THE COLLIERY GUARDIAN

AND JOURNAL OF THE COAL AND IRON TRADES.

Vol. OVII.

FRIDAY, JANUARY 16, 1914.

No. 2768.

The Testing of Air Compressors.

By H. KE AY PRATT, M.I.Mech.E.

(Continued from, page 78.)

II.

The following is a description of the arrangement
necessary to make a complete test of a steam-driven
air compressor. After the machine is fixed in its
position on the test plate, the air discharge pipe is
connected up to the pressure tank D. This tank is
then coupled to the two measuring tanks E E by means
of a swing-over valve, as seen in fig. 4. The steam
cylinders must be coupled up to the condensing plant
seen at B, and the air-pump discharge should be led to

Fig. 4.

Reference .—A = Air Compressor ; B = Condenser ; C =
Weighing Machine; D = Pressure Tank; EE = Measuring
Tanks; H'FF = Pressure Gauges; GG = Safety Valves ;
HHH = Thermometers ; J J = Air Inlet; KK = Circulating-
Water Tanks; L = Thermometer for Air; M = Throw-over
Valve for Air ; N = Throw-over Valve for Water; OOO =
Stop-valves.

the weighbridge C. Pressure gauges must be fitted
to the air tanks D and E, and thermometers must be
fixed in them. The temperature must also be taken of
the cooling water as it enters the air-cylinder jackets,
and also as it leaves them. The tanks K K must be
coupled up to the jacket circulating water discharge, and
another throw-over valve will be required here. A test
should extend over a considerable period, and during
this time the following readings must be taken regu-
larly, about every 15 or 20 minutes :—

Steam pressure.

Vacuum if working on condenser.

Revolutions per minute.

Air pressure in receiver D.

Temperature of air in receiver D.

It is very bad practice, and also untidy, to use any
odd sheets of paper for making tests. If air-compressor
tests are frequently undertaken, as in a works where
they are built, properly printed forms should be
obtained, which should subsequently be filed, so that if
any dispute arises the actual original test sheets can be
produced. Fig. 5 illustrates a suitable form upon
which these data can be recorded. This reliability test
should continue for about six hours, and during this
time two or three tests should be taken to determine the
volumetric efficiency and steam consumption, also the
consumption of circulating water in the jackets of the
air cylinders and intercooler. Sometimes the three lots
of cooling water are measured separately, but it is
generally sufficient if they are all allowed to flow into

the same tank, and are measured together. The first
efficiency test should be taken about two hours from the
start of the run. By this time it will be found that the
temperatures of the jacket water and air-measuring
tanks will have become fairly steady.

The person in charge of the test will require the
following assistants :—

One, at least, to indicate steam and air cylinders.

One in charge of the weighbridge.

One to measure the jacket water.

One to control the air-measuring tanks.

The chief himself will find plenty to do in watching his
assistants, and in noting the time and the revolutions.

When it is decided to commence the test, each
assistant should be in his place. The pressure in D
should be nicely regulated to that at which it is intended
to take the test, and the stop valves of both measuring
tanks should be firmly closed, and the safety valves
held wide open, as in the rough test already described.
The throw-over valve should be arranged to com-
municate with the measuring tanks, into which
it has been decided to start pumping first. The
assistant at the weighbridge will have to keep the
measuring tank balanced, and must start as soon
after the main test starts as he is able. The
cocks at the foot of the jacket water measuring tanks
should be wide open and the throw-over valve correctly

LOG SHEET.

