Operational beams for the LHC
OOPERATIONAL BEAMS FOR THE LHC
Y. Papaphilippou, H. Bartosik, G. Rumolo, D. Manglunki, CERN, Geneva, Switzerland
Abstract
The variety of beams, needed to set-up in the injectors asrequested in the LHC, are reviewed, in terms of priority butalso performance expectations and reach during 2015. Thisincludes the single bunch beams for machine commission-ing and measurements (probe, Indiv) but also the standardphysics beams with 50 ns and 25 ns bunch spacing andtheir high brightness variants using the Bunch Compres-sion Merging and Splitting (BCMS) scheme. The requiredparameters and target performance of special beams likethe doublet for electron cloud enhancement and the moreexotic 8b ⊕
4e beam, compatible with some post-scrubbingscenarios are also described. The progress and plans for theLHC ion production beams during 2014-2015 are detailed.Highlights on the current progress of the setting up of thevarious beams are finally presented with special emphasison potential performance issues across the proton and ioninjector chain.
INTRODUCTION
During the LHC Run 1, the LHC physics production wasbased on beams with 50 ns bunch spacing. Beams with25 ns bunch spacing were injected into LHC on few oc-casions for injection tests, Machine Developments (MDs),the scrubbing run followed by a pilot physics run [1]. Afterthe startup in 2015, apart from the LHC collision energywhich will be raised to 6.5 TeV per beam, it will be crucialto establish physics operation with the nominal 25 ns bunchspacing in order to maximise the integrated luminosity inRun 2 for the limited event pile-up acceptable by the LHCexperiments [2]. The LHC will thus request a large vari-ety of beams, including single bunches for machine com-missioning and measurements but also the standard physicsbeams with 50 ns and 25 ns bunch spacing and their highbrightness variants using the Bunch Compression Merg-ing and Splitting (BCMS) scheme [4, 5]. In addition, spe-cial beams like the doublet for electron cloud enhancement[1] and the more exotic 8b ⊕
4e beam [7], compatible withsome post-scrubbing scenarios should be also prepared andmade available from the injectors.This paper reviews the parameters of the LHC physicsbeams achieved in the injectors until 2012 and the expec-tations for their performance in the following run (for adetailed analysis see [3]). The progress and plans for theLHC ion production beams during 2014-2015 are also fi-nally presented.
SINGLE BUNCH BEAMS
During the preparation of the LHC p-Pb run in 2013, anew improved production scheme has been developed [8],with which single bunch LHC beams can be generated inthe PSB. The main ingredient was the revision of the con-trolled longitudinal blow up during first part of PSB cycle,through optimisation of C16 and C02 parameters. Thereby,the C16 voltage can be used for intensity control. This as-sures the preservation of the 6D phase space volume for dif-ferent intensities with excellent shot-to-shot reproducibil-ity and control of both intensity and longitudinal emit-tance. It is therefore expected that after Long Shutdown1 (LS1) the injectors will be able to deliver LHCPROBEbunches ( × − × p/b) and LHCINDIV bunches( × − × p/b) to the LHC with smaller intensityfluctuations compared to the operation during Run 1. TheLHCINDIV parameter range was also extended in MDsto produce single bunches with up to × p/b and/orwith lower longitudinal emittances (down to 0.15 eVs), atSPS injection. These high intensity variants can be used forimpedance or beam-beam studies. Finally, a procedure forproducing Gaussian bunches for Van der Meer scans wasestablished in 2012. It is based on longitudinal and trans-verse shaving in the PSB to obtain large emittance (morethan 2.5 µ m) single bunches with under-populated tails.Because of diffusion processes in the PS and SPS, thesebunches evolve into almost perfect Gaussian shapes at theexit of the SPS and at collision in the LHC as confirmed bythe experiments. This beam will need to be ready for thevan der Meer scans at the beginning of the 2015 run and canprofit from the newly established single bunch productionscheme in the PSB. LHC PHYSICS BEAMS
