Magnetic components, such as inductors and transformers, have important effects on the efficiency and performance of switching power supplies; their parasitic properties directly impact the high frequency properties which can cause lot-to-lot variation or unanticipated and non-ideal operation. They are also amongst the most problematic components to design, often requiring numerous design-prototype-test interactions. The electrostatic and electromagnetic analysis of wound components has become more important recently to predict their performance and frequency behavior. Accurate prediction and design of winding parasitic parameters of leakage inductance and winding capacitance for high frequency inductors and transformers in switching power supplies is fundamental to improve performance, lower cost, and speed time to market. This thesis presents a methodology and process to obtain accurate prediction of the inter- and intra-winding capacitances of high frequency magnetic components. Application examples considered are a single-winding choke, a coupled inductor filter, and a multi-winding transformer. Analytical approach for determination of parasitic capacitances in high frequency magnetic components will be covered also. Comparison of the FEA results using JMAG with experimental and empirical formula results show good agreement, supporting the method as a model-based design tool with the potential to significantly reduce the design-prototype-test cycle commonly needed with sophisticated magnetic designs.

Reducing the size of bulky magnetic components (inductors and transformers) in power converters can be achieved by increasing switching frequency and applying innovative designs of magnetic components. Ferrite is the most suitable bulk magnetic material for working at high frequencies but it is difficult to fabricate novel designs of ferrite magnetic components because of the limitations of conventional fabrication methods. Additive manufacturing (AM) has the potential to make customize ferrite magnetic components. One big challenge in 3D printing ferrite magnetic components is the lack of compatible and functional ferrite materials as printers' feedstock. This work focuses on developing ferrite feedstock for 3D printing ferrite magnetic components and providing a guideline for formulating ferrite feedstock by studying the effects of materials and processing parameters on major properties of the ferrite feedstock. The ferrite feedstock should not only be processable by a 3D printer but also make functional ferrite material that can work in power converters. To meet the requirements, the following four aspects of the feedstock are considered in this study: 1. the feedstock should be sinterable to achieve high enough magnetic permeability; 2. magnetic permeability of the feedstock can be easily tailored; 3. rheological properties of the feedstock should ensure reasonable printing resolution; 4. the feedstock can print high aspect ratio structures without slumping. Based on the four major considerations and the desired properties, materials were selected for formulating the ferrite feedstock. The effects of materials and processing variables on the major properties of the ferrite feedstock need to be studied to develop a formulation guidance of the feedstock. The effects of materials fractions and the post-printing peak sintering temperature of the feedstock on maximizing magnetic permeability were studied. The peak sintering temperature had a significant impact on permeability and solid loading (SL) and solid loading excluding diluent (SLED) had smaller impacts. Densities and microstructures of the sintered ferrite cores were characterized to illustrate how the variables affect magnetic permeability. Adding sintering additives to the feedstock was selected as an easy and effective way to tailor the permeability of the ferrite feedstock. The effect of the fractions of two types of additives, SiO2 and Co3O4, on permeability of ferrite were studied. Both SiO2 and Co3O4 can effectively reduce the permeability of the ferrite. A novel multi-permeability toroid core design was 3D-printed with ferrite feedstocks having different fractions of SiO2 to demonstrate the feasibility of fabricating special designs of ferrite magnetics using feedstocks with additives. Core-loss densities of ferrite cores fabricated with feedstocks having different fractions of the two additives were also characterized since it is another important property of ferrite cores in high-frequency converters. Adding SiO2 significantly increases the core-loss density of ferrite cores while adding proper fractions of Co3O4 decreased core-loss density at low magnetic flux densities. The mechanisms of how Co3O4 affect permeability and core-loss density were discussed. The effect of the solid loading (SL) on print-line width resolution was studied by conducting line printing tests. The experiment results showed the best print-line width resolution was achieved using the feedstock with an intermediate SL. The is, which considered both viscosity of the feedstock and coagulation in the feedstock suspension, were discussed. The effect of solid loading excluding diluent (SLED) and UV illumination time on the achievable aspect ratio of printed feedstock was studied. Yield shear strength (y) of feedstocks composition versus UV-curing time were characterized. We evaluated various phenomenological models reported in the literature for predicting the critical yield shear strength (y*) required to obtain a paste structure for a certain aspect ratio. Knowing y* would help to determine the shortest time needed for UV illumination. Applying the model that best fitted to our experimental results, we developed a processing guideline that from specified magnetic permeability and dimensions of a ferrite core, would prescribe the needed SLED and the minimal UV curing time for printing. The guideline was demonstrated by the successful fabrication of tall ferrite inductor cores commonly found in power converters. The main contributions of this study are listed below: 1. Designed, formulated, and characterized ferrite feedstock that not only has functionality for power electronics applications but is also compatible with a direct extrusion type 3D printer. The feedstock can be made into ferrite cores with relative permeability ranging from 10 to 500 which are much higher than those of soft ferrite feedstocks currently reported elsewhere. The packing densities of 950℃ sintered ferrite cores made from the feedstock can be as high as 95%. With the Hyrel 30M 3D-printer, the smallest nozzle orifice diameter that the feedstock can be extruded from is 0.42 mm. We demonstrated printing of the feedstock into a cylinders with a height of 18 mm and an aspect ratio of 3 without slumping issue. 