[en] Cross-slip is a fundamental process of screw dislocation motion and plays an important role in the evolution of work hardening and dislocation structuring in metals. Cross-slip has been widely studied in pure FCC metals but rarely in FCC solid solutions. Here, the cross-slip transition path in solid solutions is calculated using atomistic methods for three representative systems of Ni-Al, Cu-Ni and Al-Mg over a range of solute concentrations. Studies using both true random alloys and their corresponding average-alloy counterparts allow for the independent assessment of the roles of (i) fluctuations in the spatial solute distribution in the true random alloy randomness and (ii) average alloy properties such as stacking fault energy. The results show that the solute fluctuations dominate the activation energy barrier, i.e. there are large sample-to-sample variations around the average activation barrier. The variations in activation barrier correlate linearly with the energy difference between the initial and final states. The distribution of this energy difference can be computed analytically in terms of the solute/dislocation interaction energies. Thus, the distribution of cross-slip activation energies can be accurately determined from a parameter-free analytic model. The implications of the statistical distribution of activation energies on the rate of cross-slip in real alloys are then identified.

[en] The atomistic mechanisms of deformation in multicomponent random alloys are challenging to model because of their extensive structural and compositional disorder. For embedded-atom-method interatomic potentials, a formal averaging procedure can generate an average-atom EAM potential and this average-atom potential has recently been shown to accurately predict many zero-temperature properties of the true random alloy. Here, the finite-temperature thermodynamic properties of the average-atom potential are investigated to determine if the average-atom potential can represent the true random alloy Helmholtz free energy as well as important finite-temperature properties. Using a thermodynamic integration approach, the average-atom system is found to have an entropy difference of at most 0.05 k _B/atom relative to the true random alloy over a wide temperature range, as demonstrated on FeNiCr and Ni₈₅Al₁₅ model alloys. Lattice constants, and thus thermal expansion, and elastic constants are also well-predicted (within a few percent) by the average-atom potential over a wide temperature range. The largest differences between the average atom and true random alloy are found in the zero temperature properties, which reflect the role of local structural disorder in the true random alloy. Thus, the average-atom potential is a valuable strategy for modeling alloys at finite temperatures. (paper)

[en] The yield strengths of High Entropy Alloys have recently been correlated with measured or computed picometer-scale atomic distortions. Here, the root mean square microdistortion in a multicomponent alloy is shown to be nearly proportional to the misfit-volume parameter that enters into a predictive model of solute strengthening. Analysis of two model ternary alloy families, face-centered cubic Cr-Fe-Ni and body-centered cubic Nb-Mo-V, demonstrates the correlation over a wide composition space. The reported correlation of yield strength with microdistortion is thus a consequence of the correlation between microdistortion and misfit parameter and the derived dependence of yield strength on the misfit parameter.