Transcript Electron Lens at the LHC

Open Issues from the
SPS Long-Range Experiments
Frank Zimmermann
US-LARP Beam-Beam Workshop
SLAC, 2007
Gerard Burtin, Ulrich Dorda, Gijs de Rijk,
Jean-Pierre Koutchouk, Yannis Papaphilippou, Tannaji Sen,
Vladimir Shiltsev, Jorg Wenninger, + many others
outline
• motivation & scaling
• single wire as LHC LR simulator
(2002-2004)
• two wire compensation (2004)
• test of crossing schemes (2004)
• open questions and 2007 plan
Motivation: Long-Range BeamBeam Compensation for the LHC
• To correct all non-linear effects correction must be local.
• Layout: 41 m upstream of D2, both sides of IP1/IP5
Phase difference between BBLRC &
average LR collision is 2.6o
(Jean-Pierre Koutchouk)
APC meeting, 19.09.03, LRBB
J.P. Koutchouk, J. Wenninger, F. Zimmermann, et al.
1st Wire “BBLR” in the SPS
Tech. Coord.
J. Camas &
G. Burtin/BDI
Help from
many groups
two 60-cm long wires
with 267 A current
equivalent to 60 LHC LR
collisions (e.g., IP1 & 5)
Iwire=Nb e c #LR/lwire
wire current
wire length
each BBLR consists of 2 units, total length:
2x0.8+0.25=1.85 m
nominal
distance
19 mm
(in the
shadow
of the arc
aperture)
water cooling
Scaling from LHC to SPS
perturbation by wire:
y 
'
2rplw I w
ec( y  d )
relative perturbation:
y
'
 y'
 2rplw  I w 


 
~ 
ec
(

)
n


da 
for constant normalized emittance the effect in units
of sigma is independent of energy and beta function!
in simulations LHC long-range collisions & SPS
wire cause similar fast losses at large amplitudes
simulation with WSDIFF
simulation with WSDIFF
SPS wire
LHC beam
1 mm/s
1 mm/s
diffusive aperture
a few technical issues
 dedicated ion chambers and PMTs
 inductive coil to suppress wire ripple
 wire heating computed and verified
 emittance blow up by damper or injection mismatch
to reproduce LHC or to increase sensitivity
 wire scanners, scrapers
 dedicated dipole corrector to correct orbit change
locally
 always correct tune
changes in orbit & tunes (2002)
→ precise measure of beam-wire distance
d 
y orbit
change
 y I wlwrp
ecd  d  tan Qy  Qy 
Qx, y  
rp  x, y I wlw
2ec
1
2
d  d 
J. Wenninger
x tune change
y tune
change
non-linear optics
• turn-by-turn BPM data after kicks of
various amplitude
→ reduced decoherence time due to wire
→ tune shift with amplitude, roughly
consistent with
3 I wlw rp  x 2
Qx 
yˆ
4
4 ec d
3 I wlwrp  x 2
Qy  
yˆ
4
8 ec d
measuring the “diffusive” or
dynamic aperture
three types of signals:
• lifetime and background
• final emittance
• scraper-retraction
lifetime and background
lifetime vs. separation
beam loss vs. separation
drop in the lifetime and increased losses for
separations less than 9;
at 7-8 separation lifetime decreases to 1-5 h
initial & final profile
wire scans
initial/final emittance = 3.40/1.15 mm
Abel transformation of wire-scan data gives change in
(norm.) amplitude distribution:
R
 ( A)  2 A d
A
g ' ( )
 A
2
2
(Krempl, Chanel, Carli)
final emittance
mechanical scraping by edge of wire
calibration of final emittance by scraper
Calibration curve of measured final
emittance vs scraper position allows us
to estimate effective aperture due to
BBLR excitation
scaling to LHC – d.a. only 2-3?
larger emittance
variation when
wire is excited?!
effect of wire current on SPS dyn.ap.
linear dependence consistent with Irwin scaling law;
measured dynamic aperture is smaller than the simulated
scraper retraction attempt
only BBLR (at 12725 ms),
w/o scraping
BBLR at 12725 ms,
scraping at 13225 ms
PMT
PMT
BCT
BCT
on the right, scraper position is about 1;
at larger amplitudes the diffusion seems
much faster than the speed of the scraper
can we fit a
diffusion
constant?
scraper moving to target position
already intercepts halo
effect of beam-wire distance on lifetime
5th power!
d
  5 ms  
 
