arm. Whitelaw found that the arm should taper from the
centre, beginning wide and gradually narrowing to the periphery as shown in the left
hand curve, but later tests have proved that the opposite curve, widening out to the
periphery, is equally as good and more efficient.
The two forms are shown better in Figs. 39 and 40, which show one with the
narrow end of the nozzle outward and one with the wide end outward. Now, an inspec-
tion will show the difference in Fig. 40. With the narrow end n out the full pressure
will be continued up to the very end, and the jet will issue
at full speed, so that the wheel periphery must move back
with a speed about o.8*/2gh\ but the point of pressure
pushing the wheel back is down at b near the line F, and
the pressure acts on the back of the nozzle at P.
Now, in the other case (Fig. 40) the full pressure is at
line F, and the water deviates along the reversed curve all the
way, and the pressure is distributed along the curve from a
to b gradually without the water touching the other side of
the nozzle P. Here the issue of the water from the narrow
FIG. 38Diagram of Curvature t j g rest ricted by the curve a, b, and its outflow is at a
of Arms of Reaction Wheel. f , . ./. . ' '. ,
less velocity than if it issued direct into the atmosphere as
in the first case ; consequently the efficiency is higher. Another way to look at it is to
consider, in the first case the centre of back pressure is near the line F, while the centre
of back pressure in the second case is midway between a and b, farther from the centre.
Now, if we cut the nozzles across at the line F and fix the centre piece with the
one inner part of the nozzle, and mount the outer part of the nozzles on a rotatable
shaft, Fig. 39 would exactly represent pressure turbine guide blades and vanes, in which
the pressure would be due to reaction in b. And Fig. 40 would exactly represent the
guide blades and wheel vanes of an impulse turbine.
FlG. 39. Wheel with Narrow-Mouthed Nozzle.
FIG. 40. Wheel with Wide-Mouthed Nozzle.
In the first case the water fills the nozzle at full pressure, and it escapes at full
velocity, the centre of back pressure being at b.
In the second case the water enters the cut-off nozzle at n at full velocity without
pressure, glides along the curve, and falls off with little or no velocity. The centre of
back pressure is half-way between a and b.
In this Hero form of turbine, then, we see that a very simple alteration in the guide
and exit passages makes all the difference between a pressure and an impulse turbine,
and that the guide blades are not necessarily fixtures.
Now again, if we take the form (Fig. 40) in which the water issues from the nozzle
Water Turbines
39
at full velocity, and mount another wheel free to move in the opposite direction to the
reaction wheel inside, as shown in Fig. 41, both wheels would revolve in opposite
directions, the inner being a pressure turbine and the outer an impulse turbine.
FIG. 41. Combined Pressure
and Impulse Turbine.
FIG. 42. Hydraulic Experiment.
In this turbine the two opposite revolving wheels are geared to one horizontal shaft
by 3 bevel wheels.
We are now prepared to consider the various types as constructed, but first we shall
note the usual forms and facts regarding water and the hydraulic formulae.
Atmospheric pressure is roughly estimated at 14.7 Ibs. per square inch, with a
3O-inch barometric height. And the column of water this can
support in a water barometer is 34 feet.
Let H represent head of water, in this case the head is
34 feet; hence
14.7
inch ; or, pressure =
=2.3 feet head=i Ib. pressure per square
^ = 0.432 Ibs. per foot-head.
34
A cubic foot of water weighs 62.5 Ibs. if fresh, and 64 Ibs.
if sea water.
Let P = pressure per square foot, H height of column,
A its area in square feet, and W the weight of water per
FIG. 43. Hydraulic
Experiment.
cubic foot; then PA = WHA, and P = WH, and H = .
These show the connection
between head and pressure. Thus if we connect a long vertical pipe C to a water supply
main B the water will rise to a height H (Fig. 42), due to pressure P.
Now, as to velocity, if water issues from a converging nozzle (Fig. 44) on a vessel
FlG. 44. Hydraulic Experiment.
FIG. 45. Hydraulic Experiment.
as in Fig. 43, we can imagine a single drop of water of weight W falling from the
surface to the orifice. When on the top at rest it has potential energy WH, but
WV 2 WV 2
on falling it has acquired kinetic energy = , so that WH = , and V =
or approximately = 8
4
Modern Engines
The conversion of head pressure and velocity is shown very well by Froude's
experiments shown in Fig. 45. If a pipe of varying- bore is attached to a water
vessel, and small vertical pipes are attached to the discharge pipe at the different
bores, the water will rise in these small pipes and indicate the side pressure at the
different points. As the same quantity of water passes each section of the pipe the
velocity of the flow is inversely as the section, and the pressure inversely as the
velocity ; hence the indicator pipes show the difference in pressure by the height of
water. If this difference of height between the level of the water in the tank and the
V 2
pipes is measured it will be found to be equal in each case to H = , where H is the
2 S
difference between the pipe levels and tank level.
Thomson's water jet pump (Fig. 46) is sometimes explained on the hypothesis that
the high pressure jet acquires such a velocity on discharging at the nozzle that the
pressure falls below that of the atmosphere, but the experiments of Siemens on steam
FIG. 46. Thomson's Water Jet Pump.
jet pumps and others prove that friction between the high velocity jet and the air carries
off the air with the jet, and thus produces a partial vacuum behind the nozzle. This
raises the water until the two streams meet, one at high velocity of small section and
the other of large section and small velocity. The velocity of the one is decreased by
expending its energy in increasing the velocity of the other larger stream sufficient to
discharge the larger body of water through a pipe widening towards the outlet, at a
higher level than that from which it is drawn.
The same action takes place in the locomotive, where the blast pipe carries by its
high velocity jet of steam a large stream of air and gas at a slower velocity out against
the atmospheric pressure. This jet effect will be further referred to under steam power.
When water flows from a reservoir from a pipe B to a lower level n, and discharges
through an orifice of very small size compared with the pipe, the whole head is practi-
cally applied to the orifice, as the velocity in the pipe may be so small as to be neglected ;
but if the discharge is comparatively large (Fig. 47), or the orifice area a large fraction
of the pipe section, then the whole head is expended in the, three forms of H due to
P V 2
head of unexpended fall T^due to pressure, and due to velocity, and multiplying
o
each by W gives the energy in each form ; and in i ib. of water the total would be
P V 2
E = H + ^ H at the discharge.
W 2
Water Power 41
The turbines to be considered act either by one or two or all three of these forms
of energy. In the impulse turbine and jet turbine, like the Girard and Pelton wheel, we
want the full velocity at the jet ; hence the fall pipe must be large enough in area to
reduce the velocity in the pipe to a negligible quantity, so that the full pressure due to
head is at the orifice of the jet, and not expended in the pipe.
In the pressure turbine we require the water delivered to the turbine under pressure,
so that as the water passes through the wheel it losses its momentum in a forward
direction. If V is the forward component of the entering velocity, each Ib. of water
changes its momentum = ; hence gives the forward pressure on the wheel due to
& &
each Ib. of water per second. If V l is the velocity of the wheel at the entrance of
the water, the work per Ib. of water will be = 1 foot-lbs. per second.
g
And as the energy given by i Ib. of water at head H is H foot-lbs. , of which a large
fraction is given to the wheel, .. H= 1, the energy equation for turbines.
o
The guide blades are curved to give a large forward velocity. Practically, the
velocity in pressure turbines at the entrance to the wheel is half that due to the head,
and the pressure half that due to the head.
The quantity of water theoretically discharged by an orifice is the product of
velocity V and the area ejf the orifice A. Cubic feet per second = Q = AV, and
V = 8 */H .'. Q = 8A ^/H. But owing to friction this, in practice, is reduced by 0.96.
FIG. 47. Hydraulic Experiment.
In practice we take the effective head at the wheel by a pressure gauge, as no
coefficient can be given for loss due to pipe resistances. Again, we need not trouble
about the discharge through thin plates or peculiar tubes in this place. In all
turbines the orifices are shaped as near as possible to the form of a converging
nozzle, so that the area of the orifice is equal to the area of the water jet. Allowances
must also be made for friction and bends in pipes, all of which are rather beyond our
present scope.
Momentum of water, as we have seen above, is the important factor in turbine
power. In theoretical mechanics the unit of force is i Ib., and of mass 32.2 Ibs. =g,
W
or mass =
g V
Velocity V is feet per second. Vt= feet x seconds. - =/= acceleration which is the
increase of V in each second. If any force is constant the acceleration is uniform.
This force being measured by the increase of momentum it produces. Momentum
w wv
= mass x velocity, and hence force producing acceleration = -r/J also P/ = _ , and if
f> G
P is pressure and t time = i second, W weight of water flowing per second,
^X = change of momentum, and P = = change of momentum.
To apply these formula? to turbines we must take only useful cases.
42 Modern Engines
ist case, the Pelton wheel bucket or cup. The velocity of the jet is V x , and of the
bucket forward is V 2 . When the jet strikes the relative velocity is V : - V 2 forward,
and when the jet is turned back V l V 2 backward.
The absolute velocity of the jet is V x before it strikes, and after V 2 (V x V 2 ) =
2V 2 -V 1 weight of water per second is WACVj V 2 ) ; therefore the pressure on the
,. f WArVi-VjjJV, WA(V 1 -V 2 )( 2 V 2 -V 1 ) W A , W .. x ,
cup = difference of momentum = i 1 91 1 = L_i 2/v 2 y _ 2 A(V, - V 9 ) 2 .
? ff g .
From this it is easy to deduce that if V 2 = |V 1 the absolute velocity of rejected water
is o, and all the energy of the water given up to the wheel.
In all cases impulse pressure on the turbine blades is
= momentum before entering - momentum on exit.
In pressure turbines it is = (momentum + pressure) (momentum + pressure) before
and after passage through the wheel.
In the Hero reaction wheel momentum is of no account whatever, unless the outlets
are curved to form impulse wheel blades ; then in that case we get the back pressure
= /A 4- the momentum of the fluid entering the curved exit -the momentum on leaving
the curved exit. But as it is difficult to shape the orifice so as to change the momentum
entirely in its passage it is necessary, in using reaction wheels, to compound them with
an ordinary impulse wheel, so that the remaining energy in the jet is absorbed by this
additional wheel. This is more important with wheels driven by steam, air, or gas
pressure, wherein the fluid velocity is very high, and we shall deal with this under
steam turbines.
In the practical construction of turbines all these principles are applied, and upon
them the dimensions of the parts of the turbine are calculated for the best effects.
We shall now consider the practical turbines of present day manufacture. The
simple Pelton wheel has proved the best for high falls under most circumstances, and
we may begin with that turbine. It is not to be supposed that it was designed by its
inventor on scientific knowledge that the pressure was greatest on the cups when
V 2 = ^V 1 , or that completely turning back the jet gave twice the pressure, which he
obtained by striking the jet on the old flat board vanes of the primitive wheel, the
miner's "hurdy-gurdy wheel," which was built in ignorance of all hydraulic laws,
yet it did good work, and probably had an efficiency of 40 per cent. Its great advantage,
however, was that it could be made on the spot from the timber at hand, rough and
readily put up.
An interesting paper, read before the American Institute of Mining Engineers by W.
A. Doble, may be consulted on this wheel, from which the following details are obtained.
It seems that the double cup with the mid-rib was invented by several engineers,
but Mr. Pelton certainly has the credit of introducing it.
Mr. Pelton's own account of his invention is interesting. It appears that one of
the cup-shaped wheels already mentioned got loose on its shaft, and, being displaced
relatively to the jet, the latter struck the side instead of the centre of the bucket. It
was observed that under these conditions the mill which the
wheel was driving ran faster. As, however, the eccentric position
of the jet caused an end thrust on the shaft, Mr. Pelton pro-
posed first to have two sets of buckets mounted on either side
of the wheel, the side thrust on the one set of buckets being
p IG< 4 g Pelton Wheel counterpoised by that on the other, the jet playing on the
dividing line between the two sets of buckets, as indicated in
Fig. 48, for which we are indebted to Mr. Doble's paper. The step from this
arrangement to the splitting wedge was obvious, and was made by Mr. Pelton, a
mid-section through whose bucket is shown in Fig. 50, whilst Fig. 49 shows a com-
plete bucket in perspective. This type of bucket soon came into general use, and
with it the claim for very high running efficiencies. How far these claims are sub-
Pelton Wheels
43
stantiated is even yet a matter of dispute, the results obtained in different careful
experiments varying through very wide limits. The highest figure which seems at all
well authenticated is one of 91.85 per cent., which is said to have been attained in the
case of a 26-inch cascade wheel tested by Professor Hitchcock in the engineering
laboratory of the Ohio State University in 1897. This figure was obtained with a head
of 163 feet ; the flow of water was 2085 Ibs. per minute, and the bucket velocity 46.2
feet per second. The wheel ran at 447 revolutions per minute, and gave 35 horse-power
on the brake. A similar wheel of larger size, giving 80 brake horse-power, showed an
efficiency of 90.04 per cent. In the absence of complete details as to the method of
conducting the test these figures will be treated with some reserve, and perhaps the
same should be said as to certain tests made by Mr. Ross E. Browne in the laboratories
of the University of California in the year 1890. Here a Pelton wheel 15 inches in
diameter, working with a T 7 ^-inch nozzle under a head of 50 feet, is stated to have
shown an efficiency of 82.6 per cent., whilst with a jj-inch nozzle the efficiency was
82.5 per cent. On reducing the head to 8 feet the efficiency was 73 per cent. On the
other hand, in some experiments made in 1897 at the M'Gill University, by Mr. J. T.
Farmer, the wheel made a much less favourable showing, the highest efficiency recorded
being 70 per cent. The wheel tested in this case was a Pelton No. 3 motor, 18 inches
in diameter. The experiments were made with heads up to 235 feet, and with nozzles
FIG. 49. Pelton's Bucket. FIG. 50. Section of Pelton's Bucket. FlG. 51. Section of Berry's Bucket.
i-inch and f-inch in diameter, whilst the speed was varied both above and below the
point of maximum efficiency, which, as stated, was but 70 per cent. This, no doubt,
is a very fair result, but is far below those claimed in the tests made by Professor
Hitchcock and Mr. Browne. If the figures obtained by these observers can be accepted,
it would seem that there is practically no margin left for any substantial improve-
ment in the tangential water wheel. Nevertheless, as commonly constructed, there
are several possible causes of dissipation of energy. In the first place, the angle at
which the jet strikes the bucket varies as the wheel turns round, whereas in a well
constructed impulse turbine this angle is constant, and is, moreover, more nearly
tangential to the direction of the jet than is the case with the bucket of a tangential
wheel.
Further, as often constructed, the flat lower lip of the bucket slaps the jet instead
of entering it quietly. This defect has probably arisen from the makers neglecting to
take into account the relative motion of the bucket and of the jet. A type of bucket
patented by Mr. S. L. Berry, Philadelphia, has no lower lip at all, but is brought to
a sharp edge. The back of this entering lip is bevelled away so that it will stand
clear of the jet throughout its motion. The jet-splitting wedge is curved in the
vertical plane. This curve is chosen so that the vertical angle between the jet and
the wedge changes little throughout the action of the jet. This curve is easily
drawn by determining the virtual direction of the jet at three points in the path of
44
Modern Engines
the bucket and drawing- a circle through these. As shown in Fig. 51, the jet on
leaving the wheel does so in the direction of cd, thus clearing the next bucket ; it
carries away with it, therefore, residual energy
due to the transverse velocity bd. It will be
seen that in order that this should be small
it is necessary that the buckets of a tangential
water wheel should not be too closely spaced,
as otherwise the jet on leaving one bucket will
FIG. 52. Doble's Bucket. Plan.
FlG. 53. Doble's Bucket. Sectional Elevation.
FIG. 54. Doble's Bucket. Elevation.
strike on the back of the next. The bucket just described would seem to have several
points of superiority as compared with the ordinary Pelton bucket. The latter in
another way also violates hydrodynamic principles,
since it is a maxim in hydraulics to avoid any kind
of sharp corner or angle to which the flowing water
may have access. In Fig. 49 it will be seen that
there are a number of sharp corners in the bucket.
It is true that the bulk of the water will avoid these,
but since a jet striking a surface at an angle spreads
in all directions, some water must find its way into
these corners, and some loss by eddying- must thereby
arise. A bucket which appears to have special merits
in this regard is the Doble ellipsoidal bucket, which is
illustrated in Figs. 52 to 55. In this bucket the lower edge is, it will be seen, cut away
near its centre, and the jet-splitting wedge E terminates in a sharp point F, which
will enter the jet as the bucket comes round with a minimum of
disturbance. Whatever direction the water may flow in, it meets
a surface of easy curvature, and is delivered over the edges of the
bucket with a proper residual velocity. We have no particulars as
to the absolute efficiency attained with wheels thus fitted, but they
have been used to replace other buckets on the wheels of the
San Joaquin Electric Company, with an increase, it is stated, of
12 per cent, in efficiency. This statement may, of course, mean
much or little, dependent upon the efficiency of the wheels in
their original state ; but theoretical considerations, as already
mentioned, would lead us to anticipate that the Doble buckets
should prove exceedingly good whenever they are submitted to a r .,
properly conducted test. Another feature claimed for them is that
they wear very uniformly when the water supply is charged with
sand. Thus in the Mount Whitney Power Company's plant,
although the working head is 1300 feet, and no sand traps were
originally fitted, the wheel was still in excellent condition at the end of fifteen months.
FIG. 55. Doble's
Bucket. End View.
45
FIG. 56. Type of Pelton Wheel.
After this a sand trap was brought into use, and it was found that the water, though
clear, was highly charged with granite sand. A pit 100 feet long by 6 feet wide fills up
4 feet in from three to five days, so that during its 15 months' work the buckets must
have been subjected to the scour of 500 or 600 cubic feet of sand per day. As the
velocity of the jet is about 295 feet per second it is only natural that the buckets were
somewhat worn ; but this wear has been uni-
formly distributed, and the surfaces are as
smooth and perfect as when originally fitted.
One of these turbines is illustrated in Fig. 56.
The wheels are 4 feet in mean diameter, and
are designed to run at 250 revolutions per
minute. The most usual plan of governing
wheels of the Pelton kind has in the past
been that of simply deflecting the jet, so that
it failed to' strike the buckets. This plan is,
of course, wasteful ot water, since as much
is sent into the tail race when the wheel is
running light as when loaded, and in practice
is less simple to carry out than might at
first sight appear. In short, the nozzle is
mounted on a ball-and-socket joint, such as is used with hydraulic mining monitors.
This joint has to be kept tight under high pressures, and yet it must nevertheless
move easily when necessary, if "hunting" is to be avoided and the speed of the
wheel kept fairly constant. The difficulties in effecting this led to the very ingenious
system of governing by Cassel.
In this case (Fig. 57) the wheel is built up of two discs mounted on the same shaft,
along which they are free to slide, but are
normally held in contact with each other by
springs. Each disc carries a series of half
Pelton buckets as shown. So long as the two
half buckets are in contact the wheel acts as
an ordinary Pelton wheel. The discs, however,
are provided with weighted levers so arranged
that under the action of the centrifugal forces
developed on the rotation of the wheel the two
discs are forced apart as indicated in Fig. 57, so
soon as the speed rises above a certain pre-
determined limit. The jet then passes between
the wheels and is wasted in the tail race. With
this arrangement the wheel is its own governor,
and accordingly the control of speed is remark-
ably prompt and efficient.
The plan of reducing the pressure pro-
ducing flow through the nozzle by means of
a throttle valve has also been tried, but is
objectionable, in that the speed of the jet is
thereby reduced and the water no longer enters
the wheel without shock, a waste of energy thus arising. A simple plan, which
we can well believe gives very good results, is described and illustrated in Fig. 58.
In this case a sharp edged sluice is, it will be seen, fitted before each nozzle, and
the wheel is governed by moving this sluice so as to cover more or less of the opening.
This method affects the velocity of the jet very little, but changes its shape, and there
may therefore be some loss from splashing, owing to the inversion of the jet. As is
FlG. 57. Type of Pelton Wheel.
Modern Engines
FIG. 58. Sluice Governors of Pelton Wheel.
well known, to be stable a jet must be circular in section. If of any other form the
surface tension tends to bring it to this form, but overdoes the correction, so that in
point of fact the shape of the jet
oscillates about the circular sec-
tion. It is quite probable that with
heads of fair weight the velocity
of the jet will be too great for any
action of this kind to have any
perceptible effect in the passage
between the jet and the buckets.
Mr. Doble's arrangement is of
the throttling type, but the details
of the construction are such that
a solid cylindrical jet is obtained,
issuing at practically constant
velocity at all gates, from full bore
down to under one-fifth full bore.
The throttle piece is pear shaped,
and is maintained central with the jet by being mounted on a cylindrical rod passing
through the back of the casing, and which can be moved in or out. The water issues
through the annular space between the
throttle piece and the jet. The outer end
of this throttle piece projects always through
the orifice of the jet, and terminates in a
fine point connected by easy curves with the
thickest portion of the body. The water
clings to these curves, so that the jet is
always cylindrical and solid. Figs. 60 and
61 show the jet obtained with various open-
ings of the nozzle. The steadiness and trans-
parency of the jet is remarkable (Fig. 59).
Nearly all tangential wheels act much in
the same way as this Pelton wheel. The
Girard turbines being much the same in
principle, although very different in construc-
tion, in both the momentum of the jets is
converted into pressure on the wheel cups or
blades in being arrested and turned back in
its motion ; and while the angular velocity
of the wheel at the wheel blades is necessarily about half the linear velocity of the jet,
the revolutions of the wheel may be chosen and arranged for by calculating the diameter
necessary to obtain the angular velocity at the given speed, and from this the radius, or
FIG. 59. Jet from Doble's Nozzle.
FIG. 60. Jet from Doble's Nozzle. Full open.
one of the radii r of the wheel, may always be found ; thus in a wheel like Pelton's r is
the radius from centre of shaft to centre of cups. Suppose that half the velocity of
the jet is 30 feet per second, and the wheel limited to 10 revolutions per second, then
Outward Flow Turbine
47
-? = 3 feet as the circumference at radius /-. the centre of the buckets from which r
10 revs.
is found.
In the outward radial flow (Girard) there are the inner radius rand the outer r 2 ;
and having found r, r 2 is found by k x r, wherein k is a constant about = 1. 17 to 1.18.
FIG. 61. Jet from Doble's Nozzle. Partially open.
THE GIRARD TURBINE
This is an impulse turbine in which the wheel buckets are not under pressure, but
may be a full or partial flow wheel.
It is made in various forms, like all other turbines. Thus for high falls it can be
placed at the bottom of a wheel
pit and the water led down by a
pipe, so that it enters either be-
tween twin wheels or under one
wheel, and supports the weight
of the wheel and shaft, as in the
case of Niagara Falls.
And, again, for high falls
where slow rotation is desired it
can be made a partial flow wheel,
so that although the rotations are
low the peripheral speed is still
high enough for good efficiency.
Fig. 62 represents Messrs.
Gunther's turbine of the class for
deep wheel pits. When a turbine
is placed in a deep pit, and the
upright shaft is of extreme length
and of great weight, it may be
desirable to reduce the load on
the footstep as much as possible,
and in such cases a Girard tur-
bine of special design with radial
outflow is adopted. In this type
the water enters the turbine from
the underside, is admitted on the
inner circumference of the wheel, and is discharged radially outwards ; the water
pressure thus exerts thrust on the wheel, and the footstep has only to carry the weight
of the moving parts.
Fig. 63 is from a photo of one of two radial Girard turbines driving a large textile
mill abroad, each wheel being 63 inches internal diameter, and of 525 horse-power, with
95 feet fall ; the two turbines drive by steel bevel wheels on to the same second motion
shaft, and each upright shaft is about 80 feet long.
FIG. 62. Girard Turbine lor Deep Wheel Pits.
48
Modern Engines
The two diagrams, Figs. 64 and 65, show the construction of the partial flow turbine
of this type. G is the inlet nozzle, furnished at its nose with two or more guide blades
movable by rack and pinion for regulating purposes. It will be seen from this that the
wheel's diameter is independent of the velocity of water flow.
FlG. 63. Girard Turbine with Radial Outflow.
For this class of turbine we will take a partial flow example with a high fall, and
not a great quantity of water. Let us take
h as the head in feet = 400.
Q the quantity of water available in cubic feet = 4 per second.
R = 500 the limit of revolutions of the shaft per minute.
Outward Flow Turbines
A is the area of the guide passages.
Ag the area of the wheel passages at entrance or inner radius.
r is the outer radius wheel.
49
r<> the inner radius of the wheel.
C the velocity of the inner radius in feet per second.
C 2 velocity of the outer radius in feet per second.
V the velocity of water issuing from guide passages.
V 2 velocity of water from wheel.
Then V = o.94x8>/A (i).
, ,
"
L 2 = y x i. 08, as the wheel blades must never be filled.
} = C (found by taking the revolutions per second) -i-R (in this case 52? = 8.?),
60
and dividing C by this number = , the product will give the cir-
R
cumference of the inner r 2 , from which r can be found.
FlG. 64. Section of Girard Turbine.
Beginning with C, we find that C =
FIG. 65. Section of Girard Turbine.
0.94 x 8 *J h _ 0.94 x 8 V4oo 0.94 x 8 x 20
= -- feet = 75 feet per second ; and the value of r 2 = -- = - = 9 feet as circumference
of r 2 , from which we get easily 1.5 feet as the radius of r z .
The area of the guide blade passages is = A = ^= -4 = 0.0266 square feet, or
nearly 3! square inches.
Area of wheel passages A 2 = Ax i.o8=3.75x 1.08 = 4 square inches.
The circumference 9 feet, found above, would be the inner circumference, and for
the purpose of calculating the axial breadth of the wheel blades we must find the
number of blades required.
The guide blade passages A = 3. 75 square inches; and if we divide this between
three orifices in the nose ot the guide each would be 3-75 _ 1>2 g square inches area,
3
say i^ x i ; i inches axially and i inch measured circumferentially.
VOL. i. 4
5
Modern Engines
The wheel blades passages are somewhat larger = 4 inches, and axially they would be
the same as the guide blades, i^ inches ; and if we divide the 4 inches into 3 we get
4=1.33 inches as the width between the blades on the wheel at the inner radius r in
3
inches; we must add to this the thickness of the blades. In a wheel of this high velocity
for a small quantity of water it would be made as light in weight as possible and well
balanced. The blades would be no more than inch thick, hence from centre of one
blade to centre of next would be = 1.33 +0.125 = 1.455 inches ; and this into 108 inches,
the circumference = -^ = 74 blades.
1-455
If the guide nozzle had only one orifice of 3.75 square inches we could alter the
wheel blade design considerably; thus the wheel orifice, being 4 inches, could be
divided into two of 2 x 2 inches, and the nozzle, one orifice, ?i75_ 1.875 inches and
2 inches axially measured. This would give 2.125 from centre of one blade to the
next on the circumference ; hence
1 08
2.115
= 50 blades.
FIG. 66. Girard Turbine Complete.
Fig. 64 illustrates the sections
of the wheel and guide blades and
passages, and Fig. 65 the orifices,
nozzles, methods of regulation by
a sliding sluice worked by rack and
pinion.
The blades are splayed out, as
formerly explained, to enable the
water to spread as it passes through
the passages, so that the outer area
of the passages is somewhat greater
than their inner passages.
The outer radius r should be to
r 2 as 1.18 is to i. In this case r z is
1 8 inches ; hence r= 1.18 x 18 = 21^,
and the buckets would therefore be
3^ inches radially.
The efficiencyis about 75 per cent.
at best, so that as the power = ^ '-?, = 240 cubic feet per minute, and h = 400, 62.5
33,000 ..
240 x 400 x 62. 5 , , ,
the weight of one cubic foot, hence 0.75 5 = about 130 actual horse-power.
33,000
There is no pressure between the guide blades and the wheel, and as the water
enters the buckets without pressure it is freely deviated by them, and takes a course
quite independent of their shape. The action of the water on the wheel depends on
the angle through which each particle is deviated whilst freely flowing over the buckets,
and as these latter are not full there is no disturbance of the action as they pass in front
of, or away from, the jets.
The turbine illustrated in Fig. 66 is shown with part of the hood removed, which
prevents the water from flying about. The water enters through the valve shown at the
left-hand side of the Figure, and passes directly into a distributing chamber, from which
it jets on to the buckets of the wheel through gates or ports, varying in number
according to circumstances. When there is only one port it can be reduced in size
when water is scarce or less power required ; and when, as is frequently the case, there
are many ports, a sufficient number of them may be used to suit the requirements of the
moment. The arrangement by which the power is regulated is easily worked either
by hand or by a governor.
Water Turbines
THE VORTEX WHEEL
These are inward flow turbines with fixed guide blades, and include the Thomson
vortex wheel with a few movable guide blades for regulating the quantity of water
and power.
This vortex turbine has much to recommend it. The regulation is easy and
certain, due to the balancing of the inward
pressure by the outward pressure, due to
centrifugal force. In fact, it will interest
electricians to study this turbine, as it is a
hydraulic illustration of the electromotor, and
both act upon the same fundamental mechan-
ical law. This turbine, when supplied with
water under pressure, acts as a motor and
gives off power, but at the same time it acts
as a centrifugal pump, and hence produces a
counter pressure, and it works at its greatest
activity when the counter pressure is half the
working pressure. A pump is a water pres-
sure generator, a turbine is a water pressure
motor, a dynamo is an electric pressure gener-
ator, an electric motor is an electric pressure
motor. The vortex and other inward radial
flow turbines act in both capacities when at
i T .LI. i FIG. 67. Turbine at Niaeara Falls.
work. In the electromotor a counter pressure
is generated in the same way, and the greatest activity is also obtained when the
counter pressure is half the working pressure, but its efficiency is greater the nearer
the counter pressure approaches the value of the working pressure.
In considering this fact tne author advocates the inward flow radial type of turbine
for single expansion steam turbines, in which the counter pressure would moderate the
-steam flow. However, that is for another chapter.
We shall enter fully into the consideration of this type
of turbine, for with this, the Pelton wheel, and improved
Jonvals all cases can be met.
The turbine selected for illustration is the Gilke's
vortex, and after fully describing it we shall calculate its
dimensions. When these are correct and the speed
adjusted to the number of revolutions necessary to give
the counter pressure about half the working pressure, it
runs with great smoothness and silently.
Referring to the fundamental turbine, the Hero reaction
wheel, this vortex wheel can be traced back to that also ;
for it would work even if the guide blades were fixed to the
wheel blades and rotated with them ; in fact, it works well
so made. This is shown in Fig. 68.
In fact, all pressure turbines can be traced to this
simple form of wheel, and by starting from that ancestor we could treat them as
reaction wheels, from the size of the orifices at the outlet, and the pressure of water
in the wheel, and the centre of pressure.
Let us for a time depart from the orthodox treatment of pressure turbines, and
consider them from another point of view. Turbines have been very thoroughly
investigated, and the results of practice have been carefully and accurately compiled,
FIG. 68. Turbine with Rotating
Guide Blades.
Modern Engines
so that new views, while they may not advance the theory much, may, however,
have some practical results ; indeed, they have already had practical results of value.
In further developing- the fundamental turbine, suppose we make a wheel as in Fig-.
67, the inner curved part of the blades B would, in an inward flow turbine, be
fixed, and the inner reverse curves, the wheel, movable. In this case (Fig. 68),
however, they are one blade and move together, the pressure water entering on the
inner periphery and flowing 1 out at the outer A into the exhaust. We could calculate
the power by the pressure difference between the one side of the passage and the
other due to the unopposed pressure opposite the exit of the passages, just as we
could in a Barker's mill. In this case the inner orifices have a larger area than the
outer ones, otherwise there would be no pressure in the passages ; we make it
an outward flow wheel by a slight alteration, wherein the orifices on the outer
periphery are contracted, and the inner ones opened out to admit the water pressure.
Here we have the genesis of the outer and inward flow pressure turbines called reaction
turbines. First, it is observable that the inward flow has the advantage, in that it is
End View, with Front Cover Removed.
Sectional Elevation.
FIG. 69. Inward Flow Turbine.
natural that the inner orifices should be the smaller on the small circumference, and the
large orifices on the outer radius freely admit the water.
Let us see what teaching there is in this theory. First, to revert back to the
reaction wheel in its simple form, to maintain the pressure in this turbine the inlet must
be large compared with the outlet nozzles in sectional area, in order that the pressure
of the water be maintained in the wheel opposite the exit, so as to get the back driving
pressure. The exits at B in the inward flow turbine would then be small compared with
the entrances at A, and the full pressure due to the head would be maintained in the
wheel passages, and the pressure being high would tend to cause great leakage past
the sides of the wheel into B. But that can be overcome mechanically, as in the
ordinary vortex wheel. A wheel on this principle is shown in Fig. 69, an end view and
sectional elevation. The fluid enters at H and fills chamber A at full pressure ; it enters
wheel C by wide orifices on the outer radius, and is deviated tangentially and discharged
through smaller nozzles or orifices in the inner radius, from which it is exhausted
through E. These are diagrammatic figures of actual turbines ; the entrances to
the wheel passages are curved so as to save friction and eddy currents. Now, it is
evident that in this turbine only pressure is at work, and very little of that pressure is
due to momentum.
Water Turbines 53
The nearest approach to it in action is the vortex wheel, in which reaction is due
to half pressure and half momentum, and by altering this wheel so that its outlets are
very small compared with its inlets, it can be made also to work with small momentum
and large pressure in the wheel passages.
The chief advantage and only raison cP&tre for the use of guide blades in pressure
turbines of reaction type is to convert the pressure of the water into momentum in the
guide passages, so that the pressure is small at the clearance between the two sets of
blades, which must be of some width, and not subjected to high pressure to cause great
leakage. The guide blades restrict the inflow of water, and as it finds little resistance
owing to the outer or inner discharge orifices being nearly as wide as the inlets, i.e.
the guide passages, it rushes through at high velocity across the clearance space into
the wheel passages ; here it is arrested and its momentum reconverted into pressure,
and is discharged in the opposite direction from the orifices. This, then, is the object
of fixed guide blades, to reduce the pressure by converting it into velocity, so that the
pressure is relieved from the exit of the guide blades and the inlet of the wheel passages,
enabling the water to jump from one to the other at high velocity and with little side
pressures tending to leakages.
Fig. 67 is a part section ot the pressure turbines at Niagara Power Generating
station, showing the shape of blades and passages very well, and they are certainly of
correct form for fixed and movable blades. But the same turbines could be altered to
work without guide blades ; this I have shown in Fig. 68.
In both cases the pressure water would enter from the inside for mechanical reasons,
and discharges into the atmosphere. In the turbine with fixed blades (Fig. 67) there
are three conversions of energy : first, the potential energy of the fall is converted into
kinetic energy in the guide blades, and each particle is for an instant arrested in the wheel,
where its kinetic energy is converted into pressure, which again is converted into velocity
with which the water escapes. Under these conditions the velocity of the wheel blades
should be , where V is the velocity of the water as it jumps the clearance space, this
2
in order to use the momentum to best advantage.
In the case of the wheel with no fixed blades there is only one conversion of energy.
The potential energy of the water is converted into kinetic energy by the difference of
pressure caused by the unopposed pressure opposite the orifice moving the wheel round.
Theoretically, the best speed would be that due to the head of water, so that the outer
radius should have an angular velocity = V = 8 -Jh nearly in this turbine.
The question between the two is, then, whether the guide blades fixed are better and
cheaper than the packing devices required to prevent leakages of water in a wheel
without them.
The vortex wheel shows that the packing is no difficulty, and its four large guide
blades might, if it were not for their excellent regulating functions, be discarded by
curving the wheel blades differently, as shown in Fig. 68.
The reason why in all turbines so much depends upon the angles and curves of the
blade is simply due to the velocity of the water passing, and its inertia, and momentum.
It must be diverted and have its motion altered in direction gradually, and pass from
the guides to the wheel without striking a sudden blow or shock.
From these few observations the student will gather the different views on the
theories, and their discussion is interesting. We will now proceed with the discussion
of the reaction pressure turbine, the vortex wheel.
In pressure turbines the buckets of the wheel must be completely filled with
water (as otherwise there would be no pressure), and consequently the turbine must
receive the water all over the circumference in order to act efficiently. If this be not
done one of two actions takes place according as the turbine be working submerged or
not. If the turbine be submerged the buckets, after leaving the guide blade orifices,
54
Modern Engines
are full of still water, which has to be displaced when they come round to the orifices
again. Hence in each revolution a considerable amount of power is wasted in
imparting momentum to the
mass of water carried round
in the wheel. If the turbine
be not submerged a con-
siderable quantity of water
is used in filling the buckets
(which have become empty
on leaving the guide blade
orifices) before the full pres-
sure at the circumference
can be utilised. It is not
easy to find a mode of re-
gulating pressure turbines
which does not involve
shutting the water off part
of the circumference (so
making them partial ad-
mission turbines) or other-
wise greatly affecting the
conditions of the admission
of the water. The vortex
turbine is only affected by
back water to the extent
FIG. 70. -Vortex Turbine. that the water level is
affected. It can be placed
above the bottom of the
fall, so that part of the head
can be utilised by suction.
And in many cases it is
the cheapest form of tur-
bine to adopt.
The principles of the
vortex and the mode in
which the water is applied
in it will be readily under-
stood by reference to Figs.
70 and 71.
Fig. 71 shows the in-
ternal arrangement of the
vortex. The water enters
the outside casing at the
top, or in any other posi-
tion that may be conven-
ient, and passing thence
is directed by four (or more)
guide blades on to the outer
circumference of the re-
volving wheel, which is
driven round at a velocity
depending on the height of
the fall. The water, having FIG. 71. Vortex Turbine with Cover Removed.
Vortex Turbines 55
expended its energy in giving motion to the wheel, is discharged through the two
central openings, half the amount being carried away by each suction pipe. The guide
blades, it will be noticed, are movable, and turn about on a pivot placed near their
inner ends. The method in which these guides are simultaneously regulated will be
seen by reference to Fig. 70, which is an outside view of the same turbine.
The power is obtained with a slower velocity of water than in ordinary turbines.
This is effected by the balancing of the centrifugal force of the water in the revolving
wheel against the pressure due to half the head, so that only one-half of the fall or head
is employed in giving velocity to the water, the other half acting simply in the condition
of fluid pressure. Hence the velocity of the water in no part of its course exceeds that
due to one-half of the fall, and the loss from fluid friction and agitation of the water is
thus materially less than in other turbines where the water is required to act at much
higher velocities.
From the principle of injection of the water from without towards the centre,
which is adopted in the vortex, there results another saving of effect, since it admits
of the use of long and well formed channels by which the water is made gradually and
regularly to converge in passing from the outer chamber, where it is comparatively at
rest, to the point of entrance to the wheel chamber, where its velocity should be greatest.
The advantage of such convergent channels for the transmission of water with a
minimum of loss of effect, as compared with short passages such as are generally
employed in other turbines, is well known.
Further, from the same principle of injection towards the centre there is an
accordance between the velocities of all parts of the moving wheel and the proper
velocities of the water in its passage between the points of entrance and discharge.
The water when it has its greatest velocity is admitted to the circumference of the
wheel, which is the most rapidly moving part, and when it has as far as possible
imparted its power to the wheel it leaves at the centre, which has the least motion.
The water enters from the guide passages with the velocity at which the outer
circumference of the wheel is moving and without change of direction, so there is no loss
from impact.
The steadiness and regularity of the motion of the vortex are remarkable, conse-
quent upon the action of the centrifugal force of the water, which on any increase in the
velocity of the revolving wheel, increases and so checks the supply entering from
the guide passages, and on any diminution of the velocity of the wheel, diminishes and
admits the water more freely, thus counteracting in degree the irregularities of speed
arising from variations in the work to be performed. With other turbines and ordinary
water wheels the variation in speed, when the amount of work or resistance to the
motion varies, frequently is a source of great inconvenience.
Besides the adaptations referred to, which combine to give the vortex a great
superiority in efficiency, it possesses another advantage, which is in many situations of
the highest importance, namely, the simple and excellent mode of adjustment to varying
supplies of water by means of the movable guide blades.
The guide blades, as already mentioned, are made movable upon gudgeons or
centres near their points, motion being imparted to them simultaneously by a hand wheel,
which can be placed in any position easily accessible. By a very slight motion of the
guide blades the orifices may be contracted at pleasure, and are thus made to suit any
quantity of water which it may be necessary or desirable to use ; and it will be seen that,
to whatever extent these are open, the following important conditions of efficient
application of the water are fulfilled :
i st. The channels are of a gradually convergent form.
2nd. The water is uninterrupted in its course, and enters the wheel chamber from
the narrowest part of the channels, and consequently attains its maximum velocity at
the point of application.
Modern Engines
up
FIG. 72. Vortex Turbine Wheel.
3rd. The water is admitted equally to the whole circumference of the wheel.
Fig. 72 is taken from a photograph of a vortex wheel with the outer cover
removed to illustrate the form of the vanes. Some of these do not extend to the
central orifice ; the object in so making them is that they may not too much fill
the contracted part of the passages and thus impede the flow of the water.
The wheels (Fig. 73) are constructed either of steel or of rolled brass (the latter
for small sizes), and as the vanes can thus be made very much thinner than of cast metal
their number can be increased, and perfect
accuracy in the curvature secured. Hence the
water enters the wheel with less interruption
and passes through more exactly in the direc-
tion intended than is the case where the vanes
are of greater thickness and fewer in number.
The vanes are fixed on each side of a steel
or brass plate, which has a boss in the centre
to secure it upon the shaft, and there are
two outer discs or covers in which are left
circular openings through which the water
passes after it has done its work ; thus one-
half of the water is discharged on each side of
the wheel.
A separate movable cover is placed over
the wheel, so that access to it can be had
without disturbing the exterior casing. A
cover is also placed over the guide blade
chamber in the larger wheels, by means of
which access can be obtained to every part without moving the foundations or heavier
portions of the case.
The double vortex, with movable guide blades, should be adopted on all medium
and high falls in cases in which the amount of power employed varies considerably at
different times ; and the saving of water is important, so that it is necessary to use as
small a quantity as possible to do the work required ; or when the available supply of
water is at times less than the full amount for which the turbine is designed. The
consumption of water can then be economised to the utmost, as the passages can be
regulated to admit only the exact
quantity needed to do the work or to
suit the available supply.
The governing of turbines is in
many cases more difficult than the
governing of steam engines, the two
chief difficulties to be overcome being
the weight of the moving parts and
the fact that if the guides or sluice
by which the governing is effected are
worked too quickly very considerable hydraulic shock takes place, which may do
damage to the pipe line. There are many governors in the market ; most of these will
give fair results to prevent racing, that is to say, they will not allow the turbine to run
away so much as to do damage to ordinary machinery, but where great regularity of speed
is required under varying loads, for example, where a dynamo is supplying the electric
current direct on to lamps which are switched in or out from time to time, it is necessary
to provide a sensitive governor which will not admit of perceptible fluctuations in speed.
Our experience of Mr. Murray's hydraulic governor enables us cordially to
recommend its adoption for electric lighting plants such as are alluded to above, or
FIG. 73. Vortex Turbine Wheels.
Vortex Turbines
57
for any other purpose in which extreme steadiness is essential. With this governing
apparatus a full description of which will
be found below, the water is, by the action
of a high speed governor, turned to one end
or the other of a hydraulic cylinder which
works the guide blades. Such a governor
will be seen in Fig. 74, Nos. i and 2 being
sections at right angles to one another of
a 4-way valve forming a part of the appar-
atus, and No. 3 an elevation showing the
complete arrangement. In Nos. i and 2
the outer casing A is provided with a supply
port B, an escape port C, and other ports
D and E leading respectively to the top
and bottom of a regulating cylinder which
controls the supply to the motor. The action
of the apparatus may be seen from No. 3,
in which 1 is a centrifugal spring governor
driven from the turbine and connected by
the rod J with a lever K, to which the
piston H is connected by a link. When the
turbine speed increases the piston is raised
and water passes by pipe D to the upper
part of the regulating cylinder L, and acts on a piston M to reduce the supply. Should
1"
FIG. 74. Murray's Governor.
FlG. 75. Method of Fixing- Vortex Turbine,
the speed decrease the water passes by pipe E to the lower part of the cylinder L and
Modern Engines
raises the piston, and permits more water to enter the turbine so as to maintain normal
speed.
The essential difference between the single and double vortex turbines is that in the
former the water is only discharged from one side of the wheel ; the wheel of the single
vortex being, in fact, half of that of the double, the centre plate of the latter forming the
top cover of the former.
