Last year this Institute was good enough to accept some remarks by the writer regarding sound steel, entitled Plant for Hadfield Method of Producing Sound Steel Ingots, being a continuation of a research in which for some time the writer has been engaged, and full particulars of which were submitted in papers entitled A New Method of Revealing Segregation in Steel Ingots, and Method of Producing Sound Ingots, read before the Iron and Steel Institute in October, 1912. These papers gave rise to a very interesting discussion, both in England and America.

This question is one of vital importance, as unless sound steel in the form of ingots is first obtained, sound steel in the finished forms of rail billets, rails, bars, sheets, and other forms cannot be expected. The working stresses demanded by modern conditions are so great, each year they probably approach nearer to the limit of safety, that without doubt those who produce steel will be compelled to pay the attention which this subject deserves. Moreover, the problem is not an insuperable one, as shown by the writer in the papers above referred to.

In view of the information which has already been presented in these papers, it is not now necessary to go into details of the system advocated by the writer, except to repeat that the ingots and steel now referred to were produced by employing in the first place sound fluid steel, free from blowholes, this being poured into ordinary ingot molds, provided with the necessary feeding heads, the insulating slag medium, also charcoal and air blast or other suitable form of applying heat to the head portion of the ingot. The upper or head portion of the ingot is kept fluid by the intense heat generated by the air blast and charcoal, so that such fluid steel passes into the body of the ingot below the head, thus feeding and avoiding the shrinkage or piping which would otherwise occur. The result is the production of a sound steel ingot, free from blowholes, segregation and piping, and in which the waste portion, usually termed the discard, is but little over 7 or 8 per cent.

Billets produced from such ingots have shown a salable product of about 91 per cent., including the oxidation losses in heating and rolling. Plate I shows the details of the products obtained from such an ingot. While these results were obtained from a comparatively small ingot, 11 in.

square, weighing 1,680 lb., the same results have been obtained from 14, 15, 18, and 20 in. ingots.

G. Charpy of the Cie, des Forges et Acieries de Chatillon-Commentry et Neuves-Maisons, Montlugon (Allier), also made an ingot weighing 25 tons under this system, and reported to the writer the following results with regard to the system as applied to larger ingots.

M. Charpy stated that the Hadfield method of feeding the upper portion of the ingot in the special manner described in the papers mentioned above, namely, sand head, air blast, charcoal, and insulating layer of slag, appeared to him to constitute a particularly simple and practical solution of the problem met with in steel manufacture, and to overcome the difficulties met with as regards piping; also to reduce segregation and avoid the large percentage of waste material.

In Plate II, Fig. 1 represents a photograph taken by M. Charpy of the upper part of an ingot weighing 25 tons. This ingot has been treated by the writer’s method, its composition being; Carbon, 0.40; sulphur, 0.02; phosphorus, 0.04; manganese, 0.45 per cent. The material also contained a small percentage of nickel.

Fig. I shows the satisfactory’nature of the method. The portion of the ingot shown in the photograph represents about one-fourth of its total length. M. Charpy found that in order to eliminate all faults and sponginess in the ingot in question, it was only necessary to cut off about 12 to 12½ in. from the upper part; in other words, about one-twentieth of the ingot, as against one-third which is the proportion of discard ordinarily required. The waste and segregated material in this ingot only amounted to about 5 per cent, of the total weight.

Moreover, the analyses taken of the material at different parts of the ingot show this is free from segregation.

Fig. 2 shows the parts from which the analyses given in the table were taken.

Drillings were taken on the center line at the different parts shown, commencing at the top, marked A, also at the other parts as lettered.

Fig. 3 shows the dimensions of the ingot.

With reference to the question of soundness of material, it may be interesting to point out that in the smaller ingot shown in Plate I, Fig. 1, from which the billets were obtained, Figs. 2 and 3, even including the sand head portion (termed head or discard in Plate I), the total head or discard was only 2 ft. 5 in. in length. By cutting off 1 in. more, Fig. 4 shows the polished section from the top portion of this billet at the part marked A in Fig. 2; then comes the length of 1 in. removed; above this is the head or discard mentioned, 2 ft. 5 in. in length.

Fig. 5 represents the polished section from the top portion of the billet at the part marked B in Fig. 2. In other words, Fig. 5 shows the last traces of piping near the head, at the part marked B in Fig. 2, while Fig. 4, 1 in. further down, shows the complete removal of this piping, represented by the polished section from the top portion of the billet at the part marked A in Fig. 2. That is to say, the material represented in Fig. 4 and that below this represents sound material.

