skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Large-eddy simulation of the Rayleigh-Taylor instability on a massively parallel computer

Technical Report ·
DOI:https://doi.org/10.2172/86949· OSTI ID:86949
 [1]
  1. California Univ., Davis, CA (United States). Dept. of Applied Science

A computational model for the solution of the three-dimensional Navier-Stokes equations is developed. This model includes a turbulence model: a modified Smagorinsky eddy-viscosity with a stochastic backscatter extension. The resultant equations are solved using finite difference techniques: the second-order explicit Lax-Wendroff schemes. This computational model is implemented on a massively parallel computer. Programming models on massively parallel computers are next studied. It is desired to determine the best programming model for the developed computational model. To this end, three different codes are tested on a current massively parallel computer: the CM-5 at Los Alamos. Each code uses a different programming model: one is a data parallel code; the other two are message passing codes. Timing studies are done to determine which method is the fastest. The data parallel approach turns out to be the fastest method on the CM-5 by at least an order of magnitude. The resultant code is then used to study a current problem of interest to the computational fluid dynamics community. This is the Rayleigh-Taylor instability. The Lax-Wendroff methods handle shocks and sharp interfaces poorly. To this end, the Rayleigh-Taylor linear analysis is modified to include a smoothed interface. The linear growth rate problem is then investigated. Finally, the problem of the randomly perturbed interface is examined. Stochastic backscatter breaks the symmetry of the stationary unstable interface and generates a mixing layer growing at the experimentally observed rate. 115 refs., 51 figs., 19 tabs.

Research Organization:
Lawrence Livermore National Lab. (LLNL), Livermore, CA (United States); California Univ., Davis, CA (United States). Dept. of Applied Science
Sponsoring Organization:
USDOE, Washington, DC (United States); National Science Foundation, Washington, DC (United States)
DOE Contract Number:
W-7405-ENG-48
OSTI ID:
86949
Report Number(s):
UCRL-LR-119748; ON: DE95014132; TRN: AHC29521%%5
Resource Relation:
Other Information: TH: Thesis (Ph.D.); PBD: Mar 1995
Country of Publication:
United States
Language:
English