2. Identified the effects of materials and processing variales on 4 major considerations of the ferrite feedstock including maximizing sintered packing density, tailoring permeability, print-line resolution, and achievable dimensions of the printed feedstock without slumping. A deeper understanding of the mechanisms of how the variables affect main properties of the feedstock was provided. 3. Provided a preparation guideline of the ferrite feedstock that prescribe feedstock formulation and UV illumination time per print-layer from the target relative permeability and dimension of a ferrite core. / Doctor of Philosophy / To reduce the size of power electronic devices, applying novel designs of ferrite magnetic components (inductors and transformers) is a promising method. While conventional fabrication methods cannot fabricate novel designs of ferrite magnetic components that have odd or intricate geometries, additive manufacturing (AM) has the potential. One big challenge in 3D printing ferrite magnetic components is the lack of compatible and functional ferrite materials as printers' feedstock. This work focuses on developing ferrite feedstock for 3D printing ferrite magnetic components and providing a guideline for formulating ferrite feedstock by studying the effects of materials and processing parameters on major properties of the ferrite feedstock. The ferrite feedstock should not only have the desired functionalities but also be suitable for printing. Major considerations and desired properties of the feedstock were discussed. Materials were selected to formulate the feedstock based on the desired properties. To develop a formulation guidance for the feedstock, the effects of materials and processing variables on the major properties of ferrite feedstock were studied. The studies included the following 4 aspects: 1. the effects of materials fractions in the feedstock and the post-printing sintering temperature of the feedstock on maximizing magnetic permeability; 2. the effect of additives in the feedstock on tailoring permeability; 3. the effect of feedstock rheology on print-line resolution; 4. the effect of materials fraction and ultraviolet light illumination time on achievable aspect ratio of printed feedstock.

Magnetic components such as inductors and transformers are generally the largest circuit elements in switch-mode power systems for controlling and processing electrical energy. To meet the demands of higher conversion efficiency and power density, there is a growing need to simplify the process of fabricating magnetics for better integration with other power electronics components. The potential benefits of additive manufacturing (AM), or more commonly known as three-dimensional (3D) printing technologies, include shorter lead times, mass customization, reduced parts count, more complex shapes, less material waste, and lower life-cycle energy usage—all of which are needed for manufacturing power magnetics. In this work, an AM technology for fabricating and integrating magnetic components, including the design of manufacturing methodology and the development of the feedstock material, was investigated. A process flow chart of additive manufacturing functional multi-material parts was developed and applied for the fabrication of magnetic components. One of the barriers preventing the application of 3D-printing in power magnetics manufacturing is the lack of compatible and efficient magnetic materials for the printer's feedstock. In this work, several magnetic-filled-benzocyclobutene (BCB) pastes curable below 250 degree C were formulated for a commercial multi-material extrusion-based 3D-printer to form the core part. Two magnetic fillers were used: round-shaped particles of permalloy, and flake-shaped particles of Metglas 2750M. To guide the formulation, 3D finite-element models of the composite, consisting of periodic unit cells of magnetic particles and flakes in the polymer-matrix, was constructed. Ansoft Maxwell was used to simulate magnetic properties of the composite. Based on the simulation results, the pastes consisted of 10 wt% of BCB and 90 wt% of magnetic fillers—the latter containing varying amounts of Metglas from 0 to 12.5 wt%. All the pastes displayed shear thinning behavior and were shown to be compatible with the AM platform. However, the viscoelastic behavior of the pastes did not exhibit solid-like behavior, instead requiring layer-by-layer drying to form a thick structure during printing. The key properties of the cured magnetic pastes were characterized. For example, bulk DC electrical resistivity approached 107 Ω⋅cm, and the relative permeability increased with Metglas addition, reaching a value of 26 at 12.5 wt%. However, the core loss data at 1 MHz and 5 MHz showed that the addition of Metglas flakes also increased core loss density. To demonstrate the feasibility of fabricating magnetic components via 3D-printing, several inductors of differing structural complexities (planar, toroid, and constant-flux inductors) were designed. An AM process for fabricating magnetic components by using as-prepared magnetic paste and a commercial nanosilver paste was developed and optimized. The properties of as-fabricated magnetic components, including inductance and DC winding resistance, were characterized to prove the feasibility of fabricating magnetic components via 3D-printing. The microstructures of the 3D-printed magnetic components were characterized by Scanning-electron-microscope (SEM). Results indicate that both the winding and core magnetic properties could be improved by adjusting the formulation and flow characteristics of the feed paste, by fine-tuning printer parameters (e.g., motor speed, extrusion rate, and nozzle size), and by updating the curing profile in the post-process. The main contributions of this study are listed below: 1. Developed a process flow chart for additive manufacturing of functional multi-material components. This methodology can be used as a general reference in any other research area targeting the utilization of AM technology. 