De nombreuses observations astrophysiques et cosmologiques tendent à montrer que l'essentiel de la matière de notre Univers est constitué de matière sombre. À l'échelle locale, la matière sombre serait rassemblée sous la forme d'un halo statique englobant la Voie Lactée.L'idée de la détection directionnelle est de mesurer la direction de reculs nucléaires produits par l'interaction avec les particules de matière sombre.Cette stratégie de détection pourrait alors mettre en évidence une corrélation entre la distribution angulaire des reculs nucléaires et le mouvement relatif du système solaire par rapport au halo permettant ainsi de découvrir et de contraindre les propriétés de la matière sombre.Dans ce contexte, la collaboration MIMAC a développé un détecteur gazeux permettant la mesure de la trace en 3D de reculs nucléaires.Afin de démontrer le potentiel de ce détecteur, un prototype bi-chambre a été installé au LSM en juin 2012.Cette thèse porte sur l'étude de la détection directionnelle avec le détecteur MIMAC selon quatre axes de recherche.Le premier axe concerne la caractérisation du détecteur avec l'étalonnage en énergie, la mesure du facteur de quenching et de la vitesse de dérive des électrons et la mise en place de la discrimination électron/recul.Le deuxième axe porte sur l'analyse des données expérimentales acquises au LSM. Ces donnés ont permis de réaliser la première mesure de traces en 3D de reculs de noyaux fils issus de la chaine du radon.Le troisième axe concerne la simulation du bruit de fond neutron au LSM avec un modèle de propagation des neutrons dans la caverne.Cela a permis d'estimer le taux d'événements neutron attendu et l'impact de la modélisation du fond neutrons sur la limite expérimentale.Enfin, le quatrième axe porte sur l'étude de l'impact des limites LHC sur la masse des squarks sur l'interaction entre les noyaux et la matière sombre. / A large number of astrophysical and cosmological observations support the fact that the matter component of our Universe is mainly composed by dark matter.At the local scale, a static and dense dark matter halo should surround the Milky Way.The directional detection idea is to measure the direction of nuclear recoils produced by the interaction with dark matter particles.This detection strategy could highlight a correlation between the angular distribution of nuclear recoils and the relative motion of the solar system respect to the galactic halo.Such signature opens the possibility to discriminate such rare events with respect to neutron background allowing the positive direct detection of dark matter.In this context, MIMAC collaboration has developed a detector allowing the measurement of 3D nuclear recoil tracks.Since June 2012, a bi-chamber prototype is installed at the LSM to demonstrate the potential of this detector.This Ph.D. thesis presents a study of dark matter directional detection with the MIMAC detector including four different aspects.The first one consist in the characterization of the detector describing the energy calibration, the measurement of the quenching factor and the electron drift velocity and the electron/recoil discrimination.The second one focuses on the analysis of experimental data acquired at the LSM. This study shows, for the first time, the observation of low energy 3D nuclear recoil tracks from the radon progeny.The third one describes the neutron background simulation at the LSM with a neutron propagation model showingthe expected neutron event rate and the impact on exclusion limits.Finally, the fourth one is related to the study of the impact of squarks masse limits from LHC results on supersymmetric particles interaction with quarks.

京都大学 / 新制・課程博士 / 博士(理学) / 甲第22993号 / 理博第4670号 / 新制||理||1670(附属図書館) / 京都大学大学院理学研究科物理学・宇宙物理学専攻 / (主査)准教授 吉岡 興一, 教授 田中 貴浩, 准教授 髙山 史宏 / 学位規則第4条第1項該当 / Doctor of Science / Kyoto University / DFAM

京都大学 / 新制・課程博士 / 博士(理学) / 甲第23000号 / 理博第4677号 / 新制||理||1671(附属図書館) / 京都大学大学院理学研究科物理学・宇宙物理学専攻 / (主査)教授 川合 光, 教授 田中 貴浩, 准教授 吉岡 興一 / 学位規則第4条第1項該当 / Doctor of Science / Kyoto University / DFAM

