Scholarlink

<p>The present study introduces a predictive framework for the intelligent design and optimization of high-entropy alloy (HEA) and bioinspired hybrid coatings applied to carbon-nanotube-reinforced metal-matrix composites (CNT-MMCs). The research aims to address the limitations of conventional experimental coating design by developing a data-driven model capable of predicting mechanical properties with high reliability. The dataset used in this study was compiled from experimental records and open-access HEA repositories, including Kaggle, and Mendeley, containing data on composition, microstructure, processing, and performance attributes. The predictive model was developed and executed in Python using the XGBoost algorithm, employing hyperparameter tuning, and early stopping to enhance generalization. The proposed framework demonstrated predictive accuracy with R<sup>2</sup> values of 0.965 for hardness, 0.958 for yield strength, and 0.947 for elastic modulus, indicating its robustness and precision. The SHAP analysis revealed that CNT concentration, grain size, and sintering pressure are the most influential parameters controlling coating performance. Furthermore, a multi-objective optimization strategy was implemented to identify optimal HEA compositions, such as Al<sub>0.3</sub>Co<sub>0.2</sub>Cr<sub>0.2</sub>Fe<sub>0.2</sub>Ni<sub>0.1</sub>Ti<sub>0.1</sub>, exhibiting superior hardness (≈ 174 HV), yield strength (≈ 402.7&#xa0;MPa), and wear resistance (≈ 0.0046&#xa0;mm³/Nm) than MWCNTs, and MWCNTs with nano SiC particles. The optimized AI-guided configuration was translated into a bioinspired hierarchical coating design, integrating an HEA outer layer, a graded transition layer, and a chitosan–polydopamine adhesion interface. Overall, the proposed framework significantly enhances the predictive efficiency, material selection accuracy, and structural performance of CNT-MMC coatings, establishing a scalable and intelligent idea for next-generation aerospace, automotive, and energy engineering applications.</p> Graphical Abstract <p></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

Explainable Multi-objective Design of High-Entropy Alloy Coatings for CNT-Reinforced Composites Enhancing Tribological Performance and Interfacial Durability

  • 期刊论文

  • 发表日期：

    2026年6月24日

查看全文
  • Mukesh Chaudhari,
  • Natrayan Lakshmaiya

摘要

The present study introduces a predictive framework for the intelligent design and optimization of high-entropy alloy (HEA) and bioinspired hybrid coatings applied to carbon-nanotube-reinforced metal-matrix composites (CNT-MMCs). The research aims to address the limitations of conventional experimental coating design by developing a data-driven model capable of predicting mechanical properties with high reliability. The dataset used in this study was compiled from experimental records and open-access HEA repositories, including Kaggle, and Mendeley, containing data on composition, microstructure, processing, and performance attributes. The predictive model was developed and executed in Python using the XGBoost algorithm, employing hyperparameter tuning, and early stopping to enhance generalization. The proposed framework demonstrated predictive accuracy with R2 values of 0.965 for hardness, 0.958 for yield strength, and 0.947 for elastic modulus, indicating its robustness and precision. The SHAP analysis revealed that CNT concentration, grain size, and sintering pressure are the most influential parameters controlling coating performance. Furthermore, a multi-objective optimization strategy was implemented to identify optimal HEA compositions, such as Al0.3Co0.2Cr0.2Fe0.2Ni0.1Ti0.1, exhibiting superior hardness (≈ 174 HV), yield strength (≈ 402.7 MPa), and wear resistance (≈ 0.0046 mm³/Nm) than MWCNTs, and MWCNTs with nano SiC particles. The optimized AI-guided configuration was translated into a bioinspired hierarchical coating design, integrating an HEA outer layer, a graded transition layer, and a chitosan–polydopamine adhesion interface. Overall, the proposed framework significantly enhances the predictive efficiency, material selection accuracy, and structural performance of CNT-MMC coatings, establishing a scalable and intelligent idea for next-generation aerospace, automotive, and energy engineering applications.

Graphical Abstract