In the steel making practice, it is very difficult to know the true temperature of molten steel in the furnace. As the steel is covered with molten slag, it is difficult to measure the temperature with the optical method, and also the temperature is too high to be measured with the Pt-Ft-Rh thermo-couple.
The author succeeded to measure accurately and easily the temperature of molten steel in the furnace by using the Tungsten-Carbon thermo-couple with a special device.
The principle of construction of his measuring device is as follows:
1) Tungsten wire, one end of which being made flat, is inserted in a carbon tube (one member constituting the couple), and that flat end is in contact with one end of the tube, a spring holding the contact. The end makes a hot junction.
2) The hot junction is covered with a carbon cap. This carbon cap is dipped directly into molten steel. By making use of the carbon cap in place of ordinary refractory covers, the temperature can be quickly and accurately measured, as the time lag of heat conduction is very small.
3) The part which is in contact with molten slag have double protection tubes which consist of a carbon tube and a tube of refractory material. The part which is rot in contact with molten metal and slag has a protection tube of cast iron covered with asbestos paper against gas flame.
4) In this method, it is unnecessary to use inert gases (H² or N²) to prevent oxydation of the couple. The device is not heavy, so that a man can easily carry it.
The author measured the temperature of molten steel in Siemens furnaces, electric arc furnaces and high-frequency electric furnaces with the new devices, and have found various interesting facts as the results

View full abstract

An accurate consideration is given on the carburizing reaction which has heretofore been wrongly represented thus:
3Fe+2CO=Fe³C+CO².
The reaction should be denoted as follows:
(Solid sol.)+2CO (Solid sol.)+CO².
The concentration of carbon dissolved in iron increases as the partial pressure of CO increases until the solid solution becomes saturated with respect to carbon. Since Fe³C is unstable, it does not separate out even when the solid solution is saturated with carbon. The equilibrium existing over the saturated solid solution is,
2CO C+CO²
in which carbon is no longer dissolving in iron. The concentration of carbon varies with the ratio CO/CO₂, the relation being shown in Fig. 4. In this case, the concentration of CO₂ has a larger influence on the carbon content than CO. On the other hand, the reaction C+CO²→2CO, ingenearl, does not reach an equilibrium because of the comparatively inert nature of carbon. It is due to this fact that carburization with carbon alone is relatively feeble; that is, the atomosphere in contact with carbon is apart from the equilibrium 2CO C+CO².
The influence of alkali or alkali-earth carbonates (except CaCO³) in promoting the carburizing reaction, which has hitherto been erroneously considered as a catalytic action, may also be explained on the ground that the equilibrium 2CO C+CO² is established in the presence of these carbonates. On beeing heated, they easily react with carbon producing CO and oxy-carbonates or solution of oxides and carbonates. The CO evolved immediately dissociates maintaining the equilibrium 2CO C+CO². Therefore, the surface of iron in contact with this mixture becomea saturated with carbon, which further promotes the diffusion of carbon into the mass of iron. In the ease of CaCO³, however, the temperature at which the decomposition pressure is equal to 1atm., is 780°Therefore, in a vessel open to air and above this temperature a complete decomposition of CacO³ will take place. Accordingly, CaCO³ cannot have influence upon the carburizing reaction.

View full abstract

The author investigated the diffusion of Cu, Mn, Ni, Fe, Zn, Si and Mg into aluminium which are commonly added as aluminium alloys.
As to the experimental process, each cylinder of 50mm in height and 20mm in diam. made from these diffusing elements or alloys with aluminium was contacted closely with an aluminium disc of 10mm in height and 20mm in diam. Next thus prepared samples were placed in the Tammann tube or mild steel crucible of 150mm in depth and 20mm in diam., so as to let the aluminium disc go upwards, and downwards in the case that it's specific gravity is greater than that of the diffusing elements. The melting for diffusion was carried out by a nichrome resistance furnace, and held for 1·5 hours at a constant temperature. Then samples for chemical analysis were sucked up with an aspirator from the top or bottom of the molten alminium discs.
The diffusion coefficients were calculated by the following equation.
where ψ is the probability integral; h, the thickness of aluminium disc; D, the diffusion coefficient; t, the time of diffusion; CM. the concentration of diffused element at the upper or lower surface of the molten aluminium disc; and CMo, the concentration of diffusing element in the cylinder.
From the experimental results, the author found that Si has the greatest, Mg, Mn and Ni the smallest, and Cu, Zn intermediate diffusion coefficient, except Fe of which coefficient is small at a low temperature and great at a high temperature, whilst he could not find a definite law between the diffusion coefficient and the properties of each element.

View full abstract