Mémoire numérisé par la Division de la gestion de documents et des archives de l'Université de Montréal.

Dans cette thèse nous présentons la conception, l'étude des performances et les premiers tests, effectues au Cosmic Muon Telescope situe au SLAC, d'un nouveau détecteur d'identification des particules émises dans la région ''avant'' du détecteur SuperB.Ce détecteur est base une technique de temps de vol (TOF). Pour identifier les particules avec une impulsion jusqu'à 3GeV/c et une distance de vol de l'ordre de deux mètres nous avons besoin d'un détecteur TOF capable de mesurer le temps avec une précision typique de 30 ps. Pour atteindre cet objectif nous avons conçu un composant pour lequel le passage d'une particule chargée produit de la lumière Cherenkov dans un ''fused silica'' (quartz) radiator qui est ensuite détectée par des photodétecteurs tres rapides et une électronique rapide dédiée. Nous l'appelons détecteur DIRC-like TOF.Les photodétecteurs HAMAMATSU SL-10 MCP-PMT ont été caractérises sur faisceau de test au LAL et la résolution en temps d'environ 37 ps a été mesurée. La nouvelle électronique 16-canaux USB WaveCatcher développée au LAL(CNRS/IN2P3) et CEA/IRFU montre un jitter de moins de 10 ps. La géometrie du détecteur a quartz a été étudiée avec une attention particulière a l'aide d'une simulation Geant4. Celle-ci montre que la meilleure géométrie permet d'atteindre une résolution en temps d'environ 90 ps par photoélectron avec au moins 10 photoélectrons détectés, donnant en moyenne la résolution totale désirée de 30 ps.Nous avons construit un prototype d'un tel composant, utilisant les barres de quartz utilisées pour l'expérience Babar et nous l'avons installe dans le Cosmic Ray Telescope au SLAC. Une résolution en temps d'environ 70~ps par photoélectron a été obtenue, en accord avec la simulation.Cette preuve de principe a convaincu la Collaboration SuperB d'adopter un tel composant comme solution de base pour l'identification des particules émises vers l'avant dans SuperB. Le point délicat, encore ouvert, est celui de la résistance de ce détecteur aux bruits de fond de la machine.Dans cette thèse nous présentons aussi les études préliminaires de différents types de bruit de fond et leur effet sur les performances du détecteur DIRC-like TOF. Le processus Bhabha radiatif est de loin la source dominante de bruit de fond. Le taux de photoélectrons de bruit de fond principalement du aux gammas d'énergie d'environ 1.4 MeV est estimée a ~480 kHz/cm^2 ce qui correspond a 2 C/cm^2 de charge d'anode integrée sur 5 ans. Le flux de neutrons traversant l'électronique de front end du détecteur DIRC-like TOF est estimée a ~10^11/cm^2/year. Ces résultats préliminaires sont rassurants. / In this thesis, we present the conception, the performances studies and the first tests in the Cosmic Muon Telescope situated at SLAC of a new detector for the particle identification in the forward region of the SuperB detector.This detector is based on time-of-flight (TOF) technique. To identify the particles with momentum up to 3 GeV/c and flight base around two meters we need a TOF detector able to measure the time with a precision of about 30 ps. To achieve this goal we have conceived a device producing Cherenkov light in a fused silica (quartz) radiator, by a charged particle, which then detected with very fast photodetectors and dedicated ultrafast electronics. We call it, the DIRC-like TOF detector.For what concern the photodetectors, the HAMAMATSU SL-10 MCP-PMT has been characterized at LAL test bunch and the time resolution of about 37 ps has been measured. The new 16-channel USB WaveCatcher electronics developed by LAL (CNRS/IN2P3) and CEA/IRFU has shown to have a jitter of less than 10 ps. The geometry of the quartz detector has been then carefully studied with Geant4 simulation. Which shows that the best detector geometry allow to reach the time resolution of about 90 ps per photoelectron with at least 10 photoelectrons detected, giving in average the desired 30 ps total time resolution.We have constructed a prototype of such device, using the quartz bars available from the Babar experiment, and we have installed it, in the SLAC Cosmic Ray Telescope. A time resolution of about 70 ps per photoelectron was obtained, in agreement with simulation.This proof-of-principle has convinced the SuperB Collaboration to adopt such a device as the baseline for the SuperB particle identification detector in the forward region. The delicate point which is still opened is the resistance of this detector to the machine background.In this thesis we also present preliminary studies of different types of background and their effect on the performances of the DIRC-like TOF detector. Radiative Bhabha process is by far the dominant source of background. The rate of the background photoelectrons caused mainly by the gammas with energy around 1.4 MeV is estimated to be ~480 kHz/cm^2 which corresponds to 2 C/cm^2 of integrated anode charge in 5 years. The neutron flux thought the DIRC-like TOF front end electronics is ~10^11/cm^2/year. These preliminary results are reassuring.

