Laboratory investigation of flow structure and resistance of high- and low-angle dunes

A prominent control on the flow over subaqueous dunes is the slope of the downstream leeside. While previous work has focused on steep (~30°), asymmetric dunes with permanent flow separation, little is known about dunes with lower lee-slope angles for which flow separation is absent or intermittent. Here, we present a laboratory investigation where we systematically varied the dune lee-slope, holding other geometric parameters and flow hydraulics constant, to explore effects on the turbulent flow field and flow resistance. Three sets of fixed dunes (lee-slopes of 10°, 20° and 30°) were separately installed in a 15 m long and 1 m wide flume and subjected to 0.20 m deep flow. Measurements consisted of high-frequency, vertical profiles collected with a Laser Doppler Velocimeter (LDV). We show that the temporal and spatial occurrence of flow separation decreases with dune lee-slope. Velocity gradients in the dune leeside depict a free shear layer downstream of the 30° dunes and a weaker shear layer closer to the bed for the 20° and 10° dunes. The decrease in velocity gradients leads to lower magnitude of turbulence production for gentle lee-slopes. Aperiodic, strong ejection events dominate the shear layer, but decrease in strength and frequency for low-angle dunes. Flow resistance of dunes decreases with lee-slope; the transition being non-linear. Over the 10°, 20° and 30° dunes, shear stress is 8%, 33% and 90 % greater than a flat bed, respectively. Our results demonstrate that dune lee-slope plays an important, but often ignored role in flow resistance.

Flow conditions, bed form dimensions and shear stress experiments

Large asymmetric bed forms commonly develop in rivers. The turbulence associated with flow separation that develops over their steep lee side is responsible for the form shear stress which can represent a substantial part of total shear stress in rivers. This paper uses the Delft3D modeling system to investigate the effects of bed form geometry and forcing conditions on flow separation length and associated turbulence, and bed form shear stress over angle-of-repose (30 lee side angle) bed forms. The model was validated with lab measurements that showed sufficient agreement to be used for a systematic analysis. The influence of flow velocity, bed roughness, relative height (bed form height/water depth), and aspect ratio (bed form height/length) on the variations of the normalized length of the flow separation zone, the extent of the wake region (where the turbulent kinetic energy (TKE) was more than 70% of the maximum TKE), the average TKE within the wake region and the form shear stress were investigated. Form shear stress was found not to scale with the size of the flow separation zone but to be related to the product of the normalized extent of the wake region (extent of the wake region/extent of water body above the bed form) and the average TKE within the wake region. The results add to understanding of the hydrodynamics of bed forms and may be used for the development of better parameterizations of smallscale processes for application in large-scale studies.

RANS simulations of wind flow at the Bolund experiment

As part of their development, the predictions of numerical wind flow models must be compared with measurements in order to estimate the uncertainty related to their use. Of course, the most rigorous such comparison is under blind conditions. The following paper includes a detailed description of three different wind flow models, all based on a Reynolds-averaged Navier-Stokes approach and two-equation k-ε closure, that were tested as part of the Bolund blind comparison (itself based on the Bolund experiment which measured the wind around a small coastal island). The models are evaluated in terms of predicted normalized wind speed and turbulent kinetic energy at 2 m and 5 m above ground level for a westerly wind direction. Results show that all models predict the mean velocity reasonably well; however accurate prediction of the turbulent kinetic energy remains achallenge.

Comparison of particle dispersion obtained with simple stochastic models and les in pipe flow

A hybrid Eulerian-Lagrangian approach is employed to simulate heavy particle dispersion in turbulent pipe flow. The mean flow is provided by the Eulerian simulations developed by mean of JetCode, whereas the fluid fluctuations seen by particles are prescribed by a stochastic differential equation based on normalized Langevin. The statistics of particle velocity are compared to LES data which contain detailed statistics of velocity for particles with diameter equal to 20.4 µm. The model is in good agreement with the LES data for axial mean velocity whereas rms of axial and radial velocities should be adjusted.