Although the LHC is approaching the end of Run 2, a major milestone in its history, there is little sense of this in the CERN Control Centre, where the one-month 2018 heavy-ion run has just one week left to go. Much still has to be achieved before the start of Long Shutdown 2 (LS2) on the morning of 3 December and there is no let-up in the pace of activity.
The heavy-ion source has made a remarkable recovery from its emergency surgery at the start of the run. The chain of accelerators downstream, starting with LEIR, have cooled, bunched, stripped and accelerated the stream of ions emerging from Linac3. The single-bunch intensity quickly exceeded the target for the HL-LHC. Until 19 November, bunches were delivered from the PS to the SPS in batches of four separated by 100 ns. A new scheme with batches of three bunches separated by 75 ns is now in use. Besides allowing more bunches to be packed into the LHC rings, this is already delivering even higher intensities in each bunch.
Nevertheless, one should not get the impression that all is plain sailing. Taking the injectors and LHC into new domains of operation continually throws up surprises and the teams concerned are in permanent problem-solving mode. The unprecedented intensities cause collimation losses that are always close to limits that can cause the beam to be dumped to avoid magnet quenches (new installations during LS2 will help with that). An unexplained motion of the beam orbit at a frequency of 10 Hz has provoked a few premature dumps. It is now clear that this was present in the preceding proton run but it causes more trouble for lead beams.
Problem-solving is further complicated by the need to weigh the time taken for conclusive investigations against the potential gains in the remaining run time. Decisions are based on hard data, physical understanding and judgement, but are subject to the strictures of machine protection. Continual interaction with the LHC Programme Coordinators ensures that the priorities of the LHC experiments are best served.
For example, the beam size at the ALICE interaction point was found to be twice the size it should be and measures had to be taken to reduce the impact on the integrated luminosity. A textbook explanation for this is a shift of the beam waist (the place where the beam is smallest) away from the interaction point (see a picture of the beam optics in the ALICE experiment in the last LHC Report). An initial measurement (moving the collision point using the RF system) hinted that this was the case – just before it was aborted by an interlock. A later measurement was completed but refuted this hypothesis. Finally, based largely on theoretical considerations, Stéphane Fartoukh identified the cause as a strong, but highly localised, coupling of the horizontal and vertical betatron motions and proposed an elegant solution. This worked almost magically, restoring the proper beam sizes and luminosity for ALICE.
This was accomplished during the scheduled refill of the ion source. Then followed a rigorous re-validation of a new collision configuration with the polarity of the ALICE spectrometer magnet and crossing angle reversed, as requested for the second half of the data-taking. At the same time, the injectors have been switching to the new 75 ns injection scheme and the LHC has been learning to handle very high lead bunch intensities up to four times the original design value.
On 25 November, a new peak luminosity record of 6×1027cm-2s-1 (i.e. six times the original LHC design value!) was set in both ATLAS and CMS. LHCb is also receiving far greater heavy-ion luminosity than ever before and ALICE is being held at its saturation value for 8 hours in typical fills.
Peak luminosities in ATLAS and CMS are being steadily increased from fill to fill, with the goal of demonstrating the nominal value for the HL-LHC. And a few special measurements are still to come…