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
LLNL, Livermore, California
SLAC, Menlo Park, California
Funding: This work was supported by the Office of Science, U. S. Department of Energy, under Contract No. DE-AC02-05CH11231.
A new set of resonances for electron cloud dynamics in the presence of a magnetic field has been found. For short beam bunch lengths and low magnetic fields where lb << 2*π/ωc, (lb = bunch duration, ωc = non-relativistic cyclotron frequency) resonances between the bunch frequency and harmonics of the cyclotron frequency cause an increase in the electron cloud density in narrow ranges of magnetic field near the resonances. For ILC parameters the increase in the density is up to a factor of approximately 3, and the spatial distribution of the electrons is broader near resonances, lacking the well-defined density "stripes" of multipactoring found for non-resonant cases. Simulations with the 2D computer code POSINST, as well as a single-particle tracking code, were used to elucidate the physics of the dynamics. The resonances are expected to affect the electron cloud dynamics in the fringe fields of conventional lattice magnets and in wigglers, where the magnetic fields are low. Results of the simulations, the reason for the bunch-length dependence, and details of the dynamics will be discussed.
C.M. Celata is presently also a visitor in Physics, Mathematics, and Astronomy at California Institute of Technology.
LBNL, Berkeley, California
Funding: This work was supported under the auspices of the Office of Science, U.S. DOE under Contract No. DE-AC02-05CH11231.
Recently* it has been pointed out that numerical simulation of some systems containing charged particles with highly relativistic directed motion can by speeded up by orders of magnitude by choice of the proper Lorentz boosted frame. A particularly good example is that of short wavelength free-electron lasers (FELs) in which a high energy electron beam interacts with a static magnetic undulator. In the optimal boost frame with Lorentz factor gammaF, the red-shifted FEL radiation and blue shifted undulator have identical wavelengths and the number of required time-steps (presuming the Courant condition applies) decreases by a factor of gammaF-squared for fully electromagnetic simulation. We have adapted the WARP code** to apply this method to several FEL problems including coherent spontaneous emission (CSE) from pre-bunched e-beams, radiation in multi-wavelength undulators, and the effective lengths of undulators with entrance and exit matching ramps. We also discuss some preliminary results from applying the boosted-frame method to Coherent Synchrotron Radiation calculations.
*J.-L Vay Phys. Rev. Lett. 98, 130405 (2007)
**D.P. Grote, A. Friedman, J.-L. Vay, and I. Haber, AIP Conf. Proc., 749, 55 (2005).
LBNL, Berkeley, California
LLNL, Livermore, California
Funding: Supported by the US-DOE under Contracts DE-AC02-05CH11231 and DE-AC52-07NA27344, and a DOD SBIR Phase II. Used resources of NERSC, supported by the US-DOE under Contract DE-AC02-05CH11231.
The development of advanced accelerators often involves the modeling of systems that involve a wide range of scales in space and/or time, which can render such modeling extremely challenging. The Adaptive Mesh Refinement technique can be used to significantly reduce the requirements for computer memory and the number of operations. Its application to the fully self-consistent modeling of beams and plasmas is especially challenging, due to properties of the Vlasov-Maxwell system of equations. Most recently, we have begun to explore the application of AMR to the modeling of laser plasma wakefield accelerators (LWFA). For the simulation of a 10GeV LWFA stage, the wake wavelength is O[100μm] while the electron bunch and laser wavelength are typically submicron in size. As a result, the resolution required for different parts of the problem may vary by more than two orders of magnitude in each direction, corresponding to up to 6 orders of magnitude of possible (theoretical) savings by use of mesh refinement. We present a summary of the main issues and their mitigations, as well as examples of application in the context of LWFA and similar beam-plasma interaction setup.
SLAC, Menlo Park, California
LBNL, Berkeley, California
BNL, Upton, Long Island, New York
CERN, Geneva
Funding: Work supported by Department of Energy contract DE–AC03–76SF00515 and the US LARP program.
The SPS at high intensities exhibits transverse single-bunch instabilities with signatures consistent with an Ecloud driven instability. While the SPS has a coupled-bunch transverse feedback system, control of Ecloud-driven motion requires a much wider control bandwidth capable of sensing and controlling motion within each bunched beam. This paper draws beam dynamics data from the measurements and simulations of this SPS instability, and develops initial performance requirements for a feedback system with 2-4 GS/sec sampling rates to damp Ecloud-driven transverse motion in the SPS at intensities desired for high-current LHC operation. Requirements for pickups, kickers and signal processing architectures are presented. Initial lab measurements of proof-of-principle lab model prototypes are presented for the wideband kicker driver signal functions.
LBNL, Berkeley, California
BNL, Upton, Long Island, New York
SLAC, Menlo Park, California
CERN, Geneva
Funding: Supported by the US-DOE under Contract DE-AC02-05CH11231 and the US-LHC LARP. Used resources of NERSC, supported by the US-DOE under Contract DE-AC02-05CH11231.
Electron clouds impose limitations on current accelerators that may be more severe for future machines, unless adequate measures of mitigation are taken. Recently, it has been proposed to use feedback systems operating at high frequency (in the GHz range) to damp single-bunch transverse coherent oscillations that may otherwise be amplified during the interaction of the beam with ambient electron clouds. We have used the simulation package WARP-POSINST to study the growth rate and frequency patterns in space-time of the electron cloud driven beam breakup instability in the CERN SPS accelerator with, or without, an idealized feedback model for damping the instability. We will present our latest results and discuss their implications for the design of the actual feedback system.
LBNL, Berkeley, California
LLNL, Livermore, California
FZK, Karlsruhe
Funding: Supported by the US-DOE under Contract DE-AC02-05CH11231, the US-LHC LARP, and the US-DOE SciDAC program ComPASS. Used resources of NERSC, supported by the US-DOE under Contract DE-AC02-05CH11231.
At PAC05, we presented the package WARP-POSINST for the modeling of the effect of electron clouds on high-energy beams. We present here the latest developments in the package. Three new modes of operations were implemented: 1) “build-up mode” where, similarly to Posinst (LBNL) or Ecloud (CERN), the build-up of electron clouds is modeled in one region of an accelerator driven by a legislated bunch train; 2) “quasi-static mode” where, similarly to Headtail (CERN) or Quickpic (USC/UCLA), the “frozen beam” approximation is used to split the modeling of the beam and the electrons into two components evolving on their respective time scales; and 3) “Lorentz boosted mode” where the simulation is performed into a moving frame where the space and time scales related to the beam and electron dynamics fall in the same range. The implementation of modes (1) and (2) was primary motivated by the need for benchmarking with other codes, while the implementation of mode (3) fulfills the drive toward fully self-consistent simulations of e-cloud effect on the beam including the build-up phase. We also present benchmarking with other codes and selected results from its application to e-cloud effects.