Exascale Simulations of Neutron Star Mergers
David Radice, The Pennsylvania State University (Principal Investigator)

Gravitational waves and electromagnetic radiation from colliding neutron stars (NSs) encode precious information about the internal structure and composition of NSs and reveal the explosions in which some of the heaviest elements, such as gold, platinum, and uranium are formed.  Relativistic heavy-ion collisions and experiments such as PREX here on Earth are probing the nature of matter under extreme conditions.  The nuclear physics involved in the creation of the heavy elements will be more tightly constrained as FRIB comes online in the next years. However, as observations and laboratory measurements improve, so must the theoretical understanding of NS mergers in order to maximize the science return from these large scale investments.  Ab-initio supercomputer simulations are the only tool able to bridge astronomical observations and laboratory experiments and connect them to the merger dynamics.  Current simulation results are affected by large systematic errors stemming from the inability to resolve all spatial and temporal scales in mergers and by their approximate treatments of neutrinos.  This project aims to overcome these limitations by developing a new simulation infrastructure able to leverage next-generation supercomputer hardware, enabling calculations at unprecedented resolutions and extending over long timescales.  One of the key deliverables is a new neutrino transport solver including general-relativistic and quantum kinetic effects using the filtered spherical harmonics and Galerkin methods, some of the most sophisticated approaches developed in the applied mathematics and computational physics communities.  This project develops the theory foundations needed to address some of the most pressing questions in nuclear astrophysics such as the nature of matter inside NSs and the astrophysical site of production of the heavy elements.