X.IMPURITIES IN STEEL.
Any elements in steel which reduce its strength or durability in any way may be classed as impurities.
A theoretical ideal of pure steel is a compound of iron and carbon; it is an ideal that is never reached in practice, but it is one that is aimed at by many manufacturers and consumers, because experience shows that, especially in high steels, the more nearly it is attained the more reliable and safe is the product.
All steel contains silicon, phosphorus, sulphur, oxygen, hydrogen, and nitrogen, none of which add any useful property to the material. It is admitted that, starting with very small quantities of silicon or phosphorus in mild steel, small additions of either element will increase the tensile strength of the steel perceptibly up to a given amount, and that then the addition of more of either one will cause a reduction of strength. The same increase of strength can be obtained by the addition of a little carbon, producing a much more reliable material. It is not known that even such slight apparent gain in strength can be made by using oxygen, nitrogen, or hydrogen.
Manganese is present in all steel as a necessary ingredient, it gives an increase in strength in the same way as phosphorus, and when increased beyond a small limit it causes brittleness. Hadfield’smanganese steel is a unique material, not to be considered in connection with the ordinary steel of commerce.
Webster’s experiments are perhaps the most complete of any that show the effects of small increases of silicon, phosphorus, sulphur, and manganese, but as these are not completed they are not quoted here, because Mr. Webster may reach additional and different results before these pages are printed.
The chief bad qualities of steel that are caused by these impurities are known as “red-shortness,” “cold-shortness,” and “hot-shortness.”
A steel is called red-short when it is brittle and friable at what is known commonly as a low red heat—“cherry red,” “orange red.”
Red-shortnessis caused chiefly by sulphur or by oxygen; many other elements may produce the same effects; it seems probable that nitrogen may be one of these, but the real action of nitrogen is as yet obscure.
A red-short steel is difficult to work; it must be worked at a high heat—from bright orange up to near the heat of granulation—or it will crack. When hardened, it is almost certain to crack. When red-short steel is worked with care into a sound condition, it may when cold be reasonably strong, but hardly any engineer of experience would be willing to trust it.
Hot-shortsteel is that which cannot be worked at a high heat, say above a medium to light orange, but which is generally malleable and works soundly at medium orange down to dark orange, or almost black.
This is a characteristic of most of the so-called alloy steels, or steels containing considerable quantities of tungsten, manganese, orsilicon. It is claimed that chrome steel may be worked at high heats and that it is less easily injured in the fire than carbon steel. This is not within the author’s experience. It is this property of hot-shortness that makes the alloy steels so expensive; the ingots cannot be heated hot enough nor worked heavily enough to close up porosities, and therefore, there is a heavy loss from seams.
The range of heat at which they can be worked is so small that many re-heatings are required, increasing greatly the cost of working.
As compared to good carbon steel they are liable to crack in hardening, and when hardened they are friable, although they may be excessively hard.
Cold-shortsteel is steel which is weak and brittle when cold, either hardened or unhardened. Of those which are always found in steel, phosphorus is the one well-known element which produces cold-shortness.
It is clear that no one can have any use for cold-short steel.
Red-short or hot-short steel may be of some use when worked successfully into a cold condition, but cold-short steel is to be avoided in all cases where the steel is used ultimately cold.
If the theoretically perfect steel is a compound of iron and carbon, it cannot be obtained in practice, and the only safeguard is to fix a maximum above which other elements are not to be tolerated.
In tool-steel of ordinary standard excellence such maximum should be .02 of one per cent; it may be worked to easily and economically, except perhaps in silicon, which element is generally given off to some extent by the crucible; it should be kept as low as possible, however,say well under 10, one tenth of one per cent. Some people claim that a little higher silicon makes steel sounder and better; but any expert temperer will soon observe the difference between steels of .10 and .01 silicon. For the highest and best grade of tool-steel the maximum should be the least attainable. Every one hundredth of one per cent of phosphorus, silicon, or sulphur will show itself in fine tool-steel when it is hardened. It is assumed, of course, that such impurities as copper, antimony, arsenic, etc., exist only as mere traces, or not at all.
As oxygen must be at a minimum, no one has yet succeeded in making a really fine tool-steel from the products of the Bessemer or of the open-hearth process.
