STEEL, HEAT TREATMENT OF. The physical structure of steel may and does vary to a marked degree according to the heat treatment that it has undergone; and that the strength, toughness and general usefulness of a struc tural member of steel may be largely controlled by the way in which it is subjected to the action of heat. Formerly it was assumed that phos phorus gives brittleness, that sulphur makes the metal red-short, and that the other properties are almost entirely dependent upon the per centage of carbon present. With the introduc tion of the microscope and modern methods of metallography in the study of steel and other alloys, it has been found that the structure and the properties of the material may be largely modified without any change whatever in the chemical composition, and as a conse quence some very practical lessons have been learned. When steel of a coarse structure, but not necessarily brittle, is heated to a certain temperature and is then allowed to cool in air, or is quenched in oil or in water, the original structure is destroyed and is replaced by one of a very fine-grained character. Pure iron, when so coarsely crystalline as to resemble cast zinc, may be restored to excellent qualities by simply heating it to a certain temperature (known as the °critical points) • its subsequent structure resembling that which it possessed when it originally left the rolls. The old idea of the process of annealing was to reheat the material to a high temperature, and then to hold it for a time at that temperature, and subsequently to cause it to cool very slowly; but later researches showed that the temperature of reheating must not exceed F., and that the time of cooling has little or no influence upon the result. It is altogether unnecessary to reheat and re forge the specimen, and, in fact, the material is apt to be still further injured by such a proc ess. Since almost any piece of structural steel is liable to be rendered more or less crystalline during the operations through which it is neces sarily passed, it is most desirable that the proc ess of restoration be made a regular portion of the routine of manufacture. To permit this it is only necessary that proper furnaces be de signed to admit the large pieces, and to permit a uniform temperature to be maintained in all parts, reliable pyrometers being of course pro vided so that the correct temperature may be maintained. The operation may then be per formed upon every piece as the final stage in its manufacture. The result is elimination of danger of accident from weakness by crystalli zation. The microscope indicates that heating
at high temperatures causes a great develop ment in the size of the crystalline grains, and that reheating to about 1,600° F. restores the original structure, or yields an even better one. A structural steel, although good in its normal rolled or forged condition, may easily de teriorate by being heated to a temperature a little above that to which steel is most com monly heated, previously to being rolled or forged. Steel that is made brittle by such heat ing, or dangerously brittle by exposure to con siderably higher temperatures, can be com pletely restored to the best possible condition without remelting and without forging down to a smaller size. Practically all of the ex perimental results show, not only that the original good qualities of normally rolled steel can be restored after the material has been made brittle by the exceedingly simple expe dient of heating to about 1,600° F. for a very short time, but also that the steel may even be made better than it was originally.
It was long ago noted that the influence of thermal treatment increases rapidly with the carbon content of steel. In one of Howe's experiments with a muffle the following varia tions were shown: A yet higher carbon with less manganese yielded less tensile strength than the *0.92 car bon and .240 Mn,)) but a higher elastic limit. It will be noted that generally the best result was obtained by cooling from C. These experiments were made in a closed muffle, with both slow heating and slow cooling. The elastic limit reaches a minimum at about 720° C., then rises rapidly on heating to temperatures of 750° to 850 , and then decreases steadily. The slower the cooling the lower elastic limit is to be expected. In the matter of elongation it has been found that with 0.50 per cent of car bon, as the initial temperature (for slow cool ing) is raised, the elongation decreases moder ately up to or 1,300°. With higher initial temperature the elongation decreases rapidly. In the matter of tensile strength it is noted that the increase is more marked when cooling is done in the open air than in the muffle. With rapid cooling hardness and brittleness increase, without much reference to the initial temperature of cooling, so that the steel maker always has to consider his product from several different angles. If the steel is low in carbon it will be little affected by heat treatment, but if high in carbon the heat treat ment is all important.