No description available.

No description available.

No description available.

Stromrichtertransformatoren der Hochspannungsgleichstromübertragung bilden das Bindeglied zwischen Gleichspannungs- und Drehstromsystem. Um den ausfallsicheren Betrieb über die gesamte Lebensdauer zu gewährleisten, muss deren Öl-Papier-Isoliersystem entsprechend dimensioniert werden. Eine optimale Dimensionierung setzt ein detailliertes Verständnis über die Beanspruchung des Isoliersystems sowie deren zuverlässige Modellierung sowohl unter Betriebsspannung als auch bei den überlagerten, transienten Überspannungen voraus. Im Rahmen dieser Arbeit wird daher das dielektrische Verhalten des Öl-Papier-Isoliersystems in Anlehnung an dielektrische Prüfungen sowohl unter Gleichspannungsbelastung als auch einer zusammengesetzten Spannungsbelastung aus einer Gleich- und einer Blitzstoßspannung (einer sog. Mischspannungsbelastung) untersucht. Der Vergleich von numerischen Berechnungen auf Grundlage eines ladungsträgerbasierten Ansatzes nach Poisson-Nernst-Planck (PNP) mit Durchschlagexperimenten gibt dabei Aufschluss über die Beanspruchung des Öl-Papier-Isoliersystems. Weiterhin wird gezeigt, dass der in den etablierten, resistiv-kapazitiven Berechnungsmodellen vernachlässigte Ladungsträgereinfluss in Bezug auf die Beanspruchung des Isoliersystems unzureichende Ergebnisse zur Folge hat und demnach zwingend zu berücksichtigen ist. Die an realitätsnahen, Öl-Papier-isolierten Anordnungen erzielten Ergebnisse zeigen nicht nur den Einfluss der an Grenzflächen oder im Papier akkumulierten Ladungsträger auf die Beanspruchung des Isoliersystems. Ebenso werden die Annahmen des ladungsträgerbasierten Ansatzes und die Berechnungsergebnisse des PNP-Modells qualitativ bestätigt. Infolge der Ladungsakkumulation im Papier tritt die höchste Beanspruchung im Ölspalt und nicht im Papier auf. Öl-Papier-isolierte Anordnungen werden somit geringer beansprucht, als eine Strömungsfeldberechnung vermuten lässt. Dies widerspricht den Annahmen der etablierten Berechnungsmodelle und wird im Weiteren durch Polaritätseffekte an homogenen, aber unsymmetrischen, papierisolierten Elektrodenanordnungen oder durch den nachweisbaren Einfluss des Ölvolumens im Prüfgefäß auf die Beanspruchung einer Anordnung verdeutlicht. Unter Mischspannungsbelastung wird weiterhin gezeigt, dass eine Überlagerung der Gleichspannung und damit auch der Polaritätswechsel keine höhere Beanspruchung des Isoliersystems im Vergleich zur reinen Gleichspannungsbelastung zur Folge hat. Die etablierten, resistiv-kapazitiven Modelle ließen jedoch den Polaritätswechsel als kritischste Beanspruchung vermuten. Somit wird nicht nur die Anwendbarkeit der ladungsträgerbasierten PNP-Modellierung an Öl-Papier-Isolieranordnungen qualitativ verifiziert. Ebenso wird demonstriert, dass die stark vereinfachten Annahmen der etablierten Berechnungsmodelle die Beanspruchungen unter Gleich- und der untersuchten Mischspannungsbelastung nicht abbilden können. Der Einsatz klassischer Strömungsfeldberechnungen zur Nachbildung der Beanspruchung des Öl-Papier-Isoliersystems unter Gleichspannungsbelastung entspricht damit nicht mehr dem Stand der Forschung. / Converter transformers of HVDC transmission systems connect HVDC and HVAC systems. To ensure a reliable operation during the entire lifetime, their oil-paper-insulation system must be designed appropriately. An optimized dielectric design demands a fundamental understanding of the dielectric stresses as well as a reliable modeling of the insulation system both under operating voltages and under superimposed, transient overvoltages. Hence, in this work the dielectric behavior of the oil-paper-insulation system is investigated. Based on dielectric tests the investigations are performed under DC voltage stress and a composite voltage stress of a DC voltage in stationary conditions superimposed by a lightning impulse voltage. The comparison of numerical calculations using a charge-carrier-based approach according to Poisson-Nernst-Planck (PNP) with breakdown experiments clarifies the dielectric stress of the oil-paper-insulation system. Furthermore, the comparison with results determined by the established, resistive-capacitive calculation models shows that it is mandatory to take the influence of the charge carrier accumulation into account. The presented results, which were obtained at oil-paper-insulated arrangements which represent the dielectic stress of real arrangements, show the influence of the charge carriers accumulating at interfaces or in the paper insulationon on the dielectric stress. The results confirm the calculations and the assumptions according to the charge-carrier-based model as well. Due to the charge carrier accumulation, the highest dielectric stress occurs in the mineral oil and not in the paper insulation. In contrast, the findings obtained assuming an ohmic conductivity would results in a higher dielectric stress of the oil-paperinsulated arrangements. Furthermore, polarity effects in homogeneous but asymmetrical, paper-insulated electrode arrangements or the influence of the surrounding oil in the test vessel demonstrate the effects of the charge carriers. Under composite voltage stresses it is also shown, that the applied superimposed voltage as well as the fast polarity reversal does not lead to a higher dielectric stress of the arrangements compared to the pure DC voltage stress. Commonly used calculation models would determine higher stresses due to the fast polarity reversal instead. Consequently, the applicability of the charge-carrier-based PNP calculation model is verified qualitatively in the presented investigations. Furthermore, it is demonstrated that the simplified assumptions of the commonly used calculation models cannot simulate the dielectric stresses under DC voltage stress and under the investigated superimposed voltage stresses. Hence, the determination of the dielectric stresses of oil-paper-insulation systems under DC voltage stress according to the commonly used calculation models assuming an ohmic conductivity does not correspond to the current state of research.

