676

THE COLLIERY GUARDIAN

October 1, 1915.

CURRENT SCIENCE

Silica Brick for the By-Product Coke Oven.

The Transactions of the American Institute of Mining
Engineers contain a paper by Kenneth Seaver on the
manufacture of silica brick. In speaking of silica brick,
he refers to a brick having a silica content of 94 per
cent, or more and made usually from quartzite rock
with a small percentage of lime as a binder. Pennsyl-
vania manufactures probably from 75 to 85 per cent, of
all . the silica brick made in the United States. A
representative analysis of the Pennsylvania quartzite is
as follows:—Silica (SiO2), 97'80 per cent.; alumina
(A12O3), 0’90; ferric oxide (Pe2O3), 0’85 ; lime (CaO),
OTO; magnesia (MgO), 0T5; alkalies, 0*40; total,
100'00 per cent.

A typical analysis of Wisconsin quartzite, much used
in the manufacture of silica brick in the vicinity of
Chicago, follows:— Silica (SiO2), 97'15 per cent.;
alumina (A12O3), 1'00; ferric oxide (Pe3O3), 1'05; lime
(CaO), 010; magnesia (MgO), 0'25; alkalies, 0T0;
total, 99'65 per cent.

The Alabama quartzite used in silica brick manu-
factured in Birmingham, Ala., is found in the eastern
part of that State. A typical analysis follows :—Silica
(SiO2), 97’70 per cent.; alumina (A12O3), 0'96;. ferric
oxide (Fe2O3), 0’80; lime (CaO), 0 05; magnesia (MgO),
0'30 ; alkalies, 0'31; total, 100'12 per cent. In Colorado
and Montana both quartzite and sandstone are used to
some extent in the manufacture of silica brick.

The author next describes the method of manufacture.
During the grinding, 2 per cent, of lime, by weight, is
introduced as milk of lime. When the brick is burned,
the lime serves to give the necessary bond or physical
strength. The quality of the lime has no small effect
upon the resulting characteristics of the brick. As
regards ultimate refractoriness, no marked effect is
noticeable upon increasing the lime addition up to 3 per
cent. From that point on, the effect is clear in a
lowering of the fusion point. The general effect of
increasing the fineness of grind is to improve the
abrasive resistance of the brick, if it is properly burned.
The ground silica material as ready to mould is of about
the consistency of damp sand, but probably because of
the extreme angularity of its grains has less flow as it is
manipulated. As a consequence, the moulds must be
filled by pounding with heavy plank beaters covered
with metal, and all corners and shapes of irregular
outline must be rammed in by hand. Unequal ramming,
which leaves too great a variation in the density of the
green brick, will give unequal expansion in the kiln.
In view of the extreme accuracy required, steel moulds
are used exclusively for all high-grade coke-oven work.
After moulding, the shapes are placed on pallets, on
which the brick are dried either on the steam-heated
hot floor or on rack cars in some form of tunnel dryer.

The vast majority of silica brick are burned with soft
coal in the familiar type of round, down-draught kiln,
varying in diameter from 30 ft. upward. Fire is con-
tinued slowly and with slight acceleration for from nine
to 12 days, during the last two or three of which the
final temperature should be practically attained. The
kiln is then held at the required temperature, say at
26 cone (1,650 degs. Cent., or 3,002 degs. Fahr.) directly
in front of the firebox, for another day, after which the
fireboxes may be sealed and allowed to remain thus for
approximately 24 hours more. Cooling will require
about as long as the burning proper; but this varies
somewhat with atmospheric conditions.

