Mr. Chairman and Gentlemen:—I am announced to speak to you this evening upon " Iron and Steel, and the Mitis Process "—a subject manifestly too vast to be discussed comprehensively in a single lecture.
I shall, therefore, confine my remarks, relative to iron and steel, to certain observations and conclusions made during a somewhat intimate practical acquaintance with these metals for the past thirty years, which I regard as contributing to a correct understanding of their structure, and to their proper treatment in the course of manufacture. I shall then, in conclusion, offer a description of a recent and most promising improvement in the metallurgy of iron and steel called the " Mitis Process."
Allow me first to endeavor to answer the question, " What is wrought iron ? " One of the greatest obstacles to a correct apprehension of this question is this :—the way in which a mass of wrought iron is built up is not generally well understood, and the difference of its structure from that of a homogeneous material is not fully comprehended. The term wrought iron is popularly supposed to designate a metal ; but it is really the name of a mechanical admixture, which, at its best, consists of clusters of crystals (which may with propriety be regarded as compound crystals) of pure iron separated from each other, as the result of the manipulative processes employed, by films or threads of an unavoidable impurity called " cinder." In the manufacture of wrought iron, the " pig," or other variety of cast iron, is first deprived, in a more or less imperfect degree, of carbon and other impurities, by what is known as the " puddling process." This process may be briefly described as consisting of four distinct operations ; viz.—
1st. The melting of the " pig iron."
2d. The " boiling " of the melted metal in a bath of liquid " cinder " (composed mainly of silicate of protoxide of iron), until the iron (which, owing to its loss of carbon and other impurities, can no longer remain fluid at the temperature employed,) begins to solidify in the form of small granules or crystals, which can be seen moving amid the boiling " cinder " like white-hot peas in a red-hot soup. When the iron begins thus to granulate or crystallize, it is said to be "coming to nature."
3d. The collection by the puddler of these granules or crystals into distinct masses called "balls." These contain much "cinder."
4th. The " squeezing " or " hammering " of these " balls," while still at a welding heat, into more solid masses, which are called "blooms." These contain much less " cinder" and other impurities than the " balls," but are far from being uniform in structure.
The " balls " above named may with propriety be regarded as white-hot sponges of iron saturated with liquid " cinder," which fills all their accidental and irregular cavities.
When the " balls " are "squeezed " or " hammered " (this last operation is often termed "shingling"), for the purpose of expelling this " cinder " and welding the granules or crystals of iron into a homogeneous mass, the attempt is never wholly successful ; for the " cinder," as the metal cools, quickly assumes a pasty consistency and flows with difficulty, and all that portion of it enclosed in the interior cavities of the " ball " is simply flattened out or elongated. Hence it will be seen that the " bloom " is composed of a compacted mass of granules or crystals of iron, separated from each other by films or strings of "cinder" of very irregular dimensions.
When speaking of crystals of iron, I mean minute ultimate units of that metal, bounded by well-defined planes, whose intersections always form salient angles. A number of such crystals may cohere and form an aggregation, having bounding planes similar in outline and relative arrangement to those of any single crystal. Such aggregations, or compound crystals, vary in size, and are often regarded as single crystals and spoken of as such, just as we speak of crystals of galena or calc-spar, when, as a matter of fact, the ultimate crystal of each of these substances remains undiscovered, and as undiscoverable as the boundaries of space.
These large or compound crystals of wrought iron are, in themselves, practically homogeneous : that is to say, the ultimate crystals of which they are composed are not separated and kept apart by any foreign substance, but are as nearly in actual contact as the law of cohesion, in obedience to which they are formed, will admit.
Now let us see how a " bloom," the crudest form of a mass of wrought iron, differs in structure from the homogeneous compound crystals of which it is chiefly built up. Such a mass of wrought iron is an aggregation of an indefinite number of such compound crystals as have been described, which are separated from each other by films or threads of "cinder" of very variable thickness; but which, notwithstanding, are mutually attracted with a greater or less degree of force, the minimum value of which is the measure of the cohesive strength of the mass.
Now let us follow the " bloom " as it progresses towards the form of a commercial bar of wrought iron, and examine carefully the structural changes which take place during such progress. When a properly heated " bloom," or other similarly constituted mass of wrought iron, is subjected to the action of the " hammer " or " rolls," the contained "cinder " endeavors to escape from its entangling alliance with the crystals of the iron, and in so doing, each particle thereof is driven into some line of least resistance, which is always finally located in a plane at right angles to the direction of the force acting upon the metal. In other words, if the " bloom " is rolled or forged into a rod or bar, the metal will be acted upon in two directions at right angles to each other, and its compound crystals will be compressed in directions normal to the exterior surfaces of the bar, and at the same time extended in the direction of its length. Thus the ends of adjacent crystals are forced towards each other, and the intervening " cinder " is compelled to move at right angles to the axis of the bar, and to unite with the films or threads of " cinder " which have become established in parallel lines of least resistance along the flanks of the compound crystals, and at right angles to the direction of the force acting upon the bar.
Fig. I is intended to illustrate on an exaggerated scale this arrangement of the elongated compound crystals of iron with intervening films or threads of " cinder," the light spaces representing the iron crystals and the dark lines the "cinder "—the force of compression being supposed to act upon
the bar in the direction of the arrow.
Fig. 2 illustrates a method of showing by experiment the character of the structural difference between a bar of wrought iron and one of a homogeneous material such as low steel. In this figure, let A he a vertical section of a cylinder provided with an accurately fitted plunger P. The space B below this plunger we will suppose to be filled with small fragments of lead of irregular dimensions, whose surfaces are covered with a coating of oxide of lead.
If, now, sufficient force is applied to the plunger P, the lead will be forced out of the hole in the lower end of the cylinder in the form of a rod C, and every fragment of lead will have become more or less elongated, but will be prevented from actual metallic contact with adjacent fragments by a film or thread of oxide of lead. In this experiment, the elongated fragments of lead correspond to the extended compound crystals of iron before named, and the oxide of lead occupies the same relative position in the rod of lead as the "cinder" in a bar of wrought iron. If now, in place of the fragments of lead, we place in the space B a solid mass of that metal, then, on applying adequate force to the plunger P, there will be forced through the hole in the bottom of the cylinder a rod of lead, whose structural difference from the former rod, made from the oxide-covered fragments, is closely allied to that subsisting between a bar of low steel, and one made of the "cinder "-coated compound crystals of wrought iron.
