Exploring small-scale structure and anisotropy in Rayleigh-Taylor instabilities using a homogeneous, statistically stationary simulation framework

Rayleigh-Taylor (RT) instabilities account for buoyancy-generated turbulence in type 1A supernovae, wildfires, and geothermal vents, and have applications in inertial confinement fusion. Direct numerical simulations of canonical configurations have provided some insight into their global behavior, but a thorough understanding of the finer turbulent structure remains elusive. This is in large part due to RT flows’ inhomogeneity, temporal variation, and sensitivity to initial conditions impeding the collection of accurate, meaningful statistics from the mixing layer. To circumvent these obstacles, the authors use a simulation framework called homogeneous buoyant turbulent (HBT). Having been validated against DNS data of the full canonical RT configuration, HBT models the RT mechanism of turbulence production as source terms within a triply periodic box to reproduce the turbulent structure of RT flows in a homogeneous, statistically stationary environment. Aspects of the turbulence structure that are usually subject to large statistical fluctuations, such as conditional means, spectra of the kinetic energy and scalar variance budgets, and anisotropy spectra are measured within a high degree of statistical certainty. These measurements and their relationship to the parameters of RT flows, including Reynolds number and Atwood number, are discussed, as well as potential future application of these measurements to the development of subgrid scale models for buoyant turbulent flows.