Théorie

Gamma-ray bursts (GRBs) and active galactic nuclei (AGNs) launch plasma outflows moving at nearly the speed of light from their central engines, and they create relativistic shock waves. Due to their extreme energies, these shocks are expected to accelerate particles via first-order Fermi acceleration, in which particles gain energy by repeatedly crossing the shock front, potentially explaining the origin of cosmic rays, especially ultra-high-energy cosmic rays. However, whether charged particles can indeed be efficiently accelerated at relativistic shocks remains under debate. A key condition for such acceleration is the presence of strong magnetic turbulence around the shock, which allows particles to be scattered and cross the shock multiple times. Yet, the physical mechanism that generates such intense turbulence is still poorly understood. In this study, we investigate a scenario where turbulence is driven downstream of the shock through the interaction between the shock front and density fluctuations pre-existing in the upstream medium. We perform 3D relativistic magnetohydrodynamic (MHD) simulations to model the turbulent structure, followed by test-particle simulations to trace particle trajectories in the resulting fields. Our results show that strong magnetic turbulence develops in the downstream region, where the magnetic field is amplified by the turbulence through the small-scale dynamo, leading to efficient shock acceleration by particle scattering. In my talk, we will discuss the conditions under which particle acceleration occurs, the nature of acceleration within magnetic turbulence, and the properties of the generated turbulence in comparison with GRB observations.

Core-collapse supernovae host the densest neutrino environments in the Universe. In these extreme conditions, neutrinos exchange flavor with each other more rapidly than they stream out of the star, forming a collective many-body system whose dynamics are dominated by weak interactions. In this talk, I will present a unified theoretical framework showing that this system behaves as a neutrino plasma, supporting collective flavor-wave excitations—flavomons—that are the exact analogs of collective electromagnetic fields in ordinary plasmas. The stimulated emission of flavomons produces an instability, with consequent restructuring of the neutrino flavor distribution which impacts the supernova explosion and multi-messenger signature. I will show that the conditions for triggering these instabilities naturally descend from detailed balanc of neutrino-flavomon interactions. By this approach, I will show that the accretion-powered neutrino outflow of supernovae generically develops neutrino-mass-induced instabilities immediately after the neutronization burst.

High-energy physics often motivates multi-field inflationary scenarios where stochastic effects play a crucial role. Peculiar to multi-field models, the noise-induced centrifugal force results in a longer duration of inflation depending on the number of fields, even when the stochastic noises themselves are small. We show that, in such small-noise regimes, the number of fields generically discriminates whether inflation successfully terminates or lasts forever. Our results indicate that inflation with an extremely large number of fields may fail to realise our observable Universe.

I will present cuHARM, a general relativistic radiation magnetohydrodynamic solver optimized to exploit exascale computing facilities. After describing the core numerical strategies enabling efficient calculation for multi-node and multi-GPU architectures, I will discuss how radiative cooling changes the dynamics of magnetically arrested accretion disks. In the second part of the talk, I will detail how radiation and its feedback on the dynamics are modeled. In cuHARM, the specific intensity is discretized in space and momentum, and is evolved through the solution of the radiative transfer equation via the discrete ordinate method. This approach eliminates the need for a closure relation and enables to resolve the anisotropy of the specific intensity. Finally, I will present the first results obtained for a radiative accretion disk operating at 0.1 times the Eddington luminosity.

Almost four decades have passed since the generalization of vacuum Kerr solutions to higher dimensions in the form of Myers-Perry black holes, yet an exact solution generalizing their charged extension (Kerr-Newman) to higher dimensions remains unknown in Einstein-Maxwell theory. Likewise, an exact solution for charged multi-NUT spacetimes in higher dimensions is still missing. In the first part of my talk, considering the Kerr-Schild class in Einstein-Maxwell theory, I will discuss a "No-Go" result related to the charging of Myers-Perry black holes that we obtained in arXiv:2309.02900. In the second part, I will present a slightly stronger generalization of this "No-Go" result in the context of generalized Kerr-Schild spacetimes, which also addresses the charging of multi-NUT solutions. Time permitting, I will further discuss some of the different classes of solutions identified in our analyses.