Nuclear Neural Networks: Emulating Late Burning Stages in Core-collapse Supernova Progenitors

One of the main challenges in modeling massive stars to the onset of core collapse is the computational bottleneck of nucleosynthesis during advanced burning stages. The number of isotopes formed requires solving a large set of fully coupled stiff ordinary differential equations, making the simulati...

Full description

Saved in:
Bibliographic Details
Main Authors: Aldana Grichener, Mathieu Renzo, Wolfgang E. Kerzendorf, Rob Farmer, Selma E. de Mink, Earl Patrick Bellinger, Chi-kwan Chan, Nutan Chen, Ebraheem Farag, Stephen Justham
Format: Article
Language:English
Published: IOP Publishing 2025-01-01
Series:The Astrophysical Journal Supplement Series
Subjects:
Online Access:https://doi.org/10.3847/1538-4365/ade717
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849247130630225920
author Aldana Grichener
Mathieu Renzo
Wolfgang E. Kerzendorf
Rob Farmer
Selma E. de Mink
Earl Patrick Bellinger
Chi-kwan Chan
Nutan Chen
Ebraheem Farag
Stephen Justham
author_facet Aldana Grichener
Mathieu Renzo
Wolfgang E. Kerzendorf
Rob Farmer
Selma E. de Mink
Earl Patrick Bellinger
Chi-kwan Chan
Nutan Chen
Ebraheem Farag
Stephen Justham
author_sort Aldana Grichener
collection DOAJ
description One of the main challenges in modeling massive stars to the onset of core collapse is the computational bottleneck of nucleosynthesis during advanced burning stages. The number of isotopes formed requires solving a large set of fully coupled stiff ordinary differential equations, making the simulations computationally intensive and prone to numerical instability. To overcome this barrier, we design a nuclear neural network (NNN) framework with multiple hidden layers to emulate nucleosynthesis calculations and conduct a proof of concept to evaluate its performance. The NNN takes the temperature, density, and composition of a burning region as input and predicts the resulting isotopic abundances along with the energy generation and loss rates. We generate training sets for initial conditions corresponding to oxygen core depletion and beyond using large nuclear reaction networks, and compare the predictions of the NNNs to results from a commonly used small net. We find that the NNNs improve the accuracy of the electron fraction by 280%–660%, the average atomic and mass numbers by 150%–360%, and the nuclear energy generation by 250%–750%, consistently outperforming the small network across all time steps. They also achieve significantly better predictions of neutrino losses on relatively short timescales, with improvements ranging from 100% to 1,000,000%. While further work is needed to enhance their accuracy and applicability to different stellar conditions, integrating NNN-trained models into stellar evolution codes is promising for facilitating the large-scale generation of core-collapse supernova progenitors with higher physical fidelity.
format Article
id doaj-art-f51f638db2e74c849962a14d143ee6c9
institution Kabale University
issn 0067-0049
language English
publishDate 2025-01-01
publisher IOP Publishing
record_format Article
series The Astrophysical Journal Supplement Series
spelling doaj-art-f51f638db2e74c849962a14d143ee6c92025-08-20T03:58:18ZengIOP PublishingThe Astrophysical Journal Supplement Series0067-00492025-01-0127924910.3847/1538-4365/ade717Nuclear Neural Networks: Emulating Late Burning Stages in Core-collapse Supernova ProgenitorsAldana Grichener0https://orcid.org/0000-0002-2215-1841Mathieu Renzo1https://orcid.org/0000-0002-6718-9472Wolfgang E. Kerzendorf2https://orcid.org/0000-0002-0479-7235Rob Farmer3https://orcid.org/0000-0003-3441-7624Selma E. de Mink4https://orcid.org/0000-0001-9336-2825Earl Patrick Bellinger5https://orcid.org/0000-0003-4456-4863Chi-kwan Chan6https://orcid.org/0000-0001-6337-6126Nutan Chen7https://orcid.org/0000-0001-7571-0742Ebraheem Farag8https://orcid.org/0000-0002-5794-4286Stephen Justham9https://orcid.org/0000-0001-7969-1569Steward Observatory and Department of Astronomy, University of Arizona , 933 North Cherry Avenue, Tucson, AZ 85721, USA ; agrichener@arizona.edu; Max Planck Institute for Astrophysics , Karl-Schwarzschild-Str. 1, 85748 Garching, Germany; Department of Physics, Technion , Haifa, 3200003, IsraelSteward Observatory and Department of Astronomy, University of Arizona , 933 North Cherry Avenue, Tucson, AZ 85721, USA ; agrichener@arizona.eduDepartment of Computational Mathematics, Science, and Engineering, Michigan State University , East Lansing, MI 48824, USA; Department of Physics and Astronomy, Michigan State University , East Lansing, MI 48824, USAMax Planck Institute for Astrophysics , Karl-Schwarzschild-Str. 1, 