METALLURGY. 



more quickly converted. Then, while the sand 

 damages the hearth and requires additional flux, 

 the sandless pig is harmless, and the flux required 

 is a minimum, thus saving material as well as wear 

 and tear ; consequently less fuel is required to con- 

 vert the sandless pig into steel. This results in an 

 increased output. The same holds true in remelt- 

 ini; in the cupola for Bessemer steel. The sandless 

 pig requires less flux and less fuel, melts more 

 quickly, and results in appreciably less wear and 

 tear in the cupola. 



Kxperiments upon the loss by oxidation of cast 

 iron when melted in a cupola were made by 

 Thomas I). West in a " twin-shaft " cupola on sand 

 and chill castings, and also on chill castings pro- 

 tected by various coatings or washes. It was found 

 that the sand castings oxidized in melting to a 

 greater extent than either the protected or un- 

 protected chill or sandless castings. The average 

 losses were: Sand iron. 5.8 per cent.; sandless iron, 

 ::.4 |x>r cent. ; chilled iron, with lime wash, 3.8 per 

 c.nt. ; with graphite wash, 3.4 per cent.; with sili- 

 cali' of soda wash, 2.9 per cent. 



Mr. David Baker, of Chicago, has devised an im- 

 proved casting plant for blast furnaces in which 

 the iron is run into ladle cars at the furnace, then 

 hauled to the casting plant, where the metal is 

 poured into moulds and delivered into cars by 

 gravity, without any adhering sand and with a 

 minimum loss in scrap. The iron goes into the 

 cars cooled and ready for shipment. 



A method for determining the hardness of cast 

 iron described by Charles A. Bauer is based on the 

 fact that, with a drill running at a uniform rate of 

 speed and under a constant pressure, the number 

 of revolutions required to drill a given depth would 

 be directly proportionate to the hardness of the ma- 

 terial subjected to the testing. 



Hot metal direct from the blast furnace is used 

 in the open-hearth furnace at the Homestead Iron 

 Works, Pennsylvania. The operation of casting the 

 metal at the Duquesne furnaces, transporting it in 

 ladle cars by the Union Railroad 5 miles, mixing, 

 reheating, recasting in ladles, and charging the 

 open-hearth furnace, is performed in about one 

 hour. Coke dust is used to retain the heat and 

 cover the metal while in transit. Each ladle is 

 fitted with an electric tipping apparatus which 

 turns the ladle and discharges its contents into 

 an immense cask-shaped mixer of 250 tons capa- 

 city. The metal is run from the blast furnace and 

 into the mixer in forty minutes, and when ready 

 for the open-hearth furnace the mixer is turned 

 and runs the iron on the opposite side into 20-ton 

 ladle cars*, which are moved to the No. 2 open- 

 hearth furnaces. The iron is then run through 

 a spout into the furnace in ten minutes from the 

 time it leaves the mixer. 



Prom experiments on the effects of adding phos- 

 phorus to iron, especially in reference to the 

 strength and fusibility of the metal, Thomas D. 

 West learns that the strength of Bessemer iron is 

 increased from 25 to 75 percent., and that the fusi- 

 bility of solid iron is greatly increased. 



Krom experiments conducted during several years 

 to determine the bursting stress of cast-iron cylin- 

 ders under water pressure, C. H. Benjamin" has 

 drawn the conclusions that cast-iron cylinders of 

 the form ordinarily used for engines, subjected to 

 internal pressure, are as likely to fail by tearing 

 on a circumference as by splitting; that by reason 

 of local weakness and distortions the cylinder may 

 fail when the stress, as calculated by the ordinary 

 formula for thin shells, is only about one third of 

 the strength shown by a test bar ; that the prin- 

 cipal cause of weakness is the sponginess of metal 

 due to uneven cooling ; and that to insure good 



castings the flanges should not be materially 

 thicker than the shell, the cylinder should be cast 

 on end, and suitable " visers " should be provided 

 for the escape of dirt and gas. 



The results of tests of cast-iron columns by Prof. 

 W. H. Burr, of Columbia University, go to show 

 very irregular variations, and the author concludes 

 his report by saying that " they constitute a revela- 

 tion of a not very assuring character in reference to 

 cast-iron columns now standing, and which may be 

 loaded approximately up to specification amounts. 

 They further show that, if cast-iron columns are 

 designed with anything like a reasonable and real 

 margin of safety, the amount of metal required dis- 

 sipates any supposed economy over columns of mild 

 steel. As a matter of fact, these results conclu- 

 sively confirm what civil engineers have long known, 

 that the use of cast-iron columns can not be justi- 

 fied on any reasonable ground whatever." 



A series of fire tests of cast-iron columns, made 

 at Hamburg, was applied to columns lOi inches 

 in diameter, and of 1.13- or 0.5-inch metal. They 

 were loaded centrally and eccentrically, and some 

 were cased with a fire-proof covering. A hydraulic 

 press was placed below the column with its cross- 

 head above it, and a hinged oven, containing 12 

 large gas burners, was clamped about it. On an 

 average, a load of 3.2 tons per square inch, with 

 a heat of 1,400 F., produced deformation in thirty- 

 five minutes in a centrally loaded column without 

 casing. This showed itself by bulging all round in 

 the middle of the heated part, especially where the 

 metal was thinner : fracture occurred finally in the 

 middle of the thickest part of the bulge. If the 

 load was less, it occurred at a higher temperature. 

 Jets of water had no effect until deformation heat 

 was reached. The casings had the effect of increas- 

 ing the time before deformation began from half 

 an hour to four or five hours. 



In the practice of the crucible process for the 

 production of steel castings at the New Admiralty, 

 St. Petersburg, ferro-silico-rnanganese is used, ac- 

 cording to Mr. Sergius Kern, as the most powerful 

 agent to add to ready-made steel for the production 

 of sound castings, the most useful alloy containing, 

 on the average, 10 per cent, of silicon and 15 per 

 cent, of manganese. Metallic aluminum is of sec- 

 ond importance, especially if good raw materials 

 are. used: an addition of fcrro-aluminum (14 per 

 cent, aluminum) with the silico-manganese is pre- 

 ferred. For steel castings, roundings coining from 

 the punching of steel ship plates for rivet holes and 

 soft-puddled iron are used. The plates contain, on 

 the average, 0.4 per cent, of manganese. In the 

 ordinary run the various castings for shipbuilding 

 contain 0.40 per cent, of carbon, 0.42 per cent, of 

 manganese, 0.35 per cent, of silicon, 0.02 per cent, 

 of sulphur, and 0.04 per cent, of phosphorus. In 

 case of castings intended to resist water or steam 

 pressure, the use of steel stronger in carbon is pre- 

 ferred, and the metal has 0.6 per cent, of carbon. 

 Such a metal is recommended for the different 

 pieces of Belleville boilers, mostly hollow and of 

 capricious forms. 



The temperature of crucible steel prepared in 

 naphtha-firing furnaces at St. Petersburg is given 

 by Sergius Kern as, just before firing for steel cast- 

 ings, 2,200 C. The experiments, Mr. Kern adds, 

 prove that naphtha-burning furnaces for melting 

 steel are far more convenient in comparison with 

 the coke furnaces, and nearly equal to the Siemens 

 gas-crucible furnace ; but are cheaper and handier 

 than the latter when small quantities of steel are 

 melted at a time. 



From experiments made and observed by him, 

 Sergius Kern has learned that ferro-sodium used us 

 a reducing flux in crucible steel makes the slagd 



