Journal of Hydraulic Engineering

An experimental study of the central portion of a unidirectional multiport diffuser discharging in the direction of a uniform ambient current in deep nonturbulent water was conducted by means of a towed model in a laboratory tank. The trajectories of the two-dimensional plane plumes that result after merging of the individual jets can be divided into weakly deflected plumes ( F_a < 0.60), with upstream spreading upon surface interaction, and strongly deflected plumes ( F_a > 1.0), with an intermediate range that contains an upstream blocked wedge. The dominant parameters F_a is an ambient/discharge Froude number, F_a =u _a/${\mathit{j}}_{\mathit{o}}^{1/3}$, where u _a is the ambient current and j _o is the two-dimensional buoyancy flux per unit diffuser length. In each plume regime the plume trajectory is a straight line, but with a distinctly different functional dependence of the trajectory slope on F_a . For the strongly deflected case, the plume behavior, its rate of rise, its tendency to attach to the bottom, and the distance for merging of the individual jets, also depend on the detailed three-dimensional diffuser geometric characteristics, including port spacing and port height. A “leakiness parameter” has been defined that characterizes these three-dimensional influences, in particular the ability of the ambient flow to pass through the merging diffuser flow field and allow the diffuser plume to stay aloft downstream. This behavior is different from earlier slot diffuser experiments by Cederwall (1971) and Roberts (1979) with zero leakiness in which a fully mixed, bottom-attached plume was observed.

Experiments have been carried out to determine the streamwise variations of the mean flow properties of the hydraulic jump formed in sloping channels. Measurements of velocities have been made with hot-film anemometer for various Froude numbers of the flow. Strip integral method, with an assumption that the jump is a wall jet, is applied to determine the mean velocity field and the variation of depth along the jump. An empirical correction is made for the deviation of static pressures in the jump from hydrostatic pressures. Satisfactory agreement is found between the predicted and experimental results of the mean flow field. A body of data pertaining to the roller length collected by various investigators has been presented, and an explanation is provided for their scatter.

The ability to predict the stability of eroding riverbanks is a prerequisite for modeling alluvial channel width adjustments and a requirement for predicting bank-erosion rates and sediment yield associated with bank erosion. However, there are a number of limitations of existing bank-stability analyses that limit their physical basis and predictive ability. Some of these limitations are addressed through the development of a new bank-stability analysis. The new approach is applicable to steep, cohesive, nonlayered riverbanks that fail along planar failure surfaces. Pore-water and hydrostatic confining pressure terms are included in the analysis. The failure plane is not constrained to pass through the toe of the bank. The predictive abilities of four bank-stability analyses (Lohnes and Handy 1968; Huang 1983; Osman and Thorne 1988; and the present analysis) were assessed using field data. The new analysis is the most successful of the tested analyses in terms of predicting the stability of riverbanks with respect to mass failure.

Three tests on non-Newtonian fluids were used to verify the use of a vertically rotating flume (VRF) for the study of the rheological properties of debris flow. The VRF is described and a procedure for the analysis of results of tests made with the VRF is presented. The major advantages of the VRF are a flow field consistent with that found in nature, a large particle-diameter threshold, inexpensive operation, and verification using several different materials; the major limitations are a lack of temperature control and a certain error incurred from the use of the Bingham plastic model to describe a more complex phenomenon. Because the VRF has been verified with non-Newtonian fluids as well as Newtonian fluids, it can be used to measure the rheological properties of coarse-grained debris-flow materials.

Minimum bed elevation in stabilized bendways is needed in the design of bank protection. Two empirical methods have been identified for scour-depth estimation that are consistent with available resources on a typical bank-protection project. The 1990 empirical method by Watanabe, Hasegawa, and Houjyou provided a good fit of the observed data used herein. Maximum depth in the bend is a function of R _c /W, aspect ratio, channel slope, mean depth, and Darcy friction factor. An empirical scour-depth estimation method is developed herein and presented with a safety factor. Maximum depth in the bend is a function of R _c /W, aspect ratio, and mean depth. The proposed scour-depth method should be limited to R _c /W from 1.5 to 10 and W /D _mnc from 20 to 125. Application to overbank flow conditions should be limited to overbank depth less than 20% of main-channel depth.

The incipient jump condition is defined as the critical condition for the formation of a surface roller immediately upstream of a sill. The incipient jump conditions can be classified in two categories. In one case, the incipient jump occurs by increasing the sill height from the state of the splashing flow. In the other case, raising the tailwater level from the splashing flow condition yields the incipient jump. The mechanisms for both cases of the incipient jump conditions are explained. Also, the effect of the channel width on the incipient jump condition is discussed. The application of the momentum equation yields a prediction for each incipient jump condition. For wide channels, in which two-dimensional breaking occurs, the agreement between the predicted and experimental results is satisfactory. The formation of a forced jump is required for the design of a stilling basin.