An expandable closed-loop framework for ductility optimization in refractory high-entropy alloys

This study establishes an integrated computational-experimental framework for designing ductile refractory high-entropy alloys. The approach combines EMTO-CPA calculations, metaheuristic optimization methods, and a KNN surrogate model to explore a wide quinary compositional space and maximize Pugh's ratio as a ductility indicator. The optimization scheme efficiently balanced exploration and convergence, progressively improving predictive accuracy and identifying multiple compositional families that offer distinct combinations of stiffness and ductility rather than a single optimum. Thermodynamic calculations and Scheil simulations predicted microstructures dominated by BCC-based phases, which were experimentally confirmed through XRD and EBSD/EDS analyses revealing dual BCC variants with slightly different lattice parameters and systematic elemental partitioning between dendrite cores and interdendritic region. Compression testing demonstrated high strength across all alloys but a clear ductility hierarchy governed by microstructural heterogeneity, with smoother compositional gradients promoting more uniform deformation. Correlation analyses revealed trend-level positive trend between Pugh's ratio, fracture strain, and toughness, although pronounced microsegregation could suppress ductility even in high-Pugh's-ratio alloys. Overall, the results highlight the need for concurrent optimization of elastic properties and microstructural homogeneity to achieve a favorable strength-ductility balance in RHEAs. The closed-loop framework proposed in this study provides an efficient and generalizable pathway for accelerating the discovery of ductile compositions in complex multicomponent alloy systems.