Fig. 5 also represents the head or discard of this ingot. The ingot as cast weighed 1,680 lb.; the discard removed weighed 123 lb., or 7.3 per cent, of the total weight of the ingot. The total weight of sound billet was 1,513 lb. or 90 per cent.

In a similar manner, from another ingot there was obtained a sound rail having a total length, wholly free from pipe, of 62 ft. 7 in., with a discard of only 18 in. of rail, and in addition to this, the ingot head itself, representing 7 to 8 per cent, of the weight of the ingot.

In order to further illustrate this method, Plate III represents four ingots, A and C made by the writer’s system, B and D ingots made in the ordinary way. Ingot B was cast from slightly rising steel, and ingot D from piping steel but not fed. Ingots A and B, also C and D, were cast from the same heats respectively. These ingots are also shown in vertical position in Plate IV.

Another illustration is shown in Plates V and VI of ingots made in the ordinary way. These represent 18-in. ingots, weighing about 2¾ tons each, cast with the small end up, A being rising or unsound material, and B piped material but not fed.

In comparison are shown Plates VII and VIII, which also represent the same sized ingots, namely 18 in., made by the writer’s system, weighing about 2¾ tons each. These are interesting because the ingots were cast with the small end up, as in ordinary practice. Plate VII gives an excellent view of the cavity produced by the sound steel in these ingots as it settles down into the body of the ingot proper, that is below the feeding head. As will be seen, by a merely superficial examination of such an ingot, occupying but a few seconds, it can be determined whether the steel in the ingot itself is sound or not. If the cavity is there, the material below in the ingot proper will be perfectly sound and free from honeycombs, blowholes, segregation, or piping.

It will be admitted by any impartial observer that it is, on the other hand, quite impossible, by looking at either of the ingots shown in Plate V, made in the ordinary manner, that is without the head or feeding arrangement, to determine whether these are sound or unsound, or the nature of the unsoundness or piping present in them, that is below the top of the ingot. In other words, the ingots would have to be sliced

up by machining, and thus spoiled for actual work, before it could be decided whether they would give perfectly sound material, whereas in the ingots shown in Plate VII, made under the writer’s system, machining or other observation carried out by mechanical methods is entirely unnecessary, and at least 88 to 90 per cent, of sound usable and salable material is obtained. This, too, can be accomplished just as readily and easily with 10,000 as with one ingot.

Moreover, while the two ingots shown in Plate IV possess some difference, the “rising” ingot, shown by A, is probably the better, because by a superficial examination it can be at once determined that such material is unsound. While rolling or forging such an ingot may help partly to

close up not all but some of the blowholes, yet it is not easy to believe such material is as perfect or possesses the same mechanical strength as that resulting from an ingot which has been cast from steel containing no blowholes. Imagine such an ingot as represented by A in Plate V being used to produce, say, a gun forging! This would not be allowed by

any government inspector the writer has ever met, yet in a rail, upon which human life so largely depends, it seems to be taken as a matter of course that such a product may safely pass into service.

As regards steel of piping nature, but cast without a feeding head, such as shown by B in Plate V, this is probably the more dangerous of the two types shown in this plate, as in the case of a warm heat the piping will descend to a greater depth. While a certain portion of this may be got rid of in the discard, this often amounting to a large percentage, yet this means great waste. Moreover, and this is still more important to be borne in mind by the user of the rail, there are cases showing that below the portion where the pipe has been supposed to terminate, this sometimes breaks out again and is not detected. In other words, a discard may be taken and show sound material, yet below this it is easily possible to have further dangerous segregation and piping.

As regards the length of time required for ingots made under the system now described to cool down before they can be dealt with, it is estimated that probably not more than about 15 min. would be necessary.

To further differentiate between the various types of ingots as now produced and as made by the writer’s system, the reader is referred to Plate IX, which gives this in a graphic manner.

Fig. 1 represents badly rising steel, which is full of blowholes almost entirely throughout the mass; the steel rises and shows practically no piping. Although in rolling this steel the blowholes are apparently closed up, without doubt by a careful examination of its microstructure it would be found that the material can never be as strong or as dense as steel made sound in the first instance. The product is, moreover, liable to segregation in the upper portion. A large discard is necessary, in some cases as much as 20 per cent, or more.

Fig. 2 represents steel which neither rises nor settles. In this steel there are fewer blowholes, but the steel is still unsound, with a tendency to pipe. This steel contains considerable unsoundness, and owing to the tendency to pipe, troubles are liable to occur with segregation and piping. Probably at least 15 to 20 per cent, discard is necessary.

