Physics-Informed Machine Learning for Modeling Turbulence in Supernovae
Authors:
Platon I. Karpov,
Chengkun Huang,
Iskandar Sitdikov,
Chris L. Fryer,
Stan Woosley,
Ghanshyam Pilania
Abstract:
Turbulence plays an important role in astrophysical phenomena, including core-collapse supernovae (CCSN), but current simulations must rely on subgrid models since direct numerical simulation (DNS) is too expensive. Unfortunately, existing subgrid models are not sufficiently accurate. Recently, Machine Learning (ML) has shown an impressive predictive capability for calculating turbulence closure.…
▽ More
Turbulence plays an important role in astrophysical phenomena, including core-collapse supernovae (CCSN), but current simulations must rely on subgrid models since direct numerical simulation (DNS) is too expensive. Unfortunately, existing subgrid models are not sufficiently accurate. Recently, Machine Learning (ML) has shown an impressive predictive capability for calculating turbulence closure. We have developed a physics-informed convolutional neural network (CNN) to preserve the realizability condition of Reynolds stress that is necessary for accurate turbulent pressure prediction. The applicability of the ML subgrid model is tested here for magnetohydrodynamic (MHD) turbulence in both the stationary and dynamic regimes. Our future goal is to utilize this ML methodology (available on GitHub) in the CCSN framework to investigate the effects of accurately-modeled turbulence on the explosion of these stars.
△ Less
Submitted 9 August, 2022; v1 submitted 17 May, 2022;
originally announced May 2022.
The Effects of Metallicity and Abundance Pattern of the ISM on Supernova Feedback
Authors:
Platon I. Karpov,
Davide Martizzi,
Phillip Macias,
Enrico Ramirez-Ruiz,
Anne N. Kolborg,
Jill P. Naiman
Abstract:
Supernova (SN) feedback plays a vital role in the evolution of galaxies. While modern cosmological simulations capture the leading structures within galaxies, they struggle to provide sufficient resolution to study small-scale stellar feedback, such as the detailed evolution of SN remnants. It is thus common practice to assume subgrid models that are rarely extended to low metallicities, and which…
▽ More
Supernova (SN) feedback plays a vital role in the evolution of galaxies. While modern cosmological simulations capture the leading structures within galaxies, they struggle to provide sufficient resolution to study small-scale stellar feedback, such as the detailed evolution of SN remnants. It is thus common practice to assume subgrid models that are rarely extended to low metallicities, and which routinely use the standard solar abundance pattern. With the aid of 1-d hydrodynamical simulations, we extend these models to consider low metallicities and non-solar abundance patterns as derived from spectra of Milky Way stars. For that purpose, a simple, yet effective framework has been developed to generate non-solar abundance pattern cooling functions. We find that previous treatments markedly over-predict SN feedback at low metallicities and show that non-negligible changes in the evolution of SN remnants of up to $\approx 50\%$ in $cooling\; mass$ and $\approx 27\%$ in $momentum\; injection\; from\; SN\; remnants$ arise from non-solar abundance patterns. We use our simulations to quantify these results as a function of metallicity and abundance pattern variations and present analytic formulae to accurately describe the trends. These formulae have been designed to serve as subgrid models for SN feedback in cosmological hydrodynamical simulations.
△ Less
Submitted 30 April, 2021; v1 submitted 23 April, 2020;
originally announced April 2020.
