Transcript LHC experience with different bunch spacings in 2011 (25, 50 & 75 ns): Electron cloud aspects Giovanni Rumolo, G.

LHC experience with different bunch
spacings in 2011 (25, 50 & 75 ns):
Electron cloud aspects
Giovanni Rumolo,
G. Iadarola, O. Dominguez, G. Arduini
in LHC Performance Workshop (Chamonix 2012), 6 February 2012
For all LHC data shown (or referred to) in this presentation and discussions:
V. Baglin, H. Bartosik, P. Baudrenghien, G. Bregliozzi, S. Claudet, J. EstebanMüller, W. Höfle, G. Lanza, T. Mastoridis, G. Papotti, F. Roncarolo, E.
Shaposhnikova, L. Tavian, D. Valuch
Introduction
Focus of this talk  Results of the analysis of the 2011 electron cloud
observations and measurements
ex,y decreasing
towards 1.1mm
First 1380
bunches in LHC
21/02
13/03
05/04 12/04
28/06
18/07
Nb increasing
towards
1.45 x 1011 ppb
30/09
30/10
Commissioning
with beam
75ns physics
run (nominal)
Scrubbing
run 50ns
Physics run
50ns
Nominal:
1.1 x 1011 ppb
2.5 mm
Nominal 50ns
beams
2
Introduction
Focus of this talk  Results of the analysis of the 2011 electron cloud
observations and measurements
25ns MDs
(nominal)
07-14-24/10
21/02
13/03
05/04 12/04
29/06
26/08
30/10
Commissioning
with beam
75ns physics
run (nominal)
Scrubbing
run 50ns
Physics run
50ns
Nominal:
1.1 x 1011 ppb
2.5 mm
– 75ns operation  no electron cloud observations in 2011
– 50ns operation
→ Observations during scrubbing
→ Physics operation with residual electron cloud activity
– 25ns MDs: evolution of dmax in the arcs and uncoated SS
– Estimation of scrubbing time and closing remarks
3
Electron cloud observables
Electron flux to the chamber wall Fe
PRESSURE RISE
Beam chamber
POWER ON THE CHAMBER WALL
4
Electron cloud observables
Electron cloud with density re
around the beam
COHERENT INSTABILITY
– Affects only the
last bunches of
each batch
– Can be single or
coupled bunch
Beam
Beam chamber
INCOHERENT EMITTANCE GROWTH
– Causes degrading lifetime
and slow beam loss
– Typically associated to bunch
shortening and loss pattern
increasing along the batch
5
Scrubbing run in 2011
⇒ The scrubbing run took place in the week 5–12 April 2011
− Nominal 50ns spaced beams with up to 1020 bunches per beam
injected into the LHC and stored at 450 GeV/c
⇒ Very efficient machine cleaning
– The dynamic vacuum decreased by one order of magnitude over 17h
of effective beam time (i.e. 72h machine time)
– The heat load on the beam screen in the arcs
→ significant at the beginning of the scrubbing run
→ within measurement resolution at the end
– The average stable phase decreased by one order of magnitude
– Instabilities and emittance growth, visible during the first fills,
disappeared later even with low chromaticity settings
⇒ After scrubbing, physics with 50ns and stable beams with 1380
bunches per beam on 28 June 2011
6
dmax in the arcs:
estimation technique
Beam 1
13
10
x 10
09/04
Beam 2
Energy
13/04
Intensity
8
6
4
2
0
0
5
Before 50ns
scrubbing
15
Time [h]
Two snapshots before (09/04) and
after (13/04)
the heat
scrubbing
Average
load run to
reproduce the measured heat load
[x 10 mW/m/beam]
by means of simulations!
