Research

Last updated: February 2026

The Computational Stellar Astrophysics group at the University of Victoria works on problems related to the origin of the elements and a range of topics in stellar physics. Our research focuses on the development of numerical tools and simulations to study the evolution of stars, their interior hydrodynamic mixing and oscillation processes, and their impact on the Universe.

3D stellar hydrodynamics with PPMstar

We develop and use the PPMstar explicit gas dynamics code to perform large-scale 3D hydrodynamic simulations of stellar interiors on leadership-class supercomputers.

Massive main-sequence stars: We simulate core convection and convective boundary mixing in massive stars (e.g. 25 M⊙), establishing scaling relations for internal gravity wave (IGW) generation and mixing. This work connects to asteroseismic observations of stochastic low-frequency variability, with recent results showing that envelope convection excites IGW eigenmodes that dominate the brightness variations seen by space telescopes. We also investigate the effects of radiation pressure and radiative diffusion, as well as rotation, on convective core dynamics.

Red giant branch (RGB) stars: Our 3D simulations of red giant interiors, from the luminosity bump to the tip of the red giant branch, explore internal gravity wave excitation and propagation in evolved low-mass stars. Recent work published in Nature Astronomy demonstrates that rotation amplifies wave-driven mixing by more than two orders of magnitude, offering an explanation for long-standing chemical abundance anomalies in red giants such as the lithium-rich red clump star problem.

Other stellar types: We have performed the first 3D simulations of core convection in 10⁴ M⊙ supermassive stars from the early Universe (so-called Population III stars), and the first full-sphere 3D simulations of core-helium-burning stars, where partially mixed (semiconvective) layers may be erased by convective dynamics.

Dynamic nucleosynthesis

A central theme of our research is nucleosynthesis driven by dynamic mixing events in stellar interiors where convection interacts with nuclear burning — what we call dynamic nucleosynthesis.

Intermediate neutron-capture process (i-process): We study the i-process, which operates at neutron densities (~1015 cm-3) intermediate between the slow (s) and rapid (r) neutron-capture processes. Using 3D hydrodynamic simulations of hydrogen-ingestion events (e.g. in Sakurai’s object and rapidly accreting white dwarfs), we develop advective two-stream post-processing methods to compute i-process yields. These yields successfully reproduce abundance patterns observed in carbon-enhanced metal-poor (CEMP) stars that show signatures of both the r- and s-processes. We also collaborate with experimental nuclear physicists on key reaction rate measurements that constrain i-process conditions, with results published in Physical Review Letters and a comprehensive review in Nature Reviews Physics.

Oxygen-carbon shell mergers: We investigate nucleosynthesis in oxygen-carbon shell mergers in massive stars prior to core collapse. This work shows that 3D mixing uncertainties can be as significant as nuclear physics uncertainties for the production of proton-rich (p) nuclei, and that these pre-supernova events may be important sources of ⁴⁴Ti, light odd-Z elements, and ⁴⁰K — with implications for radiogenic heating in rocky exoplanets.

Stellar evolution and nucleosynthesis

We contribute to the broader field of stellar nucleosynthesis through the NuGrid collaboration, producing comprehensive asymptotic giant branch (AGB) evolution and nucleosynthesis data sets. Our work extends to heavy element production in Type Ia supernovae, calcium production in novae, and sodium-22 yields from core-collapse supernovae relevant to the composition of presolar grains found in meteorites. We also study white dwarf cooling ages and the chemical evolution of stellar populations such as the globular cluster ω Centauri.

Cyberinfrastructure

We develop Cyberhubs — virtual research environments based on Jupyter and Docker that enable collaborative data analysis, visualization, and reproducible workflows for the NuGrid and PPMstar collaborations. Our two active platforms are ppmstar.org and astrohub.uvic.ca.