Escolar Documentos
Profissional Documentos
Cultura Documentos
TURBINES,
& WATER
(SPECIALLY AMPTED
CHARLES
H.
MOTORS
FOR IHGINEERS)
INNES. M,A.
FOURTH EDITION.
Revised and Enlarged.
Price 4/6 net.
THE LIBRARY
OF
THE UNIVERSITY
OF CALIFORNIA
LOS ANGELES
GIFT OF
John S.Prell
WATER MOTORS
INCLUDING THE
JOHN
Goil
S.
PRELL
4s.
6D. NET.
1904,
Engincerag
Library
TJ
PREFACE.
I
HAVE endeavoured
in this little
work
to put in as concise a
form as
Com-
momentum
is
losses of
methods
have done
my
which
hope
will
be of
practical service.
I added a method of Turbine design, mainly
was indebted to Professor Rateau, whose work,
*
well deserves perusal, and is one of the
Traite des Turbo-Machines,"
graphical, for
"
which
made to Chapter
have added an extra chapter comprising descriptions of
Centrifugal Pumps for high lifts, and of Fans capable of producing
considerable pressures. There are several other minor additions, and
In
this,
XXVII., and
the
have
been corrected.
In the part on Turbines I have dealt with radial and axial flow types,
both of the reaction and impulse classes, the latter including the
Pelton wheel. In the chapters on Centrifugal Pumps I have considered
the Appold and Rankine types, and have
some drawings and descriptions of pumps
added to
this
edition
733389
designed by M. Rateau.
of the centrifugal
The theory
pump, and
of the fan
is
and
oil
the
a theory
not
and
a subject.
by means of copious
deemed desirable, and I hope that it
the text
clear,
will
and
how much
human race.
realise
of the
make
illustrations,
who
lies in
CHAS. H. INNES.
Rutherford College,
July, 1904.
CONTENTS.
CHAPTER
The Motion
Water
of
of
1
CHAPTER
Measurement
PAGE
I:
of the
Power
of a
Stream
CHAPTER
Form Assumed by
II.
III.
CHAPTER
IV.
15
Friction of Piping
CHAPTER V.
Losses of Energy from Sudden Changes of Velocity and Direction
19
CHAPTER VI.
24
Hydraulic Engines
CHAPTER
Theory
of the Hydraulic
VII.
28
Engine
CHAPTER VIII.
The Turbine
37
CHAPTER IX.
44
Classification of Turbines
CHAPTKR X.
54
CHAPTER XI.
Theory
of Radial-flow Reaction
Turbine
56
CHAPTER XII.
Theory of Axial-flow Reaction Turbine
.".'
64
VI.
PAOE
CHAPTER XIII.
fiQ
Construction of
the Vanes
of Turbines
CHAPTER XIV.
The Design
of Reaction
..
CHAPTER XV.
The Regulation
of Reaction Turbines
CHAPTER XVI.
Turbine Governors
CHAPTER XVII.
119
Theory
of
Impulse Turbines
CHAPTER XVIII.
Turbines at Assling, Carinthia....
CHAPTER XIX.
Turbines at
Theory
CHAPTER XX.
The Design
of
CHAPTER XXI.
Correction of the Vane Angles for Axial Turbines
The
"
Poncelet
"
l
Water-wheel
CHAPTER XXII.
The Pelton
or Tangential Water-wheel
...
CHAPTER XXIII.
CHAPTER XXV.
187
CHAPTER XXVI.
Theory
of the Centrifugal
Pump
...
192
CHAPTER XXVII.
Comparisons between Theory and Experiment of Centrifugal
Pumps 206
Pumps
Centrifugal
CONTENTS.
vii.
CHAPTER XXVIII.
PAGE
at Khatatbeh,
219
Egypt
CHAPTER XXIX.
The
Effect of the
Vane Angle
<f>
239
CHAPTER XXX.
On
Pump
245
CHAPTER XXXI.
The Balancing
of Centrifugal
Pumps
248
CHAPTER XXXII.
Method
of
Designing a Centrifugal
Pump
255
CHAPTER XXXIII.
The Fan
261
CHAPTBR XXXIV.
The Theory
of the
Fan
283
CHAPTER XXXV.
High-pressure Fans and
Pumps
289
CHAPTER XXXVI.
at Niagara
324
CHAPTER XXXVII.
Hydraulic Buffers
332
CENTRIFUGAL
PUMPS AND TURBINES.
CHAPTER
I.
J 2g =
/t
A body
8
falling a height h attains a velocity
h, and, neglecting
friction and cross motions, this is exactly what each particle
of water does.
The cause of the velocity in both cases is
2p
MOTION OF WATER
= wh foot-pounds on a weight
that eravity has done work
kinetic energy
of will, which must, if unresisted, produce
= E^l
foot-pounds,
whence
and
If in
CD
UNDER PRESSURE.
must be multiplied by the
whence,
if
coefficient of contraction,
Cc
per second,
orifices
sharp-edged orifices
C =
'582
to
'61
edged
.'__
_'..'
and
2-618
'CIS
when
1-618
A is
A-
oo,
energy
is
and so F
|-
|is
x F, where
is
wh =
pound
before the
= F^i
2<7
where A 1
come the
C,,
V2
~W~
but h - A 1
=
|1,
- A1
because A 1
is
n
Gv " h;
_,
^ ~_
? ,3
POWER OF A STREAM.
CHAPTER
II.
downwards
Fio. 6.
(figs.
5,
6,
FIG.
in the
the
7),
7.
sides
applied
For a rectangular notch
:
Q=
where
Q = cubic feet
per second
= breadth of notch
h = height of surface of still
6
If b
is
of the notch ;
a coefficient of discharge.
advisable, c
If b is the
'595.
'667.
POWER OF A STREAM.
57
fuTr'
-^
downwards.
Here
8
When b =
,,6 =
2 A, c
h, c
bh
=
=
"595,
-020,
Q=
Q =
2'54A.
5-3
Ai
CHAPTER
III.
Fm.
8.
energy
is
^-
and we
shall
is
per pound.
ENERGY OF WATER
'
or,
*
.
kinetic,
is
H!
is
and velocity
no
loss
by
H.
v.
friction, kc.,
of
the
above
Cv =
-j^xCrxCcXwx
=
-97
'64
3600,
x 8 v/l5 x 3 GOO
2500,
in round numbers.
2. A rectangular notch is 10ft.
broad, and the head of
water above the bottom of the notch is 5ft. The whole
weir is 40 ft. broad, and the total fall of the water is 15 ft.
Find the number of cubic feet per second, and the available
H.R
of the stream.
AND FALLING.
IN RISING
Q =
5' 35 c
=
=
=
H.P.
Jh
5 35 x -595 x 10 x 5
J~l^,
355,
62 5
= 625
Water
fall
in feet
550,
x 355 x 15
4-
550
605, nearly.
is
J!l
as half of this
is lost
by
( fi8 8)
Let V
trough
58
64
20
and
per second.
58j} ft.
to this velocity
friction,
H=
V2 =
.'.
The speed
at
64 (26 9 - 7}
V=
v/Iif-9
=
=
is
ft.
above the
64 x 199;
35'6.
is
such that
1-7
2?
ENERGY OF WATER
10
motion of the trough to the train enables the water to overcome the head of 7ft. The experiment made previous to
A stream of
constructing this apparatus was as follows
water was allowed to issue from a water main at the speed
of about 15 miles an hour, and a long
in. pipe was bent at
the bottom so as to face the current, and it was found that a
stream was maintained through the pipe until its top was
raised 7 ft. above the level of the water. A stream of 16
miles an hour could theoretically be maintained by a head
of 7 ft. ; hence the water could rise no higher.
:
A
lift has a ram 9 in. diameter and works
^ adirect-acting
under
constant head of 49| ft.,
allowing for weight of ram
and friction of mechanism.
Find the steady speed when
if ting a load of
1,350 lb., and also the load raised^ double
that speed, neglecting friction of
piping.
Let p
pressure per
square
W = load in pounds,
A=
area of
ram
foot,
in square feet,
3060,
near*.
IN RISING
Let
v
If
AND FALLING.
11
8 x
4.
62'5
A direct-acting
x 48
x 81
1330 Ib.
shown
in fig. 10.
is the pressure per square foot of area of the
the length of crank is 1 ft.,
a
if
of
piston
hydraulic engine,
and the number of revolutions 100 per minute, the piston
being at the middle of its stroke, and the obliquity of the
5.
lift is
What
Orfi
Let p
pressure required
P ~ rr "
Hl
626
27
60
1ft.,
and
N=
revolutions per
ENERGY OF WATER
12
.
g .fx.x^x
p
=
=
=
.coy (
_ h _ ) = 17,
62 5 (1730
171) = 62 5 x 1559,
97500 Ib. per square foot,
676 Ib. per square inch.
In the above examples the slide rule has been used for
calculations.
In the last
it
all
worked
shown in
lift
example
under a constant head. How this is obtained is
10 and 11, Mr. Ellington's lift.
The hydraulic ram which lifts the cage above is smaller
in diameter than usual, and its size is determined by the
strength required to carry the load, and not by the working
pressure of water available. The lift cylinder A is connected
to a cylinder B, beneath which is a cylinder C of larger
diameter. There is a piston in each, connected by a ram D,
Given a certain pressure of water, B is made of
fig. 11.
such a size that its full area enables the pressure above it to
balance within a very little the weight of ram and cage
when B is at the top and the ram is at the bottom of its
The annular area E, fig. 11, is sufficient to overcome
stroke.
friction and lift the net load, while the whole pressures of
E and B are communicated to the annular area J J in
hydraulic connection with the lift cylinder, the area J J
multiplied by the length of stroke giving a volume equal to
that displaced by the ram. With proper proportions the
force on the ram is the same at every point of the stroke,
because, roughly speaking, when the ram is at the top of its
stroke B and C are at the bottoms of their cylinders, and
consequently the pressure per square inch on each of them,
and therefore on the area J J, will be increased to such an
extent as to exactly compensate for the loss of pressure
caused by loss of head at the base of the ram.
The mode of action of the lift is as follows When the lift
has to rise, the rope shown by a dotted line passing round
two pulleys and up through the lift is pulled by the
attendant ; this admits water under pressure to C, while B
is always in communication with the accumulator.
Water
is then forced out from the annular area J J into the lift
figs.
and
IN RISING
AND FALLING
14
B 11 in., diameter of C
available pressure 33Hb., stroke of ram 50Ht., stroke
Useful load'lifced 8 cwt. Since "the water
8Jft.
delivered from below J J = that supplied to the lift cylinder,
the area of J J
Diameter of ram 3iin., diameter of
21f in
of
BC
50
square inches.
33i x
_^_
A= =
pi
pressure on J J per square
= pressure on ram per square
,
inch,
inch,
515
9621'
i.e.,
The additional
bottom of
its
by pressure on E E at the
obtained as follows It will be seen
ia io * ft beiow B
load, balanced
stroke
is
'
P2
33}
EE =
7854 x
(21|)>
- area of section
of
FRICTION OF PIPING.
rod
on J
15
J,
33
x 95
38 x 336
595
268
Ib.
59'5
efficiency
^_
work done
work done
8 cwt x 50
useful
total
mean pressure on E x
8 x 112 x 50|
36-25 x 336 x 8J
area of
x 8i
455
CHAPTER
IV.
FRICTION OF PIPING.
where
square
is the area in
v is the velocity in feet per second,
is a coefficient depending on the surface
feet, and
FBICTION OF PIPING.
16
of the plate
the
Experiment thus shows us that
is
friction
of the water. The most
is to the friction of pipes.
20'
This
pipe
is
where
z
= m*dlv = m lLl
d 2<J
-d 2 2g
'
=
62~5
X
62^5
'
~d
2g
=
I
12df
2g'
which gives
where
Z =
Z =
'005 for
new
cast-iron pipes,
FRICTION OF PIPING.
17
G2
">
is in inches, L in yards,
Professor Unwin's formula
where d
TTJ
77:
_ ml
G in gallons
per minute.
vn
II.,
section
5.
What
Example
F =
1.
x Z
62'5
A
2g
=
also
8000
lb.,
-M (i
4.
18-1
putting
1
12
^ x
d)
+ -^~)
12 d/
1
(
\
y2
~d
sTg
p _
=
3P
4,
Z =
-01,
ft.
5280, v
power
8000 x 4 x 60
33000
of the
Ibs.
pumps, we
x feet
33000
may
FRICTION OF PJPINO.
18
rrp _
or>
h x
per minute
33000
Ibs.
Axvx60x^d
x 62-5
33000
181 x 4 x CO x - x 3- x 62'6
33000
581, as before.
work
of
steam
'6
and
-d*
12
1\
.'.
I.H.P. of
Total head
100
pumping engine x
35 '6
135
efficiency
head x
per minute
33000
Ibs.
LOSSHS OF ENERGY.
CHAPTER
19
V.
WHENEVER
the water
Q = Av =
where
A!
VL
Q =
W = 62-5 Q
Change
of
62-5
Aj
vt
t.
momentum
62'5
A!
- vt )
Vj (v
g
all the forces acting on the water parallel to the
Let p be the pressure per square foot in the smaller
pipe, P! that in the larger, and let p be the pressure on the
annular space where the pipes meet, then
- v,)
/.
AU t = 62'5
i; t
- A)!
~ P A - p (Ai
- A, v, (v
\Pi AJL
impulse of
axis.
20
p)
Pl
62-5
Pi
1'
ff
62 '5
vv l -
TI
(1)
Let
pipe,
H=
H! =
first
Hl ~ 62o +
H-
H! =
loss of
2V
head
625
2
(t-
r,)
(2)
This
is,
Energy
In
figs.
14 and 15
= AC
B AC = 9.
changed to v x
the angle
AB
is
the velocity
?;,
which
is
suddenly
CHANGES OF VELOCITY.
Then the
loss of
=
where
-2t; 1
*'i
BD is perpendicular to AC.
Now,
for a given
change of direction
minimum when B C = B D
an important
that
BC
when
is,
vx
...
(3)
will clearly be a
=
= v cos 0,
AD
certain circumstances
;
AD)
as will
result,
change of direction
pressure
21
as before,
change of direction,
direction.
H=
and
*y
02-5
We may
/>,
625
2y
AB
AC
BC
'2g
2
BC
2.V
= A C. C D when A C
A r*
_ Al.
<AD
(fig.
TI
uu
when A C> A D
14)
r<
(fig. 15).
g
so that the gain may be found graphically.
- p, A C. C D must
Now, for a maximum value of p
maximum and AC<AD. Now, it is a well-known
be a
fact
FROM
LOSSES OF ENERGY
22
such as
obtained
when
CD
is
that the greatest product AC.
= AD. / 7^ - ?> is a maximum
=
AC, CD,
when AC
vv
t>
cos
CD
0,
and
;j
;/
t'"
2
COS 6
625
(J
results side
by
side
Example
(5)
(6)
1.
second if 1,000
a pipe 20 in. diameter suddenly enlarged to 30
~-2u.
2a
1000
"-
diameter.
say
7-65
^= 34
144
J^i'7854 x 30-
1000
x
60
(7'65
/i,
in.
3'4)
2g2
64
.
energy
lost
per second
This last
=
*
[iLL-Hl^,
4-*
and
i^
\ v
4.
Hence,
|L-80j -'^-80-6,
.-.
p,
-8031 Ib.