6
Av. heat load [W/hcell]
10
5
4
20
After 50ns
scrubbing
3
2
1
0
0
5
10
15
Time [h]
20
Bunch length [ns]
0
0
10
Bunch intensity [ppb]
Bunch intensity
5
dmax in the arcs:
estimation technique
20
60
70
20
80
30
40
50
Bunch position [us]
2
]
0
x 10
10
20
Bunch length [ns]
0
-1
30
9
1
]
-1
x 10
0
10
20
30
70
80
40
50
Bunch position [us]
9
70
80
R0=0.7,
scan
in d80max
60
70
Emax=330 eV
2
1
40
50
Time [us]
0
0
-
0
60
Simulator PyECLOUD
20
30
2
1
60
fastBCT + bunch-bybunch b-length (B2)
2
40
50
Bunch position [us]
0
0
10
e per unit length [m
-
10
5
40
50
Bunch position [us]
0
0
10
30
fastBCT
+ bunch-by1
bunch b-length (B1)
e per unit length [m
x 10
60
70
10
20
80
30
40
50
Time [us]
60
70
80
Total simulated heat load
25
20
Heat load [W/hcell]
Measured
heat load
15
10
5
0
1.8
1.9
2
2.1
d
2.2
max
2.3
2.4
2.5
8
dmax in the arcs: results (50ns)
Beam 1
13
10
x 10
09/04
Energy
Beam 2
13/04
Intensity
8
6
4
2
0
0
5
10
Before 50ns
scrubbing
15
20
After 50ns
scrubbing
Time [h]
6
Av. heat load [W/hcell]
dmax5
2.28
4
3
2
1
0
0
2.2
50ns threshold@450 GeV
2.18
2.1
50ns [email protected] TeV
5
10
15
Time [h]
20
dmax in uncoated straight sections:
estimation technique
• The evaluation of dmax is done in the field-free regions in proximity of the
pressure gauges
– Used Beam1 data from gauges (Cu): VGI.141.6L4.B and VGPB.2.5L3.B
– A solution (R0 , dmax) is found comparing the pressure rises DPi measured at
different injections with the electron fluxes Fi from simulations
Measured
pressures

Baked but uncoated: SEY ~1.6-1.9.

Length 0.3 m

Pumping speed from NEG and maximum for CH4 ≈ 10 L/s
NEG
Simulated electron fluxes
10
dmax in uncoated straight sections:
results (50ns)
• Pressure rise measurements with 50ns beam to estimate dmax in the fieldfree regions in proximity of the pressure gauges (R0 ≈0.2–0.3)
– Measurements done at the beginning and at the end of the scrubbing run
– Measurements done during the 50ns operation of LHC (19 May)
– As expected, we are approaching the dmax thresholds for 50ns beams
50ns threshold@450GeV
29 June 2011, date of
the first injections of
25ns beams in LHC
50ns [email protected]
11
LHC operation with 50ns beams
•
By end-June 2011, LHC was filled with 1380 bunches per beam
– Nominal 50ns beams not suffering from obvious electron cloud limitations, very low rate
emittance growth
– No typical pattern along the batches as from electron cloud
•
Reduction of transverse emittances and increase of bunch current (from July
onwards) did not cause any significant return of the electron cloud
– Consistent with expected electron cloud behaviour (weak dependence on transverse
emittances, decrease with bunch current in dipoles)
•
Pressure rise from electron cloud only survived in a wide common StSt beam
pipe (close to ALICE)
1380 bunches per ring
e growth < 2% /h
12
Summary 50ns run
(before 25ns beam in LHC)
dmax
dmax
(last estimated)
dmax
(threshold @450 GeV)
(threshold @3.5 TeV)
Straight section
(uncoated)
1.66
1.63
1.58
Beam screen (arcs)
2.18
2.2
2.1
Nominal beam
1.1 x 1011 ppb
*
Thresholds in the arcs do not change significantly
at least up to Nb=1.8 x 1011 ppb
13
25ns experience in 2011
Beam 1
13
20
x 10
29/06
07/10
14/10
Energy
Beam 2
24-25/10
Intensity
15
10
5
0
0
5
DATE
Heat load [W/hcell]
S12
40
29 June S23
S34
30AugustS45
26
S56
20
7 OctoberS67
S78
10
S81
14 October
0
0 October5
24-25
10
15
20
25
30
Time [h]
35
40
45
50
55
SHORT DESCRIPTION
Injections of 9 x 24b trains per beam with different spacings between them
Two attempts to inject a 48b train with damper on and off: fast instability dumps
the beam within 500 turns in both cases (SBI and CBI)
High chromaticity (Q’x,y ≈15): Injection tests with trains of 72-144-216-288 bunches
from the SPS + ramp to 3.5 TeV & 5h store with 60b (12+24+24) per beam
Scrubbing
High chromaticity: injection of up to 1020 bunches per beam in 72b trains
(decreasing spacings between trains at each fill: 6.3–3.2–1 ms)
10 of up to15
20 in Beam
25 1 and 1020
30 in Beam352 (1ms train
40 spacing)45
Injection
2100 bunches
Time [h]
50
55
dmax in the uncoated sections:
results (25ns)
13
20
x 10
29/06
07/10
14/10
24-25/10
Intensity
15
10
5
Heat load [W/hcell]
0
0
40
30
20
10
0
5
•
•
10
15
20
25
30
Time [h]
35
40
45
50
S12
Attempt made on 14 October to take pressure rise measurements on a dedicated
S23
fill with decreasing spacings (4—3—2—1 ms), but hard to use data for the dmax
S34
estimation due to rapid evolution of beam and vacuum
S45
After considerable 25ns scrubbing, on the morning of the 25 October, 8 x 72b
S56
batches with different spacings could be injected for Beam 1 and remain stable to
S67
S78
allow the pressure values to level
S81
15
55
dmax in the uncoated sections:
results (25ns)
Start of 25ns
beams in LHC
50ns threshold
@450GeV
25ns threshold
@450GeV
50ns threshold @3.5TeV
•
•
25ns threshold @3.5TeV
Scrubbing with 25ns beam (~40h) has lowered dmax to 1.35 !
