Minutes and Summary

Main purpose of the meeting: PbPb data-taking progress - Planning MD block and polarity reversal News from the LHCC 2026 questions

LPC minutes 24 November 2025

Present (P = in person): Chris Young (P), Chiara Zampolli (P), Roderik Bruce (P), Andrej Gorisek (P), Eric Torrence (P), Robert Muenzer (P), Filip Moortgat (P), Ivan Calì (P), Krjstian Roslon (P), Matthew Nguyen, Stephane Willocq, Flavio Pisani, Matteo Solfaroli Camillocci, Riccardo Longo, Joanna Wanczyk (P), Maciej Trzebinski, Anna Sfyrla, Dragoslav Lazic, Marko Milovanovic, Michi Hostletter (P), Tomasz Bold, Witold Kozanecki, Xavier Buffat

Introduction (Chris Young)

Filip Moortgat: is the worse intensity due to the power cut, maybe they [LEIR] had to restart something?

Chris Young: there are some issues in LEIR and they are trying to retune everything. There was a problem in LEIR, due to a magnet that was down but it was not reporting that it was down. They practically have to recommission LEIR. They think it can come back, but they don’t think that for the next fill it will be significantly better. 

[For intensities of 1.5, the optimal for ALICE is 6h, more than 7]

Roderik Bruce: for the ALICE polarity, we will stay in +/+ only for the first Beam-Beam MD, then we will switch back.

Robert Muenzer: we can also provide luminosity for the +/+.

Chiara Zampolli: what was the reason why the MD was moved by a week?

Chris Young: because the injectors would have not been ready for all the beams needed, and they needed more time. 

Roderik Bruce: especially for the last MD. 

Filip Moortgat: wed afternoon you will have stable beams, 280 bunches with 25 ns the shortest.

Chris Young: some will be 25, some will be 50, then there are some gaps. 

Eric Torrence: what will be the highest intensity?

Roderik Bruce: we don’t know. There will be two different types of trains, one with still 50 ns space but shorter trains, and we can hope to have a factor 2 higher intensity in the SPS. Then we have the transmission through the SPS, which might be significantly worse due to the higher intensity. We have to see how much will survive to collision in the LHC, which is part of the MD. Then you have the other type of trains where you have a mix of 25 and 50 ns in the trains, for which it would be good to have feedback from the experiments to see how bad this is, with shorter 25 ns spacing, if it is a showstopper, or if you could live with it after some work. This MD might potentially guide any upgrade in the PS. In principle there is an upgrade in PS that would allow a 25 ns spacing between all the bunches for Pb. If the experiments are not interested, we should know to not do this upgrade. This would be for Run 4, or maybe even Run 5. Surely not next year. In addition, for the first MD, the beam-beam one, we would like to have a good luminosity measurement from all the experiments. Will it be an issue for ALICE for the ZDC?

Robert Muenzer: we switched to using FT0 for the lumi, and for this the xing angle won’t matter.

Roderik Bruce: from FLUKA studies it seems that the SND bump could have some potential negative implications. To be followed up. 

Filip Moortgat: [concerning the proposal for the separate low mu runs for ATLAS and CMS] we were hoping to have very long fills at lower intensity, like the pp reference runs, to be impacted less by availability, which is very much affected by intensity.

Chris Young: but then LHCb will drop out of leveling. 

Filip Moortgat: not necessarily. 

Eric Torrence: they would not have the burn off. 

Michi Hostletter: the burn off should go down by 80%. 

Filip Moortgat: I disagree with the proposal. I think all the numbers [slide 9] are comparable within uncertainties. What we gain by having a common lower intensity beam with slower leveling is long and stable fills, without operational issues, which we have with the high intensity. So we were hoping for smooth low intensity and then PbPb.

Chris Young: ATLAS already expressed the wish to not have this at the end of the pp run. 

Michi Hostletter: even with 1.5e11 you have almost unlimited leveling time if you have mu 3 in ATLAS and CMS. After ~39 h one could even take a step in b*, till then you have so much power in separation, which is not true if you want to keep one experiment high. In this case you would need b* steps which means some percentish up excursions, which you don’t have in case you run together. It becomes a little bit less smooth. 

Filip Moortgat: we should see how this might affect LHCb. 

Michi Moortgat: if you level at your usual target, you would be levelled together with them, since none of ATLAS and CMS will burn the beam. The intensity will drop just by ~20%, because the two main consumers are leveled down. Here we’re talking of 40h. LHCb would become the main consumer of protons. 

