1. Experience and Projections for Low-Energy RHIC Operations Todd Satogata, Collider-Accelerator Department Workshop on Future Prospects in High Energy QCD July 21, 2006  Background, parameters, challenges  22 GeV proton test run preliminary findings  New issues: RHIC RF acceptance, AGS cooling  Performance projections and plans

2. July 21, 2006T. Satogata - Hot QCD Workshop2/19 Background and Challenges  There is substantial and growing interest in RHIC heavy ion collisions with c.m. energy in the range  s NN = 5-50 GeV/n  Corresponds to Au beams in RHIC of  =2.68 to 26.8  Nominal Au injection is  =10.52, already below design  =12.6  Krishna Rajagopal’s RIKEN workshop energies:  s NN = 5, 6.3, 7.6, 8.8, 12.3, 18, 28 GeV/n  Operational and strategic challenges  How does RHIC magnetic field quality degrade at lower currents and energies?  Can we fit low-energy beam in the RHIC RF system?  What are short- and long-term facility improvements?  Can we collide 32 Au to lower Z/A, and run with higher fields/currents? (Short answer: unfortunately not.)

3. July 21, 2006T. Satogata - Hot QCD Workshop3/19 Searching For The QCD Critical Point In The Press A. Cho, Science, 312 (14 Apr 2006)

4. July 21, 2006T. Satogata - Hot QCD Workshop4/19 RIKEN Workshop Accelerator Summary (April 2006)  No apparent show-stoppers for RHIC collisions at E cm = 5-50 GeV/n  Only equal energies  Unequal species possible only if minimum rigidity > 200 T-m  Without cooling  long vertex distribution  Small set of specific energies (and species?) should be a workshop deliverable for planning:  2.5,3.2,3.8,4.4… GeV/n total beam energy  Studies that should be done soon:  A ~1 day study period at low total beam energy to identify power supply, lifetime, tuning issues/limitations  Low-current superconducting magnet measurements  Pre-cooling in AGS  10+x luminosity ?  RHIC electron beam cooling would make this a fantastic facility: ~100x luminosity, small vertex distribution, long stores. Expected whole vertex minbias event rate [Hz] T. Roser, T. Satogata RHIC Heavy Ion Collisions

5. July 21, 2006T. Satogata - Hot QCD Workshop5/19 2001 9.8 GeV/u Au collisions  2 days of 9.8 GeV/u collisions  0.4  b -1 integrated luminosity   *=3m by necessity  60-90 minute stores  56 Au bunches, 0.6x10 9 /bunch  10-30 Hz ZDC rates  IBS and aperture dominated beam and luminosity lifetime  Another run at this energy may improve this by factor of 2-5  1.0-2.0x10 9 /bunch  Raise  * to improve lumi lifetime  RHIC is best used as a storage ring collider below beam energies of ~12 GeV/u

6. July 21, 2006T. Satogata - Hot QCD Workshop6/19 Initial Machine Projections ModeBeam Energy [GeV/u] N bunches Ions/bunch [  9 ]  * [m] Emittance [  m] L peak [cm -2 s -1 ] Au-Au 2001-29.8550.63158.0  10 24 Au-Au 2003-431.2451.0315-301.2  10 26 Au-Au9.8551.21015-401.0  10 25 Au-Au2.51100.31015-302.5  10 22 Au-Au25551.2315-402.0  10 26  Assumes expected luminosity scaling as  3 below 9.8 GeV/u   */aperture and integrated luminosity tradeoffs must be studied  Projections do not include potential improvements  Electron and stochastic cooling (peak and integrated luminosity)  Lattice modifications to mitigate IBS (integrated luminosity)  Total bunch intensity from vacuum improvements (peak luminosity)  Ions/bunch prediction for 2.5 GeV/u has gone down by x3-4!  ZDC rates naively on order of 0.1 Hz  RF acceptance between AGS/RHIC, discussed later in this talk

7. July 21, 2006T. Satogata - Hot QCD Workshop7/19 Low-Energy Magnetic Field Quality  Magnet currents scale with rigidity B  which scale with   Field quality deteriorates rapidly at very low currents  Currently have some magnet measurements at very low currents, enough for simulations  Extrapolating low-field behavior from simulations Total Energy  BB Dipole Current 9.8 GeV/u10.5281.11430 A 2.5 GeV/u2.6820.69110 A pp Low-Energy Test June 5-6 2006