In every burned silica brick expansion will occur on
heating. Such thermal or temporary expansion will
disappear upon cooling, and is not to be confused with
the permanent expansion produced in the body of the
brick during its manufacture. It is essential in the
construction of the coke ovens that the expansion,
whatever its nature, be as small as possible, and that,
whatever it is to be, it be known and adequately cared
for. The representative silica brick produced from the
Pennsylvania quartzite may be assumed to have the
following composition:—Silica (SiO2), 9625 per cent.;
alumina (A12O3), 0'88; ferric oxide (Fe2O3), 0'79; lime
(CaO), 1'80; magnesia (MgO), 0'14; alkalies, 0 39;
total, 100'25 per cent. What are the changes which
occur as the siliceous mass is burned ? Certain of these
are obvious, such for instance as the formation of a
bond of one or more of the lime-silica compounds such
as the meta-silicate CaO, SiO2 or the ortho-silicate
2CaO, SiO2. No less important, though less recognised,
changes are occurring in the brick as it is heated. The
evidence of the change, regardless of its reason, is easily
indicated by the most casual measurement when brick
are drawn from the kiln from time to time, as higher
temperatures are reached. The final expansion will be
very close to f in. per foot. The theory of this expan-
sion is that a change occurs in the crystalline form of
the silica, an inversion from the quartz crystal to the
crystalline form of cristobalite. The intermediate
crystalline form of tridymite does not apparently occur
here. Since these various crystal forms have varying
specific gravities, the volumes accordingly vary. Quartz
has a specific gravity of 2'65, tridymite of 2'27, and
cristobalite of 2'33. It is the difference between the
indices of refraction of these minerals that furnishes the
greatest aid in distinguishing the one from the other.

The following gives the results obtained by the
author:—

Quartz and		
	calcium	Cristobalite.
Specimen.	silicate.	
	Average,	Average,
	per cent.	per cent.
Lot 1 (1st burn) 		22'65	77'35
Lot 2 (2nd burn) 		17'13	82'87
Lot 3 (3rd burn) 		16'02	83'98
Quartz rock (1st burn) ..	50'57	48'95
Quartz rock (2nd burn)..	31'38	68'62

AND TECHNOLOGY.

The tests recorded in the above table showed that the
maximum change from quartz to cristobalite takes
place at the first burning, followed by a further appre-
ciable inversion at the second, and practically no change
at the third.

The fact, moreover, that no tridymite was deter-
minable in either the burned brick or the calcined quartz
is explained by the presence of the negligible amount of
flux. This would favour the direct inversion of quartz
to cristobalite. Thus far, we have assumed the changes
in crystal form to be desirable, because of their effect
on the expansion of the material.

In considering whether or not all the expansion
possible is burned out of the brick, it is necessary to
determine first the limits and then how closely they are
approached. Day, Sosman and Hostetter say* that the
true specific volume of cristobalite as compared with
quartz at 20 degs. Cent, is 13'4 per cent, greater.
Assuming a green brick to be wholly pure quartz
(100 per cent. SiO2), and the brick burned from it wholly
cristobalite, the volume increase would then be 13'4 per
cent, if the relative porosity of the body be disregarded.
This theoretically possible expansion is diminished by the
effect of the impurities present in the actual brick, and
by the resulting compounds formed. For instance,
4 per cent, of impurities would produce a much larger
proportion of compounds to which this expansion
phenomenon does not apply. An expansion of, say,
f to t76 inch per foot, to which by-product coke-oven
silica is burned, thus provides for somewhat more than
13 per cent, increase in volume. It must be borne in
mind, however, that by quick burning the brick may be
easily expanded by double or treble the normal amount;
but sound brick does not result. This excessive expansion
is due to minute fissures opened up in the body of the
brick by a too sudden rise in temperature during
burning. In the early stages of firing the heat must be
raised very slowly. In general, the conclusion is
justified that the present good commercial practice
carries the results of the quartz-cristobalite inversion to
the greatest degree economically possible, and that the
measure of the expansion of the brick affords a closely
approximate measure of the inversion—providing only
that the brick are sound.

There is much to be learned as to the effect of both
time and temperature as well as the influence of
pressure, which may produce a shifting of some of the
inversion points. Judging from the work thus far done,
it is impossible to expect the full inversion of quartz to
cristobalite under any practical operating conditions of
manufacture.

A New Type of Turbine Pump.

