Three weeks of intense machine development on the LHC came to a satisfying conclusion on the night of 21 September with the final validation of the machine-protection systems for operation with “bunch trains”. Less than three weeks later, the machine was running with 248 bunches per beam, giving a peak luminosity of 8.8×1031 cm–2 s–1, close to this year’s target of 1032 cm–2 s–1.
Until the beginning of September, the LHC ran with bunches spaced by 1–2 μs, injected one bunch at a time from the Super Proton Synchrotron, the final stage in the injection chain. The change to injecting bunch trains – groups of bunches – not only reduces the time required to fill the machine but also allows for further increases in luminosity. It is therefore another important step on the route to full operation of the LHC. Eventually, the collider will run with 2808 bunches per beam, with 25 ns between bunches in a train. The target for 2010 was for a bunch spacing of 150 ns (equivalent to about 45 m) in the trains.
Running with bunch trains requires the careful setting-up of crossing angles between the beams at the interaction points in order to avoid unwanted collisions on either side of the experiments. Tests showed that the minimum angle needed to avoid parasitic collisions with the 150 ns trains is 100 μrad. They also revealed that there is more dynamic machine aperture at the interaction region than predicted at the nominal crossing angle, at injection, of 170 μrad. For the subsequent physics runs the crossing angle was reduced to 100 μrad during the ramp of the beam energy and the “squeeze”.
Using crossing angles has the consequence that all the protection devices had to be set up to match the new trajectories round the machine, a process that alone took the best part of a week, but all was ready for the first physics fill with the new conditions on 22 September. For this, the operations team injected three trains of eight bunches to give 24 bunches per beam. The fill of 13.5 hours provided around 170 nb–1 of integrated luminosity. A day later, the number of bunches was increased to 56 per beam.
This initial work on bunch trains was with approximately the same total beam intensity as in August, but the first fill brought a bonus. Bunches of nominal intensity were injected into the LHC with a smaller-than-usual transverse size. While this might give a higher initial luminosity, it was expected to cause lifetime problems when the beams were brought into collision. However, the beam lifetime remained surprisingly high (around 25 hours) and the luminosity was significantly higher than expected.
The first step to higher intensity took place on 25 September, with an increase to 104 bunches per beam. The total intensity was now more than 1013 protons per beam and a single fill for physics could deliver more than 1 pb–1. At 3.5 TeV the LHC had reached a stored energy per beam of 6 MJ, the highest for any collider and exceeding the record set at the Intersecting Storage Rings at CERN many years ago.
The next increase, to 152 bunches per beam, was made on 30 September by injecting 16 bunches at a time in two 8-bunch trains. This was followed on the night of 4–5 October with the first physics fill with 200 bunches, which provided 2 pb–1 in 12 hours. Then, on 7–8 October, the fill with 248 bunches was achieved, with bunch trains injected three at a time.
The strategy for increasing the intensity is driven by the machine protection, as the stored beam energy increases with each step. The aim is to provide three successful fills for physics to deliver more than 20 hours of colliding beams before progressing to the next step. Running with protons is scheduled to stop towards the end of October, by which time the LHC should be running with 344 bunches per beam. There will then be a period to set up for the first runs with heavy ions, before a short shutdown at the end of the year.