ORNL, Oak Ridge, Tennessee
UTK, Knoxville, Tennessee
The computational approach is providing essential guidance and analysis for the commissioning and operation of SNS. Computational models are becoming sufficiently realistic that it is now possible to study detailed beam dynamics issues quantitatively. Increasingly, we are seeing that the biggest challenge in performing successful analyses is that of knowing and describing the machine and beam state accurately. Even so, successful benchmarks with both theoretical predictions and experimental results are leading to increased confidence in the capability of these models. With this confidence, computer codes are being employed in a predictive manner to guide the machine operations. We will illustrate these points with various examples taken from the SNS linac and ring.
ORNL, Oak Ridge, Tennessee
The contents, structure, implementation, benchmarking, and applications of ORBIT as an accelerator simulation code are described. Physics approaches, algorithms, and limitations for space charge, impedances, and electron cloud effects are discussed. The ORBIT code is a parallel computer code, and the scalabilities of the implementations of parallel algorithms for different physics modules are shown. ORBIT has a long history of benchmarking with analytical exactly solvable problems and experimental data. The results of this benchmarking and the current usage of ORBIT are presented.
CERN, Geneva
GSI, Darmstadt
CCLRC/RAL/ASTeC, Chilton, Didcot, Oxon
GSI, Darmstadt
PSI, Villigen
Rostock University, Faculty of Engineering, Rostock
DESY, Hamburg
Rostock University, Faculty of Computer Science and Electrical Engineering, Rostock
Precise and fast 3D space-charge calculations for bunches of charged particles are still of growing importance in recent accelerator designs. A widespread approach is the particle-mesh method computing the potential of a bunch in the rest frame by means of Poisson's equation. Recently new algorithms for solving Poisson's equation have been implemented in the tracking code Astra. These Poisson solvers are iterative algorithms solving a linear system of equations that results from the finite difference discretization of the Poisson equation. The implementation is based on the software package MOEVE (Multigrid Poisson Solver for Non-Equidistant Tensor Product Meshes) developed by G. Pöplau. The package contains a state-of-the-art multigrid Poisson solver adapted to space charge calculations. In this paper the basic concept of iterative Poisson solvers is described. It is compared to the established 3D FFT Poisson solver which is a widely-used method for space charge calculations and also implemented in Astra. Advantages and disadvantages are discussed. Further the similarities and differences of both approaches are demonstrated with numerical examples.
TEMF, Darmstadt
For devices in which the bunch dimensions are much smaller than the dimensions of the structure the numerical field solution is typically hampered by spurious oscillations. The reason for this oscillations is the large numerical dispersion error of conventional schemes along the beam axis. Recently, several numerical schemes have been proposed which apply operator splitting to optimize and under certain circumstances eliminate the dispersion error in the direction of the bunch motion. However, in comparison to the standard Yee scheme the methods based on operator splitting do not conserve the standard discrete Gauss law. This contribution is dedicated to the construction of conserved discrete Gauss laws and conservative current interpolation for some of the split operator methods. Finally, the application of the methods in a PIC simulations is shown.
LBNL, Berkeley, California
Tech-X, Boulder, Colorado
PSI, Villigen
Fermilab, Batavia, Illinois
University of Maryland, College Park, Maryland
LANL, Los Alamos, New Mexico
BNL, Upton, Long Island, New York
MaryLie/IMPACT (ML/I) is a 3D parallel Particle-In-Cell code that combines the nonlinear optics capabilities of MaryLie 5.0 with the parallel particle-in-cell space-charge capability of IMPACT. In addition to combining the capabilities of these codes, ML/I has a number of powerful features, including a choice of Poisson solvers, a fifth-order rf cavity model, multiple reference particles for rf cavities, a library of soft-edge magnet models, representation of magnet systems in terms of coil stacks with possibly overlapping fields, and wakefield effects. The code allows for map production, map analysis, particle tracking, and 3D envelope tracking, all within a single, coherent user environment. ML/I has a front end that can read both MaryLie input and MAD lattice descriptions. The code can model beams with or without acceleration, and with or without space charge. Developed under a US DOE Scientific Discovery through Advanced Computing (SciDAC) project, ML/I is well suited to large-scale modeling, simulations having been performed with up to 100M macroparticles. ML/I uses the H5Part* library for parallel I/O. The code inherits the powerful fitting/optimizing capabilities of MaryLie, augmented for the new features of ML/I. The combination of soft-edge magnet models, high-order capability, and fitting/optimization, makes it possible to simultaneously remove third-order aberrations while minimizing fifth-order, in systems with overlapping, realistic magnetic fields. Several applications will be presented, including aberration correction in a magnetic lens for radiography, linac and beamline simulations of an e-cooling system for RHIC, design of a matching section across the transition of a superconducting linac, and space-charge tracking in the damping rings of the International Linear Collider.