CHANGES OF
VELUCi'i
23
k.
AB =
BAC =
30 deg.,
AC =
Then
8.
is
= AC. CD
pl - p
625
'
cj
p = 20 x
and
*
.
if
144.
p = p + 625 AC.
CD
g x 144
=
By
calculation
BC 2
v*
Vj*
162
1-54
ft.
- 2vv L cose
82 - 2 x 16 x 8 x
64
'
= P
_u
f
62 5
'144
16,
16 x 8 x
- 82
-=
~32~-
HYDRAULIC ENGINES.
CHAPTER
VI.
HYDRAULIC ENGINES.
believe,
foremost
comes the
Brotherhood
three-cylinder
HYDRAULIC ENGINES.
25
FIG. 17.
any
and no flywheel
is
required.
when
the shaft
placed vertically, the engine being beneath the capstan.
The whole arrangement of capstan and engine can be
rotated on a horizontal axis, so as to bring the engine above
Its
is
to a capstan,
is
26
HYDRAULIC ENGINES.
on a cylindrical surface, about the axis of which the oscillation takes place. The bedplate has a corresponding hollow
HYDEAULIC ENGINES.
27
Thus (fig. 19) the piston is moving to the right, water under
pressure is entering (as shown by the arrow) from the central
port, while the exhaust is taking place through the right
On either side of the cylinder are short gudgeons or
port.
These have their
pins, whose axis is the axis of oscillation.
bearings in two levers (figs. 19, 21), pivoted at one end to
the frame, near the crank axle. By means of a hand wheel
and screw, acting on the other end of each lever, the cylinder
and by
can be pressed down upon its bearing surface
;
28
and
the revolution, but also of the water in the cylinder
vessel or accumulator
supply pipes, which, if there be no air
another
near the engine, will at one moment decrease and at
increase the effective pressure, causing at high speeds
it is the duty of the designer to
which
dangerous shocks,
We shall, however, treat of this more fully later
avoid.
on. The maker of these engines guarantees an efficiency of
80 per cent, and experiment shows that it rises as high as
90 per cent.
CHAPTER
VII.
when
10.
23.
3(j
by hydraulic
friction, &c., is
then
MN
drawn equal to
(1
a similar construction be
F)^,and
AKB
made
will be
reduction of
the
pressure in feet of water on the piston caused by
F must be obtained by
velocity and friction of the water.
caused
etc.,
frictional
losses
the
by
piping,
calculating
KB
between the engine and the accumulator. The curve
is a parabola, which may be readily seen by anyone who
knows the properties of this curve, and remembers that the
= v sin neglecting the obliquity
velocity of the piston
of the connecting rod, v being the velocity of the crank pin,
and
the angle between the line of stroke and the crank
centre line. But this is not the only alteration of head, for
the continually varying velocity of the piston makes it
necessary for the water in the pipes to increase and reduce
its velocity before and after mid- stroke respectively.
t>
W=
W =
Let
A=
2r
velocity of
But
.'.
'61
- at the end
Xo energy
at the beginning
middle
and
it is zero.
the beginning
is
W
A-gar
1
_
p
.Lc> -
v*
P2
62 5 a
where
Let
v*_
A.
l_
ga
I is
Aa =
then the
62'bga a
loss of
is
^ -.
r
g
T)
The
loss of
W
62 5
.-.
DE
total loss of
head
is
oJ'o
v2
gr
= t- inl +
gr\
A }
625 A/
CL
and
and
E G L,
DE GLC
EANKBL,EL
32
KB
- P ~ (1 +
f
where
F)
62 5
*9
= p -
64 (1
F)
62-5
^
*(/
33
We
Fio. 26.
motion that varies the mean pressure in this way. The two
remaining quantities capable of alteration are the number of
revolutions and the length of stroke, and it is this latter
that is varied in Mr. John Hastie's engine, described in the
Proceedings of the Mechanical Engineers in 1879.
Figs. 26, 27, 28, 29, 30, 31
as applied
Pro. 28.
36
The crank
as will fill three cylinders.
(figs. 26, 27, 29, 30), sliding in
pin I is attached to a block
back
At
the
of this
a radial groove in M, the crank disc.
block, revolving on pins projecting from it at either end, are
as
much water
Tilill
ill
Ilnrr7i|
runs loose on N; within this are two rollers, carried in forks
These rods, with the rollers,
at the ends of hollow rods.
are pressed outwards towards the circumference by helical
roller
over
each
while
passes a chain K, one end
springs T,
of which is attached to a snug W, projecting from the side
of the drum, the other to the hollow shaft N.
When water is admitted to the cylinder under pressure,
the crank disc begins to revolve, and with it the hollow shaft
36
is held fast by
the cam
N, while the spindle P carrying
chain
the resistance of the load, applied through the hoisting
result of this
The
barrel.
to the circumference of the chain
same
the
at
on
wound
time,
are
R
;
chains
the
up
is that
to the cam,
owing to the motion of the crank disc relatively
the
and
is
with the crank pin
pushed outwards,
the block
stroke is increased. The winding up of the chain compresses
the springs, and this compression, and the simultaneous
increase of the stroke, go on until the resistance of the
the latter is driven
springs balances that of the pulley, when
which lasts as
round, and a state of equilibrium established,
If the load
long as no change occurs in the load or pressure.
disc
the
crank
of
a
motion
is increased,
relatively to the cam
again takes place, until the turning moment on the crank
The
efficiency
427,
10,
"51,
U8ef ul
total
633, 745,
14,
10,
-54,
work
work
17,
969,
20,
-60,
'58,
857,
-5(i,
1081,
weight x
21,
'615,
lift
1193 (pounds).
22 gallons).
'65
x 6'25
THE TURBINE.
37
would be about
=
||
'87,
nearly.
CHAPTER
VIII.
THE TURBINE.
THE TURBINE.
38
vane.
/A
and Bu> 2
velocity,
Bw.2 u being
noted.
.'
and
because the
momentum
is
changed from
U
in time
by a force
P.
in time
on
the vane
P foot-pounds
-.(te,
a quantity independent of
u> 2 ),
t.
If
(-i
wa
.......
(7)
THE TURBINE.
39
C
c
v^
c
V COS a
Wi
- W t = Vi COS
= v sin (a +
sin
= v sin a
sin
w 2 = u cos P
- w 2 = v 2 cos
0)
c sin
v 2 sin 6
0-n(/3+0)f
= u sin p
)
(g)
and
- M! =
v cot
sin
o,
may now
circle
rota-
THE TVBBINK.
are sectional elevations ; the former is part of a cylindrical
vanes E and the
developed section taken through the guide
being to the left,
wheel vanes G, the direction of motion of
THE TURBINE.
41
THE TURBINE.
42
ment of the turbine shaft takes place, and this collar then
takes a bearing, and supports part of the weight so as to
The two rings
at H.
prevent there being too great friction
P Q, as before mentioned, regulate the supply of water.
six
spindles, which are
The outer ring P is suspended by
screwed at the top through nuts in the spur wheels R.
These are
together by a large centre wheel so that
geared
from a handthey can be all turned together by gearing
wheel (not shown in the figure), so that P can be raised or
lowered to adjust the opening for the passage of the water
to the outer ring. Q is raised by screwing up the outer ring
until it catches projecting brackets fixed to Q, which then
rises with P. As usual, the power is taken from the shaft by
means of a bevel wheel above H. It will be noticed that a
suction tube T is used, reaching about 4| ft. below the bottom
of the wheel, so as to allow for variations in the lower level
If this were not done, head would be lost
of the stream.
when the water is low in the river we must, however,
leave the explanation of the action of the suction tube to a
later page, merely mentioning here that as long as the lower
level of the water is above the bottom of the tube the
turbine may be placed at some height above the tail race,
Thus the
without any alteration of the effective head.
lengths of the shaft and of the six spindles may be less than
of the
reduction
of
the
would
otherwise
weight
be,
they
former lessening the loss by friction at H.
We stated above that the velocity of whirl at discharge
should be zero that is, that the water should flow out
Now u (tig. 32) is the
axially or, in this case, vertically.
total velocity of outflow, but u sin /3 is the component of the
that
the
water
of
the wheel. However
carries
out
velocity
great w 2 the other component, may be, no more water will
flow through a wheel of a given size so long as u sin ft is
fixed. Thus, for a given wheel and given quantity of water,
u sin
must have a fixed value but u must be as small as
;
,-i
possible, because
one of the
losses of
head
is
-,
the energy
hen
THE TURBINE.
43
if
'
62 5
g x 550
and the
maximum
hydraulic efficiency
?i
+ Qi = Q
cubic feet.
ewj
+Q CW
1
1 )
550(7
efficiency
CLASSIFICATION OF TUKBINES.
44
CHAPTER
IX.
CLASSIFICATION OF TURBINES,
turbines belong to one of two classes, called reaction
In the former, when working at full
power, all the guide and wheel passages are filled with
water, and the turbine is said to be drowned, and the
velocity of flow in one part can be determined when that in
any other part is known ; or, to put it mathematically, if
AU A 2 , A 3 be the cross-sections of the stream at any point,
and v lt v 2t v s the velocities perpendicular to those sections,
ALL
then v t A!
v2
A2 =
v3
A3
tt
CLASSIFICATION OF TURBINES.
45
H
H
the second we
first the water fl&ws outwards or inwards
have already described ; and in the third the water enters
leaves
and
The axes may
axially.
radially,
approximately
be in all cases vertical or horizontal. Fig. 36 shows a section
perpendicular to the axis, and through the guide apparatus
and wheel of a radial outward-flow or Fourneyron turbine.
At B are the guide vanes and at D the wheel vanes, the
arrow showing the direction of rotation.
Fig. 37 is a
;
CLASSIFICATION OF TURBINES
46
the tank or
type of wheel, A is
two
the supply cylinder, which consists of
lower slides within
concentric tubes the upper is fixed, the
lowered
or
raised
is
and
it like the inner tube of a telescope,
of the inner tube is a
by the rods b ; near the upper edge
it and the outer
leather collar to make the joint between
a
of the inner tube acts as
tube water-tight. The lower part
once.
at
orifices
the
all
for
a regulating sluice
as to round
to its internal surface woo Jen blocks, so shaped
the orifices
off the turns in the course of the water towards
formed
is
by a nxed
the
The bottom of
supply cylinder
lower end of
disc C, which is supported by hanging at the
carries
disc
a fixed vertical tube enclosing the shaft. This
v-rtical section of this
penstock
FiO. 38.
is small.
turbine.
is
CLASSIFICATION OF TUKB1NES.
47
I.
Number
of
11*81 in.
11*81 in.
16 14 in.
45 cast iron
14 84 in.
84 wrought iron
583
928
5 085 c. ft.
131 ft. 519
revolutions per
minute
Quantity of water per minute
Head of water
Available power per minute..
Power measured on brake ...
Efficiency per cent
4'52
129
c. ft.
ft.
36,440
16,498
45'27
16 in.
ft.-lb.
ffc.-lb.
53,806
28,809
in.
ft.-lb.
ft.-lb.
535
FIG. 41.
CLASSIFICATION OF TURBINES.
Fig. 42 is an outline drawing of an inward-flow turbine,
with suction tube and cylindrical sluice at the bottom ; the
lower part of the figure is a sectional plan through guide
apparatus and wheel. These two latter are marked A and
B.
is
is
Flo. 42.
entry
CLASSIFICATION OF TURBINES.
increased, so that
49
we have
of this type, is
shown
in
and 45.
being in section
The
first
shows
its sides
are
;
at O (fig. 45), and flows
round in both directions, and enters in the manner shown
by the arrows. It seems probable that had the casing
been made like that of the spiral chamber or volute of
a centrifugal pump it would have been better, because the
direction of motion here turns through almost 180 deg.
at the top of the wheel casing. The spiral casing is shown
"
by Weisbach in his Mechanics." Instead of the area being
greatest at O and least at the opposite end of the diameter,
it decreases from O uniformly round the circumference, the
water flowing in the direction of the hands of a clock, so
conical
5p
(fig. 44).
CLASSIFICATION OF TURBINES.
CLASSIFICATION OF TUEBIMES.
51
CLASSIFICATION OF TURBINES.
52
of
'
CLASSIFICATION OF TURBINES.
53
per second.
23-48
14-83
11-29
9'68
dynamo per
cent.
7500
58-65
4380
37'90
CHAPTER
X.
AK
= AC + CK
= KB + CK
-BC
is
55
ha
where A 3
34 - 5^1
20
Diameter of tube
in feet.
49
98
16
2-3
/is
3116
2952
27-88
27-88
33
49
2624
65
1476
1377
8'2
19-68
98
12-46
115
1115
13
9-84
THEORY OF
56
CHAPTER
XI.
Drop A ie it a perpendicular on
the tangential velocity of the water at
The angles o and should
O
O v^ O tv t be v, c lt v lt and w t
O c lt
then
O wt
is
Fio
Then
w =
Wt =
=
sin
=
sin
l
Cj
cx
<t>
<f>
50.
v cos
Vi COS
v sin (o
v sin a
B,
Bw
67
Then
W 2 = U COS
W 2 = V 2 COS
sin ^ = ?* sin (^ +
/3
C2
Co
<?
sin
= u
(13)
sin
FIG. 51.
water
of
velocities is
TTT
and
if
is
a right angle,
W WiCi, and
jS
fig.
tv 2
52 shows that
wz c
= O
c.2
'
.'.
the
cos
work done
is
ft
THEOKY OF
58
apply here
is
~-
efficiency.
The hydraulic
efficiency
is,
therefore,
is
because Cl Wl is the actual work done per pound, and
g
the work it would do if there were no waste of energy.
This efficiency is called hydraulic because it takta into
account hydraulic losses, but not friction of shafting. These
losses are caused by the friction of a supply pipe if there
is one, from friction and curvature of guide and wheel
passages, from leakage, and shock at entry into the wheel,
and, if a suction tube is used, from shock on leaving the
wheel and friction in the tube. The lower edge of the tube,
or sluice if there is one at its end, is another cause of waste,
which may be reduced by rounding the edge, and thus
lessening the sudden change of direction and velocity at
this point. In designing we must t ry to arrange that shock
shall not occur at entry and exit from the wheel, although
it will probably occur to a greater or less extent in practice
S3 at least experiments show.
Whenever in its flow the
stream encounters the edge of a vane a retardation takes
place, which may be lessened by making these edges thinner,
but which cannot be prevented. Like all fractional losses,
these are exceedingly variable, but it ia convenient for
purposes of design to give them values depending on certain
coefficients and velocities, the coefficients
being given
average values, deduced from experiment.
All the losses of head caused by the guide passages may
;
so that
|L,
F ia
a coefficient
In the following
we
The
by leakage and by
loss
friction of
59
represented by
is
F2
because
is
v.2
there
is
is
is
+ F3
-^- (1
if
the coefficient
),
L = 125 i +
and
Let
L=
Q=
-2
^+
~^1 + -2^ +
-125
(15)
(16)
|s!
without one
(17)
so that a, a 2 a s are the cross-sections of the stream at discharge from guide passages, wheel passages, and suction
tube.
,
The hydraulic
because ^-^-
is
efficiency
77
lb.,
and -^
is
18 )
+L=H
(19)
expended in useful
losses.
2
2
?>1
ff
work and
__^I-^L__
c2
v cos
C2
a.
