Download - Electron cloud and scrubbing in the LHC
Electron cloud and scrubbing in the LHCG. Arduini, H.Bartosik, G. Iadarola, G. Rumolo
LHC Operation Meeting, Evian, 19/12/2012
Many thanks to:Many thanks to Cryogenics, Damper, EN/STI, Injection,
Operation, Vacuum teamsSeveral ABP, BI and RF colleagues contributing to measurements
Outline
• Introduction on electron cloud (EC) and scrubbing effects
• EC effects in the LHC: 50ns vs. 25ns
• Experience with 25ns beams:
• Observations in 2011
• The 2012 Scrubbing Run (450GeV)
• 25ns beams at 4TeV• Scenarios for post-LS1 operation
• Conclusions
Outline
• Introduction on electron cloud (EC) and scrubbing effects
• EC effects in the LHC: 50ns vs. 25ns
• Experience with 25ns beams:
• Observations in 2011
• The 2012 Scrubbing Run (450GeV)
• 25ns beams at 4TeV• Scenarios for post-LS1 operation
• Conclusions
Introduction
When the an accelerator is operated with close bunch spacing an Electron Cloud
(EC) can develop in the beam chamber due to the Secondary Emission from the
chamber’s wall.
0 200 400 600 800 10000.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
Primary e- energy [eV]
Seco
ndar
y El
ectro
n Yi
eld
[SEY
]
SEYmax
Secondary Electron Yield (SEY) of the
chamber’s surface:
• ratio between emitted and impacting
electrons
• function of the energy of the primary
electron
Introduction
When the an accelerator is operated with close bunch spacing an Electron Cloud
(EC) can develop in the beam chamber due to the Secondary Emission from the
chamber’s wall.
• Strong impact on beam quality (EC
induced instabilities, particle losses,
emittance growth)
• Dynamic pressure rise
• Heat load (on cryogenic sections)
Dipole chamber @ 7TeV
Introduction: electron cloud effects and scrubbing
10-8 10-6 10-4 10-2 1001
1.2
1.4
1.6
1.8
2
SEY m
ax
Dose [C/mm2]
Scrubbing of Cu measured with
e- at 500eV (CERN TE-VSC)
Scrubbing is a mitigation for the e-cloud effects:
• Keeping a significant e- flux on the chamber’s walls causes a decrease of the
SEY (and hence the e-cloud)
• The dependence of the SEY on the accumulated dose is logarithm - like
Outline
• Introduction on electron cloud (EC) and scrubbing effects
• EC effects in the LHC: 50ns vs. 25ns
• Experience with 25ns beams:
• Observations in 2011
• The 2012 Scrubbing Run (450GeV)
• 25ns beams at 4TeV• Scenarios for post-LS1 operation
• Conclusions
Electron cloud effects in the LHC dipoles: 50ns vs. 25ns
• Main focus on the dipole magnets (~60% of the machine) they determine the performance in terms of beam quality
1 1.5 2 2.510
-6
10-4
10-2
100
102
SEY
Elec
tron
flux
[mA
/m]
25ns50ns
LHC- dipole magnet - 450GeV
• The “multipacting threshold” for 25ns beams is significantly lower than for 50ns
• In 2011, 4 days of scrubbing with 50ns beams + 2 days of tests with 25ns beams have lowered the SEY in the arcs well below 2.0 allowing an “EC free” operation also in 2012
Outline
• Introduction on electron cloud (EC) and scrubbing effects
• EC effects in the LHC: 50ns vs. 25ns
• Experience with 25ns beams:
• Observations in 2011
• The 2012 Scrubbing Run (450GeV)
• 25ns beams at 4TeV• Scenarios for post-LS1 operation
• Conclusions
Electron cloud effects with 25ns: flashback into 2011
up to ±5mm
~ bunch 25 is the first unstable
First injection tests with a train of 48b. on 26/08/2011:
• Beam unstable right after injection (dump due to losses)
• Probably driven by e-cloud in the dipoles (mainly vertical motion, trailing bunches of the train)
• Beam stable with high chromaticity settings (Q’=15)
Thanks to W. Höfle and D. Valuch
Electron cloud effects with 25ns: flashback into 2011
up to ±5mm
~ bunch 25 is the first unstable
First injection tests with a train of 48b. on 26/08/2011:
• Beam unstable right after injection (dump due to losses)
• Probably driven by e-cloud in the dipoles (mainly vertical motion, trailing bunches of the train)
• Beam stable with high chromaticity settings (Q’=15)
Thanks to W. Höfle and D. Valuch
The instability could be successfully reproduced in HEADTAIL simulations (with electron distributions from PyECLOUD)
bunch 26 is the first unstable
Electron cloud effects with 25ns: flashback into 2011
First scrubbing tests with 25ns (14 & 24-25 Oct. 2011):• Injected up to 2100b. for B1 and 1020b. for B2
• SEY in the arcs could be decreased down to 1.52
0 10 20 30 40 500
5
10
15
20 x 1013
Inte
nsity
Time [h]
0 10 20 30 40 50
1.4
1.6
1.8
2
2.2
max
Time [h]
29/06 07/10 24-25/1014/10
50ns
0 10 20 30 40 500
5
10
15
20 x 1013
Inte
nsity
Time [h]
0 10 20 30 40 50
1.4
1.6
1.8
2
2.2
max
Time [h]
25ns threshold @450 GeV
Reconstructed comparing heat load meas. and PyECLOUD sims.
