Starting in 2004 the MPI SCT group develops software needed for alignment of the SCT detector modules based on particle tracks. To achieve the best possible reconstruction of tracks the exact location of all modules needs to be known.
In a first step this is achieved by geometrical survey of the sensors within the modules, and by surveying the mounting points of the modules on the discs and barrel cylinders. The expected accuracy obtained with the external methods is better than 5 micrometer for the sensors within a module. For the module locations one expects larger uncertainies. For example for the endcap modules the expected precision is about 20-50 (100) micrometer transverse (parallel) to the beam direction. This is insufficient to achieve the physics goals of ATLAS.
Using the external measurements as starting points, the knowledge of the module locations is improved by the alignment procedure based on particle tracks. The MPI strategy uses an iterative procedure to constrain the six degrees of freedom for individual modules by means of a Chi2 minimization of the sum of unbiased track residuals with respect to the alignment parameters. The first version of an alignment algorithm has been designed and implemented into the ATLAS reconstruction software. Initial tests on Monte Carlo simulated events demonstrate that the method is functioning.
One possible way of improvement is the use of a large number of tracks which hit the regions where two modules on a structure overlap, since these give very strong constraints.
The method has been extended to also incorporate the pixel detector. For the pixels two three dimensional residuals are used for the two dimensions of the pixel within the module plane. Basic functionality has been achieved already and further studies are underway.
Related documents: