January 18, 1918.

THE COLLIERY GUARDIAN

125

s
p
Mn

Low S.
Per cent,
max.
0’05
0 20
0*60

High S.
Per cent,
max.
0*25
0*20
0*40

In both cases the sulphur would be increased on good
melting by 0*02 per cent, in the crucible and by 0-03
per cent, in the cupola, and the manganese would be
slightly reduced; in the case of the low sulphur iron
leaving a slight excess of manganese, which would
exert a malleable influence, and generally tend to
solidity of casting, as well as give greater fluidity in
the molten metal, whilst not materially—if at all—
affecting the annealing, except to assist in keeping the
grain or fracture close.

Such a form of grading would be of material assist-
ance to the ironfounder because, if the sulphur, phos-
phorus, and manganese were fairly constant and a
maximum percentage, the only thing which need be
considered by him is silicon, for the total carbon—
particularly in cupula melting—can be left to take care
of itself, as this will not materially vary in any parti-
cular brand of iron.

Under the above form of grading, supposing a
foundry using crucibles found grade lb suited best;
then mixing in equal proportions grades la and 1c
would give them grade lb. This is illustrated as
follows: —

Grade la	say Si 0*20 %	to	0*40 %	fracture	white.

„ 1c	„	Si 0*60 %	to	0*80%

2 0*80%	to	1*20%

------—-------------

Grade lb	„	0*40 %	to	0’60 %	fracture	white.

Again, what would be an equally satisfactory mix-
ture, and would be almost sure to produce a^i equally
good white fracture, would be: —

Grade la say Si 0’20 %	to 0’40 % gracture white.

„ 2a „ Si 0*40%	to 0 60%	„ hard mottled

2 I <r60% to 1’00%

0’30 % to 0’50 % fracture white.

The fact that the fracture of grade 2a has such a
low silicon indicates that the total carbon is higher
than grades la, lb, and 1c, and therefore the risk of a
slightly lower silicon in the casting is allowable. This
illustrates the value of fracture as a guide to total
carbon contents, and it is sufficiently accurate for all
practical purposes.

The same method of mixing would not do for use
on cupolas. Grade la would be, and lb might be, too
hard to use alone, because the silicon might be too low
for good annealing, for, unlike crucible melting, the
cupola materially hardens any grade of iron on melting.
For example, a very hard mottled iron—nearly spotted
white, and low in silicon—might even turn white if
melted in the cupola and cast into the1 same section—
say, 2 in. by 2 in. pig—but if the pig were 3 in. by
3 in., the fracture would be likely to be spotted white.
One reason of this is that the percentage of silicon in
the pig is reduced on melting the iron in the cupola by
nearly 0-25 per cent., and hence the solidification of the
molten metal is liable to be much quicker by being
chilled, and in mottled grades of iron, the quicker the
cooling the whiter the iron. Therefore, the following
mixing would be suitable: —

(1)	1 c cotild be used alone.

(2)	la say Si.	0’40% to 0’60% fracture white.,

2b ,, Si.	0'60% to 0*90%	,, hard mottled.

2 | 1’00% to 1’50%

0*50% to 0*75% = 0’625%

2

Deduct oxidation loss 0*20%

Contents of casting, about silicon ... 0’425%

(3)	In the same way, 2a and 2b could be mixed in equal
quantities, and give an average hard mottled.

(4)	Also lb and 2a, and so on.

Analyses of Castings.

Fairly typical analyses for the average size casting
are as follow :

Low sulphur. High sulphur.

Per cent.	Per cent.

Silicon .... 0’50	to 0’60	...	0*50 to 0*50

Sulphur.....	0*06	...	0*25

Phosphorus ..	0*15	...	0’15

Manganese ...	0*40	...	0’20

In ordinary every-day practice, however, silicons in
castings usually vary from perhaps 0-50 to 0’80 per
cent., and hence, on the same section of casting, silicon
0*80 per cent, would anneal easier than 0-50 per cent.
On the same sections this would not make material
difference, although it would be better if silicons could
be more regular. Again, if the silicon were constant,
and the thickness or sizes of castings varied somewhat,
this again would not make material difference, for in
both cases the difficulty can be and is largely overcome
by placing the larger castings in the annealing oven in
the region of the greater heat. But where the difficulty
may, and does arise, is chiefly if the large castings
contain less silicon than the smaller, as in this case
there is the risk of either under-annealing the larger
castings, or of over-annealing the latter, when both are
annealed together. Even with Siemens and air fur-
naces there are exactly the same difficulties, but as the
outputs are so large, it is the custom to anneal in one
oven the castings nearest in size, and consequently the
only risk is varying silicon in the same size castings.

Dr. E. G. D. Menzies has been appointed certifying sur-
geon under the Factory and Workshop Acts for Wolsing-
ham district, Durham. Dr. F. J. MacManus has been
similarly appointed for Killeshandra, co. Cavan. A vacancy
exists at Ramsgate.

