Machine Protection Issues affecting Beam Commissioning Beam Instrumentation
Machine Protection Issues affecting Beam Commissioning Beam Instrumentation other than BLMs linked to the Machine Protection System Chamonix Workshop XIV CERN - 17 th-21 st January 2005 Rhodri Jones (AB/BDI) from discussions with David Belohrad, Brennan Goddard, Jose Gonzalez, Patrick Odier, Hermann Schmickler, Rüdiger Schmidt, Jorg Wenninger and others
Outline • Beam Position Interlock for the LHC Beam Dump System in IR 6 • Interlock for Fast Beam Position Changes and Oscillations • Capabilities of Fast Beam Current Transformers for machine protection Machine Protection Issues affecting Beam Commissioning - Rhodri Jones (AB/BDI) 2
Beam Position Interlock for the LHC Beam Dump Why do we need an interlock? • Dump channel aperture is very tight ® particularly tight at injection • Need to avoid damage of beam dump equipment ® for all failure scenarios • Limit induced Q 4 quenches to real failure modes ® no quench of Q 4 for normal dump operation Machine Protection Issues affecting Beam Commissioning - Rhodri Jones (AB/BDI) 3
A Clean On-Axis Beam Dump at 450 Ge. V 4σ 8σ M S D 7. 2σ aperture To Dump (TDE) Q 4 TCDS 7. 8σ aperture TCDQ MKD 7. 2σ aperture M S D 8. 2σ aperture Machine Protection Issues affecting Beam Commissioning - Rhodri Jones (AB/BDI) 4
+4 mm Off-Axis Beam Dump at 450 Ge. V 4σ 8σ To Dump (TDE) 4. 6σ M S D aperture Q 4 TCDQ MKD TCDS M S D 4. 6σ aperture +4 mm M S D Machine Protection Issues affecting Beam Commissioning - Rhodri Jones (AB/BDI) 5
-4 mm Off-Axis Beam Dump at 450 Ge. V 4σ 8σ To Dump (TDE) M S D Q 4 TCDQ MKD 4. 2σ aperture TCDS M S D 5. 0σ aperture D S M -4 mm Machine Protection Issues affecting Beam Commissioning - Rhodri Jones (AB/BDI) 6
-4 mm & 14/15 MKD Beam Dump 450 Ge. V 4σ To Dump (TDE) 8σ M S D Quench Q 4 TCDS M S D Particle Shower TCDQ MKD 1. 2σ aperture -4 mm No Damage to Any Equipment M S D Machine Protection Issues affecting Beam Commissioning - Rhodri Jones (AB/BDI) 7
Summary of Acceptable Orbit Variations for the Beam Dump System (see talk by Brennan Goddard at Chamonix 2003) • 4 mm Limit at Injection (450 Ge. V) ® Safe for ALL beams ® Q 4 quench may occur for large emittances ® Q 4 quench likely to occur in 14/15 MKD case • 4 mm Limit at 7 Te. V ® Safe for MOST beams ® quench of Q 4 will only occur in 14/15 MKD case ® damage may occur for large emittance nominal beams – Can be avoided by reducing the limit to 3. 5 mm Machine Protection Issues affecting Beam Commissioning - Rhodri Jones (AB/BDI) 8
Beam Position Interlock Layout • New IR 6 layout sees the addition of 4 interlock BPMs per beam ® ® ® 2 redundant BPMs added near TCDQ and 2 near preceding Q 4 90° phase advance to minimise chance of an unfortunate orbit bump BPM diameters have to be different due to integration constraints large diameters (80 mm & 130 mm) require the use of stripline pick-ups will use standard LHC orbit BPM analogue electronics 2 x 130 mm diameter stripline BPMs 2 x 80 mm diameter stripline BPMs Machine Protection Issues affecting Beam Commissioning - Rhodri Jones (AB/BDI) 9
Acquisition – Response and Threshold To take into account all scenarios need to look at worst case: • D 1 failure gives ~ 60 mm/turn at the TCDQ ® Response time required: 1 turn (detection) + 2 turns (BIC delay & abort gap synchronisation) 200 mm movement between detection and dump. • For a single pilot bunch of 5 109 protons the BPM system has a single shot resolution of ~1. 5% of half radius: ® ~ 300 mm for 80 mm diameter BPM (at Q 4) ® ~ 500 mm for 130 mm diameter BPM (at TCDQ) • Interlock threshold: ® set to 4 - 0. 2 - 0. 5 = 3. 3 mm to give effective threshold of 4 mm Machine Protection Issues affecting Beam Commissioning - Rhodri Jones (AB/BDI) 10
Acquisition - Detection • Modified LHC orbit digital acquisition card ® Direct comparison of position & threshold performed in FPGA no dependence on external software ® Autotriggered system no dependence on external timing ® Hard wired output to BIC • Open issues ® Limiting spurious triggers • Do we require a certain number of bunches out of limit before dumping? – For single or few bunches this will imply an increased latency ® Alignment and position offsets • Do we measure these with the beam or reduce threshold to include them? ® Sensitivity switching (at around 3 1010 cpb) • Direct link from intensity card? Machine Protection Issues affecting Beam Commissioning - Rhodri Jones (AB/BDI) 11
