The mode of proceeding when coke is the fuel employed, rests upon the same principles, but the dimensions of furnace that are best suited to the different combustibles are different. As a general principle, the height of furnaces must depend upon the force of the blast and the density of the fuel. If the fuel be dense, and the blowing machine weak, the furnace must not have a great height; and even if the blast can be made strong, too high a furnace is disadvantageous for light charcoal. Coke, on the other hand, may be used in furnaces of greater height than any species of charcoal, provided the blast be of sufficient power. So long as the imperfect bellows were used in blowing, the height of the furnace was limited wholly by their action. More powerful apparatus in the form of cylinders, analogous in form and arrangement to those of steam-engines, and like them, either single or double acting, have now been introduced; the intensity of the blast is in them only limited by the moving power, which is applied to them, and when this is the steam engine, it may be said, that no limit can arise from the want of blast. We may, therefore, at the present day, regulate the height of furnaces by the nature of the fuel that is consumed in them.
The greater part of the furnaces in our country still retain the ancient and imperfect form of bellows, hence their height is restricted to the limits of from eighteen to twenty-four feet, and rarely or never reaches thirty. But when the apparatus is such as to supply a proper quantity of air, it has been found that even with light and porous charcoal, such as is given by white pine, the height ought not to be less than thirty feet, and when hard woods are used should be as great as thirty-six feet. Furnaces of even forty feet have been found to answer an excellent purpose, where the charcoal was prepared from oak. When coke is used, furnaces have been made as high as fifty, or even as seventy feet; but experience in England has shown, that from forty-five to forty-eight feet is the proper limit. This height is not at present exceeded in that country, even when the furnace has the greatest dimensions in other respects, and has been found efficacious, even when the vast quantity of eighteen tons has been furnished daily by a single furnace.
The force of the blast will depend upon the nature of the fuel, the volume of air, the quantity of mixed material the furnace holds; and thus furnaces in which coke is used, will require the most powerful blast, whether we have regard to the volume or the intensity. The latter may be measured by a column of mercury adapted in a syphon tube to the air pipes, exactly as the gauge is adapted to the pipes of the steam engine.
The reduction and liquefaction of the metal take place progressively, as the charges descend in the furnace. The separation of the oxygen is due to the presence of carbonaceous matter at high temperatures, begins at the surface of the pieces of ore, and proceeds gradually inwards; the earthy parts of the ore, of the fuel employed, and the flux, unite and melt; they are thus separated, and being sooner fused than the metal, make their way through the charcoal, and descend first to the hearth. The reduced metal, continuing in contact with the burning carbon, acquires a greater or less portion of that substance, becomes fusible, melts, and follows the liquified earths. Dropping into the hearth that already contains the liquid vitrified earths, it passes by its superior gravity to the bottom, and is protected by them from the blast. Even at the bottom of the hearth, the heat is sufficient to retain the carbureted metal in a liquid state, and this is permitted gradually to accumulate, until it rises nearly to the level of the dam.
It now becomes necessary to withdraw or cast the metal. This is done by forcing a way through a channel left beneath the dam in the masonry of the hearth, and closed with clay; the inner portion of this is baked hard, and requires to be broken through with a steel point. As soon as the passage is opened, the metal runs out, and is received in a long trench formed in the sand floor of the moulding house, to which are adapted a number of less trenches, at right angles, each containing about one hundred weight of metal. The metal in the longer trench is also broken into pieces of the same size, and the ingots thus formed are called pigs, whence the term for this variety, pig iron.
From one to three days will elapse from the time of the first charge until the furnace can be tapped, and pigs cast. From that time the casting succeeds with tolerable regularity, according to the working of the furnace, and at intervals depending upon the volume of the charge, and the capacity of the hearth.
It appears probable that the fusion of the iron is effected always by a direct chemical union of that metal with carbon, in the proportion of two atoms of the former to one of the latter. This constitutes, as we have seen, the white variety of pig iron. But as it continues, generally speaking, in the furnace, long after its fusion takes place, it acquires a temperature higher than its proper melting point, and a tendency to separation takes place, the iron retaining in combination no more of the carbon than is necessary to maintain it in a fluid state at the increased temperature. Thus the grey variety of pig iron is formed; and on casting it, the carbon, in a form similar to that of plumbago, is disseminated throughout the mass, or forms on its surface the efflorescence that is called kish, and which is always a sign of a high quality in the iron it accompanies.
In conformity with this theory, we find that a high temperature in the furnace always produces grey cast iron; and that a low temperature, from whatever cause it may arise, renders the iron more or less inclining to white. So also if the metal be not exposed to the heat for a sufficient length of time, it becomes white.
Karsten classes these several causes of whiteness in the product, in the following order:—
"In conformity with the observations that have hitherto been made, white cast iron is obtained: