Abstract
When strained, with or without subsequent heating, many metals produce twinned markings. It is customary to call the former mechanical twins and the latter annealing twins. According to Dana “ twin crystals are those in which one or more parts regularly arranged are in reverse position with reference to the other part or parts.” This must be held to apply not merely to the external symmetry of crystals which possess plane faces, but also to the internal symmetry of such substances as metals and alloys. which do not possess plane faces. Metallic crystals may result in one of several ways. They are formed, e. g., by solidification from the liquid state, a process which is known as primary crystallisation, and results in the formation of crystallites, or they are produced by rearrangement of the atoms in primary crystals giving rise to what may be termed secondary crystals. This term includes such phenomena as crystal growth, recrystallisation, polymorphic transformation, etc. Primary crystals which have solidified in such a way that they are unstrained do not contain twin crystals. The phenomenon is confined to the various types of secondary crystals just mentioned, and, therefore, appears to be related to what may be called atomic rearrangement. The present paper describes observations made on annealing twins, mainly those of copper and its alloys. It has arisen as a result of what the authors have found and described in another paper, viz., that it has proved impossible to prepare large copper crystals by the method of critical straining and annealing without at the same time producing twin crystals.
Experiments
. The copper used was commercially pure “ high conductivity ” metal in the form of cold rolled strip. The distorted crystals contained numerous twins which had been produced by repeated annealings and rollings during manufacture. When the strip was heated in an atmosphere of nitrogen at a temperature of 800° C. for six hours, complete recrystallisation occurred with the production of equiaxed crystals. Subsequently the strip was deformed to a very slight extent and again heated at 800° C. for six hours, after which the small equiaxed crystals were completely converted by crystal growth into a few large crystals. Representative microstructures of the strip as received, after crystallisation and crystal growth, are shown in photographs Nos. 1, 2 and 3, at a magnification of 80 diameters. It will be noticed that the crystal size increases markedly under the above treatments, and, in fact, No. 3 represents only a portion of one copper crystal which, however, contains twin crystals oriented in three directions. It will be noticed that all three photographs show the presence of twins. They may either appear light on a dark crystal or dark on a light crystal; they may extend right across a given crystal or over only a part of it. Further, they vary very considerably in thickness. These photographs show that the twinned crystals in copper which are present in the originally cold worked metal are neither removed by recrystallisation nor by subsequent crystal growth, although during the former process all the old crystals are replaced by new ones, and during the latter a large number of crystals are absorbed into a few large ones. Photograph No. 4 represents an area, part of which has undergone recrystallisation (small crystals), and the remainder crystal growth in addition. The multiple twinning shown in the large crystals is very striking.