机器学习辅助设计穿甲用TiZr系高熵合金的可解释模型

Machine Learning-Assisted Design of Interpretable Models for TiZr-Based High-Entropy Alloys Used in Armor-Piercing Applications

摘要:

现代军事技术对穿甲战斗部材料毁伤效能的需求持续升级，推动着高性能合金体系的创新研究。在此背景下，兼具高硬度、抗压强度及高温相稳定性的TiZr系高熵合金，因其在穿甲战斗部中的潜在应用价值受到广泛关注。本研究引入机器学习方法辅助开展高熵合金成分设计，旨在探索成分与性能之间的复杂关系，提升合金设计效率。考虑到穿甲作用对合金材料硬度的高要求，搭建了包含组元摩尔分数和5项关键描述符在内的15维特征数据集，集成157组合金硬度数据，采用八种典型机器学习模型（随机森林、K近邻、支持向量机等）进行训练，经网格搜索超参数优化及交叉验证，筛选出精度最优的预测模型，并利用SHAP可解释性框架（SHAP，Shapley Additive Explanations）揭示特征贡献机制。结果表明，XGBoost模型（XGBoost，Extreme Gradient Boosting）预测质量最高，决定系数R2达到0.73，平均绝对百分比误差为14.0%，混合焓（ΔH_mix）、Nb元素含量、原子尺寸失配（δ）是影响合金硬度的最主要因素，硬度调控需协同热力学稳定性、电子结构与几何尺寸三要素，特征间补偿效应是性能优化的关键，该方法为穿甲用TiZr系高熵合金提供了成分设计新范式，验证了机器学习在高效毁伤领域的巨大工程价值，有助于加速新型高性能材料的开发进程。

Abstract:

Modern military technology demands continuous improvement in the damage efficiency of armor-piercing projectile core materials, driving innovation in high-performance alloy systems. TiZr-based refractory high-entropy alloys (RHEAs, Refractory High-Entropy Alloys), known for their high hardness, compressive strength, and thermal phase stability, have garnered attention due to their potential application in armor-piercing warheads. This study introduces a machine learning (ML)-assisted approach to alloy design, aiming to uncover the complex relationship between composition and performance while improving design efficiency. To address the critical requirement for hardness in armor-piercing applications, a 15-dimensional feature dataset was constructed, incorporating component molar fractions and five key descriptors, based on 157 experimental hardness data points. Eight ML models—including random forest, K-nearest neighbors, and support vector machines—were trained, and XGBoost was identified as the most accurate through hyperparameter tuning via grid search and cross-validation. The SHAP (SHAP，Shapley Additive Explanations) framework was applied to interpret feature contributions. Results indicate that the XGBoost model achieves the highest predictive performance (R2 = 0.73, and the average absolute percentage error is 14.0%). The most influential factors affecting alloy hardness are mixing enthalpy (ΔH_mix), niobium (Nb) content, and atomic size mismatch (δ). Effective hardness control necessitates the synergistic regulation of thermodynamic stability, electronic structure, and geometric dimensions, where inter-feature compensation plays a critical role in optimizing overall performance. This work establishes a novel compositional design paradigm for TiZr-based RHEAs, demonstrating the engineering value of ML in enhancing material development for high-efficiency damage applications.