and to a certain extent even is still, to remelt the metal in what is
termed a _finery furnace_, a kind of forge in which a charcoal fire is
urged by a cold blast, and so regulated that an excess of oxygen is
supplied, or rather more than would suffice to convert all the carbon of
the fuel into carbonic acid; although this is perhaps not absolutely
necessary, as carbonic acid would itself supply oxygen by suffering
reduction to carbonic oxide. At any rate the melted metal is exposed to
an oxidizing atmosphere and constantly stirred. Many different
arrangements of the furnace and details of the process have been used.
For instance, where the finest quality of malleable iron was not aimed
at, coke has been the fuel employed, and many shapes of furnaces, etc.,
have been contrived, and various additions of ores, oxides, etc., made
to the charge, according to local practice and the nature of the crude
iron. One marked effect of the operation is the final removal of nearly
all the _silicon_, which is burnt or oxidized into _silica_, and this at
once unites with oxide of iron, which is also formed, to produce a
readily fusible slag of silicate of iron, and in the production of this
silicate any sand attached to the pig will also take part. Much of the
carbon, amounting sometimes to more than half, is also eliminated as
carbonic oxide, and of what is left but little remains in the graphitic
state. The action on the phosphorus is usually less marked, but there is
always a notable reduction of the quantity. The sulphur is also lessened
in some degree, although when coke is used, the fuel has the
disadvantage of itself containing sulphur, phosphates, and other
deleterious matters. Sometimes a little lime is added to the charge to
take up the sulphur from the coke. The operation lasts some hours, the
fused metal being frequently stirred with an iron rod, until it assumes
a pasty granular condition, when the workman gradually collects it upon
the end of the rod into a ball of about three-quarters of a cwt. in
weight. These balls, or _blooms_ as they are called, are removed from
the furnace while still intensely hot, and at once submitted to powerful
pressure by means of some suitable mechanical arrangement, the effect
being to squeeze out the liquid slag and force the particles of metal
together by which the whole becomes partially welded into a more compact
mass. Then this mass is, while still hot, either hammered with gradually
increased force of the strokes, or in the more modern practice, passed
between iron rollers (these we shall presently describe), by which it is
shaped into a bar. The bars are afterwards cut into lengths, reheated
without contact of fuel, again hammered or re-rolled; and this process
is several times repeated when the best product is required. During the
first treatment of the blooms, and also in the subsequent hammering or
rolling, the oxygen of the atmosphere acts on the surface of the glowing
metal, so as to cover it with thin scales of oxide, and these, carried
into the interior of the mass, will give up their oxygen to any residual
silicon, carbon, etc., producing a little more slag, carbonic oxide,
phosphate of iron, etc., which by the pressure of the hammers or rolls
are ultimately forced out of the metal. It will be observed that in
producing the pig iron the chemical action is the separation of oxygen
from the metal, while conversely an oxidizing action is set up in the
finery and subsequent treatment, in order to burn off the foreign
ingredients. But this cannot be done without at the same time
re-oxidizing some of the iron itself, of which therefore there is always
a considerable loss, by its formation into slag (silicate), cinder,
foundry scale (oxide), etc. The quantity of iron lost depends of course
on many conditions, such as the care exercised in the operations, but it
occurs in all the processes that have been devised for the conversion in
question, even in the most modern: its amount may be taken to range
between 10 and 20 per cent. The reader is requested to bear in mind the
nature of the chemical actions that have just been described, for in
even the most recently invented processes the principle is the same in
nature and effect. So completely can the foreign elements be eliminated
by this, or some analogous process, such as we shall presently mention,
that the finest Swedish bar iron contains more than 99½ per cent. of the
metal, and in some cases only a very little carbon and a mere trace of
phosphorus remain, amounting together to less than 1 part in 2000. Such
metal is made from very pure ore, containing no sulphur and scarcely any
