been published in a convenient form in a volume entitled Cast Iron-A Record of Original Research, which should be read by every one interested in the scientific aspect of iron-founding.

Mr Keep has pointed out that white iron frequently does not give sound castings, and that blowholes lead to lower specific gravity and diminished strength; but a small addition of silicon eliminates blowholes and produces sound castings. As soon as the metal is sound the greatest crushing strength is obtained. This condition gives the maximum density; further addition of silicon leads to the formation of graphite and diminished brittleness, and gives the greatest transverse and tensile strength. When the graphite is further increased, the metal is divided by the interspersed graphitic

material, which is now in a coarser state of aggregation, and the strength and hardness of the casting are decreased. The deflection also increases with the increase of graphite, but when the maximum separation of graphite has taken place, any further addition of silicon causes stiffness or brittleness.

Mr Keep also drew attention to the important fact that white iron shrinks more during solidification than grey iron, and hence on adding silicon to white iron the shrinkage is diminished. The amount of shrinkage closely follows the proportion of combined carbon in cast iron-so closely, indeed, that it is found in practice that the hardness of the metal may be judged with considerable accuracy by a shrinkage test. Hence a shrinkage test acts as a control on the proportion of graphite, and indirectly also on the percentage of silicon. When discussing the influence of silicon on

EFFECT OF SILICON ON THE PROPERTIES OF CAST IRON.

...

2

4

0 7.560 7.719 72 22,720 10.14 25,790,000 168,700 75.30
0.5 7.510* 7.670 52 27,580 12:31 28,670,000 204,800 | 91.42
7.641 7.630 42 28,490 12.72 31,180,000 207,300 92.54
1.4 7.555 7.473 31,440 14 04 23,500,000 183,900 82.08
7.518 7.350 22 35,180 15.70 23,560,000 137,300 61.29
2.5 7.422 7 388 22 32,760 14 62 25,450,000 172,900+ 77.18+
3 7.258 7.279 22 27,390 12.23 21,150,000 128,700 57.45
7.183 7.218 27 25,280 11.28 15,640,000 106,900 47.74
5 7.167 7.170 32 22,750 10 16 18,720,000 103,400 46.16
7.5 7.128 7.138 42 11,950 534 14,750,000 111,000 49.55
10 6.978 6.924 57 10,630 4.75 13,930,000 76,380 34.10

Lbs. Tons.
2,702 1.206

3,280 1.464

1.98 0.38 1.60 0.19 0.32 0.14 0.05
2.00 0.10 1.90 0.45 0.33 0.21 0.05
3,370 1.504 2.09 0.24 1.85 0.96 0.33 0.26 0.04
3,498 1.561 2.21 0.50 1.71 1.37 0.30
3,446 1.538 2.18 1.62 0.56 1.96
3,534+ 1.577+ 1.87 1.19 0.68 2.51 0.26 0.75 0.05
2,850 1.272 2.23 1.43 0.80 2.96 0.34 0.70 0.04
2,543 1.135
2.01 1.81 0.20 3.92 0.33 0.84 0.03
2,342 1.046
2:03 1.66 0.37 4.74 0.30 0.95 0.05
1,505 0.672 1.86 1.48 0.38 7.33 0.29 1.36 0.03
1,252 0.559 1.81 1.12 0.69 9.80 0.21 1.95 0.04

0.05

...

0.28 0.60 0.03

* This number is rather low, as the specimen afterwards proved to be somewhat faulty.

+ The value in this case is probably exceptionally high; a crushing strength of about 60 tons might be anticipated from its position in the series.

cast iron, it is customary in some quarters to speak as though the only effect exerted by silicon is that due to the separation of graphitic carbon. Professor Howe, for example, regards cast iron as a mixture of graphite and a metallic matrix, and its strength as dependent on the relative absence of the former, and the approximation of the latter to steel, and has elaborated this view at considerable length, and with characteristic care and ability. (I. & S. Inst. Jour., 1903, i. p. 706). But it is important to recognise two other effects produced by silicon in addition to its power of separating graphitic carbon. In the first place, both the tensile strength and hardness of carbon free iron are increased by the addition of silicon, which acts in this respect much like combined carbon in iron or steel, but in a less energetic manner.' Excess of silicon, however, like excess of carbon, causes the metal to become crystalline, so that rich ferro-silicons are sufficiently brittle to break easily when struck with a hammer, and can be readily crushed to powder. Secondly, it is noted that the size of the graphite is influenced by the judicious addition of silicon, since what may be termed freshly precipitated graphite, obtained by adding silicon to a hard iron, is smaller in size than that which exists in an ordinary soft foundry iron, and consequently the metal is closer in texture and stronger when, by the addition of silicon, a portion of the graphite has been thrown out of solution in this form. We are therefore justified in claiming a specific influence for silicon itself in addition to that which it exerts upon the state of combination of the carbon present. In order to accurately understand the constitution of cast iron, it is necessary to consider not only the total amount of graphite, and the proportion it bears to the combined carbon, but also the texture of the graphite and the effect of other elements on the matrix in which the graphite is found. It will be seen, when discussing the influence of the other elements upon cast iron, that each has a specific and beneficial influence when present in suitable proportions, though the amount which can be introduced with advantage is very much less with certain elements than with others.