The guide blades direct the water on the wheel in precisely the same manner as they
do in the double vortex, and may be either fixed or movable as is desired.
This single vortex turbine is very well suited to medium and low falls, where a
considerable body of water is to be dealt with, and where it is desirable to have the
turbine left dry when the head water is shut off.
As the water is only discharged below the wheel, and not above and below as in the
double vortex, part of the fall may be utilised by a suction pipe.
When difficulty or expense is apprehended in sinking the foundations a considerable
FIG. 76. Vortex Turbine Direct Coupled to Dynamo.
saving may be expected, as the turbine can be placed on a platform above the tail water,
and no work need be done below the water level beyond the sinking of a hole for the
suction pipe to discharge into.
The single vortex is, of course, only to be used when the wheel is placed horizontally
with a vertical spindle.
It will be seen from the illustrations that the heavy cast-iron casing that is required
for the double vortex is dispensed with.
The usual method of fixing the single vortex is shown in Fig. 75. The wheel is
placed on a floor made of timber or metal in a pit of masonry or brickwork. The
timber breasting (which for convenience is shown broken away in the engraving) is
of the same height as the sides of the watercourse. There is no way for the water
to pass from the chamber above the floor to the tail race below except through the
turbine.
The diameter of the single vortex is naturally larger than that of a double vortex
for the same power ; but so great are the advantages of dispensing with the outer casing
and keeping the turbine above the tail water level that, where the quantity of water is
large and fall low, it is adopted.
As the wheels are comparatively small in diameter, the rotational speed is pretty
high. The speed of the wheel is some large fraction of the speed ot the water as it
would flow out ot the wheel were it held stationary ; and being closed all watertight,
Vortex Turbines
59
it can very readily be coupled to a dynamo direct, and a great many of these vortex
turbines are so used. Fig 1 . 76 represents one of small size so adapted.
Below we give some particulars of interest regarding the single vortex wheel,
from which the quantity of water required, the speed, and diameter of guide blade
channel can be obtained.
TABLE IV.
FALL IN FEET.
H.P.
8
9
10
ii
12
[
Cubic feet per minute ....
883
784
706
642
588
10 '
Revolutions per minute ....
9 1
106
118
128
1 37
1
Approximate weight (cwts.) .
Diameter of guide blade channel (inches)
53
63
47
61
43
59
4i
58
39
57
(
Cubic feet per minute ....
J 3 2 5
1176
1059
962
883
iJ
Revolutions per minute ....
80
89
99
107
118
5 I
Approximate weight (cwts. ) .
63
59
55
53
49
I
Diameter of guide blade channel (inches)
6?
. 66
64
63
61
(
Cubic feet per minute ....
1765
1470
1412
1284
1176
Revolutions per minute ....
7i
83
89
96
1 06
20J
Approximate weight (cwts.) .
73
65
63
61
57
I
Diameter of guide blade channel (inches)
73
68
67
66
65
The particulars for double vortex wheels with horizontal shaft are-
TABLE V.
FALL IN FEET.
H.P.
36
38
40
45
So
55
60
"1
Diameter of inlet (inches)
Cubic feet per minute
IO
118
9
106
8
88
8
76
66
7
59
7
53
1
Revolutions per minute .
810
960
1055
"37
1389
1474
'553
r
Diameter of inlet (inches)
ii
ti
IO
9
8
8
8
20\
Cubic feet per minute
157
141
118
IOI
88
78
71
(
Revolutions per minute .
768
810
936
"37
1216
1290
1359
(
Diameter of inlet (inches)
12
12
ii
IO
9
9
8
25
Cubic feet per minute
196
I 7 6
147
126
no
98
88
I
Revolutions per minute .
730
769
886
ion
1216
1290
1359
,
Diameter of inlet (inches)
13
13
12
ii
10
IO
9
3O-{
Cubic feet per minute
235
212
I 7 6
151
132
118
106
I
Revolutions per minute .
644
680
8 4 2
958
1080
1290
1359
f
Diameter of inlet (inches)
15
IS
13
13
12
ii
n
40
Cubic feet per minute
282
235
202
I 7 6
r 57
141
(
Revolutions per minute .
595
629
744
910
IO24
1086
"44
f
Diameter of inlet (inches)
17
16
15
H
13
12
12
H
Cubic icet per minute
392
353
294
252
221
196
I 7 6
1
Revolutions per minute .
592
629
744
803
973
1932
1087
6o
Modern Engines
TABLE \ T .continued.
H.P.
FALL IN FEET.
36
38
40
45
50
55
60
6oj
Diameter of inlet (inches)
Cubic feet per minute
Revolutions per minute .
47
522
18
424
549
16
353
683
VO CO CO
H Q O
COCO
264
859
13
235
1032
13
212
1087
7j
Diameter of inlet (inches)
Cubic feet per minute
Revolutions per minute .
20
549
476
19
494
502
18
412
602
16
353
737
15
3Q
788
15
275
910
00
,{
Diameter of inlet (inches)
Cubic feet per minute
Revolutions per minute .
22
628
476
21
5 6 4
502
19
47
549
17
404
649
16
353
6 95
837
'5
282
882
IOOX
Diameter of inlet (inches)
Cubic feet per minute
Revolutions per minute .
24
784
398
23
706
462
21
5 88
505
19
54
580
18
441
635
392
673
16
353
706
The following construction is used in setting out the blades in an inward radial
flow turbine.
The guide and wheel blades for inward flow turbines are partly involutes of a circle,
to lessen the contraction of the stream as it
flows through the passages. Fig. 760, A is
the external circumference of the guide ap-
paratus, and B the internal circumference ;
draw the perpendicular CD. Make the angle
CBE 15, and draw the line CE at right
angles to BE, then describe a circle with C as
centre, through the point E. Let BF be the
inner pitch of buckets, and suppose a thread
wound round the circle E, and carrying a
pencil at B ; as this is unwound, the arc BGH
is traced by the point B, the point H being a
little to the left of the line JFG, and the width
of the passage being, therefore, uniform from
G to H. The remainder of the bucket HA[is
a portion of a circle, the tangents AD and
AK containing the angle 87.5. The buckets
in the revolving wheel are constructed similarly,
except that the angle L is made 12 instead of 15, and the angle M, instead of 87.5,
should be made an angle of 90.
In some constructions the number of guide blades are made equal to or nearly
equal to the number of wheel blades. In the vortex we have only four or six guide
blades.
The leading dimensions of a vortex wheel can be found as follows :
Let Q = cubic feet of water per second, as before.
h = the effective head.
V = velocity in feet per second due to h.
Vi = M ,, in guide blades.
V 2 = ,, ,, ,, in inner orifies.
FIG. 76. Diagram to find Curvature of
Turbine Blades.
Inward Flow Turbines 61
Let A x = cross section guide blade orifice at discharge.
A 2 = ,, ,, wheel passages.
fj = radius of circumference on which water enters wheel.
r z = ,, ,, ,, ,, leaves wheel.
Now, in setting out upon a design we have given the value of Q and h. From
this r v r% can be calculated if we fix upon a limiting number of revolutions per
minute.
There are other mechanical calculations, such as the breadth and depth of the
blades and their number and their combined sectional opening.
Then let C 2 = the velocity of inner radius of wheel, and
C : = ,, ,, outer radius of wheel.
Then it is found that the area A of the guide passages should be 1.15 times the
area A 2 of the wheel passages, and that r^ should be 1.17 times r%. And if there is a
draught tube, its area A 3 should be 1.25 times the area of A 2 .
We can now proceed with the design in figures, starting with the given data for
which a turbine is required
Q = 13 cubic feet per second.
h = 36 feet effective.
V 1 = o.57 x8 ^ = 4.56 ^36 = 27.36 feet per second.
Now, from construction the velocity C 2 at the inner radius equals V x x 0.966 . . C 2
= 27.36 x .996 = 26.44 f eet P er second.
Then, we can settle the value of r^ and r% if we know what speed in revolutions
per second is required. We will assume the speed required is 420 per minute, or 7
per second ; and if we now divide V l by speed per second we get the circumference
at 7* 2 , and from the circumference the value of r. 2 from a table ; hence -^^ = 3-77 feet, or
45 inches circumference. From a table of circumferences and diameters we find the
diameter corresponding to this circumference = 14^ inches. Again, to find r v the outer
radius, a rule in practice arrived at by experience is r 1 = r 2 x 1.17, and as ^ = 7.25, r^
will equal 7.25x1.17 = 8.6 inches, or outer diameter= 17.2 inches. That rule holds
good for inward flow wheels with multiple guide blades, nearly as many as wheel
blades, but for the vortex the rule is nearer r^ x 1.5=11 inches nearly, or 22 inches
diameter. Thus we have arrived at the speed in revolutions, and the diameters of
the wheel. We find the combined area of the wheel passages at the inner radius.
This is equal to A 2 = ^-; hence = * =0.5 square feet, or 72 square inches clear
V 2 26.44
opening of combined wheel passages.
The circumference at r 2 is = 45 inches. The number of blades is determined by the
circumference divided by 1.5 in inches ; hence ^- = 30 blades. And if each blade is | inch
* o
thick we get 3.75 inches as space occupied by these blades by their thickness on this
circumference of r z and the contraction due to the angle, as shown by the figure. If
we divide off FB, using F as a centre with radius FG, then the radius F/ cuts off on
FB a part equal to the opening, and FB is to Ff as the whole circumference is to
the actual opening ; so that if the distance Ff is o, 25 of FB, then the actual opening is
as .25 to i, or 45 x. 25 = ii. 25 inches. And deducting 3.75 for thickness of blades,
we get 11.25-3.75 = 7.5 inches as A 2 on one side; dividing the total area .5 square
feet by 7.5 we get -^ = 9.6 inches as the axial breadth of the blades at the inner
7-5
radius.
The guide blades, if numerous, must be calculated in the same way.
Again, the ratio between the radius r^ and r% we took as 1.5 to i, so that the breadth
of the wheel axially may be reduced outwards in that ratio. Hence, since the blades
6^
Modern Engines
are 9.6 inches broad at r v they will be <- - at 7^ = 6.4 inches. This reduction at the
outer periphery is shown in Figs. 72 and 73 very well.
The rest of the design is mechanical, such as the casing, the bearings, shaft, and
governor ; these come under other headings. The inward flow turbine is sometimes
made as shown in Fig. 77, an illustration of a common form, in which the wheel passages
are spiral and the discharge axial.
Fig. 78 shows how this turbine is conveniently fixed in the pentrough.
Another type of mixed flow turbine is well shown by Messrs. Gunther's illustration
of a large pair recently made.
We must consider the so-called mixed flow turbine, to which a large class belong.
Usually they are inward and axial flow, and can be easily placed in a timber or brick
wheel pit with the draught tube projecting through the floor.
In some of them the guide blades are made movable for regulating purposes, while
some have a revolving cage whereby the orifice can be reduced in area. The wheel,
usually of gun-metal or bronze, is shown in Fig. 77. The blades are curved and
FIG. 77. Mixed Flow Turbine.
FlG. 78. Method of Fixing Turbine.
spiral, leading down to outward flow ; curved also for downward flow blade. These
types of turbines are all difficult to regulate in speed, but where close regulation
is not of importance they are cheap to instal, and give a good efficiency when full
loaded.
The mixed flow turbine constructed by Messrs. Gunther has an axial and outward
radial flow set of blades, with a series of guide blades acting much as the large blades
in the vortex wheel. As it is a good type of wheel for a horizontal shaft and considerable
speeds at high powers, we may consider a good example of it recently made for a large
mill abroad.
The plant consists of two of the firm's latest type of mixed flow turbines, in which
the water enters on the outer circumference of the wheel, and is discharged partly on
the inner circumference and partly in an axial direction. The head of water available
is 35 feet, and each turbine has been designed to produce 150 horse-power with this
head at a speed of 180 revolutions per minute. Plate I. shows the complete plant
mounted on its bed-plate, consisting of two cast-iron beams, which, when everything
is bolted up, form what is practically a continuous bed-plate. It will be seen that the
two turbines are placed back to back with their shafts in alignment, and that there
Mixed Flow Turbines
is a rope pulley mounted between bearings on a small shaft in a line with the other two
shafts. It may be connected to either or both turbines by means of two claw clutches
worked by levers, as may be seen in the illustration.
A question which readily suggests itself when looking at the combined plant is,
Why were two turbines used ? In these days a 300 horse-power turbine is com-
paratively small, and size would have been no obstacle ; moreover, the expense of a
double set of piping would have been saved. In this particular instance, however, it
was deemed preferable to have two instead of one, and for two principle reasons. The
first of these had reference to the weight of any single portion of the machinery. Owing
to the locality for which the plant was destined it was desirable to keep down weights
as much as possible. Then, too, one turbine can be used alone when the full output of
300 horse-power is not required.
The rope pulley is 5 feet 3 inches in diameter, and it is grooved for eighteen
ropes, each i inch in diameter. The turbine wheels are 36 inches in diameter, and
the general construction may be seen in Fig. 79, which shows a section through
the turbine case, and a section of the guide and wheel vanes, showing the posi-
tion of the former when closed and when fully open. The casings are of mild steel
plates with cast-iron flanges riveted in, and each turbine is provided with a manhole
GUIDE BLADES
OPEN
Sectional Elevation.
Section of Guide Blades and Wheel Vanes.
FlG. 79. Large Mixed Flow Turbine.
for examination of the interior. The water inlets are 42 inches in diameter, and the
suction bends 36 inches in diameter. The wheel of either turbine can be taken out by
removing the suction bend and the cast-iron end to which it is bolted, which does not
necessitate touching anything on the pulley side. The end pressure of the shaft is said
to be very slight, and to take it up a lignum vitae footstep is provided. This is placed
in a small chamber, which may be seen bolted to the outside of the suction bend. It is
adjustable for wearing by means of the set screw at the end of the chamber, and it is
lubricated by the pressure water taken from the turbine case by means of a small pipe,
which may be seen in the engraving. This lubrication is so arranged that the pressure
water passes up the centre of the footstep and through a groove across its diameter.
It is said that this construction ensures that the whole rubbing surface is efficiently
lubricated. As the pressure water might contain grit or sand, measures are taken to
prevent this getting into the bearing.
Regulation for speed or power is, as may be gathered from Fig. 79, brought about
by movement of the guide blades. The guide blades are pivoted, and are opened or
closed simultaneously by the rotation of a ring carrying a number of equally placed pins,
one of each fits into a slot in each guide plate. Quite a small motion, or travel of the
ring, suffices to fully open or fully close the blades. By properly proportioning the
guide blades, and placing the pins on which they work, matters are so arranged that
64
Modern Engines
the blades are practically in working balance against the pressure ot the water, and
that consequently the gates are easily and quickly opened. As a result close governing
is obtained. The ring which opens or closes the guide blades is worked by two
FlG. 80. Method of Fixing Large Jonval Turbines for Low Falls.
cranks placed 180 apart and coupled by levers, the desired motion being imparted
by the screw revolution of a screwed spindle.
The spindle actuating the regulating levers is provided with a coupling at its
upper extremity. This is taken to an upper floor over the turbine house, and there
connected to the regulating pillar and governors.
Pressure Turbines 65
PRESSURE TURBINES
These are inward or outward flow pressure turbines usually, but are also made for
axial flow.
This is the type ot wheel used at Niagara Falls in the large power scheme, work-
ing without a draught tube, and they are outward flow. We shall refer to them
again.
But before the Niagara scheme there was a paper mill in Niagara with a wheel-pit
167 feet deep and 28 x 43 feet in section, with a vertical supply pipe 13^ feet in diameter,
working three inverted Jonval turbines of 1 100 horse-power each. By inverted is meant
that the guide blades were below and the wheel above, so that the upward water
pressure supported the weight of the wheel and vertical shaft, the shaft being 10 inches
diameter.
Niagara Falls are on that immense chain of inland lakes on the boundary between
Canada and the United States ; these lakes are natural reservoirs for the rainfall over
240,000 square miles. The overflow discharges at a constant rate nearly, through the
river St. Lawrence into the sea. Between Lake Erie and Lake Ontario the water flows
through the river Niagara with a fall of 326 feet in 36 miles, the quantity of flow is about
300,000 cubic feet per second ; such a vast quantity at such a high fall would, if it could
be utilised, give up about 7,000,000
horse-power night and day.
In all these large powers some
modification of the pressure wheel is
used with a vertical shaft.
The arrangement for setting large
Jonval turbines in masonry for low
falls is shown in Fig. 80. The water
enters from the left and first passes
through a screen, to intercept float- 'jJ^^JipaM^-^ .'.
ing rubbish ; it then passes the sluice
valve into the wheel pit and falls FIG. 81. Jonval Turbine with Horizontal Shaft,
through the wheel into the tail race.
And Fig. 81 shows an axial turbine of this class with a horizontal shaft made in a
convenient form for coupling to a draught tube. Again, for an inward flow turbine,
let
Q = cubic feet of water available per second.
h = effective head of water.
V = velocity of water due to fall in feet per second.
V 2 = ,, ,, across clearance between guide blades and wheel in
feet per second.
C = velocity of wheel blades at entrance of water at r r
C 2 = ,, outlet r v
A x = sectional area of jet at discharge of guide blades.
A 2 = ,, ,, wheel passages.
A 3 = ,, ,, tube or exhaust.
R = revolutions of shaft.
r^ = inner radius of wheel.
7- 2 = outer radius of wheel.
N = number of wheel blades.
N^= ,, guide blades.
Empirical rules give the following data :
-i=.
r 2 1.17'
VOL. I. 5
66 Modern Engines
2 = 0.25. V 2 = o. 5 7x8x Jh.
A 3
C 2 = o. 9 6xV 2 . A 2 =Q
V 2
Assume we wish to calculate a wheel for 100 feet head effectual. Quantity of
water Q = SO cubic feet per second ; speed of wheel, 10 per second. We shall proceed
to find the dimensions of the wheel. Thus from V 2 = o.57 x 8 ^ = 0.57 x 8 ^100 =
4. 56 x 10 = 45.6 feet per second. And C 2 the velocity of the wheel at r its outer periphery
in an inward flow turbine = C 2 = 0.96 x V 2 = 0.96 x 45.6 = 43.7 feet per second.
Q
From this and the speed 10 revolution per second we get r-^ ; for -^ will give us the
R
circumference in feet, from which r-^ can easily be found. Thus ^^-^ = 4.37 feet circum-
ference =4.37 x 12 = 52. 44 inches, equals a diameter of 16 inches; .'. r-> = = 8 inches, and
2
g
= 6.8 inches, and 8 inches for outer ^ in an inward flow wheel.
1.17
The combined area of the guide blade passages = -- = =1.15 square feet
V 2 43-7
= 165 square inches = A 1 .
From this and the number of blades and their thickness the axial depth of the
wheel can be calculated, for we know its circumference at r r But besides the thickness
of the blades and their number we want to know their contraction of passages by their
obliquity to the radial line ; this can be calculated if we know the angles. The angles
will be, however, better drawn to instructions given in Fig. 760 (p. 60) already, and the
actual opening F G measured ; before drawing it we must know the number of blades.
These should be numerous and thin. If we take a safe estimate from practice of making
them about 1.5 inches from the centre of one blade to the next on the circumference, this
circumference is 52 inches; hence -^ = 35 blades. Each blade might be ^-inch thick ;
1 '5
hence the lot will take up x 35 about 7^ inches. This amount must be deducted from
the circumference.
We must find the openings on the diagram, Fig. 760, F G, and suppose we find
them each 0.25 of what they would be if the blades were truly radial, that is
to say, the openings would be only J of what they would be with radial blades,
of 52 = 13 inches, and this - 4^, the thickness of blades = 8 inches as the one
side of the area of the outflow. A = 165 square inches ; hence -5 = 20 inches, nearly the
8 -5
axial depth of the wheel. We have now found
Diameter of wheel outside 16 inches.
,, ,, inside 13.6 inches.
Area of guide blade passages =165 square inches, and the number and thickness of
blades from which we find the length axially of the wheel.
If a draught tube is used its area should = A X x 1.25.
On this basis there would be some pressure, but very little where the water
jumps from guide blades to wheel blades, for in an inward flow turbine it is difficult
to get the r^ end of the opening larger than the guide blade openings ; they must be
smaller.
In the axial flow Jonval turbine the radius t is the mean between the outer and
inner circle filled by the wheel passages. Here the passages A p A 2 , A 3 are easily found,
A A
thus A x = 2 -, and A 3 = 2- ; so that if A x = 1.15 feet, A 2 would = 1.15 x 1.15= 1.8 square
O *
feet. And A 3 =i.8x4 = 7.2 square feet opening.
Turbine Vanes
PRESSURE
10
TURBINE VANES
In the foregoing 1 calculations values and constants have been taken to show the
working of elementary problems in turbines. The subject, however, can be treated with
mathematical precision by geometrical, graphic, and algebraic figures and formulae to
be found in classic works. While the conclusions to be arrived at by these means are
sure guides as to the practical designing and construction, they cannot be realised
entirely.
In dealing with guide bladed turbines, in which the energy is delivered mostly
by momentum, we speak of velocities as "absolute" and "relative." To explain
the meaning of the terms we may refer to the Pelton jet and wheel (Fig. 50). The
nozzle is fixed, and the jet issues at a velocity due to pressure. This velocity is absolute ;
it is the velocity with which a particle in the jet moves away from the nozzle to the
vane; "relative" velocity means the velocity of something else compared with this
absolute velocity. The vane, for instance, might have no resistance to offer to the
jet, in which case it would be carried forward with the jet at the same velocity, and
the relative velocity ot the vane would be the same as the absolute velocity. Again,
if the vane were fixed there would be no relative velocity between it and the jet. In the
first case the jet would not split and flow back out of the vanes ; the vanes would be
carried along like a shield on the advancing end of the jet; being unresisted, no work could
be got from it. In the second case the water would
split, and, if we neglected friction, would return with
same velocity, exerting a pressure only on the vane.
Now, between these extreme cases we have relative
velocities at which we obtain motion of the vane
against resistance, and there is a certain velocity
of the vane compared with that of the jet from
which we obtain the greatest amount of energy, and
that is a relative velocity midway between the two
extremes which we have considered, that is, when
the velocity of the vane is half the velocity of the
jet. When the vane had no resistance against
the jet its velocity was equal to that of the jet, and therefore no pressure could be
obtained upon it. When the vane was at rest full pressure was obtained, but no
velocity. If now the vane is allowed to move at half the jet velocity the maximum
work done against the resistance which controls the relative velocity for the pres-
sure x the vane velocity = the power obtained. As V the relative velocity of the jet
increases, P the pressure upon it decreases ; and as P the pressure on the vane
increases (by restraining the vane to a lower V), V decreases. Take a rectangular
parallelogram (Fig. 82), divide the top line to represent parts of pressure on the vane,
and the bottom line to represent parts of velocity, numbered from right to left for
pressure, and from left to right for velocity ; a perpendicular line anywhere between
the two ends will cut the area into two parts corresponding to V and P respectively in
magnitude, so that if we increase the one area we decrease the other in the same
amount. Suppose we divide it at 5 on bottom line, then P=i$ and V = 5; hence
P x V = 75. Divide it at half, that is, make P = V, then 10 x 10= 100. Any other division
than half will give a less result ; hence in a Pelton wheel the relative velocity is half
the absolute velocity for best effect.
This is due to the jet striking squarely on the vane in the direct line ot motion, and
the curves deflecting the split jet to clear the incoming jet and effect the pressure. The
jets leaving the vane have no relative velocity, for the vane is receding from them in
the opposite direction to that which they flow in.
But the case is different if the jet strikes the vane at an angle and leaves it at
to
VELOcrrr
FIG. 82. Diagram showing Relation
between Pressure and Velocity.
68
Modern Engines
an angle. The figure representing the effects is then a rhomboid, with the sides
drawn to scale showing the value of velocities.
Suppose in Fig. 83 a part of a turbine
wheel. The jet enters along line V, its
length representing its velocity.
We must resort to the resolution of
forces by the parallelogram of forces.
Suppose for illustration we take a
_ model to illustrate the motions of a body like
an inclined plane under a force acting at an
angle. It is not sufficient in practice merely
to represent the direction and magnitude of
forces by lines and angles ; it is better to see
the effects first, and take the line and angle
representation afterwards. Place a wedge
with an angle of 60, say, at A on a plane
movable on rollers (Fig. 84), and have a
guided sliding rod with a roller on the end, so that it can move easily up or down,
but not sideways. Now, if we press with a force on this rod downwards there will be a
certain amount of the force transmitted
downwards, but owing to the incline
some force will act from left to right,
and the plane will run forward in the
direction of the arrow, while the rod
will move downward. And if the height
of the wedge P from A to B is double
the length CB, as it is, then the plane
F would move twice as fast as the -
rod W.
Now, if rod W is replaced by a
FIG. 83. Diagram of Forces acting on
Turbine Blades.
FlG. 84. Experiment with Forces acting on
Inclined Plane.
water jet, CA may be considered the
vane of a turbine ; we would get the
same effect, only the water would rebound from the plane.
Referring to the Fig. 83, the water travels along line V; then, if we draw a line QB
equal in length to a scale representing the
velocity V l of the vanes, then the relative
velocity of the jet to the turbine will be
= QB. Drawing this from the end of
V parallel with the turbine motion, and
completing the parallelogram APBQ,
PB will give the relative velocity of the
jet. The absolute velocities when enter-
ing and leaving the vanes are the import-
ant points, not the velocity of flow through
the vanes, as the energy given up de-
pends upon the square of these velocities,
and the efficiency would be
d
FIG. 85. Diagram of Forces acting on
Turbine Blades.
The velocity v must be reduced to the
smallest value, and this depends on the
velocity V x of the vanes and on the angle of entrance of the vanes and the angles of exit ;
this also can be shown by the parallelogram of forces. Fig. 85, AB represents, again,
the direction and velocity of motion of the jet ; QB represents V lt the velocity of the
Turbine Vanes
6 9
F G
FIG. 86. Diagram to find Curvature of Guide Blades.
turbine blade in direction and magnitude ; then join AQ by a line, this represents the
velocity of the jet relatively to the turbine = R. Constructing the parallelogram QBed,
with angle a and ft equal, we get Qe, the magnitude and direction of z>, the issuing
velocity of the jet at the exit. From this figure it is easy to prove that the smaller
the angles a and ft the smaller v would be, and that if the jet is completely reversed in
direction v = ^V. At best effect, and whatever these angles may be, this value is never
less than ^ V, but greater the larger a is ; but even with angles a and ft = 45 the lost
effect is only 17 per cent., and in practice this angle varies between 15 and 22^, so
that the actual effect of a variation in the angles has not great influence.
To draw the vanes of a parallel flow guide, draw two lines parallel to represent
the entering face and the exit face
of the guides to any given scale
(Fig. 86). Take FG = the pitch, and
make the angle DFG = to a ; from
G draw a line GDA perpendicular to
FD.
With centre A, radius AD, de-
scribe the curve DB ; proceed with
the others in the same way.
The water is directed in the
proper direction as it enters and gains velocity in the narrowing passage of the guide
blades.
For the blades of inward flow turbines we have already given instructions on p. 61.
For outward and inward flow wheels and all wheels a different construction is required
for impulse and reaction wheels, as already described.
Fig. 87 represents a method for setting out the wheel blade curves. Draw two
lines representing the entering and exit forces of the wheel, and divide off the pitch of
the blades FG. And as before for the guides, raise a line from G perpendicular to FE,
having made the angle of exit GFE = B. At an angle of 45, from point E draw line
B EC cutting AD at C ; make the
A 7" ---,._ e q_ triangle CBE equilateral ; B will
then be the centre of the curve from
C to E.
The designing of turbines, like the
designing of all prime movers, is not
altogether controlled by theoretical
considerations. These are admirable
guides, but practice and experience
alone can produce good designs from
these theoretical considerations.
Turbine makers and designers keep their experience among themselves, and have
their own methods of working out details. The designer should keep in view the
Pelton wheel as an elementary impulse turbine, also the Hero turbine, as the elementary
pressure or reaction turbine as shown in Fig. 39, and also as shown in Fig. 40 drawn
as an impulse turbine.
The form of blades shown in Fig. 34 of a Girard wheel is more of a curve than
those seemingly dictated by theory, yet they are quite efficient.
We may now turn to the question of governing water turbines. This is no
ordinary problem. The chief difficulties arise from the inertia of the water, its weight
and momentum. To check an increase of speed due to fall of load, the water flow must
be checked. If this is done suddenly great pressure results, and something may burst
to prevent such an accident. Air vessels may be used on the mains to give some
elasticity. Again, if the water requires acceleration to increase the speed when it falls
F G
FIG. 87. Diagram to find Curvature of Wheel Blades.
Modern Engines
due to increase of load, time is required to increase its flow. Early governors were
mostly designed of great power and magnitude, in order to operate valves or sluices
on the main supply ; these, being large and subject to great pressure, were very sluggish
and useless for any work requiring steady driving.
In many cases, however, the speed was wonderfully steady, despite the poor
governor ; this was, however, due to the fact that the working load, which could be
varied to any extent, was a small fraction of the total load. The mill gearing is usually
machinery of a constant load, being in most cases about 75 per cent, of the total, so that
if loo horse-power turbine were in use its load could not have a maximum variation of
more than 25 horse-power out of 100, and this could only occur in the most unlikely
event of all the power on the variable load happening to go off at once, probably the
variation in load on a turbine working a mill, through belts, pulleys, spur wheels, and
shafting, never exceed 10 per cent, of the total. That being so, the governor in most
cases was more ornamental than useful, and really acted more as a safeguard against
dangerous racing in the event of a main belt or shaft failing.
All that, however, has been changed by electrical transmission of power, in which
the variable load is 75 per cent, of the whole, and the constant load correspondingly
small, and it is quite possible to throw off all the variable load at once by the opening
of a switch or cut-out. Hence the governor must be made to govern the turbine speed
promptly, and within a small variation. The known governors are not many, but
present-day forms of them are fairly satisfactory, and close enough governing can be
obtained by the best of them. Some turbines are easier governed than others. We
have referred to the Pelton wheel governor and the vortex governor already in con-
nection with these wheels.
TURBINE GOVERNORS
The question of turbine governing when constant speed under varying loads is one
of much importance, and presents some difficulties when the load is liable to sudden and
large fluctuations; especially is this the case when driving electrical generators for
constant electric pressures. The same difficulty arose with steam and gas engines
in early days, for until electrical engineering became a commercial success the govern-
ing of engines was not understood. The marine engine, the locomotive engine, the
factory engine, traction engine, and others got along very well without any particularly
good governing. In most cases of power users the power required was pretty constant,
for the gear used to transmit it generally took from 40 to 70 per cent, of the maximum
all the time. The marine engine cannot run above its maximum speed unless the
propeller is lost or the propeller shaft broken, or in storms, when the propeller may be
lifted out of the water. The locomotive is governed by the driver of the engine.
Steam, gas, and water powers are governed by centrifugal pendulum governors
as a rule. This type of governor, originated by James Watt, is capable of considerable
variation in design.
For water turbines, wherein the valves to be moved, or sluices to be shut, or blades
altered in angle, a hydraulic relay is necessary. The old attempts to govern by powerful
governors are all failures ; mechanical relays also are failures. A light sensitive
governor controlling a powerful cylinder and piston which operates the valves is the
only solution of the problem at all satisfactory.
The centrifugal governors we shall treat later on fully under " Engine Mechanism."
Meanwhile, the points of interest are the hydraulic principles used in the governors to
operate the sluices or gates of the turbine.
In some old style turbines a large pendulum governor was used to throw in and out
two clutches to gear into the sluices from the turbine shaft so as to open or shut them.
But such governing always occupies too much time, and hence oscillations in speed
Turbine Governors
occur at every change of load. If the load is increased the speed decreases, and the
governor engages the clutch which opens the sluices ; but it takes time to open the
sluice, so that the speed does not at once cease to fall, and the opening goes on until
the speed has increased somewhat ; it opens too much, so that the speed rises enough to
throw in the other clutch to close the sluices ; this closing is again overdone, and so
the see-saw in speed goes on, gradually settling. A heavy
fly-wheel modifies these oscillations.
The relay hydraulic governor shown in Fig. 88 was
one of the first successful ones. A piston valve A with a
free passage through the centre and an annular passage
acts as a valve, admitting water to either one or other of
two piston heads of different areas. On the smaller piston
below there is a constant water pressure, tending to raise
the piston if the water above the large piston is open to
discharge through A and the overflow ; hence if the governor
raises the valve from the position shown the water above
the large piston will escape, and the pressure on the small
piston will raise it until the valve closes in the higher position.
Again, suppose
the valve dropped
from the position
shown, the pres-
sure would be
admitted to the
larger pistonhead,
so that the piston
would drop as far
as the valve had
dropped, and
would be locked
in the new posi-
tion.
The movement
of the piston
follows, and tends
to close the port
at whatever posi-
tion the governor
shifts the valve ;
hence the move-
ments of the small
light valve are
followed by the
powerful pistons
as if the pistons
were directly con-
nected to the
governor. The action is prompt and positive. The figure shows this governor in its
elementary form, and as so constituted, is all that is required for sluices requiring
small amounts of movement ; for long movements the governor is connected to the
valve rod by a lever, the fulcrum of which rises and falls with the piston.
In Murray's governor (p. 57) the valve and piston are separate, so that a long stroke
can be made with a very short movement of the governor valve.
J
I
r
i . VI I
FIG. 88. Relay Hydraulic Governor.
FIG. 89. Willans' Governor.
Modern Engines
Willans' governor can be actuated by an electric solenoid operated from a dynamo
when the turbine is used for electrical generating. The arrangement is shown in
Fig. 89, in which the governor valve is inside the piston, which acts in the same way
as described for Fig. 88.
Many attempts at solenoid governing have been made, and Willans contains the
elements of success first achieved. Willans, instead of actuating the throttle valve or
expansion valve directly by the electromagnet or solenoid, employs the latter to actuate
a small supplementary valve, which is almost frictionless, and this in its turn controls
the supply or discharge of water, steam, or other fluid pressure to a cylinder in which a
piston works, which actuates the throttle valve or expansion gear of the engine. In
this way, although absorbing a power less than half that required for one 2o-candle
lamp, the solenoid is able to control the most powerful expansion gear.
The Willans electric governor is shown in Fig. 89, where S is a solenoid taking
the place, in incandescence lighting, of one of the lamps. In other words, the solenoid
is on a branch between the main wires. The core C of the solenoid is suspended by a
spring, and this spring is attached at the top to an adjusting screw used for regulating
FlG. 90. Electric Solenoid Governor.
the light. The other end of the core is connected with a small piston valve working
inside the main piston W, which latter piston controls the throttle valve in the casing T.
Water or other fluid pressure is admitted by the pipe P into an annular chamber
surrounding the water piston W, and also, by means of a suitable passage X, into an
annular space between the two small pistons which form the piston valve. The water,
after actuating the piston W, escapes through the pipe E, and by means of a small piece
of flexible pipe not shown.
The foregoing is a description from Engineering at the date ot the patent. The
solenoid, however, is not sensitive enough for small variations in electric pressure. To
work either this relay valve or Murray's valve, for electric governing, an electric relay
is required. And for this purpose the author can recommend a carbon plate rheostatic
relay, in which the resistance in a coil of a solenoid and plunger is varied by the pressure
caused by a shunt current.
The best form is a carbon pile rheostat, such as that first employed by Brush on his
arc dynamos, and shown in Fig. 90. There are four columns of carbon plates in series.
When the turbine and dynamos are at rest the plates rest loosely on each other and on
the metal blocks B on the lever L, so that the resistance is high and little current passes,
in the solenoid, and the core of the solenoid is at its lowest point, the valve being full
Turbine Governors
73
open for the water. But if we start the machine it increases in speed and the magnet puts
on pressure on the carbons, and the solenoid pulls upon the valve lever and begins to close
the water inlet, so that a fixed speed is soon attained where the forces balance. The
FIG. 91. Centrifugal Governor. Sectional Elevation.
weights on the lever and the tension on the spring K are so adjusted that this resistance
comes into play near the fixed speed. This device gives power with sensitiveness ; for
when properly adjusted the
slight variation in pressure
on the carbons causes a large
difference in current in the
regulating solenoid.
In the diagram the regu-
lator solenoid is in series with
the carbon resistance, and
the pulling magnet M in
parallel ; but it is more sen-
sitive when all three are in
series the solenoid, the
magnet, and the resistances.
It is shown connected to a
Murray valve, but may be
used with other valves, mov-
ing freely.
For controlling the large
inward flow turbines, gov-
erned by movable guide
blades like that shown in
Messrs. Gunther's diagram
(p. 63), wherein the ring
carrying the guide blades has to be turned through a few degrees to open or close
the guide passages.
M. Rateau, in his Turbo Machinery, describes the combination of sensitive
powerful centrifugal governor, with a differential hydraulic ram for working the
guide blades. Fig. 91 is a longitudinal section, Fig. 92 a cross section. The ram is
FlG. 92. Centrifugal Governor. Cross Section.
74 Modern Engines
jointed to the long 1 lever G on the controlling 1 shaft of the turbine. The cylinder D
is always connected to the water pressure, and C is put into connection with the pressure
or exhaust by the action of the governor. When connected to the pressure the ram
moves to the right, the ram in C being the greater in area. When connected to the
exhaust the ram moves to the left.
The valve operated by the governor is shown in Fig. 93. It is fixed to the inlet of
cylinder D. A regulating needle AB is placed to regulate the exhaust outflow. The
valve is worked by K from the lever H attached to the governor under T.
As the lower orifice of the cylinder D is connected to the water under pressure,
while the upper orifice is in connection with the exhaust to the atmosphere, and the
space between these two orifices is in communication with the large cylinder C, we
see that when the index of the governor rises and the valve S is lowered the upper
orifice is closed and the water under pressure flows into the cylinder C by the lower
orifice ; and when, on the other hand, the index descends, the valve S rises and opens
fully the upper orifice, which has a section almost double that of the lower, so that the
pressure falls in C and the water escapes from it. If, now, the valve is midway between
these two positions, so that the two orifices present the same section to the water, the
pressure in the cylinder C remains about midway between the initial pressure of
the water, and the pressures on the differential piston AB are balanced so that it does
not move. There is, however, this disadvantage, that there is an almost continual flow
of water through m. On the other hand, there are two great advantages : first, the
small valve S is only rarely forced to the end of its stroke by the governor ; generally it
is only partially moved, and from this results a speed of gate opening or closing more
or less great. The point of rotation d of the lever H is not really fixed ; it is connected
to the piston valve in such a manner that it rises or falls proportionally to the motion of
this piston. It results from this, that when the piston is at rest, and consequently the
valve S occupies its mean position as well as the point e, the index i of the governor is
obliged to fix itself in the position corresponding to that of the point d, and, reciprocally,
the point d and the piston valve are obliged to follow the movement of the index.
This device of making the fulcrum rise and fall with the movement of the governor
is an old one, and was used in the Willans' governor in 1884, and we have already
explained that it is used on the governor in Fig. 88 (p. 71).
On the whole question of water turbine governors it may be taken that all the best
makers have adopted effective hydraulic relay governing, and that close governing is
obtained under all loads ; but the load should not be varied suddenly to any large
extent, not because the governor would not act, but to avoid the effects of the inertia
of the flowing water in the pipes and valves.
In order to prevent the bursting of the pipes when the flow is suddenly checked by
the governor, ample relief valves are often provided. These are shown distinctly in
Messrs. Gunther's illustrations.
Some turbines, notably the inward flow types, are more stable than others, due to
the flow being opposite in direction to the centrifugal force on the whirling water.
In cases where water is not abundant and sometimes small in quantity it is better
to use several turbines, and to run the number sufficient to use the available water at
any time, than to use one large turbine and work it at ^, |, or f , or whole gate ; but
where economy of water is no object the large turbine is cheaper in first cost and
maintenance.
Other makers of governors exist, but it is difficult to get exact information of
their design ; but all of them working with centrifugal governors are on the relay
principles, as described herein.
Gunther's turbines are operated by separate governor valve and piston, same as
Murray's governor. This arrangement is shown in Messrs. Gunther's improved Pelton
wheel turbine or high pressure impulse wheel, shown in Fig. 94.
Turbine Governors
75
That their governor is a high-class one may be seen from the following' test results
at an installation with a Gunther turbine and governor :
Turbine at Landore Hotel, Keswick. This is a Girard turbine fitted with a Gunther
governor as described. Fall, 300 feet ; horse-power, 15; wheel, 15" internal diameter;
speed, 850 revolutions ; coupled direct to multipolar dynamo of 75 amperes at 130 volts.
Used for lighting the hotel and charging an electric launch running on Derwentwater
Lake.
The lights are driven direct without cells, and the cables run down to the pier and
charge direct into the cells on board the launch as required. The following are the
results of actual trials with the governor: No perceptible variation of speed was
noticeable on the tachometer with sudden load changes of 10 per cent. ; with a sudden
throwing off of one-third the total load by throwing out a switch, the speed rose
temporarily only about 20 revolutions in 850 revolutions, equal to z\ per cent., and then
came back to normal. The turbine is usually left running with the governor all through
the night, the turbine house being locked up at 11.30 p.m., and
the attendant coming back at about 7.30 a.m.
This plant is shown from a photo in Plate II., wherein the
governor exterior can be seen.
Much larger turbines of the same design are shown. In
Plate III. is shown a view of high pressure type of impulse
turbine with single jet, the illustration being from a photo of
one of a set of turbines recently made for the Indian Govern-
ment for utilising the Karteri Falls in the Nilghiri Hills, the
power being transmitted electrically to the Coonoor Cordite
Factory.
The plant comprises four of these turbines, each capable
of driving 250 B.H.P. with a fall of 620 feet, and two similar
but smaller turbines of 37 B.H.P. each, for driving the exciter
dynamos. The large turbines are coupled direct to three phase
generators of 5500 volts, and the smaller ones are also coupled
direct, flexible couplings being used to couple the turbines and
dynamos in each instance ; and fly-wheels are also placed on
the turbine shafts to obtain the utmost steadiness under load
changes.
In Fig. 94 we give a diagrammatic outline of the tur-
bine, showing the port and wheel vanes in section. It will
be seen that the water enters on the outer circumference of
FIG. 93. Valve of Centri-
fugal Governor.
the wheel, and that the discharge is towards the centre, but in reality the greater part
of the discharge is at the sides. As in the other view, the buckets or vanes are
cupshaped, so that the water discharges partly towards the centre and partly at the
sides. The wheel vanes are cast in bronze in a continuous ring, which is bolted
to a central cast-iron plate. The guide port is provided with a sliding hood or
shed, by which the area of the discharge is varied according to the power required.
The lower part of the shield where it cuts off the water being properly formed so
as to make it a prolongation of the guide pert and preserve the true shape of the
issuing jet, so that the efficiency is practically the same whether the port is full or only
part open.
The four large wheels are 54 inches diameter and make 400 revolutions per minute,
and the two small ones 27 inches diameter and make 800 revolutions per minute.
Each turbine is provided with Messrs. W. Gunther & Sons' hydraulic governors,
and also with relief valve to minimise the danger from concussion in the pipes when load
is suddenly thrown off and the supply of water checked by the closing or partial closing
of the regulating hood. In this hydraulic governor the rise and fall of the governor
Modern Engines
balls opens or closes a small valve which admits pressure water taken from the pipe
main to one side or other of a hydraulic cylinder, and thus moves the piston up or down,
FIG. 94. High Pressure Impulse Wheel and Governor.
and the motion of the piston is communicated by suitable lever or other connections to
the regulating- hood, closing 1 or opening the jet as load is thrown off or on. The action
of this governor is instantaneous, and brings the
speed back to normal at all loads, that is, the
speed of the turbine is kept the same at all loads
between no' load and full load.
Referring to the diagram Fig. 94, A is the pres-
sure pipe, which takes the water from the main
from a branch on the turbine inlet, a filter being
affixed to the branch so that the water is filtered
to remove injurious particles before it passes to
the governor. The pipe A (shown as a dotted
line) conducts the water through a special valve
(which is used also for stopping and starting the
turbine hydraulically by means of the handle B).
From this valve the water passes through a piston
valve contained in the valve box C, which piston
valve is worked from the governor balls, and dis-
tributes the water to one or the other end of the
cylinder F through the pipes D or E. By means
of the valve at B the turbine can be started and
stopped ; or if desired, the governor can be thrown
out of action and the turbine kept running at any
desired fixed gate opening.
