Status of the accelerator upgrade

Dec.10 2008 Mika Masuzawa

mm 12/10/2008

Contents1. Introduction

– Luminosity goal2. Strategy

– Higher currents– Smaller y

*

– Increase y

3. What we need– Components for higher currents

• R&D Status– IR & Lattice design for smaller y

* & larger y • Summary

mm 12/10/2008

1. Introduction

Luminosity goal

mm 12/10/2008

Luminosity goal

mm 12/10/2008

Target: 5-8 × 1035 cm-2s-1

=30-50 x World Record (KEKB)

mm 12/10/2008

2. Strategy

mm 12/10/2008

Strategy

(1) Increase beam currents•1.7 A (LER) / 1.4 A (HER) → 9.6 A (LER) / 4.1 A (HER)(2) Smaller y

*

•6.5(LER)/5.9(HER) mm→3.0/3.0 mm(3) Increase y

•0.09 (with Crab) → 0.29mm 12/10/2008

Crab cavity

The superconducting cavities will be upgraded to absorb more higher-order mode power up to 50 kW.

The beam pipes and all vacuum components will be replaced with higher-current design.

The state-of-art ARES copper cavities will be upgraded with higher energy storage ratio to support higher current.

SuperKEKBe- 4.1 A

e+ 9.4 A

*y= σz = 3 mm

Will reach 5-8 × 1035 cm-2s-1.

Tunnel already exists.Most of the components (magnets, klystrons,etc) will be re-used.

Comparison of parametersKEKB Design

(A)KEKB

(achieved)KEKB upgrade

(B)Gain Factor

(B/A)

y* (mm) 10 5.9/6.5

(5.9/5.9) 3 3.3

y 0.052 0.056 (0.093) 0.296

Ry 0.885 0.879(1.15) 1.11

y/Ry 0.059 0.063(0.081) 0.267 4.5

Ibeam (H/L) (A) 1.1/2.6 1.45/1.8 (0.95/1.62) 4.1/9.4 3.6

RL 0.845 0.715(0.827) 0.86 1.02

Luminosity (1034 cm-2 s-1) 1 1.71 (1.46) 55 55

z (mm) 4 ~ 7 5(LER)/3(HER)mm 12/10/2008

( ): with crab

Strategy

3. What we need

Components for higher currentsR&D Status

mm 12/10/2008

What we needfor (1) higher beam currents

mm 12/10/2008

a.Vacuum components (pipes/bellows…)b.Modification of the monitors (BPMs,SRMs…)c.Longitudinal bunch-by-bunch FB systemd.More RF cavities and klystronse.Modifications of the RF systems for higher currentsf.New Crab cavities for SuperKEKB

Vacuum components⇨talk by K.Shibata tomorrow

mm 12/10/2008

Beam current increase causes:

Intense Synchrotron Radiation power•27.8 kW/m in LER, twice as high as in KEKB•21.6 kW/m in HER, 4 times as high as in KEKBHigh photon density•Photon density ~1x1019 photons/m/s in average

•Large gas desorption•Gas load ~ 5x10-8 Pa m3/s/m (for h = 1x10-6 molecules/photon)•Average pressure ~ 5x10-7 Pa for S ~ 0.1 m3/s/m•Electron Cloud Instability (ECI) becomes a big issue in positron ring

Heating due to Higher Order Modes (HOM)• For a loss factor of 1 V/pC, loss power ~ 200 kW

a) Vacuum components

Beam ductCopper beam duct with ante-chambers

– Copper is required to withstand intense SR power

Features (compared to simple pipe):– Low SR power density– Low photoelectrons in beam pipe– Low beam impedance– Expensive

mm 12/10/2008

BeamSR

Pump

LER HER

Y.Suetsugu

a) Vacuum components

Some sections of KEKB LER have been replaced by ante-chamber type

Trial model of a copper beam duct with ante-chambers for arc section

The duct is bent with a radius of 16 m

mm 12/10/2008Y.Suetsugu

a) Vacuum components

More on suppressing photoelectronsTiN (Titanium nitride) coating on inner surfaceDecrease secondary electron yield (SEY): Max. SEY ~0.9A test stand for the coating was built in KEK, and applied to a test duct with ante-chambers.

