Neutrino driven explosions aided by axion cooling in multidimensional simulations of core-collapse supernovae

In this study, we present the first multidimensional core-collapse supernovae (CCSNe) simulations including QCD axions in order to assess the impact on the CCSN explosion mechanism. We include axions in our simulations through the nucleon-nucleon bremsstrahlung emission channel and as a pure energy-sink term under the assumption that the axions free-stream after being emitted. We perform both spherically symmetric (1D) and axisymmetric (2D) simulations. In 1D, we utilize a parametrized heating scheme to achieve explosions, whereas in 2D we self-consistently realize explosions through the neutrino heating mechanism. Our 2D results for a $20M_{\odot }$ progenitor show an impact of the axion emission on the shock behavior and the explosion time when considering values of the PecceiQuinn energy scale $f_a\le 2\times {10}^8\mathrm{GeV}$. The strong cooling due to the axion emission accelerates the contraction of the core and leads to more efficient neutrino heating and earlier explosions. For the axion emission formalism utilized, the values of $f_a$ that impact the explosion are close to, but in tension with current limits based on the neutrinos detected from SN 1987A. However, given the nonlinear behavior of the emission and the multidimensional nature of CCSNe,we suggest that a self-consistent, multidimensional approach to simulating CCSNe, including any late time accretion and cooling, is needed to fully explore the axion bounds from supernovae and the impact on the CCSN explosion mechanism.

Figure 1

Shock radius versus the time for a 1D simulation with an $18M_{\odot }$ progenitor in comparison to the results of Fischer et al. (2016) [16] (see their Fig. 3). Following Fischer et al., we use artificial neutrino heating to trigger an explosion. All simulations have been run with $g_{ap}=9\times {10}^{-10}$ and $g_{an}=0$ or a Peccei-Quinn scale $f_a\sim 5\times {10}^8\mathrm{GeV}$. The different colors correspond to different axion-emission formalisms, although they have little impact on the early shock evolution shown here. The cyan and blue lines result from the formalism of Brinkmann et al. [18] (B&T) modified following Eq. (2) and used to match the results of Fischer et al. [16], with $\xi$, the degeneracy scale, equal to 0 and 1, respectively. The yellow, red, and orange lines follow the formalism of Carenza et al. [23], modified following Eq. (2), with $\xi$ equal to 0, 1.3078, and varying depending on the degeneracy (see text), respectively. The black curve shows the shock radius for the reference simulation without axion emission.

Figure 2

Neutrino and axion luminosities versus time. The different colors correspond to different axion emission formalisms while the black curve, labeled reference, shows a simulation without axion emission. The cyan and blue lines are following the formalism of Brinkmann et al. [18] (B&T) with $\xi$, the degeneracy scale, equal to 0 and 1, respectively, modified following Eq. (2). The yellow, red and orange lines follow the formalism of Carenza et al. [23], modified following Eq. (2), with $\xi$ equal to 0, 1.3078, and varying depending on the degeneracy, respectively.

Figure 3

Shock evolution with time in 1D for a $20M_{\odot }$ progenitor using artificial heating to trigger an explosion. The different colors correspond to different values of $f_a$ with the dark blue line being the reference simulation without axion emission.

Figure 4

Total neutrino (solid lines; measured at 500 km) and axion (dashed lines; integrated emission throughout the PNS) luminosity evolution with time. Additionally, we show the level of neutrino heating occurring during the accretion phase as dotted lines. The heating is defined to be the integral of the rate of neutrino energy deposition in the gain region. The latter is defined where the energy deposition rate is positive. The color code denotes the value of $f_a$, with blue being the reference simulation.

Figure 5

Local neutrino (solid lines; all species summed together) and axion (dashed lines) luminosities at 0.182 s (left panel) and 0.982 s (right panel) postbounce for various coupling constants (denoted by colors as shown in the legend) for the 1D simulations of the $20M_{\odot }$ progenitor.

Figure 6

Radial profiles of density (solid lines; left axis) and temperature (dashed line; right axis) for three different postbounce times of 0.182 s (left panel), 0.282 s (middle panel) and 0.982 s (right panel) for the 1D simulations of the $20M_{\odot }$ progenitor. For each time we show the profiles for each of the six values of $f_a$ and the reference run with no axions. Due to numerical issues the $f_a=5\times {10}^7\mathrm{GeV}$ simulations only progressed until 0.6 s and therefore is not shown in the last panel.

Figure 7

Shock radius evolution for simulations assuming different values of $f_a$ in 2D axisymmetric simulations using a $20M_{\odot }$ progenitor.

Figure 8

Ratio of the advection timescale to the heating timescale as a function of postbounce time. This ratio is a measure of the closeness of a simulation to achieving an explosion. When the heating timescale is shorter than the advection timescale (${\tau }_{\mathrm{adv}}/{\tau }_{\mathrm{heat}}>1$) runaway shock expansion is expected.

Figure 9

Neutrino luminosities and mean energies for different values of $f_a$.

Figure 10

Total neutrino (solid lines) and axion (dashed lines) luminosities versus the time for the five simulations with axion emission as well as the reference simulation. To add clarity to the figure we divide the axion luminosity by a factor of 4.

Figure 11

Left: Grey optical depth as a function of radius for different values of $f_a$. Right: Density as function of radius for different values of $f_a$. All the optical depths are shown at their maximum, with the time indicated in the left panel. For all the runs with a value of $f_a\ge 3\times {10}^8\mathrm{GeV}$, the time of maximum is the end of the simulation as the axion emission is still not receding. For all the other values, the maximal optical depth is reached after the peak of axion emission, when the density is increasing in the core.

Figure 12

Optical depth as a function of time and radius for different values of $f_a$. The color scheme represents the time evolution while each panel represents a simulation with a different value of $f_a$.