Abstract:
We present a new general relativistic (GR) code for hydrodynamic supernova simulations with neutrino transport in spherical and azimuthal symmetry (1D/2D). The code is a combination of the CoCoNuT hydro module, which is a Riemann-solver based, high-resolution shock-capturing method, and the three-flavor, energy-dependent neutrino transport scheme VERTEX. VERTEX integrates the neutrino moment equations with a variable Eddington factor closure computed from a model Boltzmann equation and uses the ray-by-ray plus approximation in 2D, assuming the neutrino distribution to be axially symmetric around the radial direction, and thus the neutrino flux to be radial. Our spacetime treatment employs the ADM 3+1 formalism with the conformal flatness condition for the spatial three-metric. This approach is exact in 1D and has been shown to yield very accurate results also for rotational stellar collapse. We introduce new formulations of the energy equation to improve total energy conservation in relativistic and Newtonian hydro simulations with Eulerian finite-volume codes. Moreover, a modified version of the VERTEX scheme is developed that simultaneously conserves energy and lepton number with better accuracy and higher numerical stability. To verify our code, we conduct a series of tests, including a detailed comparison with published 1D results for stellar core collapse. Long-time simulations of proto-neutron star cooling over several seconds both demonstrate the robustness of the new CoCoNuT-VERTEX code and show the approximate treatment of GR effects by means of an effective gravitational potential as in PROMETHEUS-VERTEX to be remarkably accurate in 1D. (abridged)

Abstract:
We present 2D hydrodynamic simulations of the long-time accretion phase of a 15 solar mass star after core bounce and before the launch of a supernova explosion. Our simulations are performed with the Prometheus-Vertex code, employing multi-flavor, energy-dependent neutrino transport and an effective relativistic gravitational potential. Testing the influence of a stiff and a soft equation of state for hot neutron star matter, we find that the non-radial mass motions in the supernova core due to the standing accretion shock instability (SASI) and convection impose a time variability on the neutrino and gravitational-wave signals. These variations have larger amplitudes as well as higher frequencies in the case of a more compact nascent neutron star. After the prompt shock-breakout burst of electron neutrinos, a more compact accreting remnant radiates neutrinos with higher luminosities and larger mean energies. The observable neutrino emission in the direction of SASI shock oscillations exhibits a modulation of several 10% in the luminosities and ~1 MeV in the mean energies with most power at typical SASI frequencies of 20-100 Hz. At times later than 50-100 ms after bounce the gravitational-wave amplitude is dominated by the growing low-frequency (<200 Hz) signal associated with anisotropic neutrino emission. A high-frequency wave signal is caused by nonradial gas flows in the outer neutron star layers, which are stirred by anisotropic accretion from the SASI and convective regions. The gravitational-wave power then peaks at about 300-800 Hz with distinctively higher spectral frequencies originating from the more compact and more rapidly contracting neutron star. The detectability of the SASI effects in the neutrino and gravitational-wave signals is briefly discussed. (abridged)

Abstract:
The neutrino-driven explosion mechanism for core-collapse supernovae in its modern flavor relies on the additional support of hydrodynamical instabilities in achieving shock revival. Two possible candidates, convection and the so-called standing accretion shock instability (SASI), have been proposed for this role. In this paper, we discuss new successful simulations of supernova explosions that shed light on the relative importance of these two instabilities. While convection has so far been observed to grow first in self-consistent hydrodynamical models with multi-group neutrino transport, we here present the first such simulation in which the SASI grows faster while the development of convection is initially inhibited. We illustrate the features of this SASI-dominated regime using an explosion model of a 27 solar mass progenitor, which is contrasted with a convectively-dominated model of an 8.1 solar mass progenitor with subsolar metallicity, whose early post-bounce behavior is more in line with previous 11.2 and 15 solar mass explosion models. We analyze the conditions discriminating between the two different regimes, showing that a high mass-accretion rate and a short advection time-scale are conducive for strong SASI activity. We also briefly discuss some important factors for capturing the SASI-driven regime, such as general relativity, the progenitor structure, a nuclear equation of state leading to a compact proto-neutron star, and the neutrino treatment. Finally, we evaluate possible implications of our findings for 2D and 3D supernova simulations. Our results show that a better understanding of the SASI and convection in the non-linear regime is required.