V COS a
'125 v"
'2 v 2
2 v32
also
H = 2 ^ c 2 v cos a
'125
+ 125
if*
-2
v 22
^Y +
'2
2 v<?
'^
THEORY OF
60
-n;
,.,,2
2 LI
!J_
^1 COS
C03 a
COS 6 COS a
f 125 ("?)
V
/
rz a
'2
-7TT"
<
'-
r.2
a2
'J. 'L?
sin
r.,
0,
cos a
cos
'
cos^cosa+
21 >
+ 2(^)
^"s/
(20)
'125 ("-?)''
\a/
"2
sin a 6
on
We
^2
a
LL
H5
1-169;
'
from
from
^2
15
12
025.
<*3
>'2
2Li Is COB
i'z ^
COB a
(21)
(20)
=
2
2 x
1169 x 115 x
-966
_ .i^m
/5tH
16 4
ft.
57
JtJW.
per second.
x '978
-84
254,
near*.
The
velocity of
c2
Vf cos e
Q =
a 2 v2
10 4 x
c2,
Cl
r
-
c2
The parallelograms
construction, but
cx
= 1169
a v^ in
x 15-8
18-85
18
6.
50 will give
fig.
(13)
61
and from
= 15 8
066
= -LH =
a,
<t>
by graphic
<
tan $
(c l
tan
= _JL_8!5_a___ and
a
v cos a)
v sin a
a
12
W COS
Ci
18 85 x j208
18 6 - 18 85~ x~ '978
=
Let
19-6
<p
87
5'.
Kb
12
TT
r,
sin a -
nt -
n,
t,
K6 2
J2irr jsinej
1 <2
=
{
?HL? }
sin^J
......
a2
(24)
(25)
We
^^
in (24)
the quantity n^
s
t:
-^- might
sin
be omitted.
<j>
= n +
1,
or n
2.
To obviate
this
THEORY OF
g2
We
=
=
rz
whence
>\
ft.,
4 x 1169
4 676 ft
"25 in.
iron
Aa a general
iron.
and i in.
=
Applying (24)-
96
~~i:
40
and from
6,
1-255
41.
599
41 x -25 x
40 x -187
x 4 076 x '208
6
n,
12
>91>
ft.
(25)
- =
89
27rx4x-258-
1-355 ft
l^-^ )x
'9
a3
= Jl =689
*2o
x 4
^756,
a3
6 86,
and
- ?
a3
1, very nearly,
The head
lost
by shock
63
will be
<-*>-(- -5)'
In addition to this there will be a
loss
2y,
02
vS
'
:
2g
2g
,\ total loss after
is
045
v^
2g
Hence
.254
2 54
'165
2'54
'2
2 95
'045
be absolutely fixed.
To summarise the above method of calculating a numerical
example, the quantities given are Q and H, and we assume
2
)
fig. 50,
v ain
c2
or from
from
(13),
then
THEORY OF
(J4
r 2 or TI
may now
and
t,
t lt
tz
-^i
the
_L
rg
W =
I.
-83
-91
S^H
v/
(2GA)
i COS a COS
a;' ....
-^
=
generally b
We
>
shall
Assume
(24A)
20,
17, -^
An axial-flow
= H, and 212
In practice,
-97.
varies
=
=
=
=
<t>
cos a
sin a
v2
tan
'936, cos e
342, sin
19
v 2 cos0
az
18 6
vsina
- V COS
^-
18'35,
6 36
=
=-^|
loo
'97
"956,
'292,
from (26A)
^2 v
2,
=
=
11'4
<f>
x 81 4,
11 4 square feet,
11 05 square feet.
= 4-925,
If r
a
V
2
TT
r sin a
1 08ft.
108ft.
if
r= 4925 ft
-
6p
1 225 ft.
THEORY OF
66
to
feet,
r varies from
1*5
Ja
it lies
(25A),
= 115
very nearly
1 35.
The depth
Pio.
quantity of
Flo. 54.
53.
is
needed
after this
67
AB is the
<f>
z>
inner to the outer radius, because for simplicity of construction it is made BO that a radial line perpendicular to the
axis will touch the vane from the inner to the outer radius,
so that the vane forms a helical surface of variable pitch.
When a wheel is cast the spaces between the vanes are
cored, each with a separate core, so that there ought to be
very little difficulty in giving the vanes the correct angles,
In the wheel designed
if this is theoretically possible.
above it will be well to see how far theory and practice
differ from one another with regard to these angles.
Let
a subscript 1 refer to dimensions at the outer radius, and
inner.
2
at
the
Then
in
55
to
dimensions
Or
fig.
subscript
is the mean radius 4'925, O F and O E are 5 '6 and 4'25, the
outer and inner radii at the bottom of the wheel, B O r is
Fio. 55.
made equal
to 9.
Then practice would make B1
If the arc
be struck with radius
= C O E.
GH
GE and H F be joined, then
02
as
G E the greater is
Mo
GE
Br
O B,
and
U^
HT
work done
= D O F,
^-
pound
much from
68
'
2 c
which are
less
than
O m
n O
mp
<t>
much
values of 02 0n which in practice would not ditter
from O rm. It appears, then, that there would be a saving
Fio. 66.
of
2
oo, * 1
and
cos
<L
vz
must be
---!-
.?
r
less
4925
v2
1^2 =
1085
19' 2
than unity.
(J
34
25'.
it would
would be worth while to correct
make very little difference whether & was corrected or not,
as is shown by fig. 55.
We have assumed that, although a and vary, v and conare constant throughout the whole width of
sequently
Although
it
<f> t
>:,,
69
H=
-1L +
G2 o
neglecting friction
(j
hence
and
CHAPTER
XIII.
THE
AB (fig.
CAB
BCD
CONSTRUCTION OF THE
70
the wheel vanes (fig. 58), take A B equal to the pitch, and
perpendicular to AC.
make
equal to 6; draw
A point D has now to he found as the centre of the arc C E,
F
E
be
F
E
G
being tangent at E.
*,
may
so that the angle
BCD
BAG
G H L = 0, and make L H K - a
and draw C E at right angles to
CK
D may now most readily be
the
centre
and
this bisector,
found by trial, or by drawing a perpendicular through the
Take any
HL
so that.
line
bisect
H,
;
right angle
centre of C E.
Absolute path of a particle of water through the wheel. It is
advisable to draw this in order to assure one's self that there
are no rapid bends in the path. If there are any sharp
Fie. 59.
either
Let
draw
when
calculation.
(fig. 69) be the section of a vane at the mean radius ;
V Y Ac, the parallelogram of velocity at entry.
Then
the particle of water has moved through vertical
distances
the wheel has moved through
k, On,
8,
horizontal distances f<f,lm,tv; hence k has moved to ?, ;>to
VANES OF TURBINES.
71
Bu =
kq =fg,pr =
I m, and
tv.
ru is
B to M, if
therefore the path of the water. Of course we need not
limit ourselves to three points. If the axial velocity of the
water is not constant, we may proceed as follows : Let us
suppose two sections taken through the wheel at right
angles to the axis, the one where the water enters and the
other where the water leaves the wheel. Let the first area
r,
= B
square
feet,
a
~~~*
T>
B2
Oq
square feet
~_
Then
Let h
x.
depth of wheel
so that
'
.
vertical
Let
component
time from
then
-~
of the velocity
to B!.
where
B v sin
= Bv
sin
a.
During time
t,
CONSTRUCTION OF THE
72
Oqrv
of the
Fio. 60.
AD
<f>.
VANES OF TURBINES.
73
AK
AL
B+
B),
74
B =
T>
r>o
and
aa
sin a
a*
sin
where
<f>
is
stream.
Then, E,
=B+
k x, where
is
v sin
Integrating,
^=
get
The
radial
v sin a
B + kx
we
Tc
where
= B v sin a.
I tt
BJC+
B
<
B2 ~
r sin a
B) x*
'
75
CHAPTER
XIV.
is
taken from Monsieur A. Rateau's "Trait6 des TurboMachines," page 99, et seq., which o'riginally appeared in the
Revue de Mecanique from 1897 to 1900. In this type the
pressure at discharge from the guide passages is greater
is the total head
than that of the atmosphere, and if
under which the turbine works, and v is the velocity of
discharge from the guide passages, while p : and p 2 are the
pressures p?r square foot at inflow to and outflow from the
is
<">
the liquid.
where D is the weight
This degree of reaction may vary considerably, and for
The
axial, radial, and mixed flow is generally taken as |.
reason for this is that the relative velocity of the water at
of 1 cubic foot of
76
inflow to the wheel, and the loss by shock on the vane ends
due to this, is then a minimum the losses by friction in the
wheel are very nearly constant, although e varies so that by
;
making
e
........
(2a)
in impulse turbines it
is very nearly 07, while
then
of
can be shown that is very nearly 0'5. At values of
075 for reaction and 045 for impulse turbines, the
mechanical f-fficiency is still very little less than its
maximum. By varying the degree of reaction we can also
If we wish to pass
between these two values.
vary
these limits, as for example in driving dynamos when the
fall is low and a high peripheral speed is desired, we must
sacrifice mechanical efficiency.
Leakage. This is a fault that must be taken account of in
reaction turbines, which is caused by the space which must
be left between the guide and moving wheels. This, in
axial turbines, cannot be lower than 3 millimetres, or 012 in
and may amount to 5 millimetres, or 02 in., while in radial
wheels it is as small as 1 or 3 millimetres, or 04 in. to 012.
The quantity of water escaping
,
less
Qi
where Q!
(4a)
which,
is
if t is 0'5,
becomes
LOSSES BY SHOCK
AND
FRICTION.
77
This loss is rarely less than 2 per cent, and is on the average
about 4 per cent. To reduce it to a minimum, arrangement
must be made to take up the wear of the shaft bearing,
besides such contrivances as grooves in the cylindrical or
plane surfaces where the wheels meet, which, causing
sudden enlargements and subsequent contractions, reduce
the effective head which causes leakage. In axial turbines,
divided into a number of rings which work with different
degrees of reaction, there is also obviously an internal as
well as external leakage which reduces efficiency.
The
loss of
fig.
64.
loss of
head of
06
LOSSES BY SHOCK
78
AND
FKICTION.
AX
= AB, and
AC is e
lt
CD
'
2</
where
f is the loss of
27'
Replacing
;>i
by
H, or
f
and
c^
c2
by
c^
mean
2
(cr
v*_
'
1)
where
2g'
radii of inflow
unity),
LOSSES BY SHOCK
we
AND
FRICTION.
79
obtain
*2 2
Vl
+ _!!_+
20fby
jl
fr*
i) Cl
2
}
V -Ml -
2fl rf
we may now
stated above,
(0-88)
(0-97)
7a)
replace
v2 2,
8
(1-03 w a )
which in
(0-88 Wi)
this case,
with
(I'03v 2 )
+
<r
(0-88
ji^L
1, e
vj 8 +
2
(<r
0'5,
f2
l) Cl
(8a)
becomes
......
(9a)
AD
AN
ABC
0'66
The
A D
* In
more
fig.
65
clearly.
Figs. 66
80
equal to
v.
We can
e.
experience.
a.
i.JS.
Fio
67.
deg.
<t>
The
where
and
passages,
reaction.
17,
with an increase of
when we take
we
is
given by
The hydraulic
which decreases as 1
From figs. 65, 66, and
similar manner,
e.
81
82
e.
BEACTION TUBBINES.
83
Falls, are
chapter, and
outflow.
The
84
Wheel
velocity at inflow
Cl
6872.
64-312.
Wheel
22'5.
velocity at outflow
c2
81-813.
19.
v2
76-906.
vt
<
19 deg. 6 mins.
and a tangent
= no deg.
40 mins.
13 deg. 17
v 2 at outflow
mins.
of a 2
and
a,
az
a
=
=
6-02,
7-45,
and
31i,
REACTION TURBINES.
at full
85
Head
H=
of fall
41 "45 metres
135'8ft.
5,600
chevaux
=
From
this
Ci
Now,
if
250.
we deduce
that
68 5
we assume
radial,
T)
and
g H =
-*>
<a
73
JVgH.
c2
v z cos 6
of contraction the
assuming a coefficient
and 2 whence
is
C t V COS a,
>
v 2 cos a cos 6
?-
But
rz
= 2<r/H(l-e)
v2
so that
___ ____
1 _
while
--
2 cos
differs
2 cos a cos 9
is
but
little
Ct
119 x
az
approximately equal to J T
i
1
_
2
.
'
cos 9
81
and
v
= Jvg
52 very nearly,
x 136-8 x
'48
64-8,
Falls.
1895,
ExNo.
8.
CENTRIPETAL OR
86
The
be taken as
Vl
= 2245
we
value
vz
= 81-3 ft,
is
u =
w2 =
=
v>^
and as
we have a hydraulic
18 7
ft.,
295ft.,
v cos a = Cl,
efficiency
0-90.
have
108
A very fine
87
They are
H=
Head
of fall
Cubic feet of
Horse power
Ee volutions per minute
The diameter
less
head
Low
tail race.
426
32 4
1184
120
to
to
High
tail race.
361ft.
37'
1184
120
of the
0'85.
Q=
P =
n =
it
CENTRIPETAL OK
Fio.
89
Fio. 73.
small, so that
small period.
when
very
no
50,600
oscillated
(103v 2 )
where
=
1
(8a)
(H
Fig. 76 refers to
inward
flow.
The degree
of reaction is 05,
or
v
a in
706
VT^H
Fio. 76.
We now vary c
=
and
AG
A Co corresponds to
the outward flow and 705 for the inward,
or
'62 for
so that
to
AG
AC
B, or
B.
_
The
Cj
W, - Co
W2
is
92
Ad
D1
In fig. 76 the triangles of velocities of discharge
and A Gj. H 1 are also shown, the angle e being 13i deg. The
1
1
By setting out a number
corresponding hyperbola is J 1
.
maximum
efficiency.
AF
to A X
If
we can arrive at
made equal to-
is
that giving
l
A B, F E
1
7*2
are
fig.
78,
efficiency,
and
1*2.
when
is
93
to 22
and
unity.
44, with the addition of an arrangement to transform half
the kinetic energy of discharge into pressure head. This is
usually a discharge pipe whose diameter gradually increases.
The curve 55 rises to 91 per cent, and subtracting 6 per cent
for external losses, we get a net efficiency of 85 per cent.
The mechanical efficiency of inward flow turbines is still
full,
ratio
^ of
of discharge
from guide and wheel passages T22 angle
head 10 81 ft. Similar
from the guide passages 34| deg.
;
ft.,
2 23,
1
are curves of mechanical or net
1
efficiency ; the curves B, B are curves of discharge, while
1
1
C, C are curves of torque on the shaft. The ordinates B, B
give the values of
94
of C,
the weight of a
where T is the torque in foot-pounds and
and
cubic foot of liquid, here, of course, G2'5 Ib. Both
are independent of the units employed, and are mere
numbers. Fig. 80 gives curves in full lines for an inward
<f>
"
2r lf
2r 2
Francis turbine
9'35,
7'92,
Humphrey turbine.
0-85,
10,
15
1342,
a*
a
104
12'45,
30
Curve
gives values of <f> for the Boot turbine at full
gate, and the fall of the curve may be compared with its
increase for the outward flow wheels in fig. 79. The highest
efficiencies, 80 and 82 per cent, correspond to a value of
equal
to 70.
The limiting velocities at which no work is done
are, in this case, less than double those for maximum
95
line.