Interlock for Fast Beam Position Changes and Oscillations What needs to be protected? • Arc, triplets & collimators from asynchronous beam dump ® beam position relative to TCDQ needs to be maintained • Collimators from fast orbit changes • Collimators from oscillating bunches ® generated by instabilities such as electron cloud ® not seen by orbit feedback system Machine Protection Issues affecting Beam Commissioning - Rhodri Jones (AB/BDI) 12
Off-Axis Asynchronous Beam Dump at 450 Ge. V M S D Q 4 MKD d ti hin a DS TCDS ct Prote TC y b ed M S D by Protected Q 4 TCDQ W Beam Sweep -4 mm um h pc el n n a re u t r pe To Dump (TDE) TCDQ < 20 bunches r fo K O – re u rt e p Will hit arc a Oscillate & ejected at next turn M S D Machine Protection Issues affecting Beam Commissioning - Rhodri Jones (AB/BDI) 13
Protection against an Asynchronous Beam Dump Injection (450 Ge. V) – arc protection • • Arc aperture at 7. 5 s TCDQ will sit at around 7. 5 s Can accept up to around 20 bunches hitting the arc Fully protected by 4 mm beam position interlock 7 Te. V – TCDQ, triplet and collimator protection • TCDQ will sit ~0. 5 -1 s outside of secondary collimators ® avoid turning TCDQ into secondary or primary collimator ® beam position to be stabilised via orbit feedback to 0. 5 s (300 mm at TCDQ) • Require software interlock on orbit feedback system (slow ~1 s) ® will provide protection for most asynchronous dumps ® will not provide protection if there is a fast orbit change at time of the asynchronous dump (i. e. 2 simultaneous failures) e. g. D 1 failure + asynchronous dump Machine Protection Issues affecting Beam Commissioning - Rhodri Jones (AB/BDI) 14
Protection against oscillations & fast orbit changes Injection (450 Ge. V) – collimator protection • Arc aperture at 7. 5 s • TCDQ will sit at around 7. 5 s & collimators at 6 -7 s • 4 mm beam position interlock at < 2 s ® this already allows for orbit drift due to D 1 failure during dump request • Fully protected by 4 mm beam position interlock • Allows Q kicks at max kicker strength of 1. 75 s for centred beam 7 Te. V – collimator protection • Collimators will sit at 6 s • D 1 failure will result in collimator damage after ~3 ms ® loss of around 1012 protons • Orbit only moves by only ~1 mm during this time at TCDQ Machine Protection Issues affecting Beam Commissioning - Rhodri Jones (AB/BDI) 15
Protection against oscillations & fast orbit changes Can we reduce the beam position interlock to 1 mm at 7 Te. V? • only valid for nominal intensity beams ® where BPM single bunch resolution < 100 mm • will require a relative (not absolute) position threshold value ® i. e. 1 mm offset with respect to “stable orbit” • can be done by comparing current position to last orbit reading ® Global orbit reference updated every 20 ms ® Can compare single bunch position (for oscillations) & /or 1 turn orbit • can all be implemented in hardware (FPGA) with no external input • should be separate channel from the beam dump position interlock ® Can still have data treatment in the same FPGA Machine Protection Issues affecting Beam Commissioning - Rhodri Jones (AB/BDI) 16
Fast Beam Current Decay Monitoring Some orders of magnitude for fast losses (see talks by V. Kain and A. Siemko for latest figures) • At 450 Ge. V: ® Damage limit ® Quench limit ~6 1011 ~2 109 protons lost per metre in 1 ms • At 7 Tev: ® Damage limit ® Quench limit ~ 1010 ~ 107 protons lost per metre in 1 ms Things to keep in mind • Losses will normally be distributed over more than 1 m ® extra margin (e. g. pilot bunch considered below 450 Ge. V quench level) • At 7 Te. V collimators will be closed ® if correctly positioned, fast losses will impact first on collimators ® graphite can withstand > 1012 protons in 1 ms Machine Protection Issues affecting Beam Commissioning - Rhodri Jones (AB/BDI) 17