The direct consequence of the elongation of its compound crystals and the effort of the intervening " cinder " to escape in the direction of least resistance while the wrought-iron " bloom " is being forged or rolled as before described, is the establishment of that structural peculiarity in the resulting bar known as " fibre," which is one of the most conspicuous features of wrought iron, and one not found in any other variety of ferruginous materials.
When any of the films or threads of "cinder " in a bar of wrought iron are so large as to be distinctly visible on its surface to the unassisted eye, they are called " sand seams " or." cinder cracks."
If its 'compound crystals are nearly pure iron, the bar can be readily bent cold without fracture, and, if pulled asunder by a gradually augmented force, its fibrous texture is at once evident ; but in case the compound crystals have chemically combined with them some substance, such as phosphorus or silicon, which tends to diminish both the cohesive attraction between the ultimate crystals of which they are composed and the mutual attraction of the compound crystals, then the bar cannot be easily bent cold without rupture, and is said to have a " crystalline fracture." Notwithstanding this appearance, however, the mechanical structure of the bar is the same as before; that is to say, the " cinder " and elongated compound crystals are still arranged in lines parallel with the axis of the bar, although it is quite probable that the average length of the compound crystals may be much less than in the case of the bar first described.
Whenever a " bloom " is subjected to a force of compression always acting perpendicularly to the same plane, as is the case when it is rolled into a " sheet " or " plate," its compound crystals and accompanying " cinder " are each flattened and extended parallel with that plane, and the resulting " sheet" or " plate " has more of a laminated than of a fibrous structure, being built up of a number of leaves or strata of iron separated from each other b)" films of " cinder," which, when unduly thick at any point, cause defects in the plate that are called " blisters."
The foregoing discussion of the structural difference existing between a bar of wrought iron and one of homogeneous iron (often called " low steel ") naturally brings to mind an important practical question relative to the employment of wrought iron in construction which is often asked; viz.. Will a given sample of wrought iron having a decidedly fibrous texture become crystalline under the operation of a continued repetition of violent strains or shocks ? Doubtless many persons of large and varied experience will unhesitatingly answer this question in the affirmative. The sailor who sees his chain cable (known to have been made of carefully selected, thoroughly worked, and honestly tested fibrous iron,) snap short, has no doubt about the metal having become crystalline owing to lapse of time and rough usage. The practical farmer, as he examines a broken trace or plow chain, is firmly of the opinion that the iron thereof had become crystalline by use. The railway passenger who has fortunately escaped serious injury from an accident caused by a broken axle, is usually ready to testify, with emphasized confidence, that " the iron of the axle was crystalline, and entirely unfit for the purpose for which it was used." A modern fiddle-string bridge goes down under a passing train, plunging a whole community in mourning and sending a thrill of shivering horror through the land—among the various theories advanced to disguise the utter want of sufficient intelligently distributed material in the structure, is sure to be found that of the crystallization of the iron employed.
But let us return to our question. Can a bar of wrought iron of a pronounced fibrous structure be ruptured so as to exhibit a crystalline fracture? I answer, yes,—in two ways. 1st. By a sudden application of a force of extension—commonly called a "jerk." 2d. By a prolonged repetition of a force of compression—sometimes called a "jar."
The first method of rupture may be said to consist of a transverse division of the compound crystals of the bar, as distinguished from a sliding of their interlocking flanks upon each other, as is the case when the rupture presents a fibrous appearance. I have often seen crystalline fractures produced in truly fibrous iron. In the manufacture of iron rails (now nearly an extinct industry), it was always considered desirable that they should be of a hard and crystalline texture as to their tops or " heads," but soft and fibrous in their bottoms or " flanges "; but however perfectly this distribution of metal was made, it was always possible to break a rail so as to show a crystalline fracture in its " flange." This was accomplished by making a slight " nick " across the flange (to determine the point of fracture), and placing the rail ("flange" down) in the "straightening press," on supports placed a short distance on either side of the " nick," and then putting in the "gag" "heavy" just over it: the result was almost always a crystalline fracture in the "flange"—in short, the elongated compound crystals were "jerked" asunder. But, if the points supporting the rail were placed further apart, and the rail given an opportunity to yield considerably between them, then, if the "gag” was put in " light" a number of times in succession, the fracture of the "flange" would be sure to exhibit a fibrous texture, due to the fact that sufficient time had been given to break up the films of " cinder " along the flanks of the compound crystals and destroy their transverse cohesion, thus permitting them to slide apart, and exhibit the appearance of disrupted fibres.
We are indebted to a not uncommon accident to which the hammer bars of a peculiar type of steam-hammer are liable, for an excellent illustration of the second method of producing a crystalline fracture in fibrous iron, the result of the repeated action of a percussive force of compression. In Fig. 3 is represented at A the bar of such a steam hammer. As has been before stated, there exist, in a bar of fibrous iron, films of " cinder " between the ends of its elongated compound crystals (as shown exaggerated in Fig. I). These, from the nature of their formative process, cannot possibly be of uniform thickness. This, considered in connection with the fact that the greatest force of the percussive action per unit of area of any cross section of the hammer bar is exerted upon a section made by a plane cutting the bar at right angles immediately above its head, justifies the belief that at, or near, this point, fracture would be most likely to occur. It is also evident that the percussive action of the hammer would have more destructive effect upon thick than upon thin films of " cinder"; while, at the same time, the force of cohesion between the ends of adjacent compound crystals will be diminished in some inverse proportion to the thickness of the films of "cinder" between them. It therefore seems exceedingly probable that the fracture due to continued percussion will take place, if not in the plane above named, yet in one very near to it, in which the "cinder" films chance to be of greater thickness than those in that plane ; and, as a matter of fact, fractures in such bars are usually within a few inches of the point where the bar enters its head, as at G, H, Fig. 3.
