Page:Encyclopædia Britannica, Ninth Edition, v. 13.djvu/334

318 IRON from 7 65 to 2 95 per unit of pig iron made, the saving with the Welsh ore similarly represented a reduction in consumption at the Plymouth Works, Merthyr Tydvil, of from 53 to 36 cwts., or from 2 65 to 1 80 per unit of pig iron ; so that if the Scotch iron master could compete on equal terms with the Welsh one when cold blast was used, he would have a decided advantage over the latter when hot blast was employed. Accordingly it became a matter of advan tage to the Welsh smelter to decry hot blast metal, because its manufacture gave him less profit relatively to smelters in other dis tricts than that of cold blast metal. Even at this early period, consequently, it was manifest that the advantages derived from a given modification of plant and processes attainable with one class of ore were not necessarily producible with another class ; Dufrenoy found that the furnaces of La Guerche, smelting an impure ore containing 42 per cent, of iron, derived little or no benefit by the substitution of hot air for cold ; the fuel used in each case amounted to some 25 cwts. per ton, and the only dis cernible advantage derived from the hot blast was that the resulting iron became grey instead of white, doubtless from the higher tem perature of the issuing pig enabling the graphite to separate more readily whilst cooling and solidifying. Iron masters, however, have often failed to appreciate the truth of the proposition that what is advantageous for one ore may not necessarily be so for another ; and in consequence very different estimates of the value of new pro cesses and improvements have frequently been made, some condemn ing them wholly, because under the circumstances of their own particular experience no remarkable advantages accrued ; others regarding them as universally beneficial, because under the different conditions of their experience a distinct advantage was gained. The early history of the Bessemer process for producing malleable iron and semi-steel direct from pig by decarbonizing it by blowing air through it affords a good illustration of this point, the process being at first considered by some a complete success, and universally appli cable, good metal having been made by them from certain kinds of pig iron (naturally almgst free from sulphur and phosphorus) ; whilst others regarded it as a failure because the less pure pig experimented on by them yielded only an inferior product. V. CONVERSION OF CAST IKON INTO MALLEABLE IKON AND STEEL BY DECARBONIZATION. 22. Production of " Malleable Cast Iron." It has been known for upwards of a century and a half that when articles of cast iron of not too great thickness are imbedded in powdered iron oxide (a pure red haematite as free as possible from earthy matters, smithy scales, or some obtained from 100 of raw coal) was equivalent to 8 tons 1J cwt. per ton of pig (8 "06 per unit of pig iron) when cold blast was used; in 1830, when hot blast was used with the same coke as fuel, the con sumption was reduced to an amount equivalent to 5 tons Z cwts. (5 16 per unit of pig); and in 1833, when a hotter blast and actual raw coal were employed, the consumption was only 2 tons 5 J cwts. per ton of pig (2 26 per unit of pig), this being exclusive of 8 cwts. of coal used in heating up the blast, making a total of 2 tons 13 cwts. (2 66 per unit of pig) ; figures substantially the same as those of Dufrenoy, represent ing a reduction in fuel consumption nearly in the ratio of one to three parts of Scotch coal used (as coke) with cold blast and Scotch ore. Much the same figures have also been given by Mushet : thus he states that in 1797 with cold blast the consumption of coal at the Clyde works was 7 tons 3 cwts. per ton of pig (7 15 per unit), whereas in 1839 with hot blast it was only 2 tons 34 cwts. (2 175 per unit). On the other hand during the course of a lawsuit entered upon by Neilson to enforce his patent rights, attempts were made to show that the alleged saving in fuel due to the hot blast was really owing to other causes; and subsequently similar views have been urged, perhaps with not wholly disinterested motives; thus Truran states in his work that at Dowlais the coal consumption per ton of iron was reduced to the following extent between 1791 and 1831 by improvements other than the use of hot blast, cold blast being used throughout, with the follow ing consumption per ton of iron : 17 91. 1831. Coal us 6 tons 1 C cwt. 15 2 tons 10 cwt. ... 17 ,, engines and calcining Total,... 8 1 ,, 3 7 Veny probably the increased dimensions of furnaces and various other causes all contributed to this large diminution ; but it does not there fore follow that substituting hot blast for cold did not diminish the consumption still further. That such a substitution did actually save fuel with ores and coal from South Wales is shown by Dufreuoy s figures. analogous substance) and then kept at a red heat for some days (three or more according to the thickness), a diminu tion is produced in the amount of carbon contained, so that the cast iron becomes more or less converted into soft iron. When the action is pushed to the extreme all or almost all of the carbon is removed, that in the outer layers disappear ing first, but no material diminution in the amount of phos phorus, silicon, sulphur, or manganese is produced; 1 if the heating is not continued long enough to remove all the carbon, that which remains is found in the innermost layers which constitute a core of more or less decarbonized cast iron, with an outer skin of malleable iron. Owing to the non-removal of constituents other than carbon, it is essential to the production of a good malleable metal that a tolerably pure cast iron should be employed in the first instance : unless the articles are thin, so that there is no considerable inner core of cast iron, they will not bear forging so as to weld them, the concussion fracturing the brittle core ; on the other hand, the removal of the carbon from the outer skin renders this so much less readily fusible than cast iron that articles so treated (e.g., melting pots and crucibles) will bear a very much higher temperature than cast iron vessels, especially if the core is almost wholly decarbonized ; whilst a much greater degree of toughness and power of resisting fracturing influences is communi cated. Accordingly this method of preparing cheap small malleable iron articles by casting and subsequently decar bonizing is largely employed, the goods produced being known as " run steel "; whilst even with much larger castings, such as the propellers of screw steamers, the method is often adopted, especially in combination with " case hardening " or conversion of the outermost layer of all into steel by a subsequent process (vide infra). Although the process was described in 1722 by Reaumur, patents for it have been subsequently taken out, e.g., by Lucas in 1804, and Brown and Lennox some half century later. In order to carry out the conversion of cast iron into malleable iron in this way, the articles to be treated are packed in cast or wrought iron chests in iron oxide powder ; the chests are then stacked one above another in a kind of reverberatory furnace, and gradually heated up to a red heat, which is maintained for the requisite time, after which they are annealed by slow cooling ; with charcoal pig pretty free from silicon, sulphur, and phosphorus, and with fuel in the furnace free from any large quantity of sulphur, a soft but tough, tenacious, and readily malleable skin is produced ; if, how- 1 Analyses by W. A. Miller, quoted in Percy s Metallurgy ("Iron and Steel," p. Ill), seem to indicate entire removal of sulphur and partial purification from silicon ; thus : Before Treatment. After Treatment. 2-217 0-434 0-583 0-44G 0-951 0-409 trace trace 0-015 nil. trace trace 0-502 ... 1 Probably the cast iron contained irregularly distributed intermixed cinder, the silica of which was counted as silicon. When cast iron plates are slowly oxidized by hot air, according to Tiinner, the silicon diminishes, as it does during refining and in the first stage of the Bessemer process ; but recent analyses of malleable cast iron " articles made on the large scale by cementation in hromatite powder show that substantially no change whatever occurs in the phosphorus and silicon, and that what alteration there is in the sulphur is rather in the direction of increase (from the presence of siilphur in the fuel) than otherwise. For instance Original Iron. Malleable Cast Iron after two Annealings. Original . Cast Iron. Malleable Cast Iron after two Annealings. 3-4:! less than 0-10 3-48 less than 0-10 Silicon 0-445 0-614 0-585 0-449 0-059 0-102 0-105 0-083 0-315 0-295 0-280 0-315 Manganese 0-529 0-575 0-585 0-525