Minutes and Summary

Main purpose of the meeting: Preparing for first PbPb stable beams & 2025/2026 discussions

LPC minutes 4th November

Introduction (Federico Alessio)

(Roderik Bruce) Even for the first step there has to be a quick validation of the loss maps. This would need to be done in the middle of the night to go to the 22b step. Otherwise we can try to find another activity for the night but it is difficult. If the experiments want the 22b step then we can find someone and do the fill even at night. The 80b step, which is required by MPP, will be in the morning.

(Elena Dall'Occo) For LHCb we don't need the 22b fill as we can't close the velo.

(Catrin Bernius) Is the 80b step with trains? Yes. We think that the INDIVs in the 80b step can be used so we wouldn't need the 22b step, but we might also want INDIVs in the 250b step as well.

(Filip Moortgat) For CMS the 4b step is not required.

(Federico Alessio, Roderik Bruce, Catrin Bernius) To clarify, the 22b step will only happen if the previous activities are finished in the night. If the previous activities take until the morning then the 22b step will be skipped. The 80b will happen during the day. The 2nd loss maps fill needs luminosity, the first doesn't go to collision. Ideally the 2 loss map fills will only take 4h but could take longer.

(Witold Kozanecki) It has been posted on Mattermost that the BETS access will happen tomorrow. This makes it more likely that the 22b could happen in the night as it advances the timetable.

(Roderik Bruce) The other commissioning activities require day-time experts, but some MD preparation could be done if the 22b fill is not wanted so the machine is not idle.

(Filip Moortgat, Roderik Bruce, Catrin Bernius) Is Wednesday for the vdM during the checklist not getting optimistic? Yes, it is starting to become so. The ATLAS vdM can now take place after the 7th rather than the 9th. ATLAS doesn't want to do the LSC and vdM consecutively due to expert availability.

(Andrea Ferrero) When you say Thursday do you mean we are 24h later than the spreadsheet. It is more likely to be in the night.

(Andrea Ferrero, Roderik Bruce, Catrin Bernius) Do you need 2 fills of 5h each for the 450b step? No it is 5h in total, but at least two fills to stable beams. Will these be of equal size. There is no restriction on the length of the fills. The checklist must come after both so the vdM should come afterwards. The ATLAS LSC which should happen in one of these fills is around 3h so they need to be of unequal length. The time for the checklist is uncertain and depends on when it comes. If it is in the night it will take until the morning while during the day it will be done in a couple of hours. We only really need one experiment to do the vdM in this timeslot to keep things efficient though, but ideally two could get it done during this time, and could be required if the checklist is at night.

(Andrej Gorisek) Has the 80b changed from the one that was originally shown. Yes, this is because we want to test injecting 56b trains and the old one used 40b trains.

(Witold Kozanecki, Roderik Bruce) The cycles for ALICE polarity are they fixed for Thursday morning. No, these were a placeholders in case there were delays in the HI setup. If we did it at this time we would need to switch back and forward in polarity so if there aren't delays these will be done at the time the magnet polarity switch is done.

(Catrin Bernius, Roderik Bruce) ATLAS can confirm that we are also OK without the 22b step. Was this cycle setup note required by MPP? It was required without going to stable beams but this has already been achieved as part of the setup, but we haven't gone to stable beams as we can't without the loss map check.

(Andrea Ferrero) If we could be informed if stable beams are going to be declared via calls to the control rooms as far in advance as possible that would be great.

(Filip Moortgat) Wednesday is ok for the CMS vdM but Thursday is not possible due to the P2UG review that happens on Thursday.

(Federico Alessio, Chris Young) We consider ALICE & LHCb as a baseline for doing the vdM after the 450b step. If we are a little behind, such that the experts are available, then ATLAS/CMS can also do theirs afterwards before the VIP on Friday, or afterwards. CMS could also go first if we end up ready for vdM on Wednesday.

(Chris Young) Note that we have both TSs before the ion periods finish on Fridays which is not thought to be ideal.

(Catrin Bernius) Is there no contingency in case there is a problem like with the ZDC installation this year. It is very difficult as the TS cannot fall on a weekend which also means we couldn't advance the ion commissioning to two days before the TS like was done this year.