The particular point in the circumference of such a hammer bar where the imminent fracture first appears is often determined by the manual peculiarity of the "hammer-man." A left-handed man will incline his work to the left, and a man who is right-handed will be likely to use the right side of the
anvil more than the left. In this latter case the work B, Fig. 3, will tend (whenever it is in the position shown) to produce a tensile strain at the point G, which, as the work is shifted to the centre or occasionally to the left side of the anvil, becomes a compressive strain. We should therefore expect (as is, in fact, the case) that the initial manifestation of the fracture would be found at that point, and that it would gradually extend towards H, until the bar was finally "jarred " asunder. This separation would take place through films of" cinder" between the ends of the elongated compound crystals of the bar, thus exposing those ends, and exhibiting what is called a crystalline fracture;
The belief in this so-called crystallization of wrought iron, as the result of prolonged use, is, I think, altogether a mistake ; "and I am clearly of the opinion that the crystallization observed in the case of any particular fracture existed just as we see it at the time the metal was given the shape in which it was ruptured. After a bar of distinctly fibrous wrought iron has been subjected to multitudes of sudden " jerks " of extension or "jars" of percussive compression, the " cinder " in some cross section of it (in which this impurity is slightly thicker than elsewhere) gets broken up, cohesion is destroyed, and the bar breaks with a crystalline fracture.
I have had a specimen prepared for the purpose of making the foregoing explanation of the apparent crystallization of fibrous iron more evident. It is a short piece of a square bar of wrought iron. One end is decidedly crystalline in its fracture, showing distinctly that the bar was originally built up of five flat bars. The other end is, for more than one-half of its area, as decidedly fibrous as wrought iron can well be ; and this end would have been uniformly fibrous in appearance had the workman who made the specimen exercised the requisite care. Thus, in a sample not over two inches in length, we have an instance of a fracture which most observers would call very bad, and another which as certainly would be called good.
It is a well-known fact that wrought iron is improved in strength by repeated working. This may be accounted for thus:—in the initial heating and shaping of the metal, its crystals were left with a comparatively thick film of "cinder" between them; but, by each successive reworking, the crystals of metal are driven into closer order, some of the intervening "cinder" is expelled, and what remains is very much reduced in thickness, so that the cohesive attraction (whatever that may be) between these crystals, having less space to act through, acts with augmented intensity. It is well to remember, when we speak of " less space " in a matter of this kind, that we are dealing with a very small quantity indeed—one that is a near neighbor to the infinitesimal.
Time passes, and though I could fill the fleeting hour with talk about iron, yet in this lecture, as in the field of mechanical construction, it is now fitting that iron should give place to steel.
What is steel? To the many answers to this frequently asked question, I may perhaps venture to add this one more, namely, that steel is iron freed from mechanically mixed impurities (such as " cinder ") by a melting process, during which it has combined with it chemically a small percentage (not large enough to prevent the metal being forged or rolled) of other impurities, introduced for the purpose of modifying its strength, hardness, elasticity, or ductility, in such degree as to adapt it to the particular use to which it is to be applied. In short, while wrought iron is iron having (as the legitimate result of the methods employed in its manufacture) its impurities mechanically mixed, steel is iron having (as the result of the adoption of appropriate manufacturing processes) its impurities chemically combined.
A great deal of the difficulty of correctly fixing the status of any given sample of ferruginous material may be eliminated by recognizing the fact, that what is called wrought iron is not really iron, and that the only way in which pure iron can be obtained is by electrolysis; a process which is, I need hardly say, commercially impossible for all practical purposes in the present state of our technical knowledge.
I cannot on this- occasion describe at length the various processes employed for the manufacture of steel, but will call your attention to certain practical details which are of especial interest and importance.
If we break a large ingot of mild steel (say of 12" to 15'' square) at right angles to its length, and examine the fracture, we shall find at a distance of from three-quarters of an inch to two inches from its sides, a collection of cavities or " blow holes " (as they are commonly called), which are of an irregular spherical form, and of variable size, the largest seldom exceeding one-half an inch in diameter. These holes are separated from each other by partition walls of irregular thickness, and in most instances are coated on their interior surfaces with films of a more or less iridescent oxide of iron. Fig. 4 will serve to give an idea of such a fracture as has been described.
There has been a great deal of speculation as to the origin of this array of cavities. Some have supposed that they were caused by gases dissolved in the fluid steel (very much as carbonic acid is dissolved in water), and that at the moment of solidification these gases separated out, and arranged themselves in the order in which they are found—that of a hollow square (in the case of a square ingot) whose sides are parallel with those of the ingot mould. Others have asserted that oxide of iron is present in the fluid metal, and that this being reduced by the carbon in the steel, carbonic oxide, carbonic acid, or both are set free, which being unable to escape before the steel solidifies, produces the aggregation of cavities we are considering. This theory may be in a measure true in the case of steel containing oxide of iron, but we know that good metal does not contain oxygen, and if the holes were in any great degree due to the reaction mentioned, they would be likely to be evenly distributed through the mass of the ingot, and not confined, as is the fact, to a well-defined zone.
Other suggestions involving more or less occult, hypothetical, and ingenious chemical and molecular considerations have been made; but all fail to account satisfactorily for the symmetrical arrangement of the holes observed.
My own explanation of the formation and peculiar distribution of these cavities is a purely mechanical one, which I will now endeavor to make clear.
It is a well-known fact that a vertical stream of any liquid descending freely through the atmosphere drags along with it, by frictional contact, a notable quantity of the air or of any other gas that may be in its immediate vicinity. This fact was many years since taken advantage of in the construction of the blowing apparatus called the " tromp," used for furnishing the blast for the forges of Catalonia.
This apparatus consists of a vertical pipe P (usually of wood), whose height is determined by that of the head of water at the locality of the forge ; the upper end of this pipe passes through the bottom of the wooden race-way R (Fig. 5), and is closed or opened by the movable conical plug or valve V. Below the bottom of the race-way R there are several inclined apertures a, a, a, made in the sides of the pipe P. These are for the purpose of admitting air, which, when the valve V is raised, is drawn in by the descending column of water, and mixing therewith, is carried downward and discharged thereby into a receiving chamber C. Here a separation of the air and water takes place, the former passing through the tuyere pipe T, T to the forge-fire F, and the latter escaping from the receiving chamber through a hole in its side at H. The volume and pressure of the blast supplied can be regulated within certain limits by raising or lowering the valve V, by means of the cord K, acting through the lever L.
