How and Why Do the Steels Harden?

Now, from all of these facts, what, shall we say, is the cause of the hardening of steel?

The explanation most generally accepted seems to be, that, of the three allotropic forms of iron the gamma and beta varieties are very much harder than alpha iron, which is the one which we have in annealed steel at ordinary temperatures, and, of the two, the beta is harder than the gamma variety. It is thought also, that carbon, perhaps as carbide of iron, is held in solution in gamma or beta iron after the quenching, and increases the hardness proportionately with increase of the carbon content of the steel. While this carbon or iron-carbide solid solution may and probably does itself confer additional hardness to the steel, its main function is to retard or slow down the change from gamma into beta and alpha iron, which change in carbonless iron and low carbon steels is so insistent and extremely rapid that not even the most severe quenching, as in ice-water or liquid air, can prevent or stop it. Not only do we fail to get austenite, which is gamma iron, until we get 1% or more of carbon present and quench from very high to very low temperatures, but we cannot even stop the transition at beta iron, the next lower allotropic variety until we have at least .30% of carbon, and, for serviceable hardening fully .60% of carbon in the steel.

Fortunately for us, this beta form of iron is the one we want, for it is harder and more useful than the gamma form.

Our most serviceable constituent, martensite, then, is a solid solution of carbon or iron carbide in beta iron. It is magnetic but this probably results from its containing some alpha iron through incomplete stoppage of the change by the quenching.

As has been stated, “tempering,” which means careful reheating to 400° F., 500° F., or 600° F., allows the slight “slipping” of enough of the beta solution, always eager at temperatures below the point of recalescence to return to alpha condition, to relieve the excessive brittleness of the hardened steel.

No. 72. The “Sorbite” Grain Is Produced by Cooling in a Blast of Air after Annealing. It Gives Good Wearing Properties
(Magnification 70 Diameters)

Annealing is the complete release of beta iron and the “trapped” carbon which allows of their return to the normal condition of pearlite with alpha iron. To accomplish this, the hardened steel has to be heated above its point of recalescence and cooled more or less slowly. Different speeds of cooling give different grain size, structures and physical properties.

This explanation of hardening, which is known as the “allotropic” theory, is not universally accepted, conclusive evidence being lacking at more than one point.

It must be stated also that some who hold the “allotropic theory” of hardening doubt the existence of beta iron. These contend that the so-called beta iron and martensite are only decomposition or transition forms of gamma iron and austenite.

Two or three other theories have been more or less strongly advocated but these also suffer from lack of evidence. The one which perhaps ranks next in number of advocates is the “carbon theory.” Its supporters contend that by quenching, alpha iron is made to hold carbon or iron carbide in solid solution and that it is this solution of carbon or carbide which gives to the steel hardness in proportion to its carbon content. Others hold that the great density or strain under which quenched steel exists accounts for the hardness.

Like many other great problems of the universe this one is not yet conclusively or satisfactorily solved, so reluctantly does Nature yield her secrets. But while it may not have explained all of the “whys” and “wherefores,” the work which investigators have done has brought about great improvement in methods of manufacture and quality of the alloys which are making our civilization greater and this Age more wonderful.

CHAPTER XXIII
THE EQUILIBRIUM DIAGRAM OF THE IRON-CARBON ALLOYS

It was a great day for metallurgy when it was discovered that molten mixtures were governed by the same natural laws which govern our ordinary liquid solutions. Probably it was because most metallic alloys are solid at ordinary temperatures and liquid only at very high ones that for so long a time we failed to suspect their similarity.

If a sensitive pyrometer is inserted into an ordinary solution or a molten alloy which is being gradually cooled, it indicates the first instant that solidification (freezing) begins as well as the termination of the freezing period. Unlike the freezing of water or a pure metal, complete solidification ordinarily does not take place at a definite single temperature but over a greater or lesser range of temperature. By taking such upper and lower freezing-point measurements for many different percentage compositions of a binary (two metal) alloy, for instance, curves can be plotted which show accurately the habits of any and all of the possible combinations (i.e., alloys) of those two metals.