Electron cloud effects with 25ns: flashback into 2011
0 10 20 30 40 500
5
10
15
20 x 1013
Inte
nsity
Time [h]
0 10 20 30 40 50
1.4
1.6
1.8
2
2.2
max
Time [h]
29/06 07/10 24-25/1014/10
50ns
0 10 20 30 40 500
5
10
15
20 x 1013
Inte
nsity
Time [h]
0 10 20 30 40 50
1.4
1.6
1.8
2
2.2
max
Time [h]
25ns threshold @450 GeV
Reconstructed comparing heat load meas. and PyECLOUD sims.
Beam degradation still important:
• Poor bunch by bunch lifetimes• Large emittance growth
Thanks to F. Roncarolo
First scrubbing tests with 25ns (14 & 24-25 Oct. 2011):• Injected up to 2100b. for B1 and 1020b. for B2
• SEY in the arcs could be decreased down to 1.52
Electron cloud effects with 25ns: flashback into 2011
0 10 20 30 40 500
5
10
15
20 x 1013
Inte
nsity
Time [h]
0 10 20 30 40 50
1.4
1.6
1.8
2
2.2
max
Time [h]
29/06 07/10 24-25/1014/10
50ns
0 10 20 30 40 500
5
10
15
20 x 1013
Inte
nsity
Time [h]
0 10 20 30 40 50
1.4
1.6
1.8
2
2.2
max
Time [h]
25ns threshold @450 GeV
Reconstructed comparing heat load meas. and PyECLOUD sims.
An important benchmark for our models came from bunch by bunch energy loss estimation based on stable phase measurements.
2420 2440 2460 2480 25000
0.5
1
1.5
2B
unch
ene
rgy
loss
[mJ/
Turn
]
25ns bucket number
SimulatedMeasured
Thanks to J. Esteban-Müller and E. Shaposhnikova
First scrubbing tests with 25ns (14 & 24-25 Oct. 2011):• Injected up to 2100b. for B1 and 1020b. for B2
• SEY in the arcs could be decreased down to 1.52
Outline
• Introduction on electron cloud (EC) and scrubbing effects
• EC effects in the LHC: 50ns vs. 25ns
• Experience with 25ns beams:
• Observations in 2011
• The 2012 Scrubbing Run (450GeV)
• 25ns beams at 4TeV• Scenarios for post-LS1 operation
• Conclusions
Preliminary!
Outline
• Introduction on electron cloud (EC) and scrubbing effects
• EC effects in the LHC: 50ns vs. 25ns
• Experience with 25ns beams:
• Observations in 2011
• The 2012 Scrubbing Run (450GeV)
• 25ns beams at 4TeV• Scenarios for post-LS1 operation
• Conclusions
The 2012 scrubbing run
10 20 30 40 50 60 70 800
0.5
1
1.5
2
2.5x 1014
Time [h]
Tota
l int
ensi
ty [p
]
10 20 30 40 50 60 70 80
0
20
40
60
Time [h]
Hea
t loa
d [W
/hc]
3.5 days of scrubbing with 25ns beams at 450GeV (6 - 9 Dec. 2012):
• Regularly filling the ring with up to 2748b. per beam (up to 2.7x1014 p)
• Overall very good efficiency: injection rate determined by MKI vacuum interlocks (in the beginning) and by time required by the cryogenic system to adapt to the increasing heat load (mainly in stand alones)
The 2012 scrubbing run
10 20 30 40 50 60 70 800
0.5
1
1.5
2
2.5x 1014
Time [h]
Tota
l int
ensi
ty [p
]
10 20 30 40 50 60 70 80
0
20
40
60
Time [h]
Hea
t loa
d [W
/hc]
Scrubbing effects in the arcs:
• Quite rapid conditioning observed in the first stages• The SEY evolution significantly slows down during the last scrubbing
fills (more than expected by estimates from lab. measurements and simulations)
Thanks to L. TavianSector 5-6
10 20 30 40 50 60 70 800
0.5
1
1.5
2
2.5x 1014
Time [h]
Tota
l int
ensi
ty [p
]
10 20 30 40 50 60 70 800
0.5
1
1.5
2
2.5x 10-13
Time [h]
HL/
I [W
/(hc
p)]
The 2012 scrubbing run
Scrubbing effects in the arcs:
• Quite rapid conditioning observed in the first stages• The SEY evolution significantly slows down during the last scrubbing
fills (more than expected by estimates from lab. measurements and simulations)
Thanks to L. Tavian
10 20 30 40 50 60 70 800
0.5
1
1.5
2
2.5x 1014
Time [h]
Tota
l int
ensi
ty [p
]
10 20 30 40 50 60 70 801.35
1.4
1.45
1.5
1.55
1.6
Time [h]
SEY
max
Reconstructed comparing heat load meas. and PyECLOUD sims.
