1. 6/14/11 Collimation Upgrade Plan & Questions R. Assmann, CERN for the collimation team 14/6/2011 LHC Collimation Project Review

2. LHC Collimation as Staged System LHC collimation was conceived in 2003 as a staged system. Phase 1: –For initial beam commissioning and early years of LHC operation. –Predicted not adequate for nominal and ultimate intensity. –Designed, constructed and commissioned 2003 – 2009. Phase 2: –Upgrade for nominal, ultimate and higher beam intensities. –Solves issues in efficiency, impedance and radiation impact. –Originally not clear what the solution would be. –By now various upgrade solutions worked out and under design. IR upgrade: –Adaptation to changes in IR upgrades: space and losses. –Adaptation to phase space modifications (ATS, crab cavities). 6/14/11

3. Overall Collimation Upgrade Plan (as defined in 2009) 6/14/11 Initial collimation system (2009 – 2012) Inefficiency: 0.02 % (p)  * ~ 1 – 1.5 m, 3.5 TeV R2E limits in IR7? > 4 days per setup Initial collimation system (2009 – 2012) Inefficiency: 0.02 % (p)  * ~ 1 – 1.5 m, 3.5 TeV R2E limits in IR7? > 4 days per setup Full collimation system (2018 onwards) Inefficiency: 0.0004 % (p)  * ~ 0.55 m, 7 TeV L not limited (p and ions) 30 s per high accuracy setup Radiation optimization Full collimation system (2018 onwards) Inefficiency: 0.0004 % (p)  * ~ 0.55 m, 7 TeV L not limited (p and ions) 30 s per high accuracy setup Radiation optimization Interim collimation system (2014 – 2016) Inefficiency: 0.002 % (p)  * ~ 1 – 2 m, 7 TeV Gain ~100 in R2E (IR7  IR3) L ≤ 5 × 10 33 cm -2 s -1 nominal ion intensity > 2 days per setup Interim collimation system (2014 – 2016) Inefficiency: 0.002 % (p)  * ~ 1 – 2 m, 7 TeV Gain ~100 in R2E (IR7  IR3) L ≤ 5 × 10 33 cm -2 s -1 nominal ion intensity > 2 days per setup 2013 shutdown: IR3 DS combined cleaning, IR2 TCT’s, TCLP installation? 2017 shutdown: IR(1)/2/(5)/7 DS Phase 2: integrated BPM’s, robust materials, red. impedance. Radiation opt. Collimation IR Upgrade (2022 onwards) Low  *, 7 TeV TCT’s integrated into IR upgrade Compatibility with crab cavities Collimation IR Upgrade (2022 onwards) Low  *, 7 TeV TCT’s integrated into IR upgrade Compatibility with crab cavities 2021 shutdown: tbd

4. Prepared, Empty Secondary Collimator Slots for Phase 2 EMPTY PHASE II TCSM SLOT (30 IN TOTAL) PHASE I TCSG SLOT 1 st advanced phase 2 collimator CERN SLAC design 6/14/11

5. Luminosity 6/14/11 Triplet aperture and collimation setup accuracy  R. Bruce Loss limits: collimation, (UFO’s), …  D. Woll- mann, A. Rossi, G. Bellodi Beam- beam, brightness & robust- ness limits  A. Dalloc- chio (new materials)

6. Good news: –Available aperture about 50% larger than guaranteed by design (smaller orbit errors, better alignment, …). Gain here for luminosity! –Optics very well controlled (5-10% beta beat, … for  * = 1.5m). Gain here! As expected: –Very challenging to achieve collimation & protection tolerances (only infrequent setups possible, drifts over months, …)   * limited. –Addressed by collimators with integrated beam position pickups (almost all to be equipped). Not discussed in details for this review. 6/14/11

7. Good news: –Collided successfully three times nominal brightness (head-on). Long-range beam-beam soon to be checked. Gain factor 3 here, if LR beam-beam OK as well! Under study: –Robustness of collimators for the high achieved brightness. Simulation of realistic scenarios, tests in HiRadMat facility starting in autumn. –Development of more robust collimator materials (  EuCARD/ColMat program since 2009, report A. Dallocchio). –Not discussed in details for this review. 6/14/11

8. Good news: –Since middle of May: ~ complete experimental assessment at 3.5 TeV done. –Reached the design 500 kW peak beam loss (protons) at primary collimators without quench of a super-conducting magnet! –Reached 80 MJ without a single quench from stored beam losses. –Transverse damper stabilizes beam at 3.5 TeV  high impedance OK. –Reached 99.995% collimation efficiency with 50% smaller gaps than design (low emittance, high impedance) and due to much less impact of imperfections than predicted (better orbit, lower beta beat, …). –Minimum beam lifetime at 3.5 TeV is ~4 times better than specified. 6/14/11

9. Collimation of High Power Loss 6/14/11 No quench of any magnet!

10. Ultra-High Efficiency 6/14/11 99.995 % worse better MD 99.960 %

11. Achieved Stored Energy: 80 MJ 6/14/11 80 kg TNT

12. Stored Energy Compared to 2010 Goals 6/14/11

13. Therefore some questions I It runs so well: Do we really need to invest a lot of work for a better collimation efficiency in the first long LHC shutdown (2013/14)? Do operational experience and MD measurements not prove to us sufficiently well that we can reach nominal 7 TeV luminosity in 2014/15 (with the efficiency of the present collimation system)? Do the potential gains in  * and beam brightness (beam-beam) not provide an additional margin to increase luminosity (without pushing stored energy)? 6/14/11 Reference p goal 2014 – 2017: L ≥ 1 × 10 34 cm -2 s -1 at 7 TeV Could be reached with ~50% of nominal intensity?

14. On the Other Side Predicted leakage mechanisms and locations are fully confirmed, both for protons and ions. Proposed upgrade plan will gain factor ~10 in efficiency: can be used for higher stored energy and/or larger collimation gaps (relaxed tolerances and lower impedance). Lowest risk approach. All experience relies on 3.5 TeV beam energy (higher quench margin, larger collimation gaps, lower impedance, easier operation for transverse damper, lower cross-section single-diffractive scattering, …). All experience relies on operation with 1/2 of nominal emittance (50 ns)  beam core far away from jaw surface, lower loss spikes, more room to close collimator gaps. It is assumed that 7 TeV beam is as stable as 3.5 TeV, that quench limits and efficiency scale as predicted and that losses do not become more localized at 7 TeV. 6/14/11

15. Protons: Simulations vs Measurement B1v, 3.5TeV, β*=3.5m, IR7 6/14/11 B1 Losses in SC magnets understood: location and magnitude Simulated (ideal) Measured Cleaning Inefficiency

16. 3.5 TeV: Luminosity Operation Collimation 6/14/11 Fill #1645, 200 bunches, 2.4e13 p per beam, peak luminosity 2.5e32 Collimation IR3 Colli- mation IR7 ATLAS CMS LHCb Colli- mation IR6

17. Origin of Dispersion Suppressor Losses 6/14/11 Quad Dipole Coll Quad Dipole Coll Collision p – p Pb – Pb Collision p – C Coll. Mat. on energy off energy

18. Zoom IR7 (and illustration of 2013 upgrade for IR3) 6/14/11 D. Wollmann, G. Valentino, F. Burkart, R. Assmann, …

19. quench level Proton losses phase II: Zoom into DS downstream of IR7 Impact pattern on cryogenic collimator 1 Impact pattern on cryogenic collimator 2 Simulation T. Weiler Very low load on SC magnets  less radiation damage, much longer lifetime. 99.997 %/m  99.99992 %/m Cryo-collimators can be one-sided! Simulation 6/14/11

20. Phase 1 Phase 2 Gap × 1 × 1.2 × 1.5 × 2 Ideal Inefficiency [1/m] Impedance better Better Efficiency and/or Lower Impedance Acceptable Area R. Assmann T. Weiler E. Metral Target Inefficiency (nominal intensity, design peak loss rate) Installation of collimation phase II including collimators in cryogenic dispersion suppressors WARNING: Grid simulation here for non- nominal optics and perfect machine! Increase gaps by factor 1.5 Nominal I. Larger triplet/IR aperture or lower  * Impedance Target Phase 1 (full octupoles, no transv. feedback, nominal chromaticity) Impedance Target Phase 2 (full octupoles, no transv. feedback, nominal chromaticity) 6/14/11

21. Ions: Beam 2 Leakage from IR7 Collimation (much worse, as expected) 6/14/11

22. Therefore some questions II Can the upgrade of the IR3 dispersion suppressors be delayed without any danger for magnet lifetime (SC magnets as halo dumps)? Is later upgrade work feasible in dispersion suppressors (activation)? Are we sufficiently sure about 7 TeV beam behavior to give up the improvement in collimation efficiency and/or impedance for 2014? Is the presently predicted “proton” safety factor ~4 above nominal intensity big enough (  assumptions and energy scaling)? Do we need an upgrade of the IR3 dispersion suppressors for reaching nominal ion luminosity? Will a delay of the IR3 dispersion suppressors lead to unacceptable knock-on effects for other dispersion suppressor work (IR2 for ions, IR1/5 losses into dispersion suppressors, …)? Will decision force us to work with small emittances (impact on 25 ns)? 6/14/11

