Turbulence Generation from a Sweeping-Based Stochastic Model

Abstract

Stochastic methods are widely used because they constitute a low-cost computational-fluid-dynamics approach to synthesize a turbulent velocity field from time-averaged variables of a flowfield. A new combined stochastic method based on the sweeping hypothesis is introduced in this paper. This phenomenon, stating that inertial range structures are advected by the energy containing eddies, is known to be an important mechanism of the turbulent velocity field decorrelation process. The proposed method presents the advantage of being easily implementable and applicable to any three-dimensional configuration as long as a steady Reynolds-averaged Navier–Stokes computation of the flow is available and assuming that the considered turbulence physics is compatible with the hypotheses made to build the current numerical model. The developed method is applied on a subsonic round cold free jet. The validation study shows that the synthesized turbulent velocity fields reproduce statistical features of the flow, such as two-point two-time velocity correlation functions, comparable to those found experimentally and integrates shear effects of the mean flow. The mean convection velocity of the turbulent structures is also correctly modeled. In addition, the turbulent kinetic energy spatial distribution is preserved by the stochastic method.