accelerator design summary

• There are presently two designs, eRHIC and ELIC.– For eRHIC, the Ring-Ring option with an electron

ring 1/3 the size of RHIC is the present point design, however,

– the Ring-Linac option will be maintained and developed.

– ELIC is a “green-field” design being optimized for spin preservation & handling and for potentially higher luminosity than eRHIC.

Accelerator Design Summary

U. Wienands, 2nd EIC Wkshp, JlabAccel. Design Summary, 15-Mar-04

2

eRHIC Ring-Ring

AGS

BOOSTER

RHIC

e-cooling

LINAC

EBIS

recirculating linac injector5-10 GeV static electron ring

V. Ptitsyn, BNLV. Ptitsyn, BNL


3

eRHIC Ring-LinacV. Litvinenko, BNLV. Litvinenko, BNL


4

Ion Linac and pre-booster

IR IR

Beam Dump

Snake

CEBAF with Energy Recovery

3-7 GeV electrons 30- 150 GeV light ions

Solenoid


IR IR

Beam Dump

Snake


3-7 GeV electrons 30- 150 GeV light ions

Solenoid


IR IR

Beam Dump

Snake


3 -7 GeV electrons 30 -150 GeV light ions

Solenoid

Electron Injector

Electron Cooling

ELIC LayoutELIC LayoutL. Merminga, JlabL. Merminga, Jlab


5

Design Parameter comparisoneRHICRing-Ring

eRHICLinac-Ring

ELIC

Luminosity (e-p)

4.4…1.5E32

1E33…1E34

1E33…1E35

Ions …U92+ …U92+ …6Li+++

Ep (GeV) 50…250 50…250 30…150

Ielectron (A)/ppb 0.45/1E11 0.1…1E11 2.3…4.1/1E10

ppb (proton) 1E11 1…2E11 4E9

fcoll (MHz) 28 28 1500

lb (p/ion)(cm) 20 20 0.5

lb (e–) (cm) 1…2 1 0.5

p .0065 .005 .01

e .08 – .09

ß*p/e (m) .27/.27 (y) .26/0.3…1 0.005


6

Presentations• We heard presentations on

– Electron and ion sources and polarization (Farkondeh, Poelker, Roser, Derbenev & Dudnikov, Barber)

– Cold and intense beams (Skrinsky, Kroc, Derbenev, Dudnikov)

– Energy recovering linacs (Litvinenko, Calaga, Krafft)

– “Luminosity” (Hoffstaetter, Wei, Lebedev, Montag, Hyde-Wright, Wang, Masuzawa)

• I will emphasize the WG contributions


7

Sources• Polarized Proton & ion sources

– BNL has the KEK-TRIUMF-BNL OPPIS – 1.6 mA (1E12 pp) H– @ 90% polarization– EBIS for 3He++ under development at BNL,

2E11, 70…75% polarization– Atomic-beam Ion Sources (ABS) may yield

somewhat higher polarization at lower current

– 2D source: 90% Pz and Pzz vs 50…60%

– a 6Li+++ source also may be feasible.


8

•Vector spin polarized D- ion beam in excess of 1.0 mA can be produced in the OPPIS, as well as in the atomic beam source with resonant plasma ionizer.•J.Alessi and A.Zelenski proposed to use an EBIS (Electron Beam Ion Source under development at BNL) for nuclear polarized 3He gas ionization to 3He++ ions. The polarized 3He will be produced by conventional technique of optical pumping in metastable states. The expected beam intensity is about 2•1011 3He++ ions/pulse, polarization 70-75%.

Polarized D- and He++ ion sources

EBIS test stand

T. Roser, BNLT. Roser, BNL


9

Sources (cont’d)

• Electron Sources– Irradiate a GaAs cathode with circular polarized

laser light (≈800 nm) & collect pol e–.– Strained lattice & superlattice cathodes have

largely overcome the surface charge limit (e.g. SLAC), => no issue for eRHIC ring-ring

– For ring-linac, ELIC at issue are rep. rates for the laser, up to 100 MHz lasers now available, combination of power & rep. rate not yet.

