Доступ предоставлен для: Guest
TSFP DL Home Архив Исполнительный Комитет

A STOCHASTIC LOW-ORDER MODELLING APPROACH FOR TURBULENT SHEAR FLOWS

Guillaume Lehnasch
Department of Fluid, Thermal and Combustion Sciences Pprime Institute, CNRS-Universite de Poitiers, ISAE-ENSMA 43 rue de I'Aerodrome, F-86036 Poitiers CEDEX, France

Joel Delville
Department of Fluid Flow, Heat Transfer and Combustion Institute Pprime - UPR 3346 - CNRS - Universite de Poitiers - ENSMA CEAT - F86036 Poitiers cedex, France

Аннотация

A reduced-order modelling, based on the POD-Galerkin approach, is developped in a probabilistic framework in order to build robust low-dimensional dynamical systems able to reproduce the dominant long-term dynamics of turbulent flows. It combines a statistical inference procedure to determine joined probability distribution functions of the dynamical system coefficients and a Monte-Carlo approach to foresee its most probable time evolution. The method is shown to reproduce correctly the main dynamical features of a turbulent mixing layer.