The fatigue and ultimate tensile strengths of metals between 4-2 and 293° K By R. D. McCammon* and H. M. Rosenberg The Clarendon Laboratory, University of Oxford The fatigue of copper, silver, gold, aluminium, magnesium, zinc and iron has been investi gated at 4*2, 20, 90 and 293° K. Except for zinc and iron, which exhibit brittle fracture at low temperatures, the fatigue characteristics improve very considerably as the temperature is reduced. The ultimate tensile strength of all the metals was also taken at each temperature and there was shown to be a marked correlation between the increase in the tensile strength at low temperatures and the increase in the fatigue strength. The results are discussed with reference to current ideas on the mechanism of fatigue.
204 R. D. McCammon and H. M. Rosenberg (Discussion Meeting) The vibration generator consists of a moving coil in a fixed permanent magn< and hence if direct current is passed through it a steady pull or push can exerted on the specimen depending on the polarity. In this way static tens tests could be made on the specimens at each temperature and, in particular, t variation of the ultimate tensile strength with temperature could be determine This was done for all specimens. For these measurements specimens withunifoi cross-section were used.
206 R. D. McCammon and H. M. Rosenberg (Discussion Meeting) The low-temperature curves for zinc and iron have not been investigated detail (figures 90 and 91) because these metals show a decrease in the fatigi strength at low temperatures. For zinc at 20° K there was only a very narrd range of stress in which fatigue failure could be produced. Above this stress tl specimens broke immediately the load was applied, and below it they appeared j last indefinitely. The behaviour of zinc and iron was shown to be connected wii the fact that both these metals under a tensile test give a brittle fracture at lo temperatures. Since it appeared that these metals would warrant a separa investigation no details are presented.
The fatigue of metals between 4*2 and 293° K 207 o break the specimen in 105 cycles, but since we are only concerned with ratios nd the fatigue curves are parallel to one another, except for alnmininm and mag- lesium, the actual lifetime chosen is not important.) In figure 94 the ratio of the iltimate tensile strength to the fatigue strength is plotted as function of temperature, temperature (° K) Figure 92 Figure 92. The ultimate tensile strength of copper, silver, gold and aluminium between 4-2 and 293° K. Smoothed curves are shown, taken from points at 4-2, 20, 90 and : 293° K. For aluminium readings were also made at 140 and 195° K.
The fatigue of metals between 4-2 and 293° K 209 3e mechanisms might not operate at higher temperatures, but it appears that y are not essential to the phenomenon. It seems most likely therefore that ro-cracks are produced by the interaction of dislocations; perhaps the vacancies duced in one region can, with little or no movement, agglomerate into a small 3k. Another purely geometrical mechanism is suggested in the companion paper Cottrell & Hull, which follows.
210 R. D. McCammon and H. M. Rosenberg (Discussion Meeting) in copper and silver as an explanation. These workers measured the change in f electrical resistivity after deuteron irradiation at 10° K. They then heated f specimens up to see at what temperature the extra resistivity annealed out. Ti showed that a considerable amount of the resistivity recovered at about 40° B| much lower temperature than was expected. The actual type of damage whi could anneal out at such a low temperature is still not known for certain. 3 however, during fatigue we produce similar defects to those produced dud deuteron irradiation these will anneal out above about 40° K but they will present below that temperature and hence they might be an extra hindrance to f motion of dislocations. Thus the fatigue strength would increase more rapic than it otherwise would and hence the drop in the ratio below 90° K could explained on this basis, although we have no definite evidence that the suggest] is correct. Some support, however, does come from electrical-resistivity measu ments which have been taken after the specimen has been fatigued at 20° K. 1 have found that 1 or 2 % of the extra resistivity introduced by fatiguing does cover below 90° K.