LBNL, Berkeley, California
LLNL, Livermore, California
Funding: Supported by the US DOE at LBNL and LLNL under contracts DE-AC02-05CH11231 and DE-AC52-07NA27344, LARP, SciDAC, and ComPASS. Computuational resources of the NERSC were employed.
It has been shown* that the ratio of longest to shortest space and time scales of a system of two or more components crossing at relativistic velocities is not invariant under Lorentz transformation. This implies the existence of a frame of reference minimizing an aggregate measure of the ratio of space and time scales. It was demonstrated that this translated into a reduction by orders of magnitude in computer simulation run times, using methods based on first principles (e.g., Particle-In-Cell), for particle acceleration devices and for problems such as: free electron laser, laser-plasma accelerator, and particle beams interacting with electron clouds. Since then, speed-ups ranging from 75 to more than four orders of magnitude have been reported for the simulation of either scaled or reduced models of the above-cited problems. In ** it was shown that to achieve full benefits of the calculation in a boosted frame, some of the standard numerical techniques needed to be revised. The theory behind the speed-up of numerical simulation in a boosted frame, latest developments of numerical methods, and example applications with new opportunities that they offer are all presented.
* J.-L. Vay, Phys. Rev. Lett. 98, 130405 (2007).
**J.-L. Vay, Phys. of Plasmas 14, 1 (2008).
LBNL, Berkeley, California
Tech-X, Boulder, Colorado
Funding: Funded by the U.S. DOE Office of Science HEP including contract DE-AC02-05CH11231 and SciDAC, and by U.S. DOE NA-22, DARPA, and NSF
Collider and light source applications of laser wakefield accelerators will likely require staging of controlled injection with multi-GeV accelerator modules to produce and maintain the required low emittance and energy spread. We present simulations of upcoming 10 GeV-class LWFA stages, towards eventual collider modules for both electrons and positrons*. Laser and structure propagation are controlled through a combination of laser channeling and self guiding. Electron beam evolution is controlled through laser pulse and plasma density shaping, and beam loading. This can result in efficient stages which preserve high quality beams. We also present results on controlled injection of electrons into the structure to produce the required low emittance bunches using plasma density gradient** and colliding laser pulses. Tools for accurately modeling emittance and energy spread will be discussed***.
*E. Cormier-Michel et al., Proc. Adv Accel. Workshop 2008.
**C.G.R. Geddes et al., PRL 2008.
***E. Cormier-Michel et al, PRE 2008; C.G.R. Geddes et al, Proc. Adv Accel. Workshop 2008.
Tech-X, Boulder, Colorado
CIPS, Boulder, Colorado
LBNL, Berkeley, California
University of Nevada, Reno, Reno, Nevada
Funding: Supported by the US DOE Office of Science, Office of High Energy Physics under grant No. DE-FC02-07ER41499; used NERSC resources under grant DE-AC02-05CH11231.
Laser wakefield accelerators (LWFA) have accelerated ~100 pC electron bunches to GeV energies over cm scale distances, via self-trapping from the plasma. Self-trapping cannot be tolerated in staged LWFA modules for high-energy physics applications. The ~1% energy spread of self-trapped electron bunches is too large for light source applications. Both difficulties could be resolved via external injection of a low-emittance electron bunch into a quasilinear LWFA, for which the dimensionless laser amplitude is less than two. However, significant beam charge will result in nonlinear beam loading effects, which will make it challenging to preserve the low emittance. The cold, relativistic fluid model of the parallel VORPAL framework* will be used to simulate the laser-driven electron wake, in the presence of an idealized electron beam. Profiles of the electron beam density, laser pulse envelope and plasma channel will be varied to find a nonlinear beam loading configuration that approximately flattens the electric fields across the beam. Hybrid fluid-PIC simulations will be used to measure the self-consistent emittance growth of the beam.
* C. Nieter and J.R Cary, J. Comp. Phys. 196 (2004), p. 448.
Tech-X, Boulder, Colorado
LBNL, Berkeley, California
Funding: Work supported by Department of Energy contracts DE-AC02-05CH11231 (LBNL), DE-FC02-07ER41499 (SciDAC), and DE-FG02-04ER84097 (SBIR).
Simulation of laser wakefield accelerator (LWFA) experiments is computationally highly intensive due to the disparate length scales involved. Current experiments extend hundreds of laser wavelengths transversely and many thousands in the propagation direction, making explicit PIC simulations enormously expensive. We can substantially improve the performance of LWFA simulations by modeling the envelope modulation of the laser field rather than the field itself. This allows for much coarser grids, since we need only resolve the plasma wavelength and not the laser wavelength, and this also allows larger timesteps. Thus an envelope model can result in savings of several orders of magnitude in computational resources. By propagating the laser envelope in a Galilean frame moving at the speed of light, dispersive errors can be avoided and simulations over long distances become possible. Here we describe the model and its implementation. We show rigorous studies of convergence and discretization error, as well as benchmarks against explicit PIC. We also demonstrate efficient, fully 3D simulations of downramp injection and meter-scale acceleration stages.