updates of electron cloud studies at kekb

TILC09, Tsukuba 1

Updates of Electron Cloud Studies at KEKB

Topics in 2008•Measurement of electron density

in solenoid and Q field•Mitigation using clearing electrode in B field•Mitigation using groove surface in B field

2009/4/19

Y. Suetsugu, KEKB Vacuum Group

TILC09, Tsukuba 2

EC studies at KEKB Various EC studies have been carried out utilizing

the KEKB positron ring.– EC deteriorated the luminosity of KEKB

Energy 3.5 GeVCircumference 3016.26 mNominal bunch current 1~1.3 mANominal bunch charge 10~13 nCNominal bunch spacing 6~8 nsHarmonic number 5120RMS beam size (x/y) 0.42/0.06 mmBetatron tune 45.51/43.57RF voltage 8 MVSynchrotron tune 0.024Radiation damping time 40 ms

2009/4/19

TILC09, Tsukuba 3

EC studies at KEKB Diagnostics

– Beam size blow-up: by SR monitors– Beam instabilities (Head-tail instability, coherent

instability) : by BOR, Tune measurement– Electron density at drift space, in a solenoid and Q field:

by electron monitors with RFA– Secondary Electron Yield (SEY) at lab. and in situ

Mitigation– Solenoid field at drift space Very effective– Beam pipe with antechambers– Coating to reduce SEY: TiN, NEG, Graphite, DLC– Clearing Electrode in a B field– Groove surface in a B field

2009/4/19

TILC09, Tsukuba 4

Electron density in a solenoid Measurements so far have been only at drift space

or in B field. New RFA-type electron detector was installed in a

solenoid. Only high energy electrons produced near the

bunch can enter the detector.

2009/4/19

Groove

SR

Detector 1

Detector 2

K. Kanazawa and H. Fukuma

B

-15 +150

y [m

m]

+15

-15

0

Starting points of electrons measured by the detector

Only these electrons enter the detector.

Beam

Detector

Bunch size: x = 0.434mm, sy = 0.061mm, sz = 6mm, B = 50 G, NB = 7.51010

x [mm]

NormalDetector

TILC09, Tsukuba 5

Electron density in a solenoid Electron density with a solenoid field (B = 50 G)

2009/4/19

109

1010

1011

1012

1013

0 0.2 0.4 0.6 0.8 1 1.2

Effect of a Solenoid Field

Nea

r B

eam

Ele

ctro

n C

loud

Den

sity

[m-3

]

LER Bunch Current [mA]

B = 0

B = 50 G, Detector 2


Preliminary

13 November 2008 [4, 200, 3]At 7.2m from a Bend

Photon flux=3.9x1017I[A] photons/mB=0

Detector 1

109

1010

1011

1012

1013

0 0.2 0.4 0.6 0.8 1 1.2

Effect of a Solenoid Field

Near B

eam

E

lectro

n C

lo

ud

D

en

sity

[m

-3 ]


B = 0



Preliminary

13 November 2008 [4, 200, 3]At 7.2m from a Bend

Photon flux=3.9x1017I[A] photons/m

Preliminary result

B=50 G

Detector 2


The electron density decreased less than 1/100 in the solenoid field.

The difference in two detectors may be due to COD and relative position to the primary synchrotron radiation

The measured current in a solenoid field might have included electrons drifting along the wall.

800 bunches(Bs = 6 ns)

Ie ~ photoelectrons?New detector with RFA

(2009)

TILC09, Tsukuba 6

Electron density in a Q magnet New RFA-type electron detector was also installed

in a Q magnet Only high energy electrons produced near the

bunch can enter the detector.

2009/4/19

X-axisDetector 2

Detector 1

SR


B

+20

0

-20

x [m

m]

-20 0 20

Detector

y [mm]

Detector Bias = -1 keV

x = 0.35 mm, y = 0.094 mm, z = 6.0 mm, B’ = 3.32 T/m, NB

= 7.51010

Beam

Starting points of electrons measured by the detector

Only these electrons enter the detector.

