IMRAM

Institute of Multidisciplinary Research for Advanced Materials, Tohoku University

東北大学
多元物質科学研究所

LAST UPDATE 2023/08/04

  • 研究者氏名
    Researcher Name

    石上啓介　Keisuke ISHIGAMI
    助教　Assistant professor

  • 所属
    Professional Affiliation

    東北大学多元物質科学研究所
    Institute of Multidisciplinary Research for Advanced Materials, Tohoku University

    無機材料研究部門　ナノスケール磁気機能研究分野
    Division of Inorganic Material Research, Nanoscale Magnetism

  • 研究キーワード
    Research Keywords

    ハード磁性材料
    永久磁石
    ベイズ統計
    機械学習
    Hard Ferromagnetic Material
    Permanent Magnet
    Bayesian statistics
    Machine Learning
研究テーマ
Research Subject
ベイズ統計と機械学習を活用した永久磁石材料の外場応答の理解
Understanding the external field response of permanent magnet materials using Bayesian statistics and machine learning

研究の背景 Background

炭素循環型・省エネルギー社会の実現に不可欠な永久磁石材料の磁気特性向上因子の特定や特性発現機構の理解は、産業的・社会的に重要な意義を持つ。永久磁石材料の一つであるNd-Fe-B焼結磁石の磁気特性は重希土類元素の添加や熱処理による微細組織の制御により向上する。このことは開発当初から知られていたものの、その機構についてはまだ十分に理解されていない。

It is of industrial and social importance to identify the key factor and to clarify the mechanism for improving the magnetic properties of Nd-Fe-B sintered magnets, which are widely used as one of the permanent magnet materials for achieving a carbon-neutral and low-energy society. Since the beginning of its development, it has been known that the magnetic properties of Nd-Fe-B sintered magnets can be improved by doping heavy rare earth elements and controlling the microstructure by post-sinter annealing, but the mechanism is not well understood.

研究の目標 Outcome

本研究では、実験科学(放射光X線回折・分光計測)、計算科学(物理モデルシミュレーション)、データ科学(ベイズ統計・機械学習)を連携活用することで、永久磁石材料の磁気特性向上の潜在因子を特定し特性発現機構を理解する。本研究により、新規永久磁石材料の磁気特性を効率的・合理的に設計可能になることが期待される。

In this study, to identify the latent factor contributing to the enhancement of magnetic properties of permanent magnetic materials and to understand the mechanism, we will utilize experimental science (synchrotron radiation X-ray diffraction and spectroscopy), computational science (physics-based simulation), and data science (Bayesian statistics and machine learning). This study will contribute to the effective and rational design of magnetic properties of new permanent magnetic materials.

研究図Research Figure

Fig.1. The purpose of this study is to identify the phase transformation associated with coercivity enhancement during post-sinter annealing of Ga-containing Nd-Fe-B sintered magnets by supervised learning based on sparse modeling of the coercivity with the in-situ X-ray diffraction pattern and to rank the importance of the identified phase transformations [1].

Fig.2. Sparse modeling of the coercivity by the X-ray diffraction pattern shows that the generation of the Nd6Fe13Ga phase and the annihilation of the dhcp-Nd phase correlate positively with the coercivity enhancement, while the presence of NdO correlates negatively. The former has been pointed out in previous studies, while the latter is newly revealed in this study.

文献 / Publications

[1] K. Ishigami et al., “Understanding the Coercivity of Ga-containing Nd–Fe–B Sintered Magnets from Feature Extraction and Selection of X-ray Diffraction Patterns via Dimension Reduction and Sparse Modeling”, INTERMAG 2023, Accepted for publication.

研究者HP