A local correlation-based transition model for Spalart-Allmaras turbulence model

Abstract

The Spalart Allmaras (S-A) Reynolds-Averaged Navier-Stokes (RANS) model has had considerable success in application to a wide range of turbulent flow problems. However, as turbulent kinetic energy k is not available in the S-A model, the application of sub-models of additional physical phenomena may not be possible. These phenomena could include turbulent transition models, which require a freestream turbulent intensity, or combustion closures which require a turbulent time scale, or broadband noise prediction based on k. This thesis composes of three main parts. In the first part, a comprehensive study of γ-〖Re〗_θ transition model is presented. In the second part, a local correlation-based γ-〖Re〗_θ transition model is coupled with the Spalart-Allmaras turbulence model (SA-γ-〖Re〗_θ model) in a structured parallelized Navier-Stokes solver, Flamenco and in the OpenFOAM package. The detailed and complete summary of the modified governing equations and a suite of validation and verification tests are given. The results obtained prove the validity of the SA-γ-〖Re〗_θ model. In the third part, a novel turbulence model, denoted the S-A-K model, is developed coupling the Spalart-Allmaras model with a transport equation for kinetic energy k, enabling coupling with the SA model with γ-〖Re〗_θ model (SAK-γ-〖Re〗_θ). The closure strategy of S-A-K model is proposed and validated against four standard benchmark cases. Good results are obtained using the S-A-K model compared to the results from the classical S-A turbulence model and the k-ω Shear Stress Transport Turbulence Model (k-ωSST) model, and experimental data.