The removal of the last fractions of these impurities is difficult and expensive; for instance, a steel melting iron of
may be bought for 2 cents a pound or less, whereas an iron of
can hardly be bought for less than 5 cents a pound.
This difference of three cents a pound is justifiable when the highest grade of tool-steel is to be made; and it would be silly to require any such material in any spring, machinery, or structural steel.
In addition to these impurities there are other difficulties to be guarded against, chief among which is an uneven distribution of elements.
In all steel there is somesegregation; that is to say, as the liquid metal freezes, the elements are to some extent squeezed out and collected in that part of the ingot which congeals last. It is claimed that in the Bessemer and Open-hearth processes any ferro-silicon added to quiet a heat, or any ferro-manganese added to remove oxygen, are at once absorbed and distributed through the mass, and so when any serious irregularity is discovered it is charged tosegregation.
A heat may produce billets of 75 carbon and 120 carbon, and again it is called segregation.
As a rule, inertia has more to do with such differences than segregation. One crucible of steel may produce an ingot containing 90 carbon and 130 carbon. Segregation has nothing to do with this: a careless mixer has put a heavy lump of 140- or 150-carbon steel in the bottom of the pot and covered it up with iron. The steel melted first and settled in the bottom of the pot, the iron melted later and settled on top of the steel, and they did not mix. The teeming was not sufficient to cause a thorough mixing.
Segregation covers a multitude of sins.
Exactly how much is sin and how much is segregation will not be known until analyses are made of the top, middle, and bottom of the bath, and of the contents of the ladle, these to be compared to analyses of the top, bottom, and middle of the ingots. There is certainly an unavoidable amount of segregation, and as equally certain an amount of curable irregularity due to inertia.
After steel is melted, whether in a crucible, an open hearth, or a Bessemer vessel, it boils with more or less violence. This boiling is caused by ebullition of gases, and if steel be poured into moulds while it is boiling the resulting ingot will be found to be honeycombed to an extent that is governed by the degree of the boiling.
If a heat boils violently and persistently, it is said to be “wild,” and if a wild heat be teemed the ingots will be honeycombed completely; such ingots cannot be worked into thoroughly sound steel, and no melter who has any regard for his work will teem a wild heat if he knows it.
To stop the boiling is called “dead-melting,” “killing” the steel, so that it shall be quiet in the furnace and in the moulds.
A crucible-steel maker who knows his business can, and he will, always dead-melt his steel. It only requires a few minutes of application of a heat a little above melting temperature, and this can be applied by a skilled melter without burning his crucible or cutting down his furnace; this is indeed about all of the art there is in crucible-melting, the remaining operations being easy and simple.
Dead-melting in the Bessemer vessel is not possible by increase of time; wild heats are managed differently, probably by adding manganese or silicon, or both, but exactly how is not within the author’s experience.
Dead-melting in the open hearth would appear at first sight to be always possible, but there are more difficulties in the way than in the case of crucible-melting.
The heat may be wild when the right carbon is reached, and then the melter must use a little ferro-silicon, or silico-spiegel, or highly silicious pig, or aluminum, and he must use good judgment so as not to have his steel overdosed with any of these. From half an ounce to an ounce of aluminum to a ton of steel is usually sufficient, and although any considerable content of aluminum is injurious to steel there is little danger of its being added, because of its cost, and because a little too much aluminum will cause the ingots to pipe from top to bottom.
Silicon seems to be the most kindly element to use, and it is claimed that a content of silicon as high as 20 is not injurious; some people claim that it is beneficial. That it does help materially in the production of sound steel there can be no doubt, and if such steel meets all of the requirements of the engineer and of practice it would seem to be wise not to place the upper limit for silicon so low as to prevent its sufficient use in securing soundness. But the author cannot concede that as much as 20 silicon is necessary. In crucible practice high silicon is not necessary; in “melting-iron,” or iron to be melted, it means so much dirt, indicating careless workmanship; but there will always be a little silicon present which the steel has absorbed from the walls of the crucible during the operation of melting. In high tool-steel silicon should be at the lowest minimum that is attainable.
This discussion of wild heats may appear to be outside of the scope of this work, and to belong exclusively to the art of manufacturing steel, of which this book does not pretend to treat. This is true so far that it is not recommended that the engineer shall meddle in any way with the manufacturer in the management of his work; on the other hand, itis vital to the engineer that he should know about it, because wild steel may hammer or roll perfectly well, it may appear to be sound, but the author cannot believe that it is ever sound and reliable.