Date

Steam pressure.	Vacuum.	Revolutions per minute.	Air pressure in receiver.	1  Temp erature in receiver.	Remarks.
		i			

Fig. 5.

set, and also the water in the gauge glasses ought to
stand at zero, as already explained. Then a note should
be taken of the number registered on the revolution
counter. Now, at a given signal, the safety valves on
the air tanks must be dropped, the cocks on the jacket
water-measuring tanks closed, the revolution counter
clutch thrown in, and the time noted on the stop watch.
The first air-measuring tank will now be filling, and
when the pressure reaches the correct working pressure,
the throw-over valve must be immediately turned, so as
to guide the air into the second tank. Meanwhile, the
temperature of the first tank should have been taken,
and then the air allowed to escape by raising the safety
valve, so as to get the tank ready to be filled again as
soon as No. 2 tank is full. These measuring tanks will
be filled alternately throughout the test, taking care to
note the temperature and number of the tank each
time. It is very necessary to see that the pressure in D
is kept constant. A form suitable for noting these data
is shown in fig. 6, which may be extended to include
those of the circulating water, as measured in the
calibrated tanks. Meanwhile, the assistant at the weigh-
bridge should have started to measure the discharge
from the air pumps. This test for steam consumption
should last for 20 or 25 minutes, so that, if the volumetric
test is allowed to continue from 30 to 35 minutes, there
will be ample time to get this test started' and com-

pleted, while the volumetric test is being taken. The
form necessary for the steam consumption test is
illustrated in fig. 7, and the method of filling this up is
explained in the following lines. To start the con-
sumption trial the water in the tank on the weighbridge
should just overbalance the weight on the scale beam.
When this takes place the electric bell will ring, and at
this moment a stop watch must be set going, and the
weight on the scale beam noted in the proper column.
This weight must now be moved on a certain number of
pounds, so that the time interval in the third column is
from one to one and a-half minutes, and each time the
scale beam rises the electric bell will ring. At this
signal the weight and time must be noted and the time
interval filled in, which interval is found by subtracting
from the last time reading the one immediately
preceding it. A few imaginary figures have been given
in fig. 7, and an examination of these will clearly show
how to use this form. This method has the great
advantage that shoul d the test have to be stopped for any
unforeseen cause, it is not utterly thrown away, as if only
a few readings have been taken and the time intervals
agree fairly closely, say within three seconds of each
other, the steam consumption may be calculated from the
figures obtained, with a certainty of it being fairly
correct. While these measurements are being taken
both the air and steam cylinders must be indicated every
five minutes, and on the cards must be noted the neces-
sary particulars. The person in charge of the test
should signal to the assistant working the indicators
when to take the diagram. He should also take the
revolutions per minute of the compressor, by means of
a hand counter, at the time the diagrams are obtained
and should collect the latter from his assistant and
figure them up.

The same care must be exercised on stopping the test
as in starting it. When it is decided to end the test a
signal must be given, exactly at the moment that the
pressure in one of the measuring tanks reaches the
working pressure. At the same instant a last reading
should be taken of the scale on the jacket water-
measuring tanks, the revolution counter should be
thrown out of gear, and the watch used for timing the
test must be stopped. From the data obtained the
various efficiencies can be worked out.

Isothermal Compression.

Perfect compression should be isothermal—that is,
the temperature of the air in the cylinder should remain
constant. It is the aim of designers of compressors to
arrange these machines so that the cooling arrange-
ments are such that the compression is as near
isothermal as possible. The nearer to this condition
the compression becomes, the less power is required to
drive the machine. The work in foot-pounds required
to compress and deliver 1 cubic foot of air isothermally
is given by the formula

Wj = P' V' ioge 0))

where Wx = work in foot-pounds.

P' = pressure in pounds per square foot.

P" _ highey pressure in pounds per square foot.
V' — volume at P'.

Therefore the work required to compress and deliver
1 cubic foot isothermally, from atmospheric pressure to
70 lb. per square inch, is,

Wi = 14-7 x 144 x 1 x logs f 70 + 14'7 x 14

>	\ 14*7 x 144 /

= 2,117 x 1 x loge 5-76

= 2,117 X 1 X 1 75	= 3,700 foot-pounds.

It is sometimes useful to calculate the efficiency of
compression in each cylinder separately. If this is
desired, the suction and delivery pressure must be scaled
off the indicator card, and the pressures so obtained,
plus the pressure of the atmosphere, must be used when
making calculations.

When scaling the indicator diagrams, the pressure to
be taken is that immediately at the end of the com-