Yadav, Monika
(2023)
Modeling betatron radiation for diagnostics at FACET-II using machine learning algorithms.
Doctor of Philosophy thesis, University of Liverpool.
|
Text
201367654_APR2023.pdf - Author Accepted Manuscript Download (52MB) | Preview |
Abstract
Plasma accelerators provide unique opportunities to generate high-brightness, short-pulse electron beams, which are an ideal basis for high-quality radiation generation. Beam-driven plasma wakefield accelerators (PWFA) have demonstrated significant progress during the past two decades. Facility for Advanced Accelerator Experimental Tests II (FACET-II) at SLAC National Accelerator Laboratory is a state-of-the-art facility that is used to study the properties of particles and their interactions at the highest energies. In order to accomplish the experimental goals, such as reducing energy spread, high energy transfer efficiency from drive to witness beam, and emittance preservation, at FACET-II, sophisticated diagnostics are required to characterize the beam. One important diagnostic tool is betatron radiation which will measure the spectral and angular properties of the incoming photon stream produced by these intense beams in plasma. Radiation diagnostics could also be used to reconstruct the properties of the beam and photon energy distributions. This thesis presents new numerical methods, spectrometer hardware designs and beam parameter and spectrum reconstruction methods for betatron radiation. First, three numerical models are developed for characterizing beam properties using different methods of computing particle trajectories. These models are then used to numerically integrate Liénard–Wiechert (LW) potentials to compute their emitted radiation. Progress is made towards solving the computational challenges of recovering beam information from radiation diagnostics. Using the particle-in-cell (PIC) codes, simulations for planned FACET-II PWFA and Trojan horse experiments are numerically simulated. Some practical aspects of implementing radiation diagnostics using the models mentioned above for generated witness beams are discussed. Second, a photon spectrometer at FACET-II is built at UCLA to measure betatron radiation. Measuring the beam's emitted betatron radiation can be valuable in studying the matching dynamics. This spectrometer will detect high-energy photons from Compton scattering and pair production methods. The design for both spectrometers is presented to record a single-shot gamma spectrum at FACET-II for a full suite of gamma measurements. Third, data from two types of simulated scenarios -- PWFA-derived betatron radiation experiments and high-field laser-electron-based radiation production -- are used to determine which methods could most reliably reconstruct these key properties of the beam. The data from these two scenarios provide a large range of photon energies; this variation of characteristics increases confidence in each analysis method. In both cases, the performance of maximum likelihood estimation (MLE), neural networks, and a hybrid approach combining the two are compared. Furthermore, in the case of higher energy photons, above 30 MeV, where electron-positron production is used as the basis of the spectrometer, the efficacy of matrix decomposition is examined. As such, the ML-MLE hybrid approach proves to be the most generalizable of the methods. Fourth, PIC codes are also used to compare betatron radiation in various scenarios, including strong and weak blowouts, asymmetric beams relevant to experiments at FACET-II, and Argonne Wakefield Accelerator (AWA) facility. Finally, ion collapse experiment is also studied, which aims to study how ion motion in a PWFA has nonlinear focusing effects due to nonuniform ion distributions, is extensively studied.
| Item Type: | Thesis (Doctor of Philosophy) |
|---|---|
| Divisions: | Faculty of Science and Engineering > School of Physical Sciences |
| Depositing User: | Symplectic Admin |
| Date Deposited: | 14 Aug 2023 13:54 |
| Last Modified: | 14 Aug 2023 13:54 |
| DOI: | 10.17638/03169943 |
| Supervisors: |
|
| URI: | https://livrepository.liverpool.ac.uk/id/eprint/3169943 |
Dimensions
Dimensions
