IKON 333 The decarbonizing and desiliconizing of iron by the action of an oxidizing atmosphere is the essential feature of the processes of re fining pig iron ( 23) and of making natural steel ( 26) ; but prior to 1855 these processes had only been applied to the partial purifi cation and decarbonization of pig iron, the air being blown over the surface of the fused metal ; and, in consequence of the compara tively slow rate of oxidation of carbon and silicon thus brought about, the use of fuel to melt the iron and to keep it in fusion was essential. On September 15, 1855, an English patent was granted to Gilbert Martien of Newark, New Jersey, U.S., for the purpose of partially purifying cast iron by passing streams of air or steam "through and amongst the melted metal as it flows from a blast furnace" or the rernelting furnace, the object being apparently, not to con vert pig iron into wrought iron and to supersede the puddling furnace, but simply to act as au adjunct to the refinery. Shortly after, Parry made experiments at Ebbw A 7 ale on a process substan tially the same as this. On October 17, 1855, Bessemer took out his first patent for " forcing currents of air or of steam, or of air and steam, into and among the particles of molten crude iron or of remelted pig or refined iron, until the metal so treated is thereby rendered malleable and has acquired other properties common to cast steel, and still retaining the fluid state of such metal, and pouring or running the same into suitable moulds," i.e., for converting cast iron into cast steel. A succession of patents for various improvements was taken out during the next few months, in the course of which the use of steam was dropped, certain parti cular appliances described, and the production of malleable iron as well as steel claimed. It was speedily found, however, that the production of steel of uniform quality from English pig irons was impracticable, owing to the difficulty in stopping the blowing operation at exactly the right moment to produce the desired degree of carbonization, and that the production of malleable iron was equally an unsuccessful manufacturing operation, because if the " blow " continued a little too long, the product was " burnt iron," containing oxide disseminated through it which rendered it brittle ; whilst if the metal were underblown it was hard and steely. Again, contrary to expectation in view of the knownaction of the puddling process, the oxidation of the copper, sulphur, and phosphorus con tained in the pig iron was found to be so faintly marked that practically the resulting "semi-steel" contained the whole of these impurities originally present in the pig iron employed. Accord ingly the value of the new process, of which the highest expecta tions were at first formed, was speedily found to be really but small, notwithstanding the various successive improvements patented by the inventor during 1855 and 1856 ; towards the end of the latter year, however, the difficulty was solved and the whole process ren dered practical and readily controllable by Mushet, who patented the improvement in use to the present day of decarbonizing the iron by completely blowing it, and then adding melted spiegeleisen in known quantity so as to carbonize the total mass to any definite required extent, and also to introduce manganese into the com position, thereby diminishing the injurious effects of sulphur, phos phorus, &c., on its physical qualities, the character of the metal being further regulated by choosing for the operation haematite pig, or some other kind, containing only minute amounts of sulphur, copper, and phosphorus. Mushet s patent right, however, was allowed to lapse through neglect to pay the requisite fees in the third year ; and in consequence his name is all but forgotten in connexion with his improvement on Bessemer s own process, the combination being ordinarily termed " Bessemerizing. " Details as to the practical working of the combination process are given in 36. It is to be here remarked in connexion with the Bessemer pro cess proper (the blowing) that, whilst the difficulties in the way of preparing uniform products with English irons have led to the entire abandonment of the production of iron or steel thereby in England, the method is still in use to some extent in Sweden, at Seraing, and elsewhere, the proper moment when the blow should cease being determined by rapidly sampling and testing the metal, or by the colour of the slag. In Sweden the charges of metal blown at one operation are occasionally much smaller than those usually employed elsewhere where the combination-process is adopted, whilst the converters in use are sometimes of the fixed pattern adopted by Bessemer in his earlier experiments, now mostly super seded for the spiegeleisen process