HOM Damping Simulation and Measurement of JLab Ampere Class Cavity

1
HOM Damping Simulation and Measurement of JLab
Ampere Class Cavity

• Haipeng Wang
• also Robert Rimmer and Frank Marhauser
• Jefferson Lab, Newport News, Virginia 23606, USA

2
MAFIA wakefield calculation for both broad and
narrow bands HOM impedance calculation
Critical steps to get correct cavitys
impedance through FFT of wake function
WG
WG
WG
WG
WG
WG
WG

• Generate 3D cavity vacuum CAD drawing with
  waveguide ports facing to Cartesian coordinates.

WG

• Import CAD geometry, mesh volume with uniform
  steps (5mm) in beam direction.
• Define waveguide port positions and mesh in 2D.
• Solve 10 lowest eigen modes in each WG port.
• Load WG solutions to each port with zero power
  input. Check any ignored mode.
• Set up 1D current in Gaussian bunch with s3cm,
  010s long, 1q charge at a defined off-axis
  distance.
• Set up wake potential monitor in beam direction
  at a defined path.
• Calculate 1311m long wave (4days, 1millon
  meshes)
• Take max. 2n wake data set with rest of them zero
  out. Do the Cosine window tapering.
• FFT wake (get Wzfft and Gaussian beam.
• Normalized impedance with ZamWzfft/Bam
  ZphWzfftph-Bph. With correct phase wrapping
  within -90o to 90o, the real part of impedance
  Zre should be always gt0.

3
Broadband impedances contribute to power
deposition or BBU below beam pipe cut-off
frequency
bunch length 3cm
Monopole modes excited by on-axis beam bunch
monopole dipole modes excited by off-axis 5cm
beam bunch
beam excitation Frequency (GHz) monopole modes only impedance (?) dipole modes only impedance (?) at 1cm off-axis beam peak current (A) Monopole HOM power (W) Dipole HOM power at 1cm off axis (W) beam power (W)
1.497 255.566 3.289 4.000 4089 52.6 4141.7
2.994 278.308 2.232 4.000 4453 35.7 4488.6
4
Broadband HOM power calculation for above cut-off
Calculation formula
8.5 kW
Total 18kW/cavity
2 Amp
5 cell
1.33 nC
can be improved by model measurement or high
power computing
Obtained from ABCI or MAFIA
5
Copper model bench measurement

• S21 from beam pipe to beam pipe.
• Labview automation.
• Ceramic bead-pull on-axis or off-axis.
• End groups staggered relatively at 30o or 60o.
• 5, 6, 7-cell stack assembly then 5-cell welded.
• Data sets with dummy loads or shorts.
• Rotatable coupling antennas.

6
Five-peak fitting program to get real peak,
frequency and Q

• Multi-peak fitting program is try to find real
  amplitude at each Lorenz resonance frequency, and
  filter out the contribution from each neighbor
  peak. Totally five peaks cab be fitted together.
• Each resonance peak is fitted with the Lorenz
  function.
• Total amplitude Am2(Re1P2Re2P3Re3)2(Im1P
  2Im2P3Im3bpBg)2
• P2, P3,is each peak polarity1 or -1. bp is
  background polarity. Bg is total background.
• This fitting program can be also used to find
  real impedance of each HOM from overlapped HOMs
  spectrum like obtained from MAFIA wakefield FFT..

7
First measurement data do not agree simulation
well
8
RF dummy load and single cell aluminum model
bench measurement
9
E-antenna rotation changes excitation of the HOMs

• Antenna rotation relatively 90 degree can
  suppress one or other polarization of dipole
  modes, but once the beam pipe uses absorber, and
  for those above cut-off modes, this advantage
  disappear quickly
• HOM mode IDs is most complication in the HOM
  impedance measurement.

10
Multiple beam (1d current) excitation of wake in
3D structure
One beam
Two beams
Monopole modes only
Dipole modes, polar 1
Quadruple modes, polar 1

• If there is no structure asymmetry about
  mid-plane and no asymmetric meshing, wave
  excitation and reflection have to satisfy the
  enforced boundary condition. A proper mode
  suppression should be made.
• In reality of the squared (MAFIA) and finite
  number of meshes, the mode contamination of high
  Q modes always happen.
• Excitation more than three-beam is possible.

Three beams
Dipole modes, polar 2
Quadruple modes, polar 2
11
Multiple beam excitation of wakes in a
single-cell high current cavity with one Y
waveguide group
12
Mode ID can be labeled to the spectrum from the
MAFIA simulations without a bead pulling
13
MAFIA multi-beam wakefield simulation, monopole
mode excitation only
14
MAFIA multi-beam wakefield simulation, dipole
mode excitation only
15
MAFIA multi-beam wakefield simulation, quadruple
mode excitation only
16
Summary

• Multiple beam excitation technique used in MAFIA
  time domain wakefield calculation is very
  powerful simulation tool. It can selectively
  choose BC and number of beam, position and
  polarization to suppress unwanted modes.
• FFT of MAFIA wake can get both real and
  amplitude, narrow and broad band cavitys
  impedance if the wake function is properly phase
  unwrapped and normalized.
• Mode identity is most important part of HOM
  measurement, it can be improved by rotation and
  position of antennas, bead pulling, and 2D (eigen
  mode) or 3D (time domain) simulation.
• After careful mode ID, calibration and
  bead-pull, Initial data bellow cut-off
  frequencies on single-cell aluminum model agrees
  with simulations.
• By this measurement, we bunch-marked our
  simulation method and data used in the prototype
  design.
• R/Q measurement by calibrated bead-pulling is
  needed to replace MAFIA 2D R/Q for above cut-off
  modes.
• We need re-measure the 5-cell structure after
  mode ID is clear from the new simulations.