I’m a bit confused about the situation with the LHC start-up schedule, maybe someone well-informed can help. Here’s what I’ve seen recently:
At the AAAS meeting last Friday, there was a session on “Big Science”, which included discussion of the LHC with Robert Aymar, the CERN director general. Alan Boyle, at MSNBC, reports:
June was the time frame Aymar had in mind when he was asked about the start-up schedule during a Friday session on large-scale science project. But during a follow-up chat, he pointed out that you can’t just press a big red button one day and expect each of the collider’s beams to hit full power of 7 trillion electron volts immediately…
Aymar said that buildup could still start around May 21 or 22, with tests continuing for weeks after that. His aim is to have the collider conducting scientific experiments this summer.
One thing that is definite about the schedule is that there will be a ceremonial inauguration of the machine on October 21, with a wide variety of dignitaries present, include French president Sarkozy.
It’s not clear what phase of the start-up Aymar had in mind when he was referring to June, and the idea that the machine will be doing physics this summer seems hard to reconcile with information available publicly about how the things are progressing. For an official schedule from last August, see here. The current one, from October, is available here. Both schedules have beam commissioning beginning May 15, and taking about two months, so physics in July.
The LHC is divided into 8 sectors, and each sector must go through a process of flushing, cooling down to 1.8 K (which takes a month and a half), and powering tests of the magnets. The powering tests are crucial to make sure that the magnets can quench safely, dissipating the energy contained in a magnet that leaves the superconducting state unexpectedly. According to the schedules, the powering tests should take 2-3 months, and can start only once the magnets are cool. So, from beginning of cooldown to the point that a sector is ready to try and use should be a process of about 4 months or so. So, for mid-May beam commissioning, all sectors should be cool by around now and starting powering tests soon.
One can follow the actual state of affairs here. One sector (45) is cool and undergoing powering tests, but this sector still has not had its inner triplet magnets fixed, and the plan is to warm it back up before doing this, after which it will need to be cooled down again. Cooling of 3 other sectors has begun, but has been stopped in two of these to make repairs, with cooldown to resume in week 9 of the year for one sector, week 11 for another. Of the four other sectors, cooldown is supposed to start in one of them during week 15 (mid-April), dates are not given for the others. With respect to last August’s schedule, the current situation is roughly 4-5 months behind where it is supposed to be. This would suggest that, if all goes well from now on, beam commissioning would begin mid-September. Perhaps there was some slack in that schedule, and things could happen faster than planned, but I’m just not seeing how physics this summer is in the cards for the LHC. Most likely scenario seems to be a big push to get some a beam of some sort stored in the machine in time for October 21 and the big ceremony.
One thing is sure, Aymar is proof that humans exhibit the intrinsic property of spin. Dorigo, for example, doesn’t expect the LHC to begin running until the end of the year, and then only at very low luminosities, and no-one really knows how long it will take to ramp the beam up to useful luminosities.
I doubt dorigo can be trusted on these matters. For what I understand from his blog, he basically extrapolates from his past experience with Run I and Run II at the Tevatron, rather than relying on hard information he does not really seem to have access to. He might still be right, though.
However there is one important thing that is often overlooked. It is easy to understand the problems with cooling down the LHC components, with getting a stable beam, with fixing occasional incidents. But once those are straightened out, there is one major additional hindrance: even if LHC starts producing large numbers of collisions, the two detectors will have to figure out how to best trigger on them. Sure, at the very low initial luminosity a beam clock may be enough to rely upon for a complete readout of the detector. But still, getting good, meaningful data to tape is not a trivial task. If all we care is satisfy our curiosity on whether the LHC creates a black hole that eats out Geneva overnight we may be satisfied from day one, otherwise it will be very, very slow and painful.