Luminosity determination and simulation of the LUCID detector at the ATLAS experiment

Detta är en avhandling från Lund University , Department of physics

Sammanfattning: The aim of this thesis is to describe how the luminosity can be measured in the ATLAS experiment at the Large Hadron Collider. Luminosity is a fundamental quantity that is used in most physics studies at the LHC. For example in order to measure cross sections. Firstly, the detector description of LUCID which has been implemented in the global ATLAS software framework is described and validated. Secondly, algorithms to determine the luminosity are derived and the simulation of LUCID is used to test the precision of these. Thirdly, the feasibility of using Z0 production as a alternative way to measure the luminosity or to calibrate LUCID is studied. The performance of LUCID is addressed by comparing results from the early 2010 data-taking period with predictions from the simulations. Several techniques to calculate the efficiency of LUCID are described and the results for each method is presented. Several classes of luminosity algorithms studied and applied to the signals from the LUCID detector. The first class has been designed to be used online and to extract the luminosity using two different counting methods, namely event and hit counting. An empirical model is introduced as an attempt to parametrize the non-linear effect by polynomial fits to the event and hit rate. A final attempt to minimize the non-linear effects was presented by introducing a new method that uses the full hit multiplicity distribution to determine the luminosity offline. A method to measure the absolute luminosity using the production of Z0 boson is outlined . Each step of the procedure to select Z → μ+μ− events has been studied in detail using Monte Carlo simulations of the various background and signal samples. The selection procedure is applied to a data sample recorded by the ATLAS detector over a period of 5 month and the resulting integrated luminosity is determined from the number of reconstructed Z → μ+μ− events.