In dieser Arbeit wird ein erster Versuch vorgestellt, die inklusive Produktion von neutralen Pionen durch ihren Zerfall in zwei Photonen zu messen. Dazu wurde der HERA-B Detektor in der Konfiguration des Frühjahrs `99 benutzt. Für die Bestimmung des Produktionsquerschnittes der neutralen Pionen werden die Daten aus dem inneren Bereich des elektromagnetischen Kalorimeters verwendet. Die Messung wurde mit zwei verschiedenen Target-Materialien, Kohlenstoff und Titan, durchgeführt. Aus diesem wird die Materialabhängigkeit der Pionen-Produktion ermittelt. Mehrere Methoden zur Bestimmung der Luminosität bei HERA-B werden untersucht. Es wird gezeigt, dass die durchschnittliche Anzahl von Wechselwirkungen mit Hilfe der im elektromagnetischen Kalorimeter deponierten Energie gemessen werden kann. Diese Methode bietet eine Möglichkeit zur Luminositätsbestimmung mit hoher Genauigkeit. Die Durchführbarkeit, die hadronische Produktion von Leptonpaaren in hohen Massenbereichen mit HERA-B zu messen, wird in Monte-Carlo-Simulationen untersucht. Diese zeigen, dass die erwartete Massenauflösung eine Beobachtung des Ypsilon-Signals über dem Untergrund und eine Trennung der Zustände Ypsilon (1S) und Ypsilon (2S) erlaubt. Es wird gezeigt, dass eine genaue Messung der Ypsilon-Polarisation möglich ist. Diese Messung bietet eine sehr gute Möglichkeit, die verschiedenen Mechanismen für die Produktion schwerer Quarkonium-Zustände zu testen. / A first attempt to measure the cross section of inclusive production of neutral pions decaying into two photons using the HERA-B detector in the spring of '99 configuration is presented. The pions cross section is determined using data from the inner part of the electromagnetic calorimeter for two types of target material, carbon and titanium. From the data on C and Ti the nuclear dependence of pion production was extracted. Various approaches for the luminosity determination at HERA-B are investigated. The average number of interactions in the target can be determined from the energy deposition in the electromagnetic calorimeter. This method provides a high accuracy in the luminosity determination. The feasibility to study hadronic production of dileptons in the high-mass region at HERA-B is investigated with a Monte Carlo simulation. The predicted mass resolution of Upsilon mesons will allow the clear observation of the Upsilon mass peak above the background, and to separate the Upsilon(1S) and Upsilon(2S) states. The possibility to perform an accurate measurement of polarization of Upsilon mesons is demonstrated. This measurement provides an excellent opportunity to test different mechanisms of heavy quarkonium production.

