Login

Joint BLM Thresholds WG and Collimation WG meeting (TCLD, IR3, TCT)

→ Europe/Zurich
6/R-012 (CERN)

6/R-012

CERN
Belen Maria Salvachua Ferrando (CERN), Daniele Mirarchi (CERN), Sara Morales Vigo (CERN)
Joint BLM Thresholds WG and Collimation WG meeting
Zoom Meeting ID
65689083492
Host
Sara Morales Vigo
Useful links
Join via phone
Zoom URL
Hide

Participants (remote and in person): Ludovica Bertolozzi, Enrico Bravin, Roderik Bruce, Nuno Duarte, Jorge Guardia, Birk Karlsen-Baeck, Anton Lechner, Daniele Mirarchi, Sara Morales, Belen Salvachua, Dora Veres

Power deposition from TCLD pressure spikes (TCLD.11R2 beam-gas simulations) - Nuno Duarte

  • Pressure spikes in TCLD.11R2 (installed in LS2) observed since 2022, occasionally triggering beam dumps
    • FLUKA simulations performed to evaluate risk to downstream magnets from beam-gas interactions
  • Analysis of fill #11164 --> Dumped on pressure spike by vacuum, not BLMs
    • Vacuum pressure reaching 4e-6 mbar --> ~5s with at least one gauge over threshold
    • High pressure for >140cm
    • BLMs did not dump but can see losses due to the pressure spike --> 6 BLMs on Q11R2, TCLD.11R2 and MB12R2
    • Max RS09 BLM signal used to benchmark simulations
  • Beam-gas FLUKA simulations --> Because of the low probability, inelastic interactions are forced for every primary and then normalize to pressure.
    • Location of inelastic interaction randomly sampled along particle trajectory
    • Energy deposition scored on BLMs (benchmark), MQ and MBs downstream of TCLD.11R2 (power deposition on superconducting coils)
    • Accurate model geometry around TCLD
    • Simplified gas profile in interaction region --> H2 with constant pressure
    • Scale the fluka results and then investigate the factors wrt BLM signals --> Good agreement between simulated and measured BLM signals, but fitted normalization factor is ~3.5x higher than the analytical one
    • With a gas mixture (used for FCC studies), the normalization factor is just 1.4x higher than the analytical one
  • Energy scoring from the N2 simulation (more conservative)
    • Normalized for pressure and intensity for the particular moment that is taken 
    • Energy peak happens in first dipole, but very modest (0.33 mW/cm3)
    • For simulated losses with 4e-6 mbar, the estimated maximum power density in SC coils remains well below quench limits
      • Maximum vacuum pressure before quench expected at 4e-6 mbar x 30  = 0.00012
      • Even with 1e-5mbar we will not quench
  • Daniele comments that from empirical scaling there should not be a need to increase BLM thresholds
  • Belen comments that this factor of 30 is applicable to all BLMs in the area
  • Daniele says that even increasing the maximum vacuum threshold we are well below the BLM thresholds
  • Belen asks if similar simulations for the beam pipes are foreseen, for the case of vacuum spikes in bellows for example
    • Anton says that it depends strongly on the location, and 6L2 was exchanged already
  • Daniele asks about the gas mixture composition 
    • Standard mixture used for FCC studies, for LHC it should be different compositions
      • Would it change much the simulation results?

 

Validation of updated IR3 BLM thresholds - Sara Morales Vigo

  • Off-momentum losses at the start of the energy ramp reaching BLM thresholds in IR3 during some fills in Run 3
  • Loss plane decomposition algorithm developed to quantify sharing of losses between IR3 and IR7
  • Main limitation in 2024 proton run was in the Q6L3 BLM RS10 (~5s), limiting at 20kW --> Applied threshold
    • Threshold increased by a factor of ~2 as a temporary measure before further studies were completed --> Changes on 25/06/2024, limiting at 60kW 
  • IR3 BLM thresholds model not reviewed since LHC LS1, at the moment no way to disentangle IR3 and IR7 off-momentum losses
    • BLMs at TCPs and TCSGs master threshold set to 500kW (IR3+IR7 losses), with MF = 0.4 (applied threshold at 200kW)
    • Proposal to update the IR3 BLM thresholds collimation model similarly to the IR7 collimation model that was implemented during LS2
  • Collimation model requires:
    • BLM response --> Calculated from dp/p loss maps
    • Allowed number of charge impacts on the TCPs --> Up to 500kW losses at injection probed in 2025 proton MD, ~100kW losses limit expected at top energy (TCLAs)
  • BLM thresholds in IR3 updated during the YETS 25/26
    • New model for IR3 collimators
      • 5 new families
      • Master thresholds --> 500kW at injection energy, 100kW at top energy
      • MF  = 0.4 --> Applied thresholds at 200kW injection energy, 40kW at top energy
    • Q6 and warm magnets with corrections on top of the existing models to allow 500kW losses with the master thresholds, top energy untouched
      • MF = 0.4 -->  Applied thresholds at 200kW injection energy
  • Validation of new IR3 thresholds from scaling BLM signal-to-dump ratios from:
    • Off-momentum loss maps at injection and top energy
      • Only interesting when enough losses in IR3 --> B1 losses often very low or in IR7
      • Power loss in IR3 calculated from BLM decomposition algorithm
      • B1/B2 BLM ratios scaled to 100%, equivalent B1/B2 power loss at dump calculated
        • Crosstalk between beams affects the analysis, specially when one beam loses more than the other
      • Overall OK, but hard to validate
    • Start-of-ramp losses
      • For fills up to #11474, for B1 the main power loss is in IR3, later in IR7
        • In some cases, both beams with activity in IP7
        • Daniele comments that the switch to IR7 losses for B1 happens in the first fill of 75 bunches
      • Collimator hierarchy still OK in IR7 for B1, but secondary peak of losses appearing clearly
      • Not enough signal to scale for fills with power loss under ~1kW --> Only B2 with losses in IR3
        • B2 seems well aligned to 500kW for master threshold
  • Will continue to monitor with start-of-ramp losses and adapt if needed

 

AoB: Proposal for increase of BLM threshold at TCTs during the ramp - Daniele Mirarchi

  • TCTs going to much tighter settings during the ramp this year due to the combined squeeze
    • Combination of loss increase + threshold going down during the ramp (before flattop corrections) --> Some TCTs in IP1 and IP5 with signal to dump threshold above 30% in RS11 and RS12
  • Factors needed to avoid being in warning with full machine
    • TCTPV.4R1.B2 RS11 higher than 1.2, RS12  higher than 2.1
    • TCTPH.4L5.B1 RS11 higher than 1.4, RS12 higher than 2.6
  • Check of power loss allowed on TCTs for given thresholds 
    • TCT BLM response measured in 2025 at injection and top energy --> Good agreement with FLUKA simulations
    • TCT jaw is designed for a maximum power load of 2kW --> Calculate equivalent power for different thresholds --> Conservative as not all the load remains at the TCT
      • If Flat Top corrections propagated down to lower energies --> 3.8kW
      • If RS11 and RS12 propagated to RS10 --> 0.2kW --> Acceptable, safe increase --> From EL21 (5 TeV)
        • Sufficient margin to operate the machine until the TCT thresholds model is reviewed --> Plan for LS3
There are minutes attached to this event. Show them.