How a massive star ends its life depends upon how that life has been lived - the rotation, mass and composition it was born with, mass loss and exchange, and the complex convective and nuclear burning episodes it experienced along the way. In the end, the presupernova stellar core has a density structure that can be characterized by its "compactness" - essentially how fast the density declines outside the iron core. The likelihood that a massive star explodes, by any means, is sensitive to this compactness. It turns out, perhaps surprisingly, that the compactness is not a monotonic function of the star's birth mass, and, in some mass regions, whether the star explodes or not is almost random. Here the stellar physics underlying the development of compactness is explored for a fine grid of masses across a broad range of masses (9 - 120 solar masses). Using the model set generated, and with collaborators, the resulting explosions are explored assuming a neutrino-powered mechanism. Full isotopic nucleosynthesis, light curves, and remnant masses are calculated and found to be in good agreement with observations. Neglecting rotation, most stars above 20 Msun do not explode - though there are exceptions. In a related study, based on the same model set, the upper bound to the most luminous supernovae is explored considering various energy sources. The brightest possible supernova is found to be a rotationally powered explosion of a stripped core.

Chaotic fluid motion, known at the small scales as turbulence, can significantly alter the large-scale evolution of astrophysical events. For example, the growth and impact of convection to produce a successful core-collapse supernova (CCSN) depend upon the evolution of turbulence. The ideal way to investigate it in such an environment would be with high-resolution direct numerical simulations (DNS) that resolve the turbulent energy cascade down to some small scale where the energy could be safely assumed to dissipate as heat. Unfortunately, given the high Reynolds number and a large range of spatial scales, this is well beyond the current state-of-the-art computational 3D CCSN models. Since turbulence cannot be properly simulated in 1D or 2D, a subgrid-scale model (SGS) is needed to capture the unresolved 3D physics. Simple analytical SGS models often lack accuracy, though, and complex ones are difficult to tune, resulting in limited generalizability based on initial conditions. Given the recent successes in turbulence SGS modeling with Machine Learning (ML) in adjacent fields, this thesis develops an ML algorithm to analyze current simulations and study the features of turbulence in CCSN. To demonstrate its efficacy, we test our ML approach on modeling dynamic 3D HD & MHD turbulence, integrating the former into a 1D code to study the role of one specific feature of turbulence (the effective turbulent pressure) on the fate of CCSN simulations. Furthermore, our ML tools can be used to study the broader effects of turbulence, explore other ML architectures, and be integrated in the outside 1- and multi-dimensional CCSN codes. The ML framework (Sapsan) and its implementation into a 1-dimensional code (COLLAPSO1D) are open-sourced and available for public use.

The hot and dense proto-neutron star (PNS) born subsequent to core-collapse in a type II supernova explosion is an intense source of neutrinos of all flavors. It emits the 3-5 x 10^53 ergs of gravitational binding energy gained during collapse as neutrino radiation on a time scale of tens of seconds as it contracts, becomes increasingly neutron-rich and cools. While the supernova explosion mechanism and associated accretion of material is expected to influence the neutrino emission at early time (i.e. t <~ 1 s post bounce) the late time neutrino signal is shaped by the properties of the PNS, such as the nuclear equation of state (EoS), neutrino opacities in dense matter, and other microphysical properties that affect the cooling timescale by influencing either neutrino diffusion or convection. Detection of significant numbers of late time supernova neutrinos will provide a direct window into the properties of nuclear matter and neutron stars, if the neutrino signal can be modeled accurately. The average emitted neutrino energies also strongly affect nucleosynthesis in the neutrino driven wind, neutrino induced nucleosynthesis further out in the star, and the patterns of neutrino oscillations outside of the PNS.

This thesis examines a number of aspects of this environment. First, the equations of spherically symmetric general relativistic radiation hydrodynamics are discussed, a new code for calculating neutrino transport in PNSs is described, and first results from this code are presented. It is found that the NDW is neutron rich for at least a few seconds, in contrast to other recent work. This change in the expected wind electron fraction is traced to the correct treatment of the nucleon dispersion relations in an interacting medium and turns out to be influenced by the sub-nuclear density symmetry energy. Late time convection in PNSs is also studied. It is found that the density dependence of the symmetry energy may affect the duration of convective activity, which is imprinted in the neutrino luminosity evolution.

The second part of the thesis focuses on the neutrino driven wind (NDW) which is blown from the surface of the PNS. Time-dependent hydrodynamic calculations of the NDW are presented, which include accurate weak interaction physics coupled to a full nuclear reaction network. Using two published models of PNS neutrino luminosities,

predictions of the contribution of the NDW to the integrated nucleosynthetic yield of the entire supernova are made. For the neutrino luminosity histories considered, it is found no r-process occurs in the most basic wind scenario because the NDW entropy is too low, the dynamical timescale is too long, and the wind electron fraction is too high. It is possible that the wind produces the N = 50 closed shell isotopes, but this depends on the neutrino luminosities employed. The effect of a secondary heating source on the wind is then considered. The general characteristics of a secondary heating source required to produce r-process nucleosynthesis are discussed. Then gravitoacoustic power excited either by convection or g-mode oscillations of the PNS is considered as a possible source of this heating. It is found that this a viable mechanism for increasing the wind entropy and decreasing the the dynamical timescale to values that are favorable for the r-process, when the neutrino spectra found in the first part of the thesis are assumed.

Minor mixing processes—any fluid processes that mix material or transport heat other than convection or other large-scale flows—play a critical role in stellar evolution. These processes have been invoked through phenomenological models in order to explain away many issues in stellar evolution, such as the blue-red supergiant ratio problem and the progenitor problem of SN 1987A. We discuss our incompressible numerical simulations of one such process, thermohaline convection. We developed a semi-analytic formulation of the thermal and compositional transport of this process, which has been tested by other groups and implemented into MESA and the Toulouse–Geneva Evolution code. We also discuss compressible simulations of overshooting convection by Brummell et al. (2002) and the incompressible simulations of semi- convection by Wood et al. (2013), from which we developed one-dimensional models. We present some recent results of the effects of these models on a 15 solar mass star using the stellar evolution code KEPLER. These simulations put stricter limits on the extent of overshooting convection and the efficiency of semi-convection.

Type Ia supernovae (SNe Ia) are the largest thermonuclear explosions in the Universe. Their light output can be seen across great stances and has led to the discovery that the expansion rate of the Universe is accelerating. Despite the significance of SNe Ia, there are still a large number of uncertainties in current theoretical models. Computational modeling offers the promise to help answer the outstanding questions. However, even with today's supercomputers, such calculations are extremely challenging because of the wide range of length and time scales. In this paper, we discuss several new algorithms for simulations of SNe Ia and demonstrate some of their successes.

We present the discovery and monitoring of the optical transient (OT) associated with GRB 020410. The fading OT was found by Hubble Space Telescope (HST) observations taken 28 and 65 days after burst at a position consistent with the X-ray afterglow. Subsequent re-examination of early ground based observations revealed that a faint OT was present 6 hours after burst, confirming the source association with GRB 020410. A deep non-detection after one week requires that the OT re-brightened between day 7 and day 28, and further late timeHST data taken approximately 100 days after burst imply that it is very red (F_nu proportional to nu-2.7). We compare both the flux and color of the excess with supernova models and show that the data are best explained by the presence of a Type I b/c supernova at a redshift z approx. equal 0.5, which occurred roughly coincident with the day of GRB.