LHC operation during Run 1 used mainly 50 ns beamsproduced with the standard scheme of bunch splittings inthe PS. Beams with the nominal 25 ns bunch spacing havebeen used in the LHC for the scrubbing run and machinedevelopment studies. With the successful implementationof the BCMS scheme [4, 5] in the PS in 2012, the injectorswere also able to provide LHC beams with almost twice thebrightness compared to the standard production schemes.While the 50 ns BCMS beam was injected into the LHConly an emittance preservation study of a high brightnessbeam along the LHC ramp, the 25 ns BCMS beam wasused for the 25 ns pilot physics run at the end of 2012.It should be emphasised that all these LHC beams wereproduced close to the performance limits of the injectorchain: For the 50 ns beam the intensity per bunch is close a r X i v : . [ phy s i c s . acc - ph ] D ec o the limit of longitudinal instability in the PS, whereasthe brightness of the BCMS beam is at the present spacecharge limit in the SPS. For the 25 ns beam, the intensityper bunch is close to the limit of RF power and longitudinalinstability in the SPS while the brightness is at the presentspace charge limit in the PS. Figure 1 shows the beam pa-rameters for the two types of beams as achieved in 2012after the operational deployment of the Q20 low gammatransition optics in the SPS [10, 11]. The transverse emit-tances shown in these plots are deduced from combinedwire-scans at the end of the SPS flat bottom and the valueswere cross-checked with measurements in the LHC. Theerror bars include the spread over several measurements aswell as a systematic uncertainty of 10%. The bunch inten-sity is measured at the SPS flat top after the scraping of thebeam tails, as required prior to extraction into LHC. Thesolid lines correspond to the PSB brightness curve (i.e. theemittance as a function of intensity measured at PSB ex-traction) translated into protons per SPS bunch for eachbeam type assuming intensity loss and emittance growth ( ε x + ε y ) / ( µ m ) ( ε x + ε y ) / ( µ m ) Figure 1: Beam parameters achieved operationally in theSPS in 2012 with the Q20 optics for 50 ns beams (bottom)and 25 ns beams (top) extracted to the LHC. Table 1: Operational beam parameters in 2012.
Beam type Intensity Emittance
Standard (25 ns) . × p/b 2.6 µ mBCMS (25 ns) . × p/b 1.4 µ mStandard (50 ns) . × p/b 1.7 µ mBCMS (50 ns) . × p/b 1.1 µ mbudgets of 5 % in the PS and 10 % in the SPS, respectively.All beams were produced within the allocated budgets forbeam degradation along the injector chain apart from thestandard 25 ns beam, which suffers from slow losses at theSPS flat bottom and maybe also from space charge effectsat the PS injection. Nevertheless, the nominal 25 ns beam iswell within the original specifications (i.e. . × p/band . µ m transverse emittance [12]). The beam parame-ters achieved operationally in 2012 are summarized in Ta-ble 1.The first part of the re-commissioning of the LHC beamsin the injector chain in 2014 is focused on re-establishingthe beam parameters achieved before LS1. This will rely toa large extent on the successful scrubbing of the SPS in or-der to suppress the electron cloud effect, which is expectedto be a performance limitation during the first weeks afterthe start-up since large parts of the vacuum chambers havebeen exposed to air [13].Once the 2012 beam parameters are reproduced, itshould be possible to reach slightly higher beam intensityand potentially also higher beam brightness. Already dur-ing MDs at the end of 2012 a standard 25 ns beam was ac-celerated to flat top with an intensity of about . × p/band longitudinal beam parameters compatible with injec-tion into LHC. In addition, high intensity LHC beams willbenefit from the upgraded 1-turn delay feedback for the10 MHz cavities and the upgraded longitudinal coupled-bunch feedback in the PS, which was commissioned in2014. It should also be possible to enhance the beambrightness by optimising the beam production schemes asdiscussed at the RLIUP workshop [6]: the space chargetune spread in the PS can be reduced by injecting buncheswith larger longitudinal emittance, i.e. increasing the bunchlength and the momentum spread at PSB extraction. Themaximum bunch length at the PSB-to-PS transfer is deter-mined by the recombination kicker rise time. The maxi-mum longitudinal emittance is determined by the RF ma-nipulations and by the momentum acceptance at transitioncrossing in the PS cycle, but also by the constraint thatthe final bunches should not exceed 0.35 eVs for injec-tion into the SPS. Optimising the longitudinal beam pa-rameters at PS injection requires therefore controlled lon-gitudinal blow-up during the PSB cycle with the C16 cav-ity and the use of the h=1 and h=2 PSB RF harmonics inphase at extraction to keep the larger longitudinal emit-tance bunches within the recombination kicker gap. Fur-thermore, the triple splitting in the PS was recently com-missioned at an intermediate plateau of 2.5 GeV instead ofable 2: Expected performance limits after LS1. Beam type Intensity Emittance
