Synergistic properties of deep-cryogenic-treated high-entropy alloys and basalt fiber-reinforced ultra-high performance cementitious composites

Abstract

Ultra-high performance concrete (UHPC) faces persistent challenges in achieving synergistic strength and toughness under extreme loads. This study proposes a novel approach: partially substituting quartz sand with deep-cryogenic-treated (DCT) FeCoNiCrCu high-entropy alloys (HEAs) particles and incorporating basalt fibers (BF) for a synergistic effect. The key variables are DCT time, HEAs substitution rate, HEAs particle size, and BF volume fraction. Experiments demonstrate that 48 h DCT induces compressive residual stress in HEA without phase transformation, optimizing the strength-ductility balance. With 20 vol% HEAs substitution (median particle size d₅₀ = 77.59 μm) and 2 vol% BF, the composite achieves compressive and flexural strengths of 211.02 MPa and 34.47 MPa, respectively. Multi-scale characterization reveals that DCT-activated HEAs particles significantly densify the interfacial transition zone. This densification increases ITZ nano-hardness by 28.55 % via enhanced mechanical interlocking and load transfer. This study integrates machine learning with the grey wolf optimizer (GWO) to develop a GWO-optimized XGBoost model, markedly improving prediction accuracy for compressive and flexural strengths. Feature importance analysis identifies DCT time and HEA substitution rate as the dominant factors influencing HEAs reinforced UHPC (HEA-UHPC) mechanical properties. This experimental-machine learning framework offers a promising pathway for designing next-generation protective UHPC.