The direct measurement of the strength of metals on a sub-micrometre scale

Abstract

Four types of experiment have been carried out to investigate the strength properties of annealed metals when the stressed volume is small enough to lie between the existing dis­locations in a crystal. These are (i) indentation experiments of a soft metal surface with a hard stylus, (ii) blunting of a soft metal tip against a hard surface, (iii) compression of individual metal crystals, and (iv) bending of thin filaments. The experiments were performed in either a scanning electron microscope or a transmission electron microscope with the use of micro­-loading devices capable of applying loads down to 0.1 mgf (10 -6 N). In the blunting experi­ments carried out in the transmission electron microscope it was possible to observe disloca­tions directly in the tip during loading. The majority of the experiments were carried out on gold. The results showed that strengths similar to the theoretical value can be achieved but in the experiments in which the stress was applied at an external surface (experiments (i), (ii) and (iii) above) the strength was strongly dependent on the condition of the interface. The strength that could be sustained by a region of perfect crystal in contact with a hard metallic surface was about five times less than the theoretical strength. These relatively low strengths are probably due to interfacial tractions producing very high local stresses. The introduction of some polymeric or amorphous layer at the interface raised the strength to the theoretical level. It is suggested that this is due to the fact that the polymeric layer removes most of the stress concentration. Transmission electron microscopy through aluminium tips during blunting showed that plastic deformation could take place at quite low stresses in a dislocation-free crystal with no dislocations remaining in the crystal during the early stages of blunting. Dislocation build up only occurred in the later stages of deformation. The yield stress was found to decrease with plastic strain in all the experiments, and could fall to values which were not substantially greater than those observed in large specimens. Comparison of the compressive strength of two spherical gold crystals, 0.5 μ m and 2 mm in diameter respectively, showed that the small crystal was only twice as strong as the large crystal after they had both undergone equal amounts of compression. The maximum strengths observed for gold ( μ /20) are greater than those calculated by Kelly (1966) ( ca . μ /50) and are nearer the value deduced by Brown & Woolhouse (1970) for the generation of dislocations around precipitate particles in alloys. The low strengths observed on deformed crystals are considered in the context of dislocation generation in small volumes and it is concluded that although a source mechanism of the Frank–Read type may be able to operate on this scale, some other source mechanism may also exist.