Under common processing conditions, both dilute and complex concentrated alloys are often realized as random alloys, with no correlation in the occupancy of lattice sites by the constituent atom types. The current thesis primary addresses two problems in random alloys, namely (1) the application of concurrent multiscale modeling and (2) solute-strengthening.

For accurate functioning of a forced-based atomistic-continuum (a/c) couple, it is important that the underlying material description of the two domains are the same near the a/c interface. A/c coupling of random alloys violates this criterion since the atomistic domain is inherently inhomogeneous with local fluctuations in atomic configuration and elastic stiffness, while the continuum is defined with the average elastic constants of the alloy. The resulting
coupling errors are of the order of 100 MPa in long-range spurious stresses and spurious stress fluctuations near the interface. Two methods of constructing the coupled problem are proposed in this thesis to mitigate coupling errors. In Method 1, the pad atoms are relaxed with respect to the atoms in the atomistics domain in the initial construction of the a/c couple, which eliminates spurious stresses in absence of external loading. Negligible spurious stresses (â 5 MPa) arise on loading the coupled problem. Method 2 replaces the random pad atoms with average atoms, thus explicitly eliminating fluctuations in the pad region which causes coupling errors. This methods also yield coupling errors of â 5 MPa, which are negligible relative to stresses arising in realistic mechanical problems of interest.

The latter part of the thesis deals with solute-strengthening in random alloys. In random alloys, strengthening is caused by the pinning of dislocation segments in favourable solute environments. The length scales of dislocation undulation and the energy barrier for unpinning local segments are controlled by solute-dislocation interactions (SDIs) and solute-solute interactions (SSIs). The SDIs are modeled as the interaction of solute misfit volume with dislocation pressure field, under anisotropic elasticity assumption. Comparison of the anisotropic theory with isotropic theory shows that the Voigt-averaged elastic constants best reproduce the anisotropic predictions. SSIs are accounted for in the strengthening theory and it is found to cause significant additional strengthening (â 70% at 10% Al) in dilute Ni(Al) alloys with strong Al-Al repulsion, but only a negligible increase in strength (â 2%) in bcc MoNbTaW complex alloy with larger solute misfit volumes and lower SSIs. Finally, the thesis also studies the role of short-range order (SRO) on alloy strengthening and predicts the average strengthening due to SRO, in terms of solute pair interaction energies, the well-known Warren-Cowley SRO parameters and alloy composition. The theory accounting for SRO is in progress, however, the theory in its current state lays the groundwork for realizing the most comprehensive strengthening theory for alloys of any compositional complexity.