Improving the Predictive Capability of Turbulence Models Using Evidence Theory

摘要

When turbulence models are used in the simulation of a flow for which experimental results do not exist, there is, as yet, no reliable procedure for choosing a model or for quantifying the uncertainty of the results. In fact, even if experimental data are available, there appears to be no general procedure to quantify uncertainty in simulations. The present study explores the potential of employing the Dempster-Shafer theory of evidence to address these issues and proposes an approach to quantify uncertainty in turbulent flow simulations. The approach is tested using a subsonic flow around the RAE 2822 airfoil.