The 2012 scrubbing run
Scrubbing effects in the arcs:
• Quite rapid conditioning observed in the first stages• The SEY evolution significantly slows down during the last scrubbing
fills (more than expected by estimates from lab. measurements and simulations)
10 20 30 40 50 60 70 800
0.5
1
1.5
2
2.5x 1014
Time [h]
Tota
l int
ensi
ty [p
]
10 20 30 40 50 60 70 801.35
1.4
1.45
1.5
1.55
1.6
Time [h]
SEY
max
Reconstructed comparing heat load meas. and PyECLOUD sims.
The 2012 scrubbing run
Scrubbing effects in the arcs:
• Quite rapid conditioning observed in the first stages• The SEY evolution significantly slows down during the last scrubbing
fills (more than expected by estimates from lab. measurements and simulations)
1.3 1.4 1.5 1.6 1.7 1.8
50
100
150
200
250
300
SEYmax
Sim
ulat
ed h
eat l
oad
[W/h
c]
Measured
Scrubbing effects:
• Evident improvement on beam lifetime
Sunday 10 Dec
Friday 7 Dec
The 2012 scrubbing run
Scrubbing effects:
• Evident improvement on beam lifetime
The 2012 scrubbing run
Scrubbing effects:
• Evident improvement on beam lifetime
• Important losses do no appear before 1h after injection
Beam 1
Beam 2
The 2012 scrubbing run
Scrubbing effects:
• Evident improvement on beam lifetime• After the scrubbing, emittances are ok for trains of 72b. (3.6µs spacing)
but a significant blow-up is still visible when longer trains are injected
Fill with trains of 72b. Fill with trains of 288b.
Thanks to T. Rijoff and H. Maury Cuna
The 2012 scrubbing run
Outline
• Introduction on electron cloud (EC) and scrubbing effects
• EC effects in the LHC: 50ns vs. 25ns
• Experience with 25ns beams:
• Observations in 2011
• The 2012 Scrubbing Run (450GeV)
• 25ns beams at 4TeV• Scenarios for post-LS1 operation
• Conclusions
Electron cloud observations at 4TeV
0 2 4 6 8 10 12 14 16 18x 104
0
2
4
6
8
Time [h]
Inte
n. [p
x10
13],
Ener
gy [T
eV]
0 2 4 6 8 10 12 14 16 18x 104
0
20
40
60
Time [h]
Hea
t loa
d [W
/hc]
84b.
156b. 372b.
804b.804b.
After Scrubbing Run machine studies with 25ns beams at 4TeV were possible. Main observations:
• The heat-load strongly increases during the ramp since the EC is enhanced by the photoelectrons due to synchrotron radiation This violent transient on the heat load in the arcs limits the number of bunches which can be accelerated
After Scrubbing Run machine studies with 25ns beams at 4TeV were possible. Main observations:
• Despite the larger number of electrons, at high energy the beam becomes less affected by EC the beam quality achievable at collisions is determined by the EC effects at 450GeV
Thanks to T. Rijoff and H. Maury Cuna
Begin. flat top After 8h. At flat top
Electron cloud observations at 4TeV
0 5 10 15 20 25 30 35 40 45-2
0
2
4
6
8
Time [h]
Inte
n. [p
x10
13],
Ener
gy [T
eV]
108b. 204b. 396b. 396b.
780b.
A pilot physics run with 25 ns beams took place in the last two days of the 2012 run:
• High brightness BCMS 25ns beam from the injectors (batches of 48 bunches 1.1e11ppb within transverse emittances of ~1.4µm)
• Stable beams with 396b. (squeezed up to 780b.)
