Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
US LHC Accelerator Research Program bnl - fnal- lbnl - slac
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
Electron Cloud - Status and Plans
Miguel A. Furman LBNL mafurman@lbl.gov LARP CM10 Danford’s Inn, 23-25 April 2008
LARP CM10, BNL, Apr. 2008
Electron Cloud - M. Furman
1
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
Summary
NB: some of these activities not LARP funded
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
Progress
• Benchmark code WARP vs. HEADTAIL (emittance evolution) • Benchmark code POSINST (2D) vs. WARP 2D and 3D (build-up) • Ecloud build-up simulation of SPS strip detector measurements
SPS
• Dipole build-up simulations • Ecloud feedback simulations • SLAC-CERN effort on test chambers at the SPS ecloud chicane
Ecloud detection via microwave transmission
• Experiments at PEP-II through IR12 straight (Fall 2007) through the PEP-II ecloud chicane, variable dipole field (March-April 2008)
Ecloud cyclotron resonances
• measurements at the PEP-II ecloud chicane
PS2 and MI upgrade (time permitting) Plans
LARP CM10, BNL, Apr. 2008 Electron Cloud - M. Furman 2
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
Benchmark: Warp vs. Headtail-1
no synchrotron motion No dipole field
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
LHC – =479.6 – Np=1.11011 – continuous focusing
• x,y=66.0,71.54 • nx,y=64.28,59.31
– – – –
longitudinal motion OFF ne=1012 – 1014 m–3 ecloud station/turn: Nstn=10-100 mimic dipole magnetic field by freezing the x-motion of electrons – same initial distribution of macroprotons with initial offset of 0.1y
e- motion frozen in x
LARP CM10, BNL, Apr. 2008
Electron Cloud - M. Furman
3
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
Benchmark: Warp vs. Headtail-2
with synchrotron motion, exaggerated parameters
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
LHC – =479.6 – Np=1.11011 – continuous focusing
• x,y=66.0, 71.54 • nx,y=64.28, 59.31
– – – – –
longitudinal motion ON ne=1012 – 1014 m–3 ecloud station/turn: Nstn=10-100 field-free region same initial distribution of macro-protons with initial offset of 0.1y
LARP CM10, BNL, Apr. 2008
Electron Cloud - M. Furman
4
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
Benchmark: Warp vs. Headtail-3
with synchrotron motion, reasonable parameters
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
LHC • =479.6 • Np=1.11011 • continuous focusing
x,y = 66.0, 71.54 nx,y,z = 64.28, 59.31, 0.0059 = 3.4710-4 p/p = 4.6810-4 chromx,y = 2, 2
Warp
ne=1012m-3 ne=1014m-3
ne=1013m-3
ne=1011m-3
• ne=1011–1014 m–3 • Nstn ecloud station/turn=10-100 • dipole magnetic field effect: frozen xmotion of electrons • same initial distribution of macroprotons with initial offset of 0.1y • threshold 2-particle model for TMCI:
HEADTAIL
ne=1012m-3 ne=1014m-3
ne=1013m-3
ne=1011m-3
≈ 6.41011 m–3
LARP CM10, BNL, Apr. 2008 Electron Cloud - M. Furman
5
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
Benchmark: Warp build-up vs. POSINST Posinst
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
ILC build-up simulation – E=5 GeV – Np=21010
Warp
2-D
LARP CM10, BNL, Apr. 2008
Electron Cloud - M. Furman
3-D
6
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
Benchmarks conclusions
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
Good agreement between POSINST and WARP in build-up mode • 2D agrees with 3D when physical model is 2D (eg., dipole field) Good agreement between WARP and HEADTAIL For LHC, emittance growth negligible up to ~1000 turns when ne < ~1e12 m–3 (below TMCI threshold) Next steps: • better lattice description (eg. FODO arc cells instead of constant focusing model) • more self-consistency in beam-ecloud dynamics • further understand qualitative features of results • continue benchmarking incoherent emittance growth against CERN calculations
LARP CM10, BNL, Apr. 2008
Electron Cloud - M. Furman
7
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
SPS strip detector measurements
(data from M. Jiménez et al., Proc. ECLOUD’04)
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
Field free - Strip pick-ups
Dipole field field Dipole
RFD Distribution (a.u.)
Strip pick-up Distribution (a.u.)
Field free - Retarding Field Detector (fit)
Field free
-30 V -50 V
900 750
Ne- (a.u.)
0 100 200 300 400 500 Electron Energy (eV) 600 700 800
-70 V -115 V -180 V -300 V
-19 -16 -14 -11 -9 -6 -4 -1 1 4 6 9 11 14 16 19
600 450 300 150 0
80 eV
Dipole field
Strip pick-up (multicycle measurement)
dN/dE (a.u.)