8. July 21, 2006T. Satogata - Hot QCD Workshop8/19 Proton/Gold parameters tested at B  =37.4 T-m ProtonsGold (eq)  s [GeV]22.59.18 Beam energy [GeV]11.254.59 Beam kinetic energy [GeV]10.3123.660 Relativistic  11.994.93 Relativistic  0.9970.979 Momentum [GeV/c]11.2114.496 B  [T-m]37.40 Injection current scaling0.4710.384 Main dipole current [A]217.7 Main quad current [A]202.6 Revolution frequency [Hz]7792476571 RF frequency [MHz, h=360]28.05327.566  Gold energy (  s NN =9.2 GeV/n) is near middle of preferred energy list  Tested magnets and power supplies at less than half of normal injection current  Reasonably balance for initial testing, performance extrapolation  Did NOT require change to RHIC harmonic number or substantial reconfiguration of RHIC RF  Au h=366 gives RHIC RF frequency of 28.03 MHz RHIC RF frequency range is 28-28.17 MHz

9. July 21, 2006T. Satogata - Hot QCD Workshop9/19 Low-Energy Proton Test Run Results (June 5-6 2006)  First injection to circulating beam in about 30 minutes (both rings)  Circulating beam to RF capture in under 3 hours.  Single bunch intensities up to about 8x10 10 protons.  Injection efficiency was about 70-80%.  Beam lifetimes of 5-10h (not in collision) and 4-5h (in collision) achieved.  Chromaticities and instabilities can barely be controlled with existing power supplies at this energy. For this and lower energies, we should rewire unipolar sextupoles for future operations.  Proton emittances out of AGS were very good (<10  mm mrad normalized) Measured emittances in RHIC are being analyzed, but appear to be comparable as long as beams did not go unstable.  Optics measurements and correction data were taken in both rings, and will be analyzed to compare to tests. Initial analyses show that optics (beta functions, phases, dispersions) are surprisingly close to the model, so linear field quality is not limiting performance.  Two hours were spent performing vernier scans in PHENIX and STAR with beam-beam counters with no evidence of collision distributions.

10. July 21, 2006T. Satogata - Hot QCD Workshop10/19 Low-Energy Test Run Optics  Measured optics in both rings are surprisingly good  No indications of linear optics problems, quadrupole magnet or power supplies with aberrant low-current behavior Horizontal Dispersion [Blue, m] S-coordinate [m] Predicted difference orbitMeasured difference orbit

11. July 21, 2006T. Satogata - Hot QCD Workshop11/19 Low-Energy Test Run Beam Ripple  Measured beam ripple in both rings is also good  No indications of major coherent ripple from power supplies  Coherent beam oscillation spectrum comparable to injection at nominal energies  Should still dwell AGS Booster during low-energy stores 60 Hz 120 Hz 360 Hz 10 Hz Beam Position [mm] Sampled turn number FFT Frequency [x100 Hz] FFT Amplitude

12. July 21, 2006T. Satogata - Hot QCD Workshop12/19 Low-Energy Test Run Instabilities  Beam stability was a problem through the low-energy test  Power supplies did not permit necessary chromaticity control  Yellow beam was metastable at best  Some instabilities damped with strong octupoles in both rings  Sextupole reconfiguration for low-energy ops takes ~1 day Unstable blue beam “Metastable” yellow beam Stability monitor

13. July 21, 2006T. Satogata - Hot QCD Workshop13/19 Low-Energy Test Run Vernier Scans (STAR)  Collision steering attempted at STAR and PHENIX  Mostly background, likely driven by unstable (large) yellow beam  Some apparent real rates in STAR, though very low, many backgrounds  Rates consistent with luminosity down by factor 2500 from 100 GeV (1.7x10 28 )  Expected reduction from beam parameters is closer to 4000 Could not justify one-day 22 GeV pp run this year from this data Luminosity monitoring will be a challenge for low-energy program Beam-beam tune shift may be one of our best luminosity observables BBC Rates [Hz] Horizontal bump position [mm] Vertical bump position [mm] BBC Rates [Hz]