Over twenty years since (1863), I employed (in the laboratory of the experimental Bessemer steel works at Wyandotte, Michigan) the principle of the mechanism described for supplying blast for a table blow-pipe. The apparatus for this purpose consisted of an ordinary three-necked Woolf bottle B (Fig. 6) of about a half-gallon capacity, to the middle neck of which was adapted a cork through which was passed the stem of a small funnel F, which reached nearly to the bottom of the bottle. To the right-hand neck of the bottle B was fitted the discharge syphon S. The left-hand neck of the bottle B had fitted to it a bent glass tube T, to whose horizontal end was attached a rubber tube for conveying the air to the blow-pipe. To put the apparatus in operation, a stream of water was discharged from a jet-pipe J into the top of the stem of the funnel F. The diameter of this stream was slightly less than the internal diameter of the tube of the funnel, and could be regulated as regards its volume and velocity by a suitable cock C. As the stream of water from the jetpipe J descends the tube of the funnel F, it drags along with it by frictional contact a very considerable volume of air, which, on reaching the bottom of the bottle B, separates from the water and passes to the blowpipe through the rubber tube before named, the water finding an exit through the discharge syphon S.
Now let us see how the action of the "tromp" and the apparatus just described is concerned in the casting of an ingot of steel.
Let the beaker B (Fig. 7) represent an ingot mould, and the descending stream of
water W, the stream of liquid steel;— it will be seen that the stream W carries with it a large volume of air into the water (for illustrative Fig. 6. purposes regarded as liquid steel). This air, in its endeavors to escape, turns, and in the form of globules or bubbles takes an upward direction parallel with the sides of the beaker (representing the ingot mould). On the stoppage of the stream, all this air immediately escapes from the water, leaving it as homogeneous as water usually is; but, if during the filling of the beaker the water therein was rapidly frozen (the progress of the congelation being from the sides towards the centre), it is evident that the ascending bubbles of air would be entangled in the ice as it formed, and we should have as a final result a vesicular mass or ingot of ice, quite similar as regards its method of formation to the ordinary ingot of steel.
Another illustration may make the formation of vesicles in steel ingots Still more clear. If in place of water in the preceding experiment we substitute mucilage, or any other fluid of similar consistency, we approach much nearer to the actual conditions which exist in the casting of a steel ingot ; for the steel as ordinarily melted is never as fluid as water, but approximates more nearly in mobility to the character of mucilage. As, then, the stream of mucilage descends, it will be observed that it carries with it air in the same manner as the stream of water ; but that owing to the viscosity of the fluid, the air bubbles rise through it more slowly and escape with greater difficulty, and that some of them, as they approach the surface, are again dragged down by the central descending current. Hence there is a much larger collection of bubbles of air in the mucilage than there was in the water, and, consequently, if the mucilage was solidified at the moment the descending stream was stopped, we should have a much more vesicular mass than in the case of the frozen water in the last experiment.
In comparing the foregoing experimental illustrations with the actual conditions which exist during the casting of an ingot of steel, we find an ingot mould of cast iron (corresponding to the beaker), which is filled by a rapidly descending stream of molten steel (corresponding to the water or mucilage), not as liquid as water, but more nearly of the consistency of mucilage. We also find that this stream carries into the imperfectly fluid mass of steel which rapidly fills the ingot mould a large volume of air, which attempts to rise and escape from the rapidly cooling and solidifying mass of metal in precisely the same way as the bubbles of air endeavored to escape from the water and mucilage in our two illustrative experiments.
But we find another condition present in the case of the molten steel that did not exist in either experiment ; viz., the fact of a high temperature in the fluid metal. If we examine this condition,
we shall readily discover that it has a very important influence both on the size and number of the vesicles included in the ingot of steel; for it is a well-known fact that dry air, for each 480° F. increment of temperature, increases its bulk by the amount of its original volume. Now, as the fluid steel is at least of the temperature of 3300° F., dry air introduced in the manner illustrated would be so expanded as to occupy seven times the space in the ingot that it did in the atmosphere.
There is, however, yet another fact that tends still further to augment both the size and number of the so-called " blow holes" that we are considering. It is a well-known practical condition that the air in the immediate vicinity of steel-casting pits is far from being dry. The large quantity of water used for cooling ingot moulds, and for other purposes, keeps the atmosphere surrounding both casting-ladle and ingot-mould in a very moist state, and it is certain that all such vapor-laden air carried into the molten steel would increase in volume for a given increment of temperature very much more than dry air, and would therefore correspondingly increase the size and number of the " blow holes." Furthermore, this vapor of water does not act to this end altogether through its expansion under the influence of heat, for some, if not all of it, is decomposed by the high temperature, and its oxygen, together with that of the accompanying air, is absorbed by the walls of the cavities. This produces the iridescence observed, and leaves in the " blow holes " an atmosphere composed mainly of hydrogen and nitrogen ; and it is not at all improbable that in many cases this decomposition of the watery vapor did not take place until the steel was so far solidified as to prevent the walls of the cavities yielding to any great extent, and, under such circumstances, the gases named would be under a very considerable tension.
This view is confirmed by the investigations of Prof. F. C. G. Muller of Brandenburg, who found that the mean composition of the gases in the " blow holes " was
Hydrogen, 79 per cent.
Nitrogen, 19 "
Carbonic oxide 2 "
and that their average pressure was 120 pounds per square inch.
It is of course possible that some of the gases found in the " blow holes " of Bessemer steel ingots may have found lodgment in the steel during the process of " conversion," more especially when steam is admitted with the blast for the purpose of keeping down the temperature of a " hot blow." In this case the steam would certainly be decomposed, and some of the residual hydrogen might remain entangled in the metal ; although doubtless much the larger portion owing to its great levity, would escape during the pouring of the steel from the " converter " into the " casting ladle." In the case of open-hearth steel, however, there would be no such reason for the presence of hydrogen.