2. Designed, formulated and characterized low-temperature curable magnetic pastes. The pastes are physically compatible with the additive manufacturing platform and have applications in the area of power electronics integration. 3. Provided an enhanced understanding of the core-loss mechanisms of soft magnetic materials and soft magnetic composites at high frequency applications. / Ph. D. / Magnetic components such as inductors and transformers are typically the largest circuit elements in switch-mode power systems for controlling and processing electrical energy. To meet the demands of higher conversion efficiency and power density, there is a growing need to simplify the process of fabricating magnetics for better integration with other power electronics components. The potential benefits of additive manufacturing (AM), or more commonly known as three-dimensional (3D) printing, include shorter lead times, mass customization, reduced parts count, more complex shapes, less material waste, and lower life-cycle energy usage—all of which are needed for manufacturing power magnetics. In this work, an AM technology for fabricating and integrating magnetic components, including the design of manufacturing methodology and the development of the feedstock material, was investigated. A process flow chart of additive manufacturing functional multi-material parts was developed and applied for the fabrication of magnetic components. One of the barriers preventing the application of 3D-printing in power magnetics manufacturing is the lack of compatible and efficient magnetic materials for the printer’s feedstock. Therefore, several magnetic-filled-benzocyclobutene (BCB) pastes were formulated and characterized for a commercial multi-material extrusion-based 3D-printer to form the core parts. To demonstrate the feasibility of fabricating magnetic components via 3D-printing, several inductors of differing structural complexities were designed. An AM process for fabricating magnetic components by using as-prepared magnetic paste and a commercial nanosilver paste was developed and optimized. Results indicate that both the winding and core magnetic properties could be improved by adjusting the formulation and flow characteristics of the feed paste, by fine-tuning printer parameters, and by updating the curing profile in the post-process.

Magnetic components are necessary for switch-mode power electronics converters, but they are often the bulkiest and heaviest in the system. Novel magnetic designs with intricate structures lead to the size reduction of power electronics converters but pose challenges to the fabrication process and material availability. Because of their low-temperature and pressure-less process-ability, magnetic pastes would be the material of choice to make magnetic cores with complex geometries. However, most magnetic pastes reported in the literature suffer from low relative permeability (µr < 26) due to the low magnetic fraction limited by viscosity. The conventional approach of developing magnetic pastes involves experimental iterations with trial-and-error efforts to determine the optimal compositions. To shorten the development cycle and take advantage of the computational power in the current age, this work focuses on exploring, validating, and demonstrating a computer-aided methodology to correlate material's processing, microstructure, and property to guide the development of magnetic pastes. The discrete element method (DEM) simulation was explored to create materials' microstructure and the finite element method (FEM) simulation was utilized to study the magnetic permeability based on the microstructure created by DEM or taken from an actual material sample. The combination of DEM and FEM provided the linkage among processing-microstructure-property relations. Then, the methodology was verified and demonstrated by improving a starting formulation. The formulation was modeled with DEM based on multiple variables, e.g., particle shape, size, size distribution, mixing ratio, gap, gap distribution, magnetic volume fraction, etc. The optimal mixing ratio of different powders to achieve the maximum magnetic fraction was determined by DEM. Experimental results confirmed the predicted optimal mixing ratio. To further take advantage of the computational tools, the magnetic permeability of the magnetic pastes was computed by FEM based on the DEM-generated microstructures. The effects of powder mixing ratio and magnetic volume fraction on the magnetic permeability were studied, respectively. Compared with the experimental values, the microstructure-based FEM simulations could predict the magnetic permeability of the formulations with varied powder mixing ratios or magnetic volume fractions with an average error of only 10 %. Another critical aspect of employing magnetic pastes for magnetic components in power electronics is capable of tailoring their magnetic permeability to meet different design needs. The methodology was further verified and demonstrated by guiding the selection of composition parameters for tailorable magnetic permeability of a starting formulation with flaky particles. An FEM model was constructed from a microstructural image and varied parameters were explored (particle permeability, matrix permeability, particle volume fraction, etc.) to tailor the magnetic permeability. To verify the simulated results, a set of magnetic pastes with various volume fractions of flakes was prepared experimentally and characterized for their permeability. Comparing the simulated and measured permeability, the error was found