Scholarlink

<p>Increasing attention is being paid to high-entropy alloys (HEAs) owing to their remarkable mechanical characteristics. This study forecasts phases and crystal structures of HEAs by employing machine learning (ML) methods: gradient boosting (G-Boosting), decision tree (DT), eXtreme Gradient Boosting (XG-Boosting), and random forest (RF). Phases and crystal structures of HEAs are predicted with good accuracy by XG-Boosting and RF. The reliability of the ML model is confirmed through experimental fabrication of Al<sub>0.2</sub>CuFeMnNi HEA using a mechanical alloying and selective laser melting process (SLM). The XRD patterns show that finer crystallite size and greater lattice strain cause peak widening and intensity to decline. A face-centered cubic (FCC) peak of ((111), (200), (220), (311)) is clearly exhibited in the selective laser-melted (SLMed) Al<sub>0.2</sub>CuFeNiMn HEA. As a result, there is a significant level of agreement between the experimental and XG-Boosting and RF model predictions. Additionally, microstructural characterization of the cross-sectional views of the SLMed specimen shows, the top section consists of fine equiaxed grains with a size of 1.28&#xa0;µm. In the center, columnar grains appear with a size of 5.84&#xa0;µm, and additional columnar structures can be seen in the lower portion. The EDX spectrum reveals a homogeneous distribution of elements.</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

Machine Learning-Assisted Development of Face-Centered Cubic High-Entropy Alloy Fabricated via Selective Laser Melting Technique

  • 期刊论文

  • 发表日期：

    2025年12月2日

查看全文
  • Hareharen K,
  • Panneerselvam T

摘要

Increasing attention is being paid to high-entropy alloys (HEAs) owing to their remarkable mechanical characteristics. This study forecasts phases and crystal structures of HEAs by employing machine learning (ML) methods: gradient boosting (G-Boosting), decision tree (DT), eXtreme Gradient Boosting (XG-Boosting), and random forest (RF). Phases and crystal structures of HEAs are predicted with good accuracy by XG-Boosting and RF. The reliability of the ML model is confirmed through experimental fabrication of Al0.2CuFeMnNi HEA using a mechanical alloying and selective laser melting process (SLM). The XRD patterns show that finer crystallite size and greater lattice strain cause peak widening and intensity to decline. A face-centered cubic (FCC) peak of ((111), (200), (220), (311)) is clearly exhibited in the selective laser-melted (SLMed) Al0.2CuFeNiMn HEA. As a result, there is a significant level of agreement between the experimental and XG-Boosting and RF model predictions. Additionally, microstructural characterization of the cross-sectional views of the SLMed specimen shows, the top section consists of fine equiaxed grains with a size of 1.28 µm. In the center, columnar grains appear with a size of 5.84 µm, and additional columnar structures can be seen in the lower portion. The EDX spectrum reveals a homogeneous distribution of elements.