« PreviousContinue »
Mr. Hadfield says:
"Some noteworthy results are obtained, especially in samples i and j, in which the resistance to compression is extraordinarily high. These results are the more remarkable in view of the low carbon present. A still more curious point is, that notwithstanding the high resistance to compression stresses, they are comparatively soft and can be machined—though not readily. There is no doubt that as regards pure alloys of iron and nickel, that is, no carbon or other elements being present or to only a small extent, the increase in resistance to stress is a most remarkable property, apparently only possessed by these nickel-iron alloys."
The Shearing-Strength of Nickel-Steels.
In shearing the edges of rolled deck-plates, carbon-steel often cuts raggedly—small fragments breaking back from the edge of the shears. Nickel-steel, on the other hand, cuts clean, like a very mild steel.* This is also true of punching. "Careful experiments in punching show that the hole is left with clean-cut edges, with no slivers or wire-edges. Nickel-plates shear more neatly than carbon-plates."f
Prof. Rudeloff has investigated the influence of nickel upon the shearing-strength of iron, with the following results:%
Table XIX.—Effect of Nickel on Shearing-Strength.
Per cent, Lbs. per Sq. In.
The shearing-strength rapidly rises with increase in the percentage of nickel, up to 16 per cent, at which point the nickelalloy has 2.5 times the shearing-strength of pure iron.
* E. F. Wood, Carnegie Steel Co. Private communication, t "The Advantages of Nickel-Steel," Commander J. G. Eaton. % M. Rudeloff, "Vierter Bericht des Sonderauschusses fur Eisenlegirungen," Verein zur Bef. des Gewerbfl., Berlin, 1896.
Further results on the shearing of nickel-steels are given below, under the head of nickel-steel rivets.
Loss of Strength on Punching.
Tests have been made at the Parkhead Forge rolling-mill, in Glasgow, Scotland, to determine the original strength of nickelsteel boiler-plates, and also the residual strength, after a portion had been removed by punching. Two samples of plate were tested: 'No. 1 being mild steel, of usual "boiler-plate" quality, and No. 2 a steel somewhat higher in carbon. Each contained the usual 3 per cent, of nickel. The tensile strength of the whole sheets, before punching, averaged as follows:
Lbs. per Sq. In.
No. 1 84,672
No. 2, 116,480
Test-pieces of these sheets, about 3.5 inches in width, were taken, and, in the center of each, 1-inch holes were punched. The samples were then subjected to tensile tests, with the following average results:
Lbs. per Sq. In.
No. 1 71,523
No. 2 89,510
The loss of strength by punching was, for No. 1, 15.5 per cent,, and for No. 2, 20 per cent, of the original strength. On the thicker sections of No. 1 the loss in strength was only about 10 per cent,
Mr. Beardmore states the loss of strength due to punching in ordinary mild steel to be 33 per cent, of the original strength.
This is, however, higher than is usually reported. Prof. Howe says that it ranges from 9 to 34 per cent,—the average of tests on simple-steel plates showing a little over 20 per cent, loss on punching.
It is evident from this comparison that nickel-steel suffers from punching at least as little as mild boiler-steel. Mr. Beardmore claims that in this regard nickel-steel is much superior to the simple steels.*
* "Nickel Steel," Inst. Engrs. and Shipbuilders of Scotland, March, 1896; Industries and Iron, May, 1896.
The Influence of Nickel on Appearance and Tests.
In handling the charge in the open-hearth furnace, the melter relies for guidance upon the appearance of the fracture of small test-bars, taken from time to time. "With simple carbon-steels, a very accurate estimate of the amount of carbon present can be thus obtained. Nickel changes entirely the characteristic appearance of the fracture. Test-bars from a heat of nickelsteel show very pronounced crystallization, like the chill of cast-iron, the crystals meeting sharply at the angles of the testmould. Nickel-steel has a darker fracture than simple steel; and the furnace-man, judging by the appearance of the crystallization, would be apt to think the carbon was higher than is really the case. If, however, a series of test-bars be taken during a heat of nickel-steel, it will be found that the appearance changes regularly as the carbon drops; and the appearance of the fracture, once learned, forms as safe a criterion of the carbon-contents as is furnished by the corresponding changes in the case of simple steel.
