![]() |
LPC meeting summary 20-01-2025 - final |
![]() |
Main purpose of the meeting: Rehearsal for Chamonix, 2025 schedule, pO, OO, NeNe
LPC minutes 20th January
Present: Chiara, Chris, Catrin, Robert, Silvia, Paula, Silvia, Roderik, Michi, Giovanni, Jorg, Witold, Andrei Gorisek, Gerardo Vasquez, Giovanni Cavallero, Hiroaki Menjo, Klaus, Monig, Lorenzo, Bianchi, Manuella Vincter, Oscar Boente Garcia, Peter Steinberg, Qipeng Hu, Reyes Alemany Ferannaddez, Riccardo Longo, Rosen Matev, Thomas Boettcher, Witold Kozanecki, Brian, Cole, Ivan Amos Cali, Flavio Pisani, Andres Guillermo Delannoy Sotomayor, Filip, Andrea, Dragoslav Lavic
Chamonix talk (Chris Young)
No minutes taken.
2025 schedule, pO, OO, NeNe (Chris Young)
S6:
Jorg: Some preferences seem to be for 2B (vacuum). It is not scientific (it is more “Let’s integrate first and take risks later”), in the sense that the same risk is there in September of July. The vacuum modules that were consolidated now have huge margins, but there are the unknown risks.
Catrin: why do you advance PbPb by so much in scenario 3?
Chris/Jorg: to have time to do two steps with enough time. You don’t want to just do two fills for each step.
S8:
Check the comments by Witold
Witold: related to the intensity ramp up, whether to do it between the special runs and the VdM or after the VdM: for the VdM, it is not optimal from the ATLAS point of view to have the intensity ramp-up before the VdM.
Chris: not the whole intensity ramp up will be done, but the 75 bunch fill at least, that would be before. That would be one day of the intensity ramp up, one fill with 3 bunches, and one with 75. We’ll not put the full machine before the VdM.
Witold: Argued in the past that the VdM with 150 bunches could count as a ramp up step.
Michi: but it is a different cycle.
Chris: in 2023, after a TS, there was s 3 bunch fill, and then a VdM
Catrin: this because the ATLAS solenoid was down, and it was then a special agreement with MPP.
Chris: probably important to tell the injectors that we want O on a Saturday (in case the intensity ramp up is somewhere not at the beginning)
Jorg: strictly speaking, the TS ends on Friday afternoon, so O might be needed then, but we can do setup with p
Roderik: injectors said that we’ll have single bunches on Saturday with pilots. We’ll want single bunches anyway but maybe not the full injection scheme, with not the fully optimized intensity, can be just some low intensity pilot.
Reyes: when should the collimator MD (which is planned with O beam) be?
Roderik: makes sense to have it before, so that we can use it
Chris: do we prefer to use it for both pO, OO, or one of the two? Because you need to redo the collimators between the two.
Roderik: if we do crystals, we probably need to redo them because the orbit will be different. If we do standard collimation, probably you won’t have to do it, but with crystals with both you have to do it both time, and you will lose commissioning time. This is why I did not say if it is to be done before or after: this point is not clear.
Reyes: if we do it after, if we have then also Ne, we can have the collimation MD with the two ion species, one after the other. One fill for O, one for Ne.
Roderik: I am not sure it is easy. You need 8 hours per MD, and you need the beam to be there for 16 hours and you need to guarantee that you can switch to Ne in 8 hours. You cannot take a new beam if you dump. You probably want to take a fresh beam after checking some things at injection. So maybe this is not a good plan. We cannot be sure that injectors can switch to Ne in 6-8 hours.
Chris: this is a decision for the collimation team and the MD coordinators. We just need to be aware that their days are hanging around somewhere.
Roderik: I personally don’t think that it is a good idea, because you run the risk that the Ne MD is drifting if the beam does not come as quickly as we hope, and the same team has to stay there for 16 hours.
