Publication Date

2016

Document Type

Dissertation/Thesis

First Advisor

Erdelyi, Bela

Degree Name

Ph.D. (Doctor of Philosophy)

Legacy Department

Department of Physics

Abstract

Thomas Jefferson National Accelerator Facility (JLab) has proposed a new Electron-Ion Collider, JLEIC. In this collider, a polarized electron beam and a counter rotating ionbeam collide at the interaction point(s). A critical problem for the JLEIC collider is cooling the ion beam to ensure small emittance and to achieve high luminosity. Since electron cooling—a method of cooling ‘hot’ ion beams through Coulomb interactions with ‘cold’ electron beams—is one of the most effective cooling methods, it will be used by JLEIC. However, the most naive way of calculating Coulomb forces through the pair-wise method becomes infeasible even with the most high performing computers since the computational complexity grows O(N2), where N is the number of particles as large as 10¹¹. In this dissertation, we have developed new computational tools and a high performance computer code that allows, for the first time, a particle-based simulation of realistic electron cooling scenarios of heavy ion beams. Our toolset, collectively referred to as the Particles High-Order Adaptive Dynamics (PHAD), contains three specific tools. The first tool, the adaptive multi-level Fast Multipole Method, reduces the computational cost of computing Coulomb forces to only O(N). Our platform supports particles of any complex distribution (2D or 3D). The second tool, the Picard iteration-based integrator, resolves close encounters of particles efficiently and accurately. Finally, the third tool, the Strang operator splitting, reduces the runtime while maintaining the accuracy. The high performance code is comprised of these three main components. Although, the proposed toolset is both precise and fast, completely simulating the electron cooling of the ion beam still takes a long time on a modern computer cluster due to the millions of small time steps that needs to be simulated. In order to overcome this challenge, we have developed an MPI-parallel high performance computer code to speed up our simulations. The results gathered to date are for over one million time steps, which is less than 1% of the cooling time.

Comments

Includes tables, figures, and algorithms.

Extent

204 pages

Language

eng

Publisher

Northern Illinois University

Rights Statement

In Copyright

Rights Statement 2

NIU theses are protected by copyright. They may be viewed from Huskie Commons for any purpose, but reproduction or distribution in any format is prohibited without the written permission of the authors.

Media Type

Text

Download

Share

COinS