Again, we are not far from the threshold for 25ns beams, but further scrubbing is
needed
16
dmax in the arcs: results (25ns)
13
20
x 10
29/06
07/10
14/10
24-25/10
Intensity
15
10
5
Heat load [W/hcell]
0
0
40
30
20
10
5
10
15
20
Six snapshots from the 25ns
MDs to reproduce
load averaged
the Heat
measured
heat load by
simulations!
sector by sector
S12
S23
S34
S45
S56
S67
S78
S81
25
30
Time [h]
35
40
45
50
55
25
30
Time [h]
35
40
45
50 17
55
[x 10 mW/m/beam]
0
0
5
10
15
20
dmax in the arcs: results (25ns)
13
20
x 10
29/06
07/10
14/10
24-25/10
Intensity
15
10
5
Heat load [W/hcell]
0
0
40
30
20
10
5
10
15
20
25
30
Time [h]
35
40
45
50
55
40
45
50 18
55
Three snapshots from the
25ns MDs to try
disentangling aperture of
Beam1 from Beam2
S12
S23
S34
S45
S56
S67
S78
S81
0
0
5
10
15
20
25
30
Time [h]
35
dmax in the arcs: results
13
20
x 10
29/06
07/10
14/10
24-25/10
Intensity
15
10
5
0
0
10
20
30
40
50
Time [h]
2011 scrubbing history of LHC arcs
dmax has decreased from the
initial 2.1 to 1.52 in the arcs !
2.2
d max
2
1.8
25ns threshold
@450 GeV
1.6
1.4
0
10
20
30
Time [h]
40
25ns threshold @3.5 TeV
50
19
Not only heat load and pressure rise, the beam sees
the electron cloud, too, and it consequently…
⇒ Loses energy
⇒ Gets unstable and is quickly lost or exhibits emittance
growth
⇒Has a bad lifetime with a pattern degrading
towards the tail(s) of the batches
20
Not only heat load and pressure rise, the beam sees
the electron cloud, too, and it consequently…
⇒ Loses energy
⇒ Gets unstable and is quickly lost or exhibits emittance
growth
⇒Has a bad lifetime with a pattern degrading
towards the tail(s) of the batches
21
Beam observables: energy loss
Simulated
Measured
2
Bunch energy loss [mJ/Turn]
Beam 1
1.5
1
Measurements   the energy
loss per bunch is obtained from
the stable phase shift
0.5
0
0
500
1000
1500
2000
25ns bucket number
2500
3000
3500
1.4
Simulated
Measured
Simulations −  We use the test
case the last fill on the 25 October
Beam 2
1.2
Bunch energy loss [mJ/Turn]
1
0.8
0.6
0.4
0.2
0
22
0
500
1000
25ns bucket number
1500
Not only heat load and pressure rise, the beam sees
the electron cloud, too, and it consequently…
⇒ Loses energy
⇒ Gets unstable and is quickly lost or exhibits emittance
growth
⇒Has a bad lifetime with a pattern degrading
towards the tail(s) of the batches
23
Beam observables: emittance growth
• The benefits from
scrubbing have been visible
on the 25ns beam:
14 October  batches injected with 3.6 ms spacing, Q’x,y=15
– The effect of the electron
cloud has gradually
moved later later along
the trains, in spite of the
closer spacing!