Population III (Pop III) stars are the first generation stars forming after the big bang from primordial gas. This dissertation is focused on the various processes that suppress and delay the formation of Pop III stars in the universe and their implications for the observations. We studied the impacts of the Lyman-Werner (LW) radiation that dissociates molecular hydrogen, baryon-dark matter streaming velocity introduced at recombination, ionizing radiation from nearby galaxies, and a model for the composition of dark matter known as the fuzzy dark matter on the formation of Pop III stars. Firstly, we take a closer look at the critical halo mass (Mcrit) that is the typical minimum dark matter halo mass needed to host cold dense gas to form the first stars using cosmological hydrodynamical simulations. LW radiation that dissociates molecular hydrogen and the baryon-dark matter streaming velocity both delay the formation of Pop III stars by increasing the critical halo mass. We describe our simulation suite with varying levels of LW radiation and streaming velocity to provide a fit for Mcrit as a function of LW radiation, streaming velocity, and redshift which can be used in semi-analytic models of early galaxy formation to make predictions for observations. Secondly, we explore a possible mechanism for the formation of large clusters of Pop III stars: a nearby ionizing source that ionizes a late forming halo, delaying its collapse until the halo is sufficiently large enough that the core can self-shield and suffer runaway collapse. We use numerical simulations to examine the fragmentation of the gas near the runaway collapse using the simple estimates and sink particles to show that the number of fragments is generally small, at most a handful, and that the mass accretion rate on the fragments is of order 10⁻³ Msun/yr. This rate is sufficiently high enough that the descent on the main sequence (and hence the suppression of accretion) is delayed until the stellar masses are of order 100-1000 Msun, but not high enough to produce direct collapse black holes of mass ~ 10⁵ Msun. The resulting clusters are larger than those produced in minihalos but are still likely to fall short of being easily detectable in James Webb Space Telescope blind fields. Finally, we investigate the formation of the first stars and galaxies in a fuzzy dark matter cosmology. Fuzzy dark matter, made up of ultra-light axions of mass ~ 10⁻²² eV, is a proposed alternative to the standard cold dark matter to solve its apparent small-scale problems. Its large de Broglie wavelength, of the order of kpc, results in the suppression of small-scale matter power, thus delaying the formation of the first stars and galaxies to lower redshift in much more massive halos. Therefore, first stars can be used to put very strong constraints on the mass of the fuzzy dark matter. We describe our cosmological simulations that accurately evolve the fuzzy dark matter distribution to study the formation of the first stars and galaxies.

An overwhelming majority of matter in the Universe is dark matter, a substance unlike anything we know. Detecting dark matter particles requires ruling out observed phenomena caused by known particles. This thesis advances efforts toward the detection of dark matter using one of the most sensitive particle detection technologies: the dual-phase liquid xenon time projection chamber. Specifically, data from the XENON1T Experiment, located in Italy, and the Purdue small-scale ASTERiX detector are analyzed. A background of Lead-214 beta decay events can be mitigated by tracing the radioactive Radon-222 decay chain in XENON1T. However, a preliminary reduction of background has a high cost to exposure. Research on several topics was conducted with Purdue undergraduates, including a search for dark matter particles up to the Planck Mass, characterizing backgrounds due to muons, and searching for Boron-8 solar neutrino signals. XENON1T single-scatter dark matter limits were extended to a particle mass of 10¹⁸GeV/c². The ASTERiX detector was modified to characterize a significant background to the smallest detectable energy signatures: single- and few-electron ionization signals. Infrared light was determined to be ineffective at reducing this background, and their rates were observed to decrease inversely with time since an energetic interaction according to a power law. The rates of single- and few- electron backgrounds increase linearly with increased applied extraction fields and increased depth of the initial interaction in the detector. These results indicate that these backgrounds originate at the liquid-gas interface of dual-phase detectors. In exploring a single-photon threshold for initial scintillation signals, a previously unconsidered background of large dark count signals in the photosensors became apparent. The high background of small ionization signals and large dark count signals deterred a search for Boron-8 solar neutrino interactions in XENON1T. These studies are vital to mitigating backgrounds and improving the sensitivity of liquid xenon time projection chambers to new physical phenomena.

