Petascale Simulation of Turbulent Stellar Hydrodynamics

University Of Minnesota-Twin Cities, Minneapolis MN

Investigators

Abstract

This award facilitates scientific research using the large new computational resource named Blue Waters being developed by IBM and scheduled to be deployed at the University of Illinois in 2011. It provides travel funds to support technical coordination between the principal investigators, the Blue Waters project team and vendor technical team. The project involves preparing numerical codes to be run on Blue Waters in a study of the course of helium flash events in early generation giant stars near the end of their lives. This relatively short duration (e.g. two years) event of accelerated nuclear burning can dredge up material enriched with heavier elements from the interior of the star and mix them with the outer envelope of the star as the latter is expelled, forming a planetary nebula. The plan is to first execute 3D, global simulations of this violent burning phase for short periods at high resolution to capture the turbulent entrainment of unburned hydrogen into the helium convection zone. Statistical analysis of the output from these high-resolution calculations will provide subgrid-scale parameterizations of turbulent, multifluid entrainment, mixing and nuclear burning that can then be incorporated in coarser resolution models. The coarser resolution models with parameterizations of the smaller scales of entrainment and turbulent combustion will then be integrated over much longer periods to determine the evolution of the helium flash. The project also involves using a novel technique to organize the computational work on the Blue Waters processors to use their computational power more effectively. If successful, the proposed work on Blue Waters will provide insights into the origin and dispersal of heavy elements in the universe. It could also provide a computational technique that would benefit a number of areas of research that need to make efficient use of the emerging class of very large, parallel, multi-core computing systems.