EPAC08 - Classification:05 Beam Dynamics and Electromagnetic Fields/D05 Code Developments and Simulation Techniques

05 Beam Dynamics and Electromagnetic Fields

D05 Code Developments and Simulation Techniques

Paper	Title	Page
TUOBG04	A Vlasov-Maxwell Solver to Study Microbunching Instability in the FERMI@ELETTRA First Bunch Compressor System	971
	G. Bassi Liverpool University, Science Faculty, Liverpool G. Bassi Cockcroft Institute, Warrington, Cheshire J. A. Ellison, K. A. Heinemann UNM, Albuquerque, New Mexico
	Microbunching can cause an instability which degrades beam quality. This is a major concern for free electron lasers where very bright electron beams are required. A basic theoretical framework for understanding this instability is the 3D Vlasov-Maxwell system. However, the numerical integration of this system is computationally too intensive at the moment. As a result, investigations to date have been done using very simplified analytical models or numerical solvers based on simple 1D models. We have developed an accurate and reliable 2D Vlasov-Maxwell solver which we believe improves existing codes. Our solver has been successfully tested against the Zeuthen benchmark bunch compressors. In the present contribution we apply our self-consistent, parallel solver to study the microbunching instability in the first bunch compressor system of FERMI@ELETTRA. This system was proposed as a benchmark for testing codes at the September'07 workshop on microbunching instability in Trieste. PAC2007, papers TUZBC03 and THPAN084. **https://www.elettra.trieste.it/FERMI/index.php?n=Main. MicrobProgram
	Slides
TUPP040	Intra Beam Scattering in Linear Accelerators, Especially ERLs	1631
	G. Hoffstaetter, M. P. Ehrlichman, A. Temnykh CLASSE, Ithaca
	The theories of beam loss and emittance growth by Touschek and Intra Beam Scattering have been formulated for beams in storage rings. It is there that these effects have hitherto been important because of their large currents. However, there are linear accelerators where these effects become important when considering loss rates and radiation damage. Prime examples are high current Energy Recovery Linacs (ERLs), managing these scattering effects can become challenging, and not only because of the large current, but also because the deceleration of the spent beam increases relative energy spread and transverse oscillation amplitudes. In this paper we describe two ways of simulating particle loss by these scattering affects, both implemented in BMAD. One that yields the places where scattering occurs, and another that yields loss rates along the chamber walls. BMAD includes nonlinear beam dynamics, wake effects, and more, which allows a rather complete propagation of scattered particle. For the example of the ERL x-ray facility that Cornell plans to build, we demonstrate that these capabilities are very important for designing a functional radiation protection system.
TUPP041	CSR Shielding in the Beam Dynamics Code BMAD	1634
	G. Hoffstaetter, C. E. Mayes, U. Sae-Ueng, D. Sagan CLASSE, Ithaca
	Short bunches radiate coherently at wavelengths that are longer than their bunch length. This radiation can catch up with the bunch in bends and the electromagnetic fields can become large enough to significantly damage longitudinal and transverse bunch properties. This is relevant for many accelerators that relies on bunch compression. It is also important for Energy Recovery Linacs, where spent beams are decelerated by a potentially large factor. Because this deceleration increases the relative energy spread, all sources of wake fields, especially Coherent Synchrotron Radiation (CSR), become much more important. In this paper we show how the beam dynamics code BMAD computes the effect of CSR and how the shielding effect of vacuum chambers is included by the method of image charges. We compare the results to established codes: to Elegant for cases without shielding and to a numerical solution of simplified Maxwell equations as well as to analytical csr-wake formulas. Good agreement is generally found, and in cases where numerical solutions of the simplified Maxwell equations do not agree with the csr-wake formulas, we show that BMAD agrees with these analytic formulas.
