|
C. J. Joshi, R. G. Hemker, F.
S. Tsung, E. S. Dodd, and W. B. Mori,
University of California, Los Angeles
S. Lee and T. Katsouleas, University of Southern California
Research Objectives
This research attempts
to test the feasibility of various plasma-based accelerator concepts,
to model full-scale plasma-based accelerator experiments, and to help
develop new advanced accelerator concepts.
Computational Approach
We are applying particle-based
models, including fully explicit particle-in-cell (PIC) codes, ponderomotive
guiding center PIC codes, and new photon kinetic codes. We are integrating
all these algorithms into an object-oriented framework we have developed
which supports massively parallel processing.
 |
|
These images are isosurface contours of the accelerating electric
field from a 3D PIC simulation of a plasma wakefield accelerator.
The simulation was done on 64 nodes of the T3E at NERSC. It used 14
million grid cells and 56 million particles. In the upper and lower
left, the wake shown was excited by an azimuthally symmetric drive
beam, while for the upper and lower right, the drive beam was asymmetric.
The asymmetry leads to a lower peak amplitude in the wake and in an
asymmetric transverse profile for the accelerating field. The same
color maps were used for each figure. The dark blue, light blue, green,
and yellow surfaces correspond to acceleration gradients of 0.5, 0.4,
0.2, and 0.1 GeV/m, while the red surfaces correspond to a decelerating
gradient of 0.1 GeV/m. The simulation parameters were chosen to accurately
model the E-157 plasma wakefield experiment (a collaboration of SLAC,
UCLA, USC, and LBNL). The figure was rendered with the help of the
Office of Academic Computing at UCLA. |
|
Accomplishments
We have developed
a new fully parallel, multidimensional (2D or 3D) object-oriented PIC
code which is optimized for modeling plasma-based acceleration. The object-oriented
code design allowed us to implement multiple algorithms that can that
be chosen at runtime. The code is also designed to be extendable to advanced
features like dynamic load balancing and adaptive mesh refinement. We
used this code to model 1-meter plasma wakefield stages, 2D and 3D simulations
of plasma wakefield excitation, 2D and 3D simulations of the generation
of single-cycle laser pulse by photon deceleration, and 3D simulations
of Cerenkov radiation from plasma wakes. We also developed a new ponderomotive
guiding center code for efficient modeling of laser-plasma accelerator
stages.
Significance
In plasma-based acceleration,
electrons "surf" on relativistic space charge plasma waves. In such waves,
electrons can be accelerated with gradients orders of magnitude larger than
is possible with current technology. If plasma-based accelerator technology
is successfully developed, then multi-GeV stages could be miniaturized to
fit on a tabletop. Just as the advent of tabletop high-powered lasers have
had a tremendous impact, miniature tabletop accelerators could have cross-cutting
impacts in fields as diverse as high-energy physics, synchrotron radiation
sources, medicine, and biology.
Publications
B. J. Duda, R. G. Hemker, K.-C. Tzeng, and W. B. Mori, "A new long-wavelength
hosing instability for lasers propagating in plasmas," Phys. Rev. Lett.
(in press).
K.-C. Tzeng and W. B. Mori,
"Suppression of electron ponderomotive blowout and relativistic self-focusing
by the occurrence of raman scattering and plasma heating," Phys. Rev.
Lett. 81, 104 (1998).
R. G. Hemker, K.-C. Tzeng,
W. B. Mori, C. E. Clayton, and T. Katsouleas, "Cathodeless, high-brightness
electron beam production by multiple laser-beams in plasmas," Phys. Rev.
E 57, 5920 (1998).
|