– First 1 – 2 trains seem to
be hardly affected now
– In general, improvement
in vertical
• Both beams are still
unstable in the two planes,
or anyway affected by
emittance growth
24-25 October  batches injected with 1 ms spacing, Q’x=3, Q’y=15
Beam observables: emittance growth
x
8 10
max
=1.50
10
10
e per unit length [m ]
Bunch- intensity [ppb]
-1
d
10
10
6
5
4
10
0
0
0
10
10
20
20
30
30
40
50
Bunch
40 position [us]
50
Time [m s]
60
60
70
70
80
Not only heat load and pressure rise, the beam sees
the electron cloud, too, and it consequently…
⇒ Loses energy
⇒ Gets unstable and is quickly lost or exhibits emittance
growth
⇒Has a bad lifetime with a pattern degrading
towards the tail(s) of the batches
26
Beam observables: beam losses
24-25 October  first three batches injected of last three fills
Beam 1
•
•
•
Beam 2
At this point the behaviour of the two beams is very similar
This suggests similar electron cloud rise and saturation value
It is consistent with the dmax estimation made for beam 2 with the heat
load data
Summary 25ns MDs
dmax
dmax
(last estimated)
dmax
(threshold @450 GeV)
(threshold @3.5 TeV)
Straight section
(uncoated)
1.35
1.25
1.22
Beam screen (arcs)
1.52
1.45
1.37
⇒ We still need to scrub by an additional ~0.15 to ensure ecloud-less
operation of 25ns beams
⇒ Based on our models and lab measurements, we will try giving a rough
estimation of how long we need to scrub the arcs
28
Estimation of the scrubbing time
x 10
10
0
10
2
5
0
0
10
20
30
2 x 1010
]
10
-1
Av. scrubbing current [A/m ]
10
Bunch intensity [ppb]
Bunch length [ns]
Bunch intensity [ppb]
⇒ Structure of the scrubbing beam
 Conservative assumption: the beam at the end of the last 25ns fill
in 2011
⇒ Map the electron current density to the beam screen wall Je [A/mm2]
as a function of dmax
1
5
40
50
Bunch position [us]
60
70
80
1.6 70
1.8 80
-4
10
-6
0
00
0
x 10
-2
10
10
10
9
20
20
30
30
10
1 40
1.250
40 position50
Bunch
[us]
Bunch position [us]
1.4 60
60
SEY
70
max
80
2
29
Estimation of the scrubbing time
0
2
Av. scrubbing current [A/m ]
10
-2
10
-4
10
-6
10
1
1.2
1.4
SEY
1.6
1.8
2
max
R. Cimino
Estimation of the scrubbing time:
results
It took ~2.8h of equivalent beam 1
to go from dmax 1.65 to 1.52
R. Cimino
SEY max
SEY max
1.65
1.6
1.65
1.55
1.6
1.5
1.55
1.45
1.5
0
1.45
0
~2h
5
5
SEY max
-1
-1
per unit
[m ]
Max n. of e per unit lengthMax
[m n.] of e Heat
loadlength
[W/hcell]
15
10
15
Time [h]
1.65
9
50 x 10
1.6
2.5
40
1.55
302
~9h
~8h
1.5
20
1.5
1.45
101
0
0
0.5
0
9
0 x 10
0
2.5
Probably realistic to assume
at least 20h beam time
10
Time [h]
5
5
Time [h]
10
15
10
15
10
15
10
15
Time [h]
5
Time [h]
2
1.5
1
0.5
0
0
5
Time [h]
Estimation of the scrubbing time:
considerations
⇒ Based on our best knowledge at the moment, scrubbing the LHC arcs with
the 25ns beam we had during the last 2011 fill could take about 20h
effective beam time
⇒ The corresponding machine time could be about 1w
⇒ This lowers the electron cloud by “only” one order of magnitude, therefore
does not guarantee running without incoherent effects already at the end
of the calculated scrubbing time (it can take longer)
⇒ The total effective scrubbing time is actually longer because we will need
to dump and refill more times, and finally ramp, to extend the reach of the
scrubbing efficiency and cover the needed range
⇒ Towards the end of the scrubbing process, only the last trains reach
saturation and scrub. This can be optimized