The foregoing explanation of the presence and methodical arrangement of the cavities or " blow holes " in steel ingots, and the character of the gases found in them, is, I think, sufficient to account for all the facts observed. The assertion which has often been made (as a sufficient explanation), that " the gases are occluded in the steel," deserves to rank with the pompous declaration of a village pedagogue that a total eclipse of the sun which' was terrifying his neighbors was " only a phenomenon." "Words of learned length and thundering sound " discourage rather than stimulate the inquiring mind.
I now desire to call your attention to a species of cavity much too frequently found in forgings of steel, that does not originate in the manner already described.
If a large cold ingot is put into a too highly heated furnace, its exterior surface will expand so much faster than the parts at or near its axis, as to strain the metal in the interior of the ingot beyond its elastic limit, and oftentimes actually rupture its central continuity, as is shown at A, Fig. 8. Such a breach may in some cases have a diameter equal to hall that of the ingot.
An ingot thus internally fractured, if hammered or rolled down to a smaller section, will have a cavity developed in the centre of its mass, as shown at B ; and unless the existence of this cavity is discovered, serious difficulty may result from the use of such a forging as a part of any mechanism. It is not at all impossible for a number of such cavities to be formed in the same ingot, if the heating be sufficiently rapid, in which case the initial rupture would occur at A, Fig. 9, at or near the centre of the ingot ; a second and third fracture would then take place almost simultaneously at B,B, about halfway between A and the two ends of the ingot ; and, finally, a third set of internal breaks may be formed at the points C, C, C, C, thus dividing the ingot into eight nearly equal parts of solid metal. The diameters of the several ruptures would vary in the following order, viz: That at A would be the largest, those at B, B, somewhat less, and those at C, C, C, C least of all. Such an ingot—if the internal ruptures were not too large—might be forged into a propeller shaft and actually put into a vessel, without the defects being discovered until it was twisted asunder on its first voyage.
Such possibilities of carelessness in the manufacture of heavy forgings of steel as I have described make it highly desirable that some method be devised to detect the presence of such internal ruptures before much time and labor have been expended upon the forging, and also to prove its soundness when completed. About twenty years ago, a plan for this purpose was proposed by Mr. S. M. Saxby, R. N., and some extended experiments to test its practical value were made by direction of the Admiralty ; but, although the early investigations were very promising, the method has not become established as one of the acknowledged reliable means of testing forgings of iron or steel. It is possible that some method of electrical examination may be found of service in testing the soundness of forgings, and I will venture to suggest the following :—
Let A, Fig. 10, be an internal rupture in the ingot I, to the extremities of which are connected the wires P, M, of the battery B, having in the circuit a galvanometer G. Under these conditions the galvanometer needle will be deflected a certain amount, which is a function of the strength of the current and the resistance of the circuit ; and if by any means the resistance of the circuit is diminished, the deflection of the needle of the galvanometer will be increased. For instance, if, in the proposed apparatus, the wire N be moved towards the left, for each inch of movement there will be a corresponding increase of deflection of the needle of the galvanometer ; but when the wire passes a point opposite the rupture A, the law of the increase of deflection may be found to change, and to indicate the presence of an internal breach of continuity in the ingot or forging under examination. I have had no opportunity to test this method, but make the suggestion in the hope that some one having the means and leisure will give it a thorough examination, and that it, or some modification of it, may be found of practical value.
Thus far I have spoken only of the transverse internal rupture of steel ingots in consequence of too rapid heating ; but longitudinal internal ruptures can be, and often are, produced by the same cause. In Fig. 1, let ABCD represent a cross section of a steel ingot. If too rapidly heated, the opposing sides AB and CD will expand so much faster than the centre, that an internal rupture EF may be
formed ; and the expansion of the sides AC and BD may in like manner develop a similar rupture GH, located in a plane at or nearly at right angles with that already named.
Such ruptures, though generally situated in planes at or nearly at right angles to each other, are not confined to planes located as shown in Fig. ii, for the planes of rupture may coincide with the diagonal planes of the ingot, or may occupy any position between such diagonal planes and that shown in the figure. In fact, their position is fixed by the resultant action of two forces, due to the expansion of the exterior of the ingot by the sudden heating, modified by the powerful internal strains existing in the cold ingot tending to separate the metal at its centre. These strains were established at the time the metal originally solidified in the ingot mould, and are occasioned by the outside of the ingot cooling, while its interior is either fluid or plastic ; and as the whole mass becomes cold, its interior, by the force of cooling contraction, is strained in many cases beyond its limit of elasticity, which limit may with propriety be defined as the beginning of rupture. An ingot of steel thus internally strained would require but the small addition to the tension which a too rapid heating of its outside would furnish, to produce such interior longitudinal fractures as have been described. The extent of the influence of such internal strains in all stages of the manufacture of steel is very irregular and uncertain, and this fact makes them all the more worthy of consideration in all cases in which steel is to be subjected to uses which involve the application of sudden and violent shocks.
Of the effect of such strains in steel used for the construction of cannon, Col. Eardley Maitland, R. A. Assoc. Inst. C. E., Superintendent of the Royal Gun Factory at Woolwich, in a recent paper said: "On a review of the results obtained, the author, having seen so many instances of the fracture of steel, sometimes spontaneous and sometimes under stresses quite inadequate to produce the result, was of the opinion that internal strain was the gun-maker's worst enemy, and that it was a question of great moment whether it was worth while to incur the risk of setting up such strain by oil-hardening."
It is not at all improbable that in many instances (especially in the case of steel having considerable hardness) ingots may be ruptured internally both transversely and longitudinally, thus aggravating the evil of either single species of rupture. If such an ingot were forged into a heavy crank pin, its whole interior would be permeated with most irregular and intricate imperfections, though at the same time the ends and cylindrical surface of the forging might have every appearance of soundness.
