Development Of New Approaches To Atlas Detector Simulation And Dark Matter Searches With Trigger Level Analysis

Development of New Approaches to ATLAS Detector Simulation and Dark Matter Searches with Trigger Level Analysis

Elementary particles and their interactions are successfully described by the Standard Model of particle physics (SM). However, it has been observed that extensions Beyond the Standard Model (BSM) are required to account for a large part of yet undiscovered particles and interactions, such as Dark Matter (DM). To advance the knowledge of the SM and to pursue DM discoveries, CERN has the ambitious plan of further increasing the Large Hadron Collider's (LHC) energy and luminosity, thus reaching unprecedented event rates in the field of collider physics. This thesis is divided in three parts, dealing with some of the most challenging aspects of the ATLAS experiment at the LHC present and future activities. After a thorough review of the SM, BSM physics is outlined, with particular attention to DM searches. The second part of this work addresses the issue of coping with the foreseen event rates of the High-Luminosity LHC (HL-LHC) phase. Indeed, optimizations of the existing Geant4 simulation codes are a crucial step to alleviate the need for new and expensive hardware resources. With the objective of improving the efficiency of the simulation tools, an extensive study on different compilers, different optimization levels and different build types is presented. In addition, a preliminary investigation on the geometry description of the ATLAS Transition Radiation Tracker (TRT) modules is discussed. The last part of the thesis covers the DM searches carried out by the ATLAS Trigger-object Level Analysis (TLA) group. These searches are based on the analysis of the invariant mass spectrum of di-jet events and, during LHC Run 2, have been performed at energies in the 450-1800 GeV range (integrated luminosity up to 29.3 fb-1 and center of mass energy of 13 TeV). After a review of the TLA studies, a preliminary investigation on the performance of Bayesian and Frequentist statistical tools is presented. In particular, the attention is focused on the interpretation and handling of systematic uncertainties both on background and DM signals. This is of particular importance in the process of finding localized excesses, which can indicate the existence of DM signals, and setting limits on the DM event cross sections.

Search for Dark Matter with ATLAS

This thesis describes in detail a search for weakly interacting massive particles as possible dark matter candidates, making use of so-called mono-jet events. It includes a detailed description of the run-1 system, important operational challenges, and the upgrade for run-2. The nature of dark matter, which accounts for roughly 25% of the energy-matter content of the universe, is one of the biggest open questions in fundamental science. The analysis is based on the full set of proton-proton collisions collected by the ATLAS experiment at the Large Hadron Collider at √s = 8 TeV. Special attention is given to the experimental challenges and analysis techniques, as well as the overall scientific context beyond particle physics. The results complement those of non-collider experiments and yield some of the strongest exclusion bounds on parameters of dark matter models by the end of the Large Hadron Collider run-1. Details of the upgrade of the ATLAS Central Trigger for run-2 are also included.

Development of New Approaches to ATLAS Detector Simulation and Dark Matter Searches with Trigger Level Analysiis