Journal of Hydraulic Engineering

We present the results of laboratory experiments specifically designed to quantify the height and migration rate of an aggrading gravel front. The experiments were performed in sediment feed mode with constant water and sediment supply. Particular care was put into the experimental procedure and the methods to determine the quantities of interest to ensure reproducibility of the results. The celerity and height of an aggradation front were modeled as functions of the Froude number of the flow for the initial bed profile and of a load ratio defined as the ratio of the sediment feed rate to the transport capacity of the flow for the initial bed. The two control parameters (Froude number and load ratio) also determined the translational or dispersive nature of a sediment front. Two predictors were provided to estimate the dimensionless height and celerity of an aggradation front. The former was an increasing function of the load ratio and a decreasing function of the Froude number, and the latter was roughly proportional to the squared Froude number and had no evident relationship with the load ratio. The present results are of interest for scholars and practitioners needing to determine the key properties of swift gravel fronts as those developing, for example, during flash floods.

The selection of optimum sampling configurations to measure the aspects of turbulent flow of water depends on the variable being measured, the applied measurement technique, the degree of environmental noise, and flow characteristics in and near the sampling location. This work presents a method to dynamically select the exposure time (SET) during the flow-velocity measurement at each sampling location on the basis of the moving block bootstrap (MBB) technique. The MBB enables near real-time estimation of the confidence intervals and, subsequently, the COV for turbulence parameters. The dynamic SET method has been tested in this work using time series of synthetic turbulent flow-velocity signals. This method was implemented using actual flow-velocity signals recorded with acoustic Doppler velocity (ADVs) instruments and acoustic Doppler current profilers (ADCPs). The results obtained from the implementation of the dynamic SET method show that the sampling time to achieve a defined uncertainty threshold is not the same for different turbulence parameters. Consequently, using the dynamic SET method, the exposure time can be optimized to obtain the turbulence parameters with the required uncertainty level. The dynamic SET method can be implemented in the instrument’s data logger to dynamically select the exposure time during flow measurements for a given uncertainty of the required turbulence parameter.

A central element of a pipeline’s air management infrastructure is typically its set of air exchange valves (AEVs), a designation that includes air release and air/vacuum valves. Knowledge of the air mass flow rate as a function of pressure difference is essential for AEV selection and design, a relationship expressed by its somewhat complex and certainly nonlinear characteristic curve (CC). However, both measurement and simulation of this CC are often nonintuitive and, for various practical and theoretical reasons, problematic. To provide greater insight, several helpful performance scaling analyses are undertaken here with the goal of aiding system investigation, design, and operation. To this end, a performance similarity relation (PSR) for different-sized AEVs is developed and its effectiveness demonstrated in the light of commonly available air mass flow data. Also, the PSR is successfully applied to aid in the interpretation of published experimental air expulsion data. Additionally, the interaction of the CC with basic pipeline parameters on system performance is explored in the context of common hydraulic transient events. For this second application, the water-hammer pressures generated by a pump trip scenario are numerically simulated for several test pipelines, considering four AEV sizes each having three possible characteristic curves.

Theoretical and experimental results for the evolution of cavities generated by upward vertical leakage jets in porous media are presented. The formulation is based on the conservation principles of mass, momentum, and energy. Two conditions were considered: two-dimensional isotropic medium and three-dimensional isotropic medium. The results of the cavity height are given in nondimensional form and related to the Froude number that considers the inlet flow characteristics (width of the slot or diameter of the orifice, and inlet velocity). Both the two-dimensional and the three-dimensional cases provided theoretical results for the length of the cavity limited by a maximum Froude number. This characteristic implies that the cavity destabilizes for Froude numbers higher than this maximum critical value. It was observed that parameters of the porous medium (soil parameters) not present in the conservation formulation must be added a posteriori to cover a larger spectrum of soil possibilities. The experimental results show that the theoretical proposition follows the main characteristics of the observations.

Experiments were conducted in a laboratory flume in which gravel and total bed-load rates were measured continuously using independent methods that allowed evaluation of the effect of four antecedent flow conditions on the transport of bed load under standard flow. It was found that both mean total bed-load rate and gravel transport rates are related to the magnitude of antecedent flow. The changes in mean bed-load rates were attributed to changes in the stability of the coarsest fractions of the bed material, caused by the previous high-flow event. This work indicates that long-term mean bed-load transport rates for a sand and gravel bed are not just a function of grain size and flow rate but also vary with the magnitude of antecedent flow.

Hydraulic models used to simulate floods through riparian forests along rivers need to account for flow resistance caused by tree drag. Tree drag formulations have been developed for this purpose in previous studies, using drag force measurements on several broad-leaved, temperate species at tree scale to calibrate model parameters. However, tree reconfiguration and reduction in drag force in response to increased flow velocity are still not fully understood, particularly for subtropical tree species with sclerophyllous foliage. An established towing tank method was adapted to a field setting using a motorboat to tow submerged she-oak and tea trees through still water. Drag coefficients for both genera decreased with increasing velocity due to stem and foliage reconfiguration, except for the lowest velocity measurements on some she-oak trees, indicating the emergence of a rigid regime. The rigid regime has also been observed in previous studies and is important for flood modeling because drag forces could otherwise be overestimated at low velocities. Therefore, a new Cauchy number has been formulated and tested to predict regime transition for she-oak trees.

