Bayesian Analysis of Nuclear Saturation

Voiceover

Presenter(s)

Saad Bezoui

Abstract or Description

Predicting properties of atomic nuclei across the nuclear chart and the neutron-rich matter inside neutron stars is of great interest in nuclear physics and astrophysics.

Chiral effective field theory (ChEFT) has become a powerful framework to make such predictions from first principles using nuclear interactions guided by the symmetries of low-energy quantum chromodynamics, the theory of strong interactions.

To improve chiral nuclear interactions, the empirical saturation point of the symmetric nuclear matter equation of state has been an important benchmark. Saturation means here that the symmetric nuclear matter equation of state has a minimum at the density n_0 \approx 0.16 fm^{-3}, closely related to the typical density in heavy nuclei (with some adjustments), and the associated energy per particle E(n_0)/A \approx -16 MeV. 

But chiral interactions have been benchmarked with relatively simple estimates of the empirical saturation point based on a range of energy density functionals.

These simple constraints have only limited statistical meaning at best. 

This raises the question of how well do we know the empirical saturation point? 

And how large are the systematic uncertainties in state-of-the-art constraints from density functional theory (DFT)?

In this work, we use Bayesian methods combined with a set of recent constraints from DFT to learn these systematic uncertainties from the data and derive statistically robust constraints for (n_0, E(n_0)/A). 

We find that our posterior for the saturation point overlaps significantly with a simple, box-like constraint commonly used in many-body theory, but our posterior tends to allow for somewhat lower saturation densities. 

Our framework can be readily extended to include future empirical constraints to rigorously benchmark saturation properties of ChEFT interactions in the FRIB and multi-messenger astronomy era.

Mentor

Christian Drischler