As a practical illustration of the great importance of the subject we have been considering, I cannot do better than quote the description of a defective forging given by Professor Thomas Egleston in Transactions American Society of Mechanical Engineers, Vol. VII., p. 263. He says : " I have recently had occasion to examine a forged crank pin made with great care from the best of open-hearth steel. It was rough turned to 16 ½ inches. To ascertain its quality in the centre, an inch and a half hole was bored through it. This hole revealed such a number of cracks and cavities that the hole was increased to four inches, in the hopes of cutting them out. Defects of considerable size were still found. The pin was then sawn in two [planed apart longitudinally], where single horizontal cracks 10 inches in diameter and ¾ inch wide were found, and inclined ones 7 ½ inches long, in which were cavities ½ an inch wide, to say nothing of defects of minor importance. None of these defects would have been revealed but for the forethought of examining the centre of the piece. If it had been used without this examination, it would have produced great disaster."
I also have had an opportunity of examining the forging described by Professor Egleston, and was told that it was made by one of the oldest and most experienced manufacturers of such work in this country. My experience teaches me that such defective forgings are far more common than the managers of our steel works and forges are disposed to admit or even believe.
It is a common opinion that one of the reasons why steel forgings are often found hollow in their interior is the failure to work them under a sufficiently heavy hammer ; but no hammer, not even "the hammer of Thor," can do more than aggravate the evil of internal ruptures in ingots of steel.
But let us now turn from this long digression relative to internal ruptures in ingots and forgings, and resume the consideration of the matter of " blow holes " in steel ingots. Having endeavored to explain and illustrate what I regard as the principal cause of the formation of such " blow holes," allow me now to examine some of the consequences of their presence.
The first result of hammering or rolling a vesicular ingot will be nearly to close the " blow holes." I say nearly close advisedly ; for although the vesicles may become divided, and be made to change their shape and vary their capacity, even to the extent of becoming microscopically small, they never wholly disappear.
As the " blow holes " are reduced in diameter, the contained gases are therefore subjected to a very great reduction of volume, and consequent increase of pressure in the inverse proportion to such reduction. For instance, if the initial pressure of the gases in the " blow holes" was 120 pounds per square inch (as observed by Professor Muller), and these cavities were, by hammer or rolls, reduced to one-tenth of their original diameter, their capacities would be but one-thousandth as great as at first, and therefore the pressure of the contained gases would be 120 X 1000=120,000 pounds per square inch. But, in estimating the value of this pressure to produce rupture, we must bear in mind that it acts only upon an area one one-hundredth as great as it did originally.
Now let us suppose that an article made from steel in the above described condition is subjected to a heat of 1000° F. (a dull red). In that case, the gases enclosed in the cavities of the steel will, by reason of their tendency to expand, exert three times the pressure that they did when cold, and if such pressure is symmetrically distributed (a not very likely circumstance), the article will when cold retain its original contour ; but, if there is more internal pressure upon one side of the object when heated than upon its opposite, distortion will naturally result, a thing not at all uncommon in annealing articles of steel. Thus we see that throughout a bar or other forging made from a vesicular ingot of steel there may exist a vast number of magazines of force, of great though very variable intensity, but quite sufficient to aid powerfully any external strain which tends to break the bar.
Now let us see what would be the result if the " blow holes " were without gaseous contents. Were it possible to find an ingot whose "blow holes " were absolutely empty, when it was worked, the sides of the holes would come together, but, owing to the low temperature, they would not weld. At the same time, the holes would become elongated and be made to approach each other transversely, and, in fact, would finally develop into " seams " of a variable length and depth, as they chanced to originate from large or small holes. Such a state of affairs, as may be readily understood, could have very little influence upon the ultimate ability of the bar to resist a tensile strain ; but, if such a bar were subjected to compression, it is easy to see that it would yield unequally, and much sooner on the side having the greater number of such seams.
But it is when such a bar is subjected to a transverse strain, tending to pull it apart in a direction perpendicular to that in which it has been worked, that these seams are the most injurious. Take, for example, an "angle iron " made from such an ingot : it is not at all improbable that the " seams" would be so arranged in the flanges (as at A, A, Fig. 12) as nearly to separate them into a series of rods held together transversely by occasional ligatures of metal. Now, if such a bar be punched or drilled through the "seams," and another bar B be riveted to it (as shown in the figure) and subjected to a strain in the direction of the arrow, it is self-evident that the "angle bar " would be much more likely to be pulled apart transversely through the rivet-hole than if it were made of a homogeneous material.
I think I have said enough about " blow holes " to convince any one that under the ordinary methods of manufacture they are very likely to occur, and that they are exceedingly objectionable things to have in the metal ; and now let us examine the efficiency of some of the methods that have been devised to mitigate the evil of their presence.
Some years since, the late Sir Joseph Whitworth proposed and practically carried out a mechanical process of compressing steel in the ingot mould while it was still fluid or plastic, his intention being to destroy the " blow holes " by the action of the enormous pressure employed. He certainly succeeded in turning out from his works most admirable products in steel ; but I have always had a feeling that the high character of his forgings was due more, much more in fact, to the chemical constitution of the metal, and its having been skillfully heated and carefully worked, than to any qualities resulting from its having been compressed while in a fluid state.
Let us examine this matter a little more closely. Suppose /, Fig. 13, to be a vertical section of an ingot mould filled with fluid steel 6', (having more or less " blow holes" distributed through its mass, as indicated by the small circles,) which may be forcibly acted upon by the plunger P. Now, as fluids under pressure act equally in all directions, it is evident that all the " blow holes " will be reduced in size, and also that the tension of their contained gases will be increased in the inverse proportion to their reduction in volume ; but it is not so clear that there is any action that will cause their removal from the steel altogether.
The same reasoning applies to all systems of vertical compression, whether by the action of carbonic acid, as employed by Herr Krupp, or by high-pressure steam, as at one time used in this country.
Some system of closing the top of the mould and producing a vacuum above the steel would seem to be a more rational mechanical method of removing the "blow holes," than any system of compression; for, as the pressure upon the steel was reduced, the gases in the "blow holes " would expand, and their augmented levity would cause them to escape from the fluid steel with greater rapidity than is possible under any other condition. But, after all, the simplest way of getting rid of the "blow holes " in steel is so to treat the metal before it is run into the ingot mould, that its capacity for heat is increased to such a degree that it will remain fluid long enough to permit the gas-filled " blow holes " to escape as freely as bubbles from water.