to be less than 10 %. Last, the guideline was demonstrated to predict a material composition to achieve a target relative permeability of 30. From the predicted composition, the magnetic paste was prepared and characterized. The error between experimental permeability and the target was only 5 %. With the guideline, one can formulate magnetic pastes with tailorable permeability with minimal experimental effort and select the composition parameters to achieve a target permeability. After developing a series of magnetic pastes with tailorable permeability and a maximum value of 35, the feasibility of making magnetic components with magnetic pastes was demonstrated. The commonly used magnetic cores – C-core, E-core, toroid core, bar core, and plate core were fabricated by a low-temperature (< 200 °C) and pressure-less molding process. Several innovative magnetic components with intricate core structures were also fabricated to demonstrate the shape-forming flexibility. The magnetic paste can also be used as the feedstock for paste-extrusion-based additive manufacturing, which further enhances the shape-forming capability. For demonstration, a multi-permeability core was fabricated by 3D printing the magnetic pastes with tailored permeability. The feasibility of making high-performance magnetic components by additive manufacturing or low-temperature pressure-less molding of magnetic pastes opens the door to power electronics researchers to explore more innovative magnetic designs to further improve the efficiency and power density of the power electronics converters. / Doctor of Philosophy / Magnetic components are necessary for switch-mode power electronics converters, but they are often the bulkiest and heaviest in the system. To reduce the size of the power converters, it is crucial to reduce the size of magnetic components by employing innovative magnetic designs. However, the complicated geometries of the novel magnetic designs pose challenges to the availability of material feedstock and the fabrication process. Magnetic pastes would be the material of choice to make magnetic components with intricate structures because of their flexibility in shape-forming with low-temperature and pressure-less processes. However, most magnetic pastes reported in the literature suffer from low magnetic permeability due to the low magnetic fraction limited by viscosity. The conventional approach of developing magnetic pastes involves experimental trial-and-error efforts to determine the optimal compositions. To shorten the development cycle and take advantage of computational power in the current age, this project focuses on exploring, validating, and demonstrating a computer-aided way to correlate material's processing, microstructure, and property relations to guide material development. The numerical simulations were explored to generate the microstructures and study the properties. With the guidance provided by computer simulations, a series of magnetic pastes with tailorable permeability was developed. Several novel magnetic components were fabricated with the as-developed magnetic pastes via molding or additive manufacturing to demonstrate the shape-forming flexibility.

Cette thèse traite de la conception d’un convertisseur DC / DC destiné aux véhicules électriques et hybrides (2,5 kW, 400V/14V, 250kHz). Dérivé de la topologie LLC à résonance, ce convertisseur bénéficie des nombreux avantages propres à cette structure particulière. C’est ainsi que le prototype réalisé présente un rendement très élevé, une densité de puissance très forte avec des perturbations EMI très réduites. La première partie de cette thèse est consacrée à l’analyse théorique du circuit LLC afin de dégager un modèle de conversion et une stratégie de contrôle adaptée à l’application visée. Afin de conserver un rendement important sur une large plage de charge, une structure basée sur la mise en parallèle de deux modules LLC est proposée. Une nouvelle stratégie de contrôle à deux boucles est également proposée pour équilibrer le courant entre les deux modules. La seconde partie de la thèse fait appel à la simulation et à l’expérimentation. Il s’agit de minimiser la masse et l’encombrement tout en maximisant le rendement. Un composant magnétique spécial est conçu puis dimensionné pour intégrer le transformateur et diverses inductances de résonance. Ce convertisseur met également en œuvre un système de redressement synchrone robuste avec une compensation de phase, un module de puissance avec une résistance thermique très faible et un système de refroidissement efficace par air. Le rendement maximal mesuré est 95%. Le rendement demeure supérieur à 94% sur une plage de puissance s’étalant de 500 W à 2 kW. La densité de puissance est 1W/cm3. La CEM du convertisseur est développée dans cette thèse. / In this dissertation, a 2.5kW 400V/14V, 250kHz DC/DC converter prototype is developed targeted for electric vehicle/hybrid vehicle applications. Benefiting from numerous advantages brought by LLC resonant topology, this converter is able to perform high efficiency, high power density and low EMI. A first part of this dissertation is the theoretical analysis of LLC: topology analysis, electrical parameter calculation and control strategy. To arrange high output current, this thesis proposes parallel connected LLC structure with developed novel double loop control to realize an equal current distribution. The second part concerns on the system amelioration and efficiency improvement of developed LLC. A special transformer is dimensioned to integrate all magnetic components, and various types of power losses are quantified based on different realization modes and winding geometries to improve its efficiency. This converter also implements a robust synchronous rectification system with phase compensation, a power semiconductor module, and an air-cooling system. The power conversion performance of this prototype is presented and the developed prototype has a peak efficiency of 95% and efficiency is higher than 94% from 500W to 2kW, with a power density of 1W/cm3. The CEM analysis of this converter is also developed in this thesis.

