1. CERN/GSI beam dynamics and collective effects collaboration meeting 18 th February 2009 High Intensity Operation of the CERN PS Booster (PSB) Christian Carli presenting results of many people after operating and improving the PSB since years Introduction PSB within the CERN Complex Overview of the PSB High intensity operation of the PSB Dynamic working point Performance with high intensity beams Double harmonic RF system Plans and studies for the PSB with Linac4 Motivation for Linac4 and PSB Beam Dynamics Studies Longitudinal Painting Lattice perturbations due to the chicane Benchmark simulations versus observations Simulations of injection Summary

2. High Intensity Operation of the PSB 18th February 2009 Introduction - PSB within the CERN Complex First synchrotron in the CERN accelerator chain (..in fact four “independent” machines) Receives protons from Linac2 (no ions any more from Linac3):  Up to 180 mA   RMS ~2.5  m Delivers beam to:  PS  ISOLDE  Dump LHC

3. High Intensity Operation of the PSB 18th February 2009 Introduction – CERN PS Booster (PSB) Overview Recapitulation of the present Booster with Linac2:  Constructed in the 70ies to improve the performance (intensity) of the PS  Four superimposed rings (chosen instead of fast cycling synchrotron or 200 MeV Linac) with one forth of the PS circumference Four in one design dipole and main quadrupole magnets Additional windings for individual ring-to-ring adjustments Operational complications: - “distribution” system to bring the Linac4 beam into four rings and - “recombination” of bunches from four rings  16 cells with triplet focusing, phase advance >90 o per period  Large acceptances: 180   m and 120   m in the horizontal and vertical plane, respectively  Multiturn injection with betatron stacking and septum at 50 MeV  Acceleration to 1400 MeV (was 800 MeV at the beginning) Sketch of the PS Booster with: - Distribution of Linac beam into 4 rings - Recombination prior to transfer

4. High Intensity Operation of the PSB 18th February 2009 Introduction – CERN PS Booster (PSB) Overview

5. High Intensity Operation of the PSB 18th February 2009 High Intensity Operation in the PSB - Dynamic Working Point Dynamic working point (mandatory for high Intensity and brightness):  Make space in resonance diagram for beam with large tune (direct space charge) spreads  High Intensity: Maximum direct space charge tune shift  Q V > 0.5 (tricky to estimate … form factors, emittances after injection..) Injection with Q H ~4.28 and Q V ~4.55 (5.55 some years ago) Lowered within ~100 ms Ejection with Q H ~4.17 and Q V ~4.23  High Brightness LHC type beams: Maximum direct space charge tune shift  Q V > 0.3 Injection with Q H ~4.28 and Q V close to half-integer resonance4.55  Resonance compensation (3 rd order resonances and 2Q V = 9) mandatory to reduce losses Dynamic Working point with lower vertical tune – no syst. 3 rd order resonance (taken from a presentation by M. Chanel)

6. High Intensity Operation of the PSB 18th February 2009 High Intensity Operation in the PSB – Performance with high intensity beams High intensity beams obtained with the PSB (taken from a presentation by M. Chanel) >4.1 10 13 >10 13 /Ring!

7. High Intensity Operation of the PSB 18th February 2009 High Intensity Operation in the PSB – Double harmonic RF Mandatory for high Intensity and Brightness:  Principle harmonic (now in LHC era) h=1 with ~8 kV  Second harmonic h=2 with ~8 kV as well Injection of coasting beam and (not that) adiabatic capture  Injection onto slow ramp …. is compromise:  Sufficient time for capture  Beam during shorter duration with large direct space charge tune shifts Longitudinal phase reconstruction of a beam with a double harmonic RF system (taken from a presentation by M. Chanel)

