Molecular simulation-derived features for machine learning predictions of metal glass forming ability-猫眼课题宝

DOI：

10.1016/j.commatsci.2021.110728

发表时间：

2021-08

期刊：

影响因子：

通讯作者：

中科院分区：

其他

文献类型：

作者：

Ben Afflerbach;L. Schultz;J. Perepezko;P. Voyles;I. Szlufarska;D. Morgan

我们根据从分子动力学模拟中获得的新的可计算的特征，建立了金属合金玻璃形成能力的模型。自从金属玻璃被发现以来，人们一直在努力预测新合金的玻璃形成能力(GFA)。已经获得了对GFA的有效评估，但通常依赖于合金特征温度的知识，如玻璃化转变温度、晶化温度和液相线温度，但用途有限，因为这些特征需要合成和表征感兴趣的合金。最近，预测GFA的机器学习方法采用了更容易获得的模型特征，如组成元素的基本属性。然而，这些更易访问的特征通常提供的预测精度低于它们较难访问的对应特征。在这项工作中，我们证明了通过使用从分子动力学模拟获得的输入特征来增加GFA模型的预测值是可能的。这样的特征只需要相对简单和可扩展的模拟，使它们比实验测量更容易获得，成本更低。我们生成了一个分子动力学临界冷却速率的数据库，以及相关的候选特征，这些特征受到了之前对GFA的研究的启发。在9个拟议的GFA功能列表中，我们通过套索模型确定了两个对性能最重要的功能。100℃下的结晶热和类二十面体分数显示出很好的前景，因为它们能够显著提高模型的性能，而且它们可以很容易地使用灵活的从头算量子力学方法，几乎适用于所有系统。这项用于机器学习预测的可计算功能的进步GFA将使未来的模型能够更准确地预测新的玻璃形成合金。

We have developed models of metallic alloy glass forming ability based on newly computationally accessible features obtained from molecular dynamics simulations. Since the discovery of metallic glasses, there have been efforts to predict glass forming ability (GFA) for new alloys. Effective evaluations of GFA have been obtained but generally relied on knowledge of alloy characteristic temperatures like the glass transition, crystallization, and liquidus temperatures but are of limited utility because these features require synthesizing and characterizing the alloy of interest. More recently, machine learning approaches to predict GFA have employed more accessible model features such as the elemental properties of constituent elements. However, these more accessible features generally provide less predictive accuracy than their less accessible counterparts. In this work we showed that it is possible to increase the predictive value of GFA models by using input features obtained from molecular dynamics simulations. Such features require only relatively straightforward and scalable simulations, making them significantly easier and less expensive to obtain than experimental measurements. We generated a database of molecular dynamics critical cooling rates along with associated candidate features that are inspired from previous research on GFA. Out of the list of 9 proposed GFA features, we identify two as being the most important to performance through a LASSO model. Enthalpy of crystallization and icosahedral-like fraction at 100 K showed promise because they enable a significant improvement to model performance and because they are accessible to flexible ab initio quantum mechanical methods readily applicable to almost all systems. This advancement in computationally accessible features for machine learning predictions GFA will enable future models to more accurately predict new glass forming alloys.

Molecular simulation-derived features for machine learning predictions of metal glass forming ability

文献摘要

相似文献