In considering water wheel governors we have
FIG. 95 Watson, Laidlaw, & Co.'s Governor.
described the American forms of Pelton wheel gov-
ernors. Recently a much improved automatic governor, by Messrs. Watson, Laidlaw,
& Co., Glasgow, has been brought out, and is described in Engineering as follows:
Turbine Governors
77
The wheel casing" with the governor in position is shown in Fig. 95, whilst Fig. 96
represents a section through the jet nozzle.
The rod on core A is extended behind, and enlarged in diameter, to form the plunger
B, which fits closely into a liner within the body of the nozzle. The helical spring inside
the plunger exerts a constant pressure upon it, and tends to thrust the rod forward
into the water jet, while at the same time the pressure of the water which enters at D
tends to push the plunger back with greater force. The small holes E in the end of the
plunger permit a certain quantity of water to pass into the chamber behind the plunger ;
and if the valve F, which communicates with the exhaust chamber G, be closed, the
pressure behind the plunger will rise until it equals that in front, leaving the spring free
to act without opposition. It is therefore apparent that the full closing of the valve F
has the effect of reducing the water gate to its minimum
by allowing- the spring to act freely ; while the contrary
effect will result from the full opening of the valve, owing
to the effect of the water pressure on the plunger being
greater than the power of the spring.
Partial opening of the valve F will produce water
gates intermediate between the maximum and minimum,
owing to the tapered form of the rod or core A. Full
control of the wheel is therefore effected by the closing
and opening of the valve F being- made dependent upon FlG -
the higher or lower speed at which the wheel runs. This
result is attained through the operation of the governor
shown on Fig-. 97. The weights or pendulums H
are attached elastically to
the disc J, which is fixed to
the wheel spindle and re-
volves with it. They are
held together by springs of
suitable strength, to keep
them in position till a speed
above normal has been
reached. Thereupon they
fly out, engaging on their
leather - covered surfaces
with the inner rim of the
loose pulley K, which they
tend to pull round with
them. The boss of this
pulley has, however, a chain
tion of Jet
Nozzle of
Watson,
Laidlaw, &
Co.'s Gov-
ernor.
Elevation.
Cross Section.
FlG. 97. Watson, Laidlaw, & Co.'s Governor.
wound round it, attached to the spring, and the friction of the pendulums H has to over-
come the tension of this spring before they can cause the loose pulley to turn. The loose
pulley is prevented from revolving through more than one revolution by the pin L ; but
at all speeds materially over normal the action of the pendulum will tend to draw the
pulley over through an arc of a circle corresponding to the exact amount of the excess
speed. It only remains to communicate this action to the spindle of the valve F in order
to make the size of the water jet dependent upon the speed of the wheel. This is done
by leading the chain over a wheel on the spindle of the valve, as shown in Fig. 95.
The governor is very sensitive, and can be adjusted to control the speed of the wheel
within any required degree of variation of speed under great and sudden changes of
load. It will be seen that this governor can be made to govern in two distinct ways.
It can be so set as to govern the jet by regulating the flow of water proportionally to the
load, or it can govern by opening the jet full and closing it quite shut, so that the time
78 Modern Engines
during" which water plays on the wheel is proportionate to the load. In either case the
total amount of water discharged from the nozzle is proportionate to the work done.
As the Pelton wheel is specially suitable for high speeds and falls and in electrical
engineering-, an automatic, sensitive, and efficient governor is of utmost importance ; but
the difficulties which might arise from the hydraulic ram effect of opening and closing the
jet entirely and quickly have not been referred to, although some means of meeting- will
be required.
Turbine construction in Great Britain is fairly well represented by these examples,
and in recent years it has received a very considerable impulse from the perfection of
electric transmission of power, enabling the power to be carried to the works at a more
suitable place than where the waterfall exists.
It is to the engineer not much short of a public wrong that the greed of landowners
should prohibit in many cases the use of large powers given freely and gratis by Nature
in the shape of waterfalls.
A source of power running to waste is nowadays a wilful waste, and many such
exist where the power could be obtained and utilised at a distance without in any way
damaging the land or the amenities of the neighbourhood.
The case was different when the mill or factory had to be built adjoining the fall ;
whereas now a distance up to 5 miles is not an obstacle to the use of the power.
And again, in hilly countries Ireland, Scotland, North-West England, and parts of
Wales waterfalls can be produced by the engineer, in many cases at a cost well worth
the outlay, if the land could be obtained at its true value. In most instances, however,
any such proposals have the instant effect of converting the land from a worthless moor
or bog into most valuable property at a value prohibitive. A high fall with plenty of
water is a valuable gift of Nature to the landowner, and so is the side of a hill where an
engineer can form a reservoir and catch water. But as things are, they are of little use
or value to the human race or the inhabitants of the country.
FLUID-ON-FLUID PRESSURE ENGINES
These might be classed among pumps, as they are usually employed for lifting
water, but we wish to keep the hydraulic and pneumatic machinery in one book.
Many machines are retained in use by engineers although they are well known to be
inefficient ; this is only tolerable on account of great simplicity, reliability, or portability,
rendering them preferable under the circumstances to a more efficiently complex or
cumbrous machine. A piston pump properly designed is highly efficient, but it has
many delicate parts packing- glands, shafts, fly-wheel, cranks, and rods all accurately
fitted and requiring- care and attention, and only possible of construction in a well-
equipped machine shop. Water may be moved up to a higher level from a well or mine
by a waterfall, or by steam, by very simple but inefficient means. Thus in Hungarian
mines long ago water was pumped out of them to keep them dry by means of a fall of
water working- a simple arrangement of pipes and vessels, such as could be readily
obtained and set up by a country blacksmith.
The Hungarian machine was described by Professor Rankine, who investigated it
fully mathematically. It is based on the principle propounded by Hero of steam turbine
fame in the second century B.C., and is the progenitor of many modern engineering
apparatus. From Hero's description the apparatus was somewhat like the diagram Fig.
98, from which a working model may be made with two Wolffs bottles, three tubes, and
a funnel, and five or six rubber corks. To show the experiment, bottle B is filled with
water, the discharge pipe D goes to the bottom of this bottle ; F, the air pipe, just enters
bottles A and B ; K, the supply pipe, goes to the bottom of A. To start the action, a little
water is filled into the funnel, this water entering vessel A displaces the air through
pipe F into vessel B, and thus presses upon the water in B and forces it up the delivery
Hungarian Machine
79
pipe D, which pipe may be inclined so that the water falls into the funnel and maintains
the action until all the water is discharged from B. In this experiment the water is
raised from vessel B to a higher level.
Vessels A and B may be placed on different levels ; they may be on the same level,
by shortening pipe F and keeping pipe K long, as it is.
In the Hungarian machine (Fig. 98), vessel A is at the head of the pit to be drained,
B is at the bottom. In this experiment the cork V is withdrawn, and B being in a tub of
water, it fills with watef ; now shut V and allow water to fall into the funnel through K.
V x being shut, air will be forced through F
into B, and this air will force the water up
from B through D to the pit head. When
bottle B is empty bottle A will be full ; the
water is then shut off and V and V x opened,
when B will fill again and A empty, drawing
Hero's Experiment. Experimental Hungarian Machine.
FIG. 98. Hungarian Machines.
in air. V and V l are then shut, the water turned on again, and another bottle full of
water pumped up, and so on until the level of the water is reduced.
Professor Rankine's description is as follows, referring to Fig. 99 :
A is the sumpt at the bottom of the mine into which the water collects. B is a
receiver with a non-return valve C. D is the delivery pipe from the bottom of the re-
ceiver ; there should be a foot valve. G, the waste air cock, at the top of E. The dis-
charge valve at the bottom of E is for discharging the water which has performed its
work in the working barrel. K, the supply pipe, connecting reservoir with the bottom of
the working barrel E. There is an admission valve near the bottom of the supply pipe.
The valve may be opened and shut by floats in the working barrel, or by a small
8o
Modern Engines
auxiliary water pressure engine, or by a small wheel driven by the water discharged.
A single piston valve is best, as shown in Figure.
The machine is set to work by opening the air waste cock G, the supply valve at the
same time being shut. The water from the well A opens the clack C, enters and fills the
working barrel B, and drives out the air through G, so that E and F only remain filled
with air. Then G is shut, and remains shut while the machine is working ; the discharge
valve is shut and the supply from K to E opened, and the working proceeds as follows :
The driving water descends through K into E, and compresses the air contained in E
and F. The pressure so exerted
on that air is transmitted to
the water in B, and causes it
to rise in the delivery pipe D.
When the pressure has become
equal to that of the column of
water in D added to its resist-
ance, the lifted water issues
from D into the drain, and
continues to do so until E is
filled with water. Then by the
valve gearing the supply valve
is shut and the discharge
opened ; and the water in E is
made to flow out, partly by its
own weight and partly by the
pressure of the expanding air.
As soon as the air has fallen to
its original pressure more water
from the well flows through C
into B, and drives all the air
back into F and E. Then the
discharge valve is shut and the
supply opened, and the cycle of
operations recommences.
Let h Q denote the head of
water which is equivalent to i
atmosphere, or 33.9 feet on an
average.
Let h^ be the height of the
outlet of the delivery pipe D
above the surface of the water
in A ; D, the weight of a
cubic foot of water, or 62.4
Ibs. ; Q p the number of cubic feet per second to be raised ; then DQ^ is the useful
work per second.
Let h^ be the head lost by the resistance in the pipe D, hQ + h^ + h^ is the head of
water equivalent to the pressure to which the air must be compressed in E, F, and B
before the water will issue from the outlet of D. That pressure, in atmospheres, may be
& i i,
expressed thus : n = i + -i -, and the working pressure which the barrels and air
pipe must be adapted to bear is n i atmospheres.
The volume of air which must pass per second from E into B, while the water is
being forced out of B, is Q x cubic feet at the pressure of n atmospheres.
Therefore, as the original pressure of the air, before being compressed by the
FIG. 99. Hungarian Machine.
Pulsometer
81
descent of the water into E, is i atmosphere, the volume of the mass of air which
descends per second, at the original pressure, is Q = nQ l ; and this also is the volume of
water which must descend from the source per second in order to perform the work.
Let B and E be taken respectively to represent the capacities of those portions of
the pump barrel and working- barrel which are alternately filled and emptied of water
at each stroke, and let F denote the capacity of the
ar pipe
Let h.
F 4- F
then we must evidently have = n.
B + F
3 be the loss of head by the resistance of
the supply pipe, valves, etc. Then the total head
required for the fall is H = k 1 + h 2 + h 3 , so that
the total energy expended per second is DQH =
Comparing this with DQA, the efficiency is
found
Q H
factor - in
Valves of Pulsometer.
The reduction ot efficiency by the
the above, corresponding to loss of head equal to
(i--]H, is the loss of energy due to compressing
the air and water friction.
We have already referred to the Thomson jet
pump (p. 40). We may now refer to the pulsometer
pump, which is also a fluid-on-fluid pump ; steam
being directly applied to the water to force it out,
a vacuum is formed by the condensation of the steam,
which by suction draws the water up into the pump.
In this place we will only refer to it briefly to show
its principles of action ; its construction falls under
the head of Pumps.
Referring to Fig. 100, there are two side chambers
AA to receive the water alternately, and an inter-
mediate vessel H, whose purpose will be explained.
EE are suction and GG delivery valves (Fig. 100), B a
foot valve, N the delivery chamber connected to A by
short pipes FF, and Q the rising main or delivery
pipe. To start the pump, the three vessels are filled
through the hole C, the water resting on foot valve
B. The ball L being compelled to lie on one or the
other seat JJ, steam is admitted at K, and, entering,
say, the right-hand passage, displaces the water
through F, until the level falls to the upper edge of
the orifice. Steam then blows through into F with some violence, causing a partial
vacuum in A. The ball being now drawn to the right-hand seat, water rises into the
right chamber ready for the next stroke, steam enters the left chamber, and the action
is continuously repeated. The vessel H serves the purpose of an air vessel, to steady
the flow into N ; and to prevent the sudden shock caused by the rush of suction water,
air cocks DD are placed on the three vessels, and kept open. The "Grel" valve
(Fig. 101) at P is applied to economise the steam supply. It is a short hollow piston,
VOL i. 6
FIG. 100. Pulsometer.
82
Modern Engines
which rises and falls on account of the difference of pressure within and without it,
thus closing- the pipe K after a portion of the stroke has been completed ; the
remainder of the stroke being com-
pleted by steam expansion.
It is not economical by any means,
but it is nevertheless a very useful
and, in many cases, indispensable
machine.
w
The hydraulic ram next claims
FIG. ioi. Grell Valve.
FIG. 102. Hydraulic Ram.
our attention. Here we have a machine or engine not only simple, but highly efficient,
and a valuable means for providing a high pressure water supply from a low fall of water.
Hydraulic rams are divided into two
classes. In the first, the water of the
fall is employed to raise a portion of itself
to a higher level ; in the second, the fall
is used to raise some other water or fluid
to a higher level, and hence they are called
pumping 1 rams.
The principles of action are simple.
If water is allowed to flow freely through
a long pipe it acquires a certain velocity,
due to the difference in height between
the upper and lower end, and if the lower
end is suddenly closed the whole column
of moving water in the pipe, in stopping,
delivers up its momentum by pressing
violently against the stop and the sides
of the pipe ; and if we had a small valve
on the side of the pipe near the stop, the
water would be forced out with consider-
able pressure in a sudden spurt.
It is thus only necessary to have a long
pipe, with one end in the water supply
and the other end some feet below the
water level, with a stopper valve and a
delivery valve ; an air vessel is added in
order to regulate the pressure. The
stopper valve is called the pulse valve,
because it beats regularly when the
pump is in action. The construction may be gathered from Fig. 102.
Into socket A the supply pipe fits ; this pipe is called the drive pipe. B is the pulse
FIG. 103. Balanced Pulse Valve.
Hydraulic Rams
valve and D the delivery valve ; C is the air vessel and E the rising main. It will be
obvious that a great deal depends upon the weights and sizes of the valves, especially
of the pulse valve B. If it is too heavy it will not close when the water acquires its
full velocity in flowing out of it ; if it is too light, then it will close before the
water has acquired the full velocity
due to the fall. Hydraulic rams
made on this simple plan are there-
fore not easily adapted to different
falls : the pulse valve suitable for
one fall not being correct for another
fall.
Again, in order to get the full
force of the fall the pulse valve must
have a large opening to allow the
water free exit, so that it becomes
heavy. It is therefore clear that this
valve should be balanced by an
adjustable spring or weight, so that
it can readily be adjusted to obtain
the best effects under all conditions
of working. This improvement was
made by Mr. Keith, whose machine is
shown in Fig. 103. The valve B is
hung on a lever N and balanced by
a weight J, so that it can be nicely
adjusted to open and close at the
right time on any fall.
Another balanced ram is shown
in Fig. 104, an American design. The FIG. 104. Balanced Pulse Valve.
pulse valve is a piston form guided at
the lower end in a socket, which has a rubber stop to silence the beats ; it is balanced
by weight G on lever F.
In this ram there is a device for maintaining air in the air vessel. Air is absorbed
by water under pressure, and hence in time it diminishes in the air vessel, and some
means for replenishing it has to be provided. Small
valves called snifting valves have been adopted,
but they are not always effective. The air is in the
vessel under pressure, and to get any more air
into it force must be used. In this ram a long,
small bore pipe H is connected to the drive pipe A ;
it has a valve J at the top which allows air to enter,
and which closes when the pressure comes on ; the
air is then compressed and forced into the air
vessel, a small quantity at each pulse.
In Blake's hydraulic ram (Fig. 105) springs
were used instead of air. C is a piston, sliding
water-tight in a cylinder and held down by a lever
and springs ; the water entering C raises it up at
every pulse, and thus acts regularly.
Another device, by Messrs. Easton & Co., whereby air is compressed in a chamber
in connection with the drive pipe and some of it carried into the air vessel, is shown
in Fig. 106. When the pulse valve is open, and the water flows past the opening F, air
fills the small direct connected vessel C through a snifting valve D. Upon the valve B
FlG. 105. Spring-Balanced Ram.
Modern Engines
FIG. 106. Double Air Vessel Ram.
closing, some of the water forces itself into C and some into the smaller vessel F ; thus
the violence of the compression is reduced by
the air spring in F, which forces the water
back again through D into C with air ab-
sorbed during the high pressure.
We shall only refer to one form of pump-
ing ram. This machine is used for raising
clean water from a well by means of a fall
from a river or drain of unclean water.
It has some useful applications, but is no
longer a simple machine. It must be re-
membered that these hydraulic machines
are not economical, and only on account of
their simplicity can they be recommended.
When we therefore depart from a simple
construction and introduce pistons, rods,
valves, levers, and make a complicated
mechanism of it, the advantages disappear,
and we may well consider whether a turbine
and pump would not in the circumstances be
better for the purpose.
Blake's pumping ram is shown in Fig.
107. In this pump a piston C is driven up by
the recoil of the drive water. On the other
end of the piston rod is a pumping piston D ; F is the suction pipe into the well ; H, the
delivery water valve
into receiver K, and
J is the rising main.
The spring forces the
pistons back into
their lowest position
between each pulse.
In the same way
air may be com-
pressed and used in
rock drills and other
mining tools. And
great use was made
of hydraulic ram en-
gines in compressing
air for driving the
tunnel through Mont
Cenis, water being un-
limited in quantity.
In setting out a
hydraulic ram it is
necessary first to
ascertain the quantity
of water available for
the drive, and to sur-
vey the ground to ob- FIG. 107. Pumping Hydraulic Ram.
tain the position for
the ram where it may obtain the greatest fall through a pipe of at least 30 feet long. It
Hydraulic Rams
is evident that a long pipe will be required for a high lift, for the force at the moment of
closing the pulse valve is proportional to the mass of water in motion in the drive pipe
and to its rate of motion. On the other hand, it is useless to make the pipe too long ;
for while a long pipe gives a greater pressure, it beats slower than a short pipe. In
practice the length of the drive pipe is limited often by the nature of the fall and the
location of the ram, which has to be selected for the running off of the tail or waste
water. The shortest length is about five times the height of the fall, and where possible
it may be increased to ten times the height of the fall.
The table here given is one compiled by the American Engineer, gives the coefficient
for various quantities of available drive water, which, multiplied by the coefficient for
any given height to which the water is to be forced, will give the quantity delivered.
Thus if we had a 2o-foot fall to deliver 50 feet high, the quantity delivered would
be, for 100 gallons available per minute, 0.3282 x 100 = 32.82 gallons delivered per minute.
TABLE VI. ELEVATION.
bo
11
I 1
15
18
21
2 4
27
3
35
40
45
5
60
70
80
90
100
Feet.
2
.O724
OCTI
.O4.2O
.0^07
.021";
.0181
.OI 12
.0067
.OO27
3
* v / T^
1327
ooo
.1020
T^
.0807
o /
.0651
oo
0530
.0441
.0326
VW O
0243
/
.0181
0132
.0063
.0017
...
4
.1960
I 535
1234
. IO2O
.0854
.0724
.0560
.0441
.Q348
.0281
.0180
.0112
.0063
.OO27
5
.2614
.2068
.1686
.1404
.1189
.1020
.0807
.0652
0533
.0441
0307
.O2I7
.0150
.0099
.0063
6
.3282
.2614
.2146
.I800
1535
.1327
.1063
.0870
.0724
.0608
.0441
325
.0243
,0l8o
.0132
7
.3960
.3170
.26l4
.2203
.1885
. 1640
IS 2 ?
. 1096
.0920
.0782
.0580
.0441
.0340
.0264
.0205
8
.4647
3733
.3090
.2614
.2248
.1960
1595
.1327
.1121
.0960
.0724
.0560
.0441
035 1
.0281
9
5341
4303
3572
303
.2614
.2285
.1868
.1561
I 3 2 7
. 1142
.0870
.0682
545
.0441
.0360
10
.6040
.4877
.4058
345
.2984
.2614
.2145
.1800
1535
.1327
.1020
.0807
.0651
533
.0441
ii
6 745
5459
4549
3874
3357
.2947
2425
.2041
.1746
JSH
.1172
0934
.0760
.0627
.0524
12
7453
.6040
5043
.4302
3733
.3282
.2708
.2285
.1960
.1704
.1327
.1063
.0870
.0723
.0608
13
.8166
.6627
5540
4732
.4112
.3620
.2994
2532
.2177
.1896
.1483
.1194
.0983
.0821
.0604
H
.8881
.7217
.6040
.5166
4494
.3960
.3282
.2780
2395
.2090
.1640
.1327
. 1096
.0920
.0782
IS
.9600
.7809
6543
.5601
.4877
4303
3572
33
.2614
.2285
.1800
.1460
.1211
.1020
.0870
16
.8404
.7048
.6040
5263
.4647
.3863
.3282
2835
.2482
.1960
1595
.1327
.1121
.0960
17
.9001
7555
.6480
5650
4993
4157
3535
3058
.2680
2123
I73 1
.1444
.1223
.1050
18
.9600
.8064
.6921
.6040
534i
4451
3790
.3282
.2880
.2286
.1868
.1561
.1327
.1142
19
8574
7364
.6430
.5690
.4746
.4046
357
.3081
2449
.2006
.1680
.1430
.1262
20
.9086
.7808
.6823
.6040
.5042
4303
3733
.3282
.2614
2145
.I80O
'535
.1327
21
.9600
.8254
.7217
.6392
5340
.4561
.3960
.3486
.2780
.2286
.I92O
.1640
.1420
22
.8701
.7612
6745
.5640
.4820
.4188
.3688
2947
2425
.2041
.1746
1514
23
.9150
.8007
.7098
5940
.5080
.4417
.3892
3"4
.2567
.2163
1853
.1609
24
.9600
.8404
7453
.6241
534i
4657
.4097
.3282
.2708
2l8 5
.1960
.1704
The old rule for finding the diameter of the drive pipe D= \/(i-63Q), Q being the
available amount of water in cubic feet ; and if h the height to which the water is to be
lifted and H the fall, then L, the length of the drive pipe, should be L=H+^ + X2.
ri
In the case just taken, H = 20, h = 50 ; then L = 20 + 50 + 5x2 feet = 145 feet.
20
The capacity of the air vessel should be equal to the capacity of the drive pipe, and
the area of the opening of the pulse valve should be nearly the area of the drive pipe in
section ; per 100 gallons, D would be= ^(1.63 x i6) = 5 inches nearly.
The air vessels are usually cast iron, well pitched inside to fill any pores up through
which air might escape. Glass has also been used for small pumps.
Hydraulic rams require to have well made and well designed valves, and occasional
attention to see that the air is sufficient in the air vessel. In country houses they are
much used in all countries to provide a household water supply under pressure.
We now may consider other fluid-on-fluid pressure engines, in which gases act. upon
liquids and vice versa.
86 Modern Engines
STEAM JETS
The air spray is familiar to every one, and illustrates the production of a partial
vacuum produced at right angles to a jet of fluid. We have many interesting applica-
tions of this principle in engineering.
The acting part in these apparatus is a jet of air, steam, or water, which inducts
motion in a surrounding mass of fluid. In the simple cross blast jet the horizontal
jet passing across the mouth of the vertical pipe acts simply by cutting off the atmo-
spheric pressure from that mouth, and so the liquid rises in the vertical tube and is
blown away as a spray. The velocity of the jet of air, or steam, or water is sufficient
to carry it across the mouth of the tube without deflection, even although there is a
vacuum beneath.
In the same way the jet acts in nozzles. Here the velocity of the jet first carries
air out of the annular space, making a partial vacuum into which the liquid rushes ; the
liquid coming in contact with the jet is carried on also by contact.
As nozzles are the important organs of these induction machines called injectors
and ejectors working on these principles, we must consider nozzles, or hollow cones, in
various types.
A converging cone delivers a rod of fluid without contraction to the same extent
as a straight tube or through a hole in a plate.
Steam being the most important fluid at present, and more information being
collected regarding it than about other gases, we will first consider the steam jet.
Experiments have
been made with a conical
nozzle, as shown in Fig.
108. This nozzle may be
-used with either end as
the inlet and the other end
the outlet.
With the wide end as
FIG. 108. De Laval Steam Nozzle. , . , , , A ,
the inlet and the narrow
end the outlet, we get the fluid gradually increasing in velocity up to the narrow
exit, where it issues, slightly converging at the full speed due to the height or
pressure.
If we reverse the nozzle we get a very different result.
Suppose we consider the nozzle connected to a steam boiler at 200 Ibs. pressure
about atmospheric, or 214.7. According to Professor Zeuner, in a nozzle in which steam
is adiabatically expanded the potential energy (heat) is converted into kinetic energy.
The kinetic energy of i Ib. of steam at velocity V in feet per second is of course
equal to V 2 r , ,
foot-lbs. (i)
2g
Suppose steam is let down from pressure p to pressure / 2 adiabatically in an
expanding nozzle. The internal heat of the steam at p l = (the internal heat of the steam
at / 2 ) + (kinetic energy at p z in heat units) ; or, in other words, the kinetic energy is
proportional to H - h, where H is the internal heat at p l and h that at p y And if J is
the mechanical equivalent of heat, the velocity of the steam acquired by the time it had
fallen to p. 2 would be V = *j2g](U.-h). (2)
The internal heats depend on the percentage of moisture in the steam, calculated
on the assumption of constant entropy during the expansion.
Theory and experiment agree that at a certain ratio between p^ and p., a maximum
amount of steam flows through a converging nozzle. This limiting ratio is
~~ = - 577 J and hence p., = o. 577 x p r
P\
160 -
tea -|
130}-\
At
Section
B.
"itofygauge
Flow of Fluids in Nozzles
In the case cited ot a pressure p l equal to 214.7 absolute ; p v that is at section B of
the nozzle, the pressure would equal 0.577 x 214.7= 124 Ibs. (3)
Absolute whichever end of the nozzle we used as the inlet, or 109.2 gauge pressure.
The steam, therefore, which escapes from section B has still much pressure, and
this can be seen by making the large end the inlet and blowing the steam into the
atmosphere, when the issuing jet will be seen to instantly expand widely.
By using the nozzle at the narrow end as the inlet, the steam, after passing
section B, is constrained to flow forward at a still increasing velocity as its pressure
falls, until it reaches the out-
let, where it cannot expand
further, and issues as a cylin-
drical jet at the full velocity
due to the pressure differ-
ence.
Professor Jamieson has
plotted out the following
curves (Fig. 109) of the expan-
sion in his contribution to the
discussion on the De Laval
steam turbine, with a view
of explaining not only the
sudden fall of pressure near to
and in the throat of the steam
nozzle of the De Laval turbine,
but also the further fall ot
pressure along the conical part
to its mouth. The full line
curve represented the natural
loss of pressure in dry satur-
ated steam as it expanded
in accordance with Professor
Rankine's well-known formula
pv\^ = a constant ; where p
was the pressure in Ibs. per
square inch absolute, and v
the corresponding volume in
cubic feet per Ib. of steam.
This curve is drawn from 475
Ibs. per square inch at which
pressure i Ib. of the steam
occupied i cubic foot down
to i Ib. absolute, at which
it occupied 330 cubic feet.
In the Figure is included the
range of pressures specially mentioned in the paper. The dotted line represented
an adiabatic expansion curve to pv^~ = a constant, from 215 Ibs. absolute at Section
A, before entering the nozzle, down to 0.93 Ib. at Section C, where it occupied
256.8 cubic feet, and left the nozzle with a velocity of 4127 feet per second with
24 per cent, of moisture. This curve passed through the point B, where the
steam occupied 3.5 cubic feet, and has 4 per cent, of moisture, with a velocity of
1500 feet per second. Now, it was evident from this curve that if the potential
energy of each Ib. of static steam at A had been so far converted into kinetic energy
at B that it had there a velocity of 1500 feet per second, and contained 4 per cent.
60
At
Section
A.
Pressure per square inch\
in Ibs. absolute . . /
Pressure per square inchn
in Ibs. absolute above Vaoo
atmosphere . . J
Per cent, of moisture . o
Volume of steam in cubic \
feet per Ib. . . . )
Velocity in feet per second o
The black line curve
4
3-5
1500
At
Section
C.
0-93
24
256.8
4127
represents the
expansion curve of dry saturated steam to
= constant.
a*
The dotted line curve represents
the adiabatic expansion of steam as per the
above data in the De Laval turbine steam
nozzle to 6v\P-= constant.
MlJk.
eo vo eo oo too *xo sso *to aaoasss&o
VOLUMES IN .CUBIC FCET TO THE LB.
FIG. 109. Comparison of Data and Curve from the De Laval
Nozzle, with the Natural Expansion Curve for Dry Saturated
Steam.
88
Modern Engines
of moisture, with an increase of volume from 2.11 cubic feet at A to 3.5 cubic feet
at B it must of necessity have fallen in pressure from 215 to 125 Ibs. absolute pres-
sure in doing so. This is in strict accordance with the natural law for the adiabatic
expansion of steam. The temperature of the steam must also have fallen from
382 Fahr. to at least 340 Fahr. in this short passage. It might be that the steam in
passing from A to B, and from the very small throat of ^-inch diameter, at the rate of
8 Ibs. weight per second, naturally formed a vena contmcta due to "throttling." In
any case, it must there lose potential energy due to friction and increased velocity.
This would account for its expansion from i Ib. of dry saturated steam at A to that of
slightly moist steam at B, with a corresponding and natural loss of pressure in temper-
ature during its increase of velocity from o to 1500 feet per second from A to B, and
corresponding fall in pressure.
TABLE VII. THE VELOCITY OF OUTFLOW AND THE WORKING CAPACITY OF
DRY SATURATED STEAM FROM EXPANDING CONES.
KONRAD ANDERSSON.
Counter-pressure i Atmosphere.
Counter-pressure 2.4 Lbs. per Square
Inch Absolute, corresponding to
Counter-pressure 0.93 Lbs. per Square
Inch Absolute, corresponding to
25 Inch Vacuum.
28 Inch Vacuum.
Initial
Steam
Pressure,
Lbs. per
Square
Inch.
Velocity
of Outflow
of Steam,
Kinetic
Energy
Foot-lbs.
per Second.
H.P. of
550 Foot-lbs.
per Second.
Velocity
of Outflow
of Steam,
Kinetic
Energy
Foot-lbs.
per Second.
H.P. of
550 Foot-lbs.
per Second.
Velocity
of Outflow
of Steam,
Kinetic
Energy
Foot-lbs.
per Second.
H.P. of
550 Foot-lbs.
per Second.
Feet per
Feet per
Feet per
Second.
Second.
Second.
Per Lb. of Steam per Hour.
Per Lb. of Steam per Hour.
Per Lb. of Steam per Hour.
60
2421
25.29
0.046
33 2
47-57
0.087
3680
58.44
o. 106
80
2595
29.06
0.053
3423
50-56
0.092
'3793
62.08
0.113
100
2717
31.86
0.058
3520
53-47
0.097
3871
64.66
0.118
1 20
2822
34-37
0.062
3596
55-8o
O. IOI
394
66.99
0. 122
140
2913
36.62
O.o66
3661
57-84
0.105
3999
69.01
0.125
160
2992
38-63
0.070
3718
59-65
0.108
445
70.61
0.128
180
3058
4-35
0.073
3764
61.14
O. Ill
4091
72.22
O.I3I
200
3"5
41.87
0.076
3 8lO
62.64
0.114
4127
73-50
0.134
220
3166
43.26
0.079
3852
64.03
o. 116
4 J 59
74.64
0.136
280
3294
46.83
0.085
3962
67.74
0.123
4229
77.18
0.140
In Mr. Konrad Andersson's paper he gives an example of the action of the steam in
the expanding nozzle, the narrow end connected to boiler at 200 Ibs. pressure, the wide
end into a 28 inch vacuum, steam supposed to enter dry.
The general formula? for the velocity of a gas in feet per second due to a difference
of pressure is V= ^2. gh, where h is the height of a column of the gas whose weight
balances the difference of pressure. But this formulae has its limitations, for beyond a
difference of pressure less than 58 per cent, of the initial pressure the flow is constant.
The velocity of the discharge of steam from a boiler at 100 Ibs. pressure is no greater
into a vacuum than it is into another boiler at 58 Ibs. pressure.
But this is true only for a converging nozzle discharging at the narrow end,
and suddenly and unrestrainedly expanding ; if, however, the nozzle is prolonged and
gradually diverges, the velocity increases and becomes proportional to the differ-
ence of pressure between the narrow nozzle and the wide end. The final velocity
of the steam is acquired in two stages. First, it acquires a velocity of 1500 feet
per second in passing from the boiler to and through the converging parts, and a further
2627 feet in expanding in the diverging part of the nozzle, making a final velocity of
4127 feet.
Flow of Steam
TABLE VIIL PROPERTIES OF STEAM.
Absolute
Boiler Pres-
sure per
Square Foot.
Absolute
Boiler
Pressure
per Square
Tnrh
Volume of
i Lb. of
Steam.
Cubic Feet.
Weight of
i Cubic Foot
of Steam.
Lbs.
Temperature
of Steam.
Fahr.
Velocity of
Steam Jet,
calculated
from Narrow
Nozzle. Feet
Square Root
of Product
of Density
x Pressure
per Square
J.liVll.
per Second.
Foot.
144
I
330.4
.0030
102.8
1238
.6
288
2
171.9
.0058
126.3
1262
!-3
432
3
i '7-3
.0085
141.6
1277
1.9
576
4
89.51
.0112
i53-i
1291
2-5
720
5
72.56
.0138
162.4
1299
3-i
1,440
IO
37-83
.0264
J93-3
1326
6.2
2,160
15
25-85
.0386
213.0
9.2
2,880
20
19.74
.0506
227.9
12. 1
7,200
5
8-333
.1199
280.9
1398
29.4
10,080
70
6.076
.1645
302.7
40.1
11,520
So
5-358
.1866
311.8
44-8
12,960
9
4.796
.2085
320.1
52.0
14,400
100
4-342
.2302
327.6
1420
58.0
15,840
no
3-969
.2519
334-6
63.2
17,280
1 20
3-656
2735
341.0
69.0
18,720
130
3-390
2949
347-1
74-
20,160
140
3.161
3J63
352.8
80.0
21,600
150
2.962
3376
358.2
'434
S6.o
23,040
160
2.786
3568
363-3
91.0
24,480
170
2.631
.3800
368.2
...
96.0
25,920
180
2-493
.4012
372.8
...
IO2.2
27,360
190
2.368
.4223
377-3
IO8.O
28,800
200
2.256
4433
381.6
J45 2
112. 1
36,000
250
1.825
.5478
400.8
1454
140.4
Taking the three sections of the cones :
Section A
Pressure, 200 Ibs. above atmosphere.
Moisture percentage = o.
Unit quantity of steam, i Ib.
Section B (smallest section)
Pressure, no Ibs.
Quantity of steam, 0.96. Moisture, 0.4 Ibs.
Velocity V= \/2^J(H-A)= 1500 feet per second.
Volume of steam, 3.5 cubic feet.
Section C (the wider end)
Pressure, 2 inches of mercury (absolute pressure).
Percentage of moisture in the steam, 24 per cent.
Specific quantity of steam, 0.76.
Velocity of the steam, 4127 feet per second.
Specific volume of the steam, 256.8 cubic feet per Ib.
The proportion between the areas of the large and small section of this nozzle should
be as 27.2345 to i, or the proportion between the diameters of these two sections as
5.2187 to i. If, for instance, the diameter of the small section is 6 millimetres, or very
nearly of an inch, the diameter of the large section should be 31.31 millimetres, or
nearly i^ inch. Through such a nozzle there passes a certain constant weight of dry
saturated steam of 200 Ibs. pressure per hour, neither more nor less. This fact of the
nozzle passing only a certain amount of steam per hour is used as a measure of steam.
The quantity of steam passed by the nozzle is measured by the section B, and is
equal to Q = 370 AD so long as the difference of pressure is greater than f the initial
9
Modern Engines
pressure. A is the area of the section at B, and D the weight of i cubic foot at 215
absolute pressure ; hence 370 x 0.25" x 0.5 =462.5 Ibs., taking A as inch and i cubic foot
of steam at that pressure = 0.5 Ibs. weight approximately.
W x V
We have seen that the impulse per second of a fluid is equal to . From this
! the pressure blown into. Now, if we take exhaust
steam at 15 Ibs. above zero, and assume that the pressure
falls to 8 Ibs. in the combining cone, we would obtain the
maximum weight flow of steam, for T 8 r is less than f .
The volume of the steam per cubic foot is large, and so
we must make the steam cone A larger in proportion, for it
still takes i Ib. of steam to raise up 7 Ibs. of water. 0.038
is the weight of i cubic foot at atmospheric pressure ; at
100 Ibs. pressure in the example quoted the weight is 0.23,
so that to admit the same quantity of steam the orifice areas
A should be as the weights .23 is to 0.038, about 6 to i
= 6 x .235 = 1.41 square inch. The water must have a slight
fall into the injector.
Practically i Ib. of steam raises 7 Ibs. of water, and the
steam cone would be 7 to i against a steam cone at 100 Ibs.
pressure; hence weight of steam Q AD 370= 1.635 square
inch x 0.038 x 370 = 22.25 Ibs. of steam per minute delivered,
or 1335 Ibs. per hour, and 1335x6 = 9345 very little less
than the 10,000 delivered by the live steam. Where there
are non-condensing engines exhaust steam can be used with
advantage to raise large volumes of water.
The sensible heat of the exhaust steam plus the latent
heat as it enters the steam cone is 1147, and that of the steam at 100 Ibs. = 1171
per Ib., so that if we take 1.022 Ibs. of exhaust steam it will deliver the same
heat in the combining cone as i Ib. at 100 Ibs. pressure, so that the fact that the
exhaust steam will feed against 80 Ibs. pressure, lifting 7 times its own weight, is not
so mysterious.
We come now to examine a few typical injectors. For locomotive feeding they are
incomparable. There is no lift in this case, and steam pressure need not vary much, so
that a simple injector meets with great favour. Fig. 120 shows it in its simplest form.
A is the steam cone, B the steam orifice, C the combining cone. in which the steam is
seen as a cone wedging itself into the water, condensing and flowing across an air space
D into the delivery cone E.
The air space D is of importance, as it provides a relief from back pressure which
would arise if from any cause the quantity flowing at d could not get freely away though e.
The flow of fluid must be unchecked, so as to maintain the velocity. Hence if a surplus
of fluid were to flow and no opening D provided, the velocity would be checked. By
providing an overflow any surplus is shaved off and escapes, so that no obstacle is
presented to the full velocity from the combining cone.
FIG. 1 20. Loco Injector.
Injectors
99
In discussing cones and water lifting-, we have shown in Figs. 1 18 and 1 19 that the
combining cone outlet must be larger in area than the steam cone orifice, in order to start
the injector on a lift of water. Sellars was the first to patent a device for this purpose,
shown in diagram Fig. 121. He fits for this purpose a steam tight rod with a small nozzle
drilled in its end marked the lifter. This nozzle slides by a screw in the steam cone, and
gets its steam from holes drilled obliquely into the nozzle from the side of the rod in the
position shown. It closes the main steam cone
and delivers a small jet to lift the water and start
with. When started the little nozzle is withdrawn
by the screw, and its steam parts closed.
A better method was introduced by Davis
and Metcalfe in the flap nozzle injector (Fig. 122),
wherein the nozzle is split and hinged so that
when steam is turned on it opens up and presents
an orifice greater than the steam cone, and so
forms the vacuum to lift the water for starting.
When the water arrives it is driven down in a
contracting stream, and the flap closes upon it as
a partial vacuum is formed around the stream of
water. It is thus automatic, and if from any
cause the injector should stop it will restart again. This instrument is well adapted
for locomotives, as it requires no attention to keep it at work.
This restarting arrangement is also employed in the same makers' exhaust steam
injector shown in Fig. 123, so that it starts off itself every time the engine is stopped
and started. This injector is very simple and effective and easily adjusted. By turning
the nut at the bottom the combining cone can be raised or lowered so as to nicely adjust
EXHAUST STEAM
FIG. 121. Lifting- Nozzles.
FlG. 122. Self-starting
Jet.
REGULATOR -
FIG. 123. Exhaust Steam
Injector.
FlG. 124. Automatic Restarting
Injector.
the water supply. The conical rod in the steam cone causes the steam jet to enter as an
annular jet, and so concentrates it upon the annular water jet. The steam in the centre
of a solid steam jet not being so effective as that on its outer periphery, the division in
the overflow is meant for a water seal to prevent inlet of air.
Another method for automatic restarting is shown in the latest Penberthy injector,
Fig. 124. Although introduced by Mr. R. G. Brooke, we select the Penberthy as
showing also the diverging steam cone, which must, as in the De Laval cone, increase
IOO
Modern Engines
the impulse. The combining- nozzle is nearly parallel, and cut into two lengths. In
order to restart automatically, a flap valve A opens or closes a chamber around the
division of the cone to the overflow O, and this valve opens and closes automatically.
The ordinary overflow is also used at the narrow throat of
the delivery cone.
There are many other designs for restarting auto-
matically, all based on this principle of an automatic valve
opening by the pressure to allow of the steam blowing out
the air and the water to follow it. Immediately condensation
occurs on arrival of the water, the valve closes by a partial
vacuum forming.
The original injector, the Giffard, had a tapered rod to
regulate the steam, and a moving cone to regulate the
water. Wherever injectors have to work under different
conditions of water and steam supply it is desirable to
have regulation of this kind. The original Giffard in-
jector is shown in Fig. 125. The steam cone slides by
means of the side screw, and the tapered rod by means
of the central screw and
hand wheel. By moving
either or both, adjustment
can be made for a large
variety of different condi-
tions. In fact, it is an ex-
cellent machine with which
to give or get an object
lesson on the effects of
varying pressures, tempera-
tures, and lifts. It is, how-
ever, not restarting auto-
matically.
Mr. Brookes' automatic restarter, fitted with a
sliding cone, which by moving the handle adjusts
both steam and water simultaneously, is shown in
Fig. 126. The combining cone is divided and fitted
with a closing and opening automatic valve to the
overflow. To make it restart, it has also the ordin-
ary air space overflow. The steam cone is carried
on a piston working in a cylinder, and can be screwed
out or in by the quick thread screw. A conical end FIG. 126. Adjustable Injector,
on the screw regulates the steam inlet, and the
movement of the outlet orifice regulates the water, decreasing the one at the same
time that the other increases. Hence a position can be found for satisfactory working
over a long range of varying conditions.
FIG. 125. Giffard Injector.
COMPOUND INJECTORS
With a single injector the temperature of the feed cannot exceed 200 Fahr.,
because of the low pressures in the cones. Any higher temperature generates steam
and destroys the action ; but if, with one injector, we force feed water into another,
we can, on account of the higher pressures, feed into the boiler at much higher
temperatures.
We will illustrate one by Green & Boulding, of London, called the Buffalo
Ejectors
101
Injector, as a very well designed example. The first injector on the right-hand
side is designed to lift the water and force it into the high pressure injector. On
lifting the handle 13 (Fig. 128), spindle 2 lifts the steam valve through 21 and allows
STEM steam to flow through steam cones 24 and
22. The lifting injector draws water through
21 and forces it through combining and
delivery cone 23, whence it enters combin-
EXHAUST STEAM
FIG. 127. Compound Injector FIG. 128. Section of Compound Injector.
ing cone 25, and from thence to the boiler. The overflow valve of the forcer is worked
by spindle 17 by the side rod shown in elevation (Fig. 127), which allows it to close
when the machine is at work, and opens it when stopped. When the water arrives and
steam condenses a water pressure is produced in discharge of the forcer 25, which
closes the overflow valve. This automatic clos-
ing of the valve is the feature of this injector.
Exhaust injectors can be combined with live
steam injectors to feed into
very high pressures. Messrs.
Holden & Brookes' excellent
arrangement is shown in Fig.
129. A jet of live steam starts
the exhaust injector through
pipe A, and it delivers at first
partly through P and the live
steam injector. When fully
started, all the feed takes
place through the live steam
injector.
AIR EJECTORS
We have already referred
to these. For some purposes
they are admissible, such as
on a locomotive, where a fan
or pump would be objection-
able on account of requiring attention. They are not economical.
The only case of steam jets being employed with economy in prime
movers is that of the De Laval turbine, where it is utilised by special means.
One application, however, deserves notice, that of furnace blowers where steam
is plentiful and cheap, as it is where the waste heat generates it. Messrs. Korting
Brothers supply a blower as shown in Fig. 130. On the bi-conical pipe is fixed a head
FIG. 129. Compound Injector.
FIG. 130. Air
Ejector.
