Développement d'un imageur à rayons X durci pour l'environnement radiatif du Laser Mégajoule

Abstract : Thermonuclear fusion experiments are led on megajoule class laser facility by imploding a capsule filled with Deuterium and Tritium. In this context, it is necessary to diagnose the core size and the shape of the compressed target in order to provide valuable information and identify reasons for failure. State of the art X-ray imaging diagnostics cannot realize measurments without being perturbated by the nuclear background. The diagnostic that has been designed in this thesis combine high spatial resolution X-ray imaging at high energy and radiation tolerance to nuclear background. We have first guaranteed, theoretically and experimentally, survivability of X ray multilayer coating to energetic neutrons irradiation. Consequently, we have design the X-ray imaging system in order to achieve 5 µm resolution in a spectral range up to 95 keV. The X-ray image has then been converted into visible light in order to be easily transfered throught a hardened optical relay to a protected area where the optical analyser is set. This analyser, combining light amplifier and pixelised detector, has also been studied and a novel method has been developped to reduce nuclear related transient perturbations on the device. This by parts design associated with Monte-Carlo Simulation (GEANT4) and experimental campaign on FCI facility (OMEGA) led to a coherent diagnostic architecture wich will sustain high level of nuclear perturbation.
Complete list of metadatas

https://pastel.archives-ouvertes.fr/pastel-00980810
Contributor : Adrien Rousseau <>
Submitted on : Tuesday, April 22, 2014 - 3:51:06 PM
Last modification on : Wednesday, May 15, 2019 - 3:57:14 AM
Long-term archiving on : Monday, April 10, 2017 - 3:34:11 PM

Identifiers

  • HAL Id : pastel-00980810, version 1

Citation

Adrien Rousseau. Développement d'un imageur à rayons X durci pour l'environnement radiatif du Laser Mégajoule. Physique des plasmas [physics.plasm-ph]. Ecole Polytechnique X, 2014. Français. ⟨pastel-00980810⟩

Share

Metrics

Record views

608

Files downloads

295