| Project Detail |
About 30 % of all the electrical power generated passes through a power electronic converter, and the proportion is expected to rise to 80 % in 10-15 years. The amount of electricity annually wasted due to the losses in such systems in the EU corresponds to at least billions of euros. A major part of these losses arises in passive magnetic components, such as inductors and transformers, which are also the largest and heaviest components of a power electronic device. The physical phenomena related to the power losses in the magnetic cores of these components are not properly understood at the moment. In addition, the engineering community is currently lacking efficient modeling tools for analyzing the losses in the windings of such components at high frequencies.
Improvement of high-frequency magnetic components would require accurate understanding of the power loss mechanisms. However, the device-level losses are affected by physical effects taking place in the microscopic grain and domain structures and very thin conductors, which are often subject to geometrical uncertainties. Accurate geometrical models cannot be used for analyzing the devices due to the impossibly large computational burden.
In MULTIMAG, we will address these challenges by establishing a set of new multiscale numerical modeling tools, which will provide insight into the origin of the power losses and make it possible to perform statistical analysis of the electromagnetic behaviour of such components. The application potential of these new numerical tools will be demonstrated by designing working prototypes of emerging power electronic devices, such as a solid-state transformer and a wireless power transfer system. We will also develop inverse problem approaches for identifying the models from available catalog data, lowering the threshold for adopting the models into use.
As the outcome, new means for improving the energy efficiency and power density of power electronic devices will arise. |