Abstract:
We present the first two-dimensional general relativistic (GR) simulations of stellar core collapse and explosion with the CoCoNuT hydrodynamics code in combination with the VERTEX solver for energy-dependent, three-flavor neutrino transport, using the extended conformal flatness condition for approximating the spacetime metric and a ray-by-ray-plus ansatz to tackle the multi-dimensionality of the transport. For both of the investigated 11.2 and 15 solar mass progenitors we obtain successful, though seemingly marginal, neutrino-driven supernova explosions. This outcome and the time evolution of the models basically agree with results previously obtained with the PROMETHEUS hydro solver including an approximative treatment of relativistic effects by a modified Newtonian potential. However, GR models exhibit subtle differences in the neutrinospheric conditions compared to Newtonian and pseudo-Newtonian simulations. These differences lead to significantly higher luminosities and mean energies of the radiated electron neutrinos and antineutrinos and therefore to larger energy-deposition rates and heating efficiencies in the gain layer with favorable consequences for strong non-radial mass motions and ultimately for an explosion. Moreover, energy transfer to the stellar medium around the neutrinospheres through nucleon recoil in scattering reactions of heavy-lepton neutrinos also enhances the mentioned effects. Together with previous pseudo-Newtonian models the presented relativistic calculations suggest that the treatment of gravity and energy-exchanging neutrino interactions can make differences of even 50-100% in some quantities and is likely to contribute to a finally successful explosion mechanism on no minor level than hydrodynamical differences between different dimensions.

Abstract:
Using three-dimensional (3D) simulations of neutrino-powered supernova explosions we show that the hydrodynamical kick scenario proposed by Scheck et al. on the basis of two-dimensional (2D) models can yield large neutron star (NS) recoil velocities also in 3D. Although the shock stays relatively spherical, standing accretion-shock and convective instabilities lead to a globally asymmetric mass and energy distribution in the postshock layer. An anisotropic momentum distribution of the ejecta is built up only after the explosion sets in. Total momentum conservation implies the acceleration of the NS on a timescale of 1-3 seconds, mediated mainly by long-lasting, asymmetric accretion downdrafts and the anisotropic gravitational pull of large inhomogeneities in the ejecta. In a limited set of 15 solar-mass models with an explosion energy of about 10^51 erg this stochastic mechanism is found to produce kicks from <100 km/s to >500 km/s, and >1000 km/s seem possible. Strong rotational flows around the accreting NS do not develop in our collapsing, non-rotating progenitors. The NS spins therefore remain low with estimated periods of about 500-1000 ms and no alignment with the kicks.

Abstract:
We present 3D simulations of supernova (SN) explosions of nonrotating stars, triggered by the neutrino-heating mechanism with a suitable choice of the core-neutrino luminosity. Our results show that asymmetric mass ejection caused by hydrodynamic instabilities can accelerate the neutron star (NS) up to recoil velocities of more than 700 km/s by the "gravitational tug-boat mechanism", which is enough to explain most observed pulsar velocities. The associated NS spin periods are about 100 ms to 8 s without any correlation between spin and kick magnitudes or directions. This suggests that faster spins and a possible spin-kick alignment might require angular momentum in the progenitor core prior to collapse. Our simulations for the first time demonstrate a clear correlation between the size of the NS kick and anisotropic ejection of heavy elements created by explosive burning behind the shock. In the case of large NS kicks the explosion is significantly stronger opposite to the kick vector. Therefore the bulk of the Fe-group elements, in particular nickel, is ejected mostly in large clumps against the kick direction. This contrasts with the case of low recoil velocity, where the Ni-rich lumps are more isotropically distributed. Intermediate-mass nuclei heavier than Si (like Ca and Ti) also exhibit a significant enhancement in the hemisphere opposite to the direction of fast NS motion, while the distribution of C, O, and Ne is not affected, and that of Mg only marginally. Mapping the spatial distribution of the heavy elements in SN remnants with identified pulsar motion may offer an important diagnostic test of the kick mechanism. Different from kick scenarios based on anisotropic neutrino emission, our hydrodynamical acceleration model predicts enhanced ejection of Fe-group elements and of their nuclear precursors in the direction opposite to the NS recoil. (abridged)

