| 研究実績の概要 |
SmFe12-based compounds are known as potential candidates for the next generation high-performance permanent magnets. This is due to their superior intrinsic magnetic properties than those of Nd2Fe14B at the elevated temperatures. The main challenge for the industrial application of SmFe12-based magnets is how to transfer their intrinsic magnetic properties to the extrinsic ones, large coercivity and remanence. In this project, we will overcome this problem by employing a unique combinatorial research approach.
In FY2023, we have developed SmFe12-based magnetic thin films as model systems to understand the coercivity limit in an anisotropic SmFe12-based system. We have successfully increased the coercivity of anisotropic boron-doped Sm(Fe0.8Co0.2)12 thin films to 1.8 T, the highest value reported so far, by reduction in the magnetization of the intergranular phase via Al diffusion process. Based on this knowledge, we have made an effort to develop anisotropic bulk SmFe12-based sintered magnets. We have succeeded in increasing the coercivity of Cu-doped Sm(Fe,Ti,V)12-based sintered magnets to 1.4 T by precisely controlling the intergranular phase. To elucidate how to further increase the coercivity and remanence, we applied data science. We created a dataset including alloy composition, process parameters, microstructure, and magnetic properties of SmFe12-based magnets. We launched an active learning pipeline to further optimize the alloy composition, process parameters, and microstructure to further increase the remanence and coercivity closer to their theoretical limits.
|
| 現在までの達成度 (区分) |
現在までの達成度 (区分)
1: 当初の計画以上に進展している
理由
The research is progressing very smoothly towards the project goal. We have successfully achieved a world record coercivity in SmFe12-based magnetic thin films as well as anisotropic bulk sintered magnets. In parallel, our microstructure-based micromagnetic simulation developed in this study provides clear guidelines on how to further modify the microstructure to increase the coercivity. In addition, the use of a machine learning approach is accelerating the progress of the research to optimize the alloy, process parameters and microstructure to realize a larger coercivity and remanence. The success of the work so far can be manifested with 5 peer-reviewed journal papers published within one year. In addition, 4 invited talks have been given at scientific conferences.
|
| 今後の研究の推進方策 |
We will accelerate the active learning approach to optimize the process parameters, alloy composition and microstructure of SmFe12-based sintered magnets. Using the prepared dataset, a convolutional neural network will be used to predict the role of existing phases on the coercivity and remanence, and the information will be combined with thermodynamic calculations to optimize the existing phases in the magnets. We will develop anisotropic bulk sintered SmFe12-based magnets with optimal alloy composition and process parameters to further increase coercivity and remanence. Multi-scale microstructure characterization and microstructure-based micromagnetic simulations will be used to correlate the obtained magnetic properties, microstructure and coercivity mechanism.
|