COAL WASHING.*

By T. J. Drakeley.

At the present time coal washing is undertaken in a
perfunctory manner in order to produce a coal that is
saleable or suitable for coking purposes. Beyond
that, no further particular interest is taken in the
operation, except a casual inspection of the refuse to
see that there is no abnormal waste. However,
under scientific control the washing plants can be
made to produce more perfectly cleaned coals and to
work with stricter economy.

The Theory of Coal Washing.

Coal, as it arrives at the surface, is always
associated with quantities of dirt. It was found
that the extraneous matter could be eliminated
by an elaborate washing process, and that opera-
tion was attended with such success in Germany
(where there is much dirty coal), that British officials
readily adopted the method, where the production of
a clean coal was a matter of urgency. The washing
process depends on the difference in the velocities witn
which identically shaped particles of various specific
gravities settle in the water. If a piece of coal and
one of shale, of approximately the same shape, were
dropped into water, the coal, having a smaller specific
gravity than the shale, falls at a slower rate through
rhe medium. By adjusting the velocity of the
stream of liquid used for washing and the length of
the trough, so that the light particles will be carried
to the end of the trough without touching the bed,
whilst the shale, by contact with the bed, will be
retained in the trough by friction, it is possible to
effect a separation of the light particles of coal from
the denser shale particles. Furthermore the motion
of the liquid in close proximity to the bed is slower
than that of the main stream, and that tends to
make the retention of the shale particles more pro-
nounced. Of course, if suitable dams are arranged
the trough they will retain the heavier particles of
refuse, whereas the lighter coal will be carried for-
ward by the stream of liquid.

The Practice of Washing.

Coal washers may be classified into: Trough
washers, barrel washers, bumping-table washers,
inverted, cone washers, and jig washers, the latter
being subdivided into: Nut washers (refuse removed
at ends of wet screen; no felspar), ielspai? washers
(refuse removed through wet screen), plunger move-
ment, screen movement, and compressed air move-
ment. Investigation has not expended sufficiently
lor preference to be given to any particular system of
wasning. .Each type seems to be designed to meet
definite requirements, and therefore favour should be
given to tne washer which most nearly fulfils the
essential conditions.

f ±t would appear that the best results of washing a
(friable coal are obtained with jig washers. In the
Robinson washer, the trituration of the material near
the apex of the cone is often considerable, and should
the dirt in the raw coal be of a friable nature, it is
deduced to powder. In that state it is carried back
from the base of the washer and eventually finds its
exit with the washed coal. As an example of pro-
nounced levigation in a Robinson washer, raw coal of
0-lf in. diameter was fed into the washer, but owing
to the friability of the dirt and its tendency to
disintegrate in water the sizes of the refuse particles
were less than V16in. The levigation of the materials
in travelling down a trough exceeds that experienced
in jig washing, but is considerably less than in the
Robinson washer. The choice of a washer is an
extremely difficult matter. As a rule a system which
is effective at one colliery will not produce the desired
results if it be adopted in its entirety at another
colliery. At each colliery there are peculiarities
which demand special treatment. Since the decid-
ing factor in the selection of a suitable washer
should be the physical and chemical characteristics of
the raw coal, it is a matter of the utmost importance
that the properties of the raw material should be
^subjected to a minute and careful study.

\	Sizing.

The theoretical deductions concerning the sizing of
the raw material have shown that the ratio of the
diameters of the perforations in the classifying
screens may vary from 4 to 30. However, in actual
practice, the many forms of the raw material, ranging
from almost perfect cubes to pieces having round,
oblong or triangular sections and widely differing
thicknesses, make it almost impossible to build a jig
which will separate efficiently that great variety. In
view of those difficulties, the ratio of the diameters
of the perforations in the classifying screens is usually
about 2. Such close sizing would appear to be
unnecessary. In washing fine coal in a felspar washer
the felspar is of comparatively large size and the fine
coal is concentrated among the larger material,
j Therefore it would seem that that larger material
| may well be raw coal, which is itself undergoing con-
I centration, and so a modification in the washer would
\be necessary.

Drainage of Washed Coal.

The recovery of the coal is accomplished by either
a settling tank or a bin filter.

Settling Tank.—The coal and water are led to a
tank by sluices. On entering the tank the velocity
of the water is checked and the material settles to
the bottom of the tank. A suitable conveyor removes
the coal from the tank. There are several objections
to that method which are obviated to a considerable
degree by a second method. The first objection is
that although the velocity of the water is reduced
there is still sufficient movement for the water to
carry away in suspension a quantity of the finest coal.
If the water has to be used again, that slime is better
removed. The second objection is that of handling

* From a paper read before the Manchester Geological
and Mining Society on Tuesday, January 15.

wet material with elevators, etc. The final and
strongest objection to the settling tank is the excep-
tional difficulty presented in withdrawing the finer
coal from the bottom of the tank, the coal arching
over scrapers and so escaping removal.