FPGA-Centric Clusters (FCCs) with the FPGAs directly linked through their Multi-Gigabit Transceivers (MGTs) have a proven advantage over other commodity architectures for communication bound applications. To date, however, communication infrastructure for such clusters has generally only taken one of two simple approaches: nearest-neighbor-only, which is fast but of limited utility, and processor-based, which is general but slow. The overall problem addressed in this dissertation is the architecture, design, and implementation of communication networks for FCCs. These network designs should take advantage of the decades of design experience in networks for High-Performance Computing (HPC) clusters, but should also account for, and take advantage of, unique characteristics of FCCs, in particular, the configurability of the FPGAs themselves. This dissertation has seven parts. We begin with in-depth implementations of two model applications, Directional Dark Matter (DM) Detection, and Molecular Dynamics (MD). These implementations expose the necessary characteristics of FCC networks from physical through application layers. The second is the systematic exploration of communication microarchitecture for FCCs, as has been done previously for HPC clusters and for Networks on Chips (NoCs) on both FPGAs and ASICs. One outcome of this part is to find the properties of FCCs that substantially influence the router design space. Another outcome is to create a selection of candidate routers and generalize it so that it is parameterized by routing algorithm, arbitration policy, number of virtual channels (VCs), and other parameters. The third part is to use the proposed application-aware framework to evaluate the resulting design space with respect to a number of common communication patterns and packet sizes. The results from this part enable two sets of designs. One is the selection of an optimal router for a given resource budget that accounts for all the workloads. The other is to take advantage of FPGA reconfigurability to select the optimal router accounting for both resource budget and a particular workload. The fourth part is to evaluate the advantages of this approach of adapting the router design to the application. We find that the optimality of the router design varies significantly with workloads. We observe that compared with the router configuration with the best average performance, application-aware router selection can lead to substantial improvement in performance or reduction in resources required. The fifth part is application-specific optimizations in which we develop several modules and functional units that can provide specific optimizations for certain types of communication workloads depending on the application it going to serve. The sixth part explores topology emulation, e.g., when a three-dimensional network is used in the computation of an application that is logically two dimensional. We propose a generalized fold-and-cut mechanism that both preserves the locality in logical mapping, while also making use of the extra links provided by our 3D-torus fixture. The seventh part is a table-based static-scheduled router for applications with a static or persistent communication pattern. The router supports various cases, including unicast, multicast, and reduction. By making routing decisions a priori, we can bring better load-balance to network links and reduce congestion.

This work presents the search for Dark Matter particles associated with the Higgs Boson decaying into a b b-bar quark pair. The dark matter search result is based on proton-proton collision data collected at a center-of-mass energy of 13 TeV by the ATLAS detector during Run II. The results are interpreted in the context of a simplified model (Z’-2HDM) which describes the interaction of dark matter and standard model particles via new heavy mediator particles. The new powerful Higgs tagging techniques, which exploit the jet substructure and heavy flavor information to a large extent, are developed to improve the search sensitivity of the search. The target physics signals are signature with an optimized search region and interpreted with background estimation result statistically.

