Simulating the mechanisms of serrated flow in interstitial alloys with atomic resolution over diffusive timescales

Abstract

The Portevin-Le Chatelier (PLC) effect is a phenomenon by which plastic slip in metallic materials becomes unstable, resulting in jerky flow and the onset of inhomogeneous deformation. The PLC effect is thought to be fundamentally caused by the dynamic interplay between dislocations and solute atoms. However, this interplay is almost always inaccessible experimentally due to the extremely fine length and time scales over which it occurs. In this paper, simulations of jerky flow in W-O interstitial solid solutions reveal three dynamic regimes emerging from the simulated strain rate-temperature space: one resembling standard solid solution strengthening, another one mimicking solute cloud formation, and a third one where dislocation/solute coevolution leads to jerky flow as a precursor of dynamic strain aging. The simulations are carried out in a stochastic framework that naturally captures rare events in a rigorous manner, providing atomistic resolution over diffusive time scales using no adjustable parameters.

Screw dislocations in BCC transition metals: from ab initio modeling to yield criterion

“…Peierls barriers under stress were computed by applying a strain tensor to the simulation cell to produce the targeted shear stress through (2) followed by a NEB calculation [21]. Examples for a shear stress resolved in the (101) glide plane and along the [111] Burgers vector of a screw dislocation in tungsten are presented in Figure 5(a). We note that to reduce the computed cost, we performed the NEB calculations only on the first half of the paths.…”

“…Similar independence was observed in other BCC metals [21,22], Figure 6. Trajectory of a 1/2 [111] screw dislocation between two neighboring easy core positions in BCC tungsten, with corresponding Peierls potential. 2 The dislocation position was deduced from the ab initio stress variation along the minimum energy path (Equations ( 2) and ( 3)).…”

“…As already noted, the modified Schmid law correctly predicts that the twinning region with χ < 0 is easier to shear than the antitwinning region, in qualitative agreement with the experimental data. But experiments also show that the yield stress does not depend only on χ: the T/AT asymmetry is very strong for a tensile axis between [001] and [011] (filled symbols in Figure 7) whereas the yield stress is almost constant between [001] and [111] (open symbols in Figure 7). The modified Schmid law given by ( 9) cannot account for such variations between different orientations corresponding to the same MRSSP.…”

“…For the elastic field far from the dislocation and its coupling with an applied stress, it is actually sufficient to consider the relaxation volume tensorΩ = S 0ε * defined per unit length of dislocation. When the dislocation is in its ground state, the 3-fold symmetry around the [111] axis of the easy core imposes thatΩ is diagonal with only two independent components: Ω = diag(Ω 11 , Ω 11 , Ω 33 ). Values of this tensor can be directly obtained from ab initio calculations because, due to the relaxation volume, the dislocation energy varies when a stress tensorσ is applied, with a corresponding interaction energy E inter = −Ω ij σ ij .…”

“…The corresponding minimum yield stress is presented in the second row of Figure 13. With Schmid's law, only one primary slip system exists in each stereographic triangle and the yield stress follows directly the distribution of Schmid factors, with easier glide in the regions near the [001] − [011] edge and the χ = 0 line, and a maximum yield stress near the [111] the stereographic projection where the [111](101) system is activated (orange region in Figure 13) is shifted towards χ < 0 in tension, and towards χ > 0 in compression. This shift of the primary slip system is also responsible for the emergence of neighboring primary systems close to the edges of the stereographic triangle.…”