Wednesday, February 25, 2009

Chamonix Summary -- Dipole Training

I've been talking about the Chamonix workshop for the past few posts, and will spend a few more on it, mostly because I find the accelerator side interesting. Plus, ATLAS kind of depends on those protons eventually colliding ;)

Yesterday afternoon, we got a 3-hour summary of the workshop (agenda and slides), so I thought I would share some of the more interesting items. The information (and accelerator acronyms) flew by pretty quickly, so this comes with the usual disclaimer that any mistakes are mine.

The first topic is "training" the dipole magnets... For magnets, training is accomplished by gradually increasing the current through the magnet (which increases the magnetic field) until the magnet quenches. Then the current is increased again, hopefully past the previous quench point, until the magnet quenches again at a higher current and magnetic field. The magnet "remembers" how high the current got, so the next time you turn it on (days or weeks later), it should be happy with any current below the highest quench point. And by the way, the currents that we are talking about are ~10000 A (compared to your wall plug of 15-20 A).

The LHC accelerator physicists talk in terms of "number of quenches"—how many times they have to repeat this current increase to quench cycle in order to reach a given magnetic field and therefore energy. The design energy of the LHC is proton collisions at 14 TeV, which means that every magnet in the ring must be trained up to 1/2 of that energy, or 7 TeV. This fall, we are going to run at 10 TeV, so the magnets must be trained up to 5 TeV (which they are, except for the repaired magnets in the sector that had the incident last September). They would like to reach the LHC's design energy for the run in 2011.

At Chamonix, they tried to estimate how many quenches it would take to reach given energies in the magnets... keep in mind that they can train magnets in the 8 sectors in parallel, but can only do ~3 quenches per day. Their estimates are that it would take 11 quenches to reach the equivalent current for 6 TeV in each magnet, 84 quenches to reach 6.5 TeV, and nearly 1000 quenches to reach 7 TeV (design energy). 1000 quenches means training magnets for 2 months!

So you can see, it gets a lot harder to make that last step from 13 TeV to 14 TeV. Which is why 13 TeV might be the maximum center-of-mass energy reached by the LHC. But I have 2 conclusions: 1) I wouldn't be disappointed with 13 TeV and 2) I'm not going to count these accelerator physicists out just yet...

A bientôt!

1 comment:

futureboy said...

The articles says they estimate that they will require 1000 quenches to reach 7 TeV, and they can only do about 3 quenches a day, which means training magnets for 2 months.

How does that math work out? If you can do at most 3 quenches on a magnet per day, that would require almost a year. I'm not sure if I understand your discussion about doing 8 quenches in parallel. Even if can divide this by 8, that's a minimum of 111 days, which is well over three months, not two.

Can you please clarify?