Filtering potential
d N/dxdE Strip detector (single-cycle d2N/dxdE Strip-detector measurement-fit) Strip pick-up (single-cycle measurement)
2
- 500 V (exp. Decrease)
Lateral position (mm)
Heat load efficiency = HLE e-HLE(DF)= 1.7 × e-HLE(FF)
0 100 200 300 400 500 600 700 800 900
180 eV
LARP CM10, BNL, Apr. 2008
Energy (eV)
Ee->180 eV located in the centre faster beam conditioning observed
8
Electron Cloud - M. Furman
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
SPS chicane strip detector simulation
10
avWCrun (B=0) avWCrun (B=100 G)
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
5
SPS strip detector electron spatial distribution (a,b)=(7.6,2.25) cm Eb=26 GeV, Nb=1e11, tb=25 ns, sigz=0.21 m
1
4
rhistx_d1p3_B0 rhistx_d1p3_B100 rhistx_d1p4_B0 rhistx_d1p4_B100 B=100 G, dmax=1.4
A/m**2
0.1
3
x108
0.01
2
B=100 G, dmax=1.3
0.001
SPS, Eb=26 GeV, Nb=1e11 (sigx,sigy)=(3,1.7) mm, sigz=0.21 m fill pattern: 2 batches of 72 bunches each, tb=25 ns chamber: (a,b)=(7.6,2.25) cm, rectangular st. st., deltamax=1.3, Emax=292 eV
1
2.2
B=0, dmax=1.4
0.0001 1.0
1.2
1.4
1.6 deltamax
1.8
2.0
0 -0.06 -0.04 -0.02
B=0, dmax=1.3
10
-1
electron energy spectrum SPS strip detector (a,b)=(7.6,2.25) cm Eb=26 GeV, Nb=1e11, tb=25 ns, sigz=0.21 m wcek0hdet_d13_B0 wcek0hdet_d13_B100 wcek0hdet_d1p4_B0 wcek0hdet_d1p4_B100
0.00 x [m]
0.02
0.04
0.06
10
-2
B=100 G, dmax=1.4
(A/m**2)/eV
10
-3
B=100 G, dmax=1.3
10
-4
B=0, dmax=1.4
10
-5
B=0, dmax=1.3
Conclusions: • Qualitative agreement in e– intensity (dipole > FF) • Ditto in spatial distib. • Not in energy spectrum More detailed work desirable (needed?)
1000
10
-6
0
200
400
600 E0 [eV]
800
LARP CM10, BNL, Apr. 2008
Electron Cloud - M. Furman
9
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
Ecloud in SPS
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
E-Cloud in SPS has detrimental effects on LHC injection SPS emittance blowup and intensity limits translate directly to LHC intensity limits • Nominal LHC beams in SPS at edge of e-cloud stability (with chromaticity at maximum) Future injector upgrade scenarios raise intensity well beyond stability threshold Possible remedies for SPS E-Cloud instabilities • Vacuum chamber coating to reduce SEY
potentially expensive and requires significant shutdown
• Beam scrubbing
Is it enough?
• High chromaticity operation
Significant beam losses (reduction in dynamic aperture) effectiveness uncertain at higher intensities
• Active damping
damps coherent component of instability damping vertical single bunch instability is challenging
LARP CM10, BNL, Apr. 2008
Electron Cloud - M. Furman
10
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
Feedback damper of ecloud instability for SPS
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
SLAC-LBNL NI proposal made at CM9 (J. Fox and J.Byrd) We are beginning to form a collaboration SLAC: J. Fox, M. Pivi, L. Wang LBNL: J. Byrd, M. Furman, J.-L. Vay BNL: R. de Maria CERN: F. Zimmermann, W. Höfle, E. Chapochnikova
LARP CM10, BNL, Apr. 2008
Electron Cloud - M. Furman
11
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
SPS simulations-1
arc dipole
12
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
2.5x10
1.0
Aver. ecloud density vs. peak SEY
Aver. electron-wall flux vs. peak SEY
2.0
SPS arc dipole simulation LHC beam, Nb=1.1e11, 72 bunches/train aver. based on 2 batches (4 microsec)
0.8
SPS arc dipole simulation LHC beam, Nb=1.1e11, 72 bunches/train aver. based on 2 batches (4 microsec)
1.5
0.6
1.0
A/m**2
m**-3
avdenrun#0 (Eb=26 GeV) avdenrun#1 (Eb=450 GeV)
0.4
avWCrun#0 (Eb=26 GeV) avWCrun#1 (Eb=450 GeV)
0.5
0.2
0.0 1.0
1.2
1.4 peak SEY (dtotpk)
1.6
1.8
0.0 1.0
1.2
1.4 peak SEY (dtotpk)
1.6
1.8
Assume peak SEY=~1.3-1.4 (based on MI experience) Then ne~(5-10)e11 m–3 at Nb=1.1e11
LARP CM10, BNL, Apr. 2008
Electron Cloud - M. Furman
12
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
Preliminary simul. study of SPS EC feedback-1
beam distribution after 300 turns (J.-L. Vay)
No feedback
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
Feedback cutoff 0.8GHz (1/turn)
Y (cm)
centroid Y-centroid (cm) Y-centroid (cm) Time (ns) Power (a.u.) Power (a.u.)
Y (cm)
Time (ns)
LARP CM10, BNL, Apr. 2008
Frequency (Gz)
Electron Cloud - M. Furman
Frequency (Gz)
13
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
Preliminary simul. study of SPS EC feedback-2
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
SPS at injection (Eb=26 GeV) – =27.729 – Np=1.11011 – continuous focusing
• x,y= 33.85, 71.87 • nx,y= 26.12, 26.185 • nz= 0.0059
– Nstn ecloud station/turn=100 – Initial EC dist. From Posinst
LARP CM10, BNL, Apr. 2008
Electron Cloud - M. Furman
14
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
Preliminary simul. study of SPS EC feedback-3
present conclusions
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
Idea seems, in principle, to work well
• Damping the coherent vertical motion has beneficial impact on emittance growth
What next:
• Better modeling of the feedback system (bandwidth, gain, noise, separate pickup from kicker,…) • Longer runs • Look at horizontal motion • …
LARP CM10, BNL, Apr. 2008
Electron Cloud - M. Furman
15
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
Electron Cloud Studies for the SPS and LHC
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
1. GROOVE TESTS IN THE SPS: A number of electron cloud mitigation test chambers are in preparation for installation in a new dedicated 4-magnet chicane in the SPS. We are manufacturing groove insertions to fit in one of the test chamber. Collaboration: M. Venturini, M. Furman (LBNL), M. Pivi, L. Wang (SLAC) G. Arduini, E. Chapochnikova, M. Taborelli (CERN). CERN Contact: G. Arduini, E. Chapochnikova. 2. SINGLE-BUNCH INSTABILITY SIMULATIONS: code benchmarking and long term runs for the SPS and LHC. Simulation support for FDBK system (item 3 below) Collaboration: J.-L. Vay (LBNL), M. Pivi,L. Wang (SLAC),F. Zimmermann, W. Höfle (CERN), R. De Maria (BNL) CERN Contact: F. Zimmermann
3. RF MICROWAVE TRANSMISSION: Measurement of the electron cloud density in sections of the SPS by measuring the phase shift of microwave transmitted through the beam line, as recently done at SLAC and CERN.