No description available.

The Antimatter Experiment: Gravity, Interferometry, Spectroscopy (AEGIS) at the European Organization for Nuclear Research (CERN) is an experiment investigating the influence of earth’s gravitational force upon antimatter. To perform precise measurements the antimatter needs to be cooled to a temperature of 100 mK. This will be done in a Penning trap, formed by several electrodes, which are charged with several kV and have to be individually electrically insulated. The trap is thermally linked to a mixing chamber of a 3He-4He dilution refrigerator. Two link designs are examined, the Rod design and the Sandwich design. The Rod design electrically connects a single electrode with a heat exchanger, immersed in the helium of the mixing chamber, by a copper pin. An alumina ring and the helium electrically insulate the Rod design. The Sandwich uses an electrically insulating sapphire plate sandwiched between the electrode and the mixing chamber. Indium layers on the sapphire plate are applied to improve the thermal contact. Four differently prepared test Sandwiches are investigated. They differ in the sapphire surface roughness and in the application method of the indium layers. Measurements with static and sinusoidal heat loads are performed to uncover the behavior of the thermal boundary resistances. The thermal total resistance of the best Sandwich shows a temperature dependency of T-2,64 and is significantly lower, with roughly 30 cm2K4/W at 50 mK, than experimental data found in the literature. The estimated thermal boundary resistance between indium and sapphire agrees very well with the value of the acoustic mismatch theory at low temperatures. In both designs, homemade heat exchangers are integrated to transfer the heat to the cold helium. These heat exchangers are based on sintered structures to increase the heat transferring surface and to overcome the significant influence of the thermal resistance (Kapitza resistance). The heat exchangers are optimized concerning the adherence of the sinter to the substrate and its sinter height, e.g. its thermal penetration length. Ruthenium oxide metallic resistors (RuO2) are used as temperature sensors for the investigations. They consist of various materials, which affect the reproducibility. The sensor conditioning and the resulting good reproducibility is discussed as well.

The Antimatter Experiment: Gravity, Interferometry, Spectroscopy (AEGIS) at the European Organization for Nuclear Research (CERN) is an experiment investigating the influence of earth’s gravitational force upon antimatter. To perform precise measurements the antimatter needs to be cooled to a temperature of 100 mK. This will be done in a Penning trap, formed by several electrodes, which are charged with several kV and have to be individually electrically insulated. The trap is thermally linked to a mixing chamber of a 3He-4He dilution refrigerator. Two link designs are examined, the Rod design and the Sandwich design. The Rod design electrically connects a single electrode with a heat exchanger, immersed in the helium of the mixing chamber, by a copper pin. An alumina ring and the helium electrically insulate the Rod design. The Sandwich uses an electrically insulating sapphire plate sandwiched between the electrode and the mixing chamber. Indium layers on the sapphire plate are applied to improve the thermal contact. Four differently prepared test Sandwiches are investigated. They differ in the sapphire surface roughness and in the application method of the indium layers. Measurements with static and sinusoidal heat loads are performed to uncover the behavior of the thermal boundary resistances. The thermal total resistance of the best Sandwich shows a temperature dependency of T-2,64 and is significantly lower, with roughly 30 cm2K4/W at 50 mK, than experimental data found in the literature. The estimated thermal boundary resistance between indium and sapphire agrees very well with the value of the acoustic mismatch theory at low temperatures. In both designs, homemade heat exchangers are integrated to transfer the heat to the cold helium. These heat exchangers are based on sintered structures to increase the heat transferring surface and to overcome the significant influence of the thermal resistance (Kapitza resistance). The heat exchangers are optimized concerning the adherence of the sinter to the substrate and its sinter height, e.g. its thermal penetration length. Ruthenium oxide metallic resistors (RuO2) are used as temperature sensors for the investigations. They consist of various materials, which affect the reproducibility. The sensor conditioning and the resulting good reproducibility is discussed as well.