Scholarlink

<p>Eutectic high-entropy alloys (EHEAs) show great potential as high-temperature structural materials; however, their oxidation resistance has not been sufficiently investigated. Conventional trial-and-error methods pose significant challenges for designing oxidation-resistant EHEAs. Notably, B2 and Laves phases are recognized for their excellent high-temperature oxidation resistance and readily form eutectic structures with the FCC phase. In this work, a machine learning (ML) framework was employed to screen for EHEAs with target FCC + B2 and FCC + Laves eutectic structures. Consequently, two representative EHEAs, AlCoCrFeNi<sub>2.1</sub> (EH-Ni2.1) and CoCrFeNiNb<sub>0.5</sub> (EH-Nb0.5), were designed and fabricated. Their isothermal oxidation behaviors at 1000&#xa0;°C for up to 105 hours were systematically examined. Both alloys exhibited oxidation kinetics that followed a parabolic rate law. EH-Ni2.1 demonstrated superior oxidation resistance, with a specific mass gain of 1.82&#xa0;mg/cm<sup>2</sup> and a parabolic rate constant (<i>K</i><sub><i>p</i></sub>) of 0.03 mg<sup>2</sup>·cm<sup>−4</sup>·h<sup>−1</sup>. These values were significantly lower than those of EH-Nb0.5 (3.23&#xa0;mg/cm<sup>2</sup> and 0.1 mg<sup>2</sup>·cm<sup>−4</sup>·h<sup>−1</sup>, respectively). The oxide layer on EH-Nb0.5 consisted of an outer Cr<sub>2</sub>O<sub>3</sub> layer and an inner layer containing CrNbO<sub>4</sub>, whereas EH-Ni2.1 formed a more protective oxide scale primarily composed of Al<sub>2</sub>O<sub>3</sub> and Cr<sub>2</sub>O<sub>3</sub>. This study demonstrates the effectiveness of an ML-guided approach in developing oxidation-resistant EHEAs.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

High-Temperature Oxidation Behaviors of AlCoCrFeNi2.1 and CoCrFeNiNb0.5 Eutectic High-Entropy Alloys: A Machine Learning-Assisted Design and Mechanism Study

  • 期刊论文

  • 发表日期：

    2026年7月6日

查看全文
  • He Zhang,
  • Hanqing Xu,
  • Yuqing Xie,
  • Mengdi Zhang,
  • Rui Li,
  • Gong Li

摘要

Eutectic high-entropy alloys (EHEAs) show great potential as high-temperature structural materials; however, their oxidation resistance has not been sufficiently investigated. Conventional trial-and-error methods pose significant challenges for designing oxidation-resistant EHEAs. Notably, B2 and Laves phases are recognized for their excellent high-temperature oxidation resistance and readily form eutectic structures with the FCC phase. In this work, a machine learning (ML) framework was employed to screen for EHEAs with target FCC + B2 and FCC + Laves eutectic structures. Consequently, two representative EHEAs, AlCoCrFeNi2.1 (EH-Ni2.1) and CoCrFeNiNb0.5 (EH-Nb0.5), were designed and fabricated. Their isothermal oxidation behaviors at 1000 °C for up to 105 hours were systematically examined. Both alloys exhibited oxidation kinetics that followed a parabolic rate law. EH-Ni2.1 demonstrated superior oxidation resistance, with a specific mass gain of 1.82 mg/cm2 and a parabolic rate constant (Kp) of 0.03 mg2·cm−4·h−1. These values were significantly lower than those of EH-Nb0.5 (3.23 mg/cm2 and 0.1 mg2·cm−4·h−1, respectively). The oxide layer on EH-Nb0.5 consisted of an outer Cr2O3 layer and an inner layer containing CrNbO4, whereas EH-Ni2.1 formed a more protective oxide scale primarily composed of Al2O3 and Cr2O3. This study demonstrates the effectiveness of an ML-guided approach in developing oxidation-resistant EHEAs.