The hardening effect can also be removed by the process of annealing, that is, by heating to redness and cooling slowly. This process is very generally employed in practice, for relieving in ternal stresses caused by the processes of manufacture.
Experiments by J. Muir (Phil. Trans. R.S., vol. 193, 1900) have shown that temperature has an important influence on the rate of this "recovery of elasticity." In iron and steel, complete recovery can be produced in a few minutes by dipping the over strained specimen into boiling water.
When a piece of iron or steel, of ter being overstrained in ten sion, is subjected to a compressive load, the strain is not found to be proportional to stress unless recovery has been effected by rest or heating. After recovery, the elastic limit for compression is lower than it would be in the normal state but Muir's experi ments show that the reduction is less than the amount by which the elastic limit for tension has been raised. That is to say, the general effect of strain-hardening followed by recovery is to widen the total range of stress within which stress and strain are pro portional.
loop increases with the range of stress. This area may be in terpreted as measuring the work done in performing the cycle of operations: if the strain had been wholly elastic, the area of the loop would have been nil, because the work done in stretching would have been recovered during the process of unloading, but in the plastic range more work is done in stretching than is sub sequently recovered. The differ ence, of course, is absorbed by the specimen or transformed into heat.
Elongation and Reduction of Area.-57. Useful informa tion in regard to ductility is afforded by the elongation of a specimen tested in tension—that is, the total extension of the specimen at the instant of frac ture. Barba (Mem. Soc. des Ing.
Civ., 188o) has shown that this quantity depends upon the form of the specimen, so that geometrical similarity must be main tained if strictly comparable results are to be obtained from specimens of different sizes. Unfortunately, the practice of dif ferent countries as regards the standardization of test pieces does not yet satisfy this condition.
Another quantity which serves as a convenient measure of ductility is the reduction of area—that is, the decrease in the cross-sectional area of the specimen after fracture, expressed as a percentage of the original area. This quantity is less dependent than elongation on the geometrical form of the specimen but it is difficult to measure accurately. Fracture by Tension.-58. The ultimate stages of a tension test—that is, the circumstances which determine the elongation and reduction of area—are evidently governed by the stress strain relations in plastic material, of which little is understood. When a bar of uniform section is pulled, the extension is at first distributed fairly uniformly over the whole gauge length. But just before the bar breaks, a large additional extension occurs near the place of rupture, and the section is much more con tracted there than in other parts, so that the specimen forms a "waist." Fig. 21 illustrates this point ; it is taken from a photo graph of a broken specimen of mild steel, which had a uniform diameter before the test.