TUPP042	Status of the ORBIT Code: Recent Developments and Plans	1637
	J. A. Holmes, S. M. Cousineau, A. P. Shishlo ORNL, Oak Ridge, Tennessee
	We report on recent enhancements to the physics modules of the ORBIT Code and on progress toward a new implementation of ORBIT using python. We have developed the capability to track particles through general three dimensional electromagnetic field configurations. This facility has proved essential in modeling beam transport through the complicated magnetic field regions of the SNS injection chicane and injection dump line, where beam losses are high. We have also enhanced the acceleration module to provide more flexibility for synchrotron calculations. Finally, progress continues on the migration of the ORBIT physics models to a python user environment, and we present the status of this work.
TUPP075	Numerical Studies of Resistive Wall Effects	1709
	A. V. Tsakanian Uni HH, Hamburg M. Dohlus, I. Zagorodnov DESY, Hamburg
	In this paper we describe a new numerical code to calculate wakefields of resistive wall geometries. Our code is based on conformal implicit scheme. It allows to estimate wakefields of very short bunches taking into account transitive effects neglected in the European XFEL impedance budget so far.
TUPP084	Parallel Computation of Integrated Electromagnetic, Thermal and Structural Effects for Accelerator Cavities	1724
	V. Akcelik, A. E. Candel, A. C. Kabel, K. Ko, L. Lee, Z. Li, C.-K. Ng, L. Xiao SLAC, Menlo Park, California
	The successful operation of accelerator cavities has to satisfy both rf and mechanical requirements. It is highly desirable that electromagnetic, thermal and structural effects such as cavity wall heating and Lorentz force detuning in superconducting rf cavities can be addressed in an integrated analysis. Based on the SLAC parallel finite-element code infrastructure for electromagnetic modeling, a novel multi-physics analysis tool has been developed to include additional thermal and mechanical effects. The speedup from parallel computation enables virtual prototyping of accelerator cavities on computers, which would substantially reduce the cost and time of a design cycle. The multi-physics tool will be applied to the LCLS rf gun and a superconducting rf gun cavity.
TUPP085	Beam Dynamics Using Graphical Processing Units (GPUs)	1727
	R. Appleby, D. Bailey, M. D. Salt UMAN, Manchester
	Simulation of particle beam dynamics in accelerators is computationally expensive, and requires very high particle statistics and accuracy. Conventional beam tracking tools operate sequentially on particle phase space to compute the trajectories of particles through many turns of circular, and linear, machines. Graphical Processing Units (GPUs) utilise stream processing techniques to dramatically speed up parallel computational tasks, and offer considerable performance benefits to particle beam dynamics processing. In this paper, the stream processing beam dynamics code GPMAD is presented, which exploits the NVidia GPU processor, and demonstrates the considerable performance benefits to particle tracking calculations. The accuracy and speed of GPMAD is benchmarked using the Diamond Light Source BTS lattice, and the collimation system is evaluated.
TUPP086	Efficient 3D Space Charge Calculations by Self-adaptive Multigrid Methods Using the Chombo Framework	1730
	C. R. Bahls, G. Pöplau, U. van Rienen Rostock University, Faculty of Computer Science and Electrical Engineering, Rostock
	Current and future accelerator design requires efficient 3D space charge computations for high brightness bunches which should be as precise and fast as possible. One possible approach for space charge calculations is the particle-mesh-method, where the potential is calculated in the rest frame of the bunch by means of Poisson's equation. For an efficient solution of this elliptic PDE an appropriate adaptive discretization of the domain is required. Especially it has to take into account discontinuities in the distribution of the particles. The solution method we investigate in this paper is a self-adaptive multigrid method applying composite grids. To accomplish this, we use the library Chombo* which is being developed as a framework for adaptive multiresolution solvers for elliptic and hyperbolic partial differential equations. *Developed and distributed by the Applied Numerical Algorithms Group of Lawrence Berkeley National Lab., http://seesar.lbl.gov/ANAG/chombo/
TUPP087	Short Range Wakepotentials Computed in a Moving Frame	1733
	W. Bruns WBFB, Berlin
	When computing wakepotentials in a frame moving in the same direction as the exciting charge, the relativistic charge is stretched by the factor gamma of the frame's velocity. Because the needed mesh density is proportional to the length of the charge, the moving frame allows a larger gridspacing. The paper presents some implementation details of the handling of the material boundaries moving though the computational volume, and reports some results.
