Skip to main content
Back to timeline
arXivSource publication:

Refractory high-entropy alloys move past equimolar: a dislocation-genome plus dynamics framework predicts high-temperature strength and room-temperature ductility across the NbMoTaW space

Synopsis

High-temperature strength and room-temperature ductility in refractory high-entropy alloys have long traded off against each other, and most studies stay near equimolar compositions. This multiscale framework builds energy landscapes for edge-dislocation glide from atomic configurations with a dislocation-genome method, feeds the resulting local configurational forces into stochastic discrete dislocation dynamics, and obtains a linear scaling law between critical shear stress and force fluctuation. That law predicts high-temperature strength across the NbMoTaW composition space, which is then screened jointly with a ductility criterion based on the relative mobility of screw and edge dislocations.

Source-provided article image: Data-driven exploration of the strength-ductility synergy in refractory high-entropy alloys
Fig. 1 · arXiv · Page 7

Interpretation

Composition design for refractory high-entropy alloys has centered on near-equimolar compositions, leaving much of the composition space unexplored while high-temperature strength and room-temperature ductility pull against each other. Existing data-driven work either relies directly on experimental data and machine learning, limited by uncalibrated experimental bias and weak physical interpretability, or builds property databases from ab initio simulations, which makes reliable extrapolation from microscopic properties to macroscopic performance difficult; multi-objective trade-offs still largely rely on empirical adjustment. The configurational-force fluctuation was computed for about 11,000 alloys in the NbMoTaW system; the equimolar alloy ranks relatively high on this indicator but is not optimal, and increasing Nb and W concentrations is predicted to yield strength beyond the equimolar alloy.

The method treats the local atomic arrangement near the slip plane of an edge dislocation as the basic unit encoding the resistance landscape, and a neural network predicts solute-dislocation interaction energies directly from these configurations. Pinning criteria based on local generalized stacking fault energy in face-centered cubic alloys rely on dislocations dissociating into partials separated by stacking fault ribbons, whereas body-centered cubic edge dislocations have compact cores without dissociation, so those criteria do not apply. Training data were generated with an embedding model plus NEB calculations, totaling 84,127 samples: 27,336 for NbW alloys and 56,791 for equimolar NbMoTaW; on 1,000 validation samples of the non-equimolar composition Nb25Mo25Ta10W40, absent from training, the network still predicted accurately.

Adding the local configurational force, obtained by differentiating the energy landscape with respect to dislocation position, as an extra force term in discrete dislocation dynamics together with Langevin thermal fluctuations yields critical shear stresses at different temperatures. Direct NEB simulation of long dislocation segments demands very large computational resources and struggles to cover mesoscale lengths and microsecond times. Simulated critical shear stresses agree with experimental data across several body-centered cubic alloys; the 0 K critical stress varies linearly with configurational-force fluctuation, and the limit of zero fluctuation and near-zero resistance in pure body-centered cubic metals falls on the same relation.

The composition- and temperature-dependent solid-solution strengthening model derived from the scaling law predicts high-temperature yield strength across the NbMoTaW composition space and is screened jointly with a ductility criterion based on the relative mobility of screw and edge dislocations. Earlier handling of the strength-ductility trade-off relied mainly on empirical adjustment and rarely searched systematically beyond equimolar compositions. Predicted high-temperature yield strengths for TaW0.10, NbMo0.05, NbW0.056, NbTa0.28W0.10 and TaW0.04 all fall within the 95% prediction interval; the multi-objective prediction indicates that non-equimolar compositions with a large portion of Mo and W and small additions of Nb and Ta may combine high-temperature strength with room-temperature ductility.

Perspective

The model addresses the resistance to edge-dislocation motion in single-phase body-centered cubic random solid solutions, predicting high-temperature yield strength and a ductility indicator based on screw-to-edge dislocation relative mobility within the NbMoTaW composition space; the authors note that short-range order, secondary phases and grain-boundary effects in real materials must be added separately.

Critical stresses are extracted at a threshold strain rate of 1 s-1 to approximate quasi-static loading at experimental strain rates around 10-3 s-1; the fitted T0 can exceed alloy melting points and carries only theoretical meaning; near the equiatomic region, configurational-force fluctuation deviates markedly from the size-misfit parameter and from the atomic mean square shear stress of the microelasticity model, indicating that conventional misfit parameters cannot fully capture mixing effects on dislocations.