⇨Decrease of electrons at high current region was demonstrated.

mm 12/10/2008

~4 m

90 mm

Electron numbers

Y.Suetsugu

a) Vacuum components

And more on suppressing photoelectronsClearing Electrode A possible measure even inside of magnet An electrode with a low beam impedance was developed and tested with beam (in the presence of wiggler magnetic field of B = 0.77 T). Thin structure Electrons decreased to ~1/100 by applying + 300V.

mm 12/10/2008

Clearing electrode for test

Elec

tron

Cur

rent

[A]

Applied voltage [V]

Vr = 1.0 kVB = 0.77 T

V elec

[V]

I e [A]

1585 bunches(Bs ~ 6 ns)~1600 mA

Velec = 0 V

-500 V

+500 V

1 x 10-5

1 x 10-6

1 x 10-7

1 x 10-8

1 x 10-9

CollectorsY.Suetsugu

a) Vacuum components

Drift space

In wiggler magnet

And more on suppressing photoelectrons

Grooved Surface: being tested with the beam now.• Another possible measure in magnets• Decrease effective SEY structurally

TiN~50 nm

SS + TiN coating

M. Pivi

mm 12/10/2008Y.Suetsugu et.al

a) Vacuum components

Grooved surface currently tested

mm 12/10/2008

a) Vacuum components

For lower impedance: RF shielding Comb-type RF shield Features (compared to finger type):

Low beam impedance High thermal strength Applicable to complex aperture Little flexibility (offset)

Effect of RF shielding was demonstrated in KEKB. Finger-type as an option

If more flexibility is required.

Without cooling fan

Comb-type RF-shield

Temperature of bellows

a) Vacuum components

mm 12/10/2008Y.Suetsugu

More for lower impedance: Bellows and gate valves

Comb-type RF shield is adaptable to a complicated aperture of beam duct with antechambers.

Bellows Gate valve

Y.Suetsugu et.al

a) Vacuum components

mm 12/10/2008

And more for lower impedance: Movable maskBig impedance sources in the ring•Planning to use “stealth” type (Ver.6)–Low beam impedance•Present Ver.4 ~ 1V/pC ( 90 mm) 200 kW power loss–Loss factor decreases to ~1/10 ( 90 mm).•Manageable by conventional HOM absorber

Loss factor

Bunch length [mm]

Loss

fact

or [V

/pC]

Ver.4

Ver.6

a) Vacuum components

mm 12/10/2008Y.Suetsugu

And more for lower impedance: movable mask

• Trial models were installed and tested with beam.– Principle was proved experimentally.

• Temperature rise of bellows decreased to 1/2 ~ 1/3.– But, could not withstand a high intensity beam yet.

• Start with Ver.4 ?– While beam current is low.

Head of Ver.6 (trial model)

Temperature of bellows near masks

Beam

HeadSupport

Y.Suetsugu et.al

a) Vacuum components

mm 12/10/2008

Movable mask (beam test)• Principle was proved experimentally.• Improvement for high-current is currently underway

Temperatures of bellows Temperatures of SiC cooling water

Old type

New type

Old type

New type

Y.Suetsugu et.al

a) Vacuum components

mm 12/10/2008

And more for lower impedance:Connection flange

MO-type flange, which can makes very smooth inner surface low impedance

SUS flange

Copper alloy flange (under test)

Flange and gasket (copper)

Y.Suetsugu et.al

a) Vacuum components

mm 12/10/2008

Remaining Vacuum Issues• Optimization of beam duct shape

– CSR (Coherent Synchrotron Radiation) problem have a great effect on the luminosity.

• Deformation of beam duct by heating– Displacement gauge for every BPM?

(50)

Larger aperture OR Smaller aperture

Low impedance Cure for CSR(ex, resistive wall) (Low cost?)Effective pumping

(90)?

Y.Suetsugu et.al

a) Vacuum components

mm 12/10/2008

Modification of the monitors :Beam Position Monitors

• Use same front-end electronics.

• New button electrodes◆ New connector design for

improved reliability.◆ 12 mm →6 mm diameter Signal power same as at

present, at higher beam currents, to match dynamic range of existing front-end electronics.