phosphorus, while charcoal is the fuel used in all the operations. As
already mentioned, the objection to the use of coke is the sulphur,
phosphates, and siliceous matters it contains. Toward the close of the
eighteenth century an invention came into use which obviated the
disadvantages of the cheaper fuel for converting crude iron. This was
the puddling furnace, brought into use after much experimenting by Henry
Cort in 1784. In it the pig iron is fused in a _reverberatory furnace_,
the form of which will be understood from Fig. 19, which is a diagram
showing such a furnace in section, where _f_ is the fire, _a_ an
aperture at which the fuel is introduced, _p_ the ash pit, _b_ is a low
wall of refractory material called the “bridge,” over which the flame
passes, and is by the low arched roof reflected or _reverberated_
downwards upon the charge, _c_, which is laid on a _hearth_, or iron
floor, having spaces below it where air circulates in order to prevent
it becoming too hot. In Cort’s original arrangement the bed of the
hearth was formed of sand, which gave rise to much inconvenience by
producing a quantity of the very fusible silicate of iron, that speedily
attacked the masonry of the furnace, and therefore a very important
improvement was devised some years later by S. B. Rogers, who made the
bed of his furnace of a layer of oxide of iron, spread on a cast iron
plate 1½ inches thick. In later times it has become usual to cover the
iron hearth with certain other refractory mixtures varied according to
circumstances, of oxide, ore, cinder, lime, etc. There is one of these
mixtures significantly designated “_bull-dog_” by the workmen. We may
mention here that it has, in more recent times, when very high
temperatures are obtainable, been found unnecessary to cause even the
flame to come into contact with the substances on the hearth, inasmuch
as the heat radiated from the flame and the intensely heated roof of the
furnace suffices, so that in consequence of this the roofs are now
constructed nearly flat. In the puddling furnace the melted metal is
constantly stirred, and no little skill is required to regulate the fire
by the damper on the chimney, and to admit the proper amount of air to
mix with the flame. The pig iron softens and melts gradually, until at
length it becomes perfectly liquid, at which stage it swells up and
appears to boil owing to the escape of carbonic oxide in numerous jets,
which burn with the characteristic pale blue flame. The puddler then
briskly stirs the mass to cause more complete oxidation of the carbon,
silicon, etc., by bringing the superficially formed oxide of iron into
the interior. As the iron loses its carbon, it assumes much the texture
of porridge, consisting of pasty lumps of malleable iron implexed with
the liquid slag (silicate of iron, etc.) which drips from the spongy
balls as the puddler collects them at the end of his stirring rod, as in
the finery operation. The next thing is to run the mass immediately
between powerful rolls (puddling rolls) by which the slag is squeezed
out, as before, and finally through the _finishing rolls_ that shape it
into bars or plates.
When a comparatively impure pig iron is used or when a better quality of
malleable metal is desired, the crude iron is submitted to a preliminary
treatment before puddling. This treatment, by a technical distinction,
called _refinery_, is practically identical with the _finery_ process
already described, except that instead of being collected into blooms,
the fluid metal is run out to form a layer 2 or 3 inches thick, and
this, before becoming quite solid, is suddenly cooled by having water
thrown over it, the result being a white, hard, brittle mass, which
broken into pieces is ready for the puddling furnace.
The operation that has been described is known as _hand puddling_, in
contradistinction to later methods in which it has been sought to
substitute some form of machine that will produce the same result
automatically, such as revolving furnaces, etc. It has been found
difficult to maintain these in good working order, and in England at
least mechanical puddling has never found much favour, but in the great
iron works of Creusot, in France, large revolving furnaces were in use
about 1880, which could turn out 20 tons of converted iron in 24 hours,
whereas the old hand puddling furnaces could in the same period produce
only 2½ or 3 tons, with two sets of men, the puddler and one assistant.
Of these mechanical furnaces it is unnecessary to give any account,
especially as the puddling process itself has nearly gone out of use,
having been superseded by more economical methods.