TUPP088	Software Components for Electron Cloud Simulation	1735
	D. R. Dechow, P. Stoltz Tech-X, Boulder, Colorado J. F. Amundson, P. Spentzouris Fermilab, Batavia, Illinois B. Norris ANL, Argonne, Illinois
	The Synergia2 beam dynamics code is an attempt to incorporate state-of-the-art space charge models from the Impact code into the Chef accelerator tracking code. The need to add new accelerator physics capabilities to the Synergia2 framework has led to software development efforts based on the Common Component Architecture (CCA). The CCA is a specification and a toolset for developing HPC from interchangeable parts, called components. Electron cloud is a potentially limiting effect in the performance of both high-intensity electron and proton machines. The modeling of electron cloud effects is important for the Fermilab main injector. Here, electron cloud effects are expected to play a significant role when the main injector operates in the regime of a high-intensity proton source for the neutrino program. In the ideal case, computational accelerator physicists would like to be able model electron cloud generation and dynamics in a single, self-consistent simulation. As a first step towards creating component-based, electron cloud generation simulations, this work describes a CCA component created from TxPhysics, a library of impact and field ionization routines.
TUPP089	Implementation of Fringe Field Dipole Magnets into the V-Code Beam Dynamics Simulation Tool	1738
	S. S. Franke, W. Ackermann, B. Steiner, T. Weiland TEMF, Darmstadt J. Enders, C. Hessler, Y. Poltoratska TU Darmstadt, Darmstadt
	Fast online beam dynamics simulations can advantageously assist the machine operators at various particle accelerator machines because they provide a more detailed insight into the actual machine status. Based on the moment approach a fast tracking code named V-Code has been implemented at TEMF. Within the SFB 634 project the V-Code beam dynamics simulation tool is supposed to be installed at the Superconducting Darmstadt LINear ACcelerator S-DALINAC which is designed as a re-circulating linear accelerator. In order to be able to simulate the entire beam line, an implementation of fringe field dipole magnets is mandatory. Unlike a hard edged field approach the fringe fields influence the beam focusing and its inhomogeneity results in a non-circular bunch motion. For an accurate reproduction of the transverse motion specialized techniques to obtain and to handle the reference path in V-Code together with the 3D-field data along the curved trajectory had to be developed. In the paper a summary of implementation details together with simulation results will be provided.
TUPP090	A Kinetic Model of Multipaction for SRF Cavities for Accelerator Driven Sub-Critical System (ADSS)	1741
	S. Ghatak, N. Gupta IITK, Kanpur A. S. Dhavale, K. C. Mittal BARC, Mumbai
	This work simulates multipaction in a 700 MHz elliptical SRF cavity. The cavity design was optimized using SUPERFISH. Then the electromagnetic field was re-computed with FEMLAB, a package using the finite element method, to obtain a more accurate field-mapping, and to make the field values available for computation of multipaction. In the multipacting subroutine, electrons were assumed to be released into the system from various points with different initial parameters. The electrons trajectories were tracked until they hit the cavity surface. Leap-frog scheme was used to solve the Lorentz force equation for primary electrons, as it is easy to use and is accurate up to second order. The position, velocity, phase and kinetic energy of primary electrons at each time step were calculated and stored. An interpolation function was used to calculate secondary emission yield (SEY) at different impact energies. With the emission of secondary electrons, their trajectories too were tracked along with primary electrons, in order to identify parameters responsible for multipaction. By repeating this process for large number of electrons, the multipacting trajectories were identified.
