87G STEEL.— Open-Hearth Process.

been used, since all the carbon, silicon, and man-ganese are rapidly burnt out of the pig iron. As inthe Bessemer process, however, the metal at thisstage will not work, being highly red short. Inorder to produce a malleable product ferro-man-ganese is added if very soft steel be required;spiegeleisen, if such as rail or tool steel is wanted;and a mixture of ferro-manganese and siliceous pigiron (or a special alloy containing silicon and man-ganese, which may be termed inangan-siliconeisen),when steel castings are to be made. With 6-toncharges from 70 to 100 tons of ingots per week canbe made from one furnace.

In the “ore” process the charge is nearly all made ofpig iron. About 5 tons of No. 3 hematite are melteddown, and about 1 ton of iron ore, free from phos-phorus and sulphur, broken to the size of hen eggs,is added to the charge in small quantities at a time,keeping up a regular boil, until all the carbon, silicon,and manganese have been removed by the reactionbetween the oxygen of the ore and these constituentsof the pig iron. As in the previous case, smallsamples are taken from the bath from time to timeand tested by the hammer and an examination ofthe fracture, until the metal is found to be malleable,and the fracture shows a lead-like crystallization.The spiegel or ferro-manganese is then added, andthe metal tapped out rapidly to prevent waste ofthese latter substances. From 20 to 25 cwts. of oreare required for a 5-ton charge, and, according toHackney, “ of the metal contained in the ore aboutone-half passes into the steel, so that, as the carbonand silicon in the pig amounts to about 8 per cent.,the yield of ingots and scrap is generally 1 to 2 percent, more than the weight of pig iron and spiegel-eisen put in. The consumption of coal is slightlylarger than in the scrap process, as there is moreslag on the surface of the metal, and hence the flamemust be maintained at a higher heat. The wear ofthe furnace roofs is also somewhat greater than inworking scrap, and for the same reason, the moreintense heat of the flame; and in addition to this,the splashing of the slag in the ore process helps towear away the roof. A careless furnaceman, withplenty of heating power at command, may bring thefurnace roof down in a day; but in practice andwith reasonable care, a roof of good bricks will last,even in working pig iron, from one hundred to twohundred charges. The two methods of workingabove described are often combined; that is, if thereis a limited quantity of scrap to be worked, it isdivided among the furnaces, and the charges arefinished by adding ore.”

Speaking of the Pernot furnace Hackney says: —“The working of the P ernot furnace at Saint Chamondand Allevard differs in many points from that of thefixed furnace. In melting together pig iron andscrap, the former, previously heated to redness inan auxiliary furnace, is spread uniformly over thebottom, and upon this the whole quantity of steelrail ends, or other scrap employed, is charged atonce, and the furnace is left to itself. As the bedkeeps constantly moving round, each piece of scrap |

is alternately exposed to the full heat of the flame,and dipped into the bath of melted metal that soonbegins to form. The fusion is thus very rapid;the whole mass is liquid within two hours ; samplesare then taken out at intervals, and when the metalis sufficiently soft, spiegel is added, and the charge istapped. Five charges of about 4J tons each are gotout in each twenty-four hours. The proportion ofmaterials in the total charge is—pig iron, about 24per cent.; steel scrap, 67 per cent.; and spiegel, 9per cent. The loss is between 6 and 7 per cent.,and the coal used per ton of steel made is 8 to 8Jcwts. Comparing with this the average workingduring two months of furnaces of the same design,but with fixed beds, at Terre Noire, about the sameproportions of materials are used, and the waste isthe same ; but the output, instead of being nearly11 tons per twelve hours, is little more than 5-g- tons,and the coal used per ton of steel is nearly 18 cwts.”

The following table, by A. Willis of Landore,exhibits the rates of oxidation of carbon, silicon,and manganese in the open-hearth process:—

Samples Taken.

Contained—

Carbon.

Silicon.

Manganese.

Per Cent.

Per Cent.

Per Cent.

When pig was melted,...

1-90

•570

1-140

One hour later.

1-80

•233

•576

Two hours later,.

1-70

•183

■200

Three “ .

1-65

•050

*080

Four “ .

Five “

1-60

1*10

1 Mangan

ese and si ii-

Six “ .

•60

^ con

completely

Seven “ .

■20

j gone-

It is evident from the above that a considerableamount of oxidation must have gone on during theprocess of melting down, as is the case in meltingpig for puddling, since the iron would have con-tained more than 1'9 carbon and 0-57 silicon. Thisis to be expected, and it is on this account thatthere is not the same advantage in starting withmolten pig iron direct from the blast furnace asthere is in the Bessemer process; for oxidationgoes on more rapidly during the time the pig is beingmelted down than when it is all fluid, owing to themuch greater surface exposed by the lumps of pigiron than by the fluid metal.

Many attempts have been made to obtain irondirect from the ore as iron sponge, and to melt thisdown to form steel. Chenot, in France , carried onthe direct reduction for a considerable time, andmet with a fair amount of success, but failed onaccount of the difficulty in securing perfect reduc-tion. Blair has carried the direct reduction a stepfarther, and has succeeded in making a considerablequantity of steel by melting the sponge (producedas described in the article on Iron ) in an open-hearth furnace at Pittsburg; but so far the planhas not been commercially successful.

Siemens has attacked the problem in a differentway, and has taken out a series of patents for whathe terms “The direct process.” The productionof iron by this plan was described also in thearticle on Iron , and it is only necessary to treat the