102
Modern Engines
containing- three or four cones, the one blowing- into the other ; and the top one a small
steam cone. The object for using- a number of cones in succession, each succeeding
cone increasing- in size, is to induce a large volume of air at low pressure. The steam
sets up a flow in the first jet, this induces an additional flow in the next, and so on until
the accumulated blast enters the large delivery pipe.
W Ibs. of air are set in motion by i Ib. of steam with a velocity V of steam and
a velocity V l of the mixture, and the velocity of the air as it enters the cone by suction
v+wv v 2
V,, then the momentum is M = - ^rr^- The head of steam is , and of the combined
1+ W 2g
V 2
air and steam -!-. The efficiency is difficult to determine accurately ; theoretically, it
should be It therefore increases with the velocity of the air entering
each jet. Hence increasing the velocity from one to the other in succession is a gain.
EJECTOR CONDENSER
These machines are designed to condense the exhaust steam from engines to form a
working vacuum. They have the drawback of the jet
condensers and others in mixing the condensed steam
water with the condensing water. And therefore the pure
condensed water is mixed with salt or impure water ;
whereas, in the surface condenser, canal, river, or sea
water may be used for condensing, while the pure con-
densed water of the exhaust may be saved for boiler feed-
ing. However, they have a large field of usefulness where
these considerations do not apply with any force.
This invention is due to Mr. Alexander Morton of
Messrs. Morton & Thomson, Glasgow. In this condenser
the proportion of water to steam required is much the
same as in surface or jet condensers, about 27 Ibs. of water
to i of steam.
Morton, scientifically from the first, designed his
ejectors to make use of the smallest head of condensing
water, in fact, it could be submerged in a river or pond
of cold water and so draw in the condensing water. He
designed them to lift the condensing water by means of a
small supplementary live steam jet ; also to regulate the
water supply cones by movable cones ; also a special
automatic regulator to regulate the water supply. All
these devices, however ingenious and in some few cases
necessary, introduced the elements of complication, and
attention was required for them.
In practice it is found better to provide an ejector condenser of sufficient water
capacity for the maximum load, and a fair vacuum ; and to maintain a head of at least
15 feet on the condensing water, if not by a natural fall, then by a centrifugal pump
driven by power. For wherever there is an ejector condenser there will be power
available.
With a fall of 16 feet the velocity of the water = 8\/ 16 = 32 feet per second. Hence
it will balance the atmospheric pressure and produce a vacuum without the action of the
exhaust, and maintain a good vacuum under all ordinary conditions. It is therefore
usual now to employ a plain condenser, and in all cases provide a fall of condensing
FIG. 131. Morton's Simple
Condenser.
Ejectors
103
water usually by pump. Morton's plain, simple fixed capacity condenser is shown in
diagram Fig. 131. The water enters the cone B flowing out at A at about 32 feet per
second, forming a rod of cold water. Rushing at this velocity, the exhaust enters at
D, and surrounds this cold rod of water ; it quickly condenses thereon, increasing its
diameter slightly, and is carried with the rod right into G the delivery, from which
it cannot return, for the pressure due to the velocity is greater than an atmosphere.
Any air in the steam, if not of more than usual quantities in exhaust
steam, is also carried down.
The Korting form of this ejector provides a longer contact
for the steam and water, by using a long series of cones as shown
in Fig. 132. I have used this condenser with great success in many
cases, in nearly all of which the condensing water has been pumped
up by centrifugals. For each Ib. of steam we allow 27 Ibs. of water
lifted 16 feet, and allowing 25 Ibs. of steam per horse-power. Thus
we get the power required at the pump = 25 x 27 = condensing water
per horse-power = 675 Ibs. ; and 675 x 16 = foot-lbs. = 10,800 per hour,
or 1 80 foot-lbs. per minute per horse-power. As the friction and
other losses amount to nearly as much, it is safe to calculate on 360
foot-lbs. per minute per horse-power, or roughly about i per cent, of
the engine power.
A good vacuum increases the power of steam turbines very
largely, so that it is good economy even to provide an artificial head
and cooling apparatus for condensing water ; in which case a simple
ejector condenser is by far preferable to any other.
A water pressure intensifier by means of a water jet has been
used to increase the head of water in town supply pipes, in case of
a fire requiring a jet beyond the reach of the ordinary head. This
practice is, however, only possible where a high pressure supply FIG. 132. Korting-
exists alongside of the ordinary supply. The jet is shown in Ejector Conden-
Figf. 133. sen
In London, Glasgow, and other large towns a hydraulic pressure supply at 750 Ibs.
is laid on for working lifts and high pressure rams and presses. This supply can be led
into an ordinary supply pipe A in a very fine jet, as shown in E, and the flow of
water accelerated, and the pressure increased in the fire extinguishing jets. It requires
care in using the jet. The delivery jets must all be freely opened before the high pressure
jet is opened, and they must not be closed before the high pressure jet is shut off.
In Fig. 134 we give a section
of the Penberthy No. 2 water jet
lifter, showing the expanding steam
nozzle. They are convenient in some
cases for emergency and tempor-
ary use ; but all water lifters by
steam jets are wasteful of steam,
and therefore are not adopted for
regular pumping of water to higher
levels.
We shall conclude our examination of fluid-upon-fluid engines and machines by a
briet reference to ship propulsion by fluid or jets, a method which is worthy of con-
sideration along with the only alternative, the screw propeller ; which, in fact, is only
a pump outside the vessel producing a water stream, the reaction of which propels the
boat.
The propulsion ot steam vessels by water jets occupied the attention of the British
Government's Admiralty some thirty years ago. They were tried on a vessel called the
FIG. 133. Water Pressure Intensifier Jet.
104 Modern Engines
Waterwitch with some success, but the experiments for some reason or other were not
concluded, and no report ever issued of results.
Mr. J. R. Ruthven gave a brief report, which was read before the Society
of Engineers and Shipbuilders, Glasgow, 27th October 1891, from which we
gather :
The experimental trials made were those of the Waterwitch, a gunboat built for
the Admiralty, with the purpose of comparing the propelling and manoeuvring power of
the jet with the screw. For this purpose two screw ships, the Viper and Vixen, twin
screw vessels of about the same tonnage and same midship section, were built as fair
competitors with the jet propelled Waterwitch.
The water jets are obtained by a centrifugal pump which takes its supply from the
sea, or, in case of heavy leakage, from the hold of the ship. The centrifugal wheel
is placed horizontally in the ship, and is driven direct by steam engines working
horizontally.
There is an arrangement to reverse the direction of the discharge of the jets, so
that the vessel can be propelled ahead by discharging the water astern, or astern by
discharging ahead, or the jet at one side of the vessel may discharge ahead, and on the
other astern ; in this case the ship neither goes ahead nor astern, but turns round in her
own length.
The twin screw vessels Viper and Vixen and the jet propelled Waterwitch,
common ships, all of the same midship section, and nearly same displacement. The
twin screw vessels had double stern posts and two rudders at the stern, the Waterwitch
had two rudders, one at the bow and
one at the stern, each working in a
heavy frame, intended to be used as
a ram, at either end. This must have
caused considerable increased obstruc-
tion, and as the ship is longer, on
account of these rudders and frames,
~ more difficult to turn in her own
riG. 134. " Penberthy water Jet Lifter.
length.
In 1863 it was decided, by the Admiralty, to test the water jet on a large scale.
Some time before this date the Admiralty had been satisfied on the feasibility of the
plans, by themselves making and applying the jet, in an old screw gunboat, the Jackdaw,
at Devonport, by means of a series of teeth wheels and shafts, working from the old
screw shaft to the pump, which had only one outlet led to one side of the ship. With
this imperfect arrangement the ship was propelled, and a speed was attained which
satisfied the Admiralty that it gave promise of a success.
An important point in connection with the jet as a propeller must be noted, that is,
its power in stopping the vessel. This was tried with the Waterwitch, and from going
full speed ahead the order was given to reverse the action of the jets, and the vessel was
stopped in nearly her own length. This was ascertained by throwing a billet of wood
square from the bow on the order given to reverse the jets, and the ship was stopped
when the floating billet was level with the stern. This is a very remarkable property,
which the screw cannot approach.
This power of quickly stopping and reversing a ship's headway must be of great
importance to the safety of the ship, by enabling her in many cases to avoid collisions
which the continual increase of speeds makes more and more dangerous. In passing, it
may be remarked that the great stopping power is a proof of the efficiency of the water
jet as a propeller, the value of which is obvious in quickly stopping and backing, while
with the great power of turning there is no danger from loss of steerage way, as the
ship can be effectually steered by the jets when she has little or no way on her, at which
time, of course, the rudder is of little or no use.
Jet Propellers
I0 5
TABLE X.
Vessel's name
Vixen
Viper
Waterwitch
Nominal horse-power .
160
160
160
Date of trial
2nd Aug. 1867
Ft. In.
5th Aug. 1867
Ft. In.
nth Oct. 1867
Ft. In.
9-ioth Aug. '67
Ft. In.
28th Aug. 1867
Ft. In.
3rd Sept. 1867
Ft. In.
1 2th Oct. 1867
Ft. In.
Draught -j . .5 e " '
9 10
ii ii
9 ii
II IO
9 ii
II 10
IO
II
9
8
10 7J
n ij
9 9
n 9
10 7
II 2
Midship Section, square feet
336
335-7
336
347
336
333
336
Boiler power
Full
Half
Full
Half
Full
Half
Full
Half
Full i Half
Full
Half
Full
Half
Revolutions
108.6
85-1
109.2
82
110.5
83-7
40.7
26.2
4i-5
28.2
40.9
27.9
40.4
30.6
Indicated horse-power. .
657
336
652
334
696
344
777
226
801
271
769
268
759
348.5
Speed, knots
Speed 8 x Midship Section
9.06
379-9
7-347
395
9-475
438.5
7-334
9-586
435-4
7.625
431-3
9-237
351-9
6.227
369.6
9.219
328.7
6.047
273-9
8.88
303-3
6.326
3I4-3
9.299
357-8
7.206
359-6
I.H.P.
In the discussion upon this paper it is interesting to hear what her commander said
about her.
Captain Sharpe, R.N., said: " I can safely say, having commanded the Waterwitch
during the time the experiments took place, I always urged (and still think) that they
were never carried on to the extent that they might have been. Whilst in command
of the Watervoitch I felt the extreme advantage of having her always under my own
individual command. I could stop her, go astern, turn the ship, or do anything without
reference to the engineers. The only thing I have to add is that during the time I was
in the ship I was very sorry to think that so little use came of it, and that so little has
since been done."
The subject has dropped, although there are some small steamers of the lifeboat
class still made with jet propellers, and the introduction of high speed turbines has
proven that there are difficulties with propellers of the screw type ; hence jets may after
all be reverted to. It will therefore be of interest to investigate the matter a little
further, with thirty years' more experience and progress than Mr. Ruthven had to aid
him.
The late Mr. Miller, secretary of the Society of Engineers and Shipbuilders, Glasgow,
also read a paper on jet propellers. His proposal was not made as a serious attempt at
ship propulsion in general, but with a view to assist in the case of a failure of the screw,
its shaft, or engines. So long as the boilers were intact he calculated that a steam jet,
delivering astern all the steam the boilers could make, would propel her by reaction at
3 knots an hour, a speed not much but better than nothing at all. The difficulty, however,
arises, that if far from land she would burn up all the fuel long before getting to port,
and therefore never could finish the voyage. If she had coal for 12 days at 12 knots
steaming, she could not travel far at 3 knots on the coal in stock.
The screw propeller is a pump outside of the vessel, which thrusts the water away,
and the reaction propels the ship. The jet propeller driven by a pump does the same
thing exactly. The only difference is, the pump is inside in one case, and outside in the
other.
With the outside pump there are many difficulties, well known. The loss of the
propeller is not uncommon, neither is the shaft breaking uncommon. And as the screw
is necessarily a slow speed pump, the shaft has become of enormous weight and diameter.
With the inside pump and propeller jets high speed pumps can be used. There are no
propellers or shafts liable to be lost or broken. With obvious advantages it, however,
has some natural limits which may keep it from extensive adoption.
It is all a question of velocities and pressures to be properly apportioned. For
instance, in Mr. Miller's proposal to use a large steam jet, to get any efficiency, the ship
106 Modern Engines
should move at some large fraction of velocity of the jet at the orifice. This velocity is
about 1500 feet per second, or about 17 miles a minute.
The water jet, then, as a very first requisite, must have a velocity about equal to
that of speed desired, and this limits its efficiency and usefulness.
Suppose the speed to be fixed at 10 knots per hour, it has been found that at 5 knots
and upwards the power required is as the square of the speed nearly ; for our present pur-
poses that is at any rate assumed. At 5 knots it takes of a Ib. pull or push to propel
a vessel for every square foot of wetted surface, so that if we assume a ship with 6000
square feet of wetted surface it will require 1000 Ibs. thrust to keep up 5 knots. Then if
io 2 x looo 100,000
10 knots are required, we get g = = 4000 Ibs. thrust for io knots.
O
io knots is a rate of travel = 60,870 feet per hour, and - ' - = nearly 16 feet per
second. Now, this is the point where a limit comes in. This speed, 16 feet per
second, is the speed of efflux which the jet should have to get 100 per cent, efficiency.
We can calculate the head or pressure of water which will give this velocity.
V 2
H= = 4 feet head, or about 2 Ibs. pressure. Now, if we divide the total thrust
2 S
required by this pressure, - =2000 = the area of jet in section in square inches,
2000
= 13.2 square feet. A pipe of about 50 inches diameter would give this area.
144
Hence to get efficiency we require a large pump, and so large that we put it outside
in the form of a screw in screw propellers.
But the pressure is small, 2 Ibs. or 4 feet head, so that although the velocity of the
water jet issuing from the ship is limited to this velocity, we might run a small centri-
fugal pump at a high velocity, and the water may have a greater velocity in the pipes
inside the ship. 13 square feet of area x 16 feet velocity = 13 x 16 = 208 cubic feet of water
per second, and 208 x 62.5 = 13,000 Ibs. of water per second, 52,000 foot-lbs. per second.
Nearly 100 horse-power would be required in the water, and probably the efficiency
would be less than 50 per cent., so that an indicated horse-power would be required of
more than 200 in all.
Now, suppose we double the pressure to an 8-foot head, then velocity 8 x/8~= 2.83 x
8 = 22.64 feet per second velocity. With double the pressure we do with half the jet
area, 6.5 square feet, or 1000 square inches. A 32-inch pipe or jet would now suffice.
The velocity is 22 feet, and this multiplied by 6.5 = cubic feet of water per second
= 143 cubic feet, 9000 Ibs. nearly. And energy per second = 9000 x 8 = 72,000 foot-lbs.
= 130 horse-power in the water, so that the speed of the vessel would be still io
knots ; but instead of 100 horse-power, it would now require 130. The higher the
velocity of the jet the less the efficiency, unless the speed of the ship is increased at
same rate.
We thus find that to give the jet propeller a chance we require to move a very
large volume of water at a speed above that of the vessel, and the centrifugal pump
seems to be the best machine for moving these large quantities at low pressure. Now,
the steam turbine coupled to a modern centrifugal pump would be something very
different to the engines and pump used in the early tests just referred to. 80 per
cent, efficiency can be realised, that is, 100 horse-power turbine pump will give 80 in
the water horse-power. We cannot in this place enter into the whole question of
centrifugal turbines and ship resistances; we are considering water jet fluid-on-fluid
engines.
The next observation to be made on the subject is that the pump ought not to
throw the water out through the jets directly. The object of the pump is to produce the
desired pressure on the water, and for this purpose the water when it leaves the pump
Jet Propellers 107
should be delivered to a chamber or tank through widely diverging nozzles, so that it is
without agitation brought to rest or to a very slow motion.
The propulsion is due to the reaction of the jets, to the unbalanced pressure of the
water in this tank. If we have a tank of water maintained at 5 or 6 Ibs. pressure, and
open a hole in one side 2000 square inches in area, we get an unbalanced pressure of
6 x 2000 of 12,000 Ibs. on the opposite side of the tank to the opening, and that is the
force which propels the ship.
By careful design an efficiency as high as that of the screw propeller may be
expected. At any rate, it is worth investigating again practically, whether with modern
steam turbines driving modern centrifugal or screw pumps, and with fluid-flow and
pressures directed and adjusted on modern scientific principles, that the jet is not a
better instrument than the screw for propulsion. The steam turbine has so much to
recommend it for ship propulsion ; yet its full advantage can never be realised on small
high speed screws, and to gear down to large low speed screws would certainly be a
step backwards.
High speed is the order of the day, and the higher the speed the better for jet
propelled vessels, whereas at high speeds the screw falls off in efficiency. The weight
and space occupied would be much less, and the advantages of having no outside
propeller liable to be lost or damaged, and no tunnel shafting, are not to be despised.
The problem would be still further simplified if an efficient steam jet could be used
to propel these large volumes of water. The pumps and turbines would then be abolished
also. But water jets propelled by steam jets are very inefficient, pretty much for the
same reasons that the jet propeller failed of old. The steam has a speed of, say, 1500 in
an injector cone, and the water to have any efficiency would require to move at a large
fraction of this speed to get economy ; but no water jet can be induced to move at such
velocities. At 170 Ibs. pressure the water would issue from the delivery cone at 8 ijz^o
= 120 feet per second, so that if the steam jet in the injector flowing out at 1500 or more
per second, succeeds in only imparting 230 feet per second to the jet from the combining
cone, it will feed the boiler ; but 230 is a small fraction of 1500. For boiler feeding
this inefficiency matters not. The whole steam, with all its heat, minus that radiated
and conducted, and other losses, re-enters the boiler, so that heat efficiency does not
matter.
Theoretically, the velocities imparted should be inversely as the weights in motion.
If i Ib. of steam flowing at 1260 feet per second struck n Ibs. of water, the result
should be 12 Ibs. of water moving at 100 feet per second, and so the kinetic energy of
the 12 Ibs. of water at 100 feet would be 1200, same as i Ib. of steam at 1200 feet;
for the total momentum before impact must equal the total momentum after impact.
If this could be realised in practice, then it would be useless to go on making
reciprocating rotary or turbine steam engines. A Pelton wheel driven by this ideal
water jet would far and away excel the lot. But the steam condenses instantly, and so
softens the blow of impact, and the water has considerable inertia, so that it lags behind
the rushing jet. Consequently, a steam jet as a water lifter is a poor affair from an
economical point of view.
Recently Penberthy made some improvement in the direction of applying the De
Laval cone to injectors. In all early injectors and water jet lifters the steam cone
converges to the very end ; but some makers commenced to diverge the cone a little,
as shown in Fig. 126, evidently feeling their way cautiously. Penberthy has carried this
diverging of the steam cone still further, as shown in Fig. 124, in his latest injectors and
water lifters, with considerable improvement ; how much we cannot say, for it is very
difficult to get reliable information about jet pumps and injectors.
However, I made some experiments myself, using properly designed diverging
steam cones, and thereby more than doubled the water lifted by the same steam
consumpt. The reason is that the steam has a velocity nearly double of that issuing
io8
Modern Engines
from a converging nozzle, and the heat has been utilised more efficiently before
condensation.
We have seen that compound injectors have been used to feed high pressure boilers
at high temperatures.
A compound water lifter with several nozzles in series has also been devised whereby
enormous velocities could be obtained. The water is started from rest in the first
FIG. 135. Multiple Nozzle Water Lifter.
cone, accelerated in the second, third, and so on, until it acquires a velocity approach-
ing that of the steam jet in the last one ; it is hoped that an efficient water lifter will
result, water being accelerated in stages beginning with a small jet of steam. That is,
instead of all the steam passing in one annular jet of, say, T \j-inch sectional area, it is
passed in five each -^ inch area, or three of $ inch each. The cones are reduced in
proportion as the velocity is increased ; it is shown in section, Fig. 135.
In recent tests the actual results of a turbine centrifugal pump combination for
water pressure, the following table of results was obtained. It shows what might be
done now with jet propellers with efficient pumps.
TABLE XL RESULTS OF TESTS WITH DE LAVAL STEAM TURBINE PUMPS.
Type of Turbine
Pump.
Revolutions
per
Minute.
Height of
Suction
in Feet.
Height of
Delivery
in Feet.
Quantity of Water
delivered per
Second-gallons.
Water
Horse-
power.
Brake
Horse-
power.
Efficiency.
50 horse-power
^|
duplex pump
coupled in
1500
16.4
16.4
63.5
37.87
50.3
0-753
parallel .
J
635 Ibs. of water raised 32.8 feet = 37 horse-power in the water.
The pressure of the jet corresponding to the height would be 15 Ibs., the velocity
= 8 ^32.8 = 5.74 x 8 = 45. 76 feet. The quantity of water is about 10 cubic feet, velocity
45 feet ; hence =0.22 square feet, section of jet 31.68 square inches.
45
It is intended only in this brief discussion to put the jet propeller in its true position.
Because some engineers are quite satisfied with screw propellers and unreasoningly
refuse to consider anything else, that is no reason why the case for other propellers
should not be put fairly before those of a more inquiring and enterprising disposition.
Figures can only demonstrate the feasibility of the jet system, practice alone can decide
for or against it ; and what practical tests have been made were made with machinery
certainly not the best at the time, and far inferior to what would be employed to-day
to make practical trials. The subject will be more fully discussed under " Marine
Engines."
Jet Propellers
109
TABLE XII. HORSE-POWER OF i CUBIC FOOT OF WATER PER MINUTE UNDER
HEADS FROM i TO uoo FEET.
Head in
Feet.
Horse-power.
Head in
Feet.
Horse-power.
Head in
Feet.
Horse-power.
Head in
Feet.
Horse-power.
i
.0016098
170
.273666
330
531234
480
.772704
20
.032196
180
. 289764
340
547332
490
.788802
3
.048204
190
.305862
350
5 6 3430
500
.804900
40
.064392
200
.321960
360
579528
520
.837096
5
.080490
210
338058
37
595626
54
.869292
60
.096588
2 2O
354^6
380
.611724
560
.901488
70
.112686
230
370254
390
.627822
580
933684
80
.128784
240
.386352
400
.643920
000
.965880
90
. 144892
250
.402450
410
.660018
650
.046370
IOO
.160980
260
.418548
420
.676116
700
.126860
no
.177078
270
.434646
430
.692214
750
.207350
120
.193176
280
450744
44
.708312
800
.287840
130
. 209274
2OX>
.466842
450
.724410
900
.448820
140
.225372
300
.482040
460
.740508
IOOO
.609800
ISO
.214170
310
.499038
470
.756606
IIOO
.770780
160
257568
320
S'S^S
TABLE XIII. SHOWING HEAD AND PRESSURES OF WATER.
Head
Pressure of
Head
Pressure of
of
Head of
Head of
Head of
Water in
of
Head of
Head of
Head of
Water in
Water
Water
Water in
Water in
Lbs. per
Water
Water
Water in
Water in
Lbs. per
in
in Yards.
Fathoms.
Metres.
Square
in
in Yards.
Fathoms.
Metres.
Square
Feet.
Inch.
Feet.
Inch.
5
1.66
0.83
1.52
2.16
185
61.6
30.8
56.3
80. i
10
3-33
1.66
3-04
4-33
190
63-3
31-6
57-9
82.3
15
5.00
2.50
4-57
6.49
195
65.0
32.5
59-4
84.4
20
6.66
3-33
6.09
8.66
200
66.6
33-3
60.9
86.6
25
8-33
4.16
7.62
10.80
205
68.3
34-i
62.4
88.8
30
IO.OO
5.00
9.14
12.90
2IO
70.0
35-o
64.0
90.9
35
n.6o
5-83
10.60
15.10
215
71.6
35-8
65-5
93-i
40
13-3
6.66
12. 1
17-3
2 2O
73-3
36-6
67.0
95-3
45
15.0
7-5
13.7
19.4
225
75-o
37'5
68.5
97-4
5
16.6
8-33
15.2
21.6
230
76.6
38-3
70.1
99.6
55
18.3
9.16
16.7
23-8
235
78.3
39-1
71.6
101.8
00
20. o
IO.O
18.2
25-9
240
80.0
40.0
73-i
103.9
65
21.6
10.8
19.8
28.1
245
81.6
40.8
74.6
106. i
70
23-3
11.6
21.3
30-3
250
83-3
41.6
76.2
108.3
75
25.0
12.5
22.8
32-4
255
85.0
42-5
77-7
110.4
80
26.6
J3-3
24-3
34-6
305
101.6
50.8
92-9
132.1
85
28.3
14.1
25-9
36.8
310
103.3
51.6
94.4
134-3
90
30.0
15.0
27.4
38-9
3!5
105.0
52.5
96.0
136.4
95
31.6
15-8
28.9
41.1
320
106.6
53-3
97-5
138.6
IOO
33-3
16.6
30-4
43-3
325
108.3
54- !
99.0
140.8
!5
35-o
7-5
32.0
45-4
330
I IO.O
55-0
100.5
142.9
no
36-6
18.3
33-5
47.6
335
in. 6
55-8
1 02. i
I45- 1
U5
38.3
19.1
35-o
49-8
34
"3-3
56.6
103.6
H7-3
120
40.0
20. o
36-5
5 x -9
345
115.0
57-5
105.1
149.4
!25
41.6
20.8
38.1
54-i
350
1 16.6
58.3
1 06. 6
151.6
I 3
43-3
21.6
39-6
56.3
355
118.3
59-1
108.2
153-8
'35
45-o
22.5
41.1
58-4
360
I2O.O
60.0
109.7
155-9
140
46.6
23-3
42.6
60.6
365
121. 6
60.8
III. 2
158.1
"45
48.3
24.1
44.1
62.8
37
123.3
61.6
II2.7
160.3
J5o
50.0
25.0
45-7
64.9
375
125.0
62.5
"4-3
162.4
'55
51.6
25.8
47-2
67.1
380
126.6
63-3
115.8
164.6
160
53-3
26.6
48.7
69-3
385
128.3
64.1
"7-3
166.8
'65
55-0
27-5
50.2
71.4
390
130.0
65.0
118.8
168.9
170
56-6
28.3
51-8
73-6
395
I3I.6
65.8
120.3
171.1
'75
58.3
29.1
53-3
75-8
400
133-3
66.6
121.9
173-3
180
60.0
30.0
54-8
77-9
I 10
Modern Engines
THE CENTRIFUGAL PUMP
This useful apparatus might come under the head of pumps, but as its action
depends on fluid velocities and pressures it may be as well to include it in this chapter.
The principle of the machine has been given in its
simplest form by Professor Goodere, as follows.
Conceive a ring threaded on a rod, pointing
towards and at right angles from a shaft, and
revolving about it. We know that the rod will
continually press the ring in a line perpendicular to
the rod. The ring will tend always to go forward
in a straight line.
We have now to refer to a simple experiment
in the geometry of motion. Let a circular disc
(Fig. 136) be rotated about its centre C, mark that
centre with the point of a pencil, and then draw
the pencil rapidly from the centre outwards to the
circumference. A curve will be traced on the disc,
FIG. 136. Theoretical Pump Experiment, which will take the form CPQ if the disc rotates
slowly, or will become more spiral in character, as
shown by CRST, when the velocity of rotation is increased. The pencil moves radially
in a straight line, but it traces a curve by reason of the increasing linear velocity of
each point in the circle as we pass from the centre outwards. Our conclusion is,
that a curved rod will act more effectively than an inclined rod, and that if the
curvature be regulated to the velocity of rotation a sustained and uniform push may be
maintained on each portion of water
as it passes from the centre to the
circumference.
In constructing a pump on
FIG. 137. Centrifugal Pump.
FiG. 138. Construction of Centrifugal Pump Blades.
this principle we begin with a circular disc AB, shown in section (Fig. 137), which
receives the water pressure on both sides. This is an example of balanced pres-
sure, and the friction is correspondingly reduced. The disc forms the central
division of a hollow circular box, the water enters on both sides, as shown by the
arrows, and is forced outwards by the curved vanes, the three successive types of
vane are (i) radial, (2) inclined, (3) curved. In a pump by Mr. Appold the diameter
of the case is 12 inches, that of the central opening is 6 inches, also there are 6 arms,
Centrifugal Pumps 1 1 1
curved backwards and terminating nearly in a tangent to the circumference of the
bounding circle.
The pump is set into rapid rotation between flat cheeks at the bottom of a pipe,
the circumference of the box being open to the inside of the pipe, and the centre being
open to the supply of water. The height to which they lift is equal to the square of the
peripheral speeds in feet per second.
To lift the water to any given height theoretically it must be given a centrifugal force
equal to >j2gH where H is the height, or, approximately, 8 \/H. To obtain this force
the speed of the periphery of the fan blades must be run at speeds depending upon their
curvature and the angles which they make with the outer and inner peripheries. The
diagram Fig. 138 shows the usual geometrical construction and parallelograms for
the entering ends and leaving ends of the blades. Referring to the diagram
Fig. 138, the speed of the blade at A depends upon the angle < between the pro-
longation of the blade A^ and a tangent with the circle where the blade terminates.
This tangent A^ is proportional to the peripheral speed. The angle (j> varies in practice
from 1 5 to 90 ; and in a pump with a proper volute the velocities V, instead of being
equal to \/2-H, are
< = V
90 = 0.83 \/2
45 = -94 >
30 = 1.03 ,
20 = 1.89 ,
15 = 1-355 i
In the case of a pump with a diffuser or whirlpool chamber we take into account the
inner radius and outer radius of the 4 blades r v r 2 ; and we take two ratios of these
radii as examples first, where 1 = | ; second, where 1 = f.
without a diffuser or diverging delivery.
45 = 0.9 = 0.885
30 = i.o = 0.98
15 = i-33 = i."
Without going into all minute corrections and coefficients included in the mathematical
treatment, we can indicate the calculated dimensions of a centrifugal. It may be first
pointed out that, as shown in above tables, as angle decreases or increases, so does
f _
the speed. With angle = 30 and -f"f, the speed is= \/2-H ; or at 15, V= i.
r
An angle of 90 may and has been used without decreasing the efficiency, but
lowering the speed.
The head of water maintained by a centrifugal with vanes radial at the rim is found
V 2
practically by h = . h = head in feet, V = velocity of wheel rim in feet, ^-=32.2 acceler-
o
V 2
ating force. And not to the usual theoretical formula h= .
2g
This is due to the fact that the water receives two impulses in two directions from
the fan. The true centrifugal impulse is radial along the vanes, and a tangential
impulse whirling it round. The velocities of the water in each direction at the moment
it leaves the wheel are equal to the peripheral velocity. The resultant of the two
V 2 V 2
impulses and two directions is that of the two components ; hence the k = x 2 = .
*g S
I 12
Modern Engines
V 2
Therefore the head may be greater than that equal to ; but owing to friction
o
of water and obstructions to flow the calculated head is never attained in
practice.
By curving the blades back the ratio between head and velocity of wheel is altered
considerably, as will be seen in the tables.
The efficiency in all but the lowest lifts is above 60 per cent. And by careful
design and good workmanship over 80 per cent, has been obtained in lifts of 20
to 40 feet.
The lift of a pump is limited by the highest circumferential speed at which the fan
may be run, so that when the lift exceeds 40 to 50 feet it is necessary to put pumps in
series. And when we require both high lifts and high values in gallons per minute we
must put them in parallel.
The radius r^ of a fan is limited by the revolutions per minute, and the
FIG. 139. Centrifugal Pump.
FIG. 140. French Design of Pump.
speed in feet of the circumference of r r The inlet for water is limited by the
value of r 2 .
With a given speed both H and Q are limited, and if the limit of H is exceeded,
turbines are put in series ; if Q is exceeded, then they are put in parallel.
For instance, suppose a centrifugal to lift 36 feet, and V the speed in feet of the
outer circumference of the fan, and that V= \lgh= ^32.2 x 36 = 34 feet ; the revolutions
per second are, say, 5. Then the circumference must be = 6.8 feet, or 26 inches
y i
diameter. The ratio being = -; ^1 = 13 inches; r 2 = 8.6 inches. The velocity of inlet
r z 2
at the radius r z at the fan even as high as 22.5 feet would give a possible delivery of
only 10.4 cubic feet per second as the maximum at that speed and lift ; hence we would
require to parallel the pump with another if more water was required, and the higher
the speed the sooner the limit to the capacity of the pump is reached.
As r z cannot be increased without loss of efficiency, increasing the output by that
means is restricted, as the efficiency falls off rapidly with decreasing values of ^X
Centrifugal Pumps 113
Fig. 139 shows an elevation in section of the usual construction, with curved pump
blades. The water enters at both sides, and thus balances the pressures. The angle <
in this case is about 15. Fig. 140 shows an elevation in section of the huge pumps
erected in Egypt by the French, in which the angle < is 90. They are slightly curved
the opposite to that in Fig. 139, but not radial. A few particulars may be of interest.
The shaft is vertical, and the water enters by the lower side only. The fan is much
like an inward and radial flow turbine wheel, like the Hercules wheel reversed ; if that
wheel (Fig. 77) were driven by power and the lower end dipped into water, the water
would be scooped up vertically and thrown out radially, that is the action of these
pumps. Their lift is 10 feet, and the water lifted 212 cubic feet per second ; the speed
is 32 revolutions per minute. The outer diameter is 12.446 feet ; circumferential
velocity, 20.8 feet per second = o. 82 *j2g-H, a figure which does not at all agree with
mathematical deductions ; the outer case is 19 feet 8 inches diameter. The mouthpiece
of the volute is 5 feet 3 inches.
The useful work done by the pump is 65 per cent, of the indicated horse-power ;
coal consumption, 3.85 Ibs. per pump horse-power per hour.
Now, looking to this result and the results of closely reasoned and calculated
pumps with volute curves and everything according to mathematical deductions, which
are no better, and in many cases not so good, we come to the conclusion that the data
assumed is either incomplete or in error.
Most manufacturers accept the values for < and V as given above, and, having
decided upon a value of for all their pumps the same, stick to it. -i is a ratio they
also fix upon and adhere to, both being checked by careful tests on actual pumps,
The velocity of influx at the inlets is usually 8 feet per second, or a little more ; the
velocity of discharge is from 2 to 5 feet usually per second.
The fundamental formulas is the same as for turbines. If the water leaves the wheel
without velocity, then if V is the velocity of water in feet per second the pressure due
to each Ib. stopped, or the pressure required to start it up to the velocity V, is ; and
o
multiplying by velocity of wheel rim V^ the useful work per Ib. = l foot-lbs. per
J5
V 2
second. The foot-lbs. per Ib. of water is H= , of which ?iH is given to the wheel
by the water or by the wheel to the water ; hence TjH = i, the fundamental formula
o
for centrifugal pumps and turbines, 77 being the efficiency factor.
In the case of the pump, 77 is about 0.75, and the fundamental equation H, the height
to which the water is to be lifted, requires that 1?/ H.
&
Without going into all the calculations, we will take
V l as the speed in feet of the rim of the wheel per second.
V as the speed of the water due to head H.
r lt the outer radius of wheel) _ 3
r v the inner radius of wheel/ ^"
H = height to which water is to be raised, to which is added an allowance for
pipe friction if pipe is long.
U = the speed of water in feet per second at inlet = 8 feet.
B = diameter of inlet.
U l = speed of water at outlet = 2 to 5 feet.
B! = diameter of outlet.
b =area of outlet of water on rim of fan at r v not including thickness of
blades.
VOL. i. 8
114 Modern Engines
#! = area of inlet at r 2 , not including thickness of blades.
Q = cubic feet of water delivered per second.
A =area of suction pipe.
A! = area of delivery.
N = revolutions per second.
Taking, for example, a pump to lift 32 feet and deliver 10 cubic feet per
second, vanes radial at the rim.
The pump is supposed to have a volute snail delivery chamber, and either a
diverging delivery pipe or a whirlpool chamber, that is, a circular space surrounding
the rim of the wheel in which the whirl of the water is reduced before it enters the
snail. Let the following values be known :
U = 8 feet per second inlet
lli = 5 feet per second outlet
Q = 10 cubic feet per second [known quantities.
H =32 feet
N = 5 revolutions per second>
The peripheral speed will then be = V >Jgh= ^32.2 x 32 = 32 feet per second.
The revolutions are 5 per second; hence ^- = 6.4, the circumference at r lf or
O
24 inches diameter.'. r^= 12 inches.
m
-J = f ; hence ^ = 1 2, r 2 = 8 inches, and V at r z will = 20 feet.
12 O
d, the breadth of the opening in the wheel at r v = ;) wherein we multiply
by 12 to reduce to inches ; hence
, 12 x 10 ,- 12 x 10 - ,
o = 0.6, and x = = 1.4 inches.
/2X-3.I
32( - -
20l
12 12 '
Now, to find A, usually we have the water entering on both sides, so we will divide
by 2 to find A for one side. A=' Cubi feet per second x 144 e inches for
8 feet per second x 2
each side of suction pipe and side opening in the centre of turbine, io| inches. To
this must be added an allowance for shaft arc, which would bring the central openings
to 1 1 inches, allowing about 8 square inches for shaft.
! at discharge will be similarly = IO * *J^= 225 square inches.
.8 x U .8x8
The suction pipe = -- 44_ 2OO square inches.
0.9 x8
Thus we have found the leading hydraulic dimensions. The irechanical dimensions
are treated under mechanical designs of pumps.
We will now illustrate a few actual examples of recent results of tests.
The ordinary belt driven centrifugals and the direct coupled centrifugal to quick
speed engines are greatly in use, especially the direct coupled, for marine circulation and
bilge pumps. These are too well known to require space here.
De Laval pumps, shown in Plate IV., by Messrs. Greenwood & Batley, driven by
steam turbine, are of recent introduction. The high speed enables considerable lifts to
be obtained, and two pumps can be driven, the one feeding into the other in series,
so that their pressures are added ; and if each lifts 100 feet, the two would lift
200 feet.
They can also be joined in parallel. The following table gives some results :
Centrifugal Pumps
TABLE XIV. RESULTS OF TESTS WITH DE LAVAL STEAM TURBINE PUMPS.
Type of Turbine
Pump.
Revolutions
per
Minute.
Height of
Suction
in Feet.
Height of
Delivery
in Feet.
Quantity of Water
delivered per
Second Gallons.
Water
Horse-
power.
Brake
Horse-
power.
Efficiency.
50 horse-power
)
duplex pump
coupled in
j- 1500
16.4
16.4
63-5
37.87
5-3
0-753
parallel
J
50 horse-power
'
duplex pump
coupled in
parallel
Constructed for
1500
16.4
29-53
46.3
38.66
48.0
0.805
larger head
of water than
the previous .
,
50 horse-power
\
duplex pump
coupled in
\ 220O
19.7
137-8
12.3
35-22
5-3
0.700
series .
J
20 horse-power
\
duplex pump
coupled in
\ 2315
9.84
85-3
82.5
14.27
20. o
0.713
series .
J
The general tendency at present is to increase the speed of everything, and it may
perhaps be considered as a step in the wrong direction to reduce the speed of a machine.
It may therefore be interesting to draw attention to a machine now on the market
which runs at the same speed as the turbine, namely, a centrifugal pump worked direct
from the turbine and at the same speed. The machine consists of a steam turbine
driving one centrifugal pump direct, and another by means of gearing. The slow speed
pump lifts the water and presses it into a high speed pump, which in its turn forces the
water against a considerable head. The following table shows some of the results
obtained by tests with one of these machines :
TABLE XV. RESULTS OF TRIALS WITH A DE LAVAL HIGH PRESSURE
TURBINE PUMP. "TURBINE SPEED PUMP."
Height of
Delivery.
Feet.
Gallons of
Water
delivered
per Minute.
Water
Horse-power.
Lbs. of Steam
per Water
Horse-power
per Hour.
ist Trial ....
312
529
5
34-2
2nd Trial ....
443
629
84.4
29.0
3rd Trial ....
59
5 2 9
81.6
3J-3
Work for
condensing
(
640
90.7
30.6
is included.
4th Trial ....
467i
53
430
75-o
60.9
32.2
35-J
I
286
40-5
43-i
In these tests the machine was mounted at a considerable distance from the boiler,
and the admission steam was consequently supplied to the turbine through a considerable
length of piping, which gives one reason to conclude the steam was wet. Better results
n6
Modern Engines
would probably, therefore, be obtained with dry admission steam, and with a higher
steam pressure. The condensing- water was supplied through a pipe from the slow
speed pump, and the machine consequently drove its own condenser. In the figures of
the table, work for condensing is therefore included.
To raise water 500 feet by centrifugal force is an achievement of no small moment.
The special characteristic of the turbine pumps, namely, their high speed which is
greatly in excess of anything previously known in connection with centrifugal pumps
is a mechanical problem which has been solved, partly by an improvement in the
construction of the wheel and packing boxes of the pump, including their lubricating
arrangements, and'partly by paying the greatest attention to execution of all the movable
parts. The high rate of speed has enabled the diameter of the pump wheel to be
reduced, and consequently the passive resistance of the pump has been greatly lowered.
. 141. De Laval Turbine Pumps in Parallel.
This circumstance, together with the fact that no transmission belts are used, has
made it possible to obtain with the turbine pumps an efficiency considerably in excess of
the results generally obtained with centrifugal pumps.
Owing to the water in the pump running in a continual current, self-acting valves
are avoided, and therefore the turbine pumps are very easily attended to and require very
few repairs. The turbine pump runs without any shocks, and consequently air vessels
are not required either for the suction pipe or for the pressure pipe.
Another great advantage of these pumps is their absolute safety against bursting.
Should the pressure pipe suddenly be shut off during the time the pump is working at
full speed the water pressure could not go beyond the so-called "centrifugal pressure,"
which, as a rule, is about 25 per cent, above the ordinary working pressure.
The dimensions of the turbine pump are small, and the weight, when compared with
its capacity, is low. In these respects the De Laval patent turbine pump has the
advantage over all other systems of direct acting steam pumps.
High Lift Centrifugals
117
To the larger sized steam turbines, which are provided with two driving shafts, are
coupled two pumps, one to each shaft. By arranging these pumps in such a manner
that one of them draws the water and presses it into the other, the pressure in the
delivery pipe of the latter pump is doubled. Thus a high pressure is attained, and
the efficiency becomes higher than if the same pressure were to be developed by only one
pump of larger diameter.
This pump thus arranged is especially well suited for a stationary fire engine for
large works, or wherever the pump can be coupled to a system of piping provided with
standpipes ; it is also suitable as a floating fire engine, or as a mining pump.
For waterworks in small towns the high pressure pumps are particularly well
adapted, in consequence of their high efficiency and their steady pressure. The stuffing
FIG. 142. Parsons' Turbine Pump.
boxes are provided with water checks, which completely prevent the intrusion of oil into
the water.
When the two pumps are coupled parallel that is to say, when both of them draw and
deliver the water at the same pressure a most effective water raising pump is obtained,
which delivers double the quantity of water delivered by the high pressure turbine pump,
but only at half the pressure.
Messrs. Parsons also make high lift turbo pumps, specially designed for high
speeds.
Fig. 141 represents a De Laval turbine pump with the deliveries in parallel ; and
Plate V. a De Laval turbine pump with deliveries in series.
Fig. 142 represents a Parsons' turbine with one pump under tests. This pump
delivered 55,000 gallons per hour at 165 feet head.
n8
Modern Engines
Messrs. Mather & Platt, Manchester, construct pumps with fans in series inside
of one casing instead of coupling separate fans.
The feature of this pump is that it consists of one or more sets of vanes, or
impellers, each set running in its own chamber but upon a common shaft, the delivery
pressure of the liquid varying directly as the number of chambers used. Thus, if an
ordinary single pump can deliver water against a head of 30 feet, the addition of another
chamber will give a final delivery head of 60 feet, while 4 chambers will enable the
pump to discharge the same amount of water against a head of 120 feet.
With this multiple pump large quantities of water, to a height of as much as
150 feet, are delivered.
Recently, by still greater improvements, they are now able to deliver water against
as great a head as 200 feet with a single chamber and at a high efficiency, some of the
larger sizes giving out the equivalent of as much as 76 per cent, of the power put
into them.
Most of these pumps run at speeds over 1000 up to 2000 or more revolutions per
minute.
FAN BLOWERS AND EXHAUSTERS
These machines are centrifugal pumps for air, being designed for much higher
velocities. Air being so much less in weight per cubic foot, the speeds of driving are
necessarily high.
If V = velocity of the tips of a fan in feet per second, and P = pressure in Ibs. per
square inch, V= ^97,300, and P = ^ .
773
If Q = cubic feet of air per minute, and P = pressure in Ibs. per square foot, i.e. the
FIG. 143. Anemometer.