(Chris Young) It is possible to move 2-3 HI days between the years, and advancing 2 days of setup to get the TSs in both 2025 and 2026 not on Fridays but it isn't easy.

(Matteo Solfaroli Camillocci) There were some discussions on Friday of hybrid proposals for the optics. These involve changes of configuration in 2026 and give a compromise the choice between forward physics and magnet protection. However, there are also other considerations which have been brought up over the weekend that might limit the choices. This will all be discussed at the LBOC tomorrow.

(Catrin Bernius, Matteo Solfaroli Camillocci) Are the decisions on the optics and the schedule coupled? Not really, as the error on the damage limit is so unknown such that switching between 100/100 fb-1 and 120/80 fb-1 doesn't make a huge difference seeing as there are large uncertainties on the damage limit. If there is a single solution that is good for physics and also protects the machine. In the case where there is a good/bad optics for damage or physics then we will need to think about which is used for each year if we have to do one for each year.

(Roderik Bruce, Matteo Solfaroli Camillocci) If you consider the case where we run Nom-H in IP1 then this does imply a higher risk for the ion run if we put all the ions at the end of 2026. In general having it at the end has a higher risk. For D1 if we run nom-H then the dose is at a level where a preventative solution has to be found. The triplet at IP1 is also near touching the limit. There might be margin in the limit but no one will make this definite statement.

(Matteo Solfaroli Camillocci) For the nom-H it means that we need to find a preventative solution for the D1 magnet at point 1. There are ongoing discussions on the complexity and time for replacing it this YETS.

(Federico Alessio) From our side our "risk" was based on recovering lost time in 2025 by re-balancing days, and is based on the many possible issues, not solely issues relating to the dose on the magnets at IP1/5.

(Reyes Alemany Fernandez, Matteo Solfaroli Camillocci, Roderik Bruce) For the oxygen run next year do we have enough time to commission a new optics during the 4 days commissioning time. This would be slotted in during the general pp commissioning. This will therefore at least be partially completed before the start of the 4 day period.

pPb Performance Oral Update (Roderik Bruce)

1.2 pb-1 (ATLAS/CMS) and 0.6 pb-1 (ALICE/LHCb) were in the yellow book for 2 months across Run 3 & 4. Next meeting there will be more details but here is a brief preview. I've computed numbers for conservative/optimistic Pb and conservative/optimistic proton beams, giving 4 configurations. There are several machine studies that need to be done for the optimistic scenarios, particularly for the proton beam. There is also the idea of using 25ns proton beams such that all 4 experiments get a large number of collisions. There are many open questions if this is possible due to beam-beam etc. In 2026 there are 15 HI physics days. The projections are 13 nb-1 per day for ALICE, 18-25 nb-1 per day for ATLAS & CMS, 6-20 nb-1 per day for LHCb which is highly dependent on the configuration, while there is less variation for the other IPs. This means 360 / 200 / 300 nb-1 for ATLAS & CMS / ALICE / LHCb in the most optimistic scenario for 15 days such that the targets (ie. half the numbers for Run 3 & 4) would not be met by such a short run, other than possibly LHCb in the most optimistic case. This makes sense as we are considering much a much shorter time than the 1 month in Run 3 that the yellow report is based on.

(Ricardo Longo) 25ns collisions would be very difficult for ZDC. The collisions would still be at 50ns as this is the Pb beam spacing and there would be non-colliding protons in-between which could lead to beam backgrounds. This should be ok, but the backgrounds would need to be considered. It is likely that both would need to be setup in case of problems with the 25ns configuration.

(Andrea Ferrero) In these cases ATLAS/CMS are less than a factor of 2 higher than ALICE which was the assumption in the yellow report. Yes, this appears to be the case.

(Elena Dall'Occo) Do you account for the time taken to do the swap pPb/Pbp? No there is no additional commissioning time assigned for that and even without the switch the 4 days of commissioning is optimistic. This should also be considered in the physics reach in the time allocated.

(Jamie Boyd) Does the ion refill need to also be included? This is not scheduled but as the run is short we hope that this is not needed during the run.

