Uncertainty estimation in large eddy simulations of realistic hypersonic flows
Abstract
The objective of the proposed research is to develop a systematic and computationally affordable method to estimate the uncertainty in the results of turbulence resolving simulations, particularly large eddy simulations (LES). The work will center on the development of a method to compute the approximate sensitivity of engineering quantities of interest with respect to a large range of uncertain parameters, including insufficiently known boundary or initial conditions, unknown geometrical deviations between the physical and computational models, insufficiently fine computational grids, and inaccurate physics models. The significance of the work is that it constitutes an important and necessary step towards enabling the use of LES in critical decision making, since the level of confidence in a decision (e.g., “does this design produce sufficiently low drag?”) is directly tied to the level of uncertainty in the underlying prediction. The project involves two close collaborations that will help provide the large range of data needed to validate the uncertainty estimation method. Sergio Pirozzoli of the University of Rome will use DNS to compute the “true” effect of known injected perturbations of different kinds (“known uncertainties”), while Rodney Bowersox of Texas AandM University will supply experimental data of turbulent boundary layers at high Mach and Reynolds numbers.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Jan 14, 2022
- Source ID
- FA95501910210
Entities
People
- Johan Larsson
Organizations
- Air Force Office of Scientific Research
- United States Air Force
- University of Maryland