Such are called “freezing-point” curves, and, as we shall see later on, a study of them will give us much valuable and interesting information.

Such curves have been constructed for a great many alloys since the discovery of the analogy between their behavior and that of aqueous solutions led us to study alloys after the manner which physical chemists found so satisfactory for the study of ordinary solutions. Since the study of binary or two metal alloys is often very difficult, it can be readily understood why the determination and interpretation of the curves of alloys which contain three, four or more metals is a very much more serious matter. Much of it has to be done by methods which are long and tedious, such as quenching and microscopic study of innumerable specimens taken during the freezing and subsequent cooling of various alloys of each series. The value of the results depends upon the skill, devotion and clear sightedness of those who carry out the work.

The Freezing-Point Curves of the Iron-Carbon Alloys

The “freezing-point” and “decomposition” curves of the iron-carbon series of alloys have been brought to their present stage of development after something like twenty years of labor by investigators in many lands. If we look at the diagram on page [345], we note at once that the curves are quite complicated. Even yet they are not complete for all percentage combinations of iron and carbon, and those who have given the most time and study to the subject have not yet been able to interpret with entire satisfaction to all concerned all of the discoveries so far made.

Without endeavoring to take up in detail the technique of the manner of their production, which would be unprofitable for us without a great deal more of preliminary study than we have time and space to give, we will at once examine the freezing curves of the iron-carbon alloys as now developed. The works named on page [354] as references for Chapters XXII and XXIII may be consulted for the various types and methods of construction and for explanation of freezing curves by those who desire to study them.

Referring to the freezing-point diagram on page [336], the upper or broad V-shaped line, ABC, indicates the temperatures at which the alloys of various percentages of iron and carbon begin to freeze, and the lower one, AED, the temperatures at which the freezing of these alloys ends. From the diagram it is readily seen that pure iron (100%), has a very high freezing-point and solidifies at once. Iron which contains about 2% of carbon begins to freeze at a much lower temperature and has a long period of solidification, while iron with 4.3% of carbon has the lowest freezing-point of the series with an extremely short solidification period or range.

Since we have been unable to go sufficiently into the methods and technique of freezing-curve construction to be able to understand their general classification, we must accept the statement that the curve of the iron-carbon series is really a double one. The part of it that lies to the left of the dividing line UV of the diagram on page [336], is of the type exhibited by liquids which freeze from “liquid solutions” into what are known as “solid solutions,” which by aid of the microscope are found to be homogeneous mixtures of crystals. On the other hand, alloys which lie to the right of UV, are of the type which form “eutectics.” This will be described later. This dividing line UV, which occurs at about 1.7% of carbon, divides the iron-carbon alloys into these two natural divisions. It was the basis for calling those having 2% of carbon or less, “steels,” and those with over this amount, “cast irons.”

Molten iron is so greedy for carbon, that, when it can get it, it readily holds in solution from 7% to 10% of this element. But solid (frozen) iron cannot retain anything like this amount. As we learned in the last chapter, gamma iron is the only variety which can exist above our lines of loss of conductivity, magnetism and recalescence, i. e., Ar3, Ar3·2, etc. It is, too, the only variety of solid iron which is able to retain carbon in solution, and it can retain only about 1.7% of it.

So when molten steel containing 1.5% of carbon, say, cools until it reaches the temperature represented by the line, AB, which, at its intersection with the 1.5% carbon line would be at about 2582° F., particles or crystals begin to freeze out and float in the molten alloy. As the temperature falls, more crystals separate until, when the temperature determined by intersection of the 1.5% carbon line with the lower freezing curve, AE, is reached, the last of the now mushy alloy solidifies.

Alloys of all other compositions below 1.7% of carbon do just this way except that the temperatures at which freezing begins and ends are different and distinctive for each composition.[[10]] Upon freezing, every one of them retains in “solid solution” in the “gamma” iron whatever carbon it had in the liquid or molten solution. But, as stated above, it can not be over the 1.7% limit.

