The implementation of artificial intelligence into the research and development of (currently) the most economically relevant classes of engineering hard and soft magnetic materials is addressed. Machine learning is nowadays the key approach utilized in the discovery of new compounds, physical–chemical properties prediction, microstructural/magnetic characterization, and applicability of permanent magnets and crystalline/amorphous soft magnetic alloys. Future opportunities are envisioned on at least two fronts: (a) ultra-low losses materials, as well as processes that enable their manufacturing, unlocking the next step for higher efficiency electrification, power conversion, and distribution; (b) additively manufactured magnetic materials by predicting and developing novel powdered materials properties, generative design concepts, and optimal processing conditions.
A novel process to obtain electrical steel needle-shaped macroscopic magnetic powders is demonstrated, yielding a possible combination of shape anisotropy and crystallographic texture. Such a unique synergy has the potential to (virtually) unlock the manufacturing of any three-dimensional anisotropic soft magnetic structure, even by digital/additive manufacturing. An example of its technological importance is discussed for a 1-hp (~ 746 W) electrical machine correlating efficiency and magnetic loading.
The concept of self-healed permanent magnet machines is presented. In order to demonstrate its technical feasibility, the remagnetization behavior of hard ferrite magnets in closed-circuit as a function of temperature is initially discussed; next, the experimental assessment of the remagnetization of engineered parts – with distinct geometry/dimensions from those of the material assessment – installed in an interior permanent magnet motor is addressed. Although the concept is agnostic regarding the permanent magnet material implemented, this study focused on ferrites since it explores best its benefits due to a unique combination of low criticality of elements, intrinsic physical properties, and cost accessibility.
The microstructural parameter has been examined in order to investigate the influence of its components on the intrinsic coercivity of an isotropic Nd-Fe-B sintered magnet. By applying boundary conditions for the assessment of _(grain(mis)alignment) and _(exc)(exchange coupling) and combining with magnetic measurements and structural characterization results obtained by magnetic small angle neutron scattering, the parameter _(K) associated to the reduction of the magnetocrystalline field plays a major role on the reduction of coercivity of the case study. Approaches in determining are also discussed.
The temperature dependence of the magnetic field at which irreversible magnetic flux loss of ferrite sintered magnets initiates is discussed. In this material class such a field might reduce linearly comparable behavior with that observed for Nd-Fe-B-based sintered counterparts, increase linearly or even peak for temperatures above 350K. These differences are attributed to variations in the saturation polarization and magnetocrystalline anisotropy constant(s) of the hard-magnetic phase arising from changes of chemical composition. A comparison between HK and HD5, where the latter is defined in an IEC standard, is also presented and discussed.