Fig, 3 represents settling steel which is not fed in the upper portion, as in the case of those ingots made by the writer’s system. This pipe sometimes runs half the length of the ingot. Moreover, such material has sometimes a tendency to appear sound, then below this to break out again into piping. While this steel is free from blowholes, material of uncertain nature is obtained, owing to its badly piped condition. It is probably this kind of steel which gives the dangerous rails below those made from what are known as the upper or A and B portions of the ingot. In other words, it is quite possible piped or unsound C rails may be obtained from such ingots.

Fig. 4 represents an ingot made under the writer’s system. The steel

used in the first place is perfectly sound and free from blowholes. The feeding arrangement enables from 88 to 90 per cent, of the total weight of ingot to be safely used. In some cases, no less than 93 per cent, of sound billets have been obtained from ingots made in this manner; that is, 93 per cent, of the material showed neither blowholes, piping, segregation, nor other defects.

With reference to the cavities or hollow portions in the heads of the ingots made by the writer’s system, in view of the extraordinarily large nature of these cavities, the writer decided to make the following experiment with a view to determine the exact value of the settling of the steel from the head above the ingot into the ingot itself. He believes that such a method of testing ingots has not before been carried out. It will be understood this has only been made for experimental purposes, because by an examination from the top of the ingot, occupying only a few seconds, it can be easily determined whether the ingots when produced in course of manufacture, whether in small or large quantities, possess the necessary soundness. If the steel was not sound there would be no cavity, so that the quality of the steel is self-apparent.

Nine 15-in. ingots were taken (weighing about 3,600 lb. each) as they came through, each of which had the sand head and the writer’s improved method of feeding carried out on them. After the ingots had cooled down, the hollows or cavities in the sand heads were filled with water, then the water poured out and carefully measured. Table I shows the results obtained.

The average weight for the nine 15-in. ingots showed that 139 lb., with a minimum of 128 lb. and a maximum of 162 lb., passed from the head portion into the ingot itself. This percentage is represented by an average figure of 3.88 per cent. In other words, about 4 per cent., or 140 lb., of the total weight of the ingot or ingots cast passed from the upper or feeding head into the body of the ingot.

There could not be a more striking illustration of the quality of the ingots produced and of the value of this system. Even in the case of the minimum percentage weight of steel which passed from the head of the ingot itself, there is a weight of no less than 128 lb. of steel short in the head, and therefore present in the ingot itself, all adding to the soundness and proper feeding of the piping which would otherwise occur.

To show still more clearly the important information obtained from these experiments, let it be assumed that the cubic capacity of each of the 15-in. ingots in question was approximately 12,500 cu. in. Therefore, but for this feeding, there would be a general want of solidity, chiefly at the upper portion of the ingot, to the extent of say 500 cu. in., say 4 per cent, of the whole capacity. It is surely readily apparent why an ingot which is not fed must perforce be deficient in homogeneity.

It is not claimed that ingots made in the ordinary manner are deficient to the full extent, of say 4 per cent. There is a feeding effect from the steel in the upper portion of the ingot, but it cannot be done efficiently, as the steel quickly freezes on the outside of the mold and on the surface of the liquid steel exposed to the air. Moreover, there is always an uncertainty as to how good or how bad is the resulting material. If the steel is piping very much, the trouble will be worse than when it is piping less. In any case, as the steel solidifies in an ingot of this size, the natural law of contraction demands that about 500 cu. in. has to be dealt with. The writer cannot see how this can be efficiently and cheaply met except by some such method as described in this paper.

While the results necessarily vary slightly, because the sizes of the head portion nearest the top of the mold formed in sand are not always uniform in length, as the steel shrinks down slightly more on the outside in some cases than others, on the whole the maximum and minimum figures of 4½ and 3½ per cent, of the total weight of the ingot having passed from the head into the ingot itself, show very uniform working; if the heads were absolutely the same depth in each case, there would be practically no difference.

Without wishing to exaggerate, it is easy to picture to oneself what would be the character of each of these ingots if not made and treated under the method now described. The piping would have probably run down the ingot itself, requiring a discard of probably 25 to 33 per cent.

Although water cannot be poured into the cavity of a red-hot ingot, yet the cavity can be determined in each ingot by a cursory examination while at a red or yellow heat, involving only a few seconds of time. It will therefore be seen that every ingot can be readily checked by such cursory examination.