Fast Beam Current Decay Monitoring From R. Schmidt’s MAC presentation (Dec 2004) Monitoring a loss of 1011 protons within 1 ms: ® would fully protect the LHC at 450 Ge. V ® would fully protect the LHC at 7 Te. V • provided graphite collimators are hit first • possibly some damage of copper collimator surface ® would for many types of failure protect the LHC at 7 Te. V, even if collimators are not correctly positioned ® could reduce damage by more than 3 orders of magnitude Machine Protection Issues affecting Beam Commissioning - Rhodri Jones (AB/BDI) 18
What can we Reasonably Expect to Detect with BCTs? • DC Beam Transformer ® resolution of 10 m. A for integration over 20 ms ® equivalent to change of 1. 2 1012 protons in 20 ms ® can just about detect loss of 6 1010 protons per ms ® response time of 40 ms Would be pushed to detect loss of 1011 protons per ms Response time too slow to deal with such a loss Machine Protection Issues affecting Beam Commissioning - Rhodri Jones (AB/BDI) 19
What can we Reasonably Expect to Detect with BCTs? Fast Beam Current Transformer (using figures based on SPS system) • Characteristics ® 40 MHz integration ® 12 -bit bipolar digitisation (to handle baseline droop) • 2048 effective bits for 2 1011 protons ® rms noise of system around 3 bits per 25 ns integration window • 3 bits corresponds to 3 108 protons • Total Intensity (integration over 5 turns ~0. 5 ms) ® calculated by adding up all 3564 slots of 25 ns per turn ® noise limit √(5 3564) 3 108 = 4 1010 charges ® not much margin, but detection of 1 1011 loss in 1 ms possible • hope of improving LHC system based on SPS experience Machine Protection Issues affecting Beam Commissioning - Rhodri Jones (AB/BDI) 20
What can we Reasonably Expect to Detect with BCTs? Fast Beam Current Transformer • Advantages ® ® ® independent method to measure beam losses provides redundancy for BLM system can react within ms time scale will detect bunch core losses resulting in unbunched beam limited complexity (one instrument) • Open Issues (some can be studied in SPS during 2006) ® should be independent of bunch to bunch timing • will require active filtering of signal amplifiers & added noise ® 40 MHz integrators only integrate ~21 ns out of every 25 ns • how much error does this introduce on filtered signal? ® to be robust put in FPGA what is currently done in Power. PC – possible? ® influence of 50 Hz interference for integration times < 20 ms ? Machine Protection Issues affecting Beam Commissioning - Rhodri Jones (AB/BDI) 21
Summary 1 • LHC Beam Dump Interlock ® solution exists using modified LHC orbit system electronics ® will protect the dump line from damage for all scenarios ® will limit Q 4 quenches to dump extraction failure modes • Oscillation and fast orbit change interlock ® protection provided by 4 mm interlock at injection ® protection possible at 7 Te. V if 1 mm interlock is implemented • Fast beam current decay monitoring ® fast BCT seems capable of detecting a loss of 1011 protons in 1 ms ® provides redundancy to “complex” BLM system ® will fully protect the LHC at 450 Ge. V & for most cases at 7 Te. V Machine Protection Issues affecting Beam Commissioning - Rhodri Jones (AB/BDI) 22
Summary 2 • Other instruments linked to the machine protection system ® BCT for safe beam flag • probably the DCCT as this also measures unbunched beam ® Longitudinal Density Monitor • to protect against beam in abort gap – currently under development • Reliability & availability issues * see BDI disclaimer! ® The aim is to produce a redundancy in the diagnostics of machine protection without severely reducing the viability of the system as a whole ® unlike BLM system, however, none of the systems mentioned was originally designed to be used for machine protection ® currently no numbers for reliability & availability of these systems ® further studies required to make these systems as simple and robust as possible * The aim is to produce a redundancy in the diagnostics of machine protection without severely reducing the viability of the system as a whole. Unlike BLM system, however, none of the systems mentioned was originally designed to be used for machine protection. Currently no numbers for reliability & availability of these systems. Further studies required to make these systems as simple and robust as possible Machine Protection Issues affecting Beam Commissioning - Rhodri Jones (AB/BDI) 23
- Slides: 23