270 likes | 380 Views
stacking simulations for CR & C-ERL schemes. Frank Zimmermann Thanks to: Robert Chehab , Masao Kuriki , Tsunehiko Omori, Louis Rinolfi , Junji Urakawa, Alessandro Variola , Alessandro Vivoli , Vitaly Yakimenko POSIPOL 2008, June 2008. various scenarios Compton sources
E N D
stacking simulationsfor CR & C-ERL schemes Frank Zimmermann Thanks to: Robert Chehab, Masao Kuriki, Tsunehiko Omori, Louis Rinolfi, Junji Urakawa, Alessandro Variola, Alessandro Vivoli , VitalyYakimenko POSIPOL 2008, June 2008
various scenarios • Compton sources • Compton ring – CR (“pulsed”), or • Compton ERL – CERL (“continuous”) • accumulation rings • ILC damping ring • CLIC pre-damping ring
similar assumptions as for Snowmass’05 • stacking in longitudinal phase space • ingredients: sinusoidal rf, momentum compaction, radiation damping and quantum excitation • injection septum placed at location with large dispersion; septum blade << transverse beam size • between successive injections orbit at the septum is varied with fast bumper magnets • energy of injected beam is ramped such that the transverse septum position is always separated by 2 sd0 from injected beam centroid
ILC-CR injection scheme • ILC 2008: inject every second turn (80 MHz) into the same bucket - 30 times; then wait 10 ms (~450 turns, ~1 damping time) and repeat 9 times; total injections/bucket: 300; synchrotron phase advance between two injections: 0.134 • simulation result (February’08) : 35% loss / cycle, total loss 76%!
improve ILC-CR stacking effiency • possible approaches – not (yet) used: • smaller momentum compaction factor • reduced synchrotron tune • larger interval between cycles • (but nominal already too long for vertical • damping – Vitaly Yakimenko; next slide)
ILC DR “trouble” (Vitaly Yakimenko) final gey,rms=0.017 micron initial gey,rms=6 mm minimum store time Tstore follows from 0.017*exp(2 Tstore/ty) mm=6000 mm → Tstore > 6.4 ty=164 ms leaving only 36 ms for stacking
improving ILC-CR stacking efficiency • methods chosen: • energy pre-compression [x3] (R. Chehab) • additional DR wigglers for faster damping [x2] • larger rf voltage [x 1.5] • → 2008 ILC DR Compton version
simulation for ILC-CR cycle 1, after 1st injection cycle 1, after 5th injection cycle 1, after 10th injection cycle 2, after 1st injection cycle 1, after 30th injection cycle 2, before 1st injection cycle 2, after 5th injection cycle 2, after 30th injection cycle 3, before 1st injection
cycle 10, after 30th injection 10 ms after cycle 10 110 ms after cycle 10 ~ 10.6% of injected e+ are lost! similar loss fraction for single cycle → stacking efficiency ~90% for ILC DR Compton version
questions & comments for ILC-CR • 2x3 km ring is option from Andy Wolski; it could reduce the damping times by factor 2, if we do not reduce the length of the wigglers • ring parameters can be considered somewhat flexible; at present parameters are optimized for the undulator based source • can we reduce initial energy spread to 2 MeV rms? • option of pre-damping ring for ILC?
ILC-CERL injection scheme - A • continuous stacking (ERL option),~27MHz (Omori san, • Variola san) 850 injections over 5100 turns • (inject every 6th turn), followed by 5155 turns • (~100 ms) damping; damping time 6.4 ms; • inject with constant offset d=1.2% <Vivoli san’s result: ~ 2.9 MeV • sz=9 mm,sd0=1x10-4 (0.5 MeV, small!!):63.7% loss Omori san asked about “unstable point” injection offset d=1.2% or 0.4%, z=0.1 m: 99.8% loss offset d=0.2%, z=-0.1 m: 99.9% loss offset d=0.5%, z=0.01 m: 72.8% loss offset d=0.7%, z=0.01 m: 50% loss!! Method works!? offset d=0.8%, z=0.01 m: 41.8% loss! offset d=0.9%, z=0.01 m: 36.7% loss! 63% efficient
ILC-CERL injection scheme - B over 6turns synchrotron phase advance ~0.5!? perhaps better inject every 5th turn?! • again continuous stacking (ERL option), ~32 MHz, • 1020 injections over 5100 turns • (inject every 5th turn), followed by 5155 turns • (~100 ms) damping; damping time 6.4 ms; • inject with constant offset d=0.9%, z=0.01 m • sz=9 mm,sd0=1x10-4 (0.5 MeV, small!!):36% loss offset d=1.0%, z=0.01 m: 33% loss offset d=1.2%, z=0.01 m: 27% loss! 73% efficient offset d=1.3%, z=0.01 m: 23% loss! 77% efficient offset d=1.4%, z=0.01 m: 16% loss! 84% efficient offset d=1.5%, z=0.01 m: 9% loss! 91% efficient!
simulation for ILC-CERL 10 turns, 3 bunches 20 turns, 5 bunches 100 turns, 20 bunches 5000 turns, 1000 bunches 1000 turns, 200 bunches 100 ms after last injection → stacking efficiency ~91% for ILC DR Compton version
CLIC beam parameters (updated ‘08 by Louis Rinolfi): • at IP: 3.72x109 ppb, 312 bunches, 0.5 ns spacing, • 156 ns (= ) train length, 50 Hz repetition rate • request 6x109 ppb e+to account for downstream loss • 6.2x1010 e-/bunch in 1.3 GeV Compton ring • → ~7x108 photons → 1.4x107e+ ? injected in pre-DR • accumulation over 460 turnsyields target e+ number; • could split this into 20x23 injections e.g. • minimum pre-DR circumference: 47 m • Qs is ~variable pre-DR parameter; • at exit of 2.2 GeV linac: σd=2.7%?, sz=5mm (A.Latina) • CLIC damping ring circumference: 365.2 m • 7 full trains fit into DR (7x more time for stacking; • total available time 7x20 ms=140 ms!)
CLIC-CR injection scheme • CLIC February 2008: inject every 40th turn (50 MHz) into the same bucket - 6 times; then wait 3.8 ms (~4900 turns, ~2 [now 4] damping times) and repeat 9 times; total injections/bucket: 60; synchrotron phase advance between two injections: 2.384 result coming soon
(4) CLIC-CERL also coming soon
next steps & ideas: • determine “optimum” CLIC pre-DR parameters • set up CLIC simulations once parameters clarified • optimize injection offsets z,dfor minimum loss • optimize synchrotron tune • energy pre-compressor for both ILC & CLIC • combined longitudinal/transverse stacking?
stacking is helped by: • short damping time • small energy spread (which value is possible?) • large ring momentum acceptance • sufficient store time • simulation results: • 90% for ILC-CR (300 inj’s, sd=2 MeV spread) • 91% for ILC-CERL (1020 inj’s! sd=0.5 MeV) • both for Compton version of ILC DR with energy pre-compression - Compton version of ILC-DR acceptable? #e+ / pulse for C-ERL scheme? • CLIC simulations: still work in progress