CHAPTER XV.
THE REGULATION OF REACTION TURBINES.
necessary when the power required from a
IT is
turbine
decreases to lessen the quantity of water passing through
in
this
be
done
two
may
it, and, speaking generally,
ways,
by hand or by a governor of the indirect-acting type, about
which we shall speak on a future page. Methods of regulation may also be divided again into two classes
firstly,
those that close some of the guide passages completely, and
secondly, those that close all the guide passages equally
the former are usually economical, and the latter reduce the
efficiency considerably the more the sluice or gate is closed.
An exception to this statement is found when the turbine
wheel is divided by partitions at right angles to the axis in
a radial-flow turbine, when the efficiency is but slightly
reduced, as we have already shown in the case of the Hercules
turbine, in which an efficiency of 71 per cent was obtained
when the sluice opening was 379 of the whole, while the
efficiency was 84 per cent at fall gate ; this is a very fair
result. The method of the late Professor Thompson of altering the inclination of the guide vanes, and thus decreasing
the section of the guide passages, has also given excellent
results, although the direction of flow is suddenly changed
on entry to the wheel. This is probably due to the fact
that the alteration is not sufficient to seriously affect the
:
efficiency.
96
THE REGULATION OF
REACTION TURBINES.
97
Fio. 82.
of loss of efficiency.
When
of energy.
8p
THE REGULATION OF
decreased, a throttle
valve in the suction pipe can immediately lessen the flow.
for
This may also be used
rapidly starting or stopping the
wheel This type of wheel is suitable for falls up to about
If the
The work done per pound of water and the efliciency are
reduced in consequence. The resistance of a throttle valve
alone in a stream would set up eddies, and thereby waste
power. It is sometimes, however, used in the head race or
suction tube.
REACTION TUEBIXES.
99
There
__
-f
\
cam, which
strips
THE REGULATION OF
100
were open.
We have already spoken of the sub-division of axial and
radial turbines
and for the regulation by a sluice at the
bottom of the suction tube, or between guide passages and
wheel, which latter is moved parallel to the axis, we must
refer the reader to a former page.*
It will be remembered that we gave a
description of the
Victor mixed-flow turbine, but that we were unable to
give
;
See also
fig. 95,
REACTION TURBINES.
101
water flows freely to every part of the wheel, and this sluice
therefore superior to that shown in fig. 48 (page 53). It
bored out to receive the wheel, and turned to fit into the
outer case, within which it revolves and is moved, for the
purpose of admitting and shutting off water by means of the
segment at the top of fig. 88, and a pinion which gears with
it
We have explained above that it is impossible for this
or any other arrangement that throttles the flow to give a
good efficiency at part gate, and the experiments at Keswick,
above mentioned, show this, although at fall gate this is an
excellent turbine, and, according to the experiments at the
Holyoke testing flume, has given efficiencies between '8289
and '896, with heads from 1165ft. to 1834ft At part gate
there is a sudden contraction at the sluice, and a subsequent
is
is
102
THE REGULATION OP
FlO 83;
Fio. 89.
two
REACTION TURBINES.
103
the coupling from the main shaft. Fig. 90 shows the wheel
complete, and fig. 91 a pair of wheels
which latter we shall describe later.
REACTION TURBINES.
105
curve D.
Although this method throttles the flow in
each passage, it gives a fair efficiency as shown by the
following table for a 60
Gate opening.
1000
548
297
Head.
in.
Cub.
...
1656
..
...
17
,.
...
17-53
ft.
per sec.
64-88
5092
3453
..
..
Horse power.
10218
6968
3676
Efficiency.
...
...
8401
7110
5364
Fig. 93, from Kankine's "Steam Engine," shows an oldfashioned arrangement, which is somewhat similar in
principle to the above, but which wag not efficient, and
has therefore given way to some of the better plans
mentioned above.
A very efficient form of reaction turbine is shown in
figs. 94 and 95, the Hercules-Progres, constructed by MM.
Singrun Freres, of which 1,800 have already been constructed,
and with which efficiencies of over 80 per cent have been
obtained. The wheel is of the inward mixed flow type, and
has a cast-iron disc, to which steel vanes are fixed. This is
shown in fig. 94, with shaft and coupling, and one vane,
lying in front detached from the wheel, so as to show the
manner in which it is fixed thereto. It should also be
106
THE REGULATION OF
Pro. 94.
desired, in
REACTION TURBINES.
107
108
THE REGULATION OF
BEACTION TURBINES.
109
REACTION TURBINES.
110
governor.
The
sluice
and
falls.
TURBINE GOVERNORS.
CHAPTER
Ill
XVI.
TURBINE GOVERNORS.
that
is,
FIG. 99.
to the Victor
governor as applied
the
The first, fig. 99, is a perspective view, and
turbine
the principle
second? fig- 100, u an outline drawing showing vertical and
two
seen
be
shafts,
will
In fig. 99
of its action.
112
IUKB1NE GOVERNORS.
TURBINE GOVERNORS.
The
sector, or
pawl
shifter,
fig.
100, part of
113
which
is
just
in its action.
Hett's governor
9p
TURBINE GOVERNORS.
114
Fio. 101.
TURBINE GOVERNORS.
115
Fio. 102.
MM
A. Rateau.
TURBINE GOVERNORS.
116
and
104.
piston
acts on a differential
to a lever G,
joined at the centre
AB, which
is
TURBINE GOVERNORS.
differential piston
to the left. The
moves to the
right,
former motion
117
closes,
the
sluice.
moment a motion
118
TURBINE GOVERNORS.
C and the
midway between
orifices
may
be such as
is
desirable.
11!)
Generally,
motion
AB
piston
AB
which
carries at
CHAPTER
XVII.
THE two
with
at
full admission
full power and
The
use again.
come in/
/
120
We
Work done
per pound
is axial,
If
'23
and
(27)
121
should expect
but in practice we find
about
k Ja,
(28)
|
also
alteration of
r.
CWi_C V COt a
^.r ~ir"
Fig. 106
'
vt
Oc =
2v
O c.
a.
of
122
c sec
2 COS
Fo may be taken as
a.
(30 >
be more evident
will
--
(29)
COS
'1.
<f>
123
and
less
v.j,
Aj.
may
= H-
still
8-5
Ja =
be 1
ft.,
4'C6 h.2
-94
r and
/* 2
may
21'9
4'2 nearly,
'9
86
x 8 ^/ITC
22,
8 '^'
~22
be
Jh t =
'9 ft.
x 8
'94
J&72 =
142
and
left as above.
also
and F 3
vj.
lies
between
COS
Assume
'05
= *V
and
i v sec a >
'1,
and we
shall take it as
a.
j
2
N/Jw sec
/'.
r~
+7^2g h 2
'1.
rl
sec
(30)
'805
nearly
26
deg. nearly.
124
v
v
coa
COS 6
v sin
'u
i~
& y
H"^
i"T"r\
;(^T)' +S "*'
For
/Ql\
r,
<P
<P
=
=
129 deg.
6
6
=
=
=
'
13 deg.
18 deg.
23 deg.
so that 4> may lie between 128 deg. and 129 deg.,
shall take
18 deg. and 23 deg.
and
between
We
23 deg.,
<t>
129 deg.
<t>
<j>
<j>
The hydraulic
efficiency is q
~-
<
(/
=
ij
H am
=
<P
XI
x '438 x 906
32'2 x J) 5 x 777
22
ftu
The
(30o >
and
125
--
Then
b
Let
'2 in.
TT
-'-ft.
4'2
....
(24A)
n = 66
- 62 x 4 x
'4226 - 66 x
12
12
bL
a2
= K2
^^
'7771
where
K2
lies
between
1'3
and
2.
filled at
where
=.
TT
r sin
176, v 2
sin 6 = -3907,
b 2 = 27 in.
- n
sn
and
- n
1
let
2 cos a cos
K2 =
T41
'92
;
'906
last
equation
126
CHAPTER
TURBIXES AT
XVIII.
CARINTHIA.
A.SSLIN'C,
AT
Turbine
Hesd.
No
76 ft
No. 2
No. 3
796ft
...
84
ft
Cub. ft.
per sec.
118
121
124
Revolutions
per minute.
H.P.
134
137
140
772
827
894
...
...
..
Fio. 107.
127
128
129
FIG. 110.
J 2 which
130
Fio. 111.
3'59 tons axial pressure of the bevel wheels, '4 ton ; weight
of shaft, &c., 13-53 tons.
In order to reduce the friction, two
annular grooves are turned on the face of C, and oil is
;
131
catches
and not merely by oiled bronze. The cup
the overflow, which is cleaned and used over again. In fig. 114,
oil film,
132
Fio. 114.
133
CHAPTER
XIX.
WHEN the supply of water becomes very small, it is preferable to use a turbine in which there are very few guide
passages, so that only partial admission takes place, and such
turbines are generally of the radial outward-flow type, with
horizontal axes, so as to drive direct or by belt, the water
entering the wheel near the lowest point of the circumference, and leaving it as near that point as possible, so as to
waste as little head as possible.
The construction for is the same as before, but only as a
first approximation ; so that
<t>
and
cx
where
7i x
=
=
=
7T
Vi
20,
v sec
a.
'94
-8?.h2
hj.
H.
The
may be assumed as 1 ft., as before, but, of
course, experience will be our best guide. It should be
noted that h 2 does not necessarily equal r 2 - rlt but should
not differ much from it.
The outflow ought to be radial, but calculations from
existing turbines show that generally it is not so ; but
value of
h2
02
= Cja +
= #i2 -
Vl
Cj
+
+
Cl (u'i
2 Ci w\
- d)
THEORY OF
134
where
is
the wheel.
2
'
fi
'.
'.
(1
+ Fa
ci
a
a
a
) t- 2
If the outflow
Ci
+
a
were
Vi t hi
2 2c2 w2
- 2 c2 2
c2
>
2g
7/ 2
2 g h*
.....
(32)
'1.
cos 9
F2 =
3
t' 2
t>i
assuming
ci u>!
= F2 r32 + 2 ci wi
= r - er +
+
us
1.
But
if
and
tan
<p
We
<t>,
135
shall,
assuming
the term
c2
second,
V COS a
Let
178 x -956
=
fern
841
<j>
G!
v sin a
V COS a
178 x -2923
- 1701
84-1
= *
being negative,
Let
'
.
<f>
rj_
r2
is
=
=
=
Ci
97-25
From
'605,
sin
'517o
<
(32),
n 2
(97 25)
- (841) 2
11
v
= 167
(100-5)
64'4
THEORY OF
136
For radial
flow,
-91
1067
6
24 deg.
31 min.
<f>
<f>
we shall
it,
v - 2
a,
possible.
Suppose
148 deg.
y sin (a
sin
,-<;=
<t>)
<f>
v sin 15 deg.
sin 32 deg.
gg. 9
100-5
?i
v sin a
v sin 17 desr.
sin32deg.
and from
(32),
cos
105 6
*L
we obtain
-9525
tfc
0-17
deg. 44 min.
up by them
RADIAL-FLOW TURBINES.
b
sin a
7
-0536
"031
'06475
-46
4O
x 2923
X
ZtJ^O
/ir4.
616ft<
The
is
_.
f~
2^7,
137
"36
nrm "_
-616
Q24q
Then
sin
It is best to choose a probable value for 6 2)
K2
and calculate
therefrom.
b2
Then a 2
a2
Let
1-31 ft
1\31 \2*
7 '335
v _/^22 _
=
x 105'6 x 7'335
2-27.
At the
They
all
per minute.
into
two principal
THEOBY OF
138
groups
(1)
RADIAL-FLOW TURBINES.
139
140
THEORY OF
Fio. 117.
BADIAL-FLOW TUKBINES.
141
(fig.
sectors, is
flow of water
the pipes.
142
turbine
Zuppinger's tangent wheel (tig. 121) is an impulse
of the inward-flow type, which we introduce merely to show
that this type has been tried and discarded in favour of the
outward-flow class,* when partial admission is necessary.
This latter has the advantage that its axis may be horizontal,
so that it may drive direct or by belt, and the only head lost
is the clearance of the lower circumference of the wheel
above the tail race. Without a suction tube, this would be
impossible with any other type of turbine.
*
The inward-flow type with horizontal axis has again come into use and
is
143
CHAPTER XX.
THE DESIGN
and
is
whereto
,.
and
Q = K.a.v.
where a is the section of the orifice of the guide passages
and K is the coefficient of discharge, which depends on the
From
THE DESIGN OP
144
and
80 and 0'90.
lies between
CASE 1. The Pdton Wheel* The simplest case is that of
the Pelton wheel, in which a is zero, and
is nearly v.
The
triangles of velocity are then reduced to a simple straight
= ir and if c t is the mean velocity of
line if we assume
the bucket,
then X
<f>
<j>
=
u =
^i
and we
shall
assume
v2
.
It is easily
c ly WL = v
- v 2t w 2 = u
v
ci
0'85 wj.
seen that
Wj,
W2 =
(1
X) (V
Cj),
efficiency
_(! +
*)(
-c
~TH
)e i
'
See Chapter
145
where r
is
that
when
X is
85,
0925
guide passages,
n
little less
than -
loss of
head in the
0-89.
efficiency
would be obtained
at a speed
We
AB
'a-
AC
<t>
HP
THE DESIGN OP
146
CAB
CB
AH
A C is a
AD
AG
DH
DH
little less
than \
The hydraulic
9.
efficiency
.7H.
It is zero
when
ct
o,
increases
to a
maximum
To
obtain
follows
the
maximum
Suppose D, C,
efficiency.
Draw the
AN
fig.
tangent
c^
is
as c t
= AE. A
a parabola.
proceed as
value of rj we
124, are points of maximum
the hyperbola to E B.
D N of
in
then for maximum efficiency
C cut
is
the middle point of
N. Thus, by drawing a number of
these triangles of velocity, the best value for
C may be
Let
found.
AC
is
<f>
34?
We
becomes considerable
energy
for 6
25 deg.,
it
does not
is less
is
cos
a.
w2
It equals 0155 v
decreases as
increases.
when a = 0, with X = 85, and = 25 deg., but decreases
to 0'044 v when a is 45 deg.
Fourthly, u increases with a, but very little.
Fifthly, i] decreases as a increases ; very little between
and 20 deg., but more after that.
values of
From fig. 124 we deduce that
Thirdly,
_2AC
GE^
AB
v*
*
Professor Rateau considers that the backs of the vanes should be tangents to
the relative velocity x at inflow (see page 30, " Turbo-Machines ").
THE DESIGN OF
148
With a
loss of
ti
85.
air friction,
we
x 0-96
0-83,
0-80.
efficiency
083 -
0-03
Fio. 125.
If
we take
as abscissa,
and the
29
as ordinate, we shall evidently get a straight line parallel to
the axis of , because 4> is evidently constant for a given
The curve of
sluice opening.
e
IMPULSE TURBINES.
149
where
is the torque on the shaft in foot-pounds, D the
density of the liquid, r the mean radius of the wheel, and
the head, is a straight line such as A B, fig. 125, cutting the
axis of at a point A where is unity, or very nearly so, if
a is small-, and somewhat more if this angle is greater than
25 deg.
