Final Report on the ? New Directions in the Variational Multiscale Formulation of Large Eddy Simulation of Turbulence?

The work performed as part of this grant may be described in the context of developing numerical methods for solving an abstract, nonlinear variational problem for which the straightforward Galerkin approximation on a space of reasonable dimension is inaccurate. Target applications include flows with turbulence and systems with shocks. With this in mind, an alternate finite dimensional problem is proposed in which a model term (with unknown parameters) is added to the Galerkin approximation to improve its performance. We have developed new formulations for this model term especially for turbulent flows and for systems with shocks. We have also developed a dynamic approach for evaluating the parameters that appear in this model term. We have conducted several numerical studies to confirm the usefulness of the proposed approach