23. Overall Collimation Plan (possible modification, acceptable risk?) 6/14/11 Initial collimation system (2009 – 2012) Inefficiency: 0.005 % (p)  * ~ 1 – 1.5 m, 3.5 TeV R2E limits in IR7? > 4 days per setup Initial collimation system (2009 – 2012) Inefficiency: 0.005 % (p)  * ~ 1 – 1.5 m, 3.5 TeV R2E limits in IR7? > 4 days per setup Full collimation system (2018 onwards) Inefficiency: 0.0004 % (p)  * ~ 0.55 m, 7 TeV L not limited (p and ions) 30 s per high accuracy setup Radiation optimization Full collimation system (2018 onwards) Inefficiency: 0.0004 % (p)  * ~ 0.55 m, 7 TeV L not limited (p and ions) 30 s per high accuracy setup Radiation optimization Initial collimation system (2014 – 2016) Inefficiency: 0.005 % (p)  * ~ 1 – 2 m, 7 TeV Gain ~100 in R2E (IR7  IR3) L ~ 1 × 10 34 cm -2 s -1 Ion intensity and lumi limits > 2 days per setup Initial collimation system (2014 – 2016) Inefficiency: 0.005 % (p)  * ~ 1 – 2 m, 7 TeV Gain ~100 in R2E (IR7  IR3) L ~ 1 × 10 34 cm -2 s -1 Ion intensity and lumi limits > 2 days per setup IR2 TCT’s, combined cleaning IR3, TCLP installation? 2017 shutdown: IR(1)/2/3/(5)/7 DS Phase 2: integrated BPM’s, robust materials, reduced impedance. Radiation opt. Collimation IR Upgrade (2022 onwards) Low  *, 7 TeV TCT’s integrated into IR upgrade Compatibility with crab cavities Collimation IR Upgrade (2022 onwards) Low  *, 7 TeV TCT’s integrated into IR upgrade Compatibility with crab cavities 2021 shutdown: tbd

24. Conclusion Equipping the IR3 dispersion suppressors with collimators improves the performance reach for LHC and has the lowest risk for LHC performance. It was defined as a minimal plan some years ago. There are a number of recent good news at 3.5 TeV in collimation and other LHC areas that must be taken into account: –It opens the possibility to discuss delaying the IR3 collimation upgrade in the dispersion suppressors by three years. –Some important issues were summarized and some questions put up that require attention and advice. –Subsequent talks will go into more details. Predicting performance at 7 TeV is tricky and quite involved: loss spikes, quench limit, nuclear physics p/ions, energy deposition details, small collimation gaps, high impedance, … Your advice is very much welcome! 6/14/11

25. Additional Info 6/14/11

26. Origin of Losses in Dispersion Suppressor Effect understood and predicted as early as 2003. Collimators in straight sections “generate” off-momentum p and ions (effectively). Off-momentum particles pass through straight sections and are deflected by first dipoles in dispersion suppressors. Downstream magnets act as off- momentum halo beam dump. SC regions off-hands: Impossible to put collimators in dispersion suppressors (as in LEP). Clear physics sources: p have single-diffractive scattering in matter, ions dissociate/fragment! Now confirmed by experimental data (also in horizontal plane). Loose factor ~10 with non- smooth aperture (alignment)! 6/14/11

27. p – C Interaction: Multiple Coulomb & Single-Diffractive Scattering 6/14/11

28. Analytically Derived Simple Scaling Law (E 0 = 1 TeV) 6/14/11 MCS SD R. Assmann, Proc. HE-LHC Workshop

29. Monte-Carlo Simulation of Realistic Beam Halo and Interactions 6/14/11

30. Why Off-Energy Hadrons can be so Disturbing Loss pattern cannot be compared to case of point scatterers like UFO’s or wire scanners  very diluted showers. Off energy hadrons produce a very sharp impact line. BLM’s cannot distinguish the two cases! Important uncertainties about BLM response and thresholds with such a concentrated loss. Plan quench tests for this case. 6/14/11 Point scatterer (e.g. UFO) Low energy tail after V bend (A) (B) Interaction Halo/shower (A) Very diluted  Very low risk for quench  “Fixed” by relaxing BLM limits (small T) (B) Concentrated losses  High risk for quench  Protect by tight BLM limits (medium – large T)

31. 3.5 TeV: Losses in DS of IR5 (CMS) 6/14/11 Fill #1647, 200 bunches, 2.4e13 p per beam, peak luminosity 2.5e32

32. Simple Extrapolation of Losses in Dispersion Supressor of IR5 ParameterFill #1645, 3.5 TeV7 TeV scaled Luminosity0.025 × 10 34 cm -2 s -1 1 × 10 34 cm -2 s -1 Loss @ BLM3.1 × 10 -6 Gy/s2.4 × 10 -4 Gy/s Limit @ BLM5 × 10 -4 Gy/s~3 × 10 -4 Gy/s Int. loss @ BLM for 200 d at 75% efficiency 0.039 kGy/y3.1 kGy/y Int. peak loss magnet coil (must be much higher) ?? Limit for int loss in dipole?? 6/14/11 Note: Does not include significant loads from ion operation. Does not include effect of  *. Does not include steeper scaling of losses with lumi (up to factor 5 higher  paper Annika Nordt). Win with monitor factor? Should be able to gain something with TCL/TCLP collimators (cannot fix problem due to zero dispersion). In the past strong concerns about dipoles with this load (K.H. Mess). Now OK? Clear conclusion: NOT AT ALL COMFORTABLE!

33. Quench Limit vs Energy 6/14/11

34. Where to Find Links to Info (New and Old)? 6/14/11 https://espace.cern.ch/lhc-collimation-workspace Links to past meetings, minutes, presentations, …

35. Where to Find or Put Reference Info for Upgrade? 6/14/11 https://espace.cern.ch/lhc-collimation-upgrade Minutes from collimation upgrade management meetings, agreed production and installation, tables, agreed planning, safety, …