126 DEVELOPMENTS OF THE KJELLIN AND ROCHLING- DISCUSSION. Mr. F. W. Harbord considered the modified Kjellin furnace described in the Paper a useful improvement on the original furnace. He did not, however, see how a 7-ton furnace could make rail stcel in competition with a modern converter of 10 to 14 tons capacity, or a 50-ton open- hearth furnace. The electric process, in fact, was really a refining process, and should be used as such, either in conjunction with a converter or an open- hearth furnace ; this was the most promising future for the electric furnace, in connection with steel manufacture. He did not think the etched rail sections they had seen proved very much as regards the uniformity of the electric as against open-hearth steel. He would like to know whether this steel was actually being used for rails.As regards the high breaking stress of the material, there was no difficulty in making such steel, containing, for instance, only a trace of phosphorus, provided pure materials were used, as they were here. The induction type of furnace was not, in his opinion, particularly well adapted for impure materials, and its real advantage was, that it worked in a neutral and not an oxidising atmosphere. It was a crucible furnace, in which ton lots instead of pound lots could be made. He was much interested to find that calcium carbide was produced in the furnace in situ. He had seen-he believed in an Hkroult furnace- carbide added to deoxidise the bath, with extremely good results. Mr. E. Ristori said that the title of the Paper gave a wrong itnpression of the type of furnace the author had described, which was not an induction, but a combined induction and resistance furnace, and it appeared to him from the description that most of the work was done in the resistance portion of the furnace.The induction type of furnace, he considered, was only suitable as a substitute for crucibles, but was not capable of refining steel, viz., of making good steel out of impure raw materials. This could only be done by means of the resistance type of furnace, such as the Hkroult furnace, and he noted that the inventors had to fall back on the resistance principle when they wanted to refine steel. He would like to ask the author if he could give an analysis of the raw materials used and of the finished steel in order to see what amount of refining could actually be done in practice by the type of furnace described.Dr. H. Borns : We have reason to thank Mr. Hard& for his paper. The recent discussions on the performance of electric iron and steel furnaces, notably in Germany, have helped us to a good deal of useful information, and Mr. Hardkn has now made some valuable additions. May I ask a few questions? What was the depth of the metal in the trough when Mr. Hard& observed the pinch effect? I should also feel obliged if Mr. Hard& could explain more fully how the metal is kept at a somewhat higher level in the central portion of the Rodenhauser furnace than in the ring. I t is said that the iron will, in the annular trough, assume an inclined position as if it were under the influence of a centrifugal force ; the effect is aScribed to the combined action of magnetic stray fields and of gravity.IS the surface always inclined, and does Mr. Hardin accept the suggested explanation ? Mr. Harbord has already referred to the part played by calcium carbide in the purification of the steel; Hkroult and Eichhoff make a point of the action of the carbide, especially with regard to the removal of the sulphur. Mr. Hardin tells us that the carbide must be formed in situ, and that. he All rails varied very considerably in this respect.RODENHAUSEK ELECTRIC INDUCTION FURNACES 127 noticed the smell of acetylene. Have we any real evidence of this formation in situ? Dr. Richard Seligman said that as the time was so far advanced he would not detain the meeting longer than to put one or two questions to the author.He had been extremely interested in the author’s remarks on the “pinch effect,” a phenbmenon which was of importance to all of them who had to pass heavy currents through molten conductors, and one which had a habit of cropping up in the most unexpected way. Could the author tell them at what current density the “effect ” was observed in the case of molten iron? The author stated that the “sticky” property of iron was due only to high temperature, and gave as his reason the fact that when fresh metal was added to the bath the stickiness entirely vanished. But was the effect of the added metal merely to reduce the temperature of the bath ? Was it not possible that the ‘( sticky” metal was merely diluted by the addition of the fresh metal, or had the author made the experiment of lowering the temperature of the bath by other means in order to see whether the stickiness disappeared in the same way? Dr.J. A. Harker joined the previous speakers in congratulating the author on the many interesting facts revealed in the Paper. The Grondel- Kj ellin Company had last year kindly presented a small experimental induction furnace, taking some 25 kw., to the National Physical Laboratory. He hoped this would sooii be working, they had been obliged to erect a special building to place it in, for it was intended to study in it some of the interesting questions raised by the author-he would instance the increased resistance of iron at high temperatures-and to determine the physical constants of iron and steel and other metals at high temperatures, accurate data which we were so badly in need of.The speaker also referred to some experiments in which he had noticed the “pinch” effect alluded to in the Paper. In conjunction with Mr. W. A. Price he had attempted to use a resister of molten tin as the heating medium for a bath of fused salts. If the resister were made in the form of a gridiron, the metal being contained in channels machined on the surface of a fire- brick, it was found that at high-current densities, using alternating current of fairly low frequency, severance of the conductor always took place, rendering the method incapable of application. Even with a straight channel, where magnetic effects of the current would be less serious, the slightest irregularity in the shape of the channel holding the metal apparently determined the point of severance.The current densities in these experi- ments would be from 5,000 to 15,000 amperes per square inch, and