Research And Development Of Thin-gap RPCs For ATLAS Upgrades

A Toroidal LHC Apparatus(ATLAS)is a general-purpose collider detector at the Large Hadron Collider(LHC).Since the commissioning of LHC in the year of 2008,ATLAS has collected in total 160 fb-1 integrated luminosity data at proton-proton collision energies of 7,8,and 13 TeV.Based on these data,ATLAS has discovered Higgs Boson and studied its properties,precisely measured processes of the Standard Model(SM),and searched for new physics that is beyond the SM(BSM),including searches for supersymmetric particles(SUSY)and dark matter particles.ATLAS detector consists of three major detecting subsystems:the inner detector,EM/hadron calorimeters,and the muon spectrometer(MS).Resistive plate chambers(RPCs)in the MS provide the trigger for selecting muons.The high-luminosity LHC(HL-LHC)upgrade will increase the instantaneous luminosity to about 5-7.5 × 1034 cm-2s-1,about a factor of five increased as compared to the original design value.An expected integrated luminosity of 4,500 fb-1 will extend the HL-LHC physics potential for the study of Higgs properties and searches for new physics BSM.To accommodate the higher luminosity and ageing of the detector,some ATLAS subdetectors have to be upgraded,of which thin-gap RPCs technology is proposed for the upgrade of the muon trigger system to enhance the trigger efficiency,geometrical acceptance,and trigger redundancy of the MS.This dissertation focuses on the research and development of thin-gap RPC designs and prototype tests regarding cluster size,efficiency,geometrical acceptance,and time resolution.The signal propagation process in thin-gap RPC is systematically studied through simulations and measurements of cluster size,signal attention,and crosstalk.Two unconventional readout methods are studied.The results demonstrated that new methods can increase geometrical acceptance and reduce readout channels,without obviously losing spatial resolutions.The time resolution of thin-gap RPCs is studied using deep learning,which significantly improves the time resolution.A better understanding of RPC detection mechanisms has been achieved through these studies,of which some have been adopted by the collaboration to optimize thin-gap RPC designs.Based on these studies,further technical improvements and potential applications in other disciplines are discussed.The dissertation is arranged as follows:Chapter 1 introduces the LHC machine and the ATLAS detector,including their major parameters,specific upgrade plans and a detailed discussion about the thin-gap RPCs and the new RPC system.Chapter 2 briefly summarizes particle detection history and interactions of particles.Chapter 3 consists of an overview of the development history of gaseous detectors and a review of RPC detectors.In Chapter 4,a novel finite-element simulation approach that models the prototype thin-gap RPC affected by the large cluster size problem is presented.This approach introduced the graphite layer into signal propagation simulations and established surface resistivity as one of the key parameters for the RPC design.According to simulations,the optimized thin-gap RPC prototype with a more resistive graphite layer produced a reasonable cluster size.Chapter 5 describes the study of signal propagation attenuation,including simulations and experimental test results.Details of the experiment setup and data analysis are discussed.The measured signal attenuation rates are in agreement with the simulation ones.Chapter 6 describes the studies of two new readout methods:the double-end and reflection readout methods.These two approaches reduce the readout channels and the dead areas due to the occupation by the front-end electronics.The detailed readout schemes,experimental setup,and results are discussed.Both readout methods show a good spatial resolution that meets the ATLAS requirements.Deep learning is introduced to investigate RPC time resolution in Chapter 7.The time of flight is used to label the sample with the help of a convenient data augmentation method.Details of the training are discussed,including the neural network structure,loss function,and improved time resolution.Finally,in Chapter 8,assembly and test works of thin-gap RPCs are described.The author contributed to the pre-production at CERN and developed an online monitor with a graphic user interface,which performs data acquisition and real-time analysis to accelerate cosmic ray tests for thin-gap RPC quality control.The author also established test systems for the front-end electronics and readout panels.