Dynamic testing of subgrid models in large eddy simulation based on the Germano identity

Normally the Germane identity is used in large eddy simulation (LES) of turbulent flows to evaluate model coefficients so as to minimize the square error associated with replacing a particular model in the identity. However, one can also explicitly evaluate this error and thus establish the accuracy with which a particular model is capable of reproducing the flow features at scales between the grid and test filters. This idea is tested a priori based on tophat filtered experimental data in the far field of a round jet [Liu et al., J. Fluid Mech. 275, 83 (1994)]. Of the models considered, the mixed nonlinear model yields the smallest error. This approach offers the capability to dynamically choose among available parametrizations, based only on the resolved fields during LES. (C) 1999 American Institute of Physics. [S1070-6631(99)00102-6].