A new type of single stage pump suitable for dealing
with large quantities of water against medium heads
when driven at high number of revolutions for balanced
engines, and especially for a direct-coupled turbine
drive, has recently been designed and built by Messrs.
Henry Watson and Sons Limited, of Walker Gate,
Newcastle-on-Tyne. A number of impellers of small
diameter are fixed on a common shaft, and they work
in parallel on common suction and discharges. The
small diameter impellers allow of a high number of
revolutions without increase of periphery speed beyond
the medium head requirements, whilst the multiple
impeller inlets provide sufficient area for dealing with
large quantities of water. The pump efficiency is 78
per cent., and impeller efficiencies of 98 per cent, have
been obtained with pumps delivering up to 10,000 gals,
per minute against heads up to 40 ft, with pump shafts
running up to 4,000 revolutions per minute. The
driving turbine running at high speed gives its best
results as regards steam consumption. A special
feature of these “ Paraflo ” pumps is that by arrange-
ment of dummies, or valves, the pumps may be run at
reduced capacity output on a constant number of revo-
lutions and head with very little loss of efficiency.
The turbine-driven pump shafts and impellers are so
carefully balanced that with the pumps running at
4,000 revolutions per minute there is no vibration.
Starting tests have been carried out with the pump
placed 3 ft. above the water and delivering against 30 ft.
head. The time occupied from empty to full bore
discharge was 45 seconds. The testing plant at Messrs.
Watson’s works is of the calibrated test-weir type, with
raising and lowering static head arrangements. Trials
have been made with pumps delivering over 30,000 gals,
per minute against 35 ft. head. The frictional heads
were measured by mercurial gauges.

Utilisation of Peat.

A new plant for the utilisation of peat has been
lately installed at Codigoro, Italy, and a description of
it has been published in II Monitore Tecnico, a trans-
lation of the same appearing in the Journal of the
Canadian Peat Society. Early attempts to utilise the
peat of Codigoro have proved unsuccessful. It is now
decided to try the by-product recovery process by
gasification in a Mond plant. The whole process com-
prises a drying plant, a Mond sulphate of ammonia
plant, a briquetting and humus factory, and a central
electric plant. The quality of the peat enables a ton of
dry peat to be obtained from 3 to 3'5 cu. m. of excava-
tion. Excavation is done by hand, but a mechanical
excavator, capable of producing more than 600 cu. m. in
10 hours, is shortly to be installed. Drying is effected
by the products of combustion, which are blown through
the drying chambers by electric fans. By this means
200 tons of water can be evaporated daily. Six pro-
ducers are employed to obtain Mond gas, tar and
ammonium sulphate being separated. These gasify

* “ The Determination of Mineral and Rock Densities at
High Temperatures?’ American Journal of Science, fourth
series, vol. xxxvii., No. 217, pp. 1 to 39 (1914).

30 tons of peat in 24 hours. The drying operation
reduces the moisture content from 58 to 35 per cent,
before the peat is put into the producers. From 150 tons
of peat there can be produced daily about 270,000 cu. m.
of gas of about 1,400 calories. About 80 kilogs. of
ammonium sulphate is produced per ton of peat. A
sulphuric acid plant has been installed on the spot,
owing to the high cost of transport. Power is supplied
for the whole of the plant by two turbo-generators
worked by steam boilers. The generators are of the
Tasi-Oerlikon type, of 1,000 amperes, 225 volts, three-
phase, 50 cycles, 3,000 revolutions. Steam is used at a
pressure of 12 atmospheres superheated to 300 degs., the
exhaust being utilised for the gasification and also for the
concentration of the sulphate. Most of the machinery
has been made in Italy. The Codigoro plant resembles
that used at Oreptano since 1910 for working the peat
from Lake Brentina, in the province of Firenze. A
similar plant is at work at Osnabruck in Hanover. The
same journal describes the peat industry in Russia,
where 3,000,000 tons of air-dried peat fuel is now pro-
duced annually. The fuel is used entirely in manu-
facturing industries, cotton and weaving mills, glass,
iron and steel works, etc. Many machine plants are now
in operation, more than 50 being employed by the firm
of Morosoff. Son and Co., at Orehowo, near Moscow, and
the demand for these is rapidly spreading all over
Russia, where coal is scarce and expensive. Machine
peat is retailed in Moscow at 4'80 dols. per ton.

