Development and Implementation of Practical Optimal LES Models

University Of Texas At Austin, Austin TX

Investigators

Abstract

ABSTRACT PROPOSAL NO.: CTS-0352552 PRINCIPAL INVESTIGATOR: ROBERT D. MOSER INSTITUTION: UNIVERSITY OF ILLINOIS, URBANA-CHAMPAIGN DEVELOPMENT AND IMPLEMENTATION OF PRACTICAL OPTIMAL LES MODELS In many technological devices, it is of great importance to predict and/or control the effects of turbulent fluid flows, because the behavior of the turbulence may ultimately control the performance of the device. Large eddy simulation (LES) is one of the most promising techniques for reliable prediction of the high Reynolds number turbulent flows of technological interest. In LES, the largest scales of turbulence are simulated while the effects of the smallest scales are modeled. However, there are a number of challenges to overcome for LES to realize its full potential; these include the treatment of near-wall turbulence and the impact of numerical discretization, especially in flows with complex geometries. The objective of the proposed research is to produce LES models that will provide reliable predictions in a broad range of turbulent flows. The approach is to apply the "Optimal LES" formulation, in which LES models are developed through a formal optimization process. Optimal LES models are formulated in terms of small separation multi-point velocity correlations. The primary remaining challenge in developing such models is obtaining the required velocity correlations, without resorting to extensive empirical data. The required statistical correlations will be developed using dynamic techniques, in which the correlations are estimated from the running simulation and from theoretical considerations. Of particular interest in the current project will be modeling near-wall turbulence, scalar transport and compressibility. The results of this development will be a broadly applicable and accurate LES modeling technique, which will be implemented in application CFD codes. Optimal LES will be developed into a powerful predictive tool in a wide range of turbulent flows. This will have great impact on many engineering applications in which turbulence is one of the dominant phenomena. Important examples are: flow in gas turbine engines, flow over multi-element high-lift airfoils, and flow in stirred tank reactors. With reliable LES simulations, these and many other applications will enjoy improved new designs, and reduced design costs arising from the decreased need for testing.