ss 



NATURE 



[July 14, 1923 



The Production of Single Metallic Crystals and some of their Properties. 



By Prof. H. (". II. ( \ki'ENTer, F.R.S. 



METALS and alloys are composed of aggregates ol 

 crystals. These do not, as a rule, possess plane 

 faces, that is, the external forms of crystals. They 



Fig. I. — Showing diminishing size of crystals with increase of strain in 

 parallel portion of the testpieces. By permission of the Institute of Metals. 



are joined together at boundaries which have been 

 produced by the meeting of a number of crystals 

 growing simultaneously, and are usually irregular in 



NO. 2802, VOL. 112] 



outline. It is generally assumed that on cooling, 

 crystallisation starts in the liquid metal or alloy from 

 a number of centres, and proceeds with a velocity and in 

 a manner characteristic of the metal and the con- 

 ditions under which it has been cooled. The result- 

 ing boundaries may approximate to plane surfaces, 

 but are more usually curved and irregular. These 

 crystals are called " allotriomorphic " to distin- 

 guish them from " idiomorphic " crystals, which 

 do possess plane faces and are characteristic of 

 most mineral substances and artificially prepared 

 salts. Moreover, they are usually ver\' small 

 and cannot be distinguished without the aid of 

 a microscope. It is true, that in the case of 

 large castings weighing many tons, crj'stals of 

 several cubic inches capacity have been formed 

 and afterwards isolated. The great majority of 

 metals and alloys, however, which have been cast 

 and hot-worked, have from 150 to 300 cr^'stals 

 to the linear inch, corresponding to from 3,375,000 

 to 27,000,000 crystals in a cubic inch. Frequently 

 the size is even smaller, especially in the case of 

 steels. The crystals are still more minute in 

 severely drawn wires, and from figures given 

 by Sykes it appears that in molybdenum wire 

 there may be as many as 5000 to the linear 

 inch. 



The properties of metals and alloys are the pro- 

 perties of these aggregates of minute crj^stals. 

 Sauveur was the first investigator to show, about 

 eleven years ago, that by carefully straining and 

 afterwards heating metals, much larger crj'stals 

 could be produced, and he suggested that there 

 was a critical stress which produced the largest 

 crystals. Later, other investigators, notably 

 Riider, Chappell, Jeffries, and Hanson, showed 

 that if a metal was locally deformed and then 

 heated, exceptionally large crystals were formed 

 at some distance from the point at which the strain 

 is most severe. About two years ago Miss Elam 

 and the writer succeeded in converting the whole 

 of the parallel portion of aluminium testpieces, 

 whether in the form of sheets or bars, into a single 

 crystal, which indeed extended for some distance up 

 into the curved shoulders of the testpieces, forming 

 an irregular boundary line. The cr\-stals varied in 

 volume from 0-5 to 2-0 cubic inches, and it has 

 been possible to compare some of their properties 

 with those of the aggregates of small cr}-stals, of 

 which this metal is usually composed. Experi- 

 ments have also been carried out with iron, 

 copper, silver, and tin, but with less success, 

 although in all cases it has been possible to grow 

 crystals very much larger than those contained in 

 the original metal. 



In the production of large crystals of alu- 

 minium the adjustment between mechanical strain 

 and the temperature of heating is extremely impor- 

 tant. This point is clearly brought out in Fig. i, 

 which shows how the crystal size may be varied in alu- 

 minium according to the degree of strain. The eight 

 test-pieces shown, after a preliminar}' heat treatment to 



the 



