A Neutrino-Driven Core Collapse Supernova Explosion of a 15 M Star

We present results from an ab initio three-dimensional, multi-physics core collapse supernova simulation for the case of a 15 M progenitor. Our simulation includes multi-frequency neutrino transport with state-of-the-art neutrino interactions in the ray-by-ray approximation, and approximate general relativity. Our model exhibits a neutrino-driven explosion. The shock radius begins an outward trajectory at approximately 275 ms after bounce, giving the first indication of a developing explosion in the model. The onset of this shock expansion is delayed relative to our two-dimensional counterpart model, which begins at approximately 200 ms after core bounce. At a time of 441 ms after bounce, the angle-averaged shock radius in our three-dimensional model has reached 751 km. Further quantitative analysis of the outcomes in this model must await further development of the post-bounce dynamics and a simulation that will extend well beyond 1 s after stellar core bounce, based on the results for the same progenitor in the context of our two-dimensional, counterpart model. This more complete analysis will determine whether or not the explosion is robust and whether or not observables such as the explosion energy, 56Ni mass, etc. are in agreement with observations. Nonetheless, the onset of explosion in our ab initio three-dimensional multi-physics model with multi-frequency neutrino transport and general relativity is encouraging.

Recent Progress on Ascertaining the Core Collapse Supernova Explosion Mechanism

We have been working within the fundamental paradigm that core collapse supernovae (CCSNe) may be neutrino driven, since the first suggestion of this by Colgate and White nearly five decades ago. Computational models have become increasingly sophisticated, first in one spatial dimension assuming spherical symmetry, then in two spatial dimensions assuming axisymmetry, and now in three spatial dimensions with no imposed symmetries. The increase in the number of spatial dimensions has been accompanied by an increase in the physics included in the models, and an increase in the sophistication with which this physics has been modeled. Computation has played an essential role in the development of CCSN theory, not simply for the obvious reason that such multidimensional, multi-physics, nonlinear events cannot possibly be fully captured analytically, but for its role in discovery. In particular, the discovery of the standing accretion shock instability (SASI) through computation about a decade ago has impacted all simulations performed since then. Today, we appear to be at a threshold, where neutrinos, neutrino-driven convection, and the SASI, working together over time scales significantly longer than had been anticipated in the past, are able to generate explosions, and in some cases, robust explosions, in a number of axisymmetric models. But how will this play out in three dimensions? Early results from the first three-dimensional (3D), multi-physics simulation of the Oak Ridge group are promising. I will discuss the essential components of todays models and the requirements of realistic CCSN modeling, present results from our one-, two-, and three-dimensional models, place our models in context with respect to other efforts around the world, and discuss short- and long-term next steps.

Two- and Three-Dimensional Multi-Physics Simulations of Core Collapse Supernovae: A Brief Status Report and Summary of Results from the Oak Ridge Group

We summarize the results of core collapse supernova theory from one-, two-, and three-dimensional models and provide a snapshot of the field at this time. We also present results from the Oak Ridge group in this context. Studies in both one and two spatial dimensions define the necessary} physics that must be included in core collapse supernova models: a general relativistic treatment of gravity (at least an approximate one), spectral neutrino transport, including relativistic effects such as gravitational redshift, and a complete set of neutrino weak interactions that includes state-of-the-art electron capture on nuclei and energy-exchanging scattering on electrons and nucleons. Whether or not the necessarily approximate treatment of this physics in current models that include it is sufficient remains to be determined in the context of future models that remove the approximations. We summarize the results of the Oak Ridge groups two-dimensional supernova models. In particular, we demonstrate that robust neutrino-driven explosions can be obtained. We also demonstrate that our predictions of the explosion energies and remnant neutron star masses are in agreement with observations, although a much larger number of models must be developed before more confident conclusions can be made. We provide preliminary results from our ongoing three dimensional model with the same physics. Finally, we speculate on future outcomes and directions.

Ascertaining the Core Collapse Supernova Mechanism: An Emerging Picture?

Here we present the results from two sets of simulations, in two and three spatial dimensions. In two dimensions, the simulations include multifrequency flux-limited diffusion neutrino transport in the ray-by-ray-plus approximation, two-dimensional self gravity in the Newtonian limit, and nuclear burning through a 14-isotope alpha network. The three-dimensional simulations are model simulations constructed to reflect the post stellar core bounce conditions during neutrino shock reheating at the onset of explosion. They are hydrodynamics-only models that focus on critical aspects of the shock stability and dynamics and their impact on the supernova mechanism and explosion. In two dimensions, we obtain explosions (although in one case weak) for two progenitors (11 and 15 Solar mass models). Moreover, in both cases the explosion is initiated when the inner edge of the oxygen layer accretes through the shock. Thus, the shock is not revived while in the iron core, as previously discussed in the literature. The three-dimensional studies of the development of the stationary accretion shock instability (SASI) demonstrate the fundamentally new dynamics allowed when simulations are performed in three spatial dimensions. The predominant l=1 SASI mode gives way to a stable m=1 mode, which in turn has significant ramifications for the distribution of angular momentum in the region between the shock and proto-neutron star and, ultimately, for the spin of the remnant neutron star. Moreover, the three-dimensional simulations make clear, given the increased number of degrees of freedom, that two-dimensional models are severely limited by artificially imposed symmetries.