High Reynolds Number Stratified Turbulent Wakes: Internal Wave Energetics, Self-similarity and Subgrid-scale Modeling
Abstract
Executive Summary Submersible wakes in stratified waters are characterized by values of body-based Reynolds number, Re?O(108), and internal Froude number Fr? O(10) to O(103). Previous and ongoing ONR-funded research by the PI has been based on parallel implicit spectral multidomain-based Large Eddy Simulations (LES) of stratified towed-sphere wakes of non-trivial cost at Re values as high as 105 and 4×105 and at Fr=4, 16 and 64. The analysis of the resulting dataset has enabled extensive insight into the impact of increasing Re on the life-cycle of a stratified turbulent wake and the associated wakeradiated internal wave (IW) field. At the core of the PI’s findings, lies the emergence, at sufficiently high Re, of persistent energetic and highly anisotropic ``secondary’’ stratified turbulence in the intermediate wake which results from the destabilization of buoyancydriven shear layers. Nevertheless, although it does resolve the incipient phase of these shear instabilities, the PI’s costly implicit LES does not faithfully reproduce the subsequent 3-D structure and energetics of the secondary turbulence as it lacks a subgrid scale (SGS) model to reliably account for its unresolved component. The proposed research aims to shift away from expensive large-scale parallel simulations and focus on self-similarity and SGS model development for the faster-turnaround prediction of mean and fluctuating wake flow fields at high Re. To this end, this proposal will capitalize on the availability of two unique datasets from simulations of stratified wakes at high Re: the PI’s implicit LES dataset and an accompanying fully-resolved Direct Numerical Simulation (DNS) dataset confined to the wake core. Both datasets have resulted from simulations enabled by an ongoing DoD-HPC Frontier project. The proposed work will leverage the PI’s ongoing Frontier-grant-driven collaborations with three leading experts in stratified turbulence, Prof. S. de Bruyn Kops (U. Mass.), Prof. J. Riley (U. Washington) and Dr. A. Muschinski (NWRA). Combined analysis of the above DNS and implicit LES data will yield the first detailed assessment of the relative strength of IW radiation as an energy sink with respect to the turbulent dissipation and mixing inside the wake core. An existing self-similarity model of mean wake evolution, inspired by low Re lab experiments, will then be adapted in collaboration with a French colleague to account for the enhanced horizontal/vertical momentum transport and IW radiation at high Re. As a next step, a priori and a posteriori testing and assessment of the limitations of existing SGS models will be performed through extensive comparison with the DNS database. Finally, a physicsbased SGS model will be constructed and tested at reduced resolutions (where the largest scales of the above shear instabilities are still resolved) against the DNS/implicit LES database. The new SGS model design will be founded on the reliable reproduction of anisotropic stratified turbulence physics, wake interface structure and IW radiation. In terms of naval interests, the hierarchy of proposed models will enable the rapid and accurate evaluation of wake features (mean width/height and characteristic velocity, turbulent stress and kinetic energy) along its entire downstream development at Re and Fr closer to navally relevant values. The shear-instability-resolving, physics-based SGS model will serve as the foundation for future mixed LES/RANS modeling efforts at Re values where these secondary turbulence will be fully contained at the subgrid scale.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Aug 08, 2016
- Source ID
- N000141512513
Entities
People
- Peter J Diamessis
Organizations
- Cornell University
- Office of Naval Research
- United States Navy