TILC09, Tsukuba 7

Electron density in a Q magnet Electron cloud density in Q magnet (B’ = 3.32 T/m)

2009/4/19

The observed value in the Q-Magnet was close to the estimation by simulation [CLOUDLND].(dmax at 250 eV)

The difference in two detectors may be due to COD and relative position to the primary synchrotron radiation


Preliminary result800 bunches(Bs = 6 ns)

0

1 1010

2 1010

3 1010

4 1010

5 1010

0 0.2 0.4 0.6 0.8 1 1.2

D(D2-Q1)[4,200,3]D(D2-Q2)[4,200,3]

Nea

r B

eam

Ele

ctro

n C

lou

d D

ensi

ty [

m-3]


13 November 2008 [4, 200, 3]Normal Incident Energy > 1 keVB' = 3.32 T/m13.5 m from a half bend

Photon flux = 3.17x1015 I[A] photons/m

d =1.0 d =1.2

Simulation by CLOUDLAND

TILC09, Tsukuba 8

Mitigation in magnets Mitigation techniques of EC in magnets have

recently attracted attention.– Solenoid field is very effective at a drift space.

A clearing electrode and a groove had been said to be effective even in magnets from simulations.

However, no experimental demonstration in high intensity positron rings was reported so far.– Clearing electrode:

Experiments were carried out at CERN (proton ring). Impedance and heating is key problems for positron ring.

– Groove: Experiments at a drift space were carried out at SLAC.

Experimental demonstrations of the clearing electrode and the groove were tried using a wiggler magnet in the KEKB positron ring.

2009/4/19

TILC09, Tsukuba 9

Clearing electrode Electrode in a beam pipe attracts or repels the

electrons around the beam orbit.

2009/4/19

R-pipe=38mmbunch intensity=9.36e103.5 Bunch SpacingB = 0.75 T

dmax = 1.4

Electron Density in magnetic field Two

peaks

Electrode (+)

Little electrons

L. Wang et al, EPAC2006, p.1489

TILC09, Tsukuba 10

Clearing electrode New strip-line type electrode was developed. Very thin electrode and insulator;

– Insulator: ~0.2 mm, Al2O3, by thermal spray.– Electrode: ~0.1 mm, Tungsten, by thermal spray.

Low beam impedance, high thermal conductivity– Input power ~ 100 W

2009/4/19

Stainless steel

TungstenAl2O3

40 m

m

440 mm

Feed through

Y. Suetsugu, H. Fukuma, M. Pivi and L. WangNIM-PR-A, 598 (2008) 372

TILC09, Tsukuba 11

Clearing electrode The electrode and an electron monitor were set

face to face in a test chamber.

2009/4/19

R47

Beam

5

[Electrode]

[Monitor]

Mag

neti

c field

Al-alloy chamber(Not coated)

Cooling water

Monitor

Electrode

Inside view

Electron monitor with RFAand 7 strips to measurespatial distribution

• Applied voltage Collectors: +100V Retarding Grid: 0 ~ -1 kV• Measurement: DC mode

TILC09, Tsukuba 12

Clearing electrode Electric potential in the chamber by this electrode

2009/4/19

+500 V

+120 V

0 V

TILC09, Tsukuba 13

Clearing electrode Test chamber was installed into a wiggler magnet

– Wiggler magnet. Magnetic field: 0.77 T Effective length: 346 mm Aperture (height): 110 mm

Placed at the center of pole SR: 2x1017 photons/s/m

at 1600 mA Electrode voltage (Velec):

Velec = -500 ~ +500 V Repeller voltage (Vr):

Vr = 0 ~ -1 kV

2009/4/19

Test chamber

Gate Valve Gate Valve

Feed through

TILC09, Tsukuba 14

Clearing electrode Spatial and energy distribution of EC (Velec = 0 V)

– Also for checking of electron monitor

2009/4/19

Electron distribution splits to two peaks at high current.

center

1585 bunches(Bs ~ 6 ns)

High energy electrons are at the beam position.

Center

1585 bunches(Bs ~ 6 ns)~1600 mA

TILC09, Tsukuba 15

Clearing electrode Spatial growths of EC for clearing electrode

– EC for Velec = + 300 V was much smaller than the case for Velec = 0V.

2009/4/19

Preliminary result(2009)1585 bunches

(Bs ~ 6 ns)

[Linear scale]

4x10-6[Linear scale]

4x10-8

TILC09, Tsukuba 16

Clearing electrode Effect of electrode voltage (Velec)

– Drastic decrease in electron density was demonstrated by applying positive voltage.