This line also represents the moment
to a
different scale. It is easy to see from this that the net
is
curve
axis
a
P
whose
is
efiiciency
perpendicular
parabola
to that of
and which cuts the axis of in A. Its highest
point corresponds to = i. The points in fig. 125 correspond
THE DESIGN OF
150
-2 =
-
The number
of revolutions per
_{
080.
minute
is
290, so that
-^L^ = 0-48.
and 9 are
25 deg., while
The mean value
respectively 131 and 24 deg. The water is discharged upon
the wheel by a single rectangular orifice of 71 in. breadth
</>
Fio. 127.
depth when
2"28 in.
fully opened.
IMPULSE TURBINES.
151
;.
THEORY OF
152
of the water fixes itself at such a level that the valve remains
slightly open, and as much air is allowed to enter as is
drawn off by the suction tube. This arrangement has two
disadvantages. The air carried down with the water lessens
its specific gravity, so that there is thereby a loss of head,
and when the turbine is working at reduced power, the
velocity of the water in the suction tube becomes insufficient
to carry air with it, and the tube empties itself. The power
of the wheel then suddenly decreases, and the sluice is
rapidly opened by the governor, the discharge suddenly
increases, and the tube fills again, and so on.
By this, at
Certain powers, very harmful oscillations may be produced.
2
2
- V^
Cg
Cj
..
2g
2<7
Where
c lt c t
loss of
neglect friction
V V
where
ff
(a*
Cl
?3.
1)
ri
approximately equal to
85
we may then
write
IMPULSE TURBINES.
163
BT
from
T draw
cl
J^ 2 -
BC
1)
in the
first
Now
take
BS = ST 2 + TB 2 =
2
by the
ACi =
last equation.
ff
AC
Here
and therefore
BS
a-
is
v2
taken as l.
c 2,
and draw
d D equal to
at the angle
(here 25 deg.). Join AD. This is the
similar construction may
absolute velocity u of outflow.
the
inward
flow turbine, but here
be applied to
triangle, while
is
Ad
- upon
A X. Make
154
then
AE
is ^i-
Next
project
to
G on
line
AX AG
;
is
<J
w 2t
so that
c
GE = w
1
We
-i
C2
w z = DK.
maximum
value of
To do
FE
DN
We
AN
FE
Dd
<f>
r -J- which
V& i
give
maximum
ff
efficiency are
=
=
Q'445
0'47
0-458
when
when
when
i? 2
*?
i?
=
=
=
CHAPTER
155
XXI.
11
*J
'
Vj_
Fig. 131
than 5|ft.
Ci
shows a type
<f>
c2 .
.*.
Cx
sin (a
=
+
V 2 COS 9
<)
sin a
156
.*.
sin (a
.'.
Generally,
<f>)
sin a cos
- sin
DCA =
.'.
a cos
<p
about 15.
15.
CHAPTER
157
XXII.
THIS wheel
is
Fid. 133.
and
Engineering.
THE PELTON OK
158
~ M
Work done by wheel - c(Wi
where w lt w 2 are the tangential absolute
water at inflow and outflow
WL = v
-C
U>2
=V
t'i
total
wasted energy
velocities of the
VZ
- C
Vy
2y
The
2g
is
maximum
efficiency.
2c(v -c)
Jl + F*
.'
= - whatever
the value of F.
efficiency.
Some
tests
of the
following results,
the variation in speed
causing the variation in efficiency.
TANGENTIAL WATER-WHEEL.
EFFICIENCY TESTS.
159
101
may
be desired.
CHAPTER
XXIII.
term
of years.
In Mr. Parsons' iii ward -flow turbine, figs. 137, 138, and 139,
a series of inward-flow wheels are arranged on one rotating
shaft, and enclosed in a cylinder containing the guide
blades, so that steam entering the first of the series passes
successively through each before being discharged into the
atmosphere or condenser by way of the port T. The turbine
wheels consist of metallic discs, combined with bushes, which
are slipped upon the shaft, and keyed or fixed to it in any
Each disc carries one or more series
convenient manner.
of blades in fig. 138 only one series is shown so that the
velocity of whirl is reduced to zero -before leaving the
wheel.
The enclosing case carries ring projections and
guide blades, and the ring projections are arranged so as to
12P
Q .t
-h
163
164
form a
series of chambers,
rotate.
FIG. 139.
165
oil
1GG
first
is
167
made
as small as
possible.
throttle valve R.
and
thfi coil is
168
turned BO &a to slip easily over the bush and each other with
a slight freedom of fit. In the figures the looseness is some-
D 1 figs. 137 and 138. In fig. 138 the two bearings of the
turbine are shown in position at
and K
When oil gets
access to the tubes a, b, c, which it is allowed to do by
means of holes or grooves, a slight vibration set up by the
spindle J, J causes the tubes a, 6, c to move or shake within
each other. But the movement is damped by the hydraulic
and capillary resistance of the oil between the rings, which
must be squeezed out from between a, 6, c, and at the ends
and
K K
fully,
and vibration
is
avoided.
in
it
1G9
f,
g, h,
lateral
movement being
allowed.
170
I,
separately from
by constructing the thrust pieces ij,
the block L, of sufficient diameter and thickness to be
slots cut out as shown in the right-hand view of
elastic
The washers may also
fig. 145 will give greater elasticity.
be held up to their work by springs, fig. 14G.
The packing arrangements to avoid leakage of steam
fastened on the spindle and
consist (fig. 147) of a hush
contained in a block 0, which is in halves, grooves n and
1
2
both as shown, and by
projections p, p p being formed in
/.-,
171
172
173
relatively to
174
The packings
on
175
176
13p
177
178
end, thus keeping the velocity of the steam practically conThe steam is admitted through the
stant throughout.
double-beat valve H, and enters the turbine case all round
the spindle at J, thence flowing in each direction towards
the ends of the cylinder, the impact of the steam on the
moving blades causing the barrel, and consequently the
turbine shaft, to rotate.
The steam turbine may be advantageously applied to the
driving of almost any class of machinery, either by direct
coupling or by means of belts, ropes, or gearing.
On the llth, 22nd, and 28th of January, 1897, Mr. W. D.
Hunter, M.I.M.E., engineer to the Newcastle and District
Electric Lighting Company Limited, carried out a series of
tests on a 200 kilowatt continuous-current Parsons generator,
for the purpose of determining the steam consumption under
different conditions of service and various grades of output.
The generator was designed to work with a fair all-round
economy, whether exhausting into the atmosphere or into a
condenser, although in general the latter method is expected
to be employed. The difficulty of obtaining high economy
imposed by such a set of conditions can only be partially
met in ordinary engineering practice by adding to the
engine a costly automatic expansion gear, which cannot
always be relied upon when required to act within a widelyfluctuating range of load, or when called upon suddenly to
work at full power, either high pressure or condensing. In
the case of the Parsons turbo- generator the provision for
expansion is constant, and when designed to exhaust into a
condenser only, the terminal pressure would be about l Ib.
to the square inch, with an initial boiler pressure of 140 Ib.
the steam being thereby expanded about a hundred times.
The degree of economy obtained under these conditions is
already well known the last 150 kilowatt generator supplied
to the Newcastle and District Electric Lighting Company
required only 17'281b. of water per E.H.P. per hour at full
load.
When, however, the motor is required to give fullpower working, either high pressure or condensing (with a
moderate consumption of steam), the problem of how to meet
conflicting requirements presents many difficulties, and the
fact that these difficulties have been successfully met in the
;
made
179
Kilowatts.
219-2
987
54-5
2030
1061
208-0
108-4
Total water
Tot
per hour.
180
it
T\
001
Oi
09
05
OE
OS
01
J9JM tJOX
00001
0006
OOOS
OOOi
0009
0009
OOOt
OOOE
OOOS
0001
181
CHAPTER XXIV.
COMPARISONS BETWEEN THEORY AND EXPERIMENT.
IN the following examples we have chosen only such experiments as give inflow without or with very little shock.
for doing so has been to avoid as much as
possible the more complicated calculations that would be
otherwise necessary. It frequently happens that in practice
inflow without shock and radial outfl >w are not obtained at
the same speed, and we must therefore take into account
The reason
w2
at discharge.
maximum
efficiency.
The
first
COMPARISONS BETWEEN
182
2
= 2 822 ; Q = 04 88 ;
105 '5, whence we obtain
H=
16'56
-IS
c
= 27T
60
where
N=
If
<P
= 2*
x 2-085 x
tan *
= _^l5JL_
V COS a
3-89.
23 - 22 3 x '9523
<t>
23.
75
7',
which
may
shock at inflow
=
ws =
differs
trcosa
Wt
very
Tip
H P = 62 5
-
45',
- v 2 cos
'
from 77
little
so that
be neglected.
c(w l -
-550-
w9
1 '25.
62 5 x 64-88
23 x 20
650
322
105-1.
The actual H.P. was a trifle lets, 101 25 this allows for
bearing friction and for the usual disagreement between
theory and practice. The maximum efficiency was obtained
at a less speed, viz., 96 revolutions, with 64 99 cubic feet per
second, and 16 55ft. of head. This gives rise to shock at
inflow and a backward discharge, so that w 2 is negative,
while in the above example w 2 is positive, and is the cause
of a loss of over 6 H.P., which would have given maximum
efficiency if added to the above.
The next experiment is that of an outward-flow Fourneyron turbine with a Boyden difl'user, a ring whose section is
;
A B,
183
fig.
155,
pressures, then, if
we
neglect
friction,
+ Hl=
62-5
2gf
625
U25
**
+
2g
2g
But
v s is less
surface
A is increased by
the amount
= v *.
<(>
FIG. 155.
2r x =
The maximum
73'6 in., 2 r.2 = 90 in.
was obtained at 63| revolutions per minute,
corresponding to a value of c = 20 4 and a deflection of 8 31'
of the relative velocity of inflow; that is, for iiflw withshould be 98 31', and it is actually 90. The
out shock
maximum efficiency was 8017, and at 6fi and 71 12 revolutions it had only fallen to 78 79 and 78 66 respectively, while
the change of direction of the relative velocity at inflow was
reduced to 4 39' and 2 52' respectively. We are therefore
justified in comparing these two experiments with theory,
neglecting any shock.
Taking, firstly, that at 7112
square
feet,
efficiency
<t>
revolutions,
COMPARISONS BETWEEN
184
Q =
H=
v
16-64
__
a x 9
for
radial turbines
,
tan
v cos
8m
87
=
Cl
2275 x
v.,
62 '"
550
19 9
14987
9
5'(j(i x
_ 626
29 5
=
JJL
To O
cos 9
ci
Q
s5
Cl
t-&
27'9 nearly
7'j
27-9
- 295 x
'8988
27'9
1'4
= 14
26'5
friction,
~
i
gg
^a
27'9]
24g
we assume
If
by
if
24 T> x -4067
5
8',
24-5
24 5 x -9135
Neglecting shaft
HP
=
'9
, 2==
Co
14987
6814 x
62'5
x 149-87 x 16'64
OQO
283
'
550
the amount lost by the above cause
Then
248 - 8*49
is
8 49 horse power.
239'51
The actual horse power was 222 '5, and the difference between
actual and calculated is 1701, which is 6 per cent of the
available power, and 7 '66 per cent of the actual power, and
is
186
H=
Cj
tan
<t>
c2
Q = 148-32
= 24-2, Wj_ = 221
calculated = - 12'3.
= 94 39', so that the actual divergence is 4 39'.
= 261, v = 29-25, w = - 15, and H.P. = 248 nearly.
=
16-62,
21-3, v
actual horse power was 220 '28, and the difference will be
much the same as before, the greater shock at entry helping
to account for the discrepancy between theory and practice.
The third experiment given in this paper is that of the
Hercules mixed-flow turbine, of which, unfortunately,
insufficient data are given from which to calculate the
manner in which the water leaves the wheel. Not having
sufficient data as to this class of wheel, we have avoided the
theory of the subject here, the difficulty being to decide as
The
186
= 14
45',
feet, TI
<
98, n
15ft.,
7834
4752 x "72
cj
tPi
If
we
calculate
tan
we get
is
4>
= 21-45
= 229 x
-9G7
2215.
d- wt
62 5 Q e 1 MX
650 x 32'2
Hp
-
62'5
x 78'34
550
21'45 x
2216
32 2
131
we do not know
the outflow takes place, it is no use speculating as to
the causes of difference between the actual and calculated
horse powers. These are all the experiments described in
detail in Professor Thurston's paper, and we wish the reader
to understand clearly that we have not selected these in
preference to others because they give closer agreement
between theory and practice, but because they were made
with the greatest accuracy, and the wheels are of modern
construction. Professor Thurston'* paper, mentioned above,
need only be read to appreciate that accuracy.
Again, in
selecting tests at particular gates and revolutions, we have
been solely guided by the efficiencies obtained and the
absence of shock at entry.
These and other experiments show the best agreement
between theory and practice when the areas are measured
This method of measurement has
as explained above.
another advantage, viz., that there is no need to calculate
The
how
187
7t
at inflow.
CHAPTER XXV.
THE CENTRIFUGAL PUMP.
THIS type
known
of
pump
is
lifts,
but
it
to
rises.
188
Figs. 158 and 159 show two sectional elevations of a centriWhen pumping has
fugal pump with horizontal shaft.
commenced, the water enters at A, and flows up the two side
passages to the eye B of the pump disc or fan, the water
As the disc is rotating in the
entering at both sides.
opposite direction to the hands of a watch, and as it is
necessary that the water should enter it without shock, as
in the case of a turbine, the vanes must be curved back at
the inner diameter of the fan, because the water cannot have
189
disc, its
FIG. 159.
pump
disc depends
190
CO
191
FIG. 161.
a
engines start, so that the condenser is not overheated, and
good vacuum is at once obtained.
and
work
is
which
the
small,
pumps
3. The head against
therefore the efficiency of the centrifugal is greater than
that of a reciprocating pump.
4 Its action is continuous, and no valves or air vessels
are needed in consequence.
5. Its discharge may be varied by increasing or reducing
Fig. 161
pump with
cover removed.
THEORY OF THE
192
CHAPTER XXVI.
THEORY OF THE CENTRIFUGAL PUMP.
CENTRIFUGAL PUMP.
193
should also be
entry and discharge. The angles 0,
noticed, and the fact that if wz is zero /3 is a right angle.
The work done per pound during the passage through the
u, v at
<f>
disc is
so that
if
is radial,
this
becomes Hi-l
g
not radial, and a vortex has been set up in
the eye of the pump, the total work done by the disc per
and even
pound
is
water
is
if it is
Wl
Cl>
.
is
Bv 3
to
Bv 2
of v2
(v z v 3 )
14p
(tta
va)
is
and
(ftfal?
where
(v2 v 3 )
but
-
(uz v2 )
c2
- Uz cot
9.
does
between
THEORY OF THE
194
'
.
=
and
it will
L! = -L
be avoided
c2
cot
0)
....
is
&
no
(2c)
dilluser,
the direction
loss of
= u it
head
or
Av 4
is
when
-^
v4
(Ic)
if
w 2 cot
(c 2
is
w4 ,
fig.
164,
and will be
= w lt when
least
when
(v
v4 )
(v
wj
20
4c)
2gr
CENTRIFUGAL PUMP.
195
This will be at least when the first two terms on the right
are least that is, when (Av4 ) = (t/4 lo^).
Therefore, when a discharge pipe of increasing diameter
cannot be used, there will be the least loss of head when
(5c)
and
w i may
Wt
If,
(6e)
THEORY OF THE
196
of
momentum from A
and
wi
*
r3
A
A D B is an equi-angular spiral
The
its
weight
is
increased.
