November 9, 191 1] 



NATURE 



and three steel electrodes similarly arranged in the bottom 

 of the furnace, but covered by refractory material. Three- 

 phase current is used, and it is claimed that the current 

 flows from one top electrode to the others, from one bottom 

 electrode to the others, and from each top electrode to each 

 bottom electrode. 



Kjellin Induction Furnace.- — In this furnace, an example 

 of which is in the metallurgical laboratory of the University 

 of Sheflfield, and was shown working to the members of 

 the British Association, the metal charge is placed in an 

 annular hearth, almost like a steel-melting crucible in 

 •section, but in the form of a ring. The primary coil of 

 twenty-four turns is placed in the centre round a core of 

 laminated iron. The bath or ring of metal acts as a 

 secondary circuit of a single turn, and the heat is thus 

 produced in the charge itself without contact with elec- 

 trodes. In the Frick furnace the primary coil is above the 

 crucible, and in the Colby round the outside of the crucible. 



The Rochling-Rodenhauser furnace is based on the 

 Kjellin principle, but has an important addition. In its 

 simplest form, for single-phase current, there are two 

 grooves, or heating channels, corresponding to the annular 

 •crucible of the Kjellin, but these join to a central open- 

 hearth, the whole hearth forming a kind of figure 8. In 

 the central open-hearth all the distinctly metallurgical 

 operations take place, so that this form can be used for 

 refining work, for which the Kjellin is not very suitable. 

 Not only so, but a distinct secondary winding is provided 

 in which a secondary current is induced, and these wind- 

 ings are joined to steel terminal plates which are embedded 

 in the refractory material of the furnace at the ends of the 

 central hearth. At high temperatures the refractory 

 material becomes a conductor of electricity, and thus the 

 currents induced pass through the bath in the central 

 hearth, heating it still further. 



There are many others, some only on paper ; but these 

 are the principal varieties that have been tried with any 

 considerable degree of success. The loss in melting is an 

 important point, and I am informed that this amounts to 

 about I5 per cent, in the Kjellin, about 4 to 5 per cent, 

 in the Rochling-Rodenhauser, and 7 to 8 per cent, in arc 

 furnaces. 



In considering the present position of the electric steel- 

 melting industry regard must be had to the numbers and 

 capacities of the various types of furnaces in work, not in 

 work, and being built, although a complete survey should 

 also take account of the nature and quality of the materials 

 being made, for a furnace making a ton of high-speed steel 

 should obviously be credited with more importance in the 

 commercial world than one making a ton of steel for rails. 

 The progress in numbers and capacities and in output 

 •should also be considered. So far as one could ascertain, 

 about June, 1910, there were about 118 furnaces of all 

 types, of which 70 were in use, 10 not working, and 38 

 being built. There were 77 of the arc furnaces recorded, 

 of which 29 were credited as H^roult, 17 Girod, 13 Stassano, 

 (•> Keller, and q others ; besides one furnace at Domnarfvet, 

 Sweden, for the production of 2500 tons of pig-iron per 

 annum, with one in Norway and one at Trollhattan. 

 Sweden, both in course of construction, and each designed 

 to produce about 7500 tons of pig-iron annually. Of the 

 Hdroult furnaces, the total capacity per charge of those 

 ■working was about 80 tons, and of those in course of con- 

 struction about 50 tons. The Girod furnaces, the great 

 competitors of the Hdroult, were recorded at about 38 tons 

 in work and 26 tons being built. Similarly, the figures for 

 the Kellor were 13 tons and 5 tons, and for the others 

 20 tons and 13 tons respectively. 



Of the induction furnaces, the Kjellin furnaces erected 

 totalled fourteen, with 35 tons capacity; the Rochling- 

 Rodenhausor fifteen, with 30 tons in work, i ton not in 

 work, and 17 tons capacity being built ; all others about 

 18 tons in work. That gave a total rapacity of about 250 

 tons for the arc furnaces and 100 tons for the induction, 

 or agrand total of 350 tons per charge for all electric steel- 

 melting furnaces. Pressure of other work has prevented me 

 from getting the latest figures from all the firms making 

 electric furnaces, but I have obtained these from the two 

 most important firms, viz. the H^roult and the Kjellin and 

 Rorhling-Rodonhnusor, and in this connection would rrrord 

 111V 1m St (hnnk. 1.) Mr. Donald F. Campbell and Mr. !■: ('. 



