Numerical and experimental study of electron beam dynamics in linear induction accelerators

Flash radiography allows characterizing the state of a dense object subjected to extreme physical conditions, with displacement velocities of several kilometres per second. These conditions require a specific X-ray source: small size (a few millimetres), brief duration (less than 100 ns), high energy (around 20 MeV) and high current (a few kA). This source is produced using Bremsstrahlung radiation generated by the interaction of an intense and pulsed electron beam with a high atomic number metal target. A velvet cathode emits the beam and an linear induction accelerator (LIA) transports it. The quality of radiography is mainly conditioned by the physical characteristics of the X-ray source, which are closely linked to the properties of the electron beam. The work carried out in this PhD thesis aims to model the beam dynamics in LIAs by integrating its physical characteristics, including some instabilities which degrade the beam. The developed models were validated during the commissioning of the MCH3 accelerator at CEA Valduc in Epure facility.The modelling of beam dynamics is based on the Particle-In-Cell (PIC) code LSP-Slice and the transport code EVOLI (Evolution of Instabilities). The latter was developed during this thesis and models the size of the beam envelope and its charge centroid, as well as the 'temporal' propagation of the beam by its segmentation into discs.In the first place, studies were conducted on the motion of the beam centroid during its transport. In EVOLI, the equations describe the centroid with the influence of space charge. A numerical optimization procedure allowed the simulation to reproduce the results obtained in MCH3 by incorporating the influence of the Earth's magnetic field and the misalignment of various magnetic elements (solenoids and steerers). The application of this new method to other beam transports makes possible to calculate steerer settings for numerical pre-centering of the beam. This method offers the prospect of significant time savings during the experimental beam centering process, which typically requires numerous shots.Afterward the beam modelling contributed to the commissioning of MCH3 accelerator at the end of 2022. With an initial transport attempt, the entire beam charge did not reach the conversion target, despite numerical simulations predicting nominal transport based on the envelope formalism. Experimental measurements revealed significant centroid oscillations due to the Beam Break-Up (BBU) instability. Then, a simplified BBU model was incorporated into EVOLI. By using this numerical model, a new high magnetic field transport was designed, theoretically limiting BBU intensity at the accelerator end by a factor 2.5. This result was experimentally verified by a BBU reduction by a factor 3, enabling a stable and reproducible transport of the entire beam charge to the target. However, this strong magnetic field transport strategy leads to an increase of the Corkscrew motion, which increases with magnetic field strength as shown in the work initiated in this thesis. Therefore, optimizing beam transport is a compromise between the various instabilities within the accelerator. The work of this thesis opens up prospects for considering instabilities to design innovative transports, particularly in the context of multi-pulse machines, which are currently a major development in flash radiography machines. (author)