FIG. 144. Heenan & Gilbert's Fan.
The pressure P is measured by a manometer, a U tube, which indicates
difference in pressure between inlet and outlet, then the horse-power utilised by the fan
blower = ^ .
33coo
the pressure by difference of the level of water caused by the pressure. The quantity
is measured by the anemometer, as shown in Fig. 143, an air meter by Casella of
Holborn.
The Pitot tube is an instrument consisting of a bent glass tube on a scale. One end
at right angles is fixed to the blast, the other limb is vertical. The air blast raises the
column of water inside the vertical limb, and this difference of level indicates a certain
velocity of flow, the depression of the column being proportional to the square of the flow.
Air Impellers
119
Fans are made with all sorts of blades, and every maker claims his as the best.
Some are curved backwards, some curved forwards, and others not curved at all. But as
it is a centrifugal pump there is no
doubt the same reasons for curving
the blades as there are in that case.
Professor Rankine suggested the
form adopted in the Heenan & Gilbert
fan shown in Fig. 144. The vanes ^^^^
at the inlet slice off the air and
wedge it outwards, finally throwing
it off at full velocity of the fan tip,
which is radial. Most fans deliver
the air radially, but Rateau in France
and Mr. Parsons in this country have
designed axial flow fans. These
have the advantage of being readily
constructed with fans in series, so
that high pressures can be attained.
Mr. Parsons in this way, by using a
long series of axial flow fans with
guide blades between, like his steam
turbine reversed, has succeeded in
producing air pressures up to 50 to
60 Ibs. per square inch.
The turbine can be economically
used for air or gas compression.
A modified form is illustrated here
attached to an ordinary steam tur-
bine (Fig. 145) ; this particular set is
capable of developing from 300 to
400 horse-power, and of compressing _^^^^^ ^_ TBHI I
air to 25 Ibs. per square inch. Com-
pressor can be designed for practi-
cally unlimited pressure by coupling
two or more in tandem with or with-
out intermediate coolers, and where
electricity is available they may as
readily be motor driven.
Large steam sets are especially
applicable for blast furnace work ;
there are no air valves to damage by
heat, and, as is usually necessary,
the pressure can be increased by 50
per cent, temporarily by means of a
byepass valve on the steam engine
admitting high pressure steam to the
low pressure parts of the turbine.
In rock drilling or mining work,
motor driven compressors can be at-
tached to small cars, and the com-
pressors taken right up to the face
The conveyance of an equal amount of power electrically over long distances is much
less costly than by pipes from large compressors at bank.
H
b/J
s
I
Air Impellers
The weight of the set shown in the print is about 7 tons complete.
The Rateau fan wheel for axial flow is shown in Fig. 147. A number of these can
be threaded on one shaft driven at a high speed and
with guide passages between, so that the pressures
are added together at the end outlet. And as the
volume of the air diminishes as it rises in pressure,
succeeding wheels are made of small diameter and with
fewer blades.
Mr. Parsons' wheels are like his turbine wheels,
with curved blades.
The Rateau fan lends itself very nicely to electric
driving.
The great usefulness of fans in ventilating mines
is their most important work, very large and powerful
fans being required either to force air down or draw
air up through the working, to clear out gases, foul
air, smoke, steam, etc., and supply fresh air for the
miners.
At Clara Vale Colliery a screw fan driven by a
steam turbine has been fitted. It is illustrated in Fig. 146, and has considerable
interest as a more efficient and reliable form of fan than the centrifugal form.
FIG. 147. Rateau Fan.
The only other fluid pressure apparatus
we may refer to in this chapter is the air water
lift, and we give an illustration of the plant in
Fig. 148. On the right is a steam driven air
pump to give an air pressure a little more
than the pressure due to the height of the
water to be lifted = 2. 3 feet per Ib. of air pres-
sure. The air is pumped into a reservoir,
shown like a vertical boiler. This is Messrs.
Worthington's arrangement and illustration
of the lift :
In principle and operation the air lift is
extremely simple. Air is injected through a
nozzle placed below the working water level,
and, rising, carries water up with it. The
submerged parts are subject to no wear
whatever, and, except for corrosion, are
practically indestructible, even when working in the gritty water so fatal to deep well
pump pistons. The air compressor and receiver, the only parts needing attention, are at
the surface, and readily accessible. When proper submergence of the air nozzle can be
FIG. 148. Air Water Lifter.
12,2, Modern Engines
assured the yield of a well previously fitted with a deep well pump is doubled or trebled
by the use of the air lift. The sparkle and life imparted to the water raised, and the
cooling effect of the expanding air, make this method peculiarly suitable for waterworks
purposes ; while, owing to the absence of pump rod and plunger friction and the superior
steam economy of the fly-wheel compressor, better duty can be obtained than with the
old style pump. When warm water is to be raised, its heat, instead of being a dis-
advantage, is a source of increased economy.
A certain minimum ratio between submergence of the nozzle and height of lift is,
however, essential to the successful working of an air lift.
This concludes a review of wind and water prime movers, with some notice of
machines working on hydraulic and pneumatic principles, which are not strictly speaking
prime movers, such as injectors, centrifugals, fans, and fluid-on-fluid pumps. Yet, as
they play a not unimportant part in prime movers' applications and operations, they are
worthy of some attention.
The hydraulic jet has been examined carefully and fully, for it is more than likely to
play a very important part in marine propulsion, as it offers a solution of the difficulties
of high speed, driving engines direct coupled to screw propellers outside of the vessel.
The jet presents some difficulties also, due to the fact that high pressure jets cannot be
used with economy, as the speed of the jet must not exceed the speed of the vessel by
more than twice. This entails using large quantities of water, the chief objection being
the added weight of the water inside the vessel.
On the other hand, we have seen the many advantages the jet has over the screw.
CHAPTER III
THE STEAM TURBINE
THIS prime mover, long- the dream of advanced and leading scientific engineers, has
at last reached the stage of perfection at which it can with confidence be employed
to replace the reciprocating piston and cylinder engine of any but the smallest size.
The bigger and more powerful the steam turbine, the better it compares with the piston
and cylinder engine if condensing is used. And even non-condensing Parsons' turbines
compare well with piston engines non-condensing.
Steam turbines are either of the pure impulse type, like a Pelton wheel, or of the
pressure type like the pressure turbine, of which we may cite the vortex wheel ; it will
be observed that the so-called pressure turbines, are only partially pressure turbines, for
most of them have curved guide blades in order to obtain also the impulse due to
change of motion. The De Laval turbine is purely an impulse turbine, the steam being
thrown at the highest possible velocity on the curved blades of the wheel, where its
V 2
kinetic energy is converted on the blades = per Ib. of steam.
*g
In Parsons' turbine the steam acts by its velocity being arrested on the movable
blades in the same way, and also by its pressure reacting as it leaves the movable blades.
Flowing from one wheel to the next, its velocity is gradually absorbed in each, and it
mV 2
expands from one into the other wheel. The kinetic energy = taken for each wheel.
2 S
There is the third turbine, a purely pressure turbine ; the original of which was Hero's
turbine, in which the motion and power is due entirely to unbalanced pressure. The first
patent for this type was granted in 1784 to an engineer with the name of Wolfgang de
Kempelen. As a matter of fact, there is no material difference between the impulse and
reaction turbines, so that only two types may be considered.
As the steam turbine has come to be a practical prime mover of great im-
portance, we may briefly review its rise and progress. That it will be much
improved upon in construction is beyond a doubt, and some of the earlier pro-
posals may still come forward and claim attention. It is as well to know what they
are or were.
To consider them in their order of time would be of no service. We will consider
them as a class at a time, beginning with the Hero turbine, which may yet be of service.
Presently the difficulty with this type is the high speed required of the wheel to obtain
anything like efficiency ; there is also the difficulty of getting a steam-tight joint at the
admission orifice for the steam into the wheel at very high speeds, and a third difficulty
is to get a wheel made strong enough to stand the strains of centrifugal force at these
high speeds.
Fig. 149 illustrates the Hero engine in its simplest form. Little toys in this form
were common at one time, made of glass. A brass or copper hollow ball, with a neck
123
124
Modern Engines
like a door knob, has two arms at right angle to the axis attached to the neck ; it is
pivoted at the bottom, and guided at the top by a stem. On applying a flame to the
bottom, steam is generated, which in passing out of the nozzles leaves an unbalanced
pressure which drives them backwards in the opposite
direction to the steam issue. This plan of revolving the
boiler with the nozzles obviates any difficulty with steam
joints.
De Kempelen's engine was the first practical attempt
to apply this form of turbine to actual work. It is shown
in Fig. 150, wherein A is the boiler, with a stop valve C, a
safety valve B, and a vertical pipe carrying a horizontal
pipe E, which is free to revolve ; it having a sleeve joint
as shown in Fig. 151, whereby it can move freely and yet
be steam-tight.
At the extremities of the horizontal tube a hole is
made, shown at DD, through which the steam escapes.
As the inside of the pipe opposite the holes has a larger
area by the size of the hole than the side with the hole in
it, the pipe will be pushed backwards by the steam.
Thus if the size of the hole is i square inch, and the
pressure 10 Ibs. per square inch, then there will be an
unbalanced pressure of 10 Ibs. on the side of the pipe
opposite the hole, which will propel the pipe in that
direction away from the hole.
Many writers describe this action as "reaction" ; but it is not reaction, but direct
pressure which drives the Hero turbine.
The discs on the end are weights for fly-wheel purposes. It will be evident that
FIG. 149. Hero's Engine.
(200 B.C.).
FIG. 150. De Kempelen's Engine.
FiG. 151. Details of Steam-Tight Joint
of De Kempelen's Engine.
such an arrangement would be very limited in speed. The pipe would soon reach its
breaking strain as the speed increased.
The next improvement was made later on in 1849 by Mr. Nasmyth of Patricroft.
It is thus described in the Practical Mechanics Journal, vol. i., 1849 :
Early Turbine
Fig". 152 is a complete longitudinal section of the steam wheel, with its frame,
saw, and foundation plate; and Fig. 153 is a transverse section through the wheel
FIG. 152. Nasmyth's Turbine Saw.
and waste steam case. The wheel A is a hollow open disc, with a loose side
B checked into it, and bolted down by nine bolts. The hollow saw shaft C is cast in
one piece with the wheel, and is carried in the long bearing of the double pedestal PP,
0\
FIG. 153. Nasmyth's Turbine Saw, showing Nozzles.
which is bolted down to the foundation plate D. The contrary side of the steam
wheel is steadied upon the conical bearing E, on the end of the supplying steam pipe F,
which rests between the adjustable centre G of the back pedestal and the conical
126
Modern Engines
aperture in the centre of the side plate B of the wheel. The arrows show the
passage of the steam into the wheel by the pipe H ; and the escape of the waste
steam and condensed water by the upper and lower pipes I and K, connected with
the outer casing of the wheel L. A compensating spring is cleverly arranged in a
prolongation of the steam pipe F to counterbalance the pressure of the steam
upon the area of its ingress
opening at M. A helical spring
N is placed in the open end
of the pipe, so as to abut
against the closed end of the
pipe M. This spring is acted
upon by a short sliding block,
urged forward by the adjusting
centre G ; in this way it is com-
pressed to such an extent that
its reaction will just compensate
for the steam pressure upon the
supplying area, and so prevent
the possibility of undue friction
upon the conical joint of the
side plate of the steam wheel.
The waste steam case L is
cast in two halves, bolted to-
gether at the centre by means of
side lugs, the whole being firmly
bolted by flanges at the bottom
to the base plate, which has a
steadying rib cast on it to fit
the interior of the lower edge of
the case. The total area of the
four steam apertures is i^ square
inch ; steam pressure, 60 Ibs. ;
revolutions per minute, upwards
of 2000. The time taken to cut
a bar of the average section is
10 secortds, and it is to this
limited period of action that this
species of steam propeller owes
its peculiar fitness. For although
the wheel has in reality very little
power, yet its great speed ad-
mits of a large accumulation of
momentum between each sawing
action, and practically gives it
ample power for the 10 seconds
duration of work.
Another wheel designed to
run a fan blower was described
in the same volume. It is shown
in Fig. 154 in section. The
steam enters the centre of the wheel and escapes by curved arms, as shown. The
longitudinal section shows the hollow shaft through which the steam enters the wheel.
It is evident that in both these the idea of the design is to obtain a wheel which
Watt's Turbine
127
could be run at high speeds. Unfortunately, in neither case was the diameter of the
wheels stated ; but 2000 revolutions is given as the speed.
It was pointed out at the time that the speed of the wheel would require to be some
large fraction of the speed of issue of the steam from the nozzles to get any efficiency,
and that speed was calculated at 1200 per second.
About that time much attention was paid to steam turbines. Pilbrow, the great
turbine inventor, had made his experiments and patents
known.
De Kempelen had evidently seen the impossibility of
obtaining efficiency with steam, and proposes, but un-
fortunately without illustrations, a method whereby he
uses the steam to propel water alternately from two
vessels. Through his turbine, he says, the steam is to
be admitted alternately from the boiler to the two vessels,
and presses upon the surface of the water, forces it into
the turbine ; the water is to be returned to the receivers
and worked hot.
In the same year James Watt patented this same idea.
It is shown in Figs. 155 and 156 in diagram. A vessel
A, B, D, E, C is mounted upon a pivot J, and supported by
collar K at upper end. The vessel is divided into two,
with an opening HH in each at the side for the escape
of the fluid, and there are two clack valves F and G for
the entry of the fluid, which surrounds the rotating
vessels up to the level of the overflow O. Steam is sup-
plied by pipe L alternately to the two divisions of the
vessel, so that when one is filling up with fluid the other
is being emptied by the steam pressure.
Watt proposed to use mercury, oil, or water as the
acting fluid.
Both these proposals when examined show that
they had hit on the idea of the pulsometer in a primitive
form. In point of fact, a pulsometer pivoted at the bottom, and with its water inlet
under water level and its steam inlet in a stuffing box free to revolve, would constitute
a modern and more effective turbine on the Kempelen and James Watt plans for fluid
turbines. Instead of the ordinary outlet there would be two nozzles, one on each vessel
for the ejection of the water at a tangent to the vessels.
Trevithick's patent describes Hero's turbine
without much improvement upon Hero's machine.
Ericsson, in his patent of 1830, shows considerable
improvement upon Hero, as will be seen from the two
figures one a cross section, the other a view of the
wheel. He forms conical nozzles either on the side of the
wheel, or on the outer periphery as shown at rr ly Fig. 1 57.
The fixed vanes J are carried on a fixed sleeve a on the shaft, which moves easily
within the sleeve. The steam enters the nozzles from the outside or inside, but the
fixed vanes are always placed so as to prevent the rotation of the steam.
These fixed vanes reveal the fact that Ericsson had found it necessary to stop the
rotation of the steam against a fixture. The author of this book made some experiments
with a similar wheel designed for high speeds in 1890, and found that without a set
of fixed vanes in the case where the steam enters from the outside and escapes by
the axis it was not possible to get it , to work, for the steam whirled round and
round until, by its own friction against the wheel, it lost its energy and escaped,
Sectional Elevation.
FlG. 155. James Watt's Turbine.
FIG. 156. Steam Inlet of Watt's
Turbine.
128
Modern Engines
and the friction of the steam balanced the pressure on the jet, and there was no
motion.
Alexander Morton, of Glasgow, the inventor of the ejector condenser, also found this
out, for in his first patents he describes a steam turbine on Hero's
plan, in which he used a series of concentric wheels, somewhat like
Ericsson's wheel (Ericsson, by the way, was the inventor of the screw
propeller). The idea is that the steam entering the inner one will
propel it by pressure on the nozzles, then expanding", do the same in
the next outer ring-, and in the next. It was found, however, on trial
of this series of wheels, that the effect was nothing 1 ; the steam simply
expended its energy in rushing round in the opposite direction to
that which the wheels ought to have rotated, and its friction counter-
balanced the steam pressure on the outflow nozzles. In his next
patent we find fixed
blades inserted to
stop the rotation of
the steam. There
are three wheels con-
centric with diverg-
ing nozzles on their
periphery, and be-
tween these there
are fixed blades and
an outer row of
fixed blades on the
casing ; these stop
the whirl of the
steam.
On examination of these early
machines it will be seen that
although capable inventors, from
Watt's time forward, have tried
to improve upon Hero, they have
made very little progress. Von
Rathen made Hero turbines with
diverging nozzles ; Ericsson and
Morton also employed diverging
nozzles, but a little reflection
will show that the gain to be
obtained by diverging nozzles in
this type of turbine is small
in fact, nil. All of these in-
ventors failed to grasp the fact
that the improvements required
were such as would enable the
wheel to run at enormous velo-
cities at the steam orifice. We
will now take Parsons' design
of pressure turbine, Fig. 158; it is a reaction engine, consisting of one or more pairs
of arms mounted upon a spindle, and so arranged that steam is admitted to the interior
of the said arms by suitable passages in the spindle, and is discharged from each pair
by apertures at the ends of the arms arranged to discharge at or near the tangent to the
circle swept by the arms.
Parsons' Turbines 129
The arms are arranged within a hollow case or series of cases. In one con-
struction (Fig. 158) a series of circular cases D, D 1 , D 2 check into each other circum-
ferentially, and form one larger case divided by partitions into a number of chambers.
One spindle passes through the series of chambers and carries upon it a pair of arms
in each chamber. One arm, however, may be carried in each chamber provided with
a suitable balance piece or weight ; or any number of arms may be so carried, pro-
vided that balancing is carefully allowed for. One end of the spindle passes through
a high pressure steam chest B, and a number of grooves turned in the spindle work
in a bearing having corresponding grooves ; one set of such grooves E prevents
the high pressure steam from leaking into the atmosphere from the said high pressure
steam chest, and another set of grooves E 1 prevents the high pressure steam from
leaking into the first chamber. Between the two sets of grooves and the spindle
radial apertures C 1 are drilled to a hollow, bored out or formed in the spindle.
Similar apertures C 2 lead from the said hollow to the interior of the boss off the
first pair of arms ; the arms A, A 1 , A 2 are cast hollow with a passage of uniform
area and of flattened section. The exterior is also of a flattened section sharpened
towards both edges in order to provide arms capable of passing rapidly through
the steam in the casing without creating much resistance. That is, the arms are
arranged to move on edge through the steam within the casing, and the necessary
section is provided by making them sufficiently wide. These arms rotate at about 6000
revolutions per minute, and hence it is necessary that there should be a considerable
margin of strength to resist the great centrifugal force. To provide this strength the
exterior figure of each arm is conical, that is, the arm is thicker and broader at the
root near the boss than it is at the extremity. Slots a, a 1 , a? are cut in the edges of
the arms to furnish discharge orifices and permit the steam by its reaction to propel the
arms and hence the spindle. The steam passing from these orifices fills one chamber
of the casing and passes thus through the boss of the next arm, which boss is turned
to fit a bored opening in the wall of the chamber. A series of passages through the
said boss permit steam to flow from the first chamber into the interior of the second
pair of arms ; thus the steam discharges by emission apertures into the second chamber
of the casing. From the second chamber the steam passes similarly through the boss
of the third pair of arms, and by emission jets to the third casing, whence the steam
may pass to the fourth pair of arms and chamber, and so on.
As the steam passes in this manner from chamber to chamber through each pair
of arms the pressure falls, and each pair by reaction adds to the energy of rotation
of the steam wheel or turbine spindle. The interior of the passages and emission
apertures of the arms are arranged in continually increasing area in order to allow
for expansion.
It has been proved that steam cannot flow at more than 1440 feet per second from
an orifice, and that an expansion of i to 1.63 is sufficient to gain this velocity ; therefore
in a compound turbine of the reaction jet type, when working with 100 Ibs. steam and
exhausting into the atmosphere, it is necessary to employ at least five sets of arms in
separate cases, thus limiting the expansion in each issue to under the above-mentioned
ratio of expansion. For this reason reaction wheels like Hero's engine, employing one
wheel only, have not given good results.
The boss of each arm projecting through the partitions is arranged as a nice fit,
and grooves or serrations may be turned in the bored apertures in the chamber walls ;
a groove packing is used for resisting high steam pressure. Such a packing is described
in Specification of Patent, No. 1 120 of 1890. To return to the turbine, we shall quote only
two of the claims :
i. A steam wheel or engine consisting of a number of hollow reaction arms
arranged in sequence in a number of chambers, the said arms having discharge
apertures of increasing area from the high pressure to the low pressure end.
VOL. i. 9
130
Modern Engines
2. A steam wheel or engine comprising' a series of chambers in which reaction
arms are caused to rotate, the said chambers being provided with drainage apertures
HH, and drainage boxes or spaces GG, substantially as hereinbefore described.
A speed of 1400 feet per second at the periphery is necessary for the full efficiency,
whereas none of these wheels could with safety run above 100 per second in peripheral
speed in feet.
The Hero engine in the form shown in Fig. 153, but designed for a speed of about
30,000 feet per minute at the periphery, would be as efficient as the De Laval or
Parsons' turbines without any special nozzles.
We may pass over all the abortive attempts of many inventors to make a successful
Hero turbine. Until 1893 we find only Mr. Parsons' patent for a steam turbine on
Hero's method. It is shown in Fig. 158, a longitudinal and cross section. Pilbrow,
fifty years before, had shown that by putting turbine wheels in series a greater economy
Longitudinal Section.
FiG. 158. Parsons' Turbine Hero Type.
Cross Section.
could be effected, for the difference of pressures could be reduced, and the velocities also
to some extent reduced.
Mr. Parsons in this specification calls the wheels reaction engines ; strictly speaking,
they are purely direct pressure wheels.
The reasoning regarding the expansion from 100 Ibs. to atmospheric pressure
in five stages that is, in a series of five wheels is correct enough ; but while that
arrangement would be more economical in steam than one wheel, the necessity
for running the wheels at some large fraction of 1440 feet per second still
remains.
This turbine, in order to run at a much lower speed than 1400 feet per second,
would require not five wheels in series, but fifty, as the speed and pressure of the steam
would have to fall by very small steps to get the small velocity of flow.
The velocity is equal to 64. 4^* for air with a density of 0.080728 Ibs. per
cubic foot, where x = inches of water required to balance the difference of pressure
Pp\=Pr If we were working with compressed air a difference of p 100-^ = 92,
/> 2 = 8 Ibs., the water pressure column to balance i Ib. = 2. 24 feet, or about 27 inches
of water.
To calculate for steam, we take the volume of steam at the given pressure, and thus
if the difference of pressure between the first wheel and the next were 100-93 = 7, or
nearly 0.5 of an atmosphere, at 100 Ibs. pressure, the volume of a cubic foot of steam
Turbine Wheels
is 272 times the volume of water; hence we get V = 8^272 x 0.5 x 32 = 528 feet
per second. We would require to calculate each step separately as the pressure
dropped.
It is thus seen that to reduce the speed to 500 per second, or 30,000 feet per
minute, would require a series of wheels in which the steam expanded by a drop of less
than 7 Ibs. between wheel and wheel at the start, and a less and less drop towards
the exhaust for the volume of steam at atmospheric pressure equals nearly 1800 times
that of water.
The greatest fall allowable is at the high pressure end of this series, and that is
7 Ibs. in this case ; so that it is quite clear that if any efficiency is to be got out of a
series of Hero wheels their number must be great, and the difference between each in
pressure small ; for the speed in feet could in few cases be allowed to reach even 500
per second. Mr. Parsons proposes to run the wheels at 6000 revolutions per minute.
Now, if we allow an efficiency of, say, half, ^ = 264 feet per second, or 15,840 per
minute; then
= 2.64 feet would be the circumference of the circle described by
_
6,000
the wheels at the diameter of the orifices. This diameter would then be about 10 inches.
The difficulties in the way of making
a turbine successfully on this plan are
FIG. 159. Double Wheels. Hero-De Laval.
FIG. 160. Double Wheels. Hero-De Laval.
many, and no proposal up to the present time has done much, if anything, to over-
come them.
Working with an inexpansible fluid like water, it makes an excellent turbine ; the
difficulties begin when a fluid which must be worked expansively is to be used.
The best that can be done with it at present is to compound a Hero wheel with a
purely impulse wheel, that is, to combine a Hero wheel with a De Laval wheel, and thus
reduce the revolutions by one-half. This idea occurred to the author some two years
ago, and has since been patented by Mr. Parsons.
A diagram showing the principles of this turbine is shown in Fig. 159. The inner
wheel is a Hero wheel with diverging nozzles, in which the steam acquires its maximum
velocity before striking the outer wheel, with plain curved blades like the De Laval
wheel.
If we take a De Laval wheel and place it face to face with a Hero wheel in which
the noozles are made diverging as in a De Laval wheel, the two wheels when steam
is supplied to the Hero wheel will revolve in opposite directions with about equal
torque, so that they may be geared to one common shaft, or each may drive a separate
dynamo.
The turbine of Mr. Parsons combines two wheels in this way.
Modern Engines
FIG. 161. Pilbrow's Double Turbine.
Mr. Parsons' patent specification describes his arrangements of the double wheel
type (Fig. 160) as follow:
In fluid pressure turbines of the De Laval type the method of securing a high
relative velocity between jet and bucket with reduced skin frictional losses, by rotating in
opposite directions the element-carrying nozzles and the element-carrying buckets or vanes
against which the fluid impinges, is substantially as described. The arrangement is
preferably such that the working fluid, after impinging on the vanes, passes to the exhaust
without interfering with the action of succeeding jets. The shafts of the counter-
rotating elements may be co-axial, and
may respectively carry the two reacting
parts of a dynamo machine. Reversing
turbines of this class may comprise a
separate set of nozzles (which may be
fixed), supplied from a separate pressure
chest, and adapted to direct working
fluid against the reverse sides of the
buckets, or against a row of reversely
set buckets. The inventor says: "By
my new method I produce a steam
turbine of remarkably simple construc-
tion, which in operation allows great reduction in the losses from the skin frictional
resistance ; and this renders the single-expansion type for the first time admissible for
direct coupling to dynamos or screw propellers or other purposes, where the speed
of revolution is moderate. Apart from power saved in gearing dispensed with, such
turbines have only one-fourth the disc stresses and one-fifth the skin resistance of the
simple form."
This principle of construction I have carefully investigated, and my conclusions are,
that it introduces insurmountable mechanical difficulties. It only reduces the revolu-
tions by one-half, while it at the
same time introduces a hollow
wheel with steam inlet around the
shaft, subject either to much leakage
or great friction.
Pilbrow again anticipated this
placing of two wheels together in
order to get smaller speeds or a
"high relative velocity," and his
plan is shown in Fig. 161, in dia-
gram. The steam jet enters the
first wheel and propels it in one
direction, it then enters the second
wheel at full velocity and propels it in
the opposite direction, and the two
wheels can be geared together so that they add their power to one common shaft. With
the knowledge given by the De Laval specification regarding diverging nozzles we would
now make the two wheels as shown in diagram, Fig. 162. In this turbine the steam enters
by a converging jet into a wheel with diverging passages, so that there are no difficulties
about the steam inlet as there is with a hollow wheel. The steam in passing through
the first wheel propels it by unbalanced pressure, and also expands to atmospheric
pressure, and strikes the second wheel with full velocity as in the De Laval wheel ; we
thus get the double effect. In this turbine the two wheels are face to face and overhung
on the end of shafts, so that they can have some free lateral motion, and thus find their
own centre of gravity, around which they revolve with silence.
Outlet Side.
jr n i et ^^
FIG. 162. Double Face to Face Turbine.
Turbine Wheels Double
FlG. 163. Combined Wheels Type of Turbine.
A diagram of this turbine is given in Fig. 163, in which B is the one wheel and C
the other, mounted on separate shafts ; D is the steam chest and nozzles which con-
verge on the first wheel B. The two wheels are geared by helical cut gearing 1 on to
one shaft, the one wheel F being an external toothed wheel, and the other G an internal
toothed wheel ; with same pitch line they thus both drive in same direction.
This design is the author's, and experience proves it to be of considerable advantage
for smaller powers. It will be more fully described later on.
In the ordinary De Laval turbines, if we take, for instance, a 10 horse-power machine,
the peripheral speed is 617 feet per
second, the revolutions 400 per second,
or 24,000 per minute. In my combined
wheels, in which the nozzles rotate
backwards at same speed as the wheel
goes forward, the speed of each wheel
is 200 feet per second and 12,000
revolutions per minute. And as the
disc stresses are as the square of the
speed, they will be reduced to a fourth
of what they are in a single wheel, and
the skin friction to a fifth.
The gearing is objectionable for
large sizes. Another direction for ex-
periment lies in the use of long nozzles
wound into a scroll on the periphery of
a disc, or cut on the face of a disc and
closed by a steel disc. The steam
being continually deflected exerts a
tangential pressure. A rough test of this principle gave fair results, as also did a test
with a screw cut on a barrel and gradually increasing in depth towards the outlet. The
screw was cut on a brass barrel, 6 inches diameter outside, 5 inches inside. The screw
was cut with a square thread ^-inch pitch, the thread being T \- inch thick, and the space
between y\ inch ; the depth of the thread was inch. After the barrel was threaded it
was tapered down in the lathe until the thread at one end was only / T of an^inch deep,
so that it gave a long spiral taper-
ing nuzzle, tapering 16 to i from
end to end. A brass sleeve bored
to the same taper was slipped over
the barrel hot, and shrunk on to it,
thus closing the long spiral nozzle.
The barrel was mounted on a
spindle, one end of which was
hollow for steam admission to the
narrow end of the spiral. The
steam expanding and flowing 1 along
the spiral exerted a tangential pres-
sure, which propelled the barrel on its axis. This turbine is shown in Fig. 164. It was
not so effective as the scroll.
Professor Hewitt tried a screw form of turbine, but from the illustrations given
of it, of which Fig. 165 is one, the chief direction of pressure would be axial with a
long pitch screw working a fixed barrel, whereas it is tangential pressure which is
necessary to rotate the shaft.
Mr. Morton tried an experiment on a curved nozzle with steam. He let little glass
tubes A, B, C, D, E, and F into the inner and outer curved walls of the tube or nozzle
FIG. 164. Screw Type of Turbine.
Modern Engines
G, as shown in Fig. 166. The tubes A, B, and C on the greater curved side were
U-tubes, and had mercury about half filled up ; the tubes D, E, and F on the smaller curve
(-B m rn
FIG. 165. Professor Hewitt's Screw Type of Turbine.
FIG. 166. Morton's Experiment.
dipped into mercury. On blowing steam through the nozzle a considerable pressure
was found on the outer curve, and a partial vacuum on the inner curve.
The position of the Hero turbine to-day may be
summed up by saying that it is at its best when con-
structed with nozzles diverging and delivering into
another turbine wheel with curved buckets these two
wheels geared to one shaft or driving two dynamos in
series with each other, and rotating- in opposite directions.
The reaction turbine is one in which the steam
passes from the guide blades under pressure, and with
some velocity and pressure, so that the driving effort
is partly due to unbalanced pressure, as in the Hero
machine, and partly due to the impulse from the
weight of steam thrown against the curved moving
blades.
The purely impulse turbine, the De Laval type, was very thoroughly investigated by
Pilbrow, and his results given in his Patent Specification, No.
9658, 1843. He found out that a jet of steam gave a greater
impulse to a vane when the vane was some distance from the
orifice, and with 60 Ibs. pressure ; he states that the best dis-
tance is f inch ; and with an orifice of f inch diameter he
obtained an impulse of 14 Ibs. And he calculated correctly
that to obtain efficiency the speed in feet per second of the
vanes would require to be 1250. Pilbrow allowed the jet to
expand and acquire a high velocity before striking the vanes,
by shifting the nozzle back until he obtained "the best dis-
tance from the orifice ; f inch from the vanes." It will be
seen from this that the steam would have time and space to
expand and acquire velocity before entering 1 the wheel. He
also shows fixed vanes "to lead away the steam." Their
object is not very clear. Pilbrow then put wheels in series, so
that the steam entered first one wheel, and then the next, and so
on through all of them. We have already referred to his two
wheels in series going in opposite directions face to face.
Robert Wilson, of Greenock, in 1848 patented turbines in
which the steam was successively expanded in one wheel, and p IG- 167. Wilson's Turbine.
from wheel to wheel.
A part section of the first form is shown in Fig. 167. The steam enters at I, goes
through the wheel to the inside, re-enters at s l , goes through the wheel to the outside at
Turbine Wheels in Series
T 35
m 2 , re-enters again at r 2 , thence to 2 , and so on, zigzagging out and in, as shown by
the arrows, till it finally reaches the exhaust.
In another wheel (Fig. 168) Wilson employed concentric rows of curved vanes alter-
nately fixed and movable, or alternately
fixed on two discs or shafts capable of
rotating in opposite directions. This
turbine has been resurrected with some
effect in America recently. The diffi-
culty with this form is the rapid ex-
pansion of the steam as it works its
way out. Like all multiple wheel tur-
bines, this makes them more suitable
for large than smaller powers. In
larger sizes the inner radius r can be
made a larger fraction of r r
Then Wilson anticipated the parallel
flow turbine, with alternately fixed and
movable rows of blades, enlarging their
area as he reached the exhaust end.
This is shown in Fig. 169, in which
fixed vanes i, 2, 3 are fastened to the
outer casing H, and movable vanes
4, 5, 6 are fastened to a central shaft G. FIG. 168. Wilson's Turbine.
Steam enters at S, and expands through
the wheels, each successive wheel being of larger area of wheel blades than the preceding
one. There we have the whole of the elements of the best steam turbines now made.
All that was required was
to design the wheels for high
velocity, and the bearings F
also for a high velocity.
It will therefore be seen
from this brief review of the
pioneer work in steam tur-
bines that most of the prin-
ciples and designs are very
old, and that the pioneers
understood the necessities
of the matter, especially
Pilbrow and his follower
FIG. 169. Wilson's Parallel Flow Turbine. Wilson. But they could
not make the machinery
accurate and fine enough to get the high speeds necessary for practical success.
PRACTICALLY DESIGNED WORKING TURBINES
To Parsons belongs the first place in turbine making commercially and scientifically
practical. His patents from the first show that he set out to realise the dreams of his
predecessors. First, to design and construct the wheels to work successfully at the
necessary high speeds ; and second, to reduce that speed as far as possible without loss
of efficiency.
De Laval set out also with the first end in view, and accepted the high speed as
unalterable. Both succeeded in making practicable turbines of high efficiency, the first
more especially for large powers, and the second more especially for smaller powers.
136 Modern Engines
There are other practicable turbines more recently introduced, the Rateau and two
or three others worth describing.
Matters in turbine progress stood much in the stage at which Pilbrow and Wilson
left them until 1884, when Mr. Parsons commenced to wrestle with the difficulties.
Greenock men laughed at Wilson's inventions, and his efforts were fruitless to him, but
the day came which he had hoped for, when a steamship, the King Edward, touched
at Greenock propelled by a Parsons' steam turbine, a highly developed specimen of
Wilson's type shown in Fig. 169.
It had taken 60 years to develop from Wilson's crude design to the tnrbine on that
ship.
Mr. Parsons laboured himself for fourteen or fifteen years before his great improve-
ments were fully recognised by marine engineers. His patents had expired before he began
to reap his just rewards. Fortunately, his fundamental one was extended for five years.
This slow progress of inventions in the engineering line is a most remarkable
phenomenon. It is difficult to find a reason for it. We as a manufacturing nation go
on for generation after generation using and making old things in the old ways as long as
it is ever possible to sell them. Some of these things are no doubt past improvement and
excellent in their way, but the majority are subjects for improvements. But very slowly
and very reluctantly are the improvements adopted. It is wonderful !
The internal combustion engine has the same tale about it. It has in one form or
another been in use for the past fifty years, and only now has it been recognised as a
prime mover on a large scale and its possibilities recognised. The motor car dates back
to the time of James Watt and Murdoch, yet it is only now approaching the time when
it will become an important industry. The electric tramways in towns could have been
laid down by electrical engineers in quite as complete a system as they are in now in
the year 1888, and in quite a practicable manner in 1884, yet it took years and years to
introduce them, and they only became fully recognised as the best system of rapid transit
in towns about 10 years after their perfection. And then, of course, the authorities
began to tumble over each other in their eagerness to secure the advantages of the
" novel" system.
The same story could be told of the dynamo and the electric motor. The pioneers
did a lot of hard work for nothing from 1874 till 1886. And now every engineering firm
with a vertical drilling machine and a turning lathe makes " the best motors and dynamos
in the "world" " our own special design," and so on.
At this moment, when the turbine has been recognised through Mr. Parsons' efforts
to be "the steam engine of to-day," and we may say " to-morrow," there is an amusing
keenness among engineering firms to enter as rival manufacturers, and by and by each
will have a "special design of his own," even although the difference between them is,
practically, only in the colour of the paint. The history of the first adoption of the
Parsons' steam turbine on mercantile vessels is interestingly and concisely given in
Mr. Archibald Denny's remarks on Mr. Parsons' paper on "Marine Steam Turbines,"
before the Institute of Engineers and Shipbuilders in Scotland, igth February 1901.
As a matter of history it is worth recording :
"This was not the first time that he had heard Mr. Parsons lecture on this subject:
he heard him read a paper to the Institution of Naval Architects about three years ago,
and naturally he was impressed with the advantages of the turbine. Curiously enough,
his firm had used one for driving the dynamo on board the s.s. Duchess of Hamilton,
built some years ago, and it worked exceedingly well. When he heard Mr. Parsons
read his paper before the Institution of Naval Architects he was fired with the ambition
to work with him for the success of the turbine. Several months ago his firm got in
touch with Mr. Parsons, and they made up their minds that if possible they would jointly
get a turbine vessel built for the mercantile marine. They naturally approached the
railway companies in the first instance, but they affected a terrible amount of modesty,
Parsons' Turbines 137
and each company was anxious that somebody else should make the first experiment.
So the matter was hung up, and he was beginning to despair of success when Mr. John
Williamson came forward and lent them his aid. Mr. Parsons, Mr. Williamson, and
his firm, having laid their heads together, resolved to build a mercantile turbine vessel,
and he felt it was very gratifying indeed that the Clyde had been favoured in being the
pioneer in this enterprise. It was gratifying to think that this was taking place during
the era of the Exhibition, when many people would be able to see the advantages of
this system."
The attitude of the railway companies referred to is the common one.
To proceed with the turbine construction, we will take the first patent of Mr.
Parsons, as that marks an epoch in the art of turbine making.
This invention has reference to motors of the turbine type ; that is to say, to
motors in which the actuating fluid operates between fixed and moving vanes or blades.
When elastic fluids such as gas and steam are used in a motor of this description, it is
necessary for economical working that the peripheral speed of the motor should be nearly
as great as the velocity of the gas or steam, due to its effluent pressure, a speed which,
except with very low pressure, is practically impossible.
According to this invention, to obtain a low effluent or terminal pressure while
using a comparatively high initial pressure, a compound motor, or a combination of
motors, are so arranged that the same actuating fluid operates therein in a successive
manner, undergoing expansion and falling in pressure in each, until it leaves the last
at a velocity not greatly above that which is practically attainable by the motor itself,
although greatly above that practicable with a motor having oscillating or recipro-
cating parts. By this arrangement each motor, or successive portion of the compound
motor, utilises a portion of the energy of the fluid, and thus, instead of the greater part
being wasted as heretofore, it is successively drawn upon until a comparatively high
efficiency is obtained.
The motors, or successive portions of the compound motor, may be arranged
either upon one common shaft or upon different shafts. In the former case the first
will deliver directly into the second, and the second into the third, and so on, the moving
vanes of the second (say) rotating between its own fixed vanes, and those of the third,
and similarly for the others ; the space for the actuating fluid increasing either con-
tinuously, or step by step.
This increase may be conveniently gained either by an increased area, or by an
increased pitch of the blades, or by an increased area and pitch combined.
The successive motors, or portions of a compound motor, are arranged in such wise
as to form an approximately cylindrical figure, the whole being mounted by preference
upon one and the same shaft ; the first delivering into the second, the second into the
third, and so on.
Each motor or portion comprises a set of fixed and a set of moving vanes, the
direction of motion of the actuating fluid being generally parallel, or approximately so,
to the axis of the combined motors.
Conveniently, each set of moving blades can be formed out of the solid metal on
the circumference of a brass or steel disc, the blades extending only about one-third of
the breadth of the disc, and the blank portion forming part of the moving cylinder
beyond which the blades extend.
Likewise, the fixed blades can be formed by cutting each row internally on a ring
which is afterwards cut diametrically into two parts, one to be joined to the top half of
the casing, which is divided into two parts by a longitudinal joint, and the other into the
lower portion of the casing. When all the parts are put together the fixed portion
forms a hollow cylinder with projecting rings of blades, and the moving portion a solid
(or hollow) cylinder, also with projecting rings of blades.
To balance the end pressure upon the cylinder, two similar sets of rotary parts are
138 Modern Engines
mounted upon one shaft, one set being" placed at each side of the inlet for actuating 1 fluid,
in such a way that the entering stream shall divide right and left, and the exhaust take
place at both ends. Any end pressure not thus balanced, or due to external causes, can
if desired be balanced by pressure of the exhaust fluid acting between the end of the
moving cylinder and a collar of smaller diameter than the cylinder. Thus, should the
cylinder be displaced endwise in either direction, the exhaust will be checked at that end,
and in this way a compensation will be automatically effected.
As the speed of the motor will be necessarily and designedly high, and perfect
balancing of the moving parts would not be practicable, the bearings are given a certain
very small amount of elasticity or play, combined with a frictional resistance to their
motion. Thus the cylinder will be enabled to rotate around its centre of gravity, instead
of its geometrical centre, if the two be nearly coincident, and the vibration to which it
may be subject will thereby be damped or modified.
The lubrication is effected by forcing lubricant to the parts to be lubricated, and
for this purpose a pump can be employed. Conveniently it may be a centrifugal pump,
of the type in which the fan is constructed like a screw propeller mounted on the end of
the shaft. From this pump the oil will be taken to the bearings as required, with a
constant circulation. The oil can also be used as a carrier of heat, to reduce the tem-
perature of the parts liable to grow hot. If the pump be of a kind that will not lift, a
suction fan mounted on the motor shaft may be used to raise the oil on the suction side.
This fan may also be employed to govern the supply of actuating fluid, by causing
variations of pressure, according to the speed at which it is driven, on a diaphragm
or piston in connection with the throttle or supply valve. The speed of the motor may
be regulated by an adjustable spring acting against this varying pressure, or by the
admission of air through a graduated regulating tap into the exhausting side of the fan.
To prevent leakage past the shaft at the end covers of the casing, which, when
steam is the actuating fluid, would be inconvenient, annular recesses are formed in the
covers around the shaft ends, and place these recesses in communication with a pipe in
which a partial vacuum is maintained by suitable means such as a steam jet. Any
steam which enters the recesses will thus be drawn away without apparent leakage.
Motors, according to this invention, are applicable to a variety of purposes, and if
such an apparatus be driven it becomes a pump, and can be used for actuating a fluid
column, or producing pressure in a fluid. Such a fluid pressure producer can be com-
bined with a multiple motor according to this invention, so that the necessary motive
power to drive the motor for any required purpose may be obtained from fuel or com-
bustible gases of any kind. For this purpose the pressure producer is employed to force
air or combustible gases into a close furnace of any suitable kind such as used for caloric
engines, into which furnace there may or may not be introduced other fuel (liquid or solid).
From the furnace the products of combustion can be led, in a heated state, to the multiple
motor, which they will actuate. Conveniently the pressure producer and multiple motor
can be mounted on the same shaft, the former to be driven by the latter ; but the
inventor does not confine himself to this arrangement of parts.
The blades should be of material that will withstand high temperatures, or water or
other fluid may be employed to cool the blades. This may be done by providing, in the
cylinders that carry the blades, channels or passages for the circulation of the cooling
fluid, which, in the case of the rotary cylinder, may be supplied through a passage or
passages in the shaft carrying the cylinders.
It will be observed at once that the inventor tackled the points requiring improve-
ment.
First, to reduce the velocity of steam flow as far as possible, so that the turbine
speed would approach in peripheral velocity at least half the velocity of the steam. This
he does, as had been done before, by forcing the steam to pass in series through many
wheels. The more it is baffled on its way to the exit by deviating its course by wheel
Parsons' Turbines
1 39
blades, the slower is its progress, and the more of its energy is exhausted by giving- it
away to the wheels that is the whole philosophy of the wheels in series.
