XI.THEORIES OF HARDENING.
The hardening of steel is such a marked phenomenon, and one of so great importance, that it has always attracted a great deal of attention, and many theories have been put forward in explanation.
Before chemistry was brought to bear upon the subject the proposed theories were based upon assumption, and as there were no proofs one had as much right to consideration as another, and none seemed to be altogether satisfactory.
Since science has taken up the question the theories are about as numerous as the investigators, and while no one can claim as yet to have settled the matter definitely, each one has an apparent basis of reason deduced from observed facts.
Among early observations it was noted that when unhardened steel and hardened steel were dissolved in acid a much larger amount of carbon was found in the solution of the unhardened than in that of the hardened steel. This led, first, to the distinction of combined carbon and graphitic carbon, a distinction that has been maintained through subsequent investigations. It seems to be well established now that there is a definite carbide of iron,Fe₃C, and some observers believe it to be the hard substance in hardened steel.
Following this came the announcement that these conditions,combinedandgraphiticcarbon, represented two different forms of carbon, and they were designated ascementcarbon andhardeningcarbon; also asnon-hardeningandhardeningcarbon. Later investigation having established the existence of the carbideFe₃C, this was claimed to be the hard body, but this has not met universal acceptance.
Another investigator, studying by means of the pyrometer and observing heat phenomena, concludes that hardening is due to an allotropic condition of the iron itself; that when iron is heated above the recalescent-point, and presumably below granulation, it becomes in itself excessively hard; that sudden cooling prevents its changing from this form, and so, when there is carbon present, the result of quenching is great hardness.
When steel is allowed to cool slowly to below recalescence, the iron assumes another form, and one which cannot be hardened by quenching; this latter is known as α iron, and the hardening kind as β iron. A later investigator finds it necessary to have a third allotropic form to meet some of the phenomena, which he designates by another Greek letter.
Another investigator establishes independently the saturation-point, which was pointed out and published twenty years ago, viz., somewhere about 90 to 100 carbon; he fixes the saturation-point at 89 carbon and gives the formulaFe₂₄C. He assumes that this is an exceedingly unstable carbide, that it is formed between recalescence and granulation, and can only be fixed by quenching, and that when steel is quenched the fixing of this carbide is the cause of hardness.
A still later investigation establishes this saturation-point at about 100 carbon by observing that in hardened steel of 135 carbon there is a combination of 100 carbon which is the excessively hard part of the steel, and a portion containing the remaining 35 parts of carbon that is not quite so hard, and he suggests a fourth allotropic form to cover this part.
It is also suggested that steel should be considered and treated as an igneous rock; judging from the appearance of magnified micro-sections, this suggestion appears to be a happy one for the purpose of making comparisons.
The above theories of hardening, and others, are not to be regarded as antagonistic or contradictory, doubtless there are germs of truth in every one of them, or each one may be merely the individual’s way of suggesting an explanation of the same observed phenomena, so that when a final conclusion is reached each may be found to have been travelling in the same direction by a different path. It is certain that able, patient, painstaking, men are working faithfully to produce a solution of the problem, and even if their ideas, as briefly given above, do seem to be contradictory it would only evince deeper ignorance and a stupid mind in any who should attempt to ridicule or unduly criticise honest work before it is completed. While these investigations are going on, and before any definite conclusion is reached, is there any well-established safe ground for the steel-worker and the engineer to stand upon? There certainly is a good working hypothesis for all to use, and one which it is believed will always be the right one to follow no matter what the final explanation of the remarkable phenomena of hardening, tempering, and annealing may prove to be.
After many years of careful experimenting and study Prof. J. W. Langley came to the conclusion that no matter what the final result might be as to carbides, allotropic conditions, etc., that if steel were considered as iron containing carbon in solution, whether it were a chemical combination or a mere solution, and that cold steel be regarded as a congealed liquid in a state of tension, then all known phenomena could be accounted for, and all known conditions could be produced with certainty by well-known applications of heat and force.
When carbon is in the so-called combined condition, then the solution maybe compared to pure sea-water; when the carbon is partly combined and partly graphitic, the solution may be compared to muddy sea-water, the mud representing the graphitic carbon.
When the carbon is practically all graphitic, as in over-annealed steel, then the solution may be compared to thoroughly muddy fresh water.
This hypothesis of solution agrees well with the saturation noted; then about 100 carbon is all that iron will dissolve without extraneous force; and higher carbon must be forced into solution by the work of hammers, presses, or rolls.
This gives reason to the experienced tool-maker’s well-known preference for well-hammered steel.
The hypothesis of tension, probably molecular, covers all of the phenomena of excessive hardness due to high heat, which means high molecular motion checked violently by sudden quenching. It accounts for the progressive softening due to every added degree of heat, and it accounts for rupture, cracking, due to excessive heat or to any unevenness of heat.
Without this hypothesis of tension it is difficult to understand why quenching should rupture a piece of steel, no matter what the degree of heat, or how uneven it might be.
Without it, too, it is hard to see how successive additions of heat can cause gradual changes from β to α iron, or from an unstable carbide to an imperfect solution. It would seem that the allotropic changes, or the decompositions of carbides, must be more marked than the gradual changes from hard to soft which we know to take place by slow and gentle accretions of heat.
There is no property of steel known to the author which is not covered by Langley’s hypothesis, and therefore it is put forward with confidence for engineers and steel-users to work by until the scientists shall have completed their investigations, and after that it is believed that it will be a safe working hypothesis, because science does not change facts, it only collates them and reveals the laws of action.
Under this hypothesis of Langley’s we may definehardnessastension,softnessas absence of tension.
This is not stated as established fact; it is given as a simple definition to cover the known phenomena until the final solution of the problem shall lead to a better explanation.
Regarding steel as a solution of carbon in iron, one important fact may be set down as established thoroughly: that is, that the more perfect the solution under all circumstances the better the steel.
Continued application of heat in any part of the plastic condition allows carbon to separate out of solution into a condition of mere mixture; it converts the clear sea-water into muddy water; this is the reason why so much emphasis has been given in previous chapters to the harmfulness of long-continued heating.
In every case, when steel is hot enough for the purpose desired, it should be removed at once from the fire.