No description available.

No description available.

No description available.

La physique des collisions d'ions lourds réunit deux domaines : la physique nucléaire et la physique des particules. Les progrès expérimentaux de ces dernières années donne l'opportunité d'étudier la nouvelle matière créée dans les collisions d'ions lourds qui s'appelle le plasma de quarks et de gluons.L'état initial de deux ions qui se collisionnent est affecté par les fluctuations créées par les fonctions d'ondes des nucléons. Ces fluctuations provoquent l'anisotropie de moments de la matière hadronique observée par les détecteurs. Le système créé dans une collision se comporte comme un fluide, donc l'état initial est connecté avec l'état final par l'évolution hydrodynamique. L’hydrodynamique relativiste est utilisée pour décrire l'évolution du fluide créé dans les collisions d'ions lourds. Nos résultats combinés avec les données expérimentales permettent de contraindre l'etat initial donc de faire la «rétro-ingénierie» des collisions d'ions lourds.L'observable qui caractérise l'anisotropie des moments est le flot anisotrope v_n. On présente les premières mesures du premier coefficient de la distribution de Fourier v_1 pour l'accélérateur LHC. v_1 s'appelle le flot dirigé. On effectue aussi les premiers calculs de v_1 à partir de l’hydrodynamique visqueuse. On trouve que v_1 est moins dépendent de la viscosité que les coefficients v_2 et v_3 qui sont respectivement les flots elliptique et triangulaire. On présente aussi les prédictions de v_1 pour l'accélérateur RHIC. Ces résultats ont été confirmés plus tard par les mesures de v_1 par RHIC. On propose aussi deux méthodes pour contraindre les modèles d’état initial: avec les données de v_1 et les données de v_2 et v_3. Ces méthodes donnent l'unique possibilité de contraindre les modèles Monte Carlo d'état initial. A la fin de cette thèse on montre les perspectives de ce domaine et on étudie les corrélations entre les plans des évènements qui ont été mesurées récemment et qui pourraient faire la lumière sur les fluctuations de l'état initial. / Ultra-Relativistic heavy-ion physics is a promising field of high energy physics connecting two fields: nuclear physics and elementary particle physics. Experimental achievements of the last years have provided an opportunity to study the properties of a new state of matter created in heavy-ion collisions called quark-gluon plasma. The initial state of two colliding nuclei is affected by fluctuations coming from wave- functions of nucleons. These fluctuations lead to the momentum anisotropy of the hadronic matter which is observed by the detectors. The system created in the collision behaves like a fluid, so the initial state is connected to the final state via hydrodynamic evolution. In this thesis we model the evolution with relativistic viscous hydrodynamics. Our results, combined with experimental data, give non trivial constraints on the initial state, thus achieving "reverse engineering" of the heavy-ion collisions. The observable which characterizes the momentum anisotropy is the anisotropic flow vn. We present the first measurements of the first harmonic of the anisotropic flow called directed flow v1 in Pb-Pb collisions at the LHC. We then perform the first viscous hydrodynamic modeling of directed flow and show that it is less sensitive to viscosity than higher harmonics. Comparison of these experimental data with the modeling allows to extract the values of the dipole asymmetry of the initial state, which provides constraints on the models of initial states. A prediction for directed flow v1 in Au-Au collisions is also made for RHIC. We then perform a similar modeling of the second and third harmonics of the anisotropic flow, called respectively elliptic v2 and triangular v3 flow. A combined analysis of the elliptic and triangular flow data compared with viscous hydrodynamic calculations allows us to put constraints on initial ellipticity and triangularity of the system. These constraints are then used as a filter for different models of initial state. At the end of this thesis, we show perspectives in the studies of the initial state which are opened by recent measurements of event-plane correlations which could shed light on the initial state fluctuations.

No description available.

Ultra-Relativistic heavy-ion physics is a promising field of high energy physics connecting two fields: nuclear physics and elementary particle physics. Experimental achievements of the last years have provided an opportunity to study the properties of a new state of matter created in heavy-ion collisions called quark-gluon plasma. The initial state of two colliding nuclei is affected by fluctuations coming from wave- functions of nucleons. These fluctuations lead to the momentum anisotropy of the hadronic matter which is observed by the detectors. The system created in the collision behaves like a fluid, so the initial state is connected to the final state via hydrodynamic evolution. In this thesis we model the evolution with relativistic viscous hydrodynamics. Our results, combined with experimental data, give non trivial constraints on the initial state, thus achieving "reverse engineering" of the heavy-ion collisions. The observable which characterizes the momentum anisotropy is the anisotropic flow vn. We present the first measurements of the first harmonic of the anisotropic flow called directed flow v1 in Pb-Pb collisions at the LHC. We then perform the first viscous hydrodynamic modeling of directed flow and show that it is less sensitive to viscosity than higher harmonics. Comparison of these experimental data with the modeling allows to extract the values of the dipole asymmetry of the initial state, which provides constraints on the models of initial states. A prediction for directed flow v1 in Au-Au collisions is also made for RHIC. We then perform a similar modeling of the second and third harmonics of the anisotropic flow, called respectively elliptic v2 and triangular v3 flow. A combined analysis of the elliptic and triangular flow data compared with viscous hydrodynamic calculations allows us to put constraints on initial ellipticity and triangularity of the system. These constraints are then used as a filter for different models of initial state. At the end of this thesis, we show perspectives in the studies of the initial state which are opened by recent measurements of event-plane correlations which could shed light on the initial state fluctuations.