Standard (25 ns) . × p/b 2.4 µ mBCMS (25 ns) . × p/b 1.3 µ mStandard (50 ns) . × p/b 1.6 µ mBCMS (50 ns) . × p/b 1.1 µ mthe flat bottom for providing sufficient bucket area. Furtherdetails are given in Ref. [6]. A summary of the expectedperformance limits of LHC physics beams for the run in2015 is given in Table 2. SPECIAL BEAMS: DOUBLET AND 8b ⊕ The doublet beam was originally proposed for enhanc-ing the scrubbing efficiency in the SPS at low energy [14].This beam is produced by injecting a 25 ns beam with en-larged bunch length ( τ ≈ ns full length) from the PSonto the unstable phase of the 200 MHz RF system inthe SPS. By raising the SPS RF voltage within the firstfew milliseconds after injection, each bunch is captured intwo neighbouring RF buckets resulting in a train of 25 nsspaced doublets, i.e. pairs of bunches spaced by 5 ns. Verygood capture efficiency (above 90%) for intensities up to . × p/doublet could be achieved in first experimen-tal tests in 2012. Observations on the dynamic pressurerise in the SPS arcs confirmed the enhancement of the elec-tron cloud activity as expected from the lower multipactingthreshold compared to the standard 25 ns beams predictedby numerical simulations [14]. The experimental studiesperformed up to now concentrated on SPS injection energyand thus the acceleration of the doublet beam will be animportant milestone during the 2014 MDs (for more detailssee [13]).Thanks to its micro-batch train structure, the 8b ⊕ →
21 bunch pair split-ting, which results in pairs of bunches separated by emptybuckets. Each bunch is split in four at PS flat top suchthat the bunch pattern 6 × (8b ⊕ ⊕
8b is obtained. In thiscase the bunch train out of the PS is longer than the 72bunches of the standard scheme, but the remaining gap of 4empty buckets (about 100 ns) is expected to be sufficientlylong for the PS ejection kicker. Without optimization of theLHC filling pattern, the total number of bunches per LHCbeam is estimated as 1840. More details about the perfor-mance of this beam can be found in [15].The estimated beam parameters are summarized in Ta-ble 3. Finally it should be emphasized that this beam hasnot been produced in the injectors so far since it was devel-oped during LS1. First tests of this new beam productionscheme will be subject of MD studies in 2014 or at latest Table 3: Expected parameters of the 8b+4e beam.
Beam type Intensity Emittance
Standard (8b ⊕ . × p/b 2.3 µ mBCMS (8b ⊕ . × p/b 1.4 µ min the beginning of 2015, depending on the availability ofMD time in the injectors. PROGRESS IN 2014
The first part of the PSB and the PS startup in 2014 weredevoted to the setup of the beams needed for physics. Dur-ing the time of the Chamonix workshop 2014, the singlebunch beam were in good shape in PSB and PS, and shorttrains of 12 to 24 bunches were taken in SPS for the realign-ment campaign and RF setting-up (energy matching). Thesetup of the LHC beams in the PS complex was done in par-allel to physics operation and starting from re-establishingthe beam conditions from 2012 (but already with the triplesplitting in the PS at 2.5 GeV instead of the flat bottom).The PS complex is ready to deliver the LHC beams at thestartup of the SPS in September. As large parts of the SPShave been vented and exposed to air in the course of theworks performed during LS1, it is expected that the goodconditioning state of the SPS will be degraded. Therefore,two weeks of SPS scrubbing are planned for 2014 with thegoal of reconditioning the SPS to the state of before LS1.The success of this scrubbing run is the critical milestonefor the preparation of the 25 ns LHC beams for physics in2015.The setup of the doublet scrubbing beam for the use inthe LHC will be the subject of extensive MD studies in theSPS in 2014 in several dedicated MD blocks, for establish-ing accelerations and pushing the intensity to the requested . × p/doublet. During these MDs, also the behaviourof the LHC BPMs in the SPS with the doublet beam needto be tested in preparation of the LHC scrubbing, [17].At the same time, there are many requests for dedicatedMD time in the SPS for 2014 [18]. Careful planning andprioritization of studies will be crucial, as the total amountof requested dedicated MD time exceeds the MD slotsavailable. For example, although there are first successfulrecent studies in the PSB and the PS, the full qualificationof 8b ⊕
