Minutes and Summary

Main purpose of the meeting: Data taking progress & news; Oxygen run planning; VdM planning

LPC minutes 16 June 2025

Present (P = in person): Chris Young (P), Chiara Zampolli (P), Robert Münzer (P), Paula Collins (P), Andrea Massironi (P), Silvia Pisano (P), Flavio Pisani, Ivan Cali’ (P), Roderik Bruce (P), Lazla (P), Eric Torrence (P), Lorenzo Bonechi, Riccardo Longo, Simona Ilieva, Tomasz Bold, Witold Kozanecki, a guy at the end of the table that I do not know, Peter Steinberg, Andrea Delannoy, Dragoslav Lazic, Quipeng Hu, Brian Cole, Filip Moortgat (P), Matthew Nguyen, Gerardo Vasquez, Lorenzo Rossini, Jorg Wenninger, Andrej Gorisek

Introduction (Chris Young)

Robert Muenzer: for ALICE, it is better to have the lumi program in pO in the second fill because we need to setup the ZDC in the first one. In the second fill for us it is fine to do it earlier, so if something goes wrong, we have some contingency. [22:54: have I understood correctly that for them it is fine to do it early?]

Chris Young: as long as nobody else is scanning, your scan being very short, then it is fine. Let’s say that LHCb takes 2 hours, then ATLAS after 4h, there is then plenty of time.

Eric Torrence: on the TWiki, you expect the ppb [23:42 I think Eric said only pp, but I guess he meant ppb]  might be 2 to 2.5 in the pO? 

Chris Young: I think that is 5 rather than 4.2 in charges per bunch.

Eric Torrence: I think that this is a bit high for a scan, so we will prefer to be as late as possible.

Roderik Bruce: tomorrow there should be O beams in the SPS so tomorrow evening we should know a bit more where we are with intensities. Either we will do great, or we get space charge limit at injection and then we’ll be significantly lower than 5. The value will depend on how strong the space charge is. If this is a strong constraint, then it is better to wait for tomorrow’s test. Note that the test was meant for last week. 

Qipeng: in the beginning we also need commissioning work, 1-2 hours, so we cannot go directly to physics operation (max 2 hours) that the time in 0.01 might be a bit longer.

Chris Young: then we can make the first fill 9h and then the rest in the second fill. [25:30 I am not sure the accounting is correct here: Qipeng said 1-2 hours more, at 0.01, right? So if we add 1 hour to the first fill, it might not be enough. Even if you make the second number, in the second fill, be 5 hours. I think he wants a longer first fill]

Roderik Bruce: but then this time will not be for physics, so we need to add more time.

Chris Young: yes, but the time in the 2nd fill at 0.03 is 14 hours, which is above what Natalia’s estimate of when we get to the end so there is a bit of an hour contingency. 

Robert Meunzer: even if there is no detailed planning for OO, if you take the 2x 6h for OO, it means that the VdM should be subtracted from these 12h?

Chris Young: when Roderik did the calculations of the time, it was the physics time plus the VdM time, which he assumed was…?

Roderik Bruce: it was 4h for both OO and pO. We added 4h, assuming this is the prolongation of the total time with colliding beam because everyone is off. 

Chris Young: so in the end we’re looking at 3 fills for 6 hours: 2 fills of physics data and one for VdM. If VdM is shorter than 6h, you might get a bit more than 6h of physics.

 

Beam transmutation in the O-O run (Chris Young for John Jowett)

Robert Muenzer: any chance that these ions could create a background?

Roderik Bruce: Once these particles are lost, they should be lost on the momentum collimators: they will irradiate photons and they will get more and more off momentum and finally they will hit the collimation system. So most likely, it should not be lost on the TCTs, but to be seen if there is a leakage from IR6 to the ALICE TCTs. In the draft commissioning plan, I allocated 4h of preliminary background checks during the commissioning also to check this. If it could be interesting for you, in the end of the collimation setup fill, we could take a few hours and do the same thing that we did at the end of the Pb commissioning, and see with part of the detector if you see background or not. 

Roderik Bruce: [for the momentum cut: the plot on s5 where the peak of the energy is not at 2.68] talking to John, this might be the result of the ions irradiating photons and losing energy, but he is not fully convinced, which is why he tried two versions: either he trusts this, and there is no EMD, or he takes all of it. 

