Dislocation-mediated short-range order evolution during thermomechanical processing

TL;DR

This work addresses how thermomechanical processing modifies chemical short-range order (SRO) in a chemically complex TiTaVW alloy by tying dislocation-mediated atomic rearrangements to processing parameters. Using large-scale atomistic simulations with a machine-learning interatomic potential, the authors quantify SRO evolution through competing creation and annihilation rates, $\Gamma$ and $\lambda$, across temperature and strain rate, revealing two distinct processing regimes and a far-from-equilibrium steady-state SRO not accessible by thermal annealing. They provide an empirical, yet physically grounded, model for extrapolating steady-state SRO across processing spaces and map TMP-induced SRO to equivalent equilibrium temperatures $T_{eq}$, enabling targeted design of SRO states. The results establish a mechanistic link between dislocation structure and chemical ordering, offering a framework to predict and tailor SRO in chemically complex alloys during TMP, with practical implications for processing optimization and property control.

Abstract

Thermomechanical processing alters the microstructure of metallic alloys through coupled plastic deformation and thermal exposure, with dislocation motion driving plasticity and microstructural evolution. Our previous work (Islam et al., 2025) showed that the same dislocation motion both creates and destroys chemical short-range order (SRO), driving alloys into far-from-equilibrium SRO states. However, the connection between this dislocation-mediated SRO evolution and processing parameters remains largely unexplored. Here, we perform large-scale atomistic simulations of thermomechanical processing of equiatomic TiTaVW to determine how temperature and strain rate control SRO via competing creation ($Γ$) and annihilation ($λ$) rates. The simulations employ systems containing 2.4 million atoms and utilize a machine learning interatomic potential optimized to capture chemical complexity through the motif-based sampling technique. Using information-theoretic metrics, we quantify that the magnitude and chemical character of SRO vary systematically with processing parameters. We identify two regimes: a low-temperature regime with weak strain-rate sensitivity, and a high-temperature regime in which reduced dislocation density and increased screw character amplify chemical bias and accelerate SRO formation. The resulting steady-state SRO is far-from-equilibrium and cannot be produced by equilibrium thermal annealing. Together, these results provide a mechanistic and predictive link between processing parameters, dislocation physics, and SRO evolution in chemically complex alloys.

Figures (5)

• Figure 1: Thermomechanical processing of equiatomic TiTaVW alloy.a) Uniaxial compression of TiTaVW at $300\,\text{K}$ under a constant strain rate of $10^{9}\,\text{s}^{-1}$. After each unit strain, the deformation axis is rotated to align with the direction of largest dimension. Dislocation networks are shown as green lines. b) Stress-strain curve shows the onset of steady-state flow stress. c) Dislocation density also reaches steady-state with continued deformation. d) Evolution of SRO ($D_{\text{sro}}$) with strain, eventually reaching a steady-state SRO ($D^\infty_{\text{sro}}$). The dashed orange curve shows the fit to eq. \ref{['eq:sro']}, capturing the observed $D_{\text{sro}}$ evolution. During fitting, $D^{\text{o}}_{\mathrm{sro}}$ was fixed at $D_{\mathrm{sro}}(\epsilon=0)$.
• Figure 2: Effect of temperature and strain rate on steady-state SRO.a) Steady-state SRO ($D^\infty_{\text{sro}}$) as a function of temperature and strain rate. $D^\infty_{\text{sro}}$ values were obtained by averaging $D_{\text{sro}}$ from the seventh compression cycle. b) The steady-state SRO states are of far-from-equilibrium kindislam_nonequilibrium_2025 and they move further from equilibrium as temperature increases, indicated by increasing $D_{\text{eff}}$ (sec. \ref{['meth:4']}) c) Warren-Cowley parameters for different pairs indicate that there is a change in pair preference at moderate temperatures for certain elemental pairs (indicated by the purple box), marking a change in the chemical motifs forming the SRO.
• Figure 3: Effect of temperature and strain rate on SRO evolution rate. Temperature and strain rate dependence of a) SRO creation rate $\Gamma$, and b) SRO annihilation rate $\lambda$. The shaded purple region highlights the temperature range where both $\Gamma$ and $\lambda$ become sensitive to changes in temperature and strain rate. Error bars (one standard deviation) are obtained by propagating the covariance from the nonlinear fits of $\epsilon$–$D_{\mathrm{sro}}$ data to eq. 5. The error bars are
• Figure 4: Dislocation-mediated mechanisms of SRO evolution.a) Dislocation configuration at $300\,\text{K}$ and b)$2300\,\text{K}$ after reaching steady state during thermomechanical processing at a strain rate of $10^8\,\text{s}^{-1}$. Grey features indicate dislocation-induced defects originating from cross-kinks (approximately 595 at 300 K and 361 at 2,300 K.) c) Steady-state dislocation density $\rho^\infty$ decreases with temperature and strain rate. $\rho^\infty$ values were obtained by averaging $\rho$ from the seventh compression cycle. d) Probability distribution of dislocation character at $\dot{\epsilon} = 10^{8}\,\text{s}^{-1}$, where $\theta$ is the angle between the Burgers vector and line direction $\theta$. e) Ratio of edge-to-screw character $R$ (eq. \ref{['eq:edge']}) as a function of temperature and strain rate.
• Figure 5: Steady-state SRO during manufacturing processes.a) The variation of steady-state SRO $D^\infty_{\text{sro}}$ induced by thermomechanical processing as a function of temperature and strain rate. The dashed lines show the empirical model (eq. \ref{['eq:empirical']}) fit to the $T$-$D^\infty_{\text{sro}}$ data for each strain rate. b) Equilibrium SRO states obtained from annealing at different temperatures. The equivalent temperature, $T_{\text{eq}}$, is determined by matching $D^\infty_{\text{sro}}$ obtained from thermomechanical processing with its corresponding equilibrium SRO value. An example match is highlighted by the magenta circled markers; the corresponding $T_{\text{eq}}$ is indicated by the vertical dashed line in the inset. c) Map of $T_{\text{eq}}$ across a wide range of temperatures and strain rates, spanning three relevant manufacturing process categories. The dashed line segments indicate extrapolated values obtained from the empirical model fitted using $D^\infty_{\text{sro}}$ across all temperature and strain rate cases (see Supplementary Information section 5)