5
extrapolation to LHC beambeam distance, ~9.5, would
predict 6 minutes lifetime
effect at low wire excitation
BBLR logbook 4 July 2003
“…
We bumped exactly –8.2 mm at the BBLR from cycle#
330616 (this would give 12.1 mm separation between
wire center and beam [~5], corresponding to the
latest simulations). The interpolated position with wire off
was –8.6 mm. The spread in the BPM readings was
about +/-0.2 mm. … The wire current was only -10 A
[~2 LR collisions in LHC]. Nevertheless, the losses
were high, about 3x106 at the 3rd PMT (last year we
had about 106 as the maximum integrated reading). “
for 2004 two novel 3-wire BBLRs were built;
separated from 1-wire BBLR by about 2.6o
(average LR-BBLR phase advance in LHC)
G. Burtin
remotely
movable
in Y by
5 mm!
two-wire compensation: tune scan
Qx=0.31
beam lifetime
3rd
10th
no wire
7th
2 wires
4th
1 wire
vertical tune
what happens here?
nearly perfect
compensation
lifetime is recovered over a large tune range, except for Qy<0.285
two-wire compensation: distance scan
BBSIM (T. Sen): No compensation beyond ~3mm
Measurement: Compensation lost beyond ~2.5mm from optimum
“scaled” experiments
natural SPS beam lifetime
~30 h at 55 GeV/c
~5-10 min at 26 GeV/c
(physical aperture ~4 )
to improve beam lifetime at 26 GeV/c,
emittance can be reduced by scraping;
lifetime for N~1.5 mm
Iw ~  ,
improves to ~1 h
d~ 
scaled two-wire compensation: lifetime
Lifetime
7000
lifetime in s
6000
versus
cycle
number
~69 min.
compensation
no beam
5000
beam
~36 min.
4000
excitation
3000
~61 min.
2000
1000
LHC tunes
178390
178400
178410 178420 178430
cycle number
178440
178450
J.-P. Koutchouk
crossing schemes
crossing schemes – motivation 1 EPAC’04
LHC
here tunes w/o
beam-beam were
held constant
centre
of other
beam
diffusive aperture
with xx or yy
crossing
diffusive aperture
with alternating
crossing
simulation
comparing xy, xx and yy crossing for two working points
xx
yy
8
8.5
simulations for different lattice
tunes, located along red line:
xy
6.5
crossing schemes – motivation 2
det(M)>0
det(M)<0
bounded
fast escape
model
system
 Qx , y 