Le travail présenté dans ce mémoire concerne le prototypage virtuel 3D des composants électromagnétiques d’électronique de puissance, par la technique des éléments finis. La démarche correspond à la volonté de disposer d’outils de simulation multiphysiques 3D toujours plusperformants, notamment dans le contexte de l’intégration en électronique de puissance. Il s’agit de mettre au point des méthodes et desprocédures adaptées à la caractérisation d’inductances, de transformateurs ou de coupleurs multiphasés haute fréquence, dans unenvironnement de conversion statique, avec des formes d’onde de tension et de courant non sinusoïdales. Cela nécessite de connaître le comportement harmonique des composants électromagnétiques sur une large gamme de fréquence, et de tenir compte des spécificités de réalisation comme l’utilisation de bobinages en technologie feuillard ou planar et de noyaux magnétiques en matériaux ferrite. Dans le premier chapitre de ce mémoire, une analyse des limitations actuelles des modèles analytiques et numériques des composants magnétiques HF en électronique de puissance est faite afin de définir les besoins les plus importants qui seront par la suite abordés. Dans cette optique, le choix de la plateforme ouverte de simulation multiphysique en 3D, par éléments finis, COMSOL Multiphysics, a été fait. En effet, l’objectif ici n’est pas de développer un nouveau code de calcul mais de mettre au point un outil de simulation adapté aux problématiques rencontrées en électronique de puissance. Le deuxième chapitre aborde le point très important de la détermination des pertes cuivre HF en tenant compte des effets fréquentiels tels que les effets de peau et de proximité. La problématique des bobinages feuillards ou planars est résolue par l’utilisation d’éléments coques spécifiques. Le troisième chapitre traite de la détermination despertes fer, à haut niveau d’excitation et pour des formes d’onde de champ non sinusoïdales. Sur la base d’abaques de densités de pertesfournies par les constructeurs de matériaux, deux méthodes de calcul sont proposées, l’une en cours de traitement et l’autre en posttraitement. La prise en compte de la non-linéarité est analysée ainsi que la problématique de l’existence localisée de champs tournants. Le quatrième chapitre aborde l’extraction virtuelle des paramètres électriques des composants électromagnétiques multiphasés et la définition de matrices d’impédances (inductances et résistances propres et mutuelles), en fonction de la fréquence. A l’exception des pertes fer non prise en compte ici, cette formalisation permet de traduire finement le comportement harmonique large bande des composants multiphasés. Finalement, le cinquième chapitre propose trois exemples d’utilisation de ce nouvel outil. Le premier exemple aborde ledimensionnement optimal et la caractérisation virtuelle d’un coupleur triphasé de forte puissance. Le deuxième exemple montre l’intérêt dela modélisation harmonique sous la forme de matrices impédances pour simuler le comportement d’un coupleur hexaphasé in-situ dans unconvertisseur de puissance. Enfin le troisième exemple montre la possibilité de lancer des campagnes d’études paramétriques automatisées pour étudier l’évolution d’un ou plusieurs paramètres dimensionnant afin de calculer des tables de réponses d’aide au dimensionnement. / The work presented in this thesis deals with the 3D virtual prototyping of electromagnetic components used in power electronics thanks to the finite element method. The approachfollows the desire to have an always more and more powerfull 3D multiphysics simulation tools, especially in the context of power electronics integration. It consists indeveloping adapted methods and procedures to characterize inductors, transformers or multiphase highfrequency InterCell Transformers (ICT), in a static conversion environment, in which voltage and current waveforms are non-sinusoidal. This requires the knowledge of the harmonic behavior of electromagnetic components over a wide frequencyrange, and to take into account the realization specificities such as the use of foil or planar windings technology and the use of ferrite magnetic cores. In the first chapter of this thesis, an analysis of today's analytical and numerical models limitations of HF power electronics magnetic components is made in order to identify the most important needs that will be addressed later. In this context, the choice of the open platform for multiphysics simulation in 3D finite element, COMSOL Multiphysics, has been done. Indeed, the aim here is not to develop a new calculation code but to offer an appropriate simulation tool to face the problems encountered in power electronics. The second chapter broaches the very important issue of determining the HF copper losses by taking into account frequency phenonema such as skin and proximity effects. The problem of foils planar windings is resolved by the use of specific shell elements. The third chapter concerns the determination of iron losses under high level supply and non-sinusoidal waveforms. Based on losses densities charts provided by the materials manufacturers, we