IAHR World Congress, 2019

Free-surface n Stepped Spillway Skimming Flow in A CFD Simulation

author.DisplayName 1 author.DisplayName 1 author.DisplayName 1 author.DisplayName 1 author.DisplayName 2
1LOH/IPH - Instituto de Pesquisas Hidráulicas, Universidade Federal do Rio Grande do Sul (UFRGS), Brazil
2LAHE, Eletrobras Furnas, Brazil

Over the years by developing Roller-Compacted Concrete (RCC) dams construction techniques it has increased the use of stepped spillways as the overflow and safety structure of dams. At these structures, steps promote a macro-roughness that imposes flow resistance with consequent increased energy dissipation along the chute. Thus, the use of stepped spillways leads to smaller stilling basins so with economic benefits. Besides that, the macro-roughness provided by steps accelerate the development of the boundary layer in comparison with smooth spillways by anticipating the aeration beginning of air-water multiphase flow. It changes fluid properties and reduce damage caused by cavitation. Specialized literature still does not present unanimity for the correct definition of the free-surface in multiphase flow. It is usual to define free-surface in physical hydraulic structures as the depth which air concentration is 90%. On the other hand, in computational fluid dynamics (CFD) based on the concept of volume of fluid (VOF), free-surface is given by water and air balance that is the depth with void fraction of 0.5. However, contrary to what is commonly considered, recent studies show that there is not a single void fraction isoline capable to predict free-surface all over the spillway. This paper aims to propose different void fraction values to define free-surface on aerated and non-aerated zones of skimming flow throughout a stepped spillway by using CFD simulations with concept of VOF. For this purpose, we had validated a numerical model made with Ansys-CFX commercial software by comparing numerical and experimental data for a stepped spillway in 1:10 scale, 1V:0.75H sloping, 0.40 m wide and steps with 0.06 of high assembled at Laboratório de Obras Hidráulicas (IPH/UFRGS). The simulated flows varied between 0.20 m³/(s.m) and 0.32 m³/(s.m). From this, it was verified that for the correct consideration of the free surface in this structure we should consider water fractions of the order of 0.30 in the non-aerated zone, which represents average errors of the order of 7%. In the aerated zone, it was difficult to establish a value for water fraction to represent the free-surface.

Lucas C. S. Tassinari
Lucas C. S. Tassinari








Powered by Eventact EMS