2008 – Improving confidence in computational fluid dynamics for spillways

Richard John Kenny

Abstract: Physical hydraulic models have traditionally provided the input to spillway design.

However, the application of Computational Fluid Dynamics (CFD) is increasing in popularity.
The reliability of CFD depends on several complex issues including the physics of the flow regime, the mathematical formulae describing the fluid flow and the solution method. CFD reliability should be confirmed for each particular application, including ogee spillways (common discharge structures on dams).

A literature review has been undertaken to determine validation techniques and the reliability of CFD related to ogee spillways. There is a great deal of information available about the aeronautical, marine, automotive and mining industries but relatively little available about ogee spillways. The degree of validation for ogee spillways was found to vary significantly and raises the question of whether better validation can be achieved. Better validation may improve confidence in CFD techniques.

A project has been undertaken through Queensland University of Technology to model a prototype scale notional 2D ogee spillway using FLUENT CFD software under steady state conditions. The project tests the sensitivity of the CFD results to various parameters. The sensitivity of the inlet mass flow rate was tested for several CFD parameters. The investigation considered:

1.The grid spacing required for geometric convergence;

2.The effect of domain extent on the computational results;

3.The influence of a boundary layer and wall roughness;

4.The importance of the choice of turbulence model; and

5.The impact of inlet turbulence assumptions.

The recommendations of the sensitivity investigations were used for validation against the United States Army Corp of Engineers (USACE) design data.

Excellent agreement was found between the CFD predictions and the USACE values.

Keywords: Computational fluid dynamics (CFD), dam, flood routing, fluent, Navier-Stokes equations, Numerical/computational modelling, Ogee spillway crest, physical model, turbulence.