In color-carbon tests on nickel-steels, the chemist is apt to estimate the carbon too low—not because of any color given by the nickel present (which, of itself, does not interfere with the color-test), but by reason of the effect of nickel on the carbon, changing part of it to the hardening-condition, in which form its color-effect is not so visible as that of cemenkcarbon.* For accurate comparison of tests, therefore, combustion-determinations of carbon are necessary.
The color of the finished steel becomes lighter with increase in the nickel-contents. The ordinary 3.5 per cent, nickel-steels do not perceptibly differ in appearance from simple steels; at 10 per cent, of nickel, the color is noticeably lighter; and, at 18 per cent,, the steel has a soft silvery whiteness; with higher percentages the color seems again to darken, and the 25 and 30 per cent, nickel-steels are duller and less lustrous than the 18 per cent, In texture, moreover, the 18 per cent, nickelsteels appear more smooth and close-grained than iron-alloys containing a larger percentage of nickel.
* Mr. Chase, Midvale Steel Works. Private communication.
Segregation in Nickel-Steel.
The various elements which enter into the composition of steels—silicon, sulphur, phosphorus, manganese and carbon— are all thoroughly mixed in the molten steel, and, in the furnace, are evenly distributed throughout the charge. When the steel is cast in the ingotanould, each of these elements forming its own alloy with iron, and each solidifying at a different temperature and possessing a different specific gravity from the others, begins to move through the fluid steel. As the mass cools, the tendency of these compounds is towards the central and upper part of the ingot, Analysis shows that the center of an ingot contains a much greater proportion of carbon, silicon, sulphur and phosphorus than the outside; and for this reason, in the manufacture of shafting, gun-hoops and other articles in which uniformity of composition is a sine qua non, the center of the ingot is removed by boring.
"Nickel is supposed to lessen the segregation and liquation of carbon by combining with the carbon which cements the particles of iron together, and thus bringing the specific gravity of the carbon compounds nearer to that of the rest of the alloys in the fluid mixture. It also seems to cause this cementing carbon to solidify at more nearly the same temperature as the other alloys."*
Analyses from an ingot of uncompressed nickel-steel, 66.5 inches in diameter and 122 inches long, made by the Bethlehem Steel Co., gave the following result :f
Table XX.—Segregation in Nickel-Steel.
Outside. Six In. from Center. Ratio of Con
Per cent, Per cent, centration.
Nickel, . . . .3.07 3.27 100:108
Silicon 0.172 0.170 100: 98
Sulphur 0.03 0.06 100:200
Carbon, . . . .0.31 0.36 100:116
Phosphorus, . . .0.025 0.047 100:188
It will be noted that in this nickel-steel ingot the segregationexcess of carbon is only 0.008 per cent, Prof. Howe in his Metallurgy of Steel, states that in simple steels the average segregation-excess of carbon is greater than that of any other element; and, in five cases which he cites, it exceeds 0.44 per cent. Coefficient of Expansion.
* Mr. G. H. Chase, Midvale Steel Works. Private communication, f Mr. H. P. Porter, Bethlehem Steel Works. Private communication.
Simple steel, when heated or cooled 1 degree Centigrade, at ordinary temperature, expands or contracts about 11.6 parts in a million. Expressed in the usual way, the coefficient of expansion of simple steel for each degree C. is 0.0000116.
According to M. Ch. Ed. Guillaume,* the coefficients of nickel-steels are normal up to 19 per cent, of nickel, from which point the coefficients rise rapidly up to 24 per cent,, then fall to a minimum with steels containing 36 per cent, of nickel, and then, with further increase in nickel, return to a normal value. In investigating the behavior of these alloys, M. Guillaume discovered several remarkable anomalies which led him to divide
By Ch. Ed. Guillaume.
nickel-steels into two classes: those containing less than 26 per cent, of nickel, which he calls irreversible alloys, and which show entirely different behavior at ascending and at descending temperatures; and those containing over 26 per cent, of nickel (the reversible alloys), which have a regular coefficient of expansion, the same for rising as for falling temperatures.
All the irreversible alloys show practically the same behavior as the 15 per cent, nickel-steel, as shown in Fig. 3, in which the abscissas represent the temperature and the ordinates the length of the bar. When, after heating to cherry-red, the bar is allowed to cool, the metal contracts regularly, as shown by