First physics with the LHC

14 December 2009

Tracks in the Inner Detector from the first run with stable beam, showing the three layers of silicon pixel-detectors in light gray, the four layers of the SCT outlined in black, and the TRT in purple.



ATLAS is designed to observe all physics, from the highest energy interactions the LHC can produce down to the relatively low-energy soft events observed with these first beams. The trigger is designed to capture the high-energy hard events, ignoring low-energy backgrounds, but in these early stages, it is set for minimum bias. At initial, relatively low luminosities, the probability of more than one proton-proton interaction as bunches cross is close to zero. These conditions are ideal for the study of Minimum Bias physics, investigating what will be a dominant background at higher luminosities.

 “Minimum Bias is mostly about measuring the soft QCD processes,” says William Bell, coordinator for the Minimum Bias trigger, contrasting that with hard events like Higgs and supersymmetric particle production.

The rate at which proton bunches run their laps around the LHC stands at about 11 kiloHertz, but collisions occurred just once every second, on average. “You can imagine, that’s why we need a trigger,” Will points out. “The rest of the 11,000 times it goes around, nothing happens.”

The team was busy even with first beams. “We were working on Friday, the 20th of November to understand our trigger,” says Will, and they quickly corrected the level 1 trigger timing on Sunday, November 22nd, with readout timing corrected shortly after.

At first, tungsten jaws in the collimators were closed, resulting in dramatic beam splash events. These hit all of the Minimum Bias Trigger Scintillator, allowing for the speedy corrections to timing. Since the doors opened to allow circulating beams, just a few stray “halo” particles slam into the collimators upstream of ATLAS and create showers in the detector. Small amounts of gas in the imperfect vacuum of the beam pipe cause beam gas events. Already, the trigger is learning to disregard events in the detector that stem from beam gas or halo, in addition to the cosmics it has been measuring for over a year.

Although the Minimum Bias group is relatively small, just forty or so members, it has its own data quality and preparation arm, led by Prafulla Behera. The raw data is first processed at Tier-0, creating Event Summary Data files which ship out to Tier-1 centres the world over. “At this point we have to make sure that the quality of this data is usable for physics,” says Prafulla.

His crew checks the quality of the data recorded by the detectors to see whether it’s complete enough for various minimum bias studies. They grade the data for what kinds of analysis it may be used for, and save it in digested form, the D3PD files. As the first minimum bias studies are concerned with charged tracks in the Inner Detector, their data quality group is in good communication with the tracking groups, updating software to get the best tracking performance out of the data. “It’s a combined effort now; we are very much coupled with the tracking performance group,” says Prafulla.

Alongside the real data, the full Monte Carlo sample for 900 GeV collisions is being reprocessed on the Grid to validate the results.  Meetings are critical to the data preparation and data quality groups – Prafulla attends the ATLAS Daily Run Meeting at 09:30, and as of Dec 5th, the minimum bias analysis group meets daily as well.

While the beam was unstable, the Inner Detector ran with low voltage or not at all, but on Sunday, December 6th, the accelerator group announced that the beam was orbiting stably with no intent to dump it any time soon. Finally, it was safe to run the Pixel and SCT detectors with high voltage, and the Minimum Bias group could harness the full tracking power of ATLAS.

Track reconstruction is post-doc Maaike Limper’s area of expertise. Perturbative theory doesn’t describe these low-energy QCD processes well, so the expected results of such collisions are predicted by an empirical model. For now, the model is based on previous experiments on the SPS and Tevatron, which ATLAS data is expected to confirm it. But as the LHC moves into higher energies, the Minimum Bias team will tune that model to match the results observed in ATLAS.

“In Minimum Bias, you want to understand the tracking efficiency that the software has,” says Maaike.  ATLAS, like all detectors, has limitations in the form of dead channels and dead space – places a particle can pass through without being measured. The Minimum Bias team uses the known tracking efficiency to correct for this, finding the actual number of charged particles produced in proton collisions. This measurement can then be compared to the number of charged particles expected at given collision energy to correct the QCD models, according to Maaike.

“It is the very first analysis you want to do with ATLAS data,” says Will. Minimum Bias is the first analysis for the other three experiments as well, so the pressure is on to publish.

 

Katie McAlpine

ATLAS e-News