4e beam production scheme will be done in 2015.In general, it should be stressed that 2014 will be a verybusy period for the injectors: Besides the physics operationafter the beam commissioning with partially new or up-graded hardware, the setup and commissioning of the dif-ferent LHC beams including the doublet scrubbing beam,the various dedicated and parallel MD studies, substantialamount of beam time will be needed in the PS and SPS forthe first-time setup of the Ar-ion beams in preparation forthe physics run beginning of 2015.Finally, it is worth mentioning that there will be anotherperiod of dedicated scrubbing of the SPS in 2015. Whileigure 2: Ions production scheme for 2015 [21].with the scrubbing run in 2014 the scrubbing efficiency andthe time required for achieving acceptable conditioning af-ter a long shutdown will be qualified, the aim of the scrub-bing run in 2015 will be to condition the SPS for high in-tensity 25 ns beams. The outcome of these scrubbing runswill determine if the SPS vacuum chamber really need tobe coated with amorphous Carbon [19] as presently part ofthe baseline of the LIU project for suppressing the electroncloud for the future high intensity LHC beams [20].
ION BEAMS
The Pb-Pb run in 2011 initially projected an integratedluminosity of around 30-50 µ b-1 in 4 weeks, with peak lu-minosity L peak = 1 . × Hz/cm . In fact, the peakluminosity was increased to around half of the nominal( × Hz/cm ) exceeding by far the expectations to al-most 150 µ b-1 integrated luminosity at 3.5 ZTeV. This wasdue to the increased LEIR brightness with nominal bunchpopulation of . × Pb ions per bunch but smalleremittances. Additional ingredients of this success werethe preservation of the brightness at low energy in PS dueto excellent vacuum conditions, the modified productionscheme (no splitting in PS allowing half as many buncheswith twice the intensity/bunch) and the good behaviour ofbunches on SPS flat bottom (improved low level RF to re-duce noise, IBS and space charge less critical than expectedand delivered with Q20 optics after 2013). For the p-Pb runin 2013, the LEIR bunch intensity was further increasedto . × Pb ions per bunch, exceeding the nominalvalue by a factor of 1.2. Assuming the same scheme asin 2011 and the performance of 2013, a Pb-Pb peak lumi-nosity of L peak = 2 . × Hz/cm at 6.5 ZTeV can beexpected. A further 20% increase in peak luminosity can be gained by squeezing 20% more bunches in LHC. The iongeneration scheme is presented in Figure 2 (for more de-tails see [21]). A batch compression already tested in thePS in 2012 can allow a bunch spacing of 100 ns betweentwo ion bunches. Twelve of these two-bunch batches can beaccumulated for every cycle of the SPS, with a batch spac-ing of 225 ns. After 36 injections from the SPS, assumingonce again the same brightness as in February 2013, thisscheme can deliver up to 432 bunches of . × Pb ions per LHC ring, corresponding to a peak luminosity at6.5 ZTeV of peak = 2 . × Hz/cm . SUMMARY AND CONCLUSIONS
Several optimizations of the beam production schemeswill be implemented for the LHC Run after LS1. Singlebunch beams already benefit from a better control and bet-ter reproducibility of intensity and longitudinal emittance.The longitudinal parameters at PSB-to-PS transfer of the25 ns and 50 ns physics beams are optimized for allowingeven higher beam brightness and, if requested by the LHC,the intensity of the 25 ns beams can also be slightly pushedcompared to the 2012 beam parameters. The first step inthe beam commissioning of these LHC beams in 2014 willbe however to recover their 2012 performance. In this re-spect, the critical milestone will be the success of the SPSScrubbing Run, as it is expected that the good conditioningstate of the SPS will be degraded due to the long periodwithout beam operation and the venting of machine sectorsrelated to the interventions during LS1.The setup of the doublet scrubbing beam with acceler-ation in the SPS in preparation for the LHC scrubbing in2015 will be one of the main topics of MDs in 2014. Care-ful planning and prioritisation of the dedicated MDs in thePS will be crucial due to the limited MD time available.First tests of the 8b ⊕
4e beam already demonstrated the fea-sibility of the scheme and need to be tested further in 2015,in the SPS.Besides the various physics users, the commissioning ofthe LHC beams and the MDs related to the new beams re-quested by the LHC, lots of beam time will be needed in2014 for the first-time setup of Ar-ion beams. Regardingthe ion performance, a batch compression scheme in thePS can increase the projected 2013 performance by around20% in peak luminosity.