Collaboration: S. De Santis, J. Byrd (LBNL), M. Pivi (SLAC), F. Caspers, T. Kroyer (CERN) CERN Contact: F. Caspers 4. FEEDBACK SYSTEM IN THE SPS: to mitigate electron cloud. See J. Fox & J. Byrd proposal.
Balance of FY08 FTEs (simulation and experimental efforts) M&S Travel
LARP CM10, BNL, Apr. 2008
FY09 0.6 40 k$ 12.5 k$
0.4 29 k$ 7.5 k$
Electron Cloud - M. Furman
16
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
Electron Cloud Studies for the SPS and LHC
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
CERN has a chicane at the SPS for ecloud tests Goal is to have 4 types of chamber tests: • St. St. chamber (reference) • Clearing electrodes • Carbon coating • NEG (TiZrV) coating There is a joint CERN-SLAC effort to manufacture 1-mm grooved liner to insert in the chicane chamber SLAC already manufactured a 2-mm Al grooved chamber (backup option for tests) No SLAC $$$ allocated for this project • But EMEGA Corp. has offered to do it if we pay for the tools Goal is 1 prototype St. St. grooved insertion 20” long • Ready in July 2008, installed summer 2008
LARP CM10, BNL, Apr. 2008 Electron Cloud - M. Furman
17
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
SPS grooves requirements
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
Triangular Grooves Groove Width 0.35 mm Groove Depth 1 mm Overall Depth 2 mm Groove Length 0.5 m Taper Angle 20 deg Radius at Top & Bottom 0 (< 80 mm)
LARP CM10, BNL, Apr. 2008
Electron Cloud - M. Furman
18
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
Manufacturing option for 1 mm depth: metal folding
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
Metal Folding: Form multiple folds. [EMEGA Company, USA]
LARP CM10, BNL, Apr. 2008
Electron Cloud - M. Furman
19
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
Aluminum triangular groove, SLAC. Depth 1.9mm, Opening angle 20o, radius top 95um, radius valley 144um
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
Tip
Valley
=1.50,Height=1.9mm, =20
0
0
2
• 1mm depth stainlees steel groove
insertion under development: CERN/SLAC • Back-up for SPS: 2mm Aluminum+coating triangular grooves (pictures above), manufactured by SLAC. Extrusion manufacturing limited by the groove sharpness requirements.
LARP CM10, BNL, Apr. 2008
1.8 1.6 1.4 Flat surface r=0.14mm,B=2 Tesla r=0.14mm,B=0.2Tesla r=0.09mm,B=2 Tesla r=0.09mm,B=0.2Tesla average,B=2 Tesla
SEY
1.2 1 0.8 0.6 0.4
Lanfa Wang, SLAC
0.2 0 100 200 300 400 500 600 700
Electron Cloud - M. Furman
Energy (eV)
20
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
Manufacturing options for 1 mm depth: razor blades
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
LARP CM10, BNL, Apr. 2008
Electron Cloud - M. Furman
21
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
Manufacturing options for 1 mm depth: razor blades
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
Brazed-up Assembly: Use individual razor type foil blades
LARP CM10, BNL, Apr. 2008 Electron Cloud - M. Furman 22
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
Mfg. Options
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
Extrusion: Very small radii at top & bottom of grooves are difficult to mfg Machining: Mill multiple slots in solid material Metal Folding: Form multiple folds EDM: Small radii are beyond normal tolerances Brazed-up Assembly: use individual razor type foil blades Isostatic Pressing or Metal Injection Molding: uses powdered metal & binders which would probably would not be suitable for vacuum usage. Also have difficulty in forming small radii
LARP CM10, BNL, Apr. 2008
Electron Cloud - M. Furman
23
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
Microwave Transmission Through an Electron Cloud
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
Original idea from Caspers and Kroyer (CERN) • Initially tried at the SPS Experiment recently carried out twice at PEP-II: • SLAC-LBNL-CERN collaboration: • Through IR12 straight section (L~50 m) (fall 2007)
De Santis, Byrd, Caspers, Krasnykh, Kroyer, Pivi and Sonnad, PRL 100, 094801 (7 March 2008)
• Through the ecloud chicane, with adjustable dipole field (March 2008)
Paper in progress
Fundamental idea: ecloud causes a phase shift of the transmitted microwave • Phase shift D is prop. to aver. ne in the region
LARP CM10, BNL, Apr. 2008
Electron Cloud - M. Furman
24
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
Analysis and simulations
2 c 2k 2 w 2 w c w 2 p
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
Dispersion relation:
w c=pipe cutoff angular freq. w p =plasma freq. of ecloud:
e 2ne w2 p 0 m e
Phase shift per unit length (relative to ne=0): D L 2c Choose w as close as possible to wc
Simulation with VORPAL (TechX)
w2 p
2 w 2 wc
QuickTime™ and a TIFF (Un compressed) decompressor are neede d to see this picture.
fc=2 GHz
QuickTime™ and a TIFF (Uncompressed) decompre ssor are neede d to see this picture.
