UCLA PARMELA (c. 1998)

1
UCLA PARMELA (c. 1998)
E. Colby, SLAC/ARDB March 16, 2000

• SIGNIFICANT PHYSICS CHANGES
• Particle Loading/Noise Mitigation
• Hammersley sequence particle loading quiet
  start from J. Gonichon, MIT, now adopted into
  LAACG release
• Dithering of time step to reduce numerical
  microbunching
• Realistic particle loading, incorporating
  spatial and temporal density modulations and
  particle loading by direct sampling from a video
  image of the laser mode pattern.
• RF and Static Field Representation
• Arbitrary 3D Electromagnetic (or static) fields,
  via 40x40x100 point maps of (Ex,Ey,Ez,Bx,By,Bz),
  quadratically interpolated (nearest, linear and
  cubic interpolation also supported), supporting
  two simultaneous frequencies and arbitrary extent
  (N.B. Fringe fields of cavities may be
  represented)
• 1D (Bz(z)) and 2D (Br(r,z), Bz(r,z)) static
  magnetic fields, quadratically interpolated
• PPM magnetic field representation (PCM, Sturrock,
  Offset-pole).

2
E. Colby, SLAC/ARDB March 16, 2000

• Space Charge and Image Charge Representation
• Fully 3D space charge fields with arbitrary
  aspect ratio ellipsoidal charge clouds
• Partial port of Carlstens line-by-line algorithm
  incorporating coherent synchrotron radiation
  effects in bends.
• Wakefield Effect Estimation (non-self consistent)
• Loading and application of W(z) and W?(z) maps
  from ABCI, Xwake, etc. implemented as point
  transformations at element exits.
• Corrected Physics Errors from Los Alamos 1995
  Version
• Removal of bizarre field-weighting at cathode
  that caused serious underestimation of image
  charge effects (RZ Mesh method for space charge)
• Correction of error in exit fringe field term of
  BEND card (T313 element missing 1/r).
• Diagnostics
• Addition of many diagnostics for computing
  emittances, dumping RF, static magnetic and space
  charge fields, particle distributions,
  trajectories, etc.

3
E. Colby, SLAC/ARDB March 16, 2000

• Significant Non-physics Changes
• PVM implementation of PARMELA with automatic job
  generation and semi-automated data analysis to
  perform large parameter scans quickly
• All output is in columnar ASCII files, not a
  binary TAPE2 file
• Development of a large library (gt200 scripts and
  functions) of Matlab code for post-processing and
  displaying output from PARMELA.
• Physics Errors in PARMELA
• Integration is non-symplectic (Euler method)
• EM fields and space charge fields not
  self-consistently calculated in a common
  framework
• Space charge calculation is electrostatic (no
  radiative effects included no internal magnetic
  fields even at cathode where velocity sheer is
  substantial)
• Cathode image charge treatment is
  relativistically incorrect (images placed at
  present rather than retarded positions,
  underestimating the induced energy spread)
• 3D space charge evaluation technique is noisy