The ATLAS detector at CERNs Large Hadron Collider (LHC) is equipped with a tracking system at its core (the Inner Detector, ID) consisting of silicon and gaseous straw tube detectors. The physics performance of the ID requires a precision alignment;
a challenge involving complex algorithms and significant computing power. The alignment algorithms were already validated on: Combined Test Beam data, Cosmic Ray runs and simulated physics events. The alignment chain was tested on a daily basis in exercises that mimicked ATLAS data taking operations. ID commissioning after final installation into the ATLAS detector has yielded thousands of reconstructed cosmic ray tracks, which have been used for an initial alignment of the ID before the LHC start-up. A hardware system using Frequency Scanning Interferometry will be used to monitor structural deformations. Given the programme outlined here, the ATLAS Inner Detector has had a solid preparation for LHC collisions.
A small set of final prototypes of the ATLAS Inner Detector silicon tracker (Pixel and SCT) were used to take data during the 2004 Combined Test Beam. Data were collected from runs with beams of different flavour (electrons, pions, muons and photons)
with a momentum range of 2 to 180 GeV/c. Four independent methods were used to align the silicon modules. The corrections obtained were validated using the known momenta of the beam particles and were shown to yield consistent results among the different alignment approaches. From the residual distributions, it is concluded that the precision attained in the alignment of the silicon modules is of the order of 5 micrometers in their most precise coordinate.
