Physics-interpretable topological machine learning identifies domain-wall contributors to high-frequency loss in a nanocrystalline soft magnet

Abstract High-frequency core loss constrains efficiency and power density, yet loss models do not identify which parts of an evolving domain-wall network carry the dominant reversal resistance. We present a physics-interpretable topological machine-learning workflow combining micromagnetic simulations, persistent homology, principal component analysis and inverse mapping for an idealized nanocrystalline soft-magnet model driven at nominal frequencies of 100 kHz–10 MHz using α = 1.0 and frequency-specific integration. Because reversal here is dominated by domain-wall motion, we treat the shape and energy of the wall network, captured by persistent homology, as the governing factor of loss. The first coordinate tracks magnetization, whereas the second captures frequency-associated domain subdivision and correlates with the exchange contribution to domain-wall energy. Their feature-space gradient reveals a reversal barrier near coercivity that increases with frequency. Inverse mapping localizes the largest contributions to propagating reversed-domain tips and selected fine wall segments rather than uniformly across the wall network. Consistent structure–energy–mechanism relationships connect the representation to magnetic physics. Within this protocol, reversal resistance is associated not simply with wall amount, but with a spatially selective subset of high-energy wall configurations. The workflow provides testable spatial descriptors for low-loss materials screening and optimization.

Authors

Institutions

Publication Details

Journal
NPG Asia Materials
Published
2026-11-17
DOI
https://doi.org/10.1038/s41427-026-00683-2
Primary Topic
Magnetic Properties and Applications
Type
article
Field-Weighted Citation Impact
0.00

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Physics-interpretable topological machine learning identifies domain-wall contributors to high-frequency loss in a nanocrystalline soft magnet

Chiharu Mitsumata, Satoshi Okamoto, Kosuke Nishioka, Alexandre Lira Foggiatto et al.
NPG Asia Materials
Magnetic Properties and Applications
article

Physics-interpretable topological machine learning identifies domain-wall contributors to high-frequency loss in a nanocrystalline soft magnet

Chiharu Mitsumata, Satoshi Okamoto, Kosuke Nishioka, Alexandre Lira Foggiatto, Ippei Obayashi, Yasuaki Hiraoka, Masato Kotsugi, Ryunosuke Nagaoka, Michiki Taniwaki, Takahiro Yamazaki, Tomohiro Tanaka
article en

Abstract

Abstract High-frequency core loss constrains efficiency and power density, yet loss models do not identify which parts of an evolving domain-wall network carry the dominant reversal resistance. We present a physics-interpretable topological machine-learning workflow combining micromagnetic simulations, persistent homology, principal component analysis and inverse mapping for an idealized nanocrystalline soft-magnet model driven at nominal frequencies of 100 kHz–10 MHz using α = 1.0 and frequency-specific integration. Because reversal here is dominated by domain-wall motion, we treat the shape and energy of the wall network, captured by persistent homology, as the governing factor of loss. The first coordinate tracks magnetization, whereas the second captures frequency-associated domain subdivision and correlates with the exchange contribution to domain-wall energy. Their feature-space gradient reveals a reversal barrier near coercivity that increases with frequency. Inverse mapping localizes the largest contributions to propagating reversed-domain tips and selected fine wall segments rather than uniformly across the wall network. Consistent structure–energy–mechanism relationships connect the representation to magnetic physics. Within this protocol, reversal resistance is associated not simply with wall amount, but with a spatially selective subset of high-energy wall configurations. The workflow provides testable spatial descriptors for low-loss materials screening and optimization.

NPG Asia MaterialsVol. 18(1)
University of Tsukuba (JP), Tokyo University of Science (JP), Okayama University (JP), Johannes Gutenberg University Mainz (DE), Katsushika Medical Center (JP), NLT Technologies (Japan) (JP), Institute of Multidisciplinary Research for Advanced Materials, Tohoku University (JP), Kyoto University Institute for Advanced Study (JP)
Ministry of Education, Culture, Sports, Science and Technology, University of Tsukuba, Japan Society for the Promotion of Science, Japan Science and Technology Agency, Core Research for Evolutional Science and Technology
Magnetic Properties and Applications
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon