Search

The maturation of high performance computer archi- tectures and computational algorithms has permitted de- tailed numerical modeling to play an increased role in the design and development of advanced military and com- mercial systems. Many applications, which in the past have relied solely on empiricism and sparse experimen- tal measurements, can now be analyzed rigorously using time dependent solution algorithms. This has prompted the development of a new generation of models that at- tempt to combine the robustness and efficiency offered by the Reynolds-Averaged Navier-Stokes equations with the higher level of modeling offered by the equations de- veloped for Large Eddy Simulation. The present study focuses on modifications to common second-order accu- rate Reynolds-Averaged Navier-Stokes algorithms that enhance the ability to capture large eddy motions in re- gions governed by the LES equations. Both incompress- ible (Poiseuille flow) and compressible (supersonic base flow) cases were considered. A simple means of control- ling numerical dissipation within a second-order accurate framework was shown to resolve turbulence structures at scales similar to those obtained using higher-order meth- ods with similar grids. The choice of flux limiter was also found to be a critical parameter in LES regions.