Already gearing up for next beam period …

12 January 2009

General plans for ATLAS operation in 2009, towards the next beam period.



Even without beam, the first half of 2009 is going to be surprisingly busy, not the least for the people involved in running ATLAS and any of its sub-detectors. While improvements and repairs are ongoing underground, developments are also in progress in many areas, from getting more flexible ways of changing trigger prescales, developing enhanced Data Quality monitoring, to simply reorganizing shifts and 24/7 on call duties, … everything between the sky and the floor of the ATLAS cavern, which can improve how efficiently ATLAS will record data, once the collisions are here.

After the short Christmas break, every piece of ATLAS, from the gas systems to the high speed link to Tier 0, have been quickly brought back online. Why such a rush? Simply because there is so much on the to-do-list, and everybody is motivated to get ready again. And this time you can feel there is more confidence, given that the LHC has gained real experience now.

In parallel to the on-detector work, January is dedicated to migrating all ATLAS sub-detectors to a new version of the Trigger and Data Acquisition software, which incorporates many improvements and new features that will allow for more flexibility, robustness and enhanced troubleshooting capabilities.

Three weeks of February are allocated to the yearly maintenance of the ATLAS electrical, cooling and environmental infrastructures. Every sub-detector is preparing to cope with this period, such that high priority work can continue without disruption. In March, soon after the end of the maintenance period, ATLAS will perform a full scale validation of the ATLAS DAQ and Trigger software intended for 2009 data taking, using the actual ATLAS Point 1 infrastructure, from the Readout System up to the Tier 0.

Based on ATLAS history and operation without beam (yeah, we got a lot of that…), a natural structuring of ATLAS has emerged in ‘slices’, intermediate in size between a single sub-detector and a complete ATLAS detector.  The four slices: Inner Detector, Calorimeters and their Level 1 trigger, the Muon System with its Level 1 trigger and the Forward Detectors, have proved to be very useful intermediate size systems, to debug this complex experiment.

ATLAS has the capability to operate two or more of these slices in parallel, with independent data flows, all the way to Tier 0, hence allowing multi-tasking of ATLAS. It is particularly useful outside beam time, for calibration, alignment and debugging purposes. This capability, already partially implemented, will become fully available during the spring of 2009. This feature was already extensively used by the calorimeter slice allowing for example a Liquid Argon – Tile – L1Calo combined calibration run, in parallel with any other type of ATLAS combined run.

April will be the last stretch to complete the underground on-detector work. In parallel, it will be dedicated to four weeks of combined cosmic and technical runs, one week for each of the four slices (Inner Detector, Calorimeters, Muon System and Forward Systems). Each slice-week will be used in priority to check that the sub-detectors of the slice can take data together, and pass a range of performance tests, such as high rate data taking. It will also be an opportunity to gain more operational experience with the Muon Cathode Strip Chambers and the Zero Degree Calorimeter, more recently joining the ATLAS combined data taking. The slice-weeks will also be used to quantify data taking efficiency and stability, understand data quality, find and fix bugs…

At the beginning of May, the ATLAS detector will be closed and available, including the magnets, for a full ATLAS cosmic run. The first part of the run will be dedicated to check that ATLAS as a whole passes a range of performance tests, from stability to high rate running, and find bugs again… Exercising the High Level Trigger requires no less than the full ATLAS to be run. Upon needs, this exercise could be extended to encompass a substantial cosmic data taking period, for calibration, timing and alignment.

Up to four weeks before expected beam, ATLAS will enter a phase of semi-continuous running, during which priority will be given to cosmic data taking with all sub-detectors, or to the necessary adjustments prior to beam. We eagerly expect the beams by end of July 2009. ATLAS will then surely benefit from all the hard work performed during the shutdown and all the spring debugging, and will hopefully ramp up quickly to efficient recording of quality data, with a stable setup for data analysis. The Chamonix meeting in February will either confirm or update the LHC schedule and the above plan will be adapted accordingly if beams are foreseen later than July.

 

Collage of activity in the Control Room, click for full size


 

 

Christophe Clement

Stockholm University