Journal of Hydraulic Engineering

The concept of potential velocity is introduced in this paper to explain the principle of sediment effective power. The writer suggests a method to formulate this principle. On this basis, a new classification of moving sediment in an open channel is suggested and a description of the applications to river sedimentation is presented. The variation of von Karman constant in sediment‐laden flow, the distinction between wash load and bed‐material load, as well as the adequacy of sediment transportation formulas are reviewed.

Laboratory experiments were performed in a 40‐ft long, 2‐ft wide, 1‐ft deep $(12‐m\times 0.6‐m\times 0.3‐m)$ tilting flume to elucidate the effects of temperature variations on the flow and sediment‐transport characteristics of streams with beds of fine sand ($D_{50}=0.11\mathrm{mm}$ in these experiments). Nearly constant depth was maintained with three different constant discharges, and temperatures were varied from approximately 30° C to near 0° C. In the two series of experiments at higher velocities and Froude numbers, temperature reductions produced major increases in sediment discharge and uniformity of concentration distributions and smaller, but still significant, increases in friction factor and bed‐form height. An analysis of the experimental results within the framework of a theoretical sediment‐transport model revealed that the increased sediment discharge with decreasing water temperature in flows carrying significant amounts of suspended sediment are caused by increases in both bed‐layer sediment concentration and the uniformity of the concentration profiles. The sediment discharges of low‐velocity $(F=\simeq 0.30)$ flows, which were just above the threshold of sediment motion, were not significantly affected by water temperature. It was concluded that temperature affects the near‐bed concentration only if the bed‐layer Reynolds number (based on bed‐layer thickness) is greater than about 20.

Research was carried out to develop relatively simple low cost procedures for estimating chronological monthly runoff for the historical period before a gaging station was installed on the river but during which precipitation and temperature records are available. The method also supplies estimates of runoff during periods of missing records. The procedures are based on the equation for the water budget of a drainage basin which shows runoff to be a function of precipitation, water losses and the change in storage. It is shown that current and antecedent precipitation and temperature can be used as independent variables to account for the water loss and storage terms. The methods were tested by calibrating the models for the more recent years of records and then hindcasting runoff for earlier periods during which measured runoff was available for comparison.

Major river regulation works have often resulted in alternate bar formation. Experimental work is described which provides more information on this phenomenon. Alternate bar formation is closely associated with meandering, inferring that alternate bar formation is caused by flow induced periodic disturbances of the horizontal velocity profile. Based on these considerations a criterion has been developed which defines the upper limit of alternate bar formation, and this has been satisfactorily compared to other formation criteria and experimental data. The lower limit is given by the initiation of bedload transport. A simple relation between slope, channel width and grain size is then deduced to indicate whether for a given channel, alternate bars can develop for any discharge. This relation takes into account armoring effects. Another equation is proposed to predict the scour depth resulting from alternate bar formation. Such scour often induces costly river maintainance works. Finally, some interesting results on flow resistance have been found, leading to a tentative equation for the intermediate roughness range.

A computer‐based flood‐ and sediment‐routing model which simulates river channel changes is described together with its application in the case study of a disturbed river. Simulated results of this study are supported by field observations and measurements. This model incorporates the interrelated changes in channel bed profile, width, and lateral migration in channel bends. These changes reflect, in part, a river's adjustments in power expenditure. The interrelation of changes in channel bed profile and width is illustrated by a physical example and explained by the river's tendency to establish equal power expenditure along the channel. Such power transformation associated with river channel evolution tends to restore the dynamic equilibrium in sediment transport, that is, equal sediment load along the channel. The case study demonstrates that in the case of severely disturbed rivers, flood‐level computation using a fixed‐bed model may be quite inaccurate and improved accuracy can be provided by an erodible‐bed model.

An investigation of cellular secondary currents, i.e., three‐dimensional flow patterns, is very important in hydraulic engineering because these currents might cause three‐dimensional sediment distributions and bed configurations such as sand ribbons in straight rivers. It is, however, fairly difficult to measure secondary currents of water flows, even by making use of hot‐film anemometers, because the velocity of secondary currents is within 5% of the mainstream velocity. Thus, on the basis of the hypothesis that the existence of free surface may not be an essential cause of secondary currents, the present study has investigated experimentally the turbulent structure of their currents in air conduit. Considering an essential interaction between secondary currents and bedform, longitudinal ridge elements were attached onto both lower and upper bottoms of the conduit, which simulated the longitudinal ridges and trough of river bedform, i.e., sand ribbons. All three components of the velocity were measured accurately by hot‐wire anemometers. The structure of secondary currents was examined through the equations of mean flow vorticity and mean flow energy.