Chris Young: by making the cut narrower, the amount of transmutation reduces significantly (from 4.5% to 1%)

Roderik Bruce: to be fully clear, one should say that 1% of He4 could contribute to collisions in the bucket, the rest could be in other buckets. 

Roderik Bruce: the tracking could bring down the contamination potentially since you add the natural spread of the beam to the initial conditions in energy and betatron amplitude: because of that you’d have larger oscillation amplitude, so if there are contaminant particles close to being lost, they might be lost because you are adding a small smearing effects.

Note that there will be two more updates: using FLUKA + DPMJET3 and we’ll try to update the lumi profiles based on what they find in the SPS tomorrow. This could then change. The results presented are with 3e10, but it could go lower (better) or higher (worse - even if higher would be better if we level).

Filip Moorgat: what will the non-colliding bunch be for?

Chris Young: if we added an extra bunch to the scheme, it could hit SMOG, it would not have transmutation from high lumi collisions, so you can assume it is not transmutated, and you can use it to monitor the contamination with the track multiplicity over time or the other bunches relative to that one.

Roderik Bruce: of course there could be some transmutation that could collide with the SMOG but it will be much much less than the high lumi collisions at the experiments.

Chris Young: I am a bit worried about things that are monitored over time, since over time we’ll be changing many things, like beam intensity, pileup levels… There will be a lot of complications in working out whether we are seeing something on which we could change the physics program. It is more that we could use this extra bunch to analyze it afterwards. 

Chiara Zampolli: since the contamination now looks better, should we still stay with 6h long fills to be sure?

Chris Young: after 6h the total contamination would be ~8%, we need to see what the experiments say on their limits.

 

ATLAS (Eric Torrence)

Michi Hostletter: [about the difference in lumi between CMS and ATLAS] it is expected up to a certain extent: the separation level uses a tolerance band of +/-1.5% which is a deadband, so if you are inside the band, it will not do anything. If you reach the edge, then it acts to bring you back at the center of the band. In CMS it seems that there are many ups and downs, and sometimes the leveling takes a step up and sometimes down, so in average you are more around the center, while in ATLAS there is no drift and you always burn off, so then it takes a step up, and you burn off again, and it takes another step up and so on and each step brings you back at the center of the band which is your target. In CMS maybe there are some orbit drifts, and the separation collapses a bit, you hit the upper end of the band and then the leveling  brings you a bit down, so the average along a long turn is a bit higher. 

Filip Moortgat: [about mu = 64] it is not that we cannot run at that value, we fixed at 63 to be more stable, not be brought all the time back, but otherwise we could go to 64.

Michi Hostletter: the point is that in principle cryo should give us a signal of what is the max allowed limit at a certain point, but we found out 1 week ago that they increased the max allowed lumi a bit too eagerly before they actually see the effect of this high lumi: the yellow curve [see plot on s3] is their temperature, and you see a lag wrt the green curve which is the max lumi that they allow. So the yellow curve overshoots, and this is not stable regulation. They need to adapt their algorithm to go slower to allow for higher target, then we can go back to use it again. Still, from the efficiency of cooling, the triplet which is right of 1 is the most limiting in the machine, so if you both had 64, CMS would get a bit higher, because their triplets are a bit better. One thing we could do, is that at the leveling step you’re not brought to the center of the band (the target), but to 75% more. Or we could try to reduce the band from 1% to 1.5%, this should still work fine given the stability of the lumi of CMS and ATLAS, but we’d need to check that the leveling then does not fail. We can try to do this and check if it is stable, reducing the spread for both experiments. 

Chris Young: if you made the band narrower, it would be beneficial for both experiments, they would then have more consistent trigger rate.

Eric Torrence: it depends on how it reacts. If you make the band smaller, you might get more corrections and a broader rms. I prefer more stability. CMS has more spikes which we don’t see.

Filip Moorgat: yes, we get a lot of huge spikes. If I understand correctly, it is mostly due to orbit drift.

Michi Hostletter: the up and down is the orbit drift. When applying b* leveling you can get this kind of temporary [high] spikes. We can try with a narrower band and see if it is stable. If it is, it will reduce the spread for both, then the long term difference will be different. 

Filip Moortgat: it is also true that we don’t get all the extra that comes from the spike, since we have then also a lot of dead time. So it turns into a permil effect.

