Flavor Conversion Enhances or Suppresses Supernova Explodability Independent of the Progenitor Mass

Flavor conversion can affect the neutrino-driven delayed explosion mechanism of collapsing massive stars, altering the efficiency of shock revival. We perform core-collapse supernova simulations in spherical symmetry for a set of progenitors with masses of 9.75 M⊙, 11 M⊙, 16.5 M⊙, 28 M⊙, 40 M⊙, and 60 M⊙, accounting for a mixing-length treatment for convection. Flavor conversion is modeled assuming instantaneous flavor equipartition below a critical baryon density, while conserving the lepton number. Regardless of the progenitor compactness, its mass, or the nuclear equation of state, we find that flavor conversion can increase heating (cooling) and enhance (hinder) the supernova explosion, if triggered near the gain (neutrino decoupling) region. Our findings suggest that the interplay among the region of the supernova core where flavor conversion occurs, the progenitor properties, and the nuclear equation of state is crucial in determining the fate of explosion and the properties of the compact remnant.

We find that flavor conversion can change the core-collapse fate regardless of the progenitor compactness, progenitor mass, or EOS, depending on the spatial region where it is triggered. The nuclear EOS further affects the threshold value of ρc at which the CCSN models transition from exploding to non-exploding or vice versa.

For the 11 M⊙ model (ξ2.5 = 0.01), which successfully explodes in the absence of flavor conversion, the latter consistently enhances neutrino heating independent of ρc, accelerating shock expansion and lowering the remnant mass due to reduced accretion. For the 9.75 M⊙ model, flavor conversion triggered at low ρc (ρc ≲ 1010 g/cm3) enhances the explosion, while sufficiently high ρc (ρc ≳ 1011 g/cm3) quenches it. In contrast to the early findings of Refs. [57, 60, 61], suggesting that flavor transformation aids the CCSN explodability of low-mass progenitors, we find that the impact of flavor conversion on the explodability of low-mass progenitors is nuanced and depends on the region where flavor conversion is triggered.

For intermediate-mass progenitors, we find a similar trend, regardless of their compactness. The 16.5 M⊙ model (ξ2.5 = 0.159), which explodes without flavor conversion, fails to explode when flavor conversion is triggered at high densities (ρc ≳ 1010 g/cm3), while low ρc further enhances the explosion. The 28 M⊙ model (ξ2.5 = 0.222) fails to explode regardless of ρc.

As for high-mass progenitors, the high-compactness 40 M⊙ model (ξ2.5 = 0.54) explodes without flavor conversion and with flavor conversion triggered below 109 g/cm3, but fails when flavor conversion is triggered at higher ρc. The high-mass, low-compactness 60 M⊙ model (ξ2.5 = 0.174) explodes without flavor conversion and with ρc ≲ 1010 g/cm3, but fails at higher ρc. In contrast to the early findings of Refs. [57, 60], which concluded that flavor conversion is insufficient to overcome the ram pressure of infalling material in CCSNe with high accretion rate, we find that the explodability of high-mass progenitors can be enhanced or hindered by flavor conversion.

Our findings show that flavor conversion can change the CCSN fate depending on where it is triggered, regardless of the progenitor mass, compactness, or nuclear EOS. This overall behavior reflects a delicate interplay between neutrino heating and the ram pressure of the accreting material. Correspondingly, the rate of failed explosions is affected by flavor conversion as explored in Ref. [89]. The compact remnant mass is also strongly affected [89]: enhanced explosions lead to lower remnant masses, whereas failed explosions result in slightly higher remnant masses due to prolonged accretion.

The properties of the Si/O interface are crucial to further understand the trend reported above as a function of core compactness. In fact, the accretion of infalling matter onto this interface leads to a sudden drop in the mass accretion rate and ram pressure at the shock, which may facilitate shock revival [83, 90, 91]. For instance, in our 11 M⊙ model, the accretion of the Si/O interface onto the shock at ∼ 0.1 s (cf. Fig. 7) produces a sharp drop in the ram pressure. Such drop in the mass accretion rate, combined with enhanced heating from flavor conversion, results in a rapid explosion. In contrast, for high-mass progenitors, such as 60 M⊙ models, the Si/O interface typically reaches the shock at ∼ 0.15–0.2 s (Fig. 8). If flavor transformation enhances neutrino heating (e.g., ρc = 109 g/cm3) when the Si/O interface is accreted, the explosion can be triggered or accelerated. Conversely, when flavor conversion suppresses heating (e.g., ρc = 1011 g/cm3), the beneficial impact of the Si/O interface on the explosion is hindered. Hence, flavor transformation can either amplify or counteract the effect of the Si/O interface, further contributing to the non-trivial dependence of the CCSN explodability on the progenitor structure and the region of the CCSN core affected by flavor conversion.