propose two calculation methods: one in ongoing-processing and another in post-processing. The consideration of the nonlinearity is analyzed and the problem of the existence of localized rotating fields. The fourth chapter discusses the virtual extraction of the electrical parameters of multiphase electromagnetic components and the definition of frequency depedant impedance matrices (self and mutual inductances and resistances). Except for iron losses that are not taken int account here, this formalization can fine translate the broadband harmonic behavior of multiphase components. Finally, the fifth chapter presents three examples of practical application of this new tool. The first example discusses the optimal design and virtual characterization of a high power three-phased ICT. The second example shows the interest of modeling harmonic impedances in the form of matrices to simulate the behavior of a six-phased ICT in situ in a power converter. Finally, the third example shows the ability to run automated parametric study campaigns to study the evolution ofone or more sizing parameters to calculate response tables to help the designing.

La complexité croissante des dispositifs en électronique de puissance nécessite l'intervention de la conception assistée par ordinateur. Le développement de systèmes électriques/électroniques est effectué à l'aide du prototypage virtuel dans lequel les logiciels de simulation sont utilisés pour prédire le comportement des composants. De ce fait, le prototypage virtuel permet une économie de temps et d'argent pour la réalisation de prototypes. La demande croissante d'appareils à faible puissance et à haut rendement a obligé les concepteurs à analyser précisément les pertes de chaque composant constituant du système. Les composants magnétiques constituent une partie importante des appareils en électronique, par conséquent la modélisation précise des matériaux magnétiques est nécessaire afin de prédire leur comportement réaliste dans des conditions de fonctionnement variables selon l'application. Notre travail s'inscrit dans ce contexte et propose un modèle dynamique non linéaire de composants magnétiques pour une utilisation dans des simulateurs de circuits électriques. Ce modèle de composant magnétique inclut le comportement d'hystérésis non linéaire du matériau et permet une modélisation précise des pertes fer et des pertes joule avec de plus la prise en considération des effets thermiques qui, généralement, ne sont pas pris en compte par les modèles existants. Le modèle est basé sur le principe de la séparation des contributions statiques et dynamiques des pertes fer et s'appuie sur la théorie de Bertotti. Le langage de programmation VHDL-AMS est utilisé en raison de sa fonctionnalité de modélisation multidomaines, permettant un couplage avec un modèle thermique. Le modèle de composant magnétique est mis en oeuvre dans le logiciel de simulation de circuit "Simplorer". Il est ensuite testé dans une application de convertisseur de puissance, le convertisseur abaisseur qui permet de fournir une excitation non-conventionnelle. Le modèle est validé pour différents noyaux d'inductances, différentes ondulations de courant et niveaux de charge, différentes températures et une large gamme de fréquence / The increasing complexity of power electronic devices requires the intervention of computer-aided design in electrical engineering. Development of electric/electronic systems nowadays is carried out by the help of virtual prototyping, in which simulation software are used to predict components behavior without investing time and money to build physical prototypes. The increasing demand of low power, high efficiency devices forced designers to precisely analyze losses in each component constituting the system. Magnetic components constitute a major part of electronics devices. Therefore accurate modeling of magnetic materials is mandatory in order to predict their realistic behavior under variable operating conditions. Our work takes place in this context by proposing a non-linear dynamic model of magnetic components for use in circuit simulators. It includes the material nonlinear hysteretic and dynamic behaviors with accurate modeling of winding and core losses in addition to thermal effects that are not taken into account by existing models. The model is based on the principle of separation of static and dynamic contributions as well as Bertotti’s theory. VHDL-AMS is used as a modeling language due to its multi-domain modeling feature, allowing coupling with a thermal model. The magnetic component model is implemented in circuit simulation software “Simplorer” It is then tested in a widely used power converter application, the buck converter, to ensure non conventional excitation. The model is validated for different core inductors, different current ripples, different loads, different temperatures and a wide frequency range

