Valentin Rigot, doctorant chez VEDECOM dans le domaine Electrification/Electronique de puissance, soutiendra sa thèse le 21 avril 2022 à 10h. Cette thèse concerne le lot 2 du projet HiDePe mené par l’Institut VEDECOM :
Transformateur à air pour un convertisseur dc-dc bidirectionnel haute densité de puissance et haute fréquence pour l’application automobile
le jeudi 21 avril à 10h
Amphi V dans le bÄtiment Eiffel de Centrale Supélec, 3, Rue Joliot-Curie, 91190 Gif-sur-Yvette, France
Les travaux présentés contribuent à l’amélioration de la densité de puissance dans les chargeurs embarqués dans les véhicules électriques, en profitant notamment de l’essor des composants semi-conducteurs à grand gap, qui autorisent la réalisation de convertisseurs plus compacts.
- Encadrant académique Pr. Tanguy PHULPIN, CentraleSupélec
- Directeur de thèse : Pr. Daniel SADARNAC, CentraleSupélec
- Encadrant industriel : Dr. Jihen SAKLY, VEDECOM
Contact : Leyla Habarek Arioua : leyla.arioua@vedecom.fr
Thesis Abstract
The electric vehicle is currently strongly expanding to face climate change. The power density requirements for onboard power are constantly growing and need evolution to go over the present electronic limitation without decreasing the efficiency.
This thesis focuses on modeling, optimization and realization of a high-frequency coreless transformer integrated into a dedicated converter to validate a bidirectional on-board battery charger operating at 1.5MHz.
The work contributes to the increase of the on-board charger power density by taking benefit of wideband gap semiconductors emergence which opens a new range of admitted switching frequency. It consequently offers an opportunity to increase the compacity of the converters.
After a study on high power density transformers and coreless transformers, a highlight of the opportunity that an air-core transformer designed for power density is presented. The high-frequency electrical conductors have also been considered to determine the best one for our application by considering the skin and the proximity effects. The new transformer geometry is the result of an optimization program able to propose the best turns’ positions for minimizing the magnetic emission while ensuring a certain value of self-inductance. This geometry establishes the first technical breakthrough of this thesis.
Several transformers’ trials were realized on 3D printed polycarbonate supports resulting in a 7kW prototype. An electrical, magnetic, and thermal characterization was done on the transformer after its winding process and validate the prediction. Those values were taken for the start point of a dedicated DC-DC topology. The automotive and bidirectional context led us to merge on the Dual Active Bridge converter. The switching frequency was determined to ensure the required power transfer with a soft-switching operation. This frequency value in automotive chargers establishes the second technical breakthrough of this thesis.
After determining the command signals and choosing the required components for the realization of the converter, a ready-to-use inverter branch was chosen to simplify the complexity of construction. A dedicated printed circuit board was realized to link the inverter branch with the DC bus, the transformer, and the command board of the converter. Finally, experimental results on the converter respect the requirement specifications with an efficiency conversion of 96% for a power transfer of 9kW and a switching frequency of 1.5MHz. The final estimated power density is 8.5 kW/L which is higher than the existing power density in the industry.
Some complementary works must be done to validate the prototype in a final application. It is however possible to observe this kind of transformer in an electrical vehicle in the future. The results were indeed in rupture and presents better volume, better thermal management, and better efficiency.