This suggestion brings me naturally to a consideration of the "Mitis process "; but before entering upon it, I desire to speak briefly of some points relating to the very important practical operation of hardening and tempering steel. As regards the latter process, it would be for the advantage of all who use tools of steel, if the subject could be disposed of in as few words as comprised a somewhat famous account of snakes in Ireland, which simply declared that " there are no snakes in Ireland." I wish it were possible to say that there is no such thing as the present practice of tempering steel ; for I am firmly of the opinion that much better results can be attained in its use by simply hardening it at such a temperature as by practice with the particular steel used is found most satisfactory, and omitting altogether the lawless and uncertain operation known among mechanics as " drawing the temper."
I have in my possession an admirable illustration of the possibilities of the methods suggested, in the shape of a razor made from the first Bessemer steel produced in this country. This steel, judged by the ordinary standards for "razor steel," would be considered altogether too soft for the purpose; but, nevertheless, by hardening it as much as possible, and leaving it in that condition, it made so satisfactory a razor that my father shaved himself with it for fifteen years.
I have already referred to the views of Colonel Maitland relative to hardening steel, and feel sure that the practical experience of all who have had the most to do with that operation leads to a similar conclusion.
In this connection it will be instructive to quote some of the remarks made by acknowledged experts during the discussion of a paper communicated to the American Society of Mechanical Engineers by Professor John E. Sweet (" The Unexpected which often Happens "; see Transactions American Society of Mechanical Engineers, Vol. VII., pages 156 to 160). In this discussion, Mr. Henry R. Towne, President Yale & Towne Manf. Co., speaks of numerous unsuccessful attempts to harden certain castings of steel, and states that it was finally discovered " that the steel hardened beautifully inside, but that there was on the outside a thin skin of metal, about three to four hundredths of an inch in thickness, which, except by the cyanide process, did not harden at all. In all of the castings there was perfect hardening under this skin ; and finally, the moral of this is that we should look below the surface"—a moral, I will add, which should not be forgotten by those who hope to succeed in the employment of steel. In the same discussion, Mr. Geo. M. Bond, Superintendent Gauge Department, Pratt & Whitney Manf. Co., said : " We find in our experience in making taps and reamers, that in case the steel has been over-annealed and has thus been decarbonized, the hardening does not take effect except under the surface, so that frequently, taps which appear to be soft, if turned again will harden perfectly. I think perhaps the castings referred to by Mr. Towne may have been over-annealed, and in that way a percentage of the carbon eliminated so that the hardening would not take effect upon the outside surface." In the same discussion Mr. Bond further remarks : " We had occasion to make a set of gauges in which the sizes were all two ten-thousandths of an inch larger than the nominal sizes, and five days after the gauges were finished, one of them suddenly gave way in the centre, a crack extending around it spirally, but not so as to injure the ends of the gauge. Out of curiosity, I thought that I would measure the uninjured parts to see if any change had come in the diameter, and I found at both ends the diameter had enlarged forty divisions of the micrometer, which is equal to six ten-thousandths of an inch, and which, as magnified, represented a space to the eye of about three-sixteenths of an inch under the microscope. This shows, I think, that if steel hardens at all, the internal strain must be something tremendous. This will also explain why steel, in being hardened through the centre, has a tendency to shorten under certain conditions."
Professor William A. Rogers, Assistant Professor of Astronomy, Harvard University, said: "The unexpected has always happened to me in this matter of obtaining hardened steel which has a homogeneous temper throughout the entire mass. The nearest approach to an even temper which I have ever been able to obtain has been at the works of Miller, Metcalf & Co., of Pittsburgh, and of Brown & Sharpe, of Providence. A short time since I asked the latter firm to set a price upon a hollow steel cylinder six inches in diameter, three feet in length, having walls half an inch in thickness, hardened and ground on the outside only. The price which was set—from $300 to $500 without guarantee against flaws—may be taken as the estimate of the extreme uncertainty always attending any difficult case of tempering held by those who have a full comprehension of the difficulty of the problem.
"The difficulty of giving a homogeneous temper to a large mass of metal is so great, according to my experience, that it is never perfectly done. The test which I apply as the gauge of an even temper is a very severe one. If all the lines ruled upon a highly polished bar of tempered steel have the same appearance, the temper of the graduated surface is good. I have, however, never yet seen a set of graduations in which the diamond has with a constant pressure cut all the lines to the same depth. The diamond acting upon this polished surface detects the lack of homogeneity in the most perfect manner. If there is any person in this country, or in the world for that matter, who can temper a bar of steel three feet in length and for a depth of even a quarter of an inch, at any price, I should be glad to make his acquaintance."
Mr. George Ede, in that chapter of his work on "The Management of Steel" (edition of 1866) descriptive of the method of "toughening of steel in oil," as at that time practised in the Gun Factories' department of Her Majesty's Royal Arsenal, Woolwich," says, relative to hardening solid steel shot : " Thick lumps of highly carbonized steel, whether hardened in oil or pure water, or water with a film of oil upon its surface, cannot be hardened without becoming fractured either internally or externally." In this matter of hardening steel, the value of the " personal equation " of the workman is all important. It is not uncommon to find a practical mechanic who usually has good success in the use of a certain kind of steel with which his neighbors, equally skillful perhaps in other matters, can do nothing. So often have I encountered this fact, that I am inclined to believe that if a person in pursuit of information as to the proper quality of steel to use for any given article should travel through this land and obtain the honest opinion of all who were making the article in question, that " the last state of that man would be worse than the first"; for the chances are that every person consulted would have an opinion differing from those of his fellow craftsmen, and although when our traveller started on his search for technical wisdom he was positive that he knew nothing, he could not rejoice in even that negative certainty when he returned. In the present state of our knowledge, there is no recognized uniform scientific method of hardening and tempering steel : all we have is a tentative art, as crude in its development as it is obscure in its origin.
And now let me crave your indulgence for a few words relative to the "Mitis process." The word "mitis" is a Latin adjective, meaning mild, soft, or ductile, and it was selected because of its appropriate signification, as the designation of the new art, by its Swedish inventors.