L'amplitude vibratoire d'une structure mince peut être réduite grâce au couplage électromécanique qu'offrent les matériaux piézoélectriques. En termes d'amortissement passif, les shunts piézoélectriques permettent une conversion de l'énergie vibratoire en énergie électrique. La présence d'une inductance dans le circuit crée une résonance électrique due à l'échange de charges avec la capacité piézoélectrique. Ainsi, l'ajustement de la fréquence propre de ce shunt résonant à celle de la structure mécanique équivaut à la mise en œuvre d'un amortisseur à masse accordée. Cette stratégie est étendue au contrôle d'une structure multimodale par multiplication du nombre de patchs piézoélectriques. Ceux-ci sont interconnectés via un réseau électrique ayant un comportement modal approximant celui de la structure à contrôler. Ce réseau multi-résonant permet donc le contrôle simultané de plusieurs modes mécaniques. La topologie électrique adéquate est obtenue par discrétisation de la structure mécanique puis par analogie électromécanique directe. Le réseau analogue fait apparaître des inductances et des transformateurs dont le nombre et les valeurs sont choisis en fonction de la bande de fréquences à contrôler. Après s'être penché sur la conception de composants magnétique adaptés, la solution de contrôle passif est appliquée à l'amortissement de structures unidimensionnelles de type barres ou poutres. La stratégie est ensuite étendue au contrôle de plaques minces par mise en œuvre d'un réseau électrique bidimensionnel. / Structural vibrations can be reduced by benefiting from the electromechanical coupling that is offered by piezoelectric materials. In terms of passive damping, piezoelectric shunts allow converting the vibration energy into electrical energy. Adding an inductor in the circuit creates an electrical resonance due to the charge exchanges with the piezoelectric capacitance. By tuning the resonance of the shunt to the natural frequency of the mechanical structure, the equivalent of a tuned mass damper is implemented. This strategy is extended to the control of a multimodal structure by increasing the number of piezoelectric patches. These are interconnected through an electrical network offering modal properties that approximate the behavior of the structure to control. This multi-resonant network allows the simultaneous control of multiple mechanical modes. An adequate electrical topology is obtained by discretizing the mechanical structure and applying the direct electromechanical analogy. The analogous network shows inductors and transformers, whose numbers and values are chosen according to the frequency band of interest. After focusing on the design of suitable magnetic components, the passive control strategy is applied to the damping of one-dimensional structures as bars or beams. It is then extended to the control of thin plates by implementing a two-dimensional analogous network.

Les composants magnétiques planaires (inductance et transformateur) occupent une place importante dans certains circuits intégrés utilisés en haute fréquence. Leur miniaturisation et leur intégration vont de pair avec celles des circuits électroniques qui évoluent constamment surtout pour les appareils portables. Quelques travaux scientifiques ont permis d’identifier les différents mécanismes à l’origine de pertes dans les composants magnétiques planaires, afin de les limiter. Les pertes dans les enroulements sont classiquement prises en compte par une résistance r(f) fonction de la fréquence. La détermination, à partir des paramètres S obtenus par mesure ou simulation, de la résistance r(f) constitue à ce jour un sujet d’étude à part entière, les paramètres S étant les seuls paramètres que l’on peut obtenir au-delà de la centaine de MHz. Pour contribuer à la résolution de ce problème, nous avons proposé une méthode prenant en compte toutes les pertes dans le bobinage. Cette méthode de détermination de la résistance en fonction de la fréquence se fait dans trois domaines de fréquence : - en très basse fréquence, la rDC est obtenue par calcul ou mesurée à l’aide d’un matériel basse fréquence, - aux "moyennes fréquences" lorsque les impédances R et Lω ne sont pas trop différentes, les phénomènes capacitifs pouvant être négligés, - aux résonances en très haute fréquence. L’application de cette méthode sur trois structures différentes (inductance à air de plusieurs spires, à air à une spire en oméga et à une couche de matériau magnétique) a permis de : - observer une bonne corrélation entre simulation et mesure, -valider l’évolution des pertes en fonction de la fréquence, -séparer les effets de peau et de proximité, -séparer les pertes fer et les pertes cuivre pour une inductance à couche magnétique / Planar magnetic components (transformer and inductor) have become a big part in some integrated circuits used in high frequency. Miniaturization and integration of magnetic components go hand in hand with the ones of electronics that constantly evolves especially for portable devices. A few scientific studies have identified the different mechanisms of losses in planar magnetic components. Winding losses are generally taken into account using a resistance r(f) versus frequency.The use of scattering parameters S to determine resistance r(f) represents a comprehensive research project ; S parameters