Journal of Hydraulic Engineering

Experiments were conducted on alluvial beds of four sediment mixtures having different arithmetic mean diameters and standard deviations. The bed load transport rates of individual fractions were measured. Einstein's method, Proffitt and Sutherland's method, and Misri's method of calculation of the bed load transport of individual fractions for sediment mixtures were checked with carefully collected laboratory data. By analyzing the data on lines similar to those of Misri, et al., their method of calculation of the bed load transport rate for individual fraction is modified to make it applicable over a wide range of parameters. The relationship between the dimensionless grain shear stress and the dimensionless bed load transport rate developed by Misri, et al., for uniform sediments was made applicable to sediment mixtures by introducing a multiplying correction factor. The parameters affecting this correction factor were identified and a relationship obtained for the same by analysis of an extreme amount of laboratory and river data. Use of this factor along with the transport law for uniform sediments enables determination of the rate of bed load transport of the individual fractions of a mixture.

Careful experiments were conducted on alluvial beds of four sediment mixtures having different arithmetic mean diameters and standard deviations. The suspended load transport rates of individual fractions were measured and compared with both Einstein's and Holtroff's methods of calculation of suspended load for individual fractions. These methods were found to be unsatisfactory in the present range of sediment parameters. A relationship is found to exist between the dimensionless shear stress and the suspended transport rate for uniform sediment. To make this relation applicable to nonuniform sediments, a corrective multiplying factor for shear stress is introduced. The dimensionless parameters which govern this correction factor are identified and a relation for the same obtained. In this manner the relation can be applied to individual size portions of a sediment mixture.

An unsteady flow model for ice‐covered rivers is developed and applied to the upper St. Lawrence River. The model is capable of providing continuous simulation of flow and ice conditions in an ice‐covered river that consists of a network of channels. The flow conditions are determined by numerically solving the one‐dimensional equations of motion for unsteady river flows with floating ice covers. Effects of ice cover thickness and hydraulic resistance of the ice cover are considered. Computed results compare favorably with recorded data.

A parameter estimation method has been developed for a system of three‐dimensional mathematical model of flow in open channels, which does not require (primary flow) velocity data. It has broadened the applicability and effectiveness of the model in scientific investigations into the complex interaction among the primary and secondary flows, shear stress distribution, channel characteristics (roughness, slope, and geometry), and other related variables governing all transport processes in open channels. The method was applied to a study of 3‐D structure of flow in rectangular open channels. The interaction among the primary and secondary flows and the shear stress distribution was investigated under various values of Manning's n, width‐to‐depth ratio and slope of the channels. The result has answered quantitatively many questions which arise in open channel hydraulics.

A model is presented for the calculation of vertical diffusion in stratified flows dominated by bottom shear. The model includes a very simplified representation of the potential energy (rotation against buoyancy forces) and kinetic energy of turbulent eddies. The two universal model constants are derived from measurements of vertical mixing in open‐channel flow. Analytical expressions are given for the vertical diffusion coefficient ${\epsilon }_2\left(z\right)$ in the special cases of two‐layer flows and linearly stratified flows. Numerical calculations are necessary to obtain ${\epsilon }_z\left(z\right)$ for arbitrarily stratified flows. Application of the method using a numerical one‐dimensional diffusion model shows good agreement between the calculated density profiles and profiles measured in laboratory and field situations. A practical application is the prediction of the stratification evolution near a cooling water or wastewater disposal point in a receiving flow. The model needs the input of the vertically averaged flow velocity and an initial density profile.

Urban storm drain design in Montana and throughout the 11 western states is frequently based on precipitation values determined from NOAA Atlas 2. A statistical analysis of 44 yrs of data for Billings, Montana indicates significant differences from NOAA Atlas 2 for durations between 5 mins and 30 mins. The values in NOAA Atlas 2 understate the magnitude of some return period values in Billings by as much as 69%. Peak runoff rates based on precipitation data at Billings are as much as 71% greater than rates based on precipitation from NOAA Atlas 2.