Journal of Hydraulic Engineering

Near‐prototype flume studies were conducted in which riprap‐protected embankments were subjected to overtopping flows. Embankment slopes of 1, 2, 8, 10, and 20% were covered with riprap layers with median stone sizes of 1, 2, 4, 5, and/or 6 in. Each riprap layer was tested by slowly increasing the discharge to failure. Riprap design criteria for overtopping flows were developed for estimating incipient stone movement and riprap layer failure as a function of the unit discharge, stone shape, median stone size, and embankment slope. Incipient stone movement occurred at approximately 74% of the riprap layer failure unit discharge. It was determined that rounded shape stone should be oversized approximately 40% to provide comparable protection of an angular shape stone. Flow channelization was observed to occur at approximately 88% of the unit discharge at failure. A flow concentration factor of approximately 1 to 3 was introduced for sizing stone.

Comparisons are made of the overall accuracy as well as the accuracy within different ranges of sediment concentration, Froude number, and slope for seven bed‐material load formulas. Four formulas that can compute bed‐material load transport by size fraction are used to determine the particle size distribution of the bed materials in transportation. One‐thousand, one‐hundred‐nineteen sets of laboratory data and 319 sets of river data in the sand size range are used to evaluate and compare the accuracy of these formulas. The overall accuracy of formulas in descending order are those of Yang, Engelund and Hansen, Ackers and White $\left(d_{50}\right)\text{,}$ Laursen, Ackers and White ${(}_{\mathrm{d35}})\text{,}$ Colby, Einstein, and Toffaleti formulas when applied to laboratory flumes. The accuracy in descending order when applied to natural rivers are the formulas by Yang, Toffaleti, Einstein, Ackers and White $\left(d_{50}\right)\text{,}$ Colby, Laursen, Engelund and Hansen, and Ackers and White $\left(d_{35}\right)\text{.}$ However, these ratings may vary depending on the values of sediment concentration, Froude number, and slope of the data used in the comparison. The study also indicates that Yang's formula by size fraction can accurately predict the size distribution of bed material in transportation, while Einstein's hiding and lifting correction factors overcorrected the effect of nonuniform size distribution of bed material on total bed‐material transport.

Erosion of streambanks and shorelines by exfiltrating seepage is called sapping or piping (because tubular cavities are produced). This mechanism is widespread in occurrence and is very significant to bank and shore stability, but is rarely recognized. The mechanism is complex and acts in concert with other processes of bank and shore erosion and deposition. Operation of those other mechanisms often masks the processes and products of the piping/sapping mechanism. Furthermore, failures caused by this mechanism may occur during periods of stream inactivity long after storm and/or flood events have ended. This paper describes and illustrates characteristic features of piping/sapping erosion and attendant instability. Situations typical of sites where piping/sapping is significant are illustrated. Conditions necessary (but not necessarily sufficient) to cause piping/sapping are presented. Diagnostic procedures are presented in a companion paper. The purpose of this paper is to present the fundamental aspects of this erosion mechanism and to identify those combinations of circumstances under which piping/sapping is likely to be an important process affecting bank or shoreline stability.

Although erosion of streambanks and other shorelines by emergent seepage is widespread, this erosion mechanism (termed piping or sapping) has not been recognized as important to the overall erosion process. The mechanism is complex, and interactions with other bank and shore processes tend to mask the effects of piping/sapping. Direct evidence (water emerging from a soil face and carrying away soil particles) is rarely encountered. Several types of indirect evidence are presented and illustrated in this paper, including cavities formed by piping, deposits of dislocated particles below piping zones, blind gullies, staining produced by persistent seepage outflow, and particular types of localized failures (slab toppling, block shearing, and tensile falls caused by undercutting due to piping/sapping). The interactions of this erosion mechanism with other erosion‐deposition processes are described. The purpose of this paper is to facilitate identification and evaluation of piping/sapping erosion, particularly for relatively inexperienced field investigators.

A procedure to calculate the flow and pressure distributions within navigation‐lock manifolds is presented. The approach is to treat each port as a discrete element in the flow system. Values of port hydraulic coefficients are established using previously reported laboratory studies of the flow characteristics of single‐port manifolds. Energy‐loss equations are written for the flow path through each port in the manifold. These equations use the port coefficients determined from laboratory experiments and the Darcy‐Weisbach friction factor. The energy‐loss equations are used in conjunction with the continuity equation to form a system of nonlinear equations in which the unknowns are the manifold discharges. These equations are solved simultaneously using a predictor‐corrector numerical scheme to provide a solution to multiported manifolds. The solution technique is verified with model data for manifold designs in existing locks. The solution technique determines the culvert and port discharges and the pressure distributions for these manifolds within acceptable limits.

The hydrodynamics of a dropshaft‐drift tube system under two transient conditions, namely, rapid increase of dropshaft inflow and strong main tunnel surge, is studied analytically and numerically. The governing equations are derived, and then linearized to explicitly represent the system's responses to external disturbances. The linearized study shows that the water column in the dropshaft oscillates sinusoidally in response to a disturbance. The amplitude of the oscillation was found to increase with the strength of disturbance as well as the fundamental period of the system. For inflow‐caused disturbances, a typical dynamic instability phenomenon exhibited by growing amplitude of oscillations may exist under certain conditions. For both types of disturbances, the amplitude can be so large that water in the dropshaft may overshoot the ground surface, causing damages. The analytical results are verified by direct numerical solutions of the original nonlinear equations and field records.

The dynamic load and cross‐flow response of trashrack bars and single beams subject to an incident flow was investigated for a wide range of cross‐sectional shapes, bar spacings, and flow incidence angles. At zero flow incidence, rectangular profiles with a depth‐to‐thickness ratio less than six were found to be susceptible to two mechanisms of vortex‐induced vibration: one due to Karman vortices, the other due to impinging leading‐edge vortices. Longer rectangular profiles at zero flow incidence were stable. Rounding the edges, especially those facing the flow, resulted in an increased intensity of excitation due to Karman vortices. At small flow‐incidence angles up to 15°, vortex excitation became weaker and turbulent buffeting set in. Reducing the bar spacing had a similar effect, attributable to the interference of adjacent bar wakes. Design recommendations are given to avoid serious flow‐induced vibrations.