Simulating Galaxy Formation with Fewer than a Trillion Zones

点击翻译按钮获取中文摘要

Critical questions in galaxy evolution center on the processes of feedback -- the injection of energy and momentum from stars and active galactic nuclei (AGN). Due to the vast difference in spatial scales between galaxies and stars, direct simulation of the relevant physics is still impossible with current computing technology; one must resort to subgrid models for key processes such as turbulence. The collaborating team on this project will adapt the FLASH-3 hydrodynamics code to test subgrid models of high Reynolds-number (low viscosity) turbulence against laboratory experiments and direct, brute-force calculations. They will then apply the modified hydrodynamic model to three astrophysical problems: (1) The role of bubbles powered by active galactic nuclei in the centers of galaxy clusters. Simulating these bubbles will require including buoyancy-driven instabilities on scales too small to be resolved directly. (2) Globular cluster heavy-element enrichment. Globular cluster stars are chemically homogeneous and must have formed from well-mixed material. Subgrid models that include shear-driven instabilities will allow detailed simulations of mixing that can constrain when and where these clusters were formed. (3) Development of supernova-driven outflows in starburst galaxies. Due to the short cooling times in the interstellar medium, this will require the modeling of supersonic turbulence. The project?s broader impact includes graduate student and postdoctoral researcher training, the presentation of public lectures, the development of an undergraduate course in computational fluid dynamics, and the release of simulation packages to the community.