TUPP091	WISE: a Simulation of the LHC Optics Including Magnet Geometrical Data	1744
	P. Hagen, M. Giovannozzi, J.-P. Koutchouk, T. Risselada, F. Schmidt, E. Todesco, E. Y. Wildner CERN, Geneva
	The beam dynamics in LHC strongly depends on the field quality and geometry of the magnets. A model of the LHC optics has been built, based on the information available at the end of the production as well as on statistical evaluations for the missing information The pre-processor WISE generates instances of the LHC field errors for the MAD X program, with the possibility of selecting various sources. This paper describes the progress since WISE was presented in EPAC06. The slot allocation in LHC is completed since all magnets are installed and interconnected. Geometric measurements have been added for all magnets. Furthermore, some statistical data is available relative to the precision of magnet installation (alignment) and tunnel movements. In this paper the code and the data are used to update the beta-beating estimate at injection and collision energy. The relevance of misalignments of the different magnets and their impact on beta-beating is compared to the sources that have been previously considered, i.e. the spread in the gradient of the cell quadrupoles and the uncertainty associated to the knowledge of the transfer functions of the stand-alone quadrupoles.
TUPP093	Crystalline Beam Simulations	1747
	D. A. Krestnikov JINR/DLNP, Dubna, Moscow region M. Grieser MPI-K, Heidelberg M. Ikegami JAEA/Kansai, Kizu-machi Souraku-gun Kyoto-fu I. N. Meshkov, A. O. Sidorin, A. V. Smirnov, G. V. Trubnikov JINR, Dubna, Moscow Region M. Nakao, A. Noda, H. Souda, H. Tongu Kyoto ICR, Uji, Kyoto K. Noda, T. Shirai NIRS, Chiba-shi
	A new program code was elaborated for the simulation of crystalline beams on the S-LSR storage ring (Kyoto Univ., Japan) under action of the cooling system. For the investigation of ordered proton beams, which recently were observed in first time on S-LSR, a special molecular dynamics technique was used. This article presents results of the numerical simulation and comparison with experimental data.
TUPP094	Recent Improvements in the Tracking Code PLACET	1750
	A. Latina, H. Burkhardt, G. Rumolo, D. Schulte, R. Tomas CERN, Geneva E. Adli University of Oslo, Oslo Y. Renier LAL, Orsay
	The Tracking Code PLACET has recently undergone several improvements. A redesign of its internal data structures and a new user interface based on the mathematical toolbox Octave have considerably expanded its simulation capabilities. Several new lattice elements, optimization algorithms and physics processes have been added to allow for more complete start-to-end simulations. The usage of the AML language and the Universal Parser Library extened its interfacing capability.
TUPP095	Computation of Resistive Wall Wakefields with the PBCI Code	1753
	T. Lau, E. Gjonaj, T. Weiland TEMF, Darmstadt R. Maekinen TUT, Tampere
	Both geometric disturbances and resistive wall loss of accelerator cavities contribute to the impedance causing the beam to lose energy. Impedance due to arbitrary three-dimensional (3-D) geometries can be computed with the Parallel Beam Cavity Interaction (PBCI), a parallelized, 3D-wakefield code. However, the contribution of wall loss is often significant. The contribution of this work is to incorporate resistive wall loss into 3-D time-domain simulation. Surface-impedance concept is used to consider wide-band skin-effect loss of metal. In theory, the proposed approach can be extended to consider high-frequency phenomena such as frequency-dependent conductivity of metal and anomalous skin effect.
TUPP097	New Formalism in the Spin Tracking Code Spink	1756
	A. U. Luccio, F. Lin BNL, Upton, Long Island, New York
	The code Spink, in use for more than 10 years to track polarized hadrons in a synchrotron, was overhauled with the introduction of a new system of coordinates based on a generalized Frenet-Serret system in all dimensions in space, which allows a better treatment of the curvature of the reference orbit. Two more improvements are (a) treatment of tensor polarization for particles like polarized deuterons, and (b) inclusion of space charge and beam-beam effects, so the code can be used to track spin in synchrotrons with high luminosity like new generation colliders. A. U. Luccio. Proc. Adriatico Research Conf. on Trends in Colliders Spin Physics. Trieste, Italy, 12/5-8, 1995.
