Alice Lefebvre

Predicting bed form roughness: the influence of lee side angle

Flow transverse bedforms (ripples and dunes) are ubiquitous in rivers and coastal seas. Local hydrodynamics and transport conditions depend on the size and geometry of these bedforms, as they constitute roughness elements at the bed. Bedform influence on flow energy must be considered for the understanding of flow dynamics, and in the development and application of numerical models. Common estimations or predictors of form roughness (friction factors) are based mostly on data of steep bedforms (with angle-of-repose lee slopes), and described by highly simplified bedform dimensions (heights and lengths). However, natural bedforms often are not steep, and differ in form and hydraulic effect relative to idealised bedforms. Based on systematic numerical model experiments, this study shows how the hydraulic effect of bedforms depends on the flow structure behind bedforms, which is determined by the bedform lee side angle, aspect ratio and relative height. Simulations reveal that flow separation behind bedform crests and, thus, a hydraulic effect is induced at lee side angles steeper than 11 to 18° depending on relative height, and that a fully developed flow separation zone exists only over bedforms with a lee side angle steeper than 24°. Furthermore, the hydraulic effect of bedforms with varying lee side angle is evaluated and a reduction function to common friction factors is proposed. A function is also developed for the Nikuradse roughness (ks), and a new equation is proposed which directly relates ks to bedform relative height, aspect ratio and lee side angle.

Characterising natural bedform morphology and its influence on flow

Bedforms such as dunes and ripples are ubiquitous in rivers and coastal seas, and commonly described as triangular shapes from which height and length are calculated to estimate hydrodynamic and sediment dynamic parameters. Natural bedforms, however, present a far more complicated morphology; the difference between natural bedform shape and the often assumed triangular shape is usually neglected, and how this may affect the flow is unknown. This study investigates the shapes of natural bedforms and how they influence flow and shear stress, based on four datasets extracted from earlier studies on two rivers (the Rio Paraná in Argentina, and the Lower Rhine in The Netherlands). The most commonly occurring morphological elements are a sinusoidal stoss side made of one segment and a lee side made of two segments, a gently sloping upper lee side and a relatively steep (6 to 21°) slip face. A non-hydrostatic numerical model, set up using Delft3D, served to simulate the flow over fixed bedforms with various morphologies derived from the identified morphological elements. Both shear stress and turbulence increase with increasing slip face angle and are only marginally affected by the dimensions and positions of the upper and lower lee side. The average slip face angle determined from the bed profiles is 14°, over which there is no permanent flow separation. Shear stress and turbulence above natural bedforms are higher than above a flat bed but much lower than over the often assumed 30° lee side angle.

Wave-induced sand re-suspension at dredged gravel pits based upon hydrodynamic measurements (Tromper Wiek, Baltic Sea)

ABSTRACT Gravel pits created by anchor hopper dredging may affect regional sediment transport patterns by trapping sediments. In turn, this may cause -or enhance- erosion at the adjacent coastline. Reliable assessment of such impacts requires a good understanding of the hydro-sediment dynamic processes acting at dredged pits. This paper examines the processes for sand re-suspension from pressure, current and turbidity data collected inside and outside a single dredged pit, in a non-tidal environment (Tromper Wiek, Baltic Sea). The data confirm the generally weak sediment dynamics in the area, with waves being the main hydrodynamic agent for sediment re-mobilization. Comparisons with historical data indicate a small number of sediment re-suspension events (<15%), over a 37 months-long period, without significant difference inside and outside the pit. Suspended sediment concentration profiles are predicted inside the studied pit by a simplistic model, tuned to over-estimate sediment re-suspension. The results suggest that the depth of the excavation should be very shallow (<1 m) for the bed material to be frequently extracted out by waves, and redistributed over the area. With pits up to 7 m-deep within the extraction zone, we conclude that a significant fraction of sediment is trapped over the long-term period (years).