The use of rolls for treating the product of the puddling furnace, and
for making it into bars, was also an invention of Henry Cort’s, for
which he obtained a patent in 1783. This was in many respects an immense
improvement on the older system of hammering; it is still practised, and
by it shapes can be given to the metal scarcely possible on the older
system, while the tenacity of the metal is increased by the uniformity
given to the grain. The difference of chemical composition between cast
and wrought iron the reader has already been made acquainted with, and
there is quite as great a difference in their textures. The former, when
broken across, shows a distinctly crystalline structure, which we may
compare to that of loaf-sugar, while the latter exhibits grain, not
unlike that of a piece of wood. This fibrous structure depends upon the
mechanical treatment of the iron, and in rolled bars the fibres always
arrange themselves parallel to the length of the bar. Fig. 20 shows this
fibrous structure in a piece of iron where a portion has been wrenched
off. Like wood, wrought iron has much greater tenacity along the fibres
than across them; that is, a much less force is required to tear the
fibres asunder than to break them transversely. Consequently, to obtain
the greatest advantage from the strength of wrought iron, the metal must
be so applied that the chief force may act upon it in the direction of
the fibres. Near the beginning of our article on IRON BRIDGES (_q.v._)
the reader will find some illustrations of the very different resisting
powers of cast and wrought iron.
[Illustration:
FIG. 20.—_Fibrous Fracture of Wrought Iron._
]
Nothing in the way of inventions can be compared to those of Cort’s as
to the effect they have had in promoting the iron industry, until we
reach a period some years after the middle of our century; but we must
not neglect to recognize the scarcely inferior importance of Rogers’
improvement. Singularly enough, neither of these men reaped any benefit
from his inventions. Cort died in the last year of the eighteenth
century, quite a poor man, having been supported only by a niggardly
pension of some £160 from the Government, and leaving his family in
indigent circumstances. Yet a most eminent authority on iron questions
(Sir W. Fairbairn) estimated—some time about the middle of our era—that
the two inventions of Cort’s alone, the rolling-mill and the
reverberatory puddling furnace, had by that time added to the wealth of
Great Britain by an amount equivalent to six hundred million pounds
sterling. For many iron-masters had profited by these inventions,
amassing very great fortunes, in some instances also acquiring titles of
honour. Clearly to Cort and Rogers may be applied the _sic vos non
vobis_ saying.
We shall now turn to the improvements that have been effected in the
blast furnace, and of these none perhaps has been more marked than that
made by Neilson, when in 1828 he substituted heated air for the ordinary
cold air that had before always supplied the blast. It will be
remembered that the heat is due to the combination of only the oxygen of
the air with the carbon of the coke, but the greater part of the air—the
four-fifths of nitrogen—take no part in the action, beyond abstracting a
large proportion of the heat; but when the air is heated to a high
temperature before entering the furnace, the cooling effect of the
nitrogen is greatly obviated, and consequently a much higher temperature
is obtained at the place of combustion, and the requisite intensity of
heat is at once produced, which is most effective in completing the
fusion and separation from each other of the slags and iron, and also in
accomplishing the reduction of the oxide. But Neilson found that the net
result of burning some fuel to heat the air before entering the furnace
was a great economy of the total fuel required for smelting the ore. He
had to encounter many difficulties in carrying his invention into
practice; the iron ovens first used for heating the air were rapidly
oxidized; and when thick cast iron pipes were substituted, these were
liable to leak at the joints on account of the expansions and
contractions caused by changes of temperature. Then the new invention
had as usual to contend with established prejudices and misconceptions;
but it soon came into use in Scotland, where it effected a great saving;
inasmuch as it was found possible to use with the hot blast raw coal of
a certain kind, plentiful in Scotland, because the heat retained by the
ascending gases sufficed to convert the coal at the top of the charge
into coke.