M. Tejima et al.

b) Modification of the monitors

mm 12/10/2008

Modification of the monitors:Beam Size Monitors

•Chamber redesign for high current.•Source bending radius for synchrotron radiation monitors increases 3x to reduce mirror heating at high currents.

•Reduces beam image distortion at high currents. Also increases visible light flux.

•Development of x-ray monitor also being pursued based on x-ray astronomy technique of Coded Aperture Imaging for high-bandwidth/high-speed readout with low beam current dependence.

•Optics testing and detector development being carried out in collaboration with Cornell (CesrTA ILC damping ring study machine group) and U. Hawaii (Belle detector group).•Collaboration with other machines and between physics and accelerator groups.

1-D URA Mask for Coded ApertureImaging

J.Flanagan et.al

b) Modification of the monitors

mm 12/10/2008

Bunch-by-bunch Feed back system

mm 12/10/2008

Transverse feedback similar to present design →Target damping time 0.2ms-Detection frequency 2.0→2.5 GHz.-Transverse kicker needs work to handle higher currents-Improved cooling, supports for kicker plates.

Longitudinal feedback to handle ARES HOM & 0/ mode instability→Target damping time 1ms-Use DANE-type (low-Q cavity) kicker.

Digital FIR and memory board to be replaced by new Gboard.-Low noise, high speed (1.5 GHz), with custom filtering functions possible.-Extensive beam diagnostics.

A prototype of the new bunch-by-bunch feedback system (G-board / Gproto) has been developed as a result of collaboration between SLAC, KEK and INFN, and successfully tested at KEKB, KEK ATF, KEK PF, SLAC and DANE.

c) Bunch-by-bunch FB system

RF systemsNeed RF systems which can store high beam currents and handle shorter bunches.

ARES (normal-conducting cavity) for LER ARES + SCC (Single-cell Superconducting cavity) for HER

Adopt the same RF frequency as KEKB and use the existing RF system as much as possible, with improvements as necessary to meet the requirements for SuperKEKB.

→Construction cost is greatly reduced.→Technical uncertainties are relatively small.

d & e ) RF system for higher currents

mm 12/10/2008Akai et.al

The ARES CavityPassive stabilization with huge stored energy.Eliminate unnecessary modes by coupling of 3

cavities.Higher order mode dampers and absorbers.No need for longitudinal bunch-by-bunch

feedback.No transverse instability arises from the cavities.

Storage

Resonant

Acceleration

Accelerator Resonantly-coupled Energy Storage

T. Kageyama, et al

d & e ) RF system for higher currents

mm 12/10/2008

High power RF R&D

Y. Takeuchi, T. Kageyama, et al

d & e ) RF system for higher currents

mm 12/10/2008

◆ Higher HOM powerUpgrade of HOM damper

◆ Higher input RF power400 kW/cavity -> 800 kW/cavityR&D of input coupler using new test-stand.

Superconducting CavitySuperKEKB challenges:The expected power load to the HOMabsorber is 50 kW/cavity at 4.1 A,(even) with a larger beam pipe of 220 mmφ.HOM damper upgrade may be needed.

S. Mitsunobu, et al

d & e ) RF system for higher currents

mm 12/10/2008

mm 12/10/2008

Crab cavity with 10 A beam• The original cavity is designed for 1-2 A beam

– Simple structure, suitable for SC → High kick voltage is obtained by one cavity.

– Sufficient damping of parasitic modes.– Not necessarily optimized for a 10 A beam.

• Possible problems at 10 A– Large HOM power (200 kW)

• Loss factor is not very small, because the radius of coaxial beam pipe can not be widely opened.

• Additional loss factor comes from the absorber on wide beam pipe.

– Much heavier damping of HOM’s may be needed, particularly for horizontal polarization of transverse modes (large x at crab).

