UC Santa Barbara

Intermetallics have garnered significant attention for their potential as low-density, high-temperature load-bearing materials. Unfortunately, the fact that intermetallics are highly ordered arrangements of multiple elements, means there are very few crystallographic directions along which lattice sites are densely spaced. Consequently, plastic deformation of intermetallics often involves very few active slip systems, ultimately leading to the failure of polycrystalline samples before significant plastic flow. A prime example of such behavior is the \textit{bcc-derived} $B2$ intermetallic. While the closest-packed \{110\} planes of the bcc lattice contain the densely spaced collection of $\frac{1}{2}\langle 111 \rangle $ lattice points, the additional atomic ordering present within the $B2$ intermetallic results in the $\langle 001 \rangle $ direction being the shortest distance between lattice points within the closest-packed $\{110\}$ planes. Therefore most $B2$ intermetallics deform plastically \textit{via} the preferential activation of the $\langle 001 \rangle\{110\}$ slip systems, and polycrystalline samples of these $B2$ intermetallics display very little tensile elongation before failure.

Interestingly, certain $B2$ intermetallics undergo plastic deformation \textit{via} the preferential activation of $\langle 111 \rangle \{110\}$ dislocations, prompting speculation that ductile $B2$ intermetallics can be engineered by activating both the $\langle 111 \rangle \{110\}$ and $\langle 001 \rangle \{110\}$ slip systems. While certainly an enticing idea, it remains unclear how significant the $\langle 111 \rangle \{110\}$ slip systems are in determining the macroscopic plastic response of $B2$ intermetallics. To this end, the work presented in this dissertation aims to develop computational methods, informed by first principles, to quantitatively assess the core structures and slip strengths of dislocations across the family of $B2$ intermetallics. This dissertation begins with the development of a reliable platform for efficiently predicting fault energies, and their resulting fault widths, as a function of elemental substitution on a particular lattice site of the intermetallic. A phase-field approach for describing dislocation dynamics is then extended to the $B2$ intermetallic to predict, based on first principles calculations, both the core structure and dynamics of $\langle 111 \rangle \{110\}$ and $\langle 001 \rangle \{110\}$ dislocations. The formulation of this model is validated against molecular statics calculations performed on screw dislocations within the $B2$ intermetallic, NiAl. The results of the phase-field model confirm the $\langle 111 \rangle \{110\}$ screw dislocation core of the $B2$ intermetallic is indeed non-planar, and requires the inclusion of multiple slip planes at the phase-field level to accurately capture the dislocation core. While the $\langle 001 \rangle \{110\}$ screw dislocation could also, in theory, be non-planar, it is found that this dislocation extends diffusely within a single slip plane.

Finally, the extended phase-field approach is applied to predict the dislocation core structure and critical shear stress of planar $\langle 111 \rangle \{110\}$ and $\langle 001 \rangle \{110\}$ edge dislocations across a suite of $B2$ intermetallics: NiAl, CuZn, TiFe, HfRu, YAg, and YCu. Ultimately, two paths that may lead to reasonable ductility within the family of $B2$ intermetallics are proposed: (i) the $B2$ intermetallic possesses generalized stacking fault energies that are comparable in magnitude to those observed in fcc and bcc metals, ultimately leading to $\langle 001 \rangle \{110\}$ dislocations easily gliding through the crystal, or (ii) the $B2$ intermetallic possesses near-zero generalized stacking fault energies within the neighborhood of the antiphase boundary atomic configuration, leading to nearly bcc symmetry along the $\langle 111 \rangle$ $\gamma$-line, and ultimately deformation \textit{via} the motion of $\langle 111 \rangle \{110\}$ dislocations.