This thesis will present results on the measurement of Z boson production in the electron channel associated with b-jets in proton-proton collision at 13 TeV, a measurement of fundamental interest in precision measurements and physics searches. This measurement represents important precision tests of perturbative QCD and provides information related to the b-quark inside of the proton. This measurement also represents one of the main sources of background in top quark production studies, Higgs precision measurements and the search for supersymmetric particles. The results on cross section measurements and kinematic distributions presented in this work have been unfolded to particle level and are compared to the four and five flavor schemes in Monte Carlo generators, showing that the predictions are consistent within the experimental uncertainties. In this thesis it will be also shown results of the development of the new event selection system proposed for high luminosity measurements. This system will be implemented during the second LHC long shutdown (2019-2020), which involves studies on the reconstruction of electrons using the concept of supercells, where the fine granularity of the cells of the ATLAS Liquid Argon detector is exploited to mitigate the problems to be found in high luminosity conditions. One of the problems is the increase of pile-up, caused by secondary collisions of protons in the same event, generating a large number of low pT jets that can be erroneously identified as electrons. The studies in this work introduce new discriminants based on the electromagnetic shower shape, showing that is possible to further reduce the rate of low pT jets maintaining a good performance on electron reconstruction. / Esta tese apresentará resultados da medida da produção de bósons Z no canal de elétrons associado com jatos b em colisões próton-próton a 13 TeV, uma medida de fundamental interesse em medidas de precisão e buscas por nova física. Essa medida representa importantes testes de precisão da QCD perturbativa e prove informações relacionadas ao quark b dentro do próton. Essa medida também representa uma das principais fontes de background na produção do quark top, medidas de precisão do Higgs e na busca de partículas supersimétricas. Os resultados na medida da seção de choque e distribuições cinemáticas apresentadas neste trabalho são comparados com os esquemas de quatro e cinco sabores nos geradores Monte Carlo, mostrando que as predições são consistentes dentro das incertezas experimentais. Nesta tese também serão apresentados estudos do desenvolvimento de um sistema de seleção de eventos proposto para medidas em alta luminosidade. Esse sistema será implementado durante o segundo LHC long shutdown (2019-2020), envolvendo estudos na reconstrução de elétrons usando o conceito de supercélulas, onde a fina granularidade das células do detector de Argônio Líquido do ATLAS será aproveitada para mitigar as dificuldades que se terão em condições de alta luminosidade. Um dos problemas é o aumento do empilhamento (pile-up), causado por colisões secundárias de prótons no mesmo evento, gerando uma grande quantidade de jatos de baixo pT que podem ser erroneamente identificados como elétrons. Os estudos neste trabalho apresentam novos discriminantes com base na forma da cascata eletromagnética, mostrando que é possível reduzir ainda mais a taxa de jatos de baixo pT mantendo uma boa reconstrução de elétrons.

The future International Linear Collider (ILC) and Compact Linear Collider (CLIC) are intended for precision measurements of phenomena discovered at the Large Hadron Collider (LHC) and also for the discovery of new physics. In order to offer optimal conditions for such experiments, the new colliders must produce very-high-luminosity collisions at energies in the TeV regime. Emittance growth caused by imperfections in the main linacs is one of the factors limiting the luminosity of CLIC and ILC. In this thesis, various emittance preservation and luminosity tuning techniques have been tested and developed in order to meet the challenging luminosity requirements. Beam-based alignment was shown to be insufficient for reduction of emittance growth. Emittance tuning bumps provide an additional powerful preservation tool. After initial studies of tuning bumps designed to treat certain imperfections, a general strategy for design of optimised bumps was developed. The new bumps are optimal both in terms of emittance reduction performance and convergence speed. They were clearly faster than previous bumps and reduced emittance growth by nearly two orders of magnitude both for CLIC and ILC. Time-dependent imperfections such as ground motion and magnet vibrations also limit the performance of the colliders. This type of imperfections was studied in detail, and a new feedback system for optimal reduction of emittance growth was developed and shown to be approximately ten times more efficient than standard trajectory feedbacks. The emittance tuning bumps require fast and accurate diagnostics. The possibility of measuring emittance using a wide laserwire was introduced and simulated with promising results. While luminosity cannot be directly measured fast enough, it was shown that a beamstrahlung tuning signal could be used for efficient optimisation of a number of collision parameters using tuning bumps in the Final Focus System. Complete simulations of CLIC emittance tuning bumps, including static and dynamic imperfections and realistic tuning and emittance measurement procedures, showed that an emittance growth six times lower than that required may be obtained using these methods.

No description available.

