The influences of seepage flow on particle dislodgement over a coarse granular bed

Abstract

Sediment transport plays an important role in the evolution of mountain landscapes and draws great attention to the quantification of its production, rate of transport, and interactions with anthropogenic activities. The initiation of sediment transport—or more specifically, the dislodgement of sediment particles from their resting places—depends on the magnitude and duration of local hydrodynamic forces disrupting the balance of stationary particles. In engineering practice, these driving forces are usually characterized by a surrogate measure: streamflow velocity, local realization or bulk-flow-averaged value, based on the physical scale under consideration. The implication of this approach is that the forces applied upon sediment are solely controlled by the oncoming streamflow and their magnitude can be determined uniquely by flow velocity. The direct link between velocity and force appears to be a general consensus, but the most recent research shows otherwise. In fact, noticeable discrepancies have been found between instantaneous force fluctuations and simultaneous velocity fluctuations regarding their magnitude and also the timing of their local peaks. Such discrepancies can be easily explained by the uncertain nature of a turbulent flow and underplayed by the use of a mean drag or a mean lift force coefficient to connect the main trends of the two properties. However, simplification of this kind may obscure underlying mechanisms that influence particle dislodgement in a more subtle way. Seepage effects, for instance, may change local hydrodynamic features around mobile particles and yet have received less attention, partly because seepage is difficult to identify and quantify at the instant of dislodgement. This aspect would become more critical at near-threshold conditions where local turbulent structures causing intermittent particle movements could interact with seepage at various degrees and deliver much different results. Our lack of understanding on such phenomena can limit the access to better interpretation and more accurate prediction of particle dislodgement.This study was intended to fill that gap by experimentally exploring seepage effects at the moments of particle dislodgement over a coarse granular bed. Note that seepage identified in our experiments occurred spontaneously in the vicinity of a single target particle, not introduced into the granular bed by any additional facility. Dislodgement of this target particle was limited to a low range of transport rates, slightly varied with the degrees of particle exposure considered. Observations and measurements were conducted through a particle tracking velocimetry (PTV) system at a sampling rate of 500 fps, sufficiently high to fully resolve dislodgement and ambient fluid motion simultaneously. The obtained velocity data was used to analyze a set of hydrodynamic properties, including turbulence intensity, turbulence kinetic energy transport, Reynolds shear stress, velocity and pressure quadrant distribution, and most important, the relative importance of seepage on particle dislodgement. The results served as evidence supporting that, though inconspicuous in most cases, seepage can play a significant role in particle dislodgement at near-threshold conditions by changing local turbulence features.