Joint LHC Machine-Experiments Meeting on Experiment Protection from Beam Failures

1. Joint LHC Machine-Experiments Meeting on Experiment Protection from Beam Failures Summary and action list D. MacinaTS/LEA/INT LEMIC 91

2. Machine protection at the LHC :‘organization’ • Machine protection activities of the LHC and the SPS are coordinated by the LHC Machine Protection Working Group (MPWG), co-chaired by R. Schmidt & J. Wenninger. • The MPWG WEB site (only from inside CERN !) • http://lhc-mpwg.web.cern.ch/lhc-mpwg/ • The detailed description of the commissioning of the LHC Machine Protection System and the record the test results is the mandate of the Commissioning of the MachineProtection System Sub-Working-Group. It reports to the LHCCWG. Web site: • http://lhccwg.web.cern.ch/lhccwg/MPS/mps.htm • New : • The MPWG has asked each LHC experiment to provide a contact person (BISU) to interface on machine protection issues between machine and experiments. Antonello Di Mauro(ALICE) Siegfried Wenig(ATLAS) Alick Macpherson(CMS) Richard Jacobsson(LHCb) Anne-Laure Perrot(LHCf) Mario Deile(TOTEM)

3. Program

4. Safe beam The MPWG has adopted for the LHC a limit for safe beams (nom. emittance) of 1012 protons at 450 GeV ( <~ Cu melting point) 1010 protons at 7 TeV (scaled from 450 GeV) - under discussion!! Standard procedure: if empty machine => inject pilot bunch (5x109 p) BUT it is technically possible to inject a safe beam into an empty machine (no interlock forbids it) If safe beams no injection protection required and, in case of circulating beam, absorbers and collimators at relaxed settings/not required. Safe beams are used for the fine tuning of the machine protection elements at each new fill. Some interlock may/will be masked IF and ONLY IF beam is safe

5. Beam cleaning • The protons that are lost must be intercepted with very high efficiency before they can quench a superconducting magnet : collimation! • Unlike HERA, TEVATRON, RHIC.. the LHC cannot be operated without collimators (except at injection with low intensity). • At the LHC the collimators must define the aperture (primary + secondary) which has an important impact for MP : for most multi-turn failures the beam will hit collimators first !

6. Machine Aperture ATLAS • Vertical axis : • Machine aperture in units of beam sigma (s), includingalignmenterrors and othertolerances. • Horizontal axis : • Longitudinal position on leftside of ATLAS (seenfrom the ring center). 450 GeV Arc • Injection : • Aperture limit is the LHC ARCs (~ 7-8 s). • The triplet magnets in front of ATLAS/CMS are slightly behind the ARC (~ 8-9 s). •  Collimators @ ~5-6 s ! 7 TeV, b* 0.5 m • Collisions, squeeze to b* 0.5 m : • Aperture limit is defined by the triplet magnets in front of ATLAS/CMS (~ 8 s).. •  Collimators @ ~6 s ! Triplet End of the squeeze (* < 6 m) bottleneck in the triplets => TCT needs to be closed 6 Experiments-Machine WS / June 07

7. Additional general items • Software interlock • The reaction time of this system will be at the level of a few seconds. • The systems rely entirely on the machine technical network, databases, etc – clearly not as safe as HW systems ! • The Injection SIS will complement the Injection Interlock Systems: • - Active from the beginning. • - Stops injection through the Injection Interlock Systems. • -Receives the experiments injection inhibits, either through our middleware (preferred) or through DIP. • Post Mortem Analysis • All machine user systems connected to the LHC BIS must provide Post-Mortem informationfor diagnostics, even systems that have NOT triggered the dump. • To resume beam operation after a beam dump, the PM data must be analysed and ‘understood’

8. Injection - Extraction • Same principle for injection and extraction: • “kicker” magnets: fast rise time →muchless than one turn, large (~ mrad) angles • septa: two(/more) apertures with different magnetic fields • Injection: • Beam 1: IR 2 • Beam 2: IR 8 • Extraction: • Both beams in IR 6 • Injection process and extraction process can lead to single turn beam loss!! => • Everything” required for injection/(extraction) plus the LHC ring must have the right value/state • Dedicated passive protection

9. MKI Transfer line Q5 Q4 D2 MSI IP2 Injection overview • The injection septum MSI bends the beam into the LHC in the horizontal plane • The injection kicker MKI kicks the beam in the vertical plane onto the LHC orbit • Possible injection failures: • Wrong current setting in the transfer line magnets + injection septum, fast trip of power supply, failure of SPS extraction kicker (MKE) during extraction, fast trip of power supply – e.g. downstream of transfer line collimators (protection: current surveillance, fast current monitors, absorbers) • Injection kicker failures in the LHC => caught by DEDICATED MOVEABLE ABSORBER TDI behind kicker and auxiliary movable collimators TCLI in the injection regions

10. Error settings at injection • Error settings at injection (empty machine) of D1/D2 and orbit correctors in the LSS may drive in beam in the experimental areas.It affects all experimental insertions. • Not possible to hit directly the detectors if only one failure considered (at least in IR1/IR5) Ex. MCBXH wrong settings (30-100%) nominal at injection is 5%

11. Septum magnet deflecting the extracted beam MSD Beam 1 H-V kicker for painting the beam Beam Dump Block Q5L Q4L 15 kicker magnets MKD about 700 m Q4R about 500 m Q5R Beam 2 LHC Extraction (IR6) • Clean dump • Synchronisation: the kicker rise time must coincide with the particle free abort gap • Particle free abort gap • Energy Tracking – (required kick strength from the beam energy via magnet currents)

