EB_2015_10_07_Spacecharge_halfDay

LIU Space Charge Studies at the PSB
E. Benedetto, V. Forte, M. Kowalska
Acknowledgements: J. Abelleira, C. Bracco, GP Di Giovanni, B. Mikulec, D.
Quartullo, G. Rumolo, F. Schmidt
E. Benedetto, SC half Day review, 7/10/15
Outline
• Introduction
• Emittance vs. intensity LIU curve
• Studies in 2014-2015:
•
Transverse emittance optimization for OP beams
•
Longitudinal injection form Linac4
•
Compensation of chicane perturbation
•
Effect of chromaticity
•
Benchmark code vs. measurements
•
Shaving at 160 MeV for sub-micron emittance beams
•
Losses for high intensity beams
• Code development
• MDs
• Ongoing/future studies
Intro – PSB injection Upgrade and SC
• Very large SC tune spread DQ > 0.5 at injection
•
Emittance dominated by SC (and multi-turn injection process)
• Injection from Linac4 at 160 MeV
•
Twice (?) as much brightness for LHC beams  EMITTANCE
•
Possibility to deliver higher intensity to ISOLDE  LOSSES
• H- charge-exchange injection
•
Painting or inject on the closed orbit
•
Avoid losses at the septum
Brightness (Emittance vs. Intensity) curve
Measurements B. Mikulec, et al., LIU Beam
Studies Review, 2012
1.2 eVs
0.8 eVs
•
•
On Ring 3, after
optimization
New measurements
presently ongoing by B.
Mikulec and GP Di Giovanni
• @ 160 MeV the curves should scale by a ~ factor 2
• for a given longitudinal emittance
Main question LIU: Can we confirm the
brightness ~factor 2 scaling?
• Simulations PTC-Orbit
• Transversely MATCHED distribution (Gaussian) (*)
•
•
With a given emittance, scan on the Intensity
Let it evolve for ~7ms, during fall of the chicane bump
• Quadrupolar errors at the chicane magnets + Eddy currents
+ Compensation QDE3, QDE14 (time varying)
•
•
•
Beta-beating (mostly in vertical) corrected
Excitation of half-integer corrected
Excitation of the integer line
• (Quadrupolar errors measured in 2012)
(*) In longitudinal rectangular distribution evolving in an accelerating bucket
Simulations PTC-Orbit
200 SC nodes
2.5 Direct SC module
128x128x128
250k macroparticles
I=350 ppb
• Emittance reached at the end
of the chicane bump is
~“independent” of the
starting value
Simulations PTC-Orbit
1.2 eVs
~1.27 eVs
• On a straight line & depends on longitudinal emittance
• The slope for 1.20eVs is a factor 25% lower
• (in Orange: simulations adding 2012 measured set of errors)
Choice of working point
Simulations @160 MeV, H-inj
Increase of brightness for
Qx=4.43 (4.28 is the
baseline)
Confirmed by production of
OP beams
OP beams optimization - EB May-June ’15
Full blown transverse + longitudinal simulations
including multi-turn H- injection (1)
Ongoing…
• KSW painting bump function input from ABT (J. Abelleira, C. Bracco)
• Longitudinal distribution optimized (chopping factor and dp/p) (V.
Forte, D. Quartullo)
Full blown transverse + longitudinal simulations
including multi-turn H- injection (2)
Ongoing…
• KSW painting bump function input from ABT (J. Abelleira, C. Bracco)
• Longitudinal distribution optimized (chopping factor and dp/p) (V.
Forte, D. Quartullo)
Brightness curve: summary and next steps
• Linear dependence emittance vs. intensity in both measurements and
simulations
• Effect of integer line  beam blow-up
• Simulations suggests improvement by increasing horizontal and
vertical tune, confirmed by measurements (and exploited in operation)
• Full simulations including multiturn H- injection (input form ABT)
confirm the ~factor 2 brightness gain by going to 160 MeV
• Next step/Ongoing: Benchmarking with the present injection
•
Already attempted in the past, not completely successful,
probably due to big uncertainty in the injection parameters
•
New measurements will be carefully taken for the benchmark
(with GP Di Giovanni)
Longitudinal injection from Linac4
• To get optimum chopping factor and dE for injection of LHC beam (no
longitudinal painting)  61% (=609ns), 403 keV
• Including Space Charge
• Including Linac4 distribution and time structure (input from A. Lombardi)
• Benchmark with ESME and BLond (in collaboration with D. Quartullo,
RF)
• Next steps/ongoing: benchmark with present injection (S. Albright, D.
Quartullo, V. Forte)
Compensation of the chicane ramp-down induced
perturbations
• Inconel vacuum chamber in the chicane magnets:
•
Edge effects due to rectangular magnets
Eddy-currents and multipolar components varying with time
• Chicane ramp-down shape implemented in PTC-Orbit, with errors and
ripples  definition of tolerances and the function for the correctors
QDE3,QDE14.
•
Benckmark of code vs. measurements
• “Half-integer” MASTER experiment
• Good agreement simulations vs. measurements, with measured errors
into the model (black line):
Losses are
correlated with
beta-beating
• Last optics studies in 2012 (M. McAteer). Need new measurements:
•
Kick-response matrix (dedicated MD: power all the steerers in 1 ring)
•
Turn-by-turn position measurements with AC dipole excitation to
explore nonlinear optics  planned for November, need BI ready
Effect of chromaticity (measurements +
simulations)
xx=-0.73
xx=-0.15
Case for benckmark purpose:
small SC tunespread and on
the integer
• Same trend in simulations and measurements with simplified model
• Simulations for Upgrade: no improvement
•
•
Chromatic DQ~0.05 << SC DQ~0.5
Only 1 sextupole family (correction in
x gives an increase in y)
But chromaticity control can be useful if
working close to a resonance
Beam shaving at 160 MeV for emittance (intensity)
control
New shaving scheme: more robust  will be implemented also in current
operation
1micron emittance obtained with
shaving at 160 MeV
Losses for high intensity beams
Setting up pyOrbit to:

Study efficiency of newly proposed collimator in presence of SC
− Blow-up and losses in the vertical plane indicate the need of a collimator
to scrape the beam at a given location
− Single absorber, no multistage due to tight space constrains (2p phase
advance = ~ ¼ of the PSB)
− Replace Beam Scope Window (40mm Carbon, not suited to stop protons
at 160 MeV) & reduce aperture due to smaller geometrical emittance.

Check feasibility of more stringent constraint for ISOLDE beams
(input ABT):
− Bottleneck at the recombination septa
− Today ev=8um  if 2x intensity max ev= 6um, for the same amount of
losses… CAN we make it???
Code development and consolidation
PTC-Orbit

PTC tables: time-varying fields and acceleration (PSB specific)

Injection foil + painting (functions given by ABT)

Code debugging (spikes, acceleration)
pyOrbit:

Collimation, apertures and diagnostics for lost particles

Injection foil + painting (functions given by ABT)

Longitudinal dynamics with space charge

Still open points:
− SC nodes, when acceleration is present.
− Collimator added as a child node of PTC, however child nodes are defined
as zero-length elements: how to deal with the finite length of the absorber?
Computing resources
• Computing resources: during peak time, we can easily fill ~20 machines
of the spacecharge cluster
•
Short simulations (~10ms), but parametric studies: happy with the
CERN spacecharge cluster
•
Long simulations (>100ms = few weeks), we need more speed, i.e.
larger # cores with fast connection
CERN, red
CNAF-Bologna, blue (green)
Planned/Ongoing MDs (1)
Optics measurements

Optics model (linear and non linear) of the present machine as input for
Space Charge studies.

Validate the method of the AC dipole excitation to get non-linear optics
informations, in presence of space charge
− Requires turn-by-turn BPM data, electronics currently under commissioning
and an upgraded Transverse Damper to excite the beam oscillations.
Tail repopulation after shaving

Hypothesis mechanism driven by SC

Measure timescale as input for collimation studies
Transverse shaving at >160 MeV
• to produce and preserve <1um emittance beams
Planned/Ongoing MDs (2)
Measurements of the current injection process (with OP)

Transverse emittance as a function of injection parameters

For code benchmark with present injection scheme

Requires measurements of position/angle at injection (accuracy is
currently under study)
Half integer (for the master experiment), scan on:
•
•
•
Corrector strength
Working point
Beam brightness (SC tune spread)
Explore effect of multipoles on beam dynamics
• Both in measurements and simulations
Conclusions (1)
• Main question for PSB: can we achieve 2x brightness with Linac4? 
Simulations indicate yes!
•
Confirmed linear dependence emittance vs intensity & dependence on
longitudinal emittance
•
Suggested increase of horizontal tune
•
Full blown simulations to include multi-turn H- injection, fall of the chicane
bump, optimized longitudinal parameters, realistic distribution form Linac4
are about to be completed
•
Benchmark with current injection  STARTED: new set of measurements
& then simulations
• Studies for High intensity beams (ISOLDE): what is the max intensity we
can deliver within a given emittance and losses constrains?
•
Study of a new collimation to replace the existing aperture restriction
•
Study of shaving at the new collimator to achieve low emittances
Conclusions (2)
• Code development:
•
Still setting-up pyOrbit with few question marks related to PTC interface
• New MDs (and simulations) are on the plate:
•
Most of them require help from Beam Instrumentation (BLMs, Pick-up at
injection, Turn-by-turn PUs) and RF (AC dipole MD) and we thank in
advance the experts for their help!
• On the long term:
•
Complete the machine model (linear and non-linear) for input to SC
simulations and for implementing a resonance compensation
scheme
•
Optimize machine working point (including chromaticity)
…To deliver the brightest possible beam to
LHC and to minimize uncontrolled losses
SC Mitigation Measures
• Double harmonic RF: h1+h2
• Acceleration (no energy flat bottom)
– H- injection directly on accelerating bucket
– Today: MT injection in coast, then adiabatic capture + acceleration
• Transverse painting:
– Horiz. Painting + Vert. Steering
– Today: injection offset in both planes (V steering and delay of the bump
decay wrt injection timing)
• Working point variation with time
• Resonance compensation:
– Empirical: based on loss reduction
– Systematic studies: driving terms and kick response matrix
E.Benedetto, 14/11/14,HB2014 Transverse emittance preservation in PSB
Space-Charge simulations (PTC-Orbit)
Red: initial 350e10 ppb,1mm
Blue: final (after 7ms) for the same beam.
Green: initial 350e10 ppb, 1.7 mm..
H and V Tune spread
Initial DQx extending below the integer  blow-up
Measured Space Charge tune spread and losses
mitations in the PSB
njection
ms
(with
B. Mikulec, et al, HB 2012