Electron cloud observations at 4TeV
Electron cloud effects with 25ns: 2012 experience
0 5 10 15 20 25 30 35 40 45-2
0
2
4
6
8
Time [h]
Inte
n. [p
x10
13],
Ener
gy [T
eV]
108b. 204b. 396b. 396b.
780b.
From M. Lamont, LHC Morning Meeting, 17/12/2012
A pilot physics run with 25 ns beams took place in the last two days of the 2012 run:
• High brightness BCMS 25ns beam from the injectors (batches of 48 bunches 1.1e11ppb within transverse emittances of ~1.4µm)
• Stable beams with 396b. (squeezed up to 780b.)
Electron cloud effects with 25ns: 2012 experience
0 5 10 15 20 25 30 35 40 45-2
0
2
4
6
8
Time [h]
Inte
n. [p
x10
13],
Ener
gy [T
eV]
108b. 204b. 396b. 396b.
780b.
From M. Lamont, LHC Morning Meeting, 17/12/2012
From M. Lamont, LHC Morning Meeting, 17/12/2012
A pilot physics run with 25 ns beams took place in the last two days of the 2012 run:
• High brightness BCMS 25ns beam from the injectors (batches of 48 bunches 1.1e11ppb within transverse emittances of ~1.4µm)
• Stable beams with 396b. (squeezed up to 780b.)
Outline
• Introduction on electron cloud (EC) and scrubbing effects
• EC effects in the LHC: 50ns vs. 25ns
• Experience with 25ns beams:
• Observations in 2011
• The 2012 Scrubbing Run (450GeV)
• 25ns beams at 4TeV• Scenarios for post-LS1 operation
• Conclusions
Possible scrubbing scenarios for 2015
Comm. Vacuum conditioning
(50ns or short trains 25ns) (5-7 days)
Scrubbing with 25ns(2 days)
50ns physics
δarc≥2.2 δarc≈2.2 δarc≤2.
450GeV 6.5TeVRequirements for operation with 50ns beams:
Possible scrubbing scenarios for 2015
Comm. Vacuum conditioning
(50ns or short trains 25ns) (5-7 days)
Scrubbing with 25ns(2 days)
50ns physics
δarc≥2.2 δarc≈2.2 δarc≤2.
450GeV 6.5TeV
Scrubbing with 25ns(5 days)
Test ramps with 25ns(2 days)
δarc≤2. δarc≈1.45
450GeV 6.5TeV
Requirements for operation with 50ns beams:
Further requirements for operation with 25ns beams:
Same conditions as in 2012 pilot physics run but with enhanced heat load due to higher energy
Possible scrubbing scenarios for 2015
Comm. Vacuum conditioning
(50ns or short trains 25ns) (5-7 days)
Scrubbing with 25ns(2 days)
50ns physics
δarc≥2.2 δarc≈2.2 δarc≤2.
450GeV 6.5TeV
Scrubbing with 25ns(5 days)
Test ramps with 25ns(2 days)
25ns physics (intensity ramp up and further scrubbing)
δarc≤2. δarc≈1.45
450GeV 6.5TeV
Requirements for operation with 50ns beams:
Further requirements for operation with 25ns beams:
Same conditions as in 2012 pilot physics run but with enhanced heat load due to higher energy
Possible scrubbing scenarios for 2015
Comm. Vacuum conditioning
(50ns or short trains 25ns) (5-7 days)
Scrubbing with 25ns(2 days)
50ns physics
δarc≥2.2 δarc≈2.2 δarc≤2.
450GeV 6.5TeV
Scrubbing with 25ns(5 days)
Test ramps with 25ns(2 days)
25ns physics (intensity ramp up and further scrubbing)
δarc≤2. δarc≈1.45
450GeV 6.5TeV
Requirements for operation with 50ns beams:
Further requirements for operation with 25ns beams:
Remark:In case of long periods without beam de-conditioning is likely to take place:
• No action needed for 50ns beams (due to margin taken by scrubbing with 25ns beam)
• Few hours reconditioning required for 25ns beams
Summary
• The 2012 Scrubbing Run has lowered the SEY in the arcs to less than 1.45 resulting
in a reduced heat load as well as improved beam quality (lifetime, emittances)
• In spite of the high heat loads, close to the cryogenics limit, the second part of the
Scrubbing Run and the ramps to 4 TeV did not exhibit any clear improvement in
the conditioning state of the arcs (to be investigated with additional simulations
and lab measurements)
• Emittance blow-up ascribable to EC is still observed at injection energy with long
and closely spaced trains
• At 4TeV, no indication of further emittance deterioration driven by EC
• A concentrated scrubbing run will be likely to be insufficient to fully suppress the
EC from the arcs for 25ns beams in future operation.
Thanks for your attention!