Les seize détecteurs MPX constituant le réseau ATLAS-MPX ont été placés à différentes positions dans le détecteur ATLAS et sa averne au CERN dans le but de mesurer en emps réel les champs de radiation produits ar des particules primaires (protons des faisceaux) et des particules secondaires (kaons, pions, g, protons) issues des collisions proton-proton. Des films de polyéthylène (PE) et de fluorure de lithium (6LiF) recouvrent les détecteurs afin d’augmenter leur sensibilité aux neutrons produits par les particules primaires et secondaires interagissant avec les matériaux présents dans l’environnement d’ATLAS. La reconnaissance des traces laissées par les particules dans un détecteur ATLAS-MPX se fait à partir des algorithmes du logiciel MAFalda (“Medipix Analysis Framework”) basé sur les librairies et le logiciel d’analyse de données ROOT. Une étude sur le taux d’identifications erronées et le chevauchement d’amas a été faite en reconstruisant les activités des sources 106Ru et 137Cs. L’efficacité de détection des neutrons rapides a été mesurée à l’aide des sources 252Cf et 241AmBe (neutrons d’énergie moyenne de 2.13 et 4.08 MeV respectivement). La moyenne des efficacités de détection mesurées pour les neutrons produits par les sources 252C f et 241AmBe a été calculée pour les convertisseurs 6LiF et PE et donnent (0.8580 ± 0.1490)% et (0.0254 ± 0.0031)% pour LiF et (0.0510 ± 0.0061)% et (0.0591 ± 0.0063)% pour PE à bas et à haut seuil d’énergie respectivement. Une simulation du calcul de l’efficacité de détection des neutrons dans le détecteur MPX a été réalisée avec le logiciel GEANT4. Des données MPX correspondant aux collisions proton-proton à 2.4 TeV et à 7 TeV dans le centre de masse ont été analysées. Les flux détectés d’électrons et de photons sont particulièrement élevés dans les détecteurs MPX01 et MPX14 car ils sont plus près du point de collision. Des flux de neutrons ont été estimés en utilisant les efficacités de détection mesurées. Une corrélation avec la luminosité du LHC a été établie et on prédit que pour les collisions à 14 TeV dans le centre de masse et avec une luminosité de 10^34 cm-1*s-1 il y aura environ 5.1x10^8 ± 1.5x10^7 et 1.6x10^9 ± 6.3x10^7 particules détectées par les détecteurs MPX01 et MPX14 respectivement. / The sixteen detectors forming the ATLAS-MPX network have been placed in different positions inside the ATLAS detector and its cavern at CERN in order to measure, in real time, the radiation fields produced by primary particles (beam protons) and secondary particles (kaons, pions, photons, protons) resulting from the proton-proton collisions. Films of polyethylene (PE) and lithium fluoride (6LiF) cover the detectors so as to increase their sensitivity to neutrons produced by the primary and secondary particles interacting with the materials present in the ATLAS environment. The tracks identification in an ATLAS-MPX detector is obtained with the algorithms of the MAFalda software (Medipix Analysis Framework) based on the libraries and data-analysis software ROOT. A study on the mistag rate and blob overlap was made by reconstructing the activities of 106Ru and 137Cs sources. The fast neutron detection efficiency was measured with the help of 252C f and 241AmBe sources (neutrons with an average energy of 2.13 and 4.08 MeV, respectively). The detection efficiency measured for neutrons produced by 252Cf et 241AmBe sources was calculated for the 6LiF and PE converters. It averaged at low and high energy respectively (0.8580 ± 0.1490)% and (0.0254 ± 0.0031)% for LiF and (0.0510 ± 0.0061)% and (0.0591 ± 0.0063)% for PE. A simulation of the neutron detection efficiency calculation in the MPX detector was carried out with the GEANT4 software. MPX data corresponding to the proton-proton collisions at 2.4 TeV and 7 TeV at the center of mass were analyzed. The detected flux of electrons and photons are particularly high in the MPX01 and MPX14 detectors because they are closer to the point of collision. Fluxes of neutrons were estimated using the measured detection efficiencies. A correlation with the luminosity of the LHC was established. We predict that for 14 TeV collisions at the center of mass, with a luminosity of 1034 cm^2*s^1, the number of particle detected by MPX01 and MPX14 respectively will be about 5.1x10^8 ± 1.5x10^7 and 1.6x10^9 ± 6.3x10^7.

