December i6, 1915] 



NATURE 



427 



thick and efficient lagging makes the furnace 

 economical in that a comparatively small amount 

 of energy is required to maintain it at any definite 

 temperature; it, however, renders the furnace 

 rery sluggish, and for some operations this is a 

 lisadvantage. The ability to change the tem- 

 irature rapidly and to adjust the new tempera- 

 ire quickly to a definite and constant value is 

 factor which frequently outweighs any small 

 idvantage accruing from a low power con- 

 sumption. 



We may consider for a moment the question 

 >f the power consumption of an electric furnace, 

 reasonably well-lagged furnace, with a tube, 

 ly, I in. internal diameter and 24 in. long, 

 iated to the full temperature of 1000° C. over 

 le central 18 to 20 in. of its length, should con- 

 sume power at the rate of about 400 to 500 watts. 

 This refers to the furnace itself, and it must be 

 >rne in mind that there is always an unavoidable 

 lount of energy lost in the regulating resistance 

 ised in series with the electric furnace. This will 



Sectional diagram of ^lessrs. Gallenkamp's furnace. 



frequently amount to more than 50 per cent, of 

 that usefully consumed in the furnace, and the 

 loss cannot be overcome if provision is to be 

 made for a fairly wide range of furnace tempera- 

 tures. The external loss may be reduced for fur- 

 naces always run at a definite temperature ; as in 

 this case the winding may be so chosen to require 

 only a small external resistance or even to dispense 

 with it altogether. The latter procedure is not 

 generally to be recommended owing to the lack 

 of control thereby introduced. 



If it be desired to maintain the temperature of 

 ;i furnace constant to within about 20° C, it will 

 often be found that the voltage of the commercial 

 power circuit is not sufficiently steady for the 

 purpose. A fluctuation of 5 per cent, in the 

 voltage is not unusual, and the energy, and con- 

 sequently the temperature, changes are thus of 

 the order of 10 per cent. Rapid variations above 

 and below a mean value are not of great import- 

 ance, as the effect of these is damped out by the 

 lag of the furnace. The intermittent use of large 



NO. 2407, VOL. 96] 



motors, etc., on the same energy distribution 

 system is, however, a frequent source of trouble. 

 The most accurate temperature control can be ob- 

 tained by running the furnace off a large capacity 

 storage battery. 



To assist in the control of an electric furnace, 

 it is always desirable to include an ammeter in 

 the supply circuit. For a furnace wound with 

 a low temperature-coefficient alloy, such as 

 "nichrome," the current readings alone form, a 

 useful indication to the behaviour of the furnace, 

 as the resistance docs not change to any very 

 great extent. In the case, however, of a platinum 

 or nickel wound furnace this is not sufficient, and 

 the volt-drop across the furnace must also be 

 measured. It will frequently be found that 

 changes in the regulating resistance in series with 

 a platinum furnace produce no marked changes 

 in the current, the volt-drop on the furnace, how- 

 ever, being considerably affected. The use of a 

 wattmeter is hence to be recommended for fur- 

 naces wound with materials the temperature- 

 coefficient of resistance of which is high. 



In the construction of some furnaces use is 

 made of various fireclay cements to hold the wind- 



Urilife C**in 



Firecl«.y Supporr for 

 Furn&ce Tube 



Fig. 2.— Diagrammatic representation of a Aypi^l ; wire-wound electric 

 furnace. ' . \ 



■;•".•' ji'.i ' 



in^ in position on the tube. Except for furnaces 

 which are not inteAded to run above 900° or 

 1000° C. this is undesirable, and a more satis- 

 factory plan is. to employ a tube on the external 

 surface of whipb a spiral groove is moulded; 

 such tubes may now be obtained in a variety of 

 sizes from the leading makers of refractory 

 materials. The ends of the winding should be 

 held firmly in position by being bound with a few 

 extra turns of the same material. At each end 

 of the tube these wires, when tightly twisted to- 

 gether, will form multiple stranded leads serving 

 for the current supply. The use of a different 

 metal in contact with the v/inding is generally 

 I unsatisfactory, and the failure of a high tempera- 

 i ture furnace is often due to this fault in con- 

 ' struction. Another cause of failure which may 

 be overlooked is the presence of a small piece of 

 foreign matter or impurity in the lagging em- 

 j ployed. If this comes into contact with the hot 

 i winding, it sometimes leads to fluxing of the wire 

 i and consequent breakdown of the furnace. The 

 \ material used for lagging should hence be care- 

 i fully examined, and, needless to say, it must be 

 I of such a nature that it will not react chemically 



