Chuang Ren

OSIRIS: A Three-Dimensional, Fully Relativistic Particle in Cell Code for Modeling Plasma Based Accelerators

We describe OSIRIS, a three-dimensional, relativistic, massively parallel, object oriented particle-in-cell code for modeling plasma based accelerators. Developed in Fortran 90, the code runs on multiple platforms (Cray T3E, IBM SP, Mac clusters) and can be easily ported to new ones. Details on the codes capabilities are given. We discuss the object-oriented design of the code, the encapsulation of system dependent code and the parallelization of the algorithms involved. We also discuss the implementation of communications as a boundary condition problem and other key characteristics of the code, such as the moving window, open-space and thermal bath boundaries, arbitrary domain decomposition, 2D (cartesian and cylindric) and 3D simulation modes, electron sub-cycling, energy conservation and particle and field diagnostics. Finally results from three-dimensional simulations of particle and laser wakefield accelerators are presented, in connection with the data analysis and visualization infrastructure developed to post-process the scalar and vector results from PIC simulations.

A global simulation for laser-driven MeV electrons in 50-μm-diameter fast ignition targetsa)

The results from 2.5-dimensional particle-in-cell simulations for the interaction of a picosecond-long ignition laser pulse with a plasma pellet of 50-μm diameter and 40 critical density are presented. The high-density pellet is surrounded by an underdense corona and is isolated by a vacuum region from the simulation box boundary. The laser pulse is shown to filament and create density channels on the laser-plasma interface. The density channels increase the laser absorption efficiency and help generate an energetic electron distribution with a large angular spread. The combined distribution of the forward-going energetic electrons and the induced return electrons is marginally unstable to the current filament instability. The ions play an important role in neutralizing the space charges induced by the temperature disparity between different electron groups. No global coalescing of the current filaments resulted from the instability is observed, consistent with the observed large angular spread of the ene...

Enhanced stopping of macro-particles in particle-in-cell simulations

We derive an equation for energy transfer from relativistic charged particles to a cold background plasma appropriate for finite-size particles that are used in particle-in-cell simulation codes. Expressions for one-, two-, and three-dimensional particles are presented, with special attention given to the two-dimensional case. This energy transfer is due to the electric field of the wake set up in the background plasma by the relativistic particle. The enhanced stopping is dependent on the q2/m, where q is the charge and m is the mass of the relativistic particle, and therefore simulation macro-particles with large charge but identical q/m will stop more rapidly. The stopping power also depends on the effective particle shape of the macro-particle. These conclusions are verified in particle-in-cell simulations. We present 2D simulations of test particles, relaxation of high-energy tails, and integrated fast ignition simulations showing that the enhanced drag on macro-particles may adversely affect the res...

QuickPIC: a parallelized quasi-static PIC code for modeling plasma wakefield acceleration

There has been much recent interest in plasma wakefield acceleration. This is partly due to the possibility of using it as an energy doubler, i.e., an afterburner, stage at the end of an existing linear collider such as SLC. The process in this scheme is highly nonlinear and therefore particle models are required to study it. Furthermore, the key physics involves fully three-dimensional effects. Unfortunately, even on the largest computers it is still not possible to model a full energy doubler stage using existing codes such as OSIRIS. Fortunately however, for these cases the drive beam evolves on a much longer time scale than the plasma frequency. In these cases the beam appears static or frozen for long periods of time. Under these conditions one can make the quasi-static or frozen field approximation. We have recently developed a skeleton version of a parallelized quasi-static PIC code for modeling particle beam drivers. This code combines all the best features from WAKE and the work of D.H. Whittum (1997); and it is fully parallelized. We describe the basic equations and the algorithm. We will also present preliminary results, which include benchmarking it against our fully explicit code OSIRIS.