Accelerator Modeling Through High Performance Computing Z. Li - PowerPoint PPT Presentation

Accelerator Modeling Through High Performance Computing Z. Li NERSC,LBNL Advanced Computations Department Stanford Linear Accelerator Center NCCS, ORNL Presented at Jefferson Lab, 9-24-2007 Work supported by U.S. DOE ASCR, BES & HEP Divisions under contract DE-AC02-76SF00515

Contributions To This Talk A. Candel A. Kabel K. Ko Z. Li C. Ng L. Xiao V. Akcelik S. Chen L. Ge L. Lee E. Prudencio G. Schussman R. Uplenchwar Advanced Computations Department Work supported by U.S. DOE ASCR, BES & HEP Divisions under contract DE-AC02-76SF00515

Outline � DOE SciDAC Program � Parallel Code Development under SciDAC � Applications to DOE Accelerator Projects � Collaborations in Computational Science Research

SciDAC Program SciDAC : Scientific Discovery through Advanced Computing � DOE Office of Science (SC) Simulation Initiative � Promotes application of High Performance Computing to SC programs across BES/NP/HEP Offices � Multi-disciplinary approach – computational scientists (CS & AM) work alongside application scientists � Accelerator project started as Accelerator Simulation and Technology (AST) in SciDAC1, and continues as Community Petascale Project for Accelerator Science and Simulation (COMPASS) in SciDAC2 Goal – To develop next generation simulation tools to improve the performance of present accelerators and optimize the design of future machines using flagship supercomputers at NERSC (LBNL) and NLCF (ORNL)

SLAC SciDAC Activities � Parallel code development in electromagnetics and beam dynamics for accelerator design, optimization and analysis � Application to accelerator projects across HEP/BES/NP such as ILC, LHC, LCLS, SNS, etc… � Petascale simulations under SciDAC2 on DOE’s supercomputers - currently 3 allocation awards at NERSC (Seaborg, Bassi, Jacquard) and NCCS (Phoenix) � Computational science research through collaborations with SciDAC CET/Institutes’ computer scientists and applied mathematicians

SLAC Parallel Codes under SciDAC1 � Electromagnetic codes in production mode: Omega3P – frequency domain eigensolver for mode and damping calculations S3P – frequency domain S-parameter computations T3P – time domain solver for transient effects and wakefield computations with beam excitation Track3P – particle tracking for dark current and multipacting simulations V3D – visualization of meshes, fields and particles

SLAC Parallel Codes under SciDAC2 Codes under development: � Electromagnetics Gun3P – 3D electron trajectory code for beam formation and transport Pic3P – self-consistent particle-in-cell code for RF gun and klystron (LSBK) simulations TEM3P – integrated EM/thermal/mechanical analysis for cavity design � Beam dynamics Nimzovich – particle-in-cell strong-strong beam- beam simulation

SciDAC Tools for Accelerator Applications ILC � Accelerating Cavity (DESY, KEK, JLab) – TDR, Low-loss, ICHIRO & cryomodule designs � Input Coupler (SLAC, LLNL) – TTFIII multipacting studies � Crab Crossing (FNAL/UK) - Deflecting cavity design � Damping Ring (LBNL) – Impedance calculations � L-Band Sheet Beam Klystron – Gun and window modeling LHC � Beam-beam simulations LCLS � RF gun – emittance calculations using PIC codes SNS � Beta 0.81 cavity – end-group heating and multipacting

Problems and Solver Options Omega3P Lossy Periodic External Lossless Material Structure Coupling ISIL w/ ESIL/with Implicit/Explicit Self-Consistent Nonlinear SOAR refinement Restart RestartedArnoldi Loop Arnoldi/JD i Krylov Subspace Methods WSMP MUMPS SuperLU Domain-specific preconditioners - Calculating HOM damping in the ILC cavities requires a nonlinear eigensolver when modeling the coupling to external waveguides (FP & HOM couplers) to obtain the complex mode frequencies as a result of power outflow

Advances In Accelerator Simulation ILC 3-module RF unit ILC cryomodule 1.0E+10 Modeling ILC RF 9-cell ILC cavity systems under 1.0E+09 operation Degrees of Freedom conditions 1.0E+08 3D Damped Moore’s law 1.0E+07 Cell 1.0E+06 3D detuned structure with coupler 1.0E+05 2D detuned structure 1.0E+04 2D Cell 1.0E+03 1.0E+02 1990 1995 2000 2005 2010 2015 2020

End Group Design – HOM Damping

LL Cavity End-group Design LL Shape � >15% higher R/Q (1177 ohm/cavity) � >12% lower Bpeak/Eacc ratio � 20% lower cryogenic heating � Most important modes are 0-mode in 1st 1.E+06 the 3rd band R/Q (ohm/m^2/cavity) 2nd 1.E+05 � High R/Q in the 1st&2nd bands are 3rd 1.E+04 up to 1/3 of the 3rd band 4th 5th 1.E+03 � Beam pipe tapers down to 30-mm, 1.E+02 3rd band damped locally by HOM couplers 1.E+01 1.600 2.100 2.600 3.100 � Damping criteria: 3rd band mode F (GHz) Qext<10 5 (?)

