15.1. Introduction

Turbulence is the three-dimensional unsteady random motion observed in fluids at moderate to high Reynolds numbers. As technical flows are typically based on fluids of low viscosity, almost all technical flows are turbulent. Many quantities of technical interest depend on turbulence, including:

  • Mixing of momentum, energy and species

  • Heat transfer

  • Pressure losses and efficiency

  • Forces on aerodynamic bodies

While turbulence is, in principle, described by the Navier-Stokes equations, it is not feasible in most situations to resolve the wide range of scales in time and space by Direct Numerical Simulation (DNS) as the CPU requirements would by far exceed the available computing power for any foreseeable future. For this reason, averaging procedures have to be applied to the Navier-Stokes equations to filter out all, or at least, parts of the turbulent spectrum. The most widely applied averaging procedure is Reynolds-averaging (which, for all practical purposes is time-averaging) of the equations, resulting in the Reynolds-Averaged Navier-Stokes (RANS) equations. By this process, all turbulent structures are eliminated from the flow and a smooth variation of the averaged velocity and pressure fields can be obtained. However, the averaging process introduces additional unknown terms into the transport equations (Reynolds Stresses and Fluxes) that need to be provided by suitable turbulence models (turbulence closures). The quality of the simulation can depend crucially on the selected turbulence model and it is important to make the proper model choice as well as to provide a suitable numerical grid for the selected model. An alternative to RANS are Scale-Resolving Simulation (SRS) models. With SRS methods, at least a portion of the turbulent spectrum is resolved in at least a part of the flow domain. The most well-known such method is Large Eddy Simulation (LES), but many new hybrids (models between RANS and LES) are appearing. As all SRS methods require time-resolved simulations where the time step size is relatively small, it is important to understand that these methods are substantially more computationally expensive than RANS simulations.

Ansys Fluent provides the following choices of turbulence models:

  • Spalart-Allmaras model

  • - models

    • Standard - model

    • Renormalization-group (RNG) - model

    • Realizable - model

  • - models

    • Standard - model

    • Baseline (BSL) - model

    • Shear-stress transport (SST) - model (enabled by default)

    • Generalized - (GEKO) model

  • model (add-on)

  • Transition -- model

  • Transition SST model

  • Reynolds stress models (RSM)

    • Linear pressure-strain RSM

    • Quadratic pressure-strain RSM

    • Stress-Omega RSM

    • Stress-BSL RSM

  • Scale-Adaptive Simulation (SAS) model, which can be used in combination with one of the following -based URANS models:

    • SST - model

    • Standard - model

    • BSL - model

    • Transition SST model

    • -based Reynolds stress models (RSM)

  • Detached eddy simulation (DES) model, which includes one of the following RANS models.

    • Spalart-Allmaras RANS model

    • Realizable - RANS model

    • SST - RANS model

    • BSL - RANS model

    • Transition SST model

  • Shielded Detached Eddy Simulation (SDES) model, which includes one of the following RANS models.

    • SST - RANS model

    • BSL - RANS model

    • Transition SST model

  • Stress-Blended Eddy Simulation (SBES) model, which includes one of the following RANS models.

    • SST - RANS model

    • BSL - RANS model

    • Transition SST model

  • Large eddy simulation (LES) model, which includes one of the following subgrid-scale models.

    • Smagorinsky-Lilly subgrid-scale model (with or without dynamic stress enabled)

    • WALE subgrid-scale model

    • Dynamic kinetic energy subgrid-scale model

    • Wall-Modeled LES (WMLES)

    • Wall-Modeled LES - (WMLES -)