To balance the steam pressure the turbine was made double ended, steam entering
at the middle and balancing. This is still a good plan for pressure turbines ; better, in
the author's opinion, than the dummy plan later adopted. Then he proceeds to the real
problems to be solved. He knows that a perfect balancing of the rotating parts is
mechanically impossible; he therefore devises bearings with some elasticity or play in
order to allow the "wheels to settle," that is, to rotate round their centre of gravity.
FIG. 170. Parsons' First Turbine.
The penultimate paragraph in this patent is of interest ; it shows that the turbine
is reversible, and will act as a pump if driven by power. This, we have seen in Fig.
145, has been accomplished, and then he foreshadows a hot gas turbine in which the
air is to be forced into a closed furnace supplied by fuel ; the products of combustion
are then to be led into the turbine which drives the pump, and so drive the turbine.
This scheme has never been put into practice, so far as can be ascertained, and it was
dropped out of the specification. It contains, however, the germ of the gas turbine or
internal combustion turbine.
Fig. 1 70 is a longitudinal horizontal cross section of the turbine, with two sets of wheels
FIG. 171. Parsons' Turbine, with Expanding Wheels.
right and left of the steam inlet S r The b, b v 6 2 are the movable blades, and thef,f v f z
the fixed blades. These blades were originally cut out of the solid brass rings, and
were merely flat blades at an angle of 45 with the axis.
In this form the diameter of the wheels are all alike, and expansion is obtained by
making the blades deeper radially as they near the exhaust end.
Later on the plan shown in Fig. 171 was adopted, in which not only are the blades
increased in depth radially, but they are also increased in diameter in 3 sections, making
it, as it were, triple expansion. And in order to preserve the balance of steam pressure,
140
Modern Engines
steam passages are formed in the casing connecting the annular space between each set
and the exhaust ends, so that the pressures are all equalised.
For giving to the bearings a certain very small amount of elasticity or play,
combined with a frictional resistance to their motion, in the arrangement of motor now
under description, this is done by constructing the bearings of shaft s in the manner
represented in Fig. 172.
I is a light bush, outside of which are placed metal rings or washers KK 1 . The
alternate washers K are slightly larger than the washers K 1 , so that the alternate
washers K fit the casing but not the bush, and the washers K 1 fit the central bush but
not the casing. L is a spiral spring, and
M is a nut. By these the rings or washers
KK 1 are pressed tightly together. Thus the
bush is capable of a slight lateral move-
ment, but this movement is resisted and
controlled by the rings or washers and their
mutual friction. The amount of liberty given
to the rings or washers may be, say, y^
part of an inch diametrically, but it may be
more or less according to circumstances as
may be found desirable in practice. It will
be evident that any movement of the central
FIG. 172. Turbine "Ring" Bearing.
bush in a lateral sense will be opposed by the frictional resistance of the rings or
washers. In lieu of, or in conjunction with, rings or washers, such as KK 1 with spiral
spring and nut, a steel cage such as illustrated in Fig. 173 may, in some cases, be used to
surround the bush i for the purpose of maintaining it in place whilst admitting of very
slight lateral play or elasticity. This cage n as represented is formed with longitudinal
slots n 1 , so that the intervening parts n 2 will act as springs, pressing tightly against the
central projecting part z 1 of the bush, thus resisting any lateral movement principally by
elasticity of the cage. To obtain very quiet running, however, it is preferable to use a
combination of the two arrangements, constructed by placing a steel cage such as n or
its equivalent around the washers KK 1 in order to check the communication of vibrations
from the rings or washers to the bearing itself.
In Fig. 170^ are metal washers fitting easily on
shaft j to prevent leakage of steam past the
bushes y l y l . oo are drain passages from each
end of the hollow cylinder. They pass over the
top of the motor and meet at its centre. Their
purpose is to drain away any of the steam that
may escape between the shaft ^ and bushes y l
fixed in the motor. This is effected by an arrange-
ment shown clearly in Fig. 174. Where the drain
passages oo meet, there is provided an ejector or steam nozzle p, to which steam is
admitted from the supply branch. This steam flows into an exhaust pipe g, and the
partial vacuum thus created causes the steam to flow from the passages oo and to pass
away through the drain pipe q.
His next form of turbine is shown in Fig. 175, a series of radial flow turbines ending
in a large disc series of outward flow turbines like Wilson's (Fig. 168). The construction
is thus explained :
It consists of a new arrangement of balance piston designed to minimise leakage ;
a method of bolting the rings or discs together on the turbine spindle ; the use of a
large disc with blades to expand the steam to a very low pressure when a condenser is
adopted, and an application of the principle of the relay to solenoid electrical governors
for steam turbines.
FIG. 173. Turbine "Cage" Bearing.
Parsons' Turbines
141
Fig. 175 is a sectional elevation of this turbine with balance piston shown thereon,
also illustrating the method of bolting the rings or discs together on the turbine spindle,
and use of a large disc with blades to expand the steam to a low
pressure when a condenser is adopted.
Fig. A shows the balance piston E on a larger scale in part section.
Another part of the invention relates to fastening the discs on the
turbine spindle, and for this purpose the spindle A has a collar c l at one
end, and into this collar c l pass long bolts or studs cc, which pass
through holes in the turbine discs as B, B 1 , B 2 , B 3 , B 4 , B 5 , and B 6 , and
through holes in the balance piston E. The discs are thus screwed up
and are firmly held on the spindle, together with the balance piston.
The invention also relates to turbines intended for great expansion
and exhausting into a condenser. In these turbines a large disc as B ti
(Fig. 175) is fixed on one or preferably both its faces, and the steam,
after passing through a sufficient number of turbine pairs to reduce its for 74
pressure to atmosphere, or nearly so, passes by one or both faces of ste;
the large disc, and by its impact between fixed and moving blades, gives
up its energy, and passes to the condenser, being first reduced in pressure until very
little energy is left to be carried to the vacuum. The large disc is arranged as an out-
ward flow radial turbine, and the passage communicating with the condenser surrounds
it so that the steam passes on one or both sides of the disc from
A ^<<<^x tne tur ki n e through the fixed and moving blades.
This same outward flow disc turbine is shown at K in Fig.
158, already referred to in conjunction with his Hero type of
turbine.
Mr. Parsons also patented an inward flow steam turbine, much
41 i et^L" l*ke a series of vortex wheels as used in Professor James Thomson's
FlG. 175. Parsons' Radial Flow Compound Turbine.
vortex turbine. It is shown in Fig. 176, a longitudinal section and a part cross section.
In steam turbines of the kind at present in use the steam is caused to perform
work by alternate impact upon and reaction from a large number of rotating vanes or
blades, and the steam is suitably directed or guided by fixed vanes or blades of con-
venient shape and disposition. In such motors it is impossible to obtain the maximum
Modern Engines
142
economy in the consumption of steam unless the clearances between the rotating- blades
and the enclosing- conical, disc, or cylindrical surfaces and the fixed blades and rotating
surface is very small. This necessitates considerable delicacy in the adjustment and
accuracy in the construction of the motors, and renders it difficult to keep them in such
condition as to bearing- surfaces that the best economy may be obtained in practical
work through a long- term of years.
As applied to turbines of the inward flow type, a series of inward flow turbine
wheels, upon one rotating shaft or spindle and enclosed in one cylinder or case contain-
ing- the directing vanes or guide blades, in such manner that steam entering the first of
the series passes successively through each before being discharged into the atmosphere
or condenser. The turbine wheels consist of metallic discs combined with bushes, which
are slipped upon the shaft or spindle, and firmly keyed or fixed to it in any other
convenient manner. Each disc carries upon a face or faces, or upon an edge or edges,
one or more series of blades, of suitable shape and curvature, in which the general
direction is that of radiation from the centre towards the circumference. The enclosing
case or cylinder, preferably made in halves, carries ring projections and guide blades or
vanes, and the ring projections are arranged so as to form a series of chambers in which
the turbine wheels rotate. Each wheel chamber may for facility of construction consist
FIG. 176. Parsons' Turbine with Fixed and Movable Vanes.
of two rings projecting from the interior of the cylinder or case to enclose a wheel in which
the steam is caused to pass outwards by the first partition or ring, and then over a
space between it and the case to the annular space behind the fixed guide blades or
vanes, which are disposed around the circumference of the wheel and direct the flow of
steam tangentially upon the blades or vanes of the rotating wheel. After passing the
rotating wheel vanes or blades, the steam flows towards the centre and is brought
almost to rest, thus giving up the greater part of its energy of motion to the wheel.
The steam then flows through an annular space around the rotating bush or spindle, and
is directed outwards by the next projecting ring partition to the fixed vanes or guide
plates of the next wheel in the series, there to experience a further fall in pressure and a
further expansion, and so on until the desired total expansion is accomplished. In
order to prevent leakage from any wheel chamber to the next, whereby a portion of
steam would be allowed to pass without flowing through the guide blades, the first
partition ring of each chamber is placed close to the rotating bush carrying the wheel or
close to the spindle itself, and grooves or a series of grooves are cut in the portions of
bush or spindle, and ring, opposing each other, which grooves and projections may
alternately project into each part without touching or nearly touching laterally or
longitudinally. In order to reduce leakage past the faces of the rotating vanes or
blades moving near the enclosing partition, a ring is attached to the lower edge of those
vanes or blades, or otherwise shroud the blades which project under the fixed partition
Parsons' Turbines
ring, and by so bafflingf the steam diminish leakage. At all surfaces where leakage is
likely to occur, which surfaces, of course, cannot be packed or rotated in close contact,
alternate recesses and projections, or recesses only, are cut to give the steam a tortuous
path and reduce leakage. The fixed guide blades form part of one of the partition rings.
The series of turbine wheels may be mounted on a conical, stepped, or cylindrical
spindle, as circumstances may require, and the successive wheels may be of the same
diameter or different diameters, generally increasing, as determined by the desired
amount of expansion, the number of, and discharge area at, the guide blade orifices ;
and the number and configuration of moving blades is also varied to suit various condi-
tions of steam pressure and expansion. The vibration caused by slight want of balance
by means of the following contrivance is damped :
The spindle ends run in a bush fixed into another bush with slight treedom of fit,
and this into another, and so on, giving a number of concentric bushes having slight
play the one within the other. The ends of these bushes fit into a case with some nicety
FIG. 177. Single-Ended Parsons' Turbine.
of end fit, but this is not always necessary, and fill the case with oil, the outer bush
being fixed securely. Any vibrating movement is now checked or damped by the
forcing out of oil between the bushes and the ends and circumferentially around the
bushes, so that although a slight movement is possible, yet it is resisted in whatever
direction it may tend to move. Spring supported bearings may be used, and hydraulic
resistance of other kinds may be interposed such as that due to a small cylinder or ram.
The bush in which the spindle runs is packed with segments of a thin cylinder or tube,
which segments may be of greater or less curvature than the outer surface of the bush,
and so cause bearing of the segment ends upon the exterior of the bush or the interior
of the bored surface holding it. This arrangement gives a spring or elastic movement,
and when oil or other liquid is present the hydraulic resistance already hereinbefore
described is also brought into play.
In order to prevent leakage of the steam past the spindle, a series of grooves or
projecting collars are fitted upon the spindle fitting into or working easily in a similar
series of grooves in an end bush, and by using a sufficient number leakage may be
Modern Engines
reduced to any desired extent. Instead of this, metallic packing- rings may be used.
When the steam turbine is arranged so that there is end thrust in any direction, and to
obviate end wear for an inevitable small end thrust, that end thrust is taken in a thrust
block of construction similar to those used in marine engines ; but in order to secure
that each thrust collar shall take its share of the pressure, and so produce a uniform
slight pressure over a considerable surface, which uniformity is necessary to run cool at
high velocities, the collars or flanges have a slight elastic movement, bringing them to
bear with a predetermined pressure upon their respective collars on the turbine spindle.
Thus they distribute the pressure upon the surfaces accurately, and make certain that
no one surface is subject to undue pressure.
A standard type was at last arrived at, as shown in Fig. 177. This turbine is single
ended, and balanced by dummy pistons instead of by another turbine.
It was long known that a piston well fitted, but yet smaller in bore than a cylinder
in which it can move freely, offers a considerable resistance
to the leakage of fluids ; and if the piston is long and
grooved with many grooves the resistance to leakage is
very great. An air pump was used at one time largely
working on this principle ; it is shown partly in Fig. 178,
which shows the barrel and grooved piston. The piston
does not touch, the barrel being y^^ of an inch clear ; it
requires no oiling, and has no friction. Arc lamp dash
pots are made on the same principle. In later steam
turbines this principle has also been adopted instead of
packing. Mr. Parsons carries it further when the piston
is to rotate instead of reciprocate ; he grooves the cylinder
to fit the piston, thus adding still further to the resistance
to leakage. The three dummy pistons are thus fitted in the
turbine case, and very little leakage occurs, while they
balance the steam pressures.
This system of packing steam tight will be again
referred to under Cylinder and Piston Engines. With
modern machine tools it is possible to fix piston rods and
pistons without the old-fashioned packings, spring rings,
and other nostrums. Deluel's old air pump contained
this valuable principle in its construction, hence it is
worthy of reference again to it. With the construction of
this type of turbine Mr. Parsons arrived at a permanent
design for land engines.
In vessels of the mercantile marine of moderate fast speed it is of more importance
to obtain economy in coal consumption than to reduce the weight of the engines and
condensers to their lowest limits, as is usually done in torpedo-boat destroyers, where
the boilers are extremely light and heavily pressed, and the highest possible speed is the
first consideration.
For the mercantile marine, therefore, it becomes desirable to design the turbines
for the greatest possible economy in steam, consequently the ratio of expansion extends
over nearly the whole range between the boiler pressure and that in the condenser ; the
condensers are also made of ample size so as to maintain a good vacuum, and an
efficient feed-heating arrangement is provided to warm and heat the feed.
The marine steam turbine will be found to be superior or at least equal in economy
of coal to the reciprocating engine when placed in fast vessels of the mercantile marine ;
but it may be asked, What will be the economy of the turbines when, as in the case of
yachts and almost all war vessels, much steaming is done at from one-eighth to one-tenth
power, the full power being only occasionally used ? The answer is a simple one. At
FlG. 178. Deluel's Piston.
PLATE VI. SMALL DE LAVAL TURBINE MOTOR.
Marine Steam Turbines 145
cruising- speeds the revolutions of the turbines fall well within the limits of speed of
small reciprocating engines, and such small engines are then directly coupled to the
main turbines, and work in conjunction with them, these small triple - expansion
reciprocating engines taking the steam directly from the boilers and expanding it down
to about atmospheric pressure ; it then passes to the high pressure turbine, and thence
through the low pressure turbines to the condensers.
It should, however, be added that the turbines alone have their full measure of
economy from half to full power, and even at one-quarter full power the economy
is good. It is only when the cruising speed falls to nearly one-half the full speed,
and the horse-power one-eighth of full power, that the economy of the turbines requires
some assistance, and such additions are only necessary in vessels such as war vessels
and some yachts, where much running is done at these very low speeds. They are quite
unnecessary in passenger vessels and liners.
. The arrangement of turbine machinery for an Atlantic liner of 20,000 to 30,000
indicated horse-power presents no features of novelty over the preceding designs, but
its simplicity of construction, as compared with the present usual reciprocating engine,
is more apparent than in smaller vessels.
The adaptation of steam turbines to marine propulsion required two efforts. The
screw propellers in use are all slow speed, the turbines of high speed ; hence propellers
have had to be designed to bring up their speed to the lowest possible speed of the
turbines, and the turbines had to be designed to drive at the highest speed permissible
by the screw propeller.
The next improvements relate to marine engines. The steam turbine is not an
engine in which the motion can be reversed readily, and in locomotive work and marine
propulsion reversing is a necessity, while in stationary engines it is seldom necessary
except in the case of mining winding engines. Pilbrow proposed to use turbines on
locomotives, but as yet no locomotive has been successfully worked by them. And it is
curious also that the internal combustion engines are not yet made reversing. In marine
steam turbines Mr. Parsons has met the difficulty very well by employing separate
reversing turbines of smaller power than the main turbines. Mr. Parsons' views on
the marine steam turbine were given in his paper already referred to.
The most important field for the steam turbine is undoubtedly in the propulsion
of ships. The large and increasing amount of horse-power, and the greater size and
speed of the modern engines, tend towards some form which shall be light, capable
of perfect balancing, and economical in steam. The marine engine of the piston type
does not entirely fulfil all these requirements.
Though for obvious reasons up till the present time turbines have only been fitted
in vessels designed for phenomenal speeds, yet it must not on this account be assumed
that they are only applicable to such vessels. The two conditions of suitability are that
the vessel shall have a moderately fast speed and be of moderately large size. For slow
vessels of moderate and small size the conditions for turbine machinery are not at the
present time so advantageous.
This will appear clear when we consider that the turbine machinery is actuated by
the momentum of the steam, and that the rows of blades must be sufficiently numerous,
and must move at a sufficiently high velocity to secure a good efficiency from the steam.
The class of vessels that are most suitable for the application of turbine machinery are
the following : Pleasure steamers, passenger and cross-channel steamers, liners (including
Atlantic liners of the largest size), also all fast war vessels, such as torpedo-boats,
destroyers, cruisers of all sizes, protected cruisers, and all battleships of the usual
speeds.
Now, considering that turbines are not reversible in motion, and that they are best
driven at high velocities far beyond the permissible speed of a screw propeller, it seems to
my mind that Mr. Parsons might consider an intermediate transmission gearing between
VOL. i. 10
146 Modern Engines
the turbine and the propeller. The introduction of an auxiliary reciprocating engine
as proposed, along with a reversing pair of turbines and the special screw propellers,
somewhat detracts from the advantages of the simplicity of turbines ; and if all that
extra machinery could be replaced by some transmission of power system flexible enough
for speed variations and reversing, some considerable advantages might accrue. In my
own view, there are two systems which might be tried : first, an electric transmission.
Let the turbines run dynamos of small size at high speed, and let the current operate
motors on the propeller shaft at moderate speeds. With modern electric machinery
this should present no difficulties ; even in large ships the control would be perfect.
The other system is a hydraulic one, in which the turbines would pump water to
a pressure to work by water jet propellers. Properly designed, this would also be
feasible, at any rate, for small craft. In this way the turbine speed could be adapted
to the best speed for economy.
The marine turbines are much the same as the others, only they are of much larger
diameters of wheel, and, as they drive usually on three shafts, one or more of them has
a reversing turbine and a main turbine on one shaft and in one casing. Probably the
application of the steam turbine to marine propulsion has been the most important
step, and we shall therefore fully examine the various inventions of Mr. Parsons for this
purpose. The applications of the turbine to screw propellers and their arrangement
in the ship will be treated in the chapter on " Marine Engineering."
In Patent No. 14,476, 1899, we have several marine steam turbines specified,
principally with a view to combining ahead with astern turbines. As no turbine has
been designed to drive in either direction, it is necessary to combine two turbines to
do the double movement, and this has been accomplished by Mr. Parsons in several
ways. Thus from the specification, the new invention comprises a combination of
main steam turbine with reversing steam turbine in which the main turbine and
reversing turbine are enclosed within a casing formed in one piece, or within a casing
formed of several pieces bolted together, the exhaust ends of both main and reversing
turbines discharging directly into the condenser, or into one passage leading to the
condenser a combination in which the reversing turbine is telescoped within the main
turbine in order to reduce the longitudinal space occupied, and the reversing turbine
case preferably then revolves while the centre part is fixed ; and a combination in
which the reversing turbine is placed at the steam end of the main turbine, and the
exhaust from the reversing turbine passes through the interior of the main turbine
spindle or drum. In all these modifications the reversing turbine is one of ordi-
nary types, and it is run in the vacuum of the condenser while the main turbine is
in operation.
Referring to the illustrations, Fig. 179 is a sectional elevation of a main and
reversing turbine mounted on the same spindle, and arranged and constructed according
to one modification. Turbines end to end are used in this case.
Fig. 180 is a sectional elevation of a main and a reversing turbine mounted on the
same spindle, and arranged and constructed according to another modification.
Referring in the first place to Fig. 179, a is the main turbine enclosed chiefly
in the casing c, and b the reversing turbine enclosed in the inner part / of the
casing d, the casings c and d being bolted together at e. The main turbine a is of
the well-known parallel flow type, and is mounted on the spindle /", which also carries
the reversing turbine b, which has its low pressure end h turned towards the low
pressure end g of the main turbine. Both low pressure ends open into the low pressure
casing d, and discharge by the same passage k to the condenser ; or the low pressure
casing d may be made part of the condenser without special passages. The steam
supply for the main turbine enters by the passage n ; the steam supply to the reversing
turbine is carried through the low pressure casing d by a separate steam pipe m. By
this arrangement a powerful reversing turbine which rotates in the condenser vacuum
Marine Steam Turbines
147
when the main engine is operating the spindle/" in the usual direction, e.g.> for propelling
a vessel ahead, is provided.
Instead of arranging the vacuum ends together, the reversing motor may be turned
round, placing the high pressure end near the vacuum end of the main engine, while
still maintaining the separate turbine casing / within the low pressure casing d and
FIG. 179. Parsons' Marine Turbine.
discharging from the reversing turbine into the low pressure casing as before, and also
running the reversing turbine in the condenser vacuum as before mentioned.
Referring now to Fig. 180, which illustrates the second modification, the reversing
turbine b is here telescoped within the main turbine a in order to economise
longitudinal space. The spindle f carries the drums o and 7, which support the
rotating blades of the main and reversing turbines respectively. The fixed blades of
the main turbine are attached as before to the outside casings c and d> and those of
FIG. 180. Parsons' Telescoped Marine Turbine.
the reversing turbine are fixed on the outside of the inner cylinder p, which is preferably
supported from the exhaust casing d by projecting arms or webs q. Steam is supplied
to the reversing turbine by a pipe or pipes r inside, or forming part of the inner
cylinder p. These steam pipes may serve for fixing the cylinder p to the end of the
casing d. Instead of this arrangement the steam may be conveyed to the reversing
turbine through the spindle f from the passage n, which supplies steam to the main
turbine as in the previous modification.
148
Modern Engines
Another form of turbine is shown in Fig. 181. To prevent leakage from the steam
end of the main turbine to the steam end of the reversing turbine, or vice versa, at points
2 and 3 baffle grooves and rings are formed on the casing c alternating with a series of
grooves and rings on the ring v in order to check the flow of steam, and holes are
provided on the ring v to allow any leak of steam to escape into the condenser.
FIG. 181. Double-Ended Marine Turbine.
In all the modifications the reverse turbine runs idly in the vacuum when not in use.
The telescoping arrangement can, of course, be used to make a compact turbine
for land purposes.
The next improvement, made in conjunction with Mr. Wass, relates to connecting
turbines directly to condensers. They had found by experiment that a very large
FIG. 182. Parsons' Combined Turbine and Condenser. Section.
proportion of the total energy of steam can be obtained with a steam turbine by
expanding to a very low pressure such a pressure, for example, as one-thirtieth of an
atmosphere. By this a turbine is found to give considerably increased efficiency. This
increase of efficiency, however, is easily lost by introducing a slight back pressure, and
it is difficult on board ship, or in confined spaces, to obtain room for an exhaust pipe of
sufficiently large dimensions to prevent injurious throttling of the exhaust.
Marine Steam Turbines
149
The invention consists in entirely dispensing- with connecting pipes between the
low pressure turbine and the condenser, and, in fact, making the low pressure turbine
and the condenser in one part, or as parts of the structure which may be built up of
several parts, the condenser being formed of one or more groups of tubes. The steam
thus leaves the low pressure turbine blades and passes directly to the surface condenser
part of the turbine cylinder without any intermediate connecting pipes whatever.
Referring- now to the illustrations
Fig-. 182 is a sectional elevation of one modification of the invention, in which a low
pressure steam turbine is bolted to a surface condenser.
Fig-. 183 is a plan of the same.
a is the turbine, and b the condenser. The casing c of the turbine is bolted at d
to the end e of the condenser. The shaft /passes through the turbine and condenser,
and is supported in bearings at g and h. k t k are the feet or supports of the turbine
and condenser.
The steam enters the turbine at / and passes through the turbine doing work and
leaving- the last vanes, passes at m into the condenser, which is of a usual construction.
The outlet from the condenser to the air pump is shown at n. The circulating water
FIG. 183. Plan of Turbine and Condenser.
enters the condenser at r y passing into the end space p ; it then passes up through the
tubes q to the top compartment r, whence it descends through the tubes .$ to the
compartment /, and leaves the condenser at u.
The low pressure blades of the turbine thus revolve at but a short distance from the
surface condensing tubes, and no resistance is experienced in passing the enormous
volume of exhaust steam at a low pressure from the turbine blades.
By this arrangement the steam expands down to a very low pressure indeed,
without losing the effect of expansion by back pressure or undue resistance between
the turbine and the condenser.
Mr. Parsons had five years previously patented some claims for this utilisation
of very low pressures in his turbine. We will only select three figures from the
Specification, No. 367, 1894. This invention relates to a combination of a reciprocating
steam engine with a steam turbine, and its object is to increase the power obtainable
by the expansion of steam beyond the limits possible in reciprocating engines. He
discovered that pressures much too low to be utilisable in the low pressure cylinder of
a reciprocating condensing engine may be made to develop very considerable power
in a steam turbine of simple construction specially designed for low pressures. The
invention consists in combining together an expansive reciprocating steam engine
Modern Engines
FIG. 184. Parsons' Combined Engine and Turbine. Plan.
and a low pressure steam turbine in such a manner that when the high pressure steam
has acted in the reciprocating engine, and has been expanded as much as possible in
the low pressure cylinder, it passes to the low pressure steam turbine, and thence, after
further expansion, to the condenser.
The work done by the steam in the cylinders of the reciprocating engine is thus
supplemented by the work done by the steam turbine, and so the power obtained
from a given weight of steam is largely increased. It will thus be seen that the
turbine forms the last element in
the series of apparatus for utilising
the energy of the steam on its
way from the boiler to the con-
denser, thus substantially increas-
ing the efficiency of the engine.
Fig. 184 shows in plan a re-
ciprocating compound condensing
marine engine E combined with
the low pressure turbine B ; the
turbine being utilised for any power
purposes in the ship.
The turbine B may drive a
dynamo D and produce electricity
to light a mill, or be used for
working centrifugal pumps or fans
for forced blast, urging the fires, or any other purpose.
A section of the turbine B is shown at Fig. 185 coupled directly to a dynamo. B is
the low pressure turbine, having a large disc B 1 carrying rings of blades B 3 on both sides
of it ; these rings of blades alternate with rings of fixed blades in the turbine case.
The exhaust steam from the condensing engine enters the turbine by the passage B 2 as
indicated by the arrow ; it passes radially outwards through the rings of blades on both
sides of the disc B 1 , as shown by the arrows i, 2, 3, 4. The disc is perforated at the centre
to allow the steam to
pass through. From
the disc the steam
passes to the con-
denser B 4 , as shown
by the arrow 5. D _
is a dynamo coupled
directly to the turbine
spindle.
Fig. 186 shows
the parallel flow tur-
bine adapted for low
pressures.
In Fig. 184 the re-
ciprocating compound
condensing engine E is shown in plan as driving the propeller shaft H. It is
supplied with high pressure steam by the pipe E 1 , and the exhaust steam passes
by the pipes E 2 to the condenser G, a pipe E 3 from the exhaust pipe E 2 supplies low
pressure steam to the low pressure turbine B, and the exhaust from the turbine
passes by way of the pipe E 4 to the condenser G. In this case the turbine B is
utilised for the electric lighting of the ship, and it actuates the dynamo D coupled
directly to its spindle.
Instead of, or in conjunction with, the dynamo, the turbine may drive fans for
FIG. 185. Section of Exhaust Steam Turbine. Radial.
Exhaust Steam Turbine
ventilating or forced draught purposes, or pumps for circulating water or discharging
water from the ship.
Stop valves are provided to enable exhaust steam to be supplied to the turbine, so
FIG. 1 86. Section of Exhaust Steam Parallel Turbine.
that the whole of the exhaust steam from the main engines may be passed through the
turbine if required, or part passed to the turbine and part to the condenser direct.
These simple forms of turbines might, and no doubt will, be used in many trades
where boiling is a process consuming much fuel, the waste steam from
which would give considerable power if passed through a turbine into
a condenser.
And it is a moot question whether the best way to compound an
engine with high pressure reciprocating pistons would be to connect
it with a turbine.
In any case, it is important to bear in mind that a turbine, work-
ing with steam at or below atmospheric pressure into a good vacuum,
recovers a great deal of the heat energy.
By interposing a simple turbine in the exhaust steam pipe between
the engine and condensers in compound or simple reciprocating
engines it may pay to obtain electric light in that way.
BLADES OR VANES OF PARSONS' TURBINES
These at first were formed by milling teeth on brass discs,
forming flat blades with a passage between of 45 angle with the
shaft.
These, however, were not highly efficient, as the steam in being
deflected was partly thrown against the wrong side of the passages,
thus retarding the movement.
By properly curving the blades and shaping the passages the
steam in passing presses wholly on one side, so that at an early FIG. 187.
date Messrs. Parsons adopted the form of blades shown in the
illustrations.
Fig. 187 is a view of the blades seen end on as arranged on a turbine wheel.
The invention consists in forming the blades upon the back, that is, on the convex
side as compared with the concave side, with a thickening or projection as at a (Fig.
187), the said thickening or projection being nearer to the edge where the steam
enters than where it discharges. The projection a tapers off to both edges, but more
Modern Engines
FIG. i 88.
gradually to the discharge edge than the admission edge. This is clearly shown at
Fig. 187, where c are the admission edges and d the discharge edges.
By this modification in the shape of these curved blades a passage between the
blades is produced in which the minimum of loss of energy occurs presumably by the
reduction of eddy resistance, but possibly in other ways. In any case, whatever be
the correct explanation, an improvement of about at least 10 per cent, was found in
steam consumption by altering one of the steam turbines in this manner.
In one case, for ex-
ample, a moderate amount
of thickening, such as shown
in Fig. 187, at the left-hand
side, gives a gain of 12 per
cent. ; and that a greater
' e thickening, such as shown
at the right-hand side, gives
a better result still.
These blades are formed
into circles or segments of
circles by a very ingenious and simple tool patented by Mr. Parsons and others.
As applied to the parallel flow turbine, suitable strips of ductile metal, preferably
brass, are bent into a circle or sector of a circle. On one edge of the strip
teeth of a special shape are cut by mechanism. The form of the teeth is such
that when the blade sections are laid in the grooves and the teeth turned over upon
them the teeth and blades fit each other closely and form a secure fastening or
mechanical joint. This joint is sometimes made more secure by notching or perforating
the blades before insertion, so as to interlock with the strip. One or more rings or
shroudings may be put upon the blades. The blade sector so formed is inserted and
caulked into grooves in the turbine
drums and cylinders, or is other-
wise secured to them.
Referring to Figs. 188 to 190,
a is a heavy shroud having teeth
cut into one side, leaving, however,
sufficient breadth of metal uncut to
provide a strong strip to serve as
base for the blades ; b is a light
shroud which has also cut teeth,
but the uncut breadth is less than
in a. c, c, c, etc., are blades which
are laid in the spaces between the
teeth of the two shrouds, and each tooth is closed over on the blade held by it. On the
right of Fig. 188 teeth are shown in the position occupied before closing over ; the
teeth of the heavy shroud are lettered e, e, and those of the light shroud f, f\ the blades
c, c are notched at d (Fig. 189), and the light shroud engages with the notches as
shown.
The bases of the blades may be also notched or drilled to increase the grip between
the teeth of the heavy shroud. The teeth e, e,f,f(Fig. 188) are formed with a discon-
tinuous outline on the back, so that when bent over and closed on the top of the blade
the convex outline shall then be a continuous curve, which corresponds exactly with the
concave surface of the blade, and fits that surface. By this means the teeth bear solidly
all over the ends of the blades, and form a strong continuous sector or ring as shown.
We may cut the notches between the teeth to such a depth that the teeth when
bent over will project above the blades, and the surplus metal may be subsequently
FIG. i
Turbine Blades
turned off, leaving the shroud parallel, so as to exactly fill the grooves in the drums
or cylinders.
Both the grooves in the drums and cylinders, and also the shrouds, may be
dovetailed slightly for better security, and when the blade rings have been placed
in position the teeth e are separately
expanded by caulking, so as to fill the
grooves and hold the blades tightly.
Sometimes they prefer to place a strip
of soft metal alongside the shroud
in the groove, and caulk the strip to
secure tightness of the whole structure.
In this same specification plans
for casting the blades into shrouds
are described and illustrated, but the
mechanical method is seemingly much
preferable.
In Fig. 190, given in Mr. Par-
sons' Glasgow paper, is shown the
complete construction of the revolving
wheels in the three sets of wheels in
one casing, the top half of which has
been removed, showing the curved
FIG. 190. Parsons' Wheels.
blades; and Fig. 191, from the same paper, shows one of the sets of engines for
H.M.S. Destroyer, in course of erection ; and Fig. 192 the complete steam turbine now
made in larsre sizes.
FlG. 191. Parsons' Marine Turbines.
DE LAVAL STEAM TURBINES
We now pass on to the other pioneer in steam turbines who struck out in original
lines to secure efficiency. In the early 'eighties De Laval had invented cream separators
working on the principle of centrifugal force. He therefore, in order to get the highest
practicable value for this force, was compelled to drive the separating vessels at
enormous speeds. He was thus brought into contact with the difficulties due to high
speed driving, and became thus practically acquainted with them ; and, as a direct way of
obtaining a high speed from steam power, he used a Hero turbine directly coupled to his
cream separator shaft ; and from these beginnings he started out on the improvement
154 Modern Engines
of steam turbines, culminating in the successful production of a wheel and nozzles
constructed and operated on thoroughly scientific lines and with novelty therein.
With the true modesty and dignity of the genuine inventor, the first important patent
is contained in the following few lines. (Patent 7143, 1889) :
" My invention relates to an improvement in turbines which are set in motion by
means of a current of steam ; and the object of the improvement is to increase,
by complete expansion, the velocity of the steam current, thus producing the relatively
largest quantity of vis viva of the steam.
I attain this object by the construction of the steam supply pipe in such a manner
that the cross sections of the same are slowly increased near to the turbine wheel and
in the direction of the latter. The ratio of increasing the cross sections is due to the
proportion and distance between the smallest section and the largest one, in such a
manner that in the steam passage between these two sections a permanent current of
steam is produced under isoentropical expansion.
Referring to the illustration, Fig. 193, a front view and side elevation, both partly
in section, which shows the mouthpiece of a steam supply M, constructed as above
described, in combination with a turbine wheel A, b is the smallest and c the largest
cross section. Between both these sections the steam expands from the pressure 0.577 PO
(P = boiler pressure) to the pressure of the receiver ( = P 2 ).
FIG. 192. Parsons' Steam Turbine (standard large sizes).
Having now particularly described and ascertained the nature ot my said invention,
and in what manner it is to be performed, I declare that what I claim is :
In steam turbines, the combination of the turbine wheel with a steam supply,
the cross sections of which increase regularly near to the turbine wheel and in the
direction of the same, substantially as and for the purpose specified."
That's all. And note the simple explicit drawings. It is sate to say that if this
invention had originated in New York, the specification would have filled some pages
of this book and the claims filled more space than De Laval's whole specification, while
the possibilities offered in the way of drawings are unlimited. A whole series of pos-
sible and impossible turbines would have been shown, with furnaces, chimneys, boilers,
gas works, superheaters, condensers, and every conceivable detail carefully drawn.
De Laval accepted the high speed necessary with a single wheel, and formed his
nozzle on scientific lines, so that the full velocity may be attained as already described
and illustrated under " Steam Jets."
Parsons had already shown that elasticity in the bearings of the turbine wheel
shaft was necessary to damp vibrations, and allow the wheels to approach a revolution
round the centre of gravity ; but De Laval aimed at higher peripheral speeds, and so,
instead of damping vibration, he arranged his wheel on a flexible shaft so that the shaft
could bend and allow the wheel to find its own centre of revolution. This flexible
Single Wheel Steam Turbines 155
shaft was at first made with spring actions, but, owing to the fact that the torque of
a wheel at these high revolutions is small, the shaft can be made thin enough of itself to
bend sufficient for the purposes. The first specification is as follows :
The invention consists of an improved arrangement for reducing to a minimum the
reactions on the bearings by rapidly rotating bodies, arising from their imperfect balancing.
This is accomplished by fixing the rotating body either directly or indirectly on a tube or
sleeve which is fixed on and around a shaft, which tube is made elastic or yielding in some
suitable manner, so that the rotating body is rendered self-adjusting or self-balancing.
Two constructions of this arrangement are shown
Fig. 194 being a part sectional elevation of one construction, and
Fig. 195 a like view of another construction, showing a transverse sectional view
on the line cd, and a transverse sectional view on the line ab.
Similar letters of reference indicate corresponding parts.
In Fig. 194 the body A is fixed on the boss B, which again is made fast upon the
Front View. Side Elevation.
FIG. 193. De Laval Wheel and Nozzle.
central part of the tube or sleeve C, the latter being fastened on both ends D and E to
the shaft F. This shaft is shown supported at each end in bearings G and H in the
usual way, and the motion is imparted to the system by means of the driving pulley K.
A portion of the shaft F between D and E has been turned down, thus leaving an
annular space between the tube C and the shaft F. A helical slit b is cut in the tube C
between its central portion and each end, so that the central portion occupied by the
body A may, while rotating, assume such position relative to shaft F as may be required
to secure an even rotation, which saves wear and tear in the bearings.
In Fig. 195, where the body A is directly fixed upon the thicker central portion of
the tube or sleeve C, this tube is made elastic or yielding as required, either by turning
down the tube until its thickness is considerably reduced (see transverse section through
ab}, or the tube, being somewhat thicker (see transverse section through cd), is provided
with longitudinal slits a by means of which the required elasticity is obtained.
156
Modern Engines
The bearings for the ends G and H ot the shaft have in the present cases been
shown fixed. In some instances, however, it may be desirable that these also should
be able to yield to a slight extent. In such cases Seller's bearings should be used,
pivoted bearings, or bearings surrounded with indiarubber rings like the top bearing
of the cream separators. Other constructions of self-adjusting bearings may also be
used to give the requisite adjustment.
In a subsequent patent De Laval shows how to apply this flexible shaft with special
jointed bearings to other rapidly moving heavy rotating bodies, so that a balance is
obtained. This and the previous patent display true engineering science, in which a
desirable result is achieved by natural means. The old-time engineers, when they first
introduced moderately high speed engines, tried to balance the cranks by counterweights
and other means, bringing the inertia of the revolving parts into opposition ; but
they relied upon some tons of cast iron, stone, and concrete, and ten or twelve huge
foundation bolts, after all, to prevent vibration. They did not cure vibration ; they
only made it less in amplitude by mere force, and the internal strains were greatly
increased. The early engines were kept from violent vibration by brute force, not by
any scientific means. Nowadays, engineers do not depend on foundations for smooth
working : a good engine well designed and balanced should run full speed without vibra-
FIG. 194. De Laval Flexible Shaft.
FlG. 195. De Laval Flexible Shaft.
tion when standing upon a foundation without any holding down bolts. A high speed
engine which requires to be held down to a huge foundation by heavy bolts is an engine
radically wrong in design.
A theory of the flexible shaft has been given by Mr. Sandford A. Moss in Power,
which may be summed up as follows :
The popular explanation of the phenomenon is that the flexible shaft allows the
disc to "rotate about its centre of gravity." This is not true, however, since in all
cases a disc slightly out of balance rotates so that the shaft is slightly deflected and the
centre of gravity is away from the centre of rotation. The outward centrifugal force due
to this eccentricity of the centre of gravity is just in equilibrium with the inward pull
due to the deflection of the shaft. Of course this deflection causes a rotating reaction
at the bearings, which tends to vibrate them, and which usually gives trouble in high
speed machinery. This is because the greater the speed the greater the centrifugal
force, and ordinarily the greater the shaft deflection required to balance it. The
increase of deflection throws the centre of gravity out farther, which still further increases
the centrifugal force. Hence ordinarily the reaction on the bearings due to the shaft
deflection, which is of course equal to the centrifugal force, increases rapidly with
increase of speed. However, it is a remarkable fact that, after a certain critical speed
has been reached, the shaft deflection and the reaction on the bearings actually decrease
as the speed increases, in order that equilibrium may be maintained with the centrifugal
De Laval Turbines 157
force due to the corresponding 1 eccentricity of the centre of gravity. The eccentricity is
less than the deflection, instead of being greater as ordinarily. The more flexible the
shaft the less this "critical period."
By using a very flexible shaft and running it at a much greater speed than the
critical speed, the deflection required to maintain equilibrium with the centrifugal force
due to the corresponding eccentricity of the centre of gravity becomes very small indeed,
and the corresponding rotating reaction pull on the bearing is almost imperceptible.
The critical speed in radians per second is found to be 'A = \/lvf > wnerem ^
is the natural period of vibration of the shaft if deflected and let go or struck in the
middle, M is the weight of the disc in Ibs., and K the force required to deflect it unit
distance.
oo /^2i6xr2v& I k
In revolutions per minute this becomes */ ^ - , or 187.56*7 v^j where k is
. 2 7T IVl T IVJ.
load in Ibs. required to deflect the shaft i inch and M is the weight of the disc in Ibs.
Now M/ is the actual deflection of the shaft in inches, produced by the weights of
the disc ; hence the critical speed is -r- revolutions per minute, where d is the
v
deflection of the shaft in inches produced by the weight of the disc.
For a steel shaft of the same diameter throughout, with two bearings either swivelled
or so loose that the shaft may be considered as merely supported by them, and for a
single mass fixed at any point between the bearings, the formula reduces to the
following: Let / be the distance between bearings in inches, M the weight in Ibs. of
the rotating mass, a its distance from either of the bearings in inches, and d the shaft
diameter in inches ; then the critical speed at which the maximum vibration occurs, in
d 2 I~T
revolutions per minute, is 387,700- r*/ . This also gives the period of a natural
G\L d ] * JV1
vibration. If there are two rotary masses, there are two distinct vibration periods and
critical speeds. Now as to the efficiency of such wheels, as mentioned previously,
it is important that as much as possible of the kinetic energy of the steam jet
issuing" from the nozzle should be taken up by the turbine wheel, and thus trans-
formed into mechanical energy. The angle between the nozzle and the plane of
rotation of the wheel is 20 degrees, and in order to obtain the maximum efficiency
the peripheral speed of the turbine wheel, i.e. the linear velocity of the buckets,
should be 34 per cent, of the velocity of the steam. The absolute velocity of
the steam leaving the buckets is then 34 per cent, of the initial velocity, and the
energy absorbed by the turbine wheel is 88 per cent, of the kinetic energy of the
steam.
If, for instance, the speed of the steam entering the buckets of the turbine wheel is
4000 feet per second, the speed of the steam leaving 1 the buckets should be 1360 feet
per second, and the number of horse-power per Ib. of steam 2 = u ; and the
*g x 55 x 3 600
steam consumption per theoretical horse-power ^5 3 = 9.1 Ibs.
40oo 2 - 1360*
The steam nozzles are placed in very close proximity to the buckets of the turbine
wheel, in fact the distance is only 2 millimetres, or about T Y of an inch, and consequently
there is practically no loss of velocity between the steam jet leaving the nozzle and
entering the buckets of the turbine wheel, nor leakage of steam.
The speed of the turbine wheel, which for a velocity of the steam jet of 4000 feet
per second ought to be about 1880 feet per second, or about 21 miles per minute, is,
however, much lower, for several practical reasons. At the present time the peripheral
speed of the De Laval turbine wheel does not exceed 1380 feet per second, which should
Modern Engines
make a steam consumption of 9.8 Ibs. per theoretical horse-power. The following table
gives the speed of some types of turbine wheels :
TABLE XVI. SPEEDS OF THE TURBINE WHEELS.
Size of Engine.
Middle Diameter of Wheel.
Revolutions per
Minute.
Peripheral Speed.
Feet per Second.
Horse-power.
Millimetres. Inches.
5
100 4
30,000
515
15
150 f,
24,000
617
3
225 8|
20,000
774
So
300 ii J
16,400
846
100
500 19!
13,000
i"5
300
760 30
10,600
1378
Fig. 196 shows the stresses in a wheel for a 50 horse-power steam turbine.
As may be seen by this diagram, the wheel is so constructed that both P the radial
stress and S the tangential stress have their largest value at the circumference of the
wheel, just where the buckets are fixed. Consequently the wheel is not made of uniform
strength, but is strongest at the heavy part, that is, in the centre.