• A new crab cavity has been designed, which can be used at 10 A.

f) Crab cavity for higher currents

3. What we need

IR & lattice design for smaller y

*

& larger y

mm 12/10/2008

IR (Interaction Region)

mm 12/10/2008

Issues Causes Possible cures

Physical aperture Lower * Injection beam quality (Damping ring)

Dynamic aperture Lower * Injection beam quality (Damping ring)Move the IR magnets closer to IP⇨needs stronger magnets Local correction at the IR straight section in HER

Detector background Higher beam current Optimization of mask, IP beam duct shape etc. talk by M.Iwasaki⇨

Heating of components Higher beam current & shorter bunch length

Optimization of the IP beam duct, cooling systems etc.

For smaller *

Higher luminosity smaller ⇨ * causes issues such as

IR layout

•Crossing angle: 22 30 mrad⇨•Move final focus quadrupoles closer to IP for lower beta functions at IP.•Preserve current machine-detector boundary.•QCS and solenoid compensation magnets overlap in SuperKEKB.

mm 12/10/2008

For smaller *

mm 12/10/2008

IR-overviewpresent design

(based on the LoI)

QCSL(QCSL center from IP = 0.969 m)KEKB=1.6 m

QCSR (QCSR center from IP = 1.163 m)KEKB=1.92 m

ESR

ESL

IP

The cryostats were designed with keeping the KEKB boundary conditions with Belle detector.

N.Ohuchi

For smaller *

mm 12/10/2008

Cross section of QCSR

•6 layer coils (3-double pane cake coils)•Inner coil radius : 90.0 mm•Outer coil radius : 116.8 mm•Cable size : 1.1 mm 4.1 mm1.1 mm 7.0 mm (KEKB)•Number of turns : 271 in one pole1st layer = 38, 2nd layer = 393rd layer = 46, 4th layer = 475th layer = 50, 6th layer = 51•Field gradient : 40.124 T/m•Magnet current : 1186.7 A•Magnetic length : 0.299m (R), 0.358m (L)•Inductance : 69.98 mH (R), 83.79 mH (L)•Stored energy : 49.3 kJ (R), 59.0 kJ (L)•Operation temperature : 4.5 K•Operation point w.r.t. SC limit : 75% (R), 74% (L)

N.Ohuchi

IR magnets (QCS) : DesignFor smaller *

IR magnets (QCS R&D) : Field measurement results

mm 12/10/2008

Field gradient at 1186.7A

G=40.05 T/m was obtained (Design G=40.124T/m)MultipolesData at r=48mm•a3= -0.86 units, b3= 0.91 units

•a4= -1.27 units, b4= 0.40 units

•a5= 0.11 units, b5= -0.80 units

•a6= -0.55 units, b6= -0.00 units （ I units = 10-4 × b

2 ）

Design at Rref = 50 mm

•b6=0.12, b10=-0.04, b14=0.12

N.Ohuchi

For smaller *

IR magnets (QC1&QC2) : R&D on coil winding

mm 12/10/2008

Coil winding on a cone shaped bobbinR&D work required for a winding tool

N.Ohuchi

For smaller *

mm 12/10/2008

IP

4 layer structure　　　　 Integrated field gradient ＝ 17.445T　　　　 field gradient ＝ 98.9 T/m needed for new IR ⇦　　　　 Length ＝ 336mm　　　　 Cable （ the same type as QCS R&D 4.05mm×1.16mm ）　　　　 Current ＝ 2610A　　　　 Max field inside the magnet ＝ 8.8T 　（ corresponds to 10T iron bore magnet ）

N.Ohuchi

QCS @ 1.9 K (design)IR magnets (QCS) : 1.9K version for new optics

For smaller *

mm 12/10/2008

IR magnets (QCS) : 1.9K version for new optics to do list

1. IR design has to be fixed.

2. Design of Super-QC23. Design of compensation solenoids4. Optimization of QCS, QC2 and compensation solenoid

under the combined operation

5. Design of cryostats for 1.9K operation!

N.Ohuchi

For smaller *

Local chromaticity correction for widening dynamic aperture

KEKB Design Report

Effects of local correction- widen dynamic aperture- weaken synchro-betatron resonance

KEKB: local correction only in LER

KEKB upgrade:local correction also indownstream of HER

⇨ New magnets need to be designedand made/measured/installed.

Lattice Design

mm 12/10/2008

For smaller *

Dynamic aperture

Lattice DesignTravelling focus

mm 12/10/2008

We would like to have z ~ y*

Bunch length limitation due to CSRLER bunch z=5mm (not 3mm) when y

*=3mm⇨z > y*

Hourglass effect is not negligible.

