Journal of Hydraulic Engineering

The method of characteristics is applied to numerical integration of the shallow water equations written in r‐t coordinates. Instant failure of a dam, between vertical walls which either converge or diverge, is then simulated. Comparisons are made with the Ritter and Stoker solutions for parallel walls and different initial depths downstream, including the associated positive surge. Physical model test results confirm general expectations that converging walls create deeper and faster flows downstream and vice versa. Composite wall arrangements are tested, and comparison also is made with the simulation of dam‐break flow in a contracting channel by Marshall and Menendez.

Velocity distribution equations for open channel flows are derived and compared. These equations are derived by a combined application of: (1) A probabilistic formulation of the velocity distribution problem; (2) the entropy concept in the selection of the probability distribution function of velocity; (3) a geometrical technique in modeling a curvilinear coordinate and the coordinate transformation between this and the Cartesian coordinates; and (4) the basic hydrodynamics concerning the rates of transport of mass, momentum, and kinetic energy by the flow through an open channel cross section. A technique to estimate the parameters of these velocity distribution equations is also developed. The equations are capable of modeling and simulating the velocity distribution from the channel bed to the water surface, which may have the maximum velocity occurring on or below the water surface.

Using both a theoretical and an experimental approach, the main flow features in junctions with thoroughly supercritical flows are considered. The junctions are characterized by rectangular channels of equal branch width and sharpedged corners. Junction angles of 22.5° and 45° are considered. Expressions are given for the direction of the main wave front, and the wave angles in both the upstream, and the lateral branches. The prediction is based on a conventional crosswave analysis, in which streamline curvature effects were not incorporated. Yet, reasonable agreement is found between predictions and observations regarding the maximum wave height, and its position in the junction. Furthermore, the velocity field, and the flow surface are discussed for typical flow configurations. Finally, the limits of thoroughly supercritical flow conditions are verified by the momentum equation, thereby including the flow contraction generated at the inflow to the downstream branch. Conclusions are given from which a design may easily be accomplished.

Results of experimental observations in rectangular side‐channel spillways are presented, with particular attention to the effects of a nonprismatic cross section and the bottom slope on the free surface profile. For subcritical flow, the difference of surface profiles along the channel axis and the side walls is outlined. Further, flow configurations with supercritical inflow conditions and the possible formation of hydraulic jumps along the lateral inflow length are discussed. The observations are compared with predictions based on an extended theoretical approach, thereby including the streamwise momentum contribution due to the lateral flow. The agreement between the computed and the observed axial free surface profile is fair. The proposed extended model equation should be considered whenever the bottom slope of the side channel is larger than a certain percentage.

A shaft spillway is equipped with a hood in order to operate as a siphon. Appropriate hood dimensions and shape are searched by using cylindrical hoods of different dimensions for a standard bellmouthed shaft. The hood, sized and streamlined according to experimental results, showed a good performance concerning the discharge and subatmospheric pressure conditions. Design charts for discharge coefficient as well as critical hood and shaft‐crest pressures are presented for such a siphon‐shaft spillway, and a design example is given.

A desk‐top method is presented to size detention ponds and the outlet facilities for multiple return periods. The method is based on the usual reservoir routing technique, which couples the hydrologic storage equation with those describing flow through the outlet structures. The equations are rearranged in terms of several dimensionless parameters. A computer model is employed to solve the governing equations using a finite difference scheme. The results obtained for carefully chosen combinations of the dimensionless parameters are generalized based on the concept of similarity. They are presented in chart form. These charts can be used directly, i.e., without a lengthy trial‐and‐error routing procedure, to size a detention pond that satisfies a set of design criteria requiring the control of runoff for at least three return periods. The use of the design charts are illustrated through a sample application.

The hydraulic jump in U-shaped channels is investigated. The conventional momentum approach is found to slightly overestimate the sequent depth ratio. The jump length, the characteristics of the free surface profiles, and the internal flow features are experimentally determined. Particularly interesting is the formation of a bottom separation zone for flows with small relative inflow depth. The spatial flow features and the extension of the results to hydraulic jumps in circular channels flowing partially full are discussed.