Laurent Schindfessel

Modelling of the tributary momentum contribution to predict confluence head losses

ABSTRACT This paper proposes a new model to determine the head losses at confluences in one-dimensional models of open channel networks, making use of a momentum conservation approach. Momentum conservation has been applied in several theoretical models for confluence head losses, giving satisfactory results in general. However, for larger confluence angles between the main channel and the incoming tributary, the model accuracy diminished. Many authors identified that a correct estimation of the tributary momentum contribution is a prerequisite for accurate results. This work reports on the development and application of a theoretical model for the tributary momentum contribution, based on similarities with the flow upstream of a circular bend in a straight open channel. It describes the two-dimensional depth-averaged flow features in the tributary under the assumption of a 90° angle confluence in which all channels have equal widths, in order to obtain the resulting momentum contribution. The proposed model predicts head losses within the same order of accuracy as a numerical model solving the shallow water equations in two dimensions throughout the confluence.

Experimental Investigation of Free Surface Gradients in a 90° Angled Asymmetrical Open Channel Confluence

State of the art numerical models have come up with a number of possibilities to treat the free surface (e.g. rigid-lid approach, interface-tracking methods, interface-capturing methods). Depending on the case at hand, the complexity of the free surface treatment can be changed. To make a well-informed choice, the modeller should be able to test the model performance in a selection of cases with different relevant processes. In this paper, an experimental set-up is exploited to provide a selection of cases with high-resolution data of the free surface levels in an open channel confluence. This will allow to evaluate the relative importance of the surface gradients and presents data for numerical modellers to assess the capabilities of their chosen free surface treatment for the adopted flow conditions. The selected case is a 90° angled asymmetrical confluence, with subcritical flow, but with a downstream Froude number high enough to have important contributions of the free surface gradients to the overall momentum balances in the measurement domain. Complementary, Large Eddy Simulations with a horizontal rigid-lid treatment of the free surface are performed. This will allow to evaluate the performance of this numerical methodology in the selected case of the confluence flow. In general, the free surface levels are found to be of major importance to the overall momentum balances in the Confluence Hydrodynamics Zone, and thus correct modelling of the effects of the free surface proves to be a prerequisite for correct simulation of the flow. With the data presented in this paper, the performance of the numerical models can be further tested with respect to the treatment of the free surface.

Experimental determination of free surface levels at open-channel junctions

The Author is acknowledged for extending knowledge of open channel confluences by providing additional experiments. More specifically, the Authors research offers a dataset with inclined channels, including not only fully subcritical flows (Type I flow), but also flows with supercritical flow in one upstream branch (Type II flow) or upstream and downstream branches (Type III flow) of the confluence. Moreover, the experimental data are used to assess the accuracy of Taylors model for predicting the increased water levels over a confluence. In this Discussion, we would like to address some issues that are not fully covered in the paper. First, the Authors experimental data will be compared with existing datasets available in the literature. Secondly, it will be verified whether the hypotheses on which Taylors model is based are met in the Authors experiments, in order to allow a fair comparison between the model and experiments. The Discussion is focused on the Type I and II flows.

Closure to “On analytical formulae for navigation lock filling-emptying and overtravel” by LAURENT SCHINDFESSEL, TOM DE MULDER, STEPHAN CREELLE and GERALD A. SCHOHL, J. Hydraulic Res. 53(1), 2015, 134–148

In the original paper, approximate explicit formulae were proposed for the following dimensionless quantities: the filling time , the amplitude and time of the first overtravel peak, as well as the frequency of the chamber surface oscillations around the equalization level. For several reasons, the Authors only discussed the first overtravel peak: (i) conciseness; (ii) the increasing difficulty of experimentally validating gradually smaller water surface deviations of subsequent extremes; and (iii) the changing oscillation period during the overtravel due to the nonlinear damping. Therefore, the frequency was defined in Eq. (13) to be only representative for the first overtravel peak.

On analytical formulae for navigation lock filling–emptying and overtravel

ABSTRACT In culvert-based navigation lock filling–emptying systems, inertia effects have a significant influence on the filling–emptying time and cause a (damped) oscillation of the water surface in the lock chamber around its equalization level, referred to as the overtravel phenomenon. In this paper, the derivation of analytical formulae for the lock filling–emptying time and overtravel peak of systems consisting of a number of identical culverts is revisited. In comparison to earlier publications, the underlying assumptions are made explicit and the importance of accounting for the surface area ratio of lock chamber to upper reservoir in case of filling (or lower reservoir in case of emptying) is pointed out. Additionally, it is shown how the applicability of the analytical formulae can be extended to lock filling–emptying systems with more complex lay-outs by using an “equivalent culvert” approach. The validity of the analytical formulae is thoroughly assessed, first by comparing to an accurate numerical solution of the governing non-linear second-order differential equations, and second, by means of experiments in a physical model.

Discussion of “Tang, H., Zhang, H., & Yuan, S. (2018). Hydrodynamics and contaminant transport on a degraded bed at a 90-degree channel confluence. Environmental Fluid Mechanics, 18(2), 443–463”

The authors measured and computed the hydrodynamics and passive scalar dispersion in 90-degree open channel confluences over flat and degraded beds with a dominant upstream or tributary inflow. The present discussion essentially deals with the direction of rotation of the secondary currents, reported for the flat bed configuration with dominant tributary inflow. This rotation direction is indeed surprisingly opposite to the ones reported in the literature, both from calculations and measurements, even if present geometry slightly differs from literature geometries.

Experimental investigation of the flow evolution in the tributary of a 90° open channel confluence

Open channel and river confluences have received a lot of attention in hydraulic literature, because of the interesting flow phenomena observed. Features such as flow acceleration, curvature, separation, mixing and recovery are combined in the confluence area into a complex 3D flow pattern. Typically, the analysis of these features is started at the upstream corner of the confluence area, and the upstream main and tributary branches are considered to be the (uniform) upstream boundary conditions. However, several indications in literature suggest the existence of flow features upstream of the confluence corner. This paper confirms, by means of measurements in a laboratory, 90° confluence flume, considerable streamline curvature in the tributary branch, upstream of the confluence. Furthermore, it shows and quantifies velocity redistribution as well as local water surface super-elevation and depression in the tributary branch. Consequently, flow fea-ture analysis in confluences should start a considerable distance upstream of the confluence.

Dynamic mode decomposition applied to the shear layer flows in an open channel confluence

In open channel confluences, two shear layers are often present: the one delineating the merging flows, called mixing layer or mixing interface, and the other situated between the separation zone and freestream flow. For the case of a 90° open channel confluence, Dynamic Mode Decomposition (DMD) is applied to time-resolved velocity fields in these shear layers, calculated by means of Large-Eddy Simulations. A selection of modes is presented, illustrating the potential of DMD for extracting coherent motion in confluence flows.

Large Eddy Simulation of the water flow around a cylindrical pier mounted in a flat and fixed bed

In the present work, a numerical model based upon the Large Eddy Simulation approach has been set up for predicting the three-dimensional flow around a cylindrical pier, mounted on a flat and fixed bed, a generic case that is relevant for the study of flow and scour around bridge piers. This turbulent flow configuration was studied experimentally by Nogueira et al. (2008) with Particle Image Velocimetry (PIV). The main goal of this paper is a first validation of the numerical model, based upon the available data. The numerical tool is capable to qualitatively reproduce the characteristic flow features around the pier, like e.g. the horseshoe vortex system and the vortex shedding in the wake. The predicted extent of the initial scour hole, based upon the bed shear nstress magnitudes, agrees well with the observations at the onset of the souring process during the lab experiments. Further quantitative validation of the numerical model will benefit from additional measurement efforts in the experiments.