Intrinsic ductility enhancement in Mg alloys elucidated via large-scale ab-initio calculations

TL;DR

This work investigates the intrinsic ductility of Mg and dilute Mg alloys by directly computing the core-energy difference between pyramidal $\langle\textrm{\bf c}+\textrm{\bf a}\rangle$ dislocations, $\Delta E^{\textrm{I-II}}_{\textrm{Mg}}$, and site-specific dislocation–solite interactions, $U_j$, using large-scale DFT. It combines these first-principles inputs with a line-tension cross-slip model that also accounts for external macroscopic strains and solute strengthening, enabling quantitative predictions of cross-slip barriers $\Delta G_{XS}(c)$ for Mg–Y and Mg–Zn. The key finding is that relative solute strengthening of Pyr I and II dislocations and the strain dependence of core energetics drive cross-slip behavior, rather than purely stabilizing the Pyr I core; this reconciles experimental observations of ductility enhancement and slip-system transitions in Mg–Y and Mg–Zn. The framework provides a principled design route for more ductile Mg alloys and can be embedded in higher-scale crystal-plasticity models to guide alloy development.

Abstract

Magnesium is the lightest structural alloy, yet its practical use is limited by its low ductility. Recent studies suggest ductility enhancement in dilute Mg alloys may stem from favorable solute modification of <c+a> pyramidal I/II screw dislocation core energy difference, activating <c+a> slip via a double cross-slip mechanism. This work conducts large-scale DFT calculations, reaching ~6,000 atoms, of <c+a> dislocation energetics in Mg and Mg-Y/Zn alloys. We find that relative solute strengthening effects on pyramidal I and II screw dislocation glide are crucial for cross-slip enhancement in Mg-Y, in contrast to prior investigations, that find solute-mediated dislocation-core energy modification as the main driver. Our predictions align with single- and poly-crystal experimental results and also capture the transition from pyramidal II to I preferred slip in Mg-Y.

Figures (10)

• Figure 1: Schematic of simulation set up for dislocation core energetics calculations. Identical Volterra displacement boundary conditions corresponding to the perfect $\langle\textrm{\bf c}+\textrm{\bf a}\rangle$ screw dislocation is applied to both the Pyr II and Pyr I dissociated configurations in the boundary layer region of thickness $t$.
• Figure 2: DFT calculations of Pyr I and II $\langle\textrm{\bf c}+\textrm{\bf a}\rangle$ screw dislocations. (A) Differential displacement plots of relaxed dislocation cores showing non-planar nature (plane normal is $\hkl<-2113>$). (B) Cell size convergence of $\Delta E^{\textrm{I-II}}_{\textrm{Mg}}$. (C) Electron density contour of Pyr II $\langle\textrm{\bf c}+\textrm{\bf a}\rangle$ screw dislocation (plane normal is $\hkl<-2113>$).
• Figure 3: Common neighbor analysis of relaxed dislocation core structures. Left side is Pyr II $\langle\textrm{\bf c}+\textrm{\bf a}\rangle$ screw dislocation core structure and right side is Pyr I $\langle\textrm{\bf c}+\textrm{\bf a}\rangle$ screw dislocation core structure. The solute site used for the cell-size convergence study of $\Delta U_j^{\textrm{I-II}}$ in Tab. \ref{['tab:deltaUSensitivity']} is also marked.
• Figure 4: (A) and (B) Pyr I and II $\langle\textrm{\bf c}+\textrm{\bf a}\rangle$ screw dislocation Y and Zn solute interaction energies ($U_j$) at various near-core sites. $\Delta U_j^{\textrm{I-II}}$ values for Y and Zn are also plotted on the Pyr I relaxed core structure. The sites were identified using common neighbor analysis and we additionally included an outer ring of surrounding HCP sites, amounting to a total $N_s=58$ sites. (C) average solute effects on $\Delta E^{\textrm{I-II}}_{\textrm{Mg}}$ in Mg-Y and Mg-Zn, comparison between explicit DFT $\Delta U_j^{\textrm{I-II}}$ inputs and previous MEAM/DFT SF-solute based approach WuCurtinScience2018AhmadCurtin2019 (D) concentration dependent critical resolved shear stress (CRSS) values predicted using a Labusch type weak pinning model leyson2010quantitative and the computed $U_j$ inputs for Y and Zn solutes. The reference CRSS values for FPCS slip (Pyr I) in single crystal Mg-Y are taken from Ref. Rikihisa2017. Pure Mg single crystal CRSS values at room temperature are approximately 40 MPa and 53 MPa for SPCS and FPCS glide respectively ando2010deformation.
• Figure 5: Schematic of the double cross-slip mechanism in dilute Mg-Y alloy depicting the different slip preference regimes (SPCS/FPCS). The first-principles dislocation and dislocation-solute energetics inputs, and first-principles derived solute-strengthening inputs to the cross-slip energy barrier model are indicated.