ACKNOWLEDGEMENTS
The authors would like to thank G. Arduini, T. Argy-ropoulos, M. Bodendorfer, T. Bohl, R. Bruce, C. Cornelis,H. Damerau, J. Esteban-Mller, A. Findlay, R. Garoby, S.Gilardoni, B. Goddard, S. Hancock, K. Hanke, G. Iadarola,B. Mikulec, E. Shaposhnikova, R. Steerenberg and the PSB& PS & SPS Operation crews for support and helpful dis-cussion.
REFERENCES [1] G. Iadarola and G. Rumolo, “Electron cloud/scrubbing”,Proceedings of the 2014 Evian Workshop on LHC BeamOperation, CERN-ACC-2014-0319 (2014).[2] B. Gorini, “Experiments’ expectations”, , Proceedings ofthe 2014 Evian Workshop on LHC Beam Operation, CERN-ACC-2014-0319 (2014).[3] H. Bartosik and G. Rumolo, “Beams in the injectors”, Pro-ceedings of the 2014 Evian Workshop on LHC Beam Oper-ation, CERN-ACC-2014-0319 (2014).[4] R. Garoby, “New RF Exercises Envisaged in the CERN-PS for the Antiprotons Production Beam of the ACOL Ma-chine”, IEEE Transactions on Nuclear Science, Vol. NS-32.,No. 5 (1985).[5] H. Damerau et al. , “RF manipulations for higher beambrightness LHC-type beams”, CERN-ACC-2013-0210.[6] G. Rumolo et al. , “Expected performance in the injectors at25 ns without and with LINAC4”, proceedings of the Re-view of LHC and Injector Upgrade Plans Workshop (2013).[7] H. Damerau et al. , “LIU: Exploring alternative ideas”, pro-ceedings of the Review of LHC and Injector Upgrade PlansWorkshop (2013).[8] S. Hancock, CERN-ATS-Note-2013-040 MD (2013).[9] H. Bartosik and G. Rumolo, CERN-ACC-NOTE-2013-0008(2013).[10] H. Bartosik et al. , “Increasing instability thresholds in theSPS by lowering transition energy”, CERN-ATS-2012-177,IPAC2011.[11] Y. Papaphilippou et al. , “Operational performance of theLHC proton beams with the SPS low transition energy op-tics”, CERN-ACC-2013-0124, IPAC2013.[12] LHC Design Report, edited by O. Br¨uning, P. Collier, P. Le-brun, S. Myers, R. Ostojic, J. Poole, P. Proudlock, CERN-2004-003-V-1 [13] H. Bartosik et al. , “SPS scrubbing 2014”, these proceedings.[14] G. Iadarola et al. , “Recent electron cloud studies in theSPS”, CERN-ACC-2013-0115.[15] H. Rumolo et al. , “Protons: Baseline and Alternatives, Stud-ies Plan”, these proceedings.[16] H.Bartosik et al. , “Special e-cloud bunch spacing: injec-tors”, LBOC meeting of November 5th (2013).[17] T. Lefevre et al., “Special e-cloud bunch spacing: BI com-patibility”, LBOC meeting of November 5th (2013).[18] G. Rumolo and H. Bartosik, “Injector MD planning and re-quests for 2014”, MSWG meeting of May 30th (2014).[19] C. Vallgren et al. , “Amorphous carbon coatings for the mit-igation of electron cloud in the CERN Super Proton Syn-chrotron”, Phys. Rev. ST Accel. Beams, 14, 071001 (2011).[20] H. Bartosik et al.et al.