.ind in
'--&)
......
<*>
the latter
We
CASK
I.
Pump with no
one.
Head +
losses
TT
= work done by
r
_ U>1 Cj
LB -- *
H+
,
disc.
CENTRIFUGAL PUMP.
197
= u2
c2
H+
We
?7
cot 8
of
(2c)
+ M
27
doubt
it
treatise
Now,
.'.
2g
Wj_
H+
- u cot
(C L
2
!
+ UL2
The hydraulic
2c L z
u t oot
GJ_
0)
+ u^ =
1
- 2Ui
- 2Ci u t cot
<f>
2c,
(c 1
% cot<)
C0t a
g> ;
efficiency
useful
total
90
45
30
20
10
'47
58
'65
'73
"84
CJL
1'03
1-06
112
124
175
results of
THEORY OF THE
198
f agal
pump
was very
rose above
WH P
"41,
-p^-
than
'3,
<
<;>
<t>
Head
-f
losses
H,T
and
.-.
'!
2c x
(G!
2
Ci
cot
c, ?/!
and
= work
Cj
?<!
cot ^ - Wj 2 (1
v
disc.
LlA
(ct
<#>)
done by
cot #.
cot
<p)~
cot2 0)
Mi
2grH.
2 gr H,
_l_
N/2.gH
2.gH
putting
12
i
i s/2flrH.
CENTRIFUGAL PUMP.
90
45
30
20
15
The
199
'725
............
............
77
............
94
............
"80
............
............
-84
............
103
1189
............
'87
............
1'355
table given
by Professor Unwin
is
as follows
-n
90
45
30
15
and
c,
............
-0
............
............
-8
............
............
-9
............
............
-87
............
-83
-93
-98
1'2
30 gives maximum
therefore, states that <j>
whereas the efficiency increases as decreases.
reason for the increase of efficiency is that W L
he,
efficiency,
The
-83
............
decreases
decreases
<f>
when
<f>
decreases,
and
L4
decreases
when w
H+L +L =
2 g H + Wi + D =
and
MI
The value
We
of
have taken
D may
^LEt
2
d w2
V20H.
it as 7 ft.
H+
ii
lo
+ &H
D = J 2gHk =
where
Values of
r?,
&,
7.
5
8
16
24
153
-82
"096
-048
'032
'024
86
32
77
-90
91
'92
THEORY OF THE
200
is
<t>,
so,
<
<f>
d^
and even supposing
-n
as great
when
<f>
Since
r#i
Ci
cot
90 deg. as
and
Cj
when
<P
increases as
is
<f>
<t>
and
= r/H
From the above quadratic the following
calculated
results
are
=
,-^TT =
N/20H
(1-062+ fy
90
45
30
15
745
-88
'991
1-346
when
026
1-376
when
= 78
'915
ij
r,
'9
-82
H = 1C
H=5
Comparing
this
</;
<t>
>)
pipe.
CENTRIFUGAL PUMP.
To show
201
this numerically
T
.
JJ2
~~
-L^3
T
J-^4
2 ~D 2 -
when no gradually
where
But
W^
= ffH
n
= 2f/H =
2
'
rj
Cj
L - L
8-02
174 x
/5~
7 6
1 355
2 x 7 2 - (7'6) 3
40
greater than L 4
With a reduction of
the efficiency would increase, and a
discharge pipe would be an advantage theoretically, but to
so small an extent as to make its use in practice unadvisable.
For values of H, between 16 and 32, it is an undoubted
advantage, giving a gain of from 3 to 5 per cent when
= 15 deg., and experiment might show we say "might,"
because no such experiments have been made that the speed
= 30 deg. to 45 deg.
of the disc might be reduced by making
without loss of efficiency. We have the following experito
our
belief.
In Engineering,
however,
support
ment,
vol. xliii., page 93, there is given a description of centri= 90 deg.,
fugal pumps at Khatatbeh, Egypt. In this case
= 20'8,
= 10, so that c x = '82 JYg~W, D = 2, MJ = 27,
Cj
allowing a coefficient of contraction of -9, and w 4 the
velocity of the volute, is 127, while the mechanical
If we
efficiency of pumps and engines is 65 per cent.
divide this last value by '9, to allow for engine and shaft
friction, we have a hydraulic efficiency of 72 per cent, and
and obviously
L 2 + L 3 is
<f>
<f>
<f>
^5,
because
<f>
90 deg.
c<\
(20-8)
loss
THEORY OF THE
202
The
We
--lo 415
efficiency
to these causes.
745,
is
even
Let
then
L7 =
and
40
.'.
2c l WL
and
2c 1
(c 1
2(7
= m*
H+
<V (
^)
_
OT
(i
+
4(7
+ i>i 2 )
cot0)=
+
2 -
20
(2
cx
u, cot ^ (2
16
-4m"
J{2 -
2 2
)
- 2c r
u^- cot- $
(c^
cot^)
2
)
cot
cot2
(4
16
,-
c x (c x
m2 (32 + 2 m + m
4m2
2
- Mx
cot 0)
cot3
CENTRIFUGAL PUMP.
The following
tp.bles
736 v/2r/H
"
45
"903
1001
30"
15
= 90
203
'885
'939
"87
'863
'979
'97
'86
-90
"92
"89
'925
"933
"945
'952
1'33
7?
= -806
45
30
15
-846
"88
'94
-905
Comparing the above table with the last but one, we see
= there is an increase of about '07 to '08 in
that when
the efficiency for equal values of <, and that the speed is
also decreased. The decrease of
does not bring with it a
corresponding decrease of 77, and it is evident that, while a
small diffuser is an advantage, a large one is hardly worth
We know of no type of pump
the additional first cost.
= i.
In figs. 158 and 159,
with such a low value of
as
One side of this diffuser forms a flange to which the side
cover is bolted. As in practice such pumps are made with a
about 15 deg., and as we know of no experiments
value of
in which
is as large as 30 deg., we cannot speak with
certainty, but it appears from the above calculations that
than 15 deg. might be used with
a larger value of
advantage.
It is not worth while considering the case in which a
discharge pipe of increasing diameter is used when the
pump has a diffuser, as the object of both these arrangements is to reduce the velocity of flow, and so increase the
pressure, and when one is used the other is not required.
We must next consider what occurs when shock takes
place at entry. We shall suppose the direction of motion
just before entry to be radial.
<
<f>
<i>
Then
loss of
head
=
g
THEORY OF THE
204
If
hydraulic friction
to subtract a
is
term
from the right-hand side of the equation last but one, but as
there are no means of calculating the values of F from
experiment, we have omitted this term. In the case of the
turbine ample information is obtainable the reverse holds
for the centrifugal pump, as makers invariably refuse to
give any information, and to our knowledge there are only
two sets of experiments in which all necessary particulars
are given to compare theory with practice. These will be
;
Wi(2
Vi
sin
<P
and
ir
- n
TI 61 sin
=
(2
u2 =
r-L
bi
- n
6, <i)
cosec
<t>)
Q
b
tl
where
We
CENTRIFUGAL PUMP.
On
20o
Diffuser of a Centrifugal
Pump. Guide vanes similar to those of an inward-flow
turbine may be used to reduce the velocity of whirl w lt after
the water has left the wheel, more rapidly than is tne case
when the difiuser is not fitted with them. Suppose that
these vanes make an angle x with the tangent to the circle
of radius r such that
and
ike
t/te
tan
and as
MS
a s then
,
W"2,
3
rt b
COMPARISONS BETWEEN
206
CHAPTER
XXVII.
THAT an
increase of
reduces the speed we have ample
for the pumps at Khatatbeh Ci = "82 J2 g H, and
proof
= 90deg. An experiment with a Gwynne Invincible, in
<f>
<t>
1881, near
Amsterdam, gave
Ci
<p
Pump (vol.
liiL).
TABLE
No. of
experi-
ments.
A.
207
^=
*Ci
100,
Wj
TABLE
No. of
experiment.
B.
COMPAEISONS BETWEEN
208
and MO
<f>
94 u it
<j>
MI
Putting
K=
60 x 6 z5 (2 v ri 6 X - n &!
coseo
<f>)
60 x 6 25 (2 r r z b z - n b z
ta
cosec
B)
'9,
001472.
Taking
experiment (1) as
G =
and
/.
v4
=
=
an
illustration,
1012,
1-48,
2-35
UL
3-48,
*-.* -5
N=
where
12
Wl =
""
Ui cot
<f>
60
26 04,
'
100=
32 2 x 1467
loss
Let h a
loss
L-t
7<
c2
^Ci
Loss due to
in all cases,
15775, u 2
pump
friction
= 194w
A5
= F
-,
.'.
A.,
2'375.
and taking
F-
2*5
100
may
be omitted
then
HTX
100 x 14-67
1467
100 x
+ 7834 +
14-67
2'375
'4725
57'9.
25-351
i? 3 was calculated by Mr. Parsons in the following manner
The foot-pounds raised per minute were 148,461, and those
indicated 298,438 ; hence the actual efficiency per cent was
4974; but this includes friction of the strap by which the
pump was driven, of the engine, and bearings of the pump.
Also the friction of the outside of the disc is not included
in the above theory, and must therefore be subtracted from
the indicated power.
By experiments afterwards made
these losses were estimated at 45,000 foot-pounds, but this
was probably too much. Hence corrected efficiency was
:
148461
298438 - 45000
_
~
148461
253438
^3.
^ 3 is sufficiently close to
above theory.
Table C was
i? lt
i?
2,
TABLE
No. of
experiment.
C.
K=
1,
and
COMPARISONS BETWEEN
210
In Table
211
Ft In.
19 10
13TH. 1897.
COMPAKISONS BETWEEN
212
F ^? =
The
losses
0-0469 v 4 2
friction,
together with
^-
t>
^
L and
+ L,
dividing the actual head
by
we have the hydraulic efficiency, and the table below shows
that the difference between theory and practice is trifling.
The following was the method of calculating the hydraulic
The 14th ft. is chosen.
efficiencies, actual and theoretical.
Experiment gave the following: I.H.P., 2558; W.H.P.,
911
revolutions per minute, 1541 ; mean head, 13-875 ;
quantity of water discharged, 36,341 cubic feet
interval,
The velocity of flow at discharge from the
9| minutes.
is
817 ft. per second, and the horse power required
pump
to drive the engine alone at 140'5 revolutions was 10 4, and
at 160 was 16 horse power, so that at 1541 it was 14 3 by
interpolation
Calling this F.H.P., the horse power transmitted to the shaft was
losses of
head
I.H.P. - F.H.P.
.'.
the
pump
efficiency
-^-i
S.H.P.
'3775.
241-5
This
differs
from
Wi =-2J*andc,
c "
= 4416;
.'. iv !
266.
<-*)-
't
4'66
-07 *4 i,'
13J875
34'535
The
orifice
_=
213
20-66
4Q25
'025.
made
The casing
COMPARISONS BETWEEN
214
Fio. 168
power
is
to
215
is sufficient
to
make
by applying a
same position
known moment
1,200
and
1,175 revolutions,
were made,
216
COMPARISONS BETWEEN
Fio.
170.
217
W
W
the 7
in.
wheel,
W was 686,
and H was
For the 6
in.
wheel
30'48.
Hence
'95,
so that
TTT
j=
Now, the curves in fig. 171 show that the efficiency increased with the orifice, and we think it probable that had
the 6 in. wheel been tried with a great discharge, it would
have had an efficiency of more than 40'8 that of the 7 in.
wheel so that it is probable that 12 vanes would give a
better result than 22.
To test the efficiency of the vortex chamber a wheel with
radial vanes 11 in. diameter was used, and trials were made
in which the rise of pressure of the water after leaving the
wheel was observed, and the curve of actual pressure plotted.
The theoretical pressure, on the assumption of no loss of
energy, would be obtained in the following manner. Let v
be the velocity of inflow to, and v z that of outflow from, the
;
diffuser.
Then
v ri
and
v s rs,
head
3
Putting r x and r equal to 5'5 and 9 in. respectively, and
= 58 04, as in the experiment,
= (58-04)
/5j5\
a
~|
= 327 ft.
increase of pressure was from 22'92 ft. to 38 76ft.,
or a difference of 15 '84 ft., so that the efficiency of the diffuser
The actual
218
was
56 and 59'8 per cent for radial and curved wheel vanes
respectively. The speed for the former lay between 1,312
and 1,362 revolutions, and for the latter 1,313 and 1,385.
Unfortunately, the paper does not state the angle made by
the wheel vanes in the latter case. It is interesting to note
that in the wheel with radial vanes the angle of flow
relative to the wheel was not the same as the angle of vane
Had it been so the efficiency of the
i.e., was not radial.
very
little less
H ~_
7?
32-2
x 36
(407)*
than g
we have
RR
8
'
H -r c^-.
In
219
CHAPTER
XXVIII.
mouthed
is
also a
CENTRIFUGAL lUilS
FlO. 174
AT KHATATBEH, EGYPT.
221
cone of the lower base of the disc, which are curved spirally
backwards through an angle of 60 deg. From this spiral
form the vanes gradually change, until they become xadial
CENTRIFUGAL PUMPS
chamber, as we have already pointed out. The volute is
extended by a mouthpiece 5 ft. 3 in. in diameter, fig. 175. To
this is attached the discharge conduit above mentioned,
which is 50ft. long, and which changes from a circular to a
rectangular section, 8|ft. high by 13ft. wide, so that the
The
velocity of discharge is less than 2ft. per second.
opening is controlled by a sluice valve.
FIG.
176.
In order to avoid an immersed footstep, and the consequent difficulties of lubrication, the Fontaine system of
turbine pivot is adopted.
The lower part of the pump
shaft is made hollow, to admit a column or
support C,
fig. 173, solidly fixed below, and serving as a bearing above
the level of the pump itself, and nearly 5ft. above the
highest probable water level. Above the fan, the hollow
AT KHATATBEH, EGYPT.
223
224
PlO. 180.
may
be cooled.
225
FIG. 181.
1C?
226
forwards.
The manometric
efficiency
is
from
0'35 to
227
184, 185, and 189, whose sides are either plane, as in figs. 184
or 185, or conical, as in fig. 189, and the radial section of
whose outer surface is not a circle (as is usual where diffusers
are employed), but is a spiral, which is much more efficient.
and
230
not
so.
limits
231
FIG.
We
232
demanded by the
change.
(/i
we
If
charge, and
formulae.
efficiency,
power, which
The
first
is
are given
usually called
by the following
the
inane-metric
where r t
third
is
is
and the
234
Fio. 192.
Fio. 193.
235
/j.
FIG. 394.
Q
but which he
now
defines as
so that there
</>
<
236
<t>
T increases
with
types, and
lies
of vanes.
In a
5,
and the
ratio
between 1 and
pump
4,
must not
be great, otherwise, if
accidentally becomes zero, the motor
may be burnt by the current it takes. Care is taken in each
case to obviate this difficulty. Messrs. Sautter, Harl6, have
237
volumes and small lifts. Figs. 193, 194, and 195 are the
general arrangements of bilge pumps on board ship, with their
sl
valve B and the non-return
necessary valves, viz., the sluice
pump
filled
an arrangement
pump losing its water and becoming filled with air, and
consequently ceasing to work, and another for re-charging
a pump with water should that be necessary. The first
238
MN
MN
may
be.
dis-
MN
MN
239
CHAPTER XXIX.
