The purpose of the project described in this paper has been to complete construction and to determine the operating characteristics of a charging system and a corona-type controlled leakage system for the compact Van de Graaf generator now under construction at North Texas State University. Calibration data of the generating voltmeter to be used for determining the potential developed by the generator are also included in the paper.

Siemens Corporate Technology New Technology Fields Healthcare & Technology Concepts (CT NTF HTC) have proposed a novel electrostatic accelerator for nuclear medicine which aims at gradients of up to 10 MV m^-1. With beam currents of 100 μA at ≈10 MeV it might replace cyclotrons whilst being simpler, more reliable and more cost effective. The accelerator concept consists of concentric hemispherical metallic shells spaced by insulators and placed in a vacuum system. The shells are interconnected by high voltage diodes so that they form a voltage multiplier with its highest voltage in its centre. Particle beams can be accelerated towards the centre through a set of holes in the shells. In tandem mode, with a stripper in the centre and a negative ion source as injector, beams of twice the centre voltage can be achieved. This thesis presents several commissioning milestones of a test system for the novel electrostatic accelerator, thus validating the concept for commercial applications. An inter shell insulator has been designed and successfully tested to fields of 12 MV m^-1. A diode protection concept has been devised and validated in realistic breakdown scenarios. An AC drive system including control software has been developed, delivering a sinusoidal input voltage of up to 140 kV peak to peak at 80 kHz. An automatic process to carefully commission the high voltage system in vacuum has been created, implemented in a control system and successfully operated. A 4-shell prototype with these components has been successfully tested with achieved gradients of up to 5.5 MV m^-1. A negative hydrogen ion source has been constructed, commissioned and characterised with a purposely developed wire grid. Beam currents beyond 200 μA have been achieved. Beam transport from the ion source through the 7-shell system has been demonstrated in simulations which are based on experimental data from the ion source characterisation. A stripper system has been designed and constructed.

La réponse d’un détecteur de neutrons varie avec l’énergie du neutron incident. La détermination expérimentale de cette variation se réalise au moyen de champs neutroniques monoénergétiques. Ceux-ci sont produits par l’interaction entre un faisceau d’ions accélérés et une cible fine constituée d’un dépôt réactif sur un support métallique. En utilisant différentes réactions telles que 7Li(p,n), 3H(p,n), 2H(d,n) et 3H(d,n), il est possible de produire des neutrons entre 120 keV et 20 MeV dans la direction du faisceau incident (0°).Pour atteindre des énergies inférieures, il est possible d’augmenter l’angle du point de mesure par rapport à la direction du faisceau d’ions. Cependant, cette méthode présente des problèmes d’homogénéité en énergie et en fluence des neutrons à la surface du détecteur, ainsi qu’une augmentation de la proportion de neutrons diffusés. Une alternative est l’utilisation d’autres réactions nucléaires, notamment la réaction 45Sc(p,n) qui permet de descendre jusqu’à des énergies de 8 keV à 0°.Une étude complète de cette réaction et de sa section efficace a été menée au sein d’une coopération scientifique entre le laboratoire de métrologie et de dosimétrie des neutrons (LMDN) de l’IRSN, deux instituts de métrologie européens, le NPL (National Physical Laboratory, RU) et le PTB (Physikalisch-Technische Bundesanstalt, All), et l’IRMM (Institute for Reference Materials and Measurements, CEE). Parallèlement, d’autres réactions envisageables ont été étudiées : 65Cu(p,n), 51V(p,n), 57Fe(p,n), 49Ti(p,n), 53Cr(p,n) et 37Cl(p,n). Elles ont été comparées en termes d’émission neutronique et d’énergie minimale des neutrons produits. / The response of a neutron detector, defined as the reading of the device per unit of incident fluence or dose, varies with neutron energy. The experimental determination of this variation, i.e. of the response function of this instrument, has to be performed by facilities producing monoenergetic neutron fields. These neutrons are commonly produced by interaction between accelerated ions (proton or deuteron) onto a thin target composed of a reactive layer deposited on a metallic backing. Using the 7Li(p,n), 3H(p,n), 2H(d,n) and 3H(d,n) reactions, monoenergetic neutrons are obtained between 120 keV and 20 MeV in the ion beam direction (0°).To reach lower neutron energies, the angle of the measuring point with respect to the ion beam direction can be increased. However, this method presents several problems of neutron energy and fluence homogeneities over the detector surface, as well as an important increase of the scattered neutron contribution. Another solution is to investigate other nuclear reactions, as 45Sc(p,n) allowing to extend the neutron energy range down to 8 keV at 0°.A complete study of this reaction and its cross section has been undertaken within the framework of a scientific cooperation between the laboratory of neutron metrology and dosimetry (IRSN, France), two European national metrological institutes, the National Physical Laboratory (UK) and the Physikalisch-Technische Bundesanstalt (Germany), and IRMM, the Institute for Reference Materials and Measurements (EC).In parallel, other possible reactions have been investigated: 65Cu(p,n), 51V(p,n), 57Fe(p,n), 49Ti(p,n), 53Cr(p,n) and 37Cl(p,n). They were compared in terms of neutron fluence and minimum energy of the produced neutrons.