130 likes | 146 Views
Why and how to prepare for 900 ms basic period operation in the near future. Michael Benedikt AB/OP. Outline. Introduction Motivation, basic choices, impact Expected performance Overview Other aspects Implementation Summary and conclusions. Motivation.
E N D
Why and how to prepare for 900 ms basic period operation in the near future Michael Benedikt AB/OP PS & SPS Days 2005
Outline • Introduction • Motivation, basic choices, impact • Expected performance • Overview • Other aspects • Implementation • Summary and conclusions PS & SPS Days 2005
Motivation • Existing commitments (from HIP WG results) • Important shortfall of SPS cycles for physics (CNGS and FT). • ISOLDE performance 5 to 10% below request in CNGS era. • Upgrade plans of CNGS (>factor 1.5) and ISOLDE (up to factor 5). • Proposed improvement plan for SPS physics: • Significant increase of CNGS intensity per SPS pulse: • Fulfil (nominal) CNGS in shorter period and gain time for FT. • Upgrade of CNGS without (further) reduction of cycles for FT. • Consequences of higher intensity for SPS: • Requires “double batch” filling PSB to PS i.e. more PSB cycles: • Needs ~ 4 × 105 more PSB cycles, brings ISOLDE ~ 15% below request. • Needs ~ 2 × 106 more PSB cycles, brings ISOLDE ~ 35% below request. • Potential solution is an increase of PSB repetition rate PS & SPS Days 2005
Basic choices and consequences • Why 900 ms basic period and not…….e.g.1100 ms or 600 ms: • Aim was to look for a short-term, low-cost upgrade that nevertheless gives a significant impact. • Upgrade has to be largely compatible with existing hardware and control system (no change of philosophy). • Fully exploit the potential from the “PS for LHC” upgrade. • Basic period must be compatible with PS cycles to distribute the gain. • 900 ms instead of 1200 ms means: • 33% increase in PSB cycles. • All cycles on all machines have to be integer multiples of 900 ms. • Impact of 900 ms basic period operation: • Enables improvement for SPS without jeopardizing ISOLDE physics. • Improvements for LHC and PS East hall. • Generally shorter cycles on PS -> more efficient use, more time for MD. • Basis for CNGS and ISOLDE upgrades. PS & SPS Days 2005
Impact on machine operation • Linac2: • Pulsing at 900 ms (machine was designed for 2 Hz operation). • PS Booster: • New magnetic cycle for 900 ms operation (identical for all beams). • Nominal performance for high intensity operation was obtained in MD. • PS: • Choice of 900 ms allows to shorten most PS cycles by simply cutting-off unused parts at the end of cycles. • Consequently there is no change of beam dynamics for most beams except SFTPRO for FT and CNGS (acceleration on h=8 only -> to be demonstrated). • SPS: • Time between multi-batch injections reduced – shorter injection flat. • Remaining part of cycles unchanged (ramp up – down). • LHC: • Time between injections reduced – shorter injection flat. • AD, Linac3, LEIR: no significant impact. PS & SPS Days 2005
Expected performance (SPS users) • LHC nominal 25 ns and 75 ns beams: shorter LHC filling • PS: cycle 2.7 s instead of 3.6 s. • SPS: flat bottom 8.1 s instead of 10.8 s, cycle 18.9 instead of 21.6 s. • LHC: filling time reduced by 12.5%. • LHC “individual bunch” beams (TOTEM, pilot): neutral. • PS: cycle 1.8 s instead of 2.4 s. • CNGS: 5% flux reduction (single batch) or neutral (double batch) • PS: cycle 0.9 s instead of 1.2 s,compatible with Island extraction,acceleration on h=8 only, to be demonstrated for highest intensities in MD studies, i.e. >2.8E13. • SPS: flat bottom 0.9 s instead of 1.2 s, overall cycle length 6.3 sinstead of 6.0 s for PSB single batch operation, overall cycle length unchanged at 7.2 s for PSB double batch operation. • SPS Fixed Target: neutral, unchanged duty cycle (spill/cycle). • PS: cycle 0.9 s instead of 1.2 s, acceleration on h=8 only, see CNGS. • SPS: flat bottom 0.9 s instead of 1.2 s, remaining cycle unchanged. PS & SPS Days 2005
