A. Silva Lopes

Simulation of the Askervein Flow. Part 1: Reynolds Averaged Navier–Stokes Equations (k∈ Turbulence Model)

The neutrally stratified flow over the Askervein Hill was simulatedusing a terrain-following coordinatesystem and a two-equation(k - ∈) turbulence model. Calculations were performed on awide range of numerical grids to assess, among other things, theimportance of spatial discretization and the limitations of theturbulence model. Our results showed that a relatively coarse gridwas enough to resolve the flow in the upstream region of the hill;at the hilltop, 10 m above the ground, the speed-up was 10% lessthan the experimental value. The flows most prominent feature wasa recirculating region in the lee of the hill, which determinedthe main characteristics of the whole downstream flow. This regionhad an intermittent nature and could be fully captured only in the caseof a time-dependent formulation and a third-order discretization ofthe advective terms. The reduction of the characteristic roughnessnear the top of the hill was also taken into account, showing theimportance of this parameter, particularly in the flow close to theground at the summit and in the downstream side of the hill.Calculations involving an enlarged area around the Askervein Hillshowed that the presence of the nearby topography affected the flowneither at the top nor downstream of the Askervein Hill.

Large-eddy simulation of the flow in an S-duct

Large-eddy simulations of the flow in an S-shaped, two-dimensional duct were conducted using a Lagrangian dynamic eddy viscosity subgrid-scale model and a non-orthogonal grid system. The simulations were performed at Re b = 13,800 and 30,800, using up to 13.4 million grid nodes. The results show that two factors affect the boundary layer: the concave or convex wall curvature (which, respectively, enhances or dampens the turbulent motions) and the favourable or adverse pressure gradient in the regions of curvature change, which accelerates or decelerates the boundary layer. In the presence of an adverse pressure gradient the boundary layer can separate intermittently, which also enhances the turbulent motions. The strongest separation occurs between the two curves and affects the flow not only in the following curve but also in the recovery region. The flow near the walls displays a logarithmic behaviour, but its slope and intercept vary with curvature and pressure gradient. Taylor–Görtler vortices are observed near the concave surfaces; they are responsible for strong organization of the streamwise and wall-normal velocity fluctuations, and contribute significantly to the Reynolds stresses. The turbulent kinetic energy budgets show that the production and dissipation are similarly affected by the wall curvature as the turbulent motions: they are increased by the concave wall and decreased by the convex. The mean flow advection transports turbulent kinetic energy from the region with concave curvature to the region with convex curvature.

Improving a Two-Equation Turbulence Model for Canopy Flows Using Large-Eddy Simulation

Large-eddy simulations of the neutrally-stratified flow over an extended homogeneous forest were used to calibrate a canopy model for the Reynolds-averaged Navier–Stokes (RaNS) method with the

Modelling flows within forested areas using the k—ɛ RaNS model

k-\varepsilon

Large Eddy Simulation of a Flow with Multiple Streamwise Curvatures

k-ε turbulence model. It was found that, when modelling the forest as a porous medium, the canopy drag dissipates the turbulent kinetic energy (acts as a sink term). The proposed model was then tested in more complex flows: a finite length forest and a forested hill. In the finite length forest, the destruction of the turbulent kinetic energy by the canopy was overestimated near the edge, for a length approximately twice the tree height. In the forested hill, the model was less accurate inside the recirculation zone and overestimated the turbulent kinetic energy, due to an incorrect prediction of the production term. Nevertheless, the canopy model presented here provided consistent results in both a priori and a posteriori tests and improved the accuracy of RaNS simulations with the

Numerical simulation of isotropic turbulence using a collocated approach and a nonorthogonal grid system

k-\varepsilon