It will be remembered that the active agent in the reduction of the ore
is the carbonic oxide gas formed by the incomplete combustion of the
carbon of the fuel; or what comes to the same thing, the absorption by
carbonic acid first produced of another proportion of carbon. The carbon
oxide robs the iron oxide of its oxygen to become itself changed into
carbonic acid. In reality however the action is more complex than this
in its chemical relations; for instance, metallic iron will under
certain circumstances act conversely on carbonic acid, and rob it of
half its oxygen. The net result of the reactions between carbon, iron,
iron oxide, and these gases depends mainly upon the temperature and
pressure and upon the relative quantities of each substance present. In
the gases escaping from the blast furnace there is always a large
quantity (nearly one-third) of carbonic oxide. At the blast furnaces in
work during the first half of our century the combustible gases were
allowed to burn to waste as they issued from the top of the furnace, in
the manner shown in Fig. 17, and at night the flames used to form a
weird and striking feature in the prospect of an iron-smelting region.
Instead of allowing the escaping gases to burn to waste, it became the
practice about 1860, and so continues, to draw them off and burn them
under steam boilers or use their flames for heating the blast. An
effective method of withdrawing the gases is shown in Fig. 21, which is
a section through the upper part of a smelting furnace, with the “cup
and cone” arrangement. The mouth of the furnace is covered by a shallow
iron cone _a_, open at the bottom, into which fits another cone _b_,
attached to a chain _c_, sustained by an arm of the lever _d_, which is
firmly held in position by the chain _e_, and is also provided with a
counterpoise _f_. When the mouth of the furnace is thus closed, the
gases find an exit by the opening _g_, seen behind the cones, and
leading into a downward passage, through which they are drawn by the
draught of a tall chimney to the place where they are burnt. The charge
for the furnace is filled into the hopper _a_, and at the proper time
the chain, _e_, is slackened when the weight of the material resting on
the suspended cone overcomes that of the counterpoise, and the charge
slides down over the surface of the cone _b_, which is immediately drawn
up again by the counterpoise, so that the opening into the air is at
once closed.
[Illustration:
FIG. 21.—_Cup and Cone._
]
The march of improvement in the blast furnace has been characterized
particularly in Britain and the United States by a great increase of
dimensions, which is found to promote economy in fuel, etc. In the
former country the furnace of the latter part of our century is commonly
from 70 to 80 feet high, and some have even been built with a height of
more than 100 feet, while in the States the tendency to build very high
furnaces is still more marked. A single large furnace may turn out as
much as 1,500 tons of pig iron in a week, and some in America, it is
said, actually produce as much as 2,500 tons. The more usual output of a
blast furnace is however much less than these amounts; but if we say
only one-half, or even one-third of these quantities, a state of things
is indicated very different from what obtained about 1837, when the best
Welsh furnaces produced only 200 tons a week. If we go back to the
beginning of the century, the difference is much more marked, for the
blast furnaces of that period could turn out only about 30 tons in a
week.
The proportions of fuel, ore, and limestone charged into the furnace
vary greatly according to the composition of the ore, the quality of
iron aimed at, and the practice of each manufacturer. It is usual
previously to calcine the carbonate ores and others also, in order to
expel the carbonic acid and the moisture, of which last all contain a
considerable amount: and sometimes the limestone is mixed with the ore
to undergo this preliminary process. The charge being conveyed from the
roasting kilns to the blast furnace while still hot effects an obvious
economy of fuel in the latter. In the case of hæmatite ore the
quantities of materials in one charge may be something like 54 cwt. of
ore, 9 cwt. of limestone, and 33 cwt. of coke. It is quite common to use
mixtures of different kinds of ore, so as to modify the quality of the
product according to particular requirements. The use of the limestone
is to take up silica, and the slag is found to consist mainly of
silicates of lime and alumina. The amount flowing from a blast furnace
of course varies much according to the conditions, and is larger than
would commonly be supposed; for the production of one ton of pig iron
involves the production of from ½ to 1½ tons of slag.
