Journal of Hydraulic Engineering

Onsite surveys and 75 measurements of discharge were made on 21 high‐gradient streams (slopes greater than 0.002) for the purpose of computing the Manning roughness coefficient, n, and to provide data on the hydraulics of these streams. These data show that: (1) n varies inversely with depth; (2) n varies directly with slope; and (3) streams thought to be in the supercritical flow range were actually in the subcritical range. A simple and objective method was employed to develop an equation for predicting the n of high‐gradient streams by using multiple‐regression techniques and measurements of the slope and hydraulic radius. The average standard error of estimate of this prediction equation was 28% when tested with Colorado data. The equation was verified using other data available for high‐gradient streams. Regimeflow equations for velocity and discharge also were developed.

Experiments were conducted using one full‐sized and two reducedscale models of a pump sump to geometric scales of 1:2 and 1:4 to determine whether scale effects distort the predictive ability of hydraulic models of pump sumps with such large scale ratios. Possible scale effects on vortexing, pipe swirl, inlet losses, and air ingestion due to air‐drawing vortices have been examined using the experimental results. No significant scale effects on free‐surface vortexing were evident in the tested ranges. Pipe flow swirl and inlet losses were seen to exhibit some scale effects. The air‐core vortices encountered were generally weak with very low air withdrawals. Within the accuracy of measurements of air‐withdrawal rates, the model did not indicate any gross underprediction of air withdrawals.

Time sequential aerial photographs flown approximately 20–30 yr apart have been used to measure the lateral migration rates of 189 bends on 21 clay, sand and gravel rivers in Western Canada. Channel migration rates transformed to a reference bend curvature $(r/w=2.5)$ are shown to be a simple function of stream power, outer‐bank height, and a coefficient of resistance to lateral migration $Y_b\text{.}$ This coefficient in turn is shown to be a function of the texture of the outer bank materials, with a complex form similar to that of the Hjulstrom or Shields sediment entrainment function. The task of specifying the channel migration rate appears to be essentially a sediment transport problem.

The hydrodynamic response of turbulent flow in an open channel to the bed and bank roughness heterogeneity was demonstrated. Roughness‐induced secondary currents were numerically simulated in the presence of the vortex induced by the free surface, and the two vortices induced by the channel corner. The relative importance of these sources for lateral motion was examined. Turbulence was modeled by the energy dissipation model, combined with the algebraic stress model suggested by Naot and Rodi (1982). The three‐dimensional flow was solved using the parabolic pressure correction algorithm of Patankar and Spalding (1972). The examples suggest that the dominant presence of the strong vortex, which is induced by the free surface, restricts the effects of the roughness heterogeneity to the lower part of the channel. Still, major differences in the structure of turbulence near the channel bed are shown.

The configuration of open‐channel junctions controls local sedimentary processes, channel scour, and sidewall erosion through its influence on the flow patterns established as confluent streams compete for limited channel capacity. Entry of a tributary into a mainstream ensures detachment of flow from the channel sidewall immediately downstream from the junction, and flume experiments show well‐defined relationships between the dimensions of the separation zone and discharge from the tributary. Other things equal, both the width and length of the separation zone increase systematically with an increase in confluence angle, though it is shown that values of width are much less than predicted by recent mathematical modelling. The zone grows at the expense of the proportion of the channel occupied by the immediate post‐confluence flow with the consequence that near‐bottom velocity increases by a factor of 1.3 as confluence angle is increased from 15°–90°. The implications for sediment entrainment are reviewed briefly.

A deterministic finite element model is developed for streamflow routing. The simultaneous solution of the continuity and momentum equations results in a complete flow model that uses the Galerkin finite element technique and a dimensionless time‐weighting factor. The solution for the depth of flow, velocity of flow, and discharge obtained through the Newton‐Raphson iterative equation solver is found to be unconditionally stable for the time‐weighting factor between 0.55 and 1.0. The model shows great promise for the simulation of flow in natural channels as illustrated in a test problem of the Illinois River in Oklahoma. Simulated flows at 50.4 miles (80.6 km) downstream from the Watts station compared favorably with the observed flows at the Tahlequah station. Results of application of the model are in close agreement with those of an explicit finite difference model tested. Application to an idealized rectangular channel not only defines the stable range of the time‐weighting factor but also its influence in minimizing attenuation that results from the use of large time steps.

A method is presented which enables the computation of the suspended load as the depth‐integration of the product of the local concentration and flow velocity. The method is based on the computation of the reference concentration from the bed‐load transport. Measured concentration profiles have been used for calibration. New relationships are proposed to represent the sizegradation of the bed material and the damping of the turbulence by the sediment particles. A verification analysis using about 800 data shows that about 76% of the predicted values are within 0.5 and 2 times the measured values.

This paper presents simple analytical solutions to predict the temporal and spatial distribution of concentrations in one, two, or three‐dimensional fully saturated uniform flow fields caused by instantaneous or finite‐time release of contaminants from a point, line, plane, or parallelepiped source in an isotropic porous medium. The effects of the nonflux boundaries provided by bedrock, water table, or ground‐water divides are accounted for by the method of images. The same technique is used to model the contribution of a constant concentration boundary such as the one provided by a fully penetrating perennial stream. Sensitivity analyses show the effect of the number of images considered and the time‐step size used to evaluate the Duhamel integral. An example is presented to demonstrate the use of the method to analyze the distribution of concentrations due to seepage from an ashpond located in a streamaquifer system.

Flow from a stream into an adjacent aquifer is considered as a result of abrupt lowering or raising of the stream level. The governing nonlinear partial differential equation, known as the Boussinesq equation, is reduced to an ordinary one through a Boltzmann transformation. An analytical approximation is obtained by splitting the equation into two parts, of which the nonlinear one is treated as a source. The source term in each iteration step is evaluated in terms of the solution obtained at the previous step. A finite element solution is employed, combined with a Newton‐Raphson procedure. From comparisons, it is shown that the analytical approach is valid for an increase of the stream depth in the range from 1–10 times its initial value. The analytical results are presented in the form of a simple nomogram. A simple algebraic formula is provided for stored volume calculations.

This paper consists of a running commentary on outstanding contributions to hydraulics by writers from the past two centuries, beginning with a letter from Benjamin Franklin on canal‐boat resistance and ending with several articles on experimental projects that are still current. It is noted that the role played by American writers was the gradual refinement of principles previously formulated in Europe, culminating in the application of fluid mechanics to hydraulics.