Go to the Indico Home Page Oldest event Older event Up to category Newer event Newest event Printable version
  • Compact style
  • Indico style
  • Indico style - inline minutes
  • Indico style - numbered
  • Indico style - numbered + minutes
  • Indico Weeks View
    • Switch to the management area of this event
    Login

    LIU Beam Parameter WG meeting #11

    → Europe/Zurich
    874/R-018 - MEETING ROOM PRIORITY BE OP (CERN)

    874/R-018 - MEETING ROOM PRIORITY BE OP

    CERN

    20
    Show room on map
    Hide

    Follow up from last meeting

    • The beam parameter tables for protons were reviewed
    • It is decided not to refer anymore to the 'LIU reach' parameter tables, but rather stick with the parameter tables obtained from the HL-LHC target values (and propagated back to the different accelerators of the injector chain using the beam loss and emittance growth budgets) and indicate which the highest risks are that some numbers cannot be achieved even after the LIU upgrades
    • After the analysis presented at the last meeting, it seems clear that:
      • There are a few critical points on which it is necessary to follow up constantly with machine studies but which do not appear to potentially become serious showstoppers (e.g. space charge in the PS and relaxed longitudinal parameters to ease it, longitudinal instabilities in the SPS)
      • The two remaining high risk items are the PS intensity limitation due to longitudinal instabilities and the losses at low energy in the SPS. Being the two problems correlated (the PS instability threshold could be increased by using larger longitudinal emittances, but these would only increase losses in the SPS), there is a combined action to tackle them and possibly identify a common strategy (which may imply including other unforeseen items in the upgrade plans)

     

    Ion beam parameter tables (Hannes Bartosik)

    • A full analysis was conducted using the data and the outstanding outcome of machine studies in 2015 to make an educated guess on the future achievable beam parameters for the ions in the post-LS2 scenario. In particular, realistic beam loss budgets were established along the chain (20% in LEIR, 8% between LEIR and PS, 8% in PS, 16% between PS and SPS, according to a nonlinear transmission curve in the SPS), slip stacking was included, leading finally to 1152 bunches with 1.7e8 ions/b and 1.3 um emittance in LHC. While this values would lead to an integrated luminosity ~25% lower than the HL-LHC target value, the discrepancy would already reduce to ~15% if the LHC performance efficiency was assumed to be 62% (like in 2015) instead of the assumed 50%. We should also observe that in the evaluation of the LHC filling scheme we still used 900 ns for the MKI rise time and 3300 ns for the AG keeper, while these values can be reasonably updated to 800 ns and 2900 ns, respectively. By doing that (shown offline after the meeting), we find that 1280 bunches could be injected into LHC using 16 injections from the PS to the SPS. This would provide an additional ~10% gain, which would eventually bring the luminosity reach of the LIU scheme within less than 10% from the HL-LHC target (for an LHC performance efficiency of 62%).
    • To be noted that most of the upgrades in Linac3 and LEIR injection - relevant to performance - should be already in place now (source and Linac3 upgrades, higher injection rate into LEIR). While it may take a while and intensive machine studies to learn how to use these upgrades to their full optimisation/exploitation, the beam parameters at the exit of LEIR (and also PS, which does not have any upgrades for ions in the baseline) should go already in 2016 in the direction of proving the LIU assumptions or finding bottlenecks.
    • In 2016, 20% more intensity has been available in average from Linac3 (due to the removal of an aperture limitation at the entrance of the solenoid), presently the two frequency heating is being tested, which is expected to lead to an additional ~10% increase.
    • The destructive resonances have been identified in LEIR. Using larger longitudinal emittances together with a better (empirical) compensation scheme has already led to reaching the LIU target intensity out of LEIR - and even exceeding it by >10% for some time. However, the reproducibility of the machine needs to be improved, as restoring the optimum beam conditions after some days of no tuning is still long and cumbersome.
    • The benefit of the faster injection rate into LEIR hasn't been proven yet. Injecting 13 times spaced at 100ms has hardly led to higher intensities, even with double current in the electron cooling to double the cooling rate. On the other hand, using the faster cooling rate with the nominal 200 ms spacing between injections has led to record intensities accumulated in LEIR. This however did not transfer into higher intensities extracted from LEIR because these high intensities were found to suffer more strongly from space charge at RF capture, with consequent large losses. It would be desirable to try an intermediate closer injection spacing, i.e. 150 ns, but this is for now incompatible with the source pulsing (10 Hz) and requires a dedicated MD planned for 2018.
    • In the PS, the nominal production scheme based on bunch splitting has been already successfully reproduced. Machine studies in the SPS have not yet started (only set up on EARLY beam so far)
    • Following the experience of 2015-16, the main points that need to be addressed for the ion operation in general are the options to improve the reproducibility of the source, the drift of parameters at the LEIR entrance (also depending on the stripping foil), the maintenance of the LEIR performance level. On the SPS side, it is well known that slip stacking is a high risk item because of the lack of experience at CERN and the tricky implementation of the needed features in the LLRF of the 200 MHz RF system. 
    • An option has been studied on paper to have batch compression to 50 ns at the PS flat top, which would allow 25 ns in the SPS after slip stacking and finally accumulating ~50% more bunches in LHC (the experiments have confirmed that this spacing is fine with them, confirmation for the ALICE crossing angle is pending on accelerator physics side). Assuming the same single bunch parameters as in the LIU baseline scenario, this would immediately project the attainable luminosity to ~40% more than the HL-LHC target value. Besides, this could be a mitigation in the case in which slip stacking in the SPS does not work (or comes with a delay), because it would allow to get anyway to 80% of the LIU performance. However, the success of this option is far from granted, as it would need to be first proven with dedicated machine studies and detailed simulations. If proved feasible, it would certainly require additional hardware (a broad-band cavity in the PS in the range 10-20 MHz and an extra ferrite tuner for the 40 MHz cavities), additional resources (~3 MCHF and important manpower in number of FTEs), and the post-LS2 timeline would be quite tight.
    There are minutes attached to this event. Show them.