Expected performance (PS and PSB users) • PS East hall beams: 20% increased duty cycle (spill/cycle) • Cycle 2.7 s instead of 2.4 s. Spill length ~550 ms instead of ~400 ms. • Net flux gain ~20%. • n_TOF: three scenarios with different proton energies • Cycle 1.8 s instead of 1.2 s. Beam energy 20 GeV/c (up to 24 GeV/c). • Flux reduction 33% (20 GeV/c) 27% (24 GeV/c) – can be compensated. • Cycle 0.9 s instead of 1.2 s. Beam energy change to 15.3 GeV/c from 20 GeV/c. • Flux increase by 6% (15.3 GeV/c). • 15.3 GeV/c solution is preferred to 24 GeV/c by n_TOF. • AD – neutral • PS cycle 1.8 s instead of 2.4 s. • AD cycle remains unchanged. • ISOLDE beams – 33% flux increase • PSB cycle 0.9 s instead of 1.2 s. • Nominal high intensity achieved in MD. PS & SPS Days 2005
Other aspects • Reliability and lifetime: • Equipment lifetime is usually proportional to the number of pulses and will therefore decrease like the pulse rate increases (≤ 75% of present). • This lifetime decrease concerns PSB and PS but not SPS. • Exploitation budget (P+M) has to take this into account. • Radiation aspects PSB: • 33% more cycles does not mean 33% increase in losses. • Losses in PSB at increase more than proportional when pushing the intensity. The last 10 % in peak intensity nearly double the losses at high energy (extraction and recombination). • Double batch CNGS will use lower PSB bunch intensity (smaller emittances) which will help PSB, PS, SPS. • For highest intensity ISOLDE operation: the larger number of available cycles will allow to reduce the peak intensity. PS & SPS Days 2005
Implementation strategy • 2005 – Linac2 and PSB (approved by ABMB on 06/12/04): • Start and run with 1200 ms basic period - prepare 900 ms in parallel: • Adjust machine timing layout to be compatible to 900 ms. • Perform all relevant tests (DSC software, etc) and modifications. • Prepare all operational beams on 900 ms compatible cycles. • Switch to 900 ms basic period and test-run for few weeks. • Duration of 4-5 weeks (10 days dedicated) (proposed mid May – end June). • Sufficiently long time to detect long-term problems. • Verify all operational beams under full 900 ms conditions. • Switch back to 1200 ms operation for normal run and the start-up 2006. • Minimum risk strategy. • Sufficient time for preparation (no unnecessary “crash programme”). • Results from 900 ms run will facilitate preparations for PS/SPS 2006. PS & SPS Days 2005
Implementation strategy • 2006 – PS and SPS (to be reviewed after PSB test in 2005): • Start and run with 1200 ms basic period - prepare 900 ms in parallel: • Adjust PS and SPS timing layouts to be compatible to 900 ms. • Perform all relevant tests (DSC software, etc) and modifications. • Perform MDs to address all open issues (high intensity h=8 in PS, n_TOF). • Prepare all operational beams on 900 ms compatible cycles. • Switch all machines to 900 ms basic period during dedicated MD. • 2 weeks of dedicated MD time for commissioning of 900 ms operation. • Verify all operational beams under full 900 ms conditions. • Minimum risk strategy. • Avoids mixing up with eventual problems during PS/SPS restart after long shut-down and renovation of 40 PS magnets with complete realignment. • Enables MDs on PS on h=8 high intensity before switching to 900 ms. • Sufficient time for preparation while PS/SPS are running. PS & SPS Days 2005
Summary PS & SPS Days 2005
Conclusions • 900 ms basic period operation: • 33% more proton cycles from PSB. • Basis for improvement on SPS and upgrades of CNGS and ISOLDE. • Overall cost: ~1.1 MCHF, ~10 man-years. • Can be fully implemented at end of run 2006. • Next steps: • Start preparations for Linac2 and PSB run in 2005. • Ensure compatibility of new equipment and software with 900 ms. • Specifications of power converter consolidation. • Machine timings Linac3 and LEIR, SPS control and timing changes. PS & SPS Days 2005
Acknowledgements Special thanks for the support and help of: S. Baird, JC. Bau, J. Borburgh, JP. Burnet, C. Carli, E. Carlier, P. Cennini, F. Di Maio, K. Elsener, A. Fabich, T. Fowler, R. Garoby, M. Giovannozzi, M. Haase, W. Heinze, M. Hourican, E. Jensen, R. Jones, S. Hancock, C. Hill, J. Lewis, A. Mengoni, G. Metral, M. O’Neil, U. Raich, M. Rettig, JP. Royer, K. Schindl, R. Scrivens, R. Steerenberg, JL. Sanchez-Alvarez, J. Wenninger, R. Wilfinger, T. Zickler. Many thanks to all colleagues in AB that have contributed to the study. PS & SPS Days 2005