Type Ia Supernova Diversity from 3-dimensional Models

We present results from a systematic study of the effects of initial parameters on three-dimensional thermonuclear supernova models. One major challenge to numerical models of thermonuclear supernova explosions is to reproduce the diversity observed among Type Ia supernovae (SNe Ia). With the rapid advancement of three-dimensional models over the past years, systematic studies have become possible testing the effect of changes in the initial parameters of simulations on the explosion process. The ultimate goal is, of course, to explain the correlation between the peak luminosity and the light curve shape used to calibrate cosmological distance measurements. We present the first systematic study on this issue based on three-dimensional deflagration models. Here, a Chandrasekhar-mass white dwarf (WD) star ignites nuclear reactions in the center which finally form a flame. In the deflagration model this flame burns outward with a subsonic velocity, which quickly becomes dominated by the effects of interaction with a turbulent velocity field. Caused by buoyancy (Rayleigh-Taylor) and secondary shear (Kelvin-Helmholtz) instabilities inherent in the scenario, turbulence wrinkles the surface of the flame effectively boosting its propagation speed. For a comprehensive review of SN Ia models we refer to Hillebrandt & Niemeyer (2000). For the three-dimensional simulations of the explosion process we apply the scheme developed by Reinecke et al. (1999) and Reinecke et al. (2002). The final composition of the ejecta in our models is obtained via a nucleosynthetic postprocessing procedure on the basis of tracer particles advected in the explosion simulation (for details see Travaglio et al. 2004). There is a number of possible parameters of the deflagration model that have the potential to explain the SN Ia diversity. The progenitor’s carbonto-oxygen (C/O) ratio, its metallicity, and the central density at ignition are commonly suggested, but other effects like rotation and flame ignition could also play a role. In our survey we concentrate on the former three parameters. These are varied independently. Of course, this is an oversimplification since in reality they are interrelated by stellar evolution of the progenitor WD star. Nevertheless, in this first study our goal is to infer the trends of effects of each individual parameter on the explosion. In the setup of the explosion models we apply three different carbon mass fractions of the WD material, X(C) = 0.30, 0.46, 0.62, and three different central densities at ignition, ρc = [1.0, 2.6, 4.2] × 10 9 g cm. Three different metallicities of the WD, Z = [0.3, 1.0, 3.0]Z⊙ , are represented by the Ne mass fractions (see Timmes et al. 2003). This defines the 27 models of our survey.