Flavio Pisani: this will pose another issue: in more or less a couple of weeks we’d fill the buffer between HLT1 and HLT2. We’d need to split the data taking in two, with either an MD of the TS in between to give time to empty the buffer. 

Eric Torrence: from ATLAS there was not a strong request to have the low mu at any specific moment in the timetable, but we’d like to not put it straight at the end of the pp running to make sure that we collect 1fb-1, so in case of problems we can extend it.

Filip Moorgat: I will check but I think that there was not really a strong preference from CMS when to have it. 

Fili Moortgat: for the 2026 timetable, the idea is to have the MD, then the 1TeV run then the VdM, right?

Chris Young: yes, this way we will not constrain the MDs requesting low intensity.

Roderik Bruce: for the light ions: we’ll not do them in 2026, and I don’t know if there will be any request for Run4; if so, we should think in advance about which ones, so that we can test in the source during the LS, and we can prepare with the experts. 

Chris Young: on Friday at the Research Board, I will present that this was discussed but that it was not preferred for this year even if there is interest for the future, plus that ALICE would like to declare this as special runs.

Roderik: it also came out that the proposal of the second source which is not yet approved by the CERN funding, it would be good to mention that the experiments support it, if they do. We should understand how to express our support.

Chris Young: it could be expressed in the spokespeople meeting, but this would also mean that the experiments support light ions.

ATLAS (Eric Torrence)

Michi Hostletter: [concerning going down in pileup at the end of the ~next fill] This is 10x smaller than usual, so it is good to prewarn, but it should be fine. We would just need to restart the leveling because by then we would be out of leveling. 

Chris Young: will you run these [see slide 3] triggers next week with the MD with the different filling schemes?

Eric Torrence: yes. They are trying to tune them a bit to work with 25 ns (some already worked with 25 ns). 

Roderik Bruce: in the slide with the background, the plot is from HI. This might be a different problem than in pp.

Eric Torrence: yes, this is peculiar to PbPb, and it seems to be conditioning.

Roderik Bruce: for pp it is the muon background from the outgoing collision products, but if you see in the BCM then it is something else, like beam gas. 

Eric Torrence: yes, and it correlates very well with the vacuum, so it is probably some beam-gas. 

Michi Hostletter: is it at the end of leveling?

Eric It starts before the end of leveling and then it continues for a bit. 

Michi Hostletteri: in the first two fills at the end of leveling there were some issues with CMS, but since it was in other fills too, it was not that.

 

CMS (Filip Moortgat)

Chris Young: if you do the low mu together, considering the comment from before from LHCb (to split the low mu run in 2), you probably don’t want to have the MD, 1 TeV run and VdM in the middle of it. In case you want to have a consistent dataset.

Filip Moortgat: I am not sure it is a strong argument. 

Chris Young: we could put 3 weeks before the MD and then after the VdM more if needed.

Filip Moorgat: I think that there is flexibility.

 

ALICE (Robert Muenzer)

Michi Hostletter: we had a big shift that was affecting all the experiments, and this was traced to an intervention on the RF to make VXCO more stable when we had this problem with the unlocks. As a side effect thi changed the phase, and it is not particularly corrected for. Now we are more or less but maybe not quite where we were before. We corrected it such that ATLAS that was before at + some millimeters is now perfectly at zero, so we did not come back all the way. The question is that if everybody is fine with it, we could move a 5 mm more towards 0 for ALICE. Then ATLAS will be a bit 5 mm in the other position or we could do a compromise of 3 or 4, something like to center the average of all. We did not go back completely after the RF intervention, also because for HI it is a slower process, because the lumi region data, in particular for LHCb and CMS, is updated with a significantly lower rate than for proton. It was done overnight, one step every 30 minutes, and we did it very gently. For proton we have a slightly higher feedback rate so it is easier. Here we could shift a bit more if it helps ALICE and the others don’t care so much at a level of a few mm.I think that also LHCb would move in the right direction, for CMS it is hard to tell at the mm level since they have some noise in the data, while ATLAS would slightly overshoot a bit. But before the RF issue, ATLAS was a bit on the other side, and they would be moved back there.

Andrej Gorisek: so the proposal is to move by 5 mm?

Michi Hostletter: maybe a bit less. Note that there is also some oscillation during the fill. 

Robert Muenzer: a few mm would already help, to not be immediately at the beginning of the fill at the edge. 

Michi Hostletter: shifts of this order of magnitude are transparent for us. So just let us know.

Chris Young: let’s follow on Mattermost.

Roderik Bruce: the instability of the TPC, is something that you might fix during the YETS? Or will it be there for the rest of Run 3?

