Start-to-End Simulations for the TESLA LC

1
Start-to-End Simulationsfor theTESLA LC

• A Status ReportNick WalkerDESY

TESLA collaboration Meeting, Frascati, 26-28th
May 2003
2
A Mixed-Bag of Topics

• Software tools (MERLIN advertisement)
• Beam-based alignment of the TESLA linac
• DFS vs Ballistic Alignment
• S2E simulations of luminosity performance

3
Simulation and Simulation Tools

• Much progress made towards true S2E simulations
  during TRC studies
• Codes used

• PLACET (linac)
• MERLIN (LET)
• LIAR (linac)
• DIMAD (BC, BDS)
• MAD (BC, BDS)
• ELEGANT (BC)
• GUINEAPIG (for beam-beam)

90
4
Software tools
MERLIN C class library

• Used to simulate
• Bunch Compressor
• Main Linac
• BDS
• DR (A. Wolski, LBNL)
• Models
• Single-bunch wakefields
• Full 3D alignment errors
• Girders and complex geometries
• Diagnostics tuning algorithms
• Thin-spoiler scattering (used for halo and
  collimation studies)
• Synchrotron radiation
• Control system-like interface

5
Software tools
MERLIN C class library

• Two tracking modes
• Particles(ray tracing, 2nd O TRANSPORT)
• Slice macro-particles(linac, LIAR/PLACET)
• New MATLAB interface
• More details later in this talk
• Powerful and Flexible
• Allows rapid code development
• Not for the faint hearted!

6
TESLA Linac Alignment

• A Little History
• Many studies for CDR and TDR, most based on
  Dispersion Free Steering (DFS)
• P. Tenenbaum (SLAC) made an independent study
  using LIAR for EPAC-2002 Tenenbaum, Brinkmann,
  Tsakanov
• PTs results suggested 140 emittance growth on
  average using this method! budget 50
• Culprit was assumed to be cavity tilts (300mr
  RMS), but is (I believe) actually BPM resolution
  (10mm RMS)

7
DFS

• Find an orbit (trajectory) that minimises
  dispersion
• changing the lattice beam energy match
• measure difference orbit
• using known lattice model, calculate (fit) orbit
  correction to minimise difference orbit

measureddifference
quadrupoleoffsets
linear model
8
DFS

• Find an orbit (trajectory) that minimises
  dispersion
• changing the lattice beam energy match
• measure difference orbit
• using known lattice model, calculate (fit) orbit
  correction to minimise difference orbit

random
measureddifference
quadrupoleoffsets
linear model
9
DFS

• Find an orbit (trajectory) that minimises
  dispersion
• changing the lattice beam energy match
• measure difference orbit
• using known lattice model, calculate (fit) orbit
  correction to minimise difference orbit

random
measureddifference
quadrupoleoffsets
upstreamjitter
linear model
10
DFS Problems

• Fit is ill-conditioned
• with BPM noise DF orbits have very large
  unrealistic amplitudes.
• Need to constrain the absolute orbit

minimise

• Sensitive to initial launch conditions (steering,
  beam jitter)
• need to be fitted out or averaged away

11
DFS for TESLA
The effect of upstream beam jitter on DFS
simulations for the TESLA linac. 1 sy initial
jitter 10 mm BPM noise
45.0
40.0
35.0
norm. vertical emittance (nm)
30.0
25.0
uncorrected cavity tilts cause problems for TESLA
20.0
0
50
100
150
200
250
300
350
Quadrupole
average over 100 random machines
12
Ballistic Alignment

• Turn of all components in section to be aligned
  magnets, and RF
• use ballistic beam to define straight reference
  line (BPM offsets)
• Linearly adjust BPM readings to arbitrarily zero
  last BPM
• restore components, steer beam to adjusted
  ballistic line

