Context: Achieving carbon neutrality by 2050 will require massive electrification of the power production systems. Power electronics (PE) is a key-enabler that will this transformation possible (renewables, integration of energy micro-grids, development of electric mobility)
Problem: Current developments in PE converters aim at increasing the switching frequencies of large bandgap switches (SiC or GaN). At low frequencies, magnetic components remain bulky, occupying up to 40% of the total footprint. At high frequencies (HF > 100 kHz), very significant gains are expected, but only if the losses generated by these components remain under control. Today, the main class of magnetic materials applied to HF is MnZn or NiZn ferrites, due to their low cost and convenient electrical resistivity (?elec > 1 O.m). The main drawbacks of these materials are their low saturation induction (Bsat < 0.4 T), which limits their size reduction, and their mechanical fragility. Nanocrystallines materials have better Bsat (1.3 T), but their ?elec is about 1.5 µO.m (6 times less resistive than ferrites), which generates significant induced current losses at HF.
Thesis objective: To develop magnetic composites by grinding nanocrystalline ribbons, electrically insulating the powders (coating fabricated by sol-gel), compacting of the powder at high pressure (1000-2000 MPa) for the core shaping and finally by applying an annealing treatment to relax the thermal constraints.