Issues for Optimization of a SuperB Factory John
Issues for Optimization of a Super-B Factory John T. Seeman SBF Workshop SLAC June 14, 2006
High Luminosity Topics m Horizontal m Vertical m Bunch emittance length and compression m Maximum beam-beam parameter m Crab cavities m Crab waist m Damping m IR ring circumference design m AC power vs synchrotron radiation power m Beam-beam effects makes optimum “non-linear” 2
History m m m m 2003 -2005: Both KEKB and PEP-II B-Factories have reached over 1 x 1034 /cm 2/s! 2002 -2005: Both KEK and SLAC pursued Super-B upgrades with significantly higher currents and lower by*. 2004: Super-KEKB design is presented to the KEK Laboratory. March 2005: Emerged from the Super-B Workshop in Hawaii a concept of a single pass Linear Super-B Factory November 2005: Frascati workshop outlined a Super-B design with single pass collisions, bunch compression, and energy recovery linacs. March 2006: Frascati workshop outlines a Super-B Collider design that collides every turn but with bunch compression or large crossing angles. Crab waists help. May 2006: Beam-beam simulations with more slices show larger tune shifts. 3
PEP-II Best shift and 24 hr 4
m 5
m 6
Super-B Basic concepts m Higher currents and large number of bunches make more luminosity but needs higher RF power. m Faster damping allows stronger collisions but needs higher RF power. m Shorter bunches allow smaller vertical spot sizes but larger HOMs. Crab waists helps. m Lower emittance makes for smaller spot sizes but leads to single ring enlargment issues. m Need small energy spread (~10 -3) in the collisions to match Y 4 S resonance shape. 7
Luminosity m The luminosity for a linear collider is: L=Hd Np P / 4 p E sx sy Hd : disruption enhancement P : average beam power m For a storage ring is: L=2. 17 x 1034(1+r) xy EI / by I : beam current zy : vertical tune shift by : IP vertical beta function 8
Scaling laws to optimize the single pass IP parameters m Disruption: m Luminosity m Energy spread: P. Raimondi Decrease sz + decrease N Increase spotsize Increase N Decrease spotsize Contradicting requests! Increase sz + decrease N Increase spotsize 9
More Detailed Luminosity Equation xy* = beam-beam parameter = Dy/4 p. N = number of particles by* = vertical beta function s = bunch spacing E = beam energy F = Hourglass + parasitic collision factor (~<1. 0) 10
Parameters of Super-B Designs xy Collider Units N by* s E F Lumin 1010 mm m Ge. V (~Hd) 1035 PEP-II Normal 0. 07 8 10 1. 26 3. 1 0. 84 0. 11 KEKB Normal 0. 065 5. 8 6 2. 1 3. 5 0. 76 0. 16 Super. PEP-II High I low by 0. 12 10 1. 7 0. 32 3. 5 0. 81 7 Super. KEKB High I low by 0. 28 12 3 0. 59 3. 5 0. 85 8 Linear Super. B Single pass 29. 10 0. 5 250 4 1. 07 10 Super. B Bunch shorten 0. 14 6 0. 4 0. 63 4 0. 75 10 Super. B X’ing angle 0. 045 2 0. 08 0. 5 5 0. 8 10 11
Final Focus m Use ILC final focus concept m Scale Ä to B Factory energies (4 x 7 Ge. V) (3. 75 x 7. 5 Ge. V) (3. 5 x 8 Ge. V) (3. 1 x 9 Ge. V). m Use small superconducting quadrupoles for the final doublet. m Chromatic corrections done in the IR with sextupoles. 12
A. Seryi m 13
m 14
m 15
m M. Sullivan 16