In this dissertation, we develop two novel methods for studying the nature of the Milky Way's dark matter halo. In both cases, we rely on the relationship between the dark matter halo's gravitational potential and the orbital structure it supports. The first method explores the morphology of stellar streams orbiting in non-spherical gravitational potentials. When globular clusters or dwarf galaxies fall into the Milky Way, tidal forces shred them into long filaments of stars called stellar streams. We show that in non-spherical potentials, stream morphologies are heavily dependent on the characteristics of the progenitor's orbit. Flattened axisymmetric galactic potentials, for example, are known to host minor orbit families surrounding special orbits with commensurable frequencies. The behavior of orbits that belong to these orbit families is fundamentally different from that of typical orbits with non-commensurable frequencies. We show that streams evolving near the boundaries, or separatrices, between orbit families, may become fanned out, develop a bifurcation, or both. We utilize perturbation theory to estimate the timescale of this effect and the likelihood of a stream evolving close enough to a separatrix to be affected by it. Next, we study the dynamical reasons for stream fanning and bifurcations near resonances, and find that each morphological outcome has a slightly different dynamical cause. Using a novel numerical approach for measuring the libration frequencies of resonant and near-resonant orbits, we reveal that fans come about due to a large spread in the libration frequencies near a separatrix, whereas bifurcations arise when a separatrix splits the orbital distribution of the stellar stream between two (or more) distinct orbit families. We then demonstrate how these features can arise in streams on realistic galactic orbits, in realistic galactic potentials, over timescales as short as 2-3 Gyr, and discuss how this might be used to constrain the global shape of the Milky Way's gravitational potential. The second method studied in this dissertation enables dynamical tests of a dark matter candidate known as Fuzzy (or Ultra-Light) Dark Matter. Our method relies on a wave generalization of the classic Schwarzschild approach for constructing self-consistent halos -- such a halo consists of a suitable superposition of waves instead of particle orbits, chosen to yield a desired mean density profile. As an illustration, we apply the method to spherically symmetric halos. We derive an analytic relation between the particle distribution function and the wave superposition amplitudes, and show how it simplifies in the high energy (WKB) limit. We verify the stability of such constructed halos by numerically evolving the Schrodinger-Poisson system. The proposed algorithm provides an efficient and accurate way to simulate the time-dependent halo substructures from wave interference, and to test how they will affect dynamical tracers or other observables in a galaxy. The dissertation concludes with a brief discussion of the future prospects of these two methods, especially in the context of upcoming ground- and space-based missions like Rubin LSST and the Roman Space Telescope.

Even though there is a multitude of observational evidence from cosmology and astrophysics, the standard model does not include a suitable dark matter candidate and therefore physics beyond the standard model is necessary. There are hypotheses of what the particle candidate could be coming from theories such as supersymmetry or extra dimensions. The processes producing these particles are understood very well from the theoretical perspective. The problem is that these processes have not been observed in any detectors. Therefore the nature of the dark matter remains unknown. However, it is clear that the dark matter-particle interacts with ordinary matter through gravity and in general, candidates may also interact through the weak force. These candidates are called Weakly Interacting Massive Particles. Therefore, by studying weak processes (weak in the sense that the processes are interactions mediated by a force weaker than the Electro-Magnetic and Quantum-Chromo dynamical-forces, not necessarily the weak force of the standard model) in the large hadron collider it may be possible to pose constraints on the dark matter signatures. One possible process which specifically involves the standard model electroweak interaction, which will be the model for the project, is the emission of scalar dark matter particles from the Z boson,which would result in a final state characterized by a Z boson and missing transverse energy. Simulations of the model and calculations of the cross section are done for different masses, ranging from 20−680 GeV, of the scalar dark matter particle and then compared to a standard model background process. Investigations are made whether or not it would be possible to detect darkmatter signals in the background. With the assumptions made, the results indicate that a signal from dark matter with a mass of around 40−150 GeV could not be rejected up to 5σ.

The Large Hadron Collider (LHC) is the most powerful and complex particle accelerator ever built. The ATLAS detector is a general-purpose particle detector at the LHC, designed to cover a wide range of physics measurements.This thesis presents two physics studies performed using data of proton-proton collisions collected with the ATLAS detector at √s = 8 TeV. The identication of jets originating from b quarks (b-tagging) is a crucial tool for many physics analyses at the LHC. A new technique to calibrate the efficiency of b-tagging algorithms using high transverse momentum jets is described. This technique allows to perform the calibration using jets with transverse momenta up to 1200 GeV, while the current calibration methods only reach approximately 300 GeV. The results of the calibration are presented.  The second study presented in this thesis is a search for Dark Matter (DM) production in association with a pair of heavy flavour quarks. A reinterpretation of the results of the ATLAS search for DM at √s = 8 TeV has been done based on simplied models. A set of simplied models with various DM masses, masses of the exchange particle, that mediates the interaction between DM and the regular matter, and couplings is considered. This study aims to choose benchmark models to be used in future searches at √s = 13 TeV. Additionally an accomplished technical project on the development of the b-tagging ATLAS software is presented.