Mémoire numérisé par la Division de la gestion de documents et des archives de l'Université de Montréal

This thesis will present results on the measurement of Z boson production in the electron channel associated with b-jets in proton-proton collision at 13 TeV, a measurement of fundamental interest in precision measurements and physics searches. This measurement represents important precision tests of perturbative QCD and provides information related to the b-quark inside of the proton. This measurement also represents one of the main sources of background in top quark production studies, Higgs precision measurements and the search for supersymmetric particles. The results on cross section measurements and kinematic distributions presented in this work have been unfolded to particle level and are compared to the four and five flavor schemes in Monte Carlo generators, showing that the predictions are consistent within the experimental uncertainties. In this thesis it will be also shown results of the development of the new event selection system proposed for high luminosity measurements. This system will be implemented during the second LHC long shutdown (2019-2020), which involves studies on the reconstruction of electrons using the concept of supercells, where the fine granularity of the cells of the ATLAS Liquid Argon detector is exploited to mitigate the problems to be found in high luminosity conditions. One of the problems is the increase of pile-up, caused by secondary collisions of protons in the same event, generating a large number of low pT jets that can be erroneously identified as electrons. The studies in this work introduce new discriminants based on the electromagnetic shower shape, showing that is possible to further reduce the rate of low pT jets maintaining a good performance on electron reconstruction. / Esta tese apresentará resultados da medida da produção de bósons Z no canal de elétrons associado com jatos b em colisões próton-próton a 13 TeV, uma medida de fundamental interesse em medidas de precisão e buscas por nova física. Essa medida representa importantes testes de precisão da QCD perturbativa e prove informações relacionadas ao quark b dentro do próton. Essa medida também representa uma das principais fontes de background na produção do quark top, medidas de precisão do Higgs e na busca de partículas supersimétricas. Os resultados na medida da seção de choque e distribuições cinemáticas apresentadas neste trabalho são comparados com os esquemas de quatro e cinco sabores nos geradores Monte Carlo, mostrando que as predições são consistentes dentro das incertezas experimentais. Nesta tese também serão apresentados estudos do desenvolvimento de um sistema de seleção de eventos proposto para medidas em alta luminosidade. Esse sistema será implementado durante o segundo LHC long shutdown (2019-2020), envolvendo estudos na reconstrução de elétrons usando o conceito de supercélulas, onde a fina granularidade das células do detector de Argônio Líquido do ATLAS será aproveitada para mitigar as dificuldades que se terão em condições de alta luminosidade. Um dos problemas é o aumento do empilhamento (pile-up), causado por colisões secundárias de prótons no mesmo evento, gerando uma grande quantidade de jatos de baixo pT que podem ser erroneamente identificados como elétrons. Os estudos neste trabalho apresentam novos discriminantes com base na forma da cascata eletromagnética, mostrando que é possível reduzir ainda mais a taxa de jatos de baixo pT mantendo uma boa reconstrução de elétrons.

The Large Hadron Collider (LHC) at CERN is designed to search for new physics by colliding protons with a center-of-mass energy of 13 TeV. The ATLAS detector is a multipurpose particle detector built to record these proton-proton collisions. In order to improve sensitivity to new physics at the LHC, luminosity increases are planned for 2018 and beyond. With this greater luminosity comes an increase in the number of simultaneous proton-proton collisions per bunch crossing (pile-up). This extra pile-up has adverse effects on algorithms for clustering the ATLAS detector's calorimeter cells. These adverse effects stem from overlapping energy deposits originating from distinct particles and could lead to difficulties in accurately reconstructing events. Machine learning algorithms provide a new tool that has potential to improve clustering performance. Recent developments in computer science have given rise to new set of machine learning algorithms that, in many circumstances, out-perform more conventional algorithms. One of these algorithms, convolutional neural networks, has been shown to have impressive performance when identifying objects in 2d or 3d arrays. This thesis will develop a convolutional neural network model for calorimeter cell clustering and compare it to the standard ATLAS clustering algorithm. / Graduate