No abstract

In the traditional picture of plasticity in bcc metals, edge dislocations have been assumed to play a minor role due to their high mobility with respect to screw dislocations, which then control plastic flow. 1 2 111 {110} edge dislocations indeed fit this description, as it has been shown by way of numerous atomistic simulations. However, 1 2 111 {112} edge dislocations have been comparatively much less studied. The recent discovery of a possible regime where they move slowly via thermally activated kink-pair nucleation may have implications in the plastic behavior of bcc materials. Because dislocation mobilities are very difficult to measure experimentally, in this paper, we provide comprehensive mobility laws for both types of edge dislocations as a function of temperature and stress using molecular dynamics simulations. Our results confirm the existence of clearly delimited thermally activated and phonon drag dynamic regimes for 1 2 111 {112} edge dislocations and of a single viscous drag regime for their 1 2 111 {110} counterparts. We also provide an analysis to relate the difference in mobility to the dislocation core properties. Our fitted mobility laws may be used in dislocation dynamics simulations of plastic flow involving millions of segments.

The low-temperature plastic yield of α-Fe single crystals is known to display a strong temperature dependence and to be controlled by the thermally activated motion of screw dislocations. In this paper, we present molecular dynamics simulations of 1 2 111 {112} screw dislocation motion as a function of temperature and stress in order to extract mobility relations that describe the general dynamic behavior of screw dislocations in pure α-Fe. We find two dynamic regimes in the stress-velocity space governed by different mechanisms of motion. Consistent with experimental evidence, at low stresses and temperatures, the dislocations move by thermally activated nucleation and propagation of kink pairs. Then, at a critical stress, a temperature-dependent transition to a viscous linear regime is observed. Critical output from the simulations, such as threshold stresses and the stress dependence of the kink activation energy, are compared to experimental data and other atomistic works with generally very good agreement. Contrary to some experimental interpretations, we find that glide on {112} planes is only apparent, as slip always occurs by elementary kink-pair nucleation/propagation events on {110} planes. Additionally, a dislocation core transformation from compact to dissociated has been identified above room temperature, although its impact on the general mobility is seen to be limited. This and other observations expose the limitations of inferring or presuming dynamic behavior on the basis of only static calculations. We discuss the relevance and applicability of our results and provide a closed-form functional mobility law suitable for mesoscale computational techniques.

No abstract

Screw dislocations in bcc metals display non-planar cores at zero temperature which result in high lattice friction and thermally-activated strain rate behavior. In bcc W, electronic structure molecular statics calculations reveal a compact, non-degenerate core with an associated Peierls stress between 1.7 and 2.8 GPa. However, a full picture of the dynamic behavior of dislocations can only be gained by using more efficient atomistic simulations based on semiempirical interatomic potentials. In this paper we assess the suitability of five different potentials in terms of static properties relevant to screw dislocations in pure W. Moreover, we perform molecular dynamics simulations of stress-assisted glide using all five potentials to study the dynamic behavior of screw dislocations under shear stress. Dislocations are seen to display thermally-activated motion in most of the applied stress range, with a gradual transition to a viscous damping regime at high stresses. We find that one potential predicts a core transformation from compact to dissociated at finite temperature that affects the energetics of kink-pair production and impacts the mechanism of motion. We conclude that a modified embedded-atom potential achieves the best compromise in terms of static and dynamic screw dislocation properties, although at an expense of about ten-fold compared to central potentials.