Page:The American Cyclopædia (1879) Volume IX.djvu/417

 IRON MANUFACTURE 403 suitable quality, and the conversion has been uniformly effected. The following analyses by Dr. R. W. Davenport show the progress of decarburization : ELEMENTS. Original casting. After annealing. After second annealing. Silicon 0-445 0-488 0-449 0-815 0-S-.27 0-815 0-529 0-585 0-525 0-059 0-067 o-osi Carbon 3-430 1-510 0-100 These analyses likewise show that the process is simply one of decarburization, and that the other elements present are not materially af- fected. Dr. Davenport noticed in a casting J in. thick, which had been converted to the depth of in., that the remaining in. was darker in color than the original iron used, and analy- sis showed that it contained graphitic carbon. The separation of carbon from combination may therefore possibly be the first step in the process of conversion. Within the last 30 years there have been a great number of attempts to make wrought iron and steel direct from the ore without the use of the blast furnace. These direct processes differ from the bloomary process, in which the same result is accomplished, by the reduction of the ore at a temperature below fusion, and the formation of iron sponge. The sponge thus formed contains nearly all the iron in the me- tallic state besides the earthy ingredients of the ore unaltered. To remove the earthy mat- ters and consolidate the iron, it is worked up to a bloom in a reverberatory or gas furnace, or in a bloomary hearth, and then hammered or rolled in the usual way. The process of re- duction or sponge making is very simple. The ore is either mixed with the fuel (preferably charcoal) in a cylinder or cupola furnace and heated to redness, or is exposed to a current of hot carbonic oxide gas. When reduction is ended, the sponge, which oxidizes readily owing to its porous condition, should be cooled in a reducing atmosphere before removal. If the process of reduction has been sufficiently prolonged, and the sponge fully cooled before removal, the product should contain at least 95 per cent, of iron in the metallic state. A great loss of iron has hitherto been experi- enced in working up this sponge, especially when made from poor ores containing consid- erable silica. In balling the iron in a puddling furnace, the loss of iron is excessive, owing }o its porous condition. This has been the principal cause of the failure of nearly all the modern direct methods for making wrought iron. Chenot's process, which 20 years ago seemed to be successfully established as a met- allurgical and economical process, and which furnished a large amount of iron and steel of excellent quality, is now abandoned. The same fate has befallen Yates's, Renton's, Gurlt's, and other processes which promised success. A new application of iron sponge has been found in steel making in the open- hearth regenerative furnace ; and it is now probable that iron sponge will find an eco- nomic application. In the Martin process for making steel or homogeneous iron (see STEEL), wrought-iron scrap is added to a bath of molten pig iron until the percentage of carbon remain- ing, in consequence of this addition, is very low, or, with the aid of an oxidizing flame, even perfectly removed. Iron sponge added in this way to a pig-iron bath loses iron to the extent merely of saturating the silica it contains in the formation of a cinder. In very pure ores this loss is consequently very small. It is in this way that Mr. Thomas S. Blair of Pitts- burgh successfully utilizes the sponge made by his process, which is one of great simplicity, on the principle of Chenot's, with certain im- provenlents in apparatus. His reduction cyl- inders are 40 ft. high and 3 ft. in internal diameter. The upper half is exposed to a bright red heat from burning gas on the out- side, and the lower half is provided with a water jacket for cooling the reduced product. Into the top of the cylinder is inserted a thim- ble of cast iron 6 ft. long and 28 in. in di- ameter, leaving an annular space of 4 in. be- tween it and the cylinder. Into this space we charged charcoal in small fragments and pow- der, and ore in pieces not larger than an egg. The carbonic oxide resulting from the reduc- tion of the ore burns inside the thimble, and the gas from the producers outside the cylin- der. The charge thus becomes rapidly heated through in this narrow space, and when it spreads over the whole diameter of the stack, 6 ft. from the top, it is all red hot. At the bottom of the stack there is a sleeve which when raised allows the already thoroughly cooled sponge and the charcoal charged in excess to run out. The sponge is drawn at regular intervals, and in the mean time the sleeve is luted with clay. No air gains access to the stack while drawing, as the column of finely divided iron and coal forms an effectual packing. The sponge is separated as far as practicable from the charcoal and compressed by hydraulic pressure into ingots, which are added directly or after previous heating to the bath of metal in a Siemens or other form of re- generative furnace. Owing to the great sim- plicity of the process, iron sponge is a much cheaper product than pig iron. Siemens has invented a number of sponge processes, which are all connected with the use of his regen- erative furnace. None of them have ever come into general use. Since the introduc- tion of Danks's puddler he has employed a rotary cylinder similar to this for the direct production of iron. The chamber is lined with a refractory material and heated on the regene- rative principle. The ore is first melted, then the charcoal or coke is added, and the vessel ro- tated. Reduction takes place energetically, and a ball of soft iron is speedily formed. Siemens