Thermonuclear supernova simulations with stochastic ignition

We apply an ad hoc model for dynamical ignition in three-dimensional numerical simulations of thermonuclear supernovae assuming pure deflagrations. The model makes use of the statistical description of temperature fluctuations in the pre-supernova core proposed by Wunsch & Woosley (2004, ApJ, 616, 1102). Randomness in time is implemented by means of a Poisson process. We are able to vary the explosion energy and nucleosynthesis depending on the free parameter of the model which controls the rapidity of the ignition process. However, beyond a certain threshold, the strength of the explosion saturates and the outcome appears to be robust with respect to the number of ignitions. In the most energetic explosions, we find about $0.75~M_{\odot}$ of iron group elements. Other than in simulations with simultaneous multi-spot ignition, the amount of unburned carbon and oxygen at radial velocities of a few $10^{3}\,{\rm km\,s^{-1}}$ tends to be reduced for an ever increasing number of ignition events and, accordingly, more pronounced layering results.