Size effect on deformation of aluminum and duralumin micro-pillars

Abstract

It is by now well-known that micron-sized metallic crystals exhibit a smaller-being-stronger size effect: the yield strength σvaries with specimen size D approximately as a power law σ~D^(-m)_, and the exponent m has been found to vary within a range of ~0.3 to ~1.0 for different metals. However, little is known about why such a power law comes into play, and what determines the actual value of the exponent m involved. In this study, the power-law scaling of size effect on strength in micro-crystals is explained in terms of the Taylor-type resistance in the dislocation network distribution in the specimen. Theoretical analysis shows that the power-law dependence of yield strength of metallic micro-specimens is derived from a fractal geometry of the initial dislocation network, with m = 3/(q + n) where q is the fractal dimension and n the stress exponent of dislocation velocity. Moderate departures of the initial dislocation structure from an exact fractal geometry may also yield approximate power-law dependence of strength on size.

The plastic deformation of micro-pillars is also known to be affected by whether dislocations can escape easily from the material volume, and the extent to which they mutually interact during the deformation. In the present work, pre-straining and coating are used to modify the initial dislocation content and the constraints on the escape of dislocations. Aluminum micro-pillars with or without thin coating by tungsten deposition and pre-straining, were compressed using a flat-punch nanoindenter to study their plasticity behavior. The results reveal very different behavior between specimens in the size regime of a few microns and that about one micron, suggesting that the dominant hardening mechanisms are different.

As mentioned above, pure and pristine metal micro-specimens have been found to exhibit very strong size dependence of strength, but alloyed counterparts with a much refined microstructural length scale due to the precipitates present are unknown in this aspect. Here, compression tests on duralumin (aluminum 2025 alloy) micro-pillars reveal a much weaker size dependence of strength compared to pure Al, indicating the predominance of the internal length scale in determining strength. Moreover, two-dimensional dislocation dynamics simulations are used to study precipitate strengthening effects in duralumin micro-pillars. The results show that a refined microstructure may resist and slow down the movement of dislocations inside the confined volume, leading to hardening and weak size dependence of strength.

In addition to the compression behavior, the size dependence of the creep behavior of duralumin micro-pillars is also investigated at room temperature. The effects of an internal grain boundary are also investigated. The results reveal that peak-aged duralumin pillars show increasingly significant creep with increasing pillar size, with a typical creep rate of ~〖10〗^(-4) S^(-1) which is drastically larger than that of bulk at room temperature. The bi-crystalline pillars creep even faster than the single crystalline counterparts. TEM examination of the deformed microstructures reveals that the creep rate depends on the residual dislocation density, indicating that dislocations are the agents for creep. Theoretical modeling suggests that the steadystate creep rate is proportional to the lifetime of mobile dislocations, which rises with specimen size in the microns range due to the fact that the dislocations are not easily pinned in this range, therefore they spend longer time in viscous motion across the specimen, leading to a higher strain rate according to the Orowan equation.