Scholarlink

<p>To achieve the precise optimization of nickel (Ni) and silicon (Si) contents in ductile iron and elucidate the underlying strengthening and toughening mechanisms, this study investigated the correlation between Ni/Si composition and mechanical properties using a combined approach of machine learning and first-principles calculations. A composition-property prediction model was constructed utilizing ensemble learning algorithms, yielding coefficients of determination (<InlineEquation ID="IEq1"> <EquationSource Format="TEX">\(R^{2}\)</EquationSource> <EquationSource Format="MATHML"><math> <msup> <mi>R</mi> <mn>2</mn> </msup> </math></EquationSource> </InlineEquation>) of 0.8386 and 0.8305 for ultimate tensile strength (UTS) and elongation (EL), respectively. First-principles calculations revealed that single doping with Ni enhances plasticity, whereas single doping with Si improves strength and hardness. Furthermore, Ni–Si co-doping promotes the hybridization among Fe-<i>d</i>, Ni-<i>d</i>, and Si-<i>p</i> orbitals near the Fermi level and induces synergistic charge transfer. This interaction strengthens atomic bonding while maintaining favorable plasticity, thereby uncovering the macroscopic strengthening and toughening mechanisms at the electronic level. Ultimately, these findings provide a theoretical framework and robust data support for the compositional design of ductile iron.</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

Study on Composition Optimization of Ni and Si in Ductile Iron and Its Strengthening-Toughening Mechanism Based on Machine Learning and First-Principles

  • 期刊论文

  • 发表日期：

    2026年6月2日

查看全文
  • Qing Li,
  • He Wei,
  • Xinyu Zeng,
  • Xiangming Li,
  • Zhenhua Li

摘要

To achieve the precise optimization of nickel (Ni) and silicon (Si) contents in ductile iron and elucidate the underlying strengthening and toughening mechanisms, this study investigated the correlation between Ni/Si composition and mechanical properties using a combined approach of machine learning and first-principles calculations. A composition-property prediction model was constructed utilizing ensemble learning algorithms, yielding coefficients of determination ( \(R^{2}\) R 2 ) of 0.8386 and 0.8305 for ultimate tensile strength (UTS) and elongation (EL), respectively. First-principles calculations revealed that single doping with Ni enhances plasticity, whereas single doping with Si improves strength and hardness. Furthermore, Ni–Si co-doping promotes the hybridization among Fe-d, Ni-d, and Si-p orbitals near the Fermi level and induces synergistic charge transfer. This interaction strengthens atomic bonding while maintaining favorable plasticity, thereby uncovering the macroscopic strengthening and toughening mechanisms at the electronic level. Ultimately, these findings provide a theoretical framework and robust data support for the compositional design of ductile iron.