Slow and Fast Resonant Extraction Schemes for RHIC

Slow and Fast Resonant Extraction Schemes for RHIC Kevin Brown BNL

Outline • • Oct. 31, 2010 Methods for resonant extraction Parameters Possible Locations for equipment Projected Costs Possible timelines to realize new systems Possible beam studies to test ideas Summary Drell-Yang Workshop, 2010 2

Methods for Resonant Extraction Two possible approaches 1. Traditional 2/3 integer slow extraction • • Beam needs to remain bunched (for polarization data) Beam is very rigid; 250 Ge. V Require very high efficiency (< 1% beam loss) Spill structure requirements? 2. Resonant island fast extraction • • • Oct. 31, 2010 Method developed at CERN PS Drive (adiabatically) particles into 3 transverse phase space islands using nonlinear elements Can either kick all beam out in 3 turns Or can single kick each bunch out over some time period Either way, kicker systems are very challenging to build Drell-Yang Workshop, 2010 3

Parameters • Energy = 250 Ge. V • Qx = 28. 67 (86/3), Qy = 29. 685, chroms ~+2/Q • Species = polarized protons – Beam needs to remain bunched to retain polarization – (bunch pattern set by STAR/PHENIX) • Intensity ~ 2 e 11/bunch, 111 bunches/rev (max) – Total ~ 22 e 12 protons/cycle • Spill at end of RHIC store, one spill per 10 hours – About 200 spills/yr = 4. 4 e 15 protons/yr – With ~1% loss, about 4. 4 e 13, 250 Ge. V proton loss/yr (slow extraction case, fast extraction should have lower losses) [for comparison: AGS, 1. 5 e 20 p/yr, losing 2%= 3 e 18 25 Ge. V p/yr] Oct. 31, 2010 Drell-Yang Workshop, 2010 4

Possible Locations for equipment • Slow Extraction – Need long electrostatic septa (2 or 3 modules) [ES] (at 1 cm gap, 100 k. V/cm gradient, need 10 meters of septa) – Need 4 or more magnetic septa [MS] – All/most beam loss is on ES – Locate ES between abort kicker and dump (10 O’clock) – Locate MS in 2 O’clock region • Fast resonant island extraction – Need multiple c-magnet type kickers – Require MS to separate kicked beams • Both systems will require • bumps to move beam into proper locations for extracting beam • un-squeeze b* (don’t want beam hitting in triplets) Oct. 31, 2010 Drell-Yang Workshop, 2010 5

RF Jet/Polarimeter STAR Dumps Oct. 31, 2010 Drell-Yang Workshop, 2010 PHENIX 6

~30 m IP 02 Oct. 31, 2010 Drell-Yang Workshop, 2010 7

Oct. 31, 2010 Drell-Yang Workshop, 2010 8

2 O’clock Interaction Region Extraction Septa need to fit here. Blue Ring on Outside. Blue Ring on Inside. Technical problem: Septa bend out from inner ring. Oct. 31, 2010 Drell-Yang Workshop, 2010 9

~ 1 m Oct. 31, 2010 Drell-Yang Workshop, 2010 10

~ 1° Oct. 31, 2010 Drell-Yang Workshop, 2010 11

Oct. 31, 2010 Drell-Yang Workshop, 2010 12

10 O’clock Region for ES or Kickers (? ) Oct. 31, 2010 Drell-Yang Workshop, 2010 13

Qy=29+ Qx=28 2/3 Qx = 28+ Oct. 31, 2010 Drell-Yang Workshop, 2010 14

Resonant Extraction: Analytical Picture 1 3 ES 2 Separatrix just after 10 o’clock dump kickers where a possible ES could be located. Oct. 31, 2010 Drell-Yang Workshop, 2010 15

Resonant Extraction: Analytical Picture 1 3 TMS 2 Separatrix just after Q 4 in 2 nd Sextant (start of 2 o’clock IR) where a possible Thin MS (TMS) could be located. Oct. 31, 2010 Drell-Yang Workshop, 2010 16

x’ [rad] Resonant Extraction Phase Space at 10 o’clock region: Particle Tracking x [m] Oct. 31, 2010 Drell-Yang Workshop, 2010 17

