Verifying the failing supernova constraint on dark photons with two-dimensional hydrodynamic simulations

TL;DR

This work tests the failing supernova constraint on dark photons with a DP mass of $m_{\gamma'}=0.3$ MeV by performing self-consistent two-dimensional neutrino-radiation hydrodynamic simulations that include DP production via nucleon bremsstrahlung. The DP cooling is implemented as a local energy sink, with the effective mixing parameter $\epsilon_m$ driving the production rate and potential resonances in the gain region. The authors find a critical mixing parameter $\epsilon \approx 3\times10^{-9}$ above which shock revival fails, in line with prior post-processing bounds, and show that DP cooling reduces the explosion energy and nickel synthesis while the PNS mass remains largely unchanged due to stochasticity. The results corroborate the failing SN constraint at this DP mass and motivate expanding the analysis to other DP masses and to three-dimensional simulations to assess systematics and robustness of the constraint.

Abstract

Recent studies on the dark photon (DP) production in collapsing stars argue that the cooling effect induced by DPs can hinder supernova explosions and lead to a ``failing supernova" constraint on the photon-DP mixing parameter $ε$. In order to verify the idea, we perform two-dimensional neutrino-radiation hydrodynamic simulations coupled with the DP production with the mass of 0.3 MeV. We find that the shock revival does not happen until the end of the simulations when $ε\gtrsim3\times10^{-9}$. The photon-DP mixing parameter above this value can be excluded by the failing supernova argument. Interestingly, our constraint is close to the one reported by the previous studies which adopted the post-processing framework. This result motivates one to investigate a wider parameter range of DPs with self-consistent simulations and evaluate uncertainties in the constraint.

Figures (6)

• Figure 1: Our five models are shown on the parameter space for DPs. The circles indicate the exploding models and the triangles indicate the imploding models. The upper limits on $\epsilon$2025PhRvL.134o1002C, which are obtained from the SN 1987A neutrino burst, failing supernovae (SNe), and diffuse DPs, are also shown as the colored regions.
• Figure 2: The angular-averaged shock radius as a function of the post-bounce time $t_\mathrm{pb}$. The solid curve corresponds to the reference model without DPs and the other curves correspond to the models with DPs.
• Figure 3: The diagnostic explosion energy $E_\mathrm{diag}$ for the exploding models as a function of the post-bounce time $t_\mathrm{pb}$. The solid curve corresponds to the reference model without DPs and the other curves correspond to the models with DPs.
• Figure 4: The radial profile of the neutrino heating rate $Q_\nu$ and the transverse (longitudinal) DP cooling rate $Q_\mathrm{T}$ ($Q_\mathrm{L})$ in the equatorial plane for the $\epsilon_{10}=30$ model. The upper panel is the snapshot at $t_\mathrm{pb}=0.15$ s and the lower panel is the snapshot at $t_\mathrm{pb}=0.30$ s.
• Figure 5: The radial profile for the plasma frequency $\omega_\mathrm{pl}$ in the equatorial plane for the $\epsilon_{10}=30$ model. The horizontal lines show the DP mass $m_{\gamma'}=0.3$ MeV and $\sqrt{2/3}m_{\gamma'}$, which indicate the region where DPs are resonantly produced.