UC Berkeley

Dislocations are line defects present in crystalline materials that govern the plastic deformationof metals and metallic alloys. These defects are described by continuum approaches in the far field, but the center of a dislocation core requires atomistic detail. The computational efficiency of continuum approaches allows for the use of mesoscale models probing length and time scales beyond those available in atomistic simulations. However, many of the assumptions involved which sacrifice finer details associated with dislocation core reconfiguration have been left unchecked. This dissertation presents the methodology and results of atomistic simulations applied to three main dislocation processes relevant for mesoscale modeling of FCC metals and alloys. The first is the process of dislocation climb, where dislocations migrate through the absorption/emission of point defects. Mesoscale models typically assume a dislocation to be an ideal cylindrical sink. This assumption is only valid when there is a sufficient density of jogs along the dislocation line, and through the combination of atomistic calculations of jog free energies and analytical theory, it is shown that obtaining such a sufficient density of jogs requires high homologous temperatures and/or a high supersaturation of defects. The second application is on the process of dislocation cross-slip in the presence of short-range ordering (SRO). Mesoscale models typically do not include the effect of solutes on cross-slip, and if they do, assume them to be distributed in a random fashion instead of having SRO. Through the atomistic calculation of many dislocation cross-slip energy barriers with and without the presence of SRO, it is shown that the effect of SRO on planar defect energies can significantly increase the cross-slip energy barrier, potentially reducing cross-slip rates by orders of magnitude and altering work-hardening processes during deformation. The third and final application is on the process of core restructuring of Lomer/Lomer-Cottrell dislocations as a function of stress and alloy composition. Mesoscale models assume that these dislocations, which are central to the work-hardening of FCC metals and alloys, evolve under stress through only one mechanism. It is shown through atomistic simulations that a variety of evolution mechanisms, including twin nucleation, can occur through the core restructuring of these dislocations depending on the local stress and alloy composition.