While in ingots made in the ordinary way as above mentioned a certain amount of the fluid steel passes from the upper portion to the lower, still in so doing it is robbing the quality of the upper portion of the ingot itself, which has no fluid metal above it to feed or take the place and supply the deficiency thus created. It will readily be understood therefore why the upper portion of ingots is so seriously affected as regards their soundness, also why segregation occurs. The occurrence of these defects varies according to the type of steel, whether rising, semi-rising, or settling nature, as shown by Plate IX.

Again referring to this question of the cavities, if, as proved by these experiments, in the ingots made under the writer’s system the metal in the sand head portion in descending has without doubt filled or prevented the formation of what would otherwise have been unsoundness, piping, loose structure, or segregated material, in unfed ingots made in the ordinary way there must be steel of loose structure; if not, then in many cases absolute unsoundness and segregation.

Moreover, and this is a most important point, the steel in the “fed” ingots being maintained fluid in the head portion, continues to exert its ferro-static pressure, whereas with ingots made in the ordinary way the ferro-static pressure on the center portion of the ingot is so slight that it produces very little beneficial effect. Further, without the feeding head above the ingot proper, the outside of the ingot in the ordinary ingot mold becomes rapidly chilled and frozen, so that it cannot contribute its proper share to the feeding of the remaining portion of the ingot. It is not therefore to be wondered at that rails rolled from the A and B portions of an ingot made in the ordinary way are liable to unsoundness or piping or both, and are also often full of impure segregated material.

There would probably be more dangerous ingots but for the fact that the steel maker tries to avoid this type of steel, and aims to make steel which when poured into the ingot will not pipe. Nevertheless, he is still fighting against a natural law. If piping steel is checked or avoided, he runs the risk of producing unsound steel, especially in the upper portion of the ingot, more or less permeated with blowholes. Thus, owing to lack of feeding from the upper portion, the center, or that portion on the axis line of the ingot, must be of inferior nature, as the piping characteristics persist for quite a long way down the ingot. This, as before mentioned, is for the reason that owing to want of ferro-static pressure, the ingot lacks feeding from above, which in the system of casting ingots now described is maintained to a very late stage; that is, until or close upon actual solidification takes place. There is always fluid steel in the upper portion of the ingot to feed the piping and shrinkage, both of which must occur, as they follow a natural law. This, too, is the reason why there is so little segregation in ingots made under the writer’s system, and also explains why the ferro-static pressure is kept up to a very late stage. In fact, check or hinder ferro-static pressure, and segregation with its bad effects at once commences. In the case of “ fed” ingots, the smaller amount of segregation which occurs takes place outside the ingot proper, that is in the head. This is well illustrated by the results shown in Figs. 4 and 5 of Plate I.

With steel of piping nature poured into the ingot molds and not fed, it need cause no astonishment to find that rails, even from C and D portions of the ingots, may be of material with loose structure, consequently weak if not actually unsound, thus giving inferior or bad results in service. This is shown by Fig. 3 in Plate I, which shows that with piping steel not properly fed, there is extreme danger of this pipe extending a long way down the ingot; in fact, in some cases it has been found to go as much as two-thirds the length of the ingot. Ingots have also been found in which the piping has apparently stopped, only to be resumed below the sounder steel.

The experiment carried out by the writer some years ago, by the pouring of copper into the upper portion of an ingot 15 or 20 min. after casting, showed how serious is this want of ferro-static pressure in the material situated on or near the center or axis line of the ingot in ingots which have not been properly fed. The copper finds its way down to the bottom of the ingot, although added 15 min. after casting. In any case, if there is no definite pipe at the bottom portion of such ingot, there is still material of loose or open structure, which means weak steel. Although this may not be apparent by fracture to the naked eye, nevertheless it exists and can generally be detected by an examination of the microstructure. In other words, notwithstanding that the product to be used may come from the lower half of the ingot, yet in unfed ingots it will be weak and not able to stand severe stresses. This is probably the real explanation of the serious breakages which sometimes occur in even C and D rails. Imperfect material is present; only time and sufficient working stresses are wanted to develop its existence and weakness.

If is true that some portions of the cavities in ingots have been measured, but probably not in the manner described by the writer.

