Page images

open furnaces with wide tops. It thus became at once obvious that cylinders in Wales had been productive of less injurious consequences to the general working of the furnace than those in Staffordshire, because the greater width of the Welsh tops had permitted cylinders of from 8 to 10 feet to be employed, whilst in Staffordshire only cylinders of much less size were practicable.

The important effect produced on the working of the furnace, chiefly by an alteration in the arrangement of the materials in the furnace, is a point of considerable interest, but one to which little attention has been hitherto paid. In practice it has been long known to the best managers of furnaces, that wide tops were desirable, and generally accompanied by increased make; but the precise manner in which wide tops acted was not clearly known until the attempt to use the waste gases led to its evident explanation. On the Continent, the importance of such arrangement of materials as would facilitate the passage of the blast, as nearly towards the centre of the furnace as practicable, has been known for some time and acted upon; and the writer was much pleased to find from M. Tunner, Professor of Metallurgy, in Austria, and connected with the Styrian Iron Works, that great increase of make had followed the adoption of wide tops to the charcoal furnaces of that district, combined with a method of filling, by which the coke or charcoal was placed in the centre of the furnace, and the ore and limestone around the sides.

Some few months back a furnace was placed in the writer's hands, which he found provided with a cylinder and other arrangements for taking off the gases. Although apprehensive that the cylinder would materially interfere with the working of the furnace, yet, as everything was arranged for it, and as it was six feet in diameter, he determined to blow it in without alteration. This was done-the expected result following, -constant slipping, and fretting tuyeres, with all their attendant bad results. The stacks were not powerful enough to draw off the gases, unless a closed top was used; and the writer therefore adopted an arrangement somewhat similar to the Ebbw Vale one. The result was immediate; the furnace worked with great regularity, and carried a good burden; but white iron alone was produced. The burden was lightened, but the iron remained white. A farther lightening of the burden was made; but although the cinder was exceedingly gray, still the iron was white. It became evident that a greater proportion of coke would not produce the desired change, and was in fact useless. The white iron was evidently the effect of the closed top. A pipe of nine inches diameter was inserted at the filling place, but with no effect. Another pipe was inserted, and some little change appeared.

It being important to produce gray iron, it was now determined to sacrifice the use of the gases entirely, rather than continue to make white iron. A lid or valve upon the main gas-pipe, and a covering on the top of an auxiliary gas-pipe were now opened to the atmosphere, and a decided change was at once evident. The iron became gray and the furnace worked with regularity. The white iron had evidently been caused by the pressure produced by the closed top; and so extremely sensitive did the furnace appear to be to the slightest restraint, upon the free passage of the gases, that even a strong wind blowing into the open

box, through which the gases were principally escaping, would throw the furnace back to white iron.

In Wales, where the closed tops are successfully employed, the production of white iron is rather sought for, and hence the tendency in closed tops to produce that quality is no disadvantage. In many cases, however, it must be a fatal objection to their use. But for this objection closed tops would become universal; as they entirely do away with the necessity for a lofty stack, and enable all the gases to be economized. That the tendency to produce white iron has no connexion with the mere abstraction of the gases from the furnace is clearly shown by the results of the Scotch furnaces, in which they are taken off without employing closed tops. At Dundyvan, especially, great attention has been paid to this point; and the result has been shown to be that the furnaces from which the gases are taken, work with equal regularity, and produce gray iron with equal facility to those from which the gases are not taken; and it would be easy to multiply instances of the same result.

To be Continued.

For the Journal of the Franklin Institute.

On the Analysis of Cast Iron. By Professors CAMPBELL MORFIT and JAMES C. BOOTH.

(Extract from the Official Report to Captain L. A. B. Walbach, U. S. Ordnance, on Foundry Service, upon the Examination of Gun Metal.)

The associated substances always occurring with the ores, in the furnace, are silica, lime, and alumina; in addition to which, sulphur, phosphoric acid, oxide of manganese, magnesia, and the alkalies are rarely absent. Substances of more unusual occurrence are oxides of chronium, tin, titanium, vanadium, copper, arsenic, and some others. Now, although most of these substances are reduced to their metallic state with great difficulty, by the fuel and high heat of the furnace, yet their affinity for oxygen is weakened, and the contact of a large quantity of metallic iron adds its influence to reduce portions of them to metal, in which state they enter into combination or admixture with the iron. Hence crude cast iron or pig iron, besides its essential constituent carbon, contains more or less of the elements of the ingredients above named, though always in small proportion to the iron, and, in their total amount, rarely exceeding eight to ten per cent.

