Application of the turbulent potential model to complex flows

Abstract

Typical Reynolds-averaged navier–stokes (RANS) turbulence models are concerned with modeling the Reynolds stress tensor. However, solution of the partial differential equations governing the evolution of the Reynolds stress tensor can be difficult, and due to this reason this approach is rarely used in practical computational fluid dynamics (CFD) codes. The turbulent potential model is a RANS model that generally avoids modeling the Reynolds stress tensor and significantly predicts complex turbulent flow phenomena. It makes no equilibrium and algebraic assumptions about how the turbulence is related to the mean flow. However, the computational cost and implementation complexity of the model are comparable to modem low-Reynolds-number two-equation models. The chapter also evaluates performance of the turbulent potential model in a series of complex nonequilibrium turbulent flows, which include three-dimensional boundary layers, unsteady vortex shedding, rotating turbulent flows, and boundary layer transition.