L’intérêt majeur des travaux exposés dans cette thèse est d’explorer la chevelure des trous noirs dans des cadres plus généraux que celui de la Relativité Générale en tenant compte de la présence d’une constante cosmologique, de dimensions supplémentaires, de champs de matière exotiques ou de termes de courbure de rang plus élevé. Ces extensions de la Relativité Générale peuvent s’inscrire dans le cadre de la théorie des cordes. C’est en étudiant des extensions naturelles de la Relativité Générale que nous pouvons aussi mieux comprendre la théorie d’Einstein. Dans un premier temps, nous exposerons la théorie de la Relativité Générale avec notamment les principes sur lesquelles elle s’appuie et nous donnerons les éléments mathématiques dont nous avons besoin pour la suite. Puis, une première extension sera présentée avec l’introduction de dimensions supplémentaires et de champs de p-formes qui constituent la généralisation naturelle de l’interaction électromagnétique. Nous construirons dans ce cadre de nouvelles solutions statiques de trous noirs où les p-formes permettent de modeler la géométrie de l’horizon. Nous exposerons ensuite l’extension la plus générale de la théorie d’Einstein en dimension quelconque qui génère des équations du second ordre en la métrique : la théorie de Lovelock. Nous déterminerons dans ce contexte une large classe de solutions en dimension 6 pour laquelle la théorie se réduit à celle d’Einstein-Gauss-Bonnet avec toujours la présence de p-formes. Enfin, nous étudierons une généralisation de la Relativité Générale en dimension 4 dont la modification est induite par un champ scalaire couplé conformément à la gravitation. Nous exhiberons notamment une nouvelle solution de trou noir avec un horizon plat dans cette théorie en présence de champs axioniques. Pour clore cette thèse, l’aspect thermodynamique de ces théories gravitationnelles sera étudié ; ce qui permettra de déterminer la masse et les charges de ces nouvelles solutions et d’étudier des phénomènes de transitions de phase en présence d’un champ scalaire conforme. / The main interest of the work exposed in this thesis is to explore hairy black holes in a more general framework than General Relativity by taking into account the presence of a cosmological constant, of higher dimensions, of exotic matter fields or of higher curvature terms. These extensions to General Relativity can be derived in the context of String Theory. It is also by studying natural extensions to General Relativity that we can more deeply understand the theory of Einstein. Firstly, we will display the theory of General Relativity with its building blocks in particular and we will give the mathematical tools that we need afterwards. Then, a first extension will be detailed with the introduction of higher dimensions and p-form fields which constitute the natural generalization of the electromagnetic interaction. We will build in this framework new static black hole solutions where p-form fields allow to shape the geometry of the horizon. Secondly, we will present the general extension of Einstein theory in any dimension which produces second order field equations: Lovelock theory. We will determine in this context a large class of solutions in dimension 6 for which the theory is reduced to Einstein-Gauss-Bonnet theory with the presence of p-form fields. Thirdly, we will study a generalization of General Relativity in dimension 4 whose modification is induced by a conformally coupled scalar field. We will namely exhibit a new black hole solution with a flat horizon in the presence of axionic fields. To conclude this thesis, thermodynamical aspects of these gravitational theories will be studied. In this way, we will be able to determine the mass and the charges of these new solutions and we will examine phase transition phenomena in the presence of a conformally scalar field.

京都大学 / 新制・課程博士 / 博士(理学) / 甲第24869号 / 理博第4979号 / 新制||理||1711(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 向山 信治, 准教授 樽家 篤史, 教授 橋本 幸士 / 学位規則第4条第1項該当 / Doctor of Science / Kyoto University / DFAM

General relativity (GR) has stood as the most accurate description of gravity for the last 100 years, weathering a barrage of rigorous tests. However, attempts to derive GR from a more fundamental theory or to capture further physical principles at high energies has led to a vast number of alternative gravity theories. The individual examination of each gravity theory is infeasible and as such a systematic method of examining modified gravity theories is a necessity. Studying generic classes of gravity theories allows for general statements about observables to be made independent of explicit models. Take, for example, those models described by the Horndeski action, the most general class of scalar-tensor theory with at most second-order derivatives in the equations of motion, satisfying theoretical constraints. But these constraints alone are not enough for a given modified gravity model to be physically viable and hence worth studying. In particular, observations place incredibly tight constraints on the size of any deviation in the solar system. Hence, any modified gravity would have to mimic GR in such a situation. To accommodate this requirement, many models invoke screening mechanisms which suppress deviations from GR in regions of high density. But these mechanisms really upon non-linear effects and so studying them in complex models is mathematically complex. To constrain the space of actions of Horndeski type to those which pass solar-system tests, a set of conditions on the four free functions of the Horndeski action are derived which indicate whether a specific model embedded in the action possesses a GR limit. For this purpose, a new and surprisingly simple scaling method is developed, identifying dominant terms in the equations of motion by considering formal limits of the couplings that enter through the new terms in the modified gravity action. Solutions to the dominant terms identify regimes where nonlinear terms dominate and Einstein's field equations are recovered to leading order. Together with an efficient approximation of the scalar field profile, one can determine whether the recovery of Einstein's field equations can be attributed to a genuine screening effect. The parameterised post-Newtonian (PPN) formalism has enabled stringent tests of static weak-field gravity in a theory-independent manner. This is through parameterising common perturbations of the metric found when performing a post-Newtonian expansion. The framework is adapted by introducing an effective gravitational coupling and defining the PPN parameters as functions of position. Screening mechanisms of modified gravity theories can then be incorporated into the PPN framework through further developing the scaling method into a perturbative series. The PPN functions are found through a combination of the scaling method with a post-Newtonian expansion within a screened region. For illustration, we show that both a chameleon and cubic galileon model have a limit where they recover GR. Moreover, we find the effective gravitational constant and all PPN functions for these two theories in the screened limit. To examine how the adapted formalism compares to solar-system tests, we also analyse the Shapiro time delay effect for these two models and find no deviations from GR insofar as the signal path and the perturbing mass reside in a screened region of space. As such, tests based upon the path light rays such as those done by the Cassini mission do not constrain these theories. Finally, gravitational waves have opened up a new regime where gravity can be tested. To this end, we examine how the generation of gravitational waves are affected by theories of gravity with screening to second post-Newtonian (PN) order beyond the quadrupole. This is done for a model of gravity where the black hole binary lies in a screened region, while the space between the binary's neighbourhood and the detector is described by Brans-Dicke theory. We find deviations at both 1.5 and 2 PN order. Deviations of this size can be measured by the Advanced LIGO gravitational wave detector highlighting that our calculation may allow for constraints to be placed on these theories. We model idealised data from the black hole merger signal GW150914 and perform a best fit analysis. The most likely value for the un-screened Brans-Dicke parameter is found to be ω = -1:42, implying on large scales gravity is very modified, incompatible with cosmological results.