How did the researchers test it?

Full supernova simulations are among the most computationally expensive problems in astrophysics. "Simulating the death of a massive star is something that is pretty much at the frontier of what we can do computationally at the moment. That is because it is a problem involving a great deal of physics and it is extremely expensive computationally," said Irene Tamborra, professor at the Niels Bohr Institute, head of the Particle Astrophysics group, and the study's second author.

To make the problem tractable, Gogilashvili and Tamborra built a simplified model and ran it on a large sample of stars. Their approach:

• Simulated the collapse of 195 stars with masses between 9 and 120 times the mass of the sun
• Compared models with and without neutrino flavor conversion
• Tested what happened when flavor conversion kicked in at different densities within the star
• Assessed whether each star exploded as a supernova or collapsed into a black hole

The result: neutrino behavior "can significantly alter the outcome," with the strongest effects in stars between 16 and 30 solar masses.

"It was a really exciting moment when we put all 195 simulations side by side and saw a whole range of stars flip from exploding to failing. Seeing such a clear pattern across so many stars told us that neutrino flavor conversion is something we simply cannot leave out when we try to understand how massive stars end their lives," Gogilashvili said.

Could this solve the 'supernova rate problem'?

The findings may help address a long-standing mismatch between theory and observation. Astronomers call it the "supernova rate problem": they detect significantly fewer supernovae in the universe than theoretical models predict.

One possible explanation is "failed supernovae"—stars that collapse directly into black holes without a bright, visible explosion. "Normally, we detect a supernova because the explosion shines very brightly. But if a star collapses directly into a black hole without a visible explosion, or is obscured by dust, it can effectively 'disappear' from our counts. Our results therefore suggest that there is a mechanism that could make such 'failed supernovae' more likely," Gogilashvili explained.

Tamborra added: "This could therefore not only give us better tools to predict a dying star's fate, but it may also help explain why observations do not always match theoretical predictions."

What are the limitations?

The findings come from a simplified model, not full three-dimensional supernova simulations, which remain beyond current computational reach. The pattern is clear across the 195 simulations, but the results are preliminary in the sense that more detailed modeling—and eventually observations—will be needed to confirm how strongly flavor conversion shapes real stellar deaths.

Why it matters beyond black holes

Massive stars forge heavy elements during their lifetimes and scatter them into space when they explode. Those elements later form planets and living things. "When we study how massive stars live and die, we are also investigating the origins of many of the elements that make up the universe and ourselves. In this way, questions about dying stars are linked to questions about our own origins," Tamborra said.

The study, "Neutrino flavor conversion shapes the rate of failed core-collapse supernovae," by Mariam Gogilashvili et al., appears in Physical Review D (DOI: 10.1103/pz3y-3lv5) and is also available on arXiv.