THE GERMAN AND AUSTRIAN COAL
AND IRON TRADES.

We give below further extracts from German
periodicals that have reached us, showing the course of
the coal and iron trades in Germany and Austria :—

A private message to the Dag ens Nyheter (Stockholm)
from Berlin states that the fuel question is being daily
discussed in the newspapers. Strong fears are enter-
tained of a serious scarcity. In West Berlin, where
common domestic heating arrangements exist, the posi-
tion is exceedingly precarious. The production of coke
is greatly hampered by the scarcity of labour, and trans-
port facilities for the same reason are lacking. The
newspapers recommend that all export of coke should
be stopped.

German Steel Union Report for August.

The total deliveries for August amounted to 250,080
tons of raw steel (258,092 tons in July, and 94,984 tons
in August 1914), of which 59,303 tons (61,768 tons and
15,165 tons) were semi-manufactured goods, 120,057
tons (118,737 tons and 61,390 tons) railway super-
structural material, and 70,729 tons (77,587 tons and
18,429 tons) sections.

The Interim Coal Syndicate.

At a recent coal owners’ meeting, the adhesion
of the Thyssen group to the proposed agreement was
announced, and the interim Syndicate definitely estab-
lished for a period terminating on March 31, 1917. The
only concerns which did not come into line with the
rest were the Admiral and Gluckaufsegen collieries,
which, however, are not very important, the former
having a coal output of only about 112,000 tons, and
the latter 352,000 tons per annum, whilst neither has
been very successful financially.

The Mid-German Brown Coal Market in August.

The Browm Coal Association, Halle, reports that
business during August was generally good, and up to
the average of the previous month. The scarcity of
labour has, however, been intensified by the continued
drain for army service, both skilled miners, labourers,
and the supervising staff having been affected; and the
gaps have not been adequately filled by the substitutes
employed. Work has been found for a large number of
women, and extensive use has been made of the per-
mission to work overtime; but these measures have
failed to bring the output up to the level of the demand.
Increased pay and allowances have been granted to a
considerable extent.

The Luxemburg Iron Industry.

The annual report of the Luxemburg Chamber of
Commerce states that the number of ore mining con-
cerns operating in 1914 amounted to 77 (86 in 1913),
employing 4,391 (5,807) hands and producing 5,007,457
tons (7,333,372 tons) of minette ore, of a total value of
15,826,514 fr. (21,965,818 fr.), or an average of 3'16 fr.
(2'99 fr.) per ton. After poor spring and summer busi-
ness, all the ore mines closed down entirely as soon as
war commenced, 4,122 men being discharged; but in
September and October work was gradually resumed,
and by the end of the year about 80 per cent, of the
norm<4 number of miners were again employed. The
war forced up prices of all ore sold outside existing con-
tracts, good calcareous minette rising from 35 fr. to
37'50 fr., inferior grades from 30 fr. to 35 fr., and waste
ore from 25 fr. to 30 fr., the average rise of calcareous
ore being 10 per cent., and of the siliceous ore 15 to 16
per cent. The exports to Germany totalled 494,368
tons (1,887,245 tons) of minette, and 23,467 tons (73,877
tons) of ground basic slag. Manganese ore to the extent
of 99,084 tons was imported, 67,231 tons being from
Russia, and 31,850 tons from British India. The local
iron works consumed 6,137,609 tons (8,656,670 tons) of
ore. The military requisition of the railways suspended
the supply of coke to the blastfurnaces, so that all but
eight out of 50 had to close down early in August; and
in September there were only seven at work. The pig
iron output sank from 233,172 tons in July to 45,167
tons in August, and 29,960 tons in September; but
increased again to 67,406 tons in October, 95,928 tons
in November, and 111,523 tons in December (about half
the normal average), since when the upward movement
has been maintained. The production by the blast-