Wednesday 27 May 2009

News and Gossips from the LHC

Since two weeks Planck stands first of all for a satellite CMB observatory, but it's also the name for an annual series of conferences on physics beyond the standard model. This year's edition is taking place in the furnace of Padova. Since Tommaso is around, he will surely describe everything in great detail, including color of the tie of each speaker, while I should later write a summary of the interesting ideas discussed here in case there is any. But for now I'd like to share a handful of interesting facts about the progress of the LHC that I learned from a supercool talk delivered here by Jörg Wenninger. I guess most of what's below is not new and should be familiar to those closely following the LHC saga.

One interesting fact I was not aware of: a quench (a phase transition from superconductivity to normal conductivity) of an LHC magnet can be induced by just a few milijoules of energy. That energy may be provided by a bunch of strayed protons from the beam . To avoid quenching, LHC cannot lose more than a millionth part of its beam. For comparison, the Tevatron loses about one thousandth of its beam during acceleration. In that respect, Jörg was very convincing that the LHC would ever work ;-) But then, miracles do happen, sometimes.

Another interesting part of the talk was the explanation why the LHC will run at 10 TeV in the center of mass, instead of the nominal 14 TeV. The story goes as follows. Before installing, the LHC magnets have to be "trained", that is to say, to undergo a series of quenches to let their coils settle down at stable positions. After being installed in the tunnel they are supposed to come back to their test performance with no or few quenches. It turns out that the magnets provided by one of the three manufacturing companies need an extraordinary number of quenches to settle down. Although the company in question was not pointed at, everybody knows that the name is Ansaldo. In the case of that company, the number of quenches required for stable operation at 7 TeV per beam is currently unknown, it is probably somewhere between a hundred and a thousand. At the moment it is not clear if the LHC will ever reach 14 TeV; 12-13 TeV might be a more realistic goal.

The talk gave also a detailed account of the incident of September 19 known as the 9/11 of particle physics. Although the evidence has evaporated, one can quite reliably outline the sequence of events. An abnormally large resistance in one of the magnets acted as a heat source that quenched the superconducting cable at one interconnection. In case of a quench the current should start flowing or a few minutes through the copper bus-bar that encloses the cable until the energy stored in the magnet is removed. However, due to bad soldering of an interconnection the current could not flow normally and an electric arc was created. This melted copper, punctured the helium enclosure which led to spilling of 6 tons of helium into the tunnel. The logo of the company that made the faulty magnet is always erased in the pictures, although everybody knows that the name is Ansaldo.

So what's next? The repairs of the damaged sector are almost finished. The current plan is to head for collisions this year (with a caveat "depends how one defines collisions"). Beam commissioning is scheduled for September/October and the first collisions could happen in November. The schedule is very tight and, moreover, the quality control of has revealed problems like bad soldering or reduced electrical contact in a number of places, including sectors that are already cold. The rumor is that some of the LHC magnets in reality turned out to be electric kettles.

The slides here.

9 comments:

Anonymous said...

thank you very much for sharing the lastest rumors, and for pointing out this talk.

I wonder how many papers will come up this year that just redo some analysis for 10,11,12 and 13 Tev.

Cheers!

Anon2 said...

Most of the analyses have been done separately for both "early physics" and >>normal<< lhc operation. Of course this doesn't diminish the value of this excellent and informative post!

Anon3 said...

That's really annoying that Ansaldo has giving the LHC such shabby products. I hope they get their just reward for it.

Unknown said...

You would think that they would have some kind of quality assurance program in place to prevent things like this from being missed. Seems pretty basic. Current, resistance, heat, fail...

Ptrslv72 said...

Well, let's say that Italians are not famous for the reliability of their technology... Anyway, to be fair, nothing in the slides seems to suggest that one specific firm (Ansaldo) is to blame for the sloppy joints between the magnets. On the contrary, I would naively assume that the interconnections were manufactured directly at CERN. Do you have any special information about that? Cheers, Ptrslv72

Jester said...

No, I don't know if it's CERN or the dealers who are responsible for bad joints.

Anonymous Snowboarder said...

Perhaps if CERN had not been in such a rush to step things up and had actually done this quality control work and followed a proper testing schedule these problems might have been caught far earlier with less damage and hopefully less time ultimately wasted.

This is as much a failure of project management at LHC as it is the magnet manufacturer.

Anonymous said...

I've heard a number of talks similar to Wenninger's, and when the speakers dealt with bad joints they tended to blame outsourcing to smaller firms of work once done by CERN engineers. (Many of these speakers were senior people who roamed the 70s and 80s, so I assume they are knowledgeable about proper joints). While specific companies may bear responsibility, the mood was very much against CERN management; but then, this may also have a component of staff unionism...

Anonymous said...

Actually, I'd like to point out that it's a known fact that the worst magnets so far have been manifactured by Noell. These magnets have proved very difficult to train, while the other 2 manifacturers (Alstom & Ansaldo) provided ok magnets.

Indeed, if the magnets had been properly tested before turning on the LHC, the accident wouldn't have happened. But they were so eager to get collisions before the new management took control...