TUPP098	The 3D Space Charge Field Solver MOEVE and the 2D Bassetti-Erskine Formula in the Context of Beam - E-cloud Interaction Simulations	1759
	A. Markovik, G. Pöplau, U. van Rienen Rostock University, Faculty of Computer Science and Electrical Engineering, Rostock
	In this paper the fields computed with our 3D space charge field solver MOEVE are compared to those obtained by means of the Bassetti-Erskine formula, which is a widely used 2D approximation of the electric field of a Gaussian bunch. In particular we are interested in the transversal fields of very flat bunches as the ILC or the KEKB positron bunch. Supposing a longitudinal Gaussian distribution of the bunches, we compare the computed transversal fields for a certain line density of the positron bunch. It turns out that the fields from the 2D and the 3D computation coincide very good.
TUPP099	Improvement of RFQSIM	1762
	J. M. Maus, A. Schempp IAP, Frankfurt am Main
	RFQSIM has constantly been developed at IAP to simulate multiple particles dynamic of RFQs which include simulations of high current applications, decelerators, debuncher and rebuncher e.g. for the medical application. The latest work to improve RFQSIM include the option to change the modulation and or the aperatur of the simulated RFQ to produce the same acceleration and or focusing efficiency than the two term potential gives for different electrode geometries. Additional work was done to improve the graphical analysis during runtime as well as the change of all routines to be compatible with the Fortran F95 standard. The maximum number of particles to be simulated was increased to 200k. The paper shows examples of results for the high current application like the of the p-linac and the new HLI injector for GSI and for transition sections.
TUPP100	A Four-dimensional Vlasov Solver for Microbunching Instability in the Injection System for X-ray FELs	1764
	M. Migliorati, A. Schiavi Rome University La Sapienza, Roma G. Dattoli ENEA C. R. Frascati, Frascati (Roma) M. Venturini LBNL, Berkeley, California
	The phenomenon of microbunching instabilty arises from small charge-density fluctuations in the electron bunches that are amplified by the combined effect of space charge and coherent synchrotron radiation as the beam travels through magnetic compressors. In order to study the coupled longitudinal and transverse beam dynamics we propose to develop a four-dimensional grid-based Vlasov solver. The goal is to give an accurate characterization of the microbunching instability seeded by the random noise present in the initial bunch distribution. Solving directly the Vlasov equation instead of using macroparticle simulations has the advantage of avoiding the statistical fluctuations due to a limited number of macroparticles. Because a Vlasov solver in a high dimension phase-space tends to be particularly time consuming, to be practical a code implementing this method should run on parallel processors. In this paper we report progress toward the realization of such a 4D Vlasov solver.
TUPP102	Beam Transport with Scattering Using SRIM Supporting Software Routines Code	1767
	M. Pavlovic, I. Strasik STU, Bratislava
	In many situations a particle beam is transported through matter-containing components separated by ion-optical elements. The matter-containing components scatter the beam and alter its emittance diagram. In order to include accurately the scattering in beam-transport a special beam-transport module was included in the SRIM Supporting Software Modules package (S3M). It uses transfer-matrix formalism in ion-optical elements. At the entry to a scattering element a beam-generation routine converts the actual σ-matrix into an ensemble of particles and writes a special SRIM input-file. The beam-transport in the scattering element is then calculated by SRIM MC particle tracking. At the exit of the scattering element, the module imports back the SRIM output data and can either continue with transfer-matrix transformations or generate a modified σ-matrix that can be used by other ion-optical programs. It means the beam transport with scattering can either be fully calculated by S3M, or data exchange between S3M and ion-optical programs can be provided. S3M beam-transport module is described in the paper with some typical application examples. M. Pavlovic, I. Strasik. Supporting Routines for the SRIM code, Nucl. Instr. and Meth. B 257 (2007) 601-604.