the conductors about & square inch in cross-section. I t was not found possible to employ the method as originally intended. For experimental purposes it was impossible to overrate the advantage of being able to melt metals and other substances in a neutral atmosphere, such as one had in the induction furnace. The new metallic filament lamps were calling out for such a furnace which could be worked in vacuo. Tungsten, for instance, combined with almost any gas that came near it, and a vacuum induction furnace was therefore a necessity.Did the author know whether the Colby furnace, for which American patents were taken out some years ago, was being used for melting platinum and other refractory metals in vacuo? Dr. J. Harden, replying to Mr. Harbord, said the steel had been used commercially for various purposes, among others for automobile work (samples exhibited). As regards rails, the Prussian State Railways had128 DEVELOPMENTS OF T H E KJELLIN AND ROCHLING- recently ordered considerable quantities for points and crossings, paying some 40 marks a ton more than for ordinary steel. Those who could appre- ciate this fact would recognise that there must be good reasons for such willingness to pay an extra 40 niarks a ton. When discussing the extent to which the steel had been used, it must be borne in mind that the large furnace described was the first of any size built.-VVith reference to the presence of calcium carbide, if ready-made carbide is added to the slag no action takes place (it may even be injurious to the process) provided that the tem- perature is not high enough lo decompose the carbide, in which case desulphuri- sation is observed. The author’s theory, however, is that when lime, CaO, is added to the steel while the latter contains an excess of oxygen, a deoxi- dising material, such as carbon, or, still better, ferrosilicon, also being present, the oxygen then acts upon this substance in a combustive manner, whereby the temperature is considerably increased. Therefore, with aid of the carbon present in the steel, the lime is partly decomposed into calcium and carbon monoxide ; the free calcium combines instantly with the sulphur, forming calcium sulphide, while a portion of the calcium, at this temperature, takes up some of the carbon present, thus forming calcium carbide, both being found later in the slag.Consequently it is not the carbide per se which acts as a desulphuriser or deoxidiser, but its presence is merely an indicator that the desulphurisation has actually taken place. It is the ferrosilicon, &c., which deprives the steel of its oxygen, thereby raising the temperature so that the above reaction may take place. Hence also the necessity of exces- sive heat when trying to desulphurise by means of ready-made carbide, which will have to be decomposed at high temperature; this may certainly be carried out in an arc-furnace, but is to the author’s mind a waste of power, The excessive formation of carbide in any steel furnace must be a waste of power, as it is always an expensive process; it is only necessary to form sufficient carbide to indicate that the desired reaction has taken place.In the case of an arc furnace, with a very high temperature under the arcs, the possibility exists that the deoxidation may be effected by means of lime only (or carbide) ; i.e., if the lime (or carbide) is decomposed by the heat, part of the free calcium may form calcium sulphide, and, as soon as the critical point is again reached, oxygen may be taken from the steel, re-forming lime together with the calcium present in the slag.Of course, no free calcium will ever be found in the slag after it has been withdrawn from the furnace, as it is oxidised instantly by the atmospheric oxygen. The above theory is, of course, very difficult to prove, but many observa- tions tend to indicate that it may be right, and it would certainly be very desirable if the matter could be fully cleared up. Referring to DP. Harker’s remarks, the interesting little furnace at the National Physical Laboratory was the one that had been in use at the Sheffield Exhibition. It had a power- factor of 78 to 81 per cent., and worked with a frequency of 210, the core being very narrow, as it was especially adapted for melting rare substances. In one form of this the windings were outside the crucible, and on this account the power-factor was low.No large furnaces of this type had been built, and the Colby Company was now working in conjunction with the Kjellin Company in America. He knew of no vacuum induction furnace, but he thought that although considerable mechanical skill would be called for there was no serious diffi- culty in the making of one. The best vacuum electric furnace he knew of was that of the General Electric Co., U.S.A., designed by Mr. Arsem. In It would melt 7 lbs. of iron in twenty minutes. The Colby induction furnace had been mentioned.RODENHAUSER ELECTRIC INDUCTION FURNACES 129 this boron carbide had been made and the existence of two carbides had been proved. Wolfram may be produced in a crucible lined with wolfram oxide.In reply to Mr. Ristori‘s remarks, the Kjellin furnace was originally only a melting furnace, but when it was found necessary to refine it was modified to the combination form ; but as, generally speaking, two-thirds of the power was inductive-depending on the degree of refining required-and at the least one-half, it was not accurate to call it a resistance furnace. Moreover the electrodes were neutral and unacted upon. Replying to Dr. Borns, the depth of the bath, when the pinch effect was observed, was about I$ to 2 inches. As to the difference in level, this was very slight only, and effected by brickwork, covering the rings and secured by bandage of iron. As regards the level on the surface in a Kjellin furnace, the magnetic forces are tending to drive the metal outward, widening the ring, thus raising the level on the outer wall ; but the thermal effect, caused by higher current density in the inner parts, will make it rise still more on the inner wall, where also the slag may be mostly observed. In reply to Dr. Seligman, the bath section was about I B x 2 in. and the current about 9,000 amps., but the density might have been higher, due to the obstruction which started the pinch. The “stickiness” of the metal in the observed cases was certainly due to temperature only, and it was overcome by cooling.