Chris Young: on the plot for cryo [s3]: there is a bit of time before it is allowed to step up again. This should be a lot longer so that it sees the temperature react. 

Michi Hosteletter: the length of these stair steps is now 15min if the conditions are good. The conditions go back after about more than 1 hour. In 45 minutes you reach the top, and after one hour they start seeing the effect of this [1:01:16 this what?] so probably they’d need to make each of these stair steps 30 minutes. To be discussed with cryo. 

Michi Hosteletter: [about the fact that there is a ~3% difference between ATLAS and CMS during the xing angle leveling] we should not entangle multiple effects: this happens when the experiments start being head on. If this is really related to xing angle, we could try not to use the x-ing angle anti-leveling for 1 fill. 

Eric Torrence: when the beams go headon, the difference is about 2%. Then the difference increases to 5%.

Michi Hostletter: the idea is to understand if the problem comes when you are headon, or from the sing angle antileveling, so not using it in one fill, will not cost much, and will help. One other option is that maybe the actual xing angle that we change is not really that. But this would be unexpected, we should be more than 10 urad off in one IP.

Chris Young: we measured the xing angle, right?

Michi Hosteletter: yes, with both the DOROS and the cogging technique, and there is no indication that we are so far. The other option is that the beams are not round. But it would be a coincidence that you’re similar in the beginning, and not at the end. 

Eric Torrence: I also noticed in the plots of the emittance from the scan that there is a difference between the horizontal and vertical widths in CMS but there isn’t in ATLAS. This is to think about.

Agreed to do the test after the VdM. 

The situation is like last year but the opposite direction (CMS getting less), when the xing planes inverted → so probably it comes from this, and not the xing angle.

 

CMS (Andrea Massironi)

Chris Young: the max lumi in pO is even with pileup above 1?

Andrea Massironi/Ivan Calì: yes.

Chris Young: for the pO lumi program, when would you decide between the 30 min emittance scan and a longer 1-1.5h program? If the first fill goes well, for example?

Andrea Massironi: the second fill would be also good.

Chris Young: how flexible are you for the solenoid when it is down? Someone in the MDs might not like changes in conditions.

Filip Moortgat: we discussed with Georges, and he said that after Fri it should be ok. The limitation is that it has to be in daytime. 

Andrea Massironi: it will take 4 hours for us to ramp down.

 

ALICE (Robert Münzer)

Excursions during the test with stronger orbit feedback in the right direction: less instabilities in the detectors, 30% reduction of the excursions. But data need to be better understood.

Chris Young: since with the stronger orbit feedback things looked more stable, if you have a longer high rate test, and then a low field high rate test, you would use this?

Robert Muenzer: we have to see. We understood that we cannot apply these settings for a long time. 

Jorg Wenninger: yes, we should limit it to the high rate test, and not during the b* leveling.

Robert Muenzer: yes, not for the usual data taking, but if we do a longer test, we have to see if we can do it. Same for the mini-ramp up after the special runs. But we need also to digest what we saw today and digest it.

Chiara Zampolli: so we still need to finalize the source of the spikes that happen every 10-12 minutes, right?

Robert Muenzer: yes, even if it does not impact the quality of the data, it would be useful if we could understand it.

Chris Young: although we cannot do it for all fills, we can still use it at the end of the fills when we are out of the leveling even for a few hours.

Chiara Zampolli: how many tests are planned at high rate?

Robert Muenzer: one at lower field, which is the key one. Maybe 2 in total.

Roderik Bruce: have you tried to extract the correlation between background and luminosity?

Robert Muenzer: we are working on it, also to have more fills.

Roderik Bruce: how does this impact your data?

Robert Muenzer: mainly MUON is affected and we’re checking if there are some cuts that can be applied.

Chris Young: in the special runs, since there is less intensity (smaller bunch charge), this should be very small, right?

Jorg Wenninger: yes, it should be almost nonexistent, because it seems to be beam induced. 

Roderik Bruce: interesting that it goes down during the fill. If it was really a leak, it should stay high all the time. While it is something beam induced that is worse this year.

Jorg Wenninger: this area was opened in the YETS. Also looking at previous years, the level is lower, but there is always a peak.

Roderik Bruce: this is 4R2. 

Jorg Wenninger: to use the TCTs, we should see where they are compared to this chamber. 

