Turbulence, transport and confinement: from tokamaks to star magnetism

This thesis is part of the general study of self-organization in hot and magnetized plasmas. We focus our work on two specific objects: stars and tokamaks. We use first principle numerical simulations to study turbulence, transport and confinement in these plasmas. The first part of this thesis introduces the main characteristics of stellar and tokamak plasmas. The reasons for studying them together are properly detailed. The second part is focused on stellar aspects. We study the interactions between the 3D turbulent motions in the solar convection zone with an internal magnetic field in the tachocline (the transition region between the instable and stable zones in the Sun). The tachocline is a very thin layer (less than five percent of the solar radius) that acts as a transport barrier of angular momentum. We show that such an internal magnetic field is not likely to explain the observed thickness of the tachocline and we give some insights on how to find alternative mechanisms to constrain it. We also explore the effect of the environment of star on its structure. We develop a methodology to study the influence of stellar wind and of the magnetic coupling of a star with its orbiting planets. We use the same methodology to analyse the magnetic interaction between a stellar wind and a planetary magnetosphere that acts as a transport barrier of matter. Then, the third part is dedicated to fusion oriented research. We present a numerical investigation on the experimental mechanisms that lead to the development of transport barriers in the plasma. These barriers are particularly important for the design of high performance fusion devices. The creation of transport barriers is obtained in turbulent first principle simulations for the very first time. The collaboration between the two scientific teams lead to the results presented in the fourth part of this thesis. An original spectral method is developed to analyse the saturation of stellar convective dynamos and of the ion temperature gradient instability. We also develop a reduced model that tackles the interaction between turbulent motions, waves, and zonal and mean flows. The same model is applied to both the solar tachocline and tokamak turbulence, putting forth the strong analogy that exists between the two objects we studied. (author)