TUPP103	The Performance of 3D Space Charge Models for High Brightness Electron Bunches	1770
	G. Pöplau, U. van Rienen Rostock University, Faculty of Computer Science and Electrical Engineering, Rostock K. Floettmann DESY, Hamburg
	Precise and fast 3D space charge calculations for high brightness, low emittance electron beams are of growing importance for the design of future accelerators and light sources. The program package Astra (A space charge tracking algorithm) has been successfully used in the design of linac and rf photo injector systems. The Astra suite originally developed by K. Flöttmann tracks macro particles through user defined external fields including the space charge field of the particle cloud. In this paper we investigate the performance of the 3D space charge models implemented in Astra. These are the FFT-Poisson solver with the integrated Green's function and the iterative Poisson solver based on the multigrid technique. The numerical tests consider the accuracy of the solvers for model bunches as well as the performance within a typical simulation for the XFEL.
TUPP105	Theoretical Field Analysis for Superferric Accelerator Magnets Using Elliptic Multipoles and its Advantages	1773
	P. Schnizer, E. S. Fischer GSI, Darmstadt P. G. Akishin JINR, Dubna, Moscow Region B. Schnizer TUG/ITP, Graz
	FAIR will build a set of accelerators and storage rings at GSI Darmstadt. Nearly all of them transport beams of elliptical shape (SIS 100, CR, NESR, RESR, SuperFRS). Magnetic field calculations as well as magnetic measurements provide precise field information, which is used to improve the properties of machine using numerical simulations. We had developed elliptical multipoles fulfilling the Laplace equation which enable us to describe the field within the whole aperture consistently. Now the representation of these has been simplified considerably as compared to earlier ones. Meanwhile we found analytical expressions to derive circular multipoles directly from the elliptic multipoles. We illustrate the advantage of this data representation on SIS 100 magnet data and show how a concise set of harmonics can be derived from rotating coil measurements.
TUPP109	Meshless Solution of the Vlasov Equation Using a Low-discrepancy Sequence	1776
	R. L. Warnock SLAC, Menlo Park, California J. A. Ellison, K. A. Heinemann, G. Q. Zhang UNM, Albuquerque, New Mexico
	A successful method for solving the nonlinear Vlasov equation is the semi-Lagrangian method, in which the phase space density is represented by its values on a fixed Cartesian grid with interpolation to off-grid points. Integration for a time step consists of following orbits backward in time from initial conditions on the grid, with the collective force frozen during the time step. We ask whether it would be more efficient to use scattered data sites rather than grid points, namely sites from a low-discrepancy sequence as used in quasi - Monte Carlo integration. This requires a technique for interpolation of scattered data, and with such a technique in hand one can try either backward or forward orbits. Here we explore the forward choice, with the data sites themselves following forward orbits. We treat a problem well studied by the backward method, longitudinal motion in the SLAC damping rings. Over one or two synchrotron periods results are encouraging, in that the number of data sites can be reduced by a large factor. Over longer times it appears that the sites must be redistributed or changed in number from time to time, because of clustering.
THYM03	Advanced Computing Tools and Models for Accelerator Physics	2947
	R. D. Ryne LBNL, Berkeley, California
	The design of the next generation of accelerators will require a new level of simulation capability to perform high resolution, multi-physics modelling of beam dynamics phenomena and to design complex 3D electromagnetic structures. Thanks to the availability of computational resources that will soon reach the petascale, we will be able to perform simulations involving unprecedented size, complexity, and physical realism. This paper will review the state-of-the-art in scientific computing for accelerator physics.
	Slides
FRXBGM01	Impedance Computation and Measurement in Modern Storage Rings	3728
	R. Nagaoka SOLEIL, Gif-sur-Yvette
	Recent progress in the evaluation of machine impedance and instability thresholds will be reviewed, and comparisons made between measurements and predicted impedance in recently commissioned storage rings.
	Slides