Data in support of "Design and Selection of High Entropy Alloys for Hardmetal Matrix Applications using a Coupled Machine Learning and CALPHAD Methodology"

Abstract: This study aimed to utilise a combined Machine Learning (ML) and CALculations of PHAse Diagrams (CALPHAD) methodology to design hardmetal matrix phases for metal forming applications that could serve as the basis for carbide reinforcement. The vast compositional space that High Entropy Alloys (HEAs) occupy, offers a promising avenue to satisfy the application design criteria of wear resistance and ductility. To efficiently explore this space, random forest ML models are constructed and trained from publicly available experimental HEA databases to make phase constitution and hardness predictions. Interrogation of the ML models constructed revealed accuracies > 78.7% and mean absolute error of 66.1 HV for phase and hardness predictions. Six promising alloy compositions, extracted from the ML predictions and CALPHAD calculations, were experimentally fabricated and tested. The hardness predictions are found to be systematically under and over predicted depending on the alloy microstructure. In parallel, the phase classification models were found to lack sensitivity towards additional intermetallic phase formation. Despite the discrepancies identified between ML and experimental results, the fabricated compositions showed promise for further experimental evaluation. These discrepancies were believed to be directly associated with the available databases but importantly have highlighted several avenues for both ML and database development.

摘要：本研究旨在结合机器学习（Machine Learning, ML）与相图计算（Calculations of Phase Diagrams, CALPHAD）方法，设计适用于金属成形领域、可作为碳化物增强基底的硬质合金基体相。高熵合金（High Entropy Alloys, HEAs）所涵盖的广阔成分空间，为满足耐磨性与延展性的应用设计要求提供了极具潜力的途径。为高效探索该成分空间，本研究基于公开可用的高熵合金实验数据库，构建并训练了随机森林（random forest）机器学习模型，用于相组成与硬度预测。对所构建模型的分析显示，相组成预测与硬度预测的准确率均高于78.7%，硬度预测的平均绝对误差为66.1 HV。从机器学习预测结果与相图计算结果中筛选出的6种极具潜力的合金成分，已通过实验制备并开展测试。研究发现，硬度预测结果会根据合金微观结构的不同，出现系统性的偏高或偏低。与此同时，相分类模型对额外金属间相的生成缺乏敏感性。尽管机器学习与实验结果之间存在上述偏差，但所制备的合金成分仍具备进一步实验评估的潜力。这些偏差被认为与现有可用数据库直接相关，但重要的是，本研究同时为机器学习技术与数据库开发指明了若干可行的改进方向。

搜集汇总

背景与挑战

背景概述

该数据集支持一项结合机器学习与CALPHAD方法设计硬质合金基体高熵合金的研究，包含实验数据和预测结果，主要用于金属成型应用开发。数据集可公开获取，采用CC BY-NC-ND 4.0许可协议。