LARP CM10, BNL, Apr. 2008
Electron Cloud - M. Furman
25
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
Experiment
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
PEP-II IR12 straight section L=~ 50 m Several quads plus an ecloudcontrolling solenoid Solenoid was switched on and off Beam gap (~30 m) causes ecloud to clear with frequency=frev
QuickTime™ and a TIFF (Uncompressed) decompressor are neede d to see this picture.
frev sidebands when Solenoid is off
QuickTime™ an d a TIFF (Uncompressed) decompressor are need ed to see this p icture .
LARP CM10, BNL, Apr. 2008
Electron Cloud - M. Furman
26
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
Experiment at w=2.149925 GHz
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
QuickTime™ and a TIFF (Uncompressed) decompressor rev are need ed to see this picture.
f =136 kHz
frev=136 kHz
(from De Santis et. al., PRL 100, 094801 (7 March 2008))
LARP CM10, BNL, Apr. 2008 Electron Cloud - M. Furman 27
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
Conclusions
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
From D and analytic
QuickTime™ an d a TIFF (Uncompressed) decompressor are need ed to see this picture .
LARP CM10, BNL, Apr. 2008
Electron Cloud - M. Furman
28
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
Propagation in a dipole field: PEP-II chicane
March-April 2008
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
Electrons in a dipole field:
wB
If w B w there is a magnetron resonance with a large D R=4.45 cm TE mode: w=12.266 Grad/s 11 Bres= mew/e=697.4 G
eB me
VORPAL simul. (TechX)
QuickTime™ and a TIFF (Uncompressed) decompre ssor are neede d to see this picture.
experiment
QuickTime™ an d a TIFF (Uncompressed) decompressor are need ed to see this p icture .
Bres
multiply by 2296 to get Gauss
LARP CM10, BNL, Apr. 2008 Electron Cloud - M. Furman 29
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
Microwave transmission: conclusions
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
Inexpensive, relatively easy way of measuring average ecloud density Advantages: • Direct average volumetric density measurement • In a local region of the machine (~a few to ~10’s of meters) • Parasitic • In real time • Relatively simple
What’s next: • Will repeat at SPS
LARP CM10, BNL, Apr. 2008
Electron Cloud - M. Furman
30
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
Ecloud cyclotron resonances: new effect
e– flux on chamber surface simul.
PEP-II ecloud chicane Eb=3.1 GeV, N b=6e10 tb=4.2 ns, z=1.15 cm
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
14 12 10 8 6 4 2 0 0
14 12 10 8 6 4 2 0 0
meas.
QuickTime™ an d a TIFF (Uncompressed) decompressor are need ed to see this p icture.
A/m**2
200
400
600
800
1000
B field [Gauss]
PEP-II ecloud chicane Eb=3.1 GeV, N b=6e10 tb=4.2 ns, z=1.15 cm
Resonances 1st seen in ILCDR build-up simulations (C. Celata) Predicted and quickly seen at PEP-II Collaboration: • SLAC: M. Pivi, J. Ng, L. Wang, C. Spencer • LBNL: C. Celata, M. Furman, J.-L. Vay, M. Venturini, K. Sonnad
12
A/m**2
1
2 3 4 5 6 7 8 9 10 11 n = (cycl. freq.)/(bunch freq.)=(eB/2 e)*tb m
LARP CM10, BNL, Apr. 2008
Electron Cloud - M. Furman
31
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
Ecloud cyclotron resonances
simulated e– density
2.0x10 1 3
PEP-II ecloud chicane Eb=3.1 GeV, N b=6e10
tb=4.2 ns, z=1.15 cm
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
2.0x10 13
2.0x10 1 3
PEP-II ecloud chicane Eb=3.1 GeV, N b=6e10
tb=4.2 ns, z=1.15 cm
PEP-II ecloud chicane Eb=3.1 GeV, N b=6e10
tb=4.2 ns, z=1.15 cm
1.5
1.5
1.5
m**-3
1.0
m**-3
1.0
m**-3
1.0
0.5
0.5
0.5
0.0
0
100
200
300
400
500
600
700
800
900
1000
B field [Gauss]
0.0 1900
2000
2100
2200 0.0
9200
9600
10000
10400
10800
B field [Gauss]
B field [Gauss]
We understand the basic physical Resonances need: mechanism • Low B fields (fctb<~20-40), and But agreement with expt. is imperfect: • Short bunches (wc t<~a few) • Spacing is perfect • But there is a shift in the location of peaks QuickTime™ and a TIFF (Uncompressed) decompressor are needed to see this picture. • 2D vs. 3D effects? • Instrumental issues?
LARP CM10, BNL, Apr. 2008 Electron Cloud - M. Furman 32
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
PEP-II ecloud chicane in e+ beam (M. Pivi)
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
QuickTime™ an d a TIFF (Uncompressed) decompressor are need ed to see this p icture .