It will be seen from Fig. 196 that the tangential stresses S in the boss of the wheel
increase as they approach the hole in the centre of the wheel. These stresses would be
still greater on the larger sizes of wheels, and in order to avoid these greater stresses the
larger wheels are made without any hole through the centre, but the shaft is made in
two pieces fixed to the wheel by flanges and screws. Fig. 197 shows the arrangement
of small and medium
turbine wheels ; and
the arrangement of the
larger turbine wheels
is shown in Fig. 198.
The part of the
turbine which makes
it possible to run the
^
FIG. 196. Curves of Stresses.
turbine wheel at its
enormous speed is the
"flexible shaft." The
shaft on which the
turbine wheel is
mounted has bearings
on each side of the
wheel at a good distance from it, and the shaft is consequently flexible and can allow the
wheel to swing a little in its plane of rotation. No matter with what nicety the turbine
wheel may be turned and balanced, it is practically impossible to bring the centre of
gravity of the wheel exactly into the geometrical centre round which the wheel revolves.
The flexible shaft and the turbine wheel are so proportioned that the settling of the
wheel takes place very quickly, and the critical speed is from \ to of the standard
number of revolutions of the wheel. Of course the turbine wheels are very finely
balanced, and the settling of the wheel is therefore scarcely perceptible. It is the flexible
shaft that serves to transmit the power of the turbine.
The diameter of the shaft is, on account of the high speed, very small, and it is
therefore easy to make it flexible. The shaft of the 300 horse-power turbine wheel has
a diameter of 34 millimetres, or i T 5 F inch, and that of the 150 horse-power wheel 25
millimetres, or i inch : no larger diameter is required.
De Laval Turbines
I S9
The normal speed of the turbine wheel is too high for direct driving of ordinary
machinery, and it is therefore reduced by means of gearing. This gearing is made on
the double helical system, and machined with the greatest care and accuracy, as is
necessary on account of the high speed. The speed of the gearing, that is the linear
velocity of the teeth, is about 1000 feet per second. The pinion is made of hard steel
(in one piece with the shaft), and the teeth of the gear-
ing wheels are cut in a somewhat softer steel than the
pinion.
All the revolving parts of the turbine are most care-
fully balanced, and the parts mounted on the shafts are
centred by tapers.
Thebearingsofthe
slow speed shafts
are lubricated by
rings, as is usual
in this class of
machinery. The
journals of the
flexible shaft are
oiled by sight feed
lubricators. The
bearings, which
are all made as
FlG. 197. Arrangement of Small and
Medium Turbine Wheels.
FIG. 198. Arrangement of Large
Turbine Wheels.
interchangeable
bushes, are lined
with white metal,
and there is practically no wear on them if the machine is well lined up and properly
mounted from the beginning.
The turbines are generally fitted with more than one steam nozzle, and these are
arranged at intervals in a ring in close proximity to the turbine wheel, receiving the
steam from a steam chest in the turbine case. Each nozzle is usually provided with a
USD
FIG. 199. Valve and Nozzle of De Laval Turbine.
shutting-off valve, so that any nozzle can be closed or opened at any time. This
arrangement is of considerable advantage, as when the turbine is working at reduced
load some of the nozzles may be closed and a high efficiency of the machine maintained,
even although it is not working at full load. This will be more plainly understood from
the tests of steam consumption, which will be noticed subsequently. The arrangement of
a nozzle and its valve will be seen from Figs. 199 and 200.
i6o
Modern Enines
As may be seen from the foregoing table, the peripheral speed increases with the size
of the wheel, and the larger the diameter the higher also is the peripheral speed. The
300 horse-power turbine wheel runs with a peripheral speed of 1378 feet per second in the
middle of the buckets ; the outside diameter of this wheel is 800 millimetres, or 31^ inches,
and the circumferential velocity of the wheel is 1450 feet per second, or more than 16
miles per minute. At this speed the wheel would travel round the equator of the earth
in 25 hours.
On account of the peripheral speed of the turbine wheel not being so high as it
theoretically ought to be, there is, particularly at high admission pressure and good
vacuum, a slight impact when the steam enters the buckets. This is, however, allowed,
for practical reasons, and the energy due to the loss of speed by this impact is not
entirely lost by the turbine, as will be seen later.
One advantage of the action principle of the turbine is that the turbine wheel can
revolve quite freely in
the casing. This is
an essential feature
of the machine, and,
moreover, it would
not be possible to run
at the speed required
should a tightening be
necessary round the
turbine wheel. The
wheel does not touch
anywhere, and all the
steam on emerging
from the nozzles must
. pass the buckets of the
wheel, as the radial
length of the buckets is
always larger than the
diameter of the steam
jet. There is conse-
quently no possibility
of any steam leaking
through the turbine,
but it must of neces-
sity pass the buckets
and deliver its energy
to the turbine wheel.
The high peripheral speed which, as previously seen, is necessary in order to obtain
a good efficiency has been obtained by allowing the turbine wheel to run at a very high
velocity. A reference to Table XVI. will also show that the number of revolutions is
much higher than the speeds formerly used in practical engineering.
The wheel is made as a solid disc, on the circumference of which the buckets
are dovetailed in, each bucket being made and fixed separately to the wheel. The
buckets consequently load the circumference of the wheel with a radial force when
the wheel is revolving. The amount of this force may be understood when it is
mentioned that the centrifugal force on the bucket of a 300 horse-power turbine wheel,
which bucket weighs 250 grains, is 15 cwts. when the wheel is running at its standard
speed.
The stresses in the wheel are tangential and radial ; and if we call the radial stress P
and the tangential stresses S, it is evident that both P and S vary with the radius R.
FIG. 200. De Laval Wheel and Nozzles.
De Laval Turbines
161
Further, these stresses depend on the axial thickness of the wheel in each place, and
they also affect one another.
Before the admission steam can enter the steam chest and pass from thence to the
nozzles it is regulated by the governor valve, which in its turn is controlled by the
centrifugal governor of the machine. The governor valve is a balanced double-seated
valve, connected with a link motion to the centrifugal governor.
Longitudinal Section.
Sectional Plan.
FlG. 201. De Laval Turbine.
The general arrangement of the machine will be understood from Fig. 201, which
shows a turbine in section and plan, and Fig. 202, a dissected turbine.
The speed of the turbine is regulated by a very sensitive centrifugal governor,
mounted horizontally on the end of the gear-wheel shaft. The moving parts of the
governor work practically without friction, and it is therefore very quick and powerful.
This governor is 'very simple, although the construction may seem peculiar, and its
VOL. i. ii
162
Modern Engines
dimensions are very small on account of the comparatively high speed at which it works.
Fig. 203 gives an idea of the construction of the governor.
The variation of speed between full load and no load is nearly i per cent. ; the
variation is from 2 to 3 per cent, generally.
The standard sizes of steam turbines can work with any steam pressure between 50
and 200 Ibs. per square inch, and either with or without vacuum. The only parts of the
machine which have to be arranged to suit the admission pressure and the pressure in
A Turbine shaft
B Turbine wheel
C Pinion
D End bush
E Safety bearing in turbine box
F Middle bush in two parts
G Safety bearing in box covers
H Ball bush with adjusting spring
I Steam nozzle
K Stuffing box with stop valve to nozzle
L Wheel to valve spindle
M Gear wheel
N Gear wheel shaft
O Gear wheel shaft bushes in two parts
P Lubricating ring
Q Centrifugal governor
R Driving pulley
S Stop nut with pulley for connect-
ing to tachometer
T Tightening bush in two parts
U Adjusting nut with spring
V Friction gland
FIG. 202. Loose Parts of Steam Turbine.
the exhaust are the steam nozzles, which have to be shaped according to the amount of
expansion of the steam. The nozzles are made interchangeable all other parts do not
alter with the pressure, and the machine can consequently work with any pressure
between the above limits, if only the turbine case is provided with suitable nozzles. The
turbine can also be arranged with nozzles for running both condensing and non-
condensing : this is very handy and convenient, particularly if the turbine drives its own
condensing machinery, direct or electrically.
As the question of economy became of more importance, the size of the wheels and
also the number of revo-
lutions in the larger unit
of machine are so pro-
portioned that, with in-
creasing unit, the velocity
of the vanes of the wheel
approached more closely
FIG. 203. Governor of De Laval Turbine.
to what it theoretically
ought to be. The
and the sections of the
stresses are largest in the circumference of the wheel,
wheel were so proportioned that the stresses increased with the radius. This was
done in order that the wheel might be weakest in the circumference where the vanes
were fixed, and in order to be still more certain on this point a recess was turned in
this outer portion of the wheel. Should a wheel burst on account of too high a speed it
gave out at the recess, the vanes became detached, and the steam could not longer drive
De Laval Turbines
163
the wheel. The buckets of the detached parts of the wheel were so light that they
could do no damage, and the machine only stopped running. With this type of wheel the
heavy central part had never burst. Indeed it would be a very serious matter if the heavy
part ever became detached from the shaft. When the calculations for a new wheel were
completed, the material of which the wheel had to be made was tested in order to see
whether its strength corresponded to that on which the calculations were based. If
agreement existed, the wheel was then made, and run until it burst at the periphery.
Experience had proved that the speed at which the breaking of the wheel took place
could be calculated beforehand, and, if the speed obtained by experiment accorded with
the theoretical result, the wheel was adopted as a standard for that particular type for
practical use. The wheels were generally proportioned so that breaking would take
place when the wheel was running at about double the number of revolutions required in
actual work ; consequently it could hardly happen in practice that the machine increased
so much in speed that there would be any danger from the wheel breaking at its
circumference. In order to keep the stresses down in the central parts, the wheels were
made very thick towards the axis. If a wheel of the same thickness throughout were
used, the stresses would be very great at the centre, and this was a matter to avoid in
high speed machinery. It was best to do away as much as possible with the drilling
tifj
FIG. 204. Skin Friction Curve.
of holes for bolts, etc. in the boss of a wheel, as holes considerably weakened the central
parts, and it was very often the boss which had to keep the wheel together. As to the
resistance to which the revolving wheel was subjected from the surrounding medium,
this depends partly on the skin friction, and partly also to eddy making. It was found
in practice that the resistance was almost exactly proportional to the density of the
surrounding medium, and that it increased approximately with the fifth power of the
diameter and the third power of the number of revolutions. It was therefore evident
that the thinner the medium which surrounded the wheel, the less would be the resistance
offered to its motion, and this would be plainly understood from the curves in Fig. 204.
The resistance was less in saturated steam than in air of the same pressure, and it
decreased with increasing vacuum. A 150 horse-power turbine wheel was subjected to a
resistance of 35 horse-power when running in steam of i atmosphere absolute pressure,
but if it were run in a vacuum of 28 inches of mercury 2 inches of mercury absolute
pressure the resistance would be decreased in about the same proportion, and would
be 5^ *35 = 2j horse-power, a gain of 32^ horse-power. The velocity of the steam jet
into the vacuum was, moreover, higher than the velocity of outflow into the atmosphere,
and both these circumstances made it essential that the turbine machinery should be run
under vacuum. From Fig. 204 it was also evident that the resistance was less in super-
heated than in saturated steam, and that it decreased with the amount of superheating.
1 64 Modern Engines
The frontispiece of this volume illustrates two fans driven by two of these turbines
installed at Romford Gas Works.
Plate VI. illustrates the smaller sizes set up with reducing belt gearing all self-
contained for driving small powers : this illustrates a 5 horse-power set.
And Plate VII. illustrates the large Parsons' turbine with alternator attached.
The frontispiece and Plate VI. are from Messrs. Greenwood & Batley, who are the
British representatives and manufacturers of De Laval turbines.
Before passing on to notice the few turbines introduced since the great success of
Parsons and De Laval, it may be of interest to notice another effect of high velocities in
rotating bodies, namely the gyrostatic effects : the inertia of a body opposing any force
which tends to set it in motion, or to alter its motion while moving, either in direction or
velocity.
The gyroscope illustrates these properties beautifully. The gyroscope usually
consists of a fly-wheel mounted on centres inside of a ring, so that it can be spun round
at a high velocity. It is sold as an interesting toy, but some of the most intricate
mechanical problems can be found in its theory. It is also of use for showing the
earth's rotation.
A bicycle wheel may be used to demonstrate a few effects of the gyroscope. The
rear wheel is usually easily removed with its spindle. If this is done and two people
hold it up, one grasping each end of the spindle, and the wheel be rapidly spun round,
it will be found to strongly resist any attempt to twist the shaft round, out of the line it
was held in. One of the persons may let go his end of the spindle, the wheel will not
fall, and the other person will not have to exert any more support than that required to
keep the weight of one end up. In fact, the wheel, while rotating in a vertical plane,
can be suspended by a string tied to one end of the spindle. The spindle will not drop
out of the horizontal position, even although supported at one end only.
If a gyroscope be taken between the two hands, or the spinning bicycle wheel
between two persons, and the wheel turned right over, so that the spindle is turned end
for end, a considerable resistance will be offered by the wheel to this operation. The
mathematical explanation for this resistance is complicated, but the simplest way to look
at it practically is to consider the direction of the motion of the wheel, first before it is
turned over, and after. Suppose a person looking on one side of the wheel when it first
spins round, that he sees the direction of rotation is from left to right, that is clockwise,
he will find, on turning the spindle end for end against the resistance of the wheel, that
the motion or direction of rotation has from his point of view reversed. The resistance
offered by the wheel is the resistance which a body in motion offers to reversal of the
direction of the motion, just as a jet of steam or water resists reversal in the bucket of a
turbine.
In fast running machinery like steam turbines, any alteration of the plane of revolu-
tion is accompanied by resistance which bears on the shaft and bearings. In ships the
shaft is horizontal, so that any quick turning of the ship will throw pressure on the
turbine bearings resisting the turning round of the ship. Rolling of the ship would not
have any effect, for the plane of rotation would not be altered by a ship's rolling with
the turbine shaft fore and aft. Pitching would have but a slight effect, for the angle of
movement is small. A rapid deviation of the ship at full speed from her course would
have most effect ; and in ships of high speed and light weight, such as torpedo-boats and
destroyers, with powerful engines and powerful steering gear, it is quite possible that a
sudden swerve would throw a dangerous amount of strain on the turbine bearings, and
might bring about a rupture of the structure.
A rolling ship with the turbine shaft athwart the ship would also be subjected to
straining, for the rolling would considerably and quickly alter the plane of rotation.
Experiments with gyroscopes are now mostly made with electrically propelled wheels,,
the motion being steady, and prolonged as long as we please.
Rateau Turbine
There are three or four turbines which have been introduced into practice, while
there are some three or four thousand which have not been reduced to practice. Since
Parsons and De Laval established the practicability of steam turbines fitted with proper
bearings, guide blades, and wheel shafts and other things, all designed with strict regard
to the elastic fluid and high velocities necessary, there has been the usual avalanche of
patents poured into the patent offices. About fifty patents were granted in 1901, and
now in 1903 only four or five of them have ever since been heard of.
I have tried to go through all the patents since 1890 up to the latest published.
Ninety-nine out of every hundred are utterly sterile of any original ideas. Many are not
so far advanced as Pilbrow and Wilson were fifty years ago. The fact is, when we add
Parsons' improvements and De Laval's improvements to those of Pilbrow and Wilson,
we have the whole state of the art and science of steam turbine construction at the
present moment clearly founded on the published specifications of these four inventors.
The master patents having expired by lapse of time, the vast number of new patents
can only cover some
detail, such as bearings,
valves, packing, special
nozzles or blades, so
that all recent turbines
are either Parsons'or De
Laval's, with some sub-
ordinate feature altered
or changed for another.
Some of them, if
equally skilfully de-
signed and equally well
made, would no doubt
give as good results,
but that "if" is of a
large order.
The Rateau turbine
is essentially a number
of De Laval turbines in
series with one common
shaft, and is one of the
first rank, designed by
Professor Rateau. Its
scientific points are above criticism ; and, made with that mechanical skill which French
mechanics display in all their work, it has rapidly assumed an important position. In
Engineering a brief description of it as given by Professor Rateau himself is as follows:
The multicellular turbine in question is constructed by Messrs. Sautter-Harl6 &
Co., engineers, of Paris, and consists, as shown in Fig. 205, of a series of rotating wheels
keyed on the turbine shaft, and separated by circumferential diaphragms held in grooves
inside the turbine casing. The latter is in two parts, the section being on a plane which
passes through the axis of the turbine. This allows the accurate fitting of the lower halt
of the system, which is then completed by putting in place the top half of the casing.
The wheels consist of buckled steel plates, conical in shape (Fig. 205), on the periphery
of which are riveted nickel-steel blades, which are further held in place by a band. The
fixed blades of the distributors (guide blades) are fitted on the periphery of the diaphragms
(see 2, Fig. 205), opposite the wheels. On Fig. 205 the last diaphragm is hatched in
order to throw it up more clearly.
The shaft runs through the diaphragms in bushes of anti-friction metal, with but
little play originally two-tenths of i millimetre. The shaft, however, makes its play
FIG. 205. Rateau Turbine.
i66
Modern Engines
when this is insufficient. There is no inconvenience attached to the friction of the shaft
in the anti-friction bushes. Leakages of steam outside the distributors can only take
place through this small windage, the total section of which is comparatively a very
small one, owing to the small diameter of the shaft. The space between the wheels and
the fixed parts is from 3 to 6 millimetres .118 inch to .236 inch; there is thus no
fear of their coming into contact. Should this happen, however, and a number of blades
get broken, they could easily be repaired, even on board ship, as each blade can be
removed individually and be replaced by another, fixed by one rivet only.
The steam flows through the diaphragms and wheels from the inlet at AB to the
exhaust G, its pressure decreasing progressively, each successive wheel working under a
low fall in pressure. The number of wheels varies from three to four only to thirty and
over.
The foregoing shows that all loss in efficiency through leakage between the fixed
and movable parts has been prevented in a marked degree. It has been stated that with
this system of turbine the friction of the wheels in steam leads to a notable loss. This is
188 ZJ #02
ElecinvaJ/ HJP: measured/ out/
Th&Curwes correspond/ Co the? use* of steam?
FIG. 206. Results of Tests on Rateau Turbine.
not the case ; and in actual practice, with a turbine the dimensions of which have been
carefully calculated, the loss by steam friction is only 3 to 4 per cent, of the power of the
machine.
The steam consumption is also very moderate, being, it is believed, lower than
that of any other similar type of engine. These turbines are calculated with great
precision, by formulae based on theory and completed by experiments on the flow of
steam ; the difference between the calculated figures and the results obtained at the
trials does not generally exceed 2 per cent. , thus confirming the formulae on which the
etudes were first made.
The efficiency of the turbines is shown in the diagram, Fig. 206, which gives the
results obtained at trial runs made with the first of three turbines, coupled to a continuous
current dynamo, supplied by Messrs. Sautter-Hade" & Co. for the generating station of
a mine at Penarroya, in Spain. The only motive engines in this station are the three
steam turbines in question.
From the diagram, Fig. 206, it will be seen that the consumption of the turbine
and dynamo under full load or 500 horse-power the inlet pressure being 8.5 kilogrammes
Rateau Turbine 167
per square centimetre or 120 Ibs. per square inch, and the back pressure at the condenser
.115 kilogramme per square centimetre (1.63 Ib. absolute) was 7.03 kilogrammes or
15.5 Ibs. of steam per electric horse-power hour ; and under an overload, at 640 horse-
power, with 1 1 kilogrammes or 156.5 Ibs. per square inch and .128 kilogramme or 1.82 Ib.
absolute counter pressure at the condenser, the steam consumption was only 6.74 kilo-
grammes or 15 Ibs. of steam per electric horse-power hour. These are good results for a
500 horse-power engine, seeing especially that the vacuum at the condenser was not
very high. With a high vacuum, such as the Hon. Charles A. Parsons can so readily
obtain, say 0.06 kilogramme or 0.86 Ib. absolute back pressure at the condenser, the
above figures of steam consumption would be reduced by 10 per cent, at least, and
would be 14 Ibs. per electric horse-power hour normally, and 13.4 Ibs. with an overload.
The diagram, Fig. 206, shows further that the total efficiency of the set, which
rises to 58 per cent, at full load, remains approximately constant down to below half
load ; further, the total consumption of steam the dynamo remaining excited and being
self-exciting does not exceed 336 kilogrammes (740 Ibs.) per hour running empty, being
thus 10 per cent, lower than at full load. M. Rateau believes that no other engine gives
such favourable results.
It is an impulse turbine, but the distinction between impulse turbines when put in
series order and so-called reaction turbines, as Mr. Parsons has very properly pointed
out, is the difference between tweedledee and tweedledum, when the fluid is steam.
M. Rateau denies this, and refers to hydraulics to show the difference. But water and
steam are two very different fluids. While they act similarly in some respects, they
differ very materially in others, due to the fact that one is elastic and the other is not.
OTHER STEAM TURBINES
There is the Stumpf turbine, a modification of the De Laval, with a wheel more like
a Pelton wheel, and nozzles all around.
The Seger turbine is like Pilbrow's 2-wheel machine shown in Fig. 161, and my
own wheels shown in Fig. 163. A feature of this machine is the reduction of the speed
by belt gearing.
There is also the Schulz and Curtis steam turbines ; but a perusal of the specifications
reveals nothing of much novelty in either, but we shall have occasion to examine them.
The Westinghouse Company's steam turbine is a modification of Parsons'. At an
early date this company purchased the American rights and licence for the British.
The Thomson-Houston Company in this country supply the Curtis turbine as modified
by the engineers of the General Electric Company of America at Schenectady.
German turbines are known as those of Professor Stumpf, and Schulz. French
turbines as designed by Rateau are made by Sautter-Harte & Co.; while De Laval
turbines are made in many countries.
These turbines we notice here more for the purpose of showing the small details
which distinguish them from the earlier and more original turbines.
The construction of high speed wheels with curved blades has always been some-
what a difficulty. We have seen how Parsons and De Laval make their wheels.
E. Seger, of Stockholm, in his patent of 1894, No. 4611, shows a simple method for
forming wheels for high velocities ; he employs two wheels side by side, like Pilbrow's,
Fig. 161, running in opposite directions.
The following are the instructions for manufacturing the wheels, in which the vanes
or buckets are called paddles :
Referring to the illustrations
Fig. 207 is a vertical section of two turbine wheels arranged one above the other;
Fig. 207A is a front elevation of the same with the wheel ring partly removed.
Fig. 208 illustrates the form of the paddle blanks.
i68
Modern Engines
Fig. 2o8A is a sectional view of part of a turbine wheel with an inserted paddle, but
without the outer wheel ring.
Fig. 209 is a side elevation of a part of such a turbine wheel with paddles, illustrat-
ing two different steps of the operation of fastening the latter.
Fig. 2OQA shows a part of such a turbine wheel with the paddles fixed by riveting, and
Fig. 2096 shows the outside of a part of the wheel body with notches for the paddles,
some of the paddles being inserted.
The paddle blanks a, which may be punched out of thin sheet-iron plates or of
p/&P
7
L
7
1
ii
453
24.8
11,250
J 35
276
29.65
8,i75
BLACKPOOL CORPORATION.
146
70
27.1
2500
5i5
21-35
11,000
150
o
27.0
502
23.1
11,600
i35
o
27-3
497
24.0
",953
i33
66
27-3
57
21. 1
10,693
152
o
29.0
1,500
160
23.6
o
2,530
156
5
28.9
o
1,465
It will be noted from the above results that the improvement in steam consumption
resulting from a superheat of 50 F. is about 8 per cent., and from 100 F. it averages
about 12 per cent. ; also that for every inch of vacuum above 25 inches or 26 inches the
consumption of steam falls about 4 per cent.
It will be noted that in steam turbines the steam consumption closely follows a right
line law, or is proportional to the load plus a constant quantity which represents the
consumption of steam at no load.
In connection with superheated steam it may be mentioned that, as there is no
internal lubrication of the turbines, none of the usual difficulties which occur with
reciprocating engines are met with in its employment.
Also, in steam turbines the absence of internal lubrication renders the exhaust steam
absolutely free from oil, so that the water from the hot well can be returned to the boilers
direct without oil filters.
Special tests have been made from time to time on turbine engines to verify the
statement that no increase in steam consumption occurs with the age of the plant under
fair wear and tear.
As to the consumption of steam on still larger plants, a long and exhaustive
series of tests was made in January 1900 by Mr. W. H. Lindley and Professors
Schroter and Weber, on behalf of the city of Elberfeld in Germany, on one of two
1000- kilowatt turbo alternators built at Heaton Works for that city. The turbo
alternators were constructed to give 1500 kilowatts at 4000 volts 50 periodicity,
the alternators being four pole running at 1500 revolutions per minute, and directly
coupled to the turbines. The expansion of the steam was carried out in two cylinders,
a high pressure and a low pressure, the steam being expanded down to a little below
184
Modern Engines
the atmosphere in the first, and from that to the vacuum of the condenser in the
second.
We have not space here to refer to the modus operandi of these tests, nor all the
results. A summary of them will suffice.
TABLE XX.
O
-o^ E
1
[j
fe l ^
&>-
S 3; _o
<
UH^
^ H c
Oo
(O
(2)
(3)
(4)
(5)
(6)
Kilowatts.
Kgs. per
cm?.
C.
C."
C.'
Kgs.
1190. i
IO. II
179-3
189.5
IO*2
8.81
994.8
10.47
180.9
192.0
II. I
9.14
745-3
10.76
182.0
190.0
8.0
IO. 12
498.7
10.40
180.6
209.7
29.1
11.42
246.5
10.14
179.4
196.4
17.0
15.31
No load withl
excitation. J
10.34
180.3
193.0
13-3
per hour.
1844
No load with-\
out excitation. /
10.49
181.0
194-5
13-5
1183
For the following outputs in round numbers the steam consumption per hour is as
shown :
TABLE XXI.
Output.
Steam Consumption
per Hour.
Steam Consumption
per Kilowatt-hour.
Kilowatts.
Kilogrammes.
Kilogrammes.
1250
10,786
8.63
1OOO
9,189
9.19
750
7.496
9.99
500
5.707
11.41
250
3,821
15.28
On the second plant, tests were made to determine the advantages of superheating,
and also the effect of varying the vacuum.
TABLE XXII.
Pressure Stop
Valve.
Superheat.
Vacuum.
Bar. =30".
Kilowatts.
Steam per
Kilowatt-hour.
Lbs. per Square Inch.
J57-5
153
125
C.
o
o
o
Inches of Mercury.
26.97
24-45
27.10
IOIO
1041
IO22
Lbs.
23.08
25-25
20.47
These show a gain of about 12 per cent, with 55 C. superheat, and that every inch
of vacuum improves the consumption about 4 per cent.
Steam Turbine Tests
In non-condensing plants also many tests have been made, but, as will be expected,
the steam turbine compares rather more favourably with the reciprocating 1 engine in con-
densing types. In a ico-kilowatt size a consumption of 39 Ibs. per kilowatt-hour has
been attained, and in a 25o-kilowatt turbo dynamo 38 Ibs. per kilowatt-hour, both with
about 130 Ibs. steam pressure and no superheat.
In larger sizes of 1500 kilowatts, with 200 Ibs. steam pressure and 150 F. superheat,
a consumption of 28^ Ibs. per kilowatt-hour non-condensing has been guaranteed, and is
expected to be easily attained, if not surpassed.
At the discussion of these results at the International Congress in Glasgow, 1901,
Mr. Gerald Stoney added the following interesting remarks. He said : " Tests made on
turbines which have been running for a long time showed that there was no falling off
of economy. With reference to superheat, a small degree of superheat gave about 5 to
6 per cent, extra economy : 50 F. 8 per cent., and 100 F. 12 per cent. No tests had
been made at higher degrees of superheat, but there was no doubt the economy would
increase with higher superheats, and would probably reach 20 per cent, with 350 F.
There was no difficulty in working with any degree of superheat the superheater would
stand. The effect of a good vacuum was felt more in the steam turbine than in an
ordinary engine, as it expanded right down to the vacuum of the condenser, which
is not possible in an ordinary engine. In relation to the consumptions of ordinary
engines and turbines at Elberfeld, where both were installed in the same station and
therefore were tested under the same conditions, the consumption of the steam turbines
from three-quarters to full load was better than that of the Sulzer triple-expansion
engines."
These large plants are shown in Plate VII., where the high pressure and intermediate
pressure turbines are at the left-hand side in one casing. The low pressure turbine is
shown at the middle, with a bearing between it and the first two. Then comes another
bearing, and beyond that the alternator and a small exciter.
The turbines of Mr. Parsons were tested on the ship Turbinia by Professors Ewing
and Dunkerly for steam consumption. For first trial purposes the T^^rb^n^a was con-
structed her dimensions being 100 feet in length, 9 feet beam, 3 feet draught of hull,
and 44 tons displacement. She was fitted with turbine engines of 2000 actual horse-
power, with an expansive ratio of i5o-fold, also with a water-tube boiler of great power,
of the express small tube type, but with no feed heater. The turbine engines consisted
of three separate turbines the high pressure, the intermediate, and the low pressure
each driving one screw shaft independently ; to the low pressure or centre shaft the
reversing turbine was also coupled, and on each shaft were keyed three propellers of
small diameter and of normal pitch ratio. This arrangement was found to be the best
after many trials, and has since been adhered to in subsequent vessels. The maximum
indicated horse-power that has been obtained on runs of about 5 miles' duration has been
2300, giving a speed of 34^ knots ; but a speed of 31 knots can be maintained for about
2 hours' duration, and as recently as last August runs at this speed were made at Havre
for the Committee of the Paris Exhibition, the vessel at the time being heavily laden
and with a foul bottom, showing that after four years of work the turbine engines do not
deteriorate in efficiency. Since she was completed she has run several thousand miles,
sometimes in very heavy seas, and the main engines have never caused a moment's
anxiety, nor have any repairs to them been required.
The tests, which occupied more than a fortnight, were very elaborate, and com-
prised feed-water measurements, of great accuracy, at various speeds up to 31 knots.
The water measurements were made by meters which were calibrated before and after
each day's running. At the higher rates of speed two meters were placed in series
so as to check each other ; the errors as determined by the calibration were less
than 2| per cent., and the results were taken as accurate within less than this
amount.
1 86 Modern Engines
The horse-power was determined by model experiments in one of the principal test-
ing 1 tanks in this country.
At the speed of 31 knots the consumption of steam for all purposes was deter-
mined to be 14^ Ibs. per indicated horse-power, the coefficient of propulsive horse-
power to indicated horse-power being taken at 55 per cent. In other words, the
consumption of coal for all purposes with a good marine boiler, under ordinary
conditions and mild forced draught, would be less than 2 Ibs. per indicated horse-power
per hour.
The vessel's reversing turbine, which operated the centre screw shaft only, gave her
an astern speed of 6 knots, and when running at a speed of 30 knots she could be
brought to rest in 36 seconds. It was also found that she could be brought from rest
up to a speed of 30 knots in 40 seconds.
In these tests the indicated horse-power is given, but it is difficult to estimate this
indicated power in turbines.
That such a small vessel could be fitted with power equal to 2000 actual horse-power
strikingly shows the small bulk of such a marine turbine plant.
The De Laval tests were made by Messrs. Erick Andersson, Karl Wallin, and Axel
Estelle, Stockholm.
The turbine dynamo with which the trial took place was provided with a centrifugal
pump coupled to the turbine, and worked directly from one of the inductor shafts. This
pump, at a suction height of 14 feet 9 inches, forced the water at a pressure of 11.5 Ibs.
per square inch through an ejector condenser.
The turbine dynamo was placed close to the boiler used during the trial. The steam
pipe leading to the turbine was provided with a water strainer.
The boiler, which was tubular and provided with an inside furnace, was fed
with water at a temperature of 43 Fahr., and generated steam at 118 Ibs. per
square inch, which was reduced, however, to 114 Ibs. by the throttle valve of the
governor.
The trial lasted eight hours without interruption, and the fuel used was " Best South
Yorkshire Steam Coal," of which 1472 Ibs. were consumed during the trials, while the
steam consumption amounted to 9823 Ibs., consequently the ratio of evaporation was
- 3 = 6.67. The vacuum was i 1 ? Ibs.
1472
The temperature of the condensing water was raised by the steam from 34 to 50
Fahr., consequently the quantity of condensing water may be estimated at about 66
times the quantity of steam, or 1271 cubic feet per hour.
The dynamo developed during the whole trial an average electrical force of 113.4
volts x 324.0 amperes, or II 3'4 x 3 2 4- > e q ua i to 49.92 electrical horse-power, while the
73 6
number of revolutions of the inductors was 1506.
Hence the result of the trial gave a steam consumption of ^- =24.5 Ibs. per
hour per electrical horse-power.
To ascertain the steam consumption at varying loads Andersson and Wallin made
a second trial on the 4th March in the following manner :
1. With the same steam pressure, 114 Ibs., at the nozzle case i.e. in the space
between throttle valve of the governor and the steam nozzles as on I5th February, and
with all of the six steam nozzles in their places, the turbine dynamo was worked, and
developed an electrical power of 113 volts x 326 amperes, or ^ -^ = 50.05 electrical
73
horse-power, consequently the conditions were as nearly as possible the same as at the
previous trial.
2. Then one of the steam nozzles was taken out and inserted into a separate steam
Steam Turbine Tests
pipe leading from the nozzle case to a vessel containing a quantity of water, the weight
of which had been ascertained. The vessel was placed on a sensitive balance, so that any
increase in weight could immediately and accurately be ascertained. With the same
steam pressure as in Sec. i the turbine dynamo was driven with five steam nozzles,
developing an electrical effect of 113.5 volts x 264.2 amperes, equal to an average of
40.74 electrical horse-power. At the same time the stream, which during 12 minutes
rushed through the nozzle mentioned above, was condensed and was found to be 40.5
Ibs., or per hour x 40.5 = 202 Ibs.
12
At the end of the trial all of the nozzles were carefully measured and found to be of
exactly the same cross section, consequently the quantity of steam which passed through
them was 6x202 Ibs. = 1212 Ibs. per hour. If the steam consumption in Sec. i is
1 2 I 2
calculated at the same rate the result is = 24.2 Ibs. per hour per electrical horse-
50-05
power, which differs so very little from the result at the trial on i5th February during
8 hours, but the trifling difference may be owing to a slight variation in taking obser-
vations.
As the turbine dynamo, with five steam nozzles open, developed an electrical effect of
40.79 horse-power, and the calculated steam consumption was 5x202=1010 Ibs. per
hour, in this case the quantity of steam per hour per electrical horse-power would be
1010 ,. ,,
= 24.76 Ibs.
40.79
3. At this test 2 of the steam nozzles were closed, while the steam pressure remained
the same, namely, 114 Ibs. The turbine dynamo, with three steam nozzles opened, now
developed an electrical power of 113.5 volts x 140.8 amperes, equal to 21.72 electrical
horse-power, while the steam consumption, calculated at the same rate as in the previous
case, was 3 x 202, equal to 606 Ibs. per hour, or per electrical horse-power per hour
606
, equal to 27.9 Ibs.
21.72
4. The turbine dynamo was then driven with four steam nozzles open, and
the electrical load was regulated to 113.5 volts x 164.3 amperes, equal to 25.34
electrical horse-power, the steam pressure being reduced by the throttle valve of
the governor to 93.8 Ibs., with a vacuum of 13.27 Ibs. The quantity of steam
consumed was measured in the same manner as in Test II., and was found to be
174.2 X4, equal to 696.8 Ibs. per hour, consequently per hour per electrical horse-power
_5_J_, equal to 27.49 I DS -
2 5-34
5. One of the steam nozzles was next closed, so that, as in Test III., the
turbine dynamo was driven with three steam nozzles open, and the electrical load
regulated down to 113.5 volts x 83. 5 amperes, equal to 12.87 electrical horse-power,
when the steam pressure was reduced to 74 Ibs., with a vacuum of 13.5 inches.
On measuring the quantity of steam consumed it was found to be 137.28x3 = 411.84
Ibs. per hour; consequently per hour per electrical horse -power ^ L_ L_4, equal to
32.0 Ibs.
In the tests with the 300 horse-power turbine the steam was superheated, in the
first case about 60 Fahr., and in the latter about 20 Fahr. It is evident that super-
heating is advantageous to the turbine, as it gives the steam jet a higher velocity and
thus increases the kinetic energy of the steam, and it also diminishes the resistance of
the turbine wheel, as illustrated by Table XXIV.
The use of superheated steam in connection with turbines has become more general
in recent years. Practically any degree of superheating can be used, as the highly
heated steam does not come into contact with the moving parts of the machinery ; by
i88
Modern Engines
the time the steam reaches the chamber in which the turbine wheel revolves it has
already the pressure and temperature of the exhaust steam.
TABLE XXIII. RESULTS OF TESTS WITH DE LAVAL STEAM TURBINES
AT DIFFERENT LOADS.
Pressure
Lbs. of
Turbine Machine.
of Admis-
sion Steam.
Lbs. per
Square
Vacuum
Inches of
Mercury.
No. of
Nozzles
open.
Electrical
Horse-
power.
Steam per
Electrical
Horse-
power
Remarks.
Inch.
per Hour.
50 horse-power turbine f
dynamo.
The test made in April")
1895- I
113.8
113.8
93-9
74.0
26.3
26.3
26.9
27.5
6
5
4
3
49-4
40.2
25.0
12.7
24.6
25.2
27-9
32-5
} Work for
Vcondensing
j included.
100 horse-power turbine (
dynamo.
The test made in June j
1897. I
103.7
103.8
107.4
106.7
25-8
26.4
26.8
27.9
5
3
2
I
92.7
55-6
35-
15-5
22.6
22.7
24.7
27.8
} Work for
Vcondensing
not included.
Lbs. of
Brake
Steam per
Horse-
Brake
Horse-
power.
power
per Hour.
t
113.8
26.4
7
163.0
17.6
i
150 horse-power turbine
116.9
25-9
6
138.4
18.2
Work for
motor.
The test made in Nov-
ember 1897.
113.8
"4-3
112.4
26.2
26.5
27.0
5
4
3
"4-5
88.3
64.1
17.9
18.7
19.0
condensing 1
not included.
116.2
25-7
2
37-5
22.3
-
192.7
27-3
7
303-6
14.1
300 horse-power turbine
motor.
The test made in De-^
cember 1899.
196.3
196.3
196.3
190.6
196.3
27.6
27.6
27.6
27.8
28.1
6
5
4
3
2
255-5
216.9
172.6
121. 6
74-2
14.7
14.4
J 4-5
14.9
17.2
Work for
condensing
not included.
2I 3-3
28.5
I
3i-5
21.6
;
126.6
26.98
8
337-45
15.68
300 horse-power turbine
motor.
The test made in June
IQOO
126.4
125.0
125.0
125.0
26.99
27.24
27.62
27.91
7
6
4
3
293-7
249.1
162.7
118.9
'5-76
15-92
16.25
16.70
Work for
condensing
not included.
125.0
28.16
2
73-5
18.00
125.0
28.25
I
30.4
21.77
From the results of trials it is obvious that not only the steam consumption but also
the heat consumption (in thermal units per horse-power per hour) sinks with increasing 1
superheating-. With constant peripheral speed of the turbine wheel and increasing
superheating the impact is increased on account of the higher velocity of the steam ; but
this loss is instantly transformed into heat, which raises the temperature of the exhaust
steam and thus diminishes the resistance of the turbine wheel.
It has been seen that in the construction of the De Laval steam turbine it is
necessary to adopt a very high speed of the principal driving part of the machine. This
speed is afterwards reduced in the machine by means of gearing, so that the turbine can
be coupled to ordinary machinery.
Steam Turbine Tests
189
TABLE XXIV. TESTS WITH A 30 HORSE-POWER STEAM TURBINE WORKING WITH
SATURATED AND SUPERHEATED STEAM RESPECTIVELY.
NON-CONDENSING Steam pressure, 7 atmospheres absolute = 88. 2 Ibs. Speed 01 driving shaft,
2000 revolutions per minute. Speed of turbine wheel, 20,000 revolutions per minute.
Half Load.
Full Load.
Saturated
Steam.
Superheated
Steam.
Saturated
Steam.
Superheated
Steam.
Temperature of the f Cen % rade
Steam . . .[ Fahrenheit . . .
164
3 2 7
460
860
164
327
500
932
f Metrical brake horse-
Power developed . J E ^ ^^
V, power . . .
21.4
21. 1
24-5
24.2
44.1
43-5
Si-9
51.2
Steam consumption (Kilogrammes per metri-
per brake hSrse J , al b . rak , e h rse-power
, I Lbs. in English brake
power per hour .[ horse _ pow * r . .
21.6
48.3
14.1
3i-5
17.7
39-6
"5
25-7
Heat consumption per metrical brake horse-
power per hour in metrical heat units
I4,l6o
11,270
11,610
939
Temperature of ex- [Centigrade . . .
hausted steam .| Fahrenheit . . .
IOO
212
39
588
IOO
212
343
649
The combination of the De Laval turbine to other classes of machinery is receiving"
consideration, yet experiments in this line are not sufficiently advanced that these
machines can be put on the market.
The De Laval type of turbines excel in smaller units, while the Parsons' types are
essentially large power machines, and it is significant that NEW turbines are made of
large units before any public tests are given.
Small steam engines of any kind are disappearing before the gas and oil engines, so
that the prospect before steam engines is for large units in which they can at present
compete against the internal combustion engine.
There is no doubt the steam turbine has given a new lease of life to steam as a
power working fluid, and that they will altogether supersede the reciprocating engine
for all purposes where steam is used is beyond any doubt now.
And all indications point to the culmination of all the improvements in heat engines
in the evolution of an internal combustion turbine.
Attempts have been made towards this desirable end. Mr. Parsons in his original
patent proposes an internal combustion turbine.
Mr. Ferranti, in Specification No. 2565, 1895, proposes the same thing. A brief
description from this specification will show the nature of the proposal :
In one development of the invention air is used and compressed by means of
pumps through a chamber lined with refractory material and capable of standing con-
siderable pressure which contains the combustible, preferably in the form of coal ; the
working fluid then becomes heated air, and the products of combustion leaving the fire
chamber at a very high temperature and expanding through expansion tubes so acquire
Modern Engines
190
the necessary velocity and diminish in temperature accordingly. The working fluid is
then used in an impact reaction engine.
A portion of the exhaust gases may be used which are compressed with fresh air
through the fire chamber. This is desirable in some cases, so as not to get such
extremely high working temperature, the proportion of air and burnt gases being
settled by experiment according to the temperature at which it is desired to work as
a maximum.
In some cases which are worked with a nearly closed cycle the gases exhaust from
the reaction engine at a pressure of several atmospheres. They are then, by preference,
passed through a re-generator and cooler, after which they are pumped through the
re-generator and the fire chamber together with the necessary amount of fresh air to
support combustion.
In another form the fire chamber burning solid fuel is replaced by a chamber into
which powdered coal or oil is injected together with the air for the purpose of com-
bustion, and to act as a work-
ing fluid ; the chamber is kept
under a high pressure, com- //2^T~ 'h
bustion taking place immedi-
ately and completely, the pro-
ducts issuing at a high velocity
FIG. 227. Ferranti's Hot Gas Turbine.
through expansion tubes and then giving up their energy to the impact engine. In
this case a portion of the products of combustion may also be introduced with or
without cooling, and with or without a regenerator into the fire chamber. Coal or
other gas may be used for compression, and injected in place of the other combustible
into the fire chamber together with the same or other modifications in the cycle above
described.
In some cases the high speed reaction engines are combined with the necessary
pump gear driven for compressing the working fluid through the fire chambers. The
gearing is so arranged that the high speed of rotation of the motor is geared down to
a sufficiently slow speed to work the reciprocating compressors.
When air and the products of combustion are used as a working fluid, the speed
and power are regulated by the amount of the fluid which is compressed through
the fire chamber, the pressure being kept constant, or as nearly so as possible, by
bringing more or less expansion tubes into play to act upon the wheel, the object of
this being to always work with the maximum range of temperature and so secure the
maximum economy at low loads.
When working with steam condensing impact engines or turbines, ejector con-
Gas Combustion Turbines
191
densers are used, in which the necessary power required outside the engine con-
tained in the steam is supplied by means of a high pressure jet of water generated
by a high speed centrifugal pump run direct by the axis of the reaction engine
which supplies a small jet of water which acts
by induction on a larger jet of water in the
condenser.
In some cases a water spray or cooling
jackets may be used in the air compressors, but
where such is the case it is desirable so far as
possible to separate the moisture so that it should
not pass into the combustion chamber.