Fig. 22 represents in section the later type of blast furnace, which of
course is circular in plan. Its height may be taken as 80 feet, and the
diameter at the widest part of the interior as 22½ feet, narrowed to 20
feet near the top. The lowest portion, C, is called the _crucible_, the
bottom of which is the _hearth_, both formed of the most refractory
materials obtainable. The conical widening, B, above the crucible is the
_boshes_, and at the top is seen the “cup and cone” apparatus already
described, A, surmounted by the short cylindrical iron mouth, through
apertures in which the charges are tipped from the gallery, D, these
having been raised there in small trucks by hydraulic or other
elevators. The escaping gases leave the furnace by the exit, E, which
leads into the “down-come,” G, and they are conducted from it to the
“regenerative stoves” and dealt with as presently to be described. Our
section represents the masonry of the furnace as sustained by pillars,
P, at the outside of the lower part; these pillars support a strong ring
of iron plates upon which the wall rests. This arrangement has the
advantage of allowing the workmen the greatest freedom of access to
parts about the crucible, which require much attention. Here, at the
lowest part, is an aperture from which the liquid iron is allowed to run
out every five or six hours, it being plugged in the meantime by clay
and sand. The slag being much lighter than the iron, floats above it,
and runs off at a higher level over the _tympstone_. Opening into the
hearth are several orifices to admit the hot blast from the nozzles of
the _tuyères_, which of course do not project into the furnace itself;
but they are so near to the region of intensest heat that they would be
rapidly destroyed unless they were surrounded by a casing through which
a current of water is constantly running. The _tuyères_, of which there
may be 3 or 5, are supplied from the pipe seen at K. The earlier plans
of heating the air did not permit of a very high temperature being given
to the hot blast, about 600° F. being the limit; but the “regenerative”
stoves can supply a blast of more than 1,600° F., or not far below the
melting point of silver. Another great increase has been in the pressure
of the blast; 2 or 3 lbs. per square inch sufficed in the earlier
practice; but the lofty modern furnaces have to be supplied with the
blast at a pressure of 10 lbs. per square inch, and over. Even when
comparatively low pressures were the rule, a large ironworks required
much blowing power. The works formerly at Dowlais, in South Wales, for
instance, had an engine of 650 horse-power for the blowing engine, in
which a piston of 12 feet diameter moved in a cylinder 12 feet in
length. The quantity of air that passes into a blast furnace amounts to
thousands of tons per week, its weight being much greater than that of
all the ore, coke, and limestone put together.
[Illustration:
FIG. 22.—_Section of Blast Furnace._
]
It need scarcely be said that great care and expense are bestowed on the
construction of these furnaces. Only the best and most refractory
materials, such as firebricks, are used for the lining, and the exterior
is a casing of solid masonry, strengthened with iron bands. When a new
furnace is finished it takes a month or six weeks to put it into
operation; but when this is done it will remain in action night and day
continuously for a long period—perhaps for eight or ten years—before the
necessity for repairs requires a “blow out.” And the blow out and
restarting, without the cost of repairs, entail an outlay of several
hundred pounds.
The gases leaving the throat of the furnace consist mainly of nitrogen
and a little carbonic acid, together with about one-third of their
volume of the combustible gases, carbonic oxide, and some hydrogen; but
these last do not leave the furnace in an ignited state, because the
oxygen there has already been consumed. They are conducted by the
“down-come” pipe, G, Fig. 22, to a point at which, by means of a valve,
they can be directed to one or other of two circular towers entirely
filled with firebricks, arranged chequerwise, so as to form innumerable
passages between them. The furnace gases are admitted at the bottom of
the Cowper tower, or “regenerative stove,” into a flue to which a
regulated quantity of air has access, and there they are fired: the
flame ascending the flue to the upper part of the tower, thence
descends, communicating its heat to the firebricks, which soon acquire a
very high temperature, especially where the flame first enters, and the
burnt gases leave the tower for a tall chimney, leaving most of their
heat in the firebricks. When this action has continued for a sufficient
time, the connection of the regenerator with the throat of the furnace
is cut off, and the escaping gases are directed into the other
regenerator, and at the same time the blast from the blowing engine is
made to ascend among the firebricks of the first, where gaining
increasing temperature as it ascends—the stove being hottest at the
top—the air leaves the tower to be conducted to the _tuyères_ at such
high temperature as already mentioned. While the one regenerator is thus
heating the blast, the other is in its turn accumulating heat from the
flames of the escaping gases; and thus they are worked alternately, the
action being constantly reversed after suitable intervals.