According to recent studies, the retrogression phenomena seems to be caused by the re-solution of the cluster or G.P. aggregate due to a sudden increase of its solubility in aluminium above 205°C. If this is true, only the precipitation process is expected to occur without showing remarkable hardening caused by the formation of pseudo-precipitate or cluster, when Al-Cu alloy is aged above the temperature. This fact was confirmed in this report and only slight hardening effect could be observed when the alloy (3.9%Cu) was aged at 240∼270°, after being quenched from 520°C. Then the softening was observed where actual precipitation of the second phase was detected both by microscopic examination and X-ray diffraction. Therefore, the precipitation hardening process, as the author calls it alone can be observed at these temperatures. The degree of aging (
p) was calculated from the hardening curves at the respective temperatures between 240 and 270° and log
p⁄(1−
p) vs. log
t (
t: time) was plotted. The relation was proved to be linear by this method as well as by calculating
p from the data of G. Borelius and L. Ström
(4) in their study on evolution of heat during aging. The inclination of the line (
n) was 2.6 by both methods, which will again support the author’s idea on the formula of age-hardening. The relation between the type of aging process and the temperature of aging was discussed
(5): At lower temperatures (
T<
T1), neither of the two types, pseudo-precip. and precip., can occur within the time of our observation. At the temperature above
T1, there is a range (
T1<
T<
T2), where only pseudo-precip. type can occur within the time of our observation. Then the range (
T2<
T<
T3) comes where pseudo-precip. type occurs first, followed by precip. type afterwards. And at last, above
T3 (
T<
T3), only precip. type can be seen. In Al-Cu alloy, for example,
T1≈−30°,
T2≈80° and
T3≈205°. In some other alloys, these relations are not so clearly observable. The results of the two studies by other workers
(6)(8) which were published after this report was read in April, 1950, were considered at the end of this report. These studies support the idea of the author described here.
View full abstract