4 dipoles,tunable B-field (1.5 kG max), R=4.45 cm 2 chambers (1 covers 3 dipoles, the 2nd covers 1) Multiple detectors Surfaces: 1) bare Al and 2) TiN-coated • A grooved surface and a NEG-coated surface were eliminated by budget Parasitic operation (typ. Nb=6e10, Eb=3.1 GeV, z=1.15 cm, tb=4.2 ns) PEP-II stopped for good on April 7th, 2008 • Chicane will be moved to CESR-TA
LARP CM10, BNL, Apr. 2008 Electron Cloud - M. Furman 33
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
Ecloud resonances:
possible implications for LHC and injectors
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
Resonances introduce a correlation between B fields and the EC density – But, LHC bunch way too long to directly excite cyclotron resonances However: there is a possible related effect: “magnetron effect” (F. Caspers) Electron cyclotron motion may be excited by beam-induced wake fields – “largest microwave oven ever built” CERN is encouraging us to study the effect (F. Zimmermann and F. Caspers, ongoing email exchanges)
What’s next: – look at basics, eg.: energy stored in wake fields, time constants,… – no plans for simulations at this point
LARP CM10, BNL, Apr. 2008
Electron Cloud - M. Furman
34
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
C=1256 m, tb=25 ns, Nb=4e11, z=0.935 m
(F. Z. psplusetcparameters option 2)
4
PS2 simulated ecloud build-up
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
6x10
11
density
3
4
SEY max=1.20 SEY max=1.30 SEY max=1.40
heat load
m**-3
PS2, LHC beam, bending dipole Eb=50 GeV, t b=25 ns, z=0.935 m averages taken over 2 batches (1 batch=72 bunches+gap, total=2 Gröbner multip.
ms)
W/m
2
PS2, LHC beam, bending dipole Eb=50 GeV, t b=25 ns, z=0.935 m averages taken over 2 batches (1 batch=72 bunches+gap, total=2 Gröbner multip.
ms)
2
SEY max=1.20 SEY max=1.30 SEY max=1.40
1
0 0
1
2
3
4
5x10
11
0 0
1
2
3
4
5x10
11
bunch population Nb
bunch population Nb
Contrasts between PS2 and MI upgrade: • PS2 significantly above Gröbner multipacting condition, at least for the 25 ns option • MI upgrade, even at 3e11/bunch, significantly below • But ecloud density roughly comparable MI ecloud measurement efforts have been valuable What’s next: better characterization of ecloud distribution and intensity in PS2
LARP CM10, BNL, Apr. 2008 Electron Cloud - M. Furman 35
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
FNAL Main Injector
C=3319.4 m, tb=19 ns, Nb=(0.6–1)e11, z=0.19 m
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
• For this exercise, take measured RFA signal only at Eb=60 GeV
• this is the peak signal for all cases
Field-free region, R=7.3 cm, St. St. To convert RFA voltage signal to e– flux (R. Zwaska):
• assume 1 mA/V • divide by 1.5 cm2 this assumes 30% area efficiency
Measured e– flux at RFA vs. Nb for various fill patterns (Eb=60 GeV all cases; extracted from I. Kourbanis report, ~26 Aug. 2007)
10x10
- 3
measured MI peak RFA signal vs. Nb – e flux at the wall
8
(peak is at Eb~60 GeV in all cases) data from "e-Cloud MI Measurements," I. Kourbanis, ~26 Aug. 2007
[A/m**2]
6 Je_B5 (5 trains) Je_B4 (4 trains) Je_B4u (4 trains, unequal train spacing) Je_B3 (3 trains)
4
2
0 0.0
0.2
0.4
0.6 Nb
0.8
1.0x10
1 1
LARP CM10, BNL, Apr. 2008
Electron Cloud - M. Furman
36
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
MI: e– flux at wall vs. peak SEY at Eb=60 GeV
2
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
e
av 0.1 W C r u n # 0 avWCrun#1 avWCrun#2 7 6 avWCrun#3 5 avWCrun#4 avWCrun#5 4 avWCrun#6 3 avWCrun#7
2
–
10 flux
even gaps, Nb=9.7e10) even gaps, Nb=9.0e10) even gaps, Nb=8.1e10) even gaps, Nb=7.2e10) even gaps, Nb=9.5e10) even gaps, Nb=9.1e10) uneven gaps, Nb=9.5e10) even gaps, Nb=9.1e10)
1 3
[A/m**2]
(5 (5 (5 (5 (4 (4 (4 (3
trains, trains, trains, trains, trains, trains, trains, trains,
simulated
simulated number density e 10
1 2
–
10
1 1
[m**-3]
0.01
7 6 5 4 3 2
measured
10
1 0
M I , b= 6 0 G e V E double slip-stacked batches train length=81 bunches z= 0 . 1 9 m
avdenrun#0 avdenrun#1 avdenrun#2 avdenrun#3 avdenrun#4 avdenrun#5 avdenrun#6 avdenrun#7 (5 (5 (5 (5 (4 (4 (4 (3 trains, trains, trains, trains, trains, trains, trains, trains,
10
9
0.001 1.20
M I , b= 6 0 G e V E 8 10 double slip-stacked batches train length=81 bunches z= 0 . 1 9 m
1.25
1.30 peak SEY
1.35
1.40
10 1.1
7
even gaps, even gaps, even gaps, even gaps, even gaps, even gaps, uneven gaps even gaps,
1.2
1.3
1.4
1.5
1.6
1.7
1.8
peak SEY
• Nicely clustered set of solutions for max
