Improved Turbulence Modelling Across Disparate Length Scales for Naval Computational Fluid Dynamics Applications

Computational Sciences LLC will collaborate with the Rensselaer Polytechnic Institute (RPI) to develop and validate a stand-alone computational module that naturally accounts for the effects of turbulence. Such fluctuations and transitions may be associated with compressible flows and boundary layer interactions. The module will be designed for implementation in to existing legacy codes for use in characterization of unsteady vorticity-dominated flows.The approach is based on a novel, regularized set of Navier-Stokes equations (RNS) that is extended to account for turbulence effects (fluctuations) in the continuum approximation. RNS has several important features not found in classical NS equations that are of direct relevance turbulent flows: (a) Kolmogorov-scale field fluctuations resulting from a mathematical model that accounts for turbulent diffusion in a natural manner that allows direct simulation of phenomena such as laminar-turbulence transition and wall slip effects; and (b) Natural accounting of growth of small-scale turbulent structures without refining down to Kolmogorovs scale.Phase I will focus on a simplified 3D working version of the approach by removing the selected restrictions. The validation of the model will be provided by comparison of the simulation results with the experimental data for a set of representative turbulent flows. The software module will be connected to a high order compressible flow code and will be exercised and evaluated against experimental data for selected model problems that contain elements of both nearfield and farfield wakes. Phase II will refine the approach to include generalized vortical flows generated by 6 degree-of-freedom hard body interactions, and will validate it on problems of interest to the Navy.