The National Institute for Materials Science (NIMS) has harnessed the power of advanced simulations to unravel the magnetization reversal process in Nd-Fe-B magnets. By constructing large-scale finite element models based on tomographic data obtained through electron microscopy, researchers have gained unprecedented insights into the microstructural factors influencing coercivity – a key measure of a magnet's resistance to demagnetization.
The significance of this breakthrough extends far beyond scientific curiosity. Nd-Fe-B magnets are the backbone of green power generation, electric transportation, and numerous high-tech industries due to their unparalleled strength and versatility. However, their coercivity has remained below its theoretical limit, hampering their full potential.
Enter micromagnetic simulations on realistic models of these magnets. The newly proposed approach leverages tomographic data from scanning electron microscopy (SEM) combined with focused ion beam (FIB) polishing to create high-fidelity 3D finite element models. This universal methodology opens doors to a deeper understanding of coercivity mechanisms and the nucleation of magnetization reversal.
The developed models serve as digital twins of Nd-Fe-B magnets, offering a virtual representation that accurately mirrors their physical properties. This digital twin concept enables researchers to tackle the inverse problem – designing bespoke high-performance magnets tailored to specific applications. By inputting desired magnetic properties, such as traction or variable magnetic force, researchers can harness data-driven research pipelines integrated with digital twins to propose optimal compositions, processing conditions, and microstructures for these magnets.
The implications are profound. Not only does this approach expedite the development of next-generation magnets with enhanced performance, but it also streamlines the design process, significantly reducing development time and accelerating the transition to sustainable, high-tech solutions.
The research is published in the journal npj Computational Materials.
Source: National Institute for Materials Science