First Proton-Proton Collisions In CMS!
    By Tommaso Dorigo | November 23rd 2009 04:38 PM | 5 comments | Print | E-mail | Track Comments
    About Tommaso

    I am an experimental particle physicist working with the CMS experiment at CERN. In my spare time I play chess, abuse the piano, and aim my dobson...

    View Tommaso's Profile
    My blog is not a place for hot-off-the-press news - in it you are more likely to find discussions on material well digested and thought over. Nevertheless, I do not have the guts to sit on today's news. The Large Hadron Collider at CERN has produced its first high-energy proton-proton collisions, in the core of the experiments instrumenting its underground caverns.

    It has been a long way since the first design of this extraordinary machine. I was reminded of just how much effort the construction and commissioning took by a slide shown by Ives Sirois at a workshop in Turin today: it is a schedule of the construction of the LHC dated 1989!

    As you can see, the LHC construction was supposed to go on during LEP operations: they were once thinking that they could install the big proton machine in the tunnel without having to remove the electron-positron synchrotron. Things took a different turn soon, and we are now in 2009, with two years of delay with respect to the above (admittedly optimistic) schedule!

    But we are there now. Today, the first 450-GeV protons were made to coast along the beam, bent by dipole magnets but not accelerated further after their injection in the 27 km tunnel. 900 GeV of energy available for collisions are a respectable goal -suffices to say that it is the first time we see this energetic proton-proton collisions (the SppS and the Tevatron only produced proton-antiproton collisions)- but admittedly from a scientific standpoint there is not much to be excited about. It is rather what these first collisions announce that is bound to make everybody's eyes fixed on CERN today.

    So let us see what CMS collected. Below is a picture of the very first collision recorded by the experiment. It is a good idea to have a close look, because displays such as the one shown might one fine day be presented to evidence new physics!

    In the picture you see three main displays. The one on the left shows a zoom into the inner tracker, with the charged tracks reconstructed from the hits they left in the silicon microstrip sensors; outside of the inner circle you can see the energy deposits in the calorimeter.
    The outer red "bricks" show the muon drift tubes, which remained silent: good to know, since we do not expect that a random collision will produce muon candidates; it would have actually been a bad sign if the muon detector had lit up!

    The pictures on the right show respectively a side view and a 3-D view. By now you should have figured out the meaning of the various components. One thing to note, though, is the remarkably non-jet-like structure of the calorimeter energy deposits. Remember, this is a randomly picked collision, and the released energy is too small to produce very narrow, localized hadronic jets. We should all look forward to more interesting pictures in the forthcoming days!

    For more pictures from all the four experiments, please read the CERN press release.


    Thanks Tommaso,

    I've been trying to understand this display, but it is very complex.

    The yellow represents hits in the silicon strip detectors.
    Q.1: What about pixel hits?

    Green lines are reconstructed trajectories of charged particles.
    Q.2: Many silicon strip hits have no trajectory through them.
    Q.3: No (obvious) curvature to the paths, or other indication of charge +/-
    Q.4: Some of the paths do not pass through the origin. Is this significant,
    or inaccuracy?

    Red calorimeter (ECAL) means photon or electron. Blue calorimeter (HCAL) means
    hadron. So, six hadrons. The one at the bottom has no charged track, so perhaps
    a pi-0.

    To the left, 24 charged tracks are tabulated, with column headings pt, eta, phi.
    I'm assuming pt is transverse momentum in GeV/c, eta and phi are latitude and
    Q.5: why are the tracks predominantly transverse? Is this really a random collision,
    or has it been filtered for high pt? All of these detectors have end caps, but the
    side view shows all the tracks to be very transverse, and the lowest pt I see is 2.8.

    (The color table at the lower left is especially confusing, but I guess even though
    every entry in it has been checked, it is not being used for the display.)

    Hi Bill,

    congratulations for making sense of the above pictures... I can provide some help below.
    A1: the pixel detector, as well as the inner five layers of silicon (called the TIB, for tracker inner barrel) were not included in data taking for safety reasons. With unknown beam conditions, the silicon runs risks if it is powered up, because a high rate of proton losses traversing biased sensors may damage them.
    A2: I think you should not attach too much meaning to the reconstruction of tracks that is shown in this display. I do not know the details, but usually these displays have thresholds of all sorts and selection requirements applied to the objects they display. So, for instance, if I require that only tracks with at least five hits are displayed, you will see many hits left behind with no "track" plotted over them.
    A3: I am not sure what was the magnetic field during this data taking; so depending on the momentum, you might be able to see a curvature by eye, or not.
    A4: you do not expect all tracks to pass through the origin. You may always get "v-particles" which travel several inches before decaying into two charged tracks. The V-particles are neutral kaons and neutral Lambda baryons, and they contain a strange quark. They decay weakly in 10^-10 seconds or so...

    Pi zeros are hadrons, but they are detected as pairs of photons in the em calorimeter. So they for sure do not give a signal in the hadron calorimeter -the photons are stopped within the EM crystals.

    A5: the transversity of particles (or the "centrality" of the collision) is a sign that some strong interaction took place in the collision, causing a transverse acceleration of the quarks participating in the collision. At a hadron collider we in fact select these as the most informative events. Remember that quarks carry a unknown fraction of the proton's longitudinal momentum, and so the center of mass of the collision can be moving in one or the other direction along the z axis. So we care about the transverse component much more than the longitudinal momentum, when we select interesting events. Now, during this first run we did not have much of a choice, given the extremely low luminosity. However, the events more interesting to check the behavior of the detector are those that indeed showed high-Pt tracks. One of those, you get to see here.

    Hope that helps,
    My 2 cents for Bill:
    - during that very first run the magnetic field was rigorously kept off: so no field, no bending
    - on top of genuine collisions, you are likely to have so-called "beam-gas events" (a proton from the beam hitting an atom of residual gas) and maybe some cosmic. Moreover, with only the outer 6 layers of CMS tracker on (out of 13) and no bending, don't expect to find perfect pointing tracks... and don't be surprised to see something like this:,
    taken from the same run Tommaso referred to. For nicer event displays, with more objects closer to what expected... have a look to HEP blogs in the very next days ;-)

    hey! are you still alive?!! we needz moar HEPP bloggingz!

    Q1: I confirm that only the outer layers were powered.
    Q2: actually, the standard tracking reconstruction was not very efficient for this incomplete setting of the tracker, therefore very soon the "seeding" of the tracks was modified and the track efficiency increased. But since this was the very first collision picture to be circulated, it was with the standard tracking.
    Q3: I confirm, no magnetic field that day (on purpose: a few days later the solenoids of each experiment were turned on one at the time, to readjust the beam taking into account their nuisances.)
    Q4: I'm not sure
    Q5: no filtering on Pt, since there was no magnetic field. The tracker has indeed endcaps, but these were switched off during that run. The acceptance of the tracker outer barrel is |eta|<1, therefore pretty central. In typical QCD collisions the eta distribution is quite flat (which means that the particle density grows a lot as a function of the azimuthal angle theta), but here only the very central region was instrumented.