– Indicates consistency in the model and the measurements – Conclude: max~1.25–1.35 (M. Furman, CBP-TN-387, Nov. 07)
• Simulation then implies ne~1010-1011 m–3 • A mystery remains: simulations show insensitivity to Eb • Measurements are sensitive to Eb
LARP CM10, BNL, Apr. 2008
• Qualitatively similar to SPS! (G. Arduini,- M. FurmanECLOUD’04) Proc. Electron Cloud
37
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
MI upgrade goal: Nb x5 relative to today
simulation of ecloud density vs. Ntot for frf=53 and 212 MHz
12
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
4x10
MI, frf=53 MHz vs. 212 MHz, Eb=8.9 GeV 53 MHz (h=588), sigz=0.75 m 548 filled + 40 empty buckets
10 10 10 10
13
MI, frf=53 MHz vs. 212 MHz, Eb=8.9 GeV
3
212 MHz (h=2352, sigz=0.1875 m 2192 filled + 160 empty buckets peak SEY=1.2 avdenrun#0 (h=588) avdenrun#1 (h=2352) peak SEY=1.3 avdenrun#2 (h=588) avdenrun#3 (h=2352)
linear
12
11
log
peak SEY=1.2 avdenrun#0 (h=588) avdenrun#1 (h=2352) peak SEY=1.3 avdenrun#2 (h=588) avdenrun#3 (h=2352) peak SEY=1.4 avdenrun#4 (h=588) avdenrun#5 (h=2352) 53 MHz (h=588), sigz=0.75 m 548 filled + 40 empty buckets 212 MHz (h=2352, sigz=0.1875 m 2192 filled + 160 empty buckets
10
m**-3
2
peak SEY=1.4 avdenrun#4 (h=588) avdenrun#5 (h=2352) beamneutden Gröbner multip. (frf=53 MHz)
m**-3
10 10 10
9
8
1
7
present range
0 0.0
0.5
1.0 Ntot
1.5
2.0x10
14
10 0.0
6
0.5
1.0 Ntot
1.5
2.0x10
14
Present: Nb=6e10, fRF=53 MHz, M=548 (no. of bunches) Upgrade goal: Nb=3e11, or Ntot=1.64e14
Exercise: what happens if – fRF fRFx4, NbNb/4, M Mx4 (preserve Ntot)? Answer: 212 MHz clearly better than 53 MHz: – Threshold in Ntot roughly doubles
38
LARP CM10, BNL, Apr. 2008
Electron Cloud - M. Furman
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
Long-Term Simulation of Space-Charge-Driven Dynamic Emittance Exchange
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
• • • • •
• • •
MARYLIE/IMPACT code (3D parallel) Ramp longit. tune from below to above resonance Protons, sp. ch. tune shift ~0.1–0.3 Propagate beam through a linac – 1 linac period=35 deg phase adv. – Constant focusing lattice approx. Check scaling law by I. Hofmann and G. Franchetti, PAC07
106 macroparticles, 643 grid ~1.3x106 space-charge kicks, 32 hrs on 64 proc. (IBM/SP 5) LBNL would like to participate in a joint simulation effort on PSB and PS2 with other institutions: • Space-charge effects • Ecloud • Traditional impedance/instabilities • Longstanding collaboration with GSI • R. Ryne, J. Qiang, M. Furman
LARP CM10, BNL, Apr. 2008
Electron Cloud - M. Furman
39
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
Status summary and future goals
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
1. 2. 3.
4.
5. 6. 7. 8. 9.
Injector upgrade heat load: (*) continuing Effects from ecloud on beam: (*) Benchmarks: POSINST-WARP and WARP-HEADTAIL: mostly done, need to refine lattice model • 3D self-consistent simulations: challenging; continuing • Lorentz-boosted frame method: shows good agreement with QSM in benchmarks • Effects of ionized gas on heat load and beam: not started Analyze SPS data, esp. measured heat load and e– spectrum: (*) started; need better benchmarks against expts. Apply Iriso-Peggs maps to LHC: (–) delayed or deleted Simulate e-cloud for RHIC detectors and benchmark against measurements: (**) nothing to report; e– detectors broken Simulate ecloud for LHC IR4 “pilot diagnostic bench:” not started ecloud suppression at SPS by feedback: • Simulations: started ccloud suppression at SPS via specialized chambers: new proposed activity
Requested additional funding for items 8 and 9 is spelled out on slide 16
(*) endorsed by CERN AP group (**) endorsed by CERN vacuum group
LARP CM10, BNL, Apr. 2008
(–) no longer endorsed by CERN AP group
Electron Cloud - M. Furman 40
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
Additional material
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
LARP CM10, BNL, Apr. 2008
Electron Cloud - M. Furman
41
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
SPS simulations-2
arc dipole case
0.10
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
Aver. ecloud life path vs. Nb for peak SEY=1.3, 1.4
0.08
SPS arc dipole simulation LHC beam, 72 bunches/train + gap aver. based on 2 batches (~4 microsec)
0.06
m
0.04
Gröbner multip. (y)
avpathrun#0 avpathrun#1 avpathrun#2 avpathrun#3
(Eb=26 GeV, dtotpk=1.3) (Eb=26 GeV, dtotpk=1.4) (Eb=450 GeV, dtotpk=1.3) (Eb=450 GeV, dtotpk=1.4)
0.02
1.0x10
12
Aver. ecloud density vs. Nb for peak SEY=1.3, 1.4
0.00 0.0
0.2
0.4
0.6
0.8
1.0
1.2x10
11
bunch population (Nb)
0.8
SPS arc dipole simulation LHC beam, 72 bunches/train + gap aver. based on 2 batches (~4 microsec) avdenrun#0 avdenrun#1 avdenrun#2 avdenrun#3 (Eb=26 GeV, dtotpk=1.3) (Eb=26 GeV, dtotpk=1.4) (Eb=450 GeV, dtotpk=1.3) (Eb=450 GeV, dtotpk=1.4)
0.6
m**-3
0.4
Gröbner multip. (y)
0.2
0.0 0.0
0.2
0.4
0.6
0.8
1.0
1.2x10
11
bunch population (Nb)
LARP CM10, BNL, Apr. 2008
Electron Cloud - M. Furman
42
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
MI: e– flux at wall vs. peak SEY at Eb=60 GeV
avWCrun#0 avWCrun#1 avWCrun#2 avWCrun#3 avWCrun#4 avWCrun#5 avWCrun#6 avWCrun#7 (5 (5 (5 (5 (4 (4 (4 (3 trains, trains, trains, trains, trains, trains, trains, trains, even gaps, Nb=9.7e10) even gaps, Nb=9.0e10) even gaps, Nb=8.1e10) even gaps, Nb=7.2e10) even gaps, Nb=9.5e10) even gaps, Nb=9.1e10) uneven gaps, Nb=9.5e10) even gaps, Nb=9.1e10)
2
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
0.4
e
av 0.1 W C r u n # 0 avWCrun#1 avWCrun#2 7 6 avWCrun#3 5 avWCrun#4 avWCrun#5 4 avWCrun#6 3 avWCrun#7
2
–
flux
0.3
[A/m**2]
M I , b= 6 0 G e V E
[A/m**2]
0.2
double slip-stacked batches train length=81 bunches z= 0 . 1 9 m
(5 (5 (5 (5 (4 (4 (4 (3
trains, trains, trains, trains, trains, trains, trains, trains,
even gaps, Nb=9.7e10) even gaps, Nb=9.0e10) even gaps, Nb=8.1e10) even gaps, Nb=7.2e10) even gaps, Nb=9.5e10) even gaps, Nb=9.1e10) uneven gaps, Nb=9.5e10) even gaps, Nb=9.1e10)
simulated
0.01
7 6 5 4 3 2
measured
0.1
–
simulated flux e 0.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8
10
0
peak SEY
0.001 1.20
M I , b= 6 0 G e V E double slip-stacked batches train length=81 bunches z= 0 . 1 9 m
1.25
1.30 peak SEY
1.35
1.40
simulated 10
- 1
measured e
–
10
- 2
flux
• Nicely clustered set of solutions for max
• Indicates consistency in the model and the measurements • Conclude: max~1.25–1.35 (St. St.) (M. Furman, CBP-TN-387, Nov. 07)
[A/m**2]
10
- 3
M I , b= 6 0 G e V E double slip-stacked batches train length=81 bunches z= 0 . 1 9 m
avWCrun#0 avWCrun#1 avWCrun#2 avWCrun#3 avWCrun#4 avWCrun#5 avWCrun#6 avWCrun#7 (5 (5 (5 (5 (4 (4 (4 (3 trains, trains, trains, trains, trains, trains, trains, trains,
10
- 4
10
- 5
10
- 6
even gaps, Nb=9.7e10) even gaps, Nb=9.0e10) even gaps, Nb=8.1e10) even gaps, Nb=7.2e10) even gaps, Nb=9.5e10) even gaps, Nb=9.1e10) uneven gaps, Nb=9.5e10) even gaps, Nb=9.1e10)
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
LARP CM10, BNL, Apr. 2008
peak SEY
Electron Cloud - M. Furman
43
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
Preliminary simul. study of SPS EC feedback -1
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
SPS at injection • =27.729 • Np=1.11011 • continuous focusing
x,y= 33.85,71.87 nx,y= 26.12,26.185 nz= 0.0059
• Nb ecloud station/turn=10 • Initial EC dist. From Posinst
LARP CM10, BNL, Apr. 2008
Electron Cloud - M. Furman
44
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
Effects of ecloud: growth in LHC beam
code WARP (J.-L. Vay)
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
one-turn growth simulation E=450 GeV, Nb=1.1x1011, single bunch,
Fractional Y emittance growth after 1 turn
1-turn fractional emittance growth vs. Nstn for 3 values of the ecloud density
10 10 10 10 10 10 10 10 10 10 10
2
1
• Code WARP, parallel, 3D calc.
Quasi-static approx. mode (QSM) AMR, parallel 8 processors
0
continuous foc uss ing, ( x, y)=(66.0,71.54) m, ( nx, ny)=(64.28,59.31) Nb=1.1e11, =579.6, x= y=0.884 mm, z=13 cm Mp=3e5, Me=65536/slice, Lz=+-4 z, Nz=128
-1
• Beam transfer maps from EC station to next
Up to 3000 stations
-2
-3
• Actual LHC chamber shape • Constant focusing approx. • Electrons allowed to move vertically only • No synchr. oscillations • Beam launched offset by 0.1y
-4
-5
-6
-7
N e= 10
14
10
13
10
12
m
-3
-8 2 3 4 5 6 7 8 9 2 3 4 5 6 7 8 9 2 3
10
100
1000
Conclusion: need to resolve l to reach convergence, as expected (ie., # of EC stations > tune)
LARP CM10, BNL, Apr. 2008 Electron Cloud - M. Furman
# of ecloud stations per turn
45
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
Effects of ecloud: growth in LHC beam 1-turn growth vs. ne
fractional emitt. growth in 1 turn
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
Emittance growth simul. Same conditions as previous slide • except Nstn=3000=fixed Conclusion: • D/ ne as ne-->0
1-turn fractional emittance growth vs. ecloud density (Nstn=3000)
10 10 10 10 10 10 10 10
- 1
- 2
- 3
- 4
horiz. vert.