Abstract:
We present the first three-dimensional (3D) simulations of the large-scale mixing that takes place in the shock-heated stellar layers ejected in the explosion of a 15.5 solar-mass blue supergiant star. The outgoing supernova shock is followed from its launch by neutrino heating until it breaks out from the stellar surface more than two hours after the core collapse. Violent convective overturn in the post-shock layer causes the explosion to start with significant asphericity, which triggers the growth of Rayleigh-Taylor (RT) instabilities at the composition interfaces of the exploding star. Deep inward mixing of hydrogen (H) is found as well as fast-moving, metal-rich clumps penetrating with high velocities far into the H-envelope of the star as observed, e.g., in the case of SN 1987A. Also individual clumps containing a sizeable fraction of the ejected iron-group elements (up to several 0.001 solar masses) are obtained in some models. The metal core of the progenitor is partially turned over with Ni-dominated fingers overtaking oxygen-rich bullets and both Ni and O moving well ahead of the material from the carbon layer. Comparing with corresponding 2D (axially symmetric) calculations, we determine the growth of the RT fingers to be faster, the deceleration of the dense metal-carrying clumps in the He and H layers to be reduced, the asymptotic clump velocities in the H-shell to be higher (up to ~4500 km/s for the considered progenitor and an explosion energy of 10^{51} ergs, instead of <2000 km/s in 2D), and the outward radial mixing of heavy elements and inward mixing of hydrogen to be more efficient in 3D than in 2D. We present a simple argument that explains these results as a consequence of the different action of drag forces on moving objects in the two geometries. (abridged)

Abstract:
Two-dimensional simulations of a Type II and a Type Ib-like supernova explosion are presented that encompass shock revival by neutrino heating, neutrino-driven convection, explosive nucleosynthesis, the growth of Rayleigh-Taylor instabilities, and the propagation of newly formed metal clumps through the exploding star. In both cases we find very high Ni56 velocities of 17000 km/s shortly after shock-revival, and a complete fragmentation of the progenitor's metal core within the first few minutes after core bounce, due to the growth of Rayleigh-Taylor instabilities at the Si/O and (C+O)/He composition interfaces. This leads to the formation of high-velocity, metal-rich clumps which eventually decouple from the flow and move ballistically through the ejecta. Maximum final metal velocities of 3500-5500 km/s and 1200 km/s are obtained for the Type Ib model and the Type II model, respectively. The low maximum metal velocities in the Type II model, which are significantly smaller than those observed in SN 1987A, are due to the massive hydrogen envelope of the progenitor. The envelope forces the supernova shock to decelerate strongly, leaving behind a reverse shock below the He/H interface, which interacts with the clumps and slows them down significantly. This reverse shock is absent in the Type Ib-like model. The latter is in fairly good agreement with observations of Type Ib supernovae. In addition, in this case the pattern of clump formation in the ejecta is correlated with the convective pattern prevailing during the shock-revival phase. This might be used to deduce observational constraints for the dynamics during this early phase of the evolution, and the role of neutrino heating in initiating the explosion.

Abstract:
New two-dimensional, high-resolution calculations of a core collapse supernova in a 15 Msol blue supergiant are presented, which cover the entire evolution from shock revival until the first few hours of the explosion. Explosive nucleosynthesis, its dependence upon convective mixing during the first second of the evolution and the growth of Rayleigh-Taylor instabilities at the composition interfaces of the progenitor star are all modeled consistently and allow for a comparison with observational data. We confirm our earlier findings, that the perturbations induced by neutrino driven convection are sufficiently strong to seed large-scale Rayleigh-Taylor mixing and to destroy the onion-shell structure of the stellar He-core. As in our earlier calculations, the strong deceleration of the nickel clumps in the layers adjacent to the He/H interface suggests that the high velocities of iron-group elements observed in SN 1987A cannot be explained on the basis of currently favored progenitor models. Possible solutions to this dilemma and the implications of the mixing for type Ib explosions are briefly discussed.

Abstract:
High-resolution two-dimensional simulations were performed for the first five minutes of the evolution of a core collapse supernova explosion in a 15 solar mass blue supergiant progenitor. The computations start shortly after bounce and include neutrino-matter interactions by using a light-bulb approximation for the neutrinos, and a treatment of the nucleosynthesis due to explosive silicon and oxygen burning. We find that newly formed iron-group elements are distributed throughout the inner half of the helium core by Rayleigh-Taylor instabilities at the Ni+Si/O and C+O/He interfaces, seeded by convective overturn during the early stages of the explosion. Fast moving nickel mushrooms with velocities up to about 4000 km/s are observed. This offers a natural explanation for the mixing required in light curve and spectral synthesis studies of Type Ib explosions. A continuation of the calculations to later times, however, indicates that the iron velocities observed in SN 1987 A cannot be reproduced because of a strong deceleration of the clumps in the dense shell left behind by the shock at the He/H interface.