High R/Q 3 rd Band Modes Qext=1.4x10 4 Qext=4.6x10 5 10 5 � 41mm end � 38mm pipe radius 0 pipe radius � Field significantly -5 � Low field in improved the coupler -10 Er (a.u.) � End-group modified to region -15 enhance damping -20 r=37mm r=38mm -25 r=39mm -30 r=41mm -35 Coupler region -40 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 1.2 1.4 z (m)

LL Cavity End-group 3mm 37 0 54 0 150 1.0E+06 Polarization ode polarization angle 120 Qext 90 1.0E+05 60 Qext 30 1.0E+04 M 0 New LL design Initial with TTF coupler -30 1.0E+03 1.600 1.800 2.000 2.200 2.400 2.600 2.800 3.000 1.630 2.130 2.630 F (G H z ) F (GHz) Effective damping achieved by optimizing: � End-group geometry to increase fields in coupler region � Loop shape and orientation to enhance coupling � Optimized azimuthal coupler orientation for x-y mode polarization

Crab Cavity Design for ILC BDS F (Hz) 3.6E+09 3.7E+09 3.8E+09 3.9E+09 4.0E+09 4.1E+09 4.2E+09 Improved FNAL design 0 • better HOM, LOM and SOM damping notch gap=3.0mm -20 notch gap=3.1mm • reduced HOM notch gap sensitivity -40 (to 0.1 MHz/ μ m from original 1.6 MHz/ μ m) S12 (dB) -60 • eliminates LOM notch filter -80 • avoids x-y SOM coupling -100 -120 Original Design Notch gap sensitivity original Qext in Crab-cavity 1.E+08 1.E+07 1.E+06 modified Qext New Design 1.E+05 1.E+04 Qext in the original design 1.E+03 Qext in the new design SOM x-y coupling 1.E+02 2.70E+09 3.30E+09 3.90E+09 4.50E+09 5.10E+09 F (Hz) Omega3P damping calculation A copper model is being built in UK lab based on this design.

Cavity Imperfection � HOM damping � X-Y coupling � Effects on beam emittance

TESLA cavity imperfection study TDR prototype cavity Idea TDR cavity Omega3p model The actual cell shape of the TESLA cavities differ from the idea due to fabrication errors , the addition of stiffening rings and the frequency tuning process .

TESLA cavity Measurement Data Study TDR cavity: operating mode from 80 cavities TTF module 5: 1st/2nd dipole band 1.E+07 45 1st band 6th pair 2nd band 6th pair Idea: 753KHz 1.E+05 1.E+06 40 1.E+04 1.E+06 1.E+05 1.E+03 1.E+04 35 1700 1701 1702 1703 1704 1705 1706 1877 1878 1879 1880 1881 1882 Qext 1.E+05 Eacc(MV/m) 30 1.E+04 25 20 1.E+03 700 720 740 760 780 800 820 840 860 880 1600 1650 1700 1750 1800 1850 1900 f π -f 8 π /9 (KHz) • omega3p Calculation F (MHz) (Neubauer, Michael L.) Dipole mode frequencies shift The mode spacing increases. and Qext scatter.

Modeling Imperfection Of ILC TDR Cavity Red: ideal cavity Stretching cavity f Blue: deformed cavity Q ext scatter Ideal split Cavity Stiffening ring shift Welding stiffening Cell elliptically ring deforms disk f deformed � Determine shape deformation from measured cavity data, inverse and forward methods � Important to understand effect on Qext and x-y coupling of beam dynamics � Actual deformation? – geometry measurement data will be very helpful

Cylindrical Symmetric Deformation ( 200micro on top/607micro on disk ) - cause frequency shift 1 .E + 0 5 1 .E + 0 6 1st/2nd dipole band modes Ideal v.s. deformed 1.E+07 1 .E + 0 4 1 .E + 0 5 Qext Qext Cavity stretching 1 .E + 0 3 1 .E + 0 4 1.E+06 1 .8 7 9 E + 0 9 1 .8 8 1 E + 0 9 1 .7 0 0 E + 0 9 1 .7 0 5 E + 0 9 id F (H z) i F (H z) d Qext 1.E+05 1.E+04 Stiffening idea-cavity ring deform ed surface 1.E+03 1.60E+09 1.70E+09 1.80E+09 1.90E+09 F (Hz) f π -f8 π /9=772KHz within meas. Range. 1st/2nd dipole band mode freq. shift roughly fit measurement data. 8-cavity measurement v.s. simulation TTF module 5: 1st/2nd dipole band meas. Module 5: 8 cavities :1st/2nd dipole band mode frequency shift data 2 1.E+07 0 F(real)-F(idea) (MHz) -2 1.E+06 -4 Qext 1.E+05 -6 -8 ac62 ac61 1.E+04 ac65 ac66 -10 ac79 ac77 ac63 ac60 -12 defomed surface 1.E+03 0 4 8 12 16 20 24 28 32 36 1600 1650 1700 1750 1800 1850 1900 Mode Index F (MHz)

Cell elliptical deformation (dr=250micro) - cause mode Mode x-y coupling& Qext scattering TDR cavity with elliptical cell shape 1.E+07 1 st band 6 th pair 2nd band 6 th 2nd band: 6th pair 1.E+05 pair defor cell1 along x defor cell1 along x 1.E+05 defor cell4 along x defor cell4 along x defor cell1 and 4 along x defor cell1 and 4 along x idea cavity idea cavity 1.E+04 Qext 1.E+06 1.E+04 1.E+03 1704.0 1705.0 1706.0 1707.0 1708.0 1787.0 1787.5 1788.0 1788.5 1789.0 1789.5 1790.0 F (MHz) Qext 1.E+05 ideal cavity 1.E+04 elliptically deformed 1.E+03 cavity 1.60E+09 1.65E+09 1.70E+09 1.75E+09 1.80E+09 1.85E+09 1.90E+09 F (Hz)