We also note that, in our exploding models, the shock does not exhibit a stalled phase, instead undergoes relatively prompt expansion once favorable heating conditions are achieved. Although our 1D+ framework does not fully capture multidimensional effects, this behavior suggests that the development of large-scale hydrodynamic instabilities, such as the standing accretion shock instability [16–19], might be disfavored.

Finally, we emphasize that flavor conversion in this work is implemented through a parametric prescription. Although this approach allows us to isolate and understand the macroscopic impact of neutrinos on the CCSN physics, the coupling of neutrino kinetics to the CCSN hydrodynamics is crucial to assess the implications of neutrino physics on the CCSN explosion dynamics and the multi-messenger observables. Moreover, the consequences of flavor conversion on the CCSN explodability should be considered together with other ingredients potentially important for the explosion mechanism, such as magnetic field amplification [92, 93], muon production in the newly-born neutron star [94], large-scale perturbations from convective oxygen and silicon shell burning [95], corrections to neutral current neutrino-nucleon scattering at low densities [96], or the effective mass of nucleons above nuclear saturation density [86]. Future work in this direction will be essential to assess the role of flavor transformation in shaping the properties of the remnant compact object.

Based on this paper, here are specific improvements I can make to AI systems and what the improved AI system can do:

Improvement: Train the AI to predict core-collapse supernova (CCSN) explosion outcomes by incorporating the paper's key finding that flavor conversion can either enhance or suppress explosions depending on where it is triggered (near the gain region vs. neutrino decoupling region).

What the improved AI can do:

• Given a progenitor mass, compactness parameter (ξ2.5), and equation of state (EOS), predict whether flavor conversion will aid or hinder shock revival

• Classify outcomes into four categories: enhanced explosion, suppressed explosion, failed explosion, or unchanged outcome

• Estimate the critical density (ρc) threshold that separates exploding from non-exploding models for a given progenitor

Improvement: Build an AI system that maps the full parameter space (progenitor mass, compactness, EOS, ρc) to explosion outcomes, using the paper's Table I as training data.

Improvement: Develop an AI that monitors neutrino luminosity, average energy, and heating rates (like those shown in Figures 2-3) to detect whether flavor conversion is enhancing or suppressing heating in the gain region.

Improvement: Create an AI that accounts for the paper's finding that SFHo and LS220 EOSs produce opposite explosion outcomes for the same progenitor and ρc.

Improvement: Use the paper's data on compactness (ξ2.5) to build a probabilistic model of explosion success.

Improvement: Train an AI on the shock radius evolution curves (Figures 5-8) to predict whether a shock will stall, expand, or recede.

Improvement: Build an AI that maps the radial density profile (like Figure 1) to identify which regions correspond to heating-enhancing vs. cooling-enhancing flavor conversion.

Improvement: Create an AI assistant that recommends simulation parameters (ρc, EOS, progenitor mass) to achieve desired explosion outcomes.

These improvements would enable AI systems to serve as powerful tools for supernova researchers, helping them design simulations, interpret results, and predict outcomes without running expensive full hydrodynamic calculations.

Flavor conversion can affect the neutrino-driven delayed explosion mechanism of collapsing massive stars, altering the efficiency of shock revival. We perform core-collapse supernova simulations in spherical symmetry for a set of progenitors with masses of 9.75, M, 11, M, 16.5, M, 28, M, 40, M, and 60, M, accounting for a mixing-length treatment for convection. Flavor conversion is modeled assuming instantaneous flavor equipartition below a critical baryon density, while conserving the lepton number. Regardless of the progenitor compactness, its mass, or the nuclear equation of state, we find that flavor conversion can increase heating (cooling) and enhance (hinder) the supernova explosion, if triggered near the gain (neutrino decoupling) region. Our findings suggest that the interplay among the region of the supernova core where flavor conversion occurs, the progenitor properties, and the nuclear equation of state is crucial in determining the fate of explosion and the properties of the compact remnant.