by the movable converters swinging on trunnions described in 36 ; in the newer Bessemer works, however, the most improved methods and arrangements are in use. In the John Cockcrill Works (Seraing) it has been found practi cable to ensure the continuous production of pig in the blast furnaces of sensibly constant composition, Algerian and Spanish ores being employed. From these, pig of the annexed composition is smelted and run direct into the converters without solidifying ; 23 5 parts of limestone per 100 of ore are employed together with coke (containing 8 to 10 per cent, of ash) in the proportion of 22 cwts. per ton of pig. Average Composi tion of Ores used. Cinder produced. Pig. Water G-50 Carbon dioxide 2-50 Carbon 4*50 Silicon 15-00 37-00 Alumina 4-00 13-50 Lime 3 00 43-00 Magnesia 0-50 1-50 89 "40 Ferric oxide G4-00 50 Manganous oxide Sulphur 4-25 010 3-50 1 25 100-00 Phosphoric anhydride 0-075 99-925 100-25 Owing to the considerable amount of manganese present in the pig, sufficient of that metal remains unoxidized in the blown pro duct to render it unnecessary to add spiegeleisen thereto ; on this depends the practicability of the process ; the blowing is continued until a specimen of the slag (obtained during a brief intermission of the blast for the purpose) exhibits a particular colour dependent upon the amount of residual carbon required, whilst the physical characters of the globules of metal interspersed throughout the sample are also noted ; the metal is then tipped into the casting ladle, and run into ingots which are reheated when solid enough to be withdrawn from the moulds and rolled without ever cooling below a red heat. The colour scale and the corresponding carbon percentages are as follows : Colour Percentage of Carbon of Slag j n steel. Lemon yellow 0-75 or upwards. Orange ye .low close to 0-CO Light brown 0-45 J ark brown Q 30 Bluish black 15 As regards the general character of the blowing operation, it is noticeable that the generation of heat by the oxidation of silicon and carbon is so large that without the use of any fuel at all the metal is not only kept melted but increases considerably in tempera ture, so that it remains fluid whilst the decarbouization goes on, instead of becoming pasty and almost solid as it does in the puddling forge when "coming to nature." The nature of the gaseous products on blowing a considerable mass of metal, say 5 tons, is somewhat different during the different stages of the pro cess. At first when the metal is at a relatively lower temperature, a considerable amount of carbon dioxide is formed, together with carbon oxide, but later on, when the temperature is much higher, little but carbon oxide is produced. During the early stages, more over, the amount of oxygen (combined as oxides of carbon) is much less relatively to the nitrogen than in ordinary air, showing that much of the silicon and manganese present are being oxidized ; whilst in the latter half of the blow, when the silicon and manganese have largely become oxidized, the amount of oxygen in the issuing gases is much larger, nearly equal to that present in air. Thus the following series of analyses were made by Snelus during an eighteen minutes blow (Journal Iron and Steel Institute, 1871, ii. p. 247), the specimens being collected respectively after two, four, six, ten, twelve, and fourteen minutes from the commencement. Similar results have also been subsequently obtained]] by other chemists, notably Adolf Tamm (Jcrn-KontoretsAnnalcr, xxx. 257 ; also Iron, 1879), with the iron made at Westanfors from charcoal pig. Time from Commencement, expressed as a Fraction of Total Duration of Blow. i 2 T i TT 5 tr
tr Carbon oxide nil 10-71 0-92 3-95 8.^7 Of 4 CO 8-15 19-59 3-58 29-44 2-39 31-11 1-34 ,, dioxide Oxj-gen Nitrogen ) 88-37 | 8658 0-88 85-25) 2-OOj" 76-83 -( ! cc-on 2-1C) C7 55 1 Hydrogen j 100-00 100-00 j 100-00 100-00 100-00 j 100-00 From these analyses Snelus calculates that 43 per cent, of the total carbon oxidized in this blow was converted into C0 2 , and 57 per cent, into CO. It hence results that a considerable development of heat attends the operation, especially with irons moderately rich in silicon as well as carbon. Taking the heat of combustion of carbon to CO as 2400, of carbon to C0 2 as 8000 ( 20), and of silicon to Si0 2 as 7800 (Troost and Hautefeuille give 7830), and assuming that these are also the values of the combustion heats of these elements when dissolved in (or united with) iron, which is not the case, the values being really somewhat lessened by the amounts of heat evolution during the solution in (or combination with) the iron, it results that, on blowing an iron containing for instance 2 per cent, of silicon and 3 per cent, of carbon, there will be a heat development to the following extent :
