jump to navigation

The LHC vacuum is too good, hand me a drill! October 24, 2007

Posted by dorigo in humor, physics, science.

This morning I followed a meeting where CMS experimentalists were discussing how to collect early data useful for calibrations and alignment, by optimizing the trigger requirements based on the operational parameters of the accelerator.

Basically, for a few months after the LHC accelerator is turned on, we will see a period of low beam-beam collision rates: luminosities below 10^{28} cm^{-2} s^{-1} at the very beginning, corresponding to <800 Hz of inelastic proton-proton collisions.

(By the way, this is one of the easiest formulas in particle physics, dictating that if a process has a cross section \sigma and the luminosity is L, then the rate of the considered process is N=\sigma L. With \sigma=8 \times 10^{-26} cm^{2} the proton-proton inelastic cross section and L as above, you get N=800 s^{-1}.)

As the machinists optimize the beam parameters, the initial very low luminosity collision mode will be alternating with periods when a single beam is circulating (no proton-proton collisions by design), then higher collision rates, and so on until reasonable rates (1 MHz and above) are obtained. It is an incremental process which contains some trial-and-error and some deep studies of the functioning of the machine, some of which are best done with a single beam in the tunnel. 

The rate of collisions at the very lowest luminosity is still high enough that some filtering is required before data can be stored to tape: the data recording capability will top at 300 Hz, but at the start the data storage system will not be fully operational.  In any case, the very first data will be very useful to understand the response of the CMS detector to charged tracks and jets, and to verify the alignment of the various subdetectors:  the more data is collected, the better. So a very important task of triggers in the early days will be to collect data useful for these calibration and alignment tasks. 

Ideally, regardless of luminosity one would like to optimize triggers such that the maximum amount of data is collected – and this 24/7. But beam studies take their toll… As these issues were being discussed, somebody suggested to use beam-gas interactions when the LHC accelerator is used in single-beam mode – when no beam-beam collisions happen by design in the core of the detector. That is a great idea, although a bit tricky to implement: you do not know where the beam-gas interaction takes place, so you have to tune your data acquisition system somehow, and reconstruction is also tricky since you have to drop the assumption that the tracks come from the center of the detector. Despite those caveats, you can get lots of data from single beam operation periods, when you would otherwise be just collecting muon tracks from cosmic rays.  

However, beam-gas interactions are only possible if the vacuum inside the beam pipe is not perfect, and (un)fortunately, the LHC vacuum in the vicinity of the detectors  is expected to be exceptionally good, to the point that people started to joke about creating accidental small leaks in the vacuum system:

What ? No leak incidents today ? Darn it !… Jack, hand me a drill, would you ?”

Seriously, there is a striking example of skilled experimentalists exploiting the subtlest nuisances they encounter, and turn crap into gold. I love itwhen I see smart minds at work. And you see it often in CMS meetings: excellent ideas are born when people have to rely on their resorcefulness, especially if they have one more year to tinker with inventive ideas while we wait for 14-TeV collisions…



1. Louise - October 24, 2007

Hee hee, perhaps they will discover vacuum energy. There will certainly be unexpected discoveries that will be fun to read about.

2. Paolo - October 24, 2007

Pretty post, thanks. By the way, is there a way to make reasonable with a minimum of math why electrons are important in the detectors, same for muons and – AFAIK – much less tauons? Is only because of the mass? Or there is something else to it (as one would suspect for a 14 TeV machine!)?

3. dorigo - October 25, 2007

Hi Louise, yes, but the fun of making those discoveries possible will not be less!

Paolo, electrons and muons are diamonds in the gravel of proton-proton collisions. They are exceptionally rare, and they invariably signal an electroweak process taking place. High energy ones are mostly due to the decay of W and Z bosons, or possibly fancier objects. They can be easily collected by online triggers and separated from fakes by simple offline cuts. Today’s collider detectors in fact base their discovery potential on the efficient detection of electrons and muons (CMS, for instance, stands for “Compact Muon Solenoid”).

Now, the very same argument of rarity and preciousness goes with tau leptons, of course. But tau leptons have unfortunately a mass that allows them to decay into light hadrons – one or three pions, typically. And they do that very quickly (while muons decay to electrons in 2×10^-6 seconds, outside the detector).

Because of the phenomenology of tau decay, tau leptons are easy to mistake for hadronic jets due to a parton fragmentation – something which happens in every hard proton-proton collision. Triggering on them is very difficult, and fakes are usually high. Despite of that, the decay H \to \tau \tau will be important at the LHC: for once, having a large mass will favor taus over muons and electrons: the Higgs has a sizable branching fraction into tau leptons for M_H <135 GeV.

For a discussion of tau identification in the context of Higgs searches, see for instance this post.


4. Paolo - October 25, 2007

Thanks a lot Tommaso, now in my mind many pieces fit together!

Sorry comments are closed for this entry

%d bloggers like this: