Benoît Spinewine

Small-scale laboratory dam-break waves on movable beds

This paper documents a series of small-scale laboratory experiments of dam-break waves propagating over loose granular beds, realised in a dedicated flume equipped with a fast downward-moving gate. The flood wave generates intense geomorphic action as bed sediments are eroded and entrained by the flow. Two distinct bed materials, sand and Polyvinyl Chloride (PVC), and a range of initial bed configurations are explored. The sediment bed is either flat or features an upward or downward discontinuity of level across the gate. The case of an initial non-zero water depth downstream is also investigated. The flow is imaged through the sidewalls with fast digital cameras. Imaging techniques are used to identify characteristic flow regions and to derive full velocity fields by tracking the movements of individual bed sediments as well as neutrally-buoyant tracers. The data set described in the paper is provided in digital form in the electronic supplement to this special issue.

Characteristics of Velocity and Excess Density Profiles of Saline Underflows and Turbidity Currents Flowing over a Mobile Bed

Turbidity currents in the ocean and lakes are driven by suspended sediment. The vertical profiles of velocity and excess density are shaped by the interaction between the current and the bed as well as between the current and the ambient water above. We present results of a set of 74 experiments that focus on the characteristics of velocity and fractional excess density profiles of saline density and turbidity currents flowing over a mobile bed. The gravity flows include saline density flows, hybrid saline/turbidity currents and a pure turbidity current. The use of dissolved salt is a surrogate for suspended mud that is so fine that it does not settle out readily. Thus, all the currents can be considered to be model turbidity currents. The data cover both Froude-subcritical and Froude-supercritical regimes. Depending on flow conditions, the bed remains flat or bed forms develop over time, which in turn affect vertical profiles. For plane bed experiments, subcritical flow profiles have velocity peaks located higher up in the flow, and display a sharper interface at the top of the current, than their supercritical counterparts. The latter have excess density profiles that decline exponentially upward from the bed, whereas subcritical flows show profiles with a region near the bed where excess density varies little. Wherever bed forms are present, they have a significant effect on the profiles. Especially for Froude-supercritical flow, bed forms push the location of peak velocity upward, and render the near-bed fractional excess density more uniform. In the case of subcritical flow, bed forms do not significantly affect fractional excess density profiles; velocity profiles are pushed farther upward from the bed than in the case of a plane bed, but to a lesser extent than for supercritical bed forms. Overall, the relative position of the velocity peak above the bed shows a dependence upon flow regime, being lowered for increasing Froude number Fd. Gradient Richardson numbers Rig in the near-bed region increase with increasing Fd, but are lower than the critical value of 0.25, indicating that near-bed turbulent structures are not notably suppressed. At the top interface, values of Rig are above the critical value for subcritical and mildly supercritical Fd, effectively damping turbulence. However as Fd increases, Rig goes below the critical value. Shape factors calculated from the profiles for use in the depth-averaged equation of motion are evaluated for different flow and bed conditions. Normalized experimental profiles for supercritical currents scale up well with observations of field-scale turbidity currents in the Monterey Canyon, and the range of average bed slopes and Froude numbers also compare favorably with estimated field-scale flow conditions for the Amazon canyon and fan. This suggests that the experimental results can be used to interpret the kinds of flows that are responsible for the shaping of major submarine canyon-fan systems.

Dam-break induced sediment movement: Experimental approaches and numerical modelling

The present paper aims to present the issues and the scope of work conducted under the framework of the European Research Project IMPACT, in the field of dam-break induced geomorphic flows, at the Université Catholique de Louvain, Belgium. Two characteristic behaviours are described: the near- and far-field responses to the dam-break wave. This paper gives an overview of the experimental work carried out in the frame of the research programme: flat- and stepped-bed cases for the near-field, bank erosion experiment for the far-field. New developments in modelling are summarized as well for the near-field and for the far-field. The validation by comparison with experimental results highlights that a two-layer model is very efficient in modelling near-field features, while an appropriate modelling of bank failure and sediment re-distribution in the cross section is the key of far-field simulation. Some practical conclusions are given for the future of dam-break wave modelling in terms of needs for additional research.

Dam-break flows over mobile beds: Experiments and benchmark tests for numerical models

In this paper, the results of a benchmark test launched within the framework of the NSF–PIRE project “Modelling of Flood Hazards and Geomorphic Impacts of Levee Breach and Dam Failure” are presented. Experiments of two-dimensional dam-break flows over a sand bed were conducted at Université catholique de Louvain, Belgium. The water level evolution at eight gauging points was measured as well as the final bed topography. Intense scour occurred close to the failed dam, while significant deposition was observed further downstream. From these experiments, a benchmark was proposed to the scientific community, consisting of blind test simulations, that is, without any prior knowledge of the measurements. Twelve different teams of modellers from eight countries participated in the study. Here, the numerical models used in this test are briefly presented. The results are commented upon, in view of evaluating the modelling capabilities and identifying the challenges that may open pathways for further research.

Dam-break induced morphological changes in a channel with uniform sediments: measurements by a laser-sheet imaging technique

This paper presents an idealised experiment of dam–break flow in a channel where both the bed and the banks are made of uniform erodible material. The aim is to study the morphological changes induced by a dam–break flow, as well in the bed profile (longitudinal erosion) as in the shape of the cross sections (lateral erosion and bank failures). The experiments show good qualitative agreement with field observations of geomorphic flows, and have the advantage that they are repeatable for what concerns the general evolution of the shape of the cross sections. A non–intrusive laser sheet technique was used to measure the shape of the cross sections. From there, a complete survey of the time evolution of the channel shape could be obtained. This survey forms a data set that can be used for the validation of numerical models.

