Exascale Models of Astrophysical Thermonuclear Explosions

PI Michael Zingale, Stony Brook University
Co-PI Ann Almgren, Lawrence Berkeley National Laboratory
Alan Calder, Stony Brook University
Kiran Eiden, UC Berkeley
Eric Johnson, Stony Brook University
Max Katz, Stony Brook University
Andy Nonaka, Lawrence Berkeley National Laboratory
Alexander Smith Clark, Stony Brook University
Abigail Polin, Caltech
JeanSexton, Lawrence Berkeley National Laboratory
Donald Willcox, Lawrence Berkeley National Laboratory
Zingale Incite 2022

A view of the flame propagating through the thin helium layer on a rapidly rotating neutron star -- a model for an X-ray burst.  This image shows the composition; the flame started out as a small hotspot that is spreading radially across the surface. Image: AMReX Astro Team and Michael Zingale, Stony Brook University.

 

Project Summary

This INCITE project uses the team's Castro code to carry out high-performance, robust, and accurate simulations to advance our understanding ofXRBs and SN Ia, pushing to model a larger fraction of the neutron star surface.

Project Description

This project will build upon the success of earlier INCITE awards that explored astrophysical thermonuclear explosions, in particular, Type Ia supernovae (SN Ia) and x-ray bursts(XRBs). The team will use their Castro code to carry out high-performance, robust, and accurate simulations to advance our understanding ofXRBs and SN Ia, as well as related physics (thermonuclear combustion and detonations).

In the area of XRBs, the researchers will greatly expand their work to model thermonuclear flame propagation across the surface of a neutron star. They will explore larger reaction networks and the effect of magnetic fields, and push to model a larger fraction of the neutron star surface. For their SN Ia studies, the team will focus on the double-detonation model. Both XRBs and SN Ia are multiscale, multiphysics problems that rely on the interplay between reactions and hydrodynamics. The team’s open-source Castro code has a new time-integration that is designed to strongly couple these processes, enabling the team to carry out accurate and efficient simulations of reacting flows.

Domains
Allocations