Abstract:Author(s): Wang, Tao | Advisor(s): Arefiev, Alexey | Abstract: With the advent of petawatt-class laser facilities, laser intensities reach unprecedented levels enabling novel and efficient regimes of secondary particle and radiation beams. A regime involving relativistic transparency of dense plasmas and the generation of a Megatesla-level azimuthal magnetic field is shown to be promising for generating energetic electrons and collimated gamma-ray beams. This dissertation focuses on the above regime to explore the acceleration mechanism of electrons, to optimize the gamma-ray yield, to examine the application to two-photon pair production, and to investigate the feasibility of magnetic field detection.First, we investigate direct laser acceleration in the presence of a strong azimuthal magnetic field. Test-particle models are built to explain the enhanced acceleration. The magnetic fields mitigate electron dephasing and allow an efficient acceleration over a short distance. We then report the important role of the laser phase velocity on electron confinement in this acceleration regime. We investigate the emission of collimated gamma-ray beams from laser-irradiated channel targets through three-dimensional kinetic simulations. We find a strong power scaling of conversion efficiency into MeV-level photons. The electron-positron pair production via two-photon collisions directly benefits from such a power scaling. We explore two schemes of generating pairs through the linear Breit-Wheeler process: colliding two gamma-ray beams and colliding one gamma-ray beam with blackbody radiation. The strong power scaling boosts the pair yield to the level of 100 000. Our research on the hollow-channel regime corroborates the robustness of prefilled channels. Due to the influence of ion motion, electrons acceleration and photon emission degrade in hollow channels. With a broader angular spread of gamma-ray beams, the pair yield in hollow channels is shown to be less efficient.At last, we examine the feasibility of detecting Megatesla-level magnetic fields. We choose XFEL beams to detect magnetic fields, based on the magnetic field inducing a polarization rotation via the Faraday effect. A setup of structured targets with a prefilled channel which mitigates the reduction of rotation caused by relativistic transparency is necessary to achieve rotations that exceed 0.1 mrad. A study on laser focusing configurations suggests there is flexibility regarding laser intensities.

Machine learning-based analysis of experimental electron beams and gamma energy distributions

Reconstructing Gamma-ray Energy Distributions from PEDRO Pair Spectrometer Data

Accurate Prediction of Mega-Electron-volt Electron Beam Properties from UED Using Machine Learning

Opportunities in Machine Learning for Particle Accelerators

Machine learning enabling high-throughput and remote operations at large-scale user facilities

Harnessing Machine Learning for Single-Shot Measurement of Free Electron Laser Pulse Power

Adaptive Machine Learning for Time-Varying Systems: Towards 6D Phase Space Diagnostics of Short Intense Charged Particle Beams

Machine learning for analysis of plasma driven Ion source

Automatic setup of 18 MeV electron beamline using machine learning

Machine Learning Applications in Large Particle Accelerator Facilities: Review and Prospects

Kinetic Modelling of Enhanced Electron Acceleration and Gamma-Ray Emission in High-Power Laser Interactions with Structured Targets

Applying Machine Learning Methods to Laser Acceleration of Protons: Lessons Learned from Synthetic Data

Linac_Gen: integrating machine learning and particle-in-cell methods for enhanced beam dynamics at Fermilab

Initial Application of Machine Learning for Beam Parameter Optimization at the Hefei Light Source II

Machine-learning approach for operating electron beam at KEK electron/positron injector linac

Using Machine Learning to Speed Up and Improve Calorimeter R&D

Mixed Diagnostics for Longitudinal Properties of Electron Bunches in a Free-Electron Laser

A Machine Learning Method for Monte Carlo Calculations of Radiative Processes

Feature-Based Analysis of Plasma-Based Particle Acceleration Data

Machine learning surrogate for charged particle beam dynamics with space charge based on a recurrent neural network with aleatoric uncertainty

Efficient 6-dimensional phase space reconstruction from experimental measurements using generative machine learning