The 1996 Lake Ha! Ha! breakout flood, Quebec: Test data for geomorphic flood routing methods

This paper describes a set of field data suitable for the testing and comparative assessment of geomorphic flood routing methods. The data pertain to a particularly severe and unusually well–documented flood event: the Lake Ha! Ha! breakout flood of July 1996 in the Saguenay Region of Québec. In this event, heavy rains combined with the incision of a new lake outlet caused a major flood, which significantly reworked the downstream valley. Published and unpublished data from multiple sources are assembled and co–registered in a common frame of reference. These data include vertical and oblique air photos, hydrological records, surface geology information, and digital terrain models of the pre– and post–flood valley topography. The spatial coverage encompasses the drained lake as well as the full length of the downstream valley. To meet the respective needs of two– and one–dimensional approaches, the topography is sampled on a Cartesian mesh as well as interpolated along evenly–spaced cross–sections. The data set described in the paper is provided in digital form in the electronic supplement to this special issue.

Self-similar long profiles of aggrading submarine leveed channels: Analytical solution and its application to the Amazon channel

Many submarine fans are coursed by well-defined leveed channels constructed by turbidity currents. The channels aggrade in time, typically accumulating sandy deposits in their beds and muddy deposits in their levees. Periodic channel avulsion acts to build up the fan as a whole. Here a first theory for the long profile of leveed channels is offered. The theory is based on the assumption that there exists a time period, well after channel initiation but before incipient avulsion, during which the channel and its levees are in a quasi-equilibrium state, concurrently aggrading and prograding onto the surrounding fan. The currents are assumed to deposit sand on the channel bed and mud on the levees. The formulation uses a steady uniform flow assumption and a sediment transport relation inherited from rivers and yields a partial differential equation for the evolution of the channel starting from any initial condition. For the ideal case of a channel forming on an initially unchannelized sloping fan, the theory predicts self-similar long profiles for the down-channel variation of channel bed slope, bed elevation, and width, as well as flow discharge and sand/mud discharges. The time evolution of the channel then amounts to a simple rescaling of the self-similar profile as it aggrades and progrades down fan. The theory, when tested against data from the Amazon channel of the Amazon Submarine Fan, shows encouraging comparisons. The generality and shortcomings of the model assumptions are discussed based on a comparative study of mud-rich and relatively sand-rich submarine fan systems.

Inertia effects in bed-load transport models

Inertia effects are seldom considered in morphological modeling, and most of the transport models were developed from laboratory experiments in steady uniform flow conditions. This paper considers first the hysteresis effects in transient flows between discharge, velocity, and bottom shear stress. These effects can be taken into account as far as the complete shallow-water equations are used. Secondly, inertia effects linked to the sediment response to acting forces are considered. Three types of models are investigated: (i) sediment movement instantaneously adapting to hydrodynamic changes, (ii) spatial or temporal lag laws to give space or time to the sediments to progressively reach the transport capacity, and (iii) a two-layer model, able to account for the inertia of the sediment layer. Finally, three examples are presented: a scour hole downstream of an apron, a jump over a mobile bed, and a dam-break wave. Inertia effects appear significant in the modeling, especially the latter case.

Simulations of the flow in the Mahakam river–lake–delta system, Indonesia

Abstract Large rivers often present a river–lake–delta system, with a wide range of temporal and spatial scales of the flow due to the combined effects of human activities and various natural factors, e.g., river discharge, tides, climatic variability, droughts, floods. Numerical models that allow for simulating the flow in these river–lake–delta systems are essential to study them and predict their evolution under the impact of various forcings. This is because they provide information that cannot be easily measured with sufficient temporal and spatial detail. In this study, we combine one-dimensional sectional-averaged (1D) and two-dimensional depth-averaged (2D) models, in the framework of the finite element model SLIM, to simulate the flow in the Mahakam river–lake–delta system (Indonesia). The 1D model representing the Mahakam River and four tributaries is coupled to the 2D unstructured mesh model implemented on the Mahakam Delta, the adjacent Makassar Strait, and three lakes in the central part of the river catchment. Using observations of water elevation at five stations, the bottom friction for river and tributaries, lakes, delta, and adjacent coastal zone is calibrated. Next, the model is validated using another period of observations of water elevation, flow velocity, and water discharge at various stations. Several criteria are implemented to assess the quality of the simulations, and a good agreement between simulations and observations is achieved in both calibration and validation stages. Different aspects of the flow, i.e., the division of water at two bifurcations in the delta, the effects of the lakes on the flow in the lower part of the system, the area of tidal propagation, are also quantified and discussed.

Three-dimensional Voronoï imaging methods for the simultaneous measurement of velocities and concentrations in dense particulate flows

A set of recently developed stereo imaging algorithms are proposed for the combined experimental characterization of 3D velocities and concentrations in dense particulate flows, imaged by twin cameras through a transparent plane. A robust 3D-PTV algorithm is described. An estimate of volumetric concentration is proposed, based on the surface density of near-wall particles, and calibrated on a set of Monte-Carlo simulations. The originality of the methods lies in exploiting the properties of the Voronoi diagram to perform in a robust and efficient way the various steps of the analysis: stereoscopic matching, volumetric sampling and particle-tracking operations. In contradiction with traditional methods for particletracking and concentration estimation, the present methods have proven to be successful even in case of highly concentrated, rapidly sheared dispersions of many identical particles. They can handle occlusion effects, which intervene for dense dispersions and corrupt concentration estimates. The methods are applied and validated on a set of fluidization cell experiments, allowing the derivation of full sets of 3D granular trajectories and estimates of the homogeneous concentration within the fluidized bed.