PDF modeling of turbulent flows on unstructured grids

Abstract

In probability density function (PDF) methods of turbulent flows, the joint PDF of several flow variables is computed by numerically integrating a system of stochastic differential equations for Lagrangian particles. A mathematically exact treatment of advection, viscous effects and arbitrarily complex chemical reactions is possible; these processes are treated without closure assumptions. A set of algorithms is proposed to provide an efficient solution of the PDF transport equation modeling the joint PDF of turbulent velocity, frequency and concentration of a passive scalar in geometrically complex configurations. An unstructured Eulerian grid is employed to extract Eulerian statistics, to solve for quantities represented at fixed locations of the domain and to track particles. All three aspects regarding the grid make use of the finite element method. Compared to hybrid methods, the current methodology is stand-alone, therefore it is consistent both numerically and at the level of turbulence closure without the use of consistency conditions. Several newly developed algorithms are described that facilitate the numerical solution in complex flow geometries, including a stabilized mean-pressure projection scheme, the estimation of conditional and unconditional Eulerian statistics and their derivatives from stochastic particle fields, particle tracking through unstructured grids, an efficient particle redistribution procedure and techniques related to efficient random number generation. The solver has been parallelized and optimized for shared memory and multi-core architectures using the OpenMP standard. Relevant aspects of performance and parallelism on cache-based shared memory machines are discussed and presented in detail. The methodology shows great promise in the simulation of high-Reynolds-number incompressible inert or reactive turbulent flows in realistic configurations.