2009/4/19


Velec = 0 V Velec = 0 V

[Log scale] [Log scale]

+500 V+500 V

(b)

TILC09, Tsukuba 17

Clearing electrode Effect of electrode voltage (Velec)

– Smooth decrease in density for positive Velec

Electron density decreased to 1/10 at Velec = + 100 ~ 200 V, 1/100 at Velec = + ~ 500 V

Vr = -1 kV

~1x1012 e-/m3

??1585 bunches(Bs ~ 6 ns)~1600 mA

[Log scale] Different bunch filling patterns[Log scale]

– Effective for various bunch filling patterns

TILC09, Tsukuba 18

Clearing electrode Simulation for measured electron

current is undergoing.– Electrons moves only along B field.

Behaviors were roughly reproduced, but those at the negative Velec were not still understood.– Effect of the repeller grid?

2009/4/19

Measurement Simulation (dmax = 1.2)

800 bunches(Bs = 6 ns)800 mA

Model

[Log scale]

??

TILC09, Tsukuba 19

Clearing electrode A key issue: development of a reliable connection

to feed through– We had a trouble in the previous version.– A revised electrode is under test now (2009).

2009/4/19

f8

Discharging !

Electrode

Feed through

Feed through

TILC09, Tsukuba 20

Groove Groove structure can geometrically reduce the

effective SEY. Effective even in magnet.

2009/4/19

Grooved structure in magnet (simulation)

0 100 200 300 400 500 600 7000.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

2

Energy (eV)

SE

Y

0=1.50,Height=1.9mm, =200

Flat surfacer=0.14mm,B=2 Teslar=0.14mm,B=0.2Teslar=0.09mm,B=2 Teslar=0.09mm,B=0.2Teslaaverage,B=2 Tesla

L. Wang, T. O. Raubenheimer and G. StupakovNIM A571 (2007) 588.

(a) Isosceles triangular surface

(b) Sawtooth surface

(c) Rectangular surface

TILC09, Tsukuba 21

Groove The experiment was carried out under

collaboration with SLAC (M. Pivi and L. Wang) EC was measured at the same condition to that

for clearing electrode.– The same experimental setup used in the case of electrode

2009/4/19

Wiggler magnetsB = 0.77 T

R47

Beam

[Groove]

[Monitor]

Mag

neti

c field

Y. Suetsugu, H. Fukuma, M. Pivi and L. WangTo be published in NIM-PR-A

TILC09, Tsukuba 22

Groove Triangular-type groove structure, with TiN coating Compared with the data for a flat surface (TiN)

and clearing electrode (W)

2009/4/19

Designed and manufactured in SLAC

TILC09, Tsukuba 23

Groove Spatial growths of EC for groove and flat surface

– EC for the groove was much smaller than that for flat surface.

– Small change in the spatial behavior Small SEY?

2009/4/19

Preliminary result1585 bunches(Bs ~ 6 ns)

[Linear scale]

8x10-7

[Linear scale]

8x10-7

TILC09, Tsukuba 24

Groove Electron density for the groove was lower than

that for the flat surface by ~ one order.

2009/4/19

Vr = 0 kV

Preliminary result1585 bunches(Bs ~ 6 ns)~1600 mA[Log scale]

Integrated Beam Current [mA h]

Vr = 0 kV

Integrated Electron Current [mA h]

Aging was still proceeding, if plotted by electron dose (integrated electron current).

TILC09, Tsukuba 25

Groove However, the electron density is still higher than

the case of a clearing electrode with Velec > + 300 V by more than one order of magnitude.

2009/4/19

Vr = -1 kV

Preliminary result


[Log scale]

(#4)

TILC09, Tsukuba 26

Groove A new groove structure with a height of 2 mm is

now under consideration (20). It will be tested in 2009.

2009/4/19

2 mm

TILC09, Tsukuba 27

Comparison Both methods can be used in magnetic fields. Compared to methods using thin coatings, these

methods guarantee long-term stability. A common key issue is the beam impedance. Clearing electrode is more effective than TiN-

coated flat surface, and the TiN-coated groove, in reducing the electron cloud.