– Increase area on cathode to increase charge– eRHIC: using FEL => large-scale project


10

ELIC e-Beam SpecificationsTypical parameters;

• Ave injector gun current 2.5 mA (and then 25 mA)• Micropulse bunch charge 1.6 nC• Micropulse rep rate 150 MHz (and then 1.5 GHz)• Macropulse rep rate ~ 2 kHz, 0.5 ms duration.

Circulator Ring

Injector

I

t

1/fc CCR

/c ~100 CCR

/c

I C = 1.5 km

CR

t

M. Poelker, JlabM. Poelker, Jlab


11

Continuing Trend Towards Higher Average Beam Current

ELIC with circulator ring

JLab FEL program with unpolarized

beam

Year

Ave

. Bea

m C

urre

nt (

mA

)

First polarized beam from GaAs photogun

First low polarization, then

high polarization at CEBAF

Source requirements for ELIC less demanding with circulator ring. Big difference compared to past talks. Few mA’s versus >> 100 mA of highly polarized beam.

M. Poelker, JlabM. Poelker, Jlab


12

Polarization, e–

• Electron polarization (eRHIC):– Significant progress has been made (10

GeV):• Feasible design for spin rotators incl. spin

matching.• eRHIC electron ring spin tracking studies

(incl. spin rotators, excl. detector)

• may produce >80% polarization, pol≈20 min (small e– ring helps!)

• will require excellent alignment & orbit correction (50µm rms)

• at lower energy: some shift of pol. vector


13

ELIC Polarization vs Energy

QuickTime™ and aTIFF (LZW) decompressor

are needed to see this picture.

D. Barber, DESYD. Barber, DESY


14

Polarization: ELIC, e–

• “Figure 8” design elegant way to control spin– Polarization axis given by “controlled

imperfection” (e.g. solenoid).– Smaller no. of snakes required– Potentially up to 4 IPs with longitudinal poln.


15

(Slide of Fig. 8 ring with snake)


16

Polarization: p• Polarized protons demonstrated in RHIC

(30%)– dominated by AGS depolarization

• AGS upgrade program to increase polarization

• Development of helical spin rotators/snakes


18

Polarization survival in RHIC (store # 3713)

Acceleration and squeeze ramp

Spin rotator ramp

Some loss during accel/squeeze ramp(Tune too close to ¼)

No loss during spin rotator ramp and during store

T. Roser, BNLT. Roser, BNL


19

e-Cooling• eRHIC ring-linac and ELIC proposals require e-

cooling to work. eRHIC ring-ring would profit– e-Cooling is necessary to

• combat IBS (beam emittance)• shorten the bunches (5 mm for ELIC)

– Can make flat beams (ELIC)• effect of IBS greatly reduced

• All cooling schemes require high power electron beams (50…75 MeV, ≈1 A) – Factor 10 beyond FNAL Recycler cooler


20

Electron Cooling (cont’d)

• Because of the high beam power involved– Electron ring & energy recovery are

required.– Stringent requirements on beam quality &

collinearity of e and p beams– “Hollow” electron beams may help reduce

recombination rates (Skrinsky).

• Highly ambitious R&D projects, Recycler experience will be invaluable.– BNL is launching R&D project to

demonstrate feasibility for RHIC upgrade.


21

R. Calaga, BNLR. Calaga, BNL


22

Electron Cooling System Parameters Parameter Value Units

Electrostatic Accelerator Terminal Voltage 4.3 MV Electron Beam Current 0.5 A Terminal Voltage Ripple 500 V (FWHM) Cathode Radius 2.5 mm Gun Solenoid Field 600 G

Cooling Section Length 20 m Solenoid Field 150 G Vacuum Pressure 0.1 nTorr Electron Beam Radius 6 mm Beam angular spread

≤ 8

0 µrad

Schematic Layout of the Recycler Electron Cooling

T. Kroc, FNALT. Kroc, FNAL


23

Energy Recovering Linacs• ERLs are essential in most scenarios

– Directl: ELIC, eRHIC ring-linac– Indirect: in the electron cooler

• Principle demonstrated in Jlab FEL at low energy, recently at CEBAF at 1 GeV, 80 µA.