Beam-Beam Simulations m Track macro-particles through the collision to see blow up effects. (Guinea Pig code from ILC + Ohmi beam-beam code + Cai beam-beam code) m Compare single pass beam-beam effects with ring beam-beam effects. m Couple single pass collisions with damping in a ring. 17
Horizontal Collision Vertical collision Effective horizontal size during collision about 10 times smaller, vertical size 10 times larger Simulations by D. Schulte/Raimondi/Biagini 18
m 19
m May 2006 20
m 21
K. Oide May 2006 m 22
Single Pass Linear Super-B m Collide each bunch once very hard making a lot of luminosity. m Use very small beta functions at the IP like ILC. m Re-inject into the damping rings and damp for several damping times. m Collide again after the disrupted emittances have damped. m Simulations show vertical emittance blow up is about 300 times. m Need about 6 damping times between collisions. m Need very short damping times high power. 23
Linear-B scheme LER injection HER injection LER HER LER Bunch compressor and FF IP HER Bunch compressor and FF Overall ring length about 6 Km, Collision frequency about 120 Hz*10000 bunch_trains=1. 200 MHz Bunch train stays in the rings for 8. 3 msec, then is extracted, compressed and focused. After the collision the bunch is reinjected in its ring 24
Horizontal phase after the collision Vertical phase after the collision IP Parameters set considered at the workshop caused large increase of the emittance due to the collision: ex_out/ex_in=12 ey_out/ey_in=300 M. Biagini studies 25
Type 2 Energy Recover Linear Super-B m Use single pass collider idea but use energy recovery linacs (ERLs) to accelerate, collide, then deccelerate the beams to the reduce damping ring energy and energy loss to synchrotron radiation. m This technique saves energy lost in damping rings between ~E 2 and ~E 4. m Acceleration also adiabatically damps emittance and energy spread making bunch compression easier. 26
Linear Super B schemes with acceleration and energy recovery, to reduce power 2 Ge. V e+ injection 2 Ge. V e+ DR 4 Ge. V ee- Gun IP 5 Ge. V e+ SC Linac 4 Ge. V e- SC Linac 7 Ge. V e+ 1. 5 Ge. V Linac 2 Ge. V Linac m Use SC linacs to recover energy m Use lower energy damping rings to reduce synchrotron radiation m No electron damping ring m Make electrons fresh every cycle 1. 5 Ge. V Linac Damping Rings 2 Ge. V e- Dump Ä Damping time means time to radiate all energy Ä Why not make a fresh beam if storage time is greater than 1 damping time e- Gun e+ Gun 27 Linac
Type 3 Super-B with Bunch Compression m Collide m Install every turn in the damping ring. an ILC like final focus. m Choose parameters to cause small emittance blow up. m Use bunch compressor to shorten the bunches at the IP. Decompress after the IP. m Use monochromator scheme at IP to compensate the energy spread to match the small Y 4 S resonance. 28
Simplified layout in the Small Disruption Regime Collisions every Turn P. Raimondi ILC ring with ILC FF ILC Compressor, 0. 4 Ge. V S-Band or 1 Ge. V L-Band Crossing angle optional Decompressor FF IP FF Compressor 29 Compressor
m 30
Approximate AC Power m PEP-II (2200 m) (1 ring LER and HER) Ä RF AC Power LER at 3. 1 Ge. V = 1 Me. V/turn x 4 A x 2 = 8 MW Ä RF AC Power HER at 9. 0 Ge. V = 3. 6 Me. V/turn x 2 A x 2 = 14. 5 MW Ä Total RF AC power = 22. 5 MW. m Super-B factory (4400 m) (2 rings LER and HER) Ä RF AC Power LER at 4 Ge. V ~ 1. 8 Me. V/turn x 4 A x 2= 29 MW Ä RF AC Power HER at 7. 0 Ge. V ~ 2. 2 Me. V/turn x 2. 5 A x 2= 22 MW Ä Total RF AC power = 51 MW. 31