• Bin Filter.—Probably the best method is to lead the
coal and water by sluices to bins. The water drains
from the bin whilst the coal forms a bed through
which the rest of the washing water filters. In that
way even the finest coal is recovered and a compara-
tively clean water drained from the bin. The difficulty
in connection with that method is that the wet coal
packs itself into a firm solid mass, which can be
removed from the bin only by powerful mechanical
means. As the drainage of a bin takes from 20 to 40
hours, two bins are required of such a size that whilst
one fills the other can be allowed to drain. The cost
of the bins is high but the recovery is efficient. A
modification of the bin recovery is that applied in the
Baum washer, whereby the water is made to drain
through small bins containing fine washed coal. The
bins are attached to an endless band.

Control of Washers.

Considerably more attention ought to be paid to the
control of washers. Undoubtedly at many collieries,
regular float and sink tests are not made, but it is
only upon the interpretation of the results of such
test that one can base arguments in favour of any
variation in the working of the washers. That these
tests should be made is recognised generally, especially
where jig washers are used, as their control intro-
duces many factors. With trough and inverted cone
washers .the freedom of control is more restricted.
The control exercised over jig washers usually consists
in regulating the supply of coal, the supply of water,
the stroke and-speed of the eccentric action, and the
outflow of the refuse.

Supply of Raw Coal.—If the raw coal is fed too
rapidly on to the wet screen, no opportunity is given
for any considerable separation to occur before the
material is compelled to leave the washer. That
means that both the percentage of dirt (sink particles)
in the washed coal and the loss of coal in the refuse
are high. On the other hand, should the supply
of raw coal be too slow, only the best coal is carried
away as washed coal. The usual openings of the
refuse valves release a refuse which necessarily con-
tains a large percentage of coal.

Supply of Water.—If water is pumped into the
washer too rapidly it is capable of flushing away in
the washed coal an unduly large quantity of dirt, but
as only the heavier parts of the material could sink
through the rapid stream of water to the screen, the
refuse contains little coal. The trouble introduced by
admitting too little water is that at each upstroke of
the plunger there is a strong suction, in the washing
box, that tends to arrange the particles, with the
lightest near the screen. In consequence the washed
coal is composed largely of dirt, and coal forms a
considerable portion of the refuse.

Stroke and Speed of Eccentric Action.

(а)	Long and Rapid Stroke.—The more violently
the downward stroke of the plunger is made the
greater is the distance that each particle is separated
from its neighbour. Therefore, a better opportunity
is offered for the materials to settle in accordance with
their respective specific gravities. However a rapid
upward pulsation in the washing box necessitates
a rapid discharge of water from the box, and it
has been observed that too strong a stream of
water does not result in effective separation. With a
long and rapid down stroke it is essential for the
machinery to provide for a somewhat slower upstroke
of the plunger, otherwise the admission of water to
prevent a suction at the wet screen, must be very
large, and in consequence the ill-effect of a strong
current of water is aggravated.

(б)	Long and Moderately Slow Stroke.—With such
an arangement, especially where provision is made
for a slower upstroke of the plunger, the best condi-
tions are obtained for washing raw nut cpal. The
long but moderately slow downstroke of the plunger
produces just sufficient agitation for the materials to
settle freely during the slower up-stroke of the plunger.
As water is admitted below the screen, the settling
occurs in almost still water. For fine coal in a felspar
washer even a long and moderately slow stroke is not
essential, as the agitation is unnecessarily great and
the rate at which the coal is washed is very limited.
Where fine coal is being washed, with larger sizes, a
long and somewhat slow stroke is essential. Even then
the rapidity with which the stroke must be made to
agitate the larger sizes causes such a rush of water
from the box that fine dirt is carried away with the
coal. Hence, where unsized coal is washed and subse-
quently classified, the fine washed coal is improved, as a
rule, by a second washing. Obviously if the stroke is
too slow, insufficient freedom is produced in the wash-
ing box and the plant delivers, from the washed coal
and refuse outlets, products differing very little from
the original raw coal.

(c)	Short and Rapid Stroke.—A. short stroke, even
though rapid, does not produce sufficient freedom
among the particles of nut coal to secure good separa-
tion. On the other hand, a short and fairly rapid
stroke is admirable for the fine coal. The short stroke
brought about sufficient freedom for the particles to
settle easily during the up-stroke of the plunger. It
is essential that water should be admitted to prevent
a suction at the wet screen; otherwise, not only do the
materials settle in the reverse order, but a large
quantity of fine stuff is sucked through the felspar bed
and lost as refuse.

(d)	Short and Slow Stroke.—A short and slow stroke
would probably produce no movement in a washing bed
composed of large particles, and even with fine coal
it would be unsatisfactory, as insufficient separation
would be effected.

Shale Outflow.—The shale outflow must be so regu-
lated that it corresponds with the rate of accumula-
tion of the dirt. If too slow, the dirt passes out in