Stochastic Simulation of Scalar Mixing Capturing Unsteadiness and Small-scale Structure Based on Mean-flow Properties

A key limitation of Reynolds-Average Navier-Stokes (RANS) simulation of mixing and reaction in turbulent flows is the lack of resolution of small-scale structure and associated unsteadiness. Various subgrid models formulated in state space have been developed to complement the RANS solution in this regard. We here introduce a physical-space formulation that captures unsteady advective and diffusive processes at all scales of the turbulent flow. The approach is a 3D construction based on the Linear Eddy Model (LEM), involving three orthogonally intersecting arrays of 1D LEM domains, and coupled so as to capture the 3D character of fluid trajectories. To illustrate the model performance of the 3D LEM-based formulation, here termed LEM3D, multi-stream mixing in a turbulent round jet is simulated using measured mean-flow properties as input. Comparison to scalar cross-correlation coefficients and other measured properties of this mixing configuration indicate that the LEM3D approach, in conjunction with flow properties that are provided by steady-state models, is a useful representation of complex turbulent mixing processes that would otherwise be difficult to capture within a steady-state CFD framework.