THE EFFECT
OF THE
VANE ANGLE
<t>
UPON THE
DISCHARGE.
<f>
TT
C l Wj.
(C 2
g
which
may
0)
- Vj'
_
_ ( Wl
~^~
% +D
~27~
2
2g H =
where
- Mg COt
~W~
Ci
(i
A)
u ZL cot
n
u* cosec2
2 VA
(Ci
- u?
cot2
Ui cot 0)
240
This
may also
<i>
first
take
<f>
pump
Work done
A.
is
29
JL*L=
+
*ff
-906.
WiC L
and
wt =
c lf since
<t>
90 deg.
241
cx
n
n
T,
When
0,
c^ =
x 32'2
H=
=
=
=
=
g
'905
-81
724
578
and,
ct
In practice a centrifugal
pump
will
generally begin
this
242
<j>
D =
Let
and u
8,
qofi
as before.
C t (Ci
This
is
a quadratic for
ci
n =
3,
Let
u2
Let
Ui COC
MI cot
</>
=
=
.)'
43-67
43-67 - 29 84
as before, then Co
then cot
10,
<f>)
M giving
Ci
17 9
- 8 cot
Ci ( Cl
-^
n.2
2
cosec
#>
14-9
13'83.
= 14
^?
3
55.
T455,
(i
we
- J.
u^ cosec2
+ (*u*<*** +
#.
2
2
cot2 e -
2 ,4
)
a
Cl
- 2,
4Ml cot *
- 2gB.
obtain
Cl
where
Q =
1677 Qci
- ir/H - 4015Q2 =
O,
O,
Q
Q
Q
Q
Q
Q
The
=
=
=
=
=
=
10
......
25
......
34
......
40
......
d =
=
d =
d =
d =
50
......
Cl
67C>
......
ci
32'3
......
rj
32-86
......
17
43'67
......
-n
53-5
......
60
......
9?
72-1
......
by
O D,
243
results
T?
fig.
= 7
= -8
= '905
= 7875
- '74
= -61
198, is 32 '3
ft.
per
<f>
<t>
than
<t>
<t>
<t>
<t>.
244
TABLE
Gallons
discharged
per minute.
II.
TABLE IV.
Gallons
discharged per
minute.
VARIATION OF PRESSURE IN A
246
If there is
(
2g \
2
2
cr - Wi cosec ^
no further gain
is
such that
2 o h]
head in consequence of a
of
-*-
by putting
We
M =
different manner.
and by putting
CENTRIFUGAL PUMP.
w =
and
-r
625
-i-
201{
d2 -
MI cot
^-
2
!
2
1
-2c
<t>
h - hi
<,
2
Mi cot
Ax ) -
w 1 cot0 +
H+
cosec2
2gr (A
247
- u : cot
GI
29
r(A-^
- D
J
- D2
so that this is merely another method of arriving at an
lation between c ls u lt H, and ^>.
equation
w44 = ^
Next,
Text, let us suppose that v
$w lt and that the diameter
of the discharge pipe is not increased ; then there is a gain
of pressure by shock when the water leaves the disc and
enters the volute. This gain is
<t>.
Cl2
and p G
.'.
2g
- Ci 2 -
=
1
(H +
A)
cosec2
62-5
cosec ^
-c i
+ 2gh
cx
% cot
is
0.
to
THE BALANCING OP
248
CHAPTER XXXI.
THE BALANCING OF CENTRIFUGAL PUMPS.
THE reader will now perceive that the pressure on the
disc may cause a thrust along the shaft which will require
balancing. Pumps with two inlets, as in figs. 158 and 159,
are self-balancing, but when the inlet is at one side the
thrust may be considerable. In a paper on this subject, read
by Mr.
J.
Coast,
figs.
to obtain a balance.
Fig. 200 is a vertical section. The water enters by double
at the top of the pump, and then through the
inlets
nozzle
I,
which
is
not
in
On
CENTRIFUGAL PUMPS.
249
'
'
'
"
Any water
250
THE BALANCING OF
CENTRIFUGAL PUMPS.
251
10 ft. per second, the diameter of the inlet being 8 in. This
may be calculated as follows The horizontal momentum of
the water is destroyed, and therefore, if this is the cause of
a pressure of p pounds per square foot, and
is the area of
the inlet in square feet, the impulse of the pressure p
in
:
time
t is
where
and
seconds,
WV
pAt,
A
V is its velocity.
A_
inlet in
AY*
62-6
625 x 7854 x
(f)
x 100
32
68-2
lb.,
(r
If)
axis.
The
d r,
force is perpendicular to
252
THE BALANCING OF
+ dp
,+%
the mean pressure on a 6 and c e in pounds per square
foot The resultant of these fluid pressures is
is
FIG. 203.
Bat F = P for
acting in the opposite direction to F.
equilibrium, and we may neglect small quantities of the
second degree ;
r2
o>
_~ 2rdp
62-5
g
*
dr = d p
6
25'
CENTRIFUGAL PUMPS.
Integrating,
^-~
2y
P
625
=
6^5
27,
Let
Vx
and
let
+c
+
+
C>
'20
-
253
Pi
/V 1 2
^"G^B'^V
!B
,_.
62 5
vA
V
"
,.
'
62-5
W2 ^T?
Pi
-A
"V
/~25~2if\ ltl
R lf
less will
p be
also, if
=
and
if
p s be the pressure
"
same
side,
J!*(52
where
c,
radius n.
u and V
t
2g
(c2
- u* cosec 2 ^
2 g h),
1
at radius r correspond to
Ci, Ui,
and
at
254
Now, on the
V 2 );
also
cosec2 vO
2
(d -
-Mr cosec
2
<p
+ w2 cosec2 vO
u* C08e
'
With the
^'
is
usual form of vane curving hack, x cosec
greater than u cosec f, and, in tig. 201, Vi is less than c :
By a proper choice of Vi that is, of RI, the radius of the
balancing vanes we can make p greater, equal to, or less
than /'
It should be so arranged that the whole pressure
inside a circle of radius KI is less on the right of the disc
than on the left i.e., the average value of p should be less
than /' ;, for between the radii R, and r the pressure is
greater on the right than on the left of the disc ; that on
the right is p4 , and that on the left is less than p4 , except
at radius rlt because it decreases with the radius. By this
means a perfect balance can be obtained. The value of 1^
Its calculation
is, of course, best found by experiment.
would be very complicated, involving the integral calculup,
and as there will be a little slip of water past the vanes,
which we have neglected in our calculations, theory would
not be so accurate as experiment.
<t>
:;
CHAPTER
METHOD
255
XXXII.
FIG. 204
METHOD OF DESIGNING A
256
The hydraulic
Let
?!
-
Cj
H=
5,
20,
18 65 c t - 858'6
1865
V/34S
1865+^37824
O
3434 4
2r,
ft.
5!
in.
A-
ft.
12
(27rn - ntcoaect)
13 x 25
9 x (17-25 - '966; x 5
=
Wl
= d
5 73 in.
= 4007 -
- ?<,cot^
is
1865
21-42
since
M>3 r
/.
W8
-3
M'l
= *?!-
is
l^n, and
r lt
17136.
M,.
8-508.
CENTRIFUGAL PUMP.
The diameter
ED
(fig. 162).
second
cot e
There
is
The
-2
^-,
but
'
.
we may
30. 53'.
.....
take
cosectf
it
as
(2c)
3.
1'948.
is
=
35 x 12
9 x (575 - '4895) x
"8
111.
is
35 x 62 '5 x 20 x 60 foot-pounds.
18p
METHOD OF DESIGNING A
258
Let
mean
A
and
the
d = shaft diameter,
if
T = 196/d 3
where /
Then
is
the
mean
stress in
pounds
9000
35 x 625 x 20 x 60 x 12
x 2 TT x 196 x 9000 x 139 2
d =
3-08 in.
in.,
and
A=
B =
21 in.
diameter at
The side passages are ellipses whose axes are 27'4 in. and
137 in. at right angles and parallel to the direction of
the shaft These give a combined area equal to that of the
suction pipe if it is made the same diameter as the discharge
pipe, viz., 27'4 in.
and
If,
figs.
158
(17136)2
(6
x tf
2g
40
= 114 +
is
'25
1 39 ft
CENTRIFUGAL PUMP.
The hydraulic
= 75
efficiency
20
EXAMPLE
II.
The lift is 16ft., and there are 40ft. of piping, whose friction
must be taken into account. The number of cubic feet per
cot
<t>
176, cosec
c^ -
UL
G!
cot
<A
<t>
015
= 2..
~n
Let
c^ -
cx
G!
H=
x
= 75
- 730 = O
17
-88
r,
27 5 nearly.
fig.
A =
25 X i*4
Q
X &
o
The diameter at
may therefore
We
then
^i
The number
the equation
3,
as in
Example
of revolutions per
60^ =
N=
=
I.
minute
is
2^7"!
60 x 27'5 x 12
54 TT
117
obtained from
__
= v
Ku
inches
(2rrr 1 - ntcosect)
12 x 25
4 '84 inches
*-**--*
u2
Let
41
= -2 = 1145
then cot
;
cosec
9 X 8
8'
152
12 x 25
'
('
- 6 X
ll
r52
21
9 64 inches.
The
WL
=
=
ct
27'5
u t cot
"88
<(>
26-62.
This
is
Diameter at discharge
Let
A2 =
A2 =
is
no
diffuser
13J in.
This corresponds to a diameter of 24 in. nearly. The discharge pipe should increase until the velocity is considerably
reduced.
D =
A3 =
Suppcse
4,
and
its
area
is
A3
in.
THE FAN.
261
pump
is
*(J
1
&
636
ft.
64'4
The hydraulic
efficiency
'75
17
22-666'
by hydraulic
loss of 5 '03
ft.
of
head
friction.
CHAPTER
XXXIII.
THE FAN.
FANS
fan
is
'
'
'
33UOO
262
THE FAN.
.'
Q = CA
V^H;
263
FIG. 205.
Fio. 206.
264
THE FAN.
THE FAN.
265
and as
THE FAN.
266
The formula
the
THE FAN.
anemometer gives a reading more nearly approaching
267
to
2G8
THE FAN.
Fio.214.
THE FAN.
269
270
THE FAX.
THE FAN.
271
and
Fia. 219.
more than 1 in 8, to avoid the loss due to sudden enlargement of passage. Ser ventilators have a small diameter,
ft
from 1-4 metres, or 4'6 ft., to 25 metres, or 82
formed
The Capell fan is shown in figs. 211 and 212. It isaxis
and
same
the
of two fans, one outside the other, having
consists of
at the same angular velocity. The nrst
.
revolving
272
THE FAN.
cylindrical surface contains six openings, in general rectangular and spaced at equal intervals, whose area is less
than the cylindrical surface of the drum, but equal to that
of the eye at least Six vanes of steel plates, curved and
THE FAN.
fan.
273
19p
THE FAN.
274
of
an
ellipse.
S
and not S
= a 0,
as
is
= ae +bo*
circumference.
Finally, the air enters the chimney, whose sides are
inclined at about 7 deg. The volute is constructed partly in
iron and partly in brickwork. The chimney is of steel or
masonry. Fig. 215 gives a perspective view of one of these
ventilators with part of the casing removed.
Although
generally constructed with a horizoncal axis, there are
certain particular cases where it is advantageous to use a
It is then possible to make the
vertical axis, fig. 216.
diffuser and volute and one side of the wheel case in
masonry. The shaft is carried by a footstep bearing, containing means of adjusting its height, and it can be driven
by bevel wheels, ropes, or direct by a steam engine having a
vertical axis.
These fans are made with diameters of from
1 to 4 metres
that is, from 3 28 ft. to 13 1.
Small ventilators, driven by engine or by hand, are shown
in tigs. 217 and '218, and a sectional elevation of one is
shown in fig. 219.
The Waddle open-running fan, fig. 220, is also largely used
in this country.
Its vanes curve backwards at the outer
periphery in the opposite direction to that of rotation, and
the curved rim is intended to reduce the velocity of the air
at discharge, and so convert the kinetic energy of discharge
from the fan into pressure energy, although the curvature
seems rather too rapid to effect thin properly, considering the
high velocity with which the air is discharged from the fan.
The Bumstead and Chandler fan, figs. 221 and *22, runs in
a casing of spiral form. The blades are stated* to be of
modified S form, curved forwards so as to cut into the air at
entry, the velocity of the air being gradually increased until
outflow from the fan. An cvatu chimney is added to reduce
the velocity of the air to as low a value as possible.
*
20th, 1893.
THE FAN.
Fio. 222.
275
THE FAN.
276
The Heenan and Gilbert fan, tiers. 223 and 224; has vanes
which meet the air edgeways at inflow and curve backwards
at
first,
80mewhafc 8imilar
The majority
<*>
of fans, however,
especially those used in
ountry, have vanes curving in the direction opposite
THE FAN.
277
that there are guide vanes m, m, fig. 231, which give the air
a tangential motion before reaching the fan, but in an
The object
opposite direction to the direction of rotation.
THE FAN.
278
of the revolving wheel
to zero
is
FlO. 229.
air,
and
c is
is
the
mean
-^?.
velocity of
THE FAN.
279
In
fig.
Fio. 230.
FIG. 231.
FlO.
L'32.
Fro. 236.
cent.
catalogue.
283
CHAPTER XXXIV.
THE THEORY OF THE FAN.
THE
feet,
then
of air is Cl
\ where
is
and
of the velocity of discharge
x is the radial component
but where, however, space prevents a large diameter, and a
must curve backwards,
vanes
the
be
diffuser cannot
used,
otherwise the efficiency will be low.
The apparent experimental efficiencies of various fans are
with the
very misleading, as they are mostly obtained
anemometer, which we have already stated largely
where
284
"i
as the values of
FANS.
REDUCED ORIFICES
FlO. 287.
are
much
too high.
is
vol. cxil,
way
their accuracy.
to prove
Although
might be used with advantage in the case of
a centrifugal pump, the terms orifice and equivalent orifice
are only used in the case of the centrifugal fan. The latter
it
is
-7==, and
is
is
is
is
here the
number
Manometric
efficiency is ^^-,
of
== where
,
is
if
is
of a fan is
may
be thrown in
the form
fj-
+~
+
VM
+ u =
o-,
p, q, r, s, t,
orifice.
equivalent
Q
,^
..^
Actual
TABLE
III.
FAN No.
287
Actual
Calculated
manometric
Equivalent orifice
iu square feet.
manometric
efficiency.
efficiency,
Per cent.
Per cent.
570
630
57-0
59-8
59'8
04
510
416
0'5
33'0
520
416
330
08
177
17-85
1-0
12-1
6-34
160
01
02
0'3
1-5
The equation
1+
2-43
TABLE IV.
600
-P -
13-55
75
is
O2
- 1755
0.
Calculated
manometric
Equivalent orifice
iu square feet
manometric
efficiency.
efficiency.
Per cent.
Per cent.
01*
024
0-3
0-4
0-5
08
I'O
15
The equation
i
* The smallest
equivalent
was -24 square feet.
orifice at
is
O2 -
which a
test
3-51
0.
orifice
288
we
can, for a given type of fan, find the real value of 17,
we can then proceed to design the fan in the same way as a
centrifugal pump. In our opinion an efficiency of 60 to 66
per cent may be assumed for a well-designed fan, in the
If
equation
In the Rateau ventilator the values u lt the radial component of outflow from the fan, and u 2 the axial velocity
of inflow, are, when maximum efficiency is obtained, about
58 J y H, and with an efficiency of f this gives us
,
+ '58^ JJH.
ct
'97
efficiency is
0H
CHAPTER XXXV.