- 5
- 6
- 7
- 8
10
1 0
10
1 1
10 10 ne (m**-3)
1 2
1 3
10
1 4
10
1 5
LARP CM10, BNL, Apr. 2008
Electron Cloud - M. Furman
46
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
Ecloud build-up in PS2 at 50 GeV
vs. chamber radius
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
Looked only at a bending dipole Vary pipe radius keeping all else fixed
Nb=4x1011 for tb=25 ns, Nb=5.4x1011 for tb=50 ns; other parameters as specified in LUMI06 by FZ Averages taken over 2 trains PS+ also looked at Conclusions: • Low heat load wants small radius • Low e– density wants large radius • Beam-induced multipacting condition broadens and gets shifted to lower radius relative to the impulse approximation (Gröbner, )
QuickT ime™ and a T IF F (Uncom pressed) decompressor are needed to see t his pict ure.
10
W/m
ecloud heat load in a dipole PS2, Eb=50 GeV, copper tb=25 tb=25 tb=50 tb=50
heat load
ns, ns, ns, ns,
SEYmax=1. SEYmax=1. SEYmax=1. SEYmax=1.
5
Groebner's multipacting condition (tb=25 ns) Groebner's multipacti condition (tb=50 ns)
0 0.00
1.0x10
1 2
0.05
0.10
0.15
0.20
pipe radius [m]
aver. ecloud density in a dipole PS2, Eb=50 GeV, copper
m**-3
tb=25 tb=25 tb=50 tb=50 0.5
ns, ns, ns, ns,
SEYmax=1.3 SEYmax=1.5 SEYmax=1.3 SEYmax=1.5
density
0.0 0.00
0.05
0.10
0.15
0.20
pipe radius [m]
LARP CM10, BNL, Apr. 2008
Electron Cloud - M. Furman
47
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
Ecloud build-up in PS2 at 50 GeV (contd.)
vs. chamber radius
0.14
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
A/m**2
e– flux at the walls (Je) Conclusions: • Ratio Je/ne in good agreement with analytic expectation as r-->0:
QuickT ime™ and a T IF F (Uncompressed) decom pressor are needed to see this pict ure.
0.12 0.10 0.08 0.06 0.04
electron flux at wall (Je) PS2, Eb=50 GeV, copper
tb25, tb25, tb50, tb50,
maxSEY=1.3 maxSEY=1.5 maxSEY=1.3 maxSEY=1.5
(R. Zwaska)
0.02 0.00 0.00
0.05
0.10
0.15
0.20
4x10
- 1 3
pipe radius [m]
Je/ne in a dipole PS2, Eb=50 GeV, copper 3
A-m
2
1
0 0.00
tb=25, tb=25, tb=50, tb=50, anal., anal., 0.05 0.10 pipe radius [m]
SEYmax=1.3 SEYmax=1.5 SEYmax=1.3 SEYmax=1.5 tb=25 ns tb=50 ns 0.20
0.15
LARP CM10, BNL, Apr. 2008
Electron Cloud - M. Furman
48
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
Furthermore…
6
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
2.0x10 (e 1.5
[m/s]
–
10 flux)/(e harge c
–
1 3
density)
simulated number density e 10
1 2
–
s i m p l e t h e o r y ( e x a c t h0 ) a w en
10
1 1
[m**-3]
1.0
10
1 0
M I , b= 6 0 G e V E double slip-stacked batc train length=81 bunche z= 0 . 1 9 m
avdenrun#0 avdenrun#1 avdenrun#2 avdenrun#3 avdenrun#4 avdenrun#5 avdenrun#6 avdenrun#7 (5 (5 (5 (5 (4 (4 (4 (3 trains, trains, trains, trains, trains, trains, trains, trains,
rat#0 (5 trains, rat#1 (5 trains, 0.5 M I , b= 6 0 G e V r a t # 2 ( 5 t r a i n s , E rat#3 (5 trains, d o u b l e s l i p - s t a c k e d tb a t c h e s r a i n s , ra #4 (4 t t r a i n l e n g t h = 8 1 r a t # 5 e( 4 t r a i n s , bunch s rat#6 (4 trains, z= 0 . 1 9 m rat#7 (3 trains,
10 even gaps, Nb=9.7e10) even gaps, Nb=9.0e10) even gaps, Nb=8.1e10) even gaps, Nb=7.2e10) 8 even gaps, Nb=9.5e10) 10 even gaps, Nb=9.1e10) uneven gaps, Nb=9.5e10) even gaps, Nb=9.1e10) 1.7 1.8 10 1.1
7
9
even g even g even g even g even g even g uneven even g
0.0 1.1
1.2
1.3
1.4
1.5
1.6
1.2
1.3
1.4
1.5
1.6
1.7
1.8
peak SEY
peak SEY
Flux/density consistent with simple theory, as expected • Je/re≈a/(2tb) (R. Zwaska)
This becomes exact in the limit a0
From Je results (previous slide), conclude ne~1010-1011 m–3
LARP CM10, BNL, Apr. 2008 Electron Cloud - M. Furman
49
Qu i ckTi me ™ an d a TIFF (Un co mp re s se d) de co mp re s so r a re ne ed ed to se e thi s p i ctu re .
Quasi-static mode (“QSM”)
Qui ckTi me™ and a TIFF (Uncompresse d) d eco mpressor are ne eded to see thi s pi cture.
2-D slab of electrons
3-D beam s
lattice
quad drift
s0
bend
drift
1. 2-D slab of electrons (macroparticles) is stepped backward (with small time steps) through the frozen beam field • 2-D electron fields are stacked in a 3-D array, 2. push 3-D proton beam (with large time steps) using • maps - “WARP-QSM” - as in HEADTAIL (CERN) or • Leap-Frog - “WARP-QSL” - as in QUICKPIC (UCLA/USC).
LARP CM10, BNL, Apr. 2008 Electron Cloud - M. Furman 50