Within a solid-state transformer, the isolated dc-dc converter and in particular its medium-frequency transformer are one of the critical components, as it provides the required isolation between primary and secondary sides and the voltage conversion typically necessary for the operation of the system. A comprehensive optimization procedure is required to find a transformer design that maximizes power density and efficiency within the available degrees of freedom while complying with material limits, such as temperature, flux density, and dielectric strength as well as outer dimension limits. This paper presents an optimization routine and its underlying loss and thermal models, which are used to design a 166 kW/20 kHz transformer prototype achieving 99.4% efficiency at a power density of 44 kW/dm
 3
. Extensive measurements are performed on the constructed prototype in order to measure core and winding losses and to investigate the current distribution within the litz wire and the flux sharing between the cores.

Design and Experimental Analysis of a Medium-Frequency Transformer for Solid-State Transformer Applications

This is a study of potential improvements of diagnostic methods used on high-voltage generators and motors.  It considers offline electrical measurements on the main insulation of stator windings, ...

/pdf/dielectric-response-and-partial-discharge-measurements-on-3aext28uh4.pdf

Dielectric response and partial discharge measurements on stator insulation at varied low frequency

Mon travail s'articule autour des problemes de satisfaction de contraintes (CSP) et de leurs extensions comme les problemes de contraintes quantifies (QCSP), la question essentielle etant de comprendre leur complexite. Ceux-ci peuvent etre vus comme des problemes de model-checking pour certains fragments de la logique du premier ordre. Dans le cas des CSP, Feder et Vardi ont propose la conjecture de la dichotomie, a savoir que chaque probleme CSP est soit facile (dans P) soit difficile (NP-complet). Dans le cas des QCSP, il n'existe pas encore de conjecture mature, mais une trichotomie entre P, NP-complet et Pspace-complet est envisageable.Mon manuscrit comporte deux parties.Une premiere partie porte sur la complexite du model checking pour les fragments syntaxiques de la logique du premier ordre et presente de maniere uniforme des polychotomies pour tous ces fragments sauf CSP et QCSP. Notre methodologie utilise trois ingredients : algebrique (correspondance de Galois pour apprehender l'expressivite d'un ensemble de relations), combinatoire (core generalise, i.e. sous-modele satisfaisant les meme formules pour le fragment etudie) et logique (relativisation de quantificateur(s)). En particulier, nous demontrons une tetrachotomie pour le fragment positif sans egalite de la logique du premier ordre qu'on peut voir comme une extension du QCSP pour laquelle on s'autorise la disjonction.La seconde partie porte sur des questions autour de la logique de Feder et Vardi. Cette logique est intimement liee a la classe des CSP de domaine fini: elle exhibe une dichotomie si, et seulement si, la conjecture de la dichotomie est vraie. Un probleme capture dans cette logique s'exprime en terme de motifs colories interdits ou encore comme un CSP de domaine infini etudies par Bodirsky. De plus, on peut decider lorsqu'il s'agit d'un CSP fini et l'inclusion de tels problemes. Dans le cas special ou ces problemes sont aussi du premier-ordre, on retrouve des questions similaires sous le nom de dualite par homomorphisme avec les travaux en combinatoire de Nesetril et. al. De tels problemes du premier ordre apparaissent aussi en logique lorsqu'on considere les problemes du premier ordre preserves par homomorphisme inverse (theoreme de preservation de Rossman). Ces resultats de combinatoire et de logique peuvent etre releves pour s'appliquer aux problemes de motifs interdits. En particulier, pour des classes de structures suffisamment peu denses, les problemes de motifs interdits sont des CSP de domaine fini.

https://tel.archives-ouvertes.fr/tel-01096078/document

Mémoire d'habilitation à diriger des recherches

The windings insulation of electrical machines will remain a topic that is updated frequently. The criteria severity requested by the electrical machine applications increases continuously. Manufacturers and designers are always confronted with new requirements or new criteria with enhanced performances. The most problematic requirements that will be investigated here are the extremely long lifespan coupled to critical operating conditions (overload, supply grid instabilities, and critical operating environments). Increasing lifespan does not have a considerable benefit because the purchasing price of usual machines has to be compared to the purchasing price and maintenance price of long lifespan machines. A machine having a 40-year lifespan will cost more than twice the usual price of a 20-year lifetime machine. Systems which need a long lifetime are systems which are crucial for a country, and those for which outage costs are exorbitant. Nuclear power stations are such systems. It is certain that the used technologies have evolved since the first nuclear power plant, but they cannot evolve as quickly as in other sectors of activities. No-one wants to use an immature technology in such power plants. Even if the electrical machines have exceeded 100 years of age, their improvements are linked to a patient and continuous work. Nowadays, the windings insulation systems have a well-established structure, especially high voltage windings. Unfortunately, a high life span is not only linked to this result. Several manufacturers’ improvements induced by many years of experiment have led to the writing of standards that help the customers and the manufacturers to regularly enhance the insulation specifications or qualifications. Hence, in this publication, the authors will give a step by step exhaustive review of one insulation layout and will take time to give a detailed report on the standards that are linked to insulation systems. No standard can provide insurance about lifespan, nor do any insulation tests incorporate all of the operating conditions: thermal, mechanical, moisture and chemical. Even if one manufacturer uses the standards compliance to demonstrate the quality of its realization; in the end, the successful use in operation remains an objective test. Thereafter, both customer and manufacturers will use the standards while knowing that such documents cannot fully satisfy their wishes. In one 20-year historical review, the authors will highlight the duration in insulation improvements and small breakthroughs in standards writing. High lifespan machines are not the main interest of standards. A large part of this publication is dedicated to the improvements of the insulation wall to achieve the lifespan. Even if the choice of raw materials is fundamental, the understanding of ageing phenomena also leads to improvements.

The Insulation for Machines Having a High Lifespan Expectancy, Design, Tests and Acceptance Criteria Issues

The latest generation of fusion devices needs large superconducting magnets that have to withstand operational voltages of tens of kV. Their insulation typically targets a dielectric strength of about 10 times the operational voltages to cope with degradation during operation resulting from the high electromagnetic loads with cyclic components, thermal stresses due to cycles from 293 K to 4 K, and ionizing radiation. Degradation of dielectric strength is further mitigated by applying solid polyimide overlapped layers in the insulation. The level of Lorentz forces, together with the large superconducting coil sizes, makes high-voltage tests impossible to realize under exact operational conditions. The fact that insulation is very difficult or even often practically impossible to repair in case of electrical fault, requires sound insulation design to reach utmost quality. In this paper, we will discuss the build-up of the electrical insulation in large superconducting magnets both from the manufacturing and operational point of view. In particular, the impact of the insulation design and manufacturing choices on the quality will be discussed in depth. We will also discuss the implementation and the application of insulation systems.

The Electrical Insulation of Large Superconducting Magnets

Due to an increasing demand for new, or the refurbishment of power stations, high-voltage rotating machines continue to play a significant role in generating electrical energy. An important component of these machines is their electrical insulation. Although much research has been done to improve the material, about a quarter of all failures still are related to insulation problems. Research to improve winding insulation is very complex because a large number of factors influence insulation life, among them the raw materials, size of the mica tapes, geometrical characteristics of the bar and taping- or manufacturing techniques. The results show that electrical tree propagation is the main electrical degradation mechanism that leads to breakdown of the main wall insulation and that poor impregnation of the insulation and type of taping can reduce time to breakdown significantly.

R. Brutsh

Papers

Electrical breakdown in high-voltage winding insulations of different manufacturing qualities