HIGH-PRESSURE FANS AND PUMPS.
gaue
The volumetric
efficiency
V =
A
Cj
The manometric
The
coefficient of
S.
TI
efficiency
pump
T,
MV
The above
formule
are
is
Prof.
Rateau used
20P
290
Centrifugal Pump
01
0.2
03 04
0.5
06 07
0.8
number
Centrifugal Fan
'>
The
fan
291
made
of steel of
very good quality, capable of
running at a peripheral velocity of over 800 ft. per second ;
it turns in a cast-iron
casing having two openings for suction
and forming a diffuser and volute.
The turbine, which is a
steam Pelton wheel, is 11 -8 in. in
diameter, while the fan
is 10 in. ; the method of
raising the oil from a lower to a
higher reservoir is shown in fig. 243, in which the pipe
is
connected to the lower reservoir and B to the
The
higher.
small tube
a
small
amount
of
air
from the fan,
brings
which, mixing in small bubbles with the column of oil B,'
is
292
H and
also
To avoid piston
atmosphere and the lower is left as above.
The
diaphragm is sometimes used.
experiments were marie at Messrs. Sautter-Harle's Paris
works, and the measurements by M. Rateau and M.
The discharge was measured by means of a
Chatelain.
convergent discharge pipe, two of whose faces were fixed
and parallel, while the other two were movable, like a duck's
At its larger end this was fixed to the discharge pipe
beak.
of the fan, and it was thus possible to vary the discharge;
a mercury manometer was used to measure the pressure in
this convergent mouthpiece, and the discharge could be
friction a caout-chouc
293
than
half
an atmosphere.
the
294
discharge Q, in cubic feet per second, estimated at. atmospheric pressure, is equal to the product of S, thd section of
the convergent discharge pipe in square feet, and the
velocity
of flow v in feet per second.
Hirn's experiments prove that
the coefficient of discharge of a convergent pipe does not
differ more than 1 or 2 per cent from
M. Rateau
unity.
proves in his paper that without serious error
295
32-8
..2-8
0-5 H'
The work
obtainable
where
(15-20
the total section of the nozzles in square centimetres, and P is the pressure in kilogrammes per square
centimetre, while I is the quantity of steam discharged per
s is
K _
0-85 +
6-<K>-Q.92 teg P
P - log p
log
Then the
steam
theoretical
horse
is
T =
The mechanical
efficiency
and
is
As the density
296
9*7,7
jo
jSMfxI gfuoq
a9Aod owoq
II
C
T3
ii
pnooas jad
pojqo8|p
or
jo 599; ojqnQ
jj
u} o2r,o2 JaiuAV
f j|8|8|f||S||g||gi
uocij
ojtmbs
aadspanod
aerlo
891ZZOU
mV1B
JO
.1
.1
in
;>
Jtirnuu.l
>X9 jo
aaqmn
297
10 8
where
is the number of revolutions per minute ; it lies
is in metres.
It is, however,
between 1'18 and 1*42 when
not a constant for a given opening of the discharge pipe, but
in. diameter
the
fan
is
10
increases with N.
Although
only
maximum
298
condensation.
These
electro-
and
L_
300
HIGH-PRESSURE PUMPS.
The
Fio. 240.
is
MULTI-CELLULAR PUMPS.
301
HIGH-PRESSURE PDMPS.
302
tested, and the brake horse powers of the motors for any
pressure and current were determined beforehand by experiments with a Prony brake, so that the efficiency of the pump
The discharge was measured by
itself might be obtained.
means of a previously gauged vessel containing 247 cubic
feet
Table
single wheel
pump
303
>
i||
> ~J2
ililllslliiillpiis
ooooooooooooooooooc
0.
oi,^<)00-*ip^.oS-*5
ooooo
oobboooo
SSSSSSSo2SSg
222gS2Sg2^2 00 SSS:
S-doo
^-25-8
111!
11
fc
Hi
HIGH-PRESSURE PUMPS.
304
Pump
OM
with
one wheel
_ O = 208"""
305
306
H~
307
and
^mere
308
(fig.
255).
jfonaptgs
fe
'g
o
W
S
"o
1
g -
310
3H
Pumps Driven
by Steam Turbines.
Figs. 256, 257, 258,
and 261 show the steam turbine and the
centrifugal pump
that
it drives.
It is a compound engine, the steam
passing
from one wheel to the other, but in the experiments only one
ft
FIG. 267.
wheel was used, the other being taken off the shaft so that
the efficiency of the pump alone might be calculated.
312
JURE PUMPS.
Fio. 268.
313
I.
314
HIGH-PEESSURE PUMPS.
its
315
below.
9,000
12,000
15,000
18,000
....
0-19
0'24
0'275
0'31
O'Sl
0'40
0-47
0'52
o-5S5
0'595
Maximum mechanical
the pump alone p m
efficiency of
O gl
60
little
Such pumps cost little, are very light, take up very
or attention.
They may
space need very little lubrication
for sudden outbursts
be used in mines, for sinking shafts,
HIGH-PRESSURE PUMPS.
316
..
OOOOO'TOOOOOOOOOOOOOOOO
dumd
euiqam eqi
jo
osjoq uj rtrand
jo
^JOM
jo
"""xi^isss^^Jisssst!
in;n
eaqoja
pwq
qouj
Woi
uian
punod ui
^s
<lo
supply,
fire
317
town water
<fec.
318
MATHER AND
Flo. 263.
the last or the right hand of the four wheels in the latter, and
shows the discharge. In fig. 264 the water enters at the left
319
320
through the
shown
diffuser,
264, leading to the eye of the next wheel. In this wheel the
same process is repeated, except that the passages leading
MOTOR-DRIVEN PUMPS.
22P
321
322
Wheel
to the next wheel do not pass through the diffuser.
There is very little end
No. 3 is similar to the first.
and
this
fig.
264.
is
Pumps such
pumps
324
CHAPTER XXXVI.
THE HYDRAULIC WORKS AT NIAGARA.
SINCE concluding the subject of the turbine, we have obtained
As they contain the most
descriptions of the above works.
powerful reaction turbines in the world, we should consider
these articles incomplete if we omitted an account of them.
Sir William Siemens is reported to have said that if all
the daily output of coal in the world could be used in
making steam to drive pumps, it would barely suffice to
pump back the water flowing down Niagara River. Many
schemes have therefore been brought forward, discussed,
:ind discarded for various reasons, one of which was that the
picturesque aspect of the place would have been interfered
with by them ; and this was right, for there are few things
on this earth more valuable to mankind than the beauties
of nature.
Some
FIG. 269.
325
32G
friction.
327
328
FIG. S72.
329
A heavy flywheel
reduces the necessary number of guides.
of forged iron, weighing 10 tons, with a diameter of 14 ft.,
FJG. i73.
330
331
UfFALO
"<5v^
ST.j
ir
rods and levers, clearly shown in the figures, which are
controlled by a very sensitive governor of the indirectThe speed varies between a maximum of per
acting class.
cent at the normal rate of working, and about 3 per cent
if the capacity of the turbine is increased or diminished
an
This sensitiveness
is
necessary
electric installation.
HYDRAULIC BUFFEE8.
332
again below the Falls, and near the Suspension Bridge. The
velocity of flow is given by Professor Forbes as 25 ft. per
second, which we believe to be unprecedented.
Ordinarily,
about 8 ft. per second is considered as the extreme limit in
a
brick-lined sewers ; indeed,
leading waterworks engineer
allows only about 3 ft as a maximum in lind channels.
of hard brick, this having
16
is
made
of
in.
here
The lining
been found necessary because the rock passed through
disintegrated or crumbled quickly when exposed to the air.
The brick and hydraulic cement used had united to form a
mass as solid as rock in some specimens taken from the
tunnel.
Fig.
278
The Niagara
is
Falls Paper
CHAPTER XXXVIL
HYDRAULIC BUFFERS.
THE
DA
^3
moved
;
a distance vt.
then the work done
Let
= Pvt.
This equals the kinetic
energy produced, and therefore
Pvt = DAvt. (m -
iy-^L
and
P = D A (m -
I)
jf..
%9
Let
and
initial velocity,
arrested,
^_
which
is
HYDRAULIC BUFFERS.
334
(m -
and since
/.
(m -
I) is
1)
t;
a constant,
inversely proportional to
J^~
The length
We
HYDRAULIC BUFFERS.
326
INDEX.
American Methods
of
Measurement
INDEX.
Fan,
Fan,
Fan,
Fan,
s \
Pumps
at
Khatatbeh, 219.
Foesch and Piccard's
Turbines
at
Farcot's
337
Hauel
The, 2t>l.
Appold, 244.
Bietrix and Co.'s, 282.
Bumstead and Chandler, 274.
4.
276.
Niagara, 326.
Turbine,
Experiments
with, 182.
Francis Turbine, 94.
Friction Losses in Reaction Turbines,
Fourueyron
Experiments on
Pumps,
Friction of Piping, 15.
Friction of Piping, Box's Formula for,
\^"^'*-.-
IV.
of, 28,
119, 149.
Innes'
Centrifugal
206, 210.
for, 16.
Khatatbeh, Centrifugal
Pumps
126.
Pump,
at, 201,
206, 219.
193, 206.
23P
334.
76.
of
Centrifugal
of
Sup-
Pumps,
224.
Turbines, 83.
338
INDEX.
Ptimp, 298.
Mm.
Pump, The
Centrifugal, Balancing
Shape
Pump, The
Pump, The
Centrifugal, Experiments
of,
on, 206.
Pump, The
Centrifugal, Lubrication,
224.
Pump,
"
Parttous'
Experiments on Centrifugal
Pumps,
Steam Turbine,
165, 174.
P.irsonV Inward-flow
Steam Turbine,
101.
Pardons' Parallel-flow
Steam Turbine,
High-lift,
tjtt,
driven by
motors, 298.
178.
Pans,
High-lift,
driven
by Steam
Turbines. 311.
300.
70.
Formula
Radinger-s
Bearing
for
Turbines,
Professor, 132.
Ramsbottom's Scoop
Rankine's Vane,
9.
4.
of the, 5.
143. 147.
65.
Problems on Centrifugal
Pump Design
75, 77.
Rateau's
Problems on Flow
of
Water through an
Orifice, 8.
128.
"Reduced
J7-.
and Pumps,
290.
of,
INDEX.
Reaction Turbine, Axial Flow, Theory
of, 64.
of,
and
Friction, 77.
The
Richards on
Pumps,
Balancing
Centrifugal
248, 249.
from, 77.
Governor, 111.
Stanton's Experiments on Centrifugal
Snow
Pumps, 213.
Steam Turbine-driven Fans andPumps,
289.
Steam Turbine-driven
Centrifugal
Pumps, 311.
Steam Turbine, Parsons', 161.
Steam Turbine, Tests of Parsons', ITS.
Strtam, Measurement of the Power of
a, 5.
of, 54.
143,
339
Turbine,
Turbine,
Turbine,
Turbine,
Turbine,
Inward-flow,
Mulhbach, 93.
Mr. Frederick, 101.
Outward-flow, 152.
Parsons' Steam, 161.
Turbine Pumps, Experiments with,
Nell's,
316.
83,
36.
133,
Schaffliausen, 41.
Steam,
161.
Terni, 137.
Tremont, 93.
Theory of, 56,
105, 181.
!K>,
137.
Turbine,
Turbine,
Turbine,
Turbine,
Turbine,
Turbine,
Turbine,
Turbine,
of, 69.
149.
Turbine, Vevey,
Turbine, Victor, 52,
Tremont, Turbines
Turbine,
Turbine,
Turbine,
Turbine,
Unwin's Formula
Vane Angle
Vane Angle,
at, 93.
Unwiu on
39,
239.
Vanes,
Vanes,
Vanes,
Vanes,
Vanes,
Appold, 244.
Centrifugal Pump, 239.
Construction of, 6s>.
Curves for Fourneyron,
Guibal Fan, 260.
72.
INDEX.
340
37.
87.
Changes
Sudden
of, 19.
Webber.
274.
143,
147, 157.
115.
Pump Efficiencies,
255.
5.
5.
DBfB of,
145.
CATALOGUE OF BOOKS
PUBLISHED BY OR TO BE OBTAINED FROM
London, W.C.
folio, bound
Workshop Cost Forms,
Specimen
Ltd.,
(Directors of Messrs.
Crown
Wh.Sc.,
Methods
of
&c.,
Author
By
of
ARTHUR
"Graphical
H.
ENGINEERBARKER,
Calculus,"
B.A.,
"Graphic
Third Edition.
Crown
anywhere.
full
FREDERICK GROVER,
Assoc.M.Inst.C.E.
Crown
THE CHEMISTRY
'
ENGINEERING.
of Metallurgy in the
By
A.
OF MATERIALS OF
HUMBOLDT SEXTON,
Price
5s. net,
post
CURRENT
CONTINUOUS
free.
DYNAMOS
to the
Professor
By W.
R.
AND
KELSEY,
Taunton Corporation.
Price
5s. net,
post free.
"
Inspector."
Just Published.
Second Edition.
THE MANUFACTURE
TUBES.
By
EDWARD
MARKS,
As8oc.M.In8t.C.E.,
Crown
PRACTICE.
JJODERN IRONFOUHDRY
Materials
and
Foundry Equipment,
Followed. By GEO. R. BALE,
Used,
Parti.:
Processes
A.soc.M.Inst.C.E.
By W. H. ATHERTON,
M.Sc.,
OF STEAM
and A. L.
MELLANBY,
M.Sc.
Price
3s.
Auoc.M.Inst.M.E.
net, post
fee anywhere.
DIAGRAM,
By
Practically
Applied
HENRY A. GOLDING,
Crown
Third Edition.
anywhere.
ITS
DIAGRAMS: WITH
CHARLES DAY,
This book
is
Wh.Sc.
numerous
Crown
Second Edition.
anywhere.
DROBLEMS
IN
MACHINE DESIGN.
Price
4s.
Second Eiition.
Crown
OF
Author
of
By
F.
G.
BURTON,
Crown
THE
trated.
Treatment
by
Graphic
Methods
of
the
Forces and
By W. W.
F.
PULLEN,
Wh.Sc.,
Crown
THE A B C
By W.
"
of the
DIFFERENTIAL CALCULUS.
WANSBROUGH,
D.
Movement
Proportions and
Author
"
of
Portable Engines,"
By
Mech. Engineer.
HALL,
Crown
BARKER,
A. H.
CUculus,"
Author
of
"
Graphical
&.c.
Crown
TIONS
Management
of
QUALIFICA-
Marine Engines,
OONSTANTINE,
Crown
DUTIES.
i:et, jioit
free
By
E. G.
anywhere.
Practically Applied
By W.
C.
Specially
POPPLEWELL,
M.Sc.
cloth, price
WORKING. ByW.
W.
F.
PULLEN,
Wh.Sc., A8Soc.M.InstC.E.
London
JOHW HnwooD,
all Booksellers.
This book
JAN
2 5
APR
1954
954
is
IStair
TJ
72
A 000316655