• High-current high-energy operation remains to be demonstrated

• Cavities for high-current ERL are being designed.


24

R. Calaga, BNLR. Calaga, BNL


25

Luminosity Considerations• Everything else being equal, luminosity

is prop. to I•/ß*.

• The linac scenarios gain on e by allowing high tune shift (disruption) of the e– beam.

• The ring-ring scenario somewhat makes up by higher I.

• However, how much beam loss (e.g. from halo generation) can the energy recovery replenish?


26

Gedanken ExperimentFor round, equal sized beams, the following scaling applies:

e

eee

re

I

*βγ

=L

Comparing linac-ring colliders and ring-ring colliders, what can change for the better?

1. Maximum Ie/e is set by ION ring stability. The same in the two cases

2. γe set by the physics. The same in the two cases3. Minimum ß* is set by IR region design issues. Can it be too much

better for linac-ring? Should not be any worse than for ring-ring

4. re is set by (God, Yahweh, Allah, …); YOU cannot change it5. If there are to be luminosity enhancements to be found for linac-ring

designs compared to ring-ring designs, they must arise because one is allowed to make the equivalent tune shift e bigger for linac-ring colliders.

6. Finding the physical phenomena that determine the maximum e are extremely important for evaluating the linac-ring idea.

G. Krafft, JlabG. Krafft, Jlab


28

Luminosity (cont’d)• Beam-beam simulations are underway

to reach insight• Two approaches:

– Coulomb Sum– PIC

• Benchmark: HERA!


29

013.0=Δ simeyν

Simulated coherent modes

m0.4* =eyβ272.0

041.0

==

ey

ex

013.0

009.0

=Δ

=Δmey

mex

νν

pxν

exνeyν

pyν

exex ν − eyey ν −

0/ ff x 0/ ff x

0/ ff x0/ ff x

003.0=Δ simexν

+why?

how ?

(From work with Jack Shi, KU)

082.0

027.0

==

ey

ex

dQdQ

G. Hoffstaetter, CornellG. Hoffstaetter, Cornell


30

LuminosityG. Krafft, JlabG. Krafft, Jlab


31

The electron beam parametersRequirement Reason Concerns & Measures

Beam emittance (uncoupled x, nm)

40-60 (10 GeV)50-90 (5GeV)

Match ion beam Arc lattice Wiggler || superbend

Beam y/x emittance ratio

~0.2 High luminosity 70% polarization ? High Peq ~ high Ke, HERA update? study

Damping decrement Damping time < ~25 ms at 5GeV?

Less beam-beam limit reduction at low E

Wiggler || superbendfor low E operations

Bunch intensity(120 bunches)

11011

(0.45A)

High luminosity Vacuum chamber (syn. radiation), RF, instability …

Injection On energy, top-off or continues

Integrated luminosity.High e b-b limits lead to short lifetime

On energy Injection, flexible bunch-bunch filling.

Beam-beam tune shift limit

y ~ 0.08 B-factory achieved Working point near integer(spin), study

Coherent b-b effect in Unequal-circumference collider

Increase instability region ?

Study

F. Wang, BatesF. Wang, Bates


32

Scratch of a “super-bend” for radiation enhancement at 5 GeV

Red: normal bend

Blue: center bend only

All bends on

Center bend on only

(m) 70.3m 23.4

P(MW)

~0.35 ~1.06

x

(msec)

~54.5 ~18.1

y reduction ~ 20%

(Compare to 10 GeV)

20cm0-5

e-ring path length adj. requirement (with super-bends)*Total path length increase: ~4.47cm.* Linear rad. power at 10 GeV ~14kW/m

e-p(GeV) 5/250 10/250 10/50

F. Wang, MIT BatesF. Wang, MIT Bates


33

Intra-beam phenomena in RHIC• IBS: intra-beam small-angle Coulomb scattering

primary luminosity limiting factor in an heavy-ion storage ringRutherford scattering cross section ~ Z4 / A2

