UCLA

Plasma-based acceleration (PBA) has emerged as a promising candidate for the accelerator technology used to build a future linear collider and/or an advanced light source. In PBA, an intense laser (LWFA) or a particle beam (PWFA) driver is sent through a tenuous plasma, and a trailing beam surfs on the wake behind the driver. For a linear collider, a critical metric is the luminosity which measures the collision rate. Thus, achieving high luminosity with low cost and high efficiency is important. This requires high energy transfer efficiency from the drive beam to the wake and subsequently to the trailing beam, all while maintaining low emittance and minimal energy spread in the trailing bunch. In addition, in order to achieve high luminosity it is important to be able to focus beams to very small spots sizes before they collide.In this dissertation several key aspects of a PWFA based linear collider are investigated, with a particular focus on applying advanced computational techniques to optimize and design crucial PWFA and beam delivery stages. This is accomplished by developing a workflow between the quasi-static particle-in-cell code QuickPIC and a nonlinear least-squares optimization solver and an adaptive mesh refinement algorithm for QuickPIC.Using the optimization workflow, the optimum drive and witness beam current profiles for maximizing the energy transfer efficiency between drive and witness beams in nonlinear regimes is found. For the witness beam the bunch length and its position within the wakefield are held fixed, while for the drive beam the total beam charge is held constant. It is found that the current profile of a fixed-length witness beam that minimizes variations in the accelerating field around its mean closely matches the prediction of a recent analytical model. For a drive beam of fixed charge and length, it is found that the current profile that minimizes variations of the decelerating field about its mean is very similar to that predicted by linear theory. It includes a short precursor followed by a current profile that ramps linearly to a peak value. The profiles are also compared to the predictions from a phenomenological nonlinear wakefield theory. By examining other current profiles it is shown that these profiles also maximize the efficiency in the nonlinear regime.Modeling PWFA for linear collider parameters is daunting due to large disparity in scales between the size of the wakefield and the matched spot size of the witness beam. To address this a mesh refinement algorithm for the three dimensional quasi-static PIC code QuickPIC is developed and described. Very fine (high) resolution is used in a small spatial region that includes the witness beam and progressively coarser resolutions are used in the rest of the simulation domain. A fast multigrid Poisson solver is implemented for the field solve on the refined meshes and a Fast Fourier Transform (FFT) based Poisson solver is used for the coarse mesh. Solutions on each successively finer mesh are computed using Dirichlet boundary conditions derived from the solution at the previous refinement level. The code is parallelized with both MPI and OpenMP, and the parallel scalability has also been improved by using pipelining and redistributing the data from fine mesh domains to other processors. A preliminary adaptive mesh refinement technique is described to reduce the computational time in simulations with an evolving witness beam size. Several benchmark cases are demonstrated to validate simulation accuracy. Energy conservation in the quasi-static algorithm with and without mesh refinement is measured and compared to help validate the mesh refinement algorithm.The AMR framework is used to investigate the design and performance of adiabatic plasma lenses for beam focusing. For the parameters employed, the self-fields of the beam pull plasma ions inward during the transit time of the beam, significantly enhancing the lens’s focusing strength. These lenses offer strong, symmetric focusing fields and have the potential to overcome the synchrotron radiation limit imposed by conventional optics. By incorporating both ion motion and radiation reaction into the simulations, the possibility of nanometer-scale beam focusing with significant luminosity enhancement is demonstrated.