Robert Muenzer: if there is nothing we can do during winter, it is something that will be there also next year, and it would also affect pPb since it is load related.

Filip Moortgat: is it something that you are considering to do during winter?

Robert: yes, even if it is tricky, accessing the space is very difficult. For sure we’ll try.

 

LHCb (Paula Collins)

Michi Hostletter: for the beam position [in y] that changed during PbPb: 150 um for the big jump, which is between pp and PbPb is possible due to orbit correction that is a little bit different; the small jump (20 um) might be some triplet movement, e.g. caused by some cryo issue (even if in sector 12, and not directly in sector 18), which indeed we had, but we should check if the dates and events coincide. 

NB: the fill number on slide 7 should be 11318

Roderik Bruce: over time, the shape of the plot on s 7 is constant?

Paula Collins: yes, I think so. I would need to cross-check, but I think so.

Filip Moortgat: did you have it last year?

Paula Collins: a bit, but not like this. It could also be a material distribution effect since the foil moved in the H plane compared to the VELO, compared to where it was last year, because the shims have been removed but the VELO has retracted. We can’t be sure what this is.

Chris Young: B1H had the worst cleaning performance?

Roderik Bruce: B2H. But I checked back, and as expected we had to run with tighter TCTs to the smaller b* at LHCb and we had >1 order of magnitude higher losses on the TCTs in P8 this year in the loss maps, in B1. 

Chris Young: it has to be in Beam 1 otherwise you are protected from the rest of the LHCb.

Roderik Bruce: if I compare the losses, we have probably something like a factor 50 higher than last year. It is still small in absolute value for a TCT (we are nowhere close to where ALICE was in 2023, for example). Last year it was really low, and this year it went up by a factor 50. 

Paula Collins: could this be any type of particles?

Roderik Bruce: it could be some ions fragments that impact on the block of tungsten of the collimator. It could be Pb208, Pb207, or others fragments coming from IR7, then depending on how they impact on the collimator, they could completely break up and then you would have other secondary particles, or you could have other fragments scattering out, and if they strike only a corner of the collimator and they traverse only a small part inside of the material then you could have other fragments coming out of the collimator.  

Paula Collins: which energy would they have?

Roderik Bruce: it would be close to the beam energy. They would lose by ionization, but if it is a short part, it is not a difference of factors. But this we would have seen in the background test if you had closed the VELO. 

Chris Young: the losses in protons are presumably higher, so it has to be a different type of particles. 

Paula Collins: the only thing that could be different is the type of particles. 

Chiara Zampolli: could we go back to the b* of last year?

Roderik Bruce: this is possible but we would need to re-setup the cycle, do loss maps… It would take 1 shift at least. Then also the intensity ramp up might be needed. 

Chiara Zampolli: maybe LHCb is not interested in this, since the would have to put back the VELO and it could be dangerous. What about next year? 

Paula Collins: we cannot test much if we risk to break the VELO. We are trying to find a solid diagnostics that avoids breaking the VELO.

Roderik Bruce: we could do a test with lower number of bunches for which the machine would be qualified like we did for ALICE in 2023. On the other hand we usually don’t move around the collimators during Stable Beams. E.g. we could try to slowly open the TCTs.

Paula Collins: we should be confident enough to not break anything. 

Chiara Zampolli: wasn’t there anything in the data quality that was hinting at the fact that something was happening?

Paula Collins: no.

Roderik Bruce: the funny shape you see also with the VELO retracted?

Paula Collins: yes, we still see it.

Roderick Bruce: one could see if with low intensity the shape changes. 

Paula Collins: we have the shape since always. 

Chiara Zampolli: is the shape worrisome?

Paula Collins: it depends on what the red excess is. If it is damaging particles, it is worrisome. And last year we had the same but it was much less. 

Roderik Bruce: it would be good to think of some tests that we could do this year while we have ions, for next year.

Chris Young: one test can be to separate the other experiments to see if anything changes, CMS had some data separated and ATLAS will have. Or use the VdM fill where the others were separated. The hypothesis is that this halo comes from collisions in another IP or from IR7. 

Roderik Bruce: ATLAS and the experiments can drive the losses in IR7. If you separate the other experiments, the losses in the whole accelerator will go down. They could come either directly from ATLAS or from ATLAS via the IR7. The losses in LHCb could come from any experiment. With the VdM we could see all the different IPs separated, and you can see if anything changes.

Chris Young: obviously we need to follow up on the issue with the ASICs also to know what to do next year, whether to start with the VELO retracted; we can run simulations to see where these ions are coming from, what energy they have…