Getting started with CFD

On this page

• Introduction to CFD
• Turbulence
  • Boussinesq hypothesis
• Discretisation errors
  • Validation and verification
• Near wall flow

If you are embarking on a CFD project and have little experience I would highly recommend the resources on this page. Alternatively, if you have some CFD experience, but need a refresher, then this page will also be useful.

Introduction to CFD

The content of this page is based on the Coursera module Sports and Building Aerodynamics , and supplemented using resources found on this webpage. To get started, watch the first video of week 3 of the module on the Coursera webpage:

• What, why and how?

Turbulence

Turbulence is fundamental in the study of external flow. There are a number of ways to re-create turbulence in CFD ranging from highly accurate and expensive to low fidelity relatively cheap approaches.

• Approximate forms of the Navier-Stokes equations

Supplement the above video with an introduction of some widely used eddy-viscosity turbulence models used for ship hydrodynamics:

• Application of eddy-viscosity turbulence models to problems in ship hydrodynamics
• Examples of Detached Eddy Simulation for ship CFD: Prediction of the aerodynamic behaviour of a full-scale naval ship in head waves using Detached Eddy Simulation

Boussinesq hypothesis

The Boussinesq hypothesis is a fundamental idea in turbulence models. The following video explains in detail Boussinesq hypothesis:

• Turbulence modelling

Note that the gradient diffusion assumption is only relevant where energy and temperature is modelled.

Discretisation errors

Discretisation errors are one of the main sources or error in any CFD simulation. The following video explains the basics of estimating discretisation error estimation using Richardson extrapolation:

• Some aspects of discretization

Validation and verification

The following video gives a good overview of validation and verification practices.

• Errors and uncertainty, verification and validation

Supplement this video with the following paper on uncertainty estimation for ship CFD:

• A posteriori error and uncertainty estimation in computational ship hydrodynamics

You can also play around with some values using the Discretisation uncertainty app and use it predict the discretisation uncertainty of your simulation, whose documentation can be found here.

Near wall flow

Near wall modelling is of critical importance to predict the flow around any structure. The following video explains the differences, advantages and disadvantages of different modelling principles:

• Near-wall modelling

Supplement this video with the near wall modelling section of [this paper]. Your chosen methodology can be applied by using the Near wall grid set up app. That is to say, you do not need to work through the iterative process mentioned in the video above.