8. High Intensity Operation of the PSB 18th February 2009 High Intensity Operation in the PSB – various Transverse damper needed only in the horizontal plane (?):  Mandatory for to stabilize high intensity beams in the horizontal plane Beam stable after injection without damper, but becomes unstable during acceleration  Horizontal instabilities (“kind of head-tail”??) observed (G.Rumolo, D.Quatraro ….) in 2008 in ring 4: Short rise times (short compared to synchrotron period) at well defined times in cycle Excited by narrow impedance … ejection kickers? Transverse (horizontal) profiles  Without space charge: dense peak at the center and tails due to multiturn injection  In practice: beam takes on profile yielding small “Laslett tune shift” Versatile machine  Delivers not only “demanding” high intensity and brightness beams  Many different beam with very different characteristics (intensities varying by more three) produced in from one pulse to the next Conversion to h=1 for flexibility of transfer to PS within preparation for LHC  LHC type beams: h=1 and double batch transfer  High intensity: h=2 and single batch transfer (four PSB rings ) PSB to PS Transfer Energy:  Increased (in two steps) from 800 MeV to 1400 MeV to reduce space charge effects in the PS  Direct Space Charge at PSB Injection a Bottleneck

9. High Intensity Operation of the PSB 18th February 2009 Plans for the PSB with Linac4 – Motivation for Linac4 and PSB Beam Dynamics Studies Rationale for Linac4:  PS Booster limited by direct space charge effects, quantified by direct space charge tune shift ∆Q, at injection energy  Increase injection energy such that the brightness can doubled (double intensity within unchanged normalized emittances) without changing ∆Q  Raise injection energy from 50 MeV with Linac2 to 160 MeV with Linac4 Convert to H - charge exchange injection  Reach the phase space density expected in the PSB  Allows - with a Linac4 Beam chopper - for longitudinal Painting  New injection hardware for increased injection energy and charge exchange injection Aim of PSB Beam Dynamics with Linac4 studies  Answer the question: Is above argument the whole story? (e.g. the Beam will stay a bit longer with large ∆Q in the PSB)  Thorough understanding of the implications of Linac4 on the PSB  Make sure that PSB performance expected with Linac4 is reached fast:  Anticipate possible additional limitations & difficulties and devise for cures  Ensure integration of PSB with Linac4 into the complex (versatility ….) Next bottleneck around the corner: PS and SPS at low energy

10. High Intensity Operation of the PSB 18th February 2009 Plans for the PSB with Linac4 – Active longitudinal Painting With Linac4: similar RF system than at present  Double harmonic fundamental h=1 and h=2 systems to flatten bunches reduces maximum tune shifts  Injection with d(B  )/dt = 10 Tm/s (no need for injection with small ramp rate with painting any more)  Little (but not negligible) motion in longitudinal phase space.  No way for painting from synchrotron motion (large harmonic numbers and RF voltages ruled out)  Need for active painting (aim: fill bucket homogeneously) and energy modulation

11. High Intensity Operation of the PSB 18th February 2009 Plans and studies for the PSB with Linac4 – Active longitudinal Painting Principle:  Triangular energy modulation (slow, ~20 turns for LHC)  Beam on/off if mean energy inside a contur ~80% of acceptance  Nominal LHC: intensity with 41mA (!!!) after 20 turns  High intensity: several and/or longer modulation periods Potential limitations: Linac4 jitter, debunching of Linac4 structure in Booster Note that D ~ 1.4 m at the PSB injection Ideas on implementation under discussion

12. C. Carli 12/13 11 th Deceember 2008 Plans and studies for the PSB with Linac4 - Lattice Perturbations by the injection chicane Problem (thanks to F. Ostiguy for drawing our attention on limitations at FNAL)  Chicane (to merge injected with circulating beam) dipoles add focusing initially envisaged: rectangular bends without gradients  Lattice perturbation (strong in vertical plane … tune close to resonance)  Long chicane dipoles and small deflections to reduce perturbations “Passive” compensation (M. Aiba, TE/ABT team working on hardware)  Replacing rectangular bends by sector bends or adding gradients brings (part of) the perturbation into horizontal plane  Less horizontal beta-beating, because tune is further away from half-integer resonance  Pole-face windings on chicane magnets for “active” compensation? Linac4 Work Package 3.2 “PSB Beam Dynamics” 18th February 2009 High Intensity Operation of the PSB Example: long chicane dipoles & small deflections max. deflection of chicane dipoles: 66 mrad, pole face rotations: 33 mrad