*ICAP 2006 paper ID 1222, A. Adelmann et al., "H5Part: A Portable High Performance Parallel Data Interface for Electromagnetics Simulations"
DESY, Hamburg
SLAC, Menlo Park, California
Under the US DOE SciDAC project, SLAC has developed a suite of 3D (2D) Parallel Higher-order Finite Element (FE) codes, T3P (T2P) and PIC3P (PIC2P), aimed at accurate, large-scale simulation of wakefields and particle-field interactions in RF cavities of complex shape. The codes are built on the FE infrastructure that supports SLAC’s frequency domain codes, Omega3P and S3P, to utilize conformal tetrahedral (triangular) meshes, higher-order basis functions and quadratic geometry approximation. For time integration, they adopt an unconditionally stable implicit scheme. PIC3P (PIC2P) extends T3P (T2P) to treat charged particle dynamics self-consistently using the PIC approach, the first such implementation on the FE grid. Examples from applications to the ILC, LCLS and other accelerators will be presented to compare the accuracy and computational efficiency of these codes versus their counterparts using structured grids.
LBNL, Berkeley, California
SLAC, Menlo Park, California
LBNL, Berkeley, California
The IMPACT-T code is a parallel three-dimensional quasi-static beam dynamics code for modeling high brightness beams in photoinjectors and RF linacs. Developed under the US DOE Scientific Discovery through Advanced Computing (SciDAC) program, it includes several key features including a self-consistent calculation of 3D space-charge forces using a shifted and integrated Green function method, multiple energy bins for a beams with large energy spread, and models for treating RF standing wave and traveling wave structures. In this paper, we report on recent improvements to the IMPACT-T code including short-range transverse and longitudinal wakefield models and a longitudinal CSR wakefield model. Some applications will be presented including simulation of the photoinjector for the Linac Coherent Light Source (LCLS) and beam generation from a nano-needle photocathode.
Rostock University, Faculty of Computer Science and Electrical Engineering, Rostock
DESY, Hamburg
The determination of electron cloud instability thresholds is a task with high priority in the ILC damping rings research and development objectives. Simulations of electron cloud instabilities are therefore essential. In this paper a new particle tracking program is presented which includes the Poisson solver MOEVE for space charge calculations. Recently, perfectly electric conducting beam pipes with arbitrary elliptical shapes have been implemented as boundary conditions in the Poisson solver package MOEVE. The 3D space charge algorithm taking into account a beam pipe of elliptical shape will be presented along with numerical test cases. The routine is also implemented in the program code ASTRA, in addition we compare the tracking with both routines.
TEMF, Darmstadt
The moment method model has been proven to be a valuable tool for numerical simulations of a charged particle beam transport both in accelerator design studies and in optimization of the operating parameters for an already existing beam line. On the basis of the Vlasov equation which describes a collision-less kinetic approach, the time evolution of such integral quantities like the mean or rms dimensions, the mean or rms kinetic momenta, and the total energy or energy spread for a bunched beam can be described by a set of first order non-autonomous ordinary differential equations. Application of a proper time integrator to such a system of ordinary differential equations enables then to determine the time evolution of all involved ensemble parameter under consistent initial conditions. From the vast amount of available time integration methods different versions have to be implemented and evaluated to select a proper algorithm. The computational efficiency in terms of effort and accuracy serves as a selection criterion. Among possible candidates of suited time integrators for the given set of moment equations are the explicit Runge-Kutta methods, the implicit theta methods, and the linear implicit Rosenbrock methods. Various algorithms have been implemented and tested under real-world conditions. In the paper the evaluation process is documented.
ORNL, Oak Ridge, Tennessee
IREAP, College Park, Maryland
The University of Maryland Electron Ring (UMER) is a scaled electron recirculator using low-energy, 10 keV electrons, to maximize the space charge forces for beam dynamics studies. We have recently circulated in UMER the highest-space-charge beam in a ring to date, achieving a breakthrough both in the number of turns and in the amount of current propagated. As of the time of submission, we have propagated 5 mA for at least 10 turns, and, with some loss, for over 50 turns, meaning about 0.5 nC of electrons survive for 10 microseconds. This makes UMER an attractive candidate for benchmarking space charge codes in regimes of extreme space charge. This talk will review the UMER design and available diagnostics, and will provide examples of benchmarking the particle-in-cell code WARP on UMER data, as well as an overview of the detailed information on our website. An open dialogue with interested coded developers is solicited.
GSI, Darmstadt
Simulation studies of the transverse stability of the FAIR synchrotrons have been started. The simulation code PATRIC has been developed in order to predict coherent instability thresholds with space charge and different impedance sources. Some examples of code validation using the numerical Schottky noise and analytical stability boundaries will be discussed.
Fermilab, Batavia, Illinois