Wing Sheung Chan

42 The Large Hadron Collider and the ATLAS detector The remaining magnets are toroidal, which is a defining feature of the ATLAS detecto r † . One of the three toroid magnets is in the barrel region, while the other two are end-cap magnets deployed on both sides of the detector. The toroid magnets are used to supply bending power to the MS. The magnet strengths are approximately 0 . 5 T in the barrel region and 1 . 0 T in the end-cap region. 2.2.7. Crack regions Some regions of the detector are only partially and insufficiently instrumented due to various limitations. These regions are commonly known as “crack” regions. Particles that fall in the crack regions are usually reconstructed with significantly degraded efficiency and quality. One of the crack regions is the transition region between the barrel and the end-cap regions, where detector components are mechanically separated. This corresponds to the region 1 . 37 < | η | < 1 . 52 . Another crack region is the region | η | < 0 . 1 . This region is not covered by the MS chambers in order to give way to power and read-out cables for the ID and calorimeters. 2.2.8. The trigger system During the Run-2 data taking, the LHC operated using the designed parameter of 25 ns spacing between proton bunches. This means that 40 million bunch crossings occurred per second. With the overall average pile-up of h µ i = 33 . 7 , a rate of over 1.3 billion collisions per second can be expected. This amounts to a raw data acquisition rate in the order of petabytes per second, which is practically impossible to fully process and store. To resolve this, the ATLAS detector has a trigger system that selects only events that are essential to physics analyses in order to reduce the event rate to a manageable level. Throughout Run 2, a two-level trigger system is used. The Level-1 (L1) triggers are implemented in specialised hardware installed in the detector and are able to make fast decisions. They are tasked with reducing the event rate from 40MHz down to 100 kHz . The L1 trigger decisions are made based on information from the calorimeters and the trigger chambers of the MS. Information from the ID is not used because track fitting is a computationally intensive process. The L1 triggers are fired when signatures of muons, photons, electrons, jets or hadronic τ decays are detected, or when an event has large missing transverse momentum or large total transverse energy. Besides a simple binary decision, the L1 trigger system also identifies regions of interests, which are regions of the detector where data are fully read out and further processed. Decisions from the L1 triggers are made within 25 µ s of the associated bunch crossing. Events that are kept by the L1 triggers are further filtered by the software-based High- level triggers (HLT). With a reduced event rate of 100 kHz and defined regions of interests, the HLTs are able to make more elaborate decisions. Physics objects are reconstructed † ATLAS is a genius (but a bit forced in the author’s opinion) acronym for A Toroidal LHC ApparatuS.