Dislocation pinning by substitutional impurities in an atomic-scale model for Al(Mg) solid solutions

Abstract

We report our atomic-scale computations for the static depinning threshold of dislocations in Al(Mg) solid solutions. The interaction between the dislocations and the isolated obstacles is studied for different types of obstacle, i.e. single solute atoms situated at different positions, and solute dimers with different bond directions. Part of this work is used to apply different standard analytical theories for solid solution hardening, the predictions of which are finally compared with our direct atomic-scale simulations (AS) for dislocation depinning in random Al(Mg) solid solutions. According to our comparisons, the dislocation statistics in our AS is qualitatively well described by the Mott-Nabarro-Labusch theory. In agreement with earlier results about a different system, namely Ni(Al), the depinning thresholds are similar for the edge and for the screw dislocations.

Solute strengthening in random alloys

“…Simulations using EAM potentials show comparable energies and/or strengths for screw and edge dislocations in fcc solid solution alloys [38,81,109]. TEM observations of dislocation microstructures in fcc materials are also fairly isotropic [20], i.e.…”

“…Numerical simulations of solute strengthening has lead to the identification of important trends and understandings: contribution to the strength of solutes in planes away from the glide plane, role of the screw vs. edge dislocations, role of solute pairs when the solute concentration increases and the reference is taken from a pure elemental matrix, etc. [37,38,81,109,110]. However, MD simulations, even using empirical potentials, suffer from important limitations, which make difficult to compare simulation results to (i) analytical models and (ii) real material strength.…”

Robust atomistic calculation of dislocation line tension

“…It is also important to account for ellipticity, which has been neglected in all atomic scale studies of dislocations. Finally, the concept of line tension, as applied to a smooth, continuous bow-out, is inapplicable for very small bow-outs (kink density ρ <∼ 0.05; Figure 4), and results in this domain are not consistent [48].…”

“…The combination of processing and post-processing schemes employed here in computing Γ are far beyond those in any of the earlier attempts [2,[42][43][44][45][46]48,49]. First, spurious effects are caused by finite simulation volumes because the image stresses on a bowing-out dislocation scale linearly with the amount of bow-out [50], leading to error in τ app and thus Γ. Shenoy et al [44] estimated the image stress and corrected their calculated line tension but the final results remain qualitative.…”

“…As a result of this importance, and with the increasing availability of efficient parallelized molecular dynamics algorithms [41], a number of attempts have been made to calculate the line tension directly from atomistic simulations [2,20,[42][43][44][45][46][47][48]. The simplest realization of such a calculation is the periodic small bow-out problem [22] in which an infinite initially-straight dislocation is pinned at a periodic array of obstacles with a spacing L. A driving resolved shear stress τ app is then applied, causing the free regions of the dislocation to bow-out in between the pinning points, as shown in figure 1; this problem is thus closely related to models of strengthening via pinning of dislocations.…”