334 IRON T *SJ. x -j-j^ x 8000 = 103-2 due to the formation of C0. 2 T B Ax T ih r x2400= 41-0 ,, CO Tib- x 7800 = 156-0 ., ,, Si0 2 or an evolution of 300 "2 per unit weight of metal blown. Analogous but somewhat different values are calculable with irons of different composition, or with blows so conducted that less carbon dioxide is formed. The mean specific heat of the resulting fused metal and slag is probably somewhere near to 20 (at the ordinary tempera ture the specific heat of iron is 01138, Regnault ; the specific heat is probably greater as the temperature rises, and, judging by analogy with water, greater in the fluid state than in the solid ; the specific heat of the slag at the ordinary temperature is higher than that of iron, but its mass is much less); hence, pursuing the above calcula tion, the heat evolution would suffice to raise the temperature of the metal -Q^r = about 1500 C. above its initial temperature, were all the heat applied to that purpose. This is not the case, however, for the containing vessel or converter has also to be heated up, and the waste gas:is carry off a notable amount of heat with them, whilst radiation and the cooling effects of the air on the converter also take up a considerable fraction of the heat ; on the other hand, however, a certain amount of iron becomes oxidized, thus increasing the heat development ; if 5 psr cent, of metal be thus oxidized, the heat of oxidation being taken as somewhere near to 1200 per unit of weight of metal oxidized, the heat evolution due to this will be -j-g-j- x 1200 = 60 0, or about J- of that due to the joint oxidation of the carbon and silicon. Similarly manganese, if present, becomes oxidized with evolution of heat ; on the whole it is calculated by Jordan, Akermann, Snelus, and others who have specially examined this subject, that with the kinds of pig iron usually employed at the present day, and with the size of converters used (holding 5 tons and upwards of metal), the net amount of heat actually employed in heating up the metal is sufficient to raise its temperature by at least 600 ; so that if the initial temperature of the fused pig is about 1400, that of the blown metal is at least 2000, being above the melting point of platinum ; with highly silicious pig the temperature attained is notably higher than with metal containing less silicon. When the blowing has gone on for a short period, so that the iron has become perceptibly raised above its initial temperature, a reaction commences between the iron oxide or silicate already formed and the as yet unoxidized carbon, giving rise to the evolution of gas with a sort of efferves cence ; this stage is technically termed the " boil." The precise period at which it is marked varies with circum stances, a longer time elapsing from the commencement of the blow when the iron is relatively cooler at first, and also when it is richer in silicon, in the former case because the temperature requisite to produce the effervescent action is not reached until a longer time has elapsed, and in the second because the more oxidizable silicon is chiefly affected first, and the effervescent action of iron oxide, &c., on the dissolved carbon only commences when the silicon is to a considerable extent oxidized. 1 During the earlier part of the boil, whilst the silicon and manganese still present are being oxidized, a greyish or whitish kind -of smoke issues from the converter, consisting of minute particles of slag, manganese oxide, &c., mechanically diffused through the gases. When practically all the silicon, carbon, and manganese have been oxidized, and the oxidizing action of the blast is concentrated on the iron, the colour of the smoke emitted changes to brown, and the iron becomes " burnt " or " overblown " ; if to such metal containing much iron oxide diffused throughout it molten spiegeleisen be added, as in the after part of the Bessemer- Mushet process ( 36), the effervescence or "boil" due 1 In all probability the reason why the silicon is first affected is simply because more heat is evolved in the formation of silicon dioxide than of carbon oxide, so that if carbon were burnt to carbon oxide silicon would probably react upon it, thus- Since the heat of combustion of a unit of weight of carbon to CO is about 2400, and that of silicon to SiO n 7800, the transformation 2 x 12 x 2400 28 indicated by the above equation would evolve 7800 - : = 5743 units of heat per unit of weight of silicon, a tolerably high value, indicating a considerably strong tendency towards the occurrence of this transformation. to the reaction of this iron oxide on the carbon of the spiegeleisen takes place with almost explosive violence. The following analyses by Snelus (Journ. I. and S. Inst., 1871, i. 39) illustrate the gradual diminution in carbon, silicon, and manganese, and the non-removal of phosphorus during the process of bio win": Period since Com mencement of Blow, in Min- 0. Original Pig used. C. Boil just com mencing.
13. End of Blowing before adding Spiegel. Final Steel after addition of Spiegeleisen. Ingot Borings. Itftl Crop Ends. So-callfd com- | Uned carbon... ) Graphite 1-200 2-070 2-170 1-550 0097 0-566 0-519 Silicon 1-952 0-014 0-048 0-086 0-795 trace 0-051 trace 0-635 trace 0-004 trace 0020 trace 0-067 trace 0-030 trace 0-055 0-309 0039 99-001 0-033 trace 0-053 0-309 0-039 99-047 Sulphur Phosphorus Manganese Copper Iron (by difference) 94-C30 96984 97 751 99-816 100-000 100-000 1 100-000 100000 100-000 100-000 Analogous figures have been obtained by various other observers ; the results as a whole show that during the first few minutes of the blow the silicon and manganese mainly are oxidized, raising the temperature ; during this period the carbon is also being oxidized, but not so rapidly as during the next interval whilst the boil is proceeding, a result also indicated by the analyses of the gases evolved (supra). The oxidation of the silicon, however, may not be complete even when the whole of the carbon has been burnt out, especially if higTily silicious pig was used in the first instance; thus steels containing 5 and upwards of unoxidized silicon have been found to be formed ; when excessively soft steel is required, so that but little spiegeleisen (or preferably ferro-manganese) is added to the blown metal, the presence of this silicon is not only not injurious, but is desirable as it diminishes "honeycomb ing " ; but with ordinary " Bessemer steel " containing several tenths per cent, of carbon, the presence of this amount of silicon is almost fatal to its value, silicon con jointly loith carbon producing brittleness to a marked extent. Accordingly it is of importance to blend a highly silicious pig with another kind containing less silicon, or to add to the highly silicious pig crop ends and scrap metal, so as to avoid the presence of too much silicon in the steel ultimately produced ; this intermixture when judiciously performed also avoids certain practical diffi culties attending the use of pig containing too high or too low an amount of oxidizable matter other than iron ; in the former case the temperature attained during the blow is excessive, so that on casting the final steel (after adding spiegeleisen, see 36) in iron moulds these are partly fused and destroyed ; in the former case the temperature attained is not high enough to keep the metal fluid whilst pouring into the moulds, in which case it more or less solidifies in the casting ladle, forming a " skull," whilst the slag partially solidifies in the converter and gradually chokes it. In practice it is usually found desirable to have not less than 2 and not more than 3 per cent, of silicon in the metal before blowing, so as to avoid these difficulties ; when the pig is melted in a cupola there is always a liability to loss of silicon and carbon during the fusion through oxida tion, thus impoverishing the iron in combustible matter and risking the formation of skulls, owing to the metal not becoming hot enough during the blowing to remain fluid whilst casting ; on the other hand, when the metal is ian in from the blast furnace direct (by means of an intermediate ladle) there is a liability to variation from time to time in the quality of the pig ; this difficulty can be overcome by carefully regulating the working of the furnace. Durfee has proposed to tap the blast furnace into large gas-fired heaters in which the metal is kept melted for a sufficiently long time to determine its charac-