Fig. 227 shows a combination of hot air fur-
nace, compressing pumps, and reaction engine ;
and a fire chamber arranged for burning gas, oil,
or other suitable fuel is shown in the left-hand
Figure.
At Fig. 227 (z) is a hot air chamber, (/) is
an entrance for compressed air, (k) is an arrange-
ment for introducing the fuel, (/) is the exhaust
for the burnt and heated gases to the reaction
engine (m), (n) is an air compressor.
The air pump is shown driven by a worm
and wheel gear.
None of these proposals are practicable. It
will be observed that the working pressure can
never exceed the pumped air pressure ; it is the
volume of the gases which are increased by the
heat. The same proposals were made half a
century ago for reciprocating piston engines by
Joule. The fluid is taken in at one volume
and rejected at a larger volume, but the ex-
pansion takes place mostly in the furnace, and
the power is proportional to the difference in
area between the pump piston and the engine
piston.
The most successful engine of this type
working on the reciprocating principle is that
known as the Bucket Engine, described in Vol.
II. of this work. I have worked upon the idea
of the internal combustion turbine or hot gas
turbine myself for some years, and my own con-
clusion on the matter is that the combustion
must not be continuous ; and the engine must be
worked by a fluid which is taken into the com-
bustion chamber at one pressure, and sent out at a
higher pressure to expand in the turbine. Experi-
ments led to the adoption of a pressure generator,
as shown in Fig. 228, which is a diagram drawn
to explain the invention, without regard to details
of construction.
It consists of a turbine with nozzles as usual, a combustion chamber into which the
air and fuel are forced or drawn, an ignition plug to fire the mixture. In the diagram
the air and oil fuel enter the combustion chamber by one valve, and pass through a throat,
192 Modern Engines
in order that the fresh incoming charge should clear out the remains of the previously
fired charge.
The air and oil may be drawn in by a suction fan on the turbine exhaust much
as they are drawn into a motor car petrol engine cylinder, but it is preferable to
force them in under pressure. A compression air pump and reservoir is therefore
provided, and the air used at from 50 to 80 Ibs. pressure. The oil reservoir is con-
nected to this same air reservoir, so that the oil and air are at equal pressures at the
admission valve.
The action of the apparatus is as follows :
The reservoir of compressed air is opened by the check valve. Air and oil enter
by the admission valve to the combustion chamber. The ignition plug is sparking ;
immediately the mixture reaches the ignition the mixture fires, and the increase of
pressure closes the admission valve. The hot gases under pressure rush through the
nozzle and drive the turbine. As soon as the pressure falls below that of the air
reservoir, air and oil again automatically enter the admission valve, driving the burnt
gases before them until they reach the sparking plug, when another combustion occurs.
The valve again closes, and the turbine obtains another impulse, and so on the cycle
goes intermittently.
The air pump is driven by gearing from the turbine, and the half of the combustion
chamber near the throat is cooled by water jacket The nozzle is also water jacketed.
In some cases the sparking plug has been tried about the middle of the chamber, in
order to present an elastic gas cushion between the fired gases and the nozzle, and to
prevent flame passing through the nozzle.
The necessity for water cooling, of course, reduces the efficiency much in the same
degree as in the reciprocating engine, although in the turbine a higher temperature of
the working fluid is permissible.
The continuous combustion generator is best adapted for solid fuel, coal, or coke,
while the intermittent system is best for gaseous or liquid fuel.
THE DESIGNING OF TURBINES
No detail designs for steam turbines have as yet been published. Patent specifica-
tions only give the principles and methods, and generally the designs are as far from
actual practice as the Patent Law will allow. Only by actual dissection of working
turbines can the actual construction be found. For the benefit of engineers who may
desire to have a more intimate acquaintance with turbines' design and construction,
I shall here conclude this chapter by a design which has been made and tried by
myself. There are many who may find a use for small steam turbines, easily and cheaply
made, not of transcendental economy in steam, but as good as small steam engines
reciprocating, and of same output of power.
Two types of turbines were chosen : first, the Pilbrow double wheel type, described
in his Specification No. 9568 of year 1843, and shown in diagram Fig. 163, the two
wheels running in opposite direction and coupled by gearing.
The Seger steam turbine, illustrated in Fig. 210 (p. 169), is of this type, the gearing
being belting.
The proper form of the passages in the two wheels is shown in the diagram Fig. 162.
The steam enters from nozzles converging on the first wheel, into which it flows at a
pressure about 10 to 15 per cent, less than that due to the boiler, with a velocity of about
1500 feet per second. It thus drives the first wheel by impact and pressure. It expands
in passing the first wheel, and acquires more velocity up to about 2000 feet per second.
It then enters the second wheel, and by impact drives that wheel in the opposite direction.
Many attempts have been made to carry this action further, with a view to reducing the
speed of the wheels, by adding more wheels in series, but such a practice is fallacious.
Steam Turbine Design
1 93
With these partial flow wheels two wheels only give the maximum results. All other
arrangements are contrary to theory, and obviously bad. If we want slower velocities
to deal with, then the second type of turbine, herein described, must be adopted, that
described by Wilson in his Specification, No. 12,060 of year 1848, and perfected by the
Hon. C. A. Parsons, and now well known.
There is no getting away from the fact that either one or the other type is the only
choice for simple construction and good performance.
If we attempt to put up, say, a series of De Laval wheels we meet with immense
mechanical difficulties, not to speak of steam distribution difficulties. In large size
turbines these difficulties, by high-class tools and workmanship and fine designing,
can be and have been overcome to a large extent in machines over 100 horse-power.
Engineers who make turbines over 100 horse-power can generally get all the designs
400
350
300
250
s*
s
**
s
S
Q
j>
<"
s
i
s
s
*i
s
'
\
/
f
s
'<
s
f_ ^t
'
si
,s
*
<2
f
/
. S
*R
t
t
i
f
v
/
/
Is
/'
-
/
f
5
/
/
s
/
>
.
*
>
..
f
41;
/
V
/
-A
1
f
f
%
3 '
{
a
C
y
y
t
5
10 15 20 25 30 35
FIG. 229. Turbine Speed Curve.
40
45
50
they require from a highly trained and paid staff. The designs now under consideration
are intended as instructive and useful to those who may have to make their own designs
and work out smaller affairs.
Taking the first design, the two wheels are our first concern. We shall assume
a turbine is required for 5 horse-power, steam pressure 100 Ibs. at the boiler, non-
condensing. It will give much more power condensing, and may so be used by
arranging a proper packing for the shafts, to prevent air being drawn in.
In this class of turbine, with a single wheel, a peripheral speed on the middle
diameter of the buckets is less the smaller the turbine is, owing to the weakness of
smaller wheels in comparison with larger ones. Thus a De Laval 5 horse-power wheel
is made for 515 feet peripheral speed per second, while the 300 horse-power machine
runs at 1378 feet.
The revolutions in the first case are 30,000, and in the latter 10,600 per minute.
As the turbine under discussion has two wheels, each to run at half the velocity of a
single wheel, I have therefore, from the published results of the De Laval wheels and their
VOL. i. 13
Modern Engines
speeds, which have been found in practice satisfactory, made a curve for double wheels
(Fig. 229) which shows the relation between horse-power and peripheral speeds at half
the revolutions of a single wheel, the 5 horse-power machine having a speed for each
wheel equal to 250 per second.
TABLE XXV. TABLE SHOWING VELOCITIES AND HORSE-POWER OF TURBINE
WHEELS NON-CONDENSING. ( Theoretical. )
Initial
Steam
Pressure,
Lbs. per
Square
Inch.
Counter-Pressure, I Atmosphere.
Initial
Steam
Pressure,
Lbs. per
Square
Inch.
Counter-Pressure, I Atmosphere.
Velocity of
Outflow of
Steam,
Feet per
Second.
Kinetic
Energy,
Foot-lbs.
per Second.
Horse-power
of 550
Foot-lbs.
per Second.
Velocity of
Outflow of
Steam,
Feet per
Second.
Kinetic
Energy,
Foot-lbs.
per Second.
Horse-power
of 550
Foot-lbs.
per Second.
Per Lb. of Steam per Hour.
Per Lb. of Steam per Hour.
60
2421
25.29
0.046
160
2992
38.63
0.070
80
2595
29.06
0.053
1 80
3058
40.35
0.073
IOO
2717
31.86
0.058
200
3"5
41.87
0.076
1 20
2822
34-37
0.062
220
3166
43.26
0.079
140
2913
36.62
0.066
280
3294
46.83
0.085
There being two wheels running in opposite directions at half-speed, 250 each, we
may calculate the steam required at the same rate as for one wheel going at 500 revolu-
tions per second, the object of the two wheels being to reduce speed only.
From Table XXV. we find that with 100 Ibs. pressure and the theoretical velocity
2717 per second the horse-power is 0.058 per Ib. of steam, so that if the turbine is to be
5
5 horse-power the consumpt of steam would be
0.058
= 86 Ibs. per hour.
But we cannot run wheels at such high speeds in practice. The strength of material
limits the speeds, and the 5 horse-power wheel comes out at 4 inches diameter at the
mid-line of the buckets. The peripheral speeds are given in the curve (Fig. 229) up
to 50 horse-power, and these speeds multiplied by a constant, which appears to be 2,
gives the maximum revolutions of a single wheel. This constant will then be equal
to i for double wheels at maximum speed.
The steam required is inversely proportional to the peripheral velocity ; hence we
have found that, as 86 Ibs. per hour are required at the high theoretical speed, we can
find what is necessary at the lower practical speed, 500. We will reduce the steam
o/r
consumption to Ibs. per second, =.024; then 500 : 2717 :: .024 = 0.1304 Ibs. per
3600
second, 472 Ibs. per hour.
We have seen that the flow ot steam at its maximum is Q = 37o AD in Ibs. per
Q
minute, where A is the area of the nozzle at the narrow end ; hence
8
= A.
Q in this case equals 8 nearly, therefore
8
100 Ibs. pressure = 0.23. Therefore
370x0.23
370 x D
-^p: = A, and D the density of steam at
370 D
= 0.09 square inch, equal to one jet of
c
inch diameter, or two jets each of about inch.
32 32
Double Wheel Turbine
'95
The quantity of steam seems large, but for a small non-condensing engine at
100 Ibs. boiler pressure it is by no means uncommon. The turbine gains more by
condensing than a reciprocating engine, so that probably half the steam would be saved
by condensing.
The jets would be reduced in area by one-half for condensing, as may be gathered
from the following table, in which it will be seen that the horse-power is doubled at
100 Ibs. pressure, exhausting into a 28-inch vacuum:
196
Modern Engines
TABLE XXVI. TABLE OF VELOCITIES HORSE-POWER EXHAUSTING INTO VACUUM.
Steam
Pressure,
Lbs. per
Square
Inch.
Counter-Pressure 0.93 Lbs. per Square
Inch absolute, corresponding to
28-inch Vacuum.
Steam
Pressure,
Lbs. per
Square
Inch.
Counter- Pressure 0.93 Lbs. per Square
Inch absolute, corresponding to
28-inch Vacuum.
Velocity
of Outflow
of Steam,
Feet per
Second.
Kinetic
Energy,
Foot-lbs.
per Second.
Horse-power
of 55
Foot-lbs.
per Second.
Velocity
of Outflow
of Steam,
Feet per
Second.
Kinetic
Energy,
Foot-lbs.
per Second.
Horse-power
of 550
Foot-lbs.
per Second.
Per Lb. of Steam per Hour.
Per Lb. of Steam per Hour.
60
3680
58.44
o. 106
160
445
70.61
0.128
80
3793
62.08
0.113
180
4091
72.22
0.131
IOO
3871
64.66
0.118
200
4127
73-50
0.134
120
3940
66.99
0. 122
220
4159
74.64
0.136
T4O
3999
69.01
0.125
280
4229
77.18
0.140
We have now seen that the wheels' diameter must be calculated at the velocities as
given by a curve (Fig. 229), and the area of the nozzles is calculated from the steam
required per second ; and having found these, the design becomes a mechanical
problem.
This double wheel turbine, as designed, is shown, actual size, in the three Figs.
230, 231, and 232.
Fig. 230 is a sectional plan, showing the two wheels A and B carried on flexible-
ended shafts SS.
Fig. 231 is a side elevation, showing the base, the wheel gearing, and governor.
Fig. 232 is an end view, showing gearing wheels, governor, and regulating wheel
for cutting off nozzles, and governor driving band.
Referring to Fig. 230, the special features are the gearing, consisting of an external
and internal cut toothed wheel E and I. By this device the two turbine wheels are geared
to one counter-shaft, and their speed reduced and power added.
If the large wheels are made of raw hide leather, compressed into gun- metal
shrouds, the wheels are both silent and durable. The wheels are made heavy in
order to act as buffer fly-wheels, so that in the event of a sudden heavy load being
applied, such as a short circuit on a dynamo driven by the machine, the inertia of
the wheels would absorb the shock before it could be transmitted to the thin flexible
shafts SS.
The metallic packing at the bearings PP is specially designed to work tight, either
on pressure or vacuum. Rings on the shaft made of anti-friction metal slip thereon. A
good fit between these rings are cast iron or steel spring rings, which fit the bore
closely ; a spiral spring keeps the rings in contact sideways. These packings are
lubricated by gravity feed. In the practical design of larger engines, bearings are
placed beyond these packed bearings to take the strains up.
The nozzles are bored in the body of the casing, and regulated by a cover ring
sliding over them, and which is moved by hand wheel W and a rack and pinion, shown
in Fig. 233.
In Fig. 231 the counter-shaft C is seen passing through the base, as also is the
governor G fixed to the stop valve V and driven by belt.
This governor can be also driven by direct connection to the shaft C, and the valve
Double Wheel Design
197
operated by a simple lever. The exhaust is connected to the end cover, preferably at
the bottom.
(/)
The turbines work extremely well, and without attendance of any special kind.
Modern Engines
Recently an improved system of forming- buckets on the wheels has simplified the
construction (see page 168).
The sliding cover ring for regulating the nozzles is shown in Fig. 234, where it will
be seen that with three nozzles the cover has one hole same size as nozzle inlet, another
hole twice the length, and another thrice the length of the first, so that by sliding the
ring the holes may be closed in succession. This may be done by a governor and relay
in larger turbines.
This type of turbine is intended for powers from horse-power to 50 horse-power.
For larger powers the second design conies in. As it has been fully described in the
foregoing pages, under the description of Parsons' turbines, we need not go into the details.
FIG. 232. End View of Double Wheel Turbine. Scale, actual size.
Theoretically, the difference in pressure between wheel and wheel in a series of
wheels should be very small. P is the pressure at the inlet side, and P l the pressure at
the outlet side.
P! should be considerably greater than of P. The maximum flow of steam is
p
obtained when ~ l = 0.577, and P 1 = 0.577 x P. Consequently if we had a nozzle like that
shown in Fig. 108, with steam at 100 Ibs. pressure entering at the wide end, the
pressure at B, just beyond the throat, would be P : = 0.577 x 100 = 57.7 lbs -
Regulate Slide for Turbine 199
The pressure in one wheel should, in fact, differ by only a few pounds from that of
the next ; the quantity of the steam passed will then be Q= 1.75 k*J(p-p\)p\> wherein
Q = weight of steam per minute.
A = area of passages in square inches.
p = absolute pressure inside one wheel.
p^= ,, ,, ,, the next wheel.
Thus if A = o.i square inch, and p=ioo, and ^ = 95 Ibs. per square inch, then
Q = 1.75 x o.i N/(IOO - 95)95 = 3.7 Ibs. per minute, or a quantity Q ? of 16 cubic feet
of steam. From this the velocity can be found in feet per second = ^ 23 _ii
60 x A 60 x o. i
= 384 feet per second, from which it will be at once seen that the velocity of the wheel
can be brought much nearer the theoretical at not a very great velocity.
But this small drop in pressure can only be obtained by a very long series of wheels,
with very small blades and close
together : conditions easily enough
met in large turbines, but very diffi-
cult in small ones. Hence in this
class of turbine it is not advisable
to make them below 20 or 25 horse-
power.
The chief cause of their ineffici-
ency in small sizes is due to the FIG. 233. Rack and Pinion of Double Wheel Turbine,
clearance spaces required for free
rotation and the consequent leakage of steam past the blades. The clearance is much
greater in proportion to the blade area in a small turbine than it is in a large one.
The problem is to design a wheel with blades easily made and with very many
blades of thinnest material on the wheel, the blades to be increased in area by increase
of radial length, and to be very narrow axially.
In order to make wheels to meet the requirements, a plain disc is turned up true
and with a rim (Figs. 208, 209).
The disc is slotted out all around,
forming slots into which the blades
are firmly driven ; these blades are,
at first, plain steel stampings ; the
upper part is then bent into shape.
They are then inserted in radial
slots in the wheels, and the lower
FIG. 2^4. Nozzle Regulator of Double Wheel Turbine. .
part flattened back tangentially,
thus firmly fixing the blades ; they are afterwards further fastened by rivets. In this
way a light strong wheel is formed. It is important for economy of steam that the
wheels should be narrow and the blades numerous and thin. This design meets these
requirements.
CHAPTER IV
ROTARY PISTON ENGINES
THIS class of engine is but little used as a prime mover. It is, however, used to some
extent as a reversed motor for pumping" water and blowing air. And it would seem
that, even in these humble capacities, it is to be superseded by the free-running centri-
fugal or screw pumps for both air and water. However, no work on prime movers and
FIG. 235. Diagram of Rotary Piston Engine.
modern engines would be complete without including an example or two of the best
of the class and latest developments in their construction.
Perhaps more brains and money have been spent on the invention of such engines
than on steam turbines, and yet they have never reached any practicable design
200
Rotary Piston Engines 201
approaching that of the turbine. Most middle-aged engineers can recall to mind many
brave attempts to run rotary engines successfully, all introduced with high hopes and
determination to succeed, all with same discouraging end. I can remember in my
youth two vessels fitted out with rotary engines in Scotland by a determined inventor,
who fully believed he had solved the difficulties. They both sailed away, never
again to be heard of: an impressive fact, which forcibly sent home the lesson that no
uncertain experiments should be made on sea-going vessels.
The mistaken idea underlying nearly all the inventions in this direction seems to
be that, by substituting a rotary piston for a reciprocating piston, the to-and-fro
motions will be abolished, and therefore only smooth regular rotary motion obtained,
with all the advantages of such a regular turning movement. But if one turns to the
patent records of these inventions and examines them all carefully, in every case
wherever an engine has been designed at all capable of passable working for a time it
will be found that there are quite as many reciprocating parts, to-and-fro movements,
as in a simple piston and crank engine.
It will also be found that the pistons, or doors, or sliders in the rotary are subjected to
great rubbing friction at high speed, causing unequal wear and tear and eventual failure.
One diagram (Fig. 235) will illustrate the whole class. A is a shaft running through
a cylinder C and supported in stuffing boxes in the cylinder covers, between which
covers it fits steam-tight without end-play. It has a wing B, which forms the rotary
piston. This wing must fit steam-tight at the ends and at the periphery, and many
ingenious devices have been devised to pack this piston tight. E is a door or shutter,
pivoted at F, so that it can shut up into D, to clear the piston when it comes round.
G is .the steam inlet and H the exhaust pipe. Theoretically this looks all right, and,
beginning with this simple conception, the mechanical inventor adds improvements
and alterations innumerable, and produces an engine which has all the defects of the
simple affair sketched, only toned down and made partially tolerable.
BROWN'S ROTARY ENGINE
In Fig. 235 the wing piston is shown fixed to the shaft, but it may be hinged to
it, so that the steam will press it tight at the periphery. Some inventors make the
piston oval or eccentric. We shall, however, only consider briefly the last and best
attempt on a practical scale, as described by Professor Jamieson. Fig. 236 is a cross-
section and sectional plan of this engine made with double doors. The following is
an index to the parts :
SB = Steam branch. PS = Packing strips.
SP = ,, ports and passages. SS = Spiral springs.
SV = ,, valves. S = Shaft.
VS = Valve spindles. ED = Exhaust door.
C = Cylinder. EP = ,, ports.
P = Piston. EB = ,, branch.
The steam enters by steam branch SB into the steam passage SP.
This steam passage thus forms a steam jacket to half the cylinder. It may be
arranged so as to form also a "water trap" (with a return "loop" to the boiler), so
that nothing but dry steam shall enter the cylinder. The underneath side of the
cylinder is steam jacketed by the exhaust steam. The outside of both cylinder covers
and jackets are carefully lagged with felt, and covered with wood to prevent radiation.
The steam then passes through the steam port in the left-hand revolving steam
valve SV, into the cylinder, whenever the piston P has passed the nose of the left-hand
oscillating exhaust door ED. The piston is thereby forced round under full steam
pressure until it reaches the point of "cut off," when the left-hand rotating valve closes
the steam port. For the remainder of the stroke the piston is propelled by the natural
202
Modern Engines
expansive force of the steam (enclosed between it and this steam valve) until release
takes place through the exhaust port EP, in the boss of the right-\\a.nd exhaust door ED ;
from whence the steam passes along the exhaust passage EP, and through the exhaust
branch EB, to the atmosphere or to a condenser, according as the engine is working as
a noncondensing or condensing one. Precisely the same action takes place during the
remaining half of the piston's revolution by aid of the n^A/-hand steam valve and the
left-hand exhaust door.
vs
Cross Section.
Sectional Plan.
FIG. 236. Brown's Rotary Engine.
Nothing could be simpler under the circumstances, or more effective ; for full
advantage is taken of the expansive properties of steam by so arranging the "lead" of
the eccentric (which turns the steam valves) that the "cut off" may take place at any
desired proportion of the piston's stroke; or the "cut off" may be directly varied
according to the load and steam pressure by connecting a loose eccentric to the governor.
It will be observed that the front edges and surfaces of the exhaust doors
make a steam-tight fit with the circumferential surface of the piston, and that when
these doors are closed they form part of the bore of the cylinder. They are forced
Rotary Piston Engines
203
forward by the incoming steam pressing behind them, whenever the peculiarly shaped
piston has passed them, and they are pushed gently back into their seats by the piston
(during expansion and release) against the reaction of the pent-up or "cushioning"
steam retained between them and the steam valves.
The working surfaces of the piston, steam valves, and exhaust doors are all packed
by gun-metal strips, kept up to their bearings by small adjusted spiral springs. From
the perspective view of the piston (Fig. 237) it will be seen that the side and end packing
strips are checked into each other, as well as the side strips into the ring ones next the
shaft. In order to make the wear
uniform on the side packing strips
the spiral springs are made propor-
tionately shorter the farther they are
from the centre of the shaft ; and,
since each of these springs are con-
tained in the same length of hole,
their outward pressure on their strip
is inversely proportional to their FIG. 2 37 .-Brown's Piston,
radial distance from the centre of
the shaft ; or, in other words, inversely to their travel.
BROWN'S SINGLE-DOOR ENGINE
From the description which we have just given of the double-door engine the
construction and action of the single-door one will be easily understood from Fig. 238.
This form of engine has not been steam jacketed, as it is intended for rougher and
smaller work than the double-door one, such as driving fans, etc.
The points claimed for the engine are :
1. Packing strips that are expected to wear uniformly and last a long time without
requiring adjustment or renewal.
2. Early cut off with corresponding economy, due to taking full advantage of the
natural expansive property of steam.
3. A minimum of clearance space.
4. Jacketing.
5. A thoroughly well balanced strong engine, occupying a minimum of space for
the power developed, freedom from undue vibra-
tion and noise, combined with high speed and
an economy of steam that has not been reached
by any other rotary engine, or even by any
simple reciprocating engine of the non-conden-
sing and non-compound type.
6. The possibility of compounding his
engine, or even of adopting triple or further
multiple expansion, and of using the highest
practicable pressures.
The high speed of the rubbing points of the
doors, and that under great pressure, is obvious,
-PS
FIG. 238. Single-Door Engine.
even in this otherwise excellent design, but the number of reciprocating parts is great,
and it did not come into practical use.
THE ROTA ENGINE
In order to get over the difficulty of this high speed friction between the piston and
cylinder or piston and doors a promising idea was patented by Mr. Macewan Ross, in
204
Modern Engines
which the cylinder is made to revolve with the pistons, so that their relative speeds
are much reduced. He called the engine the "Rota" engine. Figs. 239 and 244
represent the engine in sectional elevation. It consists of a cylindrical casing A with
closed ends or covers, through whose axis from either end pass two central hollow
drums B fixed to and carried by their respective brackets XX, and having cast in them
ports for admission and exhaust of steam at both ends. The drums B (Fig. 244) are
A
FIG. 239. Section of Rota Engine.
divided by central webs running their entire length to form admission and exhaust
passages bb', from which the inlet and exhaust ports CC' are carried to the interior of the
casing. A series of segmental pistons EE are fitted between the casing A and drums B,
which fill up to a greater or less extent what forms the steam space within the casing.
The crank shaft D (Figs. 239 and 240) passes through bearings bored eccentrically
in the drums, and is made with four arms or cranks which also act as slides, and which
are fitted into slide blocks F fitted into their respective
pistons. The pistons are thereby held in equal and equi-
distant angular positions relatively with the axis of the
main casing A ; the slide blocks in the pistons are allowed
to oscillate in order to accommodate themselves to the
different angular positions of the cranks, which recipro-
cate radially through them relatively with the axis of the
main casing, as the distance of the piston from the axis of
the crank shaft varies according to their position round
the drum B. Their distance apart also varies, those
pistons bearing on the portion of the drums having least
eccentricity being close together, and those on the opposite side being farthest apart.
When the pistons travel round the drum their speed is highest at the point of greatest
eccentricity, and the ports CC' are formed to admit steam between each pair of pistons
as they pass the point of least eccentricity, and to exhaust as they pass the line of
greatest eccentricity of the drums. The steam is cut off at any desired point, and
FIG. 240.
Rotary Piston Engines
205
expands, while the piston in front of it is impelled forward at a greater velocity than
that following it. The intervening space increases until the point of greatest eccentricity
is reached, when the exhaust of steam between that pair of pistons takes place through
the ports C' in the drums.
Each succeeding piston is acted
upon in like manner, and a
continuous rotative motion is
thus imparted to the crank
shaft, there being no dead
point, and all parts are in
FIG. 241. Pistons of Rota Engine.
FIG. 242. Casting of Pistons.
perfect equilibrium. It must be clearly understood that the drums are fixed into the
brackets, while the casing is in no way connected either to the crank shaft or to the
pistons, but is free to rotate upon the drums and accommodate itself to the friction
within the engine. Each
cover of the casing has
cast upon it a sleeve,
which is carried along
around the drum into
a staffing box in the
bracket, and those sleeves
form the bearings upon
which the casing re-
volves. The sole plate is
so made that steam enter-
ing at branch pipe S (see
Fig. 239), to which the
governor is fixed, passes
up a steam way formed in
bracket X, also along one
side of the sole plate to
bracket X', and so is ad-
mitted into the interior of
FIG. 243Rota Engine by Macewan Ross. the casin S thr <>ugh both
drums at the same time.
Similarly the exhaust steam passes down both brackets, and exhausts at branch T.
It will be seen from Fig. 241 that the steam chamber between the bottom pair of
pistons is just opening to steam, and the chamber at the top is opening to exhaust.
2O6
Hult Rotary Engine 207
In order to describe more clearly the cycle, suppose the middle of a side chamber
to be full of steam, then the pressure would act against both pistons equally ; but
you will observe that the upper crank is much longer than the lower one, so that
the upper crank will turn in an upward direction, taking 1 the others round with
it, until the chamber reaches the top centre, when exhaust will take place through
port C' ; and meanwhile the bottom chamber will be round to half-stroke and getting
full steam, so that there are always two cranks doing work at the same time, and thus
there is no dead point. The pistons therefore act as their own valves (the ports being in
the fixed drums), so that the chambers only take what steam they require, and leave no
passages full of steam after cut off takes place.
There is a steam space formed in the outside of each piston, so that the
accumulated steam pressure acting between them and the casing is made almost to
balance the centrifugal force of the pistons. It must also be noted that there is no
side pressure on the pistons, and therefore if fitted accurately at first they should wear
well.
This accuracy is easily attained, as all the work is done on the lathe, the pistons
being cast all together, as shown on Fig. 242, with distance pieces between each, and
are not separated until they have been turned and fitted to the drum, casing, and covers.
Each piston and crank arm is fitted with suitable springs, those in the cranks being so
arranged as to mutually accommodate themselves should there be any end movement
of the shaft. Let us now consider some of the results which are produced when the
engine is working. The pistons, which are close together when in the bottom position,
are 3 inches apart when at the top, which is 3 inches travel in half a revolution, or a
speed of 6 inches per revolution.
Now, suppose the engine to be running at its normal speed 800 revolutions per
minute the relative speed of piston is 800 x .5 feet = 400 feet per minute, and as there
are four steam spaces we have a total " outward " stroke in each revolution of 12 inches,
and yet the speed of piston is only 400 feet per minute.
The speeds of the various surfaces working in contact are also very low, as will be
seen from the adjoining table :
Pressures on Surfaces Speeds of Surfaces
working in contact, in working 1 in contact,
Lbs. per Square Inch. in Feet per Minute.
Outside of pistons and casing . . 50 283
Side blocks and crank arms . . 435 266
Crank shaft journals ... 78 366
Casing journals on drums ... 40 900
Although the latter speed, namely, the bearings of the casing, seems high, still it is
under constant pressure, and is well lubricated from holes being bored through into the
exhaust passage V in the drum.
The novel principle in this engine was the cylinder revolving with the pistons.
Fig. 243 is a view of the complete engine with the shaft fitted with a coupling.
THE HULT ENGINE
Starting from the same point as Macewan Ross, but some years later, this inventor
made some improvements on the "Rota" engine type, and by introducing roller and
ball bearings reduced the friction.
Starting from this point, the inventors of the Hult engine have made the cylinder
participate in the rotary motion of the piston, thus substituting rolling friction for the
sliding friction, which had shown itself fatal to efficiency in other rotary engines, while
208
Modern Engines
they are thereby enabled to readjust the contacts between the piston and the cylinder ;
these advantages belong- to no other system.
PRINCIPLES OF THE HULT SYSTEM
The principles of the Hult engines are shown in the schematic Figs. 245 and 246.
In order to simplify this description of the principles, only one shutter or door is shown,
instead of two or three shutters placed at respectively 180 and 120, with their steam
passages for each shutter ; for some purposes as many as six shutters have been placed
in the same piston.
The piston F, placed eccentrically within the cylinder C, is shrunk on the main
shaft A, and is provided with passages for the admission of steam H and for the
exhaust I ; the sliding shutter G is placed near the admission passage.
The steam from the boiler enters through the hollow shaft A and the passage H
into the space enclosed between the piston, the cylinder, and the shutter, and pushes
FIG. 245. Single-Door Hult Engine.
round the shutter for turning the piston, the shutter having the same length as the
piston.
When the piston reaches a certain angle the admission of steam is cut off auto-
matically by a valve placed inside or at one end of the hollow shaft. The volume of
steam introduced then works by expansion upon the shutter G, until the pressure
becomes low enough, when the rotary motion of the piston, eccentric to the cylinder,
uncovers the openings for the exhaust steam. There is no organ for admission or
exhaust of steam in the cylinder itself.
Centrifugal force acting upon the shutter ensures a perfectly steam-tight fit between
the shutter in the piston and the cylinder, both revolving in the same direction.
The shaft with the piston runs on the roller bearings B, K. The cylinder runs quite
free on the roller bearings E, L ; it is drawn round by the piston in consequence of the
pressure between them at the line of contact. The piston and the cylinder have thus
the same circumferential speed, but in consequence of their eccentricity the angular
speed of the piston is slightly higher.
In the Hult engines with two shutters, placed at 180 to one another, as shown in
Figs. 247 and 248, there are two periods of steam admission during each revolution,
Hult Rotary Engine
209
and in the engines with three shutters, shown in Fig. 250, there are three such periods.
The greater number of shutters increases the power of the engine, while reducing the
condensation within the cylinder.
As the piston and the cylinder both revolve uninterruptedly in the same direction,
the piston speed can be conveniently kept up to 50 or 60 feet per second, or about twice
the speed of reciprocating engines.
The cylinder bearings and those supporting the shaft and piston are all roller
bearings with tempered steel rollers, kept at their relative distances by guide rollers,
as shown in Fig. 246; the rollers run on racers fixed in the frame and frame covers
respectively.
When easing the nuts holding the frame covers the adjusting screws on the top
Bearings of Cylinder.
Bearing's of Shaft.
FIG. 246. Roller Bearings of Hult Engine.
of the frame can press the piston and the cylinder together at the line of contact, so
as to ensure that the cylinder follows the rotation of the piston, while keeping the line
of contact steam-tight. The ends of the piston are also adjustable in relation to the
cylinder, as described farther on.
Referring to Figs. 247 and 248, the following is an index to the parts :
Cylinder.
Cylinder end on governor side.
Cylinder end on pulley side.
Adjustment disc for piston ends.
Adjustment wedges for piston ends.
Adjustment screws for piston ends.
Piston.
Sliding shutters.
!- Steam admission passages.
Exhaust passages.
Exhaust orifices in piston.
VOL. I. 14
13. Main shaft.
14. Frame.
15. Frame cover on governor side.
16. Frame cover on pulley side.
17. \Adjustment screws for line of contact between
18. ) piston and cylinder.
19. Cover screw for reaching adjustment screws 6.
2O* ~\
' | Covers for shaft roller bearings.
22. Inner racer for cylinder roller bearing.
23. Outer racer for cylinder roller bearing.
24. Inner racer for shaft roller bearing.
Hult Rotary Engine
21 I
25. Outer racer for shaft roller bearing.
26. Steel rollers in bearings.
27. Guide rollers.
28. Hoop for cylinder guide rollers.
29. Hoop for shaft guide rollers.
30. Steam admission box.
31. Flange for admission box.
32. Fixed steam distributing tube.
33. Revolving steam distributing tube.
34. Governor valve.
35. Connecting-rod between revolving distributing
tube and admission valve.
36. Revolving steam admission valve.
37. Governor box.
38. Governor disc.
39. Governor weights.
40. Governor pivot.
41. Governor lever plate.
42. Governor lever.
43. Rod for turning automatic lubricator.
44. Holding-down bolts.
45. Cover bolts.
46. Exhaust pipe.
47. Coupling.
THE HULT TWO-SHUTTER ENGINE (Fics. 247, 248, AND 249)
The steam from the boiler enters into the box 30, passes through tube 32, and
reaches the cylinder through passages 9 and 10, every time one of them has passed
the line of contact be-
tween the cylinder i and
the piston 7. The steam
acts on the sliding shutter
8, at first with full pres-
sure, until the moment of
the cut-off, and then by
expansion.
The distributing ar-
rangement, placed in the
hollow main shaft 13,
consists of an inner fixed
tube 32, an outer revolv-
ing tube 33, and a rod
35, connecting the tube 33
with the admission valve
36, which is perforated by
a number of slots, through
which the steam can enter
into the fixed tube 32.
The tube 33, the rod 35,
and the valve 36 thus
participate in the rotary
motion of the piston 7.
The revolving tube 33 is
pierced with slots opposite
to the passages 9 and 10,
u
I
FIG. 248. Steam Ports. Double-Door Section.
so that steam at full pres-
sure is admitted into the
cylinder through these
passages every time any of the slots passes the corresponding orifice in the fixed
tube 32.
The governor, placed in the box 37, is of a centrifugal type ; its disc 38, fixed
to the steam admission valve 36, at the end of the rod 35, which is bevelled so as to
allow steam to pass, turns at the same speed as the piston 7. The weights 39,
equilibrated by a spring and mounted on the pivot 40 on the disc 38, are connected
by levers 42 with the plate 41 on the rod of the governor valve 34, which has openings
212
Modern Engines
corresponding- to those on the admission valve 36. The weights 39, when separated
by centrifugal force, pull at the levers 42, and give to the governor valve 34 an addi-
tional rotary motion, independent of that of the admission valve 36.
The expanded steam escapes from the cylinder through the passages n and the
orifices 12 on the end faces of the piston.
In consequence of the eccentricity of the piston in relation to the cylinder there are
open spaces inside the end pieces of the cylinder to allow for the movement of the shaft.
As the position of the transversal surfaces of these spaces does not change during the
rotation of the engine, the escapement periods are determined by the shape and section
of the orifices 12.
The exhaust steam then spreads throughout the interior of the frame 14, and issues
finally through the exhaust pipe 46 to the condenser, or into the outer air.
The roller bearings for
the cylinder consist of tem-
pered steel rollers 26, run-
ning- on steel racers 22,
fitted on the annular parts
2 and 3 of the two cylinder
ends. The distance between
the rollers is kept constant
by guide rollers 27, mounted
on pins fixed on hoops 28.
Two other steel racers 23
are fitted in the frame 14,
outside the rollers concen-
trically to the inner racers
22.
The driving shaft 13,
carrying the piston 7, is
supported by other similar
bearings, consisting of the
rollers 26, the inner racers
24, guide rollers 27, hoops
29, and outer racers 25.
These shaft bearings,
independent of the cylinder
bearings, can be adjusted
vertically by turning the adjustment screws 17 and 18, which give to the piston 7 the
exact pressure against the cylinder necessary for rotating the cylinder, and for ensuring
a perfectly steam-tight fit at the line of contact.
A disc 4 is placed at the left end of the piston for adjusting a suitable steam-tight
fit between the ends of the piston and the cylinder ; the position of this disc is modified
by the adjustment screws 6 pressing- upon the wedges 5, which screws are reached by
a screwdriver on removing the cover screw 19.
Whatever slight friction there may be at the ends of the piston is almost inappreci-
able, as the cylinder rolls freely in the same direction as the piston, and with but slight
difference in the angular speed, while their circumferential speed is the same.
The question of lubrication is important for all high speed engines. The Hult
engine is easily lubricated, because rolling friction is almost entirely substituted for
the sliding friction existing in all reciprocating engines ; the steam on entering the
engine takes up the exact quantity of lubricating oil required from one single automatic
lubricator driven by the rod 43, set in motion by an eccentric cam on one of the
hoops 29.
FlG. 249. Steam Ports.
Hult Rotary Marine Engine 2,13
THE HULT THREE-SHUTTER ENGINE (Fio. 250)
This is in most respects similar to the two-shutter engine, except that there are
three shutters, with three admissions of steam, and the cycle of admission, expansion,
and expulsion of steam for each shutter is completed during one revolution of the shaft.
When a three-shutter piston with its arrangements for admission of steam is
substituted for a two-shutter piston in the same engine the power is increased, and
it becomes perhaps even better equilibrated than before ; while the condensation of
steam in the cylinder be-
comes, if possible, even
less detrimental.
When the exhaust com-
mences of the steam, which
has acted upon one of the
shutters, the next shutter
is driven by expanding
steam, and the third
shutter by steam at full
pressure ; while immedi-
ately afterwards, and be-
fore the said exhaust is
completed, fresh steam is
again let on to the first
shutter.
Fig. 250 shows in trans-
verse section the steam
admission arrangement of
a three-shutter engine, of
the same size as that of
which a longitudinal sec-
tion is given in Fig. 247.
The three shutters are
marked 8, the three admis-
sion passages 10, and the
three exhaust passages 1 1 ;
the steam admitted for one
shutter through the pas- FIG. 250. Three-Door Hult Engine,
sage IOA acts upon the
shutter SA, and is exhausted through the passage IIA; and in the same manner IOB,
SB, and IIB, as well as ice, 8c, and nc, co-operate together.
THE HULT REVERSIBLE ENGINE FOR MARINE PURPOSES, AUTO-
MOBILES, ETC.
Fig. 251, representing on one-tenth scale a Hult marine engine giving 45 brake
horse-power at 10 atmospheres pressure, shows its convenient form ; the 200 brake
horse-power occupies over all a space 6 feet long, 4 feet high, and 3 feet wide. The
weight of the marine engines generally vary from 30 Ibs. to 45 Ibs. per brake horse-
power, according to power and steam pressure.
Engines for these purposes must be light and occupy small space ; they must be
simple in construction, easy to manage, and handy, so that all changes of speed and
direction can be executed easily and quickly.
In spite of all recent improvements, reciprocating engines do not fulfil these
Modern Engines
conditions as perfectly as Hult engines, which work directly on the screw shaft with
higher speed than can be obtained with reciprocating engines, while their small dimen-
sions are in every way convenient.
Being free from shocks and vibration, they can be mounted in ships and boats of
light construction ; as the engine shaft is coupled in direct prolongation of the screw
shaft, without cranks or fly-wheel, their centre of gravity can be placed much lower than
with reciprocating engines.
The description given above of stationary Hult engines also applies to the marine
engines, except in the arrangements for the distribution of steam, which in the latter
are modified to suit the requirements for reversion of direction and variation of speed.
The reversion is obtained by moving lengthwise the distributing tube for admitting
the steam on one side or the other of the sliding shutters.
Fig. 251 is a section of the Hult marine engine in the position of driving "forward"
at full speed. The distributing tube C, which does not participate in the rotary motion
of the main shaft A, is moved lengthwise by a lever D, connected by a rod E with a fixed
point P on the frame.
The steam introduced from the boiler into the box F passes through the orifices G
into the distributing tube C, pierced by the slots H, which by the longitudinal movement
of the tube can be brought in line with one or the other of the steam passages M and N,
as required for the desired direction of rotation.
As the tube C is closed at one end by the stop I, the steam can only reach the
cylinder through the slots H ; the exhaust steam escapes through the orifices K or L.
In the position shown in Fig. 251 the slot H opens fully into the steam passage M,
admitting steam for driving "forward"; the exhaust steam passes through the steam
passage N and (as soon as this passage reaches the orifice K) into the space outside
stop I, in the distributing tube ; it then passes through O into the space outside the
cylinder, and finally to the exhaust pipe R.
When the lever D is set for "back," the slots H come opposite the steam passage
N for admission of steam ; the other passage M serving in this case for the exhaust
each time it opens into the orifice L, cut in the distributing tube, which is there flattened
out for the passage of the exhaust steam.
This single lever thus ensures the perfect reversibility of the engine at all positions
of the piston, and the variation of speed depends upon how far the lever D is moved
from its central "stop " position in the direction of the " forward "or " back " positions.
The following table gives the approximate power, speed, and weight of the three-
shutter engines, as well as their consumption of dry steam per brake horse-power with
exhaust into open air.
TABLE XXVII.
142 Lbs. Pressure.
Brake
Horse-Power.
Consumption of Steam
per Brake Horse-power
and Hour, with Ex-
Number of
Revolutions
per Minute.
Weight of
Complete Engine.
haust into Open Air.
Lbs.
Cwt.
8
1500
ii
12
36
1300
4*
2O
33
1250
6
40
3 1
IOOO
12
DO
29i
850
17*
80
28
700
23
1 2O
27i
55
38
20O
27
500
,52
Sectional View.
Complete View.
FlG. 251. Hult Rotary Marine Engine.
215
2 1 6 Modern Engines
The above figures of consumption of steam without vacuum are, of course, reduced
when working with condensation, with the usual ratio of gain depending upon the
vacuum given by different types of condensers.
In comparing these figures with those of other engines it must not be overlooked
that the figures here given are for brake horse-power ; if stated for indicated horse-
power, as is sometimes done, the figures would be proportionately lower.
This chapter shows the best that has hitherto been done in the direction of the
Rotary Piston Engine. It may be observed that one objection still remains, and that is
the heavy strains in other directions than that upon the shutter, and not in the direction
of motion. The push is not direct in the line of motion, but constrained into that line by
the cylinder.
Printed by MORRISON & GIBB LIMITED, Edinburgh.
THIS BOOK IS DUE ON THE LAST DATE
STAMPED BELOW
AN INITIAL FINE OF 25 CENTS
WILL BE ASSESSED FOR FAILURE TO RETURN THIS BOOK
ON THE DATE DUE. THE PENALTY WILL INCREASE TO
50 CENTS ON THE FOURTH DAY AND TO $1.00 ON THE
SEVENTH DAY OVERDUE.
UCD LIBRARY
DUE J UN 3 01983
MAY 20 1983 MC'D
^ . x
Book Slip-20wt-5,'59CA2537s4)458
- .
%
\V^
?
J
y^
//x ''
3f^ >
.-> >
^7
^
J X
'
4%
*
187331
Call Number: IBi^M^BHB
KU ' 311751]
V.I i^^^BH
Kennedy, R.
Modern engines and
<*r* venerators.
2,\r>\rve dv
TJ25C 1
K4. K^
187331
/
1 V > '
c - v w
^^'
< S x ,^f
y V'V^'
\