The influence of these associates upon the character and influence of cast iron has been determined to a limited extent, and in the case of only a few of them. Thus, a considerable proportion of manganese is said to impart a large lamellar and brilliant fracture and some degree of brittleness; phosphorus is supposed to render iron "cold short," and sulphur, "hot short." But the full extent of influence of these few is yet undecided; while the influence of most others is either unknown or at least subject to great doubt.

Although most of the above are only occasional constituents, and exist, when present, only in minute quantities, still it was expedient that VOL. XXV.-THIRD SERIES.-No. 3.-MARCH, 1853.


the examination should be made with a view to the detection of each and every one of them.

Pulverization. It was found, after repeated trials, that the only sure way of obtaining samples of uniform composition was to clip off pieces of the size of 4th to th of an inch diameter, by means of a cold chisel and hammer. These were taken from the small broken cylinders, which had been bored from the muzzle of the gun, and had served for the mechanical tests. Reduction under the pestle and sifting should be avoided; for any process, which requires that the metal shall be acted upon in a state of powder, is defective; because the slag and graphitic carbon being, as it were, mere mechanical components pervading the mass of iron and not chemically combined with it, form finer particles by pulverization, which from their low specific gravity fly off partly as dust. This loss and the difficulty of reducing the metallic granules to the same state of fineness as the siliciuretted and carburetted constituents of this metal, prevent the possibility of obtaining a mixture sufficiently uniform to afford a fair average sample. We found that equal portions of the same powder gave under the same circumstances variable results, whatever solvent was employed. Thus one gramme digested in nitro-muriatic acid, gave of insoluble residue, in two sets of experiments similarly conducted:

[blocks in formation]

These were selected indiscriminately from more than twenty experiments, all of which yielded results that are equally conclusive. In several trials, the proportion of insoluble residue obtained from the samples of dusty metal, presented a greater variation than those from the clippings or even the fine granules. The discrepancy arises from the facts just mentioned. It is also because the mixed metal, in its finest state of comminution, is peculiarly sensitive to a decomposing or transforming action of the digesting liquids, which, instead of being confined solely to effecting solution of the metallic bases, extends to the transmutation of some of the other constituents into soluble matter, as will be treated in the paragraphs relating to the estimation of carbon, slag, and silicium. It is therefore imperative that, first, in reducing the metal mechanically, it must be to clippings, and not to powder. The size should be from 3th to 16th of an inch diameter, as this bulk is better adapted to insure a gradual reaction and uniformly exact results.

Total Amount of Carbon.-The important bearing of this ingredient upon the quality of cast iron renders its accurate estimation of paramount importance; for the metal is classified into gray and white varieties, according to the condition in which the carbon exists in it. For instance, in gray iron there is both combined and free (graphitic) carbon; while in the white the carbon is wholly or nearly all combined with the metal. Now, as the only way of accurately determining the former condition is by deducting the graphitic from the totality of carbon, it is essentially important, that these last two should be estimated with great precision. The

most usual method of determining the whole of the carbon is to ignite five grammes of the finely powdered metal, in a combustion tube, with a mixture of anhydrous chromate of lead and chlorate of potassa, or with oxide of copper. By the aid of high heat and the oxygen generated by the decomposition of those highly oxydizing bodies, the carbon burns off as carbonic acid, and is condensed by caustic potassa contained in a bulbed receiver. The increase of weight which the latter acquires by the process of combustion represents the amount of carbonic acid absorbed, whence the total amount of carbon is calculated. This plan, although expeditious, is liable to many objections; the chief of which is the necessity of having the metal in an impalpable powder, in order to effect its complete combustion. This entails the disadvantages enumerated in the paragraph upon PULVERIZATION. It will also be necessary, in most cases, to repeat the ignition in order to check doubtful results.