Chris Young: is there anything that we can do? More scrubbing?

Jorg Wenninger: we’re practically scrubbing all the time. We could do the HL-LHC beam injection test full machine :-) 

Roderik Bruce: with commissioning fills where there is low intensity, we might not see it, right? Since it is beam induced.

Jorg Wenninger: we can see right now, with the probes that we’re using. 

Roderik Bruce: if we see something there, we could do some tests playing with the TCTs online. I am not sure how easily we can change the TCT in operation since there are interlocks.

Jorg Wenninger: we should first identify where the gauges are compared to the TCTs. If the vacuum is behind, there is no point in doing anything.

Chris Young: it would be interesting to compare how the same point in the fill (e.g. after 3h) is behaving, to see how it changes as a function of time, to know how quickly the conditioning is maybe happening.

Robert Muenzer: we need indeed to check. 

Roderik Bruce: you will have to normalize to the intensity. 

Roderik Bruce: the 10% contamination in OO, is it on the total lumi or the intensity?

Robert Muenzer: the total luminosity. The lower, the better. If we can split in shorter fills, it is better.

Roderik Bruce: total luminosity: this means that we can be significantly above in instantaneous luminosity at the end of the fill and even higher in intensity [1:36:15 I am not sure I understand what Roderik means here] 

Robert Muenzer: if we level lower, we’ll have less contamination.

Roderik Bruce: yes, depending on what ATLAS/CMS will do.

Roderik Bruce: how do we coordinate between the experiments for the pileup? 

Chris Young: people can level at different values. 

Eric Torrence: we might in the end do similar to CMS. We also need to consider transmutation.

Filip Moortgat: in the first part of the fill we would level.

Robert Muenzer: for the configuration, how much do we impact each other? How much are we sharing the bunches?

Chris Young: every bunch that collides in ATLAS collides also in CMS, so they share them all. Those in ALICE, each beam collides in either ATLAS or CMS.

Roderik Bruce: I think that half of your bunches will be colliding in ATLAS/CMS, and the other half in LHCb.

Robert Muenzer: does it mean that the two beams will transmute differently?

Chris Young: yes.

 

LHCb (Paula Collins)

Jorg Wenninger: when do you want to change the polarity?

Paula Collins: we need to be up for the special runs, down for the VdM. Then we stay down. This means positive polarity.

Jorg Wenninger: we might need to discuss. We’d like to come out of the TS with the same polarity to not flip 2x. Otherwise we lose what we set up in the first days. 

Flavio Pisani: but it will be the same as now, up as for the ions. The VdM is after.

Jorg Wenninger: yes, but the first thing after TS is to check the cycle. And we need to setup the VdM, which you want at reversed polarity. And we need to check the VdM with the correct polarity. On the Sat there is the plan to check the VdM, and the nominal cycle for the ramp up. Ideally this should be the same: polarity for VdM and pp. But then there is ions. 

Roderik Bruce: would it make more sense to do first the special runs, and then everything for VdM? This might help with the polarity switch. 

Jorg Wenninger: I think we’d need to flip even more. Since we want to prepare VdM and loss maps before we start. 

Roderik Bruce: yes, but could first do ions.

Jorg Wenninger: the idea was also to not start ions on Fri evening. 

Chiara Zampolli: can we use the SMOG in NeNe to measure the contamination if you inject Ne?

Paula Collins: it is being followed up. Maybe offline also with Ne we can distinguish.

Paula Collins: should the magnet be on during the MD?

Jorg Wenninger: yes.

 

LHCf (Lorenzo Bonechi)

Riccardo Longo: I am concerned about the other experiments running at high lumi. How does it impact LHCf that will run at low mu? 

Chris Young: it does increase the transmutation if there are more collisions in the other experiments. We need to work out whether this is negligible. The hadronic contribution is larger in OO because both beams can transmutate. We hope that it is small, but we should check if the other experiments might harm the LHCf physics program.

 

Conclusions on OO

OO: we’ll do 6-ish fills, 3 fills to fit the physics and the VdM. 

pO: transmutation should be negligible: we can have an estimate for after 14h, to check that the experiments at high mu do not affect the beam too much. And we’ll wait till tomorrow for the test in the SPS, that could not be done last week since they needed to change the magnet. We’ll send an email when we know the O intensity that SPS can provide.