62
13
Ballistic Alignment
62
14
New Simulations usingPLACET and MERLIN

• 14 quads per bin (7 cells, Df 7p/3)
• RMS Errors
• quad offsets 300 mm
• cavity offsets 300 mm
• cavity tilts 300 mrad
• BPM offsets 200 mm
• BPM resolution 10 mm
• CM offsets 200 mm
• initial beam jitter 1sy (10 mm)
• New transverse wakefield included(30 reduction
  from TDR)Zagorodnov and Weiland, PAC2003

wrt CM axis
15
Ballistic Alignment

• Less sensitive to
• model errors
• beam jitter

average over 100 seeds
16
Ballistic Alignment
We can tune out linear ltydgt and ltydgt correlation
using bumps or dispersion correction in BDS
average over 100 seeds
17
100 Random Machines
dispersion corrected
18
Ballistic Alignment Problems

• Controlling the downstream beam during the
  ballistic measurement
• large beta-beat
• large coherent oscillation
• Need to maintain energy match
• scale downstream lattice while RF in ballistic
  section is off
• use feedback to keep downstream orbit under
  control

large linac apertures a for TESLA
19
S2E Simulations of Dynamic Errors Ground Motion

• collaborative effort between
• Glen White (QMUL,UK)
• Nick Walker (DESY)
• Daniel Schulte (CERN)

bulk of the work

• primary objectives
• the banana effect and its correction
• intra-train fast feedback
• intra-train lumi optimisation using fast lumi
  monitor

20
Bananas
TESLA high disruption regime long. correlated
emittance growth causes excessive luminosity loss
(banana effect)
Brinkmann, Napoly, Schulte, TESLA-01-16
21
Bananas
TESLA luminosity as a function of linac emittance
growth
Note Dey will contain a correlated component
due to wakefields
D. Schulte. PAC03, RPAB004
22
Beam-Beam Issues
Rigid bunch approximation
D. Schulte. PAC03, RPAB004
23
Beam-Beam Issues
GUINEAPIG resultbanana effect
Now optimise (scan) collision offset and
angle(collision feedback)
D. Schulte. PAC03, RPAB004
24
Beam-Beam Issues
optimise beam-beam offset
D. Schulte. PAC03, RPAB004
25
Beam-Beam Issues
optimise beam-beam offset and angle
OK for static effect dynamic effects still a
problem
D. Schulte. PAC03, RPAB004
26
Simulating the Dynamic Effect
IP FFBK

• Realistic simulated bunches at IP
• linac (PLACET, D.Schulte)
• BDS (MERLIN, N. Walker)
• IP (GUINEAPIG, D. Schulte)
• FFBK (SIMULINK, G. White)
• bunch trains simulated with realistic errors,
  including ground motion and vibration

All bolted together within a MATLAB framework
by Glen White (QMC)
27
Simulating the Dynamic Effect

• Intra-train fast feedback
• modelled realistically using
• bunches from PLACETMERLIN simulations
• realistic beam-beam simulation using GUINEAPIG

Angle feedback kicker modelled correctly in MERLIN
28
Simulating the Dynamic Effect

• LINAC (PLACET, inc. multi-bunch effects)
• static alignment errors randomly added
• RMS values chosen to give design emittance growth
  (10nm) on average
• BDS (MERLIN)
• no static errors currently included
• Ground motion
• first pass effect of random 70nm RMS quadrupole
  jitter
• full correlated ground motion models implemented

29
Simulating the Dynamic Effect

• First 500 bunches of single bunch train modelled
  (18)
• Fast feedback
• first corrects angle BPM and offset beam-beam
  kick lt50 bunches
• attempt lumi optimisation by scanning offset and
  angle
• fast lumi monitor correctly modelled by tracking
  pairs (produced by GUINEAPIG)

30
Simulating the Dynamic Effect
IP beam angle
IP beam offset
31
Simulating the Dynamic Effect
2?1034 cm-2s-1
Only 1 seed need to run many seeds to gain
statistics!
32
By-product of Dynamic Studies
http//hepwww.ph.qmul.ac.uk/lcdata/

• database of beam-beam events (GUINEAPIG)
• quasi realistic beams from linac/BDS simulations
• contains
• spent e beam
• beamstrahlung
• e pairs etc.
• useful for HEP detector studies