Another and very accurate method is that proposed by Berzelius, and which consists in depositing a weighed lump of five grammes of metal upon a fused cake of 30 to 40 grammes of chloride of silver, in a vessel containing water and closely covered to prevent access of air. A little hydrochloric acid is added to promote the action. The silver salt gradually decomposing, gives up its chlorine to the bases of the metallic iron, which by the union, become soluble in the water as chlorides, while the eliminated carbon, silicium, and slag subside to the bottom upon the silver cake, which is more or less reduced to the metallic state. The liquor being filtered off through a small paper, the undecomposed portion of chloride of silver is removed with a platinum pointed pincette, freed of adhering particles by means of the wash bottle, and reserved for use at another time. The contents of the filter, while still moist, are to be rinsed out into a beaker, treated with dilute nitric acid and gently warmed to dissolve the reduced silver powder and any traces of oxide of iron.. Great care must be observed to provide against too much heat, so as to prevent portions of the carbon, from being converted into soluble artificial tannin. When the whole of the silver and oxide of iron has been taken up, the liquor is filtered on a balanced filter, which is washed with hot water, dried first in a hot air chamber and finally in vacuo, and weighed. The gross amount of carbon, silica, and slag is thus obtained. By ignition the whole of the carbon is burned off, and the amount determined by weighing the calx. The difference between the two weights, which is the amount lost by the ignition, represents the carbon.

Solution of chloride of copper has been suggested as a substitute for the solid chloride of silver, and its action is more energetic and rapid; but the objection to the plan is, that a small portion of the carbon is lost as carbo-hydrogen gas.

The above plan has the advantage of giving the total carbon in one operation; but it is tedious, time consuming, and only manageable with difficulty. Ten to twenty days are necessary for the perfect completion of the operation; and if any air should accidentally enter the digesting vessel, the consequent formation of hydrated peroxide of iron is a serious impediment to the progress of the reaction. The traces of hydrochloric acid, with which the water may be acidulated, tend to prevent the generation of this oxide, but it must be used sparingly, since, otherwise,

[ocr errors]

by taking up some of the metalic iron, a part of the carbon might be driven off. Free chlorine has also been proposed by Berzelius, but its use is less practicable than the preceding method; because bulbed glass apparatus, requiring nice arrangement and careful management, must be employed by his mode. There is, moreover, difficulty in volatilizing all the chloride of iron formed, and also in preventing the access of air, which by its oxygen, will convert some of the carbon into carbonic acid or oxide, or both.

Another process by the same distinguished chemist consists in subjecting the iron in filings to the spontaneous action of dilute but perfectly pure nitric acid added portionwise. It is time consuming, and requires great attention to insure uniform reaction; for much heat is eliminated, and may raise the temperature of the mixture above 125° Fahr., and thus promote the transformation of a portion of the carbon into soluble matter, as mentioned in the paragraph relating to graphitic carbon. Moreover, a basic salt of iron, soluble with difficulty in cold acid, is likely to be formed. Besides, if any of the foreign alloys are convertible into insoluble oxides by the nitric acid, they remain with the carbon and silicium; and in the after-separation of the former from the latter, by ignition, they may by entire or partial volatilization with the carbon, lead to error in its estimation.

Berthier's process of estimating the total amount of carbon, by slowly oxidizing the iron by exposing it to the air and moisture, is tedious and tends to error, from the fact, that the hydrochloric acid required in the after part of the operation, to dissolve out oxide of iron, as well as the subsequent evaporation to dryness to render the silica, &c., insoluble, subjects the carbon to a partial conversion into soluble humus.

After repeated trials and due consideration of the above and other , processes for estimating the total amount of carbon, we were convinced that none of them possessed that union of speed and accuracy which is necessary to a satisfactory performance of the analysis. We therefore abandoned them, and proceeded to a course of experiment, with the view of elaborating a process which would be rapid as well as reliable; and succeeded in developing the following method.

2d. One gramme of metal clippings, having been accurately weighed, is placed in a small beaker with the atomic proportion of iodine, say, 5 grammes of iodine to one of metal; a little water is added, and the glass with its cover on, left at rest in the cold. Five to six hours suffice for completing the solution of the bases; and the rapidity of action is inversely as the state of dilution of the liquid, so that only enough of water to cover the mixture should be employed. Perfect solution is recognised in the dark brown liquid by the flocculent character of the insoluble matter," as compared with the heavier and metallic particles; and if any of the latter still remain unacted upon, they will subside heavily to the bottom. In such case, a few more grains of iodine must be added, if the first portion has entirely dissolved, and the digestion arrested by filtration, as soon as the solution of the metal has been accomplished, for a prolonged action of the iodine, as well as too great dilution promote the oxidation and volatilization of the carbon.

Great care must be observed to temper the reaction, which, when violent,

« PreviousContinue »