Luminosity degradation

J. Wei, BNLJ. Wei, BNL


34

Flat colliding beams equilibrium

y

x

P

e

Multiple IBS

Touschek scattering

Luminosity is determined by the beam area

IBS effect is reduced by a factor of the beam size aspect ratio

Cooling effect at equilibrium can be enhanced by flattening the electron beam in cooling section solenoid

x – emittance is determined by the IBS vs horizontal cooling

y – emittance is limited by the beam-beam interaction

At low coupling, cooling results in flat beams

Ya. Derbenev, JlabYa. Derbenev, Jlab


35

IBS in the Tevatron

V. Lebedev, FNALV. Lebedev, FNAL


36

Tevatron Measurements

V. Lebedev, FNALV. Lebedev, FNAL


37

Crossing Angle• Non-zero crossing angle causes luminosity

loss (geometry) and synchro-betatron coupling limiting intensity, thus luminosity.

• “Crab crossing” aligns the bunches in space such that they collide head-on, albeit in a transversely moving system, thus allowing a crossing angle, simplifying IR design greatly.

• 1st test of scheme likely at KEKB.– One cavity/ring only => c.o. for head of bunch is

different than for tail, crabbing everywhere

• Integral part of ELIC (100 mr crossing angle)


38

I.R. 20

I.R. 90

I.D. 188

I.D. 120

I.D. 30

I.D. 240

Input Coupler

Monitor Port

I.R.241.5

483

866Coaxial Coupler

scale (cm)

0 50 100 150

Superconducting Crab Cavity

K.Hosoyama (MAC 2004)

Crab Cavity Group

KEK Crab Cavity R&D GroupK. Hosoyama, K. Hara, A. Kabe, Y. Kojima, Y. Morita, H. NakaiA. Honma, A. Terashima, K. NakanishiMHIS. Matsuoka, T. Yanagisawa

M. Masuzawa, KEKM. Masuzawa, KEK


39

Crab Crossing for ELIC• Short bunches also make feasible the Crab Crossing:• SRF deflectors 1.5 GHz can be used to create a proper

bunch tilt

E

leB

F

ttt

tfcr

=

==

θ

λθαα 22

mm

ml

mMVGB

GHzcm

mF

GeVE

f

t

t

1

4

)/20(600

5.1(20

3

100

=

=

==

=

=

=

σ

λ4105

1.0−⋅=

=

t

cr

θ

α2

l

Ya. Derbenev, JlabYa. Derbenev, Jlab


40

IR issues• IR design is critical for colliders

– Electron beams esp. challenging due to handling of synchrotron radiation

– It is imperative that detector and accelerator people work together to reach a feasible design• Physics equirements (e.g. small-angle

detectors, low backgrounds)• Machine requirements (beam separation,

focusing, vacuum)


41

C. Montag, BNLC. Montag, BNL


42

C. Montag, BNLC. Montag, BNL


43

V. Litvinenko, BNLV. Litvinenko, BNL


44

Physics Requirements• The IR design will need more physics

input:– (C. Hyde-Wright, ODU)

• Forward tagging• Hadron beam tagging• Recoil protons• Neutron detection• …

• Most of these require detector access to small/zero angle, spectrometer magnets etc.


45

• First few elements in lattice should be

designed with thought to detection of forward fragments

• Compact detectors near 0deg can enhance physics program.

C. Hyde-Wright, ODUC. Hyde-Wright, ODU


46

Electron-cloud effect– Positively charged beams are subject to

electron-cloud formation & emittance blowup• ISR, PSR, B-Factories, RHIC, BINP rings,…

– Solenoids work well in drift regions, but are unlikely to work in magnets.

– Electron rings are not safe either: fast ion instability can affect even the linac scenarios.


47

Models of two-stream instability

• The beam- induces electron cloud buildup and development of two-stream e-p instability is one of major concern for all projects with high beam intensity and brightness [1,2].

