TRENDS AND CHALLENGES IN SOFT MAGNETS FOR HIGHER FREQUENCY APPLICATIONS
Résumé
Today's applications in power electronics and electrical machines requires constantly increasing power density, a demand that can only be satisfied if working frequencies are pushed over the usual limits of magnetic materials. In power electronics, chopping frequencies above 100 kHz are usual for switch-mode-power-supplies (SMPS) even for several kW units and can reach 1 MHz for several 100 W units. Mn-Zn ferrites are perfectly adapted to power applications below 1 MHz, but the core losses are rather high because of their low resistivity, so skin effect is no more negligible. In addition, the Mn-Zn ferrites are quite tricky to produce (the O2 partial pressure as to be controlled during sintering) and not adapted to low temperature co-fired ceramics technology (LTCC) or spark plasma sintering (SPS). By opposition, partially Cu substituted Ni-Zn ferrites are well adapted to these technologies but rather lossy for power applications. Co-doped NiZnCu ferrites appear as a promising material compatible with the new manufacturing technologies of magnetic components and with low power loss around 1MHz. High speed multi-poles electrical machines are today feed at frequencies above 1 kHz which causes problems with classical Fe-Si sheets. Reducing the thickness is a good solution for classical machines, but more and more machines have 3D flux paths making mandatory the used of soft magnetic composites (SMC). Although SMC are produced in industry, the modelization of the losses is still difficult. The two scales nature of the eddy currents, offers no straightforward way to apply Berttoti's theory of separation of losses. An attempt to modelize losses in SMC will be discussed and new directions for improving SMCs will be discussed.