Silicon in the Foundry. From what has already been said, it will be evident that, with pig irons of what may be regarded as good or normal composition, the regulation of the proportion of silicon is the

1 Additional proof of this statement will be found in the results published since these lectures were delivered. (See T. Baker, "Silicon in Iron and Steel "— I. & S. Inst. Jour., 1903, vol. ii.

LIBRARY

OF THE

UNIVERSITY

OF

CALIFORNIA

key to successful foundry practice. The question then arises as to the most economical and convenient method of obtaining in practice the required proportion of silicon. In special cases it has been found advantageous to use ferro-silicon containing 10 per cent. of silicon and upwards, and this method was adopted with advantage, particularly in France, in 1886. Ferro-silicons, however, possess two disadvantages; in the first place, they are relatively expensive, and are sold at prices varying with the silicon content. Further, they differ considerably both in fusibility and density from the iron with which they are to be mixed, and as a result the product is apt to be irregular in composition. When using a mixture of irons which differ considerably in an analysis, there is also the irregularity caused by relatively small errors in the order, or the time, of charging the metal into the cupola. On the whole, therefore, it is found more economical to use a mixture of such irons as are commonly met with in practice, and this method of mixing has the advantage of ensuring greater uniformity both in the supply and in the product. The ideal method is for the founder to have a fairly large stock, including several kinds of iron, each separate kind being a little too hard, or a little too soft, for the general run of work, but still not very different from what is required. By mixing these irons in suitable proportions, it is then easy to obtain any composition which may be desired, it being, of course, assumed that the composition of each variety is already known.

Sulphur.

The proportion of sulphur which is met with in foundry irons is usually small, and while the carbon and silicon are expressed in percentages, the sulphur is returned by the analyst in 100ths, or at most 10ths of one per cent. It may be noted as a curious fact that American irons are generally much freer from sulphur than are many varieties produced in the United Kingdom. Despite the small proportion of sulphur which is usually present in foundry irons, its influence is so energetic that it deserves special attention, since relatively small variations may entirely change the character of the product. Briefly stated, the effect of sulphur is to counteract that of silicon; it leads to the production of combined carbon, and consequently to increased shrinkage and hardness. On the other hand, metal which is unusually free from sulphur is apt to be somewhat too soft, and a little wanting in strength and rigidity. It is not

possible to give any figures which would be of general application, but the iron-founder would usually be safe in considering that one part of sulphur will neutralise the useful effect of at least 10 parts of silicon. For soft fluid castings the sulphur should be as low as possible, or say about 03 to 04 per cent. For strong metal about 07 per cent. gives good results; but it is doubtful if it is advantageous under any circumstances to introduce more than about 0.10 per cent. sulphur into foundry mixtures. The sulphur which is present in cast iron exists as sulphide of iron, which is readily fusible, and is probably the last constituent to solidify as the metal cools. As a result, the sulphur is apt to be very unequally distributed in a mass of iron, sometimes being found chiefly in the upper and outer part of the pig, as when much manganese is present; but in other cases the sulphur collects chiefly in the centre of the casting. Considerable care is therefore necessary, in order to obtain a fair average result, when sampling pig iron for the purpose of analysis.

Phosphorus.

It has already been mentioned that the whole of the phosphorus which is present in the ore charged into the blast furnace is reduced and passes into the pig iron, and that, consequently, the content of phosphorus from a given ore mixture is approximately constant. (See page 13.)

Phosphorus exists in cast iron in the form of phosphide, which is readily fusible, and which in its pure state is brittle and comparatively hard. It has been shown by Stead' that more than one phosphide exists in iron and steel, though the one usually met with has a composition corresponding with the formula Fe,P. This is dissolved by the iron until the proportion of phosphorus reaches about 1.70 per cent., after which the phosphide separates in the free state. Carbon tends to throw the phosphorus out of solution, though this action is not complete even with 3.5 per cent. of carbon. It will be seen that a grey phosphoric iron, when viewed under the microscope, consists of three parts: (a) the uniform ground or base consisting of iron, silicon, and manganese; (b) graphite; (c) the phosphorus eutectic. (See fig. 50.)

A small quantity of phosphorus when present in pig iron is found 1 I. & S. Inst. Jour., 1900, vol. ii. p. 60.