Although in the examination of the top of an ingot cast in the ordinary manner and from steel which “settles,” there is external evidence of some piping, this is irregular and varies considerably. Therefore, in the “best” ordinary ingot proper evidence is slight as to how much or how little the steel has piped. Dr. Dudley has pointed out in his interesting paper to the Institute on Piping and Segregation, of Ingots of Steel and Ductility-Tests for Open-Hearth Steel Rails, read in February, 1913, that in such ingots the piping is divided into two kinds, the upper, or what may be termed the visible pipe, and the lower, or hidden pipe, the extent and character of which can only be determined by cutting open the ingot. In the ingots cast under the writer’s system, all the cavity or pipe is open and can readily be inspected from the top; its extent can be determined whether in the hot or cold condition. It is therefore not necessary to cut open the ingot. In other words, such cavity produced by the piping is “the outward and visible sign of an inward and spiritual grace”— “spiritual grace” in this case meaning “soundness.”

Figs. 1 and 2, Plate X, further explain this.

Fig. 1 represents an 18-in. ingot made on the writer’s system, and as shown in Plate VII already referred to. All the cavity produced by the piping is easily visible and measurable from the top. If the ingot is allowed to go cold, the cavity can be measured and its cubic contents ascertained by filling such cavity with water, then pouring this out and taking its volume. If desired, the same object could probably be accomplished while the ingot was hot, in that case using a metal of suitable melting point. Such an ingot possesses no upper or lower cavity, it is all in one. The full characteristics of the ingot in this respect, and whether hot or cold, are known and determined by such cavity.

Fig. 2 shows the section of an ordinary 21-in. ingot made from steel which is of piping nature. It will be seen that there are two separate cavities, an upper and a lower one. While the former can be measured, the information obtained would not be of much practical value. As regards the latter, it is, so to speak, covered by a metallic diaphragm and cannot be measured except by cutting or machining open the ingot when cold. Moreover, it will be seen that the center portion of the ingot, that is the metal over and surrounding the central vertical or axis line of the ingot, is not sound for quite a long way down. Below where it appears to be quite sound to the eye there are still segregation results to be dealt with. In other words, a considerable portion of this ingot must be “discarded” or “cropped,” either in the ingot itself or the rail produced therefrom, before a sound, safe rail can be obtained, that is, one which is perfectly sound not merely to the naked eye when examining a fracture of the ingot or the rail from such ingot, but is so when its microstructure is examined, or sulphur prints or etchings are taken.

This real and true soundness is obtained in ingots made under the writer’s system; that is, almost immediately under the cavity, and after a discard or crop has been taken of even as low as 7 per cent, in some cases, and certainly with 8 to 10 per cent. Therefore, such products as billets, bars, rails, sheets, etc., produced therefrom are practically sound at any portion of their mass.

This would appear to be a considerable advance as compared with ordinary practice, and even as compared with ordinary ingots of a char-

acter such as above referred to, which in themselves are certainly superior to many ordinary ingots made.

The ingot in Fig. 1, representing an ingot made under the writer’s system, shows no unsound portion, that is to say, blowholes are entirely absent. About 8½ in. from the top of such ingot, segregation has disappeared. Similar comparison with Fig. 2, representing what may be termed a high-class ordinary ingot, shows that at a point no less than 18 in. from the top there is still unsoundness. Moreover, below even the sound portion, there is segregated material which must be discarded before safe and sound structural steel can be obtained. It will be seen the discard of the former must be far lower, yet better and more sound material is produced.

The writer is aware that the above comparison is made between an ingot prepared under his system and an ordinary ingot which has been allowed to become cold. As Dr. Dudley has pointed out, there is considerable improvement in such ingots by handling them quickly while very hot. There is no doubt this is quite correct; yet in ingots cast under the writer’s system, which can also be handled while very hot, it is difficult to see how an improvement can be effected in ingots which are already practically perfect.

At any rate, there certainly could, not be a reduction in quality. In other words, if there was any difference in ingots made under the writer’s system, the difference between “ hot and cold ” ingots would be in a favorable direction. The writer, however, thinks that these differences, that is, between the same ingots hot and cold, whether as regards the ordinary or the ingots made under his system, can hardly be very considerable, for after molten steel has once passed into the “solidus” as compared with the “liquidus” condition, there can surely only then be the ordinary contraction changes, or what may be termed “linear” contractions. These do not materially affect the volume of the cavity or the unsoundness, which if they exist remain practically as they are when the steel congeals into the “solidus” condition.

In view of the greatly increased service now met with on railroads in nearly all countries, resulting from heavier and more trains, higher speeds, whether freight or passenger, it will be seen that, in the nature of things, rails and similar products must be made from sound steel and from ingots each of which must be known to be as perfect as human ingenuity can accomplish. If this and other papers on the subject help to bring us nearer the goal and also help to remove what would appear to be a slur upon the metallurgist, then the labor involved in the many researches being carried out in this country and elsewhere will be well repaid.