[10]. Temperatures of beginning and end of freezing may always be ascertained by locating on the freezing-point diagram the points at which the vertical line representing the desired composition intersects and crosses the lines of the freezing-point curves—in these cases, AB and AE.

Of the iron-carbon alloys of compositions lying to the right of the line UV, we find the case to be different, for each one of them has more than the 1.7% of carbon which is the maximum amount which “gamma” iron can retain. Now the lowest temperature at which any iron-carbon alloy can exist without freezing is slightly above 2066° F., and there is but one composition—95.7% of iron and 4.3% of carbon—which can survive until this low temperature is reached. A content of 4.3% of carbon then, is the greatest and also the least concentration which Nature will allow to remain molten down to this minimum temperature. This 4.3% carbon composition which is the lowest melting, i.e., the easiest melted alloy, is called the “eutectic” alloy from Greek words which mean “well melting.” This eutectic composition may be said to divide or rather subdivide this group of alloys into two groups, those containing between 1.7% and 4.3% of carbon, and those which have 4.3% and over.

As stated before, freezing is not an instantaneous but a progressive process. During the freezing period of any of these alloys which have over 1.7% of carbon the still liquid portion which remains after freezing begins to become smaller and smaller in quantity as freezing progresses just as it did in alloys of the “solid solution” group. And as Nature allows a concentration of 4.3% of carbon as the highest concentration at the minimum temperature the very last of the remaining liquid of every alloy eventually gets to this eutectic composition just before the alloy freezes. Those to the left of the eutectic or exact 4.3% composition do so by the gradual freezing out of iron containing the maximum or 1.7% of carbon, i.e., iron is taken out faster than carbon, hence there is gradual concentration of carbon in the remaining liquid. This goes on until 4.3% is reached. The compositions to the right of the line WX throw out the chemical compound, Fe3C, which contains 6.6% of carbon, whereby carbon is eliminated faster than iron and the desired 4.3% carbon alloy is arrived at from the other direction.

To illustrate, take, say, the composition represented by the vertical line at 3% carbon and 97% iron. As the molten alloy cools it reaches the temperature 2330° F., at which temperature the vertical line representing the 3% carbon composition cuts the line AB. Here the alloy begins to freeze by the separation of small crystals of solidifying iron containing definite amounts of carbon.[[11]] But as the carbon thus taken along by the freezing crystals of iron is always less than 1.7%, a proportionally greater amount of iron than carbon is removed from the unfrozen part of the alloy and the remaining liquid or unfrozen part, therefore, is left with slightly more than the 3% of carbon with which it started.

[11]. The percentages of carbon carried by the particles of iron freezing at any particular temperature of the solidification range may be determined from the diagram but the works named in the reference list should be consulted for method and explanation.

This we must now consider another alloy with a lower freezing-point, the reason being, of course, its higher carbon content. At the next lower temperature, more iron containing carbon is frozen out and the remaining liquid is again left a little higher in carbon than before. In this way the continually diminishing amount of remaining liquid keeps concentrating, forming thereby a continuous succession of alloys of higher and higher carbon content as the temperature continuously drops.

Eventually, of course, the concentration of this remaining liquor becomes 4.3% of carbon just before completion of the freezing at 2066° F.

The “Eutectic,” the Part of the Alloy Which Solidifies Last
(Magnification 700 Diameters)

Now with alloys containing more than 4.3% of carbon, almost the opposite occurs. Let us choose the one having 5% carbon and 95% of iron. This molten alloy cools until at 2215° F., small crystals begin to freeze and form in the molten mass. But, as the liquid already has more than the favored 4.3% of carbon, it is not free iron which freezes out, but instead, the chemical compound, Fe3C, which contains 6.6% of carbon. This, of course, takes out carbon proportionally faster than iron, hence, at each very slightly lower temperature, the liquid which remains unfrozen contains just a little less of carbon than did its predecessor. So the constantly decreasing amount of remaining liquid progresses through a succession of compositions each containing just a little less of carbon than the previous one, and eventually, just before freezing we get back to the mixture which contains 4.3% of carbon. Of course there is left unfrozen by this time only a very small amount of the alloy and it is this which has the composition stated.