Saltation trajectories of uniform bed material in a simple open-channel flow

Lift and drag forces were measured on a sphere embedded among similar spheres hexagonally packed on the bed of a flume, with steady, uniform flow conditions. Tests were conducted with the instrumented sphere flush in the bed and in four other positions, slightly above or below the level of neighboring spheres, in order to determine the effects of sheltering and protrusion. Basic statistics and probability distribution of lift and drag forces are investigated for all cases. Lift and drag force are determined to be normally distributed for both sheltered- and protruding-particle cases. Lift-force data are further analyzed to investigate the nature of flow-induced lift-impulse events, defined as the excursions of lift force above the submerged weight of the sphere, assuming a specific gravity of 2.65. Lift impulse is determined to have a probability distribution which can be represented by the 2-parameter Gamma distribution.; Saltation trajectories of a lightweight sphere were recorded with a high-speed motion-picture camera. The sphere was released from a static position in the bed, and its motion resulted solely from the effects of fluid forces acting to entrain it into the flow.; A numerical model is developed to simulate the observed saltation trajectories, based upon known sphere and fluid properties, velocity distribution and initial sphere velocity. The model is employed, using probabilistic input for the initial velocity of the sphere, to compute ensembles of trajectories. Ensemble-average trajectory height and length predictions compare favorably to experimental results and stochastic model results reported in the literature. Height and length predictions are generally significantly less than results reported in the literature for a hybrid (probabilistic/deterministic) model. Trajectory height and length and sphere mean speed are analyzed as to ensemble-average statistics and probability distribution. Both height and length of trajectory are determined to have probability distributions which can be represented by the 2-parameter Gamma distribution.