Protection layers are often used to prevent scour and erosion, e.g., prevention of scour around wind turbine foundations. However, several cases exist where installed scour protection has settled, where loss of sediment through the armor layer can explain the failure. This paper presents the use of a detailed large eddy simulation–discrete-element method (LES-DEM) model to study sediment particles in porous media. First, a simple idealized case of the removal of sediment from an idealized cavity beneath a smooth turbulent boundary layer was set up. The model showed the penetration of turbulence, mainly in the form of sweep events, into the cavity. This high momentum would at times reach the bottom and entrain the fine sediments. The sediment would subsequently roll over and form a pile and at times be suspended from the return flow of the penetrating turbulence. Finally, a more realistic armor layer was set up with a series of closely packed spheres. Fine sediments were seeded at the bed. A hydraulically rough boundary layer was developed over the armor layer, where turbulent statistics from the model compared well against experiments. Turbulent structures characteristic of the bursting process were identified in the rough wall case. The penetration of sweep events’ entrainment and suspending the finer sediments is detailed. The flushing of cavities from passing ejection events is also presented.

A numerically based investigation is carried out to determine the minimum median diameter of the riprap stone needed to avoid shear failure inside the riprap apron used to protect spill-trough abutments against erosion. The study considers identical spill-through abutments placed on the floodplains of a compound channel. Several series of Reynolds-averaged Navier-Stokes (RANS) simulations are conducted with varying floodplain width, $B_f$, ratio between the abutment length and the floodplain width, $L_a/B_f$, mean diameter of the riprap stone, $D_{50}$, and channel curvature, $R$, to estimate the maximum Froude number, Fr, at which riprap stones in aprons placed at the base of spill-through abutments will resist shear failure by the flow. A design formula for riprap size selection in aprons protecting spill-through abutments is proposed. The formula is expressed as $D_{50}/y=C^{0.5\alpha }*{\mathrm{Fr}}^{\alpha }$, where $y$ is the flow depth next to the toe of the abutment. The two model parameters are $C$ and $\alpha$. Results show that $\alpha$ is only a function of $B_f$ for abutments that do not extend until close to the main channel and $\alpha =1.85$ for abutments extending over the whole width of the floodplain.

Previous research has shown that local scour at bridge piers is severely increased by the accumulation of woody debris around piers. However, due to the unavailability of accurate information regarding the characteristics of formed debris jams, the shape and dimensions of accumulations tested in previous laboratory experiments have had to be assumed. This article provides an assessment of debris-induced scour based on recently available knowledge about the relation between the potential dimensions of debris accumulations, the characteristics of flow, and debris elements. Clear-water scour experiments (with and without debris accumulation) were conducted using debris models with shape and size that correspond to the particular flow characteristics of each experiment. The results showed that scour depths obtained with flow-dependent debris accumulations were larger than without accumulations by a factor ranging from 1.18–2.19. The analysis of the scour depths affected by the accumulations suggested similar characteristics as well as dependence on the flow intensity, blockage area ratio, and depth ratio.

The Rational Method remains one of the most widely used approaches for estimating peak discharge in small catchments. In one widely used interpretation of the Rational Method, the maximum possible peak discharge produced by a storm with a given return period is predicted by setting the storm duration equal to the time of concentration. Whether the time of concentration maximizes peak flow for a rainfall return period, however, depends on the relationship between contributing area and storm duration. Here, we show that under many conditions, using the time of concentration in the Rational Method leads to an underestimation of peak discharge. This underestimation is illustrated using two case studies on idealized hillslopes on which runoff occurs as sheet flow. We suggest that practitioners should become cognizant of the differences between the critical duration (the storm duration that maximizes peak discharge) and the time of concentration within the Rational Method and be alert to morphology and land-use patterns that are likely to cause these timescales to diverge.

Solutions for dam-break flow mainly developed for rectangular channels are not applicable to prediction of the propagation of the dam-break wave in frictional triangular channels. This study presents an approximate solution considering the frictional effect on the dam-break flow in a dry horizontal triangular channel. Wave tip velocity is solved by an implicit formula for the product of time and resistance coefficients. All other hydraulic properties in the wave tip region can be expressed as explicit functions of wave tip velocity. Meanwhile, laboratory experiments have been performed for obtaining water surface profiles of dam-break flow from which the position and velocity of the wave tip front have been derived. Results show that retardation of the wave front position is more significant with the increases in both resistance and time. The proposed analytical solution shows satisfactory agreement with measurements, and clarifies how the behavior of the dam-break wave tip is affected by channel geometry.