Journal of Hydraulic Engineering

Data is presented detailing the water motion in the Wellington Reservoir induced by the outflow from the valve at the base of the dam wall. This outflow system establishes a simple horizontal shear flow longitudinally along the axis of the reservoir, the characteristics of which are determined by the three‐dimensional sink flow dynamics at the valve. The associated gradient Richardson number of this flow is 11.2; the vertical mass diffusion coefficient, estimated from temperature measurements, is $9.0\times {10}^{-6}{m}^2/s;$ and the energy dissipation, obtained from microstructure profiles, is around $3.2\times {10}^{-9}{m}^2/s^3\text{.}$ Together these measurements imply a flux Richardson number of 0.28. The large value of the gradient Richardson number computed from the sink flow velocity profile suggests that this mean velocity shear was augmented by a field of internal waves, which locally lowered the gradient Richardson number below the critical values.

Negatively buoyant (plunging) flow entering a horizontal and diverging channel has been studied in the laboratory. Such flow is found at the heads of reservoirs and near effluent sites into lakes. Six flow regimes could be distinguished in the experiments. The flows are similar to nonbuoyant, twodimensional diffuser flow, but negative buoyancy adds several variations. The dependence of the flow regimes on the inflow channel aspect ratio, inflow densimetric Froude number, diffuser half‐angle, and channel‐to‐diffuser transition are presented. The experimental results can be used to predict if and where flow separation will occur and what flow regimes will be present.

The internal flow behavior of negatively buoyant flow in a diverging horizontal channel is inferred from visual observations and temperature distribution. Some observed flow features are explained and illustrated.

A method is presented for calculating characteristics in the upstream direction for kinematic wave routing of overland and upland channel flow resulting from time‐varying rainfall. This method allows convenient selection of a regular time step for the discharge hydrograph without sacrificing the precision of an analytical solution. The technique is useful for verification of numerical methods, or is appropriate for use in a watershed model where histogram lateral inflows are used.

The problem of discharge assessment in compound channels is described, including the difficulty of accurately modeling main channel and floodplain flows. Theoretical considerations show that the ratios of main channel velocity and discharge to floodplain values in a compound channel are independent of bed slope and are influneced only by depth and geometry. This is also true for other significant velocity and discharge ratios. Experimental confirmation of these theoretical predictions is presented using data from three symmetrical compound channel shapes. The relationships are described mathematically where possible, and the exponents and coefficients are evaluated. Comparisons are presented of measured discharges with values obtained using conventional methods that neglect the interaction between main channel and floodplains, showing the errors incurred by the conventional approach. The implications of these findings for compound channel analysis and design are assessed.

Seawater intrusion barriers are currently being constructed in the coastal aquifers of Japan based on the results of a computer model employing finite difference numerical techniques. This model provides a practical design and management tool for predicting the likelihood of seawater intrusion through a semipervious subsurface barrier in an unconfined coastal aquifer. The analysis assumes an abrupt interface, the Dupuit approximations, and various simplifying boundary conditions. The movements of both the water table and the freshwater‐saltwater interface are numerically simulated in the unsteady state. The computed results are validated with observations obtained from a vertical two‐dimensional sand box model. The application of the computer model is illustrated by a site‐specific analysis for a proposed subsurface barrier location in a limestone coastal aquifer on Okinawa‐jima Island in the Western Pacific Ocean. From this analysis, it is found that the subsurface barrier is able to delay seawater intrusion for about 2 mo under critical conditions of continual pumping of the aquifer without recharge. The subsurface barrier is technically feasible and is a viable solution to the problem of seawater intrusion in coastal aquifers.

Despite major advances in the science of pipeline transient analysis, design for unsteady flow remains largely an art. In this paper, a fast, simple, yet effective technique for valve‐motion synthesis is developed and applied. Specifically, this method, called quick stroking (QS), is presented in general and as a BASIC computer program for execution on widely available personal computers. The method is also applied to several single pipe example problems, including cases with high or low pipe friction, and upstream or downstream valve locations. These case studies show the QS procedure to be a fast and simple method of synthesizing time‐optimal valve motions. As a result, the quick‐stroking procedure should become a welcome weapon in the hydraulic engineer's fight against unwanted waterhammer.