Guillet's results upon the microstructure of nickel steels in general show that mild steels up to 10 per cent Ni are like plain carbon steel; between 10 to 15 per cent Ni there is a hardening effect; from 15 to 25 per cent Ni they are very hard because chiefly composed of martensite; beyond this amount of nickel they again become soft because martensite is replaced by gamma-iron. The higher the car bon the less nickel is required to produce these structural changes.
has been considerable discussion as to the probable state in which molybdenum exists in steels. T. Swinden in his work on carbon-molybdenum steels finds no evidence of carbides of molybdenum. J. 0. Arnold and A. A. Read believe the greater part exists in a compound of the formula Fe,Mo,C. George Mahrs quotes Nicolardot as authority for two carbides, Mo,C and MoC. Carnot and Goutal finds a double carbide Fe,C,Mo,C, the existence of which is confirmed by Williams. The, effect of molybdenum on the critical points is similar to that of chromium, namely, the position of the critical points varies as the temperature from which the material is cooled varies.
G. L. Norris states that van adium replaces iron in the cementite or car bide by increasing amounts until finally when the percentage of vanadium is about 5 per cent all the iron has been replaced by vanadium. Vanadium carbide is not as readily soluble on heating as iron carbide and consequently van adium steel requires a higher temperature to put it into the austenitic condition.
Electrical Conductivity of Steel Several years ago Professor Barrett of Dublin, Dr. Benedicks of Upsala, Sweden, and Dr. Mat hews observed independently and at about the same time that there was a connection between the resistance offered by steel wires to electric conduction and the atomic weight of the ele ments present in the iron. This, again calls at tention to the close relation existing between al loys and ordinary solutions. Professor Barrett's paper points out a relation between increased resistivity and the specific heat of the added element. But when we recall that specific heat X atomic weight = a constant, we see that the increased resistivity is as closely connected with the atomic weight as with specific heat.
In studying the relation of electrical re sistance to the constitution of the conducting alloys, the complex character of steel must be kept in mind. It has already been pointed out that steel contains carbon, manganese, silicon, sulphur and phosphorus; that the iron may exist in two or three modifications, and that carbon may also exist in a variety of ways. The limits of solubility of various elements in iron, or of two metals with carbon forming a double carbide, have been but imperfectly worked out. Yet, notwithstanding these com plications, a broad view of existing evidence leads to the belief that the atomic law is in some way connected with the problem.
Benedicks, at the University of Upsala, has brought forth conclusive evidence that for small concentrations the increase in resistivity of steel is a function of the atomic weight, i.e., equi atomic solutions of metallic elements in iron produce equal increase in electrical resistance. Benedicks determined the electrical resistance of a number of samples of steel which had been carefully analyzed. They contained vary ing quantities of carbon, silicon, manganese, sulphur and phosphorus, the last two elements being fairly low and uniform. The steels were
tested in both the hardened and annealed state. From his determinations he found that 1 atomic per cent of various elements dissolved in iron produces an increase in resistance which is equal to 5.9 microhms per centimeter cubed. He also calculated that the resistance of ab solutely pure iron would be 7.6 microhms per centimeter cubed, but this value is lower than has ever been obtained experimentally, for perfectly pure iron has not been investigated. By means of a formula it was found possible to calculate the resistance of steel with con siderable accuracy, when its analysis was known. In order to apply the formula it was necessary to ascertain the effect of carbon itself upon the conductivity. This Benedicks has done very skilfully. In annealed steel the carbon for the most part exists in separate particles of cementite, FetC. When eutectoid steel is heated to a temperature above 720° C. and suddenly cooled, this cementite disappears and the struc ture known as martensite results. The carbon of martensite may be combined or simply dis solved without combination. However, when hardened steel is reheated and slowly cooled, cementite again appears, accompanied by ferrite. This ferrite has usually been considered to be pure iron, and ferrite and cementite when exist ing in alternating bands constitute pearlite. Bene dicks shows that ferrite is not free from car bon, but that annealed steels containing from 0.40 per cent to 1.70 per cent carbon consists of cementite and iron which contain about 027 per cent dissolved or hardening carbon. Ac cording to Benedicks, the carbon segregated in the free cementite exerts little influence upon the conductivity. Le Chatelier, however, gives the resistance of ferrite as 9.5 and cementite as 45. Benedicks' work thus confirms the chemi cal researches of Osmond and Werth, Carnot and Goutal, Brustlein, Arnold and Stansfield in regard to the existence in annealed high carbon steels of 0.27 per cent of hardening car bon in solid solution. The writer is of the opinion, that, in the separating or crystallizing out of the constituents of any alloy, no pure metal ever separates, but metal containing more or less of the other constituents of the alloy in solid solution. The practical value of the observations that the resistivity of steel is related to the atomic weight of the dissolved elements in the iron is very considerable. As Barrett has mentioned, knowing the carbon con tents of a piece of steal, its relative content of other elements may be judged by determining its conductivity. Benedicks' work makes it.. possible to judge the electrical quality of dif ferent samples of iron by a study of theinoont position without testing them at all. Barrett has also called attention to the fact that physi cal hardness has nothing to do with high re sistivity. That is, for equal percentages of im purities hard tungsten and manganese steels conduct better than soft aluminum and silicon steels. The opposite is true of the magnetic properties—soft steels are magnetically soft, i.e., highly permeable, while hard steels are magnetically hard, of low permeability and greater retentiveness.