Devices for JPARC Slow Extraction (30 -50 Ge. V/c) 2 2 x 5? 7 For 250 Ge. V, mostly need long ES/TMS ES ~ 10 m TMS ~ 15 m M. Tomizawa et al. , submitted to PRST-AB, 2010. Oct. 31, 2010 Drell-Yang Workshop, 2010 18

Septa devices for JPARC Slow Extraction M. Tomizawa et al. , submitted to PRST-AB, 2010. Oct. 31, 2010 Drell-Yang Workshop, 2010 19

Extraction Septa in the CERN SPS for Slow Extraction Systems Oct. 31, 2010 Drell-Yang Workshop, 2010 20

CERN PS Electrostatic Septa Adjustable gap. Circulating beam J. Borburgh et al. , Proceedings of the 2003 Particle Accelerator Conference Oct. 31, 2010 Drell-Yang Workshop, 2010 21

SPS Septa for SPS to LHC Extraction (450 Ge. V/c) J. Borburgh, et al. , LHC Project Report 911 Oct. 31, 2010 Drell-Yang Workshop, 2010 22

Resonant island fast extraction • Same basic configuration as Slow Extraction – Replace ES with kickers – Still use thin MS and thick MS at 2 o’clock • Kickers options – Abort style = in 1 to 3 turns extract all beam – Single bunch style = 100 nsec rise/fall time, c -magnet needs to be ~ 5 to 10 m long (in sections) – Controlling orbit in long kicker is a challenge (sections may need to be movable) Oct. 31, 2010 Drell-Yang Workshop, 2010 23

Resonance Islands at 2/3 Integer Start with 111 single bunches End with 333 single bunches R Cappi & M. Giovannozzi, PRSTAB Vol. 7, 024001 (2004) Oct. 31, 2010 Drell-Yang Workshop, 2010 24

Projected Costs (educated guessing) • • Expect ~$0. 25 M/m for ES, Need ~10 m = $2. 5 M Expect ~$0. 25 M/m for Thin MS, Need ~5 m = $1. 25 M Expect ~$0. 2 M/m for Thick MS, Need ~10 m = $2. 0 M PS’s for septa? – ES ~$0. 15/modules, 4 modules = $0. 6 M – Thin MS ~$0. 5 M – Thick MS ~$0. 25 M/module, 5 modules(? ) = $1. 25 M • Sext. PS’s, Bump Magnets + PS’s = $1. 5 M Total: ~$10 M + (shielding, instrumentation, controls, etc. ) Costs for Fast Resonant System are similar. Oct. 31, 2010 Drell-Yang Workshop, 2010 25

Possible timelines to realize new systems • Define detailed specifications = 3 to 6 months – Mostly physicists time – Some initial engineering calculations • Work out detailed designs = 1 to 1 ½ years – – Assume 1 ME for each septum type (ES, TMS, MS) Assume spec’s change some as more detailed calculations are done Assume full funding and mostly FT for engineers (>50%) Assume time needed for prototyping and engineering testing • Build units = 2 years – – Many dependencies, since so many units are needed. Need full funding for entire time Assume BNL shops are readily available for full time. Given large number septa needed, 2 years may be aggressive. • Field mapping, installation, and commissioning = 1 ½ years – Field mapping overlaps with building units, so not consecutive Total est. time = 4 to 5 years (consistent with other projects - JPARC, SPS, etc) Oct. 31, 2010 Drell-Yang Workshop, 2010 26

Possible beam studies to test ideas • Study of resonance islands – study 2/3 int. resonance conditions • Moving particles into islands • Phase space reconstruction (if we have instrumentation) • Tests to compare to simulations – Could inquire if CERN would be interested in collaborating (we all know each other …) – May require ps’s for sextupole configurations Oct. 31, 2010 Drell-Yang Workshop, 2010 27

Summary • Both fast and slow resonant extraction are possible in RHIC. Standard systems, but push design limits. • For slow extraction two significant issues – Design goal would be to keep <1% loss – Could be challenging to keep track of polarization pattern during a beam spill (what is polarization of first microbunch extracted? Abort can gap be used as a reference. ) • For fast extraction there is always some beam left over – abort kickers can dump it • Both systems are expensive and take time to design and build. • Some beam experiments should be done to test the models and understand resonance conditions. Oct. 31, 2010 Drell-Yang Workshop, 2010 28