that can be obtained either by measurement or by simulation, are the only parameters which one can get at high frequencies (above 100MHz). To solve this problem, we have proposed a method taking into account all winding losses. Our approach for determining r(f) has to be applied in 3 frequency domains: - at very low frequency, r(f) = rDC and its value is either calculated or measured using low frequency equipment, - in the middle frequency range, capacitive coupling can be neglected while impedances R and Lω are in the same order of magnitude, - at very high resonance frequencies.This method has been implemented for 3 different structures (coreless inductor with several turns of coil, Omega shape coreless inductor with one turn and inductor with a magnetic layer) leads to : - observe a good correlation between simulation and measurement, - validate the evolution of losses versus frequency, - separate skin effects and proximity effects, - separate iron losses and winding losses

Cette thèse s’inscrit dans le contexte d’alimentation des systèmes électroniques portables à faible puissance (1W environ) et fonctionnant sous faible tension. Avec la demande croissante pour la conversion d’énergie dans ces systèmes, l’intégration et la miniaturisation du convertisseur DC-DC devient une zone d’intérêt fort. Des recherches récentes ont montré des convertisseurs avec des fréquences de commutation pouvant atteindre 100 MHz. Pour de faibles niveaux de tension (1 V) et des puissances aux environs du Watt, les valeurs d’inductance de lissage de ces convertisseurs envisagées sont de l’ordre d’une centaine de nanoHenry. Ceci relance l’intérêt d’étudier l’intégration des composants passifs de dimensions millimétriques au sein d’un même boîtier avec les parties actives. Dans ce contexte, les travaux présentés dans ce manuscrit sont abordés par la conception d’inductances planaires en forme de spirale avec un noyau magnétique. Les simulations ont permis d’analyser les liens entre les paramètres géométriques et les paramètres électriques de l’inductance pour établir une structure d’inductance optimale en fonction de la limite de la technologie de réalisation. Une inductance planaire prise en sandwich entre deux couches de matériau magnétique est proposée. Les simulations ont montré l’intérêt de réaliser un tel composant. Sa structure présente plusieurs avantages, elle permet d’augmenter considérablement la valeur d’inductance tout en gardant le même encombrement par rapport à une inductance sans noyau magnétique. Elle permet également de réduire les perturbations électromagnétiques avec les composants environnants. Un procédé technologique de réalisation des inductances, basé sur la croissance électrolytique de cuivre à température ambiante, a été développé et optimisé pour valider les modélisations précédentes. Ce procédé est reproductible et permet une fabrication collective de composants. Un banc de caractérisation impédance métrique a également été conçu afin de déterminer les limites du fonctionnement fréquentiel des composants réalisés et de valider les performances de ces derniers. Ce travail propose une solution pour la réalisation de la puce active sur l’inductance dans le cadre d’un SOC (System-On-Chip). Il souligne par ailleurs l’importance de l’intégration pour l’électronique de faible puissance / The work in this thesis contributes to the domain of low power (1W approximately) portable electronic systems. These systems require integrated and miniaturized of DC-DC converters. Recent studies have demonstrated converters with high switching frequency as high as 100 MHz, requiring smaller passive components. For low voltage values (1V approximately) and 1 watt output power, the inductance value of these converter filters is about a hundred nanoHenry. Such inductors can be integrated on a millimetric scale in the same package as the active die. In this context, the work presented in this thesis starts with the design of planar spiral inductors with a magnetic core. Simulations allowed to analyze the relation between geometrical and electrical parameters of the inductor in order to design an optimal inductor. A planar inductor sandwiched between two layers of magnetic material is proposed. Simulations showed the advantages of fabricating of such component. Its structure allows to increase the inductance value without modifying the inductor’s surface compared to a coreless inductor. It also allows to reduce the electromagnetic interferences with the rest of the circuit. A technological process for the fabrication of the inductors has been developed and optimized in order to valid the previous design. This process is based on copper electroplating technique which is compatible with a repeatable and a mass fabrication of inductors. A characterization bench was also developed in order to determine the operating frequency limits of the fabricated components as well as to validate their performance. This work offers a solution for the realization of the active chip on the inductor (SOC, System- On-Chip). It also emphasizes the importance of the integration for low power electronics