 Inject in trains of 288b from the SPS, or at least (288 + N*72)
 Cause more uncaptured beam (like in the SPS)
 Minimize the length of the abort gap to possibly use multi-turn effects
(e.g. with overinjection)
32
Concluding remarks
dmax
(estimated)
Threshold dmax
(50ns, 450 GeV)
Threshold dmax
(50ns, 3.5 TeV)
Threshold dmax
(25ns, 450 GeV)
Threshold dmax
(25ns, 3.5 TeV)
Straight
section
(uncoated)
1.35
1.63
1.58
1.25
1.22
Beam screen
(arcs)
1.52
2.2
2.1
1.45
1.37
⇒ After the 25ns MDs, the LHC beam chambers have been cleaned to dmax
values well below the build up threshold for nominal 50ns beams
⇒ If the present level of machine conditioning was preserved, ‘ecloud-less’
operation of LHC with 50ns beams up to high intensities should be
possible in 2012
⇒ Only 2-3 days of scrubbing with 25ns beams for 50ns operation could be
sufficient – just to clean parts of the LHC open to air and check the
conditioning of the arcs
⇒ Scrubbing of the arcs for 25ns operation could take up to 2 weeks
machine time (including also test ramps)
Thank you for your
attention
Very special thanks to G. Iadarola, H. Bartosik, O. Dominguez, J. EstebanMüller, and F. Roncarolo for their careful off-line analysis of large amounts of
MD data and the huge simulation effort that improved the general
understanding of electron cloud and scrubbing!
Many thanks to G. Arduini, V. Baglin, P. Baudrenghien, G. Bregliozzi, S. Claudet,
G. Lanza, G. Papotti, E. Shaposhnikova, L. Tavian for all the beautiful data they
kindly provided us with and the numerous discussions
Thanks to B. Goddard, V. Kain, K. Li, H. Maury-Cuna, E. Métral, S. Redaelli,
B. Salvant, F. Zimmermann, and all those who promoted and/or actively
participated in the MDs
SPARE SLIDES
Beam observables: beam losses
24 October  batches injected with 1 ms spacing
Beam 1
d
e- per unit length [m-1]
10
10
10
Losses degrading
batch by batch
8
6
=1.50
4
0
12
Accum. number of impact. e-
max
Even weaker losses due
to delayed
injection +
Weaker losses due
to
scrubbing from the
delayed injection
injection of 1st batch
(1.551.52)
x 10
10
20
30
40
Time [m s]
d
11
max
50
60
70
50
60
70
=1.50
10
8
6
4
2
0
0
10
20
30
40
Time [m s]
Instability and emittance growth:
predictions
10
1.1e11
13
10
13
450GeV
450GeV
3.5TeV
10
10
10
10
e - central density
e- central density
10
3.5TeV
12
11
10
10
10
9
12
11
10
50ns
10
25ns
8
1.9
2
2.1
2.2
d
•
•
•
max
2.3
2.4
2.5
10
9
1.35
1.4
1.45
1.5
d
1.55
max
Calculated coherent ECI threshold for central density in dipoles is around
re=1012 m-3 for nominal intensity and Q’=0 at 450 GeV (simulations were run
assuming the whole LHC made of dipoles)
It can be stabilized with chromaticities Q’x,y>15, but emittance growth due to
electron cloud + chromaticity remains!
Right plot shows that with 25ns beams stability could be achieved only for
dmax ≤ 1.5
1.6
Estimation of the scrubbing time
⇒ Curve of the decrease of dmax with the integrated electron dose deposited on the wall,
d=JeDt [C/mm2]
⇒ Depends on material and electron energy, several measurements done in the past (two
examples illustrated here)
⇒ If we use the 500eV curve (left plot) we end up with scrubbing times in the machine
much lower than those measured  perhaps an indication that the real dmax in the
machine are lower than we believe (R0=1.0 instead of 0.7?)
2
C. Yin-Vallgren, scrubbing of
Cu measured with e- at 500eV
SEY max
1.8
1.6
1.4
1.2
1
-8
10
-6
10
-4
-2
10
10
2
Dose [C/mm ]
0
10
Dose [C/mm2]
38