We present a systematic study of the diversity of three-dimensional deflagration simulations of Type Ia supernova explosions arising from variations of the initial parameters. By altering the carbon mass fraction, the central density, and the metallicity of the progenitor white dwarf star, we set up a grid of numerical explosion models. While changing the central density has the largest impact on the explosion energy, the largest variation in the 56Ni production is found by changing the metallicity of the models. Varying the carbon mass fraction hardly affects the 56Ni synthesized although it alters the energetics of the explosion. Possible consequences for the shape of light curves of Type Ia supernovae are discussed.

We present results from a systematic study of the effects of initial parameters on three-dimensional thermonuclear supernova models.

We present synthetic bolometric and broad-band UBVRI light curves of SNe Ia, for four selected 3-D deflagration models of thermonuclear supernovae. The light curves are computed with the 1-D hydro code STELLA, which models (multi-group time-dependent) non-equilibrium radiative transfer inside SN ejecta. Angle-averaged results from 3-D hydrodynamical explosion simulations with the composition determined in a nucleosynthetic postprocessing step served as the input to the radiative transfer model. The predicted model UBV light curves do agree reasonably well with the observed ones for SNe Ia in the range of low to normal luminosities, although the underlying hydrodynamical explosion models produced only a modest amount of radioactive Ni56 and relatively low kinetic energy in the explosion. The evolution of predicted B and V fluxes in the model with a Ni56 mass of 0.42 M_sun follows the observed decline rate after the maximum very well, although the behavior of fluxes in other filters somewhat deviates from observations, and the bolometric decline rate is a bit slow. Using our models, we check the validity of Arnett's rule and the accuracy of the procedure for extracting the Ni56 mass from the observed light curves. We find that the comparison between theoretical light curves and observations provides a useful tool to validate SN Ia models. The steps necessary to improve the agreement between theory and observations are set out.

The use of type Ia supernovae as distance indicators for cosmology has initiated a search for theoretical arguments supporting the empirical calibration methods applied. To this end, as a first step, a sound understanding of the origin of the observed diversity in type Ia supernova properties is needed. Here we present a first systematic study of effects resulting from changing some physical parameters of three-dimensional deflagration models of thermonuclear supernovae. In our study we vary the progenitor's carbon-to-oxygen ratio and its central density prior to ignition because both properties are not well determined by stellar evolution theory and they may change from supernova to supernova. Next we compute for these explosion models the nucleosynthesis yields in a post-processing step. This, in addition, allows us to study variations in the progenitor's metallicity by means of different $^{22}$Ne mass fractions in the initial composition. We find that the progenitor's carbon-to-oxygen ratio and its central density affect the energy release of the models and thus the expansion velocity of the supernova. Moreover, we find that changing the metallicity and the central density changes the production of radioactive $^{56}$Ni and thus affects the luminosity. In contrast, the carbon-to-oxygen ratio has little effect on the $^{56}$Ni production. Implications of the found variations of the explosion energy and the produced $^{56}$Ni mass for the type Ia supernova diversity are discussed.