14. July 21, 2006T. Satogata - Hot QCD Workshop14/19 RF Issues for Low-Energy Ion Beams  Impacts and solutions:  Eliminate h=4 rebunch in AGS. Inject 12 smaller bunches per cycle Faster refill times, but intensity down by factor 3-4  Bunch lengths look acceptable for kickers, momentum spread 10 -3 ok  Discussion about alternate frequency RF in RHIC Requires breaking vacuum, retrofit of storage cavities  At low energy, RHIC bunches are much larger than RF bucket  Bunches: 0.3-0.5 eV-s  Bucket area: ~0.08 eV-s  Must make either  bunches smaller (cooling or lower intensity)  buckets bigger (changes to RHIC RF system) RF Bucket Bunch

15. July 21, 2006T. Satogata - Hot QCD Workshop15/19 Cooling in RHIC Injectors: AGS  1 GeV/u Au injection energy electron cooling in AGS  500 keV electron beam required for cooling  Can be built by Novosibirsk cooler developers Have recently delivered similar cooler to CSR in China  Surplused coolers from CELSIUS/IUCF will not work  AGS has only ~2.5 m straight section available for cooler  Simulations currently underway to predict cooling rate, cost  Advantages and disadvantages:  At least factor 3-5 in bunch intensities  Improved emittances, RF capture, backgrounds  Still must contend with IBS in RHIC Best mitigated by RHIC electron cooling, lattice changes  Modest costs, multi-year lead time

16. July 21, 2006T. Satogata - Hot QCD Workshop16/19 IBS and Electron Cooling at 2.5 GeV/u in RHIC  IBS growth times for low-energy gold ions are short, o(4 minutes)  Correspondingly short luminosity lifetimes, need to refill often  Electron cooling with a modest electron beam can help  Electron beam from simulation: q=5nC, RMS  e,n =10  m,  p =1e-3, L=10m  Cooling times of a few seconds, dramatically improves emittances and luminosity lifetime  BUT: Cannot be done with high-energy RHIC electron cooler  RHIC lattice for IBS suppression is also being investigated emittance growth (95% normalized) due to IBS:  i,n,95% : 15  -> 48  in 30 min Cooling of transverse emittance (95% normalized):  i,n,95% : 15  -> 4  in 10 sec A. Fedotov

17. July 21, 2006T. Satogata - Hot QCD Workshop17/19 Other Cooling at RHIC Low Energies  Stochastic cooling is feasible but requires development  High energy stochastic cooling achieved in 2006 with protons Hardware for blue ring being installed for 2007 run  BUT: Different mixing regime than high energy stochastic cooling, cannot use filter cooling, as Schottky bands overlap  Higher cooling rates achievable at lower energies, but requires further study to give cooling time predictions  Palmer cooling  Under active development at C-AD as frontier of cooling Testing cutting chord from pickup to kicker in Run 7  Requires new cold pickup in arc (high dispersion, low beta)  Concerns about 10 Hz coherent signal rejection M. Brennan

18. July 21, 2006T. Satogata - Hot QCD Workshop18/19 Low-Energy Testing in 2007  IBS will make ion lifetimes worse than protons (A. Fedotov)  Consistent with 15- to 30-minute stores at 5 GeV/u  Lower energies likely require cooling or lattice modifications  Recommend 1 week of low-energy Au setup and testing  Between maintenance days to swap power supply polarities  RF synchro setup will also require few shifts of beam  CERN data suggests  s NN =7.6 GeV/u; leverage 9.2 GeV/u done  Gain experience with RF, IBS, low luminosity, fast refills  Produces definitive luminosity for performance extrapolations  Tests at multiple energies? One day at  s NN =5 GeV/u? Beam Energy [GeV/u]IBS growth time [s] 2.5250 51800 95000 (horizontal)

19. July 21, 2006T. Satogata - Hot QCD Workshop19/19 SummarySummary  There is substantial and growing interest in RHIC operations at low energies (  s NN = 5-50 AGeV)  Discovery of the QCD critical point would be another major discovery at RHIC  There are no show-stoppers to a low-energy program  Low-energy (B  =37.4 T-m) test run occurred June 5-6 2006  Addressable issues: sextupoles, instabilities, lumi monitoring  BUT RF challenges presently indicate <1Hz interaction rates  Future planning  ~1 week at low energy in 2007 with Au is natural next step  Recommended energies:  s NN =7.6 GeV/u, test at  s NN =5 GeV/u  Low-energy electron cooling in both AGS and RHIC look very beneficial in multi-year timescale, but require program support