Type 4 Super-B with Large Crossing Angles m Collide with a large crossing angle (2 x 25 mrad) m Thus, only a small longitudinal part of each bunch collides with a small longitudinal part of the opposing bunch. (luminosity loss) m However, the vertical beta function can be made very small while keep the overall bunch length long. (luminosity gain) (better for beam instabilities) m Use m Do ILC final focus. not need strong damping. 32
Simplified layout in the Small Disruption Regime Collisions every turn Uncompressed bunches Crossing angle = 2*25 mrad Crabbed Y-Waist ILC ring & ILC FF FF IP FF P. Raimondi 33
x P. Raimondi e- e+ 2 Sx/q q 2 Sz*q z 2 Sx Vertical waist has to be a function of x: Z=0 for particles at –sx (- sx/2 at low current) Z= sx/q for particles at + sx (sx/2 at low current) 34
Horizontal Plane Vertical Plane Collisions with uncompressed beams Crossing angle = 2*25 mrad Relative Emittance growth per collision about 1. 5*10 -3 eyout/eyin=1. 0015 Raimondi 35
Parameters of Super-B Designs xy Collider Units N by* s E F Lumin 1010 mm m Ge. V (~Hd) 1035 PEP-II Normal 0. 068 8 11 1. 26 3. 1 0. 84 0. 10 KEKB Normal 0. 065 5. 8 6 2. 1 3. 5 0. 76 0. 16 Super. PEP-II High I low by 0. 12 10 1. 7 0. 32 3. 5 0. 81 7 Super. KEKB High I low by 0. 28 12 3 0. 59 3. 5 0. 85 8 Linear Super. B Single pass 29. 10 0. 5 250 4 1. 07 10 Super. B Bunch shorten 0. 14 6 0. 4 0. 63 4 0. 75 10 Super. B X’ing angle 0. 045 2 0. 08 0. 5 5 0. 8 10 36
Sigx* mm Etax mm Sigy nm Betx mm Bety mm Sigz_IP mm Sige_IP Sige_Lum Emix nm Emiy nm Emiz mm Cross_angle mrad Sigz_DR mm Sige_DR Np 10 e 10 Nbunches DR_length km Damping_time msec Nturns_betwe_coll Collision freq MHz Lsingleturn 1 e 36 Lmultiturn 1 e 36 Round single pass 0. 9 0. 0 900 0. 55 0. 8 1. 0 e-3 0. 7 e-3 1. 5 0. 8 Optional 0. 8 1. 0 e-3 7. 0 12000 6. 0 20 50 12. 0 1. 5 1. 1 Flat single pass Flat Ring with compressor 30 (2 betatron) 18 (2 betatron) +-1. 5 +-5. 0 12. 6 63. 2 5. 0 0. 080 0. 500 0. 100 0. 600 2. 0 e-2 3. 5 e-3 1. 0 e-3 0. 8 0. 002 0. 008 2. 0 Optional >2*12 4. 0 0. 5 e-3 7. 0 2. 0 12000 6. 0 20 20 50 1 12. 0 600 1. 5 1. 7 1. 1 0. 9 Flat Ring no compressor 2. 67 0. 0 12. 6 8. 9 0. 080 4. 0 1. 0 e-3 0. 7 e-3 0. 8 0. 002 4. 0 2*25 4. 0 1. 0 e-3 2. 0 12000 6. 0 20 1 600 1. 3 0. 9 P. Raimondi March 37 2006
Conclusions -There has been rapid progress in the optimization of the Super-B Collider parameters and layout --Workable parameter set contains: - ILC damping ring, - ILC bunch compressor, - ILC Final Focus - The optimal collision rate seems to be every turn. 38
Conclusions (2) for large crossing angle collider -Solution with a ILC DR + ILC FF seems very promising: - Crossing angle of about 25 mrad allows low vertical beta function with no bunch compression - Thus: No compressor and no energy acceleration - Uses all the work done for ILC - Ring and FF layouts similar to ILC - 6 km circumference rings - Strong synergy with ILC - Beam stay clear about 20 sigma assuming 1 cm radius beam pipe - Beam currents about 2 Amps - Power around 70 MW, further optimization may be possible. - Reuse PEP RF system, power supplies, vacuum pumps, etc. - Standard injection system. 39
- Slides: 39