S10:
Chris: 8h @ 0.01 (mu): is this in total for LHCf, or each of the detector positions needs 8 hours?
Hiroaki: 8 hours in total for LHCf @ 0.01 (then 0.03)
Chris: You said 200 ns is allowed for LHCf. Is it better to have things further apart? We can have 2 us or 200 ns. Is there a preference?
Lorenzo: we can go to 200 ns. We can run up to 200 ns.
Reyes: do you mean bunch spacing?
Chris: yes.
Reyes: the minimum that you can assume is 200 ns.
Chris: is ALICE flipping polarity during pO and OO?
Robert: to be checked. Anyway we are head-on.
Roderik: If you have the TCT, you might want to realign the TCTs. This is true if we change the external crossing to keep the net zero crossing, when we change polarity, we might need to realign the TCT.
Chris: no leveling in the pO run for all experiments but P1 (which required 0.01 and 0.03). Pileup is generally quite low in pO because we have 70 bunches.
Robert: so changing polarity might need more commissioning time?
Roderik: yes, unless you want to run with 0 external angle, but then you have +/-72.
Robert: ok, then it is to be checked.
S11:
Chris: everybody should tell us if they need to be levelled. What is the min x-angle for ZDC?
Catrin: 150 +/- 20
Brian: for PbPb the nominal range is [130, 170]. For pO the angle is determined for us (LHCf). For OO we can do 0, with special setup, but we can do it; if this is decided not to be done, we’re back to 150 urad which is the design of the detector.
Roderik: this is important (in addition to the geometrical factor) also because with zero crossing you can compensate the aperture loss from the lower energy.
Bruce: yes, we can do zero crossing.
Jorg: we have to change the crossing in the ramp, then.
Roderik: yes, or at flat top.
Paula: the pile up in the table is for how many bunches?
Chris: 6; so the pileup can go down if we do more bunches.
S12:
Roderik: we might want to take several bunches per injection. Like for PbPb last year we took 3 bunches per injection. In each SPS injection, you take 3 single bunches from the PS, and you can inject 3 bunches at once. They can be less than 1 us apart. They cannot be too close because of the SPS injection kicker.
Michi: we also have to see that they fit our kicker, but this is for pp. For the scenario 2B, if we try to run with 5x36, the injection kicker lines (which is the train length) will be longer than last year. Last year already with 3x36 it was quite limiting, because you want to tailor the injection kicker length, so you can move the abort gap keeper further out, and if you have to change this for a few days of running, it is a bit an overhead, because you need to revalidate the abort gap keeper.
Roderik: this for sure we won’t do.
Chris: are we in agreement that the other experiments continue taking data while there are the emittance scan in the other IP?
Witold: it depends on the burn off rate
Roderik: the burn off rate is much slower than in PbPb. I will recalculate the lumi lifetime with the updated parameters.
Witold: if it is significantly better, then why not then. We should look at the numbers.
Chris: this should be also the best scenario: if each is taking two hours for the scan, and the other continue, it means to take 2 hours from the program. If each takes the scan and the others no data, it means 8 hours taken from the program.
Reyes: as injectors, we’d like to have asap the filling scheme for pO and OO to know how many cycles we have to prepare
Chris: if you want paris of bunches or not.
Reyes: yes.
Roderik: one cycle for OO and one for pO.
Reyes:this morning we said for pO a cycle with a flat bottom that can accommodate six bunch injections, but today I only heard 3.
It is a different cycle anyway, because we want lower intensity in pO. Then it depends where we lose this intensity, if in the injectors, or the SPS, or somewhere else. For pO, it is clear that if you have 36 bunches, it is much faster if you have 6 single bunches per injection than if you have three. You cut the time in half. We want many bunches with low intensity for pO and it is easier to inject if we have more bunches per injections.
Reyes: sure, we’re open to your requirements, but we need to know if you want a cycle with three injections, or 5, or if you’re happy with three, and if this cycle can be used for pO or OO.