13. Plans and studies for the PSB with Linac4 – Lattice Perturbations by the injection chicane “Active” compensation with trims on appropriate QDE quads:  Proposal (M. Aiba, B. Goddard..) for additional quads ruled out initially: complicated scheme with similar  -function in both planes for quads in straights  Trims on QDE quadrupoles are appropriate (chicane made of rectangular bends): Large (  -function and) compensation in vertical plane with little (  -function and) perturbation in the horizontal plane In practice: trims on appropriate “Q-strips” (additional windings of PSB quads) Appropriate quads for “lower vertical tune: QDE03 and QDE14 Compensation during (no compromise) whole chicane fall Preliminary discussion on Power Converters: feasible provided chicane fall time is not too short (at least a few ms) High Intensity Operation of the PSB 18th February 2009 β-beating with (dashed) and without (solid) compensation Blue: vertical; red: horizontal

14. Plans and studies for the PSB with Linac4 – Benchmark simulations versus measurements Assessment of slow blow-up by Simulations feasible?:  Benchmark comparing measurements (by M. Chanel) with simulations (by M. Aiba and M. Martini) and between ACCSIM and ORBIT (M. Aiba and M. Martini)  For the moment, no satisfactory agreement (yet?)  Example: 2 nd harmonic RF to shorten bunches to increase tune shift ACCSIM largely overestimates blow-up ORBIT overestimates blow-up as well Dependance on number of macro-part’s Numerical artefacts?  Successful benchmark is prerequisite to apply simulations for slow blow-up and loss estimates  Simulations in any case useful for injection and a few ms afterwards High Intensity Operation of the PSB 18th February 2009  n H =19.2  m  n V =7.1  m  n H =20.4  m  n V =7.3  m ACCSIM ORBIT

15. Plans and studies for the PSB with Linac4 – Simulations of Injection Impact of injection matching and painting on PSB performance Studied: Impact of dispersion mismatch  Filamentation of structure due to injection dispersion mismatch?  Investigated by M. Aiba with input from B. Goddard – longitudinal and transverse painting of an LHC type injected over 20 turns  No significant difference with and without dispersion mismatch  To do: Impact (minor ?) of transverse matching and painting on performance? Started: optimization of painting process (betatron mismatch, bump wave forms) High Intensity Operation of the PSB 18th February 2009 Evolution of rms emittances after injection - blow-up (extrapolation ?) may be too large (slightly larger relative blow-up for 99% and 100% emittances)

16. Plans and studies for the PSB with Linac4 - Simulations of injection with perturbation by chicane Evolution of beam parameters during fall of chicane with lattice perturbations and passive compensations and different fall times (simulations by M. Aiba)  “Passive” compensation (see above) by pole-face rotation of chicane magnets yield  Simulations with ORBIT and additional routine for time dependant focusing of chicane, but no synchrotron oscillations  Simulations indicate significant reduction of blow-up Chicane: L BS = 0.25 m and sector bends Chicane: L BS = 0.25 m and “tapered” bends Lattice perturbation and “active” (by trims on QDE’s) compensation  Injection painting, “static chicane and compensation” and with synchrotron motion  Similar results (compensation reduces blow-up) High Intensity Operation of the PSB 18th February 2009

17. Summary PSB is very likely limited by direct space charge  Cures to alleviate direct space charge effects are efficient Dynamic working point, Double harmonic RF Transverse profile yielding (beam takes on this shape without any particular measure) “small” direct space charge tune shifts Linac4 aims at increasing the brightness possible in the PSB by a factor 2  Increase of injection energy from 50 MeV to 160 MeV  Simultaneously conversion to H - and availability of Linac4 chopper  Intensity increase within unchanged emittances by a factor two for LHC beams (and less than a factor two for high intensity)  Plans and studies to better understand implication of Linac4 on the PSB Active longitudinal painting scheme Simulations take an important place to optimize injection Feasibility of simulations for slow blow-up unclear Optics perturbations due to chicane are an important limitation Collimation, transfer of LHC beams to the PS, instabilities, beam loading in cavities, … High Intensity Operation of the PSB 18th February 2009