Again, it has a scientific interest; that wildness is due to too much gas, and probably to carbon-gas, may be shown by an illustration.
It has its parallel in the rising of the iron in a puddling-furnace at the close of the boil, a phenomenon with which every one is familiar who has watched a heat being boiled or puddled. That all of the iron does not run out of the puddling-furnace at this stage is owing to the fact that there is not heat enough in the puddling-furnace to keep the iron liquid after it has been decarbonized.
During the running of a basic open-hearth furnace an apparently dead heat was tapped; before the steel reached the ladle there was a sort of explosion; the steel was blown all over the shop, the men had to run for their lives, and not one tenth of the steel reached the ladle. The manager was rated roundly for carelessness in not having dried his spout, and the incident closed. A few days later another quiet heat was tapped and it ran into the ladle; about the time the ladle was full the steel rose rapidly, like a beaten egg or whipped cream, and ran out on to the floor, cutting the sides of the ladle, the ladle-chains, and the crane-beams as it flowed. The men ran, and there was no injury to the person.
Again the manager was blamed, this time for having a damp ladle, and he was notified of an impending dismissal if such a thing occurred again. He protested that he knew the ladle and the stopper were red-hot, that he had examined them personally and carefully, and knew he stated the truth.
There were several reasons for looking into the matter farther: first, the man in charge was known to be truthful and careful, so that there was no reason for doubting his word; second, if the vessel and rod were red-hot, there could be no aqueous moisture there; and, finally, such an ebullition from dampness was contrary to experience, as a small quantity of waterundera mass of molten iron, or slag, results almost invariably in a violent explosion, like that of gunpowder or dynamite.
Upon inquiry it was found that prior to both ebullitions there had been a large hole in the furnace-bottom, requiring about a peck of material to fill it in each case. Magnesite was used; the magnesite was bought raw, and burned in the place. It is well known that it takes a long time and high heat to drive carbonic acid out of magnesite, and it was surmised that insufficient roasting might have caused the trouble. Samples of burned and of raw magnesite were sent to the laboratory, and the burned was found to contain about as much carbonic acid as the raw magnesite. Then the case seemed clear: This heavily charged magnesite was packed into the hole; the heat was charged and melted. The magnesite held the carbonic acid until near the close of the operation; then the intense heat of the steel forced the release of the gas, which was at once absorbed by the steel. Owing to the superincumbent weight of the steel the gas was absorbed quietly, and when the weight was removed the gas escaped, exactly as it does at the close of puddling or in the frothing of yeast.
Whether the carbonic acid remained such, or whether it took up an equivalent of carbon and became carbonic oxide, and then again took up oxygen from the bath, and so kept on increasing in volume, is not known.
The facts seem clear, and the collateral proof is that thorough burning of the magnesite, and of any dolomite that was used, prevented a recurrence of any such accidents.
Such ebullitions have occurred and caused the burning to death of pitmen, and the statement of the above case may be of use to melters in the future who have not met such an experience.
Oxygen and nitrogen are present in all steel and both are injurious, probably the most so of all impurities.
The oxides of iron are too well known to need discussion or description; they are the iron ores mixed with gangue. They are brittle, friable, hard, and weak, like sandstones. Mixed in steel they can be nothing but weakeners, elements of disintegration. Let any one take a handful of scale—or rust—oxide of iron, in his fingers and crumble it, and it will be difficult for him to imagine how such material could be anything but harmful when incorporated in steel. Langley has shown, and other scientists have confirmed him, that oxygen may exist in iron in solution, and not as oxide; the discovery was attended with the assertion that such dissolved oxygen produced excessive red-shortness. The proof that red-shortness was caused in this way was completed by the removal of the oxygen from some extremely red-short steel; the red-shortness disappeared with the oxygen and the steel worked perfectly.
When steel is melted very low in carbon, by any process, it is certain to be red-short and rotten unless the greatest care be used to preventthe introduction of oxygen. Crucible-steel of 15 carbon or less will as a rule be red-short and cold-short; it will not weld, and is generally thoroughly worthless. The same material melted to contain 18 to 25 carbon will be tough and waxlike, hot or cold. It will weld easily into tubes, and may be stamped cold into almost any desired shape.
