384 



METALLURGY. (!RON AND STEEL.) 



would show extreme brittleness, is rolled or forged 

 during cooling to bright-red heat it will exhibit 

 no brittleness when cold. The fracture of the 

 overheated steel generally shows a coarse grain, 

 although this is not necessarily always the case. 

 The single crystal grains of which the structure 

 of the iron is built up, which can be detected 

 under the microscope by suitable etching, are 

 often of considerable dimensions when in the state 

 of overheating. Nevertheless, this is not to be 

 considered as proof positive that overheating has 

 taken place, since the method of cooling also 

 exercises a great influence over the size of the 

 ferrite grains. Rapid cooling from the tempera- 

 ture causing overheating produces fine ferrite 

 grains, without reducing the brittleness appreci- 

 ably. Moreover, it is possible, by heating low- 

 carbon mild steel for days together at between 

 700 and 890 C., to bring the material into such 

 a condition that it will show exceedingly coarse 

 ferrite grains, and yet not exhibit brittleness. 



A new method *of compressing steel during 

 solidification and while still liquid in the ingot 

 mold, which the author called "wiredrawing," 

 is described by M. A. Harmet. When molten 

 steel is poured 'into the ingot mold it may suffer 

 various changes in character, being subject to 

 contraction, crystallization, and liquation, with 

 injurious effects upon its qualities. When the 

 metal begins to cool, it shrinks from the walls of 

 the mold, a solid steel shell is formed, enclosing 

 liquid, and this continuing to cool, shrinks, be- 

 comes plastic, and attaches itself progressively 

 to the shell, leaving a hollow corresponding to the 

 shrinkage, and extending along the axis of the 

 upper part of the ingot. The lower central part 

 of the ingot also has porosities and tiny cracks, 

 and fissures may be detected by the microscope 

 pervading the whole mass. Injurious stresses are 

 set up, crystals are formed having little cohesion 

 between themselves, whereby the liability to crack 

 is increased, and the metalloids that enter into 

 the composition of the steel have a tendency to 

 separate from the iron by liquation. The ingot, 

 as cast, may therefore be useless, and require 

 mechanical treatment to remedy its defects. The 

 author's method is intended to effect compression 

 on the steel while it is in the mold. Pressure is 

 applied by means of a hydraulic press to the 

 bottom of the ingot while it is liquid in the 

 mold. Owing to the form of the modern ingot 

 mold, tapering toward the top, the upper diam- 

 eter is less than that of the lower part. By 

 applying pressure from below, the ingot, which 

 has shrunk on cooling, is thrust upward into the 

 smaller part of the conical mold. The cooled 

 shell thus presses on the central part, and the 

 hollows due to shrinkage are not free to form. 

 By hastening the solidification in this way the 

 .tendency to coarse crystallization is counteracted, 

 and the tendency of carbon to accumulate in the 

 part of the ingot where solidification last takes 

 place is lessened. The process is called wiredraw- 

 ing because of a supposed similarity between the 

 pressing of the metal into the upper part of the 

 mold to forcing it through a draw-plate. Ad- 

 vantages are claimed for this method over that of 

 Sir Joseph Whitworth, who applied pressure from 

 the top, in that the pressure as applied by him is 

 more effective and thorough. The author repre- 

 sents that with it production is increased 25 per 

 cent. 



In the Blau-Thiel process as described by Mr. 

 T. W. Cabot, the fluid iron from the blast-furnace 

 was charged in the refiner with 7 per cent, of 

 quicklime; 10 per cent, of ore was added, and 

 then a second ladle of iron. The charge was 



made of 15 tons. The pig-iron contained 3.70 per 

 cent, carbon, 1.35 phosphorus, 0.90 silicon, 0.40 

 manganese, and 0.05 sulfur. After boiling in 

 the refiner two hours, 90 per cent, of the phos- 

 phorus and 95 per cent, of the silicon had been 

 removed, while more than two-thirds of the 

 carbon remained. The finishing furnace, contain- 

 ing 3J per cent, of lump lime, 7 per cent, of ore, 

 and 7 per cent, of scrap, having been brought up 

 to heat, the refined metal was tapped into it after 

 the slag had been skimmed off. After boiling two 

 and a half hours the phosphorus was brought 

 down to 0.01, and the bath was ready for tapping. 



The belief that the percentage of graphite in 

 iron is independent of the amount of silicon 

 present is attributed by H. M. Howe to a wrong 

 interpretation of the evidence. Mr. Howe shows 

 that the graphite content in normal and relatively 

 pure commercial pig-iron is influenced only in- 

 directly by the percentage of silicon, in that silicon 

 lowers the solvent power of iron for carbon, and 

 thus lessens the proportion of combined carbon 

 and increases that of graphite, provided the total 

 carbon remains constant; the decrease of com- 

 bined carbon is rapid at first, especially as the 

 silicon rises from zero to 0.75 per cent., and then 

 becomes slower and slower. The influence of 

 silicon is often masked by that of the variables. 

 Sulfur is known to raise the saturation point of 

 cast iron for carbon ; by increasing the combined 

 carbon content it lowers the graphite content. It 

 is estimated that the proportion of combined 

 carbon in pig-iron is increased 0.02 per cent, for 

 each 0.01 per cent, increase in sulfur when the 

 iron contains from 1 to 2 per cent, of silicon, and 

 0.03 per cent, for each 0.01 per cent, of sulfur 

 when the iron contains from 2 to 3 per cent, of 

 silicon. 



In well-equipped foundries, the cinder from the 

 cupola is usually crushed in a tumbler and the 

 shot separated from the pulverized cinder. C. H. 

 Putnam further passes the pulverized cinder over 

 a magnetic separator, and thus saves additional 

 iron, recovering daily from the dump of two 

 cupolas 550 pounds of sittings, which give 450 

 pounds of strongly mottled iron after melting. 

 This iron is to be worked in with the regular 

 cupola charge, in amounts to be found by ex- 

 periment. The daily saving by the combined 

 crushing and magnetic separation with two 

 cupolas amounts to $3. 



By the Giebeler process of hardening steel it is 

 claimed that all sorts of iron can be given strength 

 and hardness double that obtained by the Harvey, 

 Krupp, and Boehler processes, while the cost of 

 production is reduced 50 per cent. Experiments 

 made with it at the Technical High School, 

 Charlottenburg, Prussia, were very satisfactory. 



The objections have been made to the new 

 methods of rail production that the rail made by 

 them is so low in carbon and has so soft a head 

 that the wear makes them useless in a much 

 shorter time than the older rails of lighter sec- 

 tion. It is claimed that these difficulties are 

 obviated in the Coyen process, by which a rail is 

 produced with a hard, tough face and free from 

 scale and strain, showing a finer grain of steel in 

 the head, and having from a third to a half 

 superior durability to the usual rail. 



Lieut.-Col. Davis, of the Naval Ordnance Bureau 

 of the United States, has produced an armor- 

 plate which, when tested at the proving-grounds 

 at Bethlehem, gave results encouraging the belief 

 that the armor-plate has again overtaken the gun 

 in the struggle for supremacy. This plate is ob- 

 tained by a novel process, carbon being driven 

 directly into the surface of the hot metal by an 