85748 Garching, GermanyMax Planck Institute for Astrophysics , Karl-Schwarzschild-Str. 1, 85748 Garching, Germany; Ludwig-Maximilians-Universitat Munchen , Geschwister-Scholl-Platz 1, 80539 Munchen, GermanyDepartment of Astronomy, Yale University , New Haven, CT 06511, USASteward Observatory and Department of Astronomy, University of Arizona , 933 North Cherry Avenue, Tucson, AZ 85721, USA ; agrichener@arizona.edu; Data Science Institute, University of Arizona , 1230 N. Cherry Avenue, Tucson, AZ 85721, USA; Program in Applied Mathematics, University of Arizona , 617 North Santa Rita Avenue, Tucson, AZ 85721, USAMachine Learning Research Lab, Volkswagen AG , 38440 Munich, GermanyDepartment of Astronomy, Yale University , New Haven, CT 06511, USAMax Planck Institute for Astrophysics , Karl-Schwarzschild-Str. 1, 85748 Garching, GermanyOne of the main challenges in modeling massive stars to the onset of core collapse is the computational bottleneck of nucleosynthesis during advanced burning stages. The number of isotopes formed requires solving a large set of fully coupled stiff ordinary differential equations, making the simulations computationally intensive and prone to numerical instability. To overcome this barrier, we design a nuclear neural network (NNN) framework with multiple hidden layers to emulate nucleosynthesis calculations and conduct a proof of concept to evaluate its performance. The NNN takes the temperature, density, and composition of a burning region as input and predicts the resulting isotopic abundances along with the energy generation and loss rates. We generate training sets for initial conditions corresponding to oxygen core depletion and beyond using large nuclear reaction networks, and compare the predictions of the NNNs to results from a commonly used small net. We find that the NNNs improve the accuracy of the electron fraction by 280%–660%, the average atomic and mass numbers by 150%–360%, and the nuclear energy generation by 250%–750%, consistently outperforming the small network across all time steps. They also achieve significantly better predictions of neutrino losses on relatively short timescales, with improvements ranging from 100% to 1,000,000%. While further work is needed to enhance their accuracy and applicability to different stellar conditions, integrating NNN-trained models into stellar evolution codes is promising for facilitating the large-scale generation of core-collapse supernova progenitors with higher physical fidelity.https://doi.org/10.3847/1538-4365/ade717Core-collapse supernovaeMassive starsNucleosynthesis
spellingShingle Aldana Grichener
Mathieu Renzo
Wolfgang E. Kerzendorf
Rob Farmer
Selma E. de Mink
Earl Patrick Bellinger
Chi-kwan Chan
Nutan Chen
Ebraheem Farag
Stephen Justham
Nuclear Neural Networks: Emulating Late Burning Stages in Core-collapse Supernova Progenitors
The Astrophysical Journal Supplement Series
Core-collapse supernovae
Massive stars
Nucleosynthesis
title Nuclear Neural Networks: Emulating Late Burning Stages in Core-collapse Supernova Progenitors
title_full Nuclear Neural Networks: Emulating Late Burning Stages in Core-collapse Supernova Progenitors
title_fullStr Nuclear Neural Networks: Emulating Late Burning Stages in Core-collapse Supernova Progenitors
title_full_unstemmed Nuclear Neural Networks: Emulating Late Burning Stages in Core-collapse Supernova Progenitors
title_short Nuclear Neural Networks: Emulating Late Burning Stages in Core-collapse Supernova Progenitors
title_sort nuclear neural networks emulating late burning stages in core collapse supernova progenitors
topic Core-collapse supernovae
Massive stars
Nucleosynthesis
url https://doi.org/10.3847/1538-4365/ade717
work_keys_str_mv AT aldanagrichener nuclearneuralnetworksemulatinglateburningstagesincorecollapsesupernovaprogenitors
AT mathieurenzo nuclearneuralnetworksemulatinglateburningstagesincorecollapsesupernovaprogenitors
AT wolfgangekerzendorf nuclearneuralnetworksemulatinglateburningstagesincorecollapsesupernovaprogenitors
AT robfarmer nuclearneuralnetworksemulatinglateburningstagesincorecollapsesupernovaprogenitors
AT selmaedemink nuclearneuralnetworksemulatinglateburningstagesincorecollapsesupernovaprogenitors
AT earlpatrickbellinger nuclearneuralnetworksemulatinglateburningstagesincorecollapsesupernovaprogenitors
AT chikwanchan nuclearneuralnetworksemulatinglateburningstagesincorecollapsesupernovaprogenitors
AT nutanchen nuclearneuralnetworksemulatinglateburningstagesincorecollapsesupernovaprogenitors
AT ebraheemfarag nuclearneuralnetworksemulatinglateburningstagesincorecollapsesupernovaprogenitors
AT stephenjustham nuclearneuralnetworksemulatinglateburningstagesincorecollapsesupernovaprogenitors