Bessemer or open-hearth steel of less than 8 carbon is almost certain to be equally worthless, whereas the same material blown or melted not below 10 or 12 carbon, and re-carbonized not above 20, will be tough and good at any heat under granulation, and equally good and tough when cold.
As to Bessemer steel, the author cannot say whether it would be possible to stop the blow between 10 and 15 carbon or not, but it seems certain that if there be no overblowing red-shortness and cold-shortness may be avoided by carbonizing back to about 15 by the use of manganese or silicon, or both together.
In the open hearth it is always possible to stop the melt at 10 carbon, and to deoxidize the heat so as to avoid shortness, and not to go above 20 carbon. Such steel will be sound and tough; it will weld and stamp perfectly, and will be satisfactory for all reasonable requirements.
The reason of this seems to be simple and plain: In melting or blowing out the last fractions of carbon below 10 to 15 the same quantity of air per second or minute must be used as when burning out the higher quantities, and now there is so little carbon to be attacked that the oxygen necessarily attacks the iron in greater and greater force as the carbon decreases.
This leaves an excess of oxygen in the steel which cannot be removed by the ordinary quantities of silicon, or manganese, or aluminum.
If more manganese or silicon be used, the red-shortness and weakness can be cured largely; but then the carbon is raised considerably, and thus the steel is brought up to where it would have been without this excessive decarbonizing, with the difference that it is not quite so strong.
What good is there, then, in extremely low melting?
It must be admitted that there are tough, good-working steels in the market of carbon < 5, manganese < 20. They are made in small furnaces, worked with great care; the product is expensive, and, unless it is wanted to be welded in place of common wrought iron, it is in no case as good as well-made steel of 12 to 20 carbon; even for welding the latter is superior if the worker will only be satisfied to work at a lemon instead of a scintillating heat.
These special cases do not militate against the general fact that extremely low steel is usually red-short and weak.
The above is written for the consideration of those engineers who think they are going safe when they prescribe low tensile strength and excessive ductility. If these requirements meant the reception of pure, or nearly pure, iron, indicated by the low tenacity and high stretch, then they would be wise; but if they result, as they almost certainly do, in initially good material rotted by overdoses of oxygen the wisdom may not be so apparent.
The real influence of nitrogen is not known to the author. Percy shows that nitrogenized iron is hard, exceedingly friable, and causes a brilliant, brassy lustre. He also says nitrogen is driven out at ayellow heat; doubtless this is true of the excess of nitrogen, but it has been shown inChapter IIthat melting in a crucible will not drive the nitrogen out of Bessemer steel.
When crucible-steel not made from Bessemer scrap and Bessemer steel of equal analysis are compared in the tempered condition, there is almost invariably a yellowish tinge over the fresh Bessemer fracture which distinguishes it from the crucible-steel. The Bessemer steel is also the weaker. These differences are believed to be due to nitrogen.
Langley maintains his belief that oxygen is still the chief mischief-maker; the author believes nitrogen to be the more potent of the two; there is no known way to remove the nitrogen, and there the question stands.
It has been stated time and again that these impurities are elements of disintegration, and that it would be wise in every case to restrict the quantities allowable within reasonable limits, giving the steel-maker sufficient leeway to enable him to work efficiently and economically, and at the same time to keep the quantities of these impurities as low as possible.
On the other hand, able, successful, and conservative engineers have claimed that if the steel-maker meets their physical requirements as shown by prescribed tests they, the engineers, should be satisfied; that they should not interfere with chemical composition, as they had no fear of subsequent disintegrations.
This argument was answered by the statement that skilled steel-workers could manipulate poor steel so as to bring it up to the requirements;that the well-trained workers in the bridge-shops would not abuse the steel; that the inherent deficiencies would not be developed; the work would go out apparently satisfactory; and that it might remain so for a long time, in the absence of unusual shocks or strains, but that in an emergency such material might fail because of deterioration where a purer material would have held on. In the absence of proofs such statements have been met with a smile of incredulity.
Fortunately some proofs are now at hand, and as the method of getting them has been obtained, more will follow from time to time.