M 
 I 
 x, y 
 2H 
 2 
 I 
 x, y 
tune evolution for three
tune evolution for three
trajectories without folding; trajectories with folding;
the motion remains bounded the resonance 1:1 is a
direction of fast escape
(J. Laskar, PAC2003)
schematic of folded frequency map
(J. Laskar)
EPAC’04
crossing schemes – motivation 3
nonlinear ‘coupling’ between
the planes? but stable
little motion at small
amplitudes but particle
loss at 6 
sample trajectories projected on amplitude plane
tune spread gives incomplete characterization of the dynamics;
experimental simulations of the two crossing schemes can be
compared at the SPS
EPAC’04
crossing schemes – motivation 4
xx
frequency
maps
for nominal
LHC tunes
crossing schemes – motivation 5
xy
simulations
yy
thanks to
Yannis
Papaphilippou
for his help in
calculating
frequency
maps!
crossing schemes – motivation 6
in most cases simulated diffusive aperture along
diagonal x=y larger for equal-plane crossing
than for alternating crossing*, sensitivity to IPIP phase advance
possible explanations:
(1) different ‘folding’ since xy crossing cancels
dodecapole and 20-pole terms in addition to
linear tune shift;
(2) twice the number of resonances for xy
crossing
*(similar result for y=0 – to be revisited)
crossing scheme test – configuration 1
BBLR2x on
(strength x2) beam
xx
BBLR2x on
“xy”
&
beam
BBLR1 off
x bump -23 mm
BBLR1 on
x bump -23 mm
“xy-2”(strength x2)
BBLR2x off
“yy”
beam
BBLR1 on
(strength x2)
x bump -23 mm
xx
xy
yy
xy(x2)
simulation
simulated diffusive aperture for XX crossing is 10% larger than for
‘quasi-XY’ or ‘quasi-YY’ crossing
experiment
xx
xy
yy
xy(x2)
measured beam lifetime is best for XX crossing, second best for
‘quasi-YY’ crossing, lowest for ‘quasi-XY’ crossing
lifetime without wire excitation was comparable to xy case
crossing scheme test – configuration 2
BBLR1 (rotated) & BBLR2 (45 degrees)
BBLR2-45 on
J.-P. Koutchouk
BBLR2x-45 off
BBLR1 on
BBLR1 on (strength x2)
beam
beam
“45o135o”
“45o45o”
x bump -8.9 mm
y bump +11.4 mm
BBLR2x-45 off
x bump -8.9 mm
y bump +11.4 mm
reduced emittance
“scaled” experiment
BBLR1 on (strength x2)
beam
“yy”
x bump 0 mm
y bump +8.5 mm
yy
45o135o
45o45o
simulation
simulated diffusive aperture for ‘45o45o’ crossing is worst; at tunes
below 0.29 it is best for YY crossing & above 0.30 for ‘45o135o’
experiment
w/o BBLR
yy
45o135o
45o45o
measured beam lifetime is worst for ‘45o45o’ crossing, and at tunes
above 0.3 best for ‘45o135o’ crossing
relative beam lifetimes consistent with simulations
some open questions
•
•
•
•
•
•
•
•
•
•
•
•
•
•
scaling from SPS to LHC
strong emittance dependence of lifetime (c.f.Tevatron pbar)
discrepancies between measured & simulated dynamic aperture
breakdown of 2-wire compensation for Qy<0.285
why 5th power law? (Tevatron: 3rd power, RHIC: 2nd and 4th power); why
different & why not higher power??
some effect observed at very low wire excitation
amplitude-dependent diffusion rate
study sensitivity of final emittance to tune with and without BBLR
discrepancies between simulated and measured lifetime (improved at higher
beam energy?)
understand parameters which are out of control or introduce intentional large
perturbation (excite sextupoles, octupoles) to reconcile experiments and
measurements
wire compensation test with colliding beams (at RHIC) (essential?)
common observable in experiments & simulations? – dynamic aperture!
lifetime?
demonstrate that 10-4 stability of pulsed wire can be achieved
crossing scheme conclusions?
2007 SPS MD plan
 lifetime/emittance growth vs beam-wire distance at different
wire currents
 tune scan of wire compensation at higher energy with longer
unperturbed lifetimes
 study compromise between nominal and PACMAN bunches
by partial compensation
 use both wires as exciters at different beam-wire separation to
mimic LRBB at different beam-beam separation (crucial
issue for the early separation upgrade scheme)
 beam lifetime vs. beam-wire distance for different tunes to see
(understand) whether different power laws found at SPS
(^5), Tevatron (^3) and RHIC (^2) and (^4) are tune related
 noise studies (if more than 2 MDs) to experimentally verify the
simulated precision requirements on a pulsed device
 experiments will be performed at two different energies
(26 GeV and 55 GeV) to confirm the theoretical scaling law
for future wire
beam-beam
compensators
- “BBLRs” -,
3-m long sections
have been reserved
in LHC at 104.93 m
(center position)
on either side of
IP1 & IP5
references
 J.-P. Koutchouk, Principle of a Correction of the Long-Range Beam-Beam Effect in LHC using
Electromagnetic Lenses, LHC Project Note 223, 2000
 J.-P. Koutchouk, Correction of the Long-Range Beam-Beam Effect in LHC using Electromagnetic Lenses,
SL Report 2001-048, 2001
 F. Zimmermann, Weak-Strong Simulation Studies for the LHC Long-Range Beam-Beam Compensation,
presented at Beam-Beam Workshop 2001 FNAL; LHC Project Report 502 (2001)
 J. Lin, J. Shi, W. Herr, Study of the Wire Compensation of Long-Range Beam-Beam Interactions in LHC
with a Strong-Strong Beam-Beam Simulation, EPAC 2002, Paris (2002)
 J.-P. Koutchouk, J. Wenninger, F. Zimmermann, Compensating Parasitic Collisions using Electromagnetic
Lenses, presented at ICFA Beam Dynamics Workshop on High-Luminosity e+e- Factories ("Factories'03")
SLAC; in CERN-AB-2004-011-ABP (2004)
 J.-P. Koutchouk, J. Wenninger, F. Zimmermann, Experiments on LHC Long-Range Beam-Beam
Compensation in the SPS, EPAC'04 Lucerne (2004)
 F. Zimmermann, Beam-Beam Compensation Schemes, Proc. First CARE-HHH-APD Workshop on Beam
Dynamics in Future Hadron Colliders and Rapidly Cycling High-Intensity Synchrotrons (HHH-2004), CERN,
Geneva, Switzerland, CERN-2005-006, p. 101 (2005)
 F. Zimmermann, J.-P. Koutchouk, F. Roncarolo, J. Wenninger, T. Sen, V. Shiltsev, Y. Papaphilippou,
Experiments on LHC Long-Range Beam-Beam Compensation and Crossing Schemes at the CERN SPS in
2004, PAC'05 Knoxville (2005)
 F. Zimmermann and U. Dorda, Progress of Beam-Beam Compensation Schemes, Proc. CARE-HHH-APD
Workshop on Scenarios for the LHC Luminosity Upgrade (LHC-LUMI-05), Arcidosso, Italy (2005)
 U. Dorda and F. Zimmermann, Simulation of LHC Long-Range Beam-Beam Compensation with DC and
Pulsed Wires (Talk), RPIA2006 workshop, KEK, Tsukuba, 07-10.03.2006 (2006)
 F. Zimmermann, Possible Uses of Rapid Switching Devices and Induction RF for an LHC Upgrade (Talk),
RPIA2006 workshop, KEK, Tsukuba, 07-10.03.2006 (2006)
 U. Dorda, F. Zimmermann et al, Assessment of the Wire Lens at LHC from the current Pulse Power
Technology Point of View (Talk), RPIA2006 workshop, KEK, Tsukuba, 07-10.03.2006 (2006)