Severe plastic deformation towards magnetic hardening in SmCo based alloys

Permanent magnets play a crucial role in numerous modern technologies. As key components for sustainable electricity generation and electromobility, they are decisive for reducing greenhouse gas emissions and combating climate change. The most commonly used permanent magnets, such as Nd-Fe-B and Sm-Co, are produced by powder metallurgical sintering. In this work, the unconventional process route of high-pressure torsion (HPT) of powder blends is investigated. The application of HPT to powder blends enables the free selection of different magnetic and binder phases. Simultaneously, the process leads to consolidation and thus bulk nanocomposite materials, that are inaccessible via conventional melting-based routes. Using powder blends of SmCo₅ and Cu or Sn, the effect of increasing number of rotations during HPT was studied in detail. Structural refinement was observed, reaching particle sizes within the region of the single domain particle size of SmCo₅ and well below the limits of SmCo₅ sintered magnets. Microstructural analyses revealed that refinement of the SmCo₅ phase is carried by first fragmentation via cracking, before the phase is plastically deformed. The decreasing SmCo₅ particle size, combined with magnetic decoupling of the hard magnetic phase by the diamagnetic Cu or Sn phase, results in increased coercivity. However, further refinement does not lead to further increase of the coercivity caused by partial amorphization of the SmCo₅ phase. The detrimental character of the amorphous hard magnetic phase was revealed by micromagnetic simulations. Furthermore, a deformation induced texture formation was found, leading to anisotropic magnetic properties. By selecting binder phases such as Ni, Co and Cr, different magnetic interaction mechanisms were studied. This work demonstrates that a wide variety of composite materials can be produced by HPT of powder blends, that are not accessible via conventional powder metallurgical sintering.