I regard this process as one of the most important improvements in the metallurgy of iron and steel that has been brought forward during the past twenty years. By its means we can produce castings of melted wrought iron, or of steel of any desired hardness, that, without having been annealed, can be forged, welded, bent cold, or worked by machine tools, with as great facility as ordinary forgings of wrought iron and steel.
From among various samples I have selected two articles which will serve to give a good idea of the possibilities of the new art. One of these is a horse-shoe with " nail holes " and " creases " complete, cast of melted wrought iron in an iron mould—an impossibility by any other process. The other is a beefsteak pounder, in the condition in which it left the " dry composition " mould in which it was cast ; this also is made of melted wrought iron. On the extremity of each tooth there is what appears to be a wire. These seeming wires were made by the melted wrought iron filling the "vent holes " of the mould; a thing which I never saw to anything like the extent shown in this sample in the case of castings made of any metal by any other process. In fact, one of the most remarkable attributes of this new art is the extreme fluidity of the metal at the time of casting, and in this fact is one of its greatest technical advantages ; for this exceeding fluidity enables all the " blow holes " to escape before the casting or ingot solidifies, thus getting rid of one of the chief obstacles to sound forgings in iron and steel.
This fluidity of the melted metal is produced by the addition of a small percentage (0.05 to 0.1 of one per cent) of the metal aluminum to the melted wrought iron or steel immediately before casting: this addition at once produces a degree of fluidity in the molten metal comparable with that of water.
The reason for this result cannot be explained, but it can perhaps be made more comprehensible by an illustration.
The soft metals—tin, lead, zinc, antimony, bismuth, etc.—melt singly at temperatures varying from 600° to 1000° F., but if a proper selection and combination from these metals be made, we shall obtain an alloy that will melt at the surprisingly low temperature of 170° F.
Under ordinary conditions, melted wrought iron or low steel is pasty or semi-fluid, and it has been found that, without increasing the temperature, the addition of the small percentage of aluminum named acts upon the metal very much as the soft metals in our illustration act upon each other,—it increases the fluidity of the mass, and thus enables results to be attained that have hitherto been impossible. Castings made by this process are, as a rule, ten per cent stronger than the scrap iron melted to produce them ; and in the case of steel there is a like increase of strength without any diminution of elongation previous to rupture, a property which is of great practical value. I regard this process as of great value in the manufacture of material for great guns, as by its use sounder and stronger ingots of steel can be insured from which to forge the component parts of built-up guns; and, furthermore, I believe it quite possible, by the use of this process in connection with the Rodman system of casting, to make a solid cast-steel gun that will be more efficient and a great deal cheaper than any built-up gun can possibly be.
The question of national defense is no trivial one, and before deciding upon the method to be adopted in the manufacture of our future arms, the merits of all systems should be carefully considered. The laudable purpose of our government to provide adequately for the defense of the nation gives to the general subject of the manufacture of heavy guns great interest at this time. That people who would preserve peace must be prepared successfully to defend peace; and a nation so prepared will rarely need to use its weapons.
The industrial victories of peace make possible war's triumphs in the defense of peace. And therefore a nation's first step in its own defense should be to foster the industries of its own people, and cordially encourage those who, with thoughtful brain, willing hand, and ready purse, stand prepared to mine coal and ore, to make furnaces roar and melt with fervent heat, while mighty engines throb and ponderous hammers beat, and, anon, gigantic tools, in whose stalwart frames and cunning fingers are crystallized the brain and brawn of generations of workers in thought and substance, give final shape and proportion to the arms of our defense, that peace and concord, prosperity and happiness, shall be ours while time endures. To this end should all labor tend—
" Till the war drum throbs no longer, and the battle flags are furled,
In the Parliament of man, the Federation of the world ;
When the common sense of most shall keep a fretful realm in awe.
And the kindly earth shall slumber lapt in universal law."
DISCUSSION
Lieutenant R. R. Ingersoll.—Mr. Chairman and Gentlemen :—I have only one objection to offer to what the author has said in his able and instructive paper, and that is as to the possibility of making a modern gun with Mitis iron or steel by the aid of the Rodman process of casting. There is but little doubt that a sufficiently strong gun cannot be produced by the Rodman system of casting, for the simple reason that, however beautiful the theory, it is impossible to realize in practice the state of initial compression of the bore indicated by the theory. This is conclusively shown in the case of cast-iron guns, the process of casting which is very much easier than is the case with steel guns. The elastic limit for compression of cast iron being taken at about 26,000 pounds per square inch, the initial compression at the surface of the bore of the completed gun should equal this amount in order to compete with built-up guns ; but thus far, an initial compression of 12,000 pounds, in the case of the 12-inch cast-iron gun built at South Boston, is about the limit reached. Again, the elastic strength of Mitis iron being, as compared to gun steel, not over two-thirds the latter, the material is not sufficiently strong, especially in view of the fact that the day is not far off when 2500 f. s. and 25 tons pressure per square inch will be as common as 2000 f. s. and 15 tons pressure per square inch are today.
I am glad to hear the author take such strong ground as he does in regard to the existence of blow holes in steel castings. It is as strong an argument as can be desired against the practicability of producing sound, strong guns of cast steel by the Rodman process.
The paper is most interesting and instructive, and I regret that the author is not present, because he could, perhaps, clear up many points upon which some of us may possibly be doubtful.
The following reply to Lt. Ingersoll's remarks was received by letter:
Mr. Durfee.—Mr. Chairman and Gentlemen :—In reply to the criticism of Lieutenant Ingersoll, I will say that the behavior of melted wrought iron or steel, when treated by the "Mitis Process," is so different from that of these metals not so treated that, in the absence of actual experiment, it is hardly just to assert that this new process is not available for the manufacture of heavy guns by the Rodman system, or some modification thereof; and if a thorough trial should prove its unfitness for use in this way, it could hardly fail to be of great value in ensuring sound ingots from which to forge the parts of built-up guns.
In view of the many failures of built-up guns made by the aid of the best talent and largest experience abroad, I think it the part of a conservative wisdom to hasten slowly enough in this (for our country) new manufacture ; to "prove all things and hold fast that which is good."
On the motion of Commander W. T. Sampson, a vote of thanks was given to the author for his valuable and instructive paper.