XFEL simulations with GENESIS

1
XFEL simulations with GENESIS
Bart.Faatz_at_desy.de

• Outline
• Undulator layout (real/simulated)
• Matching
• Genesis input files
• Post-processing
• Genesis output

2
Undulator description real
Matching section
Ls
Lc
Parameters Symbol Value Unit
Undulator segment length Ls 5 m
Undulator cell length Lc 6.1 m
Undulator gap g 10 mm
Undulator period (SASE1) lu 48 mm
Quadrupole optical length 192 mm
Quadrupole gradient at 20 GeV Q 20 T/m
3
Undulator description simulated
Undulator segment
Defocusing quadrupole
Focusing

• Increased quadrupole length in order to increase
  stepsize (adjusted gradient)
• Automatic phase matching between undulators

Actual undulator geometry is defined in the
magnet lattice file
4
Matching at the undulator entrance
Electron beam information starts with output from
Elegant (In this example until the undulator
entrance with 200k particles)

• Elegant2genesis (SDDS toolkit)
• Check slices/ particles per slice
• Adjust initial a and b for optimal slice
• Adjust initial offset and angle for optimal slice
• Add wakefield along the undulator

1e-9
100
elegant2genesis ltinput.filegt ltoutput.filegt
-totalCharge ltchargegt -slices ltslicesgt
5
Genesis input

• E-beam slice file (Elegant2genesismatch)
• Elegant2genesis produces global beam parameters
  for each slice
• match adjusts ltxgt, ltygt and ltxgt, ltygt for slice
  with maximum current
• Magnetic lattice file
• Genesis namelist input (Numerical and
  IO-parameters, )

From global parameters, Genesis fills the 5D1
phase space with a Gaussian distribution
6
Input distribution
Tail of the bunch
7
Post-processing

• Xgenesis (IDL)
• Interactive (GUI)
• Gen2Spread Separate Genesis output file
• In slices along undulator
• At different undulator position temporal profile

8
Steady state results no wakefields
Unmatched angle/offset
9
Energy loss due to the wakefield
assumed an inner radius of 4 mm copper coated
stainless steel
resistive wall wake only
10
Steady state results including wakefields
(matched)
Without taper
11
Time dependent simulations no wakefields
NSLICE
NTAIL
Part of the bunch selected for time dependent
simulations
Simulate 20000 photon slices to assure overlap,
8000 particles each
12
70 GW power, 3mm FWHM pulse (10 fs)
13
Frequency spectrum
?
14
17000
15
17500
16
18000
17
Conclusion
Much more work needs to be done
Extend time window Get frequency
information Include all wakes, quantum
fluctuations,