Clearing electrode required a power supply for each electrode. The development of a reliable electrical connection of the feed-through to the electrode is a key issue

Manufacturing cost may be low for a clearing electrode if the thermal-spray method can be used. Sufficiently accurate manufacturing of grooves might increase the cost.2009/4/19

TILC09, Tsukuba 28

Summary Various EC studies are undergoing at KEKB Updates:

– Measurement of electron density in a solenoid field and Q-magnet has just started, and the preliminary values were obtained for the first time.

– Clearing electrode in bending magnetic field was found to be very effective in reducing electron density

– Groove surface in bending magnetic field was also very effective, next to the electrode.

Next step– Development of a reliable feed through for electrode– Development of a reliable manufacturing process for groove– Estimation of impedance– Suitable choice of technique

2009/4/19

TILC09, Tsukuba 29

Backup slides

2009/4/19

TILC09, Tsukuba 30

Groove Loss factors and kick factors when the groove or

electrode is tilted against a beam.

2009/4/19

sz = 8 mm

Electrode

Groove

Electrode

Groove

Preliminary result

TILC09, Tsukuba 31

Clearing Electrode Simulation

2009/4/19

10x [mm]

-10

y [m

m]

10

-10

0

0

Preliminary result

TILC09, Tsukuba 32

Clearing electrode New strip-line type electrode was developed. Very thin electrode and insulator;

– Insulator: ~0.2 mm, Al2O3, by thermal spray.– Electrode: ~0.1 mm, Tungsten, by thermal spray.

Low beam impedance, high thermal conductivity– Input power ~ 100 W

2009/4/19

Y. Suetsugu, H. Fukuma, M. Pivi and L. WangNIM-PR-A, 598 (2008) 372

440 mm

to Feed through

40 mm

Stainless steel

TungstenAl2O3

to Feed through

TILC09, Tsukuba 33

Head tail instability Measurement of synchro-betatron sidebands

– Direct evidence of Head-Tail instability due to EC

2009/4/19

Bunch Oscillation Recorder (BOR)– Digitizer synched to RF clock, plus 20-MByte memory.– Can record 4096 turns x 5120 buckets worth of data.– Calculate Fourier power spectrum of each bunch separately.

Sideband appears at beam-size blow-up threshold, initially at ~ nb + ns, with separation distance from nb increasing as cloud density increases.

Sideband peak moves with betatron peak when betatron tune is changed.

Sideband separation from nb changes with change in ns.

J.W. Flanagan et al,PRL 94, 054801 (2005)

TILC09, Tsukuba 34

Head tail instability Simulations of electron cloud induced head-tail

instability (PEHTS)– The behavior is consistent with simulation

2009/4/19

E. Benedetto and K. Ohmi

TILC09, Tsukuba 35

Head tail instability Presence of sidebands also associated with loss

of luminosity during collision. Decaying cloud, and constant bunch current

2009/4/19

J.W. Flanagan et al, Proc. PAC05

Specific luminosity of high-cloud witness bunch is lower than that of regular bunches when leading bunch current is above 0.4 mA, but is the same below 0.4 mA.

Consistent with sideband behavior, and explanation that loss of specific luminosity is due to electron cloud instability.

Sideband Height (Non-colliding bunch)

Sp

ectr

al P

ow

er (

arb

.)

Current of proceeding bunch (mA)

Sp

ecif

ic L

um

ino

sity

(ar

b.)

- . . I . . I . . . I . . I . I . IProceeding bunch

Test bunch

TILC09, Tsukuba 36

Clearing electrode RF properties (calculation by MAFIA)

– Thin electrode and insulator Low beam impedance

2009/4/19

0.2 mm electrode 0.2 mm Al2O3

Electrode only(2 electrodes).

Impe

danc

e (Z

//) [W

]

Frequency [Hz]Lo

ss F

acto

r [V

C-1]

Thickness of Insulator [mm]

• Z// ~ a few Ohm• Z// reduced to ~1/5 compared to

the case of 1 mm thick.• R/Q ~ 0.1

• k ~1.5x1010 V/C including the connection part (2 electrodes).

• Dissipated power is ~ 120 W for 1 electrode. (@1.6 A,1585 bunches)

updates of electron cloud studies at kekb

Documents