LHCb (Elena Dall'Occo)

(No questions)

ATLAS (Catrin Bernius)

(Chris Young) Definitely try to do LSC in the first fill of 450/850b as the 2nd could be shorter than 3h. Where this is done will have to depend a bit on expert availability and the vdM scan position which requires the same experts. How much in advance do ATLAS need to let the other experiments know when we want to do this? We will work on a half day notice basis. There are no more blocking checklists which we should try to put vdM programs on top of.

(Federico Alessio, Roderik Bruce) For the leveling what value will LHCb level at? This will be 1.5e27 cm-2 s-1.

CMS (Filip Moortgat)

(Chris Young, Matteo Solfaroli Camillocci, Federico Alessio) We cannot really consider it fixed, there are other options that will be shown tomorrow. This is very fluid but we are trying to move towards a decision as quickly as possible. We don't want to say what is most likely. It is good that you are getting prepared and if it is not done this year then it might be done next year. The final decision will be made at the research board? Maybe it is more an ATS management decision rather than for the full research board. Also the LHCC might want to weigh in. Last year it was finally stamped in Chamonix but hopefully it will be quicker this year. What is said on Wednesday will be a very important ingredient. For CMS we have to go ahead assuming we will rotate otherwise it will be late or not possible, but it is not good if we ask for funding for these parts for the rotation and then it doesn't get implemented.

(Chris Young, Federico Alessio) Providing a number of 7 nb-1 is very useful, although this is a very large. Could you also provide a pPb number where the dataset would become interesting to CMS. This is being looked into and can be announced later but something like at least the size of the 2016 dataset is likely. This was 200 nb-1 such that the numbers shown by Roderik mean that this might be possible within the current number of HI days in 2026.

(Federico Alessio, Catrin Bernius) Could ATLAS give a similar statement to the CMS one? There is no interest in pPb, ATLAS just wants as much PbPb as possible with no upper limit when we would want to switch. This can be checked again with management as no upper limits were discussed.

ALICE (Andrea Ferrero)

(Michi Hostettler) Where exactly it comes from still needs to be confirmed, but perhaps the knob factors are not perfectly matched, in addition due to the lower gamma the beams are bigger and the trims are bigger. It looks like there was some orbit leakage every time things were changed.

(Witold Kozanecki, Michi Hostettler) Does this length scale follow the ATLAS or CMS method? This is a scan at constant separateion. If the length scales of beam 1 and beam 2 are different then you will get this effect. This is another possibility. If the effect is the same as in the real vdM then you will correct for this with the length scale correction. Otherwise you will need a systematic to cover this. It will require some more offline analysis.

(Chris Young, Federico Alessio) If we ended up doing it next week then this is not a problem as long as LHCb does it during the checklist, but it is good to get these out of the way. We will keep in touch with all experiments about when to schedule the different scans. We don't want to interrupt a scan for the VIP visit so could do an intensity ramp-up fill if this will be the case. We do know when the visit will end so this is an obvious timeslot for one experiment.

(Andrea Ferrero) We are investigating if we would profit from collisions tonight, and whether this would speed up the readiness for the vdM from the ALICE trigger point of view.

(Catrin Bernius, Federico Alessio) Can the spreadsheet be made more realistic? We can try but some items are less/more than 1/2 a shift making it difficult. There will surely be more discussions tomorrow morning. Things can also go up to a factor 2 faster, or infinitely slower, so each day we know more about what will be happening later in the week. Helga is giving regular updates on Mattermost.

(Witold Kozanecki) For Michi/Jorg, is it ok for all experiments to go to 4.5 sigma per beam in the scans? This will be checked in the next fill but it is hoped that this will be fine. For PbPb it is the same for the LSC scans and the vdM fills as it is the same optics for PbPb. This is not the partial separation of the non-scanning IPs during the vdM at the non-scanning IP to 25% of the luminosity, but how far the scan can go in the scanning IP, clarifying the question posed.

(Chris Young) For ALICE are the background results looking ok? They are still being analyzed, but the rate is low, but the source of the components is still under study. More studies with beam are not planned for the moment.