• In the discussing models of e-p instability, transverse beam oscillations is excited by relative coherent oscillation of beam particles (protons, ions, electrons) and compensating particles (electrons,ions) [3,4,5].

• For instability a bounce frequency of electron’s oscillation in potential of proton’s beam should be close to any mode of betatron frequency of beam in the laboratory frame.

1. http://wwwslap.cern.ch/collective/electron-cloud/.2. http://conference.kek.jp/two-stream/.3. G.I.Budker, Sov.Atomic Energy, 5,9,(1956).4. B.V. Chirikov, Sov.Atomic.Energy,19(3),239,(1965).5. M.Giovannozzi, E.Metral, G.Metral, G.Rumolo,and F. Zimmerman , Phys.Rev. ST-Accel. Beams,6,010101,(2003).

V. Dudnikov, BNLV. Dudnikov, BNL

http://wwwslap.cern.ch/collective/electron-cloud/




48

Instability in RHIC, from PAC03

V. Dudnikov, BNLV. Dudnikov, BNL


49

R&D issues for ELIC and LR-eRHIC High intensity polarized and unpolarized electron gun

• Currently a few mA Up to 450 mA / 16nC

• Currently a few 100 A of polarized beam GaAs photo injector at 80% pol. Up to 450 mA electron current at 80% pol. Methods to overcome the surface charge limit for 16nC/bunch Beam emittance control for 16nC/bunch and a large source diameter (14mm) Test and improvement of cathode lifetimes

Electron Cooling at high energies

• Currently a frew 100MeV, soon 8.9GeV/c pbar at the FNAL recycler For LR-EIC: Cooling of Au or light ions up to 100GeV, p at 27GeV New technology: ERL cooling + cooling with bunched e-beam Limits to the ion emittance with e-cooling (especially vertically) and with all noise processes. Allowable beam beam parameters for ions, especially with electron cooling



50

R&D issues for ELIC and LR-eRHIC

IR design, detector integration, saturation in special magnets, optimization … Halo development by beam disruption, especially at low electron energies Impact of beam disruption on following IRs Ion-beam dynamics with crab cavities

High current ERLs

• Currently strong influence of small e-beam oscillations on p-emittance in HERA Stabilization of the e-beam + influence on the ion beam Current limits by multi-pass Beam-Breakup instability CW operation of high filed cavities, stabilization, heat loss Influence of HOMs with large frequencis (>2GHz) R/Q and Q agreement with calculations including absorbers



51

R&D specific to ELIC Spin resonances in Figure 8 rings Stability of non-vertical polarization in figure 8 rings and in the ERL Stable beam in a 100 turn circulator ring Crab cavity R&D and crab cavity beam dynamics Beam beam resonance enhancement when operating close to the hourglass effect Limits to the bunch length, since this limits the beta function

R&D specific to LR-eRHIC 1kW FEL at 840nm Heating of the cathod / problems associated with large spot size (14mm) Production of very high polarized e-beam



52

Issues for further Study

• eRHIC– Hirata-Keil coherent

beam-beam modes– Restriction on e- beam

energy (pol, lumi)– effect of different ion

energy on electron orbit

– maximize no. of bunches in RHIC

• ELIC– Parameters have

been pushed into new territory…

• ß, lb, ring shape, crab crossing,…

– benefits of circulator ring vs “real” storage ring

• Both proposals:– Interaction region, at different energies, with spin rotators


53

At last…• There has been much progress over the last

years, the eRHIC design is maturing.• ERL technology demonstrated at CEBAF at 1

GeV• A rigorous e-cooling R&D program est’d. at

BNL• ELIC proposes some very elegant and

innovative features worth further investigation.

• Thank you to all speakers and the organizers for a very lively workshop.

accelerator design summary

Documents

ringring option

erhic ringringfor ringlinac

erhic ringringv

polarized d ion beam

ringlinac option

ebis electron beam ion

spin rotators

atomic beam source