⇨A possible cure is to use the travel focusing

Beam size changes along Z

Smaller *

Larger z

Hourglass effect

Travelling focus

mm 12/10/2008

Evaluation by Koiso-san

Needs a pair of sextupole magnets at both sides of the crab cavityANDTwo crab cavities each ring…

More evaluation of the effect on the luminosity will be studied.(how much do we gain by this scheme??Awfully expensive!)

Smaller *

Recovering from the hourglass effect

Lattice Designfor higher y

KEKB KEKB upgrade

y (max) with crab from b-b simulation

0.15 0.296

x* [m] 0.9 0.2

x (LER/HER) [nm] 18/24 12/13

y/x [%]1 0.5

x (LER/HER) 0.506/0.510 0.505/0.505Improving the performance with crab crossing is a mustBeam-beam parameter, beam lifetime ⇨ Injecter upgrade⇨presentation by Funakoshi-san on KEKB status

Other efforts for higher y areLower x

*, smaller x, y, smaller coupling ,,,,⇨Better magnet alignment⇨More auxiliary coils for optics diagnostics and correction

mm 12/10/2008

For higher y

Injecter upgradeSimultaneous (pulse-by-pulse) injection of KEKB LER/HER/PF

Phase I : Construction of a new beam transport which bypasses the energy compression system (ECS).

Phase II : Fast beam switch between KEKB e- and PF e- by a pulsed bend.

Phase III : Fast beam switch between KEKB e-, PF e- and KEKB e+ by an e+ target with a hole and pulsed steerings.

e- Gun (A1 Gun; for KEKB and e- Gun (A1 Gun; for KEKB and PFPF))

e- Gun (CT Gun; for PF-Ae- Gun (CT Gun; for PF-ARR ））

#3 Switchyard

KEKB/ KEKB/ ARAR

PFPF

pulsed BendECS

Crystal tungsten target5 mm

Hole: 3 mm

4.5 mm

Phase I and II have been completed. Beam study for Phase III is underway. Phase III will be started in this autumn.

Simultaneous injection of e- and e+ is desirable in view of thermal stability of 　 components and an ability of the operation with shorter beam lifetime.

(M. Satoh, KEKB review committee 2007)Positron Target

mm 12/10/2008

Injecter upgrade

HER / 12.5HzPF / 5Hz

N.Iida

On Dec.8, 2008First pulse-by-pulse injection to PF & KEKB HER!

mm 12/10/2008

Injecter upgradeSimultaneous injection of KEKB LER/HER/PF

Injecter upgrade

mm 12/10/2008

2004 data

Measurement data were used as alignment errors and simulation was done by Ohnishi.No need for leveling the tunnel but local ups and downs better be corrected.We also found that the magnets are rotated, horizontal position moved and so on…

Tunnel level Magnet level

KEKB tunnel continues to sink and magnets follow the tunnel

Better magnet alignmentFor higher y

Summary

mm 12/10/2008

Accelerator componentsR&D work of the accelerator components (vacuum, monitors and etc) is ongoing. Some have been tested with the beam.

IR optics IR optics is being re-designed with x=20 cm⇨Koiso-san’s talk on the new IR optics parallel session tomorrow.

IR magnetsQCS R&D version has been made and its magnetic field has been measured.IR magnets (QCS,QC1 and QC2) will all be superconducting.Evaluation of QCS at 1.9K shows a possibility of higher field gradient which is needed for the new IR optics with smaller x.⇨New cryostat for 1.9K needs to be designed.

Summary

mm 12/10/2008

Lattice designBunch length of HER 3 mm/LER 5 mm (due to CSR) is the current choice.Hourglass effect will degrade the luminosity. ⇨Travelling focus?Effect of travelling focus needs to be examined.

Better alignmentMagnet rotation/level measurement data have been taken and machine performance is simulated.Realignment (smoothing locally) will be needed though a leveling of the entire tunnel may not be necessary.

We are making steady progress toward 5-8 × 1035 cm-2s-1.

mm 12/10/2008

Contribution is welcome.