NO. 2193, VOL. 88] 



Ibbotson, respectively, for their kind help and trouble in 

 getting me this information. Comparing the H^roult 

 furnaces only, as an example, we have seen that in June, 

 1910, there were twenty-nine of these furnaces with a 

 capacity of 80 tons in work and 50 tons in course of erection, 

 130 tons in all; whilst about June, 1911, there were forty- 

 three furnaces, with a total capacity of about 242 tons 



The output of electric steel in Germany, the United 

 States, and Austria-Hungary in 19 10 amounted to almost 

 112,000 tons, which is an increase of 63,000 tons over the 

 figures for 1909. These are the only countries for which 

 the exact output of electric steel is published, but there 

 is no doubt that the figures for Sweden, France, Belgium, 

 and Italy would also show large gains. The increase will 

 probably be more than maintained in 191 1, as more than 

 thirty new furnaces of various types should be started 

 during the year, and many which only started towards 

 the end of 1910 will put in a full year's work in 191 1. 

 England will also for the first time appear as a regular 

 producer. Before the beginning of the present year the 

 H6roult furnace at Edgar Allen's in Sheffield was the 

 only arc one in steady operation. In January three 

 H^roult furnaces were commenced in England : at Vickers' 

 and Thos. Firth and Sons' in Sheffield, and at Lake and 

 Elliott's in Braintree, Essex. A Gronwall furnace, for 

 demonstration and manufacturing purposes, also started 

 at about the same time in Sheffield ; and the output of 

 England for 19 11 should amount to about 13,000 tons. 

 A 15-ton H^roult furnace is to be erected at Skinningrove 

 shortly, and is expected to turn out 200 tons per day. 

 About the same period Kjellin induction furnaces have 

 been working satisfactorily at Vickers and Jessop's in 

 Sheffield and an experimental furnace at the University 

 of Sheffield. 



Great progress will be made in Germany with electric 

 furnaces during the next year, when H^roult furnaces of 

 25 and 22 tons capacity are to be constructed. At present 

 the largest size are the two 15-ton H6roult furnaces at 

 S. Chicago and Worcester, belonging to the United States 

 Steel Corporation, who have recently acquired the H^roult 

 patents for America, and will probably erect several more 

 furnaces shortly. 



The electric furnace can be used either for melting 

 scrap directly or in combination with some other form of 

 furnace, in which case it simply acts as a refiner. The 

 majority of the recent furnaces have been employed in this 

 way, in conjunction either with Bessemer or open-hearth 

 furnaces. The latter are usually of the basic tilting type, 

 part of the charge being removed to the electric furnace 

 after the pig is melted and the bulk of the phosphorus 

 removed, leaving some phosphorus and the oxygen and 

 sulphur to be eliminated by the electric furnace. In this 

 case the time required for the electric furnace is from 

 one hour to two hours, according to the degree of refining 

 required and the original condition of the steel when re- 

 moved from the basic furnace. The power used varies 

 from 100 to 300 kw. hours per ton. When cold scrap is 

 melted the time required is about six hours, and the 

 power consumption said to be from 650 to 750 kw. hours ; 

 but really, all in. more probably 800 to 1000 per ton. Of 

 the forty-four H^roult furnaces in operation or con- 

 struction twenty-one are to melt scrap, twenty to take 

 molten steel from the basic open-hearth, one from a 

 Talbot furnace, and two from converters. 



Electric furnaces are being employed in the following 

 cases : — 



(i) To replace crucibles. The gain is then one of cost 

 of production. 



(2) For foundries. Electric furnaces are bcinr; u-..-d in 

 many foundries. At Georg Fischer's and SrhalTliau-^en 

 they are the only furnaces employed, and Lake and Elliott, 

 of Braintree, are now making most of their steel electric- 

 ally. 



(3) To replace Swedish Bessemer steel, and for steel of 

 axlo and tvr*^ auality. 



(4) For wrldless ' tubes. 'I'ho Mniin. -iii.niii Company 

 has llt'rouh furnaces in Gcrinanv and fialv. 



(5) In (•oiiil)!natioii witli T.-ilIiol fiirnar.s. Owing to the 

 fact (hat tlv Ii^'at ii>i-(I not I10 sutTiiiciillv 54n\'it for toom- 

 ing on Iran'ifprrncr" lo tlio .•liM'tric furiiar.', the output nf 

 tlif Taflxit and \hr lifi-- of tln^ lining; and roof ar^^ -aid 



