Enhanced-Fidelity Simulations of Stratified and High-Reynolds-Number Turbulence Using Embedded Measurements
Abstract
Enhanced-Fidelity Simulations of Stratified and High-Reynolds-Number Turbulence Using Embedded MeasurementsSOW:Any attempt to accurately predict turbulence must contend with two key challenges: its multiscaleand its chaotic nature. Regarding the former, as the Reynolds number of a flow increases, the separationof scales between its largest and smallest eddies becomes larger. As for the chaotic natureof the flow, the dynamics can be sensitive to additional effects such as stratification, and chaos ispromoted at higher Reynolds numbers: small changes in the flow lead to exponentially divergingevolutions in time according to a Lyapunov exponent proportional to the Reynolds number. Thesedifficulties affect both experiments and simulations, which have complementary strengths and eachits own weakness. Experiments can be performed at high Reynolds numbers and can accuratelyrepresent complex effects. However, measurements must be non-intrusive or else they alter theflow and sensing is a compromise between spatiotemporal resolution and measurement region.Computations on the other hand provide non-intrusive access to the entire field. However, directnumerical simulation can be performed at relatively low Reynolds numbers only due to the computationalcost associated with resolving all the flow scales, and turbulence modeling is required forhigher values. Even high-fidelity models such as large-eddy simulations remain a matter of activeresearch in canonical configurations of wall turbulence at high Reynolds numbers and for complexphenomena such as stratification. The complementarity of experiments and simulations motivatedus to develop a framework for Enhanced-Fidelity Simulations where we fuse measurements intonumerical simulations. In doing so, we augment the measurements with computational data thatsatisfy the flow equations, and we enhance the fidelity of the simulations by guaranteeing that theyrepresent the physical flow. We have demonstrated this technique for direct numerical simulationsof wall turbulence at moderate Reynolds numbers. In this effort, we extend the framework tolarge-eddy simulations of high Reynolds numbers turbulence and stratification.Approved for public release1
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Apr 29, 2020
- Source ID
- N000142012347
Entities
People
- Tamer A. Zaki
Organizations
- Johns Hopkins University
- Office of Naval Research
- United States Navy