Roderik: this we’ll have from the lumi studies, when we have all parameters. Now only the pileup is missing.
Chris: we don’t need to be super-optimized, we can do 10 injections of 3 bunches, or 5 injections with 6.
CMS (Andrea)
S3:
Jorg: beam based alignment (for PPS) is needed for sure.
Filip: we should say that not all the PPS detector will go in, only a subset.
Jorg: but you’ll still need an extra cycle.
Filip: probably also ATLAS will then profit
Catrin: yes, AFP might be interested, but we still don’t have a final statement.
Roderik: the beam based alignment might be skipped depending on where you install the detectors. We need a statement about how close they want to go. If we can do without alignment, it’d be good for commissioning time.
Chris: we need pileup leveling for OO needed.
ATLAS (Catrin)
S2:
Catrin: ZDC is on the St Genis side
Catrin: If AFP also in pO, they might need alignment
S4:
Roderik: if everybody wants to level at something like pileup 0.02, then we probably don’t gain too much going to 6 bunches.
Chris: there are two schemes with 12 bunches; the Single12b has well separated bunches through the machine, and the one below, Split12b is in pairs, 6 pairs 200 ns apart. You seem to suffer when injecting in pairs.
Roderik: so the 12 well separated bunches seem a good compromise, especially if we do leveling, Then you don’t lose too much compared to 6.
Chris: 9 also gains quite a lot.
Chris: ideally pO at the same energy as pp means a 4th energy to setup.
Catrin: yes, but we understand that LHCf has precedence.
Filip: this is the same as for CMS, but agreed to give priority to LHCf.
ALICE (Robert)
S3:
Rayes: if you require more than 0.5 us bunch separation, does it mean that 200 ns separation is ruled-out?
Chris: this is not good for ATLAS (trigger) and ALICE, so better 3 bunches which are 1 us apart, and you inject in 3, it seems easier
Roderik: can we fit 36 in 1 us?
Chris: with 2 us I was getting 33. So with 1 us one should be able to get 36.
Rayes: I insist on 200 ns since this is a mitigation in case the intensity from PS to SPS is too high, so we can split the bunches in the PS, to help with the space charge and inter-beam scattering in the SPS, but this would mean to deliver 2 bunches separated by 200 ns. Since this is not ok, we need to decrease the intensity in some other way.
Chris: the intensities that we’re looking at are quite low. So you should be able to scrape them in the PS, or split them in the SPS and kick one out, I thought, because we’re not looking at huge intensities with 12 bunches in OO.
LHCb (Paula)
Jorg: The b* (1m), I am not sure it is consistent with our configuration. For pO, is 1 m ok for LHCb? For PbPb it is clear, but for pO, if we go only to flat top, then LHCb is at 1.5 m, not 1m.
Roderik: We would have to do another squeeze to get to 1 m, because we cannot go to 1m at end of ramp, I think.
Jorg: if this is only for LHCb, is it worth having such an overhead vs just running a few hours more to get the same lumi?
Chris: in the pO projections we assumed 1.5 m, to get to the target in 3 days, we don’t need 1 m. So we can do the old-fashioned way.
Jorg: only pO is in between, OO is fine.
Chris: We need the lumi projections, to see if it is worth commissioning the squeeze or not. For PbPb we definitely need it.
NeNe summary (Chris)
CMS strongly in favor
ALICE slightly in favour
LHCb is slightly against it
They say larger datasets are preferable, so they prefer to spend the time on OO to meet the target
ATLAS is against it
Rayes: If the NeNe is taken from PbPb, why is LHCb against?
Catrin: this would be incompatible with ALICE.
Chris: you don’t want to take NeNe from HI.
Chris: We are at an impasse, but there is pressure on the injectors’ timetable.
Proposal is to write that there is disagreement, so it would be good that spokespeople and management discuss it. Good if the timetable is made compatible with it.
The problem is that even if NeNe could happen any other day, but then there are constraints from the other experiments.