Thermonuclear explosions of Chandrasekhar-mass C+O white dwarfs

First results of simulations are presented which compute the dynamical evolution of a Chandrasekhar-mass white dwarf, consisting of equal amounts of carbon and oxygen, from the onset of violent thermonuclear burning, by means of a new two-dimensional numerical code. Since in the interior of such a massive white dwarf nuclear burning progresses on microscopic scales as a sharp discontinuity, a so-called flamelet, which cannot be resolved by any numerical scheme, and since on macroscopic scales the burning front propagates due to turbulence, we make an attempt to model both effects explicitly in the framework of a finite-volume hydrodynamics code. Turbulence is included by a sub-grid model, following the spirit of large eddy simulations, and the well-localized burning front is treated by means of a level set, which allows us to compute the geometrical structure of the front more accurately than with previous methods. The only free parameters of our simulations are the location and the amount of nuclear fuel that is ignited as an initial perturbation. We find that models in which explosive carbon burning is ignited at the center remain bound by the time the front reaches low densities, where we stopped the computations because our description of combustion is no longer applicable. In contrast, off-center ignition models give rise to explosions which, however, are still too weak for typical Type Ia supernovae. Possible reasons for this rather disappointing result are discussed.