InEngineering, Jan. 17, 1896, Mr. Thomas Andrews, F.R.S., M.Inst.C.E., gives the following cases:
A fracture of a rail into many pieces, causing a serious accident.
A broken propeller-shaft which nearly caused a disastrous accident.
Analysis of the rail:
It is clear that the sulphur is excessive, and that it was neutralized so as to make the steel workable by an excess of manganese.
Of the propeller-shaft Mr. Andrews says chemical analysis of outside and central portions of the shaft showed serious segregation.
“The percentage of combined carbon was nearly 50 per cent greater inthe inside of the shaft than on the outside; the manganese was also in excess in the inside of the shaft; the phosphorus and sulphur had also segregated in the interior of the shaft to nearly three times the percentage of these elements found near the outside of the shaft.”
Unfortunately Mr. Andrews does not give the analysis of the shaft.
A number of micro-sections of the rail and of the shaft were made and examined.
“Numerous micro-sulphur flaws were found, varying in size from 0.015 inch downward, interspersed or segregated in the intercrystalline junctions of the ultimate crystals of the steel, and being located in such a manner as to prevent metallic cohesion between the facets of the crystals, thus inducing lines of internal weakness liable to be acted upon by the stress and strain of actual wear.”
The dimensions of these flaws in the rail varied from .0150 × .0012 to .0010 × .0004 parts of an inch.
In the shaft from .0160 × .0030 to .0020 × .0016 parts of an inch.
In the rail he found as many as 14 flaws in an area of only 0.00018 square inch, equal to nearly 60,000 flaws per square inch.
In the shaft he found as many as 34 flaws in an area of only 0.00018 square inch, equal to nearly 190,000 per square inch.
In speaking of the shaft he says: “In addition to blow-holes, air-cavities, etc., the interior of the shaft was literally honeycombed with micro-sulphide of iron flaws, which were meshed about and around the primary crystals of the metal in every direction.” “The deleterious effects of an excess of manganese in interfering with the normalcrystallization of the normal carbide of iron areas were also perceptible.”
As the number of micro-sulphur flaws in the shaft were about three times as many as in the rail, we may assume that the shaft contained at least as large a percentage of sulphur as the rail, and, owing to the general honeycombed structure, it would not be a far guess to assume that the steel was teemed wild.
“The deleterious effect of these treacherous sulphur areas and other microscopic flaws, with their prolonged ramifications spreading along the intercrystalline spaces of the ultimate crystals of the metal and destroying metallic cohesion, will be easily understood.”
“Constant vibration gradually loosens the metallic adherence of the crystals, especially in areas where these micro-flaws exist. Cankering by internal corrosion and disintegration is induced whenever the terminations of any of the sulphide areas or other flaws in any way become exposed at the surface of the metal, either to the action of sea-water, or atmospheric or other oxidizing influences. In many other ways, also, it will be seen how deleterious is their presence.”
“Internal micro-flaws of various character are nevertheless almost invariably present in masses of steel, and constitute sources of initial weakness which not unfrequently produce those mysterious and sudden fractures of steel axles, rails, tires, and shafts productive of such calamitous results. A fracture once commencing at one of these micro-flaws (started probably by some sudden shock or vibration, or owing to the deterioration caused by fatigue in the metal) runsstraight through a steel forging on the line of least resistance, in a similar manner to the fracture of glass or ice.”
It is understood that similar investigations are being carried out on an extensive scale by Prof. Arnold; in the meantime the above cases should satisfy any one that these impurities are elements of disintegration, and that the less there are of them in any steel the better for the steel.
It seems clear that if 10 sulphur will cause 60,000 flaws per square inch, 01 sulphur ought not to cause more than one tenth of that number; or, if an equal number, then they could only be one tenth of the size.
The segregation found in the shaft is so excessive that it would seem probable that there was a good deal of sin there also; but, even if it were unavoidable segregation, the harm would have been just so much the less if there had been less of total impurities present to segregate.
Arsenic is known to be very harmful in tool-steel, and it is proper to assume that it can do no good in structural steel. In any case where the properties of steel do not come up to the standard to be expected from the regular analysis examination should be made for arsenic, antimony, copper, etc. These are not as universal constituents of steel as silicon, phosphorus, sulphur, and manganese, but they are present frequently, and in any appreciable amount they are bad.