12. Possible failures at extraction • Lost synchronisation with the abort gap • Pre-fire of one of the 15 kicker modules (hard-wired system triggers all remaining 14 in 0.7-1.3 µs ) • Frequency: once per year possible • Solution: dedicated passive protection; • TCDQ (movable) and TCS protect Q4 and general LHC apertures • TCDS protects the MSD septa • In case of problems during extraction coupled with TCDQ settings, orbit or optics errors, “some” beam loss at the tertiary collimators/triplets in IR5 may occur. • Difficult to quantify - detailed analysis ongoing. PhD thesis of Thomas Kramer, AB/BT. • If abort gap (re-)population => caught by TCDQ and interlocking on abort gap monitor signal

13. Possible failures with circulating beam(magnet failures including operation mistakes) • Usually slow and detected first in the aperture restrictions of the machine (cleaning insertions) • Potential danger for the experiments (VELO and RP): uncontrolled closed bumps since they could affect only experimental areas. However: • Extremely difficult to create at 7 TeV (less difficult at 450 GeV) • They build up slowly: first losses seen by the BLMs and beam extraction before damage threshold • Only critical if combined with a fast failure => low probability

14. Summary of machine failures • Errors/failures which may drive the beam into the experimental areas; • Settings errors at injection • Undetected closed bump & fast magnet failures (highly unprobable) • Errors/failures which may drive the beam in the aperture restrictions in the LSS: • Injection errors (IR2, IR8) • Asynchronous beam dump (IR5)

15. Experiments (1) • ATLAS • BCM+ LUCID used for beam abort • Injection errors: ok for pilot bunch, concern for Safe Beam. What is the structure of Safe Beam? • Questions: mechanism for beam blow up at the IP? How loss of vacuum close to IP is treated? • ALICE • BCM used for beam dump (same system developed by LHCb). Position dictated by available space. Needs failures/error simulations to show the position is valid for the most relevant scenarios • Consequences of injection errors: old analysis. Still valid? • During injection: only BCM ON and possibly a few detectors at safe setting to monitor the losses in the experimental areas • Concern about Safe Beam

16. Experiments (2) • CMS • BCM with standard LHC BLM Readout and Interface • Worry particularly about high rates of loss rather than integrated doses • Needed updated simulation (most realistic worst case scenario) for asynchronous beam dump and injection • IF BCM OFF (no protection system) => no beam allowed in the machine • IF CMS OFF but BCM ON => beam allowed • BCM and CIBU on machine power (no local UPS) • Concern about Safe Beam

17. Experiments (3) • LHCb • BCM for beam dump. Readout front end same as BLMs • Entire BCM system under UPS • Concern On Safe Beam • Needs simulation for beam failures • LHCf • Beam dump based on the detector itself. All scintillators hit => beam dump • The experiment will run only for a few days per year: disconnected from the BIC when not in the TAN • The whole system is on ATLAS UPS • Scintillators always ON at injection

18. Experiments (4) • TOTEM • T1 and T2 should be protected by the CMS system • RP: simulation for bunch hitting the pot: • For β*=1540 m only ok for bunch <= 6x1010 p • For β*=90-0.5 m beam wider, depends on number of bunches • TOTEM CIBU only used for the positioning of the RP. The other CIBU forced TRUE, may be used later

19. Open issues and action list (1) • Safe Beam • This is a concern for all experiments. If not absolutely needed, experiments ask to lower to 1.1-1.5x1011 p. Is this OK or just a feeling? Can be simulated/calculated? • Action: MPWG • Post Mortem Data • Data from the experiments for all dumps • Experiments need PM Data from the Machine? • Details to be discussed • Action: AB/CO+ AB/OP + LEADE • Safe Beam Flag to the experiments • Technically all experiments can receive it via the GMT receiver • Only LHCb has requested it • Action: BISU

20. Open issues and action list (2) • Injection inhibit • By SW: decide on protocol • New idea for HW with fiber-CIBU • Action: MPWG + AB/CO + LEADE • CIBU (and Protection System) Power • On UPS on two different power sources • Survive a power cut for 10 minutes • CMS philosophy: BCM must be ON whenever the machine is ON => machine power and not CMS power • Apply to all experiments? • Action: MPWG + LEADE + BISU

21. Open issues and action list (3) • CMS Philosophy • No beam allowed if CMS protection system is OFF • Beam allowed if CMS not operational • Adopted by all experiments? • What is the reliability of the experiment’s protection system? Is it there a backup solution? • How many unjustified dump per experiments? • Aim for 1/year/exp? • Procedure after emergency beam dump • CMS request to latch their input until PM check is done. Check re-arm time scale Action?

22. Open issues and action list (4) • How to detect direct hit in experimental area at injection if detector off? • Possible with BCM even if may be not always sufficient • A few additional detectors ON at safe settings? • Action: BISU • TCTs for 2008 • Missing Vertical TCT in IR2 and IR8. Is this a problem for small *? • Action: MPWG + Collimation WG • Bump scenarios and wrong settings at injection • Software Interlock surveillance of D1/D2/correctors in the experimental insertions at injection & collision • Action: SIS + MPWG

23. Open issues and action list (5) • Abort gap monitor • Status + connection to BIC • Action: MPWG + AB/BI • Simulations • Asynchronous beam dump (including RP) => AB/BT • Consequences of the failure scenarios at injection => (AB/BT? TS/LEA? BISU?) • Error settings at injection (in particular IR2/IR8) => (TS/LEA?) • Additional failure scenarios when circulating beam (beam steering) => (AB/ABP?, AB/CO/MI??) • RP: simulations including mechanical stresses => (TS/LEA-MME?)

24. To be done in addition… • Get more information and check consistency of the radiation damage threshold for similar detectors in different experiments • Better define the injection inhibit philosophy • Define/organize the test procedures • Define/organize the commissioning procedures together with the Commissioning of the Machine Protection System Sub-Working-Group